This report describes the identification of a transcript that contains a 135-amino-acid open reading frame, ORF-E. The ORF-E transcript is antisense to the LR gene and was consistently expressed during latency. Although we were unable to detect the ORF-E transcript by Northern analysis, the ORF-E transcript was readily detected by RT-PCR with strand-specific primers used to initiate cDNA synthesis. This is similar to the herpes simplex virus type 1 AL gene and the UL43 gene, which were not detected by Northern blotting but were detected by RT-PCR (5
). When ORF-E was fused to GFP, this fusion protein was expressed in the nucleus of transfected neuronal cells.
The herpes simplex virus type 1 AL transcript (antisense to LAT) contains an ORF that is expressed upstream and antisense of herpes simplex virus type 1 LAT (22
). The AL transcript overlaps 198 nucleotides of the LAT promoter and 158 nucleotides of the 5′ end of the primary 8.3-kb LAT transcript. The sequence of the AL ORF is highly conserved among different herpes simplex virus type 1 isolates that have been sequenced, suggesting that this protein is expressed. The finding that sera from infected rabbits recognize peptides from the AL ORF (22
) supports the prediction that this protein is expressed. Although the AL gene and the ORF-E gene occupy similar locations on their respective genomes, there is no obvious amino acid sequence similarity between the putative ORF-E protein and the AL protein (data not shown). Among the BHV-1 isolates that have been sequenced, the ORF-E amino acid sequences are highly conserved. In contrast, an ORF-E gene was not detected at the same location on the BHV-5 genome (9
). Although the genomes of BHV-1 and BHV-5 are very similar, BHV-5 causes severe central nervous disorders in cattle, whereas BHV-1 primarily induces upper respiratory tract disorders and conjunctivitis. The most dramatic difference in the genomes of BHV-1 and BHV-5 was the LR gene locus, including the ORF-E coding sequences. Since it was hypothesized that differences in the biological properties of BHV-1 and BHV-5 are linked to differences in the LR gene locus (9
), it is logical to speculate that ORF-E may contribute to these differences.
The 5′ terminus of the ORF-E transcript was localized to sequences that were adjacent to the 3′ terminus of bICP0 (Fig. ). The putative ATG for ORF-E is more than 200 bp downstream from this region. This result was unexpected because the AT-rich motif that is present within the LR promoter (Fig. ) was downstream of the 5′ terminus of the ORF-E transcript, suggesting that this was the ORF-E promoter. However, our results suggested that ORF-E promoter sequences are located within the bICP0 coding sequences. The finding that more than one 3′ end exists for the ORF-E transcript was also not expected. We suggest that virus-encoded and induced functions and neuron-specific factors regulate the selection of polyadenylation sites. It is also possible that a family of ORF-E transcripts are synthesized, which would have complicated mapping the 5′ and 3′ termini. This possibility is supported by the fact that the LR gene encodes a family of transcripts, some of which are alternatively spliced (10
). Since ORF-E and the LR gene are both consistently expressed during latency, we suggest that these genes regulate each other's expression and/or work in conjunction to promote latency. Studies to address what role the ORF-E gene plays in the latency reactivation cycle of BHV-1 are in progress.
We predict that ORF-E encodes a regulatory protein in neurons because the ORF-E/GFP fusion protein was primarily detected in the nucleus of transfected neuronal cells (Fig. ). The cells transfected with ORF-E/GFP continued to grow (data not shown). In contrast, the LR gene induced cell cycle arrest (25
), and the bICP0 gene is cytotoxic (16
). Although we do not know the function of the ORF-E gene, it seems clear that it has novel functions relative to the LR gene and bICP0. The LR gene has antiapoptotic activity (7
). In contrast, our studies indicate that the ORF-E gene does not protect cells against an apoptotic insult in vitro (data not shown). Sequence analysis of the putative protein encoded by ORF-E suggested that it can be myristoylated, glycosylated, and phosphorylated. Amino acid sequences near the N terminus of ORF-E can be found in other proteins that have anti-inflammatory properties. Infiltration of immune cells into TG plays an active role in maintaining latency (19
), suggesting that the ORF-E protein may dampen the effects of infiltrating immune cells on infected neurons. We recently expressed the ORF-E protein in a baculovirus system, and thus we should be able to generate antibodies that will be useful for detecting ORF-E protein expression after infection.