Introduction

Chemo Secrets From a Breast Cancer Survivor

Breast Cancer Survivors

Get Instant Access

Like many types of cancers, cutaneous and uveal melanoma could benefit tremendously from the development of valid predictors of the aggressive potential of these diseases in patients. Recent studies aimed at characterizing the molecular signature of melanoma tumor cells generated a classification scheme for malignant cutaneous melanoma (Bittner et al., 2000), in addition to a molecular profile for uveal melanoma (Seftor et al., 2002b). These molecular tools demonstrated how global transcript analysis could contribute to the classification of melanoma, which may hold great promise in improving diagnosis and treatment.

The microarray analysis studies of differential gene expression of highly aggressive vs poorly aggressive human cutaneous and uveal melanoma cell lines (Bittner et al., 2000; Seftor et al., 2002b) revealed the coexpression of multiple phenotype-spe-cific genes by the aggressive tumor cells, including those of endothelial, epithelial, fibroblastic, hematopoietic, kidney, neuronal, muscle, and several precursor cell types. These observations were quite intriguing and suggested that aggressive melanoma cells may undergo a genetic reversion to a pluripotent, embryonic-like phenotype. However, the biological significance of the unexpected expression of multiple molecular phenotypes by melanoma cells remains enigmatic. Indeed, these findings have prompted further investigation regarding the potential rel-

From: Stem Cells Handbook

Edited by: S. Sell © Humana Press Inc., Totowa, NJ

evance of a "plastic" tumor cell phenotype and challenge our current thinking of identifying and targeting tumor cells that may masquerade as other cell types.

27.1.1. TUMOR CELL PLASTICITY ALLOWS FOR VASCULOGENIC MIMICRY BY AGGRESSIVE MELANOMA Many of the biological properties germane to embryogenesis are also important in tumor growth. For example, during embryonic development, the formation of primary vascular networks occurs by the process of vasculogenesis—the in situ differentiation of mesodermal progenitor cells (angioblasts or hemangioblasts) to endothelial cells that organize into a primitive network (for a review, see Risau, 1997; Carmeliet, 2000) (Fig.1). The subsequent remodeling of the vasculogenic network into a more refined complex of vasculature occurs through angiogenesis—the sprouting of new capillaries from preexisting networks. Similarly, it is widely accepted that during cancer progression, tumors require a blood supply for growth (Folkman, 1995; Rak and Kerbel, 1996; Kumar and Fidler, 1998). Based on the molecular profile of the aggressive melanoma cells, together with novel in vitro observations and correlative histopathology findings, our laboratory and collaborators have introduced the concept of vasculogenic mimicry to describe the plasticity and unique ability of aggressive melanoma tumor cells to express endothelial-associated genes and form tubular structures and patterned networks in three-dimensional (3D) culture that mimic embryonic vasculogenic networks (Maniotis et al., 1999; Hendrix et al., 2001; Hess et al., 2001; Seftor et al., 2001). In animal and patient tumors, many of

Fig. 1. Overview of vasculogenesis and angiogenesis.

these networks are lined by tumor cells adjacent to channels, some containing red blood cells (RBCs) and plasma, and possibly providing a perfusion mechanism and dissemination route within the tumor compartment that functions either independently of or simultaneously with angiogenesis. Tumors exhibiting vasculo-genic mimicry have a poor prognosis (Maniotis et al., 1999); however, the biological relevance of this potential perfusion pathway remains to be elucidated (Hendrx et al., 2003).

A noteworthy example of melanoma tumor cell plasticity is illustrated in Fig. 2, which shows the coalescence of several spheroids of aggressive human cutaneous melanoma cells after 1 mo in suspension culture. A histological section through this spheroidal mass revealed that the outermost layer resembled a stratified squamouse-like epithelium, while the inner mass was organized like a mesoderm with evidence of vasculogenic-like networks. This seemingly simplistic observation is an excellent illustration of the phenotypic diversity exhibited by aggressive melanoma cells, which is greatly influenced by microenvironmental factors.

The molecular underpinnings of melanoma vasculogenic mimicry are beginning to unfold. Recent experimental evidence has shown the importance of several key molecules in the formation of vasculogenic-like networks by melanoma cells, including vascular endothelial (VE)-cadherin (CD144 or cadherin 5), EphA2 (epithelial cell kinase), and laminin (Hendrix et al., 2001; Hess et al., 2001; Seftor et al., 2001). These molecules (and their binding partners) are also essential in the formation and maintenance of blood vessels (Risau, 1997; Carmeliet, 2000). VE-cadherin is an adhesive protein, previously considered to be endothelial cell-specific, belonging to the cadherin family of transmembrane proteins promoting homotypic cell-to-cell interaction (Hynes, 1992, Kemler, 1992; Lampugnani et al., 1992; Gumbiner, 1996). EphA2 is a receptor protein tyrosine kinase that is part of a large family of ephrin receptors (Pasquale, 1997). Binding of EphA2 to its ligand ephrin-A1 results in the phosphorylation of EphA2; however, there is evidence indicating that EphA2 can also be constitutively phosphorylated in unstimulated cells (Rosenburg et al., 1997). Strong expression of EphA2 and ephrin-A1 has been associated with increased melanoma thickness and decreased survival, and the EphA2/ephrin-A1 pathway has been linked to tumor cell proliferation (Straume and Akslen,

Fig. 2. Example of melanoma tumor cell plasticity. (A) Microscopic low-power view of the coalescence of several large spheroids of aggressive human cutaneous C8161 melanoma cells after 1 mo in suspension culture; (B) a histological section through an area delineated by rectangle in (A), where outermost layer appears similar to stratified squamous epithelium and inner mass shows vasculogenic-like networks (arrows). Magnification: (A) x15; original magnification: (B) x40.

Fig. 2. Example of melanoma tumor cell plasticity. (A) Microscopic low-power view of the coalescence of several large spheroids of aggressive human cutaneous C8161 melanoma cells after 1 mo in suspension culture; (B) a histological section through an area delineated by rectangle in (A), where outermost layer appears similar to stratified squamous epithelium and inner mass shows vasculogenic-like networks (arrows). Magnification: (A) x15; original magnification: (B) x40.

2001). Laminins are major components of basement membranes and play an active role in neurite outgrowth, tumor metastasis, cell attachment and migration, and angiogenesis (Malinda and Kleinman, 1996; Colognato and Yurchenco, 2000; Straume and Akslen, 2001). Proteolytic cleavage of laminin, particularly the laminin 5 y2 chain, can alter and regulate the integrin-mediated migratory behavior of certain cells (Giannelli et al., 1997; Malinda et al., 1999; Koshikawa et al., 2000; Straume and Akslen, 2001), illustrating its potential importance as a molecular trigger in the microenvironment.

Fig. 3. Vasculogenic mimicry in aggressive human uveal and cutaneous melanoma. (A) Western blot analysis of VE-cadherin by aggressive* human cutaneous (C8161) and uveal (C918, MUM-2B), but not poorly aggressive, cutaneous (C81-61) and uveal (OCM-1A, MUM-2C) melanoma cells. (B,C) phase contrast microscopy of developing vasculogenic networks, at d 1 and 3, respectively, formed by aggressive MUM-2B cells in 3D culture. (D,E) immunohistochemical localization of laminin 5y2 chain (dark staining highlighting vasculogenic-like networks) in aggressive C8161 cells in 3D culture (D), similar to laminin staining (dark staining surrounding spheroidal nests of tumor cells) of network patterns in a patient tissue section with aggressive uveal melanoma. (F) poorly aggressive C81-61 cells cultured on a collagen I gel for 4 d (showing no vasculogenic networks) and on a gel preconditioned for 2 d (G) by aggressive C8161 cells, which were subsequently removed before addition of poorly aggressive C81-61 cells—which now form vasculogenic-like networks and express VE-cadherin, EphA2, and laminin 5y2 chain.

Fig. 3. Vasculogenic mimicry in aggressive human uveal and cutaneous melanoma. (A) Western blot analysis of VE-cadherin by aggressive* human cutaneous (C8161) and uveal (C918, MUM-2B), but not poorly aggressive, cutaneous (C81-61) and uveal (OCM-1A, MUM-2C) melanoma cells. (B,C) phase contrast microscopy of developing vasculogenic networks, at d 1 and 3, respectively, formed by aggressive MUM-2B cells in 3D culture. (D,E) immunohistochemical localization of laminin 5y2 chain (dark staining highlighting vasculogenic-like networks) in aggressive C8161 cells in 3D culture (D), similar to laminin staining (dark staining surrounding spheroidal nests of tumor cells) of network patterns in a patient tissue section with aggressive uveal melanoma. (F) poorly aggressive C81-61 cells cultured on a collagen I gel for 4 d (showing no vasculogenic networks) and on a gel preconditioned for 2 d (G) by aggressive C8161 cells, which were subsequently removed before addition of poorly aggressive C81-61 cells—which now form vasculogenic-like networks and express VE-cadherin, EphA2, and laminin 5y2 chain.

The importance of VE-cadherin, EphA2, and laminin 5 y2 chain is summarized in Fig. 3. Molecular analyses revealed that these three molecules, among others, were dramatically overexpressed in aggressive human cutaneous and uveal melanoma cells (Bittner et al., 2000; Hendrix et al., 2001; Hess et al., 2001; Seftor et al., 2001, 2002b). At the protein level, these molecules were expressed only by aggressive tumor cells, not by poorly aggressive melanoma cells. The biological relevance of these molecules in vasculogenic mimicry was further demonstrated by downregu-lating their expression independently and measuring the consequences on their ability to form vasculo-genic-like networks. This approach revealed that downregulation of VE-cadherin, EphA2, or laminin 5 y2 chain resulted in the complete inability of aggressive melanoma cells to engage in the formation of vasculogenic-like networks in 3D culture (Hendrix et al., 2001; Hess et al., 2001; Seftor et al., 2001). Additional experiments focused on tumor cell-

associated extracellular matrix (ECM) led to the discovery of the inductive potential of the microenvironment preconditioned by aggressive melanoma cells and shown to influence poorly aggressive melanoma cells to assume a vasculogenic phenotype (Seftor et al., 2001). These data, along with subsequent observations of vascular-associated gene expression (summarized in Fig. 3G), suggest that highly aggressive melanoma cells deposit molecular signals in their environment that have the potential to induce a vasculogenic phenotype in poorly aggressive cells, with respect to the formation of vasculogenic-like networks and the concomitant expression of vascular-associated genes (VE-cadherin, EphA2, and laminin 5 y2 chain).

Based on the intriguing immunohistochemical colocalization pattern of VE-cadherin and EphA2, together with immunopre-cipitation data, we have proposed a novel hypothetical model for signaling during vasculogenic mimicry (Seftor et al., 2002). This model highlights the cooperative interactions of VE-cadherin and EphA2 and suggests downstream signaling involving phosphatidylinositol-3 kinase and focal adhesion kinase, initiated by ephrin-Al binding. These molecules are the focus of ongoing investigations to elucidate important signaling pathways involved in vasculogenic mimicry. Collectively, these findings may provide the basis of new therapeutic and diagnostic methods for cancer detection and mangement.

27.1.2. BIOLOGICAL RELEVANCE OF VASCULOGENIC PHENOTYPE IN VIVO Studies addressing the biological significance of the melanoma vasculogenic phenotype in animal models have focused on three major questions: (1) Is there a fundamental difference in the vascularization of highly aggressive vs poorly aggressive tumor cell masses? (2) Can a morphological connection be identified between tumor cell-lined vascular channels within aggressive tumors and endothelial-lined vasculature? (3) Is it possible for aggressive tumor cells to provide a vascular function when challenged to an ischemic environment?

To address the first question regarding potential differences in the vascularization of highly vs poorly aggressive melanoma cell masses, we utilized a dorsal skinfold window chamber in a nude mouse model. Into the chambers were placed either aggressive human cutaneous melanoma cells (in 3D collagen gels) or poorly aggressive human cutaneous melanoma cells under similar conditions, and the data were recorded at various time intervals over 2 wk with videomicroscopy and histological analyses. As summarized in Fig. 4, the aggressive melanoma 3D culture showed the presence of RBCs and plasma in tumor-lined channels and networks—in the absence of angiogenesis within the tumor compartment. At the periphery of the tumor mass in the stromal compartment, blood vessels were evident, suggesting a possible connection between tumor cell-lined channels and endothelial-lined vasculature. By contrast, the poorly aggressive cutaneous melanoma cells in 3D collagen gels induced an angiogenic response in the mouse model. These data demonstrate a fundamental difference in the vascularization of highly aggressive vs poorly aggressive melanoma tumor cell masses.

The second question sought to identify a potential biological connection between tumor-lined vascular channels within aggressive melanoma tumors and endothelial-lined vasculature. Using a nude mouse model injected subcutaneously with red fluorescent protein (RFP)-labeled aggressive human cutaneous melanoma cells, primary tumors were allowed to develop over 42 d. As summarized in Fig. 5, a combination of routine histological analysis and microsphere perfusion of the mouse vasculature and confocal microscopy with 3D reconstruction revealed the presence of microspheres (derived from the perfused vasculature) within the human tumor compartment. The endothelial-lined mouse vasculature was seen at the periphery of the human tumor, where fluorescein isothiocyanate (FITC)-labeled vasculature overlapped with RFP-labeled tumor cells forming vasculo-genic-like networks and channels. The aggressive human tumors demonstrated vasculogenic mimicry and no necrosis within the tumor compartment which coalesced with angiogenic mouse vessels at the human-mouse interface. These data are the first demonstration of a physiological connection between tumor cell line vascular channels in aggressive melanoma tumors and mouse endothelial-lined vasculature. However, Shirakawa et al. (2002) have recently used dynamic micromagnetic resonance angiogra-phy analysis and laser-capture microdissection to demonstrate a similar vascular connection between vasculogenic mimicry within inflammatory breast cancer xenografts and mouse angio-genic/host vessels at the tumor margin. They used the strategy of laser-capture microdissection to document the expression of endothelial-associated genes by breast cancer cells within the inflammatory breast xenograft that exhibited no central necrosis, no endothelial cells, and no fibrosis.

The third question explored the plasticity of aggressive melanoma tumor cells by challenging them to an ischemic environment and assessing whether the cells would participate in neovascularization and/or form a tumor. Hendrix et al. (2002), injected intramuscularly Ad-RFP (RFP expressing adenovirus)-or enhanced green fluorescent protein (EGFP)-labeled human cutaneous metastatic melanoma cells into the hind limbs of nude mice with surgically induced ischemia (resulting in blood flow <15% of normal). Highlights of this study are presented in Fig. 6. Five days postischemia, the presence of microspheres in limb vasculature (perfused via the aorta) demonstrated reperfusion of the limb as demonstrated by the presence of perfused beads within the vasculature. Mice were perfused via the aorta with FITC-labeled Bandeira simplicifolia lectin B4 (BSLB4) and Ulex europaeus agglutinin to visualize the luminal side of the vascular wall, and confocal microscopy revealed Ad-RFP-labeled melanoma cells adjacent to the luminal label. To exclude the possibility that Ad-vector targets murine vessels, similar experiments were performed using metastatic melanoma cells stably trans-fected with EGFP. Five days postischemia, histological cross-sections of muscular tissue stained with mouse endothelial cell-specific BSLB4 (followed by streptavidin-conjugated Alexa dye 594) showed human melanoma cells adjacent to and overlapping with mouse endothelial cells in a linear arrangement. When poorly aggressive EGFP-labeled melanoma cells were introduced at the site of ischemia in separate experiments, they were not found 5 d postischemia. This evidence demonstrated the powerful influence of the microenvironment on the transendothelial differentiation of malignant melanoma cells needed for neovas-cularization and reperfusion of ischemic limbs.

In our investigation of ischemic limb reperfusion, we chose to examine selected Notch proteins known to promote the differentiation of endothelial cells into vascular networks (Gridley, 2001; Uyttendaele et al., 2001). By immunohistochemistry, Notch 3 and Notch 4 were strongly expressed in the malignant melanoma cells, but not in the nonspecific antibody control nor in the poorly aggressive melanoma cells. Notch signaling molecules are inte-

Fig. 4. Microscopic comparison of vasculogenic vs angiogenic human cutaneous melanoma tumor xenografts (stained with hematoxylin and eosin [H&E]) in 3D collagen gels in dorsal skinfold window chambers in nude mice for 10 d. Tumor cell-lined vascular channels are prominent in lower magnification (A) and higher magnification (B,C) views of C8161 human aggressive melanoma cells in 3D collagen gels in dorsal skinfold window chambers. RBCs (arrows) are seen in some vascular channels (B), and they appear to originate from vasculature in the stromal/ muscular region (C) underlying the tumor mass. (D) Poorly aggressive human cutaneous A375P melanoma cells in 3D collagen gels in a dorsal skinfold window chamber demonstrate angiogenesis and tissue necrosis associated with the stromal/muscular compartment. Angiogenic neovascularization is prevalent (higher-magnification insets) within the tumor mass. Original magnifications: (A) x20; (B,C) x63; (D) x10.

Fig. 4. Microscopic comparison of vasculogenic vs angiogenic human cutaneous melanoma tumor xenografts (stained with hematoxylin and eosin [H&E]) in 3D collagen gels in dorsal skinfold window chambers in nude mice for 10 d. Tumor cell-lined vascular channels are prominent in lower magnification (A) and higher magnification (B,C) views of C8161 human aggressive melanoma cells in 3D collagen gels in dorsal skinfold window chambers. RBCs (arrows) are seen in some vascular channels (B), and they appear to originate from vasculature in the stromal/ muscular region (C) underlying the tumor mass. (D) Poorly aggressive human cutaneous A375P melanoma cells in 3D collagen gels in a dorsal skinfold window chamber demonstrate angiogenesis and tissue necrosis associated with the stromal/muscular compartment. Angiogenic neovascularization is prevalent (higher-magnification insets) within the tumor mass. Original magnifications: (A) x20; (B,C) x63; (D) x10.

Fig. 5. Morphological assessment of the relationship between tumor cell-lined vascular channels within aggressive human cutaneous melanoma tumors (42 d after sc administration in nude mice) and endothelial-lined vasculature. (A) Confocal microscopy of mouse vasculature perfused with F-BSLB4 and U. europaeus agglutinin delineating mouse vessels (green) showing red fluorescent microspheres (arrows) delivered from mouse vasculature into the human melanoma tumor interface. (B) H&E-stained histological section reveals an area (outlined by the rectangle) containing the mouse vasculature: melanoma interface at the tumor periphery, similar to that shown in (A). (C) H&E-stained histological section within the central region of the melanoma tumor shows melanoma cell-lined vascular channels (arrow). (D) Confocal microscopy shows the perfused mouse endothelial-lined vasculature (green) at the periphery (rectangle) of the melanoma tumor (containing RFP-labeled melanoma cells), and (E) overlaps with the RFP-labeled melanoma cells. Original magnifications: (A) x20; (B,C,E) x40; (D) x10.

Fig. 5. Morphological assessment of the relationship between tumor cell-lined vascular channels within aggressive human cutaneous melanoma tumors (42 d after sc administration in nude mice) and endothelial-lined vasculature. (A) Confocal microscopy of mouse vasculature perfused with F-BSLB4 and U. europaeus agglutinin delineating mouse vessels (green) showing red fluorescent microspheres (arrows) delivered from mouse vasculature into the human melanoma tumor interface. (B) H&E-stained histological section reveals an area (outlined by the rectangle) containing the mouse vasculature: melanoma interface at the tumor periphery, similar to that shown in (A). (C) H&E-stained histological section within the central region of the melanoma tumor shows melanoma cell-lined vascular channels (arrow). (D) Confocal microscopy shows the perfused mouse endothelial-lined vasculature (green) at the periphery (rectangle) of the melanoma tumor (containing RFP-labeled melanoma cells), and (E) overlaps with the RFP-labeled melanoma cells. Original magnifications: (A) x20; (B,C,E) x40; (D) x10.

Fig. 6. Microscopic analyses of revascularization of nude mouse ischemic hind limbs 5 (A,B) and 20 (C,D) d after surgical severing of femoral arterial branches followed by inoculation with 1 x 103 Ad-RFP-labeled human metastatic cutaneous melanoma cells (C8161) intramuscularly. (A) Confocal microscopy of mouse neovasculature at 5 d postischemia perfused with F-BSLB4 and U. europaeus agglutinin delineating mouse vessels (green) and Ad-RFP-labeled tumor cells (red) with branched neovasculature containing tumor cells; (B) immunohistochemical localization of Notch 3 protein expression in C8161 melanoma cells in reperfused ischemic muscle 5 d postischemia; (C) cross-section of reperfused ischemic hind limb at 20 d postischemia showing brown immunostaining (for cytokeratins 8 and 18) of one C8161 melanoma cell external to mouse vasculature; (D) microscopic view of H&E-stained tumor (section) formed intramuscularly by C8161 metastatic melanoma cells at 20 d postischemia. Original magnifications: (A,B,D) x40; (C) x63.

Fig. 6. Microscopic analyses of revascularization of nude mouse ischemic hind limbs 5 (A,B) and 20 (C,D) d after surgical severing of femoral arterial branches followed by inoculation with 1 x 103 Ad-RFP-labeled human metastatic cutaneous melanoma cells (C8161) intramuscularly. (A) Confocal microscopy of mouse neovasculature at 5 d postischemia perfused with F-BSLB4 and U. europaeus agglutinin delineating mouse vessels (green) and Ad-RFP-labeled tumor cells (red) with branched neovasculature containing tumor cells; (B) immunohistochemical localization of Notch 3 protein expression in C8161 melanoma cells in reperfused ischemic muscle 5 d postischemia; (C) cross-section of reperfused ischemic hind limb at 20 d postischemia showing brown immunostaining (for cytokeratins 8 and 18) of one C8161 melanoma cell external to mouse vasculature; (D) microscopic view of H&E-stained tumor (section) formed intramuscularly by C8161 metastatic melanoma cells at 20 d postischemia. Original magnifications: (A,B,D) x40; (C) x63.

Table 1

Stem Cell-Associated Genes in Melanomaa

Table 1

Stem Cell-Associated Genes in Melanomaa

Gene name

Unigene

Function

Ratio

Stem cell factor receptor c-kit (KIT)

Hs.81665

Tyrosine kinase receptor; oncogene

Was this article helpful?

0 0
How To Prevent Skin Cancer

How To Prevent Skin Cancer

Complete Guide to Preventing Skin Cancer. We all know enough to fear the name, just as we do the words tumor and malignant. But apart from that, most of us know very little at all about cancer, especially skin cancer in itself. If I were to ask you to tell me about skin cancer right now, what would you say? Apart from the fact that its a cancer on the skin, that is.

Get My Free Ebook


Post a comment