DNMT3A, efferocytosis, and plaque repair converge in atherosclerosis regression
Atherosclerosis is not only a disease of lipid accumulation. It is also a disease of failed cleanup, where dead and dying cells linger inside the plaque and fuel the inflammatory cycle that destabilizes it. In this JCI commentary, Aarti Asnani examines new mechanistic work showing that efferocytosis — the macrophage-mediated clearance of apoptotic cells — is wired to a DNMT3A-dependent DNA repair program that helps macrophages proliferate and support tissue resolution.
The biology matters because plaque regression is not passive. A regressing lesion needs macrophages that can clear debris, expand locally, and help build a thicker fibrous cap. Without that repair response, lipid lowering may reduce burden without fully restoring stability.
A repair program hidden inside apoptotic cell clearance
The commentary centers on a surprising finding: when macrophages engulf dying cells, they do more than digest cellular debris. They activate base excision repair to remove oxidized DNA lesions, especially 8-OHdG, and this appears to be necessary for efferocytosis-induced macrophage proliferation. The repair pathway involves PARP1 chromatin binding, PARylation, and increased nuclear MTH1 activity, all coordinated through DNMT3A.
This is an elegant answer to a longstanding question in vascular immunology. Why should a macrophage that is clearing corpses divide more actively? The answer emerging from these data is that proliferation is coupled to genotoxic stress control. Apoptotic-cell processing creates a metabolic and oxidative state that would otherwise damage DNA; DNMT3A-linked repair allows the cell to keep functioning while expanding the pro-resolving macrophage pool.
Macrophage proliferation is not a side reaction here. It is part of the resolution machinery.
Why DNMT3A matters in the arterial wall
Asnani places this mechanism in the context of DNMT3A-mutation–driven clonal hematopoiesis, a condition already linked to higher coronary risk and accelerated atherosclerosis. The commentary emphasizes that DNMT3A is the most frequently mutated gene in clonal hematopoiesis, and that loss-of-function variants appear to blunt the inflammatory-to-resolution switch in plaques.
That point is clinically important because the same mutant clone may alter several steps in lesion biology at once. It can influence cytokine signaling, macrophage behavior, and now, based on the related study discussed in the commentary, the ability of efferocytosis to trigger DNA repair and macrophage expansion. A mutation in a hematopoietic regulator is therefore not just a genetic marker; it may be a direct modifier of how plaques heal after injury or lipid lowering.
The specific variant discussed, R882H, carries loss of function in the methyltransferase domain. In the commentary, that detail is used to frame a plausible explanation for why some DNMT3A-mutant states may track more closely with cardiovascular phenotypes than others.
Fibrous cap thickening depends on the same pathway
The commentary draws a straight line from cellular repair to plaque architecture. In the regression model discussed, LDL-lowering treatment thickened the fibrous cap in wild-type mice, while DNMT3A-mutant clonal hematopoiesis impaired that response. Less nuclear 8-OHdG suppression, reduced PARP1 and MTH1 activity, and weaker macrophage proliferation all accompanied the blunted cap response.
Fibrous cap thickening is one of the few structural endpoints that speaks directly to plaque stability. A thicker cap means a lesion is better able to resist rupture under hemodynamic stress. When that endpoint weakens in DNMT3A-mutant settings, the clinical implication is sobering: lipid lowering alone may not normalize the biology of resolution if the macrophage compartment cannot execute the repair program.
The translational logic is straightforward. If efferocytosis is defective or its downstream repair machinery is impaired, regression becomes incomplete even when LDL is lowered.
The thymus model points to a broader tissue-resolution role
The commentary also notes that the related work extended beyond atherosclerosis into a thymic repair model driven by dexamethasone-induced thymocyte apoptosis. There, DNMT3A was required for nuclear PARP1 and MTH1 increases, suppression of oxidized DNA, macrophage proliferation, and thymic repair. That broader setting strengthens the idea that DNMT3A is not merely a plaque-specific factor; it participates in a conserved tissue-resolution pathway.
This breadth matters because cardiovascular inflammation rarely exists in isolation. The same hematopoietic clone may shape inflammatory tone across tissues, and the same repair program may be reused in different organs under different forms of injury. The commentary hints at overlap with heart failure biology, where DNMT3A mutations have also been tied to inflammatory signaling and cardiac remodeling.
Where the field goes from here
Asnani closes by pointing to questions that are now hard to ignore. Which macrophage subsets are most dependent on DNMT3A during regression? How much of the effect is mediated by DNA methylation versus the downstream repair machinery? And can lipid-lowering therapy be tailored differently in patients with DNMT3A clonal hematopoiesis, especially those carrying R882H?
Those questions are not academic. They define whether clonal hematopoiesis remains a risk marker or becomes a treatment-stratifying variable. Single-cell profiling is one obvious next step, because bulk tissue analysis cannot separate the rare macrophage states that may drive resolution from the more abundant inflammatory ones that stall it.
For now, the commentary sharpens the central message from the related experimental study: efferocytosis is not simply debris removal. It activates a DNMT3A-linked oxidized DNA repair pathway that allows macrophages to proliferate, repair tissue, and support plaque stabilization. When DNMT3A function is compromised, the repair sequence weakens, and the plaque may remain biologically vulnerable even after LDL reduction.