Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Mar;25(3):1541-1569.
doi: 10.1038/s44319-023-00051-z. Epub 2024 Jan 23.

Interferon induced circRNAs escape herpesvirus host shutoff and suppress lytic infection

Sarah E Dremel #  1 , Takanobu Tagawa #  1 , Vishal N Koparde  2   3 , Carmen Hernandez-Perez  1 , Jesse H Arbuckle  4 , Thomas M Kristie  4 , Laurie T Krug  1 , Joseph M Ziegelbauer  5
Affiliations

Interferon induced circRNAs escape herpesvirus host shutoff and suppress lytic infection

Sarah E Dremel et al. EMBO Rep. 2024 Mar.

Abstract

To globally profile circRNAs, we employ RNA-Sequencing paired with chimeric junction analysis for alpha-, beta-, and gamma-herpesvirus infection. We find circRNAs are, as a population, resistant to host shutoff. We validate this observation using ectopic expression assays of human and murine herpesvirus endoribonucleases. During lytic infection, four circRNAs are commonly induced across all subfamilies of human herpesviruses, suggesting a shared mechanism of regulation. We test one such mechanism, namely how interferon-stimulation influences circRNA expression. 67 circRNAs are upregulated by either interferon-β or -γ treatment, with half of these also upregulated during lytic infection. Using gain and loss of function studies we find an interferon-stimulated circRNA, circRELL1, inhibits lytic Herpes Simplex Virus-1 infection. We previously reported circRELL1 inhibits lytic Kaposi sarcoma-associated herpesvirus infection, suggesting a pan-herpesvirus antiviral activity. We propose a two-pronged model in which interferon-stimulated genes may encode both mRNA and circRNA with antiviral activity. This is critical in cases of host shutoff, such as alpha- and gamma-herpesvirus infection, where the mRNA products are degraded but circRNAs escape.

Keywords: Circular RNAs; Herpesviruses; Host Shutoff; Interferon-stimulated Genes.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1. Human circRNAs upregulated in de novo lytic infection models.
(A) Infographic for infection models used in this study. (B) Overlap of differentially expressed circRNAs (DECs) detected by bulk RNA-Seq from HSV-1 (n = 2–4), HCMV (n = 2), and KSHV (n = 2) infection. (C) Heatmaps for DECs which overlap between viruses, with DEC clusters indicating which model the circRNA was found to be significantly upregulated within. Data is plotted as circRNA counts (log2FC), Gene counts (log2FC), or CIRCscore (circFPB (fragments per billion mapped bases)/linearFPB). Data Information: Bulk RNA-Seq data was normalized to ERCC spike-in controls. Log2FC is relative to a paired uninfected control. Differential expression p-values were calculated using RankProd non-parametric permutation tests. DECs were those with raw back splice junction (BSJ) count across the sample set >10, log2FC > 0.5 or <−0.5, and p-value < 0.05. Heatmap values are the average of biological replicates.
Figure 2
Figure 2. Global distribution shifts for mRNA, lncRNA, and circRNA during lytic infection.
(A–C) Bulk RNA-Seq data from HSV-1 infection (MRC-5 infected with KOS MOI 10, n = 2–4), KSHV lytic reactivation (iSLK-BAC16 induced with 1 µg/mL doxycycline 1 mM sodium butyrate, n = 4), and MHV68 infection (3T3 infected with H2B-YFP MOI 5, n = 2). (A) Data for protein-coding genes plotted as relative frequency distribution for log10 ERCC normalized reads or log2FC (limited to top 10,000 most highly expressed genes). Log2FC for representative protein-coding genes is plotted in column bar graphs. (B) Data for lncRNA genes plotted as relative frequency distribution for log10 normalized reads or log2FC (limited to top 100 most highly expressed genes). Log2FC for representative lncRNA genes is plotted in column bar graphs. (C) Data for circRNA plotted as relative frequency distribution for log10 normalized BSJ reads or log2FC (limited to top 100 most highly expressed circRNAs). Log2FC for representative circRNAs is plotted in column bar graphs. Data Information: Bulk RNA-Seq data was normalized to ERCC spike-in controls, log2FC is relative to a paired uninfected or uninduced control. Frequency distribution plots are the average of biological replicates. In column bar graphs, data points are biological replicates, column bars are the average, and error bars are standard deviation.
Figure 3
Figure 3. CircRNA are resistant to viral endonuclease-mediated decay.
(A) Bulk RNA-Seq data from HSV-1 (MRC-5 infected with KOS MOI 10 for 12 h, n = 4), KSHV (iSLK-BAC16 reactivated with Dox/NaB for 3 days, n = 4), and MHV68 (3T3 infected with H2B-YFP MOI 5 for 18 h, n = 2) infection. Graphs are limited to genes where raw BSJ and forward splice junction (FSJ) counts were >1 across all biological replicates. Each dot is the average log2FC of biological replicates, with each dot representing linearFPB and circFPB for a distinct gene. (B) Ectopic expression assays of viral endonucleases (vhs, BGLF5, SOX, muSOX) versus control (GFP) vector in 293T cells. RNA was collected after 24 h and transcripts quantified by qPCR (n = 3). Data is plotted as relative expression (ddCt) using 18S rRNA as the reference gene, and relative to a paired GFP transfected sample. Data points are biological replicates, column bars are the average, and error bars are standard deviation. Data Information: Paired t-tests were performed comparing circRNA and colinear mRNA expression, p-values < 0.1 are labeled and not significant (n.s.) indicates p-values ≥ 0.1. Source data are available online for this figure.
Figure 4
Figure 4. Detection of interferon-stimulated circRNAs (ISCs).
(A–E) Bulk RNA-Seq data from MRC-5, LEC, or Akata- cells were treated with recombinant interferons for 48 h (n = 3). MRC-5 and LEC were treated with IFN-β and -γ (25 ng/mL conc). Akata- were treated with IFN-β (10 ng/mL). (A,C) Volcano plots for normalized protein-coding gene (mRNA) or circRNA reads. (B,D) Venn diagrams of significantly upregulated circRNA or mRNAs. (E) Overlap of circRNAs upregulated during herpesvirus infection (Fig. 1) or interferon-stimulation. Data Information: Bulk RNA-Seq data was normalized to ERCC spike-in controls. Log2FC was calculated relative to a paired untreated sample. Adjusted p-values (Benjamini and Hochberg) were calculated using EdgeR quasi-likelihood F test method. Differentially expressed genes or circRNAs were those with raw counts across the sample set >10, log2FC > 0.5, and p-value < 0.05.
Figure 5
Figure 5. circRELL1 (hsa_circ_0001400) restricts HSV-1 lytic infection.
(A–F) circRELL1 was depleted in MRC-5 cells using siRNAs for 48 h and subsequently infected with HSV-1 strain KOS at MOI of 0.1 or 10 PFU/cell. Data is relative to a paired Non-Targeting Control siRNA (NTC). (G,H) MRC-5 cells were infected with a lentivirus expressing circRELL1 for 48 h and subsequently infected with HSV-1 strain KOS at MOI of 10 for 12 h. Data is relative to a control lentivirus expressing circGFP. (A,D,G) RNA was collected from the cell fraction and reverse transcribed. qPCR data is plotted as relative expression (ddCt) using 18 S rRNA as the reference gene (n = 2–3). (B,E) DNA was isolated from the cell fraction and assessed by qPCR for viral and host genome copies (n = 3). (C,F,H) Supernatant was collected at 12 hpi and assessed by plaque assay. Data points (C,H n = 3; F n = 6) are biological replicates, column bars are the average, and error bars are standard deviation. Data Information: Paired two-tailed t-tests were performed, p-values < 0.1 are labeled and not significant (n.s.) indicates p-values ≥ 0.1. Source data are available online for this figure.
Figure 6
Figure 6. Proposed polycistronic model for interferon-stimulated genes.
We propose a polycistronic model in which interferon-stimulated genes can produce both mRNA and circRNA with antiviral activity. This is critical in cases of host shut off, such as alpha- and gamma-herpesvirus infection, where the mRNA product is degraded but circRNA escapes. The interferon-stimulated circRNA, circRELL1, exemplifies this model. EBV, KSHV, and HCMV infection upregulates circRELL1 expression which functions to suppress lytic infection of HSV-1 and KSHV.
Figure EV1
Figure EV1. Murine circRNAs upregulated in HSV-1 infection models.
(A) Infographic for murine HSV-1 infection models used in this study. (B) Overlap of upregulated murine circRNA detected by bulk RNA-Seq from HSV-1 latency (TG, n = 4), explant-induced reactivation (TG explant, n = 3), and drug-enhanced reactivation (TG explant + JQ1, n = 2). (C) Heatmaps for upregulated murine circRNAs, plotted as circRNA counts (log2FC), Gene counts (log2FC), or CIRCScore (circFPB/linearFPB). Data Information: Bulk RNA-Seq data was normalized to ERCC spike-in controls. Log2FC is relative to a paired uninfected control. Differential expression p-values were calculated using RankProd non-parametric permutation tests. DECs were those with raw back splice junction (BSJ) count across the sample set >10, log2FC > 0.5, and p-value < 0.05. Heatmap values are the average of biological replicates.
Figure EV2
Figure EV2. Murine circRNAs upregulated in MHV68 infection models.
(A) Infographic for murine MHV68 infection models used in this study. (B) Overlap of upregulated murine circRNA detected by bulk RNA-Seq from MHV68 de novo infection (NIH 3T3, n = 2) and drug-induced reactivation (HE-RIT, n = 2). (C) Heatmaps for upregulated murine circRNAs, plotted as circRNA counts (log2FC), Gene counts (log2FC), or CIRCScore (circFPB/linearFPB). Bulk RNA-Seq data was normalized to ERCC spike-in controls. Log2FC is relative a paired uninfected or uninduced control. Differential expression p-values were calculated using RankProd non-parametric permutation tests. DECs were those with raw back splice junction (BSJ) count across the sample set >10, log2FC > 0.5, and p-value < 0.05. Heatmap values are the average of biological replicates.
Figure EV3
Figure EV3. Screenshot of interactive resource table for transcriptomic data.
(A) Representative screenshot of the GenePlotter tab where users can query any gene of interest. Supporting datasets were generated for all bulk RNA-Seq data analyzed in this study. Datasets include raw and normalized gene counts. (B) Representative screenshot of the CircPlotter tab where users can query any circRNA of interest. Supporting datasets were generated for all bulk RNA-Seq data analyzed in this study. Datasets include raw and normalized BSJ counts.
Figure EV4
Figure EV4. CircRELL1 expression changes in response to immune stimuli.
(A) Transcript quantitation for RNA collected from fibroblasts (MRC-5), lymphatic endothelial cells (LEC), or B-cell lymphoma cells (Akata-, BJAB, Daudi) treated with lipopolysaccharide (LPS), poly I:C, or CpG DNA (n = 3). (B) Transcript quantitation for RNA collected from MRC-5, LEC, or B-cell lymphoma cells treated with IFN-β and -γ (n = 3). Data Information: qPCR data is plotted as relative expression (ddCt) using 18S rRNA as the reference gene, and relative to a paired untreated sample. Data points are biological replicates, column bars are the average, and error bars are standard deviation.
Figure EV5
Figure EV5. Transcript isoform analysis of interferon-stimulated cells.
(A) Bulk RNA-Seq data from Fig. 4 plotting log2FC (+IFN/untreated) for DECs relative to their colinear gene reads. Genes with similar levels of upregulation for circRNA and colinear gene species are labeled. (B) Bulk RNA-Seq data from Fig. 4 plotted as heatmaps showing log2FC (+IFN/untreated) for circRNA species and their colinear gene or transcript isoforms. Gene and transcript quantitation was performed using Salmon and a reference transcriptome fasta. The sample (+IFN-β or +IFN-γ) where a differentially expressed circRNA (DEC) was identified are labeled on the left of each heatmaps as “DEC cluster”. Data Information: Bulk RNA-Seq data was normalized to ERCC spike-in control and log2FC was determined relative to a paired untreated sample. All data plotted is the average of biological replicates (n = 3). The relationship between circRNA and colinear gene expression changes was assessed by linear regression analysis, R2 values are given.
See this image and copyright information in PMC

Update of

Similar articles

See all similar articles

References

    1. Abere B, Zhou H, Li J, Cao S, Toptan T, Grundhoff A, Fischer N, Moore PS, Chang Y, Racaniello VR. Merkel cell polyomavirus encodes circular RNAs (circRNAs) enabling a dynamic circRNA/microRNA/mRNA regulatory network. mBio. 2020;11:e03059–03020. doi: 10.1128/mBio.03059-20. - DOI - PMC - PubMed
    1. Abernathy E, Clyde K, Yeasmin R, Krug LT, Burlingame A, Coscoy L, Glaunsinger B. Gammaherpesviral gene expression and virion composition are broadly controlled by accelerated mRNA degradation. PLOS Pathog. 2014;10:e1003882. doi: 10.1371/journal.ppat.1003882. - DOI - PMC - PubMed
    1. Abernathy E, Glaunsinger B. Emerging roles for RNA degradation in viral replication and antiviral defense. Virology. 2015;479-480:600–608. doi: 10.1016/j.virol.2015.02.007. - DOI - PMC - PubMed
    1. Ablashi DV, Berneman ZN, Kramarsky B, Whitman J, Asano Y, Pearson GR. Human herpesvirus-7 (HHV-7): current status. Clin Diagn Virol. 1995;4:1–13. doi: 10.1016/0928-0197(95)00005-S. - DOI - PubMed
    1. Abrisch RG, Eidem TM, Yakovchuk P, Kugel JF, Goodrich JA. Infection by herpes simplex virus 1 causes near-complete loss of RNA polymerase II occupancy on the host cell genome. J Virol. 2015;90:2503–2513. doi: 10.1128/JVI.02665-15. - DOI - PMC - PubMed
-