Modifying a Xenogeneic Gvhd Model to Allow for Use of CRISPR-Treated Primary Human T Cells

Background:

Allogeneic hematopoietic stem cell transplantation (alloHSCT) is a curative treatment for high-risk leukemia, immunodeficiencies, and bone marrow failure syndromes. Therapeutic use of alloHSCT remains limited by acute graft-versus-host disease (GVHD), where activated donor T cells attack and destroy host tissues. We have previously shown that GVHD-causing T cells increase activation of AMP-activated protein kinase (AMPK), a cellular energy sensor, and that T cell-specific ablation of AMPK in murine models decreases GVHD severity.

To study human T cell biology, we modified a previous xenogeneic model. Current models transplant whole peripheral blood mononuclear cells (PBMCs) from healthy human donors into lightly irradiated immunodeficient NOD-scid IL2Rgamma ^null (NSG) mice. However, in our hands, CRISPR-treatment of primary human T cells requires up to 10 days of culturing to obtain sufficient cells for transplantation. Therefore, prior to assessing GVHD severity using AMPK-deficient human T cells, we optimized the xenogeneic GVHD model to allow time for the manipulation and subsequent injection of CRISPR-engineered cells.

Aim:

To develop a xenogeneic model compatible with CRISPR/Cas9 manipulated T cells to determine whether changes in human T cells decrease GVHD severity similar to what is seen using murine T cells.

Results:

We first demonstrated that expanded T cells alone cause minimal xenogeneic GVHD, but that disease could be significantly facilitated with addition of non-T cell antigen presenting cells (APCs). Transplanting T cells plus non-T cell APCs increased numbers of human CD45+CD3+ T cells recovered on day 25 post-transplant (Figure 1A-B) and elevated levels of human interferon (IFN)-γ (Figure 1C). Liver sections from recipients of T cells + APCs subjectively had more perivascular infiltrates than mice receiving T cells alone (data not shown). Additionally, as seen in murine T cells, xenogeneic human T cells increased fatty acid oxidation, additional evidence that our model recapitulates the murine findings (Figure 1D).

We next wished to fix the number of co-administered APCs and optimize the number of activated T cells to reliably reproduce xenogeneic GVHD without inducing overt toxicity. To accomplish this goal, we performed xenogeneic transplants with serially decreasing numbers of expanded human T cells (starting with 6×10 ⁶/recipient) and administered 1×10 ⁶ APCs in all cohorts. We also trialed inclusion of recently thawed non-T cell APCs in place of freshly derived cells. Reassuringly, the number of human CD45+CD3+ cells recovered on day 25 post-transplant remained proportional to the number of cells injected (Figure 2A-B), as did levels of human IFN-γ (detected by serum ELISA (Figure 2C). These data indicate that robust xenogeneic GVHD can be induced with as few as 2×10 ⁶ expanded T cells and 1×10 ⁶ autologous APCs, with a concomitant increase in GVHD-associated proinflammatory cytokines. Importantly, these data also demonstrate that 1×10 ⁶ APCs are sufficient to cause reproducibly severe disease and may be recovered from a cryopreserved source.

Finally, we wished to extend the assessment of GVHD severity following administration of varying doses of donor T cells. Serum from recipient mice on day 25 post-transplant was analyzed for the production of murine-derived cytokines via a LEGENDplex assay. Of 13 cytokines tested, both murine MCP-1 and TNF-alpha were proportional to the number of T cells injected (Figure 3A-B), with the MCP-1 result confirmed by ELISA (data not shown). Thus, both MCP-1 and TNF-α, cytokines commonly implicated in acute GVHD pathogenesis, provide additional host-derived soluble factors that can be utilized to quantitate the severity of GVHD in our modified xenogeneic model. Expression of these proteins will serve as valuable biomarkers in the assessment of xenogeneic GVHD using CRISPR-treated cells.

Conclusions:

We have successfully adapted a xenogeneic model of GVHD using in vitro expanded T cells and cryopreserved APCs, thereby allowing for expanded testing of genetically manipulated human T cells in an in vivo model. Future studies will compare the necessity of genes in human donor T cells using CRISPR-mediated gene editing and compare CRISPR techniques with novel pharmacological inhibition using this modified xenogeneic approach.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal