Structural and functional analysis of the related

Am J Physiol Heart Circ Physiol 306: H233–H242, 2014.
First published November 8, 2013; doi:10.1152/ajpheart.00069.2013.
Structural and functional analysis of the related transcriptional enhancer
factor-1 and NF-␬B interaction
Jieliang Ma,1,3 Li Zhang,1 Aaron R. Tipton,2 Jiaping Wu,3 Angela F. Messmer-Blust,3
Melissa J. Philbrick,3 Yajuan Qi,4,5 Song-Tao Liu,2 Hongsheng Liu,1,3 Jian Li,3 and Shaodong Guo4,5
1
College of Life Science, Liaoning University, Shenyang, China; 2Department of Biological Sciences, The University of
Toledo, Toledo, Ohio; 3Cardiovascular Institute, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston,
Massachusetts; 4Division of Molecular Cardiology, Cardiovascular Research Institute, College of Medicine, Texas A & M
University Health Science Center; 5Scott & White, Central Texas Veterans Health Care System, Temple, Texas
Submitted 31 January 2013; accepted in final form 1 October 2013
related transcriptional enhancer factor-1; nuclear factor ␬B; hypoxiainducible factor-1␣; tumor necrosis factor-␣
RELATED TRANSCRIPTIONAL ENHANCER FACTOR-1
(RTEF-1), also
known as transcriptional enhancer activator (TEA) domain
family member 4 (TEAD4), is a member of the transcriptional
enhancer factor (TEF) family (21). The TEF-1 gene family,
comprised of TEF-1, RTEF-1, divergent TEF-1, and epidermal
growth factor receptor-specific TF, is responsible for the regulation of expression of multiple genes in cardiac, skeletal,
smooth muscle, and endothelial cells (15, 32, 33, 38, 41). All
four members of the TEF-1 family share an evolutionarily
Address for reprint requests and other correspondence: S. Guo, Div. of
Molecular Cardiology, Dept. of Medicine, College of Medicine, Texas A&M
Univ. Health Science Center; Scott & White; Central Texas Veterans Health
Care System, 1901 South 1st St., Bldg. 205, Temple, TX 76504 (e-mail:
[email protected]).
http://www.ajpheart.org
conserved DNA binding domain, the so-called TEA or ATTS
domain that specifically recognizes MCAT element CATN(T/
C)(T/C) found in the promoter region of many genes in cardiac
muscles (14, 37). Recently, several cofactors for TEF-1 family
members have been identified, including p160 family of nuclear receptor coactivators (steroid receptor coactivator 1, transcriptional intermediary factor 2, and Ras-related C3 botulinum
toxin substrate 3) (3), Src/Yes-associated protein (YAP)65 (40), the
transcriptional coactivator with postsynaptic density 95/Drosophila disk large/zonula occludens-1-binding motif (29), and
the vestigial-like proteins 2 and 4 (Vgl-2, -4) (5, 28). By
interacting with other transcription factors, such as serum
response factor (SRF), myocyte enhance factor 1 (MEF2), and
Max (2, 18, 19, 28), TEF-1 family members increase the
activity of cardiac troponin T in myocardium and ␣-actin gene
expression in skeletal muscle. Thus RTEF-1 transcriptional
networks for regulation of the target gene expression under
different cellular context and environmental factors are complex and incompletely understood.
Hypoxia, a common pathophysiological phenomenon with a
profound impact on the cellular responses and properties during many cardiovascular disease processes, induces gene expression of RTEF-1 (22). In response to hypoxia, cells produce
hypoxia-inducible factor (HIF)-1␣, a transcription factor that is
usually rapidly degraded via ubiquitination in normoxic conditions (hypoxic conditions promote HIF-1␣ protein stability)
(12). Moreover, HIF-1␣ also can be activated under normoxic
conditions, initiating inflammatory responses; in turn, inflamed
lesions often become severely hypoxic (12, 31). NF-␬B, a key
mediator of inflammatory responses, links innate immunity to
the hypoxic responses through transcriptional regulation of
HIF-1␣ (23). We recently demonstrated that RTEF-1 increases
HIF-1␣ gene expression that promotes angiogenesis (36) and
accelerates myocardial recovery from ischemia (22). We also
demonstrated that RTEF-1 in endothelial cells modulates blood
glucose levels in vivo (30). However, whether RTEF-1 regulates HIF-1␣ and hypoxia-related gene expression involving
inflammatory factors and NF-␬B is unknown.
In this study, we report that RTEF-1 directly interacts with
p65 subunit of NF-␬B. By building up a molecular model of
p65-NF-␬B binding domain with the TEF-1 family members,
we identified a conservative surface-exposed area on both
RTEF-1 and TEF-1 protein crucial for p65-NF-␬B binding.
Further analyses indicated that p65-NF-␬B binding to RTEF-1
resulted in a competitive and negative effect on RTEF-1mediated HIF-1␣ transcriptional regulation. These studies reveal an important combinatorial interaction between these two
0363-6135/14 Copyright © 2014 the American Physiological Society
H233
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
Ma J, Zhang L, Tipton AR, Wu J, Messmer-Blust AF, Philbrick MJ,
Qi Y, Liu S-T, Liu H, Li J, Guo S. Structural and functional analysis of the
related transcriptional enhancer factor-1 and NF-␬B interaction. Am J Physiol
Heart Circ Physiol 306: H233–H242, 2014. First published November 8,
2013; doi:10.1152/ajpheart.00069.2013.—The related transcriptional enhancer factor-1 (RTEF-1) increases gene transcription of hypoxiainducible factor 1␣ (HIF-1␣) and enhances angiogenesis in endothelium. Both hypoxia and inflammatory factor TNF-␣ regulate gene
expression of HIF-1␣, but how RTEF-1 and TNF-␣ coordinately
regulate HIF-1␣ gene transcription is unclear. Here, we found that
RTEF-1 interacts with p65 subunit of NF-␬B, a primary mediator of
TNF-␣. RTEF-1 increased HIF-1␣ promoter activity, whereas expression of p65 subunit inhibited the stimulatory effect. By contrast,
knockdown of p65 markedly enhanced RTEF-1 stimulation on the
HIF-1␣ promoter activity (7-fold). A physical interaction between
RTEF-1 and p65 was confirmed by coimmunoprecipitation experiments in cells and glutathione S-transferase (GST)-pull-down assays.
A computational analysis of RTEF-1 crystal structures revealed that a
conserved surface of RTEF-1 potentially interacts with p65 via four
amino acid residues located at T347, Y349, R351, and Y352. We
performed site-directed mutagenesis and GST-pull-down assays and
demonstrated that Tyr352 (Y352) in RTEF-1 is a key site for the
formation of RTEF-1 and p65-NF-␬B complex. An alanine mutation
at Y352 of RTEF-1 disrupted the interaction of RTEF-1 with p65.
Moreover, expression of RTEF-1 decreased TNF-␣-induced HIF-1␣
promoter activity, IL-1␤, and IL-6 mRNA levels in cells; however, the
effect of RTEF-1 was largely lost when Y352 was mutated to alanine.
These results indicate that RTEF-1 interacts with p65-NF-␬B through
Y352 and that they antagonize each other for HIF-1␣ transcriptional
activation, suggesting a novel mechanism by which RTEF-1 regulates
gene expression, linking hypoxia to inflammation.
H234
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
classes of transcription factors and enhance our understanding
of the relation between hypoxia and inflammatory pathways.
Targeting this interactive pathway may provide an effective
strategy to disrupt RTEF-1 and NF-␬B interactions in controlling hypoxia and inflammation-related gene expression.
MATERIALS AND METHODS
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
Cell culture and siRNA transfection. Human embryonic kidney
(HEK)293 cells were maintained in DMEM containing 10% fetal bovine
serum. Human microvascular endothelial cells (HMEC-1, Center of Disease
Control) were cultured in the MCDB-131 medium containing 10% fetal
bovine serum, 1% glutamine, 0.1% endothelial growth factor, and 0.1%
hydrocortisone (Invitrogen). HMEC-1 cells were treated with 20 ng/ml
TNF-␣ (Sigma) or 20 ␮M wedelolactone (Sigma) for 5 h in the culture
medium before harvest and measurement. The primers for p65 and p50
siRNA were synthesized by GenePharma, and the sequences are as follows:
siRNA1: 5=- GGACAUAUGAGACCUUCAATT -3=; siRNA2: 5=GCUGAUGUGCACCGACAAG-3=; p50: siRNA1: 5=-AAGGGGCUAUAAUCCUGGACU-3=; siRNA2: 5=-AUAAGUUACUAGAAAUUCCUG-3=. An siRNA targeting none of the genes in human/mouse genome
was used as a negative control.
Gene transfection and reporter gene analysis. HEK293 cells were
seeded in 24-well plates and transfected using polyethylenimine
(Polysciences). Human pHIF-1␣ (⫺557/⫹300)-Luc reporter construct
was kindly provided by Drs. Scot Ebbinghaus and Daekyu Sun,
University of Arizona. The pGL2-TK-HRE-Luc plasmid containing
three copies of the hypoxia response element (HRE) (5=-GTGACTACGTGCTGCCTAG-3=) on the luciferase reporter promoter region
was provided by Giovanni Melill (Developmental Therapeutics Program, National Cancer Institute). Transfections with 300 ng of either
reporter plasmid pHIF-1␣ (⫺557/⫹300)-Luc or pGL2-TK-HRE-Luc,
with or without 300 ng of pXJ40-RTEF-1, pXJ40-TEF-1, pXJ40-p65, or
pXJ40-p50 expression plasmids, were performed in HEK293 cells.
Empty vector pXJ40 was used as a control. Luciferase activity was
determined by the dual luciferase assay system (Promega). For the
endothelial cell transfection, we used the retrovirus-expressing control
green fluorescent protein or RTEF-1 through the vector pBMN, and the
protocol for infection of HMEC-1 cells was previously described (30).
Hypoxia treatment. Hypoxic exposure was performed using a
Molecular Incubator Chamber (Billups-Rothenberg) flushed with 5%
CO2-95% N2. The concentration of oxygen (1–3%) was determined
before and after incubation using an oxygen analyzer (Vascular
Technology). Cells transfected with the pGL2-TK-HRE-Luc reporter
construct were incubated for 5 h in a hypoxic chamber before harvest
and measurement of luciferase activity.
Immunoblot analysis and immunoprecipitation. Transfected cells
were washed with cold PBS twice and lysed in cold RIPA buffer
(Boston Bio-Products), which contains 50 mM Tris·HCl, pH 7.4, 150
mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and
protease inhibitor cocktail (Roche). Protein concentrations were determined with the DC Protein Standard Assay (Bio-Rad). Samples
were subjected to SDS-PAGE, transferred to nitrocellulose membranes (Whatman), and subsequently blocked in TBS-Tween 20
containing 8% non-fat milk for 1 h. The membranes were incubated
with the indicated primary antibodies: polyclonal anti-RTEF-1 antibody (1:10,000 dilution, Genemed Synthesis), monoclonal anti-vinculin
antibody (1:100,000, Sigma), polyclonal anti-p50 antibody (1:1,000),
polyclonal anti-p65 antibody (1:1,000, Santa Cruz Biotechnology), and
monoclonal anti-glutathione S-transferase (GST) (1:10,000) followed by
incubation with horseradish peroxidase-conjugated secondary antibodies
anti-mouse IgG (1:3,000), anti-goat IgG (1:2,000, Calbiochem), or antimouse IgG (1:5,000 dilution, Vector Laboratories). Blots were developed
using the chemiluminescence detection system (Thermo Fisher). Densitometric analysis was performed using the NIH software Image J. For
immunoprecipitation, nuclear extracts (50 ␮g protein) were prepared
from HEK293 cells overexpressing RTEF-1 or NF-␬B. After hypoxic
stimulation for 5 h, 500 ␮g of the cellular protein extracts were
incubated with 0.5 ␮g normal goat serum, 0.8 ␮g anti-p65, or 0.8 ␮g
anti-p50 antibodies at 4°C for 3 h. Alternatively, 50 ␮g of nuclear
extracts was incubated with 1 ␮g anti-HA antibody or 0.8 ␮g
anti-RTEF-1 antibody at 4°C for 3 h. Protein A/G-agarose was then
added and incubated for 1.5 h at 4°C. Immune complexes were
washed with NP-40 lysis buffer five times, resolved by SDS-PAGE,
and then analyzed by immunoblot using antibodies against RTEF-1,
p50, or p65.
Gene cloning and site-directed mutagenesis. TEF-1 cDNA and
RTEF-1 cDNA were generated by PCR amplification using oligonucleotides as follows and cloned into pENTR/D-TOPO (Invitrogen). The
PCR primer sequences used are as follows: TEF-1: 5=-CACCGAGAGGATGAGTGACTCGG-3= and 5=-TCAGTCCTTTACAAGCCTGTAG-3=; RTEF-1: 5=-CACCTTGGAGGGCACGGCCGG-3= and 5=CTCTCATTCTTTCACCAGCCTGTAGATGTGGTGC-3=.
RTEF-1 point mutations (T347A, Y349A, R351A, and Y352A) were
generated using the QuickChange Site-Directed Mutagenesis Kit (Stratagene). The primers used are as follows: T347A: 5=-GGTGGAGAAAGTTGAGGCAGAGTATGCTCGCTAT and 5=-ATAGCGAGCATACTCTGCCTCAACTTTCTCCACCA; Y349A: 5=GGAGAAAGTTGAGACAGAGGCTGCTCGCTATGAGAATGGA
and 5=-TCCATTCTCATAGCGAGCAGCCTCTGTCTCAACTTTCTCC; R351A: 5=-AGTTGAGACAGAGTATGCTGCCTATGAGAATGGACACTAC and 5=-GTAGTGTCCATTCTCATAGGCAGCATACTCTGTCTCAACT; Y352A: 5=-TGAGACAGAGTATGCTCGCGCTGAGAATGGACACTACTCT and 5=-AGAGTAGTGTCCATTCTCAGCGCGAGCATACTCTGTCTCA. The underlined sequences denote the
sites that were mutated, and all constructs were carried out DNA sequencing
from Genewiz.
Recombinant protein expression and purification. The p65 gene
was cloned into pGEX-4T expression vector (Pharmacia) and induced
by 0.1 mM isopropyl-␤-D-thiogalactopyranoside for expression in
Escherichia coli BL21 at 28°C, in which p65-GST recombinant
protein was purified using glutathione-agarose columns (Pharmacia).
The concentration of recombinant proteins was assessed by comparing the target band to BSA standards on the Coomassie Blue-stained
gels.
In vitro GST-pull-down binding assay. The troponin T-coupled
rabbit reticulocyte lysate system (Promega) was used to translate
wild-type and mutant RTEF-1 in vitro. The integrity of translated
proteins was analyzed by immunoblotting. For binding assays, in vitro
translated wild-type or mutant RTEF-1 was incubated with either 700
␮g of GST alone or GST-tagged p65 bound to glutathione-agarose
beads in 1⫻ protein interaction buffer (PIB) [40 mM Tris·HCl pH 7.5,
150 mM NaCl, 5 mM MgCl2, 0.5% NP-40, 5% glycerol, 1 mM
tris(2-carboxyethyl) phosphine, 200 ␮g/ml BSA] for 4 h at 4°C with
continuous shaking. Proteins on the bead were washed with 1⫻ PIB
three times, resolved by SDS-PAGE, and analyzed by immunoblotting.
Quantitative real-time PCR analysis. Total RNA was extracted from
cells overexpressing RTEF-1 using TRIzol Reagent (Invitrogen) according to
the manufacturer’s instruction. A total of 2 ␮g of RNA was reversetranscribed using a high-capacity cDNA reverse transcription kit (Applied
Biosystems) with random primers according to the manufacturer’s protocol.
Quantitative real-time PCR amplification was performed using the SYBR
Green Master Mix kit (Applied Biosystems). The PCR primer sequences
used are as follows: IL-1␤: 5=-TCCCCAGCCCTTTTGTTGA-3= and 5=TAAGAACCAAATGTGGCCGTG-3=; IL-6: 5=-GGTACATCCTCGACGGCATCT-3= and 5=-GTGCCTCTTTGCTGCTTTCAC-3=; and tubulin: 5=-CACCCGTCTTCAGGGCTTCTTGGTTT-3= and 5=-CATTTCACCATCTGGTTGGCTGGCTC-3=. Real-time quantitative PCR was
performed in ABI-7000 System (Applied Biosystems). For each reaction, the
relative expression level was calculated as 2⫺⌬Ct, where ⌬Ct is the difference
between the Ct for the gene of interest and the Ct for the housekeeping gene
tubulin.
H235
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
Molecular docking. Two widely accepted protein-protein docking
programs, ZDOCK (7) and RosettaDock (16), were used to predict and
assess the interactions between RTEF-1, TEF-1, and NF-␬B (p50, p65,
and NF-␬B heterodimer, respectively). ZDOCK uses fast Fourier transform to globally search rigid-body transformations of two proteins and is
amenable to large-scale decoy generation (7). The RosettaDock was
run in full atom mode, allowing spin around the axis to connect the
two proteins through the standard Monte Carlo movements. First, the
initial stage-docking program ZDOCK was used as a rigid-body
globally searching method to generate 2,000 poses for each docking.
The top 100 poses were chosen by their score, which is determined by
the scoring function of ZDOCK, which consists of a van der Waals
term and an electrostatic term. Second, the 100 structures were
clustered using the Rosetta Cluster application with a cluster radius
cutoff of 4 Å, and then a representative model is subsequently selected
Sequence alignments. The protein sequences of TEAD1/TEF-1(NP_
068780.2), TEAD2/TEF-4(NP_003589.1), TEAD3/TEF-5(NP_003205.2),
and TEAD4/RTEF-1(NP_003204.2) were obtained from the National
Center for Biotechnology Information (NCBI). These sequences were
subjected to multiple sequence alignment using Clustal W2.0.12
under default parameters (8).
Construction of protein structural modeling. Available mouse
RTEF-1 crystal structure (PDB code: 3JUA) with 96% amino acid
sequence identity to human RTEF-1 was analyzed by PositionSpecific Iterated BLAST search (1). Homology model of human
RTEF-1 was generated using MODELLER 9v8 based on a template
structure (13). The final refined model was assessed by PROCHECK
(26), ERRAT (9), and VERIFY-3D (11) to show the reliability. All
the structural models were visualized by PyMol Molecular Graphics
System (10).
HIF-1α promoter
300
*
*
*
200
100
+
+
+
+
+
RTEF-1
HIF-1α promoter activity
C
7.0
+
+
100
0
+
+
+ +
TEF-1
p<0.01
*
6.0
5.0
4.0
*
2.0
60
p65
Vinculin
p50
Vinculin
800
0
p50
p65
RTEF-1
*
*
600
400
200
0
p65-siRNA c
RTEF-1 +
1
2
+
+
*
20
Relative Luciferase
Activity
siRNA
1 2
Relative Luciferase
Activity
Control siRNA
Control
E
HRE heterologous
promoter
*
0
IKK
inhibitor
+
+
40
1.0
Vehicle
+
+
p65
D
Control
RTEF-1
3.0
*
200
p50
+
+
*
300
pXJ40
+
p50
p65
HIF-1α promoter
Relative Luciferase
Activity
0
pXJ40
p<0.05
Relative Luciferase Activity
Relative Luciferase Activity
B
p<0.05
+
+
+
+
+
c
1
2
+
+
+
100
75
Fig. 1. Related transcriptional enhancer factor (RTEF)-1-stimulated hypoxia-inducible
factor (HIF)-1␣ promoter activity is regulated by expression of p65-NF-␬B subunit.
A: human embryonic kidney (HEK)293
cells were cotransfected by pHIF-1␣-Luc
reporter gene together with the plasmid
DNA expressing p65, p50, or RTEF-1, as
indicated, and luciferase activity was measured 24 h posttransfection and control vector pXJ40 used as a control. *P ⬍ 0.05 vs.
control pXJ40. B: HEK293 cells were cotransfected by pHIF-1␣-Luc together with p65, p50,
and TEF-1 expression vectors, and luciferase
activity was measured 24 h posttransfection.
*P ⬍ 0.05 vs. control pXJ40. C: HEK293 cells
were cotransfected by pHIF-1␣-Luc together
with RTEF-1 or control plasmid pXJ40, treated
with 20 ␮M wedelolactone (solid bars, IKK
inhibitor) or with vehicle (0.01% DMSO) for 5
h, and luciferase activity was measured. *P ⬍
0.01 vs. control. D: HEK293 cells were
cotransfected by phospho-hypoxia response
element (HRE)-Luc together with p50, p65,
and RTEF-1 expression vectors, and luciferase activity was measured. *P ⬍ 0.05 vs.
p65. E: HEK293 cells were transfected with
siRNA against p65, p50, or a control, and
expression of endogenous p65, p50, and vinculin was determined by Western blot.
HIF-1␣ promoter activity was measured after
cotransfection with the RTEF-1 expression
vector for 18 h. Results from 3 independent
experiments are shown as means ⫾ SD. *P ⬍
0.05 vs. control siRNA (c).
50
25
0
p50-siRNA
RTEF-1
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
A
H236
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
results indicate that expression of p65-NF-␬B regulates RTEF1-mediated HIF-1␣ promoter activity.
p65-NF-␬B interacts with RTEF-1. To test whether p65NF-␬B interacts with RTEF-1 in vitro, we extracted and
purified recombinant p65-GST from bacteria and incubated
with in vitro translated RTEF-1 for GST-pull-down assay.
Incubation of p65-GST, rather than GST alone, pulled down
RTEF-1, indicating that an interaction between p65 and RTEF-1
occurred (Fig. 2A). To further determine whether the interactions occur in vivo, HEK293 cells were transfected with
plasmid DNA expressing RTEF-1 or control hemagglutininepitope, and immune complexes were isolated from nuclear
extracts and examined for the coprecipitation of p50 and p65.
In agreement with the observation from in vitro studies, p65
was coprecipitated with RTEF-1. Also, p50 was coprecipitated
in the complex purified with antibody against RTEF-1 (Fig.
2B). Thus, the reciprocal immunoprecipitation of p65 and
RTEF-1 from nuclear extracts confirmed a physical interaction
of each other.
Computational structural analysis of RTEF-1 and NF-␬B
interaction surface. We next used two protein-protein docking
programs (7, 16) to predict and assess the interactions between
RTEF-1 and p65-NF-␬B. Analysis of the crystal structures of
RTEF-1 [protein data bank identification number (PDB ID):
RESULTS
RTEF-1-stimulated HIF-1␣ promoter activity is regulated
by expression of p65-NF-␬B subunit. Many of stimuli that
induce HIF-1␣ gene expression are known to activate a number
of transcription factors such as NF-␬B (39), and RTEF-1
increases HIF-1␣ in endothelial cells, enhancing the vascular
endothelial growth factor (VEGF) gene expression and promoting
angiogenesis, as shown in our previous studies (22, 36). We next
examined whether expression of NF-␬B affects the RTEF-1mediated transcriptional regulation of HIF-1␣ in cells. HEK293
cells were transfected with a HIF-1␣ promoter (⫺557/⫹300)coupled luciferase reporter gene alone or cotransfected with
plasmid DNA expressing RTEF-1, p50, or p65 of NF-␬B
subunit. As we previously demonstrated, expression of
RTEF-1 increased the HIF-1␣ promoter activity ⬃2.5-fold
(P ⬍ 0.05), whereas expression of p50 and p65-NF-␬B individually or together had no such effect (Fig. 1A). By contrast,
expression of p65, rather than p50, significantly reduced
RTEF-1 and TEF-induced HIF-1␣ promoter activity (Fig. 1, A
and B), suggesting that NF-␬B negatively regulates the ability
of RTEF-1 to increase the HIF-1␣ promoter activity. Conversely, treatment of an IKK-␤ inhibitor (20 ␮M wedelolactone) in cells increased RTEF-1-stimulated HIF-1␣ promoter
activity threefold (P ⬍ 0.01, Fig. 1C). To further confirm the
role of p65 in regulating RTEF-1-mediated HIF-1 promoter
activity mediated by HRE, we transfected cells with a heterologous reporter gene that contains three consecutive copies of
HRE-coupled luciferase (pHRE-luc), and expression of p65,
rather than p50, reduced HRE-luc activity by 55% (P ⬍ 0.01,
Fig. 1D). To determine the role of loss of p65 in RTEF-1regulated HIF-1 promoter activity, we depleted endogenous
expression of p65 by two different siRNA preparations, each of
which increased RTEF-1-stimulated HIF-1␣ promoter activity
by sevenfold (P ⬍ 0.01, Fig. 1E), whereas depletion of p50NF-␬B had no significant effort on the RTEF-1-stimulated
HIF-1␣ promoter activity (Fig. 1E). Taken together, these
A
Anti-RTEF-1
Anti-p65
85 kDa
85 kDa -
Anti-GST
36 kDa -
B
IP:
HA
RTEF-1
IB: p50
IB: p65
IP:
IgG p50 p65
IB: RTEF-1
Fig. 2. p65-NF-␬B interacts with RTEF-1. A: RTEF-1 was incubated with
beads bound to glutathione S-transferase (GST) or GST-p65. After being
washed, proteins bound to beads were detected by anti-RTEF-1, anti-p65, and
anti-GST antibody. 10% of RTEF-1 input was used as a positive control.
B: HEK293 cells were transfected with the RTEF-1 expression vector or
control plasmid DNA expressing hemagglutinin (HA)-epitope, and immune
complexes were isolated from nuclear extracts and examined by the p50 or p65
antibody (top). Immune complexes were achieved by immunoprecipitation (IP)
against antibody of p50, p65, or normal goat serum (IgG) followed by Western
blot using RTEF-1 antibody (bottom). Results are representative from 3
independent experiments.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
according to the best score from each obvious large cluster. Then each
of the representative models was validated by using the local enrichment method of the RosettaDock program. The search procedure is
repeated from different random starting orientations to create 10,000
decoys, whose energy scores were plotted against the root mean
square distance from the starting input conformation. The models that
achieved a “docking funnel” are considered to be the most robust
complex structures (27).
Protein-protein interface analysis. The structures of the complex
were submitted to protein-protein interface analysis server (PROTORP) (34) to analyze their Interface Accessible Surface Area.
Computational alanine scanning to predict the hotspots in the
complex interface. The RTEF-1-p65 and TEF-1-p65 complex were
submitted to the Robetta Alanine scanning server to predict the free
energy changes brought about by alanine mutation at protein-protein
interfaces by using a simple free-energy function (25). The alanine
scan free-energy changes helped us identify the hotspot residues in the
complex interface. This analysis served as an initial approach to
identify the amino acid residues that are crucial for the interaction
between the TEF family members and NF-␬B.
Statistical analysis. Data presented are from at least three independent experiments. For statistical analysis, if differences were established, the values were compared using the Student’s t-test. Values
were expressed as means ⫾ SD. The results were considered significant at P ⬍ 0.05 or P ⬍ 0.01 as indicated.
H237
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
A
RTEF-1
p50
TEF-1
p65/p50
C
RTEF-1
TEF-1
p50
p65
RTEF-1:p65
RTEF-1:p65/p50
TEF-1:p65
TEF-1:p65/p50
3JUA; Fig. 3A] and p65-NF-␬B subunits (PDB ID: 2RAM),
p50 (PBD ID: 1SVC), or the p50/p65 heterodimer (PDB ID:
1VKX; Fig. 3A) predicted an interaction between RTEF-1 and
p65, rather than p50-NF-␬B (Fig. 3B). Similar results were
found between p65-NF-␬B and TEF-1, another TEF family
member (PDB ID: 3KYS) (Fig. 3C).
A conserved domain of the TEF family interacts with
p65-NF-␬B. Sequence alignment analysis indicated that the
TEF family members have two conserved regions, one is the
TEA domain located at the NH2-terminal end for DNA interaction, and the other is the NF-␬B binding domain (NBD)
located at the COOH terminus (Fig. 4, A and B). Further
analysis of the RTEF-1 and p65-NF-␬B interaction surface
revealed several key amino acid residues that are located in ␤7
and ␤8 of the IgG-like fold of RTEF-1, likely involving the
interaction with p65 (Fig. 4C). Furthermore, the NBD of
RTEF-1 appears to encompass the region between amino acids
231 and 434 (Fig. 4, B and C). Moreover, the NBD of RTEF-1
overlaps with the previously identified YAP binding domain of
TEAD2 (6).
Identification of Y352 of RTEF-1 in NBD domain required
for interaction with NF-␬B. We next used the computational
alanine scanning program to predict the hotspots in the complex
interface of RTEF-1, TEF-1, and p65, as shown in Table 1. With
the criteria that ⌬⌬G ⬎ 1.0 kcal/mol as a first round of selection
factor, two homologous residues in RTEF-1 (Y349, Y352) and
TEF-1 (Y341, F344) were picked out to be of significance. We
next used the second factor for selection, which is the percentile of Interface Accessible Surface Area as shown in Table 2.
The analysis indicated that another two more conserved residues are important within the RTEF-1 NBD located at T347
and R351 and the TEF-1 NBD located at T339 and R343, of
which percentile Interface Accessible Surface Area is greater
than 5%. To determine the role of four residues in RTEF-1
(T347, Y349, R351, and Y352) in interaction with p65-NF-␬B,
we mutated each to alanine, a neutral amino acid, and trans-
lated them in vitro and examined their interaction for p65-GST
in the GST-pull-down assay. Compared with RTEF-1-wildtype (WT), all four alanine mutants of RTEF-1 displayed
reduced binding affinity to p65-GST (Fig. 5, A and B). Moreover, RTEF-1-Y352A mutation markedly reduced the interaction between RTEF-1 and p65 (Fig. 5A), and p65 binding
activity for RTEF-1-Y352A only reached ⬍10% of RTEF1-WT (P ⬍ 0.01, Fig. 5B).
To examine the role of RTEF-1-Y352, which is required for
p65-NF-␬B interaction, in RTEF-1-stimulated HIF-1 promoter
activity in cells, we examined the HIF-1␣ promoter activity in
cells cotransfected with p65, RTEF-1-WT, or the RTEF-1
mutant. Coexpression of p65 together with RTEF-1-T347A,
Y349A, or R351A decreased HIF-1␣ promoter activity by
50%, compared with the effect of RTEF-1-WT (Fig. 5C). By
contrast, expression of RTEF-1-Y352A mutant completely
disrupted the inhibitory effect of p65 expression on the promoter activity (P ⬍ 0.01, Fig. 5C), suggesting that the amino
acid residue at Y352 of RTEF-1 interacts with p65 and mediates the inhibitory effect of p65-NF-␬B.
RTEF-1 decreases TNF-␣-induced HIF-1␣ promoter activity through Y352. p65-NF-␬B is a well-studied transcription
factor mediating inflammatory responses (31) and is rapidly
activated by cytokines, such as TNF-␣ or bacterial products.
Prolonged TNF-␣ exposure is present during inflammatory
conditions, and we investigated whether RTEF-1 regulates
TNF-␣-induced HIF-1␣ promoter activity via Y352. Treatment
of TNF-␣ inhibited RTEF-1 stimulation on the HIF-1␣ promoter activity by 40% (P ⬍ 0.05, Fig. 6A); however, the
inhibitory effect of TNF-␣ was almost completely abolished
when RTEF-1-Y352 was changed to an alanine that blocks p65
interaction (P ⬎ 0.1, Fig. 6A). To further examine whether
endogenous NF-␬B has an effect on RTEF-1-stimulated HIF-1
promoter, treatment of 20 ␮M wedelolactone for IKK inhibition blocking p65 activity increased the ability of RTEF-1 to
promote HIF-1␣ promoter activity threefold (P ⬍ 0.01, Fig.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
B
p65
Fig. 3. Computational structural analysis of
RTEF-1 and NF-␬B interaction surface. Protein
structures and interactions between RTEF-1,
TEF-1, and NF-␬B were predicted by the
ZDOCK and RosettaDock protein-protein docking programs. A: ribbon diagram (top) and
molecular surface (bottom) of RTEF-1, TEF-1,
p50, p65, and the p50-p65 complex. B: diagrams for the interaction between RTEF-1:p65
and RTEF-1:p50/p65 complex. C: diagrams for
the interaction between TEF-1:p65 and TEF-1:
p50/p65 complex.
H238
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
TEA
A
B
RTEF-1
1
TEF-1
1
P65-NFκB 1
105
38
231
TEA
82
15
TEA
NBD
294
NBD
DBD
434
TAD
415
220
C
Fig. 4. A conserved domain of TEF family members interacts with p65-NF-␬B. A: sequence alignment of the TEF family members. The highly conserved domain
in the NH2 terminus is transcriptional enhancer activator (TEA), and the other highly conserved domain in the COOH terminus is NF-␬B binding domain (NBD).
B: schematic domain organization of RTEF-1, TEF-1 and p65. The amino acid residue numbers for different domain boundaries are indicated. TEA,
DNA-binding domain; DBD, p65 DNA-binding domain; TAD, transactivation domain. C: surface area of the RTEF-1/p65-NF-␬B complex was analyzed by the
protein-protein interface analysis server (PROTORP). p65 is shown in green, RTEF-1 in red, and TEF-1 in yellow. Diagrams of the interface between
RTEF-1/p65 (left) and TEF-1/p65 (right) are shown. The putative key amino acid residues in RTEF-1 and TEF-1 are indicated for the interaction with
p65-NF-␬B.
6B); however, the stimulatory effect was abolished by RTEF1-Y352A mutation (Fig. 6B). We further assessed the effect of
RTEF-1-Y352A mutation on TNF-␣-regulated HRE-heterologous reporter gene activity in cells; TNF-␣ treatment suppressed RTEF-1-stimulated HRE-promoter activity by 50%
under either normoxia or hypoxia (P ⬍ 0.05, Fig. 6C), and
TNF-␣ was unable to suppress RTEF-1-Y352A-stimulated
HRE-promoter activities in cells (Fig. 6C).
RTEF-1 blocks TNF-␣-induced genes in endothelial cells via
Y352. Last, we examined whether RTEF-1-Y352 affects TNF␣-induced expression of genes, such as IL-1␤ and IL-6 in
endothelial cells. Either RTEF-1 or RTEF-1-Y352A mutant
was highly expressed in HMEC-1 cells after retrovirus infection of each gene, and both IL-1␤ and IL-6 mRNA levels were
induced three- to fivefold by TNF-␣ (P ⬍ 0.01), of which
induction was decreased by RTEF-1 overexpression in cells
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
NBD
H239
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
Table 1. Computational alanine scanning to predict hotspots
in the complex interface
DDG TEF-1
GLU234
ASP240
PHE313
TYR314
ASN325
0.77
⫺0.1
1.52
0
0.44
PHE305
0.1
VAL345
GLU346
THR347
GLU348
TYR349
ALA350
ARG351
TYR352
GLU353
ASN354
TYR357
ARG360
HIS362
ARG363
SER364
ARG402
0.03
0.11
⫺0.06
⫺0.18
2.31
0.01
0.53
3.48
0.08
0.06
0.84
1.81
0.28
1.2
⫺0.01
⫺0.06
ASN317
THR319
VAL337
GLU338
THR339
GLU340
TYR341
ALA342
ARG343
PHE344
GLU345
ASN346
PHE349
ARG352
ASN354
0.04
0.11
0
⫺0.02
1.12
0.41
2.52
0
0.96
2.33
0.16
⫺0.01
0.99
0.01
0.01
ARG394
0.59
A
RTEF-1 WT
RTEF-1-T347A
RTEF-1-Y349A
RTEF-1-R351A
RTEF-1-Y352A
B
List of amino acids in related transcriptional enhancer factor (RTEF)-1 or
transcriptional enhancer factor (TEF)-1 on the interface. DDG, changed free
energy in response to the amino acids of TEF with a point mutation into Ala.
DDG ⬎1.0 kcal/mol implies a loss of binding affinity upon mutation, and
DDG ⬍ ⫺1.0 implies a reduction of binding affinity upon mutation.
(P ⬍ 0.01, Fig. 7, B and C). However, expression of RTEF1-Y352A mutant completely disrupted the TNF-␣ suppression
on both IL-1␤ and IL-6 gene expression in the HMEC-1 cells
(Fig. 7, B and C).
DISCUSSION
In this study, we present three important findings: 1) RTEF-1
interacts with p65-NF-␬B subunit, and this interaction inhibits
RTEF-1-stimulated HIF1 promoter activity; 2) we identified
that Y352 in RTEF-1 has a key role in mediating the interaction between RTEF-1 and p65; and 3) RTEF-1 inhibits TNF-␣and p65-stimulated HIF-1␣ promoter activity and expression
of genes, such as IL-1␤ and IL6 in cells, whereas mutation of
Y352A in RTEF-1 abolishes the effect of RTEF-1 inhibition.
These results suggest that an interaction of RTEF-1 with
p65-NF-␬B via Y352 has important roles in control of gene
expression, hypoxia, and inflammatory responses.
% Interface Accessible Surface Area
TYR349 14.21
ARG351 6.71
TYR352 14.84
100
80
60
40
20
*
0
WT
T347A Y349A R351A Y352A
400 HIF-1α promoter
300
TEF-1-p65
THR339 7.76
GLU340 8.39
TYR341 22.63
ARG343 14.78
PHE344 19.62
PHE349 5.96
HIS362 5.04
ARG363 9.69
p<0.01
*
*
200
100
0
p65
Table 2. Protein-Protein Interface Analysis
RTEF-1-p65
PHE313 7.95
ASN325 5.59
THR347 5.09
C
Relative Binding Activity (%)
TEF-1
Relative Luciferase Activity
DDG RTEF-1
+
+
+
+
+
Fig. 5. Identification of the key amino acid residues in RTEF-1 for p65-NF-␬B
interaction. A: in vitro synthesized RTEF-1 and its mutant proteins were
incubated with beads bound to recombinant protein GST or GST-p65. After
being washed, proteins bound were immunoblotted with the antibody of
RTEF-1. 10 % of each protein was used as input for comparison. B: binding of
RTEF-1 and its mutant protein toward p65 examined in A were quantified and
normalized against that of RTEF-1-wild-type (WT). Results are from 3
independent experiments, and means ⫾ SD are shown. *P ⬍ 0.01 vs. WT.
C: HEK293 cells were cotransfected by pHIF-1␣-Luc together with control
vector pXJ40 and expression vectors for p65-NF-␬B, RTEF-1-WT, and RTEF
mutant T347A, Y349A, R351A, or Y352A. Luciferase activity was measured
24 h posttransfection, and results from 3 independent experiments are shown
as means ⫾ SD. *P ⬍ 0.05 vs. the control.
ARG394 5.84
List of the amino acids whose surface area on the interface is ⬎5%.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
RTEF-1
H240
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
A
2.0
1.0
0
-
TNFα
-
+
p<0.05
6.0
N.S.
4.0
2.0
A
0
IKK inhibitor
-
+
-
RTEF-1-Y352A
RTEF-1
HRE
Relative Luciferase Activity
C
Hypoxia
B
*
*
*
2.0
*
1.0
+
0
TNFα
-
-
+
RTEF-1
RTEF-1
RTEF-1-Y352A
Vinculin
Vinculin
N.S.
p<0.05
Normoxia
3.0
+
+
-
-
+
+
IL-1β mRNA
HIF-1α promoter
Relative Luciferase Activity
B
+
RTEF-1-Y352A
RTEF-1
p<0.05
4.0
RTEF-1 has important roles in transcriptional regulation of
genes in cardiovascular systems. Our previous reports indicate
that RTEF-1 enhances hypoxia-related angiogenesis through
the transcriptional regulation of VEGF and HIF-1␣ in endothelial cells (21, 22). Activation of gene transcription is a
multistep process that is triggered by transcription factors and
cofactors directing transcriptional initiation; the multiple co-
*
2.0
-
+
Control
RTEF-1-Y352A
Fig. 6. RTEF-1 decreases TNF-␣-induced HIF-1␣ promoter activity through
Y352. A: HEK293 cells were cotransfected by pHIF-1␣-Luc together with
expression vectors for RTEF-1 or RTEF-1-Y352A for 18 h, and luciferase
activity was measured from cells with or without 6-h treatment of TNF-␣ (20
ng/ml, solid bars). B: HEK293 cells were cotransfected by pHIF-1␣-Luc
together with expression vectors for RTEF-1 or RTEF-1-Y352A, and luciferase activity was measured after 5-h treatment of 20 ␮M IKK inhibitor
wedelolactone (solid bars), compared with no treatment (open bars).
C: HEK293 cells were transfected by pHRE-Luc reporter gene with expression
vectors for RTEF-1 or RTEF-1-Y352A for 18 h, and cells were treated with or
without 20 ng/ml TNF-␣ for 6 h under normoxic (open bars) and hypoxic
(solid bars) conditions. Results from 3 independent experiments are shown as
means ⫾ SD. *P ⬍ 0.05 vs. no TNF-␣ treatment; *P ⬍ 0.05 vs. normoxia of
each group. ⫹P ⬍ 0.05 vs. normoxia with no TNF-␣ treatment.
*
*
6.0
0
TNFα
C
IL-6 mRNA
HIF-1α promoter
Relative Luciferase Activity
3.0
-
+
RTEF-1
-
+
RTEF1-Y352A
p<0.01
4.0
*
*
3.0
2.0
*
1.0
+
0
TNFα
-
+
Control
-
+
RTEF-1
-
+
RTEF1-Y352A
Fig. 7. RTEF-1 blocks TNF-␣-induced gene expression in endothelial cells via
Y352. A: human microvascular endothelial cells (HMEC)-1 cells were infected
by the retrovirus-expressing control green fluorescent protein (GFP), RTEF-1,
or RTEF-1-Y352A for 8 h, and the same amount of cellular proteins was
analyzed by Western blot against RTEF-1 antibody and the membrane reprobed with the vinculin antibody. B and C: quantitative real-time-PCR
analysis of IL-1␤ and IL-6 mRNA levels in the HMECs infected with
retrovirus of GFP, RTEF-1, or RTEF-1-Y352A followed by 6-h TNF-␣
treatment. Results from 3 independent experiments are shown as means ⫾ SD.
*P ⬍ 0.05 vs. no TNF-␣ treatment in each group; ⫹P ⬍ 0.05 vs. control with
no TNF-␣ treatment.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
factors for TEF family members include the p160 family of
nuclear receptor coactivators (3), YAP65, Vgl-2, Vgl-4 (40),
and the muscle-specific transcription factors SRF, MEF2, and
Max (17). Here we have identified that the MCAT element, a
consensus binding site for RTEF-1, and the NF-␬B binding site
are localized in a close proximity in the HIF-1␣ promoter.
Based on the structural analysis, Thr374, Tyr349, Arg351, and
Tyr352 in the COOH-terminal region of RTEF-1 might interact
with p65 or other yet unidentified binding partners. Among
those potential residues, Tyr352 was determined as a key
residue in response to the interaction between RTEF-1 and
NF-␬B. This result suggests that the computer-based proteinprotein interaction program is a powerful tool to identify a
potential interaction between RTEF-1 and NF-␬B, which was
supported by in vitro cell-based experiments. Of note is that
tyrosine is an amino acid that can be modified, such as by
protein tyrosine kinases via phosphorylation-based mechanisms, and markedly contributes to conformational changes of
the protein structure and function; thus whether there exists
N.S.
p<0.05
H241
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
Hypoxia
352Y
RTEF-1
TNFα
HIF-1α GGGACTTGCC….CCATTGGA……..+1 DNA
promoter -216
-206
-192 -185
Hypoxia
TNFα
RTEF-1
p65-NF-kB
HIF-1α
IL-1β & IL-6
HRE
Fig. 8. A schematic diagram represents the RTEF-1 and NF-␬B interaction on
the promoter region of HIF-1␣. RTEF-1 promotes HIF-1␣ promoter activity,
which is antagonized by p65-NF-␬B via interacting with the Y352 of RTEF-1.
Arrows indicate stimulation, and T-bars stand for inhibition.
tory responses are required for the maintenance of cells and
tissue function, whereas chronic inflammation is harmful for
the cell and its function during many disease conditions,
RTEF-1 may have pivotal roles in balancing the NF-␬B signaling during inflammation upon ischemic injury.
Our work demonstrates that RTEF-1 and NF-␬B interact and
antagonize each other for the transcriptional regulation of
HIF-1␣ and inflammatory responsive gene expression (Fig. 8),
which has important implications for many disorders related to
hypoxia, including ischemia and/or inflammation. The result
will further enhance our understanding of the transcriptional
regulation relationship between RTEF-1 and NF-␬B in control
of gene expression involving hypoxia and inflammation and
may provide a novel strategy for treating disease states associated with low-oxygen tension and a variety of acute or
chronic inflammatory diseases; the future work to dissect the
different signaling and potential of feedback regulation may be
of great importance for therapeutically targeting hypoxia- and
inflammation-related disorders in vivo.
ACKNOWLEDGMENTS
We thank Dr. Lawrence M. Pfeffer (University of Tennessee) for the p65
construct, Dr. Melillo Giovanni (NIH-NCI) for the HRE construct, and Dr.
Alexandre Stewart (University of Ottawa) for the pXJ40/RTEF-1 construct.
GRANTS
This work was supported in part by NIH Grant HLR01082837 (J. Li) and
the American Heart Association BGIA7880040 (S. Guo). This material is the
result of work supported with resources and the use of facilities at the Central
Texas Veterans Health Care System, Temple, Texas.
DISCLOSURES
This work was prepared while Jian Li was employed at the Cardiovascular
Institute, Beth Israel Deaconess Medical Center, Harvard Medical School. The
opinions expressed in this article are the author’s own and do not reflect the
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
such a regulatory mechanism in RTEF-1-Y352 necessitates
further investigation.
RTEF-1 potently increases the HIF-1␣ promoter activity,
but p65-NF-␬B has limited its capacity to induce the promoter
activity. NF-␬B is a pleiotropic protein complex formed by a
family of subunits including RelA (p65), RelB, cRel, p52, and
p50 (20). Recent studies have demonstrated that stimulation
of HIF-1␣ gene by NF-␬B provides an important, parallel level of
regulation on HIF-1␣ expression (4, 35, 39). In the absence of
p65-NF-␬B, HIF-1␣ gene is not transcribed; hence no stabilization or activity is seen, even after prolonged hypoxia exposures (24, 39). There have been several studies demonstrating
cross-talk between the NF-␬B and HIF signaling pathways, but
the mechanism is unclear. In our studies, we found that
overexpression of p65 or treatment of IKK inhibitor in
HEK293 cells had minimal effects on increasing or reducing
the HIF-1␣ promoter activity, respectively (Fig. 1, A and C)
and endogenous HIF-1␣ was also barely detectable (data not
shown); however, p65 expression and the IKK inhibition had
profound effects on RTEF-1 stimulation on the HIF-1␣ promoter activity. In particular, we demonstrated that NF-␬B is a
cofactor for RTEF-1 in control of both HIF-1 promoter activation and expression of cytokines, such as IL-1␤ and IL-6.
We identified a region in RTEF-1 that directly interacts with
p65-NF-␬B in vitro by both coimmunoprecipitation and GSTpull-down assay. A loss of p65 enhanced the capacity of RTEF-1
to stimulate HIF-1␣ transcriptional regulation, whereas overexpression of p65 decreased TEF-dependent HIF-1␣ promoter activity. Moreover, activation of NF-␬B following TNF-␣ treatments inhibited both RTEF-1-stimulated HIF-1␣ promoter activity and RTEF-1-activated HRE responses under normoxic and
hypoxic conditions, supporting the notion that there is a competitive negative regulation on HIF1 by RTEF-1 and p65.
NF-␬B transcription factor plays key roles in mediating
inflammatory responses, and TNF-␣ is a proinflammatory
cytokine with important roles in regulating inflammation, cell
proliferation, and apoptosis. TNF-␣ activates p65-NF-␬B and
promotes its nuclear translocation in control of gene expression
for cell survival. The endothelial cells from mice lacking
IKK-␤ indicate that basal NF-␬B activity is required for HIF-1
expression (35). p65-NF-␬B binds a distinct element in the
proximal promoter of HIF-1␣ gene and transcriptionally induces HIF-1␣ (39), suggesting that NF-␬B promotes HIF-1␣
gene transcriptional expression. In this study, expression of
RTEF-1 markedly inhibited the ability of TNF-␣ to stimulate
HIF-1 promoter activities and endogenous gene expression of
IL-1␤ and IL-6. However, these effects were largely blocked
by RTEF-1-Y352A mutation, supporting the notion that
RTEF-1 and its regulation via the Y352 affect expression of
inflammatory responsive genes and may potentially modulate
inflammation in health and disease state of cells.
p65 NF-␬B activation is involved in the ischemic injury and
inflammation, but TNF-␣ is not a sole factor triggering the
activation of NF-␬B. It is likely that RTEF-1 can affect the
biological responses of other upstream molecules that control
NF-␬B signaling pathway. Likewise, RTEF-1 inhibits the
TNF-␣-mediated gene expression of IL-1␤ and IL-6, which are
known NF-␬B target genes, and it is likely that expression of
RTEF-1 under hypoxia may also inhibit expression of many
other p65 NF-␬B target genes responsible for limiting ischemia
and associated inflammation. Given that moderate inflamma-
H242
RTEF-1 INTERACTS WITH NF-␬B IN GENE REGULATION
view of the National Institutes of Health, the Department of Health and Human
Services, or the United States Government. No conflicts of interest, financial or
otherwise, are declared by the authors.
AUTHOR CONTRIBUTIONS
Author contributions: J.M., A.F.M.-B., S.-T.L., and J.L. conception and
design of research; J.M. and J.W. performed experiments; J.M., L.Z., A.F.M.-B.,
Y.Q., and J.L. analyzed data; J.M., A.R.T., A.F.M.-B., M.J.P., Y.Q., and J.L.
interpreted results of experiments; J.M., L.Z., J.W., Y.Q., and S.G. prepared
figures; J.M., A.R.T., A.F.M.-B., M.J.P., J.L., and S.G. drafted manuscript;
J.M., L.Z., M.J.P., Y.Q., J.L., and S.G. edited and revised manuscript; J.M.,
Y.Q., S.-T.L., H.L., J.L., and S.G. approved final version of manuscript.
REFERENCES
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00069.2013 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.4 on October 21, 2016
1. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W,
Lipman DJ. Gapped BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids Res 25: 3389 –3402,
1997.
2. Azakie A, Lamont L, Fineman JR, He Y. Divergent transcriptional
enhancer factor-1 regulates the cardiac troponin T promoter. Am J Physiol
Cell Physiol 289: C1522–C1534, 2005.
3. Belandia B, Parker MG. Functional interaction between the p160 coactivator proteins and the transcriptional enhancer factor family of transcription factors. J Biol Chem 275: 30801–30805, 2000.
4. Bonello S, Zahringer C, BelAiba RS, Djordjevic T, Hess J, Michiels C,
Kietzmann T, Gorlach A. Reactive oxygen species activate the HIF1alpha promoter via a functional NFkappaB site. Arterioscler Thromb
Vasc Biol 27: 755–761, 2007.
5. Chen HH, Mullett SJ, Stewart AF. Vgl-4, a novel member of the
vestigial-like family of transcription cofactors, regulates alpha1-adrenergic activation of gene expression in cardiac myocytes. J Biol Chem 279:
30800 –30806, 2004.
6. Chen L, Loh PG, Song H. Structural and functional insights into the
TEAD-YAP complex in the Hippo signaling pathway. Protein Cell 1:
1073–1083, 2010.
7. Chen R, Li L, Weng Z. ZDOCK: An initial-stage protein-docking
algorithm. Proteins 52: 80 –87, 2003.
8. Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG,
Thompson JD. Multiple sequence alignment with the Clustal series of
programs. Nucleic Acids Res 31: 3497–3500, 2003.
9. Colovos C, Yeates TO. Verification of protein structures: patterns of
nonbonded atomic interactions. Protein Sci 2: 1511–1519, 1993.
10. DeLano WL. The PyMOL Molecular Graphics System. Palo Alto, CA:
Delano Scientific, 2002.
11. Eisenberg D, Luthy R, Bowie JU. VERIFY3D: assessment of protein
models with three-dimensional profiles. Methods Enzymol 277: 396 –404,
1997.
12. Eltzschig HK, Carmeliet P. Hypoxia and inflammation. N Engl J Med
364: 656 –665, 2011.
13. Eswar N, Webb B, Marti-Renom MA, Madhusudhan MS, Eramian D,
Shen MY, Pieper U, Sali A. Comparative protein structure modeling
using MODELLER. Curr Protoc Protein Sci 2: 2.9, 2007.
14. Farrance IK, Mar JH, Ordahl CP. M-CAT binding factor is related to
the SV40 enhancer binding factor, TEF-1. J Biol Chem 267: 17234 –
17240, 1992.
15. Farrance IK, Ordahl CP. The role of transcription enhancer factor-1
(TEF-1) related proteins in the formation of M-CAT binding complexes in
muscle and non-muscle tissues. J Biol Chem 271: 8266 –8274, 1996.
16. Gray JJ, Moughon S, Wang C, Schueler-Furman O, Kuhlman B, Rohl
CA, Baker D. Protein-protein docking with simultaneous optimization of
rigid-body displacement and side-chain conformations. J Mol Biol 331:
281–299, 2003.
17. Gunther S, Mielcarek M, Kruger M, Braun T. VITO-1 is an essential
cofactor of TEF1-dependent muscle-specific gene regulation. Nucleic
Acids Res 32: 791–802, 2004.
18. Gupta M, Kogut P, Davis FJ, Belaguli NS, Schwartz RJ, Gupta MP.
Physical interaction between the MADS box of serum response factor and
the TEA/ATTS DNA-binding domain of transcription enhancer factor-1. J
Biol Chem 276: 10413–10422, 2001.
19. Gupta MP, Amin CS, Gupta M, Hay N, Zak R. Transcription enhancer
factor 1 interacts with a basic helix-loop-helix zipper protein, Max, for
positive regulation of cardiac alpha-myosin heavy-chain gene expression.
Mol Cell Biol 17: 3924 –3936, 1997.
20. Hayden MS, Ghosh S. Shared principles in NF-kappaB signaling. Cell
132: 344 –362, 2008.
21. Jin Y, Messmer-Blust AF, Li J. The role of transcription enhancer
factors in cardiovascular biology. Trends Cardiovasc Med 21: 1–5, 2011.
22. Jin Y, Wu J, Song X, Song Q, Cully BL, Messmer-Blust A, Xu M, Foo
SY, Rosenzweig A, Li J. RTEF-1, an upstream gene of hypoxia-inducible
factor-1alpha, accelerates recovery from ischemia. J Biol Chem 286:
22699 –22705, 2011.
23. Jung YJ, Isaacs JS, Lee S, Trepel J, Neckers L. IL-1␤-mediated
up-regulation of HIF-1␣ via an NF␬B/COX-2 pathway identifies HIF-1 as
a critical link between inflammation and oncogenesis. FASEB J 17:
2115–2117, 2003.
24. Kenneth NS, Mudie S, van Uden P, Rocha S. SWI/SNF regulates the
cellular response to hypoxia. J Biol Chem 284: 4123–4131, 2009.
25. Kortemme T, Baker D. A simple physical model for binding energy hot
spots in protein-protein complexes. Proc Natl Acad Sci USA 99: 14116 –
14121, 2002.
26. Laskowski RA, Rullmannn JA, MacArthur MW, Kaptein R, Thornton JM. AQUA and PROCHECK-NMR: programs for checking the
quality of protein structures solved by NMR. J Biomol NMR 8: 477–486,
1996.
27. Lyskov S, Gray JJ. The RosettaDock server for local protein-protein
docking. Nucleic Acids Res 36: W233–W238, 2008.
28. Maeda T, Chapman DL, Stewart AF. Mammalian vestigial-like 2, a
cofactor of TEF-1 and MEF2 transcription factors that promotes skeletal
muscle differentiation. J Biol Chem 277: 48889 –48898, 2002.
29. Mahoney WM Jr, Hong JH, Yaffe MB, Farrance IK. The transcriptional co-activator TAZ interacts differentially with transcriptional enhancer factor-1 (TEF-1) family members. Biochem J 388: 217–225, 2005.
30. Messmer-Blust AF, Philbrick MJ, Guo S, Wu J, He P, Guo S, Li J.
RTEF-1 attenuates blood glucose levels by regulating insulin-like growth
factor binding protein-1 in the endothelium. Circ Res 111: 991–1001,
2012.
31. Oliver KM, Taylor CT, Cummins EP. Hypoxia. Regulation of NFkappaB
signalling during inflammation: the role of hydroxylases. Arthritis Res
Ther 11: 215, 2009.
32. Olson EN. Gene regulatory networks in the evolution and development of
the heart. Science 313: 1922–1927, 2006.
33. Owens GK, Kumar MS, Wamhoff BR. Molecular regulation of vascular
smooth muscle cell differentiation in development and disease. Physiol
Rev 84: 767–801, 2004.
34. Reynolds C, Damerell D, Jones S. ProtorP: a protein-protein interaction
analysis server. Bioinformatics 25: 413–414, 2009.
35. Rius J, Guma M, Schachtrup C, Akassoglou K, Zinkernagel AS, Nizet
V, Johnson RS, Haddad GG, Karin M. NF-kappaB links innate immunity to the hypoxic response through transcriptional regulation of HIF1alpha. Nature 453: 807–811, 2008.
36. Shie JL, Wu G, Wu J, Liu FF, Laham RJ, Oettgen P, Li J. RTEF-1,
a novel transcriptional stimulator of vascular endothelial growth factor in
hypoxic endothelial cells. J Biol Chem 279: 25010 –25016, 2004.
37. Stewart AF, Suzow J, Kubota T, Ueyama T, Chen HH. Transcription
factor RTEF-1 mediates alpha1-adrenergic reactivation of the fetal gene
program in cardiac myocytes. Circ Res 83: 43–49, 1998.
38. Tapscott SJ. The circuitry of a master switch: Myod and the regulation of
skeletal muscle gene transcription. Development 132: 2685–2695, 2005.
39. van Uden P, Kenneth NS, Rocha S. Regulation of hypoxia-inducible
factor-1alpha by NF-kappaB. Biochem J 412: 477–484, 2008.
40. Vassilev A, Kaneko KJ, Shu H, Zhao Y, DePamphilis ML. TEAD/TEF
transcription factors utilize the activation domain of YAP65, a Src/Yesassociated protein localized in the cytoplasm. Genes Dev 15: 1229 –1241,
2001.
41. Yasunami M, Suzuki K, Ohkubo H. A novel family of TEA domaincontaining transcription factors with distinct spatiotemporal expression
patterns. Biochem Biophys Res Commun 228: 365–370, 1996.