Redefining the genetics of murine gammaherpesvirus 68 via transcriptome-based annotation

Abstract

Viral genetic studies typically focus on large open reading frames (ORFs) identified during genome annotation (ORF-based annotation). Here we describe tools for examining viral gene expression nucleotide by nucleotide across the genome. Using these tools on the 119,450 base pair (bp) genome of murine gammaherpesvirus 68 (gammaHV68) allowed us to establish that gammaHV68 RNA expression was significantly more complex than predicted from ORF-based annotation, including over 73,000 nucleotides of unexpected transcription within 30 expressed genomic regions (EGRs). Approximately 90% of this RNA expression was antisense to genomic regions containing known large ORFs. We verified the existence of previously undefined transcripts in three EGRs and determined which parts of the transcriptome depend on protein or viral DNA synthesis. This study redefines the genetic map of gammaHV68, indicating that herpesviruses contain significantly more genetic complexity than predicted from ORF-based genome annotations, and provides alternative tools and approaches for viral genetic studies.

Figure 1. Genome-wide RNA expression in γHV68

3T12 cells were infected with γHV68 at an MOI=10. Total RNA was harvested at 18 hpi, labeled, and hybridized to the γHV68 tiled array. The mean fluorescence of probes overlapping each nucleotide position is plotted on a log₂ scale underneath the annotated gene positions. The three colored lines represent the data from three replicate infections. Blue or red arrows represent annotated genes on the positive or negative strand of the genome, respectively. Gray boxes represent the two internal repeat regions of the γHV68 genome that are excluded from our analysis. EGRs 1–30, which are greater than 700 nt and expressed one standard deviation above the mean mock-infected signal in at least two replicates, are shown as diagonal-filled arrows. Consensus polyadenylation signal sequences, AAUAAA and AUUAAA, are represented as black bars above the fluorescence signal. A) Genomic positions 1–40 kb. B) Genomic positions 40–80 kb. C) Genomic positions 80–119.45 kb. See also Figure S1.

Figure 2. Expressed Genomic Regions 8 and 24

A) Tiled array results for EGR 8 reveals a considerable 5’ extension to the K3 gene. Schematic representation of the transcripts revealed by Northern analysis and RACE are shown as black and green lines, respectively. Single stranded RNA probes used in the Northern analysis are shown as the black lines labeled P1 and P2. Light red arrows at the bottom of the figure represent predicted open reading frames of greater than 20 amino acids. B) Northern analysis of K3 and EGR 8 reveals the presence of 1.1 kb and 2 kb transcripts. C) Tiled array results for EGR 24 reveals transcriptional signal between ORF-53 and ORF-55. Schematic representation of the transcripts revealed by Northern analysis and RACE are shown as black and green lines, respectively. Single stranded RNA probes used in the Northern analysis are shown as the black lines labeled P1, P2 and P3. D) Northern analysis of ORFs 52, 53, 55 and EGR 25 reveals the presence of 800 nt, 860 nt, 1.8 kb, and 2.5 kb bands. S = sense control RNA; AS = antisense control RNA; M-8 = mock infected, collected at 8 hpi; γHV68-8 = infected at MOI=10, collected at 8 hpi; γHV68-8C = infected at MOI=10 and treated with 200ng/ml cycloheximide, collected at 8 hpi; M-18 = mock infected, collected at 18 hpi; γHV68-18 = infected at MOI=10, collected at 18 hpi; γHV68-18C = infected at MOI=10 and treated with 42 µg/ml cidofovir, collected at 18hpi.

Figure 3. Expressed Genomic Region 26

A) Tiled array results for EGR 26 reveals a 9.8 kb extension 3’ to the M9 gene. Single stranded RNA probes used in the Northern analysis are shown as the black lines labeled P1-4. Light red arrows at the bottom of the figure represent predicted open reading frames of greater than 20 amino acids. B) Northern analysis of EGR 26 reveals 0.7 kb, 3 kb, 4 kb, 8 kb, and 16 kb bands. M-18 = mock infected, collected at 18 hpi; γHV68-18 = infected at MOI=10, collected at 18hpi.

Figure 4. Tiled array identifies immediate early transcripts

3T12 cells were infected in the presence, green lines, or absence, black lines, of 200 ng/ml cycloheximide and harvested 8 hpi. Three replicates of each condition were performed and are represented as individual lines. There is considerable disruption of transcription across the γHV68 genome with the exception of ORF-50, ORF-73, ORF-75a, M12, M13, and M14. Expression of ORF-61, M4, ORF-75b, and ORF-75c is decreased but still clearly detectable. A) Genomic positions 1–40 kb. B) Genomic positions 40–80 kb. C) Genomic positions 80–119.45 kb.

Figure 5. Genome-wide mapping of cidofovir sensitivity

3T12 cells were infected in the presence, green lines, or absence, black lines, of 42 µg/ml cidofovir and harvested 18 hpi. Three replicates of each condition were performed and are represented as individual lines. Those regions that are greater than 4 fold down regulated in the presence of cidofovir are plotted on a linear scale on a nucleotide-by-nucleotide basis on the black line graph below the transcriptional signal. A) Genomic positions 1–40 kb. B) Genomic positions 40–80 kb. C) Genomic positions 80–119.45 kb. See also Figure S5.