The basic domain/leucine zipper protein hXBP-1 preferentially binds to and transactivates CRE-like sequences containing an ACGT core.

The transcription factor hXBP-1 belongs to the family of basic region/leucine zipper (bZIP) proteins and interacts with the cAMP responsive element (CRE) of the major histocompatibility complex (MHC) class II A alpha, DR alpha and DP beta genes. However, the developmental expression of hXBP-1 as revealed by in situ hybridization in mouse embryos, has suggested that it interacts with the promoter of additional genes. To identify other potential target genes of this factor, we performed binding site selection experiments with recombinant hXBP-1 protein. The results indicated that hXBP-1 binds preferably to the CRE-like element GAT-GACGTG(T/G)NNN(A/T)T, wherein the core sequence ACGT is highly conserved, and that it also binds to some TPA response elements (TRE). hXBP-1 can transactivate multimers of the target sequences to which it binds in COS cells, and the level of transactivation directly correlates with the extent of binding as observed in gel retardation experiments. One target sequence that is strongly bound by hXBP-1 is the 21 bp repeat in the HTLV-1 LTR, and we demonstrate here that hXBP-1 can transactivate the HTLV-1 LTR. Further, the transactivation domain of hXBP-1 encompasses a large C-terminal region of the protein, containing domains rich in glutamine, serine and threonine, and proline and glutamine residues, as shown in transient transfection experiments using hXBP-1-GAL4 fusion proteins and a reporter gene under the control of GAL4-binding sites.

Chondrodysplasia and neurological abnormalities in ATF-2-deficient mice.

Activating transcription factor-2 (ATF-2) is a basic region leucine zipper protein whose DNA target sequence is the widely distributed cAMP response element (CRE). We report here that mice carrying a germline mutation in ATF-2 demonstrated unique actions of ATF-2 not duplicated by other ATF/CREB family members. Mutant mice had decreased postnatal viability and growth, with a defect in endochondral ossification at epiphyseal plates similar to human hypochondroplasia. The animals had ataxic gait, hyperactivity and decreased hearing. In the brain, there were reduced numbers of cerebellar Purkinje cells, atrophic vestibular sense organs and enlarged ventricles. Unlike CREB alpha/delta-deficient mice whose main defect is in long-term potentiation, the widespread abnormalities in ATF-2 mutant mice demonstrate its absolute requirement for skeletal and central nervous system development, and for maximal induction of select genes with CRE sites, such as E-selectin.

In situ hybridization studies suggest a role for the basic region-leucine zipper protein hXBP-1 in exocrine gland and skeletal development during mouse embryogenesis.

The spatial and temporal distribution of transcripts for the TRE/CRE-binding basic region-leucine zipper protein hXBP-1 was determined by in situ hybridization. Analysis of embryos from day 10.5 to 18.5 pc revealed high level expression of hXBP-1 RNA in two developing organ systems: 1) in bone and cartilage cells of the developing skeleton and toothbuds, and 2) in exocrine glands including the pancreas and the submandibular and salivary glands. High level expression was also found in whisker follicles and in selected cells in brown adipose tissue. In the developing skeleton, hXBP-1 RNA was expressed starting on day 11.5 pc in osteoblasts of newly formed intramembranous bone. Thereafter, hXBP-1 was expressed in both osteoblasts and preosteoblasts in bone formed directly by intramembranous formation as well as in bone formed during endochondral ossification. The most intense signal was observed in preosteoblasts and osteoblasts of newly forming bone. At day 11.5 pc low level hXBP-1 expression was also observed in matrix secreting chondroblasts of bones which are formed initially of cartilage, at the stage where they consist entirely of cartilage. Signal was also present in matrix producing chondroblasts of the mature zone of the growth region during endochondral ossification although at significantly lower level than in osteoblasts. hXBP-1 is thus the first transcription factor described, to our knowledge, whose level of expression is modulated during the osteoblast developmental sequence in vivo. The pattern of expression of hXBP-1 in the developing skeleton was found to be very similar to that of the genes encoding the tissue inhibitor of metalloproteinase and alkaline phosphatase throughout development. These observations suggest that hXBP-1 may play a role in regulating the expression of tissue specific genes (TIMP, osteonectin, osteopontin, osteocalcin) expressed in osteoblasts. It is intriguing that the promoter regions of several such genes contain potential hXBP-1 binding sites.

Post-transcriptional effect of ultraviolet light on gene expression in human cells. Stabilization of cytokine-induced and poly(I).poly(C)-induced messenger RNA.

It is well established that ultraviolet light modulates gene expression in mammalian cells, particularly at transcriptional and post-translational levels. The present study was undertaken to investigate whether the fate of mRNA is also altered in ultraviolet-light-irradiated human cells. In order to facilitate distinction between transcriptional and post-transcriptional effects, this analysis has focused on six genes whose transcription is conditional on the supply of exogenous inducers, interferon-alpha, interleukin-1 alpha or the double-stranded RNA, poly(I).poly(C). Human cells induced to express these genes were found to retain a significantly higher concentration of corresponding transcripts when irradiated with ultraviolet light at the end of the inducing treatment. This stimulation was due to dose-dependent ultraviolet-light stabilization of preformed mRNA, as shown by run-on and pulse/chase experiments. This work uncovers a new facet of the cellular response to genotoxic stresses, i.e. extension of the life-span of transcription products. Whether this stabilizing effect contributes to cell recovery by promoting gene expression remains to be determined.

Human thymosin-beta 4/6-26 gene is part of a multigene family composed of seven members located on seven different chromosomes.

We have isolated a cDNA encoding the human interferon-inducible gene 6-26, by screening a cDNA library with an oligodeoxynucleotide probe. Its sequence was found to be identical to that of the human thymosin-beta 4 cDNA, which encodes a protein present in most cell types, but whose function is not clear at present. By hybridization of the thymosin-beta 4/6-26 cDNA to the DNA of a panel of human-rodent somatic cell hybrids, we found that at least seven genes homologous to this cDNA are present in the human genome. We localized these genes, some of which might be pseudogenes, to seven distinct chromosomes, namely, chromosomes 1, 2, 4, 9, 11, 20, and X.

Analysis of interferon-inducible genes in cells of chronic myeloid leukemia patients responsive or resistant to an interferon-alpha treatment.

Recombinant human interferon-alpha (IFN-alpha) can induce a hematologic remission in patients with chronic myeloid leukemia. However, some patients are resistant and others develop late resistance to the IFN-alpha treatment. To understand the molecular mechanism of this resistance, we have analyzed the expression of 10 IFN-inducible genes in the cells of three resistant patients, two responsive patients, and six healthy controls. Northern blot hybridizations showed that all the genes were induced in in vitro IFN-alpha treated peripheral blood cells of the patients and healthy controls. These genes were also inducible in peripheral blood and bone marrow cells of two out of two resistant patients administered an injection of IFN-alpha. We conclude that the resistance to the IFN-alpha treatment of the chronic myeloid leukemia patients we studied is not due to (1) the absence of induction of any of the 10 IFN-inducible genes we studied, including the low-molecular-weight 2'-5'oligoadenylate synthetase; (2) the presence of an antagonist of IFN-alpha in the peripheral blood or bone marrow cells; and (3) the presence of neutralizing anti-IFN-alpha antibodies.

Chromosomal localization of two human genes inducible by interferons, double-stranded RNA, and viruses.

Human IFI-15K and 6-16 genes are transcriptionally stimulated by interferons, double-stranded RNA, and viruses. By screening a cDNA library with oligodeoxynucleotide probes, we have isolated complete copies corresponding to these two genes. These cDNA clones allowed us to localize the IFI-15K and 6-16 genes on human chromosome 1 by somatic cell hybridization.

2-Aminopurine selectively blocks the transcriptional activation of cellular genes by virus, double-stranded RNA and interferons in human cells.

Interferons, double-stranded RNA and viruses induce the transcription of partly overlapping sets of cellular genes. We have studied the regulation of 11 interferon-inducible genes by these agents and found that four of them were also directly inducible by virus and double-stranded RNA, and two by virus only. We have investigated whether an inhibitor of interferon-beta gene activation, 2-aminopurine, would block this induction process. Induction of these genes by virus and double-stranded RNA was indeed blocked by 2-aminopurine. Since a single cis-acting element can confer inducibility both to interferons, and to virus and double-stranded RNA, we tested the effect of 2-aminopurine on gene activation by interferon-alpha and interferon-gamma. Remarkably, in all the cell lines tested, these induction processes and the establishment of an antiviral state were blocked by the drug. These observations contrast with previous reports. The inhibitory effect of this drug on gene induction was exerted in a selective fashion and at the transcriptional level. This indicate that for the virus-, double-stranded-RNA-, and interferons-mediated gene induction, an early and similar step in signal transduction was affected by 2-aminopurine.

Cloning and chromosomal location of human genes inducible by type I interferon.

When cells are treated with interferon several new proteins are induced. We have isolated by differential screening two cDNA clones corresponding to human genes inducible by IFN-alpha, termed IFI-4 and IFI-54K. The accumulation of the corresponding mRNA was followed as a function of either IFN dose or of time. The IFI-4 and IFI-54K genes, as well as two previously isolated IFN-inducible genes, namely the IFI-56K and low-molecular-weight 2-5A synthetase, were localized on the human chromosomes. Using cloned probes on Southern blots of DNA from a panel of rodent-human somatic cell hybrids, we have assigned the IFI-4 gene to chromosome 1 and the gene coding for the low-molecular-weight 2-5A synthetase to chromosome 12. We also showed that the IFI-54K and IFI-56K genes, unlike most of the IFN-inducible genes, are syntenic. They are both located on chromosome 10. In addition, evidence is given for the presence of a pseudogene homologous to IFI-56K on chromosome 13.

Regulation of two interferon-inducible human genes by interferon, poly(rI).poly(rC) and viruses.

The IFI-56K and IFI-54K genes are transcriptionally stimulated when cells are treated by interferon. We have previously shown that the IFI-56K gene is in addition directly induced by poly(rI).poly(rC), and inducer of interferon-beta. Since the regulation of the IFI-56K and IFI-54K genes by interferon are very much alike, we tested whether the IFI-54K gene is also directly regulated by poly(rI).poly(rC). Treatment of various cell lines with poly(rI).poly(rC) leads to a clear accumulation of the IFI-54K mRNA to a level which sometimes even exceeds that obtained with high doses of interferon. Several interferon-resistant cell lines were investigated for the inducibility of both the IFI-56K and IFI-54K genes by interferons, poly(rI).poly(rC) and viruses (which are the natural inducers of interferon-alpha and -beta). Both genes appear to be coordinately regulated by these inducers. It was thus interesting to search for common regulatory element(s) in the control region of these two genes. The IFI-54K gene promoter region was isolated, from which a 520-base-pair segment was sequenced and compared with the promoter region of the IFI-56K gene that we had previously sequenced. The only homology was found is a well conserved 19-bp segment located just upstream of the TATA box of these genes; interestingly, this sequence is also homologous to the minimal region needed for the inducibility by poly(rI).poly(rC) of the interferon-beta gene. This conserved sequence might be responsible for the coordinate induction of the IFI-56K and IFI-54K genes by interferon, poly(rI).poly(rC) and viruses.

The IFI-56K and IFI-54K interferon-inducible human genes belong to the same gene family.

The IFI-56K and IFI-54K human genes are coordinately regulated by interferon, double-stranded RNA and viruses in a number of cell lines. These genes encode polypeptides of 56 and 54 kDa, respectively, whose function remains to be determined. We analysed the possible structural relatedness between these syntenic and similarly regulated genes. We found that they are very closely related at the protein, mRNA and promoter levels. This suggests that the IFI-56K and IFI-54K genes are members of a gene family, which probably arose from duplication of an ancestor gene.

New inducers revealed by the promoter sequence analysis of two interferon-activated human genes.

In order to investigate the molecular basis of the regulation of interferon-inducible genes, we isolated the promoter region of two such genes coding for the (2'-5')oligo(adenylate) synthetase and a 56-kDa protein (IFI-56K). The regions surrounding the cap site were sequenced and compared with the sequences of vertebrate and viral DNA present in the Genbank data bank. Small DNA segments were found in both genes which are homologous to part of the promoter region of other genes, such as those of interferon-beta, tumor necrosis factor beta, interleukin-2 and its receptor. Since these homologies were found located in functionally important regions of these genes, we tested whether their inducers also enhance the (2'-5')oligo(adenylate) synthetase and IFI-56K gene expression. We found that poly(rI).poly(rC) and interleukin-1, activators of the interferon-beta gene and of T lymphocytes respectively, are both able to enhance IFI-56K mRNA accumulation in all cell lines tested. Cycloheximide even superinduces this gene when added together with poly(rI).poly(rC) and interleukin-1 (but not when added with interferon). We showed that these inductions are direct and not mediated by interferon produced by cells in response to poly(rI).poly(rC) or interleukin-1. The promoter sequence analyses have thus led to the discovery of unexpected inducers, i.e. an interferon inducer such as poly(rI).poly(rC) is also able to directly induce a gene that is under the control of interferon.