Structural activity relationship of avonoids with
estrogen-related receptor gammaZhiyan Huang1, Fang Fang1, Junjian
Wang, Chi-Wai Wong*Guangzhou Institute of Biomedicine and Health,
Chinese Academy of Sciences, Guangzhou Science City 510663,
Chinaarti cle i nfoArticle history:Received 28 September
2009Revised 28 October 2009Accepted 7 November 2009Available online
13 November 2009Edited by Robert
BaroukiKeywords:FlavonoidERRcPGC-1aabstractIn this study, we
investigated the structural activity relationships (SARs) of
avonoids with estro-gen-relatedreceptorc(ERRc)
anditscoactivatorperoxisomeproliferators-activatedreceptorccoactivator-1a
(PGC-1a). Isoavones (genistein and daidzein) and avanones
(naringenin and hes-peretin) did not signicantly affect ERRc
activity. Flavone apigenin and avonol kaempferol directlysuppressed
the interaction between ERRc and PGC-1a functioning as inverse
agonists. In contrast,avone luteolin suppressed PGC-1a activity in
part through promoting the degradation of
PGC-1aleadingtosuppressedERRcactivity. Therefore, theseavones
andavonols utilizealternativemechanisms to inuence the
transcriptional activities of ERRc and PGC-1a.Structured
summary:MINT-7297297,
MINT-7297435:PGC-1alphaNR2(uniprotkb:Q9UBK2)binds(MI:0407)toERRgamma(uniprotkb:P62508)
by surface plasmon resonance (MI:0107) 2009 Federation of European
Biochemical Societies. Published by Elsevier B.V. All rights
reserved.1. IntroductionFlavonoids are natural polyphenolic
products that can be foundinavarietyof plants andfruits [1].
Flavonoidscanbefurtherclassied into avones (2-phenylchromen-4-one),
avonols(3-hydroxy-2-phenylchromen-4-one), avanones
(2,3-dihydro-2-phenylchromen-4-one), and isoavones
(3-phenylchromen-4-one) (Fig. 1). Flavones such as apigenin and
luteolin can be foundinceleryandgreenpepper,
whileavanonessuchasnaringeninand hesperetin are abundant in citrus
fruits. Tea is a good sourceof avonols such as kaempferol and
quercetin; whereas,
soybeanisloadedwithisoavonessuchasgenisteinanddaidzein. Sincemany
of these avonoids are anti-oxidants, high levels of
dietaryconsumption are believed to confer a multitude of health
benetsincluding reducing oxidative damages [2].The mechanisms of
action behind certain benecial effects ex-tend beyond just
antioxidant activity. Estrogen receptors a andb(ERa and ERb) are
transcription factors that bind to specic hor-mone response
elements located near their target genes and regu-late their
expression in a ligand-dependent manner. Genistein andother
avonoids are phytoestrogens, i.e. they function as
selectiveestrogen receptor modulators (SERMs) [3]. It is
hypothesized
thattheseavonoidsmodulatetheendogenousactivitiesofestrogenreceptorstoslowdownor
prevent thedevelopmentsof breastand ovarian cancers
[4].Estrogen-relatedreceptors aand c(ERRaandERRc)togetherwith their
coactivator peroxisome proliferators-activated
receptorccoactivator-1a(PGC-1a) are keyregulators of
mitochondrialfunction [5]. ERRs share high degrees of similarity in
the DNA bind-ing domain (DBD) with ERs. On the other hand, the
ligand bindingdomains (LBDs) of ERRs are divergent from that of
ERs. AlthoughERRs are constitutively active transcription factors,
we previouslyestablished that kaempferol can function as an ERRa
and ERRc in-verse agonist [6]. In this study, we investigated into
the differentialabilities of avones, avonols, avanones, and
isoavones toregulate the activities of ERRc and PGC-1a by
reporter-based assays in HeLa cells and direct interaction assay by
surfaceplasmon resonance.2. Materials and methods2.1.
Reagents17b-Estradiol, 4-hydroxytamoxifen, daidzein, apigenin,
andgenistein were obtained from Sigma. Hesperetin, kaempferol,
lute-olin, naringenin, and quercetin were purchased from shanxi
HuikeBotanical Development Co., Ltd. DY-131 was obtained from
Tocris.BIAcore 3000 system, certied SA sensor chips, and amine
couplingreagents were obtained from BIAcore, Inc.0014-5793/$36.00
2009 Federation of European Biochemical Societies. Published by
Elsevier B.V. All rights
reserved.doi:10.1016/j.febslet.2009.11.026Abbreviations: ERRc,
estrogen-related receptor c; PGC-1a, peroxisome
prolifer-ators-activated receptor c coactivator-1a; SAR, structural
activity relationship*Corresponding author. Fax: +86 20
32290606.E-mail address: [email protected] (C.-W. Wong).1These
authors contributed equally to this work.FEBS Letters 584 (2010)
2226j our nal homepage: www. FEBSLet t er s. or g2.2. Transfection
and luciferase assaysA Gal4-DBD-PGC-1a fusion was cloned as
described [7]. Trans-fection of HeLa cells by expression and
reporter vectors as well asluciferase assays were performed as
described [6]. All
transfectionswereperformedatleastthreetimeswiththemeanandSEMofthese
experiments shown in Figs. 2 and 4.2.3. Interaction between ERRc
and PGC-1a analyzed by surfaceplasmon resonance assayHuman ERRc-LBD
protein was expressed and puried asdescribed [6]. Interactions
between ERRc-LBD and PGC-1acoactivator peptides in the presence of
various compoundswere analyzed using a BIAcore 3000 system.
Biotinylated PGC-1apeptides: EEPSLAKKAALAPAN (NR2 negative
control); ENE-ANLLAVLTETLD(NR1); EEPSLLKKLLLAPAN(NR2);
andRPCSELLK-YLTTNDD(NR3) were captured onto streptavidin
immobilizedsurface of SA chip in ow cells 1, 2, 3, and 4,
respectively. ERRc-LBD incubated with various ligands in running
buffer (50 mM TrispH8.0, 150 mMNaCl, 1 mMMgCl2, 0.002%NP-40, 0.2
mMDTT)were injected into ow cells for 3 min at a ow rate of 20
lL/minat 25 C. Changes in resonance units (RU) were monitored
simulta-neously in all ow cells. Surface was regenerated at the end
of eachcycle by an injection of 0.05% SDS. Sensorgrams were
generated byBIAcontrol software 4.1 using double-referencing to
eliminate re-sponsesfromthereferencesurfaceandbuffer-onlycontrol.
Spe-cicchangesinRUweregeneratedwithnon-specicinteractionbetween LBD
and NR2 negative control mutant peptide deducted.2.4. Western blot
analysisCells were lysed using RIPA reagent (Shanghai
Shenneng)according to the manufacturers protocol and protein
extracts
wereanalyzedby10%SDSPAGEandblottedontoPVDFmembrane.Membranes were
incubated with rabbit anti-human-PGC-1a anti-body generated in
house or mouse anti-b-actin antibody
(Boster)followedbyhorseradishperoxidase-conjugatedsecondaryanti-body
(Amersham) and developed with BeyoECL Plus reagent(Beyotime).3.
Results3.1. Effects of avonoids on ERb and ERRcWe aimed to study
the structural activity relationship (SAR) ofavonoids withERRc.
Werst conrmedtheabilitiesof theseavonoids to act as ERb agonists.
By measuring the activity of a
re-portergeneluciferaseunderthecontrolofaGal4-DBD-ERb-LBDFig. 1.
Chemical structures of avonoids. Structures of avones, avonols,
avanones, and isoavones are shown with the A-, B-, and C-rings
indicated.Z. Huang et al. / FEBS Letters 584 (2010) 2226 23fusion,
we found that these avonoids at 5 lM acted as ERb ago-nists in HeLa
cells except for hesperetin (Fig. 2A).The activity of ERRc is
strongly enhanced in the presence of
co-expressedPGC-1awhileaninverseagonist4-hydroxytamoxifensuppresses
this activity (Fig. 2B) [8]. We then examined the effectsof these
avonoids on ERRc in this system and found that
isoav-onegenisteindidnotsignicantlyaffectwhereasdaidzeinverymodestly
enhanced ERRc activity (Fig. 2C). Flavanones naringeninand
hesperetin did not show any dose-dependent effects (Fig. 2C).In
contrast, avone apigenin and avonol kaempferol suppressedERRc
activity in dose-dependent manners with kaempferol beingmore potent
(Fig. 2C). Flavone luteolin fully and avonol
quercetinpartiallysuppressedERRcactivityat 25 lM(Fig. 2C).
Wethenasked if the abilities of these avonoids to suppress ERRc
activityareduetoreducedcellularviabilities.
TotalcellularlevelofATPisreectiveofviability. Wemeasuredtotal
cellularATPcontentupon avonoid treatments and found that only
quercetin partiallyreduced viability correlating to its ability to
suppress ERRc activity(Supplementary Fig. S1). This result also
implies that the abilities
ofotheravonoidstofunctionasERRcactivitymodulatorsrelyonmechanisms
other than reducing viability.3.2. Apigenin interferes with the
interaction between ERRc and PGC-1aERRc activity is dependent onits
interactionwithPGC-1athroughthereceptor interacting(NR) motifsof
PGC-1awithastrongpreferenceforNR2overNR1andNR3(Fig. 3A).
Wethenexamined this in vitro interaction in the presence of
daidzein, api-genin, luteolin, and kaempferol. Among these
modulators, we pre-viously demonstrated that kaempferol functions
as an ERRcinverse agonist by directly blocking this
interactionmeasuredby a surface plasmon resonance based assay [6].
Using this assay,we found that daidzein and luteolin failed to
enhance or
interferewiththisinteractioncomparedtopositivecontrolswithDY-131as
an agonist and 4-hydroxytamoxifen as an inverse agonist(Fig.
3B)[8,9]. Ontheotherhand, apigenin dose-dependentsup-pressedthis
interactionapparentlyfunctioningas anERRcin-verseagonist(Fig. 3C).
Genistein, naringenin, andhesperetindidnot affect
theinteractionbetweenERRcandPGC-1a(datanotshown).3.3. Luteolin
suppresses the transcriptional activating function
withinPGC-1aSinceluteolinmodulatedERRcactivitywithout
affectingitsinteractionwithPGC-1a,
weexaminedifitwoulddirectlyaffectthe transcriptional activating
function within PGC-1a. When fusedto a Gal4-DBD, PGC-1a displays a
strong transcriptional activatingfunction on a reporter under the
control of Gal4-DBD [7]. We foundthat luteolin affected PGC-1a
activity mirroring its modulation onERRcactivity (Fig. 4A).
Wenextexaminedifluteolin
suppressedPGC-1aactivitybyalertingitsproteinstability.
WefoundthatFig. 2. Effects of avonoids on the activities of ERb and
estrogen-related receptor c (ERRc). (A) HeLa cells were transfected
with an expression plasmid of Gal4-DBD-ERb-LBDtogether with a
luciferase reporter and a control Renilla luciferase plasmid. About
10 nM 17b-estradiol as a positive control or different avonoids at
5 lM were added for 24 hbefore luciferase assays. Fold induction by
compounds were calculated and shown compared to dimethyl sulfoxide
(DMSO) as a vehicle. (B) HeLa cells were transfected withexpression
plasmids of Gal4-DBD control or Gal4-DBD-ERRc-LBD with or without
pcDNA-peroxisome proliferators-activated receptor c coactivator-1a
(PGC-1a) togetherwith a luciferase reporter and a control Renilla
luciferase plasmid. DMSO or 10 lM 4-hydroxytamoxifen (4-OHT) was
added and assays performed as in (A). (C) Transfectionwere
performed as in (B) with different avonoids (1, 5, and 25 lM); %
activity indicates the normalized activities of ERRc under the
inuences of avonoids compared toDMSO control set at 100%. (AC)
Results represent mean S.E.M.**P < 0.01.24 Z. Huang et al. /
FEBS Letters 584 (2010) 2226luteolin modestly reduced PGC-1a
protein level by Westernanalysis (Fig. 4B).4. DiscussionFlavonoids
share a common core phenylchromen-4-one struc-ture consists of A-,
B-, and C-rings (Fig. 1). Both avones and iso-avones function as
ERb agonists, suggesting that the position
oftheB-ringisnotcritical(Fig. 2A). However,
substitutionsontheB-ringplayakeyrole. Althoughapigenin, luteolin,
kaempferol,and quercetin function as ERb agonists, luteolin and
quercetin withadditional hydroxyl groups at the3-positionof
theB-ringareslightly less potent (Fig. 2A). Importantly, hesperetin
with a meth-oxygroupatthe40-positionoftheB-ringinsteadof
ahydroxylgroup as in naringenin, fails to function as an ERb
agonist (Fig. 2A).Several avonoids,specically,isoavones genistein,
daidzein,biochaninAandavone6,30,40-trihydroxyavonewerereportedtofunctionasERRaandERRcagonistsbasedonvirtual
ligandscreening and reporter-based assays [10].However, in this
studywe did not observe any effect of genistein on ERRc by in vitro
puri-ed ERRc-LBD interaction with PGC-1aNR2 motif and
luciferasereporter assays (Figs. 2 and 3). Independently, another
group alsofailed to reproduce these effects [11]. On the other
hand, we
iden-tiedinthisandapreviousstudythatapigeninandkaempferolblocktheinteractionbetweenERRcandPGC-1aandsuppressthe
expression of ERRc target gene pyruvate dehydrogenase kinase4
(PDK4) (Supplementary Fig. S2) as bone de ERRc inverse ago-nists
(Fig. 3) [6].Based on these results, we conclude that isoavones
with theirB-rings at position 3 of the C-ring do not block the in
vitro ERRcinteractionwithPGC-1a. Onthecontrary,
certainavonesandavonols such as apigenin and kaempferol with their
B-rings at po-sition 2 are capable of interfering with this
interaction. While apige-nin and kaempferol function as ERRc
inverse agonists, luteolin andquercetin do not affect the ERRc
interaction with PGC-1a (Fig. 3).Comparing the structures among
these avonoids suggests that ahydroxyl group at position 3 of the
C-ring can be accepted but a
hy-droxylgroupatposition30oftheB-ringisnottolerable(Fig.
1).Although avanones such as naringenin and hesperetin have
theirB-rings at position 2 of the C-rings, they fail to affect the
interactiondue to their 2,3-dihydro substitutions on the
C-rings.Disrupting the interaction between ERRc and PGC-1a is not
theonly mechanism by which certain avones and avonols functionto
suppress ERRc activity. Despite failing to block the ERRc
interac-tion with PGC-1a (Fig. 3), luteolin down-modulates ERRc
activityby suppressing the transcriptional activating function
withinPGC-1a in part through reducing its protein level (Fig. 4).
PGC-1ais subjected to degradation by the ubiquitin-proteasome
pathway[12]whichisregulatedbyglycogensynthasekinase3b(GSK3b)and
p38MAPK [13]. Luteolin may affect the activities of these ki-nases
or the expression of E3 ubiquitin ligases to reduce PGC-1aprotein
level through promoting its degradation. Since luteolin ismore
potent in suppressing PGC-1a activity than reducing its pro-tein
level(Fig. 4),luteolin may alsoinuence otherpost-transla-tional
modicationsof PGC-1atoaffect itsactivity[14]. Inall,we established
in this study a SAR between avonoids and ERRcFig. 3. Apigenin
directly blocks the interaction between ERRc and PGC-1a. (A) 125 nM
of puried ERRc-LBD was tested for its interaction with NR1, NR2,
and NR3 motifs asdescribed [6]. (B) DMSO, 1 lM 4-OHT, 5 lM DY-131,
25 lM daidzein, or 25 lM luteolin was incubated with 50 nM ERRc-LBD
for 1 hr before interaction analysis with NR2 asin (A). (C)
Different doses of apigenin were tested as in (B).Fig. 4. Effects
of avonoids on the activity of PGC-1a. (A) HeLa cells were
transfected with an expression plasmid of Gal4-DBD or
Gal4-DBD-PGC-1a together with a luciferasereporter and a control
Renilla luciferase plasmid. DMSO or 1, 5, and 25 lM luteolin were
added for 24 h; % activity indicates the normalized activities of
PGC-1a compared toDMSO control set at 100%. (B) HeLa cell extracts
pre-treated with DMSO or 25 lM luteolin was probed with anti-PGC-1a
or b-actin antibodies for Western analysis. PGC-1aprotein levels
normalized to b-actin from three independent experiments were
quantied. Results represent mean S.E.M.**P < 0.01.Z. Huang et
al. / FEBS Letters 584 (2010) 2226 25activity; namely, certain
avones and avonols utilize
alternativemechanismstosuppressERRcactivitythrougheitherdisruptingtheinteractionbetweenERRcandPGC-1aorsuppressingPGC-1a
activity through promoting PGC-1a protein degradation.
Thesealternative mechanisms may be further subjected to cell type
spe-cic regulation; thus, more extensive analysis of the inuences
ofavonoidsonmetabolicregulationmediatedbyERRcandPGC-1a is
warranted.AcknowledgementsThe research is supported by grants from
the National NaturalScienceFoundationofChina#30672463, theNational
BasicRe-searchProgramofChina(973-Program)#2006CB50390,
andtheKnowledgeInnovationProgramof theChineseAcademyof Sci-ences
#KSCX2-YW-R-085.Appendix A. Supplementary dataSupplementary data
associated with this article can be found, inthe online version, at
doi:10.1016/j.febslet.2009.11.026.References[1] Ross, J.A. and
Kasum, C.M. (2002) Dietary avonoids: Bioavailability,
metaboliceffects, and safety. Annu. Rev. Nutr. 22, 1934.[2] Pietta,
P.G. (2000) Flavonoids as antioxidants. J. Nat. Prod. 63,
10351042.[3] Kuiper, G.G.J.M., Lemmen, J.G., Carlsson, B., Corton,
J.C., Safe, S.H., van der Saag,P.T., vanderBurg, B. andGustafsson,
J.-A. (1998) Interactionof
estrogenicchemicalsandphytoestrogenswithestrogenreceptor{beta}.
Endocrinology139, 42524263.[4] Limer, J.L. and Speirs, V. (2004)
Phyto-oestrogens and breast cancerchemoprevention. Breast Cancer
Res. 6, 119127.[5] Giguere, V. (2008) Transcriptional control of
energy homeostasis by theestrogen-related receptors. Endocr. Rev.
29, 677696.[6] Wang, J., Fang, F., Huang, Z., Wang, Y. and Wong, C.
(2009) Kaempferol is anestrogen-related receptor [alpha] and
[gamma] inverse agonist. FEBS Lett. 583,643647.[7] Puigserver, P.,
Wu, Z., Park, C.W., Graves, R., Wright, M. and Spiegelman,
B.M.(1998)Acold-induciblecoactivatorof
nuclearreceptorslinkedtoadaptivethermogenesis. Cell 92, 829839.[8]
Coward, P., Lee, D., Hull, M.V. and Lehmann, J.M. (2001)
4-Hydroxytamoxifenbinds toanddeactivates the
estrogen-relatedreceptor gamma. PNAS 98,88808884.[9] Yu, D.D.
andForman, B.M. (2005) Identicationof anagonist
ligandforestrogen-relatedreceptorsERR[beta]/[gamma]. Bioorg. Med.
Chem. Lett. 15,13111313.[10] Suetsugi, M., Su, L., Karlsberg, K.,
Yuan, Y.-C. and Chen, S. (2003) Flavone andisoavonephytoestrogens
areagonists of estrogen-relatedreceptors. Mol.Cancer Res. 1,
981991.[11] Kallen, J. et al. (2004) Evidence for
ligand-independent transcriptionalactivation of the human
estrogen-related receptor {alpha} (ERR{alpha}):Crystal structure of
err{alpha} ligand binding domain in complex withperoxisome
proliferator-activated receptor coactivator-1{alpha}. J. Biol.Chem.
279, 4933049337.[12] Sano,M. et al. (2007) Intramolecular control
of protein stability, subnuclearcompartmentalization,
andcoactivatorfunctionofperoxisomeproliferator-activated receptor
{gamma} coactivator 1{alpha}. J. Biol. Chem. 282, 2597025980.[13]
Olson, B.L., Hock, M.B., Ekholm-Reed, S., Wohlschlegel, J.A., Dev,
K.K., Kralli, A.and Reed, S.I. (2008) SCFCdc4 acts antagonistically
to the PGC-1{alpha}transcriptional coactivator by targeting it for
ubiquitin-mediated proteolysis.Genes Dev. 22, 252264.[14]
Handschin, C. and Spiegelman, B.M. (2006) Peroxisome
proliferator-activatedreceptor {gamma} coactivator 1 coactivators,
energy homeostasis, andmetabolism. Endocr. Rev. 27, 728735.26 Z.
Huang et al. / FEBS Letters 584 (2010) 2226
