ma ine d ugs
A icle
Lep olide Imp o es Insulin Resis ance in
Die -Induced Obese Mice
Pablo Villa-Pé ez 1, Me cedes Cue o 2, Ana R. Díaz-Ma e o 3ID , Ca men D. Loba ón1ID ,
Al edo Mo eno 1, Ge mán Pe domo 4and I ene Cóza -Cas ellano 1,*ID
1Ins i u o de Biología y Gené ica Molecula , Uni e si y o Valladolid-CSIC, Valladolid 47005, Spain;
[email p o ec ed] (P.V.); [email p o ec ed] (C.D.L.); [email p o ec ed] (A.M.)
2Ins i u o de P oduc os Na u ales y Ag obiología (CSIC), La Laguna 38206, Spain; [email p o ec ed]
3Ins i u o Uni e si a io de Bioo gánica “A. González”, Uni e si y o La Laguna, La Laguna 38206, Spain;
[email p o ec ed]
4School o Nu se y, Uni e si y o Bu gos, Bu gos 09001, Spain; gmpe [email p o ec ed]
*Co espondence: [email p o ec ed]; Tel.: +34-983184005; Fax: +34-983423588
Recei ed: 18 July 2017; Accep ed: 13 Sep embe 2017; Published: 15 Sep embe 2017
Abs ac :
Type 2 diabe es (T2DM) is a complex disease linked o panc ea ic be a-cell ailu e and
insulin esis ance. Cu en an idiabe ic ea men egimens o T2DM include insulin sensi ize s
and insulin sec e agogues. We ha e p e iously demons a ed ha lep olide, a membe o he
u anocemb anolides amily, p omo es panc ea ic be a-cell p oli e a ion in mice. Conside ing he
bene icial e ec s o lep olide in diabe ic mice, in his s udy, we aimed o add ess he capabili y o
lep olide o imp o e insulin esis ance associa ed wi h he pa hology o obesi y. To his end, we
es ed he hypo hesis ha lep olide should p o ec agains a y acid-induced insulin esis ance
in hepa ocy es. In a ime-dependen manne , lep olide (0.1
µ
M) augmen ed insulin-s imula ed
phospho yla ion o p o ein kinase B (PKB) by wo- old abo e ehicle- ea ed HepG2 cells.
In addi ion, lep olide (0.1
µ
M) coun e ac ed palmi a e-induced insulin esis ance by augmen ing by
ou - old insulin-s imula ed phospho yla ion o PKB in HepG2 cells.
In i o
, acu e in ape i oneal
adminis a ion o lep olide (0.1 mg/kg and 1 mg/kg) imp o ed glucose ole ance and insulin
sensi i i y in lean mice. Likewise, p olonged lep olide ea men (0.1 mg/kg) in die -induced
obese mice imp o ed insulin sensi i i y. These e ec s we e pa alleled wi h an ~50% inc eased
o insulin-s imula ed phospho yla ion o PKB in li e and skele al muscle and educed ci cula ing
p o-in lamma o y cy okines in obese mice. We concluded ha lep olide signi ican ly imp o es insulin
sensi i i y
in i o
and in obese mice, sugges ing ha lep olide may be ano he po en ial ea men
o T2DM.
Keywo ds: lep olide; insulin esis ance; obesi y; ype 2 diabe es; HepG2 cells
1. In oduc ion
Insulin esis ance is one o he hallma ks o ype 2 diabe es (T2DM) and obesi y. Imp o emen o
insulin sensi i i y is an indispensable s ep o alle ia e T2DM. The i s phase o T2DM is cha ac e ized
by panc ea ic be a-cell compensa ion, displaying hype insulinemia in esponse o insulin esis ance.
Be a-cell o e wo king equen ly end in dys unc ion and cell dea h. A his poin , dec eased blood
insulin le els exace ba e he onse o T2DM due o bo h insulin de iciency and esis ance [1].
To his day, pha macological managemen o T2DM pa ien s aims o achie e he bes possible
glycemic con ol, while a oiding hypoglycemia. Howe e , he na u al his o y o T2DM includes
mul iple dys unc ions a ec ing he
α
-cells,
β
-cells, li e , skele al muscle, adipose issue, he
gas oin es inal ac , kidney and b ain, wha has been e med he ominous oc e [
2
]. This complex
Ma . D ugs 2017,15, 289; doi:10.3390/md15090289 www.mdpi.com/jou nal/ma ined ugs
Ma . D ugs 2017,15, 289 2 o 14
scena io o he pa hophysiology o T2DM equi es a shi in he cu en pa adigm o he ea men o
he disease.
A keys one in he managemen o diabe es is nu i ional he apy along wi h egula physical
ac i i y. Conside ing ha mos people wi h T2DM a e o e weigh o obese, weigh loss is
ecommended o imp o e glycemic con ol [
3
]. In addi ion, egula physical ac i i y has demons a ed
signi ican heal h bene i s [
4
]. When li es yle in e en ions ail in he app op ia e con ol o glucose
homeos asis, T2DM is ini ially ea ed by mono he apy wi h o al agen s, and e en ually, i may equi e
he combina ion o mul iple d ugs. In his way, insulin sensi ize s and insulin sec e agogues a e sa e
he apies o T2DM ea men . Among he insulin sensi ize s, me o min is success ully used as a
i s line pha maco he apy o he ea men o T2DM pa ien s. The majo e ec o his d ug is he
acu e inhibi ion o hepa ic gluconeogenesis h ough mechanisms in ol ing he di ec inhibi ion o
he mi ochond ial espi a o y-chain complex I, leading o ac i a ion o AMP-ac i a ed p o ein kinase
(AMPK) [
5
]. The pleio opic e ec s o me o min lead o lowe as ing blood glucose and insulin le els
wi h minimal isk o hypoglycemia [6].
When glycemic con ol canno be achie ed using me o min as mono he apy, a second line o
d ugs is added o me o min ea men . Among hese ea men s, sul onylu eas imp o e insulin
sec e ion by he egula ion o he ATP-sensi i e po assium channels in he plasma memb ane o
panc ea ic
β
-cells [
7
]. Thiazolidinediones, agonis s o nuclea pe oxisome p oli e a o -ac i a ed
ecep o gamma (PPAR
γ
), inc ease insulin sensi i i y in li e and skele al muscle [
8
]. Glinides s imula e
insulin sec e ion apidly and o a sho pe iod when needed [
9
]. Inc e ins, such as glucagon-like
pep ide 1 (GLP-1) ecep o agonis s, gas ic inhibi o y polypep ide/glucose-dependen insulino opic
pep ide (GIP) and dipep idyl-pep idase-4 (DPP-4) inhibi o s, inc ease insulin sec e ion and egula e
glucose homeos asis [
10
]. As he disease p og esses, he
β
-cell unc ion declines, and insulin he apy
become necessa y. The side e ec s o hypoglycemia and weigh gain limi he e icacy o his
he apy [
11
]. Howe e , no el insulin p epa a ions and deli e y sys ems a e a ailable o op imize
insulin he apy [
12
]. Rega ding new glucose-lowe ing he apies, sodium glucose co anspo e 2
(SGLT2) ecep o inhibi o s lowe blood glucose le els in an insulin-independen manne by inc easing
enal glucose exc e ion [
13
]. P omising new pha macological a ge s a e cu en ly unde e alua ion,
such as si uin agonis s, which enhance insulin sec e ion and/o insulin sensi i i y; o he inhibi o s o
p o ein y osine phospha ase 1B, which p olong he ac ion o insulin [10].
Fu anocemb anolides a e polyoxygena ed di e penoids, isola ed om co als, in which a u anic
ing and a
β
-lac one subuni a e inse ed in a cemb ane skele on [
14
]. Lep olide is a membe o he
u anocemb anolide amily, which was isola ed o he i s ime by Gu ié ez e al. in 2005 [
14
] om he
oc oco als Lep ogo gia alba and Lep ogo gia igida on he Paci ic coas o Panama. I has been p oposed
ha membe s o his amily, such as pukalide, may unc ion in na u e as a de ensi e oxin agains
po en ial oc oco al p eda o s [15,16]. The po en ial pha macological use o his amily o compounds
is la gely unexplo ed. Al hough, ew examples ha e been epo ed. Among hem, lopho oxin is a
neu omuscula oxin ha binds selec i ely and i e e sibly wi hin he ace ylcholine- ecogni ion
si e o nico inic ace ylcholine ecep o s, he eby p e en ing ace ylcholine om ac i a ing i s
ecep o [
17
,
18
]. The an ip oli e a i e and cy o oxic ac i i ies o some o hese compounds ha e
been s udied agains he cell lines L-929, K-562, HeLA, MDA-MB-231, A-549, HT-29 and P388 showing
weak an ip oli e a i e and cy o oxic p ope ies [
19
–
21
]. Fu he mo e, he an iplasmodial ac i i y o six
u anocemb anolides and he i egula pseudop e olide isola ed om specimens o Lep ogo gia alba
and L. igida was e alua ed, and among hem, lep olide and pukalide showed no biological ac i i y
agains he pa asi e [14].
Lep olide, among o he membe s o i s amily, has al eady been shown o inc ease panc ea ic
be a-cell p oli e a ion
in i o
, in INS1cells (Insulin sec e ing be a cell de i ed line) and p ima y
cul u es o oden panc ea ic isle s [
22
]. In addi ion, epoxypukalide, ano he molecule o his amily,
has been shown o imp o e be a-cell p o ec ion
in i o
and
in i o
, in oden isle s and in a STZ
(s ep ozo ocin)-induced model o diabe es, espec i ely [
22
,
23
]. In e es ingly, epoxypukalide also
Ma . D ugs 2017,15, 289 3 o 14
alle ia es glucose in ole ance in a p eclinical model o ype 1 diabe es [
23
]. Thus, u anocemb anolides
appea o be a ac i e molecules o main ain unc ional be a-cell mass and glycemic con ol.
In his wo k, we ha e ex ended ou ini ial indings and explo ed he capabili y o lep olide
o imp o e insulin sensi i i y. To his end, we ha e assessed he capaci y o lep olide o enhance
insulin signaling in insulin- esis an hepa ocy es and in he li e and skele al muscle o die -induced
obese mice.
2. Ma e ials and Me hods
Lep olide pu i ica ion, cha ac e iza ion and molecula s uc u e we e desc ibed p e iously [
14
].
B ie ly, c ude ex ac s om oc oco als we e subjec ed o ac iona ion. Lep olide was ini ially isola ed
as a no el compound wi h an iplasmodial ac i i y, and i s s uc u e was de e mined by NMR and
con i med by single-c ys al X- ay c ys allog aphy.
2.1. Cell Cul u e
HepG2 cells we e ob ained om he Ame ican Type Cul u e Collec ion (ATCC, Manassas, VA,
USA; #HB-8065). The cell line was o iginally isola ed om a li e hepa ocellula ca cinoma o a
15-yea -old Caucasian male. Cells we e g ow h in DMEM (1X) supplemen ed wi h 4.5 g/L D-glucose,
0.6 g/L L-glu amine, 0.1 g/L sodium py u a e and 10% e al bo ine se um.
In o de o analyze he e ec s o lep olide on he in acellula insulin signaling pa hway, HepG2
cells we e ea ed wi h 0.1
µ
M lep olide o ehicle (DMSO) du ing 24 h in medium wi hou se um.
A e wa ds, 100 nM human insulin (Sigma, S . Louis, MO, USA) was added, and HepG2 cells we e
collec ed a e 0, 5, 10, 15 and 30 min. To analyze he e ec s o lep olide in he se ing o esis ance,
HepG2 cells we e ea ed wi h 0.2 mM palmi a e and 0.1
µ
M lep olide in se um- ee medium o 24 h.
A e wa ds, 100 nM human insulin (Sigma, S . Louis, MO, USA) was added, and 15 min la e , HepG2
cells we e collec ed.
2.2. Animal P ocedu es
C57Bl6J male mice we e pu chase om Cha les Ri e Labo a o y (Écully, F ance). Male mice
we e chosen o me abolic pheno yping o a oid he po en ial a iabili y ela ed o es ous cycle.
Expe imen al p ocedu es we e app o ed by he Animal Ca e and Use Commi ee o he Uni e si y
o Valladolid (UVa), Valladolid, Spain, in acco dance wi h he Eu opean and Spanish Guidelines o
he Ca e and Use o Mammals in Resea ch. Mice we e ed wi h s anda d oden chow and wa e ad
libi um in en ila ed cages in a 12:12-h ligh /da k cycle.
“Acu e” adminis a ion o lep olide was pe o med in 12-week-old males ed a s anda d die
(SD) (33% p o ein; 58% ca bohyd a e; 9% a ) (#V1535, Ssni , Soes , Ge many) a he indica ed doses
(0.1 mg/kg and 1 mg/kg o body weigh ). “Ch onic” adminis a ion o lep olide was pe o med
in 6-week-old male mice ed a 60% kcal high a die (HFD) (20% p o ein; 20% ca bohyd a e; 60%
a ) (#D12492, Resea ch Die s, New B unswick, NJ, USA) o 10 weeks. A e 6 weeks o eeding
wi h he HFD, mice we e andomly di ided in o wo g oups, which we e ea ed wi h once-daily ip
injec ion o lep olide (0.1 mg/kg o body weigh ) o ehicle (DMSO) o ano he 4 weeks. All mice
we e main ained on HFD du ing he 4-week ea men . The day be o e sac i ice, mice we e as ed
o e nigh , ollowed by an ip insulin o saline injec ion, and 10 min la e , mice we e eu hanized o
li e and skele al muscle issues dissec ion as desc ibed p e iously [24].
Fas ing o non- as ing blood was collec ed om he ail ein in o capilla y ubes p ecoa ed wi h
po assium-EDTA (Sa s ed , Nümb ech , Ge many) o he p epa a ion o plasma o he de e mina ion
o blood glucose le els using he B eeze 2 glucome e (Baye , Le e kusen, Ge many) as p e iously
desc ibed [
24
]. Insulin le els we e measu ed using he ul asensi i e mouse ELISA assay (Me codia,
Uppsala, Sweden). T iglyce ides we e measu ed using a iglyce ides ki (Biosys ems, Ba celona, Spain).
Cy okines and lep in we e measu ed using Bio-Plex Luminex Immunoassays (Bio-Rad, He cules, CA,
Ma . D ugs 2017,15, 289 4 o 14
USA). The de ec ion limi s o TNF-
α
, IL-1, IL6, insulin and lep in we e 4.0 pg/mL, 1.6 pg/mL,
0.25 pg/mL, 0.025 pg/mL and 4.9 pg/mL espec i ely.
2.3. Glucose and Insulin Tole ance Tes s, Glucose Decay and HOMA Indexes
The in ape i oneal glucose ole ance es (ip-GTT) was pe o med, 30 min a e “acu e” ea men
o 4 weeks a e “ch onic” ea men , as p e iously desc ibed [
24
]. B ie ly, mice we e as ed o e nigh
(15 h) ollowing he ecommended s anda d ope a ing p ocedu e o pheno yping mice by he
Eumo phia Conso ium [
25
]. A e wa ds, mice we e in ape i oneally injec ed wi h 2 g glucose/kg
o body weigh . Blood glucose le els we e de e mined a 0, 15, 30, 60 and 120 min and plo ed as
a unc ion o ime. Likewise, he insulin ole ance es (ip-ITT) was pe o med, 30 min a e “acu e”
ea men o 4 weeks a e “ch onic” ea men , as p e iously desc ibed [
24
]. Fo ip-ITT, non- as ed
mice we e injec ed (1 U/kg o body weigh ) wi h insulin (Lilly, Indianapolils, IN, USA). Blood glucose
le els we e de e mined a 0, 15, 30, 60 and 90 min and plo ed as a unc ion o ime. Ip-GTT and ip-ITT
expe imen s we e pe o med using he same g oup o mice. The imeline o he expe imen s was i s
he ip-GTT assays. Then, we le mice eco e o h ee days, and ip-ITT expe imen s we e pe o med.
Glucose decay was calcula ed om he glucose measu emen s ob ained du ing he insulin ole ance
es . The measu emen s we e con e ed in o na u al loga i hm (Ln); he slope was calcula ed using
linea eg ession ( ime
×
Ln[glucose]) and mul iplied by 100 o ob ain he glucose decay cons an a e
pe minu e (%/min).
The homeos asis model assessmen (HOMA) es ima es s eady s a e be a cell unc ion (%B), insulin
sensi i i y (%S), he in e se o %S and he insulin esis ance (HOMA-IR). The HOMA Calcula o
so wa e is eely a ailable a he Uni e si y o Ox o d, Uni ed Kingdom, a he web page www.d u.ox.
ac.uk/homacalcula o .
2.4. Wes e n-Blo Analysis
HepG2 cells we e p eincuba ed in he p esence o absence o lep olide a he abo e indica ed
imes and doses. A he end o he incuba ion pe iod, cul u e media we e disca ded and cells collec ed
and washed wi h ice-cold PBS, ollowed by homogeniza ion in lysis bu e (20 mmol/L T is
·
HCl, pH
7.5, 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L EGTA, 1% ( ol/ ol) T i on X-100, 2.5 mmol/L
sodium py ophospha e, 1 mmol/L
µ
-glyce ophospha e, 1 mmol/L Na
3
VO
4
, 1
µ
g/mL leupep in
and 1 mmol/L phenylme hylsul onyl luo ide) plus p o ease inhibi o s (P o ease Inhibi o Cock ail;
Sigma). A e 10 min on ice, ex ac s we e sonica ed and cen i uga ed a 18,000
×
g o 10 min a 4
◦
C.
Pelle s we e disca ded, and solubilized p o eins (~20–40
µ
g/sample) we e esol ed by 10% SDS-PAGE
o an i-p-PKB (Se 473) (1:1000; Cell Signaling, Dan e s, MA, USA) and elec o ans e ed on o
poly inylidene di luo ide il e s (PDVF Immobilon-P memb ane (Millipo e, Bille ica, MA, USA)) o
immunoblo ing by con en ional means. A e p obing wi h speci ic an ibodies, he memb anes we e
s ipped and ep obed wi h an ibody agains ac in (1:3000; Sigma) and PKB (1:1000; Cell Signaling).
Signals we e de ec ed by chemiluminescence (Immun-S a Wes e n Chemiluminescence Ki ; Bio-Rad,
Mad id, Spain), and band densi ome y was quan i ied wi h Image J so wa e (Na ional Ins i u es o
Heal h, Be hesda, MD, USA).
Fo animal issues, li e and skele al muscle om mice, s imula ed o no wi h insulin, we e
homogenized wi h a poly on (OMNI, Kennesaw, GA, USA) in cell lysis bu e (Cell Signaling, USA)
in he p esence o p o ease/phospha ase inhibi o s. As desc ibed o HepG2 cells, ~40–60
µ
g/sample
we e esol ed in 10%-SDS PAGE o an i-pPKB, PKB and ac in.
2.5. S a is ical Analysis
S a is ical analysis o da a was pe o med using he G aphPad P ism So wa e 6.0 (La Jolla, CA,
USA). Dis ibu ions we e checked wi h he Kolmogo o –Smi no es . Da a a e p esen ed as he means
±
S.E.M. Homogenei y o a iance was pe o med using he Le ene es . Compa isons be ween wo
g oups we e done using he unpai ed S uden ’s - es (i homogenei y o a iance) o he Welch es
Ma . D ugs 2017,15, 289 5 o 14
(i he e ogenei y o a iance) when a a iable was dis ibu ed no mally; in he case o a non-pa ame ic
a iable, he Mann–Whi ney U- es was used. Compa isons be ween mo e han wo g oups we e
done using he one-way ANOVA o he K uskal–Wallis es when a a iable was dis ibu ed no mally
o non-no mally, espec i ely. Fo pos -hoc analyses, he Bon e oni es o Dunne ’s es was used
i he e was homogenei y o he e ogenei y o a iance, espec i ely. Di e ences we e conside ed
signi ican a p< 0.05.
3. Resul s
3.1. Lep olide Imp o es Insulin Signaling in HepG2 Cells
To s udy he e ec s o lep olide in basal and insulin esis ance condi ions, we ha e used a human
hepa oma cell line (HepG2). To dis ega d he possibili y o a cy o oxic e ec o lep olide on HepG2
cells, 750,000 cells we e pla ed and ea ed wi h ehicle o lep olide o 24 h (n= 4 independen
expe imen s). Li ing cells we e collec ed a e wa ds, and p o eins we e ex ac ed and quan i ied
showing no di e ences in p o ein con en in lep olide- e sus ehicle- ea ed cells (4.42
±
0.53
µ
g/
µ
L
e sus 4.30
±
0.77
µ
g/
µ
L). Vehicle- ea ed cells showed a ime-dependen ac i a ion (5 min) o he
insulin signaling pa hway a e insulin s imula ion. Lep olide (0.1
µ
M) inc eased insulin signaling
~1.3–2.0- old compa ed o ehicle- ea ed cells (Figu e 1). This imp o emen in insulin sensi i i y is
sus ained unde palmi a e-induced insulin esis ance (Figu e 2). HepG2 cells we e ea ed wi h 0.2 mM
palmi ic acid o 24 h o cause insulin esis ance, a he same ime cells we e ea ed wi h ehicle o
0.1
µ
M lep olide. Con ol cells showed an inc eased p-PKB/PKB a io in basal condi ions, which was
impai ed unde insulin esis ance condi ions. In e es ingly, lep olide ea men no only inc eases
insulin signaling in basal condi ions, bu also coun e ac ed palmi a e-induced insulin esis ance
(Figu e 2). These esul s highligh ha lep olide is a molecule wi h he capaci y o enhancing sensi i i y
a basal condi ions and palmi a e-induced insulin esis ance in HepG2 cells.
Ma .D ugs2017,15,289 5o 14
Dunne ’s es wasusedi  he ewashomogenei yo he e ogenei yo  a iance, espec i ely.
Di e enceswe econside edsigni ican a p<0.05.
3.Resul s
3.1.Lep olideImp o esInsulinSignalinginHepG2Cells
Tos udy hee ec so lep olideinbasalandinsulin esis ancecondi ions,weha eusedahuman
hepa omacellline(HepG2).Todis ega d hepossibili yo acy o oxice ec o lep olideonHepG2
cells,750,000cellswe epla edand ea edwi h ehicleo lep olide o 24h(n=4independen 
expe imen s).Li ingcellswe ecollec eda e wa ds,andp o einswe eex ac edandquan i ied
showingnodi e encesinp o eincon en inlep olide‐ e sus ehicle‐ ea edcells(4.42±0.53μg/μL
e sus4.30±0.77μg/μL).Vehicle‐ ea edcellsshoweda ime‐dependen ac i a ion(5min)o  he
insulinsignalingpa hwaya e insulins imula ion.Lep olide(0.1μM)inc easedinsulinsignaling
~1.3–2.0‐ oldcompa ed o ehicle‐ ea edcells(Figu e1).Thisimp o emen ininsulinsensi i i yis
sus ainedunde palmi a e‐inducedinsulin esis ance(Figu e2).HepG2cellswe e ea edwi h
0.2mMpalmi icacid o 24h ocauseinsulin esis ance,a  hesame imecellswe e ea edwi h
ehicleo 0.1μMlep olide.Con olcellsshowedaninc easedp‐PKB/PKB a ioinbasalcondi ions,
whichwasimpai edunde insulin esis ancecondi ions.In e es ingly,lep olide ea men no only
inc easesinsulinsignalinginbasalcondi ions,bu alsocoun e ac edpalmi a e‐inducedinsulin
esis ance(Figu e2).These esul shighligh  ha lep olideisamoleculewi h hecapaci yo 
enhancingsensi i i ya basalcondi ionsandpalmi a e‐inducedinsulin esis anceinHepG2cells.

Figu e1.Lep olideaugmen sinsulinsignalinginHepG2cells.Wes e nblo analysiso PKBand
p‐PKBinHepG2cells ea edwi hlep olide(0.1μM)o  ehicleands imula edwi h100mMinsulin
o 5,10,15and30min.Lep olideac iononHepG2cellsenhancedinsulinsignalinga all imepoin s.
Valuesa e hemeans±S.E.M.o n=6pe condi ion.They‐axis ep esen s he a ioo phospho yla ed
e sus o alp o einina bi a yuni s.Ac inwasde ec edasaloadingcon ol.*p<0.05 e sus =0
byANOVA;
$
p<0.05 e sus ehiclebyANOVA.
Figu e 1.
Lep olide augmen s insulin signaling in HepG2 cells. Wes e n blo analysis o PKB and
p-PKB in HepG2 cells ea ed wi h lep olide (0.1
µ
M) o ehicle and s imula ed wi h 100 mM insulin
o 5, 10, 15 and 30 min. Lep olide ac ion on HepG2 cells enhanced insulin signaling a all ime poin s.
Values a e he means
±
S.E.M. o n= 6 pe condi ion. The y-axis ep esen s he a io o phospho yla ed
e sus o al p o ein in a bi a y uni s. Ac in was de ec ed as a loading con ol. * p< 0.05 e sus = 0 by
ANOVA; $p< 0.05 e sus ehicle by ANOVA.

Ma . D ugs 2017,15, 289 6 o 14
Ma .D ugs2017,15,289 6o 14

Figu e2.Lep olidep e en spalmi a e‐inducedinsulin esis anceinHepG2cells.Wes e nblo 
analysiso PKBandp‐PKBinHepG2cells ea edwi hlep olide(0.1μM)o  ehicle,in hep esence
o absenceo palmi a e.Cellswe es imula edwi h100mMinsulin o 15min.Lep olideac ionon
HepG2cellsenhancedinsulinsignaling,p e en ingpalmi a e’snega i ee ec .Valuesa e he
means±S.E.M.o n=4pe condi ion.They‐axis ep esen s he a ioo phospho yla ed e sus o al
p o einina bi a yuni s.Ac inwasde ec edasaloadingcon ol.*p<0.05 e susnon‐s imula ed
insulincellsbyANOVA.
3.2.Acu eT ea men wi hLep olideImp o esGlucoseTole anceandInsulinSensi i i yinLeanMice
Toco obo a eou  indingsin i o,weacu elyinjec edlep olidea  wodi e en concen a ions
(0.1and1mg/kg)inC57BL6Jmalemice.Eachdoseo lep olideandi s espec i econ olwe e
adminis e edaspa o independen expe imen s,and heywe epe o medonsepa a edays.
Vehicleda awe epooled ombo hexpe imen s.Lep olidewasinjec ed30minbe o eip‐GTTin
as edmice( imepoin −30).A  hesame ime,plasmaglucosele elswe eassessed( imepoin −30).
A e wa ds,aboluso glucosewasadminis e edin ape i oneallyasdesc ibedin heMa e ialsand
Me hodsSec ion.Plasmaglucosele elswe emoni o eda 15,30,60,90and120mina e glucose
challenge.AsshowninFigu e3A,B,bo hlep olideconcen a ionsdisplayedimp o edglucose
ole ance.Inadi e en g oupo expe imen s,lep olidewasinjec edp e ious oip‐ITTshowing
inc easedinsulinsensi i i ya bo hlep olideconcen a ions(Figu e3C,D).Likewise,acu e
adminis a iono lep olide(0.1and1mg/kg)imp o edinsulinsensi i i yinleanmice(Figu e3C,D).
Consis en wi h hesee ec so lep olideonwhole‐bodyglucosehomeos asis,non‐ as ingplasma
glucosele elswe e educed(p=0.06)inmice ea edwi hahighdoseo lep olide(Figu e3E,F).
Collec i ely, hese esul sdemons a e ha bo hdoseso lep olideha eaposi i ee ec on
imp o inginsulinsensi i i yandglucosehomeos asis.In iewo  heseda a,weha echosen he
lowes doseo lep olide ope o m henex expe imen sindie ‐inducedobesemice.
Figu e 2.
Lep olide p e en s palmi a e-induced insulin esis ance in HepG2 cells. Wes e n blo analysis
o PKB and p-PKB in HepG2 cells ea ed wi h lep olide (0.1
µ
M) o ehicle, in he p esence o absence
o palmi a e. Cells we e s imula ed wi h 100 mM insulin o 15 min. Lep olide ac ion on HepG2 cells
enhanced insulin signaling, p e en ing palmi a e’s nega i e e ec . Values a e he means
±
S.E.M. o
n = 4 pe condi ion. The y-axis ep esen s he a io o phospho yla ed e sus o al p o ein in a bi a y
uni s. Ac in was de ec ed as a loading con ol. * p< 0.05 e sus non-s imula ed insulin cells by ANOVA.
3.2. Acu e T ea men wi h Lep olide Imp o es Glucose Tole ance and Insulin Sensi i i y in Lean Mice
To co obo a e ou indings
in i o
, we acu ely injec ed lep olide a wo di e en concen a ions
(0.1 and 1 mg/kg) in C57BL6J male mice. Each dose o lep olide and i s espec i e con ol we e
adminis e ed as pa o independen expe imen s, and hey we e pe o med on sepa a e days.
Vehicle da a we e pooled om bo h expe imen s. Lep olide was injec ed 30 min be o e ip-GTT
in as ed mice ( ime poin
−
30). A he same ime, plasma glucose le els we e assessed ( ime poin
−
30). A e wa ds, a bolus o glucose was adminis e ed in ape i oneally as desc ibed in he Ma e ials
and Me hods Sec ion. Plasma glucose le els we e moni o ed a 15, 30, 60, 90 and 120 min a e
glucose challenge. As shown in Figu e 3A,B, bo h lep olide concen a ions displayed imp o ed
glucose ole ance. In a di e en g oup o expe imen s, lep olide was injec ed p e ious o ip-ITT
showing inc eased insulin sensi i i y a bo h lep olide concen a ions (Figu e 3C,D). Likewise, acu e
adminis a ion o lep olide (0.1 and 1 mg/kg) imp o ed insulin sensi i i y in lean mice (Figu e 3C,D).
Consis en wi h hese e ec s o lep olide on whole-body glucose homeos asis, non- as ing plasma
glucose le els we e educed (p= 0.06) in mice ea ed wi h a high dose o lep olide (Figu e 3E,F).
Collec i ely, hese esul s demons a e ha bo h doses o lep olide ha e a posi i e e ec on imp o ing
insulin sensi i i y and glucose homeos asis. In iew o hese da a, we ha e chosen he lowes dose o
lep olide o pe o m he nex expe imen s in die -induced obese mice.
Ma . D ugs 2017,15, 289 7 o 14
Ma .D ugs2017,15,289 7o 14


Figu e3.Acu elep olideadminis a ionimp o esglucose ole anceandinsulinsensi i i yinlean
mice.C57Bl6J12‐week‐oldmales edas anda ddie we einjec edwi h ehicleo lep olide30min
be o e hein ape i onealglucose ole ance es (ip‐GTT)o  heinsulin ole ance es (ip‐ITT).
(A)Glucose ole ance es o miceinjec edwi hlep olide(0.1mg/kglep olideo 1mg/kg)o  ehicle.
(B)A eaunde  hecu eo  heip‐GTT.Glucose ole ancewasimp o edusingbo hlep olide
concen a ionscompa ed o he ehicle.(C)Insulin ole ance es o miceinjec edwi hlep olide
(0.1mg/kglep olideo 1mg/kg)o  ehicle.(D)A eaunde  hecu eo  heip‐ITT.Insulinsensi i i y
wasimp o edusingbo hlep olideconcen a ionscompa ed o ehicle.(E)Fas ingand
(F)non‐ as ingbloodglucosele elsa e  hi yminu eso lep olideadminis a ion.Fas ingglucose
le elswe eunchanged;meanwhile,only1mg/kglep olideimp o ednon‐ as ingglucose.Valuesa e
hemeans±S.E.M.o n=12( ehicle);n=6(lep olide0.1mg/kg);n=6(lep olide1mg/kg).*p<0.05
e sus ehiclebyANOVA.

AB
CD
AUC (A bi a y uni s)
Vehicle
Lep olide 0.1 mg/kg
Lep olide 1 mg/kg
Glucose (mg/dL)
EF
Vehicle
Lep olide 0.1 mg/kg
Lep olide 1 mg/kg
Figu e 3.
Acu e lep olide adminis a ion imp o es glucose ole ance and insulin sensi i i y in lean mice.
C57Bl6J 12-week-old males ed a s anda d die we e injec ed wi h ehicle o lep olide 30 min be o e
he in ape i oneal glucose ole ance es (ip-GTT) o he insulin ole ance es (ip-ITT). (
A
) Glucose
ole ance es o mice injec ed wi h lep olide (0.1 mg/kg lep olide o 1 mg/kg) o ehicle. (
B
) A ea
unde he cu e o he ip-GTT. Glucose ole ance was imp o ed using bo h lep olide concen a ions
compa ed o he ehicle. (
C
) Insulin ole ance es o mice injec ed wi h lep olide (0.1 mg/kg lep olide
o 1 mg/kg) o ehicle. (
D
) A ea unde he cu e o he ip-ITT. Insulin sensi i i y was imp o ed using
bo h lep olide concen a ions compa ed o ehicle. (
E
) Fas ing and (
F
) non- as ing blood glucose le els
a e hi y minu es o lep olide adminis a ion. Fas ing glucose le els we e unchanged; meanwhile,
only 1 mg/kg lep olide imp o ed non- as ing glucose. Values a e he means
±
S.E.M. o n= 12 ( ehicle);
n = 6 (lep olide 0.1 mg/kg); n = 6 (lep olide 1 mg/kg). * p< 0.05 e sus ehicle by ANOVA.
3.3. P olonged Lep olide Adminis a ion Imp o es Li e and Muscle Insulin Resis ance in Die -Induced
Obese Mice
Nex , we es ed he hypo hesis ha adminis a ion o lep olide imp o es insulin sensi i i y
in obese mice. To his end, C57BL6J mice we e ed a high a die (60% kcal o a ) o en weeks.
Ma . D ugs 2017,15, 289 8 o 14
The las ou weeks, mice we e in ape i oneally injec ed wi h lep olide (0.1 mg/kg) once a day. A he
end o he ea men , insulin sensi i i y was assessed by ip-ITT and he HOMA index in all mice.
To in es iga e he impac o lep olide in he in acellula insulin signaling pa hway in li e and skele al
muscle issues, one-hal o mice (con ol and lep olide g oups) we e injec ed wi h insulin o 10 min,
whe eas he o he hal was injec ed wi h saline. A e wa ds, mice we e eu hanized and sac i iced o
he dissec ion o li e o skele al muscle issues. Da a in Figu es 4–6we e pooled om wo disc e e
expe imen s using wo di e en g oups o mice.
P olonged adminis a ion o lep olide imp o ed glucose ole ance (Figu e 4A,B) and insulin
sensi i i y (Figu e 4C,D) in obese mice. Consis en ly, glucose decay du ing he ip-ITT du ing he i s
30 min was imp o ed in he lep olide- ea ed g oup (Figu e 4E). Fu he mo e, he insulin sensi i i y
index (%S) was signi ican ly inc eased (Figu e 4F), which was pa alleled wi h a educed HOMA-IR
index (p= 0.06; Figu e 4G), al hough body weigh was no signi ican ly dec eased (Figu e 4H). In
addi ion, as ing and non- as ing plasma insulin le els we e educed in lep olide- ea ed g oup mice
(Figu e 4I,J), which is consis en wi h he no ion o imp o ing insulin sensi i i y. Finally, plasma
iglyce ide le els we e signi ican ly educed in obese mice ea ed wi h lep olide (Figu e 4K), which
nicely co ela es wi h a 30% educ ion in li e iglyce ide con en in lep olide- e sus ehicle- ea ed
mice (Figu e 4L). Taken oge he , hese da a demons a e ha ch onic lep olide adminis a ion
coun e ac ed insulin esis ance and imp o ed lipid me abolism in die -induced obese mice.
To gain insigh in o he molecula mechanisms by which lep olide imp o ed insulin esis ance,
we analyzed le els o ci cula ing p o-in lamma o y cy okines. As shown in Figu e 5, ch onic lep olide
ea men was associa ed wi h a non-s a is ically-signi ican dec ease in blood le els o IL-1
β
and
TNF-
α
, bu IL-6 le els emained unchanged (Figu e 5A–C). Likewise, lep olide signi ican ly educed
he weigh gain du ing he las week o he ea men (Figu e 5D), in pa allel wi h educed plasma
lep in le els in obese mice (Figu e 5F).
Ma .D ugs2017,15,289 8o 14
3.3.P olongedLep olideAdminis a ionImp o esLi e andMuscleInsulinResis anceinDie ‐Induced
ObeseMice
Nex ,we es ed hehypo hesis ha adminis a iono lep olideimp o esinsulinsensi i i yin
obesemice.To hisend,C57BL6Jmicewe e edahigh a die (60%kcalo  a ) o  enweeks.The
las  ou weeks,micewe ein ape i oneallyinjec edwi hlep olide(0.1mg/kg)onceaday.A  he
endo  he ea men ,insulinsensi i i ywasassessedbyip‐ITTand heHOMAindexinallmice.To
in es iga e heimpac o lep olidein hein acellula insulinsignalingpa hwayinli e andskele al
muscle issues,one‐hal o mice(con olandlep olideg oups)we einjec edwi hinsulin o 10min,
whe eas heo he hal wasinjec edwi hsaline.A e wa ds,micewe eeu hanizedandsac i iced o 
hedissec iono li e o skele almuscle issues.Da ainFigu es4–6we epooled om wodisc e e
expe imen susing wodi e en g oupso mice.
P olongedadminis a iono lep olideimp o edglucose ole ance(Figu e4A,B)andinsulin
sensi i i y(Figu e4C,D)inobesemice.Consis en ly,glucosedecaydu ing heip‐ITTdu ing he i s 
30minwasimp o edin helep olide‐ ea edg oup(Figu e4E).Fu he mo e, heinsulinsensi i i y
index(%S)wassigni ican lyinc eased(Figu e4F),whichwaspa alleledwi ha educedHOMA‐IR
index(p=0.06;Figu e4G),al houghbodyweigh wasno signi ican lydec eased(Figu e4H).In
addi ion, as ingandnon‐ as ingplasmainsulinle elswe e educedinlep olide‐ ea edg oupmice
(Figu e4I,J),whichisconsis en wi h heno iono imp o inginsulinsensi i i y.Finally,plasma
iglyce idele elswe esigni ican ly educedinobesemice ea edwi hlep olide(Figu e4K),which
nicelyco ela eswi ha30% educ ioninli e  iglyce idecon en inlep olide‐ e sus ehicle‐ ea ed
mice(Figu e4L).Taken oge he , heseda ademons a e ha ch oniclep olideadminis a ion
coun e ac edinsulin esis anceandimp o edlipidme abolismindie ‐inducedobesemice.
Togaininsigh in o hemolecula mechanismsbywhichlep olideimp o edinsulin esis ance,
weanalyzedle elso ci cula ingp o‐in lamma o ycy okines.AsshowninFigu e5,ch oniclep olide
ea men wasassocia edwi hanon‐s a is ically‐signi ican dec easeinbloodle elso IL‐1βand
TNF‐α,bu IL‐6le els emainedunchanged(Figu e5A–C).Likewise,lep olidesigni ican ly educed
heweigh gaindu ing helas weeko  he ea men (Figu e5D),inpa allelwi h educedplasma
lep inle elsinobesemice(Figu e5F).

AC
B
DE
Vehicle Lep olide
0
100
200
300
*
Vehicle Lep olide
0
2000
4000
6000
*
Vehicle Lep olide
0.0
0.5
1.0
1.5
2.0 *
F
Glucose (mg/dL)
AUC (A bi a y Uni s)
Vehicle Lep olide
0
10000
20000
30000
40000
50000
*
Figu e 4. Con .
Ma . D ugs 2017,15, 289 9 o 14
Ma .D ugs2017,15,289 9o 14

Figu e4.P olongedlep olideadminis a ionimp o esinsulinsensi i i yinap eclinicalmodelo 
insulin esis ance.C57BL6Jmalemicewe e edanHFD o  enweeks.Thelas  ou weeks,micewe e
injec edin ape i oneallywi hlep olideo salineonceaday.A e wa ds,insulinsensi i i yand
plasmale elso insulinand iglyce ideswe eassessed.(A)Glucose ole ance es o mice ea ed
wi h ehicleo 0.1mg/kglep olide.(B)A eaunde  hecu eo  heip‐GTT.Glucose ole ance
imp o edinlep olide‐compa ed o ehicle‐ ea edmice.(C)Insulin ole ance es o mice ea ed
wi h ehicleo 0.1mg/kglep olide.(D)A eaunde  hecu eo  heip‐ITT.Insulinsensi i i y
imp o edinlep olide‐ compa ed o ehicle‐ ea edmice.(E)Glucosedecaya e  i s 30mino 
insulininjec iondu ingip‐ITT.(F)Insulinsensi i i yindex(%S).(G)HOMAindex.Insulinsensi i i y
andHOMAindexesshowedimp o edinsulinsensi i i yinHFD+lep olidemice.(H)Fas ingbody
weigh .(I)Fas ingand(J)non‐ as ingplasmainsulinle els.Insulinle elswe enon‐signi ican ly
dec eased.(K)Fas ingplasma iglyce idesle elswe edec easedinmice ea edwi hlep olide.
(L)Li e TG( iglyce ide)con en wasdec easedinpa allelwi hplasma iglyce idele els.Values
a e hemeans±S.E.M.o n=12pe g oup.*p<0.05 e sus ehiclebyS uden ’s ‐ es ;$p<0.05 e sus
SDbyANOVA.
To u he in es iga e heimpac o p olongedadminis a iono lep olideonobesemice, he
ac i a iono  hein acellula insulinsignalingpa hwaywasassessedin heli e andskele almuscle
o obesemice.AsshowninFigu e6,lep olideimp o eda basalandinsulin‐s imula edcondi ions,
wi h hephospho yla iono PKBinli e andskele almuscle issue.These esul sa eingood
ag eemen wi h hee ec o lep olideoninsulinsensi i i yinobesemice.
H
G
Fas ing Insulin (mg/dL)
Non-Fas ing Insulin (mg/dL)
Vehicle Lep olide
0
50
100
150
*
I
J
Li e TG (mg/g li e )
HOMA2 IR
Weigh Fas ing (g)
SD
Vehicle
Lep olide
0
10
20
30
40
$$
KL
Figu e 4.
P olonged lep olide adminis a ion imp o es insulin sensi i i y in a p eclinical model o
insulin esis ance. C57BL6J male mice we e ed an HFD o en weeks. The las ou weeks, mice
we e injec ed in ape i oneally wi h lep olide o saline once a day. A e wa ds, insulin sensi i i y and
plasma le els o insulin and iglyce ides we e assessed. (
A
) Glucose ole ance es o mice ea ed wi h
ehicle o 0.1 mg/kg lep olide. (
B
) A ea unde he cu e o he ip-GTT. Glucose ole ance imp o ed in
lep olide- compa ed o ehicle- ea ed mice. (
C
) Insulin ole ance es o mice ea ed wi h ehicle o
0.1 mg/kg lep olide. (
D
) A ea unde he cu e o he ip-ITT. Insulin sensi i i y imp o ed in lep olide-
compa ed o ehicle- ea ed mice. (
E
) Glucose decay a e i s 30 min o insulin injec ion du ing
ip-ITT. (
F
) Insulin sensi i i y index (%S). (
G
) HOMA index. Insulin sensi i i y and HOMA indexes
showed imp o ed insulin sensi i i y in HFD + lep olide mice. (
H
) Fas ing body weigh . (
I
) Fas ing
and (
J
) non- as ing plasma insulin le els. Insulin le els we e non-signi ican ly dec eased. (
K
) Fas ing
plasma iglyce ides le els we e dec eased in mice ea ed wi h lep olide. (
L
) Li e TG ( iglyce ide)
con en was dec eased in pa allel wi h plasma iglyce ide le els. Values a e he means
±
S.E.M. o
n= 12 pe g oup. * p< 0.05 e sus ehicle by S uden ’s - es ; $p< 0.05 e sus SD by ANOVA.
To u he in es iga e he impac o p olonged adminis a ion o lep olide on obese mice, he
ac i a ion o he in acellula insulin signaling pa hway was assessed in he li e and skele al muscle o
obese mice. As shown in Figu e 6, lep olide imp o ed a basal and insulin-s imula ed condi ions, wi h
he phospho yla ion o PKB in li e and skele al muscle issue. These esul s a e in good ag eemen
wi h he e ec o lep olide on insulin sensi i i y in obese mice.