Assessment of anti-VEGF intravitreal injection effects on murine neonatal Schlemm’s canal morphology

Abstract

Anti-VEGF intravitreal injections are the standard treatment for retinopathy of prematurity (ROP). Schlemm’s canal, essential for regulating the conventional aqueous humor outflow pathway, is still in development in premature infants. Anti-VEGF injections in premature infants may negatively impact Schlemm’s canal development, potentially increasing glaucoma risk. We sought to determine the effects of anti-VEGF intravitreal injections on Schlemm’s canal morphology in neonatal mice. We therefore assessed Schlemm’s canal morphology following anti-VEGF intravitreal injections in neonatal mice subjected to the oxygen-induced retinopathy (OIR) model with features of ROP. Intraocular pressure (IOP) was measured 5 days following injections, and ex vivo analyses of retinal flatmounts and anterior eye cups were performed to assess retinal vasculature (isolectin B4) and Schlemm’s canal morphology (CD31). Intravitreal injections of anti-VEGF antibody reduced both retinal vascular and avascular areas in OIR retinas compared to control. However, anti-VEGF injections did not induce changes in Schlemm’s canal morphology or sustained changes in IOP. Our findings indicate that a single intravitreal injection of anti-VEGF does not significantly affect murine Schlemm’s canal morphology or IOP. If anti-VEGF intravitreal injections predispose infants to glaucoma later in life, these results suggest it is unlikely due to direct effects on Schlemm’s canal morphology during development.

Introduction

Premature infants requiring supplemental oxygen to support underdeveloped lungs are often exposed to the hyperoxic environment of a high-oxygen incubator, increasing the risk for developing retinopathy of prematurity (ROP). Exposure to these hyperoxic environments halts the development of retinal vasculature. Upon return to room air, the resulting relative hypoxia can trigger aberrant retinal neovascularization^1–4. Although supplemental oxygen levels are closely monitored in neonatal intensive care units (NICUs), achieving the optimal oxygen balance continues to be challenging and can still result in complications such as ROP^5,6. ROP remains the leading cause of childhood blindness affecting premature infants⁷.

Treatments for ROP have evolved over the decades, but current strategies include intravitreal injection of anti-VEGF biologics and retinal laser photocoagulation. Anti-VEGF therapy is often considered advantageous for treatment of high-risk ROP cases due to its targeted approach, high efficacy, minimized risk of adverse effects such as myopia, and preservation of peripheral retinal vasculature⁸.

While anti-VEGF agents are well established in the clinical management of ROP, their long-term effects in neonates remain unclear. Since the first anti-VEGF agents were approved in 2004, the earliest ROP patients treated with anti-VEGF agents are only now reaching adulthood. Emerging evidence suggests potential long-term complications, including a predisposition to secondary glaucoma⁹. Several studies have also noted that intravitreal injections can cause transient increases in intraocular pressure (IOP), the leading risk factor for glaucoma^10,11.

Schlemm’s canal is a key regulator of IOP and is responsible for draining approximately 90% of the aqueous humor via the conventional outflow pathway^12,13. In humans, Schlemm’s canal development is not complete until approximately 40 weeks of gestation. Schlemm’s canal may be particularly vulnerable to anti-VEGF treatment in neonates due to its vascular endothelial cell composition^13,14.

Given these concerns, we investigated whether anti-VEGF intravitreal injections induce changes in Schlemm’s canal morphology in the oxygen-induced retinopathy (OIR) mouse model, which mimics key features of ROP. We hypothesized that anti-VEGF treatment during Schlemm’s canal development might alter its structure, resulting in increased IOP and a higher risk of glaucoma. Our findings indicate that a single intravitreal injection of anti-VEGF did not induce robust morphological changes in Schlemm’s canal or elevate IOP in the short-term. However, the potential for early anti-VEGF exposure to cause glaucoma risk later in life through delayed or indirect mechanisms remains to be determined.

Results

Anti-VEGF treatment does not alter IOP or body weight in OIR mice

To explore if anti-VEGF treatment affects Schlemm’s canal morphology in neonatal mice, we utilized the OIR mouse model as an experimental analogue of ROP. Unlike humans, who complete retinal vasculature development in utero by full term, murine retinal vascularization completes approximately three weeks after birth, thus mimicking the premature retinal status of infants with ROP¹⁵. On postnatal day 7 (P7), neonatal pups and their nursing mothers were exposed to 75% O₂ (hyperoxia) for five days until P12. This initial hyperoxic environment induces vessel pruning in the central retina (vaso-obliteration). At P12, the animals were returned to room air (approximately 21% O₂), initiating a relative hypoxic transition phase that induces new vessel growth that peaks at P17 (neovascularization). Consistent with usual practices in this model, a single intravitreal injection was administered on P12 upon return to room air¹⁶. Two groups of OIR pups received a single intravitreal injection of either anti-VEGF antibody or vehicle (sham) in both eyes. A third group of OIR pups received no injection. As an additional control, a group of pups maintained in normal room air (normoxia) for the duration of the study was divided into three parallel groups: no injection control, anti-VEGF injected, and vehicle (sham) injected (Fig. 1a). The weights of the pups were monitored from P12 to P17, as a reduction in OIR pup weight has been shown to correlate with increased retinal neovascularization^17,18. No significant differences in body weights were observed among treatment groups (Fig. 1b). In addition, IOPs were measured at P17 to determine lasting changes in IOP five days after receiving a single intravitreal injection. IOP values were not significantly different among OIR groups (Fig. 1c). Together, these results indicate that intravitreal anti-VEGF administration does not induce lasting changes in IOP or affect systemic growth in neonatal OIR mice five days following intravitreal injection.

Fig. 1.

Intravitreal injection of anti-VEGF at P12 does not induce changes in IOP at P17. (a) schematic; P7 neonatal pups and nursing dams were placed at either 75% O₂ (OIR) or remained in room air (normoxia) until P12. On P12, a group from both normoxia and OIR mice received a single intravitreal injection of anti-VEGF₁₆₄ (anti-VEGF) or PBS (sham). A group of mice from OIR and normoxia did not receive either injection (no injection). On P17, IOPs were measured with a tonometer before mice were euthanized. Posterior eye cups and anterior chambers were collected for histological analysis. (b) P17 neonatal pup weights. Weights were not different among treatment groups. Two-way ANOVA; n = 3 mice per group. (c) IOPs of P17 neonatal pups were not different among treatment groups. Two-way ANOVA; n = 5–6 eyes per group (3 mice per group representing both right and left eyes from each mouse). Data represented as mean ± SEM.

Anti-VEGF decreases neovascular and vaso-obliterated areas in the OIR mouse model

Before assessing potential effects of anti-VEGF treatment on Schlemm’s canal morphology, we first sought to validate that the intravitreal injection exhibited the expected therapeutic effects in the OIR model¹⁶. At P17, after IOP measurement, mice were euthanized, eyes were enucleated, and the globes were dissected into the posterior and anterior segments. The retinas were then isolated from the posterior segment for vascular analysis. Retinal vasculature was visualized using Griffonia simplicifolia isolectin B4 (GS-IB4) staining. As expected, we observed central vaso-obliteration and peripheral neovascularization in P17 OIR retinas, in contrast to the complete retinal vasculature patterning observed in age-matched room air controls (Fig. 2a). Intravitreal injection of either anti-VEGF or vehicle exhibited no effects on the vasculature of retinas from mice that remained at room air (Fig. 2a). In OIR mice, anti-VEGF treatment significantly decreased both the extent of peripheral neovascularization (Fig. 2b,c) and the central vaso-obliterated area (Fig. 2d,e), consistent with previous studies¹⁶. Sham injection had no effect on either parameter (Fig. 2b–e). Together, these findings confirm that intravitreal anti-VEGF antibody administration effectively mitigates the pathological vascular changes characteristic of the OIR mouse model, validating the experimental conditions used for subsequent analyses of Schlemm’s canal morphology.

Fig. 2.

Validation of the OIR model and anti-VEGF treatment. (a) Representative en face view of P17 retinal flatmounts from OIR mice that either received no intravitreal injection, a sham injection, or an anti-VEGF injection, immunostained with GS-IB4 (green, vasculature). (b) Representative P17 retinal flatmounts from OIR mice immunostained with GS-IB4 (green) with neovascular (NV) tufts in white. (c) Quantification of retinal neovascular area upon sham, anti-VEGF, or no injection treatments. Anti-VEGF significantly decreased the neovascular area. One-way ANOVA with Tukey’s post hoc tests; n = 5–6 retinas representing 3 mice per treatment. (d) Representative P17 retinal flatmounts from OIR mice immunostained with GS-IB4 (green) with vaso-obliterated (VO) area in white. (e) Quantification of vaso-obliterated area upon sham, anti-VEGF, or no injection treatments. Anti-VEGF significantly decreased VO area. One-way ANOVA with Tukey’s post hoc tests; n = 5–6 retinas representing 3 mice per treatment. Scale bars = 500 μm; data represented as mean ± SEM.

Intravitreal anti-VEGF treatment does not induce robust changes in Schlemm’s canal morphology

Having confirmed that anti-VEGF treatment exerted the expected therapeutic effects in OIR mice, we next examined whether these interventions affected Schlemm’s canal morphology. The corresponding anterior segments were collected and immunostained with labeled anti-CD31 antibody to visualize Schlemm’s canal. High-resolution tile-scan images were acquired, and the total area of Schlemm’s canal was quantified for each experimental group (Fig. 3a). An aligned rank transform (ART) ANOVA to evaluate the effects of treatment (normoxia vs. OIR) and injection (no injection, sham, anti-VEGF) revealed that the main effect of injection was statistically significant (p < 0.01), indicating differences among the injection groups (Table 1). The treatment x injection interaction approached significance (p = 0.072), suggesting a possible differential effect of injection, depending on the treatment (Table 1). Pairwise contrasts revealed few differences, notably normoxia without injection vs. normoxia + anti-VEGF (p = 0.005) and normoxia without injection vs. OIR + anti-VEGF (p = 0.017) (Table 2).

Fig. 3.

Area analysis of Schlemm’s canal. (a) Anterior segment flatmount immunostained for CD31 highlights Schlemm’s canal morphology. Composite maximum projection images show a full 360° view of the vessel and detailed structure (insert; scale bar = 100 μm). Representative tissue from an OIR-treated animal that did not receive an injection. (b), Representative images of Schlemm’s canal from animals of different treatment groups; c, Area analysis of Schlemm’s canal using custom contrasts for a mixed-effects model, n = 5–6 eyes, representing 3 mice per treatment group. No significant difference between groups. Data represented as mean ± SEM.

Table 1.

Aligned rank transform ANOVA table; df = degrees of freedom; df.res = residual degrees of freedom; Sum sq = sum of squares; Sum Sq.res = residual sum of squares. Significant values are in bold (p < 0.01).

Term	df	df.res	Sum Sq	Sum Sq.res	F value	p value
Treatment	1	7	0.462963	107.83333	0.0300532	0.8672760
Injection	1	7	68.907407	37.83333	12.7493882	0.0090893
Treatment: injection	1	7	40.907407	63.83333	4.4859298	0.0719335

Table 2.

Pairwise contrasts for treatment x injection; none = no injection. SE = standard error; df = degrees of freedom. Significant values are in bold (p < 0.05).

Contrast	Estimate	SE	df	t ratio	p value
Normoxia, anti-VEGF v. normoxia, none	−7.000	1.327	7	−5.274	0.005
Normoxia, anti-VEGF v. OIR, anti-VEGF	−0.833	1.484	7	−0.562	0.940
Normoxia, anti-VEGF v. OIR, none	−4.667	1.327	7	−3.516	0.038
Normoxia, none v. OIR, anti-VEGF	6.167	1.484	7	4.155	0.017
Normoxia, none v. OIR, none	2.333	1.327	7	1.758	0.365
OIR, anti-VEGF, v. OIR, none	−3.833	1.484	7	−2.583	0.129

Within-subject retinal area measurements demonstrated a high correlation (r = 0.88, 95% confidence interval: 0.69–0.96), justifying the inclusion of a random intercept for subject in a linear mixed-effects model. Interestingly, no statistically significant changes in Schlemm’s canal morphology or area were observed among any of the groups (Table 3; Fig. 3b,c). These results indicate that none of OIR, intravitreal anti-VEGF administration, nor a single intravitreal injection itself at P12 robustly alters Schlemm’s canal morphology in P17 neonatal mice.

Table 3.

Custom contrasts for mixed-effects model. Estimate = estimated differences; SE = standard error; df = degrees of freedom; norm. = normoxia; none = no injection.

Contrast	Estimate	SE	df	t ratio	p value
Norm, none v. norm. sham	−3636.667	71327.49	11.828	−0.051	0.960
Norm, none v. norm. Anti-VEGF	−15068.833	71327.49	11.828	−0.211	0.836
Norm, none v. OIR, none	21065.667	71327.49	11.828	0.295	0.773
Norm, none v. OIR, sham	95472.333	71327.49	11.828	1.339	0.206
Norm, none v. OIR, anti-VEGF	116377.227	72160.95	12.314	1.613	0.132
Norm, sham v. OIR, sham	99109.000	71327.49	11.828	1.389	0.190
Norm, anti-VEGF v. OIR, sham	80403.500	71327.49	11.828	1.127	0.282
OIR, none v. OIR, sham	74406.667	71327.49	11.828	1.043	0.318
OIR, none v. OIR, anti-VEGF	95311.561	72160.95	12.314	1.321	0.211
OIR, sham v. OIR, anti-VEGF	20904.894	72160.95	12.314	0.290	0.777
Norm, anti-VEGF v. OIR, anti-VEGF	101308.394	72160.95	21.314	1.404	0.185

Discussion

Intravitreal anti-VEGF injections are the current standard of care treatment for ROP. However, concerns remain regarding potential long-term consequences of VEGF inhibition during ocular development. Recent clinical evidence indicates that neonates who receive intravitreal injections of anti-VEGF biologics are at an increased risk of developing secondary glaucoma spanning from early childhood to adulthood^9,10,19. This led us to hypothesize that anti-VEGF therapy may induce changes in Schlemm’s canal morphology and inhibit the aqueous outflow pathway. In this study, we assessed the effects of intravitreal injection of anti-VEGF on IOP and Schlemm’s canal morphology in neonatal mice. Our findings demonstrate that following a single intravitreal injection at P12, P17 neonatal mice exhibit no detectable changes in Schlemm’s canal morphology or IOP. As expected, we observed that intravitreal injection of anti-VEGF significantly decreased neovascular and avascular areas in P17 OIR retinas, consistent with previous studies¹⁶. P17 OIR retinas exhibit peak neovascularization, after which the neovascular phenotype naturally regresses, regardless of treatment. For this reason, we chose to analyze retinas at this timepoint, since it most accurately represents the retinal pathology that ROP patients manifest. The reduction in avascular and neovascular areas at this timepoint demonstrates that the anti-VEGF treatment promoted physiological revascularization. Importantly, treatment did not significantly affect body weight, since reduced neonatal weight has been associated with increased retinal neovascularization and avascular areas in the OIR mouse model^17,18.

Five days following the single intravitreal injection of either anti-VEGF or vehicle, IOPs were measured. While IOP was not different among the treatment groups, the vehicle treated group trended towards a slight increase in IOP. This may reflect that the intravitreal injection itself, but not anti-VEGF, might affect IOP. Indeed, several studies have noted acute increases in IOP following intravitreal injection in humans, attributed to a transient increase in vitreous volume^20–22. However, this increase in IOP is transient and usually resolves within an hour following injection. Our results demonstrate that a few days following injection, neither intravitreal injections nor anti-VEGF induced sustained changes in IOP.

We assessed the morphology of Schlemm’s canal using an ex vivo area analysis following CD31 labeling of Schlemm’s canal at P17. This timepoint was selected to capture intermediate-term effects on Schlemm’s canal. A decrease in Schlemm’s canal area would indicate an impairment of the aqueous humor outflow pathway, a key pathological feature of glaucoma. ART ANOVA analysis revealed that main effect of injection was statistically significant, indicating that the type of injection the mice received matters in terms of Schlemm’s canal area. The interaction of treatment x injection approached significance (p = 0.072), indicating that there is a trend that the effect of the injection might depend on the treatment condition (normoxia vs. OIR). Pairwise contrasts revealed differences among normoxia (no injection) vs. normoxia + anti-VEGF and normoxia (no injection) vs. OIR + anti-VEGF. When accounting for repeated measures and within-subject variability with a linear mixed-effects model, no significant change in Schlemm’s canal morphology was detected. This suggests that the observed pairwise differences do not reflect a robust or generalized effect. It is important to note that the present study’s sample size only has the power to detect substantial differences in structural changes to Schlemm’s canal. While future studies with larger sample sizes could increase resolution to detect subtle structural changes to Schlemm’s canal, these results demonstrate that anti-VEGF injections do not robustly mediate large alterations to Schlemm’s canal morphology. To our knowledge, no previous studies have investigated Schlemm’s canal morphology after intravitreal injection. However, a limitation of this study is that assessment of Schlemm’s canal morphology was restricted to measurement of Schlemm’s canal area. More detailed analyses, including evaluation of vascular and lymphatic density, trabecular meshwork morphology, and canal lumen architecture were not performed here. Such analyses could reveal structural changes not captured by the area analysis alone. Future investigations evaluating these parameters could more comprehensively evaluate whether and how anti-VEGF agents affect Schlemm’s canal.

While the neonatal IOPs and Schlemm’s canal area did not change at P17, it is possible that changes in IOP and Schlemm’s canal morphology could be observed if the parameters of the study were different. For example, to determine if temporal effects exist, future work is required to investigate both acute and long-term effects on IOP and Schlemm’s canal morphology after anti-VEGF treatment in mice. The time of receiving anti-VEGF agents in relationship to Schlemm’s canal developmental maturation may also influence glaucoma susceptibility later in life. In mice, Schlemm’s canal and functional aqueous humor dynamics are not fully mature until a few weeks into the postnatal window. Importantly, Schlemm’s canal reaches near-maturity by approximately P12, suggesting that anti-VEGF administration at this stage may not perturb canal formation^23,24. While most of the premature infants that receive anti-VEGF injections for ROP are extremely preterm, we were unable to assess if earlier administration of anti-VEGF affects Schlemm’s canal morphology and IOP in the OIR model, because the neonatal mice must remain in a hyperoxic chamber until P12 to ensure they develop a ROP phenotype. Future studies investigating earlier anti-VEGF exposure and comparative studies of Schlemm’s canal development in mice and humans could help identify developmental windows during which VEGF inhibition may pose the greatest risk. While future studies are also warranted to investigate if repeated anti-VEGF intravitreal injections result in changes in IOP or Schlemm’s canal morphology, it is important to note that treatment of ROP is commonly done with a single intravitreal injection of anti-VEGF^25,26, and the parameters of the standard OIR model require intervention at approximately P12. Furthermore, after P17, the neovascular phenotype in the OIR model naturally regresses. This natural regression limits our ability to explore long-term effects on anti-VEGF on Schlemm’s canal or IOP in this model, as this natural regression could confound any findings. Improved mouse models of ROP and future longitudinal studies will help identify sustained effects on IOP and Schlemm’s canal following intravitreal injection of anti-VEGF.

A further limitation of this study was the use of an antibody against the murine isoform VEGF₁₆₄ rather than clinically used agents. Current anti-VEGF therapeutics for ROP include bevacizumab and ranibizumab²⁷. These agents target all isoforms of VEGF-A rather than just VEGF₁₆₅^28,29. Similarly, aflibercept, the only approved treatment for ROP in the United States by the FDA, binds VEGF-A, VEGF-B, and placental growth factor (PlGF)³⁰. Recent studies have suggested that bevacizumab can lead to the development of secondary glaucoma later in life in ROP patients⁹, which raises the possibility that broader VEGF inhibition, rather than blockade of VEGF₁₆₄ (VEGF₁₆₅ in humans) alone, contributes to this pathology. Supporting this hypothesis, pharmacological inhibition of VEGFR2 has been shown to decrease Schlemm’s canal outflow in mice, and its corresponding ligand, VEGF-A, has been implicated in regulating the conventional outflow pathway^31,32. However, mouse models of ROP demonstrate limited responsiveness to clinically used agents, necessitating the use of a murine-specific anti-VEGF₁₆₄ antibody in the present study. In addition, inherent interspecies differences in ocular anatomy and physiology limit the ability to fully recapitulate human glaucomatous phenotypes in mice. Thus, potential effects of anti-VEGF agents on Schlemm’s canal development in humans may not be observed in this model. The single time-point measurement at P17 represents another limitation of this study, as future studies that repeatedly assess IOP and Schlemm’s canal morphology over time would increase the resolution of this study.

In conclusion, we observed no significant changes in Schlemm’s canal morphology or IOP at P17 following intravitreal anti-VEGF injection at P12 in mice. While these results suggest that acute alterations in Schlemm’s canal morphology or IOP are not evident at the timepoints examined in this study, the potential for long-term effects or repeated anti-VEGF exposure on conventional outflow structures remains unclear. These results do not exclude the possibility that anti-VEGF exposure during infancy could contribute to glaucoma risk later in life through delayed or indirect mechanisms. Future studies are warranted to fully understand whether intravitreal injections of anti-VEGF in neonates increase glaucoma risk later in life.

Methods

Animals

All animal experiments and procedures were conducted in accordance with the ethical and legal requirements of the Animals for Research Act of Ontario and the Canadian Council on Animal Care Guidelines for the Care and Use of Laboratory Animals. The animal experiments in these studies were approved by The Center for Phenogenomics (TCP) and University of Toronto Animal Care Committees (Animal Use Protocol 28–0428 H) and followed the Association for Research in Vision and Ophthalmology (ARVO) guidelines for the “Use of Animals in Ophthalmic and Visual Research” and the “Animal Research: Reporting of In vivo Experiments” (ARRIVE) guidelines. Timed-pregnancy C57BL/6J female mice were purchased from TCP. All mice were housed under standard conditions in TCP: 12 h:12 h light-dark cycle, temperature at 21 ± 1 °C, and humidity at 30–60%. Animals were fed Inotiv standard diet rodent chow (TD.2918X) and chlorinated water ad libitum. Treatments were randomly assigned by cage.

Oxygen-induced retinopathy (OIR) mouse model

Mice were subjected to OIR as previously described^16,33. Postnatal day 7 (P7) pups of both sexes and nursing mothers were exposed to 75% O₂ (hyperoxia) in a hyperoxia chamber (BioSpherix ProOx360, Parish, NY, USA) for 5 days (P7-P12) to initiate retinal vascular obliteration. Mice were returned to room air (normoxia; ~21% O₂) on P12. On P17, IOPs of neonatal mice were taken using an Icare TONOVET^® for animals. As per TONOVET^® instructions, IOP of both eyes was measured 6 consecutive times by placing the tonometer directly across the eye, proximal enough for the probe to make contact with the eye surface. The average readings were reported as the IOP of each eye. Across all treatment groups, one eye of one mouse was excluded from all analyses due to microphthalmia. Treatment groups were randomized among three litters, with no more than six pups per litter.

Intravitreal injections

P12 OIR and control pups were anesthetized with isoflurane, and eyelids were opened using a sterile 30-gauge needle. Tetracaine hydrochloride (0.5%) was used as a topical anesthetic. After dilation of the eyes with topical drops of 1% tropicamide and 2.5% phenylephrine ophthalmic solutions (Alcon Canada Inc, Mississauga, ON, Canada), a small incision was made at the nasal-temporal ora serrata with a sterile 30-gauge insulin syringe needle to gain access to the posterior vitreous chamber. A sterile Nanoject III microinjector (Drummond Scientific Company, Broomall, PA, USA) was used to inject either vehicle (0.3µL PBS) or murine anti-VEGF₁₆₄ antibody (0.3µL for a final dose of 5ng/eye¹⁶; R&D Systems, Minneapolis, MN, USA). Both eyes were injected with either anti-VEGF₁₆₄ antibody or vehicle for each animal. A successful injection was determined by observing mild perturbation of the anterior chamber and absence of backflushing. After the pups recovered from anesthesia, they were returned to nursing mothers remaining at room air (normoxia) under normal husbandry conditions until P17.

Immunohistochemistry of retinal flatmounts

At P17, the pups were euthanized by CO₂ and cervical dislocation. Eyes were enucleated, corneas were perforated with a 29-gauge needle, and the whole eye was fixed in 4% paraformaldehyde (PFA) (Thermo Fisher Scientific, Waltham, MA, USA) in PBS for one hour immediately after euthanasia. Fixed whole eyes were then dissected under a dissecting microscope and bisected into anterior and posterior halves. The retinas were then isolated from the posterior halves. The retinas and anterior globe halves then underwent a secondary fixation in 4% and 2% PFA in PBS overnight at 4 °C, respectively. For en face retinal flatmounts, retinas were rinsed with PBS and permeabilized with 1% SDS in PBS for 3 min and blocked with 2.5% bovine serum albumin (BSA), 0.3% Triton X-100 in PBS for one hour at room temperature. GS-IB4 biotin-XX conjugate (Cat. No. I21414, Thermo Fisher Scientific) was diluted 1:250 in 0.5% BSA 0.3% Triton X-100 in PBS and was incubated with retinas for 48 h at 4 °C. After three 20-minute PBS washes, retinas were incubated with streptavidin-DyLight 488 (Cat. No. 21832, Thermo Fisher Scientific) at a 1:400 dilution ratio in 0.5% BSA 0.3% Triton X-100 in PBS for 24 h at 4 °C. After rinsing with PBS, immunostained retinas were mounted onto glass slides, cover-slipped with Fluoromount-G (Cat. No. 0100-01, Southern Biotechnology, Birmingham, AL, USA) and stored at 4 °C protected from light until imaging. Tile scan images of whole retinas were acquired with a 10x objective on a Leica TCS SP8 Confocal microscope with LAS-X imaging software (Leica Microsystems, Mannheim, Germany).

Immunohistochemistry of Schlemm’s canal

Schlemm’s canal tissue preparation, staining, and imaging were conducted according to previous studies, with some modifications^34–36. After fixation, anterior eye tissues were washed in PBS and transferred to 70% molecular biology grade ethanol and stored at 4 °C until further processing. Anterior eye cups were dissected to remove the retinas, extraneous tissue such as tendon, muscle, conjunctiva, and any excess sclera. Tissues were placed in blocking/staining buffer (5% normal donkey serum (Cat. 017-000-121, Jackson ImmunoResearch Labs, West Grove, PA, USA), 5% BSA (Cat. 37525 Thermo Scientific), 0.5% Triton X-100, and 0.01% sodium azide in PBS) on a nutator overnight at 4 °C. The blocking buffer was replaced with a fresh staining buffer containing mouse anti-CD31 antibody diluted at 1:50 (Cat. 557355 BD Pharmingen, Franklin Lakes, NJ, USA). The tissues were placed on a nutator at 4 °C for two days. Primary antibody solution was removed and the eye cups washed three times for one hour at room temperature in PBS containing 0.01% Tween 20. Staining buffer containing goat anti-rat Alexa 568 diluted at 1:1000 (Cat. A11077, Invitrogen, Carlsbad, CA, USA) was added, and tissues were incubated on a nutator at 4 °C for 2 days. The eye cups were washed six times for one hour at room temperature in PBS containing 0.01% Tween 20. Eye cups were placed in cold PBS, and a series of relief cuts were made towards the center of the cornea to attain a flower petal-like appearance and flatten the tissues. Tissues were placed on a glass slide, cornea side up, flattened, excess fluid drained, and mounted (Immun-Mount, Cat. 9990402, Epredia, Kalamazoo, MI, USA) with a #1.5 coverslip.

Slides were dried for 16 h at 4 °C before imaging. Schlemm’s canal was imaged on a Nikon Yokogawa W1 CSU spinning disk confocal microscope running NIS Elements AR (v.5.42.03) software. Using a 20x objective and Large-Image Acquisition option, 30 μm Z-stack (1 μm/slice) images of Schlemm’s canal with 30% overlap were acquired and stitched together to generate a complete 360° representation.

OIR quantification

Retinal neovascular areas and vaso-obliterated areas were quantified as previously described^16,17,37. Briefly, neovascular and avascular areas were quantified using Adobe Photoshop (Adobe, San Jose, CA, USA). Neovascular area (NV) % was calculated by dividing the neovascular area by the total retinal area. Vaso-obliterated area (VO) % was calculated by dividing the vaso-obliterated area by the total retinal area. This analysis was performed by a masked investigator to minimize bias.

Schlemm’s canal morphology analysis

Quantitative image analysis was performed as previously described³⁶ with the following modifications. Anterior globes were imaged in four z-series acquisitions (for each petal of the flat mount) and stitched together to create a 360° view of each anterior globe immunostained with CD31. Using Nikon NIS-Elements (v.5.20.02), individual eye petals were analyzed separately, utilizing the Region of Interest tool. Z-stacks encompassing Schlemm’s canal were identified, typically consisting of 10 μm depth, and used to generate maximum intensity projections to define the 3D morphology of the canal. The Polygon Area tool within the Manual Measurement module was used to manually trace the outer edges of Schlemm’s canal and quantify its area as defined by CD31+ staining per petal. The total area of Schlemm’s canal for a given eye was determined by summing the four area measurements from each petal. Holes inside Schlemm’s canal were not quantified, as preliminary analyses indicated no significant differences among measured areas when holes were included versus excluded. All analyses were performed by a masked investigator.

Statistical analysis

All data are presented as mean ± SEM unless otherwise noted; n is listed and defined in figure legends. Values of p < 0.05 were considered statistically significant. All statistical analyses were performed using either GraphPad Prism 10 software (GraphPad, San Diego, CA, USA) or R (v4.4.0) with packages lme4 (v35.3) and emmeans (v1.10.7) for Schlemm’s canal area analysis. Statistical analyses used are defined in figure legends. Briefly, one-way ANOVA with Tukey’s post hoc tests was used to assess changes in neovascular and vaso-obliterated areas among OIR treated mice. Two-way ANOVA with Tukey’s post hoc test was used to assess differences in weights and IOPs among OIR treated (no treatment, sham, anti-VEGF) and non-OIR treated (no treatment, sham, anti-VEGF) groups. To evaluate effects on Schlemm’s canal morphology, an aligned rank transform (ART) ANOVA was used to evaluate the effects of treatment (normoxia vs. OIR) and injection (non, anti-VEGF, sham) on retinal area measurements. The ART model was implemented to enable nonparametric factorial inference with interactions while maintaining robustness to non-normality and small sample sizes. Post hoc pairwise comparisons were conducted using estimated marginal means with Tukey multiple comparisons adjustment. The inclusion of a random intercept for subject in a linear-mixed-effects model was justified after assessing correlation of within-subject measurements. The model included fixed effects for treatment (normoxia vs. OIR), injection (no injection, anti-VEGF, sham), and their interaction. Denominator degrees of freedom for the fixed effects were estimated using the Kenward-Roger approximation to improve small-sample inference. Custom contrasts of marginal measures were specified to test planned pairwise comparisons between groups.

Author contributions

Conceptualization – TLY, TWC, GDH, SWT; Methodology – TLY, TWC, GDH, SWT, SMM; Validation – GDH, AA, SMM, DW, SWT, KS, KNW, AM; Formal analysis – AA, SWT, SMM, KNW; Investigation – GDH, AA, DQ, AM, SWT, SMM; Visualization – GDH, AA, TWC, SWT, SMM, TLY; Supervision – TWC, TLY, SWT; Funding acquisition – TWC, TLY; Writing, original draft – GDH, TWC, TLY, SWT; Writing, review and editing – GDH, AA, SMM, DQ, SWT, KS, KNW, AM, TWC, TLY.

Funding

This study was supported by the Margaret Emma Williams Trust Fund to the University of Wisconsin-Madison Department of Ophthalmology and Visual Sciences, an Unrestricted Grant from Research to Prevent Blindness, Inc. to the University of Wisconsin-Madison Department of Ophthalmology and Visual Sciences, a Core Grant for Vision Research from the National Institutes for Health (NIH) to the University of Wisconsin-Madison (P30 EY016665) (to SWT, SMM, KNW and TLY), and a National Eye Institute (NEI) R21 EY034251 (to SWT). This work was also funded by the University of Wisconsin-Madison School of Medicine and Public Health Centennial Scholars Fund (to TLY). This work was also funded by F31 EY035171 (to GDH), R01 EY025641, the Canada Foundation for Innovation, and NSERC RGPIN-2025-04563 (all to TWC).

Data availability

The datasets generated during this current study are available from the corresponding authors on request.

Declarations

Competing interests

The authors declare no competing interests.