EvidenceHealthVaccinationCardiometabolic
CXCL10–IFN-γ axis and genistein in preclinical models of mRNA-vaccine-associated myocarditis
- Study type:
- Preclinical experimental work in mice and iPSC-derived human cells plus secondary analyses of human plasma proteomics; no randomised controlled trial in humans.
- Topic:
- Vaccine safety research
- Added:
- 2026-06-06
- Status:
- published
Strength of evidence
Very Low
Source
Authors (preprint/author manuscript). Inhibition of CXCL10 and IFN-γ ameliorates myocarditis in preclinical models of SARS-CoV-2 mRNA vaccination. Author manuscript submitted to Science Translational Medicine.
This preclinical work combines **mouse experiments**, **iPSC-derived human cell models** and **secondary analysis of plasma proteomics** to explore mechanisms of mRNA‑vaccine-associated myocarditis. The authors focus on the **CXCL10–IFN‑γ axis**, show that high-dose BNT162b2 vaccination in mice raises cardiac troponin and inflammatory gene expression, and report that blocking CXCL10/IFN‑γ or giving the phytoestrogen **genistein** reduces these laboratory markers in mice and cell models. However, all key outcomes are **soft surrogate endpoints** (troponin, cytokines, gene expression, proteomics, imaging) in **models**, not hard clinical endpoints in humans. There is **no randomised, controlled human trial**, no direct modulation of CXCL10/IFN‑γ in patients, and the mouse dosing/timing only partially reflect clinical scenarios. Under WorthDay Evidence Criteria, this study provides **very low-quality evidence**: it generates hypotheses about mechanisms, but cannot prove why young vaccinated people develop myocarditis, nor that genistein is a safe or effective preventive option.
Key findings
- **Design:**
- Re‑analysis of published **human plasma proteomics** data from vaccinated individuals and patients with post‑vaccine myocarditis.
- **iPSC-derived models:** Macrophages exposed to BNT162b2 or mRNA‑1273 → cytokine and gene-expression profiling. Cardiomyocytes and 3D cardiac spheroids exposed to CXCL10/IFN‑γ "cytokine cocktails" → contractility, arrhythmia-like events, NT‑proBNP, stress and inflammation markers.
- **Mouse experiments** (BALB/c, C57BL/6): Intramuscular BNT162b2 at 0.2, 2 or 5 µg on days 0 and 21. Measurements: cardiac troponin I (cTnI), cytokines, inflammatory gene expression, immune-cell infiltration, echocardiography. Interventions: neutralising antibodies against CXCL10 and/or IFN‑γ; treatment with **genistein**.
- **Main laboratory results:**
- mRNA-vaccine exposure increases CXCL10/IFN‑γ in human datasets and iPSC‑macrophages.
- High-dose (2–5 µg) BNT162b2 in mice raises cTnI and inflammatory gene expression in the heart, especially after the second dose; LV ejection fraction remains normal.
- CXCL10/IFN‑γ blockade in mice lowers cTnI, IFN‑γ-response genes and immune-cell infiltrates.
- Cytokine cocktails in iPSC‑cardiomyocytes reduce contractility, increase arrhythmia-like events and elevate NT‑proBNP.
- Genistein reduces inflammatory markers and surrogate injury markers in vitro and in mice, with preserved antibody responses.
- Authors interpret CXCL10–IFN‑γ signalling as a central driver of vaccine-associated myocardial injury and suggest a potential "class effect" for mRNA platforms and a possible protective role for genistein.
Background
- Reports of myocarditis temporally associated with COVID‑19 mRNA vaccination, particularly in young males, have led to a search for biological mechanisms. This author manuscript (submitted to Science Translational Medicine) investigates whether a **CXCL10–IFN‑γ cytokine axis** plays a central role in vaccine-associated myocardial injury and whether **genistein**, a phytoestrogen, might mitigate this effect.
Methods
- The work combines three main components:
- **1. Human data (secondary analyses):** Re‑analysis of published **plasma proteomics** from vaccinated individuals and from patients diagnosed with post-vaccination myocarditis. Focus on cytokines/chemokines, particularly CXCL10 and IFN‑γ.
- **2. iPSC-derived human cell models:** iPSC‑derived macrophages (iPSC‑MACs) exposed to mRNA vaccines (BNT162b2, mRNA‑1273) in vitro; measurement of cytokines and gene expression. iPSC‑derived cardiomyocytes (iPSC‑CMs) and 3D cardiac spheroids exposed to CXCL10/IFN‑γ "cocktails" intended to mimic cytokine levels in vaccinated individuals vs myocarditis patients. Outcomes: contractility, arrhythmia-like events (MEA), NT‑proBNP, inflammatory and stress-related gene expression. Additional experiments with endothelial cells and macrophages, with and without **genistein**.
- **3. Mouse models:** BALB/c and C57BL/6 mice receive intramuscular BNT162b2 at 0.2 µg (body‑surface area-adjusted human equivalent) or higher doses (2 and 5 µg) on days 0 and 21. Measurements at defined time points (e.g., 8 hours after second dose): serum cTnI, cytokines, cardiac gene expression, immune-cell infiltration (neutrophils, macrophages), left ventricular ejection fraction by echocardiography. Interventions: neutralising antibodies against CXCL10 and/or IFN‑γ, oral or injected **genistein**.
Results
- **Cytokines and proteomics:** Human proteomics and iPSC‑MAC data show increased **CXCL10 and IFN‑γ** after mRNA vaccine exposure.
- **Mouse myocarditis markers:** At 0.2 µg BNT162b2, mice show only a small, clinically "subtle" troponin rise. At **2–5 µg** (10–25× the human‑equivalent dose), there are larger increases in cTnI and inflammatory gene expression in the heart; however, LV ejection fraction remains within normal ranges. Neutralisation of CXCL10/IFN‑γ reduces cTnI, IFN‑γ‑responsive genes (e.g., Stat1, Ido1) and inflammatory cell infiltration.
- **iPSC‑cardiomyocyte injury model:** Cytokine cocktails representing "vaccinated" vs "myocarditis" conditions reduce contractility, increase arrhythmia-like events and raise NT‑proBNP and inflammatory markers in iPSC‑CMs and 3D spheroids.
- **Genistein:** In vitro, genistein dampens immunoproteasome activity and inflammatory markers in cardiomyocytes, endothelial cells and macrophages. In mice, genistein lowers CXCL10, IFN‑γ, cTnI and cardiac inflammatory gene expression after high-dose BNT162b2, with limited impact on vaccine-induced antibody titres.
Interpretation
- The authors propose that: a **CXCL10–IFN‑γ signalling cascade** is a major driver of mRNA‑vaccine-associated myocardial injury, this may represent a **class effect** of mRNA platforms, and **genistein** might be a candidate intervention to prevent or ameliorate vaccine‑associated myocarditis.
- From the WorthDay perspective, several points substantially limit the evidential value for human causality:
- **Study type:** all primary experiments are in **mice and iPSC-derived cells**; the human component is limited to **secondary analysis of proteomics**, not controlled intervention.
- **Endpoints:** almost all outcomes are **soft surrogates** (troponin, NT‑proBNP, cytokines, gene expression, imaging, histology). There are **no hard clinical endpoints in humans** such as observed myocarditis incidence, severity, or mortality under controlled modulation of CXCL10/IFN‑γ.
- **Dosing and timing:** robust myocardial marker changes in mice occur mainly at **2–5 µg** BNT162b2, far above the body‑surface area-adjusted 0.2‑µg "human-equivalent" dose. Sampling at 6–10 hours after the second dose does not match typical clinical timing of myocarditis onset (often 1–3 days).
- **Model extrapolation:** cytokine doses in cell culture are inferred from proteomics data and may not reflect local cardiac concentrations in humans.
- **Clinical pattern not reproduced:** the key clinical feature—overrepresentation in **young males**—is not convincingly modelled; in mice, both sexes respond similarly in the reported markers.
- **Genistein:** while it modulates surrogate markers in models, there is **no human trial** showing that genistein prevents or treats vaccine-associated myocarditis; as a hormone‑active phytoestrogen, its broader safety profile in this context is unclear.
- Under WorthDay Evidence Criteria, S0058 therefore supports **mechanistic hypotheses** about CXCL10/IFN‑γ in model systems, but provides **very low-quality evidence** regarding: why young vaccinated humans develop myocarditis, whether this is a true "class effect" of all mRNA vaccines, or whether genistein is a safe and effective clinical intervention.
Limitations
- All primary data are from **mouse models and iPSC-derived cells**, not controlled human interventions.
- Human data limited to **secondary proteomics analysis**; no direct manipulation of CXCL10/IFN‑γ in patients.
- Key mouse effects occur at **10–25× human-equivalent doses**; relevance to clinical vaccination uncertain.
- Timing of measurements (hours after dose) does not match typical human myocarditis presentation (days).
- The male preponderance seen clinically is not reproduced in these models.
- Genistein has no human trial data for vaccine-associated myocarditis prevention or treatment.
WorthDay perspective
- S0058 is a hypothesis-generating preclinical study that explores biological mechanisms potentially relevant to mRNA-vaccine-associated myocarditis. It does **not** provide evidence that these mechanisms cause human disease or that genistein is a safe clinical intervention.
- Under the WorthDay Evidence Criteria, this is **very low-quality evidence**: interesting for researchers designing future human studies, but not a basis for personal health decisions about vaccination or supplement use.
- For WorthDay users: This study illustrates how basic science explores mechanisms, but also why such work cannot directly answer clinical questions. Decisions about vaccination should be based on large-scale human safety and efficacy data, not preclinical models.
Permalink: https://www.worth-day.com/science-compass/s0058-cxcl10-ifn-axis-and-genistein-in-preclinical-models-of-mrna
Related studies
Science meets practice