Abstract
Pathological retinal neovascularization is a cause of vision loss in diseases including retinopathy of prematurity (ROP), wet age-related macular degeneration (AMD), and diabetic retinopathy. The Unfolded Protein Response (UPR) is an intracellular signal transduction mechanism that is activated by ER stress and upregulates many proteins, including angiogenesis factors like VEGF and HIF-1α. This suggests that UPR genes and pathways may drive retinal angiogenesis. Here, we tested the role of the UPR regulator Activating Transcription Factor 6 (ATF6) in pathological and developmental retinal angiogenesis. We induced pathological retinal neovascularization in Atf6−/− mice using the oxygen-induced retinopathy (OIR) model and found significantly preserved visual function, accompanied by decreased retinal neovascularization, endothelial cell proliferation, and UPR transcriptional program induction. When we chemically blocked ATF6 signaling by intraocular injection of the small molecule Ceapin-A7, we also saw suppressed retinal expression of UPR genes. Additionally, in postnatal day 7 Atf6−/− mice when the retinal vasculature is developing in response to physiologic intraocular hypoxia, there was a transient but significant defect in pruning and retinal blood vessel extension. Together, our results demonstrate ATF6’s causal role in developmental and pathological retinal angiogenesis and highlight its potential as a therapeutic target to preserve vision in retinal neovascularization diseases.
Similar content being viewed by others
Introduction
The retinal vasculature begins developing at about 14–16 weeks of gestation in humans and shortly after birth in mice. Initially, the developing eye is nourished by the hyaloid vasculature, a transient network of vitreous vessels that regresses as the retinal vasculature forms. Retinal blood vessels first sprout from the optic nerve head and grow radially toward the periphery, guided by hypoxia-triggered angiogenic signals from avascular areas1,2,3,4,5,6,7. Disruptions in these signals can impair blood vessel formation, as seen in premature infants receiving high supplemental oxygen therapy, which can slow the progression of angiogenesis and is one of the factors involved in the pathophysiology of the blinding disease retinopathy of prematurity (ROP)8,9,10. In severe cases, ROP progresses to a more advanced stage, marked by abnormal blood vessel growth into the vitreous, along with greater vessel tortuosity and dilation11,12,13. This retinal neovascularization can be experimentally modeled in mice and rats due to the postnatal development of their retinal vasculature. Mice offer the opportunity to study molecular mechanisms more easily than rats. In the mouse oxygen-induced retinopathy (OIR) model, postnatal mice are exposed to hyperoxia and then returned to ambient oxygen levels to induce “relative hypoxia” and retinal neovascularization5,14.
A growing body of evidence implicates Unfolded Protein Response (UPR) signaling as a molecular mechanism driving neovascularization15,16,17,18,19. Hypoxia, a potent inducer of neovascularization, activates UPR pathways regulated by Inositol-Requiring Enzyme 1 (IRE1), Protein Kinase R (PKR)-like Endoplasmic Reticulum Kinase (PERK), and Activating Transcription Factor 6 (ATF6) through oxidative protein misfolding and subsequent ER stress20,21,22,23,24,25,26,27,28. These pathways enhance cellular survival under stress by upregulating chaperones, protein-folding enzymes, and degradation cofactors for misfolded proteins. Additionally, UPR activation induces critical angiogenesis regulators such as Vascular Endothelial Growth Factor (VEGF)29,30. ER stress exacerbates pathological ocular neovascularization in mice16,31,32. ATF6 directly modulates VEGF gene expression and stabilizes the VEGF protein23,29 and small molecule ATF6 activation promotes blood vessel development in vitro33. Inhibition of ER stress pathways, including ATF6, has been reported to reduce retinal neovascularization23,34. To expand on these findings and assess if ATF6 impacted visual outcomes, we investigated retinal vascular development under physiological hypoxia and pathological neovascularization in Atf6−/− mice.
Atf6−/− mice do not produce the Atf6 protein due to a premature stop codon in exon IV but remain viable in a standard laboratory vivarium, making them a valuable model for studying ATF6 function in vivo35,36. These mice develop sensorineural hearing loss by 2 months of age37. Atf6−/− mice have no defects in retinal vasculature, lamination, and function in studies from P15 through ~ 1 year, but afterward, they exhibit mild late-onset retinal degeneration, which becomes more severe when crossed with the P23H rhodopsin retinitis pigmentosa model36,38. Atf6−/− mouse retinal development (P0-P14) has not been previously evaluated. Beyond the retina, adult Atf6−/− mice develop liver steatosis and pancreatic beta-cell loss under high-fat diets39,40. In the heart, Atf6 deficiency leads to myocardial damage, while in the brain, it increases susceptibility to stroke and neurodegenerative diseases41,42. In some cases, Atf6 loss has beneficial effects, such as resistance to paralysis in multiple sclerosis models43. These prior studies indicate that loss of Atf6 may cause positive or negative effects in vivo, depending on cell and tissue type.
In this study, we investigate ATF6’s role specifically in retinal blood vessel development in a pathological retinal neovascularization OIR model and during normal developmental physiological intraocular hypoxia.
Results
Preserved retinal function and reduced neovascularization after OIR in Atf6 −/− mice
To investigate the role of Atf6 in pathological retinal angiogenesis, we subjected Atf6−/− mice to the oxygen-induced retinopathy (OIR) paradigm. First, both Atf6+/+ and Atf6−/− mice were placed in 75% oxygen (hyperoxia) from P7 to P12 (Fig. 1a). This hyperoxic exposure during retinal vasculature development induced vaso-obliteration indistinguishably in both Atf6+/+ and Atf6−/− retinas (Fig. 1b, left panels). When mice were returned to room air (relative hypoxia), reparative and pathological angiogenesis occurred, and neovascular growth peaked at P17 (Fig. 1b, center panels). By P20, neovascular regression was observed5,14,44 (Fig. 1b, right panels). To assess the impact of Atf6 loss on vascular growth, we quantified the percentage of avascular area at P12, P17, and P20, and neovascular area at P17 and P20. Vaso-obliteration did not significantly differ between Atf6+/+ and Atf6−/− mice at P12 (p = 0.87) or P17 (p = 0.66) (Fig. 1c). Both Atf6+/+ and Atf6−/− OIR mice showed a significant and similar reduction in vaso-obliteration between P12 and P17 (****p < 0.0001 for both Atf6+/+ and Atf6−/− retinas). P20 showed minimal avascular zones, less than 10%, consistently with previous studies (Supplemental Fig. 1)45,46. In contrast, neovascularization was significantly reduced in Atf6−/− mice compared to Atf6+/+ mice at both P17 (**p < 0.01) and P20 (**p < 0.01). Additionally, neovascularization decreased significantly from P17 to P20 in both Atf6+/+ (*p = 0.014) and Atf6−/− mice (**p < 0.01, two-way ANOVA, Fig. 1d).
Atf6 deficiency enhances visual function and reduces neovascularization in OIR. (a) Timeline of the mouse OIR model. Neonatal mice and their nursing mother were kept in room air from birth through postnatal (P) 7. In the vaso-obliteration phase, P7 mice were transferred from room air into 75% oxygen until P12, then returned to room air. During the neovascularization phase, blood vessels developed pathologic retinal neovascularization, which peaked at P17 and subsequently regressed. (b) Wholemount retinas stained for endothelial cells with IB4 (red) in Atf6+/+ and Atf6−/− OIR mice at P12 (avascular area highlighted with solid white line; vascular area highlighted with white-dotted lines) and at P17 and P20 (neovascular tufts highlighted with white-dotted lines). (c) The percentage of avascular area at P12 and P17 showed no significant difference between Atf6+/+ and Atf6−/− mice (each point represents one retina per mouse, data represents mean ± SEM, p > 0.05, two-way ANOVA, n = 3 or 4/group). (d) The percentage of neovascular area at P17 and P20 was significantly reduced in Atf6−/− mice compared to Atf6+/+ mice (each point represents one retina per mouse, data represents mean ± SEM, **p < 0.01, two-way ANOVA, n = 3/group). (e) Representative ERG waveforms generated by scotopic intensity (40 (S) cd.s/m2) for P20 Atf6+/+ (blue line) and Atf6−/− (red line) retinas (left). Scotopic ERG responses showed significantly greater a- and b-wave amplitudes in Atf6−/− mice compared to Atf6+/+ controls (data represents mean ± SEM, **p ≤ 0.01, ***p ≤ 0.001; Welch t-test, n = 7/group). (f) Representative ERG waveform generated by photopic intensity (10 (S) cd.s/m2) for P20 Atf6+/+ (blue line) and Atf6−/− (red line) retinas (left). Photopic a- and b-wave amplitudes were significantly higher in Atf6−/− mice compared to Atf6+/+ mice (data represents mean ± SEM, *p ≤ 0.05, ***p ≤ 0.001, Welch t-test, n = 6/group). Scale bar, 500 μm.
To evaluate whether the sample sizes used for each experiment were adequate to detect biologically meaningful differences, we performed post hoc power analyses using GraphPad Prism (version 10). We calculated group means, 95% confidence intervals (CIs), and effect sizes (Cohen’s d). At P12, the mean avascular area was slightly lower in Atf6+/+ retinas (39.92%; 95% CI: 29.62–50.22) than in Atf6−/− retinas (41.60%; 95% CI: 38.77–44.43), with a small effect size (Cohen’s d ≈ 0.53), indicating no biologically meaningful difference. At P17, a modest reduction in avascular area was observed in Atf6−/− retinas (17.18%; 95% CI: 9.69–24.67) compared to Atf6+/+ retinas (19.94%; 95% CI: 13.27–26.61), with a moderate effect size (Cohen’s d ≈ 0.97), indicating no statistically significant difference between groups.
In contrast, Atf6−/− retinas exhibited a significant and biologically meaningful reduction in neovascularization. At P17, the neovascular area was 9.63% (95% CI: 7.85–11.41) in Atf6−/− retinas versus 14.24% (95% CI: 6.67–21.80) in Atf6+/+ controls, with a large effect size (Cohen’s d ≈ 2.08). At P20, neovascularization was further reduced in Atf6−/− retinas (3.91%; 95% CI: 1.83–5.99) relative to Atf6+/+ controls (10.74%; 95% CI: 7.03–14.45), with a very large effect size (Cohen’s d ≈ 5.64). Post hoc power analysis indicated high statistical power (~ 99.8%), confirming that the sample size was sufficient to detect this robust biological effect. These results indicate that Atf6 deficiency is associated with a reproducible and statistically supported reduction in pathological neovascularization, supported by statistical and effect size analyses.
To assess visual function in the absence of Atf6 in OIR mouse retinas, we performed electroretinograms (ERGs) at P20. Waveforms of the scotopic and photopic ERG responses were collected from both Atf6+/+ and Atf6−/− OIR retinas. Both scotopic (Fig. 1e) and photopic (Fig. 1f) ERG responses showed significantly greater a- and b-wave amplitudes in Atf6−/− mice compared to Atf6+/+ controls (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001; Welch’s t-test). Together, these histologic and functional imaging studies reveal that Atf6−/− mice are resistant to neovascularization and retain better visual function compared to control Atf6+/+ mice after OIR, indicating that the inhibition of Atf6 function protects against OIR-induced pathological angiogenesis and subsequent vision loss.
Reduced endothelial cell proliferation in Atf6 −/− OIR model
To understand the molecular mechanisms underlying the reduction in neovascularization in Atf6−/− OIR mice, we compared bulk RNA sequencing (RNA-Seq) on retinas collected from P17 Atf6+/+ and Atf6−/− OIR mice (n = 5/group). Retinas were collected at P17, the peak of neovascular response, consistent with previous OIR studies5,14,44. Out of 15,524 total genes identified by RNA-Seq, 2,096 (identified by DESeq2 with p ≤ 0.05 and FPKM > 0.1, out of 15,524 transcripts) were significantly differentially expressed in Atf6−/− mice compared to controls by DESeq2 analysis (Supplemental Table 1). These differentially expressed genes (DEGs) were uploaded to g:Profiler (https://biit.cs.ut.ee/gprofiler) for functional enrichment analysis using the Gene Ontology (GO) Biological Processes database to identify significantly altered pathways (Supplemental Table 2). Our analysis revealed significant enrichment of pathways associated with blood vessel growth including endothelial cell (EC) proliferation, EC differentiation, regulation of blood vessel EC migration, and regulation of angiogenesis (Fig. 2a). Endothelial cell proliferation is a hallmark of proliferative retinopathy, essential for both regenerative and pathological angiogenesis47,48,49. To further examine these angiogenic processes in Atf6−/− OIR retinas, we analyzed expression of the constituent genes within the significantly-enriched angiogenesis-associated GO terms: EC proliferation (GO:0001935, green, n = 135 genes), EC differentiation (GO:0045446, blue, n = 90 genes), regulation of blood vessel EC migration (GO:0043535, yellow, n = 80 genes), and regulation of angiogenesis (GO:0045765, red, n = 210 genes) (Fig. 2b, Supplemental Table 3). Compared to control retinas, Atf6−/− retinas exhibited a significant reduction in the mean expression of all four gene sets: **p < 0.01 for EC proliferation and ****p < 0.0001 for the other three categories, based on a Two-Tailed Wilcoxon Signed-Rank Test (n = 5/group, Fig. 2b). These findings are illustrated in the violin plots of Fig. 2b, where green indicates **p ≤ 0.01 and blue, yellow, and red indicate ****p ≤ 0.0001. These findings support that genes associated with endothelial cell function including proliferation, differentiation, migration, and angiogenesis are significantly downregulated in Atf6−/− OIR retinas. Furthermore, these transcriptomic findings align with our histologic findings that retinal neovascularization was reduced in Atf6−/− mice.
Endothelial cell proliferation is reduced in Atf6−/− OIR mice during retinal neovascularization. (a) RNA sequencing was performed on P17 (the peak of retinal neovascularization) Atf6+/+ and Atf6−/− OIR mouse retinas. Functional enrichment analysis using the Gene Ontology (GO) Biological Processes database via g:Profiler on all significantly differentially expressed genes (2,096 genes) revealed significant changes in terms for endothelial cell proliferation, endothelial cell differentiation, regulation of blood vessel endothelial cell migration, and regulation of angiogenesis (n = 5/group). The x-axis shows the -log10(p-value) for each term. (b) Violin plots show gene set expression levels for the terms identified in functional enrichment analysis (GO:0001935 endothelial cell proliferation, green; GO:0045446 endothelial cell differentiation, blue; GO:0043535 regulation of blood vessel (BV) endothelial cell (EC) migration, yellow; GO:0045765 regulation of angiogenesis, red) were all significantly downregulated in P17 Atf6−/− retina transcriptomes, expressed as log2(fold change) compared to P17 Atf6+/+ retinas. The thick dashed line indicates the median log2(fold change), the thin dashed lines demarcate the top and bottom quartiles, and the blue line indicates 1x fold change (**p ≤ 0.01, ****p ≤ 0.0001, Two-Tailed Wilcoxon Signed Rank Test, n = 5/group). The complete gene sets are shown in Supplemental Table 3. (c) Representative images of vertical retinal sections labeled with IB4 (blue), ERG (red), and Ki-67 (green) from P17 Atf6+/+ and Atf6−/− retinas (n = 3/group). In the merged image, colocalization of ERG and Ki-67 inside of IB4 areas (arrows) indicate proliferating inner retina blood vessel endothelial cells. (d) Percentage of Ki-67-immunoreative blood vessel endothelial cells was significantly lower in Atf6−/− retinas compared to Atf6+/+ retinas (data represent mean ± SEM, *p ≤ 0.05, Student’s t-test, n = 3/group). ONL outer nuclear layer; OPL outer plexiform layer; INL inner nuclear layer; IPL inner plexiform layer; GCL ganglion cell layer. Scale bar = 50 μm.
Previous studies have shown that OIR mouse retinas typically exhibit an increase in endothelial cells, heightened cell proliferation, and enhanced vaso-proliferation, particularly during the neovascular phase47,50. To further investigate whether the reduction of neovascularization in Atf6−/− OIR retinas was due to a lower number of proliferating endothelial cells in neovascular tufts, we performed triple labeling using the endothelial marker IB4, the endothelial cell nuclei marker ERG, and the proliferation marker Ki-67 (Fig. 2c). In Atf6+/+ OIR retinas, numerous Ki-67-positive endothelial cells were observed, co-localized with ERG in IB4-positive blood vessels (Fig. 2c, top panels). In contrast, Atf6−/− OIR retinas exhibited a marked reduction in the number of Ki-67-positive endothelial cells, indicating decreased endothelial cell proliferation (Fig. 2c, bottom panels). The merged images confirm a significant reduction in proliferating endothelial cells (IB4+, ERG+, and Ki-67+) in Atf6−/− retinas compared to Atf6+/+ retinas (*p ≤ 0.05, Student’s t-test; Fig. 2d). These RNA-Seq and histologic data demonstrate that inhibition of Atf6 reduces endothelial cell proliferation in the OIR mouse retina.
Phototransduction genes upregulated in Atf6 −/− OIR mice with unchanged ONL thickness
We found that Atf6 deficiency improved visual function in P20 OIR mice, as evidenced by enhanced ERG responses (Fig. 1e, f). GO comparison of the retinal transcriptome between Atf6−/− and control mice revealed significant enrichment of terms related to visual function (Supplemental Table 2), including sensory perception of light stimulus (GO:0050953, yellow, n = 144), response to light stimulus (GO:0009416, green, n = 307), and detection of light stimulus (GO:0009583, blue, n = 49; Fig. 3a). To investigate the molecular basis of this improvement, we analyzed the expression levels of the gene sets corresponding to these GO terms (Supplemental Table 4). Compared to Atf6+/+ retinas, Atf6−/− retinas exhibited significantly higher mean expression across all three gene sets: *p ≤ 0.05 for sensory perception of light stimulus and response to light stimulus, and **p ≤ 0.01 for detection of light stimulus, based on a Two-Tailed Wilcoxon Signed-Rank Test (n = 5/group). These differences are illustrated in the violin plots in Fig. 3b, where yellow and green indicate *p ≤ 0.05 and blue indicates **p ≤ 0.01. These findings support that genes associated with visual function including sensory perception of light stimulus, response to light stimulus, and detection of light stimulus are significantly upregulated in Atf6−/− OIR retinas. Furthermore, Gene Set Enrichment Analysis (GSEA) confirmed marked enrichment of the sensory perception of light stimulus gene set (FDR = 0.01) in the Atf6−/− OIR retina transcriptome (Supplemental Fig. 2). Together, these transcriptomic results suggest that the protective effect of Atf6 deficiency against neovascularization is accompanied by preservation of visual response pathways, consistent with the improved ERG responses observed in Atf6−/− retinas (Fig. 1).
Phototransduction genes upregulated in Atf6−/− retinas with no change in retinal ONL in OIR. (a) RNA sequencing was performed on P17 (the peak of retinal neovascularization) Atf6+/+ and Atf6−/− OIR mouse retinas. Functional enrichment analysis using the Gene Ontology (GO) Biological Processes database via g:Profiler on all significantly differentially expressed genes (2,096 genes) revealed significant changes in terms for sensory perception of light stimulus, response to light stimulus, and detection of light stimulus (n = 5/group). The x-axis shows the -log10(p-value) for each term. (b) Violin plots show gene set expression levels for the terms identified in functional enrichment analysis (GO:0050953 sensory perception of light stimulus, yellow; GO:0009416 response to light stimulus, green; and GO:0009583 detection of light stimulus, blue) in P17 Atf6−/− OIR retina RNA-Seq transcriptomes, expressed as log2(fold change) compared to P17 Atf6+/+ OIR retinas (*p ≤ 0.05, **p ≤ 0.01, Two-Tailed Wilcoxon Signed Rank Test, n = 5/group). The complete gene sets are shown in Supplemental Table 4. (c) Representative images of the retinal outer nuclear layer (ONL) approximately 500 μm away from the optic disc in P17 Atf6+/+ and Atf6−/− retinas, labeled with DAPI (n = 3/group). (d) Thickness of ONL was not significantly different between Atf6+/+ and Atf6−/− retinas (data represent mean ± SEM, p > 0.05, Student’s t-test, n = 3/group). ONL outer nuclear layer; OPL outer plexiform layer. Scale bar = 50 μm.
Lastly, to determine whether the thickness of the retinal outer nuclear layer (ONL) differed between Atf6+/+ and Atf6−/− OIR retinas, we measured the ONL thickness in vertical cross-sections (Fig. 3c). No significant difference in ONL thickness was observed between Atf6−/− (data represent mean ± SEM, 75.2 ± 1.9 μm) and Atf6+/+ retinas (74.2 ± 2.8 μm, p = 0.6, Student’s t-test, n = 3; Fig. 3d). This indicates that the enhanced visual response and increased gene expression in Atf6−/− mice are not due to differences in thickness of ONL.
Downregulation of UPR genes after genetic or chemical ATF6 inhibition
ATF6 is a key component of the UPR pathway, which is activated in response to ER stress. Previous studies have shown that ER stress is significantly elevated in OIR retinas, with activation of UPR signaling molecules such as GRP78, PERK, and eIF2A, some of which are also linked to pathological neovascularization32,34,51. Since ATF6 and other UPR pathways upregulate many genes aiding ER stress adaptation16,23,33 we investigated how UPR-regulated gene expression was altered in the transcriptomic data collected from our Atf6−/− OIR retinas. Using RNA-Seq data, we examined the expression of a previously published panel of 118 genes transcriptionally regulated by the IRE1/XBP1, ATF6, and PERK/ATF4 UPR pathways35,52,53,54,55,56,57,58,59,60 in P17 Atf6+/+ and Atf6−/− OIR mouse retinas (Supplemental Table 5). The 118-gene UPR gene panel was significantly downregulated in Atf6−/− versus Atf6+/+ retinas, as shown in the violin plot (red violin plot, ***p ≤ 0.001, Two-Tailed Wilcoxon Signed Rank Test, n = 5/group; Fig. 4a). These findings reveal that the UPR transcriptional program is downregulated in Atf6−/− mice compared to Atf6+/+ after OIR.
Genetic or chemical Atf6 inactivation downregulated UPR-regulated gene set. (a) Violin plot shows expression levels of 118 UPR-regulated genes in P17 Atf6−/− OIR retina RNA-Seq transcriptomes. Expression levels are shown as log2fold change compared to Atf6+/+ OIR retinas (red violin plot, ***p ≤ 0.001, Two-Tailed Wilcoxon Signed Rank Test, n = 5/group). The complete UPR gene set is shown in Supplemental Table 5. (b) Violin plot shows expression levels of 118 UPR-regulated genes in wild-type P30 retinas treated with Ceapin-A7, 24 h post-injection. The thick dashed line indicates the median log2(fold change), the thin dotted lines indicate the top and bottom quartiles, and the blue line indicates 1x fold change (purple violin plot, ****p ≤ 0.0001, Two-Tailed Wilcoxon Signed Rank Test, n = 5/group; Supplemental Table 7).
To further test if UPR gene expression was disrupted by loss of Atf6, we chemically inhibited Atf6 using the small molecule Ceapin-A761. Intravitreal injections of Ceapin-A7 were performed on P30 Atf6+/+ mice, and bulk RNA-Seq transcriptomes were analyzed from saline-treated and Ceapin-A7-treated retinas 24 h post-injection. This design allowed us to isolate direct molecular effects of Atf6 inhibition without the confounding variables present in the OIR model, such as fluctuating oxygen levels and active neovascular remodeling. We identified 15,417 (expression > 0.1 FPKM) transcripts expressed in Ceapin-A7-treated retinas (Supplemental Table 6). Similarly to our findings in the Atf6−/− mice retinas collected after OIR, the expression levels of the UPR genes (n = 118, Supplemental Table 7) were significantly downregulated in Ceapin-A7-treated versus saline-treated retinas (purple violin plot, ****p ≤ 0.0001, Two-Tailed Wilcoxon Signed Rank Test, n = 5/group; Fig. 4b). Our results indicate that either genetic or chemical inhibition of Atf6 led to a downregulation of UPR genes relative to controls, in OIR and in normal conditions. These findings support that the reduced retinal neovascularization observed in Atf6−/− mice under OIR conditions is caused by a diminished UPR transcriptional response.
Early retinal angiogenesis is delayed in Atf6 −/− mice
During early retinal development in mice (P0-P7), physiologic hypoxia in the eye triggers the formation of new retinal blood vessels, emerging from the optic nerve head and extending from the central to the peripheral retina4,6. As these new vessels form and alleviate hypoxia, vascular growth continues radially toward the retinal periphery7. Since Atf6 deficiency suppressed neovascularization in the pathologic OIR model, we investigated if Atf6 loss also affects physiologic hypoxia-driven retinal vascular development during this early postnatal period. To address this, we analyzed retinal vasculature in Atf6+/+ and Atf6−/− mice at P3, P7, P10, and P14. These time points were selected to capture key phases of retinal angiogenesis62. At P3, the superficial vascular network begins to form as a dense, immature plexus with active sprouting at the leading edge. By P7, the superficial network is expanding toward the peripheral retina and vertical sprouting from superficial veins into the deep retinal layers begins to initiate, marking the onset of vascular stratification. P10 represents a phase of active vertical sprouting and further elaboration of the intermediate and deep vascular layers, while capillary sprouting and remodeling continue within the superficial plexus. By P14, the vascular network reaches near-maximum complexity, with established layers and ongoing remodeling. These four time points represent a continuum from initial plexus formation (P3) through onset and regulation of stratification (P7-P10) to late-stage vascular maturation (P14)62. This developmental timeline provides a framework to assess how Atf6 influences each stage of retinal angiogenesis.
Retinal wholemounts at these representative timepoints were stained with IB4 to visualize vascular growth (Fig. 5a). At P3, blood vessels were visible around the optic nerve head. By P7, these vessels had expanded into a visible network and reached the retinal periphery by P10 in Atf6+/+ retinas. In contrast, Atf6−/− retinas showed reduced vascular coverage at P7, but this difference resolved by P10, with no significant difference in vascular fraction from controls at P10 and P14. This is consistent with our prior findings that showed no vascular differences in Atf6−/− mice at older ages38. While the total retinal area was similar between Atf6+/+ and Atf6−/− retinas at all time points (Fig. 5b), the vascular fraction was significantly reduced in Atf6−/− retinas at P7 (*p < 0.05, two-way ANOVA; Fig. 5c), indicating a transient delay in the radial development of superficial vasculature during this period of developmental hypoxia-driven retinal angiogenesis.
Early retinal angiogenesis is delayed in Atf6−/− mice. (a) Wholemounts of Atf6+/+ and Atf6−/− mice at P3, P7, P10, and P14 labeled with IB4 to visualize vasculature. Retina area is outlined. (b) Retina area did not differ significantly between Atf6+/+ and Atf6−/− retinas at P3, P7, P10, and P14 (data represent mean ± SEM, p > 0.05, two-way ANOVA, n = 3/group). (c) Vascular fraction was significantly reduced in Atf6−/− retinas compared to Atf6+/+ retinas at P7 (n = 6/group) but showed no significant difference at P3 (n = 6/group) or P10 and P14 (n = 3/group) (data represent mean ± SEM, *p < 0.05, two-way ANOVA). (d) Wholemounts were analyzed for branch number in fields from central (250 μm away from the optic disc), mid-peripheral (750 μm away from the optic disc), and peripheral (1.25 mm away from the optic disc) regions (black boxes, 250 μm x 250 μm) in the superior region (left). Schematic of vascular plexus development (center). ImageJ (FIJI) quantification workflow (right). (e) Confocal microscopy images of IB4-stained wholemounts at P7 (superficial layer), P10 (superficial and deep layers), and P14 (superficial, deep, and intermediate layer) in Atf6+/+ and Atf6−/− retinas. (f) Branch counts in the superficial layer at P7, P10, and P14. At P7, branch number is significantly greater in the central retina of Atf6−/− mice compared to Atf6+/+ mice (data represent mean ± SEM, *p < 0.01, two-way ANOVA, n = 7/group). At P10 and P14, no differences in the central retina (data represent mean ± SEM, p > 0.05, two-way ANOVA, n = 3–4/group). Mid-peripheral and peripheral regions showed no differences at any time points (P7, n = 7/group; P10 and P14, n = 3–4/group). Branch numbers in the deep layer at P10 and P14 and the intermediate layer at P14 showed no differences between groups. (g) Representative images of endothelial tip cells at P7 in Atf6+/+ and Atf6−/− retinas labeled with IB4 (left). Tip cell number is significantly greater in Atf6−/− retinas (right, data represent mean ± SEM, *p < 0.05, Student’s t-test, n = 3–4/group). Sup. superior; temp. temporal; inf. Inferior. Scale bar = 1 mm (A); 1 mm (D); 50 μm (E); 50 μm (G).
Next, we examined the morphology and architecture of the developing blood vessels in more detail at P7 when we found reduced angiogenesis. We quantified vascular branch density in both the superficial and deep retinal layers. Branch density was measured in retinal areas of 250 μm x 250 μm located 250 μm (central), 750 μm (mid-peripheral), and 1250 μm (peripheral) from the optic disc (Fig. 5d). Superficial vasculature was imaged at P7, P10, and P14; deep layer vasculature at P10 and P14; and the intermediate layer at P14, when the plexus becomes more clearly visualized and distinguishable (Fig. 5e). At P7, branch density in the central retina was significantly higher in Atf6−/− retinas compared to Atf6+/+ retinas, suggesting reduced branch pruning in the absence of Atf6 (***p < 0.001, two-way ANOVA; Fig. 5f). In contrast, no significant difference in branch density was observed in the mid-peripheral and peripheral superficial retina between genotypes at P7. By P10, branch pruning had normalized between the two genotypes. In the deep retinal layer, branch density did not differ significantly between genotypes at either P10 or P14 in central, mid-peripheral, or peripheral retina. Similarly, in the intermediate layer, imaged at P14 when it becomes morphologically distinct, branch density remained comparable across all regions between Atf6+/+ vs. Atf6−/− retinas (Fig. 5f). Finally, we quantified tip cells, which are essential for angiogenesis due to their role in VEGF signaling63,64,65,66. In the Atf6+/+ retinas at P7, the vascular front at the periphery appeared smoother with fewer long filopodia observed along the leading edge. In contrast, Atf6−/− retinas at P7 exhibited many long filopodia at the leading edge, indicating active sprouting and tip cell formation (Fig. 5g, left). Moreover, the number of tip cells was significantly higher in the Atf6−/− retinas compared to the Atf6+/+ retinas (**p < 0.01, Student’s t-test; Fig. 5g, right). These results identify defects in retinal blood vessel pruning, branching, morphology, and tip cell numbers that underlie a transient delay in retinal angiogenesis seen in Atf6−/− retinas during early development.
Discussion
In this study, we explored the role of ATF6 in retinal neovascularization using the OIR model, hypothesizing that ATF6 drives pathological angiogenesis. Hypoxia, a known trigger for neovascularization, activates UPR pathways like IRE1, PERK, and ATF6, which in turn elevate factors like VEGF to promote angiogenesis. Here, we demonstrate that Atf6-deficient mice exhibit reduced blood vessel formation and significantly improved visual function post-hypoxia. Our RNA-Seq analysis showed significant downregulation of angiogenesis-related gene sets in Atf6−/− retinas, along with fewer proliferating endothelial cells, suggesting that ATF6 selectively mediates neovascularization.
UPR signaling, such as through ATF6, induces angiogenesis regulators like VEGF, linking ER stress to neovascularization18,29,30. Interestingly, VEGF can also increase unfolded proteins in the ER, further activating the UPR67. In our study, Atf6−/− mice showed resistance to neovascularization post-hypoxia, likely due to decreased endothelial cell proliferation. This reduction was accompanied by downregulation of angiogenesis-related gene sets, including endothelial cell proliferation, endothelial cell differentiation, regulation of blood vessel endothelial cell migration, and regulation of angiogenesis, in the hypoxic retinas of Atf6−/− mice. These findings align with prior reports that ATF6 directly regulates VEGF transcription by binding to its promoter and stabilizes VEGF protein to prevent its degradation23,29. Additionally, loss of ATF6 may alter oxidative stress responses, potentially affecting ROS-mediated signaling pathways such as HIF-1α stabilization, which is critical for VEGF induction under hypoxia41. These findings suggest that ATF6 deficiency not only suppresses VEGF but also disrupts the oxidative stress-mediated angiogenic response in the OIR model. This highlights ATF6’s multifaceted role in the retinal response to oxidative stress and neovascularization, making it an attractive therapeutic target for conditions involving abnormal blood vessel growth.
Our findings identify a role for ATF6 in early retinal vascular development, as evidenced by delayed angiogenesis and reduced pruning in Atf6−/− mice at P7. Specifically, we observed delayed outgrowth of the radial superficial vascular plexus and increased branch density in the central retina, suggesting a shift in vessel remodeling dynamics. Pruning in hyperoxic regions is essential for transforming the primitive vascular network into a mature, organized structure6. In Atf6+/+ retinas, the vascular front at P7 appeared smooth, with fewer filopodia, reflecting advanced maturation. In contrast, Atf6−/− retinas exhibited increased filopodia, indicative of active sprouting and tip cell formation. Tip cells, driven by VEGF-mediated astrocyte signaling in hypoxic regions, facilitate the radial extension of vasculature to the periphery63,64,65,66. These observations suggest that ATF6 is important for balancing vascular sprouting and pruning, possibly through modulation of VEGF signaling or other angiogenic pathways. Notably, the developmental delay in angiogenesis observed in Atf6−/− mice was transient, with normal vascular formation restored by P10. This suggests the presence of compensatory mechanisms that mitigate the absence of ATF6 during retinal vascular development. One explanation for the initial delay lies in ATF6’s established role in promoting mesoderm formation and differentiation into vascular tissues. Studies have shown that loss-of-function mutations in ATF6 reduce mesodermal marker expression in patient-derived stem cells, indicating a direct influence of ATF6 on early vascular development33. However, the restoration of vascularization at later stages is likely due to compensatory mechanisms involving other branches of the UPR. For example, the IRE1 pathway activates stress-response genes that support cell survival and adaptation, and its functional overlap with ATF6 may help offset the loss of ATF636,68,69. IRE1 has also been shown to regulate angiogenic factors, such as VEGF and HIF-1α, by splicing XBP1 mRNA, promoting endothelial cell proliferation and survival under hypoxic conditions15,23,70,71. The essential role of IRE1/XBP1 in development is evident in Xbp-1s−/− and Ire1α−/− mice, which exhibit severe defects and die early in embryogenesis72,73. Similarly, Perk−/− mice develop skeletal dysplasia at birth, experience postnatal growth retardation, and typically die within four weeks74,75. In contrast, adult Atf6−/− mice maintain normal retinal vasculature and visual function under standard conditions until late in life38. These findings of the various UPR gene knockout mice show that ATF6 ablation causes the least systemic toxicity among the three UPR branches, supporting ATF6 inhibition as a promising, low-toxicity strategy for suppressing pathological neovascularization.
To assess changes in photoreceptor function in Atf6−/− mice with OIR, we utilized ERG. Previous studies in rat OIR models76,77,78 and patients with retinopathy of prematurity (ROP)79,80 have reported retinal dysfunction, often implicating impaired photoreceptor function due to pathological neovascularization. In our study, Atf6−/− OIR mice exhibited increased amplitudes in both scotopic and photopic a- and b-wave ERG responses, suggesting enhanced photoreceptor function compared to Atf6+/+ OIR mice. The typical reduction in visual response amplitude in OIR is often attributed to dysregulation in photoreceptor outer segment membrane channels, which has been linked to reduced outer segment length81. Interestingly, despite the enhanced ERG amplitudes in Atf6−/− mice, the outer nuclear layer (ONL) thickness remained comparable between Atf6+/+ and Atf6−/− OIR retinas. This finding aligns with preserved ONL thickness observed in some previous OIR mouse models81,82 but retinal thinning has also been observed in some OIR mouse and rat models80,83,84,85. These results suggest that the increase in scotopic and photopic b-wave amplitudes in Atf6−/− mice is not due to differences in thickness in ONL but rather reflects functional preservation of the photoreceptors. Gene expression analysis further revealed upregulation of the sensory perception of light stimulus, response to light stimulus, and detection of light stimulus gene sets in Atf6−/− OIR retinas. While we did not directly measure photoreceptor outer segment length, the functional data suggest that photoreceptor channels in the outer segment membrane are better preserved in Atf6−/− OIR retinas. This preservation may be linked to reduced neovascularization, which alleviates the detrimental effects of pathological angiogenesis on photoreceptor function. These findings highlight a potential protective role of ATF6 inhibition in mitigating angiogenesis-associated photoreceptor dysfunction.
Our findings support that UPR inhibition, demonstrated with Atf6−/− mice and the ATF6-specific inhibitor Ceapin-A7, holds promise as a therapeutic strategy for diseases linked to neovascularization. In Atf6−/− mice, pathological neovascularization was significantly reduced, accompanied by downregulation of UPR-related genes. Similarly, intravitreal injections of Ceapin-A7 in P30 Atf6+/+ mice led to reduced expression of UPR-related genes. These findings suggest that ATF6 inhibition directly suppresses ER stress pathways that support pathological angiogenesis, particularly under hypoxic conditions. This mechanism aligns with evidence from oncology, where preclinical studies have demonstrated the critical roles of PERK and IRE1 in promoting tumor angiogenesis16. These insights have driven early-phase clinical trials (https://clinicaltrials.gov) exploring anti-IRE1 (NCT03950570) and anti-PERK (NCT04834778) compounds as potential cancer therapies86,87. The observed reduction in UPR-related gene expression may reflect a disruption of pro-angiogenic signaling and a dampened ER stress response in disease conditions, contributing to the observed reduction in neovascularization and preservation of retinal function.
Our study shows that genetic ATF6 inhibition preserves visual function and reduces neovascularization, but further research is needed to evaluate its therapeutic potential. Key steps include evaluating the ocular safety and potency of compounds that inhibit ATF6 or UPR and optimizing their delivery to the retina. If ATF6-targeting approaches prove effective in preserving vision and reducing neovascularization, they could lead to novel treatments for blinding retinal diseases.
Materials and methods
Animals
Atf6-knockout (Atf6−/−) mice were originally generated by the Kaufman laboratory35 and provided directly by Dr. Randal J. Kaufman. We have maintained and bred this line in our laboratory and used it in multiple previously published studies36,37,38. Wildtype (Atf6+/+) and Atf6−/− mice on a pure C57BL/6J (B6J) background were used for the experiments. Both male and female animals were studied at postnatal (P) days 3, 7, 10, 12, 14, 17, 20, and 30. All animals were maintained in a specific pathogen-free (SPF) facility with unrestricted access to food and water and maintained under a 12-hour light/dark cycle. Mice were obtained from at least three different litters derived from separate breeding pairs to minimize litter-specific effects. Within each genotype, pups were randomly selected and assigned to experimental or control groups to ensure balanced representation across litters. Inclusion criteria consisted of genotyped Atf6+/+ and Atf6−/− mice of both sexes on a C57BL/6J background with no baseline ocular abnormalities. Mice exhibiting developmental defects or signs of systemic illness were excluded prior to experimental allocation. Investigators responsible for outcome assessments and data analysis were blinded to both genotype and treatment group to reduce potential bias. Mouse husbandry and experimental procedures were approved by and conducted strictly according to the relevant guidelines and regulations of the Institutional Animal Care and Use Committee at Stanford University, Veterans Affairs Palo Alto Health Care System (Palo Alto, CA), and EyeCRO (Oklahoma City, OK) in compliance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research and the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines.
Oxygen-induced retinopathy
The oxygen-induced retinopathy (OIR) protocol was induced in Atf6+/+ and Atf6−/− mice as previously described14,44. Briefly, at P7, mouse pups of both sexes and their nursing mothers were placed in an environment with 75% oxygen (hyperoxia) for five consecutive days. They were then returned to room air at P12. Retinas were collected at P12 (end of hyperoxia), P17 (peak of neovascularization), and P20 (regression phase of neovascularization)5,14,44.
Administration of Ceapin-A7 (Atf6 inhibitor)
Intravitreal injections in mice were conducted under anesthesia induced by intraperitoneal injection of ketamine (80–90 mg/kg; KETASET, Fort Dodge, IA, USA) and xylazine (5–10 mg/kg; X-Ject SA, Butler, Dublin, OH, USA). Ceapin-A7 was obtained from Sigma-Aldrich (St. Louis, MO, United States) and prepared at a concentration of 10 µM dissolved in sterile-filtered saline containing 0.01% dimethyl sulfoxide (DMSO). A volume of 1 µL of Ceapin-A7 was administered via intravitreal injection using a Hamilton syringe with a 33-gauge needle (Sigma-Aldrich Corp.). For the Ceapin-A7-treated group, both eyes of each animal were intravitreally injected with Ceapin-A7 at P30. In the control group, both eyes of each animal were intravitreally injected with 1 µL of saline containing 0.05% DMSO at P30. After injection, veterinary ophthalmic antibacterial ointment was administered to prevent infection and corneal drying. Retinas were collected 24 h post-injection for RNA-Seq analysis.
Retina tissue preparation
Mice were anesthetized via intraperitoneal injection of ketamine (80–90 mg/kg; KETASET, Fort Dodge, IA, USA) and xylazine (5–10 mg/kg; X-Ject SA, Butler, Dublin, OH, USA). Retinal tissue was collected from the enucleated eyes at postnatal days P3, P7, P10, P12, P14, P17, P20, and P30. Following enucleation, the animals were euthanized by administering an overdose of Euthasol (200 mg/kg, Virbac Corporation, Fort Worth, TX, USA), with cervical dislocation performed as a secondary method. The cornea and lens were removed, and eyecups were fixed in 4% paraformaldehyde prepared in 0.1 M phosphate buffer (PB) for 60 min at 4°C. Post-fixation, the eyecups were transferred to 30% sucrose solution overnight at 4°C. Subsequently, the eyecups were embedded in Optimal Cutting Temperature (OCT) medium (Tissue-Tek, Elkhart, IN, USA), frozen using liquid nitrogen, and sectioned along the vertical meridian using a Leica cryostat (Leica Biosystems Inc, Buffalo Grove, IL, USA) to a thickness of 15 μm. The 15-µm-thick retinal sections were collected and mounted on Superfrost Plus slides (Fisher Scientific). For wholemount retinal preparation, retinas were carefully dissected from fixed eyecups. The ciliary body was retained during dissection to ensure the entire peripheral retina was preserved for complete area analysis. Retinas were then flattened for staining.
Immunohistochemistry
For immunohistochemistry, 15-µm-thick cryostat sections were incubated in normal donkey serum (NDS) (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature. Following this, the sections were incubated overnight at 4°C with primary antibodies: Alexa Fluor™ 568 conjugate Isolectin GS-IB4 (Thermo Fisher Scientific, Waltham, MA, #121412, dilution 1:1000), Alexa Fluor™ 594 conjugate Isolectin GS-IB4 (Thermo Fisher Scientific, Waltham, MA, #121413, dilution 1:1000), rabbit polyclonal antibody against ERG (Abcam, Cambridge, UK, ab92513, dilution 1:200), and Alexa Fluor™ 488 conjugate Ki-67 (Thermo Fisher Scientific, Waltham, MA, #53-5698-82, dilution 1:1000). After primary antibody incubation, sections were either washed with 0.1 M phosphate buffer (PB) (for those with conjugated antibodies) or incubated for 2 h at room temperature with Cy3-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch Laboratories, dilution 1:500). For triple-label studies, sections were incubated overnight with a mixture of primary antibodies followed by a 2-hour incubation with the secondary antibody. The sections were then washed for 30 min with 0.1 M PB, cover-slipped with VECTASHIELD HardSet™ Antifade Mounting Medium with DAPI (Vector Laboratories, Newark, CA, #H-1000-10), and analyzed using a Nikon A1R confocal microscope (Nikon, NY, USA). Immunofluorescence images were processed using Nikon NIS Elements and ImageJ (National Institutes of Health, imagej.net). The brightness and contrast of the images were adjusted using Adobe Photoshop 7.0 (Adobe Systems, Inc., San Jose, CA, USA). Wholemount immunostaining was conducted in the same manner except that the incubation times with the primary antibodies were extended to 2 nights.
RNA sequencing
RNA sequencing was carried out as described in previous studies55,88. A sample size of five whole retinas each from P17 Atf6+/+ OIR, P17 Atf6−/− OIR, P30 wildtype saline-treated, and P30 wildtype Ceapin-A7-treated mice were collected, and RNA was extracted according to the manufacturer’s instructions. BGI’s Eukaryotic Strand-specific Transcriptome Resequencing service (http://biosys.bgi.com) performed the RNA sequencing, utilizing the DNBSEQ stranded mRNA library to produce paired-end 100 bp reads, achieving 30 million reads per sample. The sequencing reads were aligned to the Mus_musculus_GCF_000001635.27_GRCm39 genome using HISAT2 alignment software (v2.0.4). Gene expression levels and normalized reads in FPKM were quantified using RSEM software (v1.2.18)89. Differential expression analysis between Atf6+/+ and Atf6−/− OIR retinas, as well as between saline-treated and Ceapin-A7-treated retinas, was conducted using the DESeq2 package (v1.4.5)90.
Functional enrichment analysis
Functional enrichment analysis was performed using g:Profiler (https://biit.cs.ut.ee/gprofiler/), focusing on Gene Ontology (GO) terms in the Biological Process category. Input genes were selected from RNA-Seq data based on significant differential expression between Atf6+/+ and Atf6−/− OIR retinas (log2fold change, p ≤ 0.05, expression > 0.1 FPKM), resulting in 2,096 differentially expressed genes (DEGs) from an initial pool of 15,524. As an alternative approach, Gene Set Enrichment Analysis (GSEA) software (https://www.broadinstitute.org/gsea/) was also used. Pre-ranked gene lists, ordered by expression relative to wild-type controls, were analyzed using the GO reference database with weighted scoring. GSEA enrichment plots were generated to visualize pathway-level changes.
Electroretinogram
Full-field electroretinogram (ERG) analyses were conducted using an Espion system (Diagnosys, Lowell, MA, USA). Following a minimum of 12 h of dark adaptation, the animals were anesthetized with an intraperitoneal injection of ketamine (80–90 mg/kg; KETASET, Fort Dodge, IA, USA) and xylazine (5–10 mg/kg; X-Ject SA, Butler, Dublin, OH, USA). Under dim red light, Proparacaine HCL Ophthalmic Solution (0.5%; Bausch and Lomb, Bridgewater, NJ, USA) was applied as a local anesthetic to the pupils. Subsequently, Tropicamide Ophthalmic (1%; Bausch and Lomb, Bridgewater, NJ, USA) and Phenylephrine HCL Ophthalmic Solution (10%; AKORN, Lake Forest, IL, USA) were administered as a cycloplegic agent and a dilator, respectively. For scotopic response assessment, the dark-adapted, dilated eyes were exposed to a stimulus intensity of 40 (S) cd.s/m². To evaluate the photopic response, the animals were light-adapted for 7 min before a strobe flash with an intensity of 10 (S) cd.s/m² was presented to the dilated eyes. Fifteen repeated flashes were averaged to obtain the final waveform, with the photopic b-wave amplitude measured from the a-wave trough to the b-wave peak.
Quantification of retinal vasculature in OIR
Vaso-obliteration was assessed by outlining the central avascular area and total retinal area using the Polygonal Lasso tool in Photoshop (Adobe Systems, San Jose, CA) and expressed as the central avascular area in relation to the total retinal area (% avascular area) at P12, P17, and P20. For neovascularization, individual neovascular tufts and clusters were outlined with a white dotted line using the same software at P17 and P20, and the total neovascular area was expressed in relation to the total retinal area (% neovascular area). Wholemount images were imported into NIH’s ImageJ software (National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html) for quantification and retinal area measurement5,14.
To quantify proliferating endothelial cells in the Atf6+/+ and Atf6−/− OIR mice, retinas were examined to compare Ki-67-immunoreactive cells (a marker for proliferating cells)91 in ERG-positive cells (an endothelial cell marker)92,93 within IB4-immunoreactive blood vessels. Two regions, each located 500 μm away from the optic disc, were selected from each retina for measurement. At these locations, serial optical sections were imaged by confocal microscopy. Every Ki-67-immunoreactive cell within the ERG-immunoreactive region of the blood vessels was counted by assessing its colocalization with IB4 immunoreactivity.
Quantification of retinal vasculature in early development
The retinal areas of wholemount retinas from Atf6+/+ and Atf6−/− mice were measured at P3, P7, P10, and P14 using ImageJ software (National Institutes of Health, Bethesda, MD). To evaluate the vascular fraction, the vascular area and the total retinal area were outlined using the Polygonal Lasso tool in Photoshop (Adobe Systems, San Jose, CA), and the vascular area was expressed as a ratio to the total retinal area at P3, P7, P10, and P14.
High-resolution images of retinal wholemounts stained with IB4 were used to analyze the number of vascular branches. Three fields of 175 μm × 175 μm were sampled from the central (250 μm away from the optic disc), mid-peripheral (750 μm away from the optic disc), and peripheral (1.25 mm away from the optic disc) areas in the superior region of wholemount retinas. Using ImageJ, images of these areas were thresholded to create a binarized image in which the vascular area is positive and the background is negative. Binarized images were then processed into skeletons using the Skeletonize plugin, and branch numbers were quantified using the AnalyzeSkeleton plugin94. Endothelial tip cells were quantified along the leading edge of the developing vascular plexus using ImageJ as described in a previous study63.
Measurement of outer nuclear layer (ONL) thickness
The thickness of the ONL was measured at 100 μm intervals, beginning 500 μm away from the optic disc. For each retinal section, five measurements of ONL thickness were taken, spaced approximately 100 μm apart, and averaged. Layer thickness measurements were collected from three retinas each from Atf6+/+ and Atf6−/− mice.
Statistical analysis
All data are expressed as mean ± standard error of the mean (SEM). A p-value of less than 0.05 was considered significant and annotated as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Student’s t-test or Welch’s t-test was used to compare two group means, depending on variance equality, which was assessed by an F-test. Two-way ANOVA followed by Tukey’s multiple comparisons test was used to evaluate the effects of genotype (Atf6+/+ vs. Atf6−/−) and developmental timepoint (e.g., P12 vs. P17 or P17 vs. P20). Violin plots were generated using the log2(fold change) data from the RNA-Seq differential expression analysis. Differences in the expression of gene sets were evaluated for statistical significance using the Two-Tailed Wilcoxon Signed Rank Test. All statistical analyses were performed using GraphPad Prism 10 (GraphPad Software, San Diego, CA).
Statistical power and sample size estimation
Post hoc power analyses were performed using GraphPad Prism 10 to evaluate whether the sample sizes used in the experiments provided sufficient statistical power. For each comparison, group means, 95% confidence intervals (CIs), and standardized effect sizes (Cohen’s d) were calculated. Sample size estimates required to achieve 80% power were based on unpaired t-tests (two-tailed, α = 0.05). Analyses were conducted for neovascularization and vaso-obliteration. Detailed results of these analyses are reported in the Results section.
Data availability
RNA-Seq data can be accessed at the NCBI Gene Expression Omnibus (GSE292097). All additional data are included in the article and supporting information.
References
Provis, J. M. Development of the primate retinal vasculature. Prog Retin Eye Res. 20, 799–821 (2001).
Saint-Geniez, M. & D’Amore, P. A. Development and pathology of the hyaloid, choroidal and retinal vasculature. Int. J. Dev. Biol. 48, 1045–1058 (2004).
Selvam, S., Kumar, T. & Fruttiger, M. Retinal vasculature development in health and disease. Prog Retin Eye Res. 63, 1–19 (2018).
Chan-Ling, T., Gock, B. & Stone, J. The effect of oxygen on vasoformative cell division. Invest. Ophthalmol. Vis. Sci. 36, 1201–1214 (1995).
Stahl, A. et al. The mouse retina as an angiogenesis model. Invest. Ophthalmol. Vis. Sci. 51, 2813–2826 (2010).
Fruttiger, M. Development of the retinal vasculature. Angiogenesis 10, 77–88 (2007).
Gariano, R. F. & Gardner, T. W. Retinal angiogenesis in development and disease. Nature 438, (2005).
Solebo, A. L., Teoh, L. & Rahi, J. Epidemiology of blindness in children. Arch. Dis. Child. 102, 853–857 (2017).
Hartnett, M. E. Pathophysiology of retinopathy of prematurity. Annu. Rev. Vis. Sci. 15, 39–70 (2023).
Hartnett, M. E. & Penn, J. S. Mechanisms and management of retinopathy of prematurity. N Engl. J. Med. 367, 2515–2526 (2012).
Dammann, O., Hartnett, M. E. & Stahl, A. Retinopathy of prematurity. Dev. Med. Child. Neurol. 65, 625–631 (2023).
Hong, E. H., Shin, Y. U. & Cho, H. Retinopathy of prematurity: a review of epidemiology and current treatment strategies. Clin. Exp. Pediatr. 65, 115–126 (2022).
Hartnett, M. E. Pathophysiology and mechanisms of severe retinopathy of prematurity. Ophthalmology 122, 200–210 (2015).
Connor, K. M. et al. Quantification of oxygen-induced retinopathy in the mouse: a model of vessel loss, vessel regrowth and pathological angiogenesis. Nat. Protoc. 4, 1565–1573 (2009).
Karali, E. et al. VEGF signals through ATF6 and PERK to promote endothelial cell survival and angiogenesis in the absence of ER stress. Mol. Cell. 54, 559–572 (2014).
Binet, F. & Sapieha, P. E. R. Stress and angiogenesis. Cell. Metab. 22, 560–575 (2015).
Papaioannou, A. & Chevet, E. Driving cancer tumorigenesis and metastasis through UPR signaling. In Coordinating Organismal Physiology Through the Unfolded Protein Response Vol. 414 159–192 (Springer, 2017).
Zhang, S. X., Ma, J. H., Bhatta, M., Fliesler, S. J. & Wang, J. J. The unfolded protein response in retinal vascular diseases: implications and therapeutic potential beyond protein folding. Prog Retin Eye Res. 45, 111–131 (2015).
Zhang, S. X. The Endoplasmic reticulum: homeostasis and crosstalk in retinal health and disease. Prog Retin Eye Res 98, (2024).
Koumenis, C. et al. Regulation of protein synthesis by hypoxia via activation of the Endoplasmic reticulum kinase PERK and phosphorylation of the translation initiation factor eIF2ɑ. Mol. Cell. Biol. 22, 7405–7416 (2002).
Bi, M. et al. ER stress-regulated translation increases tolerance to extreme hypoxia and promotes tumor growth. EMBO J. 24, 3470–3481 (2005).
Koumenis, C. & Wouters, B. G. Translating’ tumor hypoxia: unfolded protein response (UPR)-Dependent and UPR-Independent pathways. Mol. Cancer Res. 4, 423–436 (2006).
Liu, L. et al. Targeting the IRE1ɑ/XBP1 and ATF6 arms of the unfolded protein response enhances VEGF Blockade to prevent retinal and choroidal neovascularization. Am. J. Pathol. 182, 1412–1424 (2013).
Liu, L., Liu, C., Lu, Y., Liu, L. & Jiang, Y. ER stress related factor ATF6 and caspase-12 trigger apoptosis in neonatal hypoxic-ischemic encephalopathy. J. Clin. Exp. Pathol. 8, 6960–6966 (2015).
Nakamura, Y. et al. Repression of adipogenesis through promotion of Wnt/β-catenin signaling by TIS7 up-regulated in adipocytes under hypoxia. Biochim. Biophys. Acta. 1832, 1117–1128 (2013).
Wan, Y. et al. New insights into ER stress mediated by ATF6 and IRE1-XBP1 signals in yellow catfish under hypoxia. Aquaculture 597, 741926 (2025).
Jian, Z. et al. Pivotal role of activating transcription factor 6ɑ in myocardial adaptation to chronic hypoxia. Int. J. Biochem. Cell. Biol. 44, 972–979 (2012).
Galdamez, A., Le, M., Chiang, W. C. J., Lee, E. J. & Lin, J. H. Vulnerability of the Nrl-/- Cone-Dominant retina to Endoplasmic reticulum stress. Adv. Exp. Med. Biol. 1468, 319–324 (2025).
Ghosh, R. et al. Transcriptional regulation of VEGF-A by the unfolded protein response pathway. PLOS ONE. 5, e9575 (2010).
Pereira, E. R., Frudd, K., Awad, W. & Hendershot, L. M. Endoplasmic reticulum (ER) stress and hypoxia response pathways interact to potentiate hypoxia-inducible factor 1 (HIF-1) transcriptional activity on targets like vascular endothelial growth factor (VEGF). J. Biol. Chem. 289, 3352–3364 (2014).
Salminen, A., Kauppinen, A., Hyttinen, J. M., Toropainen, E. & Kaarniranta, K. Endoplasmic reticulum stress in Age-Related macular degeneration: trigger for neovascularization. Mol. Med. 16, 535–542 (2010).
Nakamura, S. et al. Mild Endoplasmic reticulum stress promotes retinal neovascularization via induction of BiP/GRP78. PLOS ONE. 8, e60517 (2013).
Kroeger, H. et al. The unfolded protein response regulator ATF6 promotes mesodermal differentiation. Sci Signal 11, (2018).
Shi, S. et al. PERK Inhibition suppresses neovascularization and protects neurons during Ischemia-Induced retinopathy. Invest. Ophthalmol. Vis. Sci. 64, 17 (2023).
Wu, J. et al. ATF6ɑ optimizes Long-Term Endoplasmic reticulum function to protect cells from chronic stress. Dev. Cell. 13, 351–364 (2007).
Lee, E. J. et al. ATF6 is required for efficient Rhodopsin clearance and retinal homeostasis in the P23H Rho retinitis pigmentosa mouse model. Sci. Rep. 11, 16356 (2021).
Lee, E. J. et al. Mutations in unfolded protein response regulator ATF6 cause hearing and vision loss syndrome. J. Clin. Invest. 135, e175562 (2025).
Kohl, S. et al. Mutations in the unfolded protein response regulator ATF6 cause the cone dysfunction disorder achromatopsia. Nat. Genet. 47, 757–765 (2015).
Usui, M. et al. Atf6ɑ-null mice are glucose intolerant due to pancreatic β-cell failure on a high-fat diet but partially resistant to diet-induced insulin resistance. Metabolism 61, 1118–1128 (2012).
Howarth, D. L. et al. Activating transcription factor 6 is necessary and sufficient for alcoholic fatty liver disease in zebrafish. PLOS Genet. 10, e1004335 (2014).
Jin, J. K. et al. ATF6 decreases myocardial ischemia/reperfusion damage and links ER stress and oxidative stress signaling pathways in the heart. Circ. Res. 120, 82–875 (2017).
Yoshikawa, A. et al. Deletion of Atf6ɑ impairs astroglical activation and enhances neuronal death following brain ischemia in mice. J. Neurochem. 132, 342–353 (2015).
Ta, H. M. et al. Atf6ɑ deficiency suppresses microglial activation and ameliorates pathology of experimental autoimmune encephalomyelitis. J. Neurochem. 139, 1124–1137 (2016).
Smith, L. E. H. et al. Oxygen-Induced retinopathy in the mouse. Invest. Ophthalmol. Vis. Sci. 35, 101–111 (1994).
Mezu-Ndubuisi, O. J. et al. Intravitreal delivery of VEGF-A165-loaded PLGA microparticles reduces retinal vaso-obliteration in an in vivo mouse model of retinopathy of prematurity. Curr. Eye Res. 44, 275–286 (2018).
Mezu-Ndubuisi, O. J. et al. Long-term evaluation of retinal morphology and function in a mouse model of oxygen-induced retinopathy. Mol. Vis. 26, 257–276 (2020).
Guaiquil, V. H. et al. A murine model for retinopathy of prematurity identifies endothelial cell proliferation as a potential mechanism for plus disease. Invest. Ophthalmol. Vis. Sci. 54, 5294–5302 (2013).
Forrester, J. V., Shafiee, A., Schröder, S., Knott, R. & McIntosh, L. The role of growth factors in proliferative diabetic retinopathy. Eye 7, 276–287 (1993).
Connor, K. M. et al. Increased dietary intake of ω-3-polyunsaturated fatty acids reduces pathological retinal angiogenesis. Nat. Med. 13, 868–873 (2007).
Hartnett, M. E. & Stahl, A. Laser versus Anti-VEGF: A paradigm shift for Treatment-Warranted retinopathy of prematurity. Ophthalmol. Ther. 12, 2241–2252 (2023).
Li, J., Wang, J. J., Yu, Q., Wang, M. & Zhang, S. X. Endoplasmic reticulum stress is implicated in retinal inflammation and diabetic retinopathy. FEBS Lett. 583, 1521–1527 (2009).
Bommiasamy, H. et al. ATF6ɑ induces XBP1-independent expansion of the Endoplasmic reticulum. J. Cell. Sci. 122, 1626–1636 (2009).
Han, J. et al. ER-stress-induced transcriptional regulation increases protein synthesis leading to cell death. Nat. Cell. Biol. 15, 481–490 (2013).
Harding, H. P. et al. An integrated stress response regulates amino acid metabolism and resistance to oxidative stress. Mol. Cell. 11, 619–633 (2003).
Kroeger, H. et al. ATF6 is essential for human cone photoreceptor development. Proc. Natl. Acad. Sci. 118, e2103196118 (2021).
Lee, A. H., Iwakoshi, N. N. & Glimcher, L. H. XBP-1 regulates a subset of Endoplasmic reticulum resident chaperone genes in the unfolded protein response. Mol. Cell. Biol. 23, 7448–7459 (2003).
Lu, P. D. et al. Cytoprotection by pre-emptive conditional phosphorylation of translation initiation factor 2. EMBO J. 23, 169–179 (2004).
Shaffer, A. L. et al. XBP1, downstream of Blimp-1, expands the secretory apparatus and other organelles, and increases protein synthesis in plasma cell differentiation. Immunity 21, 81–93 (2004).
Sriburi, R. et al. Coordinate regulation of phospholipid biosynthesis and secretory pathway gene expression in XBP-1(S)-induced Endoplasmic reticulum biogenesis. J. Biol. Chem. 282, 7024–7034 (2007).
Lee, E. J. et al. Mitochondria and Endoplasmic reticulum stress in retinal organoids from patients with vision loss. Am. J. Pathol. 193, 1721–1739 (2023).
Gallagher, C. M. et al. Ceapins are a new class of unfolded protein response inhibitors, selectively targeting the ATF6ɑ branch. eLife 5, e11878. https://doi.org/10.7554/eLife.11878 (2016).
Milde, F., Lauw, S., Koumoutsakos, P. & Iruela-Arispez, M. L. The mouse retina in 3D: quantification of vascular growth and remodeling. Integr. Biol. Camb. 5, 1426–1438 (2013).
Gerhardt, H. et al. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell. Biol 161, (2003).
Benedito, R. et al. The Notch ligands Dll4 and Jagged1 have opposing effects on angiogenesis. Cell 137, 1124–1135 (2009).
Blanco, R. & Gerhardt, H. VEGF and Notch in tip and stalk cell selection. Cold Spring Harb Perspect. Med. 3, a006569 (2013).
Hellström, M. et al. Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature 445, 776–780 (2007).
Pawlowska, Z., Baranska, P., Jerczynska, H., Koziolkiewicz, W. & Cierniewski, C. S. Heat shock proteins and other components of cellular machinery for protein synthesis are up-regulated in vascular endothelial cell growth factor-activated human endothelial cells. Proteomics 5, 1217–1227 (2005).
Malhotra, J. D. & Kaufman, R. J. The Endoplasmic reticulum and the unfolded protein response. Semin Cell. Dev. Biol. 18, 716–731 (2007).
Walter, P. & Ron, D. The unfolded protein response: from stress pathway to homeostatic regulation. Science 334, 1081–1086 (2011).
Drogat, B. et al. IRE1 signaling is essential for ischemia-Induced vascular endothelial growth Factor-A expression and contributes to angiogenesis and tumor growth in vivo. Cancer Res. 67, 6700–6707 (2007).
Moszyńska, A., Collawn, J. F. & Bartoszewski, R. IRE1 endoribonuclease activity modulates hypoxic HIF-1ɑ signaling in human endothelial cells. Biomolecules 10, 895 (2020).
Reimold, A. M. et al. An essential role in liver development for transcription factor XBP-1. Genes Dev 14, (2000).
Zhang, K. et al. The unfolded protein response sensor IRE1ɑ is required at 2 distinct steps in B cell lymphopoiesis. J. Clin. Invest. 115, 268–281 (2005).
Harding, H. P. et al. Diabetes mellitus and exocrine pancreatic dysfunction in Perk-/- mice reveals a role for translational control in secretory cell survival. Mol. Cell. 7, 1153–1163 (2001).
Zhang, P. et al. The PERK eukaryotic initiation factor 2ɑ kinase is required for the development of the skeletal system, postnatal growth, and the function and viability of the pancreas. Mol. Cell. Biol. 22, 3864–3874 (2002).
Reynaud, X., Hansen, R. M. & Fulton, A. B. Effect of prior oxygen exposure on the electroretinographic responses of infant rats. Invest. Ophthalmol. Vis. Sci. 36, 1071–1079 (1995).
Dembinska, O., Rojas, L. M., Chemtob, S. & Lachapelle, P. Evidence for a brief period of enhanced oxygen susceptibility in the rat model of oxygen-Induced retinopathy. Invest. Ophthalmol. Vis. Sci. 43, 2481–2490 (2002).
Liu, K., Akula, J. D., Falk, C., Hansen, R. M. & Fulton, A. B. The retinal vasculature and function of the neural retina in a rat model of retinopathy of prematurity. Invest. Ophthalmol. Vis. Sci. 47, 2639–2647 (2006).
Hansen, R. M. & Fulton, A. B. Background adaptation in children with a history of mild retinopathy of prematurity. Invest. Ophthalmol. Vis. Sci. 41, 320–324 (2000).
Fulton, A. B., Hansen, R. M., Petersen, R. A. & Vanderveen, D. K. The rod photoreceptors in retinopathy of prematurity: an electroretinographic study. Arch. Ophthalmol. 119, 499–505 (2001).
Vessey, K. A., Wilkinson-Berka, J. L. & Fletcher, E. L. Characterization of retinal function and glial cell response in a mouse model of Oxygen-Induced retinopathy. J. Comp. Neurol. 519, 506–527 (2011).
Nakamura, S. et al. Morphological and functional changes in the retina after chronic Oxygen-Induced retinopathy. PLOS ONE. 7, e32167 (2012).
Bretz, C. A. et al. Erythropoietin receptor signaling supports retinal function after vascular injury. Am J. Pathol (2020).
Akula, J. D. et al. The anatomy of the rat eye with oxygen-induced retinopathy. Doc. Ophthalmol. 120, 41–50 (2010).
Soetikno, B. T. et al. Inner retinal oxygen metabolism in the 50/10 oxygen-induced retinopathy model. Sci. Rep. 5, 16752 (2015).
Rimawi, M. F. et al. Early efficacy evaluation of ORIN1001, a first in class IRE1 alpha inhibitor, in advanced solid tumors. J. Clin. Oncol. 41, 1092 (2023).
Dudek, A. Z. et al. A multicenter, open-label, phase 1a study of HC-5404 in patients with advanced solid tumors. J. Clin. Oncol. 42, e15118 (2024).
Grandjean, J. M. D. et al. Deconvoluting Stress-Responsive proteostasis signaling pathways for Pharmacologic activation using targeting RNA-sequencing. ACS Chem. Biol. 14, 784–795 (2019).
Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinform. 12, 323 (2011).
Love, M. I., Huber, W. & Anders, S. Moderated Estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Gerdes, J., Schwab, U., Lemke, H. & Stein, H. Production of a mouse monoclonal antibody reactive with a human nuclear antigen associated with cell proliferation. Int. J. Cancer. 31, 13–20 (1983).
Nikolova-Krstevski, V. et al. ERG is required for the differentiation of embryonic stem cells along the endothelial lineage. BMC Dev. Biol. 9, 72 (2009).
Engelbrecht, E., Metzler, M. A. & Sandell, L. L. Retinoid signaling regulates angiogenesis and blood-retinal barrier integrity in neonatal mouse retina. Microcirculation 29, e12752 (2022).
Rust, R., Grönnert, L., Dogançay, B. & Schwab, M. E. A revised view on growth and remodeling in the retinal vasculature. Sci. Rep. 9, 3263 (2019).
Funding
This research was supported by NIH awards P30EY026877, R01EY027735; R01AG046495; R01NS125674; R01NS088485; CIRM DISC2-10973; and VA Merit I01BX002284.
Author information
Authors and Affiliations
Contributions
A.B. performed experiments, analyzed data, prepared figures, and wrote the manuscript. S.P., S.Pa., K.K., A.G., H.M., and M.S.D.-A. contributed to experiments, data analysis, and manuscript editing. M.E.H. edited and reviewed the manuscript. E.-J.L. and J.H.L conceived the study, supervised the project, contributed to data interpretation, and edited the manuscript. J.H.L. provided resources. All authors reviewed and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Bradley, A., Park, S., Park, S. et al. Targeting ATF6 reduces pathological neovascularization and improves visual outcomes in retinal disease models. Sci Rep 15, 33070 (2025). https://doi.org/10.1038/s41598-025-15393-y
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41598-025-15393-y
