Kaposi’s sarcoma-associated herpesvirus expresses an
array of viral microRNAs in latently infected cells
Xuezhong Cai*†, Shihua Lu*†, Zhihong Zhang*, Carlos M. Gonzalez‡, Blossom Damania‡, and Bryan R. Cullen*§
*Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27710; and‡Department of Microbiology and
Immunology, Lineberger Comprehensive Cancer Center, University of North Carolina School of Medicine, Chapel Hill, NC 27599
Edited by Bernard Roizman, University of Chicago, Chicago, IL, and approved February 4, 2005 (received for review November 3, 2004)
MicroRNAs (miRNAs) are an endogenously encoded class of small
RNAs that have been proposed to function as key posttranscrip-
tional regulators of gene expression in a range of eukaryotic
species, including humans. The small size of miRNA precursors
cell defense pathways. Here, we demonstrate that the pathogenic
humanherpesvirus Kaposi’s sarcoma-associated
(KSHV) encodes an array of 11 distinct miRNAs, all of which are
expressed at readily detectable levels in latently KSHV infected
per cell. The KSHV miRNAs are expressed from what appears to be
a single genetic locus that largely coincides with an ?4-kb non-
coding sequence located between the KSHV v-cyclin and K12?
Kaposin genes, both of which are also expressed in latently
infected cells. Computer analysis of potential mRNA targets for
these viral miRNAs identified a number of interesting candidate
genes, including several mRNAs previously shown to be down-
regulated in KSHV-infected cells. We hypothesize that these viral
miRNAs play a critical role in the establishment and?or mainte-
nance of KSHV latent infection in vivo and, hence, in KSHV-induced
RNA interference ? viral pathogenesis ? viral latency
ferent miRNAs have so far been identified in humans (reviewed
in ref. 1). miRNAs are initially transcribed as part of one arm of
an ?80-nt-long RNA stem–loop that in turn forms part of a
longer transcript called a primary miRNA (pri-miRNA) (2). The
first step in miRNA processing is mediated by the nuclear RNase
III enzyme Drosha, which cleaves pri-miRNA stem–loops to
release the ?60-nt long pre-miRNA hairpin intermediate (3).
After export to the cytoplasm, pre-miRNAs are further pro-
cessed by the RNase III enzyme Dicer to give an ?22-bp miRNA
duplex (4, 5). One strand of this duplex is then incorporated into
the RNA-induced silencing complex (RISC), where it acts to
guide RISC to mRNAs bearing complementary sequences (6–
8). RISC binding to mRNA can inhibit gene expression by one
of two mechanisms. If the mRNA bears an essentially perfectly
complementary sequence, binding induces cleavage by the RISC
component Ago2, leading to mRNA degradation (9–11). How-
ever, if the mRNA contains an imperfectly complementary
target, RISC binding may instead induce translational inhibition
(11–13). Unlike mRNA cleavage, which can be mediated by a
single RISC, translation inhibition is highly cooperative and may
require binding by several RISCs, potentially bearing different
miRNAs, to be effective (14).
Successful viral replication depends on the ability of the
while simultaneously inactivating innate host cell defense mech-
anisms, such as the IFN pathway or apoptosis induction. Al-
though many viruses encode proteins that specifically inactivate
utility of this approach. The small size of miRNA stem–loop
precursors could offer an attractive alternative way for these
RNAs found in all metazoan eukaryotes, and ?200 dif-
intracellular pathogens to turn off specific host genes. However,
because miRNAs act exclusively on mRNAs, they cannot exert
has decayed. Because many lytic viral life cycles are very short,
miRNAs may therefore not represent a useful tool for lytic
viruses to modify the host cell environment. In contrast, in the
case of viruses that routinely establish latent infections as part of
their life cycle, miRNAs may represent an effective way to
reshape host cell gene expression to their benefit.
Among virus families, one stands out as using latent infection
as an almost invariant component of their life cycle, i.e., the
herpesviruses. Moreover, an analysis of one human family
member, Epstein–Barr virus (EBV), has recently identified five
viral miRNAs and shown that these are expressed in latently
EBV infected cells (15). Here, we extend these earlier data by
herpesvirus, termed Kaposi’s sarcoma-associated herpesvirus
(KSHV) or human herpesvirus 8, encodes at least 11 previously
undiscovered miRNAs that are all expressed in latently KSHV
infected cells. The expression of an array of these viral miRNAs
mRNAs by miRNA-mediated RNA interference may represent
a critical step in the establishment and?or maintenance of latent
infections by KSHV.
Cell Culture and RNA Preparation. BC-1, BCBL-1 and BJAB cells
FCS. Where necessary, TPA (Sigma, final concentration 25
ng?ml) was added for 48 h before RNA preparation. Total
cellular RNA was prepared by using Trizol reagent (Invitrogen).
Cloning of Small RNAs.The cloning of small RNAs from BC-1 cells
was conducted as described by Lau et al. (16) using 750 ?g of
BC-1 total RNA. Classification of small RNA cDNA sequences
was based on cDNA sequence analysis in the GenBank database
(www.ncbi.nlm.nih.gov?GenBank?index), the miRNA registry
database (www.sanger.ac.uk?Software?Rfam?mirna), the
rRNA database (www.psb.ugent.be?rRNA?blastrrna.html), and
human tRNA database (http:??rna.wustl.edu?GtRDB?Hs?Hs-
RT-PCR and Northern Blot Analyses. RT-PCR and Northern blot
analyses were performed as described (17) except that the
annealing temperature during the PCR amplification was set at
50°C and 30 PCR cycles were performed for all KSHV se-
were run for the cellular GAPDH mRNA. Thirty micrograms of
This paper was submitted directly (Track II) to the PNAS office.
complex; EBV, Epstein–Barr virus; KSHV, Kaposi’s sarcoma-associated herpesvirus; PEL,
primary effusion lymphoma.
†X.C. and S.L. contributed equally to this work.
§To whom correspondence should be addressed. E-mail: email@example.com.
© 2005 by The National Academy of Sciences of the USA
April 12, 2005 ?
vol. 102 ?
no. 15 www.pnas.org?cgi?doi?10.1073?pnas.0408192102
total RNA per sample was used in each Northern blot. The
DNA oligonucleotide. The probe used for the U6 Northern
analysis was 5?-CGTTCCAATTTTAGTATATGTGCTGC-
CGAAGCGA-3?. The primers used for RT-PCR are listed in
Supporting Text, which is published as supporting information on
the PNAS web site.
Although KSHV is primarily known as the etiologic agent of
Kaposi’s sarcoma, KSHV is also associated with two B-cell
derived cancers termed primary effusion lymphoma (PEL) and
multicentric Castleman’s disease (18–21). Several B cell lines
derived from PEL patients have been described; we selected one
of these, BC-1, for analysis (22). BC-1 was chosen because latent
KSHV infection of these cells is stable under normal growth
conditions, with only very few cells (?2%) undergoing sponta-
neous activation of lytic replication (23), and because the KSHV
genome resident in BC-1 cells has been fully sequenced. Of note,
BC-1, like many other PEL B cell lines, is also latently coinfected
with EBV (23).
We size selected and cDNA cloned 18- to 24-nt-long RNAs
from BC-1 cells (16) and sequenced 959 different cDNA clones.
As shown in Table 1, ?18% of the short RNAs obtained were
derived from KSHV, and a further ?23% were of EBV origin.
Cellular miRNAs accounted for ?18% of the short RNA clones
obtained, and this grouping included 34 different miRNAs,
several of which have been reported only in rodents (Table 3,
which is published as supporting information on the PNAS web
site). The remaining cDNAs, representing ?41% of the total,
represent breakdown products of cellular noncoding RNAs,
mRNAs, or cellular transcripts of unknown origin.
Analysis of the KSHV sequences identified 14 distinct candi-
date viral miRNAs, with the most prevalent, named miR-K7,
being cloned 51 times (Table 2). Criteria for validation of the
authenticity of candidate miRNAs have been established (24). In
addition to identification in a library of size fractionated RNA
(Table 2), these include detection of the ?22-nt RNA by
Northern analysis. Moreover, as noted above, a defining char-
acteristic of authentic miRNA is that they are derived by
processing of an ?80-nt precursor RNA hairpin.
Analysis of the KSHV sequence context of the 14 candidate
miRNAs shown in Table 2 revealed that all sequences derived
from computer-predicted RNA hairpins of the expected ?80-nt
size (Fig. 1). As predicted based on previous analyses of human
pri-miRNAs (16, 25, 26), the mature KSHV miRNAs all derived
from one arm of the RNA hairpin and the precursor stem
extended significantly beyond the mature miRNA sequence.
This extension is required for efficient processing of the pri-
miRNA precursor by Drosha (3, 25).
As noted above, miRNA processing gives rise to an ?20-bp
miRNA duplex bearing 2-nt 3? overhangs. One strand of this
duplex is then incorporated into RISC while the second is
degraded (1). It has been proposed that selection of the incor-
porated strand depends on the relative stability of the base
pairing around the 5? nucleotide of each strand, with the strand
containing the less tightly base paired 5? end being selected for
incorporation (27, 28). However, if the difference in base pairing
is small, then each strand may be incorporated into RISC.
Moreover, the nonincorporated strand, termed the star strand,
may persist long enough to be captured during cloning, albeit at
low frequency. Inspection of the predicted KSHV pri-miRNA
hairpins (Fig. 1) shows that four of these, i.e., miR-K4, miR-K6,
miR-K8, and miR-K9, give rise to two mature miRNAs. Fol-
lowing convention, the miRNA derived from the 5? arm of the
stem–loop precursor is designated by a ‘‘5p’’ suffix, whereas the
miRNA derived from the 3? arm is designated by ‘‘3p’’ (Table 2).
Inspection of Table 2 shows that three of the four KSHV
miRNAs, which derive from the RNA strand complementary to
a more common KSHV miRNA, were only recovered once
(miR-K6-5p and miR-K8-5p) or twice (miR-K9-3p). These three
duplex intermediate. However, miR-K4-3p was recovered six
times, which suggests that it may be an authentic miRNA (see
Table 1. Composition of the small RNA cDNA library obtained
from BC-1 cells
Type Number of clonesPercentage
Noncoding RNA fragments
Total number sequenced
*It is possible that there are cellular miRNAs in these classes that were not
identified in this study.
Table 2. KSHV miRNA sequences and their genomic location
miRNA Sequences 5? to 3?
HitsLength, ntKSHV sequence coordinates
Sequence variation in recovered KSHV miRNAs is indicated by parentheses surrounding the variable nucleo-
Cai et al.
April 12, 2005 ?
vol. 102 ?
no. 15 ?
We next performed Northern analyses to confirm that the
KSHV miRNAs listed in Table 2 are actually expressed in
KSHV-infected cells. For this purpose, we derived RNA from
BC-1 cells as well as from the cell lines BCBL-1 and BJAB.
BCBL-1 cells are a second, independent PEL-derived cell line
infected with KSHV, but these cells differ from BC-1 in that they
are not coinfected with EBV. Moreover, up to 5% of BCBL-1
cells spontaneously enter lytic KSHV replication under normal
culture conditions (29, 30). BJAB is a B cell line that is not
infected by KSHV or EBV and that here serves as a negative
control. In addition to BC-1 and BCBL-1 cells cultured under
normal conditions, we also analyzed BC-1 and BCBL-1 cells that
had been treated with TPA for 48 h. TPA treatment induces lytic
KSHV replication in a significant percentage of both BC-1 and
BCBL-1 cells (23, 30), and this treatment was therefore designed
to see whether lytic replication activated expression of any
To confirm that TPA treatment was indeed inducing lytic
KSHV replication, we used immunofluorescence to quantitate
the number of BC-1 and BCBL-1 cells expressing the ORF59
protein, a KSHV early lytic protein, before and after TPA
treatment (Fig. 4, which is published as supporting information
on the PNAS web site). These data revealed that ?0.9% of BC-1
cells expressed ORF59 before TPA treatment, whereas ?20%
expressed ORF59 after treatment. Similarly, ?4.3% of BCBL-1
cells expressed ORF59 before TPA treatment and ?29% after
treatment. Analogous results were also obtained by quantitative
competitive RT-PCR using primers specific for the ORF50 gene,
encoding the major activator of KSHV lytic replication (Fig. 4B).
Therefore, although TPA treatment indeed induces a substantial
increase in the level of lytic KSHV replication in both BCBL-1
and BC-1 cells, the majority of cells remain latently infected even
after TPA treatment.
To confirm that the BC-1 cells used in this analysis were
indeed expressing latent KSHV gene products under normal
culture conditions, we performed an RT-PCR analysis of RNA
preparations derived from BC-1, as well as from the BCBL-1 and
BJAB cells, using primers specific for a latently expressed KSHV
mRNA (ORF72, encoding v-cyclin), a gene expressed in latently
infected cells that is activated during lytic infection (K12), and
a gene characteristic of lytic infection (ORF25, which encodes
the major KSHV capsid protein) (31–33). As shown in Fig. 2A,
noninduced BC-1 cells expressed ORF72 mRNA and a low level
of K12 transcripts, but ORF25 mRNA was not detected, con-
sistent with latent KSHV infection. Treatment of BC-1 cells with
TPA increased expression of K12 and induced ORF25 mRNA
expression, consistent with induction of lytic KSHV replication.
BCBL-1 cells, in contrast, expressed not only ORF72 and K12,
but also a low level of ORF25 mRNA, consistent with the
tendency of BCBL-1 to enter lytic replication spontaneously
(29). However, TPA induction clearly enhanced the expression
of K12 and ORF25, as expected. The control cell line BJAB was
negative for all KSHV transcripts, but expressed a comparable
level of the cellular GAPDH mRNA (Fig. 2A).
As noted above, miRNA processing proceeds through an
?60-nt pre-miRNA intermediate before giving rise to the ma-
ture ?22-nt miRNA (1). As shown in the Northern analysis
presented in Fig. 2B, both the ?62-nt miR-K5 pre-miRNA and
the 22-nt mature miR-K5 were readily detectable in both latently
infected and TPA-treated BC-1 cells, with little or no activation
by TPA. In BCBL-1 cells, the ?62-nt pre-miR-K5 RNA was
readily detectable, but mature miR-K5 was expressed at only low
cells (Fig. 2B).
In Fig. 2C, we present a more complete Northern analysis of
the expression of the mature KSHV miRNAs listed in Table 2.
As may be readily observed, ?22-nt RNAs that were specifically
recognized by the KSHV miRNA-specific probes were detected
for all of the miRNAs analyzed in the KSHV-infected cell lines
BC-1 and BCBL-1, but not in the control BJAB cells. Of note,
both miR-K4-5p and miR-K4-3p, derived from opposite strands
of a single pri-miRNA hairpin (Fig. 1) were detected. To save
space, only the mature KSHV miRNA signals are shown,
although the predicted ?70-nt pre-miRNA precursors were also
detected in most cases. Doublet bands were detected for miR-K7
and miR-K8-3p, consistent with the recovery of more than one
size of mature miR-K7 and miR-K8-3p documented in Table 2.
For the majority of the KSHV miRNAs, there was no evidence
of enhanced expression after TPA treatment, but some TPA
induction (?2 fold) was observed for miR-K1, miR-K2, and
miR-K3. As a loading control, we also performed a Northern
analysis for the constitutively expressed snRNA U6 and for the
Predicted stem–loop structures of KSHV pri-miRNAs. These structural predictions were derived by using MFOLD (46). Mature miRNA sequences are
www.pnas.org?cgi?doi?10.1073?pnas.0408192102Cai et al.
for both of these small cellular RNAs was consistent across all
five RNA samples tested. The fact that TPA treatment, which
increases the number of BC-1 cells entering lytic KSHV repli-
cation by ?20-fold (Fig. 4), has little or no effect on KSHV
miRNA expression (Fig. 2) argues strongly against the hypoth-
esis that the KSHV miRNAs observed in uninduced BC-1 and
BCBL-1 cells originate exclusively from the small percentage of
cells undergoing spontaneous lytic KSHV replication.
An interesting question is the average level of expression of
these viral miRNAs on a per cell basis. To determine this level,
we purchased synthetic forms of the KSHV miR-K7 and
miR-K6-3p miRNAs and then performed a Northern analysis
using BC-1 RNA and BJAB RNA to which a standard curve
of the synthetic KSHV miRNAs had been added. As shown in
Fig. 5, which is published as supporting information on the
PNAS web site, this analysis revealed that latently KSHV
infected BC-1 cells express an average of ?2,200 copies of
miR-K6-3p and ?800 copies of miR-K7 per cell. These results
are comparable to expression data for four cellular miRNAs
obtained in HeLa cells, which detected between ?700 and
?10,000 copies per cell (34).
The KSHV DNA genome exists in latently infected cells in the
form of an unintegrated episome, and previous work using BC-1
and BCBL-1 cells has reported that these cell lines contain an
average of ?59 and ?70 copies of the KSHV genome, respec-
tively (35). These numbers likely represent an upper limit for
latently infected cells, because the KSHV genome copy number
increases substantially during lytic replication. To confirm that
these earlier numbers are representative of the KSHV genome
copy number present in the cells analyzed here, we performed
a quantitative competitive PCR analysis using DNA derived
from uninduced BC-1 and BCBL-1 cells and primers specific for
the viral ORF25 gene. These data (Fig. 6, which is published as
supporting information on the PNAS web site) revealed that
each BC-1 cell on average contained ?83 copies, and each
BCBL-1 cell contained ?74 copies, of the KSHV genome.
Therefore, in the case of BC-1, we can calculate that each
episomal KSHV genome on average gives rise to ?27 copies of
miR-K6-3p and ?10 copies of miR-K7. Because miRNAs are
thought to be fairly stable (1), this finding suggests that the
promoter(s) responsible for KSHV miRNA transcription are
likely to be, at most, modestly active.
KSHV was initially discovered in AIDS-associated Kaposi’s
sarcoma (19) and is now known to play a key role in all forms of
this disease (36). In addition, KSHV is specifically associated
with two forms of B cell lymphoma (20, 21). Although KSHV
infection is fairly innocuous in immunocompetent individuals,
KSHV can be a serious opportunistic pathogen in immunocom-
promised patients and is the most common cause of cancer in
human immunodeficiency disease type 1 (HIV-1)-infected pa-
tients. As a result, Kaposi’s sarcoma has become the most
commonly reported cancer in parts of Africa (37) and is likely to
emerge as a serious problem in other areas where HIV-1
infection is increasing.
Although lytic infection by KSHV has been proposed to
contribute to KSHV-associated pathogenesis (38), latently in-
fected cells are likely to play a critical role in all KSHV-induced
tumors. Therefore, understanding how KSHV is able to establish
of cell proliferation is important to understanding KSHV patho-
genesis and potentially for the future development of drugs that
disrupt KSHV replication in vivo.
of the 88 known KSHV ORFs. Specifically, all latently KSHV-
infected cells, including the PEL-derived cell line BC-1, express
normal conditions or in the presence of TPA. ORF72 (v-cyclin) is a latent KSHV gene, and K12 is a latent gene that is activated during lytic replication, whereas
complementary to mature miR-K5 (Table 2). The RNA samples used were the same as analyzed in A. Synthetic RNA markers (M) of 63 and 22 nt in length were
Cai et al.
April 12, 2005 ?
vol. 102 ?
no. 15 ?
the KSHV genes v-FLIP (ORF71), v-cyclin (ORF72), and
LANA-1 (ORF73) (31–33). Although these three genes are not
induced by activation of lytic infection, a small number of
additional KSHV genes, e.g., K12?Kaposin and vIRF2, express
low levels of mRNA in latently infected cells but are activated
during lytic infection (32, 33). Interestingly, the three genes most
are located adjacent to one another in the KSHV genome (39).
These three ORFs are separated from the K12 gene, which is
expressed at low levels during latency, by an ?4 kb KSHV
sequence that lacks any significant ORFs, representing the
largest coding gap within the unique region of the KSHV
genome (39) (Fig. 3). Remarkably, 9 of the 10 KSHV miRNAs
identified in this analysis are located within this coding gap,
whereas the 10th, miR-K10, is found within the adjacent ORF
oriented in the same genomic direction (Fig. 3).
There are two obvious questions raised by our discovery of
multiple virally encoded miRNAs within KSHV infected cells:
(i) How are these viral miRNAs expressed? (ii) What do these
viral miRNAs do? Analysis of cellular miRNA transcription has
shown that all miRNAs analyzed so far are transcribed by RNA
polymerase II and that pri-miRNAs are capped, polyadenylated
transcripts structurally analogous to pre-mRNAs (17, 40). One
pri-miRNA may contain one or a cluster of several pre-miRNA
hairpins that are individually excised by Drosha during the first
step in miRNA processing. Pre-miRNA stem–loops are gener-
ally found either within the exons of noncoding mRNAs or
within the introns of coding or noncoding mRNAs.
The fact that all 10 KSHV miRNA stem–loops are located in
one short segment in the ?141-kb KSHV genome, and are all
oriented in the same direction, suggests that all 10 miRNAs may
derive from a single pri-miRNA transcript. A candidate pre-
mRNA that could serve as the KSHV pri-miRNA has been
described by Li et al. (41), who reported a K12 mRNA that
initiated at position 123,842, i.e., near the C-terminus of ORF73,
and was polyadenylated at position 117,432, i.e., 3? to the K12
ORF (Fig. 3). Importantly, this mRNA contains a single intron
extending from 123,594 to 118,779 that would encompass all of
the KSHV pre-miRNA stem–loops except miR-K10 (Fig. 3 and
Table 2). Although the hypothesis that nine of the KSHV
miRNAs are processed out of an intron present in a K12 mRNA
is attractive, it remains possible that one or more KSHV
miRNAs are instead, or also, derived from an exonic location
within a noncoding RNA(s). Indeed, the finding that miR-K1,
miR-K2, and miR-K3 expression is enhanced ?2-fold by TPA
treatment, whereas the remaining KSHV miRNAs are essen-
tially unaffected (Fig. 2), seems difficult to reconcile with the
idea that they all derive from the same pri-miRNA precursor.
Finally, we note that the miR-K10 stem–loop is currently unique
in being located within a known coding ORF (Fig. 3). Because
K12 is certainly expressed in KSHV infected cells, we can only
of the K12 mRNA is exported to the cytoplasm before nuclear
Drosha cleavage. The low level of recovery of miR-K10 as a
cDNA clone (Table 2) may be consistent with this hypothesis,
although miR-K10 was certainly detectable by Northern analysis
A previous report documenting the expression of five viral
miRNAs in EBV-infected cells (15), together with this current
report documenting the expression of at least 11 miRNAs in
KSHV-infected cells, suggests that miRNAs may play a critical
initiated by these, and possibly other, herpesviruses. The fairly
high level of expression of the KSHV miRNAs, i.e., ?2,200
copies of miR-K6-3p and ?800 copies of miR-K7 per BC-1 cell
(Fig. 5), and the observation that KSHV and EBV each account
for a third or more of the miRNAs detected in BC-1 cells (Table
1), is certainly consistent with this hypothesis. Although we have
not observed any significant sequence similarity between the 14
KSHV miRNAs described here (Table 2) and the five previously
reported EBV miRNAs, it is certainly possible that one or more
of these viral miRNAs targets the same cellular mRNA.
Animal miRNAs generally target imperfectly complementary
target sites located in the 3? UTR of mRNAs and then induce
translational arrest (1). This process is highly cooperative, and it
has been proposed that this cooperativity has evolved to allow
the coordinate posttranscriptional regulation of gene expression
by different miRNAs acting in concert (14). Nevertheless, ver-
tebrate miRNAs are fully capable of inducing mRNA cleavage
when they encounter a perfectly complementary mRNA target
(10, 11), and this form of RNA interference is indeed the norm
for plant miRNAs (1).
coevolved to permit one miRNA to coordinately target multiple
mRNAs, this seems unlikely to be true for the cellular targets of
viral miRNAs. Therefore, we hypothesize that viral miRNAs
may, in fact, function to induce the degradation of one or a small
number of host mRNAs. Nevertheless, it is also possible that the
KSHV miRNAs do function as true miRNAs that, perhaps
acting in concert with each other or with cellular miRNAs, block
the translation of specific host mRNAs. Finally, it is possible that
one or more of the KSHV miRNAs are actually designed to
regulate viral mRNA expression, as has indeed been proposed
for one of the EBV miRNAs (15).
Based on these three possible scenarios, we have used com-
putational methods to perform searches for human and KSHV
mRNA that contain potential targets for each of the KSHV
miRNAs listed in Table 2. Cellular mRNAs that contain poten-
tial targets highly complementary to a KSHV miRNA are listed
in Table 4, which is published as supporting information on the
PNAS web site, cellular mRNAs that contain potential partially
overlaps the K12 ORF. White boxes, KSHV genes expressed during latent infection; gray boxes, KSHV genes expressed only during lytic infection; black boxes,
KSHV pri-miRNA stem–loop precursors. The genomic orientation of each gene or miRNA is indicated by the arrow at the end of each box.
Genomic location of the KSHV miRNAs. (Upper) A schematic of the 3? end of the KSHV DNA genome with known ORFs, their orientation and expression
www.pnas.org?cgi?doi?10.1073?pnas.0408192102 Cai et al.
complementary targets within the 3? UTR are listed in Table 5, Download full-text
which is published as supporting information on the PNAS web
site, and cellular mRNAs that contain potential partially com-
plementary targets within the coding region are listed in Table
6, which is published as supporting information on the PNAS
web site. Finally, KSHV mRNAs that contain potential targets
partially complementary to a KSHV miRNA are listed in Table
7, which is published as supporting information on the PNAS
KSHV miRNAs within the coding regions of the KSHV genome.
Thus far, it has proven very difficult to identify authentic
mRNA targets for vertebrate miRNAs. Indeed, only one verte-
brate target, the HOX8 mRNA cleaved by miR-196, has been
on computer analyses (1, 43). Because we have not yet analyzed
the expression level of candidate target mRNAs or proteins in
the presence or absence or specific KSHV miRNAs, the validity
of the potential mRNA targets listed in Tables 4–7 is unclear.
This is particularly true for the partially complementary targets
listed in Tables 5–7, because genuine miRNA regulatory sites
frequently do not display a statistically significant level of
complementarity (43). With this caveat in mind, we note that
Tables 4–6 include a number of interesting candidate target
genes whose products are involved in a range of cellular activ-
ities, including apoptosis and signaling. A number of B cell-
specific genes were also detected as potential targets in this
computational analysis, including the mRNAs encoding SWAP-
70, ETV5, and BLR1, which have been reported to be repressed
in KSHV latently infected cells based on microarray analysis (44,
45). Although the identification of the mRNA targets for the
KSHV miRNAs described in this manuscript is clearly a critical
goal, the fact that these KSHV miRNAs are expressed at readily
detectable levels in latently KSHV infected cells (Fig. 2) and are
fully conserved in all KSHV genome sequences present in the
database argues that they are likely to play a key role in KSHV
replication and pathogenesis.
We thank Dirk Dittmer for helpful discussions, Bala Chandran (Uni-
versity of Kansas Medical Center, Kansas City) for KSHV ORF59-
specific antibodies, Yuan Chang (University of Pittsburgh Cancer Cen-
ter, Pittsburgh) for the BC-1 cell line, and R. Renne (University of
Florida, Gainesville) for the BCBL-1 cell line. This work was sponsored
by the Howard Hughes Medical Institute and by National Institutes of
Health Grants GM071408 (to B.R.C.) and CA096500 (to B.D.).
1. Bartel, D. P. (2004) Cell 116, 281–297.
2. Lee, Y., Jeon, K., Lee, J.-T., Kim, S. & Kim, V. N. (2002) EMBO J. 21,
3. Lee, Y., Ahn, C., Han, J., Choi, H., Kim, J., Yim, J., Lee, J., Provost, P.,
Rådmark, O., Kim, S. & Kim, V. N. (2003) Nature 425, 415–419.
4. Hutva ´gner, G., McLachlan, J., Pasquinelli, A. E., Ba ´lint, E´., Tuschl, T. &
Zamore, P. D. (2001) Science 293, 834–838.
5. Grishok, A., Pasquinelli, A. E., Conte, D., Li, N., Parrish, S., Ha, I., Baillie,
D. L., Fire, A., Ruvkun, G. & Mello, C. C. (2001) Cell 106, 23–34.
6. Hammond, S. M., Bernstein, E., Beach, D. & Hannon, G. J. (2000) Nature 404,
7. Schwarz, D. S., Hutva ´gner, G., Haley, B. & Zamore, P. D. (2002) Mol. Cell 10,
8. Martinez, J., Patkaniowska, A., Urlaub, H., Lu ¨hrmann, R. & Tuschl, T. (2002)
Cell 110, 563–574.
9. Meister, G., Landthaler, M., Patkaniowska, A., Dorsett, Y., Teng, G. & Tuschl,
T. (2004) Mol. Cell 15, 185–197.
10. Hutva ´gner, G & Zamore, P. D. (2002) Science 297, 2056–2060.
11. Zeng, Y., Yi, R. & Cullen, B. R. (2003) Proc. Natl. Acad. Sci. USA 100,
12. Doench, J. G., Petersen, C. P. & Sharp, P. A. (2003) Genes Dev. 17, 438–442.
13. Olsen, P. H. & Ambros, V. (1999) Dev. Biol. 216, 671–680.
14. Doench, J. G. & Sharp, P. A. (2004) Genes Dev. 18, 504–511.
15. Pfeffer, S., Zavolan, M., Gra ¨sser, F. A., Chien, M., Russo, J. J., Ju, J., John, B.,
Enright, A. J., Marks, D., Sander, C., et al. (2004) Science 304, 734–736.
16. Lau, N. C., Lim, L. P., Weinstein, E. G. & Bartel, D. P. (2001) Science 294,
17. Cai, X., Hagedorn, C. H. & Cullen, B. R. (2004) RNA 10, 1957–1966.
18. Fakhari, F. D. & Dittmer, D. P. (2002) J. Virol. 76, 6213–6223.
19. Chang, Y., Cesarman, E., Pessin, M. S., Lee, F., Culpepper, J., Knowles, D. M.
& Moore, P. S. (1994) Science 266, 1865–1869.
20. Said, W., Chien, K., Takeuchi, S., Tasaka, T., Asou, H., Cho, S. K., de Vos, S.,
Cesarman, E., Knowles, D. M. & Koeffler, H. P. (1996) Blood 87, 4937–4943.
21. Soulier, J., Grollet, L., Oksenhendler, E., Cacoub, P., Cazals-Hatem, D.,
Babinet, P., d’Agay, M. F., Clauvel, J. P., Raphael, M. & Degos, L. (1995) Blood
22. Cesarman, E., Moore, P. S., Rao, P. H., Inghirami, G., Knowles, D. M. &
Chang, Y. (1995) Blood 86, 2708–2714.
23. Miller, G., Heston, L., Grogan, E., Gradoville, L., Rigsby, M., Sun, R., Shedd,
D., Kushnaryov, Y. M., Grossberg, S. & Chang, Y. (1997) J. Virol. 71, 314–324.
24. Ambros, V., Bartel, B., Bartel, D. P., Burge, C. B., Carrington, J. C., Chen, X.,
Dreyfuss, G., Eddy, S. R., Griffiths-Jones, S., Marshall, M., et al. (2003) RNA
25. Zeng, Y. & Cullen, B. R. (2003) RNA 9, 112–123.
26. Lagos-Quintana, M., Rauhut, R., Lendeckel, W. & Tuschl, T. (2001) Science
27. Schwarz, D. S., Hutva ´gner, G., Du, T., Xu, Z., Aronin, N. & Zamore, P. D.
(2003) Cell 115, 199–208.
28. Khvorova, A., Reynolds, A. & Jayasena, S. D. (2003) Cell 115, 209–216.
29. Renne, R., Zhong, W., Herndier, B., Mcgrath, M., Abbey, N., Kedes, D. &
Ganem, D. (1996) Nat. Med. 2, 342–346.
30. Chan, S. R., Bloomer, C. & Chandran, B. (1998) Virology 240, 118–126.
31. Dittmer, D., Lagunoff, M., Renne, R., Staskus, K., Haase, A. & Ganem, D.
(1998) J. Virol. 72, 8309–8315.
32. Jenner, R. G., Alba, M. M., Boshoff, C. & Kellam, P. (2001) J. Virol. 75,
33. Sarid, R., Flore, O., Bohenzky, R. A., Chang, Y. & Moore, P. S. (1998) J. Virol.
34. Lim, L. P., Lau, N. C., Weinstein, E. G., Abdelhakim, A., Yekta, S., Rhoades,
M. W., Burge, C. B. & Bartel, D. P. (2003) Genes Dev. 17, 991–1008.
35. Lallemand, F., Desire, N., Rozenbaum, W., Nicolas, J. C. & Marechal, V.
(2000) J. Clin. Microbiol. 38, 1404–1408.
36. Moore, P. S. & Chang, Y. (1995) N. Engl. J. Med. 332, 1181–1185.
37. Wabinga, H. R., Parkin, D. M., Wabwire-Mangen, F. & Mugerwa, J. W. (1993)
Int. J. Cancer 54, 26–36.
38. Grundhoff, A. & Ganem, D. (2004) J. Clin. Invest. 113, 124–136.
39. Russo, J. J., Bohenzky, R. A., Chien, M.-C., Chen, J., Yan, M., Maddalena, D.,
Sci. USA 93, 14862–14867.
EMBO J. 23, 4051–4060.
41. Li, H., Komatsu, T., Dezube, B. J. & Kaye, K. M. (2002) J. Virol. 76,
42. Yekta, S., Shih, I. & Bartel, D. P. (2004) Science 304, 594–596.
43. Lai, E. C. (2004) Genome Biol. 5, 115.
44. Jenner, R. G., Maillard, K., Cattini, N., Weiss, R. A., Boshoff, C., Wooster, R.
& Kellam, P. (2003) Proc. Natl. Acad. Sci. USA 100, 10399–10404.
45. Klein, U., Gloghini, A., Galdano, G., Chadburn, A., Cesarman, E., Dalla-
Favera, R. & Carbone, A. (2003) Blood 101, 4115–4121.
46. Zuker, M. (2003) Nucleic Acids Res. 31, 3406–3415.
Cai et al.
April 12, 2005 ?
vol. 102 ?
no. 15 ?