The Human Touch
Research Models
Paula Miliani de Marval

The Human Touch

Can mice with functional human immune systems help push forward the field of immunotherapy?

The first humanized mice were developed more than three decades ago by engrafting severely immunodeficient rodent models with humanhaematopoietic cells. They soon proved useful in the evaluation of antiviral drugs for HIV, a uniquely human pathogen that disrupts and eventually destroys CD4+ T cells, central players in the body's immune system.

But are humanized mice, which harbor human immune cells that can be induced in response to infection, also ideal models for the testing of compounds that manipulate the immune system to fight cancer? That is one of the hopes as drug developers pursue alternative models for a relatively novel class of drugs known as immune checkpoint inhibitors. These compounds work by blocking the molecules that switch off immune cells, thus increasing tumor immunogenicity.

The earliest humanized mouse models, which relied on strain CB17 severe combined immunodeficiency (SCID) mice, had a number of drawbacks. For one, they were only able to produce human immune cells for a short duration, in part because the mice still generated natural killer cells—a component of the innate system—that ultimately led to the production of mouse immune cells. The human immune responses sustained by later humanized mouse models are much stronger, however.

What this means for oncology, and specifically the field of immune-oncology, is that humanized mice may be uniquely valuable in helping developers choose, with greater confidence, which compounds to move forward to clinical trials.

Indeed, recent findings by our lab, presented today at the American Association of Cancer Research meeting in New Orleans, suggest that humanized mice could be a reliable player in the development of new immunotherapies.

The research

About a year ago, our laboratory began utilizing CD34+-NSG humanized mice, which are produced by injecting CD34+ hematopoetic stem cells into NOD scid gamma (NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ) or NSG mice, to evaluate the impact of immune checkpoint inhibitors in tumor growth. CD34+-NSG mice can maintain a functional human immune system for over a year, providing a suitable model for our research.

To assess the utility of the humanized mouse model, we implanted the human RKO colon carcinoma cell line, chosen because it expresses high levels of anti-PD-L1, into CD34+-NSG mice. The CD34+-NSG mice HSCs matched the human leukocyte antigen (HLA) haplotype of the RKO cells.

Then we assessed how well the model recapitulated anti-tumor responses modulated by checkpoint inhibitors—how well it performed—by measuring the efficacy of two recently approved drugs that target different receptors on the surface of cancer cells. The first, pembrolizumab, targets the programmed cell death-1 receptor and the second, ipilimumab, targets the protein cytotoxic T-lymphocyte-associated protein 4 (CTLA-4).

The results from these studies revealed that the RKO colon carcinoma tumors are responsive to pembrolizumab and/ or ipilimumab therapy alone or in combination resulting in a significant tumor growth inhibition. Yet combining the two drugs did not boost the efficacy. Flow cytometry analysis showed the targeted CD4+ and CD8+ populations have reduced level of expression of PD-L1 indicating successful therapy.

So to better understand how these checkpoint inhibitors were impacting distinct cell populations in the humanized mice, we used flow cytometry to analyze the distribution of various T cell, B cell and NK cell in the peripheral blood, spleen, and tumor tissue of the humanized mice. We found that the checkpoint inhibitors enhanced the effector functions of interferon-gamma producing CD4+ and CD8+ T cells.

While the idea of using a humanized model to test how a particular immunotherapy works has a definite appeal, the lack of myeloid derived cells in this model may impact the final outcome of these therapies. New transgenic models in the NSG (and NOG as well) are now available which show a more complete myeloid and lymphoid lineage, which need to be tested as suitable models for immune-oncology.

Additional health concerns among these animals and the high cost of these humanized mouse (between US$1,200 to $1,500 compared to just $30-$50 for a syngenic mouse) may account for the fact that syngenic models, which have a complete and functional immune system, continue to gain favor. Additional research and perhaps more optimization will be required before the humanized mouse model starts to gain more traction.

How to cite:

Miliani de Marval, Paula. The Human Touch. Eureka blog. Apr 18, 2016. Available: