Fig 1.
Structure and expression of the HTLV-1 provirus in the host genome.
(A) Definitions regarding classification of the direction of transcription. Here we use ‘same sense’ to denote transcription from the same strand of the host genome as the proviral plus strand (e.g. tax gene). We refer to flanking host transcription upstream and downstream of 5′LTR-driven plus-strand proviral expression respectively as (a) same (plus) sense, 5′ side of the provirus and (b) same sense, 3′ side of the provirus. Similarly, we refer to transcription upstream and downstream of 3′LTR-driven minus-strand proviral expression respectively as (c) antisense, 3′ side of the provirus and (d) antisense, 5′ side of the provirus. In this study we focus on the effects of plus-strand expression. (B) Diagram of HTLV-1 proviral genome and splicing pattern of the regulatory genes tax (encoded in plus strand) and HBZ (encoded in minus strand); red boxes represent exons. The 4C viewpoint containing the CTCF binding site (BS) is shown.
Table 1.
Clone list.
Table 2.
Tax expression of HTLV-1-infected T cell clones.
Fig 2.
HTLV-1 plus-strand expression results in fewer chromatin loops (number of q4C peaks) between provirus and host genome.
(A) q4C profiles of Tax−and Tax+ cells of two different clones (3.60 and HA1). For each clone, the top panel depicts the q4C profile in the 5′ and 3′ host genome flanking the provirus (two biological duplicates), quantified as the normalized frequency of ligation events in overlapping windows (window width 10 kb, step 1 kb). On the horizontal axis, positive values denote positions extending from the 3′ LTR side of the provirus; negative values denote positions 5′ of the 5′LTR. VP–viewpoint in q4C (proviral integration site). Diamonds mark the positions of reproducible chromatin contact sites identified by the peak calling algorithm. CTCF track–open arrowheads denote positions of CTCF-binding sites (BS); the filled arrowhead denotes the CTCF-BS in the provirus. Genes track shows RefSeq protein-coding genes in the flanking host genome. The q4C profiles of remaining clones are shown in S2 and S3 Figs. (B) Number of identified peaks in non-expressing (Tax- or GFP-) and expressing (Tax+ or GFP+) subsets isolated from 6 clones: total (all peaks) and (C) peaks with or without a CTCF binding site.
Fig 3.
HTLV-1 plus-strand expression results in a significant reduction in contact frequency (q4C peak height) with the host genome (A) Schematic to show quantification of change in peak frequency with plus-strand expression. First, reproducible q4C peaks were identified in the non-expressing (Tax-negative) cells, as previously described [10]. Then, peaks were sought (using the same algorithm) in each corresponding genomic location in the plus-strand-expressing cell population. (B) Normalized peak height of q4C peaks identified in all clones analysed, respectively on the 5′ and 3′ sides of the provirus. Peak height is defined as the maximum number of ligation events per region (normalised to total ligation events in a sample) calculated for each peak region. p = 0.027 and p = 9.7 * 10−13 for upstream and downstream regions, respectively (unpaired two-tailed Wilcoxon test. (C) Normalized peak height of q4C peaks identified in all clones analysed, comparing peaks that contain a CTCF site and those without CTCF sites. Peak height was significantly greater in non-expressing cells, both in peaks with a CTCF site and those without (p = 0.0018 and p = 1.8 * 10−13, respectively, unpaired two-tailed Wilcoxon test).
Fig 4.
Proviral and host transcription and splicing in live-sorted T cell clones.
(A) RNA-seq analysis of HTLV-1 proviral expression in live-sorted d2EGFP clones (TBX4B, 11.50 and 11.63). (B) Tax expression measured by qPCR with primers specific for tax mRNA or 18S ribosomal RNA (18S rRNA). qPCR plots are expression values normalized to 18S rRNA. Data represent a mean of two biological replicates; error bars are SEM. AU—arbitrary units. (C) Host RNA expression 30kb on either side of the proviral integration site. On the horizontal axis, positive values denote positions extending from the 3′ LTR side of the provirus; negative values denote positions 5′ of the 5′LTR. Each row shows the transcription density (normalized RNA-seq read count) flanking that genomic position in the clone indicated at the right-hand side. In each case, transcription orientation and positions are shown relative to the integrated provirus. Read density shown in blue shows transcription in the same orientation as the proviral plus-strand (same sense); red shows transcription in the opposite sense to the proviral plus-strand (antisense). (D) Identification of splice sites of viral-host fusion transcripts in d2EGFP-TBX4B clone cells. Coverage tracks of same sense transcription (blue) and antisense transcription (red) in Integrative Genomics Viewer (IGV). Exons of PNPLA3 in the 3′ side of the integration site are highlighted in yellow. (E) Fusion transcripts between an HTLV-1 plus-strand major splice donor (red, proviral exon H1 or H2) and the canonical splice acceptor site in the host PNPLA3 gene (blue, PNPLA3 exon 3) were identified in GFP+ (HTLV-1 plus-strand-expressing) cells. To identify splice sites of fusion transcripts, reads were aligned to a reference genome (hg19) containing the HTLV-1 provirus (AB513134) genome in the TBX4B clone integration site at chr22:44323198. Fusion transcripts are shown with fused sequences.
Fig 5.
Treatment of T cell clones with a transcriptional elongation inhibitor allows recovery of chromatin loop formation in the Tax+ population.
(A) After treatment with 1 nM flavopiridol (FP) for 1.5 hrs, total RNA was extracted from clone 11.63 and subjected to RT-qPCR for tax and three regions in the 3′ flanking host genome, respectively a: +188 bp, b: +535 bp and c: +3,198 bp from the 3′ end of the provirus. (B) Relative expression intensity (normalized to 18s rRNA) of tax and the host genome at positions a, b and c. Data are mean ± SEM. (N = 3). * P<0.05 (paired t-test). (C) q4C profiles of mock-treated Tax- cells (top track) and Tax+ cells (middle track), and flavopiridol (FP)-treated Tax+ cells (bottom track). Diamonds mark the positions of reproducible chromatin contact sites identified by the peak calling algorithm. Open arrowheads denote positions of CTCF-BS; the filled arrowhead denotes the CTCF-BS in the provirus. Gene track shows RefSeq protein-coding genes in the flanking host genome.