Amino acids of Epstein-Barr virus nuclear antigen 3A essential for repression of Jkappa-mediated transcription and their evolutionary conservation.

Epstein-Barr virus (EBV) nuclear antigen 3A (EBNA-3A) is essential for virus-mediated immortalization of B lymphocytes in vitro and is believed to regulate transcription of cellular and/or viral genes. One known mechanism of regulation is through its interaction with the cellular transcription factor Jkappa. This interaction downregulates transcription mediated by EBNA-2 and Jkappa. To identify the amino acids that play a role in this interaction, we have generated mutant EBNA-3A proteins. A mutant EBNA-3A protein in which alanine residues were substituted for amino acids 199, 200, and 202 no longer downregulated transcription. Surprisingly, this mutant protein remained able to coimmunoprecipitate with Jkappa. Using a reporter gene assay based on the recruitment of Jkappa by various regions spanning EBNA-3A, we have shown that this mutation abolished binding of Jkappa to the N-proximal region (amino acids 125 to 222) and that no other region of EBNA-3A alone was sufficient to mediate an association with Jkappa. To determine the biological significance of the interaction of EBNA-3A with Jkappa, we have studied its conservation in the simian lymphocryptovirus herpesvirus papio (HVP) by cloning HVP-3A, the homolog of EBNA-3A encoded by this virus. This 903-amino-acid protein exhibited 37% identity with its EBV counterpart, mainly within the amino-terminal half. HVP-3A also interacted with Jkappa through a region located between amino acids 127 and 223 and also repressed transcription mediated through EBNA-2 and Jkappa. The evolutionary conservation of this function, in proteins that have otherwise significantly diverged, argues strongly for an important biological role in virus-mediated immortalization of B lymphocytes.

Epstein-barr virus nuclear antigen 3C activates the latent membrane protein 1 promoter in the presence of Epstein-Barr virus nuclear antigen 2 through sequences encompassing an spi-1/Spi-B binding site.

The Epstein-Barr virus (EBV) nuclear antigen 3C (EBNA-3C) protein is a transcriptional regulator of viral and cellular genes that is essential for EBV-mediated immortalization of B lymphocytes in vitro. EBNA-3C can inhibit transcription through an association with the cellular DNA-binding protein Jkappa, a function shared by EBNA-3A and EBNA-3B. Here, we report a mechanism by which EBNA-3C can activate transcription from the EBV latent membrane protein 1 (LMP-1) promoter in conjunction with EBNA-2. Jkappa DNA-binding sites were not required for this activation, and a mutant EBNA-3C protein unable to bind Jkappa activated transcription as efficiently as wild-type EBNA-3C, indicating that EBNA-3C can regulate transcription through a mechanism that is independent of Jkappa. Furthermore, activation of the LMP-1 promoter is a unique function of EBNA-3C, not shared by EBNA-3A and EBNA-3B. The DNA element through which EBNA-3C activates the LMP-1 promoter includes a Spi-1/Spi-B binding site, previously characterized as an important EBNA-2 response element. Although this element has considerable homology to mouse immunoglobulin light chain promoter sequences to which the mouse homologue of Spi-1 binds with its dimerization partner IRF4, we demonstrate that the IRF4-like binding sites in the LMP-1 promoter do not play a role in EBNA-3C-mediated activation. Both EBNA-2 and EBNA-3C were required for transcription mediated through a 41-bp region of the LMP-1 promoter encompassing the Spi binding site. However, EBNA-3C had no effect on transcription mediated in conjunction with the EBNA-2 activation domain fused to the GAL4 DNA-binding domain, suggesting that it does not function as an adapter between EBNA-2 and the cellular transcriptional machinery. Like EBNA-2, EBNA-3C bound directly to both Spi-1 and Spi-B in vitro. This interaction was mediated by a region of EBNA-3C encompassing a likely basic leucine zipper (bZIP) domain and the ets domain of Spi-1 or Spi-B, reminiscent of interactions between bZIP and ets domains of other transcription factors that result in their targeting to DNA. There are many examples of regulation of the hematopoietic-specific Spi transcription factors through protein-protein interactions, and a similar regulation by EBNA-3C, in conjunction with EBNA-2, is likely to be an important and unique contribution of EBNA-3C to EBV-mediated immortalization.

Nuclear bodies of stage 6 oocytes of Rana temporaria contain nucleolar and coiled body proteins.

The genetically inactive stage 6 oocyte nuclei of Rana temporaria contain certain nuclear bodies that label with nucleolus-specific and coiled body (CB)-specific antibodies. We designate them multicomponent bodies (MCBs) to reflect their mixed composition. Morphologically, each MCB contains five distinct zones: zone I composed of electron-dense fibrils similar to the dense fibrillar component (DFC) of the typical eukaryotic nucleoli; zone II resembled the fibrillar material of the inactive agranular nucleoli of stage 6 oocytes; zone III consisted of fine filamentous material corresponding to the fibrillar center (FC) of lower electron density seen in the typical nucleoli; and zones IV and V contained packed coiled threads typical of CBs. Of these, zone IV was seen in the interior of MCBs and contained tightly packed coiled threads (20 nm thick), while zone V occurred at the periphery and consisted of similar threads but loosely packed and electron dense. The material of both zones IV and V resembled that of CBs. To determine the composition of these zones, we extracted oocytes with a buffer that removes chromatin and most of the soluble proteins and processed them for immunogold labeling with a variety of antibodies. Anti-p80 coilin antibody predominantly labeled zone IV and, to a lesser extent, zone V. Anti-snRNP antibody also showed a similar labeling pattern. Anti-fibrillarin antibody predominantly labeled zone I and to a lesser extent zones IV and V. Anti-B23 antibody labeled all zones. These observations suggest that MCBs contain both nucleolar and CB material. We postulate that MCBs represent storage structures which provide material needed for the early stage of embryogenesis. The demonstration of MCBs further supports the close interrelationship between nucleoli and CBs.

A conserved domain of the Epstein-Barr virus nuclear antigens 3A and 3C binds to a discrete domain of Jkappa.

EBNA-3C can affect the LMP-1 promoter in both a positive and a negative manner through distinct DNA sequence elements. The viral transactivator EBNA-2 normally binds DNA indirectly via Jkappa to activate transcription, but this activation is prevented in the presence of EBNA-3C. The DNA element recognized by Jkappa is both required and sufficient for this inhibition. Jkappa clones isolated in a yeast two-hybrid screen using EBNA-3C as bait allowed us to delineate the sequences of both proteins mediating the interaction. Two isoforms of Jkappa that differ in exon 1, Jkappa-1 and RBP-2N, interact with EBNA-3C, suggesting that exon 1 is not required for this interaction; indeed, clones with deletion of the N-terminal third of Jkappa interacted as efficiently with EBNA-3C as full-length Jkappa clones. A Jkappa domain as small as 56 amino acids was sufficient to bind to EBNA-3C. A 74-amino-acid domain of EBNA-3C, conserved in all three EBNA-3 family members, was sufficient to interact with Jkappa. A specific mutation in this conserved domain suppressed the ability of EBNA-3C to downregulate transcription. Accordingly, EBNA-3A was also able to interact with Jkappa and downregulate Jkappa-mediated transcription as efficiently as EBNA-3C. The ability of the EBNA-3 proteins to prevent Jkappa from binding to DNA in vitro and suppress transactivation via Jkappa DNA elements suggests that the EBNA-3 proteins act analogously to the Drosophila protein Hairless.

Epstein-Barr Virus DNA recombination and loss in sporadic Burkitt's lymphoma.

Epstein-Barr virus (EBV) antigens in tumor tissue define associations of virus with human malignancies and provide clues as to mechanisms of viral oncogenesis. In Burkitt's lymphoma, EBV markers are absent from 85% of sporadic cases and 4% of endemic (African) cases, raising questions as to the exact role EBV in the disease. Standard screening criteria may be insufficient to determine the EBV status of all tumors. One of 9 tumors from American patients expressed EBV nuclear antigen 1 (EBNA1) and contained standard episomal EBV DNA, making this series consistent with the 15% EBV association traditionally ascribed to sporadic Burkitt's lymphoma. Surprisingly, 3 tumors without detectable EBNA1 contained partial EBV genomes. Identification of defective, integrated viral DNA in some tumors indicates greater involvement of virus in sporadic Burkitt's lymphoma than previously documented and suggests a process of viral DNA rearrangement and loss during malignant progression most consistent with an initiating role for EBV in tumorigenesis.

Nuclear actin filaments and their topological changes in frog oocytes.

Previous morphological and biochemical studies have suggested that actin and actin-containing filaments (microfilaments) exist in the eukaryotic nucleus and that they perform important nuclear functions. However, the concept is not widely accepted. In this study, we demonstrate actin and bundles of actin in the nuclei of oocytes of Rana temporaria by immunoblotting and immunogold labeling/electron microscopy. The system and methods used here provided nuclei, free from cytoplasmic contamination. Additionally, we have compared the topological distribution of intranuclear actin filaments in two structurally and functionally distinct stages (stages 3 and 6) of oogenesis. The stage 3 nuclei are extremely active in rRNA transcription and contain multiple nucleoli located at the periphery with the central part occupied by the lampbrush chromosomes. The stage 6 nuclei are transcriptionally inert and contain both nucleoli and chromosomes confined to a small area in the central part. The nuclear lysates derived from the manually isolated stage 3 and 6 nuclei and the nuclear contents obtained by manually removing the nuclear envelope of stage 6 nucleus both contained actin as demonstrated by immunoblotting with an actin-specific monoclonal antibody. When examined by immunogold electron microscopy using the anti-actin antibody, the stage 3 oocyte nuclei showed distinct intranuclear tracks composed of bundles of actin that extended from the nucleoli and chromosomes to the nuclear envelope. The stage 6 oocyte nuclei, on the other hand, showed short stretches of actin bundles in the central part mainly in association with the nucleoli; none of these bundles extended to the nuclear envelope. Taken together, the above results suggest that actin is a structural component of the oocyte nucleus and that polymerized actin undergoes dramatic topological changes correlated with changes in the distribution of nuclear components and their function.

Radiation inactivation analysis of rat liver microsomal glucose-6-phosphatase.

Radiation inactivation analysis was utilized to estimate the sizes of the units catalyzing the various activities of hepatic microsomal glucose-6-phosphatase. This technique revealed that the target molecular weights for mannose-6-P phosphohydrolase, glucose-6-P phosphohydrolase, and carbamyl-P:glucose phosphotransferase activities were all about Mr 75,000. These results are consistent with the widely held view that all of these activities are catalyzed by the same protein or proteins. Certain observations indicate that the molecular organization of microsomal glucose-6-phosphatase is better described by the conformational hypothesis which envisions the enzyme as a single covalent structure rather than by the substrate transport model which requires the participation of several physically separate polypeptides. These include the findings: 1) that the target sizes for glucose-6-P phosphohydrolase and carbamyl-P:glucose phosphotransferase activities were not larger than that for mannose-6-P phosphohydrolase in intact microsomes and 2) that the target size for glucose-6-P phosphohydrolase in disrupted microsomes was not less than that observed in intact microsomes. These findings are most consistent with a model for glucose-6-phosphatase of a single polypeptide or a disulfide-linked dimer which spans the endoplasmic reticulum with the various activities of this multifunctional enzyme residing in distinct protein domains.

Tissue-specific binding of a nuclear factor to the insulin gene promoter.

Using a DNA-binding gel mobility shift assay, a protein-DNA complex was detected using the rat II insulin promoter and a nuclear extract from rat insulinoma cells. This complex was detected using extracts from other insulin-producing cell lines but was not found using nuclear extracts from non-insulin-producing pancreatic cell lines or other cell types. Binding of the tissue-specific protein to the labelled promoter fragment was effectively competed by both rat and human insulin genes but not by the SV40 TATA box. This protein DNA interaction may be involved in tissue-specific regulation of insulin gene expression.

Regulation of 3-hydroxy-3-methylglutaryl coenzyme A reductase mRNA levels by L-triiodothyronine.

In hypophysectomized rats, hepatic 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase activity, immunoreactive 97-kilodalton (97-kDa) protein, and mRNA were all reduced to undetectable levels. Administration of triiodothyronine (T3) resulted in large increases in all three after a 36-h lag period. HMG-CoA reductase activity, immunoreactive 97-kDa protein levels, and reductase mRNA levels were tightly correlated. Feeding hypophysectomized rats diets containing the bile acid sequestrant colestipol, together with the potent reductase inhibitor mevinolin, resulted in an increase in HMG-CoA reductase activity similar to that seen with T3 but a lesser stimulation of reductase mRNA levels. These results suggest that agents which cause depletion of mevalonate-derived products may share in part with T3 a common mechanism for increasing levels of HMG-CoA reductase activity in order to satisfy cellular needs for these products. Dexamethasone treatment, which is known to prevent the T3-mediated stimulation of reductase activity, caused a marked decrease in 97-kDa immunoreactive material but had little effect on reductase mRNA levels.

Regulation of the activity of 3-hydroxy-3-methylglutaryl coenzyme A reductase by insulin.

In hypophysectomized--diabetic rats full restoration of hepatic HMG-CoA reductase activity required administration of both triiodothyronine and insulin. Giving triiodothyronine alone resulted in full restoration of reductase mRNA and immunoreactive protein but not of enzyme activity. Increasing the ratio of dithiothreitol to microsomal protein resulted in full HMG-CoA reductase activity. With 5mM glutathione, HMG-CoA reductase activity was 5 to 10 fold higher in non-diabetic animals. The data suggest that insulin acts by increasing the portion of HMG-CoA reductase present in its active free sulfhydryl form.

Characteristics of rat liver microsomal 3-hydroxy-3-methylglutaryl-coenzyme A reductase.

A procedure for the preparation of rat liver microsomal fractions essentially devoid of contaminating lysosomes is described. When this preparation was examined by immunoblotting with a rabbit antiserum to rat 3-hydroxy-3-methylglutaryl-CoA reductase, a single band corresponding to an Mr of 100000 was observed. No evidence was found for glycosylation of rat liver-3-hydroxy-3-methylglutaryl-CoA reductase. Native rat liver microsomal 3-hydroxy-3-methylglutaryl-CoA reductase differs from the purified proteolytically modified species in that it displays allosteric kinetics towards NADPH.

Sulfhydryl/disulfide forms of rat liver 3-hydroxy-3-methylglutaryl coenzyme A reductase.

Using radiation inactivation and immunoblotting techniques, evidence for functionally active forms of rat liver 3-hydroxy-3-methylglutaryl coenzyme A reductase with molecular weights of about 100,000 and 200,000 was obtained. In liver microsomes isolated from rats fed both mevinolin and colestipol, the Mr 100,000 form was the predominant species, whereas in microsomes from animals fed only colestipol, the Mr 200,000 species was the major form. This Mr 200,000 form could be converted to the Mr 100,000 form by addition of dithiothreitol or beta-mercaptoethanol. Although both forms appear to possess catalytic activity, the Mr 200,000 species displays sigmoidal kinetics with respect to the concentration of NADPH, whereas the Mr 100,000 form exhibits typical hyperbolic kinetics.