Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


Mice with Human Liverlets Test New Drugs

Mice with human livers are better models for spotting drug-drug interactions than are mice with only their own livers.

“Scientists at Stanford have produced mice with human brains, pigs with human blood flowing through their veins, and a human born to mice parents and mice with human heads.”

So wrote a student summarizing the “Genetically Modified Organisms” chapter of my human genetics textbook a few years ago. Two of the four comments are true, sort of.

I’ve got a new example for her, also from Stanford: Using Chimeric Mice with Humanized Livers to Predict Human Drug Metabolism and a Drug-Drug Interaction, published in the Journal of Pharmacology and Experimental Therapeutics.

It’s a terrific advance, the making of mice just human enough to provide a new way to test drug toxicities before clinical trials get underway. The researchers tested a drug against hepatitis C. The work is important because mice and humans, despite our genomic similarities, do not metabolize many drugs in the same ways. And that can be costly, even tragic.

Ironically, I‘d read about the chimeric mice right after blogging here about a 1893 paper on the source of chimerism in human mothers, and was all set with the news – and then superstorm Sandy delayed publication of the paper. So now I’m back to the new mice.

When a drug fails, it’s often a metabolite – the chemical compound that forms as the drug breaks down – that’s at fault. If a mouse doesn’t make the same metabolites as a human, a drug candidate can test as safe in the rodent, yet poison a person.

“Humanizing” a mouse might make it possible to not only text toxicity of a drug, but also monitor drug-drug interactions, which happen frequently because so many of us take more than one medication. Sometimes drug-drug interactions – DDIs – don’t emerge until a drug is approved and enough people take it, while also taking something else.

Over the past few years, several groups have created mice that have bits of humanity, and used them first to test familiar chemicals, not necessarily drug candidates. But sometimes remaining mouse liver cells obscure the effects of the tested compounds.

I like to trace the trajectories of the “breakthroughs,” and in the mouse-tox story, two papers from early summer 2011 stand out (apologies if I’ve left anything out).

First, Alexander Ploss, PhD, a virologist at Rockefeller University and colleagues, created immunodeficient mice transgenic for two human immune system genes (CD81 and occludin) that enable hepatitis C to infect the rodents, which the viruses don’t naturally do. The mice in the just-published study needed to actually get hepatitis C in order to test the drug.

In a second paper, “Humanized mice with ectopic artificial liver tissues,” Sangeeta Bhatia, MD, PhD, a biomedical engineer at MIT and colleagues fashioned “human ectopic artificial livers” URL (HEALs). These implants consist of half a million human hepatocytes, mouse fibroblasts, and human liver endothelial cells to send the appropriate hormonal signals. It’s all packaged into a 20-millimeter long plastic scaffold, with pores that keep out immune system cells. The mice function for several weeks as if they harbor a human liver.

In the new work, a team from Stanford and Genentech, with collaborators in Turkey, Australia, New Zealand and Japan, transplanted human hepatocytes into the livers of 8-week-old immunodeficient mice, but added an old biotech trick from the 1980s: making the mouse liver cells display a molecule from herpes simplex type 1 viruses.

Exposing the animals – called humanized TK-NOG mice — to a drug that only kills virally-infected cells destroys their livers. “This enables transplanted human liver cells to develop into a ‘human organ’ with a characteristic 3-dimensional architecture and gene expression pattern, which could be stably maintained for a 6-month period,” explains Gary Peltz, MD, PhD, professor of anesthesiology, pain and perioperative medicine at Stanford and senior author.

The test case was clemizole, an antihistamine used widely in the 1950s and 1960s that also blocks hepatitis C replication. “The drug tends to accumulate in the liver, which is not ideal for a general-purpose antihistamine but could be very attractive for a virus like HCV that only infects the liver,” says Jeffrey Glenn, MD, PhD, associate professor of gastroenterology and hepatology, microbiology and immunology at Stanford and hepatitis expert on the team.

Clemizole is even more powerful in the presence of a second drug, the protease inhibitor ritonavir, which slows breakdown of the first, a common type of drug-drug interaction. In addition, a breakdown product of clemizole called M1 makes the drug stick around longer, and adds to the antiviral activity, but mice make only trace and transient amounts of it. The effect is missed in mice that have their own livers.

A slice of liver from a chimeric mouse shows islands of human tissue, indicated by the dashed line. (Courtesy Dr. Gary Peltz)

The researchers infected 8 humanized mice and 8 controls, then gave them both drugs. The more human the livers – from 13-70 percent — the more M1 appeared. And the ritonavir made the clemizole hang around longer. “Although this is only one example, it indicates that it is likely that the use of chimeric mice could improve the quality of pre-clinical drug assessment,” the researchers conclude in their paper.

The new mice provide an alternative to abandoning a promising drug in preclinical testing, or letting an unsafe one proceed to clinical trials or enabling a drug-drug interaction to appear only after the numbers build with marketing.

And I can add a new example to my textbook: mice with human livers.



Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top