Factors regulating baculovirus late and very late gene expression in transient-expression assays.

Eighteen genes of Autographa californica nuclear polyhedrosis virus are necessary and sufficient to transactivate expression from the late vp39 promoter in transient-expression assays in SF-21 cells. These 18 genes, known as late expression factor genes (lefs), are also required to transactivate the very late promoter of the polyhedrin gene, polh, but expression from this promoter is relatively weak compared with expression from the vp39 promoter. To further define the factors required for late and very late promoter expression, we first determined that the eighteen lefs were also required for expression from two other major baculovirus promoters: the late basic 6.9-kDa protein gene, p6.9, and the very late 10-kDa protein gene, p10. We next examined the effect of the very late expression factor 1 gene (vlf-1), a gene previously identified by analysis of a temperature-sensitive mutant, in the transient expression assay and found that vlf-1 specifically transactivated the two very late promoters but not the two late promoters. We then surveyed the Autographa californica nuclear polyhedrosis virus genome for additional genes which might specifically regulate very late gene expression; no additional vlf genes were detected, suggesting that VLF-1 is the primary regulator of very late gene expression. Finally, we found that the relative contribution of the antiapoptosis gene p35, which behaves as a lef in these transient-expression assays, depended on the nature of the other viral genes provided in the cotransfection mixtures, suggesting that other viral genes also contribute to the ability of the virus to block apoptosis.     

Binding of the baculovirus very late expression factor 1 (VLF-1) to different DNA structures

Background

Baculovirus genomes encode a gene called very late expression factor 1 (VLF-1) that is a member of the integrase (Int) family of proteins. In this report we describe the binding properties of purified Autographa californica multiple capsid nucleopolyhedrovirus (AcMNPV) VLF-1 to a number of different DNA structures including homologous regions. In addition, its enzymatic activity was examined.

Results

VLF-1 was expressed in a recombinant baculovirus as a fusion with both HA and HIS₆ tags and its binding activity to different DNA structures was tested. No binding was evident to single and double strand structures, very low binding was observed to Y-forks, more binding was observed to three-way junctions, whereas cruciform structures showed high levels of binding. VLF-1 binding was affected by divalent cations; optimal binding to three-way junctions and cruciforms was 2 and 0 mM MgCl₂, respectively. Homologous region (hr) sequences was also examined including oligomers designed to expose the hr palindrome as a hairpin, linear double strand, or H-shaped structure. Efficient binding was observed to the hairpin and H-shaped structure. No topoisomerase or endonuclease activity was detected. Sedimentation analysis indicated that *VLF-1 is present as a monomer.

Conclusions

An HA- and HIS-tagged version of AcMNPV VLF-1 showed structure-dependent binding to DNA substrates with the highest binding affinity to cruciform DNA. These results are consistent with the involvement of VLF-1 in the processing of branched DNA molecules at the late stages of viral genome replication. We were unable to detect enzymatic activity associated with these complexes.

     

Characterization of the role of very late expression factor 1 in baculovirus capsid structure and DNA processing

Very late expression factor 1 (VLF-1) of Autographa californica multiple nucleopolyhedrovirus is a putative tyrosine recombinase and is required for both very late gene expression and budded virus production. In this report, we show that a vlf-1 knockout bacmid was able to synthesize viral DNA at levels similar to that detected for a gp64 knockout bacmid that served as a noninfectious control virus. Additionally, analysis of replicated bacmid DNA by field-inversion gel electrophoresis indicated that VLF-1 is not required for synthesizing high-molecular-weight intermediates that could be resolved into unit-length genomes when cut at a unique restriction site. However, immunoelectron microscopic analysis revealed that in cells transfected with a vlf-1 knockout bacmid, aberrant tubular structures containing the capsid protein vp39 were observed, suggesting that this virus construct was defective in producing mature capsids. In contrast, rescuing the vlf-1 knockout bacmid construct with a copy of VLF-1 that carries a mutation of a highly conserved tyrosine (Y355F) was sufficient to restore the production of nucleocapsids with a normal appearance, but not infectious virus production. Furthermore, the results of a DNase I protection assay indicated that the DNA packaging efficiency of the VLF-1(Y355F) virus construct was similar to that of the gp64 knockout control. Finally, a recombinant virus containing a functional hemagglutinin epitope-tagged version of VLF-1 was constructed to investigate the association of VLF-1 with the nucleocapsid. Analysis by immunoelectron microscopy of Sf-9 cells infected with this virus showed that VLF-1 localized to an end region of the nucleocapsid. Collectively, these results indicate that VLF-1 is required for normal capsid assembly and serves an essential function during the final stages of the DNA packaging process.     

Studies of Schizosaccharomyces pombe DNA polymerase alpha at different stages of the cell cycle.

The status of Schizosaccharomyces pombe (fission yeast) DNA polymerase alpha was investigated at different stages of the cell cycle. S.pombe DNA polymerase alpha is a phosphoprotein, with serine being the exclusive phosphoamino acid. By in vivo pulse labeling experiments DNA polymerase alpha was found to be phosphorylated to a 3-fold higher level in late S phase cells compared with cells in the G2 and M phases, but the steady-state level of phosphorylation did not vary significantly during the cell cycle. Tryptic phosphopeptide mapping demonstrated that the phosphorylation sites of DNA polymerase alpha from late S phase cells were not the same as that from G2/M phase cells. DNA polymerase alpha partially purified from G1/S cells had a different mobility in native gels from that from G2/M phase cells. The partially purified polymerase alpha from G1/S phase cells had a higher affinity for single-stranded DNA than that from G2/M phase cells. Despite the apparent differences in cell cycle-dependent phosphorylation, mobility in native gels and affinity for DNA, the in vitro enzymatic activity of the partially purified DNA polymerase alpha did not appear to vary during the cell cycle. The possible biological significance of these cell cycle-dependent characteristics of DNA polymerase alpha is discussed.     

Wild-type but not mutant p53 immunopurified proteins bind to sequences adjacent to the SV40 origin of replication.

The DNA from a wide variety of human tumors has sustained mutations within the conserved p53 coding regions. We have purified wild-type and tumor-derived mutant p53 proteins expressed from baculovirus vectors and examined their interactions with SV40 DNA. Using DNAase I footprinting assays, we observed that both human and murine wild-type p53 proteins bind specifically to sequences adjacent to the late border of the viral replication origin. By contrast, mutant p53 proteins failed to bind specifically to these sequences. SV40 T antigen prevented wild-type p53 from interacting with this region. These data show that normal but not oncogenic forms of p53 are capable of sequence-specific interactions with viral DNA. Furthermore, they provide insights into the mechanisms by which viral proteins might regulate the control of viral growth and cell division.     

A sequence-specific DNA-binding factor (VF1) from Anabaena sp. strain PCC 7120 vegetative cells binds to three adjacent sites in the xisA upstream region.

A DNA-binding factor (VF1) partially purified from Anabaena sp. strain PCC 7120 vegetative cell extracts by heparin-Sepharose chromatography was found to have affinity for the xisA upstream region. The xisA gene is required for excision of an 11-kilobase element from the nifD gene during heterocyst differentiation. Previous studies of the xisA upstream sequences demonstrated that deletion of this region is required for the expression of xisA from heterologous promoters in vegetative cells. Mobility shift assays with a labeled 250-base-pair fragment containing the binding sites revealed three distinct DNA-protein complexes. Competition experiments showed that VF1 also bound to the upstream sequences of the rbcL and glnA genes, but the rbcL and glnA fragments showed only single complexes in mobility shift assays. The upstream region of the nifH gene formed a weak complex with VF1. DNase footprinting and deletion analysis of the xisA binding site mapped the binding to a 66-base-pair region containing three repeats of the consensus recognition sequence ACATT.     

Site-specific interaction of the murine pre-replicative complex with origin DNA: assembly and disassembly during cell cycle transit and differentiation

Eukaryotic DNA replication initiates at origins of replication by the assembly of the highly conserved pre-replicative complex (pre-RC). However, exact sequences for pre-RC binding still remain unknown. By chromatin immunoprecipitation we identified in vivo a pre-RC-binding site within the origin of bidirectional replication in the murine rDNA locus. At this sequence, ORC1, -2, -4 and -5 are bound in G1 phase and at the G1/S transition. During S phase, ORC1 is released. An ATP-dependent and site-specific assembly of the pre-RC at origin DNA was demonstrated in vitro using partially purified murine pre-RC proteins in electrophoretic mobility shift assays. By deletion experiments the sequence required for pre-RC binding was confined to 119 bp. Nucleotide substitutions revealed that two 9 bp sequence elements, CTCGGGAGA, are essential for the binding of pre-RC proteins to origin DNA within the murine rDNA locus. During myogenic differentiation of C2C12 cells, we demonstrated a reduction of ORC1 and ORC2 by immunoblot analyses. ChIP analyses revealed that ORC1 completely disappears from chromatin of terminally differentiated myotubes, whereas ORC2, -4 and -5 still remain associated.