How Cancer Cells Respond To A Changing Tumor Microenvironment

3.1. Osteomimicry

Cancer cells are capable of mimicking the characteristics of cells in the tumor microenvironment. Dramatic examples include osteomimicry, the ability of cancer cells to express genes normally highly restricted to bone cells before, during, or after metastasis through the synthesis, secretion, and accumulation of bone-like proteins, such as osteocalcin (OC), osteopontin, bone sialoprotein (BSP), and osteonectin, even forming mineralizing bone under certain culture conditions (29,52,53). Cancer cells are also capable of expressing receptor activator of NF-kB (RANK) ligand and parathyroid hormone related peptide, which are known to directly or indirectly increase bone turnover through increased RANK ligand (associated with cancer cells)-RANK (associated with osteoclasts) interaction and activation of osteoclastogenesis (54,55). These unusual characteristics are caused by the ability of cancer cells to respond to factors secreted by cancer cells or by host cells in the immediate microenvironment. Using human prostate cancer and bone cells as models to define the molecular basis of osteomimicry, we observed a unique key switch controlling OC and BSP gene expression that is operative in prostate cancer but not in bone cells (56,57). This switch resides at the 8-base nucleotides sequences, called the cyclic AMP (cAMP)-responsive element (CRE), within both OC and BSP promoters, and is responsible for the regulation of both endogenous OC and BSP as well as their promoter activities (57,58). We further showed that CRE activation is under the control of a soluble factor secreted by prostate cancer and host cells, with its action mediated by cAMP-dependent protein kinase A (PKA) activation. The activation of CRE binding protein (CREB), was demonstrated by the observation of CREB phosphorylation after activation of cAMP-PKA, phosphorylated CREB translocation into the cell nucleus, and the subsequent binding of CREB to CRE as shown by gel shift and supershift assays (57). Based on these data, Fig. 2 depicts a number of possible molecular pathways mediating osteomimicry in human prostate cancer cells (58):

1. The binding of a soluble factor to a putative cell surface receptor, linking with the activation of intracellular cAMP-PKA signaling pathway. The putative receptor can be a G protein-coupled receptor that mediates downstream signaling via PKA, or, alternatively, can be linked to MHC class 1 antigen or a yet-to-be-identified receptor that binds to the soluble factor and transmits intracellular signaling through the cAMP-PKA system.

2. The participation by a soluble factor, P2 microglobulin (P2M), which was shown to activate the cAMP-PKA system (58). Because P2M is known to complex with a classic MHC class 1 antigen, this could implicate the role of this complex in the downstream intracellular signaling of osteomimicry in human prostate cancer cells.

3. The direct action of P2M in activating the cAMP-PKA system, or the direct participation by P2M in CREB downstream activation of target genes.

The biological consequences of osteomimicry could be numerous. For example, the activation of OC and BSP expression could result in the recruitment of bone cells, such as osteoclasts and osteo-blasts that participate in enhanced osteoclastogenesis, that is, increased bone turn-over, or bone pitting to create new sites in support of cancer cell attachment, growth, and colonization in bone (59,60). OC and BSP activation could contribute to new bone formation and mineralization (61). Prostate cancer cells derived from LNCaP cells with increased bone metastatic potential, such as C4-2 and C4-2B, have activated OC and BSP gene expression and also are capable of forming bone nodules when subjected to mineralizing cell culture conditions in vitro (62). Activation of CREB could result in marked gene expression changes in cancer cells and cells in the cancer microenvironment that could facilitate cancer cell growth, survival, and colonization in bone. It is interesting to note that, of approx 4000 potential CREB target genes, only a fraction are expressed in a cell context-dependent manner (63,64). This again supports the concept of the importance of the host contribution to cancer cell growth, resistance to apoptosis, and resistance to therapy. A large number of other downstream genes unrelated to osteomimicry were also found to be regulated by cAMP-PKA activation. For

G proteins

G proteins

Fig. 2. Molecular mechanisms of osteomimicry in prostate cancer cells. Prostate cancer cells have the ability to mimic gene expression and behaviors of bone cells by synthesizing and depositing bone-like proteins, such as osteocalcin (OC) and bone sialoprotein (BSP) (57). Among factors that could regulate the expression of these proteins, we found that P2 microglobulin (P2M) induces the expression of OC and BSP through an activation of the cyclic AMP (cAMP)-protein kinase A (PKA) signaling pathway with its downstream activation of a cAMP-responsive element (CRE)-binding protein (CREB). P2M is considered as a housekeeping gene with uniform expression of its mRNA in many cells. Interestingly, P2M protein expression varied widely between cells. P2M could exert its action via a number of membrane receptors, such as a G protein-coupled receptor (GPCR), a major histocompatibility antigen complex (MHC), or a yet-to-be identified P 2M receptor. Alternatively, P 2M could also act intracellularly by modulating directly CREB downstream target gene expression. The resulting increased expression and deposition of OC and BSP in bone matrix could recruit osteoblasts and osteoclasts and initiate increased bone turnover (i.e., increased bone resorption and formation) through osteoclastogenesis, which facilitates prostate cancer bone colonization. RANK, receptor activator of NF-kB and RANKL, RANK ligand (58).

Bone resorption

Fig. 2. Molecular mechanisms of osteomimicry in prostate cancer cells. Prostate cancer cells have the ability to mimic gene expression and behaviors of bone cells by synthesizing and depositing bone-like proteins, such as osteocalcin (OC) and bone sialoprotein (BSP) (57). Among factors that could regulate the expression of these proteins, we found that P2 microglobulin (P2M) induces the expression of OC and BSP through an activation of the cyclic AMP (cAMP)-protein kinase A (PKA) signaling pathway with its downstream activation of a cAMP-responsive element (CRE)-binding protein (CREB). P2M is considered as a housekeeping gene with uniform expression of its mRNA in many cells. Interestingly, P2M protein expression varied widely between cells. P2M could exert its action via a number of membrane receptors, such as a G protein-coupled receptor (GPCR), a major histocompatibility antigen complex (MHC), or a yet-to-be identified P 2M receptor. Alternatively, P 2M could also act intracellularly by modulating directly CREB downstream target gene expression. The resulting increased expression and deposition of OC and BSP in bone matrix could recruit osteoblasts and osteoclasts and initiate increased bone turnover (i.e., increased bone resorption and formation) through osteoclastogenesis, which facilitates prostate cancer bone colonization. RANK, receptor activator of NF-kB and RANKL, RANK ligand (58).

example, after cAMP-PKA/CREB activation in cancer cells, the phosphorylated CREB is responsible for the recruitment of CBP/p300 to activate downstream genes such as vascular endothelial growth factor (VEGF), cyclins, and survival factors aiding the growth and survival of cancer cells (65,66). It has been shown that removing androgen from cultured prostate cancer cells elicits a neuroendocrine phenotype caused by CREB activation (42,66,67). As expected, neuroendocrine differentiation, which has been associated with increased invasion, migration, and metastasis of prostate cancer cells, was also activated by cAMP mimetics, such as dibutyl cAMP or forskolin (42,67). Activation of G protein, via cAMP-activated P2 adrenergic receptor, has been shown to compensate for the requirement of androgen to activate AR downstream genes (67). This phenomenon could have clinical importance, because it has been proposed that the activation of AR by suboptimal concentrations of androgen could be responsible for the survival of prostate cancer cells in patients subjected to androgen-withdrawal treatment regimens (45).

3.2. Vasculogenic Mimicry

Cancer cells often grow under stress conditions, because of the lack of oxygen, increased cell crowding, and the lack of sufficient nutrients. In response to these conditions, adaptive changes by cancer cells have been observed to create their own blood vessels, such as ischemia-induced vasculogenic mimicry in melanoma, breast, and prostate cancer (61,68,69). Hendrix and colleagues (31) demonstrated the plasticity of melanoma cells, which formed tubular structures with patterned matrix deposition including laminin, heparan sulfate proteoglycans, and collagens IV and VI, expressing genes normally expressed by vascular endothelial cells and interconnecting with the preexisting blood vessels (31). This adaptive capability of cancer cells toward changes in microenvi-ronmental cues could sustain their growth and survival at metastatic sites and allow them to gain further invasive and migratory potentials. A direct link between vasculogenic mimicry, the activation of focal adhesion kinase (FAK) through phosphorylation of tyrosine-397 and -576, and decreased plasminogen activation through decreased urokinase activity have been reported (70). Because FAK activation has been associated with increased cell invasion and migration, and plasminogen activation is crucial for blood clotting, the ability of cancer cells to alter FAK and plasminogen activation is consistent with observations of an increased ability to metastasize along with aberrant wound-healing properties (30,71,72).

3.3. Epithelial-to-Mesenchymal Transition

Epithelial-to-mesenchymal transition (EMT) is a fundamental cellular process whereby an epithelial cell undergoes a structural and functional transition to assume the phenotype and behavior of a mesenchymal cell, characterized by increased migratory and invasive properties in embryonic development and also on neoplastic progression (73). It is now well-accepted that EMT occurs in a number of human cancers, including prostate cancer, and that EMT is associated with increased cancer invasion and metastasis. EMT is a highly dynamic process that signals the plasticity of cancer cells (Fig. 3). Cancer metastasis is often preceded by EMT, but, on the completion of the metastatic process, cancer cells can revert their phenotype and behavior by undergoing mesenchymal-to-epithelial transition (MET), presumably to increase cancer cell adhesion to each other and to cells in host microenvironment. This suggests the importance of the epigenetic host microenvironment that could trigger EMT and its reversal, MET (74).

There are numerous well-characterized molecular pathways that describe the underlying key regulatory processes of EMT in human cancer cells. Activation of TGF-P signaling enhances receptor tyrosine kinases and Ras activities that together can drive the translocation of Smad and Snail transcription factors from the cytoplasmic to the nuclear compartment, and the activation or suppression of downstream target genes associated with EMT have been widely proposed as the key regulatory mechanisms underlying EMT in human cancer cells (75,76). Other molecular mechanisms include the activation of Wnt and P-catenin signaling, which suppresses E-cadherin and initiates the early step of EMT (77,78); activation of the Hedgehog pathway, which contributes to increased EMT and stem cell differentiation (73,79), both important features shared by invasive cancer cells; and activation of NF-kB transcription factor, which translocates into the cell nucleus, improves the survival of cancer cells, and allows them to resist apoptotic death after therapeutic intervention (80,81). Figure 3 depicts selective growth control signaling pathways of EMT and MET as a continuum of prostate cancer progression.

Fig. 3. The proposed epithelial-to-mesenchymal transition (EMT) and its reversal of mesenchymal-to-epi-thelial transition (MET) during the progression of human prostate cancer. Increased migration and invasion of prostate cancer cells before bone metastasis can be initiated through EMT under the influence of increased growth factor signaling (e.g., transforming growth factor [TGF]-P1, epidermal growth factor [EGF], or P2 microglobulin [p2M]). The sources of growth factors and cytokines can originate from resident fibroblasts or inflammatory cells within cancer-associated stroma, and these factors can complex with extracellular matrices (ECM). Cancer cells can metastasize to bone through hematogenous spread and after arriving at the bone, cancer cells have been observed to undergo MET under the influence of bone morphogenetic protein (BMP)-7 and lipocalin 2, by reexpressing epithelial cell-associated markers, such as E-cadherin, and decreased expression of vimentin and N-cadherin and increased cancer cell adhesion and colonization at metastic sites.

Fig. 3. The proposed epithelial-to-mesenchymal transition (EMT) and its reversal of mesenchymal-to-epi-thelial transition (MET) during the progression of human prostate cancer. Increased migration and invasion of prostate cancer cells before bone metastasis can be initiated through EMT under the influence of increased growth factor signaling (e.g., transforming growth factor [TGF]-P1, epidermal growth factor [EGF], or P2 microglobulin [p2M]). The sources of growth factors and cytokines can originate from resident fibroblasts or inflammatory cells within cancer-associated stroma, and these factors can complex with extracellular matrices (ECM). Cancer cells can metastasize to bone through hematogenous spread and after arriving at the bone, cancer cells have been observed to undergo MET under the influence of bone morphogenetic protein (BMP)-7 and lipocalin 2, by reexpressing epithelial cell-associated markers, such as E-cadherin, and decreased expression of vimentin and N-cadherin and increased cancer cell adhesion and colonization at metastic sites.

4. HOW HOST CELLS CONTRIBUTE TO THE GENESIS OF CANCER 4.1. Host Infiltrating Inflammatory Cells

Cancer development often coincides with active chronic and recurrent inflammatory responses caused by innate immune responses to the presence of altered cancer epithelial cells and bacterial or viral infections at the site of cancer origin. The infiltrating inflammatory cells have been shown to release ROS and reactive nitrogen species, such as hydrogen peroxide, superoxide, and nitric oxide, to protect the host cells and eradicate the "foreign" cells and invading organisms (82,83). Responses to the presence of these highly reactive oxygen and nitrogen radicals released by the inflammatory cells and cancer cells could induce DNA damage to cancer cells and host stroma, activating DNA repair and cell proliferation programs to compensate for the cell loss resulting from failure to repair and subsequent cell death (84,85). An extensive literature suggests the possible functional and signaling roles of oxygen and nitrogen radicals in eliciting cell responses to stress and escape mechanisms for survival (86). These observations collectively support the important role of the inflammatory cascade in carcinogenesis. De Marzo et al. (87,88) proposed a role for inflammation in prostate cancer development when they found foci of proliferative inflammatory atrophy (PIA) as precursor lesions before the detection of prostate intraepithelial neoplasia (PIN), a known early pathological lesion associated with human prostate cancer development. They reported compelling epide-miological evidence to suggest a link between prostate inflammation and prostate cancer in men (87). For example, a positive correlation was found between prostatitis and sexually transmitted infections

Fig. 4. The contribution of inflammatory cells to the progression of prostate cancer. Recruitment of inflammatory cells to the prostate gland can occur under a variety of physiological and pathophysiological conditions, such as wound, infection, prostatitis, and cancer. Genetically altered prostate epithelial cells, under the influence of resident fibroblasts, inflammatory (such as macrophage and lymphocytes) cells, and endothelial cells progress further through additional genetic changes triggered by reactive oxygen species (ROS) or reactive nitrogen species (RNOS). The genetically unstable prostate cancer cell clusters can form proliferative inflammatory atrophy and then proceed to prostate intraepithelial neoplasia before becoming prostate cancer cells with increased malignant potential. After metastasizing to bone, prostate cancer cells interact with bone cells, such as osteoblasts and marrow stromal cells, to increase their growth and survival in bone. Through cellular interaction with osteoclasts, prostate cancer cells also promote osteoclastogenesis and bone turnover by increased osteoclast maturation via receptor activator of NF-kB (RANK) ligand (localized on prostate cancer cell surface) and RANK (localized on the cell surface of osteoclasts) interaction. Because of increased bone turnover, there is increased release of soluble growth factors, cytokines and extracellular matrices (ECMs), which promote further prostate cancer growth and survival in bone.

Fig. 4. The contribution of inflammatory cells to the progression of prostate cancer. Recruitment of inflammatory cells to the prostate gland can occur under a variety of physiological and pathophysiological conditions, such as wound, infection, prostatitis, and cancer. Genetically altered prostate epithelial cells, under the influence of resident fibroblasts, inflammatory (such as macrophage and lymphocytes) cells, and endothelial cells progress further through additional genetic changes triggered by reactive oxygen species (ROS) or reactive nitrogen species (RNOS). The genetically unstable prostate cancer cell clusters can form proliferative inflammatory atrophy and then proceed to prostate intraepithelial neoplasia before becoming prostate cancer cells with increased malignant potential. After metastasizing to bone, prostate cancer cells interact with bone cells, such as osteoblasts and marrow stromal cells, to increase their growth and survival in bone. Through cellular interaction with osteoclasts, prostate cancer cells also promote osteoclastogenesis and bone turnover by increased osteoclast maturation via receptor activator of NF-kB (RANK) ligand (localized on prostate cancer cell surface) and RANK (localized on the cell surface of osteoclasts) interaction. Because of increased bone turnover, there is increased release of soluble growth factors, cytokines and extracellular matrices (ECMs), which promote further prostate cancer growth and survival in bone.

and increased prostate cancer risk (88-90). Intake of anti-inflammatory drugs and antioxidants has been shown to decrease prostate cancer risk (91-94). Genetic studies revealed further supportive evidence that RNASEL, encoding an interferon-inducible ribonuclease (95,96), and MSRI, encoding subunits of the macrophage scavenger receptor (96-98), are candidates as inherited susceptibility genes for familial prostate cancer. Conversely, the loss of GSTPI, encoding a glutathione-S-trans-ferase capable of inactivating ROS and, thus, reducing genome damage, has been found to occur frequently in human prostate cancer (87,98-100). Figure 4 emphasizes the potential roles of inflammatory processes and the ways to antagonize them in prostate cancer development.

4.2. Host Stem Cells

There are two pools of stem cells that are thought to contribute to local cancer growth and its distant metastasis, and one of these stem cell populations serves as progenitors. With their uncontrolled growth potential and self-renewal property, they become a constant source of cancer cells and populate the tumor mass (101-103). For example, the basal cells in the prostate gland have been referred to as stem cells that share gene expression profiles with cancer cells and could be considered as the progenitor cells of prostate cancer (102,104). Another pool of stem cells could originate from the host (22,105). Bone marrow-derived progenitor stem cells have migratory and invasive potential (106,107). On cancer cell growth and metastasis, this pool of cells has been shown to migrate into primary or metastatic cancers, creating a rich source of growth factor and a cytokine niche supporting the growth and expansion of cancer cells (25). Although there is no concrete example of this kind of mechanism in prostate cancer, a recent work by Kaplan et al. (108) showed that the ability of tumor cells to metastasize to a predetermined location can be explained by the previous "marking" of the metastatic site by bone marrow-hematopoietic progenitor cells that express VEGF receptor 1 (VEGF1, or Fltl) and VLA-4 (integrin a4p 1; refs. 109 and 110). The expression of a4p 1, a known receptor of fibronectin produced by resident fibroblasts, in response to tumor-specific factors, creates a permissive niche for the incoming migrating tumor cells (109,110). If this mechanism has general applicability, it can be proposed that a previously established bone marrow stem cell niche, in response to tumor-derived factors by resident marrow stromal cells, could also be responsible for attracting prostate cancer bone metastasis (108-111). It has already been proposed that the homing mechanism of prostate cancer cells may involve:

1. Chemokines (SDF-1 or CXCR12) derived from marrow stromal cells and chemokine receptors (CXCR4) on the cell surface of prostate cancer cells (112).

2. Cell adhesion molecules on marrow endothelial cells and integrins (avp3 and a4p 1) on the cell surface of prostate cancer cells (113,114).

3. Hedgehog produced by cancer cells, which triggers a host stromal response mediated by paracrine interaction with cell surface receptors, patched1 (115-117).

4. Complementary growth factors/growth factor receptors and/or extracellular matrices/integrins produced by prostate cancer cells and bone cells, such as marrow stromal cells, osteoblasts, osteoclasts, or bone marrow progenitor stem cells.

Understanding the molecular mechanisms at the interface of prostate cancer cells and host cells could help in the future development of novel therapies for the treatment of prostate cancer bone metastasis (5,45,60).

+1 0

Post a comment