Molecular Aspects of Varicella-Zoster Virus Latency

Abstract

Primary varicella-zoster virus (VZV) infection causes varicella (chickenpox) and the establishment of a lifelong latent infection in ganglionic neurons. VZV reactivates in about one-third of infected individuals to cause herpes zoster, often accompanied by neurological complications. The restricted host range of VZV and, until recently, a lack of suitable in vitro models have seriously hampered molecular studies of VZV latency. Nevertheless, recent technological advances facilitated a series of exciting studies that resulted in the discovery of a VZV latency-associated transcript (VLT) and provide novel insights into our understanding of VZV latency and factors that may initiate reactivation. Deducing the function(s) of VLT and the molecular mechanisms involved should now be considered a priority to improve our understanding of factors that govern VZV latency and reactivation. In this review, we summarize the implications of recent discoveries in the VZV latency field from both a virus and host perspective and provide a roadmap for future studies.

Keywords: RNA-sequencing; VZV latency-associated transcript; epigenetics; immunity; latency; open reading frame 63; reactivation; sensory ganglia; varicella-zoster virus.

Figure 1

Structure of varicella-zoster virus (VZV) particles and genome. (A) Electron microscopy image of VZV (obtained from Centers for Disease Control and Prevention (CDC)/Dr Erskine Palmer; B.G. Partin. CDC Public Health Image Library). (B) Schematic representation of the VZV virion. (C) Schematic representation of the VZV genome structure and G + C content. (D) The VZV transcriptome profile during lytic infection of ARPE-19 cells (outer track) and latent infection of human TG (trigeminal ganglia; inner track). Note that the low frequency of VZV-infected neurons in human TG necessitates the use of targeted enrichment of VZV transcripts to detect VZV latency-associated transcript (VLT) expression [27]. Circos plot of the VZV genome (U_L, TR/IR, and U_S are shown as purple, white, and grey bands, respectively; sense and antisense open reading frames (ORFs) are indicated as red and blue blocks, respectively, with ORF numbers indicated where possible). Data represent strand-specific VZV-enriched mRNA-sequencing with peaks facing outwards from the center (black) indicating reads mapping to the sense strand, while peaks facing inward (grey) originate from the antisense strand. The y-axis is scaled to the maximum read depth per library in all cases. dsDNA, double-stranded DNA; U_L, unique long; U_S, unique short; TR, terminal repeat; IR, internal repeat; R, reiterative region; Ori, origin of replication.

Figure 2

Schematic representation of the establishment of, and reactivation from, VZV latency in sensory neurons. (A) VZV gains access to sensory ganglion neurons via the infection of nerve endings in skin and retrograde axonal transport to the neuronal cell body (1) or direct infection of cell bodies via VZV-infected T-cells (2), followed by the release of the viral genome into the nucleus (3). (B) VZV reactivation results in virus replication and spread in the cell body (1), followed by transaxonal spread to the skin to cause HZ (2), possibly involving concordant virus spread to the spinal cord (3). (C) VZV latently infected sensory neurons contain viral episomal DNA in their nuclei and express VLT and/or ORF63 RNA, as shown by RNA in situ hybridization on human TG (red signal). Magnification: 400× with 2.5× digital zoom.

Figure 3

Comparison of latency-associated transcripts among alphaherpesviruses. All alphaherpesvirus latency-associated transcripts (LATs) are located antisense to the infected cell polypeptide 0 (ICP0) locus, encoding a conserved major immediate-early transactivator (ICP0 or homologs). (A) The VZV latency-associated transcript (VLT) is a 496-nucleotide, multi-exon mRNA that is partially antisense to the ORF61 coding region via exons 3 and 4. (B) Transcripts mapping antisense to simian varicella virus ORF61 are expressed during latency (dotted arrow), but their identity has not yet been defined. (C) Bovine herpesvirus 1 encodes a 2.2 kb latency-related (LR) RNA and encodes two miRNAs in exon 1 (D) The pseudorabies virus large latency transcript (LLT) is the largest characterized alphaherpesvirus latency transcript and encodes eleven distinct miRNAs within the spliced intron. (E) The 8.2 kb herpes simplex virus type 1 (HSV-1) LAT undergoes splicing which yields two highly stable intron lariats, approximately 1.5 and 2.0 kb in size (shown as circles). Note that latency transcripts are shown as grey arrows, immediate early viral transactivators as dark red arrows, and encoded miRNAs as short vertical lines. Scaling across all schematics is equal.

Figure 4

Schematic representation of the regulation of VZV latency and reactivation. The activation of neuronal signaling pathways in response to stimuli at the periphery (2) and possibly, within the spinal cord (3) may induce VZV reactivation. At the same time, adaptive T-cell-mediated immune responses in skin (1) and ganglia (4), and the local ganglionic innate immunity provided by neuron-interacting satellite glial cells (SGC) are believed to prevent symptomatic VZV reactivation.