Comparative functional analysis of Jembrana disease virus Tat protein on lentivirus long terminal repeat promoters: evidence for flexibility at its N-terminus.

Abstract

Jembrana disease virus (JDV) encodes a potent regulatory protein Tat that strongly stimulates viral expression by transactivating the long terminal repeat (LTR) promoter. JDV Tat (jTat) promotes the transcription from its own LTR as well as non-cognate LTRs, by recruiting host transcription factors and facilitating transcriptional elongation. Here, we compared the sequence requirements of jTat for transactivation of JDV, bovine immunodeficiency virus (BIV) and human immunodeficiency virus (HIV) LTRs.
In this study, we identified the minimal protein sequence for LTR activation using jTat truncation mutants. We found that jTat N-terminal residues were indispensable for transactivating the HIV LTR. In contrast, transactivation of BIV and JDV LTRs depended largely on an arginine-rich motif and some flanking residues. Competitive inhibition assay and knockdown analysis showed that P-TEFb was required for jTat-mediated LTR transactivation, and a mammalian two-hybrid assay revealed the robust interaction of jTat with cyclin T1. In addition, HIV LTR transactivation was largely affected by fusion protein at the jTat N-terminus despite the fact that the cyclin T1-binding affinity was not altered. Furthermore, the jTat N-terminal sequence enabled HIV Tat to transactivate BIV and JDV LTRs, suggesting the flexibility at the jTat N-terminus.
This study showed the distinct sequence requirements of jTat for HIV, BIV and JDV LTR activation. Residues responsible for interaction with cyclin T1 and transactivation response element are the key determinants for transactivation of its cognate LTR. N-terminal residues in jTat may compensate for transactivation of the HIV LTR, based on the flexibility.

Full text

Page 1

BioMed Central
Page 1 of 14
(page number not for citation purposes)
Virology Journal
Open Access
Research
Comparative functional analysis of Jembrana disease virus Tat
protein on lentivirus long terminal repeat promoters: evidence for
flexibility at its N-terminus
Yang Su1,2, Gang Deng1,2, Yuanming Gai1,2, Yue Li1,2, Yang Gao1,2,
Jiansen Du1,2, Yunqi Geng1,2, Qimin Chen1,2 and Wentao Qiao*1,2
Address: 1Key Laboratory of Molecular Microbiology and Biotechnology (Ministry of Education), College of Life Sciences, Nankai University,
Tianjin 300071, China and 2Key Laboratory of Microbial Functional Genomics (Tianjin), College of Life Sciences, Nankai University, Tianjin
300071, China
Email: Yang Su - tenlions@mail.nankai.edu.cn; Gang Deng - foolduck@mail.nankai.edu.cn; Yuanming Gai - sangny1234@yahoo.com.cn;
Yue Li - liyue_pg@yahoo.com.cn; Yang Gao - gaoyang875@mail.nankai.edu.cn; Jiansen Du - djs916@mail.nankai.edu.cn;
Yunqi Geng - gengyq@nankai.edu.cn; Qimin Chen - qmchen@nankai.edu.cn; Wentao Qiao* - wentaoqiao@nankai.edu.cn
* Corresponding author
Abstract
Background: Jembrana disease virus (JDV) encodes a potent regulatory protein Tat that strongly
stimulates viral expression by transactivating the long terminal repeat (LTR) promoter. JDV Tat
(jTat) promotes the transcription from its own LTR as well as non-cognate LTRs, by recruiting host
transcription factors and facilitating transcriptional elongation. Here, we compared the sequence
requirements of jTat for transactivation of JDV, bovine immunodeficiency virus (BIV) and human
immunodeficiency virus (HIV) LTRs.
Results: In this study, we identified the minimal protein sequence for LTR activation using jTat
truncation mutants. We found that jTat N-terminal residues were indispensable for transactivating
the HIV LTR. In contrast, transactivation of BIV and JDV LTRs depended largely on an arginine-rich
motif and some flanking residues. Competitive inhibition assay and knockdown analysis showed
that P-TEFb was required for jTat-mediated LTR transactivation, and a mammalian two-hybrid
assay revealed the robust interaction of jTat with cyclin T1. In addition, HIV LTR transactivation
was largely affected by fusion protein at the jTat N-terminus despite the fact that the cyclin T1-
binding affinity was not altered. Furthermore, the jTat N-terminal sequence enabled HIV Tat to
transactivate BIV and JDV LTRs, suggesting the flexibility at the jTat N-terminus.
Conclusion: This study showed the distinct sequence requirements of jTat for HIV, BIV and JDV
LTR activation. Residues responsible for interaction with cyclin T1 and transactivation response
element are the key determinants for transactivation of its cognate LTR. N-terminal residues in jTat
may compensate for transactivation of the HIV LTR, based on the flexibility.
Background
Jembrana disease virus (JDV) is a bovine lentivirus that in
Bali cattle (Bos javanicus) often causes an acute disease
endemic in parts of Indonesia. After 5-12 days incubation,
infected cattle suffer clinical signs of fever and lymphade-
nopathy, with high viral titres of 108 infectious units per
Published: 28 October 2009
Virology Journal 2009, 6:179doi:10.1186/1743-422X-6-179
Received: 20 September 2009
Accepted: 28 October 2009
This article is available from: http://www.virologyj.com/content/6/1/179
© 2009 Su et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 2

Virology Journal 2009, 6:179http://www.virologyj.com/content/6/1/179
Page 2 of 14
(page number not for citation purposes)
milliliter in plasma [1-3]. Nucleotide sequence analysis of
the JDV genome indicates that JDV is highly related to BIV
and HIV [4-6]. Generally, lentiviruses are associated with
chronic and progressive diseases involving a long period
of latent infection. Despite the high genomic similarity to
other lentiviruses, JDV infection shows an acute clinical
and pathological syndrome with a 20% fatality rate,
which is quite different from other milder lentiviruses
[6,7]. The most evident pathology of JDV infection is an
intense lymphoproliferative disorder affecting most organ
systems, including the enlarged lymph nodes and spleen,
as well as the proliferative lymphoid infiltrate in liver and
kidneys [3,8]. Recently, a tissue-derived vaccine has been
developed [9], and is currently used to control the spread
of Jembrana disease in Bali cattle. Vaccinated cattle were
found to have 96% reduction in viral load, indicating that
the vaccination may ameliorate the disease. Nevertheless,
little is known to date about the primary cause of acute
JDV pathogenesis.
A typical lentivirus genome is comprised of flanking long
terminal repeats (LTRs) and three major structural genes,
gag, pol, and env, as well as several accessory genes repre-
sented by small open reading frames (ORFs) in the central
and C-terminal regions [10]. Several lines of evidence
from the well-studied HIV-1 show that most accessory
genes are involved in viral replication and pathogenesis
[11-14]. Among the products of these accessory genes, the
transactivator of transcription (Tat) is the most important
for viral multiplication [15,16]. JDV Tat (jTat) also largely
contributes to rapid viral replication and establishment of
acute Jembrana disease [17-19]. JTat is encoded by two
exons derived from separate ORFs in the central RNA
genome with two potential splice donor (SD) sites at posi-
tions 5299 and 5335 and six potential splice acceptor (SA)
sites between nucleotides 4939 and 5007 [4,5,19].
Although the role of exon 2 is still unknown [18], jTat
exon 1 is a potent transactivator for viral gene expression
[18,20] and has been shown to modulate cellular gene
expression and induce apoptosis, based on our previous
studies [21,22]. Interestingly, jTat strongly transactivates
not only its own LTR but also the related BIV LTR and even
the primate HIV LTR, indicating that jTat has pleiotropic
functions [18,23]. Therefore, we assume that bovine len-
tiviruses have a close evolutionary relationship with pri-
mate lentiviruses and their Tat proteins share the
common roles in the viral life cycle, especially for LTR
activation.
Due to the unusual properties, several insights into jTat
function gradually emerged [24-27]. Much attention has
been paid to the jTat C-terminal RNA-binding domain
(RBD), particularly to the arginine-rich motif (ARM),
which confers capability of binding diverse species of
transactivation response element (TAR) [20,28-31]. An
earlier study demonstrates the chameleon-like property of
this 97 amino acid protein when binding to different TAR
targets [20]. Several studies report that the interaction of
jTat with the HIV TAR bulge is mediated by a single
arginine at position 70, which is a conserved residue
Arg52 in HIV Tat (hTat) [20,31-33]. In marked contrast,
the jTat RBD adopts the -hairpin conformation when
binding to BIV and JDV TARs [28,31,34]. Three conserved
arginines Arg70, Arg73 and Arg77 that are also present in
BIV Tat (bTat), and perhaps some other residues help sta-
bilize the -hairpin conformation. To achieve high RNA-
binding affinity, jTat folds to the correlative structures in
order to recognize the species-specific RNA architectures.
Structural analysis of the jTat/TAR complex has further
demonstrated that stabilization of the complex is medi-
ated by intermolecular RNA/protein contacts [31]. Taken
together, jTat RBD undergoes significant conformational
change when binding to different RNA targets, accounting
for its pleiotropic activities upon diverse LTR promoters.
The activation domain (AD) of Tat governs recruitment of
cellular transcription factors that antagonize the TAR-
induced repression of transcriptional
Recently, it has become clear that a cofactor of hTat is cyc-
lin T1 (CycT1), a component of the positive transcription
elongation factor b (P-TEFb) [35-38]. Tat/CycT1 het-
erodimer binds to TAR, allowing the cyclin-dependent
kinase 9 (CDK9) to modify the initiated RNA polymerase
II (pol II) transcription complex to a more elongation-
competent state, by phosphorylating the pol II C-terminal
domain (CTD) [39,40]. The machinery suggests that for-
mation of Tat/CycT1 is highly required for transactivation.
In addition, LTR transactivation requires that Tat/CycT1
heterodimer adopts a cooperative conformation to facili-
tate formation of Tat/CycT1/TAR ternary complex. For
instance, murine cells are non-permissive cells for hTat to
transactivate the HIV LTR. Although hTat is able to recruit
murine CycT1 (mCycT1), the resultant complex shows
weak affinity when binding to HIV TAR [41].
elongation.
Unlike well-studied hTat, little is known about the iden-
tity and potential role of the jTat cofactor. The functional
domains in jTat by which transactivation of the cognate
and non-cognate LTRs is warranted remain unclear. In this
study, the minimal protein sequences (MPS) of jTat for
HIV, BIV and JDV LTR activation are investigated. We find
that HIV LTR transactivation by jTat requires the integrity
of jTat N-terminal domain (NTD), while activation of BIV
and JDV LTRs requires the ARM and the flanking residues.
Meanwhile, we demonstrate that CycT1 and CDK9 are
obligatory factors for JDV LTR activation as shown in com-
petitive inhibition assay and knockdown analysis. In vitro
and in vivo interaction studies reveal the robust interaction
of jTat with human, murine and bovine CycT1s. N-termi-
nal fusion protein largely affects the transactivation activ-
ity of jTat but does not alter the CycT1-binding affinity.
Furthermore, substitution of hTat N-terminal residues

Page 3

Virology Journal 2009, 6:179http://www.virologyj.com/content/6/1/179
Page 3 of 14
(page number not for citation purposes)
with jTat sequence enables hTat to stimulate the non-cog-
nate LTR activities.
Results
Identification of the minimal protein sequence required for
LTR activation
Previous studies demonstrate that jTat is a potent transac-
tivator of its own LTR as well as non-cognate LTRs, such as
HIV and BIV [42]. However, the jTat MPS required for LTR
transactivation is not clear. To identify the MPS, we gener-
ated a series of jTat mutants bearing the N-terminal and
the C-terminal truncations (Figure 1A). The truncation
mutant-stimulated HIV LTR activity in HeLa cells and BIV
and JDV LTR activities in BL12 cells were analyzed. The
preliminary experiments showed that all the LTRs
achieved the maximum activities when cells were trans-
fected with 50 ng pjTat (Figure 1B). The subsequent exper-
iments were performed using the same amount unless
specified.
By contrast with wild-type jTat, the N-terminal trunca-
tions from N20 to N40 stimulated less than 6% of LTR
activatities (Figure 1C). N5, N10 and N15 simulated 73%
to 86% of BIV and JDV LTR activities but less than 23% of
HIV LTR activity. These observations indicate that residues
downstream of N15 are indispensable for transactivation
of all three LTRs. The weak activation of HIV LTR by any
N5, N10 and N15 implies that HIV LTR transactivation
requires the integrity of jTat NTD.
C-terminal truncation mutants from C80 to C93 strongly
transactivated all three LTRs, whereas deletion of His80
(C79) abolished BIV and JDV LTR activities but not the
HIV LTR activity (Figure 1D). Truncation mutants from
C78 to C70 exhibited less than 17% of LTR activity by
wild-type jTat, suggesting that residues upstream of C78
are required for transactivating all three LTRs. Recent stud-
ies have addressed the key residues responsible for HIV
and BIV TAR binding [20,31]. In addition to three
arginines located in the jTat ARM, the His80 identified
here is a novel residue essential for jTat binding to BIV
TAR. Overall, the MPS responsible for HIV LTR transacti-
vation is amino acid residues 1-79 and that for BIV and
JDV LTR transactivation is 15-80.
The jTat RNA-binding domain contains the amino acid
residues outside the jTat ARM
In vitro gel shift assays show that three arginines (Arg70,
73 and 77) in jTat are required for recognition of the BIV
and JDV TARs but Arg70 alone is sufficient for HIV TAR
recognition [20,31]. To further identify the key residues
for TAR binding in vivo, we fuse the putative jTat RBD in
different length to the competent hTat AD (Figure 2A).
The chimeric Tat, HJ69 and HJ70, showed the inability to
transactivate LTRs while HJ66, HJ67 and HJ68 fully sup-
ported LTR activation (Figure 2B), suggesting that the jTat
RBD includes Lys68 but not Arg66 or Arg67. These obser-
vations are consistent with an earlier finding that the
arginines outside the region 70-77 do not enhance TAR-
binding affinity [20]. By contrast with Arg66 and Arg67,
Lys68 is critical for LTR activation, suggesting that Lys68
probably contributes to formation of -hairpin conforma-
tion and/or recognizes the TAR bulge architecture.
To confirm the role of Arg70, Arg73, Arg77 and residues
78-81, we engineered several jTat mutants (Figure 2C).
The single-point mutants bearing R70K mutation fail to
transactivate HIV (Figure 2D), BIV and JDV LTRs (Figure
2F). By contrast, R7377K stimulated the attenuated HIV
LTR activity, (42% of the activity by wild-type jTat). It was
reported that JM1, in which the substitution of KIHY resi-
dues with bTat-derived RIRR was involved, showed weak
TAR-binding affinity [20]. Interestingly, the marked acti-
vation of all three LTRs by JM1 was observed in our exper-
iments (Figure 2D and 2F), suggesting that it is unlikely
that KIHY play an important role in functional TAR bind-
ing in vivo. HJ68 and BJ, two chimeric proteins containing
the jTat RBD (Figure 2E), exhibited stronger transactiva-
tion activity than wild-type hTat or bTat (Figure 2D and
Figure 2F). These results suggest the jTat possesses an
enhanced RBD, facilitating the higher TAR-binding affin-
ity. In addition, the JB chimeric protein simulated BIV and
JDV LTR activities in bovine cells (Figure 2F), indicating
that jTat residues 1-67 include the competent AD. Overall,
the jTat functional RBD is composed of a core region (res-
idues 70-77) as well as the flanking residues Lys68 and
His80.
LTR activation by jTat is enhanced by P-TEFb
In the case of HIV, Tat-mediated transcriptional elonga-
tion requires recruitment of P-TEFb to the LTR promoter
[38,40,43]. In this regard, Tat AD plays a role in recruiting
specific transcription factors. To test if P-TEFb is also
required for jTat-mediated transcription initiation and
elongation, we conducted the competitive inhibition
assays. Overexpression of hTat47 inhibited activation of
the HIV and JDV LTRs by their cognate Tats dose-depend-
ently (Figure 3A). Similar results were observed in the
competitive inhibition assays using overexpressed jTat67.
We reasoned that the excessive hTat AD sequestered P-
TEFb which also participated in the jTat-mediated LTR
transactivation, leading to the consequent inhibition. Our
findings demonstrate that hTat and jTat recruit the com-
mon transcription factors for LTR transactivation.
P-TEFb consists of CycT1 and CDK9, which is also known
as PITALRE, a 43-kDa protein protein kinase that phos-
phorylates the pol II CTD [44,45]. To investigate their role
in LTR activation, we employed the CycT1 and CDK9 anti-
sense plasmids in HeLa cells to deplete endogenous fac-
tors. The effect of rT1 and rCDK9 on the correlative CycT1
and CDK9 expression was monitored by semi-quantita-

Page 4

Virology Journal 2009, 6:179http://www.virologyj.com/content/6/1/179
Page 4 of 14
(page number not for citation purposes)
Activation of LTR reporters by wild-type jTat or truncation mutants
Figure 1
Activation of LTR reporters by wild-type jTat or truncation mutants. (A) Schematic representation of N-terminal
(upper) and C-terminal (lower) truncation mutants. Numbers indicate the residue positions. (B) Cells were co-transfected with
25 ng LTR-luc reporter, 50 ng pCMV-LacZ and the indicated amounts of wild-type jTat expression plasmid (pjTat). Total DNA
was adjusted to equivalent amounts with pcDNA3.1(-). Luciferase activity was measured 48 h post transfection and normalized
to -galactosidase activity. Asterisk marks the optimal pjTat levels, which are used in subsequent assays. (C) HIV LTR activation
in HeLa cells and BIV or JDV LTR activation in BL12 cells by jTat wild-type (WT) or N-terminal truncation mutants. (D) LTR
activation by WT or C-terminal truncation mutants. WT activity on each reporter is set to 100% and percentage activation is
the relative activity of the indicated mutant. Error bars represent the standard deviation from the mean obtained from more
than three independent experiments.

Page 5

Virology Journal 2009, 6:179http://www.virologyj.com/content/6/1/179
Page 5 of 14
(page number not for citation purposes)
jTat RBD residues critical for function
Figure 2
jTat RBD residues critical for function. (A) Schematic representation of the sequence near the RBD of hTat and jTat
(upper) and chimeric proteins bearing hTat AD and jTat RBD (lower). Shaded characters indicate residues evaluated by dele-
tion analysis. hTat AD; hTat residues 1-47. (B) Percentages of LTR activation by different chimeric Tat proteins are shown and
those by HJ66 are set to 100%. (C) Sequence alignment among jTat, bTat and three mutants. Boxed bold characters indicate
residues differing from wild-type jTat. (D) Activities of mutant and wild-type jTat proteins, bTat and hTat on the HIV LTR
reporter in HeLa cells. jTat activity is set to 100%. (E) Schematic representation of chimeric proteins bearing either bTat AD
and jTat RBD (BJ) or jTat AD and bTat RBD (JB). (F) Activities of mutant jTat proteins, chimeric proteins, and wild-type bTat
and jTat proteins on BIV LTR and JDV LTR reporters in BL12 cells. jTat activity is set to 100%. The dotted line in (A) and (E)
indicates where the proteins are fused.