In This Issue

Spontaneous germinal center (GC) reactions can occur in mice with lupus, often resulting in the production of autoantibodies. In a lupus-prone congenic mouse strain, a population of T follicular (TF) cells was recently found that is phenotypically similar to a population of senescence-associated T (SA-T) cells previously identified in aged mice. Tahir et al. (p. 5725) have now analyzed these SA-T cells to assess whether they play a role in spontaneous GC development in lupus. In aged mice, PD-1⁺CD44^hiCD4⁺ SA-T cells developed in a B cell–dependent manner and were found in spontaneously developing GCs. These cells showed little or no proliferation and a CD153⁺ subset spontaneously produced high levels of osteopontin (OPN) in vivo. In lupus-prone female (f)-BWF1 mice, TF cells with a similar SA-T phenotype developed as early as six months of age and, in both these and the aged mice, the development of these cells was associated with extensive CD4⁺ T cell proliferation. Relative to the CD153^– subset, senescent CD153⁺ TF cells from f-BWF1 mice showed much more stable survival in vitro. These CD153⁺ TF cells were triggered by interaction with B cells to produce large amounts of OPN in the GCs in response to a TCR-mediated signal, and this OPN could inhibit B cell apoptosis. Adoptive transfer experiments indicated that the CD153⁺ TF cell subset enhanced the expansion of GCs in f-BWF1 mice, and blockade of OPN suppressed the production of anti-nuclear Abs and the development of lupus nephritis. This subset of senescent TF cells may thus contribute to lupus pathogenesis and to the increase in autoimmunity observed during aging.

Nanotubes, which allow for the transfer of molecules and even organelles or bacteria, are a newly appreciated mechanism of cell–cell communication. It has been shown that direct contact of CD4⁺ T cells with airway smooth muscle (ASM) cells diminishes activated T cell apoptosis, which could potentially affect inflammatory processes such as asthma. Al Heialy et al. (p. 5626) hypothesized that prosurvival signals could be delivered to CD4⁺ T cells through nanotubes. Human CD4⁺ T cells isolated from peripheral blood were stimulated with PMA/ionomycin and were then cocultured with ASM cells. Nanotubes connecting the cocultured cells were confirmed by atomic force microscopy. When CD4⁺ T cells were stained with green dye and ASM cells were stained with red dye before coculture, green and red dyes were observed in both cell types, indicating that transfer of material had occurred in both directions. Administration of anti-CD44 Ab decreased the presence of nanotubes and increased the percentage of apoptotic CD4⁺ T cells. Immunofluorescence microscopy revealed that the anti-apoptotic proteins Mcl-1 and Bcl-2 were present in both types of cells as well as the nanotubes. Following transfection of ASM cells with Mcl-1-GFP, Mc1-1 was observed in nanotubes and CD4⁺ T cells, and transfection of ASM cells with Mcl-1 siRNA resulted in an increase in T cell apoptosis, confirming that CD4⁺ T cells were receiving anti-apoptotic signals (including Mcl-1) from ASM cells via nanotubes.

Celiac disease development is associated with the HLA class II molecules HLA-DQ2 and HLA-DQ8, and an immunodominant gliadin peptide bound to HLA-DQ8 (DQ8-glia-α1) has been found to stimulate T cell responses with a biased TCR repertoire. In this issue, Petersen et al. (p. 6112) used DQ8-glia-α1 tetramers to identify gliadin-specific T cells in small intestinal biopsies from celiac patients to better understand this TCR repertoire in a physiological setting. These gliadin-specific TCRs were biased toward inclusion of TRBV9 (often paired with TRAV26-1/2) and, to a lesser degree, TRBV6-1. A non-germline-encoded Arg residue important for binding to DQ8-glia-α1 was found in the CDR3α region of gliadin-specific TRBV9⁺ TCRs and in the CDR3β region of gliadin-specific TRBV9^– TCRs. The TRBV9⁺ and TRBV9^– TCRs appeared to contact common residues of the DQ8-glia-α1 peptide and both groups of TCRs responded best to deamidated peptide. Crystal structures of TRAV26-2/TRBV9 TCRs complexed with DQ8-glia-α1 revealed remarkably similar structures despite differences in CDR3 lengths and amino acid sequences. The CDR3 loops (both α and β) were key to peptide interactions, and the orientation of the non–germline-encoded Arg residue was conserved in all three complexes analyzed. Comparison of the structure of a TRAV8-3/TRBV6-1 TCR:DQ8-glia-α1 complex with these TRBV9-containing complexes revealed similarities in docking orientation, several germline-encoded features, and, strikingly, in the positioning of the non–germline-encoded Arg residue in the CDR3α region. Thus, gluten-specific T cells in celiac disease patients have a biased TCR repertoire which, despite divergent TCR sequences, form structurally similar complexes with HLA:peptide.

In the context of vaccination, stimulation of CD8⁺ T cell responses is largely dependent on cross-presentation of Ag by subsets of dendritic cells (DCs) including XCR1⁺ DCs. Terhorst et al. (p. 5895) used a portable laser to deliver vaccibodies—chimeric proteins composed of an Ag and a targeting unit—to dermal DCs. Short pulses delivered to the mouse ear created micropores in the stratum corneum and epidermis and induced mild inflammation. The vaccibody contained OVA and XCL1, the only ligand for XCR1. XCL1 vaccibodies specifically targeted XCR1⁺ DCs and did not deliver Ag to Langerhans cells, CD11b⁺ DCs, or macrophages. OVA-specific OT-I CD8⁺ T cells and OT-II CD4⁺ T cells were labeled and transferred into mice 1 d prior to immunization with XCL1-OVA vaccibodies or OVA. Cells from draining lymph nodes (LNs) were analyzed 3 d after immunization, and XCL1-OVA vaccibodies were 3 times more effective at inducing OT-II CD4⁺ T cell proliferation and 15 times more effective at inducing OT-I CD8⁺ T cell proliferation than OVA alone. Immunization of wild-type mice with XCL1 and OVA separately or Xcr1^−/− mice with XCL1-OVA was no more effective than immunization with OVA alone, demonstrating that the beneficial effect was due to specific targeting of Ag to XCR1⁺ DCs. Enhancement of T cell proliferation with XCL1-OVA required DC migration from the skin to draining LNs but was independent of TLR signaling. In a B16-OVA melanoma model, prophylactic administration of XCL1-OVA vaccibodies controlled tumor volume and induced IFN-γ production by CD4⁺ and CD8⁺ T cells in the tumors. When administered therapeutically 3 d after mice were inoculated with B16-OVA, XCL1-OVA vaccibodies induced the control of tumor growth better than did OVA. These results demonstrate the efficacy of specific targeting of Ag to DCs capable of cross-presentation, and may lead to the development of more effective vaccines.

Surfactant protein-A (SP-A) has been shown to be beneficial in Mycoplasma pneumoniae infection in mice and human SP-A alleles are associated with susceptibility to a variety of respiratory diseases. Binding of SP-A by pathogens increases pathogen uptake by phagocytic cells, so the protective mechanism of SP-A in M. pneumoniae infection may be a result of reduced bacterial burden. In this issue, Ledford et al. (p. 6123) used mycoplasma membranes (MMF) to further dissect the protective role of SP-A in mice. Compared with wild-type (WT) mice, SP-A^−/− mice challenged oropharyngeally with MMF exhibited enhanced mucin production and neutrophil recruitment in the lungs, demonstrating that the protective phenotype was not a result of reduced pathogen load. In vitro experiments demonstrated that SP-A^−/− tracheal epithelial cells secreted more KC (CXCL1) than WT cells, which is likely responsible for the enhanced neutrophil recruitment. Phosphorylated epidermal growth factor receptor, which could be responsible for activating pathways involved in KC and mucin production, was observed in lung sections from MMF-challenged SP-A^−/− mice. This idea was supported by the finding that treatment with epidermal growth factor receptor inhibitor reduced the amount of mucin produced in MMF-challenged SP-A^−/− mice. To examine the relevance of these findings in human disease, the function of two isoforms of recombinant human SP-A2 were tested in humanized knock-in mice expressing one of the two alleles on a murine SP-A^−/− background, SP-A2 223K/– and SP-A2 223Q/– mice. MMF challenge of SP-A2 223K/– mice resulted in WT levels of mucin production, whereas SP-A2 223Q/– challenged mice had a similar phenotype as SP-A^−/− mice, producing more mucin and KC and increased neutrophil recruitment in the lungs. These phenotypes correlated with binding to MMF; whereas SP-A2 223K bound MMF with high affinity, SP-A2 223Q was not able to bind MMF. These results show how SP-A genetic variance could lead to enhanced susceptibility to respiratory disorders in humans.

Macrophages can differentiate into proinflammatory (M1) or anti-inflammatory (M2) subsets, depending on the stimuli to which they are exposed. IL-4 triggers M2 macrophage polarization through activation of the cooperatively acting transcription factors STAT6 and Krüppel-like factor 4 (KLF4), but the details of the ensuing polarization remain unclear. The MCP-1–induced protein (MCPIP) has been shown to inhibit NF-κB activation and inflammatory cytokine production and to induce biological processes linked to M2 polarization, leading Kapoor et al. (p. 6011) to hypothesize that MCPIP promotes M2 polarization and inhibits M1 polarization. Following IL-4 stimulation of murine macrophages, KLF4 induced MCPIP expression, which was important both for KLF4-mediated M1 suppression and for stimulation of M2 polarization. MCPIP also mediated the reactive oxygen species (ROS) production, endoplasmic reticulum stress, and autophagy involved in M2 polarization. The authors next analyzed the relative roles of MCPIP’s two types of catalytic function, deubiquitinase and RNase activity, using MCPIP mutants deficient in only one of these activities. Both mechanisms were required for M2 polarization and ROS production, whereas RNase activity was more important than deubiquitinase activity for induction of both autophagy and the transcription factors PPARγ and C/EBPβ. The RNase activity of MCPIP also promoted expression of M2-associated microRNAs and suppressed expression of M1-associated microRNAs. In vivo analysis of mice either specifically lacking or overexpressing MCPIP in myeloid cells confirmed this protein’s key role in promoting M2 macrophage polarization while inhibiting M1 polarization. These data introduce MCPIP as a key player in macrophage differentiation that could eventually be manipulated as a therapeutic target.