Astrocytes as targets for Venezuelan equine encephalitis virus infection.

Venezuelan equine encephalitis virus (VEE) produces an acute infection in humans and induces a well-characterized cytopathic effect in neurons of the central nervous system (CNS). However, little is known about the role of glial cells in response to VEE infection of the CNS. Our results demonstrate that VEE is capable of a productive infection in primary astrocyte cultures and that this infection is cytotoxic. Further, there were significant differences in the growth kinetics comparing virulent and attenuated strains of VEE. Additionally, VEE infection of astrocyte cultures induced gene expression of two neuro-immune modulators, tumor necrosis factor-alpha (TNF-alpha) and inducible nitric oxide synthase (iNOS). Assays for TNF-alpha protein and nitric oxide (NO) demonstrated high levels of TNF-alpha protein and low levels of NO in response to VEE infection of astrocytes. These observations suggest an important role of astrocytes in this virus-induced encephalitis, and that interactions between astrocytes, other glial cells, and neurons may be important in VEE pathogenesis. Such interactions, which could impact neuronal survival, may include loss of functional changes in astrocytes or, alternatively, their production of neurotoxic molecules.

Autocrine/paracrine IFN-alphabeta mediates the lipopolysaccharide-induced activation of transcription factor Stat1alpha in mouse macrophages: pivotal role of Stat1alpha in induction of the inducible nitric oxide synthase gene.

We have examined the role of Stat1alpha in the induction by LPS of the mouse inducible nitric oxide synthase (EC 1.14.13.39) gene. LPS induced both the tyrosine phosphorylation of Stat1alpha and the production of nitric oxide in a time- and dose-dependent manner. The phosphorylation of Stat1alpha elicited by LPS differed from that observed using IFN-gamma or IFN-beta, in that LPS induced less phosphorylated protein and the time course of induction was much delayed (2-4 h compared with 30 min). Cycloheximide inhibited LPS-mediated Stat1alpha phosphorylation. In addition, cell culture supernatants derived from macrophages treated with LPS for 4 h could be transferred to naive macrophage cultures resulting in rapid (30 min), rather than delayed (4 h), phosphorylation of Stat1alpha. Together, these results implicated an autocrine/paracrine effector protein(s) in the phosphorylation process. LPS stimulated phosphorylation of Stat1alpha in peritoneal macrophages derived from IFN-gamma-knockout mice, negating any possibility that IFN-gamma was the mediator. By contrast, neutralizing Ig raised against mouse IFN-alphabeta inhibited both the delayed LPS-mediated phosphorylation of Stat1alpha and the rapid induction of phosphorylation induced by supernatants from LPS-stimulated cultures. Collectively, these results show that LPS-induced IFN-alphabeta production, Stat1alpha activation, and nitrite accumulation closely parallel one another, suggesting that indirect activation of transcription factor Stat1alpha by IFN-alphabeta is a critical determinant of LPS-mediated inducible nitric oxide synthase gene expression.

Analysis of the antagonist effect of IFN-alpha on IFN-gamma-induced interferon consensus sequence binding protein messenger RNA in murine macrophages.

Interferon consensus sequence binding protein (ICSBP), a member of the interferon regulatory factor family of DNA binding proteins, was recently demonstrated in ICSBP knockout mice to play a critical role in hematopoiesis and virus susceptibility. In macrophages, ICSBP mRNA and protein are strongly induced by IFN-gamma, but only marginally by IFN-alpha/beta. When present concurrently, IFN-alpha/beta antagonizes IFN-gamma-induced ICSBP mRNA and protein synthesis. The unknown mechanism(s) underlying this antagonism may involve competitive interactions between these two IFN species or between molecules of their respective signaling cascades. The data presented demonstrate that IFN-alpha does not interfere with initiation of transmembrane signaling by IFN-gamma and that inhibition of ICSBP mRNA expression by IFN-alpha is independent of new protein synthesis. Nuclear proteins from IFN-gamma-treated or from IFN-alpha plus IFN-gamma-treated cells showed identical binding patterns in EMSAs using a palindromic interferon response element (pIRE) from the ICSBP promoter. These proteins were primarily reactive with antibodies directed against STAT1 alpha and, to a lesser extent, against STAT2 and ISGF3 gamma. However, when a second, upstream IRE-like sequence was evaluated by EMSA, a DNA binding pattern distinct from that seen following exposure to IFN-gamma alone was observed after prolonged stimulation with both IFN-alpha and IFN-gamma. These data suggest a possible novel mechanism for IFN-alpha-induced inhibition of IFN-gamma-induced gene expression.

Differential expression of interferon regulatory factor 1 (IRF-1), IRF-2, and interferon consensus sequence binding protein genes in lipopolysaccharide (LPS)-responsive and LPS-hyporesponsive macrophages.

Macrophages secrete interferon (IFN), as well as other cytokines, following lipopolysaccharide (LPS) stimulation. The interferon regulatory factors (IRFs) comprise a family of DNA-binding proteins that have been implicated in the transcriptional regulation of IFN and certain IFN-inducible genes. We therefore characterized basal and LPS-inducible levels of IRF-1, IRF-2, and interferon consensus sequence binding protein (ICSBP) mRNA in LPS-responsive macrophages and compared the expression of these genes in macrophages that typify two murine models of LPS hyporesponsiveness. In the first model, the LPS-hyporesponsive phenotype of the C3H/HeJ mouse is genetically determined and maps to the Lps locus on mouse chromosome 4. In the second model, normally LPS-responsive macrophages acquire a transient LPS-hyporesponsive phenotype following a prior exposure to LPS, a phenomenon referred to as "endotoxin tolerance." Using reverse transcription PCR, we detected basal levels of IRF-1 mRNA in LPS-responsive (Lpsn) macrophages that were approximately 15 times higher than those found in LPS-hyporesponsive (Lpsd) macrophages. Conversely, Lpsd macrophages expressed basal levels of IRF-2 mRNA that were approximately 18 times higher than those expressed in Lpsn macrophages. LPS stimulation resulted in a dose- and time-dependent accumulation of IRF-1, IRF-2, and ICSBP mRNA only in Lpsn macrophages. Cycloheximide inhibited the accumulation of LPS-stimulated IRF-2 and ICSBP mRNA, but not IRF-1 mRNA, thus designating IRF-1 an immediate-early, LPS-inducible gene. Finally, macrophages rendered tolerant to endotoxin expressed elevated but nonmaximal mRNA levels for all three transcription factors that are not reinduced upon secondary challenge with LPS. Thus, the IRFs may represent yet an additional molecular pathway in the complex response to LPS.

Induction of IFN-gamma in macrophages by lipopolysaccharide.

In this paper we report that macrophages can be stimulated to express detectable levels of IFN-gamma-specific mRNA. Macrophages from lipopolysaccharide (LPS)-responsive, C3H/OuJ mice are induced by LPS to increase steady-state levels of IFN-gamma-specific mRNA, while those from LPS-hyporesponsive C3H/HeJ mice are not. This interstrain variation is apparently the result of LPS-specific signal differences since macrophages derived from both Lpsn and Lpsd mouse strains are able to produce comparable levels of IFN-gamma-specific mRNA following stimulation with polyinosinic-polycytidylic acid. The identity of the cell type responsible for this IFN-gamma message appears to be the macrophage as IFN-gamma-specific mRNA was also detectable following T and natural killer cell depletion, in the LPS-stimulated RAW 264.7 cell line, and in a homogeneous population of mature macrophages propagated in vitro by stimulation of bone marrow progenitors with recombinant colony stimulating factor-1. Immunofluorescent staining of fixed and permeabilized LPS-stimulated macrophages confirmed the presence of immunoreactive IFN-gamma protein. The possible significance of IFN-gamma production by macrophages is discussed in the context of normal macrophage differentiation as well as the inflammatory immune response.

Interferon alpha-induced changes in Fc gamma R-specific mRNA expression and isotype-specific, Fc gamma R-mediated phagocytosis in C3H/OuJ (Lpsn) and C3H/HeJ (Lpsd) macrophages.

Peritoneal exudate macrophages from C3H/OuJ (Lpsn) and C3H/HeJ (Lpsd) mice were treated with interferon alpha (IFN-alpha), and isotype-specific phagocytosis and gene expression were measured. IFN-alpha selectively augments phagocytosis of IgG2a, with a highly significant induction of Fc gamma RI mRNA expression, while parallel Fc gamma RIII alpha measurements show a relatively small increase.

The physical separation of Lps and Ifa loci in BXH recombinant inbred mice.

Several reports in the literature suggest that many of the phenotypic defects of LPS-hyporesponsive C3H/HeJ mice may be attributed to decreased IFN production by their macrophages. The physical proximity on chromosome 4 of the gene which encodes sensitivity to LPS (Lps) and the structural gene cluster which encodes IFN-alpha (Ifa), suggests the possibility that the Lps gene, whose product is unknown, may actually be a part of the Ifa gene cluster. The C57BL/6J and C3H/HeJ mouse strains carry distinct alleles at both the LPs and the Ifa loci. In this study, we have analyzed these parental strains, as well as 12 recombinant inbred strains derived from these parental strains (e.g., BXH strains), for inheritance of these distinct alleles. The results show the segregation of these two loci: in 5 of 12 BXH RI strains, the IFN-alpha restriction fragment length polymorphism characteristic of one parental strain was discordant with the predicted LPS response phenotype. Therefore, we conclude that the Lps and the Ifa genes are physically distinct despite the apparent cause and effect relationship which is observed phenotypically.

Altered expression of the Salmonella typhimurium-specific B-cell repertoire in mice chronically treated with antibodies to immunoglobulin D.

Using a modification of the splenic focus assay, we analyzed the Salmonella typhimurium-specific B-cell repertoire in salmonella-susceptible BALB/c mice. Although these mice normally succumbed to salmonella infection before antibody was produced, they appeared to have splenic S. typhimurium-specific B-cell precursors that could be activated to differentiate and secrete antibody in a manner which was quantitatively and qualitatively identical to that of salmonella-resistant mouse strains. We also analyzed the primary S. typhimurium-specific B-cell repertoire in BALB/c mice that had been chronically treated with antibodies to immunoglobulin D (IgD) and therefore had no surface IgD-positive B cells. Although the frequency of S. typhimurium-specific precursors in these mice was similar to that of control mice, there was an apparent alteration in the isotype distribution pattern in anti-IgD-treated mice. Control mice generated a significantly greater proportion of IgG-secreting clones than did anti-IgD-treated mice. In addition, a greater proportion of S. typhimurium-specific clones from control mice secreted IgG2 than secreted IgG1, and those clones that secreted IgG2 but not IgM, IgG3, or IgG1 were greater than 20-fold more common in control than in anti-IgD-treated mice. Finally, we analyzed the immune response of control and anti-IgD-treated mice to a live avirulent vaccine, S. typhimurium SL3235. Although both groups were protected after challenge with a live virulent S. typhimurium strain, only the control mice made serum antibodies to this vaccine. Taken together, these results show that (i) salmonella-susceptible BALB/c mice have S. typhimurium-specific B cells, (ii) the S. typhimurium-specific B cells in anti-IgD-treated mice may have a restricted capacity to switch heavy-chain classes, (iii) the similarity observed in the frequency of the S. typhimurium-specific precursors for these two groups of BALB/c mice is not reflected in the serum, and (iv) the failure of anti-IgD-treated mice to generate a serum antibody response to SL3235 in the face of complete protection suggests that this model may be used to study cell-mediated immune mechanisms in the apparent absence of humoral immunity.

Comparison of the clonal diversity of the B cell repertoires in adult mice that differ in the expression of cell surface IgD.

To determine whether the expression of surface IgD (sIgD) influences the extent of the expressed B cell repertoire, the clonal diversity of the B cell population in mice treated chronically with anti-IgD (delta) antibodies has been compared with the B cell repertoire observed in control animals, using the splenic focus limiting dilution B cell assay. The results show that the phosphorylcholine (PC)-specific B cell precursor frequency in anti-delta antibody-treated mice is increased when compared with that of control mice. Isotype and idiotype (T15) analyses of PC clonal products from anti-delta antibody-treated and control mice revealed no distributional differences. Analyses of the 2,4-dinitrophenyl (DNP)- and fluoresein isothiocyanate-specific B cell repertoires confirmed that the equal or increased precursor frequencies observed in anti-delta antibody-treated mice are not specific for the PC antigen. The increased precursor frequency of B cells from anti-delta antibody-treated mice was not the result of increased homing of B cells from anti-delta antibody-treated mice to recipient spleens, since B cells from control mice homed twice as well to recipient spleens as did B cells from anti-delta antibody-treated mice. Other studies demonstrated that (a) on average, antibody-secreting clones were generated more slowly when B cells from anti-delta antibody-treated mice were used as a source of precursors than B cells of control mice and (b) both sIg- spleen cells and sIg+ spleen cells from anti-delta antibody-treated mice generated a higher frequency of specific antibody-secreting clones than did the corresponding populations from control mice. These observations suggest that a population of sIgM+sIgD- B cells exists that resembles sIgD+ B cells rather than neonatal or xid B cells in its ability to generate responses to PC and suggests that the sIgM+sIgD- B cells from anti-delta antibody-treated mice are more responsive than are sIgM+IgD+ B cells, regardless of antigenic specificity, to the stimuli provided in the splenic focus system. Finally, this study suggests that the expression of sIgD does not influence the extent of the expressed B cell repertoire.

Early development of the T cell repertoire. In vivo treatment of neonatal mice with anti-Ia antibodies interferes with differentiation of I-restricted T cells but not K/D-restricted T cells.

Monoclonal antibodies to I-Ak were injected into neonatal H-2k mice for a period of 3 wk. The spleens of such mice are devoid of Ia-positive cells. Allo- and trinitrophenyl (TNP)-self-specific cytotoxic T lymphocyte (CTL) responses in such anti-I-A-treated mice were almost completely abrogated at the end of the 2-3 wk in vivo treatment period. Development of suppressor cells, carry-over of blocking antibodies, lack of responder accessory cells, or defective CTL function were not responsible for the observed defect. As concanavalin A supernatant could restore the defect, it is more likely that the defect is due to the absence of competent Ia-specific T helper cells. In addition, anti-I-A-treated mice exhibit reduced I-A antigen expression in the thymus and defective Ia-bearing accessory cell function in the spleen. It is postulated that, for development of Ia-specific T cells to occur, precursor T cells need to interact with Ia-encoded products in the thymus, and anti-Ia treatment interferes with this process. Finally, the mechanism of this interference was shown to be due to actual removal or functional inactivation of those I-A-positive elements responsible for the education of I-A-recognizing T cells, since in (H-2b X H-2k)F1 mice, treatment with anti-I-Ak antibodies results in abrogation of CTL responses to TNP in association with both parental haplotypes, while in the thymus of these mice expression of both I-Ak and I-Ab was reduced.

Neonatal suppression with anti-Ia antibody. I. Suppression of murine B lymphocyte development.

To investigate the role of sIa in B lymphocyte development, mice were injected from birth with hybridoma anti-Ia antibody. There was on the average a 95% reduction in the number of sIgD+ cells, and 85% diminution in the number of sIgM+ cells, a small but significant increase in the number of sIgM+ sIgD- sla- B cells, as well as a dramatic increase in the number of sIgM- cells bearing a B lineage antigen. Exposure of cells to high doses of anti-Ia antibody in vitro or in vivo for 24 hr did not affect the number of sIg+ lymphocytes. Thus, it appears that chronic anti-Ia administration critically alters normal B lymphocyte development.