# Polyunsaturated Fatty Acid Balance Modulates Microglial State in a Murine Model of Oxygen-Induced Neovascularization

## TL;DR

This study shows that a balanced intake of omega-6 and omega-3 fatty acids can protect against retinal damage in a mouse model of retinopathy.

## Contribution

The study reveals that fat-1 mice with endogenous omega-3 enrichment show reduced neovascularization and microglial activation in oxygen-induced retinopathy.

## Key findings

- Fat-1 mice showed a 75.5% relative reduction in vaso-obliteration and a 71.8% relative reduction in neovascularization.
- Transcriptional responses to OIR were significantly dampened in fat-1 mice compared to wild-type mice.
- Microglial activation and inflammation were attenuated in fat-1 mice, with no upregulation of specific oxylipins.

## Abstract

Background/Objectives: The retina is enriched in polyunsaturated fatty acids (PUFAs) which are indispensable for normal vision, and recent clinical studies have shown that dietary supplementation of ω-6-and ω-3-polyunsaturated fatty acids (PUFAs) can provide a protective role against retinopathy of prematurity (ROP). Our study aims to understand the mechanisms by which altering ω-6-and ω-3-polyunsaturated fatty acids (PUFAs) in the eye can protect against pathologic retinal neovascularization (NV). Methods: We interrogated the effects of endogenous ω-3-PUFA enrichment using transgenic fat-1 mice which convert ω-6-PUFAs to ω-3-PUFAs in the oxygen-induced retinopathy (OIR) murine model. In the OIR model, mice are exposed to 75% oxygen from postnatal day 7 (P7) to P12, then returned to room air (RA). We used a combination of immunofluorescence, bulk retinal RNA sequencing, and lipid mediator profiling by UHPLC-MS/MS in P17 mouse retinas to identify mechanisms underlying the protective effect against NV seen in fat-1 mice exposed to OIR. Results:
Fat-1 OIR mice were protected against the development of retinopathy, demonstrating 15.1% less vaso-obliteration (75.5% relative reduction) after OIR and a 6.1% reduction in neovascularization (71.8% relative reduction) at P17 (p < 0.0001 for both). We found a dampened transcriptional response to OIR in the retina of fat-1 mice as compared to WT mouse retinas (198 vs. 782 genes, adjusted p-value < 0.01). Pathway analyses confirmed these findings, with significant OIR-induced transcriptional shifts in angiogenesis (adjusted p-value < 10−27), inflammation (adjusted p-value < 10−25), and microglial activation pathways (adjusted p-value < 10−9) in WT mouse retina that were not observed in fat-1 mice. Enrichment scores obtained through the integration of our bulk transcriptomics data with cell-resolved retina data indicate that the protective phenotype observed in fat-1 mice could be associated with intrinsic differences in microglia cell subtypes between WT and fat-1 mice. In situ, WT OIR mice demonstrated an increase in Iba1+ microglia compared to WT RA mice, whereas fat-1 OIR mice showed no difference when compared to fat-1 RA mice. Three ARA-derived oxylipins, 12-hydroxyeicosatetraenoic acid (12-HETE), prostaglandin D2 (PGD2), and thromboxane B2 (TXB2) demonstrated a pattern of upregulation in WT OIR compared to WT RA, but no upregulation in fat-1 OIR mice compared to fat-1 RA. Two EPA-derived specialized pro-resolving mediators and two LA-derived oxylipins were also differentially expressed. Conclusions: These findings show that a lower ω-6:ω-3 protects against neovascularization and is associated with attenuation of hyperoxia-induced microglial recruitment and activation, as well as inflammation and angiogenic signaling.