miRNAs in the pathogenesis of oncogenic human viruses

Abstract

Tumor viruses are a class of pathogens with well established roles in the development of malignant diseases. Numerous bodies of work have highlighted miRNAs (microRNAs) as critical regulators of tumor pathways and it is clear that the dysregulation of cellular miRNA expression can promote tumor formation. Tumor viruses encode their own miRNAs and/or manipulate the expression of cellular miRNAs to modulate their host cell environment, thereby facilitating their respective infection cycles. The modulation of these miRNA responsive pathways, however, often influences certain signal transduction cascades in ways that favor tumorigenesis. In this review, we discuss the roles of virally-encoded and virally-regulated cellular miRNAs in the respective viral life-cycles and in virus associated pathogenesis.

1. Introduction

MiRNAs (microRNAs) are a newly-discovered class of gene regulators that are single-stranded non-protein-coding RNAs with an average length of 21-22 nucleotides. The first characterized miRNA, lin-4, was originally discovered in roundworm (Caenorhabditis elegans) by Lee and colleagues in 1993 [1]. Lin-4 was shown to regulate roundworm development and was predicted to function through an RNA-RNA interaction [1]. At that time, however, lin-4 was only viewed as a unique regulator in worms and the importance of this discovery did not receive broad recognition. The dawn of the miRNA era occurred approximately eight years later when three groups simultaneously reported the identification of miRNAs as a class of regulators in the journal Science [2; 3; 4]. Since this time, the list of identified miRNAs has grown rapidly and miRNAs are now appreciated as a highly conserved and evolutionarily important component of living organisms [3]. Currently, miRNAs have been identified in almost all studied multi-cellular eukaryotes in the plant and animal Kingdoms. Recent studies demonstrated that miRNAs can also be encoded by certain single-cell eukaryotes and by viruses [5; 6]. The current total number of identified miRNA genes is ∼14197 including ∼940 human miRNAs (according to the Sanger miRBase 2010 Release 15). A much higher number of miRNAs is predicted by computational approaches [7] and it is estimated that miRNAs account for 1-5% of expressed genes in animal genomes [8; 9]. At this level, miRNAs would become one of the most abundant classes of gene regulators in higher eukaryotes [8; 9].

Since the discovery of miRNAs, the way in which they execute their function has been the subject of intense study. The central dogma states that miRNAs regulate their targets primarily through base complementary interactions. Under the guidance of the “seed” region (the 5′ nucleotides 2-8 of the mature miRNA), miRNAs typically interact with the 3′-UTR (untranslated region) of their mRNA targets and less frequently through target sites within the coding region or the 5′-UTR. This interaction occurs in the context of a protein nucleic acid complex referred to as miRISC (miRNA induced silencing complexes) and the formation of this complex on target mRNAs results in the inhibition of protein production through a number of mechanisms including degradation of the target mRNA and inhibition of translation [7; 8].

Genetically, miRNA genes are located within introns of protein-coding or non-protein-coding genes or in intergenic regions as stand alone genes. MiRNAs can be found as either a single unit or as clusters of more than one miRNA. Clustered miRNAs are usually expressed together as polycistronic transcripts and their targets are often predicted to have similar functions. Based on the similarity of the 5′ seed sequence, some miRNAs are separated into families, such as the miR-200 family of miRNAs which includes miR-200a/b/c, miR-141, and miR-429 which are transcribed from two distinct genomic loci.

Although there is only a small percentage of human miRNAs that are currently being investigated, miRNAs are predicted to regulate approximately one third of human mRNAs [7; 10]. Each miRNA can have multiple target transcripts, and each transcript can be silenced by multiple miRNAs leading to an extremely complicated regulatory system. Due in part to the pleiotropic nature of miRNA targeting, it is likely that they are involved in the regulation of nearly every signal transduction pathway utilized by the cell. To date, miRNAs have already been shown to exhibit regulatory functions in numerous cellular pathways and cellular processes including proliferation, differentiation, and apoptosis [10].

The development of cancer is a multi-factorial and multi-step process. As widespread regulators of many if not all signaling pathways, it is not a stretch to expect miRNAs to be involved in the regulation of oncogenic pathways. In fact, accumulating evidence indicates that disruption of miRNA expression correlates with most types of human malignancies including breast cancer [11], ovarian cancer [12], cervical cancer [13], prostate cancer [14], colorectal cancer [15], gastric cancer [16], nasopharyngeal carcinoma [17], lung cancer [18], renal cancer [19], and hematopoietic malignancies [20]. Human miRNAs are frequently located in genomic regions linked to cancer such as fragile sites, minimal regions of loss of heterozygosity, minimal regions of amplification, and common breakpoint regions, further implicating them in tumorigenesis [21]. MiRNAs can function as either oncogenes (oncomirs) or tumor suppressors through the inhibition of critical protein-coding genes involved in tumorigenesis. Informatics studies have suggested the possibility that miRNAs may also regulate the development of certain cancers by targeting non-protein-coding RNAs such as those transcribed from ultraconserved regions (UCRs) [22]. While these studies may have important implications in cancer biology, experimental validation of this concept is nevertheless pending. Overall, the study of detailed mechanisms through which miRNAs contribute to cancer is in its nascent stages, yet it is clear that the discovery of miRNAs has created a new dimension for us to understand the mechanism of tumorigenesis.

2. Viruses, miRNAs and cancer

A hallmark of viruses is their incredible mastery of the art of cellular hijacking to achieve their own tenets of self-preservation and proliferation. With their limited genetic capacity, they rely exquisitely on a wide range of host functions to achieve these tenets. Nevertheless, they seem never content to simply utilize existing cellular functions in their natural state but instead modulate and fine-tune each of these pathways/functions in ways that provide a greater advantage for the virus. Given the now well established role of miRNAs in regulating almost every aspect of the cellular regulatory machinery, it is not surprising to find viruses interacting with these pathways as a means of modulating the host environment to suit their needs. Although the study of how viruses utilize miRNA pathways in their life cycle is in its infancy, we know that this intersection with miRNA pathways occurs through virus-mediated modulation of existing cellular miRNA expression and/or through the evolution of a virus-encoded repertoire of miRNAs.

From an evolutionary standpoint, the genesis of a virus-encoded miRNA seems quite tractable since there are only a handful of nucleotides that are needed to confer an appropriately processed functional miRNA molecule. Yet this relatively simple evolutionary process can potentially capacitate virus mediated regulation of hundreds of cellular (and/or viral) genes. This in part, may explain the high relative abundance of virus-encoded miRNAs for some viruses such as the Epstein-Barr virus which encodes at least 44 mature miRNAs in a genome that is more than 16,000 times smaller than the human genome. This de novo evolution of functional virus-encoded miRNAs is a little more difficult than it may seem, because, in addition to the presence of seed complementarity of the target mRNAs, there are flanking sequence and secondary structure constraints on the target RNAs that determine whether the respective site is potentially susceptible to miRNA targeting. Therefore, finding targetable sequences common to a repertoire of cellular mRNAs whose inhibition provides an advantage to the virus may be a challenge. Although many viruses have, in fact, evolved with their own functional miRNA genes, they also utilize the mechanism of modulating cellular miRNAs to influence a repertoire of mRNA targets.

There is a third situation which is a hybrid of the two scenarios mentioned above. In this case, the virus appears to have evolved to encode miRNAs that target an already existing miRNA/targetome pathway. Salient examples of this is the Kaposi's sarcoma herpes virus (KSHV) encoded miR-K12-11 and Marek's disease virus (MDV-1) encoded miR-M4 which have the identical seed sequence as that of the cellular miRNA, miR-155. Having identical seed sequences, these virus-encoded miRNAs target a highly overlapping set of genes with that of miR-155 [23; 24; 25]. This scenario is even more interesting, however, because another herpesvirus, the Epstein-Barr virus, has evolved with a mechanism to simply modulate the host miR-155 expression to suit its life cycle.

It appears, then, that the miR-155 pathway is particularly well suited to the life cycle needs of these three herpesviruses but these viruses have intersected with this pathway in two different ways. Notably, all three of these herpesviruses are oncogenic while the inappropriate expression of cellular miR-155 is similarly known to be oncogenic outside of any viral context. It is not too hard to imagine then, that while a miR-155-like function is integral to some aspect of the life cycle of these viruses, the inappropriate activation of a miR-155 function during the virus life cycle may partly explain the oncogenic tendency of these viruses.

Of all the viruses studied to date, herpesviruses are the hands down winners in terms of the number of miRNAs encoded, with herpes viral miRNAs accounting for more than 97% of all known virus encoded miRNAs. The explanation for this extreme bias may lie at least partly in a defining characteristic of herpesviruses: their ability to exist in a latent phase. In this state, herpesviruses need to remain relatively undetectable to the immune system in order to avoid total virus clearance. They cannot, therefore, express foreign viral proteins that can potentially be processed and presented on the cell surface to attract an immune response. Nevertheless, herpesviruses still need to control/regulate themselves as well as the cellular host during latency. MiRNAs may be a particularly well suited means for herpesviruses to regulate themselves as well as their host cell environments during latency when they need to exist in an immunologically transparent state.

Viruses integrate into the cellular miRNA machinery in a variety of ways to achieve pathogenic success. As miRNAs regulate pathways that are such critical mediators of cellular signaling and cancer, pathogenic success is not always that far away from oncogenesis. In conjunction with host cell genetic abnormalities and/or other exogenous agents, the balance between pathogenic success and cancer can be tipped towards the latter. Below we have outlined many of the known miRNAs and affected pathways that are influenced by viruses and we discuss their possible relevance in the genesis of virus induced cancers.

3. Epstein-Barr virus

Epstein-Barr virus (EBV, Human Herpesvirus 4 (HHV-4)) is closely associated with a number of human tumors including Burkitt's lymphoma, Hodgkin's disease, lymphoproliferative disease in immuno-suppressed patients, nasopharyngeal carcinoma, gastric carcinoma, and renal cell carcinoma [26]. EBV is a member of the human herpesviridae family, of which Herpes Simplex virus (HSV), Cytomegalovirus (CMV), and Kaposi's sarcoma herpes virus (KSHV) are also members. After initial infection, EBV, like other herpesviruses, exhibits life-long latency in its host during which the viral genome exists as an episomal closed circular plasmid DNA and only a restricted set of viral genes are expressed. Viral reactivation (the lytic cycle) is initiated periodically in response to a variety of physiological signals. Reactivation leads to the expression of ∼100 viral genes and culminates in the production of infectious virions that are released and available to infect new host cells. The virus is transmitted mainly through salival exchange and currently, more than 90% of the world's adult population are persistent carriers of EBV [27]. Although the exact mechanism for EBV primary infection is still not clear, current theory holds that during primary infection, EBV initially infects the oropharyngeal epithelial cells which are permissive for viral lytic infection. Productive replication in this epithelial cell population leads to the latent infection of nearby B-cells in which the full complement of EBV latency proteins are expressed (referred to as type III latency) [28]. Expression of this full repertoire of latency genes induces B-cell activation/proliferation and facilitates a transient virus driven expansion of the latently infected B-cell pool. In a small population of circulating B-cells, EBV establishes a more restricted form of latent infection in which either one (type I latency) or no (type O latency) viral protein is expressed. Once an immune response is mounted, the lytic infected epithelial cells, as well as the expanding type III latency B-cell population, are mostly eliminated. The remaining small population of infected circulating B-cells, exhibiting restricted latency gene expression, will then persist for the lifetime of the host. It is thought that there are occasional transitions to type III latency in some B-cells, as well as sporadic reactivation events in epithelial tissue, whereby EBV is produced and secreted into the saliva for transmission to a new host.

EBV encodes its own repertoire of miRNAs and it modulates the expression of cellular miRNAs in a concerted effort to regulate its life cycle. While EBV uses miRNAs to regulate pathways in ways that promote its life cycle, virus survival has many of the same needs as tumor survival. In conjunction with additional factors including host genetic alterations and/or a compromised host immune system, the expression of EBV encoded miRNAs and/or the dysregulation of cellular miRNAs can lead to virus associated tumors through the miRNA dependent modulation of cancer pathways.

3.1. EBV-encoded miRNAs

Due to the non-immunogenic property of these tiny RNA molecules, viral miRNAs are good candidates to execute various latency associated tasks once an immune response to viral antigens has been mounted by the host. This may partly explain the common occurrence of virus-encoded miRNAs among the herpesvirus family. In addition, however, the biosynthesis cascade involved in the production of miRNAs is preferentially suited to nuclear DNA viruses (although miRNAs generated by RNA retrovirus (e.g. HIV-1) have also been identified (although this assertion remains controversial) [29]). To date, virus-encoded miRNAs have been found in the genomes of a number of human and non-human viruses including EBV, KSHV, CMV, HSV, SV40, rLCV, rhesus monkey rhadinovirus, herpes B virus, MHV68, mouse CMV, MDV, and HIV-1. As shown in Table 1, the herpesvirus family encodes most of the currently identified viral miRNAs and EBV, the first human virus reported to express its own miRNAs [6], is one of the leading viral miRNA encoders.

Table 1. List of currently identified virus encoded microRNAs.

Viruses	Virus-encoded microRNAs	References
Adenovirus	svaRNAs (VA RNA derived microRNAs)	[139; 140]
BK virus	bkv-mir-B1	[141]
Bovine herpesvirus 1	bhv1-mir-B1 to bhv1-mir-B10	[142]
CMV	hcmv-mir-UL22A, hcmv-mir-UL36, hcmv-mir-UL112, hcmv-mir-UL148D, hcmv-mir-US5-1, hcmv-mir-US25-1, hcmv-mir-US5-2, hcmv-mir-US25-2, hcmv-mir-US33, hcmv-mir-US4, hcmv-mir-UL70	[112; 143; 144]
EBV	miR-BART1 to 22, miR-BHRF1-1 to 1-3	[6; 41; 44; 45; 46]
HSV-1	hsv1-miR-H1 to H8, hsv1-miR-H11 to H18	[145; 146; 147; 148]
HSV-2	hsv2-mir-H2 to H7, hsv2-mir-H9 to H13, hsv2-mir-H19 to H25	[148; 149; 150]
Heliothis virescens ascovirus	HvAV-miR-1	[151]
Herpes B virus	hbv-mir-B2RC, hbv-mir-B4, hbv-mir-B20	[152]
Herpesvirus of Turkeys	hvt-mir-H1 to hvt-mir-H5, hvt-mir-H7 to hvt-mir-H18	[153; 154]
HIV	hiv1-mir-N367, hiv1-mir-TAR, hiv1-mir-H1	[155; 156; 157; 158; 159]
Infectious laryngotracheitis virus	iltv-mir-l1 to iltv-mir-l7	[153; 160]
JC virus	jcv-mir-J1	[141]
KSHV	kshv-miR-K12-1 to kshv-miR-K12-12	[44; 111; 112; 115; 161]
MDV	mdv1-mir-M1 to mdv1-mir-M13, mdv1-mir-M31	[162; 163; 164]
Merkel cell polyomavirus	mcv-mir-M1	[165]
Mouse CMV	mcmv-mir-m01-1, mcmv-mir-m01-2, mcmv-mir-m01-3, mcmv-mir-m01-4, mcmv-mir-m21-1, mcmv-mir-m22-1, mcmv-mir-M23-1, mcmv-mir-M23-2, mcmv-mir-M44-1, mcmv-mir-M55-1, mcmv-mir-m59-1, mcmv-mir-m59-2, mcmv-mir-M87-1, mcmv-mir-m88-1, mcmv-mir-M95-1, mcmv-mir-m107-1, mcmv-mir-m108-1, mcmv-mir-m108-2	[166]
MHV68	miR-M1-1 to miR-M1-9 (derived from tRNA precursors)	[112]
Murine polyomavirus	MR1	[167]
rLCV	rlcv-mir-rL1-1 to rlcv-mir-rL1-35	[41; 42; 43]
rhesus monkey rhadinovirus	rrv-miR-rR1-1 to rrv-miR-rR1-7	[168]
Polyomavirus SA 12	Not yet named	[169]
SV40	sv40-mir-S1	[170]

EBV encoded miRNAs are derived from two different clusters within the viral genome. Three viral miRNAs are encoded within the BamHI H rightward open reading frame (BHRF) region and the remaining miRNAs are found within the BamHI A rightward transcript (BART) region. The EBV BART gene was originally identified by Hitt and colleagues in the EBV-infected nasopharyngeal carcinoma cell line, C15 [30]. In the years following, BART transcripts were detected in both healthy viral carriers and in many other EBV-associated diseases including oral hairy leukoplakia, Burkitt's lymphoma, Hodgkin's lymphoma, gastric carcinoma, as well as nasal natural killer/T-cell lymphomas, [31; 32; 33; 34; 35]. BART sequences are largely conserved between EBV strains with the exception of a large 12-kb deletion spanning most of the BART locus which was identified in the B95-8 laboratory EBV strain [36]. For many years, BART transcripts were thought to function primarily through open reading frames within the spliced forms of these transcripts. Ectopic expression of open reading frames encoding two of these proteins, RPMS1 and A73, has been shown to modulate cellular signaling pathways relevant to tumorigenesis [37; 38]. To date, however, the endogenous expression of neither of these proteins has been observed in virus infected cells [37; 38; 39].

EBV BART miRNAs are derived from a large intron within BART transcripts prior to splicing [40]. The identification of BART miRNAs was first accomplished by Pfeffer and colleagues [6] in the EBV B95-8 strain. This group successfully cloned five mature EBV encoded miRNAs, including miR-BART1 and miR-BART2, from the BART locus, and the three BHRF miRNAs, miR-BHRF1-1, miR-BHRF1-2, and miR-BHRF1-3, located 5′- and 3′ of the BHRF1 (BamHI H Rightward ORF1) open reading frame [6]. Later, using wild-type EBV infected cell lines, another group extended the number of EBV encoded miRNAs to seventeen, including fourteen BART miRNAs and all three of the previously identified BHRF1 miRNAs [41]. Interestingly, the identification of miRNAs from the rhesus lymphocryptovirus (rLCV), an evolutionarily distant relative of EBV (∼23 million years evolutionary divergence or ∼13 million years ago), revealed a number of miRNAs that show strong conservation in terms of both miRNA sequence and genomic positions [41; 42; 43]. This intimates an evolutionarily conserved role for these miRNAs in the life cycles of these viruses.

In a separate study, Grundhoff and colleagues [44] identified a total of twenty-two mature miRNAs in wild-type EBV infected cell lines through a combination of computational bioinformatics, microarray, and northern blot approaches. Recently, an additional three miRNAs, miR-BART21-5p, BART21-3p, and miR-BART22, were identified by other groups [45; 46] bringing the total number of EBV encoded mature miRNAs to ∼40.

3.2. Regulation of EBV-encoded miRNA expression

EBV exhibits tissue-specific expression of its miRNAs. By studying nasopharyngeal carcinoma (NPC) tissues, Zhu and colleagues found that EBV-miR-BARTs were expressed in all NPC tissues examined, but EBV-miR-BHRF1s were not [45]. Similarly, in another epithelial cell model system, Kim do et al. [47] found that EBV-miR-BARTs, but not EBV-miR-BHRF1s were expressed in EBV-associated gastric carcinoma cell lines and gastric carcinoma tumor tissues.

In B-cells, the pattern of EBV miRNA expression appears to be somewhat dependent on the latency type. In one report, expression of miR-BART2 but not miR-BHRF1-3 was detected in some primary Burkitt's lymphoma samples (which exhibit type I latency), whereas miR-BHRF1-3 expression was found to be high in type III latency cell lines and AIDS-related large B-cell lymphoma samples (which typically exhibit type III latency gene expression) [48]. Consistently, another study has demonstrated that BHRF1 miRNAs are expressed at high levels in type III latency cells but are undetectable in cells displaying a viral type I/II latent infection [41].

In addition to an apparent latency type dependency of EBV miRNA expression, induction of the EBV lytic cycle has been shown to correlate with the induction of certain EBV miRNAs (such as miR-BART1-3p, miR-BART3-3p, miR-BART7, miR-BART10-3p, and miR-BHRF1-2, and to a lesser extent miR-BHRF1-1) in the type I latency Burkitt's lymphoma cell line Mutu I [41]. These observed expression differences between different stages of the EBV life cycle may reflect stage specific miRNA function. Nevertheless, future studies will be needed to assess this contention.

Theoretically, expression of miRNAs can be regulated at every step within its biosynthesis pathway (e.g. in the transcriptional initiation step or during the processing of miRNA precursors). Currently, the mechanisms involved in regulating EBV miRNA expression are not all that clear. Some studies have demonstrated that all the miR-BARTs are derived from the same transcript [40] and there is a coordinate expression of the BART mRNAs and the miR-BARTs [41]. Therefore, it is likely that the activity of the BART promoter is a primary determinant of the overall expression levels of miR-BARTs. Despite this, however, it is clear that different BART miRNAs are expressed at different levels within the same cell, probably due to varying accessibility of individual BART miRNA sequences to the microprocessor machinery [49]. Since nucleotide polymorphisms can affect the processing of miRNAs, variable relative abundances of individual miR-BARTs from one cell to another may result from sequence variations of miR-BARTs precursors [40; 50].

In contrast to miR-BARTs, miR-BHRF1s appear to be processed from two different precursors with miR-BHRF1-1 being derived from EBNA (Epstein-Barr nuclear antigen) transcripts and miR-BHRF1-2 and 1-3 being generated from BHRF1 transcripts [41]. The expression of miR-BHRF1-1 is therefore regulated largely independently from the expression of BHRF1-2 and BHRF1-3.

3.3. Biological roles of EBV-encoded miRNAs

The evolutionary conservation of a number of EBV miRNAs with miRNAs encoded by the rhesus lymphocryptovirus implies that these miR-BARTs play critical roles in EBV's life cycle. In addition, there is accumulating evidence that some EBV encoded miRNAs may contribute to the oncogenic properties of the virus such as: (1) the induction of epithelial cell proliferation; (2) providing survival signals for virus infected cells in vivo (e.g. under hypoxic conditions inside the tumor); and (3) escaping host immune surveillance.

Choy and colleagues [51] have shown that miR-BART5 targets the transcript encoding the cellular factor PUMA (p53 up-regulated modulator of apoptosis) and they propose a role of miR-BART5 in the inhibition of apoptosis. MiR-BART5 may improve the overall survival of infected cells during the natural infection course of the virus and it may contribute to cell survival in virus associated cancers. Interestingly, miR-BART5's seed sequence contains significant homology with the cellular miRNAs, miR-18a and miR-18b, which display elevated expression in a number of cancers including liver cancer [52].

Like other members of the herpesvirus family (human CMV, HSV-1, and KSHV), EBV encoded miRNAs are also predicted to suppress virus immediate early (IE) lytic genes [53]. For instance, miR-BART15 is predicted (through computational methods) to down-regulate the EBV immediate-early transactivators, Zta and Rta [53]. Since these transactivators are the molecular switches controlling the transition into the lytic cycle, targeting Zta and Rta transcripts would represent a cost efficient means of suppressing the entire repertoire of lytic genes during latency.

MiR-BART2 has been shown to down-regulate the EBV encoded DNA polymerase gene, BALF5 [54], which is expressed during the lytic phase of the virus life cycle. Strikingly, miR-BART2 has perfect complementarity to the 3′-UTR of BALF5, and miR-BART2 thereby causes degradation of the BALF5 mRNA [54]. During the transition from viral latency to lytic reactivation, miR-BART2 is downregulated and enforced expression of miR-BART2 impairs EBV lytic replication [54]. These experiments support a model in which miR-BART2 functions to enforce viral latency thereby maintaining expression of the oncogenic viral latency genes [54]. At another level, the down-regulation of BALF5 by miR-BART2 may also help to reduce the amount of this antigenic viral protein to facilitate evasion of host immune surveillance. EBV miR-BART2 has also recently been shown to inhibit expression of the stress-induced immune molecule MICB to prevent Natural Killer (NK) cell mediated cell killing and facilitate immuno-evasion during EBV infection [55].

In addition to inhibiting entry into the lytic cycle, viral miRNAs also appear to be involved in regulating the expression of certain latency genes. For instance, EBV BARTs Cluster1 miRNAs (namely, miR-BART1, miR-BART16, and miR-BART-17) have been shown to target the 3′-UTR of the viral latent membrane protein LMP1 [56]. Although LMP1 is generally viewed as a key oncogene in the development of EBV-associated malignancies, when expressed inappropriately and/or at high levels, it can also cause growth arrest and induce apoptosis. The adjusting/fine-tuning of LMP1's expression levels by miR-BARTs may help to attenuate the inhibitory properties associated with high level or inappropriate expression of LMP1. On the other hand, the downregulation of LMP1 may also be part of an immuno-evasion strategy during certain parts of the viral life cycle. In the same vein, the mRNA of another viral latent membrane protein, LMP2A, was shown to be targeted by miR-BART22 [57]. LMP2A is a potent immunogen that is capable of inducing a cytotoxic T-cell response. The modulation of LMP2A levels may similarly help alleviate host immune surveillance.

MiR-BHRF1-3 has been shown to down-regulate expression of the interferon-inducible T-cell attracting chemokine, CXCL11/I-TAC, in B-cells [48]. Inhibition probably occurs by targeting the transcripts for degradation since the 3′ UTR of CXCL11 transcripts have perfect complementarity to miR-BHRF1-3. CXCL11 functions through binding to the chemokine receptor CXCR3, which is found predominantly on T-cells. CXCL11 is an important effector of Cytotoxic T-Lymphocyte (CTL) activation and is reported to be a potent anti-tumor factor in animal models [58; 59]. Through this mechanism, miR-BHRF1-3 may help suppress immunosurveillance to EBV antigens [48]. MiR-BHRF1-3 is also predicted to target the transcripts encoding the EBV immediate-early transactivators, Zta and Rta, and may therefore represent additional viral miRNAs that help enforce latency [53].

3.4. EBV-regulated cellular miRNAs

In addition to utilizing its own repertoire of miRNAs, EBV has tapped into existing cellular miRNA pathways to facilitate its complex life cycle. A number of studies have reported a correlation between EBV type III latency (which drives B-cell activation/proliferation) and expression of the oncogenic cellular miRNA, miR-155 [60; 61], and we have shown that type III latency genes drive miR-155 expression [62]. We have also shown that the cellular miRNA, miR-146a, is similarly induced by EBV type III latency gene expression and that this occurs at least in part through the viral latent protein, LMP1 [63; 64]. A number of other cellular miRNAs are also differentially expressed (although typically more moderately) in type I and type III latently infected cells including miR-15a, miR-21, miR-28, miR-34a, members of the miR-23a cluster [miR-23a, miR-24, and miR-27a], and miR-146b [64]. Similarly, Mrazek and colleagues [65] showed that EBV regulates miR-155 and miR-146a as well as other cellular miRNAs. Below, we will focus on a number of EBV-regulated cellular miRNAs, particularly those showing larger magnitudes of induction by EBV.

3.4.1. MiR-155

MiR-155 is encoded within the BIC (B cell integration cluster) gene which was originally identified as a frequent integration site in avian leukosis virus induced-B cell lymphomas [66; 67; 68]. Although BIC transcripts were shown many years ago to play a critical role in lymphomagenesis [66; 67; 68], the mechanism through which the BIC gene elicited a tumor phenotype remained a mystery since there were no discernable open reading frames encoded by these transcripts. In 2005, Eis et al. [69] demonstrated that BIC transcripts were processed into the functional miRNA, miR-155. Since this time, elevated miR-155 expression has been shown to correlate with an array of different tumor types and is now considered one of the most highly implicated miRNAs in cancer.

3.4.1.1. Regulation of miR-155 expression by EBV

Cellular miR-155 appears to be regulated by multiple signaling pathways. For instance, miR-155 can be transcriptionally induced by B-cell receptor (BCR) activation through an AP-1 element located near the transcription start site of its promoter region [70]. As mentioned above, EBV type III latency genes facilitate high level miR-155 expression [62], but the mechanism through which this occurs is still under investigation. Nevertheless, two studies implicated the viral latent protein, LMP1, as playing a role in the transcriptional up-regulation of miR-155 in B-cells through an NF-kB mediated mechanism [71; 72]. We similarly observed LMP1 mediated induction of miR-155 but at a comparatively low level. We proposed that other EBV latency gene(s) may provide unique and essential cooperative functions necessary to sustain the robust miR-155 expression observed in type III latency cells [62]. In line with this hypothesis, Lu et al. [72] demonstrated that the EBV latency protein, EBNA-2, can also induce BIC/miR-155 expression in B-cells.

3.4.1.2. MiR-155's function

MiR-155 is involved in immunological processes in normal tissues such as immunity, haematopoiesis, and inflammation. The dysregulation of miR-155 expression plays a role in the development of a number of diseases including cancers, viral infections, and cardiovascular diseases.

Accumulating studies have identified a wide array of miR-155 targets that may contribute to the modulation of cell proliferation, differentiation and apoptosis pathways. Our group reported that miR-155 can target the 3′UTRs of numerous transcription factor transcripts [62; 73; 74]. For example, we and others demonstrated that miR-155 directly targets the transcript encoding the transcriptional repressor, BACH1 [23; 24; 62], which binds to consensus AP-1 promoter elements [75]. Considering the known activation of AP-1 by EBV latency proteins as well as the role of AP-1 in the activation of lymphocytes during immune responses, it is reasonable to believe that the suppression of AP-1 inhibitors such as BACH1 may de-repress AP-1 promoter elements and increase their accessibility to activating AP-1 factors.

Through microarray studies and 3′ UTR reporter studies, we also identified DET1 as a miR-155 target [62]. Since DET1 mediates ubiquitin dependent degradation of the AP-1 family member, c-Jun [76], the inhibition of DET1 expression may further support increased AP-1 signaling through an increase in c-Jun levels.

In addition to BACH1, we also showed that miR-155 can target two other transcriptional repressor gene transcripts, HIVEP2 and ZNF652 [62]. Like BACH1, ZNF652 and HIVEP2 repress transcription through the direct binding to promoter/enhancer elements [77; 78]. HIVEP2 is downregulated in breast cancers [79] and, mechanistically, HIVEP2 inhibits the expression of MYC through binding to an NF-kB element in the MYC intron 1 regulatory element [78]. The suppression of HIVEP2 may support cell survival and proliferation by increasing the expression of NF-kB responsive promoters such as the promoter driving expression of the MYC gene. ZNF652 binds DNA and recruits the ETO family member CBFA2T3 [77]. CBFA2T3 is a transcriptional repressor that functions as a putative breast tumor suppressor. By inference, the inhibition of ZNF652 expression by miR-155 is expected to mitigate the impact of CBFA2T3's inhibitory influences on ZNF652 binding sites and contribute to tumor development.

As alluded to above, while EBV has evolved with mechanisms to facilitate high level expression of cellular miR-155, two other oncogenic herpesviruses, KSHV and MDV-1, have been found to encode their own miR-155 functional homologues. Although it has yet to be proven, the well established link between miR-155 and cancer leads to the obvious extension that the miR-155 functions of these viruses contribute to their respective oncogenic nature. In support of this contention, it is intriguing that the non-oncogenic strain of Merek's Disease virus, MDV-2, does not encode any miR-155 ortholog [80]. It seems that EBV, KSHV and MDV-1 all utilize miR-155 function to facilitate signaling cascades that presumably support their life style but which also may partly explain their oncogenic propensity. Notably, the utilization of miR-155 in the virus life cycle is not unique to herpesviruses, as the oncogenic retrovirus REV-T (reticuloendotheliosis virus strain T) has recently been shown to up-regulate miR-155 expression in both REV-T-infected chicken embryo fibroblasts and REV-T-induced B-cell lymphomas [81]. In this setting, miR-155 was found to play a critical pro-survival/anti-apoptotic role by targeting transcripts of the cell-cycle regulatory gene, JARID2 [81].

Outside of a viral context, many studies have reported that miR-155 induces tumorigenesis through an interplay with multiple cancer-related pathways. For instance, miR-155/BIC cooperates with MYC to induce Burkitts lymphoma [82]. A recent report demonstrated that miR-155 is involved in tumor metastasis through the enhancement of TGF-beta mediated Epithelial-Mesenchymal Transition (EMT) and cell migration [83]. MiR-155 may also be responsible for the dysregulation of certain ncRNA that are transcribed from ultraconserved regions (UCRs), a class of sequences that are commonly disrupted in cancers [22]. Lastly, miR-155 is implicated in facilitating tumorigenesis by modulation of the NF-kB pathway and evasion of host innate immunity [72].

3.4.2. MiR-146

In humans, there are two miR-146 genes that are located on chromosomes 5q33 and 10q24 and encode miR-146a and miR-146b, respectively. MiR-146a and miR-146b are predicted to have similar target repertoires since they share a high homology (with only a two-nucleotide difference in their 3′ non-seed sequence). In response to various immune-mediators such as LPS, IL-1b, and TNFa, miR-146a expression can be transcriptionally activated through an NF-kB dependent mechanism [84]. The upregulation of miR-146a can be achieved by the EBV encoded LMP1 and this induction similarly occurs through NF-kB signaling [63; 85].

3.4.2.1. Role of miR-146 in tumorigenesis

Previously, a number of studies have demonstrated the dysregulation of miR-146 expression in human cancers, such as thyroid carcinomas [86; 87], cervical cancers [88], breast cancers [89], pancreatic cancers [89], and prostate cancers [89; 90]. Nevertheless, the link between miR-146 and cancer is a little murky since there is both evidence that miR-146 promotes a cancer phenotype and there is evidence that miR-146 suppresses the tumor phenotype.

Like miR-155, there is evidence that miR-146 mediates cancer development by targeting key genes involved in proliferation, apoptosis, inflammation, and immunity. For example, over-expression of miR-146a in cervical cancer cell lines significantly enhances cell proliferation rate [88]. A recent study has demonstrated that miR-146a can prevent cytokine-induced apoptosis in human bronchial epithelial cells (HBECs), probably through a mechanism involving the upregulation of BCL2L1 (Bcl-XL) and phosphorylation of STAT3 [91]. In addition, miR-146 is involved in modulating immune responses. MiR-146 targets transcripts encoding factors in the NF-kB signaling pathway, such as IRAK1 (IL-1 receptor associated kinase 1) and TRAF6 (TNF receptor associated factor 6). By extension, miR-146 may function as a negative feedback regulator to fine-tune the immune response [84]. It is also predicted that the dysregulated expression of miR-146a could lead to inhibition of the interferon response pathway in tumors, therefore helping to suppress host immune surveillance through interferon signaling. In the context of EBV infection, induction of miR-146a could result in the suppression of the interferon-mediated antiviral response, protecting the virus from host immunity and facilitating the development of EBV-associated tumors [63].

Evidence that miR-146a also functions as a tumor suppressor is derived from studies showing a down-regulation of miR-146a levels in androgen-independent prostate cancers [90]. The loss of miR-146a function may contribute to this transformation, probably due to the loss of miR-146a mediated suppression of ROCK1, a key factor promoting the androgen-dependent to androgen–independent transformation of prostate cancer [90].

Constitutively active NF-kB signaling is often utilized by tumor cells to promote proliferation, survival, and metastasis. Studies in human breast cancer cell lines have shown that miR-146a/b acts as a negative regulator of NF-kB activity to inhibit tumor cell invasion and migration [92]. Breast cancer metastasis suppressor 1 (BRMS1) plays an important role in the suppression of metastasis in multiple human cancer cells. BRMS1 significantly induces miR-146a and miR-146b levels which, in turn, downregulate the expression of epidermal growth factor receptors and inhibit invasion and migration of tumor cells [93].

A recent study demonstrated that a single nucleotide polymorphism (SNP rs2910164) in the pre-miR-146a shows a close association with papillary thyroid cancer [86]. This common G/C polymorphism is located at position +60 of pre-miR-146a and somatic mutations to a C are commonly found in papillary thyroid tumors. Functionally, this mutation affects the processing of miR-146a and leads to an impact on the overall level of mature miR-146a (the G allele generally gives rise to a higher level of mature miRNAs than the C allele). The reduction in mature miR-146a in papillary thyroid cancers due to G → C mutation leads to a weaker inhibition of target genes involved in the NF-kB signaling pathway (e.g. TRAF6 and IRAK1), and CCDC6 (a gene frequently rearranged with RET proto-oncogene in papillary thyroid cancer). In these settings it appears that miR-146a suppresses tumorigenesis. This suppression is alleviated by the G → C transversion, mitigating the miR-146a mediated negative feedback regulation of the NF-kB signaling pathway [86; 94].

3.4.3. Other EBV-regulated cellular miRNAs

Besides miR-155 and miR-146, miR-127 is induced by EBV infection [95; 96] and miR-21, -23a, -24, -27a, and -34a are expressed at higher levels, whereas miR-28 is expressed at lower levels in EBV-infected type III latency cells compared to EBV-infected type I latency cells [64]. Some of these miRNAs may similarly contribute to the tumorigenic phenotype of EBV-infected cells.

Leucci and colleagues found that miR-127 expression remains high in EBV-positive Burkitt's lymphoma cells, but not in EBV-negative BL cells [96]. The induction of miR-127 may inhibit B-cell differentiation and promote lymphomagenesis by targeting BLIMP1 and XBP1 transcripts, the protein products of which are considered to be the master regulators of plasma cell differentiation [96]. The induction of miR-21 by EBV latency III infection has significant implications for both the EBV life-cycle and EBV-mediated tumorigenesis. As an oncomir, elevated expression of miR-21 has been demonstrated in a number of malignancies including metastatic breast cancers, colorectal cancers, hepatocellular cancers, cervical cancers, pancreatic cancers, and glioblastoma [89; 97; 98; 99; 100; 101]. In vitro, over-expression of miR-21 promotes cell proliferation, migration and invasion, and inhibition of miR-21 suppresses cell growth, migration, and invasion, and increases apoptosis [97; 98; 99; 102]. Several studies have demonstrated that miR-21 down-regulates expression of the tumor suppressor genes, programmed cell death 4 (PDCD4) and tropomyosin 1 (TPM1) [97; 99; 103]. It is possible that in the setting of EBV-infected B lymphocytes, miR-21 may similarly inhibit PDCD4 and/or TPM1 expression.

Recently, Shinozaki and colleagues have reported that EBV infection results in the down-regulation of miR-200 family members [104]. We and others have shown that the miR-200 family of miRNAs can trigger the EBV lytic switch, causing the transition from latency to the lytic cycle [105; 106]. The suppression of miR-200 family members by EBV may facilitate the maintenance of EBV latency thereby supporting establishment and/or maintenance of the tumor phenotype.

4. KSHV

Kaposi's sarcoma-associated herpesvirus (KSHV) is the causative agent of Kaposi's sarcoma (KS) and is associated with the pathogenesis of primary effusion lymphoma (PEL) and multicentric Castleman's disease [107; 108; 109]. Like EBV, KSHV encodes its own miRNAs. To date, at least 24 mature miRNAs have been identified in KSHV [110; 111; 112; 113]. All of these miRNAs are located in a single cluster within the viral latency-associated region of the KSHV genome. KSHV miRNAs can be readily detected in latently infected B-cell lines (e.g. primary effusion lymphoma cell lines BC-1, BC-3, and BCBL1) [110; 111; 112; 113] and have been detected in biopsy samples from patients with KS and multicentric Castleman's disease [114].

Accumulating studies have shown that KSHV miRNAs regulate a number of cellular genes whose inhibition may contribute to tumorigenesis. Through microarray studies, Samols et al. [115] found that ectopic expression of a portion of the KSHV miRNA cluster results in the down-regulation of 65 cellular genes. Among them, thrombospondin 1 (THBS1) shows a significant decrease at both the mRNA and protein levels. A 3′-UTR reporter assay demonstrated that THBS1 is directly targeted by multiple KSHV miRNAs including miR-K12-1, miR-K12-3-3p, miR-K12-6-3p, and miR-K12-11. This observation fits nicely into previous studies showing that THBS1 is a strong tumor suppressor, an anti-angiogenic factor, and an immune-modulator that is commonly decreased in KS lesions [116].

Recently, Qin and colleagues [117] have shown that miR-K12-3 and miR-K12-7 selectively activate the transcription and secretion of IL6 and IL10 by macrophages and myelomonocytic cells. They further showed that the induction of IL6 and IL10 expression was achieved, at least partially, through down-regulating a dominant-negative isoform of the transcription factor, C/EBPB (referred to as LIP). IL6 and IL10 are known to play an active role in the pathogenesis of KSHV-associated malignancies by promoting tumor cell growth, angiogenesis, and by suppressing T-cell activation.

As discussed above, KSHV miR-K12-11 shares the same seed sequence as the cellular oncomir, miR-155 [23; 24]. Further, the enforced expression of miR-K12-11 or miR-155 causes the down-regulation of a common group of targets indicating that miR-K12-11 is, in fact, a functional ortholog of miR-155 [23; 24]. It is likely that miR-K12-11 provides many of the same needs for the KSHV life cycle as miR-155 does for EBV. Nevertheless, there may be unique aspects of the respective viral life cycles that partly explain the different mechanisms used by these two viruses to achieve a miR-155/miR-K12-11-like function. Although miR-155 has been found to be dysregulated in a variety of tumor types, its normal expression appears to be preferentially targeted to immune cells. KSHV may benefit from having its own version of miR-155 to more easily enforce a miR-155-like function in settings, such as in endothelial cells, in which the expression of cellular miR-155 may be more difficult to accomplish through normal signal transduction pathways.

Other KSHV encoded miRNAs have also been shown to target cellular pathways that may promote oncogenesis. The cyclin-dependent kinase inhibitor CDKN1A (p21) was recently shown to be a direct target of miR-K12-1 [118] indicating that the KSHV miRNAs can inhibit core components of the cell cycle regulatory machinery. Stable inhibition of miR-K12-1 in virally infected B-cells was found to cause de-repression of CDKN1A and increased cell cycle arrest after p53 (encoded by TP53) activation [118].

KSHV miRNAs appear to also play a direct role in preventing apoptosis through the targeting of apoptosis regulators. Ziegelbauer et al. showed that multiple KSHV encoded miRNAs (miR-K12-5, -9, -10a, and -10b) inhibit the expression of BCLAF1, a gene encoding a Bcl2-associated factor that promotes apoptosis [119]. The targeting of BCLAF1 by multiple miRNAs may be a strategy to enforce a more robust inhibition to better ensure host survival and/or reduce host susceptibility to apoptosis [119]. KSHV miR-K12-7 has been shown to inhibit expression of the stress-induced immune regulatory gene, MICB, thereby preventing NK cell mediated cell killing during KSHV infection [55].

Like EBV, the ability to establish a primarily latent infection is critical for KSHV tumor pathologies. Recently, Bellare and colleagues have reported that KSHV miR-K12-9* directly targets the lytic immediate early transcription factor, RTA (ORF50), and prevents viral reactivation [120]. In another study, Lu et al. [121] found that a mutant KSHV with deleted viral miRNAs displayed higher lytic activity and that RTA is targeted by miR-K12-5 (either directly or indirectly). Another viral miRNA, miR-K12-4-5p, was shown to enhance DNA methyltransferase activity and overall CpG methylation by inhibiting expression of the cellular repressor protein, RBl2 [121]. The maintenance of DNA methylation within certain regions of the viral genome, such as the RTA promoter, is believed to help stabilize viral latency [121]. Through yet another mechanism, KSHV miR-K12-1 was found to promote viral latency by activating the NF-kB pathway, in part, through targeting the mRNA of the NF-kB inhibitor NFKBIA (IkBa) [122]. Together, these studies illustrate that like EBV, KSHV-encoded miRNAs support the maintenance of latency through a variety of mechanisms.

KSHV has now been shown to regulate a number of cellular miRNAs that may specifically play roles in its life cycle. Punj et al. showed that the KSHV viral protein K13 induces miR-146a which suppresses expression of the chemokine receptor gene, CXCR4 [123]. CXCR4 is normally expressed at high levels on immature endothelial cells in bone marrow where it plays a key role in holding these progenitor cells for their further maturation. Thus, the downregulation of CXCR4 by miR-146a may facilitate the premature release of these KSHV-infected endothelial progenitors into the circulation, possibly contributing to the subsequent development of Kaposi's sarcoma from these immature cells.

The KSHV latent protein K15M has been reported to up-regulate miR-21 and miR-31 expression. In turn, miR-21 and miR-31 are believed to regulate key factors involved in cell migration and contribute to K15M induced cell motility and invasion [124].

5. HPV

High-risk human papillomaviruses (such as HPV-16 and HPV-18) are the causative agents of human cervical, penile, and anal cancers. Although no HPV-encoded miRNA has been identified to date, HPV has been found to regulate cellular miRNAs that appear to modulate cellular signal transduction pathways in a way that facilitates a more optimal environment for viral replication. First, the HPV E7 oncoprotein suppresses the cellular miRNA, miR-203, which is normally expressed in differentiating epithelial cells [125]. MiR-203 targets the TP53 family member, TP63 [126; 127], which unlike TP53, appears to promote cellular proliferation. The HPV mediated down-regulation of miR-203 and the resulting up-regulation of TP63 protein is believed to facilitate more robust viral genome replication by forcing differentiating cells into S phase where nucleotide pools and DNA replication associated factors are elevated [125].

Wang and colleagues [128] have reported that the tumor suppressor miRNA, miR-34a, is down-regulated in both clinical cervical cancer samples and cervical cancer derived cell lines. The down-regulation of miR-34a here is likely the result of HPV infection since the high-risk HPV E6 oncoprotein was shown to inhibit miR-34a expression [128]. The inhibition of miR-34a by E6 occurs indirectly through destabilizing the cellular TP53 protein which otherwise binds and activates the miR-34a promoter [128]. Evidence that inhibition of miR-34a is important for HPV-associated cancers has been inferred from experiments showing that overexpression of miR-34a in HPV-infected cells induces cell growth arrest and apoptosis [128].

Lastly, Martinez et al. [129] have shown that E6 (from HPV-16) can also inhibit the cellular miRNA, miR-218. In these studies, miR-218 was found to inhibit expression of the cellular gene, LAMB3, which was previously shown to enhance cell migration, tumorigenesis, and viral infection. An HPV mediated downregulation of miR-218 and the subsequent up-regulation of LAMB3 is therefore expected to promote viral infection as well as HPV's tumor promoting functions [129].

6. HCV

Hepatitis C virus (HCV) is a hepatotropic positive-stranded RNA virus that belongs to the Flaviviridae family. Patients with HCV infection usually end up with chronic active viral infection which, in turn, leads to cirrhosis and hepatocellular carcinoma. Cellular miRNAs have been shown to mediate critical interactions between HCV and host cells. Among these miRNAs, miR-122 is probably one of the most highly studied. MiR-122 is expressed at high levels in liver tissues and is considered to be principally a liver-specific miRNA. Interestingly, miR-122 is thought to play a critical role in HCV's life-cycle as a positive regulator of HCV replication. Although the underlying mechanism is unclear, miR-122 has been shown to bind to the 5′-UTR of the HCV RNA genome and increase HCV RNA abundance as well as the production of infectious viral particles [130; 131; 132; 133]. Clinically, the strong influence of miR-122 on HCV replication makes miR-122 an attractive potential pharmaceutical target for HCV treatment and it has recently been reported that therapeutic silencing of miR-122 (using an anti-miR-122 locked nucleic acid inhibitor) indeed showed promising results in the treatment of HCV infection in animals [134].

7. HBV

Hepatitis B virus (HBV) is a DNA virus that is closely associated with the development of hepatocellular carcinoma. Although virally encoded miRNAs have not been identified for HBV, a number of cellular miRNAs have been reported to be involved in HBV-associated malignancies. The expression of miR-602 was found to be elevated in HBV-infected liver and in hepatocellular carcinoma, and miR-602 may contribute to the HBV-mediated tumorigenesis through targeting of the tumor suppressor gene, RASSF1A [135]. Recently, Zhang et al. showed that miR-143 was dramatically increased in HBV related hepatocellular carcinoma samples. Although miR-143 is generally considered to be a tumor suppressor, these investigators suggested that the up-regulation of miR-143 may contribute to tumor metastasis by inhibiting the expression of the fibronectin type 3 domain containing 3B (FNDC3B) gene, whose protein product is a regulator of cell motility [136].

8. Concluding remarks

Despite their extremely small size, miRNAs have profound influences on cell signaling pathways. It is no wonder that viruses, the kings of efficiency in the biological world, would evolve to include miRNAs in their arsenal of regulators to achieve the various and sundry tasks necessary to facilitate their respective life cycles. As a field, we now have a pretty good cataloguing of miRNA repertoires encoded by most of the more commonly studied viruses, although there will certainly be further additions in the years to come. We also have at least some appreciation for some of the functions of these virus-encoded miRNAs. Nevertheless, given that only a handful of viral miRNAs have been studied functionally, and considering that miRNAs often regulate hundreds of genes, we haven't yet scratched the surface in our understanding of the biological functions of these miRNAs. Progress in this area will be ramped up substantially in the near future, however, with the advent of new technologies to globally assess miRNA targeting, such as cross linking-argonaute/RNA immuno-precipitation methods [137], and miRNA induced transcriptome [74] and proteome changes [138]. These data sets will lay the foundation to get at the biological roles of these miRNAs so that we can better understand the functions they play in virus and tumor biology. Along the way, we can anticipate unique clinical opportunities for the diagnosis, prognosis and treatment of virus associated cancers.

Table 2. Summary of herpesvirus microRNAs and their biological effects.

Viruses	microRNAs	Biological Effects
EBV	miR-BART2	Down-regulation of EBV BALF5. Inhibits expression of the stress-induced immune molecule, MICB.
	miR-BART5	Down-regulation of the cellular factor PUMA (p53 up-regulated modulator of apoptosis).
	BARTs Cluster1 microRNAs (namely, miR-BART1, miR-BART16, and miR-BART-17)	Down-regulation of EBV LMP1.
	miR-BART15	Predicted to down-regulate expression of the EBV immediate-early transactivators, Zta and Rta.
	miR-BART22	Down-regulation of EBV LMP2A.
	miR-BHRF1-2	Down-regulates BHRF1 (EBV-encoded homolog to cellular anti-apoptotic factor Bcl-2).
	miR-BHRF1-3	Down-regulates interferon-inducible T-cell attracting chemokine CXCL11/I-TAC.
		Predicted to target EBV immediate-early transactivators, Zta and Rta.
KSHV	KSHV-microRNA cluster (namely, miR-K12-1, miR-K12-3-3p, miR-K12-6-3p, and miR-K12-11)	Down-regulates expression of the cellular factor, thrombospondin 1 (THBS1).
	miR-K12-1	Down-regulates expression of the cellular cyclin-dependent kinase inhibitor CDKN1A (p21) and the NF-kB inhibitor NFKBIA (IkBa).
	miR-K12-3	Down-regulates expression of the cellular factor, LIP; activates transcription and secretion of IL-6 and IL-10.
	miR-K12-4-5p	Down-regulates expression of the cellular repressor protein, RBl2.
	miR-K12-5	Directly or indirectly down-regulates expression of the KSHV Immediate-early transactivator RTA (ORF50).
	miR-K12-7	Down-regulates expression of the cellular factor, LIP; activates transcription and secretion of IL-6 and IL-10.
		Down-regulates expression of the stress-induced immune molecule, MICB.
	miR-K12-9*	Down-regulates expression of the KSHV immediate-early transactivator RTA.
	miR-K12-5, -9, -10a, and -10b	Down-regulates expression of the cellular factor, BCLAF1.
	miR-K12-11	Down-regulates expression of the cellular factor, BACH1 and other miR-155 targets.
rLCV	miR-rL1-1, miR-rL1-2, miR-rL1-5-5p, miR-rL1-5-3p, miR-rL1-6-5p, miR-rL1-6-3p, miR-rL1-8, miR-rL1-9, miR-rL1-13, miR-rL1-14-5p, miR-rL1-14-3p, miR-rL1-15, miR-rL1-16-5p, miR-rL1-16-3p, miR-rL1-19, miR-rL1-20-5p, miR-rL1-20-3p, miR-rL1-23-5p, miR-rL1-23-3p, miR-rL1-24-5p, miR-rL1-24-3p, miR-rL1-25, miR-rL1-27, miR-rL1-29, miR-rL1-32-5p, miR-rL1-32-3p, miR-rL1-33	EBV homologues
MDV	mdv1-miR-M4	Share seed sequence with miR-155.