This invention relates to the fields of molecular biology, virology and gene therapy. In particular, the invention relates to the treatment or prophylaxis of a retinal disorder due to mutations in Crumbs homologue-1 (CRB1) in a human subject, using a gene therapy vector comprising a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein.

Leber's congenital amaurosis (LCA) is the earliest and most severe group of inherited retinopathies, with an incidence of two to three in 100 000 people, and is the most common cause of congenital blindness in children. It is an autosomal recessive or dominant condition recognized in infants during the first months of life. LCA type 8 (LCA8) is inherited recessively. Symptoms of LCA8 patients include nystagmus, slow pupil response, retinal dysfunction, impaired vision and ultimately blindness. The eyes of these blind or severely visually impaired infants show an apparent normal fundus but lack of retinal activity as measured by electroretinography (ERG). At least some LCA8 patients show a thicker retina than normal or other LCA patients. LCA8 patients have mutations or DNA alterations in or affecting the CRB 1 gene locus. LCA8 patients account for 10-15% of all LCA patients.

Retinitis pigmentosa (RP) is an inherited and severe group of degenerative eye disease that occurs at 1 in 3 000 people and causes severe vision impairment and often results in complete blindness. Severe recessive progressive retinitis pigmentosa occurs in young children that have mutations or DNA alterations in or affecting the Crumbs homologue-1 (CRB1) gene locus. These young children become gradually blind before their twentieth birth day. There is no clear relation between genotype (type of mutation) and phenotype (LCA or RP). RP due to mutations in the CRB1 gene (RP12) account for 3-5% of all RP patients.

Mutations in CRB1 cause recessive retinitis pigmentosa with para-arteriolar preservation of the RPE (PPRPE), recessive retinitis pigmentosa, recessive Leber congenital amaurosis, or dominant pigmented paravenous chorioretinal atrophy. The symptoms may include Coats-like exudative vasculopathy; mutations result in a thickened retina with abnormal lamination. Genetic analyses showed that LCA is mostly monogenic, but caused by more than 20 genes, including CRB1 (∼10-15% of all cases), CEP290 (∼20% of all cases), GUCY2D (∼15% of all cases), IMPDH1 (∼10% of all cases), RPE65 (∼5% of all cases), and the less frequently occurring AIPL1, RPGRIP1, RDH12, NMNAT1, SPATA7, LCA5, CRX, TULPI, MERTK, LRAT, RD3, OTX, CABP4, KCNJ13, IQCB1 and others (den Hollander et al., 2008. Prog Retin Eye Res. 27(4):391-419). Mutations in the CRB1 gene are a leading cause of LCA (10-15% of all cases). Genetic analyses showed that RP is caused by more than 50 genes. Mutations in the CRB1 gene account for 3-5% of all cases of RP (RP12). The number of patients with LCA8 or RP12 due to mutations in CRB1 is about the same as patients with LCA1 due to mutations in the guanylate cyclase 2D (GUCY2D) gene and even twice that of patients with LCA2 due to mutations in the gene coding for retinal pigment epithelium 65 kD protein (RPE65).

It is estimated that more than 100 000 world citizens have type 8 Leber's congenital amaurosis or RP12. The CRB1 gene encodes Crumbs homologue-1 (CRB1) which is expressed in the human retina in photoreceptors and Müller glia cells, as well as in retinal progenitor cells, and localizes adjacent to the adherens junction complex at the outer limiting membrane. CRB1 regulates directly or indirectly the physical interaction between retinal cell types. In the adult human retina, loss of CRB1 results in loss of adhesion between Müller glia cells and photoreceptors, resulting in structural changes such as loss of regular lamination. Ultimately, this leads to loss of rod photoreceptors followed by loss of cone photoreceptors by cell death. In the developing retina, loss of CRB 1 results in loss of adhesion between retinal progenitor cells and newly differentiated photoreceptors and Müller glia cells. Ultimately, the misplaced cells do not form a functional neuronal network and undergo cell death. Loss of CRB1 in the developing retina also results in an increase in number of late born retinal cells (rod photoreceptors, Müller glia cells, bipolar cells, late born sub-types of amacrine cells) and an increase in mislocalized retinal cells causing an immature appearance of the retina.

Presently, there are no therapeutics or effective treatments available to prevent, delay or treat LCA8 or RP12 in humans. Therefore, there is a need in the art for methods and means for the treatment of retinal disorders due to mutations in CRB 1. Preferably, the methods and means have no toxicity or almost no toxicity. In particular, the present invention has sought to provide a gene therapy vector to be used for the treatment of retinal disorders due to mutations in CRB 1.

In a first aspect, the present invention relates to a gene therapy vector for use in treatment or prophylaxis of a retinal disorder due to mutations in CRB1 gene in a human subject, wherein the gene therapy vector comprises a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein.

In one embodiment, the retinal disorder is Leber's congenital amaurosis or retinitis pigmentosa, preferably LCA8 or RP12.

In one embodiment, the CRB2 protein is a eumetazoan CRB2 protein, preferably a CRB2 protein of human, non-human primate, murine, feline, canine, porcine, ovine, bovine, equine, caprine, or lupine origin, more preferably the CRB2 protein is a human CRB2 protein.

In one embodiment, the gene therapy vector is a recombinant parvoviral vector or a lentiviral vector, more preferably wherein the vector is a recombinant adeno-associated virus (rAAV) vector.

In one embodiment, the gene therapy vector is a recombinant adeno-associated virus vector selected from the group consisting of recombinant adeno-associated virus serotype 1 (rAAV1), recombinant adeno-associated virus serotype 2 (rAAV2), recombinant adeno-associated virus serotype 3 (rAAV3), recombinant adenoassociated virus serotype 4 (rAAV4), recombinant adeno-associated virus serotype 5 (rAAV5), recombinant adeno-associated virus serotype 6 (rAAV6), recombinant adeno-associated virus serotype 7 (rAAV7), recombinant adeno-associated virus serotype 8 (rAAV8), recombinant adeno-associated virus serotype 9 (rAAV9), serotype variants, for example for enhanced transduction of Müller glia cells, such as rAAV6 ShH10 and ShH10Y, and combinations thereof.

In one embodiment, the CRB2 protein comprises or consists of an amino acid sequence that has at least 80% sequence identity with the amino acid sequences of any one of SEQ ID NO: 40-63 or 65-83, more preferably any one of SEQ ID NO: 40-42, and wherein preferably the CRB2 protein is functionally active as measured by electroretinography.

In one embodiment, the nucleotide sequence encoding CRB2 is operably linked to expression control elements comprising a promoter that produces sufficient expression of CRB2 to obtain a therapeutic effect, wherein the promoter preferably is selected from the group consisting of: truncated CMV promoter, CMV promoter, truncated human RLBP1 promoter, human photoreceptor specific rhodopsin kinase promoter, and human rod photoreceptor specific rhodopsin promoter, wherein preferably the promoter is selected from the group consisting of: CMV promoter according to SEQ ID NO: 121, truncated human RLBP1 promoter according to SEQ ID NO: 122, human photoreceptor specific rhodopsin kinase promoter according to SEQ ID NO: 123, human rod photoreceptor specific rhodopsin promoter according to SEQ ID NO: 124 and truncated CMV promoter according to SEQ ID NO: 133.

In a second aspect, the present invention relates to an AAV vector comprising a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein and at least one parvoviral inverted terminal repeat (ITR) sequence, wherein the nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein is operably linked to expression control elements comprising a promoter that is capable of sufficient CRB2 protein expression to obtain a therapeutic effect.

It also relates to the AAV vector comprising a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein and at least one parvoviral inverted terminal repeat (ITR) sequence, for use as a medicament, wherein the nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein is operably linked to expression control elements comprising a promoter that is capable of sufficient CRB2 protein expression to obtain a therapeutic effect.

In a third aspect, the present invention relates to a virion comprising the AAV vector according to the invention.

In a fourth aspect, the present invention relates to a host cell comprising the AAV vector according to the invention.

In a fifth aspect, the present invention relates to a pharmaceutical composition comprising an AAV vector according to the invention, or a virion according to the invention, and a pharmaceutically acceptable excipient.

In a sixth aspect, the present invention relates to a kit comprising: (a) an AAV vector according to the invention, a virion according to the invention, or a pharmaceutical composition according to claim the invention; and, (b) optionally, instructions for using the AAV vector, virion or pharmaceutical composition according to (a) in the prevention, treatment, or amelioration of one or more symptoms of a retinal disorder due to mutations in CRB1 gene.

In this document and in its claims, the verb "to comprise" and its conjugations is used in its non-limiting sense to mean that items following the word are included, but items not specifically mentioned are not excluded. In addition, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one".

An "insect cell", as used herein, refers to an insect cell which allows for replication of a recombinant parvoviral (rAAV) vector and which can be maintained in culture. For example, the cell line used can be from Spodoptera frugiperda, Drosophila cell lines, or mosquito cell lines, e.g., Aedes albopictus-derived cell lines. Preferred insect cells or cell lines are cells from the insect species which are susceptible to baculovirus infection, including, e.g., Se301, SeIZD2109, SeUCR1, Sf9, Sf900⁺, Sf21, BTI-TN-5B1-4, MG-1, Tn368, HzAm1, Ha2302, Hz2E5, High Five (Invitrogen, CA, USA) and expresSF^+® (US 6,103,526; Protein Sciences Corp., CT, USA). Growing conditions for insect cells in culture, and production of heterologous products in insect cells in culture are well-known in the art and described, e.g., in the following references on molecular engineering of insect cells. Methodology for molecular engineering and expression of polypeptides in insect cells is described, for example, in Summers & Smith (1987). A manual of methods for baculovirus vectors and insect culture procedures (Vol. 1555, Bulletin - Texas Agricultural Experimental Station). College Station, TX; Luckow (1991). Cloning and expression of heterologous genes in insect cells with baculovirus vectors. InProkop et al., Recombinant DNA Technology & Applications (pp. 97-152). New York, NY: McGraw-Hill, Inc.; King & Possee (1992). The baculovirus expression system - A laboratory guide. Chapman and Hall, London; O'Reilly, Miller & Luckow (1992). Baculovirus expression vectors: A laboratory manual. New York, NY: Oxford University Press; Freeman & Richardson (1995). Baculovirus Expression Protocols (Vol. 39, Methods in molecular biology). Totawa, NJ: Humana Press; US 4,745,051; US 2003-0148506; and WO 2003/074714.

As used herein, the term "operably linked" refers to a linkage of polynucleotide (or polypeptide) elements in a functional relationship. A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For instance, a transcription regulatory sequence is operably linked to a coding sequence if it affects the transcription of the coding sequence. Operably linked means that the DNA sequences being linked are typically contiguous and, where necessary to join two protein encoding regions, contiguous and in reading frame. The phrase "under control of" is used interchangeably herein.

"Expression control sequence" refers to a nucleic acid sequence that regulates the expression of a nucleotide sequence to which it is operably linked. An expression control sequence is "operably linked" to a nucleotide sequence when the expression control sequence controls and regulates the transcription and/or the translation of the nucleotide sequence. Thus, an expression control sequence can include promoters, enhancers, internal ribosome entry sites (IRES), transcription terminators, a start codon in front of a protein-encoding gene, splicing signal for introns, and stop codons. The term "expression control sequence" is intended to include, at a minimum, a sequence whose presence are designed to influence expression, and can also include additional advantageous components. For example, leader sequences and fusion partner sequences are expression control sequences. The term can also include the design of the nucleic acid sequence such that undesirable, potential initiation codons in and out of frame, are removed from the sequence. It can also include the design of the nucleic acid sequence such that undesirable potential splice sites are removed. It includes sequences or polyadenylation sequences (pA) which direct the addition of a polyA tail, i.e., a string of adenine residues at the 3'end of a mRNA, sequences referred to as polyA sequences. It also can be designed to enhance mRNA stability. Expression control sequences which affect the transcription and translation stability, e.g., promoters, as well as sequences which effect the translation, e.g., Kozak sequences, are known in insect cells. Expression control sequences can be of such nature as to modulate the nucleotide sequence to which it is operably linked such that lower expression levels or higher expression levels are achieved.

As used herein, the terms "promoter" or "transcription regulatory sequence" refer to a nucleic acid fragment that functions to control the transcription of one or more coding sequences, and is located upstream with respect to the direction of transcription of the transcription initiation site of the coding sequence, and is structurally identified by the presence of a binding site for DNA-dependent RNA polymerase, transcription initiation sites and any other DNA sequences, including, but not limited to, transcription factor binding sites, repressor and activator protein binding sites, and any other sequences of nucleotides known to one of skill in the art to act directly or indirectly to regulate the amount of transcription from the promoter. A "constitutive" promoter is a promoter that is active in most tissues under most physiological and developmental conditions. An "inducible" promoter is a promoter that is physiologically or developmentally regulated, e.g., by the application of a chemical inducer. A "tissue-specific" promoter is only active in specific types of tissues or cells.

The terms "substantially identical", "substantial identity", "% identity" or "essentially similar" or "essential similarity" mean that two peptide or two nucleotide sequences, when optimally aligned, such as by the programs GAP or BESTFIT using default parameters, share at least a certain percentage of sequence identity as defined elsewhere herein. GAP uses the Needleman and Wunsch global alignment algorithm to align two sequences over their entire length, maximizing the number of matches and minimizes the number of gaps. Generally, the GAP default parameters are used, with a gap creation penalty = 50 (nucleotides) / 8 (proteins) and gap extension penalty = 3 (nucleotides) / 2 (proteins). For nucleotides, the default scoring matrix used is nwsgapdna and for proteins the default scoring matrix is Blosum62 (Henikoff & Henikoff, 1992. Proc Natl Acad Sci U S A. 89(22):915-919). It is clear than when RNA sequences are said to be essentially similar or have a certain degree of sequence identity with DNA sequences, thymine (T) in the DNA sequence is considered equal to uracil (U) in the RNA sequence. Sequence alignments and scores for percentage sequence identity may be determined using computer programs, such as the GCG Wisconsin Package, Version 10.3, available from Accelrys Inc., 9685 Scranton Road, San Diego, CA 92121-3752, USA or the open-source software Emboss for Windows (current version 2.7.1-07). Alternatively, percent similarity or identity may be determined by searching against databases such as FASTA, BLAST, etc.

As used herein, the term "nucleic acid construct" is intended to mean a nucleic acid molecule (typically comprised of DNA) operably linked to expression control elements, such as, for example, a promoter that is capable of expression of the nucleic acid molecule.

As used herein, the term "gene therapy vector" is generally intended to mean a nucleic acid molecule (typically comprised of DNA) capable of replication in a host cell and/or to which another nucleic acid segment can be operatively linked so as to bring about replication of the attached segment. A virus is a gene therapy vector. As used herein, the term "vector" refers to a genetic construct that is composed of genetic material (i.e., nucleic acids). Vectors may include one or more genetic elements as described herein arranged such that an inserted coding sequence can be transcribed and translated in a suitable expression cell. In addition, the vector may include one or more nucleic acid segments, genes, promoters, enhancers, activators, multiple cloning regions, or any combination thereof, including segments that are obtained from or derived from one or more natural and/or artificial sources. A gene therapy vector disclosed herein may be comprised within an infectious viral particle. The terms "viral particle" and "virion" are used interchangeably herein. Thus, the present invention also encompasses virions, as well as host cells that comprise a nucleic acid construct or a gene therapy vector of the invention.

As used herein, the terms "protein", "polypeptide", and "peptide" are used interchangeably, and include molecules that include at least one amide bond linking two or more amino acid residues together. Although used interchangeably, in general, a peptide is a relatively short (e.g., from 2 to about 100 amino acid residues in length) molecule, while a protein or a polypeptide is a relatively longer polymer (e.g., 100 or more residues in length). However, unless specifically defined by a chain length, the terms peptide, polypeptide, and protein are used interchangeably.

As used herein, the term "subject" (also interchangeably referred to as "patient") refers to any subject that can serve as a recipient for a gene therapy vector, a pharmaceutical composition, or a virion of the present invention. In one embodiment, the subject is a human being). In certain embodiments, a "patient" refers to any animal host, including, but not limited to, human and non-human primates, bovines, canines, caprines, cavines, corvines, epines, equines, felines, hircines, lapines, leporines, lupines, murines, ovines, porcines, racines, vulpines, and the like, including, without limitation, domesticated livestock, herding or migratory animals, exotics or zoological specimens, as well as companion animals, pets, and any animal under the care of a veterinary practitioner.

As used herein, "an effective amount" would be understood by those of ordinary skill in the art to provide a therapeutic, prophylactic, or otherwise beneficial effect to a recipient subject.

The term "isolated" refers to material that is substantially, or essentially, free from components that normally accompany the material as it is found in its native state. Thus, isolated polynucleotides in accordance with the invention preferably do not contain materials normally associated with those polynucleotides in their natural, or in situ, environment.

The present invention relates to the treatment of a retinal disorder or a disorder associated with cellular changes in the retina, in particular changes due to one or more mutations in CRB1. More specifically, the present invention relates to the treatment of Leber's congenital amaurosis (LCA), in particular Leber's congenital amaurosis-8 (LCA8), and to the treatment of progressive retinitis pigmentosa (RP), in particular progressive retinitis pigmentosa 12 (RP12), or alternatively said early onset RP12. The invention provides a method for at least in part decreasing loss of retinal activity and structural integrity in an animal, wherein the loss of retinal activity and structural integrity comprising at least in part loss of Crumbs homologue (CRB) function in said animal. Said decreasing of loss of retinal activity and structural integrity is accomplished via a recombinant adenoassociated viral (rAAV) expression vector expressing a first nucleic acid segment encoding a first therapeutic gene product that express a biologically-functional Crumbs homologue (CRB) peptide, polypeptide, or protein for use in one or more investigative, diagnostic and/or therapeutic regimens, including for example, the treatment of one or more disorders or diseases of the mammalian eye, and in particular, for treating congenital retinal blindness including, retinal dystrophy such as Leber's congenital amaurosis type 8 (LCA8) and retinitis pigmentosa (RP) due to lack of sufficient biological Crumbs homologue (CRB) function, in humans. It is preferred that the treatment is not or almost not toxic.

In a first aspect, the present invention relates to a gene therapy vector for use in treatment or prophylaxis of a retinal disorder due to mutations in CRB1 gene in a human subject, wherein the gene therapy vector comprises a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein.

The full length CRB1 (for example, SEQ ID NOs: 1-2) cannot usually be used in adenoassociated virus because of its size, although recently it has been shown that it is possible to express full length CRB 1 using AAV9 when using a small promoter, such as a truncated CMV (minimal CMV; preferably SEQ ID NO: 133) or hGRK1 promoter (Pellissier et al., 2014. Mol Ther Methods Clin Dev. 1:14009). The size of normal full length CRB 1 cDNA is about 4.22 kb. A vector comprising this sequence and also comprising other expression elements such as a CMV promoter, inverted terminal repeats and 5' untranslated region would approximately be 5.2 kb. Since the normal genome of the preferred vector of this invention, AAV, is 4.7 kb, the recombinant genomes that are larger than 4.9 kb are not packaged correctly in the capsid, thereby often resulting in defective viruses. As a consequence, it was considered that also the shortened version of CRB1 may be useful in gene therapy, since the short variant occurs in several species and with respect to its protein structure resembles CRB2. However, the Inventors found unexpectedly that both the naturally occurring short variant of CRB 1 (lacking EGF domains as compared to full-length CRB 1; short CRB 1 or sCRB 1 is shown in SEQ ID NO: 3) and also CRB3A (SEQ ID NO: 84) were toxic (tested in AAV5 and AAV9 capsid, see Example 1 as presented herein, wherein it is shown that expression of short human CRB1 in immune naive CRB1 knockout retina is toxic). Therefore, it was preferred that the gene therapy vector of the present invention does not comprise a nucleotide sequence encoding naturally-occurring CRB 1 and/or CRB3 protein and does not comprise a nucleotide sequence encoding a naturally occurring short variant of CRB 1 (lacking EGF domains as compared to full-length CRB 1). Examples of these sequences are provided in the sequence listing (e.g., SEQ ID NO: 1-39 and 64 for CRB 1 and SEQ ID NO: 84-120 for CRB3). Even more surprisingly, the Inventors found that CRB2 did not result in a significant toxic effect as seen with the naturally occurring short variant of CRB 1 and with CRB3.

The Inventors' analysis of mice lacking CRB1, mice lacking CRB2, mice lacking CRB 1 with reduced levels of CRB2, mice lacking CRB2 with reduced levels of CRB 1, and mice lacking both CRB 1 and CRB2 suggest very similar functions for CRB 1 and CRB2. Similarly, the functions of Crumbs homologue (CRB) proteins are exchangeable e.g., the human CRB1 protein can rescue partially the phenotype in fruit flies lacking Crumbs (Crb) protein (den Hollander et al., 2001. Hum Mol Genet. 10(24):2767-73), and the zebrafish CRB2B protein can rescue the phenotype in zebrafish lacking CRB2A protein (Omori & Malicki, 2006. Curr Biol. 16(10):945-57). Other advantages of endogenously or exogenously increasing levels of human CRB2 protein are the following:

1. A) human CRB2 cDNA is small, about 3.9 kb, resulting in an expression cassette comprising the CRB2 cDNA and expression elements of typically only about 4.9 kb. It was found that expression of human CRB2 using a parvoviral vector can be obtained in the retina.
2. B) Native CRB2 is present in the retina of mice both in photoreceptor cells and in Müller glia cells, and possibly also in retinal pigment epithelial cells. Also, in other species, native CRB2 is present and functional in photoreceptor cells. However, in humans, native CRB2 only is present in Müller glia cells, more specifically at the subapical region adjacent to adherens junctions at the outer limiting membrane in Müller glia cells, but not in photoreceptor cells. It is acknowledged that the situation between mice and man differs (Figs. 1 and 2). Several mouse models have been developed and experiments were performed in several mouse models. In the Examples presented herein, it has been illustrated that conditional knockout mice have been developed that have a similar phenotype as presented in humans suffering from RP12.
3. C) The immune system in humans who are deficient in CRB 1 may recognize recombinant CRB 1 as a non-self protein and possibly an immuno reaction against the recombinant CRB 1 is incurred. CRB2 is recognized as a self protein since it is already expressed and immune-tolerated in the retinas and epithelia of other organs of these patients and will not result in an immune response.

The cause of the toxicity of full length and short CRB1 as referred to above is yet unknown and may be on the level of DNA, RNA or protein. Without wishing to be bound by any theory, it is for example possible that overexpression of short CRB1 protein scavenges essential proteins or that the RNA transcript of short CRB1 cDNA results in a disbalance in microRNAs.

Toxicity of CRB proteins was tested using human-derived retinal pigment epithelial cells as indicated in Example 3, or using electroretinography on mouse retinas as indicated in Example 6 presented herein.

In one embodiment, mutations in the CRB1 gene result in loss of CRB 1 functional protein, as can be determined using electroretinography (ERG), multi-focal ERG, optical coherence tomography (OCT), microperimetry, visual evoked potention (VEP) test, functional magnetic resonance imaging test, or behaviour maze-test (see, for example: Bainbridge et al., 2008. N Engl J Med. 358(21):2231-9; Annear et al., 2011. Gene Ther. 18(1):53-61; Maguire et al., 2008. N Engl J Med. 358(21):2240-8; Testa et al., 2013. Ophthalmology. 120(6):1283-91; Cideciyan et al., 2008. Proc Natl Acad Sci USA. 105(39):15112-7; Watkins et al., 2012. Brain. 135(Pt 5):1566-77). These methods provide a quantifiable way to measure the regression or progression of retinal visual function in the examined eye. Preferably, the retinal disorder is Leber's congenital amaurosis or retinitis pigmentosa, more preferably LCA8 or RP12.

Retinitis pigmentosa (RP) is an autosomal recessive or dominant group of diseases that represent progressive or late severe forms of inherited retinal dystrophies affecting initially rod photoreceptors and subsequently cone photoreceptors. The twelveth gene, located on the long (q) arm of chromosome 1 between positions 31 and 32.1, implicated in the onset of this genetically and clinically heterogeneous, and therefore assigned to the RP12 gene locus (van Soest et al., 1994. Genomics. 22(3):499-504), was Crumbs homologue-1 (CRB1) ( den Hollander et al., 1999. Nat Genet. 23(2):217-21). The RP12 gene caused RP with preserved para-arteriolar retinal pigment epithelium (PPRPE) (Heckenlively et al., 1982. Br J Ophthalmol. 66(1):26-30). Some genes that cause retinitis pigmentosa also cause Leber congenital amaurosis. The CRB1 gene was also implicated in Leber congenital amaurosis (LCA) type 8 (LCA8) and progressive types of RP without PPRPE (den Hollander et al., 2004. Hum Mutat. 24(5):355-69). LCA is an autosomal recessive or dominant group of diseases that represent the earliest and most severe form of all inherited retinal dystrophies. The RP12 or LCA8 gene encodes for Crumbs homologue-1 (CRB1) which is expressed predominantly in photoreceptors and Müller glia cells at a subapical region adjacent to adherens junctions at the outer limiting membrane in the retina. CRB 1 plays a role in the formation and maintenance of adhesion between photoreceptors and Müller glia cells. Without CRB 1 protein, adhesion between these cells is weakened, leading to loss of normal retinal lamination. Without CRB1 protein, the subapical CRB 1/PALS 1/MUPP1 and CRB 1/PALS 1/PATJ protein complexes, required for maintaining cellular polarization and maintenance of adhesion between photoreceptors and Müller glia cells, are destabilized. Mutations in the CRB1 gene reduce or abolish the ability of CRB 1 protein to maintain the subapical CRB 1/PALS 1/MUPP1 and CRB 1/PALS 1/PATJ protein complexes and to maintain the adhesion between photoreceptors and Müller glia cells, as in RP with mutations in the CRB1 gene or LCA8. It is unclear why some people with CRB1 gene mutations have severe, early visual impairment associated with Leber congenital amaurosis, and other people experience more gradual vision loss and other eye problems associated with retinitis pigmentosa. Other genetic factors (such as CRB2; Alves et al., 2013. Hum Mol Genet. 22(1):35-50) may modify the effects of CRB1 gene mutations to influence the severity of these conditions.

The first report of LCA was published in 1869 by Theodor Leber (Leber, 1869. Albrecht von Graefes Arch Ophthal. 15:1-25). Currently, at least twenty genes have been reported to cause LCA. Mutations in CRB1 account for -15% of all cases of LCA making it one of the leading causes of LCA. Diagnosis of LCA8 is typically made within the first few months of life in an infant with severely impaired vision or total blindness, a flat electroretinogram (ERG) and involuntary eye movements (nystagmus) (Hufnagel et al., 2013. Graefes Arch Clin Exp Ophthalmol. 250(8):1117-28). Loss of normal retinal structure in LCA8 is unlike other forms of the disease which exhibit marked retinal thinning that generally worsens with age (Pasadhika et al., 2010. Invest Ophthalmol Vis Sci. 51(5):2608-14) or exhibit preserved retinal structure with loss of retinal activity as is the case for LCA1 due to mutations in the Gucy2d gene. Using spectral-domain optical coherence tomography (SDOCT) to scan the central macular and perifoveal areas, a study revealed that LCA8 patients typically show a thicker retina with loss of the 6 retinal layers and immature appearance, compared to people without eye-disease or patients with other types of LCA such as LCA2 (Jacobson et al., 2003. Hum Mol Genet. 12(9):1073-8).

Less severe retinal degeneration is observed in visually-impaired patients lacking a functional CRB1 gene due to retinitis pigmentosa. Retinitis pigmentosa is the leading cause of inherited retinal degeneration-associated blindness. Retinitis pigmentosa (RP) is a disease condition that was first identified and named by Dr. Donders in 1857. Retinitis pigmentosa is a group of related conditions that are inherited, progressive and clinically distinctive and share a similar feature of dystrophy or damage to the photoreceptors of the retina and of the pigment epithelium underneath the photoreceptors. Currently, at least 50 genes have been reported to cause dominant or recessive RP. Around 30-40% are autosomal dominant, 50-60% are autosomal recessive and 5-15% are X-chromosome linked. The prevalence is 1 in 4 000 among all age groups and 1 in 3 000 persons in population younger than 65 years of age. Mutations in CRB1 account for -3-5% of all cases of RP, making it one of the leading causes of RP. The number of patients affected by recessive mutations in the CRB1 gene (RP and LCA) is approximately double that affected by mutations in the RPE65 gene (LCA type 2 or LCA2), for which successful AAV-mediated gene therapy trials have been described. Diagnosis of RP patients is typically made within the first decades of life with initial vision problems especially in dim light. This manifests as a loss of vision around the peripheries, known as tunnel vision. The central vision is spared until the later stages of the disease. RP12 patients typically show preserved para-arteriole retinal pigment epithelium (PPRPE) (Heckenlively, 1982. Br J Ophthalmol. 66(1):26-30). The preservation of retinal structure in patients with RP compared to patients with LCA8 due to mutations in the CRB1 gene does suggest that they are better suited for future therapeutic strategies, but timely expression of the CRB1 gene in LCA8 patients will rescue the structure and function of LCA8 retinas as well.

Loss of CRB1 function in humans leads to progressive RP12 or LCA8, though loss of CRB 1 function in mice leads to relative mild retinal disorganization and degeneration. It is unclear why some people with CRB1 gene mutations have severe, early visual impairment associated with Leber congenital amaurosis, and other people experience more gradual but progressive early onset vision loss and other eye problems associated with retinitis pigmentosa. It is also unclear why mice lacking CRB 1 show a relative mild phenotype compared to humans lacking CRB1. Other genetic factors (such as CRB2; Alves et al., 2013. Hum Mol Genet. 22(1):35-50) may modify the effects of Crb1 gene mutations to influence the severity of these conditions. Indeed, mice lacking CRB2 in the retina show a phenotype mimicking progressive RP detected in human patients lacking CRB1 (Alves et al., 2013. Hum Mol Genet. 22(1):35-50), and mice lacking CRB2 and CRB1 mimic LCA8 detected in human patients lacking CRB1 (Pellissier et al., 2013. PLoS Genet. 9(12):e1003976). Other factors involved are light exposure; exposure to moderate levels of white light significantly increased the level of retinal disorganization and degeneration in mice lacking CRB1 (van de Pavert et al., 2004. J Cell Sci. 117(Pt 18):4169-77; van de Pavert et al., 2007. J Neurosci. 27(3):564-73; van de Pavert et al., 2007. Glia. 55(14): 1486-97).

In part, the phenotypes in mice and humans may differ because of different localization of CRB1 and CRB2 proteins. In the mouse retina, immuno electron microscopy showed that CRB1 localizes in the apical villi of Müller glia cells at the subapical region (SAR) adjacent to adherens junctions (AJ) at the outer limiting membrane (OLM). In the mouse retina, CRB2 localizes at two regions: the inner segments of photoreceptors at the subapical region (SAR) adjacent to adherens junctions (AJs) at the outer limiting membrane (OLM), as well as at the apical villi of Müller glia cells at the subapical region (SAR) adjacent to adherens junctions (AJs) at the outer limiting membrane (OLM) (van Possum et al., 2006. Hum Mol Genet. 15(18):2659-72). Loss of CRB 1 in the mouse retina therefore leaves functional CRB2 protein in photoreceptors and Müller glia cells, resulting in a mild phenotype.

In the human retina, immuno electron microscopy showed that CRB2 localizes in the apical villi of Müller glia cells at the subapical region (SAR) adjacent to adherens junctions (AJ) at the outer limiting membrane (OLM). In the human retina, CRB 1 localizes at two regions: the inner segments of photoreceptors at the subapical region (SAR) adjacent to adherens junctions (AJs) at the outer limiting membrane (OLM), as well as at the apical villi of Müller glia cells at the subapical region (SAR) adjacent to adherens junctions (AJs) at the outer limiting membrane (OLM) (Pellissier et al., 2014. Hum Mol Genet. 23(14):3759-71). Loss of CRB1 in the human retina therefore leaves functional CRB2 protein at the SAR in Müller glia cells but not in photoreceptors, resulting in a severe phenotype.

Three mouse models carrying mutations in the Crb1 gene have been described. Homozygote mice with the naturally occurring rd8 mutation in the Crb1 gene (Crb1^rd8/rd8) show mild retinal degeneration, preferentially in one quadrant of the retina, the inferior (ventral) nasal quadrant (Aleman et al., 2011. Invest Ophthalmol Vis Sci. 52(9):6898-910; Mehalow et al., 2003. Hum Mol Genet. 12(17):2179-89). Homozygote knock-out mice lacking any CRB 1 protein (Crb1^-l-) show mild retinal degeneration, preferentially in one quadrant of the retina, the inferior (ventral) temporal quadrant (van de Pavert et al., 2004. J Cell Sci. 117(Pt 18):4169-77; van de Pavert et al., 2007. J Neurosci. 27(3):564-73). Heterozygote knock-in/heterozygote knock-out mice expressing no wild-type mouse CRB1 but expressing CRB1 with a substitution at position 249 of amino-acid tryptophane (W) for a cystein (C) (Crb1^C249W/-) show very late-stage mild retinal degeneration (van de Pavert et al., 2007. Glia. 55(14):1486-97). Crb1^C249W/ mice were developed as a mouse model for RP12 patients with a C250W substitution in both alleles.

Importantly, in none of the 3 mouse models (Crb1^rd8/rd8, Crb1^-/-, or Crb1^C249W/-) was there a significant decrease in retinal function as measured by electroretinography (Aleman et al., 2011. Invest Ophthalmol Vis Sci. 52(9):6898-910; Mehalow et al., 2003. Hum Mol Genet. 12(17):2179-89; van de Pavert et al., 2004. J Cell Sci. 117(Pt 18):4169-77; van de Pavert et al., 2007. J Neurosci. 27(3):564-73; van de Pavert et al., 2007. Glia. 55(14):1486-97). Therefore, these mice appeared not suitable for testing efficacy of CRB gene therapy vectors as measured by electroretinography.

Several other newly developed mouse models are useful for evaluating gene replacement therapy. First, the Crb2 conditional knock-out mouse (Crb2 cKO) lacking CRB2 in all retinal cells except the retinal pigment epithelium (e.g., the Crb2^flox/floxChx10Cre) (Alves et al., 2013. Hum Mol Genet. 22(1):35-50). Second, the Crb2 conditional knock-out mouse lacking CRB2 in photoreceptors (e.g., the Crb2^flox/floxCrxCre) (Alves et al., 2014. Hum Mol Genet. 23(13):3384-401). Third, the Crb2 conditional knock-out mouse lacking CRB2 in Müller glia cells (e.g., Crb2^flox/floxPdgfrαCre) (Alves et al., 2014. Hum Mol Genet. 23(13):3384-401). Fourth, the homozygote Crbl/heterozygote Crb2 conditional knock-out mouse lacking CRB 1 in all retinal cells and having reduced expression of CRB2 in all retinal cells except the retinal pigment epithelium (e.g., Crb1^-/- Crb2^flox/+Chx10Cre) (Pellissier et al., 2014. Hum Mol Genet. 23(14):3759-71). Fifth, the homozygote Crbl/heterozygote Crb2 conditional knock-out mouse lacking CRB 1 in all retinal cells and having reduced expression of CRB2 in photoreceptors (e.g., Crb1^-/- Crb2^flox/+CrxCre). Sixth, the homozygote Crb1/heterozygote Crb2 conditional knock-out mouse lacking CRB 1 in all retinal cells and having reduced expression of CRB2 in Müller glia cells (e.g., Crb1^-/-Crb2^flox/+PdgfraCre). Seventh, the homozygote Crb1/homozygote Crb2 conditional knock-out mouse lacking CRB 1 in all retinal cells and CRB2 in all retinal cells except the retinal pigment epithelium (e.g., Crb1^-l-Crb2^flox/floxChx10Cre) (Pellissier et al., 2013. PLoS Genet. 9(12):e1003976). Eighth, the homozygote Crbl/homozygote Crb2 conditional knock-out mouse lacking CRB1 in all retinal cells and CRB2 in photoreceptors (e.g., Crbl^-/-Crb2^flox/floxCrx10Cre). Ninth, the homozygote Crbl/homozygote Crb2 conditional knock-out mouse lacking CRB1 in all retinal cells and CRB2 in Müller glia cells (e.g., Crbl^-/-Crb2^flox/floxPdgfrαCre).

The Crb2 conditional knock-out mouse lacking CRB2 in all retinal cells except the retinal pigment epithelium (e.g., the Crb2^flox/floxChx10Cre) (Alves et al., 2013. Hum Mol Genet. 22(1):35-50) was used to evaluate gene replacement therapy. The Crb2^flox/floxChx10Cre exhibits progressive retinal degeneration and scotopic (rod-mediated) and photopic (cone-mediated) loss of retina function as measured by ERG from 1 to 6 months of age (Alves et al., 2013. Hum Mol Genet. 22(1):35-50). The mouse is blind at 12-18 months of age. AAV-mediated transfer of CRB2 to Crb2 cKO retina restored vision to these animals as evidenced by ERG. AAV-mediated transfer of CRB2 to the postnatal Crb2 cKO retina expressed CRB2 in photoreceptors and Müller glia cells and caused preservation of retinal structure at the time of expression of CRB2. These experiments showed the feasibility of preserving retinal structure after a single dose of AAV-CRB2 even in severely degenerating Crb2 cKO retinas.

For testing efficacy of a gene therapy vector of the invention in human patients, several standard as well as state of the art techniques are available, such as electroretinography, pupillometry, scanning laser ophthalmoscopy, optical coherence tomography, and behavior tests. Especially, the use of functional magnetic resonance imaging (fMRI), as biomarker for early improvements in cortical visual function upon treatment of the retina with the AAV-CRB gene therapy vector, will be most useful for analyzing and interpreting different dosing regimes.

In one embodiment, a gene therapy vector of the present invention may be used in a combination therapy, for example in combination with:

1. a) addition of protective, nurturing or growth factors such as e.g., GDNF or CTNF,
2. b) addition of drugs that normalize the intraocular pressure in eyes such as e.g., eye drops containing prostaglandin analogs, beta blockers, alpha agonists, and/or carbonic anhydrase inhibitors,
3. c) addition of drugs or tools that decrease the light-sensitivity of eyes such as e.g., prosthetic contact lenses,
4. d) addition of drugs that normalize the retinoid cyclus in the retina such as e.g., retinoids,
5. e) addition of drugs that increase the strength of adherens junctions at the retinal outer limiting membrane such as e.g., magnesium and calcium salts.

In one embodiment, the gene therapy vector of the present invention may be applied only once to a subject suffering from the retinal disorder due to mutations in CRB1 gene. Re-application of the same or a similar vector, for example with the same or another capsid, is expected to become advantageous with signs of decreased vision in the dark. The same or a similar vector may be re-applied, because injection of the AAV vector subretinally provokes low immune response. In contrast, intravitreal injection has been demonstrated to result in an immune response, but in such cases, another suitable vector may be used (Li et al., 2008. Mol Vis. 14:1780-1789). It is envisioned that maze experiments, but preferentially fMRI experiments with visual tasks in dimmed light, by testing the visual cortex, will be most instrumental in determining the time point at which a re-application becomes advantageous.

According to the present invention, the subject suffering from the retinal disorder due to mutations in CRB1 gene and to be treated using a gene therapy vector, an AAV vector, a virion, a host cell or a pharmaceutically composition according to the present invention is a human.

In one embodiment, the CRB2 protein is a eumetazoan CRB2 protein, preferably a CRB2 protein of human, non-human primate, murine, feline, canine, porcine, ovine, bovine, equine, caprine, or lupine origin, more preferably the CRB2 protein is a human CRB2 protein.

In one embodiment, the gene therapy vector is a recombinant parvoviral vector or a lentiviral vector.

Viruses of the Parvoviridae family are small DNA animal viruses. The Parvoviridae family may be divided between two subfamilies: the Parvovirinae, which infect vertebrates, and the Densovirinae, which infect insects. Members of the Parvovirinae subfamily are herein referred to as the parvoviruses and include the genus Dependovirus. As may be deduced from the name of their genus, members of the Dependovirus genus are unique in that they usually require coinfection with a helper virus, such as adenovirus or herpes virus, for productive infection in cell culture.

The Dependovirus genus includes adeno-associated virus (AAV), which normally infects humans (e.g., serotypes 1, 2, 3A, 3B, 4, 5, and 6) or primates (e.g., serotypes 1 and 4), and related viruses that infect other warm-blooded animals (e.g., bovine, canine, equine, and ovine adeno-associated viruses). Today, it is possible to differentiate between the serologically distinguishable types of at least AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8 and AAV9. AAV vectors constitute a single-stranded DNA with an outer icosahedral coat of structural protein having a diameter of 18 to 26 nm, typically about 25 nm. Further information on parvoviruses and other members of the Parvoviridae family is described in Kenneth (1996). "Parvoviridae: the viruses and their replication". In Fields & Knipe, Fields' virology (3rd ed.) New York, NY: Raven Press. For convenience, the present invention is further exemplified and described herein by reference to AAV. It is however understood that the invention is not limited to AAV but may equally be applied to other parvoviruses. It is also understood that the invention extends to AAV chimeric viruses, comprising chimeric capsid proteins and/or AAV hybrid viruses (or pseudotyped viruses) that also have a similar size as found for the wild-type parvoviruses (18-26 nm diameter). A description and some examples are given in WO 2000/028004. Examples of AAV chimeric and/or hybrid viruses are, for example, AAV2/1, AAV2/3, AAV2/4, AAV2/5, AAV2/5.2, AAV2/6, AAV2/7, AAV2/8 and AAV2/9.

The AAV genome consists of rep genes encoding proteins required for replication of the virus and cap genes encoding the viral structural proteins. One or more of the rep genes which are required for replication (e.g., rep40, rep52, rep68 and/or rep78) or the cap genes which are required for the capsid structure (e.g., VP-1, VP-2 and/or VP-3) can, for example, be replaced in the virus with a transgene when preparing adeno-associated vectors. The ITR regions which are still present at the 5' and 3' ends are needed, as cisactive elements, for packaging the transgene into infectious, recombinant AAV particles and for the replication of the DNA of the recombinant AAV genome (Kotin, 1994. Hum Gene Ther. 5(7):793-801). A "recombinant parvoviral or AAV vector" (or "rAAV vector") herein refers to a vector comprising one or more polynucleotide sequences of interest, genes of interest or "transgenes" that are flanked by parvoviral or AAV inverted terminal repeat sequences (ITRs). Such rAAV vectors can be replicated and packaged into infectious viral particles when present in an insect or mammalian host cell that is expressing AAV rep and cap gene products (i.e., AAV Rep and Cap proteins). When an rAAV vector is incorporated into a larger nucleic acid construct (e.g., in a chromosome or in another vector, such as, a plasmid or baculovirus used for cloning or transfection), then the rAAV vector is typically referred to as a "pro-vector" which can be "rescued" by replication and encapsidation in the presence of AAV packaging functions and necessary helper functions.

Thus, in one embodiment, the gene therapy vector of the present invention is an rAAV vector. In one embodiment, the rAAV is selected from the group consisting of recombinant adeno-associated virus serotype 1 (rAAV1), recombinant adeno-associated virus serotype 2 (rAAV2), recombinant adeno-associated virus serotype 3 (rAAV3), recombinant adenoassociated virus serotype 4 (rAAV4), recombinant adeno-associated virus serotype 5 (rAAV5), recombinant adeno-associated virus serotype 6 (rAAV6), recombinant adeno-associated virus serotype 7 (rAAV7), recombinant adeno-associated virus serotype 8 (rAAV8), recombinant adeno-associated virus serotype 9 (rAAV9), serotype variants, for example for enhanced transduction of Müller glia cells, such as rAAV6 ShH10 (Klimczak et al., 2009. PLoS One. 4(10):e7467) and ShH10Y (Dalkara et al., 2011. Mol Ther. 19(9):1602-8), and combinations thereof.

In one embodiment, the nucleotide sequence encodes for a CRB2 protein comprising an amino acid sequence that has at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% preferably at least 100% sequence identity with the amino acid sequences of any one of SEQ ID NO: 40-63 or 65-83, more preferably any one of SEQ ID NO: 40-42, most preferably SEQ ID NO: 40.

Such a CRB2 protein preferably has an intracellular domain of 37 amino acid residues or alternatively an intracellular domain plus transmembrane domain of 63 amino acid residues. Without wishing to be bound by any theory, these domains in particular are considered most relevant for membrane localization and formation of the Crumbs homologue (CRB) protein complex linked to the actin cytoskeleton of the cell, which are believed to be important to rescue the phenotype and non-toxicity.

In one embodiment, the nucleotide sequence encodes for a CRB2 protein consisting of an amino acid sequence that has at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% preferably at least 100% sequence identity with the amino acid sequences of any one of SEQ ID NO: 40-63 or 65-83, more preferably any one of SEQ ID NO: 40-42, most preferably SEQ ID NO: 40.

In one embodiment, the nucleotide sequence encodes for a CRB2 protein comprising or consisting of an amino acid sequence as shown in any one of SEQ ID NO: 40-63 or 65-83, more preferably any one of SEQ ID NO: 40-42, most preferably SEQ ID NO: 40.

In one embodiment, the CRB2 protein comprises a contiguous amino acid sequence that is at least 95% identical to the carboxy (C)-terminal region of 37 contiguous amino acids of a sequence as set forth in SEQ ID NO: 40-63 or 65-83, more preferably any one of SEQ ID NO: 40-42, even more preferably SEQ ID NO: 40.

In one embodiment, a CRB2 protein comprising or consisting of an amino acid sequence as defined herein is a functional or, alternatively said, active CRB2 protein. To test whether a protein comprising or consisting of an amino acid sequence as defined herein is a functional CRB2 protein, electroretinography is performed.

In short, an AAV vector, preferably AAV2/9 or AAV2/5, wherein the capsid is AAV9 or AAV5 and the ITRs are AAV2, is generated to allow expression of the CRB2 protein comprising or consisting of an amino acid sequence as defined herein operably linked to a CMV promoter. A construct can be made according to the Examples as presented herein. The AAV vector is administered subretinally to the retina of Crb2 cKO mice (Crb2^F/FChx10Cre) on postnatal day 14. The contralateral eye receives a control AAV vector which comprises GFP instead of the CRB2 protein to be tested. A positive control animal receives a recombinant AAV expressing CRB2 protein according to SEQ ID NOs: 40, 41 or 42. At 3 months of age or later, i.e., at least 2.5 months after application of the virus, a-wave and b-wave electroretinograms are made as described in Tanimoto et al. (2013. Methods Mol Biol. 935:69-78). Briefly, retinas of anesthetized mice are exposed to light flashes at different intensities (on the x-axis the light intensity expressed as log (cd^∗s/m²)). A CRB2 protein comprising or consisting of an amino acid sequence as defined herein is considered to have CRB2 activity (or to be a functional CRB2 protein) if the maximal b-wave and/or a-wave amplitude (in microvolts) in the electroretinogram is increased by at least 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9 or 2.0-fold as compared to the AAV-GFP treated contralateral retina. More preferably, the AAV vector is administered not on postnatal day 14, but on postnatal day 3 or 4, since this may significantly increase the efficacy of treatment.

In one embodiment, in the gene therapy vector according to the invention, the nucleotide sequence encoding CRB2 is operably linked to expression control elements comprising a promoter that produces sufficient expression of CRB2to obtain a therapeutic effect, wherein the promoter preferably is selected from the group consisting of: CMV promoter, preferably according to SEQ ID NO: 121, CMV promoter, truncated CMV or minimal CMV promoter, truncated human RLBP1 promoter, human photoreceptor specific rhodopsin kinase promoter and human rod photoreceptor specific rhodopsin promoter.

The nucleic acid sequence of an illustrative human Müller glia specific retinaldehyde binding protein-1 (RLBP1) promoter (Pellissier et al., 2014. Mol Ther Methods Clin Dev. 1:14009; similar in part to the published mouse RLBP1 promoter described in Vazquez-Chona et al., 2009. Invest Ophthalmol Vis Sci. 50(8):3996-4003; Vogel et al., 2007. Invest Ophthalmol Vis Sci. 48(8):3872-7) which is preferred for use in the present invention is shown in SEQ ID NO: 122.

The nucleic acid sequence of an illustrative human GRK1 specific promoter (Khani et al., 2007. Invest Ophthalmol Vis Sci. 48(9):3954-61; Boye et al., 2012. Hum Gene Ther. 23(10):1101-15) which is preferred for use in the present invention is shown in SEQ ID NO: 123.

The nucleic acid sequence of an illustrative human RHO specific promoter (Pellissier et al., 2014. Mol Ther Methods Clin Dev. 1:14009) which is preferred for use in the present invention is shown in SEQ ID NO: 124.

The nucleic acid sequence of an illustrative truncated CMV promoter (Pellissier et al., 2014. Mol Ther Methods Clin Dev. 1:14009) is shown in SEQ ID NO: 133.

Particularly preferred gene therapy constructs of the present invention are the following: AAV-hGRK1-CRB2 (specific for rod and cone photoreceptors); AAV-hRHO-CRB2 (specific for rod but not cone photoreceptors); AAV-CMV-CRB2 (allows expression in rod+cone photoreceptors, Müller glia cells, and retinal pigment epithelium); AAV-CMV-CRB2-miRT (that reduces transcription in retinal pigment epithelium); AAV-truncatedRLBP1-CRB2; AAV-truncatedCMV-CRB2. AAV-hGRK1-CRB2 (specific for rod and cone photoreceptors) and AAV-hRHO-CRB2 (specific for rod but not cone photoreceptors) and AAV-CMV-CRB2 are most preferred.

The nucleic acid sequence of illustrative microRNA target sites (miRT's) to lower the expression of AAV transcript containing the miRT sequence in retinal pigment epithelium cells (Karali et al., 2011. PLoS One. 6(7):e22166) which can be used in combination with the present invention is shown in SEQ ID NOs: 125-127. These sequences are the predicted and functional target sites in e.g., the AAV-CMV-CRB2-miRT vector, not the miRNA sequences themselves. miRNAs, that recognize and interfere with the translation of, or degrade, the target CRB2-miRT mRNA transcript, are expressed in the RPE. The skilled person is capable of using such miRTs in the present invention (see, for example, Karali et al., 2011. PLoS One. 6(7):e22166).

In one embodiment, recombinant CRB2 protein is expressed in rod and cone photoreceptor cells, but not in retinal pigment epithelium or Müller glia cells. This can be achieved, for example, by applying the human photoreceptor specific rhodopsin kinase promoter according to SEQ ID NO: 123. As a result, the retina is protected against degeneration. A preferred gene therapy vector of the invention is hGRK1-hCRB2(In5)-spA using AAV2 ITR and AAV5 capsid proteins.

In one embodiment, the gene therapy vector may further comprise, operably linked to the nucleotide sequence encoding the CRB2 protein, one or more of the following: inverted terminal repeats, such as, for example, of any wild-type or mutant AAV; a promoter/enhancer, such as, for example, the CMV promoter/enhancer; a wild-type or synthetic transcription splice donor/acceptor site, such as, for example, In5; a wild-type or synthetic transcription poly-adenylation site, such as, for example, spA; one or more micro RNA target sites to reduce transcriptional activity in retinal cell types, such as, for example, the retinal pigment epithelium. In one embodiment, the gene therapy vector according to the invention comprises a wild-type, mutant or codon-optimized DNA sequence encoding wild-type or mutant CRB2 proteins of any species.

In one embodiment, a wild-type or synthetic transcription splice donor/acceptor site, such as, for example, synthetic intron (In5) may be inserted in the gene therapy vector for stable transcript processing of CRB2. A preferred nucleic acid sequence of an illustrative synthetic intron (In5) in the coding sequence of the Crumbs homologue (CRB) gene is shown in SEQ ID NO: 128. The intron is preferably inserted into CRB2 cDNA between two adjacent exons with a sequence of exon NNNAG/intron/GNNN exon, where G, A, T, C stands for one of the four nucleotides, and N stands for any of the four nucleotides.

In one embodiment, the gene therapy vector according to the invention comprises a nucleotide sequence encoding a CRB2 protein comprising an amino acid sequence that has at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% preferably at least 100% sequence identity with the amino acid sequence of SEQ ID NO: 40 and wherein the promoter is the CMV promoter according to SEQ ID NO: 121.

In one embodiment, the gene therapy vector according to the invention comprises a nucleotide sequence encoding a CRB2 protein comprising an amino acid sequence that has at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% preferably at least 100% sequence identity with the amino acid sequence of SEQ ID NO: 40 and wherein the promoter is the human photoreceptor specific rhodopsin kinase promoter according to SEQ ID NO: 123.

The similarity between the CRB family member sequences from different species is readily recognized by those with skills in the art of the field.

CRB homologue consensus regions, as illustrated by reference to amino acid positions of consensus sequence SEQ ID NO: 77, are: amino acid positions 265-1515, 1555-2068, and 2083-2146. CRB variable regions are amino acid positions 1-264, 1516-1554, and 2069-2082. Other notable regions of the CRB consensus alignment may or may not include:

1. (1) epidermal growth factor like domains at amino acid positions 265 to 301, 304 to 341, 346 to 384, 386 to 423, 425 to 461, 462 to 498, 499 to 530, 543 to 579, 580 to 609, 607 to 644, 646 to 683, 685 to 721, 723 to 759, 761 to 798, 800 to 836, 838 to 900, 902 to 938, 940 to 976, 978 to 1019, 1205 to 1241, 1479 to 1515, 1756 to 1792, 1794 to 1830, 1832 to 1868, 1871 to 1912, 1912 to 1948, 1950 to 1987, 1989 to 2027, 2028 to 2068 of the consensus sequence (known to be essential for activity in polarized cells such as photoreceptors and Müller glia cells - see, e.g., Richard et al., 2006. Hum Mol Genet. 15 Spec No 2:R235-43; van de Pavert et al., 2007. Glia. 55(14):1486-97);
2. (2) laminin G-like domain at amino acid positions 1021 to 1203, 1248 to 1478, and 1555 to 1755 of the consensus sequence (known to be essential for cell adhesion, signalling, migration, assembly, and differentiation activity);
3. (3) a transmembrane domain at amino acid position 2083 to 2109;
4. (4) a highly conserved 37-amino acid C-terminal region containing a FERM-protein and C-terminal PDZ-protein binding motifs at amino acid position 2110 to 2146 (Richard et al., 2006. Hum Mol Genet. 15 Spec No 2:R235-43).

In a second aspect, the present invention relates to a nucleic acid construct, preferably an isolated nucleic acid construct, comprising a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein and at least one parvoviral inverted terminal repeat (ITR) sequence, wherein the nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein is operably linked to expression control elements comprising a promoter that is capable of sufficient CRB2 protein expression to obtain a therapeutic effect.

In a third aspect, the present invention relates to a virion, comprising an AAV vector according to the present invention. In one embodiment, the virion is an AAV virion.

In a fourth aspect, the present invention relates to a host cell comprising an AAV vector according to the invention. In one embodiment, the host cell is a mammalian or an insect host cell as defined herein above. If the host cell is a mammalian host cell, then preferably that host cell is a human host cell.

In a fifth aspect, the present invention relates to a pharmaceutical composition comprising an AAV vector according to the invention, or a virion according to the invention, and a pharmaceutically acceptable excipient.

Pharmaceutically acceptable excipients are well-known by the person skilled in the art. Examples of pharmaceutically acceptable excipients are a buffer, a carrier, a vehicle or a diluent. Preferably, the pharmaceutical composition further comprises one or more of the following: a lipid, a liposome, a lipid complex, an ethosome, a niosome, a nanoparticle, a microparticle, a liposphere, a nanocapsule, or any combination thereof.

In one embodiment, the pharmaceutical composition may be formulated for administration to the human eye. Typically, the pharmaceutical composition is administered by direct injection into the retina or the surrounding tissue. More specifically, the pharmaceutical composition needs to be suitable for subretinal or intravitreal injection and thus needs to be a sterile and isotonic fluid, using NaCl or sugars. In this regard, reference is made to International publications WO 2012/114090 and WO 2011/133933.

In a sixth aspect, the invention relates to a kit comprising: (a) an AAV vector according to the present invention, a virion according to the present invention, or a pharmaceutical composition according to the present invention; and (b) instructions for using the AAV vector or pharmaceutical composition according to (a) in the prevention, treatment, or amelioration of one or more symptoms of a retinal disorder due to mutations in CRB1 gene. SEQ ID NO: Species Gene UniProtKB/Swiss-Prot accession number 1 Homo sapiens CRB 1 isoform 1 P82279-1 2 Homo sapiens CRB 1 isoform 2 P82279-2 3¹ Homo sapiens CRB 1 isoform 3 P82279-3 4 Homo sapiens CRB 1 isoform 4 P82279-4 5 Homo sapiens CRB 1 isoform 5 P82279-5 6 Mus musculus CRB1 isoform 1 Q8VHS2-1 7 Mus musculus CRB 1 isoform 2 Q8VHS2-2 8² Mus musculus CRB 1 isoform 3 Q8VHS2-3 9 Mus musculus CRB 1 isoform 4 Q8VHS2-4 10 Rattus norvegicus CRB 1 isoform 1 D3ZZL8 11 Spermophilus tridecemlineatus CRB1 isoform 1 I3MCW9 12 Pongo abelii³ CRB1 isoform 1 H2N4A7 13 Pan troglodytes CRB1 isoform 1 H2R3X3 14⁴ Nomascus leucogenys CRB1 isoform 1 G1S8V2 15 Gorilla gorilla CRB1 isoform 1 G3RZV2 16 Gorilla gorilla CRB1 isoform 2 G3SK60 17 Macaca mulatta CRB1 isoform 1 H9FS33 18⁵ Macaca mulatta CRB1 isoform 3 F7DFY1 19 Equus caballus CRB1 isoform 1 F7CHP4 20 Ailuropoda melanoleuca CRB1 isoform 1 D2HBN8 21 Ailuropoda melanoleuca CRB1 isoform 2 G1MGT9 22 Bos taurus CRB1 isoform 1 F1N3A5 23 Cavia porcellus CRB1 isoform 1 H0VFY3 24 Callithrix jacchus CRB1 isoform 1 F7IHH9 25 Callithrix jacchus CRB1 isoform 2 F6ULV9 26 Callithrix jacchus CRB1 isoform 3 F7IHI1 27 Gallus gallus CRB1 isoform 1 E1BT21 28 Taeniopygia guttata CRB1 isoform 1 H0Z1J9 29 Meleagris gallopavo CRB1 isoform 1 G1N252 30 Danio rerio CRB1 Q1A5L3 31⁶ Takifugu rubripes CRB1 H2RRM4 32 Pelodiscus sinensis CRB1 K7G2S0 33 Monodelphis domestica CRB1 F6UYP3 34 Canis familiaris CRB1 F1Q0H7 35 Oryctolagus cuniculus CRB1 G1TAK8 36 Caenorhabditis elegans CRB1 Q19350 37⁷ Xenopus tropicalis CRB1 isoform 1 F6URK2 38⁸ Xenopus tropicalis secretory protein CRB1 isoform 2 Q9IBG4 39 Anolis carolinensis CRB1 not yet known 40 Homo sapiens CRB2 isoform 1 Q5IJ48-1 41 Homo sapiens CRB2 isoform 2 Q5IJ48-2 42 Homo sapiens CRB2 isoform 3 Q5IJ48-3 43 Pan troglodytes CRB2 isoform 1 H2R133 44 Macaca mulatto CRB2 isoform 1 H9F357 45 Macaca mulatto CRB2 isoform 2 F7H6F4 46 Sus scrofa CRB2 isoform 1 F1SKU3 47 Spermophilus tridecemlineatus CRB2 isoform 1 I3NFS4 48 Otolemur garnettii CRB2 isoform 1 H0XLX5 49 Bos taurus CRB2 isoform 1 F1N2V0 50 Bos taurus CRB2 isoform 2 G3MYX7 51 Loxodonta africana CRB2 isoform 1 G3SL69 52 Loxodonta africana CRB2 isoform 2 G3U9Y6 53 Callithrix jacchus CRB2 isoform 1 F7H2F5 54 Callithrix jacchus CRB2 isoform 2 F7H7L7 55 Cavia porcellus CRB2 isoform 1 H0VG27 56 Gorilla gorilla CRB2 isoform 1 G3SAT2 57 Gorilla gorilla CRB2 isoform 2 G3RKD4 58 Mus musculus CRB2 isoform 1 Q80YA8 59 Rattus norvegicus CRB2 isoform 1 D4A3W2 60 Ailuropoda melanoleuca CRB2 isoform 1 G1LX36 61 Myotis lucifugus CRB2 isoform 1 G1Q9U3 62 Oryctolagus cuniculus CRB2 isoform 1 G1SWT6 63 Monodelphis domesti CRB2 isoform 1 F6UYP3 64 Sarcophilus harrisii CRB1 isoform 1 G3W2U0 65 Sarcophilus harrisii CRB2 isoform 1 G3WDZ2 66 Sarcophilus harrisii CRB2 isoform 2 G3WDZ3 67 Otolemur garnettii CRB2 isoform 1 H0XLX5 68 Danio rerio CRB2a Q1A5L2 69 Danio rerio CRB2b I3VKD7 70 Latimeria chalumnae CRB H3BHZ4 71⁹ Takifugu rubripes CRB2 isoform 1 H2UG 11 72 Gallus gallus CRB2 isoform 1 F1P3N1 73 Gallus gallus CRB2 isoform 2 E1BYW1 74 Taeniopygia guttata CRB2 isoform 1 H0Z9G5 75 Pelodiscus sinensis CRB2 isoform 1 K7FFW6 76¹⁰ Xenopus tropicalis CRB2 isoform 1 F6QPR5 77 Drosophila melanogaster CRB P10040 78 Daphnia pulex CRB E9GB00 79 Acyrthosiphon pisum CRB J9JPN0 80 Acromyrmex echinatior CRB F4WXJ1 81 Branchiostoma floridae CRB2 isoform 1 C3Y2J2 82 Strongylocentrotus purpuratus CRB2 isoform 1 H3J1J1 83 Anolis carolinensis CRB2 not yet known 84 Homo sapiens CRB3 isoform A Q9BUF7-1 85 Mus musculus CRB3 isoform A Q8QZT4-1 86 Callithrix jacchus CRB3 isoform A F7I870 87 Macaca mulatto CRB3 isoform A F6WKB4 88 Ailuropoda melanoleuca CRB3 isoform A G1LKJ7 89 Sus scrofa CRB3 isoform A F1SBS8 90 Sus scrofa CRB2 isoform 3 F1SKU4 91 Otolemur garnettii CRB3 isoform A H0WXN3 92 Myotis lucifugus CRB3 isoform A G1PSN7 93 Loxodonta africana CRB3 isoform A G3TBW3-1 94 Spermophilus tridecemlineatus CRB3 isoform A I3M0B8 95 Homo sapiens CRB3 isoform B Q9BUF7-2 96 Equus caballus CRB3 isoform A isoform 1 F6RUX8 97 Equus caballus CRB3 isoform A isoform 2 F6Z0G7 98 Rattus norvegicus CRB3 isoform A Q4V8I0 99 Cavia porcellus CRB3 isoform A H0VZN0 100 Sarcophilus harrisii CRB3 isoform A G3WE86 101 Cricetulus griseus CRB3 isoform A G3HCL1 102 Danio rerio CRB3 isoform A isoform 1 Q1A5L0 103 Danio rerio CRB3 isoform A isoform 2 Q1A5K9 104 Gasterosteus aculeatus CRB3 isoform A G3P8A8 105¹¹ Xenopus tropicalis CRB3 isoform A Q5EGD4 106 Tetraodon nigroviridis CRB3 isoform A Q4SE18 107 Strongylocentrotus purpuratus CRB3 isoform A H3I971 108 Mus musculus CRB3 isoform B Q8QZT4-2 109 Callithrix jacchus CRB3 isoform B F6Z7P9 110 Macaca fascicularis CRB3 isoform B G8F358 111 Macaca mulatta CRB3 isoform B F6WKC2 112 Ailuropoda melanoleuca CRB3 isoform B D2HY77 113 Pan troglodytes CRB3 isoform B H2QF51 114¹² Nomascus leucogenys CRB3 isoform B G1S1D6 115 Gorilla gorilla CRB3 isoform B G3QVX7 116¹³ Pongo abelii CRB3 isoform B H2NX84 117 Heterocephalus glaber CRB3 isoform B G5C4F3 118 Brugia malayi CRB3 A8PUD4 119 Caenorhabditis elegans CRB3 A0FLQ5 120 Caenorhabditis remanei CRB3 E3MAT4 ¹ Also known as: short CRB1, sCRB1, CRB1dE3/4, CRB1Δ3/4, CRB1ΔE3/4 (Kantardzhieva et al., 2005. Invest Ophthalmol Vis Sci. 46(6):2192-201)
² Also known as: Crb1s
³ Pongo pygmaeus abelii
⁴ Hylobates leucogenys
⁵ Also known as: short CRB1 or sCRB1 or CRB1dE3/4 or CRB1Δ3/4 or CRB1ΔE3/4
⁶ Fugu rubripes
⁷ Xenopus sp.
⁸ Xenopus sp.
⁹ Fugu rubripes
¹⁰ Xenopus sp.
¹¹ Xenopus sp.
¹² Hylobatus leucogenys
¹³ Pongo pygmaeus abelii SEQ ID NO: Description 121 CMV promoter flanked at the 5' end with a BglII restriction site (AGATCT) 122 Human RLBP1 promoter 123 Human GRK1 specific promoter 124 Human RHO specific promoter 125 miR-31-5p target sites 126 miR-126-3p target sites 127 miR-211 / miR-204 target sites 128 illustrative synthetic intron (ln5) in coding sequence of the CRB gene 129 illustrative synthetic polyadenylation region flanked at the 3' end with a BglII restriction site (AGATCT) 130 Illustrative 5' untranslated region 131 AAV2 inverted terminal repeat flanked at the 3' end with a BglII restriction site (AGATCT) 132 AAV2 inverted terminal repeat flanked at the 5' end with a BglII restriction site (AGATCT) 133 Truncated CMV promoter or minimal CMV promoter

• Figure 1. A representative presentation of the localization of CRB 1 in the human retina.
• Figure 2. A representative presentation of the localization of CRB 1 in the mouse retina.
• Figure 3. A representation of the mammalian Crumbs homologue protein family.
• Figure 4. Degeneration in the ventral but not in the dorsal retina of Crb1^-/-Crb2^F/+ Chx10Cre/+ mice. A-E, Technovit sections. Left panels, dorsal (superior) retina. Right panels, ventral (inferior) retina.
• Figure 5. Electroretinogram b-waves of Crb1^-/-Crb2^F/+Chx10Cre/+ mice (on 50% C57BL/6J and 50% 129/Ola genetic background) showing loss of retinal activity at 3 months of age. Note that wild-type, Crb1^-/- and Crbl^-l-Crb2^F/F mice (not containing Chx10Cre) do not show differences in retinal activity (data not shown). Panel a, scotopic ERG showing loss of rod photoreceptor activity in Crb1^-/-Crb2^F/+Chx10Cre retinas (light grey; lower line) vs. Crb1^-/-Crb2^F/F retinas (dark grey; upper line). Panel b, photopic ERG showing loss of cone photoreceptor activity in Crbl^-/-Crb2^F/+Chx10Cre retinas (light grey; lower line)) vs. Crbl^-l-Crb2^F/F retinas (dark grey; upper line).
• Figure 6. Loss of a separate photoreceptor layer in Crb1^-l-Crb2^F/FChx10Cre /+ retinas. A-E, Technovit sections. Left panels, control retina. Middle panels, Crb1^+l- Crb2^F/FChx10Cre/+ retinas. Right panels, Crb1^-/-Crb2^F/FChx10Cre/+ retinas showing absence of a separate photoreceptor layer and mislocalized retinal cells.
• Figure 7. Loss of retinal activity in Crb1Crb2 cKO compared to Crb1 and Crb2 cKO retinas. Panels a and c, measured ERGs at 3 months of age. Panels b and d, measured ERGs at 1 months of age. Panels a and b, scotopic. Panels c and d, photopic. Note, at 3 months of age, the very good separation of confidence intervals in b-wave amplitude between Crb1 KO and Crb2 cKO retinas. The lines in the figures represent the following: the upper most line concerns Crb1 KO, second from the top is Crb2 KO, third from the top is Crb1^+/-Crb2^F/FChx10Cre (heterozygote Crb1^+l- homozygote floxed Crb2^F/F heterozygote ChxlOCre), the bottom line is Crb1^-/-Crb2^F/FChx10Cre (homozygote Crb1^{- /-} homozygote floxed Crb2^F/F heterozygote Chx10Cre).
• Figure 8. Upon subretinal injection, AAV9-CMV-GFP and ShH10Y-CMV-GFP infect Müller glia cells and photoreceptors. Abbreviations: GCL, ganglion cells; PRC, photoreceptor cells; RPE, retinal pigment epithelium cells.
• Figure 9. Expression of short CRB 1 (SEQ ID NO: 3) in Crb1 KO retinas using subretinal injection of AAV2/9-CMV-sCRB1 vectors. Panel a, control Crb1 KO retina. Panel b, Crb1 KO retina expressing sCRB1 upon transduction with AAV2/9-CMV-sCRB 1 viral particles. Abbreviations: OLM, outer limiting membrane; OPL, outer plexiform layer. Note: Expression of sCRB1 caused retinal degeneration in about half of the transduced retina (degeneration data not shown).
• Figure 10. Representative experiment showing rescue of loss of retinal activity. Crb2 cKO retinas were injected at postnatal day 23 subretinally with 10¹⁰ AAV2/ShH10Y-CMV-CRB2 or AAV2/ShH10Y-CMV-GFP viral particles and analyzed for ERG and immunohistochemistry at 3 months of age. Panel a, electroretinogram scotopic b-wave showing rescue of retinal activity in the right Crb2 cKO eye transduced with AAV2/ShH10Y-CMV-CRB2 (dark line), compared to the left eye of the same Crb2 cKO transduced with AAV2/ShH10Y-CMV-GFP (faint line). The scotopic a-wave is also rescued (data not shown). Panel b, immunohistochemistry showing expression of sCRB 1 in the right eye of the animal used in panel a. No expression of CRB2 was detected in the left eye of the same animal. Abbreviations: OLM, outer limiting membrane; RPE retinal pigment epithelium.
• Figure 11. Specific transduction of Müller glia cells using 10¹⁰ AAV2/6-RLBP1-GFP viral particles containing the human RLBP1 promoter (SEQ ID NO: 122) upon intravitreal injection (specific infection of Müller glia cells). Retinas were collected 3 weeks post-infection. Panel a, scanning-laser-ophtalmoscopy (SLO). Panel b, SLO showing fluorescent cells. Panel c, immunohistochemistry showing specific expression of GFP in Müller glia cells.
• Figure 12. AAV6 and ShH10Y capsids transduce adult human Müller glia cells. 1 µL 10¹³ genome copies per mL of AAV2/6-CMV-GFP-WPRE-pA (panel a) or AAV6 variant AAV2/ShH10-CMV-GFP-WPRE-pA (panel b) was applied to pieces of cultured adult human retina. GFP expression was detected in Müller glia cells.
• Figure 13. GFP and CRB1 protein expression in cell lines. Western Blotting of HEK293T cell lysates transfected with the calcium phosphate method and 10 µg of pAAV-CMV-GFP-WPRE-pA or pAA V-CMV-hCRB1-pA vectors showed subsequent CRB 1 and GFP protein levels. However, whereas RPE-derived ARPE-19 cells expressed normal amount of GFP, CRB 1 protein is just above detection level in three times overloaded protein lysates.
• Figure 14. Rescue of loss of retinal function by subretinal injection of AAV2/9-CRB2 viral particles in Crb mutant mouse eyes; failure of rescue of retinal function by AAV2/9-CRB1 viral particles. Crb1^-/-Crb2^F/+Chx10Cre^Tg/+ (Crb1Crb2^F/+ cKO; a-f) and Crb2^F/FChx10Cre^Tg/+ (Crb2 cKO; g-h) mouse retinas injected subretinally at 2 weeks of age with 1 µL of 2 × 10¹⁰ genome copies of 4.9 kb AAV2/9-CMV-CRB2-In5-spA (briefly AAV2/9-CRB2), i.e., CRB2 flanked by AAV2 ITRs and packaged in AAV9 capsid proteins and in the contralateral control eye with AAV2/9-CMV-GFP (a-c, g-h), or with 1 µL of 1 × 10¹⁰ genome copies of 4.8 kb AAV2-minimalCMV-CRB 1-spA (minimal CMV presented as SEQ ID NO: 133 in the sequence listing) containing AAV9 viral particles and in the contralateral control eye with AAV2/9-minCMV-GFP (d-f), and analyzed at 3 or 4 months of age by electroretinography under scotopic (dark-adapted; a-b, d-e, g-h) or photopic (light-adapted; c, f) conditions. Scotopic b-wave amplitudes (a, d, g) and a-wave amplitudes (b, e, h), and photopic b-wave amplitudes (c, f) are indicated. CRB2 vectors rescued loss of retinal function in two different Crb mutant mouse models (a-c, g-h), whereas CRB 1 vectors did not rescue loss of retinal function (d-f).
• Figure 15. Toxicity of CRB proteins tested by intravitreal injection of AAV2/ShH10Y-minimalCMV-CRB 1-In5-spA or AAV2/ShH10Y-CMV-CRB2-In5-spA viral particles in Crb mutant mouse eyes, i.e., CRB2 or CRB1 DNA operably linked to the promoter that is indicated, flanked by AAV2 ITRs and packaged in ShH10Y capsid proteins. Crb1^-/- Crb2^F/+Chx10Cre^Tg/+ (Crb1Crb2^F/+ cKO; a-e) mouse retinas injected intravitreally at 2 weeks of age with 1 µL of 10¹⁰ genome copies of 4.9 kb AAV2/ShH10Y-CMV-CRB2-In5-spA and in the contralateral control eye with AAV2/ShH10Y-CMV-GFP (a-b), or with 1 µL of 5 × 10⁹ genome copies 4.8 kb AAV2/ShH10Y-minimalCMV-CRB1-In5-spA containing ShH10Y viral particles and in the contralateral control eye with AAV2/ShH10Y-minimalCMV-GFP (c-e). The eyes were analyzed at 3 months of age by electroretinography under scotopic (dark-adapted; a-b, c-d) or photopic (light-adapted; e) conditions. Scotopic b-wave amplitudes (a, c) and a-wave amplitudes (b, d), and photopic b-wave amplitudes (e) are indicated. No statistically significant differences in retinal function were detected for intravitreally applied CRB2 vectors compared to GFP control vectors (a-b). Intravitreally applied CRB1 vectors showed strongly reduced retinal responses upon expression of CRB1 vectors, suggesting toxic effects by CRB1 vectors.

Retinas of Crbl^-/+Crb2^F/FChx10Cre/+ mice (heterozygote for Crb1, homozygote for floxed Crb2) show to some extent a similar but more severe phenotype than observed in Crb2^F/FChx10Cre/+ retinas. Electroretinography showed a significant loss of retinal activity at 1 month of age that progressed quickly. The phenotype starts already at E15.5 (at this time point similar to E17.5 in Crb2^F/FChx10Cre/+ retinas), with disruptions at the outer limiting membrane, and rosettes of retinal cells can be detected. A major difference of these mouse retinas compared to the Crb2^F/FChx10Cre/+ retinas is the aberrant localization of several retinal cell types. E.g., some amacrine cells ectopically localize in the photoreceptor layer, and some cone and rod photoreceptors ectopically localize at the ganglion cell layer. Nevertheless, in these retinas there is still three nuclear layers (outer and inner, and ganglion) and two plexiform layers (outer and inner) suggesting that the lamination of the retina is grossly normal.

Retinas of Crb1^-/-Crb2^F/FChx10Cre/+ mice (homozygote for Crb1, homozygote for floxed Crb2; also indicated as Crb1Crb2 cKO) show the most severe phenotype (Fig. 6). Electroretinography showed a severe loss of vision at 1 month of age (though there is still some retinal activity; Fig. 7). These retinas do not show a separate photoreceptor layer (no outer and inner segment, nuclear or outer plexiform layer) and no outer plexiform layer but a single broad nuclear layer, an inner plexiform layer, and a ganglion cell layer. The nuclear layer contains nuclei of rod and cone photoreceptors, bipolar, horizontal, amacrine and Müller glia cells, but surprisingly also nuclei of ganglion cells. The inner plexiform layer only occasionally contains cell nuclei. The ganglion cell layer that normally contains nuclei of ganglion and displaced amacrine cells contains in addition nuclei of rod photoreceptors, bipolar, horizontal, Müller glia cells. So, whereas there is a laminated retina, several early as well as late born cells localized ectopically. Furthermore, there was a significant increase in dividing retinal progenitor cells at E15.5, E17.5, P1 and P5. Concomitant, there is an increase in late born cell types such as rod photoreceptors, bipolar, Müller glia, and late-born amacrine cells, but not in early born cell types such as ganglion, cone photoreceptors, horizontal, and early born amacrine cells. Increased apoptosis was detected at E13.5, E17.5, P1, P5, P14 and at 3 months of age. These data suggest that CRB proteins (CRB2 and CRB1) play a role in suppressing proliferation of late born retinal progenitor cells or timely exiting the cell cycle, in addition to maintaining the adherens junctions between retinal progenitor cells, rod and cone photoreceptors, bipolar and Müller glia cells.

In Crbl-1-Crb2F1+ChxlOCre/+ mice the morphological phenotype starts at P10 (Fig. 4), and a significantly decreased ERG is detected at 3 months of age (data not shown) and is very clear at 6 months of age (whereas no decrease is detected in Crb1^-/- retinas). In these retinas, the dorsal (superior) part of the retina does not show retinal degeneration, whereas the ventral (inferior temporal and nasal) part does. This is in part reminiscent to Crb1^-/- retinas in which only one quadrant (inferior temporal) and Crb1^rd8/rd8 retinas in which only one quadrant (inferior nasal) part of the retina shows (limited) retinal degeneration. These mice are useful for functionally testing our AAV CRB gene therapy vectors by electroretinography (ERG) since control double heterozygote Crb1^+/- Crb2^F/+Chx10Cre/+ retinas do not show a morphological or ERG phenotype. Unfortunately, the confidence intervals for control and mutant mice (on 50% C57BL/6J and 50% 129/Ola mixed genetic background) at 3 and 6 months of age are very close to each other, rendering the model difficult for interpreting (partial) rescue studies. Crb1^-/- Crb2^F/+Chx10Cre/+ mice on 99.9% C57BL/6J background are being produced and will provide less inter-mouse variation. As described below, we consider Crb2^F/FChx10Cre/+ retinas, which mimic loss of CRB1 in retinitis pigmentosa patients, as the best for rescue studies since their electroretinograms are easier to interpret.

It is important to note that there are alternative transcripts of the human CRB1 gene. One transcript, lacking exons 3 and 4 but maintaining the open reading frame, encodes a shorter form than full length CRB1 but is present in human (SEQ ID NO: 3), and e.g., apes, monkeys, canine, equine, feline and many other species, however not in mice. Notably, the sequence of CRB2 (SEQ ID NO: 40) is very similar to the sequence of this naturally occurring short variant of CRB1 (sCRB1 or sCRB1ΔE3/4). In our initial trials we generated AAV vectors with the CMV promoter, the short CRB1, a synthetic intron (In5) in the short CRB 1 cDNA sequence, and a synthetic spA. Upon subretinal injection of this vector packaged in AAV serotype 9 (AAV9), we detected significant expression of short CRB 1 at the "outer limiting membrane" in Müller glia cells and photoreceptors. Similarly, results were obtained with vector packaged in AAV serotype 5 (AAV5). We subretinally injected 1 µL of AAV2/9-CMV-hCRB1ΔE3/4In5-spA (1.00 × 10¹⁰ delivered vector genomes) plus a ten-fold lower dose of AAV2/9-CMV-GFP-WPRE-pA (1.00 × 10⁹ delivered vector genomes) into the left eye of retinas lacking CRB1 with reduced levels of CRB2 (Crb1^-/-Crb2^flox/+Chx10Cre retinas). The contralateral control eye received 1 µL of AAV2/9-CMV-GFP-WPRE-pA (1.00 × 10¹⁰ delivered vector genomes). Crb1^-/-Crb2^flox/+Chx10Cre retinas show progressive loss of retinal function from 1 to 3 to 6 months of age (data not shown). The treated eyes showed expression of short CRB 1 in a large region of the retina at the "outer liming membrane" of Müller glia cells and photoreceptors, and in retinal pigment epithelium. However, using two independently generated batches of the viral particles, we detected loss of the photoreceptor layer as well as retinal pigment epithelium layer due to expression of the short variant of CRB 1 in Müller glia cells or photoreceptors or retinal pigment epithelium both by histochemistry and by immunohistochemistry. The cause of these toxic effects is to be further analysed and may for example be e.g., an immune-response in the CRB1 naive Crb1 knockout retina, ectopic expression effect, incompatibility of mouse and human CRB 1 protein, differences between short and full length CRB1, interference of short CRB 1 with the expression of other CRB 1 transcripts or proteins, dose-dependent toxicity, untimely expression of short CRB1), and it might be related to the inability in producing continuous high level expression of short (or full length) CRB 1 in cultured cell lines. Preliminary studies expressing the short CRB 1 in wild-type C57BL/6J retina showed toxicity as well, suggesting that the toxicity is not only due to the expression of short human CRB1 in immune-naive Crb1 knockout retina. This urged us to test expression of CRB2 in a therapeutic vector, since CRB2 expression was well tolerated in cell lines. Expression of CRB2 in Müller glia cells or photoreceptor cells or retinal pigment epithelium did not result in toxic effects. More specifically, expression of CRB2 in Muller glia cells or photoreceptor cells or retinal pigment epithelium did not result in a detectable loss of the photoreceptor layer and/or the retinal pigment epithelium layer. This lack of toxic effects of CRB2 expression in Müller glia cells and photoreceptor cells and retinal pigment epithelium is relevant to the development of future clinical applications. Note that we used very high levels of AAV-CRB2 vector (10¹⁰ delivered vector genomes) but toxic effects were not detected.

In this example, the Inventors evaluated whether delivery of a species-specific version of Crumbs homologue (CRB) (i.e., human) to Müller glia cells and photoreceptors of the postnatal Crb2 cKO mouse could restore function to these cells. Serotype 6 (variant ShH10Y) AAV vectors were used to deliver human CRB2 subretinally to Müller glia cells and photoreceptors of postnatal day 23 (P23) Crb2 cKO mice. Electroretinogram (ERG) and behavioral testing were used to assess visual function and immunocytochemistry was used to examine therapeutic transgene expression, Crumbs homologue (CRB) complex protein localization and preservation of retinal structure in treated and untreated eyes.

This example demonstrates that an AAV vector subretinally delivered to the left eyes of P23 Crb2 cKO mice facilitated expression of wild-type CRB2, restoration of visual function and behavior, and preservation of rod and cone photoreceptors. Ten weeks following injection, retinal function (ERG) was analyzed in treated and untreated eyes. In some experiments, ERG was performed every two weeks after 4 weeks until 10 weeks post injection (the latest time point evaluated). At 10 weeks post injection, all animals were sacrificed and their treated and untreated retinas were evaluated for expression of CRB2 and localization of Crumbs homologue (CRB) complex proteins.

The results confirm that rod-mediated and cone-mediated function was restored to treated eyes of Crb2 cKO mice (ERG a-wave and b-wave amplitudes were about twice better than in the untreated eyes). Moreover, the treatment effect was stable for at least 10 weeks post-administration. Histology revealed AAV-mediated CRB2 expression in Müller glia cells and photoreceptors and a restoration of Crumbs homologue (CRB) complex protein location in treated mice. In addition, cone cell densities were higher in treated eyes than untreated contralateral controls. This result suggests that treatment is capable of preserving cone and rod photoreceptors for at least 10 weeks post treatment. This is the first demonstration that postnatal gene therapy is capable of restoring visual function and behavior to, and preserving retinal structure in, a mammalian model of RP due to mutations in the Crumbs homologue gene. Importantly, results were obtained using a well characterized, clinically relevant AAV vector; the in vivo animal model data thus obtained provide the foundation for an AAV-based gene therapy vector for treatment of children affected with LCA8 and/or RP due to mutations in the CRB 1 gene.

Crb2^flox/flox mice were generated at the Inventor's facilities. Chx10Cre heterozygote embryos were obtained from a living stock at The Jackson Laboratory (Bar Harbor, ME, USA). Heterozygotes were mated at the Inventors' facilities to produce Crb2^flox/floxChx10Cre homozygous mice and isogenic Crb2^flox/+Chx10Cre control offspring (both heterozygous for Chx10Cre). All mice were bred and maintained in a centralized facility at the Inventors' institution under a 12 hr/12 hr light/dark cycle. Food and water were available ad libitum. All animal studies were approved by the local Institutional Animal Care and Use Committee and conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and KNAW (Koninklijke Nederlandse Akademie van Wetenschappen) regulations.

AAV vectors with serotype 6 variant ShH10Y capsid proteins and AAV2 ITR and Rep proteins (AAV2/ShH10Y) were used to deliver human CRB2 (hCRB2) as they have been shown to exhibit robust transduction efficiency and a faster onset of expression in retinal Müller glia cells as well as photoreceptors than other AAV serotypes. The serotype 6 variant ShH10Y AAV capsid was provided by Dr. John Flannery (University of California, Berkeley, CA, USA). AAV serotype 5 was obtained from Plasmid Factory. AAV serotype 9 was obtained from Dr. Joost Verhaagen (Netherlands Institute for Neuroscience). A ubiquitous cytomegalovirus (CMV) promoter was selected to drive expression of hCRB2. The nucleic acid sequence of an illustrative ubiquitous CMV promoter which was used in the studies is shown in SEQ ID NO: 121. The CMV promoter is flanked at the 5' sequence with a BglII restriction site (AGATCT). A synthetic intron (In5) inserted in the CRB2 cDNA was used for stable transcript processing of CRB2. The nucleic acid sequence of an illustrative synthetic intron (In5) in the coding sequence of the Crumbs homologue (CRB) gene is shown in SEQ ID NO: 128. The intron was inserted into CRB2 cDNA between two adjacent exons with a sequence of exon NNNAG/intron/GNNN exon, where G, A, T, C stands for one of the four nucleotides, and N stands for any of the four nucleotides. A synthetic poly-adenylation (spA) sequence was used for efficient termination of transcription. The nucleic acid sequence of an illustrative synthetic polyadenylation region (Levitt et al., 1989. Genes Dev. 3(7):1019-25) in between the stop codon behind the translated region of the Crumbs homologue (CRB) gene and the 3'flanking inverted terminal repeat which was used is shown in SEQ ID NO: 129. The synthetic polyadenylation site is flanked at the 3'sequence with a BglII restriction site (AGATCT). The nucleic acid sequence of an illustrative 5' untranslated region located in between the CMV promoter and the translated region of the Crumbs homologue (CRB) gene which was used is shown in SEQ ID NO: 130.

The CMV-hCRB2In5-spA fragment, containing BglII restriction sites at the 5' and 3' ends, with sequence identified in SEQ ID NO: 40 was synthesized by GenScript (Piscataway, NJ, USA). The BglII CMV-hCRB2In5-spA fragment was cloned into pUC57 (Thermo Fisher Scientific, Waltham, MA, USA) containing two inverted terminal repeats (ITRs) of AAV2 flanked by BglII restriction sites (SEQ ID NO: 131 and 132). The resulting AAV-hCRB2 plasmid of 4.9 kb contained the sequence identified in SEQ ID NO: 40 and was sequence verified.

AAV vectors were packaged and purified by iodixanol gradient ultra-centrifugation according to previously published methods (Zolotukhin et al., 1999. Gene Ther. 6(6):973-85; Hermens et al., 1999. Hum Gene Ther. 10(11):1885-91; Ehlert et al., 2010. BMC Neurosci. 11:20). Viral particles were diluted, washed and concentrated using an Amicon 100 kDa MWCO Ultra-15 device (Millipore, Billerica, MA, USA) in Dulbecco's Balanced Salt Solution (Life Technologies, Bleiswijk, Netherlands) and titered by quantitative real-time PCR (Aartsen et al., 2010. PLoS One. 5:e12387). Resulting titers were 1.00 × 10¹³ viral genomes per mL (vg/mL) for AAV2/ShH10Y-CMV-hCRB2 or AAV2/9-CMV-hCRB2 (AAV2 ITR and Rep proteins; AAV9 capsid proteins) or AAV2/5-CMV-hCRB2 (AAV2 ITR and Rep proteins; AAV5 capsid proteins).

In a typical experiment, 1 µL of AAV2/ShH10Y-CMV-hCRB2 (1.00 × 10¹⁰ delivered vector genomes) plus a ten-fold lower dose of AAV2/ShH10Y-CMV-GFP-WPRE-pA (1.00 × 10⁹ delivered vector genomes) was delivered subretinally at postnatal day 23 (P23) to the left eye of each Crb2^flox/floxChx10Cre mouse. The contralateral control right eye was injected with 1 µL of AAV2/ShH10Y-CMV-GFP-WPRE-pA (1.00 × 10¹⁰ delivered vector genomes). Subretinal injections were performed as previously described (Aartsen et al., 2010. PLoS One. 5:e12387). Further analysis was carried out on all animals, not only the ones which received comparable, successful injections (> 60 % retinal detachment and minimal complications). It is well established that the area of retinal detachment corresponds to the area of viral transduction (Cideciyan et al., 2008. Proc Natl Acad Sci U S A. 105(39):15112-7; Timmers et al., 2001. Mol Vis. 7:131-7).

In a representative experiment, electroretinograms (ERGs) of treated Crb2 cKO (n = 3) and isogenic controls (n = 2) were recorded using a PC-based control and recording unit (Toennies Multiliner Vision; Jaeger/Toennies, Hochberg, Germany) according to methods previously described with minor modifications (Haire et al., 2006. Invest Ophthalmol Vis Sci. 47(9):3745-53). Initial ERG measurements were recorded at 4 weeks' postinjection, and each subsequent 2 weeks thereafter, until 10 weeks postinjection (the latest time point evaluated in the study). Age matched isogenic controls were recorded alongside treated animals at every time point. Mice were dark-adapted overnight (more than 12 hours) and anesthetized with a mixture of 100 mg/kg ketamine, 20 mg/kg xylazine and saline in a 1:1:5 ratio, respectively. Pupils were dilated with 1 % tropicamide and 2.5 % phenylephrine hydrochloride. A heated circulating water bath was used to maintain the body temperature at 38°C. Hydroxypropyl methylcellulose 2.5 % was applied to each eye to prevent corneal dehydration. Full-field ERGs were recorded using custom, gold wire loop corneal electrodes. Reference and ground electrodes were placed subcutaneously between the eyes and in the tail, respectively. Scotopic rod recordings were elicited with a series of white flashes of seven increasing intensities (0.1 mcds/m² to 1.5 cds/m²). Interstimulus intervals for low intensity stimuli were 1.1 second. At the three highest intensities (100 mcds/m², 1 cds/m² and 5 cds/m²), interstimulus intervals were 2.5, 5.0 and 20.0 seconds, respectively. Ten responses were recorded and averaged at each intensity. Mice were then light adapted to a 100 cds/m² white background for 2 min. Photopic cone responses were elicited with a series of five increasing light intensities (100 mcds/m² to 12 cds/m²). Fifty responses were recorded and averaged at each intensity. All stimuli were presented in the presence of the 100 cds/m² background. B-wave amplitudes were defined as the difference between the a-wave troughs to the positive peaks of each waveform.

Alternatively, ERGs recordings were elicited with a series of light pulses of increasing intensities (2.7 cds/m² to 25 cds/m², logarithmically spread over 10 levels. Pulse lengths ranged from 0.5 to 5 msec. Between pulses there was a delay of approximately 2 seconds (0.5 Hz). Thirty responses were recorded and averaged at each intensity. No extra delay was introduced for the transition from one intensity level to the next. Between pulses, no background lighting was present. The a-wave trough was defined as the minimum response between 0 and 30 milliseconds after stimulus onset. The b-wave peak was defined as the maximum response between 15 and 100 milliseconds after stimulus onset. The a-wave amplitude was defined as the difference between the baseline and the a-wave trough, whereas the b-wave amplitude was defined as the difference between the b-wave peak and the a-wave trough.

Photopic b-wave maximum amplitudes (those generated at 12 cds/m²) of all CMV-hCRB2- treated (n = 3) Crb2 cKO (both treated and untreated eyes) and isogenic control mice were averaged and used to generate standard errors. These calculations were made at every time point (4 weeks' to 10 weeks' post-injection). This data was imported into Sigma Plot for final graphical presentation. The paired t-test was used to calculate P-values between treated and untreated eyes within each group over time (4 weeks postinjection vs. 10 weeks post-injection). Significant difference was defined as a P-value < 0.05.

Ten weeks post-injection, P23-treated Crb2 cKO mice and age matched isogenic controls were dark adapted for 2 hr. Immediately following dark adaptation, mice were sacrificed under dim red light (> 650 nm). The limbus of injected and un-injected eyes was marked with a hot needle at the 12:00 position, facilitating orientation. Enucleation was performed under dim red light and eyes were placed immediately in 4 % paraformaldehyde. Eyes that were to be used for cryo sectioning were prepared according to previously described methods (Haire et al., 2006. Invest Ophthalmol Vis Sci. 47(9):3745-53). Briefly, corneas were removed from each eye, leaving the lens inside the remaining eye cup. A small "V" shaped cut was made into the sclera adjacent to the burned limbus to maintain orientation. After overnight fixation, the lens and vitreous were removed. The remaining retinal RPE-containing eyecup was placed in 30 % sucrose in PBS for at least 1 hr at 4°C. Eyecups were then placed in cryostat compound (Tissue Tek OCT 4583; Sakura Finetek, Inc., Torrance, CA, USA) and snap-frozen in a bath of dry ice/ethanol. Eyes were serially sectioned at 10 µm with a cryostat (Microtome HM550; Walldorf, Germany). Eyes that were to be used for whole mount analysis were prepared according to previously described methods (van de Pavert et al., 2007. J Neurosci. 27(3):564-73). Orientation was achieved as previously mentioned. After overnight fixation, cornea, lens, vitreous and retinal pigment epithelia were removed from each eye without disturbing the retina. A cut was made in the superior (dorsal) portion of the retina adjacent to the original limbus bum to maintain orientation.

Retinal cryosections and whole mounts were washed 3 times in 1X PBS. Following these washes, samples were incubated in 0.5 % Triton X-100^® for 1 hr in the dark at room temperature. Next, samples were blocked in a solution of 1 % bovine serum albumin (BSA) in PBS for 1 hr at room temperature. Retinal sections were incubated overnight at 37°C with a rabbit polyclonal CRB2 antibody EP13 or SK11 (1:1000 and 1:200, respectively; provided by Dr. Penny Rashbass, University of Sheffield, UK) diluted in 0.3 % Triton X-100^®/1 % BSA. Following primary incubation, retinal sections and whole mounts were washed 3 times with 1X PBS.

Retinal sections were incubated for 1 hr at room temperature with IgG secondary antibodies tagged with Cyanine dye Cy5 (Molecular Probes, Eugene, OR, USA) diluted 1:500 in 1X PBS. Following incubation with secondary antibodies, sections and whole mounts were washed with 1X PBS. Retinal sections were counterstained with 4',6'-diamino-2-phenylindole (DAPI) for 5 min at room temperature. After a final rinse with 1X PBS and water, sections were mounted in an aqueous-based medium (DAKO) and cover-slipped. Retinal whole mounts were oriented on slides with the superior (dorsal) portion of the retina positioned at the 12:00 position. Samples were mounted in DAKO and cover-slipped.

Retinal sections were analyzed with confocal microscopy (Leica TCS SP5 AOBS Spectral Confocal Microscope equipped with LCS Version 2.61, Build 1537 software, (Bannockburn, IL, USA). All images were taken with identical exposure settings at either 20x or 63x magnification. Excitation wavelengths used for DAPI and CRB2 stains were 405 nm and 650 nm, respectively. Emission spectra were 440-470 nm and 670 nm, respectively. Retinal whole mounts were analyzed with a widefield fluorescent microscope (Axioplan 2) (Zeiss, Thornwood, NY, USA) equipped with a QImaging Retiga 4000R Camera and QImaging QCapture Pro software (QImaging, Inc., Surrey, BC, Canada). Quadrants of each whole mount were imaged at 5x under identical exposure settings and then merged together in Photoshop^® (Version 7.0) (Adobe, San Jose, CA, USA).

It was previously reported that rod and cone responses in the Crb2 cKO mouse are significantly decreased at 1 month of age and progressively decreased at 3 months of age (Alves et al., 2013. Hum Mol Genet. 22(1):35-50). Here, the inventors have shown that P23-treatment of this mouse with an AAV vector carrying the human CRB2 gene (SEQ ID NO:40) under the control of a ubiquitous (CMV) promoter led to substantial restoration of rod photoreceptor function as measured by electroretinography (ERG). Representative rod traces from CMV-hCRB2-treated and control CMV-GFP treated eyes showed that rod function in CMV-hCRB2 treated eyes was restored to approximately 40 % of normal at 10 weeks post-injection. Similar to previous reports, rod responses in contralateral, untreated eyes were about 20% of normal by this time point. Importantly, restoration of rod photoreceptor a-wave and b-wave function remained stable at 3 months (the latest time point evaluated in this study (see Fig. 10). Rod retinal function (ERG) is partially preserved in the Crb2 cKO mouse. Studies have shown that even very small ERG amplitudes translate into robust visual behavior (Williams et al., 2006. PLoS Med. 3(6):e201). In fact, LCA2 patients who received AAV-RPE65 therapy were found to exhibit behavioral restoration despite a complete lack of ERG response (Maguire et al., 2008. N Engl J Med. 358(21):2240-8). So, the rescue of loss of retinal function in Crb2 cKO retinas by the AAV-hCRB2 vector is very promising for future gene therapy studies. This is the first example of rescue of loss of retinal function in mammalians lacking Crumbs homologue (CRB) function using a candidate clinical gene therapy vector.

Analysis was carried out on all animals, not only the ones which received comparable, successful injections (> 60 % retinal detachment and minimal complications). It is well established that the area of retinal detachment corresponds to the area of viral transduction (Cideciyan et al., 2008. Proc Natl Acad Sci USA. 105(39):15112-7; Timmers et al., 2001. Mol Vis. 7:131-7). Mice with unsuccessful subretinal injections showed lack or limited expression of hCRB2 and GFP in combination with lack of rescue of scotopic b-wave or a-wave ERG function (see Fig. 10). Due to the inter-mouse variability in untreated Crb2 cKO rod responses (60-80 % of WT by 3 months of age), statistical comparison of average rod responses of treated vs. untreated eyes is problematic. However, within an animal, rod ERG amplitudes are nearly equal between partner eyes, therefore we calculated the average intra-mouse rod a- and b-wave amplitude ratios for treated versus untreated eyes and then plotted these ratios over time.

CRB 1-deficiency affects both Müller glia and photoreceptors in LCA8 and RP patients due to mutations in the CRB1 gene. The ubiquitous CMV promoter was therefore chosen for this study as a means of targeting both cell types. The AAV6 variant ShH10Y capsid was chosen because it infects upon subretinal injection efficiently in vivo mouse Müller glia and photoreceptors (and infects e.g., in vitro human retinal Müller glia cells, see Fig. 12). Immunostaining of Crb2 cKO retinas 10 weeks posttreatment with AAV-CMV-hCRB2 revealed that this promoter drove robust hCRB2 expression in inner segments of photoreceptors and apical villi of Müller glia cells. Typically, a retinal cross section from an eye injected with this therapeutic vector shows intense hCRB2 staining at the outer limiting membrane whereas the contralateral, mock GFP treated eye from the same mouse lacks any hCRB2 expression. Levels of CMV-mediated hCRB2 expression approached that seen in isogenic control eyes. hCRB2 expression in CMV-hCRB2-treated neural retina was restricted to the outer limiting membrane. hCRB2 expression was occasionally found in the retinal pigment epithelium. In normal mammalian retinas, the retinal pigment epithelium also expresses Crumbs homologue (CRB) complex members such as PALS 1 (Park et al., 2011. J Neurosci. 31(47):17230-41), albeit at lower levels than at the outer limiting membrane (Pellissier et al., 2014. Hum Mol Genet. 23(14):3759-71). Overexpression of hCRB2 in the wild-type RPE cells in the Crb2 cKO did not result in noticeable altered morphology or function of retinal pigment epithelium. Notably however, the CMV promoter construct did not drive therapeutic hCRB2 expression outside the photoreceptor cells, Müller glia cells and retinal pigment epithelium. This lack of off target expression is relevant to the development of future clinical applications. If required, overexpression in retinal pigment epithelium can be decreased by the use of micro-RNA target sites (miRT's) specific for miRNAs expressed in retinal pigment epithelium cells (Karali et al., 2011. PLoS One. 6(7):e22166).

It is important to note that while CRB 1-deficiency in humans causes LCA8 and progressive RP very well detectable by ERG, CRB 1-deficiency in mice causes late-onset retinal degeneration and degeneration limited to one quadrant of the retina and not detectable by ERG. Our immuno-electron microscopy data showed that in mice CRB 1 is restricted to the "outer limiting membrane" of Müller glia cells, whereas in humans CRB 1 is localized to the "outer limiting membrane" of Müller glia cells and photoreceptors. Our immuno-electron microscopy data showed that in mice CRB2 is localized to the "outer limiting membrane" of Müller glia cells and photoreceptors, whereas in humans CRB2 is restricted to the "outer limiting membrane" of Müller glia cells. Our analysis of mice lacking CRB1, mice lacking CRB2, mice lacking CRB 1 with reduced levels of CRB2, mice lacking CRB2 with reduced levels of CRB 1, and mice lacking both CRB 1 and CRB2 suggest very similar functions for CRB 1 and CRB2. Similarly, the functions of Crumbs homologue (CRB) proteins are exchangeable e.g., the human CRB 1 protein can rescue partially the phenotype in fruit flies lacking Crumbs (Crb) protein (den Hollander et al., 2001. Hum Mol Genet. 10(24):2767-73), and the zebrafish CRB2B protein can rescue the phenotype in zebrafish lacking CRB2A protein (Omori & Malicki, 2006. Curr Biol. 16(10):945-57).

Prior to Examples 1 and 2, several plasmids were transfected as naked plasmid DNA in cell lines (e.g., HEK293, MDCKII and ARPE19 cell lines) as described in section 3.1 MATERIALS AND METHODS. It was apparent that the transfected cell lines with short or full length CRB1 cDNA consistently resulted in low CRB1 expression. Also, cell lines (e.g., HEK293 and MDCKII cell lines) that stably express full length CRB1 cDNA (SEQ ID NO: 1) or short CRB1 cDNA (CRB1 lacking the entire extracellular domain; SEQ ID NO: 3) had a low expression. In contrast, cell lines expressing CRB2 cDNA resulted in high expression of CRB2 protein. These observations indicate that cells handle increased expression of CRB2 better than increased expression of CRB 1.

Experiments have been carried out in several mouse models.

Short human CRB 1 was overexpressed in retinas lacking CRB1 protein expression and with reduced levels of CRB2 protein. Thus, these mice still have functional native CRB2 protein in Müller glia cells and photoreceptor cells since CRB2 in mouse retina is present in both cell types. It is conceivable that this remaining mouse CRB2 protein is capable of taking over the function of the CRB1 protein. These mice on 50 % C57BL/6J and 50 % 129/Ola genetic background were less suitable to test rescuing of the phenotype in the retina. Control mice and mutant mice are significantly different in retina activity as measured using electroretinography. However, there is quite some variation in experimental animals and as a consequence the confidence intervals are close to one another. As far as rescuing the phenotype is concerned, the mouse model is still suboptimal and could be further optimized by backcrossing to 99.9 % C57BL/6J. Recently, trials were initiated in mice (on 75 % C57BL/6J and 25 % 129/Ola genetic background) lacking CRB 1 and having reduced levels of CRB2 using human CRB2 in AAV9 conform to the experimental setting as outlined above. As with the described AAV2/ShH10Y-CMV-CRB2 experiments, ERG rescue results were obtained using AAV2/9-CMV-CRB2 (1.00 × 10¹⁰ delivered vector genomes) viral particles subretinally injected into P14 Crb1^-/-Crb2^F/+Chx10Cre retinas (on 75 % C57BL/6J and 25 % 129/Ola genetic background) that were analyzed at 4 months of age.

Human CRB2 was overexpressed in retinas of mice lacking CRB2. These mice still have functional CRB1 protein in Müller glia cells, but lack functional CRB protein in photoreceptor cells. This situation most closely resembles the situation as seen in patients suffering from RP12 or LCA8. In these patients (lacking functional CRB1), CRB2 is present in Müller glia cells, but not in photoreceptor cells (see also Figs. 1 and 2). Retinas of Crb2 conditional knock-out mice show a big difference in retina activity at 1 and 3 months of age. The retinas of Crb2 mutant mice are rescued phenotypically, and the confidence intervals are separated and well interpretable.

Human CRB2 was overexpressed in retinas of mice lacking CRB 1 and with reduced levels of CRB2 (Crbl^-/-Crb2^F/+Chx10Cre). These mice lack CRB1 in the retina, but still have reduced levels of functional CRB2 protein in Müller glia cells and photoreceptor cells. This situation resembles mice lacking CRB 1 (on a genetic background with reduced levels of CRB2). Retinas of control Crb2^F/+ conditional knock-out mice on 75 % C57BL/6J and 25 % 129/Ola genetic background do not show loss of retina activity compared to wild-type mice. Retinas of Crb1 Crb2^F/+ conditional knock-out mice on 75 % C57BL/6J and 25 % 129/Ola genetic background show a big difference in retina activity at 3 months of age. The retinas of Crb1 Crb2^F/+ mutant mice are rescued phenotypically, and the confidence intervals are separated and well interpretable. These experiments show that CRB2 can rescue a CRB 1 phenotype in a mammalian disease model.

The present Example indicates that the phenotype, measured as retina activity using electroretinography, in the eyes that show expression of recombinant human CRB2 is rescued. In absence of expression of recombinant human CRB2 the phenotype is not rescued.

Experiments have been performed using several promoters. We have used the following promoter-gene constructs:

• full length CMV-CRB2 (in rescue experiments in Crb2 cKO and Crb1Crb2^flox/+ cKO mice)
• full length CMV-sCRB 1 (in rescue experiments in Crb1Crb2^flox/+ cKO mice)
• full length CMV-GFP (in expression experiments)
• truncated CMV-GFP (in expression experiments)
• truncated CMV-CRB1 (in rescue and toxicity experiments in Crb1Crb2^flox/+ cKO mice)
• hGRK1-CRB1 (in expression experiments in Crb1 KO mice; the rescue and toxicity experiments will follow)
• hRHO-CRB 1 (in expression experiments in Crb2 KO mice; the rescue and toxicity experiments will follow)
• hGRK1-CRB2 (in expression experiments in Crb2 cKO mice; the rescue experiments will follow)
• hRHO-CRB2 (in expression experiments in Crb2 cKO mice; the rescue experiments will follow)
• RLBP1-GFP (in expression experiments)

Long-term therapy is achievable in a mammalian model of Crumbs homologue (CRB) deficiency, the Crb2 cKO mouse, the Crb1Crb2^F/+ cKO mouse, using the rAAV vector CRB2 constructs disclosed herein. Importantly, these results could not be obtained by the use of short-CRB 1 or full-length CRB 1 constructs because of toxicity, whereas the results could be obtained with the non-toxic CRB2 constructs. Importantly, tools are present to test CRB2 gene therapy vectors in the mice lacking CRB1 and/or CRB2 which mimic different degrees of the LCA8 and RP due to loss of CRB1 phenotype. These results provide evidence for the successful use of rAAV-based CRB2 gene therapy vectors for treatment of retinal dystrophies, and LCA8 and RP due to loss of CRB1 in particular. Experiments have also been performed using AAV2/9-CMV-hCRB2-spA in rescue experiments, and AAV2/5-CMV-hRHO-CRB2-spA and AAV2/5-hGRK1-CRB2-spA in expression experiments, and also using the AAV2/9 vector, or hRHO or hGRK1 promoters, no toxicity was detected when overexpressing human CRB2, whereas overexpression of human short CRB 1 in AAV5 or in AAV9 vectors was toxic.

Toxicity of CRB proteins can be tested using human-derived retinal pigment epithelial cells according to the following Example.

ARPE19 cells (ATCC CRL-2302) are transfected with one of the different (modified) CRB constructs (e.g., CRB 1, sCRB 1, CRB2 isoform 1, CRB2 isoform 2, CRB2 isoform 3, CRB3, etc.) together with a control GFP construct (Aartsen et al., 2010. PLoS One. 5:e12387; UniProtKB/Swiss-Prot sequence P42212) using the calcium phosphate method (described e.g., in Sambrook & Russell (2001). Molecular cloning: a laboratory manual (3rd ed.), Cold Spring Harbor, NY: Cold Spring Harbor Laboratory). As a control, a CRB2 construct is used (CRB2 sequence: SEQ ID NO: 40). The CRB constructs are used in equimolar amounts and a total amount of 20 µg of DNA is added per petridish. CRB constructs are made as described in Example 2.1. Briefly, CRB constructs are made by chemical synthesis and subcloned into pUC57. These constructs comprise AAV2 ITRs (SEQ ID NO: 131 and 132), CMV promoter (SEQ ID NO: 121), CRB cDNA to be tested (e.g., SEQ ID NO: 40 or other CRB sequence, Intron 5 (SEQ ID NO: 128), and synthetic pA (SEQ ID NO: 130).

The GFP construct is used as internal transfection control in a fixed amount. For example, 18 µg of CRB construct plus 2 µg of GFP construct is used. In this way, a series of equimolar plasmid concentrations can be tested while adding the same amount of DNA, such as for example 2, 4, 8 or 16 µg of CRB construct, plus 18, 16, 12 or 4 µg of GFP construct, respectively.

On the day before transfection, ARPE19 cells are plated in duplicate at 30% of confluence in a 10 cm petridish in DMEM supplemented with 10% Fetal Bovine Serine and penicillin/streptomycin. After refreshing the medium 2 hours before transfection, the transfection mix is prepared with 20 µg of DNA in 500 µL of 0.25 M CaCl₂ and TE (10 mM Tris, 1 mM EDTA pH 8) buffer per dish. While constantly vortexing, 500 µL of 2X HBS (281 mM NaCl, 100 mM Hepes, 1.5 mM Na₂HPO₄, pH 7.12) are added drop wise to the transfection mixture and the complete mix is directly added to the cells for overnight incubation. The medium is refreshed in the following morning. Two days later (i.e., 72 h after transfection), the attached and floating cells are harvested separately (one duplicate) and together (the second duplicate) and after centrifugation, resuspended in 1 mL of Phosphate Buffer Saline (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄ and 1.76 mM KH₂PO₄). Subsequently, cells are tested for:

• cell number and viability with a Luna Automated Cell Counter (Logos Biosystems, Inc.; Annandale, USA). The counter determines the number of cells and via Trypan Blue staining discriminates between viable and non-viable cells. Trypan Blue staining was performed using the Standard protocol by Life Technologies as outlined below.
• protein expression by Western Blotting. Proteins from the cell lysates are separated by SDS-page electrophoresis. After transfer to nitrocellulose membrane, the nitrocellulose membrane is immunostained for CRB, GFP and Actin proteins and analyzed by Odyssey Infrared Imaging System (LI-COR; Westburg BV, Leusden, the Netherlands). This method is described in the manual for Western Blot Analysis developed for Aerius, and Odyssey Family of Imagers by Li-Cor, published 2003, revised January 2012. As primary antibodies anti-CRB 1 (AK2, AK5 and AK7; van de Pavert et al., 2004. J Cell Sci. 117(Pt 18):4169-77) and anti-CRB2 (SK II from Pen Rashbash, described in van de Pavert et al., 2004. J Cell Sci. 117(Pt 18):4169-77) and anti-GFP (Becton Dickinson and Company) were used. Secondary antibodies (IRDye 800-CW goat anti chicken, mouse or rabbit, or donkey anti goat) were from Li-Cor.

The following procedure will enable you to accurately determine the cell viability. Cell viability is calculated as the number of viable cells divided by the total number of cells within the grids on the hemacytometer. If cells take up trypan blue, they are considered non-viable.

1. 1. Determine the cell density of your cell line suspension using a hemacytometer.
2. 2. Prepare a 0.4% solution of trypan blue in buffered isotonic salt solution, pH 7.2 to 7.3 (i.e., phosphate-buffered saline).
3. 3. Add 0.1 mL of trypan blue stock solution to 1 mL of cells.
4. 4. Load a hemacytometer and examine immediately under a microscope at low magnification.
5. 5. Count the number of blue staining cells and the number of total cells. $$\%\mathrm{viable}\mathrm{cells}=\left[1.00-\left(\mathrm{Number}\mathrm{of}\mathrm{blue}\mathrm{cells}÷\mathrm{Number}\mathrm{of}\mathrm{total}\mathrm{cells}\right)\right]\times 100$$

To calculate the number of viable cells per mL of culture, use the following formula: $$\mathrm{Number}\mathrm{of}\mathrm{viable}\mathrm{cells}\times {\mathrm{10}}^4\times 1.1=\mathrm{cells}/\mathrm{mL}\mathrm{culture}$$ (Remember to correct for the dilution factor).

Transfection of full length CRB 1 in ARPE-19 cells lead to high number of detached/dead cells and resulted in less than 20% viable CRB1 transfected cells. Transfection of CRB2 in ARPE-19 cells resulted in more than 95% viable transfected cells. This indicates that full length CRB 1 is toxic and/or inhibits cell growth. The amount of CRB 1 expressed in the attached cells is almost undetectable by Western Blot in contrast to GFP in ARPE-19 or CRB 1 in HEK293T cells (Fig. 13). Furthermore, even overloaded three times, the reference protein level (Actin) in CRB1-transfected ARPE-19 is still lower than GFP-transfected ARPE-19. This demonstrates that full length CRB 1 is toxic and/or inhibits cell growth.

For further analysing the toxicity effects by full length CRB 1 compared to CRB2, we use the following constructs: AAV-truncatedCMV-CRB1; AAV-hGRK1-CRB1; AAV-hGRK1-sCRB1; AAV-hRHO-sCRB1.

The Crb1Crb2^ƒlox/+ conditional knock-out mouse lacking CRB 1 in all retinal cells and with reduced levels of CRB2 in all retinal cells except the retinal pigment epithelium (e.g., the Crb1^-/-Crb2^flox/+Chx10Cre on 75% C57BL/6J and 25% 129/Ola genetic background) and Crb2 cKO mice (99.9% C57BL/6J background) were used to evaluate gene replacement therapy using AAV2/9-CMV-CRB2-In5. The Crb1^-/- Crb2^flox/+Chx10Cre mice on 75% C57BL/6J and 25% 129/Ola genetic background exhibit progressive retinal degeneration and scotopic (rod-mediated) and photopic (cone-mediated) loss of retina function as measured by ERG from 3 to 6 months of age (Pellissier et al., 2014. Hum Mol Genet. 23(14):3759-71). The mouse is blind at 12-18 months of age.

AAV-mediated transfer of CRB2 using AAV2/9-CMV-hCRB2-In5 to Crb1Crb2^flox/+ cKO retina restored vision to these animals as evidenced by ERG. AAV-mediated transfer of CRB2 to the postnatal Crb1Crb2^flox/+ cKO retina expressed CRB2 in photoreceptors and Müller glia cells and caused preservation of retinal structure at the time of expression of CRB2.

Subretinal AAV-mediated transfer of CRB2 using 1 µL of 2 × 10¹⁰ genome copies of AAV9 viral particles containing 4.9 kb AAV2-CMV-hCRB2-In5 to Crb1Crb2^flox/+ cKO retina or Crb2 cKO retina restored vision to these animals as evidenced by ERG, Fig. 14 (a-c, g-h). Subretinal AAV9-mediated transfer of CRB2 to the postnatal Crb1Crb2^flox/+ cKO retina or Crb2 cKO retina expressed CRB2 in photoreceptors and Müller glia cells and caused preservation of retinal structure at the time of expression of CRB2.

These experiments showed the feasibility of preserving retinal structure after a single dose of AAV2/9-CMV-hCRB2-In5 (in short AAV-CRB2) even in severely degenerating Crb1Crb2^flox/+ cKO or Crb2 cKO retinas. These data demonstrate that loss of CRB1 in the Crb1^-/-Crb2^ƒlox/+Chx10Cre retinas can be compensated by rescue using AAV-CRB2. In other words, these data demonstrate that elevating levels of CRB2 by using AAV-CRB2 in the Crb1^-/-Crb2^ƒlox/+Chx10Cre retinas can rescue the degeneration phenotype in retinas lacking CRB 1 and having reduced levels of CRB2.

Subretinal AAV-mediated transfer of CRB 1 using 1 µL of 10¹⁰ genome copies of AAV9 viral particles containing 4.8 kb AAV2-minimalCMV-hCRB1 expression vector (i.e., hCRB 1 operably linked to the minimalCMV promoter and flanked by AAV2 ITRs, packaged in AAV9 capsid proteins) to Crb1Crb2^flox/+ cKO retina (99.9% C57BL/6J background) did not restore vision to these animals as evidenced by ERG, Fig. 14 (d-f). As evidenced by immunohistochemistry experiments, subretinal AAV-mediated transfer of CRB 1 to the postnatal Crb1Crb2^flox/+ cKO retina expressed CRB 1 in photoreceptors and Müller glia cells but did not cause preservation of retinal structure at the time of expression of CRB1 (data not shown). These experiments showed the lack of capacity of wild type CRB1 in preserving retinal structure after a single dose of AAV-CRB1 in severely degenerating Crb1Crb2^ƒlox/+ cKO retinas. Example 4 showed that wild type CRB2 can work as a gene replacement therapy, whereas Example 5 demonstrated that wild type CRB 1 in the Crbl^-/-Crb2^flox/+Chx10Cre retinas cannot.

Toxicity of CRB proteins can be tested using Crb1^-/-Crb2^flox/+Chx10Cre mice according to the following Example.

Crb^(-/-Crb2^flox/+Chx10Cre mouse retinas are intravitreally injected with a (modified) CRB construct (e.g., CRB1, short CRB 1, CRB2 isoform 1, CRB2 isoform 2, CRB2 isoform 3, CRB3 etc.) in a recombinant AAV expression vector in one eye, whereas the contralateral eye receives a control AAV-GFP construct (Aartsen et al., 2010. PLoS One. 5:e12387; UniProtKB/Swiss-Prot sequence P42212). The eyes are treated with the vectors using the AAV transduction method (described e.g., in Aartsen et al., 2010. PLoS One. 5:e12387). Control animals receive an AAV-CRB2 construct in one eye and the control AAV-GFP construct in the contralateral eye (CRB2 sequence: SEQ ID NO: 40). The AAV-CRB constructs are intravitreally injected into the eyes of Crbl^-/-Crb2^flox/+Chx10Cre mice in equimolar amounts and a total amount of 1 µL of 5 × 10⁹ to 10¹⁰ genome copies of AAV2/ShH10Y-(CMV or minimalCMV)-CRB and in the contralateral control eye with the same amount of AAV2/ShH10Y-(CMV or minimalCMV)-GFP. AAV-CRB constructs are made as described in Example 2.1. Briefly, CRB constructs are made by chemical synthesis and subcloned into pUC57. These constructs comprise AAV2 ITRs (SEQ ID NO: 131 and 132), CMV promoter (SEQ ID NO: 121) or minimal CMV promoter, CRB cDNA to be tested (e.g., SEQ ID NO: 40 or other CRB sequence, synthetic pA (SEQ ID NO: 130) and an optional Intron 5 (SEQ ID NO: 128). The GFP construct is used as internal transduction control in a fixed amount. Plasmids are packaged in AAV serotype ShH10Y capsids. Intravitreal ShH10Y-mediated transfer of genes to the mouse retina expressed proteins in Müller glia cells and other inner retinal cell types (Pellissier et al., 2014. Mol Ther Methods Clin Dev. 1:14009) as well as the retinal ciliary body.

Three to seven Crb1^-/-Crb2^F/+Chx10Cre^Tg/+ (Crb1Crb2^F/+ cKO) are injected at 2 weeks of age intravitreally with 1 µL of 5 × 10⁹ to 10¹⁰ genome copies of CRB or control GFP viral particles. In vivo retinal function is to be analyzed at 3 to 5 months of age by electroretinography under scotopic (dark-adapted overnight) or photopic (light-adapted with a background illumination of 30 cd/m2 starting 10 minutes before recording) conditions. Mice are anaesthetized using ketamine (66.7 mg/kg body weight) and xylazine (11.7 mg/kg body weight). The pupils are dilated and single royal blue-flash stimuli range from -3 to 1.5 log cd s/m². Twenty responses are averaged with inter-stimulus intervals of 2 s. A-wave responses revealed direct photoreceptor functions (rods and cones under scotopic and only from cones under photopic conditions) and B-waves revealed the retinal activities. A representative experiment is shown in Fig. 15.

Potential toxicity (represented by a decreased retinal activity as determined by ERG) of CRB proteins is measured in comparison to GFP contralateral eyes. Significant reduction of the ERG average responses will be considered as toxicity. An example is shown in Fig. 15, CRB1 protein showed signs of toxicity whereas CRB2 does not.

Retinal expression of CRB proteins upon intravitreal transduction in Crb1Crb2^F/+ cKO or Crb2 cKO eyes is examined by standard immunohistochemistry using antibodies against the respective CRB proteins (e.g., anti-CRB2 or anti-CRB1 or anti-CRB3 as in van de Pavert et al., 2004. J Cell Sci. 117(Pt 18):4169-77).

Intravitreal transduction of full length CRB 1 into Crb1Crb2^F/+ cKO eyes lead to a significant reduced b-wave and a-wave in electroretinograms. Similar experiments using full length CRB2 do not show decreases in b-waves of a-waves. This indicates that full length CRB 1 is toxic (reduces the a- and/or b-waves in electroretinograms) to the Crb1Crb2^F/+ cKO retina when applied intravitreally using 1 µL of 5 × 10⁹ to 10¹⁰ genome copies of capsid ShH10Y particles, whereas CRB2 is not toxic.