The ICOS:ICOSL Costimulatory Pathway Plays an Important Role in GVHD and Bone Marrow (BM) Graft Rejection.

ICOS, a CD28/CTLA-4 family member, is expressed on activated T cells. ICOS Ligand, a B7 family member, is constitutively expressed on B cells, monocytes and some T cells. Through the use of blocking anti-ICOS mAb and ICOS deficient (−/−) mice, we found that ICOS:ICOSL interactions play an important role in GVHD and BM graft rejection. Anti-ICOS mAb (given d-1 to d28 post BMT) significantly delayed or reduced mortality at 2 different T cell doses in a full MHC-disparate GVHD model. ICOS−/− T cells led to delayed or reduced mortality at 3 different cell doses compared to wild-type T cells. ICOS−/− CD4+ or CD8+ T cells infused into class II- or class I-disparate recipients, respectively, revealed that ICOS:ICOSL interactions regulate both CD4+ and CD8+ T cell alloresponses. Anti-ICOS inhibited GVHD in a CD28-independent fashion. Anti-ICOS inhibited GVHD mediated by either stat 4−/− or stat 6−/− T cells indicating that the ICOS pathway regulates both Th2 and Th1-mediated GVHD. In contrast to blockade of the B7:CD28/CTLA-4, CD40L:CD40 or the OX40:OX40L pathway, anti-ICOS mAb inhibited GVHD even when delayed until d5 post BMT, a time when substantial T cell expansion has occurred. A TCR transgenic model of GVHD was used to further study effects of ICOS:ICOSL blockade. All CB6 F1 recipients of anti-host alloreactive 2C CD8+ and TEa CD4+ T cells succumbed to GVHD mortality by d18 after transfer of cells. In contrast, 88% of anti-ICOS-treated mice survived long-term. Evaluation of spleens early after transplant revealed that anti-ICOS mAb reduced the number of TEa CD4+ cells by 44% and 2C CD8+ cells by 83%. Green fluorescent protein (GFP) 2C CD8+ and GFP TEa CD4+ T cells were infused into irradiated CB6 F1 mice and irrelevant or anti-ICOS mAb was administered. Mice were imaged on d4, 7 and 12 after T cell transfer. By d7, pronounced infiltration of GFP+ cells was noted in the peripheral and mesenteric LN, spleen, Peyer’s patches (PP), skin, gingiva, liver, kidney, lung, ileum, and colon of GVHD control mice. In contrast, there were fewer GFP+ cells in the spleen, ileum, colon, kidney, lung, skin and gingiva of anti-ICOS-treated mice, although there was no decrease in GFP+ cells in LNs or PP. To study the role of host ICOS expression in BM graft rejection, wild-type or ICOS−/− mice were sublethally irradiated and given allogeneic BM and evaluated for donor chimerism at 6 weeks post BMT. Five of 10 wild type mice engrafted (ave − 26% donor) in contrast to all 10 of ICOS−/− mice (ave − 71% donor). Collectively, these data indicate that ICOS:ICOSL interactions play an important role in GVHD, whether mediated by CD4+ Th1 or Th2 T cells or CD8+ T cells. Importantly, blockade of ICOS:ICOSL after initiation of alloresponses inhibited GVHD, in contrast to blockade of other costimulatory pathways, suggesting that the ICOS pathway may be a novel therapeutic target in primed transplantation situations. Anti-ICOS interfered with expansion of donor T cells in the spleen early after transplant and reduced the number of effector cells in several GVHD target tissues. These data suggest this pathway may be indicated for therapeutic targeting for the inhibition of GVHD and BM graft rejection.