Blog Post June 21, 2018

Preclinical testing of a new type of immunotherapeutic using JAX NSG mice


The complexity of cancer includes the interaction among multiple cell types, complex signaling pathways and numerous molecules secreted by the immune system and surrounding tissues. Many of these complex interactions can be harnessed to combat cancerous cells, while others can hinder the development of an effective immunotherapeutic. Typically, once a new therapy has been identified and proven effective through in vitro testing, the next phase of testing involves an appropriate in vivo platform. A recent study published in Nature Communications by Ravi and colleagues, explored a new type of therapeutic developed by the authors—a so-called Y-trap that targets immune checkpoints, CTLA-4 or PD-L1, and TGFβ signaling in the target cell microenvironment. For their in vivo testing, the authors used tumor-bearing NSG™ mice from JAX co-engrafted with HLA matched human immune cells. The humanized NSG™ mice support the growth of human cancer cell lines and patient-derived xenografts (PDX; Shultz et al., 2014 and Wang et al. 2018).

Building a better Y-trap

The authors hypothesized that their bifunctional antibody-ligand traps, targeting either CTLA-4 or PD-L1 fused to the C-terminus of the heavy chain specific for the TGFβRII ectodomain sequence, would activate T cells while simultaneously deactivate autocrine and paracrine TGFβ signaling in the microenvironment. By disrupting TGFβ-mediated differentiation of T regulatory cells (Tregs) and immune tolerance, they could generate a more effective immunotherapeutic strategy.

The authors first validated the correlation between TFGβ activation and FOXP3 expression, and then performed in vitro testing of their engineered therapeutic using human peripheral blood mononuclear cells (PBMCs). For their in vivo studies, they chose to use NSG™ mice from JAX which they co-engrafted with HLA-A2+ human CD34+ bone marrow cells and matched human tumors. They tested their therapeutic, α-CTLA4-TGFβRII, using a human melanoma cell line and a patient-derived xenograft. The α-CTLA4-TGFβRII Y-trap showed greater suppression of tumor growth than either α-CTLA4 or α-TGFβRII alone. Following treatment, immunophenotype analysis of remaining tumor tissue revealed the Y-trap treatment significantly diminished the presence of CD4+CD25+FOXP3+ Tregs in the tumor microenvironment.

In addition to analyzing Tregs, the authors evaluated the effect of their antibody on IFNγ expression in CD8+ T cells found in the tumor. They observed that Y-trap-treated A375 melanoma tumor-bearing mice exhibited a significant reduction in tumor progression and had a significant increase in tumor reactive IFNγ-expressing CD8+ cells. Furthermore, in PDX-bearing mice, they also confirmed antibody efficacy: the authors observed higher tumor-reactive IFNγ-expressing CD8+ cells and increased the percentage of CD4+ and CD8+ T cells with central memory phenotypes. These results indicated that the Y-trap antibody effectively activated CD8+ cytotoxic T cells and blocked CD4+ T cells from responding to the autocrine/paracrine TGFβ signaling in their immediate microenvironment.

The authors then examined the use of their antibody against cancers that are unresponsive to current checkpoint inhibitors, such as triple-negative breast cancer (TNBC), which has an increased risk of metastases. Using NSG™ mice bearing the TNBC cell line that expresses elevated PD-L1, TGFβ, and shows enhanced lung metastases, the authors observed that treatment with their immunotherapeutic resulted in significant inhibition of tumor progression, reduction of Tregs and lung metastases, while increasing tumor-reactive IFNγ-expressing CD8+ cells, and CD4+ and CD8+ central memory T cells.

In addition to examining the CTLA-4 checkpoint inhibitor, the authors tested a bifunctional therapeutic developed against PD-L1. The α-PDL1-TGFβRII Y-trap antibody evaluated against either A375 or TNBC-bearing mice reconstituted with human CD34+ HSC exhibited inhibition of tumor growth, diminished FOXP3+ expressing Tregs and increased the percentages of tumor-reactive IFNγ-expressing CD8+ cells. In addition to immunological efficacy, the authors also observed normal liver function and body weight in treated mice during the study indicating a lack of toxicity.

What about these JAX NSG™ mice?

Using mice as a model organism for human disease has several advantages, such as size and both genetic and physiological similarity to humans. While choices of research mice were previously limited to nude mice, more immuno-deficient strains, such as the NSG™ mice from JAX, are more effective to study tumor pathogenicity and drug efficacy.

At JAX, we offer the non-obese diabetic (NOD)-scid gamma mouse (NOD.Cg-prkdcscidIl2rgtm1w1; NSG™), which as noted by Hosur and colleagues, is the most widely used immunodeficient Il2rgnull mouse model that supports the development of a human hematopoietic and immune system (Hosur et al., 2017). This type of humanized mouse model has been relied on to study tumor progression, metastasis, and other immunological functions (Wang et al., 2018). These mice are the gold standard host for engraftment of human tumors or the establishment of human immune components following hematopoietic stem cell transplantation.


The biology of cancer involves several pathways and mechanisms that help evade the hosts’ immune system and spread cancer cells to other organs or tissues. In the study by Ravi and colleagues, they engineered a therapeutic, a so-called Y-trap, that targets either CTLA4 or PD-L1 and is fused to a TGFβRII ectodomain sequence, which disrupts autocrine or paracrine TGFβ signaling within the targeted microenvironment. They used tumor-bearing, humanized NSG™ mice from JAX engrafted with either human melanoma or TNBC for their in vivo testing. They observed a reduction in tumor-infiltrating Tregs, as well as an inhibition of tumor progression. Based on their in vivo and in vitro study observations, they concluded that the engineered therapeutic was an effective strategy against cancers, and warrants further testing.

JAX offers a wide range of mouse models, including highly advanced, genetically modified strains of mice that are capable of engrafting both human immune system cells and human tumors, as well as strains lacking specific immune cell types. These model systems are vital in analyzing therapeutic mechanism of action that promotes immune mediated attack on cancer cells.

Researchers can also freely access the Mouse Tumor Biology (MTB) database, which catalogs multiple spontaneous and genetically engineered mouse models of human cancer. The database provides a wealth of information, including tumor frequency and latency data, genomic data, pathology reports and images, associations between models and scientific literature, and links to additional online cancer resources. The MTB database also includes a section on PDX tumors available from JAX and data that includes mutational patterns in specific cancers and genes that are commonly mutated across a spectrum of cancer types.

The extensive mouse model portfolio from JAX for in vivo testing includes humanized NSG™ and NSG™-SGM3, as well as tumor-bearing Onco-Hu® mice, which are powerful preclinical models for studying immuno-modulators, either alone or in combination with other treatments. Additionally, JAX recognizes that some applications may require generating specific models on the most appropriate background strain. The JAX model generation service makes use of techniques such as CRISPR/Cas9 to develop the most translationally-relevant mouse models, including custom genetically humanized models that express a specific human immune target of interest.

Visit the JAX immunology page to learn more about how JAX products and services empower your immunology research.

Recommended Resources

Mouse Tumor Biology Database

Humanized Mice Page

Guide: Humanized Mice for Immuno-Oncology


Hosur V, Low BE, Avery C, et al.,2017. Development of Humanized Mice in the Age of Genome Editing. J Cell Biochem. 118(10):3043-3048. doi: 10.1002/jcb.26002 [PMID: 28332231]

Ravi R, Noonan KA, Pham V., et al., 2018. Bifunctional immune checkpoint-targeted antibody-ligand traps that simultaneously disable TGFβ enhance the efficacy of cancer immunotherapy. Nat Commun. 9(1):741. DOI: 10.1038/s41467-017-02696-6

Shultz LD, Goodwin N, Ishikawa F, et al., 2014. Human cancer growth and therapy in immunodeficient mouse models. Cold Spring Harb Protoc. 1(7):694-708. doi: 10.1101/pdb.top073585 [PMID: 24987146]

Wang M, Yao LC, Cheng M, et al., 2018. Humanized mice in studying efficacy and mechanisms of PD-1-targeted cancer immunotherapy. FASEB J. 32(3):1537-1549. DOI: 10.1096/fj.201700740R