ORIGINAL RESEARCH
published: 30 May 2022
doi: 10.3389/ neu .2022.844841
F on ie s in Neu ology | www. on ie sin.o g 1May 2022 | Volume 13 | A icle 844841
Edi ed by:
Jean-cha les Sanchez,
Uni e si é de Genè e, Swi ze land
Re iewed by:
Amie Hille ,
O egon Heal h and Science Uni e si y,
Uni ed S a es
Anas asia Bougea,
Na ional and Kapodis ian Uni e si y
o A hens, G eece
Yuhu Zhang,
Guangdong P o incial People’s
Hospi al, China
*Co espondence:
Jo ge Sáiz
[email p o ec ed]
Albe o Be ga eche
jesusalbe o.be ga echeya za@
osakide za.eus
Special y sec ion:
This a icle was submi ed o
Neu ological Bioma ke s,
a sec ion o he jou nal
F on ie s in Neu ology
Recei ed: 28 Decembe 2021
Accep ed: 19 Ap il 2022
Published: 30 May 2022
Ci a ion:
Konje od M, Sáiz J, Ba bas C,
Be ga eche A, A danaz E, Hue a JM,
Vinag e-A agón A, E o ME,
Chi laque MD, Abillei a E, Iba luzea JM
and Amiano P (2022) A Se o Reliable
Samples o he S udy o Bioma ke s
o he Ea ly Diagnosis o Pa kinson’s
Disease. F on . Neu ol. 13:844841.
doi: 10.3389/ neu .2022.844841
A Se o Reliable Samples o he
S udy o Bioma ke s o he Ea ly
Diagnosis o Pa kinson’s Disease
Ma cela Konje od1,2, Jo ge Sáiz2*, Co al Ba bas2, Albe o Be ga eche3,4,5*,
E a A danaz6,7,8, José MaHue a8,9, Ana Vinag e-A agón3, MaElena E o7,10,
MaDolo es Chi laque8,9, Euna e Abillei a11,12, Jesús MaIba luzea8,13,14,15 and
Pila Amiano8,11,12
1Di ision o Molecula Medicine, Rudje Bosko ic Ins i u e, Zag eb, C oa ia, 2Facul ad de Fa macia, Cen o de Me abolómica
y Bioanálisis, Uni e sidad San Pablo-CEU, CEU Uni e si ies, Mad id, Spain, 3Depa men o Neu ology, Uni e si y Hospi al
Donos ia, San Sebas ián, Spain, 4Neu oscience A ea, Biodonos ia Heal h Resea ch Ins i u e, San Sebas ián, Spain,
5Biomedical Resea ch Ne wo king Cen e Conso ium o he A ea o Neu odegene a i e Diseases (CIBERNED), Ins i u o de
Salud Ca los III, Mad id, Spain, 6Na a a Public Heal h Ins i u e, Pamplona, Spain, 7Na a a Ins i u e o Heal h Resea ch
(IdiSNA), Pamplona, Spain, 8Spanish Conso ium o Resea ch on Epidemiology and Public Heal h (CIBERESP), Ins i u o de
Salud Ca los III, Mad id, Spain, 9Ins i u o Mu ciano de In es igación Biosani a ia, Mu cia, Spain, 10 Depa men o Neu ology,
Na a a Hospi al Complex, Pamplona, Spain, 11 Minis y o Heal h o he Basque Go e nmen , Public Heal h Labo a o y in
Gipuzkoa, San Sebas ián, Spain, 12 Epidemiology o Ch onic and Comunnicable Diseases A ea, Biodonos ia Heal h Resea ch
Ins i u e, San Sebas ián, Spain, 13 Minis y o Heal h o he Basque Go e nmen , Sub Di ec o a e o Public Heal h and
Addic ions o Gipuzkoa, San Sebas ián, Spain, 14 En i onmen al Epidemiology and Child De elopmen A ea, Biodonos ia
Heal h Resea ch Ins i u e, San Sebas ián, Spain, 15 Facul y o Psychology, Uni e si y o he Basque Coun y UPV/EHU, San
Sebas ian, Spain
Backg ound: Pa kinson’s disease (PD) is a p og essi e neu odegene a i e diso de ,
diagnosed acco ding o he clinical c i e ia ha occu in al eady ad anced s ages o
PD. The de ini ion o bioma ke s o he ea ly diagnosis o PD ep esen s a challenge ha
migh imp o e ea men and a oid complica ions in his disease. The e o e, we p opose
a se o eliable samples o he iden i ica ion o al e ed me aboli es o ind po en ial
p ognos ic bioma ke s o ea ly PD.
Me hods: This case–con ol s udy included plasma samples o 12 pa ien s wi h PD and
21 con ol subjec s, om he Spanish Eu opean P ospec i e In es iga ion in o Cance
and Nu i ion (EPIC)-Na a a coho , pa o he EPIC-Spain s udy. All he case samples
we e p o ided by heal hy olun ee s who we e ollowed-up o 15.9 (±4.1) yea s and
de eloped PD disease la e on, a e he sample collec ion. Liquid ch oma og aphy
coupled o andem mass spec ome y was used o he analysis o samples.
Resul s: Ou o 40 ha we e selec ed and s udied due o hei in ol emen in
es ablished cases o PD, se en signi ican ly di e en me aboli es be ween PD cases
and heal hy con ol subjec s we e ob ained in his s udy (benzoic acid, palmi ic acid,
oleic acid, s ea ic acid, myo-inosi ol, so bi ol, and quinolinic acid). These me aboli es
a e ela ed o mi ochond ial dys unc ion, he oxida i e s ess, and he mechanisms o
ene gy p oduc ion.
Conclusion: We p opose he samples om he EPIC s udy as eliable and in aluable
samples o he sea ch o ea ly bioma ke s o PD. Likewise, his s udy migh also be
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
a s a ing poin in he es ablishmen o a well- ounded panel o me aboli es ha can be
used o he ea ly de ec ion o his disease.
Keywo ds: Pa kinson’s disease, bioma ke s, me abolomics, a y acids, EPIC-Na a a
INTRODUCTION
Me abolomics is an impo an well-es ablished ool, able o
p o ide use ul insigh s o unknown biochemical mechanisms
and possible bioma ke s o a ious diso de s (1). Unde s anding
al e ed me abolic pa hways and me aboli es p o ides a be e
knowledge o unde lying biological al e a ions. This in o ma ion
migh imp o e he ea men s a egies and de ine no el
he apeu ic app oaches, as well as acili a e disease p edic ion
and diagnosis. The me abolic p o iles o neu odegene a i e and
neu opsychia ic diso de s and he ela ed me abolic pa hways
a e s ill unclea (2,3), as i is he case o Pa kinson’s disease
(PD). Howe e , some s udies ha e shown an associa ion be ween
ce ain me aboli es and se e al me abolic pa hways in PD (2–4).
Al e a ions in he yp ophan and kynu enine me abolism ha e
been associa ed wi h he appea ance o psychia ic symp oms
and he de elopmen o PD. Ce ain me aboli es, as pa o
his me abolic pa hway, showed dec eased le els in se e al
biological luids, such as yp ophan (4,5), kynu enic acid
(KA) (5,6), and quinolinic acid (QA) (5,6), while kynu enine
(3), 5-hyd oxy yp ophan (3), xan hu enic acid (3), and 3-
hyd oxykynu enine (7) showed an ele a ion in PD. Due o
he signi icance o his me abolic pa hway, i is assumed ha
dis up ion o yp ophan and kynu enine me abolism migh
lead o neu o oxici y associa ed o PD (8). Dopamine and
no epineph ine me abolism play an impo an ole in PD
de elopmen and p og ession. I is known ha dopamine gic loss
in he subs an ia nig a is one o he bigges hallma ks o PD
(4,5), while me aboli es o he ica boxylic acid cycle (TCA)
migh be associa ed wi h dopamine gic loss due o mi ochond ial
dys unc ion and al e a ions in ene gy p oduc ion (5). Se e al
me aboli es associa ed o hese me abolic pa hways ha e been
obse ed o be al e ed in PD (5–9). Suga s and hei de i a i es
(4–11) showed inc eased le els, while amino acids and hei
de i a i es we e al e ed in subjec s wi h PD (4,5,12,13).
Changes in b anched-chain amino acids migh also implica e
changes in p o ein syn hesis, mi ochond ial biogenesis, and
au ophagy (4). These indings la gely indica e he in ol emen
o mi ochond ial dys unc ion in PD pa hogenesis. Fu he mo e,
me aboli es in ol ed in a y acid and lipid me abolisms,
such as a y acids, glyce ophospholipids, ca ni ines, and bile
acids, showed al e ed le els in subjec s wi h PD, implying on
he po en ial in ol emen o in lamma ion, inc eased a e o
oxida i e s ess, impai ed b ain me abolism, and mi ochond ial
dys unc ion in PD pa hogenesis (2,4,14–17). Fa y acids mos ly
show educed le els in pa ien s wi h PD, possibly due o hei
ulne abili y o he oxida i e s ess ha causes lipid pe oxida ion
and s uc u al damage o a y acids (18). Among o he a y
acids, s ea ic acid (4,5), oleic acid (4,5), and palmi ic acid (4,5)
showed signi ican changes in subjec s wi h PD, compa ed o
con ol subjec s. Mo eo e , i has been epo ed ha u ic acid,
oge he wi h o he me aboli es o pu ine me abolism, has a
p o ec i e ole agains cell dea h and damage caused by oxida i e
s ess (4,5,19). I is obse ed ha people wi h lowe le els o
u ic acid in he b ain, se um, o plasma ha e highe isks o
de elop PD. The e o e, low le els o u ic acid migh be a po en ial
bioma ke o he ea ly diagnosis o PD (19–21). In addi ion,
o he compounds, including alcohols, hyd oxy acids, and amines
(5,9,22,23), ha e also been al e ed in pa ien s wi h PD. Howe e ,
he mechanisms ha lead o hese changes a e s ill unclea .
I is no ewo hy ha hese me abolic pa hways ha e been
ela ed o PD in subjec s wi h an al eady es ablished diso de ,
meaning ha hese me aboli es ha e been ound in pa ien s
who had al eady de eloped PD a he ime when hose s udies
we e ca ied ou . Unsu p isingly, o PD as well as o o he
neu odegene a i e diso de s, he ea ly diagnosis o he diso de
is he g ea es challenge. The main poin is o de ec he diso de
be o e he neu odegene a ion s a s o p oceed wi h adequa e
ea ly ea men . In his ega d, he scien is s ha e no been able
o ind a se o me aboli es ha a e al e ed be o e he disease is
es ablished. The complica ion o inding bioma ke s o de ec ing
disease be o e i eme ges is o ind eliable samples o be used. The
s udies published so a a e mainly ocused on bioma ke s o he
ea ly s ages o PD (24–26). The di icul y o hese s udies elies on
he iden i ica ion o pa ien s who ha e ecen ly de eloped PD, in
he ea lies possible s age.
The plasma samples used in his s udy we e ob ained om he
Eu opean P ospec i e In es iga ion in o Cance and Nu i ion
(EPIC), a p ospec i e, mul icen e , coho s udy. This s udy
ocuses on in es iga ing he e ec s o se e al ac o s and he
incidence o cance and o he ch onic diseases, mo e han hal a
million pa icipan s om 23 di e en cen e s and 10 Eu opean
coun ies we e ec ui ed and ollowed-up o almos 15 yea s.
I is possible o s udy he bank sample o ind hose dono s
ha we e no diagnosed o a pa icula disease a he ime o
sample collec ion and de eloped he disease a e wa d du ing
he moni o ing pe iod o his s udy. We used a subse o
hese samples om heal hy pa icipan s (no diagnosed o PD
o showing any PD- ela ed symp oms) a baseline ha we e
esea ched by expe epidemiologis s o iden i y dono s who
had no a diagnosed PD and did no show any PD- ela ed
symp oms a he ime o sample collec ion, bu who de eloped
PD la e on. These samples a e, he e o e, o an ex ao dina y
alue o s udying bioma ke s ha a e al e ed be o e PD shows
any symp oms and he neu odegene a ion begins. Coun ing on
hose samples, we conside ed ha in es iga ing hose me aboli es
ha a e known o be al e ed when PD is es ablished can be a good
s a ing poin o inding bioma ke s o he ea ly diagnosis o
he disease.
F on ie s in Neu ology | www. on ie sin.o g 2May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
TABLE 1 | Co esponding sys ems, columns, and mobile phases o analy ical me hods de eloped in his s udy.
Ion-pai ing HILIC A HILIC B RPLC A RPLC B
Sys em 1200 In ini y
6460 QqQ
1260 In ini y II
6470 QqQ
1260 In ini y II
6470 QqQ
1260 In ini y II
6470 QqQ
1260 In ini y II
6470 QqQ
Column Zo bax Ex ended
C18 (2.1 ×
150 mm, 1.8 µm)
XB idgeBEH
Amide 2.5 mic on
(2.1 ×100 mm,
2.5 µm)
XB idgeBEH
Amide 2.5 mic on
(2.1 ×100 mm,
2.5 µm)
Zo bax Eclipse,
XDB, C18 (4.6 ×
150 mm, 5 µm)
Zo bax Eclipse,
XDB, C18 (4.6 ×
150 mm, 5 µm)
Mobile phase A 97% wa e and
3% me hanol,
10 mM TBA,
15 mM ace ic acid
0.1% Fo mic acid
p epa ed in wa e ,
pH 9 (NH3)
5 mM Ammonium
o ma e p epa ed
in wa e
0.1% Fo mic acid
p epa ed in wa e
0.5% Fo mic acid
p epa ed in wa e
Mobile phase B 10 mM TBA,
15 mM ace ic acid
in me hanol
0.1% Fo mic acid
p epa ed in ACN
Ace oni ile 0.1% Fo mic acid
p epa ed in
me hanol
0.5% Fo mic acid
p epa ed in
me hanol
The e o e, his a icle aims o s udy he EPIC samples using
liquid ch oma og aphy and mass spec ome y me hodologies
o inding me aboli es ha can be used as ea ly bioma ke s o
PD. I is based on he abili y o he EPIC samples o e eal
ea ly bioma ke s o PD, conside ing he unique na u e o
hose samples.
MATERIALS AND METHODS
Subjec Rec ui men
This s udy was conduc ed a he Cen e o Me abolomics
and Bioanalysis (CEMBIO) in Mad id, Spain. This case–con ol
s udy included plasma samples om one Spanish coho
(EPIC-Na a a), pa o he EPIC s udy. Biological samples,
including plasma, se um, e y h ocy es, and leukocy es, we e
collec ed and s o ed in liquid ni ogen. The exclusion c i e ia
included physically o men ally incapabili y o pa icipa e in
his s udy, as well as p egnancy and lac a ion o women. All
he pa icipan s we e ee o cance a he ime o diagnosis.
De ailed collec ed da a abou die a y in ake and li es yle o
pa icipan s ha e been published p e iously. Sociodemog aphic
cha ac e is ics and an h opome ic measu es we e collec ed
om all he pa icipan s using s anda d p ocedu es (27). This
s udy included 12 case plasma samples om subjec s who
had de eloped Pa kinson’s disease om he sampling ime o
June 2011 and 21 co esponding con ol samples. Bo h he
g oups consis ed o 18.2% women, aged be ween 46 and 63
yea s old, and 81.8% men aged be ween 51 and 64 yea s
old. Inciden PD cases we e asce ained by eco d linkage
wi h heal h da abases o iden i y po en ial cases, ollowed by
indi idual e ision o he medical his o y by expe neu ologis s
in o de o es ablish he diagnosis based on he a ailable clinical
eco ds (28). Diagnosis o Pa kinson’s disease was es ablished o
pa icipan s ul illing a leas wo o he ollowing c i e ia: (1)
p ima y ca e eco ds using ei he codes 332 o he In e na ional
Classi ica ion o Diseases-9 (ICD-9) o codes N87 o he
In e na ional Classi ica ion o P ima y Ca e; (2) egis a ion o
p esc ip ions, including subjec s wi h a leas one p esc ip ion
o any o he N04-an ipa kinsonian d ugs o he Ana omical
The apeu ic Chemical/De ined Daily Dose (ATC/DDD) index
(N04-an ipa kinsonian d ugs, N04A-an icholine gic agen s, and
N04B-dopamine gic agen s); (3) mo ali y eco d using codes
332 o he ICD-9 o PD; (4) he minimum basic da a se
(CMBD) using codes 332 o he ICD-9 o he PD; and (5)
dea h ce i ica es wi h he G20 code o he ICD-10. Likewise,
ully desc ibed p ocedu es used o asce aining PD cases in he
EPIC s udy a e p e iously published by Gallo e al. (28). The
diagnosis o each pa icipan was based on a combina ion o
wo a iables, including he deg ee o neu ologis ’s expe ise and
con idence in e alua ing he da a and quali y/amoun o he da a
(29). Acco ding o he a o emen ioned c i e ia, diagnoses we e
classi ied as “de ini e” ( he highes deg ee o con idence and da a
quali y), “ e y likely” (high deg ee o con idence, good/low da a
quali y), “p obable” (mode a e deg ee o con idence and g ea
da a quali y), and “possible” (29). The Na a a cen e e i ied all
he po en ial causes.
Sample P epa a ion
The s aws con aining he plasma o each sample we e
emo ed om he eeze (−80◦C) and slowly hawed on ice.
Subsequen ly, hey we e opened and ans e ed o 500 µl
Eppendo ubes, which we e kep cons an ly on ice. They
we e o exed o 2 min. A o al o 100 µl o plasma we e
ans e ed o an Eppendo ube and 300 µl o a cold
mix u e (−20◦C) o me hanol:e hanol (1:1, / ) we e added
o dep o einiza ion. A e s i ing he samples o 1 min, hey
we e incuba ed on ice o 5 min and o exed o ano he
1 min. Samples we e cen i uged o 20 min a 13,200 pm
and a 4◦C. Finally, 100 µl o supe na an we e ans e ed
o a high-pe o mance liquid ch oma og aphy (HPLC) ial
o analysis.
P epa a ion o Blanks and Calib a ion
Samples
Blank samples we e p epa ed in he same way as he plasma
samples. A o al o 300 µl o me hanol:e hanol (1:1, / )
we e added o 500 µl Eppendo ubes wi h 100 µl o
Milli-Q wa e . The p o ocol con inued wi h cen i uga ion
a 13,200 pm o 20 min. The supe na an was ans e ed
o he HPLC ial o analysis. Calib a ion samples we e
F on ie s in Neu ology | www. on ie sin.o g 3May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
TABLE 2 | Lis o s udied me aboli es and hei e en ion imes and ansi ions in he assigned ch oma og aphic me hod.
Me abolic pa hway/class Me aboli es RT T ansi ions Me hods
Amino acids and de i a i es C ea inine 1.193 112.0 ⇒41.2 Ion-pai ing
Py oglu amic acid 6.937 128.1 ⇒84.0
Benzoic acids and de i a i es Benzoic acid 15.113 121.1 ⇒77.0
Bile acids Deoxycholic acid 20.894 391.6 ⇒345.2
Pu ine me abolism Hypoxan hine 1.950 135.0 ⇒92.0
Xan hine 2.523 151.0 ⇒108.0
Inosine 4.623 267.1 ⇒135.0
Guanosine 4.896 282.1 ⇒150.0
Fa y acid and dica boxylic acid me abolism Palmi ic acid 22.446 255.2 ⇒256.3
Oleic acid 22.559 281.5 ⇒282.3
S ea ic acid 23.252 283.5 ⇒283.5
Sube ic acid 14.951 173.1 ⇒111.1
Me hylmalonic acid 12.097 117.1 ⇒71.0
E hylmalonic acid 13.059 131.1 ⇒87.1
Suga s and de i a i es Myoinosi ol 1.294 239.1 ⇒179.0
TCA cycle Succinic acid 12.097 117.0 ⇒73.1
Malic acid 12.806 133.0 ⇒115.0
T yp ophan and kynu enine me abolism T yp ophan 7.513 203.1 ⇒116.0
Kynu enic acid 14.589 188.2 ⇒144.0
Amino acids and de i a i es Valine 12.068 118.2 ⇒55.1 HILIC A
Alanine 10.207 90.1 ⇒44.1
Amines Me hylhis amine 11.941 126.2 ⇒109.0
T ime hylamine 4.409 60.1 ⇒44.2
Pu ine me abolism U ic acid 8.763 169.0 ⇒141.0
Amino acids and de i a i es D-me hionine 3.298 150.0 ⇒61.1 HILIC B
Se ine 4,155 106.1 ⇒60.1
Th eonine 3.893 120.1 ⇒74.1
T yp ophan and kynu enine me abolism 3-hyd oxykynu enine 3.439 225.2 ⇒208.0
Alcohols and polyols P opylene glycol 3.375 77.1 ⇒59.1 RPLC A
Dopamine and no epineph ine me abolism Dopamine 3.482 154.2 ⇒137.0
3,4-dihyd oxyphenylace ic acid 5.345 167.1 ⇒123.1
TCA cycle Py u ic acid 2.625 87.0 ⇒43.2
a-ke oisocap oic acid 5.339 129.1 ⇒85.0
Suga s and o he s Th eonic acid 2.314 135.1 ⇒75.0
Gamma bu y olac ones Dehyd oasco bic acid 6.229 175.1 ⇒88.1 RPLC B
Suga s and o he s Galac i ol 2.227 183.2 ⇒129.0
So bi ol 2.227 183.2 ⇒129.0
D-Gluconic acid 2.249 195.1 ⇒129.0
T yp ophan and kynu enine me abolism Quinolinic acid 3.807 168.1 ⇒78.0
Kynu enine 5.312 209.2 ⇒192.0
RT, e en ion ime exp essed in minu es.
p epa ed om 1,000 ppm s ock solu ions in me hanol:e hanol
(1:1, / ).
Analy ical Se up
This s udy was conduc ed using 5 di e en liquid
ch oma og aphy-mass spec ome y (LC-MS/MS) me hods
acco ding o he de ec abili y o he analy es (see
Table 1 and Sec ion “Ch oma og aphic Me hod” in he
Supplemen a y Ma e ial). B ie ly, one ion-pai ing me hod,
wo hyd ophilic in e ac ion liquid ch oma og aphy (HILIC)
me hods, and wo Re e sed Phase Liquid Ch oma og aphy
(RPLC) me hods we e used combined wi h andem mass
spec ome y in a iple quad upole. These me hods
we e pa ially alida ed acco ding o he in o ma ion
p o ided in he Sec ion “Me hods Valida ion” in he
Supplemen a y Ma e ial. The analy ical pe o mance o
he me hods has been e alua ed by s udying he linea i y,
he epea abili y, he in e media e p ecision, and he
sensi i i y in e ms o limi o de ec ion (LOD) and limi o
quan i a ion (LOQ).
F on ie s in Neu ology | www. on ie sin.o g 4May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
Me aboli es and Me hods
Acco ding o he in o ma ion p o ided in he in oduc ion,
he ollowing analy es we e included in his s udy. Due o
he chemical di e si y o hese me aboli es, he analy es
we e dis ibu ed in he 5 analy ical me hods abo e-
desc ibed based on hei de ec abili y and selec i i y,
as shown in Table 2.
Da a T ea men and S a is ical Analysis
A e he ch oma og am inspec ion, he ob ained da a we e
ea ed wi h he MassHun e Quan i a i e Analysis so wa e
(Agilen MassHun e Quan i a i e Analysis e sion 10.0) o
he de e mina ion o he a ea o each peak. Mic oso O ice
Excel was used o quan i a ion and blank sub ac ion. The
p- alues ( he - es o he Wilcoxon/Mann–Whi ney U es ,
Mic oso O ice Excel, and SPSS, espec i ely) we e calcula ed
o each me aboli e. Log2FC was calcula ed acco ding o he
ollowing o mula:
log2FC =log2(a e age CASES/a e age CONTROLS)
Mul i a ia e s a is ics we e also pe o med in his s udy.
Supe ised o hogonal pa ial leas squa e-disc iminan analysis
(OPLS-DA) was pe o med. Volcano plo s plo ing a iable
impo ance in p ojec ion (VIP) in he OPLS-DA model
agains co ec ed p- alues [p(co ), loading alues scaled as
co ela ion coe icien s alues] we e gene a ed. Va iables wi h
absolu e p(co ) lowe han 0.3 show a low co ela ion, while
a iables be ween 0.3 and 0.5 show an in e media e co ela ion.
Me aboli es wi h p- alues <0.05, VIP sco e >1, and p(co ) ≥
0.3 we e conside ed signi ican .
The ecei e ope a ing cha ac e is ic (ROC) cu es and he
a ea unde he cu e (AUC) o he s udied me aboli es we e
ob ained in Me aboanalys e sion 5.0 (30).
RESULTS
Demog aphic Da a
This s udy included 12 case plasma samples om subjec s
who had de eloped Pa kinson’s disease and 21 co esponding
con ol samples. Ou o 12 PD subjec s, 83.3% we e male
and 16.7% we e emale, while in he con ol g oup, 80.95%
we e male and 19.05% we e emale (Table 3). The e was
no di e ence in body mass index (BMI) (p=0.615),
age (p=0.782), gende (p=0.865), and smoking s a us
(p=0.632) be ween pa ien s wi h PD and heal hy con ol
subjec s (Table 3). Educa ion in o ma ion showed ha 51.5%
o o al subjec s had p ima y school comple ed, 33.3% had
none, 6.1% had longe educa ion, i.e., uni e si y deg ee, 6.1%
had echnical/p o essional school, and 3% did no speci y hei
educa ion s a us.
Ta ge ed Me abolomic Analysis
Ou o 40 analy es ha we e analyzed, se en signi ican
me aboli es we e obse ed. Benzoic acid, palmi ic acid, oleic
acid, s ea ic acid, myo-inosi ol, so bi ol, and quinolinic acid
TABLE 3 | Demog aphic cha ac e is ics o heal hy con ol subjec s and pa ien s
wi h Pa kinson’s disease (PD) in ol ed in his s udy.
PD pa ien s
N=12
Con ol subjec s
N=21
S a is ics
Male (%) 10 (83.33) 17 (80.95) χ2=0.029;
d =1; p=0.865
Female (%) 2 (16.66) 4 (19.05)
Age (yea s)
[median (25 h;
75 h)]
60.50 (55; 62.50) 60.00 (54.50;
62.50)
U=118.5;
p=0.782
BMI (kg/m2) (mean
±SD)
28.45 ±3.74 29.16 ±3.91 = −0.508;
p=0.615
Smoke s (%) 5 (41.7) 7 (33.3) χ2=0.229; d =
1; p=0.632
Non-smoke s (%) 7 (58.3) 14 (66.7)
we e signi ican ly changed in subjec s ha la e de eloped
PD, compa ed o con ol subjec s. While a y acids, myo-
inosi ol, and so bi ol we e signi ican ly dec eased, benzoic
acid and quinolinic acid we e signi ican ly inc eased in PD
subjec s. Despi e he assump ion ha dec eased le els o u ic
acid ep esen a isk o PD de elopmen (21), in his s udy,
he di e ence in u ic acid le els be ween subjec s ha la e
de eloped PD and heal hy con ol subjec s was no obse ed
(Table 4).
The ROC cu es o signi ican me aboli es wi h he AUC >
0.7 (palmi ic acid, oleic acid, s ea ic acid, myo-inosi ol, so bi ol,
and quinolinic acid) a e shown in Figu e 1. OPLS-DA and
olcano plo s o se en signi ican compounds a e gi en in
Figu e 2.
The analy ical pa ame e s o signi ican compounds a e
shown in Table 4. All he signi ican compounds showed
good epea abili y and in e media e p ecision, wi h a
coe icien o a ia ion >0.99 (Table 5). The analy ical
pe o mance o he i e ch oma og aphic me hods is
shown in he Supplemen a y Ma e ial in he Sec ion
“Me hod Valida ion.”
DISCUSSION
Pa kinson’s disease is a complex, he e ogeneous
neu odegene a i e diso de wi h an expec ed ise in p e alence
up o 9 million in 2030 (31). Thus, an inc easing numbe
o pa ien s wi h PD migh cause a high inancial and social
bu den (32). Clinical diagnosis o PD is usually es ablished
when i s pa kinsonian symp oms appea and when he e
is al eady a signi ican dopamine gic loss (10). The e o e,
due o he lack o bioma ke s o ea ly diagnosis o PD,
a ge ing me aboli es ha could be in ol ed in PD de elopmen
and p og ession migh imp o e he apeu ic e iciency and
p o ide a be e unde s anding o he unde lying molecula
mechanisms ha lead o PD de elopmen (31), as well as
imp o ing he quali y li e o pa ien s and elie e p essu e on
medical se ices.
F on ie s in Neu ology | www. on ie sin.o g 5May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
TABLE 4 | Lis o analyzed me aboli es oge he wi h hei p- alues, log2FC alues, a iable impo ance in p ojec ion (VIP), p(co ) sco es, co ec ed p- alues, and he
a ea unde he cu e (AUC) sco es.
Me aboli es p- alue log2FC VIP p(co ) q AUC
Amino acids and de i a i es
D-me hionine 0.3653 0.10 0.28 0.09 0.6958 0.606
Se ine 0.6553 0.03 0.89 0.26 0.8191 0.577
Th eonine 0.3898 −0.08 0.63 0.22 0.6779 0.614
Valine 0.3725 −0.06 0.40 0.13 0.6773 0.571
Alanine 0.3450 −0.19 0.90 0.35 0.6900 0.613
C ea inine 0.0619 0.21 1.34 0.31 0.2251 0.688
Py oglu amic acid 0.2814 −0.13 0.71 0.28 0.6621 0.651
T yp ophan and kynu enine me abolism
T yp ophan 0.9767 0.01 0.11 0.05 1.0017 0.503
Kynu enic acid 0.9877 −0.01 0.68 0.12 0.9877 0.505
3-hyd oxykynu enine 0.0789 0.67 1.28 0.38 0.2630 0.701
Quinolinic acid 0.0440 0.38 1.62 0.34 0.2514 0.728
Kynu enine 0.1335 0.15 1.43 0.33 0.3814 0.690
Benzoic acids and de i a i es
Benzoic acid 0.0385 0.31 1.30 0.34 0.2567 0.693
Bile acids
Deoxycholic acid 0.3085 0.24 0.32 0.03 0.6856 0.619
Pu ine me abolism
U ic acid 0.5773 −0.10 0.81 0.25 0.8553 0.536
Hypoxan hine 0.6508 −0.71 0.72 0.28 0.8397 0.553
Xan hine 0.9366 −0.01 0.36 0.01 1.0704 0.548
Inosine 0.4830 −1.53 0.76 0.26 0.8050 0.582
Guanosine 0.7943 −0.08 0.31 0.00 0.9345 0.503
Fa y acid and dica boxylic acid me abolism
Palmi ic acid 0.0013 −1.31 1.91 0.68 0.0520 0.893
Oleic acid 0.0036 −1.25 1.86 0.68 0.0720 0.854
S ea ic acid 0.0076 −1.02 1.62 0.62 0.0760 0.792
Sube ic acid 0.0593 0.32 1.03 0.28 0.2636 0.720
Me hylmalonic acid 0.2769 0.04 1.30 0.28 0.6923 0.627
E hylmalonic acid 0.5404 −0.10 0.95 0.32 0.8314 0.571
Suga s and de i a i es
Galac i ol 0.0570 −0.90 1.21 0.29 0.2850 0.722
So bi ol 0.0040 −2.20 1.13 0.42 0.0533 0.836
D-Gluconic acid 0.5855 −0.10 0.67 0.06 0.8364 0.545
Th eonic acid 0.7249 0.07 0.60 0.09 0.8787 0.582
Myo-inosi ol 0.0211 −0.27 1.45 0.52 0.1688 0.765
TCA cycle
Py u ic acid 0.4958 −0.37 0.40 0.02 0.7933 0.614
a-ke oisocap oic acid 0.6346 0.03 0.76 0.02 0.8461 0.556
Succinic acid 0.6186 0.03 0.95 0.11 0.8532 0.558
Malic acid 0.0936 0.11 1.08 0.14 0.2880 0.656
Amines
Me hylhis amine 0.9611 −0.03 0.15 0.06 1.0390 0.506
T ime hylamine 0.3191 −0.18 1.17 0.45 0.6718 0.610
Alcohols and polyols
P opylene glycol 0.0607 −0.49 1.50 0.40 0.2428 0.683
Dopamine and no epineph ine me abolism
Dopamine 0.9609 −0.06 0.66 0.21 1.0677 0.505
(Con inued)
F on ie s in Neu ology | www. on ie sin.o g 6May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
TABLE 4 | Con inued
Me aboli es p- alue log2FC VIP p(co ) q AUC
3,4-dihyd oxyphenylace ic
acid
0.2210 −0.23 0.78 0.10 0.5893 0.643
Gamma bu y olac ones
Dehyd oasco bic acid 0.9639 0.06 0.33 0.10 1.0146 0.505
p- alue - le el o signi ican ob ained wi h - es o Mann-Whi ney U es ; log2FC - was calcula ed acco ding o he ollowing o mula: log2FC - log2(a e ageCASES/a e ageCONTROLS);
VIP, a iable impo ance in he p ojec ion in OPLS-DA model; p(co ), loading alues scaled as co ela ion coe icien s alues in OPLS-DA model; q, Benjamini-Hochbe g (FDR, alse
disco e y a e) co ec ion; AUC, a ea unde he cu e.
Signi ican me aboli es ound in his s udy we e bolded.
FIGURE 1 | The ecei e ope a ing cha ac e is ic (ROC) cu es o signi ican ly di e en me aboli es be ween Pa kinson’s disease (PD) cases and con ol subjec s, wi h
he a ea unde he cu e (AUC) >0.7.
Al e ed Me aboli es
Ou o he 40 s udied compounds ela ed o PD, se en s a is ically
signi ican (p<0.05) me aboli es ha e been obse ed in PD
subjec s compa ed o heal hy con ol subjec s. Benzoic acid,
palmi ic acid, oleic acid, s ea ic acid, myo-inosi ol, so bi ol, and
quinolinic acid we e signi ican ly changed in PD subjec s. Al e ed
le els o suga alcohols migh indica e al e a ions in he suga
me abolism. I is assumed ha inc eased glucose le els could
o e come glycolysis capaci y, which causes con e sion o glucose
o so bi ol. Ano he al e ed me aboli e ha was signi ican ly
al e ed is myo-inosi ol, which also plays an impo an ole in
suga me abolism. I is known ha al e ed le els o myo-inosi ol,
oge he wi h al e ed le els o so bi ol, migh be associa ed wi h
changes in glucose me abolism and glycolysis (9). Al e a ions o
glycolysis and suga me abolism imply on po en ial in ol emen
o me abolic pa hways ha pa icipa e in he ene gy p oduc ion
in pa hogenesis o PD (11). Besides, malabso p ion o so bi ol
migh be associa ed wi h gas oin es inal dys unc ion. Bac e ial
o e g ow h causes changes in he gu mucosa ha leads o
suga malabso p ion (33). Bac e ial o e g ow h, as well as o he
gas oin es inal dys unc ions, is common among subjec s wi h
PD. Up o 80% o pa ien s wi h PD show some signs o
F on ie s in Neu ology | www. on ie sin.o g 7May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
FIGURE 2 | (A) O hogonal pa ial leas squa e-disc iminan analysis (OPLS-DA) plo o signi ican ly di e en compounds be ween PD cases and con ol subjec s
(R2X =0.563, R2Y =0.402, and Q2 =0.165). Cases a e ma ked as o ange and con ols a e ma ked as g een; (B) Volcano plo plo ing VIP in OPLS-DA model
agains p[co ] alue o signi ican ly di e en compounds be ween PD cases and con ol subjec s. Colo ed acco ding o iden i ie s.
TABLE 5 | The analy ical pa ame e s o signi ican ly di e en me aboli es be ween PD cases and con ol subjec s.
Me aboli es T ansi ion Linea i y (ppm) R Repea abili y (%RSD) In e media e p ecision (%RSD) LOD LOQ
Palmi ic acid 255.2 ⇒256.3 1–25 0.9960 5.4 26.0 475.1 1,583.7
Oleic acid 281.5 ⇒282.3 1–25 1.0000 6.5 17.5 247.9 826.5
S ea ic acid 283.5 ⇒283.5 1–25 0.9998 9.3 27.7 655.3 2,184.4
Benzoic acid 121.1 ⇒77.0 0.25–2 0.9996 2.4 6.3 80.1 267.0
Myoinosi ol 239.1 ⇒179.0 0.025–0.5 0.9970 1.3 6.0 23.6 78.6
So bi ol 183.2 ⇒129.0 0.002–0.5 0.9990 1.6 10.1 1.6 5.2
Quinolinic acid 168.1 ⇒78.0 0.002–1 0.9999 0.9 19.0 1.7 5.7
*linea i y exp essed in ppm; R, coe icien o a ia ion; epea abili y and in e media e p ecision exp essed as %RSD; LOD, limi o de ec ion; LOQ, limi od quan i a ion.
gas oin es inal impai men s, which usually appea in he ea ly
s age o PD (34). Al e ed le els o galac i ol and so bi ol in
subjec s wi h PD ha e been obse ed in o he s udies as well
(11). Dec eased le els o myo-inosi ol, galac i ol, and so bi ol
we e epo ed in his s udy, while Ahmad e al. (9) showed
dec eased le els o galac i ol bu inc eased le els o so bi ol
and myo-inosi ol. Howe e , inconsis en esul s migh indica e
a lack o possible associa ion be ween PD de elopmen and
impai men s o suga me abolism and ene gy p oduc ion in
he ea ly s age o Pa kinson’s disease o e en be o e he illness
has appea ed. Fu he mo e, dec eased le els o a y acids in
subjec s wi h PD ha e been obse ed in his s udy. Palmi ic,
oleic, and s ea ic acids we e signi ican ly dec eased. Such a esul
is in co espondence wi h o he s udies ha ound a educ ion
in a y acids le els in pa ien s wi h PD (4). Ha elund e al.
(4) ound, among o he s, dec eased le els o palmi ic, oleic,
and s ea ic acids in subjec s wi h PD. The me abolism o a y
acids has epea edly been associa ed o he de elopmen and
pa hogenesis o PD. Changes in a y acids migh be associa ed
wi h mi ochond ial dys unc ion (4,5,35), neu oin lamma ion
(14), al e a ions in apop o ic signaling (36), as well as wi h
oxida i e s ess (13). E en mo e, hese indings ag ee wi h ecen
disco e ies published by ou g oup (29) in which a di e en
coho o a simila numbe o subjec s was used o he blind
disco e y o p ognos ic bioma ke s o PD using un a ge ed
me abolomics. In his s udy, se e al a y acids, including he
ones epo ed he e, we e also ound s a is ically dec eased. This
suppo s ha changes in he me abolism o a y acid could
help o unde s and he p og ession o PD and he in ol ed
species could be s ing candida es as bioma ke s o ea ly
diagnosis o he disease. Quinolinic acid is an in e media e
compound in he yp ophan-kynu enine me abolic pa hway,
which has al eady been associa ed wi h PD de elopmen (8).
Kynu enine is he main in e media e compound and i can be
me abolized in wo ways o kynu enic acid, which ac s as a
neu op o ec i e agen o o 3-hyd oxykynu enine and quinolinic
acid, which a e neu o oxic. Inc eased le els o quinolinic acid
cause neu on exci a ion by ac i a ion o N-me hyl-D-aspa a e
(NMDA) ecep o s, which consequen ly lead o exci o oxici y,
inc eased in lamma ion, and e en ually o neu onal dea h (37).
I is known ha degene a ion o dopamine gic neu ons in he
subs an ia nig a in Pa kinson’s disease is a esul o exci o oxici y.
Recen s udies (6,38) showed ha al e ed kynu enine pa hways
and hei me aboli es a e p esen in plasma and ce eb ospinal
luid, espec i ely, in subjec s wi h PD. This is in co espondence
wi h ou inding o al e ed le els o quinolinic acid. In his s udy,
inc eased concen a ion o quinolinic and 3-hyd oxykynu enine
has been ound in pa ien s wi h PD, compa ed wi h heal hy
F on ie s in Neu ology | www. on ie sin.o g 8May 2022 | Volume 13 | A icle 844841
Konje od e al. Ta ge ed Me abolomics in Pa kinson’s Disease
FIGURE 3 | Schema ic s uc u e o he Eu opean P ospec i e In es iga ion in o Cance and Nu i ion (EPIC) s udy and he samples used in s udy.
con ol subjec s, while simila indings we e as well ound by
Heilman e al. (7). Howe e , 3-hyd oxykynu enine was no
signi ican ly inc eased, al hough i showed a end in subjec s
wi h PD. I is assumed ha al e a ions in hese me aboli es a e
associa ed wi h he se e i y o PD symp oms (7). Dys unc ion
o yp ophan and he kynu enine pa hway migh esul in
inc eased oxida i e s ess, as well as neu oin lamma ion ha
would lead o neu odegene a i e p ocesses cha ac e is ic o
PD. The e o e, inc eased le els o quinolinic acid in he
subjec s ha la e de eloped PD migh indica e al e a ions in
he kynu enine pa hway in he ea ly s age, be o e i s PD
symp oms appea and migh ep esen po en ial bioma ke
o ea ly diagnosis o lead o de elopmen o he apeu ic
app oach ha could a ge kynu enine o 3-hyd oxykynu enine
con e sion (8).
Globally, he me aboli es ha ha e been ound o be
signi ican ly al e ed in his s udy a e ela ed o mi ochond ial
dys unc ion, oxida i e s ess, and he mechanisms o ene gy
p oduc ion. Conside ing ha all he olun ee s en olled in
his s udy we e heal hy a he ime o sample collec ion,
hese indings migh imply ha hese p ocesses begin o be
a ec ed be o e PD shows any symp oms. This in o ma ion
is o g ea ele ance o a disease, such as PD, in which
he de ini ion o a me aboli e panel ha can be used
o he ea ly diagnosis o he disease has been sough
o decades.
F on ie s in Neu ology | www. on ie sin.o g 9May 2022 | Volume 13 | A icle 844841
