Control of Archetype BK Polyomavirus MicroRNA Expression

Abstract

BK polyomavirus (BKPyV) is a ubiquitous human pathogen, with over 80% of adults worldwide being persistently infected. BKPyV infection is usually asymptomatic in healthy people; however, it causes polyomavirus-associated nephropathy in renal transplant patients and hemorrhagic cystitis in bone marrow transplant patients. BKPyV has a circular, double-stranded DNA genome that is divided genetically into three parts: an early region, a late region, and a noncoding control region (NCCR). The NCCR contains the viral DNA replication origin and cis-acting elements regulating viral early and late gene expression. It was previously shown that a BKPyV microRNA (miRNA) expressed from the late strand regulates viral large-T-antigen expression and limits the replication capacity of archetype BKPyV. A major unanswered question in the field is how expression of the viral miRNA is regulated. Typically, miRNA is expressed from introns in cellular genes, but there is no intron readily apparent in BKPyV from which the miRNA could derive. Here, we provide evidence for primary RNA transcripts that circle the genome more than once and include the NCCR. We identified splice junctions resulting from splicing of primary transcripts circling the genome more than once, and Sanger sequencing of reverse transcription-PCR (RT-PCR) products indicates that there are viral transcripts that circle the genome up to four times. Our data suggest that the miRNA is expressed from an intron spliced out of these greater-than-genome-size primary transcripts.IMPORTANCE The BK polyomavirus (BKPyV) miRNA plays an important role in regulating viral large-T-antigen expression and limiting the replication of archetype BKPyV, suggesting that the miRNA regulates BKPyV persistence. However, how miRNA expression is regulated is poorly understood. Here, we present evidence that the miRNA is expressed from an intron that is generated by RNA polymerase II transcribing the circular viral genome more than once. We identified splice junctions that could be generated only from primary transcripts that contain tandemly repeated copies of the viral genome. The results indicate another way in which viruses optimize expression of their genes using limited coding capacity.

Keywords: BKPyV; RNA splicing; miRNA; polyomavirus.

FIG 1

A circular viral template expresses higher levels of miRNA than a linear template. (A) Model of miRNA biogenesis, showing the pre-miRNA coming from the intron (blue line) of a primary transcript that is longer than the genome. The exon in this example (red), when spliced, forms a VP1 mRNA. D, splice donor; A, splice acceptor; Late pA, late polyadenylation signal site. The mature miRNA is processed from the putative intron. The location of the primer NCCR-r, which was used to perform reverse transcription to confirm transcription of the NCCR, is shown. (B) Total RNA was extracted from transfected 293 cells and assayed for BKPyV miRNA expression using stem-loop RT qPCR. The results are normalized to the cellular miRNA control, hsa-let-7a-1. The circular-template value was arbitrarily set to 1. ***, P < 0.005.

FIG 2

Viral RNA transcripts contain the NCCR region. RPTE cells were infected with archetype Dik virus (A) or rearranged variantDunlop virus (B) or mock infected. Total RNA was extracted 3 dpi. Reverse transcription (RT) and PCR amplification with specific NCCR primers (see Materials and Methods and Fig. 1A) were performed, and the PCR products were electrophoresed on a 2% DNA agarose gel. Representative gels from three independent repeats with each virus are shown. (C) RNA-seq read coverage of the Dunlop genome. Data are from one representative sample. Read coverage on the late strand (red) and early strand (blue) graphed along the length of the viral genome is shown. The locations of the noncoding control region (NCCR), agnoprotein gene (Agno), early genes (EG), late genes (LG), and miRNA gene are indicated. The enlarged area shows the coverage for NCCR and Agno regions.

FIG 3

Evidence for viral primary mRNAs that circle the genome more than once. (A) (Line 1) Linear representation of a primary tandem genomic transcript resulting from the RNA polymerase transcribing the genome more than once. The relevant genomic regions are shown in different colors (not to scale), and the locations of putative splice donors (D) and acceptors (A) based on the RNA-seq data are shown. A1 and D1 are in the first tandem copy, and A2 and D2 are in the second copy. Also shown are the two PCR primers (P1 and P2) used to analyze the RNA products. cDNA from this unspliced RNA would not be amplified by primers 1 and 2 in our assay because the product would be >5 kb. (Line 2) Spliced product resulting from excision of a genome-sized intron. Primers 1 and 2 can now amplify cDNA from this RNA, yielding a product of 99 bp, or larger if the transcript circled the genome more than twice and was subsequently spliced. (Line 3) Intron resulting from splicing to give the product in line 2. (B) TA cloning to confirm the splice junctions detected in the RNA-seq data. Specific primers shown in panel A were used to amplify the cDNA transcribed from RNA of archetype-infected RPTE cells. PCR products were cloned into pCR2.1-TOPO vector, and 12 random clones were chosen for DNA minipreps. The DNA was digested with BamHI and XbaI and electrophoresed on a 2% DNA agarose gel. The bands are 94 bp longer than the cDNA insert due to the location of the restriction sites used to excise the inserts. The diagrams to the left of the gel indicate the structures of the spliced products represented by the different-sized bands. Lane 10 is empty pCR2.1-TOPO vector digested with BamHI and XbaI, which produces a 94-bp band because there is no insert.