Interactions between bone marrow-derived cytokines, growth factors, and tumors play a critical role in both the homing of tumors to the bone and the development of bone metastasis. Bone is a storehouse of latent growth factors produced by stromal cells and osteoblasts that, when activated during osteoclastic bone resorption, can enhance the growth of tumor cells.
This article reviews the role these factors may play in bone metastasis.
Several studies have shown that breast carcinoma cells, which induce osteoclastic bone resorption, release growth factors that enhance tumor growth. In addition, bone-derived growth factors and chemokines, such as stromal cell-derived factor 1 and monocyte chemoattractant protein 1, can act as chemoattractants to attract tumor cells to bone. Finally, the interaction between tumor cells and bone marrow stromal cells can result in increased production of cytokines and growth factors, such as interleukin 6 or the ligand for the receptor activator of nuclear factor κB, that can enhance bone destruction, tumor growth, and angiogenesis.
Bone is a frequent site of metastasis in patients with breast carcinoma, prostate carcinoma, and myeloma. This reflects both the propensity of these tumors to home to bone and the capacity of bone marrow to support the growth of the tumor. The metastatic process, which involves bone destruction and/or formation, enhances both the growth of the tumor due to the release of activated growth factors from bone during the bone resorptive process and the production by bone marrow stromal cells and osteoblasts of growth factors and cytokines, which enhance the survival and growth of the tumors. Many of these stromal cell-derived growth factors and cytokines are up-regulated due to adhesive interactions between tumor cells and bone marrow stromal cells (Fig. 1, Table 1). In addition, soluble factors produced by tumors themselves further enhance the production of cytokines by bone marrow stromal cells, which also increase tumor survival and growth. Cytokines that have been shown to be active in this process are listed in Table 2. This article is focused on stromal cell-derived factors that enhance the bone metastatic process and tumor growth.
Table 1. Adhesive Interactions Involved in Bone Metastases
Adhesion and progression
Growth factor production
Table 2. Stromal/Osteoblast-Derived Factors that May Affect Bone Metastasis
PDGF: platelet-derived growth factor; RANK-L: ligand of the receptor activator for nuclear factor κB; VEGF: vascular endothelial growth factor; IL-6: interleukin 6; FGFs: fibroblast growth factors; BMPs: bone morphogenic proteins; IGFs: insulin-like growth factors; SDF-1: stromal cell-derived factor 1; TGF-β: transforming growth factor β; MCP-1: monocyte chemoattractant protein 1.
Adhesive Interactions In Bone Metastasis
Interactions between specific cell surface molecules on bone cells, bone marrow cells, and tumors are critical to both tumor invasion and the metastatic process. The importance of these interactions has been demonstrated clearly by studies in human breast carcinoma, myeloma, and prostate carcinoma. Van der Pluijm and coworkers1 found that the urokinase receptor and β1 integrins formed functional adhesion complexes at distinct sites at the cell surface of metastatic human breast carcinoma cells and that the urokinase receptor is capable of regulating the adhesive function of integrins on breast carcinoma cells. That study showed that the addition of a blocking peptide for the urokinase-integrin complex inhibits the attachment of breast carcinoma cells to vitronectin. Using a mouse model of breast carcinoma metastasis, those authors reported that transplantation of nude mice with MDA-231 breast carcinoma cells, which over express this blocking peptide, results in a significant reduction in tumor progression in bone compared with empty vector-transfected cells. Furthermore, mice that were transplanted with MDA-231 cells and that received continuous administration of the peptide for 28 days had significantly reduced tumor progression in bone compared with animals that were treated with a scrambled control peptide. These data show that breast carcinoma progression in bone requires adhesive interactions between molecules that are expressed in the bone and molecules that are expressed in tumor cells. Similarly, Sung and coworkers2 have shown human breast carcinoma cells adhere, proliferate, and migrate to bone through the interactions between αvβ3 and αvβ5 integrins and bone sialoprotein.
In patients with multiple myeloma, adhesive interactions between the α4β1 and α5β1 integrins and vascular cell adhesion molecule 1 (VCAM-1) or fibronectin appear to play an important role in up-regulating the expression of cytokines and growth factors by bone marrow stromal cells, further enhancing tumor growth and chemotherapy resistance of the tumor. Damiano and Dalton3 have shown that these adhesive interactions play an especially important role in the capacity of myeloma cells to resist standard chemotherapeutic agents, including doxorubicin and melphalan. It is very likely that adhesive interactions between a variety of tumor cells and bone marrow stromal cells result in the release of growth factors by stromal cells and osteoblasts, also further enhancing tumor growth. In support of the importance of these adhesive interactions between tumor cells and bone marrow stromal cells are our recent studies with macrophage inflammatory peptide 1α (MIP-1α) and the myeloma cell-derived cell line ARH-77.4 We previously reported that about 70% of myeloma cells from patients produce MIP-1α5 and that MIP-1α is a potent osteoclast stimulatory factor that induces human osteoclastogenesis independent of the ligand for the receptor activator of nuclear factor κB (RANK-L).6 Furthermore, when ARH-77 cells are transplanted into severe combined immunodeficiency (SCID) mice, the mice develop all of the clinical features of human myeloma, including lytic bone lesions, hypercalcemia, and bone destruction, only at sites adjacent to the myeloma-derived cells. Most important, when an antisense construct to MIP-1α was transfected into ARH-77 cells, the cells displayed both decreased tumor cell homing and decreased tumor growth in the bone marrow compared with cells that were transfected with the empty vector.4 This decreased tumor growth was due to down-regulation of the α5β1 integrin on the ARH-77 cells that resulted in decreased binding to human bone marrow stromal cells and decreased growth in vivo. Thus, adhesive interactions between tumor cells and stromal cells play a critical role in the homing of the tumor to the bone, the growth of the tumor in the bone, and the up-regulation of growth factor production by stromal cells required for tumor cell survival.
Growth Factors and Cytokines Involved in Bone Metastasis
Transforming growth factor β
Several laboratories have shown that, when breast carcinoma cells metastasize to bone and induce bone resorption, transforming growth factor β (TGF-β) is released in active form from bone. In particular, Chirgwin and Guise7 have shown that breast carcinoma cells produce parathyroid hormone-related peptide (PTHrP), which induces osteoclastic bone resorption and releases TGF-β from the bone matrix. TGF-β is produced by osteoblasts and bone marrow stromal cells. TGF-β then increases PTHrP production further, creating a vicious cycle in which tumor cells induce bone destruction and, through this process, release growth factors that enhance the growth of the tumor. TGF-β is a potent, multifunctional cytokine that is produced by many cells, including osteoblasts and bone marrow stromal cells that can regulate cell growth and stimulate matrix production. It has been demonstrated that TGF-β is a major factor in bone remodeling, and tumor-derived agents that enhance TGF-β production have been associated with increased bone formation.8 TGF-β normally functions as a suppressor of tumor growth. Lang and coworkers9 have shown that mice lacking TGF-β due to haploinsufficiency are more susceptible to tumors. Furthermore, TGF-β is immunosuppressive, which can increase tumor survival further by suppressing the immune system. TGF-β also can stimulate normal stromal cells and osteoblasts to secrete growth factors that enhance tumor growth. In patients with multiple myeloma, Brown and coworkers10 reported that TGF-β appears to depress dendritic cell function in these patients, further enhancing the growth of these tumors and protecting them from immune surveillance. Guise and coworkers11 showed that the administration of a neutralizing monoclonal antibody to PTHrP inhibited the development of breast carcinoma bone metastasis by MDA-MB-231 cells in nude mice. It is noteworthy that those authors also showed that inhibiting TGF-β responsiveness of the tumor by using a dominant negative TGF-β receptor12 reduced bone metastasis.
Platelet-derived growth factor
Platelet-derived growth factor (PDGF) is a polypeptide produced by osteoblasts in the bone microenvironment that shows extensive sequence homology with the oncogene c-Cis. PDGF increases cell replication, bone resorption, collagen degradation, and collagenase expression as well as inhibiting osteoblast function. The mitogenic activity of PDGF results in the growth of tumor cells as well as the enhancement of osteoclast activity. Yi and coworkers13 reported that, using MCF-7 cells in a model of breast carcinoma metastasis to bone in nude mice, breast carcinoma cells that overexpressed HER-2-NEU produced large amounts of PDGF and showed an enhanced propensity to metastasize to bone. Furthermore, they suggested that PDGF played a causative role in the development of osteosclerotic bone metastasis in this model. Thus, up-regulation of PDGF may enhance osteoblast formation and activity in bone metastasis and may enhance the growth of tumor cells through its mitogenic effects on tumors.
Fibroblast growth factors
Fibroblast growth factors (FGFs) consist of a gene family that contains 23 members, although only FGF-1 and FGF-2 have been studied extensively in bone. FGF-2 is produced by osteoblasts and stimulates bone cell replication. FGF-2 also has been shown to increase osteoclast-like cell formation in mouse bone marrow cultures14 Beck and coworkers15 showed that FGF can increase the growth of neuroblastoma cells as well as enhancing the expression of adhesion molecules on neuroblastoma cells, thereby increasing tumor growth.
Vascular endothelial growth factor
Vascular endothelial growth factor (VEGF) induces the growth of vascular endothelial cells as well as enhancing osteoclast formation and activity. Investigators in Japan have shown that VEGF can rescue the osteopetrotic phenotype of the op/op mouse, suggesting that VEGF is an osteoclast stimulatory factor.16 VEGF is produced by stromal cells rather than by the vascular endothelium of bone. VEGF, FGF, TGF-β, and the insulin growth factors increase VEGF production by several cell types, and VEGF production is up-regulated by myeloma cells when they bind to bone marrow stromal cells.17 This up-regulation of VEGF results in increased angiogenesis that enhances the proliferation of the myeloma cells. Thus, VEGF appears to be an important cytokine regulating the growth of myeloma cells and probably the growth of other tumor cells that bind to bone marrow stromal cells. Gupta and coworkers17 have shown that, in cocultures of myeloma cells with bone marrow stromal cells, VEGF significantly increased interleukin 6 (IL-6) secretion by bone marrow stromal cells and that stromal cells from myeloma patients and normal donors secreted VEGF, FGF, and IL-6. Thus, VEGF produced by bone marrow stromal cells has multiple effects on tumor growth, including increased angiogenesis, increased growth of tumor cells, and up-regulation of IL-6 (another growth factor for myeloma cells), as well as increased osteoclast formation.
Insulin-like growth factors
Insulin-like growth factors (IGFs) are produced by osteoblastic cells and are regulated by a number of factors produced by bone marrow stromal cells in the bone marrow microenvironment. These include TGF-β, FGF, PDGF, and prostaglandins. IGFs induce proliferation of osteoblasts and play a major role in stimulating differentiation of osteoblasts.18 In addition, IGFs play a role in bone resorption by stimulating the formation of osteoclasts and activating preexisting osteoclasts. However, the mechanism responsible for the effects of IGFs on osteoclasts remains to be clarified. It is possible that IGFs induce osteoclast activity by their indirect effects on osteoblasts. IGFs also are potent mitogens for tumor cells. For example, Ferlin and coworkers19 showed that IGFs induce the survival and proliferation of myeloma cells independent of IL-6 production by bone marrow stromal cells. Those authors found that IGF acts as a potent survival/proliferation factor for myeloma cells, and they strongly suggested a role for IGFs in the pathophysiology of multiple myeloma.
IL-6 is a pleiotropic cytokine that has multiple effects on cell function. IL-6 enhances osteoclast activity in human bone marrow and in murine bone marrow cultures and appears to have direct effects on osteoclast precursors in human systems.20 IL-6 is also a growth factor for myeloma cells, and its production is up-regulated when myeloma cells bind to bone marrow stromal cells.21 This up-regulation of IL-6 appears to be result from adhesive interactions between very late activation antigen 4 (VLA-4) and VCAM-1. Induction of IL-6 production by bone marrow stromal cells has multiple effects on myeloma growth in bone. It enhances the growth of the myeloma cells and enhances osteoclastic bone resorption, which results in the release of growth factors from the bone matrix. IL-6 can act synergistically or additively with other osteoclast-activating factors produced in the myeloma microenvironment to further increase bone destruction and tumor growth. Recently, Yaccoby and coworkers reported that there appeared to be an important link between the growth of myeloma cells and bone disease in a human SCID (SCID-hu) model of myeloma.22 Those investigators reported that blocking osteoclastic bone resorption decreased the growth of primary myeloma cells from patients in fetal bones implanted in SCID mice. Blocking IL-6 production may have important effects on the growth of myeloma cells and bone destruction in patients with myeloma.
RANK-L is a recently described factor produced by osteoblasts and stromal cells that is a potent stimulator of osteoclastogenesis. RANK-L expression is increased when myeloma cells or breast cancer cells bind to marrow stromal cells.23 RANK-L then induces osteoclastogenesis, which results in the release of growth factors that further enhance the growth and survival of tumor cells. In addition, factors produced by tumor cells directly enhance RANK-L expression. Using the 5T2 model of myeloma, Oyajobi and coworkers24 showed that MIP-1α can induce RANK-L expression in bone marrow stromal cells. In addition, IL-6, which is induced when murine myeloma cells bind to bone marrow stromal cells, also enhances RANK-L production. It has been shown that IL-6 enhances the survival of myeloma cells. RANK-L has no direct effects on the growth of tumor cells, although its increased production enhances bone destruction, which further enhances the bone metastatic process.
Chemoattractants Produced by Bone Marrow Stromal Cells
A number of factors produced by bone marrow stromal cells and osteoblasts can act as chemoattractants to enhance the migration of tumor cells to bone (Fig. 2). Monocyte chemoattractant protein 1, which is secreted by bone marrow endothelial cells, produces chemoattraction of the 5T2 multiple myeloma cell line.25 Recently, Muller and coworkers26 reported that breast carcinoma cells from malignant breast tumors and metastases expressed the chemokine receptors CXCR4 and CCR-7, which mediated the chemoattraction and invasive responses in the tumors. These findings suggest that chemokines and their receptors may play a critical role in determining the metastatic destination of tumor cells. Geminder and coworkers27 reported a possible role for CXCR4 and its ligand, stromal cell derived factor-1 (SDF-1), in the development of bone marrow metastasis by neuroblastoma. Those authors reported that CXCR4 expression may be a general characteristic of neuroblastoma cells and that SDF-1 induces the migration of CXCR4-expressing neuroblastoma cells. These neuroblastoma cells then interact with components of the bone marrow microenvironment to promote neuroblastoma cell adhesion to bone marrow stromal cells and subsequent neuroblastoma cell proliferation. It also has been shown that SDF-1 plays an important role in myeloma cell adhesion to fibronectin and VCAM-1.28 Bone marrow stromal cells express SDF-1, and SDF-1 expression up-regulates VLA-4-mediated myeloma cell adhesion to fibronectin.28
Interactions between tumor cells and bone marrow stromal cells result in enhanced production of cytokines and growth factors that increase osteoclastic bone resorption and tumor growth and survival. Enhanced osteoclastic bone destruction results in the release of growth factors from the bone matrix, further enhancing the growth of tumor cells and increasing the production of PTHrP, which further increases osteoclastic bone resorption. Interruption of these tumor-stromal interactions and down-regulation of the production of cytokines and growth factors produced by bone marrow stromal cells should have beneficial therapeutic effects for patients with bone metastases. For example, osteoprotegerin, which blocks RANK-L-induced osteoclast formation, also decreased the growth of myeloma cells in an SCID-hu model of myeloma,23 whereas blocking IL-6 expression by bone marrow stromal cells can effect the growth and survival of myeloma cells. Therapeutic interventions that block the expression of growth factors on bone marrow stromal cells should have profound effects on the treatment of patients with bone metastasis.