Of Mice and Men

Will those hard-working humanized mice help get us to an AIDS vaccine? Scientists are sounding more optimistic.

By Regina McEnery

It’s difficult to imagine how an animal that fits in the palm of one’s hand could be rejiggered to behave like Uncle Harry or Aunt Jo—or, more accurately, Uncle Harry or Aunt Jo with a raging viral infection. But some mice that have been genetically engineered to lack an immune system can do just that because they can accept almost any kind of transplant. This means that they can be made to carry functioning human genes, cells, tissues, and organs, and used to study human diseases in ways that would be ethically unacceptable or technically impossible in humans.

The first humanized mice were created more than two decades ago. Since then, substantial improvements have been made to their transplanted immune systems, improving their reliability as preclinical animal models. There are now four major types of humanized mouse models being used to study everything from diabetes and autoimmunity to cancer and a wide array of infectious diseases.

But no other infectious agent has been more extensively studied in humanized mice than HIV. Though primates are still considered the best model for studying HIV infection, humanized mice have the advantage of being far less costly. As their quality improves, they are becoming integral to HIV research. They have been used, for instance, to test new HIV drugs and the systemic delivery of neutralizing antibodies—highly specific proteins that bind viruses and prevent them from infecting host cells.

In recent years, scientists have designed humanized mice that appear to recapitulate a particularly troublesome aspect of HIV infection: the persistence of HIV in reservoirs of latently infected CD4+ T cells—even after treatment has suppressed the virus to virtually undetectable levels in the blood. Such mice are likely to prove valuable to growing efforts to find a cure for HIV, which have lately focused on reactivating such latent reservoirs so that they can be targeted and destroyed. 


A bone marrow-liver-thymus (BLT) humanized mouse engrafted with tissues and cells that generate a functional human immune system. Several recent research breakthroughs suggest the humanized mouse model may be helpful in AIDS vaccine research. Photo by Vladimir Vrbanac from the lab of Harvard Medical School scientist Andrew Tager.

Humanized mouse models have also long been sought to aid in the development of an AIDS vaccine. However, limitations in the ability of these models to develop functional T-cell responses against the virus that mimic those in humans—a critical arm of a vaccine-induced response to HIV—has tempered enthusiasm for these small animal models. Moreover, difficulties in infecting humanized mice through their mucosa due to the lack of sufficient human cells in the vaginal, rectal, and gastrointestinal tracts have similarly impeded efforts to use the mice to study HIV transmission and pathogenesis.

But a series of papers published this year suggests researchers have found a way around these barriers—most notably with the creation of the bone marrow-liver-thymus (BLT) humanized mouse. Those mice took a starring role at an all-day symposium at Harvard Medical School in Boston on Nov. 5 devoted to the application of humanized mouse models to AIDS vaccine development. “The immune responses in these models are very similar to what we see in human infection,” said Todd Allen, co-chair of the symposium and principal investigator at The Ragon Institute of Massachusetts General Hospital (MGH), Massachusetts Institute of Technology (MIT), and Harvard. “But we don’t know yet how well that will play out following vaccination of these mice. The biggest limitation is that this remains a model of a human immune system in a mouse environment.”

A flurry of findings

Allen led a recent study that caused a small stir in AIDS vaccine research circles. He and his colleagues found that BLT mice infected with HIV mounted cellular immune responses that closely mirrored those observed in HIV-infected humans, and moreover that HIV also escaped from those responses in a manner very similar to natural infection. Finally, Allen and his team found that BLT mice carrying a human immune-related gene associated with enhanced control of viral replication suppressed the virus in a way that was virtually identical to how humans who express that same gene control the virus. Allen said his lab is now looking at the potential to induce human HIV-specific immune responses in the humanized mice through vaccination.

Though mice are much smaller than people, they can shed light on how HIV makes its way around the body—as was vividly illustrated by Allen’s Harvard colleague Thorsten Mempel. He and his team recently tracked HIV-infected human T cells in the lymph node of a humanized mouse using a high-tech surveillance tool called intravital microscopy. This was the first time scientists have visualized the behavior of such cells in a live animal. The study found that HIV-infected T cells migrate robustly in lymph nodes. A small subset of these infected cells derive from either multiple cell fusions or through multiple adhesions to other CD4+ T cells in the lymph node. These interactions resulted in the formation of long continuous membrane surfaces that increased the length of infected cells some ten-fold. The researchers suggest that all this may facilitate cell-to-cell transmission of the virus and promote widespread HIV dissemination. 

In yet another study, scientists injected humanized mouse muscle cells with a modified viral vector optimized for the production of various broadly neutralizing antibodies (bNAbs)—those that target a broad range of HIV’s many genetic variants. They found that the antibodies prevented infection even when the animals were challenged with high doses of HIV. Alex Balazs, a researcher in David Baltimore’s lab at California Institute of Technology, where the experiments were conducted, said it remains to be seen whether the results seen in BLT mice can be replicated in humans. “History has shown us that humans don’t behave like mice,” said Balazs. “We have to be prepared for surprises.” 

Humanized mice are contributing to research on novel therapies as well. Rockefeller University scientist Michel Nussenzweig has been testing cocktails of potent bNAbs as a therapy in humanized mice infected with HIV. He and his team have found that giving a single bNAb or even as many as three did not produce durable results; the virus rebounded weeks after the antibody treatment ceased. But when they increased the number of bNAbs used, the virus had still not rebounded in seven of the eight mice after two months. Researchers suspect that the expanding arsenal of more potent antibodies might improve the chances of this strategy working and, if so, might provide an alternative to the daily grind of antiretroviral therapy.

The origins of BLT

The BLT mouse was initially developed by virologist Victor Garcia-Martinez, who is now at the University of North Carolina, in conjunction with a team at the University of Minnesota. Scientists make the mice by surgically implanting them with human organoids, which are fetal liver and thymic tissue that mimic organs—in this case organs that are essential to the development of immune cells. The mice are then irradiated and given transplants of stem cells taken from human fetal livers. These cells take up residence in the bone marrow, establishing a source for the human immune system borne by BLT mice. Mice altered this way were found to have a wide range of human immune cells in their peripheral blood; the cells also infiltrated tissues and organs in the lungs, GI tract, and liver, just as they would in the human body. 

Garcia-Martinez and his team showed that these mice developed human T cells at a furious pace after being injected with the bacterial toxin that causes toxic shock syndrome, or Toxic 1—one sign that their immune systems were similar to that of humans. The researchers also measured the amount of time it took the mice to produce cytokines and found that it corresponded with the time taken to induce human inflammatory responses. 

But the transplanted BLT immune system is not identical to a human’s. One challenge, for example, is that antibody-producing cells, known as B lymphocytes, don’t mature properly in the bodies of these mice. Dale Greiner, a University of Massachusetts scientist who has authored two reviews on the impact of humanized mouse models on the study of human disease, said this may be because the lymphoid organs in such mice are disorganized. It is in these organs that the immune responses are amplified and refined, especially those involving the production of neutralizing antibodies—which are these days a central focus of HIV vaccine research. 

In humans, he said, all of the components are “where they need to be.” In humanized mice, “it is like walking into a warehouse, where everything is scattered.” Greiner says that the genetic engineering required to remove the immune system in these mice, so that it can be replaced by a human one, might inadvertently disrupt the genes required to “organize” their lymphatic system in an immunologically functional manner. 

Still, researchers are optimistic about the future of humanized mice in AIDS vaccine research and appear to believe the BLT model, in particular, can be tweaked and improved to that end. “What I think would really catalyze the field,” said Andrew Tager, a Harvard Medical School scientist who collaborated with Allen on his recent study, “is if there could be funding for a consortium to focus on making this a better model with an eye toward answering more questions about HIV. How can we make the immune responses of the model even better? Do we need to put more human genes in the mice? We have shown we are on track. The time is now.”