Recipient Hematopoietic, but Not Non-Hematopoietic, Antigen-Presenting Cells Are Indispensable for CD4+ T-Cell-Mediated Xenogeneic Graft-Versus-Host Disease

Background

As a major drawback of conventional GVHD models using inbred mouse strains, they are not suitable for assessing the safety and effectiveness of current human-specific therapies. To overcome this limitation, xenogeneic GVHD models using highly immunodeficient mice are currently being developed and are widely used to study human diseases. However, the underlying mechanisms of human immune responses against host are not fully understood. We have previously reported that human CD4⁺ T-cells and, to a lesser extent, CD8⁺ T-cells play a critical role in the induction and development of xenogeneic GVHD. (Kawasaki et al. Blood Abstract 2016 #701). Donor antigen-presenting cells (APCs) were not necessarily required for T-cell activation. In this context, the present study focused on the immunobiology of CD4⁺ T-cell-mediated GVHD, and defined the individual role of host hematopoietic, and non-hematopoietic, APCs in inducing human immune responses in NOD/Shi-scid-IL2rg ^null (NOG) mice.

Methods

A xenogeneic GVHD model was established by intravenously injection of human CD4⁺ T-cells into NOG mice. Bone marrow (BM) chimeras were generated as summarized in Figure 1A. Briefly, BM cells from wild-type (MHC^+/+) or MHC^−/− NOG mice, respectively, were transplanted into other mice, in which only non-hematopoietic cells including target organs, or only hematopoietic cells including APCs, respectively, were deficient in MHC molecules. After human CD4⁺ T-cells were injected into these BM chimeras, survival and clinical GVHD scores were monitored.

Results

Control mice (MHC^+/+ → MHC^+/+) uniformly developed progressive GVHD, and they all died within 12 weeks following the injection of human CD4⁺ T-cells. The former chimeras (MHC^+/+ → MHC^−/−) showed similar GVHD severity and mortality, whereas the latter chimeras (MHC^−/− → MHC^+/+) showed significantly less body weight loss and lower clinical GVHD scores than the control (Figure 1B, C), which resulted in longer survival (Figure 1D). These results indicate that host hematopoietic, but not non-hematopoietic, APCs play a critical role in the development of xenogeneic GVHD.

To verify these results, we then analyzed T-cell subsets in the lungs and spleen of NOG mice one week after the injection of purified human naïve (CD45RA⁺CCR7⁺) CD4⁺ T-cells. Flow cytometric analysis showed a markedly increased percentage of central memory (CD45RA⁻CCR7⁺) and effector memory (CD45RA⁻CCR7⁻) CD4⁺ T-cells in the lungs and spleen of MHC^+/+ mice. Notably, the memory differentiation was also observed in MHC^−/− mice. However, the proportion of terminally differentiated effector memory (CD45RA⁺CCR7⁻) CD4⁺ T-cells was significantly increased in MHC^−/− mice, suggesting that a certain number of effector T-cells failed to develop a memory phenotype due to a lack of antigen-specific stimulation. Splenectomized NOG mice also showed rapid memory differentiation of CD4⁺ T-cells in the lungs during early GVHD. These findings suggest that neither MHC molecule expression on host tissues nor host hematopoietic APCs are required for the memory differentiation of CD4⁺ T-cells in lungs during early GVHD in xenogeneic chimeras.

Focusing on memory T-cells during early xenogeneic GVHD, we revealed that most memory CD4⁺ T-cells in lungs were less activated and showed a slowly proliferative phenotype (CD69^low and CFSE^low) compared to those in spleen. Intracellular IFN-γ production was also decreased in these cells. These findings suggest that most memory T-cells in lungs have less potential to cause sustained inflammation, and are therefore less associated with the pathogenesis of GVHD. Increased susceptibility of T-cells to peripheral apoptosis appears to be one of the characteristic properties of spontaneous proliferation. Consistent with this, we revealed that the memory CD4⁺ T-cells in lungs showed a significantly decreased expression of Bcl-2 compared to those in the spleen, indicating that these cells spontaneously acquire the memory phenotype, but not in response to non-hematopoietic APCs.

Conclusions

Our data clearly indicate that the immunobiology of xenogeneic GVHD is highly dependent on human CD4⁺ T-cells and mouse hematopoietic, but not non-hematopoietic, APCs. We believe that our findings provide a better understanding of the immunobiology of humanized mice and support the development of novel options for the prevention and treatment for GVHD.

View largeDownload slide

View largeDownload slide

Close modal