Tat-SF1 Is Not Required for Tat Transactivation but Does
Regulate the Relative Levels of Unspliced and Spliced
Heather B. Miller1,2, Kevin O. Saunders1,3, Georgia D. Tomaras1,3, Mariano A. Garcia-Blanco1,2,4*
1Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, North Carolina, United States of America, 2Center for RNA Biology, Duke
University Medical Center, Durham, North Carolina, United States of America, 3Department of Surgery, Duke University Medical Center, Durham, North Carolina, United
States of America, 4Department of Medicine, Duke University Medical Center, Durham, North Carolina, United States of America
Background: HIV-1 relies on several host proteins for productive viral transcription. HIV-1 Tat-specific factor 1 (Tat-SF1) is
among these cofactors that were identified by in vitro reconstituted transcription reactions with immunodepleted nuclear
extracts. At the onset of this work, the prevailing hypothesis was that Tat-SF1 was a required cofactor for the viral regulatory
protein, Tat; however, this had not previously been formally tested in vivo.
Methodology/Principal Findings: To directly address the involvement of Tat-SF1 in HIV-1 gene expression, we depleted
Tat-SF1 in HeLa cells by conventional expression of shRNAs and in T- Rex -293 cells containing tetracycline-inducible shRNAs
targeting Tat-SF1. We achieved efficient depletion of Tat-SF1 and demonstrated that this did not affect cell viability. HIV-1
infectivity decreased in Tat-SF1-depleted cells, but only when multiple rounds of infection occurred. Neither Tat-dependent
nor basal transcription from the HIV-1 LTR was affected by Tat-SF1 depletion, suggesting that the decrease in infectivity was
due to a deficiency at a later step in the viral lifecycle. Finally, Tat-SF1 depletion resulted in an increase in the ratio of
unspliced to spliced viral transcripts.
Conclusions/Significance: Tat-SF1 is not required for regulating HIV-1 transcription, but is required for maintaining the
ratios of different classes of HIV-1 transcripts. These new findings highlight a novel, post-transcriptional role for Tat-SF1 in
the HIV-1 life cycle.
Citation: Miller HB, Saunders KO, Tomaras GD, Garcia-Blanco MA (2009) Tat-SF1 Is Not Required for Tat Transactivation but Does Regulate the Relative Levels of
Unspliced and Spliced HIV-1 RNAs. PLoS ONE 4(5): e5710. doi:10.1371/journal.pone.0005710
Editor: Olivier Schwartz, Institut Pasteur, France
Received February 27, 2009; Accepted April 21, 2009; Published May 27, 2009
Copyright: ? 2009 Miller et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by a grant (R01-GM071037) to M.A.G-B., and a grant (R01-AI052779) to G.D.T. Work done through the Duke CFAR was
supported by a grant (P30 AI 64518) and for the GHRB PhosphorImager Core by a NIAID contract (UC6 AI058607). The funders had no role in study design, data
collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com
The human immunodeficiency virus type 1 (HIV-1), like all
other complex retroviruses, tightly regulates transcription from its
genome. This regulation is mediated by both viral and cellular
factors [1,2,3,4]. The viral regulatory protein, Tat, stimulates
transcription elongation of HIV-1 through a series of events
termed Tat transactivation [5,6,7,8,9,10,11,12]. Tat recruits the
human positive transcription elongation factor b (P-TEFb) to the
TAR RNA element at the 59 end of nascent transcripts [2,5]. Tat
interacts directly with cyclin T1 (CCNT1), a component of P-
TEFb, which allows recognition of TAR . P-TEFb recruitment
has been proposed to be necessary and sufficient for transcrip-
tional elongation . The CDK9 kinase activity of P-TEFb
results in hyperphosphorylation of the carboxyl-terminus domain
(CTD) of the largest subunit of RNA Polymerase II (RNAPII),
leading to efficient elongation [7,8,9,10,14,15].
Many groups have investigated the mechanism by which HIV-1
utilizes P-TEFb as a cellular cofactor for Tat transactivation. These
studies suggest that P-TEFb is part of a multiprotein complex that
associates with RNAPII at the HIV-1 promoter and that other
cellular factors also assist in transactivation [16,17,18]. Previous
studieshaveused nuclearextract fractionsfromTataffinitycolumns
to reconstitute Tat transactivation in vitro [16,18,19,20]. One of
these studies identified a cellular activity that was required for Tat-
specific, TAR-dependent activation of HIV-1 transcription in vitro,
and it was termed Tat stimulatory factor (Tat-SF) [18,21]. Further
affinity purification of this activity identified a novel, 140-kDa
protein that was sequenced and named Tat-SF1 (accession number
NP_055315; HUGO gene name HTAT-SF1). Immunodepletion of
this protein from nuclear extracts resulted in a reduction in Tat
transactivation [16,18,19], and overexpression of Tat-SF1 resulted
in a small increase in Tat transactivation . The increase in Tat
transactivation, however, was primarily due to a decrease in basal
transcription, and only a small increase in Tat-dependent
transcription . In addition to the usual caveats of overexpression
data, this result was not recapitulated by the same group when using
a different plasmid system . These studies, which used the best
technology available at the time, justifiably concluded that Tat-SF1
was a likely cofactor for Tat transactivation in vitro and in vivo.
PLoS ONE | www.plosone.org1 May 2009 | Volume 4 | Issue 5 | e5710
Tat-SF1 has also been proposed to be a general elongation
factor [18,19]. It can associate with Tat:P-TEFb transcription
elongation complexes in nuclear extracts  through a direct
CCNT1 interaction [14,23]. Two other transcription elongation
factors, hSPT5 and the RAP30 protein of TFIIF, associate with
Tat-SF1 . Tat-SF1 has been shown to be a component of an
RNAPII-containing complex that also contains other HIV-1
cellular cofactors such as P-TEFb and hSPT5, and these factors
were shown to be recruited to the HIV-1 promoter in HeLa
nuclear extract . In a separate study, immunoprecipitation
experiments showed that Tat-SF1, along with P-TEFb, TCERG1
(CA150), and TFIIF all associate in an RNAPII-containing
In addition to the associations with transcription factors, Tat-
SF1 has also been found to interact with several components of the
spliceosome. Large RNAP II-containing complexes that associate
with 59-splice sites contain Tat-SF1 . Tat-SF1 also interacts
with snRNP proteins U1 70 K, U2B0, and Sm proteins B and B9.
In addition, Tat-SF1 associates with all five spliceosomal U
snRNAs, and this interaction depends on its RNA recognition
motifs (RRMs) . Moreover, the yeast homologue of Tat-SF1,
CUS2, helps refold U2 snRNAs to aid in prespliceosome assembly
. The association with both elongation and splicing factors has
led to the suggestion that Tat-SF1 can couple these two processes
. Indeed, another transcription-splicing coupling factor,
TCERG1, binds Tat-SF1 directly through multiple interactions
with FF domains in the former .
Insight into the role of Tat-SF1 in the HIV-1 lifecycle has
previously been limited to immunodepletions and in vitro analyses
or transient overexpression experiments. In this manuscript, we
present studies that utilize RNA interference (RNAi) to reevaluate
Tat-SF1’s role in Tat transactivation and HIV-1 replication in vivo.
We found that Tat-SF1 depletion did not affect transcription from
the HIV-1 LTR and did not alter the overall level of viral
transcripts; however, Tat-SF1 depletion resulted in a significant
decrease in viral replication. This study demonstrates that the
major effect upon knockdown of Tat-SF1 was a change in the ratio
of unspliced to fully spliced HIV-1 RNAs. Based on our data, we
propose a novel activity for Tat-SF1 as a post-transcriptional
regulator of viral pre-mRNAs.
Materials and Methods
Two different short hairpin RNA (shRNA) sequences targeting
Tat-SF1 transcripts in the pSuper vector backbone (OligoEngine)
were gifts from Dr. Bryan Cullen (Duke University). Tat-SF1(A)
consists of a hairpin targeting the following Tat-SF1 sequence:
GAATCTGTGGAACTTGC. A third shRNA targeting Tat-
SF1, Tat-SF1(C), was expressed from the pSM2 vector (Open
Biosystems), obtained from the Duke RNAi Facility. Tat-SF1(C)
targets the following Tat-SF1 sequence: GGCCTTCTAGAG-
CAAGGCATTT. The GFP targeting and non-silencing control
shRNAs were also in pSM2. Empty pSuper, used as a negative
control in transient knockdowns, was a gift from Dr. Vann Bennett
(Duke University). To create tetracycline inducible shRNA
constructs, hairpin sequences were subcloned from pSuper by
digestion with BamH1 and XhoI and ligated into the correspond-
ing sites in pcDNA5/FRT/TO (Invitrogen). Hairpin sequences
targeting GFP and Tat-SF1 in the pSM2 vector were PCR
amplified using the following oligonucleotides: 59-CCA TGG
GGA TCC CAG CAC ATA TAC TAG TCG AC 39 AND 59
CCA TGC GGC CGC TAA TTC AGC TTT GTA AAA ATG-
39. Gel-purified PCR products were digested with BamH1 and
NotI and ligated into the corresponding sites in pcDNA5/FRT/
TO. The structure of all constructs was verified by DNA sequence
The HIV-CAT reporter and the CMV-driven Tat plasmid,
pcTat have been described previously [29,30]. The SV40-driven
luciferase reporter, pGL3-Control Vector (Promega), was used to
control for transfection efficiency. The HIV-1 proviral indicator
plasmid, pNL-Luc-HXB was described in . The HIV-1 pNL4-
3 plasmid, pSG3DEnv, was described in . pSG3DEnv contains
a four nucleotide insertion mutation that encodes a truncated
gp120, without disrupting any of the known splicing regulatory cis
elements. The plasmid expressing the VSV-G envelope, pHIT/G
was described in . The plasmid expressing the AMLV
envelope, pSV-A-MLV-env was described in .
The Flp-InTMT- RexTM-293 cell line is a stable, tetracycline-
inducible, mammalian expression cell line. It was used here to
express shRNAs under the control of a CMV promoter that
contained two tandem repeats of the tet operator 2 sequence. T-
Rex-293 cells that harbored the pFRT/lacZeo and pcDNA6/TR
plasmids were cultured in DMEM (high glucose) supplemented
with 10% fetal bovine serum, 1% penicillin/streptomycin and 1%
L-glutamine. Zeocin (100 mg/mL) and blasticidin (15 mg/mL)
were included to select for cells containing a single integrated FRT
site and the Tet repressor plasmid. A GFP-expressing plasmid,
Gint , was linearized with DraIII and transfected into T-Rex-
293 cells using Lipofectamine 2000 (Invitrogen). Genetecin
(500 mg/mL) was added to culture media and was changed every
three days until a stable, polyclonal population was obtained.
These cells were sorted by fluorescence activated cell sorting
(FACS) to isolate the population with high GFP expression.
shRNAs described above in pcDNA5/FRT/TO and empty vector
pcDNA5/FRT/TO were each cotransfected with pOG44 (to
express Flp recombinase) into GFP-positive T-Rex -293 cells, and
stably transfected cells were selected with hygromycin (200 mg/
mL), blasticidin (15 mg/mL), and geneticin (500 mg/mL). Tetra-
cycline-reduced FBS (Hyclone) was used for all cultures that
contained a tetracycline-inducible plasmid to minimize basal
shRNA expression. Tetracycline concentrations and induction
times were optimized for each shRNA expressed to achieve
maximal knockdown. Cell viability was assessed at various time
pointstime points after shRNA induction using trypan blue
exclusion. The dye was added to trypsinized cells to 0.2% (v/v),
and viable cells were counted on a hemocytometer. Cell
confluency was evaluated at various time points by imaging cells
on an Olympus (Melville, NY) IX71 epifluorescence microscope.
Images were acquired with an Olympus DP70 digital camera, and
images were processed with DP Controller software (Olympus).
HeLa cells were cultured in DMEM (high glucose) with 10%
FBS and 1% penicillin/streptomycin. HeLa cells stably expressing
the non-silencing shRNA or Tat-SF1(C) were created by
linearizing the plasmids with BstBI and ApaI, respectively, before
transfection into HeLa cells using Lipofectamine 2000. Positive
transfectants were selected with puromycin. 293T and TZM-bl
cells  were cultured in DMEM (high glucose) with 10% FBS
and 1% penicillin/streptomycin.
Western blot analysis
Protein concentrations of cell lysates were determined by a
Bradford assay (Bio-Rad) and 30 mg of protein was separated on
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org2 May 2009 | Volume 4 | Issue 5 | e5710
SDS-PAGE gels before transferring to PVDF membranes
(BioRad). The membranes were probed with rabbit polyclonal
antibodies to CCNT1 (AbCam) at a 1:500 dilution, PTB (Intronn,
LLC, Durham, NC) at a 1:5000 dilution or antiserum to Tat-SF1
(Research Genetics Inc, Huntsville, AL) at a 1:500 dilution. Rabbit
secondary antibody (Amersham) was used at a 1:5000 dilution and
proteins were detected with ECL (Amersham) or SuperSignal
West Chemiluminescence Substrate (Pierce).
Pseudotyped virus production and HIV-1 replication
293T cells were plated in 100-mm dishes and cotransfected with
1 mg pHIT/G (for VSV-G envelope production) or 1 mg pSV-A-
MLV-env (for AMLV envelope production) and 10 mg pNL-Luc-
HXB using the calcium phosphate method. Media was changed
the following day and supernatants were passed through a 0.45-
mm pore size filter 48 hours post transfection. Viral supernatants
not used immediately were frozen at 280uC.
T-Rex-293 cell lines expressing either empty pcDNA5/FRT/
TO or a plasmid with an shRNA targeting GFP or Tat-SF1 were
induced with 2.5 mg/mL tetracycline for 48 hours. Then, 26105
cells were plated per well of a 6-well plate, keeping tetracycline
present in the media. Cells were imaged in phase contrast and
fluorescence before and after infection to confirm equal con-
fluency. 72 hours post induction, media was removed and 2 mL of
viral supernatant was added to the cells. HeLa cells stably
expressing either the non-silencing shRNA or an shRNA targeting
Tat-SF1 were infected in the same manner. Since 293 and HeLa
cells do not express the HIV-1 specific CD4 receptor, only VSV-G
or AMLV pseudotyped viruses were able to infect the target cells.
After 24 or 48 hours, a luciferase assay was performed on the cell
lysates. All values were normalized to the protein concentration of
the lysate, determined by the BCA Protein Assay (Pierce).
TZM-bl assays of infectivity
TZM-bl cells were used for replication-competent HIV-1
infection because they express endogenous CXCR4, transgenic
CD4 and CCR5, and integrated Tat-dependent beta-galactosidase
and luciferase reporter genes. They have also been shown to be
susceptible to RNAi . Empty vector, a non-silencing shRNA,
or Tat-SF1-targeting shRNAs were co-transfected with a GFP-
expressing plasmid into TZM-bl cells using FuGene6 (Roche).
Forty-eight hours post-transfection, GFP-positive cells were
collected by FACS. 86103sorted cells were plated per well in a
96-well plate in DMEM supplemented with DEAE (15 ng/mL).
For reverse transcriptase assays, 1.66104sorted cells were plated
per well in a 12-well plate. Plated cells were infected with
replication-competent HIV-1TT31, HIV-1JRFL, or HIV-1IIIB at
multiplicities of infections of 0.004, 0.02, or 0.05 respectively.
HIV-1TT31is a chimeric virus composed of the env gene from an
early-transmitted primary virus isolate  and the genome of
HIV-1NL4-3. HIV-1TT31was constructed by cotransfecting 293T
cells with a single-genome-amplified env gene from patient TT31,
and the recombinant HIVNL4-3backbone, pHIVenvBstEIInef-hisD
. Supernatants from the transfected cells were harvested 3 days
post-transfection and used to generate viral stocks in peripheral
blood mononuclear cells. At 3 or 6 days post-infection with
replication-competent virus, lysates were prepared for luciferase
assays with the Britelite Gene Assay System (PerkinElmer)
according to the manufacturer’s instructions. Luciferase values
were read on a Victor3(PerkinElmer) reader. Values shown are
the means of three independently transfected, FACS sorted, and
infected wells. The error bars represent standard error. Lysates
from GFP-sorted, uninfected cells were prepared in SDS lysis
buffer to analyze levels of Tat-SF1 by western blot both at the time
of infection (48 hours after knockdown) and at the end of the
Reverse transcriptase assays of infectivity
Virion-associated reverse transcriptase was measured as de-
scribed previously [40,41]. Briefly, cell culture supernatants from
cells infected with HIV-1IIIBwere harvested by centrifugation for
5 min at 1500 rpm and treated with Triton X-100 at a final
concentration of 1%. 10 mL of treated supernatant was added to
40 ml of reaction cocktail containing: 50 mM Tris-HCl pH 7.8,
75 mM KCl, 2 mM Dithiothreital, 5 mM MgCl2, 5 mg/mL Poly
A, 0.03 units/mL oligo dT12–18, 0.05% Igepal CA-630, 10 mM
EGTA, and 10 mCi/mL dTTP [32P] (MP Biomedical, Solon,OH).
Reactions were incubated at 37uC for 90 min. 40 mL of the
reverse transcription reaction was blotted onto DE 81 membrane
(Thermal Scientific, Odessa, TX) using a Schleicher and Schuell
Minifold and a vacuum. The blotted membranes were washed
briefly with 26 SSC (0.3 M NaCl, 0.03 M Sodium Citrate) at
room temperature and exposed to a storage phosphor screen
overnight. The screen was scanned with a Typhoon phosphor
image, and the data analyzed with ImageQuant 5.2 software (GE
Healthcare, Piscataway, NJ).
Tat transactivation assays
HeLa cells were seeded in a 24-well plate before transient
transfection with 0.4 mg of either empty pSuper plasmid, a non-
silencing shRNA , or Tat-SF1(A) and Tat-SF1(B) in combination
using Lipofectamine 2000. After 24 hours of knockdown, cells
were passaged to a 6-well plate before transient transfection of
reporter plasmids. T-Rex-293 cells were induced with 2.5 mg/mL
tetracycline for 48 hrs and seeded in a 6-well plate before transient
transfection of reporter plasmids. The calcium phosphate method
was used to transfect 0.5 mg of HIV-CAT, 50 ng of SV40-LUC as
an internal control for transfection efficiency, and the indicated
amount of either pcTat or a CMV-driven empty vector. Twenty-
four hours later, lysates were prepared using 500 mL of Cell
Culture Lysis Reagent (Promega). Luciferase assays were per-
formed with 40 mL of lysate using a Luciferase Assay Kit
(Promega) according to the manufacturer’s instructions. Readings
were obtained using a Lumat LB 9507 Luminometer (EG&G
Berthold). CAT assays were performed using the diffusion method
on the same volume of lysate . Both reporter gene assays were
background corrected with values obtained from lysis buffer alone.
All CAT values were normalized to luciferase values.
Northern blot analysis
T-Rex-293 cells were induced with 2.5 mg/mL of tetracycline
for 48 hours, plated in 6-well plates (46105cells/well), and after
an additional 24 hours, transiently transfected with 1 mg of
pSG3DEnv by the calcium phosphate method. Each shRNA-
expressing cell line was transfected in triplicate and tetracycline
was present in the media at all times. Media was replaced the day
following transfection, and 48 hours post-transfection, cells were
washed and total RNA was extracted with TRIzol (Invitrogen).
4 mg of total RNA was subjected to a glyoxylation reaction for
1 hour at 55uC. RNAs were immediately chilled for 10 minutes on
ice before separation on a 0.8% agarose gel. RNA was transferred
overnight to BrightStar-Plus positively charged nylon membrane
(Ambion) by capillary action in 106 SSC. After crosslinking the
membrane in a UV Stratalinker 2400 (Stratagene), transfer
efficiency was assessed by staining the membrane with methylene
blue and the agarose gel with ethidium bromide. Membranes were
prehybridized for 2 hours in 0.5 M Na3PO4, pH 7.2, 7% (w/v)
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org3 May 2009 | Volume 4 | Issue 5 | e5710
SDS, 1 mM EDTA, pH 7.0 at 68uC. An HIV-1 LTR radiola-
belled probe was generated from a 25 ng KpnI-HindIII fragment
of the HIV-1 LTR described in  using [a-32P]dCTP and the
Random Primers DNA Labeling Kit (Invitrogen). Denatured
probe (.16107cpm) was added to hybridization solution and
incubated with the membrane for at least 1 hour at 68uC.
Membranes were washed once at room temperature in 16
SSC+0.1% SDS, then three times at 68uC in 0.56 SSC+0.1%
SDS, followed by overnight exposure to film with an intensifying
screen at 280uC or a phosphor screen. Membranes were stripped
with boiling 0.5% SDS and reprobed with a random primed
GAPDH PCR product. Northern blots were quantified with
ImageQuant by first background correcting each band. The
proportion of each of the three HIV-1 RNA classes in cells
depleted of Tat-SF1 was expressed relative to the proportion of the
same RNA class in cells treated with the control shRNA.
Quantification of total HIV-1 transcripts levels was achieved by
adding the values of the 9 kb, 4 kb, and 2 kb signals and
normalizing to the GAPDH signal. Statistical significance was
determined using a paired Student’s t test, and significance was set
Reverse transcription and real-time PCR detection of HIV-
Total cellular RNA was isolated using TRIzol per the
manufacturer’s protocol. RNA samples were then subjected to
two rounds of DNase digestion to remove residual DNA using
DNA-free (Ambion). To verify the removal of DNA, each RNA
sample was tested in triplicate by real-time PCR using SYBR Green
PCR Master Mix (Applied Biosystems) and HIV-1 LTR specific
primers (HIVshortF and HIVshortR) described elsewhere .
System (Applied Biosystems) with the following cycling profile: one
cycle of 50uC for 5 min, one cycle of 95uC for 10 min, 40 cycles of
95uC for 15 s and 60uC for 1 min. Reverse transcription was
performed on300 ng ofeachRNAsample for1 h at 37uC usingM-
MLV Reverse Transcriptase (Invitrogen), 6 mg of random hexamer
primers (Invitrogen), 0.67 mM dNTPs (Invitrogen), 40 U RNAse
Out (Invitrogen), and First Strand Buffer (Invitrogen) in a 45 mL
final reaction volume. The enzyme was then inactivated by
incubating the reactions at 70uC for 15 min. The resulting cDNAs
were tested in real-time PCR reactions to determine the absolute
quantity of total and unspliced HIV-1 transcripts using the PCR
master mix and cycling profile described above. Total initiated
HIV-1 transcripts were amplified using primers HIVshortF and
HIVshortR. PCR primers LA8 and LA9  were modified to
match the sequence of HIVNL4-3, and used to amplify unspliced
HIV-1 transcripts. Dissociation curves were performed on all PCR
reactions to determine the specificity of each PCR reaction. The
Software v1.2.2 (Applied Biosystems).
Tat-SF1 depletion does not affect T-Rex-293 cell viability
To test the importance of Tat-SF1 in the HIV-1 life cycle in vivo,
we used RNAi to specifically silence its expression. Four T-Rex-
293 cell lines were constructed containing four unique shRNAs
that targeted Tat-SF1, which were expressed from tetracycline-
inducible promoters (see Materials and Methods). This system
minimized continuous expression of the shRNAs until tetracycline
was added to the media, thus reducing selection based on shRNA
expression. All of the cell lines were tested for shRNA-mediated
Tat-SF1 depletion, and the two most efficient, Tat-SF1(A) and
Tat-SF1(B), were used in all subsequent experiments. In addition,
two control cell lines were made that stably integrated either an
empty vector (no shRNA) or an shRNA targeting GFP.
Tetracycline concentration and induction time were optimized
to achieve the maximum level of knockdown. Analysis of cell
lysates from induced cells shows that both cell lines that contain an
shRNA targeting Tat-SF1 have significantly lower Tat-SF1
protein levels compared to the empty vector control (Figure 1A).
The knockdown of Tat-SF1 by either shRNA did not affect the
levels of CCNT1, a direct binding partner of Tat-SF1, or the
polypyrimidine tract binding protein (PTB), an unrelated splicing
factor (Figure 1A). Flow cytometry demonstrated that the cell line
containing an shRNA targeting GFP produced an active shRNA,
as GFP levels were reduced by approximately 50% (Figure 1B)
while Tat-SF1 levels were unchanged (data not shown). Tetracy-
cline induction for 72 hours was chosen for the functional assays,
although 96 hours also showed persistent knockdown.
We first sought to evaluate the effect of Tat-SF1 knockdown on
T-Rex-293 cell viability. Trypan blue exclusion was used to
quantify viability in the control and knockdown cells at 24, 48, and
72 hours post-tetracycline induction. Figure 1C shows the average
number of viable cells per mL in each cell line. The number of
viable cells per mL in the Tat-SF1 knockdown cell lines was not
significantly different from the control cells at the end of the
experiment. The Tat-SF1(B) shRNA shows slightly lower cell
viability initially, however, this difference is not significant at later
time points. The density of the Tat-SF1 knockdown cells was also
indistinguishable from that of the control cells at each time point
(Figure 1D). These data indicate that Tat-SF1 depletion did not
affect T-Rex-293 cell viability, and therefore, we proceeded to
assay the role of Tat-SF1 as a cellular cofactor of HIV.
Tat-SF1 depletion inhibits HIV-1 infectivity
If Tat-SF1 was indeed required for Tat transactivation in vivo,
then RNAi-mediated depletion of Tat-SF1 should decrease the
ability of HIV-1 to replicate. To test this hypothesis, control and
Tat-SF1 knockdown T-Rex-293 cells were infected with a
vesicular stomatitis virus G-protein (VSV-G) pseudotyped virus
in a single-round replication assay. Although this pseudotyped
virus does not produce infectious progeny due to its lack of the
HIV-1 envelope, we use the term replication here to refer to the
reproduction of viral genomes. Serial dilutions of the viral stock
were added to cells and lysates were assayed for luciferase activity,
which indicates LTR-dependent transcription, processing, and
expression. To control for any differences in cell number between
control and knockdown cell lines, cells were imaged before and
after infection to ensure equal confluency, and luciferase values
were normalized to the protein content of the lysate. Surprisingly,
both of the Tat-SF1 knockdown cell lines showed levels of HIV-1
replication that were no different than those in the empty vector
and GFP control cells after 24 hours of infection (Figure 2A). This
remained true even with a 1:100 dilution of the viral stock. HIV-1
replication was also unaffected by Tat-SF1 knockdown after
48 hours of infection (data not shown).
cells so efficiently that any deficiency due to Tat-SF1 depletion was
being masked. To address this possibility, we also assessed replication
using an AMLV-pseudotyped virus, which infects target cells with a
lower efficiency. Again, there was no significant difference in HIV-1
replication after Tat-SF1 knockdown (Figure 2B).
Earlier studies of Tat-SF1 function were performed in HeLa
cells, so we next tested our hypothesis by infecting HeLa cells that
stably expressed a third shRNA targeting Tat-SF1, Tat-SF1(C).
Tat-SF1 depletion did not significantly affect cell viability in HeLa
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org4 May 2009 | Volume 4 | Issue 5 | e5710
Figure 1. Tat-SF1 depletion does not affect T-Rex-293 cell viability. (A) Analysis of Tat-SF1 knockdown by Western blot. T-Rex-293 cells were
induced with tetracycline and lysates were made at the time points indicated. Equal protein amounts were loaded for Western blot analysis with
antibodies against Tat-SF1, CCNT1 and PTB. Lanes 1–3 are from cells that express an empty vector control and lanes 4–9 are from cells that express
one of two unique shRNAs targeting Tat-SF1. (B) Analysis of GFP knockdown by flow cytometry. T-Rex-293 cells were induced with tetracycline for
72 hours, and resuspended cells were fixed with formaldehyde for flow cytometry analysis. The bar graph quantifies the mean GFP level of each
sample. (C) Analysis of cell viability by trypan blue exclusion. Equivalent numbers of T-Rex-293 cells (expressing either an empty vector control, an
shRNA targeting GFP, or one of two shRNAs targeting Tat-SF1) were induced with tetracycline for the times indicated. Viable cells per mL are reported
as a means of three independent experiments. Error bars represent standard error. (D) Analysis of cell viability by fluorescence microscopy. Equal
numbers of T-Rex-293 cells were induced with tetracycline for the times indicated and imaged with fluorescence microscopy.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org5 May 2009 | Volume 4 | Issue 5 | e5710
cells (data not shown). Again, replication in Tat-SF1 depleted cells
was comparable to control cells (Figure 2C). This led us to
conclude that either Tat-SF1 had a minimal effect on HIV-1
replication, or it affected a step not interrogated by the
experiments with these pseudotyped viruses. Since Brass et al.
identified Tat-SF1 as a host factor required for HIV-1 propagation
using a replication-competent virus , we favored the latter
In order to test if Tat-SF1 played a role in any step of the viral
life cycle, we decided to assess the effect of Tat-SF1 depletion with
a replication-competent virus. TZM-bl cells were used because
they express CXCR4, CD4 and CCR5, and also contain
integrated Tat-dependent beta-galactosidase and luciferase re-
porter genes. These cells are amenable to treatment with siRNAs
 and have equal susceptibility to HIV-1 infection as compared
to human peripheral blood mononuclear cells (PBMCs) [36,46].
TZM-bl cells were transiently transfected with an empty vector, a
non-silencing shRNA, or Tat-SF1 specific shRNAs. These cells
were cotransfected with a GFP expressing plasmid to sort GFP-
positive cells. GFP-positive transfectants were plated in 96-well
plates and infected with one of two different replication-competent
HIV-1 strains, HIV-1TT31 or HIV-1JRFL. HIV-1JRFL is a
laboratory adapted strain, whereas HIV-1TT31 encodes an
envelope from an early-transmitted primary isolate . After 3
days of infection, a luciferase assay was performed on the cell
lysates. Upon Tat-SF1 depletion, the level of HIV-1 infection
Figure 2. Tat-SF1 depletion does not affect pseudotyped HIV-1 replication. (A) VSV-G replication assay. Serial dilutions of pseudotyped
virus (VSV-G envelope) were used in a single-round replication assay with T-Rex-293 cells expressing an empty vector, an shRNA targeting GFP, or
shRNAs targeting Tat-SF1. A western blot confirming knockdown is shown below the chart. Luciferase values of the cell lysates were read after
24 hours. Luciferase values were background corrected and normalized to protein content. (B) AMLV replication assay. An AMLV-pseudotyped virus
was used in the same type of experiment described in (A) except that lysates were harvested after 48 hours. Values reported are the means of
triplicate wells. Error bars represent standard error. The western blot in (A) also corresponds to the cells used in this experiment. (C) VSV-G-replication
assay. Pseudotyped virus was used on HeLa cells stably expressing a non-silencing shRNA or a third shRNA targeting Tat-SF1. Luciferase values of the
cell lysates were read after 24 hours. Luciferase values were background corrected and normalized to protein content. Values are reported as the
means of triplicate wells. Error bars represent standard error.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org6 May 2009 | Volume 4 | Issue 5 | e5710
decreased approximately 3-fold compared to the control cells
(Figures 3A and B). The decrease in infectivity was similar for both
HIV-1 strains tested. A comparable decrease in infectivity after 6
days of infection was also observed (data not shown). Lysates from
uninfected cells showed persistent knockdown of Tat-SF1
throughout the length of the experiment (data not shown).
As an independent, more direct measurement of viral
infectivity, a third strain, HIV-1IIIB, was used to infect control
and Tat-SF1 knockdown TZM-bl cells, and the viral supernatants
from these cells were assayed for reverse transcriptase activity.
Luciferase readings from the infected cell lysates recapitulated the
decrease in infectivity upon Tat-SF1 depletion that was seen for
the two HIV-1 strains used previously (Figure 3C). After 3 days of
infection, both control and Tat-SF1 knockdown cells produced
background levels of reverse transcriptase. After 6 days of
infection, the most efficient knockdown cells, Tat-SF1(B), showed
an inability to produce reverse transcriptase above background
levels (Figure 3D).
Together, these data indicated that Tat-SF1 plays a positive role
in regulating the lifecycle of HIV-1 at a step not interrogated by
pseudotyped viruses. This conclusion was not easily reconciled
with an effect on Tat transactivation.
Tat-SF1 depletion does not affect Tat transactivation in
Although Tat-SF1 has been implicated in vitro as a cofactor for
the viral protein Tat, this had yet to be demonstrated in vivo. To
examine Tat transactivation when Tat-SF1 was depleted, control
and Tat-SF1 knockdown HeLa cell lines were transfected with a
chloramphenicol acetyltransferase (CAT) reporter under the
control of the HIV-1 LTR. This plasmid, along with an internal
luciferase control for transfection efficiency, was cotransfected with
either the Tat-expressing plasmid, pcTat , or an empty vector.
The amount of pcTat transfected was experimentally determined
so that Tat transactivation was in the linear range (Figure 4A). As
expected, Tat stimulated transcription of the CAT reporter gene
Figure 3. Tat-SF1 depletion inhibits HIV-1 infectivity. (A) TZM-bl assay using HIV-1TT31. Replication-competent HIV-1TT31was used to infect
TZM-bl cells that were transiently transfected with either an empty vector, a non-silencing shRNA control, or shRNAs targeting Tat-SF1. A western blot
confirming knockdown is shown below the chart. This western blot corresponds to cells used in Figures 3A–D. Luciferase values of the cell lysates
were read 3 days after infection. Luciferase values were reported as the means of triplicate wells. Error bars represent standard error. (B) TZM-bl assay
using HIV-1JRFL. Replication-competent HIV-1JRFLwas used in the same type of experiment described in (A). (C) TZM-bl assay using HIV-1IIIB.
Replication-competent HIV-1IIIBwas used in the same type of experiment described in (A) and (B), except that cell lysates were read 6 days after
infection. (D) Reverse transcriptase assay. Supernatants from infected cells in (C) were measured for reverse transcriptase activity 3 and 6 days after
infection. Values from the phosphor screen image were reported as the means of triplicate wells. Error bars represent standard error.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org7 May 2009 | Volume 4 | Issue 5 | e5710
by approximately 40-fold in control cells (Figure 4B and Table 1).
Surprisingly, similar levels of Tat transactivation were seen when
Tat-SF1 was depleted. Furthermore, basal transcription of the
LTR (in the absence of Tat) was unaffected by RNAi-mediated
depletion of Tat-SF1. In addition, the same results were
recapitulated in T-Rex-293 cells (Figure 4C and Table 2). These
findings indicate that, contrary to the in vitro results, Tat-SF1
depletion did not affect basal or Tat-mediated transactivation of
the HIV-1 LTR in vivo.
Tat-SF1 maintains the ratios of HIV-1 RNAs
Although Tat-SF1 depletion resulted in a decrease in HIV-1
infectivity, it did not affect Tat transactivation. Rather than having
a role in transcription elongation of HIV-1 RNA, it seemed
possible that Tat-SF1 could post-transcriptionally regulate viral
gene expression. To test this hypothesis, we used Northern blots to
analyze HIV-1 RNAs in control and knockdown cells. The
pSG3DEnv plasmid was transfected into GFP control and Tat-SF1
knockdown T-Rex-293 cell lines, and total RNA was harvested
48 hours later. Hybridization was performed using an HIV-1
LTR-specific radiolabeled probe that detects all three classes of
viral RNAs: unspliced (,9 kb), singly spliced (,4 kb), and fully
spliced (,2 kb). A western blot of T-Rex-293 cell lysates confirms
efficient knockdown of Tat-SF1 (Figure 5A). Figure 5B shows the
results of a representative Northern blot probing RNA from GFP
control and Tat-SF1 knockdown cell lines 48 hours after
transfection with the pSG3DEnv plasmid. In the mock transfected
lane, no viral pre-mRNA signal is detected, indicating that the
Figure 4. Tat-SF1 depletion does not affect basal or Tat-dependent transcription from the HIV-1 LTR in vivo. (A) Tat titration. HeLa cells
were cotransfected with the indicated amount of pcTat, an HIV-1 CAT reporter, and an SV40-Luciferase reporter. Lysates were subjected to a CAT
assay and normalized with values obtained from a luciferase assay 24 hours after cotransfection. All CAT and Luciferase values were background
corrected. Values shown are the means of duplicate transfections. (B) Tat transactivation assay in HeLa cells. HeLa cells were transiently transfected
with either an empty vector, a non-silencing shRNA, or two shRNAs targeting Tat-SF1. A western blot confirming knockdown is shown below the
chart. Cells were cotransfected with the same reporters as in (A) plus 0.3 ng pcTat or an empty vector. All CAT and Luciferase values were background
corrected, and CAT was normalized to Luciferase. Values shown are the means of duplicate transfections. Raw values are shown in Table 1, Exp. 1. (C)
Tat transactivation in T-Rex-293 cells. T-Rex-293 cells harboring either an empty vector control, an shRNA targeting GFP, or one of two shRNAs
targeting Tat-SF1 were induced with tetracycline for 48 hours. A western blot confirming knockdown is shown below the chart. Cells were
cotransfected with the same reporters as in (A) and (B). All CAT and Luciferase values were background corrected, and CAT was normalized to
Luciferase. Values shown are the means of triplicate transfections. Raw values are shown in Table 2. Error bars represent standard error.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org8 May 2009 | Volume 4 | Issue 5 | e5710
bands seen in the other lanes are HIV-1-specific. Quantification of
each RNA class demonstrates that intron-containing unspliced
and singly spliced transcripts were elevated and fully spliced
transcripts were reduced when Tat-SF1 was depleted (Figure 5C).
This corresponds to an unspliced/fully spliced ratio of ,0.7 in
GFP control cells and ,1.6 in Tat-SF1 depleted cells. Similar
changes in ratios were observed when another viral plasmid, pNL-
Luc-HXB, was transfected into Tat-SF1 depleted T-Rex-293 cells
(data not shown). These data further suggest a post-transcriptional
role for Tat-SF1.
To test whether or not the total amount of HIV-1 transcripts
was altered by Tat-SF1 depletion, total viral RNAs were
quantified from triplicate Northern blot experiments and normal-
ized to GAPDH. Indeed, total HIV-1 RNA levels were not
significantly different between the control and Tat-SF1 knock-
down cell lines (Figure 5D).
As an independent test of a change in the splicing pattern, the
same RNAs used in the Northern blot experiments were analyzed
by qRT-PCR. We attempted the specific amplification of the 9 kb,
4 kb and 2 kb classes using previously published oligos .
Nonetheless, separation the PCR products on an acrylamide gel
revealed that, although the 9 kb primers produced one distinct
band, the 4 kb and 2 kb primers produced multiple amplicons
larger than the expected one (likely including unspliced RNAs). In
addition, the qPCR melting curves indicated more than one
product for the 4 kb amplification. In fact, careful inspection of the
design of the oligos used in a previous report  suggests that the
reported RNA-class specificity would be exceedingly difficult to
achieve. Nevertheless, the quantification of the unspliced RNAs
over all initiated transcripts confirmed that there was an increase
in the 9 kb, unspliced transcripts upon Tat-SF1 knockdown, in
concordance with the Northern blots (Figure 5E). This change in
the viral splicing pattern suggests that Tat-SF1 plays a post-
transcriptional role in regulating the ratio of unspliced to spliced
HIV-1 RNAs in vivo.
The results presented here shed new light on the mechanism by
which the human protein, Tat-SF1, functions in HIV-1 replica-
tion. The first important conclusion was that Tat-SF1 was not
required for basal or Tat-dependent transcription from the HIV-1
LTR in vivo. This contradicts previously reported conclusions
[16,22]; however, careful reevaluation of the published data
suggests concerns about these earlier inferences. First, as part of a
multiprotein complex, immunodepletion of Tat-SF1 from nuclear
extract may have resulted in co-depletion of other proteins
essential for Tat transactivation, most notably, cyclin T1, which
directly binds Tat-SF1. As a subunit of P-TEFb, co-depletion of
cyclin T1 could certainly affect levels of transcription. Equally,
RNAi-mediated depletion of CA150, which we previously
hypothesized would be a Tat cofactor  and which is also
known to interact with P-TEFb, did not show any effect on basal
or Tat-dependent transcription from the HIV-1 LTR (data not
shown). In light of these RNAi-mediated depletion experiments, it
seems very possible that the identification of both Tat-SF1 and
CA150 as Tat cofactors from Tat-affinity column experiments
could be explained by their association with P-TEFb.
A previous report on Tat-SF1 function showed a small increase
in Tat transactivation when Tat-SF1 was overexpressed . As
described earlier, this change was primarily due to a decrease in
basal transcription, however, even this effect was not reproduced
Table 1. Effect of Tat-SF1 depletion on basal and Tat-dependent HIV-1 transcription in HeLa cells*.
Reporter GeneExp.Empty Vector Non-silencingTat-SF1
CAT1 269 1361650.6364 1255434.5301 1131037.6
2 110 2449 22.396 251126.265 234236.0
3 722 1066114.8392 1350034.4 4371066024.4
LUC1 1652890 19977331.21 203498022039471.10 203498018462920.91
2 190617 2909751.53 1385612332801.68 1729593285681.90
3 17814069635900.54 10896826917380.63 1316443769614 0.58
*Values shown are the means of duplicate wells for 3 independent experiments. Row 1 is calculated from the same experiment shown in Figure 4B. CAT activity was the
value of the slope of the linear function obtained by plotting cpm of acetylated chloramphenicol versus time. Luciferase activity of the same sample was measured with
a luminometer. For both reporter assays, background values obtained from lysis buffer alone were subtracted from each sample.
Table 2. Effect of Tat-SF1 depletion on basal and Tat-dependent HIV-1 transcription in T-Rex-293 cells*.
Reporter Gene Empty VectorGFP Tat-SF1(A)Tat-SF1(B)
CAT 30280 9.325188 7.521 1497.1 29993.4
LUC 3255164915441.5 4350652816960.6 327725218257 0.74650791981630.4
*Values shown are the means of triplicate wells, calculated from the same experiment shown in Figure 4C. CAT activity was the value of the slope of the linear function
obtained by plotting cpm of acetylated chloramphenicol versus time. Luciferase activity of the same sample was measured with a luminometer. For both reporter
assays, background values obtained from lysis buffer alone were subtracted from each sample.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org9 May 2009 | Volume 4 | Issue 5 | e5710
Figure 5. Tat-SF1 maintains the levels of unspliced and spliced HIV-1 RNAs. (A) Western blot analysis confirming knockdown in T-Rex-293
cells. (B) Representative Northern blot analysis of HIV-1 RNA classes. T-Rex-293 cells were transfected with pSG3DEnv 72 hours after tetracycline
induction. At 48 hours post-transfection, total RNA was isolated for electrophoresis and Northern blotting. A DNA probe specific to the HIV-1 LTR
detected the ,9 kb unspliced, ,4 kb singly spliced, and ,2 kb fully spliced RNAs. Lane 1 contains RNA from mock-transfected GFP control cells, lane
2 from transfected GFP control cells, and lanes 3 and 4 from transfected Tat-SF1 shRNA cells. The lower panel shows the same membrane, stripped
and reprobed for GAPDH. (C) Tat-SF1 depletion alters the levels of HIV-1 RNA classes. Values are reported as the mean proportion of each RNA class,
relative to the GFP control cells from three independent Northern blot experiments. Error bars represent standard error. Statistically significant
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org10 May 2009 | Volume 4 | Issue 5 | e5710
using a different promoter to drive expression of Tat-SF1 .
Finally, chromatin immunoprecipitation (ChIP) data indicated
that Tat-SF1 was not present at an integrated HIV-1 LTR-driven
reporter gene during Tat transactivation in vivo, while RNAPII and
P-TEFb were . This finding is difficult to reconcile with Tat-
SF1 being a required Tat cofactor. Nonetheless, when a cell line
harboring integrated proviral DNA was utilized for ChIP
experiments, Tat-SF1 was detected at the promoter-proximal
region . In light of our data, we propose that the presence of
viral splice sites or other elements in the HIV-1 transcripts explains
the recruitment of Tat-SF1 to the proviral locus.
An alternative explanation for the different conclusions drawn
from previous overexpression and our knockdown studies is that
lack of functioning Tat-SF1 can be compensated by another
cellular protein. SPT5 has also been reported to stimulate Tat
transactivation when overexpressed  and inhibits transactiva-
tion when immunodepleted from nuclear extract  or depleted
by RNAi . CA150 did not compensate for Tat-SF1 depletion,
as a double knockdown of both proteins did not affect HIV-1
transcription (data not shown). Our findings do not eliminate the
possibility that Tat-SF1 plays some role in Tat transactivation,
although it is not absolutely required. A similar caveat, which can
be leveled against all silencing experiments, is that a knockdown is
not tantamount to a knockout, and the remaining level of the
silenced gene product is sufficient for activity. Nonetheless,
depletion of Tat-SF1 did result in the reduced viral replication
and altered ratios of viral transcripts, so the preponderance of the
evidence strongly suggests that Tat-SF1 is not a required Tat
Another important and novel finding presented here is that Tat-
SF1 depletion increased the ratio of unspliced to spliced viral
transcripts. It should be noted that splicing patterns of HIV
transcripts among different cell lines are highly uniform, suggesting
that the tight regulation of these transcript ratios is crucial for the
viral lifecycle [51,52,53].
Careful investigation of the changes in HIV-1 spliced ratios over
time showed that accumulation of unspliced and singly spliced
RNAs and reduction of fully spliced RNAs did not occur until
48 hours post-transfection (data not shown). These data highlight
the importance of timing when analyzing HIV-1 spliced RNAs.
Thus, we have focused our efforts for determining Tat-SF1’s role
in maintaining RNA ratios to later time points.
While an effect on transcript ratios can be explained by several
different mechanisms, we propose that Tat-SF1 regulates
alternative splicing of HIV-1 RNAs. This is consistent with
previous data regarding the yeast protein CUS2, which is
structurally similar to human Tat-SF1. Both proteins contain
two RRMs in their N-terminus, but Tat-SF1 has a large acidic C-
terminus that is absent in CUS2. The first RRMs of these proteins
are 37% identical and 59% similar. The second RRMs are 30%
identical and 56% similar. CUS2 associates with U2 snRNA in
splicing extracts and co-immunoprecipitates PRP11, which is a
subunit of SF3a. When anti-Tat-SF1 antibodies were used for
immunoprecipitation, the human homologue of PRP11, SF3a66
(SAP62), was also immunoprecipitated . An effect of Tat-SF1
depletion on HIV-1 RNA ratios is also consistent with recent
unpublished data from our laboratory that demonstrate that Tat-
SF1 depletion changes the relative levels many alternatively spliced
transcripts in human cells without affecting the total amount of
these transcripts (H.B. Miller et al., unpublished results). An effect
on splicing could be direct, as proposed for CUS2. Thus, Tat-SF1
could help in the folding and activity of splicing factors such as U2
snRNAs, but it could also rework the folding of the HIV
transcripts leading to efficient splicing. Tat-SF1 may also have
an indirect effect on HIV-1 pre-mRNA splicing by regulating the
processing of transcripts encoding other HIV dependency factors
(HDFs) [37,54,55]. In fact, we analyzed the HDFs published by
the Brass et al. and Konig et al. screens and found approximately 2-
fold enrichment over chance alone in genes that also had evidence
of Tat-SF1-regulated alternative splicing (p-values of 0.02 and
0.05, respectively) (H.B. Miller et al, unpublished results). Tat-SF1
could also be involved in virion packaging. The decrease in
infectivity in Tat-SF1 depleted cells could be explained if Tat-SF1
was a chaperone protein, helping fold the viral pre-mRNA
genome into productive virions. Such a role in viral RNA
packaging would be consistent with Tat-SF1’s role in influenza
virus replication. Tat-SF1 was identified as a stimulatory host
factor, possibly aiding in the formation of RNA-nucleoprotein
complexes by acting as a molecular chaperone . It remains to
be seen whether Tat-SF1 binds HIV-1 pre-mRNA and helps
package viral genomes into virions. An increase in the unspliced
RNA upon Tat-SF1 knockdown could also be explained by Tat-
SF1 having a role in RNA export from the nucleus, although this
has not yet been tested. Additionally, Tat-SF1 may play a role in
regulating the levels of unspliced and spliced HIV-1 RNAs by
affecting their stability. The virus may rely on Tat-SF1 to
destabilize unspliced RNAs in order to maintain an optimal ratio
of unspliced and spliced RNAs.
Other studies that support a role for Tat-SF1 in HIV replication
found that Tat-SF1 was upregulated in either HIV-1 gp120
stimulated primary T cells  or HIV-1 Nef overexpression in T
cells . The strongest validation of Tat-SF1’s role in the HIV-1
lifecycle was revealed when it was identified in a large-scale,
RNAi-based screen for HIV-dependency factors (HDFs) .
Despite the fact that these studies refer to Tat-SF1 as a Tat
cofactor, their findings remain consistent with this protein being
involved in the post-transcriptional control of HIV-1 RNA levels
and further support the idea that Tat-SF1 depletion has a negative
effect on the viral life cycle in vivo.
We thank Hal Bogard, Dr. Jennifer Lin Umbach, and Dr. Bryan Cullen for
valuable reagents and technical advice. We thank Dr. Caroline LeSommer
for her advice in the creation of the T-Rex-293 cell lines, and other
members of the Garcia-Blanco lab for useful discussions and suggestions
during the preparation of this manuscript. We also thank Tim Robinson
and Dr. Matthew Marengo for their advice on stastical analysis. We thank
Dr. Daniel Kuritzkes for providing the HIV-1 plasmid, pHIVenvBstEIInef-
hisD, and Dr. Pinghuang Liu and Glenn Overman of the Duke Center for
AIDS Research (CFAR) Molecular Virology Core for the replication-
competent viruses and reverse transcriptase assays. The following reagents
were obtained through the AIDS Research and Reference Reagent
Program, Division of AIDS, NIAID, NIH: pSV-A-MLV-env from Dr.
differences between GFP control and Tat-SF1 knockdown conditions are indicated with asterisks. (D) Tat-SF1 depletion does not alter total HIV-1 RNA
levels. Levels of the 3 RNA classes quantified from triplicate Northern blots were totaled and normalized to GAPDH levels. Values are reported as the
means relative to the GFP control cells from three independent experiments. Error bars represent standard error. (E) Tat-SF1 depletion results in an
increase in unspliced HIV-1 transcripts. qRT-PCR was performed on the same RNA samples used for Northern blot experiments. The medians of
triplicate amplifications (both unspliced products and all initiated HIV-1 transcripts) were calculated and means of unspliced transcripts/all initiated
HIV-1 transcripts ratios from triplicate samples are reported. Error bars represent standard error.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org11 May 2009 | Volume 4 | Issue 5 | e5710
Nathaniel Landau and Dr. Dan Littman, pSG3Denv from Drs. John C.
Kappes and Xiaoyun Wu, and TZM-bl cells from Dr. John C. Kappes, Dr.
Xiaoyun Wu and Tranzyme Inc.
Conceived and designed the experiments: HM KS GT MGB. Performed
the experiments: HM KS. Analyzed the data: HM KS GT MGB. Wrote
the paper: HM KS GT MGB.
1. Jones KA, Peterlin BM (1994) Control of RNA initiation and elongation at the
HIV-1 promoter. Annu Rev Biochem 63: 717–743.
2. Cullen BR (1998) HIV-1 auxiliary proteins: making connections in a dying cell.
Cell 93: 685–692.
3. Cullen BR (1992) Mechanism of action of regulatory proteins encoded by
complex retroviruses. Microbiol Rev 56: 375–394.
4. Goff SP (2007) Host factors exploited by retroviruses. Nat Rev Microbiol 5:
5. Jones KA (1997) Taking a new TAK on tat transactivation. Genes Dev 11:
6. Wei P, Garber ME, Fang SM, Fischer WH, Jones KA (1998) A novel CDK9-
associated C-type cyclin interacts directly with HIV-1 Tat and mediates its high-
affinity, loop-specific binding to TAR RNA. Cell 92: 451–462.
7. Herrmann CH, Rice AP (1995) Lentivirus Tat proteins specifically associate
with a cellular protein kinase, TAK, that hyperphosphorylates the carboxyl-
terminal domain of the large subunit of RNA polymerase II: candidate for a Tat
cofactor. J Virol 69: 1612–1620.
8. Herrmann CH, Rice AP (1993) Specific interaction of the human immunode-
ficiency virus Tat proteins with a cellular protein kinase. Virology 197: 601–608.
9. Isel C, Karn J (1999) Direct evidence that HIV-1 Tat stimulates RNA
polymerase II carboxyl-terminal domain hyperphosphorylation during tran-
scriptional elongation. J Mol Biol 290: 929–941.
10. Mancebo HS, Lee G, Flygare J, Tomassini J, Luu P, et al. (1997) P-TEFb kinase
is required for HIV Tat transcriptional activation in vivo and in vitro. Genes
Dev 11: 2633–2644.
11. Taube R, Fujinaga K, Wimmer J, Barboric M, Peterlin BM (1999) Tat
transactivation: a model for the regulation of eukaryotic transcriptional
elongation. Virology 264: 245–253.
12. Yang X, Gold MO, Tang DN, Lewis DE, Aguilar-Cordova E, et al. (1997)
TAK, an HIV Tat-associated kinase, is a member of the cyclin-dependent family
of protein kinases and is induced by activation of peripheral blood lymphocytes
and differentiation of promonocytic cell lines. Proc Natl Acad Sci U S A 94:
13. Bieniasz PD, Grdina TA, Bogerd HP, Cullen BR (1999) Recruitment of cyclin
T1/P-TEFb to an HIV type 1 long terminal repeat promoter proximal RNA
target is both necessary and sufficient for full activation of transcription. Proc
Natl Acad Sci U S A 96: 7791–7796.
14. Zhou Q, Chen D, Pierstorff E, Luo K (1998) Transcription elongation factor P-
TEFb mediates Tat activation of HIV-1 transcription at multiple stages. Embo J
15. Dahmus ME (1996) Reversible phosphorylation of the C-terminal domain of
RNA polymerase II. J Biol Chem 271: 19009–19012.
16. Zhou Q, Sharp PA (1996) Tat-SF1: cofactor for stimulation of transcriptional
elongation by HIV-1 Tat. Science 274: 605–610.
17. Sune C, Hayashi T, Liu Y, Lane WS, Young RA, et al. (1997) CA150, a nuclear
protein associated with the RNA polymerase II holoenzyme, is involved in Tat-
activated human immunodeficiency virus type 1 transcription. Mol Cell Biol 17:
18. Parada CA, Roeder RG (1999) A novel RNA polymerase II-containing complex
potentiates Tat-enhanced HIV-1 transcription. Embo J 18: 3688–3701.
19. Li XY, Green MR (1998) The HIV-1 Tat cellular coactivator Tat-SF1 is a
general transcription elongation factor. Genes Dev 12: 2992–2996.
20. Wu-Baer F, Lane WS, Gaynor RB (1998) Role of the human homolog of the
yeast transcription factor SPT5 in HIV-1 Tat-activation. J Mol Biol 277:
21. Zhou Q, Sharp PA (1995) Novel mechanism and factor for regulation by HIV-1
Tat. Embo J 14: 321–328.
22. Kim JB, Yamaguchi Y, Wada T, Handa H, Sharp PA (1999) Tat-SF1 protein
associates with RAP30 and human SPT5 proteins. Mol Cell Biol 19: 5960–5968.
23. Fong YW, Zhou Q (2000) Relief of two built-In autoinhibitory mechanisms in P-
TEFb is required for assembly of a multicomponent transcription elongation
complex at the human immunodeficiency virus type 1 promoter. Mol Cell Biol
24. Sanchez-Alvarez M, Goldstrohm AC, Garcia-Blanco MA, Sune C (2006)
Human transcription elongation factor CA150 localizes to splicing factor-rich
nuclear speckles and assembles transcription and splicing components into
complexes through its amino and carboxyl regions. Mol Cell Biol 26:
25. Kameoka S, Duque P, Konarska MM (2004) p54(nrb) associates with the 59
splice site within large transcription/splicing complexes. Embo J 23: 1782–1791.
26. Fong YW, Zhou Q (2001) Stimulatory effect of splicing factors on transcriptional
elongation. Nature 414: 929–933.
27. Yan D, Perriman R, Igel H, Howe KJ, Neville M, et al. (1998) CUS2, a yeast
homolog of human Tat-SF1, rescues function of misfolded U2 through an
unusual RNA recognition motif. Mol Cell Biol 18: 5000–5009.
28. Smith MJ, Kulkarni S, Pawson T (2004) FF domains of CA150 bind
transcription and splicing factors through multiple weak interactions. Mol Cell
Biol 24: 9274–9285.
29. Berger J, Hauber J, Hauber R, Geiger R, Cullen BR (1988) Secreted placental
alkaline phosphatase: a powerful new quantitative indicator of gene expression in
eukaryotic cells. Gene 66: 1–10.
30. Malim MH, Hauber J, Fenrick R, Cullen BR (1988) Immunodeficiency virus rev
trans-activator modulates the expression of the viral regulatory genes. Nature
31. Tokunaga K, Greenberg ML, Morse MA, Cumming RI, Lyerly HK, et al.
(2001) Molecular basis for cell tropism of CXCR4-dependent human
immunodeficiency virus type 1 isolates. J Virol 75: 6776–6785.
32. Li M, Gao F, Mascola JR, Stamatatos L, Polonis VR, et al. (2005) Human
immunodeficiency virus type 1 env clones from acute and early subtype B
infections for standardized assessments of vaccine-elicited neutralizing antibod-
ies. J Virol 79: 10108–10125.
33. Fouchier RA, Meyer BE, Simon JH, Fischer U, Malim MH (1997) HIV-1
infection of non-dividing cells: evidence that the amino-terminal basic region of
the viral matrix protein is important for Gag processing but not for post-entry
nuclear import. Embo J 16: 4531–4539.
34. Landau NR, Page KA, Littman DR (1991) Pseudotyping with human T-cell
leukemia virus type I broadens the human immunodeficiency virus host range.
J Virol 65: 162–169.
35. Bonano VI, Oltean S, Garcia-Blanco MA (2007) A protocol for imaging
alternative splicing regulation in vivo using fluorescence reporters in transgenic
mice. Nat Protoc 2: 2166–2181.
36. Derdeyn CA, Decker JM, Sfakianos JN, Wu X, O’Brien WA, et al. (2000)
Sensitivity of human immunodeficiency virus type 1 to the fusion inhibitor T-20
is modulated by coreceptor specificity defined by the V3 loop of gp120. J Virol
37. Brass AL, Dykxhoorn DM, Benita Y, Yan N, Engelman A, et al. (2008)
Identification of host proteins required for HIV infection through a functional
genomic screen. Science 319: 921–926.
38. Keele BF, Giorgi EE, Salazar-Gonzalez J, Decker JM, Pham KT, et al. (2008)
Identification and characterization of transmitted and early founder virus
envelopes in primary HIV-1 infection. PNAS 105: 7552–7557.
39. Lu J, Sista P, Giguel F, Greenberg M, Kuritzkes DR (2004) Relative replicative
fitness of human immunodeficiency virus type 1 mutants resistant to enfuvirtide
(T-20). J Virol 78: 4628–4637.
40. Willey RL, Smith DH, Lasky LA, Theodore TS, Earl PL, et al. (1988) In vitro
mutagenesis identifies a region within the envelope gene of the human
immunodeficiency virus that is critical for infectivity. J Virol 62: 139–147.
41. Chen CH, Weinhold KJ, Bartlett JA, Bolognesi DP, Greenberg ML (1993)
CD8+ T lymphocyte-mediated inhibition of HIV-1 long terminal repeat
transcription: a novel antiviral mechanism. AIDS Res Hum Retroviruses 9:
42. Neumann JR, Morency CA, Russian KO (1987) A novel rapid assay for
chloramphenicol acetyl transferase gene expression. BioTechniques 3: 444–447.
43. Lin J, Cullen BR (2007) Analysis of the interaction of primate retroviruses with
the human RNA interference machinery. J Virol 81: 12218–12226.
44. Overman RG, Llorens AL, Greenberg ML, Garcia-Blanco MA, Tomaras GD
(2007) Initiation of Human Immunodeficiency Virus Type 1 (HIV-1)
Transcription is Inhibited by Noncytolytic CD8+ Suppression. The Open
Virology Journal 1: 1–7.
45. Arrigo SJ, Weitsman S, Rosenblatt JD, Chen IS (1989) Analysis of rev gene
function on human immunodeficiency virus type 1 replication in lymphoid cells
by using a quantitative polymerase chain reaction method. J Virol 63:
46. Wei X, Decker JM, Liu H, Zhang Z, Arani RB, et al. (2002) Emergence of
resistant human immunodeficiency virus type 1 in patients receiving fusion
inhibitor (T-20) monotherapy. Antimicrob Agents Chemother 46: 1896–1905.
47. Exline CM, Feng Z, Stoltzfus CM (2008) Negative and positive mRNA splicing
elements act competitively to regulate human immunodeficiency virus type 1 vif
gene expression. J Virol 82: 3921–3931.
48. Bres V, Gomes N, Pickle L, Jones KA (2005) A human splicing factor, SKIP,
associates with P-TEFb and enhances transcription elongation by HIV-1 Tat.
Genes Dev 19: 1211–1226.
49. Zhou M, Deng L, Lacoste V, Park HU, Pumfery A, et al. (2004) Coordination of
transcription factor phosphorylation and histone methylation by the P-TEFb
kinase during human immunodeficiency virus type 1 transcription. J Virol 78:
50. Ping YH, Chu CY, Cao H, Jacque JM, Stevenson M, et al. (2004) Modulating
HIV-1 replication by RNA interference directed against human transcription
elongation factor SPT5. Retrovirology 1: 46.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org12May 2009 | Volume 4 | Issue 5 | e5710
51. Sonza S, Mutimer HP, O’Brien K, Ellery P, Howard JL, et al. (2002) Selectively
reduced tat mRNA heralds the decline in productive human immunodeficiency
virus type 1 infection in monocyte-derived macrophages. J Virol 76:
52. Purcell DF, Martin MA (1993) Alternative splicing of human immunodeficiency
virus type 1 mRNA modulates viral protein expression, replication, and
infectivity. J Virol 67: 6365–6378.
53. Gorry PR, Howard JL, Churchill MJ, Anderson JL, Cunningham A, et al. (1999)
Diminished production of human immunodeficiency virus type 1 in astrocytes
results from inefficient translation of gag, env, and nef mRNAs despite efficient
expression of Tat and Rev. J Virol 73: 352–361.
54. Konig R, Zhou Y, Elleder D, Diamond TL, Bonamy GM, et al. (2008) Global
analysis of host-pathogen interactions that regulate early-stage HIV-1 replica-
tion. Cell 135: 49–60.
55. Zhou H, Xu M, Huang Q, Gates AT, Zhang XD, et al. (2008) Genome-scale
RNAi screen for host factors required for HIV replication. Cell Host Microbe 4:
56. Naito T, Kiyasu Y, Sugiyama K, Kimura A, Nakano R, et al. (2007) An
influenza virus replicon system in yeast identified Tat-SF1 as a stimulatory host
factor for viral RNA synthesis. Proc Natl Acad Sci U S A 104: 18235–18240.
57. Misse D, Gajardo J, Oblet C, Religa A, Riquet N, et al. (2005) Soluble HIV-1
gp120 enhances HIV-1 replication in non-dividing CD4+ T cells, mediated via
cell signaling and Tat cofactor overexpression. Aids 19: 897–905.
58. Simmons A, Aluvihare V, McMichael A (2001) Nef triggers a transcriptional
program in T cells imitating single-signal T cell activation and inducing HIV
virulence mediators. Immunity 14: 763–777.
Tat-SF1 Affects HIV RNA Ratios
PLoS ONE | www.plosone.org13 May 2009 | Volume 4 | Issue 5 | e5710