Thyroid-hormone-induced browning of white adipose tissue does not contribute to thermogenesis and glucose consumption

A icle
Thy oid-Ho mone-Induced B owning o Whi e
Adipose Tissue Does No Con ibu e o
The mogenesis and Glucose Consump ion
G aphical Abs ac
Highligh s
dThy oid ho mone induces b owning independen o
sympa he ic ne ous sys em h ough TRb
dDespi e high UCP1, he beige a lacks ad ene gic inpu and is
inac i e
dThy oid ho mone’s me abolic and he mogenic e ec s a e
main ained in UCP1 knockou mice
dThy oid ho mone induces hype he mia and ele a es he
body empe a u e se poin
Au ho s
Ko nelia Johann, Anna Lena C eme ,
Alexande W. Fische , ...,
Je ey W. Dalley, Heiko Backes,
Jens Mi ag
Co espondence
jens.mi [email protected]
In B ie
Thy oid ho mone induces b owning o
whi e a , bu i is unclea whe he his
con ibu es o he mogenesis. He e,
Johann e al. show ha hy oid-ho mone-
induced beige a is me abolically inac i e
due o lack o cen al s imula ion and ha
he me abolic and he mogenic e ec s o
he ho mone a e independen o UCP1.
Johann e al., 2019, Cell Repo s 27, 3385–3400
June 11, 2019 ª2019 The Au ho (s).
h ps://doi.o g/10.1016/j.cel ep.2019.05.054
Cell Repo s
A icle
Thy oid-Ho mone-Induced B owning
o Whi e Adipose Tissue Does No Con ibu e
o The mogenesis and Glucose Consump ion
Ko nelia Johann,
1
Anna Lena C eme ,
2
Alexande W. Fische ,
3
Ma kus Heine,
3
E a Rial Pensado,
4,5
Julia Resch,
1
Sebas ian Nock,
1
Samuel Vi ue,
6
Lisbe h Ha de ,
1
Rebecca Oelk ug,
1
Ma iana As iz,
7
Geo g B aban ,
1
Amy Wa ne ,
6
An onio Vidal-Puig,
6
Hen ik Os e ,
7
Ani a Boelen,
8
Miguel Lo
´pez,
4,5
Joe g Hee en,
3
Je ey W. Dalley,
9,10
Heiko Backes,
2
and Jens Mi ag
1,11,
*
1
In e nal Medicine I, Molecula Endoc inology, Cen e o B ain, Beha io and Me abolism, Uni e si y o L€
ubeck, 23562 L€
ubeck, Ge many
2
Mul imodal Imaging o B ain Me abolism, Max Planck Ins i u e o Me abolism Resea ch, 50931 Cologne, Ge many
3
Depa men o Biochemis y and Molecula Cell Biology, Uni e si y Medical Cen e Hambu g-Eppendo , 20246 Hambu g, Ge many
4
Neu Obesi y G oup, Depa men o Physiology, CIMUS, Uni e si y o San iago de Compos ela-Ins i u o de In es igacio
´n Sani a ia, San iago
de Compos ela 15782, Spain
5
CIBER Fisiopa ologı
´a de la Obesidad y Nu icio
´n (CIBERobn), San iago de Compos ela 15782, Spain
6
Uni e si y o Camb idge Me abolic Resea ch Labo a o ies, Wellcome T us -MRC Ins i u e o Me abolic Science, Camb idge CB2 0QQ, UK
7
Ins i u e o Neu obiology, Cen e o B ain, Beha io and Me abolism, Uni e si y o L€
ubeck, 23562 L€
ubeck, Ge many
8
Labo a o y o Endoc inology, Ams e dam Uni e si y Medical Cen e s, 1105 Ams e dam, he Ne he lands
9
Depa men o Psychology, Uni e si y o Camb idge, Camb idge CB2 3EB, UK
10
Depa men o Psychia y, Uni e si y o Camb idge, Camb idge CB2 2QQ, UK
11
Lead Con ac
*Co espondence: jens.mi [email p o ec ed]
h ps://doi.o g/10.1016/j.cel ep.2019.05.054
SUMMARY
Regula ion o body empe a u e c i ically depends
on hy oid ho mone (TH). Recen s udies e ealed
ha TH induces b owning o whi e adipose issue,
possibly con ibu ing o he obse ed hype he mia
in hype hy oid pa ien s and po en ially p o iding
me abolic bene i s. He e, we show ha b owning
by TH equi es TH- ecep o band occu s indepen-
den ly o he sympa he ic ne ous sys em. The beige
a , howe e , lacks su icien ad ene gic s imula ion
and is no me abolically ac i a ed despi e high le els
o uncoupling p o ein 1 (UCP1). S udies a di e en
en i onmen al empe a u es e eal ha TH ins ead
causes hype he mia by ac ions in skele al muscle
combined wi h a cen al body empe a u e se -poin
ele a ion. Consequen ly, he me abolic and he mo-
genic e ec s o sys emic hype hy oidism we e
main ained in UCP1 knockou mice, demons a ing
ha nei he beige no b own a con ibu es o he
TH-induced hype he mia and ele a ed glucose
consump ion, and unde lining ha he me e p es-
ence o UCP1 is insu icien o d aw conclusions on
he he apeu ic po en ial o b owning agen s.
INTRODUCTION
Obliga o y he mogenesis is de ined as hea esul ing om
main enance o i al unc ions. In homeo he mic species li ing
in he moneu al condi ions, his ype o he mogenesis is su i-
cien o main ain body empe a u e. Whene e he ambien
empe a u e is educed, hea -sa ing mechanisms, such as
asocons ic ion, dec eased mo emen , and piloe ec ion,
a e ac i a ed, and adap i e (o acul a i e) he mogenesis is
induced. Among hese adap i e he mogenic mechanisms is
he ac i a ion o non-shi e ing he mogenesis in b own adipose
issue (BAT). BAT he mogenesis elies on he exp ession o
uncoupling p o ein 1 (UCP1) ha dissocia es mi ochond ial sub-
s a e oxida ion om ATP syn hesis by in oducing a p o on
leak, esul ing in hea elease (Cannon and Nede gaa d, 2004).
BAT is ac i a ed by he sympa he ic ne ous sys em (SNS)
eleasing no epineph ine (NE) ha binds o b3-ad ene gic e-
cep o s (ADRB3) on b own adipocy es, leading o an inc ease
in in acellula cyclic AMP (cAMP) (Zhao e al., 1994). This ise
in cAMP ac i a es, o ins ance, he con e sion o he hy oid
ho mone hy oxine (T4) o he mo e biologically ac i e o m
3,30,5- iiodo hy onine (T3) by he enzyme deiodinase ype 2
(DIO2) (Sil a and La sen, 1983). T3 in u n inc eases Ucp1
exp ession (Rabelo e al., 1996). BAT he mogenesis has been
shown o ha e nume ous me abolic bene i s, like inc eased
glucose and lipid up ake (Ba el e al., 2011; Be be
´ee al.,
2015; G eco-Pe o o e al., 1987; Shiba a e al., 1989; Heine
e al., 2018), an i-diabe ic e ec s, and amelio a ion o obesi y
(Ba el e al., 2018; Kajimu a e al., 2015; S ensson e al.,
2016). Addi ionally, i was ound ha sus ained cold exposu e
o ce ain compounds can induce b own adipocy e-like (b i e
o beige) cells in subcu aneous whi e adipose issue o mice,
a p ocess called b owning (Ba el and Hee en, 2014; Ishibashi
and Seale, 2010; Pe o ic e al., 2010; Young e al., 1984). I is
cu en ly assumed ha his beige a migh con ibu e o he -
mogenesis and may ha e bene icial me abolic e ec s, al hough
he majo i y o s udies ha e ailed o demons a e whe he
inc eased UCP1 in beige a also esul s in highe he mogene-
sis o me abolism on he sys emic le el (Ba el and Hee en,
Cell Repo s 27, 3385–3400, June 11, 2019 ª2019 The Au ho (s). 3385
This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
(legend on nex page)
3386 Cell Repo s 27, 3385–3400, June 11, 2019
2014; Ha ms and Seale, 2013; Kajimu a e al., 2015; Keipe and
Jas och, 2014; Wa ne and Mi ag, 2016).
Thy oid ho mones a e essen ial in he egula ion o me abolic
and he mo egula o y pa hways in mammals (Mullu e al., 2014;
Sil a, 2006; Yen, 2001). This becomes e iden in hype hy oid
pa ien s, which show impai ed hea ole ance and inc eased
body empe a u e (Sil a, 2003). The e ec s o hy oid ho mone
on he mogenesis ha e been s udied o decades, es ablishing
he pa adigm ha he ho mone inc eases obliga o y he mogen-
esis in se e al issues, including muscle, which leads o a highe
body empe a u e (Sil a, 2006). Thy oid ho mones exe hei ac-
ions by he nuclea hy oid ho mone ecep o s (TRs) alpha1
(TRa1), be a1 (TRb1), o be a 2 (TRb2), al e ing gene exp ession
(Sap e al., 1986; Ta a, 1986; Thompson e al., 1987). TRs a e
exp essed in almos e e y issue and o en ha e o e lapping
unc ions; howe e , he e a e issues ha p edominan ly ely
on TRa1o TRb, espec i ely (Yen, 2001). Wi h ega d o he mo-
genesis, TRbhas been shown o play a ole in he egula ion o
adap i e he mogenesis by egula ing Ucp1 mRNA exp ession,
and TRa1 seems o modula e ad ene gic sensi i i y (Ribei o
e al., 2001; Weiss e al., 1998). Howe e , mice de oid o all
TRs s ill display compe en BAT ec ui men and Ucp1 gene
exp ession bu dep essed he mogenesis, sugges ing a com-
plex in e play be ween hy oid ho mone and body empe a u e
egula ion (Golozoubo a e al., 2004). Mo eo e , i was ecen ly
disco e ed ha TRa1 also con ols ail hea loss wi h seconda y
e ec s o BAT he mogenesis (Wa ne e al., 2013). The e o e,
hy oid ho mones a e ele an o almos all aspec s o he mo-
egula ion, including he mal conduc ance as well as obliga o y
and acul a i e he mogenesis. In addi ion o he pe iphe al
ac ions, ecen s udies e ealed ha hy oid ho mone also in-
duces acul a i e he mogenesis h ough cen al mechanisms,
as cen al hype hy oidism leads o di ec ac i a ion o BAT in
an AMP-ac i a ed p o ein kinase (AMPK)-dependen manne in
a s (Al a ez-C espo e al., 2016; Lo
´pez e al., 2010; Ma ı
´nez-
Sa
´nchez e al., 2017b) and induces b owning in mice (Ma ı
´-
nez-Sa
´nchez e al., 2017a). Mos in e es ingly, a ecen s udy
e ealed ha he TRbagonis GC-1 di ec ly igge s b owning
(Lin e al., 2015), an e ec ha has been epo ed o hy oid ho -
mone as well (Hoe ig e al., 2016; Weine e al., 2016). The da a
sugges ed ha he beige a ec ui men could con ibu e o
he body empe a u e inc ease obse ed in hype hy oidism
and migh p o ide a o able me abolic e ec s. The indings
he e o e challenge he well-es ablished model ha hy oid ho -
mone only modula es basal me abolic a e and obliga o y he -
mogenesis (Sil a, 2006) and indica e ha beige a migh be a
p ominen playe in he he mogenic and me abolic e ec s o
he ho mone. The e o e, we aimed o es he ole o he mogenic
adipose issue in body empe a u e egula ion and o he me a-
bolic bene i s obse ed in sys emic hype hy oidism.
RESULTS
Hype hy oidism Imp o es Glucose Tole ance and
Induces B owning o Inguinal Whi e Adipose Tissue
To es ablish he e ec s o hype hy oidism, wild- ype (WT) mice
we e ea ed o 14 days wi h ei he 0.5 mg/L T3 o 1 mg/L T4 in
d inking wa e (Figu e 1A). T ea men wi h T3 induced a subs an-
ial hype hy oidism wi h an 8.5- old inc ease in se um o al T3
( T3) and a 90% dec ease in se um T4 (Figu es 1B and 1C).
Hype hy oidism was ully achie ed a e 24 h (no shown),
pe sis en ly obse ed o e he cou se o a day (Figu es S1A
and S1B), and leads o a comple e supp ession o pi ui a y hy-
oid-s imula ing ho mone (TSH) (Figu e S1C). I was accompa-
nied by an inc ease in ene gy expendi u e, a lowe espi a o y
quo ien , and an inc ease in body weigh and leng h as expec ed
(Neilsen, 1953; Rako e al., 2016; G een, 1975), as well as
hype phagia (Figu es S1D–S1H). Ou model o T4-induced
hype hy oidism was milde , wi h no signi ican ele a ion in T3
and a 7.4- old inc ease in se um T4 (Figu es 1B and 1C) bu
no signi ican changes in body weigh , leng h, o ood in ake (Fig-
u es S1F–S1H). As hype hy oidism is associa ed wi h imp o ed
glucose ole ance, we pe o med an in ape i oneal glucose
ole ance es (ipGTT) and ound educed blood glucose le els
in T3- ea ed, bu no T4- ea ed, mice (Figu es 1D and 1E; p <
0.01 o T3; p = 0.34 o T4; epea ed measu es [RM]-ANOVA).
Co e body empe a u e in hy oid-ho mone- ea ed mice s a ed
o inc ease a e 1 week o ea men (Figu e 1F) and was signi -
ican ly ele a ed in T3- ea ed, bu no T4- ea ed, mice a e
2 weeks (T3 p < 0.01; T4 p = 0.09; Figu e 1G). Skin empe a u e
abo e he in e scapula b own adipose issue (iBAT) as
measu ed by in a ed he mog aphy did no signi ican ly change
Figu e 1. Hype hy oidism Imp o es Glucose Tole ance and Induces B owning o ingWAT
(A) Scheme o ea men egimen o induc ion o hype hy oidism in mice.
(B) Se um o al T4 le els a e 14 days o ea men .
(C) Se um o al T3 le els a e 14 days o ea men .
(D) ipGTT a e 12 days o ea men wi h THs.
(E) A ea unde cu e (AUC) o ipGTT (D).
(F) Rec al empe a u e o con ol and T3- and T4- ea ed mice o e 14 days o ea men .
(G) Rec al empe a u e a he end o ea men .
(H) iBAT empe a u e as measu ed by in a ed he mog aphy o e 14 days o ea men .
(I) iBAT empe a u e as measu ed by in a ed he mog aphy a he end o ea men no malized o ec al empe a u e.
(J) Tail empe a u e o con ol and T3- and T4- ea ed mice as measu ed by in a ed he mog aphy o e 14 days o ea men .
(K) Hepa ic PK ac i i y in con ol and T3- and T4- ea ed mice.
(L) Hepa ic PEPCK ac i i y in con ol and T3- and T4- ea ed mice.
(M) Hepa ic glycogen con en o con ol and T3- and T4- ea ed mice.
(N) Rep esen a i e H&E s ainings o iBAT and ingWAT o con ol and T3- and T4- ea ed mice. Scale ba 10 mm.
(O and P) Gene exp ession analysis in (O) ingWAT and (P) iBAT in con ol and T3- and T4- ea ed mice.
(Q) Di ec compa ison o ingWAT UCP1 p o ein exp ession o T3- ea ed mice o con ol iBAT.
Da a a e p esen ed as mean ±SEM; n = 6–8. *p < 0.05; **p < 0.01; ***p < 0.001. See also Figu e S1.
Cell Repo s 27, 3385–3400, June 11, 2019 3387
o e he cou se o he 14 days o ea men (Figu es 1H and S1I)
bu was signi ican ly dec eased upon T3 a he end o he ea -
men , in absolu e alues bu also when no malized o body
empe a u e (Figu es 1H and 1I). In pa allel, skin empe a u e
abo e he inguinal whi e adipose issue (ingWAT) was un-
changed (Figu e S1J). In e es ingly, ail empe a u e was signi i-
can ly ele a ed o e he cou se o he T3 ea men indica i e o
asodila ion o dissipa e excess hea s a ing a day 4 o he
ea men (Figu e 1J). To u he in es iga e glucose me abolism,
we measu ed he ac i i ies o a e-limi ing s eps in glycolysis
(py u a e kinase [PK]) and gluconeogenesis (phosphoenolpy -
u a e-ca boxykinase [PEPCK]) in li e . PK ac i i y was signi i-
can ly dec eased in T3- and T4- ea ed mice (Figu e 1K), and
PEPCK ac i i y was inc eased in T3- ea ed mice (Figu e 1L).
Hepa ic glycogen con en was dec eased in T3- ea ed mice
(Figu e 1M), indica ing a highe sys emic need o glucose. O gan
weigh s de e mined a he ime o sac i ice showed he expec ed
ca diac hype ophy in T3- ea ed, bu no T4- ea ed, mice (Fig-
u es S1K and S1L). In bo h T3- and T4- ea ed animals, iBAT de-
po s we e signi ican ly la ge and appea ed pale , sugges ing
highe lipid deposi ion (Figu e S1M), which was con i med by
H&E s aining (Figu e 1N). In e es ingly, when we in es iga ed in-
gWAT mo phology (Figu e 1N), an inc ease in smalle , mul ilocu-
la a cells was obse ed, indica ing induc ion o b owning by
hy oid ho mone, as epo ed p e iously (Al a ez-C espo e al.,
2016; Finan e al., 2016; Weine e al., 2016). To es his on
he molecula le el, we measu ed gene exp ession o he mo-
genic ma ke genes (uncoupling p o ein 1 [Ucp1]; deiodinase 2
[Dio2]; pe oxisome p oli e a o -ac i a ed ecep o gamma coac-
i a o 1 alpha [Ppa gc1a]; mi ochond ial ansc ip ion ac o A
[T am]; PR domain con aining 16 [P dm16]; cell-dea h-inducing
DFFA-like e ec o a [Cidea]; b3-ad ene gic ecep o [Ad b3];
Figu es 1O and 1P) and UCP1 p o ein exp ession (Figu e 1Q).
In ingWAT, we ound an ele a ion o se e al he mogenic
ma ke s, including Ucp1,Dio2,Ppa gc1a, and Cidea in T3-
and T4- ea ed animals, and Ad b3 mRNA was educed upon
T3 ea men (Figu e 1O). In con as , iBAT gene exp ession anal-
ysis showed a dec ease in mos he mogenic and b own a
ma ke genes (Figu e 1P). In e es ingly, despi e educed mRNA
le els, UCP1 p o ein was ele a ed in T3- ea ed iBAT wi h no
change in he p o ein le el o he mi ochond ial espi a o y chain
complexes (Figu es S1N–S1P). In T3- ea ed ingWAT, UCP1
p o ein eached le els compa able o abou one- hi d o he
amoun ound in un ea ed iBAT (Figu e 1Q), sugges ing a
possibly high he mogenic po en ial. B owning was also induced
by T3 and T4 in gonadal WAT (gWAT) on he mRNA le el (Fig-
u e S1Q); howe e , no signi ican amoun o UCP1 p o ein was
de ec able in his issue (Figu e S1R). Gi en ha he T3 ea men
en i ely supp essed TSH, hus po en ially blocking he oscilla ing
hy hm o hy oid ho mone sec e ion, we es ed whe he he lack
o a ci cadian hy hm would by i sel cause b owning. Using
clock-de icien Pe 1/Pe 2 double mu an (double knockou
[DKO]) mice, we con i med ha his was no he case, as Ucp1
mRNA exp ession was no induced in ingWAT (Figu e S1S).
Taken oge he , hese da a show ha T3 and T4 a e capable
o b owning ingWAT on he molecula le el; howe e , he sys-
emic e ec s o hype hy oidism a e mo e p onounced in he
T3- ea ed g oup.
Ucp1 Exp ession in ingWAT Is Dependen on TRb
As i was epo ed ha b owning o ingWAT can be induced by
TRb-selec i e compounds (Lin e al., 2015), we in es iga ed he
me abolic pheno ype o global TRbknockou (KO) mice (Fig-
u e 2A). These mice showed ele a ed se um T3 and T4 le els
(Figu es S2A and S2B) due o he impai ed eedback o he hypo-
halamus-pi ui a y- hy oid axis (Fo es e al., 1996). Al hough
ipGTT was no al e ed in TRbKO mice compa ed o hei WT
li e ma es (Figu es 2B and 2C; p = 0.41 o TRb; RM-ANOVA),
hei basal me abolic a e (BMR) (de ined he e as me abolism a
he moneu ali y) was signi ican ly inc eased (Figu es 2Dand
2E). This did, howe e , no ansla e o an inc ease in daily ene gy
expendi u e a oom empe a u e (Figu es S2C and S2D), sug-
ges ing a compensa o y adap a ion o non-shi e ing he mogen-
esis. Despi e being hype hy oid, TRbKO mice did no ha e
ele a ed body o iBAT empe a u e (Figu es 2F–2H), and he his-
ology o iBAT and ingWAT was simila o con ols (Figu e 2I).
When in es iga ing gene exp ession in ingWAT, we ound ha
Dio2 and Ad b3 exp ession we e no al e ed in hype hy oid
TRbKO mice, bu Ucp1 mRNA exp ession was e en lowe han
in con ols (Figu e 2J). In iBAT, Ucp1 exp ession was no al e ed
on mRNA (Figu e 2K) o p o ein le el (Figu es S2E and S2F). These
da a demons a e ha b owning o ingWAT by hy oid ho mone
on he molecula le el depends on in ac TRbsignaling.
Hype hy oidism Dec eases Adipose Tissue Glucose
and Lipid Up ake
Ac i a ion o he mogenesis in adipose issues is associa ed wi h
inc eased glucose up ake (Cannon and Nede gaa d, 2004). To
es his di ec ly, we pe o med [18] luo o-2-deoxyglucose–
posi on emission omog aphy (
18
FDG-PET)/compu ed omog a-
phy (CT) scans o T3- and T4- ea ed mice. Despi e ele a ed
UCP1 le els, ingWAT glucose up ake was no inc eased bu
dec eased upon sys emic T3 o T4 ea men (Figu es 3Aand
3B). Likewise, glucose up ake in iBAT was educed (Figu es 3C
and 3D), and no change was obse ed in gWAT o soleus muscle
(Figu es S2G and S2H).
18
FDG-PET/CT may, howe e , no accu-
a ely e lec b own o beige a he mogenic ac i i y (Hanki e al.,
2017; Olsen e al., 2017), as he mogenesis is mainly ueled
by a y acids (Ba el e al., 2011; Cannon and Nede gaa d,
2004). We he e o e also measu ed up ake o iglyce ide- ich
lipop o eins (TRLs) in hype hy oid mice. In iBAT, TRL up ake
signi ican ly dec eased o abou one- hi d in sys emically T3-
and T4- ea ed mice, and no di e ence was obse ed in ingWAT
(Figu e 3E). Gi en he lack o me abolic ac i a ion in he p esence
o ele a ed UCP1 p o ein le els, we hypo hesized ha he he mo-
genic capaci y o he issue migh no be ully exploi ed due
o insu icien ad ene gic s imula ion. The e o e, we injec ed
mice wi h 1 mg/kg NE o s imula e maximum non-shi e ing
he mogenesis. Con ol and T3- ea ed animals showed an
ele a ion in oxygen consump ion a e NE injec ion (Figu e 3F);
howe e , he esponse was mo e p onounced in T3- ea ed
mice, indica ing a highe he mogenic capaci y. Ou hypo hesis
o insu icien ad ene gic s imula ion was u he suppo ed on
he molecula le el, as we ound no signi ican ele a ion in ingWAT
and iBAT o in acellula cAMP (Figu es 3G and 3H), phospho-
ho mone sensi i e lipase (HSL; Figu es 3I and 3J), o ee a y
acids (Figu es 3K and 3L), which a e equi ed o ac i a ion o
3388 Cell Repo s 27, 3385–3400, June 11, 2019

UCP1 (Shabalina e al., 2010). In e es ingly, exp ession o genes
in ol ed in lipogenesis and lipolysis (pe oxisome p oli e a o -
ac i a ed ecep o gamma [Ppa g]; a y acid anslocase [Cd36];
ca bohyd a e- esponsi e elemen -binding p o ein be a [Ch ebpb];
lipop o ein lipase [Lpl]; lep in [Lep]; lipase A [Lipa]; s ea oyl-coen-
zyme A [CoA]desa u ase 1 [Scd1]; a y acid syn hase [Fasn])
was mos ly unal e ed in ingWAT (Figu e S2I) while being educed
in iBAT o hype hy oid mice (Figu e S2J). Collec i ely, hese
da a sugges ha , despi e a molecula b owning inge p in
including ele a ed UCP1 p o ein, he me abolic u no e is no
ele a ed in he mogenic adipose issues o hype hy oid animals.
Thy oid-Ho mone-Induced Hype he mia Is
Independen o SNS Signaling in Adipose Tissues
T3 cen ally ac i a es he mogenesis ia he SNS (Al a ez-
C espo e al., 2016; Lo
´pez e al., 2010; Ma ı
´nez-Sa
´nchez
e al., 2017b). In e es ingly, we obse ed a down egula ion o
Ad b3 exp ession in iBAT and ingWAT in sys emic hype hy oid-
ism, sugges ing educed ad ene gic esponsi eness o hese
issues. We hen measu ed NE u no e in hype hy oid mice
(Figu es S3A and S3B). The da a e ealed ha baseline NE le els
(0 h) and NE u no e we e no signi ican ly di e en be ween
con ols and T3- o T4- ea ed mice a oom empe a u e in in-
gWAT bu lowe in iBAT (Figu es 4A, 4B, S3C, and S3D). This
was accompanied by no mal le els o y osine hyd oxylase
p o ein in bo h issues (Figu es S3E–S3G) and sugges s ha
he ele a ion o body empe a u e in sys emically hype hy oid
animals migh no be caused by a cen al ac i a ion o he
SNS. To elucida e his in g ea e de ail, we housed mice a he -
moneu ali y (TN) (Figu e 4C), whe e he mogenic adipose issue
is only minimally inne a ed due o he lack o acul a i e he mo-
genesis (Cannon and Nede gaa d, 2004; Sjo
¨g en e al., 2007).
Figu e 2. Ucp1 Exp ession in ingWAT Is Dependen on TRbExp ession
(A) Adul TRbKO mice and hei WT li e ma es we e me abolically cha ac e ized.
(B) ipGTT o WT and TRbKO mice.
(C) A ea unde cu e o ipGTT (B).
(D) BMR o WT and TRbKO mice no malized o body weigh .
(E) BMR o WT and TRbKO mice plo ed agains body weigh .
(F) Rec al empe a u e o WT and TRbKO mice.
(G) iBAT empe a u e as measu ed by in a ed he mog aphy.
(H) Rep esen a i e in a ed pic u es o iBAT o WT and TRbKO mice.
(I) Rep esen a i e H&E s ainings o ingWAT and iBAT o WT and TRbKO mice. Scale ba 10 mm.
(J and K) Gene exp ession analysis o (J) ingWAT and (K) iBAT o WT and TRbKO mice.
Da a a e p esen ed as mean ±SEM; n = 5. *p < 0.05; **p < 0.01; ***p < 0.001. See also Figu e S2.
Cell Repo s 27, 3385–3400, June 11, 2019 3389
Figu e 3. Hype hy oidism Dec eases Adipose Tissue Glucose and Lipid Up ake
(A) Ra e o glucose me abolism in ingWAT o con ol and hype hy oid mice a e 14 days o ea men as measu ed by
18
FDG-PET/CT. n = 7–14.
(B) Rep esen a i e
18
FDG-PET/CT pic u es o ingWAT o con ol and hype hy oid mice. Whi e a ows indica e ingWAT depo analyzed.
(C) Ra e o glucose me abolism in iBAT o con ol and hype hy oid mice a e 14 days o ea men as measu ed by
18
FDG-PET/CT. n = 7–14.
(D) Rep esen a i e
18
FDG-PET/CT pic u es o iBAT o con ol and hype hy oid mice. Whi e a ows indica e iBAT depo analyzed.
(E) Up ake o adioac i ely labeled TRLs in o ingWAT and iBAT o con ol and T3- and T4- ea ed mice a e 14 days o ea men . n = 7–8.
(legend con inued on nex page)
3390 Cell Repo s 27, 3385–3400, June 11, 2019
In e es ingly, TN-housed mice s ill displayed a signi ican ly
ele a ed body empe a u e al eady a e 1 day o ea men
wi h T3 (Figu es 4D, 4E, S3H, and S3I). As expec ed, NE le els
in ingWAT (Figu es 4F and S3J) as well as iBAT (Figu es 4G
and S3K) we e s ongly educed a TN as compa ed o oom
empe a u e. Only when no malized o issue weigh , bu no
p o ein con en , T3 caused a signi ican ly u he decline o NE
in iBAT a bo h empe a u es (Figu es 4G and S3K). In bo h is-
sues, no signi ican change was obse ed in in acellula cAMP
(Figu es 4H and 4I) and ee a y acids (Figu es 4J and 4K). In e -
es ingly, ingWAT o T3- ea ed mice a TN s ill showed his olog-
ical signs o b owning (Figu e 4L), oge he wi h inc eased
exp ession o Ucp1, albei a much lowe le els (Figu e 4M;
Table S1). As expec ed, we ound ‘‘whi ened iBAT’’ in con ol
and T3- ea ed mice a TN cha ac e ized by bigge , unilocula
cells (Figu e 4L). The lowe Ucp1 gene exp ession in iBAT by sys-
emic T3 ea men was e e sed a TN (Figu e 4N; Table S1),
which oge he wi h ele a ed UCP1 p o ein (Figu e 4O) sugges s
a ‘‘ e-b owning’’ o he whi ened iBAT by T3. Toge he , ou da a
implica e ha he inc ease in body empe a u e and he b own-
ing o ingWAT in sys emic hype hy oidism occu independen ly
o SNS ac i a ion. Mo eo e , he beige a seems o lack he mo-
genic ac i i y due o insu icien sympa he ic s imula ion.
In e es ingly, his is opposi e o wha is obse ed in cen al hy-
pe hy oidism, which is cha ac e ized by ma ked SNS ac i a ion
(Lo
´pez e al., 2010; Ma ı
´nez-Sa
´nchez e al., 2017b). We he e-
o e es ed whe he he AMPK pa hway in he en omedial nu-
cleus o he hypo halamus (VMH) is igge ed by he sys emic
T3 ea men . Ou da a e ealed ha he e was no e ec o he
sys emic T3 ea men a 23Co 30
C on AMPKaphospho yla-
ion in he VMH (Figu es 4P–4S), despi e T3 eaching he b ain as
e idenced by he T3-induced hypo halamic genes Dio3,H , and
Kl 9 (Figu e S3L), as well as signi ican ly inc eased hypo halamic
T3 con en (Figu e S3M).
Thy oid-Ho mone-Media ed Inc ease in Body
Tempe a u e Pe sis s in he Cold
To es whe he alle ia ing he hea s ess o he T3 ea men
no malizes body empe a u e egula ion, we epea ed he ea -
men a 10C(Figu es 5A, S3N, and S3O). In e es ingly, he
ele a ion o body empe a u e by T3 pe sis ed in hese animals
(Figu e 5B) wi h no mal iBAT (Figu es 5C and 5D) and ail empe -
a u e (Figu e 5E), indica ing he absence o hype he mia and a
cen ally ele a ed body empe a u e se poin . In e es ingly, T3
s ill induced b owning o ingWAT a he his ological le el, and
iBAT con ained mo e a d ople s, compa able o he si ua ion
a highe empe a u es (Figu e 5F). Howe e , on he molecula
le el (Figu es 5G and S3P), Ucp1 mRNA was no signi ican ly
highe in he T3- ea ed g oup, mos likely because he ingWAT
o con ol animals was b owned by he lowe empe a u e.
Only Dio2 mRNA le els we e s ill ele a ed by he T3 ea men
(Figu e 5G). In con as , iBAT showed lowe exp ession o he -
mogenic genes and Ad b3 in he T3 g oup (Figu es 5H and
S3Q), wi h no mal UCP1 p o ein (Figu e 5I). The NE s imula ion
es showed a s onge esponse in bo h g oups and a smalle
di e ence be ween T3- ea ed and un ea ed animals (Figu e 5J),
sugges ing a lowe addi ional e ec o T3 on he mogenic capac-
i y a 10C as compa ed o 23C. We obse ed compa able
le els o NE and in acellula cAMP in ingWAT and iBAT o
bo h g oups (Figu es 5K–5N, S3R, and S3S). In e es ingly, ee
a y acids we e signi ican ly ele a ed in ingWAT o T3- ea ed
mice, and no di e ence was ound in iBAT (Figu es 5O and
5P). When we analyzed he VMH, we obse ed ha now, in he
absence o hea s ess as e idenced by he no mal ail empe -
a u e, AMPKaphospho yla ion was educed by he T3 ea men
(Figu es 5Q and 5R).
Thy oid-Ho mone-Media ed Inc ease in Body
Tempe a u e Is Independen o UCP1
As nei he beige no b own a appea ed o be ac i e in hype hy-
oid mice a oom empe a u e, we hypo hesized ha he
obse ed hype he mia migh no depend on UCP1-media ed
he mogenesis. To es his di ec ly, we ea ed UCP1 KO mice
and hei WT li e ma es a oom empe a u e wi h T3 (Figu e 6A),
esul ing in bo h geno ypes in a dec ease in se um T4 and a
s ong inc ease in se um T3 (Figu es S4A and S4B) and educed
blood glucose le els in he ipGTT (Figu es 6B and 6C; ipGTT: p
ge-
no ype
= 0.466, p
T3
< 0.0006, and p
in e ac ion
= 0.892). The ela i e
amoun o lean mass was inc eased in T3- ea ed WT and UCP1
KO mice, and he ela i e amoun o a mass was no signi i-
can ly al e ed (Figu es S4C and S4D; lean mass: p
geno ype
=
0.721, p
T3
< 0.0008, and p
in e ac ion
= 0.065; a mass: p
geno ype
=
0.458, p
T3
= 0.565, and p
in e ac ion
= 0.069). Body empe a u e
was signi ican ly ele a ed in bo h g oups a e ea men wi h
T3 (Figu e 6D; body empe a u e: p
geno ype
= 0.794, p
T3
<
0.0001, and p
in e ac ion
= 0.213). In e es ingly, we ound mul iloc-
ula cells in ingWAT o UCP1 KO mice a e T3 ea men as a his-
ological sign o b owning, and he mul ilocula cells in he iBAT
o hese mice appea ed bigge (Figu e 6E). Mo eo e , we ound
an inc eased exp ession o Cidea a e T3 ea men in UCP1 KO
mice, and mos o he ma ke s o b owning we e no al e ed in
ingWAT (Figu e 6F; Table S2). iBAT gene exp ession o WT
and UCP1 KO mice showed a dec ease in Ad b3,Dio2, and
Ppa gc1ain sys emic hype hy oidism (Figu e 6G; Table S2). I
was epo ed ha ingWAT ea u es o he he mogenic mecha-
nisms, like calcium o c ea ine cycling (Ikeda e al., 2017; Kazak
e al., 2015, 2017); howe e , we only ound a mino inc ease in
Se ca2 (A p2a2) gene exp ession in ingWAT o T3- ea ed WT
(F) Change in oxygen consump ion o con ol and T3- ea ed mice housed a 23C a e NE injec ion. n = 6.
(G) cAMP le els in ingWAT o mice housed and ea ed a 23C. n = 4.
(H) cAMP le els in iBAT o mice housed and ea ed a 23C. n = 4.
(I) P o ein exp ession o HSL and phospho HSL (pHSL) in ingWAT o mice housed and ea ed a 23C. n = 4.
(J) P o ein exp ession o HSL and pHSL in iBAT o mice housed and ea ed a 23C. n = 4.
(K) F ee a y acids in ingWAT o mice housed and ea ed a 23C. n = 4.
(L) F ee a y acids in iBAT o mice housed and ea ed a 23C. n = 4.
Da a a e p esen ed as mean ±SEM. *p < 0.05; **p < 0.01; ***p < 0.001. See also Figu e S2.
Cell Repo s 27, 3385–3400, June 11, 2019 3391
Figu e 4. Thy oid-Ho mone-Induced Hype he mia Is Independen o SNS Signaling
(A and B) NE u no e in (A) ingWAT and (B) iBAT o con ol and T3- and T4- ea ed mice a e 14 days o ea men no malized agains issue weigh . n = 4 pe ime
poin and g oup.
(legend con inued on nex page)
3392 Cell Repo s 27, 3385–3400, June 11, 2019
Li le, A.G., and Seebache , F. (2014). The e olu ion o endo he my is ex-
plained by hy oid ho mone-media ed esponses o cold in ea ly e eb a es.
J. Exp. Biol. 217, 1642–1648.
Lo
´pez, M., Va ela, L., Va
´zquez, M.J., Rod ı
´guez-Cuenca, S., Gonza
´lez, C.R.,
Velagapudi, V.R., Mo gan, D.A., Schoenmake s, E., Agassandian, K., Lage,
R., e al. (2010). Hypo halamic AMPK and a y acid me abolism media e
hy oid egula ion o ene gy balance. Na . Med. 16, 1001–1008.
Lo
´pez, M., Al a ez, C.V., Noguei as, R., and Die
´guez, C. (2013). Ene gy bal-
ance egula ion by hy oid ho mones a cen al le el. T ends Mol. Med. 19,
418–427.
Ma ı
´nez-Sa
´nchez, N., Mo eno-Na a e e, J.M., Con e as, C., Rial-Pensado,
E., Fe nø, J., Noguei as, R., Die
´guez, C., Fe na
´ndez-Real, J.M., and Lo
´pez, M.
(2017a). Thy oid ho mones induce b owning o whi e a . J. Endoc inol. 232,
351–362.
Ma ı
´nez-Sa
´nchez, N., Seoane-Collazo, P., Con e as, C., Va ela, L., Villa oya,
J., Rial-Pensado, E., Buque
´, X., Au ekoe xea, I., Delgado, T.C., Va
´zquez-Ma -
ı
´nez, R., e al. (2017b). Hypo halamic AMPK-ER s ess-JNK1 axis media es
he cen al ac ions o hy oid ho mones on ene gy balance. Cell Me ab. 26,
212–229.e12.
Mullu , R., Liu, Y.Y., and B en , G.A. (2014). Thy oid ho mone egula ion o
me abolism. Physiol. Re . 94, 355–382.
Nede gaa d, J., and Cannon, B. (2013). UCP1 mRNA does no p oduce hea .
Biochim. Biophys. Ac a 1831, 943–949.
Neilsen, E.L. (1953). S udies on he edi a y dwa ism in mice. XIV. E ec o
hy oxin and g ow h ho mone on g ow h. Ac a Pa hol. Mic obiol. Scand. 32,
316–334.
Olsen, J.M., Csikasz, R.I., Deh a i, N., Lu, L., Sands o
¨m, A., O
¨be g, A.I., Ne-
de gaa d, J., S one-Elande , S., and Beng sson, T. (2017). b
3
-Ad ene gically
induced glucose up ake in b own adipose issue is independen o UCP1 p es-
ence o ac i i y: Media ion h ough he mTOR pa hway. Mol. Me ab. 6,
611–619.
Os e , H., Dame ow, S., Hu , R.A., and Eichele, G. (2006). T ansc ip ional
p o iling in he ad enal gland e eals ci cadian egula ion o ho mone biosyn-
hesis genes and nucleosome assembly genes. J. Biol. Rhy hms 21, 350–361.
Pe o ic, N., Walden, T.B., Shabalina, I.G., Timmons, J.A., Cannon, B., and
Nede gaa d, J. (2010). Ch onic pe oxisome p oli e a o -ac i a ed ecep o
gamma (PPARgamma) ac i a ion o epididymally de i ed whi e adipocy e cul-
u es e eals a popula ion o he mogenically compe en , UCP1-con aining
adipocy es molecula ly dis inc om classic b own adipocy es. J. Biol.
Chem. 285, 7153–7164.
Puigse e , P., He on, D., Giano i, M., Palou, A., Cannon, B., and Nede -
gaa d, J. (1992). Induc ion and deg ada ion o he uncoupling p o ein
he mogenin in b own adipocy es in i o and in i o. E idence o a apidly
deg adable pool. Biochem. J. 284, 393–398.
Rabelo, R., Schi man, A., Rubio, A., Sheng, X., and Sil a, J.E. (1995). Delinea-
ion o hy oid ho mone- esponsi e sequences wi hin a c i ical enhance in he
a uncoupling p o ein gene. Endoc inology 136, 1003–1013.
Rabelo, R., Reyes, C., Schi man, A., and Sil a, J.E. (1996). In e ac ions among
ecep o s, hy oid ho mone esponse elemen s, and ligands in he egula ion
o he a uncoupling p o ein gene exp ession by hy oid ho mone. Endoc i-
nology 137, 3478–3487.
Rako , H., Engels, K., Ho
¨nes, G.S., S ucksbe g, K.H., Moelle , L.C., Ko
¨h le, J.,
Zwanzige , D., and F€
uh e , D. (2016). Sex-speci ic pheno ypes o hype hy-
oidism and hypo hy oidism in mice. Biol. Sex Di e . 7,36.
Ribei o, M.O., Ca alho, S.D., Schul z, J.J., Chiellini, G., Scanlan, T.S., Bianco,
A.C., and B en , G.A. (2001). Thy oid ho mone–sympa he ic in e ac ion and
adap i e he mogenesis a e hy oid ho mone ecep o iso o m–speci ic.
J. Clin. In es . 108, 97–105.
G een, E.L. (1975). Biology o he Labo a o y Mouse (Do e Publica ions).
Rubio, A., Raasmaja, A., and Sil a, J.E. (1995). Thy oid ho mone and no epi-
neph ine signaling in b own adipose issue. II: Di e en ial e ec s o hy oid
ho mone on be a 3-ad ene gic ecep o s in b own and whi e adipose issue.
Endoc inology 136, 3277–3284.
Sal a o e, D., Simonides, W.S., Den ice, M., Za acki, A.M., and La sen, P.R.
(2014). Thy oid ho mones and skele al muscle–new insigh s and po en ial
implica ions. Na . Re . Endoc inol. 10, 206–214.
San hanam, P., Ahima, R.S., Mammen, J.S., Gio anella, L., and T eglia,
G. (2018). B own adipose issue (BAT) de ec ion by 18F-FDG PET
and hy oid ho mone le el(s)-a sys ema ic e iew. Endoc ine 62,
496–500.
Sap, J., Mun
˜oz, A., Damm, K., Goldbe g, Y., Ghysdael, J., Leu z, A., Beug, H.,
and Venns o
¨m, B. (1986). The c-e b-A p o ein is a high-a ini y ecep o o
hy oid ho mone. Na u e 324, 635–640.
Shabalina, I.G., Os , M., Pe o ic, N., V backy, M., Nede gaa d, J., and Can-
non, B. (2010). Uncoupling p o ein-1 is no leaky. Biochim. Biophys. Ac a
1797, 773–784.
Shabalina, I.G., Pe o ic, N., de Jong, J.M., Kalino ich, A.V., Cannon, B., and
Nede gaa d, J. (2013). UCP1 in b i e/beige adipose issue mi ochond ia is
unc ionally he mogenic. Cell Rep. 5, 1196–1203.
Shiba a, H., Pe
´ usse, F., Valle and, A., and Bukowiecki, L.J. (1989). Cold expo-
su e e e ses inhibi o y e ec s o as ing on pe iphe al glucose up ake in a s.
Am. J. Physiol. 257, R96–R101.
Sil a, J.E. (2003). The he mogenic e ec o hy oid ho mone and i s clinical
implica ions. Ann. In e n. Med. 139, 205–213.
Sil a, J.E. (2006). The mogenic mechanisms and hei ho monal egula ion.
Physiol. Re . 86, 435–464.
Sil a, J.E., and La sen, P.R. (1983). Ad ene gic ac i a ion o iiodo hy onine
p oduc ion in b own adipose issue. Na u e 305, 712–713.
Sjo
¨g en, M., Alkemade, A., Mi ag, J., No ds o
¨m, K., Ka z, A., Rozell, B., Wes -
e blad, H., A ne , A., and Venns o
¨m, B. (2007). Hype me abolism in mice
caused by he cen al ac ion o an unliganded hy oid ho mone ecep o
alpha1. EMBO J. 26, 4535–4545.
S ensson, K.J., Long, J.Z., Jed ychowski, M.P., Cohen, P., Lo, J.C., Se ag, S.,
Ki , S., Shinoda, K., Ta aglia, J.A., Rao, R.R., e al. (2016). A sec e ed Sli 2
agmen egula es adipose issue he mogenesis and me abolic unc ion.
Cell Me ab. 23, 454–466.
Tamucci, K.A., Namwanje, M., Fan, L., and Qiang, L. (2018). The da k side o
b owning. P o ein Cell 9, 152–163.
Ta a, J.R. (1986). The sea ch o he mechanism o ho mone ac ion. Pe spec .
Biol. Med. 29, S184–S204.
Thompson, C.C., Weinbe ge , C., Lebo, R., and E ans, R.M. (1987). Iden i ica-
ion o a no el hy oid ho mone ecep o exp essed in he mammalian cen al
ne ous sys em. Science 237, 1610–1614.
Va gas-Cas illo, A., Fuen es-Rome o, R., Rod iguez-Lopez, L.A., To es, N.,
and To a , A.R. (2017). Unde s anding he biology o he mogenic a : is
b owning a new app oach o he ea men o obesi y? A ch. Med. Res. 48,
401–413.
Vujo ic, M., No ds o
¨m, K., Gau hie , K., Flaman , F., Visse , T.J., Ven-
ns o
¨m, B., and Mi ag, J. (2009). In e e ence o a mu an hy oid ho -
mone ecep o alpha1 wi h hepa ic glucose me abolism. Endoc inology
150, 2940–2947.
Wa ne , A., and Mi ag, J. (2012). Thy oid ho mone and he cen al con ol o
homeos asis. J. Mol. Endoc inol. 49, R29–R35.
Wa ne , A., and Mi ag, J. (2016). B eaking BAT: can b owning c ea e a be e
whi e? J. Endoc inol. 228, R19–R29.
Wa ne , A., Rahman, A., Solsjo
¨, P., Go schling, K., Da is, B., Venns o
¨m, B.,
A ne , A., and Mi ag, J. (2013). Inapp op ia e hea dissipa ion igni es b own
a he mogenesis in mice wi h a mu an hy oid ho mone ecep o a1. P oc.
Na l. Acad. Sci. USA 110, 16241–16246.
Weine , J., K anz, M., Klo
¨ ing, N., Kuna h, A., S einho , K., Rijn jes, E., Ko
¨h le,
J., Zeisig, V., Hanki , M., Gebha d , C., e al. (2016). Thy oid ho mone s a us
de ines b own adipose issue ac i i y and b owning o whi e adipose issues
in mice. Sci. Rep. 6, 38124.
Weiss, R.E., Mu a a, Y., Cua, K., Hayashi, Y., Seo, H., and Re e o , S.
(1998). Thy oid ho mone ac ion on li e , hea , and ene gy expendi u e
Cell Repo s 27, 3385–3400, June 11, 2019 3399

in hy oid ho mone ecep o be a-de icien mice. Endoc inology 139,
4945–4952.
Yen, P.M. (2001). Physiological and molecula basis o hy oid ho mone ac ion.
Physiol. Re . 81, 1097–1142.
Young, P., A ch, J.R., and Ashwell, M. (1984). B own adipose issue in he
pa ame ial a pad o he mouse. FEBS Le . 167, 10–14.
Zhang, Z., Bisschop, P.H., Foppen, E., an Bee en, H.C., Kalsbeek, A., Boelen,
A., and Flie s, E. (2016). A model o ch onic, in ahypo halamic hy oid ho -
mone adminis a ion in a s. J. Endoc inol. 229, 37–45.
Zhao, J., Unelius, L., Beng sson, T., Cannon, B., and Nede gaa d, J. (1994).
Coexis ing be a-ad enocep o sub ypes: signi icance o he mogenic
p ocess in b own a cells. Am. J. Physiol. 267, C969–C979.
3400 Cell Repo s 27, 3385–3400, June 11, 2019
STAR+METHODS
KEY RESOURCES TABLE
CONTACT FOR REAGENT AND RESOURCE SHARING
Fu he in o ma ion and eques s o esou ces and eagen s should be di ec ed o and will be ul illed by he Lead Con ac
P o . D . Jens Mi ag ([email p o ec ed]).
REAGENT o RESOURCE SOURCE IDENTIFIER
An ibodies
an i-UCP1 Jas och e al., 2012 N/A
an i-Se ca2 Cell Signaling Technology Ca #4388; RRID:AB_2227684
an i-HSL Cell Signaling Technology Ca #4107; RRID:AB_2296900
an i-pHSL(Se 660) Cell Signaling Technology Ca #4126S; RRID:AB_490997
an i-OxPhos cock ail In i ogen Ca #45-8099; RRID:AB_2533835
an i-Ty osine Hyd oxylase Sigma Ca #C2928
an i-b-ac in Sigma Ca #A1978; RRID:AB_476692
an i-HSP90 Cell Signaling Technology Ca #4877; RRID:AB_2233307
an i-pAMPKa(Th 172) Cell Signaling Technology Ca #2535; RRID:AB_331250
an i-b-ac in(AC-74) Sigma Ca #A5316; RRID:AB_476743
an i- abbi polyclonal HRP-conjuga ed an ibody DAKO, Denma k Ca #P0448; RRID:AB_2617138
an i-mouse polyclonal HRP-conjuga ed an ibody DAKO, Denma k Ca #P0447; RRID:AB_2617137
Chemicals, Pep ides, and Recombinan P o eins
L-Thy oxine (T4) Sigma Ald ich, Ge many Ca #T2376
3,30,5-T iiodo-L- hy onine (T3) Sigma Ald ich, Ge many Ca #T6397
A e enol (No epineph ine) Sano i Ca #03870227
a-me hyl-DL- y osine me hyl es e hyd ochlo ide (AMPT) Sigma Ald ich, Ge many Ca #M3281
C i ical Comme cial Assays
Pie ce BCA Ki The mo Scien i ic, Ge many Ca #23227
TGX S ain F ee Fas Cas Ac ylamide Ki Bio-Rad Labo a o ies, Ge many Ca #1610183, Ca #1610185
Ad ans a Wes e nB igh Quan um Ad ans a, USA Ca #K-12041-D20
Se um o al T4 DRG Diagnos ics, Ge many Ca #EIA-1781
Se um o al T3 No aTec Immundiagnos ica GmbH, Ge many Ca #DNOV053
Ame sham cAMP Bio ak EIA GE Heal hca e Li e Sciences Ca #RPN225
F ee Fa y Acid Quan i ica ion Ki abcam plc, UK Ca #ab65341
Expe imen al Models: O ganisms/S ains
C57BL/6NC Cha les Ri e Labo a o ies, Ge many RRID:MGI:2160593
B6.129S1-Th b m1D /J (TRbko) Fo es e al., 1996 RRID:IMSR_JAX:003462
B6.129-UCP1 mKZ/J (UCP1 ko) Ene ba
¨ck e al., 1997 N/A
Pe 1/2 dko Husse e al., 2012 N/A
Oligonucleo ides
P ime sequences see Table S3 N/A
So wa e and Algo i hms
Phenomas e so wa e TSE Sys ems, Ge many N/A
Minispec Plus So wa e 6.0 B uke Co p., Bille ica, MA, USA N/A
Mic oso O ice Excel 2013 Mic oso N/A
P ism 7 G aphPad N/A
Ci cWa e .1.4 Os e e al., 2006 N/A
Cell Repo s 27, 3385–3400.e1–e3, June 11, 2019 e1
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Unless s a ed o he wise, animals we e housed in g oups a 23 ±1C a cons an 12 hou ligh /da k cycle wi h ad libi um access o ood
and wa e . Expe imen s we e conduc ed in male mice a he age o h ee o six mon hs. Wild- ype C57/BL6NC (RRID:MGI:2160593)
we e pu chased om Cha les Ri e Labo a o ies (Cha les Ri e , Ge many), TRbko mice (RRID:IMSR_JAX:003462), UCP1 ko mice and
Pe 1/Pe 2 dko mice we e gene a ed as desc ibed be o e (Ene ba
¨ck e al., 1997; Fo es e al., 1996; Husse e al., 2012) and knockou
mice we e always compa ed o WT li e ma e con ols. All animal p ocedu es we e app o ed by he MELUR Schleswig-Hols ein,
LANUV No d hein-Wes alen and he BGV Hambu g, Ge many.
METHOD DETAILS
S udy design
Hype hy oidism was induced by ea men wi h THs in d inking wa e as ecommended by he Ame ican Thy oid Associa ion guide-
lines wi h a dose low enough o a oid cachexia (Bianco e al., 2014), i.e., 1 mg/L L-Thy oxine (T2376, Sigma Ald ich, Ge many) in
0.01% BSA (A7906, Sigma Ald ich, Ge many) o 0.5 mg/L 3,30,5-T iiodo-L- hy onine (T6397, Sigma Ald ich, Ge many) in 0.01%
BSA o 14 days (Figu e 1A). Con ol mice ecei ed 0.01% BSA. Basic me abolic p o iling was pe o med by measu ing body weigh ,
ood and wa e in ake, measu emen o co e body empe a u e using a ec al p obe (BAT-12, Physi emp, USA), in a ed he mog-
aphy (T335, FLIR, Sweden) (Wa ne e al., 2013), and glucose ole ance es s (ipGTT, 2 g/kg body weigh ). Body composi ion
was measu ed using Minispec LF110 and Minispec Plus So wa e 6.0 (B uke Co p., Bille ica, MA, USA). O gan collec ion was always
pe o med in he middle o he ligh phase, unless s a ed o he wise.
PET-CT
[
18
F]-FDG PET imaging and kine ic modeling was pe o med as desc ibed be o e (Jais e al., 2016). Fo he analysis o he pa ame ic
images o he me abolic a e o glucose (MRglc) olumes o in e es s (VOIs) con aining iBAT o ingWAT, espec i ely we e de ined o
each indi idual animal and each measu emen .
TRL Up ake
Up ake o iglyce ide- ich lipop o eins and [
3
H]2-Deoxyglucose (
3
HDG) in hype hy oid C57/BL6NC mice was in es iga ed as
desc ibed be o e (Ba el e al., 2011). B ie ly, mice we e as ed o 4 h be o e ecei ing an in a enous injec ion o 100 mL adiola-
belled ecombinan TRLs (80 mg iglyce ides/kg) labeled wi h
14
C- iolein (0.6 MBq/kg) and
3
HDG. O gans om anes he ized mice
we e ha es ed a e sys emic pe usion wi h PBS-hepa in (10 U/ml) ia he le hea en icle. Tissues we e homogenized using
Sol able (Pe kin Elme ) and adioac i i y was de e mined using liquid scin illa ion coun ing.
Indi ec Calo ime y and nonshi e ing he mogenic capaci y
Measu emen s o daily ene gy expendi u e and basal me abolic a e in WT and TRbko li e ma es was pe o med in single-housed
animals using Phenomas e (TSE Sys ems, Ge many). Oxygen consump ion (VO
2
), ca bon dioxide p oduc ion (VCO
2
), espi a o y
exchange a e (RQ), and ene gy expendi u e (EE) we e calcula ed wi h Mic oso O ice Excel and Phenomas e so wa e (TSE
Sys ems, Ge many). Basal me abolic a e was measu ed o 1 hou in animals as ed o 6 hou s a 30C.
Ene gy expendi u e, espi a o y exchange a e and nonshi e ing he mogenic capaci y in as ed con ol and T3- ea ed mice we e
pe o med using he CaloBox sys em (PhenoSys GmbH, Ge many). Fo measu emen o daily ene gy expendi u e a 23C and es ing
me abolic a e a 30C, mice we e acclima ed o he chambe s indi idually o a leas 60 minu es, be o e basal O
2
consump ion and
CO
2
p oduc ion was analyzed o up o 60 minu es in 20 s in e als. Fo analysis o nonshi e ing he mogenesis capaci y in mice
housed and ea ed a 10Co 23
C, espec i ely, mice we e acclima ed o chambe s o a leas 60 minu es. Then basal ene gy
expendi u e a 23C was measu ed o 30 minu es, be o e injec ing mice subcu aneously wi h 1 mg/kg NE (A e enol, Sano i) and
measu ing O
2
consump ion and CO
2
p oduc ion in in e als o 20 s o a leas 60 minu es a e injec ion. The change in oxygen con-
sump ion (DVO
2
) was calcula ed by sub ac ing basal O
2
consump ion be o e NE injec ion om O
2
consump ion a e NE injec ion.
NE u no e
In o de o de e mine NE u no e in hype hy oid WT mice, n = 16 mice pe g oup we e ea ed o wo weeks wi h T3 o T4, espec-
i ely. On he las day o ea men mice we e injec ed wi h he y osine hyd oxylase inhibi o a-me hyl-DL- y osine me hyl es e
hyd ochlo ide (i.p. 120 mg/kg) (AMPT, M3281, Sigma Ald ich, Ge many) and sac i iced a e 0, 2, 4 and 6 hou s (n = 4 pe g oup
and ime poin ), espec i ely. Ca echolamines we e ex ac ed in 2% pe chlo ic acid (PCA, 48%–50%, 44464, Al a Aesa , MA,
USA) and quan i ied as desc ibed be o e (Dalley e al., 2002). P o ein con en o he samples was quan i ied using DC P o ein Assay
Ki acco ding o manu ac u e ’s p o ocol (Bio-Rad Labo a o ies, Inc.).
Gene exp ession analysis
Fo gene exp ession analysis, RNA was isola ed using QIAGEN RNeasy Ki s (QIAGEN, Ge many), ansc ibed in o cDNA (Molecula
Biology Re e Aid S and cDNA Ki , The mo Fishe Scien i ic, Ge many) ollowing manu ac u e ’s ins uc ions. qPCR analysis was
e2 Cell Repo s 27, 3385–3400.e1–e3, June 11, 2019
pe o med using SYBR G een PCR Mas e Mix (Roche, Ge many) and Quan S udio Applied Biosys ems (The mo Fishe Scien i ic,
Ge many). E iciency o he PCR was calcula ed using s anda d cu es and le els o gene exp ession we e no malized o a house-
keeping gene ( ibosomal p o ein la e al s alk subuni P0,Rplp0 and pep idylp olyl isome ase a,Ppia o adipose issues; Ppia and
hypoxan hine phospho ibosyl ans e ase,Hp o M. Soleus and M.Gas ocnemius, and Hp o pi ui a y and hypo halamus) using
he DDC
T
me hod. P ime sequences a e lis ed in Table S3.
Wes e n Blo
P o ein isola ion om snap- ozen issue was pe o med by homogenizing issues in RIPA bu e (150 mM NaCl, 50 mM T is-HCl pH
7.5, 0.1% w / ol SDS, 0.5% w / ol sodium deoxychola e, 1% ol/ ol Nonide P40, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium
py ophospha e, 1 mM NaVO
4
and 10 mM NaF), wi h eshly added P o ease inhibi o s (5892970001, Roche Diagnos ics GmbH, Ge -
many). P o ein concen a ion was de e mined using Pie ce BCA Ki (23227, The mo Scien i ic, Ge many). Elec opho esis was pe -
o med using SDS Gels (TGX S ain F ee Fas Cas Ac ylamide Ki (1610183, 1610185, Bio-Rad Labo a o ies, Ge many). P o eins we e
hen ans e ed on o PVDF memb ane (IPVH00010, Me ck Millipo e, L d, Co k, I eland), blocked o 1 hou in 5% milk in TBS-T and
incuba ed wi h p ima y an ibodies (an i-UCP1; Jas och e al., 2012); an i-Se ca2, 4388, Cell Signaling Technology, Inc.; an i-HSL,
4107, Cell Signaling Technology, Inc.; an i-pHSL(Se 660), 4126S, Cell Signaling Technology, Inc.; an i-OxPhos cock ail, 45-8099,
In i ogen; an i-Ty osine Hyd oxylase, C2928, Sigma; an i-b-ac in, A1978, Sigma; an i-HSP90, 4877, Cell Signaling Technology,
Inc.) o 16 hou s a 4C. A e wa d memb anes we e washed 4x10 minu es, incuba ed wi h seconda y an ibodies (an i- abbi poly-
clonal HRP-conjuga ed an ibody, P0448, DAKO, Denma k; an i-mouse polyclonal HRP-conjuga ed an ibody, P0447, DAKO,
Denma k) o 1 hou a oom empe a u e, and washed again o 4x10 minu es. Chemiluminescence was eco ded using Ad ans a
Wes e nB igh Quan um (K-12041-D20, Ad ans a, USA) and ChemiDoc Touch Imaging Sys em (Bio-Rad Labo a o ies, Ge many).
Quan i ica ion o band in ensi ies was pe o med using ImageLab
TM
So wa e (Bio-Rad Labo a o ies, Ge many). P o ein exp ession
analysis in he VMH (an i-pAMPKa(Th 172), 2535, Cell Signaling Technology, Inc.; an i-b-ac in(AC-74), A5316, Sigma) was pe o med
as desc ibed be o e (Al a ez-C espo e al., 2016). P o ein exp ession was no malized o o al p o ein con en ans e ed on o he
memb ane, o housekeeping p o eins, espec i ely.
Enzyme ac i i y and glycogen measu emen
Glycogen con en in li e and M. Gas ocnemius, as well as ac i i ies o Py u a e Kinase and hepa ic Phosphoenolpy u a e
Ca boxykinase o snap- ozen issue we e de e mined as desc ibed be o e (Vujo ic e al., 2009).
ELISA
Se um le els o T4 (EIA-1781, DRG Diagnos ics, Ge many) and T3 (DNOV053, No aTec Immundiagnos ica GmbH, Ge many) we e
de e mined acco ding o manu ac u e ’s ins uc ions. Hypo halamic T3 con en was measu ed in pooled hypo halamic punches o
he la e al and basal hypo halamus as desc ibed p e iously (Zhang e al., 2016). cAMP le els in ingWAT and iBAT we e de e mined
acco ding o manu ac u e ’s ins uc ions (RPN225, GE Heal hca e, UK). F ee a y acids in ingWAT and iBAT samples we e analyzed
acco ding o manu ac u e ’s ins uc ions (ab65341, abcam plc; UK).
His ology
Tissues we e ixed in 4% pa a o maldehyde o 48 hou s, dehyd a ed by washing in inc easing concen a ions o e hanol and xylol
be o e embedding in o pa a in. Tissues we e cu in 5 mm slides and s ained wi h Hemalun and Eosin acco ding o he manu ac u e ’s
p o ocol (X883.2, T865.3, Ca l Ro h GmbH&Co KG, Ge many).
QUANTIFICATION AND STATISTICAL ANALYSIS
Fo analysis o all da a Mic oso O ice Excel and G aphPad P ism 7 so wa e was used. Analysis o ci cadian hy hms was pe -
o med using Ci cWa e .1.4 so wa e (Os e e al., 2006). The a iances be ween g oups in es iga ed we e simila and app op ia e
es s we e pe o med o analyze di e ences be ween he espec i e g oups. Fo expe imen s wi h epea ed-measu emen s a RM
ANOVA was pe o med wi h indi idual pos - es s, he expe imen s o he UCP1 ko mice and T3 ea men we e analyzed using a
2-WAY ANOVA wi h pos - es s. All alues a e ep esen ed as mean ±SEM. Numbe o animals (n) pe expe imen a e depic ed in
each igu e.
Cell Repo s 27, 3385–3400.e1–e3, June 11, 2019 e3