When fighting off an infection, the immune system can sometimes have an extreme response to the pathogen and end up turning on body tissue itself.
There is no cure or treatment for this condition, called sepsis, and the results can be deadly: more than 1.7 million people die from sepsis in the U.S. each year.
Researchers like Assoc. Prof. Nicolas Chevrier at the University of Chicago Pritzker School of Molecular Engineering (UChicago PME) are working to understand the mechanisms behind sepsis—with the hope that that information could ultimately be used to develop new treatments.
Using advanced technologies, Chevrier and his team found that a certain kind of enzyme was found in the gut of mice with sepsis. Further study showed that the enzyme essentially chewed open the membranes of red blood cells in the body, releasing toxic chemicals and causing anemia.
In fact, human sepsis patients with higher levels of this enzyme in their blood had worse outcomes overall, the team found. The results were published in the Journal of Clinical Investigation.
“What’s surprising is that this specific enzyme has never been associated with sepsis, and it has never been associated with gut biology,” Chevrier said.
“This study not only provided new insights into sepsis but also demonstrated that our technologies can be used to uncover novel findings in complex biological processes like sepsis,” said Michihiro Takahama, a former postdoctoral fellow in Chevrier’s lab who led the research. Takahama is now an assistant professor at the University of Osaka.
Finding ‘self-venom’ in the gut
The research team began the study by conducting spatial transcriptomics of a whole mouse—a technique that was made possible through a system developed in the Chevrier lab. The technique combines high-resolution microscopy and genetic sequencing to examine gene expression across molecules, cells, tissues and organs of a whole mouse, allowing the team to gain insights into how disease affects an organism across the entire body.
When the research team used their technique on a mouse model with sepsis, they found that more than 10,000 genes were regulated in sepsis across the body. They used computational methods to narrow down which might be the most important.
That’s how they found PLA2G5, a gene that encodes an enzyme. PLA2G5 is part of a family of enzymes found in both mammalian tissue and in snake and bee venom. The enzymes break open lipids—an action that can be both helpful and harmful, depending on the situation.
What was surprising is that the team found PLA2G5 in the colon and small intestine, a place where it hadn’t been seen before. To find out what effect it might have on the body, the team treated a sepsis mouse model with an antibody against PLA2G5.
They found that when this enzyme was blocked, the mouse model had reduced biomarkers in the blood that are associated with organ injury. “Mice treated with an antibody also had a higher survival rate,” Takahama said.
The team also examined the gene expression of the mouse model that had been treated with the antibody. Interestingly, it had more macrophages that are known to help clear up red blood cells.
Further tests showed that the PLA2G5 chewed through red blood cells and released their toxic material into the blood, causing tissue damage.
“People who have snake bites sometimes have anemia from an attack on their red blood cells, so we thought it was surprising that an enzyme in your own body could have the same effect—a sort of ‘self-venom,’” Chevrier said.
A potential biomarker for sepsis outcomes
To understand whether it might have the same effect on humans, the team tested blood samples from COVID-19 patients who had sepsis. Those with higher levels of PLA2G5 had worse outcomes.
“That suggests that PLA2G5 levels might also have predictive value for sepsis prognosis,” Takahama said.
Next, the team hopes to further examine the enzyme to better understand where exactly it comes from and whether it attacks other cells in the body beyond red blood cells.
“We need to better understand the mechanism and whether it triggers other problems, as well,” Chevrier said.
Other authors include Krysta Wolfe, who provided the blood samples; Makoto Murakami, who provided lipidomic analysis, Steven Dudek, who provided the mouse models; as well as Gabriella Richey, Madison Plaster, Anna Czapar, Fabian Hernandez, Denis Cipurko, Tatsuki Ueda, Yoshimi Miki, Yuki Nagasaki, Yoshitaka Taketomi, Tatsuya Saitoh, and Tadafumi Kawamoto.
Citation: “Secreted phospholipase PLA2G5 acts as a hemolytic factor in sepsis,” Takahama et al. Journal of Clinical Investigation. May 1, 2026. DOI: 10.1172/JCI195001.
Funding: National Institutes of Health