H5 — Autoimmune-Genetic-Risk Chain (candidate, subset hypothesis)

Status: project synthesis document, version 0.1, 2026-05-24 Drawn from: ontology entities staged 2026-05-24 (H5 module — autoimmune_genetic_risk_module, 9 gene entities, celiac_spectrum_disease, regulatory_T_cell_dysfunction); bridges B-Gen-1 and B-Gen-2; open questions Q-Gen-1..9; Kerrebijn 2025 GWAS re-examination 2026-05-24; existing H1 chain evidence (Goebel 2021, Krock 2023, Seefried 2025, Hanani 2026, Sanchez 2025, Jakobsson 2026, Dorta-Aguilar 2023). Purpose: define H5 as a candidate subset FM mechanism distinct from H1/H2/H3/H4. H5 is anchored on the recognition that a specific 9-gene set — HLA-DQ2/DQ8 + CTLA4 + PTPN22 + IL2/IL21 + IL2RA + TNF + MYO9B + CLDN2 + TGM2 — is the canonical celiac-disease + T1DM autoimmune-genetic signature and, applied to FM, defines a genetic-risk-stratifiable subgroup that maps cleanly onto the project's existing H1 chain at the upstream end.

TL;DR

H5 says: within the anti-SGC-IgG-positive 37% of FM (Seefried 2025), there exists a genetically-defined subgroup whose H1 chain operates with a celiac-spectrum-like upstream architecture — HLA-DQ2/DQ8-presented antigens drive autoreactive T-cell responses, tolerance-failure variants (CTLA4 + PTPN22) let autoreactive B-cells escape, T-cell-regulation variants (IL2/IL21 + IL2RA) drive germinal-center hyperactivity, and barrier-failure variants (MYO9B + CLDN2) provide the persistent antigen-flux substrate. The output is the same FM-IgG → MRGPRX2 → mast-cell-mediated effector phenotype already established by H1. The framework adds (a) a genetic-risk stratifier for trial design, (b) FDA-approved drug repurposing candidates with established safety profiles, and (c) a celiac-spectrum cross-condition anchor that is mechanistically — not just statistically — justified.

Confidence at chain level: bridging. No FM-direct evidence at any of the 9 loci yet. The 2026-05-24 Kerrebijn 2025 re-examination confirms none of the 9 genes appear in the published top-26 FM GWAS loci, but the partitioned-heritability analysis is exclusively brain-tissue — interpretable as subset-mechanism dilution at biobank scale rather than refutation. The right test instrument is case-only PRS within FM-IgG-positive cohorts, not biobank-scale GWAS.

What changed from H1 to H5

H5 does not replace H1. It supplies a genetic-substrate layer for the subset of H1 patients whose autoimmune cascade operates through the celiac/T1DM-genetic-architecture pathway. Three substantive additions:

1. Genetic-risk stratification for the existing H1 cohort. H1 currently has no genetic stratifier — the anti-SGC-IgG-positive 37% is defined serologically. H5 proposes that within that 37%, the high-genetic-risk module carriers (HLA-DQ2/DQ8 + tolerance triad + barrier-failure variants) define a phenotype-sharper subgroup that responds preferentially to gut-targeted and T-cell-targeted interventions.

2. Mechanistic upstream for Jakobsson 2026 (24 microbial bile acids ↔ anti-SGC IgG) and Dorta-Aguilar 2023 (gut permeability + zonulin/LPS/sCD14 elevation). The H1 chain currently treats barrier failure as acquired-only (driven by dysbiosis). H5 supplies the genetic substrate (MYO9B + CLDN2 variants) that explains why some patients develop the barrier failure and others don't given the same dysbiotic insult.

3. Cure-path-arm expansion via drug repurposing. Each gene class in the H5 module maps to an existing FDA-approved drug or established dietary intervention: abatacept (CTLA4), low-dose IL-2 (IL2RA / Treg restoration), anti-TNF biologics (TNF), larazotide acetate (CLDN2 / barrier), gluten-free dietary trial (HLA-DQ2/DQ8). Five candidate H1 cure-path-arm extensions where currently we have three.

The H5 chain (vertical, top-to-bottom)

``` LEVEL 1 — GENETIC RISK SUBSTRATE (heritable, present from birth) ├── HLA-DQ2 / HLA-DQ8 — class II haplotypes; gluten-peptide presentation ├── CTLA4 — T-cell coinhibitory checkpoint; tolerance failure ├── PTPN22 R620W — TCR/BCR signaling threshold dysregulation ├── IL2 / IL21 (4q27 locus) — Treg maintenance + Tfh cytokine ├── IL2RA (CD25) — high-affinity IL-2 receptor on Tregs ├── TNF promoter — pro-inflammatory cytokine arm shared with RA ├── MYO9B — Rho-GAP regulating epithelial barrier integrity ├── CLDN2 — tight-junction claudin (paracellular cation pore) └── TGM2 — tissue transglutaminase, celiac autoantigen

LEVEL 2 — TRIGGER INTERSECTION (environmental hit on genetic background) ├── gluten exposure (HLA-DQ2/DQ8 carriers, lifelong) ├── gut dysbiosis (existing project trigger) ├── viral infection (existing project trigger) └── stress / trauma (existing project triggers)

LEVEL 3 — PROCESS LAYERS (where the genetics get expressed) ├── Barrier failure ← MYO9B + CLDN2 variants ├── Tolerance failure ← HLA-DQ2/DQ8 + CTLA4 + PTPN22 └── Treg incompetence ← IL2RA + CTLA4 + IL2/IL21

LEVEL 4 — DOWNSTREAM EFFECTORS (where H5 merges into H1) ├── Bile acid translocation (Jakobsson 2026 — anchor) ├── Antigen flux to peripheral lymphoid tissue ├── Germinal-center hyperactivity (IL-21-driven, Talwar 2025 LC framework) ├── Plasma-cell expansion └── FM-IgG + anti-SGC IgG production (Goebel 2021, Seefried 2025)

LEVEL 5 — CONVERGENCE WITH H1 (Sanchez 2025 mechanism) ├── FM-IgG → MRGPRX2 → mast-cell degranulation ├── MC mediators (IL-6, histamine, tryptase, TNF-α, chymase) ├── Satellite glia + sensory neuron sensitization (Hanani 2026 multi-ganglia) └── Central sensitization machinery (existing project layer)

LEVEL 6 — SYMPTOMS ├── Widespread pain └── Multimodal hypersensitivity ```

The architectural claim of H5: levels 1–4 are H5-specific (the genetic upstream); levels 5–6 are shared with H1 (the existing chain). H5 is therefore not a parallel chain — it is the genetic-substrate-stratified upstream of H1.

The 9 genes — what each contributes

Tier 1 — Autoimmune-tolerance triad

HLA-DQ2 / DQ8 — MHC class II haplotypes conferring ~95% of celiac genetic risk and the strongest T1DM risk haplotype. Present deamidated gluten-derived peptides (deamidation by TGM2) to autoreactive CD4+ T cells. In the H5 framework, HLA-DQ2/DQ8 carriage predicts enrichment within FM-IgG-positive patients. Cheapest immediate test: saliva-based HLA-DQ2/DQ8 typing ~$50–100/sample (Q-Gen-1). FM cohort carriage frequency is uninvestigated in the project corpus.

CTLA4 — T-cell coinhibitory checkpoint maintaining peripheral tolerance. CTLA4 risk variants impair Treg function and permit autoreactive T-cell activation → B-cell help → autoantibody production. Druggable via abatacept (CTLA-4-Ig fusion protein, FDA-approved for RA) — adds a T-cell-tolerance intervention layer to the H1 cure-path distinct from CD38 plasma-cell depletion (daratumumab) and FcRn blockade (Q-Gen-6).

PTPN22 — strongest non-HLA autoimmunity risk gene. The R620W variant (rs2476601) alters TCR/BCR signaling thresholds so autoreactive lymphocytes escape tolerance. No direct drug target; functions as a genetic-risk stratifier. PTPN22 + HLA-DQ2/DQ8 + CTLA4 together define the high-tolerance-failure subset.

Tier 2 — T-cell regulation and the cytokine arm

IL2 / IL21 locus (chr4q27) — adjacent genes encoding IL-2 (T-cell growth factor, Treg maintenance) and IL-21 (T-follicular-helper cytokine, drives germinal-center B-cell help and IgG class switching). The IL-21 arm is directly upstream of the B-cell / plasma-cell activation step already in the H1×H2 graphic — IL-21 drives the germinal-center hyperactivity producing FM-IgG. Low-dose IL-2 therapy, trialed in lupus and T1DM to selectively expand Tregs, is a candidate H5 intervention.

IL2RA (CD25) — high-affinity IL-2 receptor alpha chain, highly expressed on Tregs. Risk variants associated with T1DM, MS, alopecia areata, Graves disease. The MS cross-condition connection is meaningful — the project's existing HERV-W ↔ FM/MS bridge gains a parallel Treg-dysfunction axis. Druggability via anti-CD25 mAb has a problematic history (daclizumab withdrawn due to autoimmune hepatitis); low-dose IL-2 is the safer Treg-restoration route.

TNF — TNF-α already lives in the project's ontology as a mast-cell mediator and a microglial cytokine. The H5 contribution is at the genetic level: TNF promoter polymorphisms (e.g., -308A) stratify response to anti-TNF biologics (etanercept, infliximab, adalimumab — all FDA-approved RA/IBD/AS) — adds a cytokine-targeted intervention layer to the H5 cure-path.

Tier 3 — Gut-barrier and the celiac-autoantigen axis

MYO9B — Rho-GAP regulating epithelial barrier integrity. Originally a celiac susceptibility gene. Variants → barrier dysfunction → bacterial and bile-acid translocation. Supplies a genetic substrate for the Dorta-Aguilar 2023 (gut permeability in FM/ME-CFS) and Jakobsson 2026 (24 microbial bile acids ↔ anti-SGC IgG) findings already in the project. The H1 chain treats barrier failure as acquired; H5 adds the heritable predisposition.

CLDN2 (claudin-2) — tight-junction protein forming paracellular cation pores. CLDN2 upregulation is documented in IBD; polymorphisms are IBD risk factors. Druggable via larazotide acetate — a tight-junction-restoring peptide in celiac clinical trials; candidate H1 upstream intervention alongside FMT and colesevelam (Q-Gen-7).

TGM2 — tissue transglutaminase 2; the celiac autoantigen (anti-tTG IgA is the diagnostic biomarker). Catalytic function: deamidation of gluten peptides increases their HLA-DQ2/DQ8-binding affinity. H5 testable prediction: anti-tTG titers in the FM-IgG-positive subset would either reveal genuine celiac comorbidity or document molecular mimicry between gluten-derived peptides and FM autoantigens. Cheapest H5 stratification screen (~$50–100/sample commercial serology, Q-Gen-3).

Intervention break-points

Six H5-specific cure-path-arm candidates, ordered roughly from upstream (genetic-stratification → dietary intervention) to downstream (cytokine and same-as-H1 effector interruption):

Note: H5 patients also remain candidates for the existing H1 cure-path-arm interventions — daratumumab (CD38 plasma cell depletion), FcRn blockade (efgartigimod), MRGPRX2 antagonism (barzolvolimab), and cromolyn (MC stabilization). H5 adds five upstream-tier interventions where currently we have one (FMT/colesevelam at the dysbiosis layer).

The Kerrebijn 2025 sharpening result

A 2026-05-24 re-examination of the Kerrebijn 2025 FM GWAS (2.5M individuals; 54,629 FM cases) was performed against the 9-gene H5 module. None of the 9 genes appear in the published top-26 FM-GWAS loci, which are all neural: HTT (Huntington's gene), GPR52, CAMKV, DCC, DRD2/NCAM1, MDGA2, CELF4. Stratified-LDSC partitioned-heritability is exclusively brain-tissue-enriched with no immune-tissue enrichment.

This is a sharpening result, not a refutation. The same paper has a self-flagged caveat: subset mechanisms operating in ~37% of FM (Seefried 2025 anti-SGC IgG positive) are systematically diluted to invisibility in biobank-scale top-loci analysis. The H1, H2, and H5 subsets all fall in this category — none of them surface in the Kerrebijn top hits.

The corollary is methodological: the correct test instrument for H5 is not biobank-scale GWAS but case-only PRS within FM-IgG-positive cohorts. A celiac+T1DM polygenic risk score computed on existing FM cohorts (Q-Gen-2) would partially close B-Gen-1 without requiring new genotyping. Q-Gen-9 captures the lower-cost variant: check the Kerrebijn supplementary tables for sub-threshold (sub-genome-wide-significant) signal at the H5 loci.

H5 vs H4 — overlap or orthogonal?

The H4 chain (immuno-metabolic, built 2026-05-24) and H5 (autoimmune-genetic-risk) both propose subset FM mechanisms with peripheral upstream sources. The H5-direction-decision question (Q-Gen-8) is whether they identify the same patients, mostly the same, or orthogonal subgroups:

• H4 enrichment: FM patients with metabolic-syndrome / obesity / T2DM comorbidity; expected to show elevated peripheral leptin, elevated CSF leptin, microglial-activation signature on PET-TSPO.
• H5 enrichment: FM patients with celiac-spectrum / T1DM comorbidity OR familial autoimmunity load; expected to carry HLA-DQ2/DQ8 + tolerance-failure variants; FM-IgG-positive subset enrichment.

The hypothesis: H4 and H5 are largely orthogonal subgroups sharing only the final-common-pathway microglial-activation step. A stratified analysis on a sufficiently-large FM cohort with both metabolic-syndrome status and HLA-DQ2/DQ8 typing would close Q-Gen-8 directly.

Open questions that close H5

The full Q-Gen block (1–9) is in synthesis/open-questions.md. The priority-ordered subset:

1. Q-Gen-1 — HLA-DQ2/DQ8 carriage frequency in FM-IgG-positive vs FM-IgG-negative cohorts. Cheapest test (~$50–100/sample); predicts enrichment in IgG-positive subset; would partially close B-Gen-1. 2. Q-Gen-3 — anti-tTG IgA + anti-DGP IgA seroprevalence in FM cohorts. Same price band; commercially available serology; tests for genuine celiac comorbidity or molecular mimicry. 3. Q-Gen-2 — celiac+T1DM PRS on FM-IgG-positive samples vs FM-IgG-negative samples. Computable retrospectively on any existing genotyped FM cohort; bypasses Kerrebijn-scale dilution. 4. Q-Gen-6 — abatacept pilot in CTLA4-risk-carrier FM-IgG-positive patients. Highest cure-path-arm impact; FDA-approved RA biologic. 5. Q-Gen-5 — 6-week gluten-free trial in HLA-DQ2/DQ8-positive FM patients. Cheapest interventional study; no IND barrier. 6. Q-Gen-8 — H5 vs H4 cohort overlap. Determines whether H4 and H5 should be modeled as one mechanism stratified two ways or as two distinct mechanisms.

Cross-condition anchors

H5 establishes two new mechanistically-justified (not just statistical) cross-condition links:

Celiac-spectrum disease (B-Gen-2). Shared genetic architecture (HLA-DQ2/DQ8 + MYO9B + CLDN2 + TGM2) + shared barrier-failure mechanism + shared gut-derived antigen-flux substrate. The deferred candidate 2025-s-non-coeliac-wheat-sensitivity-symptoms in the queue becomes worth promoting under the H5 direction. FM-celiac comorbidity prevalence in the project corpus is uninvestigated — the cheapest near-term gap to close.

T1DM (existing project anchor — Cano-Cano 2026 multimodal-biomarker template). The H5 framework re-justifies the T1DM cross-condition anchor as mechanistically — not just statistically — defensible. The HLA-DQ2/DQ8 + CTLA4 + PTPN22 + IL2RA risk-factor overlap between T1DM and the H5 FM module makes the Cano-Cano multimodal biomarker panel (sNfL + GFAP + RNFL + thalamic volumetry) transferable on stronger grounds than statistical-template borrowing alone.

Confidence-tier framing

No claim above bridging tier is asserted; no FM-direct evidence anchors are populated.

What promotes H5 from bridging to emerging

A single positive finding at any of the following gates moves H5 from bridging to emerging:

1. HLA-DQ2/DQ8 carriage enrichment in FM-IgG-positive vs FM-IgG-negative cohorts (Q-Gen-1) — promotes B-Gen-1. 2. Anti-tTG IgA seropositivity above population baseline in FM-IgG-positive subset (Q-Gen-3) — promotes B-Gen-2. 3. Celiac+T1DM PRS elevation in FM-IgG-positive subset (Q-Gen-2) — promotes B-Gen-1. 4. Symptom-load reduction in HLA-DQ2/DQ8-positive FM patients on 6-week gluten-free dietary trial (Q-Gen-5) — promotes the H5 chain itself. 5. Symptom-load reduction in CTLA4-risk-allele-carrier FM-IgG-positive patients on abatacept pilot (Q-Gen-6) — promotes H5 + adds cure-path arm at emerging tier.

The lowest-cost gate is (1): a single retrospective HLA-DQ2/DQ8 typing study on existing FM cohort serum/saliva banks costs ~$5–10K for a 100-patient pilot. That is the natural starting point for any operational H5 program.

What would refute H5

H5 would be refuted (not merely sharpened) by:

1. HLA-DQ2/DQ8 carriage frequency in FM-IgG-positive patients indistinguishable from the general population, in a 200-patient or larger cohort. 2. Celiac+T1DM PRS distribution in FM-IgG-positive samples indistinguishable from FM-IgG-negative samples and from healthy controls, in a comparably-powered analysis. 3. Anti-tTG IgA seropositivity in FM-IgG-positive patients at population baseline rate. 4. Negative result on the gluten-free dietary trial (Q-Gen-5) in well-stratified HLA-DQ2/DQ8-positive FM-IgG-positive cohort.

Two of the four are cheap: HLA-DQ2/DQ8 typing (Q-Gen-1) and anti-tTG screening (Q-Gen-3). The framework is therefore disciplined — it makes near-term-falsifiable predictions, not deferred speculation.

Position in the project's chain inventory

H5 is structurally distinct from H4 (different upstream source, different subset enrichment) but structurally related to H1 (same downstream MRGPRX2 → MC effector arm). The natural cure-path integration is: stratify FM-IgG-positive patients on H5-module carriage, then assign upstream-arm interventions (gluten-free, larazotide, abatacept, low-dose IL-2) to high-carriers and downstream-arm interventions (FcRn, MRGPRX2 antagonism, cromolyn) to low-carriers.

Status

• Synthesis document: drafted 2026-05-24, version 0.1
• Ontology entities: staged at bridging tier (11 entities)
• Bridges: B-Gen-1, B-Gen-2 staged at bridging tier
• Open questions: Q-Gen-1..9 logged
• FM-direct evidence: none at any locus
• Direction-decision pending: whether to commit operational resources to H5 (Q-Gen-1 pilot, candidate-paper promotion, agent query expansion) vs. hold at this bridging-tier framework