Modeling Antileukemic Adoptive Immunotherapy In Mouse-Humans Chimeras To Identify Novel Mechanisms Of Cancer Immunoediting

In spite of the documented efficacy of allogeneic Hematopoietic Stem Cell Transplantation (allo-HSCT) in the cure of Acute Myeloid Leukemia (AML), post-transplantation relapse remains an unmet clinical need. According to the “leukemia immunoediting” hypothesis relapse may be due to the outgrowth of immune-resistant leukemic variants upon the selective pressure of the transplanted immune system. We provided a clinically relevant proof-of-principle of this model in transplanted patients (Vago et al, N Engl JMed, 2009), but ex vivo studies often lack mechanistic insights, due to high inter-individual variability and lack of suitable controls. Conversely, “mouse-in-mouse” models of cancer immunoediting can provide precious and reproducible insights into molecular mechanisms, but those results are often difficult to translate into clinical practice due to species-specific factors. To overcome these limitations, here we set up a novel mouse-human chimeric model of antileukemic adoptive immunotherapy, to dissect how T cell immune pressure can sculpt leukemia gene expression profile.

Purified primary human AML blasts were infused into non-irradiated immunodeficient NOD/SCID γ-chain null (NSG) mice. Upon documentation of leukemia engraftment, mice received serial infusions of human T cells, either autologous or allogeneic (HLA-identical, HLA-haploidentical or HLA-disparate) to the leukemic cells to mimic immune pressure. Absolute counts of human leukemic and T cells were monitored weekly in mice peripheral blood. At sacrifice, leukemic cells were FACS-purified, total RNA was extracted and gene expression profile was analyzed using Illumina microarray. Deregulated genes and signatures were identified by pairwise LIMMA analysis. Gene Ontology (GO) and Gene Set Enrichment Analysis (GSEA) curated databases were interrogated to identify deregulated processes.

Infused leukemic cells stably and reproducibly engrafted into the murine bone marrow, and could be detected circulating in the peripheral blood of treated mice from three weeks after infusion. Leukemic cells exponentially expanded, could be transferred to secondary and tertiary recipients, and, in the absence of immune pressure, displayed a stable gene expression profile amongst littermates and upon serial transfer. HLA-disparate and HLA-haploidentical T cells eradicated AML from 10/10 treated mice, and complete eradication was confirmed by no disease recurrence in second transplant recipients. HLA-identical T cells granted only temporary control in 6/6 mice, while autologous T cells were completely inefficacious in 6/6 mice. Leukemic blasts subjected to T cell-mediated immune pressure showed a specific and reproducible gene signature. GO and GSEA demonstrated the selective deregulation of genes involved in immune processes. Among the top-ranked upregulated genes we identified genes related to response to interferons, comprising proteasome and immunoproteasome subunits, as well as molecules and receptors involved in antigen processing and presentation, comprising classical and non-classical HLA Class I and II molecules (HLA-DMA, HLA-DRA, CD74, B2M, TAPBP, TAPBPL).

Our findings provide further proof of the leukemia immunoediting hypothesis, demonstrating that leukemic cells modify their expression profile, and their immunogenicity, in response to T cell-mediated immune pressure, and that antigen presentation pathways represent key targets in these processes. The model we set up provides a novel and valuable tool to investigate these mechanisms in detail. Experiments with T cells genetically modified to express a suicide gene are currently ongoing, for a time-wise control of alloreactivity aimed at modeling a longer phase of equilibrium between adoptively transferred immune cells and leukemia.

Bordignon:MolMed SpA: Employment. Bonini:MolMed SpA: Consultancy.