Identification of an Alloreactive CD8⁺ T Memory Stem Cell Population in Recipient Mice with Acute Graft-versus-Host Disease.

Memory CD8⁺ T cells are an important component of long-term immunity against infectious pathogens because of their higher frequency of antigen-specific CD8⁺ T cells as well as their ability to proliferate, produce inflammatory cytokines, and kill target cells more rapidly upon secondary antigen encounter than naïve CD8⁺ T cells. How the pool of memory CD8⁺ T cells is generated and maintained is key issue to understanding and perhaps manipulating long-term memory response, such as graft-versus-host disease (GVHD) where host antigens persist. Using a major histocompatibility complex (MHC)-identical but minor histocmpatibility (miHA)-mismatched mouse model of human allogeneic BM transplantation (allo-BMT), we recently identified alloreactive memory CD8⁺ T cells responsible for persistent GVHD. We found that donor CD44^hiCD62L^lo effector memory and CD44^hiCD62L^hi central memory CD8⁺ T cells recovered 42 days after allo-BMT (d42-CD8⁺ T cells) from B6 mice receiving normal C3H.SW CD44^loCD8⁺ T cells and T cell-depleted (T⁻BM) caused lethal GVHD in secondary B6 recipient mice.

Interestingly, in addition to these classical memory phenotypes, a third population of donor CD44^loCD62L^hiCD8⁺ T cells, which accounted for 2% to 6% of whole d42-donor CD8⁺ T cells, was identified in the spleens and livers of these B6 recipients with ongoing GVHD. When cultured in the presence of B6 dendritic cells (DCs)+IL-2+IL-15, these d42-CD44^loCD62L^hiCD8⁺ T cells rapidly and vigorously proliferated as compared to d42-CD44^hiCD62L^lo and d42-CD44^hiCD62L^hiCD8⁺ T cells. By day 15 following this ex vivo culture, d42-CD44^loCD62L^hiCD8⁺ T cells expanded as many as 92.0-fold, whereas CD44^hiCD62L^lo effector/effector memory and CD44^hiCD62L^hi central memory CD8+ T cells only expanded 2.1-fold and 11.0-fold, respectively. Furthermore, ex vivo stimulation of d42-CD44^hiCD62L^loCD8⁺ T cells with B6 DCs+IL-2+IL-15 only induced the generation of CD44^hiCD62L^lo effector/effector memory CD8⁺ T cells, whereas d42-CD44^hiCD62^hiCD8⁺ T cells generated both CD44^hiCD62L^lo and CD44^hiCD62L^hi cells. In contrast, d42-CD44^loCD62L^hiCD8⁺ T cells generated all three T cells subsets, e.g., CD44^hiCD62L^lo, CD44^hiCD62L^hi and CD44^loCD62L^hi CD8⁺ T cells. These data suggest that d42-CD44^loCD62L^hi CD8⁺ T cells have more potent ability than any other CD8⁺ T memory cell subsets to proliferate and differentiate into effector/memory T cells upon re-exposure to specific host miHAs as well as the ability to self-renew, resembling to the property of stem cells. Of note, these d42-CD44^loCD62L^hi CD8⁺ T cells expressed much higher levels of CD122 and CD127 than donor naive CD44^loCD8⁺ T cells. When cultured in the presence of B6 DCs+IL-2+IL-15 for 5 days, there were significantly more donor CD8⁺ T cells recovered from d42-CD4^loCD62L^hiCD8⁺ T cell cultures than that from truly naive CD44^loCD8⁺ T cells(7.9-fold vs. 2.1 fold). Thus, d42-CD44^loCD62L^hiCD8⁺ T cells found in B6 mice with GVHD are distinguishable from truly naïve CD44^loCD8⁺ T cells of normal C3H.SW mice and are responsible for sustaining the generation of both alloreactive effector memory and central memory CD8⁺ T cells. In summary, these data identify a heretofore unrecognized population of CD8⁺ T memory stem cells that may be key cellular targets for the prevention and treatment of persistent acute and chronic GVHD.