“By blocking this secretory protein with an antibody, we can treat the tumor by targeting multiple cellular types – fibroblasts and blood vessels -- in the tumor microenvironment,” Mok says.
The researchers are taking steps to advance the approach to clinical trials, including commercial manufacture of the antibody and additional toxicity studies.
“As with digital medicine, we are seeing more such marriage of data analysis and modeling with experimental biology to derive new understanding of disease mechanisms and novel drug targets,” the co-corresponding author, Stephen T.C. Wong, Ph.D., P.E., professor and associate director at Houston Methodist Cancer Center, says.
Targeting fibroblasts with antibodies
Mok’s research focuses on cancer-associated fibroblasts – supportive cells that make and maintain connective tissue, or stroma, that provides a scaffolding for organs. Stroma in tumors has been shown to block distribution of drugs to cancer cells. Mok has been working closely with Wong to study the tumor microenvironment. Last year, they repositioned an old drug targeting fibroblast-tumor crosstalk for ovarian cancer.
Research by Mok and colleagues had previously shown that MFAP5 promotes angiogenesis – blood vessel formation – and its overexpression is associated with poor prognosis in ovarian cancer. They also showed in preclinical research that down-regulating the protein in fibroblasts suppressed tumor growth and angiogenesis.
The team set out to find whether an antibody approach might successfully block MFAP5 to treat the disease. They developed monoclonal antibodies to MFAP5, identifying three with strong affinity for the protein and choosing clone 130A due to its specificity for a peptide sequence common to both human and mouse MFAP5.
They applied the antibody to blood samples from healthy individuals and to age-matched pre-surgical patients with advanced high-grade serous ovarian cancer. Samples from the cancer patients had much higher levels of MFAP5, confirming the antibody recognizes the protein and is useful for detecting circulating MFAP5.
Cell line experiments showed that ovarian cancer cells cultured with ovarian fibroblasts treated with MFAP5 had significantly increased motility, activity that was suppressed when those cells were treated with any of the three antibodies. The antibodies also suppressed activity of endothelial cells, which line the interior of blood vessels.
High MFAP5 tied to reduced survival in pancreatic cancer
Before launching experiments in mouse models, the team evaluated the feasibility of MFAP5 as a potential target for pancreatic cancer, which has a dense fibrous component. They stained tissue samples from 64 patients and found MFAP5 virtually absent from surrounding tissue but significantly higher in pancreatic cancer fibroblasts.
Analysis of 91 pancreatic cancer patients showed those with high stromal expression of the protein had median survival of 8 months, compared to 31 months for those with low expression of MFAP5.
Human cell line experiments showed that MFAP5 had the same energizing effect on pancreatic cancer cells as it did in ovarian cancer cells, and the three antibodies had a similar impact blocking that effect.
The team used the 130A antibody against mouse models of ovarian and pancreatic cancer derived from human tumor xenografts and found that treated mice had much lower tumor burdens than controls.
Combining the 130A antibody with paclitaxel chemotherapy significantly reduced ovarian tumor burden in mice, and experiments showed the likely cause of this effect was suppression of angiogenesis.
Fibrotic gene expression associated with shorter survival
To evaluate the prognostic significance of the fibrotic microenvironment in ovarian cancer, the team ran correlation studies of fibrotic genes in cancer-associated fibroblasts and patient survival. They found patients whose cancer stroma had the fibrotic gene signature had median survival of 19 months, while those with low levels of fibrotic gene expression had median survival of 33 months.
Mok and colleagues identified 176 genes whose expression was positively correlated with MFAP5 expression in cancer-associated fibroblasts.
Further analysis of the 176 genes by Wong and colleagues at Houston Methodist using their proprietary systems biology modeling program identified a novel collagen-enriched signaling network.
Two key fibrosis-related genes, COL1A1 and COL11A1, were found to be correlated with expression of MFAP5. Fibroblasts treated with the antibody showed lower levels of the two collagen genes.
In the mouse experiments, the researchers noted no observable toxic effects by the antibody.
“Targeting stroma is another approach to treat cancer,” Mok says. “These cells have less genetic heterogeneity than tumor cells, so they’re less likely to adapt to treatment. Maybe every patient can respond.” Mok and Wong are now working on synthesizing humanized antibodies for animal studies in preparation for clinical trials.