Type I interferons are essential mediators of apoptotic death in virally infected cells.

BACKGROUND: The interferons (IFNs) have been extensively studied in the context of host defence against viral infection. In the established model of IFN action, virally infected cells secrete type I IFNs (IFN-alpha/beta) which induce an antiviral state in uninfected cells. However, it is not clear how IFNs function on the infected cells. It has been reported that cells infected by some viruses die by apoptosis. RESULTS: In the present study, we found that three types of viruses commonly induce apoptosis in primary cell cultures. Importantly, we observed that virus-induced apoptosis was inhibited by anti-IFN-alpha/beta antibodies, and in cells lacking either the type I IFN receptor 1 (IFNAR1) or its downstream mediator, Stat1 (Signal transducer and activator of transcription 1). IFN-alpha treatment by itself did not induce apoptosis unless it was combined with transfection by double-stranded RNA (dsRNA), which is normally generated during the course of viral infection. CONCLUSION: These results indicate a novel antiviral function of the type I IFNs, i.e. the selective induction of apoptosis in virally infected cells. In effect, these IFNs have a bifunctional role in limiting the spread of virus; eliciting an antiviral state in uninfected cells while promoting apoptosis in infected cells. Our results may help explain why IFNs are sometimes useful in the treatment of viral diseases and will provide further insight into the mechanisms of virus-induced pathogenesis.

DNA damage-induced apoptosis and Ice gene induction in mitogenically activated T lymphocytes require IRF-1.

Lymphocytes are highly sensitive to DNA damage-induced apoptosis. In thymocytes, the tumor suppressor p53 has been shown to be required for this type of apoptosis. However an as yet unknown, p53-independent pathway(s) appears to mediate the same event in mitogenically activated mature T lymphocytes. By using mice with a null mutation in the IRF-1 gene, we revealed that DNA damage-induced apoptosis in the latter cell type is dependent on the anti-oncogenic transcription factor interferon regulatory factor-1 (IRF-1). Thus two different anti-oncogenic transcription factors, p53 and IRF-1, are required for distinct apoptotic pathways in T lymphocytes. Furthermore, we found that mitogen induction of the interleukin-1 beta-converting enzyme (Ice) gene, a mammalian homolog of the Caenorhabditis elegans cell death gene ced-3, is also IRF-1-dependent. An IRF-1 binding sequence was identified in the 5' flanking region of the Ice gene. In addition, ectopic overexpression of IRF-1 results in the activation of the endogenous Ice gene and enhances the sensitivity of cells to radiation-induced apoptosis. Thus, induction of Ice gene may be involved in IRF-1 dependent DNA damage-induced apoptosis in activated mature T lymphocytes.

Cooperation of the tumour suppressors IRF-1 and p53 in response to DNA damage.

Normally growing cells promptly cease DNA synthesis when exposed to genotoxic stresses, such as radiation, and this cell-cycle arrest prevents the accumulation of mutations. The transcription factor interferon regulatory factor (IRF)-1 is essential for the regulation of the interferon system, inhibits cell growth, and manifests tumour-suppressor activities. Here we show that mouse embryonic fibroblasts (EFs) lacking IRF-1 are deficient in their ability to undergo DNA-damage-induced cell-cycle arrest. A similar phenotype has been observed in EFs lacking the tumour suppressor p53 (refs 8, 9), although the expression of IRF-1 and p53 are independent of one another. Furthermore, we show that transcriptional induction of the gene encoding p21 (WAF1, CIP1), a cell-cycle inhibitor, by gamma-irradiation is dependent on both p53 and IRF-1, and that the p21 promoter is activated, either directly or indirectly, by both in a transient cotransfection assay. These two tumour-suppressor transcription factors therefore converge functionally to regulate the cell cycle through the activation of a common target gene.

An IRF-1-dependent pathway of DNA damage-induced apoptosis in mitogen-activated T lymphocytes.

Lymphocytes are particularly susceptible to DNA damage-induced apoptosis, a response which may serve as a form of 'altruistic suicide' to counter their intrinsic high potential for mutation and clonal expansion. The tumour suppressor p53 has been shown to regulate this type of apoptosis in thymocytes, but an as yet unknown, p53-independent pathway(s) appears to mediate the same event in mitogen-activated mature T lymphocytes. Here we show DNA damage-induced apoptosis in these T lymphocytes is dependent on the antioncogenic transcription factor interferon regulatory factor (IRF)-1. Thus two different anti-onco-genic transcription factors, p53 and IRF-1, are required for distinct apoptotic pathways in T lymphocytes. We also show that mitogen induction of the interleukin-1 beta converting enzyme (ICE) gene, a mammalian homologue of the Caenorhabditis elegans cell death gene ced-3, is IRF-1-dependent. Ectopic overexpression of IRF-1 results in the activation of the endogenous gene for ICE and enhances the sensitivity of cells to radiation-induced apoptosis.

Accelerated exon skipping of IRF-1 mRNA in human myelodysplasia/leukemia; a possible mechanism of tumor suppressor inactivation.

The transcription factor IRF-1 has been shown to function as a tumor suppressor. Here we report that a significant proportion of the IRF-1 mRNA detected in normal human hematopoietic cells and cultured cell lines lacks exon 2 (containing the AUG initiation codon) and 3 as a result of exon skipping. Surprisingly, when we examined the bone marrow and peripheral mononuclear cells from patients with myelodysplastic syndrome (MDS) or leukemia secondary to MDS, we could still detect the exon-skipped form but little or none of the intact IRF-1 mRNA. This appears to be the result of accelerated exon skipping since we could find no mutations within the exons and splicing junctions from these patients. The exon-skipped form of IRF-1 lacking exons 2 and 3 displayed neither DNA binding nor tumor suppressive activities. Thus this accelerated exon skipping may cause the inactivation of IRF-1 and thereby contribute to the development of human hematopoietic malignancies.

Cellular commitment to oncogene-induced transformation or apoptosis is dependent on the transcription factor IRF-1.

The transcriptional activator interferon regulatory factor 1 (IRF-1) and its antagonistic repressor IRF-2 are regulators of the interferon (IFN) system and of cell growth. Here we report that embryonic fibroblasts (EFs) from mice with a null mutation in the IRF-1 gene (IRF-1-/- mice) can be transformed by expression of an activated c-Ha-ras oncogene. This property is not observed in EFs from wild-type or IRF-2-/- mice but is still observed in EFs from mice deficient in both genes. The transformed phenotype of ras-expressing IRF-1-/- EFs could be suppressed by the expression of the IRF-1 cDNA. Thus, IRF-1 functions as a tumor suppressor. Furthermore, expression of the c-Ha-ras oncogene causes wild-type but not IRF-1-/- EFs to undergo apoptosis when combined with a block to cell proliferation or treated by anticancer drugs or ionizing radiation. Hence, IRF-1 may be a critical determinant of oncogene-induced cell transformation or apoptosis.

The oncogenic transcription factor IRF-2 possesses a transcriptional repression and a latent activation domain.

IRF-1 and IRF-2 are two structurally related transcription factors originally identified as regulators of the type I interferon (IFN) system. IRF-1 functions as an activator whereas IRF-2 binds to the same cis-elements and can repress IRF-1 action. More recently these two factors have been shown to act in a mutually antagonistic manner to regulate cell growth; overexpression of the repressor IRF-2 leads to cell transformation, whereas concomitant overexpression of IRF-1 leads to reversion. Previous studies have identified DNA-binding domains in IRF-1 and IRF-2 and an activation domain in IRF-1. In the present study we show that IRF-2 also possesses a transcriptional repression domain in its carboxyl terminal region. We further observe that a LexA-IRF2 fusion can inhibit the function of an activator positioned nearby in the promoter. Thus, repression by IRF-2 may involve both competition with IRF-1 for binding to the promoter as well as the 'silencing' of nearby activators. Furthermore, we demonstrate the presence of a latent activation domain in the central region of IRF-2 and speculate that IRF-2 may contribute to gene activation under certain conditions.

Multiple mechanisms mediate glucose repression of the yeast GAL1 gene.

Several mechanisms contribute to the glucose repression of the GAL1 gene in Saccharomyces cerevisiae. We show that one mechanism involves the transcriptional down-regulation of the GAL4 gene and a second requires the GAL80 gene. We also examine the contribution of cis-acting negative elements in the GAL1 promoter to glucose repression. In an otherwise wild-type strain disruption of any one of these three mechanisms alleviates repression of GAL1 only 2- to 4-fold. However, in the absence of the other two mechanisms the transcriptional down-regulation of GAL4 is sufficient to repress GAL1 expression 40- to 60-fold and the GAL80-dependent mechanism is sufficient to repress GAL1 expression 20- to 30-fold. These first two mechanisms constitute a functionally redundant system of repression and both must be disrupted in order to abolish glucose repression of GAL1. In contrast, negative elements in the GAL1 promoter are effective in repressing GAL1 expression 2- to 4-fold in glucose medium only when at least one of the other two mechanisms of repression is present. Thus, glucose repression of GAL1 is mediated primarily by the first two mechanisms, whereas the third mechanism supplements repression severalfold.