Plasticity of Arterial and Venous Endothelial Cell Identity – Some Nerve!

The vascular endothelium, the innermost lining of the blood vessel wall, forms a critical barrier between blood and all tissues. Endothelial injury in large caliber arteries contributes to coronary artery disease, stroke, and peripheral artery disease, whereas endothelial injury in veins may predispose to venous thromboembolic disease such as deep vein thrombosis and pulmonary embolism. Accumulating studies over the past few decades revealed that endothelial cells from arteries, veins, and lymphatic vessels exhibit important differences with respect to their appearance, function, regenerative capacity, and molecular repertoire. The identification of specific markers for arterial (EphrinB2) and venous (EphB4) endothelial cells lead to the appreciation that upstream signaling pathways may govern arterial versus venous endothelial cell fate specification.^1–4 For example, VEGF-mediated induction of ERK activation facilitates the induction of Dll4/Notch signaling leading to arterial EphrinB2 expression.^2–4 In contrast, transcriptional activation of Coup-TFII inhibits the VEGF receptor neuropilin to suppress notch signaling, thereby inducing venous marker expression.⁵ The role of VEGF in promoting arterial endothelial cell identity has been well-recognized from genetic ablation studies across species. However, VEGF-independent mechanisms may also be contributory as suggested by the observation that arterial VEGF expression is significantly reduced after birth.⁶ In addition, the heterogeneity of endothelial cells within the vessel wall⁷ raises the question of flexibility of EC identity towards arterial or venous differentiation. In keeping with this, arterial cells contribute to venous vessels, and venous cells may differentiate into arteries or lymphatic endothelial cells in zebrafish embryos.^{8, 9} Collectively, these studies suggest that extrinsic factors, potentially independent of VEGF, may govern endothelial cell fate plasticity during development and in their maintenance in adult vessels.

In this issue of Circulation Research, Parfanuad et al.¹⁰ studied how arterial endothelial cells are different than their venous counterparts and specifically revealed how arterial-venous endothelial cell fate decisions are correlated to the presence of sympathetic innervation. Endothelial cell fate was examined using avian chimera preparations in which the aortic fragments of quails are grafted to the coeleomic cavity of chick embryos at embryonic day 2 (E2). Remarkably, the majority of cells (>90%) emigrating from embryonic day 15 (E15) aorta fragments were found in host arteries, and not veins. Intriguingly, embryonic aorta grafts developmentally earlier than embryonic day 11 (E11) lost this preferential arterial host engraftment and instead ~40% of cells engrafted into host veins. Because the total number of endothelial cells originating from grafted E15 or E8 aorta was similar, the authors hypothesized the existence of extrinsic signals that skewed cell colonization towards arteries and away from veins. Most notably, one key event that occurs around E11 is the onset of innervation of the aortic vessel wall by sympathetic nerves.¹¹ Consistent with this, detection of catecholamine released by perivascular nerve terminals was negligible at E8 and steadily rose to abundant positive glyoxyic acid staining at E15 in aortae. Similar findings for sympathetic innervation were observed on E15 carotid and femoral arteries. To prove that sympathetic innervation altered arterial endothelial cell fate specification, the authors used chemical ablation using 6-hydroxydopamine (6HDOPA) or surgical ablation by removing the neural tube and notochord after the 5^th somite level from E2 quail embryos, the latter resulting in embryos lacking most neural crest derivatives including peripheral nerves. Both approaches resulted in a reduction of arterial markers (ephrinB2, Dll4, and Nrp1) and loss of the arterial phenotype. Furthermore, as a gain-of-function strategy, Pardanaud et al. co-cultured E8 aortic quail fragments with catecholamine-producing E15 sciatic nerves, and grafted into chick hosts. They observed that more than 90% of the grafted ECs acquired an arterial fate and colonized to host arteries.

To delineate the mechanism by which sympathetic innervation can regulate EC fate, the authors studied the role of the neurotransmitter norepinephrine (NE) that is synthesized and released by both the central and peripheral nervous system where it acts on target cells by binding and activating noradrenergic receptors. When the arteries from E8 aortae, that would normally drive ECs to a venous fate, or chorioallantoic membrane (CAM) arteries were exposed to NE, there was a marked increase in arterial colonization. Using pharmacological agonists and antagonists of the adrenergic α1 and α2 receptors, they found that NE increases endothelial ERK activation via adrenergic α1 and α2 receptor signaling in mice and cultured human ECs. Previous studies have linked ERK activation to arterial gene expression–loss of ERK signaling inhibits arterial specific gene expression and arterial branching in zebrafish embryos.^{12, 13} These findings are important as they suggest a direct link by which sympathetic innervation and neurotransmitters drive EC commitment to an arterial cell fate via a NE-α1R/α2R-ERK signaling axis (see Figure).

Figure. Sympathetic nerve activation of α-adrenergic signaling regulates endothelial cell plasticity.

Norepinephrine, a catecholamine synthesized and released by the sympathetic nervous system, binds and activates α1 and α2 adrenergic receptors (α1R and α2R) on target cells. This process results in the recruitment of the G proteins Gq and Gi and subsequent downstream activation of phospholipase-C (PLC) by α1R or inhibition of adenylate cyclase/protein kinase A (PKA) signaling pathways by α2R, effects that both promote ERK activation. α1R- and α2R-mediated activation of ERK induces arterial gene programming, in part, through increasing the expression of EfnB2, Dll4, Nrp1 and negatively regulating the expression of venous-associated genes EphB4, CoupTFII, and Apj. VEGF-mediated ERK activation is also known to positively regulate arterial endothelial cell fate, although how sympathetic innervation interacts with this pathway is not clear.

Collectively, these findings raise several provocative scientific questions and potential clinical implications. For example, if cells are plastic to adopt an arterial fate and engraft into arterial hosts, then are homing markers, (e.g. integrins, chemokines, chemokine receptors, etc) involved in a sympathetic innervation-dependent manner? Chemical or surgical denervation led to a more modest increase (~10%) in venous colonization implicating that there might be additional factors (e.g. VEGF) that cooperatively regulate arterial and venous cell plasticity with sympathetic innervation. Consistent with this notion, VEGF released from peripheral sensory nerves induce arterial differentiation of nascent vessels in skin of mouse embryos, and genetic ablation of VEGF from nerves suppresses arterial differentiation.¹⁴ In addition, since VEGF and angiogenesis are induced in response to hypoxia. sympathetic innervation may either alter VEGF-mediated effects on endothelial cell plasticity developmentally or in response to postnatal pathological injury.

Sympathetic innervation-mediated arterial endothelial cell plasticity may also be relevant to human pathophysiology. In support, arterial endothelial cells of human blood vessels express both α1- and α2-ARs.^{15, 16} Interestingly, when Parfanuad et al. exposed human umbilical vein endothelial cells to NE, it induced expression of arterial markers to levels observed on human umbilical artery endothelial cells. Furthermore, previous studies indicate that catecholamines favorably contribute to collateralization and arteriogenesis in response to hindlimb ischemia in mice independent of effects on blood pressure.^17–20 Moreover, the sympathetic cotransmitter neuropeptide Y also promotes angiogenesis in ischemic skeletal muscles in rats.²¹ Interestingly, lumbar sympathectomy significantly attenuated neuropeptide Y release after hindlimb ischemia in rats further linking sympathetic innervation to angiogenesis.²¹ In addition to sympathetic innervation of the vasculature favorably contributing to angiogenesis, the sympathetic nervous system also regulates bone marrow (BM)-derived hematopoietic cell mobilization through stromal eNOS activation.¹⁸ These studies suggest that catecholamines at physiological concentrations that do not cause vasoconstriction, may provide a conducive environment for induction and maintenance of arterial endothelial cell identity, an effect important for angiogenesis and arteriogenesis after injury.

More broadly, how can these findings be exploited for future scientific exploration for clinically relevant diseases? First, decades of cardiac transplant studies indicate that these organs exhibit only partial re-innervation over time,²² an effect that may reflect their significantly impaired coronary endothelial responses, a key event in chronic transplant rejection.²³ Future studies investigating cell-based expression of α1- and α2-AR subtypes, sympathetic innervation, and the role of physiological levels of catecholamines may lead to a better understanding of transplant endothelial function and the survival of these precious grafts. Second, vascular graft failure from use of venous conduits contributes significantly to morbidity and mortality of patients receiving venous AV fistulas for hemodialysis, saphenous veins for coronary artery bypass graft surgery, and venous conduits for peripheral artery disease, among others. Might methods to increase sympathetic innervation and catecholaminergic stimulation improve the survival of these venous grafts as they are exposed to the harsh systemic circulation for the first time attempting to “arterialize” their venous vascular endothelium? Finally, emerging studies highlight the importance of both sympathetic and parasympathetic innervation for effective cardiac regeneration in young neonatal mice.^{24, 25} As the role of either of these nervous systems on cardiac angiogenesis were not specifically explored, future studies will be required to delineate if sympathetic or parasympathetic innervation is more dominantly important for angiogenesis in response to various forms of myocardial injury and, in particular, in older-aged mammals when regenerative capacity is diminished.

Taken together, the findings by Parfanuad et al. extend our understanding of the sympathetic nervous system for arterial vs. venous endothelial cell plasticity and provide insights for maintenance of endothelial arterial identity in adult. Future investigations in this field between the crosstalk of the sympathetic (and parasympathetic) nervous system on arterial, venous, and lymphatic endothelial cell fate may provide the rationale for designing new therapeutics for a range of ischemic cardiovascular disease states.