(19)
(11)EP 2 305 171 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
29.12.2021 Bulletin 2021/52

(21)Application number: 10184018.9

(22)Date of filing:  29.03.2004
(51)International Patent Classification (IPC): 
A61D 19/00(2006.01)
G01N 15/14(2006.01)
C12N 5/00(2006.01)
(52)Cooperative Patent Classification (CPC):
Y10T 436/255; C12N 5/0612; G01N 15/1459; G01N 15/147; G01N 15/1404; G01N 2015/1415; G01N 2015/1006; Y10T 436/2575; Y10T 436/10; Y10T 436/115831; Y10T 436/2525; G01N 2015/1406; G01N 2015/149; Y10T 436/25; Y10T 436/12; Y10T 436/25375

(54)

Apparatus and methods for providing sex-sorted animal sperm

Vorrichtung und Verfahren zur Bereitstellung von geschlechtssortierten Tierspermien

Appareil et procédés pour fournir du sperme d'animaux triés selon le sexe


(84)Designated Contracting States:
AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PL PT RO SE SI SK TR

(30)Priority: 28.03.2003 US 458607 P
28.03.2003 US 458731 P

(43)Date of publication of application:
06.04.2011 Bulletin 2011/14

(62)Application number of the earlier application in accordance with Art. 76 EPC:
09014407.2 / 2194379
04749513.0 / 1608963

(73)Proprietor: Inguran, LLC
Navasota, TX 77868 (US)

(72)Inventors:
  • Durack, Gary
    Mahomet, IL 61853 (US)
  • Hatcher, Jeremy, T.
    Urbana, IL 61801 (US)
  • Westfall, Lon A.
    Mahomet, IL 61853 (US)
  • Helbing, David R.
    Mahomet, IL 61853 (US)
  • Wallace, Jeffrey D.
    Rantoul, IL 61866 (US)
  • Vandre, Gary P.
    Mahomet, IL 61853 (US)
  • Didion, Bradley
    Mt. Horeb, WI 53572 (US)
  • Nayak, Niraj V.
    Redondon Beach, CA 90277 (US)
  • Anzar, Muhammad
    Chesterfield, MO 63017 (US)
  • Ludwig, Cindy L.
    St. Louis, MO 63146 (US)
  • Graham, Jeffrey A.
    Chesterfield, MO 63017 (US)
  • Crowley, Kathleen S.
    Webster Groves, MO 63119 (US)

(74)Representative: Jacob, Reuben Ellis et al
Maucher Jenkins 26 Caxton Street
London SW1H 0RJ
London SW1H 0RJ (GB)


(56)References cited: : 
WO-A1-99/58955
  
  • SEIDEL G E JR ET AL: "Current status of sexing mammalian spermatozoa.", REPRODUCTION, vol. 124, no. 6, December 2002 (2002-12), pages 733-743, XP002626262,
  • GARNER D L: "Sex-sorting mammalian sperm: Concept to application in animals", JOURNAL OF ANDROLOGY, vol. 22, no. 4, July 2001 (2001-07), pages 519-526, XP002627778,
  • SHAPIRO H M ET AL: "Multistation multiparameter flow cytometry: some influences of instrumental factors on system performance.", CYTOMETRY, vol. 4, no. 1, July 1983 (1983-07), pages 11-19, XP002627529,
  • MANSBERG H P ET AL: "The Hemalog D white cell differential system.", THE JOURNAL OF HISTOCHEMISTRY AND CYTOCHEMISTRY, vol. 22, no. 7, July 1974 (1974-07), pages 711-724, XP002627530,
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description

Background of the Invention



[0001] This invention relates to method of handling non-human semen having cells therein. This sperm cell handling method relates to animal semen collection, and more particularly to apparatus and methods using various techniques, including flow cytometry, to yield sperm populations that are enriched with sperm cells having one or more desired characteristics, such as viable populations of sperm cells sorted according to DNA characteristics for use by the animal production industry to preselect the sex of animal offspring.

[0002] The fertilization of animals by artificial insemination (Al) and embryo transplant following in vitro fertilization is an established practice. In the livestock production industry, the ability to influence the reproductive outcome toward offspring having one or more desired characteristics has obvious advantages. By way of example, there would be an economic benefit in the dairy industry to preselect offspring in favor of the female sex to ensure the production of dairy cows. Efforts have been made toward achieving this goal by using flow cytometry to sort X and Y sperm cells, as evidenced by the disclosures in US Patents Nos. 6,357, 307 (Buchanan, et al.), 5,985, 216 (Rens, et al.), and 5,135, 759 (Johnson). However, none of these efforts has resulted in the introduction of a commercially successful high-throughput system capable of producing production volumes of relatively pure sexed sperm cells having a motility sufficient for effective fertilization.

[0003] Accordingly, there is a current need in the animal production industry for a viable high-speed system for efficiently isolating sperm cells based on a specified DNA characteristic (or other characteristics) to produce quantities of such cells, which can be used on a commercial scale. Also needed is a sperm handling system that preserves the viability of such isolated sperm as it is processed by the isolating system and that allows for preservation of such isolated sperm until such time that it is ready for use. The present invention addresses these needs.

[0004] Flow cytometry may broadly be defined as measuring characteristics of individual particles as they pass generally single file in a fluid stream through a measuring device which, typically, provides information for classifying the particles according to selected characteristics. Optionally, the particles may then be separated into populations using any number of techniques, including droplet sorting, droplet interference sorting, and fluid switching. Another option is to selectively destroy unwanted particles, for example by photo ablation.

[0005] In an optically-based flow cytometry system, optics are used to direct and focus a beam of light (e.g., visible light or UV light) on the stream containing the particles, and to collect emissions from the particles, including scattered light and/or fluorescence emissions from the particles. In one common optic system, for example, a beam of light (e.g., a laser beam) is focused on the stream and emissions are collected by a pair of collection units, one positioned forward of the laser for collecting scattered light emissions and another positioned orthogonally to both stream and the laser for collecting fluorescence emissions. Each collection unit includes a separate photodetector, which increases the cost of the system. Further, in traditional optic systems the photodetectors translate the collected emissions into electrical signals, which are analyzed using analog systems to classify the particles according to selected characteristics of the particles. Analog systems are relatively inexpensive, but only limited information can be derived from the signals.

[0006] Others have tried to develop technology that can be used to process sperm cells to obtain populations of sperm cells that are enriched with sperm that have a desired sex chromosome. However, the existing technology falls short of the inventive technologies described herein.

[0007] For example, Johnson et al. (U. S. Patent No. 5,135, 759) describe the separation of intact X and Y chromosome-bearing sperm populations according to DNA content using a flow cytometer/cell sorter into X and Y chromosome-bearing sperm enriched populations. As described, the sperm is combined with a DNA selective dye at a temperature of 30 to 39°C for a period of 1 hour (39°C) to 1.5 hours (30°C). A flow cytometer is then used to measure the amount of fluorescent light emitted as the sperm passes through a laser beam that excites the dye. Because the X chromosome-bearing sperm contains more DNA than the Y chromosome- bearing sperm, with most species of mammal having about 3 to 5% difference, the X chromosome-bearing sperm emits more fluorescent light than the Y chromosome-bearing sperm. In order to account for the fact that the fluorescence measurement may vary depending on the rotational orientation of the sperm cells, two photo detectors are used. The first determines whether the sperm cells are properly oriented, while the second takes a measurement that is used to classify the sperm as having an X or Y chromosome. An oscillator is used to cause the stream containing the sperm to break into droplets downstream of the place where the sperm pass through the laser beam. Droplets containing single sperm of a predetermined fluorescent intensity are given a charge and electrostatically deflected into collection vessels. The collected, gender enriched sperm population, is then used for microinjection, in vitro fertilization, or artificial insemination.

[0008] Seidel et al. (WO 02/43574) also describe separation of sperm into gender enriched populations of X and Y chromosome-bearing cells using flow cytometry. Seidel et al. describe staining the cells at a temperature between 30°C and 40°C.

[0009] United States Patent Application Pub. No. 2003/0157475 A1 (Schenk, August 21, 2003) describes a method of cryopreserving sperm cells that have been sorted according to X or Y chromosome content. As noted therein, it is desirable to add a cryoprotectant to sperm cells before they are cryopreserved to protect the sperm cells during the cryopreservation process. For example, glycerol is one cryoprotectant that is commonly added to bovine sperm cells prior to cryopreservation. However, in order to obtain better protection from the cryoprotectant, it is desirable to wait for the cryoprotectant to equilibrate with the sperm cells before subjecting the sperm cells to temperatures below 0°C. During the equilibration period, the cryoprotectant penetrates the cell membrane to provide intra-cellular protection in addition to any extra-cellular protection provided by the cryoprotectant. Thus, the cryopreservation methods described in United States Patent Application Pub. No. 2003/0157475 A1 specify that an extender containing glycerol is added to the sperm cells after they have been cooled to about 5°C. Then the sperm cells and glycerol are allowed to equilibrate at 5°C for anywhere between 1 and 18 hours before the sperm cells are subjected to lower temperatures. The disclosure recommends an equilibration period of between three and six hours in order to obtain the best results.

[0010] Unfortunately, the time and expense involved in a 3 to 6 hour equilibration period will have a negative impact on profitability of a commercial sperm sorting process. Furthermore, in the context of a commercial sperm sorting process, it is believed that the health of the sperm is generally improved by reducing the time between collection of the sperm and cryopreservation (other factors being equal). From this standpoint as well, it would be desirable to have access to cryopreservation technology that does not require a long equilibration period to obtain the optimal benefits of a cryoprotectant. Moreover, the known cryopreservation technology is reported to have a detrimental impact on sperm motility, which is indicative of decreased sperm fertility. Thus, there is a need for cryopreservation techniques that preserves sperm health compared to conventional techniques.

[0011] Seidel G E Jr et al, REPRODUCTION, 124(6), 2002, 733-743 describes a review of spermatozoa sexed by flow cytometry and cell sorting.

[0012] Garner D L, J ANDROLOGY, 22(4), 2001, 519-526 describes a review of spermatozoa sexed by flow cytometry and cell sorting.

[0013] Shapiro H M et al, CYTOMETRY, 4(1), 1983,11-19 describes a multistation multiparameter flow cytometry system for sperm cell sorting.

[0014] WO 99/58955 A1 substantially simultaneous and multiplexed, multi-analyte diagnostic system and/or method for performing assays using a flow analyzer.

[0015] Mansberg H P et al, J HISTOCHEM CYTOCHEM, 22(7), 1974, 711-724 describes a Hemalog D system performs automated differential white ceil counts utilizing principles of cytochemistry, electro-optical measurement and signal logic processing.

[0016] WO 01/37655 describes a method for cryopreserving sex-selected sperm cells.

Summary of the Invention



[0017] According to a first aspect of the invention, there is provided a method of handling non-human semen having cells therein, the method including providing a supply of semen obtained from a non-human male animal, using a programmable machine to conduct a plurality of integrated flow cytometry operations, the operations including: a) receiving the supply of semen into the machine, b) forming multiple streams containing the cells, and c) sorting at least one portion of the cells into a first population of cells having a characteristic A and a second population of cells having a characteristic B, and distributing the first or the second population for commercial use, wherein the step of forming multiple streams containing the cells is enabled by having multiple flow cytometry units linked to form a single integrated multi-channel flow cytometry unit, and wherein the single integrated multi-channel flow cytometry unit utilizes a common supply of cells for the multiple streams; and a common system for delivering carrier fluid, the common carrier fluid delivery system being capable of supplying one or more of the units at a first flow rate and one or more of the units at a second flow rate; a common operations control input; and a common processor for receiving and processing information from the multiple flow cytometry units, wherein the operational parameters of each of the multiple flow cytometry units can be set independently of the other flow cytometry units.

Brief Description of the Drawings



[0018] 

Fig. 1 is a work flow diagram for an exemplary sperm sorting process;

Fig. 2 is a schematic diagram of a of a flow cytometry droplet sorting system;

Fig. 3 is a side view of a flow cytometry apparatus for droplet sorting showing an epi-illumination optic assembly focusing an excitation beam on an upward moving fluid stream generated by a nozzle system;

Fig. 4 is an end view of a nozzle and nozzle holder suitable for the inventive system;

Fig. 5 is a sectional view of the nozzle and nozzle holder of Fig. 4 taken through cutting plane 5--5 on Fig. 4;

Fig. 6 is a schematic diagram of a sperm cell entrained in a fluid stream being interrogated by an elliptically shaped beam spot;

Fig. 7 is a schematic diagram showing the angular envelope for the desired orientation of a sperm cell in which the light beam from the optics system will strike a wide face of the cell generally broadside;

Fig. 8 is a cross sectional view a nozzle body suitable for the inventive system;

Fig. 9 is a side view of the nozzle body shown in Fig. 8 showing a series of cutting planes (A-A through H-H and J-J through K-K) through the nozzle body;

Figs. 9A-9H and 9J-9K are sectional views of the nozzle body shown in Figs. 8 and 9 along the corresponding planes (A-A through H-H and J-J through K-K) of Fig. 9;

Fig. 10 is a perspective view of a cross section of a nozzle system having an orienting baffle in the nozzle;

Fig. 11 is a cross sectional view of the nozzle system shown in Fig. 10;

Fig. 12 is an enlarged partial cross sectional view of a portion of the nozzle system shown in Figs. 10 and 11;

Fig. 13 is an enlarged partial cross sectional view similar to the view shown in Fig. 12, but taken from a direction that is perpendicular to the viewing direction in Fig. 12;

Fig. 14 is a side view of a baffle holder holding a baffle plate;

Fig. 15 is a top view of the baffle holder and baffle plate shown in Fig. 14;

Fig. 16 is a top view of a baffle holder rotationally oriented in a nozzle so that the legs of the baffle plate intersect in a line that is parallel to the major axis of ellipse D in the nozzle;

Fig. 17 is a top view of a baffle holder rotationally oriented in a nozzle so that the legs of the baffle plate intersect in a line that is perpendicular to the major axis of the ellipse D in the nozzle;

Fig. 18 is a side cross sectional view of a nozzle system including a baffle showing a series of cutting planes (A-A through E-E) through the nozzle and baffle;

Figs. 18A-18E show the cross sectional flow areas at various points in the nozzle system shown in Fig. 18;

Fig. 19 is a cross sectional view similar to Fig. 12 taken through a nozzle having a baffle plate that is perpendicular to the longitudinal axis of the nozzle;

Fig. 20 is a cross sectional view of the nozzle shown in Fig. 19 taken through the cutting plane 20-20 shown on Fig. 19;

Fig. 21 is a cross sectional view similar to the cross sectional view of Fig. 18 showing a nozzle system having a sample introduction conduit at an offset location;

Fig. 22 is a perspective view of a nozzle system mounted on a nozzle mount;

Fig. 23 is schematic diagram of a plurality of aligned sperm cells being rotationally oriented as they pass through an orifice member toward the interrogation location;

Fig. 24 is a schematic diagram showing the droplet break-off location downstream from the nozzle;

Fig. 25 is a schematic diagram of a break-off sensor system;

Fig. 26 is a front elevation of one flow cytometry system;

Fig. 27 is an enlarged perspective view of a portion of the system shown in Fig. 26 with parts of the system removed for clarity;

Fig. 28 is a schematic diagram of an epi-illumination optics system;

Fig. 29 is a perspective view of an epi-illumination optics system;

Fig. 30 is a side view of the epi-illumination optics system shown in Fig. 29;

Fig. 31 is a top view of the epi-illumination optics system shown in Figs. 29 and 30;

Fig. 32 is a sectional view of the epi-illumination optics system along the plane 32-32 of Fig. 30;

Fig. 33 is a sectional view of a portion of the epi-illumination optics system along the plane 33-33 of Fig. 31;

Fig. 34 is a perspective view showing only elements of the optical filtering system that are rearward of the dichroic filter of the epi-illumination optics system shown in Fig. 29;

Fig. 35 is a perspective view of another epi-illumination optics system including translational adjustment of the cylindrical lens;

Fig. 36 is a schematic diagram of an interrogation location of showing a laser beam focused on a fluid stream downstream of the nozzle at a skewed angle of incidence;

Fig. 37 is a schematic diagram of a sort calibration system of;

Fig. 38 is a schematic diagram of an epi-illumination sensor for use with the sort calibration shown in Fig. 37;

Fig. 39 is a block diagram of a digital cell analyzer (DCA) and processor controller.

Fig. 40 is a schematic diagram of a multi-channel sorter showing two channels;

Fig. 41 is a work flow diagram of a multi-channel sorter showing four channels;

Fig. 42 is block diagram of an analog cell analyzer (ACA);

Fig. 43 is a graph illustrating a stream of waveform pulses from a photodetector output detecting fluorescent pulses from cells streaming at an average rate of 10,000 cells/second;

Fig. 44 is an exploded view of Fig. 43 illustrating the stream from a photodetector output detecting three fluorescent pulses from three cells streaming at an average rate of 10,000 cells/second; a square wave of a 100MHz droplet clock has been superimposed on the illustration to show the synchronization between the three pulses and the square wave pulses of the droplet clock;

Figs. 45-48 illustrates movement of a sperm cell relative to a laser beam spot having a narrow width;

Fig. 49 is an exemplary illustration of the digital information corresponding to a time-varying analog output from a photodetector detecting a single fluorescence pulse based on 122 samples at a 105MHz continuous sampling rate;

Fig. 50 is a schematic diagram illustrating the timing relationship between laser pulses, fluorescence emissions from a cell resulting from the laser pulses and the digital samples of the photodetector output;

Fig. 51 is a schematic diagram illustrating how the digital samples shown in Fig. 50 form a pulse waveform;

Fig. 52 is a schematic diagram of a pulse waveform from an X sperm cell synchronized with the pulse waveform of a Y sperm cell showing higher peak intensity in the pulse waveform for the X sperm cell;

Fig. 53 is a schematic diagram of a pulse waveform showing a threshold and integration window that can be used for pulse analysis;

Fig. 54 is a histogram of a sample containing X and Y sperm cells showing the high resolution attainable with slit scanning techniques;

Fig. 55 is histogram of a sample containing X and Y sperm cells showing the relatively poor resolution attained with standard illumination;

Figs. 56-59 show fluorescence histograms and scatter plots of peak vs. area for sperm nuclei and live sperm cells;

Figs. 60-61 illustrate a four-component model of a fluorescence intensity histogram for sperm cells-Fig. 60 shows raw data and Fig. 61 shows model curves generated by of an iterative algorithm based on the data shown in Fig. 60;

Figs. 62-63 illustrate a three-component model of a fluorescence intensity histogram for sperm cells-Fig. 62 shows raw data and Fig. 63 shows model curves generated by of an iterative algorithm based on the data shown in Fig. 62;

Fig. 64 illustrates the non-linear nature of the CSD feature; the top panel shows average M plots for X-bearing and Y-bearing sperm cells; the middle panel shows a graph of the first derivatives of these average M plots (i.e. M') for signal amplitude values less than the peak height of the average Y-bearing fluorescence emission pulse; and the bottom panel shows the difference between the first derivatives (M'x-M'y) as a function of signal amplitude;

Fig. 65 illustrates a system in which the CSD feature is the computed slope of a line that passes through two points on the fluorescence emission pulse;

Figs. 66-69 illustrate improved discrimination achieved by use of CSD feature extraction;

Fig. 70 illustrates a bi-variate sort region set on a scatter plot of CSD vs. pulse area scatter;

Fig. 71 illustrates flow cytometry re-analyses for a test in which the left panel corresponds to the high recovery/coincident accept sort strategy (no coincidence abort strategy) and the right panel corresponds to the high purity/coincident reject sort strategy (coincident abort strategy);

Fig. 72 is a work flow diagram of digital signal processing;

Fig. 73 is an example of a k-Means clustering strategy;

Fig. 74 is a conceptual illustration and graphical representation of application of a Bayes Minimum Error decision rule to pulse feature data;

Fig. 75 is graphical representation of results obtained using a Bayes Minimum Error decision rule and Mahalonobis distance thresh holding;

Fig. 76 is a conceptual illustration of moving window statistics to provide "forgetting";

Fig. 77 is a graphical representation drift compensation;

Fig. 78 illustrates a fluid stream containing an exemplary distribution of particles;

Fig. 79 is a graph showing purity as a function of fluid delivery rate with a coincident accept sort strategy;

Fig. 80 is a graph showing the percentage of desired particles successfully sorted into the usable population as a function of fluid delivery rate with a coincident reject sort strategy;

Fig. 81 is a graph showing the inverse relationship between the percentage of coincident droplets accepted for sorting into a population of desired particles compared to the percentage of coincident droplets rejected for sorting into that population;

Fig. 82 is a decision flow diagram showing the overall operation of a sorting apparatus;

Fig. 83 is a side elevation of a cytometer oriented to produce a stream of droplets having a horizontal velocity component and a collection system to collect the droplets;

Fig. 84 is an enlarged perspective view of the collection system shown in Fig. 83 shown relative to the nozzle system and deflector plates;

Fig. 85 is a schematic diagram of a collection system;

Fig. 86 is a front elevation of an intercepting device of the collection system shown in Fig. 83;

Fig. 87 is a side elevation of an intercepting device of the collection system shown in Fig. 83;

Figs. 88-95 show graphical results of several sperm centrifugation experiments;

Figs. 96-98 are schematic diagrams illustrating the steps in a filtration method;

Fig. 99 is a schematic diagram of a filtration system used to filter sperm cells;

Fig. 100 is a schematic diagram of another filtration system used to filter sperm cells;

Figs. 101 and 102 show graphical results of sperm cell filtration experiments;

Fig. 103 is a work flow diagram for cryopreservation method;

Fig. 104 shows graphical results for a sperm cell cryopreservation experiment;

Fig. 105 is a work flow diagram for a method of processing sperm cells;

Fig. 106 is a perspective view of a multi-channel particle sorter with parts broken away to show internal features of the sorter;

Fig. 107 is a perspective view of a manifold system that may be used for fluid delivery in the multi-channel particle sorter of Fig. 106;

Fig. 108 is a perspective view of the manifold system of Fig. 107 showing internal fluid connections of the manifold system;

Fig. 109 is a perspective view of the particle sorter shown in Fig. 106 with additional elements removed or partially removed to better show internal features of the sorter;

Fig. 110 is a front elevation of the particle sorter shown in Fig. 106;

Fig. 111 is a side elevation of the particle sorter shown in Fig. 106 with the side wall of the housing removed to show internal features of the sorter;

Fig. 112 is a side elevation of the particle sorter shown in Fig. 106 (taken from the side opposite the side from which Fig. 107 was taken) with the side wall removed to show internal features of the sorter;

Fig. 113 is a perspective view of the particle sorter shown in Figs. 106 taken from an angle behind the sorter and with the back cover removed to show internal features of the sorter;

Fig. 114 is a perspective view of a portion of the particle sorter shown in Fig. 106 showing the mounting of multiple nozzle systems to a cross bar;

Fig. 115 is a perspective view of a portion of the particle sorter shown in Fig. 106 showing the relative positions of the collection system and other parts of the particle sorter;

Fig. 116 is a schematic diagram of a fluid delivery system for a multi- channel sorter;

Figs. 117 and 118 are schematic diagrams of two different laser beamsplitting systems;

Figs. 119 and 120 are perspective views of another multi-channel system;

Figs. 121-134 show graphical results of various experiments;

Fig. 135 is a schematic diagram of a nozzle system wherein the nozzle directs the fluid stream through a capillary tube;

Fig. 136 is a schematic diagram of a photo damage sorting system;

Fig. 137 is a schematic diagram of an alternative sorting system based on fluidic switching; and

Fig. 138 is a schematic diagram of an alternative sorting system based on a high-speed droplet interference stream that diverts selected discrete segments of the fluid stream carrying the analyzed particles.



[0019] Corresponding parts are designated by corresponding reference numbers throughout the drawings. A parts list with associated reference numerals for each part follows. The parts list is provided with section headings generally corresponding to section headings in the specification to facilitate use of the parts list. Generally, each section of the parts list provides a reference numeral for the parts that are introduced for the first time in the corresponding section of the Detailed Description.

Parts List with Associated Reference Numerals for Each Part


General Overview



[0020] 
39
Semen Collection
41
Label Semen
41A
Add Buffer
43
Quality Control
47
Washing
48
Staining Fluid
49
Staining
51
Incubation
53
Load into Sample Introduction Device of Flow Cytometer
54
Add Sheath Fluid Through Flow Cytometry
55
Sorting
57
Collecting Sorted Sperm
58A
Add Collection Fluid
58B
Concentrate Sperm Cells
58C
Add Cryoextender
59
Load Sorted Sperm into Straws
61
Cryopreservation
63
Packing in Liquid Nitrogen
65
Distribution
67
Sales
69
Storage
71
Artificial Insemination

Flow Cytometry



[0021] 
1
System (Overall)
3
Supply of Carrier Fluid
7
Supply of Sheath Fluid
9
Flow Cytometry Apparatus Having Sorting Capabilities
15
Fluid Delivery System
17
Carrier Fluid
19
Sheath Fluid
21
Stream of Fluid
23
Stream of Particles
25
Beam of Electromagnetic Radiation
31
Electromagnetic Radiation Emission from Particles
33
Droplets
35
Particles Contained in Droplets

Flow Cytometry Apparatus (Single Channel)



[0022] 
101
Nozzle System
103
Nozzle Orifice
105
Transducer
107
Droplet Break-off
109
Optics System
115
Interrogation Location
117
Photodetector
119
Sorting System
123
First Different Group or Population of Droplets
125
Second Different Group or Population of Droplets
2201
Collection System
131
Processor

Nozzle System



[0023] 
133
Cylindrical Flow Body
135
Central Longitudinal Bore
137
Nozzle
139
Funnel-shaped Nozzle Body
141
Passage Through Nozzle Body
145
Internally Threaded Counterbore
149
Threaded Projection or Stud
155
O-ring Seal
157
Conduit (Tubular Needle)
167
Annular Space (Gap)
173
Radial Bore in Flow Body (Sheath Fluid)
183
Second Radial Bore (Additional Sheath Fluid)
189
Central Core of Carrier Fluid
191
Outer Co-axial Sheath of Fluid

Cell Orientation



[0024] 
201
Bovine Sperm Cell
205
Paddle-shaped Head
207
Flat Wide Opposite Faces
209
Narrow Edges
211
Sperm Equator
213
Nucleus
215
Tail
217
Nucleus Length
219
Head Length
221
Head Width
223
Overall Length
225
Localized Region Within Nucleus
227
Direction of Stream Flow
229
Angular Envelope in Which Light Beam Strikes Wide Face
R1
Angular Range
P
Plane

Nozzle Design



[0025] 
231
Interior of Nozzle Body
233
Interior Surface of Nozzle Body
235
First Axially Tapered Region
237
Second Axially Tapered Region
239
Third Axially Tapered Region
247
Longitudinal Axis of Nozzle
249
Fourth Region Interior of Nozzle
251
Axial Length of Fourth Region
255
Orifice Member
257
Counterbore at Front End of Nozzle
259
First Torsional Zone
261
Second Torsional Zone
263
Surface of First Torsional Zone
267
Surface of Second Torsional Zone
271
Torsional Forces
273
Axial Length of First Torsional Zone
275
Axial Length of First Tapered Region
277
Axial Length of Second Tapered Region
279
Axial Length of Second Torsional Zone
309
Conical Upstream Surface of Orifice Member
315
Cylindrical Downstream Surface of Orifice Member
317
Axial Length of Conical Upstream Surface
327
Axial Length of Downstream Surface

Orienting Baffle



[0026] 
2001
Orienting Baffle
2003
Baffle Plate
2005
Baffle Holder
2007
Upstream Leg
2009
Downstream Leg
2015
Line of Intersection
2017
Central Axis of Nozzle Body
2019
Curved Edge of Upstream Leg
2025
Distance Lower Leg Extends Downstream
2027
Overall Length of Baffle Holder
2029
Exterior Diameter of Baffle Holder
2031
Interior Diameter of Baffle Holder
2033
Distance Between Line of Intersection and Center of Nozzle
2035
Upstream End of Baffle
2037
Inclined Surface of Baffle Holder
2039
Side Edges of Downstream Leg
2041
Downstream Edge of Downstream Leg
2049
Gap Between Baffle Plate and Baffle Holder
2051
Inside surface of Baffle Holder
2053
Volume Behind Baffle Plate
2055
Interior Volume of Nozzle
2057
Longitudinal Axis of Cylindrical Baffle Holder
2059
Line Through Major Axis of Ellipse D
2061
Distance Between Injection Needle and Baffle
2067
Downstream End of Baffle Holder
2069
Contact Points Between Baffle Holder and Nozzle
2071
O-Rings
2077
Downstream End of Nozzle Holder (Boss)
2079
Interior Diameter of Boss
2081
Portion of Sheath Fluid Between Core Stream and Nozzle Surface
2087
Cross Section Upstream (A)
2089
Cross Section at Baffle (B)
2091
Cross Section at Baffle (C)
2093
Cross Section at Baffle (D)
2094
Cross Section Downstream of Baffle (E)
2097
Perpendicular Baffle System
2095
Air Bubble
2099
Perpendicular Baffle Plate
2101
Curved Edge of Perpendicular Baffle Plate
2103
Straight Edge of Perpendicular Baffle Plate
2105
O-ring
2107
Annular Shoulder (Shelf) in Nozzle
2109
Outer Diameter of Sample Injection Needle (Conduit)
2151
Nozzle System Having an Offset Sample Introduction Conduit

Nozzle Mounting and Adjustment



[0027] 
331
Nozzle Mount
333
First Linear Stage
337
Second Linear Stage
339
X Axis
341
Y Axis
343
Third Rotational Stage
345
Z Axis
347
Fixed First Stage Member (Not Shown)
349
Frame for First Fixed Stage Member
355
Movable First Stage Member
357
Actuator (Micrometer) for First Stage
359
Fixed Second Stage Member
361
Movable Second Stage Member
363
Actuator (Micrometer) for Second Stage
365
Fixed Third Stage Member
371
Movable Third Stage Member
373
Actuator (Micrometer) for Third Stage
375
Generally Upward Direction of Stream Containing Cells
377
Angle of Upward Direction

Transducer and Droplet Formation



[0028] 
379
Collar
381
Piezoelectric Element (Not Shown)
383
Terminals
D
Diameter of Stream

Break-off Sensor



[0029] 
389
Break-off Sensor
391
Microprocessor
393
Light Source
395
Linear Photoarray (Photodiodes)
401
Lens for Droplet Break-off Sensor
405
Current to Voltage Op-amp Circuits
407
Track/hold Amplifiers
409
Sinewave Generator (Track/hold Signal)
411
A/D Converter
412
Camera System
413
Strobe
414A
Mask
414B
Slit-Shaped Opening in Mask

Epi-illumination Optics System



[0030] 
415
Epi-illumination System
417
Epi-illumination Instrument
419
Longitudinal Optical Axis
425
Beam Spot
427
Axis of Focused Illumination Beam
429
Rectangular Base
431
Reflecting Filter
435
Laser or Arc-lamp
437
Conditioning Lens Assembly
439
Opening in
441
Side Wall of a Dichroic Chamber
443
Dichroic Chamber
445
Retaining Ring
447
Neutral Density Filter
449
Cylindrical Lens
455
Lens Holder
457
Jam Nut
459
Elliptical Cross Section of Beam Spot
461
Clips for Reflecting Filter
463
Filter Holder
465
Angular Face of Filter Holder
467
Openings in Filter Holder
469
Linear Stage for Filter Holder
471
X Axis
473
Outrigger
475
Actuator for Linear Stage
477
Dichroic Filter
479
Clips for Dichroic Filter
485
Frame for Dichroic Filter
487
Forward Direction
489
Longitudinal Optical Axis of the Optical Instrument
491
Focusing Lens Assembly
497
Fluorescent Pulse Waveform or Signal Emitted by Cell
498
Excitation Spatial Function
501
Microscope Adapter
503
Opening in Front Wall of Dichroic Chamber
505
Front Wall of Dichroic Chamber
507
Focusing Barrel
509
Lens Mount Barrels
511
Focusing Lens
513
Rearward Direction
515
Telescoping Focus Adjustment
517
Collimated Emitted Light
519
Filtering System
521
Emission Filter
523
Emission Filter Holder
525
Opening in Back Wall of Dichroic Chamber
527
Back Wall of Dichroic Chamber
529
Alignment Pellicle Assembly
531
Slider of Alignment Pellicle
533
Rail for Filter Assembly Components
535
Filter Holder for Alignment Pellicle
539
Pellicle Filter Element
541
Clips for Securing Filter Element to Filter Holder
543
Angle for Alignment Pellicle Relative to Optical Axis
545
Fasteners for Securing Slider to Base
547
Parallel Slots in Base
549
Aspheric Lens
551
Holder for Aspheric Lens
553
Frame for Aspheric Lens
557
Fasteners for Aspheric Lens
559
Spatial Filter
561
Aperture Plates
563
Frame for Spatial Filter Plates
567
Vertical Slit
571
Horizontal Slit
573
Aperture
575
Vertical Dimension
577
Horizontal Dimension
579
Collection Volume
583
Plate Holder
587
Fasteners for Plate Holder
589
Backing Member for Aperture Plates
449A
Adjustable Mounting Assembly
449B
Slots
449C
Slots
450
Epi-illumination that Reflects Fluorescence Emissions
451
Dichroic Filter

Photodetector



[0031] 
591
Mounting Plate for Photodetector
595
Fasteners for Photodetector

Angle of Beam Incidence



[0032] 
605
Distance Between Interrogation Location and Nozzle Orifice
609
Beam Axis
A
Angle of Incidence

Focused Beam Spot



[0033] 
L1
Length along Major Axis
W1
Width along Minor Axis

Sorting System



[0034] 
627
Charging Device
629
Charged Deflector Plates
631
Charging Element
633
Opening in Charging Element
635
Power Supply for Deflector Plates
5001
Adjustable Mounting Assembly
5003
Mounting Assembly Adjustment Board
5005
Mounting Assembly Backing
5007
Fasteners
5009
Slots
5011
Translation Axis
5013
Translation Axis
5015
Mounting Assembly Adjustment Board
5017
Fasteners
5019
Slots
5021
Fixed Support
5023
Fasteners
5025
Spring

Automat Sort Calibration



[0035] 
4201
Calibration System
4203
Epi-illumination Sensor
4205
Fiber Optic Cable
4207
Dichroic Filter
4209
Lens System
4211
Fluorescent Emission from Particle in Droplet
4213
Photodetector
4215
Beam Stop

Sort System Fault Correction



[0036] 
5047
Debris Removal System for Charging Element
5049
Debris Removal System for Deflector Plates
5051
Support for Charging Element
5053
Vacuum Passage
5055
Vacuum Line
5057
Opening Adjacent Charging Element
5058
Fitting
5059
Compressed Gas Line
5061
Manifold
5063
Air Passages
5064
Openings
5065
Fitting
5066
Side of Deflector Plate

Protection of Sorted Sample



[0037] 
4033
Collection Vessel
4041
Contamination Prevention Mechanism
4043
Pneumatic Actuator
4045
Swing Arm
4047
End of Swing Arm

Fluid Delivery System



[0038] 
645
Syringe Pump
647
Flow Line from Pump to Carrier Supply
649
Vessel for Containing Supply of Carrier Fluid
651
Line from Pump to Injection Needle
657
Supply Line from Syringe Pump to Needle
659
Variable Speed Motor
661
Second Vessel-for Supply of Sheath Fluid
667
Supply Line for Connecting Sheath Fluid to Radial Bore in Nozzle
669
Control Valve in Supply Line
671
Gas Pressure System for Sheath Fluid
675
Source of Pressurized Gas
679
Air Line for Pressurized Gas
681
Regulator for Controlling Pressure Supplied to Sheath Fluid Tank
683
Two-way Valve in Air Line

Control



[0039] 
689
A/D Converter
693
Relative Beam Intensity Experienced by Point Moving Through Beam Spot
695
Relative Emitted Pulse Intensity From Sperm Traversing Beam Spot
d
Distance Between Nozzle and Droplet Break-off Location

Signal Processing



[0040] 
701
Output Signal From Photodetector
703
Droplet Generation Clock Signals
705
Digital Signal Processing (Digital Cell Analyzer)
707
Digital Signal from A/D
735
PC/Computer Terminal
737
Master Clock (128 x Clock Signal)
739
Data Acquisition (HH1')
741
Initializing Detection Parameters (HH1)
745
Initializing Discrimination Parameters (HH2)
747
Digital Pulse Detection (HH3)
749
Digital Pulse Analysis - Feature Extraction (HH4)
753
Pulse Area (HH5)
755
Pulse Peak (HH6)
757
Pulse Discrimination (HH7)
759
Sorting (HH8)
761
Drift Analysis (HH9)
763
Decision Boundary for Bayes Rules
769
Initialize
771
System Check
773
User Interaction
775
Retry (Up to Three Times)
777
Flush
779
Bead Quality Control
781
Aspirate Sample
783
Sample Quality Control
785
Start Sample
787
Sort On
789
Sample Complete
791
Continue Sample
793
Sort Off
795
X/Y Discrimination Optimum
797
Set X/Y Discrimination
799
Discrimination OK
801
Rate Optimum
803
Set Syringe Rate
805
Rate OK
807
System Check
809
System Reset
811
System OK
813
Exemplary Overall Operational Flow
825
Integrator
827
Width/Area Comparator
829
Dynamic Threshold Calculator
831
Pulse Discrimination
833
JTAG Port I/O
837
Window Comparator (Area)
839
Pulse Width and Trigger Logic
841
Sort Decision
843
I/O Controllers
845
Slave Controllers
847
Sort Controller Board
849
USB
851
DSP Board SDRAM
853
Sort Signal
854
Low-Pass Filter
855
I/O Board SDRAM
857
Processor I/O
859
Peripheral l/O Bus
861
Sort Pulse Generator
863
Data Management Processor
865
Pulse Detection Processor
867
Feature Extraction Processor
873
Sort Processor
875
DSP Board RAM
OL
Inverse Relationship Between Coincident Droplets in Usable Population Compared to Coincident Droplets in Unusable Population
P1
Point on Line OL Corresponding to 85% Purity
LL
Point on Line OL Corresponding to 60% Collection of Desired Particles
OR
Operating Range (Segment of OL Between PI and LL)
6000
Raw Data
6001
1st Population of Non-aligned Cells
6003
2nd Population of Non-aligned Cells
6005
Aligned Y Population
6007
Aligned X Population
6010
Raw Data
6011
Population of Non-aligned Cells
6015
Aligned Y Population
6017
Aligned X Population

Multi-channel System



[0041] 
1001
Multi-channel System
1003
Flow Cytometry Units
1005
Common Particle Supply
1007
Common Source of Electromagnetic Radiation
1009
Common Housing
1011
Common Input for Control
1019
Common Output
1021
Common Fluid Delivery System
1023
Common Temperature Control System
1025
Common Power Source
1027
Common Waste Recovery System
1029
Common Deflector Plate System
1031
Common Cleaning System

Common Housing



[0042] 
1069
Base
1071
Two Side Walls
1073
Lower Pair of Shoulders
1075
Lower Cover Panel
1077
Front of Housing
1081
Upper Pair of Shoulders
1083
Upper Cover Panel
1085
Rear of Housing
1087
Framework for Mounting Multiple Cytometry Units
1089
Cross Bar Affixed to Side Walls of Housing (For Attaching Nozzle Mounts)
1093
Angled Mounting Plate Extending Between Side Walls

Common Fluid Supply



[0043] 
1105
Pump for Carrier Fluid
1107
Common Supply of Carrier Fluid
1115
Gas Pressure System for Sheath Fluid
1117
Common Supply of Sheath Fluid
1121
Manifold System
1123
Vessel Containing Common Supply of Carrier Fluid
1125
Holder for Vessel
1133
Holding Block
1135
Cavity for Receiving Vessel
1137
Second Cavity for Buffer Material
1139
Vessel for Buffer Material
1141
Syringe Pump
1147
Supply Line from Syringe Pump to Manifold
1149
Three-way Valve Controlling Carrier and Buffer Fluid
1155
Vessel for Common Supply of Sheath Fluid
1157
Supply Line from Sheath Fluid Vessel to Manifold
1161
Source of Pressurized Gas
1163
Gas Line
1165
Regulator in Gas Line
1167
Two-way Valve for Gas Line Between Gas Source and Sheath Fluid Tank
1169
Gas Line for Pressurizing a Supply of Cleaning Solution
1173
Tank for Cleaning Solution
1175
Two-way Valve for Gas Line for Cleaning Solution
1177
Manifold
1179
Laminated Block
1181
Passages
1185
Fluid Flow Circuit
1189
Inlets Connected to Syringe Pump
1191
Inlets Connected to Supply of Sheath Fluid
1193
Outlets for Carrier Fluid and Sheath Fluid
VI-V6
Valves for Controlling Flow Through Manifold Passages
1203
Frame Member (For Attaching Manifold Block)
1205
Fittings Threaded into Block
1207
Sample Reservoir
V1A-V1D
Two-way Valves (For Controlling Flow of Sample Fluid to Nozzles)
1217
Needle of Sample Reservoir
1221
Waste System
1223
Waste Tank (Receptacle)
1225
Mechanism Such as Vacuum Pump (For Generating Vacuum)
1227
Waste Lines (Connecting Valves V1A-V1D to Waste Tank)
1233
Hydrophobic Filter (In Line Connecting Waste Tank and Vacuum Pump)
1235
Fluid Circuit for Sheath Fluid
V2A-V2D
Two-Way Valves (For Controlling Flow of Sheath Fluid to Nozzles)
1241
Sheath Supply Line
1247
Waste Lines Connecting Sheath Fluid Flow Circuitry to Waste Tank

Common Power Supply and Controls



[0044] 
1249
Common Power Supply
1251
Common Power Delivery Systems
1253
Common Input (GUI)
1255
Common Output (To Microprocessor)

Common Temperature Control



[0045] 
1257
Temperature Control System
1259
Fluid Flow Circuit (For Temperature Control)
1263
Fluid Passages (For Temperature Control in Holding Block)
1265
Control Unit
1269
Fluid Passages (For Temperature Control in Manifold)
V6
Shut off Valve

Common Light Beam and Beam Splitting System



[0046] 
1270
Beam splitter
1270A
First Beam from Beam splitter
1270B
Second Beam from Beam splitter
1271
Second Beam splitter
1271A
First Beam from Second Beam splitter
1271B
Second Beam from Second Beam splitter
1272
Third Beam splitter
1272A
First Beam from Third Beam Splitter
1272B
Second Beam from Third Beam splitter
1273
Beam Guidance System
1279
Lower Filter Assembly
1281
Upper Mirror Assembly
1285
Base (For Lower Filter Assembly)
1289
Stage (For Lower Filter Assembly)
1291
Mechanism for Moving Stage (Micrometer)
1293
Tiltable Platform on the Stage
1295
Mirror (On Platform)
1297
Base (For Upper Mirror Assembly)
1299
Stage (For Upper Mirror Assembly)
1301
Tiltable Platform (For Upper Mirror Assembly)
1303
Mirror (For Upper Mirror Assembly)
1305
Mechanism for Moving Upper Stage
1309
Target Plates (Affixed to Side Wall of Housing)
1311
Vertically Aligned Holes (In the Target Plates)
1315
1st Reflecting Filter
1317
2nd Reflecting Filter
1319
3rd Reflecting Filter
1321
4th Reflecting Filter

Common Deflector Plates



[0047] 
1331
Two Common Deflector Plates
1333
Frame (For Mounting Common Deflector Plates on Housing)

Modular Multi-Channel System



[0048] 
4001
Multi-Channel System
4009
Modular Cytometry Unit
4011
Housing for Modular Unit
4013
Common Housing
4015
Laser
4017
Beam Splitting and Guidance System
4021
Hole for Laser to Enter Modular Housing
4023
Plate to Cover Exit Hole
4025
Collection System for System

Capillary Tube Nozzle System



[0049] 
1335
Capillary Tube Nozzle System
1337
Capillary Tube
1341
Chamber Filled with Light-transmitting Medium

Alternative Sorting Systems



[0050] 
1351
Photodamage Sorting System
1353
Second Laser
1355
Collection Receptacle
1357
Fluid Switching System
1359
Fluid Switching Device
1361
Capillary Branch to First Collection Vessel
1365
Capillary Branch to Second Collection Vessel
1367
Transducer (For Creating Pressure Waves for Selectively Controlling Direction of Fluid Flow)
1369
Capillary Tube on End of Nozzle
1371
Droplet Interference Stream Sorting System
1373
High-Speed Droplet Interference Stream
1375
Droplet Generation System for High-Speed Droplet Stream
1377
High-Speed Nozzle System
1379
High-Speed Fluid Stream
1381
Transducer for Droplet Interference Stream Generation
1383
High-Speed Droplets
1387
Electric Deflection Plate for High-Speed Droplet Deflection
1389
Uncharged Droplets
1391
Charged Droplets
1397
Diverted Segment of Fluid Stream
1399
Intersection of High-Speed Droplet Stream with Coaxial Fluid Stream
1403
Collection Capillaries

Collection System



[0051] 
2201
Collection System
2203
Intercepting Device
2205
Impact Surface
2207
Collection Vessel
2211
Droplet Entryway
2213
Bulb of Pipette
2215
Pipette
2217
Inside Wall of Pipette
2225
Guide Tube
2227
Collection System Frame
2229
Circular Holder
2231
Set Screw for Intercepting Device Height
2233
Mounting Plate
2235
Set Screws for Lateral Adjustment
2241
Lateral Slot
2243
Tray for Holding Collection Vessels
2245
Exit Window
2247
First Intercepting Device
2249
Second Intercepting Device
2265
Stray Droplets

Collection Fluid



[0052] 
2301
Collection Fluid

Filtration



[0053] 
2401
Filter
2403
Collection Vessel for Filtration
2405
Concentrated Slurry Containing Sperm Cells
2409
Syringe Mechanism
2411
Cannula Filter
2413
Resuspension fluid
2419
Second Container
2421
Syringe for Filtration Experiment
2423
Sample for Filtration Experiment
2425
Filter for Filtration Experiment
2427
Vacuum Pump for Filtration Experiment
2431
Syringe for Filtration Experiment II
2433
Sample for Filtration Experiment II
2435
Filter for Filtration Experiment II
2437
Filter Holder for Filtration Experiment II

Cryopreservation



[0054] 
2501
Adjust Concentration
2503
Add Cryoprotectant
2505
Add Protein Source
2507
Load in Straws
2509
Cool to Holding Temperature
2511
Maintain at Holding Temperature
2513
Cool to Temperature Approaching Critical Zone
2515
Cool Through Range of Ice Crystal Formation
2517
Immerse in Liquid Nitrogen

Common Collection System



[0055] 
2801
Common Collection System
2803
Common Frame for Intercepting Devices
2805
Waste Trough
2807
Tray for Collection Vessels

Pulsed Laser System



[0056] 
3001
Pulsed Laser
3003
Laser Pulse Sensor
3005
Laser Pulse
3007
Fluorescence Pulse Lifetime Decay
3009
Digital Sample

Detailed Description



[0057] The embodiments described below relate to collection and processing of animal semen, particularly to processing semen from a domestic animal to sort the sperm cells according to a specified DNA characteristic (e.g., X/Y chromosome content to preselect the gender of offspring).

General Overview



[0058] Figure 1 is a work flow diagram providing an overview of the steps in one exemplary sperm sorting process. The process starts with collection of neat semen samples from one or more male animals (e.g., bulls) at step 39. The semen samples are labeled for identification at step 41, contacted with a buffer, at step 41A and transported to a processing facility. In addition to the buffer, additives may also be added at step 41A, including, for example, an energy source, a protein source, an antibiotic, and/or a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly. An optional quality control test may be performed at step 43 to insure that the quality of each sample (e.g., sperm motility) is sufficient to indicate that the final product is likely to meet minimal quality criteria. An optional washing step may be performed at step 47. At step 47A the staining protocol that will be used for processing is selected by using various staining protocols to stain aliquots of the sample and then analyzing the sortability of each aliquot to identify a desired staining protocol for that particular sample. Staining according to the selected staining protocol is performed at step 49 by adding a staining fluid 48 containing a chemical dye (e.g., a DNA selective fluorescent dye) to each sample. In addition to the staining fluid, additives may also be added at step 48, including, for example, an energy source, a protein source, an antibiotic, and/or a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly. The samples are incubated at step 51 to allow for uptake of the dye by the sperm. Then a sample is loaded into the sample introduction device of a flow cytometer at step 53. The sample fluid is introduced into the flow cytometer along with a sheath fluid at step 54. In addition to the sheath fluid, additives may also be added at step 54, including, for example, an energy source, a protein source, an antibiotic, and/or a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly. At step 55 the flow cytometer sorts the sperm cells according to a specified DNA characteristic, as will be described below. As the sorted sperm cells are collected by the collection system of the flow cytometer at step 57, they are added to a collection vessel that contains a collection fluid or cryoextender at step 58A. In addition to the collection fluid, additives may also be added at step 58A, including, for example, an energy source, a protein source, an antibiotic, and/or a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly. By this time the sperm cells are in a solution that has been diluted by the various fluids added throughout the process. Accordingly, the population of sperm cells having the desired DNA characteristic are concentrated at step 58B for use in commercial artificial insemination. A cryoextender is added to the concentrated sorted sperm cells at step 58C. In addition to the cryoextender, additives may also be added at step 58C, including, for example, an energy source, a protein source, an antibiotic, and/or a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly. The sperm cells are then packed in tubular containers (referred to in the breeding industry as"straws") at step 59 and cryopreserved at step 61. The cryopreserved sperm are packed for storage in liquid nitrogen at step 63. The cryopreserved sperm are then distributed through a commercial distribution system at step 65 and sold to animal breeders at step 67. The animal breeders may store the cryopreserved sperm at step 69 until they are ready to use the sperm to artificially inseminate a female animal (e.g., cow) at step 71.

Sample Collection and Dilution


Sample Collection



[0059] The sperm sample to be sorted may be a freshly collected sample from a source animal, such as bovine, equine, porcine, or other mammalian source, or a thawed, previously cryopreserved sample. Moreover, the sample may be a single ejaculate, multiple pooled ejaculates from the same mammal, or multiple pooled ejaculates from two or more animals.

[0060] Various collection methods are known and include the gloved-hand method, use of an artificial vagina, and electro-ejaculation. The sperm are preferably collected or quickly transferred into an insulated container to avoid a rapid temperature change from physiological temperatures (typically about 35°C to about 39°C). The ejaculate typically contains about 0.5 to 15 billion sperm per milliliter, depending upon the species and particular animal.

[0061] Regardless of the method of collection, an aliquot may be drawn from the sperm sample and evaluated for various characteristics, such as for example, sperm concentration, sperm motility, sperm progressive motility, sample pH, sperm membrane integrity, and sperm morphology. This data may be obtained by examination of the sperm using, for example, the Hamilton-Thorn Motility Analyzer (IVOS), according to standard and well known procedures (see, for example, Farrell et al. Theriogenology (1998) 49 (4): 871-9; and U. S. Patent Nos. 4,896, 966 and 4,896, 967).

Dilution



[0062] The sperm sample may be combined with a buffer (in the form of a solid or solution) to form a sperm suspension. Among other things, the buffer may enhance sperm viability by buffering the suspension against significant changes in pH or osmotic pressure. Generally, a buffer is non-toxic to the cells and is compatible with the dye used to stain the cells. Exemplary buffers include phosphates, diphosphates, citrates, acetates, lactates, and combinations thereof. Presently preferred buffers include TCA, TEST, sodium citrate, HEPES, TL, TES, citric acid monohydrate, HEPEST (Gradipore, St. Louis, MO), PBS (Johnson et al., Gamete Research, 17: 203-212 (1987) ), and Dulbecco's PBS (Invitrogen Corp. , Carlsbad, CA).

[0063] One or more buffers may be combined together or with additives as discussed below to form a buffered solution, and the buffered solution combined with the sperm sample to form a sperm suspension. A buffered solution may also contain one or more additives, as described in greater detail below. Exemplary buffered solutions are described in Table I. Preferred buffered solutions include a solution comprising 3% TRIS base, 2% citric acid monohydrate, and 1% fructose (w/v) in water at a pH of about 7.0, a solution designated as TCA #1 in Table 1, and a solution designated as TCA #2 in Table I.
TABLE I. Buffered Solutions
COMPONENTSTCA# 1TCA#2TESTNa CitrateHEPESTL
Sodium chloride (NaCI)         7.6g 5.84g
Potassium chloride (KCI)         0.3g 0.23g
Sodium bicarbonate (NaHCO3)           2.1g
Sodium phosphate monobasic (NaH2PO4-H20)           0.04g
(+)-2-hydroxyproprionic acid (Na Lactate)           3.68ml
Magnesium chloride (MgC12)         0.lg 0.08g
N-(2-hydroxyethyl)piperazine-N'-(2-ethansulfonic acid) (HEPES)         2.38g 2.38g
tris(hydroxymethyl) amimo nethane (TRIS base) 30.3g 32.02g 10.28g      
Citric Acid Monohydrate 15.75g 18.68g        
Na Citrate Dihydrate       29g    
2-[(2-hydroxy-1,1-bis[hydroxymethyl] ethyl) aminoethanesulfonic acid (TES)     43.25g      
Fructo se 12.5g 2.67g   lOg 2.52g  
D-Glucose     2g      
Steptamycin     0.25g      
Penicillin-G     0.15g      
Water 1 liter 1 liter 1 liter 1 liter 1 liter 1 liter
Target pH 7.35 7.35 7.35 7.35 7.35 7.35
Target osmolality (milliosmols/kg ~314 ~300 ~302 ~316 ~298 ~296
             


[0064] Alternatively, the sperm may be combined with a metabolic inhibitor to form an inhibited sperm suspension. Metabolic inhibitors cause the sperm cells to emulate sperm cells of the epididymis of a mammal, such as for example a bull, by simulating the fluid environment of the epididymis or epididymal tract of the mammal. Such an inhibitor would reduce or inhibit the motility and metabolic activity of the sperm. Exemplary inhibitors of this class include carbonate based inhibitors, such as for example those disclosed in Salisbury & Graves, J. Reprod. Fertil., 6: 351-359 (1963). A preferred inhibitor of this type comprises NaHCO3, KHCO3, and C6H807•H2O. A more preferred inhibitor of this type comprises 0.204g NaHCO3, 0.433g KHCO3, and 0.473g and C6H807•H2O per 25mL of purified water (0.097 moles/L of NaHCO3, 0.173 moles/L of KHCO3, 0.090 moles/L and C6H807•H2O in water).

[0065] In addition to a buffer, the sperm suspension may also contain a range of additives to enhance sperm viability or motility. Exemplary additives include energy sources, protein sources, antibiotics, and compositions which regulate oxidation/reduction reactions intracellularly and/or extracellularly. One or more of these additives may be introduced into the buffer or buffered solution before the formation of the sperm suspension or, alternatively, may be separately introduced into the sperm suspension.

[0066] One or more energy sources may be added to minimize or inhibit the sperm cells from oxidizing intracellular phospholipids and other cellular components. Exemplary energy sources include monosaccharides, such as fructose, glucose, galactose and mannose, and disaccharides, such as sucrose, lactose, maltose, and trehalose, as well as other polysaccharides. For example, the resulting sperm suspension may include about 1% (w/v) to about 4% (w/v) of the energy source(s). If included, the energy source is preferably fructose and the sperm suspension contains about 2.5% (w/v).

[0067] To minimize dilution shock, provide support to the cells, or disperse the cells throughout the suspension, a protein source may also be included in the buffer, buffered solution, or sperm suspension. Exemplary protein sources include egg yolk, egg yolk extract, milk (including heat homogenized and skim), milk extract, soy protein, soy protein extract, serum albumin, bovine serum albumin, human serum substitute supplement, and combinations thereof. Albumin, and more particularly bovine serum albumin (BSA), is a preferred protein source. For example, if included, BSA may be present in the sperm suspension in an amount of less than about 5. 0% (w/v), preferably less than about 2% (w/v), more preferably less than about 1% (w/v), and most preferably in an amount of about 0. 1 % (w/v).

[0068] The use of a protein source, such BSA, alone may initiate the process of capacitation in a percentage of the sperm cells in the suspension. It is preferred that this process take place in the female reproductive tract. Therefore, in order to inhibit the initiation of capacitation during dilution, as well as during the subsequent staining and sorting, an alternative protein source or a protein substitute may be included in the sperm suspension. The alternative protein source or protein substitute possess the advantageous effects of a typical protein source, such as BSA, in addition to the ability to inhibit the initiation of capacitation in a larger percentage of the cells in the sperm suspension. Examples of alternative protein sources include human serum substitute supplement (SSS) (Irvine Scientific, Santa Ana, CA) and cholesterol enhancer BSA, while an example of a protein substitute includes a polyvinyl alcohol, such as for example, a low to medium viscosity polyvinyl alcohol generally of a molecular weight of about 30,000 to about 60,000. Generally, if included, these compositions will be present in the same amounts as disclosed above with respect to BSA, with the total albumin content of the buffer or buffered solution generally not exceeding about 5.0% (w/v).

[0069] An antibiotic may be added to the sperm suspension in order to inhibit bacterial growth. Exemplary antibiotics include, for example, tylosin, gentamicin, lincomycin, spectinomycin, Linco- Spectin® (lincomycin hydrochloride-spectinomycin), penicillin, streptomycin, ticarcillin, or any combination thereof. The antibiotics may be present in a concentration of about 50µg to about 800µg per ml of semen, regardless of whether the semen is neat, buffered, or contains additional substances, such as for example, any of the additives mentioned herein. The Certified Semen Services (CSS) and National Association of Animal Breeders (NAAB) have promulgated guidelines regarding the use of antibiotics with respect to sperm collection and use.

[0070] A composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly may also be included in the sperm suspension. Such a composition may provide a protective effect to the sperm cells, such as for example by maintaining sperm viability or progressive motility. Examples of such a composition include, for example, pyruvate, vitamin K, lipoic acid, glutathione, flavins, quinones, superoxide dismutase (SOD), and SOD mimics. If included in the sperm suspension, such a composition may be present in a concentration sufficient to effect the protective effect without detrimentally affecting sperm health. Exemplary concentration ranges include from about 10µM to about 50µM depending upon such factors as the particular composition being used or the concentration of sperm in the suspension. For example, pyruvate may be present in the sperm suspension in a concentration from about 1mM to about 50mM, preferably from about 2.5mM to about 40mM, more preferably from about 5mM to 25mM, even more preferably from about 10mM to 15mM, still more preferably about 15mM, and most preferably about 10mM. Vitamin K may be present in the sperm suspension in a concentration from about 1µM to about 100µM, preferably from about 10µM to about 100µM, and more preferably about 100µM. Lipoic acid may be present in the sperm suspension in a concentration from about 0.1mM to about 1mM, preferably from about 0.5mM to about 1mM, and more preferably about 1mM.

Staining of the Cells to be Sorted



[0071] Generally, sperm cells may be stained by forming a staining mixture comprising sperm cells, a buffer, and a dye. The sperm cells may be derived from a freshly obtained semen sample, as discussed above with respect to sample collection and dilution, or from a thawed cryopreserved semen sample.

[0072] If the semen sample is a thawed, previously cryopreserved sample, the sperm are preferably thawed immediately prior to staining. Generally, a straw or other cryopreservation vessel containing the frozen sperm may be placed in a water bath, the temperature of which is preferably in excess of the glass transition temperature of the sperm cell membrane (i.e., about 17°C), but not so great as to adversely impact sperm health. For example, frozen sperm may be thawed by immersing the cryopreservation vessel in a water bath maintained at a temperature of about 17°C to about 40°C for a period of about 30 seconds to about 90 seconds.

[0073] Once obtained, the sperm cells may be introduced into the staining mixture in the form of neat semen or in the form of a suspension derived therefrom, e.g., a sperm suspension as discussed above with respect to sample collection and dilution.

[0074] The dye may be in the form of a neat solid or a liquid composition. The dye may also be dissolved or dispersed in an unbuffered liquid to form a dye solution. Alternatively, the dye may be in the form of a dye suspension comprising a dye and a buffer or buffered solution that is biologically compatible with sperm cells. A range exemplary buffers and buffered solutions are discussed above with respect to sample collection and dilution. For example, among the buffers which may be used is a TCA buffer solution comprising 3% TRIS base, 2% citric acid monohydrate, and 1% fructose in water at a pH of about 7.0, or a carbonate-based inhibitor solution comprising 0.204g NaHCO3, 0. 433g KHCO3, and 0. 473g C6H807•H2O per 25mL of purified water (0.097 moles/L of NaHCO3, 0.173 moles/L of KHCO3, 0.090 moles/L C6H807. H20 in water). Thus, for example, a staining mixture may be formed by combining neat semen with a dye. Alternatively, the staining mixture may be formed by combining neat semen with a buffer or buffered solution and a dye. Additionally, the staining mixture may be formed by combining a sperm suspension with a dye.

[0075] The staining mixture may be formed by using one or more UV or visible light excitable, DNA selective dyes as previously described in U. S. Patent No. 5,135, 759 and WO 02/41906. Exemplary UV light excitable, selective dyes include Hoechst 33342 and Hoechst 33258, each of which is commercially available from Sigma-Aldrich (St. Louis, MO). Exemplary visible light excitable dyes include SYBR-14, commercially available from Molecular Probes, Inc. (Eugene, OR) and bisbenzimide-BODIPY® conjugate 6-{[3-((2Z)-2-{[1-(difluoroboryl)-3,5-dimethyl-1H-pyrrol-2-yl] methylene}-2H-pyrrol-5-yl) propanoyl]amino}-N-[3-(methyl{3-[({4-[6-(4-methylpiperazin-1-yl)-1H,3'H-2,5'-bibenzimidazol-2'-yl] phenoxy}acetyl)amino]propyl}amino) propyl] hexanamide ("BBC") described in WO 02/41906. Each of these dyes may be used alone or in combination; alternatively, other cell permeant UV and visible light excitable dyes may be used, alone or in combination with the aforementioned dyes, provided the dye does not detrimentally affect the viability of the sperm cells to an unacceptable degree when used in concentrations which enable sorting as described elsewhere.

[0076] The preferred concentration of the DNA selective dye in the staining mixture is a function of a range of variables which include the permeability of the cells to the selected dye, the temperature of the staining mixture, the amount of time allowed for staining to occur, and the degree of enrichment desired in the subsequent sorting step. In general, the dye concentration is preferably sufficient to achieve the desired degree of staining in a reasonably short period of time without substantially detrimentally affecting sperm viability. For example, the concentration of Hoechst 33342, Hoechst 33258, SYBR-14, or BBC in the staining mixture will generally be between about 0.1µM and about 1.0M, preferably from about 0.1µM to about 700µM, and more preferably from about 100µM to about 200µM. Accordingly, under one set of staining conditions, the concentration of Hoechst 33342 is preferably about 100 µM. Under another set of staining conditions, the concentration of Hoechst 33342 is about 150 µM. Under still another set of staining conditions the concentration is preferably about 200 µM.

[0077] In addition to buffer, other additives may be included in the staining mixture to enhance the viability or motility of the sperm; these additives may be provided as part of the sperm source, the dye source, or separately to the staining mixture. Such additives include energy sources, antibiotics, compositions which regulate oxidation/reduction reactions intracellularly and/or extracellularly, and seminal plasma, the first three of which are discussed above with respect to sample collection and dilution, and the last of which is discussed below with respect to collection fluids. Such additives may be added during the staining techniques in accordance therewith.

[0078] In particular, it has been observed that the inclusion of a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly in the staining mixture may help to maintain sperm viability at elevated staining temperatures, at elevated dye concentrations, at increased staining periods, or any combination thereof. Examples of these compositions and the use of the same are discussed above with respect to buffers and diluents. Such compositions may be added during the staining techniques in accordance therewith.

[0079] The staining mixture may be maintained at any of a range of temperatures; typically, this will be within a range of about 4°C to about 50°C. For example, the staining mixture may be maintained at a "relatively low" temperature, i.e., a temperature of about 4°C to about 30°C. Alternatively, the staining mixture may be maintained within an "intermediate" temperature range, i.e., a temperature of about 30°C to about 39°C. In addition, the staining mixture may be maintained within a "relatively high" temperature range, i.e., a temperature of about 40°C to about 50°C. Selection of a preferred temperature generally depends upon a range of variables, including for example, the permeability of the cells to the dye(s) being used, the concentration of the dye(s) in the staining mixture, the amount of time the cells will be maintained in the staining mixture, and the degree of enrichment desired in the sorting step.

[0080] Uptake of dye by the sperm cells in the staining mixture is allowed to continue for a period of time sufficient to obtain the desired degree of DNA staining. That period is typically a period sufficient for the dye to bind to the DNA of the sperm cells such that X and Y chromosome-bearing sperm cells may be sorted based upon the differing and measurable fluorescence intensity between the two. Generally, this will be no more than about 160 minutes.

Sheath Fluid



[0081] To sort the sperms cells, the stained cells are introduced as a sample fluid into the nozzle of a flow cytometer as described below. As part of the process, the sample fluid is typically surrounded by a sheath fluid. The sheath fluid permits the sperm cells in the sample fluid to be drawn out into a single file line as discussed below. The sheath fluid is collected along with the sperm cells by the collection system of the flow cytometer and therefore forms part of the post-sort environment for the sperm cells. Thus, it is desirable that the sheath fluid provides a protective effect to the cells upon contact of cells by the sheath fluid.

[0082] The sheath fluid generally comprises a buffer or buffered solution. Examples of buffers and buffered solutions and illustrative concentrations of the same that may be used in the sheath fluid are disclosed above with respect to sample collection and dilution. In a particular example, the sheath fluid comprises 0.96% Dulbecco's phosphate buffered saline (w/v), 0.1% BSA (w/v), in water at a pH of about 7.0.

[0083] Optionally, the sheath fluid may also contain a range of additives that are beneficial to sperm viability or motility. Such additives include, for example, an energy source, a protein source, an antibiotic, a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly, an alternative protein source, and polyvinyl alcohol. Each of these additives, and examples of the same, is discussed above with respect to sample collection and dilution. Such additives may be added to the sheath fluid in accordance therewith.

[0084] The sheath fluid may optionally be filtered prior to the sorting step. Contaminants that may be present in the sheath fluid, such as non-soluble particulates, may interfere with sorting. Therefore, the sheath fluid may be filtered prior to its introduction into a flow cytometer. Such filters and methods of using the same are well known in the art. Generally, the filter is a membrane of about 0.1 microns to about 0.5 microns.

[0085] The stained cells may be introduced into the sheath fluid at any time subsequent to staining. Typically, a stream of the stained cells in the sample fluid is injected into a stream of sheath fluid within the nozzle of the flow cytometer. Initially, there is substantially no contacting of the sample fluid and the sheath fluid due to laminar flow of the fluids as discussed in more detail below. It is desirable that the sample fluid and the sheath fluid remain as substantially discrete flowing streams until after the particles (e.g., the stained sperm cells) in the sample fluid have been analyzed. At some point, however, the sheath fluid and the cells of the sample fluid come in contact with one another. For instance in a droplet sorting flow cytometer (discussed below) the sheath fluid and sample fluid begin contacting one another as droplets are being formed downstream of the interrogation location.

[0086] At the time of the introduction of the stained cells and the sheath fluid, both the stained cells and the sheath fluid may be at a temperature from about 4°C to about 50°C. The sheath fluid and the stained cells may be at the same or at different temperatures, with either being at a higher temperature than the other. Accordingly, at the time of the introduction of the stained cells and the sheath fluid, both the cells and the sheath fluid are at the same temperature; for example, at a "relatively low" temperature, such as for example at about 5°C to about 8°C; at an "intermediate" temperature, such as for example at about 25°C to about 30°C; or at a "relatively high" temperature, such as for example at about 40°C to about 43°C.

FLOW CYTOMETRY



[0087] Inventive technologies in flow cytometry may be used for sorting x and Y sperm cells. Referring now to Figs. 2 and 3, one example of a flow cytometry system is designated in its entirety by the reference numeral 1. As will appear, the flow cytometry system 1 is useful for classifying and sorting particles, such as sperm cells, according to selected characteristics. In general, the system 1 comprises a supply 3 of carrier fluid 17 containing particles to be sorted, a supply 7 of sheath fluid 19, flow cytometry apparatus having sorting capabilities, generally designated 9, and a fluid delivery system 15 for delivering the carrier 17 and sheath fluids 19 from respective supplies 3, 7 under pressure to the flow cytometry apparatus 9. The flow cytometry apparatus 9 is adapted for receiving the carrier 17 and sheath 19 fluids, for combining the fluids 17, 19 to create a stream of pressurized fluid 21, for directing the stream 21 carrying the particles through a focused beam of electromagnetic radiation 25 (e.g., UV laser light), and for analyzing the electromagnetic radiation 31 (e.g., fluorescent light) emitted by particles passing through the focused beam 25. The apparatus 9 also functions to break the stream 21 up into droplets 33 containing particles to be evaluated, and to sort the droplets 33 based on the aforesaid measurements according to one or more characteristics of the particles contained in the droplets 33. While this system may be used to analyze and preferably sort any type of particle, it has particular application to sorting cells according to one or more desired characteristics of the cells (e.g., size, DNA content, shape, density, gene sequence, etc.). This system is especially suited for sorting animal sperm cells for commercial use by the animal production industry for in vivo or in vitro artificial insemination.

SINGLE-CHANNEL SORTING APPARATUS AND METHOD


Flow Cytometry Apparatus



[0088] The flow cytometry apparatus 9 shown in Fig. 3 comprises a nozzle system, generally designated 101, for delivering a fluid stream 21 containing particles (e.g., stained sperm cells) through a nozzle orifice 103 under pressure with the cells substantially in single file and, in the case of sperm cells, with asymmetric heads of the sperm cells substantially in a desired orientation which will be described. As in conventional flow cytometry droplet sorting systems, a transducer 105 is provided opposite the nozzle orifice 103 for introducing acoustical energy into the fluid stream 21 which causes the stream 21 to break into droplets 33 containing individual cells at a "droplet break-off' location 107 spaced from the nozzle orifice 103. The system 1 also includes an optics system, generally designated 109, for focusing a beam of electromagnetic radiation 25 (e.g., 350-700 nm UV or visible laser light) on the fluid stream 21 at an "interrogation" location 115 which, in the described example, is between the nozzle orifice 103 and the droplet break-off location 107. Thus, the described example is a jet-in-air system. In other examples, the interrogation location 107 could be inside the nozzle orifice 103 or upstream from the orifice 103. In any event, the cells are adapted to pass through the beam of light 25 at the interrogation location 107, resulting in excitation of a chemical stain (or other reporting medium) in the cells to cause fluorescence emissions 31 having a wavelength different from that of the beam 25 (e.g., if the illumination light 25 has a wavelength of about 350 to 370 nm, the fluorescent emissions 31 may have a wavelength of about 460 nm). A photodetector 117 is operable to detect these emissions 31 and to convert them into electrical signals which are processed and used to classify the cells according to selected characteristics, such as the X/Y chromosome content of sperm cells. The flow cytometry apparatus 9 further comprises a sorting system, generally designated 119, for sorting the droplets 33 into different groups or populations (e.g., two populations 123,125) according to the classification of the cells contained in the droplets 33 and a collection system, generally designated 2201 (Fig. 2), for collecting the droplets 33 and maintaining the segregation of the different populations 123,125.

[0089] Operation of the system 1 is controlled by a processor 131, such as microprocessor or other digital or analog control and/or processor, or combinations thereof, which controls the various functions of the components of the system 1 in a manner to be described. Significantly, the processor 131 is also responsive to particle analysis information to control the output of the system 1 based on selected control and sorting strategies involving different parameters, including the desired purity of one of the sorted populations of particles, the acceptable quantity (or percentage) of desired particles one of the populations as compared to the quantity (or percentage) of desired particles in one or more of the other populations, and other parameters, as will be discussed later.

Nozzle System



[0090] Referring to Figs. 4 and 5, the nozzle system 101 may comprise, a generally cylindrical flow body 133 having a central longitudinal bore 135 through it, and a nozzle 137 on the flow body 133 having a funnel-shaped nozzle body 139. A passage 141 extends through the nozzle body 139 co-axial with the bore 135 in the flow body 133 and terminates in the aforementioned nozzle orifice 103 at the forward end of the nozzle 137. The nozzle body 139 has an internally threaded counterbore 145 at its rearward end for threadably receiving a threaded projection or stud 149 at the forward end of the flow body 133 to removably connect the nozzle 137 to the flow body 133, the connection being sealed by an O-ring seal 155. It will be understood that the nozzle can be removably connected to the flow body in other ways or, alternatively, the parts could be integrally formed as one piece.

[0091] Particles are delivered to the nozzle 137 by means of a conduit 157 positioned co-axially in the bore 135 of the flow body 133. The outside diameter of the conduit 157 is less than the inside diameter of the bore 135 so that an annular space 167 is formed around the conduit 157. Conduit 157 is a tubular needle (e.g., a 16-ga. needle having an inside diameter of 0.01 in.) having a front end which extends into the counterbore 145 at the back of the nozzle 137. The back end of the conduit 157 is connected to the fluid delivery system 15 for delivery of carrier fluid 17 (e.g., a staining mixture containing sperm cells) to the conduit 157. The annular space 167 surrounding the conduit 157 is connected by means of a radial bore 173 in the flow body 133 to the fluid delivery system 15 for delivery of sheath fluid 19 into the annular space 167. As shown in Figs. 3 and 5, an optional second radial bore 183 may be provided in the flow body 133 connecting the annular space 167 to another line (not shown) for supply of additional sheath fluid 19 to the nozzle 137.

[0092] As in conventional flow cytometry systems, sheath fluid 19 is introduced into the annular space 167 surrounding the conduit 157. The velocity of the sheath fluid 19 as it flows past the tip of the conduit 157 is much higher that the velocity of the carrier fluid 17 exiting the conduit 157, so that the carrier fluid 17 and cells (e.g., sperm cells) contained therein are accelerated by the sheath fluid 19 toward the orifice 103 of the nozzle 137. This acceleration functions to space the cells out generally in a single file arrangement for separate analysis by the optics system 109. The sheath fluid 19 surrounds the carrier fluid 17, resulting in the fluid stream 21 having a central core 189 of carrier fluid 17 and an outer co-axial sheath 191 of sheath fluid 19 surrounding the central core 189 (see Fig. 6). As will be understood by those skilled in flow cytometry, the laminar flow and hydrodynamic focusing of the central core 189 tends to confine the particles to the core 189, with little mixing of the sheath 19 and carrier fluids 17 in the nozzle 137. Further, the central core 189 remains essentially intact within the sheath 191 as the stream 21 moves through the nozzle system 101, until such time as droplets 33 are formed at the break-off location 107. This type of co-axial flow is particularly suited for flow cytometry, because the particles to be analyzed are confined within the relatively narrow core 189 of the stream. As a result, a beam of light 25 focused on the center or core 189 of the stream 21 will illuminate the particles so that they may be analyzed substantially one at a time. By confining the core 189 within a sufficiently narrow diameter, one can obtain more uniform illumination of the particles in the core fluid 189. For good analytical results, the diameter of the core containing the particles should desirably be within a range of 7 to 20 microns, and more desirably within a range of 7 to 14 microns. The diameter of the core stream 189 can be increased or decreased by adjusting the rate of delivery of the carrier fluid 17 relative to the rate of delivery of the sheath fluid 19.

Cell Orientation



[0093] For optimizing analytical results, it is desirable that particles having asymmetric shapes be in a desired orientation when they pass through the light beam from the optics system. As is known to those skilled in the art, fluorescence emissions from asymmetric particles tend to anisotropic (i.e., the intensity of the emissions are not uniform in all directions). As used herein, the term "desired orientation" means an orientation which allows the processing system to discriminate between cells having different characteristics with an accuracy in a range of 70% to 100%.

[0094] To illustrate the point, a bovine sperm cell 201 is illustrated in Figs. 6 and 7. Typically, the cell has a paddle-shaped head 205 with relatively flat wide opposite faces 207 and narrow edges 209, a nucleus 213 in the head 205 containing the chromatic DNA mass of the cell, and a tail 215 extending from the head 205 providing the motility necessary for effective fertilization. The average bovine sperm cell 201 has a head length 219 of about 8µm, a head width 221 of about 4µm, and an overall length 223 from the front of the head to the end of the tail of about 100µm. In the average bovine sperm cell 201, the nucleus 213 occupies most of the head volume and is only slightly smaller than the sperm head 205. Thus, the nucleus length 217 is almost equal to the head length 219, again being about 8µm in length. It has been observed that in the bovine the X/Y chromosomes of the sperm cells 201 are localized in a region of the nucleus 225 (Fig. 6) below and immediately adjacent the longitudinal midline or equator 211 or center of the head 205. More specifically, this sub-equatorial region 225 extends no more than about 20% of the nucleus length 217 on the lower half (toward the tail 215) of the nucleus 213, and still more specifically no more than about 1.0-1.5µm below the equator 211 of the nucleus 213.

[0095] When sperm cells pass through the excitation beam 25, it is desirable that the cells be substantially in single file and that the head 205 of the each cell 201 be substantially similarly oriented to reduce orientation variability from cell to cell and thus provide for a more uniform measurement of the cells. It is also desired that the cells have an orientation which will enable accurate discrimination between X and Y cells. Desirably, this orientation is one where the length of the sperm cell 201 is generally aligned with the direction of stream flow 227 (either head leading (shown Fig. 6) or head trailing) and where the head 205 of the sperm cell 201 is rotated on its longitudinal axis so that the head 205 falls within an angular envelope 229 in which the light beam 25 from the optics system 109 will strike a wide face 207 of the cell 201 generally broadside, as shown schematically in Fig. 7, rather than a narrow edge 209 of the cell. Preferably, the envelope 229 defining the desired orientation is generated by rotation of a sperm cell 201 through an angular range of R1 relative to a plane P which is generally perpendicular to the incoming light beam 25, as viewed in a cross section taken transversely through the stream 21. The range R1 is preferably 0 to 90 degrees. The nozzlemay be configured to achieve this desired orientation with an accuracy of up to 90% or more.

[0096] The tolerance for sperm orientation (i.e., the size of the envelope 229 defined by angular range R1) is related to the numerical aperture of the lens used to collect fluorescence emissions 31 from the sperm cells. In the example shown Fig. 7, for example, the optics system 109 has a fluorescence emission 31 detection volume 579 defined by a solid angle of 55 degrees. When the rotational orientation of a sperm head 205 is outside the envelope 229 defined by R1 as the sperm moves through the beam 25, a relatively stronger fluorescence emission 31 from an edge 209 of the sperm head 205 will be collected by the optic system 109, preventing the processor 131 from correlating the intensity of the fluorescence emission 31 with the X/Y chromosome content of the sperm cell 201. However, the optics system 109 does not collect the relatively stronger fluorescence emissions 31 from the narrow edge 209 of the sperm heads 205 as long as the rotational orientation of a sperm head 205 is within the envelope 229 as it passes through the interrogation location 115. Thus, in the example shown Fig. 7, the orientation of the sperm cell does not result in collection of the relatively stronger edge-wise fluorescence emissions as long as the narrow edges 209 of the sperm head 205 are confined within angle R1. The solid angle of the collection volume 579 can be decreased by using a lens with a smaller numerical aperture, thereby increasing angle R1 and the tolerance for poorly oriented sperm. However, this also decreases the number of photons that can be collected by the optics system 109, which can impact the measurement of fluorescence emissions 31 by reducing the intensity of the emissions 31 detected by the photodetector. Likewise, if the optics system 109 collects fluorescence emissions 31 with a high numerical aperture lens to obtain a stronger intensity of the fluorescence emissions detected by the photodetector, then the tolerance for sperm orientation decreases. Thus, in designing a system, one needs to strike a balance between the tolerance for sperm orientation and the numerical lens aperture. The optimal balance will depend on the orienting capabilities and optical sensitivity of the system. For example, a lens having a numerical aperture 0.65 is used.

Nozzle Design



[0097] As shown in Figs. 8 and 9, the interior 231 of the nozzle body 139 downstream from the counterbore 145 may have an interior surface 233 comprising first, second and third axially tapered regions 235, 237, 239 for progressively accelerating the speed of the fluid stream 21 in a downstream direction toward the nozzle orifice 103. As noted previously, this acceleration functions to space the particles (e.g., cells) in the stream 21 so they assume a generally single file formation so they can be analyzed substantially one particle at a time. At least two of these regions, and preferably all three 235, 237, 239, have generally elliptical (oval) shapes in cross sections taken at right angles to the longitudinal axis 247 of the nozzle 137, as is shown in Figs. 9A-9H and Figs. 9J-9K. The interior surface 233 of the nozzle body 139 also has a fourth region 249, not tapered, downstream from the first three regions 235, 237, 239 and immediately upstream of the nozzle orifice 103 which may be formed in a separate orifice member 255 secured in a counterbore 257 at the front of the nozzle body 139. The generally elliptical cross sectional shapes of the first 235 and second 237 regions may be oriented in substantially the same direction to define a first torsional zone 259, and the generally elliptical cross sectional shape of the third region 239, constituting a second torsional zone 261, is oriented at an angle (e.g., about 90 degrees) relative to the generally elliptical cross sectional shapes of the first 235 and second 237 regions. The orientation is such that the interior surface 233 of the nozzle body 139 applies torsional forces to the fluid stream 21 and thereby tends to orient the sperm cells 201 in the aforestated desired orientation as they pass through the nozzle orifice 103. Preferably, the first torsional zone 259 has an axial length 273 of 3.0-4.5 mm and the first 235 and second 237 tapered regions making up the zone 259 have approximately equal axial lengths 275, 277 (e.g., about 1.8 mm). The second torsional zone 261 has an axial length 279 of 3.5-5.0 mm. The fourth region 249 is preferably generally cylindrical in shape. Each generally cross-sectional elliptical shape A-D (Fig. 8) at the boundaries of the first 235, second 237 and third 239 regions has a major axis diameter and a minor axis diameter, exemplary dimensions of which are shown in Fig. 8 and Table II below.
TABLE II
EllipseMajor Axis Diameter (mm)Minor Axis Diameter (mm)Ratio
A 7.0 6.0 1.2
B 6.1 5.3 1.15
C 2.1 2.1 1
D 0.9 0.2 1.45


[0098] It will be understood that the above dimensions are exemplary, and that other dimensions and shapes may also be suitable. Functionally, the changes in the ratios between the major and minor diameters, and the different orientations of the elliptical shapes of the regions, create side forces which act on each cell 201 and apply a torsional force 271 tending to rotate the cell 201 on its longitudinal axis so that its wide faces 207 align with the minor axis in the first torsional zone 259 and as the cell is gently twisted (e.g., 90 degrees) to align with the minor axis of the second torsional zone 261. Each of the tapered surfaces 235, 237, 239 also serves to accelerate the stream 21 (and cells) flowing through the nozzle 101. The acceleration may increase more gradually in the first 235 and third 239 regions and more rapidly in the second region 237. Again by way of example, the taper of the first region 235 may range from about 11-14 degrees; the taper in the second region 237 may range from about 42-48 degrees; and the taper in the third region 239 may vary from about 8-12 degrees. The nozzle body 139 is formed from a suitable material such as molded plastic (ABS) or metal.

[0099] The orifice member 255 (Fig. 8) is preferably formed from a hard, wear resistant material, such as sapphire, which is capable of being machined or otherwise formed with precise dimensions. The orifice member 255 itself has, in one example, a conical upstream surface 309 of generally circular cross section which decreases in diameter from about 0.92 mm to about 0.060 mm and has an axial length 317 of about 0.54 mm and a taper angle of about 39 degrees. The orifice member 255 also has a generally cylindrical downstream surface 315 with a diameter of about 0.060 mm and an axial length 327 of about 0.36 mm. These dimensions are exemplary only, and it will be understood that the orifice member 255 may have other sizes and shapes. For example, the shape of the upstream surface 309 may be generally elliptical (oval) in cross section, and the diameter of the orifice 103 at the downstream end of the nozzle 137 may range from 40 to 100 microns or more. It is desirable that the size of the orifice 103 be such that the cells exiting the nozzle 101 are substantially in single file formation within the core 189 of the stream 21 and substantially in the desired orientation, as described previously. For example, in the case of sperm cells an orifice 103 having a diameter of about 60-62 microns at the downstream end has been found to be suitable. Preferably, the nozzle orifice 103 serves to further accelerate the stream 21 and to shape and size the stream 21 for optimum cell spacing, cell orientation and droplet 33 formation, as will be described.

[0100] The velocity of the cells as they exit the nozzle 137 will depend on various factors, including the pressure at which sheath fluid 19 is introduced into the nozzle system 101. At a pressure of 1.88 bar (20 psi), the cells will exit the nozzle orifice 103 of the above example at a velocity of about 16.6 m/s as a generally cylindrical stream 21 containing cells which are substantially similarly oriented at the core 189 of the stream 21. At a sheath pressure of 2.07 bar (30 psi), the cell velocity will be about 20.3 m/s. At different sheath fluid 19 pressures, the velocity of the stream 21 will vary.

Introduction of Core Stream to Torsional Zone



[0101] Improved orientation of particles may be obtained by altering the flow of the fluid stream 21 through an orienting nozzle so that the core stream 189 containing the particles to be oriented (e.g., sperm cells) is directed along a flow path, at least a portion of which is offset from the center of the nozzle so that the particles are subjected to the hydrodynamic orienting forces generated by a nozzle while they are at a location that is offset from the center of the nozzle. Directing the core stream 189 along an offset flow path may also improve orientation of particles in a traditional nozzle (i.e., one that does not have any torsional zones). In many nozzles, one can determine that a given position is offset from the center of the nozzle because it is displaced from a longitudinal axis of the nozzle. One can also recognize that a particular position is offset from the center of a nozzle because it is displaced from the geometric center of a cross sectional area of the nozzle through which the fluid stream flows.

[0102] A number of techniques may be used to direct the core stream 189 along a flow path that is offset from the center of the nozzle. For example, an orienting baffle may be positioned in the nozzle to deflect the core stream to one side of the nozzle. Similarly, the conduit 157 for introducing the core stream 189 containing the sample particles may be relocated from the traditional center of the nozzle to an offset location. Furthermore, it is contemplated that an offset sample introduction conduit 157 may be used in combination with an orienting baffle. Examples of use of an orienting baffle and use of an offset sample introduction conduit are discussed below.

[0103] The improved orientation of particles (e.g., sperm cells) achieved by use of an orienting baffle and/or offset sample introduction conduit 157 may be due to a number of factors. One factor is that the deflection of the core stream 189 and/or a change in the size and shape of the cross sectional flow area results in application of hydrodynamic forces that tend to orient asymmetric particles. (Kachel, et al., Histochemistry and Cytochemistry, 25 (7): 774-80 (1977)). Another factor is that it has been found that asymmetric particles (in particular sperm cells) tend to orient as they flow in a fluid stream in close proximity to a solid surface. Thus, by directing the core stream 189 so that it is in close proximity to the interior surface of a nozzle or a baffle surface one can obtain improved orientation of the particles. Furthermore, a baffle and/or offset sample introduction conduit can be used in conjunction with an orienting nozzle which applies additional orienting forces (e.g., torsional forces) to the asymmetric particles. In that case, the baffle can operate to direct the fluid stream so that the core stream containing the particles to be oriented flows along a path that is offset from the center of the nozzle while the particles are subjected to the torsional forces generated by one or more of the torsional zones.

Orienting Baffle



[0104] Figs. 10-13 show one exemplary orienting baffle, generally designated 2001, positioned in the orienting nozzle 137 described above. However, the baffle 2001 could be used in conjunction with a different nozzle, including a non-orienting nozzle. The baffle 2001 is positioned in the nozzle upstream from the orifice 103 and downstream from the sample injection needle 157. Referring to Figs. 14 and 15, the baffle comprises a baffle plate 2003 that is held in place by a baffle holder 2005. As shown, the baffle plate 2003 is generally L-shaped and constructed of a substantially rigid, durable and corrosion-resistant material (e.g., stainless steel). The L-shaped plate 2003 has an upstream leg 2007 and a downstream leg 2009, which are desirably substantially perpendicular to each other (e.g., within about 5 degrees of being perpendicular). As shown in the drawings, the two legs 2007, 2009 of the L-shaped plate 2003 intersect at a line 2015 that is perpendicular to the longitudinal axis 2017 of the nozzle 137 (Fig. 11). As shown in Fig. 14, the line of intersection 2015 is also spaced a short distance 2033 (e.g., about 0.3 mm) away from the longitudinal axis 2057 of the baffle holder 2005. The upstream leg 2007 of the L-shaped plate 2003 extends from the line of intersection 2015 away from the longitudinal axis 2017 of the nozzle 137 all the way to the edge of the baffle holder 2005, as shown in Fig. 15. Thus, the upstream leg 2007 is formed with a curved edge 2019 that closely matches the shape of the baffle holder 2005. As shown in Fig. 14, the upstream leg 2007 is inclined at an angle AA of about 15-25 degrees from perpendicular to the longitudinal axis 2057 of the baffle holder 2005. The downstream leg 2009 of the L-shaped plate 2007 extends downstream from the line of intersection 2015 of the two legs 2007, 2009 a distance 2025 of about 2.0-2.5 mm at an angle BB that is in the range of about 60-80 degrees from perpendicular to the longitudinal axis 2057 of the baffle holder 2005.

[0105] The baffle holder 2005 is sized and shaped to fit inside the nozzle 137, as shown in Figs. 10-13. The baffle holder 2005 is preferably made of a moldable material (e.g., polypropylene) although the baffle holder 2005 may be constructed from other materials. The baffle holder 2005 , shown in Figs. 14 and 15, is generally shaped as a hollow cylindrical shell about 4.0-4.5 mm in overall length 2027. The baffle holder 2005 has an exterior diameter 2029 of about 5-6 mm and an interior diameter 2031 of about 2.5-3.5 mm. If the baffle holder 2005 is to be molded, a minor draft (not shown) can be provided on the surfaces of the holder 2005 (e.g., to allow the baffle holder to be easily removed from an injection molding machine). The upstream end 2035 of the exemplary baffle holder 2005 has an inclined surface 2037 which is inclined at the same angle AA as the upstream leg 2007 of the L-shaped plate 2003. The upstream leg 2007 of the L-shaped plate 2003 abuts against and is supported by the inclined surface 2037 of the baffle holder 2005. The side edges 2039 (Fig. 15) of the downstream leg 2009 of the L-shaped plate 2003 are partially embedded (e.g., received in slots) in the baffle holder 2005 to hold the baffle plate 2003 in a position in which the downstream leg 2009 spans generally from one side of the baffle holder 2005 to the other. The downstream edge 2041 of the downstream leg 2009 forms a straight line which is generally perpendicular to the longitudinal axis 2057 of the baffle holder 2057. There is a gap 2049 (Fig. 14) between the downstream edge 2041 of the downstream leg 2009 and the interior cylindrical surface 2051 of the baffle holder 2005. The gap 2049 provides fluid communication between a volume 2053 defined by the legs 2007, 2009 of the L-shaped plate 2003 and the interior cylindrical surface 2051 of the baffle holder 2003 and the rest of the interior volume 2055 of the nozzle 137.

[0106] The baffle holder 2005 is desirably positioned inside the nozzle with the longitudinal axis 2057 of the baffle holder 2005 generally aligned with the longitudinal axis 2017 of the nozzle 137 so that it holds the L-shaped plate 2003 in the position described above. Desirably, the exemplary baffle plate 2003 is rotationally oriented so that the line of intersection 2015 of the two legs 2007, 2009 of the plate 2003 is parallel to a line 2059 running through the major axis of ellipse D, as shown in Fig. 16. However, the exemplary baffle 2001 also performs well when the intersection 2015 of the two legs 2007, 2009 of the L-shaped plate 2003 is perpendicular to the line 2059 running through the major axis of ellipse D, as shown in Fig. 17. Furthermore, the baffle may have any rotational orientation. As shown in Fig. 12, the sample injection needle 157 is desirably a distance 2061 of about 0.25-1.0 mm upstream from the most upstream portion 2035 of the baffle 2001. More desirably, the sample injection needle 157 is about 0.55-0.65 mm upstream from the most upstream portion 2035 of the baffle 2001.

[0107] The baffle holder 2005 may be held in a desired position relative to the nozzle in any number of ways. Referring to Fig. 14, the downstream end 2067 of the baffle holder 2005 is stepped so that it fits farther downstream in the nozzle 137. The stepped downstream end 2067 of the holder 2005 is circular in shape and abuts against the elliptically shaped interior surface 233 of the nozzle 137. Thus, the contact between the interior surface 233 of the nozzle 137 and the baffle holder 2005 is generally limited to two points 2069, as shown in Fig. 13. A pair of O-rings 2071 are positioned around the baffle holder 2005 between the nozzle 137 and the threaded projection 149 of the flow body 133 (Figs. 11-13) and seal the nozzle system 101 against leakage. The O-rings 2071 may be made of Viton®, or any other similar materials. The two O-rings 2071 are compressed as the nozzle 137 is screwed onto the threaded projection 149 to provide a fluid-tight seal. Two O-rings 2071 are used because a single O-ring cannot be compressed within the space between the nozzle 137 and the flow body 133 due to the length 2027 of the baffle holder 2005. Any number of O-rings or a different type of seal could be used, provided that the number of O-rings or other type of seal is selected so that there will be a fluid-tight seal when the nozzle 137 is screwed onto the flow body 133. This will depend on a number of factors, including the size and shape of the nozzle 137, flow body 133, baffle holder 2005, and O-rings 2071 as well as the type of seal. The O-rings 2071 also help hold the baffle holder 2005 in the desired position. The O-rings 2071 occupy the space around the baffle holder 2005, thereby restricting side-to-side movement of the baffle holder 2005 inside the nozzle 137. Frictional forces between the O-rings 2071 and the baffle holder 2005 also resist rotational movement of the baffle holder 2005.

[0108] When the nozzle 137 is tightened on the flow body 133 as shown in Fig. 12, the downstream end 2077 of the threaded projection 149 from the flow body 133, in the form of a boss, is approximately even with the most upstream portion 2035 of the baffle 2001. As a result, the baffle holder 2005 is held axially captive between the flow body 133 (at the upstream end 2035 of the baffle holder 2005) and the interior surface 233 of the nozzle 137 (at the downstream end 2067 of the baffle holder 2005). Other retaining mechanisms may be used. As shown in the drawings, the interior diameter of the boss 2079 (Fig. 12) at the downstream end of the threaded projection 149 is roughly equal to the internal diameter 2031 of the baffle holder 2005.

[0109] Those skilled in the art will recognize that the flow through the nozzle system 101 remains laminar notwithstanding the baffle 2001 because the small cross sectional area through which the fluids must flow results in a low Reynolds number for the flow. As is shown in Fig. 11, the baffle deflects the core stream 189 and sheath stream 191 away from the central longitudinal axis 2017 of the nozzle 137 and toward an interior surface 233 of the nozzle 137. The core stream 189 may also flow very close to the interior surface 233 of the nozzle 137 as the core stream 189 passes between the transition between the first 259 and second 261 torsional zones. However, a portion 2081 of the sheath fluid stream 191 remains between the core stream 189 and the interior surface 233 of the nozzle 137 so the particles in the core stream 189 do not actually impact or contact the interior surface 233 of the nozzle 137. Farther downstream in the nozzle 137, the hydrodynamic forces push the core stream 189 back toward the center of the nozzle 137 (e.g., in alignment with the longitudinal axis 2017 of the nozzle 137).

[0110] Referring to Figs. 18A-18E, the baffle 2001 changes the shape and reduces the size of the cross sectional flow area in the nozzle 137. (For the sake of clarity, Figs. 18A-18E do not show any nozzle structure downstream from the baffle. The flow area in each of the Figs. 18A- 18E is outlined in bold for clarity.) Upstream from the baffle 2001 (Fig. 18A), the cross sectional flow area 2087 is generally circular or elliptical. At the upstream end 2035 of the baffle 2001, the flow area begins to change from a circular shape to a generally semi-circular shape 2089 at the intersection 2015 of the legs 2007, 2009 of the baffle plate 2003 (Fig. 18B), although other shapes may be suitable. There the cross sectional flow area 2089 is smaller than the flow area 2087 upstream from the baffle. Fig. 18C illustrates the flow area 2091 as fluid flows through a part of the baffle holder 2005, and Fig. 18D illustrates the flow area 2093 farther downstream at the downstream end 2041 of the downstream leg 2009 of the baffle plate 2003. It will be observed that flow area 2093 is somewhat larger than flow area 2091 due to the angular orientation of the downstream leg 2009 of the baffle plate 2003. Downstream from the baffle plate 2003 (Fig. 18E) the flow area 2094 through the baffle corresponds the shape of the interior surface 2051 of the baffle holder 2005, which is circular in the illustrated drawing (other shapes may be suitable). Downstream from the baffle holder 2005 the torsional zones 259,261 of the nozzle 137 desirably provide torsional forces as discussed above.

[0111] As shown in Fig. 11, it has been observed that one or more air bubbles 2095 may become trapped in the volume 2053 between the downstream leg 2009 of the L-shaped plate 2003 and the baffle holder 2005. Furthermore, a portion of a bubble 2095 may extend through the gap 2049 between the edge 2041 of the downstream leg 2009 and the baffle holder 2005. Thus, the air bubble(s) 2095 can occupy a portion of the cross sectional flow area downstream of the downstream leg 2009 of the L-shaped plate 2003, perhaps affecting the flow of fluid through the nozzle 137. The exemplary baffle 2001 has been found to work well both with and without the air bubble(s) 2095. Thus, a baffle can be used to orient sperm cells without involvement of any bubbles.

[0112] Another exemplary orienting baffle, generally designated 2097, is shown in Figs. 19 and 20. The baffle 2097 comprises a flat generally semi-circular baffle plate 2099 in the orienting nozzle 137 discussed above. The baffle plate 2099 is positioned in the nozzle 137 downstream of the sample introduction conduit 157 and generally perpendicular to the longitudinal axis 2017 of the nozzle 137. The baffle plate 2099 has a curved edge 2101 that generally matches the curvature of the interior surface 233 of the nozzle 137 so that there are no large gaps between the curved edge 2101 of the baffle plate 2099 and the interior surface 233 of the nozzle 137. The baffle plate 2099 also has a straight edge 2103 that extends a short distance past longitudinal axis 2017 of the nozzle 137 so that it is approximately aligned with the outer diameter 2109 of the sample introduction conduit 157. The baffle plate 2099 is held in position by friction resulting from compression of the baffle plate 2099 between an o-ring seal 2105, which is similar to the o-ring seals 2071 described in connection with the L-shaped baffle 2001 above, and an annular shoulder or shelf 2107 formed on the interior of the nozzle 137. As shown in Fig. 19 the orienting baffle 2099 operates by deflecting the fluid stream so that the core stream 189 containing the particles to be analyzed is offset from the central longitudinal axis 2017 of the nozzle 137 along a portion of its flow path. For example, the core stream 189 may be directed along a flow path that is offset from the longitudinal axis 2017 of the nozzle 137 as it flows through the first torsional zone 259, as well as at least a portion of the second torsional zone 261. Consequently, the particles (e.g., sperm cells) are subjected to the torsional forces generated by the torsional zones 259, 261 while they are in a position that is offset from the central longitudinal axis 2017 of the nozzle 137.

[0113] Those skilled in the art will recognize that substantial changes may be made to the exemplary baffles 2001, 2097 described above. All that is required is that the baffle be configured to deflect the core stream 189 and sheath stream 191 toward an interior surface of the nozzle or to cause the core 189 and sheath stream 191 to flow through a cross sectional area that changes in size and/or shape. Further, it is understood that the orienting baffle structure may be integrally formed with the nozzle or integrally formed with the nozzle and flow body.

Offset Sample Introduction Conduit



[0114] The core stream 189 may be directed along a flow path that is offset from the central longitudinal axis 2017 of the nozzle 137 by repositioning the sample introduction conduit 157 from its traditional position at the center of the nozzle 137 to an offset position. For example, Fig. 21 shows an exemplary offset sample introducing nozzle system 2151 having an offset sample introduction conduit 157. Except as noted, the nozzle system 2151 is substantially the same as the nozzle system 101 shown in Figs. 4 and 5. The significant difference is that the sample introduction conduit 157 has been moved away from the center of the nozzle 137 so that it is no longer aligned with the nozzle's longitudinal axis 2017. Thus, the core stream 189 is directed into the torsional zones 259, 261 of the orienting nozzle 137 along a flow path that is offset from the longitudinal axis 2017. Although the exemplary nozzle system 2151 shown in Fig. 21 uses the exemplary orienting nozzle 137 describe above, it is contemplated that offset sample introduction conduit 157 could be used with a different orienting nozzle or a nonorienting nozzle to orient particles in the core stream 189.

Nozzle Mounting and Adjustment



[0115] The flow body 133 and nozzle 137 are mounted in a selected orientation and position by means of a nozzle mount, generally designated 331. The mount 331 may comprise a plurality of stages (Fig 22), including first and second linear stages 333, 337 providing linear adjustment of the flow body 133 and nozzle 137 along X and Y axes 339, 341, respectively, and a third rotational stage 343 providing rotational adjustment about a Z axis 345 corresponding to the longitudinal axis 2017 of the flow body 133 and nozzle 137. These stages 333, 337, 343 may be conventional in design, suitable stages being commercially available, for example, from Newport Corporation of Irvine CA. In particular, the first linear motion stage 333 comprises a fixed first stage member (not shown) mounted on a frame 349, a movable first stage member 355 slidable on the fixed first stage member along the X axis 339, and an actuator 357, e.g., a micrometer, for precisely moving the movable first stage 355 member to a selected X-axis position. The second linear motion stage 337 comprises a fixed second stage member 359 mounted on the movable first stage member 355, a movable second stage member 361 slidable on the fixed second stage member 359 along the Y axis 341, and an actuator 363, e.g., a micrometer, for precisely moving the movable second stage member 361 to a selected Y-axis position. The rotational (third) stage 343 comprises a fixed third stage member 365 mounted on the movable second stage member 316, a movable third stage member 371 rotatably mounted on the fixed third stage member 365 for rotation about the Z-axis 345, and an actuator 373, e.g., a micrometer, for precisely rotating the movable third stage member 371 to a selected angular position relative to the Z-axis 345. The three-axis adjustment provided by these stages 333, 337 343 allows the nozzle 137 and the fluid stream 21 exiting the nozzle orifice 103 to be precisely positioned relative to the optics system 109. Rotation of the nozzle 137 about the Z-axis 345 is particularly helpful because it enables the stream 21 exiting the nozzle 137 to be rotated to bring the cells (e.g., sperm cells) oriented by the nozzle 137 into a position in which the light beam 25 from the optics system 109 will fall on the desired surfaces of the cells (e.g., the flat faces 207 of sperm heads 205), as illustrated schematically in Fig. 23. Other nozzle mounts may be suitable. For example, a 4-axis nozzle mounting system can also be used, providing linear adjustment along X, Y and Z axes and rotational adjustment along the Z axis. Further, it may be desirable to use one or more stages having an automated alignment feature, such as a servo or stepper motor controlled microtranslation stage (e.g., part number M-1 10.2DG from Polytech PI, Inc. of Auburn, Michigan).

[0116] In Fig. 36, for example, the nozzle 137 is oriented to direct a stream 21 containing cells to be analyzed in a generally upward direction. The angle 377 between the direction of the fluid stream 21 and horizontal is preferably in the range of 5 to 85 degrees, . This orientation is advantageous in that any air trapped in the nozzle system 101 is readily removed. Also, the velocity of the fluid stream 21 decreases gradually under the force of gravity prior to collection of the droplets 33. A more gradual deceleration of the droplets 33 is believed to be less stressful to the cells being analyzed which, in the case of sperm cells, can result in higher motility of the sorted sperm after collection. Of course, , the nozzle 101 may be positioned so that the fluid stream 21 has a substantially downward velocity when it exits the orifice 103 as is conventional for jet-in-air cytometers.

[0117] Optionally, components of the nozzle system 101 such as the flow body 133 and nozzle 137 are coated with a non-reflective, non-emissive material (e.g., a dull dark paint or epoxy which does not emit light when subjected to UV laser light) to reduce any reflected and/or emitted light off these elements 133,137 which might otherwise cause signal noise or have other adverse effects on the optics system 109.

Transducer and Droplet Formation



[0118] The transducer 105 for introducing energy into the fluid stream 21 comprises, in one example, a collar 379 containing a piezoelectric element (not shown) secured around the flow body 133 of the nozzle system 101 (Figs. 3 -5). The transducer is of conventional design, such as is available from Beckman Coulter, Inc. as part No. 6858368. The transducer has terminals 383 for connection to a suitable source of acoustical energy so that energy can be delivered to the fluid stream 21 at a frequency which will cause it to break into droplets 33 at the droplet break-off location 107 downstream from the nozzle 137 a distance d (Fig. 24). As will be understood by those skilled in flow cytometry, the characteristics of the droplet formation are governed by the following Equation 1:



[0119] Where V is the velocity of the stream 21; f is the frequency applied to the fluid stream 21 through the nozzle 137; and λ is the "wave length" or distance between the droplets 33. It is a known principle of flow cytometry that droplets 33 will form in a regular pattern with the distance between droplets 33 being 4.54 times the diameter of the stream 21. Since the diameter D of the stream 21 close to the nozzle 137 generally corresponds to the diameter of the nozzle orifice 103 at its downstream end, the frequency at which the stream 21 (and nozzle 137) must be vibrated to form the droplets 33 can be easily calculated using the following Equation 2:



[0120] The transducer 105 may be operated to generate in the range of 30,000 - 100,000 droplets 33 per second. For example, the transducer 105 may generate 50,000 - 55,000 droplets per second. Assuming the frequency is 55,000 cycles per second (55 kHz), and further assuming that the concentration of cells in the stream 21 is such that cells exit the nozzle 137 at a substantially matching rate of 55,000 cells per second, then there will be, on average, one cell per droplet 33. (In reality, some droplets 33 will contain no cells, some will contain one cell, and some will contain more than one cell.) Of course, any of various factors can be changed to vary this average, including a change in frequency (f), stream 21 (orifice 103) size (D) and stream 21 velocity (V). Ideally, these factors should be such as to reduce the amount of stress imparted to the cells during the course of the process, especially in the case of sperm cells where the preservation of motility is important.

Break-off Sensor



[0121] Referring to Fig. 2, a break-off sensor 389 may be employed to determine the location (e.g., break-off location 107) at which the stream 21 begins to form free droplets 33. The break-off location 107 will vary depending on several factors including stream 21 viscosity, surface tension of the fluid and the amplitude of vibration of the transducer 105. By monitoring the break- off location 107, the amplitude of the transducer 105 may be varied to maintain the break-off location 107 within a given range so that the time at which each droplet 33 breaks off can be more accurately predicted by the microprocessor 131. This allows the microprocessor 131 to accurately control the electrical charge of the droplet 33 which is accomplished by selectively controlling the charge of the stream 21. Since the charge of the droplet 33 will be the same as the charge of the stream 21 immediately before droplet 33 formation, the microprocessor 131 controls the sorting of the droplets 33 by selectively charging the stream 21, as noted below.

[0122] In general, a break-off sensor is for use with any continuous stream of fluid which is breaking into droplets at a break-off location. (In Fig. 2, the break-off sensor 389 is located downstream from the nozzle 137 and interrogation location 115. ) One exemplary break-off sensor 389 is shown schematically in Fig. 25. A light source 393 is positioned on one side of the stream 21 to illuminate the stream 21 within the given range at which the break-off location 107 will be maintained. A linear photoarray 395 positioned on the other side of the stream 21 is adapted to be oriented along an axis substantially parallel to the stream 21. As a result, the photoarray 395 detects light from the light source 393 which passes through the droplets 33 and provides output signals corresponding to the detected light.

[0123] The output signals are processed to determine the position of the break-off location 107. For example, the output signals may be digitized and provided to the processor 131 for processing. Alternatively, as shown in Fig. 25, the light source 393 may be an LED or other source which generates a near-infrared portion of the visible spectrum. The light passing between the droplets 33 is magnified by a lens 401 and directed toward an 8 by 1 linear array of photodiodes 395. Each photodiode generates a current that is proportional to the light intensity impinging thereon. This current is fed into 8 current to voltage op-amp circuits 405. The output voltage from the op-amps is AC coupled into 8 track/hold amplifiers 407. The track/hold signal 409 used by the amplifiers is taken from the transducer 105. The output from the track/hold amplifier is fed into the A/D converter 411 of a microprocessor unit (MPU) 391. The digital values computed by the MPU 391 will be provided to the system control microprocessor 131. A lookup table and/or algorithm may be used by the system control microprocessor 131 to convert between break-off location 107 drift and voltage adjustment to the transducer 105. Alternatively, the output from the MPU 391 may be an analog signal such as a DC voltage having an amplitude corresponding to a change in the amplitude of vibration of the transducer 105. The dc voltage can be applied to the high voltage amplifier input driving the droplet transducer 105 to vary the amplitude of vibration. Thus, such a processor 391 would constitute a control for receiving the output signal from the photoarray 395 and providing a location signal corresponding to a location of the break-off location 107. Such a processor 391 would also constitute a control for receiving the output signal indicative of the position of the break-off location 107 of the droplets 33 and varying operation of the transducer 105 as a function of the position of the location 107.

[0124] Alternatively, as is well known to those skilled in the art, a video camera and strobe light may be used to monitor and control the droplet break-off location. Thus, as shown in Figs. 26-27, a video camera system 412 and strobe 413 may be provided to monitor the break-off location 107. It is desirable to place the strobe 413 behind a mask 414A (e.g., a cover with a small slit; shaped opening 414B) to limit the amount of light produced by the strobe 413 that enters the optics system 109 (Fig. 27).

Epi-Illumination Optics System



[0125] The optics system 109 is adapted for focusing a beam of electromagnetic radiation 25 (e.g., a laser beam) on the fluid stream 21 as a beam spot, so that the cells to be analyzed pass through the spot. The beam 25 may be laser light in the visible or ultraviolet portion of the spectrum, for example, having a wavelength of about 350-700 nm, although other wavelengths may be used. The wavelength of the laser light may be selected so that it is capable of exciting a particular fluorochrome used to analyze particles. If the optics system 109 is used to analyze sperm cells stained with Hoechst 33342, for instance, the wavelength may be selected to be in the range of about 350-370 nm. The power output of the laser may vary between 50 and 300 mW. Sperm cells may be analyzed using a 200 mW laser, for example. Referring to Figs. 28-34, the system 109 is an epi-illumination system 415 comprising an instrument, generally designated 417, having a longitudinal optical axis 419. As used herein, the term "epi-illumination" means an optics system where at least some of the fluorescence emissions from cells passing through the beam spot are directed back through the optical instrument along the same axis as the focused beam 25, but in the opposite direction. This type of system is advantageous in that only one set of optics is required, including only one photodetector 117, unlike conventional systems which detect forward and side fluorescence and which use two or more photodetectors. However, it will be understood that while an epi-illumination system is preferred, many of the aspectsmay be applied regardless of the type of optics system used.

[0126] The epi-illumination instrument 417 may comprise a rectangular base 429 supporting a plurality of optical elements. These optical elements are described below, with specific examples of relevant dimensions, focal lengths, and part numbers. As will be understood by those skilled in the art, this information is exemplary only, and alternative optical elements can be used.

[0127] Referring to Figs. 28-34, the optical elements include a reflecting filter 431 which reflects a collimated beam 25 of light from a laser or arc lamp 435, for example, through a conditioning lens assembly 437 mounted in an opening 439 in a side wall 441 of a dichroic chamber 443 extending up from the base 429. In this particular example, the conditioning lens assembly 437 comprises a retaining ring 445, neutral density filter 447, cylindrical lens 449, lens holder 455 and jam nut 457. The cylindrical lens 449 introduces a one-dimensional divergence into the beam 225 and directs it toward optical elements (described below) which shape the beam to have a desired cross sectional shape 459, preferably generally elliptical. By way of example, the cylindrical lens 449 may be a plano-convex lens having a focal length of 16 mm. A beam expander (not shown) can optionally be installed in the instrument 417 to allow adjustments to be made to the shape of the elliptical beam spot 459.

[0128] The reflecting filter 431 is mounted by clips 461 on the angular face 465 of a filter holder 463 which has openings 467 in it to permit the beam 25 to reflect off the filter 431 toward the optics of the instrument 417. The holder 463 is fastened to a linear stage 469 movable along an X-axis 471 relative to an outrigger 473 secured to the base 429 and dichroic chamber 443, the stage 469 being movable by suitable means 475 (e.g., a micrometer) to precisely locate the holder 463 and reflecting filter 431 to reflect the beam 25 into the instrument 417 at the proper location. A dichroic filter 477 is held by clips 479 on a frame 485 mounted in the dichroic chamber 443 and functions to reflect the shaped beam 25 in a forward direction 487 along an axis 489 which, in this particular example, corresponds to the longitudinal optical axis 419 of the instrument. The beam 25 passes through a focusing lens assembly 491 which focuses the beam 25 on the fluid stream 21 as a beam spot having the aforementioned generally elliptical shape 459 (Fig. 6) with the major axis of the ellipse extending generally perpendicular to the direction of flow 227 of the stream 21. As each cell passes through the beam spot 459, the fluorescing dye (or other reporting agent) in the cell is activated to emit fluorescent light 31 (Fig. 23). In the case of sperm cells stained with a DNA selective fluorescing dye, X cells have more DNA than Y cells, include more fluorescing dye, and emit a stronger signal than Y cells (e.g., 3.8%), which provides a basis for discriminating and sorting cells, as will be described. The focusing lens assembly 491 includes, in one example, a microscope adapter 501 mounted in an opening 503 in a front wall 505 of the dichroic chamber 443, a focusing barrel 507, a pair of lens mount barrels 509, and the lens 511 itself, which may be a 12.5 mm diameter, plano-convex lens with a focal length of 16 mm, available from Oriel Corporation as part number 41209, and is anti-reflective coated for light having a wavelength in the range of 340-550 nm. The lens 511 may be made of fused silica. Other focusing lenses may also be suitable, such as an infinitycorrected fluorescence microscope objective. The focusing lens assembly 491 has a conventional telescoping focus adjustment 515 to focus the elliptically-shaped beam spot 459 on the core 189 of the stream 21.

[0129] The outgoing fluorescent light 31 emitted by the cells as they pass through the beam spot 459 is of a different (longer, due to the Stoke's shift principle) wavelength than the incoming laser light 25. Some of the fluorescence emissions 31 are transmitted in a rearward direction 513 along the incoming beam axis back through the focusing lens 511 which collects and collimates the fluorescence emission 31. The collimated fluorescence emissions 517 pass in a rearward direction from the lens 511 to the dichroic filter 477, which transmits the fluorescence emission 517. By way of example, the dichroic filter 477 may be a filter available from Omega Optical as part number XF2001, 400DCLP.

[0130] The optics system 415 includes a filtering system 519 positioned rearward of the dichroic filter 477 along the optical axis 419 of the instrument 417. In one example, the filtering system 519 includes an emission filter 521 in a holder 523 mounted in an opening 525 in a back wall 527 of the dichroic chamber 443. The emission filter 521 attenuates any laser light scatter or other undesired electromagnetic radiation that is transmitted through the dichroic filter 477. By way of example and not limitation, the emission filter 521 can be a thin film, long-pass filter adapted to transmit more than 90% of light having a wavelength greater than 408 nm, as is available from Omega Optical as part number XF3097. An alignment pellicle assembly 529 is spaced rearwardly along the optical axis 419 from the emission filter. This assembly includes a slider 531 movable on a rail 533 extending longitudinally of the base 429 parallel to the longitudinal optical axis 419 of the instrument 417, a filter holder 535 secured to the slider 531, a pellicle filter element 539, and clips 541 for securing the pellicle filter element 539 to the filter holder 535 at an angle 543 relative to the optical axis 419 of the instrument 417. The pellicle filter element 539 has the same thickness as the dichroic filter 477 and functions to translate the collimated fluorescence emission 517 back onto the optical axis 419 of the instrument 417. Fasteners 545 extending up through parallel slots 547 in the base 429 on opposite sides of the rail 533 secure the slider 531 to the base 429 in the desired position along the optical axis 419. Spaced to the rear of the alignment pellicle assembly 529 is an aspheric lens 549 held by a holder 551 mounted in a frame 553 which is also slidable on the rail 533 and secured in selected position by suitable fasteners 557. The aspheric lens 549 focuses the collimated fluorescence emission 517 onto a spatial filter, generally designated 559, which filters out reflection or emission from sources other than the cells to be analyzed. The aspheric lens 549 may be, for example, a 12.5 mm diameter aspheric lens having a focal length of 15 mm, as is available from Oriel Corporation. The lens 549 is preferably anti-reflective coated for visible emission wavelengths but made of a material (e.g., flint glass) which further attenuates transmission of laser light scatter.

[0131] As shown in Fig. 34, the spatial filter 559 may comprise, a pair of aperture plates 561 releasably held by a frame 563 mounted on the base 429 of the instrument 417. Each of the plates 561 has a slit 567, 571 therein, one slit 567 preferably being generally vertical and the other 571 preferably generally horizontal, the arrangement being such that the slits 567, 571 intersect to form an aperture 573. The aperture 573may be generally rectangular in shape andmay have a vertical dimension 575 of 100 microns and a horizontal dimension 577 of 500 microns. The size and shape of the aperture 573 may vary (or even be adjusted by changing aperture plates), so long as it functions to remove reflections and light from any source other than the collection volume 579. The frame 563 holding the aperture plates 561 preferably has two parts, namely, a plate holder 583 slidable on the rail 533 of the base 429 and secured in selected position by fasteners 587, and a backing member 589 for securing the aperture plates 461 in position on the plate holder 583.

[0132] The smaller (vertical) dimension 575 of the aperture 573 in the spatial filter 559 may be sized (or adjusted) to enable use of a "slit scanning" technique to evaluate the cell. This technique is described in more detail in the "Focused Beam Spot "section of this specification.

[0133] Another example of an epi-illumination optics system, generally designated 450, is shown in Fig. 35. This epi-illumination system is substantially the same as the system shown in Figs. 28-34, except as noted. One significant difference is that the dichroic filter 477 has been replaced with a different dichroic filter 451 that transmits (rather than reflects) the illumination beam 25 and reflects (rather than transmits) the fluorescent emissions 31. Also, because the fluorescence emissions 31 are reflected by the dichroic filter 451 rather than transmitted, there is no need for an alignment pellicle 539 in this example of an epi-illumination optics system 450. Thus, the epi-illumination system 450 is just one example of how the optics system can be reconfigured if desired.

[0134] Further, the cylindrical lens 449 is mounted on an adjustable mounting assembly 449A. The mounting assembly 449A allows two-axis translational movement of the cylindrical lens 449 in a plane perpendicular to the illumination beam 25. Releasable fasteners (e.g., screws (not shown)) extend through slot-shaped holes 449B (only one of which is visible on Fig. 35). Release of the fasteners allows translational movement of the lens 449 in a first direction perpendicular to the beam 25. Similar fasteners (not shown) extend through slot-shaped holes 449C, allowing translational movement of the lens 449 in a second direction perpendicular to the first direction. This allows minor adjustment of the relative positions of the cylindrical lens 449 and beam 25 so that the intersection of the beam 25 and lens 449 can be moved across the surface of the lens 449, thereby causing slight changes to the focusing provided by the cylindrical lens 449. Once the lens 449 is in the desired position, the fasteners can be tightened to hold it there.

Photodetector



[0135] The emitted fluorescence passing though the spatial filter 559 falls upon a photodetector 117 fastened to a mounting plate 591 slidable on the rail 533 of the base 429 at the rear of the epi-illumination instrument 417 and securable in fixed position by fasteners 595 (Fig. 32). The photodetector 117 detects the fluorescent emissions 31 and converts them into electrical signals which can be processed to analyze the desired characteristics of the cells, as will be described in more detail later. The photodetector 117 may be a conventional device, such as a photodetector available from Hammamtsu. The photodetector 117 preferably includes a preamplifier and PMT gain which is optimized for emission intensity produced by the epi-illumination system for the particular stained cells being analyzed.

[0136] In general, the PMT gain is optimized when between about 200 and 2000 volts are applied to the vacuum tube. In the case of detecting fluorescent emissions from Hoechst 33342, for instance, the PMT gain is optimized when between about 400-800 volts are applied to the vacuum tube. One particularly desirable photodetector includes a PMT having a spectral range of 185-830 nm (530 nm peak), a 0.01 mA maximum average anode current, a cathode radiant sensitivity of 70 mA/W typical, a cathode luminous sensitivity of 140, µA/lm, anode luminous sensitivity of 300 A/lm, max anode dark current of 1 nA (0.1 nA typical), and a 1.4 nanosecond risetime. The PMT is DC coupled amplifier demonstrating a flat gain to > 37 MHz, having a 1 V peak output into a 50 Ω load and a recovery time of less than 400 nanoseconds. It is also desirable for the amplifier to allow high voltage adjustment for compensation of PMT efficiency variations without decreasing the signal-to-noise ratio to less than 800 dB.

Angle of Beam Incidence



[0137] Fig. 36 schematically illustrates one desirable orientation of the intersection of the light beam and the fluid stream. Several points are of note. As shown, the light beam 25 is focused on the stream 21 at a location 115 that is only a short distance 605 from the exit orifice 103 of the nozzle 137, preferably less than 1.0 mm, or even inside the nozzle 137, so that the cells pass through the spot 459 while they are still substantially in desired orientation, as previously described. This is particularly important for cells which are mobile in the fluid stream 21, including sperm cells.

[0138] Another point of note is that the beam 25 of this example may be directed toward the fluid stream 21 along a beam axis 609 which intersects the fluid stream 21 at an angle of incidence A which is skewed (off 90 degrees) relative to a longitudinal axis of the fluid stream 21, as viewed from a side of the stream 21 (see Fig. 36). When sorting certain particles, it has been found that better discrimination of the different types of particles may be obtained by illuminating the stream 21 at an angle of incidence other than 0°. Sperm nuclei, for instance, are desirably illuminated at an angle of incidence A that is in the range of 5 to 45 degrees. Other particles (e.g., live sperm cells) are easier to interrogate when the light beam 25 is generally perpendicular to the fluid stream 21 (i.e., when angle A is about 0°). Thus, it is contemplated that angle A may be any angle.

[0139] The proper selection of angle A results in improved signal to noise discrimination in certain particles and thus more accurate discrimination based on different characteristics of those particles (e.g., sperm nuclei with X and Y chromosomes sperm cells). This improvement may be due to a number of factors, including reduced laser light scatter entering the focusing lens 511. Because the focused beam spot 459 is preferably wider than the stream 21, a diffraction pattern is created at the intersection 115 of the beam 25 and the stream 21. When angle A is greater than about 12 degrees, the reflected diffraction pattern does not fall on the lens 511. Another factor may be that the skewed angle A allows the beam 25 to be focused very close to the nozzle orifice 103, so that the nozzle body 139 does not interfere with the lens 511. Relatedly, the cells are more uniformly aligned closer to the nozzle 137, so that focusing the beam spot 459 closer to the nozzle 137 results in an improved signal. Further, the more "head on" profile of the cell presented to the lens 511 (beam 25) at the skewed angle A reduces the variation of total fluorescence intensity caused by any misalignment of the cells. In this regard, in the case of sperm cells it is preferable that the beam 25 fall on one of the wide faces 207 of each sperm cell 201, as discussed above, and that the nozzle 101 and optics system 109 be positioned to achieve this result.

[0140] While a skewed angle of incidence A is believed to be beneficial in sorting some particles, it is contemplated that the angle of intersection between the beam axis and the stream may be 90 degrees or any skewed angle. It is also expected that the optimal angle of incidence may vary widely depending on the properties of the particular particles being analyzed.

Focused Beam Spot



[0141] Referring to Fig. 6, the focused beam spot of one example is shown as having a generally elliptical (oval) shape 459 with a length L1 along a major axis extending generally at right angles to the direction of fluid stream flow 227 and a width W1 along a minor axis extending generally parallel to the direction of fluid stream flow 227. The width W1 may be less than the length of the head of the sperm cell 219, and even more preferably less than the length of the region 225 containing the chromatic DNA mass of the cell, which in the case of a bovine sperm cell 201 has a length of less than about 1µm. For a stream 21 having sheath stream 191 that is about 60µm in diameter and a core stream 189 containing bovine sperm cells 201, an exemplary length L1 is about 80µm and an exemplary width W1 is about 1.5µm. By focusing the beam spot 459 to a width W1 which is less than the length of the head 205 of the sperm cell 201, or any other cell or particle being analyzed, and even more preferably less than the diameter of the DNA region 225 of the head 205 of the sperm cell 201, greater signal resolution is achieved, as will be understood by those familiar with "slit scanning" techniques. This is a technique by which a beam 25 is narrowed to have a width less than the length of a cell (i.e., the dimension of the cell in the direction of stream flow) so that as the cell moves through the narrow beam, photon emissions 31 from the cell are measured over the length of the cell, as will be discussed later. In this way, information can be obtained about variations in structure, including DNA material, along the length of the cell. The slit-scanning technique is also helpful in identifying "coincident" cells, that is, cells which are overlapping or very close together.

[0142] As mentioned previously, slit scanning can also be carried out by sizing the aperture 573 of the spatial filter 559 to have a vertical dimension 575 such that only a portion of the light emitted from a cell, corresponding to a fraction of the cell length in the direction of stream flow, passes through the aperture to the photodetector 117. Further, signal resolution can be optimized by adjusting the width of the beam and/or the size of the aperture of the spatial filter to work together to provide a beam spot that is suitably shaped for slit scanning.

[0143] One way to adjust the shape of the beam spot 459 is by changing to a different cylindrical lens and/or by making an adjustment to a beam expander in the optics system 109. Further any method of shaping the beam 25 to form an elliptically shaped beam spot 459 may be contemplated. Beam spots of other shapes and sizes may also be used and may be contemplated.

Sorting System



[0144] Fig. 2 illustrates an example of the sorting system 119. The sorting system 119 comprises an electrostatic charging device 627 for charging and/or not charging the droplets 33 depending on the classification of the particles contained in the droplets 33 (e.g., the X/Y chromosome content of sperm cells), and a pair of electrostatic charged deflector plates 629 for sorting the droplets 33 into different groups 123, 125, according to their charge. It is desirable to coat the deflector plates 629 with a dull, low-emissive coating (e.g., epoxy or paint) to limit light reflected or emitted by the deflector plates 629. The deflector plates 629 may be charged by any suitable power supply 635. It is generally desirable for the electrical potential between the two fully charged deflector plates 629 to be in the range of 2000-4000 volts. However, the electrical potential between the deflector plates 629 may be anywhere between about 1000 and 6000 volts.

[0145] The charging device 627 comprises a charging element 631 having an opening 633 therein through which the stream 21 passes at a location near the droplet break-off location 107 (e.g., within five droplet lengths or closer). It is desirable to mount the charging element 631 with a mechanism that facilitates adjustment of the position of the charging element 631 with respect to the droplet break-off location 107. As shown in Figs. 26 and 27, for example, the charging element 631 and deflector plates 629 may be attached to an adjustable mounting assembly 5001 that allows three-axis translation and tilt adjustment of the charging element 631 and deflector plates 629 with respect to the nozzle system 101. For translation along an axis 5011 parallel to the stream 21, the mounting assembly 5001 includes a board 5003 fastened to a backing 5005 by releasable fasteners 5007 passing through slots 5009 in the board 5003, the slots 5009 being oriented generally parallel to axis 5011. For translation in an axis 5013 perpendicular to the stream 21, a second adjustment board 5015 is fastened to the first board 5003 by releasable fasteners 5017 passing through slots 5019 in the second adjustment board 5015, the slots 5019 being oriented generally parallel to axis 5013. The charging element 631 and deflector plates 629 are secured to the second adjustment board 5015. Thus, by releasing the fasteners 5007 and/or 5017, one can adjust the position of the charging element 631 and deflector plates relative to the nozzle system 101 in a plane parallel to the fluid stream 21 and then tighten the fasteners 5007 and/or 5017 to secure the mounting assembly 5001.

[0146] For translation along a third axis perpendicular to the first two axes 5011, 5013, the backing 5005 is fastened to a fixed support 5021 by adjustable fasteners 5023 (e.g.., threaded bolts screwed into tapped holes in the fixed support 5021). Each adjustable fastener 5023 may pass through a spring 5025 positioned between the backing 5005 and the fixed support 5021. The amount of compression of any spring 5025 can be adjusted by tightening or loosening the respective fastener 5023. Adjusting the compression of all springs 5025 in the same amount results in translation along the third axis. The mounting assembly 5001 can be tilted in virtually any direction by changing the relative compression of one or more of the springs 5025 with respect to one or more other springs 5025.

[0147] The relative positions of the charging element 631 and deflector plates 629 remain fixed with respect to one another because they are all fastened to the same adjustment board 5015. This prevents adjustment of the mounting assembly 5001 from affecting alignment of the changing element 631 with respect to the deflector plates 629.

[0148] The charging element 631 is connected to a suitable electrical circuit (e.g., a 90 volt selectively charging circuit) under the control of the processor 131 and coupled to a power supply for applying an electrical charge to the charging element 631. The circuit is used to charge or not charge the stream 21 immediately prior to the formation of a droplet 33 at the break-off location 107 depending on whether the droplet 33 contains a particle having the desired characteristics (e.g., at least one live X-chromosome sperm cell). The charging element 631 is positioned electrostatically near the stream 21 or near the droplets 33 formed from the stream 21 for providing an electrical reference with respect to the electrostatic polarity of the stream 21. The droplets 33 carry the same charge as the stream 21 at the instant the droplet 33 breaks from the stream 21. The charged or uncharged droplets 33 then pass between the deflector plates 629 and are sorted by charge into collection vessels 2207 of the collection system 2201. While sorting produces two groups or populations of droplets 123, 125 in Fig. 2, the particles may be separated into any number of populations from 1 to N sorted by placing different charges on the droplets 33 in respective groups, any by supplying the appropriate number of collection vessels, each being positioned to collect a different population of droplets.

Automated Drop Delay Calibration



[0149] In the sorting system 119 described above, the processor 131 must estimate the time it takes for a particle to move from the interrogation location 115 to the droplet break-off location 107 so that the charge (or lack of charge) to be applied to the droplet 33 containing that particle is applied when the particle is in the last attached droplet 33 at the break-off location 107. If the delay setting used by the processor 131 is wrong, the droplets 33 will not be sorted according to their contents. Similarly, if the application of electrical charges to the droplets 33 is even slightly out of phase with droplet 33 formation this can degrade sorting because none of the droplets 33 will be fully charged and droplets 33 that are supposed to have neutral charge will carry a small positive or negative electrical charge. This will alter the paths of the droplets 33 through the electric field between the deflection plates 629.

[0150] The best way to verify that the processor 131 is using the appropriate delay setting or to adjust the drop delay setting (i.e., calibrate the system's 9 drop delay setting), is to sort a number of droplets 33 and examine the results. By incrementally varying the delay setting and monitoring the results, one can select the optimal delay setting. Traditionally, this sort calibration is performed manually. Recently, automated calibration systems have been designed to sample or examine the contents of the droplets in the sorted droplet streams and automatically adjust the delay setting without human intervention. For example U. S. Patent Nos. 6,372, 506 (Norton) and 5,643, 796 (van den Engh), both disclose automated sort calibration systems. The purported advantages of these systems are that they are less labor intensive and are capable of verifying the delay setting throughout the sorting process rather than just during initial set up. The drawbacks are that they are cumbersome and take up valuable space unnecessarily.

(i) Epi-illumination Sensors



[0151] Referring to Fig. 37, an automated continuous calibration system 4201 for a fluorescence activated droplet sorting cytometry system comprises one or more epi-illumination sensors 4203 positioned to sense the contents of droplets 33 to verify the delay setting for droplet charging. Referring to Fig. 38, each epi-illumination sensor includes a light source (not shown), a fiber optic cable 4205, a dichroic filter 4207, a lens system 4209, a photodetector 4213, and a control system. The processor 131 may serve as the control system, but other processors or controls could be used instead.

[0152] The light source may be a low-power solid state laser dedicated solely to the automated calibration system 4201. Alternatively, a beam splitter (not shown) may be used to divert a portion (e.g., about 5%) of the energy in the beam 25 used for interrogation of particles in the fluid stream 21 to one or more epi-illumination sensors 4203. Similarly, the fiber optic cable 4209 can be positioned in a beam stop 4215 (Fig. 26) to gather light from beam 25 after it passes through the interrogation location 115. The light from the light source must include light having a wavelength capable of exciting fluorescent molecules in the particles being sorted, thereby causing fluorescence emissions 4211 from the particles. If the particles are stained with Hoechst 33342, for instance, the light source can provide light having a wavelength of about 350 nm, about 407 nm or any other wavelength capable of exciting the Hoechst 33342 molecules.

[0153] The fiber optic cable 4205 extends from the light source to a location downstream of the interrogation location 115. In one example, the fiber optic cable 4205 leads to a location adjacent the trajectory of one of the droplet streams as it moves through the electric field between the deflector plates 629. The dichroic filter 4207 is positioned in front of the end of the fiber optic cable 4205. The dichroic filter 4207 transmits light having the spectral characteristics of the light conducted by fiber optic cable 4205, but reflects light having the spectral characteristics of the fluorescence emissions 4211. Thus, the dichroic filter 4207 may have the same specifications as the dichroic filter 477 described above in connection with the epi- illumination optics instrument 417. The focal length of the lens system 4209 is selected based on the expected distance of the sensor 4203 from the droplets 33 so that the illumination/detection volume of each sensor 4203 is about equal to the volume of the droplets 33.

[0154] Referring to the example shown in Fig. 37, an epi-illumination sensor 4203 is positioned adjacent the trajectory of each of the three sorted droplet streams 4225, 4227, 4229 to sense the contents of droplets 33 in a respective stream. The cytometer system 9 includes an electrically insulated support 4221 for mounting the two deflection plates 629. The support has three holes 4223, one adjacent the trajectory of each sorted droplet stream 4225, 4227, 4229. An epi-illumination sensor 4203 is positioned at each hole 4223 to observe droplets 33 in one of the droplet streams 4225, 4227, 4229 through the respective hole 4223. This compact configuration takes up relatively little space and keeps components of the calibration system 4201 out of the way, providing better access to other parts of the cytometer 9.

[0155] If a droplet containing a fluorescent particle passes through the illumination/detection volume of the sensor 4203, this will result in a flash of fluorescence emissions 4211, some of which will be collected by the lens system 4209 and reflected off from the dichroic filter 4207 to the photodetector 4213. Signals from the photodetector 4213 are provided to the processor 131. Based on the signals received from the photodetectors 4213, the processor 131 can determine the contents of the droplets 33 in each of the sorted droplet streams 4225, 4227, 4229.

[0156] If a sensor 4203 fails to detect a flash of fluorescence emission 4211 when the processor 131 expects a droplet 33 containing a fluorescent particle to pass by that sensor 4203, the processor 131 can use that information to adjust the delay setting or adjust the location of the droplet break-off location 107. Likewise, the processor 131 can make an adjustment if a sensor 4203 detects a fluorescent emission 4211 when the processor 131 does not expect a droplet 33 containing a particle to be passing by the sensor 4203. Furthermore, the processor can compare the relative frequency of fluorescent emissions 4211 from the sorted streams 4225, 4227, 4229 to see if the frequency of detected fluorescent emissions 4211 matches the expected frequency. The processor 131 can also adjust the amplitude of the charge applied to the charging element 631 to increase or decrease the amount by which a sorted stream 4225, 4229 is deflected to maximize the intensity of the detected fluorescence emissions 4211. This will maintain the alignment of the trajectory of the deflected droplet streams 4225, 4229 so the droplets pass directly through the collection volume of the epi-illumination sensor. Because the sensors 4203 are positioned to observe the streams 4225, 4227, 4229 as they move through the electrical field between the deflector plates 629, the calibration system has a shorter response time than it would if it observed the streams 4225, 4227, 4229 in the freefall area downstream of the deflection plates.

(ii) Empty Droplet Test Stream



[0157] One sensitive indication of the quality of the calibration can be arranged by creating and monitoring a calibration test stream that contains substantially only empty droplets 33. Referring to the sort calibration system 4201 shown Fig. 37, droplets 33 containing desired particles are sorted into stream 4225 and droplets 33 containing any other particles and most of the empty droplets 33 are sorted into stream 4229 (i.e., the waste stream). The test stream 4227 is created by applying a neutral charge to at least a fraction (e.g., 1 out of every 10) of the empty droplets 33. Many droplets 33 that are considered "empty" for traditional sorting purposes are actually droplets 33 for which there is a low probability that the droplet 33 contains a particle, based on the arrival time of particles at the interrogation location 115 and estimated droplet formation boundaries in the fluid stream 21. These "empty" droplets should not be sorted into the test stream 4227 because this would inevitably result in detection of some particles in the test stream 4227.

[0158] Instead, for the test stream 4227 the processor 131 should select only droplets 33 that the processor 131 believes have substantially zero probability of containing a particle in order to create a substantially particle-free test stream 4227. The probability that any randomly selected droplet 33 contains a cell is known and is approximately the average cell analysis rate divided by the droplet generation rate. This means that by monitoring the rate of mis-sorts in the test stream 4227 it is possible to estimate fractional adjustment of the phase relationship of droplet charging needed to match the phase of droplet 33 formation. For example the processor 131 may select droplets that it estimates have about 15% or lower probability of containing a particle, The probabilistic cutoff for substantially zero probability may be selected based on sort-speed, tolerance for impurity, or other sort parameters, with the cutoff including higher probabilities that a droplet will include a particle for high-speed sorting or when there is more tolerance for impurity.

[0159] Failure of the processor 131 to create a substantially particle-free test stream 4227 (i.e., a test stream 4227 in which the ratio of droplets 33 containing particles to the total number of droplets 33 agrees with the probabilistic cutoff used to select droplets 33 for the test stream 4227), as indicated by detection of more than a threshold number of droplets 33 containing particles in the test stream 4227, is a definitive indication of sub-optimal sorting and prompts the processor 131 to adjust the drop delay setting. The threshold level is determined in relation to the probabilistic cutoff used to select droplets 33 for the test stream 4227 and the total number of droplets 33 selected for the test stream 4227. Ideally, some droplets 33 can be selected for the test stream 4227 even though one or more particles in the fluid stream 21 are relatively close to an estimated drop formation boundary for the respective droplet 33 to make the system 4201 more sensitive to slightly sub-optimal drop delay settings.

[0160] Of course, the sort calibration system could apply a non-neutral charge to and deflect droplets selected for the test stream. The relative order of the streams 4225, 4227, 4229 could also be rearranged, although interposing the test stream 4227 between the waste stream 4225 and the stream of desired particles 4229 reduces the risk of crossover contamination of the sorted sample by the waste stream. Further, if the particles do not emit fluorescent light, different sensors can be used to detect any scattered light caused by particles in the test stream.

(iii) Impact of Sort Calibration System



[0161] The automated calibration system 4201 may be operable to automatically determine and set the phase relationship between droplet formation and droplet charging to within about 5% of the optimal phase (i.e., within +/- about 18 degrees. The system 4201 may operable to automatically determine and set the phase relationship to within about 1% of the optimal phase (i.e., within +/- about 3.6 degrees)). Furthermore, the calibration system 4201 may be operable to continuously monitor a high- speed droplet sorting system and automatically maintain the phase relationship within about 10% of the optimal phase (i.e., within +/- about 36 degrees). In addition, the system 4201 may be operable to continuously monitor a high-speed droplet sorting system and automatically maintain the phase relationship without about 3% of the optimal phase (i.e., within +/- 10.8 degrees).

Sort System Fault Correction



[0162] From time to time, a droplet 33 will stray from its normal trajectory and hit the charging element 631 or the deflector plates 629. If one or more droplets 33 hit the charging element 631, the charging element 631 may not be able to charge droplets 33 properly. Further, the normal droplet 33 trajectory through the charging element 631 can become obstructed causing even more droplets 33 to accumulate on the charging element 631. Also, if stray droplets 33 strike a deflector plate 629, they can distort or otherwise disrupt electrical field lines between the deflector plates 629, thereby changing the trajectory of the sorted droplet steams 123, 125.

[0163] Thus, it is desirable to have a debris removal system to remove debris from the charging element 631 and/or the deflector plates 629. In one example, shown Figs. 26 and 27, the system 9 includes a debris removal system 5047 for the charging element 631 and a debris removal system 5049 for the deflector plates 629.

[0164] Referring to Fig. 27, the charging element 631 is held in position by a support 5051 secured to board 5015 of the adjustable mounting assembly 5001. A vacuum passage 5053 (shown in phantom) extends through the support 5051 to an opening 5057 adjacent the charging element 631. The vacuum passage 5053 is connected to a suitable vacuum source (not shown) by a vacuum line 5055 attached to a fitting 5058 on the support 5051. Suitable controls are provided for selectively applying a vacuum in the passage 5053 to vacuum any undesired material (e.g., stray droplets 33) off the charging element 631 and restore proper function of the charging element 631.

[0165] Relatedly, as shown in Fig. 27, a manifold 5061 fastened to the mounting assembly 5001 has a network of air passages 5063 therein (shown in phantom) connected via an air line 5059 and fitting 5065 to a source of compressed air or other gas (not shown). The passages 5063 have openings 5064 positioned along a side 5066 of each deflector plate 629 and the portions 5067 of the passages 5063 leading to the openings 5065 are oriented so compressed air blown through the manifold 5061 will clear any stray droplets 33 or other debris off the deflector plates 629. Any material blown off the deflector plates 629 will hit a cover panel (not shown) and drain into a suitable waste collection device (not shown).

[0166] If the processor or other sensor determines that stray droplets 33 have hit the charging element 631 or deflector plates 629, as indicated by the sort calibration system described above for example, the processor can automatically initiate a fault correction procedure or mode, which can include applying a vacuum to passage 5053 to vacuum material from the charging element 631 and/or sending compressed gas through passages 5067 to blow material off the deflector plates 629.

Protection of Sorted Sample During Fault Mode



[0167] The system 9 may also includes a contamination prevention mechanism 4041 (Fig. 26), which can be activated by the processor 131 to limit or prevent contamination of the sorted sample any time the sorting system is in the fault correction mode. The contamination prevention mechanism includes a pneumatic actuator 4043 operable to selectively move a swing arm 4045 between a shielding position (shown Fig. 26) and a non-shielding position (not shown). In the shielding position, the end 4047 of the swing arm 4045 covers the opening of the collection vessel 4033, thereby preventing collection of droplets 33 by the collection vessel 4033. In the non-shielding position, the collection vessel 4033 is uncovered. Normally, the swing arm 4045 is in the non-shielding position, but the processor 131 causes the actuator 4043 to move the swing arm 4045 into the shielding position any time the processor 131 determines that there is a risk of contamination (e.g., the nozzle system 101 becomes clogged, the droplet break-off location 107 becomes unstable, or stray droplets 33 have hit the charging element 631 or deflector plates 629). The end 4047 of the swing arm 4045 is trough-shaped to drain any fluid collected by the swing arm 4045 into the waste container 4035.

Fluid Delivery System



[0168] The system 1 described above is capable of effectively producing quantities of particles (e.g., X-sperm cells) sorted by selected characteristics. The rate of production can be increased or decreased by varying the rates at which the fluid delivery system 15 (Fig. 2) delivers carrier fluid 17 and sheath fluid 19 to the nozzle 137. In one example, the fluid delivery system includes a syringe pump 645, one example of such a pump being MICROLAB® Model PSD/3 available from Hamilton Company. The pump 645 is operable to deliver carrier fluid 17 to the nozzle 137 at a rate of about 20 µl/min. In general, the pump 645 should be operable to deliver sample fluid 17 to the nozzle 137 at a rate in the range of 10-50 µl/min. The pump 645 is connected by a flow line 647 to the supply 3 of carrier fluid 17, which may be a suitable vessel 649 containing a volume of material to be analyzed and sorted. Where the temperature of the particles being analyzed is a factor, as in the case of sperm cells, for example, the temperature of the vessel 649 may be controlled by a suitable temperature control system, such as heating/cooling bath (not shown). The syringe pump 645 is movable through an intake stroke to aspirate carrier fluid from the supply vessel and through a discharge stroke to dispense carrier fluid 17 through a supply line 651 to the injection needle 157 of the nozzle system 101. The pump 645 is preferably driven by a variable speed motor (not shown) under the control of the processor 131. By way of example, the pump 645 may be driven by a stepper motor which operates at selectively variable rates to pump carrier fluid 17 to the needle 159 at rates necessary to obtain the desired throughput. Other types of fluid delivery devices can be used instead of a syringe pump. To provide just one example, the vessel 649 can be pressurized by a pressurized gas source. Furthermore, it is desirable to keep the lines 647, 651 as short as is practically possible because the line environment is not conducive to the health of sensitive cells (e.g., sperm cells) that may be in the carrier fluid 17.

[0169] The supply 7 of sheath fluid 19 comprises a second vessel 661, e.g., a tank in Fig. 2, holding an appropriate volume of sheath fluid 19 connected to the radial bore 173 in the flow body 133 of the nozzle system 101 by a supply line 667 having a control valve 669 therein. In the example of Fig. 1, the sheath fluid vessel 661 is pressurized by a gas pressure system 671 comprising a source 675 of pressurized gas (e.g., air or other gas, such as nitrogen) communicating with the tank 661 via an air line 679 having a regulator 681 in it for controlling the pressure supplied to the tank 661. A two-way valve 683 in the air line 679 is movable between a first position establishing communication between the tank 661 and the gas source 675 and a second position venting the tank 661. The gas pressure regulator 681 is a conventional regulator preferably under the control of the processor 131. By controlling the tank 661 pressure, the pressure at which sheath fluid 19 is delivered to the flow body 133 may also be controlled. This pressure may range from 1.1 to 6.9 bar (16 to 100 psi).

[0170] In Fig. 26 the fluid delivery system 15, includes a sheath fluid tank (not shown) and a sample station 4051. The sample station includes a two-part pressure container 4053 adapted to hold a sample tube 4055. The bottom section 4057 of the pressure container is moveable up and down relative to the upper section 4059 of the pressure container 4053 between an open position (shown Fig. 26), in which the sample tube 4055 may be loaded or unloaded, and a closed position (not shown) in which the two parts 4057, 4059 of the pressure container 4053 come together to form a seal to contain pressurized gas used to pump carrier fluid 17 from the sample tube 4055 to the nozzle system 101.

[0171] When the pressure container is open a spring-biased swing arm 4071 moves to a position beneath the line 651 that delivers carrier fluid 17 to the nozzle system 101 (See also Fig. 119 #4071'). The swing arm 4071 is trough-shaped and adapted to collect fluid backflushed through the line 651 and to drain the backflushed fluid to the waste container through port 4073. As the pressure container 4053 moves from its open position to its closed position, a cam plate 4075 attached to the bottom section 4057 of the pressure container 4053 moves the swing arm 4071 against its spring bias to clear the area between the two sections 4057, 4059 and allow the pressure container 4053 to close.

Control



[0172] Referring again to Fig. 2, the microprocessor 131 (or other digital or analog control and/or processor, or combinations thereof) controls the operation of the system 1. As noted below with regard to Fig. 39, the microprocessor may be implemented as a system control processor and four processors for handling signal processing. Alternatively, some or all functions may be integrated into one or more processors. For example, the system control microprocessor (see Fig. 36) may be implemented by using one of the four signal processing processors. In addition, as noted below, the signal processing may be implemented by an analog circuit (e.g., an analog cell analyzer as shown in Fig. 39) or a combination of analog and digital circuitry.

[0173] The microprocessor 131 provides output signals to control the fluid delivery system 15 (noted below) in response to input signals received from the epi-illumination system 415, provides output signals to control the transducers 105 in response to input signals received from the break-off sensors 389, and provides output signals to control the sorting system 119 (noted below) in response to input signals received from the epi-illumination system 415. The microprocessor 131 may provide output signals to other parts of the cytometry system 9 as noted elsewhere herein. Further, the microprocessor 131 may be adapted to process information and provide output signals in real time. Broadly speaking, the term "real time" refers to operations in which the operation of the processor 131 matches the human perception of time or those in which the rate of the operation of the processor 131 matches the rate of relevant physical or external processes. In one context, the term "real time" can indicate that the system reacts to events before the events become obsolete.

[0174] In general, electrical signals from the epi-illumination system 415 are converted to digital information by an A/D converter 689 which supplies the corresponding digital information to the microprocessor 131. In response to the information, the microprocessor 131 controls a sorting system 119 and a fluid delivery system 15, both described above.

[0175] The electrical signals output from the photodetector 117 of the epi-illumination system 415 are time-varying analog voltage signals indicative of the amplitude of the emitted fluorescence 31 at any instant in time generated by each cell as it is illuminated by the laser beam 25. Thus, the analog signals (also referred to as analog output) are in the shape of time-varying waveform pulses 497 as illustrated schematically in Figs. 52 and 53. In general a waveform pulse 497 is defined as a waveform or a portion of a waveform containing one or more pulses or some portion of a pulse. Thus, the amplitude of each waveform pulse 497 at any instant in time represents the relative rate of photon emission 31 of each cell at that instant in time as the cell passes through the laser beam 25. X chromosome bovine sperm cells have a higher DNA content than Y chromosome bovine sperm cells (e.g., about 3.8%). As a result, live X cells labeled with a fluorescent stain as noted above will produce a different waveform pulse 497 than pulses from any other labeled cells. By analyzing the pulses 497 as noted below (see Signal Processing, Slit Scanning, and Critical Slope Difference), each cell can be identified as an X cell or not identified as an X cell (~X). In general, as used herein, X cells refers to live X cells, Y cells refers to live Y cells and ~X cells refers to the combination of live Y cells and cells which otherwise produce a detectable fluorescence emission 31 but which cannot be identified with a reasonable probability as being live X cells.

[0176] The timing of each waveform pulse 497 indicates the position of each cell in the stream 21. Since the rate at which the sheath fluid 19 is being delivered through the nozzle 137 remains constant, and since the distance d (in Fig. 25) between the nozzle 137 and the droplet break-off location 107 is known, the position of each droplet 33 is known and the cells, if any, within each droplet 33 are known. Thus, the microprocessor 131 can calculate the instant at which each forming droplet passes through the charging collar 631 and can control the polarity of the collar 631 and thus control whether a droplet 33 is charged for deflection by the charging elements 631 of the sorting system 119. Since the microprocessor 131 knows the droplet formation rate and identifies the cells within a droplet as X or ~X, the microprocessor 131 knows the cell content of each droplet 33 and keeps track of (or enumerates) the number of cells in each population 123, 125. Depending on the sort strategy, see below, the microprocessor 131 determines which droplets 33 are charged for deflection and which droplets 33 are not charged so that they are not deflected.

Signal Processing


A. Digital Sampling Introduction



[0177] As previously described, the interaction between the laser beam 25 and the particle produce a "pulsed" photon emission 31 (e.g., a fluorescence emission) that is captured by the collection lens 511 of the optics system 109 and delivered to a photodetector 117. The photodetector 117 converts the photon energy at any instant in time to an analog voltage output of time-varying amplitude. This output is a series of waveform pulses 497 (Figs. 43 and 44) which contain many features that can be used to discriminate among populations of particles. Among these features are the total photon emission, the rate of photon emission as a function of the particle's spatial transit through the laser beam, the maximum rate of photon emission during the transit, the average rate of photon emission during the transit, and the time required for transit. The combination of laser beam geometry 459, particle size, distribution of the emission source through the particle volume and particle velocity determine the frequency spectrum of waveform pulse 497. For example the system 1 used with bovine semen described previously it has been determined that each cell 201 produces a waveform pulse 497 of between 800 ns and 1200 ns in duration. It has also been determined that as a function of frequency, more than 97% of the power in the waveform pulse 497 is delivered at frequencies below 30 MHz. This frequency spectrum will be discussed later as it related to the Nyquist sampling theorem. Taken together these waveform pulses 497 form an output signal 701 from the photodetector 117 that is a continuous, time varying, signal that represents the transit of the particle stream through the apparatus. In addition to features of individual pulses that are used to discriminate among populations, the time varying signal provides a precise record as to the relative spacing (time and position) among the individual particles that pass through the apparatus and relative velocity of the particles moving through the apparatus. This precise time, position and velocity record can be synchronized with the droplet generation clock signals 703 as shown in Fig. 44 to determine which particles are members of a particular droplet 33 formed by the droplet generation apparatus 105. This information can be used as the basis for determining "coincidence" or the occurrence of a desired and undesired particle in a single droplet 33. The ability to accurately determine the number and classification of each particle in a droplet 33 allows for accurate, efficient sorting.

[0178] Digital signal processing 705 as illustrated in Fig. 72 may be employed to analyze detection of fluorescence pulses 31 as indicated by synchronously sampled output signals 701 from the photodetector 117. This processing would be implemented in pulse analysis software employing instructions and/or algorithms, as noted herein. The time-varying analog output signal 701 of the photodetector 117 is provided to an A/D (analog/digital) converter 689 which synchronously samples it. Synchronously sampling means sampling to produce digital information corresponding to the analog output. Synchronously sampling is also referred to as continuously sampling or streaming acquisition. As noted below, the sampling rate depends on the frequency spectrum of the analog output.

[0179] Converter 689 provides an output including digital information 707 which is provided to the microprocessor 131 or other digital analysis device which executes the pulse analysis software to analyze the digital information 707. In general, the pulse analysis software would include digital pulse detection, pulse feature extraction and pulse discrimination.

B. Sampling Frequency & Signal Frequency Spectrum



[0180] The signal output 701 from the PMT 117 is captured by a high speed analog to digital converter 689 (ADC) that samples the output 701 continuously at a frequency of 105 MHz. It is well understood that when sampling a time varying signal it is necessary for the sampling frequency to be at least twice the maximum frequency contained in the signal being sampled. This is known as the Nyquist sampling theorem. For this reason the output signal 701 from the PMT 117 is first sent through a 40 MHz low-pass filter 854 (see Fig. 39) to ensure that the maximum frequency contained in the signal 701 is under the 52.5 MHz limit imposed by the sampling rate. It is important to note that the optical 109, fluidic 15 and detection systems of the apparatus 1 have been tuned to produce a pulse waveform 497 having optimum frequency characteristics for sampling at the 105 MHz rate. The sampling rate may be varied between about 25 and 200 MHz.

C. Pulse Processing



[0181] Pulse processing takes place in four (4) TigerSharc DSP processors that share memory and are connected to one another by high-speed parallel ports. As illustrated in Fig. 39, the four processors are: 1) a data management processor 863 which receives data from a high-speed ADC 689 which digitizes the output signals 701 from the photodetector 117; 2) a pulse detection processor 865 which detects the waveform pulses 497 represented by the digital information; 3) a feature extraction and discrimination processor 867 which extracts features from the detected pulses 497 and discriminates the pulses 497 based on the extracted features; and 4) a sort processor 873 which determines a sort classification for each pulse 497 based on the extracted features and the discrimination, which determines sort decisions for the corresponding cells and droplets 33 and which is synchronized with droplet formation 105. In general a processor 863, 865, 867, 873 completes a task and sets a "flag" so that companion processors know there is data available to process.

[0182] Each processor 863, 865, 867, 873 runs independently of the others, maximizing the overall throughput because they do not interrupt each other. Thus, any processor 863, 865, 867, 873 may be capable of performing any function and one or more processors or functions may be combined into a single processor or spread out over a plurality of processors. The processor 863, 865, 867, 873 labels as used above and this application are used for convenience only and are not intended to be limiting in any way.

[0183] All four processors 863, 865, 867, 873 are linked to a DSP board SDRAM 851 for exchanging information and are linked to a processor input/output (I/O) 857 for synchronization and communication with a peripheral I/O bus 859 connected to the PC 735 and the sort pulse generator 861. The processor I/O 857 may be implemented by two or more SharcFIN I/O processors connected by a communication link. Sort signals 853 are provided to the PC 735 via the peripheral I/O bus 857 and are used to control the sort pulse generator 861 controlling the charging of droplets 33.

[0184] The processor I/O 857 receives the output 707 from the analog/digital converter (ADC) 689, e.g., Bitware Corp. 105MHz/2-channel, 14 bit capable of 105MHz/1-channel sustained. The ADC 689 is connected to the photodetector 117 output for converting its time varying analog output signals 701 into digital information 707 and is also connected to an I/O board SDRAM 855 for storing the blocks of digital information from the ADC 689.

[0185] In general, the analog output signals 701 from the photodetector 117 are indicative of characteristic A or characteristic B (e.g., X or ~X). The A/D converter 689 converts the analog output signals 701 from the photodetector 117 of the flow cytometry system 1 into corresponding digital information 707. The processors 863,865, 867,873 analyze and classify the digital information 707 and provide a sorting signal to the sorting system 119 as a function of the detected and classified digital information.

D. Data Acquisition



[0186] As previously stated, the signal output 701 from the photodetector 117 is captured by a high speed analog to digital converter (ADC) 689 that samples the output continuously at a frequency of 105 MHz. Data (digital information 707) are transferred immediately into high-speed memory blocks (I/O Board SDRAM) 855 which serve to buffer the incoming data. These memory blocks 855 are organized in a manner to maintain the integrity and sequence of the data stream 707. These memory blocks 855 are also accessible by the digital signal processing (DSP) processors 863, 865, 867, 873 by direct memory access (DMA). In this manner the processors 863, 865, 867, 873 can access the incoming data 707 without interrupting the ADC 689. This facilitates efficient transfer of data 707 to these processors 863, 865, 867, 873 for feature extraction, analysis and sort classification. Throughout this process, the data management processor 863 keeps the pulse samples 707 in order and time indexed (relative to the master clock 737, which is 128 times the droplet 33 frequency) to preserve their reference to "real time" or the actual time that the cell passed through the laser beam 25. The ADC 689 ping-pongs back and forth between two inputs, continuously sampling the time varying analog output signals 701 including the waveform pulses 497 and converting them into digital information 707 which is provided in blocks 855 to the I/O Board SDRAM under the control of the data management processor 863. Processor 863 assembles the information 707 into a continuous stream.

E. Initializing Detection Parameters



[0187] In order to effectively distinguish over background noise, the digital pulse detection software 747 should be provided with information indicating signal background second order statistics, i.e. knowledge of the behavior of the output voltage signal 701 from the photodetector 117 when there is no fluorescence pulse 497. These statistics can be learned by software for initializing detection parameters 741 in an unsupervised manner during the initialization period immediately following startup of the system 1. In general, a pulse may be defined as 2 or 3 standard deviations from the background level.

[0188] Due to the possibility that introduction of the carrier fluid 17 into the sheath fluid stream 191 may cause a change in background fluorescence emission, the carrier fluid 17 should be present for the initialization of the detection parameters. Simple computation of the second order statistics of a time sequence of output voltage signal values may overestimate the standard deviation of the background (due to the possible presence of fluorescence pulses 497 in the sequence). An iterative procedure is therefore preferred to gradually eliminate this effect. The pulse detection software 747 accomplishes this by computing the statistics of the total signal 701 (background + pulses), using these values to apply pulse detection logic, re-computing the signal statistics without samples detected to be within pulses, and repeating this procedure until the background statistic estimates converge (or a fixed maximum number of iterations occurs). By evaluating the background with cells present, a more accurate indication of the expected correct pulse 497 amplitude can be determined. Table III summarizes the detection initialization procedure for determining detection parameters for use by the pulse detection software.



[0189] In general, the A/D converter 689 converts the analog output signals 701 from the photodetector 117 into corresponding digital information 707 indicative of characteristic A or characteristic B (e.g., X or ~X). The digital signal processor 865 determines background characteristics of the time-varying output signals 701 from the digital information 707 corresponding thereto, detects waveform pulses 497 from the digital information 707 as a function of the determined background characteristics, and provides a sorting signal 853 to the sorting system 119 as a function of the detected pulses 497.

F. Initial Discrimination Parameters



[0190] Similar to the detection parameters (and subsequent to their initialization as shown in Table III), parameters for use in a discrimination algorithm may be initialized in an unsupervised fashion. Unlike the detection algorithm parameters, however, an iterative procedure is not necessary. In this case, software for initializing the discrimination parameters 745 detects a preset number (e.g., 100,000) of fluorescence pulses 497, computes the features to be used for discrimination for each detected pulse 497, and uses a clustering procedure (see Table IV for a summary of candidate clustering procedures) to assign these pulses 497 to populations of interest (e.g. X, ~X).
Table IV. Summary of clustering approaches being considered for use in discrimination algorithm parameter initialization.
Algorithm NameAlgorithm Approach
k-Means Iterative (local) minimization of sum of squared distance (Euclidean or Mahalanobis) between points within each population [1]
Fuzzy k-Means Expectation-Maximization of (Gaussian) mixture model [2]
Agglomerative Hierarchical Merging of "nearest" clusters (starting with each data point as its own cluster) until desired number of clusters is reached. Various measures for determination of "nearest" clusters include distance between closest points, distance between furthest points, distance between cluster means, and average distance between points. [1]


[0191] Fig. 73 contains an example of the results of application of a k-means clustering procedure to define population 1 and population 2 based on statistics of distribution. The second order statistics of these populations are then used to set the parameters necessary for discrimination (the coefficients of a 1st or 2nd order polynomial decision function). Table V summarizes the discrimination initialization procedure.



[0192] In general, the A/D converter 689 converts the analog output signals 701 from the photodetector 117 into corresponding digital information 707 indicative of characteristic A or characteristic B (e.g., X or ~X). The digital signal processor 867 generates initial discrimination parameters corresponding to the digital information 707, discriminates the digital information as a function of the initial discrimination parameters, and provides a sorting signal 853 to the sorting system 119 as a function of the discriminated digital information.

G. Digital Pulse Detection



[0193] The first processing step is pulse detection performed by pulse detection processor 865 to determine whether a particular waveform is a waveform pulse 497 corresponding to a fluorescence emission 31 of a cell. The processor 865 executes a pulse detection algorithm which identifies sample sets that are likely to represent either particles targeted for sorting into a population or particles targeted to be avoided because they are potential contaminants to a population. In the case of bovine sperm sorting, a dye is added to quench the emission 31 of nonviable cells, causing their associated pulse intensities to be -1/3 the intensity of a live cell. Nonviable cells are not considered as sorting targets or potential contamination. They are not considered detected pulses 497. Pulses 497 from live cells are detected by monitoring the intensity of samples for a successive number of samples that rise above the background levels. Once this level crosses a statistically determined threshold the processor 865 jumps to a later time that is approximately 75% of the expected pulse 497 width for a live cell. If the level is still above the threshold, the series of samples are considered to be a pulse 497. Samples from detected pulses 497 are moved to a block of memory used by the feature extraction processor 867.

[0194] A statistical anomaly detection approach is one example which may be employed by digital pulse detection software 747 although it is contemplated that other approaches for identifying and/or isolating digitized pulses 497 may be used. Essentially, digital samples 707 of the output voltage signals 701 from the photodetector 117 detecting fluorescence which are statistically anomalous from the background are considered to part of a pulse 497. For additional robustness (to minimize noise detections), additional temporal criteria may be included.

[0195] Pulse detection proceeds as follows. When the voltage output signal 701 from the photodetector 117 is not a pulse, the Mahalanobis distance from the background of incoming samples 707 of the signal 701 is computed and compared with a preset threshold. If the distance of a given sample exceeds the threshold, it is considered to be the potential start of a pulse 497, and the pulse detection software begins to buffer the incoming samples. If the next predetermined number of samples (e.g., 25) also exceed the threshold, a pulse 497 is considered to have started and buffering continues until the pulse end criteria are met; otherwise, the buffer is reset and checking for the start of a pulse resumes. While in a pulse 497, if a sample is below the threshold, then it is considered to be the potential end of a pulse and the buffer location is recorded (but sample buffering continues). If the next predetermined number of samples (e.g., 25) are also below threshold, the pulse 497 is considered to have ended and the pulse 497 consists of the buffered samples up to the recorded location. Table VI summarizes the pulse detection algorithm, and Fig. 49 provides an illustration of the results of pulse detection on a digitally acquired fluorescence pulse 497.



[0196] In general, the A/D converter 689 converts the analog output signals 701 from the photodetector 117 into corresponding digital information 707 indicative of characteristic A or characteristic B (e.g., X or ~X). The digital signal processor 865 analyzes the digital information and processor 873 provides a sorting signal 853 to the sorting system 119 as a function of the detected digital information.

H. Feature Extraction and Discrimination



[0197] The next processing step is feature extraction performed by the feature extraction and discrimination processor 867. This processor responds to flags set by the pulse detection processor 865. Samples from detected pulses are placed in memory shared with the feature extraction processor 867. Features such as area, pulse width, pulse height, Gaussian correlation coefficient and/or other features are determined for each pulse 497. In some cases pulses 497 are determined to be "doublets" or invalid and features are not extracted. For the case of bovine sperm 201 features are only extracted for pulses 497 that have the general amplitude and width of a live X or Y cell. Typically, the pulse amplitude for a live sperm cell is in the range of about 700-900 mV, although this range may be as wide as 500-1000 mV. Once the features are extracted they are compared to the feature spaces defined for the population(s) selected for sorting. If the features match the feature spaces identified for sorting, then processor 867 sets a flag indicating a positive sort command to the sort processor 873. In general, the classification of a particular cell is made by the discrimination processor 867 and the sort decision is made by the sort processor 873.

[0198] Digital information 707 representing fluorescence emissions 31 (and thus the characteristics of corresponding cells which created them) are discriminated by software 757 based on specific features or characteristics which exhibit distinguishably different statistical behavior in feature space (the n-dimensional orthogonal space formed by n features as the axes) for the different populations of interest. Therefore, the first step in analyzing digital information 707 for the purposes of discrimination is computation of these features, a process called feature extraction performed by pulse analysis software 749 executed by the processor 867. Table VII lists the several candidate features which software 749 may use for this application. One or more of these features will be selected to form the feature space for classification. It should be noted that there are additional features providing enhanced separation so that this list is exemplary, not comprehensive. For example, the software 749 may employ a subroutine 753 to determine pulse 497 area and/or may employ a subroutine 755 to determine pulse 497 peak.
Table VII. Summary of candidate features currently being considered for use in digital pulse analysis relating to feature extraction.
Feature NameFeature Description
Pulse Area Approximated by sum (or average) of pulse samples
Pulse Peak Maximum value of pulse samples
Pulse "Inner" Area Sum (or average) of inner TBD samples of pulse (centered on pulse mean)
Pulse Width Number of samples in pulse.
Pulse "Gaussianity" MSE or correlation coefficient of pulse with a Gaussian shape with the same 2nd order statistics.
Pulse "Lagging Peak" Pulse value at TBD samples past peak (or mean)
Critical Slope Difference (CSD) Slope of pulse at a point along the pulse at which the difference between the first derivative of a pulse produced by particles having characteristic A and the first derivative of a pulse produced by particles having characteristic B is at or near a maximum

I. Slit Scanning



[0199] In general, the elliptical spot 459 provided by the illumination system 109 measures the relative DNA content differences in cells. Resolution can be improved further by analyzing the fraction of the pulse 497 of the fluorescence emission 31 detected by the photodetector 117 more likely to contain characteristics which are being evaluated. A biological phenomenon of certain cells (e.g., bovine sperm cells) is the localization of the X/Y chromosomes in a sub-equatorial region 225 which is immediately adjacent the longitudinal midline or equator or center of the nucleus 213 of the cell 201 and which has a length of about 1 µm. (See Fig. 6). In fact, the X/Y chromosomes are not necessarily centered in the nucleus 213. Thus, resolution can be improved by converting the time-varying analog output 701 of the photodetector 117 into digital information 707 and analyzing a portion of the digital information corresponding to the fraction of the pulse 497 of the fluorescence emission 31, e.g., corresponding to the light emitted from the circumequatorial region 225 such as 20-60% and particularly 20-30% of the waveform pulse centered around the pulse 497 peak.

[0200] As noted above, slit scanning can be employed to obtain the fluorescence measurement from a portion of each cell's chromatin rather than from the chromatin as a whole. The elliptical spot 459 provided by the epi-illumination system 415 noted above measures the relative DNA content differences in cells from specific sections of the chromatin, so that the resolution of X cells and ~X cells relative to one another is improved. As noted above, the slit scanning measurement technique is a fluorescence measurement approach that focuses the excitation beam 25 so that a dimension of the focused spot size 459 is much less than a cell diameter as shown in Fig. 6. In this way, the cell 201 is scanned by the laser beam 25 as the cell passes through the elliptically-shaped beam spot 459. The resulting waveform pulse 497 produced by the photodetector 117 output 701 detecting the fluorescence emission 31 resulting from slit scan illumination contains information about the localization of fluorescence along the length of the cell 201. As shown in Figs. 45-48, as the cell 201 traverses the elliptically-shaped beam spot 459, the time-varying waveform pulses 497 (red/orange line) are the convolution of the relative beam intensity (blue line) and the relative emitted pulse intensity (which corresponds to the fluorescence emissions from stain excited by the elliptical spot as the cell traverses the beam and which varies because the fluorescence distribution along the axis of the cell varies).

[0201] By illuminating only a fraction of the cell's chromatin at one time, the resulting time-varying analog output 701 from the photodetector 117 contains information specific to the localization of fluorescence within the chromatin along the longitudinal axis of the cell 201. Although the detected fluorescence emission 31 from slit scanning is less than the detected emission 31 from scanning by a beam 25 having a spot width comparable to the cell diameter, resulting in waveform pulses 497 from slit scanning having a lower pulse amplitude, the majority of difference between the X-chromosome bearing cells and the Y-chromosome bearing cells appears in the center 20-30% to 20-60% of the waveform pulse 497. If only the rectangular area 725 in Fig. 53 is considered for discriminating X-Y sperm cells, then a larger relative difference can be measured between the localized variation in DNA content within the section of chromatin that corresponds to the rectangular region 725 due to the presence of the X and Y chromosomes within that region as compared to the total DNA content of the cells. For example, bovine X-Y sperm cells have a difference in total DNA content of about 3.8%. The fluorescence emission 31 from the X and Y chromosomes will be contained in the rectangular region 725. If this rectangular region 725 accounts for 20% of the total waveform pulse 497 corresponding to a fluorescence emission 31, then a 14% difference in relative DNA content within the region will exist. By measuring the relative DNA content differences from specific sections of the chromatin, the resolution of X-Y sperm cell differentiation is improved (e.g., from 3.8% to 14%). Fig. 54 illustrates the resolution attainable using slit scanning illumination and processing the areas from only the center 20% of the pulse 497 (i.e., the rectangular region 725 of Fig. 53). The histogram of Fig. 54 allows a very high percentage (e.g., 98%) of the X chromosome bearing sperm and Y chromosome bearing sperm to be identified with a high degree of confidence (e.g., 95%). In comparison, the histogram of Fig. 55, which illustrates the resolution obtainable when using standard illumination techniques, shows that slit scanning offers a significant improvement over the results obtained using standard illumination techniques.

[0202] Two approaches which can be employed to obtain the area 725 of the center portion of the waveform pulse 497 as illustrated in Fig. 53 are digital signal processing (DSP) of digitized photodetector 117 time-varying analog output 701, as discussed in this section, or analog integration using an analog threshold trigger, as noted below. As noted herein, DSP processing involves continuously sampling the time-varying analog output 701 from the photodetector 117 to obtain digital information 707 corresponding to the output 701 and applying DSP algorithms to the digital information 707 to extract features, such as area size, from the digital information corresponding to the center portion 725 of the waveform pulse 497 which corresponds to the difference in DNA content due to the presence of an X or Y chromosome in different cells 201. As a simple example, the center 20% of the total area of each waveform pulse 497 would be determined by analyzing the digital information 707 corresponding thereto. The analysis would be used to generate a histogram such as illustrated in Fig. 53.

J. Pulsed Laser Scanning



[0203] In one example, it is contemplated that for example the system 1 include a pulsed laser to illuminate the cells. The slit scanning (as described above) may or may not be employed. For example, a mode-locked solid-state laser can be used to emit a train of electromagnetic pulses having a pulse width (duration) of 1-100 picoseconds at a pulse frequency of about 50-150 MHz and at an average power output of about 100-500 milliwatts. One suitable laser is a Vanguard 350 mode-locked solid-state laser (available from Spectra- Physics, Mountain View, CA 94039), which is operable to emit a series of pulses about 12 picoseconds in width (duration) at a frequency of about 85 million pulses per second and at an average power of about 350 milliwatts. Because the 350 mW of power is delivered over extremely short bursts of only 12 picoseconds, the peak power output of such a laser is several hundred times (e.g., about 800 times) greater than the average power.

[0204] The output of such a laser can be described as quasi continuous wave (quasi-cw) because, for many applications, the pulse repetition rate is fast enough to approximate a continuous wave (cw) output. Indeed it is possible to operate the system as described above with a quasi-cw laser in much the same manner as one would operate with a cw laser. This provides certain advantages because solid-state lasers typically operate more efficiently, require less extensive cooling systems, and require less maintenance than most other lasers.

[0205] A quasi-cw pulsed solid-state laser can also result in significantly improved signal-to-noise ratios using digital signal processing techniques. A timing circuit may be included and is operable to produce a timing signal indicative of the arrival of laser pulses at the interrogation location 115 (i.e., the area where the laser beam 25 illuminates the stream 21). For example, the timing circuit may be a laser pulse sensor 3003 as shown in Fig. 40 for detecting light corresponding to the laser pulse including scattered light generated by the interaction of each laser pulse with the fluid stream 21 and/or including light from the laser pulses. Alternatively, for lasers which may be triggered, a triggering signal may be provided to the microprocessor 131 and/or the A/D converter 689 to synchronize either or both to the laser pulses, as noted below with regard to Fig. 50. In either example, the laser pulse timing would provide a clock signal for the system.

[0206] Referring to Fig. 50, a timing diagram illustrates the timing relationship between the laser pulses LP, the fluorescence emissions FE from a cell as a result of repeated excitation by the laser pulses LP as the cell passes through the beam spot 459 and the digital samples DS of the photodetector output 701. As shown in FIGS. 45-49, as a cell passes through the laser beam spot 459 the fluorescence emission 31 varies depending upon the amount of illumination of the portion of the cell which generates the fluorescence emission 31. Fig. 50 illustrates twenty (20) laser pulses LP1-LP20 which impinge upon a cell as the cell passes through the interrogation zone 115 of a flow cytometer 1. Each laser pulse LP1-LP20 corresponds to a fluorescence emission FE1-FE20, respectively, which exponentially decays after substantially instantaneous excitation by the laser pulse.

[0207] The microprocessor 131 may control the A/D converter 689 (see Fig. 40) so that the converter 689 samples the output signal 701 of the photodetector 117 at or near peak of each fluorescence emission FE1-FE20, as indicated by digital samples DS1-DS20, respectively. In other words, the timing circuit synchronizes the sampling rate of the A/D converter 689 with the fluorescence emissions FE1-FE20. The resulting digital signal produced by transit of a particle through the interrogation zone 115 is the functional equivalent of the digital signal that would have been produced by the digitization of a pulse waveform 497 from a continuous wave laser. As shown in Fig. 51, for example, by considering only the fluorescence intensity during the digital samples DS1-DS20 and disregarding fluorescence intensity drop-off between laser pulses LP1-LP20, the fluorescence intensity as a function of time is a pulse waveform 497. This permits feature extraction by the microprocessor 131 from the digital signal 707 generated by the pulsed laser in order to analyze the cell providing the fluorescence emissions FE1-FE20. A more sensitive photodetector having relatively fast response time of about 2 nanoseconds or less may be used to more accurately detect the fluorescence emissions.

[0208] Thus, the pulsed laser provides advantages in a flow cytometry system 1 in that it is possible to use a lower power pulsed laser to obtain substantially the same analysis that would be obtained with a cw laser operating at an average power much higher than the average power of the pulsed laser. Further, the high peak power from a pulsed laser tends to saturate the fluorophores so that the fluorescence emissions are maximized thereby reducing the signal-to-noise ratio of the output signals of the photodetector. In other words, by using a laser pulse that contains much more energy than is required to saturate the fluorophore, variations in the output of the laser do not result in variations in the fluorescent emissions 31.

[0209] Those skilled in the art will recognize that there are many ways to cause a laser to emit a series of pulses. It is understood that other pulsed lasers, including other mode-locked lasers, Q-switched lasers, and cavity dumping lasers, could be used in place of the mode-locked laser discussed above. Similarly, many other ways to time the digital sampling and process the resulting information will be apparent from the foregoing disclosure. For example, the digital sampling could be timed so there is a different delay (or no delay) between a laser pulse and a digital sample. Likewise, the number of digital samples per pulse or the number of pulses that elapse between digital sampling can also be varied.

K. Estimation of Population Characteristics



[0210] As noted above, flow cytometry can be used to discriminate X-bearing bovine sperm cells from Y-bearing bovine sperm cells based on their relative 3.8% difference in DNA content. Discrimination is achieved through analysis of characteristics of the time-varying signal 701 that is produced by the photodetector 117 used to record the fluorescence emission 31 as the stained cell passes through the interrogation location 115. This interaction is illustrated in Figs. 45-48. Figs. 45-48 illustrate how a pulse waveform 497 is generated by the fluorescence emissions 31 resulting from the interaction between the laser beam 25 and a stained sperm cell 201. The emission pulse 497 is the convolution integral of the excitation spatial function 498 and the emission spatial function of the cell 201. Characteristics of the fluorescence pulse waveform 497 are used to classify a cell as X, Y or undetermined. In one example, X-Y discrimination relies on two pulse characteristics: peak pulse height and pulse area.

[0211] These characteristics are illustrated on the example pulse that appears in Figs. 52 and 53. Figures 52 and 53 are examples of pulses 497 from X-bearing and Y-bearing sperm cells. The pulses 497 were generated from a computer model that assumed excitation illumination with a laser beam 25 having an elliptically-shaped beam spot 459 having a 2µm Gaussian beam waist W1 (Fig. 6) and that the DNA content difference was distributed uniformly across the center 20 percent of the cell 201. These assumptions are representative of slit scanning illumination of bovine sperm cells 201 having a localized DNA difference as discussed in more detail above. Integration of the pulses 497 results in a 3. 8% average difference between the pulse 497 area for an X cell and the pulse 497 area for a Y cell.

[0212] It is possible to generate histogram and scatter plots of the pulse 497 peak and area characteristics for stained cells and nuclei. Figs. 56-59 contain histograms of the pulse area characteristic for stained nuclei and live cells, plus scatter plots of the pulse 497 area and peak characteristics for stained nuclei and live cells. Some of the items that may limit live cell discrimination, and ultimately the cell sorting rate are evident in these plots. Notably, the live cell histogram of Fig. 57, and to a lesser extent the nuclei histogram of Fig. 56, have a left shoulder that is typical of fluorescence intensity histograms for mammalian sperm cells. It has been determined that the left shoulder is generated by one or more populations of slightly unaligned cells (i.e., cells that generate relatively weaker fluorescence emissions due to slight deviations from the optimal alignment, but that are not so far out alignment to cause the relatively brighter fluorescence emission 31 from the narrow edge 209 of the sperm head 205 to be collected by the optics system 109). Only about half of the X cells can be easily identified in the live cell area histogram. The rest overlap with the Y cell population and the non-aligned cell populations. Even when peak pulse height is added, as shown in the scatter plots in Figs. 56-59, X-bearing cell classification may be significantly limited.

[0213] Figs. 60-61 illustrate the overlap of the X and Y population distributions. In Figs. 60-61, a four-component computer model has been applied to raw data 6000 (Fig. 60) to estimate population statistics for two populations of non-aligned cells (6001, 6003), aligned live Y cells (6005) and aligned live X cells (6007) (Fig. 61). It is desirable to discriminate the X and Y populations as a function of the coefficient of variation (CV) of the X and Y populations. For example, it is desirable to minimize the coefficient of variation (CV) of the X and Y populations in order to improve discrimination. In particular, when a population of sorted X cells is desired, it is desirable for the CV of the X cell population to be less than 1.5%, more desirably about 1.3%, and even more desirably less than 1.3%. Traditionally, the CV of a distribution of fluorescence intensity of sperm cells has been considered with respect to the distribution functions for only two populations (X and Y). Quality control with respect to CV has been limited to crude subjective estimation of the CV of the X and Y populations to decide whether continued analysis or sorting is worthwhile.

[0214] One function of the microprocessor 131 is to provide an automated estimation of the CV of the X population using the four-component model illustrated in Figs. 60-61. In order to estimate the CVs of the populations present in a feature (e.g. pulse area) distribution, it is necessary to estimate the 2nd order statistics of the population distributions. This may be achieved by applying a model of a known form and finding the best fit of that model to the observed data.

[0215] Given the expectation of normally-distributed data, an approach consisting of Parzen Window based non-parametric density estimation (utilizing a Gaussian kernel function) followed by application of a Gaussian mixture parametric model has been chosen. Specifically, the four-component model illustrated in Figs. 60-61 consists of a sum (or mixture) of four uni-variate Gaussian distributions, with these four components being the feature distributions corresponding to aligned X cells, aligned Y cells, and a two-component aligned cell population. The parameters characterizing the model then are the population means (averages) (4), population standard deviations/variances (4), and prior probabilities (expected % of the overall distribution) (4). These 12 parameters can then be varied to achieve a best fit of the model to the observed data histogram. With the model component parameters thus estimated, an estimate of the CV of a population of interest (in particular, X cells) may be determined from the estimated population standard deviation and mean:



[0216] In order to reduce computational complexity, constraints have been placed on the model to reduce the dimensionality of the parameter space. In particular, the standard deviations of the model components corresponding to the aligned X and aligned Y populations have been constrained to be the same. Also, the aligned X and aligned Y components have been constrained to make up the same percentage of the overall mixture - thus the non-aligned populations are assumed 50% X cells and 50% Y cells.

[0217] Non-parametric density estimation is applied prior to model fitting to obtain an improved estimate of the total density function (being the sum of the component densities) underlying the raw histogram data. The specific technique applied is known as "Parzen Windows" (Duda, Hart, and Stork, 2001), here utilizing a Gaussian kernel or window function due to the assumed sumof-Gaussian nature of the underlying density. The standard deviation of the Gaussian kernel is chosen to be 1% of the number of populated histogram bins; this value has been empirically observed to provide adequate but not excessive smoothing of the histogram. Each data point in the histogram then contributes a kernel function centered on the histogram bin containing the data point. The density estimate is then obtained as the sum of the kernel functions.

[0218] The methodology chosen for variation of the model parameters to achieve the best fit to the data is known as Expectation Maximization (See Duda R.O. , Hart, P.E., and Stork, D.G., 2001, Pattern Classification 2nd Ed., John Wiley & Sons; and Moore, A., "Very Fast EM-based Mixture Model Clustering using Multiresolution kd-trees," in M. Kearns and D. Cohn, Eds., Advances in Neural Information Processing Systems, pages 543-549,1999, Morgan Kaufman).

[0219] The specific algorithmic implementation utilized is as follows:
  1. 1) Initial conditions for the model parameters are set. The top two local maxima in the Parzen density estimate are used as the initial Y and X mean locations (the initial amplitude of the maxima for both X and Y populations being estimated as the amplitude of the right peak). The initial X and Y population variance is estimated as the variance of the data to the right of the local minimum occurring between the left and right peaks relative to the right peak location. Also, the initial X and Y population prior probabilities are set as the percentage of all points falling to the right of this local minimum. The initial X and Y population Gaussian density estimates are then computed using these parameters and subtracted from the total Parzen density estimate. The mean and variance of the remaining data points are then computed and used to initialize the two-component aligned population model as follows. The two population means are assumed (arbitrarily) to be 5% apart (2.5% above and below the overall aligned mean). Given an (initial) assumption of equal priors and equal variances, then, the component variances are given by:

    where σ2 is the variance and µ is the mean of the respective population.
  2. 2) Updated estimates of component population statistics (means, standard deviations, and priors) are computed using the Parzen density estimate. Each histogram bin location is weighted in the statistical computations by the Parzen density estimate in that bin. Additionally, each data point contributes to all component population statistic computations weighted by the degree to which that point is believed to belong to a given population, based on the current component population parameters. This degree of membership is determined as the ratio of a given component population (Gaussian) probability density value to the sum of all component population probability density values at the data point. Thus we have (for all data points x in the histogram, populations cp ∈ {cx, cy, cu}, and population parameter vector θp=[µp, σp, Pp]) population component memberships used in the computation of updated parameter estimates given by:



    Updated means and variances are then computed using the Parzen density estimate values weighted by these membership values, with updated priors given by the average membership for each component population over all data points.
  3. 3) Parameter updating procedure continues until all parameters reach steady-state (i.e., stop changing significantly from one update iteration to the next (or a maximum allowed number of iterations occurs).


[0220] As previously mentioned the aligned X and Y populations are constrained in this procedure to have the same variance and prior probability. This constraint is achieved by using the average of the X and Y variance and prior values computed via the above procedure at each iteration.

[0221] Alternatively, a similar modeling approach can be applied to a three-component model (Figs. 62-63) in which the cells comprising the two unaligned populations 6001, 6003 in the four- component model are treated as a single Gaussian distribution 6011 rather than two distinct subpopulations. The non-aligned cells can be modeled as third Gaussian distribution (shown Fig. 63) having a mean and standard deviation determined by a best fit of the left shoulder and left major peak of the raw data 6010 (shown Fig 62). The three-population model also has estimated statistics for the aligned Y Population 6015 and aligned X Population 6017. One advantage of the three-component model is that it requires only 9-dimensional parameter space (compared to 12-dimensional parameter space required for the four-component model). However, it has been found that the four-component model typically results in an estimated density function that more closely matches the raw data.

[0222] Those skilled in the art will recognize that a wide variety of statistical techniques can be used to estimate the characteristics of the aligned X and aligned Y populations. Thus, the four-component model, the three-component model, or other models may be implemented by any parametric or non-parametric computer algorithms to estimate the characteristics of the aligned X cell and/or aligned Y cell populations.

L. CV-Based Selection of Staining Conditions



[0223] Several factors affect the efficiency of sorting stained cells within a population into enriched subpopulations of cells. Among these factors is the amount of differential fluorescence between the various subpopulations of cells within a stained population. Differential fluorescence is affected by dye uptake, which varies based upon staining factors, such as for example, the concentration of the stain, the length of the staining period, the temperature at which staining occurs, and the number and concentration of any additives that may be included with the stain or added to the staining mixture. Accordingly, adjustments to any or all of these factors may be made to increase the sorting efficiency (the rate at which cells may be sorted into at least one enriched subpopulation of cells with certain degree of purity and/or a minimal loss of desired cells) of the population of stained cells. Further, one can increase efficiency of a multisample sorting system by adjusting one or more of these factors for each sample, thereby countering any sample-to-sample variations. In the context of bovine sperm sorting, for example, sorting efficiency can be improved by adjusting one or more of the foregoing staining factors from one semen sample to the next to counter bull-to-bull variations or sample-to-sample variations within the same bull.

[0224] A determination of the coefficient of variation ("CV") for a given fluorescence emission characteristic of a population of cells to be sorted is one manner in which to determine if adjustments to the staining conditions could be made to achieve a desired sorting efficiency. For example, one may adjust the staining conditions as a function of the CV of any feature extracted from the pulse waveform generated by movement of a cell through the interrogation location, such as any feature indicative of total fluorescence intensity or peak fluorescence intensity (including total fluorescence intensity and peak fluorescence intensity). As previously discussed in greater detail, CV is an indicator of the homogeneity or consistency of a distribution of a measurable property or characteristic of a population, such as for example a fluorescence emission characteristic of a particular subpopulation of a given population. CV may be determined by dividing the standard deviation of the measured characteristic of a sample by the sample mean. CV can also be determined automatically by the flow cytometry system 9, such as by implementation of the iterative CV estimation algorithm discussed in detail above. The lower the CV, the greater the homogeneity or consistency of the distribution of the measured characteristic.

[0225] As applied to the staining and separation of sperm cells, the CV of a particular fluorescence emission characteristic for a sample of X and Y chromosome bearing sperm cells may be affected by the staining conditions. The concentration of the stain, the length of the staining period, the temperature of the staining mixture, and/or the number and concentration of additives affect the CV of a given fluorescence emission characteristic. Increasing the concentration of the stain, the length of the staining period, and the temperature of the staining mixture and/or decreasing the number and concentration of additives will generally lower the CV. Such conditions may be altered individually or in combination. In addition, if any one of these factors is such that it would tend to increase the CV of a fluorescence emission characteristic, such as for example, by shortening the staining time, then any one or more of the other conditions may be adjusted such that it counteracts the effect of the first, such as for example, by increasing the dye concentration, with the overall result being a decrease in the CV of the fluorescence emission characteristic to a level sufficient to achieve a desired sorting efficiency.

[0226] Accordingly, by manipulating any one or any combination of these factors in this manner, the CV of a fluorescence emission characteristic of the X and Y chromosome bearing populations may be decreased to a value that enables sorting of the sperm sample into a subpopulation of gender enriched semen comprising a desired percent purity of X chromosome bearing cells.

[0227] Unfortunately, changes that tend to lower the CV of the X bearing sperm may have negative consequences such as increased cost or decreased sperm motility or fertility. For example, other things being equal it is desirable to use lower stain concentrations and shorter staining periods to minimize any harmful impact of the staining process on the sperm. With this in mind, one may predetermine a CV at which an acceptable sorting efficiency will be achieved. Thereafter, a fraction of the cell sample to be sorted is stained and subjected to flow cytometric analysis. A fluorescence emission characteristic of the fraction is determined, and the fraction is classified into subpopulations based upon the characteristic. The CV of the fluorescence characteristic is determined with respect to the cells of one of the subpopulations (an enriched subpopulation). If the CV of the fluorescence emission characteristic of the cells of the enriched subpopulation is equal to or less than the predetermined CV at which sorting is to occur, then the remainder of the cell sample is stained according to the conditions under which the fraction was stained. The sample is thereafter sorted, for example, according to the methods disclosed herein. If the CV of the particular fluorescence emission characteristic of the cells of the enriched subpopulation is greater than the predetermined CV at which sorting is to occur, then another fraction of the same sample is analyzed in a similar manner, but under staining conditions believed to achieve a yet lower CV. In such a situation, the CV may be lowered by, for example, increasing the length of the staining period, increasing the concentration of the dye, increasing the temperature at which the fraction is stained, or any combination thereof. This series of steps (i.e., removal of a fraction from the sample to be sorted, adjustment of the staining conditions, and a determination of the CV) is repeated until the CV of the particular fluorescence emission characteristic of the cells of the enriched subpopulation is determined to be equal to or lessthan the predetermined CV. Thereafter, the remainder of the sample is stained accordingly and may be sorted. The cell sample may comprise a semen sample, and the cells of the enriched subpopulation comprise X chromosome bearing sperm cells.

[0228] Accordingly, to one example there is a process for evaluating a set of conditions for staining a population of cells for sorting, the population comprising a first type and a second type of cell. The process comprises (a) staining a fraction of the population of cells with a fluorescent dye under a set of staining conditions; (b) exposing the stained cells to electromagnetic radiation as the stained cells are passed through an interrogation location of a flow cytometer at a rate, R; (c) determining a fluorescence emission characteristic of the exposed cells; (d) using the determined fluorescence characteristic to classify the cells into two or more sub-populations, one of the subpopulations being an enriched subpopulation of the first cell type; (e) determining a coefficient of variation for the fluorescence emission characteristic of the cells of the enriched subpopulation; and (f) determining whether to modify any staining condition under which the cells are to be stained or the rate, R, at which the stained cells are passed through the interrogation location of the flow cytometer. In another example, another fraction of the population of cells is stained under a different set of staining conditions and steps (b) through (e) are repeated with that fraction. This process may be performed on two, three, four or any number of additional fractions. In yet another example, multiple fractions of cells are drawn from the sample at the same time. Each fraction may be stained simultaneously, or each may be stained subsequent to the previous fraction being passed through the flow cytometer. In the former case, each fraction may be stained with its own unique set of staining conditions and step (f) may comprise using the respective CVs to determine a set of staining conditions to be used to stain additional cells. In the later instance, the staining conditions of the subsequently stained fractions may be altered according to the determination of step (f) with respect to a previously analyzed fraction. The process is repeated until the CV may be repeated until the CV is determined to be about equal to or less than a specified CV (e.g., 1.3%).

[0229] Alternatively, once one has predetermined a CV at which an acceptable sorting efficiency will be achieved, the entire cell sample may be stained. A fraction of the cell sample is removed and subjected to flow cytometry analysis. A fluorescence emission characteristic of the fraction is determined and used to classify the cells into two or more sub-populations. The CV of the fluorescence characteristic is determined with respect to the cells of an enriched subpopulation. If the CV of the fluorescence emission characteristic of the cells of the enriched subpopulation is equal to or less than the predetermined CV at which sorting is to occur, then the remainder of the cell sample is thereafter sorted. If the CV of the particular fluorescence emission characteristic of the cells of the enriched subpopulation is greater than the predetermined CV at which sorting is to occur, then a second fraction from the same sample is analyzed in a similar manner and the CV of the same fluorescence characteristic is determined. The CV of the second fraction may be lowered by, for example, increasing the length of the staining period, increasing the concentration of the dye, or any combination thereof. This series of steps (i.e., removal of a fraction from the sample to be sorted and a determination of the CV) is repeated until the CV of the particular fluorescence emission characteristic of the cells of the enriched subpopulation is determined to be equal to or lesser than the predetermined CV. Thereafter, the remained of the sample may be sorted, for example, according to the methods disclosed herein. In a particular n, the cell sample comprises a semen sample, and the cells of the enriched subpopulation comprise X chromosome bearing cells.

[0230] Accordingly, another example is a process for evaluating a set of conditions for staining a population of cells for sorting, the population comprising a first type and a second type of cell. The process comprises (a) staining a fraction of the population of cells with a fluorescent dye under a set of staining conditions; (b) exposing the stained cells to electromagnetic radiation as the stained cells are passed through an interrogation location of a flow cytometer at a rate, R; (c) determining a fluorescence emission characteristic of the exposed cells; (d) using the determined fluorescence emission characteristic to classify the cells into two or more subpopulations, one of the subpopulations being an enriched subpopulation of the first cell type; (e) determining a coefficient of variation for the fluorescence emission characteristic of the cells of the enriched subpopulation; (f) determining whether to modify any staining condition under which the fraction of cells are to be stained or the rate, R, at which the stained cells are passed through the interrogation location of the flow cytometer; and (g) applying the modified staining condition to the remainder of the population of cells. In another example, steps (a) through (f) are repeated at least once with at least one other fraction of the population of cells. Steps (a) through (f) may be repeated once, twice, three times, four times or a greater number of times. In another example, multiple fractions of cells are drawn from the sample at the same time. Each sample may be stained simultaneously, or each may be stained subsequent to the previous fraction being passed through the flow cytometer. In the later instance, the subsequent staining of the fractions may be altered according to the determination of step (f) with respect to a previously analyzed. In still another example, the process further comprises prior to step (g), selecting the modified staining condition that results in the lowest coefficient of variation for the fluorescence emission characteristic. In yet another example, the process comprises the repetition steps (a) through (e) until the coefficient of variation for the fluorescence emission characteristic of at least one of the fractions is about 1.3% or less. In another example, the process further comprises prior to step (g), selecting the modified staining condition that results in a coefficient of variation of about 1.3 or less.

[0231] In addition to performing such an analysis before sorting the entire sample as detailed above, a similar analysis may be performed while the staining and sorting of the sample is occurring in an effort to ensure that sorting efficiency is maintained. Accordingly, in another example, the CV of a fluorescence emission characteristic of the cells of an enriched subpopulation of a fraction of a sample that has been previously stained, a portion of said sample which is in the process of being sorted, is determined as described above. Adjustments to the staining conditions under which these samples were stained are made according to the methods discussed above with respect to the presort adjustments.

[0232] The selection of a predetermined CV at which an acceptable sorting efficiency will be achieved is based upon several factors, including for example, the type of cell being sorted, the rate of sorting, and the degree of purity desired with respect to sorting of the population into enriched subpopulations. Generally, a CV is selected that will allow for sorting to the desired percent purity of the enriched subpopulation while minimizing the amount of time necessary to achieve the same, such as for example, by achieving an 85% degree purity of the enriched subpopulation while minimizing the length of the staining period. With these factors in mind, the CV of the fluorescence emission characteristic of the cells of an enriched subpopulation is generally between about 2.0% and about 1.0%.

M. Critical Slope Difference Feature Extraction



[0233] The microprocessor 131' with digital signal processing (DSP) illustrated in Fig. 40 employed as part of a cell sorter makes it possible to extract features of the time resolved fluorescence emission pulse, particularly features that cannot be easily or inexpensively obtained using analog technology. In particular, a pulse feature which exhibits non-linear properties and which significantly improves the separation and thus the resolution of particles A and B (e.g., improves the discrimination of live, aligned X sperm cells) is a feature referred to as critical slope difference (CSD). CSD is a quantification of the slope of the fluorescence emission pulse at a signal amplitude where the difference between the first derivative of a pulse produced by particle A (e.g., a X-bearing cell) and the first derivative of a pulse produced by particle B (e.g., a Y-bearing cell) approaches a maximum.

[0234] Functions that describe fluorescence emission pulses may be expressed in terms of signal amplitude as a function of time: y = x (t). Within the context of detecting CSD features, a function may be defined that describes the fluorescence emission pulses in terms of pulse duration time as a function of signal amplitude. This function may be referred to as an M function. The M function is obtained by transposing the fluorescence emission pulse function as shown below.

Fluorescence Emission Pulse function : y=x(t)

M Function: t=M(y)

t= pulse duration

y= signal amplitude



[0235] Comparison of the M functions for typical X and Y bovine sperm cells illustrates the discriminating power of the CSD feature. The top panel of Fig. 64 shows average M plots for X-bearing and Y-bearing sperm cells. The middle panel in Fig. 64 shows a graph of the first derivatives of these average M plots (i.e. M') for signal amplitude values less than the peak height of the average Y-bearing fluorescence emission pulse. It can be seen in this plot that as signal amplitude approaches the average peak height of the Y-bearing pulse, the difference between the first derivatives (M'y and M'x) increases significantly. Plotted in the bottom panel of Fig. 64 is the difference between the first derivatives (M'x-M'y) as a function of signal amplitude. The CSD feature quantifies the slope of M (M') for an individual pulse near the region where the maximum difference in first derivatives occurs (or the slope at a corresponding point on the fluorescence emission pulse function). For the purpose of discriminating X and Y bearing sperm cells, CSD is determined for each pulse at the point where the leading edge of the pulse intersects the CSD threshold, as shown in Figs. 62-63. In some examples, CSD may depend upon the characteristics of the illuminating beam such as beam width whether the beam is continuous or pulsed. An algorithm for determining CSD is discussed below with regard to Fig. 65.

[0236] Figure 64 illustrates that in some cases the CSD feature has a non-linear nature, such as in the case of sorting X-Y sperm cell populations. The difference between the derivatives (M'x-M'y) increases as the CSD threshold approaches the peak of the Y pulse. The nonlinear characteristic of this difference places the mean value of the nonaligned cells and the aligned Y cells 45% lower than the mean value of the aligned X cells in the CSD feature space. The standard deviation in the CSD feature space of the aligned X cells is largely unaffected (i.e. similar to that seen in the peak or area feature spaces). It is this nonlinear, high gain nature of the CSD feature that increases the number of aligned X cells that can be accurately discriminated.

[0237] One computational efficient method for determining the CSD value for a given pulse is illustrated in Fig. 65. A CSD threshold may be determined as a function of a peak height of the fluorescence emission pulses. In particular, it may be determined based on the average peak height of the fluorescence emission pulses. The CSD threshold is maintained at a point where about 25% of the pulse peaks from live, aligned cells fall at or below the threshold. Therefore, the CSD threshold is adjusted dynamically during the sort based on a running peak height distribution (i.e., relative to an average peak height). For example, the threshold may be based on a weighted running average of peak height (with more recent measurements being given more weight than older measurements). The CSD value is the slope of a line that passes through two points on the pulse. These points are the CSD pulse threshold and the pulse peak. Thus, in this example the CSD value is only an approximation of the slope of the pulse waveform 497 at the intersection of the leading edge of the pulse and the CSD threshold. However, other methods of computing the CSD value for a given pulse are readily apparent, some of which can provide more precise CSD values if desired.

[0238] In another example, the CSD threshold may be dynamically adjusted as a function of the CV of the CSD feature extraction for a subpopulation of particles. In the case of sorting sperm cells for example, by increasing the CSD threshold from a relatively low level (e.g., the pulse detection threshold) the CSD threshold will reach a level that results in a substantial increase in the CV of the CSD of the Y cells but is still low enough that the increase in the CV of the CSD for the X cells is significantly lower in comparison to the CV increase in the Y cells. This effect can be observed in the CSD distribution as a fanning out of one subpopulation in the overall CSD distribution. Good discrimination from the CSD feature can be achieved by maintaining the CSD threshold at this level.

[0239] It should be noted that the discriminating power of the CSD feature is enhanced by use of slit scanning approach to flow cytometry. The shape of the beam spot 459 can influence the shape of the pulse waveform 497. For example, by using a beam spot 459 having a relatively small width W1, a localized fluorescence difference in a sample particle (e.g., the localized fluorescent intensity difference resulting from localization of the X or Y chromosome in the central region 225 of a sperm nucleus 213) has a greater influence on the first order derivative of the pulse waveform. Accordingly, the slit scanning techniques may be used in combination with CSD feature extraction. Conversely, using a laser having a beam waist that is too large (e.g., equal to or greater than the diameter of the particles) may prevent effective use of the CSD feature to discriminate particles. The acceptable range for the width of the beam waist of the focused illumination beam will depend on a number of factors including the size and shape of the particles, the distribution of dye within the particles being analyzed, and the amount of difference between the typical waveform pulses for the particles to be discriminated. In the case of sperm cells, CSD feature extraction from waveform pulses 497 generated by excitation of bovine sperm cells 201 with a laser having a beam waist of less than 3 µm has worked well as indicated below. Of course, CSD feature extraction with any form of slit scanning discussed in the slit scanning section is considered.

[0240] Use of the CSD feature substantially increases the yield of the system, particularly in the case of sorting X-Y sperm cell populations because it allows collection of many more aligned X cells. Due to the overlap in the populations defined in peak vs. area or rise-time vs. area feature spaces, no more than about 70% of the aligned X cells can be discriminated with a certainty about or greater than 85%. When the CSD feature is used, 95% or more of aligned X cells can be discriminated, which significantly increases the percentage of live X cells that can be collected without reducing the purity of the population of collected X cells below a desired level of purity.

[0241] This is seen graphically in the live cell data shown in Figs. 66-69. The non-linear nature of the CSD feature allows X cells to be isolated for sorting. The gross selection on CSD applied in the scatter plot shown Fig. 68 results in a 70% pure X area population. When bi-variate sort discrimination is applied in the area and CSD feature spaces (Fig. 68), > 95% of the aligned X cells can be discriminated for sorting. The data in Figs. 66-69 were collected at a total cell throughput of about 22,000 live cells per second on one channel of a four-channel cytometry system (see multi-channel system discussion below). Even with coincidence detection enabled (high purity), over 6,000 X cells per second were collected at a purity level of at least 85% purity.

[0242] Figures 66-69 illustrate one advantage of the CSD feature when used to discriminate X-bearing and Y-bearing sperm cells. Figures 66 and 67 are histograms of the area feature of the fluorescence emission pulses for the feature space defined in the scatter plots shown Figs. 68 and 69. In Fig. 68, the CSD feature has moved most of the non-aligned and aligned Y cells completely out of the display of the scatter plot. This leaves a 70% pure X population in the \ frame of the scatter plot, which is what is shown in the pulse area histogram in Fig. 66. Non-CSD discrimination is shown in the pulse area/rise time scatter plot shown in Fig. 69. Aligned X cells make up about 30% of the corresponding area histogram (Fig. 67). More than 95% of the aligned X cells can be collected at > 85% purity using the CSD feature for discrimination. By comparison, no more than 70% of the aligned X cells can be discriminated using the traditional feature space on the right.

[0243] Several live cell sorts have been completed using the CSD vs. pulse area, bi-variate discrimination technique. Figure 70 is an example of how a sort region set in this two dimensional feature space can be used to exclude non-aligned cells and Y cells. Figure 70 illustrates a bi-variate sort region set on a scatter plot of CSD vs. pulse area scatter. Notice how the sort region drops lower on the area feature for high values of CSD (CSD increase from left to right and area increases from bottom to top) and moves higher on the area feature as CSD drops to lower values. The bi-variate sort region set on the above CSD vs. pulse area scatter plot was used to sort X cells at a sort decision rate > 6000 X cells per second with a input live cell rate of 26,000 cells per second. Purity based on flow cytometry re-analysis was 87%.

[0244] The CSD feature makes possible a high yield, no-coincidence abort (i.e., coincident accept or high recovery) sorting strategy. A pulse feature could provide nearly baseline separation and thus 100% accurate classification of live X and Y sperm cells. This condition would make it possible to sort cells at reasonably high rates without aborting droplets that contain both a cell classified as X and non-X (either unknown or Y). This sorting strategy is referred to as the high recovery or coincidence accept strategy. An experiment was performed to test this using the CSD feature. Coincidence accept sorts were performed with an input rate of 12,000 live X cells per second on one channel of a four-channel flow cytometer. 77% of the X cells were properly aligned, making 4,600 X cells per second potentially available for sorting. Under these conditions, 4,300 cells per second were sorted into the population of X cells. Subsequent purity analysis indicated a purity from this sort of > 87% without correction for dead cells and 89% with correction for dead cells. A high purity, coincidence reject detection sort was performed immediately after this sort under the same conditions. A collection rate of 3200- 3500 cells per second was observed. Purity analysis indicated a purity of 92% without correction for dead cells and a purity of 94% with dead cell correction.

[0245] The results of the above experiment indicate that at 12,000 live cells per second input, > 92% of aligned X cells can be collected at a purity > 85%. This is an indication that the CSD feature provides 95% accurate classification of all aligned X cells. Under these circumstances, yield from the cell sorter is limited primarily by correct cell alignment.

[0246] Figure 71 illustrates the flow cytometry re-analyses for a test in which the left panel corresponds to the high recovery coincident accept sort strategy (no coincidence abort strategy) and the right panel corresponds to the high purity/coincident reject sort strategy (coincident abort strategy). The left panel (87% pure) was for an output of 4,400 X cells per second without coincidence aborts. The right panel was from a sort completed under the same conditions except droplets containing contaminating cells were aborted. Purity for this sort was about 92%. These sorts demonstrate that high recovery, no coincidence abort sorts are possible when the CSD feature is used for discrimination.

[0247] Use of the CSD feature is not limited to sorting of sperm cells or any particular species of sperm cells. As those skilled in the art will appreciate from the foregoing disclosure, the CSD feature can be adapted to improve discrimination between any groups of particles that generate signal pulses having different first order derivative characteristics regardless of the cause of the difference.

N. Discrimination



[0248] Once the features of the pulses have been extracted by pulse analysis software 749, discrimination (e.g., classification) of pulses is accomplished by pulse discrimination software 757 executed by processor 867 employing a logic application such as Bayes Minimum Risk decision rule. This rule is a modification of a Bayes Minimum Error decision rule that allows assignment (and adjustment for) differing costs associated with making different erroneous classification (e.g., discrimination) decisions.

[0249] Bayes Minimum Error computes the decision boundary 763 or decision surface as the surface of equal a posteriori probability between populations in feature space. For the case of (assumed) Gaussian probability distributions this surface is in general quadratic, although in certain conditions may be linear (or be able to be closely approximated by a hyper-plane). The classification (e.g., discrimination) decision is made by first computing the a posteriori probabilities for a given point in feature space (generally from class-conditional probability densities and known/assumed a priori population probabilities using Bayes Rule) then choosing the class label as that of the population having the highest a posteriori probability.

[0250] Bayes Minimum Risk includes a factor to allow adjustment of the decision boundary 763 in the case when it is desired to assign different costs for making different classification errors (e.g. it may be costlier to classify "Y" cells as "X" cells than vice versa). In this application, this allows a trade-off between sorted sample purity and recovery. In this decision rule, the "risk" of assigning each possible class label to a point in feature space is computed as the sum of the a posteriori probabilities of membership in each population times the cost associated with classifying as the current population given true membership in each other population. Table VIII summarizes the procedure for Bayes Minimum Error classification. Note that for multi-variate Gaussian densities, evaluation of Bayes rule to obtain the a posteriori probabilities may be reduced to evaluation of the quadratic function seen in Table VIII, given that the coefficients W, w, and wo are as computed in the discrimination algorithm parameter initialization procedure given in Table V. Fig. 74 shows a graphical example of classification by this procedure. The illustration on the left is a schematic illustration of the two populations 1 and 2 and the decision boundary 763 separating the populations. The histogram on the right shows two concentric sets of ellipses defining the X and Y regions, with the decision boundary 763 being a line defined by the intersection of the ellipses.



[0251] For additional robustness, an additional step is taken in the classification of digital fluorescence pulses. The Mahalanobis distance of a pulse in feature space from the population assigned via Bayes Minimum Error is computed, and if greater than a threshold, the pulse is labeled as "not classified" or some other appropriate indication that it is not likely a member of any known population. Fig. 75 illustrates the effect of this additional step, again using features computed from digitally acquired fluorescence pulse data.

[0252] In general, the A/D converter 689 converts the analog output signals 701 from the photodetector 117 into corresponding digital information 707 indicative of characteristic A or characteristic B (e.g., X or ~X). The digital signal processor 865 extracts features from the digital information and processor 873 provides a sorting signal 853 to the sorting system as a function of the extracted features.

O. Sort Classification and Droplet Synchronization



[0253] The fourth, sort processor 873 manages droplet classification, implements sorting strategy and delivers a sort trigger pulse 853 that is synchronized with the droplet selected for sorting. This processor 873 receives cell classification information from the discrimination processor 867 and relates that information to the droplet generation clock 703 (i.e. aligns the position of particles classified for sorting into a population with the formation of droplets). It determines if there is coincidence within a droplet and manages that coincidence based on predetermined sort strategies. It maintains a FIFO of all cell classifications and droplet sort decisions that sets the correct delay between when the particle was observed in real time and when the particle arrives at the last attached droplet. It will produce a properly timed output pulse 853 of appropriate polarity and amplitude for each droplet selected for sorting.

[0254] In general, the A/D converter 689 converts the analog output signals 701 from the photodetector 117 into corresponding digital information 707 indicative of characteristic A or characteristic B (e.g., X or ~X). The digital signal processor 867 discriminates the digital information 707 as indicative of characteristic A or as indicative of characteristic B (e.g., X or ~X) and provides a sorting signal 853 to the sorting system 119 as a function of the discriminated digital information.

[0255] In general, the digital signal processors 863, 865, 867, 873 include instructions for detecting waveform pulses represented by the digital information, instructions for extracting features in the detected pulses and instructions for discriminating the detected pulses as a function of their extracted features. In addition, the processors include instructions for defining a decision boundary 763 discriminating between the extracted features representing characteristics A and the extracted features representing characteristic B. Further, the processors 863, 865, 867, 873 may optionally adjust the relative location of the decision boundary 763 with respect to the extracted features representing characteristic A and with respect to the extracted features representing characteristic B as a function of at least one of the following: (1) the purity of the at least one population with respect to either characteristic A particles or characteristic B particles, and (2) the quantity of characteristic A particles or characteristic B particles in the at least one population relative to the total quantity of characteristic A particles or characteristic B particles in the stream. For example, the processor may move the decision boundary 763 to include less of population 1 and more of population 2, or visa versa, based on the output of a particular sample or based on the desired output (e.g., as noted above with respect to the Bayes Minimum Risk decision rule to adjust the decision boundary for differing costs).

P. Drift Compensation



[0256] Given that over time the waveform pulses corresponding to the fluorescence emissions may vary or exhibit drift over time (due to staining variations, temperature change, sample age and/or other factors, for example), the system may optionally employ drift analysis software 761 (Fig. 72) defining dynamic thresholds which vary to compensate for any effects of drift. In particular, the pulse detection thresholds employed by software 747 may be adjusted for any slow variations in the signal background characteristics, and the discrimination algorithm of software 757 may adjust the decision boundary 763 (Fig. 74) to account for any drift in the populations in feature space.

[0257] In the case of the algorithm(s) employed by pulse detection software 747, the drift compensation software 761 accomplishes drift compensation by updating the background mean and standard deviation estimates based on sample statistics estimates computed within a moving window of a given length of samples (e.g., 10-100 samples) ending with the current sample. Given the (assumed) slow drift rate relative to the data acquisition frequency, the background statistics need not be updated every sample; rather, background statistic updates may occur periodically (e.g., every 6 seconds; see reference character 795 and Fig. 82). Additionally, the window may contain less than unity weighting to allow a "forgetting" rate to de-weight older samples relative to newer samples in the statistics computations. Fig. 76 illustrates the concept of statistic (mean) computation within a moving window without and with a "forgetting" rate.

[0258] Similar to the detection algorithm drift compensation, the discrimination algorithm(s) employed by pulse discrimination software 757 achieve drift compensation by periodic updates of the 2nd order statistics of the populations in feature space. In this case, however, only those feature values from pulses assigned to a given population are used to update the statistics of that population. Again, non-unity weighting may be used to include a "forgetting" rate. Fig. 77 shows a conceptual illustration of the effects of applying this technique to populations in feature space. Fig. 77 illustrates an example of drift compensation for population statistics in feature space. Yellow denotes population 1 (X), green population 2 (Y), diamonds the class mean estimates (with an exaggerated illustration of drift), and block arrows changes in the population covariance estimates reflected in deformation of the constant-sigma ellipses.

[0259] In general, the digital signal processor 863 employs a detection threshold for analyzing the digital information, which threshold is a function of a background mean estimate and a standard deviation of the sampled time-varying output signals computed within a moving window of samples ending with the current sample.

Q. Advantage of All Digital Techniques Over Analog Techniques



[0260] One of the main advantages for using an all digital system for sorting is that there is no "dead time" associated with the detection and analysis of a pulse. With analog systems there is always a finite "switching time" required for electronics to reset after the occurrence and detection of a pulse. This time is usually on the order of at least one microsecond. Since the digital system captures a continuous stream it really has no dead time.

[0261] Another advantage of a digital system is the ability to look forward and backward in time around a pulse classified for sorting. In general, the digital signal processing requires about five (5) droplet periods for analysis. Preferably, the time delay between droplet illumination 115 and droplet formation 107 is about seven (7) droplet periods. This allows the system to classify a particular particle based on the probability that it will contaminate the usable population as indicated by the features of the particular particle and based on the proximity of the particular particle to another classified particle. As an example, the sort processor 873 may reject a particle viewed as having a 50% probability of being a live X cell whereas the sort processor 873 may accept a particle viewed as having a 50% probability of being a live X cell when the particle is coincident with a second particle viewed as having a 95% probability of being a live X cell.

R. Analog Cell Analysis



[0262] It is also contemplated that the time-varying output signals from the photodetector may be processed by analog circuitry 819, such as a field programmable gate array, which may be less expensive than a digital cell analyzer. Fig. 42 is a block diagram of one example of an analog cell analyzer which may be employed as part of the system. Fig. 53 graphically illustrates the analog analysis. A threshold is set to produce a trigger based on pulse height. The threshold opens an integration window which gates an analog integrator to accumulate charge. The window remains open either for a fixed period or until the pulse amplitude fall below the trigger threshold. In this way, only the area of the portion of the pulse within the integration window is used for relative fluorescence measurements.

[0263] Referring to Fig. 42, the output 701 of the photodetector 117 is supplied to an integrator 825 which integrates the output signal 701 in synchronization with the droplet clock 703. The integrated signal is provided to a width/area comparator 827 for comparing the level of the integrated signal to a threshold level defining a pulse (e.g., a pulse may be defined as an integrated signal with 40% of a certain threshold. ). A dynamic threshold calculator 829 functions similarly to the drift compensation noted above in that monitors the integrated signal level and it varies the threshold level which the width/area comparator uses as a function of variations in the average integrated signal level.

[0264] The pulse discriminated signal is provided to a window comparator 837 to confirm that the pulse area is within an acceptable range. The pulse discriminated signal is also provided to a pulse width and trigger logic circuit 839 to confirm that the pulse width is within an acceptable range. If the area and width are acceptable, the logic provides a trigger signal to an I/O controller 843 which indicates the sort decision 841. Thus, the window comparator 837 and the pulse width and trigger logic 839 make the decision as to whether a cell should be classified as an X cell or a ~X cell.

[0265] The I/O controller 843 provides the sort decision 841 to the sort controller board 847 in the form of a X or ~X signal. The I/O controller 843 also includes a Universal Serial Bus (USB) interface 849 for connecting to the PC 735 and may have I/O port for connecting to slave controllers 845 of the other channels. The analog cell analyzer also includes a Joint Test Access Group (JTAG) port 833 for programming the width/area, comparator, the window comparator and the pulse width and trigger logic.

[0266] It is also contemplated that the analog cell analyzer may be employed simultaneously with the digital cell analyzer 705. For example, the analog analyzer may be used to adjust voltage thresholds used by the digital analyzer. On the other hand, the digital analyzer may be used to identify various features of the pulse and this feature information may be used to configure the analog cell analyzer, particularly if it is implemented with a gate array.

Control Strategies



[0267] In general, the microprocessor 131 is programmed to implement control and sorting strategies which are intended to optimize the efficiency of the system 1 in terms of throughput and/or loss of desirable particles to meet any cost requirements of the sorted product. This may involve, for example, balancing the need for high purity of at least one collected population and the need to recover at least a minimum percentage of desirable particles from the sample being sorted. Achieving such a balance is important, particularly in the context of commercial applications where cost and profitability are important considerations.

[0268] To this end, the microprocessor 131 implements a control strategy which is a series of instructions and/or algorithms that control system variables such as fluid delivery rate and/or sort parameters. The microprocessor also implements a sorting strategy which defines the decision process for determining how each particle or group of particles is sorted. Each particular control strategy may employ one or more sort strategies. Various sorting strategies may be used depending on such factors as the selected control strategy, the particle detection system and/or information relating to the particle distribution in the fluid stream.

[0269] Regarding particle distribution, Fig. 78 illustrates a fluid stream containing an exemplary distribution of particles. In this particular example, the stream is formed by a nozzle similar to the nozzle described above and contains a mixture of particles having different characteristics A and B, e.g., X and Y sperm cells. As shown, the cells follow generally one after another in a series which can be viewed as comprising sequential sets of particles. These sets include first particle sets each comprising one or more particles having a characteristic A (e.g., indicating a live X sperm cell), second particle sets each comprising one or more particles having a characteristic B (e.g., indicating a Y sperm cell or, more generally, a sperm cell which is not a live X cell (~X), such as a Y cell, or a dead X or Y cell), and third particle sets each comprising two or more closely spaced particles at least one of which has characteristic A and at least one of which has characteristic B (e.g., one more X sperm cells and one or more-X sperm cells). Third particle sets are also hereinafter referred to as "coincident" particle sets.

[0270] Whether a particular particle is considered as constituting a set by itself or part of another set will depend primarily on its spatial position and/or separation relative to adjacent particles. In a droplet sorting system, the various particle sets will be defined by the particles in the droplets. In a photo-damage sorting system where a laser is used to ablate (kill or otherwise damage) selected particle sets to provide a collected population having a desired content, as discussed below in the "Photo-Damage Sorting" section, the various particle sets will be defined by the spatial proximity of the particles, i.e., whether the spatial separation between particles is sufficient to enable accurate classification of the particles and/or the ablation of one or more undesired particles by the laser without also ablating one or more desired particles. Similarly, in a fluid-switching sorting system where portions of the stream containing selected particles are diverted to provide a collected population having a desired content, as is discussed below in the "Fluid Switching Sorting" system, the various particle sets will be defined by the spatial proximity of the particles, i.e., whether the spatial separation between particles is sufficient to enable accurate classification of the particles and/or diversion of selected particles.

[0271] It will be observed from the foregoing that sort decision applied to the different particle sets may be varied, depending on the desired result or throughput of the system. For example, in a droplet sorting system, the sorting strategy used may depend on the treatment of "coincident" droplets, i.e., droplets containing third particle sets. In the handling of bovine sperm cells in a flow cytometry droplet sorting system and method as described herein, for example, to enhance the number of X sperm cells in at least one collected population, it may be desirable to use a strategy where each coincident droplet containing an X sperm cell is accepted and sorted as if it contained only X sperm cells, even though the droplet may also contain an ~X sperm cell (coincident accept strategy). On the other hand, to enhance the purity of X sperm cells collected in the stated population, it may be desirable to reject each coincident droplet containing a ~X sperm cell even though the same droplet may also contain an X sperm cell (coincident reject strategy). In general, as will be pointed out below, there are many control strategies which may be employed to maximize particle throughput and there are many sorting strategies that may by employed with each particular control strategy. The strategies can be applied to various sorting techniques using flow cytometry, such as droplet sorting, photo-damage sorting, and fluid-switching sorting. Further, the above strategies can be used to sort any type of particle according to any desired characteristic or characteristics of the particle.

[0272] According to one example, the microprocessor controls the rate at which the fluid delivery system delivers the fluid containing the particles as a function of other variables of the system. The microprocessor can control the fluid delivery rate as a function of a desired output result. Since the microprocessor determines the identity of each particle and determines whether such is directed to at least one collected population, the microprocessor can determine and control the output result by varying the control strategy and/or by varying the sorting strategy. A desired output result may generally be defined as at least one of the following: (1) the purity of at least one collected population with respect to characteristic A particles or characteristic B particles ("high recovery"), and (2) the quantity of characteristic A particles in the stated population relative to the total quantity of characteristic A particles in the stream, or the quantity of characteristic B particles in the stated population relative to the total quantity of characteristic B particles in the stream ("high purity"). Alternatively, the system may employ a substantially constant fluid delivery rate and the microprocessor can the control the sort parameters as a function of a desired output result. The desired output result may generally be defined as a combination of (1) the purity of the particles in at least one collected population and (2) the quantity of desired particles available in the stream but not included in the stated population ("constant flow rate").

[0273] In general, it may be assumed that when sorting two populations an identified cell could have a 50/50 probability of being part of one population or the other. However, it is also contemplated that an unidentified cell may in fact have some other probability other than a 50/50 probability of being part of one population or the other. This other probability may be determined by empirical analysis or from other characteristics regarding the sample being sorted.

[0274] Several different control strategies are discussed in more detail below.

A. High Recovery Control Strategy



[0275] One type of control strategy may be referred to as a "high recovery' control strategy. The objective of this strategy is to maximize the number of desired particles sorted into the population of desired particles as long as the purity of that population is at or above an acceptable purity.

[0276] Pursuant to this strategy, the first particle sets described above are sorted into the population of desired particles because each of these sets contains one or more particles having a desired characteristic A. The third particle sets are also sorted into the population of desired particles (coincident accept) because each of these sets also contains one or more particles having a desired characteristic A, albeit accompanied by one more particles having characteristic B. On the other hand, the second particle sets are rejected (i.e., not sorted into the population of desired particles) because they do not contain a particle having the desired characteristic. To optimize throughput using this strategy, the microprocessor increases the fluid delivery rate as long as the purity of the collected population is at or above an acceptable level. Stated in the converse, the fluid delivery rate is increased as long as the probable level of contamination of the population of desired particles is at or below an acceptable level.

[0277] As an example, consider the use of a high recovery control strategy for sorting X and Y sperm cells in the fluid stream of Fig. 78. The desired result may be to sort all of the X cells into a population of X cells so long as the population remains at or above an acceptable purity, e.g., so long as X/(X+Y) is greater than 85% or some other acceptable level. To obtain this result, the first particles sets are sorted into a population of X cells because they contain only one or more X cells. The third particle sets are also sorted into the population of X cells because they also contain one or more X cells, even though they may also contain a Y (or ~X) cell. The second particle sets are sorted into some other population because they do not contain one or more X cells. In this example, the rate at which the fluid delivery system delivers the fluid containing the cells to the nozzle would continue to be increased as long as the purity of the population of X cells is greater than 85%. Conversely, if the purity of the population of X cells falls below 85%, the fluid delivery rate is decreased.

[0278] In the context of a droplet sorting system, it is known from Poisson's equation that for any given droplet generation rate, the number of multiple-cell droplets will increase as the cell delivery rate increases. In other words, increasing the delivery rate of fluid containing the cells will increase the number of multiple-cell droplets. Therefore, if the coincident accept sorting strategy is used and coincident droplets containing third particle sets are sorted into the population of desired particles, increasing the fluid delivery rate will result in a decrease in the purity of the collected population because at higher fluid delivery rates more coincident droplets are being generated and collected. Figure 79 illustrates this result for a two (2) particle fluid so that 100% of the desired particles are captured. As shown, at very low fluid delivery rates (FDR along x axis), the purity (y axis) of the resulting collected population is very high because very few coincident droplets containing third particle sets are being generated and collected. As the fluid delivery rate increases (FDR increases to the right along the x axis), the percentage of coincident droplets generated increases resulting in more coincident droplets being collected and reducing purity of the usable population (along the y axis). In the specific example illustrated, the fluid delivery rate is 30K particles/second at about 80% purity.

[0279] The results of using a high recovery strategy can be dramatic, as illustrated by a simple example where X and Y sperm cells are sorted using a droplet sorting process. Assume, for example that droplets are generated at a rate of 60K/sec, and that sperm cells are delivered to the interrogation location at a rate of 30K/sec. According to Poisson's equation, if all droplets containing X cells are sorted into the population of X cells, including coincident droplets containing X and Y cells, about 15,000 X cells will be collected every second. The collected population will include about 2,600 Y cells, reducing the purity of the population with respect to X cells to about 85.2%. However, the number of collected X cells (15,000) represents a substantial increase relative to a strategy where coincident droplets are not collected, as in the high purity strategy or mode discussed below. In the high purity mode, operating at a droplet frequency of 40K/sec and cell delivery rate of 40K/sec (10K cells/sec more than in the high recovery mode example above), only about 11,800 X cells are collected every second, or about 3,800 X cells less than in the high recovery strategy. Further, when the high purity strategy is used, about 9,200 X cells are lost or wasted because coincident droplets are not sorted into the population of X cells. Therefore, if less than 100% purity is acceptable, it may be desirable to use the high recovery mode to increase the number of X cells collected or, stated conversely, to decrease the number of X cells lost.

[0280] In summary, in the high recovery control strategy using the coincident accept sorting strategy, the particle delivery rate is inversely related to the purity of the collected population of desired particles (sometimes referred to as the "usable" population).

B. High Purity Control Strategy



[0281] A second type of control strategy may be referred to as a "high purity" control strategy. The objective of this strategy is to maintain the purity of the collected population with respect to particles having a desired characteristic at high level, so long as the quantity of desired particles in the collected population relative to the total number of desired particles available in the stream is at or above an acceptable quantity (i.e., so long as the quantity of desired particles in the stream which are not collected remains below an acceptable quantity). Pursuant to this strategy, the first particle sets described above are sorted into the population of desired particles because each of these sets contains one or more particles having a desired characteristic A, and because these sets contain no contaminating particles. On the other hand, the second and third particle sets are sorted into one or more "unusuable" populations (coincident reject) because they contain contaminating particles (i.e., characteristic B particles). To optimize throughput using this "high purity" strategy, the microprocessor increases the fluid delivery rate as long as the quantity of desired particles that are sorted into the usable population relative to the total number of desired particles available in the stream remains at or above an acceptable quantity.

[0282] As an example, consider the use of a high purity control strategy for sorting X and Y sperm cells in the fluid stream of Fig. 78. The desired result may be to sort all of the X cells into a population of X cells so long as the quantity of X cells collected from the stream remains at or above an acceptable quantity, e.g., at least 60%. To obtain this result, the first particles sets are sorted into the usable population because they contain only one or more X cells. On the other hand, the second and third particle sets are sorted into one or more unusable populations because they contain one or more contaminating (~X) cells. In this example, the rate at which the fluid delivery system delivers the fluid containing the cells to the nozzle would continue to be increased as long as the quantity of X cells collected in the usable population as a percentage of the total available quantity of X cells that have been sorted remains at or above 60%. Conversely, if the quantity of X cells not collected in the usable population rises above 40% of the total number of available X cells that have been sorted, the fluid delivery rate is decreased.

[0283] As noted above in the context of a droplet sorting system, it is known that increasing the fluid delivery rate will increase the number of multiple-cell droplets, and thus the number of coincident droplets containing third particle sets. Since coincident droplets are not sorted into the population of collected X cells when using a coincident reject sorting strategy, this means that increasing the fluid delivery rate will result in a increase in the quantity of live X cells lost to the unusable population.

[0284] Figure 80 illustrates this result for a two (2) particle fluid so that the usable population has 100% purity of desired particles. As shown, at very low fluid delivery rates (FDR along x axis), the percentage of desired particles in the usable population is very high because very few coincident droplets are being generated and rejected. As the fluid delivery rate increases (FDR increases to the right along the x axis), the percentage of coincident droplets containing third particle sets increases and more such sets are rejected. This reduces the quantity of desired particles that are sorted into the usable population relative to the total quantity of desired particles available in the stream (i.e., the percentage of desired particles from the stream which are collected in the usable population). In the specific example illustrated, the fluid delivery rate is about 40K particles/second and about 75% of the desired particles are sorted into the usable population.

[0285] In summary, in the high purity control strategy implementing the coincident reject sorting strategy, the particle delivery rate is inversely related to the percentage of desired particles in the collected population (i.e., high purity of desired particles in the usable population).

C. Constant Flow Rate Control Strategy



[0286] A third type of control strategy may be referred to as a constant flow rate control strategy. In this strategy, the microprocessor maintains the fluid delivery rate constant (or within a constant range) and varies the percentage of collected (or rejected) coincident droplets as long as the purity of at least one collected population is at or above an acceptable level and as long as the quantity of desired particles in that population is at or above an acceptable quantity relative to a total quantity of desired particles that have been processed. Stated in the converse, the fluid delivery rate is constant and the percentage of accepted (or rejected) coincident droplets varies as long as the probable level of contamination of the usable population is at or below an acceptable level of purity and as long as the probable quantity of desired particles that is lost to a population other than the stated (usable) population is at or below an acceptable quantity.

[0287] As an example, consider the use of a constant flow rate control strategy for sorting the fluid stream shown in Fig. 78. The desired result may be to sort X sperm cells into a usable population having a purity of at least 85% and to collect at least 60% of the X cells in the stream so that no more than 40% of the X cells are sorted into the unusable population. In this example, the rate at which the fluid delivery system delivers the particles would be maintained constant and the percent of collected or rejected third particle sets (coincident sets) would be varied as long as the purity of the usable population with respect to X cells is equal to or greater than 85%, and as long as the percentage of X cells sorted into the unusable population is less than 40% so that the percentage of desired particles sorted into the usable population is equal to or greater than 60% (variable coincident accept sorting strategy). As the percentage of accepted third particle sets increases, the purity of the usable population decreases, but the quantity of desired particles (e.g., X cells) sorted into the unusable population decreases. Conversely, as the percentage of accepted third particle sets decreases, the purity of the usable population increases, but the quantity of desired particles (e.g., X cells) that are sorted in the unusable population increases.

[0288] As noted above, it is known from Poisson's equation that the number of multiple-cell droplets (and thus the number of coincident droplets containing third particle sets) is constant for a constant fluid (cell) delivery rate. Since the number of coincident droplets is constant in this control strategy, the percentage of coincident droplets sorted into the usable population will impact both the purity of the usable population and the quantity of X cells that are wasted by being sorted into an unusable population. This is because the percent of unwanted Y (or ~X) cells in coincident droplets which are accepted and sorted into the collected unusable population is inversely related to the percent of X cells in coincident droplets which are rejected and thus not sorted into the collected usable population.

[0289] Figure 81 illustrates the constant fluid delivery rate control strategy in a flow cytometry droplet sorting system and method as described herein implementing a variable coincident reject sorting strategy for a two (2) particle fluid. As shown, line OL illustrates the inverse relationship between the percentage of rejected coincident droplets (x axis) compared to the percentage of accepted coincident droplets (y axis). When there is a very low percentage of accepted coincident droplets, there is a very high percentage of rejected coincident droplets. Conversely, when there is a very high percentage of accepted coincident droplets, there is a very low percentage of rejected coincident droplets. Line OL illustrates this inverse relationship and represents the operating line of the variable coincident accept sorting strategy at a given constant particle flow rate. At point P in Fig. 81 along operating line OL, the purity of the usable population is a given percentage depending on particle flow rate, e.g., 85% purity. As the percentage of accepted coincident droplets increases (to the left and upward) along operating line OL, the number of undesired particles that are sorted into the usable population increases and the purity drops below 85%, which may be unacceptable. As the percentage of accepted coincident droplets decreases (to the right and downward) along operating line OL, the purity goes above 85% and is acceptable.

[0290] At point LL along operating line OL, 75% of the coincident droplets are rejected (i.e., sorted into the unusable population) so that the percentage of desired particles that are wasted by being sorted into the unusable population is a given percentage based on the particle delivery rate, e.g., 40%. As the percentage of rejected coincident droplets increases (to the right and downward) along operating line OL, the percentage of desired particles that are sorted into the usable population decreases (e.g., to <60%), which may be unacceptable. As the percentage of rejected coincident droplets (to the left and upward) along operating line OL, the percentage of desired particles sorted into the usable population increases (e.g., to >60%) and is acceptable. Thus, for a constant flow rate control strategy implementing a variable coincident accept sorting strategy, the microprocessor may operate the system so the percentage of accepted and rejected coincident droplets varies in an operating range between point PI and LL as indicated by arrow OR. Note that operating range OR may encompass more or less of the operating line, depending on the level of tolerance for impurity and loss of desired particles to the unusable population.

[0291] In summary, in the constant flow rate control strategy using the variable coincident accept sorting strategy, the percentage of third particle sets which are accepted is inversely related to the purity of the usable population and inversely related to the quantity of desired particles wasted by being sorted to a unusable population.

D. Summary of Control Strategies



[0292] The following Table summarizes the control strategies noted above.
CONTROL STRATEGYHIGH RECOVERYHIGH PURITYCONSTANT FLOW RATE
Controlled parameter Fluid delivery rate Fluid delivery rate Sort parameters
Controlling parameter: Purity of population Quantity of desired particles in population Purity of population AND Quantity of desired particles in population
Sorting strategy Coincident accept Coincident reject Variable coincident accept
X/Y Sorting strategy Collect X droplets and X+ ~X droplets; reject ~X droplets Collect X droplets; reject X+~X droplets and ~X droplets Collect X droplets; vary percentage of collected X+~X droplets; reject ~X droplets
Definition The fluid delivery rate is increased as long as the purity of the population with respect to X particles is at or above an acceptable level The percentage of coincident droplets in the population is increased as long as the quantity of desired particles in the usable population relative to the total quantity of X particles in the stream is at or above an acceptable quantity The percentage of coincident droplets in the population is increased as long as the purity of the population with respect to X particles is at or above an acceptable level; to continue operation, the quantity desired particles in the usable population relative to the total quantity of X particles in the stream must be at or above an acceptable quantity
Converse Definition The fluid delivery rate is increased as long as the probability of contamination of the usable population is at or below an acceptable level of purity The fluid delivery rate is increased as long as the probability of loss of the quantity of desired particles in an unusable population is at or below an acceptable quantity The fluid delivery rate is increased as long as the probability of contamination of the usable population is at or below an acceptable level of purity AND as long as the probability of loss of the quantity of desired particles in the unusable population is at or below an acceptable quantity
Desired result > minimum acceptable purity; e.g., >85% purity >minimum acceptable quantity; e.g., >60% of desired particles captured (<40% of desired particles lost) > minimum acceptable purity and >minimum acceptable quantity; e.g., >85% purity and >60% of desired particles captured (<40% of desired particles lost)


[0293] Relatedly, a sorted sample obtained using one of the above control strategies can be combined with a second sample to obtain a final (e.g., commercial) sample having the desired characteristics. For example, a sample sorted according to the high purity strategy to produce a 100% pure population can be combined with a population of the same volume sorted to 80% purity to produce a final sample having a purity of 90%. Or in the case of animal sperm sorted to a high purity, an aliquot amount of the sorted sperm can be combined with an aliquot amount of unsorted sperm to produce a final sample of desired purity at lower cost than sorting the entire amount of sperm using any of the above sorting methods.

[0294] The above description of the control strategies assumes accurate identification and sorting of each droplet including each coincident droplet. In practice, 100% accuracy is not possible for any number of reasons. In order to minimize contamination, therefore, it may be desirable to reject particles which cannot be classified with certainty as belonging to the desired population. On the other hand, if certain particles can be identified and classified as being in the desired population within a certain selected probability (e.g., greater than 50% in the case of sperm cells), it may be desirable to classify the particles as belonging to the desired population so that they are not lost to the unusable population. Thus, as discussed earlier, particles such as sperm cells may be accepted or rejected for sorting into a population of desired cells based on the probability that such particles belong in the usable population.

[0295] The terms "usable" and "unusable" as used in the above table and this application are used for convenience only and are not intended to be limiting in any way. Thus, a "usable" population includes any "first" population, regardless of how or whether it is used, and an "unusable" population includes any "second" population different from the usable population, regardless of how or whether it is used. Similarly, a "desired" population means any population which is sorted according to selected particle characteristics.

[0296] The microprocessor and its signal processing software constitutes a system for processing the electrical signals from the photodetector to classify particles (e.g., particles in general and sperm particles in particular) according to characteristics of the particles and to obtain information relating to the distribution of the particles as described above with respect to Fig. 78. Furthermore, the microprocessor constitutes a control system responsive to the signal processing software for varying the rate at which the fluid delivery system delivers particles to the nozzle system as a function of the obtained information relating to the distribution of the particles. Furthermore, the microprocessor constitutes a control system responsive to the signal processing software for varying the sorting strategy as a function of the obtained information relating to the distribution of the particles.

[0297] In general, the microprocessor constitutes a control system responsive to information received from the flow cytometry apparatus for controlling the sorting system to vary its sorting strategy or for controlling the fluid delivery system. In other words, the microprocessor is capable of operating in a first mode to vary the sorting strategy, is capable of operating in a second mode for controlling the fluid delivery system, is capable of operating in a third mode to vary the sorting strategy and for controlling the fluid delivery system, and may be capable of operating in other modes. When operating in the first or third mode, the microprocessor is capable of varying the rate at which fluid is delivered as a function of at least one of the following: (1) the purity of the at least one population with respect to either characteristic A particles or characteristic B particles, and (2) the quantity of characteristic A particles or characteristic B particles in the at least one population relative to the total quantity of characteristic A particles or characteristic B particles in the stream.

Collection System



[0298] A collection system is needed to collect the droplets after they pass between the deflector plates. The collection system for a conventional cytometer may be no more than collection vessels disposed to catch the droplets in the various droplet streams after they pass between the deflection plates. Similar conventional collection systems can be used.

[0299] However, it may be difficult to use a conventional collection system in which the nozzle is oriented to direct the fluid stream at an upward angle, thereby giving the droplets a horizontal velocity component. One issue is that the droplets would travel some horizontal distance along their arched trajectories before they begin downward movement that would be suitable for landing in a collection vessel. For example, if the nozzle is pointed upward at a range of 45° to 60° and the droplets exit at a velocity between 15 m/s and 20 m/s, the droplets will be a horizontal distance of several meters away from the nozzle before they reach the apex of their trajectories and begin downward movement. Thus, a good deal of lab space would be occupied by the droplet streams. Furthermore, at a range of several meters it could also be difficult to make sure the droplets land in the proper collection vessels. The trajectories of the droplet streams can change whenever one or more operating conditions for the cytometer change (e.g., adjustment to the fluid delivery rate resulting in a change in the fluid velocity at the nozzle orifice). Changes in the trajectories of the droplet streams will be magnified by the distance that the droplets travel. Thus, changes in the trajectories that do not result in an appreciable change in droplet location at a point relatively near the nozzle could result in a significant change in location of the droplets at a location that is farther away from the nozzle. As discussed above, some systems employ feedback to the droplet formation and/or sample fluid delivery systems that could result in droplet streams that constantly alter their trajectories. One may also want to vary the pressure at which the sheath fluid is delivered to the nozzle. Air currents, temperature variations, and humidity variations could also alter the trajectories of the droplet streams. Any factor that could change the trajectory of the droplet streams could also require the collection vessels to be repositioned so the droplets land in the appropriate collection vessels. In contrast, the trajectories of droplets streams in a conventional cytometer having a downward pointing nozzle are less susceptible to variation. For example, the fact that the droplet streams have a substantially downward initial velocity means that variation in fluid velocity at the orifice does not result in any significant variation in the trajectories. Furthermore, the collection vessels are relatively close to the nozzle which makes the collection system more tolerant to trajectory variations in the droplet streams.

[0300] Figs. 83-85 show of a collection system, generally designated 2201, that may be used to collect sorted droplets in the systemn. The collection system is particularly suited for collection of droplets 33 when the cytometer nozzle system 101 is oriented to direct the fluid stream 21 at an upward angle or any other angle having a horizontal component. As the droplets pass between the deflector plates 629, they are sorted into multiple droplet streams 123, 125 (e.g., two) having different arched trajectories. As shown in Figs. 84 and 85, the trajectory of each droplet stream leads to one of two droplet intercepting devices 2203. If the droplets are being sorted into more than two droplets streams, a separate intercepting device would need to be provided for each additional droplet stream. Thus, the number of intercepting devices in a collection system will depend on the number of streams into which the droplets are being sorted.

[0301] Each intercepting device 2203 in the exemplary collection system has an impact surface 2205 positioned to span the trajectory of one of the droplet streams to divert droplets moving along that trajectory to a collection vessel 2207 positioned beneath each intercepting device. The impact surfaces are preferably made of a pliable material. Without being bound by a particular theory, it is believed that pliable materials cushion the impact of droplets striking the surface of the intercepting device, thereby reducing damage to the particles (e.g., sperm cells) in the droplets. For example, the intercepting devices may be constructed of polypropylene, polyethylene, or other similar polymers. Referring to Figs. 86 and 87, the intercepting devices 2203 have been constructed by cutting a droplet entryway 2211 (e.g., rectangular window) in one side of the bulb 2213 of a pipette 2215. Thus, a portion of the inside wall 2217 of the bulb opposite the droplet entryway forms a curved impact surface 2205 which spans the trajectory of the respective droplet stream. Conveniently, the tube of the pipette serves as a guide 2225 for directing fluid from the impact surface to the collection vessel.

[0302] Referring to Fig. 84, the intercepting devices are fastened to a collection system frame 2227, which holds them in place. In order to account for variability in the trajectories of the droplet streams, it is desirable to allow adjustment of the positions of the intercepting devices. For example, the vertical height of each intercepting device may be adjusted by sliding the guide tube up and down in a circular bore through a holder 2229. When the intercepting device is at the desired height, a set screw 2231 may be tightened to hold it at that height. The holder may be attached to a mounting plate 2233 which is attached to the collection system frame, for example by set screws 2235 (e.g., two set screws). The set screws 2235 pass through a horizontal slot 2241 in the mounting plate to allow lateral adjustment of the intercepting device. After adjustment, the set screws may be tightened to hold the circular holder in the desired location. Those skilled in the art will recognize that a variety of other fastening devices could be used to adjust the position of the intercepting device. The collection vessels 2207 are held beneath the intercepting devices in slots 2241 in a tray 2243 for holding the collection vessels. Thus, each collection vessel may be moved within a respective slot as necessary for it to remain in position under the respective intercepting device. Also, a water bath (not shown) may be provided around the collection vessel if desired to control the temperature of the contents of the collection vessel.

[0303] Referring to Fig. 85, an exit window 2245 (e.g., a rectangular window) has been cut in the back of one of the intercepting devices 2247 to allow one or more droplet streams to pass through the intercepting device. A second intercepting device 2249 is positioned behind the exit window to intercept the droplets that pass through this window. An exemplary entry window for the second intercepting device may be approximately the same size as the exemplary exit window for the first intercepting device. For reasons that will be apparent, it is desirable for the exit window to be significantly smaller than the entry window for the first intercepting device. For instance, an exemplary entry window for the first intercepting device is about 1.95cm (5/8 of an inch) high and about 0.95cm (3/8 of an inch) wide. In contrast, an exemplary exit window may be 0.32cm (1/8 of an inch) high and 0.79cm (5/16 of an inch) wide.

[0304] During operation of the cytometer, the collection system operates to intercept the droplets in the sorted streams. The intercepted droplets then drain down through the guide tubes 2225 of the intercepting devices 2203 and into the collection vessels 2207. In a case in which a cytometer has an upward pointing cytometer nozzle that directs droplet streams along arched trajectories, for example, the intercepting devices allow the droplets to be intercepted at a point on their trajectory that is significantly closer to the nozzle in comparison to the point at which the droplets would be collected by a conventional collection system (i.e., a collection system without intercepting devices).

[0305] Intercepting the droplet streams relatively early along their arched trajectories (e.g., while they are still moving upward) reduces the amount of variation in the location of the droplets at the time the droplets first encounter the collection system. Accordingly, the collection system can tolerate more variation in the trajectories of the droplet streams than a convention collection system could tolerate. Similarly, the droplets are less likely to be buffeted by air currents because of their shorter paths to the collection system.

[0306] A balance must be struck between moving the intercepting devices 2203 closer to the nozzle 101 to increase tolerance for trajectory variations and moving the intercepting devices farther away from the nozzle orifice to reduce or minimize the force of impact when droplets impact the intercepting devices, as by positioning the intercepting devices so they intercept the droplet streams substantially at the apex of their trajectories. Accordingly, the best location for the intercepting devices will depend on the durability of the particles (e.g. sperm cells) being sorted, the droplet velocities, and the expected magnitude of variation in the droplet stream trajectories. In the case of droplets containing bovine sperm cells having a velocity at the nozzle orifice of about 16 to 20 m/s, for example, the intercepting devices may be positioned in the range of 4-6 inches from the nozzle orifice. In a system in which a first intercepting device has a exit window and a second intercepting device is positioned behind the first intercepting device, for example, the first intercepting device may be in the range of about 10.2cm to 12.7cm (4 and 5 inches) from the nozzle. The second intercepting device may be in the range of about 12.7cm to 15.2cm (5 to 6 inches) from the nozzle. More desirably, the second intercepting device is about 5.5 inches from the nozzle.

[0307] The configuration in which one intercepting device 2203 is positioned to intercept the droplets that pass through an exit window of another intercepting device is particularly advantageous when one is not concerned about the purity of one of the sorted populations (e.g., Y chromosome-bearing sperm in the case of sperm sorted for use in breeding dairy cattle). Those skilled in the art will know that a number of stray droplets 2265 (e.g., a mist of stray droplets) having unknown contents may be produced by the cytometer in addition to the droplets in the sorted streams as shown in Fig. 85. The first intercepting device should be positioned so that the stream of droplets that are being sorted into the population for which there is the greatest tolerance for impurities hit the impact surface 2205 of the first intercepting device and the stream for which purity is most critical passes through the exit window 2245 to hit the impact surface of the second intercepting device. This way the majority of the stray droplets are collected into the collection vessel 2207 containing the population for which there is less concern about purity, as shown in Fig. 85, and will not contaminate the population for which purity is critical. Also by intercepting and collecting the stray droplets, one avoids the need to clean as often as if the stray droplets escape the collection system. In contrast to the first intercepting device, the second intercepting device only diverts droplets that pass through the smaller exit window. This facilitates maintenance of the purity of the population collected by the second intercepting device.

[0308] Those skilled in the art will recognize that the exemplary collection system could readily be modified in a number of ways. For example, it would be possible to construct a droplet intercepting device having an integrally formed (or otherwise attached) collection vessel beneath it. Similarly, although the intercepting devices in the example shown in Figs. 83- 87 are modified pipettes, it is understood that the intercepting devices 2203 can be any of a variety of shapes. For example, each intercepting device may comprise a flat or curved sheet, a spoon, a bowl, or other similar shape. The only requirement is that the intercepting device is operable to intercept droplets moving along a respective trajectory of a droplet stream and to divert the intercepted droplets into a collection vessel. However, one advantage to constructing the intercepting devices out of a readily available and relatively inexpensive product, such as a pipette, is that it may be more economical to replace and dispose of the used intercepting devices after each sample run rather than clean the intercepting devices between sample runs. This could help reduce costs of operating the collection system.

Collection Fluid



[0309] The sorted sperm are collected in a vessel that contains a collection fluid 2301 (Figs. 56 and 57). Generally, the purpose of the collection fluid includes cushioning the impact of the sperm cells with the collection vessel or providing a fluid support for the cells. Consistent with these considerations, the collection fluid may comprise a buffer or buffered solution and a protein source.

[0310] If included, examples of buffers or buffered solutions that may be used in the collection fluid are disclosed above with respect to sample collection and dilution. Typically, these buffers or buffer solutions will be in a concentration of about 0.001M to about 1.0M and have a pH of about 4.5 to about 8.5, preferably of about 7.0. For example, the collection fluid contains buffer comprising 0.96% Dulbecco's PBS (w/v) at a pH of about 7. 0 Moreover, the collection fluid may contain a metabolic inhibitor comprising 0.204g NaHCO3, 0.433g KHCO3, and 0. 473g C6H807·H20 per 25mL of purified water (0.097 moles/L of NaHCO3, 0.173 moles/L of KHCO3, 0.090 moles/L C6H807·H20 in water).

[0311] If included, the protein source may be any protein source that does not interfere with the viability of the sperm cells and is compatible with the particular buffer or buffered solution being used. Examples of common protein sources include milk (including heat homogenized and skim), milk extract, egg yolk, egg yolk extract, soy protein and soy protein extract. Such proteins may be used in a concentration from about 1% (v/v) to about 30% (v/v), preferably from about 10% (v/v) to about 20% (v/v), and more preferably about 10% (v/v). While milk may be used in combination with a buffer or buffered solution, generally milk is used in the absence of the same, as milk is a solution itself that may serve the same purpose of a buffer or buffered solution. In such instances, the collection fluid may contain about 80% (v/v) to about 90% (v/v) milk.

[0312] In addition to or in lieu of the protein source, the collection fluid may also comprise seminal plasma. Seminal plasma serves the dual benefits of improving sperm viability and motility and of stabilizing the sperm membrane (thereby preventing capacitation during the collection and storage of the sperm). Maxwell et al., Reprod. Fert. Dev. (1998) 10: 433-440. The seminal plasma may be from the same mammal from which the semen sample was obtained, from a different mammal of the same species, or from a mammal of a different species. If included in the collection fluid, typically the percentage of seminal plasma will be in the range of about 0.5% (v/v) to about 10% (v/v). If used in combination with a protein source, such as for example egg yolk or milk, the total percentage of seminal plasma and protein source will range from about 1 % (v/v) to about 30% (v/v). In such instances, the percentage of seminal plasma will be inversely proportional to the percentage of the protein source. Optionally, the collection fluid may also contain a range of additives that are beneficial to sperm viability or motility. Examples of such additives include an energy source, an antibiotic, and a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly, each of which is discussed above with respect to sample collection and dilution. Such additives may be added to the collection fluid in accordance therewith.

[0313] Alternatively, and in lieu of the use of a collection fluid, the sorted cells may be collected into a vessel containing or coated with a cryoextender used in the subsequent cryopreservation stepsThe sorted cells are collected into a cryoextender. In another example, the collected cells are sorted into a cryoextender comprising water, Triladyl® (Minitube, Verona, Wl, comprising glycerol, tris, citric acid, fructose, 5mg/100ml tylosin, 25mg/100ml gentamycin, 30mg/100ml Spectinomycin, and 15mg/100ml Lincomycin), egg yolk, and pyruvic acid. In yet another example, the collection fluid is the cryoextender comprising 25g Triladyl®, 25g egg yolk, and 10mM pyruvic acid in 75mL of water.

[0314] It is to be understood that the percent concentrations of protein in the collection fluid disclosed herein are those prior to the addition of the flow sorted cells. The addition of the flow sorted cells will dilute the final concentration of the collection fluid to about 1/20 that of what it was prior to the addition of the flow sorted cells. Therefore, for example, the collection fluid may initially contain about 10% (v/v) egg yolk. After the flow sorted cells are collected in the collection vessel containing the collection fluid, the final concentration of egg yolk will be reduced to about 0.5% (v/v).

Pre-Treatment of Intercepting Devices and/or Collection Vessels



[0315] In order to minimize possible damage to particles (e.g., sperm cells) that may be sorted , the intercepting devices 2203 and/or collection vessels 2207 (Figs. 56-60) may be treated prior to use. Such pre-treatment may comprise, for example, contacting or soaking the intercepting devices and collection vessels in a bath containing a composition that will serve to minimize the impact between the particle and the intercepting device. Upon removal of the intercepting devices and collection vessels from the bath, a certain amount of the composition will remain on the intercepting devices and collection vessels and serve as a cushioning agent for the particles in the droplets. The composition, therefore, should have characteristics suitable for providing the desired cushioning effect. In addition, the composition should also be compatible with the particle or cell being sorted, the sheath fluid, and the collection fluid. Consistent with these considerations, the composition used to treat the intercepting devices and collection vessels may comprise a buffer or buffered solution, a sheath fluid, a collection fluid, a cryoextender, any components contained in the buffered solution, sheath fluid, collection fluid, or cryoextender, or any combination thereof. Buffers, buffered solutions, sheath fluids, and collection fluids used for the staining and separation of sperm cells are described above. Accordingly, in an example, the intercepting devices and collection vessels are contacted with (e.g., soaked in or brushed with) sheath fluid or the intercepting devices and collection vessels are contacted with collection fluid. It is also possible that the intercepting devices and collection vessels are contacted with a cryoextender described below.

[0316] The contacting or soaking of the intercepting devices and collection vessels with the composition preferably occurs for a period of time sufficient to allow the composition to adhere to the surfaces of the intercepting devices and collection vessels. Such a period of time is generally less than about 90 minutes, preferably less than about 60 minutes, more preferably about 30 to about 60 minutes, and most preferably about 60 minutes. The intercepting devices and collection vessels may be merely contacted with the composition prior to use.

[0317] In lieu of or in combination with the contacting of the intercepting devices and collection vessels with the above-described composition, the intercepting devices and collection vessels may also be contacted with specific components contained in the sheath fluid, the collection fluid, and/or the cryoextender, such as for example, BSA, SSS, egg yolk, egg yolk extract, milk (including heat homogenized and skim), milk extract, soy protein, and soy protein extract. Accordingly, in one example, the intercepting devices and collection vessels are contacted with sheath fluid and subsequently contacted with 0.1 % (v/v) bovine serum albumin. In another example, the intercepting devices and collection vessels are contacted with sheath fluid and subsequently contacted with 10% (v/v) egg yolk. In another example, the intercepting devices and collection vessels are soaked in collection fluid and subsequently contacted with 0.1 % (v/v) bovine serum albumin. In another example, the intercepting devices and collection vessels are soaked in collection fluid and subsequently contacted with 10% (v/v) egg yolk.

[0318] Although the intercepting devices and collection vessels receive the same pre-treatment in each example described above, it is possible to use different pre-treatment protocols for the intercepting devices and the collection vessels. Likewise, some of the intercepting devices or collection vessels could receive one pre- treatment and others of the intercepting devices or collection vessels could receive a different pre-treatment. Certain advantages of the pre-treatment can also be obtained by pre-treating only the intercepting devices or only the collection vessels.

Concentration



[0319] As noted above, the sorted sperm collected by the flow cytometer have been diluted by the addition of various buffers and extenders, the staining fluid, the sheath fluid, and the collection fluid. Typically, the concentration of sperm cells after sorting by flow cytometry as described above is in the range of about 0.7 - 1.4 × 106 sperm cells/ml. Therefore, it is important to concentrate the sorted sperm cells to minimize the dilution shock to the sperm and to attain the proper concentration of sperm for cryopreservation and artificial insemination. Standard practice in the animal breeding industry, for example, is to perform artificial insemination with sperm at a concentration of either about 20 × 106 or about 40 × 106 sperm cells/ml. One way to concentrate the sperm cells is through centrifugation of the fluid collected by the cytometer. Another way to concentrate the sperm is to pass the fluid collected by the cytometer through a filtration system. These methods are discussed in more detail below.

A. Centrifugation



[0320] Any conventional centrifuge can be used to concentrate sperm. However in a commercial operation it is preferable to use a centrifuge having the capacity to centrifuge a large batch of sperm cells at once. During centrifugation a majority of the sperm cells will collect in a pellet at the bottom of the centrifuge tube due to the centrifugal force acting on the sperm cells. The magnitude of the centrifugal force is conventionally stated as the number of times the centrifugal force exceeds the gravitational force. Because the centrifugal force is the critical parameter and because the magnitude of the centrifugal force at any given speed (angular velocity) will vary depending on the length of the radius of curvature, the speed of centrifugation is typically specified by stating the magnitude of the centrifugal force. For example, a 600g force means the angular velocity of the centrifuge is selected so the resulting centrifugal force will be 600 times the force of gravity. The majority of the fluids and any sperm cells that escape being centrifuged into the pellet will be in the supernatant. Generally, the supernatant is removed and the sperm cells in the pellet are resuspended for further processing as described below. It is important to maximize the percentage of sperm that are concentrated in the pellet, while at the same time minimizing damage to the sperm cells.

[0321] According to one example, a centrifuge tube containing about 10 × 106 sorted sperm cells is placed in a centrifuge. To facilitate concentration, centrifuge tubes may be used as the collection vessels in the collection system of the cytometer. This avoids the need to transfer the sorted sperm cells to a centrifuge tube before centrifugation. The tube is centrifuged at a speed and for a duration that is sufficient to cause a pellet of concentrated sperm cells to form in the bottom of the tube. The speed and duration of the centrifugation is desirably selected in consideration of several factors, including: the fact that sperm cells are fragile and can be damaged by centrifugation at an excessive speed; the size of the centrifuge tube will affect the time required for sperm cells to move to the bottom of the tube; and the sperm cells are more likely to be damaged by centrifugation at a given speed the longer the centrifugation continues. Thus, in one examplethe centrifuge tube is centrifuged at 550-800g for a period of about 6-10 minutes. According to another example , the centrifuge tube is centrifuged at 650-750g for a period of about 6-10 minutes. In still another example, the centrifuge tube is centrifuged at 700g for a period of about 6-10 minutes. In yet another example, the centrifuge tube is centrifuged at 700g for a period of about 7 minutes.

[0322] As demonstrated in the following experiments, the speed of the centrifuge and the duration of centrifugation may affect the percentage of sperm cells recovered and the motility of the recovered sperm cells. The experiments were conducted without actually sorting the sperm cells. Instead, various fluids including buffers, extenders, sheath fluids and a staining fluid were added to semen samples to simulate the sorting process. The samples were then centrifuged in an attempt to concentrate the sperm cells.

Centrifuge Example 1



[0323] In centrifuge example I bovine semen was collected and evaluated as described above. The semen sample was diluted with a quantity of Tris-citric acid ("TCA") having a pH of 7.3 to attain a concentration of 150 × 106 sperm cells/ml. Spermatozoa were stained with Hoechst 33342 (100, µM) at 41 ° C for twenty minutes. Two 15 ml tubes were prepared with buffers for the simulation. Tube 1 was partially filled with 750 µl of phosphate buffered saline ("PBS") with 10% egg yolk and 14.25 ml PBS with 0.1 % bovine serum albumin ("BSA"). Tube 2 was partially filled with 750 µl TCA with 10% egg yolk and 14.25 ml PBS with 0. 1% BSA. Each of the two tubes received 100 ul of the solution containing the stained spermatozoa, which were then incubated at room temperature for 20 minutes. The two tubes were then divided into two aliquots of 7 ml each. One aliquot from each tube was centrifuged at 2250 rpm (about 540g) for 7 minutes in a fixed bucket centrifuge. The other aliquot from each of the two tubes was centrifuged at 2500 rpm (about 660 g) for 7 minutes. Immediately after centrifugation, 10 ml pipettes were used to remove and save the supernatant from each aliquot. The pellets were resuspended in 200 ul of TCA with 10% egg yolk (pH 7.0). Pre-and post-centrifuge sperm motility was observed under a phase contrast microscope. Fifty ul of a fixative (0.1 % glutarldehyde in 3.4% Na citrate) was added to each pellet and supernatant to immobilize the sperm for concentration determination with a hemacytometer. Total numbers of spermatozoa were calculated on the basis of volume used/recovered multiplied by the corresponding sperm concentration as determined by the hemacytometer. The recovery rate was calculated as the total number of sperm in the pellet divided by the sum of the total number of sperm in the pellet and the total number of sperm in the supernatant.

[0324] The results, as shown in Figs. 88 and 89 show there is little difference in sperm cell motility caused by varying the centrifuge speed. The results also show that motility was slightly better using TCA compared to PBS.

Centrifuge Example II



[0325] In centrifuge example II semen samples from three bulls were collected and evaluated as described above. One of the samples was disqualified for failure to meet initial quality control standards. The other two semen samples were diluted with a quantity of TCA having a pH of 7.3 in order to obtain a sperm concentration of 150 × 106 sperm/ml. The spermatozoa were stained with a 10µM Hoechst 33342 solution at 41°C for twenty minutes. A simulated buffer containing 1500 µl PBS with 10% egg yolk and 28.3 ml PBS with 0.1% BSA was added to each of two tubes. Two hundred µl of the stained spermatozoa (30 × 106 sperm cells) were added to each tube and incubated at room temperature for twenty minutes. Three 9 ml aliquots of semen mixture were taken from each of the two tubes for centrifugation. One aliquot from each of the two samples was centrifuged for seven minutes in a 15 ml centrifuge tube at each of the following speeds: 550g; 650g; and 750g. The temperature during centrifugation was 22°C. Immediately after centrifugation, supernatant was removed with a 10 ml pipette, leaving about 200-300 µl supernatant in the pellet. The pellets were resuspended with 200 µl of TCA having 10% (v/v) egg yolk having a pH of 7.0. Pre-and post-sort sperm motility was observed under a phase contrast microscope. Severe sperm agglutination was noted in the post-centrifuge samples from one of the two bulls. Fifty µl of a fixative (0.1 % glutardehyde in 3.4% Na citrate) was added to each supernatant and pellet to immobilize the sperm for concentration determination. Recovery rate was determined according to the formula set forth in centrifuge experiment 1.

[0326] The results are shown in Fig. 90. The results show improved recovery rate of sperm cells at 650g compared to 550g. However, there was little difference in recovery rate between 650g and 750g. There was no significant difference in sperm cell motility caused by varying the speed of the centrifuge.

Centrifuge Example III



[0327] For centrifuge example III, the procedure of centrifuge example II was substantially repeated with the same three bulls on a different day. The results are shown in Fig. 91. The results confirm that there is little difference in the recovery rate at 650g compared to 750g.

Centrifuge Example IV



[0328] Semen was collected from two different bulls on two different days. Semen was transported and evaluated in the manner described above. Based on sperm concentration of raw semen, spermatozoa were diluted with Tris-citric acid (TCA, pH 7.3) plus 10mM pyruvate, to a concentration of 150×106 sperm/ml. The spermatozoa were stained with 10 µM Hoechst 33342 at 41 °C for 20 min. After staining, 267 µl of the solution containing the stained spermatozoa were diluted to a concentration of 1×106 sperm/ml by addition of the following simulated buffers: 2 ml PBS with 10% (v/v) egg yolk; and 37.733 ml PBS with 0.1% (w/v) bovine serum albumin (BSA). The stained spermatozoa and simulated buffers were incubated at room temperature for at least 20 minutes. Four 9 ml aliquots were taken from the stained spermatozoa and simulated buffer mixture obtained from each bull. The four aliquots from the first bull were centrifuged at varying combinations of centrifuge speed and duration in the following sequence:
  1. (1) 700g for 7 minutes for the first aliquot;
  2. (2) 700g for 10 minutes for the second aliquot;
  3. (3) 650g for 10 minutes for the third aliquot; and
  4. (4) 650g for 7 minutes for the fourth aliquot.


[0329] The four aliquots from the second bull were centrifuged at varying combinations of centrifuge speed and duration in the following sequence:
  1. (1) 700g for 10 minutes for the first aliquot;
  2. (2) 700g for 7 minutes for the second aliquot;
  3. (3) 650g for 10 minutes for the third aliquot; and
  4. (4) 650g for 7 minutes for the fourth aliquot.


[0330] All centrifugation was performed in 15 ml centrifuge tubes in a swing head centrifuge (Allegra 6R, Beckman Coulter Inc. Fullerton, CA) at 22°C. The time interval between semen collection at farm and centrifugation in lab was 4-5 hours. Immediately after centrifugation, supernatant was removed with 10 ml pipettes, leaving -250 µl supernatant with each pellet. The pellets were resuspended in 250 µl of Delbecco's PBS (pH 7.0). Sperm motility and progressive motility were observed using a Hamilton-Thorn Motility Analyzer (two slides per sample; two chambers per slide) after staining but before centrifugation and again after centrifugation. Sperm concentration was determined by hemacytometer measurement of a 100 µl aliquot of the pre-centrifuge stained spermatozoa and simulated buffer mixture that had been placed in the freezer and a 10 µl aliquot of the resuspended pellet mixed with 90 µl fixative (0.1 % glutaraldehyde in 3.4% Na citrate). Recovery rate was determined as in Centrifuge Sample 1. The results are shown in Figs. 92 and 93.

[0331] The data indicate that > 85% of the spermatozoa can be recovered after centrifugation at 650g or 700g, for 7 or 10 minutes (Fig. 92). However, recovery rate was slightly better (95%) at 700g. The decline in motility after centrifugation (compared to before centrifugation) in all treatments could be due to the presence of dead/abnormal/fragile spermatozoa which could not withstand the stress of centrifugal force. Sperm motility declined by 10-14% (Fig. 93) in all treatments. The higher decline in sperm motility (14%) at 650g for 7 min might be due to the longer exposure of sperm to simulated buffer as centrifugation at 650g was conducted after 700g. Centrifugation did not show any adverse effect on progressive motility of spermatozoa, rather there was improvement by 2-3%.

Centrifuge Example V



[0332] Semen was collected from one bull on two different days. Semen was evaluated, diluted and stained with Hoechst 33342, and further diluted in simulated buffers as described in Centrifuge Example IV. Four 9 ml aliquots of the stained spermatozoa and simulated buffer mixture were obtained for each of the two semen samples. The aliquots from the first sample were centrifuged at one of the following combinations of centrifuge speed and duration in the following sequence:
  1. (1) 750g for 10 minutes for the first aliquot;
  2. (2) 750g for 7 minutes for the second aliquot;
  3. (3) 700g for 10 minutes for the third aliquot; and
  4. (4) 700g for 7 minutes for the fourth aliquot.


[0333] For the aliquots obtained from the second sample, the combinations of centrifuge speed and duration were the same, but the sequence was modified as follows:
  1. (1) 750g for 7 minutes for the first aliquot;
  2. (2) 750g for 10 minutes for the second aliquot;
  3. (3) 700g for 7 minutes for the third aliquot; and
  4. (4) 700g for 10 minutes for the fourth aliquot.


[0334] Centrifugation was conducted in a 15 ml centrifuge tube in a swing head centrifuge (Allegra 6R, Beckman Coulter Inc. Fullerton, CA) at 22°C. The interval between semen collection at farm and centrifugation in laboratory was about 6.5 hours for the first sample, and about 4 hours for the second sample. Post centrifugation processing, i.e. removal of supernatant, resuspension of pellet, determination of sperm concentration, and motility estimation via Hamilton-Thorn Motility Analyzer, were conducted following the same procedure as described in Example IV. The results are shown in Figs. 94 and 95.

[0335] The results show that > 85% of the sperm population in highly diluted suspension can be recovered with 700g or 750g in 7 minutes or 10 minutes (Fig. 94). An increase in g force to 750g did not improve the recovery rate significantly. As was the case in Centrifuge Example IV, the decline in motility after centrifugation (as compared to before centrifugation) was observed in all treatments. In the present experiment, sperm motility declined by 13-20% (Fig. 95) which is little higher than in Centrifuge Example IV. The variation could be due to variation in semen sample and longer time interval from semen collection to centrifugation (6 hours) in one replicate. As explained in Example IV, the decline in sperm motility (about 20%) at low speed centrifugation (700xg, for 7 or 10 min) might be due to the longer exposure of sperm to simulated buffer as they were centrifuged after 750g centrifugation. The decline in progressive motility was negligible (1- 5%).

B. Secondary Centrifugation



[0336] In order to recover sperm that might otherwise be lost in the supernatant, it is possible to centrifuge the supernatant after it has been separated from the pellet. Without being bound by a particular theory, applicants believe the pellet/supernatant interphase impedes movement of spermatozoa into the pellet. Removal of the interphase by separating the pellet from the supernatant will allow further centrifugation of the supernatant to cause sperm cells that would have remained in the supernatant to form a second pellet. The second pellet can be resuspended and added to resuspended sperm from the first pellet.

C. Filtration



[0337] An alternative concentration method that may be used to avoid loss of sperm cells in the supernatant is filtration. As shown in Fig. 96, according to an examplet a filter 2415 is incorporated in a collection vessel 2403. The size of the pores in the filter are desirably in the range of about 0.2-1 microns. It is also desirable that the filter is not a depth filter (e.g., a filter having tortuous passages in which sperm tails can be caught). Rather it is desirable that the filter be as thin as possible. For example, it is desirable that the filter thickness be in the range of 50µm to 500µm; more desirable that the filter thickness be in the range of 75µm to 250µm; and most desirable that the filter thickness be in the range of 100µm to 150µm. A low level vacuum 2417 is applied to remove the fluids through the filter as the droplets 33 are collected. It is important to use a low level vacuum (less than 20 inches of mercury, e.g., 15 inches of mercury) to avoid inflicting damage to the sperm cells. In one example the vacuum is low enough that the fluid removal rate is about 1.0 ml/15 seconds. According to another example, the vacuum is applied intermittently to allow the sperm cells a chance to recover. In still another example, the filter 2415 is constructed of a material that is compatible with sperm cells, yet has no binding affinity for them. At the completion of the sort, about 80-90% of the fluids will have been removed through the filter. However, enough fluid remains (about 10-20%) that the sperm cells are in a concentrated slurry 2405, thereby preventing the sperm cells from forming a filter cake. The concentrated suspension may be transferred to another container 2419, as shown in Fig. 97 for example. A syringe mechanism 2409 with a cannula-tip filter 2411 can be used to remove some of the remaining liquid from this container 2419. However, enough fluids are left in the container to prevent the sperm cells from caking on the filter 2411. The same considerations apply to the cannula tip filter 2411 as the filter 2415 in the collection vessel. Thus, the cannula filter 2411 pore size is desirably in the range of about 0.2-1.0 microns and the cannula filter is relatively thin to avoid having sperm tails getting caught in tortuous passages in the filter. For example, a DynaGard® hollow polypropylene fiber syringe tip filter, which is commercially available from Spectrum Laboratories, Inc. of Rancho Dominguez, CA may be used for the cannula tip filter. As shown in Fig. 98, a resuspension fluid 2413 is flushed through the cannula-tip filter to wash cells that may be sticking to the filter surface back into the slurry. The resuspension fluid may include a quantity of the filtered fluid and/or a suitable extender. After a quantity of resuspension fluid sufficient to remove sperm cells from the filter has been back flushed through the filter, additional resuspension fluid may be added if desired. The total quantity of resuspension fluid is selected to bring the concentration to a desired concentration (e.g., about 20 × 106 sperm cells/ml). Thus, the filtration process of this example is a three-step process involving the use of a filter in the collection vessel, filtration using a cannula filter, and resuspension to obtain the desired concentration.

[0338] In an alternative two-step filtration process, the first and second steps of the three-step process described above are combined so that removal of all fluid is through a cannula filter. In this process the sorted sperm cells are directed to a collection vessel that does not have a filter. The fluids are removed by low vacuum and/or intermittent vacuum as described above that is applied through the cannula-tip filter 2411. When the sperm cells are in a concentrated slurry, a resuspension fluid, such as for example, an extender, is flushed back through the cannula filter to obtain the desired concentration of sperm cells.

Filtration Example I



[0339] Filtration example I shows the recovery rate and motility of sperm cells after concentration by a three-step filtration process. Semen samples were collected from three bulls and evaluated as provided in the sample preparation section above. One of the three semen samples was disqualified for failing to meet minimum initial quality criteria. Two remaining samples were diluted with a quantity of TCA (pH 7.3) necessary to attain a concentration of 150 × 106 sperm cells/ml. Five hundred ul PBS with 10% egg yolk and 9.5 ml PBS with 0.1 % BSA was added to each of two 15 ml test tubes. Sixty-seven ul of semen sample (about 10 × 106 sperm cells) was added to each test tube and incubated for twenty minutes at room temperature. Referring to Fig. 99, a vacuum pump 2427 was used to apply negative pressure to draw a four ml aliquot of the diluted semen 2423 through a filter 2425. The filtrate 2429 was collected in a syringe 2421. After filtration sperm cells on the filter were flushed back with 1 ml TCA buffer in a 15 ml tube. Sperm motility was assessed visually. Pre-and post-filtration samples were mixed with a fixative (0.1% glutardehyde in 3.4 Na citrate) to immobilize the sperm cells. Sperm concentration was determined using a hemacytometer. Total number of sperm cells was calculated on the basis of volume multiplied by the concentration of sperm cells. The recovery rate was calculated as the total number of sperm cells in the flushed back portion divided by the total number of sperm cells in the aliquot prior to filtration. The process was repeated with a different filter. The experiment tested both of the following filters: (1) a 1.0, µm PTFE (not FTPE) membrane disc (syringe) filter (which is available from Pall Corporation, Life Science Group, Ann Arbor, MI, Cat # PN4226T or VWR, Batavia, IL, Cat. # 28143-928); and (2) 0.8 SFCA (surfactant free cellulose acetate) membrane disc (syringe) filter (Corning, Inc., Corning, NY, Cat. # 431221; VWR Batavia, IL, Cat. # 28200-028). The results are shown in Fig. 101. More spermatozoa were recovered with cellulose acetate filters as compared to PTFE filter, i.e. 67 vs 33% due to low protein binding affinity of cellulose acetate. Visual motility of spermatozoa recovered ranged from 63% (PTFE) to 68% (Cellulose acetate).

Filtration Example II



[0340] Filtration example II shows the recovery rate and motility of sperm cells after concentration by at two-step filtration process. Semen samples were collected from three bulls and evaluated as provided in the sample preparation section above. The three samples were diluted with a quantity of TCA (pH 7.3) necessary to attain a concentration of 150 × 106 sperm cells/ml. One and one half ml of PBS with 10% egg yolk and 28.3 ml PBS with 0.1 % BSA was added to each of 50 test tubes. Two hundred µl of semen sample (about 30 × 106 sperm cells) was added to each test tube and incubated for twenty minutes at room temperature. Referring to Fig. 100, a syringe 2431 was used to apply negative pressure to draw a 6 ml aliquot of the diluted semen 2433 from each test tube through a filter 2435. The filter was placed in a filter holder 2437 (a Swinnex filter holder from Millipore Corporation, Billerica, MA Cat # SX0002500,). After filtration, the filtration holder 2437 was disconnected from the syringe and the tubing, keeping the filter holder intact. Spermatozoa on the filter were collected by turning the filter assembly upside down and back flushing 1 with ml of TCA buffer using a 3 ml syringe having a small piece of tubing at the tip in a 15 ml test tube. Sperm motility was assessed visually. Pre-and post-filtration samples were mixed with a fixative (0.1 % glutardehyde in 3.4 Na citrate) to immobilize the sperm cells. Sperm concentration was determined using a hemacytometer. Total number of sperm cells and the recovery rate were calculated as specified in Filtration example 1. The process was repeated twice to test different filters. The experiment tested both of the following filters : (1) a 0. 2 µm Teflon membrane filter (which is available from X-Partek, P. J Cobert Associates Inc. St. Louis Cat. # 944106; and (2) a 0.8 cellulose acetate membrane filter (Millipore Corporation, Billerica, MA Cat. #AAWP 02500). The results are shown in Fig. 102. In both filters, the recovery rate of spermatozoa was low (-25%). It was low in Teflon filter as in example 1. However, low recovery rate and visual motility of flushed back spermatozoa in cellulose acetate filter might be due to the material used by different vendor and/or ability of spermatozoa to attach with filter holder/assembly.

D. Dense Medium Concentration



[0341] Another alternative method of concentrating the collected sperm relies on flotation of sperm cells in a high-density medium. According to this method, a high-density medium is added to the collected sperm cells to raise the specific gravity of the suspension above about 1.3. For example, a colloidal silica suspension such as is available under the Percoll® and Isolate® tradenames may be used to increase the specific gravity of the suspension. The sperm cells will float to the top of the suspension, where they can be skimmed or otherwise collected, because of the increased specific gravity of the suspension. A resuspension fluid is added to the cells that have been collected from the surface to bring the final concentration to about 20 × 106 sperm cells/ml. Some of the suspension fluid may be removed by one of the filtration methods described above prior to addition of the high density medium to reduce the quantity of high density medium required to attain the desired specific gravity.

Cryoextension


A. Cryoprotection



[0342] Once the sperm have been sorted and collected in the collection vessels, they may be used for inseminating female mammals. This can occur almost immediately, requiring little additional treatment of the sperm. Likewise, the sperm may also be cooled or frozen for use at a later date. In such instances, the sperm may benefit from additional treatment to minimize the impact upon viability or post-thaw motility as a result of cooling and freezing.

[0343] Generally, a cryoextender comprises a buffer or buffered solution, a protein source, and a cryoprotectant. Examples of buffers and buffered solutions that may be used in the cryoextender are disclosed above with respect to sample collection and extension. Typically, these buffers will be in a concentration of about 0.001M to about 1.0M and have a pH of about 4.5 to about 8.5, preferably of about 7.0.

[0344] If included, a protein source may be added to provide support to the cells and to cushion the contact of the cells with the collection vessel. The protein source may be any protein source that does not interfere with the viability of the sperm cells and is compatible with the particular buffer or buffered solution being used. Examples of common protein sources include milk (including heat homogenized and skim), milk extract, egg yolk, egg yolk extract, soy protein and soy protein extract. Such proteins may be found in a concentration from about 10% (v/v) to about 30% (v/v), preferably from about 10% (v/v) to about 20% (v/v), and more preferably about 20% (v/v). While milk may be used in combination with a buffer or buffered solution, generally milk is used in the absence of the same, as milk is a solution itself that may serve the same purpose of a buffer or buffered solution. In such instances, the cryoextender would contain about 80% (v/v) to about 90% (v/v) milk.

[0345] A cryoprotectant is preferably included in the cryoextender to lessen or prevent cold shock or to maintain fertility of the sperm. Numerous cryoprotectants are known in the art. Selection of a cryoprotectant suitable for use with a given extender may vary, and depends upon the species from which the sperm to be frozen were obtained. Examples of suitable cryoprotectants include, for example, glycerol, dimethyl sulfoxide, ethylene glycol, propylene glycol, trehalose, Triladyl® and combinations thereof. If included, generally, these cryoprotectants are present in the cryoextender in an amount of about 1% (v/v) to about 15% (v/v), preferably in an amount of about 5% (v/v) to about 10% (v/v), more preferably in an amount of about 7% (v/v), and most preferably in an amount of about 6% (v/v).

[0346] In one particular example, the cryoextender comprises water, Triladyl®, egg yolk, and pyruvic acid. In yet another example, the cryoextender comprises 25g Triladyl® 25g egg yolk, and 10mM pyruvic acid in 75mL of water.

[0347] Optionally, the cryoextender may also contain a range of additives that are beneficial to sperm viability or motility and that prevent or lessen the detrimental side effects of cryopreservation. Such additives may include, for example, an energy source, an antibiotic, or a composition which regulates oxidation/reduction reactions intracellularly and/or extracellularly, each of which is discussed above with respect to sample collection and dilution. Such additives may be added to the cryoextender in accordance therewith.

B. Cryopreservation of Sorted Sperm Cells



[0348] In most cases, it will not be possible to use the sperm cells that have been sorted as described above for immediate artificial insemination. Particularly in the case of a commercial sperm sorting operation, the sorted sperm cells must be stored and/or transported before they can be used for artificial insemination. This will usually require cryopreservation of the sperm cells. The sorted sperm may be loaded into elongate cylinders (known as "straws" in the breeding industry) and cryopreserved to preserve the sperm during transportation and storage. Cryopreserved sperm cells can be stored for long periods of time in liquid nitrogen. To use the cryopreserved sperm, the straw may be immersed in a heated water bath to thaw the sperm. Then the straw is loaded into an artificial insemination gun which is used to inseminate a female animal. Several precautions must be taken to protect the sperm cells during cryopreservation. Otherwise the sperm cells will be so damaged (as indicated by a low post-thaw motility rate of 5-10%) that they are not suitable for use in artificial insemination.

[0349] Conventional cryopreservation methods involve sequentially adding a protein source (e.g., egg yolk), cooling the sperm to a temperature of about 4-5°C, adding a cryoprotectant (e.g., glycerol), maintaining the sperm and cryoprotectant at a steady temperature in the range of about 4-5°C for a period of time sufficient to allow the sperm cells to equilibrate with the cryoprotectant, and then supercooling the sperm, as by immersing the sperm cells in liquid nitrogen at-196 °C for storage. Those skilled in the art will recognize that the purpose of the protein source is to protect sperm from damage as they cool from about 14°C to about 8°C, which is the temperature at which sperm cells are most susceptible to cold shock. In contrast, the cryoprotectant protects the sperm cells from damage at temperatures below 0°C. Even though the temperatures involved in cryopreservation are well below freezing and the term "freezing" is sometimes used to describe cryopreservation, those skilled in the art will also know that cryopreserved sperm are not actually frozen. To be precise, the cryopreserved sperm are in a supercooled state. The conventional period during which sperm cells and cryoprotectant are maintained at a steady temperature can last anywhere from 60 minutes to many hours. The overall time to complete cryopreservation using conventional methods generally exceeds four hours. Furthermore, it is believed that up to 50% of the sperm cells are killed in conventional cryopreservation processes. Though sperm are cryopreserved using conventional methods according to some examples, other examples as described herein may employ improved cryopreservation methods to reduce the time required for cryopreservation and/or to improve the health of the cryopreserved sperm.

[0350] Fig. 103 shows a work flow diagram outlining the steps of one exemplary example of an improved method of cryopreserving spermn. At step 2501, the concentration of a solution containing sorted sperm cells is adjusted to be in the range of about 1 million - 40 million sperm/ml, depending on the standard used by the targeted consumer (e.g., breeding association). For example, the sperm concentration may be adjusted to be in the range of about 20 million to 24 million sperm/ml. Adjustment of the sperm concentration may include addition of resuspension fluid, buffers and/or extenders to concentrated sperm as described above. At step 2503, a cryoprotectant (e.g., glycerol) is added before the sperm are cooled. The sperm cells begin equilibrating with the cryoprotectant as soon as they come into contact with the cryoprotectant. At step 2505, a protein source (e.g., egg yolk) is also added to the solution containing the sperm cells as described above.

[0351] The sperm cell solution, protein source, and cryoprotectant are loaded into conventional 0.5 or 0.25 ml artificial insemination straws using a conventional loading machine at step 2507. Those skilled in the art will be familiar with a number of conventional apparatus and techniques that may be used to load semen into straws. For example, United States Patent No. 5,249, 610, issued October 5,1993 to Cassou, et al. and incorporated herein by reference, provides instruction about the filling of straws with bovine semen using a disposable injector nozzle. Moreover, equipment for filling straws is commercially available from Minitube of America, located in Verona Wl. Any of these or similar conventional loading methods and apparatus can be used to load the sorted sperm cells into straws.

[0352] After loading, the sperm cells are cooled to a holding temperature at step 2509. In general, the holding temperature should be selected with the following considerations in mind: holding sperm cells at a temperature that is too high (e.g., 10°C) may cause unnecessary damage from cold shock; equilibration of sperm cells with a cryoprotectant (e.g., glycerol) is believed to be most active at temperatures in the range of 4-5°C; and holding sperm cells at temperatures that are too low (e.g., < 0°) is believed to be damaging to the sperm cells. Thus, according to one example, the holding temperature is in the range of 0-8°C. More desirably, the holding temperature is in the range of 2-6°C. Even more desirably, the holding temperature is in the range of 4-5°C. The cooling rate used for this step 2509 is selected to minimize damage to the sperm cells. The cooling rate may be controlled (e.g., substantially constant) to provide homogenous cooling and to prevent the sperm from suffering temperature shock. The cooling rate should also cool the sperm quickly enough to reduce their metabolism before they incur membrane damage, but slowly enough that they do not suffer from temperature shock. One can control the cooling rate by placing the straws containing the sperm cells in a programmable freezer (e.g., an IceCube 1810CD freezer which is available commercially from Minitube of America, located in Verona, WI) to cool them. According to one example, the programmable freezer cools the sperm from about room temperature (typically in the range of about 22 and 24°C) at a constant cooling rate of 0.1 and 0.3 °C/minute. In one example, the cooling rate is in a range of about 0.15 and 0. 25 °C/min. In another example, the cooling rate is about 0. 2 °C/min. In yet another example, the cooling rate is selected so the sperm are cooled from their initial temperature to the holding temperature in about 90 minutes. In still another example, the cooling rate is selected to cool the sperm from their initial temperature to the holding temperature at a constant cooling rate in about 90 minutes. The cooling rates referred to above actually refer to the rate of the cooling of the chamber of the programmable freezer, but because of the thin walls and long, thin shape of the straw (e.g., about 5.25 inches long, less than 3 mm in diameter, and about 0.15 mm in wall thickness) and the conductive properties of the straw, the temperature difference between the sperm cells and the cooling chamber is not significant.

[0353] After the sperm cells have been cooled to the holding temperature, at step 2511 they are kept at or near that temperature for a period to allow substantial completion of their equilibration with the cryoprotectant. For example, the programmable freezer described above can be programmed to hold the sperm cells at a steady temperature during the period. According to another example , the sperm cells are held at the holding temperature for a period that is shortened compared to conventional methods because the sperm have already been equilibrating with the cryoprotectant during the cooling process. For example, the period may be in the range of about 10 and 60 minutes. In another example, the period is in the range of about 20 and 40 minutes. In yet another example, the period is about 30 minutes. In another yet a further example the period is less than 60 minutes. In yet another further example, the period is less than 40 minutes. The relatively short holding period offers a number of advantages in a commercial sperm sorting process. First, it reduces the time required to process sorted sperm which can translate to cost savings. Also, the sperm cells still perform metabolic processes at temperatures in the range of 0-8°C so reducing the time for which sperm need to be held at this temperature can improve the health of the sperm cells, which will increase the value of the sperm cells to animal breeders who are concerned about artificial insemination success rates.

[0354] After the sperm cells have been held at the holding temperature for a period described above, the sperm cells are cooled at step 2513 to a temperature that approaches the critical temperature zone for sperm cryopreservation. Those skilled in the art will know that the critical temperature zone is the zone at which ice crystal formation and changes in osmotic pressure damage the sperm cells. This temperature may vary depending on the solution in which the sperm cells are cryopreserved, but the critical temperature zone is generally in the range of -18 and -35°C. Sometimes this critical temperature zone is reported to be in the range of about -18 and -30°C. Thus, in an alternativethe cooling rate used to cool the sperm cells from the holding temperature to a temperature that approaches -18°C (e.g., -15°C) is selected to protect the health of the sperm. Relevant factors to consider include that fact that the sperm cells are still equilibrating with the cryoprotectant during this period, the fact that sperm are still performing some metabolic functions, and the fact that the sperm are still somewhat sensitive to rapid temperature change. Again, it is desirable that the cooling rate be a controlled rate, such as a rate that may be programmed into the programmable freezer described above. More desirably, the cooling rate used to cool the sperm from the holding temperature to a temperature that approaches about -18°C is a constant cooling rate. Thus, according to another example, the sperm cells are cooled from the holding temperature to about -15°C at a cooling rate in the range of about 1.0 - 5.0 °C/min. In yet another example, the cooling rate is in the range of about 2.0 - 4.0 °C/min. In yet a further example, the cooling rate is about 3.0 °C/min.

[0355] Step 2515 involves rapidly cooling the sperm cells through the critical temperature zone to limit the time sperm cells dwell therein. Thus, according to one example, the cooling rate through the critical temperature zone about (e.g., -18°C to about -30°C) is selected to be much faster than the cooling rate used to cool sperm cells to the holding temperature and the cooling rate used to cool sperm cells to the temperature approaching the critical temperature zone. Thus, the steeper cooling rate is desirably in the range of from about 8 - 40°C per minute. More desirably, the steeper cooling rate is in the range of from about 8-12°C per minute. Most desirably, the steeper cooling rate is about 10°C per minute. The temperature range over which the steeper cooling rate is used may extend beyond the critical temperature zone. Thus, in yet another example, the sperm cells are cooled at one of the steeper cooling rates described above from about -15°C to about -40°C. In still another example, the sperm cells are cooled at one of the steeper cooling rates described above from about -15°C to about -80°C. The step of cooling the sperm through the critical temperature zone at a steeper rate may be accomplished in the programmable freezer described above.

[0356] After the sperm cells have been cooled below the critical temperature zone (e.g., to - 80°C), the straws containing the sorted sperm are immersed in liquid nitrogen (-196°C) at step 2517 to provide maximum useful life of the sorted sperm cells. The use of liquid nitrogen to store cryopreserved sperm is widespread in the animal breeding industry in the context of unsorted sperm. Thus, those skilled in the art will be familiar with technologies involving the transportation and storage of sperm in liquid nitrogen, which need not be discussed in great detail herein. It is sufficient to note that conventional containers are available to provide for long term storage of bulk quantities of artificial insemination straws in liquid nitrogen and that smaller and more portable containers are also available for providing storage of artificial insemination straws in liquid nitrogen for transport to customers and/or for transport to a farm having one or more female animals to be inseminated with cryopreserved sperm.

[0357] One advantage of the cryopreservation methods described herein is that the cryopreservation can be completed in less time than is required according to conventional methods. Perhaps relatedly, the decline in motility due to cryopreservationmay only about 5-11%, as indicated by the example discussed below. Thus, cryopreservation as described herein preserves the health of the sperm cells as indicated by tests showing that sperm cells cryopreserved as described herein have greater than 50% (e.g., about 60%) motility after they are thawed in a 37°C water bath for about 50 seconds. As discussed above, sperm motility may be analyzed by an automatic machine (e.g., the IVOS sperm analyzer from Hamilton Thorn Research) or by visual examination.

[0358] It should be noted that the cryopreservation methods described above are contemplated as being used in a commercial scale sperm sorting process. Thus, according to one example, the steps of the inventive methods described herein are performed simultaneously on a batch of sorted sperm cells to quickly cryopreserve the entire batch of sperm cells in a manner that preserves their health. For example, by using the multi-channel flow cytometry apparatus described below, it is possible to obtain about 840 × 106 sorted X chromosome-bearing sperm cells in the collection system of the apparatus in about 20 minutes. This is enough sperm cells to fill several dozen straws. Moreover, a batch can include the combined sperm cells by two or more different sorting cytometers. After being concentrated as described above, the sperm cells can be loaded into any number of straws and cryopreserved as a batch. For example, it takes about 5 minutes to add an extender (including both a protein source and a cryoprotectant) to a batch of sperm cells, and about 15 minutes to load the sperm cells into artificial insemination straws using an automatic loading machine. All the straws in the batch are cooled simultaneously in a programmable freezer. Furthermore, the capacity of some programmable freezers allows simultaneous cryopreservation of thousands of artificial insemination straws. For example, the IceCube 1810CD freezer referred to above has the capacity to cryopreserve simultaneously over 2,500 0.5 ml straws or over 3,800 0.25 ml straws. Thus, one could wait to start the cooling step until multiple batches have been obtained. Alternatively, multiple batches could be obtained substantially at the same time by running multiple multi-channel flow cytometry machines (see below) in parallel and simultaneously cooling multiple batches obtained therefrom together in a programmable freezer. In example, it takes a period of less than 220 minutes to cool the sperm cells from room temperature to a supercooled state and immerse them in liquid nitrogen (-196°C). In another example, the supercooling period is less than 190 minutes. In still another example, the supercooling period is less than 150 minutes.

[0359] Those skilled in the art will recognize that substantial modifications may be made to the foregoing exemplary methods. For example, the sperm cells may be cryopreserved in a container other than an artificial insemination straw. Likewise, the steps in the method that involve changing or maintaining temperature may be performed by any suitable means, including water baths, liquid nitrogen vapors, and conventional programmable or non-programmable freezers, for example. Furthermore, a wide variety of substances or combinations of substances could be used as the protein source and/or the cryoprotectant. These substances include substances and concentrations of substances listed above in connection with the discussions regarding buffers, extenders, cryoprotectants, sheath fluids, and collection fluids. Moreover, the order of some steps in the method may be varied. Although Fig. 95 indicates that the cryoprotectant is added after the concentration of the sorted sperm is adjusted, it is also contemplated that a cryoprotectant can be added before the concentration is adjusted. For example, the cryoprotectant may be provided in the collection fluid or in the sheath fluid used in connection with a flow cytometer. Some of the benefits may also be obtained by partially cooling the sperm cells and then adding the cryoprotectant. Likewise, the order in which the protein source is added may be varied as long as the protein source is effective to protect the sperm cells from cold shock as they pass through the temperature range of about 14 to 8°C.

Cryopreservation Example I



[0360] Bovine semen was collected, transported, and evaluated as described above. Two test tubes containing 5 ml each of TCA buffer (pH 7.3) were placed in one of two water baths for at least five minutes. One water bath was at a temperature of 35°C and the other water bath was at 41 °C. Spermatozoa at 24°C were added to each tube so that the final concentration in each tube was 150 x 106 sperm/ml. The two tubes were each divided into two aliquots which were kept in respective water baths. After the sperm had equilibrated with the TCA buffer for five minutes, 80µM Hoechst 33342 was added to one of 35°C aliquots and one of the 41 °C aliquots. After addition of the Hoechst 33342, all four aliquots were incubated for 20 minutes in their respective water bath. After incubation, the test tubes were removed from the water baths and left at room temperature (about 25°C) for five minutes. Then the contents of each test tube were diluted with a TCA extender containing 20% egg yolk and 6% glycerol (v/v) (pH 7.0) to a final concentration of 20 x 106 sperm/ml. The contents of each test tube were then used to fill a 0.5 ml artificial insemination straw. Each of the four straws was placed in a programmable freezer (an IceCube 1810CD freezer from Minitube of America, Wl). The following cooling sequence was programmed into the programmable freezer: (1) 22°C to 4°C @ -0.2 °C/min ; (2) hold at 4°C for 30 min; (3) 4°C to -15°C @ -3.0 °C/min; and (4) -15 °C to -80 °C @ -10.0 °C/min. After reaching -80 °C, the straws were immersed in liquid nitrogen (-196°C) for 45 minutes. Then the straws were immersed in a 37°C water bath for 50 seconds to thaw. Sperm motility was checked under a phase contrast microscope both before and after cryopreservation. The results are shown in Fig. 104. The post-thaw motility was generally on the order of 60%. This represents a decline in motility of only about 5-11% compared to before cryopreservation. Analysis of variance revealed no significant effect of either Hoechst 33342 or incubation at 41 °C on post-thaw sperm motility.

Operation of the System



[0361] The overall operation 813 of the flow cytometry system 9 will now be described with reference to Fig. 82 and in the specific context of sperm cells (e.g., bovine sperm cells), but it will be understood that the description is exemplary only, and that the system can be used to process other types of particles.

[0362] The first series of steps leading up to the six second repeat loop involve calibration of the system. After initializing 769, a system check 771 is performed to confirm, among other things, that the processor 131 or processors are operational. If an error is detected after three failed system checks 775, user interaction 773 is requested. If the system check is positive, the microprocessor directs the system to flush 777 the nozzle system with a suitable fluid, and then a quality control material 779, such as beads or bovine nuclei, are run through the system to initialize the detection parameters (see 739 in Fig. 72) and confirm that the system is operating within an acceptable quality control. This involves an evaluation of the control material to test the sensitivity and precision of the system to confirm that the system can adequately discriminate a sample. If the quality control is not confirmed after three attempts 775, user intervention 773 is requested.

[0363] If the quality control material indicates an acceptable level of quality control, a sample 781 is aspirated and a portion or aliquot of the sample to be sorted is checked for quality 783. Sample quality may be determined by a calculation of a quality factor (Q-factor) of the sample. For example, the type of cells may be detected in a first aliquot of the sample. During this detection, the initialized detection parameters (741) are rechecked and the initial discrimination parameters (745) are generated. If the type of cells detected in the aliquot indicates that the sample meets or exceeds a preset standard (e.g., that the sample can be discriminated to yield a certain purity or motility and, in particular, that there are sufficient live X cells available for processing), then the system continues operation. If sample quality fails three times 775, user interaction is requested.

[0364] Continued operation involves sorting 785 of the remainder of the sample employing a six second repeated loop. At the beginning of the loop, the microprocessor confirms that sorting of the sample is not complete 789. If the sorting of the sample is complete 789, the microprocessor proceeds to aspirate the next sample 781 if it is available or to turn off the sorting operation 793 if additional sample is not available. If the sample is not complete 789, the microprocessor initially checks the X/Y discrimination 795 of the sample to confirm that it is within an optimum range. In other words, drift analysis as noted above (761 in Fig. 72) is conducted. If any changes should be made, such changes are implemented and the discrimination 795 is again checked. If the discrimination is still unacceptable at this point, the sort is turned off 793 and user interaction is requested.

[0365] Otherwise, the system proceeds to determine whether the fluid delivery system is delivering fluid and cells at a rate which is within an optimum range 801. This determination depends on the type of control strategy used. For the high recovery control strategy, the optimum rate would be determined by evaluating purity or looking at x/x+~X of the collected population. If the determined purity is higher than a required purity level, the feed input rate of the cells is increased by increasing a rate control signal provided to the syringe pump 803. This would tend to increase coincident cells and decrease purity because more coincident cells including ~X cells would be collected with the X cells. If the determined purity is lower than the required purity, the feed input rate of the cells is decreased by decreasing a rate control signal provided to the syringe pump to reduce the frequency of coincident cells 803. Thus, the cell input rate is a function of the determined purity of the collected population as compared to a desired purity level, e.g., a function of the identified ~X sperm cells collected.

[0366] For the high purity control strategy, the optimum rate would be determined by calculating lost X cells, e.g., discarded X/discarded X + collected X. If the quantity or percentage of lost X cells are less than an acceptable level, the input rate of the cells is increased by increasing a rate control signal provided to the syringe pump 803. This would tend to increase coincident cells and increase the number of discarded X cells because more cells including X cells would be discarded with the Y cells. If the quantity or percentage of lost X cells is higher than the acceptable level, the input rate of the cells is decreased by decreasing a rate control signal provided to the syringe pump 803 to decrease coincident cells. Thus, the cell input rate is a function of the determined lost X cells of the discarded population as compared to number of X cells in the collected population, e.g., a function of the number of X sperm cells not collected.

[0367] If this modified rate is acceptable 805, the system proceeds to another system check 807. If the system check is acceptable 807, the sort continues in the six second loop. If not, the system is reset 809. If after reset the system is not acceptable or if the revised feed rate is not acceptable 811, the sort is turned off 793 and user intervention is requested 773.

[0368] The sorted droplet streams are collected by the collection system 2201. Droplets that are sorted into the population of X cells pass through the exit window 2245 in the first intercepting device 2247 to be intercepted by the second intercepting device 2249. From there, the droplets containing the X cells flow into a collection vessel 2207. Other droplets are intercepted by the first intercepting device 2247 and directed to the waste trough 2805. Of course droplets intercepted by the first intercepting device could also be saved, as noted above. When a suitable amount of X-bearing sperm cells have been collected in the collection vessel, sorting may be interrupted to allow concentration of sperm cells in the collection vessel 2207. A new collection vessel may be placed under the first intercepting device 2247 or the collected fluid may be poured into a different container and the collection vessel replaced. Then sorting may resume. The sperm cells in the collected fluid are concentrated, loaded in straws, and frozen as described above.

Temperature Control During Operation



[0369] Temperature control throughout the process may be used to improve the results of the process. As has already been discussed above, the temperature of the sperm may be controlled during various steps in the process (e.g., staining and cryopreservation).

[0370] For example, Fig. 105 is a work flow diagram of one example of a possible method of temperature control. The temperature of semen samples at the time they are collected will be determined by the body temperature of the animal from which they are collected. For example, at step 2601 bovine semen samples are collected at about 37°C. An insulated container is used for transportation of the semen samples to the lab from the collection site at step 2603. The insulated container retards cooling of the sperm.

[0371] During sample evaluation at step 2605, the temperature is maintained below the collection temperature, but in excess of a temperature corresponding to a glass transition temperature below which the sperm cells suffer membrane damage. For example the temperature may be maintained in the range of about 18-37 °Cor the temperature may be maintained in the range of about 24-37°C during sample evaluation. The sperm cells may be placed in an environment having a temperature in the range of about 22-25°C during sample evaluation. Depending on the temperature of the sperm upon arrival at the lab, the effect of placing them in an environment having a temperature in the range of about 22-25°C may be to continue slow cooling of the sperm, to maintain the temperature of the sperm, or to slightly raise the temperature of the sperm. The temperature may be elevated (e.g., to 40°C or higher) for staining at step 2607The temperature of the sperm cells during the staining step may be in the range of about 20-40°C, as is also discussed above.

[0372] At step 2609, the stained semen mixture is held in a water bath until such time that the mixture is introduced into a flow cytometer. The temperature of the water bath may be similar to the temperature used for the staining step. The temperature of the water bath may be in the range of about 40-47°C, or about 20-37°C. The temperature of the water bath may be in the range of about 20-25°C. After being held in the water bath for any time between one minute and two hours, the stained sperms cells are sorted by flow cytometry as discussed above at step 2611. At step 2613, the collected sperm cells are concentrated. Concentration may be performed in an environment that has a temperature that will not significantly change the temperature of the sperm cells. For example, concentration may be performed in an environment having a temperature in the range of about 20 and 25°C. An extender, protein source, and cryoprotectant are added to the concentrated sperm at step 2615. Then, at step 2617 the sperm cells are loaded into artificial insemination straws. The loading step may be performed in an environment having a temperature that will not significantly change the temperature of the sperm cells. Finally, at step 2619 the temperature of the sperm is controlled during cryopreservation as discussed above.

[0373] The sperm cells may be stained at still lower temperatures . For example, it may be desired to sort the sperm cells in a flow cytometer at a relatively low temperature (e.g., about 0°C to about 8°C). This may require modification of the overall temperature control. First, when cooling the sperm cells prior to introduction into a flow cytometer, egg yolk and other common protein sources that protect the sperm cells from cold shock at temperatures below the glass transition temperature generally may not be used as such protein-containing substances tend to foul and/or clog the fluidics of the flow cytometer. Thus, it is desirable to cool the sperm cells before performing the staining step in order to take advantage of natural cold shock protectants found in neat semen, such as for example, the seminal fluid. Any attempt to stain the sperm cells prior to cooling would require addition of buffers to protect the sperm which would dilute the neat semen and reduce the natural protection against cold shock.

[0374] Accordingly, for sorting the sperm cells at a temperature in the range of about 0°C to about 8°C includes placing the sperm cells in an environment having a temperature less than about 8°C to cool the sperm cells to a temperature in the range of about 0°C to about 8°C prior to staining. Any method may be used to cool the sperm cells, but it is desirable to use a method that protects against rapid temperature fluctuations of the sperm cells during the cooling process. For example, a container holding the sperm cells may be placed in a room temperature water bath, which in turn is placed in an environment having a temperature less than about 8°C. The temperature of the sperm cells may be monitored and ice is added to the water bath to further cool the sperm cells. The staining step may be performed as described above except that the staining mixture is subjected to a temperature in the range of about 0°C to about 8°C. Due to the lower temperature, the incubation period required to stain the cells is considerably longer. Once the sperm cells have been cooled to 8°C or below, it is desirable to avoid warming them. Thus, it might also be possible to operate the flow cytometer in an environment having a temperature in the range of about 0°C to about 8°C. Similarly, it might also be possible to collect the sorted sperm cells in a collection vessel that is surrounded by an environment having a temperature in the range of about 0°C to about 8°C. Moreover, any extenders, cryoprotectants, buffers, protein sources, antibiotics, antioxidants, or other additives may be added at a temperature in the range of about 0°C to about 8°C. With respect to addition of the cryoprotectant, it may be desirable to add slightly more of the cryoprotectant than would be added absent sorting the sperm cells at a temperature in the range of about 0°C to about 8°C. Thus, a cryoprotectant containing 7% glycerol (v/v) may be added to sperm cells after the sperm cells have been sorted at a temperature in the range of about 0°C to about 8°C.

[0375] Supercooling of the sperm cells from the temperature in the range of about 0°C to about 8°C proceeds generally as described in the cryopreservation section above. However, the sperm cells will need to be held at a temperature in the range of about 0°C to about 8°C for a period of time after addition of the cryoprotectant before supercooling to allow time for the sperm cells to equilibrate with the cryoprotectant. Thus, the sperm cells are allowed to equilibrate with the cryoprotectant for a period in the range of about 30 minutes to about 3 hours. Alternatively, the sperm cells are allowed to equilibrate with the cryoprotectant for a period in the range of 1-2 hoursor the sperm cells are allowed to equilibrate with the cryoprotectant for a period of about 90 minutes.

[0376] Conventional temperature control apparatus and methods (e.g., water baths, incubators, coolers, and freezers) may be used to heat or cool the sample to attain or maintain the specified temperatures in the foregoing examples. It is understood that placing a sample in an environment having a different temperature than the sample, will cause the temperature of the sample to change over time. There may even be temperature variations within the sample. As has been mentioned, it is desirable to change the temperature of the sample gradually to help maintain the health of the sperm. Gradual temperature change also serves to reduce the temperature variation within the sample. As is well known by those skilled in the art, the rate of temperature change of the sample will be influenced by many factors, including the volume of the sample, the size and shape of the sample container, and the magnitude of the temperature difference between the sample and the environment. However, those skilled in the art will readily be able to select an appropriate method and apparatus to achieve the desired temperature control after considering all the relevant factors.

[0377] The time at which sperm cells remain at any temperature can also impact the health of the sperm. Thus, in processing the temperature may be controlled throughout the process is desirably completed within a timeline as discussed below.

Timeline for Operation



[0378] Generally, it may be desirable to complete the sperm sorting process in the least amount of time possible to reduce the damage to the sperm. The improved staining method described may reduce the time require for staining to about 10 minutes. Likewise, the novel cytometer described above may be used to sort sperm cells in less time than would be required by a conventional cytometer. For example, a flow cytometer using the technology discussed above can collect between 2,000 and 10,000 sperm cells having a desired DNA characteristic per second. Furthermore, the cryopreservation process may be used to reduce the time needed to complete cryopreservation of the processed sperm cells compared to conventional cryopreservation methods. Accordingly, one example involves processing sperm pursuant to an overall method to take advantage of one or more of the timesaving innovations to reduce the time required to complete the entire process. For example, a batch of sperm cells (e.g., an ejaculate) is collected from a male mammal (e.g., bull), evaluated for quality control, stained, sorted according to a specified DNA characteristic, loaded into one or more containers (e.g., straws), and cryopreserved within a period of about 12 hours from the time of collection is less than about 8 hours, less than about 6 hours,less than about 3 hours,less than about 2 hours, or less than about 1 hour.

MULTI-CHANNEL SORTING APPARATUS AND METHOD



[0379] In order to sort more sperm in less time, it is possible to use more than one cytometry unit in parallel to sort that same sperm sample. One way to do this is to simply divide the stained sperm cells into multiple aliquots and run each aliquot through a different cytometer. However, as will be discussed below, certain advantages may be obtained by designing an apparatus that comprises multiple cytometry units in a single integrated multi-channel cytometry unit.

Multi-Channel System Sharing Integrated Platform



[0380] Figs. 106-116 show one example comprising a multi-channel cytometry system, generally designated 1001, where multiple single-channel flow cytometry units, designated 1003, are ganged together as an integrated system to produce sorted product. Four such units are illustrated in this particular example, but this number can vary. The units may be integrated in various ways, as by sharing an integrated platform comprising one or more of the following elements (1) a common supply of particles 1005; (2) a common source of electromagnetic radiation 1007; (3) a common housing 1009; (4) a common input for controlling operation of the units 1011; (5) a common output 1019 allowing evaluation of the operation of one unit relative to another unit; (6) a common fluid delivery system 1021; (7) a common temperature control system 1023; (7) a common power source 1025; (8) a common waste recovery system 1027; (9) a common deflector plate system 1029; and (9) a common cleaning system 1031. The use of common elements is beneficial because it allows the system to be run more efficiently and profitably, achieves more consistent results among channels by reducing the number of variables, facilitates any trouble-shooting that may be needed, and is economical. The multi-channel approach also makes the sorting system more amenable to scale up or scale- down.

[0381] Each of the cytometry units 1003 has components similar to certain components of the flow cytometry apparatus 9 of the previous example and, for convenience, corresponding parts are designated by the same reference numbers with the addition of a prime ('). In general, each unit comprises a nozzle system 101', a mount for mounting the nozzle system 331', a transducer 105', and an epi-illumination optics instrument 417' for focusing a beam of light 25' on the fluid stream 21' exiting the nozzle orifice 103', all as previously described. Each unit further comprises a photodetector 117' operable as in the first example to detect fluorescence emissions 31' from the particles in the stream 21' and to convert the emissions 3 1' to electrical signals 701' which are processed to classify the particles by a specified DNA characteristic. Each unit 1003 is also equipped for sorting the droplets 33' into different groups or populations 123', 125' according to the classification of particles contained in the droplets 35'. The populations of droplets sorted by the units are collected by the collection system 2201.

A. Common Housing and Modularity



[0382] The flow cytometry units are mounted in a modular arrangement in a common housing 1009. In the example shown in Figs. 106 and 109-113, the housing has a base 1069 and two side walls 1071 extending up from the base. The side walls have a lower pair of shoulders 1073 for supporting a lower cover panel 1075 at the front of the housing 1077, and an upper pair of shoulders 1081. A lower cover panel 1075 at the front of the housing 1077 is mounted between the lower shoulders 1073. The upper shoulders 1081 support an upper cover panel 1083 at the rear of the housing 1085. The front and rear of the housing 1077, 1085 are substantially open to provide access to the equipment inside. It will be understood that the housing 1009 may have other configurations. Further, it will be understood that the various units could be installed in separate housings.

[0383] The flow cytometry units 1003 are mounted side-by-side as modules on an appropriate framework 1087 in the housing 1009. Specifically, the nozzle mounts 331' for positioning the nozzles 101' are releasably attached to a cross bar 1089 (Fig. 106) affixed to the side walls 1071 of the housing, and the bases 429' of the epi-illumination instruments 417' are releasably fastened to an angled mounting plate 1093 extending between the side walls 1071 of the housing toward the rear of the housing 1085 (Fig. 109), the arrangement being such that a particular unit can be installed or removed as a module. This modularity facilitates installation, removal for maintenance and/or replacement, and enables any number of flow cytometry units 1003 to be readily added as needed or desired to increase the throughput capacity of the system.

B. Common Fluid Supply and Delivery Systems



[0384] The fluid delivery system 1021 of this example is equipped to provide appropriate fluids to each of the cytometry units 1003. As illustrated schematically in Fig. 108, the system generally comprises a pump 1105 for conveying carrier fluid 17' from a common supply of carrier fluid 1107 under pressure, a gas pressure system 1115 for conveying fluid from a common supply 1117 of sheath fluid 19' under pressure, and a manifold system 1121 for receiving the fluids from respective supplies and delivering the fluids under pressure to the various cytometry units 1003, as needed. In the specific example of Fig. 116, the supply of carrier fluid comprises a vessel 1123 containing a suitable volume of such fluid (e.g., 5 ml.). The vessel is held by a holder 1125, which may be a block 1133 having a cavity 1135 sized to receive the vessel 1123. The block also has a second cavity 1137 for holding a vessel 1139 containing a suitable buffer material for conditioning the system during use, as will be described later.

[0385] The pump 1105 for delivering carrier fluid from the vessel is desirably (but not necessarily) a syringe pump 1141 as previously described. The plunger of the pump is movable through an intake stroke to aspirate a selected volume of carrier fluid 17' from the vessel 1139 and through a discharge stroke to dispense carrier fluid through a supply line 1147 to the manifold 1177 and from there to the various nozzles 101' of the system. The syringe pump is also operable to aspirate fluid from the vessel 1139 containing buffer and to pump the buffer through the system in a manner to be described. A three-way valve 1149 controls the flow of carrier and buffer fluids to and from the pump 1141. The pump is driven by a variable speed motor under the control of the processor 131'. By way of example, the pump may be driven by a stepper motor which operates at selectively variable speeds to pump carrier fluid to the manifold system 1121 at rates necessary to obtain the desired throughput from the units 1003. Multiple syringe pumps or other types of fluid delivery devices can be used instead of a single syringe pump.

[0386] The supply 1117 of sheath fluid may comprise a vessel 1155, e.g., a tank connected to the manifold 1177 by means of a supply line 1157. The gas pressure system 1115 is operable to pressurize the tank and comprises a source of pressurized gas 1161 (e.g., air or nitrogen) communicating with the tank via a gas line 1163 having a regulator 1165 in it for controlling the pressure supplied to the tank, and a two-way valve 1167 which, in a first position, establishes communication between the tank and the gas source, and in a second position, is operable to vent the tank. The gas pressure regulator 1165 is a conventional regulator adjustable to control the pressure supplied from the air source. The gas pressure system 1115 also includes a gas line 1169 for pressurizing a supply 1173 of cleaning solution (e.g., de-ionized water in a tank) which can be used to flush the fluid circuitry in a manner to be described hereinafter. Flow through the gas line is controlled by a two-way valve 1167 operable in the same manner as valve 1167.

[0387] The manifold 1177 may comprise a laminated block 1179 (Fig. 112) of material having passages 1181 formed in it to define a fluid flow circuit 1185 such as that shown diagrammatically in Fig. 116. (The passages may be formed by machining grooves in faces of the laminations prior to assembly of the laminations to form the block.) The fluid circuit includes inlets 1189, 1191 connected to the syringe pump 1141 and to the supply 1117 of sheath fluid, and sets of outlets 1193, for providing such fluids to the flow cytometry units 1003, each such set including a carrier fluid outlet and a sheath fluid outlet. Flow through the various flow passages 1181 is controlled by valves VI-V6 which, in one example, are solenoid-operated valves in housings attached to the manifold block 1179. The block is desirably of substantially transparent material (e.g., acrylic plastic) to facilitate monitoring of the system 1121 and trouble-shooting. The manifold 1177 is attached to a frame member 1203 extending between the side walls 1071 of the housing 1009 adjacent the bottom of the housing below the nozzle systems 101'. The inlets and outlets 1193, of the manifold 1177 may comprise fittings 1205 threaded into the block, such as flangeless nut and ferrule fittings available from Upchurch Scientific, a division of Scivex. It will be understood that the design and construction of the fluid circuit 1185 in general and the manifold 1177 in particular may vary .

[0388] Referring to Fig. 116, the manifold fluid circuit 1185 for the carrier fluid 17' includes a sample reservoir 1207 for holding a limited supply of carrier fluid (e.g., 1.0 ml). If the carrier fluid contains sperm cells, for example, providing such a reservoir close to the nozzles 101' is beneficial to the viability and motility of the sperm cells, since the storage of such cells, even for short periods of time, in small spaces may be detrimental to motility. Flow of carrier fluid from the sample reservoir 1207 to the nozzles 101' is controlled by a series of two-way valves V1A-V1D, one for each nozzle. Each valve V1A-V1D has a first position establishing fluid communication between the needle 1217 of the sample reservoir and the needle 157' of a respective nozzle for delivery of carrier fluid 17' to the needle under pressure generated by the syringe pump 1141, and a second position establishing fluid communication between the needle 1217 and a waste system, generally designated 1221, which is common to all of the flow cytometry units 1003. In the example shown, the waste system 1221 comprises a waste tank 1223 for holding waste material, a mechanism 1225 such as a vacuum pump for generating a vacuum in the waste tank, and waste lines 1227 connecting the valves V1A-V1D and the waste tank. A valve V3 is provided in the waste line upstream from the waste tank for opening and closing the waste line, as needed. A suitable hydrophobic filter 1233 is provided in the line connecting the waste tank 1223 and the vacuum pump 1225.

[0389] The manifold fluid circuit 1185 for the sheath fluid 19' includes a plurality of valves V2A- V2D. Each valve has a first position establishing fluid communication with the supply 1117 of sheath fluid in the tank for delivery of sheath fluid 19' to a respective flow body 133' via a sheath supply line 1241, and a second position establishing fluid communication between the flow body and the waste tank via a waste line 1247. The pressure at which the sheath fluid is delivered to the flow bodies 133' will depend on the sheath tank pressure (as controlled by the regulator 1165) which may range from 0.07 to 6.9 bar (1 to 100 psi), more desirably from 0.69 - 3.45 bar (10 to 50 psi), even more desirably 1.3 to 2.76 bar (15 to 40 psi), and even more desirably from about 1.58 to 2.07 bar (20 to 30 psi).

[0390] While the use of a common supply for all of the units has various advantages, it is contemplated that at least some of the flow cytometry units could be supplied with sample material from separate sources.

C. Common Power Supply and Input and Output Controls



[0391] The flow cytometry units 1003 also share a common power supply 1025, common power delivery systems, a common input (GUI) 715' for controlling operation of the channels by the microprocessor 131', and a common output provided to the microprocessor allowing evaluation of the operation of one channel relative to another channel. For example, the common output includes providing the digitized signals from each epi-illumination system to the microprocessor for an indication of the fluorescence intensity measured by each channel, for an indication of the rate at which each channel is separating particles, for an indication of the staining variations (which may be indicated by the intensity difference of fluorescence pulses from X and Y cells) and for an indication of the decision boundaries 763 used by each channel for discriminating between particles. As another example, the common output includes providing the output signals from the break-off sensors 389' to the microprocessor for an indication of the droplet break-off location 107' of each channel.

D. Common Temperature Control



[0392] Optionally, a temperature control system, generally designated 1257, is provided to regulate the temperature of the contents of the vessels 1123 in the holding block 1133 and the temperature of the manifold 1177. Such temperature control reduces the variability of the system, thus providing more consistent measurements between channels and, for certain types of cells (e.g., sperm cells), helping to maintain the viability of the cells.

[0393] The temperature control system 1257 may comprise a fluid flow circuit 1259 comprising fluid passages 1263 in the holding block 1133 and fluid passages 1269 in the manifold block 1179, and a control unit 1265 for circulating a thermal fluid (e.g., water) through the circuit at a selected temperature. The temperature is desirably such as to maintain the fluid, especially the carrier fluid, at an optimal temperature to maximize cell viability and, if sperm cells are involved, sperm motility. A valve shut-off V6 is positioned in the circuit for controlling flow through the circuit. The temperature control unit may be used to maintain the sperm cells at the desired temperature prior to sorting as discussed above.

[0394] All of the valves in the fluid delivery system 1021 are operated by conventional means, such as solenoids, under control of an operator or suitable programming. The various fluid flow lines connecting the components of the system outside the manifold block 1179 are desirably of substantially transparent plastic tubing for observing any blockages. For example, the tubing may be 0.0625 in. OD tubing of FEP polymer. The flow lines of the temperature control system 1257 are desirably somewhat larger (e.g., 0.125 in. OD) to provide greater flow capacity.

E. Common Light Beam and Beam Splitting System



[0395] As previously noted, the multi-channel system shares a common source of electromagnetic radiation or beam light 1007. By way of example , the source may be a laser beam from a UV multiline laser primarily having wavelengths of 351.1 nm and 363.8 nm. Alternatively, it may be desirable to use a pulsed laser (e.g., a mode-locked laser), particularly to synchronize digital sampling with a pulsed laser (as discussed in the pulsed laser section) in order to increase the effective power delivered to each cytometry unit, thereby increasing the number of cytometry units that can be operated with a single laser.

[0396] The power required to generate the laser beam will vary depending on the requirements of each flow cytometry unit and the number of units. For example, if there are N units and each unit requires a light beam having an effective power of W watts, then it will be necessary to generate a laser beam having a power of (W x N) + L, where L equals the system power loss among the optical elements of the system. Using a single laser to supply all of the flow cytometry units is economical compared to a system using multiple lasers. It is also efficient and provides for more consistent measurements from one channel to the next, because there is no variability on account of different beam characteristics (e.g., beam intensity, light polarity, beam divergence) or electrical noise resulting from the use of multiple lasers.

[0397] According to oneexample, a beam splitting and guidance system is used to split a single laser beam into three or more separate beams. As shown in Fig. 117, for example, a 50/50 beam splitter 1270 (i.e., a beam splitter that is operable to divide a single beam into two separate beams having approximately equal intensity) can be used to split a single beam 25' into two beams 1270A, 1270B. By using a second 50/50 beam splitter 1271 to split one of the two beams 1270B into two additional beams 1271A, 1271B, one can generate a total of three beams 1270A, 1271A ,1271B from a single beam 25'. Each beam can be directed into the optics system of a flow cytometer, for example an epi-illumination optics system 415' as shown in Fig. 117. One could also use additional 50/50 beam splitters to split the single laser beam into any number of additional separate beams. As shown schematically in Fig. 118, for example, a third beam splitter 1272 can be added to the 3-way beam splitting system (Fig. 117) so that the three 50/50 beam splitters 1270,1271, 1272 can be used to split a single 25' beam into four separate beams 1271A, 1271B, 1272A, 1272B. From this one can readily appreciate that the single beam can be split into any number of separate beams. Other beam splitting arrangements may be used to split the incoming source beam into multiple light beams for the various units.

[0398] A beam splitting system is shown in Figs. 106 and 109. A beam guidance system 1273 is provided for guiding the common beam 1007 to the optics instruments 417' of the various flow cytometry units 1003. In Figs. 106 and 111, the guidance system 1273 includes a lower mirror assembly 1279 mounted on one side wall 1071 of the housing 1009, an upper mirror assembly 1281 mounted on the side wall 1071 above the lower mirror assembly, and a series of reflecting filters 1283, one associated with each optics instrument 417'. The lower mirror assembly is operable to reflect a beam 1007 from a suitable source upwardly to the upper mirror assembly, and the upper mirror assembly is operable to reflect the beam through an opening in the side wall 1071 to the reflecting filters 431' of the various instruments 417'.

[0399] The lower mirror assembly may include a base 1285 fastened to the side wall 1071 of the housing 1009, a stage 1289 movable vertically on the base by a suitable mechanism 1291, such as a micrometer, a tiltable platform 1293 on the stage (e.g., a kinematic optical mount Model P100-P available from Newport), and a mirror 1295 on the platform, the position of the mirror being adjustable by moving the stage and the mirror platform to the appropriate locations. The upper mirror assembly is similar to the lower assembly, comprising a base 1297, a vertically movable stage 1299, a tiltable platform on the stage 1301, and a mirror 1303 on the platform. A pair of target plates 1309 are affixed to the side wall of the housing 1009 between the upper and lower mirror assemblies. The target plates 1309 have vertically aligned holes 1311 therein to facilitate adjustment of the upper and lower mirrors so that an incoming beam 1007 is precisely reflected toward the reflecting filters 431' of the instruments 417', all of which filters are aligned with the incoming beam.

[0400] Each of the first three reflecting filters 1315, 1317, 1319 functions as a beam splitter, i.e., it functions to reflect a specified percentage of the beam and to pass the remaining percentage of the beam. For example, in the case of four epi-illumination instruments, the reflecting filters 431' of the first three instruments each reflect a percentage of the laser light 1007, so that each of the first three units of the series receives 25% of the electromagnetic radiation of the original beam 1007. For example, the reflecting filters of the first, second and third units may reflect 25%, 33% and 50% of the incident light, respectively. The last reflecting filter 1321 of the series desirably reflects all of the remaining light (about 25% of the original beam) to the last instrument of the series. As a result, each of the four instruments should receive the same intensity of radiation (light) to interrogate the cells in respective streams.

[0401] Depending on the beam splitting devices used in the above system 1273, it may be desirable that the laser beam have a particular polarization. The transmittance-to-reflectance ratio of dielectric filters can vary depending on the polarization of the light. Further, when dealing with linearly polarized light, dielectric filters (which are manufactured for use at a specified angle of incidence) can be too sensitive to variations in the angle of incidence. Circularly or elliptically polarized light alleviates this problem to some extent because the polarization vector of the light is in a variety of different orientations with respect to the optical axis of a dielectric filter as the light interacts with the filter. Thus, elliptically or circularly polarized light simulates randomly polarized light, which provides more tolerance for variations in the angle of incidence on a dielectric filter. Accordingly, if the laser described above generates a beam of light having a vertical polarization, for example, it may be advantageous to convert the light to circularly polarized light before it is split. As will be understood by those skilled in the art, this can be accomplished by passing the beam through a 1/4-wave retardation plate (filter) of polarizing material having its optical axis rotated 45 degrees relative to the plane of the laser polarization. The beam thus transmitted by the waveplate will have approximately circular polarization, and it can be more easily split to provide multiple beams to the optics systems of respective flow cytometer units.

[0402] Moreover, by rotating the wave retardation plate to alter the angle between the laser polarization and the optical axis of the material used to make the waveplate, eccentricity can be introduced into the approximately circular polarization of the beam (i.e., the polarization can be made more elliptical). Changing the eccentricity of the elliptical polarization of the beam can change the transmittance-to-reflectance ratio of the dielectric filters by causing the polarization vector for a greater percentage of the light to have a particular angle with respect to the optical axis of the dielectric filter. Accordingly, if the balance of light among the multiple cytometry units is outside the desired range, one can rotate the waveplate to increase or decrease the eccentricity of the elliptically polarized light, thereby altering the transmittance-to-reflectance ratios of the various filters until a better balance is achieved. Similarly, if the waveplate is transmitting elliptically polarized light, one can influence the transmittance-to-reflectance ratio of one of the filters by rotating that filter.

[0403] Regardless of the method used to split the single beam into multiple separate beams. Balance of the power delivered to each cytometry unit can be achieved by selectively blocking a percentage of the light to bring all the cytometry units down to the same level of power. For example, the neutral density filter 447' of each epi-illumination system 415' can be selected to block more or less of the light to balance the illuminating power delivered by the beam splitting and guidance system to each individual cytometry unit. If one channel of a multi-channel unit receives significantly more illumination from the beam splitting and guidance system, a neutral density filter 467' that transmits less light can be used in the epi-illumination system 415' of that channel to bring the illumination power for that channel more in line with the other channels. It is desirable, though not essential, that channel-to-channel variations in the illuminating power be less than about 10%. It is even more desirable that the channel-to-channel variations be less than about 5%.

[0404] It will also be appreciated that pulsed laser scanning, as described above, may be desirable for multi-channel flow cytometry. For example, the UV multiline laser can be replaced with a mode-locked pulsed laser operating at about 85 MHz to allow more flow cytometry channels to be powered by a single laser. For example, the peak power provided in each pulse of a mode-locked laser emitting pulses having a width (duration) of about 12 picoseconds at a frequency of about 85 MHz is approximately 800 times the average power output of the laser. Thus, a mode-locked laser (e.g., a Vanguard 350 from Spectra-Physics) can provide enough illumination energy to operate a few dozen cytometry units (e.g., 32 cytometry units) while operating at only about 350 milliwatts.

[0405] The use of fiber optics for supplying light to the units is also contemplated. The fibers may be used to direct light from the laser to respective units, thus eliminating the need for the guidance system described above.

F. Common Deflector Plates



[0406] In Figs. 106 and 108-116, the sorting system 119' of each flow cytometry unit 1003 is substantially identical to the sorting system 119 except that the units desirably share two common deflector plates 1331 extending across the width of the housing 1009 at the front of the housing. There are advantages to using a common set of deflector plates, including a consistent charge from one channel to the next, the use of a common power supply, a larger plate area providing a more stable electric field and more uniform droplet deflection, and a consistent angle of deflection for collection of sorted samples. The deflector plates 1331 are mounted on a frame 1333 fastened to the housing 1009.

[0407] Alternatively, separate plates could be provided for each unit.

G. Common Collection System



[0408] In Figs. 107 and 116, a common collection system 2801 includes two intercepting devices for each cytometry unit as described above in connection with the collection system 2201 for the single unit. However, a common frame 2803 is provided to hold all eight of the intercepting devices. Also, one of the two intercepting devices for each cytometry unit directs fluid into a common waste trough 2805 rather than a collection vessel. The waste trough makes it easier to discard sorted droplets that contain particles of little value (e.g., Y-chromosome bearing sperm cells for breeding dairy cows). If it is desirable to retain all the sorted droplets, the waste trough can be removed and collection vessels can be placed under each intercepting device. The four collection vessels in the example shown in Figs. 107 and 116 rest in openings in the surface of a collection tray 2807. A common water bath (not shown) may be provided under the surface of the collection tray to control the temperature of the contents of the collection vessels.

H. Multi-Channel Control



[0409] The various flow cytometry units are controlled by the microprocessor 131', (or other suitable processing system) which desirably has a common input and a common output as discussed above.

[0410] Desirably, the operational parameters of each flow cytometry unit 1003 can be set independently of the other units so that such parameters can be varied as between units. These parameters may include, for example, the frequency of droplet formation, the control and sorting strategies utilized by a particular unit, the criteria used by each unit to classify and sort particles in the fluid supplied to the unit, and other parameters. For example, in certain situations it may be desirable to supply one or more units with carrier fluid 17' at a first flow rate and other units a second (different) flow rate. Similarly, it may be desirable to use one control sorting strategy (e.g., a "high efficiency" strategy) for one or more units while using a different strategy (e.g., a "low loss" strategy) for other units. By controlled variation of these parameters among the units, based on historical data and data collected on a real-time basis, the throughput of the units can be managed and the results of the system optimized. The capability of independent operation also allows selected units to be operated in the event fewer than all of the units are needed or available.

I. Operation of Multi-Channel System



[0411] The operation of the multi-channel system may be similar to that described previously, except that the multiple flow cytometry units are adapted to conduct flow cytometry operations in parallel (i.e., during the same time period or overlapping time periods) for higher throughput.

[0412] Prior to the start of a run, the fluid delivery system 1021 is flushed, if necessary, with cleaning solution from the tank 1173 by moving the valve V5 to its cleaning position. The system is then conditioned with buffer fluid using the syringe pump 1141. During this procedure, the valves V1A-V1D and V2A-V2D are moved to establish communication with the waste receptacle 1223 which is under vacuum. As a result, the cleaning solution and/or buffer fluid flows through the system to waste. This process cleans the system 1021, primes the syringe pump 1141 and removes air bubbles from the system.

[0413] With the three-way valve 1149 suitably positioned, the syringe pump 1141 is operated through an intake stroke to aspirate a quantity of carrier fluid 17' containing particles, e.g., sperm cells, following which the valve 1149 is moved to establish communication with the manifold 1177 and the syringe pump moves through a discharge stroke to pump a volume of carrier fluid into the sample reservoir 1207 to fill it. The temperature of the carrier fluid 17' is controlled by the temperature control system 1257 to maintain the cells in the carrier fluid at the desired temperature. With the valves V1A-V1D positioned to establish communication with the sample reservoir 1207, further operation of the syringe pump 1141 forces carrier fluid through the lines to the needles of respective nozzle assemblies for flow through the nozzles 101', as previously described. At the same time, and with the valves V2A-V2D positioned to establish communication with the sheath fluid tank 1155, sheath fluid 19' is forced through the supply lines to respective flow bodies and through the nozzles, also as previously described. This process continues for an appropriate length of time sufficient to pump a suitable volume of fluid through the system 1001. The duration of a particular run will vary depending on the quantity of carrier fluid in the supply vessel, the rate at which the carrier fluid is pumped through the system, and the number of channels in the system. For example, a run may continue for only a limited period of time (e.g., 15 minutes during which about one ml of carrier fluid is delivered to each nozzle) or it may continue indefinitely, with the supply of fluid being replenished as needed.

[0414] In the event a needle 157' becomes clogged, the appropriate valve V1 is moved to establish communication with the waste receptacle 1223. Sheath fluid 19' entering the flow body 133' will then flow under the force of the vacuum 1225 back through the needle 157' to waste, thus flushing and clearing the needle. If there is a need to shut off the flow to a particular nozzle, the valves V1 and V2 are simply switched to their waste positions.

[0415] Although the system described herein with respect to both the single channel and multi-channel configurations has been described with regard to particle separation, such as the separation of X and Y cells, it is contemplated that such particles include any particles having different characteristics which may be arbitrarily noted as characteristic A and characteristic B. Further, the sorting function may be eliminated entirely, so that the flow cytometry apparatus (single-channel or multi-channel) operates only to classify the particles and not to sort them.

[0416] While the multi-channel system is described above in the context of operating the flow cytometry units in parallel, it will be understood that the units could also be operated in series. For example, it is contemplated that particles in one stream could be sorted by one unit into multiple populations, and that one or more of such sorted populations could then be passed through one or more other units in series to perform additional sorting operations to sort different particles using the same or different sorting strategies.

J. Upright Multi-Channel System



[0417] Figs. 119 and 120 show a multi-channel flow cytometry system. This system, generally designated 4001 comprises four cytometry units 4009 ganged together. The nozzle system 101', epi-illumination optics system 450', deflector plates 629', sample station 4051, contamination prevention mechanism 4031 and other components of each unit 4009 are mounted on a shared vertical mounting board 4011. Referring to Fig. 120, a single laser 4013 and a beam splitting and guidance system 4015, which is substantially similar to the beam splitting and guidance system 1273 described above, provide illumination for each epi- illumination system 450'. The laser 4013 passes through a hole 4019 (Fig. 119) in a common housing 4021 containing the beam splitting and guidance system 4115. The beam splitting and guidance system 4115 and epi-illumination systems 450' are on one side of the board 4011. The focusing lens assembly 491' of each epi-illumination system 450' extends through the board 4011 to the other side (similarly to the configuration show in the single channel system shown Figs. 26 & 27), on which the remainder of the components for the units 4009 are mounted.

[0418] The units 4009 are all oriented so that their nozzle systems 101' direct the fluid streams 21' downward. Each unit 4009 also has a collection system 4031, which includes a collection vessel 4033 for collecting droplets 33 containing a desired population of particles and a waste container 4035 for collecting other droplets 33. A water bath (not shown) or other temperature control may be used to control the temperature of the collection vessel 4033.

[0419] The multiple flow cytometry units 4009 can also share a common power supply (not shown), a common input for controlling operation of the units (not shown), and a common output (not shown) allowing comparative evaluation of the operation of the units 4009 relative to one another. As demonstrated by comparison of the two exemplary multi-channel examples 1001, 4001, the nature of the integrated platform and the sharing of features between or among multiple flow cytometry units in a multi-channel system can be varied extensively.

Impact of Multi-Channel Processing on Overall Process



[0420] The overall process described above can be performed with multi-channel sperm sorting to decrease the time required to sort the sperm cells. With few exceptions, the method does not change. One minor change is that sorted sperm cells will be collected in multiple collection vessels. The contents of the multiple collection vessels can be combined for concentration if desired. Alternatively, the contents of each collection vessel can be concentrated separately. It will be appreciated that the time required to sort a batch of sperm cells (e.g., an ejaculate) from collection to completion of the cryopreservation step can be significantly reduced by using multiple cytometry units to process the batch. For example, if four cytometry units operate in parallel to process the batch of sperm cells, the time required to complete sorting is reduced to approximately one quarter of the time required to sort the batch using a single cytometry unit. Thus, by substituting the step of sorting sperm with four cytometry units operating in parallel with the step of sorting sperm with a single cytometry unit, the exemplary timeline for completion of the method from collection to completion of the freezing step can be reduced. The time can be reduced even further by increasing the number of cytometers operating in parallel to sort the sperm cells in the sample, subject to the practical limitations involved in operating a parallel system having more than four such units. Thus, the sorting step in the overall process described above is performed by sorting the sperm cells according to a specified DNA characteristic in a multi-channel flow cytometry apparatus. For example the sperm processing method may comprise the step of sorting sperm cells according to a specified DNA characteristic in a multi-channel flow cytometry apparatus in which each channel collects in the range of about 2,000-10,000 sperm cells having a desired DNA characteristic per second.

Multi-Channel Sorting Example I



[0421] Bull semen was collected from a sexually mature bull using an artificial vagina and the sample transported to a nearby staining facility in a temperature-controlled container at 37°C. Upon receipt, the semen was analyzed for concentration, visual motility, motility and progressive motility by the Hamilton-Thorn Motility Analyzer (IVOS), according to standard and well known procedures (Farrell et al. Theriogenology, 49 (4): 871-9 (Mar 1998)).

[0422] Six tubes of 1 mL of 150 X 106 sperm/mL sperm suspension were prepared by suspending an aliquot of semen in 41 °C TCA #2 buffer containing 10mM pyruvate bringing the overall pH to 7.35. Then varying amounts of 10 mM Hoechst 33342 solution in water were added to the sperm samples to yield final dye concentrations of 200, 300, 400, 500, 600, & 700 µM Hoechst 33342. Each of the six samples was incubated at 41 °C for approximately 30 minutes. The samples were analyzed by flow cytometry and the % CV of the X cell population was estimated by iterative computer algorithm for the 200, 300, and 400 µM Hoechst 33342 samples. The %CV for the 300 and 200 µM Hoechst 33342 were both ascertained to be within the acceptable range near 1.3 % CV. Accordingly, it was determined that a concentration of 250 µM Hoechst 33342 would be used to stain a batch of sperm cells for further processing.

[0423] Two tubes containing 2mL each of 150 X 106 sperm/mL sperm suspension were prepared by suspending an aliquot of semen in 41 °C TCA #2 buffer containing 10mM pyruvate (again bringing the overall pH to 7.35). Then 10 mM Hoechst 33342 solution in water was added to each of the two sperm suspensions to yield a final dye concentration of 250µM Hoechst 33342. The sperm suspensions were maintained in a 41°C water bath for 30 min. After 30 minutes, the sperm suspensions were removed from the 41°C water bath and 4µL of 25mg/mL FD&C #40 was added to one of the suspensions. The other was stored at ambient temperature to provide comparison samples for the assessment assays.

[0424] The stained and quenched sperm suspension was loaded onto the sample port of one channel of a four channel droplet sorting flow cytometer. Delbecco's PBS was used as the sheath fluid. The cytometer was equipped with an orienting nozzle as described above and having a 60 micron orifice. A semicircular baffle plate was installed perpendicular to the longitudinal axis of the nozzle as described above. The transducer was operated at 54 KHz and the droplet break-off location was controlled manually. An epi-illumination optics system as described above was used to direct approximately 25% of the beam of a continuous wave laser to intersect the fluid stream at a perpendicular angle. The focusing and collection lens had a 0.65 numerical aperture. The beam was focused to a spot having a width less than 3 µm for slit scanning the sperm cells. Digital signal processing was used to extract the critical slope difference and pulse area for each detected pulse waveform. Classification parameters for classification of X cells, Y cells, and undetermined cells in the two-dimensional CSD and pulse area feature space were manually entered into the processing system for classifying sperm cells according to chromosome content.

[0425] Sperm were sorted according to X and Y chromosome content using a coincidence accept sort strategy for collection of X cells, assigning a 50/50 probability that each unclassified sperm was an X cell or Y cell. The sample fluid rate was manually adjusted to maintain purity of collected X cell population (as indicated by the GUI) at 85% or better and to maintain the rate of X cell collection above a minimum rate. After approximately fifteen million X sperm had been collected in a tube that had been soaked in sheath fluid for at least one hour and then coated with 0.5mL of 10% egg yolk in TCA #2 buffer at pH 7.0, the tube was removed and replaced with an additional tube that has been similarly prepared.

[0426] Immediately after removing a collection tube from the flow cytometer, a comparison sample from the stained, but not sorted, sperm suspension was prepared. The sorted and comparison samples were centrifuged for 7 min @ 750g in a 15 mL tube. The supernatants were removed using a transfer pipette to yield a concentration of approximately 40 million sperm/mL. TCA #2 buffer pH 7.0 was added to the sperm suspensions to yield a final concentration of approximately 20 million sperm/mL. This process continued until the flow cytometer had produced four collection tubes (A2-A5). The sorted samples and "non-sorted" comparison samples were assessed by IVOS. Sorted sample A3 and its non-sorted comparison sample were tested for % intact acrosomes by differential interference contrast microscopy. All the sorted samples were counted by hemacytometer to determine the output rate of sorted sperm per hour. The % X chromosome bearing sperm was confirmed by flow cytometer reanalysis. Results of the IVOS assessment for the sorted and "non-sorted" comparison samples are provided in Figs. 121 (motility) and 122 (progressive motility). The total number of sperm sorted into each collection tube is shown in Fig. 123. The rate of sperm sorted per hour for each collection period is shown in Fig. 124. Percentage of X chromosome bearing sperm for each sorted sample is listed in Fig. 125. Results of the assessment of the acrosome integrity were 72% intact acrosomes for the sorted sample and 78% for the non-sorted comparison sample.

[0427] The results demonstrate the technical ability to yield more than 5,000 sorted X cells per second at greater than 85% purity per channel of multi-channel flow cytometry system for sustained periods. The results also show the technical ability to yield more than 7,000 X cells per second at greater than 85% purity for sustained periods under ideal conditions. Further, the results indicate that samples of sorted sperm cells obtained by such high-speed flow cytometric sorting will suffer only slight declines in motility, indicating that the sorted sperm will have good fertility.

Multi-channel Sorting Example II



[0428] Bull semen was collected from a sexually mature bull using an artificial vagina. The ejaculate was split into two aliquot. The first aliquot of 250µL of semen was suspended in 5mL of 37°C Triladyl®. The second aliquot, which comprised the remained of the ejaculate, was suspended in two parts 37°C carbonate buffer (pH 6.1-6.2). Both aliquots were transported at 37°C in a temperature-controlled container to a processing facility. At the processing facility, the first aliquot was floated in ~ 120mL of 37°C water in a 200mL beaker and placed in a cold room to slowly cool to 5°C. The second aliquot was analyzed for concentration, motility and progressive motility by the Hamilton-Thorn Motility Analyzer (IVOS), according to standard and well known procedures (Farrell et al. Theriogenology, 49 (4): 871-9 (Mar 1998)).

[0429] Three 1mL tubes of 150 X 106 sperm/mL sperm suspension were prepared by transferring sub-aliquots containing 150 million sperm from the second aliquot to empty tubes, centrifuging at 500g for 5 min, removing the supernatants, and re-suspending the sperm pellets in 1mL of 28°C TCA #2 buffer containing 10mM pyruvate pH 7.35. Ten mM Hoechst 33342 solution in water was added to each of the three tubes in various amounts to yield final dye concentrations of 100, 150, & 200 µM Hoechst 33342. Each of the three tubes was held at 28°C for approximately 60 minutes. Sperm from each of the three tubes was analyzed by flow cytometry and the CV of total fluorescence intensity of the X population was determined for the 100, 150, and 200 µM Hoechst 33342 staining conditions using an interactive computer algorithm. The CVs for the 150 and 200 µM Hoechst 33342 were both within the acceptable range near 1.3%. Thus, it was determined to use staining conditions including 150 µM Hoechst 33342 concentration for sorting.

[0430] One tube containing 5mL of 150 X 106 sperm/mL sperm suspension was prepared by transferring a sub-aliquot containing 750 million sperm from the second aliquot, centrifuging at 500g for 5 min, removing the supernatant, and re-suspending the sperm pellet in 28°C TCA #2 buffer containing 10mM pyruvate (pH 7.35). Ten mM Hoechst 33342 solution in water was added to the tube in an amount yielding a final dye concentration of 150µM Hoechst 33342. The tube was maintained in a 28°C water bath for 60 min. After 60 minutes, the tube was removed from the 28°C water bath and 10µL of 25mg/mL FD&C #40 was added.

[0431] The now stained and quenched sperm suspension was loaded onto the sample port of one channel of a multi-channel droplet sorting flow cytometer system. The sperm suspension was maintained at 28°C. Using substantially the same instrument settings as set forth in Multi-channel Example I, X & Y chromosome bearing sperm were separated by the flow cytometry system using a coincidence abort sort strategy for a period necessary to place an enriched X cell population of approximately eighteen million sperm into a collection tube that had been prepared by soaking with sheath buffer for at least one hour and then adding 0.5mL of Triladyl® cryo-preservation media containing 10 mM pyruvate pH 6.6. The sperm cells were introduced into the flow cytometry system at a rate of between about 25,000 and 30,000 cells/second. An enriched population of X cells was collected at a rate varying from 4,500 per second to 6,000 per second. When approximately eighteen million sperm had been sorted into a collection tube, the tube was removed and replaced with another tube that had been similarly prepared. Immediately after removal of a collection tube from the flow cytometer, the sorted sperm suspension was centrifuged for 7 min @ 700g. The supernatant was removed using a transfer pipette to yield a concentration of approximately 100 million sperm/mL. Triladyl® cryo-preservation media containing 10 mM pyruvate (pH 6.6) was added to the sperm suspensions to yield a final concentration of approximately 50 million sperm/mL. This process continued until the flow cytometer had produced three collection tubes (D1-D3). Approximately 52 million sperm were sorted in 259 min yielding an overall collection rate of about 12 million enriched X sperm per hour of sorting. The re-suspended sorted sample tubes were floated in ~120mL of 28°C water in a 200mL beaker and placed in a 5°C cold room to slowly cool.

[0432] After the sorted samples reached 5°C, the three tubes of sorted sperm were combined into one tube. The pooled sample was analyzed by IVOS to determine the % motility, % progressive motility, and concentration. Additional Triladyl® cryo-preservation media containing 10 mM pyruvate pH 6.6 was added to the sample to yield a final concentration of approximately 50 million sperm per mL. The % X-chromosome bearing sperm in the sorted pooled sample was 87% as determined by flow cytometer re-analysis. A summary of the IVOS assessment compared to the non-sorted sample of the same ejaculate is illustrated in Fig. 126.

[0433] The pooled sorted sample and the first aliquot were loaded into standard 0.25cc straws in a 5°C cold room. The loaded straws were transferred to a programmable freezer and frozen by the following program: 5 min @ 5°C, cool from 5°C to-12°C @ 4°C/min, cool from -12°C to - 100°C @ 40°C/min, cool from -100°C to -140°C @ 20°C/min, hold at -140°C. After the straws had reached -140°C, they were quickly removed from the freezer and plunged into liquid nitrogen.

[0434] Thawed straws were analyzed by IVOS for % motility and % progressive motility after incubation at 37°C for 30 and 120 minutes. Results from a set of two sorted and unsorted straws are summarized in Fig. 127 and Fig. 128.

Multi-channel Sorting Example III



[0435] Bull semen was collected from a sexually mature bull using an artificial vagina and the ejaculate split into two aliquot. A first aliquot of 250µL of semen was suspended in 5mL of 37°C Triladyl®. A second aliquot, which comprised the remainder of the ejaculate, was suspended in two parts 37°C carbonate buffer (two parts 0.097 moles/L of NaHCO3, 0.173 moles/L of KHCO3, 0.090 moles/L C6H8O7·H2O in water) (pH 6.1-6. 2). Both aliquots were transported at 37°C in a temperature-controlled container to the processing facility. At the processing facility, the first aliquot was floated in ~120mL of 37°C water in a 200mL beaker and placed in a cold room to slowly cool to 5°C. The second aliquot was analyzed for concentration, motility and progressive motility by the Hamilton-Thorn Motility Analyzer (IVOS), according to standard and well known procedures (Farrell et al. Theriogenology, 49 (4): 871-9 (Mar 1998)).

[0436] Two tubes of 150 X 106 sperm/mL sperm suspension were prepared by transferring into each of two empty tubes a fraction containing 900 million sperm from the second aliquot, centrifuging each tube at 500 X g for 5 minutes, removing the supernatant from each tube, and re-suspending each sperm pellet in 6mL of 28°C TCA #2 buffer containing 10mM pyruvate (pH 7.35). 10mM Hoechst 33342 solution in water was added to each of the two tubes to yield final dye concentrations of 200µM Hoechst 33342 in one tube and 400µM Hoechst 33342 in the other tube. Each of the two tubes was held at 28°C for approximately 120 minutes. Sperm from each of the tubes was analyzed by flow cytometry and the CV of total fluorescence intensity of the X population was determined for the 200µM and 400µM Hoechst 33342 staining conditions using an interactive computer algorithm. The CVs for the 200µM and 400µM Hoechst 33342 were both within the acceptable range of about 1.3%. The sperm suspension stained with a concentration of 200ptM Hoechst 33342 was chosen for sorting. 10µL of 25mg/mL FD&C #40 was added to this tube of stained sperm suspension just prior to sorting.

[0437] The stained sperm suspension was loaded onto the sample port of one channel of a multi-channel droplet sorting flow cytometer system. The sperm suspension was maintained at 28°C. Using substantially the same instrument settings as set forth in Multi-channel Example I, X & Y chromosome bearing sperm were separated by the flow cytometry system using a coincidence abort sort strategy for a period of time necessary to place an enriched X chromosome bearing cell population of approximately eighteen million sperm into a collection tube that had been prepared by soaking with sheath buffer for at least one hour and then adding 0.5mL of Triladyl® cryo-preservation media (pH 6.6). The sperm cells were introduced into the flow cytometry system at a rate of between about 25,000 and 30,000 cells/second. An enriched population of X chromosome bearing cells was collected at a rate varying from 4,500 per second to 6,000 per second. When approximately eighteen million sperm had been sorted into a collection tube, the tube was removed and replaced with another tube that had been similarly prepared. Immediately after removal of a collection tube from the flow cytometer, the sorted sperm suspension was centrifuged for 7 min @ 700 X g. The supernatant was removed using a transfer pipette to yield a concentration of approximately 100 million sperm/m L. Triladyl® cryo-preservation media (pH 6.6) was added to the sperm suspensions to yield a final concentration of approximately 50 million sperm/mL. This process continued until the flow cytometer had produced two collection tubes (C1-C3). Approximately 35 million sperm were sorted in 193 minutes yielding an overall collection rate of 11 million enriched X chromosome bearing cells per hour of sorting. The re-suspended sorted sample tubes were floated in ~120mL of 28°C water in a 200mL beaker and placed in a 5°C cold room to slowly cool.

[0438] After the sorted samples reached 5°C, the three tubes of sorted sperm were combined into one tube. The pooled sample was analyzed by IVOS to determine the % motility, % progressive motility and concentration. Additional Triladyl® cryo-preservation media (pH 6.6) was added to the sample to yield a final concentration of approximately 50 million sperm per mL. The % X-chromosome bearing sperm in the sorted pooled sample was 88% as determined by flow cytometer re-analysis. A summary of the IVOS assessment compared to the non-sorted sample of the same ejaculate is illustrated in Fig. 129.

[0439] The pooled sorted sample and unsorted sample (i.e., the first aliquot from above) were loaded into standard 0.25cc straws in the 5°C cold room. The loaded straws were transferred to a programmable freezer and frozen by the following program: 5 min @ 5°C, cool from 5°C to - 12°C @ 4°C/min, cool from -12°C to -100°C @ 40°C/min, cool from -100°C to -140°C @ 20°C/min, hold at -140°C. After the straws had reached -140°C, they were quickly removed from the freezer and plunged into liquid nitrogen.

[0440] Thawed straws were analyzed by IVOS for % motility and % progressive motility after incubation at 37°C for 30 and 120 minutes. Results from a set of sorted and unsorted straws are summarized in Fig. 130 and Fig. 131.

Multi-channel Sorting Example IV



[0441] Bull semen was collected from a sexually mature bull using an artificial vagina and the ejaculate split into two aliquot. The first aliquot of 250µL of semen was suspended in 5 mL of 37°C Triladyl®. The second aliquot, which comprised the remained of the ejaculate, was suspended in two parts 37°C carbonate buffer (two parts 0. 097 moles/L of NaHCO3, 0.173 moles/L of KHCO3, 0.090 moles/L C6H8O7·H2O in water) (pH 6.1-6. 2) and held under CO2. Both aliquots were transported at 37°C in a temperature-controlled container to the processing facility. At the processing facility, the first aliquot was floated in ~120mL of 37°C water in a 200mL beaker and placed in the cold room to slowly cool to 5°C. The second aliquot was analyzed for concentration, motility and progressive motility by the Hamilton-Thorn Motility Analyzer (IVOS), according to standard and well known procedures (Farrell et al. Theriogenology, 49 (4): 871-9 (Mar 1998)).

[0442] A 5 mL tube of 150 X 106 sperm/mL sperm suspension was prepared by transferring a fraction containing 750 million sperm from the second aliquot (pH 6.1-6.2) to an empty tube and adding 28°C carbonate buffer (pH 7.35) to a final volume of 5 ml. To this sperm suspension, 10 mM Hoechst 33342 solution in water was added to yield a final dye concentration 150µM Hoechst 33342. The suspension was held at 41°C under CO2 for approximately 40 minutes and then placed at 28°C for sorting. Ten µL of 25mg/mL FD&C #40 was added to the tube of stained sperm suspension just prior to sorting.

[0443] The stained sperm suspension was loaded onto the sample port of one channel of a multi-channel droplet sorting flow cytometer system. The sperm suspension was maintained at 28°C. X & Y chromosome bearing sperm were separated by the flow cytometry using a coincidence abort sort strategy for a time period necessary to place an enriched X chromosome bearing cell population of approximately eighteen million sperm into a collection tube that had been prepared by soaking with sheath buffer for at least one hour and then adding 0.5mL of Triladyl® cryo-preservation media (pH 6.6). The sperm cells were introduced into the flow cytometry system at a rate of between about 25,000 and 30,000 cells/second. An enriched population of X chromosome bearing cells was collected at a rate varying from 4,500 per second to 6,000 per second. When approximately eighteen million sperm had been sorted into a collection tube, the tube was removed and replaced with another tube that has been similarly prepared. Immediately after removal of a collection tube from the flow cytometer, the sorted sperm suspension was centrifuged for 7 min @ 700 X g. The supernatant was removed using a transfer pipette to yield a concentration of approximately 100 million sperm/mL. Triladyl® cryo-preservation media and pyruvate (pH 6.6) was added to the sperm suspensions to yield a final concentration of approximately 50 million sperm/mL. This process continued until the flow cytometer had produced two collection tubes (C2-C3). The re-suspended sorted sample tubes were floated in ~120mL of 28°C water in a 200mL beaker and placed in a 5°C cold room to slowly cool.

[0444] After the sorted samples reached 5°C, the two tubes of sorted sperm were combined into one tube. The pooled sample was analyzed by IVOS to determine the % motility, % progressive motility and concentration. Additional Triladyl® cryo-preservation media and pyruvate (pH 6.6) was added to the sample to yield a final concentration of approximately 50 million sperm per mL. A summary of the IVOS assessment compared to the non-sorted sample of the same ejaculate is illustrated in Fig. 132.

[0445] The pooled sorted sample and unsorted sample (i.e., the first aliquot from above) were loaded into standard 0.25cc straws in the 5°C cold room. The loaded straws were transferred to a programmable freezer and frozen by the following program: 5 min @ 5°C, cool from 5°C to - 12°C @ 4°C/min, cool from -12°C to -100°C @ 40°C/min, cool from -100°C to -140°C @ 20°C/min, hold at -140°C. After the straws had reached -140°C, they were quickly removed from the freezer and plunged into liquid nitrogen.

[0446] Thawed straws were analyzed by IVOS for % motility and % progressive motility immediately after thawing and after incubation at 37°C for 30 minutes. Results from a set of sorted and unsorted straws are summarized in Fig. 133 and Fig. 134.

CAPILLARY TUBE NOZZLE SYSTEM



[0447] Fig. 135 illustrates an alternative nozzle system, generally designated 1335, similar to that described above except that a capillary tube 1337 (of quartz or fused silica, for example) is connected to the nozzle 137 so that fluid exiting the nozzle orifice 103 is directed into and through the tube. The optics system 109 of the flow cytometer is optically coupled to the side of the tube in a suitable manner, as by a chamber 1341 filled with a light-transmitting medium such as oil or gel having a known index of refraction. The use of a capillary tube, compared to the jetting of the fluid stream through open space, has the benefit of reducing the lensing of the stream 21 due to the acoustical energy supplied by the transducer 105, and enabling the focusing lens 1343 to be positioned immediately adjacent the fluid stream for increasing resolution of the emission signals.

[0448] After the particles have been interrogated and classified, they may be sorted using any conventional techniques known to those skilled in the art, as by use of a fluid switching device shown in Fig. 137 or other suitable devices such as photo-damage systems or droplet sorting systems.

SORTING TECHNIQUES OTHER THAN DROPLET SORTING


Photo-damage Sorting



[0449] The flow cytometry improvementsmay be applicable not only to droplet cell sorting as described above, but also to other sorting techniques, such as sorting by photo-damage (laser ablation). Photodamage sorting is discussed in United States Patent No. 4,395, 397. Fig. 136 schematically illustrates one example of a single-channel flow cytometry photo-damage system.

[0450] As shown in Fig. 136, the photo-damage sorting system 1351 is similar to the droplet sorting system of Fig. 2, and corresponding parts are designated by corresponding reference numbers with the addition of a double prime ("). In general, the system comprises the same components as the system of Fig. 2, except that the droplet sorting components are eliminated (e.g., the transducer 105, the charging device 627, the deflector plates 629, and associated power sources 635). Instead these components are replaced by a laser 1353 or similar device which is responsive to instructions received from the microprocessor 131" to ablate undesired particles in the fluid stream 21". As a result, the stream collected in a collection receptacle 1355 contains a desired population of particles. For example, if the particles being analyzed are sperm cells and the intended result is to collect sperm cells having a characteristic A (e.g., a desired chromosome content), then the microprocessor receives signals from the epi-illumination system 415" which identifies cells not having characteristic A and selectively activates the laser to ablate such cells or otherwise render them ineffective.

[0451] Different control sorting strategies can be employed in a photo-damage system, including the "high recovery" and "high purity" sorting strategies discussed above in the context of a droplet sorter. In a photo-damage system, particles contained in the fluid stream are spaced at various intervals along the stream and generally follow one after another in single file. The particles have different characteristics, some having a characteristic A, for example, and others having a characteristic B. The sequence of particles is random, so viewed as a continuous procession, the particles can be divided into different particle series, one following another, including a first particle series consisting only of one or more particles having characteristic A, a second particle series consisting only of one or more particles having characteristic B and a third particle series consisting of two or more closely spaced particles at least one of which has characteristic A and at least one of which has characteristic B. The latter (third) group generally corresponds to the closely spaced particles in a "coincident" droplet discussed previously, at least for sorting strategy purposes. Thus, the two or more particles in the third group may be closely spaced in the sense that the spatial separation between the particles is insufficient to allow accurate discrimination/classification of the particles, or because such separation is insufficient to permit one particle in the series to be ablated by the laser without damaging the other particle(s) in the same series. In any event, the closely spaced particles in each (or at least some) of the third series of particles can be ablated or not abated, depending on the sorting strategy employed. It should be noted that multiple particles in a first series or multiple particles in a second series could be "closely spaced", but since the particles in any such series have the same characteristic (A or B), they are treated as a single-particle series, at least for sorting strategy purposes.

[0452] The photo-damage system can be a single-channel system or a multi-channel system, as described above.

Fluid Switching Sorting



[0453] Flow cytometry systems using fluid switching techniques, are disclosed, for example, in U. S. Patents Nos. 6,432,246 (Adair), 4,756,427 (Gohde, et al.), and 3,791,517 (Friedman). Fig. 137 is a partial view showing such a system, generally designated 1357. It is substantially identical to the system shown in Fig. 2 except that the nozzle system 101" includes a capillary tube 1369 (e.g., see Fig. 135), and the sorting system comprises a fluid-switching device 1359 coupled to the capillary tube 1369 downstream from the interrogation location 115'. The construction and operation of the fluid- switching device can incorporate any conventional fluid switching technology such as disclosed in the above-referenced patents. In general, the device functions to sort desired particles from undesired particles in response to instructions received from the processor by intermittently diverting portions of the fluid stream containing the desired/undesired particles along separate flow paths 1361, 1365 for collection in vessels or the like. The switching is commonly achieved by selectively actuating a transducer 1367 in one of the flow paths.

[0454] The various sorting strategies described above in regard to droplet sorting and photo-damage sorting can also be employed in a fluid-switching system. In the fluid-switching system, particles contained in the fluid stream are also spaced at various intervals along the stream and generally follow one after another in single file. The particles have different characteristics, some having a characteristic A, for example, and others having a characteristic B, and the sequence of particles is random. Therefore, as discussed above in regard to the photo-damage system, the procession of particles can be divided into different particle series, one following another, including a first particle series comprising one or more particles having characteristic A, a second particle series comprising one or more particles having characteristic B and a third particle series comprising two or more closely spaced particles at least one of which has characteristic A and at least one of which has characteristic B. The latter (third) group generally corresponds to the closely spaced particles in a "coincident" droplet discussed previously, at least for sorting strategy purposes. Thus, the two or more particles in the third group may be closely spaced in the sense that the spatial separation between the particles is insufficient to allow accurate discrimination/classification of the particles, or because such separation is insufficient to permit one particle in the series to be diverted by the fluid-switching device separate from the another particle in the same series. In any event, the closely spaced particles in each (or at least some) of the third series of particles can be diverted to one collection location or another, depending on the sorting strategy employed. As explained above in connection with photo-damage sorting, multiple particles in a first series or multiple particles in a second series could be "closely spaced", but since the particles in any such series have the same characteristic (A or B), they are treated as a single-particle series for the purpose of sorting strategy.

[0455] The fluid switching system can be a single-channel system or a multi-channel system, as described above.

Droplet Interference Sorting



[0456] A high-speed droplet interference sorting system 1371, shown schematically in Fig. 138, may be used to sort particles by diverting selected segments of the coaxial carrier and sheath fluid stream.

[0457] In contrast to some other sorting techniques, the droplet interference sorting technique does not require the coaxial carrier and sheath stream to be formed into droplets. Thus, there is no need to couple the nozzle system 101"' used for delivery of the carrier and sheath fluids with a droplet generation system. By way of example only, passing the carrier and sheath fluids through a nozzle system at 60 psi to create a 50 micron diameter stream is one suitable arrangement for formation of a laminar coaxial fluid stream for delivery of particles to the droplet interference sorting system. Particles in the coaxial fluid stream are analyzed and classified by the optics system 109"' and processor 131"' as they move through the interrogation location 115"', as has been described above for the other sorting systems. Sorting occurs downstream from the interrogation location, at a location where the coaxial fluid stream intersects a high-speed droplet interference stream 1373.

[0458] The droplet interference stream 1373 is generated by a droplet generation system 1375 similar to the droplet generation system used for droplet sorting. A high-speed fluid stream 1379 passes through a high-speed nozzle system 1377 that is coupled to a piezoelectric transducer 1381 or other source of acoustical energy for causing the high-speed fluid stream to break into droplets 1383 downstream from the high-speed nozzle. For example, a particle-free fluid at 1500 psi may be passed through the high-speed nozzle to form a 70 micron diameter high-speed fluid jet. The high-speed nozzle may be oscillated at 400 KHz to form high-speed droplets. The high-speed droplets 1383 pass through an electric field generated by one or more electric deflection plates 1387 so that the path of the high-speed droplets may be controlled by selectively applying an electric charge to the droplets, as was done to control the path of droplets in the droplet sorting system. The high-speed droplet interference stream is directed so some high-speed droplets intersect the coaxial fluid stream at a point 1399 downstream from the interrogation location. For example, uncharged droplets 1389 may be directed to collide with the fluid stream while charged droplets 1391 are deflected away from the coaxial fluid stream. When a high-speed droplet collides with the coaxial fluid stream, a segment 1397 of the fluid stream and any particles contained therein are diverted from the path they would have otherwise taken. The application of a charge or no charge to a high-speed droplet may be timed so the arrival of that droplet at the intersection 1399 with the coaxial fluid stream coincides with the arrival of a particular segment of the coaxial fluid stream. Thus, by selectively charging high-speed droplets depending on the classification of particles contained within the coaxial stream segments, one can sort particles by diverting all coaxial fluid stream segments that contain one or more selected particles and not diverting other coaxial stream segments or vice-versa. Collection capillaries 1403 having a slight vacuum may be used to collect both the diverted 1397 and undiverted coaxial stream segments. The droplet interference sorting system may be set up so the high-speed droplets merge with diverted coaxial stream segments or so the high-speed droplets remain separate from the diverted stream segments after collision with the coaxial stream segments.

[0459] Because there are no particles or cells in the high-speed droplet interference stream 1373, it is possible to use very high pressures and very high droplet frequencies without damaging the particles or cells to be sorted. This allows sorting of stream segments each having less volume (e.g., four times less volume) than the volume of a droplet in the droplet sorting system. This greatly increases the maximum throughput of the system while also reducing the dilution factor of the sorted particles. Moreover, because finely filtered liquid with no cells or particles is used to form the droplet interference stream, more consistent droplet formation is possible because the droplet formation nozzle is less likely to become clogged or suffer from protein buildup than the nozzle system used in the droplet sorting system. Another advantage is that the distance between particle analysis at the interrogation location and the sorting point 1399 can be reduced (e.g., by a factor of four), allowing more accurate prediction of the time of arrival of a particular particle at the sorting point. Furthermore, the droplet interference system allows more flexibility in adjustment of nozzle size or pressure for the coaxial fluid stream. If desired, the droplet interference sorting system can be combined with the capillary tube nozzle system. A multi-channel droplet interference sorting system may use a high-pressure fluidic pump to supply multiple droplet interference stream generating nozzles with fluid from a common fluid supply.

[0460] When introducing elements from the example(s) thereof, the articles "a," "an," "the," and "said" are intended to mean that there are one or more of the elements. The terms "comprising," "including," and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements. The term "or" is intended to include "and/or" and is intended to mean "one or another or both." Thus, an indication of "ABC or DEF" means (1) ABC, or (2) DEF, or (3) both ABC and DEF. The term "and/or" is intended to have the same meaning as "or" as defined above. Thus, the term "and/or" is intended to include "or" and is intended to mean "one or another or both." For example, an indication of "ABC and/or DEF" means (1) ABC, or (2) DEF, or (3) both ABC and DEF.


Claims

1. A method of handling non-human semen having cells therein, the method including:

providing a supply of semen obtained from a non-human male animal;

using a programmable machine to conduct a plurality of integrated flow cytometry operations, the operations including:

a) receiving the supply of semen into the machine;

b) forming multiple streams containing the cells; and

c) sorting at least one portion of the cells into a first population of cells having a characteristic A and a second population of cells having a characteristic B; and

distributing the first or the second population for commercial use;
wherein the step of forming multiple streams containing the cells is enabled by having multiple flow cytometry units linked to form a single integrated multi-channel flow cytometry unit, and wherein the single integrated multi-channel flow cytometry unit utilizes a common supply of cells for the multiple streams; and a common system for delivering carrier fluid, the common carrier fluid delivery system being capable of supplying one or more of the units at a first flow rate and one or more of the units at a second flow rate; a common operations control input; and a common processor for receiving and processing information from the multiple flow cytometry units, wherein the operational parameters of each of the multiple flow cytometry units can be set independently of the other flow cytometry units.


 
2. A method as claimed in claim 1, wherein operational parameters of each of the multiple flow cytometry units can be set independently of the other units so that such parameters can be varied as between units, and wherein the operational parameters comprise frequency of droplet formation, control and sorting strategies, or criteria used to classify and sort particles.
 
3. A method as claimed in claim 1, characterised in that the programmable machine is operable to conduct the operations in parallel.
 
4. A method as claimed in claim 1 or claim 3, characterised in that the step of sorting at least one portion of the cells into first and second populations further includes the step of sorting the remaining portion of the cells of the supply into a first population containing cells having characteristic A and a second group containing cells having characteristic B, provided that the cells sorted from the at least one portion of the cells meet or exceed a preset standard.
 
5. A method as claimed in claim 4, characterised in that the preset standard is a minimum recovery rate of cells having characteristic A or characteristic B or a minimum purity for at least one of the populations.
 
6. A method as claimed in claim 1, characterised in that the step of sorting at least one portion of the cells into first and second populations further includes the steps of:

a) sorting the cells from a second portion of the supply into different populations, including a first population containing cells having characteristic A and a second population containing cells having characteristic B; and

b) blending one of the populations of the first portion with one of the populations of the second portion to obtain a blended population.


 
7. A method as claimed in claim 6, characterised in that the first population of the at least one portion has an unacceptable purity, and in that the first population of the second portion has an acceptable purity, and in that the blended population has an acceptable purity.
 
8. A method as claimed in claim 1, characterised in that the step of sorting at least one portion of the cells further includes the step of splitting a laser beam using a beam splitting and guidance system into either:

a) three or more;
or

b) any number of separate beams which are each directed into an optics system associated with each of the multiple flow cytometry units.


 
9. A method as claimed in claim 1, characterised in that the flow cytometry units share
a common housing for the multiple flow cytometry units.
 


Ansprüche

1. Verfahren zur Handhabung von nicht-menschlichem Samen, der Zellen darin aufweist, wobei das Verfahren Folgendes einschließt:

Bereitstellen einer Zufuhr von Samen, die von einem nicht-menschlichen männlichen Tier erhalten wurde;

Verwenden eines programmierbaren Geräts, um eine Vielzahl von integrierten Durchflusszytometrie-Vorgängen durchzuführen, wobei die Vorgänge Folgendes einschließen:

a) Aufnehmen der Zufuhr von Samen in das Gerät;

b) Bilden von mehreren Strömen, die die Zellen enthalten; und

c) Sortieren von mindestens einem Teil der Zellen in eine erste Population von Zellen, die ein Merkmal A aufweisen, und eine zweite Population von Zellen, die ein Merkmal B aufweisen; und

Ausgeben der ersten oder der zweiten Population für kommerzielle Verwendung;
wobei der Schritt des Bildens von mehreren Strömen, die die Zellen enthalten, dadurch ermöglicht wird, dass mehrere Durchflusszytometrie-Einheiten vorliegen, die verknüpft sind, um eine einzige integrierte Multikanal-Durchflusszytometrie-Einheit zu bilden, und wobei die einzige integrierte Multikanal-Durchflusszytometrie-Einheit Folgendes nutzt: eine gemeinsame Zufuhr von Zellen für die mehreren Ströme; und ein gemeinsames System für die Abgabe von Trägerflüssigkeit, wobei das gemeinsame Trägerflüssigkeitsabgabesystem eine oder mehrere der Einheiten bei einer ersten Flussrate und eine oder mehrere der Einheiten bei einer zweiten Flussrate versorgen kann; einen gemeinsamen Vorgangssteuereingang; und einen gemeinsamen Prozessor für den Empfang und die Verarbeitung von Informationen von den mehreren Durchflusszytometrie-Einheiten, wobei die Betriebsparameter von jeder der mehreren Durchflusszytometrie-Einheiten unabhängig von den anderen Durchflusszytometrie-Einheiten eingestellt werden können.


 
2. Verfahren nach Anspruch 1, wobei die Betriebsparameter von jeder der mehreren Durchflusszytometrie-Einheiten unabhängig von den anderen Einheiten eingestellt werden können, sodass derartige Parameter zwischen Einheiten variiert werden können, und wobei die Betriebsparameter die Frequenz der Tröpfchenbildung, Steuer- und Sortierstrategien oder Kriterien, die zur Klassifizierung und Sortierung von Partikeln verwendet werden, umfassen.
 
3. Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass das programmierbare Gerät betriebsfähig ist, um die Vorgänge parallel durchzuführen.
 
4. Verfahren nach Anspruch 1 oder Anspruch 3, dadurch gekennzeichnet, dass der Schritt des Sortierens von mindestens einem Teil der Zellen in erste und zweite Populationen weiter den Schritt des Sortieren des verbleibenden Teils der Zellen der Zufuhr in eine erste Population, die Zellen enthält, die Merkmal A aufweisen, und eine zweite Gruppe, die Zellen enthält, die Merkmal B aufweisen, einschließt, vorausgesetzt, dass die Zellen, die aus dem mindestens einen Teil der Zellen sortiert wurden, einem voreingestellten Standard entsprechen oder diesen übertreffen.
 
5. Verfahren nach Anspruch 4, dadurch gekennzeichnet, das der voreingestellte Standard eine minimale Rückgewinnungsrate von Zellen, die Merkmal A oder Merkmal B aufweisen, oder eine minimale Reinheit für mindestens eine der Populationen ist.
 
6. Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass der Schritt des Sortierens von mindestens einem Teil der Zellen in erste und zweite Populationen weiter die folgenden Schritte umfasst:

a) Sortieren der Zellen aus einem zweiten Teil der Zufuhr in verschiedene Populationen, die eine erste Population, die Zellen enthält, die Merkmal A aufweisen, und eine zweite Population, die Zellen enthält, die Merkmal B aufweisen, einschließen; und

b) Vermischen von einer der Populationen des ersten Teils mit einer der Populationen des zweiten Teils, um eine gemischte Population zu erhalten.


 
7. Verfahren nach Anspruch 6, dadurch gekennzeichnet, dass die erste Population des mindestens einen Teils eine inakzeptable Reinheit aufweist, und dadurch, dass die erste Population des zweiten Teils eine akzeptable Reinheit aufweist, und dadurch, dass die gemischte Population eine akzeptable Reinheit aufweist.
 
8. Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass der Schritt des Sortierens von mindestens einem Teil der Zellen weiter Folgendes einschließt: den Schritt des Teilens eines Laserstrahls unter Verwendung eines Strahlteilungs- und -führungssystems in entweder:

a) drei oder mehr;
oder

b) jedwede Anzahl von getrennten Strahlen, die jeweils in ein optisches System, das mit jeder der mehreren Durchflusszytometrie-Einheiten verbunden ist, geleitet werden.


 
9. Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass die Durchflusszytometrie-Einheiten ein gemeinsames Gehäuse für die mehreren Durchflusszytometrie-Einheiten teilen.
 


Revendications

1. Procédé de manipulation de sperme non humain contenant des cellules, le procédé comprenant les étapes consistant à :

fournir un approvisionnement de sperme obtenu auprès d'un animal mâle non humain ;

utiliser une machine programmable pour effectuer une pluralité d'opérations intégrées de cytométrie de flux, les opérations comprenant :

a) la réception de l'approvisionnement de sperme dans la machine ;

b) la formation de multiples flux contenant les cellules ; et

c) le tri d'au moins une partie des cellules dans une première population de cellules ayant une caractéristique A et une seconde population de cellules ayant une caractéristique B ; et

distribuer la première ou la seconde population à des fins commerciales ;
l'étape de formation de multiples flux contenant les cellules étant permise par le fait d'avoir de multiples unités de cytométrie de flux liées pour former une unité multicanaux intégrée de cytométrie de flux individuelle, et l'unité multicanaux intégrée de cytométrie de flux individuelle utilisant un approvisionnement commun de cellules pour les multiples flux ; et un système commun pour distribuer le fluide porteur, le système commun de distribution de fluide porteur étant capable de fournir une ou plusieurs des unités à un premier débit et une ou plusieurs des unités à un second débit ; une entrée commune de contrôle des opérations ; et un processus commun pour recevoir et traiter les informations provenant des multiples unités de cytométrie de flux, les paramètres opérationnels de chacune des multiples unités de cytométrie de flux pouvant être configurés indépendamment des autres unités de cytométrie de flux.


 
2. Procédé selon la revendication 1, les paramètres opérationnels de chacune des multiples unités de cytométrie de flux pouvant être configurés indépendamment des autres unités de telle sorte que ces paramètres peuvent être modifiés entre les unités, et les paramètres opérationnels comprennent la fréquence de formation de gouttelettes, les stratégies de contrôle et de tri ou les critères utilisés pour classer et trier les particules.
 
3. Procédé selon la revendication 1, caractérisé en ce que la machine programmable est exploitable pour effectuer les opérations en parallèle.
 
4. Procédé selon la revendication 1 ou 3, caractérisé en ce que l'étape de tri d'au moins une partie des cellules dans des première et seconde populations comprend en outre l'étape de tri de la partie restante des cellules de l'approvisionnement dans une première population contenant des cellules ayant la caractéristique A et un second groupe contenant des cellules ayant la caractéristique B, à condition que les cellules triées à partir de l'au moins une partie des cellules satisfassent ou dépassent une référence prédéfinie.
 
5. Procédé selon la revendication 4, caractérisé en ce que la référence prédéfinie est un taux de récupération minimal des cellules ayant la caractéristique A ou la caractéristique B ou une pureté minimale pour une moins une des populations.
 
6. Procédé selon la revendication 1, caractérisé en ce que l'étape de tri d'au moins une partie des cellules dans des première et seconde populations comprend en outre les étapes consistant à :

a) trier les cellules à partir d'une seconde partie de l'approvisionnement dans différentes populations, notamment une première population contenant des cellules ayant la caractéristique A et une seconde population contenant des cellules ayant la caractéristique B ; et

b) mélanger une des populations de la première partie avec une des populations de la seconde partie pour obtenir une population mélangée.


 
7. Procédé selon la revendication 6, caractérisé en ce que la première population de l'au moins une partie a une pureté inacceptable, et en ce que la première population de la seconde partie a une pureté acceptable, et en ce que la population mélangée a une pureté acceptable.
 
8. Procédé selon la revendication 1, caractérisé en ce que l'étape de tri d'au moins une partie des cellules comprend en outre l'étape d'amincissement d'un faisceau laser à l'aide d'un système d'amincissement et de guidage du faisceau dans soit :

a) trois ou plus ;
soit

b) un nombre quelconque de faisceaux distincts qui sont chacun dirigés dans un système optique associé à chacune des multiples unités de cytométrie de flux.


 
9. Procédé selon la revendication 1, caractérisé en ce que les unités de cytométrie de flux partagent un boîtier commun pour les multiples unités de cytométrie de flux.
 




Drawing























































































































































































































































































































































































































Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description




Non-patent literature cited in the description