New Feedback Loop Between Lactate and RNA Methylation Drives Macular Degeneration
Scientists uncover how ALKBH3, an RNA demethylase, hijacks retinal metabolism in AMD—and how blocking it may restore vision.
Summary
Researchers identified a self-amplifying molecular loop driving age-related macular degeneration (AMD). In diseased retinal pigment epithelium (RPE), the RNA demethylase ALKBH3 is overexpressed, stripping m1A methyl marks from HK2 and VEGFA mRNAs, boosting their stability and ramping up glycolysis and angiogenesis. Excess lactate from accelerated glycolysis lactylates histone H3K18, which then transcriptionally amplifies ALKBH3 expression—completing a vicious positive feedback cycle. Breaking this loop with the ALKBH3 inhibitor HUHS015 reduced RPE degeneration and, combined with anti-VEGF drug Aflibercept, synergistically suppressed choroidal neovascularization (CNV) in mouse models. The findings reframe AMD as a metabolic-epigenetic disease and open new therapeutic avenues.
Detailed Summary
Age-related macular degeneration (AMD) is the leading cause of irreversible blindness in older adults, yet no treatment effectively targets the retinal pigment epithelium (RPE) degeneration that initiates the disease. Approximately 10% of AMD patients progress to wet AMD, characterized by choroidal neovascularization (CNV) and rapid vision loss. Current anti-VEGF therapies slow CNV but do not address the upstream RPE dysfunction, leaving a critical unmet need.
This study investigated whether aberrant N1-methyladenosine (m1A) RNA modification contributes to AMD pathology. Using single-cell RNA sequencing data from wet AMD patients and age-matched controls, the team identified ALKBH3—an m1A eraser enzyme—as uniquely and progressively upregulated in RPE cells along the pseudotemporal trajectory of AMD progression. Elevated ALKBH3 was also confirmed in fetal RPE cells under hypoxia (1% O2 or CoCl2 treatment), in laser-induced CNV mouse models, and in aged mice, aligning with globally reduced m1A levels across all AMD models.
Mechanistically, ALKBH3 demethylates m1A sites on HK2 (hexokinase 2, the rate-limiting glycolytic enzyme) and VEGFA mRNAs, preventing their recognition and degradation by the m1A reader YTHDF2. This stabilizes both transcripts, amplifying glycolysis in RPE and increasing VEGFA secretion. The dm1ACRISPR system—a targeted RNA demethylation tool—confirmed that site-specific m1A removal from HK2 and VEGFA was sufficient to phenocopy ALKBH3 overexpression. Excess lactate produced by hyperactive glycolysis promotes histone lactylation specifically at H3K18 (H3K18la), and chromatin immunoprecipitation assays showed that H3K18la directly occupies the ALKBH3 promoter to enhance its transcription—closing a positive feedback loop. Alkbh3 overexpression in mouse RPE caused visual impairment, RPE structural anomalies, and CNV, while Alkbh3 knockout suppressed RPE glycolysis and CNV formation.
Therapeutically, the small-molecule ALKBH3 inhibitor HUHS015 disrupted the H3K18la–ALKBH3–HK2/VEGFA feedback loop, mitigating hypoxia-induced RPE degeneration in vitro and reducing CNV in vivo. Strikingly, HUHS015 combined with Aflibercept (an anti-VEGF agent) produced synergistic suppression of CNV, suggesting complementary mechanisms of action—HUHS015 targeting upstream metabolic dysregulation while Aflibercept blocks downstream angiogenic signaling.
These findings reframe AMD pathogenesis as a metabolic-epigenetic disorder, where RNA modification, glycolytic reprogramming, and histone lactylation co-conspire to drive disease. The H3K18la–ALKBH3–HK2/VEGFA network represents a promising multi-node therapeutic target, and the synergy between HUHS015 and Aflibercept suggests a combination strategy that could outperform current monotherapy approaches for wet AMD.
Key Findings
- ALKBH3 is the only m1A regulator significantly upregulated in RPE cells of wet AMD patients versus controls.
- ALKBH3 stabilizes HK2 and VEGFA mRNAs by removing m1A marks, boosting glycolysis and choroidal neovascularization.
- Excess lactate from heightened glycolysis causes H3K18 histone lactylation, which transcriptionally amplifies ALKBH3—forming a positive feedback loop.
- ALKBH3 inhibitor HUHS015 reduces RPE degeneration and synergizes with Aflibercept to suppress CNV in mice.
- Alkbh3 knockout mice show suppressed RPE glycolysis and reduced CNV, confirming its causal role.
Methodology
The study combined single-cell RNA sequencing of human AMD clinical samples (GSE135922, GSE203499), in vitro fetal RPE hypoxia models, laser-induced CNV mouse models, Alkbh3 knockout and overexpression mice, and the dm1ACRISPR targeted demethylation system. Mechanistic validation used chromatin immunoprecipitation, RIP assays, YTHDF2 reader studies, metabolomics, and pharmacological inhibition with HUHS015 alone and in combination with Aflibercept.
Study Limitations
The clinical single-cell RNA sequencing analysis included only two wet AMD patients and two controls, limiting statistical power and the ability to exclude confounders such as sex. Results rely heavily on animal and cell-culture models; human validation of the feedback loop and therapeutic efficacy requires larger clinical studies. The long-term safety and ocular bioavailability of HUHS015 have not yet been assessed in humans.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.