COMPOSITIONS AND METHODS FOR TREATMENT OF DISORDERS RELATED TO CEP290

[0015] FIGS. 1A to 1E illustrate that CEP290 aa 1 to 380, i.e., aa 1 to 380 of SEQ ID NO: 2, mediate membrane association. FIG. 1A is a scale representation of human CEP290 SEQ ID NO: 2. Known human CEP290 mutations are noted by tick marks. The indicated domains and mutations are adapted from reference 30. FIG. 1B are fluorescence microscopy images showing the localization pattern of GFP fusions of full length (FL) and truncated CEP290 constructs in hTERT-RPE1 cells. Cells were stained for ARL13B and with DAPI. Scale bars are 5 μm. Insets show 3× magnified views of areas (in boxes) of colocalization between CEP290 truncations and ARL13B. FIG. 1C show representative membrane flotation assay performed on hTERT-RPE1 cells overexpressing CEP290 aa 1 to 580 of SEQ ID NO: 2. Equal amounts of each of five fractions, beginning from the top (1) and ending at the bottom (8) of the sucrose gradient were analyzed. The sucrose percentage and the expected protein composition of each fraction are indicated. Blots were probed for GFP to detect CEP290 aa 1 to 580 and for the indicated controls. Relevant molecular weight markers are shown in kDa. FIG. 1D shows that ARL13B positive vesicles were immunoprecipitated from postnuclear supernatants of hTERT-RPE1 cells expressing the indicated CEP290 truncations or GFP alone. The input, unbound fraction (UB) and immunoprecipitated fraction (IP) were probed with an anti-GFP antibody to detect our truncations. Relevant molecular weight markers are shown in kDa. FIG. 1E show the peripheral membrane (Peri.) and integral membrane protein (Int.) fractions of hTERTRPE1 cells expressing the indicated CEP290 truncations that were isolated and probed for GFP to detect CEP290 truncations and for the indicated controls. Relevant molecular weight markers are shown in kDa.

[0016] FIGS. 2A to 2F illustrate that CEP290 directly binds membranes in vitro and contains a highly conserved membrane binding amphipathic α-helix. FIG. 2A are fluorescence microscopy images showing the localization pattern of GFP fusions of full length (FL) and truncated CEP290 constructs in hTERTRPE1 cells stained for LAMP2 and with DAPI. Scale bars are 5 μm. Insets show 3× magnified views of areas (in boxes) illustrating lack of colocalization between CEP290 truncations and LAMP2. FIG. 2B shows the results of subcellular fractionation experiments performed on 293T cells expressing CEP290 constructs. Cells were fractionated into cytoplasmic (C), membrane (M), nuclear (N), and cytoskeletal (S) fractions and analyzed by western blotting. Relevant molecular weight markers are shown in kDa. FIG. 2C shows the percent of each truncation present in the membrane fraction. Data are presented as mean±SD, n=3. Asterisks indicate statistical significance over GFP alone. FIG. 2D show the results of liposome co-flotation assays performed on purified CEP290 aa 1 to 580. Data is shown for assays performed with CEP290 aa 1 to 580, both with and without liposomes, and with BSA as a control. Equal amounts of each of five fractions, beginning from the top (1) and ending at the bottom (5) of the sucrose gradient were analyzed. The sucrose percentage and the expected protein composition of each fraction are indicated. Relevant molecular weight markers are shown in kDa. FIG. 2E illustrate helical wheel projection of CEP290 aa 257 to 292. Darkest circles represent negatively charged amino acids; slightly less dark circles represent positively charged amino acids; the lightest grey circles represent polar, uncharged amino acids, and medium grey circles represent nonpolar amino acids. By aligning CEP290's predicted amphipathic helix from Gallus SEQ ID NO: 35, Meleagris SEQ ID NO: 36, Rattus SEQ ID NO: 37, Mus SEQ ID NO: 38, Pongo SEQ ID NO: 39, Macaca SEQ ID NO: 40, Homo sapiens SEQ ID NO: 41, Felis SEQ ID NO: 42, Ailuropoda SEQ ID NO: 43 and Danio SEQ ID NO: 45, one can see there is usually conservation of polarity and charge between the divergent residues.

[0017] FIGS. 3A to 3C show that CEP290 colocalizes with microtubules via region M, and truncation of its N- and C-termini enhances this colocalization. FIG. 3A are fluorescence confocal microscopy images showing the localization pattern of GFP fusions of full length (FL) and truncated CEP290 constructs expressed in hTERT-RPE1 cells or in HEK293T cells. Samples were stained for α-tubulin or acetylated α-tubulin, pericentrin or with DAPI. Scale bars are 10 μm. Insets show 10× magnified views of areas of colocalization (in boxes) between CEP290 truncations and pericentrin. In all cases, truncations partially colocalize with pericentrin. In some cases, truncations colocalize with tubulin. Region M is necessary and sufficient for tubulin colocalization. FIG. 3B are scale representations of the library of CEP290 truncations tested in FIG. 3A. Included is a summary of the extent of colocalization with tubulin, rated on a scale ranging from negative (−) to highly positive. Truncations were also in vitro transcribed and translated and subjected to a microtubule MT sedimentation assay. The recombinantly expressed and purified microtubule binding region of CEP290 was similarly subjected to a MT sedimentation assay. FIG. 3C illustrates that for each truncation, the percent of transfected hTERT-RPE1 cells in which the GFP fused truncation displayed a fibrillar localization pattern was determined. At least 100 transfected cells were counted per experiment. Data are presented as mean±SD, n=3. Quantification of the different localization patterns of each truncation from three independent experiments. Region M of CEP290 is necessary and sufficient for fibrillar localization. The inclusion of the N- and C-termini diminishes the extent to which truncations assume a fibrillar pattern of localization.

[0018] FIGS. 4A to 4F show that CEP290 directly binds to and bundles microtubules via region M, and its N- and C-termini inhibit microtubule binding and bundling. FIG. 4A shows representative confocal fluorescence microscopy images of GFP fusions of full length (FL) and truncated CEP290 constructs expressed in hTERT-RPE1 cells stained for acetylated α-tubulin and with DAPI. Scale bars are 10 μm. Truncations containing CEP290 aa1695 to 1966 of SEQ ID NO: 2 perturb the acetylated tubulin staining pattern in hTert-RPI and 293T cells, and deletion of the N- and C-termini enhance the effect. Confocal images of CEP290 truncations overexpressed in hTERT-RPE1 cells stained with an anti-acetylated tubulin antibody. Truncations containing aa1695 to 1966 of SEQ ID NO: 2 increase the intensity of acetylated tubulin staining and result in more bundles of acetylated MTs compared to untransfected cells. FIG. 4B show percent of transfected 293T cells showing an increase in the intensity, and perturbation in pattern, of acetylated α-tubulin staining compared to untransfected cells, following transfection with CEP290 truncations. Deletion of both the N- and C-termini, but not either alone, significantly increases the intensity of acetylated tubulin staining. 100 transfected cells were counted per experiment. Data are presented as mean±SD, n=5. Means with different letters are significantly different. FIG. 4C are Western blots of microtubule (MT) co-sedimentation assays for in vitro transcribed and translated CEP290 truncation mutants. The resulting supernatants (S) and microtubule pellets (P) were probed for the presence of the GFP-fused truncations and are shown for assays performed both with (+MT) and without (−MT) microtubules. Relevant molecular weight markers are shown in kDa. FIG. 4D shows the percent of each truncation co-sedimenting with microtubules. Quantification of the percent of total truncation present in the MT pellet for three independent experiments is shown. Only minimal pelleting is observed in the −MT controls. AA1695 to 1966 of SEQ ID NO: 2 are necessary for significant MT binding, and deletion of both the N- and C-termini, but not of either alone, significantly increases MT binding. Data are presented as mean±SD, n=3. Means with different letters are significantly different. The N- and C-termini cooperatively inhibit MT binding. FIG. 4E are Coomassie-stained gel of microtubule co-sedimentation assays performed with recombinantly expressed and purified CEP290 truncation M subjected to a MT sedimentation assay with increasing concentrations of microtubules (tubulin). Truncation M directly binds MTs with an apparent K_d in the nanomolar range, comparable to other MT binding proteins. The supernatants (S) and microtubule pellets (P) are shown. Relevant molecular weight markers are shown in kDa. CEP290 aa1695 to 1966 of SEQ ID NO: 2 directly bind MTs with a K_d in the nanomolar range. FIG. 4F is a binding curve of microtubule cosedimentation assays as in FIG. 4E. The fraction of truncation M present in the pellet in the absence of microtubules was subtracted from all data points. Data points are presented as means±SD, n=3. A curve was fit to the data points using the non-linear regression functionality of the GraphPad Prism program.

[0019] FIGS. 5A to 5G show that the overexpression of either the N- or C-terminal regulatory regions of CEP290 ablates the normal inhibition of CEP290. FIG. 5A shows fluorescence microscopy fields of hTERT-RPE1 cells transduced with lentiviral empty vector (EV), or vectors encoding either the N- (aa 1 to 580) or C-terminus (aa 1966 to 2479) of CEP290. Cells were stained for acetylated α-tubulin and pericentrin to detect primary cilia and with DAPI. Arrowheads indicate primary cilia. Arrows indicate cells with multiple axonemes originating from the same focus of pericentrin. Scale bars are 10 μm. FIG. 5B show the percent of lentivirus-treated cells forming primary cilia. Data are presented as mean±SD, n=3. 100 cells were counted per experiment. FIG. 5C show fluorescence microscopy images of single hTERT-RPE1 cells transduced with lentiviral vectors as in FIG. 5A. Cells were stained for acetylated α-tubulin to detect primary cilia and with DAPI. Scale bars are 5 μm. FIG. 5D shows the average primary cilium length for hTERT-RPE1 cells as in FIG. 5C. Data are presented as mean±SD, n=3. A total of at least 150 cilia were measured per condition. FIG. 5E are fluorescence microscopy images of hTERT-RPE1 cells that formed multiple cilia after transduction with lentiviral vector encoding the N-terminus of CEP290. Cells were stained with acetylated α-tubulin, ARL13B, and with DAPI. Scale bars are 5 μm. FIG. 5F are fluorescence microscopy images of hTERT-RPE1 cells that formed multiple cilia after transduction with lentiviral vector encoding the N-terminus of CEP290. Cells were stained with acetylated α-tubulin, pericentrin and with DAPI. Scale bars are 5 μm. FIG. 5G shows the percent of lentivirus treated hTERT-RPE1 cells forming multiple cilia. At least 100 cells were counted per experiment. Data are presented as mean±SD, n=3.

[0020] FIGS. 6A to 6F show that an in-frame deletion in Cep290 in the rd16 mouse ablates Cep290's microtubule binding activity. FIG. 6A is a schematic representation of the microtubule binding region of human CEP290 in relation to the rd16 mouse deletion (17). Truncations of CEP290 representing the part of the microtubule binding region deleted in the rd16 mouse and the part of the microtubule binding region maintained in the rd16 mouse were created as shown. FIG. 6B are confocal fluorescence microscopy images showing the localization pattern of the GFP-tagged “Maintained” and “Deleted” CEP290 truncations. Cells were stained for α-tubulin, pericentrin and with DAPI. Scale bars are 10 μm. FIG. 6C are immunoblots of representative microtubule co-sedimentation assays for in vitro transcribed and translated “Maintained” and “Deleted” CEP290 truncations. The supernatant (S) and microtubule pellet (P) fractions are shown in assays performed both with (+MT) and without (−MT) the addition of microtubules. Relevant molecular weight markers are shown in kDa. FIG. 6D shows the percent of each truncation co-sedimenting with microtubules. Data are presented as mean±SD, n=2. FIG. 6E are representative western blots of microtubule co-sedimentation assays performed with full length WT and rd16 Cep290 from mouse brain homogenate. The supernatant (S) and microtubule pellet (P) fractions are shown in assays performed on samples induced to polymerize microtubules (+MT) and samples treated to prevent microtubule polymerization (−MT). Relevant molecular weight markers are shown in kDa. FIG. 6F shows percent of WT and rd16 CEP290 co-sedimenting with microtubules. Data are presented as mean±SD, n=3. Asterisks indicate statistical significance over −MT samples.

[0021] FIG. 7A to 7E show rd16 mouse fibroblasts are deficient in primary cilium formation. FIG. 7A show fluorescence microscopy fields of WT primary dermal fibroblasts that were stained for acetylated α-tubulin to identify primary cilia, and stained with DAPI. Fibroblasts were grown in media with (Fed) or without (Starved) serum. Arrowheads indicate primary cilia. Scale bars are 10 μm. FIG. 7B show fluorescence microscopy fields of rd16 primary dermal fibroblasts stained for acetylated α-tubulin to identify primary cilia, and stained with DAPI. Fibroblasts were grown in media with (Fed) or without (Starved) serum. Arrowheads indicate primary cilia. Scale bars are 10 μm. FIG. 7C show the percent of WT and rd16 primary dermal fibroblasts that form primary cilia in serum fed and serum starved conditions. Quantification was based on separate experiments on fibroblasts coming from 3 different animals per genotype. At least 100 cells were counted per experiment. Data are presented as mean±SD, n=5. FIG. 7D show high magnification fluorescent microscopy images of representative serum-starved WT and rd16 primary dermal fibroblasts stained for acetylated α-tubulin to identify primary cilia, and stained with DAPI. Scale bars are 5 μm. FIG. 7E are average cilium length of serum starved WT and rd16 primary dermal fibroblasts. Quantification was based on separate experiments on fibroblasts coming from 3 different animals per genotype. At least 50 cilia were measured per experiment, and a total of 400 cilia were measured for both the WT and rd16 fibroblasts. Data are presented as mean±SD, n=5.

[0022] FIG. 8 shows a speculative model for CEP290 activity at the primary cilium in four panels. CEP290 is maintained in a closed, inhibited state by its N- and C-termini and CP110 during the cell cycle; CP110 bound to CEP290 with the N- and C-termini binding each other (panel A). Upon entry into G₀ CP110 is degraded at the mother centriole (panel B), destabilizing the closed conformation. This conformational change in the protein frees and activates CEP290's membrane binding and microtubule binding domains (panel C). In its open conformation, active CEP290 is able to recruit additional interacting partners including MTs and to initiate IFT and ciliogenesis (panel D).

[0023] FIG. 9A is the nucleic acid sequence encoding full length naturally occurring human CEP290 SEQ ID NO: 1.

[0024] FIG. 9B is the amino acid sequence encoding full length naturally occurring human CEP290 SEQ ID NO: 2.

[0025] FIG. 10A is a synthetic nucleic acid sequence for codon optimized human CEP290 SEQ ID NO: 3.

[0026] FIG. 10B is a synthetic amino acid sequence for codon optimized human CEP290 SEQ ID NO: 4.

[0027] FIG. 11A is a synthetic construct which is a minigene for CEP290 SEQ ID NO: 5, which encodes the following CEP290 amino acid fragments, spliced together in a single open reading frame: aa130 to 380, aa700 to 1040, aa1260 to 1605, and aa1695 to 1990.

[0028] FIG. 11B is the synthetic amino acid sequence construct SEQ ID NO: 6 encoded by FIG. 11A.

[0029] FIGS. 12A to 12C demonstrate the generation and testing of CEP290 shRNA constructs. FIG. 12A shows hTERT-RPE1 cells transiently transfected with three different CEP290 shRNA constructs, which were lysed and analyzed by western blotting for CEP290 levels. Blots were reprobed for GAPDH as a loading control. FIG. 12B shows densitometric quantification of CEP290 protein levels in hTERT-RPE1 cells as in FIG. 12A. CEP290 levels were normalized using GAPDH as a loading control and the percent of CEP290 remaining, compared to untransfected hTERT-RPE1 cells, was determined Immunofluorescence microscopy images of fields of hTERT-RPE1 cells, either untransfected or transfected with CEP290 shRNA construct 2, were stained for acetylated tubulin as a marker of the primary cilium (data not shown). FIG. 12C shows the percent of control and sh2 transfected hTERT-RPE1 cells that formed cilia upon serum starvation.

[0030] FIGS. 13A to 13C show the isolation and testing of clonal CEP290 knockdown cell lines. FIG. 13A shows clonal retrovirus-transduced hTERT-RPE1 cell lines expressing CEP290 shRNA 2 that were lysed and analyzed by western blotting for CEP290 levels. Blots were reprobed for GAPDH as a loading control. FIG. 13 B shows densitometric quantification of CEP290 protein levels in these hTERT-RPE1 cells. CEP290 levels were normalized using GAPDH as a loading control and the percent of CEP290 remaining, compared to untransfected hTERT-RPE1 cells, was determined Immunofluorescence microscopy images of fields of control and sh2.8 hTERT-RPE1 cells stained for acetylated tubulin as a marker of the primary cilium were obtained (data not shown). FIG. 13C shows the percent of control and sh2.8 hTERT-RPE1 cells that formed cilia upon serum starvation.

[0031] FIGS. 14A and 14B show the construction and testing of a miniCEP290 construct. FIG. 14A is a schematic representation of the CEP290 protein with identified functional domains, interacting domains, and protein motifs labeled. Grayed areas represent regions of the protein not included in the miniCEP290 construct. FIG. 14B is a schematic representation of miniCEP290 Immunofluorescence images of hTERT-RPE1 cells transiently transfected with GFP-fused miniCEP290 and stained for acetylated tubulin and pericentrin, and with DAPI were obtained, and showed the localization of miniCEP290 to the ciliary transition zone (data not shown).