IN THIS ISSUE

The Mountz group found that one C57BL/6J (B6) × DBA/2J recombinant inbred strain (BXD2) developed autoimmune arthritis and/or renal disease and had high levels of autoantibodies with multiple reactivities. In a continuation of their studies, Hsu et al. (p. 5357 ) showed that arthritis severity was greater in BXD2 mice injected with hybridomas producing multireactive vs single-reactive autoantibodies; the multireactive autoantibodies had more mutations and were of IgG isotype. BXD2 mice had higher levels of mRNA for activation-induced cytidine deaminase (AID) in their follicular splenic B cells and had more proliferating germinal center CD86⁺ B cells compared with B6 controls. Stimulated CD4⁺ T cells from BXD2, but not B6, mice induced high levels of AID expression in BXD2 or B6 B cells and had a high proliferative response that was abrogated by addition of CTLA4-Ig to the culture medium. AID expression in B cells and production of IgG autoantibodies in BXD2 mice were reduced to B6 levels by injection of 4-wk-old mice with a virus vector expressing CTLA4-Ig; the effect lasted up to 28 wk of age after a single injection. The authors conclude that CD28-stimulated CD4⁺ T cells induce increased AID expression in germinal center B cells that results in multireactive autoantibodies.

The number and phenotypes of macrophage subsets are influenced by colony-stimulating factors (CSFs) that induce their differentiation. However, the molecular properties of the subsets and the interplay of CSFs inducing them are not known fully. Fleetwood et al. (p. 5245 ) generated two macrophage populations, GM-BMM and BMM, by exposure of mouse bone marrow precursors to GM-CSF and M-CSF, respectively, for 7 days in vitro. GM-BMM had greater stimulation of T cell proliferation in an MLR and produced IL-12p70 and IL-23 and more TNF-α and IL-6 after LPS stimulation than BMM, whereas LPS-stimulated BMM produced more IL-10 and CCL2. IκBα degradation was rapid, and RelA nuclear levels increased quickly in LPS-treated GM-BMMs vs BMMs. EMSAs demonstrated three NF-κB DNA binding complexes and one AP-1 complex in both cell populations after exposure to LPS but with different kinetics for formation of the complexes within each population. Exposure of BMM to GM-CSF plus M-CSF for 16 h resulted in a changed phenotype whereby LPS stimulation resulted in increased production of TNF-α, production of IL-12p70 and IL-23, and greater nuclear localization of RelA. Conversely, exposure of GM-BMM with M-CSF plus GM-CSF led to diminished LPS-induced production of TNF-α and IL-23 and an increase in CCL2. Kinetics of formation of nuclear complexes and their compositions were altered in both populations treated with both CSFs. The experiments show that differential CSF treatment of two macrophage subsets alters their cytokine profiles following LPS stimulation via changes in transcription factor binding. These results may be useful in controlling localized inflammatory reactions.

Patients with Fanconi anemia (FA) have progressive bone marrow failure, have high serum levels of TNF-α, and often progress to acute myeloid leukemia. Sejas et al. (p. 5277 ) developed a mouse strain lacking FA complementation group C protein (FANCC) that is important in oxidative stress signaling. Fancc^−/− mice had higher mortality and greater cytopenia due to LPS-induced septic shock than wild-type controls. Bone marrow cells from LPS-injected mutant mice formed fewer colonies in vitro, had increased apoptosis, and were unable to reconstitute lethally irradiated wild-type animals. Serum levels of TNF-α, IL-6, and MIP-2 were higher in Fancc^−/− mice than in controls after LPS treatment; expression of other proinflammatory genes was higher in mutant bone marrow cells in the presence of LPS. The inflammatory phenotype was corrected in LPS-treated lethally irradiated mice given mutant bone marrow transduced with a DNA repair-proficient FANCC gene, in LPS-treated mutant mice lacking TNF-α by genetic deletion or by injection of anti-TNF-α Ab, and in LPS-treated mutant mice pretreated with a reactive oxygen species (ROS) scavenger. LPS-induced intracellular levels of ROS in bone marrow cells were higher in mutant vs wild-type mice. Inhibitors of JNK and p38 kinase rescued progenitor cell growth in vitro, but only the JNK inhibitor suppressed LPS-induced ROS production. Retroviral gene transfer of a dominant-negative mutant of the JNK upstream activator into mutant bone marrow cells also suppressed ROS production. The authors demonstrate that the functional defect in their mouse model of FA is due to an inhibition of hemopoiesis by TNF-α-induced ROS.

The pro- and anti-inflammatory activities of macrophages at sites of tissue injury are well known but not well understood. Miyake et al. (p. 5001 ) developed a mouse model of human acute respiratory distress syndrome (ARDS) to study the role of alveolar macrophages in acute lung injury. Their transgenic mice expressed the human diphtheria toxin receptor (DTR) gene under control of the lysozyme M (LysM) gene promoter in macrophages. Alveolar and peritoneal macrophages in LysM-DTR, but not wild-type, mice died after DT injection, and only transgenic animals died after receiving a high dose of DT. Extensive damage in the lungs of injected transgenic mice was seen by macroscopic and histological examinations.Surprisingly, alveolar epithelial type II (AE2) cells that also express the LysM gene were absent or apoptotic in lungs of injected LysM-DTR mice but not in controls. Loss of surfactant proteins in DT-injected transgenic mice was demonstrated by immunoblotting on lung extracts and by immunohistochemistry on lung sections. Lethally irradiated LysM-DTR mice injected with wild-type bone marrow lost AE2 cells but retained alveolar macrophages after DT injection, and, although they all died, lung damage and congestion was less severe than controls reconstituted with LysM-DTR bone marrow. Bronchoalveolar lavage fluid levels of hepatocyte growth factor (produced by bronchial epithelial cells to act on AE2 cells) in DT-treated irradiated transgenic mice receiving LysM-DTR bone marrow were higher than in controls that received wild-type bone marrow. The authors demonstrate that loss of AE2 cells in their mouse model of human ARDS is responsible for respiratory failure and that alveolar macrophages reduce injury.

Although TLRs are important in the innate immune response to pathogens, the role of TLR-MyD88 signaling in protecting a host against viral infections has not been determined. In this issue, two groups used mice lacking MyD88 to explore the role of the adaptor protein. In the first report, Zhou et al. (p. 5173 ) infected mice with cytopathic vesicular stomatitis virus (VSV). Intranasal (i.n.) infection resulted in hind limb paralysis or death in >60% of MyD88^−/− mice compared with only a few instances in wild-type mice; surprisingly, no MyD88^−/− or wild-type mice died after i.v. VSV infection. Serum levels of IFN-α were lower in infected MyD88^−/− mice after i.n. infection but comparable to those of wild-type mice after i.v. infection. MyD88^−/− mice produced fewer neutralizing Abs after i.n. VSV infection and less IgM and more IgG1 Abs after i.v. or i.n. infection than wild-type controls. Only infected wild-type mice had high levels of neutralizing Abs up to 70 days postinfection (p.i.), and their splenic CD4⁺ T cells had higher VSV-specific responses 7 days p.i.; the results were independent of the route of infection. The data indicate that the impact of MyD88 is dependent on the route of infection. Their finding that MyD88 is essential for initial IgM induction and for maintenance of long-term antiviral neutralizing Ab is corroborated by the work of Guay et al. (p. 5124 ). This second group injected mice i.p. with polyoma virus (PyV) and found greatly reduced viral-specific IgM and IgG Ab-secreting cells in spleen and bone marrow and lower serum levels of viral-specific IgG Abs in PyV-infected MyD88^−/− vs wild-type mice at 21 days and 1.5 years p.i.; MyD88^−/− and wild-type mice had similar numbers of germinal center B cells at 7 and 14 days p.i. At 14 days and 1.5 years p.i., infected mice lacking MyD88 had normal levels of only virus-specific IgG1 and IgG3 Abs but greatly reduced levels of IgG2a and IgG2b Abs. Infected wild-type or SCID mice reconstituted with T cell-depleted MyD88^−/− splenocytes generated anti-PyV IgM and IgG or only IgM responses, respectively. Infected B cell-deficient mice that had been reconstituted with MyD88^−/− B cells made small amounts of anti-PyV IgG compared with controls given wild-type T cells, but infected T cell-deficient mice given MyD88^−/− or wild-type T cells had comparable Ab levels. These experiments show that MyD88 is required in B cells for long-term humoral immunity to PyV infection. Both groups shed new light on the involvement of MyD88 in the antiviral adaptive immune response in mice.

The germinal center (GC) reaction requires helper T cell activation of B cells for Ab production and memory B cell formation. However, it is not known what prevents expansion of CD4⁺ T cells within the GC. Marinova et al. (p. 5010 ) found that purified CD4⁺CD57⁺ T cells from human tonsils promoted survival, Ab production, and higher CD40, CD80, and CD86 surface expression of GC B cells in vitro than did CD4⁺CD57⁻ T cells. In contrast, GC CD4⁺CD57⁺ T cells inhibited proliferation and IL-2 production of activated GC CD4⁺CD57⁻ T cells, especially those that were activated under Th1-polarizing conditions. Proliferation of activated responder cells was partially restored by separation from the GC CD4⁺CD57⁺ T cells in Transwells, by treating the cells with anti-CD95L Ab, or by adding TGF-β- or IL-10-neutralizing Ab. The combination of anti-CD95L, anti-IL-10, and anti-TGF-β Abs completely abolished GC CD4⁺CD57⁺ T cell suppression of activated GC CD4⁺CD57⁻ T cell proliferation. Among other cell surface molecules studied, only CTLA-4 was expressed more highly on GC CD4⁺CD57⁺ T cells compared with conventional effector CD4⁺ T cells, and its expression was comparable to that of conventional regulatory T cells. GC CD4⁺CD57⁺ T cells expressed glucocorticoid-induced TNF-like receptor to the same extent as regulatory T cells but lacked Foxp3 expression. The authors describe a phenotypically unique GC CD4⁺CD57⁺ T cell that provides help for B cell responses but suppresses expansion of conventional effector CD4⁺ T cells within the GC.

Several lines of evidence suggest that TLR2 has both immunostimulatory and immunosuppressive roles in regulating humoral and/or cellular responses to pathogenic microbes. To clarify the role of TLR2 in the cellular innate immune response, Watanabe et al. (p. 4917 ) exposed peritoneal macrophages from wild-type or TLR2-deficient mice to FITC-labeled Staphylococcus aureus or Escherichia coli. Phagocytosis of both bacteria was equivalent in both macrophage populations, whereas killing of S. aureus, but not E. coli, by TLR2^−/− macrophages was lower as measured by in vitro colony-forming activities of recovered phagocytosed bacteria. JNK phosphorylation levels were low in TLR2^−/− macrophages incubated with S. aureus but high when incubated with E. coli. The colony-forming ability of S. aureus phagocytosed by wild-type or by LPS-treated TLR2^−/− macrophages was reduced by a JNK inhibitor. S. aureus-phagocytosing TLR2^−/− macrophages released more superoxide than wild-type macrophages, and a superoxide inhibitor increased colony counts from S. aureus-infected wild-type macrophages. The JNK inhibitor further increased superoxide production in S. aureus-phagocytosing wild-type, but not TLR2^−/−, macrophages. The experiments show that engulfed S. aureus survive in mouse macrophages by using TLR2 to activate JNK and suppress superoxide production.

Summaries written by Dorothy L. Buchhagen, Ph.D.