[Frontiers in Bioscience 3, d657-664, July 1, 1998]
Reprints
PubMed
CAVEAT LECTOR




Table of Conents
 Previous Section   Next Section

CD44 EXPRESSION AND GROWTH FACTORS

Jun-ichi Hamada1, Yutaka Sawamura2, and Erwin G. Van Meir3

1 Division of Cell Biology, Cancer Institute and 2 Department of Neurosurgery, Hokkaido University School of Medicine, Sapporo, Japan, 3 Laboratory of Molecular Neuro-Oncology, Neurosurgery Department and Winship Cancer Center, Emory University, Atlanta, Georgia, 30322, U.S.A. and 3 Laboratory of Tumor Biology and Genetics, Neurosurgery Department, University Hospital, 1011 Lausanne, Switzerland

Received 4/18/98 Accepted 5/15/98

7. THE ROLES OF CD44 IN PERITONEAL DISSEMINATION

As mentioned above, the evidence that some growth factors stimulate hyaluronan production by mesothelial cells, and that reactive mesothelium expresses CD44, leads us to consider the importance of CD44 in peritoneal or pleural metastasis of cancer (34,35,37-40). This section considers the roles of CD44 in adhesion of tumor cells to mesothelium in peritoneal dissemination.

Some in vitro studies have demonstrated that tumor cells use CD44 molecules to attach to mesothelial cells. Cannistra et al. have shown that CD44-positive ovarian cancer cells exhibit significant mesothelial binding which is partly blocked by anti-CD44 antibodies (41). Pretreatment of mesothelium with hyaluronidase also inhibits the binding of ovarian tumor cells to mesothelium, suggesting that tumor CD44 recognizes hyaluronate on mesothelium. A series of reports by Turner and his colleagues also indicate that ovarian cancer cells adhere to hyaluronic acid on mesothelial cells via CD44 (42-44).

CD44 is also involved in the peritoneal dissemination of gastric cancer cells. Human scirrhous gastric cancer cells, which have a high potential to form peritoneal dissemination in nude mice, express higher levels of CD44 than cells which do not cause peritoneal dissemination (45). The binding ability of these highly metastatic cells to mesothelial cells or hyaluronic acid is partly inhibited by treatments with anti-CD44 antibodies or hyaluronidase. The in vivo inoculation of highly metastatic cells treated with anti-CD44 antibodies results in a significant prolongation of survival time as compared to control mice inoculated with non-treated tumor cells. These observations suggest that CD44 facilitates peritoneal dissemination of this gastric cancer cell line. In an experiment using other human gastric cancer cell lines, the treatment with antibodies to both CD44 and beta1 integrin inhibits the dissemination of gastric cancer cells in the peritoneal cavity of nude mice and prolongs their survival time (39). Further, the report shows that TGF-beta1 increases the expression of CD44 in both the tumor cells and in mesothelial cells, which results in the enhancement of adhesion and invasion to the mesothelial cell monolayer. Thus, these findings suggest that CD44 is one of the mediators involved in the attachment of gastric cancer cells to mesothelial cells and that TGF-beta1 may participate in the promotion of the dissemination.

In vivo observations strengthen the importance of the interaction of tumor cells with mesothelial cells via tumor expressed CD44 (47). When mouse ovarian or mammary ascites tumor cells are injected intra-peritoneally into mice, hyaluronan accumulates at the initial site of attachment of tumor cells and cells clump to the mesenteric surface. Immunohistochemistry using anti-CD44 antibodies reveals that the great majority of the cells that initially attach to the mesentery are strongly CD44-positive. These histopathological findings suggest the involvement of hyaluronan-rich matrix in tumor cell attachment to the mesentery through interaction with tumor cell CD44.

A weakly malignant cloned cell line, ER-1, has been derived from a rat mammary carcinoma cell line in order to study factors mediating tumor progression (48). It was found that growth factors such as EGF and TGF-beta enhance the malignancy of the ER-1 cells, which is assessed by their ability to achieve peritoneal dissemination (49). Both TGF-beta and EGF enhance the in vitro invasion of mesothelial cell monolayers by ER-1 cells (table 3). However, ER-1 cells treated with TGF-beta but not EGF adhere to cultured mesothelial cells or hyaluronic acid-coated plates at a higher rate compared to non-treated ER-1 cells. Analyses by flow cytometry using anti-rat CD44 antibodies revealed the increased expression levels of the stimulation with TGF-beta but not EGF (figure 1). These findings suggest that enhanced invasiveness of ER-1 cells to mesothelial cell monolayers by TGF-beta or EGF is induced by distinct mechanisms. TGF-beta-induced, but not EGF-induced, invasiveness is probably related to the upregulation of CD44 expression.

Table 3. Enhanced effects of EGF or TGF-beta on tumorigenicity and in vitro invasion of mesothelial cell monolayers by rat ER-1 mammary carcinoma cells

CYTOKINEa

AMOUNT (ng/ml)

NUMBER OF PENETRATED COLONIES/CM2± SDb

NUMBER OF RATS WITH TUMOR/NUMBER OF RATS USEDc

EGF

0

2.3± 1.1

0/5

1

35.7± 4.5

2/5

10

95.7± 8.9

5/5

100

103.3± 10.4

5/5

TGF-beta

0

2.3± 1.1

0/5

0.1

5.6± 2.3

1/5

1

38.3± 4.1

4/5

10

63.9± 5.3

5/5

a: ER-1 cells were treated in vitro with EGF or TGF-beta for 24 hours. b: ER-1 cells were seeded on the rat mesothelial cell monolayer. Five days after the tumor cell seeding, the number of colonies per 1 cm2 formed under the mesothelial cell monolayer were counted under a phase contrast microscope. c: ER-1 cells (1x105/rat) were injected intraperitoneally into syngenetic rats. Three weeks later, the rats were killed for examination of tumor formation in the peritoneum (peritoneal dissemination).

Figure 1. Enhanced expression of CD44 on rat ER-1 mammary carcinoma cells after treatment with TGF-beta. ER-1 cells were treated in vitro with TGF-beta (left ) or EGF (right) for 24 h. CD44 expression was measured by flow cytometry using anti-rat CD44 antibodies. TGF-beta but not EGF enhances the CD44 expression on ER-1 cells in a concentration-dependent manner.

Considered together, these findings indicate that CD44 may play a key role in tumor dissemination in body cavities. Strategies to directly block CD44 function or eliminate microenvironmental factors which upregulate expression and function of CD44 and its ligand hyaluronan may result in the decreased spread of tumor cells exfoliated into body cavities.