Decoding murine cytomegalovirus

Abstract

The genomes of both human cytomegalovirus (HCMV) and murine cytomegalovirus (MCMV) were first sequenced over 20 years ago. Similar to HCMV, the MCMV genome had initially been proposed to harbor ≈170 open reading frames (ORFs). More recently, omics approaches revealed HCMV gene expression to be substantially more complex comprising several hundred viral ORFs. Here, we provide a state-of-the art reannotation of lytic MCMV gene expression based on integrative analysis of a large set of omics data. Our data reveal 365 viral transcription start sites (TiSS) that give rise to 380 and 454 viral transcripts and ORFs, respectively. The latter include >200 small ORFs, some of which represented the most highly expressed viral gene products. By combining TiSS profiling with metabolic RNA labelling and chemical nucleotide conversion sequencing (dSLAM-seq), we provide a detailed picture of the expression kinetics of viral transcription. This not only resulted in the identification of a novel MCMV immediate early transcript encoding the m166.5 ORF, which we termed ie4, but also revealed a group of well-expressed viral transcripts that are induced later than canonical true late genes and contain an initiator element (Inr) but no TATA- or TATT-box in their core promoters. We show that viral upstream ORFs (uORFs) tune gene expression of longer viral ORFs expressed in cis at translational level. Finally, we identify a truncated isoform of the viral NK-cell immune evasin m145 arising from a viral TiSS downstream of the canonical m145 mRNA. Despite being ≈5-fold more abundantly expressed than the canonical m145 protein it was not required for downregulating the NK cell ligand, MULT-I. In summary, our work will pave the way for future mechanistic studies on previously unknown cytomegalovirus gene products in an important virus animal model.

Fig 1. Overview of applied omics approaches.

MCMV gene expression was analyzed in Swiss murine embryonic fibroblasts (NIH-3T3) infected with BAC-derived wild-type MCMV at an MOI of 10. Viral transcription start sites (TiSS) and splicing events were determined through total RNA-seq, 4sU-seq, cRNA-seq and dSLAM-seq (n = 2; including one biological replicate for dSLAM-seq with cycloheximide (CHX; 4 h) or phosphonoacetic acid (PAA; 24 h) treatment). To decipher the MCMV translatome, four biological replicates of ribosome profiling were performed. Enrichment of reads at translation start sites (TaSS) was improved by pre-treating cells with Harringtonine–Harr. (two biological replicates) or Lactimidomycin–Lacti. (one biological replicate) for 30 min. The available time points and conditions are indicated by stars for any given approach.

Fig 2. Characterization of the MCMV transcriptome.

A. Screenshot of MCMV gene expression showing annotated transcripts and ORFs in the M25 locus with 5’ read enrichment at TiSS as depicted by cRNA-seq and dSLAM-seq as well as Ribo-seq data, respectively. Four viral transcripts, which initiate within the M25 region of the MCMV genome, are highlighted in yellow (TiSS scores in S4 Table). The schematic depicts translation of the 130 and 105 kDa M25 protein isoforms validated in a recent study [27] and validated by our Ribo-seq data. The M25 RNA *1 also encodes four small ORFs (M25 uORFs 1–3 and M25 uoORF) of 6, 11, 8 and 63 aa, respectively whose expression levels and kinetics (6 hpi) correspond to their respective transcript (M25 RNA *1) and were hence annotated. dSLAM-seq data are depicted in linear scale, Ribo-seq data in logarithmic scale. B. Graphical representation of 5’ read enrichment obtained by dSLAM-seq and cRNA-seq approaches. C. Venn diagram depicting the number of TiSS identified by both cRNA-seq and dSLAM-seq. TiSS included in the final annotation are depicted in the green circle as ‘annotated’. TiSS labelled as ‘Required for an ORF’ represent TiSS that are required to explain the translation of a downstream ORF (no other TiSS within 500 nt upstream of the ORF).

Fig 3. Identification of MCMV splicing events.

Mapped reads from 4sU-seq and total RNA-seq identified 366 putative splicing events in the MCMV transcriptome. The y-axis displays the number of reads occurring at a spliced region, further categorized into reads spanning exon-exon junctions (red) by at least 10 nt as well as non-exon-spanning reads upstream (green) and downstream (blue). Putative splicing events were sorted based on the ratio of spliced (red) to unspliced (green + blue) reads. Only 28 of the 366 putative splicing events were included into our new reference annotation because they (i) had already been identified by others (16/28; S2 Table), (ii) were highly abundant, or (iii) affected the coding sequence of an MCMV ORF or sORF. To avoid unnecessary complexity in the revised annotation of the MCMV transcriptome, we excluded the other (putative) splicing events from our new reference annotation.

Fig 4. dSLAM-seq reveals distinct core promoter motifs associated with viral gene expression kinetics and a novel viral ie gene (ie4).

A. Depiction of core promoter motifs of viral TiSS clustered according to their maximal transcription rates (new RNA derived from the dSLAM-seq data) in three equally sized bins (high, mid and low). The TATA box and initiator element (Inr) are shown. B. Ratio of new RNA levels at 4 hpi with and without cycloheximide (CHX) treatment (n = 1) are shown for the 365 MCMV TiSS (grey dots). This identified three immediate early TiSS, namely ie1/ie3, ie2 and ie4 (m166.5), highlighted in colored dots. C. Validation of the m166.5 RNA (ie4) as a so far unknown MCMV immediate early gene by qRT-PCR. qRT-PCR was performed on total RNA isolated from MCMV-infected NIH-3T3 cells harvested at 4 hpi with and without CHX treatment. GAPDH was used as a housekeeping gene and results were plotted as fold change relative to MCMV infection under DMSO treatment for three biological replicates. D. Graphical depiction of 4 clusters (CL1-4) of viral TiSS obtained by unsupervised clustering on new RNA derived from the dSLAM-seq data (relative expression levels shown). Relative levels were calculated for each TiSS on the basis of new RNA levels which were normalized (reads per million) and scaled such that the maximum across any given time point is 1. E. Graphical and sequence logo depiction of core promoter motifs identified for CL1-4 clusters through MEME motif analysis. Please note that the TATA-/TATT-box motif in cluster CL3 is shifted by 2 nt to the left compared to cluster CL1. TBM: TATA-box like motif, Inr: Initiator element.

Fig 5. Transcription kinetics of viral TiSS.

A. Graphical depiction of 6 clusters (TR0-5) of viral TiSS obtained by manual classification (for criteria see Methods) based on new RNA derived from the dSLAM-seq data (relative expression levels shown) computed similarly as in Fig 4D. B. Graphical and sequence logo depiction of core promoter motifs identified for TR0-5 clusters through MEME motif analysis computed as in Fig 4E. Please note that the TATA-/TATT-box motif in cluster TR4 is shifted by 2 nt to the left compared to cluster TR1. The TR3 cluster did not enrich for a specific motif, which is at least in parts due to the small number of TiSS (n = 13) in this cluster. TBM: TATA-box like motif, Inr: Initiator element, NS: Not significant.

Fig 6. Converting a TATA box to a TATT box is sufficient to alter viral gene expression kinetics.

A. NIH-3T3 cells were infected with a two-color MCMV reporter virus (MCMV_TATA-Δm152-eGFP_SCP-IRES-mCherry) and the TATA>TATT mutant thereof (MCMV_TATT-Δm152-eGFP_SCP-IRES-mCherry) at an MOI of 5 for the indicated time points with and without PAA treatment. mCherry and eGFP expression were analyzed through fluorescence microscopy. Representative images of three biological replicates (n = 3) are shown. B. Cells were fixed and eGFP and mCherry levels were analyzed quantitatively through flow cytometry and mean fluorescent intensity (MFI) values were plotted for three biological replicates (n = 3) along with standard deviation (S.D.) with and without PAA pre-treatment.

Fig 7. The MCMV translatome.

A. Venn diagram depicting the number of MCMV ORFs in our revised MCMV genome annotation as detected by ribosome profiling compared to the Rawlinson et al. reference annotation [13]. B. Total number of viral ORFs annotated by ribosome profiling grouped into CDS (Rawlinson reference annotation), large ORFs, short ORFs (sORFs), upstream ORFs (uORFs), upstream overlapping ORFs (uoORFs), iORFs (internal ORFs) and downstream ORFs (dORFs). N-terminal extensions (NTEs) or truncations (NTTs) may span any of these defined groups of ORFs. C-F. Start codon usage for annotated novel large ORFs, small ORFs, NTEs and NTTs, respectively. ORFs in gray depict orphan ORFs, for which no TiSS could be identified. Each graph depicts the number of ORFs on the y-axis and the start codon usage on the x-axis.

Fig 8. Characterization of N-terminally truncated ORFs in the m145 locus.

A. ORFs and transcripts expressed from the m145 locus. This includes the so far unknown m145 ORF #1 and #2 expressed from m145 RNA #1. Coordinates for the TiSS and ORF start codon are shown for each transcript and ORF. dSLAM-seq data are shown in linear scale, Ribo-seq data in logarithmic scale. Aggregated reads across all time points mapping to the m145 locus are shown. B. Schematic representation of the MCMV mutants generated to characterize novel viral gene products encoded by the m145 locus. Mutant viruses were generated based on a reporter virus with a V5-tag inserted at the C-terminus of the canonical m145 CDS. The viruses were generated by en passant BAC mutagenesis as described in methods on this backbone. The Δm145 CDS harbored a STOP codon at the 40^th codon to skip additional AUGs downstream of the m145 CDS signal peptide which may have resulted in additional products, hindering accurate analysis of the locus. The Δm145 TATA RNA #1 mutant included a mutation in the TATA box of the respective transcript to abrogate gene expression downstream while the Δm145 ORF #1 mut mutant was created by mutating the start codon of m145 ORF #1. The Δm145 virus is a previously created virus where the entire m145 locus (i.e. m145 CDS) was replaced by a kanamycin cassette. C. SVEC 4–10 murine endothelial cells were infected with the indicated viruses at an MOI of 1 for 24 and 48 h. V5-tagged m145 gene products were characterized by Western blot. Parental WT MCMV infection was used as negative control. D. SVEC 4–10 cells were infected for 48 h with the m145-V5 virus at an MOI of 1. Cells were harvested and treated with or without EndoH_f (E) or O-glycosidase (O) to qualitatively analyze glycosylation patterns of m145 gene products via Western blot. The m145 CDS gene product of 70 kDa shifted to 55 kDa upon EndoH_f treatment justifying its actual predicted weight. E. SVEC 4–10 cells were infected with m145 virus mutants at an MOI of 1 for 18 h and stained with rat anti-MULT-I and mouse anti-m04 antibodies following cell surface MULT-I analysis through flow cytometry by gating on infected cells (m04+). Anti-rat and anti-mouse isotype antibodies were utilized as negative controls. Western blots and flow cytometry histograms are a representative for two (n = 2) and three biological replicates (n = 3), respectively.

Fig 9. MCMV uORFs/uoORFs tune viral gene expression.

Ribo-seq data (aggregated reads in logarithmic scale) of the respective viral genomic loci and their validation by dual luciferase assays are shown for m169 uORF (A), m119.3 uORF (B), M35 uoORF (C) and M48 uoORF (D). The number of AUG codons for the respective viral u(o)ORFs are indicated. Coordinates represent the start codons of the u(o)ORFs and ORFs. Both ORFs and u(o)ORFs were annotated based on translational start site profiling data, including Harringtonine and Lactimidomycin pre-treated samples and PRICE analysis. Manual curation assessed the presence of upstream TiSS, start codon usage and presence of STOP codons to identify relevant ORFs. psiCheck-2 reporter plasmids harbored the indicated MCMV u(o)ORFs (WT) and AUG start codon mutants thereof (Mut) upstream of firefly-luc reporter gene. Luciferase assay data at 48h post transfection are shown as mean RLU (Firefly/Renilla ratio) of three biological replicates (n = 3) plotted along with the standard error (S.E.M.).