**1. Introduction**

Obstructive disease, i.e., stroke and atherosclerosis of the coronary and peripheral arteries, is the leading cause of morbidity and mortality. Collateral circulation, which is composed of anastomotic vessels called collaterals, is the most important system capable of mitigating the effects of obstructive disease [1–5]. However, the number and diameter of collaterals in tissues vary greatly among individuals for reasons that are only beginning to be understood [1]. Moreover, little is known about the basic biology of collaterals, in part because of their small diameter and low density in tissues, difficulty in distinguishing them from other vessels, and lack of methodologies allowing study of their endothelial cells (ECs) and smooth muscle cells (SMCs) in cell culture. It is known, however, that collaterals are unique with regard to their: mechanism of formation during development (collaterogenesis) and its high sensitivity to genetic background-dependent variation, anatomic location in the circulation, hemodynamic forces acting on their ECs and SMCs, hallmark tortuosity, high susceptibility to risk factor-induced rarefaction, robust shear stress-dependent outward remodeling, and protective function in ischemic disease [1–5].

In mice, collaterals form late during gestation by the sprouting of ephrin-B2<sup>+</sup> ECs o ff of a small number of distal arterioles that subsequently undergo a tip cell-led proliferation, migration, and lumenization process to establish anastomoses between adjacent arterial trees [6–8]. This process varies greatly due to naturally-occurring polymorphisms in certain genes in the collaterogenesis pathway, resulting in wide di fferences in collateral extent and thus collateral-dependent blood flow and tissue injury in experimental models of occlusive arterial disease [9–11]. Collateral blood flow also varies widely in humans su ffering stroke [3,4], coronary, and peripheral artery diseases [12,13], with recent evidence supporting involvement of the same key variant gene identified in the mouse [14]. Since collaterals interconnect adjacent arterial trees, there is little or no pressure drop across them in the absence of obstructive disease. Thus, blood flow within collaterals is near zero, oscillating slowly toward one or the other tree that they anastomose [6,15]. Endothelial cells and SMCs that compose the collateral wall are therefore continuously exposed to low and disturbed shear stress, high circumferential wall stress, and low blood oxygen content. Accompanying this adverse hemodynamic environment, which favors vascular inflammation and endothelial dysfunction elsewhere in the circulation, collaterals in mice undergo a progressive decline in number and a loss of diameter with aging, presence of vascular risk factors, and in models of Alzheimer's disease [16–21]. Supportive evidence, based on measurement of collateral-dependent flow induced by acute ischemic stroke, has been reported in humans with aging and presence of vascular risk factors [22–24]. This so call "collateral rarefaction" has been linked in mice to chronic low-level inflammation, endothelial dysfunction, and increased proliferation of collateral ECs and SMCs [16–20,25]. Collaterals are also capable of undergoing robust anatomic outward remodeling in steno-occlusive disease, compared to similarly sized arterioles [2,5,9,10]. Additionally, di fferent from arterioles in the trees that they interconnect, collaterals lack myogenic responsiveness and have less SMC tone at baseline [26,27].

Despite these important unique features of collaterals, no studies have examined cellular and molecular aspects of their endothelial and smooth muscle cells to determine whether they di ffer from those of similarly sized arterioles. For example, blood flow in arteries and arterioles of healthy tissues, and thus fluid shear stress experienced by their ECs, is laminar, orthograde, and high velocity, with the exception of the inner curvature of the aortic arch and sites immediately downstream of arterial bifurcations where flow varies from high-velocity and orthograde to low-velocity with transient flow reversals in a fraction of the fluid laminae during each cardiac cycle (i.e., "disturbed" flow) [28–31]. Endothelial cells at these sites of disturbed shear stress are cobblestone in shape, as opposed to elsewhere where they are elongated and aligned with the vessel axis. They also evidence higher levels of proliferation, apoptosis, permeability, lipid uptake, oxidative stress, and markers of inflammation and aging, i.e., displaying the well-known pro-atherogenic EC phenotype specific to these sites. Flow and shear stress in collaterals in the absence of obstruction share similarly disturbed conditions [6,15]. Yet this is the normal environment in which collateral ECs and SMCs reside.

The purpose of this study was to examine the hypothesis that collateral ECs and SMCs express unique morphological and functional phenotypes that serve as adaptations or specializations to maintain the integrity of the collateral wall and balance against collateral rarefaction favored by the low and disturbed shear stress, high wall stress, and low blood oxygen levels—the latter caused by low flow-induced hemoglobin unloading to tissue. Such di fferences, if present, could o ffer insights into how collaterals are able to persist in healthy individuals but also why they are so susceptible to rarefaction with aging and other vascular risk factors. In addition, knowledge concerning the molecular features of collateral wall cells is fundamental to better understand how collaterals form during development, remodel in obstructive disease, and undergo risk factor-induced rarefaction, and to investigate possible treatments to augmen<sup>t</sup> or intervene with these processes.
