Melanoma Progression in the Brain: Role of Pericytes, the Basal Lamina, and Endothelial Cells in Tumor Vascularization

Introduction

Melanoma is a malignancy that originates primarily in the skin ( ), although melanomas can also arise from melanocytes in other pigmented tissues such as the uvea of the eye. Although there have been many improvements in the diagnosis, excision, and treatment of primary melanomas, the lethality of these tumors stems from their aggressive metastasis to secondary sites such as brain, bone, and lung ( ). Melanoma progression and metastasis are regulated both by intrinsic, tumor cell-autonomous factors, and by extrinsic factors in the tumor microenvironment ( ). Intrinsic factors include mutations that affect tumor suppressor mechanisms, growth factor receptors, and growth factor production in transformed cells. Extrinsic factors include elements of normal host stroma such as the vasculature, fibroblasts, myeloid cells, and extracellular matrix (ECM) ( ; ). It seems likely that improved melanoma therapy will require better understanding of both intrinsic and extrinsic mechanisms of tumor progression.

Tumor vascularization is one of the most widely recognized stromal influences on the progression of melanomas and other types of tumors. Accordingly, tumor blood vessels have become attractive targets for cancer therapy. Although most therapeutic efforts have been focused on the vascular endothelium ( ; ), it should also be possible to attack the tumor vasculature at other levels. Similar to other types of microvessels, tumor blood vessels are composed of three main elements: pericytes, endothelial cells, and the vascular basal lamina (basement membrane) ( ). These components are highly interactive and communicate extensively with each other to determine vessel development, maturation, and function ( ). Thus, uncoupling pericyte-endothelial-basal lamina interaction at any level may have the potential to disrupt tumor vascularization. As part of our effort to understand the respective contributions of pericytes, endothelial cells, and the basal lamina to vascular morphogenesis and function, we have used a model of B16F10 melanoma microinjection into the mouse brain. By using mice with genetic modifications that affect different elements of tumor microvessels, we have attempted to identify molecular aspects of vascular interaction that may be susceptible to therapeutic attack. In addition, tumors in this location may provide useful information regarding the behavior of metastatic melanomas in the brain.

Role of pericytes in melanoma vascularization in the brain

Throughout the history of vascular biology, studies focused on endothelial cells have outnumbered studies on pericytes by a wide margin. Nevertheless, it has become clear that pericytes are critical, early participants in the neovascularization process, and that the pericyte–endothelial cell relationship persists throughout the life of microvessels ( ; ). Pericytes ensheath the outer surfaces of endothelial cells that form the vascular lumen, promoting vascular development and maturation via their interactions with the endothelial cells ( ; ) ( Figure 12.1 ). Pericytes expressing platelet-derived growth factor receptor-β (PDGFR-β) are recruited to blood vessels by PDGF-B secreted by endothelial cells ( ). The importance of the PDGF/PDGFR signaling pathway in pericyte biology is comparable to the importance of the VEGF/VEGFR signaling pathway in endothelial cell biology. Mice deficient in expression of PDGFR-β or PDGF-B exhibit severely decreased pericyte coverage of endothelial cells, leading to increased microvascular leakage ( ). Depletion of pericytes via inhibition of the PDGF-B/PDGFR-β signaling pathway leads to enlargement of tumor vessels in a mouse model of pancreatic cancer, accompanied by greatly increased endothelial cell death ( ).

Along with PDGFR-β, we have found that the NG2 proteoglycan is another important constituent of the pericyte cell surface. In addition to its utility as a reliable pericyte marker ( ), NG2 is also functionally important for the recruitment of pericytes to blood vessels ( ) and for pericyte interaction with endothelial cells ( ). Our development of an NG2-null mouse ( ) has allowed us to explore NG2 function by determining the negative effects of NG2 ablation on the development of several types of cell populations, including pericytes, oligodendrocyte progenitor cells, keratinocytes, macrophages, and both brown and white adipocytes.

In extending our studies to the role of pericytes in melanoma vascularization in the brain, we have benefited from the fact that the density of pericyte ensheathment of endothelial cells is higher in the central nervous system (CNS) than in other organs ( ). Pericytes in CNS microvessels play an essential role in the structural integrity of vessels, including the properties of the blood–brain barrier ( ), thus magnifying the functional importance of pericytes in vessel function ( Figure 12.1 ). When we microinjected B16F10 melanoma cells (C57Bl/6 origin) into the corpus callosum of wild-type and NG2-null C57Bl/6 mice, we found that tumor progression was reduced roughly threefold by the ablation of NG2 ( ). Because B16F10 cells do not express NG2, the observed difference in tumor growth in the two mouse lines must be due to the presence versus absence of NG2 in the respective host microenvironments. Pericytes, macrophages, and oligodendrocyte progenitors are the host cell populations that normally express NG2 in the brain. Based on our previous finding of pericyte deficits in the vasculature of the NG2-null mouse ( ), we initially focused our attention on the pericyte/vascular compartment of the melanoma stroma.

Comparison of microvessels in the tumors from wild-type versus NG2-null mice revealed that pericyte recruitment was unaffected by NG2 ablation, but that pericyte ensheathment of endothelial cells was reduced twofold in the absence of NG2 ( ). This change in pericyte–endothelial cell interaction resulted in developmental deficits in both vascular cell populations, as well as in the basal lamina. In melanoma vessels in NG2-null mice, maturation of pericytes was retarded roughly 10-fold, as judged by acquisition of the maturation marker α-smooth muscle actin (α-SMA). In addition, the ability of pericytes and endothelial cells to assemble the vascular basement membrane was impaired by a factor of 3 in NG2-null melanoma tumor vessels, as measured by quantifying the deposition of collagen IV, the major collagen species of the basal lamina. From a functional standpoint, tumor vessel patency was reduced twofold in the NG2-null mouse, whereas vessel leakiness was increased by a factor of 4 ( ). These deficits in vessel function led to a 20-fold increase in intratumoral hypoxia in the NG2-null mouse. We have observed a very similar spectrum of vascular deficits in mammary tumors growing in NG2-null mice ( ), accompanied by retarded progression of mammary tumors in the NG2-null environment. These findings demonstrate the importance of NG2 in pericyte biology, and emphasize the extremely tight nature of the interactions that exist between pericytes, endothelial cells, and the basal lamina. In addition, these results suggest NG2 as a possible candidate for targeting pericyte function as a means of disrupting tumor vascularization.