Alleviation of Acute Graft-Verus-Host Disease Following Regulatory Dendritic Cell Treatment in Young and Older Mice.

Abstract 3550

Poster Board III-487

Older patients experience increased severity of graft-versus-host disease (GVHD) following allogeneic bone marrow transplant and host dendritic cells (DC) are a vital initiator of this disease.

To evaluate the effect of age on GVHD severity and establish a model that can be used to study treatment of GVHD, a complete MHC mismatched bone marrow transplant (BMT) model was used to induce GVHD. Older (≥14 month) and young (3-4 month) mice succumbed to GVHD, but older mice experienced faster onset of weight loss and earlier mortality. Prophylactic treatment with syngeneic (host strain-derived) regulatory DC (DCreg) has been shown to decrease mortality of GHVD in young mice (Immunity 2003 18: 367-379). To determine whether this treatment also decreases the morbidity from GVHD in young mice and to evaluate its efficacy in older mice, day +2 BMT mice were treated with DCreg (hereafter referred to as DCreg treated BMT mice). As previously described, DCreg were generated by culture of bone marrow cells (BM) with IL-10, GM-CSF, and TGF-β1 (Immunity 2003 18: 367-379). All DCreg treated mice exhibited a brief period of GVHD and survived at least 45 days post-BMT; overall, older mice had increased morbidity and decreased long-term survival than young mice. In vivo analysis of regulatory and co-stimulatory DC molecules (PD-L1, PD-L2, CD103, CD40, CD80, and CD86) on host and donor-derived DC revealed no statistical differences between young and older DCreg treated BMT mice. Paired Ig-like receptor B (PIR B) is an inhibitory receptor expressed on DC and GVHD is exacerbated in PIR B^−/− mice (Nat. Immunol. 2004 5:623-629). An antibody that recognizes both PIR A and PIR B was used to look at expression on DC in DCreg treated BMT mice. PIR A/B expression was increased in vivo on young host DC compared to older host DC. Transferred DCreg are host-strain derived and it is unclear if the PIR A/B expression in these experiments was on residual host DC or the transferred DCreg. Future utilization of green fluorescent protein positive (GFP+) DCreg will determine the origin of upregulated PIR A/B expression in vivo. In contrast to cultured stimulatory DC, young cultured DCreg revealed increased PIR A/B expression and were negative for co-stimulatory DC molecule expression. This difference in PIR A/B expression potentially provides an explanation for the disparate outcome of older DCreg treated BMT mice.

Previous studies have implicated Treg in the induction of tolerance in DCreg treated mice (Immunity 2003 18: 367-379). However, the mechanism of tolerance maintenance has not been investigated. Abrogation of Treg mediated suppression in vivo with anti-GITR antibody (DTA-1) did not result in recurrence of GVHD or mortality in young mice, implying maintenance of the GVHD-free state is not Treg mediated in young mice.

Human transplant recipients have a compromised immune system from both the transplant and immunosuppressive drug regimen that results in increased susceptibility to infection. Upon infection, GVHD recurrence often occurs. A murine influenza infection model was used to assess the donor immune system's capacity to respond to infection and remain free of GVHD in young DCreg treated BMT mice. Forty-five days or more post-BMT, once completely recovered from GVHD and demonstrated to be allo-reconstituted, the DCreg treated BMT mice were intra-nasally infected with 0.1LD₅₀ of the H1N1 mouse adapted A/PR/8/34 influenza virus. DCreg treated BMT mice did not have a resurgence of GVHD following infection and displayed similar morbidity to control mice.

In conclusion, DCreg treatment of BMT mice results in decreased morbidity and mortality from acute GVHD in both young and older mice relative to untreated BMT mice. However, the older mice still do not fully recover and exhibit increased morbidity relative to young mice. The maintenance of tolerance observed in young DCreg treated BMT mice does not appear to be Treg mediated. DCreg treated BMT mice experience similar morbidity and mortality as control mice during influenza infection and do not experience a resurgence of GVHD, implying an intact immune response to influenza infection.

Future studies will specifically examine the requirement of PIR B on DCreg for successful inhibition of GVHD. Additional work will continue to investigate the mechanisms of induction and maintenance of tolerance in DCreg treated BMT mice.

S.M.S. is supported by the National Institutes of Health (grant #5T32 AI07485).