Neuroprotective Effect of Intraperitoneal Humanin-G in Retinal Degeneration of Royal College of Surgeons Rats

This study aimed to examine whether Humanin-G (HNG), a mitochondrial derived peptide with cytoprotective properties, could improve the retinal function and gene expression profiles after intraperitoneal injections to Royal College of Surgeons (RCS) rats with Retinal Pigment Epithelium (RPE) dysfunction and retinal degeneration. Starting at postnatal day 21 (p21), RCS rats received twice a week intraperitoneal injections of either Low Dose HNG (0.4 mg/kg), High Dose HNG (4mg/kg), or sham- saline for 1 or 4 weeks. Visual function was tested with full field scotopic & photopic electroretinography (ERG) and optokinetic testing (OKT) 1 and 4 weeks after first injection (WAFI). The rats were euthanized after the ERG and OKT (1 or 4 WAFI) and the dissected retinas and RPE were collected for RNA, cDNA and Quantitative Real- time PCR (qRT-PCR) analysis. The results of our study showed that high dose (4mg/kg) HNG at 4 WAFI was associated with the largest change in gene expression in the RPE and retina of treated animals, altering expression of genes involved in apoptosis, oxidative stress, inflammation and retinal/RPE function. Analysis of a and b waves from scotopic and photopic ERG showed no difference between either low or high dose of HNG and sham injection at 4 WAFI. However, at 4 WAFI, the visual acuity in rats treated with high dose HNG showed significant improvement as compared to the rats treated with low dose of HNG or saline. Most significantly, our findings support that HNG administered IP can modulate RPE/neuroretina cells and improve vision, thus may be a potential treatment for retinal degeneration diseases.

RCS model; Humanin-G; Mitochondria Derived Peptide; Retinal Degeneration .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 3 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 4

1 Mitochondrial DNA (mtDNA) (inherited maternally) can be classified into haplogroups, 2 according to the accumulation of single nucleotide polymorphisms (SNPs). 3 Mitochondria, with their small, but significant mtDNA, play important roles in human 4 diseases, even when the disease is not caused directly by mitochondrial mutations 1,2. 5 mtDNA variants can exert major influences on cellular energy pathways as well as 6 mediate expression of nuclear genes related to complement, inflammation, apoptosis, 7 autophagy, methylation and cell signaling transduction pathways 3-11. 8 Mitochondrial Derived Peptides (MDPs) are biologically active, short peptides (20-27 9 AAs), which are encoded from short open reading frames (sORF) in the human mtDNA 10 12-15. Humanin (HN), a 24-amino acid MDP, has been shown in vitro and in vivo to have 11 anti-apoptotic, neuro-protective properties supporting cell survival 14-17. Circulating HN 12 levels decrease with age and are associated with age-related diseases in mice, rats and 13 humans 18,19. Furthermore, offspring of centenarians have higher HN levels compared to 14 the aging population18. Altogether, our findings suggest that retaining HN levels with age 15 may promote healthy aging. HN has been shown to be cyto-protective in Alzheimer’s 16 disease, atherosclerosis, myocardial and cerebral ischemia, and Type 2 diabetes 17,20. 17 HN and SHLP2 (Small Humanin-Like Peptide 2) can lower plasma amino acids and lipid 18 metabolites associated with metabolic aging disorders 21. 19 In vitro studies have shown protective effects of HN against hypoxia-induced toxicity in 20 RGC-5 cells 22. Pre- or co-treatment with HN protected SH-SY5Y neuroblastoma cells 21 from toxicity induced by silver nanoparticles (AgNPs) 23. The HNG peptide, the HN 22 analogue, improved mitochondrial function and decreased cell death in PC12 cells 23 stressed with amyloid β 25-35 peptides 24. In human ARPE-19 retinal cybrids, HNG 24 reduced amyloid- β induced cell stress 25. HNG has a glycine replacing serine at 25 position 14, making it 1000-fold more potent in its protective functions. 26 An in vitro animal study showed rescue of cortical neuron viability after NMDA damage 27 following 10 µmol/L HN treatment 26. In vivo, HN-treated diabetic mice demonstrated 28 improved glucose tolerance and lower pancreatic-beta cell apoptosis due to activation 29 of the Stat3 pathway 27. HNG treatment improved the cognitive functions in aging 30 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 5 C57B1/6N mice and reduced the levels of inflammatory markers 28. While studies have 31 shown HN or HNG improve retinal cells in vitro 16,29,30, there are no studies showing 32 their in vivo effects in retinal degeneration models. 33 The Royal College of Surgeons (RCS) rat is an established model for retinal 34 degeneration. This model has dysfunctional RPE due to a MerTK mutation leading to 35 the accumulation of outer segment debris and photoreceptor death31, is associated with 36 ER stress 32, and activation of mitochondrial and cytosolic calpain 33. The RCS model 37 has been frequently used in the preclinical testing of drugs and stem cells for retinal 38 degeneration 34-40. Recent studies have identified early mitochondrial dysfunction in 39 RCS rats 41 and mtDNA deletions in the late stages of RCS retinal degeneration (>200d) 40 42. 41 The present study investigates the effects of Humanin-G (HNG) intraperitoneal (IP) 42 injections on the Royal College of Surgeons rat. We used intraperitoneal injections to 43 deliver either low dose (0.4 mg/kg) or high dose (4mg/kg) HNG to the RCS rats and 44 assessed the molecular changes in RPE cells and neuroretina, as well as vision change 45 at 1 WAFI and 4 WAFI. 46 47

48 HNG Peptide Preparation: The custom synthesized HNG peptide with guaranteed TFA 49 removal services was purchased from Genscript Inc. (Piscataway, NJ) and diluted with 50 sterile 0.9% saline before the injection. PLGA (Poly(lactic-co-glycolic acid)) and PVA 51 (Polyvinyl alcohol) polymers were purchased from Akina Inc (West Lafayette, IN). 52 Methylene Chloride was purchased from Sigma-Aldrich (St. Louis, MO). Acetonitrile, 53 HPLC water, LC-MS water, and formic acid were purchased from Fisher Scientific 54 (Waltham, MA). Analytical grade solvents were used in all experiments. 55 HNG Treatments for RCS Rats with Retinal Degeneration 56 Experimental animals : For all experimental procedures, animals were treated in 57 accordance with the NIH guidelines for the care and use of laboratory animals, the 58 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 6 ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and under 59 a protocol approved by the Institutional Animal Care and Use Committee (AUP-17-097 60 and AUP 20-055). Non-nude rats of the immunodeficient Royal College of Surgeons 61 (RCS) rat strain with the MERTK mutation (dysfunctional RPE) were used for the 62 experiments 39. 63 Intraperitoneal Injection of HNG : Starting at postnatal day 21 (p21), rats were given 64 an intraperitoneal injection of either “Low Dose” (0.4mg/kg) HNG; “High Dose” (4mg/kg) 65 HNG; or sham-saline. Injections using either 22g or 27g needles were repeated twice 66 weekly. The experiments were concluded after either 1 WAFI (short-term study) or 4 67 WAFI (long-term study) after the ERG and OKT tests were completed. 68 Full field Scotopic & Photopic Electroretinography (ERG) : For low dose injection, 69 the rats were tested with ERG, using a Rodent ERG system (Diagnosys Celeris), at 4 70 WAFI. For high dose injection, the rats were evaluated for changes in visual function at 71 1 WAFI and 4 WAFI. After overnight dark-adaptation, rats were anesthetized by 72 Ketamine/Xylazine (40-55 mg/kg Ket, 6 – 7.5 mg/kg Xyl), then the eyes were treated 73 with 0.5% tetracaine (Bausch & Lomb) and 1% atropine eye drops (Akorn 74 Pharmaceuticals, Lake Forest, IL). and then anesthesia was maintained by 1% 75 isoflurane. The rats were placed onto a heating pad on a rodent exam table and 76 positioned in front of a monocular mini-ganzfeld photostimulator. After applying GenTeal 77 Lubricant Eye Gel, electrodes were placed onto the corneal surface. Visual responses 78 with scotopic and photopic stimuli were recorded for both eyes, to obtain simultaneous 79 recordings of the same animal using established protocols. 80 Optokinetic Testing (OKT): For low dose injection, the rats were tested with OKT at 4 81 WAFI. For high dose injection, 1 and 4 weeks after the first injection of HNG (low dose 82 and high dose) or saline, the visual acuity of RCS nude rats was measured by recording 83 videos of optomotor responses to a virtual cylinder with alternating black and white 84 vertical stripes (Optomotry, Cerebral Mechanics Inc., Alberta, Canada). The testing was 85 described previously36,43. 86 Rats were dark-adapted for at least 1 hour prior to testing. Optomotor responses were 87 recorded at 6 different spatial frequencies for one minute per frequency by testers 88 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 7 blinded to the experimental condition. Both the left and right eyes were tested by 89 alternating the direction of the moving stripes. Two independent tests were performed 90 at each time point, with at least one hour time in between; one test going from lowest to 91 highest frequency, and the other from highest to lowest frequency. The best visual 92 acuity of the two tests was used for analysis. All tests were video recorded and 93 evaluated off-line by two independent observers blinded to the experimental conditions. 94 Any discrepancies between the two observers resulted in re-analysis of videos by a 95 third observer, and data discussion before giving a final score, and prior to decoding the 96 experimental condition. 97 Isolation of tissue, RNA and Amplification of cDNA : Upon removal of the anterior 98 segment of the eye including the cornea, iris and lens, the retina was removed and 99 immediately snap frozen in liquid nitrogen. The RPE was isolated by utilizing the 100 “Simultaneous RPE cell Isolation and RNA Stabilization” (SRIRS method) 44. The 101 posterior eye cup including the choroid, sclera and RPE was removed and immediately 102 placed in a microcentrifuge tube containing 400µl of RNAprotect cell reagent (Qiagen). 103 After a minimum of 10 minutes, the tube was agitated and RPE cells observed to be 104 released into solution. The eye cup was then removed and the dissociated RPE cells 105 pelleted. This provided selective isolation of the RPE and protection against RNA 106 degradation in one step. RNA was isolated from RPE cells and the neuroretina using 107 the PureLink RNA mini kit. cDNA was produced using SuperScript IV VILO 108 (ThermoFisher) according to the manufacturer’s instructions. 109 Quantitative Real-time PCR (qRT-PCR) Analyses : qRT-PCR was performed on 110 individual samples using PowerUp SYBR Green Master Mix (ThermoFisher Scientific) 111 on an Applied Biosystems QuantStudio 5 Real-Time PCR system real time quantitative 112 PCR detection system. Primers (QuantiTect Primer Assay (Qiagen) or KiCqStart 113 Primers (Sigma)) were used to analyze 24 different genes: Inflammation and oxidative 114 stress ( Ddit3, HSP α 5, Il6, Il1 β , Tnf α , Ccl2 ); Apoptosis ( Bax, Casp3, Casp7, Casp9, 115 Bcl2l1, Bcl2l13); Antioxidation ( Sod2); Photoreceptor markers ( Crx, Gngt1, Nrl, Rom1, 116 E2f1); and RPE markers ( Best1, Rlbp1, Rpe65, Tjp1) (Table 1 ). Primers were 117 standardized with the HMBS housekeeping gene. All analyses were performed in 118 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 8 triplicate. The fold values were calculated using the 2^(- ΔΔ Ct) formula. All fold value 119 changes of the HNG-IP treatments are calculated compared with saline control. 120 Statistical Analysis: Statistical analysis was performed using Graphpad Prism, using 121 t-tests (paired and unpaired), Mann-Whitney U, one-way ANOVA for multiple 122 comparisons and/or Tukey post-hoc analysis. The significance level was set at p<0.05. 123 124

125 Intraperitoneal HNG Injection Studies 126 Animals were weighed prior to each injection (HNG or saline) and any side effects were 127 recorded. No side effects were observed throughout the study. No change in behavior 128 was found. We found no significant changes in the weights between the saline and 129 HNG (low or high dose) groups at the end of the study (Figure 1). 130 Retinal Gene Expression Levels after Intraperitoneal HNG Injection 131 Gene Expression for IP Low Dose HNG (0.4mg/kg) at 1 WAFI: The RPE showed a 132 significant increase in expression of Tnfα (1.789-fold P=0.0023) and a significant 133 decrease in expression of Rlbp1 (0.6781-fold, P=0.0271) in animals treated with low 134 dose HNG for one week (Supplemental Table 1). There were no statistically significant 135 changes in gene expression in the neuroretina of animals treated with low dose HNG for 136 one week. 137 Gene Expression for IP Low Dose HNG (0.4mg/kg) at 4 WAFI: The neuro-retina 138 showed decreased levels of proapoptotic Ddit3 (0.7658-fold, P=0.0159) and increased 139 levels of Crx (1.468-fold, P=0.0357) and Tjp1 (1.567-fold, P=0.0286) ( Figure 2 , 140 Supplemental Table 2 The RPE cells demonstrated no significant changes in gene 141 expression. 142 Gene Expression for IP High Dose HNG (4mg/kg) at 1 WAFI: I n t h e R C S r a t s 143 treated with High Dose HNG (n=9), the RPE showed decreased expression levels of 144 apoptotic genes (Bcl2l1, 0.587-fold, P=0.0190) and Casp7 (0.5737-fold, P=0.0321) 145 compared to the saline treated samples (n=9) ( Figure 3A, Supplemental Table 3) The 146 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 9 neuro-retina showed increased levels of Crx (1.309-fold, P= 0.0103) (Figure 3B). There 147 were no statistically significant changes in other genes’ expression. 148 Gene Expression for IP High Dose HNG (4mg/kg) at 4 WAFI: The RPE showed 149 increased expression levels of pro-apoptotic genes ( Ddit3, 1.842-fold, P=0.0317; 150 Casp3,1.65-fold, P=0.0079); inflammation genes (Il6, 1.782-fold, P=0.0286); antioxidant 151 genes (Sod2, 1.797-fold, 0.0286); and RPE genes important for normal function ( Best1, 152 1.698-fold, P=0.0317 (Figure 4a , Supplemental Table 4). There was also a lower 153 expression level of E2f1 (0.569-fold, P=0.016) in the RPE. The neuroretina showed 154 increased levels of proapoptotic Casp7 (1.309-fold, P=0.0159) ( Figure 4b , Table 4). 155 There were no statistically significant changes in the expression of other genes. 156

of ERG at 1 and 4 WAFI : The ERG analyses were performed first on Low 157 Dose HNG (0.4mg/kg), and Sham-Saline at 4 WAFI. There was no difference between 158 Low Dose HNG and Saline in a-wave and b-wave of scotopic and photopic full field 159 ERG (Figure 5). Later, High Dose HNG (4mg/kg) and Saline groups were tested with 160 ERG. At 1 and 4 WAFI, there was no significant difference in scotopic and photopic a-161 wave (data not shown), or scotopic and photopic b-wave between High dose HNG and 162 Sham-Saline groups (Figure 5). 163 Visual function improvement evaluated by optokinetic testing (OKT) : At 4 WAFI, 164 optokinetic testing in RCS rats showed that the visual acuity of eyes with high dose 165 HNG (n=11) showed significant improvement from that of saline groups (n=9. P<0.05). 166 (Figure 6) There was no significant difference in any other time points or doses tested. 167 168

169 HNG has demonstrated cytoprotective effects in human cybrid cell lines and animal 170 models of neurodegeneration. In a study evaluating an analog of Humanin in 171 ameliorating streptozotocin-induced diabetic nephropathy in Sprague Dawley rats, 172 [S14G]-humanin was administered intra-peritoneal (once daily for a course of sixteen 173 weeks) at a dosage of 0.4 mg/Kg of body weight 45. Another study reported that 174 humanin exhibits neuroprotective effects in vitro in human cell culture models and 175 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 10 enhances cognition in vivo in aged mice (Kelvin Yen 2018). In the present study, we 176 used intraperitoneal injections to deliver either low dose (0.4 mg/kg) or high dose 177 (4mg/kg) HNG to the RCS rats and showed that HNG IP injections can modulate in RPE 178 and neuroretina gene expression levels and in the short-term, improve retina function in 179 the RCS rat model. 180 181 Low Dose HNG: 182 RPE: The Low dose IP injections at 1 WAFI resulted in upregulation at Tnf α and 183 downregulation of Rlbp1 in the RPE cells, whereas by 4 WAFI, no significant changes in 184 gene expression were observed across all pathways. In a rat model of ulcerative colitis, 185 intraperitoneal HNG (10 or 20 µM) reduced expression of Tnfα and Il-1β and decreased 186 caspase 3 activities46. Similarly, in a rat model of pituitary tumors, 5 µM HN inhibited the 187 proapoptotic effect of Tnf α on cultured anterior pituitary cells 47. In ApoE deficient mice 188 on a high-cholesterol diet, after 16 weeks of intraperitoneal HN (4 mg/kg/day) treatment, 189 Tnfα , MCP-1 and osteopontin were downregulated, along with decreased apoptosis, 190 compared to the untreated ApoE-/- mice 48. Similar decline in the pro-inflammatory Tnfα 191 and IL-1β were seen after HN treatment in HUVEC culture 49. 192 TNF-α is a key pivotal mediator of inflammation in AMD, known for disrupting 193 endothelial integrity and initiating inflammatory pathways that accelerate disease 194 progression 50,51. However, studies assessing TNF- α level in aqueous humour or in 195 serum remain inconclusive 52-55. Although TNF-α is theoretically expected to be elevated 196 in AMD, most comparisons between AMD and control groups have shown no significant 197 differences 53,55. furthermore, TNF- α was significantly higher in patients who improved 198 than in those who deteriorated 53. 199 Overall, systemic and local measurements of TNF- α remain inconsistent. While TNF- α 200 inhibition is effective for some ocular diseases such as uveitis 56, clinical outcomes for 201 anti–TNF-α therapy in AMD, whether systemic 57 or intravitreal 58,59, have been variable. 202 Future studies should clarify these discrepancies. 203 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 11 Our observation of elevated TNF α at 1 WAFI may reflect either the uncertain role of 204 TNFα or the short duration and low dose of treatment, particularly since no change was 205 observed at 4 WAFI. Further investigation is warranted. 206 RLBP1 encodes cellular retinal-binding protein (CRALBP), an 11-cis-retinal–binding 207 protein expressed in Müller and RPE cells 60, and is essential for the visual cycle. 208 RLBP1 mutations cause three clinical subtypes—Bothnia dystrophy, retinitis punctata 209 albescens, and Newfoundland rod-cone dystrophy 61. Gene therapy using AAV vectors 210 expressing RLBP1 has improved the visual cycle in Rlbp1−/− mice 62 and in patients 211 with RLBP1-associated retinal dystrophy 63. We observed downregulation of Rlbp1 in 212 RPE cells at 1 WAFI but not at 4 WAFI. Further studies are needed. 213 214 Neuro-retina: Interestingly, at 1 WAFI, no gene expression changes were seen for any 215 of the neuroretinal genes, suggesting that either higher HNG doses or a longer 216 exposure is required to modulate the neuroretinal genes. By 4 WAFI, Ddit3 was 217 significantly downregulated (p=0.0159), while Tjp1 (p=0.0286) and Crx (p=0.0357) were 218 upregulated. The DNA damage inducible transcript 3 (DDIT3, also known as 219 transcription factor C/EBP-homologous protein, CHOP) is an ER stress effector that 220 promotes ER stress and the subsequent inflammation and apoptosis induced by 221 lipopolysaccharide (LPS) exposure 64,65. 222 DDIT3 is implicated in neurodegeneration. Human RPE cybrids harboring AMD 223 mitochondria show markedly elevated DDIT3 (also known as CHOP, 633.9%), Caspase-224 3 (125.7%). Caspase-7 (181.3%) levels, with reduced E2F1 (66.2%) and SOD2 levels 225 (23.1%) compared to the cybrids with age-matched normal mitochondria. HNG 226 treatment significantly reduces these pro-apoptotic and ER stress markers 25. In primary 227 human RPE cell cultures, HN pretreatment decreases ER-stress induced apoptosis by 228 elevating mitochondria glutathione and lowering CHOP (DDIT3) levels 30. Similarly, 229 Sreekumar et al. demonstrated that HN localizes to RPE cells and protected against 230 oxidative stress by improving mitochondrial biogenesis and bioenergetics 16. The 231 observed reduction of Ddit3 expression by HNG in our study further supports DDIT3’s 232 role in retinal degeneration and highlights HNG’s therapeutic potential. 233 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 12 Tight junction protein-1 (TJP1), also known as the tight junction marker ZO-1, is a 234 peripheral membrane phosphoprotein of the zonula occludens family 66 and a key 235 component linking junctional proteins to the cytoskeleton to maintain epithelial integrity 236 67. Reduced ZO-1 expression in RPE cells is associated with epithelial–mesenchymal 237 transition and compromised retinal support 67,68, leading to pathologies such as age-238 related macular degeneration (AMD) and diabetic retinopathy 67,68. Thus, our data, 239 HNG-induced upregulation of Tjp1, suggests its potential to preserve RPE integrity and 240 treat AMD. 241 Crx (cone–rod homeobox) is an transcription factor critical for photoreceptor 242 development and differentiation, regulating numerous phototransduction and metabolic 243 genes 69. Dysregulated Crx–Otx2 interaction contributes to early-onset retinal 244 degeneration 70. Mutations in upstream regulators of Crx–Otx2 disrupt gene activation 245 balance, causing aberrant photoreceptor gene expression and apoptosis 70. Crx also 246 partners with Nrl and Nr2e3 to promote rod-specific gene expression while suppressing 247 cone genes during differentiation 69. Loss or mutation of Crx results in photoreceptor 248 degeneration and disorders such as Leber congenital amaurosis (LCA) and cone–rod 249 dystrophy 71. In our study, HNG-induced upregulation of Crx further supports its 250 potential as a therapeutic agent for AMD. 251 High Dose HNG: 252 As expected, high-dose HNG altered the expression of more genes than low-dose 253 HNG. 254 RPE: At 1 WAFI, HNG-treated RPE cells showed downregulation of the pro-apoptotic 255 gene Casp7 (p = 0.0321) and the anti-apoptotic Bcl2l1 (p = 0.0189). While the latter 256 appears inconsistent with HNG’s beneficial effects, Bcl2l1 has multiple cellular roles. 257 BCL2L1 encodes the anti-apoptotic protein BCL-XL 72, which provides a strong selective 258 advantage to hPSCs under stress conditions such as thawing, expansion, and cloning 259 72. BCL-XL localizes mitochondria to prevent cytochrome C release and caspase-260 dependent apoptosis through interactions with other BCL-2 family members 73,74. 261 Beyond its anti-apoptotic role, BCL-XL contributes to metabolism, mitochondrial 262 dynamics, and calcium homeostasis 73,74, and has been implicated in regulating cell fate 263 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 13 by inhibiting neuroectodermal differentiation 75. In RPE tissue and ARPE19 cells, BCL-264 XL supports cell survival 76, but persistent expression in senescent RPE cells 265 contributes to apoptosis resistance and tissue dysfunction in aging and AMD models 266 75,77,78. Inhibiting BCL-XL in senescent RPE is therefore a promising therapeutic 267 approach for AMD 77,78. Given its roles in mitochondrial and calcium homeostasis—268 critical to RPE physiology—the downregulation of Bcl2l1 by HNG may also contribute to 269 its protective effects. 270 At 4 WAFI, high-dose HNG altered more genes than the low dose (6 vs 0) in RPE cells, 271 including upregulation of pro-apoptotic (Casp3), ER stress (Ddit3), inflammatory (Il-6), 272 antioxidant (Sod2), and RPE marker (Best1) genes, while E2f1 expression was 273 significantly reduced. E2F1 is a transcription factor regulating genes involved in DNA 274 replication, repair, the cell cycle, and apoptosis 79. Its biological functions are modulated 275 by post-translational modifications 80. IL-6 is a pleiotropic cytokine that primarily acts as 276 a pro-inflammatory factor 81. In RPE cells, IL-6 inhibits Sirt1 through PI3K/AKT/mTOR-277 mediated phosphorylation, suppressing the E2F1/HMGA1/G6PD pathway and thereby 278 increasing oxidative stress and cell death 82. Although HNG increased Casp3, Ddit3, 279 and Il-6 while reducing E2f1, these changes may reflect alternative functions and 280 warrant further investigation. 281 Superoxide dismutases (SODs) are critical antioxidant enzymes that prevent oxidative 282 stress by metabolizing reactive oxygen species (ROS) 83. SOD2, localized in 283 mitochondria, detoxifies superoxide radicals generated during oxidative phosphorylation 284 at complexes I and III, converting them into hydrogen peroxide and oxygen 84. 285 Upregulation of Sod2 reduces oxidative stress, supporting mitochondrial protection, a 286 key factor in neurodegenerative disease mitigation 85. Thus, HNG-induced Sod2 287 upregulation suggests a potential therapeutic mechanism. 288 The Best gene family encodes Ca²⁺ -activated anion channels with diverse physiological 289 roles in multiple organs, including the eye 86. Best1 is predominantly expressed in RPE 290 cells and is genetically associated with various retinal degenerations 87,88, 291 encompassing over 350 known mutations leading to progressive vision loss and 292 blindness 89,90. Given Best1’s critical function in retinal health and its link to untreatable 293 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 14 vision disorders, the observed Best1 upregulation by HNG further supports its 294 therapeutic potential in retinal degenerative diseases. 295 296 Neuro-retina: At 1 WAFI, the neuroretina exhibited elevated Crx expression 297 (p=0.0103), while other genes remained comparable to saline-treated controls. As 298 described above, Crx regulates multiple photoreceptor-specific genes essential for 299 photoreceptor development and differentiation. Its upregulation supports the beneficial 300 effect of HNG in retinal degeneration (RD) diseases. At 4 WAFI, Casp7 expression 301 increased (p=0.0159), whereas other genes remained unchanged relative to controls. 302 Overall, more genes were modulated in RPE cells (six genes) than in neuroretina (one 303 gene), likely due to the reduced neuroretinal thickness in RCS rats by 30 days or a 304 greater sensitivity of RPE cells to HN-mediated protection. Sreekumar et al. reported 305 that polarized human RPE cells contain high intracellular HN co-localized with 306 mitochondria 16. Exogenous FITC-labeled HN peptides showed robust uptake and 307 mitochondrial translocation in RPE cells compared to controls, indicating a 308 mitochondria-targeted protective role. Moreover, RPE cells express all three HN 309 receptors—ciliary neurotrophic factor receptor (CNTFR α ), transmembrane glycoprotein 310 gp130 (GP130), and cytokine receptor WSX1 16,30—which may account for the stronger 311 HNG effects observed in RPE cells than in neuroretina. 312 Functional Testing with High Dose HNG: 313 No significant differences were observed in scotopic or photopic a- and b-wave 314 amplitudes between high-dose, low-dose HNG, and sham-saline groups. The absence 315 of detectable ERG changes at 1 WAFI may reflect the single early injection and the time 316 required for HNG to exert measurable effects. By 4 WAFI, the rapid retinal degeneration 317 characteristic of the RCS rat resulted in undetectable ERG responses across all groups; 318 therefore, a potential beneficial effect of HNG cannot be excluded, as ERG may lack the 319 sensitivity to detect subtle improvements. Interestingly, while no improvement in OKT 320 was seen at 1 WAFI following high-dose intraperitoneal HNG, visual function improved 321 at 4 WAFI. Molecular analyses revealed increased expression of several genes in both 322 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 15 neuroretina and RPE tissues, suggesting that intraperitoneally administered HNG may 323 cross the blood–retina barrier or induce systemic changes affecting these tissues. 324 Overall, intraperitoneal HNG administration in the RCS rat model (a) modulated gene 325 expression in RPE and neuroretina, indicating possible blood–retina barrier penetration 326 or systemic regulation, and (b) improved visual performance at 4 WAFI as measured by 327 OKT. Given the variability in molecular responses and the limited clinical applicability of 328 intraperitoneal delivery for retinal diseases, future studies will focus on evaluating 329 alternative HNG delivery methods in this model of retinal degeneration. 330 331 Acknowledgments: The authors want to acknowledge the contribution of Dr. M.C. 332 Kenney, UCI (who passed away in December 2023) to the initiation, funding and 333 completion of this project. The authors would like to acknowledge the technical 334 assistance of Shari R. Atilano, Mithalesh Singh (Kenney lab), and Seiler lab staff 335 members Robert Sims, Karla Echeverria, Jay Santoso, Reeva Reyes, Jeffrey Delgado; 336 and students Alice Avetyan, Candice Wu, Caroline Lee, Jennie Xu, Kevin Wu, Angel 337 Blanquel, Adeline Cheng, Winnie Luong, Julianne Frances Agapinan, Ceci Zhang, Ani 338 Petrosyan, Ani Khachigian, Ryan Pavey. Christopher Quimpo, Maihan Phan, and other 339 students. This research was funded by the National Eye Institute, grant R01 EY027363 340 and to a small part with grant R01 EY031834. The authors acknowledge support to the 341 Gavin Herbert Eye Institute at the University of California, Irvine from an unrestricted 342 grant from Research to Prevent Blindness and from NIH core grant P30 EY034070. 343 344 Conflicts of Interest: The authors declare no conflict of interest. 345 Author contributions: 346 BL: performed experiments, collected & analyzed data; wrote the paper; approved the 347 final version 348 KS: performed experiments, collected & analyzed data; wrote the paper; approved the 349 final version 350 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 16 MO: performed experiments, collected & analyzed data; wrote the paper; approved the 351 final version 352 NI: performed experiments, collected data; approved the final version 353 MJS: performed experiments, analyzed data; wrote the paper; edited figures; provided 354 funding; approved the final version 355 356 Figure Captions: 357 Figure 1. No change in animal weight at 4 WAFI. 358 359 Figure 2. Effect of Low dose (0.4mg/kg) HNG on gene expression in neuroretina 4 360 WAFI. Expression of Ddit3 was significantly decreased and expression of Tjp1 and Crx 361 were significantly increased 4 WAFI with low dose HNG compared with saline in the 362 neuroretina 363 Figure 3. Effect of High dose (4.0mg/kg) HNG on gene expression in the RPE and 364 Neuroretina 1 WAFI. (a) Expression of Bcl2l1 and Casp7 were both significantly 365 decreased 1 WAFI with high dose HNG compared with saline in the RPE. (b) 366 Expression of Crx was significantly increased 1 WAFI with high dose HNG compared 367 with saline in the Neuroretina. 368 Figure 4. Effect of High dose (4.0mg/kg) HNG on gene expression in the RPE and 369 Neuroretina 4 WAFI. (a) Expression of Ddit3 , Il6, Sod2, Casp3, and Best1 were all 370 significantly increased, and E2f1 were significantly decreased 4 WAFI with high dose 371 HNG in the RPE. (b) Expression of Casp7 was significantly increased 4 WAFI with high 372 dose HNG in the neuroretina. 373 Figure 5. ERG Results from low (0.4mg/kg) and high dose (4mg/kg) IP Injections at 1 374 and 4 WAFI. No significant change was found. 375 Figure 6. Effect of high dose IP injecti ons of HNG on OKT at both 1 WAFI and 4 376 WAFI. OKT was improved at 4 WAFI. Asterisk indicates significant difference at P>0.05. 377 Table 1. Description of Genes Analyzed by qRT-PCR 378 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 17 379 Supplemental Table 1. qPCR Results from low dose (0.4mg/kg) IP Injections 1 WAFI 380 (RPE and Neuroretina) 381 382 383 Supplemental Table 2. qPCR Results from low dose (0.4mg/kg) IP Injections 4 WAFI 384 (RPE and Neuroretina) 385 386 Supplemental Table 3. qPCR Results from high dose (4mg/kg) IP Injections 1 WAFI 387 (RPE and Neuroretina) 388 389 Supplemental Table 4 . qPCR Results from high dose (4mg/kg) IP Injections 4 WAFI 390 (RPE and Neuroretina) 391 392

393 394 395 1 Cz arneck a, A. M . & Bartnik, E . The r ole of th e mit och ondrial g enom e in ag eing an d 396 c ar cinogenesis . J Agi ng R es 2011 , 136435 (2011). h t tps://doi.or g /10.4061/2011/136435 397 2 W allace, D . C. & F an, W . Ener g e tics, epigenetics, mit och ondrial genetics. Mi t o cho ndrio n 10, 12-398 31 (2010). h t tps://doi.or g /10.1016/j.mito.2009.09.006 399 3 A tilano , S . R. e t a l . Dif f er e n ti al Epig en e tic Sta tus and R espo nses t o St r essor s b etw ee n R etin al 400 Cybrids Cells with Afric an v er sus E ur opea n Mi t ochond rial DN A: I nsigh ts int o Diseas e 401 Susceptibiliti es. Cells 11 (2022). h t tps://doi.or g /10.3390/ cells11172655 402 4 A tilano , S . R . e t a l . Mi t och ondrial D NA variants c a n medi at e meth yl a tio n st atus o f inflamma tion, 403 angiog en esis and signaling g enes . Hum Mol Ge net 24, 4491-4503 (2015 ). 404 h ttps://doi.or g /10.1093/hmg /ddv173 405 5 Dolink o , A. H., Chw a, M., A tilano , S. R. & K enney , M. C. Afric an and Asian Mit ochondrial DNA 406 Haplogr oups Conf er Resis tance Ag ain s t Diabetic Stresses on R e tinal Pigment Epithelial Cybri d 407 Cells In Vitr o. M ol Neur obio l 57, 1636-1655 (2020). h t tps://doi.or g /10.1007/ s12035-019-01834-z 408 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 18 6 Dolink o , A. H . e t a l . Hypo xia-induced t r anscriptional di f f e r ences in Afric an a n d Asian v er sus 409 Eur opean di abetic cybrids . Sci R ep 13, 3 818 (2023). h t tps://doi.o r g /10.1038/ s41598-023-30518-410 x 411 7 K enney , M. C. e t a l . Inhe ri t ed mi t ochond rial DNA v a riants can af f ec t c ompl ement , inflamma tio n 412 and apop t osis pathw a y s: insigh ts into mit ochond rial-nuclea r int er ac tions. H u m Mol Gene t 23, 413 3537-3551 (2014 ). h t tps://doi.o r g /10.10 93/hmg /ddu065 414 8 K enney , M. C. e t a l . Molecular and bioenerg etic dif f e r ences bet w ee n cells with Afric an v er sus 415 Eur opean inh eri t ed mi t ochond rial DNA h aplogr oups: implica tio ns f or popula ti on susceptibility to 416 diseases. Bio chi m Biop h ys Act a 1842 , 208-219 (2014 ). 417 h ttps://doi.or g /10.1016/j.bbadis .2013.1 0.016 418 9 K enney , M. C. e t a l . Mitochondri al DN A v aria n ts medi at e ene r gy pr oduc tion and expression 419 lev els f or CFH, C3 and EFEMP1 g enes: implic a ti ons f or ag e-related macular de genera tion . PLoS 420 One 8 , e54339 (2013). h t tps://doi.o r g /10.1371/journal.pon e.0054339 421 10 Malik, D . e t a l . Human r etinal tr ansmit ochond rial cybrids with J or H m tDNA haplogr oups 422 r espond d if f e r e n tly t o ult r a violet radiatio n: implica tions f or r etin al diseas es. PLoS One 9 , e9900 3 423 (2014). h t tps://doi.or g /10.1371/journal. pone.0099003 424 11 Nashine, S. & K enney , M. C. E f f ec ts of Hu manin G (HN G) on angiogenesis a nd ne u r odegenera ti on 425 mark er s in Age-r elat ed Macul ar Dege ner ation (AMD). Mi t oc ho ndrio n 74, 1 01818 (2024). 426 h ttps://doi.or g /10.1016/j.mito.2023.11 . 001 427 12 Gong , Z. & T asset, I. Humanin enhan ces the cellular r espons e t o s t r ess by activ a tion of 428 chaper on e-mediated au t ophagy . Onc o t ar get 9 , 10832-10833 (2018). 429 h ttps://doi.or g /10.18632/ onc o t a r g e t.24 396 430 13 Guo , B. e t a l . Humanin peptid e suppres ses apopt osis by in terf ering with Bax ac tiv a ti on. Na t u re 431 423 , 456-461 (2003 ). h t tps://doi.o r g /10. 1038/na ture01627 432 14 Lee, C. e t a l . IGF- I regulat es the ag e-d ep ende n t signaling p eptid e humanin . A gin g Cell 13, 958-433 961 (2014). h t tps://doi.or g /10.1111/ acel.12243 434 15 T ajima, H. e t a l . E vidence f or in viv o pr oduction of Humanin peptid e, a neurop r otectiv e f ac t or 435 ag ains t Alzh eimer 's disease- r el a ted insults. Neur osci Let t 324 , 22 7-231 (2002 ). 436 h ttps://doi.or g /10.1016/ s0304-3940(02) 00199-4 437 16 Sr eek um ar , P . G. e t a l . The Mitochondr ial-Deriv ed P eptide Humanin Pr ot ects RPE Cells Fr om 438 Oxidativ e S tress, Senesce nce, and Mi tochondrial Dy s functio n. Inves t O phth al mol Vis Sci 57, 439 1238-1253 (2016 ). h t tps://doi.o r g /10.11 67/iovs.15-17053 440 17 Y en, K., Lee, C., Meht a, H . & Cohen, P . T he emer ging r ole of th e mit ochon drial-d eriv ed pep tid e 441 humanin in stress resist ance . J M o l E ndo c r i no l 50, R11-19 (2013). h ttps://doi.org /10.1530/ JME-442 12-0203 443 18 Muzumdar , R . H . e t a l . H umanin: a nov el cen t r al r egul a tor of p eriph er al insulin ac tion. PLoS On e 444 4 , e6334 (2009). h t tps://doi.o r g /10.1371/journal.pone .0006334 445 19 Y en, K. e t a l . The mit ochondri al deri v ed peptide humanin is a r egulat or of lif espan and 446 healthsp an. A gin g (Albany NY) 12, 11185-11199 (2020 ). h t tps://doi.o r g /10.18632 / aging.103534 447 20 Kin, T . e t a l . Humanin e xpr ession in sk elet al muscles of pa tients with chr onic pr ogr essiv e extern al 448 oph th almoplegia . J Hu m Ge net 51, 555-558 (2006). h t tps://doi.or g /10.1007/ s10038-006-0397-2 449 21 Meht a, H . H. e t al. Metabolomic pr ofil e o f diet-induced obesi ty mice in r esp onse to humanin and 450 small humanin-lik e peptide 2 tr e a tme n t . Metabol omi cs 15, 88 (2019 ). 451 h ttps://doi.or g /10.1007/ s11306-019-154 9-7 452 22 Men, J ., Zhang , X., Y ang , Y . & Gao , D . An AD-r ela ted neu r opr o tector rescues t r ans f ormed ra t 453 r etinal g anglion cells fr om CoCl(2) -induced apoptosis. J Mol Ne urosci 47, 1 44-149 (2012 ). 454 h ttps://doi.or g /10.1007/ s12031-011-970 1-5 455 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 19 23 Gurun a th an, S. , Jey ar aj, M. , K ang , M. H. & Kim, J . H. Mit ocho ndrial P eptid e Humanin Pr ot ec ts 456 Silv er Nano par ticles-Induce d Neuro t o xici ty in Human Neur obla s toma Cancer Cell s (SH-S Y5Y ). In t 457 J Mol Sci 20 (2019). h t tps://doi.o r g /10.33 90/ijms20184439 458 24 Jin, H. e t a l . Pr ot ec tiv e e f f ects of [Gly14]-Humanin on bet a-am yloid-induced PC1 2 cell dea th by 459 pr ev e n ting mit och ondrial dy s funct ion. Neuro che m Int 56, 417-423 (2010 ). 460 h ttps://doi.or g /10.1016/j.neui n t .2009.1 1.015 461 25 Nashine, S. e t a l . Hum anin G (H NG) pr o t ects a g e-related macular degen er ation (AMD) 462 tr ansmitochondri al ARPE-19 cybrids fr o m mit ochond rial an d cellula r dama g e. Cell Death Dis 8 , 463 e2951 (2017). h t tps://doi.o r g /10.1038/ cddis.2017.348 464 26 Cui, A. L. e t a l . Humani n r escues culture d r a t c o rti c al neu r ons fr om NMD A-induc ed t oxicity not 465 by NMD A r eceptor . Scie n tific W orldJ our nal 2014 , 341529 (2014). 466 h ttps://doi.or g /10.1155/2014/341529 467 27 Hoang , P . T . e t a l . The neur osur viv al f act or Humani n inhibits beta-cell apoptosis via signal 468 tr ansduce r and activ ator of tr anscrip tio n 3 activ a tion an d dela y s and ameli or at es diabe tes in 469 nonobese diabe tic mice. Met a bolism 59, 343-349 (2010). 470 h ttps://doi.or g /10.1016/j.metabol.2009 . 08.001 471 28 Y en, K. e t a l . Humanin Pr ev e n ts Age-R elat ed Cognitiv e Decl ine in Mic e and is A ssocia ted with 472 Impr ov ed Cognitive Age in Humans. Sci R ep 8 , 14212 (2018). h ttps://doi.or g /1 0.1038/ s41598-473 018-32616-7 474 29 Ma tsu nag a, D . e t a l . Humanin Pr ot ects RPE Cells fr om Endoplasmic R eticulum Str ess-Induce d 475 Apoptosis by Upr egul a ti on of Mi t och ondrial Glutathion e. PLoS One 11, e0165150 (2016). 476 h ttps://doi.or g /10.1371/journal.pon e.01 65150 477 30 Minas y an, L., Sreek umar , P . G., Hi n ton , D . R. & K annan, R. Pr otectiv e M ech anisms of the 478 Mit och ondrial-Derived P eptide Humanin in Oxida tive and Endoplasmic R eticulu m Str ess in RPE 479 Cells. Oxid Med Cell Lo ngev 2017 , 16752 30 (2017). h t tps://doi.or g /10.1155/2017/1675230 480 31 D'Cruz, P . M. e t a l . Mut ation of the r ec ep t or tyrosine kinase g en e Mertk in the r e t inal dy s tr op hic 481 RC S rat . Hum M ol Ge net 9 , 645-651 (2000). h t tps://doi. or g /10.1093/hmg /9.4.645 482 32 K r oeg er , H. e t a l . Inducti on of endoplas mic r eticulum stress g e nes, BiP and cho p, in genetic an d 483 en vir onment a l models of r e tinal d eg en e r a tio n. In vest Op hth almol Vis Sci 53 , 759 0-7599 (2012). 484 h ttps://doi.or g /10.1167/iovs.12-10221 485 33 Mizuk oshi, S. e t a l . Activa tion of mit och ondrial c alpain and r ele ase of ap optosis-inducing f act o r 486 fr om mit ochondri a in R CS ra t r e tina l deg en er ation . Exp E ye R es 91, 35 3-361 (2010 ). 487 h ttps://doi.or g /10.1016/j.ex e r .2010.06 .0 04 488 34 Cuenc a, N ., Pinilla, I., Sa uv e, Y . & Lun d, R. E arly chan g es in s ynaptic c onn ec tivity f ollowing 489 pr ogr essive pho t orecep t or d eg en era tio n in R CS r a ts . E ur J Neur osci 22, 105 7-1072 (2005 ). 490 h ttps://doi.or g /10.1111/j.1460-9568.200 5.04300.x 491 35 LaV ail, M. M. Leg acy of the R CS ra t: imp act of a seminal s tu dy on r etin al cell biol ogy and r etina l 492 deg en er ativ e diseas es. Prog Brai n R es 131 , 617-627 (2001). h t tps ://doi.or g / 10.1016/ s0079-493 6123(01)31048 -8 494 36 Lin, B. e t a l . R etina Or g anoid T r anspla n ts Dev elop Phot orecep t or s and Impr ov e Visual Function in 495 R CS Rats With RPE Dy s function . In vest Op h t halm ol Vis Sci 61, 34 (2020). 496 h ttps://doi.or g /10.1167/iovs.61.11.34 497 37 Oz aki, T . e t a l . I n tr a vit r e al injection or t o pic al ey e-dr op appli c ation of a mu-c alpa in C2L domain 498 peptide p r o t ec ts ag ains t pho t orecep t or cell de a th in R oy al Colle g e of Su r g e ons' ra ts, a model o f 499 r etini tis pigmentosa. Bioc him Biop h ys A cta 1822 , 178 3-1795 (2012). 500 h ttps://doi.or g /10.1016/j.bbadis .2012.0 7.018 501 38 Thomas, B. B. e t a l . Co-gr afts of Human Embry onic St em Cell Derived R etin a Or g anoids an d 502 R etinal Pigmen t Epi thelium f or R etinal R ec onstruc tion in Immunod e ficient R eti nal Deg enerate 503 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 20 R oy al Colleg e of Sur g e ons Ra ts . F ro n t N euros ci 15, 752958 (2021). 504 h ttps://doi.or g /10.3389/fnins.2021.7529 58 505 39 Thomas, B. B. e t a l . A new im munode ficien t r e tinal dy s t r ophic r at model f or tr ansplanta tio n 506 s tudies using human-de riv ed cells. G r a e f e s A r c h C l i n E x p O p h t h a l m o l 256 , 21 13-2125 (2018 ). 507 h ttps://doi.or g /10.1007/ s00417-018-413 4-2 508 40 V ollr ath, D . e t a l . Corr ection of the r eti nal dy s tr oph y pheno type of the R CS r at by vir al g ene 509 tr ans f er of Me rtk. Proc Na tl Acad Sci U S A 98, 12584-12589 (2001). 510 h ttps://doi.or g /10.1073/pnas.22136419 8 511 41 Zhu, Z. H., Fu , Y ., W eng , C. H. , Zhao , C. J. & Y in, Z. Q. Pr oteomic profiling of earl y deg ene r ative 512 r etina of R CS r a ts . I n t J O phth almol 10, 8 78-889 (2017 ). h t tps://doi.o r g /10.18240 /ijo.2017.06.08 513 42 Br a v o-Nuev o, A ., Williams, N., G eller , S. & S t one , J . Mitochondri al dele tions in normal an d 514 deg en er ating r at r etin a. A dv Exp Me d Bi ol 533 , 241-248 (200 3). h t tps://doi.or g / 10.1007/978-1-515 4615-0067-4_30 516 43 Seiler , M. J. e t a l . Vision R ec overy a nd Connectivity by F e t al R e tinal Shee t T r ans plant ation in a n 517 Immunode ficie n t R e tinal Degenerat e R at Model . I n ve s t O ph t h al m o l V i s S c i 58, 614-630 (2017 ). 518 h ttps://doi.or g /10.1167/iovs.15-19028 519 44 Xin-Zhao W ang , C., Zhang , K., Ar edo , B. , Lu, H. & Ufr et-Vincen ty , R. L. Nov el meth od f or the r apid 520 isola ti on of RPE cells specific ally f or RNA e x tr acti on and analy sis. Ex p E ye R es 102 , 1-9 (2012 ). 521 h ttps://doi.or g /10.1016/j.ex e r .2012.06 .0 03 522 45 Moin, H. e t a l . E f f ectiveness of a nalo g of Humanin in ameliora ti ng s t r ep t o zot ocin-induc ed 523 diabetic nep hr op a th y in Spr agu e Dawley r a ts. P eptides 165 , 1 71014 (2023). 524 h ttps://doi.or g /10.1016/j.pep tides.2023 . 171014 525 46 Gultekin, F . A . e t a l . E f f ec ts of humanin on e xp erime nt al c oli tis induced by 2,4,6- trinit r obe nz ene 526 sulphonic acid in r a ts. Sau di J Gastroe nt erol 23, 105-111 (2017 ). 527 h ttps://doi.or g /10.4103/ sjg.SJG_318 _16 528 47 Gott a r do , M. F . e t a l . A n ti apop t otic f act o r humanin is e x pr ess ed in normal and tu mor al pitui t a ry 529 cells and pr o t ects them fr om TNF-alp ha-induced apop t osis. PLoS One 9 , e1 11548 (2014). 530 h ttps://doi.or g /10.1371/journal.pon e.01 11548 531 48 Zhang , X. e t a l . Humanin prev e n ts intra-r enal micr ov ascular r emod eling and i nflamma tion in 532 h yper cholester ol emic ApoE de ficie n t mice. Lif e Sci 91, 199 - 206 (2012 ). 533 h ttps://doi.or g /10.1016/j.lf s.2012.07 .01 0 534 49 W ang , X. e t a l . Humanin p r ev e n ts high gluc ose-induced mon ocyte adhesion t o e ndotheli al cells 535 by t ar g eting KLF2. Mol Imm un ol 101 , 245-25 0 (2018 ). 536 h ttps://doi.or g /10.1016/j.molimm.2018. 07.008 537 50 Rajesw ar en , V . e t a l . Elev at ed t umor necrosis f act or alpha and v ascula r endo thelia l gr owth f act o r 538 in in termedi at e a g e-related macular de g enera ti on and g e ogr aphic atr oph y . F r on t O ph t ha l m ol 539 (Lausanne) 4 , 1356957 (2024). h t tps://do i.or g /10.3389/f oph t .2024.1356957 540 51 T w ar og , M. e t a l . TNF alp ha induced by DNA-sensing in macr ophag e c omp r omises r etin al 541 pigmen t epi theli al (RPE) barrier function. Sci R ep 13, 1 4451 (2023). 542 h ttps://doi.or g /10.1038/ s41598-023-416 10-7 543 52 Khan, A . H. et a l. The e f f ec t of s y st emic l ev els of TNF-alpha and c ompl eme n t p a t hw a y activity on 544 out comes of VE GF inhibition in n e ov ascular AMD . E ye (Lond) 36, 219 2-2199 (2022 ). 545 h ttps://doi.or g /10.1038/ s41433-021-018 24-3 546 53 Nassar , K. e t a l . Se rum cyt okines as bio mark er s f or ag e-r elat e d macul ar degene r a tio n. Gr aef es 547 Arch Clin Ex p Op h t hal mol 253 , 699-704 (2015). h t tps://doi.o r g /10.1007/ s00417-014-2738-8 548 54 R e z a r-Dr eindl, S. e t a l . The I n tr a ocular Cyt okine Pr ofile and Therapeu tic R espons e in P er sis t e nt 549 Neov ascular Ag e-R e lat e d Macula r Deg en er ation. In v es t O phth almol Vis Sci 57, 41 44-4150 (2016). 550 h ttps://doi.or g /10.1167/iovs.16-19772 551 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 21 55 Spindler , J., Zandi, S., Pfist er , I . B ., Gerh a r dt, C . & Ga rweg , J. G. Cy t okine pr ofiles in the aque ous 552 humor and s erum of p a ti ents with d ry a nd treated w e t age-r elat ed macul ar d egenera tion . PLoS 553 One 13, e0203337 (2018). h t tps://doi.o rg /10.1371/journal.pone .0203337 554 56 Cor der o-Coma, M . & Sobrin, L. A nti-t umor necr osis f actor-alpha therapy in uv eitis. Sur v 555 Ophth almol 60, 575-589 (2015). h t tps://doi.or g /10.1016/j.sur v op h thal.2015.06 .0 04 556 5 7 Fe r n a n d ez - V e ga , B . e t a l . Block ade of T umor Nec r osis F ac t o r-Alpha : A R ol e f or Adalimumab in 557 Neov ascular Ag e-Related Macul ar Dege ner ation Re fr ac t o ry t o A n ti-A ngiog en esis Ther apy? Case 558 R ep Op h t halm ol 7 , 154-162 (2016). h t tps://doi.or g /10.1159/000445102 559 58 R osenf eld, P . J. & Goodman , K. W . When is of f-label drug use in th e p a tie n t 's best in teres t ? Am J 560 Ophth almol 147 , 761-763 (2009). h t tps:/ /doi.or g /10.1016/j.ajo.2009.01 .011 561 59 Theodossiadis, P . G ., Liar ak os, V . S. , Sfik akis, P . P ., V er g ados, I. A. & Theod ossiadis, G. P . 562 In tra vitreal adminis tra t ion of the anti-tu mor necrosis f act o r agent infliximab f or n eov ascular a g e-563 r ela ted macul a r degenera ti on. A m J O p h t h a l m o l 147 , 825-830, 830 e821 (2009). 564 h ttps://doi.or g /10.1016/j.ajo.2008 .12.00 4 565 60 P alcz ew ski, K. & Ki ser , P . D . Shedding new ligh t on the g enera ti on of the visual chr omophor e . 566 Proc Natl Acad Sci U S A 117 , 19629-19638 (2020). h t tps://doi.or g /10.1073/pnas.2 008211117 567 61 Bocquet, B . e t a l . R e tinitis Punctat a Albescens and RLBP1-Allied Pheno typ es: Phenotype-568 Geno type Correla tion and N a tural Hist o r y in the Aim of Gen e Ther apy . Op hth alm ol Sci 1 , 100052 569 (2021). h t tps://doi.or g /10.1016/j. x ops.2 021.100052 570 62 Choi, V . W . e t a l . A A V-medi a ted RLBP1 g ene t her apy impr ov es t he ra te of dark adapta tio n in 571 Rlbp1 knock out mice. Mol Ther Met ho ds Clin Dev 2 , 15022 (2015 ). 572 h ttps://doi.or g /10.1038/m tm .2015.22 573 63 K v an t a, A . e t a l . Int e rim saf ety and ef fic acy of g ene therapy f or RLBP1-associat ed r etin al 574 dy s tr o ph y: a phase 1/2 trial. Nat Comm u n 15, 7438 (2024). h t tps://doi.o r g /10.10 38/ s41467-024 -575 51575-4 576 64 Gorb a tyuk, M. S ., S t a rr , C. R. & Go rbatyu k, O . S. End oplasmic r e ticulum str ess : Ne w insigh ts into 577 the p a th og en esis and treatme n t of retin al degenera tive diseas es. Prog R etin E ye R es 79, 100860 578 (2020). h t tps://doi.or g /10.1016/j.pret eyer es .2020.100860 579 65 Li, H. e t a l . High dose expression of hem e o xigenase-1 induc es r e tinal d eg en era ti on through ER 580 s tress-r elat ed DDIT3. Mol Neur ode gene r 16, 16 (2021). h t tps://doi.o r g /10.118 6/ s13024-021 -581 00437-4 582 66 F anning , A. S ., J ameson, B. J., J esai tis, L . A . & And er son, J. M . The tigh t junc tio n pr o t ein ZO-1 583 es tablishes a link be tw e en th e tr ansmem br ane pr o tein occludin and the ac tin cyto sk elet on . J Bi ol 584 Chem 273 , 29745-29753 (1998). h t tps://doi.or g /10.1074/jbc.273.45 .29745 585 67 Geo r giadis, A. e t a l . The tight junction a ssocia ted signalling pr o teins Z O-1 and ZONAB r egul ate 586 r etinal pigment epit helium home o s t asis in mice . PLoS One 5 , e 15730 (2010). 587 h ttps://doi.or g /10.1371/journal.pon e.00 15730 588 68 Zhou, M. e t a l . R ole of Epithelial-M e sench ymal T r ansition in R etinal Pigmen t Epith elium 589 Dy s function. F ro nt Cell Dev Biol 8 , 501 (2020). h t tps://doi. or g /10.3389/f cell.2020. 00501 590 69 Zheng , Y . & Chen, S. T r anscriptional precision in phot o r ecep t o r dev elopme n t and diseases - 591 Lessons fr om 25 y ear s of CRX r e sear ch. F ro n t Cell Neur osci 18, 13 47436 (2024). 592 h ttps://doi.or g /10.3389/fncel.2024.1347 436 593 70 Langouet, M. e t a l . M uta ti ons in BC OR, a c o-repressor of CRX/O TX2, a r e associat ed with e arly-594 onset r e tinal de g en er ation. Sci Adv 8 , e a bh2868 (2022). h t tps://doi.or g /10.1126/sciadv .abh2868 595 71 P an, D ., Zhang , X., Jin, K. & Jin, Z. B . CRX haploinsuf ficiency c ompr omises photor ec eptor 596 pr ecur so r tr anslo c ation and dif f e r e n ti a ti on in human r etinal or g anoi ds. St em Ce ll R es Ther 14, 597 346 (2023). h t tps://doi.or g /10.1186/ s13287-023-03590-3 598 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 22 72 A v ery , S. e t a l . BCL-XL media t es th e s t r ong selective adv ant a g e of a 20q11.2 1 amplific a tio n 599 c ommonly f ound in human embry onic s t em cell cultu r es . St em Cell R ep orts 1 , 3 79-386 (2013 ). 600 h ttps://doi.or g /10.1016/j.st emcr .2013 .1 0.005 601 73 Chong , S. J . F . e t a l . Nonc a nonic a l Cell F at e R egul a ti on by Bcl-2 Pr ot eins . T rends Cell Biol 30, 537-602 555 (2020). h t tps://doi.or g /10.1016/j. t cb .2020.03.004 603 74 Gr oss, A. & K a tz, S . G. Non-apop t o tic functions of BCL-2 f amily pr ot eins. Cell Death Diff er 24, 604 1348-1358 (2017 ). h t tps://doi.o r g /10.10 38/ cdd.2017.22 605 75 Abbasaliz a deh, S . & Bahar v and, H. T echnologic al p r ogr ess an d challenges t ow a r ds cGMP 606 manuf acturing of human pluri po t e n t stem cells based therapeu tic pr oduc ts f or allog en eic an d 607 aut ol ogous cell ther a pies. Biotech nol A dv 31, 1600-1623 (2013). 608 h ttps://doi.or g /10.1016/j.bio t echa dv .20 13.08.009 609 76 Zhang , N. e t a l . The import ance of Bcl-xL in the surviv al of human RP E cells. In ves t Ophth almol 610 Vis Sci 48, 3846-3853 (2007 ). h t tps://doi. or g /10.1167/iovs.06-1145 611 77 Lee, K. S., Lin, S., Copland, D . A., Dick, A. D . & Liu, J . Cellular senescence in the aging r etina and 612 dev elopme n ts of seno ther ap ies f or ag e- r el a ted macular d eg en er ation . J Ne uroin flamma tion 18 , 613 32 (2021). h t tps://doi.or g /10.1186/ s12974-021-02088-0 614 78 R yu, W ., P ark, C. W ., Kim, J ., L ee, H . & Chung , H. The Bcl-2/Bcl-xL Inhibi t or AB T-263 A ttenuates 615 R etinal Deg en er ation by Selectiv ely Ind ucing Apopt osis in Senescent R etinal Pigmen t Epitheli al 616 Cells. Mol Cells 46, 420-429 (2023). h t tps://doi.or g /10.14348/molcells.2023.2188 617 79 DeGr egori , J . & Johnson, D . G. Dis tinct and Ov erlapping R oles f or E2F F amil y Member s in 618 T r anscription, Pr olif era ti on and A popt osis. C urr Mol M ed 6 , 73 9-748 (2006). 619 h ttps://doi.or g /10.2174/1566524010606070739 620 80 Putz er , B. M. & Eng elmann, D . E2F1 apop t osis c ou n ter att ack ed : evil s trik es back. T rends Mol Me d 621 19, 89-98 (2013 ). h t tps://doi.o r g /10.101 6/j.molmed.2012.10.009 622 81 Klein, R . et al . Ma rk er s of inflammation, o xidativ e s tress, an d endo theli al dy s func tion and the 20-623 y ear cumul a tive incidenc e of e arly age-r e lat ed macul ar d eg en er ation : th e B ea v e r Dam E y e Study . 624 J AMA O phth almol 132 , 446-455 (2014). h ttps://doi.or g /10.1001/jamaop h thalmo l.2013.7671 625 82 Gong , C. e t a l . IL-6-induced acetylation of E2F1 ag gr avate s o x i d at i ve d a m a g e o f re t i n a l p i g m e n t 626 epithe lial cell line . Exp E ye R es 200 , 1082 19 (2020). h t tps://doi.or g /10.1016/j.ex e r .2020.108219 627 83 Miller , A . F . Super oxide dismutases: anci en t en z ymes and new insigh ts. FEBS Let t 586 , 585-595 628 (2012). h t tps://doi.or g /10.1016/j. f ebslet . 2011.10.048 629 84 Z elk o , I. N., M ariani , T . J. & F olz, R. J . Su per o xid e dismut as e multigene f amily: a c omparison of 630 the CuZn-SOD (SOD1), Mn-SOD (SOD2) , and E C-SOD (SOD3) g ene s tructures, ev olution , and 631 e xp r ession . F ree R adic Biol Me d 33, 337-349 (2002). h t tps://doi.or g / 10.1016/ s0891-632 5849(02)00905 -x 633 85 Mueller-B uehl, A. M . e t a l . R ed uced R et inal Deg eneration in an Oxid a tive St r ess Or g an Culture 634 Model th r ough an iNOS-I nhibit o r . Biol ogy (Basel) 10 (2021 ). 635 h ttps://doi.or g /10.3390/biology10050383 636 86 Owji, A. P ., Kittredge, A., Zha ng , Y . & Y an g , T . Structu r e and Func tion of the Bestrophin f amily of 637 c alcium-activ at ed chlorid e channel s. Cha nnels (Austin) 15, 604- 623 (2021). 638 h ttps://doi.or g /10.1080/19336950.2021. 1981625 639 87 Johnson, A . A. e t a l . Bestr ophi n 1 and r etinal disease . Prog R etin E ye R es 58, 45-69 (2017 ). 640 h ttps://doi.or g /10.1016/j.pr e t eyer es .201 7.01.006 641 88 K ong , M. D . e t a l . Div e r g e n t Manif est ations in Bi allelic V er sus M onoall elic V ariants of RP1-, 642 BE ST1-, and PR OM1-Associat e d R etinal Disor de r s. I n t J Mol Sci 26 (2025 ). 643 h ttps://doi.or g /10.3390/ijms26146615 644 89 Y ang , T ., Jus tus, S., Li, Y . & T san g , S. H. B E ST1: the Bes t T ar g e t f or Gene and Cell Ther apies. Mo l 645 Ther 23, 1805-1809 (2015 ). h t tps://doi.o r g /10.1038/m t.2015.177 646 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint 23 90 K ar aosmanoglu, B ., Im r en, G., Utin e, E. , T a ylan Sek eroglu, H. & T askir an , E. Z. Allele-specific 647 an tis ense oligonucleo tides f or the t r e a tme n t of BE ST1-r el a ted domina n tly inheri t ed r e tina l 648 diseases: An in vitr o mode l. Exp E ye R es 241 , 109 833 (2024 ). 649 h ttps://doi.or g /10.1016/j.ex e r .2024.109 833 650 651 .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint Table 1. Description of Genes Analyzed by qRT-PCR Symbol Gene Name Gene RefSeq Number Functions INFLAMMATION PATHWAY Ddit3 DNA-damage inducible transcript 3 NM_001109986 This gene encodes a member of the CCAAT/enhancer -binding protein (C/EBP) family of transcription factors. The protein functions as a dominant- negative inhibitor by forming heterodimers with other C/EBP members and preventing their DNA binding activity. The protein is implicated in adipogenesis and erythropoiesis, is activated by endoplasmic reticulum stress, and promotes apoptosis. Fusion of this gene and FUS on chromosome 16 or EWSR1 on chromosome 22 induced by translocation generates chimeric proteins in myxoid liposarcomas or Ewing sarcoma. Multiple alternatively spliced transcript variants encoding two isoforms with different length have been identified. Il6 interleukin 6 NM_012589 Enables cytokine activity and interleukin-6 receptor binding activity. Involved cellular response to lipid; positive regulation of cell communication; and response to peptide hormone. Located in cytoplasm and extracellular space. Il1b interleukin 1 beta NM_031512 Enables cytokine activity. Involved in positive regulation of intracellular signal transduction; regulation of gene expression; and regulation of neurogenesis. Located in extracellular space. Tnf tumor necrosis factor NM_012675 Enables tumor necrosis factor receptor binding activity. Involved in positive regulation of cell communication; positive regulation of macromolecule metabolic process; and positive regulation of neuron death. Located in several cellular components, including external side of plasma membrane; extracellular space; and neuronal cell body. Hspa5 heat shock protein family A (Hsp70) member 5 NM_013083 Enables misfolded protein binding activity and unfolded protein binding activity. Involved in cellular response to cAMP; cellular response to gamma radiation; and cellular response to metal ion. Acts upstream of or within response to endoplasmic reticulum stress. Located in membrane; mitochondrion; and smooth endoplasmic reticulum. Ccl2 C-C motif chemokine ligand 2 NM_031530 Enables chemokine activity and heparin binding activity. Involved in cellular response to cytokine stimulus; cellular response to lipid; and leukocyte migration. Located in several cellular components, including perikaryon; perinuclear region of cytoplasm; and rough endoplasmic reticulum. Colocalizes with C-fiber. .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint ANTIOXIDANT PATHWAY Sod2 superoxide dismutase 2 NM_017051 Enables several functions, including identical protein binding activity; manganese ion binding activity; and superoxide dismutase activity. Involved in hydrogen peroxide biosynthetic process; negative regulation of membrane hyperpolarization; and positive regulation of hydrogen peroxide biosynthetic process. Located in mitochondrial nucleoid. APOPTOSIS PATHWAY Bax bcl2 associated x NM_017059 Enables BH domain binding activity; chaperone binding activity; and heat shock protein binding activity. Involved in glial cell apoptotic process; mitochondrion organization; and response to corticosterone. Located in cytosol; mitochondrial outer membrane; and perinuclear region of cytoplasm. Bcl2l1 bcl2-like 1 NM_001033672 Enables several functions, including GTPase binding activity; MDM2/MDM4 family protein binding activity; and cysteine- type endopeptidase inhibitor activity involved in apoptotic process. Involved in cellular response to cytokine stimulus; cellular response to lipid; and positive regulation of transport. Located in cytosol; mitochondrial outer membrane; and presynapse. Colocalizes with clathrin-coated pit; mitochondrial membrane; and synaptic vesicle membrane. Bcl2l13 bcl2-like 13 NM_001398849 Over-expression leads to apoptosis. Mitochondria specific protein. Casp3 caspase 3 NM_012922 Enables death receptor binding activity; phospholipase A2 activator activity; and protease binding activity. Involved in luteolysis; nervous system development; and positive regulation of amyloid-beta formation. Located in cytosol; membrane raft; and neuro nal cell body. Part of death-inducing signaling complex. Casp7 caspase 7 NM_022260 Enables cysteine-type peptidase activity. Involved in aging; leukocyte apoptotic process; and striated muscle cell differentiation. Located in cytosol and intracellular membrane- bounded organelle. Casp9 caspase 9 NM_031632 Enables cysteine-type peptidase activity. Involved in glial cell apoptotic process; response to estradiol; and response to ischemia. Located in cytosol; mitochondrion; and nucleus. PHOTORECEPTOR MARKERS Crx cone-rod homeobox NM_021855 Enables DNA binding activity. Involved in circadian rhythm and positive regulation of photoreceptor cell differentiation. Located in nucleus. E2f1 E2F transcription factor 1 NM_001100778 Enables protein kinase binding activity and sequence-specific DNA binding activity. Involved in cellular response to hypoxia; cellular response to nerve growth factor stimulus; and positive regulation of glial cell proliferation. Located in cytoplasm and nucleus. .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint Rom1 retinal outer segment membrane protein 1 NM_001009690 Predicted to enable protein homodimerization activity; be involved in cell adhesion; act upstream of or within detection of light stimulus involved in visual perception; protein complex oligomerization; and retina development in camera-type eye. Predicted to be located in photoreceptor outer segment membrane and an integral component of plasma membrane. Nrl neural retina leucine zipper NM_001106036 Predicted to enable DNA -binding transcription activator activity, RNA polymerase II -specific; leucine zipper domain binding activity; and promoter -specific chromatin binding activity. Predicted to be involved in regulation of transcription by RNA polymerase II and retinal rod cell development and to act upstream of or within positive regulation of gene expression and positive regulation of transcription by RNA polymerase II. Predicted to be located in cytosol and nucleoplasm and to be active in nucleus. Gngt1 G protein subunit gamma transducin 1 NM_001135777 Predicted to enable G-protein beta-subunit binding activity. Acts upstream of or within cardiac muscle cell apoptotic process and cellular response to hypoxia. Located in photoreceptor inner segment and photoreceptor outer segment. Part of heterotrimeric G-protein complex. RPE MARKERS Rpe65 retinoid isomerohydrolase RPE65 NM_053562 Enables all -trans-retinyl-ester hydrolase, 11 -cis retinol forming activity. Involved in circadian rhythm; neural retina development; and retinoid metabolic process. Located in cell body. Tjp1 tight junction protein 1 NM_001106266 Enables protein C-terminus binding activity. Involved in cellular response to glucose stimulus; negative regulation of vascular permeability; and response to lipopolysaccharide. Located in bicellular tight junction; cytoplasm; and gap junction. Best1 bestrophin 1 NM_001011940 Predicted to enable identical protein binding activity; to contribute to chloride channel activity; to be involved in transepithelial chloride transport; to act upstream of or within chloride transport; detection of light stimulus involved in visual perception; and regulation of calcium ion transport. Predicted to be located in basolateral plasma membrane. Predicted to be part of chloride channel complex. HOUSEKEEPER Hmbs hydroxymethylbilane synthase NM_013168 Part of the hydroxymethylbilane synthase superfamily in which the gene product is the third enzyme in the heme biosynthetic pathway and serves as a catalyst for the head to tail condensation of four molecules of porphobilinogen into the linear hydroxymethylbilane. .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint .CC-BY-NC 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 March 24, 2026. ; https://doi.org/10.64898/2026.03.20.713049doi: bioRxiv preprint