Identification and genetic mapping of differentially expressed genes in mice differing at the If1 interferon regulatory locus.

A subtractive cDNA library was used to identify differentially expressed genes in mouse strains that differ at If1, a locus that regulates response to interferon induction by Newcastle Disease Virus infection. Among the isolated clones, sequence analysis identified the ribosomal proteins L37a and S8 as well as cDNAs for thymosine beta4, the QM transcriptional factor, and a novel genetic sequence. Analysis of two multilocus mouse crosses showed that the thymosine beta4 gene, Ptmb4, is present as a single-copy gene that maps to distal Chr X. The L37a, S8, and QM clones are all members of large multilocus families. These five clones were used to determine the map locations for 37 loci, of which 31 had not previously been described. The novel genetic sequence, D3Ppr1, mapped to distal Chr 3 near the position of the If1 locus, suggesting it may be a candidate for this regulatory gene.

Regulation of CD34 expression in differentiating M1 cells.

CD34 is a cell surface glycoprotein expressed on hematopoietic stem and progenitor cells, but not on mature blood cells. In the present study we found that CD34 downregulation during hematopoiesis occured at the level of transcriptional initiation. Two transcription initiation sites (TISs) were identified in each of three different CD34+ cell lines; these TISs were located at 120 and 80 bp 5' of the translation start site, respectively. The promoter lacks TATA elements and, like other TATA-less promoters, the TISs conform to the consensus sequence for an INR (PyPyCAPyPyPyPy). An additional 3000 bp of upstream genomic DNA were sequenced and found to contain consensus sites for transcription factors, suggesting their potential role in gene regulation. Transient transfection assays using CD34 promoter-luciferase reporter constructs, containing sequences up to 3 kb upstream and inclusive of the TIS, indicate that this promoter drives transcription in hematopoietic CD34+ cells but not CD34+ nonhematopoietic cells. Both cell type specific expression and full promoter activity are maintained in constructs that contain as little as 454 bp upstream of the TISs. Optimal promoter activity requires the 5' untranslated region of exon 1, which contains a 51-bp element that has the potential to form an extensive secondary structure. In the plasmid DNA, however, this secondary structure was not detectable by P1 nuclease digestion. At least three proteins present in uninduced M1 nuclear extracts bind to this element. Two of the three proteins were identified as Sp 1 and Sp 3 based on supershift experiments. These data suggest that CD34 expression by hematopoietic stem and progenitor cells involves hematopoietic cell-specific factors that interact with regulatory elements within the first 230 bp of the promoter and that optimal expression requires a 60-bp segment of the 5' untranslated region.

Stimulation of interferon and cytokine gene expression by imiquimod and stimulation by Sendai virus utilize similar signal transduction pathways.

The imidazoquinolineamine derivative 1-(2-methyl propyl)-1H-imidazole [4,5-c]quinoline-4-amine (imiquimod) has been shown to induce alpha interferon (IFN-alpha) synthesis both in vivo and in peripheral blood mononuclear cells in vitro. In this study, we show that, in these cells, imiquimod induces expression of several IFNA genes (IFNA1, IFNA2, IFNA5, IFNA6, and IFNA8) as well as the IFNB gene. Imiquimod also induced the expression of interleukin (IL)-6, IL-8, and tumor necrosis factor alpha genes. Expression of all these genes was transient, independent of cellular protein synthesis, and inhibited in the presence of tyrosine kinase and protein kinase C inhibitors. Infection with Sendai virus led to expression of a similar set of cytokine genes and several of the IFNA genes. Imiquimod stimulates binding of several induction-specific nuclear complexes: (i) the NF-kappa B-specific complexes binding to the kappa B enhancer present in the promoters of all cytokine genes, but not in IFNA genes, and (ii) the complex(es) binding to the A4F1 site, 5'-GTAAAGAAAGT-3', conserved in the inducible element of IFNA genes. These results indicate that imiquimod, similar to viral infection, stimulates expression of a large number of cytokine genes, including IFN-alpha/beta, and that the signal transduction pathway induced by both of these stimuli requires tyrosine kinase and protein kinase activity.

Virus-mediated induction of interferon A gene requires cooperation between multiple binding factors in the interferon alpha promoter region.

Transcriptional activation of interferon A (IFNA) gene in virus-infected cells is controlled by a 35-nucleotide inducible element that is cell type specific. Within this region, two elements, alpha F1 and IRF-1 binding sites, were shown by mutation analysis to play a crucial role in the expression of inducible element. In this study, we have analyzed the binding of nuclear proteins to the alpha F1 sequence and have shown that the induction is associated with the formation of a novel complex alpha F1/B, which contains at least two DNA binding proteins of 68 and 96 kDa. In contrast, no binding of the purified interferon regulatory factor 1 (IRF-1) either to the alpha F1 or IRF-1 binding sites could be detected in vitro. However, the oligonucleotides corresponding to alpha F1 or IRF-1 binding sites competed efficiently for the induction of IFNA4 promoter region in a transient transfection assay. We suggest that the induction of IFNA promoter region requires cooperation between alpha F1 binding proteins and IRF-1. Interestingly, our data also show that the inability of IFNA6 promoter to be expressed in infected L-cells may be a result of a viral-induced repressor, which could act by binding and inactivating alpha F1 or by competing for the IRF-1 binding site. These results suggest that cell-specific expression of IFNA genes results from core-cruitment of trans-acting factors that bind to alpha F1 and the IRF-1 binding site with the cell-specific virus-induced activator or repressor.

Lack of role of TGF-beta 1 in decreased lymphoproliferative response in HIV-1 infection.

We investigated whether reduction of the phytohemagglutinin (PHA)-induced proliferative response of lymphocytes from HIV type-1 (HIV-1)-infected (HIV+) individuals could be explained by overproduction of transforming growth factor-beta 1 (TGF-beta 1), a strong inhibitor of T cell proliferation. PHA-stimulated PBMC from 40 HIV- and 42 HIV+ homosexual men from the Baltimore Center of the Multicenter AIDS Cohort Study (MACS) were studied using Northern blot analysis of expression of TGF-beta 1 mRNA and determining the effects of anti-TGF-beta 1 neutralizing antibody on PHA-induced proliferative responses. Compared to the HIV- donors, HIV+ donors did not show increased expression of TGF-beta 1 mRNA in unstimulated or PHA-stimulated PBMC. Furthermore, a neutralizing antibody to TGF-beta 1 did not reverse the decreased proliferative response of PBMC from HIV+ individuals to PHA or interleukin-2. These results indicate that TGF-beta 1 is not involved in T cell proliferation defects seen in HIV+ donors.

65-kDa protein binds to destabilizing sequences in the IFN-beta mRNA coding and 3' UTR.

The transient expression of human interferon (IFN-beta) gene in response to viral induction is regulated both at the transcriptional and posttranscriptional levels. The decrease in levels of IFN-beta mRNA, which requires protein synthesis, is due to transcriptional repression as well as a rapid turnover of beta mRNA. Previous studies have shown the presence of two destabilizing sequences, one in the 3' untranslated region (UTR) and the other in the coding region. We have shown in this study that the coding region destabilizing element resides in the 3' end of the coding region (+538 to +637) and that the degradation does not require the translation of IFN-beta mRNA through its coding region. In addition, we have identified three domains of 19, 20, and 29 nucleotides long that specifically bind a 65-kilodalton (kDa) cytoplasmic protein. One of the binding sites is in the 3' end of the coding region and the other two in the 3' UTR. All these regions are AU-rich and show considerable homology to each other. Interestingly, the levels of the 65-kDa protein was increased after poly rI.rC induction. We suggest that this 65-kDa protein is a component of the IFN-beta mRNA destabilizing complex or plays a role in the degradation of IFN-beta mRNA.

Relationship between IL-2 receptor expression and proliferative responses in lymphocytes from HIV-1 seropositive homosexual men.

Previous studies have shown that exogenous IL-2 does not correct the reduction in phytohaemagglutinin (PHA)-induced proliferation of lymphocytes from HIV-1 infected (HIV+) individuals. We investigated the mechanism of this reduction to determine if reduced expression of the complete IL-2 receptor (IL-2R) was responsible. In a series of experiments, PHA-stimulated lymphocytes from a total of 89 HIV- and 93 HIV+ homosexual men from the Baltimore Multicentre AIDS Cohort Study (MACS) were studied to determine the expression of messages for the alpha and beta subunits of the IL-2R, the binding of 125I-IL-2 to high affinity IL-2R, and the effect of IL-2 on cell proliferation. Compared to HIV- donors, PHA-stimulated lymphocytes from most HIV+ donors demonstrated (i) a reduction in high affinity IL-2R expression that correlated with the reduction in the IL-2-induced proliferative response; and (ii) a reduction in expression of both IL-2R alpha- and beta-chain mRNA which may be responsible for decreased high affinity IL-2R expression. However, lymphocytes from some HIV+ individuals had borderline low IL-2-induced proliferation despite normal or elevated expression of high affinity IL-2R. These results suggest that decreased expression of IL-2R may account, at least in part, for the lower proliferative response of cells from HIV+ donors.

Detection of a neuron-specific 9.0-kb transcript which shares homology with antisense transcripts of HIV-1 gag gene in patients with and without HIV-1 infection.

The pathogenesis of the neurologic abnormalities associated with the acquired immune deficiency syndrome (AIDS) is poorly understood. Although human immunodeficiency virus type 1 (HIV-1) transcripts have been detected in endothelial cells and macrophages of the central nervous system in patients with AIDS, infection of neuronal cells by HIV-1 has not been established. The purpose of this study was to localize HIV-1 transcripts in the central nervous system. 3H and digoxigenin-UTP-labeled riboprobes generated from a 942-bp fragment of DNA from the 5' end of the HIV-1 gag sequence were used for in situ hybridization. The antisense riboprobe hybridized to lymphoid cells in the sections of kidney and spleen obtained from patients with AIDS, as well as to the HIV-1-infected A3.01 cell line. The control sense probe did not hybridize to these same cells. In contrast, no detectable hybridization was observed to neuronal cells when the antisense probe was applied to sections of brain obtained from patients with and without AIDS. To our surprise, however, specific hybridization was observed to neuronal cells when the control sense probe was applied. This hybridization with the control sense probe was seen in both patients with and without HIV-1 infection. Northern blot analysis confirmed the in situ hybridization results; a unique 9.0-kb transcript was detected exclusively in brain tissue. These data suggest that there is a neuron-specific 9.0-kb transcript that shares extensive homology with antisense gag HIV-1 sequences and that this transcript is expressed in neuronal cells of both HIV-1-infected and noninfected individuals. The biological significance of this 9.0-kb transcript is unknown, but it may play an important role in the interactions of HIV-1 with neuronal cells.

Distinct activation of murine interferon-alpha promoter region by IRF-1/ISFG-2 and virus infection.

Virus infection in mouse L929 cells activates expression of interferon-alpha 4 (IFN-alpha 4), but not IFN-alpha 6. The integrity of a symmetrical sequence, GTAAAGAAAGT (alpha F1 site); (-103 to -93), present in the 35 nucleotide (nt) long inducible element (IE) (-109 to -75) of the alpha 4 promoter region is essential for the virus-induced expression. In the present study, we have shown that the interferon regulatory factor 1 (IRF-1) can induce expression of both IFN-alpha 4 and -alpha 6 in a transient expression assay. Virus infection cooperates with IRF-1 and further enhances transcription from the alpha 4 promoter, but inhibits the IRF-1-mediated expression from the alpha 6 promoter. The virus-mediated induction is determined by both IRF-1 and alpha F1 sites, while activation by IRF-1 in a cotransfection assay is not greatly influenced by the alpha F1 sequence. The activation of IFN-alpha gene promoters by IRF-1 was limited to the transient expression assay. The integrated alpha 4 promoter or the endogenous IFN-alpha genes could not be induced by transfection with IRF-1 expressing plasmid and IRF-1 did not up-regulate expression of the endogenous IRF-1 gene. However, expression of IRF-1 alone was sufficient to up-regulate the expression of two IFN stimulated genes, 2',5' oligoadenylate synthetase (OAS) and interferon stimulated (ISG)-15 gene. These results suggest that induction of IFN-alpha gene expression by virus infection requires cooperation between IRF-1 and another factor(s) that binds to the alpha F1 sequence.

Mouse genotype affects inducible expression of cytokine genes.

The levels of circulating IFN in mice infected with Newcastle disease virus (NDV) are regulated by the If-1 locus. In this study we show that in NDV-infected C57BL/6 mice, which carry the If-1h allele and produce high levels of IFN, high levels of both IFN-alpha and -beta mRNA can be detected in the spleen. In contrast, only very low levels of IFN mRNA could be detected in spleens of infected BALB/c mice containing the If-1l allele and producing low levels of IFN or in B6.C-H28c mice that are congenic for the If-1l allele. The relative levels of all individual IFN-alpha 1, alpha 4, and alpha 6 mRNA in spleens of infected BALB/c were lower than in spleens of infected C57BL/6 mice, indicating that the If-1 locus affects the expression of all IFN-alpha subtypes and is not associated with the deletion or inactivation of a specific IFN gene. The relative levels of IFN regulatory factor-1 mRNA in infected mice carrying the If-1l and If-1h loci were comparable, suggesting that the If-1 regulation is not associated with the altered expression of the IFN regulatory factor-1 gene. Quantitative difference in the expression of IFN-alpha and -beta genes was also observed in in vitro-infected peritoneal macrophages isolated from either C57BL/6 or BALB/c mice. A surprise finding was that the If-1 locus also affected the NDV-induced expression of two other cytokine genes, TNF-alpha and IL-6. Priming of the macrophage cultures with murine IFN enhanced the expression of all cytokine genes, and the relative levels of IFN, TNF-alpha, and IL-6 mRNA induced by NDV in macrophages derived from C57BL/6 and BALB/c mice were comparable. We propose that the If-1 locus affects the early stages of a signal transduction pathway which are common to the virus-mediated induction of IFN, TNF-alpha, and IL-6 genes.

Identification of a novel virus-responsive sequence in the promoter of murine interferon-alpha genes.

We have previously shown that the infection of mouse L-cells with Newcastle disease virus activates transcription of the alpha 4 but not the alpha 6 interferon gene and that the induction is mediated by a 35-base pair inducible element (IE) found in the alpha 4 promoter (-109 to -75). In the present study, we show that the inactivity of the alpha 6 promoter can be mapped to 2 out of 6 nucleotides in which the alpha 6 differs from alpha 4 IE. The symmetrical sequence, GTAAAGAAAGT (-103 to -93), present in the alpha 4 IE is essential for its inducibility and binding of nuclear protein(s) to the alpha 4 IE.

Virus infection and interferon can activate gene expression through a single synthetic element, but endogenous genes show distinct regulation.

Virus inducible elements (IE) in promoters of mouse alpha-interferon and human beta 1-interferon genes contain multiple copies of the hexanucleotide sequence AGT-GAA or its variants which are also found in the interferon-stimulated response element of genes transcriptionally induced by interferon. We have examined the similarities between virus and interferon induction of gene expression and the role of AGTGAA and AAT-GAA hexamers in these responses. Hybrid plasmids were constructed by inserting the IE region, the alpha 4 promoter, or the multiple copies of AGTGAA or AAT-GAA 5' to the inactive-45 human immunodeficiency-chloramphenicol acetyltransferase hybrid gene, and their inducible expression was studied in a transient expression assay. In L-cells, multiple hexamers were efficiently induced both by infection with Newcastle disease virus and by interferon treatment; while the alpha 4 promoter and the IE inducible region were induced predominantly by virus rather than by interferon. In order to dissociate the effect of virus and endogenous interferon on the induction process, we examined the gene expression in Vero cells, which have undergone homozygous deletion of type 1 interferon genes, and in VNPT-159 cells, which were derived from Vero cells by insertion of an inducible human interferon beta 1 gene. The results show that while the alpha 4 promoter was efficiently induced only by virus in both cell types, the constructs containing shorter segments of the IE were induced by both virus and interferon in Vero cells. However, the inducibility by interferon was not detected in VNPT-159 cells, suggesting that the presence of endogenous interferon suppresses interferon-induced expression of hexanucleotide repeats and the short inducible region. In contrast, virus inducibility of endogenous interferon-stimulated genes, ISG-15 and ISG-54, was about 100-fold more efficient in VNPT-159 cells than in Vero cells, suggesting that this induction is largely mediated through synthesis of endogenous interferon. Hence, endogenous interferon may play a role in the autoregulation of both interferon genes and interferon-stimulated genes.

Inhibition of human immunodeficiency virus (HIV) replication by HIV-trans-activated alpha 2-interferon.

We have prepared stable cell lines, derived from Vero cells and A3.01 cells, that express a hybrid human alpha 2-interferon gene under control of the human immunodeficiency virus (HIV) long terminal repeat. These cells constitutively produced low levels (50-150 units/ml) of alpha 2-interferon. However, high levels of interferon (10(3) units/ml) could be induced upon trans-activation by the product of the tat gene (pIIIextatIII), and de novo infection by HIV resulted in a moderate increase (400 units/ml) in alpha 2-interferon synthesis. In contrast to the fully permissive HIV replication, in transfected Vero cells or infected A3.01 cells, the transcription and replication of HIV in Vero or A3.01 cells containing the HIV long terminal repeat--alpha 2-interferon hybrid gene (VN89 and A3N89 cells, respectively) was completely inhibited. These data suggest that virus-trans-activated alpha 2-interferon synthesis can be used as a selective inhibitor of HIV replication.

Upstream regulatory elements of murine alpha 4-interferon gene confer inducibility and cell type-restricted expression.

We have identified and functionally characterized DNA sequences that are required for the inducible and cell-restricted expression of the murine alpha 4-interferon gene. Hybrid plasmids in which the alpha 4 promoter region or its 5' deletions were inserted upstream of the CAT gene were constructed, and the expression of these hybrid genes was studied in mouse L-cells both in permanent and transient assays with comparable results. Inducible expression was not affected by deletions up to -109; however, when the deletion was extended to -96, inducibility by Newcastle disease virus was abolished; however, this hybrid plasmid was expressed constitutively. Further deletion to -88 did not permit either constitutive or inducible expression. Insertion of the 35-base pair-long sequence (-109 to -75 base pairs) from the alpha 4 promoter region 5' of the minimal alpha 4 or human immunodeficiency virus promoter region, conferred inducibility to these two inactive promoters. The 5' deleted hybrids or plasmids containing the inducible element were induced only at low levels in transfected NIH/3T3 cells that do not express endogenous alpha 4 gene efficiently, indicating that the inducible region also determines the cell-specific expression. A tandem repeat of AGTGAA, which is present in the -109 to -88 region of alpha 4 in two copies, showed both basal levels of expression and inducibility in L-cells, while its analogue AATGAA was highly inducible but was not expressed constitutively. The inducibility of the synthetic hexamer repeats did not show cell type-restricted expression, suggesting that their response does not fully reflect the range of expression observed for the inducible region and the endogenous alpha genes.

Differential and cell type specific expression of murine alpha-interferon genes is regulated on the transcriptional level.

In mouse cells induced with virus infection or dsRNA, the relative levels of alpha-4 interferon mRNA were higher than the levels of alpha-1 and alpha-6 mRNAs; the ratio between relative levels of alpha-4 and alpha-1 or alpha-6 mRNA was, however, dependent on the cell type. Recombinant plasmids, in which the expression of chloramphenicol acetyltransferase (CAT) gene was directed by the promoter regions of alpha-1, alpha-4 or alpha-6 interferon genes were constructed and their inducible expression was studied either in transient assay or in permanently transfected mouse cells. The highest levels of CAT activity and CAT mRNA were observed with alpha-4 CAT plasmid, while the expression of alpha-1 CAT was consistently higher than that coded by alpha-6 CAT plasmid; the ratio between CAT activities coded by alpha-4 CAT and alpha-1 CAT was dependent on cell type. However, in heterologous Vero cells, the transfected alpha-1 and alpha-4 genes were expressed constitutively, and the levels of mRNAs were comparable. These results show that the difference in the relative levels of individual alpha-1 and alpha-4 mRNAs reflects the transcriptional inducibility of the respective promoter regions.

Synthesis, antiviral activity, and conformational characterization of mouse-human alpha-interferon hybrids.

Reciprocal hybrids were constructed between human and mouse interferons (IFNs), and their antiviral activity was examined on different target cells and compared to the activity of the parental molecules. In addition, we used a number of predictive algorithms on a data base of the available alpha-interferon sequences to propose a working model for the overall conformation of the alpha-interferon molecule that is consistent with the structural predictions. Remarkable conservation within the predicted alpha-helical segments of the interferon molecule was observed. We propose that the observed changes in the activity and specificity of the hybrids obtained are largely due to the sequences present in the loops at the ends of the major helical structures; these are less conserved, contain beta-bends, and are generally hydrophilic and flexible. The data on the constructed mouse-human hybrids have shown that the activity on human cells is contributed by determinants present in the N-terminal 122 amino acids of human IFN, thus implicating one or more loops within this region (e.g. loops 1-12, 25-38, 70-74, and 103-113). The activity on bovine cells appears to be localized mainly in sequence 60-121, implicating the role of loops 70-74 and/or 103-113 of the human IFN molecule. The specificity of mouse IFN for mouse cells is in some or all of the loops (70-74, 103-113, 134-139, and 163-166) in the C-terminal sequence. The proposed working model should provide guidelines for the study of the specificity of action in molecular terms.

Human fibroblast but not lymphoid cells have unusually long polyadenylated interferon-beta 1 mRNAs.

The transcription unit of human interferon-beta 1 (IFN-beta 1) mRNA was examined by chain elongation of nascent RNA in isolated nuclei of human fibroblasts and lymphoid cells induced to produce IFN. In fibroblasts, transcription proceeds beyond 2,400 nucleotides downstream from the poly(A) site of mature mRNA and appears to terminate in the region rich in Alu sequences. Northern hybridization showed the presence of a minor polyadenylated RNA species, about 3,200 nucleotides long, that hybridized to the probes derived from 3'-flanking regions of IFN-beta 1 mRNA. S1 nuclease analysis established that this long polyadenylated transcript represents a mixture of three RNA molecules with defined 3' termini. In all three mRNAs, as in mature IFN-beta 1 mRNA, the polyadenylation site was located within a few nucleotides downstream from the AAUAAA hexanucleotide consensus sequence. Surprisingly, in Namalva lymphoblastoid cells no transcription beyond the polyadenylation site of mature IFN-beta 1 mRNA could be detected either in isolated nuclei or total RNA.

Activation of human immunodeficiency virus by herpesvirus infection: identification of a region within the long terminal repeat that responds to a trans-acting factor encoded by herpes simplex virus 1.

Herpes simplex virus 1 (HSV-1) infection induces transcription of the chloramphenicol acetyltransferase (CAT) gene directed by the long terminal repeat (LTR) of human immunodeficiency virus (HIV) in both transiently and permanently transfected cells containing the HIV-LTR/CAT hybrid gene. To define the mechanism by which HSV-1 stimulates the HIV LTR, we examined the effects of isolated regulatory genes from HSV-1. The results of cotransfection assays with the immediate-early (IE) genes of HSV-1, IE110 (ICP0) and IE175 (ICP4), showed that the IE110 protein, either alone or in combination with the IE175 protein, can activate the HIV LTR. Cotransfection with the IE175 gene alone or with the Vmw65 gene (coding for a virion transcription factor) alone did not lead to HIV-LTR activation. The lack of requirement for the IE175 or Vmw65 gene products in transient-expression assays was confirmed in permanent cell lines containing the HIV-LTR/CAT hybrid gene by using temperature-sensitive mutants defective in the IE175 gene product or in uncoating functions. By deletion analysis, we localized a 73-bp-long region (positions -104 to -32) from the HIV LTR that responded to HSV-1 activation; when this region, which is distinct from the previously identified trans-activating responsive (TAR) region, was ligated to a heterologous, HSV-1-nonresponsive gene (alpha 4-interferon/CAT), it conferred inducibility by both HSV-1 infection and IE110/175 cotransfection. Both simian and human cytomegalovirus also induced the HIV-LTR/CAT hybrid gene. However, we failed to detect specific upstream sequence requirements for induction by cytomegalovirus. Our results indicate that infection with unrelated viruses can alter the expression of HIV in an infected cell.