cells
A icle
Pa ien s wi h Cholangioca cinoma P esen Speci ic
RNA P o iles in Se um and U ine Ex acellula
Vesicles Mi o ing he Tumo Exp ession:
No el Liquid Biopsy Bioma ke s
o Disease Diagnosis
Ainhoa Lapi z 1, Ande A belaiz 1, Colm J. O’Rou ke 2, Jose L. La in 3, Adelaida La Cas a 1,
Cesa Iba a 4, Juan P. Jimeno 5, Al a o San os-Laso 1, Lau a Izquie do-Sanchez 1,6,
Ma cin K awczyk 7,8 , Ma ia J. Pe ugo ia 1,6 , Raul Jimenez-Ague o 1,
Albe o Sanchez-Campos 4, Ioana Riaño 1, Espe anza Gónzalez 9, F ank Lamme 7,
Ma co Ma zioni 10, Rocio I.R. Macias 6,11 , Jose J.G. Ma in 6,11 , Tom H. Ka lsen 12,
Luis Bujanda 1,6, Juan M. Falcón-Pé ez 6,9,13 , Jespe B. Ande sen 2, Ana M. A ansay 3,6 ,
Ped o M. Rod igues 1,* and Jesus M. Banales 1,6,13,*
1Depa men o Li e and Gas oin es inal Diseases, Biodonos ia Heal h Resea ch Ins i u e,
Donos ia Uni e si y Hospi al, Uni e si y o he Basque Coun y (UPV/EHU), 20014 San Sebas ian, Spain;
[email p o ec ed] (A.L.); [email p o ec ed] (A.A.);
[email p o ec ed] (A.L.C.); [email p o ec ed]g (A.S.-L.);
[email p o ec ed]g (L.I.-S.); [email p o ec ed]g (M.J.P.);
[email p o ec ed] (R.J.-A.); [email p o ec ed] (I.R.);
[email p o ec ed] (L.B.)
2Depa men o Heal h and Medical Sciences, Bio ech Resea ch & Inno a ion Cen e (BRIC),
2200 Copenhagen, Denma k; [email p o ec ed] (C.J.O.); jespe [email p o ec ed] (J.B.A.)
3CIC bioGUNE, Genome Analysis Pla o m, 48160 De io, Spain; [email p o ec ed] (J.L.L.);
[email p o ec ed] (A.M.A.)
4Hospi al o C uces, 48903 Bilbao, Spain; cesa [email p o ec ed] (C.I.);
[email p o ec ed] (A.S.-C.)
5“Complejo Hospi ala io de Na a a”, 31008 Pamplona, Spain; [email p o ec ed]
6Ca los III Na ional Ins i u e o Heal h, Cen e o he S udy o Li e and Gas oin es inal
Diseases (CIBERehd), 28220 Mad id, Spain; [email p o ec ed]
7Depa men o Medicine II, Saa land Uni e si y Medical Cen e, Saa land Uni e si y,
66421 Hombu g, Ge many; Ma [email p o ec ed] (M.K.); [email p o ec ed] (F.L.)
8Depa men o Gene al, T ansplan and Li e Su ge y, Labo a o y o Me abolic Li e Diseases,
Cen e o P eclinical Resea ch, 02-091 Wa saw, Poland
9Cen e o Coope a i e Resea ch in Biosciences (CIC bioGUNE), Basque Resea ch and Technology
Alliance (BRTA), Exosomes Labo a o y, 48160 De io, Spain; [email p o ec ed]
10 Depa men o Gas oen e ology, “Uni e si àPoli ecnica delle Ma che”, 60121 Ancona, I aly;
m.ma zioni@s a .uni pm.i
11 Expe imen al Hepa ology and D ug Ta ge ing (HEVEFARM), Biomedical Resea ch Ins i u e
o Salamanca (IBSAL), 37007 Salamanca, Spain; [email p o ec ed] (R.I.R.M.); [email p o ec ed] (J.J.G.M.)
12 Di ision o Cance Medicine, Su ge y and T ansplan a ion, No wegian PSC Resea ch Cen e ,
Oslo Uni e si y Hospi al, 0372 Oslo, Spain; [email p o ec ed]
13 IKERBASQUE, Basque Founda ion o Science, 48013 Bilbao, Spain
*Co espondence: ped o. [email p o ec ed] (P.M.R.); [email p o ec ed]g (J.M.B.);
Tel.: +34-9-4300-6125 (P.M.R.); +34-9-4300-6067 (J.M.B.)
Recei ed: 17 Feb ua y 2020; Accep ed: 9 Ma ch 2020; Published: 14 Ma ch 2020
Abs ac :
Cholangioca cinoma (CCA) comp ises a g oup o he e ogeneous bilia y cance s wi h dismal
p ognosis. he e iologies o mos CCAs a e unknown, bu p ima y scle osing cholangi is (PSC) is a isk
Cells 2020,9, 721; doi:10.3390/cells9030721 www.mdpi.com/jou nal/cells
Cells 2020,9, 721 2 o 33
ac o . Non-in asi e diagnosis o CCA is challenging and accu a e bioma ke s a e lacking. We aimed
o cha ac e ize he ansc ip omic p o ile o se um and u ine ex acellula esicles (EVs) om pa ien s
wi h CCA, PSC, ulce a i e coli is (UC), and heal hy indi iduals. Se um and u ine EVs we e isola ed
by se ial ul acen i uga ions and cha ac e ized by nanopa icle acking analysis, ansmission
elec on mic oscopy, and immunoblo ing. EVs ansc ip ome was de e mined by Illumina gene
exp ession a ay [messenge RNAs (mRNA) and non-coding RNAs (ncRNAs)]. Di e en ial RNA
p o iles we e ound in se um and u ine EVs om pa ien s wi h CCA compa ed o con ol g oups
(disease and heal hy), showing high diagnos ic capaci y. he compa ison o he mRNA p o iles o
se um o u ine EVs om pa ien s wi h CCA wi h he ansc ip ome o umo issues om wo coho s
o pa ien s, CCA cells
in i o
, and CCA cells-de i ed EVs, iden i ied 105 and 39 commonly-al e ed
ansc ip s, espec i ely. Gene on ology analysis indica ed ha mos commonly-al e ed mRNAs
pa icipa e in ca cinogenic s eps. O e all, pa ien s wi h CCA p esen speci ic RNA p o iles in EVs
mi o ing he umo , and cons i u ing no el p omising liquid biopsy bioma ke s.
Keywo ds: bioma ke s; cholangioca cinoma; ex acellula esicles; liquid biopsy; ansc ip omics
1. In oduc ion
Cholangioca cinomas (CCAs) a e he e ogeneous bilia y malignancies cha ac e ized by dismal
p ognosis. he incidence and mo ali y a es o hese cance s a e apidly inc easing globally, cu en ly
accoun ing o ~15% o all p ima y li e cance s and ~3% o gas oin es inal malignancies [
1
–
3
].
Acco ding o hei ana omical localiza ion, CCAs a e classi ied in o in ahepa ic (iCCA), pe ihila
(pCCA), o dis al (dCCA). he e iology o mos CCAs is unknown. Howe e , se e al isk ac o s
ha e been desc ibed, including he p esence o p ima y scle osing cholangi is (PSC: 5–15% de elops
CCA), a ch onic choles a ic li e disease ha is associa ed wi h au oimmune phenomena agains
he in a- and ex ahepa ic bile duc s [
1
,
4
,
5
]. Impo an ly, 70–80% o pa ien s wi h PSC concomi an ly
p esen in lamma o y bowel disease (IBD), mainly ulce a i e coli is (UC), which is hough o p ecede
he de elopmen o he li e disease [4].
CCAs a e gene ally asymp oma ic in ea ly s ages, being he e o e commonly diagnosed in
ad anced phases when he disease is dissemina ed. La e diagnosis combined wi h he chemo esis an
na u e o hese umo s [
6
] highly comp omise he cu en he apeu ic op ions, mainly based on
su ge y, signi ican ly impac ing on pa ien ’s wel a e and ou come [
1
,
2
]. he diagnosis o CCA is
usually conduc ed by combining clinical, biochemical, adiological, and his ological in o ma ion.
Imaging echniques usually ely on compu ed omog aphy (CT), magne ic esonance imaging
(MRI), magne ic esonance cholangiopanc ea og aphy (MRCP), posi on emission omog aphy (PET),
pe cu aneous anshepa ic cholangiog aphy (PTC), endoscopic e og ade cholangiopanc ea og aphy
(ERCP), o endoscopic ul asound depending on he umo loca ion [
1
,
5
,
7
]. Howe e , imaging
p oceedings ha e impo an limi a ions, as hey a e no accu a e enough o de e mine he maligni y o
he umo masses, pa icula ly in ea ly s ages, as well as o di e en ia e be ween he main p ima y
li e cance s, i.e., iCCA, hepa ocellula ca cinoma (HCC) o HCC-CCA mixed umo s, which is
undamen al o p o ide he app op ia e s anda ds o ca e. On he o he hand, MRCP and his ological
analysis (biopsy o b ushing cy ology) comp ise he majo diagnos ic ools o PSC [
4
,
8
,
9
]. In addi ion,
he measu emen o non-speci ic se um umo bioma ke s [i.e., ca bohyd a e an igen 19-9 (CA19-9)
and ca cinoemb yonic an igen (CEA)] is commonly conduc ed in o de o help in he diagnosis o
CCA, bu hei low sensi i i y (pa icula ly in ea ly s ages o he disease) and speci ici y (also ele a ed
in some PSC pa ien s wi hou cance ), aise impo an conce ns ega ding hei clinical u ili y [
7
,
10
].
The e o e, umo biopsy is cu en ly manda o y o con i m he diagnosis and s aging o CCA, guiding
he clinical managemen o hese pa ien s [
11
]. Based on all o hese diagnos ic conce ns, he e is
Cells 2020,9, 721 3 o 33
an u gen need o de e mine new accu a e, non-in asi e bioma ke s o he ea ly diagnosis o CCA,
pa icula ly in pa ien s a isk.
In he las decade, ex acellula esicles (EVs) ha e been en isioned as p omising ools in he ques
o umo bioma ke s and as impo an media o s o disease pa hogenesis [
12
]. EVs cons i u e
a he e ogeneous popula ion o lipid bilaye ed sphe es (30 nm – 2
µ
m in diame e ) con aining di e se
biomolecules (e.g., p o eins, nucleic acids, lipids, and me aboli es), which a e eleased by cells and ound
in all bio luids (e.g., blood and u ine) [
13
,
14
]. Taking in o accoun hei biogenesis, EVs may be classi ied
as exosomes, mic o esicles (MV), and apop o ic bodies. These small esicles pa icipa e in cell- o-cell
communica ions, modula ing signaling pa hways in pa hobiology [
13
–
17
]. We p e iously epo ed
a di e en ial p o eomic p o ile o se um EVs om pa ien s wi h CCA, HCC, o PSC, as well as om
heal hy indi iduals, iden i ying accu a e candida e bioma ke s o he di e en ial diagnosis o hese
diseases [
18
]. Conside ing ha umo cells can also elease RNAs encapsula ed wi hin EVs, and ha
hei p o iles can mimic he cellula s a e/al e a ions, an ex ensi e cha ac e iza ion o he RNA con en
om se um and u ine EVs om pa ien s wi h CCA (and con ol condi ions) migh p o ide new
diagnos ic bioma ke s as well as he apeu ic a ge s.
In his s udy, we aimed o cha ac e ize he RNA p o ile o se um and u ine EVs om pa ien s
wi h CCA, PSC, o UC, as well as heal hy indi iduals, and iden i y candida e diagnos ic bioma ke s
mi o ing hei umo cell exp ession wi hin he liquid biopsy concep . Fo his pu pose, he exp ession
o selec ed candida es was e alua ed in human CCA umo and su ounding heal hy issues om wo
independen coho s o pa ien s [The Cance Genome A las (TCGA) and Copenhagen], as well as in
cell cul u es (CCA s. no mal) and EVs eleased by no mal o umo human cholangiocy es in i o.
2. Ma e ials and Me hods
2.1. Pa ien s
Se umandu inesamples ompa ien swi hCCA(n=12and23, espec i ely),PSC(n=6and5, espec i ely),
UC (n =8 and 12, espec i ely), and heal hy indi iduals (n =9 and 5, espec i ely) we e ob ained om Donos ia
Uni e si yHospi al(SanSebas ian,Spain),C ucesUni e si yHospi al(Bilbao,Spain),and“ComplejoHospi ala io
de Na a a” (Pamplona, Spain). he E hical Commi ees o Clinical Resea ch om each pa icipa ing ins i u ion
app o ed all he esea ch p o ocols and all pa ien s accep ed o pa icipa e in he s udy and signed he w i en
consen s o allow he use o hei samples o biomedical esea ch. Clinical cha ac e is ics o pa ien s and umo s
a e summa ized in Supplemen a y Table S1. he diagnosis o PSC was based on he Eu opean Associa ion
o he S udy o he Li e (EASL) guidelines [
19
] by demons a ing he p esence o bile duc al e a ions
(s ic u es o i egula i ies in in ahepa ic and ex ahepa ic bile duc s) using MRCP a e excluding seconda y
causes o cholangi is. he diagnosis o UC was pe o med by combining endoscopic and his ological s udies,
mainly colonoscopy in pa allel wi h hema oxylin and eosin (H&E) s aining, a e excluding o he po en ial
diseases. Finally, CCA diagnosis was con i med by his ological analysis o umo samples and/o h ough
he combina ion o clinical, biochemical and adiological app oaches. Tumo s age was de e mined based on
he 7 h edi ion o he Ame ican Join Commi ee on Cance (AJCC) classi ica ion.
RNA-seq da a om he TCGA coho (36 CCAs and 9 su ounding li e samples) [
20
],
downloaded as le el 3 da a h ough Fi eB owse po al [BROAD Ins i u e o MIT & Ha a d, MA, USA
(sou ce: h ps://gdac.b oadins i u e.o g/)], and whole ansc ip ome p o iling [Human T ansc ip ome
(HT) BeadChips (Illumina Inc., San Diego, CA, USA)] o he “Copenhagen coho ” including 217 CCA
su gical specimens (153 iCCA, 43 pCCA, 15 dCCA, 6 unknown loca ion), 143 no mal su ounding li e
samples, and 9 no mal in ahepa ic bile duc s (GSE26566) [
21
,
22
] we e used o e alua e he exp ession
o se um and u ine bioma ke s in umo issue.
2.2. Cell Cul u es
No mal human cholangiocy es (NHCs) we e isola ed om no mal li e issue and cha ac e ized
as p e iously desc ibed [
23
–
25
]. Fu he mo e, wo comme cial human CCA cell lines (EGI1 and TFK1,
Cells 2020,9, 721 4 o 33
Leibniz Ins i u e DSMZ-Ge man Collec ion o Mic ooganism and Cell Cul u es, Ge many) we e used.
NHC and EGI1 cells we e cul u ed in ully-supplemen ed DMEM/F-12 medium, as p e iously
desc ibed [
23
–
25
], while TFK1 cells we e cul u ed in DMEM/F-12 supplemen ed wi h 10% e al bo ine
se um (FBS; Gibco, The mo Fishe Scien i ic, Wal ham, MA, USA) and 1% penicillin/s ep omycin
(P/S; Gibco). Cells we e seeded in 150 mm collagen-coa ed issue cul u e dishes (4
×
10
6
cells)
wi h each espec i e cell cul u e medium and le o pla e a achmen o e nigh . A e wa ds,
cells we e washed wi h phospha e-bu e ed saline (PBS) and incuba ed wi h “EV ecollec ion media”
(DMEM/F-12+Glu amax supplemen ed wi h 1% P/S, and wi hou se um). A e 48 h, cells we e
ha es ed o RNA isola ion and cell cul u e media was collec ed and s o ed a
−
80
◦
C o subsequen
EVs isola ion. Cells we e g own a 37
◦
C in a humidi ied chambe o 5% CO
2
. Du ing all he expe imen s,
mycoplasm es was pe o med by con en ional PCR and cells we e es ed as mycoplasm nega i e.
2.3. Isola ion o EVs om Se um, U ine, and Cell Cul u es
Se um, u ine, and cell-de i ed EVs we e isola ed as p e iously desc ibed [
18
]. B ie ly,
1 mL o se um, 50 mL o u ine, o 300 mL o cell cul u e media ( ozen a
−
80
◦
C) we e hawed a
oom empe a u e and u he p ocessed h ough se ial di e en ial ul acen i uga ion s eps a 4
◦
C.
Fi s , in o de o emo e cell deb is, se um, u ine and cell cul u e media we e cen i uged a 10,000
×
g
o 30 min and subsequen ly ul acen i uged a 100,000
×
g o 75 min, o pelle he EVs, which we e
hen washed wi h PBS and pelle ed again a e ul acen i uga ion a 100,000
×
g o 75 min. Finally,
he pelle ed EV ac ion was esuspended in 20
µ
L o PBS and hen s o ed a
−
80
◦
C o u he analysis.
2.4. T ansmission Elec on Mic oscopy (TEM)
Fo he cha ac e iza ion o EVs, he isola ed ac ion o EVs was s ained nega i ely and analyzed
by TEM. EV samples we e di ec ly adso bed on o glow-discha ged (60 seg low discha ging using
a PELCO easy-glow de ice) ca bon-coa ed coppe g id (300 mesh). A e wa ds, g ids we e ixed
wi h 2% pa a o maldehyde (PFA) in phospha e bu e (PB 0.2M pH 7.4) o 20 min and washed wi h
dis illed wa e . Then, he con as s aining was made by incuba ing he g ids wi h 4% u anyl ace a e
(UA) a 4
◦
C o 15 min. TEM images we e ob ained by using TECNAI G2 20 C-TWIN high- esolu ion
ansmission elec on mic oscope, a an accele a ion ol age o 200 kV.
2.5. Immunoblo ing
P o ein le els o bo h EV and endoplasmic e iculum ma ke s (i.e., CD63 and CD81 s. GRP78,
espec i ely) we e e alua ed in se um and u ine EVs and in whole-cell ex ac s (WCEs) by immunoblo ing.
To al p o ein concen a ion was calcula ed wi h he Mic o BCA p o ein assay ki (The mo Fishe Scien i ic,),
ollowing he manu ac u e ’s ins uc ions. Loading bu e [50 mM T is-HCl, 2% SDS, 10% glyce ol and 0.1%
b omophenol blue, wi hou
β
-me cap oe hanol o di hio h ei ol (DTT)] was added o p o ein samples,
ollowed by hea dena u a ion a 95
◦
C o 5 min. Then, 10 and 4
µ
g o o al p o ein om se um and u ine
EVs, espec i ely, we e sepa a ed by 12.5% sodium dodecyl sul a e-polyac ilamide gel elec opho esis
(SDS-PAGE) and elec o- ans e ed on o a ni ocellulose memb ane (GE Heal hca e, Chicago, IL, USA)
and blocked wi h 5% skim milk powde / is-bu e ed saline (TBS)-0.1% ween (TBS-Tween) o 1 h.
A e wa ds, memb anes we e p obed o e nigh a 4
◦
C wi h he app op ia e p ima y an ibodies
[an i-CD81 (BD Biosciences), an i-CD63 (DSHB), and an i-GRP78 (BD Biosciences, San Jose, CA, USA)]
a 1:500 dilu ion in blocking solu ion and, a e h ee washes wi h TBS-Tween (5 min each), ho se adish
pe oxidase-conjuga ed seconda y an ibody (an i-mouse; Cell Signaling, Dan e s, MA, USA) a a dilu ion
o 1:5000 (in milk blocking solu ion) we e incuba ed o 1 h a oom empe a u e. Memb anes we e
de eloped o p o ein de ec ion using ECL plus (The mo Fishe Scien i ic), wi h he iB igh FL1500
Wes e n Blo Imaging Sys em (The mo Fishe Scien i ic).
Cells 2020,9, 721 5 o 33
2.6. EV Size and Concen a ion
Size dis ibu ion and concen a ion o EVs we e e alua ed by nanopa icle acking analysis
(NTA) using a NanoSigh LM10 Sys em (Mal e n, UK) u he equipped wi h as ideo cap u e
and a pa icle- acking so wa e. NTA pos -acquisi ion se ings we e kep cons an o all samples.
Each ideo was analyzed o ob aining he mean and mode esicle size as well as pa icle concen a ion.
2.7. To al RNA Isola ion
A e EVs isola ion, o al RNA was ex ac ed using he miRCURY
™
RNA Isola ion Ki
(Qiagen, Hilden, Ge many) ollowing manu ac u e ’s speci ica ions. A e wa ds, o al RNA was
esuspended in 20
µ
L o dis illed H
2
O and la e used o ansc ip omic analysis. Rega ding cell samples,
o al RNA was ex ac ed using he TRIzol
®
eagen acco ding o he manu ac u e ’s ins uc ions
(Li e Technologies Co p., Ca lsbad, CA, USA).
2.8. Illumina Gene Exp ession A ay
Illumina HumanHT-12 WG-DASL V4.0 R2 exp ession beadchips we e used o cha ac e ize gene
exp ession [messenge RNAs (mRNAs) and non-coding RNAs (ncRNAs)]. he quali y o RNA samples
was measu ed using a RNA Pico Chip Bioanalyze (Agilen Technologies, San a Cla a, CA, USA).
200 ng o RNA samples we e used o he a ay. he cDNA syn hesis, p equali ica ion, ampli ica ion,
labeling and hyb idiza ion o he samples we e pe o med ollowing he WG-DASL HT Assay Lab
p o ocol (Illumina Inc.). he ampli ied cDNAs we e hyb idized o he di e se gene-p obes o he a ay
and gene exp ession le els we e de ec ed by a HiScan scanne (Illumina Inc.). Raw da a we e ex ac ed
wi h GenomeS udio analysis so wa e (Illumina Inc.), in he o m o GenomeS udio’s Final Repo .
Raw exp ession da a we e backg ound-co ec ed, log
2
- ans o med and quan ile-no malized using
he lumi R package [
26
] (Bioconduc o eposi o y, Chicago, IL, USA). To pe o m he Venn diag ams,
all he ansc ip s iden i ied in a leas one sample wi h a “de ec ion p- alue” <0.01 we e selec ed.
A e wa ds, in he compa isons be ween g oups, ansc ip s ha we e signi ican ly iden i ied in a leas
20% o he samples (wi h a bila e al p- alue <0.05; independen samples wo- ailed - es , no assuming
equal a iances) we e conside ed o subsequen analysis.
2.9. Func ional En ichmen Analysis
Func ional analysis o candida e liquid biopsy RNA bioma ke s was de e mined by gene on ology
(GO) en ichmen o biological p ocesses, molecula pa hways and unc ions, by using he Func ional
En ichmen analysis ool (FunRich) e sion 3.1.3 (Fun ich Indus ial Co. L d, Hong Kong) [27,28].
2.10. S a is ical Analysis
S a is ical analysis was pe o med using G aphPad P ism e sion 6.0 (G aphPad So wa e,
San Diego, CA, USA). Da a a e shown as boxes and whiske s (min o max). When compa ing wo
g oups, non-pa ame ic Mann-Whi ney o pa ame ic -S uden es s we e conduc ed. Fo compa isons
be ween mo e han wo g oups, non-pa ame ic K uskal-Wallis es ollowed by a pos e io i Dunns
es o he pa ame ic one-way analysis o a iance (ANOVA) es ollowed by a pos e io i Tukey’s pos
hoc es we e used. In o de o calcula e he diagnos ic alues o se um and u ine RNA bioma ke s,
allowing o disc imina e be ween pa ien s wi h CCA, PSC and UC, and heal hy indi iduals, a ea unde
he ecei e ope a ing cha ac e is ic cu e (AUC) alues we e de e mined using he SPSS 20.0 so wa e
(IBM, Ehningen, Ge many), ollowed by he calcula ion o sensi i i y (SEN) and speci ici y (SPE)
alues, posi i e p edic i e alue (PPV), nega i e p edic i e alue (NPV), posi i e likelihood a io
(PLR), nega i e likelihood a io (NLR), and accu acy index (AI). Di e ences we e conside ed signi ican
when p<0.05.
Cells 2020,9, 721 6 o 33
3. Resul s
3.1. Cha ac e iza ion o Se um and U ine EVs om Pa ien s wi h CCA, PSC, o UC, and Heal hy Indi iduals
A e isola ion, se um and u ine EVs we e cha ac e ized by TEM, immunoblo ing and NTA.
In esemblance wi h ou p e ious indings using he same isola ion p o ocol [
18
], TEM images showed
a ypical ounded mo phology in he isola ed esicles om bo h se um and u ine (~100–200 nm),
co esponding o exosomes and/o small mic o esicles (Figu e 1A). By immunoblo ing, he EV p o ein
ma ke s CD63 and CD81 we e highly en iched in he isola ed EV ac ion, when compa ed o o al se um
o NHC whole-cell ex ac s (WCE), while he endoplasmic e iculum ma ke 78 kDa glucose- egula ed
p o ein (GRP78) was comple ely absen in isola ed se um and u ine EVs bu only ound exp essed
in WCE om NHCs (Figu e 1B), subs an ia ing a p ope isola ion and a high pu i y o he ob ained
EVs. Rega ding he size o EVs, NTA e ealed no signi ican di e ences in he size o se um and u ine
EVs among g oups, p esen ing an a e age size o ~180 nm, in esemblance wi h se um and u ine EV
concen a ion, which was ound simila in he s udy popula ion (Figu e 1C).
Figu e 1.
Cha ac e iza ion o se um and u ine EVs om pa ien s wi h CCA, PSC, UC, and heal hy
con ols. In o de o alida e he p o ocol o EVs isola ion, we used blood se um and u ine om heal hy
indi iduals. (
A
) TEM images o blood se um (le ) and u ine ( igh ) EVs om heal hy indi iduals
showcasing he ypical ound shape (~150 nm) and mo phology. (
B
) Rep esen a i e immunoblo s o
he EV ma ke s CD63 and CD81 (posi i e con ols) and GRP78 (nega i e con ol) om EVs isola ed
om se um (le ) and u ine ( igh ) o heal hy indi iduals ha indica e an en ichmen o EV ma ke s
and a comple e absence o he endoplasmic e iculum (ER) ma ke GRP78, compa ed o o al se um
and whole cell ex ac s (WCEs) o no mal human cholangiocy es (NHC). (
C
) Nanopa icle acking
analysis (NTA) o se um (up) and u ine (down) EVs e ealing no di e ences in EV concen a ion
be ween CCA, PSC, UC, and heal hy indi iduals and a simila EV mode (~180 nm).
Cells 2020,9, 721 7 o 33
3.2. Di e en ial RNA P o iles o Se um EVs om CCA, PSC, UC, and Heal hy Indi iduals
The ansc ip omic p o iles o se um and u ine EVs isola ed om pa ien s wi h CCA, PSC,
UC, and heal hy con ols we e de e mined by RNA mic oa ay-based ansc ip omics (Illumina Inc.).
T ansc ip omic da a a e a ailable in GSE144521. Conside ing all he ansc ip s ha we e iden i ied
in a leas one sample included in any o he s udy g oups (de ec ion p- alue <0.01), a o al o
25,084 ansc ip s we e iden i ied in se um EVs. Among hem, 10,104 ansc ip s we e iden i ied
in se um EVs isola ed om heal hy indi iduals, in pa allel wi h he iden i ica ion o 11,124, 4204,
and 24,264 ansc ip s in se um EVs om pa ien s wi h UC, PSC, and CCA, espec i ely, wi h 1617
o he iden i ied ansc ip s being sha ed among all g oups (Figu e 2A). In all he s udy g oups, he g ea
majo i y o he iden i ied ansc ip s we e mRNAs (9516, 10,526, 3949, and 23,029 ansc ip s ound
in heal hy indi iduals and pa ien s wi h UC, PSC, o CCA, espec i ely), ollowed by non-coding
RNAs such as non-coding RNAs (mainly including pseudogenes, long non-coding RNAs (lncRNAs),
among o he s), mic oRNAs (miRNAs o miRs), and small nucleola RNAs (snoRNAs) (Figu e 2B;
Supplemen a y Table S2). O he ypes o RNAs, including small nuclea , miscellaneous, guide,
small cy oplasmic, an isense, RNase MRP, ibosomal, and elome ase RNAs we e also de ec ed.
Nex , he ansc ip ome o se um EVs om he ou s udy g oups was de e mined and compa ed.
Speci ically, 1932 ansc ip s we e di e en ially iden i ied in CCA s. heal hy indi iduals, 2888 in CCA
s. PSC, and 2807 in CCA s. “PSC, UC, and heal hy indi iduals” combined as one unique con ol
(disease and heal hy) g oup (Figu e 3). Meanwhile, 866 ansc ip s we e di e en ially iden i ied in
se um EVs om pa ien s wi h PSC compa ed wi h a g oup comp ised o pa ien s wi h UC and heal hy
indi iduals (Supplemen a y Figu e S1).
The analysis o candida e RNA bioma ke s in se um EVs om CCA s. heal hy indi iduals poin ed
ou ing inge and FYVE like domain con aining E3 ubiqui in p o ein ligase (RFFL), ol ac o y ecep o
amily 4 sub amily F membe 3 (OR4F3), and he amily wi h sequence simila i y 107 membe B
(FAM107B) as he mRNAs wi h he highes diagnos ic capaci y, p esen ing AUC alues o 1.00,
1.00, and 0.991, espec i ely, along wi h he non-coding RNAs PMS1 homolog 2 misma ch epai
sys em componen pseudogene 4 (PMS2L4), miR-604, and SNORA58 (AUC: 0.991, 0.944, and 0.926,
espec i ely) (Figu e 3A). Since PSC is a well-known isk ac o ha inc eases he odds o de eloping
CCA, he ansc ip omic p o iles o se um EVs om pa ien s wi h CCA s. PSC we e also compa ed.
In pa icula , he mRNA ansc ip s pa aoxonase 1 (PON1), ac i a ing ansc ip ion ac o 4 (ATF4),
and phosphoglyce a e dehyd ogenase (PHGDH) s ood ou as he bes candida e bioma ke s o
he di e en ial diagnosis o CCA and PSC, all wi h AUC alues o 1.00 (Figu e 3B). Simila ly,
he lncRNAs me as asis associa ed lung adenoca cinoma ansc ip 1 (MALAT1) and LOC100190986,
and he snoRNA SNORA11B (AUCs: 1.00) also p esen ed a high accu acy o he iden i ica ion o CCA
s. PSC (Figu e 3B). O no e, se e al mRNA and non-coding RNAs p o ided excellen diagnos ic
alues (AUC alues up o 0.931 and 0.902, espec i ely) o he diagnosis o PSC, when compa ed wi h
pa ien s wi h UC and heal hy indi iduals (Supplemen a y Figu e S1). Finally, gene al CCA ansc ip
bioma ke s (compa ed o PSC, UC and heal hy indi iduals combined as one unique con ol g oup)
we e also iden i ied and RFFL, zinc inge p o ein 266 (ZNF266) and OR4F3 cons i u ed he mRNA
ansc ip s wi h he highes AUC alues (1.00, 0.976, and 0.960, espec i ely) while miR-551B, PMS2L4,
and LOC643955 we e he ncRNAs p esen ing he highes diagnos ic capaci y, displaying AUC alues
o 0.909, 0.880, and 0.873, espec i ely (Figu e 3C).
3.3. Selec i e mRNAs P esen in Se um EVs om Pa ien s wi h CCA Mi o Thei Le els in Human Tumo
Tissue, CCA Cells In Vi o and EVs-De i ed om Tumo Cholangiocy es
A e iden i ying 2807 ansc ip s signi ican ly al e ed in se um EVs om pa ien s wi h CCA
compa ed o pa ien s wi h PSC, UC, and heal hy indi iduals, we e alua ed i he exp ession o
hese ansc ip s we e also signi ican ly changed in human CCA issue, compa ed o non- umo
su ounding issue, in wo independen in e na ional coho s o pa ien s (TCGA and he “Copenhagen”
coho s). Impo an ly, 901 ou o he 2807 selec i e RNA ansc ip s we e also al e ed in he TCGA
Cells 2020,9, 721 8 o 33
coho , p esen ing he same end o exp ession when compa ed o se um EVs, wi h 765 ansc ip s
being up egula ed while 136 ansc ip s we e down egula ed in compa ison o non- umo issue
(Figu e 4A, le ). These 765 ansc ip s we e hen c oss- alida ed in he Copenhagen coho ,
in which we we e able o iden i y 479 sha ed ansc ip s wi h he same exp ession endency, wi h 391
being up egula ed and 88 educed when compa ed wi h su ounding li e issue (Figu e 4A, igh ).
A e selec ing he common mRNAs ha sha e he same end o exp ession in se um EVs and umo
issue o pa ien s wi h CCA compa ed o con ols, we nex e alua ed hei exp ession le els in wo
human CCA cell lines (EGI1 and TFK1) compa ed o NHCs
in i o
, ob aining 156 commonly al e ed
ansc ip s (Figu e 4B). Finally, we isola ed EVs om hese wo CCA cell lines and om NHCs and,
a e hei cha ac e iza ion (Supplemen a y Figu e S2) [
18
], we e alua ed hei ansc ip omic con en
and c oss- alida ed he p e ious candida e ansc ip s, esul ing in 105 mRNAs wi h sha ed al e ed
le els in se um EVs, umo issue, CCA cell lines, and in CCA-de i ed EVs (Figu e 4C).
Figu e 2.
Compa a i e ansc ip omic analysis o se um EVs om pa ien s wi h CCA, PSC, o UC,
and heal hy indi iduals. (
A
) Venn diag ams showing he numbe o ansc ip s iden i ied pe
g oup. Venn diag ams we e gene a ed using In e ac iVenn web-based ool [
29
]. (
B
) Numbe o
ansc ip s iden i ied wi hin each s udy g oup, subclassi ied acco ding o hei ype [messenge RNA
(mRNA), non-coding RNA (including mos ly pseudogenes and long non-coding RNAs, among o he s),
miscelaneous RNA (miscRNA), guide RNA, mic oRNA (miRNA), small nucleola RNA (snoRNA),
small nuclea RNA (snRNA), ibosomal RNA, elome ase RNA, small cy oplasmic RNA, an isense
RNA and RNase MRP RNA]. In all g oups, mRNAs cons i u e he mos abundan ly iden i ied RNAs,
ollowed by non-coding RNAs (pseudogenes, lncRNAs, and o he s), miRNAs, and snoRNAs.
Cells 2020,9, 721 9 o 33
Figu e 3. Con .
Cells 2020,9, 721 16 o 33
Figu e 5.
Selec ed liquid biopsy bioma ke s o CCA. F om he 105 mRNAs commonly al e ed in
se um EVs, CCA human umo s, CCA cells and in cell-de i ed EVs compa ed o hei co esponding
con ols, 5 bioma ke s we e selec ed based on hei diagnos ic capaci y. Box plo diag ams wi h
he mRNAs abundance in se um EVs (le ), CCA umo s om he TCGA, and “Copenhagen” coho s,
CCA cells and cell-de i ed EVs ( igh ), compa ed o hei espec i e con ols, o (
A
)c-Ma inducing
p o ein (CMIP), (
B
)glu ama e deca boxylase 1 (GAD1), and (
C
)NME/NM23 nucleoside diphospha e kinase
1 (NME1); (
D
)CDP-diacylglyce ol syn hase 1 (CDS1), and (
E
)CDC28 p o ein kinase egula o y subuni
1B (CKS1B). (
F
) Diagnos ic p edic ion (ROC cu es and AUC alues) o he selec ed se um liquid
biopsy bioma ke s and o he combina ion o CMIP, NME1 and CKS1B o he diagnosis o CCA
in compa ison wi h
(PSC +UC +Heal hy indi iduals).
Abb e ia ions: AUC, a ea unde he ecei e
ope a ing cha ac e is ic (ROC) cu e; EVs, ex acellula esicles; NHC, no mal human cholangiocy e;
NBD, no mal bile duc s; SL, su ounding li e ; TCGA, he cance genome a las.
Cells 2020,9, 721 17 o 33
Figu e 6.
Compa a i e ansc ip omic analysis o u ine EVs om pa ien s wi h CCA, PSC, o UC,
and heal hy indi iduals. (
A
) Venn diag ams showing he numbe o ansc ip s iden i ied pe
g oup. Venn diag ams we e gene a ed using In e ac iVenn web-based ool [
29
]. (
B
) Numbe o
ansc ip s iden i ied wi hin each s udy g oup, subclassi ied acco ding o hei ype [messenge RNA
(mRNA), non-coding RNA (including mos ly pseudogenes and long non-coding RNAs, among o he s),
miscelaneous RNA (miscRNA), guide RNA, mic oRNA (miRNA), small nucleola RNA (snoRNA),
small nuclea RNA (snRNA), ibosomal RNA, elome ase RNA, small cy oplasmic RNA, an isense
RNA and RNase MRP RNA]. In all g oups, mRNAs cons i u e he mos abundan ly iden i ied RNAs,
ollowed by non-coding RNAs (pseudogenes, lncRNAs, and o he s), miRNAs, and snoRNAs.
Cells 2020,9, 721 18 o 33
Figu e 7. Con .
Cells 2020,9, 721 19 o 33
Figu e 7. Con .
Cells 2020,9, 721 20 o 33
Figu e 7.
Di e en ial ansc ip omic p o ile o u ine EVs and diagnos ic capaci y. Volcano plo
(
−
log
10
(p- alue) and log
2
( old-change); up le ), hea map o he di e en ially exp essed ansc ip s
(up igh ) and diag ams wi h he diagnos ic capaci y wi h he highes AUC alues o he 10 selec ed
mRNAs and 5 selec ed non-coding RNAs in se um EVs om (
A
) CCA s. Heal hy indi iduals;
(
B
) CCA s. PSC; (
C
) CCA s. (PSC +UC +Heal hy indi iduals). Abb e ia ions: AI, accu acy index;
AUC, a ea unde he ecei e ope a ing cha ac e is ic cu e; CI, con idence in e al; miRNA, mic oRNA;
lncRNA, long non-coding RNA; miscRNA, miscellaneous RNA; NLR, nega i e likelihood a io;
NPV, nega i e p edic i e alue; PLR, posi i e likelihood a io; PPV, posi i e p edic i e alue;
SEN, sensi i i y; snRNA, small nuclea RNA; snoRNA, small nucleola RNA; SPE, speci ici y; RNA,
aul RNA.
Cells 2020,9, 721 21 o 33
3.5. Selec i e mRNAs P esen in U ine EVs om Pa ien s wi h CCA Mimic Thei Le els in Human Tumo
Tissue, CCA Cells In Vi o, and EVs-De i ed om Tumo Cholangiocy es
Simila o ou p e ious analysis on he se um EV bioma ke s iden i ied in pa ien s wi h
CCA, compa ed o all he o he s udy g oups, we now selec ed he 1329 RNA ansc ip s ha
we e signi ican ly al e ed in u ine EVs om pa ien s wi h CCA and compa ed hei exp ession
le els wi h CCA and su ounding umo issues om bo h TCGA and Copenhagen coho s.
A e pe o ming a comp ehensi e analysis o hese mRNAs in he TCGA coho , we we e able
o iden i y 390 dys egula ed ansc ip s (305 up egula ed and 85 down egula ed) ha a e commonly
al e ed in u ine EVs and in umo samples om pa ien s wi h CCA (Figu e 8A, le ). Addi ionally,
compa ed o he changes obse ed in u ine EVs, 259 ansc ip s also sha ed he same pa e n o
al e a ion in he Copenhagen coho , wi h 206 ansc ip s p esen ing inc eased exp ession while
53 ansc ip s we e educed, when compa ed wi h su ounding li e (Figu e 8A, igh ). he compa ison
o hese 259 ansc ip s wi h he di e en ial ansc ip ome o CCA cell lines compa ed o NHCs,
e ealed 84 sha ed mRNAs ha we e al e ed in CCA cells, wi h 69 being up egula ed and 15 displaying
dec eased exp ession (Figu e 8B). In EVs isola ed om CCA and NHC cell cul u es, we we e able
o iden i y 39 mRNAs (34 up egula ed and 5 down egula ed) commonly al e ed wi h u ine EVs,
umo issue, and CCA cells (Figu e 8C).
Figu e 8. Con .
Cells 2020,9, 721 22 o 33
Figu e 8.
mRNAs commonly de egula ed be ween u ine EVs, CCA umo s om wo independen
coho s o pa ien s, umo cells
in i o
and in CCA-de i ed EVs. mRNAs di e en ially abundan
in se um EVs om pa ien s wi h CCA s. (PSC +UC +Heal hy indi iduals) we e compa ed
wi h he ansc ip ome o : (i) pa ien s wi h CCA om he Cance Genome A las TCGA
(n =36)
and “Copenhagen”
(n =217)
coho s, (ii) CCA cells (EGI1 and TFK1) and cell-de i ed EVs compa ed
o hei espec i e con ol g oups, u he selec ing he ones ha a e commonly exp essed. Hea map
o he di e en ially exp essed ansc ip s in (
A
) TCGA (le ) and “Copenhagen” coho s ( igh );
(
B
) Whole-cell ex ac s om CCA cells and no mal human cholangiocy es (NHCs); (
C
) Cell-de i ed
EVs; (
D
) Gene on ology (GO: FunRich da abase [
27
]) analysis o he 105 ansc ip s commonly al e ed in
se um EVs, CCA human umo s, CCA cells and in cell-de i ed EVs, highligh ing he biological p ocesses
and pa hways in which he iden i ied ansc ip s a e in ol ed, as well as hei biological unc ion.
Abb e ia ions: EVs, ex acellula esicles; NHC, no mal human cholangiocy e; SL, su ounding li e ;
TCGA, he cance genome a las.
Cells 2020,9, 721 23 o 33
In o de o e alua e he ole o hese ansc ip s in ca cinogenesis, we conduc ed a GO analysis
and obse ed ha , in esemblance wi h wha we p e iously ound in se um EVs, hese mRNA ansc ip s
code o p o eins ha a e p edominan ly ela ed wi h umo de elopmen and p og ession, namely
me abolic pa hways (nucleic acids and p o ein me abolism), signal ansduc ion, cell communica ion,
EMT and immune esponse. Al hough less ep esen ed, some ansc ip s we e also linked o
ene gy and cell g ow h/main enance pa hways (Figu e 8D). In his ega d, he ansc ip s ubiqui in
conjuga ing enzyme E2 C (UBE2C) and se ine p o ease inhibi o B1 (SERPINB1) a ose as po en ial liquid
biopsy bioma ke s, being inc eased in u ine EVs isola ed om pa ien s wi h CCA in compa ison
o a g oup con aining pa ien s wi h PSC, UC and heal hy con ols (Figu e 9A,B). Combining hese
u ine bioma ke s in o one panel inc eased hei diagnos ic accu acy, p o iding an AUC alue o
0.812 o he diagnosis o CCA (Figu e 9C). No ewo hy, he exp ession le els o hese ansc ip s
we e also ma kedly up egula ed in CCA umo samples om he TCGA and Copenhagen coho s,
when compa ed wi h bo h no mal su ounding li e specimens and/o no mal in ahepa ic bile
duc s, p esen ing also inc eased exp ession in CCA cells and in CCA-de i ed EVs, when compa ed
wi h NHCs (Figu e 9A,B). Impo an ly, SERPINB1 mRNA le els inc eased wi h disease se e i y in
he Copenhagen coho , being pa icula ly o e exp essed in ad anced umo s ages compa ed wi h
ea ly s age CCAs (Supplemen a y Figu e S5A). Fu he mo e, al hough no being p esen ed as one o
he bes liquid biopsy candida e, Tc ex1 domain con aining 2 (TCTEX1D2) le els we e ound up egula ed
in poo ly-di e en ia ed umo s compa ed wi h well-di e en ia ed ones (Supplemen a y Figu e S5B).
Figu e 9.
Po en ial u ine liquid biopsy ma ke s o CCA. F om he 39 ansc ip s commonly ound
in se um EVs and di e en ially exp essed in pa ien samples, CCA cells and in cell line-de i ed
EVs, 2 po en ial u ine liquid biopsy ma ke s wi h he bes diagnos ic capaci y we e selec ed.
Box plo diag ams wi h he mRNA ansc ip abundance in u ine EVs (le ) and he exp ession
in he TCGA and “Copenhagen” coho s, cholangiocy e cell lines and cell line-de i ed EVs ( igh )
o (
A
)Ubiqui in conjuga in enzyme E2 C (UBE2C) and (
B
)Se ine p o einase inhibi o B1 (SERPINB1).
(
C
) Diagnos ic p edic ion (ROC cu es and AUC alues) o he selec ed u ine liquid biopsy ma ke s
and om he combina ion o UBE2C and SERPINB1 o he diagnosis o CCA in compa ison
wi h
(PSC +UC +Heal hy indi iduals).
Abb e ia ions: AUC, a ea unde he ecei e ope a ing
cha ac e is ic cu e; EVs, ex acellula esicles; NHC, no mal human cholangiocy e; NBD, no mal bile
duc s; SL, su ounding li e ; TCGA, he cance genome a las.
Cells 2020,9, 721 24 o 33
4. Discussion
In he las decade, a conside able e o has been made o iden i y no el non-in asi e bioma ke s
o he ea ly and accu a e diagnosis o CCA [
7
]. He e, we epo o he i s ime he di e en ial
RNA p o ile o se um and u ine EVs om pa ien s wi h CCA, PSC, o UC, and heal hy indi iduals,
iden i ying new po en ial bioma ke s wi h high diagnos ic capaci y. No ewo hy, some o he al e ed
mRNAs we e simila ly changed in CCA umo s om wo independen coho s o pa ien s, and in
umo cells and CCA-de i ed EVs
in i o
, highligh ing hei u ili y as liquid biopsy bioma ke s as
well as hei po en ial alue as a ge s o he apy.
High- h oughpu omic app oaches ha e been o g ea help in o de o ind po en ial new candida e
bioma ke s. In ac , he iden i ica ion o he p o eomic con en o EVs, as well as ce ain ci cula ing
p o eins, in bio luids ha e al eady p o ided candida e bioma ke s in bile, se um, and u ine om
pa ien s wi h CCA [
18
,
30
–
32
]. We ha e ecen ly desc ibed he di e en ial p o eomic p o iles o
se um EVs om pa ien s wi h CCA, compa ed o HCC, PSC, and heal hy indi iduals, epo ing new
po en ial p o ein bioma ke s wi h high diagnos ic capaci y [
18
] ha mus be in e na ionally alida ed
by ELISA echnology. Ne e heless, a ull ansc ip omic analysis in dis inc body luids (in pa icula
EVs) om hese pa ien s has ne e been conduc ed and migh esul in he iden i ica ion o no el,
accu a e bioma ke s o he diagnosis o CCA. Al hough p o eins a e usually mo e s able han mRNAs,
hei p esence wi hin EVs p o ides hem p o ec ion om deg ada ion; mo eo e , RNAs a e usually
easie o de ec and quan i y, e en when ound a e y low le els, which may help in hei as e
ansla ion in o he clinic [
33
]. In ac , ci cula ing small ncRNAs a e ound in all bio luids (including
se um and u ine), mainly due o hei ema kable esis ance o RNase deg ada ion. High ci cula ing
RNase le els con ibu e o a low abundance o o he ypes o RNAs (mRNAs) and signi ican ly
comp omise hei easy and eliable de ec ion [
34
]. S ill, speci ic RNAs can be eleased om cance
cells in o bio luids, allowing hei iden i ica ion and u he de e mina ion o hei po en ial alue as
bioma ke s, as hey may mi o he cellula s a e wi hin he concep o liquid biopsy. Fo ins ance, some
RNA ansc ip s we e al eady e idenced and ound inc eased in plasma, as is he case o elome ase
e e se ansc ip ase (hTERT) ha showed diagnos ic and p ognos ic alue o p os a e cance , being
a good p edic o o ecu ence [
35
,
36
]. Simila ly, he le els o he long non-coding RNA p os a e cance
associa ed 3 (PCA3) we e abundan ly ound in u ine o pa ien s wi h p os a e cance , cons i u ing
a p omising non-in asi e bioma ke o he diagnosis o ha cance [
37
,
38
]. he le els o se e al
miRNAs we e also epo ed al e ed in se um, plasma, and u ine o pa ien s wi h gas oin es inal
cance s, including CCA, cons i u ing also po en ial no el bioma ke s o cance diagnosis [
7
,
33
,
39
–
41
].
Taking ad an age om ou p e iously epo ed EV isola ion p o ocol [
18
], we ha e he e also
se led up he p o ocol o he isola ion o u ine EVs. By analyzing he ansc ip omic p o ile o
se um and u ine EVs om pa ien s wi h CCA, he p esen s udy was pionee in iden i ying no el
po en ial RNA bioma ke s wi h high diagnos ic capaci y o CCA. No ewo hy, some o hese new
RNA bioma ke s we e also signi ican ly al e ed in CCA issue om he wo in e na ional coho s
o pa ien s and u he dis u bed in CCA cell lines and in EVs sec e ed om hese umo cell lines,
when compa ed wi h NHCs. Consequen ly, hese bioma ke s a e mi o ing wha is happening in
he umo issue, since he dis u bances obse ed in umo biopsies (and CCA cell lines) ha a e la e
sec e ed in EVs and eleased in o he bloods eam a e amenable o de ec ion ei he in se um o u ine
(Figu e 10). This cons i u es a no el and inno a i e liquid biopsy app oach whe e we a e able o de ec
speci ic al e a ions ha a e obse ed in umo issue wi hou ob aining umo samples, ha bo ing
a high diagnos ic alue. In he u u e, e alua ing he ele ance o hese bioma ke s in p edic ing
p ognosis and in guiding he apeu ic decisions is en isioned.
Cells 2020,9, 721 25 o 33
Figu e 10.
No el liquid biopsy app oach o cholangioca cinoma. CCA umo cells display dis inc RNA
exp ession p o iles which a e la e eleased in o ci cula ion in EVs, con aining po en ial bioma ke s
o CCA, ha a e amenable o de ec ion in se um and u ine, hus cons i u ing a no el liquid
biopsy app oach.
Taking in o conside a ion he 105 po en ial liquid biopsy bioma ke s ha we e iden i ied in se um,
we he ein epo ed he bes i e se um bioma ke s ha display excellen diagnos ic accu acy: CMIP,
GAD1,NME1,CDS1, and CKS1B. Impo an ly, hese no el po en ial bioma ke s migh cons i u e be e
CCA bioma ke s han CA19-9 since he AUC alues ha we he ein ob ained (up o 0.891) a e highe
han he diagnos ic capaci y epo ed in a sys ema ic e iew and a me a-analysis, in which he AUC
alue o CA19-9 was 0.830 [
42
]. Despi e he po en ial diagnos ic alue o hese ansc ip s, conside ing
ha hey a e concomi an ly inc eased in umo issue in wo in e na ional independen coho s,
we pos ula e ha hey also migh play a pi o al pa hological ole du ing cholangioca cinogenesis. S ill,
no s udies ha e cu en ly add essed he in ol emen o hese biomolecules in CCA, al hough se e al
wo ks al eady alued hei ole in o he ypes o cance . Fo ins ance, he ansc ip ion ac o CMIP
was p e iously shown o be inc eased in human gas ic cance and glioma umo s, con ibu ing o
umo p oli e a ion and me as asis [
43
,
44
]. Fu he mo e, high CMIP le els we e associa ed wi h wo se
p ognosis ( ecu ence- ee and o e all su i al) in gas ic and b eas cance s [
43
,
45
] and we e ela ed
wi h he cep in esis ance in HER2-posi i e gas ic cance cells [
46
]. Simila ly, he enzyme encoded
by GAD1 gene, which ca alyzes he con e sion o L-glu amic acid o
γ
-aminobu y ic acid, has been
ound o e exp essed in se e al ypes o umo s, including lung adenoca cinoma [
47
], nasopha yngeal
ca cinoma [
48
], o al squamous cell ca cinoma [
49
], p os a e cance [
50
], and b ain me as asis [
51
],
being also ound up egula ed in colon and HCC cells
in i o
[
52
]. In pa allel, high GAD1 le els
we e also shown o co ela e wi h he pa hological s age o pa ien s wi h lung adenoca cinoma,
posi i ely co ela ing wi h me as asis and wi h wo se ecu ence- ee su i al [
47
]. Rega ding NME1,
i s ele ance in cance is s ill con o e sial. a me a-analysis e alua ed he p ognos ic alue o NME1
in pa ien s wi h diges i e sys em neoplasms (including pa ien s wi h HCC and gallbladde cance ,
bu no CCA) and epo ed ha high NME1 le els a e co ela ed wi h well-di e en ia ed umo s
and wi h less-se e e cance s ages, wi h no e iden co ela ion wi h p ognosis [
53
]. In ag eemen
wi h he epo ed me as asis-supp essing ole o NME1, pa ien s wi h HCC p esen ing lowe p o ein
le els o NME1 displayed inc eased me as a iza ion [
54
]. In pa ien s wi h HCC [
55
] and in wo animal
models o HCC [
56
], NME1 exp ession was ound up egula ed in compa ison wi h non- umo issue,
Cells 2020,9, 721 32 o 33
66.
Shi, L.; Wang, S.; Zanga i, M.; Xu, H.; Cao, T.M.; Xu, C.; Wu, Y.; Xiao, F.; Liu, Y.; Yang, Y.; e al. O e -exp ession
o CKS1B ac i a es bo h MEK/ERK and JAK/STAT3 signaling pa hways and p omo es myeloma cell
d ug- esis ance. Onco a ge 2010,1, 22–33. [C ossRe ]
67.
Huang, C.W.; Lin, C.Y.; Huang, H.Y.; Liu, H.W.; Chen, Y.J.; Shih, D.F.; Chen, H.Y.; Juan, C.C.; Ke , C.G.;
Huang, C.Y.; e al. CKS1B o e exp ession implica es clinical agg essi eness o hepa ocellula ca cinomas
bu no p27(Kip1) p o ein u no e : an independen p ognos ica o wi h po en ial p27 (Kip1)-independen
oncogenic a ibu es? Ann. Su g. Oncol. 2010,17, 907–922. [C ossRe ]
68.
Zhang, J.; Liu, X.; Yu, G.; Liu, L.; Wang, J.; Chen, X.; Bian, Y.; Ji, Y.; Zhou, X.; Chen, Y.; e al. UBE2C is
a po en ial bioma ke o in es inal- ype gas ic cance wi h ch omosomal ins abili y. F on . Pha macol.
2018
,
9, 847. [C ossRe ]
69.
Li, J.; Zhi, X.; Shen, X.; Chen, C.; Yuan, L.; Dong, X.; Zhu, C.; Yao, L.; Chen, M. Deple ion o UBE2C educes
o a ian cance malignancy and e e ses cispla in esis ance ia down egula ing CDK1. Biochem. Biophys.
Res. Commun. 2020,523, 434–440. [C ossRe ]
70.
Jin, Z.; Zhao, X.; Cui, L.; Xu, X.; Zhao, Y.; Younai, F.; Messadi, D.; Hu, S. UBE2C p omo es he p og ession o
head and neck squamous cell ca cinoma. Biochem. Biophys. Res. Commun. 2020,523, 389–397. [C ossRe ]
71.
Wang, X.; Yin, L.; Yang, L.; Zheng, Y.; Liu, S.; Yang, J.; Cui, H.; Wang, H. Silencing ubiqui in-conjuga ing
enzyme 2C inhibi s p oli e a ion and epi helial-mesenchymal ansi ion in panc ea ic duc al adenoca cinoma.
FEBS J. 2019,286, 4889–4909. [C ossRe ]
72.
Wei, Z.; Liu, Y.; Qiao, S.; Li, X.; Li, Q.; Zhao, J.; Hu, J.; Wei, Z.; Shan, A.; Sun, X.; e al. Iden i ica ion o
he po en ial he apeu ic a ge gene UBE2C in human hepa ocellula ca cinoma: an in es iga ion based on
GEO and TCGA da abases. Oncol. Le . 2019,17, 5409–5418. [C ossRe ]
73.
Xiong, Y.; Lu, J.; Fang, Q.; Lu, Y.; Xie, C.; Wu, H.; Yin, Z. UBE2C unc ions as a po en ial oncogene by
enhancing cell p oli e a ion, mig a ion, in asion, and d ug esis ance in hepa ocellula ca cinoma cells.
Biosci. Rep. 2019,39. [C ossRe ]
74.
Wu, Y.; Jin, D.; Wang, X.; Du, J.; Di, W.; An, J.; Shao, C.; Guo, J. UBE2C Induces Cispla in Resis ance ia
ZEB1/2-Dependen Up egula ion o ABCG2 and ERCC1 in NSCLC Cells. J. Oncol.
2019
,2019, 8607859.
[C ossRe ]
75.
Le man, I.; Ma, X.; Sege , C.; Maolake, A.; Ga cia-He nandez, M.L.; Rangel-Mo eno, J.;
Acke man, J.; Nas iuk, K.L.; Susia jo, M.; Hammes, S.R. Epigene ic supp ession o SERPINB1 p omo es
in lamma ion-media ed p os a e cance p og ession. Mol. Cance Res. 2019,17, 845–859. [C ossRe ]
76.
Huasong, G.; Zongmei, D.; Jian eng, H.; Xiaojun, Q.; Jun, G.; Sun, G.; Donglin, W.; Jianhong, Z. Se ine
p o ease inhibi o (SERPIN) B1 supp esses cell mig a ion and in asion in glioma cells. B ain Res.
2015
,
1600, 59–69. [C ossRe ] [PubMed]
77.
Cui, X.; Liu, Y.; Wan, C.; Lu, C.; Cai, J.; He, S.; Ni, T.; Zhu, J.; Wei, L.; Zhang, Y.; e al. Dec eased exp ession o
SERPINB1 co ela es wi h umo in asion and poo p ognosis in hepa ocellula ca cinoma. J. Mol. His ol.
2014,45, 59–68. [C ossRe ]
78.
Tseng, M.Y.; Liu, S.Y.; Chen, H.R.; Wu, Y.J.; Chiu, C.C.; Chan, P.T.; Chiang, W.F.; Liu, Y.C.; Lu, C.Y.; Jou, Y.S.;
e al. Se ine p o ease inhibi o (SERPIN) B1 p omo es o al cance cell mo ili y and is o e -exp essed in
in asi e o al squamous cell ca cinoma. O al. Oncol. 2009,45, 771–776. [C ossRe ] [PubMed]
79.
Willmes, C.; Kuma , R.; Becke , J.C.; F ied, I.; Rachakonda, P.S.; Poppe, L.M.; Hesbache , S.; Schadendo , D.;
Sucke , A.; Sch ama, D.; e al. SERPINB1 exp essionis p edic i e o sensi i i y and ou come o cispla in-based
chemo he apy in melanoma. Onco a ge 2016,7, 10117–10132. [C ossRe ] [PubMed]
80. Razumila a, N.; Go es, G.J. Cholangioca cinoma. Lance 2014,383, 2168–2179. [C ossRe ]
81.
Hou, Z.H.; Xu, X.W.; Fu, X.Y.; Zhou, L.D.; Liu, S.P.; Tan, D.M. Long non-coding RNA MALAT1 p omo es
angiogenesis and immunosupp essi e p ope ies o HCC cells by sponging miR-140. Am. J. Physiol.
Cell Physiol. 2019,318, C649–C663. [C ossRe ]
82.
Sonoha a, F.; Inokawa, Y.; Hayashi, M.; Yamada, S.; Sugimo o, H.; Fujii, T.; Kode a, Y.; Nomo o, S. P ognos ic
alue o long non-coding RNA HULC and MALAT1 ollowing he cu a i e esec ion o hepa ocellula
ca cinoma. Sci. Rep. 2017,7, 16142. [C ossRe ] [PubMed]
83.
Hou, Z.; Xu, X.; Zhou, L.; Fu, X.; Tao, S.; Zhou, J.; Tan, D.; Liu, S. he long non-coding RNA MALAT1
p omo es he mig a ion and in asion o hepa ocellula ca cinoma by sponging miR-204 and eleasing SIRT1.
Tumou Biol. 2017,39, 1010428317718135. [C ossRe ] [PubMed]
Cells 2020,9, 721 33 o 33
84.
Konishi, H.; Ichikawa, D.; Yamamo o, Y.; A i a, T.; Shoda, K.; Hi amo o, H.; Hamada, J.; I oh, H.; Fuji a, Y.;
Koma su, S.; e al. Plasma le el o me as asis-associa ed lung adenoca cinoma ansc ip 1 is associa ed
wi h li e damage and p edic s de elopmen o hepa ocellula ca cinoma. Cance Sci.
2016
,107, 149–154.
[C ossRe ] [PubMed]
85.
Dai, X.; Chen, C.; Xue, J.; Xiao, T.; Mos o a, G.; Wang, D.; Chen, X.; Xu, H.; Sun, Q.; Li, J.; e al. Exosomal
MALAT1 de i ed om hepa ic cells is in ol ed in he ac i a ion o hepa ic s ella e cells ia miRNA-26b in
ib osis induced by a seni e. Toxicol. Le . 2019,316, 73–84. [C ossRe ] [PubMed]
86.
Tian, W.; Du, Y.; Ma, Y.; Gu, L.; Zhou, J.; Deng, D. MALAT1-miR663a nega i e eedback loop in colon cance
cell unc ions h ough di ec miRNA-lncRNA binding. Cell Dea h Dis. 2018,9, 857. [C ossRe ]
87.
Cheong, J.Y.; Shin, H.D.; Cho, S.W.; Kim, Y.J. Associa ion o polymo phism in mic oRNA 604 wi h
suscep ibili y o pe sis en hepa i is B i us in ec ion and de elopmen o hepa ocellula ca cinoma. J. Ko ean
Med. Sci. 2014,29, 1523–1527. [C ossRe ]
88.
Bai, S.Y.; Ji, R.; Wei, H.; Guo, Q.H.; Yuan, H.; Chen, Z.F.; Wang, Y.P.; Liu, Z.; Yang, X.Y.; Zhou, Y.N. Se um
miR-551b-3p is a po en ial diagnos ic bioma ke o gas ic cance . Tu k. J. Gas oen e ol.
2019
,30, 415–419.
[C ossRe ]
89.
Zhang, Y.; Yan, L.; Han, W. Ele a ed le el o miR-551b-5p is associa ed wi h in lamma ion and disease
p og ession in pa ien s wi h se e e acu e panc ea i is. The . Aphe . Dial. 2018,22, 649–655. [C ossRe ]
90.
Song, G.; Zhang, H.; Chen, C.; Gong, L.; Chen, B.; Zhao, S.; Shi, J.; Xu, J.; Ye, Z. miR-551b egula es
epi helial-mesenchymal ansi ion and me as asis o gas ic cance by inhibi ing ERBB4 exp ession. Onco a ge
2017,8, 45725–45735. [C ossRe ]
©
2020 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access
a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion
(CC BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).
