Tire Chemicals Linked to Heart Rhythm Disorder in Landmark Study

A ubiquitous family of tire-derived chemicals that has already gained notoriety for killing salmon in urban streams may now pose a threat to the human heart. In a sweeping integrative study published in the Journal of Advanced Research, researchers report that 6PPD, a rubber antioxidant added to virtually every automotive tire, and its ozone-formed transformation product 6PPD-quinone (6PPD-Q) are associated with molecular signatures of atrial fibrillation, the world’s most common sustained cardiac arrhythmia. Combining network toxicology, machine learning, single-cell transcriptomics, molecular docking, and laboratory experiments, the team built a multi-layered case that these roadside pollutants may help create the electrical and structural conditions in which the atria fall into chaotic, stroke-promoting rhythm.

The stakes are considerable. Atrial fibrillation afflicts tens of millions of people worldwide and dramatically raises the risk of ischemic stroke, heart failure, and death from any cause. Classical culprits, including hypertension, aging, and metabolic disease, explain much of the burden, but epidemiologists have repeatedly observed that traffic-related air pollution and fine particulate matter also increase AF incidence. That observation has fueled the concept of the cardiac exposome, the idea that external chemical agents sculpt an arrhythmia-prone atrial environment by driving oxidative stress, inflammation, and the electrical and structural remodeling of heart tissue. What remained unknown was whether tire wear chemicals, which are shed continuously onto roads and washed into waterways, belong on that list of cardiac hazards.

6PPD is added to tires to prevent rubber from degrading on contact with ozone in air. As tires abrade against pavement, particles loaded with the compound accumulate on road surfaces, and oxidation rapidly converts a fraction of it into 6PPD-Q. The transformation product shot to scientific fame in 2021 when it was identified as the agent behind urban runoff mortality syndrome in coho salmon. Since then, monitoring studies have detected both chemicals in soils, waterways, and air, while human biomonitoring has confirmed their presence in urine and blood. In a South Chinese cohort, for example, median urinary 6PPD-Q reached 2.91 nanograms per milliliter in pregnant women. Experimental work has already tied the compounds to oxidative stress, mitochondrial dysfunction, and tissue injury in several organ systems, including the cardiovascular system, but a direct molecular link to arrhythmia had never been mapped.

To build that map, the researchers first assembled a comprehensive picture of atrial fibrillation’s molecular landscape. They integrated three independent transcriptomic datasets of human atrial tissue, corrected for batch effects, and identified 249 genes significantly dysregulated in AF. Weighted gene co-expression network analysis then grouped thousands of genes into modules correlated with disease status, and a curated set of high-confidence AF genes from the GeneCards database was layered on top. The union of these approaches yielded 522 candidate AF-associated genes. In parallel, three complementary target-prediction platforms, the Similarity Ensemble Approach, PharmMapper, and SwissTargetPrediction, generated 496 potential human protein targets for 6PPD and 6PPD-Q based on chemical similarity and pharmacophore mapping. Where the two lists intersected, sixteen candidate genes emerged as the molecular crossroads between tire pollution and atrial disease.

The team then turned to machine learning to separate signal from noise. Eleven classical algorithms were combined into 113 predictive classification models, trained and tested with five-fold cross-validation on the transcriptomic data. Ensemble models built around naïve Bayes performed best, and the explainable-AI technique SHAP, which quantifies each feature’s contribution to individual predictions, revealed four genes that dominated the models: DDR1, BCHE, HTR2B, and KCNJ3. Each carries plausible cardiovascular credentials. DDR1 is a collagen-activated receptor tyrosine kinase implicated in cardiac fibrosis and structural remodeling. BCHE encodes butyrylcholinesterase, an enzyme of cholinergic signaling that can influence atrial electrophysiology through autonomic regulation. HTR2B is a serotonin receptor linked to cardiac hypertrophy, fibroblast activation, and inflammatory myocardial responses. KCNJ3 encodes GIRK1, a potassium channel subunit central to the acetylcholine-activated current that shapes atrial action potentials and AF susceptibility.

Single-cell RNA sequencing of human atrial tissue from patients with and without AF added a spatial dimension to the analysis. The four hub genes were not scattered uniformly across the heart’s cellular mosaic. Instead, their expression formed dense hotspots within cardiomyocyte, smooth muscle, and fibroblast clusters, while lymphocytes and myeloid cells showed little to no signal. Gene set enrichment analysis reinforced the biological coherence of the findings: high DDR1 expression tracked with cardiomyopathy and remodeling pathways, high BCHE with protein turnover and neurodegenerative modules, high HTR2B with immune activation, and high KCNJ3 with energy metabolism and proteolysis. Molecular docking simulations, using AlphaFold and UniProt protein structures, predicted favorable binding of both 6PPD and 6PPD-Q to all four proteins, with Vina scores between roughly minus 8.7 and minus 7.5 kilocalories per mole, a range consistent with meaningful structural compatibility.

Crucially, the computational predictions were then tested at the bench. Human cardiac fibroblasts exposed for 24 hours to 6PPD at 2.5 or 5.0 micromolar, or to 6PPD-Q at 0.25 or 0.50 micromolar, showed significantly elevated mRNA and protein levels of all four candidate targets compared with vehicle controls. The authors are careful to note that these expression changes do not prove the docked compound-protein interactions caused them, but the concordance between in silico and in vitro results strengthens the plausibility of the proposed mechanism.

The most striking evidence came from living animals. Fifty male C57BL/6 mice were divided into groups, and an AF-susceptible model was established by four weeks of daily tail-vein injections of acetylcholine and calcium chloride. Mice that simultaneously received intraperitoneal 6PPD or 6PPD-Q every three days developed markedly higher AF inducibility than susceptible mice given vehicle alone, and their arrhythmic episodes lasted significantly longer. Cardiac tissue from the exposed animals showed upregulated expression of Ddr1, Bche, Htr2b, and Kcnj3, mirroring the human cellular data. Notably, 6PPD-Q produced effects comparable to 6PPD at less than half the administered dose, hinting that the transformation product may be the more potent cardiac actor, although the study was not designed to disentangle compound-specific mechanisms.

The authors are candid about the limits of their work. The transcriptomic cohorts are retrospective and modest in size; bulk tissue profiling can dilute cell-type-specific changes; the fibroblast line used in vitro does not fully recapitulate human atrial myocytes; and the mouse model, along with the acetylcholine-calcium challenge, cannot capture the full complexity of human AF. The exposure doses were chosen from prior rodent toxicology for hazard identification, not to reproduce real-world human exposure, and no direct measurements of circulating or cardiac compound concentrations were made. Toxicokinetic data suggest 6PPD-Q distributes rapidly and is eliminated within hours to a day, but how long its molecular effects persist in the heart remains unknown. Direct target-engagement assays, gene-silencing experiments, and human atrial cardiomyocyte or organoid models will be needed to confirm causality and define the electrophysiological consequences.

Even with those caveats, the study marks a conceptual advance. It demonstrates a template for exposome science in which computational toxicology, artificial intelligence, single-cell biology, and animal experimentation converge on a concrete set of molecular suspects. If future work validates DDR1, BCHE, HTR2B, and KCNJ3 as genuine mediators of pollution-driven arrhythmia, the findings could reshape how regulators weigh tire-derived contaminants, a pollutant class largely invisible to current air quality standards, and open new diagnostic or preventive avenues for a disease that already burdens health systems worldwide. Every mile driven sheds these chemicals onto the road; the question now being asked, with increasing urgency, is how much of that chemical dust ends up echoing in the human heartbeat.

Subject of Research: Molecular links between tire-derived contaminants 6PPD and 6PPD-quinone and atrial fibrillation

Article Title: Network toxicology, machine learning, single-cell analysis, and molecular docking reveal 6PPD/6PPD-Q-associated atrial fibrillation signatures

Article References: Zhan, W., Zhao, J., Li, Y., Ran, S., Lai, L., Wu, J., Zhou, S., Li, R., Cui, J., & Xia, J. (2026). Network toxicology, machine learning, single-cell analysis, and molecular docking reveal 6PPD/6PPD-Q-associated atrial fibrillation signatures. Journal of Advanced Research. https://doi.org/10.1016/j.jare.2026.10.032

Image Credits: AI Generated

DOI: 10.1016/j.jare.2026.10.032

Keywords: 6PPD, 6PPD-quinone, atrial fibrillation, tire wear particles, network toxicology, machine learning, single-cell RNA sequencing, molecular docking, cardiac fibrosis, environmental health, cardiac exposome, air pollution