One specific area of interest in vascular signal transduc-tion has been the recent identification of receptor tyrosine kinases (RTKs), which coordinate signaling cascades integral to angiogenic phenomena. At least five transmembrane receptor kinases have been shown to possess an expression pattern that is predominantly endothelial, and these can be conceptually subdivided into two groups. One group, or subfamily, encodes the VEGF receptors, flt-1, flk-1, and flt-4, whereas the second group encodes the TEK and Tie receptors. Both groups share structural homology with extracellular Ig and cytoplasmic split-kinase motifs. Because expression patterns vary with developmental time and space, there has been considerable speculation as to whether the differential expression of each RTK is important in dictating when or how the vascular endothelial lineage is specified. Based on the very early and predominantly overlapping expression patterns of flk-1 compared with the subsequent onset of TEK and tie-1 expression during embryonic development, it has been suggested that these signaling pathways are necessary for establishing the embryonic vasculature. Analysis of dominant-negative mutations in the murine TEK gene reveal its essential role in fostering endothelial proliferation in vivo, but not in the derivation of the endothelial lineage from angioblastic precursors. When considering the recurrent rounds of neovascularization seen during normal development or accompanying adult-onset, ischemia-induced ocular angio-genesis, it is also quite likely that the regulated expression of the fibroblast, platelet-derived growth factor, as well as the vascular endothelial growth factor (VEGF) families are also of pivotal importance for endothelial as well as pericyte proliferation [3, 4].
Receptor Tyrosine Kinases: Endothelial-Pericyte Interactions?
Biochemical analyses and gene cloning studies have clearly demonstrated the importance of the fibroblast growth factor family in fostering endothelial growth and angiogen-esis. Although it has been established that bFGF (FGF-2) is a component of select subendothelial basement membranes in vivo (as well as a component of the endothelial extracellular matrix in vitro), its mode of release as well as its accumulation and function in the matrix remain somewhat equivocal. Because matrix (heparin)-bound FGF envelops populations of quiescent endothelial cells, others and we have suggested that it acts as a wound healing or survival agent, being released following tissue injury. However, if FGF-2 acts only in an autocrine or paracrine role to stimulate endothelial proliferation, then it is difficult or impossible to explain the exceedingly low rate of endothelial turnover observed in microvascular beds that possess FGF-2-enriched basement membranes, let alone its presence in the extracellular compartment despite any convincing evidence of local injury or cell death. These important observations have led us to postulate other critical roles for FGF signaling in vascular cells . Specifically, is it possible that the recruitment and/or local differentiation of retinal peri-cytes are regulated by endothelial-produced and matrix-associated growth regulators (i.e., FGF-2).
As has been well studied, the pericyte resides within the basement membrane in association with the microvascular endothelial cell. Although the pericytes' contribution to the basement membrane, per se, has not been critically established, the strategic location of this mural cell population within the basement membrane places it in a position capable of participating in unique heterocellular signaling, events that may be essential for their sustained growth and/or survival [6, 7]. Work in the field has demonstrated a role for PDGF in potentiating vascular smooth muscle and pericyte proliferation.Vascular endothelial cell (EC) production of PDGF is seemingly regulated by injury and repair, because quiescent EC populations fail to produce measurable PDGF levels in vitro. In vivo, PDGF and PDGF RTK mRNA have been localized within pericyte-deficient capillaries, but it is unknown whether such an autocrine signaling loop is responsible for stimulating endothelial cell cycle progression. Recently, work from our laboratory has revealed that FGF-2 acts as a principal regulator of pericyte growth during the development and/or remodeling of the microvasculature . These observations demonstrate a critical role for FGF signaling not only in microvascular development but in the pathological morphogenesis occurring during disease as well.
Modulation of pericyte contractile phenotype is by serine/threonine kinase signal transduction, TGFb-1, and the extracellular matrix. TGFb-1, which is a member of the transforming growth factor gene family and represents the prototype of this multimember family, coordinates a variety of activities, including the regulation of growth, differentiation, and extracellular matrix production. Its widespread expression during embryonic development points to the important role that TGFb-1 plays in orchestrating epithelial-mesenchymal interactions, angiogenesis, chrondro-, and osteogenesis. Only recently, through the targeted disruption of TGFb-1, have these essential role(s) been mapped. TGFb-1 signals through two transmembrane serine/threonine kinases, STKI and STKII, which function together and share about 40 percent homology with each other . Each STK possesses an extracellular cysteine-rich domain responsible for fostering ligand binding, a single transmembrane domain, and an intracellular serine/threo-nine kinase domain responsible for intracellular signaling. It has been demonstrated that binding of TGFb-1 to STKII causes receptor phosphorylation, allowing association and phosphorylation of STKI, therein forming a functional het-eromeric complex capable of downstream signaling. In vivo phosphorylation occurs within a stretch of five clustered threonines and serines (TTSGSGSG) within the GS domain, a conserved stretch of 30 amino acids that are unique to STKI. Mutations that either (1) inactivate the STKI kinase domain, (2) eliminate the STKI phosphorylation sites, or (3) alter the STKII cysteine-rich region involved in recognition of TGFb-1 all prevent STKI phosphorylation and TGFb-1 signal transduction. Elucidation of the downstream targets for STKI and STKII may unveil important new insights in TGFb-1 signaling during embryonic development, cellular differentiation, and the response to injury. Our recent discovery that TGFb-1 signaling may orchestrate the onset of pericyte differentiation as well as control pericyte extracellular matrix production not only indicates the putative importance of STK signaling during retinal microvascular development, but, in consideration of the known alterations that occur in retinal capillary permeability, basement membrane composition, and vascular cell growth/differentiation during diabetes, it seems evident that aberrations in TGFb-1 signal transduction may also be critically involved in regulating pericyte de-differentiation during disease onset [5, 8].
Despite recent in vitro demonstrations that pericytes can dramatically inhibit endothelial growth in a TGF-b-dependent pathway, little evidence indicates that the reverse is also true (i.e., that pericyte growth or contractile pheno-type can be modulated by endothelial-produced, matrix-associated growth regulators). Early work helped establish the important role that the endothelial-derived extracellular matrix plays in modulating retinal pericyte growth state and contractile phenotype [5, 9]. Most interesting and perhaps critical to our more current understanding of matrix dynamics and pericyte signal transduction is the demonstration that collagen IV-enriched, endothelial-synthesized extracellular matrix fosters pericyte proliferation. In these growth-potentiated pericyte cultures, smooth muscle contractile proteins cannot be detected [5, 9]. Similarly, when pericytes are stimulated to proliferate under the control of exoge-nously added FGF-2 and heparin (both of which are components of the endothelial-derived matrix), a-smooth muscle actin expression is also repressed (as cells proliferate).
Not only do these observations suggest that the growth state of the cells and their resultant contractile phenotype are inversely related, but the data also suggest the phenotypic similarity that exists between vascular smooth muscle and pericytes, both with respect to the molecular events controlling contractile phenotype as well as cell cycle progression. It is tempting to speculate that control of either pericyte growth or differentiation is the sum total of the positive and negative signals that originate in the vascular basement membrane, with each signaling pathway having as targets the regulators of cell cycle progression. In fact, reports in the literature state that stimulation of growth by FGF-2 is mediated by phosphorylation of the retinoblastoma susceptibility gene product, RB, which is thought to restrict cell cycle progression through late G1. It has also been demonstrated that TGFb-1 may inhibit cell proliferation by preventing RB phosphorylation. We presume that both of these important signaling pathways are in place in the microvas-culature (Figure 1). Furthermore, in consideration of the role that TGFb-1 signal transduction plays in fostering retinal pericyte differentiation (presumably via induction of smooth muscle gene transcription), it seems possible that other downstream targets are also affected. For example, TGFb-1 signaling in pericytes and/or endothelial cells could account for the excessive accumulation of type IV collagen in diabetic basement membranes. In turn, reduplication of the basement membrane may cause the loss of contact of endothelial cells from pericytes. Concomitantly, pericyte phenotypic modulation would ensue, resulting in de-differentiation, loss of smooth muscle contractile proteins, and a failure to restrain endothelial cells in their growth-arrested state. These de-repressed endothelial cells would then begin proliferating and retinal angiogenesis would ensue. In this model (Figure 1), we would postulate that pericyte pheno-typic modulation (de-differentiation), together with local alterations in the extracellular matrix (rather than pericyte death), serve as early signals that elicit the recurrent rounds of endothelial proliferation observed during diabetes-induced neovascularization. This hypothesis is supported by the observations that, in humans, only a modest decrement in pericyte number is seen in proliferative diabetic retinopa-thy (~20% to 30%). Our collective data would strongly suggest that the pericytes are present in the proliferative retinopathic vasculature, but these cells are neither fully contractile nor in intimate contact with their endothelial neighbors (Figure 1).
ß actin-mediated motility EC Growth Pericyte absence
Matrix remodeling Pericyte Recruitment
Pericyte/EC contact Myf-5, TGFb-1,aVSM-actin and Rho GTPase-mediated pericyte differentiation EC growth arrest
Matrix remodeling Loss of EC/pericyte contact b actin-mediated EC migration
Rho GTPase-mediated repression of pericyte aVSM actin
Pericyte loss/ 'dedifferentiation EC growth
ß actin-mediated motility EC Growth Pericyte absence
Matrix remodeling Pericyte Recruitment
Was this article helpful?