The regulation of KSHV lytic reactivation by viral and cellular factors

Abstract

Kaposi’s sarcoma-associated herpesvirus (KSHV) is an oncogenic herpesvirus that exhibits two distinct phases of infection in the host- latent and lytic. The quiescent latent phase is defined by limited expression of a subset of viral proteins and microRNAs, and an absence of virus production. KSHV periodically reactivates from latency to undergo active lytic replication, leading to production of new infectious virions. This switch from the latent to the lytic phase requires the viral protein regulator of transcription activator (RTA). RTA, along with other virally encoded proteins, is aided by host factors to facilitate this transition. Herein, we highlight the key host proteins that are involved in mediating RTA activation and KSHV lytic replication and discuss the cellular processes in which they function. We will also focus on the modulation of viral reactivation by the innate immune system, and how KSHV influences key immune signaling pathways to aid its own lifecycle.

Introduction

Herpesviruses are double-stranded DNA viruses that establish life-long, latent infection in their hosts, a hallmark state marked by an absence of active viral replication and virion production. Kaposi’s sarcoma-associated herpesvirus (KSHV), a member of the gammaherpesvirus family, is the causative agent of several neoplasms, namely, the endothelial cell malignancy Kaposi’s sarcoma, and B-cell lymphoproliferative disorders, primary effusion lymphoma (PEL) and multicentric Castlemans’s disease [1]. During latency, the viral genome, maintained as an episome, is copied into newly-divided progeny cells and expresses a repertoire of viral proteins and microRNAs from a specific latency-associated locus [2,3]. Specifically, a major player is latency-associated nuclear antigen (LANA) protein, which is expressed to ensure maintenance of the viral genome [2]. In addition, other latent viral proteins and microRNAs aid in promoting latent cell survival and immune evasion [4]. However, various stimuli can cause the virus to exit this latent phase. The key viral protein regulating this switch from latency to lytic replication is regulator of transcription activator (RTA). RTA is a nuclear DNA-binding protein required for KSHV viral reactivation, and its expression initiates the lytic cascade of KSHV gene expression. Subsequently, the lytic genes express proteins to facilitate genome replication and virion production [5]. Importantly, these lytic proteins promote the various diseases caused by KSHV and are essential for the production of new progeny virions that are required for viral spread through new infections. However, LANA and RTA do not act alone and engage various cellular and viral factors to promote these distinct phases of the viral lifecycle. In this review, we highlight the key findings aimed at understanding the cellular factors that modulate KSHV lytic reactivation and we discuss the host immune proteins that play a role in controlling lytic replication.

RTA: master regulator of lytic reactivation

It is well-established that the expression of RTA is indispensable for KSHV lytic reactivation and its expression alone is enough to prompt the initiation of the lytic cycle [6]. Once expressed, RTA binds to its own promoter to enhance its expression and transactivates other viral promoters, making RTA critical in the switch from latency to lytic infection [7]. Hence, processes that regulate the induction and stability of RTA play pivotal roles in this process. While the physiological conditions that cause KSHV reactivation are not fully understood, various environmental cues like hypoxia, and reactive oxygen species (ROS) can trigger the lytic cycle by inducing transcription factors to bind to the RTA promoter and, as a consequence, stimulating its transcription (Figure 1) [8,9]. For example, depletion of the transcription factor forkhead box O1 (FoxO1), an antioxidant protein, promoted KSHV lytic reactivation by inducing ROS [10]. Similarly, hypoxia-induced factor 1α (HIF1α) binds to the RTA promoter during hypoxia [11]. Furthermore, the activation of mitogen-activated protein kinase pathways (MAPK), important for cell survival under diverse cellular and environmental signals, can also promote KSHV lytic reactivation [12]. This pathway induces RTA expression by direct binding of the transcription factors, activator protein-1 (AP-1) and early growth response-1 (EGR1), to the RTA promoter [12,13]. These factors are phosphorylated downstream of MAPK signaling. Additionally, inflammatory mediators, such as histamine, also influence KSHV reactivation via the activation of the MAPK signaling pathways [14].

Figure 1: Induction of RTA and RTA-mediated regulation of lytic replication.

Transcription factors downstream of various stimuli induce RTA transcription and trigger lytic reactivation. Once transcribed, RTA mRNA is modified by m6A modifications, which is added and read by host proteins. RTA protein can also induce expression of host genes in addition to viral genes. Pro-viral genes are depicted in green, antiviral genes in red and genes that can act both pro-viral and antiviral are depicted in grey.

Once transcribed, the RTA mRNA transcript, encoded by the gene ORF50, is stabilized by post-transcriptional modifications conferred by host factors as illustrated by the addition of N6-methlyadenosamine (m⁶A) on the RTA mRNA [15,16]. While the enzyme N6-adenosine-methyltransferase 70 kDa subunit (METTL3) is responsible for depositing m⁶A modifications, YTH N6-methyladenosine RNA-binding protein-2 (YTHDF2) and Staphylococcal nuclease domain-containing protein-1 (SND1) are the reader proteins that recognize these m⁶A marks, and all these proteins are required for robust KSHV lytic reactivation [15,16]. However, evidence shows that METTL3 and YTHDF2 have opposing roles in different cell lines. Depletion of METTL3 and YTHDF2 reduced ORF50 mRNA in the latently-infected epithelial cell line iSLK219, but increased its expression during lytic reactivation in TREx-BCBL1-RTA cells, a PEL cell line with doxycyline-inducible RTA [16]. Thus, m⁶A modifications on RTA mRNA serve a differential role that is cell-type-dependent. Similar to RTA mRNA stability, the stability of RTA protein is also regulated by cellular proteins such as RTA-interacting protein nuclear receptor coactivator 2 (NCOA2). NCOA2 binds to protein abundance regulatory signal II (PARS II) domain of RTA and prevents its proteasomal degradation [17]. As such, knockdown or deletion of NCOA2 in latently infected cells reduces the expression of viral lytic proteins and virus production [17]. Also, deamidation of RTA by phosphoribosylformyl-glycinamidine synthetase (PFAS), a nucleotide metabolism enzyme, impairs its function by blocking its nuclear import and consequently reducing lytic reactivation [18].

Many host factors directly interact with RTA to induce transactivation of viral genes. Recombination signal sequence-binding protein jκ (RBP-jκ), a transcription factor functioning in the Notch signaling pathway, acts as an important mediator for RTA-induced lytic gene expression. Several viral genes directly activated by RTA contain RBP-jκ binding sites, and ablation of RBP-jκ reduces expression of such genes [19]. Similarly, during viral reactivation, RTA induces expression of numerous host genes in an RBP-jκ dependent manner. Two core RTA-induced host genes required for lytic reactivation are Geminin (GMNN), a cell cycle regulator, and GGT6, a glutathione metabolism enzyme [20]. Knockdown of either protein reduced the expression of viral lytic genes, viral DNA replication and production of infectious virions [20]. However, the mechanism of action of these RTA-induced genes has not yet been described. In summary, RTA can be modulated at various steps of its synthesis by cellular proteins and plays an integral role by inducing other host and viral genes during lytic reactivation (Figure 1).

Epigenetic modifications leading to activation of the RTA promoter

Epigenetic modifications at genomic sites modulate chromatin architecture and allow genome accessibility for transcription. Transcription-promoting epigenetic marks include histone acetylation and specific histone methylation patterns (H3K4me3 for promoters) on histone lysines. The latent KSHV genome is heavily chromatinized and contains both repressive and active epigenetic marks. Depletion of chromatin organization proteins, CTCF and Rad21, increases general viral transcription and enhances virion production during lytic replication in an RTA-independent manner [21,22]. However, reactivation requires an increase in active epigenetic marks on the viral genome, including the RTA/ORF50 promoter. Treatment of latently-infected cells with sodium butyrate, a histone deacetylase inhibitor, increases histone acetylation at the RTA promoter [23]. As such, histone acetylation and removal of nucleosomes from the RTA promoter increase KSHV lytic reactivation in latently-infected cell lines [24]. On the other hand, polycomb repressive complexes 1 and 2 (PRC), mammalian complexes that mediate transcriptional repression in cells, inhibit KSHV lytic reactivation [25,26]. Hence, depletion of PRC members Bmi1 and Enhancer of zeste homolog 2 (EZH2) results in KSHV reactivation from latency [24,27]. In addition, similar to the action of sodium butyrate, bacterial metabolites drive the reactivation of KSHV via inhibition of histone deacetylases [23,28]. Finally, other LANA-bound chromatin modulators like Kruppel-associated-box associated protein-1 (KAP1) and nuclear factor erythroid 2-related factor-2 (Nrf2) are recruited to the ORF50 promoter to repress transcription and their depletion result in KSHV lytic reactivation [29].

Histone demethylases also play a significant role in KSHV lytic reactivation. Overexpression of Jumonji domain-containing protein D3 (JMJD3) and ubiquitously transcribed tetratricopeptide repeat X (UTX), histone demethylases that remove the repressive H3K27me3 mark, causes lytic reactivation of KSHV in latent cells [27]. Interestingly, JMJD3 and UTX interact with PAN RNA, a highly abundant viral RNA transcribed in the lytic cycle, which recruits these factors to the RTA promoter to remove repressive chromatin marks [30]. Conversely, lysine demethylase 2B (KDM2B), a histone demethylase that removes active histone marks, binds to the RTA promoter to inhibit its expression [31]. Additionally, the DNA-sensing protein interferon-gamma inducible protein 16 (IFI16) has been shown to suppress lytic reactivation, as its knockdown in latent PEL cell lines induced lytic gene transcription [32]. IFI16 acts by increasing repressive histone methylation (H3K9me3) on the viral genome to promote latency [33]. Likewise, polo-like kinase-1 (PLK1) helps maintain the inactive methylation mark (H3K27me3) at RTA and other viral promoters to ensure latency [34]. Interestingly, depletion of sirtuin-1 (SIRT1), a histone deacetylase, caused an increase in active methylation marks and a concomitant decrease in repressive methylation marks at the RTA promoter without altering acetylation levels [35]. Histone phosphorylation is another epigenetic modification that influences latent to lytic cycle transition. Furthermore, depletion of tousled-like kinase-2 (TLK2) induced reactivation of KSHV in latently-infected 293 cells and PEL cell lines as a result of decreased histone H3 phosphorylation due to loss of TLK2 at the ORF50/RTA promoter [36]. Taken together, the increase in activating epigenetic marks results in enhanced RTA expression and subsequent lytic viral gene expression (Figure 2A).

Figure 2: Epigenetic modifications, DNA damage response and cellular stress pathways influence lytic reactivation.

(A) Epigenetic factors regulate the addition of histone modifications on RTA and the viral genome. Host factors shown on the left inhibit lytic reactivation by adding H3K27me3, H3K9me3 and promoting phosphorylation of histone H3. Transcriptional repressors (KAP1 and Nrf2) inhibit transcription and IFI16 promotes repressive H3K9me3 marks on the viral genome. Conversely, on the right, JMJD3 and UTX remove the transcription-suppressive H3K27me3 mark and promote lytic reactivation. Dotted lines depict the removal of histone mark. (B) Host DNA repair proteins and the DNA damage response promote lytic reactivation. Cellular ER stress sensors are pro-viral for lytic replication and reactivation while non-sense mediated decay proteins suppress lytic reactivation. Pro-viral genes are depicted in green and antiviral genes in red.

DNA replication and DNA damage response proteins modulate KSHV RTA expression

The DNA damage response (DDR) is a coordinated network of host signaling proteins that recognize DNA lesions and double-stranded breaks during DNA replication. As such, viral replication provides substrates to activate the DDR response which can influence KSHV reactivation [37,38]. Poly (ADP ribose) polymerase-1 (PARP1), a key mammalian DNA repair protein recruited to DNA lesions, binds to and modifies RTA resulting in inhibition of RTA activity [39]. During KSHV lytic reactivation, PARP1 is degraded by ORF59 through recruitment of a host E3 ubiquitin ligase, checkpoint with FHA and RING finger domains (CHFR), to allow for efficient RTA function [40,41]. Consequently, knockdown of CHFR reduces expression of KSHV lytic proteins during lytic reactivation [41]. Hence, KSHV usurps a host E3 ubiquitin ligase to degrade PARP1 and allow for efficient KSHV lytic reactivation [41]. Additionally, KSHV requires the tumor suppressor and DDR protein p53 and its downstream targets p21 and TRIML2 to overcome cellular stress during lytic replication [38,42]. While the exact mechanism for TRIML2 is unclear, p21 acts by promoting cell cycle arrest upon DDR activation [38,42]. However, other DDR proteins such as MRE11 and Ku80 localize to the viral replication compartments and have opposing roles in KSHV lytic reactivation. MRE11 is a member of a double-stranded break sensing complex and activates the ataxia-telangiectasia mutated (ATM) DNA repair pathway. Inhibition or knockdown of MRE11 decreases KSHV lytic replication upon reactivation in B cells [43]. On the other hand, decreasing Ku80, a non-homologous end joining (NHEJ) DNA repair protein, increases KSHV virus production during reactivation [43].

In addition to DNA repair proteins, host factors like the mini chromosome maintenance proteins (MCM), components of a DNA helicase complex that is required for cellular genome replication, also influence KSHV lytic reactivation. This is seen with the depletion of MCM6, a binding partner of viral processivity factor ORF59, which led to an enhancement in viral genomes and late gene expression during lytic reactivation [44]. While KSHV and other herpesviruses encode their own DNA replication machinery for viral replication, host factors can also regulate this process. For example, RTA induces N-myc downstream regulated gene 1 (NDRG1) to bind and stabilize the viral DNA helicase ORF44 [45]. As expected, abrogation of NDRG1 during lytic reactivation reduced viral genome copy numbers intracellularly and decreased virion production [45]. Thus, cellular factors regulate lytic reactivation through the DDR pathway and by modulating viral replication protein expression (Figure 2B).

Cellular stress and non-sense mediated decay pathways that control RTA activation

Herpesviruses are adept at commandeering the cellular machinery to promote translation of viral proteins. As the endoplasmic-reticulum (ER) is the site for membrane-associated protein folding, the increased production of viral proteins during the lytic cycle of KSHV can induce ER stress. In order to alleviate the ER stress and reduce misfolding of proteins, cells initiate the unfolded protein response (UPR) via activated transcription factor 6 (ATF6), protein kinase R-like ER kinase (PERK) and inositol-requiring enzyme 1(IRE1). All three proteins are activated upon KSHV lytic reactivation, and depletion of these UPR proteins reduced virus replication and production [46]. However, pathways downstream of these UPR proteins were not activated during lytic reactivation [46]. In addition, the cellular UPR pathway also directly transactivates RTA. The induction of IRE1 causes the X-box binding protein (XBP) to be spliced to its active isoform (sXBP) and directly bind and activate the RTA promoter [47,48]. Additionally, the mammalian target of rapamycin complex (mTORC), a nutrient-sensing pathway that modulates protein translation, is important for early viral protein synthesis, but not for late proteins during lytic replication [49].

During translation, premature stop codons and long 3’ untranslated regions (UTRs) in mRNA stall ribosomes that activate the non-sense mediated decay (NMD) to degrade aberrant mRNA transcripts. Lytic reactivation of KSHV is enhanced by inhibition of UPF1 and UPF3X, components of the NMD pathway [50,51]. Interestingly, NMD factor UPF1 is directly recruited to the RTA transcript due to its 3’UTR structure and acts in concert with UPR-activated protein sXBP1 [50]. Hence, cellular stress proteins are optimally utilized by KSHV for lytic replication during reactivation (Figure 2B).

Innate immune modulation of KSHV reactivation and replication

The innate immune sensors in cells detect pathogen-associated molecular patterns (PAMPs) during viral infection and initiate downstream signaling pathways to generate antiviral responses. While most limit viral infection, KSHV has evolved to utilize some of these pathways. During KSHV lytic reactivation, nucleic acid-sensing pathways are activated and modulate KSHV lytic reactivation. The cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) axis is a major DNA-sensing pathway, in which cGAS senses double-stranded DNA and catalyzes the production of 2’,3’ cyclic GAMP to serve as an intracellular signal and activate ER-resident STING protein [52]. Upon STING activation, the NFκB and tank-binding kinase-1 (TBK1)/interferon regulatory factor-3 (IRF3) pathways are stimulated, leading to the production of type I interferon [52]. While knockdown of cGAS or STING increases KSHV lytic reactivation, highlighting the antiviral activity of this pathway against KSHV, the virus also encodes vIRF1/K9 to block STING activity during lytic reactivation by interfering with the TBK1-STING interaction [53]. Furthermore, the protein phosphatase catalytic subunit-6C (PPP6C), an interacting partner of KSHV-encoded ORF48, facilitates KSHV lytic reactivation by dephosphorylating and inactivating STING [54].

RNA-sensing pathways also function in KSHV lytic reactivation due to the production of double-stranded RNA (dsRNA) that serves as a PAMP for RNA sensors [55,56]. Consequently, cytosolic RNA sensors, retinoic acid-inducible gene 1 (RIG-I) and melanoma differentiation-associated-5 (MDA5), limit KSHV lytic reactivation via mitochondrial antiviral signaling (MAVS)-dependent signaling, and knockdown of either protein leads to an increase in KSHV lytic gene expression, lytic proteins and virion production [55,56]. On the other hand, other RNA-binding proteins, such as RNA-editing enzyme adenosine deaminase acting on RNA-1 (ADAR1), play a pro-viral role during KSHV lytic reactivation [57]. The ablation of ADAR1 caused decreased lytic gene transcription and viral replication due to enhanced interferon β (IFN-β) expression through the RIG-I- and MDA5-MAVS axis [57]. In addition to cytosolic sensors, Toll-like receptors (TLRs), endosomal or cell membrane-associated proteins, also sense various PAMPs, and stimulation of single-stranded RNA (ssRNA)-sensing TLRs 7 and 8 induced RTA promoter activation and KSHV lytic reactivation in PEL cells [58]. Furthermore, the activation of TLR3, a double-stranded RNA sensor induced upon KSHV infection, increases RTA expression mediated by the adaptor protein Toll-interleukin-1 receptor (TIR) domain-containing adaptor inducing β-interferon (TRIF) [59,60]. Interestingly, once expressed, RTA degrades TRIF suggesting a regulatory loop where TLR3 signaling supports lytic reactivation via RTA and KSHV subsequently shuts down the TLR3 signaling to prevent antiviral innate immune responses [61].

Nucleotide binding and oligomerization domain (NOD)-like receptors (NLRs) are a family of innate immune proteins that oligomerize into signaling complexes upon ligand sensing to produce pro-inflammatory cytokines. While most members function in inflammasome activation to induce IL-1 and IL-18, others signal through alternative pathways. NLRP1, an inflammasome-activating member, plays an antiviral role during KSHV lytic reactivation by inducing IL-1 and IL-18. Hence, knockdown of NLRP1 results in increased viral gene expression and virus production [62]. However, another NLR member, NLRX1, is pro-viral during KSHV lytic reactivation, as it modulates the cytosolic RIG-I-like receptor (RLR) adaptor, MAVS, to suppress type I interferon (IFN-β) expression, and NLRX1 knockdown hampered KSHV lytic reactivation in both latent epithelial and PEL cell lines [63]. During KSHV lytic reactivation, apoptosis-inducing family of caspase proteases also serve as a modulator of IFN-β production. As a consequence, blocking caspase activity impaired KSHV lytic reactivation by enhancing IFN-β expression and interferon-stimulated genes (ISGs) [64].

Once interferons are induced by activation of PRR signaling, they activate transcription of many ISGs in both an autocrine and paracrine manner. The wave of ISGs expressed downstream of interferons restrict viruses at different stages of replication and as a consequence acts on KSHV during lytic reactivation. For example, interferon-induced tetratricopeptide (IFIT) proteins, induced upon reactivation, restrict KSHV lytic replication by augmenting ISG expression, and as a result, knockdown of these proteins led to an increase in virion production [65]. However, this IFIT-induced restriction of KSHV lytic replication was only limited to the epithelial iSLK219 cells [65]. In addition, lytic reactivation causes an increase in global protein ISGylation, an IFN-induced modification of proteins with the addition of ubiquitin-like ISG15. Thus, knockdown of ISG15 and HERC5, the cellular enzyme required for ISGylation, led to increased virus production [66]. Thus, innate immune pathways are activated and regulate lytic replication during KSHV reactivation (Figure 3).

Figure 3: Innate immune proteins and KSHV lytic reactivation.

The major DNA (cGAS, STING) and RNA (RIG-I, MDA5, MAVS, TLR7/8) sensing pathways and their components act to limit KSHV lytic reactivation. PPP6C modulates STING phosphorylation while ADAR1 and NLRX1 modulate the RIG-I/MDA5/MAVS pathway. NLR protein, NLRP1, also acts antivirally through pro-inflammatory cytokines. Caspase-8 has been demonstrated to limit the expression of IFN-β. Downstream of IFN, the ISGs, IFITs and ISG15, inhibit viral reactivation. Pro-viral genes are depicted in green and antiviral genes in red.

Conclusions

The studies discussed in this review highlight the virus-host interactions required for KSHV reactivation and their mechanism of action as well as pathways that control KSHV RTA expression. Lytic reactivation is crucial for virus production and shedding light on the processes that are essential for reactivation will help to identify new targets for therapeutic intervention. Hence, it is imperative to continue to investigate the intricate relationship between KSHV and the host to find suitable ways to inhibit this persistent oncogenic virus, curb its production and, thus, reduce the KSHV-associated disease burden.