Housing Mice At Sub-Thermoneutral Temperatures Influences Severity Of Gvhd In Mouse Models

While a very high proportion of humans develop serious and fatal GVHD following bone marrow transplantation (BMT), mice typically do not develop lethal graft versus host disease (GVHD) when receiving bone marrow alone, even when the donor and recipient mice are MHC mismatched. Thus, the major pre-clinical model in which novel treatments for GVHD are being tested has a fundamental resistance to GVHD. Researchers therefore add donor-derived T cells from spleen or lymph node in order to generate a similar incidence and severity of GVHD as seen in humans, thus using a model system that has fundamental differences from the clinical situation.

Recent work from our lab has revealed that the standard cool (sub-thermoneutral) temperature in which all laboratory mice must be housed (creating a chronic condition of metabolic cold stress) results in impaired CD8⁺ immune function, while bolstering the action of immunosuppressive cells. Current work suggests that the excessive stimulation of cold-stress induced sympathetic nerve activity, and subsequent production of norepinephrine is the cause of this immunosuppressive environment. Because GVHD depends upon the activation and function of donor CD8⁺ T cells in the host environment, we hypothesized that mouse models do not develop expected amounts of GVHD following a bone marrow only transplant because the host environment, which is metabolically cold-stressed, blocks the function of effector T cells.

To test our hypothesis we first housed BALB/c mice at standard (22°C) and thermoneutral temperatures (30°C) for two weeks. To generate a MHC disparate model, we lethally irradiated the mice and transplanted C57BL/6 bone marrow without additional splenocytes. We used both weight loss and survival following BMT as a readout for GVHD severity. As expected, under standard conditions of housing at 22° mice were able to recover from the transplant and return to their initial weight within the same 17 day post-transplant timeframe (percent initial weight at 22°: 101% ± 1%; N=10). In support of our hypothesis we saw weight loss in the mice maintained at 30° as early as 17 days post-transplant (percent initial weight at 30°: 93% ± 2%; N=10). Interestingly, weight loss continued in the mice maintained at thermoneutrality (percent initial weight loss 48 days post-transplant at 30°: 0.84 ± 0.03, N=10). Beginning 49 days post-transplant some mice at thermoneutrality began to succumb to GVHD. Thus, housing mice at thermoneutrality decreased survival suggesting a greater incidence/more severe GVHD (Figure 1; p=0.01, N=10). As our previous data in solid tumor models suggests that numbers of immunosuppressive myeloid derived suppressor cells (MDSCs) are increased when mice are metabolically stressed at 22°, we next examined differences in this cell population in the spleens of transplanted mice. We found an obvious trend towards decreased splenic MDSCs in mice housed at 30° compared to 22°.

View largeDownload slide

View largeDownload slide

Close modal

To retest this hypothesis in a clinically relevant minor antigen mismatch model we housed male C57BL/6 mice at ST or TT for two weeks. We then lethally irradiated the mice and transplanted female 129/SvJ bone marrow. Again, we found a greater weight loss (p=0.02; N=10) in the mice receiving bone marrow only maintained at 30° compared to 22° as early as 28 days post-transplant (percent initial weight at 22°: 103% ± 2%, N=10; percent initial weight at 30°: 96% ± 2%, N=10). We also found increased numbers of MDSCs in the spleens of mice housed at 22° compared to 30° (p=0.09, N=10).

The increasing number of patients receiving BMT for various malignancies clearly highlights the critical importance of improving mouse models for understanding GVHD as this problem remains a major barrier to achieving greater survival for patients. These data demonstrate the existence of a fundamental relationship between cool standard housing temperature used in all research institutes and a resistance to GVHD, which is likely linked to a reduced immune response and/or increased immunosuppression. Achieving a better understanding of this potential confounding environmental factor could help to improve the use of mouse models for identifying the best therapies to attenuate or eliminate GVHD in humans, while at the same time, promoting effective immune reconstitution and graft versus tumor effect.