Sleepless latency of human cytomegalovirus
- 1.8k Downloads
As with all human herpesviruses, human cytomegalovirus (HCMV) persists for the lifetime of the host by establishing a latent infection, which is broken by periodic reactivation events. One site of HCMV latency is in the progenitor cells of the myeloid lineage such as CD34+ cells and their CD14+ derivatives. The development of experimental techniques to isolate and culture these primary cells in vitro is enabling detailed analysis of the events that occur during virus latency and reactivation. Ex vivo differentiation of latently infected primary myeloid cells to dendritic cells and macrophages results in the reactivation of latent virus and provides model systems in which to analyse the viral and cellular functions involved in latent carriage and reactivation. Such analyses have shown that, in contrast to primary lytic infection or reactivation which is characterised by a regulated cascade of expression of all viral genes, latent infection is associated with a much more restricted viral transcription programme with expression of only a small number of viral genes. Additionally, concomitant changes in the expression of cellular miRNAs and cellular proteins occur, and this includes changes in the expression of a number of secreted cellular proteins and intracellular anti-apoptotic proteins, which all have profound effects on the latently infected cells. In this review, we concentrate on the effects of one of the latency-associated viral proteins, LAcmvIL-10, and describe how it causes a decrease in the cellular miRNA, hsa-miR-92a, and a concomitant upregulation of the GATA2 myeloid transcription factor, which, in turn, drives the expression of cellular IL-10. Taken together, we argue that HCMV latency, rather than a period of viral quiescence, is associated with the virally driven manipulation of host cell functions, perhaps every bit as complex as lytic infection. A full understanding of these changes in cellular and viral gene expression during latent infection could have far-reaching implications for therapeutic intervention.
KeywordsLatency Human cytomegalovirus Cell survival
Human cytomegalovirus (HCMV) is typical of human herpesviruses in that it has both a latent and a lytic phase of its life cycle, which are bridged by periodic reactivation events. In the immune competent, primary infection is rarely symptomatic, and following the establishment of latency, the virus maintains a persistent infection effectively controlled by the immune system. It is now established that, in vivo, one site of HCMV latency is in bone marrow resident CD34+ myeloid progenitor cells as well as in their derivative CD14+ monocytes present in peripheral blood. Whilst there is little consensus, likely due to differences in cell types analysed and models of latency used, in the spectrum of viral genes expressed during latent infection, it is clear that latency is associated with a much restricted virus gene transcription programme and, in general, an absence of expression of viral major lytic genes [1, 2, 3, 4, 5]. This is also true for cytomegalovirus infection in other species, although the mechanisms by which latency is established vary. A number of studies have demonstrated that murine cytomegalovirus (MCMV) establishes latency, which suggests that some parallels may be drawn between the two species [6, 7]. This inhibition of lytic gene expression during latency is likely affected through repressive chromatin marks around the promoter of the viral major immediate early promoter (MIEP). Following differentiation of progenitor myeloid cells into terminally differentiated dendritic cells or macrophages, however, this repressive chromatin structure around the MIEP is relieved, resulting in changes in post-translational modifications of histones around the MIEP associated with transcriptional activation and concomitant induction of viral lytic immediately early (IE) gene expression [5, 8, 9, 10, 11, 12]. These data imply that reactivation routinely occurs in vivo, but this is sub-clinical due to robust host immune responses and is supported by the recent observations that macrophages and DCs, in vivo, are sites of virus reactivation in the healthy carrier [13, 14].
There are numerous effects on the host cell during HCMV latency
Given the relative paucity of viral genes expressed during HCMV latency [1, 2, 3, 4, 5] compared to lytic infection, it may initially seem surprising that such profound changes in the cell result from latent infection. However, another level of regulation of gene expression is via microRNAs (miRNAs). These are highly conserved small (approximately 21 nucleotides in length) RNA molecules encoded in the genomes of plants and animals, which generally regulate the expression of genes by binding to the 3′-untranslated regions (3′-UTR) of specific mRNAs.
Although the first published description of an miRNA was in 1993 , the understanding of the functions of many of these small non-coding RNA molecules is still being elucidated and is complicated by the fact that each miRNA is thought to be able to regulate multiple genes. This, coupled with the fact that there are hundreds of miRNAs transcribed in the cells of higher eukaryotes [41, 42, 43], reflects the enormous complexity in levels of regulation of gene expression afforded by miRNAs. Various lines of research suggest that miRNAs may act as key regulators of processes as diverse as early development , cell proliferation and cell death , apoptosis, and fat metabolism , as well as cell differentiation [47, 48, 49]. There is also evidence to suggest that miRNA expression is involved in the pathogenesis of a number of diseases, including cancer [50, 51] and viral infection [31, 52, 53, 54, 55].
HCMV-induced latency leads to changes in a number of cellular miRNAs
Fold change during latency compared to mock
Downregulation of a cellular miRNA can lead to upregulation of cellular proteins during latency
One of the cellular miRNAs which is known to be downregulated during latent infection of CD34+ progenitor cells is hsa-miR-92a ( and Table 1). Predictive algorithms and biochemical analysis have shown that this miRNA can target the myeloid transcription factor GATA2, and as predicted, during experimental HCMV latency, there is an increase in levels of this myeloid cellular transcription factor [31, 34].
Cellular IL-10 is important for the maintenance of latency and immune evasion
Cellular IL-10 is a secreted cytokine known to have immune modulatory properties as well as having pro-life effects on myeloid progenitors, such as CD34+ cells, by driving the expression of the anti-apoptotic factor Bcl2 (Fig. 2a, c, d and ). During HCMV latency, cIL-10 is upregulated, and for this reason, the specific functions of cIL-10 in the latent secretome have been analysed . For example, consistent with this observation that cIL-10 is known to play an anti-apoptotic role in CD34+ cells (Fig. 2d), [31, 61], antibody depletion of latency-associated secretion of cIL-10 from the supernatants of latently infected cells results in increased cell death and loss of latent viral genome carriage . Additionally, latency-associated secretion of cIL-10, together with virally induced increases in cellular TGF-beta, results in the establishment of an immune suppressive microenvironment around latently infected cells. This, in turn, inhibits CD4+ cytotoxic T-cell effector functions and thereby suppresses host immune surveillance of the latently infected cell .
We know that latent infection of primary CD34+ progenitor cells by HCMV results in their increased survival in the face of pro-apoptotic signals, and this, at least in part, appears to involve the known latency-associated increase in the expression of cIL-10 [31, 59]. However, how cIL-10 mediated this protection was unclear, but the changes in the latent apoptome would suggest that the cIL-10 upregulated during latency modulates cIL-10-mediated suppression of extrinsic and intrinsic pro-apoptotic signals (Fig. 2d).
Engagement of the cIL-10 receptor by cIL-10 is known to induce signalling via STAT3 phosphorylation. This results in positive autoregulation of cIL-10 expression as well as expression of intrinsic death-signalling pathway inhibitors such as Bcl2 [31, 59] and HSP70 . Consistent with this, and the known increase in cIL-10 during HCMV latent infection of CD34+ cells , latently infected CD34+ cells also showed extensive increases in STAT3 phosphorylation (Fig. 2b, c) and concomitant increases in the expression of Bcl2 and HSP70 (Fig. 2a, c). HSP70 plays a role in the negative regulation of the intrinsic pathway due to the ability to target pro-caspase 3 [63, 64], present in the FAS-mediated signalling pathway. Thus, during latency, the FAS-mediated signalling pathway is targeted at different stages of FAS-mediated apoptosis via extrinsic apoptosis signalling (Fig. 2d). Interestingly, HSP70 also plays a significant role as a potent inhibitor of the formation of the mitochondrial apoptome  and the intrinsic pathway of programmed cell death (Fig. 2d).
Not all of the changes in proteins we have identified during HCMV latency are in pro-life factors. Figure 2a, c shows that there was a 3.5-fold increase in levels of p53 phosphorylated at serine 15. Ser 15-phosphorylated p53 is known to have pro-apoptotic properties as it is able to upregulate the transcription of the pro-apoptotic factor Bax and concomitantly decrease the expression of anti-apoptotic factor Bcl2 . However, the expression of Bcl2 clearly increased during HCMV latency in our studies (see Fig. 2a, c and ) despite elevated levels of phosphorylated p53. Consequently, our view is that other functions associated with latent infection also counter the transcriptional regulation of Bcl2 by p53. One possibility is that HSP70 increased during latent infection and can act to stimulate the expression of Bcl2 via AKT . Therefore, the expression of HSP70 may be strong enough to overcome the repressive effects of phosphorylated p53. Similarly, formation of pro-apoptotic Bax homodimers can be prevented by HSP70 . Therefore, it is likely that HSP70 works at multiple levels to help check and balance the levels of anti-apoptotic factors in the latently infected cell, although this needs to be formally addressed.
Latency-associated viral IL-10 can cause downregulation of the cellular miRNA hsa-miR-92a
It appears, then, that the changes in cellular miRNA expression resulting from latent infection can have important downstream effects on both intrinsic cell survival and host immune evasion. However, until recently, the mechanism by which HCMV latent infection caused such changes in cellular miRNAs expression was far from clear. There are a number of viral genes expressed during latency, which could potentially affect cellular miRNA expression, although the latency-associated functions of many of these latency-associated genes are only just beginning to be unravelled.
One of these is a viral cIL-10 homologue, known as LAcmvIL-10, encoded by the UL111A gene. It is interesting that although latent HCMV infection robustly induces the expression of cIL-10, as discussed above, the virus also expresses LAcmvIL-10 during this life cycle. However, HCMV is not unique amongst the herpesviruses in encoding a cIL-10 homologue . However, HCMV actually encodes two IL-10 homologues. These two isoforms of virus-encoded IL-10 are generated by alternative splicing from the viral UL111A gene as depicted graphically in Fig. 2e. One of these is a protein of 175 amino acids, termed cmvIL-10, which is expressed during lytic infection and has the expression kinetics of a late gene. The second isoform, predicted to consist of 139 amino acids and termed LAcmvIL-10, has a C-terminal truncation and is expressed during both lytic and latent infection [68, 69, 70]. The encoding of a cIL-10 homologue is not unique to HCMV. An IL-10 homologue encoded by the UL111A open reading frame (ORF) has also been identified in rhesus macaque CMV (RhCMV). Although it has a slightly different gene structure than cmvIL-10, like cmvIL-10, it shows low amino acid identity to host cIL-10 .
The full-length cmvIL-10 gene shares 27 % amino acid homology with cIL-10  and has a number of functions in common with cIL-10: it forms homodimers and binds the cIL-10 receptor [69, 72]; it triggers STAT3 phosphorylation and activation of the JAK/STAT signalling pathway [73, 74]; it signals via the phosphoinositide-3-kinase pathway, contributing to cytokine suppression [74, 75, 76] and cIL-10-positive autoregulation ; and it shares the ability of cIL-10 to prevent NF-κB activity via inhibition of IKK [75, 78].
In contrast, LAcmvIL-10 appears quite dissimilar to cIL-10 and cmvIL-10. Although, like cIL-10 and cmvIL-10, it can downregulate major histocompatibility complex (MHC) class II in latently infected GMPs , either it does not signal through the IL-10 receptor (IL-10R) or it engages the receptor in a different way than cIL-10 and cmvIL-10. Therefore, the mode of action and function during HCMV latency is uncertain.
Interestingly, analysis of viruses lacking the UL111A gene locus was found to be impaired in their ability to induce cIL-10 upon latent infection . Similarly, consistent with these observations, viruses lacking UL111A did not downregulate cellular hsa-miR-92a. Furthermore, recombinant LAcmvIL-10 was found both to induce cIL-10 and to cause the downregulation of hsa-miR-92a .
Taken together, then it appears that expression of LAcmvIL-10 during latent infection results in downregulation of cellular hsa-miR-92a, which in turn leads to the upregulation of the myeloid transcription factor GATA2. This increase in GATA2 then drives the expression of cIL-10, which inhibits intrinsic cell death signals and aids immune evasion of the latently infected cell (Fig. 3).
It is also worth emphasising that GATA2 is known to be involved in the hematopoiesis and myeloid cell differentiation . The extent to which latent infection of CD34+ cells, in itself, drives the latently infected progenitor cell down the myeloid lineage, rather than the lymphocyte lineage, is unclear. However, this could, in part, explain the fact that latent viral genomes have not been detected in T and B cells even though these presumably derived from the same CD34+ progenitor cells giving rise to cells of the myeloid lineage.
The ramifications of latency-associated changes in a number of other cellular miRNAs that have been identified during latent (Table 1 and ) infection are unclear. Similarly, latent infection also results in changes in a number of other secreted cellular proteins , and the effects of these changes on the latently infected cell will be enlightening. Regardless, it is now clear that latent infection imparts on the latently infected cells a plethora of phenotypic changes through an orchestrated manipulation of cell gene expression and cell functions. These are likely necessary for efficient carriage and reactivation of latent viral genomes, but they may also provide an ‘Achilles heel’ to allow the development of novel therapeutics to target and clear latent infection, at least in some clinical settings.
We thank Linda Teague, Roy Whiston, and Stuart McGregor Dallas for data and technical support and the MRC for funding, Grant Number G0701279. This research was supported by the Cambridge NIHR BRC cell phenotyping hub.
- 5.Sinclair J, Poole E (2014) Human cytomegalovirus latency and reactivation in and beyond the myeloid lineage. Future Virol 6:7Google Scholar
- 7.Seckert CK, Griessl M, Buttner JK, Scheller S, Simon CO, Kropp KA, Renzaho A, Kuhnapfel B, Grzimek NK, Reddehase MJ (2012) Viral latency drives ‘memory inflation’: a unifying hypothesis linking two hallmarks of cytomegalovirus infection. Med Microbiol Immunol 201(4):551–566. doi: 10.1007/s00430-012-0273-y CrossRefPubMedGoogle Scholar
- 11.Reeves M, Murphy J, Greaves R, Fairley J, Brehm A, Sinclair J (2006) Autorepression of the human cytomegalovirus major immediate-early promoter/enhancer at late times of infection is mediated by the recruitment of chromatin remodeling enzymes by IE86. J Virol 80(20):9998–10009CrossRefPubMedCentralPubMedGoogle Scholar
- 12.Groves IJ, Reeves MB, Sinclair JH (2009) Lytic infection of permissive cells with human cytomegalovirus is regulated by an intrinsic ‘pre-immediate-early’ repression of viral gene expression mediated by histone post-translational modification. J Gen Virol 90(Pt 10):2364–2374CrossRefPubMedGoogle Scholar
- 16.Rossini G, Cerboni C, Santoni A, Landini MP, Landolfo S, Gatti D, Gribaudo G, Varani S (2012) Interplay between human cytomegalovirus and intrinsic/innate host responses: a complex bidirectional relationship. Mediators Inflamm 2012:607276. doi: 10.1155/2012/607276 CrossRefPubMedCentralPubMedGoogle Scholar
- 27.Chaumorcel M, Lussignol M, Mouna L, Cavignac Y, Fahie K, Cotte-Laffitte J, Geballe A, Brune W, Beau I, Codogno P, Esclatine A (2012) The human cytomegalovirus protein TRS1 inhibits autophagy via its interaction with Beclin 1. J Virol 86(5):2571–2584. doi: 10.1128/JVI.05746-11 CrossRefPubMedCentralPubMedGoogle Scholar
- 32.Mason GM, Poole E, Sissons JG, Wills MR, Sinclair JH (2012) Human cytomegalovirus latency alters the cellular secretome, inducing cluster of differentiation (CD)4+ T-cell migration and suppression of effector function. Proc Natl Acad Sci USA 109(36):14538–14543CrossRefPubMedCentralPubMedGoogle Scholar
- 42.Friedlander MR, Lizano E, Houben AJ, Bezdan D, Banez-Coronel M, Kudla G, Mateu-Huertas E, Kagerbauer B, Gonzalez J, Chen KC, LeProust EM, Marti E, Estivill X (2014) Evidence for the biogenesis of more than 1000 novel human microRNAs. Genome Biol 15(4):R57. doi: 10.1186/gb-2014-15-4-r57 CrossRefPubMedCentralPubMedGoogle Scholar
- 43.Chiang HR, Schoenfeld LW, Ruby JG, Auyeung VC, Spies N, Baek D, Johnston WK, Russ C, Luo S, Babiarz JE, Blelloch R, Schroth GP, Nusbaum C, Bartel DP (2010) Mammalian microRNAs: experimental evaluation of novel and previously annotated genes. Genes Dev 24(10):992–1009. doi: 10.1101/gad.1884710 CrossRefPubMedCentralPubMedGoogle Scholar
- 49.Koga Y, Yasunaga M, Takahashi A, Kuroda J, Moriya Y, Akasu T, Fujita S, Yamamoto S, Baba H, Matsumura Y (2010) MicroRNA expression profiling of exfoliated colonocytes isolated from feces for colorectal cancer screening. Cancer Prev Res 3(11):1435–1442. doi: 10.1158/1940-6207.CAPR-10-0036 CrossRefGoogle Scholar
- 58.Hsu AP, Sampaio EP, Khan J, Calvo KR, Lemieux JE, Patel SY, Frucht DM, Vinh DC, Auth RD, Freeman AF, Olivier KN, Uzel G, Zerbe CS, Spalding C, Pittaluga S, Raffeld M, Kuhns DB, Ding L, Paulson ML, Marciano BE, Gea-Banacloche JC, Orange JS, Cuellar-Rodriguez J, Hickstein DD, Holland SM (2011) Mutations in GATA2 are associated with the autosomal dominant and sporadic monocytopenia and mycobacterial infection (MonoMAC) syndrome. Blood 118(10):2653–2655. doi: 10.1182/blood-2011-05-356352 CrossRefPubMedCentralPubMedGoogle Scholar
- 60.Hahn CN, Chong CE, Carmichael CL, Wilkins EJ, Brautigan PJ, Li XC, Babic M, Lin M, Carmagnac A, Lee YK, Kok CH, Gagliardi L, Friend KL, Ekert PG, Butcher CM, Brown AL, Lewis ID, To LB, Timms AE, Storek J, Moore S, Altree M, Escher R, Bardy PG, Suthers GK, D’Andrea RJ, Horwitz MS, Scott HS (2011) Heritable GATA2 mutations associated with familial myelodysplastic syndrome and acute myeloid leukemia. Nat Genet 43(10):1012–1017. doi: 10.1038/ng.913 CrossRefPubMedCentralPubMedGoogle Scholar
- 61.Todaro M, Zerilli M, Ricci-Vitiani L, Bini M, Perez Alea M, Maria Florena A, Miceli L, Condorelli G, Bonventre S, Di Gesu G, De Maria R, Stassi G (2006) Autocrine production of interleukin-4 and interleukin-10 is required for survival and growth of thyroid cancer cells. Cancer Res 66(3):1491–1499. doi: 10.1158/0008-5472.CAN-05-2514 CrossRefPubMedGoogle Scholar
- 67.Liu M, Aneja R, Liu C, Sun L, Gao J, Wang H, Dong JT, Sarli V, Giannis A, Joshi HC, Zhou J (2006) Inhibition of the mitotic kinesin Eg5 up-regulates Hsp70 through the phosphatidylinositol 3-kinase/Akt pathway in multiple myeloma cells. J Biol Chem 281(26):18090–18097. doi: 10.1074/jbc.M601324200 CrossRefPubMedGoogle Scholar
- 72.Ouyang P, Rakus K, van Beurden SJ, Westphal AH, Davison AJ, Gatherer D, Vanderplasschen AF (2014) IL-10 encoded by viruses: a remarkable example of independent acquisition of a cellular gene by viruses and its subsequent evolution in the viral genome. J Gen Virol 95(Pt 2):245–262. doi: 10.1099/vir.0.058966-0 CrossRefPubMedGoogle Scholar
- 79.Poole E, Avdic S, Hodkinson J, Jackson S, Wills M, Slobedman B, Sinclair J (2014) Latency-associated viral interleukin-10 (IL-10) encoded by human cytomegalovirus modulates cellular IL-10 and CCL8 secretion during latent infection through changes in the cellular microRNA hsa-miR-92a. J Virol 88(24):13947–13955. doi: 10.1128/JVI.02424-14 CrossRefPubMedCentralPubMedGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.