|
Fig. 1. Rod outer segment and cone outer segment morphology in wild-type and prom1-null X. laevis. (A) Wild-type (WT) (n=43) ROS have a cylindrical and ordered structure, whereas prom1-null (n=42) ROS are very dysmorphic. Structures of interest are stacked membrane whorls of different sizes (white arrowheads, region i) and overgrown folded OS membranes that are oriented vertically along the outside of the ROS (black arrowhead, region ii). Green, WGA. (B) Incubating retinas in Lucifer Yellow dye verified that the overgrown and dysmorphic ROS discs of prom1-null mutants were not open to the extracellular space. COS are indicated by the white arrowheads and nascent ROS discs that are open to the extracellular space are indicated by black arrowheads. It is normal for there to be open discs at both the base and the tip of the ROS. Yellow, Lucifer Yellow (n=5). (C) Wild-type COS are ordered and cone-like, with no significant inner segment localization of cone opsin besides some small puncta. In prom1-null COS, cone opsin-positive membranes were elongated and fragmented (black arrowhead). Tendrils of cone opsin-positive membrane were observed to be wrapped around the base of adjacent ROS (white arrowheads) or draped over the ROS in sheets (white asterisks). COS membranes also appeared to attach themselves to the ROS for support (v). Significant cone opsin mislocalization to the inner segment generally did not occur, but was present in a small subset of animals (vii, F0, n=6/42 samples). Green, cone opsin; blue, Hoechst; magenta, WGA. Seven replicates. The green channel (cone opsin) was adjusted non-linearly in Cii, iv and vii to show the tendrils of cone opsin-positive membrane and cone opsin inner segment localization with greater intensity. Scale bars: 10 µm.
|
|
Fig. 2. TEMs demonstrating the principal changes in ultrastructure of prom1-null, cdhr1-null, and prom1 plus cdhr1 double-null mutant X. laevis compared to wild-type controls. (A) Wild-type ROS ultrastructure is highly ordered and consists of stacked OS membrane discs with properly formed hairpins (white arrowheads) that are meticulously aligned and enclosed within a plasma membrane (black arrowhead) (n=3). (B) Wild-type COS feature highly ordered stacks of disc lamellae. The leading edges of the lamellae opposite the connecting cilium are neatly aligned, whereas the hairpins near the connecting cilium are not as well aligned (n=3). (C) Prom1-null ROS sometimes lacked hairpins in the basal discs and had a severely convoluted structure where the disc membranes were overgrown and folded in on themselves. There were some instances of disc membrane invagination into the rod inner segment (dashed box) (F0, n=4). (D) Prom1 plus cdhr1 double-null ROS were indistinguishable from prom1-null ROS and featured whorls of overgrown membranes that contained disc rims (white arrowheads) (F0, n=3). (E) Prom1-null COS were difficult to observe, but commonly seen features were loops of disc membrane that appeared unattached to the cone inner segment (CIS; black arrows) and the presence of thin, loosely folded membranes above the CIS (white arrows; F0, n=3). (F) The principal feature of cdhr1-null ROS was overgrown and misaligned discs that were oriented vertically instead of horizontally within the ROS plasma membrane. Areas of overgrowth were commonly associated with large (asterisks) or small (yellow arrowheads) bubbles of membranes around the area where disc orientation changes occurred. Overgrowth of disc membrane was easily seen when the ROS was in a coronal orientation, and these areas of disc overgrowth or disorganization appeared to be contained within the plasma membrane (black arrowhead). Hairpins (white arrowheads) were still present in overgrown disc membranes (F0, n=5). (G) COS ultrastructure appeared mostly normal in cdhr1-null X. laevis, but some COS had overgrown lamellae membranes that resulted in the loss of alignment at the ciliary and non-ciliary sides of the outer segment, giving it a frayed appearance (black arrows) (F3, n=2/5 samples studies). (H) COS in prom1 plus cdhr1 double-null X. laevis were similarly difficult to visualize to prom1-null animals, and they also comprised loose, convoluted and looped thin membranes (white arrows; F0, n=3). From three replicates. Scale bars: 800 nm, except in right-hand panel for C (2 µm).
|
|
Fig. 3. Averaged scotopic single-flash recordings, photopic single-flash recordings and photopic 5 Hz flicker recordings from wild-type and prom1-null F0 animals. (A) Averaged scotopic single-flash recordings from wild-type (n=8) and prom1-null animals (n=7). (B) Averaged photopic single-flash recordings from wild-type (n=7) and prom1-null animals (n=7). (C) Averaged photopic 5 Hz flicker recordings from wild-type (n=7) and prom1-null animals (n=7). Waterfall plots (left) and transformed linear regression curves (right) were used to visualize and compare wild-type and prom1-null ERG waveforms and A- and B-wave amplitudes. Data are plotted as mean±s.e.m. From one replicate. *P<0.05, ***P<0.001, ****P<0.0001 (two-way ANOVA with Sidak's post hoc test).
|
|
Fig. 4. Rod outer segment morphology in wild-type and F3 cdhr1-null X. laevis. (A) Wild-type (n=6) basal ROS have clear incisures and a uniform appearance all around the ROS. (B) A maximum intensity projection (left) and two optical sections (middle, right) demonstrating the structure of an overgrown membrane as seen in a coronal section of cdhr1-null basal ROS (F3, n=9). Features of interest were the overgrown membrane (white arrowheads) and the holes (‘pock-marking’) at the base of the OS looking up from the inner segment (black arrowhead). (C) A side-view of overgrown disc membranes which appeared to comprise a large overgrowth that folds back onto itself (white arrowhead). (D) A long ‘tail’ of overgrown disc membrane that extended from, and then looped under, the basal ROS. Green, WGA; magenta, cone opsin. (E) Lucifer Yellow staining of cdhr1-null ROS in the coronal (left) and sagittal (right) orientations; nascent discs at the basal ROS are normally open to the extracellular space as are discs near the tip of the ROS (black arrowheads), but there was no abnormal Lucifer Yellow dye penetration into the middle of the cdhr1-null ROS (n=7; white arrowhead, overgrown membrane). From two replicates. Green and yellow, Lucifer Yellow; magenta, WGA. Scale bars: 5 µm.
|
|
Fig. 5. Averaged scotopic single-flash recordings, photopic single-flash recordings, and photopic 5 Hz flicker recordings from wild-type and cdhr1-null F3 animals. (A) Averaged scotopic single-flash recordings from wild-type (n=8) and cdhr1-null animals (n=8). (B) Averaged photopic single-flash recordings from wild-type (n=10) and cdhr1-null animals (n=10). (C) Averaged photopic 5 Hz flicker recordings from wild-type (n=10) and cdhr1-null animals (n=10). Waterfall plots (left) and transformed linear regression curves (right) were used to visualize and compare wild-type and cdhr1-null A-wave and B-wave responses. The large positive A-wave values at 75 cd s/m2 (photopic single-flash recordings) were likely an artifact introduced by the large early receptor potential response measured by the electrode used in this experiment. Data analysis utilized a two-way ANOVA with Sidak's post hoc test. Data are plotted as mean±s.e.m. From one replicate.
|
|
Fig. 6. Rod outer segment and cone outer segment morphology in wild-typeand F0 prom1 plus cdhr1 double-null X. laevis. Similar to prom1-null animals, ROS were shortened and bulbous and COS were elongated and fragmented. There were occasional instances of mislocalization of cone opsin to the inner segment and tendrils of cone opsin-positive membrane (white arrowheads), and sheets of cone opsin-positive membrane (asterisk) that were closely associated with the ROS (WT, n=15; F0, n=12). From three replicates. Green, cone opsin; magenta, WGA; blue, Hoechst. Scale bars: 10 µm.
|
|
Fig. 7. Averaged scotopic single-flash recordings, photopic single-flash recordings, and photopic 5 Hz flicker recordings from wild-type and prom1 plus cdhr1 double-null F0 animals. (A) Averaged scotopic single-flash recordings from wild-type (n=4) and prom1 plus cdhr1 double-null animals (n=3). (B) Averaged photopic single-flash recordings from wild-type (n=5) and prom1 plus cdhr1 double-null animals (n=4). (C) Averaged photopic 5 Hz flicker recordings from wild-type (n=5) and prom1 plus cdhr1 double-null animals (n=4). Waterfall plots (left) and transformed linear regression curves (right) were used to visualize and compare wild-type and prom1 plus cdhr1 double-null A-wave and B-wave responses. Data are plotted as mean±s.e.m. From one replicate. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 (two-way ANOVA with Sidak's post hoc test).
|
|
Figure S1. Immunoreactivity for prom1 protein in wildtype and prom1-null X. laevis. (A) In wildtype animals, prom1 protein was localized to the base of the ROS (white arrows), small puncta on the outside of the ROS (black arrowheads), and along one edge of the COS (white arrowheads; n = 9). (B) Prom1 immunoreactivity was lost in prom1-null retinas and the ROS/COS were dysmorphic (F1, n = 14). (C) Western blot for prom1-null F0 animals (n = 10) showed a significant reduction in prom1 protein immunoreactivity in bands of the correct molecular weight for the prom1 protein (~95 kDa) compared to wildtype animals (n = 11); p < 0.0001, unpaired studentâs t-test. One animal showed successful editing by Sanger sequencing but had a smaller reduction in prom1 immunoreactivity than the others (Ç). The upper panel shows prom1 protein and the lower panel is a duplicate blot with α-acetylated tubulin (~49 kDa) used as a loading control. Two replicates. Channels: magenta = WGA, green/white = prom1. Abbreviations: WT = wildtype.
|
|
Figure S2. Immunolabelling of various photoreceptor outer segment proteins in wildtype (14 dpf, top, n = 18) and F0 prom1-null (14 dpf, middle, n = 26; 42-43 dpf, bottom, n = 26) retinas. There was no mislocalization of any protein surveyed into the inner segment of the photoreceptors, with the exception of some rare instances of mislocalization of cone opsin to the inner segment (see Fig. 1). As animals aged, there was an increase in Hoechst-stained and autofluorescent deposits in the outer segment layer (white arrowheads). Seven replicates. Channels: green = protein of interest, blue = Hoechst. Abbreviations: OpS = optical section (centre of stack). Scale bar = 50 μm.
|
|
Figure S3. Immunoreactivity for cdhr1 protein in wildtype and cdhr1-null X. laevis. (A) In wildtype animals, cdhr1 protein was localized to a band at the base of the ROS (white arrowheads) and within the ROS plasma membrane (black arrowheads; n = 10). (B) Cdhr1 immunoreactivity was lost in the ROS in cdhr1-null animals, but significant signal remained in the cone outer and inner segments and in presumptive RPE microvilli (cdhr1-null F3, n = 11). (C) Western blot for cdhr1-null animals (F3, n = 12) showed a complete reduction in cdhr1 protein immunoreactivity in bands of the correct molecular weight for the cdhr1 protein (~99 kDa) compared to wildtype animals (n = 10); p < 0.0001, unpaired student’s t-test. The upper panel shows cdhr1 protein and the lower panel is a duplicate blot with α-acetylated tubulin (~49 kDa) used as a loading control. Two replicates. Channels: magenta = WGA, green/white = cdhr1. Abbreviations: WT = wildtype.
|
|
Figure S4. Immunolabelling of photoreceptor outer segment proteins in wildtype (14 dpf, top; n = 10) and F3 cdhr1-null (14 dpf, middle, n = 11; 42-43 dpf, bottom, n = 7) retinas. There was no mislocalization of any protein surveyed into the inner segment of the photoreceptors. Five replicates. Channels: green = protein of interest, blue = Hoechst. Scale bar = 50 μm. Abbreviations: dpf = days post-fertilization, prph-2 = peripherin-2, WT = wildtype.
|
|
Figure S5. Immunolabelling of various photoreceptor outer segment proteins in wildtype (14 dpf, top; n = 11) and F0 prom1 + cdhr1-null (14 dpf, middle, n = 14; 42-43 dpf, bottom, n = 7) retinas. Similar to prom1-null animals, there were occasional instances of mislocalization of cone opsin to the inner segment (black arrowheads) and an increase in condensed nuclei in the outer segment layer in older animals (white arrowhead). Three replicates. Channels: green = protein of interest, blue = Hoechst. Scale bar = 50 μm. Abbreviations: dpf = days post-fertilization, prph-2 = peripherin-2, WT = wildtype.
|
|
Figure S6. Averaged scotopic single-flash recordings, photopic single-flash recordings, and photopic 5 Hz flicker recordings from prom1-null F1 animals.
(A) Averaged scotopic single-flash recordings from wildtype (n = 9) and prom1-null F1 animals (n = 11). (B) Averaged photopic single-flash recordings from wildtype (n = 7) and prom1-null F1 animals (n = 5). (C) Averaged photopic 5 Hz flicker recordings from wildtype (n = 7) and prom1- null F1 animals (n = 5). Waterfall plots (left) and transformed linear regression curves (right) were used to visualize and compare wildtype and prom1-null F1 A-wave and B-wave responses. Data analysis utilized a Two-Way ANOVA with Sidakâs post hoc test. Data are plotted as mean ± SEM. One replicate. Statistics: p < 0.05 *, p < 0.01 **, p < 0.001 ***, p < 0.0001 ****.
|
