Endothelial Tenascin-X is Going with the Flow

Atherosclerosis is the leading cause of death and disability globally.¹ The progressive deposition of lipids and the inflammatory mediators that drive the formation of atherosclerotic plaques are most evident at branches and bifurcations of blood vessels – sites of disturbed flow. These observations prompted foundational studies by Peter Davies of flow mediated mechano-transduction in the vasculature.² These and subsequent studies characterized the differential alignment of endothelial cells (ECs) under conditions of laminar (LF) and disturbed flow (DF) and the attendant differences in gene expression. For example, EC Prostaglandin G/H synthase -2 (Ptghs2), barely detectable under static conditions, is upregulated by LF, but not DF.³ Studies in hyperlipidemic mice lacking EC Ptghs2⁴ or the receptor (IPr)⁵ for its predominant product in ECs, prostacyclin (PGI₂), reveal a focal atherogenesis reflective of this differential flow mediated gene expression.

Here, Liang et al⁶ have addressed the role of the EC response to DF in regulating endothelial to mesenchymal transition (EndMT), recently shown to be evoked by DF, and a critical regulator of cellular polarity and interactions relevant to atherogenesis.⁷ They found that TNXB, which encodes extracellular matrix glycoprotein tenascin – X (TN-X), to be one of the most upregulated genes by LF, but not by DF, in human umbilical vein endothelial cells. They showed the LF dependent upregulation of TNXB to be dependent on induction of KLF4. Interestingly, KLF4 appears to regulate Ptghs-2 dependent PGI₂ formation.⁸ These observations translated ex vivo as TN-X was differentially expressed in human aortic segments subject to LF vs DF. EndMT markers such as fibronectin, Vim, and Mmp2 and transforming growth factor (TGF)-β-mediated phosphorylation of Smad2 were upregulated in ECs deficient of TNXB.

Reverting to model systems, the authors then showed that that EC depletion of Tnxb in hyperlipidemic mice accelerated atherosclerosis and amplified vascular remodeling after carotid artery ligation. Furthermore, administration of an antibody that neutralized TGF-β to these mice reduced aortic phosphorylation of Smad2 and expression of VCAM1 and CD68 concomitant with attenuation of vascular remodeling and restraint of atherogenesis. Further, immunoprecipitation experiments in vitro are consistent with direct binding of TGF-β to the fibrinogen-like domain of TN-X.

Liang et al have elegantly unpicked a role for EC TN-X in regulating EndMT in a TGF-β dependent manner, but this may just be the end of the beginning. Both tenascins and TGF-β have contrasting, context dependent effects. For example, in these experiments, depletion of Tnxb in mouse ECs leaves expression of Tnxb in vascular smooth muscle cells (VSMCs) intact. The role of Tnxb in these and other cells relevant to atherogenesis remains to be defined. While the physiological role of TN-X is classically involved in the maintaining the structural integrity of collagen fibrils, it has also been implicated in tumor suppression, promotion of angiogenesis and osteoclast differentiation, in addition to TGF-β activation.⁹ Its deficiency in humans results in the Ehlers Danlos Syndrome, characterized by hyper-elastic joints and occasional cardiovascular complications.¹⁰ The related tenascin -C (TN-C) has been reported both to promote and restrain atherosclerosis and is highly context dependent in its function, possibly due to its modular structure and multiple binding partners.⁹ TN-C can act as a scaffold along which VSMCs migrate: VSMC TN-C is upregulated in response to injury. Deletion of the microsomal PGE synthase (mPges) -1 in ECs or VSMCs suppresses PGE₂ but augments PGI₂, which restrains the response to injury, including the expression of TN-C in VSMCs.^11,12

Similarly, the role of TGF-β signaling in atherogenesis is context dependent. While deletion of TGF-β in ECs restrains inflammation and atherosclerosis,¹³ T-cell specific depletion of TGF-β receptor 2 accelerates atherosclerosis in mice.¹⁴ It is unsurprising, as the authors note, that TN-X can inhibit or promote TGF signaling. Like most things in life, it all depends on context.

One suspects that single cell and spatial approaches will feature in the next chapter of this intriguing story which, one hopes, will eventually conclude with therapeutic opportunity.