UCSF studies illuminate possible new landscape for targeting cancer

Dual UCSF findings dramatize what scientists have just begun to recognize: While so-called oncogenic cells drive the development of cancerous tumors, other, seemingly innocent cells -- inflammatory cells - within emerging tumors can influence the fate of tumors, their aggression and their response to therapy.

The findings, both reported in recent publications, suggest that blocking the action of these 'innocent' cells - or the factors they release - may interrupt the progression of cancerous tumors at early stages of development, the researchers say.

"Tumor biology has historically focused on the cells that contain oncogenic or tumor-suppressor gene mutations. But the combined implication of our two studies suggests that we need to look beyond the mutated tumor cells and into their micro-environment, to take into account the regulation of inflammatory cells," says the lead author of one study, published in Cell, Lisa Coussens, PhD, assistant professor of pathology and a member of the UCSF Comprehensive Cancer Center.

In the Cell study, the researchers report that an enzyme known as matrix metalloproteinase-9 (MMP-9) significantly influences the progression of cancer development in mice engineered to express oncogenes, or tumor-causing genes, that predispose animals to develop squamous cell carcinomas, a common skin cancer in humans. Moreover, the researchers demonstrate that the primary source of the enzyme is not the oncogenic epithelial cells that drive the cancer, but immune inflammatory cells within the tumor's developing mass.

In the other study, published in Nature Cell Biology, the researchers demonstrate that the same enzyme is a key component of the "switch" that initiates angiogenesis, the growth of blood vessels, in mice genetically engineered to develop carcinomas of the pancreatic islets. Angiogenesis supplies emerging tumors with the nutrients and oxygen that are essential for fueling their progression beyond the pre-malignant state. And once again, the researchers detected evidence that the enzyme sparking angiogenesis was expressed not by oncogenic cells but by inflammatory cells in the tumor's micro-environment.

"Our findings reveal new factors influencing the process by which cells become malignant," says Zena Werb, PhD, UCSF professor of anatomy and a principal participant in both studies. "They show that factors originating outside the oncogenic cells are specifying whether there is progression in pre-malignant lesions, the nature of the malignancies that develop, in terms of aggressiveness, and such characteristics as what factor is promoting angiogenesis. And this is occurring in tumors containing very powerful oncogenes."

As the primary source of the MMP-9 protein appears to be inflammatory cells in both models, the discoveries could explain why other studies have suggested that anti-inflammatory drugs, such as COX-2 inhibitors (in mice) and nonsteroidal anti-inflammatory drugs and aspirin (in humans), appear to reduce the onset and progression of colon cancer. Several epidemiological studies have detected a 40 to 50 percent decrease in the risk of colorectal cancer in individuals who regularly use aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs). COX-2 inhibitors are generally prescribed to treat the pain associated with inflammatory diseases such as arthritis.

Inflammatory cells, part of the normal immune response, migrate into developing tumor lesions, apparently in response to tissue abnormalities (or "damage'') caused by oncogene expression in the tumor cells. But ironically, says Coussens, the two studies show that the inflammatory response that occurs during carcinogenesis can help boost the angiogenic response and proliferation of oncogenic cells, by releasing MMP-9.

"This is one of the first studies that very clearly demonstrates that inflammatory cells are a critical part of the tumor pathway," says Coussens.

More broadly, says Werb, the two studies suggest that, even in the absence of MMP-9, there are multiple mechanisms that influence the progression of a cancer. "This is incredibly important to recognize in terms of how human cancers behave," she says. "There are a progressive series of events occurring, and the question of what factors propel that progression haven't been addressed much.

"Now we're seeing for the first time that there are these layers of hierarchies or events involved. In these two studies we've highlighted just one that alters the tumor micro-environment."

--------------------------------------------------------------------------------------------------

In the skin carcinoma study, (Cell, Vol. 103, 481-490, October, 2000) the researchers discovered that, in most cases, MMP-9 increased the proliferation of oncogene-expressing cells, enhanced conversion of developing cancers into frank carcinomas, and reduced the frequency with which tumor cells terminally differentiated, or matured, a key step in normal cell health.

However, underscoring the complexity of cancer development, the researchers also determined that when the same type of tumors developed in the absence of MMP-9, those tumors were less differentiated, indicative of a higher grade tumor.

This conflicting aspect of the finding could explain why some MMP inhibitors have proven ineffective against late-stage cancers in clinical trials.

"It's possible that the same enzyme does the opposite thing at different stages of a tumor's growth, because its targets are different," says Werb. "Maybe the same things that help set up the tumor are then used by surveillance cells to try to do it in.

"Usually mouse model studies indicate how many mice or which mice develop tumors, or how big a tumor will get. In our study, we've gone beyond that to ask, `What is special about the 20 or 50 percent that made it?'" says Werb.

Co-authors of the Cell paper were Christopher L. Tinkle, BA, of the UCSF Cancer Research Institute, Douglas Hanahan, and Zena Werb.

The study was funded by the National Cancer Institute and the American Cancer Society. In the study published in Nature Cell Biology (Vol. 2 October 2000), the researchers demonstrate that MMP-9 is a key constituent of the "switch" that initiates angiogenesis in pre-malignant cells. They demonstrate that it does so in part by releasing the potent growth factor VEGF, or vascular endothelial growth factor, from the tumor matrix, enabling it to stimulate angiogenesis.

The researchers made this observation in mice genetically engineered to develop carcinomas of the pancreatic islets. In the study, every mouse developed islet tumors of the pancreas; only 1 to 2 percent of these tumors developed into adenomas and carcinomas. Of these, 50 percent developed hyper-proliferative islets with inactive vasculature; subsequently, a subset, 20 percent, "switched on" angiogenesis

The revelation regarding MMP-9's activation of VEGF helped solve a conundrum: The researchers had previously discovered that VEGF was found at similar levels in both normal islets and in cancerous islets, indicating that the potent growth factor has to be activated by some mechanism to actually activate angiogenesis. MMP-9, they discovered, is part of that mechanism. "VEGF is stored in the extracellular matrix, and the protease, released by inflammatory cells, makes VEGF available to its receptor by somehow releasing it from the matrix," says the lead author of the study, Gabriele Bergers, PhD, a visiting postdoctoral fellow in the laboratory of Douglas Hanahan, PhD, UCSF professor of biochemistry and senior author of the study. The study suggests that MMP-9 is not the sole inducer of angiogenesis, as 40 to 50 percent of mice still developed angiogenic islets in its absence. "Another protease or another angiogenic factor must kick in and switch on angiogenesis," says Bergers.

Moreover, while the absence of MMP-9 led to the development of fewer angiogenic islets, and therefore fewer tumors, the study indicated that another protease, MMP-2, is also involved in promoting tumor growth. The absence of MMP-2 did not affect angiogenic switching, but it did reduce tumor size.

"Specific inhibitors of MMP-9 may be particularly effective as chemo-preventive agents that target the initial angiogenic switch in quiescent normal vasculature found in pre-malignant lesions or in normal tissue beds freshly seeded with disseminated tumor cells," says senior author Hanahan.

Yet there may be hope for broader applicability as well, he says, by combining inhibitors of MMPs with inhibitors of the as-yet-undefined alternative regulators that elicit angiogenesis in the absence of MMP-9 activity. Such combinatorial attacks might be able to completely shut down angiogenesis, blocking tumor growth and rendering MMP inhibitors efficacious for treating late stage cancers."

Co-authors of the Nature Cell Biology paper were Rolf Brekken, PhD, Department of Vascular Biology, The Hope Heart Institute, Seattle; Gerald McMahon, SUGEN Inc, South San Francisco; Thiennu H. Vu, MD, PhD, UCSF Department of Medicine; Takeshi Itoh, Institute for Virus Research, Kyoto University and Shionogi Institute for Medical Science, Osaka, Japan; Kazuhiko Tamaki, Sankyo Co. Ltd, Tokyo; Kazuhiko Tanzawa, Sankyo Co. Ltd., Tokyo; Philip Thorpe, PhD, Harold C. Simmons, Comprehensive Cancer Center, University of Texas Southwestern Medical Center, Dallas; and Shigeyoshi Itohara, Brain Science Institute, Wako, Japan.

The study was funded by the National Cancer Institute.