In This Issue

Blocking the interaction of programmed death-1 (PD-1) on tumor-specific CD8⁺ T cells with PD-1 ligand has shown promising results in clinical trials for the treatment of cancer. Whereas most studies to date have focused on the effects of PD-1 inhibition on CD8⁺ T cells that recognize immunodominant determinants (IDDs) of tumor Ags, no studies have investigated whether PD-1 targeted therapies affect CD8⁺ T cells that recognize subdominant determinants (SDDs). In this issue, Memarnejadian et al. (p. 3348) used an in vitro model in which priming of B6 mice with SV40-encoded T Ag⁺ tumor cells (C57SV) generates CD8⁺ T cell clones that recognize four well-defined peptide epitopes of T Ag. Although PD-1 expression was observed on both SDD-specific (site I) and IDD-specific (site IV) CD8⁺ T cells, PD-1 expression was significantly higher on site I–specific CD8⁺ T cells, and anti–PD-1 tripled the numbers of these CD8⁺ T cells while only minimally increasing the site IV–specific CD8⁺ T cell population. This expanded population expanded at a late time point (day 6 after C57SV injection) and was fully functional, maintaining both IFN-γ production and MHC I–restricted cytotoxicity of SDD-pulsed splenocytes. The authors also showed that PD-1 blockade increased site I–specific CD8⁺ T cells responses in cohorts of B6 mice that were immunized with cell lines that cannot form peptide:MHC I complexes, supporting previous observations that cross-priming plays a significant role in antitumor immunity. Further immunization studies using peptide-pulsed DCs, recombinant vaccinia viruses encoding full-length T or truncated T Ag, and tumor cells expressing variants of T Ag demonstrated that anti–PD-1 treatment enhanced SDD-specific T cell responses not by promoting T cell proliferation, but by preventing site I–specific CD8⁺ T cell lysis, thereby relieving suppression by immunodominant CD8⁺ T cells. In summary, this work demonstrates a role for PD-1 in suppression of subdominant CD8⁺ T cell responses and shows that interference with this checkpoint can broaden antitumor responses. The therapeutic benefit of this broadened response may be further augmented, as SDD CD8⁺ T cells may be more likely to escape tolerance mechanisms, thereby contributing to long-term protective immunity.

Bim and Nur77 are proteins that have been characterized as regulators of thymocyte fate in response to high-affinity TCR signaling, in part by promoting apoptosis and contributing to central tolerance through clonal deletion. Hu et al. (p. 3147) examine Nur77 during development of MHC class I–restricted thymocytes in the presence and absence of Bim to better understand its function independent of its proapoptotic role. They observed that Bim and Nur77 each contribute to double-positive and single-positive thymocyte development by examining knockout mice lacking Bim, Nur77, or both proteins. Nur77-deficient mice had a higher proportion of Foxp3⁺CD25⁺CD4⁺ regulatory T cells (Tregs) in the thymus compared with wild-type (WT) mice, as did Bim^−/−Nur77^−/− mice compared with Bim^−/−Nur77^+/+ mice. The frequency and number of anergic phenotype (Foxp3^–FR4^hiCD73^hi) CD4⁺ T cells was also significantly higher in Bim^−/−Nur77^−/− mice compared with single knockouts or WT mice, suggesting that Nur77 expression negatively regulates the development of tolerance-associated cells. Reconstitution of irradiated Rip-mOva mice with Nur77 and Bim WT or deficient OT-I bone marrow was used to assess the impact of Nur77 on CD8⁺ T cell tolerance to a model self-antigen. In these chimeras, Nur77 deficiency did not significantly affect clonal deletion in the absence or presence of Bim, but only Bim^−/−Nur77^−/− (and not Bim^−/−) mice developed diabetes, indicating a role for Nur77 in regulating tolerance. These data reveal multiple roles for Nur77 that are critical to promoting self-tolerance, functions that are independent of its role in clonal deletion, and may include modulating the quality or strength of TCR signaling to impact Treg selection in the thymus and T cell activation in the periphery.

Activation-induced cytidine deaminase (AID) is known to be involved in Ig gene somatic hypermutation (SHM) and class switch recombination (CSR). AID activity is tightly regulated to prevent unintended mutations, and previous studies have shown that efficient SHM can be induced in germinal center (GC) B cells or B-like cell lines but not in ex vivo–activated B cells or lymphoma cells. In this study, Liu et al. (p. 3023) examine factors that induce SHM and CSR under in vitro conditions to better understand this disparity in AID activity. They observed that ectopic expression of AID in ex vivo–activated Aicda^−/− but not Aicda^+/+ splenic B cells promoted SHM in the Ig V gene, and that the extent of SHM did not correlate with high AID expression. Ectopic expression of AID was able to induce CSR to IgG₁ in Aicda^−/− B cells, but this occurred at a lower efficiency compared with Aicda^+/+ B cells. In confirmation, ectopic expression of AID promoted efficient SHM in a CH12 B lymphoma cell line only after it was modified to be AID deficient, and these cells were also able to undergo CSR, albeit at a lower level of efficiency compared with CH12 cells with endogenous AID. Taken together, these results show that exogenous AID can induce SHM in the absence of endogenous AID in ex vivo–activated primary B cells and B lymphoma cells and suggest that SHM is impacted by the spatiotemporal distribution and/or posttranslational modifications of endogenous AID.

Human roseoloviruses, including human herpesvirus (HHV)6A, HHV6B, and HHV7, are pervasive human pathogens that typically persist as latent viruses but can be reactivated under immunosuppressive conditions. A murine roseolovirus (MRV) was identified recently as a homolog of human roseolovirus that infects the mouse thymus, and in this issue, Patel and Yokoyama (p. 3212) show that control of MRV infection in neonatal mice requires CD8⁺ T cells. MRV infection of neonatal C57BL/6 (B6) mice caused severe thymic necrosis and destruction of nearly all CD4 single-positive (CD4 SP) and CD4⁺CD8⁺ double-positive (DP) thymocytes, but these subsets recovered to levels similar to those in uninfected mice by 7 wk postinfection. Unexpectedly, MRV viral burdens were higher in CD8 single-positive (CD8 SP) and CD4⁻CD8⁻ double-negative thymocytes than in CD4 SP or DP thymocytes. CD8 SP cells within infected thymi showed evidence of an effector T cell phenotype, including expression of activation markers, intracellular effector molecules, and a restricted TCR repertoire, whereas CD8^−/− mice showed higher viral burdens and greater lethality following MRV infection compared with B cell deficient mice. Together, these results show that MRV infection of B6 mice has the potential to be used to better understand roseolovirus immunopathogenesis in B6 background mutant mouse strains, and reveal CD8⁺ T cells as a critical element in protective immunity.