Here we report that high mobility group protein N2 HMGN2, a member of the high mobility group superfamily that affects chromatin function, modulates the expression of HBD-2 in A549 cells
Trang 1lipopolysaccharide-induced expression of b-defensins in A549 cells
Lu-Xia Deng1,2, Gui-Xia Wu1, Yue Cao1, Bo Fan1, Xiang Gao1, Lin Luo3and Ning Huang1,4
1 Research Unit of Infection and Immunity, West China School of Preclinical and Forensic Medicine, Sichuan University, Chengdu, China
2 Chongqing Lummy Pharmaceutical Co Ltd, Chongqing, China
3 Department of Anesthesiology, West China College of Stomatology, Sichuan University, Chengdu, China
4 State Key Laboratory of Biotherapy, West China Hospital, Sichuan, Chengdu, China
Introduction
Antimicrobial peptides are produced by many different
tissues of the body when the innate immune system is
activated by lipopolysaccharide (LPS) or other
inflam-matory cytokines and chemokines Defensins are small
cationic antimicrobial peptides which have usually been
categorized into three major families based on the
molecular structure: a-, b-, and h-defensins [1–3] They
regulate both the innate and adaptive immune
responses, exhibiting broad-spectrum antimicrobial
activity against most Gram-positive and Gram-negative
bacteria, fungi and viruses such as influenza viruses
b-defensins are largely expressed in epithelial cells
on the surface of respiratory, gastrointestinal and geni-tourinary tracts and skin [3,4] Human b-defensin (HBD) 1 is constitutively expressed in the epithelial surface of the respiratory and genitourinary tracts [5]
In contrast, the expression of HBD-2, HBD-3 and HBD-4 is inducible in response to LPS, tumour necro-sis factor a (TNF-a), interleukin 1b or Gram-negative pathogens, underscoring their crucial role in epithelial host defence under inflammatory conditions [4,6] Inducible expression of b-defensins by human epithelial
Keywords
high mobility group protein N2; human
b-defensin-2; lipopolysaccharide; p65;
regulation
Correspondence
N Huang, Research Unit of Infection and
Immunity, West China School of Preclinical
and Forensic Medicine, Sichuan University,
Chengdu, China
Fax ⁄ Tel: +86 28 8506 8208
E-mail: huangpanxiao@126.com
(Received 13 August 2010, revised 5 April
2011, accepted 11 April 2011)
doi:10.1111/j.1742-4658.2011.08132.x
Human b-defensin-2 (HBD-2) is an antimicrobial peptide produced by the epithelial cells that plays an important role in innate and adaptive immu-nity Here we report that high mobility group protein N2 (HMGN2), a member of the high mobility group superfamily that affects chromatin function, modulates the expression of HBD-2 in A549 cells treated by lipo-polysaccharide Mechanistically, HMGN2 prolongs the retention time and enhances the accumulation of nuclear factor jB p65 in the nucleus, and promotes the acetylation of p65 through increasing histone acetyltransfer-ase activity and enhancing p65-Ser536 phosphorylation Additionally, chro-matin immunoprecipitation reveals that HMGN2 and p65 synergistically promote their specific binding to HBD-2 promoter, thereby affecting the downstream transcription Taken together, these results suggest that HMGN2 acts as a positive modulator of nuclear factor jB signalling to promote lipopolysaccharide-induced b-defensin expression
Abbreviations
AA, anacardic acid; ChIP, chromatin immunoprecipitation; Co-IP, co-immunoprecipitation; Glut2, type 2 glucose transporters; HAT, histone acetyltransferase; HBD, human b-defensin; HDAC, histone deacetylase; HMG, high mobility group; HMGA, HMG-AT-hook family; HMGB, HMG-box family; HMGN2, high mobility group protein N2; IjBa, NF-jB inhibitor a; LPS, lipopolysaccharide; MAPK, mitogen-activated protein kinase; NF-jB, nuclear factor jB; PE-H, pEGFPN1-HMGN2; Psi-H, pSilencer-HMGN2-2; siRNA2, small interfering RNA2; TNF-a,
tumour necrosis factor a; TSA, trichostatin.
Trang 2cells has been shown to be regulated through several
signalling pathways, such as the nuclear factor jB
(NF-jB) pathway, the p38 mitogen-activated protein
kinase (MAPK) pathway, the c-Jun N-terminal kinase
pathway and the phosphatidylinositol-3-kinase⁄ AKT
pathway [7–11]
The chromatin-associated high mobility group
(HMG) superfamily includes the HMG-AT-hook
(HMGA) family, HMG-box (HMGB) family and
HMG nucleosome binding (HMGN) family and
mod-ulates a wide range of DNA-dependent activities such
as chromatin structure, post-translational
modifica-tions, and rates of transcription, replication and
recombination [12–15] Interestingly, recent studies
have demonstrated that some members of the HMG
family such as HMGA1 and HMGB1 are directly
involved in the regulation of NF-jB activation and its
downstream gene transcription [3,16–18] It is well
known that NF-jB activation is regulated by
RelA⁄ p65 acetylation and deacetylation which are
mediated by histone acetyltransferases (HATs) and
de-acetylases (HDACs), respectively [19,20]
HAT-medi-ated acetylation of RelA⁄ p65, particularly at lysine 310
and to a lesser extent at lysine 221, enhances NF-jB⁄
DNA binding and attenuates its interaction with
NF-jB inhibitor a (IjBa) [21] Conversely, acetylated
RelA⁄ p65 is deacetylated by HDACs, leading to the
repression of downstream gene transcription and
IjBa-dependent nuclear export of IjBa⁄ NF-jB
com-plex [20,22] The NF-jB activity and the expression of
its downstream gene alter when HAT and HDAC
inhibitors respectively are used [23–26]
HMGN2 has been shown to partially decrease
HDAC activity [27] and also to enhance the H3K14ac
level in chromatin by stimulating HAT activity [13]
Nevertheless, it is unclear whether HMGN2 modulates NF-jB activity by changing the activity of HATs and⁄ or HDACs and thereby regulates the expression
of downstream target HBD-2 Therefore, in the present study we initially supposed that (a) HMGN2 elevates the activity of NF-jB, enhances the amount of NF-jB p65 in the nucleus and regulates the equilibrium between HATs and HDACs thereby affecting p65 acet-ylation; and (b) HMGN2 itself or through interaction with p65 binds to the chromatin in the promoter region of b-defensin genes to enhance HBD-2 expres-sion This research aimed to confirm the initial hypoth-esis through transient transfection and luciferase experiments, the activity of HAT and HDAC blocking experiments, chromatin immunoprecipitation (ChIP) assay and co-immunoprecipitation (Co-IP) etc
Results
Gene expression profiles after HMGN2 knockdown or⁄ and LPS stimulation in A549 cells
As the first step to characterizing the function of HMGN2 in transcriptional regulation in response to LPS, we prepared knockdown HMGN2 in A549 cells and found that HMGN2 expression at both mRNA and protein levels was significantly downregulated in A549 cells treated by HMGN2-specific small interfer-ing RNA2 (SiRNA-HMGN2-2) (Figs 1A,B and S1) Next we employed a cDNA microarray to examine the effect of reduced endogenous HMGN2 level and⁄ or LPS treatment on gene expression profiles in A549 cells (Figs S2 and S3) The results showed that HMGN2 downregulation altered the expression of over 4% of 31 000 genes by twofold or more (Table 1)
0.8 0.6 0.4 0.2 0
*
1.0
HMGN2 RT-PCR western blot
-actin HMGN2
HMGN2 Histone H3
*
Western blot RT-PCR
PE-H PE
Psi-H blank Psi Psi-H/PE-H Psi-H/PE
PE-H
PE
Psi-H blank
Psi
PE-H PE Psi-H blank Mak
er Psi
Psi-H/PE-H Psi-H/PE
Fig 1 Efficient knockdown of HMGN2 in
A549 cells Representative RT-PCR (A) or
western blot (B) results showing that
HMGN2 mRNA and protein levels were
reduced by over 80% after siRNA2
treat-ment b-actin and histone H3 served as the
loading control for RT-PCR and western blot,
respectively B, blank group; N, normal
group (without LPS treatment); M, marker.
1–3, different HMGN2 siRNA (C)
Represen-tative RT-PCR and western blot results
showing the expression of HMGN2 mRNA
and protein in the different established
sta-ble A549 cells (D) Values are presented as
mean ± SD for at least five independent
experiments performed in triplicate.
*P < 0.01 versus blank group.
Trang 3Table 1 List of genes with changed expression in microarray analysis.
GenBank
CGAP gene
symbol
Log 2 ratio
Entrez gene description
NM_007315.2 STAT1 3.5 )2.0 1.7 Signal transducer and activator of transcription 1, 91 kDa NM_139276.2 STAT3 5.3 )2.4 2.1 Signal transducer and activator of transcription 3
(acute-phase response factor)
NM_001556.1 IKBKB )4.1 3.1 )1.9 Inhibitor of kappa light polypeptide gene enhancer in B-cells,
kinase beta
NM_021975.2 RELA 5.8 )3.1 1.2 v-rel reticuloendotheliosis viral oncogene homolog A, nuclear
factor of kappa light polypeptide gene enhancer in B-cells
3, p65 NM_000346.2 SOX9 )0.2 )4.2 )4.1 SRY (sex determining region Y) box 9 (campomelic dysplasia,
autosomal sex-reversal)
NM_014261.1 TRIF 2.8 )1.4 1.9 TIR domain containing adaptor inducing interferon-beta
NM_003225.2 TFF1 1.2 3.2 4.8 Trefoil factor 1 (breast cancer, estrogen-inducible sequence
expressed in)
BU739862 MED19 )0.9 )0.5 )1.1 Mediator of RNA polymerase II transcription, subunit 19
homolog (yeast) NM_003187.3 TAF9 )0.5 )1.1 )1.2 TAF9 RNA polymerase II, TATA box binding protein (TBP)
associated factor, 32kDa
homolog (yeast)
NM_001904.2 CTNNB1 4.8 3.4 7.2 Catenin (cadherin-associated protein), beta 1, 88kDa
Trang 4Among these genes, only about 2% were upregulated,
while the rest were downregulated All these genes have
been annotated in the NCBI Reference Sequence
data-base In contrast, LPS stimulation obviously activated
several signalling pathways including the NF-jB
path-way to induce the expression of cytokines and
antimi-crobial peptides About 5% of genes were upregulated
and 3% were downregulated in group C compared with
group A In addition, 4% of genes were upregulated
and 2% were downregulated in group D compared with
group A (Table 1) Significantly, the expression of
DEFB4 (HBD-2) and DEFB103A (HBD-3) was
chan-ged and the ratios of log 2 were increased after LPS
induction and decreased after HMGN2 knockdown
The microarray analysis of the human genome chip
containing 31 000 genes revealed differential expression
of 3–5% in the four groups (with a false discovery rate
corrected P£ 0.05 and fold change ‡ 1) The changed
gene expression profiles could be categorized into
fol-lowing gene ontology: MAPK, focal adhesion, Toll-like
receptor, epithelial cell, regulation of actin cytoskeleton,
vascular endothelial growth factor, Fc epsilon, Wnt,
cytokine–cytokine receptor interaction, apoptosis,
adhe-rens junction, dorso-ventral axis formation, T cell
receptor, insulin and JAK–STAT signalling pathways
HMGN2 regulates LPS-induced HBD-2 expression
in A549 cells
Since microarray analysis showed that the expression
of HBD-2 was significantly changed after LPS
induc-tion and HMGN2 knockdown, next we aimed to
con-firm these results by RT-PCR and western blot We
employed a variety of A549 cells with stable
transfec-tions of pSilencer-HMGN2-2 (Psi-H),
pEGFPN1-HMGN2 (PE-H), control siRNA (Psi, PE), wild-type A549 cells (blank), and reintroduction of HMGN2 expression vector or control vector to HMGN2 knock-down A549 cells (Psi-H⁄ PE-H or Psi-H ⁄ PE, respec-tively; Figs 1C,D, S4 and S5) These cells were treated with 0, 20, 40, 60, 80 or 100 lgÆmL)1 LPS for 24 h and then the RNA and protein were isolated for RT-PCR and western blot analysis The results demon-strated that LPS induced HBD-2 expression in a dose-dependent manner which can be abolished by HMGN2 knockdown However, reintroduction of HMGN2 into HMGN2 knockdown cells recovered LPS-induced expression of HBD-2 (Fig 2A–D) Taken together, these data suggest that HMGN2 is crucial for LPS-induced HBD-2 expression
To verify that the decreased levels of HBD-2 protein and transcripts are indeed linked to the expression levels of HMGN2, the plasmid expressing double-point mutant HMGN2-S24, 28E was prepared (this protein enters the nucleus but does not bind to chromatin); then we examined the levels of HBD-2 expression level
in HMGN2) A549 cells that were stably transformed with plasmid expressing either the wild-type HMGN2 protein (PEGFPN1-HMGN2) or the double-point S24, 28E HMGN2 protein (PEGFPN1-HMGN2-S24, 28E deletion mutants were generated by PCR amplification
of the corresponding part of HMGN2 cDNA) In A549 cells, replenishment of HMGN2 protein upregu-lated the protein levels of HBD-2, an indication that the levels of this protein are indeed linked to the cellu-lar levels of HMGN2 In contrast, replenishment of the S24, 28E HMGN2 double-point mutant, which does not bind to chromatin, did not change the levels
of HBD-2, suggesting that the interaction of HMGN2 with chromatin regulates HBD-2 expression
signifi-Table 1 (Continued).
GenBank
CGAP gene symbol
Log 2 ratio
Entrez gene description
NM_001791.2 CDC42 4.0 2.5 7.1 Cell division cycle 42 (GTP binding protein, 25kda)
NM_003387.3 WASPIP 0.1 )1.8 )1.1 Wiskott–Aldrich syndrome protein interacting protein
NM_000539.2 RHO 4.3 1.2 6.1 Rhodopsin (opsin 2, rod pigment) (retinitis pigmentosa 4,
auto-somal dominant) NM_005406.1 ROCK1 0.6 )2.4 )2.5 Rho-associated, coiled-coil containing protein kinase 1
NM_018890.2 RAC1 3.1 1.6 3.7 Ras-related C3 botulinum toxin substrate 1 (rho family, small
GTP binding protein Rac1)
Trang 5cantly (Fig 2E) Overall, the results support an
impor-tant role of HMGN2 with chromatin regulation in
inducible HBD-2 expression
HMGN2 regulates NF-jB activity in A549 cells
To reveal potential mechanisms by which HMGN2
regulates LPS-induced HBD-2 expression, we first
examined whether HMGN2 could modulate NF-jB
activity in A549 cells because the promoter region of HBD-2 contains four NF-jB binding sites [4] We observed that LPS led to increased NF-jB levels in the nucleus and decreased NF-jB levels in the cytoplasm However, the LPS-induced change of NF-jB distribu-tion was markedly attenuated by HMGN2 knockdown, indicating that loss of HMGN2 inhibits LPS-induced NF-jB accumulation in the nucleus On the other hand, gain of HMGN2 due to overexpression
D C
F E
PE-H group
PE group
HBD2 β-actin
HBD-2 β-actin
HBD2 β-actin
0 20 40 60 80 100
HBD2 β-actin
HBD-2 β-actin HBD-2 β-actin
Psi-H/PE-H group
Psi-H/PE group
HBD2 β-actin
Psi-H group
B group
HBD-2 real time PCR
0
4 6
2
*
*
*
*
*
*
*
*
0.2
0.8
0.6
0 0.4
1.0
HBD-2 Western blotting
*
*
*
*
*
*
*
*
*
* Psi-H/PE group
PE-H group
PE group
B group Psi-H group Psi group Psi-H/PE-H group
Psi-H/PE group
PE-H
PE group
B group Psi-H group Psi group Psi-H/PE-H group
Psi group
Psi-H/PE-H group
β-actin HBD2 β-actin HBD2 β-actin HBD2 β-actin HBD2 β-actin HBD2
Psi-H group
B group
Psi-H/PE group
β-actin HBD2
PE-H group
β-actin HBD2
PE group
Psi group
0 20 40 60 80 100
Cell
Transfected
HMGN2
HBD-2
β-actin
(PEGFPN1-HM
GN2)
(PEGFPN1-HMG N2-S24,28E)
0 0.2 0.4 0.6 0.8
HMGN2 HMGN2-S24,28E
Transfected–
Transfected+
Fig 2 HMGN2 is crucial for LPS-induced HBD-2 expression in A549 cells The cells were incubated in DMEM with 10% FBS containing 0, 20, 40, 60, 80 or 100 lgÆmL)1 LPS for 24 h and then HBD-2 mRNA and protein levels were detected by RT-PCR and western blot (A) Representative RT-PCR results showing the expression of HMGN2 mRNA in the different established stable A549 cells b-actin served as the loading control (B) Values of RT-PCR are expressed
as relative expression compared with the blank group and presented as mean ± SD for at least five independent experiments performed in triplicate *P < 0.01 versus blank group (C) Representative western blot results showing the expression of HMGN2 protein in the different established stable A549 cells b-actin served as the load-ing control (D) Photodensitometric analysis
of western blot is presented as mean ± SD for at least five independent experiments performed in triplicate *P < 0.01 versus blank group (E) Western blot analysis of stably transfected HMGN2)A549 cells expressing either HMGN2 or the HMGN2-S24, 28E A549 HMGN2)denotes control, non-transformed HMGN2)cells (F) Values
of western blots are expressed as relative expression compared with b-actin and pre-sented as mean ± SD for at least five inde-pendent experiments performed in triplicate.
Trang 6ducibly prompted the accumulation of NF-jB in the
nucleus (Fig 3A–D)
To further confirm that HMGN2 modulates
LPS-induced NF-jB activation in A549 cells, A549 cell lines
harbouring a transient NF-jB-dependent luciferase
reporter were utilized Following LPS (100 lgÆmL)1)
treatment for 4 h, a 20- to 25-fold increase in luciferase
activity from the baseline level was seen in normal
A549 cells (Fig 3E) Compared with normal A549 cells
(group B), luciferase activity was increased 25% in HMGN2 overexpressing cells (PE-H) and decreased 50% in HMGN2 knockdown cells (Psi-H) Signifi-cantly, restoration of HMGN2 expression in HMGN2 knockdown cells increased luciferase activity to a level comparable with normal A549 cells (Psi-H⁄ PE-H) (Fig 3E) Taken together, these data prove that HMGN2 regulates LPS-induced NF-jB activity in A549 cells
B
LPS stimulated concentrations (µg·mL–1) 0
0.2 0.4 0.6 0.8 1.0
*
*
*
*
*
*
*
*
*
Psi-H/PE group
PE-H group
PE group
B group Psi-H group Psi group Psi-H/PE-H group
A
B
Psi-H/
PE-H
PE
Psi
Psi-H/
PE Psi-H
PE-H
C
LPS stimulated time (min)
B
Psi-H/
PE-H
PE
Psi
Psi-H/
PE Psi-H PE-H
0 5 1 3 6 1 0 1 0 0 5 1 3 6 1 0 1 0
D
0.2
0.4
0.6
0.8
1.0
40
30
20
10
0 LPS stimulated time (min)
0
*
*
*
*
*
*
Psi-H/PE group
PE-H group
PE group
B group
Psi-H group
Psi group
Psi-H/PE-H group
E
*
*
* Psi-H/PE group
PE-H group
PE group
B group Psi-H group Psi group Psi-H/PE-H group
PE -H PE
B
Ps i-H Ps i
Ps i-H /P E -H
Ps i-H /P E
Fig 3 HMGN2 regulates NF-jB activity in A549 cells (A–D) HMGN2 promotes the nuclear accumulation of p65 At 24 h after transfection, the cells were incubated in DMEM with 10% FBS containing 0, 20, 40, 60, 80 and 100 lgÆmL)1LPS Fresh medium was added 1 h later (A) Representative western blot results showing the cytoplasmic and nuclear distribution of p65 in the different established stable A549 cells (B) Photodensitometric analysis of western blot is presented as mean ± SD for at least five independent experiments performed in triplicate *P < 0.01 versus blank group The cells were incubated in DMEM with 10% FBS containing 100 lgÆmL)1LPS Fresh medium was replaced 0, 5, 15, 30, 60, 120, 180 min later (C) Representative western blot results showing the cytoplasmic and nuclear distribution of p65 in the different established stable A549 cells (D) Photodensitometric analysis of western blot is presented as mean ± SD for at least five independent experiments performed in triplicate *P < 0.01 versus blank group (E) HMGN2 promotes the transcription activity of NF-jB The cells were transfected with NF-jB luciferase reporter and treated by LPS The luciferase activity is presented as mean ± SD for at least five independent experiments performed in triplicate *P < 0.01 versus blank group The luciferase activity in the blank group untreated by LPS (mock) was utilized as the control value.
Trang 7HMGN2 modulates the acetylation of p65
The transcription factor NF-jB activity is known to
be regulated by reversible acetylation through HATs
and HDACs Anacardic acid (AA) inhibits HAT
activ-ity while trichostatin (TSA) inhibits HDAC activactiv-ity
Therefore, we used these reagents to treat A549 cells
Western blot analysis for p65-Lys310 acetylation in
AA-treated or TSA-treated cells demonstrated clearly
(Table 2) that the acetylation of p65-Lys310 was decreased in HMGN2 knockdown cells with
100 lgÆmL)1 LPS for 5, 15, 30, 60 or 120 min and increased in HMGN2 overexpressing cells with
100 lgÆmL)1LPS for 60 and 120 min but the later dif-ference was not statistically significant (Fig 4A,B), suggesting that HMGN2 promotes the acetylation of p65-Lys310 mainly by enhancing HAT activity Since p65-Lys310 acetylation depends on p65-Ser536 phos-phorylation, we also examined the effect of HMGN2
on p65-Ser536 phosphorylation and found that HMGN2 could enhance the phosphorylation of p65
on Ser536 as well Furthermore, western blot analysis demonstrated that the phosphorylation of p65 on Ser536 was upregulated in the PE-H group and down-regulated in the Psi-H group compared with the blank group However, inhibition of HDAC or HAT did not affect p65 phosphorylation (Fig 4C,D) These results demonstrated that HMGN2 promoted p65-Lys310 acetylation via enhancing p65-Ser536 phosphorylation and increasing HAT activity
Table 2 Cell grouping in HAT and HDAC activity experiments.
PE-H group A549 cells stably transfected with
PEGFPN1-HMGN2 PE-H ⁄ AA group A549 cells stably transfected with
PEGFPN1-HMGN2 and then adding AA
Psi-H group A549 cells stably transfected with
pSilencer-HMGN2-2 Psi-H ⁄ TSA group A549 cells stably transfected with
pSilencer-HMGN2-2 and then adding TSA
B group
Psi-H/TSA group
PE-H group
5 15 30 60 120
PE-H/AA group
Psi-H group
B A
a-acetyl-p65 whole cell extract
0.4 0.6
0.2
0 0.8
*
*
1.0
*
*
*
*
D C
Psi-H/TSA group
B group PE-H group PE-H/AA group
LPS stimulated time (min) LPS stimulated time (min)
*
*
*
Psi-H/TSA
PE-H group
5
PE-H/AA group
Psi-H group
a-phosph-p65 whole cell extract
0.4 0.6
0.2
0 0.8
*
*
1.0
*
Nucleus
Psi-H group Psi-H/TSA group
B group PE-H group PE-H/AA group
LPS stimulated time (min)
LPS stimulated time (min)
*
*
60
Fig 4 HMGN2 increases the HAT activity
to promote NF-jB activation The A549 cells were pre-incubated with 25 l M AA for 4 h
or 100 n M TSA for 18 h and treated with
100 lgÆmL)1LPS for 5, 15, 30, 60 or
120 min (A) Representative western blot results showing the acetylation of p65-Lys310 in the indicated A549 cells (B) Photodensitometric analysis of western blot
is presented as mean ± SD for at least five independent experiments performed in triplicate *P < 0.01 versus blank group.
P < 0.01 PE-H ⁄ AA versus PE-H group (C) Representative western blot results show-ing the phosphorylation of p65-Ser536 in the indicated A549 cells (D) Photodensitometric analysis of western blot is presented as mean ± SD for at least five independent experiments performed in triplicate.
*P < 0.01 versus blank group.
Trang 8HMGN2 binds to HBD-2 promoter upon LPS
stimulation
Members of the HMGN family have been reported to
affect the chromatin structure to alter the recruitment
of transcription factors to promoter [28] To explore
the mechanism by which HMGN2 affects LPS-induced
expression of HBD-2, we performed ChIP analysis of
the 13 kb region of HBD-2 gene using antibodies to
HMGN1 or HMGN2 and demonstrated a twofold to
fourfold enrichment of HMGN2 over a )3 kb region
5¢ to the start of transcription, suggesting that the
region for P1 amplification contains the enrichment
region for HMGN2 and HBD-2 chromatins (Fig 5)
More importantly, in the HBD-2 chromatin derived
from the HMGN2) A549 cells, the level of HMGN1
was enriched in several regions, especially those
span-ning primer sets P13–P6 and P5–P2 (Fig 5) The
increased level of HMGN1 in the HBD-2 chromatin of
the HMGN2) A549 cells suggested that HMGN1 and
HMGN2 had functional redundancy among them
Furthermore, to address the possibility that
HMGN2 promotes the interaction of NF-jB p65 with
HBD-2 promoter, we produced knockdown HMGN2
and performed ChIP analysis to show that depletion
of HMGN2 affected the chromatin binding of NF-jB
(Figs 6A and S6), indicating that HMGN2 enhances
the interaction between p65 and HBD-2 promoter
Thus we hypothesized that HMGN2 and p65 may
pre-assemble into a regulatory complex on HBD-2
pro-moter However, Co-IP experiments demonstrated that
p65 and HMGN2 did not assemble into a complex in
the nuclear compartment because HMGN2 antibody
efficiently precipitated HMGN2 from nuclear extracts
of A549 cells but the precipitates did not contain p65 (Fig 6B) Likewise, anti-p65 immunoprecipitated p65 efficiently and the precipitates did not contain HMGN2 (Fig 6B) Next, we examined whether HMGN2 and p65 share the same binding sites on HBD-2 promoter ChIP analysis indicated that the antibodies to HMGN2 and p65 efficiently immunopre-cipitated chromatin containing the respective proteins (Fig 6C), but the chromatin with HMGN2 enrichment does not contain p65 (Fig 6C, upper left panel) and the chromatin with p65 enrichment does not contain HMGN2 (Fig 6C, lower right panel) Based on these results we conclude that HMGN2 could not interact with p65 in nuclear extracts and they did not share the same chromatin binding site
In contrast, results of reciprocal sequential ChIP analysis revealed that chromatin that was sequentially immunoprecipitated with HMGN2 and p65 antibody was enriched in the DNA of HBD-2 promoter (P1) while the HBD-2 promoter was enriched in chromatin that was sequentially immunoprecipitated with anti-p65 and anti-HMGN2 (Fig 6D,E), suggesting that HMGN2 and p65 mutually promote their binding to the promoter of HBD-2
Discussion
The mechanism underlying the regulation of the expression of antimicrobial peptides including b-defen-sins has not been elaborated at the transcriptional level Our studies indicate that chromatin binding pro-tein HMGN2 mediates the LPS-induced expression of
P1
B A
–4
Ex1
M 1 2 The ratio of chromosome immunoprecipitation
HMGN1/INPUT HMGN2/INPUT 8.0
6.0
4.0
2.0
0.0
-/HMGN1/INPUT
P18 P17 P16 P15 P14 P13 P12 P11 P10 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 G1 G2 G3 G4 G5 G6 G7 G8
Fig 5 HMGN2 is enriched in HBD-2
pro-moter chromatin in A549 cells (A)
Enrich-ment of each DNA sequence in the HMGN2
or HMGN1 immunoprecipitate relative to
the input DNA is normalized and plotted as
the position of the PCR primer pair within
the HBD-2 gene locus Each point is an
averaged value from three independent
experiments (B) The electrophoretogram
chromosome ultrasonication M, maker; 1,
the whole chromosome; 2, chromosome
analysis following ultrasonication.
Trang 9HBD-2 in A549 cells Our microarray analysis
indi-cated that depletion of HMGN2 protein altered the
expression level of over 4% of genes by twofold or
more in A549 cells Significantly, the HBD level in the
LPS group was fivefold higher than that in the control
group, while the HBD-2 level in the HMGN2
knock-down group was threefold less than in the LPS group
The results of the microarray are reliable because they
agree well with the expression of genes known to be
modulated by HMGN2, such as N-cadherin,
Sry-related HMG-box gene 9 (Sox9), pituitary homeobox
2, heat shock proteins and type 2 glucose transporters
(Glut2) [4,29–35] Pathway analysis of the microarray
results showed that HMGN2 modulates the Toll–
NF-jB pathway upon LPS stimulation because the
expression of RELA (p65), IKBKB (IjB) and myeloid
differentiation primary response gene 88 (MyD88),
which were the main members on the Toll–NF-jB
pathway, was changed in HMGN2 knockdown plus
LPS groups compared with the LPS group The changes of gene expressions observed in the microarray were identified by RT-PCR and western blotting Finally, the results detected in the microarray were consistent with those achieved through RT-PCR and western blotting, including the change of HBD-2 expression Compared with A549 cells stimulated by LPS, HBD-2 expression was 50% less in HMGN2 knockdown cells and was over 30% higher in HMGN2 overexpressing cells In addition, reintroduction of HMGN2 re-expression led to the recovery of HBD-2 expression by over 70% in HMGN2 knockdown cells Overall, these findings prove that HMGN2 plays an essential role in LPS-induced HBD-2 expression in A549 cells
Next we aimed to elucidate the molecular mecha-nism by which HMGN2 regulates HBD-2 expression HBD-2 promoter contains several binding sites for transcription factors including NF-jB, NF-IL-6 and
p65
HMGN2
IP
:
A
Control IgG
Contr
Control lgG Control lgG α-HMGN2 10% input α-P65 α-HMGN21% input Control lgG α-P65
Control lgG
α-p65
2 nd ChIP
HBD-2 P1
Input
control IgG
1 st
: α-p65
2 nd
st
2 nd
: α-p65
2 nd
ChIP
HBD-2 P1
Input
Control IgG
WB:
1.5
1.0
0.5
0
Psi Psi-H
Fig 6 HMGN2 and p65 independently bind HBD-2 promoter in A549 cells (A) Depletion
of HMGN2 reduced the binding of p65 to HBD-2 promoter (B) Co-IP assay showing that HMGN2 and p65 did not form a com-plex in the nucleus IP, antibody used for immunoprecipitation; WB, antibody used for western blot (C) ChIP assay showing that HMGN2 and p65 were not co-localized in the chromatin (D) HMGN2 and p65 were co-localized in the promoter region of HBD-2 chromatin, with the IgG as a negative con-trol (E) ChIP assay showing that HMGN2 or p65 bound to HBD-2 chromatin at P1.
Trang 10AP-1 [36–38] Previous studies found that the
upregu-lation of HBD-2 promoter activity was mainly
depen-dent on NF-jB in A549 cells, and LPS is known to
induce the activation of NF-jB Therefore, we
pro-posed that HMGN may mediate LPS-induced HBD-2
expression through the NF-jB signalling pathway To
examine this possibility we first performed western blot
analysis to show that the accumulation of p65 protein
in the nucleus, indicative of NF-jB activation, was
increased in HMGN2 overexpresssing A549 cells but
decreased in HMGN2 knockdown cells Next, we
employed NF-jB luciferase reporter assay to quantify
NF-jB activation [39] The NF-jB-induced luciferase
activity was significantly diminished in HMGN2
knockdown cells and increased in HMGN2
overex-pressing cells Based on these data we could conclude
that HMGN2 is crucial for LPS-induced NF-jB
acti-vation
NF-jB activation is known to be reciprocally
regu-lated by RelA⁄ p65 acetylation and deacetylation
medi-ated by HATs and HDACs HDACs and HATs are
enzymes that influence transcription by selectively
de-acetylating or de-acetylating the e-amino groups of lysine
located near the amino termini of core histone
pro-teins Acetylation of p65 at lysines 218, 221 and 310
by HATs including the general transcriptional
coacti-vators CBP and P300 would impair the association
between IjBa and p65, thus enhancing the binding
affinity of p65 for DNA [40] The effect of HATs is
compromised by HDACs that deacetylate p65 and
thus promote the interaction between p65 and IjB
HDACs are categorized into two classes: class I
HDAC 1, 2, 3, 8 and 11, and class II HDAC 4, 5, 6,
7, 9 and 10 Previous studies reported that the
associa-tion of NF-jB with HDAC1 and HDAC2 co-repressor
proteins functions to repress the expression of NF-jB
regulated genes [19,20,22]
Interestingly, several studies suggested the
relation-ship between HMGN proteins and the activity of
HATs and HDACs An in vitro assay showed that
HMGN1 and HMGN2 partially inhibit the
endoge-nous mouse HDAC activity [27] HMGN1 enhances
the acetylation level of lysine 14 in the tail of H3,
while HMGN1 and HMGN2 increase the acetylation
through enhancing the activity of HATs [13] AA is
an HAT inhibitor and inhibits the nuclear
transloca-tion and acetylatransloca-tion of p65, repressing TNF-induced
NF-jB dependent reporter gene expression [25] In
contrast, TSA, an HDAC inhibitor, enhances p65
acet-ylation induced by Gram-negative bacteria and
trans-forming growth factor b1 [41] In addition, HDAC
inhibitor SFN led to a time- and dose-dependent
upregulation of HBD-2 mRNA and protein expression
in Caco-2, HT-29 and SW480 cells [42] This is paral-leled by changes in the acetylation of distinct core pro-teins, H4 and HMGN2, resulting in the induction of LL-37, a member of the antimicrobial peptide that protects the urinary tract against invasive bacterial infection [43]
In the current study, AA (HAT inhibitor) was selected to test whether blocking HATs diminishes the acetylation level of the p65-Lys310 subunit in the PE-H transfected A549 cells If HMGN2 promotes the acety-lation of p65 through increasing HAT activity, AA pre-treatment would reduce the acetylation level of p65-Lys310 in the nucleus in the PE-H⁄ AA group compared with the PE-H group On the other hand, TSA (HDAC inhibitor) was used to examine the acetylation level of p65-Lys310 in the Psi-H transfected A549 cells If HMGN2 promotes the acetylation of p65 through inhibiting HDAC activity, the TSA pretreatment would augment the acetylation level of p65-Lys310 in the nucleus in the Psi-H⁄ TSA group compared with the Psi-H group Because p65-Lys310 acetylation was only allowed when p65-Ser536 was phosphorylated, the p65 global phosphorylation status on Ser536 was also tested
by western blotting in five groups
The results demonstrated that the amount of p65-Lys310 acetylation in the nucleus of A549 cells in the PE-H group was low compared with that in the pres-ence of AA (Fig 4A,B) Adding AA to the HMGN2 overexpressing cells helped to reduce the acetylation level of p65-Lys310 in the nuclei to the normal level, while adding TSA to the HMGN2 knockdown cells did not bring the acetylation level of p65-Lys310 in the nucleus to normal; the difference was not statistically significant Results of western blotting for p65-Lys310 acetylation in AA-treated or TSA-treated cells indi-cated that HMGN2 increased acetylation of p65-Lys310 mainly by enhancing the activity of HATs Since p65-Lys310 acetylation depends on p65-Ser536 phosphorylation, we also examined the effect of HMGN2 on p65-Ser536 phosphorylation Subse-quently, thePE-H plus AA group or the Psi-H plus TSA group did not modify p65-Ser536 phosphoryla-tion compared with the PE-H or Psi-H group respec-tively, the difference not being statistically significant However, western blot analysis demonstrated that the phosphorylation of p65 on Ser536 was upregulated in the PE-H group and downregulated in the Psi-H group compared with the blank group Collectively, these results indicated that HMGN2 promotes p65-Lys310 acetylation mainly via increasing HAT activity and enhancing p65-Ser536 phosphorylation
Lastly, we performed ChIP analysis to demonstrate that HMGN2 enhances HBD-2 transcription directly by