| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The Clinical Connection |
Author's Affiliation: Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, Nebraska
Requests for reprints: James E. Talmadge, University of Nebraska Medical Center, 987660 Nebraska Medical Center, Omaha, NE 68198-7660. Phone: 402-559-5639; Fax: 402-559-4990; E-mail: jtalmadg{at}unmc.edu.
Activin, a member of the transforming growth factor-β (TGF-β) superfamily, and its antagonist, follistatin, can regulate cell proliferation, differentiation, angiogenesis, apoptosis, and potentially tumor growth. Follistatin, a glycosylated single-chain protein (1, 2), is biologically active due, in part, to its ability to bind and neutralize activin (3). Relevance of these molecules to tumor progression includes the finding that activin-expressing mammary carcinomas (4) grow faster than follistatin-expressing tumors, which grow slower than activin-transduced or mock-transduced tumors. In this issue of Clinical Cancer Research, Ogino et al. (5) expand the profile of follistatin bioactivity to include suppression of experimental metastasis. The authors used multiple variants from two small-cell lung cancer cell lines, which they transfected with follistatin and injected intravenously into severe combined immunodeficiency disease mice to form experimental metastasis. One interesting aspect of these studies from the laboratory of Dr. Sone at the University of Tokushima is their novel observation of experimental bone metastasis (6).
Metastasis is the dissemination and growth of tumor cells at sites discontinuous to the primary tumor and represents the most devastating and challenging aspect of tumor growth. Many patients with advanced lung, breast, or prostate cancers develop osteolytic bone metastasis, resulting in significant morbidity and occasionally mortality (7). Recently, Kang et al. (8) found follistatin to be one of a limited number of overexpressed genes that were part of a bone metastasis signature, which included overexpression of CXCR4, MMP1, FGF5, CTGF, and IL-11. They also reported that this bone metastasis signature differed from that of metastasis gene signatures. This is an important concept (9) that suggests that metastatic efficacy is dependent on the coordinated expression of multiple genes and that the expression of metastasis-associated genes can be dependent on the microenvironment such as the bone.
Patients with bone metastasis can have multiple complications, including osteolysis, spinal cord compression, hypercalcemia, fracture of long bones, and unrelenting and intractable pain (10). Furthermore, patients who have developed bone metastasis are rarely cured. Unfortunately, our understanding of bone metastasis is incomplete, which limits the development of effective therapeutics (11). However, recent studies from Italy (12, 13) have reported that circulating levels of activin in cancer patients is associated with the presence of bone metastasis. These studies included patients with breast cancer or prostate cancer who were compared with female patients with age-related osteoporosis, male patients with benign prostatic hypertrophy, and healthy blood donors of both sexes. Serum activin levels were reported to be significantly increased in breast cancer or prostate cancer patients as compared with age-matched control donors (P < 0.0001) and patients with osteoporosis or benign prostatic hypertrophy. Further, the activin levels in breast cancer or prostate cancer patients with bone metastases were significantly higher than patients without bone metastases. A significant correlation was also observed in breast cancer patients between activin levels and the number of bone metastases, whereas in prostate cancer patients, activin levels correlated with the Gleason score, prostate-specific antigen levels, and, to a lesser extent, bone metastases. Based on these observations, activin seems to have a role in the pathogenesis of bone metastasis, suggesting that activin inhibitors such as follistatin might have therapeutic activity. This hypothesis is supported by the observation of therapeutic activity by Ki26894 for a bone metastatic variant of human breast cancer MDA-MB-231 cells (14). In these studies, Ki26894 blocked TGF-β signaling by MDA-231-D cells and TGF-β–induced motility and invasion of MDA-231-D cells. X-ray radiography revealed that systemic Ki26894 treatment initiated 1 day before the injection of MDA-231-D cells into the left ventricle of BALB/c nu/nu female mice decreased the incidence of bone metastasis. Moreover, Ki26894 treatment prolonged the survival of mice inoculated with MDA-231-D cells as compared with vehicle-treated mice. Taken together, these findings suggest that TGF-β kinase inhibitors such as Ki26894 may be useful for blocking metastasis to bone.
The formation of metastases in bone consists of multiple steps, including proliferation of tumor cells and endothelial cell formation of hematologic and lymphatic vasculature; intravasation and circulation of cancer cells in the primary tumor vasculature; cellular arrest and extravasation of the circulating tumors cells; and subsequent growth and expansion of the tumor cells resulting in a micrometastatic site. If tumor cells arrest in the bone marrow, they may egress from a central sinus and attach to a bone surface, frequently stimulating osteoclastic bone destruction and resulting in the formation of a microfocus of tumor cells (Fig. 1 ). The bone microenvironment and the cellular and molecular signatures associated with bone metastases are better understood compared with metastasis to other organs (8). Interactions between tumor cells and osteoclasts are unique and critical to the formation of bone metastases, providing targets for designing specific therapeutic interventions against bone metastases. Indeed, drugs such as bisphosphonates (15) or neutralizing antibodies to parathyroid hormone–related protein and osteoprotegerin can disrupt these interactions (16) and have been shown to suppress bone metastases. Despite these successes, our understanding of cancer molecular signatures and signatures critical to bone metastases remains limited. Determination of molecules critical to bone metastasis should facilitate the development of effective and specific therapeutic interventions targeting bone metastases. Yoneda et al. (17) showed that chemokine (C-X-C motif) receptor 4, a receptor for the stromal cell–derived factor-1 ligand with high expression levels in bone, is highly expressed by a bone metastatic variant of the MDA-MB-231 breast cancer cell line, an observation (18) consistent with clinical data in breast cancer (19). One current hypothesis is that tumor cells respond to stromal cell–derived factor-1 chemoattractant gradients, resulting in preferential migration to bone (19). Interestingly, many of the genes expressed in bone metastasizing MDA-MB-231 variants are responsive to TGF-β, which is highly expressed in bone (20).
|
Based on these results, follistatin joins a long list of chemokines, cytokines, enzymes, and growth factors whose overexpression has been associated with malignancy and metastasis (22–24). Whereas there is always excitement with the identification of a new surrogate associated with experimental metastasis, there are redundancies within the metastatic process such that any one molecule is unlikely to be critical to this pathologic process. Nonetheless, each new potential regulator expands our understanding of the metastatic process and provides a potential new therapeutic target for our armatarium against metastatic disease. Because experimental metastasis assays reflect only steps that occur late in the metastasis process, we look forward to reports of follistatin expression effects on primary tumor growth and spontaneous metastasis, especially metastasis to bone sites. The finding that transfection with follistatin has inhibitory activity for metastasis by two different small-cell lung cancer cell lines suggests that this observation may not be model dependent and studies extending these observations to other tumor histiotypes are warranted. One remaining critical question is whether those metastatic foci that occur following experimental metastasis retain the follistatin transgene or if it has been deleted or down-regulated. Taken together, these studies support therapeutic strategies focused on mediators that down-regulate activin levels such as follistatin. Further, it seems likely that if such therapeutics can be targeted to the medullary space, they will have the potential to control metastatic growth in bones (25).
Acknowledgments
I thank Kirsten Stites for critical review and editing of the manuscript, as well as the development of the graphic.
Footnotes
Commentary on Ogino et al., p. 660
Received 10/ 2/07; accepted 10/ 5/07.
References
Key Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
