The Ugly Side of NO: Erythrocyte-derived NO and Atherosclerosis

Nitric oxide (NO) has long occupied the “good” side of vascular biology, reflecting the longstanding view that NO is protective and promotes healthy vascular function. Indeed, a generalist’s view of NO typically focuses on the “good” that the endogenous gas can do in the vasculature, being critical not only for vasorelaxation but also for preventing thrombosis, inflammation, and deleterious vascular cell proliferation. Although decades of research have made clear that loss of NO promotes atherogenesis, NO has a Janus face in which excessive or dysregulated NO signaling can be maladaptive.¹ The idea that too much NO could promote atherosclerosis was intimated in seminal studies showing that eNOS overexpression accelerates atherogenesis in lesion prone mice.² Sun and colleagues extend on this general concept, providing further evidence of a novel route by which NO generation can turn “ugly”—via erythrocyte-sourced NO that rewires smooth muscle metabolism to favor lipogenesis and lipid storage, thereby contributing to atherosclerosis.³

The possibility that erythrocytes can synthesize NO was raised nearly 30 years ago by Jubelin and Gierman,⁴ although the field remained controversial until later studies provided stronger evidence for functional erythrocyte eNOS.⁵ More recent studies, ranging from human erythrocyte physiology to cell-specific mouse genetic experiments, have shown that erythrocyte eNOS influences a range of functions including red cell deformability, platelet reactivity, endothelial function, blood pressure homeostasis, and ischemic preconditioning.^6–9 Consistent with the current study,³ previous findings demonstrate that erythrocyte arginase-1 (Arg1) influences vascular calcification and the cell composition of atherosclerotic lesions, pointing to erythrocyte-derived NO as a mediator of the effects.¹⁰ Extending these findings, Sun and colleagues provide an arc of evidence—spanning from not only erythrocyte-specific genetic perturbation, but also from primary smooth muscle phenotyping, -omics and pharmacologic dissection of pathway logic, and atherosclerosis in vivo—to show that lack of Arg1 in erythrocytes increases NO production, which reprograms smooth muscle cells to store lipid.

While the presence of Arg1 in erythrocytes has been known since the mid-1980s,¹¹ the functional effects of its presence have remained unclear. Later studies tied erythrocyte Arg1 activity directly to NOS activity, demonstrating tight control of red cell NO biosynthesis by Arg1.¹² The purported mechanism by which Arg1 regulates NO in erythrocytes is based in substrate competition: because Arg1 and NOS compete for the same substrate (L-arginine), reducing Arg1 activity shifts arginine metabolism toward greater NO generation. Consistent with this mechanism, Sun and colleagues showed that erythrocyte Arg1 deletion increases NO production, which appears to prime smooth muscle cells to assume a foam cell-like phenotype and promote atherosclerosis.

A major strength of the study is that the central claim is not carried by a single model, but by a chain of complementary models that afford mechanistic depth. Rather than identifying a single downstream effector, the authors outline a signaling cascade by which erythrocyte-derived NO reprograms smooth muscle lipid handling. Smooth muscle cells from erythrocyte-specific Arg1-deficient mice accumulated lipid droplets and exhibited a coordinated program of enhanced lipid uptake and de novo lipogenesis. Mechanistically, the authors place NO-sensitive sGC→cGMP→PKG signaling upstream of CD36 induction, supported by increased VASP phosphorylation, increased CD36 expression, and the observation that NO donors or sGC agonism phenocopies the lipid-laden phenotype, whereas inhibition of sGC, PKG, or CD36 blunts foam cell formation. RNA-seq and metabolic analyses further point to activation of the lipogenic machinery, including ACC and FASN, and identified loss of PDE2A as a potential molecular switch that couples chronic cGMP signaling to increased cAMP-PKA activity. In turn, PKA-dependent inhibition of AMPK releases ACC from its normal restraint, thereby favoring fatty acid synthesis and storage. Consistent with this mechanistic framework, erythrocyte Arg1 deletion in vivo increased lipid-rich atheroma burden in the aortic arch and thoracic aorta of apoE-deficient mice and was accompanied by greater vascular permeability, more intraplaque erythrocytes, and enrichment of plaque cholesterol esters, linking smooth muscle metabolic reprogramming to accelerated plaque formation. Thus, the study by Sun and colleagues advances a model in which erythrocyte-derived NO does not merely expose smooth muscle cells to more lipid, but metabolically rewires them to become more competent lipid-storing foam cells within the atherosclerotic vessel wall (Fig. 1).

Fig. 1: The Good, the Bad, and the Ugly of NO.

Conceptual models for roles of NO in vascular health and disease: (a) the Good – endothelium-derived NO preserves vascular health; (b) the Bad – oxidative and nitrosative stress promotes vascular injury and dysfunction; and (c) the Ugly – erythrocyte-derived NO drives smooth muscle cell (SMC) foam cell formation and atherosclerosis. As depicted in panel c, the findings of Sun and colleagues show that low arginase 1 activity in intraplaque erythrocytes promotes NO generation from erythrocyte-resident NO synthase, which diffuses to neighboring SMCs, triggering a cascade of responses that increase CD36 and upregulate acetyl CoA carboxylase (ACC) and fatty acid synthase (FASN), thereby altering metabolism to promote de novo lipogenesis, foam cell formation, and plaque growth. Schematics were created using BioRender (https://biorender.com).

A broader translational implication of this work is that it intersects with a growing literature placing erythrocytes within the atherogenic microenvironment rather than outside it. In human atherosclerotic lesions, erythrocyte membranes accumulate within the necrotic core, and intraplaque hemorrhage has been linked to more rapid plaque progression,^{13, 14} supporting the idea that erythrocytes can actively contribute to lesion expansion or instability once endothelial leak or neovessel rupture permits red cell entry. Viewed in this context, the study by Sun and colleagues adds a new dimension: beyond delivering cholesterol-rich membrane material and prooxidant cargo (e.g., heme), erythrocytes may also reshape plaque biology through NO-dependent metabolic reprogramming of neighboring smooth muscle cells. At the same time, an important caveat is species. Human and other primate erythrocytes express high Arg1 activity, whereas mouse erythrocytes express comparatively less,^{11, 12} a distinction recognized decades ago. Thus, erythrocyte Arg1 deletion in mice may model a pathway that is potentially quite relevant to human disease, even if the quantitative importance of this axis in murine physiology differs from that of humans.

The findings of Sun and colleagues are particularly notable for shifting attention toward erythrocytes as active participants in plaque biology and toward smooth muscle metabolism as a determinant of atheroma expansion. Nevertheless, several questions now follow. It will be important to determine how strongly this erythrocyte-smooth muscle axis operates in human atherosclerosis, and whether it is most relevant to advanced, erythrocyte-rich plaques marked by endothelial leak or intraplaque hemorrhage rather than to all legions uniformly. The species difference in erythrocyte Arg1 expression warrants caution, but it also underscores opportunity: if human erythrocytes are indeed more enriched in Arg1 and more dependent on this axis than mouse erythrocytes, then the pathway described by Sun and colleagues could have greater translational relevance than standard murine physiology might imply. Moreover, the fact that smooth muscle cells are major contributors to human atherosclerotic lesions, accounting for roughly 30–50% of cells expressing macrophage markers,¹⁵ underscores how important the differentiation of smooth muscle cells into foam-like cells may be to atherosclerosis.

This study also raises a translational possibility: if erythrocyte Arg1 proves to be dynamically regulated in human disease, then defining the endogenous mechanisms that govern its activity may open the door to therapeutic modulation. In that setting, targeted approaches, perhaps via erythrocyte- or plaque-directed delivery systems rather than systemic arginase manipulation, could offer a way to blunt maladaptive red cell-vascular signaling while preserving the broader physiological functions of NO. Based on the mechanistic findings, it would seem that identifying ways of increasing Arg1 activity in erythrocytes within the diseased vascular wall could mitigate NO generation by siphoning arginine toward ornithine rather than leaving it available to be acted upon by eNOS. Although selective pharmacologic activators of Arg1 are not currently established, Arg1 activity can be enhanced by factors that favor its active state, including manganese loading and trimer stabilization, raising the intriguing but still speculative possibility that local augmentation of erythrocyte Arg1 activity within plaques could restrain red cell-derived NO signaling and mitigate the progression of atherosclerosis. In the end, the “ugly” side of NO may lie not in the molecule itself, but in the pathological context that turns a vasoprotective signal into a driver of plaque growth.