Characterization of an abundant, unique 1.7-kilobase bovine herpesvirus 4 (BHV-4) late RNA and mapping of a BHV-4 IE2 transactivator-binding site in its promoter-regulatory region

The Expression and Nuclear Retention Element of Polyadenylated Nuclear RNA Is Not Required for Productive Lytic Replication of Kaposi's Sarcoma-Associated Herpesvirus

Kaposi's sarcoma-associated herpesvirus (KSHV) is an oncogenic human gammaherpesvirus and the causative agent of Kaposi's sarcoma (KS), primary effusion lymphoma (PEL), and multicentric Castleman's disease (MCD). During reactivation, viral genes are expressed in a temporal manner. These lytic genes encode transactivators, core replication proteins, or structural proteins. During reactivation, other viral factors that are required for lytic replication are expressed. The most abundant viral transcript is the long noncoding RNA (lncRNA) known as polyadenylated nuclear (PAN) RNA. lncRNAs have diverse functions, including the regulation of gene expression and the immune response. PAN possesses two main cis-acting elements, the Mta response element (MRE) and the expression and nuclear retention element (ENE). While PAN has been demonstrated to be required for efficient viral replication, the function of these elements within PAN remains unclear. Our goal was to determine if the ENE of PAN is required in the context of infection. A KSHV bacmid containing a deletion of the 79-nucleotide (nt) ENE in PAN was generated to assess the effects of the ENE during viral replication. Our studies demonstrated that the ENE is not required for viral DNA synthesis, lytic gene expression, or the production of infectious virus. Although the ENE is not required for viral replication, we found that the ENE functions to retain PAN in the nucleus, and the absence of the ENE results in an increased accumulation of PAN in the cytoplasm. Furthermore, open reading frame 59 (ORF59), LANA, ORF57, H1.4, and H2A still retain the ability to bind to PAN in the absence of the ENE. Together, our data highlight how the ENE affects the nuclear retention of PAN but ultimately does not play an essential role during lytic replication. Our data suggest that PAN may have other functional domains apart from the ENE. IMPORTANCE KSHV is an oncogenic herpesvirus that establishes latency and exhibits episodes of reactivation. KSHV disease pathologies are most often associated with the lytic replication of the virus. PAN RNA is the most abundant viral transcript during the reactivation of KSHV and is required for viral replication. Deletion and knockdown of PAN resulted in defects in viral replication and reduced virion production in the absence of PAN RNA. To better understand how the cis elements within PAN may contribute to its function, we investigated if the ENE of PAN was necessary for viral replication. Although the ENE had previously been extensively studied with both biochemical and in vitro approaches, this is the first study to demonstrate the role of the ENE in the context of infection and that the ENE of PAN is not required for the lytic replication of KSHV.

PAN RNA: transcriptional exhaust from a viral engine

Kaposi’s sarcoma-associated herpesvirus (KSHV), also designated human herpesvirus 8 (HHV-8), has been linked to Kaposi’s sarcoma, as well as to primary effusion lymphoma (PEL), and a subset of multicentric Castleman’s disease. KSHV genomes are maintained as episomes within infected cells and the virus exhibits a biphasic life cycle consisting of a life-long latent phase during which only a few viral genes are expressed and no viral progeny are produced and a transient lytic reactivation phase, in which a full repertoire of ~ 80 lytic genes are activated in a temporally regulated manner culminating in the release of new virions. Lytic replication is initiated by a single viral protein, K-Rta (ORF50), which activates more than 80 viral genes from multiple resident viral episomes (i.e., viral chromosomes). One of the major targets of K-Rta is a long non-coding nuclear RNA, PAN RNA (polyadenylated nuclear RNA), a lncRNA that accumulates to exceedingly high levels in the nucleus during viral reactivation. K-Rta directly binds to the PAN RNA promoter and robustly activates PAN RNA expression. Although PAN RNA has been known for over 20 years, its role in viral replication is still incompletely understood. In this perspective, we will briefly review the current understanding of PAN RNA and then describe our current working model of this RNA. The model is based on our observations concerning events that occur during KSHV lytic reactivation including (i) a marked accumulation of RNA Pol II at the PAN promoter, (ii) genomic looping emanating from the PAN locus, (iii) interaction of a second viral lytic protein (ORF57) with K-Rta, PAN RNA and RNA Pol II, (iv) the essential requirement for PAN RNA expression in cis for optimal transcriptional execution needed for the entire lytic program, and (v) ORF57 recruitment of RNA Pol II to the PAN genomic locus. Together our results generate a model in which the PAN locus serves as a hub for sequestration/trapping of the cellular transcriptional machinery proximal to viral episomes. Sequestration at the PAN locus facilitates high levels of viral transcription throughout the viral genome during lytic replication. ORF57 acts as a transcription-dependent transactivator at the PAN locus by binding to both Rta and PAN to locally trap RNA Pol II. The resulting accumulation of high levels of nuclear PAN RNA created by this process is an inducible enhancer-derived (eRNA) by-product that litters the infected cell nucleus.

Two herpesviral noncoding PAN RNAs are functionally homologous but do not associate with common chromatin loci

During lytic replication of Kaposi’s sarcoma-associated herpesvirus (KSHV), a nuclear viral long noncoding RNA known as PAN RNA becomes the most abundant polyadenylated transcript in the cell. Knockout or knockdown of KSHV PAN RNA results in loss of late lytic viral gene expression and, consequently, reduction of progeny virion release from the cell. Here, we demonstrate that knockdown of PAN RNA from the related Rhesus macaque rhadinovirus (RRV) phenocopies that of KSHV PAN RNA. These two PAN RNA homologs, although lacking significant nucleotide sequence conservation, can functionally substitute for each other to rescue phenotypes associated with the absence of PAN RNA expression. Because PAN RNA is exclusively nuclear, previous studies suggested that it directly interacts with host and viral chromatin to modulate gene expression. We studied KSHV and RRV PAN RNA homologs using capture hybridization analysis of RNA targets (CHART) and observed their association with host chromatin, but the loci differ between PAN RNA homologs. Accordingly, we find that KSHV PAN RNA is undetectable in chromatin following cell fractionation. Thus, modulation of gene expression at specific chromatin loci appears not to be the primary, nor the pertinent function of this viral long noncoding RNA. PAN RNA represents a cautionary tale for the investigation of RNA association with chromatin whereby cross-linking of DNA spatially adjacent to an abundant nuclear RNA gives the appearance of specific interactions. Similarly, PAN RNA expression does not affect viral transcription factor complex expression or activity, which is required for generation of the late lytic viral mRNAs. Rather, we provide evidence for an alternative model of PAN RNA function whereby knockdown of KSHV or RRV PAN RNA results in compromised nuclear mRNA export thereby reducing the cytoplasmic levels of viral mRNAs available for production of late lytic viral proteins.

Herpesviruses produce noncoding RNAs, some of which are essential to the viral life cycle. One such noncoding RNA from Kaposi’s sarcoma-associated herpesvirus is the polyadenylated, nuclear (PAN) RNA, which is required for production and release of progeny virions from infected cells. In this study, we demonstrate that although lacking nucleotide sequence conservation, PAN RNAs from two related viruses–when knocked down–exhibit the same phenotype, loss of late lytic viral gene expression and progeny virion production. Moreover, they can functionally substitute for each other to rescue this phenotype. We demonstrate that, in contrast to published literature, the reduction in viral gene expression upon PAN RNA knockdown is not due to loss of PAN RNA association with conserved, specific chromatin loci, nor does PAN RNA expression affect the viral transcription factor complex required for generation of the late lytic viral mRNAs. We present data suggesting that PAN RNA instead serves as a binding platform to sequester cellular proteins that are mislocalized to the nucleoplasm. These herpesviral noncoding RNAs can serve as models for the mechanistic study of human noncoding RNAs.

Efficient heterologous antigen gene delivery and expression by a replication-attenuated BoHV-4-based vaccine vector

Bovine Herpesvirus 4 (BoHV-4) is a gammaherpesvirus belonging to the Rhadinovirus genus and due to its biological characteristics has been proposed as a vaccine vector for veterinary vaccines. Because viral vector-associated risk is a major concern for viral vector applications, attenuation is a desirable feature. Therefore, efforts are directed toward the development of highly attenuated viral vectors. BoHV-4 naturally exhibits limited pathogenicity and a further attenuation, in terms of replication, was obtained by disrupting the late gene encoding the 1.7-kb polyadenylated RNA (L1.7). An L1.7 deleted mutant BoHV-4 (BoHV-4-A-KanaGalKΔL1.7), as well as its revertant (BoHV-4-A-Rev), was generated by homologous recombination from the genome of a BoHV-4 isolate (BoHV-4-A) cloned as a bacterial artificial chromosome (BAC). BoHV-4-A-KanaGalKΔL1.7 showed attenuation in terms of competence to reconstitute infectious virus, viral replication, and plaque size when compared to BoHV-4-A, BoHV-4-A-Rev, and BoHV-4-A-KanaGalKΔTK, a recombinant control virus where the KanaGalK selectable marker was inserted into the thymidine kinase open reading frame. The capability of BoHV-4-A-KanaGalKΔL1.7 to deliver and express a heterologous antigen was investigated by replacing the KanaGalK cassette with a vesicular stomatitis virus glycoprotein (VSVg) expression cassette to generate BoHV-4-A-EF1αVSVgΔL1.7. BoHV-4-A-EF1αVSVgΔL1.7 infected cells robustly expressed VSVg, thus confirming that the replication deficiency resulting from L1.7 disruption did not prevent heterologous gene delivery and expression. Although further work is needed to identify the specific function of the BoHV-4 L1.7 gene, the L1.7 gene may represent an ideal targeting locus for the integration of a heterologous antigen expression cassette, resulting in attenuation of the viral vector.

Conservation of a triple-helix-forming RNA stability element in noncoding and genomic RNAs of diverse viruses

Abundant expression of the long noncoding (lnc) PAN (polyadenylated nuclear) RNA by the human oncogenic gammaherpesvirus Kaposi's sarcoma-associated herpesvirus (KSHV) depends on a cis-element called the expression and nuclear retention element (ENE). The ENE upregulates PAN RNA by inhibiting its rapid nuclear decay through triple-helix formation with the poly(A) tail. Using structure-based bioinformatics, we identified six ENE-like elements in evolutionarily diverse viral genomes. Five are in double-stranded DNA viruses, including mammalian herpesviruses, insect polydnaviruses, and a protist mimivirus. One is in an insect picorna-like positive-strand RNA virus, suggesting that the ENE can counteract cytoplasmic as well as nuclear RNA decay pathways. Functionality of four of the ENEs was demonstrated by increased accumulation of an intronless polyadenylated reporter transcript in human cells. Identification of these ENEs enabled the discovery of PAN RNA homologs in two additional gammaherpesviruses, RRV and EHV2. Our findings demonstrate that searching for structural elements can lead to rapid identification of lncRNAs.

Establishment of an immortal turkey turbinate cell line suitable for avian metapneumovirus propagation

Until recently, there has not been a homologous avian cellular substrate which could continuously produce high titer avian metapneumovirus (AMPV); development of such a cell line should provide an excellent model system for studying AMPV infection. We have established a non-tumorigenic immortal turkey turbinate cell line (TT-1) to propagate sufficiently high AMPV titers. Currently, immortal TT-1 cells are growing continuously at 1.2-1.4 population doublings per day and are at passage 160. Kinetic analysis suggests that AMPV can infect and replicate more rapidly in TT-1 compared to Vero cells, although both cell types undergo apoptosis upon infection. The non-tumorigenic, reverse transcriptase negative TT-1 cell line can serve as an excellent homologous cellular substrate for virus propagation.

Genome sequence of bovine herpesvirus 4, a bovine Rhadinovirus, and identification of an origin of DNA replication

Bovine herpesvirus 4 (BoHV-4) is a gammaherpesvirus of cattle. The complete long unique coding region (LUR) of BoHV-4 strain 66-p-347 was determined by a shotgun approach. Together with the previously published noncoding terminal repeats, the entire genome sequence of BoHV-4 is now available. The LUR consists of 108,873 bp with an overall G+C content of 41.4%. At least 79 open reading frames (ORFs) are present in this coding region, 17 of them unique to BoHV-4. In contrast to herpesvirus saimiri and human herpesvirus 8, BoHV-4 has a reduced set of ORFs homologous to cellular genes. Gene arrangement as well as phylogenetic analysis confirmed that BoHV-4 is a member of the genus Rhadinovirus. In addition, an origin of replication (ori) in the genome of BoHV-4 was identified by DpnI assays. A minimum of 1.69 kbp located between ORFs 69 and 71 was sufficient to act as a cis signal for replication.

Establishment of a cell line persistently infected with bovine herpesvirus-4 by use of a recombinant virus

Bovine herpesvirus-4 (BHV-4), a gammaherpesvirus lacking a clear disease association, productively infects multiple cell lines of various species and causes cell death. A human rhabdomyosarcoma cell line, RD-4, infected with BHV-4 produced low levels of early and late viral RNAs and infectious virus, but exhibited no cytopathic effect. Using a recombinant BHV-4 containing a neomycin-resistance gene, we established RD-4-derived cell lines persistently infected with BHV-4. The viral genome in these cells was predominantly circular. Because of drug selection, every cell contained a viral genome. In addition, all cells stained with a BHV-4-specific antiserum. Therefore, these cell lines are not carrier cultures. These cells produced infectious virus at all passages tested. Even though cells were selected and maintained at a concentration of geneticin at least 2.5 times that necessary to kill uninfected RD-4 cells, selected cells contained only approximately one viral genome per diploid host cell genome. Persistently infected cells grew more slowly than uninfected cells, even in the absence of drug. The slower growth of these cells suggests that any growth advantage conferred by multiple copies of the neomycin-gene-carrying viral genome might be offset by the detrimental effects of viral gene expression. This situation contrasts with other gammaherpesviruses, which are able to growth-transform cells.

Gene expression from the ORF50/K8 region of Kaposi's sarcoma-associated herpesvirus

The ORF50 gene of Kaposi's sarcoma (KS)-associated herpesvirus, or human herpesvirus 8 (KSHV), activates viral replication and is weakly homologous to the herpesvirus family of R transactivators; therefore, the transcription and translation events from this region of KSHV are key events in viral reactivation. We demonstrate that ORF50 is expressed in a bicistronic message after induction of the viral lytic cycle. ORF50 migrated as a series of polypeptides: the major ones as 119 and 101 kDa, respectively. Using 3' rapid amplification of cDNA ends, RT-PCR, and cDNA library screening, we demonstrate that the major ORF50 transcript also encodes K8. The ORF50/K8 transcript was resistant to cyclohexamide, whereas the K8 transcript was only partially resistant to cyclohexamide at early timepoints. Both transcripts showed partial resistance after 12 h of phorbol ester induction. Using a GAL4-ORF50 fusion protein expression vector, we demonstrate that the transactivation domain of ORF50 resides within a 160-amino-acid region of the carboxyl portion of the ORF. Upstream regions of both ORF50 and K8 have basal promoter activity in KSHV-infected cells. K8, which had sequence homology to Bzip proteins, did not activate either promoter. However, both promoters were activated after cotransfection of ORF50 in BCBL-1 cells.