Effector CD8 + T cells are important mediators of adaptive immunity, and receptor-ligand interactions that regulate their survival may have therapeutic potential. Here, we identified a subset of effector CD8 + T cells that expressed the inhibitory fragment crystallizable (Fc) receptor FcgRIIB following activation and multiple rounds of division. CD8 + T cell-intrinsic genetic deletion of Fcgr2b increased CD8 + effector T cell accumulation, resulting in accelerated graft rejection and decreased tumor volume in mouse models. Immunoglobulin G (IgG) antibody was not required for FcgRIIB-mediated control of CD8 + T cell immunity, and instead, the immunosuppressive cytokine fibrinogen-like 2 (Fgl2) was a functional ligand for FcgRIIB on CD8 + T cells. Fgl2 induced caspase-3/7-mediated apoptosis in Fcgr2b + , but not Fcgr2b À/À , CD8 + T cells. Increased expression of FcgRIIB correlated with freedom from rejection following withdrawal from immunosuppression in a clinical trial of kidney transplant recipients. Together, these findings demonstrate a cell-intrinsic coinhibitory function of FcgRIIB in regulating CD8 + T cell immunity.

SUMMARYSince the discovery of the CD40-CD154 costimulatory pathway and its critical role in the adaptive immune response, there has been considerable interest in therapeutically targeting this interaction with monoclonal antibodies in transplantation. Unfortunately, initial promise in animal models gave way to disappointment in clinical trials following a number of thromboembolic complications. However, recent mechanistic studies have identified the mechanism of these adverse events, as well as detailed a myriad of interactions between CD40 and CD154 on a wide variety of immune cell types and the critical role of this pathway in generating both humoral and cell-mediated alloreactive responses. This has led to resurgence in interest and the potential resurrection of anti-CD154 and anti-CD40 antibodies as clinically viable therapeutic options. KeywordsTransplantation; CD40; CD40L; thromboembolism; costimulation; Fc receptor A myriad of studies over the past two decades have revealed the importance of CD40-CD154 interactions in the generation of alloreactive responses, and as such it represents an extremely attractive target for therapeutic intervention in transplantation. This pathway requires the interaction of the CD40 molecule with its ligand, CD40L (CD154). These molecules belong to the tumor necrosis factor (TNF) superfamily and are expressed on a wide range of tissues and cell types, with CD40 constitutively expressed on antigenpresenting cells (APC), including B cells, macrophages and dendritic cells (DC), and CD154 inducibly expressed mainly on CD4 + T cells and endothelial cells following activation [1]. The first descriptions of this pathway focused on the importance of CD40-CD154 interactions in promoting T cell-dependent humoral responses [2], but it quickly became apparent that these interactions were also essential for generating cell-mediated immunity. Seminal studies showed that the crosslinking of CD40 on dendritic cells by CD154 on activated CD4 + helper T cells led to "licensing" of the APC, leading to the upregulation of

| INTRODUC TI ONAdvances in HLA typing and HLA antibody testing over the past two decades transformed our ability to assess donor/recipient compatibility in the context of organ transplantation. Beginning from serologic donor/recipient HLA matching through emphasis on avoidance of preformed donor-specific HLA antibodies (DSAs), current approaches delve into amino acid sequences of HLA alleles, assessing the so-called molecular mismatch between donor and recipient, using one or more of the currently available tools: HLAMatchmaker, [1][2][3][4] Amino-Acid comparison, Electrostatic Mismatch Score (EMS 5-7 ), or PIRCHE-Predicted Indirectly ReCognizable HLA Epitopes presented by recipient HLA-Class II antigens. 8-10 A short description of each method is provided in the supplemental material.It is notable that only minor variations were found when comparing the ability of these different approaches to predict poor graft outcome (correlation ranging between R 2 of .85-.96). 8,11 In addition, when different patient populations were studied (eg, pediatric vs adult; kidney vs lung), different cutoff values were reported, 12-19 indicating the need Funding information Paul I Terasaki Research FoundationMolecular mismatch load analysis was recently introduced as a means for performing risk stratification following organ transplantation. However, although good correlation was demonstrated between molecular mismatch load and generation of de novo donor-specific HLA antibody (DSA), quite a few exceptions exist, and the underlying factors that define HLA immunogenicity remain unclear. Herein, we present a new paradigm to interrogate differences between molecular mismatches that lead to the generation of de novo DSA and those that do not (the 2MM1DSA cohort).Specifically, patients transplanted across 2 HLA-DQ mismatches, who formed de novo DSA only to one mismatch (foe) but not the other (friend), provide a unique environment in which patient-specific factors that affect the immune response other than immunogenicity, such as infection and immunosuppression, can be controlled for. It further permits focusing on mismatches uniquely exhibited by the de novo DSA allele, rather than mismatches shared by both DSA and non-DSA alleles. This concept paper illustrates several examples, highlights the need for center-specific or population-specific cutoff values for posttransplant risk stratification, and mostly argues that if there is no direct correlation between molecular mismatch load and immunogenicity, then molecular mismatch load must not be adopted as an approach for equitable organ allocation. K E Y W O R D Salloantibody, antigen presentation/recognition, clinical research/practice, histocompatibility, organ allocation

Currently, the ability to predict or monitor the efficacy of HLA antibody-removal therapies is deficient. We previously reported that titration studies are a consistent and accurate means of assessing antibody strength. To test whether titration studies can also predict which patients are better candidates for desensitization, we studied 38 patients from 3 centers (29 receiving plasmapheresis/low-dose intravenous immunoglobulin [IVIg]; 9 patients receiving high-dose IVIg). For patients undergoing plasmapheresis/low-dose IVIg, antibody titer reduction correlated with number of treatment cycles for both class I and II antibodies but only up to approximately 4 cycles. Reduction in titer slowed with additional cycles, suggesting a limit to the efficacy of this approach. Furthermore, initial titer (predesensitization) can guide the selection of candidates for successful antibody-removal treatment. In our experience, patients with antibodies at an initial titer >1:512 could not be reduced to the goal of a negative lymphocyte crossmatch, corresponding to a 1:16 titer, despite a significant increase in the number of treatment cycles. Change in mean fluorescence intensity (MFI) value did not correlate with success of treatment if initial MFI values were >10 000, likely due to single antigen bead saturation. Overall, we present a potential prognostic tool to predict candidacy and a monitoring tool to assess efficacy of desensitization treatment.