Induction of Alloantigen Specific CD4 T Cell Anergy by B7: CD28 Mediated Costimulatory Blockade Significantly Alters the Functional Program of Bystander CD8, Monocytes, NK and B Cells.

Retention of antigen (Ag) specific immunity to pathogens and tumor cells in the context of adequate control of alloreactivity remains a challenge for allogeneic hematopoietic stem cell transplantation (HSCT). Global or subset depletion of T cells achieves control of alloreactivity at the cost of loss of non-allospecific repertoire. Induction of Ag specific anergy has been explored as a way to selectively impact alloreactivity. Anergy is induced when Ag is presented to CD4 T cells without CD28:B7 mediated costimulation. We and others have shown that CD4 anergy results from a positive signaling cascade, rendering Ag specific T cells unable to proliferate and produce cytokines on Ag specific rechallenge. When this concept was translated to a haploidentical HSCT clinical trial, all donor marrow harvest mononuclear cells were subjected to inhibition of CD28 mediated costimulation by CTLA-4-Ig, a fusion protein binding both B7.1 and B7.2 costimulatory molecules. In our ongoing clinical trial, unfractionated donor peripheral blood mononuclear cells (PBMC) are cultured with allostimulators in the presence of costimulatory blockade (CSB) provided by anti-B7.1 and anti-B7.2 humanized monoclonal antibodies. To understand the impact of inducing alloAg specific CD4 T cell anergy on bystander PBMC, we investigated the effects on CD8 T cells, monocytes, B and NK cells. Mimicking our ex vivo clinical anergization protocol, primary MLR using PBMC from fully HLA mismatched healthy donors (n=12) were performed in the presence or absence of CSB. CSB resulted in 73% mean inhibition of proliferation after 72 hrs of primary MLR. CD4 and CD8 T cells, monocytes, NK and B cells from these MLRs were isolated by negative selection as were control unmanipulated CD4 and CD8 T cells. From each population, RNA was extracted and global gene expression profiling performed using Affymetrix human 133plus2 chips. Unsupervised analysis was performed using DNA-Chip Analyzer and supervised analysis using Significance Analysis for Microarrays. While the frequency of alloreactive CD4 T cells in human PBMC is only 1–10%, we observed global gene expression variance (436 genes) between unstimulated CD4 cells vs. CD4 cells isolated from MLR in the presence or absence of CSB (P≤0.05). Even more surprising was the impact of CSB on bystander CD8, monocytes, NK and B cells. Between cells from MLR with or without CSB, there were 632 differentially expressed genes in CD8 T cells, 105 differentially expressed in NK cells, 85 differentially expressed in monocytes and 1781 in B cells (P≤0.05 for all populations tested). We observed not only expected alterations (e.g, in inflammatory cytokines and receptor mediated signaling) but also observed changes [e.g. in proteasome degradation (i.e. CD4, CD8, NK and B cells) and in expression of genes regulating cell motility (i.e. CD8 T and NK cells)]. These results demonstrate that induction of alloAg specific anergy within a small population of alloreactive CD4 T cells results in dramatic alterations of the genetic repertoire of bystander cells. Thus, blockade of the CD28:B7 costimulatory pathway not only inactivates alloreactive CD4 T cells but also alters the functional program of other immune cells that will repopulate the host post-HSCT.