Functions of KSHV microRNAs and circular RNAs

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

We have discovered and published around 40 validated cellular mRNA targets of KSHV miRNAs. Importantly, many of these targets have not previously been studied in the context of KSHV infection. Any given miRNA is repressing not one, but multiple genes at the same time. Our general question specific aim is which miRNA targets are functionally significant for a given phenotype of interest. To answer this goal, we have designed assays to repress around 60 miRNA targets with multiple siRNAs (to control for off-target effects). We have completed the first rounds of screening miRNA targets for their roles in changing a type of cell behavior that is observed in KSHV-infected cells. The specific objectives of this project are to understand how specific microRNA targets are more important than other targets for controlling cell behavior and we have accomplished this objective. We are continuing with a more complicated screen to answer additional questions. We have been continuing aims (1) to understand how viral infection influences microRNA biosynthesis and degradation (2) how regulation of the cholesterol biosynthesis pathway by viral microRNAs. Additional manuscripts on these aims will be submitted in 2019. A new aim has been to investigate a new class of non-coding RNAs called circular RNAs in the context of KSHV infection. We began this circular RNA project using various methods to identify human circular RNAs whose levels are changed upon KSHV infection. Our first hypothesis was that specific human circular RNAs may function to repress KSHV miRNA activity. In this case, we further hypothesized that potential antiviral circular RNAs may be induced upon infection with KSHV. To measure expression changes with KSHV infections, we used two methods. First, we used new microarrays that measure expression of 5,396 human circular RNAs. For these experiments, we used three biological replicates from mock-infected or KSHV-infected primary endothelial cells (HUVECs) or a B cell line, MC116. Previous reports demonstrated that human circular RNAs can inhibit human miRNAs by acting as a decoy or sponge to bind to miRNAs and prevent the miRNAs from targeting mRNAs for repression. Given this information, we are initially interested in circular RNAs that meet three characteristics: 1. Circular RNAs that are abundant, since highly expressed circular RNAs are likely to serve better as miRNA inhibitors. 2. Circular RNAs that contain predicted KSHV miRNA binding sites. 3. Circular RNAs that are elevated in KSHV-infected cells compared to control cells. We found dozens of circular RNAs that meet these characteristics. In addition to using microarrays, we also used RNA-sequencing (RNA-seq) to measure circular RNA changes in the same RNA samples and to potentially discover new circular RNAs that were not represented on the microarrays. These RNA samples were treated in two ways before RNA-sequencing: mock treatment, or treatment with RNase R, which cleaves linear mRNAs, but does not cleave circular RNAs. This RNase R treatment method has been used by another group to enrich for circular RNAs. These RNA samples were then sequenced using stranded, paired-end, long reads and analyzed for linear RNA (mRNA, lncRNA, etc.) expression changes, as well as changes in circular RNAs. We have integrated our data from microarrays, RNA-seq, and predicted miRNA binding site information to arrive at a handful of newly identified circular RNAs. We analyzed these specific human circular RNAs first, as they meet the three criteria mentioned above and may represent human circular RNAs that may function to inhibit viral miRNA function. We used specific PCR primers that will only amplify the circular RNAs (not linear mRNA) and verified our genomic expression profiling results with these qPCR assays. Additionally, through collaboration with Dr. Mileidy Gonzalez and Dr. John Spouge in the Computational Biology Branch, National Library of Medicine, we have identified specific circular RNAs that are enriched for predicted KSHV miRNA target sites. They developed new software to test the enrichment of KSHV miRNA binding sites in human circular RNAs and found multiple examples of enrichment of KSHV miRNA targets sites, with statistical significance. These findings suggest that the density of these potential KSHV miRNA targets is very unlikely to be a result of random sequences in certain circular RNAs. These human circular RNAs could represent a new type of innate immunity, which may have arisen from evolutionary pressure for human adaptation to chronic viral herpesvirus infections throughout human evolution. We initially focused on one human circular RNAs, called hsa_circ_0001400, because expression of it increased with KSHV infection, it was abundant in relevant cell types for KSHV infections, and it contained a predicted binding site for a KSHV miRNA. We found that transient and stable expression of hsa_circ_0001400 inhibited expression of two KSHV genes, but did not decrease intracellular levels of KSHV DNA. This suggested that hsa_circ_0001400 could inhibit KSHV infection at a post-entry step of infection. We also answered whether KSHV encodes viral circular RNAs using the same methods used to identify human circular RNAs in our RNA sequencing data. We discovered and validated that at least seven circular RNAs are expressed by KSHV. We found that three of these KSHV circular RNAs can change cell growth kinetics. We also found that these KSHV circular RNAs were strongly induced in the lytic cycle of infection and that these circular RNAs are detected in biopsies from patients with KSHV infections. Understanding the mechanisms of how these circular RNAs modify cell behavior and infection fits into our goal of researching how non-coding RNAs regulate host-virus interactions.