Memory CD8+ T cells mediate immune surveillance to protect against secondary infections in nonlymphoid organs. To assess the involvement of these cells in immune surveillance of the brain, Young et al. (p. 1192) investigated CD8+ T cell trafficking upon systemic Listeria monocytogenes infection. During infection, Ag-specific CD8+ T cells migrated into the brain in a CD49d/VCAM-1–dependent manner. Whereas Ag-specific CD8+ T cells that migrated to the spleen differentiated into memory cells, those in the brain consistently expressed early effector cell markers. CD8+ T cells in the brain protected against intracranial rechallenge with Listeria; however, their protective ability was shorter lived in the brain than in other organs. During primary infection, Ag-specific CD8+ T cells were found in the choroid plexus, and upon intracranial rechallenge, they migrated to cerebrospinal fluid-filled areas and, to a lesser degree, the brain parenchyma. Stimulated Ag-specific cells injected directly into the parenchyma could migrate to the periphery, where they persisted long term and differentiated into memory cells. Interestingly, the cells that remained in the brain did not differentiate but instead were gradually deleted. Thus, effector, rather than memory, CD8+ T cells may mediate brain immune surveillance, thus protecting the brain from the potential undesirable side effects of resident memory T cells.

Interaction between the chemokine receptor CXCR4 and its ligand, SDF-1 (stromal cell-derived factor-1), can induce signaling outcomes responsible for a vast array of important biological functions. In T cells, CXCR4 associates with the TCR to modulate T cell activation and migration through the integration of TCR- and GPCR-associated signaling pathways. To better understand how this receptor complex acts, Kremer et al. (p. 1440) examined its utilization of different phospholipase C (PLC) isoforms. A general requirement for PLC signaling was identified for CXCR4-TCR–mediated ERK activation, CXCR4 internalization, and T cell migration. Although PLC-γ1 induces calcium mobilization following TCR stimulation, it was not required for calcium mobilization, ERK or Ras activation, CXCR4 internalization, or CXCR4–TCR complex formation in response to treatment with SDF-1. Instead, PLC-γ1 was important for SDF-1–induced T cell migration through a mechanism independent of TCR ligation and the adaptor protein LAT. In contrast, PLC-β3 was not needed for T cell migration but was necessary for SDF-1–induced calcium mobilization and activation of ERK, N-Ras, and K-Ras. The authors propose that these two PLC isoforms mediate separable, nonredundant functions and could be targeted therapeutically to modulate T cell migration without affecting other signaling.

The broadly expressed coinhibitory molecule programmed death-1 ligand 1 (PD-L1) was initially characterized as a ligand for the inhibitory receptor programmed death-1 (PD-1). However, PD-L1 can also bind to B7-1 to inhibit T cell activity in vitro. Two articles in this issue seek to distinguish the in vivo functions of the PD-L1:PD-1 and PD-L1:B7-1 interactions using selective blocking Abs. In the first article, Paterson et al. (p. 1097) used a NOD mouse model to compare the effects of blocking both PD-L1:B7-1 and PD-L1:PD-1 interactions with blockade of the PD-L1:B7-1 interaction alone during diabetes development. Both the “dual-blocking” (9G2) and “single-blocking” (2H11) Abs strongly precipitated diabetes when given to 13-week-old mice, but 2H11 was less effective than 9G2 in precipitating disease in younger mice. These data suggested that the PD-L1:B7-1 interaction inhibited autoreactive effector T cells and was especially active during the later phase of diabetes development. This idea was further supported in an adoptive transfer model of diabetes, in which 2H11 accelerated disease onset in NOD SCID recipients of diabetic, but not pre-diabetic, NOD T cells, whereas 9G2 promoted disease in both situations. Moreover, both Abs strongly induced diabetes in adoptive transfer disease models mediated by islet-specific effector CD4+ or CD8+ T cells. Thus, the PD-L1:B7-1 interaction protects the host against self-reactive CD4+ and CD8+ T cell responses to limit diabetes development in vivo.

In the second article, Yang et al. (p. 1113) addressed the role of the PD-L1:B7-1 interaction in alloimmune responses. Previous studies demonstrated that PD-L1 blockade accelerated graft rejection, but it was not known which of PD-L1’s receptors were involved. In a model of cardiac transplantation, PD-1–deficient, but not B7-1–deficient, recipients demonstrated accelerated allograft rejection in response to PD-L1 blockade, suggesting involvement of the PD-L1:B7-1 interaction in graft survival. Indeed, specific blockade of this interaction using the 2H11 Ab promoted allograft vasculopathy and rejection. 2H11 treatment also increased alloreactive Th1 and Th2 cells while decreasing regulatory T cells in the spleens of allograft recipients, compared with untreated recipients. Because PD-L1 and B7-1 are expressed on both T cells and APCs, an in vitro culture system was used to tease apart the mechanism of the inhibitory PD-L1:B7-1 interaction. Binding of PD-L1 on APCs to B7-1 on T cells was found to outweigh other possible combinations in providing the inhibitory signal implied by the effects of 2H11 treatment. Taken together, these two studies reveal an important role for an interaction between PD-L1 and B7-1 in preventing autoimmunity and transplant rejection in vivo.

The IgG receptor FcγRIIB inhibits activating FcRs on myeloid cells and controls B cell Ab production. Conflicting data have questioned the involvement of FcγRIIB in systemic lupus erythematosus (SLE). In mice, the autoimmune phenotype of FcγRIIB−/− mice may be influenced by the genomic region flanking the fcγr2b gene, which differs between the 129 and B6 backgrounds and has been designated Sle16. To distinguish the role of fcγr2b in SLE from that of its flanking region, Boross et al. (p. 1304) generated FcγRIIB−/− mice on a pure B6 background (FcγRIIBB6−/−) and compared them with mice generated using 129 embryonic stem cells and then backcrossed onto the B6 background (FcγRIIB129−/−). Both mouse strains demonstrated a hyperactive phenotype and had increased susceptibility to the induction of autoimmunity. However, the FcγRIIB129−/− mice, but not the FcγRIIBB6−/− mice, spontaneously developed antinuclear Abs, suggesting involvement of the Sle16 region in SLE. Mice deficient in FcγRIII that also bore the 129-derived Sle16 region spontaneously developed systemic autoimmunity, further supporting the role of Sle16 in SLE susceptibility. Crossing FcγRIIBB6−/− mice with mice bearing the SLE-prone Yaa locus resulted in a synergistic increase in lupus development. These data suggest that fcγr2b is a disease-modifying gene that amplifies autoimmunity induced by dominant SLE susceptibility genes such as Yaa and Sle16.

Study of the immune response to infections with dimorphic fungi has been limited by a lack of knowledge regarding the fungal Ags responsible for protective T cell immunity. Wüthrich et al. (p. 1421) generated TCR transgenic mice, designated Bd 1807, with CD4+ T cells specific for an immunodominant Ag shared between Blastomyces dermatitidis and Histoplasma capsulatum as a potential tool to study immunity to fungal infections. In support of the utility of this model, T cells from these 1807 mice were activated and recruited to the lung in response to infection with four different systemic dimorphic fungi. Adoptive transfer of 1807 T cells into wild-type mice, followed by fungal vaccination, demonstrated that these T cells were able to develop memory, migrate to the lung during a recall response, and differentiate into Th1 effectors able to confer protection against multiple fungal infections. The adoptive transfer system was also used to demonstrate the long-term persistence of fungal Ag presentation and the requirements for vaccine-induced protective CD4+ T cell-mediated antifungal responses. This transgenic system will facilitate analysis of CD4+ T cell immunity to dimorphic fungi that may lead toward new treatments for these common infections.

Innate and adaptive immune responses are traditionally considered to differ based on speed of initiation, effector cell precursor frequency, receptor specificity, and the existence of memory. Although NK cells have long been assumed to belong to the innate immune system, recent studies have suggested that these cells establish a form of long-term memory. By comparing the kinetics of NK and T cell responses to viral infection, Schlub et al. (p. 1385) may compel a reconsideration of the relative roles of these cells in immunity. Surprisingly, NK cell expansion following viral infection did not occur earlier than T cell expansion, and the growth rate of responding NK cells was similar to, or even slower than, that of T cells. In examining the contraction phase of the antiviral response, three models of population decay were assessed. A purely exponential decay pattern suggested an absence of memory, whereas a pattern of either exponential decay followed by a plateau or biphasic decay suggested the establishment of a population of long-term memory cells. T cells, as expected, followed one of the latter two patterns, depending on cell subset and organ location. In contrast, NK cells decayed exponentially, albeit more slowly than T cells. Whereas the rate of T cell decay slowed over time, NK cell decay remained constant. These data suggest that the apparently long-lived NK cells detected following an infection may represent cells that are slowly decaying, rather than a true “memory” population.

Cross-presentation of Ags from dead cells and the consequent priming of a CTL response require dendritic cell (DC)-mediated engulfment of the dead cells. Granzymes A and B (GrAB) are known to have an important role in CTL-mediated cytotoxicity, but it is not known how their involvement in cell death may affect the priming of an immune response. Hoves et al. (p. 1166) analyzed the effects of GrAB deficiency on the induction of CTL responses to a model tumor Ag. Compared with wild-type animals, mice lacking GrAB demonstrated delayed CTL-mediated tumor cell killing with no significant difference in the level of CTL activation. Cross-presentation of Ags from cells killed by GrAB-deficient CTLs was markedly impaired, leading to reduced priming of Ag-specific CD8+ T cell responses both in vitro and in vivo. Impaired phagocytosis of dead tumor cells by CD8α+ DCs was responsible for the reduction in cross-presentation observed in the absence of GrAB. The authors propose that, in addition to their involvement in perforin-mediated cell death, GrAB are important in the cross-priming that is required for the generation of antitumor effector T cell responses.

Summaries written by Jennifer Hartt Meyers, Ph.D.