|Home | About | Journals | Submit | Contact Us | Français|
Vertebrate photoreceptors have a modified cilium composed of a basal body, axoneme and outer segment. The outer segment includes stacked membrane discs, containing opsin and the signal transduction apparatus mediating phototransduction. In photoreceptors, two distinct classes of vesicles are trafficked. Synaptic vesicles are transported down the axon to the synapse, while opsin-containing vesicles are transported to the outer segment. The continuous replacement of the outer segments imposes a significant biosynthetic and trafficking burden on the photoreceptors. Here, we show that Ahi1, a gene that when mutated results in the neurodevelopmental disorder, Joubert syndrome (JBTS), is required for photoreceptor sensory cilia formation and the development of photoreceptor outer segments. In mice with a targeted deletion of Ahi1, photoreceptors undergo early degeneration. While synaptic proteins are correctly trafficked, photoreceptor outer segment proteins fail to be transported appropriately or are significantly reduced in their expression levels (i.e., transducin and Rom1) in Ahi1−/− mice. We show that vesicular targeting defects in Ahi1−/− mice are cilium-specific, and our evidence suggests that the defects are caused by a decrease in expression of the small GTPase Rab8a, a protein required for accurate polarized vesicular trafficking. Thus, our results suggest that Ahi1 plays a role in stabilizing the outer segment proteins, transducin and Rom1, and that Ahi1 is an important component of Rab8a-mediated vesicular trafficking in photoreceptors. The retinal degeneration observed in Ahi1−/− mice recapitulates aspects of the retinal phenotype observed in patients with JBTS, and suggests the importance of Ahi1 in photoreceptor function.
The primary cilium is a microtubule-based organelle found on the surface of most quiescent cells (Pan and Snell, 2007; Pugacheva et al., 2007). During ciliogenesis, the mother centriole is delivered to the apical cell surface via attachment to a Golgi-derived centriolar vesicle (Sorokin, 1962). Accessory structures are recruited that ultimately form the basal body; the immature axoneme projects from the mother centriole; and vesicular fusion events produce the ensheathment of the elongating axonemal shaft. The centriolar vesicle then fuses with the plasma membrane (Gilula and Satir, 1972; Satir and Christensen, 2007). Subsequent elongation and maintenance of the cilium are regulated by intraflagellar transport (Rosenbaum and Witman, 2002; Scholey and Anderson, 2006; Scholey, 2008).
The rod and cone photoreceptor cells of the retina are examples of cells that utilize specialized sensory cilia for sensation, in this case, light (Adams et al., 2007; Liu et al., 2007). Failure of cilium formation and/or function has recently been implicated in many diseases; such diseases are referred to as ciliopathies (Badano et al., 2006). An important example of ciliopathies is inherited retinal degeneration, in which the outer segments of the rod and cone photoreceptor cells are affected. The connection here is that the light sensitive outer segments are specialized sensory cilia. Like other cilia, these photoreceptor sensory cilia contain an axoneme, which begins at the basal bodies, passes through a transition zone (also called the “connecting cilium”) and into the outer segment (Beisson and Wright, 2003; Liu et al., 2007).
Joubert syndrome (JBTS) is considered to be one such ciliopathy (Badano et al., 2006; Hildebrandt and Zhou, 2007). JBTS is an autosomal recessive disorder classically characterized by brainstem malformations (i.e., the “molar tooth” sign) and cerebellar vermis aplasia/hypoplasia, but also retinal dystrophy (Parisi et al., 2007; den Hollander et al., 2008; Zaki et al., 2008). Multiple genes are known to cause JBTS, with 10–15% of all cases of JBTS attributed to mutations in the Abelson-helper integration site-1 (AHI1) gene (Dixon-Salazar et al., 2004; Ferland et al., 2004; Parisi et al., 2006). AHI1 encodes a cytoplasmic multi-domain protein that is thought to serve as a scaffolding protein (Jiang et al., 2002). Only recently has the function of Ahi1 begun to be elucidated in a recent report that demonstrated the requirement of Ahi1 for the formation of primary non-motile cilia (Hsiao et al., 2009).
Here, we show that in Ahi1−/− mice on day E18.5, basal bodies are found to be normally distributed in the outer retina, and immature axonemes are beginning to form. Post-natal Ahi1−/− mice fail to develop complete photoreceptor sensory cilia, and they exhibit degeneration of photoreceptor cells. Proteins that are trafficked through the transition zone to the outer segments are mis-targeted in Ahi1−/− mice, with many of the aberrantly targeted proteins being seen instead in the photoreceptor cell bodies. Our data demonstrate that the absence of Ahi1 results in a dramatic failure of photoreceptor sensory cilia formation, and emphasizes that inherited retinal degenerative disorders are a subclass of the ciliopathies.
Generation of Ahi1 knockout mice on a C57BL/6J genetic background (B6-Ahi1+/−) has been previously described (Hsiao et al., 2009). However, since Ahi1−/− mice on a C57BL/6J (B6) background die at birth, we performed backcrossing of our B6-Ahi1+/− mice onto multiple, independent inbred lines; the result was that one inbred line, BALB/cJ, was found to be capable of overcoming the post-natal lethality phenotype. The BALB/cJ-Ahi1+/− animals were then repeatedly backcrossed to BALB/cJ mice. To obtain tissue for analysis, we examined animals from Ahi1+/− X Ahi1+/− crosses, with wildtype littermates serving as controls. Mice were maintained on a normal 12 hour light-dark cycle (06:00 to 18:00) with unlimited access to food and water. All mouse procedures were performed under approval from the Institutional Animal Care and Use Committees of the Albany Medical College, Rensselaer Polytechnic Institute, and the Wadsworth Center (NY State Department of Health), in accordance with The National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Ahi1−/−, Ahi1+/−, and Ahi1+/+ mice at various ages were overdosed with avertin or sodium pentobarbital, and then perfused transcardially with phosphate-buffered saline (PBS; pH 7.4) at room temperature, followed by cold 4% paraformaldehyde (PFA) made in PBS. Eyes were removed and post-fixed in 4% PFA, followed by cryoprotection in a 30% sucrose solution made in PBS. Enucleated mouse eyes were sectioned at a thickness of 10–20 μm on a Microm cryostat (Richard-Allan Scientific, Kalamazoo, MI), and were then mounted on positively charged glass slides (Superfrost Plus, Fisher, Pittsburgh, PA). Sections were allowed to air dry for 30 min before being stored at −80°C.
To obtain embryonic day 18.5 (E18.5) pups, we overdosed pregnant females with avertin or sodium pentobarbital, and removed the embryos and placed them into ice-cold 0.1M PBS. Embryos were perfused under a Zeiss dissecting microscope (Carl Zeiss, Stuttgart, Germany) and processed as described above.
TUNEL staining, using the In Situ cell death detection kit (Roche Diagnostics, Germany), was performed on retinas from Ahi1−/−, Ahi1+/−, and Ahi1+/+ mice at various ages, to establish when the retinal degeneration was occurring in the photoreceptor layer. Briefly, 4% PFA was applied to slides containing retinal sections that had been previously perfused as described above. The slides were then washed for 30 min in PBS, followed by incubation in 0.1% Triton X-100 and 0.1% sodium citrate, for 2 min on ice. The slides were rinsed briefly in PBS before being incubated in the TUNEL reaction mixture (prepared according to the manufacturer’s instructions) at 37°C for 1 h in the dark. Slides were rinsed in PBS, and coverslipped with Fluoromount-G; they were analyzed on a Zeiss AxioImager-Z1 microscope and imaged with an AxioCam MRm camera (Carl Zeiss).
Ahi1−/−, Ahi1+/−, and Ahi1+/+ mice at various ages were overdosed with avertin or sodium pentobarbital and perfused as described above. The only method differences were: (i) for immunostaining, mice were perfused with either cold 1% or 4% PFA; and (ii) the eyes were enucleated and, after removal of the cornea and lens, were fixed for an additional 2 to 3 h at 4°C (Liu et al., 2004). The eyecups were then transferred to a 30% sucrose solution in PBS, incubated overnight at 4°C, and embedded in OCT freezing medium, for cryosectioning. Eyes were sectioned at a thickness of 10–20 μm. Sections were blocked in 10% fetal bovine serum (FBS) in PBS with 0.04% Triton X-100 (PBS-TX) for 1 h, before application of primary antibodies. Primary antibodies were applied overnight at 4°C in 1% FBS. Secondary antibodies were applied for 2 h at room temperature in 1% FBS. All antibody solutions were prepared in PBS-TX. Vectashield mounting medium with DAPI (Vector Labs, Burlingame, CA) was applied after staining. Slides were visualized using either (i) a Zeiss AxioImager-Z1 microscope and a Zeiss Apotome, with imaging by an AxioCam MRm camera and AxioVision software release 4.5 (Carl Zeiss); or (ii) a Zeiss LSM 510 Meta confocal microscope, with image processing using the Zeiss Meta 510 software (Carl Zeiss). All images were processed in Adobe Photoshop CS2 (v 9.0.2; Adobe Systems, San Jose, CA). Contrast and brightness of images were adjusted through linear level adjustments, as needed, to optimize the intensity range of the images.
Retinas were dissected from post-natal day 12 (PN12) eyes (Ahi1−/− (n=3) and Ahi1+/+ (n=4)) and homogenized in RIPA buffer [50 mM Tris (pH 8), 150 mM NaCl, 1% NP-40, 0.5% Na-deoxycholate, 0.1% SDS, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride and a 1X protease inhibitor cocktail (Roche Applied Science, Indianapolis, IN)]. Following incubation on ice for 30 min, the homogenized retina was centrifuged at 10,000 x g for 30 min at 4°C. Protein concentrations of the supernatant for each retina sample were determined with the Advanced Protein Assay Reagent kit (Cytoskeleton Inc, Denver, CO). The protein lysates (5 μg) were resolved on an SDS/PAGE gel and transferred onto PVDF membrane (Millipore). The membrane was blocked in 5% skim milk/TBSTx [100 mM Tris (pH 7.4), 150 mM NaCl, and 0.01% Triton-X100] for 1 h at room temperature. The membrane was then incubated with primary antibody (loading controls consisted of using chicken anti-βIII tubulin antibodies (1:1000; Millipore)) diluted in blocking solution at 4°C overnight. Primary antibodies (transducin (1:5000) and Rom1 (1:1000)) were detected with either the SuperSignal West Femto Maximum Sensitivity Substrate Chemiluminescence kit (Pierce, Rockford, IL) or with fluorescent secondary antibodies (Alexa Fluor 488; Invitrogen (Molecular Probes), Carlsbad, CA). Signals were analyzed with a Syngene G:Box iChemi XT imaging system and the GeneTools analysis software (Synoptics, Frederick, MD). In order to quantify and compare the signal intensities of each sample, the detected signals were unsaturated and in the linear range of detection.
The following primary antibodies were used for immunostaining, at the indicated final dilutions: Ahi1 (rabbit IgG, 1:1000; (Doering et al., 2008)); rhodopsin (mouse IgG, clone 4D2, 1:1000; kindly provided by Dr. RS Molday and mouse IgG, clone 1D4, 1:1000, Sigma, St. Louis, MO (Molday and MacKenzie, 1983)); synaptotagmin (mouse IgG2a, 1:100; Calbiochem, San Diego, CA (Matthew et al., 1981)); anti-γ-tubulin (mouse IgG1, 1:300; Sigma (Hsiao et al., 2009)); Rab8a (mouse IgG2b, 1:50; BD Biosciences (Hsiao et al., 2009)); α-transducin (rabbit IgG, 1:1000; Santa Cruz Biotechnology, Santa Cruz, CA (Chen et al., 2007)); Rom1 (mouse IgG, 1D5, 1:40; kindly provided by Dr. RS Molday (Bascom et al., 1992)); Rp1 (chicken IgY, 1:2000, (Liu et al., 2002)); Rpgrip1 (rabbit IgG, 1:5000; kindly provided by Dr. Tiansen Li (Hong et al., 2001)); and anti-cyclic nucleotide gated channel alpha 1 (mouse IgG, 1D1, 1:10; kindly provided by Dr. RS Molday (Cook et al., 1989)). Detection of the primary antibodies was determined via the appropriate fluorophore-labeled secondary antibodies (Alexa Fluor (all from Invitrogen (Molecular Probes)), or Jackson ImmunoResearch, West Grove, PA).
Retinas from Ahi1−/− mice and Ahi1+/+ littermate control mice at PN 12 were prepared for light and electron microscopy as described previously (Liu et al., 2003; Liu et al., 2009). Briefly, animals were sacrificed and perfused with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M PBS. Eyes were enucleated, and the corneas were removed. Eyecups were then fixed in the same fixative for an additional 4 h at 4°C. Retinas were trimmed into 2-mm pieces, and the pieces were postfixed in 1% OsO4, dehydrated through a graded ethanol series followed by propylene oxide infiltration, and embedded in Epon (EMbed812; Electron Microscopy Sciences, Hatfield, PA). Semi-thin (0.5 μm) sections were cut and stained with alkaline toluidine blue for light microscopy. Ultrathin sections (60 nm) were cut and stained with 2% uranyl acetate and lead citrate, and imaged in an FEI Tecnai transmission electron microscope.
We carried out localization studies of Ahi1 in the retina of the post-natal mouse. In the wildtype (Ahi1+/+) mouse, Ahi1 immunolabeling was observed in the outer nuclear layer photoreceptors (Fig. 1A). More specifically, Ahi1 protein was found in the transition zone of the photoreceptor sensory cilium, co-localizing with the transition zone marker, Rpgrip1 (Fig. 1C; top row). To determine whether Ahi1 localizes to the axoneme, we compared the immunostaining patterns of Ahi1 and the axonemal marker Rp1. Ahi1 immunostaining was observed proximal to Rp1 staining, indicating that Ahi1 is present at the proximal end of the axoneme (Fig. 1C; bottom row). Ahi1 immunolabeling was also observed in the inner retina of the wildtype mouse (Fig. 1A). More specifically, Ahi1 immunostaining was found to occur at the edge of horizontal, bipolar and ganglion cell bodies. These immunostaining patterns were not observed in mice with a targeted deletion of Ahi1 (Ahi1−/−)(Fig. 1B). Moreover, no Ahi1 protein was detected in dissected Ahi1−/− retina by Western blotting (Supplementary Fig. 1).
Although individuals with JBTS have been reported to have retinal dystrophy (Parisi et al., 2006), little is known about the pathogenesis of this process in JBTS. Histological analysis of retinas from Ahi1−/− mice revealed complete absence of the photoreceptor outer segments in post-natal day 12 (PN 12) animals (Fig. 2A; compare layer at black arrow in Ahi1+/+ retina with the lack of this layer (at the black arrow) in Ahi1−/− retina). By PN 24, significant loss of photoreceptor cells was observed in the outer nuclear layer (ONL)(Fig. 2A). The photoreceptor nuclear layer showed progressive decline in thickness, and was absent in mice at PN 200 (Fig. 2A). At PN 22, apoptosis was apparent in the photoreceptor cells in retinas from Ahi1−/− mice, but not in retinas from Ahi1+/+ littermates (Fig. 2B).
Studies have demonstrated that rod photoreceptor cells are born late in embryonic development in rodents, and further, that rhodopsin expression can be detected in late embryonic development in a small number of rod photoreceptors (Morrow et al., 1998; Rutherford et al., 2004). Therefore, to differentiate whether the absence of the photoreceptor layer in Ahi1−/− mice was due to an error in retinal developmental, or whether it was due to a degenerative process, we examined the retinas of mice at earlier time points. Examination of immature photoreceptor cells in E18.5 pups, using rhodopsin antibodies, showed rhodopsin immunostaining in some cells both in Ahi1+/+ and in Ahi1−/− mice, consistent with the idea that early photoreceptor cell development is not disrupted in the Ahi1−/− mice (Fig. 3A). Similarly, no major defects were observed in the developing Ahi1−/− retinas at PN 5 (Fig. 2A), suggesting that the retinal degeneration occurs after cell differentiation. In addition, the appearance of rhodopsin containing immature outer segments suggests that, at least initially, immature outer segments could form in the Ahi1−/− retina (Fig. 3A; white arrows).
To determine whether primary cilium formation is initiated in the absence of Ahi1, we examined the retinas of Ahi1−/− and Ahi1+/+ pups at day E18.5. Basal bodies (visualized via antibodies to the basal body marker, γ-tubulin) were found to be properly targeted in the apical region of the photoreceptor neuroblasts, in Ahi1−/− retinas (Fig. 3B). Since γ-tubulin was distributed at the outer edge of the outer nuclear layer in both Ahi1−/− and Ahi1+/+ pups, we concluded that basal body mis-positioning is not the cause of the impaired development of the photoreceptor outer segments.
We examined further whether photoreceptor sensory cilia were present in the PN 12 retina. While outer segments were discernible in light micrographs of epoxy-embedded retinal sections from Ahi1+/+ mice, no outer segments were observed in sectioned Ahi1−/−retinas (Fig. 4A,B). These results were confirmed at the ultrastructural level. Outer segments in the Ahi1+/+ retina appeared, at PN 12, as straight, parallel cylinders projecting from inner segments toward the retinal pigment epithelium (Fig. 4C). In contrast, outer segments were completely absent in PN 12 Ahi1−/− mice (Fig. 4D). No disc structures can be seen in the space between the inner segments and the retinal pigment epithelium (Fig. 4D). We further examined the structure of the photoreceptor sensory cilium components in Ahi1+/+ and Ahi1−/− retinas. The stacks of nascent discs were organized and were oriented perpendicular to the long axis of the axoneme that extended from the basal body and transition zone in Ahi1+/+ mice (Fig. 4E). In the Ahi1−/− mice, the basal body and transition zone showed structures comparable to that in the Ahi1+/+ retina (Fig. 4F). Strikingly, the axoneme was normal in appearance and extended into the space between the inner segment and the retinal pigment epithelium (Fig. 4F). However, no disc membranes were seen along the axoneme (Fig. 4F). Instead, the axoneme was surrounded, at its distal end, by disorganized membranous material (Fig. 4F).
Ahi1+/+ mice exhibited a high level of rhodopsin expression that was localized specifically to the outer segments by PN 11 (Fig. 5A,B). The outer segments were seen to be more pronounced at the PN 22 and PN 42 time points (Fig. 5A). While Ahi1−/−retinas also exhibited strong rhodopsin expression, rhodopsin was not present in the outermost layers of the outer nuclear layer (outer segments). Instead, it was aberrantly distributed to the inner side of the outer nuclear layer (cell bodies), as well as in the inner segments of the photoreceptors (Fig. 5A,B). Photoreceptor degeneration, as indicated by a thinning of the outer nuclear layer, was observed in retinas from Ahi1−/−mice by PN 22, and was nearly complete by PN 42; at the latter time point, only a few rhodopsin-positive cells remained, and the outer nuclear layer was only approximately one cell thick (Fig. 5A (see high magnification image)).
To determine the fate of other outer segment proteins in Ahi1−/− retinas, we carried out localizations of several outer segment proteins. In Ahi1+/+ retinas, the outer segment proteins Cnga1, Rom1 and transducin were found in their normal locations in the outer segments (Fig. 6C and Fig. 7A,C). However, as seen with rhodopsin localization in Ahi1−/− retinas, Cnga1 and Rom1, were mis-targeted to the inner segments and cell bodies of photoreceptor cells (Fig. 6B,D and Fig. 7B). Interestingly, the levels of Rom1 and transducin were notably lower in Ahi1−/− retinas than in Ahi1+/+ retinas, both immunohistochemically (Fig. 7A v. B and C v. D) and by Western blotting (Fig. 7E-H).
To determine whether the mis-targeting of the outer segment proteins that we observed was due to a general defect in vesicular trafficking, we investigated the localization of a second class of vesicles that are targeted to a distinct cellular domain, the synaptic vesicles. For this, we determined whether these vesicles were being correctly targeted to the axon terminals. We performed a localization of the synaptic vesicle marker, synaptotagmin, in the retina, at a time point prior to the onset of retinal degeneration in Ahi1−/− mice. Synaptotagmin immunostaining of Ahi1+/+ retinas was strong in the plexiform (synaptic) layers of the retina, at PN 11, consistent with the known distribution pattern of pre-synaptic markers in the retina (Fig. 5C, left). Ahi1−/− retinas showed an identical synaptotagmin pattern at PN 11 (Fig. 5C, right). This result indicates that Ahi1 is not required for the transport of synaptic vesicles to the cell periphery, nor is it required for targeting of the pre-synaptic vesicles to the axon and pre-synapse. Apparently, Ahi1 is not required for synaptic vesicle trafficking or targeting, but it is required for specific targeting of vesicles for outer segment formation.
We examined Rab8a expression and distribution in retinas from day E18.5 Ahi1−/− pups. Rab8a was found to localize to photoreceptor neuroblasts in both Ahi1+/+ and Ahi1−/− retinas (Fig. 8A,B). However, Rab8a levels were reduced in the photoreceptors from Ahi1−/− mice (Fig. 8A, B). This result and previous work from our lab indicates that Rab8a is destabilized in the absence of Ahi1 (Hsiao et al., 2009).
The data that we have presented demonstrate that Ahi1 is required for photoreceptor outer segment development, and that the lack of Ahi1 results in early photoreceptor degeneration. The failure of outer segment development in Ahi1−/− mice is complete, and may be due to a ciliary trafficking defect, as evidenced by mis-targeting of several outer segment proteins to the inner segments and cell bodies of photoreceptor cells in Ahi1−/− mice. The findings that basal bodies are positioned correctly, that axonemes begin to form, and that Rab8a levels are notably decreased in the photoreceptor cells of the Ahi1−/− mice are all consistent with the concept of such a ciliary trafficking defect. In contrast, non-ciliary trafficking appears to be normal in the retinas of Ahi1−/− mice, as indicated by the normal distribution seen for synaptotagmin. The requirement of Ahi1 for photoreceptor outer segment formation is consistent with, and provides insight into, the retinal degeneration phenotype in patients with JBTS.
Basal bodies, which are a clear marker of cellular polarity and are also required for cilium formation, were found to be in their proper apical position in the photoreceptors of Ahi1−/− mice. This suggests that Ahi1 is not required for the positioning of the basal body to the photoreceptor membrane, or for cell polarity. Interestingly, at late embryonic time points, rhodopsin-containing extensions from the photoreceptors were observed that morphologically resembled immature cilia. Such features could indicate that in the retina, Ahi1 is not required for initiation of cilium formation, but rather for ciliary growth and maturity, and/or outer segment maintenance.
The role just proposed for Ahi1 in outer segment growth/maturity could arise from the interaction of Ahi1 with Rab8a. Knockdown of Ahi1 in ciliated cells has been shown to result in significant decreases in Rab8a expression (Hsiao et al., 2009). We confirm here that loss of Ahi1 results in a reduction in Rab8a in photoreceptors. This result implicates the involvement of Ahi1 in stabilization of Rab8a levels in photoreceptor cells (Hsiao et al., 2009). It is likely that the reduction in Rab8a, induced by loss of Ahi1, plays a role in the failure of the process involved in outer segment development and maintenance. Moreover, since the loss of Rab8a results in ciliary defects (Nachury et al., 2007; Hsiao et al., 2009), our present results implicate the involvement of both Rab8a and Ahi1 as critical mediators of the growth and maturation of photoreceptor outer segments.
It is important to note that the outer segments of rod and cone photoreceptor cells are specialized sensory cilia. Membrane-bound vesicles carrying newly synthesized outer segment proteins are transported from the post-Golgi membranes and fuse with the plasma membrane near the transition zone, through which the proteins are delivered to the outer segments (Young, 1968; Deretic, 2006). Rab8a is sited near the base of the transition zone in photoreceptor cells and is thought to be necessary for proper rhodopsin-containing vesicle docking and transport from the post-Golgi to the plasma membrane of the transition zone (Deretic et al., 1995). This idea is supported by studies in which alterations in Rab8a function induced accumulation of rhodopsin along the outer edge of the rod inner segments, with no transport of rhodopsin occurring through the transition zone into the outer segments (Deretic et al., 1995; Deretic, 1997; Moritz et al., 2001). Such a loss of rhodopsin transport ultimately leads to photoreceptor cell degeneration (Deretic et al., 1995; Deretic, 1997; Moritz et al., 2001). The abnormal pattern of rhodopsin distribution in photoreceptors from Ahi1−/− mice, followed by loss of the outer nuclear layer, strongly resembles the Rab8a phenotype in Xenopus (Moritz et al., 2001). Moreover, we have recently shown that knockdown of Ahi1 results in impairments of trafficking from the Golgi to the plasma membrane (Hsiao et al., 2009), and that Ahi1 is involved (likely through its effects on Rab8a) in vesicular trafficking in inner medullary collecting duct (IMCD3) cells. Nevertheless, general polarized vesicle transport is unlikely to be affected, given that synaptotagmin was seen to localize to the plexiform layers without any extraneous mis-targeting to the outer nuclear layer in Ahi1−/− mice. Moreover, our synaptotagmin result indicates that Ahi1 is not required for the transport of synaptic vesicles to the axon and pre-synapse. As a whole, these results suggest that polarized vesicular trafficking is affected in ciliary, but not synaptically, targeted vesicles. Taken together, our results demonstrate that trafficking of proteins/vesicles to the outer segments is altered in the absence of Ahi1, suggesting that disruptions of this trafficking pathway are responsible for the failure of proper rhodopsin transport, and/or the growth and maintenance of the photoreceptor outer segments.
The absence of Ahi1 in photoreceptor cells results in a significant decrease in levels of Rab8a, as well as in decreases in levels of some outer segment proteins (i.e., transducin and Rom1), although not all of them (i.e., rhodopsin and Cnga1). These findings suggest either an involvement of Ahi1 in the stabilization of proteins in photoreceptors, or else an effect of Ahi1 on the transcription of the genes that encode these proteins. Future studies will be aimed at elucidating which of these two roles Ahi1 is actually playing in the cell.
The retinal degeneration observed in our Ahi1−/− mice recapitulates aspects of the retinal abnormalities that are observed in patients with JBTS (Parisi et al., 2006; Utsch et al., 2006; Parisi et al., 2007), and demonstrates the importance of Ahi1 in retinal function. Recently, Cep290, another JBTS-related protein, was found to associate with Rab8a, and to be required for Rab8a targeting to the centrosome and to the basal body of the primary cilium (Kim et al., 2008; Tsang et al., 2008); such a distribution is similar to our result for Ahi1 (Hsiao et al., 2009). Both Ahi1−/− mice and Cep290-deficient mice show early post-natal photoreceptor degeneration (Chang et al., 2006). Interestingly, Cep290-deficient mice, in contrast to our Ahi1−/− mice, do appear to form outer segments (Chang et al., 2006) suggesting that photoreceptor degeneration that occurs in Ahi1−/− and Cep290-deficient mice is possibly mediated by two distinct pathways. In support of this, it is becoming increasingly clear that photoreceptor degeneration can occur through the presence of active opsin in the photoreceptor cell body (Alfinito and Townes-Anderson, 2002; Lee and Flannery, 2007; Chinchore et al., 2009). Given the common retinal degeneration phenotype between our Ahi1−/− mice and Cep290-deficient mice, this suggests the importance of polarized trafficking of rhodopsin for maintaining photoreceptor health; both Ahi1−/− and Cep290-deficient mice have abnormal rhodopsin expression in the photoreceptor cell body. It further suggests the possibility that the activity of Rab8a, in controlling rhodopsin localization, is regulated together by Ahi1 and Cep290 in the photoreceptor sensory cilia. However, since Cep290-deficient mice appear to form outer segments, this suggests a unique role of Ahi1 in the formation of the photoreceptor sensory cilia, possibly through a role in the trafficking of membrane constituents necessary for the formation of outer segments (Hsiao et al., 2009).
Many ciliopathies are characterized by loss of vision, and many other systemic disorders that cause retinal degeneration can have a ciliopathic component (Adams et al., 2007). In only a few of these have the intercellular transport defects been characterized. In Bardet-Biedl syndrome, opsin is mis-targeted to the outer nuclear layer; however, outer segments are present, are only slightly stunted, and contain opsin (Nishimura et al., 2004; Abd-El-Barr et al., 2007). In Usher syndrome, there is, depending on which gene is mutated, either an impairment or else a complete block of opsin transport to the outer segment, with opsin accumulation seen in the connecting cilium. The onset of degradation in this case is seemingly a function of the severity of the transport defect (Liu et al., 1998; Liu et al., 1999). In both of the above disorders, rhodopsin is either able to traffic to the cilium or else it collects in or at the base of the cilium. The phenotype that we have seen in our Ahi1−/− mice is distinct from the above pattern, in that a loss of outer segments occurs with little accumulation of rhodopsin at the distal ends of the inner segments. Rhodopsin-bearing vesicles may initially be docking and starting to accumulate; however, they are cleared through an endocytotic mechanism that randomly distributes rhodopsin across the cell. We have previously shown that targeting of endocytotic vesicles is impaired in Ahi1-knockdown cells (Hsiao et al., 2009). Another explanation could be mis-targeting of rhodopsin-containing post-Golgi vesicles, causing most vesicles to shuttle indiscriminately, thereby preventing accumulation in the inner segments, and also causing the high level of rhodopsin accumulation seen at the inner side of the photoreceptors. Such mis-targeting could be due to loss of the interaction between Ahi1 and Rab8a, that we have previously shown to exist (Hsiao et al., 2009).
We have shown that Ahi1, a gene that causes Joubert syndrome when mutated, is required for the maintenance of photoreceptor outer segments, through the proper preservation of polarized vesicular trafficking to the cilia, and when absent results in retinal degeneration. This is consistent with findings from other mouse models of photoreceptor degeneration that show failure of outer segment formation. A classic example is the homozygous retinal degeneration slow (rds) mice, that has a mutation in the peripherin 2 (Prph2) gene, which encodes the peripherin 2 protein that is thought to be required for outer segment disc formation (Molday, 1998). Mutations in PRPH2 cause retinitis pigmentosa and several dominant macular dystrophies in humans (Farrar et al., 1991; Kajiwara et al., 1991; Nichols et al., 1993; Wells et al., 1993). Lastly, our current findings not only suggest a role for Ahi1 in Rab8a-mediated vesicle trafficking (Hsiao et al., 2009), but also provide clues toward a better understanding of the pathogenesis of ciliopathies.
This work was supported in part by the National Institutes of Health [EY12910 to E.A.P and MH71801 to R.J.F.]; the Kirby Foundation [to E.A.P]; the Foundation Fighting Blindness [to E.A.P]; Research to Prevent Blindness [to E.A.P]; the Rosanne Silbermann Foundation [to E.A.P]; and the March of Dimes Foundation [5-FY09-29 to R.J.F.]. The authors wish to thank Dr. Adriana Verschoor (Wadsworth Center) for critical reading and input on our manuscript as well as the Wadsworth Center for allowing us to maintain our animal colonies there since 2005.