Regulation of cytokine gene expression by reactive oxygen and reactive nitrogen intermediates.

Reactive oxygen intermediates (ROI), reactive nitrogen intermediates (RNI), and cytokines are frequent companions at sites of acute inflammation. Previous work has established a clear link between the production of cytokines and the subsequent generation of ROI and RNI. However, more recent data indicates that ROI and RNI not only serve as end-stage effector molecules of pathogen destruction and tissue injury, but also as initiators of acute inflammation. Specifically, ROI and RNI will upregulate cytokine gene expression since antioxidants inhibit interleukin 8 (IL-8) production and do not decrease production of other cytokines. Treatment with hydroxyl radical scavengers such as dimethyl sulfoxide (DMSO) will decrease the production of IL-8 in stimulated human whole blood, fibroblasts, type II epithelial cells, and hepatoma cells, but not other cytokines. Addition of exogenous ROI will increase IL-8 production in these same cells. Inhibition of nitric oxide synthase will decrease production of IL-8, whereas addition of nitric oxide (NO)-generating compounds will increase production of IL-8. The hydroxyl radical appears to be the final common pathway of cell activation for IL-8 synthesis, since DMSO will inhibit the NO-driven production of IL-8. Our data indicate that ROI and RNI can serve as intracellular second messengers to induce IL-8 gene expression.

Nitric oxide regulation of interleukin-8 gene expression.

While the regulation of nitric oxide (NO) by inflammatory cytokines or lipopolysaccharide (LPS) has received considerable attention, NO modulation of cytokine expression has yet to be fully explored. The NO synthase inhibitor, NG-nitro-L-arginine methyl ester (L-NAME), inhibited interleukin (IL)-8 and IL-6 production in LPS-stimulated human whole blood in a dose-dependent manner. In the presence of 1 microgram/mL LPS, L-NAME blocked IL-8 release (72 +/- 4% inhibition at 20 mM (mean +/- SEM, p < .05)) 24 h post-LPS without affecting cellular viability. IL-6 production was significantly inhibited only with the highest dose of L-NAME used. L-NAME inhibition of IL-8 production was also observed at the mRNA level. Conversely, direct exposure of whole blood to NO with the spontaneous NO liberator DETA NONOate caused a dose-dependent stimulation of IL-8, but had no effect on IL-6 release. IL-8 concentrations rose from 8.3 +/- 1.9 ng/mL at 24 h to 31.7 +/- 7.6 ng/mL at 72 h with a single stimulation of 10 mM DETA NONOate. The hydroxyl radical scavenger dimethyl sulfoxide (DMSO) prevented the DETA NONOate induction of IL-8, suggesting the participation of the hydroxyl radical in the NO-induced IL-8 production. These results provide important evidence substantiating a role for NO as a regulator of cytokine expression.

Functional characterization of monoclonal antibodies to interferon-tau.

Interferon tau (IFNtau) produces an array of biological effects, including antiluteolytic, antiviral, antiproliferative and immunomodulatory activities, without the consequent cytotoxicity associated with other type I IFNs. Four anti-IFNtau monoclonal antibodies (MAbs) have been characterized by determining regional epitopes and observation of their effects on IFNtau binding, antiviral and antiproliferative activity. Using an enzyme-linked immunoadsorbent assay (ELISA) developed against six overlapping synthetic peptides representing the entire linear sequence of IFNtau, three antibodies, HL-98, HL-100 and HL-127, were found to react with the carboxy terminal peptide, while HL-129 bound the penultimate amino terminal peptide. Binding studies indicated that MAbs directed against either region could effectively inhibit the binding of alkaline phosphatase labeled IFNtau to cells expressing type I IFN receptors. While only two of the MAbs significantly reversed IFNtau-induced growth inhibition, the antiviral activity of IFNtau was significantly inhibited by MAbs that bound the amino and carboxy termini, confirming the functional importance of these domains in the binding and subsequent activity of IFNtau.

Tissue-mediated selection of viral variants: correlation between glycoprotein mutation and growth in neuronal cells.

Viral variants with different biological properties predominate in the central nervous system (CNS) and lymphoid tissues of carrier mice infected at birth with the Armstrong strain of lymphocytic choriomeningitis virus. The CNS isolates have the same phenotype as the parental strain and cause acute infections in adult mice, while the spleen-derived isolates cause chronic infections associated with suppressed T-cell responses and susceptibility to opportunistic infections. Our previous studies have identified a single amino acid change in the viral glycoprotein, a phenylalanine-to-leucine (F-->L) mutation at residue 260, that correlates with the tissue-specific selection and the persistent and immunosuppressive phenotype of the spleen isolates (R. Ahmed, C.S. Hahn, T. Somasundaram, L. Villarete, M. Matloubian, and J. H. Strauss, J. Virol. 65:4242-4247, 1991). In this study, we screened viral isolates obtained from the spleen, liver, kidney, and brain of carrier mice for the presence of this mutation and determined the temporal selection of variants as they appear in these organs. We found that this F-->L amino acid change is common to > 90% of the spleen and liver isolates and is selected for rapidly by day 32 postinfection (p.i.). Although the kinetics observed in the kidney are relatively slower than in the spleen and liver, this F-->L mutation predominates in the kidney-derived isolates by 250 days p.i. In contrast, the majority of the CNS isolates retain the parental sequence up to 250 days p.i. In addition, most of the brain isolates replicated efficiently in a neuronal cell line, and this enhanced growth phenotype in neurons correlated with the parental F genotype. This linkage with neurotropism, along with our earlier finding that the F-->L mutation is necessary for enhanced infection of macrophages (M. Matloubian, S. R. Kolhekar, T. Somasundaram, and R. Ahmed, J. Virol. 67:7340-7349, 1993), provides a cellular basis for the molecular changes associated with tissue-specific selection. Taken together, these results suggest that tropism for macrophages is a critical determinant in selection of variants with the F-->L mutation in tissues such as spleen and liver, and tropism for neurons is important in retention of the F genotype in the CNS.

Nitric oxide regulation of IL-8 expression in human endothelial cells.

The regulatory signals required to induce the production of IL-8, an important neutrophil chemoattractant and activator, have yet to be clearly defined. We examined the role of nitric oxide (NO) in IL-8 regulation. The NO synthase inhibitor, (L)-NG-nitroarginine methyl ester (L-NAME), inhibited the TNF-stimulated IL-8 production in the human endothelial cell line, ECV304, in a dose-dependent manner without affecting cellular viability (TNF alone, 5.5 +/- 0.9 ng/ml; TNF + 5 mM L-NAME, 2.4 +/- 0.5 ng/ml). Moreover, exogenously added NO produced by the spontaneous NO generating compounds, S-Nitroso-N-acetyl-D,L-pennicillamine (SNAP) and Ethanamine, 2,2'-(hydroxynitrosohydrazono)bis- (DETA NONOate), induced a dose-dependent release of IL-8 from these cells. Maximal stimulation of IL-8 was found to be 1.2 +/- 0.1 ng/ml with the 1 mM concentration of SNAP and 1.6 +/- 0.1 ng/ml with the 2 mM concentration of DETA NONOate. These results provide key evidence substantiating a regulatory role of NO in IL-8 expression.

Transcriptional and post-transcriptional regulation of interleukin-8.

Steady-state mRNA for interleukin (IL) 8 persists significantly longer than mRNA for tumor necrosis factor (TNF)-alpha in lipopolysaccharide (LPS) stimulated human whole blood. Nuclear run-ons demonstrated consistent levels of transcriptional activity of the IL-8 gene at 2 and 26 hours after LPS stimulation when compared with the rapid induction and arrest of the TNF-alpha gene. Inhibition of cellular transcription with actinomycin D added at 2 hours after LPS resulted in the substantial decrease of both IL-8 and TNF-alpha mRNAs, demonstrating half-lives of 4.6 and 2.3 hours, respectively. In contrast, inhibition of transcription at 23 hours after LPS revealed extremely stable IL-8 mRNA with a half-life of > 10 hours. The half-life of beta-actin in the same actinomycin-D-treated samples did not vary significantly from the half-life calculated at the 2-hour time point (5.5 hours versus 5.6 hours), indicating that the observed IL-8 mRNA stability was not an artifact of the system. Both IL-8 and TNF-alpha protein levels decreased when actinomycin D was added 2 hours after LPS stimulation. However, no effect in IL-8 protein levels was evident when actinomycin D was added at 23 hours after LPS. These results demonstrate that IL-8 mRNA stability is controlled at both the transcriptional and posttranscriptional levels of gene regulation.

High-yield expression and purification of human interferon alpha-1 in Pichia pastoris.

For several years, interferon alpha-1, also known as interferon alpha-D, has been studied for treatment of various viral diseases, such as hepatic fibrosis caused by hepatitis B, herpes simplex virus keratitis, and bovine respiratory diseases in calves. Currently, recombinant human interferon alpha-D (rHuIFNalphaD) is expressed intracellularly in Escherichia coli or secreted by Bacillus subtilis and Saccharomyces cerevisiae. In this report, we describe the process of obtaining a relatively high-yield secretion of biologically active recombinant rHuIFNalphaD using the Pichia pastoris system. The process produced as high as 0.7 mg of purified protein per 20 ml of shake culture of rHuIFNalphaD with better bioactivity than the commercially available rHuIFNalphaD molecule produced in E. coli.

IFN-tau inhibits IgE production in a murine model of allergy and in an IgE-producing human myeloma cell line.

BACKGROUND: IFN-tau, a type I IFN, is an antiviral, immunomodulating, and antiproliferative agent similar to IFN-alpha and IFN-beta, but IFN-tau lacks the toxicity associated with high concentrations of these IFNs in tissue culture and in animal studies. We have previously shown that IFN-tau inhibits antibody production in a murine model of an autoimmune disease. OBJECTIVE: We investigate the effectiveness of ovine IFN-tau and other type I IFNs in suppressing the development of allergic sensitization in a murine model of allergy by using ovalbumin (OVA) antigen as an allergen and in suppressing IgE production by using a human IgE-producing myeloma cell line. METHODS AND RESULTS: Mice that were treated with IFN-tau in vivo before and after intraperitoneal immunization with aluminum hydroxide-precipitated OVA had significantly lower OVA-specific IgE levels than the PBS-treated group. IFN-tau-treated mice had reduced inflammatory cell infiltration into the lung tissue. Furthermore, in vitro IFN-tau treatment of splenocytes taken from OVA-immunized mice suppressed OVA-induced proliferation. Also, treatment of the IgE-producing human myeloma cell line U266BL with IFN-tau-reduced IgE production and inhibited cell proliferation compared with media controls. Similar suppression of proliferation and inhibition of IgE production was seen with other type I IFNs, as well as a humanized IFN-tau/IFN-alphaD chimeric that consists of residues 1 to 27 of the ovine IFN-tau and residues 28 to 166 of the human IFN-alphaD. The chimeric was not toxic to human peripheral white blood cells at concentrations as high as 10(5) U/mL, whereas human IFN-alphaD was toxic at 10(3) U/mL. CONCLUSION: These data suggest that IFNs may be useful in preventing allergic sensitization by suppressing the production of allergen-specific IgE antibodies without toxic side effects.

Impaired responsiveness to gamma interferon of macrophages infected with lymphocytic choriomeningitis virus clone 13: susceptibility to histoplasmosis.

Lymphocytic choriomeningitis virus clone 13 (LCMV clone 13), a variant isolated from the spleens of neonatally infected mice, causes persistent infections in mice infected as adults. Such persistently infected mice succumb to a normally sublethal dose of Histoplasma capsulatum, and their macrophages contain overwhelming numbers of yeast cells of the fungus. Both LCMV clone 13 and H. capsulatum yeast cells target and replicate in macrophages of the host. We sought to study the effects of LCMV clone 13 on the ability of macrophages to control growth of H. capsulatum in vitro. We show that the growth of H. capsulatum within macrophages was not directly affected by the presence of LCMV clone 13. However, macrophages containing LCMV clone 13 did not respond fully to gamma interferon (IFN-gamma) stimulation. Such unresponsiveness resulted in proliferation of the fungus within macrophages cultured in the presence of IFN-gamma. The addition of anti-IFN-alpha/beta antibodies to LCMV clone 13-infected macrophage cultures restored macrophage responsiveness to IFN-gamma. These results indicate that production of IFN-alpha/beta by LCMV clone 13-infected macrophages antagonizes their responsiveness to IFN-gamma. Such antagonism may be one of the mechanisms by means of which certain viruses cause immune suppression and susceptibility to opportunistic infections.

Molecular basis of organ-specific selection of viral variants during chronic infection.

Viral variants of different phenotypes are present in the central nervous system (CNS) and lymphoid tissues of carrier mice infected at birth with the Armstrong strain of lymphocytic choriomeningitis virus. The CNS isolates are similar to the parental virus and cause acute infections in adult mice, whereas the lymphoid isolates cause chronic infections associated with suppressed T-cell responses. In this study, we provide a molecular basis for this organ-specific selection and identify a single amino acid change in the viral glycoprotein that correlates with the tissue specific selection and the persistent and immunosuppressive phenotype of the variants. This phenylalanine (F)-to-leucine (L) change at position 260 of the viral glycoprotein was seen in the vast majority (43 of 47) of the lymphoid isolates, and variants with L at this residue were selected in spleens of persistently infected mice. In striking contrast, isolates with the parental sequence (F at residue 260) predominated (48 of 59 isolates) in the CNS of the same carrier mice. Complete nucleotide sequence analysis of the major structural genes of several independently derived (from different mice) spleen isolates showed that these variants were greater than 99.8% identical to the parental virus. In fact, the only common change among these spleen isolates was the F----L mutation at residue 260 of the glycoprotein. These results show that an RNA virus can exhibit minimal genetic drift during chronic infection in its natural host, and yet a single or few mutations can result in the organ-specific selection of variants that are markedly different from the parental virus.