Immunodeficient mice reconstituted with a human immune system (HIS mice) give rise to human T cells, which make them an attractive system to study human immune responses to tumors. However, a major challenge has been that HIS mice typically exhibit sub-optimal T cell responses, including failure to help B cells, respond to antigen challenge or viral infections, and develop T cell memory to antigens. However, we have demonstrated the capability of establishing T cell memory to Raji tumor.
The discovery of this phenomenon came to fruition serendipitously. We were validating a new HIS mouse model with human IL-15 (SRG-15) that exhibits increased human NK cells. We surmised that the increased NK cells would elicit greater anti-CD20 antibody-mediated killing against implanted Raji tumors in SRG-15 than HIS mice with mouse IL-15 (SRG) that have less human NK cells. Surprisingly, we noticed a spontaneous regression of Raji tumor after palpable growth in both control and experimental animals across different hematopoietic stem cell (HSC) donors and saw consistent tumor regression. This regression was dependent on HIS reconstitution since non-reconstituted mice had uncontrolled tumor outgrowth. Our paper shows how this Raji tumor control is mediated by infiltrating human T cells with subsequent established T cell memory that even control a Raji rechallenge. We worked with our molecular profiling/bioinformatics team at Regeneron to perform single cell transcriptomics concurrently sequencing of T cell receptor (TCR) repertoires. Interestingly, this was the first use of the 10x genomics platform for single cell RNAseq of human cells in our molecular profiling department. Molecular analysis of tumor-infiltrating T cells showed greater activated T cells in the tumor than concomitant spleen and allowed us to clone Raji-specific TCRs.
Raji cells can also act as stimulant to human T cells from non-tumor challenged HIS mice in vitro. We did an ELISpot to measure the IFNg activity of human PBMC derived T cells in comparison to the HIS T cells. In the paper you will see human T cells produce a significantly larger IFNg response in comparison to the HIS T cells. We co-culture anti-CD3/anti-CD28 with T cells as a positive control because they are co-stimulation molecules that stimulate T cells. However, when we add the anti-CD3/anti-CD28 to our HIS T cells, we do not see a response. In comparison when you add anti-CD3/anti-CD28 to the human PBMC-derived T cells you see a very strong human IFNg response. This illustrates the sub-optimal abilities of the HIS T cells. However interestingly, when you co-culture HIS T cells with anti-CD3/anti-CD28 and Raji, there is a strong IFNgresponse. This strong stimulation cannot be seen in Ramos, a lymphoma line, or a mouse lymphoma line A2 or an irrelevant tumor line like HEK293. The Raji cells act as third stimulus to the sub-optimal HIS T cells.
Our work inspired the work of our colleagues (Lin et al. 2023. Commun. Biol.) whom saw spontaneous tumor regression in a different tumor model – HT29, a colorectal tumor, in HIS mice. Our studies show how Raji lymphoma cells elicit both T cell-mediated tumor control. This led them to study regressing HT29 tumors and demontrate an increase in CD4 cytotoxic T lymphocyte (CTL) infiltration. Lin et al. determine the important role of human CD4 CTLs in mediating tumor clearance, independent of CD8 T cells.
The unique ability of Raji and HT29 tumor cells to elicit T cell-mediated tumor control in vivo highlights two important implications: First, the potential to use these cell lines as a platform to generate antigen-specific T cell responses in vivo with HIS mice. Indeed, our work in cloning Raji-specific TCRs emphasizes the usefulness of this platform. Second, these results highlight the need to determine what Raji or HT29-specific factors drive such potent human T cell activation in HIS mice. By determining what factors are essentially “normalizing” human T cell function in HIS mice, it will be possible to exploit such factors, e.g. genetically express them in HIS mouse models, to build a more credible in vivo HIS model that can be used to test T cell based immunotherapy.