Syntaxin 3 is essential for photoreceptor outer segment protein trafficking and survival
Edited by John E. Dowling, Harvard University, Cambridge, MA, and approved July 15, 2020 (received for review May 26, 2020)
Significance
Rod and cone photoreceptors are highly polarized cells which rely on incompletely understood trafficking pathways to deliver and sort proteins into multiple outer segment subcompartments (disc, rim, and plasma membrane). Here we identify the SNARE protein syntaxin 3 as a trafficking component for both disc and rim proteins but not plasma membrane proteins. Our results suggest that syntaxin 3-mediated vesicle fusion is a convergence point for disc and rim protein trafficking and sorting. Furthermore, we find no evidence that eliminating syntaxin 3 affects cone photoreceptor trafficking, suggesting that divergent trafficking pathways in the two cell types may underlie their different outer segment structure. These studies significantly enhance our understanding of the unique and elegant cell biology of the photoreceptor.
Abstract
Trafficking of photoreceptor membrane proteins from their site of synthesis in the inner segment (IS) to the outer segment (OS) is critical for photoreceptor function and vision. Here we evaluate the role of syntaxin 3 (STX3), in trafficking of OS membrane proteins such as peripherin 2 (PRPH2) and rhodopsin. Photoreceptor-specific Stx3 knockouts [Stx3f/f(iCre75) and Stx3f/f(CRX-Cre)] exhibited rapid, early-onset photoreceptor degeneration and functional decline characterized by structural defects in IS, OS, and synaptic terminals. Critically, in the absence of STX3, OS proteins such as PRPH2, the PRPH2 binding partner, rod outer segment membrane protein 1 (ROM1), and rhodopsin were mislocalized along the microtubules to the IS, cell body, and synaptic region. We find that the PRPH2 C-terminal domain interacts with STX3 as well as other photoreceptor SNAREs, and our findings indicate that STX3 is an essential part of the trafficking pathway for both disc (rhodopsin) and rim (PRPH2/ROM1) components of the OS.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
Proper assembly of the photoreceptor outer segment (OS), a modified primary cilium, is essential for vision. The OS is a unique structure consisting of stacks of membranous discs containing the machinery for phototransduction. It also represents a unique challenge for the biosynthetic machinery because the OS discs are constantly shed from the distal tips and renewed from the base (1, 2). Approximately 10% of OS material is turned over every day, thus large quantities of protein and lipid must be constantly synthesized and trafficked to the OS. This process is made even more complicated by the fact that rods and cones have different OS structure and different trafficking as well as by the precise organization of the OS. Newly synthesized proteins must be correctly sorted into either the plasma membrane domain, the disc domain, and/or the disc rim domain. This sophisticated, high-capacity trafficking and sorting system is essential for photoreceptor function but remains poorly understood, despite significant recent advances in the area. Membrane proteins destined for the OS are synthesized and packaged in the inner segment (IS), then delivered to the plasma membrane at the base of the OS. These trafficking vesicles fuse with the plasma membrane, after which proteins transport to the OS via the plasma membrane of the connecting cilium and incorporate into newly evaginating OS discs (3–6).
Peripherin 2 (PRPH2 also known as RDS), is a membrane protein essential for disc rim structure (7, 8). Mutations in PRPH2 cause various rod- and cone-dominant retinal degenerations in patients (9), and mice lacking PRPH2 do not form OSs (10). Studies suggest that PRPH2 traffics to the OS in different vesicles than other OS components such as rhodopsin or cone opsins (11–15). This idea is underscored by the finding that PRPH2 is trafficked in large part via an unconventional secretory pathway that bypasses the trans-Golgi and may involve the endosomal recycling system (16–19). However, the precise mechanisms that regulate the trafficking of PRPH2 remain unclear. In previous mass spectrometry experiments, we identified and validated PRPH2 interacting partners, including syntaxin 3 (STX3) and SNAP25 (17). STX3 and SNAP25 are members of the soluble NSF attachment protein receptor (SNARE) family. These proteins are involved in vesicle fusion with the photoreceptor plasma membrane and are known to participate in trafficking rhodopsin (20) as well as in synaptic vesicle release. Interestingly, we found that rhodopsin also interacts with STX3 (17), so we hypothesize that STX3-mediated fusion events might be a common mechanism by which vesicles trafficked through both conventional and unconventional secretory pathways converge on the way to the OS.
To evaluate the role of STX3 in trafficking OS disc proteins such as PRPH2 and rhodopsin, we eliminated STX3 in photoreceptors using a conditional knockout mouse line. In the absence of STX3, photoreceptors undergo rapid degeneration and OS proteins including PRPH2, the PRPH2 binding partner ROM1, and rhodopsin are mistrafficked. These findings confirm the importance of STX3 for trafficking of OS disc proteins and significantly enhance our understanding of the essential interplay between IS trafficking and OS structure.
Results
CRX-Cre and iCre75 Lead to Effective Knockdown of STX3 in the Retina.
Global knockout of Stx3 is lethal (21), so we generated a conditional knockout, with LoxP sites flanking exons 4 and 5 (SI Appendix, Fig. S1) which are required for all four STX3 isoforms (STX3A–D) (22) (SI Appendix, Fig. S1A). STX3 is the major syntaxin in photoreceptors (23), with STX3B being the only STX3 isoform expressed.
STX3 is expressed throughout postnatal development (SI Appendix, Fig. S2A), and in the adult retina is distributed in all photoreceptor compartments except the OS (17, 23). To knock down STX3 in specific retinal cells, we used either the cone-rod homeobox (CRX)-Cre (24, 25) or the rhodopsin Cre (iCre75) (26). CRX is initially expressed at embryonic day 12.5 (24, 27, 28) (SI Appendix, Fig. S2B), and the promoter fragment used to generate the CRX-Cre line generates early knockdown in both rods and cones as well as some bipolar cells (25). The iCre75 promoter leads to Cre activation in rods at approximately postnatal day 9 (P9) (26) (SI Appendix, Fig. S2B), but is not expressed in cones or other retinal cells. In the Stx3f/f(iCre75) mice, STX3 protein levels in the retina dropped dramatically as Cre levels increased (SI Appendix, Fig. S2 C and E), while in Stx3f/f(CRX-Cre) retinas, the earlier onset of Cre expression led to low STX3 protein levels from birth (SI Appendix, Fig. S2 D and F). Overall there was 76% knockdown in STX3 levels in Stx3f/f(iCre75) retinas at P30 compared to Stx3f/f littermates (referred to as wild type [WT]), and 94% knockdown in Stx3f/f(CRX-Cre) retinas at P15 (Fig. 1B). As expected, Stx3f/f(iCre75) retinas exhibit no STX3 expression in rods, in contrast to cones and inner retinal cells (Fig. 1 A and C). In the Stx3f/f(CRX-Cre), no STX3 was found in photoreceptors (Fig. 1A). Consistent with developmental expression of CRX-Cre in bipolar cells (25), little labeling for STX3 was present in the inner retina in the Stx3f/f(CRX-Cre) (Fig. 1C).
Fig. 1.

Elimination of STX3 Leads to Profound Degeneration and Functional Impairment in the Retina.
Significant photoreceptor degeneration was evident by P15 in Stx3f/f(CRX-Cre) retinas (Fig. 2 A and B), and no photoreceptors remained by P30 (Fig. 2 A and B, Middle and SI Appendix, Fig. S3B). Degeneration was slower in Stx3f/f(iCre75) retinas: progressive loss of photoreceptors was observed from P21 to P60 (Fig. 2 A and B and SI Appendix, Fig. S3A). The OS layer was also thinner in both Stx3f/f(CRX-Cre) and Stx3f/f(iCre75) compared to WT (Fig. 2A, red bars and SI Appendix, Fig. S3), and by P21 no OSs were present in the Stx3f/f(CRX-Cre) retina. By P45, remaining OSs in the Stx3f/f(iCre75) were extremely short. Inner nuclear layer (INL) thickness was also reduced in the Stx3f/f(CRX-Cre) retina at P21 (Fig. 2C), consistent with STX3 knockout from the inner retina in this line, but was unaffected in Stx3f/f(iCre75) mice (Fig. 2C). Heterozygous mice did not exhibit degeneration and were not evaluated further (SI Appendix, Fig. S3).
Fig. 2.

Transmission electron microscopy (TEM) showed that ISs had an abnormal accumulation of vesicles and membranous debris and a rounder, swollen appearance in the Stx3f/f(CRX-Cre) (Fig. 2 D, arrows and Inset, P15) and Stx3f/f(iCre75) (Fig. 2F, P21) retinas compared to WT. OSs in the Stx3f/f(CRX-Cre) retina had irregular disc size, stacking, and alignment at P15 (Fig. 2E, arrows). At P21, OSs and ISs were rare in the Stx3f/f(CRX-Cre) retina, and those that remained exhibited severe structural abnormalities (Fig. 2 F and G, Insets). In the P21 Stx3f/f(iCre75), some OSs were largely normal (though shorter than WT), while others exhibited severe abnormalities in disc alignment and sizing, as well as accumulation of debris (Fig. 2G, arrows). By P45, Stx3f/f(iCre75) OSs shortened further with significant empty space in the OS layer due to photoreceptor loss (Fig. 2 H and I).
Connecting cilia existed at ages when OSs were still present [P15 and 19 for Stx3f/f(CRX-Cre) and P21 and 30 for Stx3f/f(iCre75)] though they often appeared swollen (white arrows highlight cilia in Fig. 2J). Measurement of the acetylated region of the axoneme was performed and the length of the acetylated region was found to be significantly reduced in Stx3f/f(CRX-Cre) and Stx3f/f(iCre75) retinas compared to WT (Fig. 2 K and L) at P15 and P21, respectively.
STX3 is also found at photoreceptor synapses, so we evaluated synaptic structure. WT terminals were normal, showing long synaptic ribbons, synaptic vesicles (Fig. 3A and SI Appendix, Fig. S4A, arrows), and postsynaptic horizontal and bipolar cell processes in rods and cones and flat contacts in cones (arrowheads, Fig. 3A). Rods and cones in the Stx3f/f(iCre75) formed ribbon synapses, but showed progressive synaptic pathology, more severe in rods than cones, and were characterized by extensive synaptic ribbon remodeling, and some ribbons appeared to be free floating (at P21, black arrows, Fig. 3B, and SI Appendix, Fig. S4). Cone terminals from Stx3f/f(iCre75) mice had less synaptic disruption than rod terminals. Cones formed ribbon synapses and flat contacts (black arrowheads, Fig. 3B) until at least P45 (SI Appendix, Fig. S4B), although some synaptic ribbons showed a decline in structural integrity.
Fig. 3.

Rod and cone terminals in the Stx3f/f(CRX-Cre) retina showed abnormalities from the earliest time examined, P15 (SI Appendix, Fig. S4C), and progressive rod terminal loss was followed by cone terminal loss. Many rod terminals showed abnormally small, free floating ribbons (arrows, Fig. 3C), no discrete arciform densities, and free-floating presumptive ribbon material. Others lacked visible ribbons and contained abnormal vesicles, vacuoles, and multivesicular bodies, which became more pronounced with age (Fig. 3C, * and SI Appendix, Fig. S4C). Cone terminals also showed abnormalities in the Stx3f/f(CRX-Cre) retina, but were less affected than rods. There was little evidence of invagination by postsynaptic processes at ribbon contacts in rods or cones, and when present, appeared to originate from horizontal cells rather than bipolar cells (Fig. 3C). These findings were confirmed by immunofluorescence (IF) labeling: little colocalization between the rod spherule marker plasma membrane Ca2+ ATPase (PMCA) (29, 30) and the bipolar cell marker protein kinase C (PKC) (Fig. 3E) or the horizontal cell marker calbindin (SI Appendix, Fig. S5A) was observed in terminals in the Stx3f/f(CRX-Cre). Gross outer plexiform layer (OPL) defects were also visible. In the P15 Stx3f/f(CRX-Cre), the OPL was thinner than in WT, fewer photoreceptor terminals were present (Fig. 3D), and there was increased PMCA labeling within the outer nuclear layer (ONL), a common feature of degenerating photoreceptors, which retract their terminals out of the OPL as they die (Fig. 3 D and E, arrowheads). However, there was proper stratification of rod bipolar cell terminals (PKC) and VGLUT3 amacrine cell processes suggesting that the inner plexiform layer was undisturbed in Stx3f/f(CRX-Cre) and Stx3f/f(iCre75) retinas (SI Appendix, Fig. S5 B–D). Together, these results suggest that the absence of STX3 is deleterious to rod and cone synapses, and that rods are more severely affected than cones.
Retinal function measured by full-field electroretinography (ERG) was also reduced in STX3 knockouts. Scotopic (rod-driven) a-wave ERGs were significantly reduced by ∼55% at P30 and ∼66% at P60 (Fig. 4 A and C). Scotopic b-waves were similarly reduced, and scotopic a-/b-wave ratios were similar in WT and Stx3f/f(iCre75) mice, consistent with the overall lack of degeneration in the inner retina in the Stx3f/f(iCre75). Photopic (cone-driven) b-wave amplitudes in the Stx3f/f(iCre75) retina did not differ from WT controls at any time point (Fig. 4 B and D), although oscillatory potentials were reduced (Fig. 4B). Preservation of cone function is consistent with rod-specific STX3 knockout in Stx3f/f(iCre75) mice. In keeping with the more severe structural defects in the Stx3f/f(CRX-Cre) retina, these animals exhibited no scotopic or photopic ERG responses at any time points examined (Fig. 4).
Fig. 4.

Eliminating STX3 Impairs Photoreceptor OS Protein Trafficking.
We examined the localization of the OS proteins PRPH2, the PRPH2 binding partner ROM1, rhodopsin, and cone opsins, at ages when OSs were still present [P21 for Stx3f/f(iCre75) and P15 for Stx3f/f(CRX-Cre)]. PRPH2, ROM1, and rhodopsin are all proteins that interact with STX3 (17). In both knockout lines, a portion of PRPH2, ROM1, and rhodopsin mislocalized away from the OS to the IS, cell body, and synaptic terminals of photoreceptors (arrows, SI Appendix, Fig. S6 A–C and Fig. 5A). Even though rhodopsin and PRPH2 traffic to the OS through separate pathways (15), PRPH2, ROM1, and rhodopsin showed similar patterns of mislocalization in the Stx3f/f(CRX-Cre) retina and were largely colocalized (Fig. 5 A and B, arrowheads, Fig. 5B, and SI Appendix, Fig. S7, arrows) suggesting they were shunted to a common abnormal location in the absence of STX3. This observation suggests the possibility of the presence of an alternative pathway that could be a default one that uses multiple SNARE complexes and multiple redundancies to get these proteins to their destination. ROM1 showed less mislocalization than PRPH2 or rhodopsin, but mislocalized ROM1 was found in the same punctae as PRPH2 and rhodopsin. Cone opsins trafficked properly to the OS in knockdown retinas (SI Appendix, Fig. S6D).
Fig. 5.

RPGR, which is restricted to the base of the connecting cilium and the IS, did not colocalize with mislocalized PRPH2 (SI Appendix, Fig. S7C, arrowheads) suggesting that mistrafficked OS proteins did not accumulate at the base of the connecting cilium but localized to a different site in the IS. Microtubules move cargo within the cell, and indeed we find that mislocalized PRPH2, ROM1, and rhodopsin in the Stx3f/f(CRX-Cre) colocalize with acetylated α-tubulin within the photoreceptor cell bodies (Fig. 5 C–E), suggesting that these OS proteins traffic improperly along microtubules to the synaptic terminals in the absence of STX3.
To understand whether mislocalization of OS proteins in Stx3 knockout was restricted to disc proteins that interact with STX3, we labeled sections from Stx3f/f(CRX-Cre) retina (which showed more severe defects), with an antibody that recognizes both the rod cyclic nucleotide gated channel (CNGB1) and its nonmembrane bound isoform, glutamic acid rich protein (GARP). CNGB1 and GARP associate with the OS plasma membrane rather than the disc membrane and do not interact with STX3 (17). Critically, CNGB1/GARP did not mislocalize in the Stx3f/f(CRX-Cre) retina (SI Appendix, Fig. S8).
The PRPH2 C Terminus Interacts with SNARE Proteins.
Previously we identified STX3, SNAP25, and syntaxin binding protein 1 (STXBP1, also known as munc18) as PRPH2 binding partners (17). As expected, in anti-PRPH2 immunoprecipitations, STX3, SNAP25, and STXBP1 coprecipitated with PRPH2 in WT retinas, but were not pulled down from Prph2−/− retinas (confirming interactions were specific, Fig. 6 A and I; I, input; FT, flow through; E, bound protein eluted from beads; B, beads with uneluted protein). Both SNAP25 and STXBP1 came down with PRPH2 in Stx3f/f(CRX-Cre) retinas (Fig. 6A), though not to the same extent as in WT, suggesting that STX3 was not essential for interactions between PRPH2 and SNAP25/STXBP1. SNAP25 and STXBP1 are normally part of the SNARE complex at sites of vesicle fusion and are found throughout the IS, cell body, and synaptic terminals (Fig. 6 B–E). This pattern is altered in the Stx3f/f(CRX-Cre) retina: instead of even, diffuse labeling in the IS, SNAP25 and STXBP1 colocalized with PRPH2-positive (Fig. 6 B and C, arrows) and rhodopsin-positive punctae (Fig. 6D, arrows) in the IS.
Fig. 6.

To identify interacting regions of PRPH2 and STX3, we generated GST fusion constructs consisting of GST (control), GST with the cytoplasmic PRPH2 C-terminal (GST-PRPH2-CT), or GST with the STX3 SNARE domain (GST-STX3-SNARE). The PRPH2-CT does not pull down PRPH2 or ROM1 (the second intradiscal loop mediates these interactions) (31). However, GST-PRPH2-CT pulled down STX3, SNAP25, and STXBP1 from WT retinal lysates (lanes 4 and 6, Fig. 6F). To confirm that pulldown of STX3/SNAP25/STXBP1 with GST-PRPH2-CT was not due to coprecipitation with full-length PRPH2, we pulled down with GST-PRPH2-CT from the Prph2−/− retina (lane 5, Fig. 6F). STX3, STXBP1, and SNAP25 came down with GST-PRPH2-CT, indicating that full-length PRPH2 is not needed for these interactions, and that they are mediated by the PRPH2 C-terminal.
Previously we found that rhodopsin immunoprecipitated from the retina with STX3 antibodies (17). Similarly, rhodopsin also came down with STX3, SNAP25, and STXBP1 in the GST-PRPH2-CT pulldown from WT retina (Fig. 6F). Rhodopsin was absent in pulldowns from the Prph2−/− retina (Fig. 6F, lane 5), likely reflecting the fact that rhodopsin levels are undetectably low in the Prph2−/− (Fig. 6F, lane 2). No STX3, SNAP25, or rhodopsin were pulled down with the negative control (GST, Fig. 6G). To identify the region of STX3 involved in interactions with PRPH2, complexes pulled down by the STX3 SNARE domain were evaluated (Fig. 6 H and I). GST-STX3-SNARE pulled down PRPH2, ROM1, STXBP1, and rhodopsin from WT retina (Fig. 6H, lane 4). These four proteins were also pulled down with GST-STX3-SNARE from Stx3f/f(CRX-Cre) retinas (Fig. 6H, lane 6), confirming that full-length STX3 was not needed for these interactions to occur, rather they are mediated directly or indirectly by the STX3 SNARE domain. ROM1 and rhodopsin were not pulled down from the Prph2−/− retina, but again, these proteins are virtually undetectable in the absence of Prph2−/−.
Discussion
Here we present evidence supporting the importance of STX3 for photoreceptor OS protein trafficking and health. Knockout of STX3 in photoreceptors leads to early-onset, rapid degeneration. In the absence of STX3, photoreceptors exhibit defects in both the IS/OS where STX3 is involved in protein targeting to the OS, and at the synaptic terminals where STX3 is essential for proper synaptic vesicle release. Critically, in the absence of STX3 in rods, photoreceptor disc proteins such as PRPH2, ROM1, and rhodopsin mislocalize, as do vesicle fusion components STXBP1 and SNAP25, while OS plasma membrane proteins such as CNGB1/GARP do not. Given our observation that PRPH2, ROM1, and rhodopsin (but not CNGB1/GARP) interact with STX3 and other SNAREs such as SNAP25, and regulators of SNARE function, such as STXBP1, these findings suggest that STX3-mediated vesicle fusion is critical for proper delivery of disc proteins to the OS.
One interesting finding is that cone opsins traffic to the OS normally in the absence of STX3, while mislocalized PRPH2 colocalizes with mislocalized rhodopsin, suggesting that PRPH2 in cones was not mistrafficked. These observations suggest that cones traffic OS proteins differently from rods, an idea that is supported by other literature (32, 33). For example, mutations that affect PRPH2 oligomerization lead to mislocalization of PRPH2 and cone opsins, but do not affect rhodopsin trafficking (34, 35). These differences in rod vs. cone trafficking pathways may be part of the mechanism which leads to differences in OS assembly in rod vs. cone. We have previously found that in rods, disc formation is not initiated without PRPH2, while cone OS lamellae evaginate as long as cone opsins are present, with PRPH2 being important for disc rim formation as a second step (36–38). However, the linkage between differences in OS protein targeting and differences in OS assembly in rods vs. cones (as well as pathways for cone OS targeting) is an area where further exploration is needed.
Rhodopsin trafficking relies on small GTPases (Rab11, Rab8, and Arf4) and associated regulators (reviewed in refs. 39 and 40), STX3, SNAP25, VAMP7 (20, 41), and the Smad Anchor for Receptor Activation (SARA) (42), another STX3 interacting partner. SARA is important for vesicle fusion at early endosomes, and rhodopsin mislocalizes when SARA levels are reduced (42). Our understanding of PRPH2 trafficking lags significantly behind our understanding of rhodopsin trafficking, and no critical targeting GTPases have yet been identified. A significant amount of PRPH2 traffics via an unconventional route that bypasses the trans-Golgi (16, 17), but the molecular regulators of this pathway remain unknown. Recently, a role for Rab11a and late endosomes in PRPH2 ciliary/OS targeting has been proposed (19). Although these data may fit with our findings, there is no solid evidence that late endosomes are involved in the trafficking of OS membrane proteins in vertebrate rods. For such a system to exist, those proteins have to be delivered to the plasma membrane first, then endocytosed through early endosomes into the recycling endosome. The absence of such a system is supported by the lack of OS proteins in the IS plasma membrane in our study.
Despite the mislocalization of PRPH2, rhodopsin, and ROM1 in the absence of STX3, it is worth noting that a substantial fraction of these proteins is still delivered to the OS. It is possible that other syntaxins not usually expressed in photoreceptors are up-regulated in the Stx3 knockouts or that other alternative pathways exist that are syntaxin-independent mechanisms to deliver membrane proteins to the OS. It is known that proteins lacking specific targeting signals for other cellular compartments can be delivered to the OS by default (43). Much remains to be understood about how OS proteins are delivered to the OS.
An exciting finding is that interactions between PRPH2, STX3, SNAP25, and STXBP1 are mediated by the C-terminal of PRPH2 and the SNARE domain of STX3. The PRPH2 C-terminal is an attractive candidate interaction domain. It consists of amphipathic alpha-helices and intrinsically disordered domains, is known to interact with other proteins, and promotes OS targeting, disc curvature, and membrane fusion (11, 12, 44–48). Our data support the previously proposed hypothesis that the PRPH2 C-terminal may take the place of the v-SNARE in vesicle fusion during OS targeting (46, 49, 50). However, it is also possible that PRPH2 is an essential regulator of membrane fusion, but does not itself directly participate in a membrane fusion complex. In this case, the putative v-SNARE could be VAMP7, which has been shown in a frog model to be a v-SNARE involved in targeting rhodopsin vesicles (41). There is also support for tetraspanins (like PRPH2) to play a critical regulatory role with VAMP7 in other systems. For example, the tetraspanin protein CD82 regulates clathrin-mediated endocytosis of epidermal growth factor receptor (EGFR), and delivery of CD82 to the cell surface is dependent on VAMP7 (51). Clathrin vesicles are a part of STX3-mediated photoreceptor trafficking pathways (52) and, given the emerging importance of endosomal pathways in the trafficking of PRPH2 (19), this represents an exciting area for future exploration.
Retinal degeneration in Stx3 knockouts was severe and rapid. Photoreceptor death often follows malformation of OSs, (e.g., Rho−/− and Prph2−/−), so defects in OS protein trafficking and abnormal OS structure may contribute to photoreceptor degeneration in the Stx3 knockouts. However, degeneration in both Stx3 knockout lines was more rapid than in either the Rho−/− or Prph2−/− (10, 53), suggesting other mechanisms, such as synaptic dysfunction, also contribute to the degeneration. In addition to its role in OS protein trafficking, STX3 is essential for synaptic vesicle exocytosis/release in photoreceptors (23, 54). Consistent with a role for STX3 in synaptic release, we find that Stx3 knockout rod and cone synapses are highly abnormal. In addition to synaptic defects arising due to impaired synaptic vesicle release (i.e., direct consequences of STX3 loss in synaptic terminals), the terminals may suffer additional effects from STX3 loss. Many rod terminals in Stx3f/f(CRX-Cre) retinas accumulate abnormal membranous vacuoles/vesicles. This accumulation of membranous material may represent mistrafficked PRPH2/rhodopsin or disruption of normal synaptic vesicle cycling mechanisms, and thus contribute to accelerated degeneration.
Understanding the differences in photoreceptor protein sorting and trafficking mechanisms for different photoreceptor compartments and subcompartments is an essential unanswered cell biological question. Our studies provide strong evidence that STX3 is an essential piece of the trafficking puzzle and is critical for photoreceptor health. Critically, our findings suggest that cones and rods rely on different trafficking components, a finding which has significant implications for their differing OS structure. In addition, our data suggest that though initially trafficked separately, OS disc membrane and disc rim (but not OS plasma membrane) components converge at a STX3-dependent fusion event, providing increased mechanistic understanding for how proteins are sorted during OS biogenesis. These studies significantly expand our understanding of this process and open up several exciting areas for future exploration.
Materials and Methods
Additional information for each section is in SI Appendix.
Animals.
Animal experiments were approved by the University of Houston Institutional Animal Care and Use Committee and adhered to recommendations in the NIH Guide for the Care and Use of Laboratory Animals. Stx3 flox/flox (Stx3f/f) mice were generated by inGenious Targeting Laboratory, Inc., Ronkonkoma, NY and crossed with iCre75 transgenic mice (generated by Ching-Kang Chen, Baylor College of Medicine, Houston TX and made available with permission by Wolfgang Baehr, University of Utah, Salt Lake City, UT) (26) or CRX-Cre transgenic mice (generated by Tom M. Glaser, University of California, Davis, CA and made available with permission by Anand Swaroop, National Eye Institute, NIH, Bethesda, MD).
Immunofluorescence, Immunoblotting, Immunoprecipitation, and GST Pulldown.
Immunofluorescence, immunoprecipitation, SDS/PAGE, and Western blotting using retinal extracts were performed as described previously (17, 35, 55). GST pulldowns were performed as described previously using GST-STX3-SNARE (generously provided by Roger Janz, McGovern Medical School, UTHealth, Houston, TX) (56), GST-PRPH2-CT (31), and GST (expressed from pGEX4T2). Details on antibodies used are provided in SI Appendix, Table S1.
Electroretinography.
Full-field dark-adapted and light-adapted ERGs were recorded as previously described (57).
Transmission Electron Microscopy and Light Microscopic Histology.
Statistical Analysis.
Statistical testing was performed using GraphPad Prism, v.8.3.0 (GraphPad Software) and post hoc testing in cases with multiple groups was employed. All experiments (e.g., IF labeling, GST pulldown, ERG) were done on at least three separate occasions regardless of the number of individual animals used for each study.
Data Availability
All data are included in the manuscript and SI Appendix.
Acknowledgments
We thank Drs. Ching-Kang Chen, Wolfgang Baehr, Tom M. Glaser, Anand Swaroop, Hemant Khanna, Vadim Arshavsky, Roger Janz, and Robert Molday for the provision of tools as indicated in the text. This work was supported by the NIH (R01EY10609 to M.I.N. and M.R.A.-U., and National Institute of General Medical Sciences 1 P20 GM12552801A1 to S.M.C.).
Supporting Information
Appendix (PDF)
- Download
- 5.98 MB
References
1
R. W. Young, The renewal of photoreceptor cell outer segments. J. Cell Biol. 33, 61–72 (1967).
2
R. W. Young, D. Bok, Participation of the retinal pigment epithelium in the rod outer segment renewal process. J. Cell Biol. 42, 392–403 (1969).
3
T. Burgoyne et al., Rod disc renewal occurs by evagination of the ciliary plasma membrane that makes cadherin-based contacts with the inner segment. Proc. Natl. Acad. Sci. U.S.A. 112, 15922–15927 (2015).
4
J. D. Ding, R. Y. Salinas, V. Y. Arshavsky, Discs of mammalian rod photoreceptors form through the membrane evagination mechanism. J. Cell Biol. 211, 495–502 (2015).
5
R. H. Steinberg, S. K. Fisher, D. H. Anderson, Disc morphogenesis in vertebrate photoreceptors. J. Comp. Neurol. 190, 501–508 (1980).
6
S. Volland et al., Three-dimensional organization of nascent rod outer segment disk membranes. Proc. Natl. Acad. Sci. U.S.A. 112, 14870–14875 (2015).
7
M. L. Milstein et al., Multistep peripherin-2/rds self-assembly drives membrane curvature for outer segment disk architecture and photoreceptor viability. Proc. Natl. Acad. Sci. U.S.A. 117, 4400–4410 (2020).
8
K. Arikawa, L. L. Molday, R. S. Molday, D. S. Williams, Localization of peripherin/rds in the disk membranes of cone and rod photoreceptors: Relationship to disk membrane morphogenesis and retinal degeneration. J. Cell Biol. 116, 659–667 (1992).
9
C. J. Boon et al., The spectrum of retinal dystrophies caused by mutations in the peripherin/RDS gene. Prog. Retin. Eye Res. 27, 213–235 (2008).
10
S. Sanyal, A. De Ruiter, R. K. Hawkins, Development and degeneration of retina in rds mutant mice: Light microscopy. J. Comp. Neurol. 194, 193–207 (1980).
11
B. M. Tam, O. L. Moritz, D. S. Papermaster, The C terminus of peripherin/rds participates in rod outer segment targeting and alignment of disk incisures. Mol. Biol. Cell 15, 2027–2037 (2004).
12
R. Y. Salinas, S. A. Baker, S. M. Gospe 3rd, V. Y. Arshavsky, A single valine residue plays an essential role in peripherin/rds targeting to photoreceptor outer segments. PLoS One 8, e54292 (2013).
13
P. Avasthi et al., Trafficking of membrane proteins to cone but not rod outer segments is dependent on heterotrimeric kinesin-II. J. Neurosci. 29, 14287–14298 (2009).
14
M. M. Abd-El-Barr et al., Impaired photoreceptor protein transport and synaptic transmission in a mouse model of Bardet-Biedl syndrome. Vision Res. 47, 3394–3407 (2007).
15
R. N. Fariss, R. S. Molday, S. K. Fisher, B. Matsumoto, Evidence from normal and degenerating photoreceptors that two outer segment integral membrane proteins have separate transport pathways. J. Comp. Neurol. 387, 148–156 (1997).
16
G. Tian et al., An unconventional secretory pathway mediates the cilia targeting of peripherin/rds. J. Neurosci. 34, 992–1006 (2014).
17
R. Zulliger et al., SNAREs interact with retinal degeneration slow and rod outer segment membrane protein-1 during conventional and unconventional outer segment targeting. PLoS One 10, e0138508 (2015).
18
S. M. Conley et al., Prph2 initiates outer segment morphogenesis but maturation requires Prph2/Rom1 oligomerization. Hum. Mol. Genet. 28, 459–475 (2019).
19
W. Otsu, Y. C. Hsu, J. Z. Chuang, C. H. Sung, The late endosomal pathway regulates the ciliary targeting of tetraspanin protein peripherin 2. J. Neurosci. 39, 3376–3393 (2019).
20
J. Mazelova, N. Ransom, L. Astuto-Gribble, M. C. Wilson, D. Deretic, Syntaxin 3 and SNAP-25 pairing, regulated by omega-3 docosahexaenoic acid, controls the delivery of rhodopsin for the biogenesis of cilia-derived sensory organelles, the rod outer segments. J. Cell Sci. 122, 2003–2013 (2009).
21
E. Sanchez et al., Syntaxin 3, but not syntaxin 4, is required for mast cell-regulated exocytosis, where it plays a primary role mediating compound exocytosis. J. Biol. Chem. 294, 3012–3023 (2019).
22
L. Curtis et al., Syntaxin 3B is essential for the exocytosis of synaptic vesicles in ribbon synapses of the retina. Neuroscience 166, 832–841 (2010).
23
D. M. Sherry, R. Mitchell, K. M. Standifer, B. du Plessis, Distribution of plasma membrane-associated syntaxins 1 through 4 indicates distinct trafficking functions in the synaptic layers of the mouse retina. BMC Neurosci. 7, 54 (2006).
24
A. Nishida et al., Otx2 homeobox gene controls retinal photoreceptor cell fate and pineal gland development. Nat. Neurosci. 6, 1255–1263 (2003).
25
L. Prasov, T. Glaser, Pushing the envelope of retinal ganglion cell genesis: Context dependent function of Math5 (Atoh7). Dev. Biol. 368, 214–230 (2012).
26
S. Li et al., Rhodopsin-iCre transgenic mouse line for Cre-mediated rod-specific gene targeting. Genesis 41, 73–80 (2005).
27
A. Furukawa, C. Koike, P. Lippincott, C. L. Cepko, T. Furukawa, The mouse Crx 5′-upstream transgene sequence directs cell-specific and developmentally regulated expression in retinal photoreceptor cells. J. Neurosci. 22, 1640–1647 (2002).
28
T. Furukawa, E. M. Morrow, C. L. Cepko, Crx, a novel otx-like homeobox gene, shows photoreceptor-specific expression and regulates photoreceptor differentiation. Cell 91, 531–541 (1997).
29
J. Johnson et al., Vesicular glutamate transporter 3 (VGLUT3) expression identifies glutamatergic amacrine cells in the rodent retina. J. Comp. Neurol. 477, 386–398 (2004).
30
J. E. Johnson Jr. et al., Spatiotemporal regulation of ATP and Ca2+ dynamics in vertebrate rod and cone ribbon synapses. Mol. Vis. 13, 887–919 (2007).
31
X. Q. Ding, H. M. Stricker, M. I. Naash, Role of the second intradiscal loop of peripherin/rds in homo and hetero associations. Biochemistry 44, 4897–4904 (2005).
32
S. Karan, H. Zhang, S. Li, J. M. Frederick, W. Baehr, A model for transport of membrane-associated phototransduction polypeptides in rod and cone photoreceptor inner segments. Vision Res. 48, 442–452 (2008).
33
R. K. Raghupathy et al., Rpgrip1 is required for rod outer segment development and ciliary protein trafficking in zebrafish. Sci. Rep. 7, 16881 (2017).
34
S. M. Conley, D. Chakraborty, M. I. Naash, Mislocalization of oligomerization-incompetent RDS is associated with mislocalization of cone opsins and cone transducin. Adv. Exp. Med. Biol. 723, 657–662 (2012).
35
D. Chakraborty, X. Q. Ding, S. M. Conley, S. J. Fliesler, M. I. Naash, Differential requirements for retinal degeneration slow intermolecular disulfide-linked oligomerization in rods versus cones. Hum. Mol. Genet. 18, 797–808 (2009).
36
R. Farjo et al., Retention of function without normal disc morphogenesis occurs in cone but not rod photoreceptors. J. Cell Biol. 173, 59–68 (2006).
37
S. M. Conley et al., Insights into the mechanisms of macular degeneration associated with the R172W mutation in RDS. Hum. Mol. Genet. 23, 3102–3114 (2014).
38
D. Chakraborty, S. M. Conley, M. R. Al-Ubaidi, M. I. Naash, Initiation of rod outer segment disc formation requires RDS. PLoS One 9, e98939 (2014).
39
J. Wang, D. Deretic, Molecular complexes that direct rhodopsin transport to primary cilia. Prog. Retin. Eye Res. 38, 1–19 (2014).
40
J. N. Pearring, R. Y. Salinas, S. A. Baker, V. Y. Arshavsky, Protein sorting, targeting and trafficking in photoreceptor cells. Prog. Retin. Eye Res. 36, 24–51 (2013).
41
V. Kandachar, B. M. Tam, O. L. Moritz, D. Deretic, An interaction network between the SNARE VAMP7 and Rab GTPases within a ciliary membrane-targeting complex. J. Cell Sci. 131, jcs222034 (2018).
42
J. Z. Chuang, Y. Zhao, C. H. Sung, SARA-regulated vesicular targeting underlies formation of the light-sensing organelle in mammalian rods. Cell 130, 535–547 (2007).
43
S. A. Baker et al., The outer segment serves as a default destination for the trafficking of membrane proteins in photoreceptors. J. Cell Biol. 183, 485–498 (2008).
44
M. L. Milstein, V. A. Kimler, C. Ghatak, A. S. Ladokhin, A. F. X. Goldberg, An inducible amphipathic helix within the intrinsically disordered C terminus can participate in membrane curvature generation by peripherin-2/rds. J. Biol. Chem. 292, 7850–7865 (2017).
45
N. Khattree, L. M. Ritter, A. F. Goldberg, Membrane curvature generation by a C-terminal amphipathic helix in peripherin-2/rds, a tetraspanin required for photoreceptor sensory cilium morphogenesis. J. Cell Sci. 126, 4659–4670 (2013).
46
K. Boesze-Battaglia et al., The tetraspanin protein peripherin-2 forms a complex with melanoregulin, a putative membrane fusion regulator. Biochemistry 46, 1256–1272 (2007).
47
M. Damek-Poprawa, J. Krouse, C. Gretzula, K. Boesze-Battaglia, A novel tetraspanin fusion protein, peripherin-2, requires a region upstream of the fusion domain for activity. J. Biol. Chem. 280, 9217–9224 (2005).
48
T. C. Edrington 5th, P. L. Yeagle, C. L. Gretzula, K. Boesze-Battaglia, Calcium-dependent association of calmodulin with the C-terminal domain of the tetraspanin protein peripherin/rds. Biochemistry 46, 3862–3871 (2007).
49
K. Boesze-Battagliaa, F. P. Stefano, Peripherin/rds fusogenic function correlates with subunit assembly. Exp. Eye Res. 75, 227–231 (2002).
50
K. Boesze-Battaglia, O. P. Lamba, A. A. Napoli Jr., S. Sinha, Y. Guo, Fusion between retinal rod outer segment membranes and model membranes: A role for photoreceptor peripherin/rds. Biochemistry 37, 9477–9487 (1998).
51
M. Chaineau, L. Danglot, T. Galli, Multiple roles of the vesicular-SNARE TI-VAMP in post-Golgi and endosomal trafficking. FEBS Lett. 583, 3817–3826 (2009).
52
S. Veleri et al., REEP6 mediates trafficking of a subset of Clathrin-coated vesicles and is critical for rod photoreceptor function and survival. Hum. Mol. Genet. 26, 2218–2230 (2017).
53
J. Lem et al., Morphological, physiological, and biochemical changes in rhodopsin knockout mice. Proc. Natl. Acad. Sci. U.S.A. 96, 736–741 (1999).
54
C. L. Hays et al., Simultaneous release of multiple vesicles from rods involves synaptic ribbons and syntaxin 3B. Biophys. J. 118, 967–979 (2020).
55
M. W. Stuck, S. M. Conley, M. I. Naash, The Y141C knockin mutation in RDS leads to complex phenotypes in the mouse. Hum. Mol. Genet. 23, 6260–6274 (2014).
56
X. Liu, R. Heidelberger, R. Janz, Phosphorylation of syntaxin 3B by CaMKII regulates the formation of t-SNARE complexes. Mol. Cell. Neurosci. 60, 53–62 (2014).
57
W. L. Cheng, Z. F. Qian, J. Su, [Effect of huoxue tablet on the rheology of erythrocyte in primary hypertension]. Zhongguo Zhong Xi Yi Jie He Za Zhi 17, 718–720 (1997).
58
H. M. Stricker, X.-Q. Ding, A. Quiambao, S. J. Fliesler, M. I. Naash, The Cys214–>Ser mutation in peripherin/rds causes a loss-of-function phenotype in transgenic mice. Biochem. J. 388, 605–613 (2005).
Information & Authors
Information
Published in
Classifications
Copyright
© 2020. Published under the PNAS license.
Data Availability
All data are included in the manuscript and SI Appendix.
Submission history
Published online: August 10, 2020
Published in issue: August 25, 2020
Keywords
Acknowledgments
We thank Drs. Ching-Kang Chen, Wolfgang Baehr, Tom M. Glaser, Anand Swaroop, Hemant Khanna, Vadim Arshavsky, Roger Janz, and Robert Molday for the provision of tools as indicated in the text. This work was supported by the NIH (R01EY10609 to M.I.N. and M.R.A.-U., and National Institute of General Medical Sciences 1 P20 GM12552801A1 to S.M.C.).
Notes
This article is a PNAS Direct Submission.
Authors
Competing Interests
The authors declare no competing interest.
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
Cite this article
Syntaxin 3 is essential for photoreceptor outer segment protein trafficking and survival, Proc. Natl. Acad. Sci. U.S.A.
117 (34) 20615-20624,
https://doi.org/10.1073/pnas.2010751117
(2020).
Copied!
Copying failed.
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.