Induction of GADD45α expression contributes to the anti-proliferative effects of polymethoxyflavones on colorectal cancer cellsJ O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4
.sc iencedi rect .com
journal homepage: www.elsevier .com/ locate / j f f
Induction of GADD45a expr ession contr ib utes to the anti-pr olifer ati ve effects of pol ymetho xyflavones on color ectal cancer cells
And y Chi -Lung Lee a,1,2, Wei-Chun Hsiao a,1, Duncan Edw ar d Wright a, Shin Yen Chon gb, Siew Keng Leow b, Chi-T ang Ho c, Cheng-Fu Kao a, Yi-Chen Lo b,*
aInstitute of Cellular and Or ganismic Biolog y, Academia Sinica, Nankang, Taipei 11529, Taiw an bGr aduate Institute of Food Science and Tec hnolog y, National Taiw an Univer sity , Taipei 10617, Taiwan cDepartment of Food Science , Rutger s Univer sity , New Brunswic k, NJ 08901-85 20, USA
A R T I C L E I N F O A B S T R A C T
Article his tory:
Receiv ed in re vised form
4 Jan uary 2013
Availa ble online 15 Fe bruary 2013
Keywor ds:
1756-4646/$ - see front matter 2013 Else vi http://dx.doi.org/ 10.1016/j.jf f.2013.01.003
* Corresponding author . Address: Graduate Roose velt Rd., Da’an Dist., Taipei 10617, Taiw
E-mail address: lo yichen@ntu.edu.tw (Y.- 1 These authors contribute d equal ly to this 2 Current address: Genomic Research Cent
Abbreviations: PMFs, polymethoxyflavon monometh ylation; H3K4me3, his tone H3 ly lysine 27 acetylation; ChIP , chro matin immu inducible protein a; iNOS, nitric oxide synth
Polymethoxyflavones (PMFs), such as nobiletin and tang eretin, ha ve a high cellular uptake
rate , on account of their hydrophobicity . Nobiletin has been sho wn to do wn-re gulate
inflammatory gene expression, and is thus considered a promising anti- inflammatory
and anti-carcinogeni c agent. Most proteins kno wn to be affected by PMFs are cell cycle re g-
ulators and kinases. In the present stud y, we demonstrate that PMF-med iated induction of
GADD45a partially underlies the anti-prolifer ative effect of PMF on colorectal cancer cells.
PMF-med iated induct ion of GADD45a was found to be associated with chr omatin remodel-
ing events at a no vel enhancer element residing at intron 1 of GADD45a. Moreo ver ,
GADD45a transcription correlated with differential acetylation of lysines 14 and 27 of his-
tone H3 (H3K14/27) follo wing tr eatm ent with either nobiletin or tang eretin. PMF- induced
changes in H3K27 ace tylation, H2A.Z and CB P wer e observed at intron 1, but not at p53
response-elements of GADD45a, sugg esting that nobiletin/tang er etin-m ediated GADD45a
activ ation ma y be independent of p53.
2013 Else vier Ltd. All right s reserve d.
1. Intro duction
Pol ymethoxyflavones (PMFs), whic h include nobileti n and
tangere tin, are compoun ds found in citrus fruits that ha ve
been used to trea t se ver al diseases (revie wed by Li et al.,
er Ltd. All rights reserv ed
Institute of Food Scien an. Tel.: +886 2 3366 412
C. Lo).
work.
er , Academia Sinica, Nan es; 5-HTMF , 5-h ydroxy sine 4 trimeth ylat ion; H
nopre cipitatio n; p53RE, p ase; CO X2, cytochr ome c
2009). PMFs contai n se ve ral full y meth ylated hydr oxyl gro ups
(Fig . 1), which confer high hydroph obicity that resul ts in a
higher cellular uptake rate , as compar ed to non-meth ylated
polyh ydr oxylated flavo noids (Li et al., 2009; Mur akami et al.,
2002; Ohigashi et al., 2001).
.
ce and Technology , Nation al Taiwan University , No . 1, Sec. 4, 3; fax: +886 2 2362 0849.
kang, Taipei 11529, Taiwan. -6,7,8,4 0-tetramethoxyflavone; H3K4me1, histone H3 lysine 4 3K14ac, histo ne H3 lysine 14 acetylation; H3K27ac, histone H3
53 response element s; GADD45a, gro wth arrest and DNA-dama ge- oxidase subunit II.
5-h ydr oxy-6,7,8,4 0-tetr ametho xyflavon e (5-HTMF).
J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4 617
Nobiletin and its meta bolites ha ve anti-inflammator y and
anti-tum ourig enic effects in mouse skin, mediated thro ugh
do wn-r egu lation of the inducib le nitric oxide synthase (iNOS)
and cytoc hro me c oxidase subunit II (COX2) genes (Li, Car ey, &
Workman, 2007; Li & Sang, et al., 2007; Li & Pan, et al., 2007;
Pan et al., 2008). Nobiletin also exhibits anti-pr olifer ativ e
activ ity in a human squamous cell carc inoma cell line
(Kandas wami, Perkins, Soloniuk , Drze wieck i, & Middleton,
1991). Furthermo re, nobiletin has been re ported to cause
G2/M arr est and apoptosis in he patocellula r carc inoma cells
(Asamot o et al., 2004 ). Others ha ve sho wn that nobileti n and
tangeretin induce G1 arr est in bre ast and colon cancer cells
(Koropa tnic k, Morle y, & Fer guson, 2007; Pan, Chen, Lin-Shiau,
Ho , & Lin, 2002; Zheng et al., 2002). Suc h inhibition of cell
cycle pro gre ssion was postulated to arise fro m modulation
of the activ ities of se ver al ke y G1 regu lator y pro teins, includ-
ing CDK2 and CDK4, and the CDK re gulators p21 and p27 (Pan
et al., 2002). Recent studies ha ve sho wn that hydrox ylated
PMFs (such as 5-h ydr oxy-6,7,8,4 0-tetr amethoxyflavone (5-
HTMF), a hydr oxylated counterpa rt of tangeretin) (Fig . 1) iso-
lated fro m aged citrus peels ha ve mor e potent anti-canc er
and anti-inflammator y acti vities than their PMF counterpar ts
(Li, Lo , & Ho , 2006; Li & Car ey, et al., 2007; Li & Sang, et al.,
2007; Li & Pan, et al., 2007; Li et al., 2009; Qiu et al., 2010, 2011).
Gene expre ssion can be modified by ep ig enetic mec ha-
nisms, includin g DN A meth ylation, histone modification
and chr omatin remodelling . Suc h modifications at the pro -
moter or gene coding re gions can resul t in chro matin struc-
tur e chan ges and tr anscriptiona l activ ation or inacti vation
of tumour suppr essor genes (Ba ylin & Ohm, 2006; Estelle r,
2007; Lu et al., 2009). Histone modifications, particula rly
meth ylation and acetylatio n, ar e linked to gene activ ation. It
is kno wn that tri-meth ylation of histone H3 lysin e 4
(H3K4me3) and acetylatio n of lysin e 14 (H3K14ac) are associ-
ated with tr anscriptio nall y-acti ve genes. In addition , acetyla-
tion of H3 lysin e 27 (H3K27ac) and mono-meth ylation of
lysin e 4 (H3K4me1) freq uentl y distingui sh acti ve fro m inac-
ti ve enhance r elements (Clayton, Hazzal in, & Mahade van,
2006; Cre yghton et al., 2010; Heintzm an & Ren, 2009;
Heintzm an et al., 2009; Rada-Ig lesias et al., 2011). Thus,
histone modifications that silence tumour suppre ssor genes
ma y be consider ed as potential dia gnostic markers, and drugs
that counter act suc h modifications ma y be candidates for
futur e studies into possib le anti-ca ncer ther apies.
Gr owth arre st and DN A-dam age-inducib le pro tein a
(GADD45a) pla ys an importan t role in cell cycle check point
activ ation, and functions as a tumour suppr essor by inducing
apoptosis via the JNK and p38 pathw ays (Bula vin, Kova lsky ,
Hollander , & Fornace , 2003; Harris et al., 1999). GADD45a gene
expr ession has been re ported to be up-r egu lated by the p53
pathw ay in respo nse to DN A dama ge str ess (An, Kim, &
Roeder , 2004), and do wn-r egu lated in br east cancer as
compar ed to normal br east epi thelium cells (Marks et al.,
2005). Mor eo ve r, GADD4 5a-nu ll mice prese nt with enhanced
genomic insta bility , as well as tumour initiation and pro gre s-
sion (Hollander et al., 2001 ). The in vo lve ment of GADD45 a in
the metab olic pathw ays that reg ulate cancer cell prol iferat ion
ma y make it a suita ble marker for cancer pro gnosis (An et al.,
2004; Melino et al., 2000; Somasunda ram et al., 2008; Zerbini
et al., 2004).
In the prese nt stud y, we demonstr ate that expr ession of
GADD45a is induced by PMFs, and this is possib ly mediate d
thro ugh chan ges in chro matin structur e at GADD45a regu la-
tory /enhan cer regions. Furthermo re , we demonstr ate that
inductio n of GADD45a expre ssion underlies the anti-pr olifer a-
ti ve effects of PMFs on a color ectal cancer cell line .
2. Ma ter ials and methods
2.1. Cell cultur e and reagents
HCT116 cells were culture d in DMEM supplemen ted with 10%
heat-inacti va ted FBS and 2 mM L-glutamine . Cells were main-
tained at 37 C in a 5% CO 2 air -humidified atmosphe re .
2.2. Cell gr owth assa y
HCT116 cells wer e incubated in an xCELLigence system incu-
bator (Roche-Applied-S cience , Indianapolis, IN , USA) with or
without PMFs (including nobiletin, tangeretin and 5-h ydr o-
xy-6,7,8 ,4 0-tetr ametho xyflavo ne (5-HTMF; a hydr oxylated
counte rpart of tangere tin)) for 3–4 da ys. The system monitors
cellular eve nts (including cell num ber , cell via bility and
morpho logy) in rea l time , by meas uring electrical impedan ce
acro ss inter dig itated micr oelectr odes inte gra ted on the
bottom of tissue cultur e E-Plates.
2.3. Western blot analysi s
Cells were washed with phosphate -buffer ed saline (PBS), prior
to lysis in RIP A lysis buffer containing NP-40 and pro tease
inhibitors. Fifty micr ogr ams of prote in were sep ara ted by
10% SDS–PAGE, befor e being transf erre d to a PVDF membr ane
(Immobilon-P , Millipor e, Billerica , MA, USA). The membr ane
was then incubate d with antibod ies ag ainst Gadd45 a (Cell Sig-
naling, Den vers, MA, USA) or a-tubulin (Sigma–Aldric h, St.
Louis, MO , USA).
Table 1 – List of pr imers used in qPCR.
Primer Forwa rd sequences (5 0–3 0) Re ve rse sequenc es (50–30)
GADD45a TTTGC AATATGA CTTTGGA GGA CATCCCCCA CCTT ATCC AT
GAPDH AGCCA CATCGCTC AGA CAC GC CCAATACG ACCAAATCC
P1 TGTGGAA GGT AACGA CATACA GA CAACCATCTG ACACCC
P2 CACG ATTTC CCAGTT ACA C CCCGTTTG TATTC AGGGA T
P3 CAACA CTTCTGA GGTAAACTTTG TCC AGGGTT AATGCA CTG AA
P4 AAAGTG CCGC AACT ATTATAGA GC TTTAGCA GA GGCT AGA G
P5 GCTGA TTTGC ATAGCCC A CTCTC CGGGGTT ATCCT
P6 GCTCTCTC CCTGGGC GA CTC ACCTTTC GGTCTTC TG
P7 CTTCTCT TACCT ACCCCG C TTGC CCTGTG CAAACTT
P8 GGGA CTTCTCA CGGGA C CACCTGGGC TCT ACGAA
P9 TGTGGA TCTGTGGT AGGT CACATCTCGGCTG AAAGT
P10 ACATTAAGA GGATAGA GTG GAA TTT TTCCC TACCT ACCAAAGGTTT
P11 ACACAGCA TGGAAA TCTTTCT TTTGC ATGA GA GT ATC ATTCA CC
P12 CCCAACT ACCTTGAA GAAA CTTT A ACACTTGA ACATTCT ATATTTTGG C
P13 GGTGT AGAAA GA TAAGGGA GC AGGGT AACA CAGAA TGCC
618 J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4
2.4. shRNA knockdown
HCT116 cells were tr ansfected with lenti vira l ve ctors contain-
ing shRN A tar geting either GADD45a, or GFP as a nega tiv e
contr ol. Vectors were acquir ed from the National RN Ai core
facility platform at Academia Sinica, Taiw an. Contr ol or GAD-
D45a knoc kdo wn cells were treate d with DMSO- or tangeretin
(50 lM) for 16 h, befor e being collected for use in the cell
gr owth assa y.
2.5. Quantitative PCR analysi s
RN Awas extr acted using Trizol (Invitr oge n, Carlsbad, CA, USA),
and re ve rse tran scribed using In vitro gen Superscrip t III. The
resulting cDN A was then subject to RNase H digestion.
Quanti tativ e PCR was subsequ entl y perform ed using the Real
Time read y Human Cell Cycle Re gulation Panel (Roche-
Applied -Science , Indianapoli s, IN , USA), follo wing the
man ufactur er’s pro tocol.
2.6. Chr omatin immunopr ecipitation (ChIP)
ChIP was performed as describe d pre viousl y (Cheng et al.,
2011). Briefly, cells were harv ested by tr ypsinisation and
resuspend ed in PBS solutio n (containing 20 mM sodium
butyr ate) at room temper ature . Cells were gentl y mixed
with 1% formaldeh yde to cros s-link DN A and pro teins.
The cros s-linking reacti on was terminated by the addition
of gly cine to a final concentr ation of 125 mM. Cells were
mixed for a further 5 min at room temper atur e, and were
then washed twice with PBS . Cells were subsequ entl y lysed
with ly sis buffer consisting of 50 mM Tris–HCl, pH 8, 10 mM
eth ylenedi aminetetr aacetic acid (EDTA), 1% SDS , broa d-
rang e pro tease inhibitor cock tail, 1 mM PMSF and 20 mM
butyr ate . Chr omatin was fra gmented by sonication using a
MISONIX Sonicato r 3000 (MISONIX, Farmingd ale, NY , USA).
After sonication , the cell lysa te was centrifuged at 10,000 g
for 10 min at 4 C, and the supernat ant was collected.
DN A–protein complexes wer e imm unopr ecipitate d using
antibod ies ag ainst the indicated pro tein, and were then
subjecte d to reverse cros s-linking. The imm unopr ecipited
DNA was reco ve red using a PCR purification kit (Qiagen,
Valencia, CA, USA) accor ding to the manu factur er’s instruc-
tions. Antibodi es ag ainst H3K4me1, H3K4me3, H3K14ac ,
H3K27ac and H2A.Z were purc hased fro m Abcam (Cam-
bridg e, MA, USA). CBP antibod y was purc hased fro m Santa
Cruz Biotec hnology (Santa Cruz, CA, USA). Quantitati ve
PCR was used to determine the occupanc y of eac h mark
at reg ulatory regions of GADD45a, using the primer pairs
listed in Tab le 1. Eac h experim ent was performed twice .
For eac h sample , PCR was performed in triplicate , and the
bound fra ction was normalise d to the input (IP/Input). To
determine fold enric hment, the data were further norma-
lised to a non-sp ecific binding contro l.
2.7. Statistical analysi s
Data wer e analy sed by One-w ay or Two- way ANO VA. A p va lue
of <0.05 was conside red to be statistical ly significant.
3. Results
3.1. PMFs inhibit pr olifer ation of human epithelial color ectal cancer cells, and alter expr ession of cell cycle genes
In or der to determine the effects of PMFs on color ectal cancer
cell gro wth, we treate d HCT116 cells with the PMFs nobiletin,
tange retin and 5-HTMF . After 18 h of tr eatment with an y of
these compou nds, the cell index was redu ced as compar ed
to the contr ol (Fig. 2A). The effect of 5-HTMF on the cell index
(79% redu ction of cell via bility) was, ho we ve r, mu ch str ong er
than that of nobiletin and tangeretin (33% and 26% reduction,
respe ctiv el y) (Fig . 2A).
To identify chan ges in gene expr ession that ma y underlie
the observ ed effects on the cell index, we pro filed the
express ion of 84 genes in vo lve d in cell cycle regu lation
(Supplemen tary data Fig . S1). These included genes in vo lve d
in check point contr ol, as well as cell cycle phase reg ulators,
cyclins, CDKs and cell cycle inhibitors. As sho wn in Fig . 1B
and C, a num ber of important tr anscriptio n factors and cell cy-
cle reg ulators were affected by 10 and 18 h PMF trea tment. Both
nobiletin (30 M) and tangeretin (50 M) gre atly up-re gulated
Fig. 2 – Pol ymetho xyflavones ex ert an inhibitory effect on gro wth and differ entiall y affect gene expr ession. (A) Normalised
cell index of HCT116 follo wing 18 h of tr ea tment with nobiletin (NOB), tanger etin (TAN) or 5-HTMF . Eac h tr ea tment
significantl y reduced cell index compar ed to a DMSO ve hicle contr ol (*p < 0.05). Th e cell index was significantl y less for cells
tr ea ted with 5-HTMF , as compar ed to tha t of cells tr ea ted with nobileti n or tanger etin (#p < 0.05). One-wa y ANO VA was
perform ed to determ ine sta tistical significance, and err or bars repr esent the standa rd erro r of the mean of 2–3 indi vidual
samples. (B and C) Th e expr ession of in volv ed in cell cycle regula tion wer e anal ysed in HCT116 cells tr eate d with nobiletin
(NOB, gr ey box), tanger etin (TAN , blac k box ) or 5-HTMF (light gr ey box) for (B) 10 and (C) 18 h. The expr ession lev el of each
gene was normalis ed to tha t in the DMSO contr ol, and GAPDH was used as an internal contr ol.
J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4 619
expre ssion of GADD45a and CHEK1 (a cell cycle regu lator)
(Fig . 2B and C), while inducing a smaller incre ase in p53
(Supplem entar y data Fig. S1). Con ve rsely , most of the genes
up-re gulated by nobiletin and tangere tin were either do wn-
regu lated or unaffecte d by 5-HTMF treatm ent. Thus, nobiletin
and tangeretin possib ly affect cell gro wth thro ugh a similar
mec hanism, wher eas 5-HTMF ma y act thro ugh a sepa rate
pro cess.
3.2. Nobiletin and tanger etin cause changes in chr omatin str uctur e at regulator y regions of GADD45 a
Expr ession of GADD45a is rapi dly induced by DN A dama ge
(Zerbini, Wang, Corr ea, Cho , & Liberm ann, 2005; Zhan, Chen,
Antinor e, & Fornace , 1998), and suc h induction is de pendent
on the modification of histones at p53 respo nse elemen ts
(p53RE) within GADD45a (An et al., 2004; Gamper & Roeder ,
2008). We thus asked whether nobiletin and tangeretin
up-re gulate GADD45a expre ssion throu gh p53-de pendent
chr omatin remodelling at GADD45a reg ulatory regions
(Fig . 3A). We first measur ed p53 occupanc y at these regions
by chr omatin imm unopre cipitation- quantitati ve PCR (ChIP-
qPCR). No significant cha ng es in p53 occupanc y were detected
follo wing trea tment with nobileti n or tangeretin (data not
sho wn) consistent with our observ ation that p53 express ion
was onl y mildl y affected by these PMFs (Supplemen tary data
Fig . S1). Next, we measur ed the le ve l of tw o histone modifica-
tions, H3K4m e3 and H3K14ac , both of which ar e associated
with tr anscriptiona lly-a cti ve genes (Li & Care y, et al., 2007; Li
& Sang, et al., 2007; Li & Pan, et al., 2007 ). No appare nt chan ge
in H3K4me3 was observ ed follo wing PMF tr eatment (Fig . 3B).
The le ve l of H3K14ac at intro n 1 (P6–P8) was incre ased follo w-
ing treatm ent with tangeretin , but onl y a mild incr ease was
observ ed at the p53RE sequenc e (P9–P12; Fi g. 3C). As suc h,
we suggest that histone acetylatio n at a putati ve regu lator y re-
gion near intro n 1 is associate d with PMF-m ediated induction
of GADD45a, with p53RE pla ying a minor role .
3.3. Tanger etin-mediated up-r egulation of GADD45 a is associated with active enhancer mar ks and chr omatin remodelling at intr on 1
We speculate d that PMF-media ted GADD45a inductio n might
also in vo lve chan ges in chr omatin structur e at enhancer ele-
ments. As H3K4m e1 and H3K27ac ha ve been report ed to mark
suc h elemen ts (Cre yghton et al., 2010; Heintzman & Ren,
2009; Heintzman et al., 2009; Rada-Igles ias et al., 2011 ), we
used ChIP-qPCR to examine occupanc y of these marks
Fig. 3 – Th e chro mati n structur e of intr on 1 of GADD45a is modula ted by PMFs, suggest ing the pr esence of a no vel regula tory
region. (A) Map of GADD45a gene structur e. Th e loca tions of primers used for ChIP-qPC R ar e indica ted. These pr imers wer e
used to determine occupanc y of (B) H3K14ac, (C) H3K27ac, (D) H3K4m e3, (E) H3K4me1 and (F) IgG. Cells wer e tr ea ted with
DMSO (striped bars), nobiletin (white bars) or tanger etin (black bars).
620 J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4
thro ughout GADD45a. H3K4me1 appear ed to be decr eased at
most loci by treatm ent with nobiletin or tangeretin (Fig . 3E),
while the le vel of H3K27ac ar ound intro n 1 was incr eased
follo wing tangereti n tr eatment (Fig . 3D). H3K27 is acetylate d
by CBP (Tie et al., 2009 ), and as suc h, we addr essed whether
the incre ase in H3K27ac arises fro m recru itment of CBP to
Fig. 4 – A tanger etin-r esponsi ve enhancer element within Gad d45 a is loca ted near intr on 1. ChIP-qPCR was used to determine
the occupanc y of (A) CBP and (B) H2A.Z, using the pr imers indica ted in Fig. 3A.
Fig. 5 – GADD4 5a knoc kdo wn antagonise s tanger etin-medi ated cell gro wth suppr ession. HCT116 cells wer e infected with
lenti vir al vectors contain ing shRN A agains t either GADD4 5a or GFP . Cells wer e subseque ntl y tr ea ted with DMSO or tanger etin
(50 M). (A) Le ve l of GADD45a mRN A, as determ ined by RT -PCR. (B) Le ve l of Gad d45 a pro tein, as determine d by Western blot of
whole cell lysa te. (C) Stac ked columns illustr ate the normalis ed cell index of GFP- or GADD4 5a-knoc kdo wn cells, follo wing
tr ea tment with DMSO or tanger etin. Error bars repr esent the standar d erro r of the mean of 2–3 indi vidual samples. *p < 0.05.
J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4 621
intro n 1 (P6). CBP was indeed increa sed at this re gion upon
tangeretin tr eatment (Fig. 4A). Thus, tangeretin ma y induce
chr omatin remodeling by facilitating recru itment of CBP . In
addition, the histone variant H2A.Z pla ys an essential role
in the response of enhance rs to ph ysiological and en viron -
mental stimu li (Gaudr eau et al., 2009; Wigge & Kumar , 2010 ),
622 J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4
and thus we pre dicted that it ma y be incorpor ated into the
pro moter re gions of GADD45a upon PMF treatm ent . We found
that H2A.Z was incre ased at P6 upon tangeretin trea tment,
but not follo wing nobiletin trea tment (Fig . 4B). This ma y ac-
count for the gre ater induction of GADD45a by tangeretin as
compar ed to nobiletin.
3.4. GADD45 a pla ys a centr al role in tanger etin-mediated cancer cell gr owth inhibition and pr olifer ation
Finall y, we in ve stigated whether the incr ease in GADD4 5a
expr ession induce d by tangeretin contributed to the observ ed
gr owth suppr ession of HCT116. Cells were tr ansfected with
shRN A dire cted ag ainst GADD45a. Both GADD45a RN A
(Fig. 5A) and pro tein (Fig . 5B) were reduced by 50% at 48 h
post-tr ansfection , as compar ed to cells transf ected with
shRN A against GFP . Follo wing tr ansfection with either sh GAD-
D45a or shGFP , cells were exposed to tangeretin for 18 h.
Silencing GADD45a significantly incr eased the cell index, as
compar ed to the shRN A GFP contr ol (Fig. 5C). This result sup-
ports a dire ct role for GADD45a induction by tangereti n in cell
gr owth suppr ession.
4. Discussion
In the pre sent stud y, we demonstr ate that GADD45a expr es-
sion is induced by the PMFs nobiletin and tangereti n. Such
up-r eg ulation is associate d with chr omatin remodeli ng at a
no vel putati ve re gulator y element within intr on 1 of GAD-
D45a, as indicate d by an incr ease in the occupanc y of
H3K27ac and H2A.Z. The incre ase in the enhance r mark
H3K27ac was accompani ed by an equi va lent incre ase in CBP ,
the acetyltr ansfera se responsib le for this modification.
Nobiletin did not affect H3K27ac , perhaps accounti ng for
the stronger effect of tangereti n on GADD45a expr ession.
H3K4m e1 was not incr eased by tangereti n; this is perhaps
in ag reeme nt with the recen t finding that H3K27ac , but not
H3K4m e1, is a signatur e of acti ve enhance rs (Jae nisc h et al.,
2010; Wysoc ka et al., 2011). Furthermo re, tw o findings suggest
that PMF-media ted up-r eg ulation of GADD45a ma y be inde-
pendent of p53: chro matin remodelling does not occur at
the p53RE, and p53 occupanc y is unaffe cted by PMF
tr eatment.
Gadd4 5a has recentl y been re ported to pla y a role in the
pro motion of genome sta bility by DN A-demeth ylation (Jin,
Guo , & Pfeifer , 2008; Ma ye r et al., 2009; Niehrs et al., 2007).
GADD45a knock do wn results in DN A hypermeth ylation, and
the Gadd45 a pro tein is recruited to sites of demeth ylation
via an inter action with a DN A re pair endon uclease , XPG (Bar -
reto et al., 2007). In addition to the prese nted results , we ob-
serv ed that expre ssion of XPG was also enhanced during
tangere tin-mediate d GADD45a inductio n (data not sho wn).
Hence , it remains possib le that GADD45a‘s role in PMF-m edi-
ated contro l of cell pro lifer ation ma y in vo lv e DN A
demeth ylation.
Se ve ral studies ha ve indicate d that 5-h ydrox y PMFs ha ve
muc h str ong er inhibitor y effects on the gro wth of colon can-
cer cells as compar ed to their PMF counterparts (Li et al., 2009;
Qiu et al.,2010, 2011). In ag reement with these earlier findings,
we observ ed that the normali sed cell index of HCT116 cells
treate d with 5-HTMF was significantly reduced as compar ed
to that of cells tr eated with nobile tin or tangeretin . Ho weve r,
5-HTMF tr eatment does not induce GADD45a expre ssion, indi-
cating that 5-HTMF affects cell gro wth thro ugh differ ent
mech anisms to nobiletin and tangeretin (Supplementa ry data
Fig. S1). We observ ed that 5-HTMF trea tment resulte d in de-
crease d expr ession of genes in vo lve d in DN A dama ge-signal-
ling (including ATM , ATR , and BRC A1 ) and rep air (including
BRC A2 and RAD51). It is possib le that the reduced transc rip-
tion of these genes ma y contrib ute to the lo wer survi va l of
cells tr eated with 5-HTMF , but a compr ehensi ve under stand-
ing of the mec hanisms in volv ed awaits further in ve stigation.
Pre vious studies ha ve also demon stra ted that 5-HTMF can
cause cell cycle arr est and apoptosis thro ugh up-r egu lation of
p53, p21 and Bax pro teins and do wn-re gulation of CDK-2 and
CDK-4 acti vity (Qiu et al., 2010, 2011 ). Ho we ve r, in the curr ent
stud y we did not observ e an y effect of 5-HTMF on the le vel of
mRNA encoding these pro teins (Supplemen tary data Fi g. S1).
We postulate that 5-HTMF tr eatment ma y instead regu late
these factors at a post-tr ansla tional le ve l. Fi nall y, the rela-
tiv el y high lipophilicity and permea bility of 5-HTMF as com-
pare d to nobilein and tangereti n ma y also promo te its
uptake into the cell and hinder cell gro wth (Li et al., 2009;
Mur akami et al., 2002; Ohigashi et al., 2001; Qui et al., 2010).
In summar y, our data suggest that nobiletin and tangere tin
ma y affect cell gro wth throu gh a similar mech anism, wher eas
gro wth inhibition by 5-HTMF ma y be mediated thro ugh a sep-
arate proce ss, possib ly in vo lving the modula tion of signalling
prote ins rela ted to cell pro lifer ation and apoptosis (Qiu et al.,
2010).
6. Conclusion
We ha ve identified GADD45a, a putati ve tumour suppr essor ,
as a no ve l tar get of nobiletin and tangeretin . Up-r egu lation
of GADD4 5a contribut es to the anti-pr olifer ativ e effects of
these PMFs on color ectal cancer cells. More over , we ha ve
identified a no vel re gulator y re gion within intro n 1 of GAD-
D45a, at whic h H3K27ac and H2A.Z are enric hed in respo nse
to nobiletin and tangeretin tr eatment. Consequ ently , expre s-
sion le vel s of GADD45a ma y serv e as a potential dia gnostic
marker for color ectal cancer , and as a possib le tar get for fu-
tur e studies into anti-cancer ther apies.
Ackno wledgements
The HCT116 cell line was kindl y pro vided by Dr . Han-Chun g
Wu (Academia Sinica, Taiw an). This resear ch was supporte d
by the National Science Council, Taiw an, under Gr ant NSC
100-2313-B-0 02-021 and by internal support fro m the College
of Bior esour ces and Agricultur e, National Taiw an Univ ersity
and Academia Sinica, Taipei, Taiw an.
Appendix A. Supplementary data
Supplementa ry data associated with this article can be found,
in the online ve rsion, at http://dx .doi.or g/10.1016/ j.jff.2013.
R E F E R E N C E S
An, W., Kim, J., & Roeder , R. G. (2004). Or dered cooperati ve functions of PRMT1, p300 and CARM1 in transcriptional activ ation by p53. Cell, 117 , 735–748.
Asamoto, M., Ohn ishi, H., Tujimur a, K., Hoka iwa do , N., Takahashi, S., Og awa, K., Kurib ayashi, M., Og iso , T., Okuyama, H., & Shirai, T. (2004). Inhibition of cell prolifer ation by nobil etin, a dietary ph ytochemical, associa ted with apoptosis and char acteristic gene expression, but lack of effect on early rat hepatocarcinogenesis in viv o. Cancer Science, 95 , 936–942.
Barreto , G., Schafer , A., Marhold, J., Stach, D., Swaminathan, S. K., Handa, V., Doderlein, G., Maltry , N., Wu, W., Lyko , F., & Niehrs, C. (2007). Gadd 45a promotes epigenetic gene activ ation by re pair -mediated DNA dem eth ylat ion. Nature , 445, 671–675.
Ba ylin, S. B., & Ohm, J. E. (2006). Epig enetic gene silencing in cancer – A mechanism for early oncog enic pathwa y addiction? Nature Reviews Cancer , 6, 107–116.
Bula vin, D. V., Kovalsk y, O., Hollander , M. C., & Fornace, A. J. Jr ., (2003). Loss of oncog enic H-ras-induced cell cycle arrest and p38 mitog en-activ ated protein kinase activ ation by disruption of Gadd45a. Molecular and Cellular Biology , 23 , 3859–3871.
Cheng, P. Y., Lin, Y. P., Chen, Y. L., Lee, Y. C., Tai, C. C., Wang, Y. T., Chen, Y. J., Kao , C. F., & Yu, J. (2011). Interplay between SIN3A and ST AT3 med iates chr omatin conformat ional changes and GF AP expression during cellular differentiation. PLoS One, 6, e22018.
Cla yton, A. L., Hazzalin , C. A., & Mahade van, L. C. (2006). Enhanced histone acetylation and transcription: A dynamic perspectiv e. Molecular Cell, 23 , 289–296.
Cre yghton, M. P., Cheng, A. W., Welstead, G. G., Kooistra, T., Care y, B. W., Stei ne, E. J., Hanna, J., Loda to , M. A., Fr ampto n, G. M., Sharp , P. A., Bo yer , L. A., Young, R. A., & Jaenisc h, R. (2010). Histone H3K27ac separa tes activ e from poised enhancers and predicts de velopmental state. Proceedings of the Nationa l Academy of Sciences of the United Stat es of Ameri ca, 107 , 21931–21936.
Esteller , M. (2007). Epig enetics pro vides a ne w generation of oncog enes and tumour -suppressor genes. British Journal of Cancer , 96 (Suppl.), R26–R30.
Gamper , A. M., & Roeder , R. G. (2008). Multiva lent binding of p53 to the ST AGA complex mediates coactiv ator recruitment after UV damage. Molecular and Cellu lar Biology , 28 , 2517–2527.
Gaudreau, L., Ge vry , N., Har dy, S., Jacques, P. E., Laflamme, L., Svotelis, A., & Robert, F. (2009). Histone H2A.Z is essential for estrogen rece ptor signaling. Genes & Developme nt, 23 , 1522–1533.
Harris, C. C., Wang, X. W., Zhan, Q. M., Course n, J. D., Khan, M. A., Kontn y, H. U., Yu, L. J., Hollander , M. C., O’Conn or , P. M., & Fornace, A. J. (1999). GADD45 induction of a G(2)/M cell cycle chec kpoin t. Proceedings of the National Academy of Sciences of the United States of America , 96 , 3706–3711.
Heintzman, N. D., Hon, G. C., Ha wkins, R. D., Kheradpour , P., Stark, A., Harp , L. F., Ye, Z., Lee, L. K., Stuart, R. K., Ching, C. W., Ching, K. A., Antosie wicz-Bo ur get, J. E., Li u, H., Zhang, X., Gr een, R. D., Lobanenkov , V. V., Ste wart, R., Thomson, J. A., Cra wfor d, G. E., Kellis, M., & Ren, B. (2009). Histone modifications at human enhancers reflect globa l cell-type-spec ific gene expression. Nature , 459 , 108–112.
Heintzman, N. D., & Ren, B. (2009). Finding distal re gulatory elements in the human genome. Current Opinion in Genetics & Development , 19 , 541–549.
Hollander , M. C., Kovalsk y, O., Salvador , J. M., Kim, K. E., Patterson, A. D., Raines, D. C., & Fornace, A. J. (2001). Dimeth ylbenz anthracene carcinogenesis in Gadd 45a-null mice is associated with decreased DNA re pair and increased mutation frequenc y. Cancer Resear ch, 61 , 2487–2491.
Jaenisc h, R., Cre yghton, M. P., Cheng, A. W., Welstead, G. G., Kooistra, T., Care y, B. W., Steine, E. J., Hanna, J., Loda to , M. A., Fr ampton, G. M., Sharp , P. A., Bo yer , L. A., & Young , R. A. (2010). Histone H3K27ac separ ates activ e from poised enhancers and predicts de velopmental state. Proceedings of the National Academy of Sciences of the United States of America , 107 , 21931–21936.
Jin, S. G., Guo , C., & Pfeifer , G. P. (2008). GADD45A does not promote DNA dem eth ylat ion. Plos Genetics, 4, e1000013.
Kandas wami, C., Perkins, E., Soloniuk, D. S., Drze wiecki, G., & Middleton, E. (1991). Antiprolifer ativ e effects of citrus flavonoids on a human squa mous-cell carcinoma in vitro . Cancer Letters , 56 , 147–152.
Koropatnic k, J., Morle y, K. L., & Fer guson, P. J. (2007). Tang eretin and nobiletin induce G1 cell cycle arrest but not apoptosi s in human breast and colon cancer cells. Cancer Letters , 251 , 168–178.
Li, B., Care y, M., & Workman, J. L. (2007). The role of chr omatin during tr ansc ription. Cell, 128 , 707–719.
Li, S., Lo , C. Y., & Ho , C. T. (2006). Hydroxylated polymethoxyflavones and meth ylated flavonoids in sweet orange (Citrus sinensis ) peel. Journal of Agricultur al and Food Chemistry , 54 , 4176–4185.
Li, S., Pan, M. H., Lai, C. S., Lo , C. Y., Dushe nko v, S., & Ho , C. T. (2007). Isolation and syntheses of polymethoxyflavones and hydroxylatedpoly methox yflavones as inhibitors of HL-60 cell lines. Bioor ganic & Medicinal Chemistry , 15 , 3381–3389.
Li, S., Pan, M.-H., Lo , C.-Y ., Tan, D., Wanga, Y., Shahidi, F., & Co , C.- H. (2009). Chemistry and health effect s of polymethoxyflavonesand hydroxylated polymethoxyflavones. Journal of Functional Foods, 1, 2–12.
Li, B., Sang, S., Pan, M. H., Lai, C. S., Lo , C. Y., Yang, C. S., & Ho, C. T. (2007). Anti-inflammatory property of the urinary metabolites of nobil etin in mouse. Bioorg anic & Medicinal Chemistry Letters , 17, 5177–5181.
Lu, T. Y., Kao , C. F., Lin, C. T., Huang, D. Y., Chiu, C. Y., Huang, Y. S., & Wu, H. C. (2009). DNA meth ylation and histo ne modification re gulate silencing of OPG duri ng tumor progre ssi on. Journa l of Cellular Bioc hemistry , 108 , 315–325.
Marks, J. R., Wang, W., Huper , G., Guo , Y. Q., Murph y, S. K., & Olson, J. A. (2005). Analysis of meth ylat ion-sensitive transcriptome identifies GADD45a as a frequently meth ylat ed gene in breast cancer . Oncogene, 24 , 2705–2714.
Ma yer , C., Schmitz, K. M., Schmitt, N., Hoffmann-Roh re r, U., Schafer , A., & Grum mt, I. (2009). TAF1 2 recruits Gadd 45a and the nucleotide exci sion re pair complex to the promoter of rRNA genes leading to activ e DNA demeth ylation. Mole cular Cell, 33, 344–353.
Melino , G., Le vrer o, M., De Laurenzi, V., Costan zo , A., Sabatini, S., Gong, J., & Wang, J. Y. J. (2000). The p53/p63/p73 family of transcription factors: Overlapping and distinct functions. Journal of Cell Science, 113 , 1661–1670.
Murakami, A., Koshim izu, K., Ohigashi, H., Kuwahara , S., Kuki, W., Takahashi, Y., Hosota ni, K., Kaw ahara, S., & Matsuoka, Y. (2002). Characteristic rat tissu e accu mulation of nobiletin, a chemopre venti ve polymethoxyflavonoid, in comparison with luteolin. Biofactor s, 16 , 73–82.
Niehrs, C., Ba rreto , G., Schafer , A., Marhold, J., Stach, D., Swaminathan, S. K., Handa, V., Doderlein , G., Maltry , N., Wu, W., & Lyko , F. (2007). Gadd 45a promotes ep ig enetic gene activ ation by re pair -mediat ed DNA demeth ylation. Nature, 445 , 671–675.
Ohigashi, H., Murakami, A., Kuwahara , S., Takahashi, Y., Ito , C., Furuka wa, H., Ju-Ic hi, M., & Koshimizu, K. (2001). In vitro absorption and metabolism of nobiletin, a chemopr eventi ve polymethoxyflavonoid in citrus fruits. Bioscience Biotec hnology and Bioc hemistry , 65 , 194–197.
624 J O U R N A L O F F U N C T I O N A L F O O D S 5 ( 2 0 1 3 ) 6 1 6 – 6 2 4
Pan, M. H., Chen, W. J., Lin-Shiau, S. Y., Ho , C. T., & Lin, J. K. (2002). Tang eretin induces cell-cy cle G1 arr est through inhibiting cyclin- dependent kina ses 2 and 4 act ivities as well as elev ating Cdk inhibitors p21 and p27 in hu man colorectal carcinoma cells. Car cinogenes is, 23 , 1677–1684.
Pan, M. H., Lai, C. S., Li, S. M., Chai, C. Y., Lo , C. Y., Dushe nko v, S., Ho , C. T., & Wang, Y. J. (2008). Anti-inflammatory and antitumor promotional effects of a no vel urinary metabolite , 30,4 0-didemeth ylnobilet in, deriv ed from nobil etin. Car cinogenes is, 29 , 2415–2424.
Qiu, P., Dong, P., Guan, H., Li, S., Ho , C. T., Pan, M. H., McClements, D. J., & Xiao , H. (2010). Inhi bitory effect s of 5-h ydroxy polymethoxyflavones on colon cancer cells. Molecular Nutrition & Food Resear ch, 54 (Suppl. 2), S244–S252.
Qiu, P., Guan, H., Dong, P., Li, S., Ho , C. T., Pan, M. H., McClements, D. J., & Xiao , H. (2011). The p53-, Ba x- and p21-dependent inhibition of co lon cancer cell gr owth by 5-h ydroxy polymethoxyflavones. Molecular Nutrition & Food Resea rch, 55, 613–622.
Rada-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S. A., Flynn, R. A., & Wysocka, J. (2011). A unique chr omatin signature unco vers early de velopmental enha ncers in hu mans. Nature, 470, 279–283.
Somasundaram, K., Redd y, S. P., Britto , R., Vinnako ta, K., Aparna, H., Sree pathi, H. K., Tho ta, B., Kumari, A., Shilpa, B. M., Vrinda , M., Umesh, S., Samuel, C., Shetty , M., Tando n, A., Pande y, P., Hegde , S., Hegde , A. S., Balasubramaniam, A., Chandramouli, B. A., Santosh, V., Konda iah, P., & Rao, M. R. S. (2008). No vel glioblastoma markers with diagnostic and prognostic value identified through transcriptome analysis. Clini cal Cancer Resear ch, 14 , 2978–2987.
Tie, F., Banerjee, R., Stratton, C. A., Prasad-Sinha, J., Stepanik, V., Zlobi n, A., Diaz, M. O., Scacheri, P. C., & Harte, P. J. (2009). CBP- med iated ace tylation of histone H3 lysine 27 antagonizes Drosophila Polycomb silencing. Development , 136 , 3131–3141.
Wigg e, P. A., & Kum ar , S. V. (2010). H2A.Z- containing nucleosomes med iate the thermosensory response in ara bid opsis. Cell, 140 , 136–147.
Wysocka, J., Rad a-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S. A., & Flynn, R. A. (2011). A unique chr omatin signature un co vers early de velopmental enhancers in humans. Nature , 470, 279–283.
Ze rbini, L. F., Wang, Y., Correa, R. G., Cho , J. Y., & Libermann, T. A. (2005). Blocka ge of NF-kappaB induce s serine 15 phosp horylation of mutant p53 by JNK kinase in prostate cancer cells. Cell Cycle, 4, 1247–1253.
Ze rbini, L. F., Wang, Y. H., Czibere , A., Correa, R. G., Cho , J. Y., Ijiri, K., Wei, W. J., Jose ph, M., Gu, X. S., Grall, F., Goldring, M. B., Zh ou, J. R., & Libermann, T. A. (2004). NF-kap pa B- mediated re pression of gro wth arrest- and DNA-dama ge-inducible proteins 45 alph a and -gamma is essential for cancer cell surviva l (vol 101, pg. 13618, 2004). Proceedings of the National Academy of Science s of the Unit ed States of America, 101 , 15271.
Zh an, Q., Chen, I. T., Antinore, M. J., & Fo rnace, A. J. Jr ., (1998). Tumor suppressor p53 can partici pate in transcriptional induct ion of the GADD45 promoter in the absence of direct DNA binding. Molecular and Cellular Biology , 18 , 2768–2778.
Zh eng, Q., Hirose , Y., Yoshim i, N., Murakami, A., Koshim izu, K., Ohigashi, H., Sakata , K., Matsumoto , Y., Say ama, Y., & Mori, H. (2002). Furt her in vestigation of the modifying effect of various chemopre venti ve agents on apoptosis and cell prolifera ti on in hu man colon cancer cells. Journa l of Cancer Resear ch and Clinical Oncology , 128 , 539–546.
Induction of GADD45α expression contributes to t
1 Introduction
2.2 Cell growth assay
2.3 Western blot analysis
3 Results
3.1 PMFs inhibit proliferation of human epithelial colorectal cancer cells, and alter expression of cell cycle genes
3.2 Nobiletin and tangeretin cause changes in ch
3.3 Tangeretin-mediated up-regulation of GADD45α
3.4 GADD45α plays a central role in tangeretin-m
4 Discussion
6 Conclusion