GC1 Deletion Prevents Light-Dependent Arrestin Translocation in Mouse Cone Photoreceptor Cells

Abstract

purpose. Light-driven translocation of phototransduction regulatory proteins between the inner and outer segments of photoreceptor cells plays a role in the adaptation of these cells to light. The purpose of this study was to examine the effects of the absence of guanylate cyclase 1 (GC1) on light-driven protein translocation in rod and cone cells. Both cell types express GC1, but differ in sensitivity, saturation, and response times to light.

methods. Immunohistochemical techniques employing antibodies specific for cone and rod transducin α (Tα) subunits and arrestins were used to examine light-driven translocation of these proteins in the retinas of wild-type and GC1 knockout (KO) mice.

results. Translocation of cone arrestin from cone outer segments to the inner cell regions was disrupted in the absence of GC1, whereas translocation of arrestin and Tα in rods was not affected. Cone Tα did not translocate in wild-type and GC1 KO mice, but differed in its subcellular distribution in GC1 KO retina, remaining in the cone outer segment in light and in dark.

conclusions. These results suggest that multiple, independent pathways regulate the translocation of phototransduction proteins and that GC1, and presumably cGMP, are of key importance in signaling the translocation of cone arrestin.

Rod and cone photoreceptors subserve vision under dim and bright light conditions, respectively, and are able to respond to a remarkable range of light intensities through a process called light adaptation. ¹ Calcium is perhaps the best-characterized secondary messenger in photoreceptor adaptation. Recently, another mechanism involving the physical transport of signaling molecules to and from the discrete subcellular compartments of photoreceptor cells has been implicated in light adaptation. ² This phenomenon, which was first described in biochemical and immunohistochemical studies of rod transducin and rod arrestin in the retinas of dark- and light-adapted rats ³ and mice, ⁴ has recently been substantiated in vertebrate photoreceptors by using biochemical ² and genetic approaches. ⁵  

The signaling cascade responsible for triggering light-driven translocation of proteins in vertebrate photoreceptors is unknown. Studies of mice lacking proteins involved in the initiation and recovery of the visual phototransduction cascade have provided important clues about the nature of the cascade that drives translocation in rod photoreceptor cells. ⁶ In mice deficient in rhodopsin signaling, transducin fails to move from the outer segments in response to light, and arrestin does not completely translocate to the outer segments in the light. Arrestin translocation occurs normally in mice that lack the transducin α (Tα) subunit, suggesting that the normal visual phototransduction pathway is not necessary to signal its movement. Both transducin and arrestin translocate in the absence of each other, suggesting that translocation of these proteins is not codependent. Phosphorylation events associated with the recovery phase of phototransduction are also not necessary to signal the translocation of these proteins. 

Recent studies have shown that cone arrestin also undergoes light-induced translocation in mouse and bovine retinas ^{7 8} ; however, less is known about the functional significance of protein translocation in these cells. As in rod cells, light-induced translocation of cone arrestin does not appear to require cone photopigment phosphorylation in mammalian retina. ⁹ Although this result suggests that light-induced protein translocation may serve similar functions in rods and cones, the recent observations that the localization of cone transducin in zebrafish retina is not altered by either light or changes in cytoplasmic calcium ¹⁰ suggest that the presence of this phenomenon may, in some cases, be cell and species dependent. 

In this study, we examined light-driven translocation of arrestin and transducin in the rod and cone cells of the GC1 knockout (KO) mouse. ^{11 12} The absence of GC1 in these mice abolishes the ability of the cone cells to respond to light and severely compromises the light response of the rod cells. ¹² The results of our analyses show that this mutation also selectively disrupts light-induced translocation of arrestin in cone cells. In addition, we observed that the subcellular localization of cone Tα is abnormal in these retinas. These data show that translocation processes in mouse rod and cone cells are independent events and suggest that in the absence of GC1, cone cells are unable to reset the light-signaling cascade that triggers arrestin translocation. 

Methods

Mice

Homozygous male GC1 (or GCE) KO mice, ¹² originally obtained from the University of Texas Southwestern Medical Center, Dallas (generously provided by David Garbers), were rederived at the University of Florida on a C57BL6 background, as previously described. ¹¹ All animals were handled in accordance with the University of Florida College of Medicine’s policies and with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Genotyping was performed as previously described. ¹¹  

Light Exposure

Mice were born and maintained in an isolator unit under a 12-hour light–dark cycle. On the day of experimentation, 5-week-old GC1 KO and wild-type (WT) mice were either dark adapted for at least 2 hours or were light adapted by exposing them to diffuse white light (∼2000 lux) for 30 minutes (n = 4, GC1 KO; n = 3, WT). The pupils of the mice that were exposed to light were dilated before exposure with 1% atropine sulfate. Immediately after dark or light adaptation, the eyes were enucleated and submerged in 4% paraformaldehyde (PFA). Enucleation of eyes in all dark-adapted mice was performed under a low-intensity red safe light. 

Immunohistochemistry and Fluorescence Microscopy

Eyes were fixed in 4% PFA overnight at 4°C. The corneas and lenses were then removed, and the eyecups were cryoprotected by immersing them in 30% sucrose (wt/vol) in PBS overnight at 4°C. The eyecups were embedded in optimal cutting temperature (OCT) medium (Tissue-Tek; Sakura Finetec, Torrance, CA) and sectioned (16 μm) for immunohistochemical analyses. Sections designated for immediate analyses were allowed to dry overnight at room temperature. The remaining slides were stored at −20°C until use. 

Retinal sections were rinsed in PBS and incubated in blocking solution (10% goat serum in PBS) for 30 minutes at room temperature. Primary antibodies were diluted in primary dilution buffer (0.3% Triton X-100 and 1% BSA in PBS) and applied to the sections that were then incubated overnight at 4°C. For detection of rod arrestin, a polyclonal antibody raised against a peptide corresponding to amino acid residues 2-18 from rat visual arrestin was used (1:1000; PA1-731; Affinity Bioreagents, Golden, CO). Cone arrestin was detected with the LUMIJ polyclonal antibody, an antibody previously described and characterized by Zhu et al. ⁸ (1:1000; generously provided by Cheryl Craft and Xumei Zhu, Doheny Eye Institute, Keck School of Medicine, Los Angeles, CA). Transducin was detected with polyclonal antibodies specific for the rod Tα subunit (1:1000; anti-Gαt1, sc-389; Santa Cruz Biotechnology, Santa Cruz, CA) and the cone Tα subunit (1:500; anti-Gαt2, sc-390; Santa Cruz Biotechnology), the specificities of which have been examined in our laboratory. ¹¹ After incubation with the primary antibodies, the sections were rinsed in PBS and incubated for 1 hour at room temperature with goat anti-rabbit IgG secondary antibodies tagged with Alexa-488 fluorophore diluted 1:500 in PBS (Molecular Probes, Eugene, OR). The sections were then rinsed a second time with PBS and incubated with peanut agglutinin lectin (PNA) tagged with Alexa-594 fluorophore (25 μg/mL in PBS; Molecular Probes) for 20 minutes After a final rinse step, the sections were counterstained with 4′,6′-diamino-2-phenylindole (DAPI), mounted in an aqueous-based medium (Gel Mount; Biomedia, Foster City, CA), and coverslipped. For immunocytochemical analyses, three to four eyes were processed for each genotype and for each lighting condition. Approximately 40 retinal sections from each eye were examined by routine fluorescence microscopy. The retinal region examined extended approximately 200 μm nasally and 200 μm temporally to the vertical meridian of each eye. The images shown in all the figures are representative of the retinas examined. All images were acquired with a digital imaging system. To facilitate comparisons between WT and GC1 KO immunostained sections, the images were acquired with identical camera settings and exposure times. 

Results

The Absence of GC1 Alters the Subcellular Localization of Tα in Cones

In a study to characterize the retinas of GC1 KO mice, we noted that the cone Tα immunostaining patterns in the retinas of dark-adapted WT and KO mice were significantly different from each other. ¹¹ In WT retinas, Tα protein was strongly localized to the cone outer segments, whereas in KO retinas, Tα immunoreactivity was prominent in the cone inner segments and synaptic regions. This immunolocalization pattern was observed as early as 4 weeks of age in all cone cells examined in the KO retinas and persisted up to 6 months of age (the latest time we examined). These observations prompted us to examine the location of cone transducin in these retinas under light- and dark-adapting conditions. The pattern of cone Tα labeling in dark-adapted WT retina was identical with that observed in light-adapted retina and was strongly localized to the cone outer segments, indicating that the cellular location of cone Tα in these retinas was not altered by our light adaptation paradigm (Fig. 1) . The patterns of cone Tα labeling in both dark- and light-adapted GC1 KO retinas were also similar to each other (Fig. 1) ; however, under both adaptation conditions, the level of cone Tα immunostaining in the inner segments and synapses of the cone cells was significantly higher than that observed in the WT retinas. 

Effect of the Absence of GC1 on Light-Driven Translocation of Arrestin in Cone Cells

In normal mouse retina, light triggers rapid movement of arrestin into the outer segments of cone photoreceptor cells. ^{7 8} In this experiment, we tested the hypothesis that knockout of GC1, which compromises rod cell responses and eliminates cone cell responses to light, ¹² would disrupt translocation of arrestin in cone cells. We compared arrestin immunostaining patterns in dark- and light-adapted WT and GC1 KO retinas. Comparisons of the cone arrestin antibody staining patterns of dark- and light-adapted WT and GC1 KO retinas revealed that cone arrestin in both light- and dark-adapted GC1 KO retinas was localized to the cone outer segments and the synaptic terminals (Fig. 2) . This staining pattern was identical in cone cells located in both the central and peripheral regions of superior and inferior retina and was significantly different from that observed in WT retina. In dark-adapted WT retina, cone arrestin was present in the outer and inner segments, and in the synaptic regions of the cone cells. Exposure of WT mice to light produced a shift of cone arrestin from the inner cellular compartments to the outer segments, consistent with previous analyses of light-induced translocation of cone arrestin in mouse retina. ⁹ We did not observe any differences in staining patterns between superior and inferior retinal regions. 

Effect of the Absence of GC1 on Light-Driven Translocation of Arrestin and Tα in Rod Cells

Light stimulation of WT rod photoreceptors induces rapid translocation of both Tα and arrestin between the compartments of these cells. In rods, Tα moves from the outer segments into the inner segments, outer nuclear layer, and outer plexiform layer, ² and arrestin moves from the inner regions of the rod cells to the outer segments. ^{4 6} In this experiment, we examined the effects of GC1 knockout on the subcellular distribution of arrestin and Tα in dark- and light-adapted rod cells. Comparisons of rod Tα (Fig. 3A)and rod arrestin (Fig. 3B)immunostaining patterns in dark- and light-adapted WT and GC1 KO retinas revealed that the staining patterns of these two transduction proteins were identical in WT and GC1 KO mice under both lighting conditions. 

Discussion

In our recent morphologic characterization of the retinas of GC1 KO mice, we noted that cone Tα immunoreactivity in dark-adapted GC1 KO retinas was shifted to the inner segments and synaptic regions of the cone cells. ¹¹ This observation, together with our previous characterization of the GUCY1*B retina that suggests that the absence of GC1 is likely to induce the biochemical equivalent of light exposure in photoreceptors, ¹³ prompted the current investigation of light-driven translocation of transducin and arrestin in the GC1 KO mouse retina. The process of protein translocation in rod cells has been functionally linked to light adaptation in these cells ² and may serve to protect photoreceptors exposed to extreme lighting conditions. The results of our investigation show that (1) the subcellular localization of Tα in the cone cells of GC1 KO retinas is abnormal regardless of adaptation state, (2) that light-driven translocation of cone arrestin is disrupted in GC1 KO retinas, and (3) that translocation of both Tα and arrestin in rod cells is normal in these retinas. 

Recent immunohistochemical analyses of the subcellular location of cone Tα in dark- and light-adapted zebrafish retinas have shown that, in this species, Tα is highly concentrated in the cone outer segments and, unlike rod Tα, does not move to the inner portions of these cells in response to light stimulation. ¹⁰ The results of our analyses of WT mice retinas under dark- and light-adapting conditions are consistent with these observations and suggest that the absence of Tα translocation in cone cells may be a general characteristic of these cells. In GC1 KO retinas, we observed that the levels of cone Tα are persistently elevated in the inner regions of the cone cells, regardless of adaptation state. This abnormal distribution pattern resembles that which would be predicted in light-adapted retinas if, as occurs with rod Tα, cone Tα moved to the inner regions of the cone cells in response to light. Overexpression of cone Tα in GC1 KO cone cells could also alter the localization pattern of this protein in GC1 KO cone cells; however, Western blot analyses of Tα performed in a previous study of the GC1 KO mouse retina ¹¹ showed that the levels of Tα in GC1 KO are not increased over WT. This observation, together with the staining patterns observed in this study, suggest that Tα may be actively redistributed in GC1 KO cone cells. The abnormal distribution of cone Tα in GC1 KO retinas could reflect the abnormal physiological state of the cones in these retinas. In both GC1 KO mouse and GUCY1*B chicken retinas, which also carry a GC1-null mutation, cone cells do not transduce light into visual signals. ^{12 13} Biochemical analyses of GUCY1*B retinas show that the absence of GC1 in these retinas is associated with significant reductions in the levels of GCAP1 protein ¹⁴ and cGMP ¹³ in the photoreceptor cells. The abnormally low cGMP levels in these cells have been hypothesized to induce chronic hyperpolarization. ¹³ Physiological ¹² and biochemical ¹¹ analyses of GC1 KO retinas suggest that the cone cells in these retinas may also be chronically hyperpolarized. If the cone cells in GC1 KO retinas are in a persistent state of light adaptation, then the increased levels of cone Tα in the inner portions of these cells, although abnormal, could represent a protective mechanism. Increased sequestration of cone Tα in the inner compartments of these cells would reduce the amount of cone Tα in the outer segments, thereby reducing the amount of Tα available to interact with photoactivated opsin proteins under the most extreme lighting conditions. 

Our data show that translocation of arrestin in rod cells was normal in GC1 KO retinas, whereas translocation of arrestin in cone cells was disrupted. This observation, together with the observation that rod cells in GC1 KO retinas retain their ability to respond to light whereas cone cells do not, ¹² suggests that light transduction and protein translocation are interdependent processes in photoreceptor cells. The biochemical intersection of these processes is not clear. In mouse rod cells, translocation of arrestin requires rhodopsin activation, but activation of transducin, the primary effector G protein of the visual phototransduction cascade, does not appear to be necessary in the translocation-signaling pathway. ⁶ In addition, the translocation processes in both rod ⁶ and cone ^{9 15} cells do not appear to be dependent on phosphorylation of activated opsins. Over the course of a normal phototransduction event, the levels of intracellular cGMP and Ca²⁺ decrease in the photoreceptor outer segments and return to prestimulus levels through Ca²⁺-sensitive feedback loops. ¹⁶ The inability of the cone cells in the GC1 KO retina to produce membrane potential changes in response to light is likely to lead to relatively static levels of intracellular Ca²⁺ in the outer segments. Our data do not directly implicate GC1 in the transport signaling cascade, but they do suggest that the dynamic changes in either membrane potential or intracellular ion concentrations that normally occur after light stimulation are necessary for transport signaling. 

Unlike cone cells in the GC1 KO mouse retina, rod cells in this retina retain their ability to respond dynamically to light. ¹² This observation and the results of previous studies in the GC1 KO mouse ¹¹ indicate that there is another GC present in mouse rod cells that is capable of supporting cell function. A likely candidate for this enzyme is GC2, an isoform of GC1 that is expressed in photoreceptor cells. ^{17 18 19} In contrast, mouse cone cells do not appear to possess redundancy in the phototransduction cascade on a GC1-null background. It is noteworthy that both cone and rod function are severely compromised in GUCY1*B chicken retina ²⁰ and in patients with Leber congenital amaurosis type-1, ²¹ a blinding retinal disease that has been linked to null mutations in the GC1 gene. ²² The relatively greater impact that the loss of GC1 has on rod cells in chicken and human retina could reflect species differences or could represent a more general feature of retinas that have evolved for vision under photopic conditions. Further studies are needed to examine protein transport processes in rod and cone cells and the roles they have in facilitating the physiological function and survival of these cells and to identify the biochemical cascades that initiate light-driven transport in these cells. 

²  Present affiliation: The Pioneer Center for Learning and Memory, Howard Hughes Medical Institute, Massachusetts Institute of Technology, Cambridge, Massachusetts.

 Supported by National Eye Institute Grant EY11388 and a Research Grant from the University of Florida McKnight Brain Institute.

 Submitted for publication June 14, 2004; revised July 27 and September 20, 2004; accepted September 26, 2004.

 Disclosure: J.E. Coleman, None; S.L. Semple-Rowland, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Subcellular distribution of the Tα subunit in cone cells in dark- and light-adapted WT and GC1 KO mouse retinas. Shown are cone Tα (cTα) immunostaining patterns (green channel) observed in sections taken from dark- and light-adapted WT and GC1 KO retinas. Sections were colabeled with PNA lectin (red channel) and DAPI (blue channel) to highlight the position of the cone inner segments (IS) and the outer plexiform layer (OPL). Insets: higher magnifications of dark-adapted WT and GC1 KO cone cells. Arrows: IS region of the cone cells. Scale bars: (low magnification images) 25 μm; (insets) 12 μm. OS, outer segments; ONL, outer nuclear layer.

Figure 2.

 

Comparisons of the distribution of cone arrestin in dark- and light-adapted WT and GC1 KO mouse retinas. (A) Staining patterns for PNA (red channel), cone arrestin (CAR; green channel) and PNA+CAR (merged image) are shown. Sections were colabeled with PNA lectin to highlight the position of the cone inner segments. (B) Close-ups of the boxed regions of WT and KO retinas (CAR only) shown in (A). Dotted lines: borders between the various compartments of the cone cells. Please note that all sections shown are on axis and are at the level of the optic nerve head; however, in some cases, during tissue processing, the RPE separated from the ONL and, in doing so, pulled some of the fragile cone OS away from the cone cell body, giving the impression that the cone sheaths were lengthened (KO dark panel and WT light panel). OS, outer segments; IS, inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer. Scale bar, 25 μm.

Figure 3.

 

Comparisons of the distribution of rod Tα (rTα) subunit and rod arrestin in dark- and light-adapted WT and GC1 KO mouse retinas. (A) rTα immunostaining in dark- and light-adapted WT and GC1 KO retinas. (B) Rod arrestin immunostaining in dark- and light-adapted WT and GC1 KO retinas. Abbreviations are as in Figure 2 . Scale bars, 25 μM.

The authors thank Kathy Laughlin for excellent technical assistance and Clay Smith and Wolfgang Baehr for helpful discussions and comments on the manuscript.