EFA6A regulates retinal function through the control of photoreceptor cell activity and structure

TL;DR

EFA6A is crucial for retinal function by controlling photoreceptor cell activity and structure, and its absence causes vision-related defects.

Contribution

EFA6A's role in regulating retinal function and photoreceptor structure is newly identified.

Findings

EFA6A depletion in the retina causes morphological and functional defects in photoreceptors.

EFA6A is essential for retinal pigment epithelium phagocytic activity.

Loss of EFA6A disrupts retinal homeostasis and neuronal organization.

Abstract

The initial step in visual perception, i.e., the collection of light, is carried out by the outer segments of photoreceptor cells. These outer segments, specialized primary cilia, house the photopigments that absorb photons. Here, we report that EFA6A, an exchange factor for the small G protein Arf6, previously described as a regulator of early ciliogenesis by promoting distal appendage vesicle fusion, is essential for visual function. We demonstrate that EFA6A is present in various layers of the mouse retina, including the photoreceptors and the retinal pigment epithelium (RPE). Accordingly, depletion of EFA6A in the mouse retina leads to both morphological and functional defects in photoreceptors, resembling the phenotypes observed in retinal ciliopathies. We also show that EFA6A depletion in RPE cells severely impairs their phagocytic activity. Thus, our results point out a key role for EFA6A in the control of visual activity.

•EFA6A contributes to the regulation of retinal function•Loss of EFA6A alters retinal physiological responses•EFA6A impacts retinal neuronal organization•EFA6A is required for retinal homeostasis

EFA6A contributes to the regulation of retinal function

Loss of EFA6A alters retinal physiological responses

EFA6A impacts retinal neuronal organization

EFA6A is required for retinal homeostasis

Molecular biology; Sensory neuroscience; Cell biology

Figures

Localization of EFA6A in the mouse retina(A) Immunolabelling of wild-type retinas using an anti-EFA6A antibody revealed specific EFA6A expression in the RPE, photoreceptors (both inner and outer segments), certain cells in the inner nuclear and plexiform layers, and the ganglion cell layer. Incubation with the antibody preabsorbed with purified EFA6A protein served as a negative control and confirmed labeling specificity. Scale bar 20 μm.(B) *In situ* hybridization using an EFA6A antisense probe, with a sense probe as control, showed specific labeling in the ganglion cell layer, inner and oute

Morphological changes resulting from EFA6A depletion in the retina(A) Morphology of retinas injected with control siRNA or EFA6A siRNA. Semi-thin retinal sections (1 μm), counterstained with methylene blue-Azur II, show normal retinal architecture in si-control (a) and disrupted morphology in si-EFA6A#1 (b) retinas. Scale bars, 20 μm.(B) Transmission electron micrographs of photoreceptor outer segments (OSs) from si-control (a) and si-EFA6A#1 (b) retinas. si-EFA6A treatment results in a loss of well-organized OS and signs of degeneration. Scale bar 2 μm.(C) Mislocalization of rhodopsin in EFA6

Effect of EFA6A depletion on scotopic luminance-response function(A) Representative dark-adapted ERG traces recorded at flash intensities ranging from −4.72 to 1.87 log scot td/s in untreated controls, control siRNA, si-EFA6A#1, and si-EFA6A#2 groups, 7 days post-injection. Each trace represents the average response of at least nine mice.(B) Intensity-response curves of a-wave amplitudes in control (black squares), control siRNA (gray inverted triangles), si-EFA6A#1 (black circles), and si-EFA6A#2 (gray triangles) groups. Data are presented as mean ± SD (*n* = 3); ∗∗∗*p* < 0.001.(C) Intensity-

Effect of EFA6A depletion on RPE functions(A and B) Representative ERG c-wave recordings at different flash intensities in untreated controls, control siRNA-, si-EFA6A#1-, and si-EFA6A#2-injected mice at day 7 post-injection. Each trace represents the average response of at least nine mice. ERG results were analyzed by two-way ANOVA. Bonferroni post hoc testing was used to evaluate amplitude differences among intensities. Data were considered significant if *p* < 0.05.(C) Phagocytosis assay in ARPE19 cells treated with control siRNA or si-EFA6A pool for 48 h, followed by a 6 h incubation with

Full text

Introduction

The retina is responsible for converting light signals from the environment into neural signals that the brain can interpret. Vision begins when photons are captured by the outer segments (OSs) of photoreceptor cells, highly specialized primary cilia (PC) that convert light into electrical signals through a process known as phototransduction.1^,^2 These dynamic cilia undergo constant renewal to balance the shedding and phagocytosis of their distal tips by the retinal pigment epithelium (RPE). Key proteins, including the light-sensitive receptor rhodopsin, are continually transported from the inner segment (IS) to the OS via the connecting cilium (CC). Most ciliopathies, such as Joubert, Meckel, Senior-Løken, and Bardet-Biedl syndromes, lead to retinal degeneration, often alongside systemic manifestations such as renal cystic disease, obesity, and mental retardation.3 The genes responsible for these retinal disorders also function in primary cilia across other tissues, explaining their pleiotropic effects.

The EFA6 family comprises four tissue-specific isoforms, each encoded by a distinct gene, that function as nucleotide exchange factors (GEFs) for the small G protein Arf6. All EFA6 proteins share a Sec7 domain responsible for their GEF activity, as well as a common C-terminal region, but differ markedly in their N-terminal domains. Despite their high homology and frequent overlapping expression, whether the EFA6 isoforms have distinct functional roles remains unclear. EFA6A is predominantly expressed in the colon, small intestine, and brain. It regulates intracellular trafficking of various cargoes, including GPCRs, the transferrin receptor, ion channels4^,^5^,^6^,^7 and directs vesicle transport to form apical lumens in mammary epithelial cells. Through interactions with F-actin and α-actinin, EFA6A also participates in remodeling the cortical actin cytoskeleton,8^,^9^,^10 which likely underpins its role in neurite formation and dendritic and axonal regeneration. Indeed, earlier work identified EFA6A as a key regulator of dendritic branching and spine formation.11^,^12

Despite earlier reports of its abundant expression in the mouse eye13 and its presence in the outer plexiform layer,14 the role of EFA6A in retinal function remained largely unexplored. Proteomic analyses previously identified EFA6A in the photoreceptor sensory cilium complex,15 further suggesting a potential role in retinal ciliary structures. Recently, we demonstrated that EFA6A plays a critical role in the early stages of primary cilium assembly by regulating the fusion of distal appendage vesicles to form the ciliary vesicle.16 Ciliogenesis is a polarized, tightly regulated process that begins with the recruitment of pre-ciliary vesicles at the distal appendages of the mother centriole, followed by their fusion into a ciliary vesicle.17 Loss of EFA6A leads to the accumulation of Arl13B-positive vesicles around the distal appendages and blocks ciliogenesis.16 Arl13B, a small G protein that regulates membrane trafficking, ciliation, and cilia length, is essential for photoreceptor development, function, and maintenance; mutations in ARL13B are associated with Joubert syndrome, a ciliopathy characterized by vision loss due to photoreceptor defects.18

Given that photoreceptors are highly specialized primary cilia assembled through conserved molecular mechanisms, we investigated the role of EFA6A in photoreceptor function. Here, we report that EFA6A is an essential regulator of visual function. In the mouse retina, EFA6A controls both the morphology and electrical activity of photoreceptor OS. Moreover, EFA6A is required for the phagocytic activity of RPE cells.

Together, our findings establish EFA6A as a key player in visual activity, acting through the control of photoreceptor function and maintenance.

Results

EFA6A localizes to retinal and retinal pigment epithelium cells in the mouse retina

EFA6A is highly enriched in the brain, yet its function in neurons remains incompletely understood.5 Here, we investigated the presence and potential role of EFA6A in the mouse retina. Immunohistochemistry (IHC) analysis revealed EFA6A protein expression in RPE cells, the OS and IS of photoreceptors, as well as in some cells within the inner nuclear layer (INL), inner plexiform layer (IPL), and ganglion cell layer (GCL) (Figure 1A; see Figure S1A for anti-EFA6A antibody validation). In situ hybridization using an antisense oligonucleotide probe further confirmed the presence of EFA6A mRNA in the GCL, INL, outer nuclear layer (ONL), and, to a lesser extent, the RPE (Figure 1B). Notably, a signal is also visible in the RPE in the sense probe control, which most likely represents background rather than specific labeling, consistent with the high melanin content of RPE cells. This observation has also been documented previously.19 These findings indicate that EFA6A is expressed across multiple retinal layers, including photoreceptors and the RPE. To further characterize EFA6A localization, we performed co-immunolabelling with specific markers for distinct retinal cell types. EFA6A co-localized with Peanut Agglutinin (PNA), a marker for cone OS and IS, confirming its presence in these structures (Figure 1C). Co-labelling with anti-rhodopsin antibodies revealed EFA6A expression in rod OS (Figure 1C). In the INL, EFA6A co-localized with anti-calbindin D-28K in horizontal cells (soma and processes) and with anti-PKCα in bipolar cells (soma and dendrites) (Figure S1B). EFA6A expression was also detected in the soma and dendrites of some amacrine cells, as shown by co-labelling with anti-syntaxin antibodies (Figure S1B). In the GCL, EFA6A expression was low in the soma of ganglion cells (marked by anti-Thy1.2), but higher in their dendritic processes (Figure S1B). Finally, EFA6A co-localized with RPE65 in RPE cells (Figure 1C).Figure 1. Localization of EFA6A in the mouse retina(A) Immunolabelling of wild-type retinas using an anti-EFA6A antibody revealed specific EFA6A expression in the RPE, photoreceptors (both inner and outer segments), certain cells in the inner nuclear and plexiform layers, and the ganglion cell layer. Incubation with the antibody preabsorbed with purified EFA6A protein served as a negative control and confirmed labeling specificity. Scale bar 20 μm.(B) In situ hybridization using an EFA6A antisense probe, with a sense probe as control, showed specific labeling in the ganglion cell layer, inner and outer nuclear layers, and RPE. Diaminobenzidine (DAB) was used for signal detection. Scale bar 20 μm.(C) Confocal immunofluorescence microscopy demonstrates the co-localization of EFA6A with various retinal markers. Co-labelling with peanut agglutinin (upper panels) indicates EFA6A presence in the inner and outer segments of cone photoreceptors (arrows). Co-labelling with anti-rhodopsin antibodies (middle panels) revealed EFA6A expression in the outer segments, particularly in the CC (arrows). Co-labelling with anti-RPE65 antibodies (lower panels) confirmed EFA6A expression in the RPE (arrows). Scale bars, 50 μm (upper panels), 25 μm (middle and lower panels).

In summary, EFA6A is expressed in multiple retinal cell types, including cone and rod OS and IS, RPE, and specific interneurons across the INL and GCL.

Depletion of EFA6A leads to morphological alterations in the mouse retina

To investigate the role of EFA6A in retinal morphology and ultrastructure, we depleted EFA6A using intravitreal injection of small interfering RNA (siRNA). The efficiency of EFA6A knockdown was confirmed by Western blotting and RT-PCR (Figures S2A and S2B). Additionally, immunohistochemical analysis of Cy3-labeled EFA6A-siRNA 48 h post-injection showed broad distribution across retinal layers, including robust uptake by both rod and cone photoreceptors, as demonstrated by co-labeling with rhodopsin and cone opsin, respectively (Figure S2C upper and lower panels). Light microscopy of retinal sections from EFA6A-siRNA treated mice revealed severe retinal detachment (Figure 2A, panel b, white arrowhead), along with the loss of some OS (Figure 2A, panel b, black arrowhead) and a reduction in nuclei within the ONL (ONL; Figure 2A, panel b vs. a), indicating that EFA6A depletion effectively reached the ONL and induced specific structural defects. Electron microscopy (EM) further confirmed profound alterations in photoreceptor OS architecture at 7 days post-injection. In contrast to the organized, tightly packed OS of control retinas, the OS in EFA6A-depleted retinas appeared disorganized and loosely aligned (Figure 2B, panel b vs. a), with a disrupted interface between OS and RPE (Figure S2D, panels a, b). Large vacuole-like holes were also evident within the OS layer (Figure 2B, panel b, white arrowhead), suggesting defective OS growth. These structural defects were accompanied by alterations in rhodopsin labeling, as shown by immunostaining (Figure 2C, panel b vs. a). Rhodopsin distribution appeared less uniform, suggesting potential impairment in trafficking from the inner segment (IS) to the outer segment (OS). This altered distribution may be associated with disorganization of the connecting cilium (CC), which connects the IS and OS (Figure 2B, panel b, black arrowheads). The CC appeared less distinct in EFA6A-depleted tissue compared to controls (Figure 2B, panel b vs. a). Supporting these observations, acetylated tubulin staining (normally enriched in the CC and RPE) was significantly reduced and disorganized in EFA6A-depleted retinas (Figure 2D, panel b vs. a). Quantification of acetylated tubulin signal distribution revealed that, in control retinas, approximately 64% of the signal localized to the IS + OS region (which contains the CC), whereas in EFA6A-siRNA-treated retinas this proportion dropped to an average of ∼19%, consistent with a substantial loss of CC integrity (see STAR Methods). Ultrastructural analysis of the ONL revealed degenerating rod and cone photoreceptor cell bodies in EFA6A-siRNA-treated retinas compared to controls (Figure S2D, panels c, d), indicating that EFA6A depletion leads to photoreceptor degeneration.Figure 2. Morphological changes resulting from EFA6A depletion in the retina(A) Morphology of retinas injected with control siRNA or EFA6A siRNA. Semi-thin retinal sections (1 μm), counterstained with methylene blue-Azur II, show normal retinal architecture in si-control (a) and disrupted morphology in si-EFA6A#1 (b) retinas. Scale bars, 20 μm.(B) Transmission electron micrographs of photoreceptor outer segments (OSs) from si-control (a) and si-EFA6A#1 (b) retinas. si-EFA6A treatment results in a loss of well-organized OS and signs of degeneration. Scale bar 2 μm.(C) Mislocalization of rhodopsin in EFA6A-depleted retinas. Immunohistochemistry of retinal cryosections using anti-rhodopsin antibodies (green) and Peanut Agglutinin (PNA, red) shows correct rhodopsin localization in si-control (a) and its mislocalization in si-EFA6A#1 (b) retinas. DAPI (blue) stains nuclei. Abbreviations: RPE, retinal pigment epithelium; OS, outer segment; IS, inner segment; ONL, outer nuclear layer; CC, connecting cilium. Scale bars, 20 μm.(D) Mislocalization of acetylated tubulin in EFA6A-depleted retinas. Immunohistochemistry with anti-acetylated α-tubulin antibodies (green) in si-control (a) and si-EFA6A#1 (b) cryosections. Nuclei are counterstained with DAPI (blue). Scale bars, 20 μm.(E) Schematic representation summarizing the morphological alterations observed in si-EFA6A-treated retinas.

In summary, EFA6A depletion in the mouse retina causes profound morphological defects, particularly in the OS and RPE layers. These include OS loss and degeneration, rhodopsin mislocalization, disrupted CC architecture, and photoreceptor cell death (Figure 2E for illustration). Together, our results demonstrate that EFA6A plays a critical role in organizing the CC and maintaining OS structure in photoreceptor cells.

Depletion of EFA6A leads to functional alterations of the mouse retina

The observed structural defects in EFA6A-depleted retinas prompted us to assess whether these changes were associated with functional impairments in retinal physiology. To this end, we recorded electroretinograms (ERGs) seven days after siRNA injection, measuring responses under scotopic (rod-dominated) and photopic (cone-dominated) conditions (Figures 3 and 4, respectively). We compared untreated controls, control siRNA-injected retinas, and two distinct EFA6A-targeting siRNAs.Figure 3. Effect of EFA6A depletion on scotopic luminance-response function(A) Representative dark-adapted ERG traces recorded at flash intensities ranging from −4.72 to 1.87 log scot td/s in untreated controls, control siRNA, si-EFA6A#1, and si-EFA6A#2 groups, 7 days post-injection. Each trace represents the average response of at least nine mice.(B) Intensity-response curves of a-wave amplitudes in control (black squares), control siRNA (gray inverted triangles), si-EFA6A#1 (black circles), and si-EFA6A#2 (gray triangles) groups. Data are presented as mean ± SD (n = 3); ∗∗∗p < 0.001.(C) Intensity-response curves of dark-adapted b-wave amplitudes in the same experimental groups and conditions as in (B). Mean ± SD values are shown.(D) Quantification of the maximum saturated b-wave amplitude (Vmax) for untreated (black squares), control siRNA (green squares), si-EFA6A#1 (blue triangles), and si-EFA6A#2 (purple triangles) groups at day 7 post-injection.(E) Dark-adapted b/a-wave ratios recorded at flash intensities from −1.12 to 1.87 log scot td/s in untreated (black squares), control siRNA (purple inverted triangles), si-EFA6A#1 (red circles), and si-EFA6A#2 (blue triangles) groups. Values are presented as mean ± SD (n = 3); p < 0.001. Statistical analysis of ERG data was performed with statistical software (Prism 4; GraphPad). ERG results were analyzed by two-way ANOVA. Bonferroni post hoc testing was used to evaluate amplitude differences among intensities. Data were considered significant if p < 0.05.Figure 4. Effect of EFA6A depletion on photopic luminance-response function(A) Representative light-adapted ERG traces recorded at flash intensities ranging from −2.32 to 1.87 log scot td/s in untreated, control siRNA-, si-EFA6A#1-, and si-EFA6A#2-injected mice at day 7 post-injection. Each trace represents the average response of at least nine mice.(B) Intensity-response curves of a-wave amplitudes from untreated mice (black squares), control siRNA-injected (gray inverted triangles), si-EFA6A#1-injected (black circles), and si-EFA6A#2-injected (gray triangles) retinas. Data are presented as mean ± SD (n = 3; ∗∗∗p < 0.001).(C) Intensity-response curves of b-wave amplitudes under photopic conditions for the same groups as in (B). Data are shown as mean ± SD.(D) Quantification of the maximum saturated b-wave amplitude (Vmax) for untreated (black squares), control siRNA (green squares), si-EFA6A#1 (blue triangles), and si-EFA6A#2 (purple triangles) groups at day 7 post-injection.(E) Light-adapted b/a-wave amplitude ratios measured at flash intensities ranging from 0.47 to 1.87 log scot td/s in untreated mice (black squares), control siRNA-injected (red inverted triangles), si-EFA6A#1-injected (red circles), and si-EFA6A#2-injected (blue triangles). Values are presented as mean ± SD. Statistical analysis of the ERG data was performed with statistical software (Prism 4; GraphPad). ERG results were analyzed by two-way ANOVA. Bonferroni post hoc testing was used to evaluate amplitude differences among intensities. Data were considered significant if p < 0.05.

As expected, no significant differences were observed between the untreated and control siRNA-injected groups. In contrast, retinas treated with EFA6A-targeting siRNAs displayed a pronounced reduction in ERG responses under both scotopic and photopic conditions (Figures 3A and 4A).

Under scotopic conditions and in response to high-intensity stimuli (1.87 log scot td·s), the a-wave amplitude—primarily representing rod photoreceptor activity—was markedly reduced to 38% and 45% of control levels in siRNA#1 and siRNA#2-treated groups, respectively, at day 7 post-injection (Figure 3B). This decrease aligns with the structural degeneration and dysfunction observed in the photoreceptor layer (Figure 2). The b-wave amplitude, which reflects bipolar cell responses, was also significantly reduced following EFA6A depletion (Figure 3C). The maximum (saturated) scotopic b-wave amplitude (Vmax) decreased by approximately 40% (siRNA #1) and 43% (siRNA #2) compared to controls (p < 0.001; Figure 3D). Importantly, the b/a-wave ratio remained unchanged across all groups (Figure 3E), suggesting that the primary defect lies within the photoreceptors rather than secondary to broader retinal degeneration.

Similarly, under photopic conditions and high-intensity stimuli (1.87 log scot troland·s), the a-wave amplitude was significantly reduced to 31% and 27% of control levels for siRNA #1 and #2, respectively (p < 0.001; Figure 4B). The b-wave amplitude also showed a significant reduction (Figure 4C). The maximum (saturated) photopic b-wave amplitude (Vmax) decreased to 27% and 30% of control levels (p < 0.001; Figure 4D). As in the scotopic analysis, the photopic b/a-wave ratio remained unaffected, with no evidence of a negative-type ERG (b/a ratio <1; Figure 4E). This further confirms that the observed deficits primarily result from photoreceptor dysfunction rather than downstream retinal circuitry alterations.

In conclusion, our results demonstrate that EFA6A is essential for maintaining both the morphological integrity and functional activity of photoreceptors in vivo.

EFA6A is required for the phagocytosis of photoreceptor outer segment by retinal pigment epithelium cells

Alongside the impairment of the a- and b-waves, we observed that the c-wave originating from the hyperpolarization of the RPE apical membrane in response to decreased subretinal K^+^ concentration, was also significantly affected in EFA6A-depleted retinas (Figure 5A). c-wave recordings across various light intensities revealed a substantial reduction in amplitude in siEFA6A-treated groups compared to controls at day 7. For instance, at high-intensity stimuli (0.87 log scot troland·s), the c-wave amplitude was reduced to 50% (siEFA6A#1) and 54% (siEFA6A#2) of control levels (Figure 5B). No significant differences were detected between untreated and control siRNA-injected retinas (Figure 5A). These findings indicate that EFA6A depletion induces RPE dysfunction and/or degeneration.Figure 5. Effect of EFA6A depletion on RPE functions(A and B) Representative ERG c-wave recordings at different flash intensities in untreated controls, control siRNA-, si-EFA6A#1-, and si-EFA6A#2-injected mice at day 7 post-injection. Each trace represents the average response of at least nine mice. ERG results were analyzed by two-way ANOVA. Bonferroni post hoc testing was used to evaluate amplitude differences among intensities. Data were considered significant if p < 0.05.(C) Phagocytosis assay in ARPE19 cells treated with control siRNA or si-EFA6A pool for 48 h, followed by a 6 h incubation with FITC-labelled latex beads opsonized with photoreceptor outer segment (POS) fragments. Cells were subsequently fixed and stained for F-actin using phalloidin (red). Scale bar 10 μm.(D) Quantification of POS-opsonized bead uptake as described in (C). Results are presented as mean ± SD from four independent experiments, each analyzing 40–250 cells per treatment condition (n = 4; unpaired two-tailed Student’s t test; Data were considered significant if p < 0.05; ∗∗∗p < 0.001).

Given that one of the critical functions of the RPE is the daily phagocytosis of shed photoreceptor outer segment (POS) tips essential for maintaining photoreceptor homeostasis, we next investigated whether EFA6A is involved in this process. To assess phagocytic capacity, we used the human RPE cell line ARPE-19, a widely accepted model for RPE phagocytosis studies, and quantified the internalization of POS fragment-opsonized fluorescent beads as described by Klettner et al.20 Strikingly, siRNA-mediated depletion of EFA6A led to a profound reduction in bead internalization, demonstrating that EFA6A is essential for RPE phagocytic activity (Figures 5C and 5D).

Taken together, these results reveal a critical role for EFA6A in visual function, extending beyond photoreceptor integrity to include the phagocytic activity of the RPE. EFA6A is thus a key regulator of the interdependent processes that sustain retinal homeostasis and vision.

Discussion

In this study, we set out to define the localization and functional roles of EFA6A in the retina, expanding on previous research that primarily focused on its neuronal functions in the brain. Here, we provide the first functional evidence that EFA6A is essential for vision and retinal integrity.

Our data demonstrate that EFA6A is widely expressed in the mouse retina, including in photoreceptors and RPE cells. Loss of EFA6A function resulted in severe retinal defects, characterized by dramatically reduced ERG responses implicating both rod and cone photoreceptors as primary sites of dysfunction. While we cannot fully exclude contributions from other EFA6A-expressing retinal cell types, the combined electrophysiological, morphological, and ultrastructural analyses highlight a central role for EFA6A in photoreceptor and ciliary function.

Morphologically, EFA6A depletion disrupted the organization of the photoreceptor layer, leading to OS loss, altered rhodopsin distribution, and disorganization of rod outer segments characterized by enlarged interdiscal spaces and reduced disc compaction and alignment. These phenotypes are consistent with defects in lipid and protein trafficking toward the OS, in line with EFA6A’s established role in regulating Arf6-mediated vesicle trafficking and actin cytoskeleton remodeling^l^. In addition to trafficking defects, the shortening of the OS likely reflects the combined impact of impaired RPE phagocytic activity and perturbed membrane delivery. Reduced phagocytosis can disrupt the daily renewal cycle of the OS, leading photoreceptors to limit the production of new discs to prevent excessive accumulation.19 Concurrently, defects in membrane and cargo transport from the inner segment may further compromise disc biogenesis, amplifying the OS-shortening phenotype.21 Notably, we also observed a striking loss of acetylated tubulin labeling and ultrastructural abnormalities in the CC, implicating EFA6A in the assembly and stability of this critical structure. Given that the CC serves as the conduit for transporting essential proteins such as rhodopsin to the OS, defects in ciliary integrity likely underlie the progressive photoreceptor degeneration observed in the absence of EFA6A.

Importantly, our findings align with and extend previous studies on ARL13B, a key ciliary protein. Dilan and colleagues demonstrated that ARL13B is critical for OS development and retinal morphogenesis,18 while we identified EFA6A as an interactor of ARL13B.16 This functional link between EFA6A and ARL13B suggests that EFA6A may contribute to ARL13B-dependent ciliary processes, including the maintenance of ciliary architecture and rhodopsin trafficking. Disruption of this interaction may explain the severe ciliary phenotypes observed in our EFA6A-deficient retinas.

Beyond the retina, EFA6A is increasingly recognized as a key regulator of cell polarity across diverse systems, including neurons (dendritic and axonal specification),12^,^22 epithelial cells (tight junction formation and apical-basal polarity),23^,^24 and primary cilia assembly.16 These broad roles underscore the importance of EFA6A as a multifunctional scaffold protein coordinating vesicle trafficking, cytoskeletal dynamics, and polarity establishment.

Our findings thus position EFA6A as a critical node in retinal and ciliary homeostasis, with broader implications for human disease. Given its fundamental role in ciliary assembly and maintenance, we propose that dysregulation of the PSD gene, which encodes EFA6A, could contribute to a wide range of ciliopathies. While individual ciliopathies are rare (e.g., ∼1 in 1,000 for autosomal dominant polycystic kidney disease, ∼1 in 150,000 for certain syndromic forms), they collectively represent a significant public health burden.25

Mutations in PSD or altered regulation of EFA6A protein levels could underlie pleiotropic disorders affecting multiple organ systems, including the retina, kidney, and brain. Furthermore, given the evolutionary conservation of ciliary machinery, our results raise the possibility that EFA6A dysfunction may contribute to unexplained cases of retinal degeneration, neurodevelopmental disorders, and systemic ciliopathies.

In summary, our study reveals that EFA6A is a critical component of the retinal ciliary machinery, essential for the proper assembly of the connecting cilium, the delivery of proteins such as rhodopsin to the OS, and the maintenance of photoreceptor and RPE function. These findings highlight EFA6A’s broader role in cell polarity and vesicular trafficking and suggest that aberrations in its expression or function could underlie a spectrum of ciliopathies. Future work should focus on identifying PSD mutations in patients with inherited retinal diseases, elucidating the systemic impact of EFA6A dysfunction, and exploring therapeutic strategies to modulate EFA6A pathways in ciliopathies and related disorders.

Limitations of the study

This study identifies EFA6A as a regulator of retinal function. However, several limitations should be considered. First, our analyses primarily relies on the loss of function/mRNA depletion/animal model, which does not allow us to fully discriminate between developmental and adult-specific roles of EFA6A in the retina. Second, while our data support a functional impact of EFA6A on retinal physiology, the precise molecular mechanism and downstream effectors mediating these effects remain to be fully elucidated. In particular, whether EFA6A acts through Arf6-dependent trafficking pathways or additional signaling networks in specific retinal cell types requires further investigation. Finally, although our functional assays revealed altered retinal responses, the broader relevance of these findings to human retinal physiology and disease will require validation in complementary models. Future studies combining cell-type-specific manipulations and mechanistic approaches will be necessary to address these limitations.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Michel Franco (franco@ipmc.cnrs.fr).

Materials availability

This study did not generate new unique reagents.

Data and code availability

This study did not generate new datasets or code. All data supporting the findings of this study are available within the article and its supplemental information.

Acknowledgments

We would like to express our sincere gratitude to Dr. M. Ettaiche for his expertise and his help in ERG experiments. We thank M. Partisani for her technical assistance and Dr. Hiroyuki Sakagami (Kitasato University, Japan) for the EFA6B specific anti-sera. This work is supported by the 10.13039/501100004794Centre National de la Recherche Scientifique. ChatGPT was used to improve language style. We acknowledge the Electron Microscopy facility CCMA (Center Commun de Microscopie Appliquée) from the Université Côte d’Azur, the flow cytometry or microscopy facility from the « Institut de Pharmacologie Moléculaire et Cellulaire », all part of the « Microscopie Imagerie Cytométrie Azur » GIS IBiSA labeled platform. We acknowledge SABLESPlatfomes financed by the 10.13039/501100000780European Union through the 10.13039/501100008530European Regional Development Fund.

This work is dedicated to Dr T. Léveillard who kindly encouraged and advised us during the realization of this project.

Author contributions

S. A., S. P., F. B., M. P., V. S., S. L-G., C. R., A. C., L. D-C., and M.F. performed research. S. A., S. P., M. P., V. S., S. L-G., C. R., C. D., F. L., and M. F. analyzed data. M.F. designed the research. M. F. wrote the manuscript.

Declaration of interests

The authors declare no competing interest.

STAR★Methods

Key resources table

REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesMouse monoclonal anti-acetylated TubulinSigma-AldrichCat# T7451; RRID:AB_609894Mouse monoclonal anti-SyntaxinSigma-AldrichCat# S0664; RRID: AB_477019Mouse monoclonal anti-PKCalphaSigma-AldrichCat# P5704; RRID: AB_477375Mouse monoclonal anti-Calbindin D-28KSigma-AldrichCat# C9848; RRID:AB_476894Mouse monoclonal anti-RhodopsinMilliporeCat# MAB5356; RRID:AB_2178961Mouse monoclonal anti-Arf6 (clone 8A6-2)Gift from S. BourgoinN/ARabbit anti-RPE65Novus BiologicalsCat# NB100-355; RRID:AB_10003038Goat anti-Peanut Agglutinin (PNA)Vector LaboratoriesCat# RL-1072; RRID: AB_2336873Rabbit anti-OpsinSigma-AldrichCat# AB_5407; RRID: AB_177456Rat Monoclonal anti-Thy1.2AbcamCat# ab91082; RRID:AB_2040590Purified rabbit anti-EFA6BGift from Dr. SakagamiN/APurified rabbit anti-EFA6A (C-terminal peptide)EurogentecCustom antibodyDonkey anti-mouse Alexa Fluor secondary antibodiesJackson ImmunoResearchCat# varies; RRID variesDonkey anti-rabbit Alexa Fluor secondary antibodiesJackson ImmunoResearchCat# varies; RRID variesAlexa Fluor–conjugated secondary antibodiesInvitrogenCat# varies; RRID variesChemicals, peptides, and recombinant proteinsEpoxy embedding resin (EPON 812)Electron Microscopy SciencesCat# L14120FITC-Labelled latex beads 1 μm diameter)Sigma-AldrichCat# L1030DAPISigma-AldrichCat# D9542Fluorescent PhalloïdinInvitrogen (Molecular Probes)Cat# A22283Experimental models: Cell linesBHK-21 cellsATCCRRID:CVCL_1915ARPE-19 cellsATCCCat# CRL-2302RRID:CVCL_0145Experimental models: Organisms/strainsMouse: C57BL/6JCharles River/Jackson LaboratoryRRID:IMSR_JAX:000664OligonucleotidesMouse EFA6A forward primer 5′atctctgttgcgcccc3′This studyN/AMouse EFA6A reverse primer 5′gggcggcggaagccctga3′This studyN/AsiRNA targeting human EFA6A (SMARTpool)DharmaconCat#M-010155Control siRNASigma-AldrichCat#SIC001antisense oligonucleotide EFA6A5′gcaggtagaggatcatcgccttgaggatcccgtggaagctcttcc3′This studyN/Asense oligonucleotide unrelated to EFA6A as negativecontrol5′ggaagagcttccacgggatcctcaagggcatgatcctctacctgc3′This studyN/AsiRNA targeting mouse EFA6A#1ccaagugggaauucuuEurogentecSequence providedsiRNA targeting mouse EFA6A#1ggugcuaccgagagacEurogentecSequence providedCy3-labeled siRNA targeting mouse EFA6A#1Sigma-AldrichCustomRecombinant DNAvsv-g-EFA6A plasmidPreviously describedN/AEGFP-EFA6B plasmidPreviously describedN/ACritical commercial assaysSuperScript III Reverse TranscriptaseInvitrogenCat#18080044HotStartTaqDNA PolymeraseQ-BiogenN/ASoftware and algorithmsPrism 4GraphPad Softwarehttps://www.graphpad.comAdobe AcrobatAdobe Systemshttps://www.adobe.com/productsLAS AF softwareLeica Microsystemshttps://www.leica-microsystems.comZen sofwareZeisshttps://www.zeiss.com/Microsoft WordMicrosofthttps://www.microsoft.com/wordMicrosoft ExcelMicrosofthttps://www.microsoft.com/excelImageJ/FijiNIHRRID:SCR_002285iTEM imaging softwareOlympus Soft Imaging Solutionshttps://www.olympus-lifescience.comUTAS Big-Shot softwareLKC TechnologiesN/AOtherRNAiMAX transfection reagentInvitrogenCat#13778150jetPEI transfection reagentPolyplus-transfectionCat#101-10NHamilton microsyringe (34G)HamiltonCat#65460-02Confocal Laser MicroscopeZeissLSM-780Confocal Laser MicroscopeLeicaSP5Transmission electron microscope (JEM-1400)JEOLJEM-1400Morada digital cameraSIS/OlympusMorada

Experimental model and study participant details

Animals

Wild-type C57BL/6 mice (male and female, 4 months old) were usd for all in vivo experiments. Animals were housed under standard conditions with a 12h light/dark cycle and had ad libitum access to food and water. All experiment procedures were performed in accordance with institutional guidelines and were approved by the appropriate animal care and use committee. (Reference number of the project: 00537.01; Agreement number: C061525 -16/04/2015).

Cell lines

Baby hamster kidney cells (BHL-21) and human retinal pigment epithelial cells (ARPE-19) were used. BHK-21 cells were maintained in BHK-21 medium supplemented with 5% fetal calf serum (FCS), 10% tryptose phosphate broth, 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin. ARPE-19 cells were cultured in DMEM/F12 (1:1). All cells were maintained at 37°C in a humidified incubator with 5% CO_2_.

Cultured retinal pigment epithelium (ARPE-19) cells were used for in vitro phagocytosis assays. Cell line authentication was not performed. Cells were routinely tested and found negative for mycoplasma contamination.

Method details

Intravitreal injection of siRNA

Mice were anesthetized by intraperitoneal injection of ketamine/xylazine (ketamine, 100–125 mg/kg; xylazine, 10–12.5 mg/kg). Intravitreal injections were performed using a Hamilton syringe equipped with a blunt 34-gauge needle. One microliter of siRNA solution was slowly (3–5 s) into the vitreous chamber following pressure equilibration.

Transfection reagents (Transit-TKO, Mirus Bio; jetPEI, Polyplus-transfection) were tested for intravitreal delivery but did not improve knockdown efficiency or phenotypic outcome compared to siRA injection alone. Therefore, siRNAs were delivered without transfection reagents.

Cell culture and transfection

siRNA transfections in ARPE-19 and BHK cells were performed using RNAiMAX (Invitrogen) according to the manufacturer’s instructions. Cells were analyses 48h post-transfection. cDNA transfections in BHK cells were performed using JetPEI(Polyplus Transfection) and analyses after 24h.

In situ hybridization, immunochemistry and double labeling

Mice were dark-adapted overnight, anesthezised, and euthanized. Eyes were enucleated, punctured at the limbus, and perfused with ice-cold 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). After 30 min of fixation, the cornea, lens and vitreous were removed. Eyecups were cryoprotected in 20% sucrose in PBS for 1h, embedded in Tissue-Tek OCT compound (Sakura Finetek), and snap-frozen in liquid nitrogen-cooled isopentane. Cryosections (14 μm) where then prepared.

For in situ hybridation, DIG-labelled antisense oligonucleotide probes targeting EFA6A were used with a DIG-labelled sens probe serving as a negative control. Probes were labeled at the 3′ end using terminal deoxynucleotidyl transferase. Retinal sections were post-fixed for 20 min in 4% PFA, permeabilized with 0.1% tween 20 in PBS, and washed three times in PBS. Sections were acetylated for 10 min in 0.25% acetic anhydride prepared in 0.1 M triethanolamine (pH 7.5), prehybridized for 10 min at 37°C in 4xSSC containing 12.5% formamide, and hybridized overnight at 42°C in hybridation buffer consisting of 4xSSC 12.5% formamide, 2.5x Denhardt’s solution, herring sperm DNA (250 μg/mL), yeast tRNA (125 μg/mL), and DIG-labelled probe (22 ng/μL). The slides were washed briefly with 4X SSC and then washed for 10 min with 1XSSC. DIG-labeled probes were detected according to the protocol from Roche Diagnostics. Briefly, the sections were incubated with anti-DIG-alkaline phosphatase for 2 h at 25 °C, rinsed with washing buffer, and incubated with nitroblue tetrazolium (NBT)–5-bromo-4-chloro-3-indolyl-phosphate (BCIP) for 2 h in the dark.

Immunohistochemistry and fluorescence microscopy

For retinal tissue analysis, mouse eyes were enucleated and fixed in 4% paraformaldehyde (PFA), in phosphate-buffered saline (PBS) for 2h at room temperature. After fixation, tissues were cryoprotected and sectioned following standard procedures. Retinal sections were permeabilized with 0.1% Triton X-100 in PBS for 15 min and blocked for 1h at room temperature in PBS containing 10% horse serum.

For immunofluorescence on cultured cells, ARPE-19 cells grown on glass coverslips were washed with PBS and fixed in 4% PFA for 20 min at room temperature. Cells were then permeabilized with 0.1% Triton X-100 in PBS for 5 min and blocked for 1h in PBS containing 10% horse serum.

Primary antibodies were diluted in blocking buffer and incubated overbigth at 4°C. After washing, samples were incubated with species-appropriate fluorescent secondary antibodies for 1h at room temperature in the dark. Fluorescently labeled siRNAs were detected directly. F-actin was visualized using fluorescently labeled phalloïdin and nuclei were counterstained with DAPI.

Fluorescence images were acquired using a confocal laser-scanning microscopes (SP5 Leica Microsystems or LSM-780 Zeiss) under identical acquisition settings for control and experimental samples. Images were then analyzed using ImageJ and Adobe Photoshop software.

Electron Microscopy

Eyes enucleated from mice killed by anesthetic overdose were placed in fresh fixative (2.5% glutaraldehyde in 100 mM cacodylate, pH 7.4) for 1 h at 25°C. Tissues were rinsed in the same buffer and then post fixed with 1% OsO4 in cacodylate buffer for 2 h. Samples were then rinsed in distilled water, dehydrated in a graded acetone series, and infiltrated and embedded in epoxy resin (EPON 812 EMS Hatfield PA 19440). Ultrathin sections (80 nm) were cut with an ultramicrotome (Reichert ultracut E), collected onto copper grids, stained with 4% uranyl acetate and lead citrate, and imaged in a transmission electron microscope (JEOL JEM 1400) at 100 kV equiped with a Morada SIS camera. Ultrastructural analyses focused on the organization of the photoreceptor outer segments, connecting cilium and retinal pigment epithelium.

Electroretinography (ERG)

Scotopic and photopic ERGs were recorded using a stainless ring electrode that made contact with the corneal surface through a thin layer of 0.7% methylcellulose. Needle electrodes placed in the cheek and the tail served as reference and ground leads, respectively. Responses were differentially amplified (0.3–500 Hz), averaged, and stored using a UTAS Big-Shot signal averaging system (LKC Technologies, Gaithersburg, MD). In scotopic conditions, ERGs were recorded to flash stimuli ranging from −4.72 to 1.87 log (sc td∗s) presented, in order of increasing intensity to the dark-adapted eye, in an LKC ganzfeld flashes. Cone ERGs were obtained to flash stimuli ranging between −2.32 and 1.87 log (sc td∗s) superimposed on a steady adapting field (44 cd/m2) after a 7-min adaptation period.24 The a-wave amplitude was measured from the pre-stimulus baseline to the trough of the a-wave. The b-wave amplitude was measured from the trough of the a-wave to the positive peak. The amplitude of the c-wave was measured from the pre-stimulus baseline to the peak of the c-wave. Oscillatory potentials (OPs), which are superimposed over the ascending phase of the ERG b-wave, were extracted for the highest and the lowest stimulus intensity using EM software with two filtering operations equivalent to applying a bandpass filter with corner frequencies of 73 and 500 Hz. For each of the four parameters, the data were plotted against flash luminance to generate the scotopic and photopic luminance response. Two other parameters were derived from the scotopic and photopic luminance response of the b-wave, the V max and logσ.

RNA isolation and RT-PCR

Total RNA was isolated using the Chomczynski method with a FastPrep apparatus. Reverse transcription was performed using Superscript III and oligo(dT) primers PCR amplification were carried out using gene-specific primers for mouse EFA6A and GAPDH. Products were analyzed agarose gel electrophoresis.

Bovine retinal rod outer segment membranes

Bovine retinal rod outer segment membranes (ROS) were prepared under dim red light as described by Kühn22 and stored at −80°C.

Opsonization of FITC-Labeled latex beads

In order to opsonize FITC-labeled latex beads (Sigma), 15 mg of ROS protein per 10 μL of beads was agitated for 1 h at room temperature. The beads were washed twice in 0.9% sterile NaCl and resuspended in 100 μL of NaCl.

Phagocytosis assays

Phagocytosis experiments were performed essentially as described elsewhere (Klettner et al. 2011).19 Briefly, Arpe-19 cells were grown on 11 mm round glass coverslips, treated with si-RNA control or EFA6A) for 48h before incubation with ROS-opsonized FITC latex beads for 6h. Then, cells were fixed in 3%paraphormaldehyde and a nuclear and F-actin staining was performed using Dapi and phalloidin respectively. Fluorescence image stacks of Arpe-19 cells obtained by laser scanning confocal microscopy (LSM780, Carl Zeiss, France) were analyzed to count ingested beads.

Western blot (WB) analysis

Cells were washed twice in PBS then scrapped and resuspended in Lysis buffer (20 mM TriEthanolAmine pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% sodium dodecyl sulfate and protease inhibitor PMSF 0.25 mM). Cell lysates were centrifuged and supernatants were diluted with Laemmli Sample Buffer. The proteins were heat denatured and processed by SDS-PAGE, transferred to nitrocellulose membranes, stained with Ponceau Red and immunoblotted with the indicated antibodies. Proteins were detected with SuperSignal western lightning chemiluminescence reagents (ThermoScientific) following the manufacturer’s instructions.

For the analysis of the protein expression in siRNA-treated mouse retinas by WB, retinas were dissected and washed in PBS and then homogenenized in Lysis Buffer. Samples were boiled for 10 min, briefly sonicated (10 s) and centrifuged to remove insoluble material. The resulting lysates were diluted 1:10, and the total protein concentrations were determined by the Bradford assay. Equal amounts of protein (40 μg/lane) were then resolved by SDS/PAGE and analyzed by Western blotting as described above.

Quantification and statistical analysis

Image quantifications were performed using ImageJ software (MIH). For acetylated tubulin labeling, signal distribution was quantified by measuring fluorescence intensity within two defined retinal regions: the ONL and the combined IS/OS region wich includes the connecting cilium. Data represent the mean of three to five independent experiments and are presented as mean ± SD. Statistical analyses were performed using two-tailed Student’s t-tests with Microsoft Excel software. Statistical analysis of ERG data were performed with statistical software (Prism 4; GraphPad). ERG results were analyzed by two-way ANOVA. Bonferroni post hoc testing was used to evaluate amplitude differences among intensities. Statistical significance (p) was determined as indicated in the figure legends, and p values <0.05 was considered statistically significant.

Additional resources

No additional resources were generated or required for this study.