The upstream promoter region of the Autographa californica multicapsid nuclear polyhedrosis virus (AcMNPV) gp64 gene contains five copies of TAAG, the conserved sequence found at the transcriptional initiation sites of almost all baculovirus late genes. In AcMNPV-infected Sf9 cells, late transcription initiation is detected from only two upstream TAAG sites and not from three downstream TAAG sites. To examine several models for preferential TAAG site utilization, we constructed a series of recombinant AcMNPV baculoviruses that contain promoter region sequences from the gp64 gene fused to a chloramphenicol acetyl transferase reporter gene. Promoter-reporter constructs were inserted in the polyhedrin locus. To test a scanning model in which TAAG sites are sequentially selected according to their location in the region, we generated recombinant viruses in which the highly transcribed sites were inactivated by point mutations. Transcription from the mutant promoter constructs was compared qualitatively and quantitatively to transcription from the wild-type gp64 promoter. Inactivation of the upstream TAAG sites did not result in increased transcription from the downstream TAAG sites, suggesting that immediate context, rather than position, determines promoter utilization. To test this hypothesis, we made a series of minimal promoter constructs containing decreasing quantities of the sequences immediately flanking one of the active gp64 TAAG sites. Reporter constructs containing a gp64 TAAG site and > or = 12 bp of flanking sequence on both sides were transcribed at near wild-type levels. Constructs with less flanking sequence (9 or 6 bp of flanking sequence) were accurately transcribed, but at substantially lower levels, and transcription was not detected from constructs containing only 3 bp of flanking sequence. These results suggest that nucleotides immediately flanking the TAAG site (4-6 bp) are necessary for basal promoter activity while additional flanking sequences (> or = 12 bp) are required for late promoter activation and regulation. To further examine late promoter selection, we constructed recombinant AcMNPV baculoviruses that contain heterologous late promoters from the gp64 gene of the related virus Orgyia pseudotsugata MNPV (OpMNPV). TAAG sites that serve as functional late promoters in OpMNPV were found to mediate transcription initiation at only basal levels in the context of the AcMNPV genome, suggesting that late promoter activation may be virus specific within the family Baculoviridae.

In a previous study of translational regulation of a baculovirus gene, we observed that translation initiated at an unexpectedly high efficiency from an AUG codon found in what was believed to be a poor context (M.-J. Chang and G. W. Blissard, 1997, J. Virol. 71, 7448-7460). In the current study, we examined the roles of nucleotides flanking a baculovirus AUG initiator codon in modulating translation initiation in lepidopteran insect cells. The roles of nucleotides flanking the AcMNPV gp64 initiator codon were examined by site-directed mutagenesis and functional assays in transfected Sf9 cells. To eliminate potential cis-acting sequences and effects, the gp64 initiator context was cloned in-frame with a chloramphenicol acetyl transferase reporter gene and under the control of a heterologous promoter. All possible single-nucleotide substitutions were generated in positions -6 to -1 and +4 to +6, relative to the A of the initiator AUG codon, which was designated +1. Constructs were transfected into lepidopteran cells and translation products were quantified by an enzyme-linked immunosorbent assay procedure. Substitutions of pyrimidines or other nucleotides at the -3 position resulted in little or no detectable effect on translation efficiency. In contrast, specific substitutions at the +4 and +5 positions resulted in approximately 2- to 3-fold increases in translation. Substitution of A in the +4 position resulted in an approximately 3-fold increase in translation, and substitution of any nucleotide for T in the +5 position resulted in approximately 1.9- to 2.8-fold increases. Substitutions at other positions (-6 to -1 and +6) resulted in no detectable increase or decrease in translation efficiency. These experimental results suggest an optimal initiator context of 5'-N N N N N N A U G A a/c/g N-3' for efficient translation initiation in lepidopteran cells. Consensus translation initiation contexts were generated from baculovirus genes and lepidopteran genes, then compared with the experimental results from the gp64 initiator context.

Small upstream open reading frames (ORFs) or minicistrons located in the 5 leader of eukaryotic mRNAs have been shown to play a role in translational regulation of some eukaryotic genes, particularly mammalian proto-oncogenes. A survey of the baculovirus Autographa californica multicapsid nuclear polyhedrosis virus genome suggests that at least 10 transcripts from late genes contain potential minicistrons, and at least three of these minicistrons appear to be conserved in homologous genes of the related Orygia pseudotsugata MNPV. The position of the minicistron from one of these genes, gp64, is also conserved in gp64 genes from several baculoviruses, suggesting a potential regulatory function. To identify the potential role of the gp64 minicistron in regulating translation from gp64 late mRNAs, we generated a series of recombinant viruses containing the gp64 promoter and minicistron in combination with a chloramphenicol acetyltransferase reporter gene (cat) inserted into the polyhedrin locus. We first fused a cat reporter in frame with the minicistron coding region to demonstrate that the minicistron initiator ATG was in a context suitable for translational initiation. In subsequent experiments, a cat reporter was fused in frame to the downstream gp64 ORF, and various constructs containing point mutations that inactivated the minicistron were examined. Translational efficiency in the presence and absence of the minicistron was measured by quantitative analysis of gp64-cat RNA and the GP64-CAT protein. In the absence of a functional minicistron, translational efficiency from the downstream gp64-cat reporter ORF increased. Surprisingly, single-point mutations that inactivated the minicistron initiator ATG also resulted in utilization of an upstream in-frame ATG that is found within the minicistron coding region and that is in a poor translational initiation context. Double-point mutation constructs that inactivated both the minicistron initiator ATG and the upstream in-frame ATG also resulted in increased translational efficiency from the downstream gp64-cat ORF. Thus, the gp64 minicistron serves as a negative regulatory element that decreases translation of the gp64 ORF on late mRNAs.