Cytomegalovirus latency-the sum of subtleties

Abstract

Human cytomegalovirus (HCMV) is a betaherpesvirus, which, like all herpesviruses, establishes a life-long latent infection while retaining the ability to reactivate its replicative program. While HCMV likely reactivates frequently and sporadically in healthy individuals and typically without disease, reactivation poses a serious disease threat in the immunocompromised. The latent program of HCMV is complex and has been challenging to define due to limitations in appropriate experimental model systems related to virus-host species specificity, limited identification of in vivo latent reservoirs, and the dynamic cellular differentiation of the hematopoietic latency reservoir that is directly linked to latency maintenance and reactivation phenotypes. Here, we review the current understanding of HCMV latency, with a focus on cross-cutting principles derived collectively from in vitro experimental culture models and in vivo animal models using the corresponding orthologs (CMVs) to HCMV.

Keywords: cytomegalovirus; herpesvirus; latency; virus-host interactions.

Fig 1

Schematic of latency and reactivation. The nature of the latent infection in HCMV remains elusive. However, our mechanistic insights have grown dramatically over the past two decades due to innovation in experimental models, in vitro and in vivo, and viral genetics relying on bacterial artificial chromosome clones of HCMV. The establishment of latency or reactivation from latency depends on the biology of the infected cell and a multitude of interconnecting virus-host interactions (Fig. 2) that remain to be completely defined. A central theme emerging from research carried out by many groups is that latency is established by host restrictions that reversibly restrict viral gene expression without aborting infection following virus entry. Reactivation is stimulated through differentiation- or stress-induced reduction in host restrictions (red) coupled with increases in viral gene expression and host factors (green) that accumulate to thresholds to cross checkpoints for commitment to virus reactivation. The virus has evolved to sense and respond to changes in host signaling, as evidenced by host transcription factor control of virus gene expression, susceptibility of viral factors to selective host degradation, and a plethora of virus-host interactions to tweak biology of the infected cell. Fundamental questions remain to be fully defined, including: (i) how is the viral genome silenced and maintained, (ii) what is the latent transcriptome, (iii) what are the additional reservoirs of HCMV latency, and (iv) how are diverse signals integrated to make decisions around the maintenance of or reactivation from latency?

Fig 2

Hematopoietic differentiation-associated host-virus interactions regulating latency and reactivation. HCMV latency and reactivation are tied to hematopoietic differentiation and corresponding changes in chromatinization of the viral genomes and patterns of viral gene expression (gray gradients). HCMV latency and reactivation decisions are modulated by the integration of diverse cellular information through a series of complex virus-host interactions, which are remarkably defined by hematopoietic differentiation. While this figure is not intended to be comprehensive, we represent key host pathways with gradients to reflect their general regulation through hematopoietic differentiation. Viral proteins (spheres) and miRNAs (blue boxes) and their proposed modulation of these pathways and impact infection outcomes—latency or reactivation—are shown. Some of these interactions are known to influence hematopoietic differentiation, particularly in the case of US28 and viral miRNAs, which target major signaling pathways important to stress responses and cellular differentiation. The EGFR→AKT→PI3K and related-Rho and MAPK signaling axis is important in regulating latency/reactivation decisions (red gradients). These pathways are regulated by US28, UL138, and UL135, as well as v-miRNAs. Type-1 IFN signaling and expression of ISGs are also thought to promote a latent state and can be regulated by US28 and UL138. In the case of UL136, a determinant important for reactivation, its levels are regulated by the IDOL host E3 ubiquitin ligase, which is regulated by LXR signaling through differentiation (orange gradient). High levels of IDOL present in undifferentiated hematopoietic cells maintain low levels of UL136p33 for latency, and loss of IDOL with differentiation allows UL136p33 to accumulate and drive increased IE gene expression and virus reactivation. LUNA desumoylates MORC3 (purple gradient), which represses IE and stimulates UL138 gene expression (46, 47). UL8 interacts and stabilizes components of the Wnt/β-catenin pathway for reactivation (green gradient). Finally, expression of the host transcription factor FOXO3 is increased with differentiation (blue gradient) and is important for activation of alternative intronic promoters regulating IE expression. FOXO3 is maintained at low levels during latency by UL7 and v-miRNAs. Most interactions on the side of reactivation directly or indirectly contribute to increased activation of IE genes. Much remains to be understood about the coordination of and interplay between these host pathways and viral regulators. US28 is particularly complex, and some phenotypes have been defined in CD34+ cells (solid sphere outline) or monocytes (indicated by dashed sphere outline), which we have tried to differentiate.