RECOMBINANT AAV-CRUMBS HOMOLOGUE COMPOSITION AND METHODS FOR TREATING LCA-8 AND PROGRESSIVE RP

Abstract

The present invention relates to a Crumbs homologue (CRB) therapeutic for use as a medicament or in a method of treatment or prophylaxis, for example in the treatment or prophylaxis of a retinal disorder due to mutations in the Crumbs homologue-1 (CRB1) gene, such as Leber's congenital amaurosis 8 (LCA8) or retinitis pigmentosa 12 (RP12). In particular, the present invention relates to a recombinant viral vector comprising CRB2 or modified non-toxic forms of either CRB1 or CRB3 that resemble CRB2.

Claims

1.-15. (canceled)

16. A gene therapy vector comprising at least one of: a) a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein; and b) a nucleotide sequence encoding a modified Crumbs homologue-I (CRB1) protein or a modified Crumbs homologue-3 (CRB3) protein, wherein the modified CRB1 protein or the modified CRB3 protein comprises a modification in the C-terminal part of the protein and is less toxic or not toxic in an in vivo assay, which assay comprises the steps of: i) intravitreal transduction in one eye of a Crb2 conditional knock-out (cKO) or Crb1Crb2F/+ cKO mouse with recombinant adeno-associated virus (rAAV) comprising the nucleotide sequence encoding the modified CRB1 or modified CRB3 protein; ii) intravitreal transduction in the other eye of the mouse with rAAV comprising a nucleotide sequence encoding a wild-type CRB2 protein, wild-type CRB1 or wild-type CRB3 as a control; and, iii) making an electroretinogram one and three months after transduction, wherein an increased percentage in the maximum a-wave and/or b-wave amplitude in the electroretinogram in the modified CRB1 or modified CRB3 transduced retinas compared to the electroretinogram in the wild-type CRB1 or wild-type CRB3 transduced retinas indicates that the modified CRB1 protein or the modified CRB3 protein is less toxic; or wherein a maximum a-wave and/or b-wave amplitude in the electroretinogram of at least 60% of the difference of maximum a-wave and/or b-wave amplitude with wild type CRB2 subtracted with the maximum a-wave and/or b-wave amplitude with modified CRB1 or modified CRB3 protein indicates that the modified CRB1 protein or the modified CRB3 protein is not toxic.

17. A gene therapy vector according to claim 16, wherein at least one of: a) the CRB2 protein is a eumetazoan CRB2 protein; b) the modified CRB1 protein is a modified eumetazoan CRB1 protein; and c) the modified CRB3 protein is a modified eumetazoan CRB3 protein.

18. A gene therapy vector according to claim 16, wherein the gene therapy vector is a recombinant parvoviral vector or a lentiviral vector.

19. A gene therapy vector according to claim 18, wherein the vector is a recombinant adeno-associated virus (rAAV) vector.

20. A gene therapy vector according to claim 19, wherein 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 adeno-associated 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 (rAA9), or a serotype variant of any thereof.

21. A gene therapy vector according to claim 16, wherein 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, 65-83 and wherein the CRB2 protein is functionally active as measured by electroretinography.

22. A gene therapy vector according to claim 16, wherein the nucleotide sequence encoding CRB2, modified CRB1 or modified CRB3 is operably linked to expression control elements comprising a promoter that produces sufficient expression of CRB2, modified CRB1 or modified CRB3 protein, respectively, to obtain a therapeutic effect.

23. A gene therapy vector according to claim 16, wherein the modification in the C-terminal part of the amino acid sequence of the modified CRB1 or modified CRB3 protein is selected from the group consisting of: i) the PDZ binding domain of CRB1 or CRB3 is replaced by amino acid residues 1282-1285 of SEQ ID NO:40; ii) the FERM binding domain of CRB1 or CRB3 is replaced by amino acid residues 1251-1264 of SEQ ID NO:40; iii) the transmembrane domain of CRB1 or CRB3 is replaced by amino acid residues 1225-1247 of SEQ ID NO:40; iv) the 16 C-terminal amino acid residues of CRB1 or CRB3 are replaced by amino acid residues 1270-1285 of SEQ ID NO:40 or by conserved amino acid substitutions of amino acid residues 1270-1285 of SEQ ID NO:40; v) an amino acid sequence of CRB1 or CRB3 consisting of the FERM binding domain of CRB1 or CRB3, two N-terminal amino acid residues and five C-terminal amino acid residues is replaced by amino acid residues 1249-1269 of SEQ ID NO:40 or by conserved amino acid substitutions of amino acid residues 1249-1269 of SEQ ID NO:40; vi) an amino acid sequence of CRB1 or CRB3 consisting of the transmembrane domain of CRB1 or CRB3, two N-terminal amino acid residues and one C-terminal amino acid residue is replaced by amino acid residues 1223-1248 of SEQ ID NO:40 or by conserved amino acid substitutions of amino acid residues 1223-1248 of SEQ ID NO:40; vii) an amino acid sequence of CRB1 or CRB3 comprising any of a serine, threonine or tyrosine at a position that corresponds to position 1243, 1254, 1257, 1258, 1259, 1261 and/or 1274 of SEQ ID NO:40; and, viii) combination of one or more i)-vii).

24. A nucleic acid construct comprising at least one of: a) a nucleotide sequence encoding a Crumbs homologue-2 (CRB2) protein and at least one parvoviral inverted terminal repeat (ITR) sequence, wherein preferably 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; and b) a nucleotide sequence encoding a Crumbs homologue-1 (modified CRB1) protein or a Crumbs homologue-3 (modified CRB3) protein, and at least one ITR sequence, wherein preferably the nucleotide sequence encoding a modified CRB1 protein or a modified CRB3 protein is operably linked to expression control elements comprising a promoter that is capable of sufficient modified CRB1 or modified CRB3 protein expression to obtain a therapeutic effect and wherein the modified CRB1 protein or the modified CRB3 protein comprises a modification in the C-terminal part of the protein and is less toxic or not toxic in an in vivo assay, which assay comprises the steps of: i) intravitreal transduction in the one eye of a Crb2 conditional knock-out (cKO) or Crb1Crb2F/+ cKO mouse with recombinant adeno-associated virus (rAAV) comprising the nucleotide sequence encoding the modified CRB1 or modified CRB3 protein; ii) intravitreal transduction in the other eye of the mouse with rAAV comprising a nucleotide sequence encoding a wild-type CRB2 protein, wild-type CRB1 or wild-type CRB3 as a control; and, iii) making an electroretinogram one and three months after transduction, wherein an increased percentage in the maximum a-wave and/or b-wave amplitude in the electroretinogram in the modified CRB1 or modified CRB3 transduced retinas compared to the electroretinogram in the wild-type CRB1 or wild-type CRB3 transduced retinas indicates that the modified CRB1 protein or the modified CRB3 protein is less toxic; or wherein a maximum a-wave and/or b-wave amplitude in the electroretinogram of at least 60% of the difference of maximum a-wave and/or b-wave amplitude with wild type CRB2 subtracted with the maximum a-wave and/or b-wave amplitude with modified CRB1 or modified CRB3 protein indicates that the modified CRB1 protein or the modified CRB3 protein is not toxic.

25. A pharmaceutical composition comprising a gene therapy vector according to claim 16 and a pharmaceutically acceptable excipient.

26. A kit comprising: (a) a gene therapy vector in accordance with claim 16, or a pharmaceutical composition comprising the gene therapy vector and a pharmaceutically acceptable carrier; and, (b) instructions for using the gene therapy 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.

27. A method for treatment or prophylaxis of a retinal disorder due to mutations in a CRB1 gene, comprising administering an effective amount gene therapy vector according to claim 16 to a subject in need thereof.

28. A method according to claim 27, wherein the retinal disorder is Leber's congenital amaurosis or retinitis pigmentosa.

29. A method according to claim 28, wherein the retinal disorder is LCA8 or RP12.

30. A gene therapy vector according to claim 17, wherein: a) the CRB protein is a CRB2 protein of human, non-human primate, murine, feline, canine, porcine, ovine, bovine, equine, epine, caprine, or lupine origin; b) the modified CRB1 protein is a modified CRB1 protein of human, non-human primate, murine, feline, canine, porcine, ovine, bovine, equine, epine, caprine, or lupine origin; or c) the modified CRB3 protein is a modified CRB3 protein of human, non-human primate, murine, feline, canine, porcine, ovine, bovine, equine, epine, caprine, or lupine origin.

31. A gene therapy vector according to claim 30, wherein: a) the CRB2 protein is a human CRB2 protein; b) the modified CRB1 protein is a modified human CRB1 protein; or c) the modified CRB3 protein is a modified human CRB3 protein.

32. A gene therapy vector according to claim 20, wherein the serotype variant comprises rAAV6 ShH10, rAAV6 ShH10Y, or a combination thereof.

33. A gene therapy vector according to claim 21, wherein the CRB2 protein comprises or consists of an amino acid sequence that has at least 80% sequence identity with the amino acid sequence of any one of SEQ ID NO: 40-42.

34. A gene therapy vector according to claim 22, wherein the promoter is selected from the group consisting of: a truncated CMV promoter, a CMV promoter, a truncated human RLBP1 promoter, a human photoreceptor specific rhodopsin kinase promoter, and a human rod photoreceptor specific rhodopsin promoter.

35. A gene therapy vector according to claim 34, wherein the promoter is selected from the group consisting of: a CMV promoter according to SEQ ID NO: 121, a truncated human RLBP1 promoter according to SEQ ID NO:122, a human photoreceptor specific rhodopsin kinase promoter according to SEQ ID NO:123, a human rod photoreceptor specific rhodopsin promoter according to SEQ ID NO:124, and a truncated CMV promoter according to SEQ ID NO:133.

36. A gene therapy vector according to claim 23, wherein the amino acid residues at positions 1243, 1254, 1259, 1261 and 1274 are serine; the amino acid residue at position 1257 is threonine; and the amino acid residue at position 1258 is tyrosine.

Description

DESCRIPTION OF THE FIGURES

[0113] FIG. 1. A representative presentation of the localization of CRB1 in the human retina.

[0114] FIG. 2. A representative presentation of the localization of CRB1 in the mouse retina.

[0115] FIG. 3. A representation of the mammalian Crumbs homologue protein family.

[0116] FIG. 4. Degeneration in the ventral but not in the dorsal retina of Crb1.sup.−/−Crb2.sup.F/+Chx10Cre/+ mice. FIGS. 4A-4E, Technovit sections. Left panels, dorsal (superior) retina. Right panels, ventral (inferior) retina.

[0117] FIG. 5. Electroretinogram b-waves of Crb1.sup.−/−Crb2.sup.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.sup.−/− and Crb1.sup.−/−Crb2F/F mice (not containing Chx10Cre) do not show differences in retinal activity (data not shown). FIG. 5A, scotopic ERG showing loss of rod photoreceptor activity in Crb1.sup.−/−Crb2.sup.F/+Chx10Cre retinas (light grey; lower line) vs. Crb1.sup.−/−Crb2.sup.F/F retinas (dark grey; upper line). FIG. 5B, photopic ERG showing loss of cone photoreceptor activity in Crb1.sup.−/−Crb2.sup.F/+Chx10Cre retinas (light grey; lower line)) vs. Crb1.sup.−/−Crb2.sup.F/F retinas (dark grey; upper line).

[0118] FIG. 6. Loss of a separate photoreceptor layer in Crb1.sup.−/−Crb2.sup.F/FChx10Cre/+ retinas. FIGS. 6A-6E, Technovit sections. Left panels, control retina. Middle panels, Crb1.sup.+/−Crb2.sup.F/FChx10Cre/+ retinas. Right panels, Crb1.sup.−/−Crb2.sup.F/FChx10Cre/+ retinas showing absence of a separate photoreceptor layer and mislocalized retinal cells.

[0119] FIG. 7. Loss of retinal activity in Crb1Crb2 cKO compared to Crb1 and Crb2 cKO retinas. FIGS. 7A and 7C, measured ERGs at 3 months of age. FIGS. 7B and 7D, measured ERGs at 1 months of age. FIGS. 7A and 7B, scotopic. FIGS. 7C and 7D, 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.sup.+/−Crb2.sup.F/FChx10Cre (heterozygote Crb1.sup.+/− homozygote foxed Crb2.sup.F/F heterozygote Chx10Cre), the bottom line is Crb1.sup.−/−Crb2.sup.F/FChx10Cre (homozygote Crb1.sup.−/− homozygote floxed Crb2.sup.F/F heterozygote Chx10Cre).

[0120] FIG. 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.

[0121] FIG. 9. Expression of short CRB1 (SEQ ID NO. 3) in Crb1 KO retinas using subretinal injection of AAV2/9-CMV-sCRB1 vectors. FIG. 9A, control Crb1 KO retina. FIG. 9B, 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).

[0122] FIG. 10. Representative experiment showing rescue of loss of retinal activity. Crb2 cKO retinas were injected at postnatal day 23 subretinally with 10.sup.10 AAV2/ShH10Y-CMV-CRB2 or AAV2/ShH10Y-CMV-GFP viral particles and analyzed for ERG and immunohistochemistry at 3 months of age. FIG. 10A, 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). FIG. 10B, immunohistochemistry showing expression of sCRB1 in the right eye of the animal used in FIG. 10A. No expression of CRB2 was detected in the left eye of the same animal. Abbreviations: OLM, outer limiting membrane; RPE retinal pigment epithelium.

[0123] FIG. 11. Specific transduction of Müller glia cells using 10.sup.10 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. FIG. 11A, scanning-laser-ophtalmoscopy (SLO). FIG. 1113, SLO showing fluorescent cells. FIG. 11C, immunohistochemistry showing specific expression of GFP in Müller glia cells.

[0124] FIG. 12. AAV6 and ShH10Y capsids transduce adult human Müller glia cells. One μl 10.sup.13 genome copies per ml of AAV2/6-CMV-GFP-WPRE-pA (FIG. 12A) or AAV6 variant AAV2/ShH10-CMV-GFP-WPRE-pA (FIG. 12B) was applied to pieces of cultured adult human retina. GFP expression was detected in Müller glia cells.

[0125] FIG. 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 pAAV-CMV-hCRB1-pA vectors showed subsequent CRB1 and GFP protein levels. However, whereas RPE-derived ARPE-19 cells expressed normal amount of GFP, CRB1 protein is just above detection level in three times overloaded protein lysates.

[0126] FIG. 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.sup.−/−Crb2.sup.F/+Chx10Cre.sup.Tg/+ (Crb1Crb2.sup.F/+ cKO; FIGS. 14A-14F) and Crb2.sup.F/F Chx10Cre.sup.Tg/+ (Crb2 cKO; FIGS. 14G-14H) mouse retinas injected subretinally at 2 weeks of age with 1 μL of 2.Math.10.sup.10 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 (FIGS. 14A-14C; 14G-14H), or with 1 μl of 1×10.sup.10 genome copies of 4.8 kb AAV2-minimalCMV-CRB1-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 (FIGS. 14D-14F), and analyzed at 3 or 4 months of age by electroretinography under scotopic (dark-adapted; FIGS. 14A-14B, 14D-14E, 14G-14H) or photopic (light-adapted; FIGS. 14C, 14F) conditions. Scotopic b-wave amplitudes (FIGS. 14A, 14D, 14G) and a-wave amplitudes (FIGS. 14B, 14E, 14H), and photopic b-wave amplitudes (FIGS. 14C, 14F) are indicated. CRB2 vectors rescued loss of retinal function in two different Crb mutant mouse models (FIGS. 14A-14C, 14G-14H), whereas CRB1 vectors did not rescue loss of retinal function (FIGS. 14D-14F).

[0127] FIG. 15. Toxicity of CRB proteins tested by intravitreal injection of AAV2/ShH10Y-minimalCMV-CRB1-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.sup.−/−Crb2.sup.F/+Chx10Cre.sup.Tg/+ (Crb1Crb2.sup.F/+ cKO; a-e) mouse retinas injected intravitreally at 2 weeks of age with 1 μL of 10.sup.10 genome copies of 4.9 kb AAV2/ShH10Y-CMV-CRB2-In5-spA and in the contralateral control eye with AAV2/ShH10Y-CMV-GFP (FIGS. 15A-15B), or with 1 μL of 5109 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 (FIGS. 15C-15E). The eyes were analyzed at 3 months of age by electroretinography under scotopic (dark-adapted; FIGS. 15A-15B, 15C-15D) or photopic (light-adapted; FIG. 15E) conditions. Scotopic b-wave amplitudes (FIGS. 15A, 15C) and a-wave amplitudes (FIGS. 15B, 15D), and photopic b-wave amplitudes (FIG. 15E) are indicated. No statistical significant differences in retinal function were detected for intravitreally applied CRB2 vectors compared to GFP control vectors (FIGS. 15A-15B). Intravitreally applied CRB1 vectors showed strongly reduced retinal responses upon expression of CRB1 vectors, suggesting toxic effects by CRB1 vectors.

EXAMPLES

Description of the Mouse Models

[0128] Crb1.sup.−/+Crb2.sup.F/FChx10Cre/+ and Crb1.sup.−/−Crb2.sup.F/FChx10Cre/+ Mice

[0129] Retinas of Crb1.sup.−/+Crb2.sup.F/FChx10Cre/+ mice (heterozygote for Crb1, homozygote for floxed Crb2) show to some extent a similar but more severe phenotype than observed in Crb2.sup.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.sup.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.sup.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.

[0130] Retinas of Crb1.sup.−/−Crb2.sup.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. Crb1.sup.−/−Crb2.sup.F/+Chx10Cre/+ Mice

[0131] In Crb1.sup.−/−Crb2.sup.F/+Chx10Cre/+ 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.sup.−/− 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.sup.−/− retinas in which only one quadrant (inferior temporal) and Crb1.sup.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.sup.+/−Crb2.sup.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.sup.−/−Crb2.sup.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.sup.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.

Example 1

[0132] Expression of Short Human CRB1 in Immune Naïve CRB1 Knockout Retina is Toxic

[0133] 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 sCRB11ΔE3/4). In our initial trials we generated AAV vectors with the CMV promoter, the short CRB1, a synthetic intron (In5) in the short CRB1 cDNA sequence, and a synthetic spA. Upon subretinal injection of this vector packaged in AAV serotype 9 (AAV9), we detected significant expression of short CRB1 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 (1.00×10.sup.10 delivered vector genomes) plus a ten-fold lower dose of AAV2/9-CMV-GFP-WPRE-pA (1.00×10.sup.9 delivered vector genomes) into the left eye of retinas lacking CRB1 with reduced levels of CRB2 (Crb1.sup.−/−Crb2.sup.flox/+Chx10Cre retinas). The contralateral control eye received one μl of AAV2/9-CMV-GFP-WPRE-pA (1.00×10.sup.10 delivered vector genomes). Crb1.sup.−/−Crb2.sup.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 CRB1 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 CRB1 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 naïve Crb1 knockout retina, ectopic expression effect, incompatibility of mouse and human CRB1 protein, differences between short and full length CRB1, interference of short CRB1 with the expression of other CRB1 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) CRB1 in cultured cell lines. Preliminary studies expressing the short CRB1 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-naïve 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 (10.sup.10 delivered vector genomes) but toxic effects were not detected.

Example 2

[0134] AAV-Mediated Gene Therapy Restores Visual Function and Behavior in a Mouse Model of Retinitis Pigmentosa (RP) Due to Loss of Crumbs Homologue (CRB) Function

[0135] 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.

[0136] 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.

[0137] 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 CRB1 gene.

[0138] 2.1 Materials and Methods:

[0139] Experimental Animals:

[0140] 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/flox)Chx10Cre 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.

[0141] Construction of AAV Vectors:

[0142] 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, Calif., 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) 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.

[0143] 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, N.J., USA). The BglII CMV-hCRB2In5-spA fragment was cloned into pUC57 (Thermo Fisher Scientific, Waltham, Mass., 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.

[0144] AAV vectors were packaged and purified by iodixanol gradient ultra-centrifugation according to previously published methods (Zolotukhin et al., 1999; Hermens et al., 1999; Ehlert et al., 2010). Viral particles were diluted, washed and concentrated using an Amicon 100 kDa MWCO Ultra-15 device (Millipore, Billerica, Mass., USA) in Dulbecco's Balanced Salt Solution (Life Technologies, Bleiswijk, Netherlands) and titered by quantitative real-time PCR (Aartsen et al, 2010). Resulting titers were 1.00×10.sup.13 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).

[0145] Subretinal Injections:

[0146] In a typical experiment, one μl of AAV2/ShH10Y-CMV-hCRB2 (1.00×10.sup.10 delivered vector genomes) plus a ten-fold lower dose of AAV2/ShH10Y-CMV-GFP-WPRE-pA (1.00×10.sup.9 delivered vector genomes) was delivered subretinally at postnatal day 23 (P23) to the left eye of each Crb2(flox/flox)Chx10Cre mouse. The contralateral control right eye was injected with one pi of AAV2/ShH10Y-CMV-GFP-WPRE-pA (1.00×10.sup.10 delivered vector genomes). Subretinal injections were performed as previously described (Aartsen et al., 2010). 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; Timmers et al., 2001).

[0147] Electroretinographic Analysis:

[0148] 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). Initial ERG measurements were recorded at 4 weeks' postinjection, and each subsequent 2 weeks thereafter, until 10 weeks post-injection (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/m.sup.2 to 1.5 cds/m.sup.2). Interstimulus intervals for low intensity stimuli were 1.1 second. At the three highest intensities (100 mcds/m.sup.2, 1 cds/m.sup.2 and 5 cds/m.sup.2), 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/m.sup.2 white background for 2 min. Photopic cone responses were elicited with a series of five increasing light intensities (100 mcds/m.sup.2 to 12 cds/m.sup.2). Fifty responses were recorded and averaged at each intensity. All stimuli were presented in the presence of the 100 cds/m.sup.2 background. B-wave amplitudes were defined as the difference between the a-wave troughs to the positive peaks of each waveform.

[0149] Alternatively, ERGs recordings were elicited with a series of light pulses of increasing intensities (2.7 cds/m.sup.2 to 25 cds/m.sup.2, 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.

[0150] 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 post-injection vs. 10 weeks post-injection). Significant difference was defined as a P-value<0.05.

[0151] Tissue Preparation:

[0152] 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 cryosectioning were prepared according to previously described methods (Haire et al., 2006). 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, Calif., USA) and snap-frozen in a bath of dry ice/ethanol. Eyes were serially sectioned at 10 m 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). 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.

[0153] Immunohistochemistry and Microscopy:

[0154] Retinal cryosections and whole mounts were washed 3× in 1×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 SKi1 (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× with 1×PBS.

[0155] Retinal sections were incubated for 1 hr at room temperature with IgG secondary antibodies tagged with Cyanine dye Cy5 (Molecular Probes, Eugene, Oreg., USA) diluted 1:500 in 1×PBS. Following incubation with secondary antibodies, sections and whole mounts were washed with 1×PBS. Retinal sections were counterstained with 4′,6′-diamino-2-phenylindole (DAPI) for 5 min at room temperature. After a final rinse with 1×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.

[0156] Retinal sections were analyzed with confocal microscopy (Leica TCS SP5 AOBS Spectral Confocal Microscope equipped with LCS Version 2.61, Build 1537 software, (Bannockburn, Ill., USA). All images were taken with identical exposure settings at either 20× or 63× 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, N.Y., 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 5× under identical exposure settings and then merged together in Photoshop® (Version 7.0) (Adobe, San Jose, Calif., USA).

[0157] 2.2 Results

[0158] Photoreceptor Function (ERG) and Restored in AAV-Treated Crb2 cKO Mice:

[0159] 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). 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). 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). 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.

[0160] 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; Timmers et al., 2001). 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.

[0161] The Ubiquitous CMV Promoter Drives hCRB2 Transgene Expression in Müller Glia Cells and Photoreceptors of Crb2 cKO Mice:

[0162] CRB1-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 PALS1 (Park et al., 2011), albeit at lower levels than at the outer limiting membrane (Pellissier L P, Lundvig D M, Tanimoto N, Klooster J, Vos R M, Richard F, Sothilingam V, Garcia Garrido M, Le Bivic A, Seeliger M W, Wijnholds J. Hum Mol Genet. 2014 Jul. 15; 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).

[0163] It is important to note that while CRB1-deficiency in humans causes LCA8 and progressive RP very well detectable by ERG, CRB1-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 CRB1 is restricted to the “outer limiting membrane” of Müller glia cells, whereas in humans CRB1 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 CRB1 with reduced levels of CRB2, mice lacking CRB2 with reduced levels of CRB1, and mice lacking both CRB1 and CRB2 suggest very similar functions for CRB1 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), and the zebrafish CRB2B protein can rescue the phenotype in zebrafish lacking CRB2A protein (Omori & Malicki, 2006).

[0164] 2.3 Discussion

[0165] 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 CRB1.

[0166] Experiments have been carried out in several mouse models.

[0167] Short human CRB1 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 CRB1 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 (1.00×10.sup.10 delivered vector genomes) viral particles subretinally injected into P14 Crb1.sup.−/−Crb2.sup.F/+Chx10Cre retinas (on 75% C57BL/6J and 25% 129/Ola genetic background) that were analyzed at 4 months of age.

[0168] 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.

[0169] Human CRB2 was overexpressed in retinas of mice lacking CRB1 and with reduced levels of CRB2 (Crb1.sup.−/−Crb2.sup.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 CRB1 (on a genetic background with reduced levels of CRB2). Retinas of control Crb2.sup.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 Crb1Crb2.sup.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 Crb1Crb2.sup.F/+ mutant mice are rescued phenotypically, and the confidence intervals are separated and well interpretable. These experiments show that CRB2 can rescue a CRB1 phenotype in a mammalian disease model.

[0170] 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.

[0171] Experiments have been performed using several promoters. We have used the following promoter-gene constructs: [0172] full length CMV-CRB2 (in rescue experiments in Crb2 cKO and Crb1Crb2(flox/+) cKO mice) [0173] full length CMV-sCRB1 (in rescue experiments in Crb1 Crb2(flox/+) cKO mice) [0174] full length CMV-GFP (in expression experiments) [0175] truncated CMV-GFP (in expression experiments) [0176] truncated CMV-CRB1 (in rescue and toxicity experiments in Crb1Crb2(flox/+) cKO mice) [0177] hGRK1-CRB1 (in expression experiments in Crb1 KO mice; the rescue and toxicity experiments will follow) [0178] hRHO-CRB1 (in expression experiments in Crb2 KO mice; the rescue and toxicity experiments will follow) [0179] hGRK1-CRB2 (in expression experiments in Crb2 cKO mice; the rescue experiments will follow) [0180] hRHO-CRB2 (in expression experiments in Crb2 cKO mice; the rescue experiments will follow) [0181] RLBP1-GFP (in expression experiments)

[0182] 2.4. Conclusion

[0183] Long-term therapy is achievable in a mammalian model of Crumbs homologue (CRB) deficiency, the Crb2 cKO mouse, the Crb1Crb2F/+ cKO mouse, using the rAAV vector CRB2 constructs disclosed herein. Importantly, these results could not be obtained by the use of short-CRB1 or full-length CRB1 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 CRB1 in AAV5 or in AAV9 vectors was toxic.

Example 3

[0184] Toxicity Test of CRB Proteins in the ARPE-19 Cell Line by Cell Counting and Western Blotting

[0185] 3.1. Materials and Methods

[0186] 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. CRB1, sCRB1, 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; GFAP-driven transgene expression in activated Müller glial cells following intravitreal injection of AAV2/6 vectors; UniProtKB/Swiss-Prot sequence P42212) using the calcium phosphate method (described e.g. in Sambrook and Russell (2001) “Molecular Cloning: A Laboratory Manual (3.sup.rd edition), Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, New York). 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.

[0187] 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 0.25M CaCl.sub.2) and TE (10 mM Tris, 1 mM EDTA pH 8) buffer per dish. While constantly vortexing, 500 μl of 2×HBS (281 mM NaCl, 100 mM Hepes, 1.5 mM Na.sub.2HPO.sub.4, 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 Na.sub.2HPO.sub.4 and 1.76 mM KH.sub.2PO.sub.4). Subsequently, cells are tested for: [0188] 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. [0189] 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 (http://biosupport.licor.com/docs/Wester_Blot_Analysis_11488.pdf. As primary antibodies anti-CRB1 (AK2, AK5 and AK7; van de Pavert et al., 2004) and anti-CRB2 (SK II from Pen Rashbash, described in van de Pavert et al., 2004) 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.

[0190] Trypan Blue Staining Using the Standard Protocol by Life Technologies:

[0191] Protocol 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. [0192] 1. Determine the cell density of your cell line suspension using a hemacytometer. [0193] 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). [0194] 3. Add 0.1 mL of trypan blue stock solution to 1 mL of cells. [0195] 4. Load a hemacytometer and examine immediately under a microscope at low magnification. [0196] 5. Count the number of blue staining cells and the number of total cells.

% viable cells=[1.00−(Number of blue cells−Number of total cells)]×100

[0197] To calculate the number of viable cells per mL of culture, use the following formula:

Number of viable cells×10.sup.4×1.1=cells/mL culture (Remember to correct for the dilution factor).

[0198] 3.2 Results

[0199] Transfection of full length CRB1 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 CRB1 is toxic and/or inhibits cell growth. The amount of CRB1 expressed in the attached cells is almost undetectable by Western Blot in contrast to GFP in ARPE-19 or CRB1 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 CRB1 is toxic and/or inhibits cell growth.

[0200] For further analysing the toxicity effects by full length CRB1 compared to CRB2, we use the following constructs: AAV-truncatedCMV-CRB1; AAV-hGRK1-CRB1; AAV-hGRK1-sCRB1; AAV-hRHO-sCRB1.

Example 4

[0201] Gene Replacement Therapy in Crb1(−/−) Crb2(Flox/+) Chx10Cre and Crb2(Flox/Flox)Chx10Cre Mice Using AAV2/9-CMV-CRB2-In5

[0202] The Crb1Crb2.sup.flox/+ conditional knock-out mouse lacking CRB1 in all retinal cells and with reduced levels of CRB2 in all retinal cells except the retinal pigment epithelium (e.g. the Crb1.sup.−/−Crb2.sup.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.sup.−/−Crb2.sup.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 L. P. (2014) CRB2 acts as a modifying factor of CRB1-related retinal dystrophies in mice. Hum Mol Genet. July 15; 23(14):3759-71). The mouse is blind at 12-18 months of age.

[0203] AAV-mediated transfer of CRB2 using AAV2/9-CMV-hCRB2-In5 to Crb1Crb2.sup.flox/+ cKO retina restored vision to these animals as evidenced by ERG. AAV-mediated transfer of CRB2 to the postnatal Crb1Crb2.sup.flox/+ cKO retina expressed CRB2 in photoreceptors and Müller glia cells and caused preservation of retinal structure at the time of expression of CRB2.

[0204] Subretinal AAV-mediated transfer of CRB2 using 1 μL of 2.Math.10.sup.10 genome copies of AAV9 viral particles containing 4.9 kb AAV2-CMV-hCRB2-In5 to Crb1Crb2.sup.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.sup.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.

[0205] 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.sup.flox/+ cKO or Crb2 cKO retinas. These data demonstrate that loss of CRB1 in the Crb1.sup.−/−Crb2.sup.flox/+Chx10.sup.Cre 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.sup.−/−Crb2.sup.flox/+Chx10Cre retinas can rescue the degeneration phenotype in retinas lacking CRB1 and having reduced levels of CRB2.

Example 5

[0206] Lack of Gene Replacement Therapy in Crb1(−/−)Crb2(Flox/+)Chx10Cre Using AAV2/9-CMV-CRB1

[0207] Subretinal AAV-mediated transfer of CRB1 using 1 μL of 10.sup.10 genome copies of AAV9 viral particles containing 4.8 kb AAV2-minimalCMV-hCRB1 expression vector (i.e., hCRB1 operably linked to the minimalCMV promoter and flanked by AAV2 ITRs, packaged in AAV9 capsid proteins) to Crb1Crb2.sup.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 CRB1 to the postnatal Crb1Crb2.sup.flox/+ cKO retina expressed CRB1 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.sup.flox/+ cKO retinas. Example 4 showed that wild type CRB2 can work as a gene replacement therapy, whereas example 5 demonstrated that wild type CRB1 in the Crb1.sup.−/−Crb2.sup.flox/+Chx10Cre retinas cannot.

Example 6

[0208] Toxicity Test of CRB Proteins in Crb1(−/−)Crb2(Flox/+)Chx10Cre Mice

[0209] Toxicity of CRB proteins can be tested using Crb1(−/−)Crb2(flox/+)Chx10Cre mice according to the following Example.

[0210] 6.1. Materials and Methods

[0211] Crb1(−/−)Crb2(flox/+)Chx10Cre mouse retinas are intravitreally injected with a (modified) CRB construct (e.g. CRB1, short CRB1, 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; GFAP-driven transgene expression in activated Müller glial cells following intravitreal injection of AAV2/6 vectors; 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; GFAP-driven transgene expression in activated Müller glial cells following intravitreal injection of AAV2/6 vectors). Control animals receive a 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 Crb1(−/−)Crb2(flox/+)Chx10Cre mice in equimolar amounts and a total amount of 1 μL of 5.Math.10.sup.9 to 10.sup.10 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) Molecular Therapy Methods & Clinical Development 1:14009; Specific tools for targeting and expression in Müller glial cells) as well as the retinal ciliary body.

[0212] Three to seven Crb1.sup.−/−Crb2.sup.F/+Chx10Cre.sup.Tg/+ (Crb Crb2.sup.F/+ cKO) are injected at 2 weeks of age intravitreally with 1 μL of 5.Math.10.sup.9 to 10.sup.10 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 interstimulus 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.

[0213] 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.

[0214] Retinal expression of CRB proteins upon intravitreal transduction in Crb1Crb2.sup.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., J. Cell Science, 2004).

[0215] 6.2 Results

[0216] Intravitreal transduction of full length CRB1 into Crb1 Crb2.sup.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 CRB1 is toxic (reduces the a- and/or b-waves in electroretinograms) to the Crb1Crb2.sup.F/+ cKO retina when applied intravitreally using 1 μL of 5.Math.10.sup.9 to 10.sup.10 genome copies of capsid ShH10Y particles, whereas CRB2 is not toxic.

REFERENCES

[0217] The following references, to the extent that they provide exemplary procedural or other details supplementary to those set forth herein, are specifically incorporated herein by reference. [0218] International Publication Number WO 2011/133933 and references to patents mentioned herein. [0219] Aartsen. W. M. (2010). PLoS One 5, e12387. [0220] Acland, G. M. (2005). Mol Ther 12, 1072-1082. [0221] Acland. G. M. (2001). Nat Genet 28, 92-95. [0222] Aleman, T. S. (2011). Invest Ophthalmol Vis Sci 52, 6898-6910. [0223] Alexander. J. J. (2007). Nat Med 13, 685-687. [0224] Allocca, M (2007) J Virol 81, 11372-11380. [0225] Alves, C. H. (2013). Hum Mol Genet 22, 35-50. [0226] Alves C. H. (2014) Hum Mol Genet. 2014 Jul. 1; 23(13):3384-401. [0227] Amado, D. (2010). Sci Transl Med 2, 21ra16. [0228] Annear. M. J. (2011). Gene Ther 18, 53-61. [0229] Ashtari, M. (2011). J Clin Invest 121, 2160-2168. [0230] Bainbridge, J. W (2009). Eye (Loud) 23, 1898-1903. [0231] Bainbridge, J. W. (2008). N Engl J Med 358, 2231-2239. [0232] Bainbridge, J. W. (2009). J Gene Med 11, 486-497. [0233] Bazellieres, E. (2009). Front Biosci 14, 2149-2169. [0234] Beltran, W. A. (2010). Gene Ther 17, 1162-1174. [0235] Bennett, J. (2012). Sci Transl Med 4, 120ra15 [0236] Bennicelli, J. (2008). Mol Ther 16, 458-465. [0237] Bianchi, S. (2010). J Neurol 257, 1039-1042. [0238] Blits, B. (2010). J Neurosci Methods 185, 257-263 [0239] Booij, J. C. (2005). J Med Genet 42, e67. [0240] Boutin, S (2010) Hum Gene Ther 21, 704-712 [0241] Boye, S. E. (2012) Hum Gene Ther 23, 1101-1115. [0242] Boye, S. E. (2013a). Mol Ther. [0243] Boye, S E. (2010). PLoS One 5, e11306. [0244] Boye, S. L. (2011). Invest Ophthalmol Vis Sci 52, 7098-7108. [0245] Boye, S. L. (2013b). Hum Gene Ther 24, 189-202. [0246] Buch, P. K. (2008). Gene Ther 15, 849-857. [0247] Bujakowska, K. (2012). Hum Mutat 33, 306-315 [0248] Bulgakova, N. A. (2009). J Cell Sci 122, 2587-2596. [0249] Chuing, D C. (2009). JAAPOS 13, 587-592 [0250] Cideciyan, A. V. (2010). Prog Retin Eye Res 29, 398-427. [0251] Cideciyan, A. V. (2008). Proc Natl Acad Sci USA 105, 15112-15117. [0252] Cideciyan, A. V. (2009). Hum Gene Ther 20, 999-1004. [0253] Cideciyan, A. V. (2013). Proc Natl Acad Sci USA 110. E517-E525. [0254] Corton, M. (2013). Orphanet J Rare Dis 8, 20. [0255] Cremers, F. P. (2004). Novartis Found Symp 255, 68-79. [0256] Cremers, F. P. (2002). Hum. Mol Genet 11, 1169-1176. [0257] Dalkara, D. (2009). Mol Ther 17, 2096-2102. [0258] Dalkara, D. (2011). Mol Ther 19, 1602-1608. [0259] Davis, J. A. (2007). J Blot Chem 282, 28807-28814. [0260] den Hollander, A. 1. (2010). J Clin Invest 120, 3042-3053. [0261] den Hollander, A. I. (2004). Hum Mutat 24, 355-369. [0262] den Hollander, A I. (2002). Mech Dev 110, 203-207 [0263] den Hollander, A. I. (2001a). Am J Hum Genet 69, 198-203. [0264] den Hollander, A. I. (2001b). Hum Mol Genet 10, 2767-2773. [0265] den Hollander, A. 1. (2007). Invest Ophthalmol Vis Sci 48, 5690-5698. [0266] den Hollander, A. I. (2008). Prog Retin Eye Res 27, 391419. [0267] den Hollander, A. 1. (1999). Nat Genet 23, 217-221. [0268] Drack, A. V. (2009). JAAPOS 13, 463-465. [0269] Dong, B. (2010). Mol Ther 18, 87-92. [0270] Fischer, M. D. (2009). PLoS One 4, e7507 [0271] Galvin, J. A. (2005). Ophthalmology 112, 349-356 [0272] Gao, G. (2011). Hum Gene Ther 22, 979-984. [0273] Gerber, S. (2002). Ophthalmic Genet 23, 225-235. [0274] Gillespie, F. D. (1966). Am J Ophthalmol 61, 874-880. [0275] Glushakova, L. G. (2006)Mol Vis 12, 298-309. [0276] Gosens, I. (2008). Exp Eye Res 86, 713-726. [0277] Hanein, S. (2004). Hum Mutat 23, 306-317. [0278] Hanein, S. (2006) Adv Exp Med Blot 572, 15-20 [0279] Hauswirth, W. W. (2008). Hum Gene Ther 19, 979-990. [0280] Heckenlively, J. R. (1982). Br J Ophthalmol 66, 26-30. [0281] Heilbronn, R. (2010). Handb Exp Pharmacol, 143-170. [0282] Henderson, R. H. (2011). Br J Ophthalmol 95, 811-817. [0283] Hermens, W. T. (1999). Hum Gene Ther 10, 1885-1891. [0284] Hirsch, M. L. (2010) Mol Ther 18, 6-8 [0285] Hsu, Y. C. (2010). BMC Cell Biol 11, 60. [0286] Izaddoost, S (2002). Nature 416, 178-183 [0287] Jacobson, S. G. (2006a). Mol Ther 13, 1074-1084. [0288] Jacobson, S. G. (2006b). Hum Gene Ther 17, 845-858. [0289] Jacobson, S. G. (2003). Hum Mol Genet 12, 1073-1078 [0290] Jacobson, S. G. (2008). Invest Ophthalmol Vis Sci 49, 4573-4577. [0291] Jacobson, S. G. (2012). Arch Ophthalmol 130, 9-24. [0292] Kantardzhieva, A. (2005). Invest Ophthalmol Vis Sci 46, 2192-2201. [0293] Karali, M. (2011). PLoS One 6, e22166. [0294] Khani, S. C. (2007). Invest Ophthalmol Vis & ci 48, 3954-3961. [0295] Klimczak, R. R. (2009) PLoS One 4, e7467. [0296] Koenekoop, R. K. (2004). Sury Ophthalmol 49, 379-398. [0297] Kolstad, K. D. (2010). Hum Gene Ther 21, 571-578. [0298] Komaromy, A. M. (2010) Hum Mol Genet 19, 2581-2593. [0299] Lai, Y. (2010). Mol Ther 18, 75-79. [0300] Lambert, S. R. (1993). Am J Med Genet 46, 275-277. [0301] Leber, T., (1871) Albrecht von Graefes Arch. Ophthal., 17:314-340. [0302] Leber, T., (1869) Albrecht von Graefes Arch. Ophthal. 15: 1-25. [0303] Lemmers, C (2004). Mol Biol Cell 15, 1324-1333 [0304] Levitt, N. (1989). Genes Dev 3, 1019-1025. [0305] Li, W. (2009). Mol Vis 15, 267-275. [0306] Li, W. (2011a). Virology 417, 327-333. [0307] Li, X. (2011b). Invest Ophthalmol Vis Sci 52, 7-15. [0308] Livak, K J. (2001). Methods 25, 402-408 [0309] Lotery, A. J. (2001a). Arch Ophthalmol 119, 415-420. [0310] Lotery, A. J. (2001b). Ophthalmic Genet 22, 163-169. [0311] Lotery, A. J. (2003). Hum Gene Ther 14, 1663-1671. [0312] Maguire, A. M. (2009). Lancet 374, 1597-1605. [0313] Maguire, A. M (2008). N Engl J Med 358, 2240-2248. [0314] Mancuso, K. (2009) Nature 461, 784-787. [0315] McKay, G. J. (2005). Invest Ophthalmol Vis Sci 46, 322-328. [0316] Mehalow, A. K. (2003) Hum Mol Genet 12, 2179-2189. [0317] Moore, A. T. (1984). Br J Ophthalmol 68, 421-431. [0318] Mowat, F. M. (2012). Gene Ther [0319] Mussolino, C. (2011). Gene Ther 18, 637-645 [0320] Omnori, Y. (2006). Curr Blot 16, 945-957. [0321] Pang, J. (2010) Gene Ther 17, 815-826. [0322] Pang, J. J (2006) Mol Ther 13, 565-572. [0323] Pang, J. J. (2011). Mol Ther 19, 234-242. [0324] Park, B (2011) J Neurosci 31, 17230-17241. [0325] Park, T. K. (2009). Gene Ther 16, 916-926. [0326] Pasadhlka, S. (2010). Invest Ophthalmol Vis Sci 51, 2608-2614. [0327] Pawlyk, B. S. (2010). Hum Gene Ther 21, 993-1004 [0328] Pawlyk, B. S. (2005). Invest Ophthalmol Vis Sci 46, 3039-3045. [0329] Pellissier L. P. (2013) Targeted ablation of CRB1 and CRB2 in retinal progenitor cells mimics Leber congenital amaurosis. PLoS Genet. 2013 December; 9(12):e1003976. [0330] Pellissier L. P. (2014) CRB2 acts as a modifying factor of CRB1-related retinal dystrophies in mice. Hum Mol Genet. July 15; 23(14)-3759-71. [0331] Pellissier L. P. (2014) Specific tools for targeting and expression of Muller glia cells. Mol Ther Methods & Clincal Dev 1:14009 [0332] Pellikka, M. (2002). Nature 416, 143-149. [0333] Perrault, I. (1999). Mol Genet Metab 68, 200-208. [0334] Petersen-Jones, S. M. (2012). Vet Ophthalmol 15 Suppl 2, 29-34. [0335] Petit, L. (2012). Mol Ther 20, 2019-2030. [0336] Provost, N. (2005). Mol Ther 11, 275-283. [0337] Rapti, K. (2012). Mol Ther 20, 73-83 [0338] Richard, M. (2006). Hum Mol Genet 15 Spec No 2, R235-R243. [0339] Roh, M. H. (2002) J Cell Biol 157, 161-172. [0340] Rolling, F. (2006). Bull Mem Acad R Med Belg 161, 497-508. [0341] Salegio, E. A. (2012). Gene Ther. [0342] Schappert-Kimmijser, J. (1959). AMA Arch Ophthalmol 61, 211-218. [0343] Schroeder, R. (1987). Arch Ophthalmol 105, 356-359. [0344] Schuil, J (1998) Neuropedatrics 29, 294-297. [0345] Simonelli, F. (2010). Mol Ther 18, 643-650. [0346] Simonelli, F. (2007). Invest Ophthalmol Vis Sci 48, 4284-4290. [0347] Stieger, K (2008) Mol Ther 16, 916-923. [0348] Stieger, K. (2011). Methods Mol Biol 807, 179-218. [0349] Stieger, K. (2010). Discov Med 10, 425-433. [0350] Stieger, K. (2009). Mol Ther 17, 516-523 [0351] Sun, X. (2010). Gene Ther 17, 117-131. [0352] Sundaram, V (2012) Eur J Pediatr 171, 757-765. [0353] Surace, E. M. (2008). Vision Res 48, 353-359 [0354] Tan, M. H. (2009). Hum Mol Genet 18, 2099-2114. [0355] Tanimoto N. Sothilingam V, Seeliger M W; Functional phenotyping of mouse models with ERG. Methods Mol Biol. 2013; 935.69-78 [0356] Timmers, A. M. (2001). Mot Vis 7, 131-137. [0357] Vallespin, E (2007). Invest Ophthalmol Vis Sci 48, 5653-5661. [0358] van de Pavert, S. A. (2004) J Cell Sci 117, 4169-4177. [0359] van de Pavert, S. A. (2007a). J Neurosci 27, 564-573. [0360] van de Pavert, S. A. (2007b). Glia 55, 1486-1497. [0361] van den Hurk, J. A. (2005). Mot Vis 11, 263-273. [0362] van Rossum, A G (2006) Hum Mol Genet 15, 2659-2672. [0363] van, Soest, S. (1996). Cytogenet Cell Genet 73, 81-85. [0364] van, Soest, S. (1994). Genomics 22, 499-504. [0365] Vazquez-Chona, F. R. (2009) Invest Ophthalmol Vis Sci 50, 3996-4003. [0366] Vogel, J. S. (2007). Invest Ophthalmol Vis Sci 48, 3872-3877. [0367] Waardenburg, P J (1963). Acta Ophthalmol (Copenh) 41, 317-320. [0368] Wagner, R. S. (1985). Arch Ophthalmol 103, 1507-1509. [0369] Walia, S. (2010). Ophthalmology 117, 1190-1198. [0370] Wu, Z. (2010) Mol Ther 18, 80-86. [0371] Yang, G. S. (2002). J Prol 76, 7651-7660. [0372] Yin, L. (2011). Invert Ophthalmol Vis Sci 52, 2775-2783. [0373] Yzer, S (2006) Invest Ophthalmol Vis Sci 47, 3736-3744. [0374] Zemant, J. (2005). Invest Ophthalmol Vis Sci 46, 3052-3059. [0375] Zhong, L. (2008a). Virolog, 381, 194-202. [0376] Zhong, L. (2008b). Proc Natl Acad Sci USA 105, 7827-7832. [0377] Zolotukhin, S. (1999). Gene Ther 6, 973-985. [0378] Zolotukhin, S. (2002). Methods 28, 158-167