Building better blood vessels: new insight on the molecular control of arteriogenesis

See article by Boengler et al. [2] (pages 573–581) in this issue. Our understanding of the initial steps in blood vessel formation has been advanced through studies employing transgenic and knock-out mouse models [1] as well as isolated cells cultivated under conditions of altered oxygen level and shear stress. From the standpoint of identifying basic biological controls, examination of the tumor vascularization process also has provided insight concerning the molecular pathways required to form and sustain the microvasculature. However, much remains to be discovered about the far slower, adaptive process of arteriogenesis whereby pre-existing arterioles adopt new structural features required for sustaining blood flow around vascular obstructions. Especially fascinating are molecular events that occur in the vicinity of obstruction points where oxygen levels are near normal but flow rates significantly altered. An interesting report by Boengler and colleagues, appearing in this issue of Cardiovascular Research[2], has helped in this regard by identifying links between biomechanical stress and expression of specific transcription factors in the arterial wall. Importantly, two of these factors, the cardiac ankyrin repeat protein (CARP) and Egr-1, previously have been implicated in cardiovascular diseases associated with both vascular and myocardial tissue remodeling [1,3]. Arteriogenesis is proposed to consist of two phases [4,5]. The initiating event appears to be endothelial cell damage produced by high fluid shear forces generated through pre-existing collaterals that by-pass flow around an obstructed vessel segment. Biomechanical stress on the tunica intima within by-pass vessels promotes inflammatory cell adhesion to activated endothelial cells with subsequent release of monocyte chemotactic and survival factors such as MCP-1 and GM-CSF [6]. Once inflammation begins, a second remodeling phase follows driven by unknown agents but seemingly down-stream of activated TGFβ1 that accumulates within the tunica media through the actions of continuously infiltrating monocytes and macrophages. Developmentally, TGFβ1 has been associated with perivascular cell recruitment and neovessel stabilization through coordinated regulation of genes encoding smooth muscle contractile proteins such as vascular smooth muscle α-actin and extracellular matrix proteins such as type I collagen [1]. Mice defective for TGFβ1 have numerous deficiencies associated with poor host defense responses and microvessel wall weakness. Importantly, TGFβ1 elicited a strong, local arteriogenic response within the rabbit hindlimb collateral circulation when infused directly into the obstructed femoral artery [4]. Boengler et al. [2] extend these observations by showing that local infusion of activated TGFβ1 into the obstructed hindlimb circulation evokes more than a three-fold increase in CARP mRNA expression within a 24 h period. What does elevated expression of CARP in the collateral bed tell us about the molecular control of arteriogenesis? Currently, our information on the biochemical properties of this nuclear protein comes primarily from the cardiac morphogenesis literature where CARP has been identified as a negative regulator of ventricular-specific myosin light chain-2 (MLC-2v) gene transcription [3]. CARP-mediated transcriptional repression in the developing mouse heart appears to be based on its ability to sequester the YB-1 cold shock domain protein (also known as MSY1) required for MLC-2v promoter activation. Additionally, CARP is a down-stream target of the mammalian Nkx 2.5 gene required for specification of the ventricular chambers and cardiac morphogenesis. Nkx 2.5 knock-out mice do not express normal levels of CARP, an observation that may explain why these mice have a lethal defect linked to the failure of the heart tube to undergo looping morphogenesis [3]. CARP controls YB-1/MSY1 interaction with a region of the MLC-2v promoter that specifies positional information along the anterior-to-posterior gradient required for ventricular chamber morphogenesis. In view of this proposed function, it is tempting to speculate that spatial information also might be relayed by local CARP expression in remodeling collateral arteries. These signaling ‘landmarks’ may demarcate points of altered gene transcription in the vessel wall necessary for incremental structural changes in arterial diameter and/or vessel trajectory. By no means a passive bystander, the YB-1/MSY1 protein partner of CARP also may contribute to arterial remodeling given its demonstrated ability to suppress transcription of the vascular smooth muscle α-actin gene in aortic smooth muscle cells [7] and its unique perivascular localization in the remodeling coronary vasculature of accepted cardiac allografts [8]. Silencing of the smooth muscle α-actin gene may represent one aspect of arterial remodeling that depends on YB-1/MSY1, a proven CARP protein partner in the heart, to positionally titrate α-actin levels within the vessel wall thus permitting local changes in smooth muscle cell migration. Several investigative groups have demonstrated inverse relationships between smooth muscle α-actin expression, cellular phenotypic modulation, and migration [9]. While CARP typically is considered as a cardiac muscle-specific protein [3], there certainly is nothing ‘fishy’ about arterial CARP as described by Boengler et al. [2]. A highly homologous protein referred to as c193 previously was identified in human endothelial cells where its level of expression was markedly enhanced by pro-inflammatory lipopolysaccharide and TNFα [10]. In this context, it is not unreasonable to imagine that biomechanical stress, with attendant inflammatory response of the collateral endothelium, could trigger expression of vascular CARP-like proteins. Subsequent formation of CARP:YB-1/MSY1 protein complexes may then function as positional agents for guiding local changes in smooth muscle gene transcription required for vessel turning, enlargement, and increased blood flow around an obstruction. Finally, the significance of Egr-1 induction as a consequence of CARP over-expression described by Boengler et al. [2] should not be overlooked in the emerging mechanistic model for arteriogenesis. Egr-1 expression has been closely linked with arteriosclerosis and may assist in arteriogenesis by amplifying CARP-mediated gene reprogramming events in the collateral bed. While Egr-1 appears to influence smooth muscle cell phenotype, it also can augment expression of matrix remodeling proteases such as uPA required to clear connective tissue from the perivascular space and allow vessel enlargement. While not examined by Boengler et al. [2], a connection between mechanical stretch, up-regulation of the Ras-Raf-Mek-Erk Map kinase signaling pathway, and increased Egr-1 expression has been discussed in the literature [11]. Currently, it remains to be determined if Egr-1 levels are elevated in parallel with CARP in the intact collateral artery wall and whether this process requires an Erk intermediary. The studies performed by Boengler et al. [2] utilizing COS-1 cell over-expression of CARP certainly are informative but will require further evaluation in animal models or stretched arterial segments in the presence and absence of various MAP kinase inhibitors. The identification of novel gene regulatory proteins associated with the process of arteriogenesis represents a new advance in our understanding of the molecular regulation of collateral formation. These initial observations no doubt will provide the basis for many more exciting and informative physiological and biochemical studies.