(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Super-resolution mapping in rod photoreceptors identifies rhodopsin trafficking through the inner segment plasma membrane as an essential subcellular pathway [1]

['Kristen N. Haggerty', 'Department Of Ophthalmology', 'Visual Sciences', 'Department Of Biochemistry', 'Molecular Medicine', 'West Virginia University', 'Morgantown', 'West Virginia', 'United States Of America', 'Shannon C. Eshelman']

Date: 2024-01

Photoreceptor cells in the vertebrate retina have a highly compartmentalized morphology for efficient phototransduction and vision. Rhodopsin, the visual pigment in rod photoreceptors, is densely packaged into the rod outer segment sensory cilium and continuously renewed through essential synthesis and trafficking pathways housed in the rod inner segment. Despite the importance of this region for rod health and maintenance, the subcellular organization of rhodopsin and its trafficking regulators in the mammalian rod inner segment remain undefined. We used super-resolution fluorescence microscopy with optimized retinal immunolabeling techniques to perform a single molecule localization analysis of rhodopsin in the inner segments of mouse rods. We found that a significant fraction of rhodopsin molecules was localized at the plasma membrane, at the surface, in an even distribution along the entire length of the inner segment, where markers of transport vesicles also colocalized. Thus, our results collectively establish a model of rhodopsin trafficking through the inner segment plasma membrane as an essential subcellular pathway in mouse rod photoreceptors.

Funding: This work was supported by the National Institute of Health ( https://www.nih.gov/ ): P20-GM144230 (MAR); Knights Templar Eye Foundation to MAR and Discovery Grant (RGPIN-2022-02982) from the Natural Sciences and Engineering Research Council of Canada to MAA. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Overall, considering the current uncertainty regarding the impact of protein regulators of Rho biosynthesis and trafficking in mice, the complete route of Rho trafficking within the IS remains incompletely characterized despite the importance of trafficking within this critical subcellular compartment in mouse rods. In this study, we used super-resolution fluorescence microscopy to perform a quantitative spatial analysis of Rho localization in mouse rods ISs. Super-resolution modalities are powerful tools for testing the subcellular localization of protein targets in mouse rod domains, such as the IS. Both structured illumination microscopy (SIM), and stochastic optical reconstruction microscopy (STORM), a single-molecule localization microscopy, were previously used to localize ciliary proteins to the nanometer-scale subcompartments of the CC in mouse rods [ 24 , 57 ]. More recently, an alternative 3D STORM mode, named rapid imaging of tissues at the nanoscale STORM (RAIN-STORM), was developed to localize proteins in rod photoreceptor presynaptic terminals and postsynaptic dendrites within the mouse OPL [ 58 ]. To enable super-resolution localization of Rho in the mouse IS, we applied techniques to reliably immunolabel Rho in the IS of mouse retina, including an OS peeling approach, nanobody targeting of Rho-GFP in Rho-GFP/+ knock-in retinas, and surface labeling for SIM and STORM followed by a quantitative localization analysis.

The network of small GTPase protein regulators of the Rho secretory pathway in the IS has been systematically defined in frog rods. In this pathway, Arf4 GTPase binds to Rho in the trans-Golgi network via the Rho C-terminal VxPx motif [ 44 ], a sequence that was shown to be essential for OS targeting in frogs and zebrafish [ 45 , 46 ]. ASAP1, the Arf GTPase activating protein (GAP), and small GTPases FIP3 and Rab11a are then recruited to the complex to form post-Golgi Rho carrier vesicles [ 47 , 48 ], which are then targeted to the plasma membrane by Rab8a GTPase and Rabin8, the Rab8 guanine nucleotide exchange factor [ 49 – 51 ]. In mice, Arf4, Rab8, and Rab11a are not essential for Rho trafficking [ 52 , 53 ], suggesting the existence of redundant regulators or compensatory trafficking networks in mouse rods. One common Rho trafficking mechanism is the role of IS SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) proteins. Rho vesicle docking to the IS plasma membrane in frog rods requires the proteins syntaxin 3 (STX3) and SNAP25 [ 54 ]. In the retinas from rod-specific STX3 knockout mice, Rho and the OS disc rim protein peripherin-2 are mislocalized to the IS and ONL [ 55 ], demonstrating that the functional role for STX3 is maintained in mice. Furthermore, STX3 was also shown to form in vivo protein interactions with Rho [ 56 ]. After SNARE-mediated docking, Rho-containing vesicles putatively fuse with the plasma membrane and Rho protein gets inserted into the membrane, an event that is classically mapped near the BB and CC in the apical IS of frog rods [ 33 ], but this key event has not been thoroughly examined in the mouse rod IS.

In the IS of frog rod photoreceptors, Rho has been localized with immunoelectron microscopy (immuno-EM) with the Golgi, in post-Golgi Rho carrier vesicles, and sporadically at the IS plasma membrane [ 32 – 34 ]. In early immuno-EM mammalian retina studies, Rho was localized to the IS plasma membrane in both mouse and cow rods [ 35 ], and in immunolabeled thin sections of pig retina, Rho was grossly localized to the IS plasma membrane, Golgi, and ONL [ 36 ]. More recent images of post-embedding immuno-EM or cryo-immuno-EM localization of Rho in adult mouse rods have shown minimal Rho labeling in the IS [ 26 , 37 , 38 – 42 ]. Despite the lack of comprehensive Rho IS localization details from these studies, Rho has been consistently localized at mouse CC plasma membrane in mouse rods [ 38 – 40 , 42 ]. In fluorescence localization studies using mammalian retina, including in hRho-GFP fusion knock-in mouse retinas, it has to date been challenging to develop methodology to visualize the small population of Rho molecules in the IS compared to the overwhelming amount of Rho in OS discs [ 43 ].

The rod IS compartment is the biosynthetic domain of rods that is filled with endoplasmic reticulum (ER) and Golgi secretory organelles, mitochondria, and cytoplasmic microtubules all surrounded by a plasma membrane. The rod IS microtubular network is proposed to nucleate from the basal body (BB) [ 8 , 25 ], which is a region of the apical IS composed of 2 centrioles, one of which—the mother centriole—is continuous with the axoneme of the CC. Cytoplasmic dynein 1 has been demonstrated to be essential for rod health as the putative motor protein complex for intracytoplasmic movement toward the minus end of microtubules in the IS [ 26 – 29 ]. The BB is also the site of the distal appendages (DAPs), which are 9 radially symmetrical pinwheel-shaped blades that link the mother centriole to the plasma membrane and may serve as a gate or barrier structure at the critical IS/CC interface (reviewed in Wensel and colleagues’ study [ 30 ]). The mouse rod IS also contains a ciliary rootlet, which is a filamentous cytoskeletal element that is linked to the BB and extends to the rod presynaptic terminal [ 31 ]. Still, there is uncertainty regarding the subcellular localization and integration of Rho and other nascent proteins within the IS compartment architecture.

Mislocalization of Rho is a hallmark phenotype reported in many animal models of retinal disease, including models for retinitis pigmentosa (RP) and other retinal ciliopathies. RP is an inherited neurodegeneration affecting 1:4,000 in the United States of America [ 10 ] that causes neuronal cell death in rod photoreceptors and subsequent loss of vision. In animal models, a wide range of Rho RP mutations have been demonstrated to cause Rho protein mislocalization to the rod IS region, as well as some degree of Rho mislocalization to the photoreceptor outer nuclear layer (ONL) and outer plexiform layer (OPL) [ 11 – 15 ]. RP mutations to other rod-specific genes also result in Rho mislocalization [ 16 – 19 ], and mouse models for syndromic retinal ciliopathies and Leber congenital amaurosis (LCA) are caused by mutations in cilia-related genes that lead to rod degeneration in the retina due to defective ciliary ultrastructure that is comorbid with Rho mislocalization [ 20 – 24 ]. Collectively, rod dystrophies caused by any number of disruptions to normal rod homeostasis lead to Rho mislocalization, such that restorative therapeutic strategies to treat these retinal diseases must account for proper trafficking of Rho to preserve long-term rod stability. Therefore, a thorough understanding of the cellular mechanisms of Rho synthesis and trafficking in the IS of rods is essential for the development of effective retinal therapies.

In the vertebrate retina, vision is initiated by the phototransduction cascade in the outer segment (OS) sensory cilia of rod and cone photoreceptor cells. The light-absorbing G-protein coupled receptor (GPCR) proteins rhodopsin (Rho) in rods and the cone opsins in cones are densely packaged into flattened membrane discs that are stacked within the OS cilium. In the mouse retina, each rod OS contains approximately 800 discs [ 1 ] with a Rho packaging density of approximately 75,000 Rho molecules per disc [ 2 – 4 ]. Rod OS discs are completely renewed in about 10 days through the process of disc shedding from the distal rod OS tips [ 5 , 6 ]. As such, new discs are formed at the OS base, and these nascent discs must be filled with newly synthesized protein, primarily Rho, which is trafficked through the rod photoreceptor inner segment (IS) to the OS by way of the connecting cilium (CC) [ 7 , 8 ]. In mouse rods, the CC is a thin, approximately 300-nm-diameter ciliary bridge that spans 1.1 μm between the IS and OS and is composed of an axoneme core of 9 microtubule doublets that extend into the OS [ 9 ]. A coordinated homeostasis of OS disc shedding, nascent disc formation, IS protein delivery, and CC trafficking must be maintained in photoreceptors for a lifetime of proper vision.

Results

Rhodopsin immunofluorescence labeling strategies in mouse retina Previous studies established that immunolabeling the N-terminus or C-terminus of Rho labels different fractions of Rho molecules in rods from mammalian retinal tissue preparations [36,59]. The immunofluorescence staining pattern of Rho was tested using the following antibodies: (1) the 1D4 monoclonal antibody (hereafter Rho-C-1D4), which targets the last 8 amino acids in C-terminus/cytoplasmic tail of Rho, i.e., the 1D4 sequence [60]; (2) the N-terminus targeting 4D2 monoclonal antibody (Rho-N-4D2) [59]; and (3) an N-terminus targeting Rho-N-GTX polyclonal antibody. For immunofluorescence, the Rho antibodies were first tested in immunolabeled mouse retinal cryosections imaged with confocal scanning microscopy. With this method, all the Rho antibodies targeted the OS layer almost exclusively with strong fluorescence labeling (Fig 1A). Rho immunofluorescence localization in mouse retinas with SIM and STORM was also tested, which required a whole-retina immunolabeling approach modified from previous STORM imaging studies [24,57] for both improved labeling density and thin resin sectioning of stained retinas (see Materials and methods for details). By whole-retina labeling and SIM, the Rho-C-1D4 antibody only labeled the distal rod OS tips with fluorescence, while both N-terminal antibodies, Rho-N-4D2 and Rho-N-GTX, predominantly labeled the base of the OS where new discs are formed (Fig 1B). N-terminal Rho immunolabeling at the OS base is consistent with the N-termini of Rho in newly formed evaginating discs being exposed to the extracellular space [61]. In each condition, Rho protein in the IS was not reliably detectable with immunofluorescence. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Mouse rod inner segments are nonpermissive to conventional Rho labeling strategies. (A) Diagram of mouse rod photoreceptor layers adjacent to z-projection confocal fluorescent images of WT mouse retinal cryosections immunolabeled for Rho (magenta), STX3 (cyan)—which labels the IS and OPL—and DAPI nuclear staining (blue) in the ONL. Rho prominently labels rod OS. The STX3 channel is removed from part of the images for clarity. (B) Z-projection SIM images of thin (1 μm) plastic sections of WT retinas stained for immunofluorescence as whole retinas. The white arrow indicates the rod that is magnified in the Rho-N-GTX example, and the white asterisk indicates Rho-N-GTX labeling of the new OS discs in the magnified rod. Centrin (yellow) immunolabeling was used to localize the rod CC and BB. (C) Z-projection confocal image of an adult Rho-GFP/+ cryosection adjacent to z-projection SIM images of 2 μm Rho-GFP/+ cryosections. (D) Z-projection confocal image of a Rho-GFP-1D4/+ cryosection and z-projection SIM images of 2 μm Rho-GFP/+ cryosections. In both (C) and (D), a magnified region of a SIM image is shown with raised contrast and brightness (intensity) levels to depict faint IS GFP fluorescence in both heterozygous knock-in mouse lines (yellow arrows). Inverted images are shown to highlight this pattern. (E) Z-projection SIM images of Rho-GFP/+ thin plastic retina sections that were immunolabeled for NbGFP-A647 (magenta) and centrin (yellow) as whole retinas. Raised intensity images are shown to depict less intense NbGFP-A647 labeling in some proximal OSs and surrounding the CC. (F) SIM images of thin sections from Rho-GFP/+ and WT retinas that were stained as in (E) but with twice the amount of NbGFP-A647 (x2). Lack of staining in WT retinas demonstrates NbGFP-A647 specificity. BB, basal body; CC, connecting cilium; IS, inner segment; ONL, outer nuclear layer; OPL, outer plexiform layer; OS, outer segment; Rho, rhodopsin; SIM, structured illumination microscopy; STX3, syntaxin 3; WT, wild-type. https://doi.org/10.1371/journal.pbio.3002467.g001 Next, 2 knock-in Rho fusion mouse lines were used: Rho-GFP and Rho-GFP-1D4. In each knock-in, GFP is fused to human Rho directly after the C-terminal 1D4 sequence [43]. In the Rho-GFP-1D4 version of the knock-in, an additional 1D4 sequence, which contains the VxPx targeting motif, is appended to the C-terminus of GFP [15]. Although both knock-in mice, as heterozygotes, have normal photoreceptor morphology and Rho OS localization, the addition of the extra 1D4 sequence was previously shown to be essential for GFP and Rho-Dendra2 OS targeting in frog and fish retinas [45,46,62], and so the Rho-GFP-1D4 knock-in mice was included in this mouse study to account for any possible subcellular localization requirement of the 1D4 sequence. In confocal images of Rho-GFP/+ and Rho-GFP-1D4/+ retinas, GFP fluorescence was confined to in the OS layer, and only a faint, noise-like GFP signal was visualized in the IS layer in thin (2 μm) retinal cryosections from both knock-in mice using SIM (Fig 1C and 1D). Therefore, a GFP-specific nanobody was obtained, purified, and conjugated to Alexa 647, a bright and stable fluorophore compatible with both SIM and STORM, to both enhance the GFP signal in our knock-in mouse retinas and to overcome the issue of GFP bleaching during our whole retina processing technique. Nanobodies are small (approximately 15 kDa) single-chain camelid immunoglobulins that are ideal for super-resolution fluorescence microscopy [63]. The specificity of the GFP nanobody Alexa 647 conjugate (abbreviated NbGFP-A647) was validated with western blotting and confocal immunofluorescence (S1A–S1C Fig), in which WT retinas were used as controls for Rho-GFP labeling with NbGFP-A647. In each test, NbGFP-A647 properly labeled Rho-GFP or Rho-GFP-1D4 with no evidence of nonspecific labeling in WT controls. Notably, the Rho-C-1D4 antibody properly labeled the Rho-GFP-1D4 protein band in western blots of Rho-GFP-1D4/+ retinal lysates (S1B Fig). To validate these mouse lines further, we used a deglycosylation assay to confirm that Rho-GFP and Rho-GFP-1D4 fusion proteins, from Rho-GFP/+ and Rho-GFP-1D4/+ retina lysates, respectively, were properly glycosylated like the endogenous mouse Rho protein from the same lysates (S1D and S1E Fig). Next, whole retina immunolabeling of Rho-GFP/+ and WT retinas with NbGFP-A647 was performed. In SIM images of thin sections from these Rho-GFP/+ retinas, NbGFP-A647 fluorescence was again predominantly limited to the distal OS tips, and only a few rods had any Rho-GFP labeling in the proximal OS near the centrin-positive CC (Fig 1E). Immunolabeling of centrins, which are calcium-binding proteins localized specifically to the CC and BB centrioles in mouse rods [57,64], was used to label the positions of the CC and BB as the apical edge of the IS throughout the study. Even when Rho-GFP/+ retinas were stained with double the amount of NbGFP-A647 (Fig 1F), the Rho-GFP-localized fluorescence was still predominantly found in the distal OS tips of Rho-GFP/+ retinas with no detectable fluorescent signal in the IS, indicating that NbGFP-A647 could not penetrate into the IS.

Outer segment peeling of mouse retinas enables immunofluorescence detection of rhodopsin in the photoreceptor inner segment layer To improve the penetration of our immunoreagents to the rod IS, a retinal peeling technique was used in which mouse retinas were iteratively adhered to and removed from filter paper to physically detach OSs from the retina in order to generate IS-enriched retina samples [65,66] (Fig 2A). OS removal using between 4 and 8 paper peeling cycles was tested using standard transmission electron microscopy (TEM). Eight peeling cycles resulted in increased OS removal, and so this number was used throughout this investigation to generate IS-enriched retinas (S2A Fig). After generating IS-enriched retinas from Rho-GFP/+ mice, the enrichment of IS proteins was validated with western blotting, where the amount of OS proteins, Rho-GFP, endogenous mouse Rho (labeled with Rho-C-1D4 antibody), and ROM1, were reduced compared to full retina control samples, while the levels of IS proteins STX3 and Rab11a were unchanged between conditions (Fig 2B). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. OS peeling generates IS-enriched mouse retina samples. (A) Diagram of the mouse retina OS peeling method. Mouse retina slices are iteratively peeled from filter paper 8 times to remove rod OSs (yellow) from the IS layer (magenta). (B) Western blot analysis comparing control full retina samples vs. retinas after OS peeling/IS-enriched retinas; both from Rho-GFP/+ mice. Approximately 2% of total volume from lysates from 1 retina (either control or peeled) were used from each condition for SDS-PAGE. Molecular weight marker sizes are indicated in kDa. Immunolabeled bands from antibodies targeting OS proteins, including NbGFP-A647, Rho-C-1D4, and ROM1 (black arrow points to the ROM1 monomer band), were reduced in peeled lysates, whereas IS proteins STX3 and Rab11a were roughly equal demonstrating IS enrichment. Tubulin immunoblotting was performed as a loading control. (C) Control full retina slices and (D) IS-enriched retinal slices from Rho-GFP/+ mice were fixed and stained for resin embedding, ultrathin sectioning, and TEM ultrastructure analysis. TEM images were pseudocolored to point out key rod structures as follows: OS = yellow, IS = magenta, CCs/BBs = blue. Single rod examples are magnified from both conditions to emphasize intact CC ultrastructure and the preservation of the IS plasma membrane (magenta arrows). The locations of the DAPs and rootlet are annotated in (D). (E) Z-projection SIM images of Rho-GFP/+ IS-enriched retinas immunolabeled with NbGFP-A647 (magenta) and centrin antibody (green). Fluorescence in the single rod example corresponding to the OS, IS, and CC are annotated. (F) Control SIM image of a WT IS-enriched retina immunolabeled and imaged as in (E) to demonstrate NbGFP-A647 labeling specificity. Scale bar values match adjacent panels when not labeled. BB, basal body; CC, connecting cilium; DAP, distal appendage; IS, inner segment; NbGFP-A647, GFP nanobody Alexa 647 conjugate; OS, outer segment; Rho, rhodopsin; STX3, syntaxin 3; SIM, structured illumination microscopy; TEM, transmission electron microscopy; WT, wild-type. https://doi.org/10.1371/journal.pbio.3002467.g002 Next, Rho-GFP/+ retinas were fixed for TEM immediately after peeling/IS enrichment to evaluate if the IS and CC ultrastructure were maintained after OS removal. Compared to the overall morphology of unpeeled Rho-GFP/+ rod photoreceptors (Fig 2C), IS-enriched retinas contained fewer OSs with a generally intact IS layer (Fig 2D). In higher-magnification TEM micrographs, IS-enriched rod examples had preserved CC structure that were typically attached to a remaining OS, and these rods had an intact IS plasma membrane and cytoplasm (Figs 2D and S2B). Because there was evidence that some ISs lacked CC as a result of the OS peeling, all IS-enriched retinas analyzed in this study were stained with CC/BB markers to ensure that only ISs that were not damaged by the peeling were analyzed. IS-enriched retinas from Rho-GFP/+ mice were then immunolabeled with NbGFP-A647 using whole-retina immunolabeling and SIM as before. In this case, presumably due to the removal of many of the OSs in these retinas that would otherwise soak up all available immunoreagents, NbGFP-A647 fluorescence was now observed in the rod CC, IS, and ONL regions among the sparse remaining OSs in these sections (Fig 2E). WT IS-enriched retinas were also stained with NbGFP-A647 to demonstrate labeling specificity (Fig 2F).

Rhodopsin surrounds the distal appendages at the inner segment–connecting cilium junction Since we observed a continuous string of Rho fluorescence between the IS plasma membrane and into the CC membrane in mouse rods (Figs 3B and 5A), we next tested Rho molecule localization relative to the DAPs, which are the BB structures at the IS/CC interface in rods. The DAPs are 9 proteinaceous blades that project radially from the mother centriole at the base of the CC and extend to the plasma membrane [30]. In mammalian primary cilia, the DAPs were shown to function as binding sites for IFT (intraflagellar transport) proteins, which facilitate active ciliary transport, and to function in the gating of the ciliary GPCRs Smoothened and SSTR3 [81]. Here, an antibody targeting the DAP protein CEP164 was used to mark the position of the DAPs in rods, and in SIM images, the CEP164+ DAPs were localized as a well-defined line of fluorescence at the proximal end of the CC, which is the IS-CC interface (Fig 9A). Next, Rho-GFP/+ IS-enriched retinas were immunolabeled with NbGFP-A647 along with CEP164 and centrin antibodies to localize Rho at the DAPs using SIM and STORM. In these data, Rho-GFP molecules were again localized as continuous strings of fluorescence between the IS and CC plasma membranes but in a pattern surrounding the CEP164+ DAP blades (Fig 9B and 9C) with only a few STORM molecules colocalized with the CEP164+ DAPs. These results suggest that Rho is not integrated into the DAPs but is likely associated with the surrounding plasma membrane between the IS and CC. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 9. Rho localizes around the DAPs in mouse rods. (A) SIM z-projection images of a WT full retina section immunolabeling with a centrin antibody that marks the location of rod CC and the DC of the BB in white, along with a CEP164 antibody that labels the DAPs in green. The SIM retina image is also shown after 3D-deconvolution processing (SIM + 3D decon.), and a single rod cilium example is magnified. (B) SIM image from peeled Rho-GFP/+ mouse labeled with NbGFP-A647 (magenta), anti-CEP164 antibody (green), and anti-centrin antibody (white). A single rod is magnified, and the DAPs region is further magnified. The remaining OSs and the IS region are indicated. (C) STORM reconstruction single rod examples and magnified DAPs regions from Rho-GFP/+ IS-enriched retina immunolabeled with NbGFP-A647 (magenta) and a CEP164 antibody (green). Centrin-2 antibody labeling marks the position of the CC (white), and residual NbGFP-A647 signal labels the OS; both are captured as widefield fluorescence images. White arrows indicate the regions that are magnified in the adjacent image. Scale bar values match adjacent panels when not labeled. BB, basal body; CC, connecting cilium; DAP, distal appendage; DC, daughter centriole; NbGFP-A647, GFP nanobody Alexa 647 conjugate; IS, inner segment; OS, outer segment; Rho, rhodopsin; SIM, structured illumination microscopy; STORM, stochastic optical reconstruction microscopy; WT, wild-type. https://doi.org/10.1371/journal.pbio.3002467.g009

Golgi and post-Golgi trafficking proteins are localized at the inner segment plasma membrane in mouse rods We next tested the super-resolution localization of proteins that have previously been associated with the Rho secretory pathway in the rod IS. Rab11a, a post-Golgi small GTPase, was shown to mediate Rho trafficking in post-Golgi vesicles in frog rods [47,48] and was previously localized to the IS in mouse retinas in a semi-diffuse, puncta-like pattern [52,73,82]. Here, WT retina cryosections were immunolabeled with a specific Rab11a antibody [83], and Rab11a was localized with confocal imaging to the IS, in the ONL surrounding the nuclei, and in the OPL (Fig 10A). With SIM, Rab11a was localized as bright puncta throughout the OS, IS, and ONL, and in individual rod ISs, many Rab11a+ puncta were observed to be colocalized with the STX3+ IS plasma membrane (Fig 10B). After puncta counting, the rate of IS plasma membrane associated Rab11a+ puncta was 45.8% ± 14.3% (SD) per mouse rod IS (n = 955 puncta, n = 32 rods); the rest of the Rab11a+ puncta were internal/cytoplasmic. In STORM reconstructions, Rab11a molecules were localized into molecule clusters within the IS, which were isolated for visualization using a Voronoi tessellation clustering algorithm (Fig 10C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 10. Rho colocalizes with Rab11a in mouse rod ISs. (A) Z-projection of a WT retinal cryosection co-immunolabeled with centrin-2 and Rab11a antibodies and counterstained with DAPI. (B) SIM z-projection image of a WT retina co-immunolabeled with centrin-2, STX3, and Rab11a antibodies. The SIM retina image is also shown after 3D-deconvolution processing (SIM + 3D decon.), and Rab11a+ puncta are localized in the IS layer. A single rod example is magnified, and a threshold image of the Rab11a channel is shown pseudocolored to depict puncta that are localized at the STX3+ IS membrane hull as magenta, and puncta that are localized internally or at the CC as white. (C) STORM reconstruction of a single rod from a WT retina immunolabeled as in (B). A magnified region is shown, and in the adjacent image, Rab11a+ clusters identified with Voronoi tessellation (see Materials and methods) are in white. (D) Z-projection of a WT retinal cryosection co-immunolabeled with centrin-2 and GMII antibodies and counterstained with DAPI. (E) SIM images, including 3D decon. processed images, of a WT retina co-immunolabeled with centrin-2, STX3, and GMII antibodies. As in (B), a single rod example is shown, and in the adjacent image, membrane-localized GMII+ puncta are pseudocolored magenta, and internal (and ciliary) puncta are white. (F) Frequency plots for FWHM measurements of individual Rab11a and GMII IS-localized puncta from the SIM data represented in (B) and (E). For Rab11a FWHM values, n (number of puncta) = 67. For GMII FWHM values, n (number of puncta) = 58. (G) Confocal z-projection image of a WT retina cryosection immunolabeled with centrin and DYNC1H1 antibodies, as well as DAPI counterstaining. DYNC1H1+ fluorescence fills the IS layer. In the adjacent panel, a STORM reconstruction of a WT retina immunolabeled with centrin-2, STX3, and DYNC1H1 antibodies is depicted. A single rod example is shown. (H) STORM reconstruction of a WT retina immunolabeled with centrin-2, STX3, and rootletin antibodies. A single rod example is shown, and the ciliary rootlet is indicated. (I-M) STORM images of a (I) Rho-GFP/+ IS-enriched retina or (J) Rho-GFP-1D4/+ IS-enriched retina, each co-immunolabeled with NbGFP-A647, and centrin, STX3, and Rab11a antibodies. STORM reconstruction channels—NbGFP-A647 (magenta) and Rab11a (cyan)—are superimposed with the matching widefield fluorescence image of centrin/STX3 immunolabeling (combined, yellow). IS regions are indicated. A single rod example is shown, and the CC is indicated. In the adjacent image, Rab11a+ clusters identified with Voronoi tessellation are in white, and the STX3+ IS hull is outlined in yellow. A white arrow indicates a further magnified region of Rho-GFP molecules localized around Rab11a clusters; however, there was a relatively low degree of colocalization between Rho and Rab11a. Next, Rho-GFP was co-immunolabeled with (K) DYNC1H1 antibody or (L) Rootletin antibody. In both, the locations of the CC and the IS outline are indicated, and in (K) areas where Rho-GFP and DYNC1H1 molecules overlap are indicated with white arrowheads. In (L), the location of the ciliary rootlet is indicated. (M) STORM data from (I-L) conditions were used to perform Mosaic interaction analyses to test the colocalization between Rho-GFP molecules and the other immunolabeled target from the same rod IS. Interaction strength values are compared as violin plots (circles = median values and dashed lines = mean values). N values, corresponding to the number of rods from each condition, are Rab11a vs. Rho-GFP, n = 15; Rab11a vs. Rho-GFP-1D4, n = 11; DYNC1H1 vs. Rho-GFP, n = 10; Rootletin vs. Rho-GFP, n = 15. In all panels, white arrows indicate regions that are magnified. Scale bar values match adjacent panels when not labeled. Numerical values corresponding to all graphical data are provided in Table E in S1 Data. CC, connecting cilium; FWHM, full width half maximum; GMII, Golgi alpha-mannosidase II; IS, inner segment; NbGFP-A647, GFP nanobody Alexa 647 conjugate; Rho, rhodopsin; SIM, structured illumination microscopy; STORM, stochastic optical reconstruction microscopy; STX3, syntaxin 3; WT, wild-type. https://doi.org/10.1371/journal.pbio.3002467.g010 Next, although the GM130+ cis-Golgi was prominently localized in the myoid region of the IS (Fig 5B); the localization of Golgi alpha-mannosidase II (GMII), a medial/trans-Golgi glycoside hydrolase that was previously shown to process Rho protein in Golgi [84], was localized throughout the IS of mouse retinas using confocal microscopy (Fig 10D). In SIM images, GMII was also localized in discrete puncta within rod ISs, albeit less densely than Rab11a+ IS puncta (Fig 10E); however, more GMII+ puncta were colocalized with the IS plasma membrane. The rate of IS membrane associated GMII+ puncta was 73.8% ± 19.2% (SD) (n = 303 puncta, n = 27 rods). Using SIM with 3D deconvolution, the diameters (⌀) of single, isolated IS Rab11a+ and GMII+ puncta were measured. Based on diameter, Rab11a+ puncta were distributed in 2 groups: <200 nm ⌀ puncta (mean ⌀ = 126.7 nm ± 23.5 nm, n = 34 puncta) and >200 nm ⌀ puncta (mean ⌀ = 320.8 nm ± 33.8 nm, n = 34 puncta), while GMII+ puncta were normally distributed as 1 group (mean ⌀ = 119.6 nm ± 14.5 nm, n = 58 puncta) (Fig 10F). Together, these results indicate that vesicular Golgi and post-Golgi trafficking organelles are targeted to the IS plasma membrane in mouse rods. For comparison, cytoplasmic dynein-1 and the ciliary rootlet were also localized in the mouse IS with super-resolution fluorescence. Cytoplasmic dynein-1 is essential in rods as the putative motor complex for intracellular cytoplasmic transport. An antibody targeting the force-generating heavy chain of the dynein-1 complex, DYNC1H1, was used to localize dynein-1 throughout the entire rod IS layer, and with STORM, DYNC1H1 molecules were localized in a homogenous distribution throughout the IS (Fig 10G). An antibody targeting rootletin, the core protein of the ciliary rootlet, was used to reconstruct the rootlet in rods with STORM (Fig 10H).

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002467

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.