Oltra 2023 — HERV expression segregates ME/CFS, FM, co-diagnosed, and healthy

One-paragraph summary

The single most important paper for the project's viral-genome-modification framework. The authors profiled HERV expression patterns across four cohorts: pre-pandemic ME/CFS, pre-pandemic fibromyalgia, patients meeting criteria for both conditions, healthy controls — plus a separate post-COVID condition cohort. Hierarchical clustering of HERV expression profiles allowed perfect participant assignment into four distinct groups, demonstrating disease-specific HERV "fingerprints." HERV-W ENV protein detection rates by cohort: 48.5% in post-COVID, 33.3% in FM, 25% in ME/CFS, 8.9% in healthy controls. The pre-pandemic FM cohort also showed elevated HERV-W ENV positivity, demonstrating that the mechanism predates SARS-CoV-2 — it is a general post-viral phenomenon, not a long-COVID artifact. The paper establishes HERV reactivation as a candidate upstream durable mechanism for chronic post-viral conditions, with HERV-W ENV detection in patient serum as a tractable biomarker. Implications: a candidate genetic-level explanation for why these conditions persist long after the triggering infection has cleared, and an entry point for anti-HERV-W ENV biologic therapy (temelimab in MS clinical development).

Claims as triples

• viral_infection — causes → HERV_reactivation [evidence: cross-cohort HERV expression profiling; confidence: emerging]
• HERV_reactivation — causes → HERV_W_ENV [evidence: detection of envelope protein product; confidence: emerging]
• HERV_W_ENV — modulates → neuroinflammation_glial [evidence: bridges from MS literature where HERV-W ENV TLR4 activation is established; confidence: emerging]
• serum_HERV_W_ENV_level — predicts → fm_central_only [evidence: 33.3% FM vs 8.9% HC; confidence: emerging]
• HERV_reactivation — bridges → fm_autoimmune [evidence: 33% prevalence suggestively similar to Seefried 2025 anti-SGC IgG 37%; overlap unmeasured; confidence: bridging]
• HERV_reactivation — bridges → post_covid_syndrome [evidence: 48.5% HERV-W ENV positive; confidence: emerging]
• HERV_reactivation — bridges → me_cfs [evidence: 25% HERV-W ENV positive; confidence: emerging]

Methods note

Cross-sectional observational. Pre-pandemic ME/CFS, FM, and co-diagnosed cohorts from Spanish patient population; separate post-COVID cohort. HERV expression by RT-qPCR or RNA-seq across multiple HERV families (HERV-W, HERV-K, others). HERV-W ENV protein detection by ELISA / Western blot. Hierarchical clustering on expression profiles. Authors: Elisa Oltra and Anna Mensa-Vilaró — established Spanish ME/CFS and HERV research groups.

Limitations

• Reviewed preprint at eLife (not yet final journal publication). Status: peer review completed; final formal publication pending.
• Cross-sectional — cannot establish that HERV reactivation precedes FM symptom onset (could be reverse causation if FM-related immune dysregulation reactivates HERVs as a downstream consequence).
• HERV-W ENV detection methodology varies across labs; replication using the same protocol in independent cohorts is needed before tier promotion.
• Sample sizes per cohort moderate; the perfect classification accuracy may not generalize to larger heterogeneous cohorts.
• Did not stratify by anti-SGC IgG status, MC activation markers, or other H1/H2 biomarkers — cannot directly test whether HERV-W positivity overlaps with the autoimmune or mast-cell subtypes (Q30).
• Did not include longitudinal data from before viral exposure → through chronic disease — would be the gold standard for causation.

Open questions raised

• Q30 (added 2026-05-08): Does HERV-W ENV positivity correlate with anti-SGC IgG positivity in the same FM patients? The 33% / 37% prevalence overlap is suggestive; co-measurement in a single cohort would directly test whether HERV reactivation is the upstream cause of the H1 chain.
• Q31: Does the HERV-W-positive FM subset respond to temelimab? Direct cure-path test for an HERV-defined FM subtype.
• Q33: Is post-COVID FM a primarily HERV-driven phenomenon (mechanism 1 in viral-genome-modification-hypothesis.md) or a SARS-CoV-2 RNA-integration phenomenon (mechanism 3)? HERV-W positivity rate (48.5%) is substantial but not 100%; the residual is consistent with mechanism 3 or with non-genome-modifying mechanisms.
• What proportion of the 9% HERV-W-positive healthy controls go on to develop FM, ME/CFS, or long COVID over 5–10 years? Would test predictive (pre-disease) vs reactive (post-disease) interpretations.
• Are there specific HERV-K (HML-2) subtypes that cluster with FM specifically vs ME/CFS specifically? The "different fingerprint per disease" finding suggests this and would refine subtype mapping.

Triangulation notes

• Anchors the viral-genome-modification framework introduced in synthesis/viral-genome-modification-hypothesis.md. This paper is the empirical cornerstone of that framework's H1-chain-relevant evidence.
• Reshapes the H1×H2 subtype matrix by introducing a candidate fourth axis (HERV-W ENV positivity). The matrix in causal-chain-hypotheses.md should be extended in v0.2 to a 2×2×2 cube once HERV-W co-measurement data are available.
• Provides a candidate upstream cause for both H1 and H2. HERV-W ENV is a TLR4 ligand → MC activation (H2) AND drives chronic immune activation that could underlie B-cell dysregulation (H1 via EBV-B-cell lineage). One mechanism, two downstream chains.
• Therapeutic axis added. Temelimab (anti-HERV-W ENV mAb) becomes a candidate cure-path biologic. The chain HERV_reactivation → HERV_W_ENV → mast_cell activation + neuroinflammation → FM phenotype is now testable end-to-end with a single agent's effect on the proximal-most actor.
• Cross-condition implications. The 48.5% / 33.3% / 25% / 8.9% gradient (long COVID > FM > ME/CFS > HC) supports the project's framing of these as related but distinct conditions on a shared substrate. The "co-diagnosed" intermediate group with its own HERV fingerprint is consistent with B6 (predictive-coding / interoceptive-inference failure) being the common downstream phenotype across all three.

Bridges

• B6 strengthened: post-infectious nociplastic conditions share HERV reactivation as candidate upstream substrate.
• NEW B7 candidate: HERV reactivation ↔ FM autoimmune subtype — the 33% / 37% prevalence overlap is the load-bearing observation. Promote to formal bridge in bridges.md when overlap is empirically measured.
• B1 (Long COVID ↔ FM): strengthened by shared HERV mechanism; not just shared SFN phenotype.
• B4 (ME/CFS ↔ FM): strengthened by HERV fingerprint similarity with FM-specific differences.