Analysis of Virus and Host Proteomes During Productive HSV-1 and VZV Infection in Human Epithelial Cells

Leveraging biotin-based proximity labeling to identify cellular factors governing early alphaherpesvirus infection

Neurotropic alphaherpesviruses, including herpes simplex virus type 1 and pseudorabies virus, establish a lifelong presence within the peripheral nervous system of their mammalian hosts. Upon entering cells, two conserved tegument proteins, pUL36 and pUL37, traffic DNA-containing capsids to nuclei. These proteins support long-distance retrograde axonal transport and invasion of the nervous system in vivo. To better understand how pUL36 and pUL37 function, recombinant viral particles carrying BioID2 fused to these proteins were produced to biotinylate cellular proteins in their proximity (<10 nm) during infection. Eighty-six high-confidence host proteins were identified by mass spectrometry and subsequently targeted by CRISPR-Cas9 gene editing to assess their contributions to early infection. Proteins were identified that both supported and antagonized infection in immortalized human epithelial cells. The latter included zyxin, a protein that localizes to focal adhesions and regulates actin cytoskeletal dynamics. Zyxin knockout cells were hyper-permissive to infection and could be rescued with even modest expression of GFP-zyxin. These results provide a resource for studies of the virus-cell interface and identify zyxin as a novel deterrent to alphaherpesvirus infection.IMPORTANCENeuroinvasive alphaherpesviruses are highly prevalent with many members found across mammals [e.g., herpes simplex virus type 1 (HSV-1) in humans and pseudorabies virus in pigs]. HSV-1 causes a range of clinical manifestations from cold sores to blindness and encephalitis. There are no vaccines or curative therapies available for HSV-1. A fundamental feature of these viruses is their establishment of lifelong infection of the nervous system in their respective hosts. This outcome is possible due to a potent neuroinvasive property that is coordinated by two proteins: pUL36 and pUL37. In this study, we explore the cellular protein network in proximity to pUL36 and pUL37 during infection and examine the impact of knocking down the expression of these proteins upon infection.

Subconfluent ARPE-19 Cells Display Mesenchymal Cell-State Characteristics and Behave like Fibroblasts, Rather Than Epithelial Cells, in Experimental HCMV Infection Studies

Human cytomegalovirus (HCMV) has a broad cellular tropism and epithelial cells are important physiological targets during infection. The retinal pigment epithelial cell line ARPE-19 has been used to model HCMV infection in epithelial cells for decades and remains a commonly used cell type for studying viral entry, replication, and the cellular response to infection. We previously found that ARPE-19 cells, despite being derived from an epithelial cell explant, express extremely low levels of canonical epithelial proteins, such as E-cadherin and EpCAM. Here, we perform comparative studies of ARPE-19 and additional epithelial cell lines with strong epithelial characteristics. We find that ARPE-19 cells cultured under subconfluent conditions resemble mesenchymal fibroblasts, rather than epithelial cells; this is consistent with previous studies showing that ARPE-19 cultures require extended periods of high confluency culture to maintain epithelial characteristics. By reanalyzing public gene expression data and using machine learning, we find evidence that ARPE-19 cultures maintained across many labs exhibit mesenchymal characteristics and that the majority of studies employing ARPE-19 use them in a mesenchymal state. Lastly, by performing experimental HCMV infections across mesenchymal and epithelial cell lines, we find that ARPE-19 cells behave like mesenchymal fibroblasts, producing logarithmic yields of cell-free infectious progeny, while cell lines with strong epithelial character exhibit an atypical infectious cycle and naturally restrict the production of cell-free progeny. Our work highlights important characteristics of the ARPE-19 cell line and suggests that subconfluent ARPE-19 cells may not be optimal for modeling epithelial infection with HCMV or other human viruses. It also suggests that HCMV biosynthesis and/or spread may occur quite differently in epithelial cells compared to mesenchymal cells. These differences could contribute to viral persistence or pathogenesis in epithelial tissues.

Viral proteogenomic and expression profiling during productive replication of a skin-tropic herpesvirus in the natural host

Efficient transmission of herpesviruses is essential for dissemination in host populations; however, little is known about the viral genes that mediate transmission, mostly due to a lack of natural virus-host model systems. Marek's disease is a devastating herpesviral disease of chickens caused by Marek's disease virus (MDV) and an excellent natural model to study skin-tropic herpesviruses and transmission. Like varicella zoster virus that causes chicken pox in humans, the only site where infectious cell-free MD virions are efficiently produced is in epithelial skin cells, a requirement for host-to-host transmission. Here, we enriched for heavily infected feather follicle epithelial skin cells of live chickens to measure both viral transcription and protein expression using combined short- and long-read RNA sequencing and LC/MS-MS bottom-up proteomics. Enrichment produced a previously unseen breadth and depth of viral peptide sequencing. We confirmed protein translation for 84 viral genes at high confidence (1% FDR) and correlated relative protein abundance with RNA expression levels. Using a proteogenomic approach, we confirmed translation of most well-characterized spliced viral transcripts and identified a novel, abundant isoform of the 14 kDa transcript family via IsoSeq transcripts, short-read intron-spanning sequencing reads, and a high-quality junction-spanning peptide identification. We identified peptides representing alternative start codon usage in several genes and putative novel microORFs at the 5' ends of two core herpesviral genes, pUL47 and ICP4, along with strong evidence of independent transcription and translation of the capsid scaffold protein pUL26.5. Using a natural animal host model system to examine viral gene expression provides a robust, efficient, and meaningful way of validating results gathered from cell culture systems.

SUN2 Modulates the Propagation of HSV-1

Herpesviruses assemble new viral particles in the nucleus. These nucleocapsids bud through the inner nuclear membrane to produce enveloped viral particles in the perinuclear space before fusing with the outer nuclear membrane to reach the cytoplasm. This unusual route is necessary since viral capsids are too large to pass through nuclear pores. However, the transient perinuclear nucleocapsids (250 nm in diameter) are also larger than the width of the perinuclear space (30 to 50 nm). Interestingly, linker of the nucleoskeleton and cytoskeleton (LINC) components SUN and KASH connect the inner and outer nuclear membranes and regulate their spacing. Previous work by others on the related pseudorabies virus and human cytomegalovirus showed that they functionally interact with SUN proteins. To clarify the role of SUN proteins, we explored their impact on herpes simplex virus 1 (HSV-1), another herpesvirus. Using dominant negative SUN mutants and RNA interference, we show that HSV-1 propagation is dependent on the LINC complex. In contrast to pseudorabies virus, SUN2 disruption by either approach led to increased HSV-1 extracellular viral yields. This SUN2 dependency may be linked to its greater impact on perinuclear spacing in infected cells compared to SUN1. Finally, the virus itself seems to modulate perinuclear spacing. IMPORTANCE The large size of herpesviruses prevents them from travelling across the nuclear pores, and they instead egress across the two nuclear membranes, generating short-lived enveloped perinuclear virions. This poses a challenge as the perinuclear space is smaller than the virions. This implies the separation (unzipping) of the two nuclear membranes to accommodate the viral particles. The LINC complex bridges the two nuclear membranes and is an important regulator of perinuclear spacing. Work by others hint at its functional implication during pseudorabies virus and cytomegalovirus propagation. The present study probes the importance for HSV-1 of the SUN proteins, the LINC components found in the inner nuclear membrane. Using dominant negative constructs and RNA interference (RNAi), the data reveal that SUN2 exhibits antiviral propriety toward HSV-1, as disrupting the protein leads to increased viral yields. This is in contrast with that reported for pseudorabies and suggests that differences among herpesviruses may, once again, prevail.

The crystal structure of the varicella-zoster Orf24-Orf27 nuclear egress complex spotlights multiple determinants of herpesvirus subfamily specificity

Varicella-zoster virus (VZV) is a human pathogen from the α-subfamily of herpesviruses. The VZV Orf24-Orf27 complex represents the essential viral core nuclear egress complex (NEC) that orchestrates the egress of the preassembled virus capsids from the nucleus. While previous studies have primarily emphasized that the architecture of core NEC complexes is highly conserved among herpesviruses, the present report focuses on subfamily-specific structural and functional features that help explain the differences in the autologous versus nonautologous interaction patterns observed for NEC formation across herpesviruses. Here, we describe the crystal structure of the Orf24-Orf27 complex at 2.1 Å resolution. Coimmunoprecipitation and confocal imaging data show that Orf24-Orf27 complex formation displays some promiscuity in a herpesvirus subfamily-restricted manner. At the same time, analysis of thermodynamic parameters of NEC formation of three prototypical α-, β-, and γ herpesviruses, i.e., VZV, human cytomegalovirus (HCMV), and Epstein–Barr virus (EBV), revealed highly similar binding affinities for the autologous interaction with specific differences in enthalpy and entropy. Computational alanine scanning, structural comparisons, and mutational data highlight intermolecular interactions shared among α-herpesviruses that are clearly distinct from those seen in β- and γ-herpesviruses, including a salt bridge formed between Orf24-Arg167 and Orf27-Asp126. This interaction is located outside of the hook-into-groove interface and contributes significantly to the free energy of complex formation. Combined, these data explain distinct properties of specificity and permissivity so far observed in herpesviral NEC interactions. These findings will prove valuable in attempting to target multiple herpesvirus core NECs with selective or broad-acting drug candidates.

Mutagenesis of the Varicella-Zoster Virus Genome Demonstrates That VLT and VLT-ORF63 Proteins Are Dispensable for Lytic Infection

Primary varicella-zoster virus (VZV) infection leads to varicella and the establishment of lifelong latency in sensory ganglion neurons. Reactivation of latent VZV causes herpes zoster, which is frequently associated with chronic pain. Latent viral gene expression is restricted to the VZV latency-associated transcript (VLT) and VLT-ORF63 (VLT63) fusion transcripts. Since VLT and VLT63 encode proteins that are expressed during lytic infection, we investigated whether pVLT and pVLT-ORF63 are essential for VZV replication by performing VZV genome mutagenesis using CRISPR/Cas9 and BAC technologies. We first established that CRISPR/Cas9 can efficiently mutate VZV genomes in lytically VZV-infected cells through targeting non-essential genes ORF8 and ORF11 and subsequently show recovery of viable mutant viruses. By contrast, the VLT region was markedly resistant to CRISPR/Cas9 editing. Whereas most mutants expressed wild-type or N-terminally altered versions of pVLT and pVLT-ORF63, only a minority of the resulting mutant viruses lacked pVLT and pVLT-ORF63 coding potential. Growth curve analysis showed that pVLT/pVLT-ORF63 negative viruses were viable, but impaired in growth in epithelial cells. We confirmed this phenotype independently using BAC-derived pVLT/pVLT-ORF63 negative and repaired viruses. Collectively, these data demonstrate that pVLT and/or pVLT-ORF63 are dispensable for lytic VZV replication but promote efficient VZV infection in epithelial cells.

Targeted RNA Sequencing of VZV-Infected Brain Vascular Adventitial Fibroblasts Indicates That Amyloid May Be Involved in VZV Vasculopathy

BACKGROUND AND OBJECTIVES

Compared with stroke controls, patients with varicella zoster virus (VZV) vasculopathy have increased amyloid in CSF, along with increased amylin (islet amyloid polypeptide [IAPP]) and anti-VZV antibodies. Thus, we examined the gene expression profiles of VZV-infected primary human brain vascular adventitial fibroblasts (HBVAFs), one of the initial arterial cells infected in VZV vasculopathy, to determine whether they are a potential source of amyloid that can disrupt vasculature and potentiate inflammation.

METHODS

Mock- and VZV-infected quiescent HBVAFs were harvested at 3 days postinfection. Targeted RNA sequencing of the whole-human transcriptome (BioSpyder Technologies, TempO-Seq) was conducted followed by gene set enrichment and pathway analysis. Selected pathways unique to VZV-infected cells were confirmed by enzyme-linked immunoassays, migration assays, and immunofluorescence analysis (IFA) that included antibodies against amylin and amyloid-beta, as well as amyloid staining by Thioflavin-T.

RESULTS

Compared with mock, VZV-infected HBVAFs had significantly enriched gene expression pathways involved in vascular remodeling and vascular diseases; confirmatory studies showed secretion of matrix metalloproteinase-3 and -10, as well increased migration of infected cells and uninfected cells when exposed to conditioned media from VZV-infected cells. In addition, significantly enriched pathways involved in amyloid-associated diseases (diabetes mellitus, amyloidosis, and Alzheimer disease), tauopathy, and progressive neurologic disorder were identified; predicted upstream regulators included amyloid precursor protein, apolipoprotein E, microtubule-associated protein tau, presenilin 1, and IAPP. Confirmatory IFA showed that VZV-infected HBVAFs contained amyloidogenic peptides (amyloid-beta and amylin) and intracellular amyloid.

DISCUSSION

Gene expression profiles and pathway enrichment analysis of VZV-infected HBVAFs, as well as phenotypic studies, reveal features of pathologic vascular remodeling (e.g., increased cell migration and changes in the extracellular matrix) that can contribute to cerebrovascular disease. Furthermore, the discovery of amyloid-associated transcriptional pathways and intracellular amyloid deposition in HBVAFs raise the possibility that VZV vasculopathy is an amyloid disease. Amyloid deposition may contribute to cell death and loss of vascular wall integrity, as well as potentiate chronic inflammation in VZV vasculopathy, with disease severity and recurrence determined by the host's ability to clear virus infection and amyloid deposition and by the coexistence of other amyloid-associated diseases (i.e., Alzheimer disease and diabetes mellitus).

Quantitative Temporal Viromics

The abundance, localization, modifications, and protein-protein interactions of many host cell and virus proteins can change dynamically throughout the course of any viral infection. Studying these changes is critical for a comprehensive understanding of how viruses replicate and cause disease, as well as for the development of antiviral therapeutics and vaccines. Previously, we developed a mass spectrometry-based technique called quantitative temporal viromics (QTV), which employs isobaric tandem mass tags (TMTs) to allow precise comparative quantification of host and virus proteomes through a whole time course of infection. In this review, we discuss the utility and applications of QTV, exemplified by numerous studies that have since used proteomics with a variety of quantitative techniques to study virus infection through time. Expected final online publication date for the Annual Review of Virology, Volume 8 is September 2021. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Translational adaptation of human viruses to the tissues they infect

Viruses need to hijack the translational machinery of the host cell for a productive infection to happen. However, given the dynamic landscape of tRNA pools among tissues, it is unclear whether different viruses infecting different tissues have adapted their codon usage toward their tropism. Here, we collect the coding sequences of 502 human-infecting viruses and determine that tropism explains changes in codon usage. Using the tRNA abundances across 23 human tissues from The Cancer Genome Atlas (TCGA), we build an in silico model of translational efficiency that validates the correspondence of the viral codon usage with the translational machinery of their tropism. For instance, we detect that severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is specifically adapted to the upper respiratory tract and alveoli. Furthermore, this correspondence is specifically defined in early viral proteins. The observed tissue-specific translational efficiency could be useful for the development of antiviral therapies and vaccines.

Interrogating Host Antiviral Environments Driven by Nuclear DNA Sensing: A Multiomic Perspective

Nuclear DNA sensors are critical components of the mammalian innate immune system, recognizing the presence of pathogens and initiating immune signaling. These proteins act in the nuclei of infected cells by binding to foreign DNA, such as the viral genomes of nuclear-replicating DNA viruses herpes simplex virus type 1 (HSV-1) and human cytomegalovirus (HCMV). Upon binding to pathogenic DNA, the nuclear DNA sensors were shown to initiate antiviral cytokines, as well as to suppress viral gene expression. These host defense responses involve complex signaling processes that, through protein–protein interactions (PPIs) and post-translational modifications (PTMs), drive extensive remodeling of the cellular transcriptome, proteome, and secretome to generate an antiviral environment. As such, a holistic understanding of these changes is required to understand the mechanisms through which nuclear DNA sensors act. The advent of omics techniques has revolutionized the speed and scale at which biological research is conducted and has been used to make great strides in uncovering the molecular underpinnings of DNA sensing. Here, we review the contribution of proteomics approaches to characterizing nuclear DNA sensors via the discovery of functional PPIs and PTMs, as well as proteome and secretome changes that define a host antiviral environment. We also highlight the value of and future need for integrative multiomic efforts to gain a systems-level understanding of DNA sensors and their influence on epigenetic and transcriptomic alterations during infection.

Decoding the Architecture of the Varicella-Zoster Virus Transcriptome

Varicella-zoster virus (VZV), a double-stranded DNA virus, causes varicella, establishes lifelong latency in ganglionic neurons, and reactivates later in life to cause herpes zoster, commonly associated with chronic pain. The VZV genome is densely packed and produces multitudes of overlapping transcripts deriving from both strands. While 71 distinct open reading frames (ORFs) have thus far been experimentally defined, the full coding potential of VZV remains unknown. Here, we integrated multiple short-read RNA sequencing approaches with long-read direct RNA sequencing on RNA isolated from VZV-infected cells to provide a comprehensive reannotation of the lytic VZV transcriptome architecture. Through precise mapping of transcription start sites, splice junctions, and polyadenylation sites, we identified 136 distinct polyadenylated VZV RNAs that encode canonical ORFs, noncanonical ORFs, and ORF fusions, as well as putative noncoding RNAs (ncRNAs). Furthermore, we determined the kinetic class of all VZV transcripts and observed, unexpectedly, that transcripts encoding the ORF62 protein, previously designated Immediate-Early, were expressed with Late kinetics. Our work showcases the complexity of the VZV transcriptome and provides a comprehensive resource that will facilitate future functional studies of coding RNAs, ncRNAs, and the biological mechanisms underlying the regulation of viral transcription and translation during lytic VZV infection.IMPORTANCE Transcription from herpesviral genomes, executed by the host RNA polymerase II and regulated by viral proteins, results in coordinated viral gene expression to efficiently produce infectious progeny. However, the complete coding potential and regulation of viral gene expression remain ill-defined for the human alphaherpesvirus varicella-zoster virus (VZV), causative agent of both varicella and herpes zoster. Here, we present a comprehensive overview of the VZV transcriptome and the kinetic class of all identified viral transcripts, using two virus strains and two biologically relevant cell types. Additionally, our data provide an overview of how VZV diversifies its transcription from one of the smallest herpesviral genomes. Unexpectedly, the transcript encoding the major viral transactivator protein (pORF62) was expressed with Late kinetics, whereas orthologous transcripts in other alphaherpesviruses are typically expressed during the immediate early phase. Therefore, our work both establishes the architecture of the VZV transcriptome and provides insight into regulation of alphaherpesvirus gene expression.