Family 20: Inflammation Resolution and Lipid Mediators
Family overview. Inflammation is normally self-limiting — the transition from acute inflammation to resolution is an active, biochemically programmed process mediated by specialized pro-resolving mediators (SPMs): resolvins (E-series from EPA, D-series from DHA), protectins, maresins, and lipoxins (Serhan, Libreros, and Nshimiyimana 2022). Unlike anti-inflammatory agents that suppress immune responses, SPMs actively clear neutrophils, stimulate efferocytosis, enhance antimicrobial activity, and promote tissue repair (Serhan, Libreros, and Nshimiyimana 2022). Chronic inflammation results when resolution mechanisms fail — not simply because pro-inflammatory signals persist.
Concrete mechanisms and ME/CFS evidence:
SPM deficiency in post-viral syndromes. Engert et al. (2026) propose that sleep disturbance in Long COVID dysregulates SPM biosynthesis, producing a state of failed resolution rather than excessive inflammation (Engert et al. 2026). Resolvin levels have not been measured in ME/CFS plasma or CSF — this is an open gap (Engert et al. 2026) (Rauf, Naveed, and Asghar 2026).
Post-acute sequelae as resolution failure. Rauf et al. (2026) characterize PASC as a disorder of impaired innate immune resolution: residual viral antigens, host alarmins, and reactivated latent viruses sustain chronic TLR/RIG-I/NLR/cGAS-STING signaling, but the primary defect may be failure to resolve rather than failure to suppress (Rauf, Naveed, and Asghar 2026).
SPM-enriched supplementation. A 12-week open-label trial of SPM-enriched marine oil in Post-COVID Syndrome improved eicosanoid profiles and inflammatory resolution markers (Gracia Aznar et al. 2024). No such trial exists in ME/CFS.
Cholinergic anti-inflammatory pathway and SPMs. Vagal signaling via α7-nAChR stimulates resolvin production, linking autonomic dysfunction (Family 8) to resolution capacity. Reduced vagal tone in ME/CFS (reduced HRV) would predict impaired vagal-SPM coupling.
Dietary precursor dependence. SPM synthesis depends on omega-3 precursor availability (EPA for E-resolvins, DHA for D-resolvins/protectins/maresins) and aspirin-acetylated COX-2 for aspirin-triggered resolvins. Dysregulated fatty acid metabolism and altered lipid profiles in ME/CFS could impair precursor availability (Bao and Yao 2026).
Gut microbiota-SPM axis. Gut microbiota composition shapes the lipid mediator landscape; dysbiosis shifts profiles toward pro-inflammatory eicosanoids (PGE2, LTB4) at the expense of SPMs (Bao and Yao 2026). ME/CFS gut dysbiosis (Chapter Gastrointestinal and Microbiome Dysfunction) would therefore predict a pro-resolution deficit via this gut-immune axis.
Full discussion: hypothesis registry (H1: SPM deficiency hypothesis); cross-reference to autophagy and inflammation resolution in Family 13.
Evidence status: Theoretical for ME/CFS (no SPM measurements exist in ME/CFS cohorts); Probable for Long COVID (Engert 2026, Rauf 2026); Established for SPM biology generally (Serhan laboratory, 30+ years).
Lithium as a potential SPM biosynthesis enhancer. A separate line of evidence suggests lithium may upregulate SPM biosynthesis. Basselin et al. (2010) demonstrated that chronic lithium treatment in rats increases brain 17-HDHA levels 1.9-fold — 17-HDHA is the committed intermediate for D-series resolvins (RvD1-6) and protectins (PD1), generated by 15-lipoxygenase from DHA (Basselin et al. 2010). Critically, the 17-HDHA increase was observed only in LPS-stimulated rats; lithium alone did not increase 17-HDHA over control — the effect is lithium augmentation of a stress response, not lithium-initiated SPM production at baseline. Moreover, Basselin 2010 used chronic therapeutic lithium dosing (serum ~0.6–0.8 mM); relevance to ultralow-dose lithium (2 mg/day) in humans is unknown and the direction of effect may differ. This is the only study directly linking lithium to SPM precursor production at any dose. Toricelli et al. (2021) showed that microdose lithium (2—20 µM Li₂CO₃) shifts the cytokine profile toward pro-resolution: reduced IL-1α, IL-6, MIP-1β/CCL-4 and NF-κB, with increased IL-10 — a profile consistent with resolvin activity, though SPMs were not measured (Toricelli et al. 2021). Acute lithium also inhibits COX-2 expression and PGE₂ production in activated microglia, which could favor the COX-2 substrate switch from prostaglandin biosynthesis toward SPM biosynthesis under the right cofactor conditions, though the aspirin-mediated acetylation required for the “AT-resolvin” branch would not be replicated by lithium alone (Stachowicz 2023).
The connection has not been demonstrated at any dose — no study has directly measured resolvins, protectins, or maresins after lithium treatment, and no study has tested lithium’s modulation of SPM receptors (ChemR23, GPR32, ALX/FPR2). The complete chain from Li⁺ → 15-LOX → 17-HDHA → RvD1-6/PD1 → GPCR → M2/pro-resolving shift is assembled from separate literatures. Whether the 1.9-fold 17-HDHA increase observed in rats at chronic lithium doses translates to meaningful SPM increases at ultralow-dose lithium (2 mg/day) in humans is completely unknown. The consistent direction of the Toricelli cytokine shift and the Basselin precursor increase make this a testable hypothesis, not an established mechanism. Importantly, the Toricelli cytokine profile (IL-10↑, IL-6↓, IL-1α↓) is also consistent with lithium’s well-established GSK-3β inhibition, which suppresses NF-κB transcriptional activity through multiple mechanisms (increased IκB stability, decreased p65 transactivation, reduced β-catenin degradation) (Beurel and Jope 2014). This simpler, canonical pathway requires no novel NCS-1, InsP3R1, or SPM hypothesis — lithium inhibits GSK-3β at IC50 ~2 mM, which is plausibly relevant at the 2–20 µM concentrations used by Toricelli only if lithium accumulates intracellularly or if GSK-3β inhibition is indirect (e.g., via inositol depletion affecting PKC-Akt-GSK3 signaling). The SPM hypothesis and the GSK-3β/NF-κB hypothesis are not mutually exclusive, but the GSK-3β route has substantially more precedent and should be considered the parsimonious default. See also Gut-Microbiome Lipid Mediator Axis: SPM Precursor Conversion Deficiency for the gut-microbial axis of SPM synthesis, and Aspirin + Omega-3 for SPM Precursor Provision for the aspirin-triggered resolvin pathway.
Testable predictions. (1) In rats, 4-week lithium at therapeutically relevant serum concentrations (0.6—1.2 mM) increases brain RvD1 and PD1 by ≥1.5-fold vs vehicle (LC-MS/MS lipidomics). Falsified if neither rises. (2) In ME/CFS patients, 8-week ultralow-dose lithium (2 mg/day) raises plasma 17-HDHA by ≥1.3-fold and shifts the lipid mediator profile toward D-series resolvins (RvD1, RvD2) by targeted LC-MS/MS. Falsified if 17-HDHA does not increase or SPM:pro-inflammatory eicosanoid ratio does not increase. (3) Lithium’s cytokine shift (↓IL-1α, ↓IL-6, ↓MIP-1β; ↑IL-10, per Toricelli 2021) is mediated via SPM receptors: blocked by WRW4 (ChemR23 antagonist) or absent in GPR32-knockdown macrophages. Falsified if cytokine shift persists despite receptor blockade.
If ME/CFS patients have deficient SPM biosynthesis — whether from reduced omega-3 precursor availability, impaired enzymatic conversion, or disrupted vagal-SPM coupling — the failure of active resolution would convert every inflammatory trigger (infection, exertion, stress) from a self-limited episode into a sustained inflammatory event. This would explain the prolonged, disproportionate response to minor triggers that characterizes the illness, and it does not require ongoing antigenic stimulation. The prediction is testable: targeted lipidomics of SPMs (RvD1–6, RvE1–3, LXA4, MaR1, PD1) in ME/CFS plasma versus matched controls, ideally before and after a standardized CPET protocol. If SPM levels are low at baseline and fail to rise post-exertion (as they should physiologically), this would support resolution failure as a chronicity mechanism.
Certainty: 0.45. SPM biology is well-established; SPM measurement is technically feasible (LC-MS/MS lipidomics); no ME/CFS-specific data exist. The hypothesis synthesizes documented ME/CFS features (prolonged inflammation, post-exertional exacerbation, dysregulated lipid profiles, reduced vagal tone) under the resolution-failure umbrella.
Testable predictions: (a) Baseline SPM levels (RvD1, RvE1, LXA4) will be lower in ME/CFS than matched controls; (b) SPM levels will fail to rise appropriately 2–24h post-CPET in ME/CFS versus controls; (c) SPM deficiency severity will correlate with PEM duration and severity.
High-mobility group box 1 (HMGB1) is one of several danger-associated molecular patterns (DAMPs) that may bridge cellular stress to sustained neuroinflammation (Ibrahim, Wasim, and Rahman 2026). Nuclear HMGB1 translocates to the cytoplasm and is released extracellularly in response to cellular stress, where it signals via TLR4, TLR9, and RAGE receptors. Critically, HMGB1’s redox state determines its bioactivity: disulfide-HMGB1 acts as a pro-inflammatory cytokine, while fully reduced HMGB1 is chemotactic (Ibrahim, Wasim, and Rahman 2026). This redox switching provides a plausible mechanism for exertion-dependent inflammatory state changes in ME/CFS — exercise-induced oxidative stress could shift HMGB1 toward the pro-inflammatory disulfide form, triggering the PEM cascade. The cGAS-STING pathway (Family 4) and mtDNA release (Section Immunosenescence Evidence Remains Indirect) are better-evidenced DAMP mechanisms in ME/CFS; HMGB1 should be considered a complementary candidate rather than a central hub.
HMGB1 is particularly relevant in conditions involving both cellular stress and barrier disruption: the increased blood-brain barrier permeability documented in some ME/CFS patients would allow peripheral HMGB1 to access CNS TLR4/RAGE receptors on microglia and astrocytes. The HMGB1-pCTS-L (cathepsin L) complex described by Chen et al. provides an additional amplification mechanism where HMGB1 binding to cathepsin L sustains inflammatory signaling beyond free HMGB1 alone (Chen et al. 2026).
HMGB1 intersects with the cGAS-STING pathway (Family 4): HMGB1 binds immunogenic DNA and enhances cytosolic DNA sensing, providing a second route to type I interferon induction alongside mtDNA leakage. No study has measured extracellular HMGB1 or its redox isoforms in ME/CFS — this is a direct experimental gap.
Certainty: 0.55. HMGB1 biology is well-established in neuroinflammation; redox-dependent activity switching is documented; BBB disruption is documented in some ME/CFS patients. Direct ME/CFS evidence is absent. The hypothesis generates falsifiable predictions that require only plasma and CSF samples from existing cohorts.
Testable predictions: (a) Total extracellular HMGB1 will be elevated in ME/CFS plasma versus controls; (b) The disulfide:fully-reduced HMGB1 ratio will increase 2–24h post-exertion, correlating with PEM severity; (c) HMGB1 levels will correlate with TLR4 activation markers and IL-1β in CSF.
S100A8 and S100A9 (calprotectin) are Ca2+-binding DAMPs released by activated neutrophils, monocytes, and macrophages that signal through TLR4 and RAGE to amplify inflammation. Nunes et al. (2024) demonstrated significant S100-A9 upregulation in ME/CFS plasma proteomics (Nunes et al. 2024). Given that S100A8/A9 activates microglia via TLR4, S100A8/A9 released from peripheral immune cells — or from microglia themselves — could sustain the microglial activation state documented in ME/CFS PET studies.
Calprotectin has additional clinical utility: Hetland et al. showed that calprotectin levels correlate strongly with NET burden (\(r \ge 0.745\)) in vaccine-associated thrombotic complications (Hetland et al. 2022), establishing it as a clinically accessible NET proxy. In ME/CFS, serial calprotectin measurement before and after standardized exertion could serve as a low-cost index of exercise-induced innate immune activation, monitoring NET-mediated thromboinflammation and microglial priming simultaneously without specialized assays.
Certainty: 0.55. S100-A9 upregulation is directly documented in ME/CFS (Nunes 2024). S100A8/A9-TLR4-microglia axis is established in neuroinflammation. NET-correlation via calprotectin is validated. The specific microglial-priming link and exertion-dynamics are untested. A critical caveat: plasma S100A8/A9 elevation does not guarantee CNS penetration — whether calprotectin crosses the BBB in ME/CFS, or activates microglia indirectly via peripheral TLR4 signaling affecting vagal afferents, is unknown.
Testable predictions: (a) Plasma calprotectin will rise 6–24h post-CPET in ME/CFS, correlating with PEM severity; (b) Calprotectin levels will correlate with NET markers (MPO-DNA, H3cit) and complement activation products; (c) In patients with the largest post-exertional calprotectin rise, measured microglial activation (TSPO PET) will be highest.