25-Hydroxycholesterol Inhibits Kaposi's Sarcoma Herpesvirus and Epstein-Barr Virus Infections and Activates Inflammatory Cytokine Responses

Abstract

Oncogenic gammaherpesviruses express viral products during latent and lytic infection that block the innate immune response. Previously, we found that Kaposi's sarcoma herpesvirus (KSHV/human herpesvirus-8) viral microRNAs (miRNAs) downregulate cholesterol biogenesis, and we hypothesized that this prevents the production of 25-hydroxycholesterol (25HC), a cholesterol derivative. 25HC blocks KSHV de novo infection of primary endothelial cells at a postentry step and decreases viral gene expression of LANA (latency-associated nuclear antigen) and RTA. Herein we expanded on this observation by determining transcriptomic changes associated with 25HC treatment of primary endothelial cells using RNA sequencing (RNA-Seq). We found that 25HC treatment inhibited KSHV gene expression and induced interferon-stimulated genes (ISGs) and several inflammatory cytokines (interleukin 8 [IL-8], IL-1α). Some 25HC-induced genes were partially responsible for the broadly antiviral effect of 25HC against several viruses. Additionally, we found that 25HC inhibited infection of primary B cells by a related oncogenic virus, Epstein-Barr virus (EBV/human herpesvirus-4) by suppressing key viral genes such as LMP-1 and inducing apoptosis. RNA-Seq analysis revealed that IL-1 and IL-8 pathways were induced by 25HC in both primary endothelial cells and B cells. We also found that the gene encoding cholesterol 25-hydroxylase (CH25H), which converts cholesterol to 25HC, can be induced by type I interferon (IFN) in human B cell-enriched peripheral blood mononuclear cells (PBMCs). We propose a model wherein viral miRNAs target the cholesterol pathway to prevent 25HC production and subsequent induction of antiviral ISGs. Together, these results answer some important questions about a widely acting antiviral (25HC), with implications for multiple viral and bacterial infections. IMPORTANCE A cholesterol derivative, 25-hydroxycholesterol (25HC), has been demonstrated to inhibit infections from widely different bacteria and viruses, including severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, its mechanism of activity is still not fully understood. In this work, we look at gene expression changes in the host and virus after 25HC treatment to find clues about its antiviral activity. We likewise demonstrate that 25HC is also antiviral against EBV, a common cancer-causing virus. We compared our results with previous data from antiviral screening assays and found the same pathways resulting in antiviral activity. Together, these results bring us closer to understanding how a modified form of cholesterol works against several viruses.

Keywords: EBV; KSHV; cholesterol; innate immunity.

FIG 1

25HC suppressed viral transcription during KSHV de novo infection in primary endothelial cells. (A) Experimental outline of 25HC treatment and KSHV de novo infection of human umbilical vein endothelial cells (HUVECs). (B) Volcano plot showing repressed KSHV differentially expressed genes (DEGs) in HUVECs at 2 dpi with or without 25HC (+/−25HC) (n = 4). DEGs with log₂ fold change (FC) of >1.5 and P value (pval) of <0.01 are indicated in red dots to the left of the vertical dotted line and above the horizontal dotted line. This plot was generated using the EnhancedVolcano package from Bioconductor using viral DEGs (see Table S1 in the supplemental material) as input. (C) Heat map of viral transcript reads (average of four biological replicates) normalized to transcripts per million (TPM) with and without 25HC at 2 dpi. Viral transcripts are classified as immediate early (IE), early, delayed early, and late transcripts. Latency transcripts are annotated with the dark red bars on the right.

FIG 2

25HC induced an inflammatory response. (A) Heat map of top 500 genes by variance, with each column representing a biological replicate with the designated infection and treatment. (B) Volcano plot of host DEGs in infected HUVECs with (+) or without (−) 25HC. (C) Ingenuity Pathway Analysis (Qiagen) of differential gene expression in infected HUVECs with or without 25HC with corresponding Z-score and P value of cholesterol-related pathways (orange text) and inflammatory pathways (green text). (D) Validation of uninfected RNA samples for gene expression changes by qPCR relative to Vehicle control samples (n = 4). Unpaired t test was performed. (E) Quantitation of secreted cytokines in the supernatant after 25HC treatment of HUVECs (n = 3). Statistical analysis using an unpaired t test was performed, and P values are indicated as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. hpt, hours posttreatment.

FIG 3

IL-8, IL-1α, and FOSB contribute to 25HC antiviral activity against KSHV de novo infection. (A) Schematic diagram of experiment with siRNA depletion, treatment (with or without 25HC), and KSHV de novo infection of HUVECs. (B) Knockdown by siRNA of candidate target genes on the day of infection. A statistical test using one-way analysis of variance (ANOVA) and post hoc test with Bonferroni correction was performed. (C) Viral gene expression of LANA after siRNA depletion of candidate target genes (n = 4 to 6) and with control samples (infection only, n = 5). One-way ANOVA and post hoc t test with Bonferroni correction were performed, and P values are indicated as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P < 0.0001.

FIG 4

25HC is antiviral against EBV de novo infection. (A) Schematic diagram of experimental outline showing B cell enrichment of PBMCs and pretreatment (pre-tx) with 25HC, followed by EBV infection. Flow cytometry (Flow cyto), RNA sampling, and addition of 25HC are seen at the indicated time points. (B) Flow cytometry of pretreated and infected B cell-enriched PBMCs to measure cell viability (left) and apoptosis rate (right). Unpaired t test was performed for each time point. (C) Flow cytometry of B cell-enriched PBMCs treated with 25HC after EBV infection and measured for cell viability (left) and apoptosis (right). Unpaired t test was performed for each time point. P values are as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 5

25HC blocks EBV-driven transformation in B cell-enriched PBMCs. (A) Scatterplot of flow cytometry analysis of PBMCs with annexin V and 7-AAD staining after EBV infection and pretreatment with either vehicle (left) or 25HC (right). (B) Scatterplot and gating strategy of PBMCs infected with EBV. (C) Based on annexin V and 7AAD staining, cells in the “LCL” and “PBMC” gates were designated healthy or apoptotic subsets. These subsets were quantified as a percentage of total cells in the designated gates. Percentages are averages from three biological replicates of pretreated PBMCs. Statistical analysis using an unpaired t test was performed for each subgroup, P values are as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 6

25HC selectively suppressed some viral transcripts in EBV de novo infection. (A) Volcano plot showing repressed EBV differentially expressed genes (DEGs) in MACSExpress purified B cells at 10 to 12 dpi with or without 25HC (n = 4). DEGs with log₂ fold change of >1.5 and P value of <0.01 are indicated in red dots to the left of the vertical dotted line and above the horizontal dotted line. (B) Heat map of EBV viral transcript reads (average of four biological replicates) normalized to transcripts per million (TPM) with and without 25HC at 10 to 12 dpi. (C) Histogram of Integrated Genome Viewer showing total viral reads at indicated loci in the EBV genome. (D) Validation of viral transcript levels using RT-qPCR. Statistical analysis using unpaired t test was performed, and P values are indicated as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 7

25HC targets LMP-1-related host gene expression and upregulates cytokine pathways in the context of EBV de novo infection. (A) Heat map of top 500 host genes by variance in purified B cells during EBV de novo infection with or without 25HC at 10 to 12 dpi (n = 4). (B) Volcano plot of host DEGs in the context of EBV de novo infection of purified B cells with or without 25HC at 10 to 12 dpi. (C) Ingenuity Pathway Analysis of host DEGs (cholesterol and mevalonate-related genes [orange text]; cytokine and inflammation pathways [green text]). (D) Validation of candidate host transcript levels using RT-qPCR. Statistical analysis using unpaired t test was performed, and P values are indicated as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 8

CH25H is induced in PBMCs by type I interferons. (A and B) Time course experiment of CH25H gene expression in PBMCs relative to vehicle control (VehCtrl) samples after IFN-α (A) and IFN-β (B) treatment and assayed with RT- qPCR. (C and D) Corresponding time course experiment of ATF3 gene expression in PBMC after IFN-α (C) and IFN-β (D) treatment and assayed with RT-qPCR. (E) Schematic diagram of antiviral activity of 25HC and proposed modulators of CH25H expression.