Two mRNA species that produce biologically active interferon were isolated from human fibroblasts and studied by size fractionation and cloning in Escherichia coil plasmid pBR322. The major fibroblast interferon (Hu IFN-B1) is coded for by the smaller of the two mRNAs, an 11S species, 900 nucleotides long, which in cell-free systems yields a 20,000 M, protein. The second interferon mRNA species (Hu is 14S, about 1300 nucleotides long, and codes for another protein of 23,000-26,000 Mr. The two interferon mRNAs do not cross-hybridize. Both are induced by poly(rIrC), but IFN-,2 mRNA is induced to about 10% in cells by cycloheximide treatment alone whereas under these conditions IFN-P1 is not induced.The major interferon produced by human fibroblasts was recently purified and its amino-terminal amino acid sequence determined by Knight et al. (1). An interferon mRNA 12S fraction from human fibroblasts was cloned in Escherichia coli, and the recombinant DNA nucleotide sequence of the resulting clones was shown to contain the codons corresponding to the structure determined by protein sequencing (2, 3). During the course of similar studies, we observed that human fibroblasts actually contain two interferon mRNAs. As shown here, these two RNAs differ in their size and in their translation products; studies of cDNA recombinant clones in E. coli show that the two mRNAs do not readily cross-hybridize. Sehgal and Sagar (4) have also recently succeeded in separating the two interferon mRNAs from human fibroblasts by mercury-agarose gel electrophoresis.MATERIALS AND METHODS mRNA Preparation and Translation. Poly(rI-rC) superinduction of human FS11 fibroblasts, preparation of poly(A)+ mRNA, sucrose gradient purification, and mRNA translation in micrococcal nuclease-treated rabbit reticulocyte lysates were described in detail (5). Antibodies to human fibroblast interferon were produced and used for immunoprecipitation of the [35S]methionine-labeled translation products with staphylococcal protein A-Sepharose, as before (5). Interferon mRNA injection into Xenopus laevis oocytes was carried out according to Raj and Pitha (6) Assay of Interferon by (2'-5')Oligo(A) Synthetase E. Translation products from reticulocyte lysates or from the medium of mRNA-injected oocytes were diluted 1:10 to 1:15 and 0.1 ml was added to a 96-well microtiter plate. After 18 hr at 37°C, cells were lysed with Nonidet P40, and the lysates were adsorbed on poly(rI-rC)-agarose beads, which were then incubated with a-[32P]ATP (0.3 Ci/mmol; 2.5 mM; 1 Ci = 3.7 X 1010 becquerels) as detailed (7). After 20 hr at 30°C, the supernatant was digested by bacterial alkaline phosphatase and the amount of (2'-5')ApA formed was analyzed by paper electrophoresis at pH 3.5.Cloning of cDNAs. Procedures established by Rougeon (8) were used. cDNA was prepared from sucrose gradient RNA fractions (4 ,ug) with reverse transcriptase (RNA-dependent DNA polymerase) from avian myeloblastosis virus (J. Beard), (dT)12_18, and 4 mM pyrophosphate (9). Double-stranded cDNA was ...

In human cells treated with interferons, there is an increase in the amount of HLA-ABC and (32-microglobulin exposed on the cell surface. We have used a cloned HLA-A,B,C cDNA probe to demonstrate by molecular hybridization that this effect of interferon is preceded by a large increase in the amount of HLA mRNA in the cell. Thiseffect was found in five different human cell lines, with purified leukocyte and fibroblast interferons. The increase in HLA mRNA is comparable in its Idnetics and dose-response to the induction of (2'-5') oligo(A) synthetase mRNA by interferons. Therefore, interferons seem to activate at least two cellular genes which have different biochemical functions. Studies on the mechanism by which interferons (IFNs) inhibit virus multiplication have led to the discovery of several enzymes, inhibitors ofprotein biosynthesis, which are induced in cells exposed to IFNs (for review, see refs. 1-3). An assay for the mRNA of one of these enzymes, (2'-5') oligoadenylate synthetase (synthetase E), was developed and it was shown that IFNs induce, within a few hours, the accumulation of specific synthetase E mRNA in the cytoplasm of treated cells (4). Inhibitors of transcription and translation block the induction of synthetase E (5-8). The same is true of the IFN-induced eukaryotic initiation factor 2 protein kinase (5) and of 5-10 other proteins which can be detected by one-or two-dimensional polyacrylamide gel electrophoresis (9-13). Induction of these proteins by IFN therefore is likely to take place at the gene level.In addition to the above effects, IFNs also cause complex changes in the plasma membranes of cells (14). These changes alter the electric charge ofthe cells (15), decrease their motility (16), and influence cell-cell recognition events involved in the immune response (14,17). Biochemically, these membrane changes affect both the lipids (18) and several surface protein antigens (19,20). Thus, human IFNs increase the amounts of HLA-A,B,C antigens and of P2-microglobulin available for interaction with specific antibodies on the surface of human cells (21,22). These changes are specific because there is no increase in HLA-DR or in many other surface antigens. We wondered whether these membrane effects of IFNs were topical modifications in the exposure ofantigens on the cell surface or involved changes in gene expression similar to those found for synthetase E.To investigate this question we used a cloned HLA cDNA probe, pHLA-1 (23), to study the HLA-A,B,C mRNA levels in IFN-treated cells. Our results show that, in five human cell lines of lymphoblastoid and fibroblastic origins, both leukocyte IFN (IFN-a), and fibroblast IFN (IFN-,B) produce a large and rapid increase in HLA-A,B,C mRNA, which often precedes the increase in synthetase E mRNA. air. All cultures were grown at 37°C in the presence of 100 units of penicillin and 100 pg of streptomycin per ml. Human IFN-a was purified (31) from Sendai virus-infected chronic myelogenous leukemic cells, obtained from Institut Merieux (Lyon, F...

Two cis-acting elements, the enhancer and the promoter,, independently contribute to the cell-specific expression of the rat insulin 1 gene. The activities of these elements are presumably mediated by trans-acting factors. We have performed intracellular competition experiments that suggest the presence of a negative factor(s) that represses the enhancer activity in cells that do not express the insulin gene.In these experiments fibroblast cells (COS-7) were transfected with two plasmids: a test plasmid containing the gene for chlorampheiiicol acetyltransferase under the control of the thymidine kinase promoter and the insulin enhancer; and a competitor plasmid containing insulin enhancer sequences and the simian virus 40 origin of replication to permit its replication in the recipient cells. The presence of the competitor plasmid led to a 5-to 6-fold -increase in chloramphenicol acetyltransferase activity as compared with the activity detected when insulin enhancer was absent from either the competitor or the test plasmid. A 5-fold increase in chloramphenicol acetyltransferase activity was also seen when the rat amylase enhancer was present on the competitor plasmid; in contrast the simian virus 40 enhancer exerted no effect. Efficient derepression required additional sequences downstream from those essential for enhancer activity. We propose that the activity ofthe rat insulin 1 enhancer is modulated by a negative tbans-acting factor(s) that is active in cells not expressing insulin but is overridden by the dominant positive trans-acting factor(s) present in insulinproducing cells.During cellular differentiation, cell types acquire the ability to stably express a characteristic set of genes. The. molecular mechanisms by which this occurs are poorly understood. Several genes whose expression is restricted to a specific subset of cells contain cis-acting sequences required for efficient transcription in the appropriate differentiated cells (1-6). We have demonstrated (7) that 5' flanking DNA sequences of the rat insulin 1 gene contain two such cellspecific elements. The properties of these elements are consistent with the hypothesis that differentiated cells contain positive trans-acting proteins (differentiators) that interact with the specific cis-acting sequences to stimulate the expression of the associated genes. This idea is also consistent with the observations and concepts of other workers (8-12). On the other hand, "extinction" of the differentiated phenotype after fusion of differentiated cells with fibroblasts is frequently observed (13). We have also noted that specific sequence deletions in the insulin gene 5' flanking sequences can lead to increases in expression in nondifferentiated cells (unpublished observations). These results are consistent with a role for repressor-like molecules in nonexpressing cells.

DNA from a human adult was fragmented by partial digestion with restriction endonuclease EcoRI and cloned in λ Charon 4A. Clone C15, with a human DNA insert of 17 × 103 bases, was identified as containg a gene for the fibroblast, interferon, interferon β1. Restriction mapping shows that this gene, located on a 1840‐base EcoRI fragment, is not interrupted by introns. Moreover, we show that this human genomic DNA fragment is able to direct the synthesis of active human interferon β1 in Escherichia coli. Interferon activity of up to 7 × 106 U/1 was recovered from phage lysates by chromatography on CiBacron blue–Sepharose, and had the same immunological properties and species specificity as interferon produced by human fibroblasts.

Induction of IFN-beta 1 RNA was studied in the mouse cell line SR117-21E transformed by a BPV episome containing the human IFN-beta 1 gene deleted of promoter sequences upstream from position -40. Nuclei isolated from these cells synthesize constitutively IFN-beta 1 RNA from the partially deleted promoter. The IFN-beta 1 RNA synthesized by nuclei of uninduced SR117-21E cells is similar to that made by nuclei of poly(rI):(rC)-induced cells, but does not accumulate and hence no IFN is produced unless the cells have been treated either by ds RNA or by cycloheximide. We conclude that the IFN-beta 1 gene has, in addition to the transcription control due to upstream promoter sequences, an additional post-transcriptional control acting on mRNA accumulation and linked to sequences close to the TATA box and RNA start site. Both controls are relieved by ds RNA.