Pancreatic Cancer Gets Lost in Translation
Through systematic testing of the protein-making machinery of pancreatic cancer cells, researchers at the Herbert Irving Comprehensive Cancer Center (HICCC) have discovered a surprising mechanism underlying the most dangerous forms of the disease. The findings may provide a new framework for identifying patients with more aggressive disease and understanding why some tumors respond differently to emerging therapies.
Pancreatic ductal adenocarcinoma is one of the deadliest human cancers, but not all cases are equally dangerous. “Under the microscope, less aggressive pancreatic tumors often retain features of pancreatic ducts, whereas tumors that lose this duct-like identity tend to be more prone to metastasis, or spreading beyond the local site,” says Christine Chio, PhD, senior author on the new paper. That suggests that the physiological program that allows a cell to become cancerous in the first place is separate from the program that subsequently enables some tumors to metastasize. Chio, an assistant professor of genetics and development at Columbia Vagelos College of Physicians and Surgeons and a member of the HICCC, has been searching for the mechanisms that push pancreatic tumors from localized growth toward metastatic disease.
Searching the genome for clues about pancreatic cancer's lethality
Working with researchers at Columbia and the University of California, San Diego, Chio’s lab used a genome-wide CRISPR screening approach to generate a pool of mouse pancreatic cancer cells in which different genes were systematically switched off. By transplanting this library into mice, the team could ask which gene losses made pancreatic tumors grow or spread more aggressively. “We took an unbiased approach, using a genome-wide screen,” says Chio, adding that “going in, our question simply was ‘are there differences in proteins needed for a primary tumor versus metastatic outgrowth?’”
Patient tumor samples show that eIF4G2 levels decrease as pancreatic cancers become more aggressive (from left to right). The finding supports the study’s discovery that eIF4G2 helps maintain a less aggressive tumor state and that loss of this protein may allow cancer cells to spread.
A protein called eIF4G2 rose to the top of the list; eliminating it yielded highly aggressive, metastatic tumors. “Finding eIF4G2 as our top hit was a complete surprise, and it made us realize that the rules we identified for the primary tumors in prior research aren’t the same as the ones for metastasis,” says Chio.
The protein eIF4G2 is a non-canonical translation factor, meaning it’s not part of the usual machinery for translating messenger RNA into proteins; only some mRNAs require it. “We identified a unique subset of mRNAs, 79 in total, whose translation depends on eIF4G2. Our data suggest that reduced translation of this program allows pancreatic cancer cells to adopt a more aggressive, metastatic state,” says Chio. The researchers then looked for this translational signature in a database of tumor profiles from patients. “From this computational analysis, we found that those tumors with this signature in patient datasets are more prone to develop metastasis, and that patients with these tumors had poorer survival,” says Chio.
Next steps: improving treatment responses
It’s a timely finding. A new class of drugs called KRAS inhibitors has shown promising results in some, but not all, cases of pancreatic cancer. “These translation rules may be useful in the context of emerging KRAS-targeted therapies, because they could help subsets of patients who may be more responsive to these inhibitors,” says Chio.
The results may also lead to better treatments for metastatic tumors, though that will require an indirect approach. As Chio explains, “eIF4G2 is a tumor suppressor, which means we will need to find ways to restore its expression in aggressive tumors.” She and her colleagues have already applied for funding to identify factors in the tumor microenvironment that suppress eIF4G2 expression. Targeting those factors might restore this tumor-suppressive translational program and make aggressive tumors more vulnerable to treatment.
Chio’s lab is also working to define the molecular features that make certain mRNAs dependent on eIF4G2. “One thing we are particularly interested in is whether there are features within the mRNA itself that determine this dependency,” she says. Defining those rules could make it possible to recognize when pancreatic tumors have lost this metastasis-restraining translational program and, eventually, to identify ways to restore it.
“The long-term goal is to understand not just which genes are present in a tumor, but how cancer cells choose which messages to translate as they become more aggressive,” says Chio.
