Transforming endothelial cells in atherosclerosis

Cells contributing to atherosclerotic disease are highly plastic and can shift their phenotype in a changing microenvironment. A study in Nature Metabolism now reveals that transforming growth factor-β (TGF-β) can transform endothelial cells into pro-inflammatory cells and that inhibition of TGF-β-receptor signalling in the endothelium can reverse atherosclerosis in mice.

Endothelial cells (ECs) line all blood vessels and are critical mediators of inflammatory responses. In the setting of atherosclerosis, ECs become chronically activated through a combination of turbulent blood flow, lipid accumulation in the vessel wall and exposure to inflammatory mediators (for example, IL-1β)1. The activated endothelium in turn orchestrates the recruitment and maintenance of inflammatory cells in the expanding atherosclerotic plaque. Understanding how ECs respond to the complex inflammatory microenvironment driving plaque initiation, progression and maintenance is requisite for effective therapy. Recent clinical trials have revealed that globally antagonizing IL-1β signalling significantly (though modestly) decreases cardiovascular events in patients with elevated systemic inflammation despite effective lipid control2, thus providing proof-of-concept evidence that targeting inflammation may be a feasible approach to combat atherosclerosis. However, IL-1β antagonism is also associated with an elevated risk of fatal infections2. Targeted therapy is clearly warranted. The response of the endothelium to inflammatory stimuli has been identified as one such target. Beyond expressing leukocyte adhesion molecules and producing chemokines that communicate to circulating and resident inflammatory cells, activated ECs can transform into mesenchymal-like cells in a process known as endothelial-to-mesenchymal transition (EndMT). These cells either co-express EC and mesenchymal markers (a state referred to as partial EndMT) or fully lose EC identity, delaminate from the endothelium and contribute to populations of fibroblasts and, to a lesser extent, smooth muscle cells in the plaque interior3,4. Because EndMT has been linked to vulnerable atherosclerotic plaques and clinical events4, understanding how this process is regulated could lead to the development of effective therapeutic approaches.

As a key driver of EndMT (Fig. 1), the cytokine TGF-β has recently been posited to be pro- atherosclerotic3,4. However, the role of TGF-β signalling in atherogenesis is controversial, because human studies have also implicated TGF-β in plaque stabilization5. This controversy reflects the varied cell- and context-dependent roles of ligands, receptors and downstream transcription factors in this pathway6. Although global inhibition of TGF-β signalling or specific modulation in non-ECs has suggested that TGF-β suppresses atherosclerosis in vivo7,8,9,10, the contribution of TGF-β signalling in the endothelium itself has not been directly tested. In this issue of Nature Metabolism, Chen et al. report that TGF-β stimulation uniquely induces markers of inflammation in cultured ECs, but not smooth muscle cells, T-cells or macrophages11. By specifically deleting TGF-β receptor 1 and 2 (Tgfbr1/2) in the endothelium before exposure of atherosclerosis-prone Apoe−/− mice to a high-cholesterol, high-fat diet, the authors demonstrate delayed onset of atherosclerosis, decreased plaque size and macrophage content and attenuation of necrotic core formation. The translational potential of this finding was enhanced by the use of established atherosclerosis models, through which the authors determined that inducible EC-specific deletion of Tgfbr1/2 not only impedes further plaque progression in mice fed a high-cholesterol, high-fat diet but also augments plaque regression when the animals are switched to a normal diet, thus suggesting that continual endothelial TGF-β signalling is required for plaque maintenance. Furthermore, a therapeutic approach using EC-specific nanotherapy to deliver high-affinity Tgfbr1/2 short interfering RNA in mouse models of atherosclerosis progression and regression yielded similar encouraging results. Together, these studies illustrate that targeting TGF-β signalling specifically in the endothelium may mediate plaque regression.

Fig. 1: Endothelial TGF-β signalling drives atherogenesis.

Signalling through TGFβR1/2 enhances the expression of inflammatory and mesenchymal genes, while modulating ECM gene expression and suppressing EC identity genes. As a result, ECs are activated, undergo EndMT and promote the formation of atherogenic lesions. This TGF-β-mediated transformation of the EC phenotype is likely to affect monocyte recruitment as well as paracrine signalling within the atherosclerotic niche. Therapeutic inhibition of TGFβR1/2 signalling in ECs, for example through delivery of short interfering RNAs to the endothelium, has been shown to elicit atherosclerotic-plaque regression in mouse models.

Although the effects on atherosclerosis are provocative, perhaps most intriguing are the findings from the single-cell RNA-sequencing of aortic ‘ECs’ obtained from these models. Of the CD31+ cell clusters identified, two were characterized by diminished levels of EC identity genes, and with intact TGF-β signalling, cells had features of a transformed EndMT phenotype, including expression of mesenchymal markers, genes involved in modulation of the extracellular matrix (ECM) and elevated adhesion molecules and chemokines. In the absence of TGF-β signalling, these EndMT and inflammatory signatures were diminished. Additional clusters were identified representing partial-EndMT phenotypes, in which cells retained EC markers but co-expressed mesenchymal markers and had pro-inflammatory gene signatures. Again, ablating TGF-β signalling appeared to decrease these inflammatory/ECM features. Although previous studies have suggested that TGF-β may drive EndMT3,4, Chen et al. provide further insight revealing that a ‘transformed’ pro-inflammatory endothelial phenotype exists in atherosclerotic plaques and is partially corrected when TGF-β signalling is blocked in the endothelium11.

Fundamental questions regarding the roles of TGF-β and EndMT in atherosclerosis remain to be addressed in future studies: To what extent is delamination and migration of EndMT cells required for their contribution to atherosclerosis? Is full EndMT required, or is partial EndMT sufficient? Does this process drive plaque vulnerability or calcification? Is EndMT a consequence or a cause of the pro-inflammatory phenotype? In future studies, it will be important to determine whether antagonism of endothelial TGF-β signalling impairs migration of endothelial-derived cells into the plaque. It remains possible that the expression of adhesion molecules, chemokines or ECM proteins by cells undergoing EndMT in situ may modulate the plaque niche in the absence of migration by enhancing the recruitment of inflammatory cells or by altering the survival, proliferation or phenotype (for example, efferocytosis) of recruited or resident cells.

Although Chen et al. provide a compelling rationale to target TGF-β signalling exclusively in the endothelium in atherosclerosis, much work remains to be done before clinical translation. With any intervention that chronically suppresses endothelial activation, the possibility of increased susceptibility to infection and/or delayed response to injury exists. Atherosclerotic regression should also be examined in large-animal models after EC-specific Tgfbr1/2 knockdown to determine whether treatment will be effective in advanced, unstable plaques. Finally, it is worth establishing whether blocking TGF-β signalling in the endothelium might be effective in treating other diseases involving inflammation-driven EndMT, such as organ fibrosis, transplant vasculopathy and pulmonary arterial hypertension12. This endothelial TGF-β signalling axis may well be a critical regulatory node that can be therapeutically targeted in multiple vascular inflammatory diseases.


  1. 1.

    Gimbrone, M. A. Jr. & García-Cardeña, G. Circ. Res. 118, 620–636 (2016).

    CAS  Article  Google Scholar 

  2. 2.

    Ridker, P. M. et al. N. Engl. J. Med. 377, 1119–1131 (2017).

    CAS  Article  Google Scholar 

  3. 3.

    Chen, P. Y. et al. J. Clin. Invest. 125, 4514–4528 (2015).

    Article  Google Scholar 

  4. 4.

    Evrard, S. M. et al. Nat. Commun. 7, 11853 (2016).

    CAS  Article  Google Scholar 

  5. 5.

    Cipollone, F. et al. Stroke 35, 2253–2257 (2004).

    CAS  Article  Google Scholar 

  6. 6.

    Toma, I. & McCaffrey, T. A. Cell Tissue Res. 347, 155–175 (2012).

    CAS  Article  Google Scholar 

  7. 7.

    Chen, P. Y., Qin, L., Li, G., Tellides, G. & Simons, M. EMBO Mol. Med. 8, 712–728 (2016).

    CAS  Article  Google Scholar 

  8. 8.

    Lievens, D. et al. Eur. Heart J. 34, 3717–3727 (2013).

    CAS  Article  Google Scholar 

  9. 9.

    Mallat, Z. et al. Circ. Res. 89, 930–934 (2001).

    CAS  Article  Google Scholar 

  10. 10.

    Robertson, A. K. et al. J. Clin. Invest. 112, 1342–1350 (2003).

    CAS  Article  Google Scholar 

  11. 11.

    Chen, P.-Y. et al. Nat. Metab. (2019).

  12. 12.

    Kovacic, J. C. et al. J. Am. Coll. Cardiol. 73, 190–209 (2019).

    Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Kathryn L. Howe or Jason E. Fish.

Ethics declarations

Competing interests

The authors declare no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Howe, K.L., Fish, J.E. Transforming endothelial cells in atherosclerosis. Nat Metab 1, 856–857 (2019).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing