Analy ica Chimica Ac a 1279 (2023) 341848
A ailable online 27 Sep embe 2023
0003-2670/© 2023 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-
nc-nd/4.0/).
Explo a o y op imisa ion o a LC-HRMS based analy ical me hod o
un a ge ed me abolomic sc eening o Cannabis Sa i a L. h ough
Da a Mining
M. San Nicolas
a
,
b
,
c
,
*
, A. Villa e
a
,
b
, M. Oli a es
a
,
b
, N. E xeba ia
a
,
b
, O. Zuloaga
a
,
b
, O. Aizpu ua-
Olaizola
c
, A. Usobiaga
a
,
b
a
Depa men o Analy ical Chemis y, Facul y o Science and Technology, Uni e si y o he Basque Coun y (UPV/EHU), 48940, Leioa, Basque Coun y, Spain
b
Resea ch Cen e o Expe imen al Ma ine Biology and Bio echnology (PIE), Uni e si y o he Basque Coun y (UPV/EHU), 48620, Plen zia, Basque Coun y, Spain
c
So e eign Fields S.L., 20006, San Sebas ian, Basque Coun y, Spain
HIGHLIGHTS GRAPHICAL ABSTRACT
•Empi ical mul is ep op imisa ion o
un a ge ed me abolomics s udy o
cannabis.
•Sol en combina ions inducing phase
sepa a ion p o ide inc eased me aboli e
ex ac ion capaci y.
•The e e se-phase ch oma og aphic
column p o ided g ea e me abolic
co e age while keeping good sepa a ion
o he majo pola compounds.
•Rep esen a i e and complemen a y
me abolic in o ma ion was ound in
lowe and lea issues, in con as o he
s em.
ARTICLE INFO
Keywo ds:
Plan me abolomics
Expe imen al ac o s
Cannabis sa i a L.
LC-HRMS
Me abolomic co e age
Da a mining
ABSTRACT
Backg ound: Recen inc ease in public accep ance o cannabis as a na u al medical al e na i e o ce ain
neu ological pa hologies has led o i s app o al in di e en egions o he wo ld. Howe e , due o i s p e ious
illegal backg ound, li le esea ch has been conduc ed a ound i s biochemical insigh s. The e o e, in he cu en
amewo k, me abolomics may be a sui able app oach o deepening he knowledge a ound his plan species.
Ne e heless, expe imen al me hods in me abolomics mus be ca e ully handled, as sligh modi ica ions can lead
o me abolomic co e age loss. Hence, he main objec i e o his wo k was o op imise an analy ical me hod o
app op ia e un a ge ed me abolomic sc eening o cannabis.
Resul s: We p esen an empi ically op imised expe imen al p ocedu e h ough which he b oades me abolomic
co e age was ob ained, in which ex ac ion sol en s o me aboli e isola ion, ch oma og aphic columns o LC-
* Co esponding au ho . Ba io Sa iena s/n, 48940, Leioa, Vizcaya, Spain.
E-mail add ess: [email p o ec ed] (M. San Nicolas).
Con en s lis s a ailable a ScienceDi ec
Analy ica Chimica Ac a
jou nal homepage: www.else ie .com/loca e/aca
h ps://doi.o g/10.1016/j.aca.2023.341848
Recei ed 27 June 2023; Recei ed in e ised o m 22 Sep embe 2023; Accep ed 23 Sep embe 2023
Analy ica Chimica Ac a 1279 (2023) 341848
2
qO bi ap analysis and plan - ep esen a i e biological issues we e compa ed. By explo a o y means, i was
de e mined ha he sol en combina ion composed o CHCl
3
:H
2
O:CH
3
OH (2:1:1, / ) p o ided he highes
numbe o ea u es om di e se chemical classes, as i was a wo-phase ex ac an . In addi ion, a e e se phase
2.6
μ
m C18 100 Å (150 ×3 mm) ch oma og aphic column was de e mined as he app op ia e choice o
adequa e sepa a ion and u he de ec ion o he di e se me aboli e classes. Apa om ha , o e all ch o-
ma og aphic peak quali y p o ided by each column was obse ed and he need o ba ch co ec ion me hods
h ough quali y con ol (QC) samples was con i med. A las , lea and lowe issues esul ed o p o ide com-
plemen a y me abolic in o ma ion o he plan , o he de imen o s em issue, which esul ed o be negligible.
Signi icance: I was concluded ha he op imised expe imen al p ocedu e could signi ican ly ease he pa h o
u u e esea ch wo ks ela ed o cannabis me abolomics by LC-HRMS means, as he wo k was based on p e ious
plan me abolomics li e a u e. Fu he mo e, i is c ucial o highligh ha an op imal analy ical me hod can a y
depending on he main objec i e o he esea ch, as changes in he expe imen al ac o s can lead o di e en
ou comes, ega dless o whe he he esul s a e be e o wo se.
1. In oduc ion
Up o cu en imes, Cannabis Sa i a L. has been an unau ho ised
plan species in mos coun ies o he wo ld [1]. As a consequence, li le
esea ch on i s insigh s has been ca ied ou in he pas . Howe e , in he
USA, o ins ance, 28 s a es ha e al eady passed he Medical Ma ijuana
Laws (MML) [2], and e en hough s a e and Fede al laws cu en ly
appea o be in con lic , his ou s ands an inc easing endency o
cannabis applica ion o medical pu poses, gi ing a sligh no ion o how
widesp ead i s use could be in he u u e. Due o his ac , li e a u e
a ound he inne biochemis y o cannabis is slowly expanding, since
his plan species wa an s u he in es iga ion, on behal o u he
esea ch in medicine o na u al p o enance. Thus, ea ly documen ed
esea ch wo ks in cannabis ocused in he cha ac e isa ion o i s bioac-
i e compounds, cannabinoids and e penes [3–6], which seem o ha e a
p omising u u e in pallia i e medicine [7–11], none heless, as hese
compounds a e na u ally biosyn hesised in he cannabis plan , cu en
s udies a e also ying o documen he dis inc i e ea u es o i s inne
me abolic pa hways. This way, some p ima y me aboli es as amino
acids, ca bohyd a es o o ganic acids ha e al eady been s udied o
knowledge deepening in o he plan ’s beha iou al esponse mechanisms
o chemo a s di e en ia ion, whe eas some seconda y me aboli es as
alkaloids, plan ho mones o phy os e oids ha e been acknowledged o
pa icipa ing in signalling mechanisms [12–21]. Ne e heless, es ab-
lished analy ical p ocedu es o he ensu ing o quali y con ol wi h ull
ans e abili y o esul s a e s ill lacking.
In his h ead, “omics”- ela ed s udies occupy one o he mos no ed
posi ions among cu en analy ical ends, as he decons uc ion o
molecula mechanisms which make up a biological o ganism can help o
o ally comp ehend i s inne unc ioning [22]. Mo eo e , in he
me abolome, he las downs eam ung o he e e ed “ome”-s (genome,
ansc ip ome, p o eome and me abolome) [23], he inge p in o bo h
in e nal and ex e nal pe u ba ions can be acked, making me ab-
olomics a e y help ul ool o unde s and he minu iae o he en i on-
men ha su ounds a li ing being [24]. Due o his ac , me abolomics
is b oadly applied o sea ching bioma ke s in li ing o ganisms [25].
Ne e heless, in he plan kingdom, specially, he me abolic di e si y is
as , as beyond 200,000 di e en me aboli es can compose he me ab-
olome o a plan indi idual [26]. The e o e, inno a i e bio echnological
app oaches a e necessa y, o ins ance, in p ojec s ocusing cul i a ions’
adap abili y o uneasy su oundings [27], and he applica ion o
me abolomics could be a majo upg ade [28]. Ne e heless, unless
p ope analy ical me hods a e applied o bioma ke s sea ch, me abo-
li es o in e es migh be missed o misiden i ied.
In me abolomics, he domina ing analy ical pla o m is mass spec-
ome y, commonly coupled o a ch oma og aphic sys em [29]. Among
such sepa a ion echniques, liquid ch oma og aphy (LC) is he mos
widely used one, as i is a e sa ile ool o bo h pola and non-pola
me aboli e analysis, while gas ch oma og aphy (GC) is used o he
analysis o ola ile me aboli es and p ima y me aboli es a e de i a i-
sa ion [29]. In ega d o mass spec ome y de ec ion, high- and
low- esolu ion pla o ms can be dis inguished. B oadly speaking,
low- esolu ion pla o ms (e.g., andem mass spec ome y, MS/MS) a e
usually add essed o quan i a i e a ge me abolomics, add essing
me aboli es belonging o ce ain me abolic pa hways, while
high- esolu ion mass-spec ome y (e.g., O bi ap) is used o unknown
me aboli e iden i ica ion in non- a ge ed me abolomics, p o iding a
b oade scope conce ning co e age o me abolic pa hways [30]. Hence,
LC coupled o High Resolu ion Mass Spec ome y (LC-HRMS) o e s
e sa ili y and high esolu ion o an accu a e desc ip ion o he
me abolome [31]. Ne e heless, his app oach is ne e uly unbiased,
as di e en ac o s o pa ame e s mus be de ined p io o analysis, such
as s a iona y phases o ionisa ion modes [30]. Consequen ly, p ope ly
de eloped analy ical me hods should be applied when pe o ming
me abolomics esea ch.
A p ope ly de eloped un a ge ed analy ical me hod should p ima ily
o e a b oad me abolomic co e age. As s a ed, di e en ac o s o he
whole expe imen al p ocedu e mus be op imised, in o de o build a
obus analy ical me hod o be applied in u u e wo ks ela ed o his
ield. In his ega d, wo ks ela ed o me hodological op imisa ion
ac ually exis in he ield o me abolomics. Fo ins ance, he e a e ex-
amples in he li e a u e ocussed on he op imal selec ion o ex ac ion
sol en s o he de e mina ion o as many me aboli es as possible in he
explo a o y analysis o human luids [32] o he s anda disa ion o he
de i a isa ion s ep o ola ile me aboli es in biological samples [33].
Indeed, his las op imisa ion app oach was la e applied in plan
me abolic p o iling s udies [34]. Fu he mo e, s anda dised p o ocols
o plan me abolomics ha e been p e iously de eloped based on he
objec i e o co e ing as many me aboli es as possible [35–38].
Among he expe imen al ac o s a ec ing he me aboli e co e ing
scope in cannabis, sol en s o e icien me aboli e ex ac ion should be
conside ed. Mo eo e , i he analy ical pla o m o choice is LC-HRMS,
he ch oma og aphic s a iona y phase should also be aken in o ac-
coun , as e icien sepa a ion be ween compounds would be necessa y.
In addi ion, i is s a ed ha plan s possess a complex me abolome, hus,
explo ing he me abolic in o ma ion in di e en biological issues would
also be a equi emen , as in me abolomic s udies o ganism-
ep esen a i e in o ma ion should be e ie ed.
Bea ing all ha in mind, mul i a ia e da a mining echniques, such
as P incipal Componen Analysis (PCA) and Hie a chical Clus e ing
Analysis (HCA), could be he sui able pa h o ollow o co esponding
expe imen al ac o op imisa ion in he me abolome analysis o
cannabis. Bo h echniques a e unsupe ised echniques, meaning he
a iance wi hin he ac ual da a o de ec ed me aboli es in he analysed
samples is explo ed, excluding a p edic ion ou come [39]. Conc e ely,
using hose da a mining echniques, ends explaining he a iance o
la ge da ase s, which migh no be no iced a i s glance due o hei
dimensionali y, can be spo ed and ele an in o ma ion ex ac ed [40].
These a e Thus, o ac o op imisa ion, he applied selec ion c i e ion
was based on he explained a iance by he subjec s unde s udy. In his
wo k, expe imen al ac o s such as ex ac ion sol en s, ch oma og aphic
columns and biological issues we e s udied o app op ia e un a ge ed
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
3
me abolomic sc eening o Cannabis Sa i a L.
2. Ma e ials and me hods
2.1. Reagen s and sol en s
Sol en s o ch oma og aphic analysis, ace oni ile (ACN), me hanol
(MeOH) and wa e (H
2
O) o HPLC g ade, we e acqui ed in Pan eac/
AppliChem (ITW Reagen s, S.R.L., I aly). Fo mic acid (FA) o LC-MS
was pu chased om Fische Chemicals (The mo Fishe Scien i ic Inc.,
USA) and ammonium ace a e (NH
4
Ac, o molecula biology ≥98%) was
acqui ed in Sigma Ald ich (Me ck G oup, USA). Calib a ion o he
qO bi ap was pe o med wi h Pie ce™ LTQ Velos ESI Posi i e Ion
Calib a ion Solu ion and Pie ce™ ESI Nega i e Ion Calib a ion Solu ion
by The mo Scien i ic (The mo Fishe Scien i ic Inc., USA).
Fo sample ex ac ion, MeOH (anhyd ous), ichlo ome hane
(CHCl
3
, syn hesis g ade) and me hyl e -bu yl e he (MTBE, HPLC
g ade) we e ob ained in Mac on Fine (A an o , Inc., USA), Scha lau
(Scha lab S.L., Spain) and Pan eac/AppliChem (ITW Reagen s S.R.L.,
I aly), espec i ely. Wa e was il e ed o a esis i i y o 18.2 MΩ cm @
25 ◦C and a o al o ganic con en (TOC) less han 3 ng mL
−1
using a Q-
POD wa e dispense and a Millipak exp ess 40 (0.22
μ
m il e ) by Me ck
(Me ck G oup, USA).
2.2. Plan samples
The samples we e ha es ed om cannabis plan s cul i a ed in he
acili ies o So e eign Fields (So e eign Fields S.L., Spain). The plan s
we e cul i a ed indoo , in an isola ed 24 m
2
(6 m ×4 m) g ow h oom,
buil inside he g eenhouse. Inside he oom, 6 Ce amic Me al Halide
(CMH) lamps (315 W/37000 lm) by Luma ek (Luma ek L d., UK) we e
ins alled in he oo , a 2.10 m heigh , uni o mly dis ibu ed h ough he
oom, coupled o Adjus -A-Wings la ge en o ce e lec o s.
Two simila cul i a ions we e un o he op imisa ion p ocedu e. In
he i s one, jus one plan was ha es ed o he i s wo expe imen s
desc ibed in sec ion 2.4, while in he second one, 10 plan s we e ha -
es ed o he hi d expe imen desc ibed in sec ion 2.4. A chemo ype III
cannabis cul i a was cul i a ed in bo h (C
o al THC
/C
o al CBD
<0.1) [41].
Plan s we e g own in 11-L black po s con aining a soil/hummus/
nu ien mix u e. Speci ically, he mix u e was composed o 80% o
Ligh mix soil o Biobizz (Biobizz Wo ldwide S.L., Spain), 20% o
hummus, and 10 g/L o a me mix nu ien solu ion by Lu pe (Lu pe
Na u al Solu ions, Spain), which is composed o ba guano, bone meal,
kelp meal, Azomi e®, o ganic al al a, insec ass, blood meal, dolomi e,
langbeini e humic and ul ic acids, and a complex blend o hizobac e ia
and T ichode ma. The o al cul i a ion ime in bo h cases was 12 weeks,
being he i s 4 weeks he ege a i e s age du ing which he plan s
g ew, and he nex 8 weeks he lowe ing s age. The ege a i e and
lowe ing s ages we e de ined by he pho ope iod egime. In he i s 4
weeks o g ow h, he pho ope iod was 18 h ligh /6 h da k, and in he
lowe ing s age, 12 h ligh /12 h da k. Mo eo e , he lamps we e se o
wo k a 50% and 80% o he o al lamp in ensi y du ing he ege a i e
and lowe ing s ages, espec i ely. Du ing he cul i a ion ime, he oom
empe a u e was kep be ween 22 and 25 ◦C du ing he day ime, and i
did no dec ease below 18 ◦C a nigh . The ela i e humidi y o he oom
was con olled a 60% (±5%).
Finally, when he cul i a ion ime elapsed, he samples we e ha -
es ed and ins an ly ozen in liquid ni ogen, ollowing he quenching
me hod. In he i s cul i a ion, a lowe sample was aken om a single
plan , while in he second cul i a ion, lea , lowe , and s em samples
we e collec ed om each indi idual (n =10). The collec ed samples had
o be ep esen a i e o he whole plan indi idual, as each one o hem
was a biological eplica e. So, he ollowing sampling design was
de ined:
- Lea samples: A pool o a o al 12 lea es om each plan indi idual
we e collec ed, 6 o hem om he uppe hal pa o he plan , and
he 6 om he lowe hal , all o hem uni o mly dis ibu ed ac oss he
plan .
- Flowe samples: Samples om each biological eplica e we e
composed o 5 di e en lowe s. The i s lowe sample was aken
om he apical end o he plan , in he uppe poin o i . The second
one was collec ed 30 cm lowe han his poin , nex o he s em. F om
his poin , he wo nea es ami ica ions, which we e opposi e o
each o he , we e ollowed un il he end-poin o hem, and he
lowe s si ua ed he e we e collec ed. Finally, he lowes lowe si -
ua ed nex o he s em was collec ed. The collec ed 5 lowe s we e
pooled o a ep esen a i e lowe sample o each plan indi idual.
- S em sample: The lowes poin o he plan s em was se as he
e e ence poin . 5 cm abo e he e e ence poin , he s em was ho i-
zon ally cu , cu ing down he whole plan indi idual. F om his
poin , a 10 cm long s em piece was aken as a sample. The possible
ami ica ions on he s em sample we e cu down, hus, only
emaining he p incipal s em.
A e wa ds, he samples we e kep a −80 ◦C empe a u e p io o
analysis.
2.3. Gene al wo k low
Fig. 1 shows he gene al wo k low ollowed o achie e op imal pa-
ame e s o an un a ge ed analy ical me hod o he de e mina ion o as
many me aboli es as possible in Cannabis Sa i a L.
The op imisa ion p ocedu e was di ided in h ee main s eps.
The i s s ep was he assessmen o he ex ac ion sol en o
ep esen a i e isola ion o he me aboli es om he plan ma ix. In his
con ex , 5 di e en sol en s (i. e., me hanol-wa e , me hanol-wa e
acidi ied wi h o mic acid, me hanol-chlo o o m, chlo o o m-me hanol-
wa e , me hyl e -bu yl e he -me hanol-wa e ) and li e a u e-based
ex ac ion p o ocols we e compa ed on he same lowe sample, using
he same analysis condi ions such as ch oma og aphic sepa a ion and
de ec ion se up. Since mos o he li e a u e wo ks ega ding plan
me abolomics use e e se phase liquid ch oma og aphy condi ions, he
ex ac s we e analysed using a e e se phase column (i. e., 2.6
μ
m C18
100 Å (150 ×3 mm) column). Once he ex ac ion p o ocol was se , he
second s ep was he e alua ion o h ee ch oma og aphic columns (i.e.,
2.6
μ
m C18 100 Å (150 ×3 mm), 2.6
μ
m PS-C18100 Å (150 ×3 mm) and
BEH 1.7
μ
m Amide Hyd ophilic In e ac ion Liquid Ch oma og aphy
(100 ×2.1 mm) columns). In he las s ep, a small-scale cul i a ion
expe imen was designed o he me abolomic-like s udy o 10 plan -
clones, in which lea , lowe and s em issues o he plan we e ana-
lysed using he condi ions se in he p e ious expe imen s.
2.4. Expe imen al p ocedu e
P io o any ex ac ion he ozen samples we e g ound by Spex
SampleP ep F eeze Mill 6770 (The mo Fishe Scien i ic Inc. USA),
using liquid ni ogen, which wo ked unde he ollowing condi ions: 1
min o sample p ecooling ollowed by 1 min g inding cycle a 8 cycles
pe second (cps). The esul was a homogeneous-powde - ozen sample.
As p e iously men ioned, in he i s op imisa ion s ep, 5 di e en
li e a u e-based me aboli e ex ac ion sol en s (see Table 1) we e
compa ed o choose he ex ac ion sol en o sol en combina ion ha
p o ides he b oades me abolomic co e age o cannabis. The ex ac-
ions we e ca ied ou in a P ecellys/C yollys homogenise sys em
(Be in echnologies, F ance). 100 mg g ound homogeneous lowe
sample was weighed, 8 Z O beads (1.4 mm diame e ) we e added o
each ial and 1.2 mL o he es ed sol en we e added acco ding o he
p o ocols speci ied in Table 1. All he assays we e ca ied ou in
iplica e.
The ex ac ions we e pe o med a a 6400- pm a e, wi h 3 cycles o
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
4
60 s un and 15 s es be ween each un (6400-3x60-15). The empe -
a u e was kep a 4 ◦C wi h liquid ni ogen du ing he ex ac ion ime. In
s udy cases Nº 1, Nº 2 and Nº 3 (see Table 1), he espec i e mix u e o
sol en s esul ed in o al miscibili y a he speci ied con en s. On he
con a y, in s udy cases Nº 4 and Nº 5 (see Table 1), he con en s o hei
main non-pola sol en s (CHCl
3
and MTBE, espec i ely) caused hei
immiscibili y wi h he H
2
O:MeOH mix u e. As a consequence, phase
sepa a ion occu ed in he la e cases, whe e 50% ( / ) o he o al
ex ac an consis ed o he non-pola phase (CHCl
3
in case Nº 4 and
MTBE in case Nº 5) and he o he 50% ( / ) consis ed o he pola phase
(H
2
O:MeOH, 1:1 / ). In he single-phase ex ac an cases (cases Nº 1, Nº
2 and Nº 3), he espec i e sol en s we e added in o de o pola i y, om
he leas pola o he mos pola . In he case o he wo-phase ex ac ions
(i.e., Nº 4 and Nº 5, see Table 1), he p ocedu e was pe o med in wo
s eps. Fi s , he sample was ea ed wi h 300
μ
L MeOH and 300
μ
L H
2
O
( he pola phase), and a e one ex ac ion un, 600
μ
L o he co e-
sponding main non-pola sol en (CHCl
3
o MTBE) we e subsequen ly
added o sequen ial ex ac ion o he sample.
Once he ex ac ion was comple ed, he ials we e cen i uged in he
Alleg a X–30R cen i uge (Beckman Coul e , USA) o 5 min, a 21,000 g
o ce and 4 ◦C. The supe na an s we e decan ed o ano he 2 mL ial o
a second cen i uga ion un (15 min a 21,000 g and 4 ◦C). A las , p io
o analysis, he pola and non-pola ex ac s we e sepa a ed in he case
o wo-phase ex ac ions, and hey we e ans e ed o ch oma og aphic
ials and dilu ed a a 1:20 a io in MeOH, as well as he single-phase
ex ac s. All he ex ac s in hese assays we e analysed by means o
LC-qO bi ap using a e e se phase Kine ex 2.6
μ
m C18 100 Å (150 ×3
mm) wi h AJ0-8782 C-18 p e-column (Phenomenex, USA), as de ailed
in sec ion 2.5.
Ha ing se he ex ac ion p ocedu e, he app op ia e ch oma o-
g aphic column was selec ed h ough explo a o y analysis o a lowe
sample. The compa ed columns we e: i) he p e iously used e e se
phase Kine ex 2.6
μ
m C18 100 Å (150 ×3 mm) wi h AJ0-8782 C-18 p e-
column (Phenomenex, USA) o enhanced hyd ophobic e en ion, ii) an
aqueous s able wi h posi i e su ace cha ged e e se phase Kine ex 2.6
μ
m PS-C18 100 Å (150 ×3 mm) wi h AJ0-8950 PS-C18 p e-column
(Phenomenex, USA) o imp o ed hyd ophilic a ini y, and, iii) a
no mal phase Acqui y UPLC BEH 1.7
μ
m Amide Hyd ophilic In e ac ion
Liquid Ch oma og aphy (HILIC) 130 Å (100 ×2.1 mm) wi h Acqui y
UPLC BEH 1.7
μ
m Amide Vangua d 130 Å (2.1 ×5 mm) p e-column
(Wa e s, USA) o enhanced sepa a ion and selec i i y o pola
compounds.
Apa om ha , he ch oma og aphic d i and signal luc ua ion o
Fig. 1. Gene al expe imen al wo k low o e iew.
Table 1
S udied ex ac ion-sol en combina ions o suspec sc eening o Cannabis.
Ex ac ion Sol en s (%, / ) Re e ence
H
2
O MeOH CHCl
3
MTBE
1 MW (Single-
phase)
20 80 – – [42]
2 MWFA (Single-
phase)
20 (0.1%
FA)
a
80 – – [35]
3 MC (Single-phase) – 80 20 – [30]
4 CMW (Two-
phases)
25 25 50 – [43]
5 M MW (Two-
phases)
25 – 25 50 [44]
a
The ex ac ion p ocedu e Nº 2 is equal o he Nº 1, bu acidi ied a 0.1% ( / )
wi h FA. Ac onyms: M: Me hanol; W: Wa e ; FA: Fo mic Acid; C: Chlo o o m;
M : Me hyl e -Bu yl E he .
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
5
he mos sui able column sequence we e explo ed ac oss he analysis
sequence. Fo doing so, h ee eplica es o he lowe sample we e
andomly analysed 10 imes each, and a pooled sample was pe iodically
analysed be ween hem, se ing as a quali y con ol (QC) e e ence.
Va iances o he samples we e explo ed h oughou he analysis
sequence, compa ing he esul s ob ained wi h QC co ec ion and
wi hou co ec ion.
In he hi d op imisa ion s ep, as he ex ac an and he ch oma o-
g aphic column a e se up, he small-scale me abolomic s udy o he 10
plan -clones om he second cul i a ion (see 2.2. Plan samples sec ion)
was conduc ed o ep esen a i e biological issue explo a ion,
compa ing lea , s em and lowe samples.
2.5. Ch oma og aphic me hods and gene al MS se ings
Un a ge ed analysis in he plan ex ac s was pe o med by The mo
Scien i ic Dionex Ul ima e 3000 liquid ch oma og aph coupled o a
The mo Scien i ic Q Exac i e Focus quad upole-O bi ap mass spec-
ome e (UHPLC-q-O bi ap), equipped wi h a hea ed elec osp ay
ionisa ion sou ce (HESI, The mo-Fishe Scien i ic, CA, USA). Analysis
we e pe o med wi h 5
μ
L injec ion om each ex ac in bo h ionisa ion
modes, posi i e and nega i e, wi h he au oma ic injec o (a 5 ◦C). To
a oid any ca yo e , he injec ion needle was cleaned be o e and a e
e e y injec ion wi h 50
μ
L o H
2
O:MeOH (90:10, / ).
Rega dless o he column used, ch oma og aphic analysis consis ed
o wo mobile phases, wa e (sol en A) as pola sol en and ACN (sol-
en B) as non-pola sol en , bo h o hem con aining ei he 0.1% ( / )
o FA o 10 mM NH
4
Ac, o espec i e measu emen s in posi i e o
nega i e ionisa ion.
In he case o he ch oma og aphic sepa a ion using he C18 column,
he eluen low a e was kep a 0.3 mL/min. Du ing he analysis ime,
he sol en g adien s a ed a 95% A (held o 1 min), hen linea ly
dec eased o 5% An un il minu e 16 (held o 10 min) and inally
e u ned o he ini ial condi ions in 5 min, whe e hey we e main ained
o 1 min. Rega ding he sepa a ion condi ions using he PS-C18 column,
he eluen low a e was also 0.3 mL/min and he g adien was s ab-
lished as ollows. The mobile phase g adien s a ed a 97% A, which
was held o he i s 2 min, hen dec eased o 85% An un il minu e 5,
ollowed by a change o 5% An un il minu e 11 (held o 14 min), be o e
e u ning o ini ial condi ions in 4 min (held o 1 min). A las , o he
sepa a ions using HILIC column, bo h mobile phase low and composi-
ion we e changed. The low g adien s a ed a 0.2 mL/min (held o 3
min), hen i inc eased o 0.3 mL/min (held o 21 min) and i was again
dec eased o 0.2 mL/min un il minu e 35, when he analysis un ended.
Rega ding he mobile phase composi ion g adien , he analysis me hod
s a ed a 3% A o he i s 3 min, hen inc eased o 15% An un il minu e
5, changing o 25% An un il minu e 14 and a inal inc ease o 60% An
un il minu e 17 (held o 3 min), be o e e u ning o ini ial condi ions in
3 min (held o 12 min).
Rega dless o he ionisa ion mode, he HESI ionisa ion sou ce was
ope a ed unde he ollowing pa ame e s: sp ay ol age o 3.20 kV,
sp ay cu en o 0.50
μ
A, he capilla y empe a u e was kep a 320 ◦C
and S-lens RF le el a 55.0, he shea h gas (ni ogen) low a e a 48
a bi a y uni s (au), auxilia y gas low a e a 11 (au), sweep gas low
a e a 2 (au), and auxilia y gas hea e empe a u e a 310 ◦C. E e y
h ee days Pie ce LTQ ESI Calib a ion Solu ions we e used o ex e nal
calib a ion o he ins umen .
Da a acquisi ion in he high- esolu ion mass spec ome e (qO bi-
ap) was done in Full MS-Da a dependan MS
2
(Full-MS-dd-MS
2
) dis-
co e y acquisi ion mode o e e y analysis o he op imisa ion p ocess.
Full-scan mass spec um was collec ed in a scan ange be ween 70 and
1000 m/z wi h a esolu ion o 70,000 FWHM o an m/z o 200. Th ee
addi ional scans we e pe o med in dd-MS
2
mode wi h a esolu ion o
17,500 FWHM, an isola ion window o 0.8 m/z, and applying an s epped
no malized collision ene gy (nce) o 10%, 35% and 75% in he collision
cell, whe e he i s mass ixed was 50.0 m/z and in ensi y o AGC a ge
o 2.00⋅10
5
. The so wa e used was Xcalibu 4.0 (The mo-Fishe -
Scien i ic).
2.6. Da a p ocessing
Da a p ocessing was pe o med using Compound Disco e e 3.3
so wa e by The mo Fische . A minimum signal- o-noise a io (S/N) o 3
was s ablished in o de o conside a ch oma og aphic peak easible and
subsequen ly in eg a e i s co esponding unde -peak a ea. In he
de ec ed ch oma og aphic ea u es, MS1 was used as p ecu so o
compound de ec ion, wi h a mass ole ance o 5 ppm a e pe o ming 5
scans pe easible peak. Compounds we e also g ouped ac oss samples
wi h a mass ole ance o 5 ppm and a e en ion ime (RT) ole ance o
0.2 min based on he p e e ed ions [M+H]
+1
and [M −H]
−1
o posi i e
ionisa ion and nega i e ionisa ion, espec i ely. Mo eo e , a maximum
h eshold o 1 min peak wid h a hal peak heigh was s ablished in e e y
de ec ed ea u e. Addi ionally, possible gaps we e illed wi h 1.5 S/N
a io and also 5 ppm mass ole ance.
In he s udy cases o his wo k, only signi ican ea u es we e
conside ed. To be deemed signi ican , he de ec ed ea u es had o pass
he ollowing il e s: g ea e ch oma og aphic peak a ea han 1,000,000
in ensi y coun s, less a iance han 30% be ween co esponding epli-
ca es and, a leas , pa ial pu a i e spec al ma ch wi h a minimum
single candida e compound among p o ided spec al agmen a ion li-
b a ies. These e e ed lib a ies we e Cannabis Sa i a L. endogen sus-
pec lis , e ie ed om Plan Me abolic Ne wo k da abase [40],
LipidMaps s uc u e da abase [41] and Endogenous me aboli es da a-
base o 4400 compounds p o ided by The mo Scien i ic. Fu he mo e,
de ec ed ea u es we e also il e ed acco ding o ch oma og aphic peak
quali y, which was de ined by ou c i e ia: he jaggedness, he zig-zag
index, he FWHM2base and he modali y o he peaks [45]. Each o
hese me ics was measu ed a 5/10, being his alue, he ep esen a ion
o i s con ibu ion compa ed o he o he pa ame e s con ibu o s. Each
measu ed me ic con ibu ed equally o he o e all peak quali y.
The e o e, he peak quali y il e h eshold was se a 5/10, as only peaks
su passing his o e all alue would be accep ed as signi ican o he
s udy. Fo me aboli e iden i ica ion in he signi ican ch oma og aphic
ea u es, mzLogic da a analysis algo i hm and Mass F on ie 7.0 spec al
in e p e a ion so wa e we e used, bo h om The mo Fishe Scien i ic.
In each o he signi ican ch oma og aphic ea u es, me aboli e candi-
da es om spec al lib a ies we e anked acco ding o he simila i y
be ween expe imen al and es ima ed MS2 spec um and he F agmen
Ion Sea ch (FISh) sco e was calcula ed in he i e candida es wi h
highes simila i y sco e o s uc u al elucida ion and pu a i e me abo-
li e iden i ica ion. The FISh sco e was calcula ed wi h a high accu acy
mass ole ance o 2.5 mmu, low accu acy mass ole ance o 0.5 Da and a
S/N h eshold o 3. Fea u e anno a ion was pe o med acco ding o he
con idence le els de ined in Sch impe-Ru ledge e al. [46]. Since no pu e
e e ence s anda d was used, he highes con idence le el was 2 (pu a-
i e iden i ica ion, MS2 ma ch). The ea u es ul illing all hose p e ious
de ined c i e ia we e la e analysed h ough unsupe ised mul i a ia e
da a-analysis o de e mine he a iances and ends be ween he ana-
lysed samples and anno a ed signi ican ea u es. These da a-analysis
app oaches we e ca ied ou using Me aboanalys 5.0 [47–49].
3. Resul s and discussion
3.1. Ex ac an selec ion
To de e mine he mos sui able ex ac an o op imal me abolomic
sc eening o cannabis, he ange o co e age o each ex ac ion sol en
combina ion was explo ed. None heless, o a oid he assessmen o
hypo he ical alse posi i es, he p ocessed da a was il e ed acco ding o
he cons ain s s a ed in sec ion 2.6., so only ch oma og aphic a eas o
signi ican ea u es we e e alua ed h ough mul i a ia e da a analysis.
The esul s a e p esen ed in Table S1 o he Suppo ing Ma e ial (SM),
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
6
whe e a o al o 171 ea u es we e de ec ed in posi i e ionisa ion mode
and 35 ea u es in nega i e ionisa ion mode. The a ea alues we e au o
scaled and ans o med o loga i hmic scale (Log
10
), p io o analysing
he da a by P incipal Componen Analysis (PCA), and he ends o
g ea e a iance we e de e mined hough explo a o y means (see
Fig. 2).
Acco ding o he PCA sco es plo , PC1 explained 71.3% and 82.1% o
he o al a iance in he posi i e and nega i e ionisa ion, espec i ely.
Rega ding PC1, single-phase ex ac an s (MC, MW and MWFA) seemed
o ha e no ele ance a all since mos o he a iables a e ela ed o he
an agonis phases (pola and non-pola ) o he wo-phase ex ac ions
(CMW and M MW) ha loca ed a opposed endpoin s o PC1. PC2,
which explained 17.6% and 9.4% o he o al a iance in posi i e and
nega i e ionisa ions, espec i ely, seemed o be ela ed wi h he
dis inc ion o single-phase and wo-phase ex ac an s. Acco ding o
hese ac s, an agonis phase ex ac s o he wo-phase ex ac an s p e-
sen ed g ea e signi icance o e he o al a iance in he s udy, which
sugges ed ha a wo-phase ex ac an would be mo e app op ia e o
ep esen a i e me abolomic co e age o cannabis. Indeed, phase di -
e en ia ion could be he eason o he enhancemen o he ex ac ion
yield, a he expense o single-phase ex ac ions, as bo h e y pola and
e y non-pola me aboli es could be quan i a i ely isola ed om he
Fig. 2. PCA sco es and loadings o he il e ed esul s in he ex ac an op imisa ion s ep a) Sco es plo o posi i e ionisa ion esul s b) Loadings plo o posi i e
ionisa ion esul s c) Sco es plo o nega i e ionisa ion esul s d) Loadings plo o nega i e ionisa ion esul s.
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
7
plan ma ix.
Going deepe in o he esul s ob ained using wo-phase ex ac an s, i
could be no ed ha sco es co esponding o he ex ac s o CMW appea
o be u he han he co esponding M MW ex ac s om he null
ele ance poin o he a iable space acco ding o PC1, which could
indica e a g ea e s onge s a is ical in luence o he CMW ex ac ion.
Ne e heless, a i s glance, i was di icul o asse ha he e is ac u-
ally a signi ican di e ence be ween bo h wo-phase ex ac ions. Thus,
he ela i e ex ac ion yield o he signi ican ch oma og aphic ea u es
go using wo-phase ex ac an s was assessed (see Fig. 3). Acco ding o
hese esul s, e en hough he di e ence among he yield o bo h
ex ac an s was ba ely signi ican , in e e y s udy case g ea e a e age
peak a ea was ob ained using CMW as ex ac an . As he composi ion o
he pola phase was equal in bo h cases (50% me hanol and 50% wa e ,
/ ), he di e ence esided in he co esponding non-pola sol en .
Chlo o o m would p esen g ea e ex ac ion yield o non-pola com-
pounds han me hyl e -bu yl e he , as chlo o o m possesses g ea e
pa i ion coe icien (LogP
CHCl3
=2.3, LogP
C5H12O
=0.9). The e o e,
heo e ically, chlo o o m p esen s 2.55 imes g ea e a ini y owa ds
non-pola compounds, esul ing in an enhanced ex ac ion yield o
me aboli es o his na u e. On he o he hand, g ea e pola i y g adien
be ween pola and non-pola phases could also enhance pola compound
ex ac ion capabili y in i s co esponding phase, esul ing in a g ea e
ex ac ion yield o hese me aboli es o his na u e as well. Hence, based
on he explo a o y esul s, he ex ac ion sol en combina ion composed
o chlo o o m-me hanol-wa e (50 %-25 %–25%, / ) was conside ed
he mos sui able o ep esen a i e un a ge ed me abolomics sc eening
o cannabis. I should also be no ed ha g ea e coun o signi ican
ea u es was de ec ed in posi i e ionisa ion analysis (171) han in
nega i e ionisa ion analysis (35), hus gi ing he o me case g ea e
s a is ical weigh in he explo a o y op imisa ion p ocess.
3.2. Ch oma og aphic column selec ion
A e pe o ming he da a p ocessing and il e ing o he esul s o he
s udied ch oma og aphic condi ions as desc ibed in sec ion 2.6,
ob ained esul s o pola and non-pola ex ac s a e esumed in Table S2
and Table S3 o SM, espec i ely. De ec ed ea u e coun s and co e-
sponding a eas ob ained wi h each o he s udied ch oma og aphic
columns a e displayed in Fig. 4.
The i s ac o no e was ha , ei he in posi i e o nega i e ionisa-
ion modes, g ea e numbe o signi ican peaks we e anno a ed when
using he C18 ch oma og aphic column. A o al o 195 signi ican ea-
u es we e de ec ed using he C18 column, whe eas he HILIC and he
PS-C18 columns yielded 85 and 117 signi ican ea u es, espec i ely.
Secondly, we obse ed ha , ega dless o he ch oma og aphic column
used, a highe numbe o ea u es we e de ec ed in he posi i e ionisa-
ion mode, as i happened in he p e ious op imisa ion s ep. A i s
sigh , his ac acknowledges ha analysis h ough posi i e ionisa ion
mode, ega dless o he analy ical condi ions applied, could p o ide
b oade in o ma ion, as mo e ch oma og aphic ea u es could be
anno a ed.
In addi ion o he numbe o ea u es, he a eas o he ch oma o-
g aphic peaks we e also a ac o o be conside ed. As expec ed, he HILIC
column p o ided good sepa a ion and la ge peak a eas o cha ac e is ic
pola compounds, such as amino acids, among o he s. Conc e ely,
choline (m/z 104.1071), p oline (m/z 116.0707), be aine (m/z
118.0863), leucine (m/z 132.1018), aspa agine (m/z 133.0607), ig-
onelline (m/z 138.0549) and glu amine (m/z 147.0763) and we e
de ec ed wi h he la ges peak a eas in he posi i e ionisa ion mode
acquisi ion, all o hem pu a i ely iden i ied (iden i ica ion con idence
le el 2), while in he esul s o nega i e ionisa ion mode hyd oxy h e-
onine (m/z 134.0460), caldi ol (m/z 253.0936) and cou a ic acid (m/z
295.0457) s ood ou , iden i ied as en a i e s uc u es (iden i ica ion
con idence le el 3). On he o he hand, he C18 column p o ided la ge
ch oma og aphic peak a eas in he analysis o he non-pola ex ac ,
whe e cannabinoids highligh ed, along wi h lipids and s e oids. In his
case, he la ges peak a eas belonged o
α
- a nesene (m/z 205.1948), 6-
[(1E,3E,5E,7E,9E,11E)-9,11-dime hyl e adeca-1,3,5,7,9,11-hexaenyl]-
5-e hyloxane-2,4-dione (m/z 341.2104), cannabidiolic acid (m/z
359.2210), a milla in (m/z 415.2109) and myxopy onin B (m/z
432.2373), pu a i ely iden i ied in posi i e ionisa ion (iden i ica ion
Fig. 3. Rela i e co e age o he de ec ed signi ican ea u es o he pola and non-pola ex ac s o he CMW and M MW ex ac an s: a) Pola ex ac s h ough
posi i e ionisa ion b) Non-pola ex ac s h ough posi i e ionisa ion c) Pola ex ac s h ough nega i e ionisa ion d) Non-pola ex ac s h ough nega i e ionisa ion.
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
8
con idence le el 2), while 11-Deoxyco icos e one (m/z 329.2123, le el
3), 17
α
-hyd oxyp egnelone (m/z 331.2280, le el 3) cannabidiolic acid
(m/z 357.2071, le el 2) and esol in D6 (m/z 359.2230, le el 2) high-
ligh ed in nega i e ionisa ion.
Acco ding o Fig. 5, i was also de e mined ha he PS-C18 column,
could be deemed as he leas ele an among he s udied columns o be
used in un a ge ed me abolomics. As i s s a iona y phase elied on a
midway pola i y na u e be ween he C18 and he HILIC column, i ailed
o p o ide mo e in o ma ion han he al eady ob ained wi h he o he
columns. In pa icula , he PS-C18 column does no allow he de ec ion
o non-pola compounds ei he in numbe o ype o hose al eady
de ec ed by he C18 column, while o he de ec ion o pola compounds
i is p e e able o use he HILIC column.
Mo eo e , i is wo h men ioning ha mos o he pola compounds
de ec ed in he pola ex ac using he HILIC we e also de ec ed in he
measu emen s using C18 column. O cou se, i should no be igno ed he
ac ha he pola i y o he e e sed-phase column implied a lowe
e en ion a ini y owa ds compounds o pola na u e, so ha bo h he
a eas o hei co esponding ch oma og aphic peaks and he di e ence
in sepa a ion ime be ween hem we e signi ican ly lowe wi h he C18
column; ne e heless, a highe numbe o compounds we e de ec ed. Fo
ins ance, some amino acids such as L-me hionine (m/z 150.0582), ami-
noadipic acid (m/z 162.0760) o L- yp ophan (m/z 205.0970), alkaloid
de i a i es such as nico inamide (m/z 123.0553), 4-Indoleca baldehyde
Fig. 4. Numbe o signi ican ea u e coun (SFC) de ec ed by each o he ch oma og aphic columns and hei co esponding peak a eas (CPA). a) SFC in pola phase,
posi i e and nega i e ionisa ion modes b) SFC in non-pola phase, posi i e and nega i e ionisa ion modes c) CPA s SFC in pola phase, posi i e ionisa ion mode d)
CPA s SFC in non-pola phase, posi i e ionisa ion mode e) CPA s SFC in pola phase, nega i e ionisa ion mode ) CPA s SFC in non-pola phase, nega i e ion-
isa ion mode.
M. San Nicolas e al.
Analy ica Chimica Ac a 1279 (2023) 341848
9
(m/z 146.0599) o ans-3-indoleac ylic acid (m/z 188.0704), me abo-
li es ha play an impo an ole in he inne egula ion o he plan s (all
o hem we e iden i ied a con idence le el 2) we e no de ec ed in he
pola ex ac by he HILIC column, ins ead hey we e anno a ed wi h he
C18 column. Tha did no happen in he case o he analysis o he non-
pola ex ac wi h he HILIC column, as only 7 and 10 ea u es we e
de ec ed in posi i e and nega i e ionisa ion modes, espec i ely.
Apa om he anno a ed ea u e numbe and hei co esponding
peak a eas, hei quali y should also be a ac o o conside . Fu he
da a-analysis o he esul s o un a ge ed me abolomic esea ch would
depend on he ob ained peak a eas, hence, in o de o a oid undue de-
ia ions, de ec ed peak should pass a minimum quali y h eshold, as
poo ch oma og aphic peaks may de ia e esul s om hei ue alues.
In Fig. 6, peak quali ies o he de ec ed ea u es a e displayed, and he
a e age peak quali y was calcula ed a 95% con idence o each ch o-
ma og aphic column. The a e age peak quali ies o e ed by each column
we e simila in he ou s udy cases shown in he igu e. In ac , he
in e nal de ia ion wi hin each column was g ea e han he di e ence
be ween columns, hus, he quali y o he de ec ed ea u es was ully
compa able in he di e en s udy cases. The PS-C18 column o e ed he
smalles de ia ion in he di e en s udy cases, none heless, i was p e-
iously obse ed ha i s con ibu ion, acco ding o he objec i e o he
wo k, was he leas among he h ee columns. Apa om ha , he C18
column had g ea e s a is ical weigh in he peak quali y explo a ion, as
i p o ided a highe numbe o signi ican ea u es.
Hence, based on all hose esul s and obse a ions, analysis h ough
he C18 column is sugges ed o he in ended pu pose, bea ing in mind
ha i s s a iona y phase will p esen g ea e e en ion a ini y owa ds
non-pola compounds. The e o e, i was concluded ha C18 column
would be he mos sui able o a b oade me abolomic co e age.
Mo eo e , i is known ha in me abolomic s udies, whe e la ge
sample se s a e analysed in a single un, ch oma og aphic signal d i
can occu along he sequence. This phenomenon can lead o misleading
conclusions since he ch oma og aphic signals may depend on he
acquisi ion- ime. Hence, o ensu e accu a e da a acquisi ion, he ch o-
ma og aphic a ea d i was assessed suing he C18 column o de ine he
QC co ec ion e ec h ough he analysis ime. The da a o un il e ed
Log
10
ans o med esul s can be seen in Fig. 6. This un il e ed aw da a
is p esen ed in Table S4 (Pola ex ac , QC co ec ed da a), Table S5
(Pola ex ac , non-co ec ed da a), Table S6 (Non-pola ex ac , QC
co ec ed da a) and Table S7 (Non-pola ex ac , non-co ec ed da a) o
SM. Since no il e was applied o he de ec ed ch oma og aphic ea-
u es, he ea u e anno a ion in his sec ion was limi ed o en a i e
s uc u e o he me aboli es (le el 3), molecula o mula ma ch (le el 4)
o o jus a unique ch oma og aphic ea u e (le el 5), acco ding o he
me aboli e iden i ica ion le el speci ied in Sch impe-Ru ledge e al. [46].
A i s glance, an appa en di e ence be ween QC-co ec ed and
non-co ec ed esul s can be dis inguished. On he one hand, he Com-
pound Disco e e 3.3 so wa e allowed he co ec ion o he e en ion
ime (RT) shi s in he alignmen o ch oma og aphic peaks ac oss he
samples h ough Ch omAlign algo i hm [50], ha ing he QC sample as a
e e ence. As i can be seen in he RT shi s co ec ion plo s, in some o
he cases RT shi s o up o 0.2 min we e co ec ed, om whe e i could
be deduced ha he lack o RT shi s co ec ion could also ha e a sig-
ni ican impac on he subsequen ch oma og aphic peak a ea in eg a-
ion, hus, also a ec ing he a ea alues and, he e o e, he subsequen
da a-analysis. On he o he hand, i could also be obse ed ha he
cumula i e non-co ec ed signal o he QC sample exhibi ed signi ican
luc ua ions h ough he ime span o he analysis sequence, which
appea ed o be independen o he da a acquisi ion ime, as hey did no
ollow a de ined end. Signal co ec ion h ough he SERRF (Sys em-
a ical E o Remo al using Random Fo es ) QC no maliza ion me hod
[51], which was he one implemen ed in he Compound Disco e e 3.3
da a p ocessing so wa e, esul ed in mo e cons an a ea signal esul s in
he ch oma og aphic ea u es h ough sequence ime. Hence, hese wo
co ec ion me hods led o lowe in a-g oup a iance o he da a, as i
can be seen in he co esponding PCA sco e plo s. In he case o he pola
ex ac , i was clea ha co ec ing he da a using QC samples allowed a
clus e ing o eplica es wi h educed a iance. Since he analysed sam-
ples we e echnical eplica es o h ee biological eplica es o he same
lowe sample, ideal esul s would be expec ed o p esen null a iance
be ween ei he o he eplica es, so, based on he empi ical esul s, he
smalle he a iance be ween eplica es, he mo e eliable he ob ained
esul s would be. In he case o he non-pola ex ac , his was no so
e iden , as he a iance be ween eplica es did no di e so much be-
ween QC-co ec ed and non-co ec ed da a, so i would be di icul o
asce ain om jus he PCA plo s ha he e was an appa en di e ence
be ween he wo cases. Ne e heless, as s a ed be o e, he QC signal
luc ua ion plo e ealed luc ua ions in he summed signals o he QC
sample in di e en analysis imes. Wha i could be de e mined was ha
Fig. 5. Peak quali ies o he de ec ed signi ican ea u es and hei a e age alues in each o he s udied cases, in a 95% con idence in e al a) pola phase, posi i e
ionisa ion mode b) non-pola phase, posi i e ionisa ion mode c) pola phase, nega i e ionisa ion mode d) non-pola phase, nega i e ionisa ion mode.
M. San Nicolas e al.
