Fig 1.
BAC-cloning of CCMV strain Heberling.
(A) Targeted integration of the BAC cassette into the CCMV Heberling genome was mediated by homology arms left (HL) and right (HR), replacing the region between US1 and US7. The BAC cassette contains a replication origin (ori), a centromeric region (sopC), genes encoding replication (repE) and partitioning factors (sopA, sopB), as well as resistance genes for eukaryotic (xanthine phosphoribosyltransferase—gpt) and prokaryotic selection (chloramphenicol acetyltransferase—CAT). TR—terminal repeats; UL—unique long region; IR—internal repeats; US—unique short region. (B) Five BAC clones showed an AgeI restriction pattern matching the predicted in silico digest. Arrowheads indicate bands that were missing due to an unintended deletion of IRS1 to US1 and US8 to US13 regions. (C) All selected CCMV BAC clones were infectious upon transfection into fibroblast but differ in their titer output performance. Means (center of the error bar) and standard errors of the mean of n = 3 are depicted.
Fig 2.
Genotypic analysis of CCMV clones.
(A) Drop in sequencing read coverage indicates that a large proportion of viral genomes within the parental CCMV Heberling pool have a deletion event between IRS1 and US10. (B) Whole genome sequencing revealed variations in genomic sequences of selected CCMV BAC clones and the parental CCMV Heberling. Positions of missense, frameshift, deletion and insertion mutations for BAC clones are plotted in the grey tracks of the Circos plot. Intra-strain variations of the CCMV Heberling isolate are plotted in the green track; the frequency of each variation is indicated. Outer tracks: CCMV ORFs (red) and genomic regions. (C) The individual BAC clones were compared with the CCMV Heberling reference sequence for the absolute counts of single nucleotide variations (SNVs). White stacked bars indicate number of variations detected based on the alignment to the reference genome sequence of Heberling (NC_003521.1) but found in the re-sequenced parental CCMV virus stock pool. Colored bars indicate de novo variations in each clone that are neither detected in the original reference sequence nor the parental CCMV virus stock. CCMV BAC-177 showed the lowest number of SNVs and contained no frameshift mutations. (D) Similarly, the BAC clones were analyzed for the spectrum of transition and transversion mutations.
Fig 3.
UL128 repair rescues CCMV’s tropism for epithelial and endothelial cells.
(A) In CCMV Heberling, a frameshift mutation in the first exon of UL128 leads to a premature stop codon in exon 2. Sanger sequencing chromatograms of BAC-177 and the UL128-repaired BAC-177 document the successful removal of the frameshift by BAC mutagenesis. (B, C) The indicated cell types were infected with viruses derived either from the Heberling isolate, BAC-177 or the UL128-repaired BAC-177, using an MOI of 0.1. At 24 h post infection, the cells were analyzed for IE1/2 gene expression by flow cytometry (B). Virus growth in epithelial cells was monitored over a 33-day period. Means (center of the error bar) and standard errors of the mean of n = 3 are depicted (C).
Fig 4.
Restoration of deleted US regions and floxing of BAC cassette.
(A) Schematic overview of US2-US6 and US8-US13 regions that were restored by BAC mutagenesis in CCMV BAC-177. Region IRS1 to US1 was not reinstated. LoxP sites were inserted at both ends of the BAC cassette to remove prokaryotic sequences by Cre recombinase expression during virus reconstitution. The final CCMV-BAC construct was named BAC-Phan9. Whole genome sequencing verified the integrity of BAC-Phan9. Coverage of DNA sequencing and positions of minor sequence variations in the US and TR regions are plotted. Sanger sequencing of DNA from reconstituted virus confirms the expected Cre-mediated recombination outcome at the US6-US7 region. (B) One-step growth curves of CCMV BAC-177, BAC-Phan9 and CCMV Heberling were analyzed in human fibroblasts. Means (center of the error bar) and standard errors of the mean of n = 3 are depicted. (C) Cells infected with either CCMV Heberling or BAC-Phan9 were analyzed for protein expression by mass spectrometry. The observed differences at 6 and 24 hpi are depicted as volcano plots. Proteins enriched in BAC-Phan9 infection and proteins orginitating from the US2-US10 region are labeled in blue, proteins enriched in Heberling infection in red. (D) Flow cytometry analysis of MHC-I surface expression during infection with CCMV BAC-Phan9, HCMV Merlin (upper row) and CCMV Heberling (lower row).
Fig 5.
Re-evaluation of the CCMV coding potential.
(A) To annotate potential novel coding sequences, the CCMV genome was scanned for unknown six-frame translation products, for homologs of non-canonical HCMV ORFs [6], for splice variants and for small ORFs (smORFs). To validate the candidate ORFs (numbers are indicated), proteomic data of CCMV-infected cells was analyzed for matching peptides. In total, 14 novel gene products were unambiguously identified by at least two unique peptides. (B) UL91.1 as an example of a newly found non-canonical CCMV ORF. RNA sequencing (RNA-seq) of infected cells detected a splicing event between the UL88 and UL91 locus. Multiple peptides matched the annotated ORF sequence, including the N-terminal peptide and a peptide spanning over the exon/exon junction, thus confirming translation of this splice variant. (C) smORF1 is another unique ORF with yet unknown function that overlaps the lytic replication origin of CCMV including the 5’-end of the long non-coding RNA4.9.
Fig 6.
Temporal classes of CCMV protein expression.
CCMV-infected cells were harvested at the indicated times post infection and subjected to a mass spectrometry-based proteomic analysis. K-means clustering classified viral gene expression into five temporal profiles (TP). The left-hand panel displays the average cluster profiles of TP classes 1–5. On the right hand, the expression profiles of individual CCMV proteins are plotted as a heatmap.
Fig 7.
Gene ontology (GO) analysis of the most deregulated host proteins during CCMV and HCMV infection.
(A) For each time point post infection, fold-changes in host protein expression were calculated relative to uninfected control cells (0 h). The top and bottom 3% of protein fold-changes were selected from each time point and used for multi-list GO enrichment analysis. (B) Heatmap showing the GO terms for strongly increased or depleted proteins during HCMV and CCMV infection (LC: long chain). The HCMV data set WCL2 was taken from [47].
Fig 8.
Comparative analysis of CCMV-induced changes in host gene expression at the transcript and protein level.
At the indicated time points total RNA and protein were prepared from CCMV- infected cells and analyzed by RNA sequencing and tandem mass spectrometry respectively. K-means clustering of 4,748 transcripts and proteins identifies seven individual clusters representing different modes of host gene regulation during CCMV infection. Shown are the average mRNA and protein profiles of each cluster (left panel), a heatmap of the corresponding gene expression data (middle panel) and selected representative profiles from each cluster (right panel).