Rasquinha MT, Lasrado N, Petro-Turnquist E, Weaver E, Venkataraman T, Anderson D, Laserson U, Larman HB, Reddy J.
Biology (MDPI) · July 13, 2022 · DOI: 10.3390/biology11071055
How to cite
Using Phage ImmunoPrecipitation Sequencing (PhIP-Seq) across a 419,915-peptide mouse proteome library, the authors mapped the autoantibody repertoire in the coxsackievirus B3 (CVB3) mouse model of viral myocarditis and uncovered 32 previously unreported autoantibody reactivities — including antibodies to COA4 and PIK3AP1 that were independently validated by ELISA. The pattern was shared with CVB4 but absent in influenza-infected and naïve controls, pointing to a broader, ubiquitously-expressed autoantigen landscape that classical candidate-based assays had missed and reframing post-viral myocarditis as a molecular-mimicry paradigm with autoantigen reach far beyond the heart-specific candidates that dominated three decades of literature.
In this publication:
- PhIP-Seq detected 32 previously unreported autoantibody peptides from 25 proteins in CVB3-infected mice across a 419,915-peptide mouse proteome library.
- ELISA confirmed COA4 and PIK3AP1 reactivity in PhIP-Seq-positive sera; 70% of infected animals had PIK3AP1 antibodies.
- CVB3 and CVB4 reactivity profiles were similar but absent in influenza-A and naïve controls — ruling out generic post-viral signal.
- Findings expand the autoantigen landscape beyond heart-specific candidates that explain reactivity in only 20–31% of dilated cardiomyopathy patients.
Myocarditis — inflammation of the heart muscle — can progress to dilated cardiomyopathy (DCM), a leading reason for heart transplant. Viruses are commonly implicated, but live virus is rarely found in patient tissue by the time disease becomes apparent; pathology is instead driven by autoantibodies the immune system produces against the heart's own proteins. The problem is that the field has historically scanned for autoantibodies one antigen at a time — cardiac myosin, β1 adrenergic receptor, troponin I — and even then, only 20–31% of DCM patients show reactivity to any single candidate. That leaves the autoimmune story dramatically incomplete.
The authors took the unbiased approach. Using PhIP-Seq — a bacteriophage display platform that presents 56-amino-acid peptide tiles spanning the entire mouse proteome (419,915 peptides) — they screened sera from A/J mice infected with coxsackievirus B3 (CVB3), the standard rodent model of viral myocarditis. PhIP-Seq let them ask, in a single assay, which of every mouse protein the immune system was making antibodies against. They cross-checked against naïve controls, against influenza-infected mice, and against CVB4-infected animals (a related enterovirus), then orthogonally validated top hits by ELISA using recombinant protein.
The result: 32 novel autoantibody specificities not previously linked to viral myocarditis, including strong, polyclonal reactivity against COA4 (cytochrome c oxidase assembly factor 4) and PIK3AP1 (phosphoinositide-3-kinase adaptor protein 1) — both ubiquitously expressed housekeeping proteins, both validated by ELISA, both shared between CVB3 and CVB4 but absent in influenza or naïve animals. The work expands the autoantigen landscape of post-viral myocarditis well beyond the heart-specific candidates that dominated 30 years of literature.
