The usherin mutation c.2299delG leads to its mislocalization and disrupts interactions with whirlin and VLGR1

Abstract

Usher syndrome (USH) is the leading cause of combined deafness-blindness with type 2 A (USH2A) being the most common form. Knockout models of USH proteins, like the Ush2a^-/- model that develops a late-onset retinal phenotype, failed to mimic the retinal phenotype observed in patients. Since patient's mutations result in the expression of a mutant protein and to determine the mechanism of USH2A, we generated and evaluated an usherin (USH2A) knock-in mouse expressing the common human disease-mutation, c.2299delG. This mouse exhibits retinal degeneration and expresses a truncated, glycosylated protein which is mislocalized to the photoreceptor inner segment. The degeneration is associated with a decline in retinal function, structural abnormalities in connecting cilium and outer segment and mislocaliztion of the usherin interactors very long G-protein receptor 1 and whirlin. The onset of symptoms is significantly earlier compared to Ush2a^-/-, proving expression of mutated protein is required to recapitulate the patients' retinal phenotype.

Fig. 1. Ush2a^delG/delG protein is stable, developmentally regulated, and mislocalized.

A Developmental steady-state levels of total usherin transcripts are comparable in WT and Ush2a^delG/delG retinas. Shown are mean ± SEM from three independent samples. Statistical significance was determined by one-way ANOVA. P values: P15: p = 0.0043 (**). N value: 3 mice per time point and genotype and measurements were made in triplicates. B Immunoblotting shows that the mutant protein and GFP are expressed only in the KI retinas. C Immunoblot shows that the mutant protein is unable to form complexes through electrostatic interactions. D Immunoblot under reducing and non-reducing conditions shows that truncated usherin is unable to form a covalent complex (upper panel). PRPH2 was used as a positive control (lower panel). E Full-length usherin is located at the ciliary base (upper left and middle images) and OPL (lower left image). The truncated protein is absent from the P30 WT retina (right images). F Co-staining of retinal sections from P30 Ush2a^delG/delG for usherin and acTub shows the absence of full-length usherin (left and middle images). Labeling for FLAG reveals the truncated protein to be localized in IS (right images). G Co-labeling for FLAG and acTub at P30 shows a dispersed localization of the mutant protein in the IS, without distinct localization in the periciliary region. H Co-labeling for FLAG and STX3 demonstrates the accumulation of the truncated protein in the IS (left images). 3D-remodeling showing truncated usherin organized in compartments (right images). I–K Immunoblots of fractionated WT and Ush2a^delG/delG retinal extracts showing full-length usherin localized in the membrane fractions in the WT and absence in Ush2a^delG/delG retina (I). Immunoblot probed with anti-FLAG antibody revealed the major portion of the mutated protein is in the cytosolic fraction while a minor portion is present in the membrane fraction (K). GAPDH was used as a cytosolic marker, IRBP as a S-IPM marker, while PRPH2 was used as a membrane marker. L Immunoblot probed with anti-FLAG antibody showing the mutant protein is glycosylated like PRPH2. OS outer segment, IS inner segment, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, S-IPM soluble inter photoreceptor matrix, Cyt. cytoplasm, Memb. membrane, Glyco. glycosylated, Deglyco. deglycosylated, Red. reducing, Non-Red. non-reducing.

Fig. 2. The Ush2a^delG/delG model exhibits a late-onset decline in visual function.

A Optomotor response is significantly reduced at P180 in Ush2a^delG/delG mice and persists at P500. Shown are mean ± SEM. N values: WT P45: 7; P180: 9; P500: 8. Ush2a^delG/delG P45: 6; P180: 13; P500: 12. Statistical significance was determined by a two-tailed unpaired t-test. P values: P180: p = 0.0015 (**), P500: p = 0.0197 (*). B Maximum ERG a-wave and b-wave amplitudes plotted against the log of the eliciting flash intensities. Animals (P200) were flashed from weakest to brightest light intensities. Shown are mean ± SEM. N value: 7 WT mice were tested for each intensity in the measurement of both scotopic a- and b-waves. 5 Ush2a^delG/delG mice were tested for each intensity in the measurement of both scotopic a- and b-waves. C–G Full-field ERGs were performed on WT and Ush2a^delG/delG mice at the indicated ages. C, D Representative scotopic and photopic ERG waveforms are shown from P360 Ush2a^delG/delG and WT mice. E–G Maximum scotopic a-, scotopic b-, and photopic b-wave amplitudes were measured at the indicated ages. Amplitudes are significantly decreased in the P360 and P500 Ush2a^delG/delG mice. Shown are mean ± SEM. Significance was determined by a two-tailed unpaired t-test. P values: Scotopic a-wave: P360: p = 0.0001 (***) P500: p = 0.0006 (***). Scotopic b-wave: P360: p = 0.0001 (***) P500: p < 0.00001 (****). Photopic b-wave: P360: p = 0.0079 (**) P500: p = 0.0292 (*). N values: Scotopic a-wave: WT P30: 18; P180: 11; P270: 9; P360: 17; P500: 5. Ush2a^delG/delG P30: 15; P180: 17; P270: 9; P360: 13; P500:12. Scotopic b-wave: WT P30: 18; P180: 11; P270: 9; P360: 17; P500:5. Ush2a^delG/delG P30: 15; P180: 17; P270: 9; P360: 12; P500:12. Photopic b-wave: WT P30: 18; P180: 10; P270: 9; P360: 17; P500: 5. Ush2a^delG/delG P30: 13; P180: 17; P270: 9; P360: 12; P500:12.

Fig. 3. Expression of mutant usherin leads to late-onset photoreceptor degeneration and mislocalization of rhodopsin.

A Representative retinal light micrographs are shown for P180, P360, and P500 Ush2a^delG/delG and WT mice. B The number of photoreceptors in the ONL was determined by counting the number of nuclei at the indicated distances from the optic nerve and data are presented as means ± SEM. N values: 4 for WT P360, 3 for all other tested genotypes and time points. A significant decrease in the ONL thickness is observed at P360 persisting at P500. Significance was determined by two-way ANOVA with a Bonferroni post hoc test. P values: P360: Superior: −2.1 mm: p = 0.0222 (*), −1.5 mm: p = 0.0012 (**), −0.9 mm: p = 0.0095 (**). Inferior: 0.9 mm: p = 0.0204 (*). P500: Superior: −2.1 mm: p = 0.0420 (*), −1.5 mm: p = 0.00486 (*), −0.9 mm: p = 0.0313 (*), −0.3 mm: p = 0.0420 (*). Inferior: 0.3 mm: p = 0.0078 (**), 0.9 mm: p = 0.0232 (*), 2.1 mm: p = 0.0270 (*). C Retinal sections from WT (P500) and KI animals at P60, P200, and P500 were labeled with an antibody against rhodopsin. The mislocalization of rhodopsin is observed in the Ush2a^delG/delG retinas. Shown are single-plane images with lower images taken at higher magnifications. White arrowheads highlight the weak basic rhodospin signal located in the ONL in P500 WT. White arrows point to mislocalized rhodopsin. OS outer segments, IS inner segments, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, S superior, I inferior.

Fig. 4. The Ush2a^delG/delG retina shows a decline in cone number, mislocalization of cone opsins, and an upregulation of the stress marker GFAP.

A Cone cell density count from four 0.016 mm² areas per retinal whole mount from P500 Ush2a^delG/delG and WT mice (one in each petal) taken at a 300 µm distance from the optic nerve (see the Method Section) and stained with PNA, showing that number of cones is significantly reduced in the Ush2a^delG/delG retina mirroring the functional decline. Shown are means ± SEM from N = 3 animals. P value: p = 0.0221 (*) by a two-tailed unpaired t-test. B Retinal sections from the indicated genotypes at P500 were co-labeled with PNA (green) and a mixture of anti-S- and M-opsins (red) antibodies. Nuclei in all sections were counterstained with DAPI (blue). White arrows show an accumulation of cone opsins in the ONL and IS. Two panels on the right side of each genotype showing 3D construction of the OS/IS and ONL/OPL areas where mislocalization of cone opsins is observed. Shown are representatives of N = 3 animals per genotype. C Total number of cones counted from the entire retinal cross sections shown in B. Also shown are the number of cones with mislocalized cone-opsins either in the IS/ONL or at the synaptic terminals. N values: WT: Presented are average cone counts from two sections per eye from five independent mice. Ush2a^delG/delG: Presented are average cone counts from three sections per eye from six independent mice. Significance was determined by a two-way ANOVA with a Bonferroni post hoc test. P values: total: p < 0.0001 (****), IS/ONL: p < 0.0001 (****), synaptic: p < 0.0001 (****). D Paraffin retinal sections from WT and Ush2a^delG/delG at P60, P200, and P500 were labeled for PNA (green) and immunolabeled either for S- or M-opsin (red) and counterstained with DAPI (blue). On the right side of each panel a 3D reconstruction of 8-12 confocal stacks (1 µm thickness) is shown, highlighting the photoreceptor cell layer. White arrows pointing to S- or M-opsin mislocalized in the IS, ONL, or synaptic terminal. E Retinal sections from WT and KI animals at P30, P180, and P360 were labeled with antibodies against GFAP. At P360, upregulation of GFAP was observed. Shown are single-plane images. OS outer segments, IS inner segments, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, IPL inner plexiform layer, GCL ganglion cell layer.

Fig. 5. Expression of mutant usherin leads to shortening of the photoreceptor cilium.

A Retinal sections labeled with the CC marker Cen2 revealed a decrease in CC length in the Ush2a^delG/delG retina. B Measurement of the ciliary length revealed statistically significant shortening at P30 Ush2a^delG/delG mice (359 cilia were measured from three independent WT and Ush2a^delG/delG mice). The acetylated region of the photoreceptor cilium is also significantly reduced (177 and 175 cilia were measured from three separate WT and Ush2a^delG/delG mice, respectively). Significance was determined by a two-tailed unpaired t-test, p < 0.0001 (****). C Measurements of the ciliary length revealed the shortening to be significant in P300 Ush2a^delG/delG mice (208 and 210 cilia were measured from two separate WT and Ush2a^delG/delG mice, respectively). Significance was determined by a two-tailed unpaired t-test, p < 0.0001 (****). Data presented in B and C are means ± SD. D Representative TEM images of the OS/IS junction (upper images) and higher magnification images of the CC (lower images) from P360 WT and Ush2a^delG/delG mice are shown. E A photoreceptor schematic denoting the areas at which measurements of OS and CC widths were made. F The width of OS and CC was quantified in P360 WT and Ush2a^delG/delG retinas. The width of the OS was significantly reduced in the Ush2a^delG/delG retina, while the width of CC remained unchanged (68 cilia were measured for WT and 90 cilia were measured for the Ush2a^delG/delG). Shown are means ± SD. Significance was determined by a two-tailed unpaired t-test, p < 0.0001 (****) for OS width. G High magnification TEM images are taken from P360 WT and Ush2a^delG/delG mice showing a cross-section of the CC. OS outer segment, CC connecting cilium, IS inner segment, PR periciliary region, OSW outer segment width, CCW connecting cilium width.

Fig. 6. VLGR1 and WHRN are mislocalized in the Ush2a^delG/delG retinas. P30 WT and Ush2a^delG/delG retinas were used.

VLGR1 co-localizes with usherin (A) and WHRN (B) adjacent to the ciliary base in WT retinas. C VLGR1 localizes between the IS and OS and in the OPL (left two images). VLGR1 is localized in the ciliary region (second image from the right) and rightmost image shows a high-magnification image demonstrating VLGR1 localization at the ciliary base. D VLGR1 is mislocalized to the IS (leftmost image) while its localization in OPL is maintained (second image from the left) in Ush2a^delG/delG. Co-staining for acTub shows the absence of VLGR1 from the CC (right images). E WHRN is localized between the IS and OS, and at the OPL in the WT retina (left image). Co-labeling for acTub and WHRN demonstrates the localization of WHRN at the CR (second image from the right). A high-magnification image of a single photoreceptor cilium demonstrates WHRN localizing at the ciliary base (rightmost image). F WHRN is mislocalized to the proximal part of the OS in the Ush2a^delG/delG retina (left panels). Co-staining for acTub and WHRN revealed the absence of WHRN from the CC (second image from the right) as well as the mislocalization to the proximal part of the OS (examples highlighted by arrows). A high magnification image demonstrates the absence of WHRN from the CC, and its mislocalization towards the proximal OS (rightmost image, highlighted by arrow). G VLGR1 and FLAG are independently localized in IS in the Ush2a^delG/delG retina. H WHRN in the Ush2a^delG/delG retina is localized in the OS, while the truncated usherin is localized in IS of Ush2a^delG/delG retina. OS outer segment, CR ciliary region, CC connecting cilium, IS inner segment, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer.

Fig. 7. Proper localization of other USH proteins in the Ush2a^delG/delG retinas.

A–F P30 retinal sections from WT and Ush2a^delG/delG mice were labeled with antibodies against other USH proteins as listed above. The localization of all proteins tested was unaltered in the Ush2a^delG/delG retina when compared to WT. OS outer segment, CR ciliary region, IS inner segment.

Fig. 8. Preservation of periciliary localization of VLGR1 and WHRN in the Ush2a^delG/+ retina.

A, B Labeling of the retinal section from P30 Ush2a^delG/+ mice for usherin (USH2A) and acTub demonstrates its localization at the periciliary base (A, B) and the OPL (A, right image) in Ush2a^delG/+ mice. C, D Labeling for VLGR1 and acTub revealed its localization at the ciliary base (C), where VLGR1 co-localizes with usherin (D) in the Ush2a^delG/+ retina. E, F Staining for WHRN demonstrates its localization at the ciliary base, where it co-localizes with VLGR1 (F) in the Ush2a^delG/+ retina. G, H Staining for FLAG in Ush2a^delG/delG (G) and Ush2a^delG/+ (H) retinas revealed a portion of the truncated usherin (FLAG) to be localized at the photoreceptor cilium in the Ush2a^delG/+ retina but not in Ush2a^delG/delG. I High magnification images of single photoreceptor cilia demonstrating the co-localization of truncated and full-length usherin at the ciliary base in the Ush2a^delG/+ retina. J Immunoblot of extracts from cellular fractionation of Ush2a^delG/+ retinas demonstrated that mutant usherin is found in the cytosol and membrane fractions. K Co-labeling of retinal sections from P30 Ush2a^delG/delG (left images) and Ush2a^delG/+ (right images) for truncated usherin (FLAG) and the ER marker calreticulin. The truncated protein showed a prominent accumulation at the ER in the Ush2a^delG/delG retina, while in the Ush2a^delG/+ retina this accumulation was less pronounced. OS outer segment, CR ciliary region, IS inner segment, ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer.

Fig. 9. Schematic summarizing the role of mutant usherin in impacting the structure of the periciliary membrane complex.

A In the WT photoreceptor cells, the PMC is intact, facilitating the efficient loading of cargo, including rhodopsin, into the CC. Rhodopsin is destined to the OS. B In the Ush2a^delG/delG photoreceptor, the PMC is disrupted. WHRN and VLGR1 are trapped in the inner segment in vesicles, no longer able to localize at the ciliary pocket. WHRN is mislocalized towards the OS and inner segment. The disruption of the PMC decreases the loading efficacy of rhodopsin in the CC and thus impacts its transport toward the OS. Additionally, a huge portion of the mutant usherin is trapped in the ER, causing cellular stress. PMC periciliary membrane complex, OS outer segment, IS inner segment, Rho rhodopsin, ER endoplasmic reticulum, CC connecting cilium. Created with BioRender.com.