Zhou 2026 — The gut-retina axis: dysbiosis → microglial priming → retinal inflammation

One-paragraph summary

Mechanistic review consolidating the emerging literature on the gut-retina axis in age-related macular degeneration (AMD), describing the chain: gastrointestinal dysbiosis → imbalanced microbial diversity → depletion of protective metabolites (short-chain fatty acids, bile acids) → microglial priming → inflammasome activation → pro-inflammatory + pro-angiogenic cytokine secretion (IL-6, IL-1β, TNF-α, VEGF) → retinal inflammation + neovascularization. The review identifies Mediterranean and high-fiber diets, fecal microbiota transplantation (FMT) in animal models, and drug repurposing as candidate therapeutic strategies for modulating disease severity through this axis. For the FM project, this paper extends the B2 bridge (gut-brain-axis-disruption → glia → inflammation) directly to retinal tissue using identical mechanistic elements — the same SCFAs and bile acids the project's existing B2 chain invokes (Jakobsson 2026, Kishore 2026), the same microglial priming mechanism (extended from brain microglia to retinal microglia), the same downstream cytokine output (IL-6, IL-1β, TNF-α already in the project ontology), and the same FMT interventional candidate already in the project. The project's gut-brain framework therefore extends cleanly to gut-CNS (encompassing both brain and retinal microglia) without requiring a new mechanistic framework — only the explicit recognition of retinal microglia as a substrate.

Claims as triples

• gut_microbiome_dysbiosis — causes → retinal_inflammation [evidence: review of multiple primary studies; confidence: emerging in AMD context, bridging for FM]
• short_chain_fatty_acid — modulates → retinal_microglia [evidence: microbial-metabolite → microglial-priming chain; confidence: bridging]
• bile_acid — modulates → retinal_microglia [parallel chain to SCFAs; confidence: bridging]
• retinal_microglia — present_in → retina [extends existing project microglia entity to retinal substrate; confidence: established (in retinal-biology literature)]
• retinal_microglia — causes → retinal_inflammation [via inflammasome activation + cytokine release; confidence: emerging]
• fmt — modulates → retinal_inflammation [animal-model evidence cited in review; confidence: bridging]
• mediterranean_diet — modulates → gut_microbiome_dysbiosis [established in dietary-microbiome literature, extends to retina here; confidence: bridging]
• gut_brain_axis_disruption — bridges → retinal_inflammation [the B2 bridge extension claim; confidence: bridging]

Methods note

Narrative review synthesizing primary studies on gut-microbiome composition in AMD (16S rRNA + metabolomics analyses across multiple cohorts), animal-model FMT experiments demonstrating causality, and dietary intervention studies (Mediterranean and high-fiber diets, AMD progression). No new primary data. From a US ophthalmology group (Skondra lab); positions the gut-retina axis as a translational target.

Limitations

• AMD is an age-related neovascular/atrophic condition; FM is a chronic-pain syndrome. Direct mechanistic continuity between the two requires that the same gut-microglia-inflammation chain operates with FM-relevant triggers in FM-relevant retinal tissue. This is supported by parallel project anchors (Tian 2026 gut-brain-microglia chain in AD; Jakobsson 2026 bile-acid → anti-SGC-IgG in FM; Kishore 2026 gut microbiome scoping review in FM) but is not directly demonstrated in FM patients.
• Review-tier, not primary; the underlying primary literature is heterogeneous and includes substantial animal-model component.
• The cytokine output (IL-6, IL-1β, TNF-α, VEGF) overlaps with non-specific systemic inflammation; the retina-specific component versus systemic-inflammation-spillover-to-retina is not always cleanly separated in the primary literature.

Open questions raised

• Does the gut-retina chain operate in FM patients? Specifically, do FM patients with elevated zonulin / LPS / sCD14 (Dorta-Aguilar 2023 markers, already in project) show preferentially altered retinal inflammatory signal on imaging?
• Does FMT in nociplastic-pain conditions (Martín Pérez 2024, already in project) affect retinal microglial state? (No published data; tractable as secondary analysis in BASIS-FM-style trials.)
• Is the retinal microglial phenotype in FM concordant with the TSPO-PET-positive brain microglial signal (Coluzzi 2025)? Co-imaging would resolve whether retinal microglia state tracks brain microglia state in the same patient.
• What is the cleanest retinal-microglia biomarker for clinical deployment? (Plasma cf-mtDNA + ISG signature already in project as systemic readout; retinal-imaging-specific readout for retinal microglia is at the frontier.)

Triangulation notes

• Extends the B2 bridge (gut-brain-axis-disruption → glia → inflammation → FM symptoms) from brain to retinal microglia. The same mechanism, the same metabolite intermediates (SCFA, bile acid), the same effector cell type (microglia, lineage-shared between brain and retina), the same cytokine output. Bridges B2 to a retinal readout substrate.
• Strengthens Tian 2026 (gut → microglia → TREM2/SYK/NF-κB) at a different tissue site. The Tian mechanism was demonstrated in brain microglia in an AD model; Zhou 2026 supplies the parallel evidence at retinal microglia in AMD. Two independent literatures point to the same gut-microglia chain at different tissue substrates.
• Connects to Kishore 2026 + Jakobsson 2026 + Dorta-Aguilar 2023 (the project's existing gut-brain-axis cluster): these papers established the gut chain in FM at the brain-microglia level; Zhou 2026 supplies the retina-microglia extension. Retinal imaging therefore becomes a candidate readout for the same gut-brain framework the project has been working with.
• Aligns with I-5 calcium intersection (calcium-dysregulation-hypothesis.md v0.2): retinal microglia, like brain microglia, signal through P2X7 (ATP-gated Ca²⁺ entry) and SOCE. The same drug class (P2X7 antagonists, STING inhibitors) should affect both populations.
• Connects to BASIS-FM trial design: colesevelam (bile acid sequestrant) is the lead intervention in the BASIS-FM stage 1 design. Zhou 2026 supports the broader prediction that bile-acid-modulating interventions affect microglial state across CNS sites including retina.

Bridges

• B-Re-3 (Gut microbiome ↔ retinal inflammation extension of B2). Endpoints: gut dysbiosis (already FM-anchored via Dorta-Aguilar 2023, Jakobsson 2026, Kishore 2026); retinal inflammation (AMD-anchored here; FM-bridging). Confidence: bridging. Closing this bridge: retinal-imaging assessment in FM patients with measured intestinal-permeability markers; or FMT-treatment retinal-imaging change in nociplastic-pain patients.

Cure-path implications

The gut-retina axis adds a retinal-imaging trial endpoint to the project's existing microbiome-modulating cure-path arms (FMT, colesevelam, SCFA supplementation, prebiotic strategies). The microbiome-targeting interventions in BASIS-FM stage 1 could include retinal imaging as a secondary outcome at low marginal cost. If positive, this would close part of B2 fully and supply a non-invasive readout for downstream microbiome-modulating clinical trials.