A researcher’s data is only as good as the experimental techniques used to obtain those results. And a Stanford University study published yesterday in Cell Reports, calls into question the accuracy of a widely used method in mice that helps scientists gauge the human immune system’s response to stem cell-based therapies. The findings, funded in part by CIRM, urge a healthy dose of caution before using promising results from these mouse experiments as a green light to move on to human clinical trials.

Humanized mice aren’t quite human. Illustration: Pascal Gerard

Immune rejection of stem cell-based products is a major obstacle to translating these therapies from cutting-edge research into everyday treatments for the general population for people. If the genetic composition between the transplanted cells and the patient are mismatched, the patient’s immune system will see that cell therapy as foreign and will attack it. Unlike therapies derived from embryonic stem cells or from another person, induced pluripotent stem cells (iPSC) are exciting because scientists can potentially develop stem cell-based therapies from a patient’s own cells which relieves most of the immune rejection fears.

But manufacturing iPSC-derived therapies for each patient can take months, not to mention a lot of money, to complete. Some patients with life-threatening conditions like a heart attack or stroke don’t have the luxury of waiting that long. So even with these therapies, many researchers are working towards developing non-matched cell products which would be available “off-the-shelf. In all of these cases, immune-suppressing drugs would be needed which have their own set of concerns due to dangerous side effects, like serious infection or cancer. So, before testing in humans begins, it’s important to be able to test various immune-suppressing drugs and doses in animals to understand how well a stem cell-based therapy will survive once transplanted.

But how do you test a human immune response to a human cell product in an animal? Believe it or not, researchers – some of whom are authors in this Cell Reports publication – developed “humanized mice” back in the 1980’s. These mice were engineered to lack their own immune system to allow the engraftment of a human immune system. Over the years, advances in this mouse experimental system has gotten it closer and closer to imitating a human immune system response to transplantation of mismatched cell product.

Close but no cigar, it seems.

The team in the current study performed a detailed analysis of the immune response in two different strains of humanized mice. Both groups of animals did not mount a normal, healthy immune response and so they could not completely reject transplants of various human stem cells or stem cell-based products. Now, if you didn’t know about the abnormally weak immune response in these humanized mice, you might conclude that very little immunosuppression would be needed for a given cell therapy to keep a patient’s immune system in check. But conclusively making that interpretation is not possible, according to team lead Dr. Joseph Wu, director of Stanford’s Cardiovascular Institute:

Joseph Wu. Photo: Steve Fisch/Stanford University

“In an ideal situation, these humanized mice would reject foreign stem cells just as a human patient would”, he said in a press release. “We could then test a variety of immunosuppressive drugs to learn which might work best in patients, or to screen for new drugs that could inhibit this rejection. We can’t do that with these animals.”

To uncover what was happening, the team took a step back and, rather than engrafting a human immune system into the mice, they engrafted immune cells from an unrelated mouse strain. Think of it as a mouse-ified mouse, if you will. When mouse iPSCs or human embryonic stem cells were transplanted into these mouse, the engrafted mouse immune system effectively rejected the stem cells. So, compared to these mice, some elements of the immune system in the humanized mouse strains are not quite capturing the necessary complexity to truly reproduce a human immune response.

More work will be needed to understand the underlying mechanisms of this difference. Other experiments in this study suggest that signals that inhibit the immune response may be elevated in the humanized mouse models. Dr. Leonard Shultz, a pioneer in the development of humanized mice at Jackson Laboratory and an author of this study, is optimistic about building a better model:

“The immune system is highly complex and there still remains much we need to learn. Each roadblock we identify will only serve as a landmark as we navigate the future. Already, we’ve seen recent improvements in humanized mouse models that foster enhancement of human immune function.”

Until then, the team urges other scientists to tread carefully when drawing conclusions from the humanized mice in use today.