Humanized animal models allow researchers to see how drugs work or affect humans without needing human test subjects. However, they aren’t always as accurate or as efficient as they sound. Researchers at Duke University School of Medicine, though, have developed a way to make them better. Results were published in Nature Communications on March 4.
“Adeno-associated viruses are the most popular and prominent gene therapy vectors,” said Karl-Dimiter Bissig, MD, PhD, senior author and Chen Family Associate Professor of Pediatrics. “But the problem of gene therapy remains the same: delivery.”
Adeno-associated viruses (AAVs) serve as a delivery mechanism carrying cargo into specific locations in the body, primarily, the liver.
Humanized mouse models have both mouse and human liver cells. However, much more of the AAVs end up in the mouse liver cells than the human ones — up to 80%. “This is unfortunate because that means you can’t really use humanized mice with a lot of accuracy,” Bissig said.
Bissig and team removed the murine AAV receptor, allowing the AAVs to get where they are needed: in the human liver cells.
This new model has the potential to impact a number of areas, especially clinical trials.
“It will help us identify problems before we get to testing in humans, and it will help us pick the right drug candidates for the clinical trials,” Bissig said.
Often, researchers may have many drug candidates and need to create a short list of options for clinical trials. With this improved model, they will have better data to help them pick the most promising candidates sooner.
To further test the utility of this new model, the team looked at teratoma assays. Teratomas are benign germ cell tumors, and by putting induced pluripotent stem cells under the mice’s skin, tissues form for all other parts of the body. By co-staining those tissues, researchers can identify what tissue the virus goes to.
If this works for normal human tissue, it should also work for cancer. “You can now study AAV gene therapy for genetic disease or cancer in the human context,” Bissig said, “and you could put a library of AAV onto it and see which one goes preferentially to your cancer of choice.”
This pioneering work has opened a promising new chapter in biomedical research. By refining humanized animal models to more accurately mimic human physiology, researchers can navigate drug development with greater confidence and efficiency, ultimately paving the way for more effective treatments and improved patient outcomes.