|
Figure 1. Percent of reads mapped to functionally different regions on FV3 genome. The FV3 genome regions are functionally classified as exons, introns, or intergenic regions based on annotation of the reference genome (NC_005946.1). As intronic regions (introns) are lacking in ranaviral coding genes, about 50 intergenic regions are interspersed between ORFs. The intergenic regions take about 20% of the FV3 genome with a length of 20-900 bp. Transcriptomic reads in most infected tissues are also remarkably mapped within these intergenic regions, indicating that these intergenic regions are transcribed and probably function as regulatory RNA species.
|
|
Figure 2. Transcriptomic comparison and distribution of TATA-Box-like cis-element in intergenic regions of the FV3 genome. (A) Line chart depicts cross-tissue averages of RNA-Seq reads differentially mapped to intergenic regions and almost all annotated FV3 coding ORFs labeled on the top. Note the X-Axis tick labels on the top for even-numbered ORFs (such as FV3gorf2L between FV3gorf1R and FV3gorf3R) are omitted due to the space limitation. Bar chart depicts the position weight matrix (PWM) scores of the TATA-box, a cis-regulatory element (CRE) marking core promoters of eukaryotic genes significantly detected across all FV3-genome intergenic regions (labeled as FV3UTR start-end nt position along the FV3 reference genome). (B) Mean PWM scores of TATA-box CRE in FV3 intergenic regions that are intermediately upstream of top-ten highly expressed FV3 coding genes (ORFs) in each temporal class of immediate early (IE), delay early (DE), or late (L) genes as revealed by transcriptomic analyses. Mean PWM scores were calculated using tools at https://ccg.epfl.ch/pwmtools/pwmscore.php. In both (A) and (B), the cross-panel mPWM scores of the TATA-box CRE is averagely (Ave) shown as data-labeled black bar at the right. (C) The matrix of TATA-box that interacts with a transcription factor of TATA-box binding protein (TBP) is from MEME-derived JASPAR CORE 2020 vertebrates affiliated with the PWM tools.
|
|
Figure 3. Comparison of position weight matrix (PWM) scores of key cis-regulatory elements (CREs) detected in FV3-genome intergenic regions, and that interact with vertebrate transcription factors potently in immune regulation. Shown are mean PWM scores of CREs in FV3 intergenic regions that were significantly detected to bind (A) IRF-like, (B) NF-κB2-like, (C) STAT1-like, (D) CEBP-like, (E) CREB-like, and (F) PU.1 (a.k.a. SPI1) transcription factors. Mean PWM scores were calculated using tools at https://ccg.epfl.ch/pwmtools/pwmscore.php with CRE Matrices (indicated by Matrix or Cluster numbers, and schematics in
Figure 4
) are from MEME-derived JASPAR CORE 2020 vertebrates or JASPAR CORE 2018 vertebrates clustering affiliated with the PWM tools. The genome-wide mPWM scores across all intergenic regions for each CRE are averagely shown (Ave) as data-labeled black bars at the right for overall comparison. CEBP, CCAAT enhancer binding protein beta; CREB, cAMP-response element binding protein; IRF, interferon regulatory factor; NF-κB, Nuclear factor-κB; SPI1 or PU.1, a TF binding PU-box, a purine-rich DNA sequence; and STAT, signal transducer and activator of transcription.
|
|
Figure 4. Intergenic regions immediately upstream of highly expressed FV3 genes serve as putative core promoters with enhanced capacity to bind vertebrate transcription factors of (A) IRFs, (B) NF-κB2-like, and (C) STAT1-like, but not much enhanced for (D) CEBPA, (E) CREB1, and (F) SPI1 transcription factors. Shown are mean PWM scores of cis-regulatory elements (CREs) in FV3 intergenic regions that are immediately upstream of top-ten highly expressed FV3 coding genes (ORFs) in each temporal class of immediate early (IE), delay early (DE), or late (L) genes. Mean PWM scores were calculated using tools at https://ccg.epfl.ch/pwmtools/pwmscore.php with CRE Matrices are from MEME-derived JASPAR CORE 2020 vertebrates or JASPAR CORE 2018 vertebrates clustering affiliated with the PWM tools. The cross-panel average mPWM scores (Ave) of each CRE are shown as data-labeled black bars at the right for overall comparison. Abbreviations of TFs are as in
Figure 3
.
|
|
Figure 5. Intergenic regions immediately ahead of highly expressed FV3 genes containing cis-regulatory elements (CREs) exhibit higher likelihood of binding vertebrate IRFs, NF-κB2-like, and STAT1-like transcription factors. (A) Shown are overall averages of PWM scores per compared CREs in all FV3 intergenic regions (All) and those are immediately upstream of top-ten highly expressed FV3 coding genes (Top10) in each temporal class of immediate early (IE), delay early (DE), or late (L) genes as revealed by transcriptomic analyses. Mean PWM scores were calculated as in previous figures. *p < 0.001 and n = 10, compared to the All group. (B) The CRE PWM enhancing index was adopted to compare fold changes of mean PWM scores between the Top10 and All groups after normalization with the PWM evolution of TATA-box between the two groups as baseline (indicated by the dash line). Abbreviations of TFs are as in
Figure 3
.
|
|
Figure 6. Comparison of transcriptomic and enrichment of putative microRNA (miRNA) sequences in intergenic regions of FV3 genome. (A) As line chart in
Figure 2
, mean RNA-Seq reads are differentially distributed among intergenic regions and almost all annotated FV3 coding ORFs. A distribution plot between the vertical Axis and gene labels, shows the median of read density (Log2 Unit) of mapped reads along the FV3 genome as in the FV3-Δ64R-infected kidney to show the full-genome coverage at both positive (green) and negative (orange) strand orientations. Transcription of the intergenic regions along the higher read density spanning the ORF coding genes is shown using the shaded blue curve indicating mean read counts across the eight infected tissues tested. (B) The prediction of miRNA-like sequences in most intergenic regions (marked as UTR start-end site along FV3 reference genome including the 5’- and 3’-untranslated regions), which are especially enriched in five regions (named as C, I, R, AF and AT per putative miRNA density/Kb) as marked using blue dash line. The sequence information of all predicted miRNAs is listed in
Supplemental Excel Sheet
. The miRNA prediction and target validation were performed using three RNA analysis programs through an online BiBiServ Service.
|
|
Figure 7. Transcriptomic analysis of the viral genome and X. laevis mRNA encoding interferon receptor subunits in the control (Ctrl) and FV3-infected kidney. (A) The virus-targeted transcriptome analysis shown as a distribution plot of mapped reads in FV3 genome (GenBank Accession No. NC_005946.1). The X-axis shows the length of the genome (in Mb, 0.105 Mb of FV3), and the Y-axis indicates the log2 of the median of read density. Green and red indicate the positive and negative strands, respectively. Note, no FV3 transcript read was obtained from the control (Ctrl) mock-infected kidney, and the full coverages of both positive and negative reads on the FV3 genome in the infected kidney. (B) Family-wide transcriptomic analysis of X. laevis mRNA encoding interferon receptor subunits for type I (ifnar1/2), II (ifngr1/2), and III (ifnlr1/il10rb) IFNs to show the differential expression of these IFN receptor genes in the kidney (Blue bars against the left Axis for FPKM, Fragments Per Kilobase of transcript per Million mapped reads) and regulated expression in FV3-infected kidney (Orange bars against the right Axis for Log2 fold changes). Note the significant reduction of the beta-subunits of type II and type III IFN receptors (indicated by red arrows), which may putatively result from a higher enrichment of the intergenic miRNA species as shown in
Table 1
. *p (FDR) < 0.05 relative to the control, n = 5.
|
|
Figure 8. Transcriptomic comparison of the viral genome and X. laevis mRNA encoding interferon receptor subunits in the mock, FV3-Δ64R, and FV3-WT infected intestine (A) and thymus (B). The distribution plots of mapped reads alone FV3 genome (GenBank Accession No. NC_005946.1) were shown as in
Figure 3
. Partial coverages of the viral genome were determined forFV3-Δ64R-infected intestine, and for both FV3-WT and FV3-Δ64R in the infected thymus. Comparative alignments showed that transcripts of some ORFs and miRNA-enriched intergenic regions were defective (labeled and framed using blue line) as compared between two virus strains. Red arrows indicate further repression of some IFN-receptor genes corresponding to potential higher expression of respective miRNA by FV3-WT in the intestine. The putative miRNA-mediated repression of IFN receptor genes is not detected in FV3-Δ64R infected thymus. Abbreviations and gene accession numbers are listed in
Table 1
. *p (FDR) < 0.05 relative to the control, n = 5.
|
|
Figure 9. Transcriptomic comparison of the viral genome and X. laevis mRNA encoding interferon IFN regulatory factors (irf) in the mock, FV3-Δ64R and FV3-WT infected kidney (A), intestine (B), and thymus (C). The distribution plots of mapped reads alone FV3 genome (GenBank Accession No. NC_005946.1) is shown as in
Figure 3
with a full-genome coverage for the infected kidney samples. Partial coverages of the viral genome were determined in the FV3-Δ64R infected intestine and for both FV3-WT and FV3-Δ64R in the infected thymus. Comparative alignments indicates that transcripts of some ORFs and miRNA-enriched intergenic regions are defective (labeled and framed using blue line) as compared between two virus strains. Analysis shows reduced expression of IRF genes corresponding to potential higher expression of respective miRNA by FV3-Δ64R in kidney and FV3-WT in intestine (indicated by red arrows). However, miRNA-mediated reduction of IRF genes is not detected in FV3-Δ64R infected thymus. This suggests a tissue- and virus strain-dependent expression of miRNA and interference on host gene targets. Abbreviations and gene accession numbers are listed in
Table 1
. *p (FDR) < 0.05 relative to the control.
|
|
Figure 10. Examples of miRNA that are predictably targeted on 3’-UTR regions of Xenopus mRNA encoding interferon receptor subunits. (A) Hybridization of individual miRNA with its mRNA targets was performed using a program of RNAhybrid with its accompanying programs RNAcalibrate and RNAeffective as described. The hybridization structures and minimum free energy (Mfe) are given, the thresholds of Mfe was set as -28.0 kcal/mol to reflect typical Mfe of miRNA (like let7) hybridization to mRNA targets. MiRNA C-20 or AT-20, the twentieth miRNA in the C or AT groups, respectively, as illustrated in
Figure 2
and
Supplement Excel sheet
for sequence detail. (B) Functional validation using synthetic siRNA with identical sequences to the mature C-20 and AT-20 miRNAs. Synthesis of siRNAs and transfection of X. laevis A6 cells were performed as described, and gene specific RT-PCR was used to quantify the expression of target genes. Top panel: Line chart representing numbers of predicted sites targeted by miRNA on the 3’-UTR of each template target. Bottom panel: Bar chart of relative gene expression obtained with mature C-20 (gray histogram) and AT-20 (hachured histogram) miRNAs. The GenBank Accession numbers of the tested transcripts are listed in
Table 2
. *p < 0.05, n = 5 relative to the sample transfected using a scramble siRNA.
|
|
Figure 11. Examples of miRNA that are predictably targeted on 3’-UTR regions of X. laevis mRNA encoding several IRF genes. (A) Hybridization of individual miRNA with its mRNA targets was performed using a program of RNAhybrid with its accompanying programs RNAcalibrate and RNAeffective as described. The hybridization structures and minimum free energy (Mfe) are given, the thresholds of Mfe was set as -28.0 kcal/mol to reflect typical Mfe of miRNA (like let7) hybridization to mRNA targets. MiRNA C-20 and AF-8, the twentieth and eighth miRNA in the C and AF group, respectively, as illustrated in
Figure 2
and
Supplement Excel sheet
for sequence detail. (B) Schematic shows validation using synthetic siRNA with identical sequences to representative miRNAs. Synthesis of siRNAs and transfection of X. laevis A6 cells were performed as described, and gene specific RT-PCR was used to quantify the expression of target genes. Top panel: Line chart representing numbers of predicted sites targeted by miRNA AF-8 (black triangles) and C-20 (blue cicles) on the 3’-UTR of each template target. Bottom panel: Bar chart of relative gene expression obtained with C-20 (blue histogram) and AF-8 (hachured histogram) miRNAs. The GenBank Accession numbers of the tested transcripts are listed in
Table 2
. *p < 0.05, n = 5 relative to the sample transfected using a scramble siRNA.
|
