Photo by Aaron Logan
Lab mice
Researchers say they have developed a new approach to model human immune variation that overcomes the limitations of traditional mouse models.
With this approach, the team identified genetic markers that directly correlate with the outcome of inflammatory and malignant diseases in humans, including chronic lymphocytic leukemia and Burkitt lymphoma.
The findings suggest that accounting for immune diversity is critical to the success of predicting disease outcomes based on immune cell measurements.
The team reported these findings in Nature Communications.
Traditionally, researchers have relied on inbred mouse strains to gain insight into human diseases while reducing experimental noise.
“If you take a black, a brown, or a white mouse, each one will give you a different answer in the same assay,” said Klaus Ley, MD, of La Jolla Institute for Allergy and Immunology in California.
“For example, if you vaccinate them, their responses will be different, which creates a lot of experimental noise. However, when you think about patients, or even healthy people, we are all different.”
To mine those differences for information, the researchers embraced the experimental noise. Instead of analyzing a single inbred mouse strain, they turned to the hybrid mouse diversity panel (HDMP).
The HDMP is a panel of about 100 different inbred mouse strains that mirror the breadth of genetic and immunological diversity found in the human population.
The researchers studied the natural variation in the activation pattern of abdominal macrophages. Macrophages isolated from 83 different mouse strains from the HDMP were exposed to lipopolysaccharide (LPS), a major component of the outer wall of gram-negative bacteria.
“Fundamentally, when the immune system is confronted with gram-negative bacteria, it can deal with the situation in 2 ways—either it gets very angry and tries to kill the bacteria or it can wall them off in an attempt to live with it,” explained Dr Ley. “Both strategies carry a certain risk, but a long evolutionary history has ensured that mice and people can survive with either strategy.”
The LPS-induced reactions of the macrophages covered the whole spectrum—from very aggressive (LPS+) to very tolerant (LPS-), depending on the mouse strain.
Next, the researchers asked which genes were active during each type of response to identify gene signatures that correlated with LPS responsiveness.
The team then ran these gene signatures across various human gene expression data sets and discovered they strongly correlated with human disease outcomes.
For example, macrophages isolated from healthy joints were enriched in LPS-tolerant genes, whereas macrophages from rheumatoid arthritis patients were strongly skewed toward LPS-aggressive.
Since it was known that mice and people with the aggressive phenotype are better at fighting cancer, the researchers specifically asked whether the level of LPS responsiveness could predict tumor survival.
After analyzing data from 18,000 biopsies across 39 different tumor types, the team found the LPS+ gene signature strongly correlated with survival, while the LPS- signature correlated with cancer death.
The pattern was significant across many types of cancer, including chronic lymphocytic leukemia, Burkitt lymphoma, osteosarcoma, melanoma, and large-cell lung carcinoma.
