Studies of T cells that recognize peptides from the model Ag cytochrome c have contributed extensively to our understanding of TCR function. However, efforts to define the structural basis of TCR specificity in this system have not been successful. In this issue, Newell et al. (p. 5823) have solved and compared the crystal structures of moth cytochrome c (MCC)-reactive TCRs (2B4 and 226) bound to MCC/I-Ek. The 226 TCR, a natural variant of the 5c.c7 TCR, has a high affinity for the MCC peptide and was found to cross-react with MCC peptides bearing a wide variety of amino acid substitutions. Despite large differences between the CDR3 regions of 2B4 and 226, both TCRs bound to MCC/I-Ek in a very similar manner and demonstrated a nearly identical docking mode. Interestingly, the germline-encoded regions of the Vα-chains of these TCRs dominated the Vβ-chains in imposing the binding mode. Detailed analysis of these structures provided hypotheses to explain many of the conserved features that previous studies have identified in MCC/I-Ek–reactive TCRs. Structures of the 2B4 and 5c.c7 TCRs in the absence of ligand revealed that both TCRs have extremely flexible CDR3 loops. In providing explanations for a great deal of previous biochemical and cellular data, this study reveals much about the structural basis of TCR specificity and provides a framework for future insights into T cell biology.

The liver is considered a tolerogenic organ, yet little is known about the factors responsible for the differentiation of the induced regulatory T cells (iTregs) important to its suppressive nature. Hepatic stellate cells (HSCs) store the vitamin A metabolite retinoic acid (RA), which has been shown to promote iTreg differentiation and inhibit the induction of proinflammatory Th17 cells. To address the potential contributions of HSCs to the liver’s tolerogenic state, Ichikawa et al. (p. 5549) analyzed the involvement of these cells in iTreg induction. Surprisingly, highly purified HSCs did not serve as professional APCs for CD4+ T cells and could only present Ag to NKT cells with low efficiency. Instead, in the presence of dendritic cells, HSCs promoted the TGF-β–induced differentiation of iTregs and suppressed TGF-β–mediated Th17 differentiation. Use of an RA receptor antagonist or vitamin A-deficient HSCs demonstrated that RA production was required for the HSC-mediated enhancement of iTreg differentiation. Thus, HSCs act as regulatory bystanders that indirectly support the immune suppressive state of the liver by promoting the differentiation of iTregs while inhibiting Th17 cell development.

Although they are considered part of the innate immune system, NK cells can develop memory-like responses. Antitumor dendritic cell (DC)-based immunotherapy can lead to the long-term maintenance of NK cells in an antitumor state and consequent tumor regression. To determine the mechanism by which this antitumor activity is sustained, Shimizu et al. (p. 5927) analyzed the requirements for NK cell activation in response to tumor challenge following DC immunization. Tumor-induced reactivation of NK cells 1 mo after DC immunization required CD4+ T cell interactions with endogenous DCs. CXCL10 produced by splenic DCs recruited NK cells to the T cell-rich splenic white pulp, where these T cell–DC interactions occur. The long-term reactivation of NK cells required IL-2 produced by effector memory CD4+ T cells (CD4+ TEM), which additionally expressed CD40L and induced DC maturation. This DC maturation and subsequent production of IL-12 were also necessary for NK cell activation. Thus, long-term antitumor “memory” responses by NK cells can be established by DC immunization, but only with the involvement of endogenous DCs and CD4+ TEM cells. These data support the utility of DC-based immunotherapy in the generation of thorough immune protection against tumors.

Chronic inflammation or infections such as HIV can lead to regenerative bone marrow (BM) failure. Infection with the opportunistic fungal pathogen Pneumocystis, which causes severe pneumonia in AIDS patients, induces BM failure in RAG-deficient mice that also lack the type I IFNR (IFrag−/− mice). Taylor et al. (p. 5956) sought to identify the mechanism responsible for BM failure in these mice and found that its prevention required IFN signaling in BM-derived cells, but not stromal cells. During the course of Pneumocystis lung infection, apoptosis of neutrophils and myeloid precursors was observed in the BM of IFrag−/− mice. This apoptosis was associated with oxidative stress and caspase activation representing both the extrinsic and the intrinsic apoptotic pathways, although the extrinsic pathway dominated in neutrophils. The expression of antiapoptotic genes, as well as the growth factors GM-CSF and G-CSF, was reduced in the BM of Pneumocystis-infected IFrag−/− mice compared with RAG−/− mice, whereas TNF-α expression was increased. Inhibition of TNF-α could enhance myeloid precursor survival in IFrag−/− mice, indicating the functional relevance of the expression data. Signaling via type I IFN thus maintains the balance between pro- and antiapoptotic signals in the BM during Pneumocystis infection, and this knowledge may be applied to the prevention of regenerative BM failure in patients.

Inflammatory mediators such as TNF play a central role in the pathogenesis of arthritis and other autoimmune diseases. In a search to identify regulators of TNF activity, Gottar-Guillier et al. (p. 6014) identified the nonreceptor tyrosine kinase BMX (bone marrow kinase on chromosome X) as an important player in multiple pathways of inflammatory signaling. IL-8 secretion in response to TNF, IL-1β, or TLR agonists was reduced by BMX depletion. BMX was found to be required for TNF-mediated activation of JNK, p38 MAPK, and NF-κB signaling and acted either at or downstream of the TAK1/TAB complex, which is critical for the activation of these pathways. Activation of IL-8 transcription in vitro was abolished in the presence of BMX mutants lacking the kinase or pleckstrin homology domains, indicating that kinase activity and membrane localization were required for BMX activity. In vivo relevance of BMX-mediated modulation of inflammatory signaling was assessed in a mouse model of inflammatory arthritis. BMX-deficient mice were protected from arthritis development, but this protection was eliminated by a catalytically inactive BMX mutant. BMX is thus an essential regulator of inflammation through both kinase-dependent and -independent processes.

Memory CD8+ T cells can protect against malaria-causing Plasmodium infections. To determine how different Plasmodium species and host-specific factors affect the threshold of memory CD8+ T cells necessary for antimalarial protection, Schmidt et al. (p. 5873) analyzed the infection of different mouse strains with P. berghei and P. yoelii. In BALB/c mice, protection against P. yoelii challenge required more circumsporozoite (CS) Ag-specific memory CD8+ T cells than did protection against P. berghei challenge. Although the cause is unclear, it is typically more difficult to successfully immunize C57BL/6 and B10.D2 mice than BALB/c mice against P. berghei challenge. CS-specific memory CD8+ T cells in these three strains did not differ significantly in phenotype or function; instead, host factors outside the MHC influenced the threshold of memory T cells required for protective immunity. Generation of reciprocal bone marrow chimeras showed that these factors were CD8+ T cell extrinsic. However, obstacles to protective immunization in B10.D2 and C57BL/6 mice could be overcome by multiple booster immunizations that induced extremely high numbers of CS-specific memory CD8+ T cells. The increased threshold of CD8+ T cells required for protection against P. yoelii infection in BALB/c mice or P. berghei infection in B10.D2 mice correlated with higher rates of Plasmodium replication in the liver. The variability this study reveals in the contributions of different host genotypes and Plasmodium strains to protection against malaria reflects the complex challenges facing the development of malaria vaccines for the genetically variable human population.

Regulatory B cells (Bregs) produce IL-10 and can suppress autoimmune responses. To assess the suppressive abilities of endogenous Bregs in vivo, Carter et al. (p. 5569) studied bone marrow chimeras lacking IL-10–producing B cells. Compared with controls, these mice developed exacerbated Ag-induced arthritis. Enhanced disease was accompanied by a reduction in the numbers and FoxP3 expression levels of regulatory T cells (Tregs) in the draining lymph nodes and increases in proinflammatory Th1 and Th17 responses. The transitional 2 marginal zone Breg subset of IL-10–producing B cells was responsible for these effects, as adoptive transfer of these cells—but not other B cell subsets—into IL-10−/− mice reduced arthritis and normalized the numbers of Tregs and Th1/Th17 cells. A mechanistic basis for this Breg activity was suggested by the observation that IL-10–producing B cells formed longer-lasting conjugates with CD4+ T cells than did IL-10−/− B cells. This longer contact time was associated with augmented Treg differentiation and a reduction in the differentiation of Th17 cells. IL-10–producing Bregs can thus promote tolerance by inducing Treg development and shifting the balance between regulatory and inflammatory T cells.

Tcell-dependent (TD) immune responses lead to the generation of memory B cells able to rapidly mount a robust, often IgG-mediated immune response upon re-exposure to the TD Ag. The importance of augmented signaling through the IgG BCR in this recall response has not been determined, leading Haniuda et al. (p. 5620) to analyze whether type II T cell-independent (TI-II) Ags that directly crosslink the BCR can elicit recall responses to epitope-matched TD Ags in vivo. Mice were primed with a TD Ag and then challenged with TI Ags bearing the same epitopes. A type I T cell-independent Ag that activated both the BCR and TLR4 could induce an IgG1 memory response, but, surprisingly, a TI-II Ag induced no detectable response. Tertiary challenge with the TD Ag led to a reduced IgG1 response in mice that had received the TI-II Ag, compared with controls, suggesting that high-affinity Ag-specific memory B cells were tolerized by the TI-II Ag. This tolerance was induced in a B cell-intrinsic manner, and mechanistic analysis revealed that high-affinity Ag-specific IgG+ memory B cells were not anergized but were instead specifically eliminated, at least partially via apoptosis, in response to the TI-II Ag. This study identifies a checkpoint in memory B cell differentiation that may help prevent the development of autoimmunity as a by-product of a productive immune response.

Summaries written by Jennifer Hartt Meyers, Ph.D.