The Oncogenic Kaposi's Sarcoma-Associated Herpesvirus Encodes a Mimic of the Tumor-Suppressive miR-15/16 miRNA Family

Abstract

Many tumor viruses encode oncogenes of cellular origin. Here, we report an oncoviral mimic of a cellular tumor suppressor. The Kaposi's sarcoma-associated herpesvirus (KSHV) microRNA (miRNA) miR-K6-5p shares sequence similarity to the tumor-suppressive cellular miR-15/16 miRNA family. We show that miR-K6-5p inhibits cell cycle progression, a hallmark function of miR-16. miR-K6-5p regulates conserved miR-15/16 family miRNA targets, including many cell cycle regulators. Inhibition of miR-K6-5p in KSHV-transformed B cells confers a significant growth advantage. Altogether, our data show that KSHV encodes a functional mimic of miR-15/16 family miRNAs. While it is exceedingly well established that oncogenic viruses encode oncogenes of cellular origin, this is an unusual example of an oncogenic virus that encodes a viral mimic of a cellular tumor suppressor. Encoding a tumor-suppressive miRNA could help KSHV balance viral oncogene expression and thereby avoid severe pathogenesis in the healthy host.

Keywords: KSHV; miR-15/16; microRNA; tumor virus.

Figure 1.. KSHV miR-K6-5p Mimics miR-16-Induced Cell Cycle Arrest

(A) Sequences of miR-K6-5p, the miR-15/16 family miRNAs, and miR-214. miRNA seed sequences (nucleotides 2–7) are in red. (B and C) Primary lymphatic endothelial cells (LEC) were transfected with mimics of miR-16, miR-K6-5p, or a negative control (ctrl) and subjected to growth curve analyses (B) or cell cycle analysis by propidium iodide (PI) staining on day 2 after transfection (C). n = 3. (D and E) JSC-1 cells were transfected with miRNA mimics and subjected to growth curve analysis (D) or cell cycle analysis by anti-BrdU/PI staining on day 2 after transfection (E). n = 3. *p < 0.05, **p < 0.01, ***p < 0.001. Data are represented as mean ± SEM. See also Figure S1.

Figure 2.. miR-K6-5p Mimics miR-16-Induced Gene Expression Changes

(A) Principle component analysis of mRNA-seq data in 293T/NoDice. (B) Pearson’s correlation compares mRNA log2 fold changes caused by miR-16 or miR-K6-5p. (C) Pearson’s coefficients from other comparisons in the mRNA-seq dataset, as in (B). (D–F) Cumulative distribution frequency (CDF) plots depicting regulation of the top 250 TargetScan-predicted targets of the listed miRNAs by mimics of miR-16(D), miR-214 (E), or miR-K6-5p(F) in the mRNA-seq data. Numbers in parentheses are gene set sizes and p values for comparisons to all mRNAs, calculated using 2-sample Kolmogorov-Smirnov (K-S) tests. (G) Heatmap showing Z scores for mRNAs among the top 250 TargetScan-predicted miR-15/16 family miRNA targets that contribute to the enrichment in cell-cycle-related categories detected by DAVID. See also Figure S2 and Tables S1 and S2.

Figure 3.. miR-K6-5p Regulates Target mRNAs via miR-16 Binding Sites

(A) Diagram of canonical target sites (green) expected to be shared or preferential for miR-16 and miR-K6-5p. Nucleotides 2–7 seed sequences are in red. V denotes A, C, or G. (B) The miR-16 binding site in a firefly (FLuc) luciferase 3′UTR reporter vector for BCL2 was mutated and tested for regulation by mimics of miR-16 or miR-K6-5p in dual luciferase reporter assays in 293T/NoDice cells. Fluc data were sequentially normalized to data from a co-transfected Renilla luciferase (RLuc) control, the empty FLuc vector, and negative control mimic (ctrl). n = 3. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data are represented as mean ± SEM. See Figure S3A. (C) Heatmap showing Z scores for mRNAs that represent different types of targets of miR-16 and/or miR-K6-5p and were chosen for validation experiments presented in Figures S3B–S3M.

Figure 4.. miR-K6-5p Confers a Competitive Disadvantage in the KSHV-Transformed PEL Cell Line BC-3

(A) Proportion of small RNA reads for miR-15/16 family miRNAs and miR-K6-5p in small RNA sequencing datasets from the PEL cell lines BC-3, BC-1, and BCBL-1 (Gottwein et al., 2011). (B) Western blots for KSHV vCyc, LANA, vIRF3, and for GAPDH in the same PEL cell lines and KSHV-negative Ramos cells. Representative of n > 3. (C) Diagram of the lentiviral miR-K6-5p-sponge (8SK6-5p). (D) Diagram of experimental design for competition assay in (F). (E) TaqMan qRT-PCR to assess miR-16 or miR-K6-5p expression in sponge-transduced and sorted BC-3 cells. n = 3. (F) Results from competition experiments in BC-3. n = 5. Throughout the figure, *p < 0.05, **p < 0.001. Data are represented as mean ± SEM.