1
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
www.na u e.com/scien i ic epo s
The cannabinoid ligand LH-21
educes anxie y and imp o es
glucose handling in die -induced
obese p e-diabe ic mice
Sil ana Y. Rome o-Ze bo 1,2, Inmaculada Ruz-Maldonado1,2,3, Vanesa Espinosa-Jiménez 1,
Alex Ra acho 4, Ana I. Gómez-Conde5, Lou des Sánchez-Salido5, Nadia Cobo-Vuilleumie 6,
Benoi R. Gau hie 6, F ancisco J. Tinahones1,7, Shan a J. Pe saud3 & F ancisco J. Be múdez-
Sil a 1,2
LH-21 is a iazol de i a i e ha has been desc ibed as a low-pe mean neu al CB1 an agonis , hough
i s pha macology is s ill unclea . I has been associa ed wi h an i-obesi y ac ions in obese a s. Howe e ,
i s ole in p e en ing ype 2 diabe es (T2D) onse ha e no been s udied ye . Gi en CB1 ecep o s
emain as po en ial pha macological a ge s o igh agains obesi y and T2D, we wan ed o explo e
he me abolic impac o his compound in an animal model o obesi y and p e-diabe es as well as he
lack o ele an ac ions in ela ed cen al p ocesses such as anxie y. C57BL/6J mice we e ende ed obese
and p e-diabe ic by eeding a high- a die o 15 weeks and hen ea ed wi h LH-21 o ehicle o wo
weeks. Food in ake, body weigh and glucose handling we e assessed, oge he wi h o he ele an
pa ame e s. Beha iou al pe o mance was e alua ed by he open ield es and he ele a ed plus
maze. LH-21 did no a ec ood in ake no body weigh bu i imp o ed glucose handling, displaying
issue-speci ic bene icial ac ions. Unexpec edly, LH-21 induced anxiolysis and e e ed obesi y-induced
anxie y, appa en ly h ough GPR55 ecep o . These esul s sugges ha LH-21 can be a new candida e
o igh agains diabe es onse . Indeed, his compound shows po en ial in coun e ac ing obesi y-
ela ed anxie y.
Obesi y and ela ed co-mo bidi ies a e eaching pandemic p opo ions wo ldwide. In 2014, 11% o men and
15% o women age 18 and olde we e obese, while mo e han 42 million child en unde he age o i e yea s we e
o e weigh in 20131. The leading isk ac o s o ype 2 diabe es (T2D) a e excess body weigh and physical inac-
i i y, and T2D is highly co ela ed wi h he global p e alence o obesi y, which has nea ly doubled since 19801.
Cu en ly obesi y- ela ed diseases cons i u e a hea y bu den o heal h sys ems, so i is no su p ising he e is an
inc easing demand o inno a i e he apeu ic in e en ion o alle ia e hese disabling condi ions.
Since he disco e y o he ano exigenic ac ions o cannabinoid CB1 ecep o blockade and he bene icial
impac i has on body weigh homeos asis and me abolic con ol, he e has been an in ense basic and clinical
esea ch e o aiming a de eloping and ma ke ing new an i-obesi y agen s based on his ecep o 2.
1Unidad de Ges ión Clínica In e cen os de Endoc inología y Nu ición, Ins i u o de In es igación Biomédica de
Málaga (IBIMA), Hospi al Regional Uni e si a io de Málaga, Uni e sidad de Málaga, Málaga, Spain. 2Cen o de
In es igación Biomédica en Red de Diabe es y En e medades Me abólicas Asociadas (CIBERDEM), Málaga, Spain.
3Diabe es Resea ch G oup, Di ision o Diabe es & Nu i ional Sciences, Facul y o Li e Sciences & Medicine, King’s
College London, Guy’s Campus, London, SE1 1UL, UK. 4Depa men o Physiological Sciences, Cen e o Biological
Sciences, Fede al Uni e si y o San a Ca a ina (UFSC), 88040-900, Flo ianópolis, SC, B azil. 5Bioimaging acili y,
Ins i u o de In es igación Biomédica de Málaga (IBIMA), Hospi al Regional Uni e si a io de Málaga, Uni e sidad
de Málaga, Málaga, Spain. 6S em Cells Depa men , Andalusian Cen e o Molecula Biology and Regene a i e
Medicine (CABIMER), Se ille, Spain. 7Cen o de In es igación Biomédica en Red de Obesidad y Nu ición
(CIBEROBN), Málaga, Spain. Inmaculada Ruz-Maldonado and Vanesa Espinosa-Jiménez con ibu ed equally o his
wo k. Co espondence and eques s o ma e ials should be add essed o S.Y.R.-Z. (email: yanina. ome o@ibima.
eu) o F.J.B.-S. (email: ja ie [email p o ec ed])
Recei ed: 6 Decembe 2016
Accep ed: 26 Ap il 2017
Published: xx xx xxxx
OPEN
www.na u e.com/scien i ic epo s/
2
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
The p o o ypical molecule o his ype o compound is imonaban , which was in ac ma ke ed in Eu ope and
many o he coun ies in 2006 as an an i-obesi y ea men (Acomplia®), as an adjunc o die and exe cise o he
ea men o obese pa ien s o o e weigh pa ien s wi h associa ed isk ac o s. Howe e , due o psychia ic side
e ec s, imonaban was wi hd awn om he ma ke in 20092, 3. In e es ingly, soon a e wa ds i became e iden
ha blockade o cen al CB1 ecep o s was no equi ed o achie e me abolic bene i s4, 5, and he de elopmen
o a second gene a ion o CB1-based d ugs ocused on pe iphe al an agonism was in ensi ied6. O e he las ew
yea s se e al pe iphe al-ac ing CB1 an agonis s as well as o he cannabinoid-based agen s ha e been desc ibed.
Examples a e he compounds AM65457, JD50378, TM388379 and LH-2110.
LH-21, 5-(4-chlo ophenyl)-1-(2,4-dichlo ophenyl)-3-hexyl-1H-1,2,4- iazole, was ini ially desc ibed as an
in i o CB1 an agonis wi h a pa adoxic low a ini y in i o o CB1 ecep o s, low pene a ion in o he b ain
and being de oid o in e se agonis p ope ies10. Howe e , i s pha macology isn’ clea and o he au ho s ha e
epo ed ex a-CB1 e ec s o LH-21 and i s abili y in c ossing he blood-b ain-ba ie (BBB)11. This compound
has been ound o dec ease eeding beha iou in ood-dep i ed a s and o educe ood in ake and body weigh
gain in a gene ic model o obesi y, he obese Zucke a 12. Fu he mo e, i educed eeding and body weigh gain
in Wis a a s ed a high- a die 13. Howe e , he bene icial ac ions o his compound in diabe es onse as well as
he pu a i e mechanisms in ol ed ha e ye o be cla i ied. Mo eo e , despi e his compound being able o c oss
he BBB o some ex en , i s pu a i e modula o y ac ions o obesi y- ela ed anxie y a e unknown.
To add ess hese unexplo ed a eas we ha e ea ed a mouse model o die -induced obesi y and p e-diabe es
o wo weeks wi h LH-21 (3 mg/Kg). Body weigh gain, ood in ake, glucose homeos asis, sys emic in lamma o y
ma ke s as well as panc eas and li e his opa hology we e s udied. In addi ion, a beha iou al s udy o he e ec s
o LH-21 on obesi y- ela ed anxie y was pe o med.
Resul s
LH-21 imp o es glucose handling in obese p e-diabe ic mice. The p esen wo k was designed o
e alua e he ac ions o he cannabinoid ligand LH-21 in diabe es p e en ion, wi h a ocus on i s e ec s on he
panc eas and li e , and de e mine i s e ec s on obesi y- ela ed anxie y. A gene al o e iew o he s udy design is
depic ed in Fig.1. In b ie , acco ding he li e a u e a mouse model o die -induced obesi y14 was used o assess
he e ec s o subch onic ea men wi h LH-21 (3 mg/Kg/day o wo weeks) on body weigh , ood in ake, glu-
cose homeos asis and in lamma o y s a us, and a beha iou al s udy was also pe o med on hese animals. LH-21
dosage was selec ed acco ding o p e ious wo ks using his compound12, 13, 15. As expec ed, mice ed a high- a -
die (HFD; 45% ene gy in ake om sa u a ed a ) became obese when compa ed o hose ed a s anda d- a die
(10% ene gy in ake om a ), wi h di e ences in body weigh clea ly e iden a week 11 o he die a y in e en-
ion (Supplemen a y Figu e1A), being close o 20% inc ease o e con ol a week 15, as expec ed16. Mo eo e ,
impai ed glucose ole ance and educ ion in insulin sensi i i y we e de ec ed a e 8 and 11 weeks o HFD,
espec i ely, wi h no changes in as ing glucose (Supplemen a y Figu e1B,C), sugges ing al e a ions o glucose
homeos asis compa ible wi h a p e-diabe ic s age. Obese p e-diabe ic mice ea ed daily o up o 2-weeks wi h
LH-21 (3 mg/Kg b.w.) did no display weigh loss no dec eased ood in ake as compa ed o ehicle ea ed mice
(Fig.2A,B). They emained glucose in ole an when compa ed o ehicle-injec ed mice as assessed by glucose
a ea-unde - he-cu e (AUC) calcula ion (ANOVA analysis o he en i e ime ame, Fig.2C inse ), hough
dec eased glucose alues we e ound a ime poin 30 min a e glucose o e load (S uden ’s es ), i.e. he e was a
lowe pla eau in glucose excu sion (Fig.2C). Fu he mo e, a endency o LH-21 o dec ease as ing glucose was
ound (Vehicle: 108 ± 4 mg/dL e sus LH-21: 98 ± 7 mg/dL, p = 0.193 S uden ’s es ) and inc eased insulinemia
and HOMA-IR (homeos a ic model assessmen -insulin esis ance) we e no p esen in LH-21-injec ed mice
(Fig.2D,E). Rega ding endoc ine panc eas unc ion, isle s isola ed om HFD mice ha had been ea ed o wo
weeks wi h LH-21 we e submi ed o glucose-s imula ed insulin sec e ion (GSIS) expe imen s. Two-way ANOVA
Figu e 1. S udy design. Ou line ep esen ing he design o he p esen s udy. D awings a e unde C ea i e
Commons license CC0 om Pixabay.com.
www.na u e.com/scien i ic epo s/
3
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
Figu e 2. Me abolic cha ac e iza ion o obese p e-diabe ic LH-21- ea ed mice. (A) Body weigh changes du ing
he wo weeks o ea men wi h 3 mg/Kg LH-21. Body weigh was no a ec ed by he ea men and bo h ehicle-
and LH-21-injec ed mice emained obese when compa ed o he con ol g oup. (B) Food in ake was moni o ed
daily, he line and he ba g aphs ep esen he mean daily in ake and he mean o al in ake o ene gy (Kcal),
espec i ely, du ing he wo week pe iod. No changes in calo ic in ake was de ec ed be ween ehicle- and LH-21-
injec ed mice, bo h g oups displaying a highe calo ic in ake han he con ol g oup. (C) Glucose ole ance was
assessed by an i.p. GTT in as ed mice injec ed wi h 2 g/Kg glucose. AUC analysis o he en i e ime ame e ealed
glucose in ole ance in bo h ehicle- and LH-21- injec ed mice, hough lowe plasma glucose le els we e ound in
LH-21-injec ed mice a ime poin 30 minu es a e glucose challenge (S uden ’s es ). (D) Insulin plasma le els in
as ed mice as assessed by ELISA. 45% HFD- ehicle mice had highe insulinemia bu no hose ea ed wi h LH-
21. (E) HOMA-IR was used as a sub oga e measu e o insulin esis ance (HOMA-IR = insulin (mU/l) × glucose
(mmol/l)/22.5). 45% HFD- ehicle mice had highe HOMA-IR bu no hose ea ed wi h LH-21. (F) Glucose-
s imula ed insulin sec e ion om isle s isola ed om obese p e-diabe ic mice injec ed wi h ehicle o LH-21.
Isle s isola ed om LH-21- ea ed mice sec e ed less insulin han hose injec ed wi h ehicle wi h his di e ence
being pa icula ly ma ked in he high glucose condi ion. (A–E) n = 8–10 mice each g oup. One-way ANOVA and
Bon e oni’s pos - es , *p < 0.05, **p < 0.01 e sus con ol; #p < 0.05 e sus 45% HFD-Vehicle. (F) Isle s om wo
di e en mice om each g oup, wo independen expe imen s, n = 6 expe imen al wells o each condi ion, wo-
way ANOVA wi h Bon e oni’s pos - es ; *p < 0.05 e sus 3 mM Glc, ##p < 0.01 and ###p < 0.001.
www.na u e.com/scien i ic epo s/
4
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
analysis e ealed ha isle s om LH-21- ea ed mice showed a educ ion in insulin elease when compa ed o
isle s om ehicle-injec ed mice (Fig.2F). Indeed, dec eased sec e ion was ound in esponse o 11 mM glucose
in isle s om mice ea ed wi h LH-21 (Fig.2F, Bon e oni’s pos - es ). These indings, oge he wi h dec eased
insulinemia, dec eased insulin esis ance and lowe glucose le els a speci ic ime poin s du ing he glucose ol-
e ance es (GTT) (Fig.2C–E) sugges LH-21 imp o es glucose handling by a ou ing pe iphe al glucose up ake.
LH-21 displays se e al an i-in lamma o y and cy op o ec i e ac ions. A panel o in lamma o y
cy okines was assayed on plasma samples om hese mice oge he wi h ELISA measu emen o he le els o
he adipokines lep in and adiponec in (Table1). Plasma lep in, high-molecula -weigh (HMW) adiponec in,
IL-6 and he chemokine (C-X-C mo i ) ligand 1 (CXCL1) le els we e ound o be nega i ely impac ed in obese
p e-diabe ic mice, sugges ing a sys emic p o-in lamma o y s a e (Table1). Bo h lep in and CXCL1 le els we e
sligh ly dec eased a e LH-21 ea men , he la e no being s a is ically di e en when compa ed o he con-
ol g oup while no e ec on adiponec in and IL-6 we e de ec ed. Mo eo e , no changes among g oups we e
de ec ed o IL-5, IL-2, IL-12p70, IL-10 and IFN-γ (Table1). LH-21 ea men was no associa ed wi h a dec ease
in obesi y-induced mac ophage in il a ion o he isle s (Fig.3A) and did no e e he obesi y-induced dec ease
in M2 mac ophages, as assessed by M c-1 and CD163 immunos aining (Supplemen a y Figu e3A,B). Howe e ,
obesi y-induced apop osis in isle s was e e ed by LH-21 (Fig.3B). In he li e , inc eased ed-oil s aining was
obse ed in HFD-mice when compa ed o con ol mice, sugges ing ec opic accumula ion o a (Fig.3C). Indeed,
bo h mac ophage in il a ion and apop osis we e inc eased in he li e s o obese p e-diabe ic mice (Fig.3D,E).
Mac ophage in il a ion was dec eased in he li e by LH-21 (Fig.3D) wi hou modi ying he le els o he M2
sub ype (Supplemen a y Figu e3C,D), bu nei he li e s ea osis no li e apop osis we e amelio a ed by LH-21
(Fig.3C and E).
LH-21 p omo es anxiolysis in obese p e-diabe ic mice and e e s obesi y-induced anxi-
e y. Gi en he con lic ing epo s on he abili y o LH-21 o c oss he blood-b ain-ba ie , and in he con ex
o s udying he me abolic ac ions o LH-21 on obese p e-diabe ic mice, we decided o explo e he pu a i e ac ions
o his compound in beha iou , a bo h he acu e and subch onic le els. Locomo ion and anxie y we e assessed
in d ug naï e obese p e-diabe ic mice as well as a e acu e and subch onic ea men wi h LH-21 by analyzing
pe o mance on he open ield es and he ele a ed plus maze es . No changes in locomo ion we e de ec ed in
obese p e-diabe ic mice (Figs4A–D and S2A–D). Howe e , obese p e-diabe ic mice displayed dec eased ime and
dis ance a elled in open a ms du ing he ele a ed plus maze es (Figs4F,G and S2F,G), sugges ing he p esence
o anxie y in hese mice. Mo e impo an ly, acu e adminis a ion o LH-21 inc eased bo h ime spen and dis ance
a elled in he cen e du ing he open ield es pe o mance (Fig.4B,C), sugges ing an anxioly ic e ec . In ac ,
he anxie y ai s displayed by obese p e-diabe ic mice we e o ally e e ed by acu e adminis a ion o LH-21
(Fig.4E–G). No e ec o LH-21 a e subch onic ea men was ound in open ield es pe o mance (Fig.5A–D)
in con as o wha was ound a e LH-21 acu e injec ion (Fig.4A–D). Howe e , i is no ewo hy ha subch onic
ea men o obese p e-diabe ic mice wi h LH-21 e e ed he obesi y-induced anxie y, as did acu e injec ion o
LH-21, inc easing numbe o en ies, ime spen and mean dis ance a elled in open a ms when compa ed o
ehicle-injec ed obese p e-diabe ic mice (Fig.5E–G).
LH-21 induces beha iou al changes in mice h ough GPR55 ecep o . The abili y o LH-21 in
coun e ac ing obesi y-induced anxie y is appa en ly media ed h ough CB1-independen mechanisms, gi en CB1
blockage is known o p omo e anxie y. Indeed, we ha e p e ious e idence o GPR55 media ing some me abolic
ac ions o LH-2117. The e o e, we pos ula ed ha cen al LH-21-d i en ac ions migh be media ed h ough his
ecep o . To challenge his hypo hesis we ca ied ou pha macological expe imen s in naï e heal hy mice ha
we e injec ed wi h 3 mg/kg o LH-21 and 3 mg/kg o he GPR55 an agonis CID16020046. Subsequen beha -
iou al analysis was pe o med in he open ield and he ele a ed plus maze. LH-21 dosage was selec ed acco ding
o p e ious wo ks using his compound, as s a ed abo e, and CID dosage was selec ed acco ding o p e ious
expe imen s pe o med in ou lab. LH-21 dec eased en ies and ime in o open a ms in he ele a ed plus maze,
Sys emic In lamma o y Ma ke s CONTROL 45% HFD VEH 45% HFD LH-21
Lep in (pg/ml) 4121 ± 749 19550 ± 8501** 13720 ± 7917*
HMWAdiponec in (ng/ml) 5397 ± 2679 2864 ± 746*2942 ± 1100*
IL-6 (pg/ml) 51,7 ± 13,0 211,4 ± 107,6*183,4 ± 165,9*
CXCL1 (pg/ml) 137,8 ± 29,9 383,9 ± 209,8*232,4 ± 169,5
IL-5 (pg/ml) 5,46 ± 1,78 8,99 ± 4,26 12,95 ± 11,67
IL-2 (pg/ml) 5,31 ± 1,59 5,20 ± 1,03 5,56 ± 1,46
IL-12 p70 (pg/ml) 61,41 ± 10,53 78,09 ± 19,51 95,73 ± 31,61
IL-10 (pg/ml) 23,23 ± 16,48 22,45 ± 11,86 26,11 ± 11,12
IFN-γ (pg/ml) 1,31 ± 0,71 1,45 ± 0,44 1,77 ± 0,68
Table 1. Sys emic in lamma o y ma ke s in obese p e-diabe ic LH-21- ea ed mice. The le els o lep in and
HMW adiponec in in plasma we e measu ed wi h a comme cial ELISA. A mouse p o-in lamma o y panel ki
was used o he quan i a i e de e mina ion o IFN-γ, IL-2, IL-5, IL-6, CXCL1, IL-10 and IL-12p70 by mul i-
a ay elec ochemiluminescence de ec ion echnology. n = 8–10 mice each g oup. One-way ANOVA and
Bon e oni’s pos - es , *p < 0.05, **p < 0.01 e sus con ol.
www.na u e.com/scien i ic epo s/
5
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
Figu e 3. E ec s o LH-21 in he isle s o Lange hans and he li e . (A) Mac ophage in il a ion in he isle s was
assessed by immunohis ochemis y wi h he speci ic ma ke F4/80. Quan i ica ion o immunos aining wi hin
isle s e ealed inc eased mac ophage in il a ion in 45% HFD mice, wi h no di e ences be ween ehicle- and
LH-21-injec ed mice. The images a e ep esen a i e o isle s om ou mice each g oup and ou di e en sec ions
om each panc eas; one-way ANOVA and Bon e oni’s pos - es , *p < 0.05 e sus con ol. (B) Apop osis in isle
cells was assessed by a speci ic apop osis ki . The numbe o apop o ic cells was inc eased in isle s om 45% HFD-
ehicle mice, while he e was no inc eased apop osis in isle s om 45% HFD-LH-21 mice when compa ed o
con ol mice. (C) Fa accumula ion in he li e was s udied by ed-oil O s aining o hepa ic sec ions. Quan i ica ion
o his speci ic ma ke e ealed inc eased a y con en in he li e o 45% HFD mice, bo h ehicle- and LH-21-
injec ed mice. (D) Mac ophage in il a ion in he li e was assessed by immunohis ochemis y wi h he speci ic
ma ke F4/80. These sec ions we e coun e s ained wi h haema oxylin. Quan i ica ion o immunos aining in he
li e e ealed inc eased mac ophage in il a ion in 45% HFD- ehicle when compa ed o con ol- ehicle mice, bu
no inc ease was de ec ed in 45% HFD-LH-21 mice. (E) Apop osis in he li e was assessed by a speci ic apop osis
ki . The numbe o apop o ic cells was inc eased in he li e s o bo h 45% HFD- ehicle and 45% HFD-LH-21
mice when compa ed o con ol- ehicle mice. (A,B) The images a e ep esen a i e o isle s om ou mice in each
g oup and ou di e en sec ions om each panc eas; one-way ANOVA and Bon e oni’s pos - es , ***p < 0.001
e sus con ol, ###p < 0.001 e sus 45% HFD- ehicle. (C-E) The images a e ep esen a i e o li e s om ou mice
in each g oup and ou di e en sec ions om each li e ; one-way ANOVA and Bon e oni’s pos - es , *p < 0.05,
**p < 0.01 e sus con ol- ehicle, #p < 0.05 e sus 45% HFD- ehicle.

www.na u e.com/scien i ic epo s/
6
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
Figu e 4. Beha iou al s udy in obese p e-diabe ic mice acu ely ea ed wi h LH-21. (A–D) Explo a o y ac i i y
was eco ded o en minu es in he open ield maze in con ol, 45% HFD- ehicle and 45% HFD-LH-21 mice.
En ies in o cen e (A), ime in cen e (B), dis ance in cen e (C) and o e all dis ance a elled (D) we e
moni o ed. No changes we e de ec ed be ween con ol and 45% HFD- ehicle mice. Howe e , acu e LH-21
injec ions inc eased explo a o y ac i i y in he cen e in 45% HFD mice, as assessed by inc eased ime and
dis ance a elled in cen e ; (E–G) Anxie y was analysed by he ele a ed plus maze es in hese mice. En ies
in o open a m (E), ime in open a ms (F) and dis ance a elled in open a ms (G) we e eco ded o i e
minu es. 45% HFD- ehicle mice showed an anxie y-like beha iou wi h dec eased ime in open a ms and
dis ance a elled in open a ms when compa ed o con ol mice. Acu e injec ions o LH-21 in 45% HFD mice
inc eased he numbe o en ies in o open a ms and e e ed he HFD-induced dec ease in ime in open a ms
and dis ance a elled in open a ms. n = 8–10 mice each g oup. One-way ANOVA and Bon e oni’s pos - es ,
*p < 0.05 e sus con ol; #p < 0.05, ##p < 0.01 e sus 45% HFD-Vehicle.
www.na u e.com/scien i ic epo s/
7
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
Figu e 5. Beha iou al s udy in obese p e-diabe ic mice subch onically ea ed wi h LH-21. (A–D) Explo a o y
ac i i y was eco ded o en minu es in he open ield maze in con ol, 45% HFD- ehicle and 45% HFD-LH-21
mice. En ies in o cen e (A), ime in cen e (B), dis ance in cen e (C) and o e all dis ance a elled (D) we e
moni o ed. No changes among g oups we e de ec ed. (E–G) Anxie y was analysed by he ele a ed plus maze es
in hese mice. En ies in o open a m (E), ime in open a ms (F) and dis ance a elled in open a ms (G) we e
eco ded o i e minu es. 45% HFD- ehicle mice showed an anxie y-like beha iou , wi h dec eased ime in
open a ms and dis ance a elled in open a ms when compa ed o con ol mice. Subch onic ea men o obese
p e-diabe ic mice wi h LH-21 induced an anxioly ic-like beha iou wi h inc eased numbe o en ies in o open
a ms and LH-21 e e ed he HFD-induced dec ease in ime in open a ms and dis ance a elled in open a ms.
n = 8–10 mice each g oup. One-way ANOVA and Bon e oni’s pos - es , *p < 0.05 e sus Con ol-Vehicle,
#p < 0.05 e sus 45% HFD-Vehicle.
www.na u e.com/scien i ic epo s/
8
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
sugges ing anxiogenesis (Fig.6B). Indeed, a endency o dec ease dis ance a elled in open a ms in he plus
maze as well as less cen al squa e ac i i y in he open ield es was e iden (Fig.6). In addi ion, LH-21 dec eased
o al locomo ion in he open ield es (Fig.6A). These beha iou al e ec s o LH-21 poin s o a di e en esponse
in heal hy and obese p e-diabe ic mice. Howe e , and in e es ingly, all hese beha iou al e ec s o LH-21 we e
p e en ed by adminis a ion o he GPR55 an agonis CID16020046, sugges ing ha cen al LH-21 ac ions a e
media ed h ough his ecep o (Fig.6).
Discussion
LH-21 was desc ibed as a silen CB1 an agonis wi h poo b ain pe meabili y10. This pha macological p o ile has
led o LH-21 being conside ed pa o he new class o p omising pe iphe al CB1 an agonis s3. Howe e , i s pha -
macology is s ill unclea and i s po en ial ole in delaying T2D onse as well as he unde lying mechanisms ha e
no been explo ed ye . He ein we demons a e ha subch onic LH-21 ea men : (1) does no induce body weigh
loss in a mouse model o obesi y and p e-diabe es, (2) imp o es diabe es isk ac o s such as glucose handling
and issue in lamma ion, (3) displays cy op o ec i e ac ions on bo h panc ea ic isle s and he li e by dec easing
apop osis and mac ophage in il a ion in hese issues, espec i ely, which sugges s issue-speci ic ac ions ha
con e esis ance in he long- e m o he onse o diabe es and s ea ohepa i is and, (4) con eys he abili y o coun-
e ac obesi y- ela ed anxie y, which could ha e implica ions in he managemen o obesi y-induced anxie y and
ela ed diso de s.
The C57Bl/6J s ain o mice used in hese s udies is p one o de elop obesi y and me abolic dis u bances a e
being ed a HFD o se e al weeks14. We ha e main ained his mouse s ain on a HFD o 15 weeks, which led
o he de elopmen o a pheno ype mimicking ha o human obesi y and p e-diabe es. Thus, mice de eloped
obesi y, glucose in ole ance, insulin esis ance, hype insulinemia, hype lep inemia, hypoadiponec inemia, sys-
emic low-g ade in lamma ion (inc eased le els o he p o-in lamma o y cy okines IL-6 and CXCL1), a y li e ,
mac ophage in il a ion in he li e and isle s o Lange hans and inc eased apop osis in hese issues. These obese
p e-diabe ic mice we e subch onically ea ed wi h 3 mg/Kg/day LH-21, a dose ha has been p e iously used
in o he s udies ocused on an i-obesi y ac ions o LH-2112, 13, 15. We did no ind body weigh loss o dec eased
ood in ake du ing LH-21 ea men , in con as o wha has p e iously been epo ed ollowing LH-21 deli -
e y o a s ed a HFD13. This disc epancy migh be ela ed o he di e en animal model (Wis a a s e sus
C57Bl/6J mice) and/o he obesi y induc ion p o ocol used (60% HFD o en weeks e sus 45% HFD o i een
weeks). Howe e , and in ag eemen wi h his p e ious s udy, we did no de ec no maliza ion in a con en in
he li e be ween LH-21- and Vehicle- ea ed mice, as assessed by changes in ed-oil O s aining. In e es ingly,
despi e LH-21 no inducing educ ions in ood in ake and body weigh in ou model, ou esul s poin s o LH-21
Figu e 6. Beha iou al s udy in heal hy mice acu ely ea ed wi h LH-21 and CID16020046. (A) Explo a o y
ac i i y was eco ded o en minu es in he open ield maze in Veh-, LH-21- and CID + LH21-injec ed heal hy
mice. En ies in o cen e , ime in cen e , dis ance in cen e and o e all dis ance a elled we e moni o ed. Acu e
LH-21 injec ions ended o dec ease explo a o y ac i i y in he cen e , and dec eased o al explo a o y ac i i y.
These e ec s we e o ally e e ed by CID16020046 p e-injec ion. (B) Anxie y was analysed by he ele a ed
plus maze es in hese mice. En ies in o open a m, ime in open a ms and dis ance a elled in open a ms
we e eco ded o i e minu es. LH-21-injec ed mice showed an anxie y-like beha iou wi h dec eased en ies
and ime in o open a ms when compa ed o con ol Veh-injec ed mice. These anxiogenic e ec s o LH-21 we e
o ally e e ed by CID16020046 p e-injec ion. n = 6–8 mice each g oup. One-way ANOVA and Bon e oni’s
pos - es , *p < 0.05 e sus con ol; #p < 0.05, ##p < 0.01 e sus LH-21-injec ed mice.
www.na u e.com/scien i ic epo s/
9
Scien i ic RepoR s | 7: 3946 | DOI:10.1038/s41598-017-03292-w
inducing a me abolic imp o emen in ele an pa ame e s such as glucose handling, dec eased insulin sec e-
ion, dec eased hype insulinemia, dec eased HOMA-IR index and a endency o dec ease as ing glucose. In
addi ion, LH-21 ea men also sligh ly amelio a ed he HFD-induced low-g ade in lamma ion, wi h dec eased
le els o he cy okines lep in and CXCL1. Fu he mo e, lep in is also an impo an me abolic egula o wi h
ano exic and weigh - educing e ec s whose sensi i i y is blun ed du ing hype -lep inemia (lep in esis ance)
and is es o ed when plasma lep in is educed18. Taken oge he , hese indings sugges ha LH-21 could con e
some deg ee o p o ec ion agains diabe es de elopmen du ing obesi y. In u he suppo o his hypo hesis, we
ound se e al cy op o ec i e ac ions o LH-21 in li e and panc ea ic isle s. Speci ically, HFD-induced apop osis
was e e ed in he isle s, and mac ophage in il a ion was dec eased in he li e . LH-21 did no a ec he numbe
o epai e M2 mac ophages in any issue, hus sugges ing ha bene icial ac ions o LH-21 in he li e migh be
media ed h ough down egula ion o kille M1 mac ophages. In ag eemen wi h ou indings, lowe le els o
clea ed caspase-3 ha e been ound in isle s o s ep ozo ocin -injec ed mice a e ea men wi h he CB1 an ag-
onis /in e se agonis AM25119 and CB1 in mac ophages has also been linked o bo h isle and li e damage20, 21.
Howe e , he eason o he di e en ial esponse be ween hese wo issues is unknown and could be ela ed o
he le el o exp ession and/o he speci ic cell ype exp essing he a ge ecep o . In e es ingly, a ecen s udy wi h
JD5037, a pe iphe al an agonis /in e se agonis o CB1 ecep o s, has shown po en and speci ic ac ions o his
compound a hepa ocy es and β-cells by a ge ing he CB1b iso o m22, unde lining he a ie y o ac ions o CB1
an agonis s.
An ea lie s udy wi h AM6545, a neu al CB1 an agonis wi h educed b ain pene ance, indica ed ha i was
less e ec i e han imonaban in educing body weigh , adiposi y, insulin esis ance, and hype lep inemia, and
i had minimal e ec on ood in ake7. La e , JD5037 was ound o obus ly educe ood in ake, body weigh , and
adiposi y wi hou a ge ing b ain CB1 ecep o s8. F om he abo e s udies, i has been concluded ha in e se ago-
nism is a bene icial pha macological p ope y in pe iphe al CB1 an agonis s o imp o ing me abolism8. LH-21
was ini ially desc ibed as a neu al CB1 an agonis wi h poo b ain pene ance and a pa adoxically low a ini y
in i o10, hough i was la e epo ed as a b ain-pe mean , weak CB1 in e se agonis 11. Ou esul s pinpoin s
o speci ic an i-diabe ic ac ions o LH-21 ha a e p obably de i ed om i s abili y o an agonize, wi h lowe
a ini y han o he an agonis s, he CB1 ecep o s, also possibly ha ing a ole hei weak in e se agonism p op-
e ies. Fu he mo e, we ha e ecen ly ound ha some me abolic ac ions o LH-21 could be media ed h ough
he cannabinoid-sensi i e ecep o GPR5517 and his ecep o is known o be exp essed in pe iphe al me abolic
issues and in he hypo halamus among o he b ain egions23–26, being egula ed by nu i ional s a us and o he
physiological p ocesses in ol ed in ene gy balance27. Indeed, lack o GPR55 has been linked o inc eased adipos-
i y and insulin esis ance28. In e es ingly, acu e ood in ake expe imen s wi h a single high dose o LH-21 (60 mg/
kg) in WT and CB1-KO mice ha e e ealed a CB1-independen dec ease o o e nigh eeding and body weigh
gain11. So, i is plausible ha , oge he wi h CB1-media ed ac ions, some o he me abolic e ec s o LH-21 a e
media ed h ough GPR55.
Obesi y is associa ed wi h an inc eased isk o anxie y in humans29. This ela ionship has also been demon-
s a ed in obese mice ed a HFD30, 31, which also displayed unc ional al e a ions and neu al adap a ions in
GABAe gic do somedial hypo halamic neu ons and b ain ewa d ci cui y, espec i ely. P e ious s udies ha e
sugges ed ha LH-21 has lowe BBB pe meabili y han o he CB1 an agonis s, like imonaban , and ha acu e
e ec s o LH-21 on eeding a e no associa ed wi h anxie y-like beha iou s12, 15. Howe e , he e is e idence o
LH-21 c ossing he BBB o some ex en . Fo example, i.p. adminis a ion o LH-21 was ound o sligh ly dec ease
e hanol sel -adminis a ion12, LH-21 pa ially an agonized mo o dep ession induced by cen al adminis a ion
o a cannabinoid ecep o agonis (CP55940)12 and ela i ely high concen a ions o LH-21 we e de ec ed in b ain
ollowing an in a enous injec ion o he compound11. In e es ingly, he ex en o which LH-21 can modula e
cen al p ocesses such as anxie y has no been explo ed ye . He e we show ha LH-21, whe he unde acu e o
subch onic ea men , educed HFD-induced anxie y beha io by inc easing he ime and dis ance a elled in
he open a ms du ing he ele a ed plus maze pe o mance. Mo eo e , he numbe o en ies in o he open a ms
we e also inc eased in LH-21- ea ed mice. In he case o acu e injec ions, LH-21 e en inc eased ime in he
cen e and dis ance in he cen e in he open ield es when compa ed o con ol mice, hus showing a po en
anxioly ic e ec in obese p e-diabe ic mice. These e ec s we e no due o hype -locomo ion as he o e all dis-
ance a elled was no a ec ed by he ea men . Taken oge he , hese expe imen s sugges ha LH-21 can c oss
he BBB in su icien amoun s o induce pe se beha iou al changes, also modula ing obesi y-induced anxie y.
Gi en ha CB1 an agonis s ha e been unequi ocally desc ibed as anxiogenic d ugs and, speci ically, hey ha e
been ound o dec ease explo a ion on he ele a ed plus-maze32 ou esul s sugges ha hese cen al e ec s o
LH-21 a e no media ed h ough CB1 ecep o s. In e es ingly, as s a ed abo e, we ha e ecen ly ound ha some
me abolic ac ions o LH-21 could be media ed h ough GPR5517, and his ecep o is also known o be exp essed
a he cen al ne ous sys em25, 26, 33. Speci ically, he nucleus accumbens and he hypo halamus a e among
he b ain egions wi h highe GPR55 exp ession25, 26, bo h o hem being in ol ed in anxiogenesis30, 31. Thus,
anxie y- ela ed e ec s o LH-21 could be media ed h ough cen al GPR55 ecep o s. To challenge his hypo h-
esis we pe o med a beha iou al s udy in heal hy mice acu ely ea ed wi h LH-21 and he GPR55 an agonis
CID16020046. In e es ingly, CID16020046 p e en ed LH-21-induced beha iou al changes, hus s ongly sugges -
ing a key ole o GPR55 in media ing he LH-21-d i en changes in anxie y. Ne e heless, LH-21 induced anxio-
genesis in heal hy mice, an e ec ha was he opposi e o ha exe ed on obese p e-diabe ic mice. Fu he mo e,
LH-21 dec eased explo a o y ac i i y in heal hy mice, in con as wi h obese p e-diabe ic mice. These opposing
e ec s be ween heal hy and obese p e-diabe ic mice could be ela ed o changes in he exp ession o GPR55
du ing obesi y de elopmen , in pa allel o he well-known changes ha ake place in he endocannabinoid sys-
em du ing his disease’s de elopmen 34–36. Indeed, he endocannabinoid sys em is unc ionally connec ed wi h
GPR5537–39. Mo eo e , in ag eemen wi h ou indings in heal hy mice, al e a ions in physical ac i i y ha e been
epo ed in lean mice lacking GPR5528. Al hough ou esul s highly suppo he in ol emen o GPR55 in cen al