Researchers are increasingly drawn to the study of metabolism, the collective suite of chemical reactions occurring in cells, for answers to questions about cancer. A new searchable database aids that quest.
Scientists from Memorial Sloan Kettering and the Dana-Farber Cancer Institute have published a pioneering, cross-cancer analysis of changes in metabolism that occur during cancer progression. The database they created will help researchers who wish to test hypotheses about how metabolic changes fuel cancer growth. The resource is being made available publicly online.
Based on the analysis of more than 900 tumor samples across seven different types of cancer, the resulting collection is the largest of its kind and sets a benchmark for future research. A paper describing the results appears today in the journal Cell Systems.
According to MSK computational biologist Ed Reznik, one of the paper’s corresponding authors, the road map does for metabolism what genome-wide genetic studies have done for the search for cancer-related genes.
“We know that there are very potent genomic drivers that often arise in different tumor types, such as KRAS in pancreatic cancer, lung cancer, and others,” Dr. Reznik says. “Similarly, we find universal common signatures of metabolic changes that seem to happen in nearly every cancer type that we have in our data set.”
We find universal common signatures of metabolic changes that seem to happen in nearly every cancer type that we have in our data set.
But there were some striking differences too. Certain metabolic changes were associated with more-aggressive and metastatic tumors, for example. This raises the question of whether these changes might be fueling the transition to aggressive disease.
Deciphering Metabolic Traffic Patterns
To conduct their study, the team of scientists analyzed metabolic data from 11 previously published research papers. They looked specifically at measurements of tumor metabolites, or the small-molecule products of chemical reactions in cells. Augustin Luna, a bioinformatics scientist in the lab of computational biologist Chris Sander in the cBio Center at Dana-Farber, did much of the standardization work necessary to allow comparisons across the different data sets.
Studying metabolites offers a revealing glimpse into how quickly metabolic reactions in a cell are proceeding, even if what’s driving the changes isn’t always clear.
Dr. Reznik likens it to traffic patterns in Manhattan. “There might be a ton of traffic on the FDR Drive one day,” he says. “An explanation for that could be that there’s some construction work being done on the FDR, and it’s blocking traffic. But it could also be that all the way on the other side of town on Riverside Drive there is a huge accident, and all the traffic is overflowing onto the FDR Drive.”
It’s similar inside a cell, he says. The metabolite results raise interesting questions that then can be investigated further in the lab.
Usual and Unusual Suspects of Tumor Metabolism
Some of the metabolic changes found in cancers were not that surprising. For example, lactate levels tended to be higher in cancers. That makes sense because many tumors use glucose for fuel, and lactate is a byproduct of that reaction, which is called glycolysis.
Others were more unexpected. One unusual suspect was kynurenine, a derivative of the amino acid tryptophan. Kynurenine was higher in tumors compared with normal tissues. Not only that, it was also higher in more aggressive tumors than in lower-grade tumors across many cancer types, the researchers found.
Kynurenine is known to suppress the activity of certain immune cells. One possible explanation for why it is elevated in aggressive tumors is that it’s helping them evade detection by the immune system. But this is just a hypothesis, the researchers are quick to point out.
Another interesting finding from the study is that the normal tissues surrounding a tumor also sometimes display changes characteristic of the grade or stage of the adjacent tumor. “This might be pointing to what lab scientists might want to study in the tumor microenvironment,” Dr. Reznik notes.
Dr. Luna hopes the new analysis will be a jumping-off point for other scientists in the field, including those who are looking for potential new avenues for treatment. “We really see it as a benchmark or starting place for other researchers,” he says.
The searchable database can be accessed at www.sanderlab.org/pancanmet.
This work was supported by the National Resource for Network Biology from the National Institute of General Medical Sciences, the National Institutes of Health, and the National Cancer Institute.