Production and in vivo biologic actions of recombinant mouse interferon alpha 2

We have expressed a recombinant mouse interferon alpha (r.Mu-IFN alpha 2) in Escherichia coli under the control of a tryptophan promoter using a synthetic adaptor formed by annealing two partially complementary oligonucleotides which introduced an ATG start codon and re-established the complete coding sequence of the mature IFN alpha 2 protein in the expression vector. Levels of up to 10(7) reference units of Mu-IFN alpha 2 per liter of culture were obtained using this construction. This recombinant mouse interferon alpha 2 exhibited antiviral activity in mice infected with EMC virus and antitumor activity in mice inoculated with Friend leukemia cells in a manner similar to that of natural mouse interferon alpha/beta.     

Characterization of the murine alpha interferon gene family

Mouse and human genomes carry more than a dozen genes coding for closely related alpha interferon (IFN-alpha) subtypes. IFN-alpha, as well as IFN-beta, IFN-kappa, IFN-epsilon, and limitin, are thought to bind the same receptor, raising the question of whether different IFN subtypes possess specific functions. As some confusion existed in the identity and characteristics of mouse IFN-alpha subtypes, the availability of data from the mouse genome sequence prompted us to characterize the murine IFN-alpha family. A total of 14 IFN-alpha genes were detected in the mouse genome, in addition to three IFN-alpha pseudogenes. Four IFN-alpha genes (IFN-alpha1, IFN-alpha7/10, IFN-alpha8/6, and IFN-alpha11) exhibited surprising allelic divergence between 129/Sv and C57BL/6 mice. All IFN-alpha subtypes were found to be stable at pH 2 and to exhibit antiviral activity. Interestingly, some IFN subtypes (IFN-alpha4, IFN-alpha11, IFN-alpha12, IFN-beta, and limitin) showed higher biological activity levels than others, whereas IFN-alpha7/10 exhibited lower activity. Most murine IFN-alpha turned out to be N-glycosylated. However, no correlation was found between N-glycosylation and activity. The various IFN-alpha subtypes displayed a good correlation between their antiviral and antiproliferative potencies, suggesting that IFN-alpha subtypes did not diverge primarily to acquire specific biological activities but probably evolved to acquire specific expression patterns. In L929 cells, IFN genes activated in response to poly(I*C) transfection or to viral infection were, however, similar.     

Mechanism of interferon action. Kinetics of induction of the antiviral state and protein phosphorylation in mouse fibroblasts treated with natural and cloned interferons

The induction of phosphorylation of both protein P1 and protein synthesis initiation factor eIF-2 alpha and the inhibition of virus replication were examined in mouse L929 fibroblasts treated with either natural mouse or individual cloned human interferons (IFN). Natural mouse IFN synthesized in Newcastle disease virus-induced L929 cells and two cloned human leukocyte IFN subspecies synthesized in Escherichia coli, IFN-alpha D and IFN-alpha A/D, possessed antiviral activity in L929 cells as measured by single cycle virus yield reduction with both vesicular stomatitis virus and reovirus. Natural L929 IFN and cloned IFNs, alpha D and alpha A/D, also induced the protein kinase that catalyzed the phosphorylation of endogenous ribosome-associated protein P1 and the alpha subunit of purified initiation factor eIF-2. Two other cloned human IFNs, alpha A and alpha D/A, were poor inducers of both the antiviral state and the phosphorylation of P1 and eIF-2 alpha in mouse L929 cells. The ability of individual human IFN-alpha subspecies to induce P1 and eIF-2 alpha phosphorylation in mouse L929 cells correlated with their ability to induce an antiviral state. Furthermore, the detailed kinetics of induction, in mouse L929 cells, of P1 and eIF-2 alpha phosphorylation and of the antiviral state by the heterologous cloned human IFN-alpha A/D were equivalent to the kinetics of induction by the homologous natural mouse L929 IFN. These results suggest that different subspecies of biologically active IFN induce equivalent antiviral activities and biochemical changes in mouse L929 cells, and that protein phosphorylation may play a major role in the antiviral mechanism of IFN action in mouse L929 fibroblasts.     

Comparison of the lipoprotein gene among the enterobacteriaceae. DNA sequence of Morganella morganii lipoprotein gene and its expression in Escherichia coli

A DNA sequence of 532 base pairs encompassing the entire Morganella morganii lipoprotein gene (lpp) was determined. Sequence comparisons of the M. morganii lpp gene with the lpp genes from Escherichia coli, Serratia marcescens, and Erwinia amylovora reveal that the M. morganii lpp gene is more distantly related to the E. coli lpp gene than any of the other lpp genes examined. Between the E. coli and M. morganii lpp genes, the following homologies were found: 44% in the promoter region (bases, -45 to -1), 88% in the 5'-end untranslated region of the mRNA, 58% in the signal sequence coding region, 75% in the coding region for the first 51 and 43% for the last 7 amino acid residues. Upstream of the promoter region and downstream of the termination codon, there are extensive insertions, deletions, and base substitutions. In spite of the differences in the DNA sequences, the lipoprotein structure was found to be highly conserved except for the carboxyl-terminal sequence of 7 amino residues. The coding region of the M. morganii lpp gene including the signal sequence was inserted into an expression cloning vector so that the production of the M. morganii lipoprotein could be induced in E. coli by a lac inducer, isopropyl-beta-D-thioglactoside. It was found that when induced, the M. morganii prolipoprotein was apparently secreted normally across the E. coli cytoplasmic membrane, modified with glycerol and palmitic acid, processed to the mature lipoprotein, and assembled in the E. coli outer membrane. The bound form covalently linked to the peptidoglycan was also found.     

Organization, structure and expression of murine interferon alpha genes.

Using a human interferon-alpha probe we have isolated recombinant phages containing murine interferon-alpha (Mu IFN-alpha) genes from a genomic library. One of these phages contained two complete Mu IFN-alpha genes and part of a third gene. The insert of a second phage held two IFN genes. This indicates that the Mu IFN-alpha genes are clustered in the genome as is the case for the analogous human genes. The nucleotide sequences of these 5 genes were determined. They show that the genes are all different, albeit highly homologous. The deduced amino acid sequences show that four of the five genes contain a putative glycosylation site. Three genes were transiently expressed in COS cells and they gave rise to protein products showing antiviral properties. The expression of the five Mu IFN-alpha genes and the Mu IFN-beta gene was studied in virus-induced mouse L cells. The individual mRNAs were visualized in a nuclease S1 experiment, using a specific probe for each gene. In RNA preparations from induced cells mRNAs for each of the five alpha genes and the beta gene were present. However, substantial differences in the amounts of the individual mRNAs were observed.     

Pseudogene IFN-alpha L: removal of the stop codon in the signal sequence permits expression of active human interferon

Biologically active interferon (10(6)-10(7) units/liter) was produced in Escherichia coli from modified human alpha interferon (IFN-alpha) pseudogene L. IFN-alpha pseudogene L has a stop codon in the signal peptide coding region. The region that contains the stop codon was replaced with the corresponding region of another human IFN-alpha gene, WA, that does not have a stop codon and was previously engineered for expression by fusion to the M13mp11 lac promoter. The interferon L fusion product was induced with IPTG after infecting E. coli JM103 with the M13 bacteriophage that contained the modified human IFN-alpha pseudogene L. Hence, the IFN-alpha L mature interferon coding sequence, which is not identical to any other alpha-interferon gene, has been conserved for active interferon coding information.     

Mechanism of interferon action. Kinetics of decay of the antiviral state and protein phosphorylation in mouse fibroblasts treated with natural and cloned interferons

The kinetics of decay of the antiviral state and protein phosphorylation induced with natural mouse interferon (IFN) and with cloned human IFN were examined in monolayer cultures of mouse Ll929 fibroblast cells. The antiviral state measured by single cycle virus yield reduction with either vesicular stomatitis virus or reovirus decayed significantly within 2 to 3 days following removal of IFN and by 5 to 8 days virus yields had returned to the level of untreated control cells. Trypsinization of IFN-treated cells did not detectably alter the rate of decay of the antiviral state; however, the decay occurred slightly more rapidly in actively growing as compared to stationary cell cultures. The decay of the IFN-induced protein kinase which catalyzes the phosphorylation of endogenous protein P1 and purified initiation factor eIF-2 alpha correlated with the decay of the antiviral state. The decay rates of the antiviral state and protein kinase observed in mouse L929 cells that had been treated with natural mouse IFN synthesized in Newcastle disease virus-induced L929 cells were comparable to the decay rates observed in L929 cells that had been treated with recombinant human IFN-alpha A/D synthesized in Escherichia coli. The induction and decay of the antiviral state and protein kinase following treatment with a single dose of IFN did not significantly affect the sensitivity of the cell population to a subsequent treatment with a single dose of IFN. However, continuous treatment of L929 cells with natural mouse IFN or recombinant human IFN prevented the decay of both the antiviral state and protein kinase but also ultimately lead to cell death. The results suggest that protein phosphorylation may play an important role in the mechanism of IFN action in mouse L929 fibroblasts.     

Specific residues within an amino-terminal domain of 35 residues of interferon alpha are responsible for recognition of the human interferon alpha cell receptor and for triggering biological effects

Bovine interferon alpha C (IFN-alpha C) manifest at least 10(5)-fold lower antiviral activity on human cells than on bovine cells (Velan, B., Cohen, S., Grosfeld, H., Leitner, M., and Shafferman, A. (1985) J. Biol. Chem. 260, 5498-5504). By oligonucleotide site-directed mutagenesis within the coding region for the NH2-terminal 44-residue domain of BoIFN-alpha C, we replaced up to 18 residues by the corresponding HuIFN-alpha J1 residues. (HuIFN-alpha J1 is less than 60% homologous in sequence to BoIFN-alpha C.) The nine different bovine-human-IFN alpha hybrids obtained were compared to BoIFN-alpha C and HuIFN-alpha J1 with respect to their potential to induce an antiviral state, synthesis of 2-5A-synthetase, and their specific binding to human and bovine cells. Relative to BoIFN-alpha C, a gradual increase in biological activities (antiviral or 2-5A-synthetase) of approximately 10-, 10(2)-, 10(3)-, and approximately 10(4)-fold is obtained, depending on the number and positions of the residues substituted. A direct correlation exists between biological response and ability of IFN alpha to bind specifically to human cells. A BoIFN alpha molecule mutated in the 10-44 NH2-terminal domain was obtained which is 15, 8, and 35% as active as HuIFN-alpha J1 on human cells in specific binding, induction of antiviral, and 2-5A-synthetase activities, respectively. We concluded that at least 5 of the 12 residues at positions 10; 21, 22, 24; 27; 31, 34, 35, 37, 40; 42, 43 in the 10-44 NH2-terminal domain are critical for recognition of the human IFN-alpha cell receptor and for biological activity. These residues are found among 10 strictly conserved residues in all reported mammalian IFN alpha S, and they act in a cooperative manner to induce a biological response in human cells. The gap between the extent of improvement in binding capacity of the BoIFN alpha mutants on human cells and the corresponding biological response suggests that the primary signal of binding to the cell receptor is amplified within the cell. On bovine cells, HuIFN-alpha J1 and BoIFN-alpha C also compete for the same receptor, and it seems that at least part of the 10-44 NH2-terminal domain on IFN alpha is also involved in interaction with the bovine IFN alpha cell receptor.     

Identification of three interferon-inducible cellular enzymes that inhibit the replication of hepatitis C virus

Hepatitis C virus (HCV) infection is a common cause of chronic hepatitis and is currently treated with alpha interferon (IFN-alpha)-based therapies. However, the underlying mechanism of IFN-alpha therapy remains to be elucidated. To identify the cellular proteins that mediate the antiviral effects of IFN-alpha, we created a HEK293-based cell culture system to inducibly express individual interferon-stimulated genes (ISGs) and determined their antiviral effects against HCV. By screening 29 ISGs that are induced in Huh7 cells by IFN-alpha and/or up-regulated in HCV-infected livers, we discovered that viperin, ISG20, and double-stranded RNA-dependent protein kinase (PKR) noncytolytically inhibited the replication of HCV replicons. Mechanistically, inhibition of HCV replication by ISG20 and PKR depends on their 3'-5' exonuclease and protein kinase activities, respectively. Moreover, our work, for the first time, provides strong evidence suggesting that viperin is a putative radical S-adenosyl-l-methionine (SAM) enzyme. In addition to demonstrating that the antiviral activity of viperin depends on its radical SAM domain, which contains conserved motifs to coordinate [4Fe-4S] cluster and cofactor SAM and is essential for its enzymatic activity, mutagenesis studies also revealed that viperin requires an aromatic amino acid residue at its C terminus for proper antiviral function. Furthermore, although the N-terminal 70 amino acid residues of viperin are not absolutely required, deletion of this region significantly compromises its antiviral activity against HCV. Our findings suggest that viperin represents a novel antiviral pathway that works together with other antiviral proteins, such as ISG20 and PKR, to mediate the IFN response against HCV infection.     

Structure and expression of a new murine interferon-alpha gene: MuIFN-alpha I9

A new murine alpha interferon gene, MuIFN-alpha I9, isolated from a BALB/c genomic clone, was characterized. It encodes a mature polypeptide of 167 amino acids (aa), presenting from 77 to 86% homology with the seven other MuIFN-alpha I aa sequences previously described. When compared to the latter, pre-IFN-alpha I9 has 13 distinctive aa, and, remarkably, ten of these occur in pairs. The coding region, fused to the SV40 early promoter and introduced into COS monkey cells, directed the transient secretion of an acid-stable functional IFN of 18-21 kDa. The production in this system reached levels of 300 000 units per 0.15 ml. A comparison of the aa sequence of different murine, rat, bovine, and human alpha and beta IFNs revealed certain common features allowing us to propose a putative secondary structure of the IFN proteins. A detailed analysis of results previously published by us and by others showed that the MuIFN-alpha I9 gene is, together with a least twelve other MuIFN-alpha I genes, located on chromosome 4.     

A large component of the antiviral activity of mouse interferon-gamma may be due to its induction of interferon-alpha

The finding that interferon-gamma (IFN-gamma) may require two rounds of protein synthesis to induce the antiviral state raises the possibility that this IFN may not be directly antiviral. We, therefore, examined the possibility that IFN-gamma induces one or both of the other IFNs (alpha and/or beta) which in turn induce the antiviral state. Evidence is presented showing that under certain conditions a large portion of IFN-gamma's antiviral activity in mouse L-929 cells is mediated by its induction of IFN-alpha based on the findings that: 1) the antiviral activity of IFN-gamma in cells at low densities can be blocked by poly and monoclonal antibody to IFN-alpha and, 2) IFN-alpha can be demonstrated in the supernatant fluids of IFN-gamma treated cells. This report raises the possibility that a major antiviral mechanism of IFN-gamma is via induction of IFN-alpha in the mouse system. If the majority of the antiviral activity of IFN-gamma is via induction of other IFNs, then the role and mechanism of IFN-gamma might have to be reevaluated.     

Inhibitory activity of constitutive nitric oxide on the expression of alpha/beta interferon genes in murine peritoneal macrophages.

We investigated the role of the constitutive nitric oxide (NO) in the expression of interferon (IFN) genes in mouse peritoneal macrophages (PM). The treatment of PM with L-arginine-N(G)-amine (AA), a potent inhibitor of NO-producing enzymes, resulted in a marked accumulation of IFN-alpha4 mRNA and, to a minor extent, of IFN-beta mRNA. In contrast, the expression of IFN-gamma mRNA, as well as tumor necrosis factor alpha and interleukin-6 mRNA, was not affected. Furthermore, a remarkable increase in the expression of the IFN regulating factor 1 (IRF-1), but not of IRF-2, mRNA was detected in AA-treated PM. To investigate whether the AA-induced activation of the IFN system correlates with the production and antiviral activity of IFN, the extent of encephalomyocarditis virus (EMCV) replication was monitored in AA-treated PM with respect to control cultures. AA treatment strongly inhibited, in a dose-dependent manner, EMCV yields in PM. Likewise, similar results were obtained by the addition of the NO-scavenger carboxyphenyl-tetramethylimidazoline-oxyl-oxide. In addition, inhibition of NO synthesis by N(G)-mono-methyl-L-arginine in PM strongly decreased virus replication in coculture of PM and EMCV-infected L929 cells, whereas no antiviral effect was observed in L929 cells alone. Moreover, the AA-mediated antiviral activity was abrogated in the presence of antibody to IFN-alpha/beta, whereas antibody to IFN-gamma was completely ineffective. Taken together, these results indicate that low levels of NO, constitutively released by resting PM, negatively regulate the expression and activity of IFN-alpha/beta in PM. We suggest that NO acts as a homeostatic agent in the regulation of IFN pathway expression in macrophages.     

Correlation between the lipopolysaccharide response of mice and the capacity of mouse peritoneal cells to transfer an antiviral state. Role of endogenous interferon

Freshly harvested mouse peritoneal cells, from normal lipopolysaccharide (LPS)-responsive (Lpsn) mice, were capable of transferring an antiviral state (to vesicular stomatitis virus) to "in vitro aged" mouse macrophages permissive for viral replication. The transfer of the antiviral state was completely abrogated by addition of antibody to interferon (IFN)-alpha/beta in the co-culture medium. In contrast, even large numbers of donor peritoneal cells from LPS-hyporesponsive (Lpsd) C3H/HeJ and C57BL/10ScCR mice did not transfer an antiviral state to target cells. Although peritoneal macrophages from Lpsd mice did not transfer an antiviral state to target cells, they were nevertheless found to be in an antiviral state when first placed in culture. Injection of mice with antibody to mouse IFN-alpha/beta rendered peritoneal macrophages from both Lpsn and Lpsd mice permissive for vesicular stomatitis virus. The decay of this initial antiviral state in peritoneal macrophages during in vitro culture was far more rapid for Lpsd mice than for normal mice. Addition of antibody to mouse IFN-alpha/beta markedly enhanced the in vitro decay of the antiviral state of peritoneal macrophages. Treatment of total peritoneal cells from Lpsn mice with LPS resulted in IFN production, whereas IFN was not detected in the cellfree medium of LPS-treated peritoneal cells from Lpsd C3H/HeJ and C57BL/10ScCR mice. Genetic studies with F1 hybrids between Lpsn and Lpsd mice and with Lpsn and Lpsd recombinant inbred strains revealed a striking correlation between the capacity of peritoneal cells to transfer an antiviral state and their capacity to produce IFN after stimulation with LPS, suggesting that closely linked, if not identical, genes are in some way involved in the transfer of antiviral state as well as in the LPS response by peritoneal cells of normal mice.     

Interferon activity is not detected in blastocyst secretions and does not induce decidualization in mice.

Mouse trophoblast cells synthesized and secreted proteins during the peri-implantation period, some in the molecular size range of alpha interferons (IFN-alpha), known mediators of the maternal recognition of pregnancy in sheep and cows. However, conditioned media samples containing secreted proteins from Day-5 mouse blastocysts or from trophoblast outgrowths did not contain detectable levels of antiviral activity indicative of IFN. In addition, it was not possible to induce a decidual reaction in suitably sensitized uteri with intraluminal instillation of IFN-alpha. The results indicate that IFNs are probably not involved in the maternal recognition of pregnancy at the time of implantation in the mouse.     

Nucleotide sequence of the chicken cardiac alpha actin gene: absence of strong homologies in the promoter and 3'-untranslated regions with the skeletal alpha actin sequence

The entire nucleotide sequence of the chicken cardiac alpha-actin (CC alpha A) gene has been determined. This is the first complete sequence of a cardiac actin gene that includes the promoter region, cap site, all the introns, and the polyadenylation site. The gene contains six introns, five of which interrupt the coding region at amino acids (aa) 41, 150, 204, 267, and 327. The first intron is in the 5'-noncoding region and is 438 bp in length. The CC alpha A gene encodes an mRNA of approx. 1400 bp with 5'- and 3'-untranslated region of 59 and 184 nucleotides (nt), respectively. Like the chicken skeletal alpha-actin gene, the CC alpha A gene has the codon for the aa cysteine between the initiator ATG and the codon for the N-terminal aspartic acid residue of the mature protein. There are no strong homologies (less than 13 consecutive nt) in the promoter or 3'-untranslated regions between the CC alpha A and chicken skeletal alpha-actin genes even though both are expressed in skeletal muscle during development. However, the 3'-untranslated region of the CC alpha A gene demonstrates significant sequence homology (76% over a 200-nt region) with the same region in the partial sequence of the human cardiac gene. The conservation of these sequence homologies between identical isoforms rather than the different alpha actin genes suggests these conserved regions may have a role in regulation rather than tissue-specific expression, as previously proposed.     

Short duration of elevated vIRF-1 expression during lytic replication of human herpesvirus 8 limits its ability to block antiviral responses induced by alpha interferon in BCBL-1 cells

Human herpesvirus 8 (HHV-8) encodes multiple proteins that disrupt the host antiviral response, including viral interferon (IFN) regulatory factor 1 (vIRF-1). The product of the vIRF-1 gene blocks responses to IFN when overexpressed by transfection, but the functional consequence of vIRF-1 that is expressed during infection with HHV-8 is not known. These studies demonstrate that BCBL-1 cells that were latently infected with HHV-8 expressed low levels of vIRF-1 that were associated with PML bodies, whereas much higher levels of vIRF-1 were transiently expressed during the lytic phase of HHV-8 replication. The low levels of vIRF-1 that were associated with PML bodies were insufficient to block alpha IFN (IFN-alpha)-induced alterations in gene expression, whereas cells that expressed high levels of vIRF-1 were resistant to some changes induced by IFN-alpha, including the expression of the double-stranded-RNA-activated protein kinase. High levels of vIRF-1 were expressed for only a short period during the lytic cascade, so many cells with HHV-8 in the lytic phase responded to IFN-alpha with increased expression of antiviral genes and enhanced apoptosis. Furthermore, the production of infectious virus was severely compromised when IFN-alpha was present early during the lytic cascade. These studies indicate that the transient expression of high levels of vIRF-1 is inadequate to subvert many of the antiviral effects of IFN-alpha so that IFN-alpha can effectively induce apoptosis and block production of infectious virus when present early in the lytic cascade of HHV-8.     

IFN-lambda (IFN-lambda) is expressed in a tissue-dependent fashion and primarily acts on epithelial cells in vivo

Interferons (IFN) exert antiviral, immunomodulatory and cytostatic activities. IFN-alpha/beta (type I IFN) and IFN-lambda (type III IFN) bind distinct receptors, but regulate similar sets of genes and exhibit strikingly similar biological activities. We analyzed to what extent the IFN-alpha/beta and IFN-lambda systems overlap in vivo in terms of expression and response. We observed a certain degree of tissue specificity in the production of IFN-lambda. In the brain, IFN-alpha/beta was readily produced after infection with various RNA viruses, whereas expression of IFN-lambda was low in this organ. In the liver, virus infection induced the expression of both IFN-alpha/beta and IFN-lambda genes. Plasmid electrotransfer-mediated in vivo expression of individual IFN genes allowed the tissue and cell specificities of the responses to systemic IFN-alpha/beta and IFN-lambda to be compared. The response to IFN-lambda correlated with expression of the alpha subunit of the IFN-lambda receptor (IL-28R alpha). The IFN-lambda response was prominent in the stomach, intestine and lungs, but very low in the central nervous system and spleen. At the cellular level, the response to IFN-lambda in kidney and brain was restricted to epithelial cells. In contrast, the response to IFN-alpha/beta was observed in various cell types in these organs, and was most prominent in endothelial cells. Thus, the IFN-lambda system probably evolved to specifically protect epithelia. IFN-lambda might contribute to the prevention of viral invasion through skin and mucosal surfaces.