Pannexin-2 deficiency disrupts visual pathways and leads to ocular defects in zebrafish

25 Pannexin-2 (Panx2), unlike the other pannexin channels Panx1 and Panx3, exhibits a unique 26 intracellular distribution, localizing at ER-mitochondria contact sites. These specialized 27 microdomains are crucial for important cellular functions, including calcium homeostasis, lipid 28 transfer, inflammation, and apoptosis. Despite their presence in neurons and glial cells, the 29 function of Panx2 at ER-mitochondria contact sites in neurotransmission remains unclear. Here, 30 we used TALEN technology to develop a Panx2 knockout (Panx2-/-) zebrafish model, to 31 investigate its role in neuronal communication. In 6 days, post fertilization TL (Panx2+/+) larvae, 32 Panx2 expression was observed in the inner and outer plexiform layers of the retina and the 33 arborization fields of the optic tract. Transcriptome profiling of Panx2-/- larvae by RNA-seq 34 analysis revealed down-regulation of vision-related genes, specifically those involved in visual 35 and sensory perception and lens development. Behavioral tests showed that loss of Panx2 leads to 36 altered visual motor response (VMR); Panx2-/- larvae exhibited reduced locomotor activity 37 during light phases and increased activity during dark phases. Additionally, the knockout larvae 38 displayed significantly impaired optomotor response (OMR). When we tested the geometric and 39 refractive properties of adult eyes, optical coherence tomography (OCT) analysis of Panx2-/- fish 40 revealed a longer mean axial length and a negative shift in retinal refractive error (RRE) values, 41 both indicative of myopia. Furthermore, the increased corneal thickness observed in Panx2-/- fish 42 corroborated the molecular and behavioural alterations. Our findings highlight a novel role of 43 Panx2 in retinal development, visual perception, and ocular health. 44 45 46 47 48 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 3

49 Pannexin-2 is the least studied member of the vertebrate-specific family of three proteins 50 Pannexin-1 (Panx1), Pannexin-2 (Panx2), and Pannexin-3 (Panx3) (1). Early reports 51 demonstrated that mammalian Panx2 exhibited a particularly dynamic expression pattern, with 52 low levels during prenatal development and a significant increase postnatally, suggesting its 53 involvement in CNS expansion (2). In line with this, another early report pointed toward a role of 54 Panx2 in regulating the timing of neuronal differentiation, via an S-palmitoylated modification 55 (3). Although originally believed to be confined to the central nervous system (4–6), Panx2 has 56 since been detected in sixteen different tissues; highest in skin, skeletal muscles, and the eye (7). 57 Recent structural studies have revealed Panx2 as a channel-forming protein with a unique pore 58 architecture that allows the passage of small molecules, including ATP (8,9). The protein 59 undergoes post-translational modifications, including N-glycosylation at asparagine (N86) within 60 its first extracellular loop (10). Unlike Panx1 and Panx3, N-glycosylated Panx2 is absent from the 61 plasma membrane (11,12). While Panx2 was known to localize within the cytoplasm, its specific 62 organelle distribution was not consistently defined. Studies reported its presence in various 63 intracellular structures, including the ER, Golgi apparatus, and endolysosomes (10,13,14). 64 Notably, Le Vasseur et al. identified Panx2 as a novel component of mitochondria-associated 65 membranes (MAMs), using compelling live-cell imaging and electron microscopy (15). The 66 authors showed that approximately 60% of endogenous Panx2 directly associates with 67 mitochondria under physiological conditions, suggesting specialized roles that differentiate it 68 from other pannexins. MAMs are dynamic ER-mitochondria contact sites crucial for cellular 69 processes such as lipid transfer, calcium homeostasis, immune regulation, and cell death (16). We 70 know that there are numerous ER-mitochondria contact sites in neurons (17), however the 71 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 4 specific roles of Panx2 within these junctions, particularly in the context of neurotransmission, 72 neurodegenerative diseases, and cancer remains to be elucidated (18,19). 73 Multiple Panx2 isoforms exist in humans and zebrafish, with conserved regions primarily 74 localized within the transmembrane domains and intracellular loops (6). Notably, the Panx2 C-75 terminal domain exhibits significant sequence variation compared to Panx1 and Panx3 (9). To 76 investigate whether these paralog differences correlate with distinct expression patterns, we 77 performed HCR-fluorescent in situ hybridization of whole larval zebrafish. Our results revealed 78 Panx2 mRNA expression in regions of the forebrain, cerebellum, and optic tectum. To further 79 elucidate the role of Panx2 in neuronal communication and MAMs function, we generated a 80 global Panx2 knockout zebrafish model using TALEN technology. A custom-made antibody 81 identified Panx2 in the inner segments of photoreceptors and the outer retina and confirmed the 82 loss of Panx2 in the knockout. Transcriptome analysis revealed differential expression of genes 83 involved in visual perception, circadian rhythm, and both innate and adaptive immune responses. 84 Subsequent behavioral studies demonstrated that loss of Panx2 impairs visual motor responses 85 and optomotor behavior, highlighting its critical role in visual information processing. Lastly, 86 optical coherence tomography of adult zebrafish eyes revealed alterations in the shape and size of 87 the lens and eye, suggesting that early-life molecular changes induced by the absence of Panx2 88

in a myopia-like phenotype. We concluded that Panx2’s novel association with myopia is a 89 contributing factor to the observed visual-motor phenotype. 90

91 Panx2 is expressed broadly in the brain of zebrafish larvae. 92 RNA fluorescence in-situ hybridization (RNA-FISH) was used to determine panx2 93 expression in 6 days post-fertilization (dpf) whole-mount TL larvae. We found panx2 mRNA 94 widely expressed in the forebrain, cerebellum, optic tectum and retina (Fig. 1). 95 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 5 96 Figure 1: panx2 is expressed broadly in the brain of zebrafish larvae. The spatial distribution of 97 panx2 mRNA was detected by fluorescence in-situ hybridization. Scale bars:100 µm. 98 99 Generating a Panx2 global knockout zebrafish line. 100 The panx2 mutant allele was generated by TALEN-mediated genome editing, targeting 101 the BamHI restriction endonuclease site in the first exon of the panx2 gene (Fig. 2a). A DNA 102 sequence analysis of microinjected embryos identified short 4 to 11 base pairs (bp) long 103 nucleotide deletions in exon one of the panx2 gene. An adult founder fish (F0) with a 11 base pair 104 deletion (panx2Δ11) was selected for further experimentation (Fig. 2b). In this fish the 11 bp 105 deletion caused a frameshift at amino acid (aa) G28, resulting in a premature stop codon. The 106 consequence of this stop codon was a truncated reading frame for a 27-aa protein, lacking 640-aa 107 of the 651-aa long Panx2-C protein sequence (Fig. 2c). The Panx2-/- fish were viable and breed 108 for multiple generations. When comparing age-matched male and female adults to wild type 109 siblings a modest but significant decrease in body length was noted (Fig. 2d,e). The gene-editing 110 event in exon1 effectively eliminated the shared start codon for the three known Panx2 protein 111 isoforms A: 580-aa (ABY78016), B: 604-aa (ABY78018), and C: 651aa (ABY78017). The 112 Panx2-C isoform was the most abundant Panx2 mRNA detectable in 6 dpf TL (Panx2+/+) larvae 113 by RT-qPCR suggesting that it may play a dominant role in Panx2-mediated functions (Fig. 2f). 114 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 6 115 Figure 2: Generating a Panx2 global knockout zebrafish line. a) Exon-intron organization of the 116 zebrafish Panx2 gene (ENSDARG00000063019, chromosome 18 (GRCz11:CM002902.2). The 117 positions of the targeting Left and Right arms with a unique BamHI restriction endonuclease site 118 in a spacer region between the TALEN pair are highlighted. b) Representation of the 11bp 119 deletion causing a premature stop codon. c) Graphical representation of the predicted Panx2 120 structure. The region in blue indicates the residual sequence of 27 amino acids followed by a new 121 in-frame stop codon. d,e) Comparison of the body length of age-matched adult zebrafish of both 122 sexes. f) Real Time PCR quantification and presentation of the normalized expression of the three 123 Panx2 isoforms. Statistics in e,f: Welch’s test. Sample sizes were n=26 for Panx2+/+ and n=20 for 124 Panx2-/-. Significance: ***P-value<0.001 and *P-value<0.05. Error bars = SD. 125 126 Panx2 is localized in the outer retina of the zebrafish. 127 To demonstrate the loss of Panx2 protein expression in larvae a custom polyclonal anti-128 peptide Panx2 antibody was generated. Panx2+/+ larvae revealed a low, ubiquitous Panx2 129 immunoreactivity, which was most pronounced in the outer retina, the inner plexiform layer, and 130 the lens epithelium (Fig. 3a, see open triangles). Panx2 knockout larvae exhibited a significant 131 reduction in fluorescence with residual signals attributed to collagen autofluorescence in the 132 sclera and at the margins of the lens epithelium (Fig. 3b). At higher magnification of the outer 133 retina Panx2 localization was prominent in the axons of the inner segments of photoreceptor cells 134 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 7 and the lens epithelium (Fig. 3c,d, see open triangles). No immunoreactivity was detected in 135 horizontal cells which express two other pannexins, the Panx1a and Panx1b isoforms (Prochnow 136 et al. (2012). 137 138 Figure 3: Panx2 localization in outer retina of the zebrafish. a) Panx2 proteins are ubiquitously 139 expressed throughout the retina. The expression was enriched in the outer retina, the inner 140 plexiform layer, and the lens epithelium. b) After Panx2 ablation the expression was greatly 141 reduced across the retina, with residual staining left in tissue known for high autofluorescence. c) 142 Higher magnification of the outer retina showing the pronounced staining in sections of the inner 143 photoreceptor segments. d) Higher magnification of the lens epithelium. Abbreviations: SCL 144 sclera, LE lens epithelium, GCL ganglion cell layer, IPL inner plexiform layer, INL inner nuclear 145 layer, OPL outer plexiform layer, ONL outer nuclear layer, PRL photoreceptor layer. Scale bars: 146 (a, b) 60 µm; (c, d) 10 µm. 147 148 Panx2 ablation regulates vision processes and structural components of the lens. 149 A comparison of the transcriptomes of 6 dpf Panx2+/+ and Panx2-/- larvae identified 5355 150 differentially expressed genes (DEG) when the adjusted P-value (padj) was <0.05. In total, 2512 151 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 8 genes were up-regulated, and 2843 were down-regulated (Fig. 4a). The Gene Ontology analyzer 152 for RNA-seq (GOSeq 1.56.0; (20)) identified Biological Processes (BP) and Molecular Functions 153 (MF) that were over/under-represented in the DEG data (Fig. 4b). Vision-related categories like 154 Sensory Perception of Light Stimulus (GO:0050953; 61 genes down) or Sensory Perception 155 (GO:0007600; 70 genes down) were significantly down-regulated (Fig. 4c,d). The GO 156 annotations refer to the events required for an organism to receive a sensory light stimulus, 157 convert it to a molecular signal, and recognize as well as characterize it. The most up-regulated 158 process was the Humoral Immune Response representing immune responses mediated through a 159 body fluid (GO:0006959; 17 genes up) (Fig. 4e). The Structural Component of the Lens 160 (GO:0005212; 38 genes down) was the most significantly down-regulated Molecular Function. 161 This GO annotation refers to molecules that contribute to the structural integrity of the lens (Fig. 162 4f). The most up-regulated MFs were Endopeptidase Regulator Activity (GO:0061135) 163 encompassing molecules that modulate the activity of peptidases, which can control critical 164 functions of innate and adaptive immune responses like antigen processing and presentation of 165 immunogenic peptides. 166 Next, a STRING analysis with k-means clustering of the 61 differentially down-regulated 167 genes annotated as Sensory Perception of Light Stimulus (GO:0050953) showed that both the 168 retina and the lens were affected by loss of Panx2. Cluster I represented genes with essential roles 169 in normal vision and signal transduction in the retina. Genes belonging to the retinal arrestin 170 family (arr3a and arr3b), rhodopsin family (rho and rhol), as well as the guanylyl cyclase family 171 (guca1g and gucy2d), were significantly down-regulated (Fig. 4g). Cluster II represented lens-172 specific proteins belonging to the two alpha- or beta/gamma groups of the crystallin super gene 173 families, including members of the gamma-N and gamma-M subgroups. The group represented 174 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 9 the M-subfamily of gamma-crystallin, which is specific to fish. These genes play essential roles 175 in the lens structure; mutations have been shown to cause cataracts. 176 Genes in the up-regulated category Humoral Immune Response (GO:0006959) (Fig. 4g) 177 showed notable up-regulated genes in Panx2-/- larvae. Genes such as si:dkey-22f5.9, c8a, c8g, c9, 178 are expressed in the liver. Interestingly, genes implicated in human eye diseases (cfb) and age-179 related macular degeneration (c2/si:ch1073-280e3.1 (F1QJB3_DANRE) were also found to be 180 up-regulated. The significant differential expression of retina and lens genes, as well as genes 181 implicated in human eye disorders, led to us prioritize biological processes related to vision. 182 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 10 183 184 Figure 4: Panx2 ablation regulates vision processes and structural components of the lens. a) A 185 DESeq2 analysis of RNAseq data from Panx2+/+ and Panx2-/- 6dpf zebrafish larvae identified 186 5355 differentially expressed genes with an adjusted P-value < 0.05 b) Analysis of differentially 187 expressed genes using GOSeq (v1.56.0). The Gene Ontology Classifier Biological Process (BP) 188 showed significant downregulation of several vision-related categories. c,d) The most 189 significantly down-regulated biological processes included Sensory Perception of Light Stimulus 190 (GO:0050953) and Sensory Perception (GO:0007600). e) The top up-regulated BP was Humoral 191 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 11 Immune Response (GO:0006959). f) The most significantly down-regulated Molecular Function 192 (MF) was the Structural Component of the Lens (GO:0005212). g) STRING analysis of 61 193 regulated genes annotated as Sensory Perception of Light Stimulus (GO:0050953) and 17 194 regulated genes annotated as Humoral Immune Response (GO:0006959). K-means clustering 195 identified two clusters which were either retina- or lens-specific. Cluster I represents genes 196 implicated in the phototransduction processes of the retina. Cluster II represents lens-specific 197 proteins belonging to the alpha- or beta/gamma supergene families. A third group represents the 198 humoral immune response. 199 200 Light-ON and light-OFF conditions affect free swimming of Panx2-/- larvae. 201 202 We examined visually guided behaviors to explore consequences of the differential 203 expression of genes with roles in visual function. Baseline spontaneous swimming activity of 204 both genotypes was analyzed at 6 dpf, during light-ON (1200 lux) and light-OFF (0 lux) phases. 205 The representative images illustrate locomotor patterns within a single well in the light-ON 206 condition (Fig. 5a). Knockout larvae preferred swimming close to the circumference of the well. 207 In the light, Panx2-/- larvae swam a shorter distance (t=7.755, df=113.1, P-value <0.001; n=24 208 larvae) at a lower velocity (t=7.177, df=117.6, P-value <0.001; n=24 larvae) than the Panx2+/+ 209 wildtype group (Fig. 5b,c). In the absence of light, both Panx2+/+ and Panx2-/- larvae exhibited a 210 preference for the perimeter and avoided the central zone of the well (Fig. 5d). However, in the 211 light-OFF condition, Panx2-/- larvae swam further (t=4.147, df=88.91, P-value <0.001; n=24 212 larvae) at a lower steady velocity than Panx2+/+ larvae (t=4.095, df=96.20, P-value <0.001; n=24 213 larvae) (Fig. 5e,f). 214 215 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 12 216 Figure 5: Light-ON and light-OFF conditions affect free swimming of Panx2-/-. Representative 217 images of a single well showing examples of Panx2+/+ and Panx2-/- larvae locomotion patterns in 218 the light (a). Medium (20 mm/sec) are visualized with 219 green and red colors, respectively. Graphs demonstrate mean experimental traces ± 95% CI of 220 distance traveled (mm) and speed (mm/sec) of 6 dpf larvae for 30 min (b,c) and their 221 corresponding mean values, in the light. d) Representative images for each genotype, in the dark. 222 Tracked experimental traces of distance and speed under dark conditions (e,f) and their 223 corresponding mean ± SD. Sample sizes were n=24 larvae for each genotype. Significance: 224 ****P-value<0.0001 and *P-value<0.05. Error bars = SD. 225 226 Panx2 ablation affects the visual motor response to light-ON and light-OFF stimuli. 227 228 Visual motor response (VMR) analysis was performed to compare the stereotypical motor 229 responses of Panx2+/+ and Panx2-/- larvae. The kinematic data collected provided information 230 about seven parameters: duration of bouts, duration of bursts, freeze duration, average activity 231 duration, bout counts, burst counts, and freeze counts. A principal component analysis identified 232 “freeze duration” (dimension 1) and “average activity duration” (dimension 2) as capturing 233 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 13 73.62% and 15.14% of the data variance, respectively (Fig. 6a). Both parameters were selected to 234 compare the genotypes divergence. We found that Panx2-/- larvae displayed strikingly different 235 activity patterns for freeze duration, which represents the amount of time that the larval were 236 immobile (Fig. 6b) and average activity duration (Fig. 6c). Panx2 knockout larvae were less 237 active during light-ON phases, and more active in light-OFF phases. A Welch’s t-test established 238 statistical differences in mean activity of Panx2+/+ and Panx2-/- in both light conditions (P-value 239 <0.0001 for light-ON (Fig. 6d), P-value <0.0001 for light-OFF (Fig. 6e), n=24 larvae). 240 A modified VMR assay was performed next where the intensity of light was gradually 241 increased; locomotor activity was measured at 10% (400 lux), 20% (800 lux) and 30% (1200 lux) 242 light intensity. With the onset of light, Panx2+/+ larvae exhibited a progressive increase in 243 average activity duration. In contrast, Panx2-/- larvae activity was less variable as the light 244 stimulus intensified (Fig. 6f). The variance analysis revealed significant differences (P-values 245 <0.0001) in activity between Panx2+/+ and Panx2-/- larvae across all levels of luminosity (Fig. 246 6g,h,i). Altogether, the differential movement responses demonstrated that Panx2-/- larvae can 247 sense changes in light but their observed response to luminance is altered. 248 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 14 249 250 Figure 6: Panx2 ablation affects the response to light-ON and light-OFF stimuli. Principal 251 component analysis was performed using multivariable activity data. a) The variable correlation 252 plot indicates the coordinates of seven variables within the first two dimensions, where PC1 and 253 PC2 capture 73.62% and 15.14% of the data variance, respectively. b) Line graph showing 254 patterns of immobility with mean freeze duration activity for Panx2+/+ (red) and Panx2-/- (blue) 255 over alternating 20-minute periods of light-ON and light-OFF conditions. c) Line graph showing 256 the mean average activity. Mean activity, with error bars showing SD for d) light-ON and e) 257 light-OFF conditions, ****P-value<0.0001. f) Line graph showing the mean average activity for 258 Panx2+/+ (red) and Panx2-/- (blue) during an initial light-OFF period, followed by 20-minute 259 increments of increasing light intensities. Mean activity, with error bars showing SD during g) 260 10% light intensity, h) 20% light intensity and i) 30% light intensity. ****P-value<0.0001. 261 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 15 GraphPad Prism was used to generate the figure and perform statistical analysis. Data shown for 262 n=24 larvae for each genotype. 263 264 Panx2 ablation alters the optomotor response of larvae. 265 266 Optomotor response (OMR) assays evaluated the visual perception of Panx2-/- larvae. The 267 assay enabled us to test an innate behavioral response that many animals use to stabilize 268 themselves with respect to a visual environmental stimulus (21). Both Panx2+/+ and Panx2-/- 269 larvae responded to visual stimuli, composed of moving black and white stripes. Previously 270 optimized parameters (22) for three spatial frequencies (SF: 64, 128, and 256 pixels/cycle), two 271 velocities (72 and 144 pixels/sec), and two contrast levels (10% and 100%) were tested. The line 272 graphs in Fig. 7 summarize the percentage of positive response (PPR); corresponding to the 273 number of larvae that showed the expected response by swimming in the direction of the stimulus 274 (n=16 for genotype, for each distinct parameter). Altogether, Panx2-/- larvae demonstrated lower 275 PPR and hence lower visual acuity in comparison to the wildtype group for all permutations of 276 SF, speed, and contrast. 277 Under low contrast settings (10%), Panx2-/- larvae showed an average PPR of 6.94 in 278 comparison to 45.15 for Panx2+/+. The variance was most prevalent for the rightward direction at 279 a speed of 72 pixels/sec and SF of 64, where Panx2-/- larvae did not respond while PPR for TL 280 was 66.7 (P-value 0.00061, n=16 for each genotype). Similarly, when the stimulus was presented 281 in the leftward direction at the same speed and SF, Panx2-/- larvae had a PPR of 8.3 in 282 comparison to a much higher value of 58.3 for Panx2+/+ (P-value 0.006, n=16 for each genotype). 283 No statistically significant difference was observed between the two groups at the SF of 256 at 72 284 pixels/sec; and with the SF of 128 at 144 pixels/sec, regardless of directionality. 285 In the high contrast setting (100%), Panx2-/- larvae showed an average PPR of 26.38. In this 286 luminous environment, a statistically significant difference was observed between Panx2+/+ and 287 the knockout larvae at all combinations of SF and speed. Furthermore, peak response deficiency 288 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 16 was observed for Panx2-/- larvae at a SF of 64; reflecting the trend seen under low contrast 289 conditions at the same SF. While there was an increase in PPR for Panx2+/+ at high contrast, 290 when the edges of the moving stripes were well defined, there was still a considerable decrease in 291 OMR response with the loss of Panx2. The observed decline in visual acuity of Panx2-/- larvae 292 was found to be independent of direction (P-values: see Table 1). 293 Table 1: Comparison of OMR positive response percentages between Panx2+/+ and Panx2-/- 294 larvae. 295 Leftward Rightward Spatial frequency (pixels/cycle) Contrast Speed (pixels/sec) 64 128 256 64 128 256 Low (10%) 72 0.006 0.003 0.08 0.00061 0.004 0.29 144 0.013 0.176 0.037 0.03 0.575 0.287 High (100%) 72 0.001 0.04 0.03 0.00061 0.03 0.001 144 0.0014 0.036 0.012 0.0014 0.031 0.008 296 297 Figure 7: Optomotor response (OMR) impaired in Panx2-/- larvae. OMR was used to assess optic 298 flow processing in 6 dpf Panx2+/+ and Panx2-/- larvae, at two light intensities (10% and 100%). 299 ** ** * * * ** *** * **** * * Positive response rate (%) Low contrast; 10% luminosity Left Right 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 ** ***ns ** ** ns * * ns ns * panx2+/+ panx2-/- 72 pixels/sec Spatial frequency 144 pixels/sec 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 High contrast; 100% luminosity Left Right 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 0.0 20.0 40.0 60.0 80.0 100.0 64 128 256 ns .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 17 Both groups were shown visual stimulus of moving stripes at three spatial frequencies (64, 128, 300 and 256), which were presented at 72 pixels/sec and 144 pixels/sec. The line graphs show the 301 percentage of larvae that indicated the expected positive response, from n=16 for each genotype. 302 Panx2-/- larvae had significantly lower visual acuity when compared to Panx2+/+. ***P-303 value<0.001, **P-value<0.01, *P-value<0.05, ns = not a significant difference. 304 305 Panx2 loss leads to lens malformation and characteristics consistent with myopia. 306 Whole-eye Optical Coherence Tomography (OCT) images of one-year-old Panx2+/+ and 307 Panx2-/- zebrafish were acquired using a custom-developed 1310nm Spectral Domain OCT (SD-308 OCT) system. An analysis of each eye was performed from a microstructural perspective and 309 using geometric parameters, including axial length, lens diameter, retinal radius, and corneal 310 thickness, as illustrated in Fig. 8a. Detailed examinations of whole-eye images revealed 311 disruptions to the lens epithelium in Panx2-/- fish. These disruptions manifested as protruding 312 rings within the lens, unlike the smooth, well-organized epithelium observed in Panx2+/+ fish 313 (Fig. 8b,c,d). We speculated that the defects could impair the lens's light-focusing ability by 314 disrupting the normal light pathway. 315 The geometric data, normalized to body length, demonstrated that Panx2-/- fish had a 316 significantly longer mean axial length (Panx2+/+ = 47.57 ± 1.39 µm/mm; Panx2-/- = 51.97 ± 2.26 317 µm/mm; P-value <0.0001; Fig. 8e). Moreover, using the high-resolution whole eye OCT images 318 we calculated the Relative Refractive Error (RRE) to investigate anomalies in refractive 319 properties. RRE is a widely used standardized metric calculated as 1- (retinal radius/F), where F 320 is an idealized focal length equal to lens radius × 2.324. Examination of calculated RREs revealed 321 that Panx2 ablation led to a decrease in RRE values (Panx2+/+ = 0.02 ± 0.03; Panx2-/- = -0.03 ± 322 0.03; P-value <0.0001; Fig. 8f). Eyes with a distance from the lens center to the retinal pigment 323 epithelium (RPE) greater than the expected retinal radius exhibit negative RRE values, indicating 324 myopic shifts. Conversely, eyes with a shorter distance (RPE closer to the lens) have positive 325 RRE values, signifying hyperopic shifts (23). The observed statistically significant reduction of 326 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 18 RRE suggested myopia-like abnormalities in Panx2-/- fish. Finally, we detected an increase in 327 corneal thickness in Panx2-/- fish, when normalized to lens diameter (Panx2+/+ = 0.07 ± 0.01 328 µm/mm, n=38 eyes; Panx2-/- = 0.08 ± 0.01 µm/mm, n=40 eyes; P-value <0.001; Fig. 8g,h). For 329 the above-mentioned parameters, we examined a total of 38 eyes from 19 Panx2+/+ control and 330 40 eyes from 20 Panx2-/- age-matched adult fish. Altogether, our analysis demonstrated lens 331 abnormalities in Panx2-/- fish, which suggested that Panx2 loss-of-function may contribute to 332 myopic vision deficits in adult zebrafish. 333 334 335 336 Figure 8: Loss of Panx2 leads to lens defects. Eyes of adult fish were imaged using a SD-OCT 337 system. a) Schematic definition of four geometric parameters used for quantification of high-338 resolution OCT Whole-eye images. Representative images of Panx2+/+ (b) and Panx2-/- (c,d), 339 showing disruptions in lens epithelium. Scale bars = 500 μm. Bar plots showing spread of 340 quantified geometrical parameters for the axial length normalized to body length (e), relative 341 refractive error (f), corneal thickness (g), and corneal thickness normalized to lens diameter (h), 342 collectively showing significant differences between the genotypes. Sample sizes for Panx2+/+ 343 and Panx2-/- were 38 and 40 eyes from 19 and 20 fish, respectively. Significance: ***P-value 344 <0.001 and ****P-value <0.0001. 345 346 347 348 349 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 19

350 351 This study demonstrates roles of Panx2 for visual function in zebrafish. We provide three 352 lines of evidence. RNAseq data show the differential expression of genes with biological 353 functions in the sensory perception of light. The testing of behavior in response to different light 354 stimuli demonstrates that Panx2-deficient larvae exhibit both altered locomotor behavior in 355 response to illumination and impaired optomotor responses, indicating defects in both visual 356 perception and acuity. Finally, in old zebrafish, loss of Panx2 is correlated with changes to the 357 geometry and refractive properties of eyes. 358 The novel results advance our previous work on pannexin functions in the visual system. 359 We had shown previously functions of two other pannexin family members. Panx1a is 360 predominantly expressed in the outer retina and functions in light decrement detection (24,25). 361 Our work on Panx1b had revealed that the isoform is localized to the inner retina and ganglion 362 cell layer (22,26). We also showed that Panx1b is expressed in the end-feet of Muller glial cells 363 which was first evidence for a unique cell type specific expression, - neuronal versus glial -, that 364 is unique to fish (22). Behavioral analyses have revealed distinct phenotypes for Panx1a and 365 Panx1b knockout zebrafish. Panx1a KOs exhibited altered visual motor responses (VMR), 366 characterized by a general reduction in activity and specific deficits in dark phase locomotion 367 (25). Panx1b knock outs, on the other hand, showed impairments in optic flow direction-368 selectivity, particularly at high spatiotemporal frequencies and low contrast surroundings (22). It 369 was also demonstrated that Panx1a and Panx1b function in the outer retina as the slow negative 370 feedback signal from HCs to cones to generate the center/surround organization of bipolar cell 371 receptive fields (27). Here, we show that Panx2 is expressed in both the inner and outer retinal 372 layers, with prominent expression in the photoreceptor inner segment. Our data aligns with a 373 previous report, in the murine model, which showed that Panx2 is expressed in the photoreceptor 374 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 20 inner segment and outer plexiform layer of the mouse retina (7). Panx2 KO larvae exhibit a 375 pronounced reduction in locomotor activity during light phases and an increase in locomotor 376 activity during dark phases, a pattern that is markedly different from the VMR phenotypes 377 observed in Panx1a/b KOs (22,25). Furthermore, the OMR response in Panx2 KOs is severely 378 impaired during both low and high contrast conditions, regardless of direction. Thus, our findings 379 for Panx2 KO larvae suggest a severe and distinct role in retinal function compared to Panx1a/b. 380 Consistent with the behavioral data, RNAseq analysis identified a marked decrease in the 381 expression of genes involved in the detection of light stimuli in Panx2 knockouts. We speculate 382 that Panx2 plays a critical role in retinal signaling that is not redundant with Panx1a or Panx1b. It 383 is likely that their specific functions may be complementary or synergistic, contributing to 384 different aspects of visual processing. 385 The observed expression patterns and widespread changes in gene expression and 386 behaviour highlight a pivotal role of Panx2 in retinal physiology. Its strategic localization at ER-387 mitochondria contacts (MAMs) and its abundance in the neuropil, a critical region for synaptic 388 transmission, suggest that the neuronal deficits in Panx2 knockout fish could stem from 389 disruptions in bioenergetics and synaptic function. Given their highly compartmentalized 390 nature, neurons rely on precise calcium control, which is essential for their function. The 391 endoplasmic reticulum (ER) and mitochondria are critical for this regulation (28). 392 Mitochondria, in addition to providing energy to postsynaptic spines, also help maintain calcium 393 homeostasis (29). We reason that the loss of Panx2 function in the inner segments of 394 photoreceptors affects the biosynthetic machinery essential for photoreceptor function. Inner 395 segments are rich in mitochondria, and together with mitochondria-associated membranes 396 (MAMs) play a crucial role as contact sites between the endoplasmic reticulum (ER) and 397 mitochondria, facilitating the transfer of calcium and other signals. Studies indicate that 398 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 21 dysfunction in the inner segment's ability to produce energy can lead to vision loss and 399 degenerative diseases like retinitis pigmentosa and age-related macular degeneration (30). 400 Additionally, disruptions in the inner segment's metabolic processes are associated with stress 401 responses that may lead to photoreceptor cell death, impacting vision over time (31). The 402 intricate network of ER-mitochondria connections, as revealed by recent imaging techniques, 403 underscores the importance of MAMs in regulating calcium concentrations and energy dynamics 404 within neurons (17). We know that the metabolic demands of synaptic transmission are 405 substantial, requiring a constant supply of ATP (32). It is possible then that Panx2-mediated 406 processes within the ER-mitochondrial axis influence energy production or utilization, leading to 407 deficits in neuronal function. Additionally, the enriched expression of Panx2 in the neuropil 408 suggests its involvement in regulating synaptic transmission, potentially through its role in 409 calcium signaling or mitochondrial dynamics, which are known to influence the efficiency of 410 energy production. Studies have shown that ER releases calcium in dendrites upon synaptic 411 stimulation, which is subsequently buffered by mitochondria. Moreover, the tethering of 412 mitochondria to the ER might influence the duration of ER residence in spines, potentially 413 affecting synaptic plasticity (33). Computational modeling has further emphasized the role of 414 MAMs in regulating calcium and ATP dynamics within neurons, highlighting the importance of 415 MAMs communication in shaping synaptic signaling (28). Interestingly, prior overexpression 416 studies revealed that both paralogs of Panx2, Panx1 and Panx3, can form calcium-permeable ER 417 channels (34,35). These possibilities allude to the role of Panx2 in supporting retinal function; 418 and indicate that its influence likely extends beyond the visual signaling pathway. 419 Both Panx1 and Panx2 expression has been identified previously in the adult murine lens 420 epithelium (36). In adult zebrafish, Panx2 ablation resulted in lens malformations consistent with 421 myopia. RNA-seq data collaborated this finding, indicating differential expression of genes 422 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 22 associated with the structural components of the eye lens early in life. Significantly 423 downregulated genes were those for lens-specific proteins, including alpha- and beta/gamma-424 crystallins, as well as fish-specific M-subfamily gamma-crystallins. These genes are essential for 425 maintaining lens structure, and mutations in them can lead to cataracts (37,38). These results 426 reinforce the possibility that Panx2 is critical for cellular energy homeostasis. The loss of Panx2 427 could disrupt the energy supply to lens cells, hindering their growth and differentiation and hence 428 potentially resulting in a myopic phenotype. Additionally, cellular stress induced by metabolic 429 disruptions may contribute to the development of myopia-related ocular changes, such as altered 430 lens curvature or axial elongation. Collectively, our study underscores the multifaceted role of 431 Panx2 in retinal physiology. 432 Beyond its impact on vision, transcriptome profiling also suggested that Panx2 is a novel 433 regulator of immune responses, with immune activation potentially exacerbating retinal and lens 434 degeneration in aged animals. While we have not prioritized investigating this result it implies 435 that Panx2 plays multifaceted roles in maintaining visual acuity, regulating ocular morphology, 436 and moderating neuro-immune interactions crucial for retinal and lens development and health. In 437 this regard this study's implications extend to understanding the neuro-immune axis in vision and 438 suggest that Panx2 could be a therapeutic target for conditions involving retinal degeneration, 439 immune dysregulation, cataracts, or myopia. 440 441

442 Zebrafish lines. 443 Zebrafish (Danio Rerio) of strain Tubingen long fin (TL) were maintained in groups with mixed 444 sex in a recirculation system (Aquaneering Inc., San Diego, CA) at 28⁰C on a 14hr light/10hr 445 dark cycle. All animal work was performed at York University’s zebrafish vivarium and in an S2 446 biosafety laboratory following the Canadian Council for Animal Care guidelines after approval of 447 the study protocol by the York University Animal Care Committee (GZ#2019-7-R2). 448 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 23 Generating Panx2-/- zebrafish. 449 Potential TALENs target sites were identified using Mojo Hand software (http://talendesign.org) 450 (11,12). TALENs targeted exon 1 of the panx2 gene (NM_001256641). The TALEN constructs 451 were synthesized in Dr. Stephen Ekker’s lab (Mayo Clinic Cancer Center, Rochester, MN). 452 TALEN assemblies of the repeat-variable di-residues (RVD-containing repeats) were conducted 453 using the Golden Gate approach (39). pT3TS-GoldyTALEN expression vectors (40,41). Plasmids 454 were linearized with the SacI restriction endonuclease (ThermoFisher Scientific, Canada) for 15 455 min at 37°C and used as templates for in vitro transcription. 456 Capped cRNAs were synthesized from TALEN pairs mixed 1:1 using the mMESSAGE 457 mMACHINE T3 Transcription kit (Life Technologies, Canada) and purified using the Oligotex 458 mRNA Mini Kit (Qiagen Inc., Toronto, Canada). TALEN cRNAs were diluted in DNase/RNase-459 free water (Life Technologies) to the final concentration of 1 µg/μL and stored at −80 °C before 460 microinjection. 461 One-cell stage zebrafish embryos were microinjected with TALEN cRNAs pair at a dose 462 of 25 pg/nl. Genomic DNA (gDNA) was extracted from injected embryos at 4 dpf to examine the 463 TALEN mutagenesis efficiency. Individual larvae were incubated in 100 mM NaOH at 95℃ for 464 15 min. After cooling to room temperature, one-tenth of the 1 M Tris (pH8.0) volume was added 465 to the extracts to neutralize the NaOH. Finally, 1 volume TE buffer pH 8.0 was added, and 466 gDNAs were stored at -20℃. gPCR was used as a screen to detect small indel mutations by 467 BamHI restriction enzyme (RE) digests. PCR primers for genotyping: 5’-468 CGAATGCAGAATATCCTCGAGCAG-3’; reverse, 5’-GTGACGACCCGGTCAAAGG-3’. 469 Indel mutations were confirmed by sequencing (Eurofins Genomics LLC, KY, USA) of gel-470 purified PCR products cloned into the pJet1.2 cloning vector (Life Technologies). 471 Adult mosaic zebrafish (F0) were anesthetized in pH-buffered ethyl 3-aminobenzoate 472 methanesulfonate solution (0.2 mg/ml; MS-222, Sigma-Aldrich). A section of the caudal fin was 473 removed using dissecting scissors (WPI Inc., FL, USA) and placed into 1.5 ml collecting tubes. 474 The fin gDNA was isolated and screened for indel mutations as described (40). Adult F0 475 zebrafish were out-crossed to wild-type (WT) TL zebrafish and their F1 offspring were analyzed 476 by PCR and BamHI restriction digestions to verify germline transmission of mutations. 477 Heterozygous Panx2+/- F1 mutants were in-crossed to establish homozygous F2 mutants Panx2-/-. 478 All experiments described were performed with progenies of F3 or alter generations. Age 479 matched siblings served as controls. 480 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 24 NGS RNA-sequencing. 481 The transcriptomes of the Panx2-/- and sibling controls (Panx2+/+) were analyzed by RNA-seq 482 (NGS-Facility, The Center for Applied Genomics, SickKids, Toronto, ON, Canada). The raw 483 data represented the sequencing of three independent pools of ≈30 age-matched larvae at 6 dpf. 484 The RNA-seq data are deposited at the NCBI - Gene Expression Omnibus (GEO) database 485 repository (ID pending). Total RNAs were extracted using RNeasy Plus Mini Kit (Qiagen). The 486 RNA quality was determined using a Bioanalyzer 2100 DNA High Sensitivity chip (Agilent 487 Technologies, Mississauga, ON, Canada). The RNA library preparation was performed following 488 the NEB NEBNext Ultra II Directional RNA Library Preparation protocol (New England Biolabs 489 Inc., Ipswich, MA, USA). RNA libraries were loaded on a Bioanalyzer 2100 DNA High 490 Sensitivity chip to check for size, quantified by qPCR using the Kapa Library Quantification 491 Illumina/ABI Prism Kit protocol (KAPA Biosystems, Wilmington, MA, USA). Pooled libraries 492 were paired-end sequenced on a High Throughput Run Mode flow cell with the V4 sequencing 493 chemistry on an Illumina HiSeq 2500 platform (Illumina, Inc., San Diego, CA) following 494 Illumina’s recommended protocol to generate paired-end reads of 126-bases in length. 495 Differential Gene Expression and Functional Enrichment Analyses. 496 The post-sequencing processing to final read counts, normalization, and differential gene 497 expression analysis used multiple software packages, including a two-condition differential 498 expression analysis using the edgeR R-package, v.4.0.16 (42,43) and DESeq2 R package version 499 1.42.1 (44). Genotypes were incorporated into the statistical model and multiple hypothesis 500 testing was performed. The default filter for DESeq2 used a threshold of p<0.05 and the 501 Benjamini-Hochberg procedure to determine the false discovery rate (FDR), and the adjusted P-502 value (padj). 503 A Gene Set Enrichment Analysis (GSEA) tested whether a defined set of genes shows 504 statistically significant, concordant differences between genotypes. The data were filtered at padj 505 < 0.05 and analyzed for enrichment using R GOseq (v1.56.0) (20). Other tools used to process 506 RNA-seq data were HCOP (https://www.genenames.org/tools/hcop/) and db2db at bioDBnet 507 (https://biodbnet.abcc.ncifcrf.gov/db/db2db.php) to convert curated human or mouse to zebrafish 508 genes (45,46). Curated gene lists were generated based on GOseq and analyzed in STRING v12.0 509 (string-db.org) (47). The top-scoring categories were visualized using ggplot2 (v3.5.1) from the 510 tidyverse package (v2.0.0) in R (48). 511 512 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 25 Hybridization chain reaction RNA-fluorescence in situ hybridization (HCR RNA-FISH). 513 Larvae were raised under standard conditions in egg water. When embryos reached 12 hours 514 post-fertilization (hpf), egg water was replaced with egg water containing 0.003% of 1-phenyl 2-515 thiourea (PTU). Fresh egg water containing 0.003% PTU was replaced every 24 hours until the 516 larvae reached 5 dpf. At 6 dpf, the larvae were euthanized and fixed in 4% ice-cold 517 paraformaldehyde (PFA). After 24 hours, the PFA was washed out with 1× PBS, and the samples 518 were gradually dehydrated, permeabilized with methanol, and stored at −20 °C for several days 519 until HCR in situ labelling was performed. 520 Staining was performed according to the manufacturer's protocol for whole-mount 521 zebrafish larvae (49). Specifically, the samples were separated into 5 larvae per well in a 24-well 522 plate. Rehydration steps were performed by washing for 5 minutes each in 75% methanol/PBST 523 (1× PBS + 0.1% Tween-20), 50% methanol/PBST, 25% methanol/PBST, and 5 times with 100% 524 PBST. The samples were permeabilized with 30 µg/ml proteinase K for 45 minutes at room 525 temperature, followed by post-fixation with 4% PFA for 20 minutes at room temperature, and 526 five washes in PBST for 5 minutes each. The samples were prehybridized in 500 µl of probe 527 hybridization buffer (Molecular Instruments) for 30 minutes at 37 °C. Hybridization was 528 performed by adding 2 pmol of each probe set to the hybridization buffer and incubating for 16 529 hours at 37 °C. Probe sets for panx2 were purchased from and designed by Molecular Instruments 530 using proprietary HCR methodology to detect and fluorescently label target RNA transcripts. To 531 maximize targeting, the probe was designed against shared regions of known variants found on 532 NCBI Gene and the Ensembl database. Each probe set consisted of 20 split-initiator probe pairs 533 per target and utilized a B1 amplifier with a 546 nm fluorophore label. To remove excess probes, 534 the samples were washed 4 times for 15 minutes each with a wash buffer (Molecular Instruments) 535 at 37 °C, followed by 2 washes of 5 minutes each with 5× SSCT (5× SSC + 0.1% Tween-20) at 536 room temperature. Pre-amplification was performed by incubating the samples for 30 minutes in 537 an amplification buffer (Molecular Instruments) at room temperature. The fluorescently labelled 538 hairpins (B2-488) were prepared by snap cooling: heating at 95 °C for 90 seconds and then 539 cooling to room temperature for 30 minutes. The hairpin solution was prepared by adding 10 µl 540 of the snap-cooled hairpins (3 µM stock concentration) to 500 µl of amplification buffer. The pre-541 amplification buffer was removed, and the samples were incubated in the hairpin solution for 16 542 hours at room temperature. Excess hairpins were washed three times with 5× SSCT for 20 543 minutes each. Following HCR RNA-FISH, larvae were stained with DAPI (1:12,000) overnight 544 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 26 at 4 °C, followed by three 10-minute washes in 5× SSCT. The samples were then stored in 5× 545 SSCT in the dark at 4 °C until imaging. 546 A z-mold or an 8-teeth mold for zebrafish larvae was 3D-printed using 2% low melting 547 agarose on glass-bottom culture dishes (MatTek, P35G-0-10C) (50). The larvae were positioned 548 in larva-shaped slots and embedded dorsal side down in 0.5% low melting agarose. Images were 549 acquired on the Nikon A1R confocal system with either a 20XW or 40XO objective using the 550 546 nm (B1 amplifier) and 468 nm lasers. Image analysis was performed using ImageJ (v2.9.0). 551 Quantitative real-time PCR (qRT-PCR) analysis. 552 Total RNA (1 µg) was extracted and purified from pools of 30 zebrafish larvae at 6 dpf using the 553 RNeasy Plus Mini Kit (Qiagen, Germantown, MD, United States of America) as per the 554 manufacturer’s protocols. Fish were homogenized by bead beating in 1xTE buffer (pH 8.0). 555 RNAs were reverse transcribed into cDNA using the iScript Reverse Transcription Supermix 556 (Bio-Rad, Mississauga, ON, Canada) as per the manufacturer’s instructions. Gene expression was 557 analyzed by qRT-PCR using the SsoAdvanced universal SYBR Green Supermix (Bio-Rad, 558 Mississauga, ON, Canada) in the CFX96™ Real-Time PCR Detection System (Bio-Rad, 559 Mississauga, ON, Canada). Thermal cycling was carried out for 39 cycles of the following: 94°C 560 for 30 s, 50°C for 30 s, and 72°C for 1 min. The housekeeping gene, 18S, was used to determine 561 the quality of the samples and for normalization purposes. CT values were averaged and exported 562 from the CFX Manager Software (Bio-Rad, Mississauga, ON, Canada). The fold-difference for 563 the relative gene expression was calculated in the Relative Expression Software Tool VS. 2009 564 (Pfaffl et al., 2002). Three technical replicates per gene were performed from three biological 565 replicates. 566 Immunohistochemistry. 567 Larvae (6 dpf) were humanely euthanized in MS-222 solution (0.02% w/v, Sigma-Aldrich) and 568 fixed in 4% paraformaldehyde (PFA) in 1xPBS overnight at 4oC, followed by cryoprotection in 569 30% sucrose in 1xPBS. After embedding in Tissue-Tek O.C.T compound 15µm sections were cut 570 on a cryotome (Thermofisher). Samples were washed three times for 5 min with 1xPBS 571 containing 0.1% Tween-20 (PBST) at RT. Unspecific binding sites were blocked with freshly 572 prepared 5% normal goat serum (NGS, Sigma-Aldrich) in PBST for 1hr at RT℃. Following 573 blocking, samples were incubated with a custom-made primary rabbit anti-Panx2 antibody 574 (1:200, catalog #A01856, GenScript) overnight at 4°C. Subsequent washes with PBST were for 575 one hour at 4 ℃. The Alexa 488 goat anti-rabbit secondary antibody (1:20,000 in 1% NGS PBST, 576 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 27 catalog#A11034, Life Technologies) was applied for one hr at RT. After 3 washes with PBST 577 followed by one wash with water, specimens were mounted on microscope slides using ProLong 578 Antifade with DAPI (Thermofisher). Confocal images were collected using LSM-ZEN2 software 579 (Zeiss LSM700 system; Carl Zeiss MicroImaging, Oberkochen, Germany) with Plan-580 Apochromat 20x/0.8 or Plan-Apochromat 63x/1.3 oil DIC M27 objectives. For comparison of 581 wild type and knockout tissues settings for raw image collection were identical. Post image 582 collection composite figures were created using Adobe Photoshop 2021. 583 Zebrafish Locomotion Assays. 584 The Zebrabox behavior system and the ZebraLab analytical suite (ViewPoint Life Technology, 585 Lyon, France, http://www.viewpoint.fr) were used for automated extraction of behavioral outputs 586 and video tracking. Tracking videos were recorded at 30 frames per second (fps) under infrared 587 (for light-OFF recording) or visible light (light-ON) illumination using a Point Grey Research 588 Dragonfly2 DR2-HIBW camera (Teledyne-FLIR, Burlington, ON, Canada). Inside the Zebrabox 589 a lightbox provided visible and infrared light from below for recordings using 0% to 30% light 590 intensity (visible range, 0-1200 lux). 6 dpf larvae were observed in 24 or 48-well plates 591 maintained at 28⁰C throughout the experiment. All experiments were performed between 12:00 to 592 2 pm, as larvae (6 dpf) activity was previously reported to reach a stable level by early afternoon. 593 Spontaneous free-swimming assay. 594 Larval swimming activity under constant light-ON/OFF conditions was tested using 24-well 595 plates. Locomotor behavior was tracked for 30 min. For analysis of locomotion, three thresholds 596 were defined: slow (20mm). The mean total distance 597 traveled (mm) and velocity (mm/sec) in two swim speeds (medium and fast) were used for 598 statistical analysis. 599 Visual Motor Response (VMR) assay. 600 Zebrafish larvae (n=24 per genotype) were acclimatized to darkness for 2 hours in a 48-well 601 plate. Baseline activity was recorded for 21 minutes under light-OFF conditions. For the first 602 experiment, larvae were subjected to alternating 20-minute light-on (30%, 1200 lux) and light-off 603 periods, totaling 1 hour and 41 minutes. For the second experiment, a modified protocol involved 604 incremental 10% light intensity increases every 20 minutes, culminating in 30% (1200 lux) light 605 intensity after 1 hour and 21 minutes. Data acquisition for all experiments was conducted using 606 the Quantization® mode of Zebrabox. Data collected at each time point included freeze count, 607 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 28 freeze duration, mid count, mid duration, burst count, and burst duration. For analysis, the 608 average duration (mean of mid and burst durations) was calculated for each larva. 609 Opto Motor Response (OMR). 610 OMR assays were performed using a custom-built system, which follows the design described 611 (Stih, Petrucco et al. 2019). Briefly, stimuli were presented from below using an ASUS P3B 800-612 Lumen LED portable projector (https://www.asus.com/ca-en/Projectors/P3B/). The fish 613 movements were recorded using a USB 3.1 high-speed camera (XIMEA GmbH, Germany) 614 equipped with a 35mm C Series Fixed Focal Length Lens (Edmund Optics Inc., USA). An 830 615 nm long-pass filter (Edmund Optics Inc., USA) was used to block infrared illumination coming 616 from a source at the bottom of the test environment. The videos were recorded using the XIMEA 617 Windows Software Package (https://www.ximea.com/support/wiki/apis/XI- 618 MEA_Windows_Software_Package). Visual stimuli were generated with an online stimulus 619 generator program called “Moving Grating” (available at 620 https://michaelbach.de/sci/stim/movingGrating/index.html). 621 During experiments four larvae in a 3 cm dish (Thermo Scientific) were allowed to 622 acclimatize for 5 min before starting the video recording. The visual stimulus consisted of 623 sequences of alternating white and black stripes generated with 64/128/256 pixels/cycle spatial 624 frequency. The speed rate was set to 72/144 pixels/sec. The contrast was set to 10%, where the 625 stripes appeared as white and grey, or 100%, where the stripes appeared as white and black. 626 Stimuli were presented to larvae (n = 4, for each of the four independent experimental repeats) 627 for 30 seconds in the left or right direction. Once the larvae oriented towards the moving stimulus 628 and initiated a sustained swimming motion in the direction of the stimulus, it was counted as a 629 positive response. The positive rate of response was used for analysis. This value was expressed 630 as a percentage: number of larvae that swam in the direction of the stimulus/ total number of 631 larvae in the dish. 632 Optical coherence tomography (OCT) assay. 633 A custom-developed spectral-domain optical coherence tomography (SD-OCT) system was built 634 in a Michelson configuration, employing a super luminescent laser diode centered at 1,310 nm (± 635 75 nm at 10 dB; Exalos, Switzerland) and a 2048-pixel line scan camera spectrometer with a 636 maximum acquisition rate of 147kHz (Wasatch Photonics; United States of America). A 50/50 637 fiber coupler splits the source light into the reference and sample arms. In the sample arm, the 638 output light illuminates the sample surface after passing through a reflective beam collimator 639 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 29 (Thorlabs; United States of America), a 2-DOF galvo mirror, and an objective lens (LSM02, 640 Thorlabs; United States of America). The 2-DOF Galvo mirror allows for collection of reflected 641 light from the sample while raster scanning sample surface. In the reference arm, a polarization 642 controller, a dispersion compensation block, and a gold-coated reference mirror were installed. 643 The back- reflected light of these two arms is subsequently merged after passing through the 644 beam splitter and redirected to the spectrometer by the optical circulator. The formed interference 645 pattern in the spectrometer is captured by a line scan camera. The captured signal is digitized and 646 sent to the computer for processing. To form an A-line (i.e., depth profile of sample reflections at 647 a given point on sample surface), the tomograms of the sample is background subtracted and 648 mapped to k-space before applying Fourier transformation to calculate depth profile of reflectors 649 in the z-space (physical depth space). Processes were repeated for data acquired during raster 650 scanning of beam on sample surface to eventually form 3-D OCT volumetric images of zebrafish 651 eye. The axial and lateral resolutions of the system in tissue were measured as 8.5 μm and 10 μm, 652 respectively. Before imaging, age-matched adult zebrafish were humanely euthanized using 653 MS222. Fish were then placed in a silicon mold to orient the eye toward OCT system’s objective 654 lens. To minimize specular reflections from sample surface, a thin layer of PBS (~80 μm) was 655 placed over the eye before imaging. All captured OCT images were calibrated for image pixel 656 size in axial and lateral directions before interrogation in ImageJ software and quantification of 657 the geometrical parameters. 658 Statistics and data reproducibility. 659 Statistical analyses were performed in GraphPad Prism VS10.2.3. Results are represented as the 660 mean ± standard deviation (SD) or standard error of the mean (SEM) for behavioural data. For 661 molecular analysis (RT-qPCR, RNA-seq), a minimum of n ≥ 3 independent experimental 662 replicates were generated. For behavioral testing G*power analysis determined the number of 663 larvae. The normality and homogeneity of the data variance were determined by the “Shapiro-664 Wilk Test” and “Levene’s Test.” Experimental groups were compared using unpaired t-tests or 665 Welch’s t-tests. A P value <0.05 was considered statistically significant. For all experiments, 666 sample sizes, statistical tests, and when appropriate P-values are indicated in the figure legends. 667 668 669 670 671 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 30

672 We thank two members of York University Zebrafish Vivarium, Janet Fleites-Medina, and 673 Veronica Scavo for outstanding zebrafish husbandry. We also wish to thank the Center for 674 Applied Genomics, SickKids, Toronto, ON, Canada for the RNA-seq service. 675 676 Funding 677 This research was supported by the Natural Sciences and Engineering Research Council 678 (NSERC) discovery grant RGPIN-2019-06378 (GRZ). 679 680 Author contributions 681 Conceptualization, RS, GRZ; data analysis, RS, GSZ, FN, SS, AB; investigation, all authors; 682 writing – original draft preparation, RS, GRZ; writing – review and editing, all authors; 683 visualization, RS, GSZ, FN, SS, AB; supervision, NT, GRZ; project administration, GRZ; 684 funding acquisition, GRZ. 685 686 Consent for publication 687 All authors have read and agreed to the published version of the manuscript. 688 689 Ethics approval 690 All animal work was performed at York University’s zebrafish vivarium and in an S2 biosafety 691 laboratory following the Canadian Council for Animal Care guidelines after approval of the study 692 protocol by the York University Animal Care Committee (GZ#2019-7-R2). 693 694 Data Availability 695 The RNA-seq data are deposited at the NCBI - Gene Expression Omnibus (GEO) database 696 repository (ID pending at time of submission). Additional information necessary for the 697 reanalysis of the data reported in this manuscript is available from the corresponding author upon 698 request. 699 700 Competing interests 701 The authors declare no competing interests. 702 703 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 31

& Correspondence 704 Correspondence and material requests should be addressed to Riya Shanbhag, Georg R. Zoidl. 705 706

707 1. Panchina Y , Kelmanson I, Matz M, Lukyanov K, Usman N, Lukyanov S. A ubiquitous family 708 of putative gap junction molecules. Current Biology. 2000 Jun 1;10(13):R473–4. 709 2. V ogt A, Hormuzdi SG, Monyer H. Pannexin1 and Pannexin2 expression in the developing 710 and mature rat brain. Molecular Brain Research. 2005 Nov 18;141(1):113–20. 711 3. Swayne LA, Sorbara CD, Bennett SAL. Pannexin 2 is expressed by postnatal hippocampal 712 neural progenitors and modulates neuronal commitment. J Biol Chem. 2010 Aug 713 6;285(32):24977–86. 714 4. Baranova A, Ivanov D, Petrash N, Pestova A, Skoblov M, Kelmanson I, et al. The 715 mammalian pannexin family is homologous to the invertebrate innexin gap junction proteins. 716 Genomics. 2004 Apr 1;83(4):706–16. 717 5. Bond SR, Wang N, Leybaert L, Naus CC. Pannexin 1 Ohnologs in the Teleost Lineage. J 718 Membrane Biol. 2012 Aug 1;245(8):483–93. 719 6. Zoidl G, Kremer M, Zoidl C, Bunse S, Dermietzel R. Molecular diversity of connexin and 720 pannexin genes in the retina of the zebrafish Danio rerio. Cell Commun Adhes. 2008 721 May;15(1):169–83. 722 7. Le Vasseur M, Lelowski J, Bechberger JF, Sin WC, Naus CC. Pannexin 2 protein expression 723 is not restricted to the CNS. Frontiers in Cellular Neuroscience [Internet]. 2014 [cited 2024 724 Feb 21];8. Available from: https://www.frontiersin.org/articles/10.3389/fncel.2014.00392 725 8. He Z, Zhao Y , Rau MJ, Fitzpatrick JAJ, Sah R, Hu H, et al. Structural and functional analysis 726 of human pannexin 2 channel. Nat Commun. 2023 Mar 27;14(1):1712. 727 9. Zhang H, Wang S, Zhang Z, Hou M, Du C, Zhao Z, et al. Cryo-EM structure of human 728 heptameric pannexin 2 channel. Nat Commun. 2023 Mar 3;14(1):1118. 729 10. Sanchez-Pupo RE, Johnston D, Penuela S. N-Glycosylation Regulates Pannexin 2 730 Localization but Is Not Required for Interacting with Pannexin 1. Int J Mol Sci. 2018 Jun 731 22;19(7):1837. 732 11. Ambrosi C, Gassmann O, Pranskevich JN, Boassa D, Smock A, Wang J, et al. Pannexin1 and 733 Pannexin2 channels show quaternary similarities to connexons and different oligomerization 734 numbers from each other. J Biol Chem. 2010 Aug 6;285(32):24420–31. 735 12. Penuela S, Bhalla R, Nag K, Laird DW. Glycosylation Regulates Pannexin Intermixing and 736 Cellular Localization. Mol Biol Cell. 2009 Oct 15;20(20):4313–23. 737 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 32 13. Boassa D, Nguyen P, Hu J, Ellisman MH, Sosinsky GE. Pannexin2 oligomers localize in the 738 membranes of endosomal vesicles in mammalian cells while Pannexin1 channels traffic to 739 the plasma membrane. Front Cell Neurosci. 2014;8:468. 740 14. Wicki-Stordeur LE, Boyce AKJ, Swayne LA. Analysis of a pannexin 2-pannexin 1 chimeric 741 protein supports divergent roles for pannexin C-termini in cellular localization. Cell Commun 742 Adhes. 2013 Aug;20(3–4):73–9. 743 15. Le Vasseur M, Chen VC, Huang K, V ogl WA, Naus CC. Pannexin 2 Localizes at ER-744 Mitochondria Contact Sites. Cancers (Basel). 2019 Mar 11;11(3):343. 745 16. Wang N, Wang C, Zhao H, He Y , Lan B, Sun L, et al. The MAMs Structure and Its Role in 746 Cell Death. Cells. 2021 Mar 16;10(3):657. 747 17. Lee A, Hirabayashi Y , Kwon SK, Lewis TL, Polleux F. Emerging roles of mitochondria in 748 synaptic transmission and neurodegeneration. Curr Opin Physiol. 2018 Jun;3:82–93. 749 18. Markovinovic A, Greig J, Martín-Guerrero SM, Salam S, Paillusson S. Endoplasmic 750 reticulum–mitochondria signaling in neurons and neurodegenerative diseases. Journal of Cell 751 Science. 2022 Feb 1;135(3):jcs248534. 752 19. Missiroli S, Perrone M, Gafà R, Nicoli F, Bonora M, Morciano G, et al. PML at 753 mitochondria-associated membranes governs a trimeric complex with NLRP3 and P2X7R 754 that modulates the tumor immune microenvironment. Cell Death Differ. 2023 Feb;30(2):429–755 41. 756 20. Young MD, Wakefield MJ, Smyth GK, Oshlack A. Gene ontology analysis for RNA-seq: 757 accounting for selection bias. Genome Biology. 2010 Feb 4;11(2):R14. 758 21. LeFauve MK, Rowe CJ, Crowley-Perry M, Wiegand JL, Shapiro AG, Connaughton VP. 759 Using a variant of the optomotor response as a visual defect detection assay in zebrafish. J 760 Biol Methods. 2021;8(1):e144. 761 22. Safarian N, Houshangi-Tabrizi S, Zoidl C, Zoidl GR. Panx1b Modulates the Luminance 762 Response and Direction of Locomotion in the Zebrafish. International Journal of Molecular 763 Sciences. 2021 Jan;22(21):11750. 764 23. Collery RF, Veth KN, Dubis AM, Carroll J, Link BA. Rapid, Accurate, and Non-Invasive 765 Measurement of Zebrafish Axial Length and Other Eye Dimensions Using SD-OCT Allows 766 Longitudinal Analysis of Myopia and Emmetropization. PLOS ONE. 2014 Oct 767 21;9(10):e110699. 768 24. Prochnow N, Hoffmann S, Vroman R, Klooster J, Bunse S, Kamermans M, et al. Pannexin1 769 in the outer retina of the zebrafish, Danio rerio. Neuroscience. 2009 Sep 15;162(4):1039–54. 770 25. Safarian N, Whyte-Fagundes P, Zoidl C, Grigull J, Zoidl G. Visuomotor deficiency in panx1a 771 knockout zebrafish is linked to dopaminergic signaling. Sci Rep. 2020 Jun 12;10(1):9538. 772 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 33 26. Kurtenbach S, Prochnow N, Kurtenbach S, Klooster J, Zoidl C, Dermietzel R, et al. 773 Pannexin1 Channel Proteins in the Zebrafish Retina Have Shared and Unique Properties. 774 PLOS ONE. 2013 Oct 23;8(10):e77722. 775 27. Cenedese V , de Graaff W, Csikós T, Poovayya M, Zoidl G, Kamermans M. Pannexin 1 Is 776 Critically Involved in Feedback from Horizontal Cells to Cones. Front Mol Neurosci. 777 2017;10:403. 778 28. Leung A, Ohadi D, Pekkurnaz G, Rangamani P. Systems modeling predicts that mitochondria 779 ER contact sites regulate the postsynaptic energy landscape. npj Syst Biol Appl. 2021 Jun 780 2;7(1):1–14. 781 29. Tsuboi M, Hirabayashi Y . New insights into the regulation of synaptic transmission and 782 plasticity by the endoplasmic reticulum and its membrane contacts. Proc Jpn Acad Ser B 783 Phys Biol Sci. 2021 Dec 10;97(10):559–72. 784 30. Liu H, Prokosch V . Energy Metabolism in the Inner Retina in Health and Glaucoma. 785 International Journal of Molecular Sciences. 2021 Jan;22(7):3689. 786 31. Hanna J, David LA, Touahri Y , Fleming T, Screaton RA, Schuurmans C. Beyond Genetics: 787 The Role of Metabolism in Photoreceptor Survival, Development and Repair. Front Cell Dev 788 Biol. 2022 May 18;10:887764. 789 32. Li S, Sheng ZH. Energy matters: presynaptic metabolism and the maintenance of synaptic 790 transmission. Nat Rev Neurosci. 2022 Jan;23(1):4–22. 791 33. Perez-Alvarez A, Yin S, Schulze C, Hammer JA, Wagner W, Oertner TG. Endoplasmic 792 reticulum visits highly active spines and prevents runaway potentiation of synapses. Nat 793 Commun. 2020 Oct 8;11(1):5083. 794 34. Ishikawa M, Iwamoto T, Nakamura T, Doyle A, Fukumoto S, Yamada Y . Pannexin 3 795 functions as an ER Ca2+ channel, hemichannel, and gap junction to promote osteoblast 796 differentiation. Journal of Cell Biology. 2011 Jun 20;193(7):1257–74. 797 35. Vanden Abeele F, Bidaux G, Gordienko D, Beck B, Panchin YV , Baranova A V , et al. 798 Functional implications of calcium permeability of the channel formed by pannexin 1. 799 Journal of Cell Biology. 2006 Aug 14;174(4):535–46. 800 36. Dvoriantchikova G, Ivanov D, Panchin Y , Shestopalov VI. Expression of pannexin family of 801 proteins in the retina. FEBS Letters. 2006;580(9):2178–82. 802 37. Berry V , Ionides A, Pontikos N, Georgiou M, Yu J, Ocaka LA, et al. The genetic landscape of 803 crystallins in congenital cataract. Orphanet J Rare Dis. 2020 Nov 26;15:333. 804 38. Song IK, Na S, Paek E, Lee KJ. Cataract-Associated New Mutants S175G/H181Q of βΒ2-805 Crystallin and P24S/S31G of γD-Crystallin Are Involved in Protein Aggregation by 806 Structural Changes. Int J Mol Sci. 2020 Sep 5;21(18):6504. 807 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint Shanbhag et al. 2024 34 39. Cermak T, Doyle EL, Christian M, Wang L, Zhang Y, Schmidt C, et al. Efficient design and 808 assembly of custom TALEN and other TAL effector-based constructs for DNA targeting. 809 Nucleic Acids Res. 2011 Jul;39(12):e82. 810 40. Bedell VM, Wang Y , Campbell JM, Poshusta TL, Starker CG, Krug II RG, et al. In vivo 811 genome editing using a high-efficiency TALEN system. Nature. 2012 Nov;491(7422):114–8. 812 41. Bedell VM, Ekker SC. Using Engineered Endonucleases to Create Knockout and Knockin 813 Zebrafish Models. In: Pruett-Miller SM, editor. Chromosomal Mutagenesis [Internet]. New 814 York, NY: Springer; 2015 [cited 2024 Oct 1]. p. 291–305. Available from: 815 https://doi.org/10.1007/978-1-4939-1862-1_17 816 42. Chen Y , Lun ATL, Smyth GK. From reads to genes to pathways: differential expression 817 analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline. 818 F1000Res. 2016;5:1438. 819 43. Chen Y , Chen L, Lun ATL, Baldoni PL, Smyth GK. edgeR 4.0: powerful differential analysis 820 of sequencing data with expanded functionality and improved support for small counts and 821 larger datasets [Internet]. bioRxiv; 2024 [cited 2024 Oct 1]. p. 2024.01.21.576131. Available 822 from: https://www.biorxiv.org/content/10.1101/2024.01.21.576131v1 823 44. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-824 seq data with DESeq2 [Internet]. bioRxiv; 2014 [cited 2024 Oct 1]. p. 002832. Available 825 from: https://www.biorxiv.org/content/10.1101/002832v3 826 45. Mudunuri U, Che A, Yi M, Stephens RM. bioDBnet: the biological database network. 827 Bioinformatics. 2009 Feb 15;25(4):555–6. 828 46. Yates B, Gray KA, Jones TEM, Bruford EA. Updates to HCOP: the HGNC comparison of 829 orthology predictions tool. Briefings in Bioinformatics. 2021 Nov 1;22(6):bbab155. 830 47. Szklarczyk D, Kirsch R, Koutrouli M, Nastou K, Mehryary F, Hachilif R, et al. The STRING 831 database in 2023: protein-protein association networks and functional enrichment analyses 832 for any sequenced genome of interest. Nucleic Acids Res. 2023 Jan 6;51(D1):D638–46. 833 48. Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, et al. Welcome to the 834 Tidyverse. Journal of Open Source Software. 2019 Nov 21;4(43):1686. 835 49. Choi HMT, Schwarzkopf M, Fornace ME, Acharya A, Artavanis G, Stegmaier J, et al. Third-836 generation in situ hybridization chain reaction: multiplexed, quantitative, sensitive, versatile, 837 robust. Development. 2018 Jun 15;145(12):dev165753. 838 50. Geng Y , Peterson RT. Rapid Mounting of Zebrafish Larvae for Brain Imaging. Zebrafish. 839 2021 Dec;18(6):376–9. 840 841 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 3, 2024. ; https://doi.org/10.1101/2024.10.01.616190doi: bioRxiv preprint