ORIGINAL RESEARCH
published: 08 Sep embe 2020
doi: 10.3389/ med.2020.00455
F on ie s in Medicine | www. on ie sin.o g 1Sep embe 2020 | Volume 7 | A icle 455
Edi ed by:
Pe e Ko s en,
Uni e si y Medical Cen e
Gö ingen, Ge many
Re iewed by:
Ja ie Rod íguez-Ca io,
Uni e sidad de O iedo Mie es, Spain
Ka a ina Baue o a,
Cen e o Expe imen al Medicine,
Slo ak Academy o Sciences
(SAS), Slo akia
*Co espondence:
B uno Vidal
[email p o ec ed]
†These au ho s sha e
senio au ho ship
Special y sec ion:
This a icle was submi ed o
Rheuma ology,
a sec ion o he jou nal
F on ie s in Medicine
Recei ed: 11 Decembe 2019
Accep ed: 08 July 2020
Published: 08 Sep embe 2020
Ci a ion:
Cascão R, Vidal B, Ca alho T,
Lopes IP, Romão VC, Goncal es J,
Moi a LF and Fonseca JE (2020)
Celas ol E icacy by O al
Adminis a ion in he
Adju an -Induced A h i is Model.
F on . Med. 7:455.
doi: 10.3389/ med.2020.00455
Celas ol E icacy by O al
Adminis a ion in he
Adju an -Induced A h i is Model
Ri a Cascão1, B uno Vidal1*, Tânia Ca alho2, Inês Pascoal Lopes1, Vasco C. Romão 1,3,
João Goncal es4, Luis Fe ei a Moi a5† and João Eu ico Fonseca 1,3†
1Unidade de In es igação em Reuma ologia, Faculdade de Medicina, Ins i u o de Medicina Molecula -João Lobo An unes,
Cen o Académico de Medicina de Lisboa, Uni e sidade de Lisboa, Lisbon, Po ugal, 2Compa a i e Pa hology Uni , Ins i u o
de Medicina Molecula -João Lobo An unes, Faculdade de Medicina, Uni e sidade de Lisboa, Lisbon, Po ugal, 3Se iço de
Reuma ologia e Doenças Ósseas Me abólicas, Hospi al de San a Ma ia, Cen o Hospi ala Uni e si á io Lisboa No e,
Lisbon, Po ugal, 4Faculdade de Fa mácia, iMed – Resea ch Ins i u e o Medicines, Uni e sidade de Lisboa, Lisbon,
Po ugal, 5Inna e Immuni y and In lamma ion Labo a o y, Ins i u o Gulbenkian de Ciência, Oei as, Po ugal
Backg ound: We p e iously demons a ed ha celas ol has signi ican
an i-in lamma o y and bone p o ec i e e ec s when adminis e ed ia he in ape i oneal
ou e. Fo u he p eclinical e alua ion, an e ec i e o al adminis a ion o celas ol is
c ucial. He e we aimed o s udy he he apeu ic dose ange o i s o al adminis a ion.
Me hods: Celas ol (1–25 µg/g/day, N=5/g oup) was adminis a ed o ally o emale
adju an -induced a h i is (AIA) a s a e 8 days o disease induc ion o a pe iod o 14
days. A g oup o heal hy (N=8) and a h i ic (N=15) gende - and age-ma ched Wis a
a s was used as con ols. Du ing he ea men pe iod, he in lamma o y sco e, ankle
pe ime e , and body weigh we e measu ed. A he end o he ea men , he animals
we e sac i iced, blood was collec ed o clinical pa hology, nec opsy was pe o med wi h
collec ion o in e nal o gans o his opa hological analysis, and paw samples we e used
o disease sco ing.
Resul s: Doses highe han 2.5 µg/g/day o celas ol educed he in lamma o y sco e
and ankle swelling, p ese ed join s uc u e, hal ed bone des uc ion, and diminished
he numbe o syno ial CD68+mac ophages. Bone eso p ion and u no e we e
also educed a 5 and 7.5 µg/g/day doses. Howe e , he dose o 7.5 µg/g/day was
associa ed wi h hymic and li e lesions, and highe doses showed se e e oxici y.
Conclusion: O al adminis a ion o celas ol abo e 2.5 µg/g/day amelio a es a h i is.
This da a suppo s and gi es ele an in o ma ion o he de elopmen o a p eclinical
es o celas ol in he se ing o a ch onic model o a h i is since heuma oid a h i is is
a long- e m disease.
Keywo ds: heuma oid a h i is, adju an -induced a h i is, celas ol, dose, e icacy
Cascão e al. E icacy and Toxici y o Celas ol
INTRODUCTION
The he apeu ic e ec o celas ol has been demons a ed
in se e al in lamma o y diseases. Celas ol is a pen acyclic-
i e pene compound ha can be ound in oo ex ac s om
T ip e ygium wil o dii (TW) (1), an he b used in Chinese
medicine (2–4).
In he las 5 yea s, inc easing e idence o he he apeu ic
po en ial o celas ol in he ea men o heuma oid a h i is
(RA) has eme ged. S udies ha e sugges ed ha he an i-
in lamma o y p ope ies o celas ol can be mainly a ibu ed o
he egula ion o cy okine p oduc ion (5–8), he modula ion o
in lamma o y cells (8–13), he inhibi ion o os eoclas ogenesis,
and bone p o ec ion (8,14–16), mos ly due o i s abili y o
down egula e he NF-kB pa hway.
Speci ically, we ha e demons a ed ha he in ape i oneal
adminis a ion o celas ol supp essed in lamma o y signs (7),
p ese ed join s uc u e, wi h ab oga ion o he in lamma o y
in il a e and cellula p oli e a ion (7,8), and hal ed ocal bone
damage in he adju an -induced a h i is (AIA) a model (8,17).
This inhibi o y e ec o cellula in il a ion and p oli e a ion may
p o e o be o in e es o ea he de elopmen o he syno ial
umo -like pannus issue cha ac e is ic o RA, one o he main
con ibu o s o bone damage. Impo an ly, we ha e epo ed
ha his compound is able o signi ican ly dec ease he numbe
o sublining CD68+syno ial mac ophages (8), a bioma ke o
ea men e icacy in RA (18–20). So a , we obse ed ha he
in ape i oneal adminis a ion o celas ol o AIA a s was no
associa ed wi h o e signs o oxici y (8).
Despi e he he apeu ic po en ial o celas ol, u he clinical
applica ion seems o be limi ed by poo wa e solubili y (21),
low o al bioa ailabili y (22), possible side e ec s (23–26), and
a iabili y in dose egimens (9,27–30).
The e o e, in o de o ad ance he p eclinical de elopmen o
celas ol as a candida e he apeu ic compound o RA ea men ,
we analyzed he he apeu ic dose ange o o al adminis a ion o
pu e celas ol using he AIA a model.
MATERIALS AND METHODS
Animals
The AIA model has been he mos ex ensi ely used a h i ic
a model o s udy an i-a h i ic agen s because i has an
excellen ack eco d o p edic ing bo h ac i i y and oxici y.
AIA a s sha e key ea u es ela ed o RA, making hem a
c i ical ool in d ug de elopmen , and exhibi he g ea es
magni ude o disease when compa ed wi h o he models o
a h i is (31). Eigh -week-old emale Wis a AIA a s weighing
230–250 g we e pu chased om Cha les Ri e Labo a o ies
In e na ional (Massachuse s, USA). Cha les Ri e Labo a o ies
pe o med he induc ion o adju an disease using F eud’s
comple e adju an , supplemen ed wi h mycobac e ium, and
injec ed in he igh oo pad. The AIA a s we e main ained
unde speci ic-pa hogen- ee condi ions, andomly housed pe
g oup unde s anda d labo a o y condi ions (a 22◦C unde
10-h ligh /14-h da k condi ions), and gi en ee access o ood
(RM3, SDS Die s, UK) and wa e (ul apu e). In addi ion,
o minimize animal discom o , pape sha ings we e used as
bedding ma e ial in Double Decke GR1800 cages (Techniplas ,
UK) wi h i e animals each. The c i e ia o a humane sac i ice
we e de e mined as p e iously published (8), and he animals
we e sac i iced when p esen ing he maximum in lamma o y
sco e in mo e han wo limbs o when weigh loss exceeded 20%.
In acco dance wi h Di ec i e 2010/63/EU, all animal p ocedu es
we e app o ed by he ins i u ional animal wel a e body (ORBEA-
iMM) and licensed by he Po uguese compe en au ho i y
(DGAV—Di ecção Ge al de Alimen ação e Ve e iná ia, license
numbe : 0421/000/000/2016).
Celas ol P epa a ion and Adminis a ion
Celas ol (Sigma, Missou i, USA) s ock solu ion o 10 mg/ml was
p epa ed using e hanol 100% as sol en ( ehicle). A ecen s udy
has es ed he solubili y o celas ol in di e en ehicles, and we
ha e p e iously demons a ed ha e hanol is one o he mos
e icien sol en s o his compound (21). This celas ol s ock
solu ion was u he dissol ed in PEG400 (Sigma, S . Louis, USA)
(1, 2.5, 5, 7.5, 12.5, and 25 µg/g in 1 ml) and adminis a ed by o al
ga age o AIA a s o 14 consecu i e days (N=5 a s/g oup).
The sample size in each g oup was calcula ed using ee sample
size calcula ing G∗Powe e sion 3.1.9.2 so wa e [ ype o powe
analysis: a p io i;αe o p obabili y: 0.05; powe (1-βe o
p obabili y): 0.95; e ec size d: 2.59; ac ual powe : 0.976]. This
calcula ion was based on ou own p e ious da a (7,8,17).
Ou s udy was app o ed by he ins i u ional animal wel a e
body, licensed by he Po uguese compe en au ho i y, and
complied wi h good e hical, scien i ic, legal, and economic
easons o using labo a o y animals, including he 3R p inciple
( eplace, educe, and e ine). Focused on he “ educe” ule, we
used he minimum numbe o animals, calcula ed by he ee
sample size calcula ing G∗Powe so wa e, in o de o pe o m
his s udy.
The need o daily adminis a ions is suppo ed by he s udy
o Zhang e al., which showed ha he hal -li e o pu e celas ol is
∼10 h (22). Based on his publica ion, we ha e also calcula ed he
o al dose o 2.5 µg/g/day as he equi alen o he in ape i oneal
dose o 1 µg/g/day ha we had p e iously ound o be e ec i e
in he ea men o a h i is in he same a model (7,8).
The calcula ion was based on he ac ion o he celas ol dose
abso bed in he po al blood a e o al adminis a ion, which was
17.6%. In ape i oneal dose calcula ion ook in o conside a ion
ha he in apo al dose will ha e highe bioa ailabili y.
The e o e, he ela ionship be ween he a ea unde he cu e
and he doses adminis e ed in he o al and he in apo al dose
was used o de e mining he in ape i oneal dose. T ea men
was ini ia ed a e 8 days o disease induc ion, a he acu e
clinical s age o a h i is p og ession ( he apeu ic in e en ion)
(32). Heal hy non-a h i ic (N=8) and a h i ic un ea ed
(N=15) emale age-ma ched Wis a a s we e used as con ols.
The a h i ic un ea ed g oup ecei ed an equal olume o ehicle
h ough o al ga age. The ehicle p opo ion o e hanol and
PEG400 used was in he same p opo ion as he one used in he
celas ol- ea ed g oups.
The a s we e sac i iced a e 22 days o disease
p og ession by CO2na cosis, and blood, in e nal
F on ie s in Medicine | www. on ie sin.o g 2Sep embe 2020 | Volume 7 | A icle 455
Cascão e al. E icacy and Toxici y o Celas ol
o gans, as well as paw samples we e collec ed. S udies
using he AIA model a e gene ally comple ed a his
ime poin due o a pla eau e ec o in lamma o y
mani es a ions (7,31).
A h i is Se e i y E alua ion
Disease ac i i y was clinically e alua ed du ing he pe iod
o ea men by wo independen in es iga o s using an
in lamma o y sco e and by measu ing he ankle pe ime e
as eadou o a icula swelling. The in lamma o y sco e
was measu ed by coun ing he sco e o each limb join
in a scale o 0–3 (0—absence, 1—e y hema, 2—e y hema
and swelling, and 3—de o mi ies and unc ional impai men ).
The o al sco e o each animal was de ined as he sum
o he pa ial sco es o each a ec ed join (7,33). Body
weigh was also egis e ed, e e y 2 days, h oughou he
expe imen al p ocedu e.
Clinical Pa hology and His ological
Analysis
Blood was collec ed om he hea and used o se um
biochemis y measu emen o c ea ine kinase (CK),
u ea, lac a e dehyd ogenase (LDH), alanine ansaminase
(ALT) (BioAssay Sys ems, Cali o nia, USA), and p o-
ANP (Biomedica, Viena, Aus ia) by enzyme-linked
immunoso ben assay (ELISA) echnique, acco ding o he
manu ac u e ’s ins uc ions. The ELISA measu emen was
pe o med using he pla e eade In ini e M200 (Tecan,
Mannedo , Swi ze land).
Nec opsy was pe o med, and he le hind paw was
collec ed o ibio a sal join his opa hological analysis and
disease sco ing, and he li e , spleen, kidney, lung, hymus,
hea , gas oin es inal ac , and long bone (hume us)
we e collec ed o ou ine his opa hological analysis
o assess signs o celas ol-induced oxici y. B ie ly, all
o gans and issues we e immedia ely ixed in 10% neu al
bu e ed o malin, he bones we e u he decalci ied
in 10% o mic acid, and all samples we e p ocessed o
pa a in embedding.
Fo his ological disease ac i i y sco ing, se ial 5-µm sec ions
o he ibio a sal join s we e s ained wi h hema oxylin and
eosin (H&E) and immunohis ochemis y was pe o med
using he ollowing an ibodies: mouse monoclonal an i-CD68
(Abcam, Camb idge, UK), mouse monoclonal an i-os eocalcin
(os eoblas ma ke ; indica o o os eoblas ic ac i i y; Abcam,
Camb idge, UK), abbi polyclonal an i-ca hepsin K (os eoclas
ma ke ; ma u e os eoclas enzyme; Bio by , Camb idge,
UK), and abbi polyclonal an i-Ki67 (Abcam, Camb idge,
UK) an ibodies. The issue sec ions we e incuba ed wi h
he p ima y an ibody and wi h EnVision+(Dako, Glos up,
Denma k). Colo was de eloped in a solu ion con aining
diaminobenzadine- e ahyd ochlo ide (Sigma, Missou i, USA)
and 0.5% H2O2in phospha e-bu e ed saline bu e (pH
7.6). The slides we e coun e s ained wi h hema oxylin and
moun ed. His ological disease ac i i y sco ing in he ibio a sal
join s was pe o med by an independen esea che blinded
o he expe imen al g oups using ou semi-quan i a i e
sco es, as p e iously published: sublining laye in il a ion
sco e (0—none o di use in il a ion, 1—lymphoid cell
agg ega e, 2—lymphoid ollicles, and 3—lymphoid ollicles
wi h ge minal cen e o ma ion), lining laye cell numbe
sco e (0— ewe han h ee laye s, 1— h ee o ou laye s,
2— i e o six laye s, and 3—mo e han six laye s), bone
e osion sco e (0—no e osions, 1—minimal, 2—mild, 3—
mode a e, and 4—ma ked), and global disease se e i y sco e
(0—no signs o in lamma ion, 1—mild, 2—mode a e, and
3—se e e) (7,34,35). The p oli e a ion o syno iocy es was
also analyzed using a semi-quan i a i e sco e (0–4) o Ki67
immunos aining (0—no s ained cells, 1−0–25% s aining,
2−25–50% s aining, 3−50–75% s aining, and 4—mo e han 75%
s ained cells) (7). Images we e acqui ed in a Leica DM2500 (Leica
Mic osys ems, We zla , Ge many) coupled o a Leica MC170
HD mic oscope came a.
Fo he assessmen o celas ol-induced oxici y, 4-µm
sec ions o he li e , spleen, kidney, lung, hymus, hea ,
gas oin es inal ac , and hume us we e s ained wi h H&E and
analyzed by a pa hologis (TC) blinded o he expe imen al
g oups. The slides we e scanned and images we e acqui ed by a
Hamama su NanoZoome SQ slide scanne . The classi ica ion o
lesions ollowed p e iously published c i e ia (36–42).
Measu emen o Se um Bone Tu no e and
Reso p ion Ma ke s
Bone u no e was analyzed by quan i ying he le els o
a a e- esis an acid phospha ase 5b (TRACP-5b), p ocollagen
ype 1 amino- e minal p opep ide (P1NP), and C- e minal
c oss-linked elopep ide o ype I collagen (CTX-I) in se um
using ELISA (Immunodiagnos ic Sys em, Boldon, UK).
All comme cial assays we e pe o med acco ding o he
manu ac u e s’ ins uc ions, and s anda d cu es we e gene a ed
using he supplied e e ence concen a ions. Measu emen
was pe o med using a pla e eade In ini e M200 (Tecan,
Mannedo , Swi ze land).
S a is ical Analysis
No mali y dis ibu ion was assessed by D’Agos ino and Pea son
es . The ea ed g oups (celas ol 1, 2.5, 5, and 7.5 ug/g)
we e compa ed agains he un ea ed a h i ic g oup wi h he
Mann–Whi ney es wi h Bon e oni co ec ion o accoun
o mul iple compa isons, as p e iously epo ed (17). Thus,
applying he Bon e oni co ec ion, we di ided he global
signi icance le el a 0.05 by he numbe o independen es s
(n=4) o ge he Bon e oni c i ical alue o p<0.0125,
below which a es would be signi ican . In addi ion, he
Mann–Whi ney es was also used o compa e di e ences
be ween he o he wo independen g oups: heal hy non-
a h i ic and un ea ed a h i ic a s. Fo pai ed samples
(e.g., di e en ime poin s), we ha e used he Wilcoxon
ma ched-pai s signed- ank es . In hese cases, p<0.05
we e conside ed o be signi ican . All s a is ical analyses we e
pe o med using he G aphPad P ism V.5.01 (G aphPad,
Cali o nia, USA). Da a we e p esen ed as median wi h
in e qua ile ange.
F on ie s in Medicine | www. on ie sin.o g 3Sep embe 2020 | Volume 7 | A icle 455
Cascão e al. E icacy and Toxici y o Celas ol
RESULTS
O al Celas ol Imp o ed he Clinical
Ou come and Amelio a ed he
His opa hological Aspec s o AIA Ra s
The onse o a h i is in he con ala e al ankle join o he
injec ion si e occu ed app oxima ely a day 8 pos -disease
induc ion. As demons a ed in Figu e 1A, all animals showed
clinical signs o a h i is by he 4 h day o disease induc ion, and
a e 10 days, he un ea ed a h i ic g oup showed accele a ed
disease p og ession. In con as , a e 6 and 7 days o ea men ,
he a s ea ed wi h 2.5 and 5 µg/g/day (p=0.0111 and
p=0.006) and 7.5 µg/g/day(p=0.010) o celas ol showed
a signi ican lowe in lamma o y sco e compa ed o un ea ed
a h i ic a s. O no e, o al celas ol a 1 µg/g/day had no e ec
in a h i is p og ession. A e 3 days o ea men , all animals
(5 ou o 5) ea ed wi h a dose o 25 µg/g/day and 3 animals
(3 ou o 5) ea ed wi h 12.5 µg/g/day we e eu hanized due
o p og essi e weigh loss (o e 15% o body weigh ), onse o
dia hea and espi a o y dis ess. In he g oup o animals ea ed
wi h 7.5 µg/g/day o celas ol, only one animal (1 ou o 5) was
eu hanized a e 10 days o ea men wi h he same clinical signs.
A e 14 days o ea men , he dosage anging om 2.5 o
7.5 µg/g/day showed a signi ican an i-in lamma o y e ec , as
assessed by he e alua ion o he clinical in lamma o y sco e
(p=0.0023 wi h 2.5 µg/g/day, p=0.0016 wi h 5 µg/g/day, and
p=0.0028 wi h 7.5 µg/g/day s. a h i ic animals, as shown
in Figu e 1B) and also by he measu emen o ankle pe ime e
(p=0.0053 wi h 2.5 µg/g/day, p=0.0053 wi h 5 µg/g/day and
p=0.0118 wi h 7.5 µg/g/day s. un ea ed a h i ic animals, as
shown in Figu e 1C).
O no e, no di e ences we e obse ed in body weigh a e
14 days o ea men , compa ing each dose o celas ol (up o
7.5 µg/g/day) wi h he ehicle con ol animals (Figu e S1A).
Addi ionally, when compa ing, wi hin he same g oup, he i s
and las day o ea men , we we e able o obse e an inc ease
in body weigh in heal hy a s (p=0.0078, Figu e S1B) and
a signi ican weigh loss in he un ea ed a h i ic g oup, as
expec ed (p=0.0074, Figu e S1C). Howe e , no weigh loss was
obse ed in AIA animals ea ed wi h celas ol using doses up o
7.5 µg/g/day (Figu es S1D–G).
As shown in Figu e 2, syno ial hype plasia and ma ked
in lamma o y cell in il a ion we e seen in ibio a sal join s o
un ea ed a h i ic a s (p=0.0019 and p=0.0001 s. heal hy
con ols, espec i ely), also associa ed wi h signi ican bone
e osion (p=0.0008 s. heal hy con ols). In con as , in celas ol-
ea ed a s, in lamma o y in il a es we e educed in all dose
egimens, excep o he 1 µg/g/day g oup (Figu e 2A,p=0.0010
in 2.5 µg/g/day, p=0.0006 in 5 µg/g/day and p=0.0016 in 7.5
µg/g/day s. a h i ic a s). We also obse ed a educ ion in he
numbe o cells p esen in he syno ial lining laye wi h he dose
o 2.5 µg/g/day (Figu e 2B,p=0.0110 s. a h i ic a s), and a
ma ked endency o dec ease in he 5 and 7.5 µg/g/day doses
(p=0.0220; p=0.0225, espec i ely, s. a h i ic a s). These
concen a ions o celas ol we e also e ec i e in p e en ing bone
a icula des uc ion (Figu e 2C,p=0.0053 in 2.5 µg/g/day,
p=0.0069 in 5 µg/g/day, and p=0.0118 in 7.5 µg/g/day s.
a h i ic a s), wi h animals p esen ing a no mal join s uc u e
a he end o he s udy pe iod (Figu e 2D,p=0.0028 in 2.5
µg/g/day, p=0.0048 in 5 µg/g/day, and p=0.0041 in 7.5
µg/g/day s. a h i ic a s).
In addi ion, we obse ed ha animals ea ed wi h celas ol
in he dose ange be ween 2.5 and 7.5 µg/g/day ha e educed
syno ial cell p oli e a ion, as assessed in he ibio a sal join s by
Ki67 immunos aining (p=0.0020, p=0.0055, and p=0.0026
s. a h i ic animals, espec i ely) (Figu e 2E).
Finally, we obse ed ha un ea ed a h i ic a s had a
highe numbe o in il a ing CD68+syno ial mac ophages as
compa ed o heal hy con ols (p=0.0001) (Figu e 2F). Celas ol
adminis a ion was associa ed wi h a signi ican dec ease in he
numbe o CD68+mac ophages in il a ing he a h i ic join
(p=0.0014 in 2.5 µg/g/day, p=0.0095 in 5 µg/g/day s. a h i ic
a s; o he 7.5 µg/g/day he e is a s ong endency al hough no
eaching signi icance, p=0.0213).
O al Celas ol Reduced he Numbe o
Join Os eoclas s and Os eoblas s
The immunolocaliza ion o os eoclas s and os eoblas s in
subchond al bone issue a he ibia/ alus egion was pe o med
o e alua e a possible celas ol dose-dependen e ec on
bone emodeling. Un ea ed a h i ic a s showed inc eased
os eoclas numbe s (ca hepsin k+cells) in he ibio a sal bones
(p=0.004 s. heal hy con ols, Figu e 3A). Impo an ly, celas ol
adminis a ion was associa ed wi h a signi ican dec ease in
he numbe o os eoclas s, o le els simila o heal hy con ols
(p=0.004 in 5 µg/g/day and p=0.004 in 7.5 µg/g/day s.
a h i ic a s). Un ea ed a h i ic a s also showed inc eased
numbe s o os eoblas s (os eocalcin-posi i e cells) (p=0.0005
s. heal hy con ols, Figu e 3B), a pheno ype also e e sed by
celas ol adminis a ion, wi h a signi ican educ ion in he
numbe o os eoblas s as compa ed o AIA a s (p=0.0005 in 5
µg/g/day and p=0.0005 in 7.5 µg/g/day) eaching le els simila
o heal hy con ols. In he case o a s ea ed wi h celas ol a he
dose o 2.5 µg/g/day he e is a endency owa d a dec ease in he
numbe o os eoblas s ha did no each s a is ical signi icance
(p=0.02 s. un ea ed AIA a s).
O al Celas ol Reduced Bone Tu no e and
Reso p ion Ma ke s in AIA Ra s
In un ea ed a h i ic a s, he e was no signi ican inc ease in
TRACP-5b le els a he end o he s udy pe iod (Figu e 4A).
Impo an ly, bo h 5 and 7.5 µg/g/day doses o celas ol educed
TRACP-5b le els, when compa ed wi h un ea ed a h i ic
a s (p=0.0031 and p=0.0065, espec i ely), sugges ing a
dec ease in bone eso p ion. In addi ion, he ele a ed le els o
P1NP obse ed in un ea ed a h i ic a s in compa ison wi h
heal hy con ols (p=0.0074, Figu e 4B) p esen ed a ma ked
endency o dec ease wi h he celas ol dose o 5 µg/g/day
(p=0.0145 s. un ea ed a h i ic animals). In acco dance,
celas ol adminis a ion was able o signi ican ly educe he
CTX-I le els (p=0.0079 o 5 µg/g/day, Figu e 4C) and also
induced a s ong endency o dec ease wi h he 7.5 µg/g/day dose,
F on ie s in Medicine | www. on ie sin.o g 4Sep embe 2020 | Volume 7 | A icle 455
Cascão e al. E icacy and Toxici y o Celas ol
FIGURE 1 | O al celas ol signi ican ly imp o ed he clinical ou come in adju an -induced a h i is a s. The in lamma o y sco e measu ed h oughou he ea men
pe iod (A), he in lamma o y sco e (B), and he ankle pe ime e (C) e alua ed by day 22 a e disease induc ion showed ha he ange o dosage be ween 2.5 and 7.5
µg/g/day was e ec i e in p e en ing a h i is p og ession. Heal hy N=8, a h i ic N=17, celas ol 1 µg/g/day N=5, celas ol 2.5 µg/g/day N=5, celas ol 5
µg/g/day N=5, celas ol 7.5 µg/g/day N=5, celas ol 12.5 µg/g/day N=5, and celas ol 25 µg/g/day N=5. The da a a e shown as median wi h in e qua ile
ange. The di e ences we e conside ed as s a is ically signi ican o p<0.05, acco ding o Mann–Whi ney es s o compa isons be ween heal hy and a h i ic
g oups and p<0.0125 acco ding o Mann–Whi ney es s wi h Bon e oni co ec ion o compa isons be ween a h i ic and celas ol- ea ed g oups. *p<0.05 and
p<0.0125.
F on ie s in Medicine | www. on ie sin.o g 5Sep embe 2020 | Volume 7 | A icle 455

Cascão e al. E icacy and Toxici y o Celas ol
FIGURE 2 | O al celas ol ea men amelio a ed he his opa hological aspec s o adju an -induced a h i is (AIA) a s and educed syno ial CD68+mac ophages.
Celas ol signi ican ly impai ed in lamma o y cell in il a ion (A), was associa ed wi h a numbe o syno ial lining laye s simila o no mal alues (B), and educed bone
e osions (C), hus p ese ing he no mal join s uc u e (D). The celas ol- ea ed AIA a s showed a signi ican educ ion in syno ial cell p oli e a ion as assessed by
he Ki67 ma ke (E) and in he numbe o syno ial CD68-posi i e cells (F). Heal hy N=8, a h i ic N=17, celas ol 1 µg/g/day N=5, celas ol 2.5 µg/g/day N=5,
celas ol 5 µg/g/day N=5, and celas ol 7.5 µg/g/day N=4. The da a a e exp essed as median wi h in e qua ile ange. The di e ences we e conside ed as
s a is ically signi ican o p<0.05, acco ding o Mann–Whi ney es s o compa isons be ween heal hy and a h i ic g oups and p<0.0125 acco ding o
Mann–Whi ney es s wi h Bon e oni co ec ion o compa isons be ween a h i ic and celas ol- ea ed g oups. *p<0.05 and p<0.0125.
F on ie s in Medicine | www. on ie sin.o g 6Sep embe 2020 | Volume 7 | A icle 455
Cascão e al. E icacy and Toxici y o Celas ol
FIGURE 3 | O al celas ol educed he numbe o os eoclas s and os eoblas s in a h i ic join s. Ca hepsin k-posi i e cells (A) and os eocalcin-posi i e cells (B) we e
iden i ied in a h i ic join s by immunohis ochemis y. Celas ol ea men signi ican ly educes bo h ypes o cells in he doses o 5 and 7.5 µg/g/day. The paw samples
we e collec ed a he ime o sac i ice. Heal hy N=8, a h i ic N=17, celas ol 1 µg/g/day N=5, celas ol 2.5 µg/g/day N=5, celas ol 5 µg/g/day N=5, and
celas ol 7.5 µg/g/day N=4. The da a a e exp essed as median sco e wi h in e qua ile ange. The di e ences we e conside ed as s a is ically signi ican o
p<0.05, acco ding o he Wilcoxon signed- ank es s o compa isons be ween heal hy and a h i ic g oups and p<0.0125 acco ding o Mann–Whi ney and
Wilcoxon signed- ank es s wi h Bon e oni co ec ion o compa isons be ween a h i ic and celas ol- ea ed g oups. *p<0.05 and p<0.0125.
FIGURE 4 | O al celas ol was associa ed wi h he educ ion o bone u no e and eso p ion ma ke s. TRACP-5b (A), P1NP (B), and CTX-I (C) le els we e quan i ied
in a se um samples collec ed a he ime o sac i ice. Celas ol seems o educe he le els o TRACP-5b, P1NP, and CTX-I in he ea ed animals in compa ison wi h
he a h i ic a s. Heal hy N=8, a h i ic N=17, celas ol 1 µg/g/day N=5, celas ol 2.5 µg/g/day N=5, celas ol 5 µg/g/day N=5, and celas ol 7.5 µg/g/day
N=4. The da a a e exp essed as median wi h in e qua ile ange. The di e ences we e conside ed as s a is ically signi ican o p<0.05, acco ding o
Mann–Whi ney es s o compa isons be ween heal hy and a h i ic g oups and p<0.0125 acco ding o Mann–Whi ney es s wi h Bon e oni co ec ion o
compa isons be ween a h i ic and celas ol- ea ed g oups. *p<0.05 and p<0.0125.
(p=0.0159) when compa ed wi h un ea ed a h i ic a s, which
showed an accele a ed bone u no e wi h high le els o CTX-I
(p=0.0016 s. heal hy con ols). Also, he CTX-I se um le els
had a endency owa d a dec ease a 2.5 µg/g/day, bu i did no
each s a is ical signi icance.
Highe Dose o Celas ol Was Associa ed
Wi h Signs o Toxici y
A his opa hological analysis o selec ed o gans showed
his ological changes associa ed wi h he disease model and
his ological changes associa ed wi h he es compound.
F on ie s in Medicine | www. on ie sin.o g 7Sep embe 2020 | Volume 7 | A icle 455
Cascão e al. E icacy and Toxici y o Celas ol
Conce ning he disease- ela ed his opa hological aspec s,
gian -cell g anulomas we e seen in li e , lung, and spleen
in all g oups, wi h a ied se e i y un ela ed wi h ea men
o dose le els (Figu e S2). G ossly, mul iple g ay o g ay-
b own oci we e obse ed o e he pleu al su ace o he
TABLE 1 | His opa hological indings in Wis a a s upon celas ol ea men a
di e en dose le els.
Ta ge
o gan
E ec Se e i y
g adea
Celas ol (µg/g/day)
0 1 2.5 5 7.5
Thymus Nec osis,
lymphocy e
(+2) 0/16 0/5 0/5 0/6 1/6
(+3) 0/16 0/5 0/5 0/6 1/6
Li e In lamma o y
cell in il a ion,
pe ibilia y
(in ahepa ic)
(+1) 0/16 0/5 2/5 5/6 2/6
(+2) 0/16 0/5 0/5 0/6 2/6
(+3) 0/16 0/5 0/5 0/6 2/6
Bile duc
hype plasia
(+1) 0/16 0/5 3/5 4/6 4/6
(+2) 0/16 0/5 0/5 1/6 2/6
aSe e i y g ade: (+1), mild; (+2), mode a e; (+3), ma ked.
lung and he su ace o he li e and spleen. Mic oscopically,
hese changes co esponded o he a iably sized mul i ocal
o coalescing g anulomas, cha ac e ized by dense agg ega es o
mos ly mac ophages admixed wi h equen mul inuclea ed gian
cells and ewe lymphoid cells (Figu e S3). In he li e , hese
g anulomas we e p esen in he pa enchyma; in he lung, hey
we e mos ly a anged a ound b onchioles, a ound o adjacen o
la ge essels; and in he lung in e s i ium and in he spleen hey
we e mos ly seen in he ed pulp.
Rega ding possible celas ol-induced his opa hological
aspec s, lesions we e seen exclusi ely in he hymus and he li e
o medium- o high-dose g oups, and he se e i y g ade was
ela ed wi h dose (Figu e S4). The hymus showed lymphocy e
nec osis, mul i ocal, and mild o mode a e deple ion o
lymphocy es in he co ical zone. Se e i y anged om mode a e
o ma ked and was seen exclusi ely in wo o six animals o
he high-dose g oup (7.5 µg/g/day) (Table 1). In he li e , mild
lymphocy e- ich pe ibilia y in lamma o y cell in il a ion, and
bile duc hype plasia we e only seen in wo and h ee ou o
i e animals, espec i ely, in he 2.5 µg/g/day dose-le el g oup
(Table 1). Highe oxici y incidence was obse ed in he 5 and
7.5 µg/g/day dose g oups, wi h a oxici y se e i y g ade anging
om mild o mode a e in he medium dose and om mild o
ma ked in he high dose. Due o he design o he expe imen
and o he use o an acu e a model o a h i is, which apidly
p og esses and esol es, i was no possible o assess i changes
associa ed wi h celas ol ea men we e e e sible as no analysis
FIGURE 5 | Se um biochemis y o li e , enal, and ca diac ma ke s upon o al adminis a ion o celas ol. The se um le els o ALT (A), CK (B), LDH (C), u ea (D), and
p o-ANP (E) we e measu ed by ELISA o e alua e li e , kidney, and ca diac oxici y. Only he le els o LDH we e inc eased in animals ea ed wi h 7.5 µg/g/day o
celas ol. Heal hy N=8, a h i ic N=10, celas ol 1 µg/g/day N=5, celas ol 2.5 µg/g/day N=5, celas ol 5 µg/g/day N=5, and celas ol 7.5 µg/g/day N=4.
The da a a e exp essed as median wi h in e qua ile ange. The di e ences we e conside ed as s a is ically signi ican o p<0.05, acco ding o Mann–Whi ney es s
o compa isons be ween heal hy and a h i ic g oups and p<0.0125 acco ding o Mann–Whi ney es s wi h Bon e oni co ec ion o compa isons be ween a h i ic
and celas ol- ea ed g oups. *p<0.05 and p<0.0125.
F on ie s in Medicine | www. on ie sin.o g 8Sep embe 2020 | Volume 7 | A icle 455
Cascão e al. E icacy and Toxici y o Celas ol
was done a e suspension o he compound. No signi ican
his ological changes we e seen in any o he o gans.
Mo eo e , he se um biochemis y o ALT, u ea, p o-ANP,
and CK showed le els wi hin he e e ence ange o he
labo a o y o his s ain/species (Figu es 5A,B,D,E). LDH was
inc eased a he dosage le el o 7.5 µg/g/day (as shown in
Figu e 5C,p=0.0062 s. un ea ed a h i ic a s, espec i ely).
The elec oca diog am emained unal e ed in celas ol- ea ed
animals (da a no shown).
DISCUSSION
In his s udy, we aimed o de e mine he he apeu ic o al dose
o pu e celas ol and he associa ed oxici y signs in he AIA
a model.
The doses o celas ol we e ex apola ed om he
in ape i oneal dose used in ou p e ious s udies (7,8,17),
also aking in o conside a ion he bioa ailabili y and he
pha macokine ics o pu e celas ol upon o al adminis a ion in
a s (22).
We obse ed ha , a e 3 days o ea men , all animals
ea ed wi h 25 µg/g/day o celas ol and mo e han 50% o hose
ea ed wi h 12.5 µg/g/day showed a signi ican decline in heal h
s a us, wi h p og essi e weigh loss, dia hea, and espi a o y
dis ess, all o hese un ela ed wi h he disease model, and we e
eu hanized. These da a sugges ha he dose o 25 µg/g/day is
he le hal dose (LD) and he dose o 12.5 µg/g/day is he LD50 o
he o al adminis a ion o pu e celas ol, and hese dose g oups
we e he e o e excluded om his s udy. A e 10 days unde
celas ol ea men , one ou o i e a s ea ed wi h 7.5 µg/g/day
also p esen ed simila clinical signs and was eu hanized.
We ha e obse ed ha he o al dose o 1 µg/g/day o celas ol
was no e ec i e in he ea men o a h i is. In con as , he
o al doses o 2.5, 5, and 7.5 µg/g/day showed signi ican an i-
in lamma o y p ope ies as assessed by he e alua ion o he
in lamma o y sco e and ankle swelling. Mos o he in lamma o y
and bone- ela ed pa ame e s showed a dose-dependen e ec .
Howe e , in some o he analyses, his e ec did no each
s a is ical signi icance, which migh be due o he smalle sample
size in hese g oups. The celas ol doses o 2.5, 5, and 7.5 µg/g/day
we e able o educe syno ial cell in il a ion and p oli e a ion and
also dec eased bone e osions in join s. Impo an ly, a he doses
o 2.5 and 5 µg/g/day and wi h a s ong endency a he dose o
7.5 µg/g/day, celas ol educed he numbe o sublining CD68+
syno ial mac ophages, a bioma ke o clinical esponse (18–20).
These da a sugges ha , a hese o al dosages, celas ol is e ec i e
o he ea men o a h i is. Addi ionally, he doses o 5 and
7.5 µg/g/day educed he numbe o os eoclas s and os eoblas s
p esen in join issues. This obse a ion is in ag eemen wi h
he educ ion o he bone eso p ion ma ke TRACP-5b and also
wi h he dec ease o he bone eso p ion ma ke CTX-I in he 5
µg/g/day ea ed g oup and a endency o diminish also a he
7.5 µg/g/day dose, sugges ing he con ol o he accele a ed bone
u no e cha ac e is ic o a h i is. Ou esul s a e in line wi h
o he s, whe e ci cula ing le els o TRACP-5b a e no a ec ed
by a h i is induc ion, con a ily o he ma ked os eoclas ic
ac i i y occu ing in his a h i ic model. Howe e , he e a e
inc eased le els o TRACP-5b in he p o ein ex ac s ob ained
om in lamed join s (43). These indings sugges ha TRACP-5b
e lec s bone eso p ion mo e accu a ely when measu ed locally
a he han sys emically (43,44). Rega ding TRACP-5b, CTX-I,
and P1NP esul s, celas ol migh ha e a di ec e ec on bone
me abolism in he se ing o in lamma ion. Impo an ly, hese
esul s sugges ha he o al adminis a ion o celas ol is also
able o con ibu e o he p e en ion o bone damage, as we ha e
p e iously demons a ed using AIA a s unde ea men wi h 1
µg/g/day o celas ol ia in ape i oneal ou e (17).
Some in lamma o y dis u bances we e s ill no ed in he
ea ed animals. This is a consequence o he in lamma o y
p ocess ha occu ed du ing he i s days a e disease induc ion
( om day 0 up o day 8) bu be o e celas ol ea men . This
is in acco dance wi h ou p e ious obse a ions showing ha
in lamma ion induces changes since he i s days o a h i is
de elopmen (45,46).
Se e al s udies ha e demons a ed ha celas ol has cellula
a ge s in he con ex o RA, such as TAK1/IKK and MAPK/MEK
pa hways as well as MMP-9, STAT3, RANKL, and MD2/TLR4,
in e e ing wi h he p oduc ion o cy okines, chemokines,
and in lamma o y media o s; inhibi ing cell in asion and
p oli e a ion; and supp essing bone eso p ion (5,9,14,47).
In acco dance, Liu e al. ha e obse ed in a mouse model o
dexame hasone-induced seconda y os eopo osis ha celas ol
no only imp o es lipid me abolism and educes hype calciu ia
bu also mi iga es a icula ca ilage lesions, dec eases NF-kB,
MMP-1, and MMP-9 exp ession, and educes se um PTH,
TRACP-5b, CTX-I, as well as deoxypy idinoline (48). In 2012,
ou g oup showed ha celas ol dec eases he sec e ion o
bo h IL-1βand TNF in he THP-1 mac ophage-like cell line
associa ed no only wi h NF-kB inhibi ion bu also wi h caspase-1
inac i a ion (7).
Despi e i s p omising an i-a h i ic e ec s, celas ol has been
epo ed o induce weigh loss in mice models o cance (28,49–
51). A hese dosages o celas ol, no body weigh a ia ions
we e obse ed, sugges ing no majo oxicological e ec s induced
by celas ol o by he sol en s (e hanol and PEG400) used.
These wo agen s a e he mos commonly employed (52) and
we e iden i ied as he adequa e ones o dissol ing celas ol
(22). Howe e , some s udies ha e sugges ed ha e hanol and
PEG400 may be dele e ious o he gas oin es inal ac (53,
54), cons i u ing a isk ac o o oxicological side e ec s such
as body weigh loss. Since no a ia ions in body weigh we e
obse ed in celas ol- ea ed a s and due o he sho - e m
ea men du a ion o his s udy, oge he wi h he ac ha
we ha e used he minimum PEG and e hanol concen a ions
equi ed in o de o ge celas ol solubili y, we can exclude
his dele e ious side e ec in he gas oin es inal ac . No
clinical chemis y o his opa hological oxici y o his compound
has been p e iously shown. In pa allel wi h he assessmen
o he he apeu ic e ec o celas ol a di e en dose le els,
we also assessed i s possible oxic e ec s a he same doses
(1–7.5 µg/g/day). No changes in ALT and blood u ea we e
de ec ed, sugges ing ha i did no induce majo li e o
enal damage. Celas ol blocks he ion conduc ion o ca diac
F on ie s in Medicine | www. on ie sin.o g 9Sep embe 2020 | Volume 7 | A icle 455