Date Published: December 26, 2013
Publisher: Public Library of Science
Author(s): Jon Pavelin, Natalie Reynolds, Stephen Chiweshe, Guanming Wu, Rebecca Tiribassi, Finn Grey, William J. Britt.
Recent advances in microRNA target identification have greatly increased the number of putative targets of viral microRNAs. However, it is still unclear whether all targets identified are biologically relevant. Here, we use a combined approach of RISC immunoprecipitation and focused siRNA screening to identify targets of HCMV encoded human cytomegalovirus that play an important role in the biology of the virus. Using both a laboratory and clinical strain of human cytomegalovirus, we identify over 200 putative targets of human cytomegalovirus microRNAs following infection of fibroblast cells. By comparing RISC-IP profiles of miRNA knockout viruses, we have resolved specific interactions between human cytomegalovirus miRNAs and the top candidate target transcripts and validated regulation by western blot analysis and luciferase assay. Crucially we demonstrate that miRNA target genes play important roles in the biology of human cytomegalovirus as siRNA knockdown results in marked effects on virus replication. The most striking phenotype followed knockdown of the top target ATP6V0C, which is required for endosomal acidification. siRNA knockdown of ATP6V0C resulted in almost complete loss of infectious virus production, suggesting that an HCMV microRNA targets a crucial cellular factor required for virus replication. This study greatly increases the number of identified targets of human cytomegalovirus microRNAs and demonstrates the effective use of combined miRNA target identification and focused siRNA screening for identifying novel host virus interactions.
Human cytomegalovirus (HCMV) is a highly prevalent infectious disease, infecting greater than 30% of the population. Although normally asymptomatic in healthy individuals, HCMV infection is a significant cause of morbidity and mortality in immunocompromised populations, individuals with heart disease and recipients of solid organ and bone marrow transplants –. HCMV is also the leading cause of infectious congenital birth defects resulting from spread of the virus to the unborn fetus. Reactivation of virus from a latent infection, rather than primary infection, is often responsible for HCMV associated pathologies –.
RISC-IP techniques have recently been used to identify targets of viral miRNAs , . The approach relies on the stable interaction of the miRNA associated RISC protein complex with the targeted transcript. Following lysis of cells the RISC complexes are immunoprecipitated using direct antibodies that recognize Argonaute 2. RNA is then isolated, labeled and analysed by microarray to identify transcripts which are significantly enriched due to miRNA targeting. In a previous study we used this technique to identify targets of a single miRNA, US25-1, in the context of HEK293 cells . Here we used the same basic approach to identify targets in primary human fibroblast cells infected with either the laboratory adapted AD169 strain of HCMV or the clinical strain TR. In both cases cells were infected at a high multiplicity of infection (MOI) of three and cells harvested three days post infection. Following lysis and immunoprecipitation, RNA was isolated by trizol extraction and analysed by microarray using the Illumina HumanRef-8 platform which contains probes for approximately 24,000 well annotated genes (Figure 1A). In addition to uninfected cells, RNA was analysed from infected cells immunoprecipitated with pre-immune serum instead of anti-Argonaute 2 antibody. To determine the level of enrichment, lysate was sampled before immunoprecipitation to establish total levels of transcript expression. Enrichment was then calculated as the transcript level of the IP sample divided by the total RNA sample. To determine transcripts specifically targeted by HCMV miRNAs, the level of enrichment from infected samples was divided by the level of enrichment in uninfected samples. However, infection with HCMV results in significant perturbation of total levels of many cellular transcripts through mechanisms unrelated to miRNA expression. False positive enrichment attributed to viral miRNA targeting can therefore occur due to down regulation of total RNA levels in the infected sample, where IP background levels remain relatively unchanged. To overcome this, a correction calculation was introduced using the results from the control IP using the pre-immune serum pull down. As enrichment values in this sample would only occur through changes in total levels due to AD169 infection, rather than any effective enrichment through specific immunoprecipitation, false enrichment could be effectively subtracted from the data sets generated with anti-Argonaute 2 pull downs. Example calculations are shown in supplemental figure S1.
Despite recent advances in our understanding of miRNA transcript interaction, identification of valid targets remains challenging. The nature of miRNA targeting, where functional effects may rely on multiple miRNAs targeting a single transcript or multiple genes within single pathways being targeted, requires a system wide approach to elucidate the functions of miRNAs. Recent studies have used such approaches to identify targets of gamma-herpesvirus miRNAs , –. However, no systematic screening approach has been presented for HCMV in the context of viral infection. Here we use a RISC-IP approach to identify putative targets of HCMV miRNAs in the context of viral infection, an important step towards generating a global understanding of the role these small regulatory RNAs play in the biology of HCMV and herpes viruses in general.