Characterization of herpes simplex virus 2 primary microRNA Transcript regulation

Models of Herpes Simplex Virus Latency

Our current understanding of HSV latency is based on a variety of clinical observations, and in vivo, ex vivo, and in vitro model systems, each with unique advantages and drawbacks. The criteria for authentically modeling HSV latency include the ability to easily manipulate host genetics and biological pathways, as well as mimicking the immune response and viral pathogenesis in human infections. Although realistically modeling HSV latency is necessary when choosing a model, the cost, time requirement, ethical constraints, and reagent availability are also equally important. Presently, there remains a pressing need for in vivo models that more closely recapitulate human HSV infection. While the current in vivo, ex vivo, and in vitro models used to study HSV latency have limitations, they provide further insights that add to our understanding of latency. In vivo models have shed light on natural infection routes and the interplay between the host immune response and the virus during latency, while in vitro models have been invaluable in elucidating molecular pathways involved in latency. Below, we review the relative advantages and disadvantages of current HSV models and highlight insights gained through each.

MicroRNA Regulation of Human Herpesvirus Latency

Herpesviruses are ubiquitous human pathogens. After productive (lytic) infection, all human herpesviruses are able to establish life-long latent infection and reactivate from it. Latent infection entails suppression of viral replication, maintenance of the viral genome in infected cells, and the ability to reactivate. Most human herpesviruses encode microRNAs (miRNAs) that regulate these processes during latency. Meanwhile, cellular miRNAs are hijacked by herpesviruses to participate in these processes. The viral or cellular miRNAs either directly target viral transcripts or indirectly affect viral infection through host pathways. These findings shed light on the molecular determinants that control the lytic-latent switch and may lead to novel therapeutics targeting latent infection. We discuss the multiple mechanisms by which miRNAs regulate herpesvirus latency, focusing on the patterns in these mechanisms.

Identification of miRNAs encoded by Autographa californica nucleopolyhedrovirus

Two Autographa californica nucleopolyhedrovirus (AcMNPV) encoded miRNAs, AcMNPV-miR-1 and AcMNPV-miR-3, have been reported by us in 2013 and 2019, respectively. Here, we present an integrated investigation of AcMNPV-encoded miRNAs, which include the above two miRNAs and three additional newly identified miRNAs. Six candidate miRNAs were predicted through small RNA deep sequencing and bioinformatics, of which, five were validated. Three miRNAs are located opposite the coding sequences, the other two are located in the coding sequences of viral genes. Targets in both virus and host were predicted and subsequently tested using dual-luciferase reporter assays. The validated targets were found mainly in AcMNPV, except for the targets of AcMNPV-miR-4, which are all host genes. Based on reporter assays, the five miRNAs predominantly function by down-regulating their targets. The transcription start sites of these miRNAs were bioinformatic screened based on known baculovirus promoter motifs. Our study reveals that AcMNPV-encoded miRNAs function as fine modulators of the interactions between host and virus by regulating viral and/or host genes.

Analysis of G-quadruplexes upstream of herpesvirus miRNAs: evidence of G-quadruplex mediated regulation of KSHV miR-K12-1-9,11 cluster and HCMV miR-US33

BACKGROUND

G-quadruplexes regulate gene expression, recombination, packaging and latency in herpesviruses. Herpesvirus-encoded miRNAs have been linked to important biological functions. The presence and the biological role of G-quadruplexes have not been studied in the regulatory regions of virus miRNA. We hypothesized that herpesvirus-encoded miRNAs are regulated by G-quadruplexes in their promoters.

RESULTS

We analyzed the 1 kb regulatory regions of all herpesvirus-encoded miRNAs for the presence of putative quadruplex-forming sequences (PQS). Over two-third (67%) of the regulatory regions of herpesvirus miRNAs had atleast 1 PQS. The 200 bp region of the promoter proximal to herpesvirus miRNA is particularly enriched for PQS. We chose to study the G-quadruplex motifs in the promoters of miR-K12 cluster in Kaposi's sarcoma-associated Herpesvirus (KSHV miR-K12-1-9,11) and the miR-US33 encoded by Human Cytomegalovirus (HCMV miR-US33). Biophysical characterization indicates that the G-quadruplex motifs in the promoters of the KSHV miR-K12 cluster and the HCMV miR-US33 form stable intramolecular G-quadruplexes in vitro. Mutations disrupting the G-quadruplex motif in the promoter of the KSHV miR-K12 cluster significantly inhibits promoter activity, while those disrupting the motif in the promoter of HCMV miR-US33 significantly enhance the promoter activity as compared to that of the respective wild-type promoter. Similarly, the addition of G-quadruplex binding ligands resulted in the modulation of promoter activity of the wild-type promoters (with intact G-quadruplex) but not the mutant promoters (containing quadruplex-disrupting mutations).

CONCLUSION

Our findings highlight previously unknown mechanisms of regulation of virus-encoded miRNA and also shed light on new roles for G-quadruplexes in herpesvirus biology.

Hidden regulation of herpes simplex virus 1 pre-mRNA splicing and polyadenylation by virally encoded immediate early gene ICP27

In contrast to human cells, very few HSV-1 genes are known to be spliced, although the same pre-mRNA processing machinery is shared. Here, through global analysis of splice junctions in cells infected with HSV-1 and an HSV-1 mutant virus with deletion of infectious cell culture protein 27 (ICP27), one of two viral immediate early (IE) genes essential for viral replication, we identify hundreds of novel alternative splice junctions mapping to both previously known HSV-1 spliced genes and previously unknown spliced genes, the majority of which alter the coding potential of viral genes. Quantitative and qualitative splicing efficiency analysis of these novel alternatively spliced genes based on RNA-Seq and RT-PCR reveals that splicing at these novel splice sites is efficient only when ICP27 is absent; while in wildtype HSV-1 infected cells, the splicing of these novel splice junctions is largely silenced in a gene/sequence specific manner, suggesting that ICP27 not only promotes accumulation of ICP27 targeted transcripts but also ensures correctness of the functional coding sequences through inhibition of alternative splicing. Furthermore, ICP27 toggles expression of ICP34.5, the major viral neurovirulence factor, through inhibition of splicing and activation of a proximal polyadenylation signal (PAS) in the newly identified intron, revealing a novel regulatory mechanism for expression of a viral gene. Thus, through the viral IE protein ICP27, HSV-1 co-opts both splicing and polyadenylation machinery to achieve optimal viral gene expression during lytic infection. On the other hand, during latent infection when ICP27 is absent, HSV-1 likely takes advantages of host splicing machinery to restrict expression of randomly activated antigenic viral genes to achieve immune evasion.

Little is known regarding to how HSV, a large DNA virus and known to contain very few spliced genes, escapes host pre-mRNA splicing machinery. Here, by establishing a high throughput splice junction identification platform and quantitative analysis method to assess splicing efficiency based on high throughput data, we find that HSV-1 encodes hundreds of previously unknown alternative splice junctions; however, splicing of these novel spliced genes is largely silenced in wild-type HSV-1 infected cells, explaining why only very few spliced genes have been previously identified in HSV-1. Moreover, ICP27 is required for splicing inhibition and 3’ end formation of ICP34.5, the major viral neurovirulence factor and also the major target of latently expressed viral miRNAs. These findings not only fundamentally change the view of HSV gene structure, but also reveal a mechanism by which HSV employs host splicing and polyadenylation machineries to achieve optimal gene expression during acute infection and may also contribute to immune evasion during latency when ICP27 is not expressed.

HSV-2-encoded miRNA-H4 Regulates Cell Cycle Progression and Act-D-induced Apoptosis in HeLa Cells by Targeting CDKL2 and CDKN2A

MicroRNAs (miRNAs) encoded by latency-associated transcript are associated with both latent and acute stages of herpes simplex virus 2 (HSV-2) infection. In this study, miRNA-H4-5p and miRNA-H4-3p were ectopically expressed in HeLa cells to explore potential cellular targets of viral miRNAs and demonstrate their potential biological functions. The results showed that miRNA-H4-5p could reverse apoptosis induced by actinomycin D (Act-D) and promote cell cycle progression, but miRNA-H4-3p had no such obvious functions. Bioinformatics analysis, luciferase report assay, quantitative reverse transcription polymerase chain reaction (qRT-PCR), and Western blotting demonstrated that miRNA-H4-5p could bind to the 3'-untranslated region (UTR) of cyclin-dependent kinase inhibitor 2A (CDKN2A) and cyclin-dependent kinase-like 2 (CDKL2) to negatively regulate their expression. We verified that these two targeted genes were associated with cell apoptosis and cell cycle. Furthermore, in HeLa cells infected with HSV-2, we detected significantly reduced expression of CDKN2A and CDKL2 and demonstrated the negative regulation effect of miRNA-H4-5p on these two target genes. Our findings show that viral miRNAs play a vital role in regulating the expression of the host's cellular genes that participate in cell apoptosis and progression to reshape the cellular environment in response to HSV-2 infection, providing further information on the roles of encoded herpesvirus miRNAs in pathogen-host interaction.

Herpes Simplex Virus 1 Lytic Infection Blocks MicroRNA (miRNA) Biogenesis at the Stage of Nuclear Export of Pre-miRNAs

Various mechanisms have been identified by which viruses target host small RNA biogenesis pathways to achieve optimal infection outcomes. Herpes simplex virus 1 (HSV-1) is a ubiquitous human pathogen whose successful persistence in the host entails both productive (“lytic”) and latent infection. Although many HSV-1 miRNAs have been discovered and some are thought to help control the lytic/latent switch, little is known about regulation of their biogenesis. By characterizing expression of both pre-miRNAs and mature miRNAs under various conditions, this study revealed striking differences in miRNA biogenesis between lytic and latent infection and uncovered a regulatory mechanism that blocks pre-miRNA nuclear export and is dependent on viral protein ICP27 and viral DNA synthesis. This mechanism represents a new virus-host interaction that could limit the repressive effects of HSV-1 miRNAs hypothesized to promote latency and may shed light on the regulation of miRNA nuclear export, which has been relatively unexplored.

Herpes simplex virus 1 (HSV-1) switches between two infection programs, productive (“lytic”) and latent infection. Some HSV-1 microRNAs (miRNAs) have been hypothesized to help control this switch, and yet little is known about regulation of their expression. Using Northern blot analyses, we found that, despite inherent differences in biogenesis efficiency among six HSV-1 miRNAs, all six exhibited high pre-miRNA/miRNA ratios during lytic infection of different cell lines and, when detectable, in acutely infected mouse trigeminal ganglia. In contrast, considerably lower ratios were observed in latently infected ganglia and in cells transduced with lentiviral vectors expressing the miRNAs, suggesting that HSV-1 lytic infection blocks miRNA biogenesis. This phenomenon is not specific to viral miRNAs, as a host miRNA expressed from recombinant HSV-1 also exhibited high pre-miRNA/miRNA ratios late during lytic infection. The levels of most of the mature miRNAs remained stable during infection in the presence of actinomycin D, indicating that the high ratios are due to inefficient pre-miRNA conversion to miRNA. Cellular fractionation experiments showed that late (but not early) during infection, pre-miRNAs were enriched in the nucleus and depleted in the cytoplasm, indicating that nuclear export was blocked. A mutation eliminating ICP27 expression or addition of acyclovir reduced pre-miRNA/miRNA ratios, but mutations drastically reducing Us11 expression did not. Thus, HSV-1 lytic infection inhibits miRNA biogenesis at the step of nuclear export and does so in an ICP27- and viral DNA synthesis-dependent manner. This mechanism may benefit the virus by reducing expression of repressive miRNAs during lytic infection while permitting elevated expression during latency.

Herpes Simplex Virus 2 Latency-Associated Transcript (LAT) Region Mutations Do Not Identify a Role for LAT-Associated MicroRNAs in Viral Reactivation in Guinea Pig Genital Models

Despite the long-standing observation that herpes simplex virus (HSV) latency-associated transcript (LAT) promoter deletion viruses show impaired recurrence phenotypes in relevant animal models, the mechanism by which these sequences exert this phenotypic effect is unknown. We constructed and evaluated four mutant HSV-2 isolates with targeted mutations in the LAT promoter and LAT-associated microRNAs (miRNAs) affecting (i) the LAT TATA box; (ii) the LAT ICP4-binding site; (iii) miRNA I (miR-I) and miR-II (miR-I/II), which both target ICP34.5; and (iv) miR-III, which targets ICP0. While the LAT TATA box mutant caused milder acute infections than wild-type (WT) virus, there was no difference in the recurrence phenotype between these viruses. LAT and miRNA expression during latency was not impaired by this mutation, suggesting that other promoter elements may be more important for latent HSV-2 LAT expression. Mutation of the LAT ICP4-binding site also did not cause an in vivo phenotypic difference between mutant and WT viruses. Acute infection and reactivation from latency of the miR-I/II mutant were similar to those of its rescuant, although the acute infection was slightly reduced in severity relative to that caused by the wild-type virus. The miR-III mutant also exhibited WT phenotypes in acute and recurrent phases of infection. While they do not rule out an effect of these elements in human latency or reactivation, these findings do not identify a specific role for LAT or LAT-associated miRNAs in the HSV-2 LAT promoter deletion phenotype in guinea pigs. Thus, other sequences in this region may play a more important role in the long-studied LAT-associated phenotype in animals.IMPORTANCE While it has been known for several decades that specific HSV-1 and HSV-2 sequences near the LAT promoter are required for efficient viral reactivation in animal models, the mechanism is still not known. We constructed four mutant viruses with the goal of identifying critical sequence elements and of specifically testing the hypothesis that microRNAs that are expressed during latency play a role. Determination that specific LAT promoter sequences and miRNA sequences do not influence viral reactivation of HSV-2 helps to narrow down the search for the mechanism by which the virus controls its latency and recurrence phenotype.

Varicella-Zoster Virus Expresses Multiple Small Noncoding RNAs

Many herpesviruses express small noncoding RNAs (sncRNAs), including microRNAs (miRNAs), that may play roles in regulating lytic and latent infections. None have yet been reported in varicella-zoster virus (VZV; also known as human herpesvirus 3 [HHV-3]). Here we analyzed next-generation sequencing (NGS) data for small RNAs in VZV-infected fibroblasts and human embryonic stem cell-derived (hESC) neurons. Two independent bioinformatics analyses identified more than 20 VZV-encoded 20- to 24-nucleotide RNAs, some of which are predicted to have stem-loop precursors potentially representing miRNAs. These sequences are perfectly conserved between viruses from three clades of VZV. One NGS-identified sequence common to both bioinformatics analyses mapped to the repeat regions of the VZV genome, upstream of the predicted promoter of the immediate early gene open reading frame 63 (ORF63). This miRNA candidate was detected in each of 3 independent biological repetitions of NGS of RNA from fibroblasts and neurons productively infected with VZV using TaqMan quantitative PCR (qPCR). Importantly, transfected synthetic RNA oligonucleotides antagonistic to the miRNA candidate significantly enhanced VZV plaque growth rates. The presence of 6 additional small noncoding RNAs was also verified by TaqMan qPCR in productively infected fibroblasts and ARPE19 cells. Our results show VZV, like other human herpesviruses, encodes several sncRNAs and miRNAs, and some may regulate infection of host cells.IMPORTANCE Varicella-zoster virus is an important human pathogen, with herpes zoster being a major health issue in the aging and immunocompromised populations. Small noncoding RNAs (sncRNAs) are recognized as important actors in modulating gene expression, and this study demonstrates the first reported VZV-encoded sncRNAs. Many are clustered to a small genomic region, as seen in other human herpesviruses. At least one VZV sncRNA was expressed in productive infection of neurons and fibroblasts that is likely to reduce viral replication. Since sncRNAs have been suggested to be potential targets for antiviral therapies, identification of these molecules in VZV may provide a new direction for development of treatments for painful herpes zoster.

Alphaherpesvirus Latency: A Dynamic State of Transcription and Reactivation

Alphaherpesviruses infect a variety of species from sea turtles to man and can cause significant disease in mammals including humans and livestock. These viruses are characterized by a lytic and latent state in nerve ganglia, with the ability to establish a lifelong latent infection that is interrupted by periodic reactivation. Previously, it was accepted that latency was a dominant state and that only during relatively infrequent reactivation episodes did latent genomes within ganglia become transcriptionally active. Here, we review recent data, focusing mainly on Herpes Simplex Virus type 1 which indicate that the latent state is more dynamic than recently appreciated.