Early Mitochondrial Changes Emerge Before Disease in Arrhythmogenic Cardiomyopathy Model

In arrhythmogenic cardiomyopathy, a heritable heart muscle disorder most often tied to faulty desmosomal proteins, the heart’s molecular machinery begins to fail long before any symptom appears. A new study published in Nature Cardiovascular Research has now mapped that hidden phase in unprecedented detail, using time-resolved proteomics in a mouse model of PKP2 deficiency to show that mitochondrial remodeling is one of the earliest detectable changes — and that a compound called honokiol can broadly blunt the pathological protein shifts.

The research team, led by Jonathan S. Achter, Mikkel Svenningsen, John Mulvey and Giorgia Bertoli in the laboratories of Alicia Lundby and Mario Delmar, used an inducible, cardiomyocyte-specific knockout of Pkp2, the gene encoding plakophilin-2, a core structural component of the desmosome, the junctional complex that mechanically ties heart muscle cells together. Adult mice received tamoxifen injections to trigger the knockout, and the researchers then collected the right and left ventricular free walls separately at five time points: 7, 10, 14, 17 and 21 days post-induction, with tissue from Cre-negative littermates serving as controls.

The proteomic workflow was unusually deep. Across 72 cardiac tissue samples, the mass spectrometry pipeline identified more than 4,000 proteins per sample, 6,527 unique proteins in total, quantified from 117,767 precursor ions. Crucially, PKP2 abundance declined progressively and in parallel in both ventricles, falling to roughly half of control levels by day 10 and about a quarter by day 21. Because the depletion was comparable between chambers, any difference in remodeling between them could be attributed to intrinsic ventricular properties rather than unequal loss of the trigger protein.

That asymmetry turned out to be one of the study’s central findings. Principal component analysis showed that in both ventricles, time — that is, progression of the phenotype — was the dominant source of proteome variation, explaining more than 40 percent of the variance along the first principal component. But the right ventricle’s trajectory shifted earlier than the left’s, despite identical PKP2 loss. This mirrors the human disease, in which the right ventricle is typically affected first and most severely, and it suggests that chamber-specific vulnerability is written into the ventricles’ baseline molecular differences rather than into the knockout itself.

The earliest remodeling was strikingly coordinated. Proteins whose abundance tracked most tightly with PKP2 included the desmosomal components junction plakoglobin, desmoplakin, desmocollin-2 and desmoglein-2, along with adherens junction proteins such as catenin alpha-3 and N-cadherin, with Pearson correlations ranging from 0.91 to 0.97. Loss of PKP2 therefore destabilizes the entire mechanical junction complex rather than a single subunit, and the components decay in an ordered hierarchy: in the right ventricle, desmoglein-2 declined at the same rate as PKP2, followed by desmocollin-2, desmoplakin and junction plakoglobin, with the same rank order appearing at modestly lower rates in the left ventricle.

Interwoven with this structural collapse was a metabolic one. Among the proteins most strongly co-regulated with PKP2 were mitochondrial enzymes, including acyl-CoA thioesterase 2 and acyl-CoA dehydrogenase family 11, both involved in fatty acid metabolism. Regression analysis across the time course revealed extensive remodeling in both ventricles — 614 proteins significantly downregulated and 823 upregulated in the right ventricle, with comparable numbers in the left — and hierarchical clustering resolved six distinct temporal trajectories. The downregulated clusters were dominated by cardiomyocyte proteins involved in aerobic respiration, fatty acid oxidation and muscle contraction, while the upregulated clusters carried markers of myeloid and B cells early on and of fibroblasts and endocardial cells later, consistent with progressive cardiomyocyte loss and inflammatory, fibrotic replacement.

The mitochondrial changes were not uniform but staged. Proteins involved in fatty acid oxidation and ATP synthase declined persistently from early on, whereas a broader set — spanning fatty acid and amino acid metabolism and respiratory complex I — began falling only around day 17. Meanwhile, mitochondrial DNA copy number increased, mitochondria were significantly larger at day 10 (a linear mixed model across 2,286 segmented organelles gave P = 0.0038), citrate synthase activity rose, and PINK1, a marker of mitophagy, was elevated, even as most mitochondrial proteins, including mitochondrially encoded subunits, declined. Mitochondrial superoxide showed only a non-significant trend at day 10, suggesting that structural and molecular mitochondrial damage precedes overt oxidative dysfunction, which appears later in the disease course.

To test whether this early mitochondrial remodeling could be therapeutically intercepted, the researchers turned to honokiol, a polyphenolic compound reported to activate the mitochondrial NAD-dependent deacetylase sirtuin-3, which their own earlier work had shown ameliorates the phenotype in these mice. Starting at day 7, Pkp2 knockout mice received honokiol or vehicle for 14 days. Proteomic analysis of right ventricular tissue at day 21 showed that honokiol shifted protein abundances in directions opposite to the disease trajectory: proteins in the persistently and delayed downregulated clusters were relatively preserved, while disease-upregulated clusters were reciprocally reduced. Notably, the protective effect extended beyond the mitochondrial module to the full range of declining functional groups.

Thermal proteome profiling in isolated adult mouse cardiomyocytes offered a window into how honokiol engages its targets. Of more than 6,700 proteins whose melting behavior was resolved across nine temperatures, 60 showed significant thermal stability shifts upon honokiol treatment, with several inner mitochondrial membrane proteins — including the complex III subunit UQCR11, IMMP2l and NME4 — stabilized, and oxidative phosphorylation, particularly complex III, among the strongest enriched signatures. Interestingly, sirtuin-3 itself showed no significant thermal shift, and the authors note that although honokiol can bind SIRT3 directly, that interaction acts as a net steady-state inhibitor, implying that honokiol’s sustained cellular effects likely arise through indirect mechanisms, possibly including a shift in the mitochondrial NAD+/NADH ratio that boosts SIRT3 flux.

The authors are careful to frame these findings appropriately. All results come from a mouse model with tamoxifen-induced knockout in 3-to-6-month-old male mice, and molecular changes — however early and coordinated — are model estimates of disease progression, not demonstrations of efficacy in humans. Arrhythmogenic cardiomyopathy still lacks therapies that prevent arrhythmia or halt cardiomyopathy progression, and the study’s contribution is a spatiotemporal map of the concealed stage: molecular changes precede functional and anatomical abnormalities, unfold in a defined sequence from desmosome to mitochondria to organ-level manifestations, and can be pharmacologically modulated at their earliest detectable point. Whether honokiol or related mitochondrial-targeting strategies can translate into interventions that delay disease onset in patients remains a question for future work, but the study establishes that the window for such intervention opens far earlier than symptoms ever suggest.

Subject of Research: Temporal proteomics of PKP2-arrhythmogenic cardiomyopathy in a mouse model

Article Title: Temporal proteomics maps concealed-stage remodeling and reveals early mitochondrial alteration as an intervention target in a model of PKP2-arrhythmogenic cardiomyopathy

Article References: Achter, J. S., Svenningsen, M., Mulvey, J., Bertoli, G., Meyer, E., Yepuri, G., Zhang, M., Ramasamy, R., Delmar, M., & Lundby, A. (2026). Temporal proteomics maps concealed-stage remodeling and reveals early mitochondrial alteration as an intervention target in a model of PKP2-arrhythmogenic cardiomyopathy. Nature Cardiovascular Research. https://doi.org/10.1038/s44161-026-00894-4

Image Credits: AI Generated

DOI: 10.1038/s44161-026-00894-4

Keywords: arrhythmogenic cardiomyopathy, plakophilin-2, proteomics, mitochondria, honokiol, right ventricle, desmosome, mouse model, thermal proteome profiling, sirtuin-3, fatty acid metabolism, early intervention