CD157 in Ovarian Carcinoma: How Does It Help Us? -- Annunziata and Birrer, 10.1093/jnci/djq269 -- JNCI Journal of the National Cancer Institute

© The Author 2010. Published by Oxford University Press.

----------
EDITORIAL


CD157 in Ovarian Carcinoma: How Does It Help Us?
Christina M. Annunziata, Michael J. Birrer
Affiliations of authors: Medical Oncology Branch, National Cancer Institute, Bethesda, MD (CMA); Department of Medicine, Massachusetts General Hospital Cancer Center, Boston, MA (MJB)

Correspondence to: Michael J. Birrer, MD, PhD, Department of Medicine, Massachusetts General Hospital Cancer Center, 55 Fruit St, Boston, MA 02114 (e-mail: mbirrer@partners.org).

CD157 is the second member of a family of nicotinamide adenine dinucleotidases that also includes CD38. These cell surface markers are ectoenzymes that cleave extracellular nicotinamide adenine dinucleotide to produce ADP-ribose and cyclic ADP-ribose at the cell surface. They also act as receptors for CD31 (aka "platelet–endothelial cell adhesion molecule-1") expressed on endothelial cells. The signaling regulated by CD38 and CD157 may serve to regulate calcium homeostasis, as well as to modulate the function of poly-ADP ribosyl transferases involved in cell signaling, DNA repair, and apoptosis.

In this issue of the Journal (1), Ortolan et al. present an intriguing story involving CD157 that suggests a role for this protein in the progression and aggressiveness of ovarian tumors. CD157 was detected in half of human primary ovarian cancers. Expression of CD157 was higher in primary tumors than in nonadherent ascites cells and was associated with more aggressive ovarian cancer, as reflected by decreased overall patient survival. Exogenous expression of CD157 in an ovarian cancer cell line increased cell migration in a scratch assay in vitro. Conversely, inhibiting CD157 function with an antagonist antibody decreased migration, invasion, and adhesion of primary cultured ovarian cancer cells and ovarian cancer cell lines. These findings were consistent with the known ability of CD157 to interact with CD31 on endothelial and stromal cells.

What is the potential clinical utility of this observation? CD157 is anchored to the plasma membrane by glycophosphatidylinositol and can be shed into serum. In rheumatoid arthritis, high serum CD157 is associated with more severe joint destruction (2). Therefore, levels of CD157 could be potentially measured in the serum of ovarian cancer patients, as well as in the tumors themselves. CD157 is unlikely to be sensitive or specific enough as a biomarker for diagnosis of ovarian cancer but could confer increased invasive capacity to some ovarian cancers and therefore could indicate a particular line of therapy. CD157 was detected in ovarian cancers of each histology and may impart its function independent of histological classification.

Could CD157 be prognostic of outcome in ovarian cancer? In multiple myeloma and chronic lymphocytic leukemia, high expression of the related molecule CD38 on tumor cells indicates a worse prognosis (3). CD157 expression appears to be associated with poor outcome in ovarian cancer because tumors that expressed high levels of CD157 as assessed by histological scoring gave worse overall survival in comparison with those with lower levels of CD157 in this study. The effect was more profound in the tumors of serous histology. However, given the myriad of prognostic markers for this disease (CA-125, gene signatures, and cyclin E to name a few), it is not clear whether another one will greatly assist the field or the management of these patients.

It may be more valuable to look closely at the biology. It is interesting to speculate how the function of CD157 might contribute to a more aggressive ovarian cancer or to the generation of a chemoresistant phenotype. CD157 binds to its ligand, CD31, that is expressed on endothelial cells, thereby anchoring the cancer cell and possibly activating signaling in either cell type. CD157 does not have a transmembrane domain, but it interacts with other key cell surface molecules to increase their signaling (2). For example, CD31-mediated ligation of CD157 induces colocalization and tyrosine phosphorylation of focal adhesion kinase (2), which potentially triggers intracellular signal transduction downstream of src-family kinases. Similarly, binding of the family member CD38 to CD31 triggers phosphorylation of c-abl and extracellular signal-regulated kinase/mitogen-associated protein kinase (4). The focal adhesion kinase–mitogen-associated protein kinase signaling cascade could mediate increased cellular migration.

Could CD157 provide a target for therapeutic intervention? A monoclonal antibody that blocks CD157 binding to CD31 or similar counterreceptors might prevent invasion of malignant ovarian epithelial cells, but it is not likely to affect ovarian cancer cells that are in the ascites because CD157 was not expressed by the nonadherent ascites cells in the current study. If CD157 promotes invasion in the metastatic ovarian cancer cells, but is not expressed when the cells are dispersed in the ascites, it is unclear whether CD157 must be reexpressed to promote heterotypic cellular adhesion or whether the reexpression of CD157 on adherent cells is triggered by the microenvironment after the ovarian cancer cells have metastasized. Each of these possibilities in the timing of CD157 expression in the process of ovarian cancer dissemination would have different implications for the development of CD157 as a potential therapeutic target. It also remains to be explored whether the interaction between CD157 and CD31 activates the endothelial cell through CD31, and whether this triggers angiogenesis. This activity would provide solid support for CD157 as a therapeutic target.

The enzymatic activity of CD157 may be as important as its role as a receptor. When CD157 cleaves nicotinamide adenine dinucleotide to generate ADP-ribose and cyclic ADP-ribose (5), the reaction generates substrate for ADP-ribosyl transferases and polymerases. ADP-ribosylation of cell surface proteins may modulate their function and promote aggressiveness in ovarian cancer. This potential outcome of CD157 activity warrants further examination, given the recent development of poly-ADP-ribose polymerase inhibitors in the clinic. Specifically, which proteins become ADP-ribosylated at the cell surface? How is this affected by CD157 activity? How might this affect the CD31-expressing cells?

As with most interesting biomarker studies, this one generates more questions than answers. To be clinically important, these results need additional validation and better understanding from a biological and biochemical standpoint. Once these stromal–epithelial interactions are mapped out, it is likely that novel therapeutic approaches will evolve leading to more effective treatment of women with ovarian cancer.

open here please (full-text on right side of the screen):
CD157 in Ovarian Carcinoma: How Does It Help Us? -- Annunziata and Birrer, 10.1093/jnci/djq269 -- JNCI Journal of the National Cancer Institute