A better understanding of how a colon tumor progresses to metastasis could lead to improved methods of diagnosis and treatments of colorectal cancer that has spread to other organs, such as the liver or brain.
Two recently published studies by Washington University School of Medicine researchers at Siteman Cancer Center have outlined the landscape of colorectal cancer evolution from the primary tumor to metastasis. The researchers discovered the landscape of RNA genes altered in metastasis and delineated the functional role and clinical utility of a long non-coding RNA (lncRNA) gene that contributes to this process. These studies are the culmination of a cross-disciplinary research collaboration between the labs of co-senior authors, Associate Professor of Internal Medicine and Assistant Director of the McDonnell Genome Institute Christopher Maher, PhD, and Professor of Surgery and Chief of Surgical Oncology Ryan Fields, MD.
The first study, published in Nature Communications, led to the discovery of 150 long noncoding RNAs that may contribute to metastasis.
“While existing efforts have focused on protein-coding genes, our study is the first to assess the biological and clinical significance of RNA genes contributing to metastatic progression by analyzing normal tissue, colon tumors, and metastatic tumors from the same patient,” said co-senior author Maher. “Notably, we found that levels of a single RNA molecule, called RAMS11, associate with metastatic progression and is a marker of poor overall patient outcome.”
To characterize the biological function and mechanism of RAMS11, the researchers used a gene editing technology called CRISPR to turn off RAMS11 in colorectal cancer cells. The loss of RAMS11 caused the cells to become less aggressive cells which could be reversed upon reintroducing RAMS11. Furthermore, gene edited cells lacking RAMS11 injected into mice were no longer able to metastasis to the lung and liver.
“These findings demonstrate that manipulating a single RNA could alter the metastatic potential of a cell,” Maher said.
Further, the team performed a drug screen revealing that high levels of RAMS11 associate with treatment resistance to chemotherapy drugs and topoisomerase inhibitors. Subsequent experiments confirm the mechanisms by which RAMS11 activates a protein called topoisomerase II alpha (TOP2α), thereby explaining the resistance to topoisomerase inhibitors.
“We were excited to dissect how RAMS11 causes resistance to topoisomerase inhibitors in metastatic colon cancer cells.” said first author Jessica Silva-Fisher, PhD. “Given the widespread use of FDA-approved topoisomerase inhibitors across cancer types, the findings from our study may have broader clinical impact.”
Fields said, “There is a significant unmet need in clinical oncology to identify new markers of cancer that can reliably predict and stratify low- and high-risk patients. This will allow oncologists to move from a ‘one size fits all’ to a ‘personalized and precision-based’ approach that will reserve aggressive and higher risk treatments to those who need it most, sparing those who do not need it the unnecessary side effects,” he said. “We hope to explore further the ability of RAMS11 and other biomarkers to do just that.”
Although this study focused on the role of RAMS11 in colorectal cancer, the researchers also revealed high levels of RAMS11 in multiple solid tumors and therefore could potentially have a broad role in regulating cancer metastasis. To further demonstrate this, they show that RAMS11 promotes aggressive phenotypes and metastasis in lung cancer, using similar experimental techniques.
“Our lab prioritizes long noncoding RNA genes, like RAMS11, that are misbehaving in multiple cancer types which we refer to as ‘onco-lncRNAs’”, Maher said. “The consistent deregulation of onco-lncRNAs further emphasizes their essential roles in promoting cancer. In this case, RAMS11 is a critical regulator of metastasis”.
In the second study, published in Science Advances, the researchers comprehensively sequenced the genome of nearly 100 tumor samples collected from multiple regions and sites across 11 patients with metastatic colorectal cancer who underwent treatments at Siteman Cancer Center. By tracking the somatic mutations throughout the tumors within patients, the researchers were able to detail the intricate heterogeneity and reconstruct how the cancer evolves in these patients – from initiation to progression and formation of metastasis – and in the mouse models.
Colorectal cancer is a highly complex and heterogeneous entity. A tumor typically consists of multiple cancer cell subpopulations (or clones) that could exhibit distinct behaviors. While some of the clones may be more aggressive and have a tendency to move away from the primary site or possess better survival when circulating via the bloodstream giving them more potential to metastasize, some other clones may possess genetic makeups that allow them to overcome the most advanced treatments.
“Tumor heterogeneity is a challenge in treatment of advanced colorectal cancer,” Fields said. “The more complex the tumors are, the more difficult to treat them. Some of the underrepresented clones at diagnosis may be among those that are toughest to remove and subsequently fuel the emergence of treatment resistance. The ultimate goal of treating this disease is to eradicate all tumor clones, leaving no trace for them to come back”.
The researchers have outlined the intricate heterogeneity of the metastatic colorectal cancer in unprecedented detail. In contrast to previous studies that only sequenced single sites or small coverage of the tumor genome using exome or targeted sequencing, the current study, led by Ha Dang, PhD, a Washington University School of Medicine researcher at Siteman Cancer Center, utilized whole genome sequencing, coupled with exome and deep targeted validation sequencing of specimen collected from multi organs and multi tumor regions.
“Reconstruction of tumor evolution is challenging particularly in solid cancer,” he said. “Bias and under sampling the complexity of tumors are common in cancer sequencing studies due to spatial heterogeneity.”
By sampling a large number of sites and regions within individual patients, the researchers vastly reduced such bias. They were able to identify and track a large number of mutations across tumors and comprehensively reconstructed the common models of clonal evolution. The authors noticed that metastasis can be formed from multiple clones from the primary tumors, a model called polyclonal metastasis seeding. Moreover, they found in several cases that the metastasis seeding clones were shared between distinct distant metastases, yet they found no trace of these metastasis seeding clones in the primary tumors that were exhaustively sequenced using a multi-region sequencing strategy, suggesting that one metastasis could have arisen from another metastasis.
“Our study dissects the complexity of metastasis by demonstrating the cooperation of multiple cells from different subclones and that it can occur in multiple rounds of seeding that involve distinct clones,” Maher said. “These findings will impact future strategies to target, and ultimately inhibit, the establishment and progression of metastases.”
The researchers also found that very often the metastasis evolved from subclones presented at low frequency in the primary tumors. Additionally, the subclones presented at low frequency could carry critical mutations conferring treatment resistance. For example, in a patient with 17 samples sequenced, a KRAS mutation was found in a subclone presented in only about 5 percent of the cancer cells in the primary tumor and was missed in standard-of-care pathological analysis that only looked at a single tumor biopsy.
“Patients with KRAS mutations have poor response to a commonly used drug called EGFR inhibitor,” Fields said. “Knowing that the patient possesses this mutation, although at low prevalence, is useful to designing the most effective treatment plan for this patient. However, this could not have been achieved without comprehensively evaluating the patient tumors.”
In addition to outlining how colorectal cancer metastasizes, the researchers went further to study this process in the mouse model. They implanted the patient tumor specimen in mice and made them grow into a tumor called patient-derived xenograft (PDX), and subsequently characterize how the subpopulation of patient cancer cells survived and continued to evolve in the mice. Patient-derived xenografts have attracted great attention recently, as a model for precision oncology. Many drugs and treatment methods may be tested for efficacy in the PDXs before they are given to patients.
“While PDX models are a promising tool for designing personalized treatments, it is unknown whether the PDX recapitulates the patient tumor composition and behaviors,” Fields said. “This is the first time PDXs have been studied in mCRC in conjunction with matched patient tumors.”
The researchers found that the clonal complexity of the patient tumors were vastly underrepresented in the PDXs. The practice of utilizing PDXs for treatment decision-making, therefore, needs to be fully evaluated.
Taken together, these two companion studies from the collaboration between Fields and Maher provide novel insights into the biology of colorectal cancer. Their ongoing work will further explore and validate these findings and may need to novel diagnostics and therapeutics in solid tumors.
Funding: This research was supported by the Alvin J. Siteman Cancer Center Siteman Investment Program, The Foundation for Barnes-Jewish Hospital Cancer Frontier Fund, the National Cancer Institute Cancer Center Support Grant P30 CA091842, the Barnard Trust, the Washington University PDX Development and Trial Center (U54CA224083), the American Surgical Association Foundation Fellowship, the American Cancer Society Institutional Review Grant, the Society of Surgical Oncology James Ewing Foundation Clinical Investigator Award, a Sidney Kimmel Translational Science Scholar Award, the David Riebel Cancer Research Fund, a Research Scholar Grant (130878-RSG-17-058-01-RMC) from the American Cancer Society, NIH National Cancer Institute R21 (R21CA185983-01), and NIH CTSA (UL1 TR002345), the Washington University Surgical Oncology Training Grant (T32 CA009621), a NIH R35 CA197561, and the Washington University Digestive Diseases Research Core Center (P30 DK052574).