The rapid inactivation of nuclear tyrosine phosphorylated Stat1 depends upon a protein tyrosine phosphatase.

After interferon-gamma (IFN-gamma) treatment of cells the appearance of tyrosine phosphorylated Stat1 in the nucleus was maximal within 20-30 min, remained for 2-2.5 h and activated molecules disappeared by 4 h. In the absence of continued signaling from the receptor (imposed by staurosporine treatment) previously activated Stat1 disappeared completely within 60 min, implying continuous generation and removal of active molecules during extended IFN-gamma treatment. Proteasome inhibitors prolonged the time of activation of Stat1 by prolonging signaling from the receptor but not by blocking removal of already activated Stat1 molecules. By analyzing with 35S labeling the distribution of total Stat1 and activated Stat1, we concluded that the Stat1 molecules promptly cycle into the nucleus as tyrosine phosphorylated molecules and later return quantitatively to the cytoplasm as non-phosphorylated molecules. Therefore, the removal of the activated Stat1 molecules from the nucleus appears not to be proteolytic but must depend on a protein tyrosine phosphatase(s).

Tyrosine-phosphorylated Stat1 and Stat2 plus a 48-kDa protein all contact DNA in forming interferon-stimulated-gene factor 3.

Interferon alpha induction of transcription operates through interferon-stimulated-gene factor 3 (ISGF), a transcription factor two components of which are members of the newly characterized Stat family of transcription factors. Interferon alpha induces tyrosine phosphorylation of Stat1 and Stat2 proteins that associate and, together with a 48-kDa protein, form ISGF3. Evidence is presented that a heterodimer of Stat1 and Stat2 is present in ISGF3 and that Stat1 and the 48-kDa protein make precise contact, while Stat2 makes general contact, with the interferon-stimulated response element, the binding site of the ISGF3.

Persistence of liver-specific messenger RNA in cultured hepatocytes: different regulatory events for different genes.

Normal adult rat hepatocytes plated on rat tail collagen-coated dishes and fed a chemically defined medium supplemented with epidermal growth factor and dimethylsulfoxide (DMSO) were examined over a 40-d culture period for (a) the amount of albumin secreted; (b) steady-state albumin mRNA levels; (c) steady-state mRNA levels for six other liver-specific genes and three common genes; and (d)transcription of several liver-specific and common genes using isolated nuclei. DMSO-treated hepatocytes in culture for 40 d expressed albumin mRNA at 45% the level of normal liver and five other liver-specific genes at levels ranging from 21% to 72% of those in normal liver. The rate of synthesis of ligandin RNA using nuclei from 40-d hepatocytes in a nascent chain extension assay was 130% of the value obtained for normal liver, indicating that liverlike transcriptional activity for ligandin was maintained in this in vitro culture system. In contrast, the rates of synthesis of albumin and phosphoenolpyruvate carboxykinase (PepCK) mRNAs using nuclei from 40-d hepatocytes were 8% and less than 1%, respectively, and, therefore, were at levels that were much lower than was expected given the steady-state mRNA levels for these two genes. The discrepancy between the steady-state mRNA levels and rates of synthesis of RNA was analyzed, and the results suggest that the albumin and PepCK mRNAs from hepatocytes in culture may be more stable than those from liver. A plateau period for secretion of albumin, expression of albumin, alpha 1-antitrypsin, ligandin, phenylalanine hydroxylase, and PepCK mRNAs, and synthesis of albumin RNA using isolated nuclei was observed from days 6 to 40. The usefulness at a biological and molecular level of a hepatocyte culture system in which liver-specific genes are expressed over a long plateau period is discussed.

Transcription termination occurs within a 1000 base pair region downstream from the poly(A) site of the mouse beta-globin (major) gene.

Transcription of RNA from the beta-globin (major) gene in nuclei from induced mouse erythroleukemia cells terminates within the region between 700 to 2000 bases downstream from the poly(A) addition site, but not at particularly favored sites. In addition we present the first analysis of in vivo labeled RNA from a cell transcription unit that shows RNA termination in the same region as diagnosed by analysis of in vitro labeled RNA. The region in which termination occurs is contained in 1414 bp that were sequenced beginning 600 bases downstream from the poly(A) site. There is an increased frequency of the sequence AATAAA at the beginning and a stem and loop structure followed by a string of Ts near the end of this region.

Induced transcription of the mouse beta-globin transcription unit in erythroleukemia cells. Time-course of induction and of changes in chromatin structure.

The transcription of the beta-globin genes in mouse erythroleukemia cells has been examined by hybridizing labeled RNA obtained from isolated nuclei after chain elongation in the presence of [alpha-32P]UTP. There is induction of at least 30-fold of beta maj globin transcription after cells are treated with either dimethylsulfoxide or hexamethylene bisacetamide. The induction requires 36 to 48 hours to be maximal, during which time the cells double about three to four times. During this time, a site in the beta maj DNA region becomes hypersensitive to DNase. The development of this hypersensitive site is co-ordinate with the transcriptional increase. The induced transcripts in the beta-globin region are alpha-amanitin-sensitive (and therefore are RNA polymerase II products). An examination of weak transcriptional signals to DNA fragments upstream of the beta maj globin gene in uninduced mouse erythroleukemia cells and in cells that do not make globin is also reported. The low level of hybridization to the upstream regions in uninduced erythroleukemia cells, in L cells (a fibroblast) and in a strain of erythroleukemia cells that no longer make globin are not equally sensitive to alpha-amanitin as in the induced signal. These experiments help define the inducible transcription unit for beta maj globin mRNA production.

A precise termination site in the mouse beta major-globin transcription unit.

Nascent labeled RNA from induced, globin-producing mouse erythroleukemia cells was hybridized to cloned regions of the beta major-globin gene. Transcription ceases about 1,000 bases downstream from the poly(A) site as indicated by protection from nuclease digestion of a discrete-sized RNA fragment that it shorter than the protecting cloned DNA fragment. This defines an apparently unique termination site for a protein-coding gene that is transcribed by RNA polymerase II.

Further evidence that the majority of primary nuclear RNA transcripts in mammalian cells do not contribute to mRNA.

Nuclear RNA from Chinese hamster ovary cells was effectively separated into polyadenylic acid [poly(A)]-containing [poly (A)+] and non-poly(A)-containing [poly(A)-] fractions so that -90% of the poly(A) was present in the (A)+ fraction. Only 25% of the 5'-terminal caps of the large nuclear molecules were present in the (A)+ class, but about 70% of the specific mRNA sequences (assayed with cDNA clones) were in the (A)+ class. It appears that many long capped heterogeneous nuclear RNA molecules are of a different sequence category from those molecules that are successfully processed into mRNA.

Large heterogeneous nuclear ribonucleic acid has three times as many 5' caps as polyadenylic acid segments, and most caps do not enter polyribosomes.

The rate of synthesis in Chinese hamster cells of 5' cap structures, m7 GpppNmp, in large (greater than 700 bases) heterogeneous nuclear ribonucleic acid (RNA) molecules is two to three times faster than the synthesis of 3'-terminal polyadenylic acid segments. As judged by presence of caps, newly synthesized polysomal messenger RNA, exclusive of messenger RNA the size of histone messenger RNA, is more than 90% in the polyadenylated category. It appears, therefore, that between half and two-thirds of the long capped heterogeneous nuclear RNA molecules do not contribute a capped polysomal derivative to the cytoplasm. There are capped, nonpolysomal, non-polyadenylated molecules with a rapid turnover rate that fractionate with the cytoplasm. These metabolically unstable molecules either could represent leakage into the cytoplasm during fractionation or could truly spend a brief time in the cytoplasm before decay.

Addition of poly(A) to nuclear RNA occurs soon after RNA synthesis.

A kinetic analysis of the appearance of [3H]uridine label in RNA sequences that neighbor poly(A), as well as the incorporation of [3H]adenosine label into both the RNA chain and the poly(A) of poly(A)-containing molecules, shows that poly(A) is added within a minute or so after RNA chain synthesis in Chinese hamster ovary cells and HeLa cells. Previous conclusions by several groups (5-7) that poly(A) might be added as long as 20-30 min after RNA synthesis appear to be in error, and the present conclusion seems much more in line with several different types of recent studies with specific mRNAs that suggest prompt poly(A) addition (13-16).

Short capped hnRNA precursor chains in HeLa cells: continued synthesis in the presence of 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole.

The labeling of m7GpppN1mpN2p caps with L-[methyl-3H]methionine on short (100-500 nucleotides) heterogeneous nuclear RNA (hnRNA) chains of HeLa cells is increased 2-3 times but the labeling of caps on longer (greater than 2000 nucleotides) hnRNA chains is decreased by approximately 80% by treatment of the HeLa cells with 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB). The experimental conditions were as follows: HeLa cells were treated with 75 muM DRB for 40 min before labeling and also during the 30-min pulse of L-[methyl-3H]methionine; actinomycin D (0.05 microgram/mL) was used to suppress ribosomal RNA synthesis. Control cells received no DRB. The RNA was separated in Me2SO gradients to ensure no aggregation. Labeling of cells with [3H]uridine for 10 min and separation of RNA by these techniques reconfirmed the findings [Tamm, I., Hand, R., & Caliguiri, L. A. (1976) J. Cell Biol. 69, 229-240; Sehgal, P. B., Darnell, J. E., Jr., & Tamm, I. (1976) Cell 9, 473-480] that 70-80% of the synthesis of hnRNA (GREATER THAN 1000 NUCLEOTIDES) IS SENSITIVE TO INHIBITIOn by DRB but that 20-30% is resistant. This analysis of the methyl-labeled caps provides evidence that DRB causes early termination of a large fraction (approximately 70-80%) of hnRNA precursor chains. In contrast to the finding of continued synthesis and accumulation of short m7GpppN1mpN2p-capped chains in the presence of DRB, the synthesis of m2,2,7GpppN1mpN2mp-capped small nuclear RNAs was inhibited by approximately 70% by DRB.

The addition of 5' cap structures occurs early in hnRNA synthesis and prematurely terminated molecules are capped.

After cells were labeled by brief exposure to 3H-methyl-L-methionine, the majority of labeled 5' terminal cap I (m7GpppN1mpN2p) oligonucleotide structures were in nuclear RNA (hnRNA) molecules approximately 750 nucleotides or less in length. After longer label times, the proportion of cap I structures in nuclear molecules longer than mRNA rose to approximately 60% of the total, but approximately 40% of the cap I structures were still in molecules shorter than approximately 750 nucleotides. The cap I structures in both long and short hnRNA chains contained all four 2' methylated nucleotides in the N1 position in about the same proportion as in mRNA. None of the large hnRNA molecules could be demonstrated to contain 5' pppX p termini; the only such terminus in high molecular weight RNA was pppAp which was decreased markedly by low doses of actinomycin and is presumably the terminus of pre-rRNA. These results raise the possibilities that hnRNA chains can initiate with any of the four nucleotides, that capping occurs very close to or at the start of hnRNA chain synthesis and that approximately 40% of the hnRNA chains may be prematurely terminated.

Production of mRNA in Chinese hamster cells: relationship of the rate of synthesis to the cytoplasmic concentration of nine specific mRNA sequences.

We constructed cloned DNA sequences complementary to unselected mRNAs [poly(A)+ cytoplasmic RNA] from Chinese hamster ovary cells and used them in RNA:DNA hybridization experiments. Each cloned DNA hybridized a single mRNA from 1.3-3.5 kb in length. The relative rates of labeling (transcription rates) of nuclear RNA complementary to each individual DNA segment varied approximately 10 fold. The relative cytoplasmic concentration of the same specific RNA sequences in the mRNA after an equilibrium labeling of the cells varied approximately 100 fold. In addition, we estimated the sizes of the nuclear RNA precursor molecules to these cytoplasmic mRNAs. Four main conclusions arise from these studies. First, the primary RNA transcripts, which range in size from 2.4-13.5 kb, are 2-6 times larger than the mRNAs; second, each cloned DNA segment is complementary to only one species of mRNA; third, for the RNA complementary to at least three of the nine cloned DNA segments, the relative cytoplasmic content is considerably different from the relative rate of nuclear RNA synthesis, suggesting the post-transcriptional events are involved in the determination of the cytoplasmic concentrations of some mammalian mRNAs; and fourth, the fraction of total nonribosomal nuclear RNA complementary to the nine cloned DNA segments is in most cases 10 fold less than the fraction of cytoplasmic mRNA complementary to the same cloned DNA segments, suggesting the synthesis of many hnRNA molecules that are qualitatively different from those which eventually contribute mRNA to the cytoplasm.

Oligonucleotides in heterogeneous nuclear RNA: similarity of inverted repeats and RNA from repetitious DNA sites.

A comparison has been made by oligonucleotide analysis of three fractions of HeLa cell hnRNA: (1) the "snap-back" fraction (ds-hnRNA, 5% of the total); (2) the fraction that self-anneals during prolonged incubation (25% of total); and (3) the fraction that hybridizes most rapidly to an excess of HeLa cell DNA (rep-hnRNA, 10% of the total). T1 fingerprints of each of these hnRNA fractions were similar to one another and featured the largest T1 oligonucleotides of known sequence previously isolated from ds-hnRNA (Robertson, H.D., et al. (1977) J. Mol. Biol. 115, 571--590; Jelinek, W. (1977 J. Mol. Biol. 115, 591--602). When hybridized to DNA either in solution or immobilized on filters, the isolated ds-hnRNA and the rep-hnRNA fractions showed similar hybridization kinetics in the COt range of "intermediate" repetitive DNA sequences; the ds-hnRNA and the rep-hnRNA also self-annealed to equal extents in the absence of any DNA. DNA of all buoyant density classes contained the T1 oligonucleotides diagnostic of the ds-hnRNA and the rep-hnRNA. While hnRNA is rich in inverted repeated sequences, cytoplasmic mRNA contains far fewer such sequences.

The definition of transcription units for mRNA.

The detailed analysis of TUs with purified Ad2 DNA and the analysis of TU size for the bulk HeLa cell hnRNA compared to the size of mRNA of infected and uninfected cells supports the conclusion that mRNA in mammalian cells is generally derived by the processing of primary transcripts. In the context of this volume, these results indicate that the RNA transcription products of chromatin which are related to mRNA are longer than the mRNA itself. Proper functioning of chromatin in vitro must eventually take these results into account.

Adenovirus type 2 mRNA in transformed cells: map positions and difference in transport time.

Adenovirus type 2 rat transformed cells produced two polyadenylic acid-terminated mRNA's with approximate coordinates 1.5-4.4 and 4.4-11.0 on the physical map of the adenovirus type 2 genome. These mRNA's were also formed early during lytic infection in addition to one or more smaller mRNA's from the 4.4-11.0 region. In transformed cells, the 1.5-4.4 mRNA appeared in the cell cytoplasm without detectable lag, whereas the 4.4-11.0 mRNA required at least 20 to 30 min for the maximal rate of accumulation.

The methylation of adenovirus-specific nuclear and cytoplasmic RNA.

Each poly(A) containing cytoplasmic AD-2 MRNA contains at its 5' terminus the general structure m7 GpppN1 pN2p or m7 GpppN1mpN2mpNp as well as an average of 4 m6A and 0.5-1 m5C residues per molecule. Almost all of the N1m residues are adenine derivatives including Am, m6Am and probably m26,6Am. The N2m is mostly Cm but small amounts of the other three methylated bases are also present. All the methylated constitutents of mRNA are distant from the 3' terminal poly(A). The amount of m6A appears to be greater in larger mRNA than in smaller mRNA. Nuclear Ad-2 specific RNA also contains caps, m6A, and m5C with about twice as much m6A relative to caps as cytoplasmic mRNA. The similarity of Ad-2 nuclear and mRNA to HeLa hnRNA and mRNA suggests that adenovirus mRNA production is a good model for eukaryotic mRNA production.

Methyl labeling of HeLa cell hnRNA: a comparison with mRNA.

The majority of the mRNA molecules in HeLa cells contain 1-2 residue(s) of m6Ap and one blocked, methylated 5' terminal "cap" structure. The hnRNA, which is longer than mRNA, contains both m6Ap and caps but 4-6 times as many m6Ap residues per chain. In addition, nuclear molecules contain T2 RNA ase-resistant, methyl-labeled oligonucleotides ("di-" and "tri-" nucleotides) which are not found in mRNA. Some of the dinucleotides may be precursors to the 2'-0-methylated nucleotides in the cap structures. These results are compatible with internal methylation of hnRNA molecules (both m6Ap and 2'-0-methyl) followed by hnRNA cleavage and the addition of the cap structure to generate at least some of the HeLa cell mRNA. It also appears that some hnRNA molecules, which are longer than most mRNA molecules, contain cap structures suggesting the derivation of some mRNA molecules from the 5' regions of hnRNA.

Methylated, blocked 5 termini in HeLa cell mRNA.

Poly(A)-containing HeLa cell mRNA prepared from cells labeled with [methyl-(3)H]methionine or [(32)P]phosphate was found to contain a variety of methylated, blocked 5'-terminal structures of two general types: m(7)GpppN(7)-Np and m(7)GpppN(m)-N(m)-Np. In addition, about one-third of the [(3)H]methyl label was present in the N(6)-methyladenosine; this labeled nucleoside was not found in the 3'-terminal one-third of the mRNA chain and thus may also be in the 5' portion of the mRNA.