Ci a ion: Maimó-Ba celó, A.;
Ma ín-Saiz, L.; Fe nández, J.A.;
Pé ez-Rome o, K.; Ga ias-A jona, S.;
La a-Almúnia, M.; Pié ola-Lope egui, J.;
Bes a d-Escalas, J.; Ba celó-Coblijn, G.
Polyunsa u a ed Fa y Acid-En iched
Lipid Finge p in o Glioblas oma
P oli e a i e Regions Is Di e en ially
Regula ed Acco ding o Glioblas oma
Molecula Sub ype. In . J. Mol. Sci.
2022,23, 2949. h ps://doi.o g/
10.3390/ijms23062949
Academic Edi o : Gio anni Luca
G a ina
Recei ed: 4 Feb ua y 2022
Accep ed: 4 Ma ch 2022
Published: 9 Ma ch 2022
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2022 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 ps://
c ea i ecommons.o g/licenses/by/
4.0/).
In e na ional Jou nal o
Molecula Sciences
A icle
Polyunsa u a ed Fa y Acid-En iched Lipid Finge p in o
Glioblas oma P oli e a i e Regions Is Di e en ially Regula ed
Acco ding o Glioblas oma Molecula Sub ype
Albe Maimó-Ba celó1,2,† , Lucía Ma ín-Saiz 3,† , JoséA. Fe nández 3, Ka im Pé ez-Rome o 1,2,
San iago Ga ias-A jona 4,5,6 , Mónica La a-Almúnia 7,8, Ja ie Pié ola-Lope egui 1,2, Joan Bes a d-Escalas 9,*
and Gwendolyn Ba celó-Coblijn 1,2,*
1Ins i u d’In es igacio Sani a ia Illes Balea s (IdISBa, Heal h Resea ch Ins i u e o he Balea ic Islands),
07120 Palma, Spain; [email p o ec ed] (A.M.-B.); [email p o ec ed] (K.P.-R.);
ja ie [email p o ec ed] (J.P.-L.)
2Resea ch Uni , Uni e si y Hospi al Son Espases, 07120 Palma, Spain
3Depa men o Physical Chemis y, Facul y o Science and Technology, Uni e si y o he Basque
Coun y (UPV/EHU), Ba io Sa iena s/n, 48940 Leioa, Spain; [email p o ec ed] (L.M.-S.);
[email p o ec ed] (J.A.F.)
4Qui ónsalud Medical Cen e , 07300 Inca, Spain; d [email p o ec ed]
5Son Ve íQui ónsalud Hospi al, Balea ic Islands, 07609 Son Ve i Nou, Spain
6Hospi al de Lle an , 07680 Po o C is o, Spain
7Depa men o Neu osu ge y, Jimenez Diaz Founda ion Uni e si y Hospi al, Reyes Ca olicos A ., No 2,
28040 Mad id, Spain; [email p o ec ed]
8Rube In e na ional Hospi al, Maso S ., No 38, 28034 Mad id, Spain
9Bioanalysis and Pha macology o Bioac i e Lipids Resea ch G oup, Lou ain D ug Resea ch Ins i u e,
Uni e si éCa holique de Lou ain, 1200 B uxelles, Belgium
*Co espondence: [email p o ec ed] (J.B.-E.); [email p o ec ed] (G.B.-C.)
† These au ho s equally con ibu ed o his wo k.
Abs ac :
Glioblas oma (GBM) ep esen s one o he deadlies umo s owing o a lack o e ec i e
ea men s. The ad e se ou comes a e wo sened by high a es o ea men discon inua ion, caused
by he se e e side e ec s o emozolomide (TMZ), he e e ence ea men . The e o e, unde s anding
TMZ’s e ec s on GBM and heal hy b ain issue could e eal new app oaches o add ess chemo he apy
side e ec s. In his con ex , we ha e p e iously demons a ed he memb ane lipidome is highly cell
ype-speci ic and e y sensi i e o pa hophysiological s a es. Howe e , li le emains known as o
how memb ane lipids pa icipa e in GBM onse and p og ession. Hence, we employed an ex i o
model o assess he impac o TMZ ea men on heal hy and GBM lipidome, which was es ablished
h ough imaging mass spec ome y echniques. This app oach e ealed ha bioac i e lipid me abolic
hubs (phospha idylinosi ol and phospha idyle hanolamine plasmalogen species) we e al e ed in
heal hy b ain issue ea ed wi h TMZ. To be e unde s and hese changes, we in e oga ed RNA
exp ession and DNA me hyla ion da ase s o he Cance Genome A las da abase. The esul s enabled
GBM sub ypes and pa ien su i al o be linked wi h he exp ession o enzymes accoun ing o he
obse ed lipidome, hus p o ing ha explo ing he lipid changes could e eal p omising he apeu ic
app oaches o GBM, and ways o amelio a e TMZ side e ec s.
Keywo ds:
glioblas oma; MALDI-IMS lipidomics; emozolomide; modula gene exp ession;
molecula
sub ypes; lipid me abolism
1. In oduc ion
Glioblas omas (GBM), he mos agg essi e ype o as ocy oma, a e he mos equen
malignan p ima y b ain umo , accoun ing o 15% o all b ain umo s and p esen ing a
wo ldwide incidence o 3–4 pe 100,000 people [
1
]. S anda d ea men consis s o su ge y
ollowed by chemo he apy and adio he apy [
2
]. Un o una ely, his ea men ende s
In . J. Mol. Sci. 2022,23, 2949. h ps://doi.o g/10.3390/ijms23062949 h ps://www.mdpi.com/jou nal/ijms
In . J. Mol. Sci. 2022,23, 2949 2 o 22
a low median su i al o i een mon hs [
3
]. The e e ence chemo he apeu ic agen is
emozolomide (TMZ), a DNA alkyla ing agen . Reg e ably, i induces se e e side e ec s
such as dizziness and blu ed ision and, consequen ly, a ound 20% o pa ien s ea ed
wi h TMZ discon inue ea men [
4
]. Due o i s high- a e o discon inui y and poo o e all
su i al imp o emen , TMZ ea men is conside ed pallia i e (non-cu a i e) [5].
As in many o he cance s, he lack o models accu a ely mimicking gene ic he e ogene-
i y and umo mic oen i onmen hinde s he s udy o glioblas oma [
5
]. Howe e , he la es
ad ances in genomics enabled TP53, ATRX, TERT, NF1, PTEN, and EGFR o be iden i ied as
GBM gene d i e s, p o iding new insigh s in o he de elopmen o GBM [
6
,
7
]. Mo eo e ,
mu a ions on IDH1 o IDH2 genes ha e been ela ed o be e pa ien ou comes [
8
]. Fu -
he mo e, h ee ansc ip omic molecula sub ypes a e de ined based on di e en genomic
al e a ions: p oneu al, classical, and mesenchymal; hus enabling he iden i ica ion o
speci ic epigene ic al e a ions as well as molecula sub ype-dependen in e ac ions wi h
he immune mic oen i onmen [
9
]. In his sense, changes in cell- ype composi ion, sec e ed
ex acellula esicles, and soluble ac o s con ibu e o GBM mic oen i onmen he e o-
genei y and a e all in ol ed in ea men esis ance and umo ecu ence [
10
]. In e ac ion
be ween glial cells and GBM cells is key o umo g ow h and p og ession. While umo -
associa ed as ocy es can in e ac wi h endo helial cells and pe icy es om he b ain–blood
ba ie
[11–13]
, hey also appea o be in ol ed in he limi ed esponse o adia ion o
emozolomide chemo he apy [
13
,
14
]. GBM cells also show he abili y o communica e wi h
inna e immune cells (mic oglia), by changing hei pheno ype o enhance umo g ow h
and su i al [
5
]. Hence, he de ini ion o hese molecula sub ypes and he di e en ial con-
ibu ion o he mic oen i onmen componen s is a c i ical s ep owa ds he de elopmen
o mo e accu a e ea men s a egies [9].
Second o adipose issue, he b ain is he mos lipid en iched o gan in he body,
pa icula ly in memb ane lipids, phospholipids, and sphingolipids. Cell memb ane lipid
composi ion, o memb ane lipidome, includes hund eds o molecula species, each o which
ha e speci ic oles ha emain mos ly unknown. Consis en ly, he lipidome has p o en
o be highly sensi i e o pa hophysiological p ocesses, and al e a ions in cell lipid p o ile
a e associa ed wi h mul iple pa hophysiological p ocesses such as di e en ia ion [
15
–
18
],
p oli e a ion [
19
–
21
], and cance de elopmen [
22
–
24
]. In ac , memb ane lipid species a e
sensi i e enough o be used as bioma ke s o se e al cance ypes, such as o a ian cance ,
p os a e cance , and b eas cance [
25
–
27
]. Impo an ly, he i up ion o imaging MS (IMS)
echniques in o he lipidomic ield ha e clea ly demons a ed how speci ic and sensi i e
he lipidome is o bo h physiological changes and pa hological insul s. In his con ex ,
b ain issue emains one o he mos analyzed issues by di e en IMS echniques, showing
concise lipid species dis ibu ion be ween g ey and whi e ma e in human samples
[28–30]
and highligh ing he po en ial o lipidomics o s udy GBM. Thus, imaging lipidomic
echniques oge he wi h machine lea ning p o ocols a e capable o apidly classi ying
gliomas based solely on hei lipid p o iles, o e ing a po en ial ool o in aope a i e
examina ion and apid classi ica ion [31].
In addi ion, he egula ion o se e al genes in ol ed in lipid me abolism, such as
SCD and ELOVL6 which pa icipa e in mono- and polyunsa u a ed a y acid (MUFA,
PUFA, espec i ely) syn hesis, is al e ed in GBM [
32
,
33
]. One o hese PUFAs is a achidonic
acid, he p ecu so o a la ge amily o bioac i e molecules in ima ely in ol ed in in lam-
ma ion. Impo an ly, a s udy analyzing human GBM iden i ied signi ican co ela ions
be ween he high exp ession o mPGES1 and PTGR1, enzymes in ol ed in he syn hesis
o p os aglandins, and was ela ed o poo pa ien su i al [
34
]. Con e sely, highe gene
exp essions o 15-HPGD, in ol ed in p os aglandin ca abolism, was associa ed wi h be e
ou comes in cance pa ien s, including GBM [34].
In . J. Mol. Sci. 2022,23, 2949 3 o 22
Taking in o accoun he close ela ionship al eady es ablished be ween lipid me abolism
and GBM de elopmen , we conside ed he s udy o i s lipidome and he impac on lipid
composi ion o TMZ es ablished wi h spa ial esolu ion which could help o e eal new
a ge s o ea GBM o add ess TMZ side e ec s.
2. Resul s
2.1. Impac o GBM and TMZ T ea men on B ain Tissue Lipidome
An ex i o app oach was employed o assess he impac o TMZ on GBM lipidome.
Thus, su gically heal hy and GBM biopsies we e ob ained om he same pa ien , imme-
dia ely placed in o DMEM-F12 cell cul u e medium and incuba ed in he p esence o
absence o TMZ (10 mg/mL, 4 h) o DMSO ( ehicle). A e he incuba ion pe iod, biopsies
we e snap- ozen in he absence o c yop o ec i e subs ances, and heal hy b ain and GBM
sec ions we e analyzed using MALDI-IMS in nega i e- and posi i e-ion mode a 50
µ
m
la e al esolu ion.
2.1.1. Cha ac e iza ion o Heal hy B ain and Glioblas oma Lipidome
Fi s , he p esence o he egions di e ing in hei lipidome wi hin he b ain and
GBM biopsies we e in es iga ed o es ablish he di e ences be ween heal hy and GBM
lipidome. We p e iously demons a ed he high co ela ion exis ing be ween IMS lipid
clus e s s ablished by K-means and HD-RCA algo i hms and his ological issue s uc u es,
cell ypes, o e en cell pa hophysiological s a es [
18
,
22
,
35
,
36
]. B ie ly, he so wa e conside s
he dis ibu ion o all lipids de ec ed by applying a clus e ing o segmen a ion analysis.
Then, i ende s a isual ep esen a ion whe e pixels wi h a simila o p oximal lipid
composi ion a e g ouped in o he same egion acco ding o a clus e ing/segmen a ion
algo i hm. In his pa icula s udy, one o he challenges was o co ela e lipid dis ibu ion
wi h i s ana omical coun e pa . Un o una ely, he maligniza ion p ocess by i sel al eady
implies a loss o issula a chi ec u e, while he ine a chi ec u e is ine i ably comp omised
du ing incuba ion in a cell cul u e medium. Thus, we ocused on dis inguishing cells
exhibi ing a high p oli e a i e a e, as highly p oli e a i e cells a e equi ed o issue
main enance and he p og ession o cance [
37
,
38
]. Ki-67 (MKI67) is a cell p oli e a ion
ma ke and p ognos ic ma ke in GBM [
39
–
41
] in ol ed in he pe ich omosomal laye
du ing mi osis [
42
]. The e o e, MKI67+ s aining was used o de ine egions o in e es by
immuno luo escence (IF). These MKI67+ egions we e employed o iden i y he lipid clus e
ha o e lapped he mos wi h he mos highly p oli e a i e egion in bo h heal hy and
GBM biopsies. While MKI67 s aining e ealed se e al egions wi h di e en IF in ensi y in
umo issue, s aining in ensi y was homogenous and low in heal hy issue (Figu e 1).
Nex , we compa ed he lipid p o ile o he highly p oli e a i e a eas o heal hy and
GBM samples. Lipidome analysis de ec ed 124 di e en lipid species belonging o 11 lipid
classes. Unsupe ised PCA using all lipid species demons a ed ha he lipidome o he
highly p oli e a i e egions (MKI67+ clus e s) disc imina es be ween heal hy and GBM
issue (Figu e 2a).
In . J. Mol. Sci. 2022,23, 2949 4 o 22
In . J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 4 o 23
Figu e 1. B ain p oli e a ion zones and MALDI-IMS clus e ing compa ison. Rep esen a i e images
o IF and MALDI-IMS o heal hy and umo ea ed and non- ea ed a e shown. Heal hy b ain and
GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO ( ehicle) o TMZ (10
mg/mL, 4 h) and analyzed by MALDI-IMS a 50–100 µm la e al esolu ion. B ain p oli e a ion zones
we e de e mined by MKI67+ IF s aining and used o selec he MALDI-IMS clus e , gene a ed by
HD-RCA om consecu i e issue sec ions [43]. HD-RCA clus e ing enabled he iden i ica ion o he
IMS egions o in e es (Clus e *) wi h g ea e co ela ion (e alua ed by di ec isual inspec ion)
wi h he MKI67+ IFs, based on he simila i y o lipidomic con en in each MALDI-IMS expe imen .
P oli e a ion zones a e ma ked in o ange in MKI67+ IFs. DAPI was used as a nucleus ma ke
(ma ked in blue on IF MKI67 images). The dis ibu ion o PI 38:4 (885.55 m/z) is shown as a ep e-
sen a i e MALDI-IMS lipid dis ibu ion. Colo scale indica es he in ensi y o he PI 38: 4 -H dis i-
bu ion (0, black; 1, whi e). HD-RCA numbe o segmen s was se om 2 o 5, wi h p io backg ound
noise il a ion using in-house MATLAB algo i hms [43].
Nex , we compa ed he lipid p o ile o he highly p oli e a i e a eas o heal hy and
GBM samples. Lipidome analysis de ec ed 124 di e en lipid species belonging o 11 lipid
classes. Unsupe ised PCA using all lipid species demons a ed ha he lipidome o he
highly p oli e a i e egions (MKI67+ clus e s) disc imina es be ween heal hy and GBM
issue (Figu e 2a).
Figu e 1.
B ain p oli e a ion zones and MALDI-IMS clus e ing compa ison. Rep esen a i e images
o IF and MALDI-IMS o heal hy and umo ea ed and non- ea ed a e shown. Heal hy b ain
and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO ( ehicle) o TMZ
(10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50–100
µ
m la e al esolu ion. B ain p oli e a ion
zones we e de e mined by MKI67+ IF s aining and used o selec he MALDI-IMS clus e , gene a ed
by HD-RCA om consecu i e issue sec ions [
43
]. HD-RCA clus e ing enabled he iden i ica ion o
he IMS egions o in e es (Clus e *) wi h g ea e co ela ion (e alua ed by di ec isual inspec ion)
wi h he MKI67+ IFs, based on he simila i y o lipidomic con en in each MALDI-IMS expe imen .
P oli e a ion zones a e ma ked in o ange in MKI67+ IFs. DAPI was used as a nucleus ma ke (ma ked
in blue on IF MKI67 images). The dis ibu ion o PI 38:4 (885.55 m/z) is shown as a ep esen a i e
MALDI-IMS lipid dis ibu ion. Colo scale indica es he in ensi y o he PI 38:4 -H dis ibu ion
(0, black; 1, whi e). HD-RCA numbe o segmen s was se om 2 o 5, wi h p io backg ound noise
il a ion using in-house MATLAB algo i hms [43].
In . J. Mol. Sci. 2022,23, 2949 5 o 22
In . J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 5 o 23
Figu e 2. Impac o umo igenesis on MKI67+ clus e s a he le el o lipid class composi ion. Heal hy
b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO ( ehicle) o
TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50 µm la e al esolu ion. (a) PCA analysis
showed a clea sepa a ion be ween heal hy b ain (blue), and GBM ( ed) issue, implying dis inc
lipid inge p in s. Va iance is explained by he wo componen s: 80.4%. (b) Rela i e in ensi y a i-
a ion o he main memb ane lipid class was analyzed in he posi i e-ion mode in heal hy b ain and
GBM. (c) Rela i e in ensi y a ia ion o he main memb ane lipid class was analyzed in he nega i e-
ion mode in heal hy b ain and GBM. Values a e exp essed as ela i e peak in ensi y no malized o
o al ion cu en and ep esen he mean ± SEM (n = 5). S a is ical analysis was assessed using - es
analysis. * p < 0.05; ** p < 0.01. Abb e ia ions: HexCe , hexosylce amides; PE, phospha idyle hano-
lamine; PE P-, PE plasmalogen; PG, phospha idylglyce ol; PI, phospha idylinosi ol; PS, phospha i-
dylse ine; SM, sphingomyelin; Sul , sul a ide; PC, phospha idylcholine; DG, diacylglyce ol; TG, i-
acylglyce ol; Ce , ce amide.
S a is ical compa ison o lipid classes demons a ed ha phospha idyle hanolamine
(PE), PI, and sphingomyelin (SM) we e signi ican ly inc eased in GBM (1.5-, 3.6-, 1.8- old
inc ease, espec i ely). Con e sely, a s a is ical dec ease was obse ed in sul a ide con en
in GBM issue (2.9- old dec ease). I is wo h men ioning ha he alues ep esen ed in
Figu e 2 accoun o he in ensi y de ec ed du ing MALDI-IMS analysis, which depends
on he ioniza ion capabili y o each compound. Consequen ly, he in ensi y alues o di -
e en lipid classes canno be compa ed. Fo ins ance, sul a ides show high in ensi y al-
ues despi e only accoun ing o app ox. 4% o o al lipids in whi e ma e [44]. Thus, com-
pa isons mus be made exclusi ely wi hin he same lipid classes.
Figu e 2.
Impac o umo igenesis on MKI67+ clus e s a he le el o lipid class composi ion. Heal hy
b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO ( ehicle) o
TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50
µ
m la e al esolu ion. (
a
) PCA analysis
showed a clea sepa a ion be ween heal hy b ain (blue), and GBM ( ed) issue, implying dis inc lipid
inge p in s. Va iance is explained by he wo componen s: 80.4%. (
b
) Rela i e in ensi y a ia ion o he
main memb ane lipid class was analyzed in he posi i e-ion mode in heal hy b ain and GBM. (
c
) Rela i e
in ensi y a ia ion o he main memb ane lipid class was analyzed in he nega i e-ion mode in heal hy
b ain and GBM. Values a e exp essed as ela i e peak in ensi y no malized o o al ion cu en and
ep esen he mean
±
SEM (n= 5). S a is ical analysis was assessed using - es analysis. * p< 0.05;
** p< 0.01
. Abb e ia ions: HexCe , hexosylce amides; PE, phospha idyle hanolamine; PE P-, PE plas-
malogen; PG, phospha idylglyce ol; PI, phospha idylinosi ol; PS, phospha idylse ine; SM, sphingomyelin;
Sul , sul a ide; PC, phospha idylcholine; DG, diacylglyce ol; TG, iacylglyce ol; Ce , ce amide.
S a is ical compa ison o lipid classes demons a ed ha phospha idyle hanolamine
(PE), PI, and sphingomyelin (SM) we e signi ican ly inc eased in GBM (1.5-, 3.6-, 1.8- old
inc ease, espec i ely). Con e sely, a s a is ical dec ease was obse ed in sul a ide con en
in GBM issue (2.9- old dec ease). I is wo h men ioning ha he alues ep esen ed in
Figu e 2accoun o he in ensi y de ec ed du ing MALDI-IMS analysis, which depends
on he ioniza ion capabili y o each compound. Consequen ly, he in ensi y alues o
di e en lipid classes canno be compa ed. Fo ins ance, sul a ides show high in ensi y
alues despi e only accoun ing o app ox. 4% o o al lipids in whi e ma e [
44
]. Thus,
compa isons mus be made exclusi ely wi hin he same lipid classes.
Nex , we analyzed he changes occu ing a he molecula species le el. O special
in e es was he di e en ial impac on diacyl PE and PE P- species. In he heal hy b ain,
he mos abundan PE diacyl species we e 36:1, 40:6, and 38:4 (35.6, 13.9, and 11.3%,
espec i ely). In GBM issue, he mos abundan species we e 36:1, 38:4, and 36:2 (30.0,
13.4, and 10.1%, espec i ely). To al diacyl PE le els we e g ea ly inc eased in GBM
issues (Figu e 2), especially PE 34:0 and 36:4 (Figu e 3a). In his s udy, we conside ed
PE e he lipids we e mos ly PE plasmalogens as his is he mos abundan subg oup in
he b ain [
45
]. While PE plasmalogen o al le els we e simila in bo h s udy g oups, he
In . J. Mol. Sci. 2022,23, 2949 6 o 22
disease had a p o ound e ec a he PE plasmalogen molecula species le el. In heal hy
issue, PE P-36:2 and 38:4 ollowed by 34:1 and 40:6 species (20.8, 14.7, 13.7, and 10.2%
o o al PE plasmalogen, espec i ely) we e he mos abundan PE plasmalogen species.
Con e sely, 38:4, 36:4, 40:6, and 38:6 we e he mos abundan PE plasmalogen species in
he MKI67+ GBM clus e (22.0, 14.8, 10.4, and 10.3% o o al PE plasmalogen, espec i ely).
Thus, he esul s showed a solid endency o 36C- and 38C:PUFA-con aining species o
inc ease in de imen o 40C:PUFA- and MUFA/DUFA-con aining species when compa ing
GBM o heal hy issue. This shi was s a is ically signi ican o 36:2 (20.8 s. 3.6%), 36:4
(
1.9 s. 14.8%
), 38:4 (14.7 s. 22.0%), 38:6 (2.8 s. 10.3%), and 40:5 (7.1 s. 4.3%) (Figu e 3b).
Consis en wi h he li e a u e, he mos abundan PI species in heal hy and GBM issue was
38:4 (54.9 and 75.7%, espec i ely), which in u n was he only species o show a signi ican
inc ease in GBM compa ed o heal hy issue (Figu e 3c).
In . J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 6 o 23
Nex , we analyzed he changes occu ing a he molecula species le el. O special
in e es was he di e en ial impac on diacyl PE and PE P- species. In he heal hy b ain,
he mos abundan PE diacyl species we e 36:1, 40:6, and 38:4 (35.6, 13.9, and 11.3%, e-
spec i ely). In GBM issue, he mos abundan species we e 36:1, 38:4, and 36:2 (30.0, 13.4,
and 10.1%, espec i ely). To al diacyl PE le els we e g ea ly inc eased in GBM issues
(Figu e 2), especially PE 34:0 and 36:4 (Figu e 3a). In his s udy, we conside ed PE e he
lipids we e mos ly PE plasmalogens as his is he mos abundan subg oup in he b ain
[45]. While PE plasmalogen o al le els we e simila in bo h s udy g oups, he disease had
a p o ound e ec a he PE plasmalogen molecula species le el. In heal hy issue, PE P-
36:2 and 38:4 ollowed by 34:1 and 40:6 species (20.8, 14.7, 13.7, and 10.2% o o al PE plas-
malogen, espec i ely) we e he mos abundan PE plasmalogen species. Con e sely, 38:4,
36:4, 40:6, and 38:6 we e he mos abundan PE plasmalogen species in he MKI67+ GBM
clus e (22.0, 14.8, 10.4, and 10.3% o o al PE plasmalogen, espec i ely). Thus, he esul s
showed a solid endency o 36C- and 38C:PUFA-con aining species o inc ease in de i-
men o 40C:PUFA- and MUFA/DUFA-con aining species when compa ing GBM o
heal hy issue. This shi was s a is ically signi ican o 36:2 (20.8 s. 3.6%), 36:4 (1.9 s.
14.8%), 38:4 (14.7 s. 22.0%), 38:6 (2.8 s. 10.3%), and 40:5 (7.1 s. 4.3%) (Figu e 3b). Con-
sis en wi h he li e a u e, he mos abundan PI species in heal hy and GBM issue was
38:4 (54.9 and 75.7%, espec i ely), which in u n was he only species o show a signi ican
inc ease in GBM compa ed o heal hy issue (Figu e 3c).
Figu e 3. Impac o umo igenesis on MKI67+ clus e s a he le el o molecula species composi ion.
Heal hy b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO
( ehicle) o TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50 µm la e al esolu ion. Each
g aph ep esen s he pe cen age o each molecula species wi hin each phospholipid class. (a) PE,
(b) PE P-, (c) PI, (d) SM, and (e) Sul . Values a e exp essed as he pe cen age o o al a y acid
(mole%) and ep esen he mean ± SD, n = 5. Fo simplici y, species accoun ing o less han 5% we e
no included in he g aph. De ailed esul s o all lipid species iden i ied a e included in Supplemen-
a y Table S1. S a is ical signi icance was assessed using - es analysis, * p < 0.05; ** p < 0.01; *** p <
0.001. Abb e ia ions: PE, phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylino-
si ol; SM, sphingomyelin; Sul , sul a ide.
Figu e 3.
Impac o umo igenesis on MKI67+ clus e s a he le el o molecula species composi ion.
Heal hy b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO
( ehicle) o TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50
µ
m la e al esolu ion. Each
g aph ep esen s he pe cen age o each molecula species wi hin each phospholipid class. (
a
) PE,
(
b
) PE P-, (
c
) PI, (
d
) SM, and (
e
) Sul . Values a e exp essed as he pe cen age o o al a y acid (mole%)
and ep esen he mean
±
SD, n= 5. Fo simplici y, species accoun ing o less han 5% we e no
included in he g aph. De ailed esul s o all lipid species iden i ied a e included in Supplemen a y
Table S1. S a is ical signi icance was assessed using - es analysis, * p< 0.05; ** p< 0.01; *** p< 0.001.
Abb e ia ions: PE, phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol; SM,
sphingomyelin; Sul , sul a ide.
Rega ding sphingolipids, he main SM species in he heal hy b ain we e d36:1, d42:2,
and d34:1 (41.0, 26.8, and 10.6%, espec i ely), whe eas in GBM hey we e d34:1, d36:1 and
d36.2 (37.5, 26.6, and 8.3%, espec i ely). The mos s iking changes be ween s udy g oups
we e he sha p inc ease in d34:1 and d40:1 molecula species and he dec ease in d42:2 in
GBM issue (Figu e 3d). Finally, in he heal hy b ain, he main sul a ide species we e d42:2,
42:1, and d42:1 (50.8, 15.9, and 11.3%, espec i ely), whe eas in he umo hey we e d42:2,
In . J. Mol. Sci. 2022,23, 2949 7 o 22
d36:3, and d36:4 (23.9, 21.0, and 11.4%, espec i ely). In his s udy, sul a ide species showed
he mos p o ound changes in composi ion in GBM issue. Compa ed o heal hy issue,
GBM p esen ed lowe le els o d42:2, d42:1, d44:2, and 42:1, wi h signi ican alues o he
la e h ee, while d34C and d36C-species we e inc eased, signi ican ly o d36:2 (Figu e 3e).
2.1.2. E ec s o Temozolomide T ea men on he Lipidome o he P oli e a i e A eas in
GBM and Heal hy B ain
Al hough TMZ is he s anda d ca e ea men o GBM, he impac i migh ha e on
he lipidome emains unknown. Thus, o u he unde s and his aspec , we compa ed he
lipid p o iles o he clus e s o e lapping wi h he mos p oli e a i e a eas, i.e., he MKI76+
a eas in he biopsies incuba ed wi h TMZ.
PCA e ealed ha he wo expe imen al g oups, TMZ ea ed and non- ea ed GBM,
could be success ully di e en ia ed based on he lipid p o ile o he p oli e a i e egions
(Figu e 4). Howe e , TMZ ea men did no bea a s a is ically signi ican impac a he
le el o lipid class, and only PE 38:5 was signi ican ly inc eased a he molecula species
le el (Figu e 5).
In . J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 7 o 23
Rega ding sphingolipids, he main SM species in he heal hy b ain we e d36:1, d42:2,
and d34:1 (41.0, 26.8, and 10.6%, espec i ely), whe eas in GBM hey we e d34:1, d36:1
and d36.2 (37.5, 26.6, and 8.3%, espec i ely). The mos s iking changes be ween s udy
g oups we e he sha p inc ease in d34:1 and d40:1 molecula species and he dec ease in
d42:2 in GBM issue (Figu e 3d). Finally, in he heal hy b ain, he main sul a ide species
we e d42:2, 42:1, and d42:1 (50.8, 15.9, and 11.3%, espec i ely), whe eas in he umo
hey we e d42:2, d36:3, and d36:4 (23.9, 21.0, and 11.4%, espec i ely). In his s udy, sul-
a ide species showed he mos p o ound changes in composi ion in GBM issue. Com-
pa ed o heal hy issue, GBM p esen ed lowe le els o d42:2, d42:1, d44:2, and 42:1, wi h
signi ican alues o he la e h ee, while d34C and d36C-species we e inc eased, sig-
ni ican ly o d36:2 (Figu e 3e).
2.1.2. E ec s o Temozolomide T ea men on he Lipidome o he P oli e a i e A eas in
GBM and Heal hy B ain
Al hough TMZ is he s anda d ca e ea men o GBM, he impac i migh ha e on
he lipidome emains unknown. Thus, o u he unde s and his aspec , we compa ed he
lipid p o iles o he clus e s o e lapping wi h he mos p oli e a i e a eas, i.e., he MKI76+
a eas in he biopsies incuba ed wi h TMZ.
PCA e ealed ha he wo expe imen al g oups, TMZ ea ed and non- ea ed GBM,
could be success ully di e en ia ed based on he lipid p o ile o he p oli e a i e egions
(Figu e 4). Howe e , TMZ ea men did no bea a s a is ically signi ican impac a he
le el o lipid class, and only PE 38:5 was signi ican ly inc eased a he molecula species
le el (Figu e 5).
Figu e 4. Impac o TMZ ea men on memb ane lipid class composi ion o he MKI76+ egion in
GBM. GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO ( ehicle) o
TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50 µm la e al esolu ion. (a) PCA analysis o
he p oli e a ion (MKI67+) clus e s iden i ied in non- ea ed GBM (blue), and TMZ- ea ed GBM
( ed). Va iance is explained by he wo componen s: 71.3%. (b) Lipid class ela i e in ensi y a ia ion
Figu e 4.
Impac o TMZ ea men on memb ane lipid class composi ion o he MKI76+ egion
in GBM. GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO ( ehicle) o
TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50
µ
m la e al esolu ion. (
a
) PCA analysis
o he p oli e a ion (MKI67+) clus e s iden i ied in non- ea ed GBM (blue), and TMZ- ea ed GBM
( ed). Va iance is explained by he wo componen s: 71.3%. (
b
) Lipid class ela i e in ensi y a ia ion
analyzed in he posi i e-ion mode o he p oli e a i e egions in non- ea ed GBM (DMSO) and
TMZ ea ed GBM. (
c
) Lipid class ela i e in ensi y a ia ion analyzed in he nega i e-ion mode o
he p oli e a i e egions in non- ea ed GBM (DMSO) and TMZ ea ed GBM. Values a e exp essed
as he ela i e abundance o peaks and ep esen mean
±
SEM (n= 4). S a is ical analysis was
assessed using - es analysis. Abb e ia ions: HexCe , hexosylce amides; PC, phospha idylcholine; PE,
phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol; PS, phospha idylse ine;
SM, sphingomyelin; Sul , sul a ide; PC, phospha idylcholine; DG, diacylglyce ol; TG, iacylglyce ol;
Ce , ce amide.
In . J. Mol. Sci. 2022,23, 2949 8 o 22
In . J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 8 o 23
analyzed in he posi i e-ion mode o he p oli e a i e egions in non- ea ed GBM (DMSO) and
TMZ ea ed GBM. (c) Lipid class ela i e in ensi y a ia ion analyzed in he nega i e-ion mode o
he p oli e a i e egions in non- ea ed GBM (DMSO) and TMZ ea ed GBM. Values a e exp essed
as he ela i e abundance o peaks and ep esen mean ± SEM (n = 4). S a is ical analysis was as-
sessed using - es analysis. Abb e ia ions: HexCe , hexosylce amides; PC, phospha idylcholine;
PE, phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol; PS, phospha idyl-
se ine; SM, sphingomyelin; Sul , sul a ide; PC, phospha idylcholine; DG, diacylglyce ol; TG, iacyl-
glyce ol; Ce , ce amide.
Figu e 5. Impac o TMZ ea men on memb ane lipid species o he MKI76+ egion in GBM biop-
sies. Heal hy b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO
( ehicle) o TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50 µm la e al esolu ion. Ba
diag ams compa e changes in lipid composi ion o (a) PE, (b) PE P-, (c) PI, (d) SM, and (e) sul .
Values a e exp essed as he pe cen age o o al lipid species in ha lipid class (mol%) and ep esen
mean ± SEM, n = 5. S a is ical signi icance was assessed using - es analysis. *: p < 0.05. Fo cla i y,
species accoun ing o less han 5% we e excluded om he analysis. De ailed esul s o all compa -
isons and all lipid species a e included in Supplemen a y Table S1. Abb e ia ions: PE, phospha i-
dyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol; SM, sphingomyelin; sul , sul-
a ide.
2.1.3. Temozolomide Exe s Mul iple E ec s o e he Heal hy B ain Lipidome
One o he mos ele an clinical issues o GBM ea men is he discon inui y o
chemo he apy due o side e ec s. In his s udy, we in es iga ed he impac o TMZ on
heal hy b ain lipid composi ion. The esul s showed ha he ea men induced changes
a he le el o bo h lipid classes and molecula species composi ion. Thus, PCA clea ly
sepa a ed he s udy g oups based on lipid class composi ion (Figu e 6a). S a is ical com-
pa ison e ealed a signi ican inc ease in he ea ed g oup in PI (2.2- old inc ease) and a
dec ease in hexosylce amide, and sul a ides (1.8- and 1.6- old dec ease, espec i ely).
Figu e 5.
Impac o TMZ ea men on memb ane lipid species o he MKI76+ egion in GBM biopsies.
Heal hy b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in DMSO
( ehicle) o TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50
µ
m la e al esolu ion. Ba
diag ams compa e changes in lipid composi ion o (
a
) PE, (
b
) PE P-, (
c
) PI, (
d
) SM, and (
e
) sul .
Values a e exp essed as he pe cen age o o al lipid species in ha lipid class (mol%) and ep esen
mean
±
SEM, n= 5. S a is ical signi icance was assessed using - es analysis. *: p< 0.05. Fo
cla i y, species accoun ing o less han 5% we e excluded om he analysis. De ailed esul s o
all compa isons and all lipid species a e included in Supplemen a y Table S1. Abb e ia ions: PE,
phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol; SM, sphingomyelin;
sul , sul a ide.
2.1.3. Temozolomide Exe s Mul iple E ec s o e he Heal hy B ain Lipidome
One o he mos ele an clinical issues o GBM ea men is he discon inui y o
chemo he apy due o side e ec s. In his s udy, we in es iga ed he impac o TMZ on
heal hy b ain lipid composi ion. The esul s showed ha he ea men induced changes a
he le el o bo h lipid classes and molecula species composi ion. Thus, PCA clea ly sepa-
a ed he s udy g oups based on lipid class composi ion (Figu e 6a). S a is ical compa ison
e ealed a signi ican inc ease in he ea ed g oup in PI (2.2- old inc ease) and a dec ease
in hexosylce amide, and sul a ides (1.8- and 1.6- old dec ease, espec i ely).
In . J. Mol. Sci. 2022,23, 2949 9 o 22
In . J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 9 o 23
Figu e 6. Impac o he TMZ ea men on memb ane lipid classes in he MKI67+ egion o heal hy
b ain issue. Heal hy b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies
in DMSO ( ehicle) o TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50 µm la e al esolu ion.
(a) PCA analysis o he p oli e a ion clus e s selec ed. The e was a clea sepa a ion be ween con ol
heal hy b ain (DMSO) (blue), and TMZ ea ed heal hy b ain ( ed), implying dis inc lipid inge -
p in s. (b) Lipid class ela i e in ensi y was analyzed in posi i e-ion mode a ia ion o selec ed e-
gions in con ol heal hy b ain (DMSO) and TMZ ea ed heal hy b ain. (c) Lipid class ela i e in en-
si y was analyzed in nega i e-ion mode a ia ion o selec ed egions in con ol heal hy b ain
(DMSO) and TMZ ea ed heal hy b ain. Values a e exp essed as mean ± SEM (n = 4). S a is ical
analysis was assessed using a - es analysis. The as e isk (*) indica es a signi ican di e ence be-
ween bo h condi ions * p < 0.05; *** p < 0.001. Abb e ia ions: HexCe , hexosylce amides; PC, phos-
pha idylcholine; PE, phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol;
PS, phospha idylse ine; SM, sphingomyelin; Sul , sul a ide; PC, phospha idylcholine; DG, diacyl-
glyce ol; TG, iacylglyce ol; Ce , ce amide.
Rega ding diacyl PE species, he mos abundan species in heal hy TMZ ea ed b ain
we e 36:1, 40:6, and 38:4 (27.1, 23.2, and 15.8%, espec i ely). In his class, he only signi -
ican changes we e a sligh inc ease in 34:0 (3.2 s. 5.5%) and a dec ease in 38:1 (6.1 s.
4.0%) (Figu e 7a). The mos abundan PE P- species we e 40:6, 38:4, and 40:4 (23.5, 20.4,
and 11.5%, espec i ely). T ea men inc eased many o he PUFA-con aining PE P- species
(Figu e 7b), signi ican ly o 38:4, 38:6, and 40:7 (14.7 s. 20.4%, 2.8 s. 6.5%, and 1.8 s.
3.2%, espec i ely). These inc eases in PUFA-con aining species we e compensa ed by a
dec ease in MUFA and DUFA-con aining ones, which we e signi ican o 34:1 and 36:2
(13.7 s. 5.0%, and 20.8 s. 7.3%, espec i ely). Finally, TMZ ea men aised he PUFA-
con aining PI le els signi ican ly o 38:4 and 38:5 (54.9 s. 80.4%, and 5.1 s. 6.3%) (Figu e
7c).
Figu e 6.
Impac o he TMZ ea men on memb ane lipid classes in he MKI67+ egion o heal hy
b ain issue. Heal hy b ain and GBM biopsy sec ions we e p epa ed a e incuba ing esh biopsies in
DMSO ( ehicle) o TMZ (10 mg/mL, 4 h) and analyzed by MALDI-IMS a 50
µ
m la e al esolu ion.
(
a
) PCA analysis o he p oli e a ion clus e s selec ed. The e was a clea sepa a ion be ween con ol
heal hy b ain (DMSO) (blue), and TMZ ea ed heal hy b ain ( ed), implying dis inc lipid inge p in s.
(
b
) Lipid class ela i e in ensi y was analyzed in posi i e-ion mode a ia ion o selec ed egions in
con ol heal hy b ain (DMSO) and TMZ ea ed heal hy b ain. (
c
) Lipid class ela i e in ensi y was
analyzed in nega i e-ion mode a ia ion o selec ed egions in con ol heal hy b ain (DMSO) and TMZ
ea ed heal hy b ain. Values a e exp essed as mean
±
SEM (n= 4). S a is ical analysis was assessed
using a - es analysis. The as e isk (*) indica es a signi ican di e ence be ween bo h condi ions
*p< 0.05; *** p< 0.001. Abb e ia ions: HexCe , hexosylce amides; PC, phospha idylcholine; PE,
phospha idyle hanolamine; PE P-, PE plasmalogen; PI, phospha idylinosi ol; PS, phospha idylse ine;
SM, sphingomyelin; Sul , sul a ide; PC, phospha idylcholine; DG, diacylglyce ol; TG, iacylglyce ol;
Ce , ce amide.
Rega ding diacyl PE species, he mos abundan species in heal hy TMZ ea ed
b ain we e 36:1, 40:6, and 38:4 (27.1, 23.2, and 15.8%, espec i ely). In his class, he only
signi ican changes we e a sligh inc ease in 34:0 (3.2 s. 5.5%) and a dec ease in 38:1
(
6.1 s. 4.0%
) (Figu e 7a). The mos abundan PE P- species we e 40:6, 38:4, and 40:4 (23.5,
20.4, and 11.5%, espec i ely). T ea men inc eased many o he PUFA-con aining PE P-
species (Figu e 7b), signi ican ly o 38:4, 38:6, and 40:7 (14.7 s. 20.4%, 2.8 s. 6.5%, and
1.8 s. 3.2%, espec i ely). These inc eases in PUFA-con aining species we e compensa ed
by a dec ease in MUFA and DUFA-con aining ones, which we e signi ican o 34:1 and
36:2 (13.7 s. 5.0%, and 20.8 s. 7.3%, espec i ely). Finally, TMZ ea men aised he
PUFA-con aining PI le els signi ican ly o 38:4 and 38:5 (54.9 s. 80.4%, and 5.1 s. 6.3%)
(Figu e 7c).
In . J. Mol. Sci. 2022,23, 2949 16 o 22
wi h TMZ—which p esen ed non-local ecu ence—ha e poo e o e all su i al han
pa ien s wi h local ecu ence [
77
]. I migh be o g ea in e es o measu e he exp ession
o mesenchymal ma ke s [
78
] in he subse o non-local ecu ence pa ien s, in o de o
desc ibe a posi i e co ela ion. Wi h his in mind, he genomic and me agenomic analysis
o lipid- ela ed enzymes could help in unde s anding how he obse ed lipid changes a e
egula ed a he gene le el.
In his s udy, we employed he molecula sub ype classi ica ion es ablished by Ve -
haak e al. [
47
] o del e in o he co-exp essed genes coding o lipid enzymes ha could
accoun o he GBM lipidomic pheno ype desc ibed. The analysis using TCGA-GBM
da abases iden i ied nine lipid- ela ed genes whose exp ession u ned ou o be egula ed in
a molecula sub ype-dependen manne . Fu he , i e o hese, namely PLA2G5, PLA2G2A,
FAPB5, FABP7, ELOVL2, and ALOX5AP, we e signi ican ly associa ed wi h poo o e all
and disease-speci ic su i al. PLA2G5, FABP7, and ALOX5AP gene exp ession le els also
showed conspicuous epigene ic egula ion acco ding o he speci ic gene egion me hyla ion
le els. Pa icula ly ema kable was he high me hyla ion o FABP7 and PLA2G5 in he
P oneu al sub ype, oge he wi h lowe gene exp ession. Con e sely, FABP7 and PLA2G5
exhibi ed low me hyla ion and highe gene exp ession in he Classical sub ype when com-
pa ed o he o he sub ypes. Finally, he ALOX5AP gene showed lowe me hyla ion and
highe gene exp ession in he Mesenchymal sub ype compa ed o he o he s.
The Classical sub ype gene signa u e is cha ac e ized, among o he s, by ele a ed
EGFR and NOTCH3 exp ession. Recen s udies ha e s eng hened he in ol emen o
ELOVL2 in EGFR signaling main enance and GBM p oli e a ion, h ough i s con ibu ion o
PUFA syn hesis and memb ane composi ion. Fu he , ELOVL2 exp ession was also ound
o be associa ed wi h poo su i al [
51
]. FABP7, a p o ein in ol ed in he mobiliza ion
and anspo o a y acids and wi h a high a ini y o a achidonic and docosahexaenoic
acid, is in ol ed in b ain de elopmen and has been desc ibed o imp o e cell mig a ion
and in il a ion in malignan glioma cells [
79
,
80
]. In addi ion, FAB7 modula es he ac i i y
o PKC unde a achidonic and docosahexaenoic acid supplemen a ion in GBM pa ien -
de i ed neu osphe es and cell lines [
81
]. These s udies suppo he ole o FABP7 in he
inc ease in PUFA-con aining phospholipids and hei engagemen in GBM p og ession,
especially in he Classical sub ype. Ul ima ely, he ac i i y o PLA2G5, a sec e ed PLA2 wi h
a high a ini y o unsa u a ed a y acids, may also be in ol ed in he PUFA me abolism.
Meanwhile, he ele a ed me hyla ion o ALOX5AP in he P oneu al sub ype and i s lowe
me hyla ion in he Mesenchymal sugges he in ol emen o he eicosanoid me abolism in
s omal cells, which could accoun o umo mic oen i onmen en ichmen in a achidonic
acid. P oneu al o Mesenchymal ansi ion is a desc ibed mechanism o esis ance o
chemo he apy in GBM elapse. Thus, adio he apeu ic ea men a o s he p esence o a
Mesenchymal om a P oneu al pheno ype. The ans o ma ion o P oneu al o selec ion
o Mesenchymal cells, mo e esis an o adio he apy, could ep esen he mechanisms
esponsible o his acqui ed esis ance [
82
,
83
]. In his scena io, he pi o al exp ession
and me hyla ion le els o ALOX5AP s ongly indica e a ele an ole o a achidonic acid-
con aining phospholipids and he de i ed me abolism in he umo mic oen i onmen and
GBM p og ession.
4. Ma e ials and Me hods
Ma e ials and eagen s: 2-Me cap obenzo hiazole (MBT) and 1,5-diaminonaph alene
(DAN), hema oxylin and eosin o his ological s aining, Ki67 p ima y an ibody ma ked
wi h FITC (The moFishe Scien i ic, Wal ham, MA, USA), DMEM-F12, FBS, penicillin–
s ep omycin, and TMZ we e pu chased om Sigma–Ald ich (Ba celona, Spain).
Human sample collec ion: Sample collec ion o his s udy was speci ically app o ed
by he E hics Resea ch Commi ee o he Balea ic Islands (n
º
IB 3626/18 PI). In o med con-
sen in w i ing was ob ained o each pa ien en olled in he s udy. Fou pa ien s ha bo ing
b ain umo s sugges i e o GBM, newly diagnosed a e neu ological symp oms in which
su gical esec ion o open biopsy we e indica ed, we e included. Ana omopa hological
In . J. Mol. Sci. 2022,23, 2949 17 o 22
analysis con i med he diagnosis o glioblas oma in he ou pa ien s ec ui ed. Pa ien s
ecei ed he pha macy-compounded solu ion o 5- aminolae ulinic acid 1 h be o e anes-
he ic induc ion. To al dose was acco ding o pa ien weigh (20 mg/kg). The Pen e o
®
(Zeiss
®
, Obe kochen, Ge many) su gical mic oscope wi h BLUE 400TM in eg a ed luo es-
cence module was used o su gical esec ions. Be ween 3 and 4 h a e adminis a ion o
5-aminole ulinic acid, posi i e umo samples and heal hy samples we e ob ained du ing
su gical esec ion om he su gical ma gin and immedia ely incuba ed in DMEM-F12 in
he p esence o absence o TMZ (10 mg/mL, Sigma–Ald ich (Ba celona, Spain)) o 4 h a
37 ◦C and 5% O2. These we e hen collec ed and s o ed a −80 ◦C un il p ocessing.
His ological sec ions: Sec ions 10
µ
m hick we e ob ained wi h a c yos a (Leica
CM3050S, Leica Biosys ems, We zla , Ge many) a
−
20
◦
C wi hou using c yop o ec i e
subs ances o embedding ma e ial. Sec ions we e placed on plain glass mic oscope slides
o MALDI-IMS analysis and consecu i e sec ions o IF analysis we e placed on posi i e
cha ged adhe en plain glass mic oscope slides. Samples we e s o ed a
−
80
◦
C un il
subsequen MALDI-IMS o immuno luo escence analysis.
Immuno luo escence analysis: His ological sec ions we e ixed wi h
−
20
◦
C p echilled
100
µ
L me hanol–ace one (50/50, / ) and hen incuba ed wi h 1:500 MKI67 p ima y
an ibody ma ked wi h FITC (10 mg/mL) (The moFishe Scien i ic, Wal ham, MA, USA) in
0.2% BSA-PBS, ollowing he p e iously desc ibed p o ocol in Bes a d-Escalas e al. [
18
].
Nuclei s aining was pe o med wi h 4
0
6-diamidino-2-phenylindole (DAPI, 1 mg/mL, BD
biosciences, Ba celona, Spain) a 1:10,000 dilu ion o 1 min a oom empe a u e. Finally,
samples we e obse ed wi h Axioscope Cell Obse e mic oscope and/o Zeiss LSM 700
con ocal mic oscope (Ca l Zeiss, Obe kochen, Ge many).
MALDI-IMS analysis: A o al o 32 ( ou sec ions o each ea men and ioniza ion mode)
his ological sec ions ob ained om ou di e en pa ien s we e p epa ed and analyzed by
MALDI-IMS as desc ibed in Ga a e e al. [
36
]. B ie ly, MBT o DAN we e used as ma ix o
posi i e- o nega i e-ion de ec ion, espec i ely, and deposi ed wi h he aid o ou in-house
designed sublima o de ice, which allows pe ec con ol o all he pa ame e s in ol ed in he
sublima ion p ocess [
84
]. Sec ions o b ain biopsies om di e en indi iduals we e scanned
in posi i e- and nega i e-ion mode using he o bi ap analyze o a MALDI-LTQ-O bi ap
XL (The mo Fishe , San Jose, CA, USA). The MALDI sou ce used in his s udy was he
one p o ided by he manu ac u e , which is equipped wi h a N2 lase (LTB, Be lin, model
MNL 100, 100
µ
J max powe , ellip ical spo , 60 Hz epe i ion a e), and a e y simple op ical
a angemen , consis ing o wo mi o s and a single ocusing lens o = 125 mm.
Da a we e acqui ed wi h a mass esolu ion o 60,000 in he scanning ange o 550–1000 o
nega i e-ion mode and 480–100 Da o posi i e-ion mode. Two mic oscans o 10 lase sho s
we e eco ded o each pixel and he as e size used was 50 mic ons. Spec a we e aligned
and analyzed using in-house p og ams de eloped in Ma lab (Ma hWo ks, Na ick, MA, USA).
Lipid assignmen was based on he compa ison be ween he expe imen al m/zand he species
in he so wa e’s da abase (<33,000 lipid species plus adduc s) and in he lipid maps da abase
(www.lipidmaps.o g (las accessed on 10 Janua y 2022)). Mass accu acy always measu ed
be e han 9 ppm and was ypically be e han 3 ppm. In his ype o analyze s, mass accu acy
depends somehow on he in ensi y o he peaks, he e o e, he m/zwi h highe in ensi y
p esen be e mass accu acy. Fo m/zchannels wi h se e al lipid assignmen s, “On- issue”
MS/MS and MS3 was ca ied ou in o de o unequi ocally assign hem.
Fo he sake o cla i y, only species p esen in a leas 80% o he analyzed samples
we e conside ed o u he s a is ical analysis.
In e oga ion o GBM gene exp ession and me hyla ion da ase s: Human TCGA GBM
A yme ix U133a and Me hyla ion27k da ase s we e in e oga ed using Xena B owse . The
da a ela i e o he selec ed genes shown in Figu e 10 we e downloaded and s a is ical
di e ences we e measu ed independen ly.
Co-Exp ession Modules iden i ica ion Tool (CemiTool): CEMiTool was used o iden i y
co-exp ession o lipid ela ed genes associa ed wi h he GBM molecula sub ypes, acco ding
o Ve haak e al. [
47
]. To his end, he TCGA-GBM A yU133a da ase and associa ed clinical
In . J. Mol. Sci. 2022,23, 2949 18 o 22
da a we e downloaded and analyzed. The CEMiTool analysis e u ned se en modules wi h
a di e en gene numbe and composi ion. Six o he modules we e posi i ely co ela ed,
acco ding o he no malized en ichmen sco e (NES), wi h some o he ou molecula
sub ypes applied as pheno ype labels (Figu e 8, Supplemen a y Table S3). The pa ame e s
used in CEMiTool we e he ollowing: alue o Be a chosen = 9; Pea son co ela ion
coe icien , dissimila i y h eshold used as cu o on hie a chical clus e ing = 0.8; simila
modules we e me ged, he numbe o module e u ned = 8; a ea unde cu e / o al a ea
in he Be a s R squa ed g aph = 0.802; de e mina ion coe icien (“scale- eeness” o he
esul ing ne wo k) = 0.939.
Gene on ology (GO) analysis: This was ca ied ou using he DAVID bioin o ma ic
esou ces 6.8 se e [85].
S a is ical analysis: To es ablish issue clus e s, an in-house p og ammed clus e ing
algo i hm was used, se ing he numbe o segmen s om 2 o 8. To s a is ically e alua e
he di e ences in he lipid inge p in s be ween he iden i ied a eas, - es , ANOVA, and
Pos Hoc analysis we e compu ed using SPSS S a is ics 25.0 o Windows (IBM, A monk,
NY, USA). PCA analysis and sepa a ion models we e ca ied ou by O ange Biolab 2.7.8
(Ljubljana, Slo enia) [
86
]. Fo TCGA GBM gene exp ession and me hyla ion, mul iple
compa ison o dina y one-way ANOVA wi h pos -hoc Tukey es was compu ed using
G aphPad P im ( e sion 8.0).
5. Conclusions
Al oge he , he p esen s udy p o ides solid e idence ega ding he sensi i i y o
he memb ane lipidome in he de elopmen o GBM and desc ibes he mul iple e ec s
occu ing a he le el o lipid composi ion in b ain issue upon TMZ ea men . The esul s
sugges a scena io whe e, o e sho ime pe iods, he umo issue lipidome is ini ially
impe ious o ea men , while he heal hy b ain is mo e sensi i e o i . The unexpec ed
sensi i i y o heal hy issue o his ea men could be ela ed o he side e ec s o blu y
ision o dizziness ha e en ually lead o high discon inui y a es. The e is no doub
ha knowledge o he side e ec s could help educe discon inui y a es, which cu en ly
a ec 20% o he pa ien s, and inc ease he e ec i eness o u u e he apies. Howe e , his
s udy con ains some limi a ions; namely, he numbe o samples included was limi ed,
due o he di icul y in ob aining hese ypes o samples and he challenges associa ed
wi h he ex i o model, which can apidly comp omise issue s uc u e. I is in he ligh
o hese limi a ions, ha he esul s o he in silico app oach gain mo e ele ance, as
he ansc ip omic and genomic esul s suppo he lipidomic esul s desc ibed he ein.
Thus, he in e oga ion o he TCGA-GBM ansc ip ome da abase highligh ed he ole o
PUFAs in GBM p og ession, especially in he Classical and Mesenchymal sub ypes. The
desc ip ion o how lipid enzymes a e di e en ly egula ed acco ding o molecula sub ype
was disco e ed o be in line wi h he changes in PUFA-con aining phospholipids desc ibed
in his wo k. This genomic analysis e ealed a speci ic lipid gene signa u e depending
on he molecula classi ica ion o GBM. Globally, he esul s showed newly coo dina ed
lipid–gene ic changes ha could se he base o o he app oaches o GBM ea men .
Supplemen a y Ma e ials:
The ollowing a e a ailable online a h ps://www.mdpi.com/a icle/10
.3390/ijms23062949/s1.
Au ho Con ibu ions:
Concep ualiza ion, J.B.-E., A.M.-B. and G.B.-C.; me hodology, J.B.-E., A.M.-B.,
G.B.-C., L.M.-S., J.A.F., K.P.-R., S.G.-A., M.L.-A. and J.P.-L.; so wa e, A.M.-B., L.M.-S., J.A.F., K.P.-R.
and J.P.-L.; alida ion, J.B.-E., G.B.-C., A.M.-B., L.M.-S., J.A.F., K.P.-R. and J.P.-L.; o mal analysis,
J.B.-E., G.B.-C., A.M.-B., L.M.-S. and J.A.F.; in es iga ion, J.B.-E., A.M.-B., G.B.-C., L.M.-S., J.A.F.,
K.P.-R. and J.P.-L.; esou ces, L.M.-S., J.A.F., S.G.-A. and M.L.-A.; da a cu a ion, J.B.-E., G.B.-C.,
A.M.-B., L.M.-S., J.A.F. and K.P.-R.; w i ing—o iginal d a p epa a ion, J.B.-E., G.B.-C. and A.M.-B.;
w i ing— e iew and edi ing, J.B.-E., G.B.-C. and A.M.-B.; isualiza ion, J.B.-E., G.B.-C., A.M.-B.,
K.P.-R., L.M.-S. and J.A.F.; supe ision, J.B.-E. and G.B.-C.; p ojec adminis a ion, J.B.-E. and G.B.-C.;
unding acquisi ion, J.B.-E., G.B.-C. and A.M.-B. All au ho s ha e ead and ag eed o he published
e sion o he manusc ip .
In . J. Mol. Sci. 2022,23, 2949 19 o 22
Funding:
This s udy was suppo ed in pa by he Resea ch Uni o he Uni e si y Hospi al Son
Espases (“Aju s a la In es igacióde l’Hospi al Son Espases 2017—Aplicación del lipidoma en el di-
agnós ico, p onós ico y a amien o del glioma”), Basque Go e nmen (IT1162-19), he Ins i u e
o Heal h Ca los III (PI16/02200), and he EC (Eu opean Regional De elopmen Fund, ERDF,
CP12/03338). A.M.B. and J.B-E. hold p edoc o al ellowships o he Go e n Balea (Di eccióGene al
d’Inno aciói Rece ca, FPI/2160/2018 and FPI/1787/2015, espec i ely), co- unded by he ESF (Eu o-
pean Social Fund). K.P.-R. con ac was suppo ed by he Go e n Balea (Se ei d’Ocupacióde les
IIles Balea s and Ga an ia Ju enil, JQ-SP 18/17), co- unded by he ESF. G.B.-C.’s was suppo ed by
he Ins i u e o Heal h Ca los III, co- unded by ERDF (Miguel Se e II p og am, CPII17/00005).
Ins i u ional Re iew Boa d S a emen :
The s udy was conduc ed acco ding o he guidelines o he
Decla a ion o Helsinki and app o ed by he E hics Commi ee o he Balea ic Islands (n
º
IB 3626/18
PI, da e o app o al, 5 Ap il 2018).
In o med Consen S a emen :
In o med consen was ob ained om all subjec s in ol ed in he s udy.
Da a A ailabili y S a emen :
Da a was ob ained om TCGA Resea ch Ne wo k and a e publicly
a ailable h ps://www.cance .go / cga.
Acknowledgmen s:
The esul s shown he e a e in whole o pa based upon da a gene a ed by he
TCGA Resea ch Ne wo k: h ps://www.cance .go / cga.
Con lic s o In e es : The au ho s decla e no con lic o in e es .
Re e ences
1.
Young, R.M.; Jamshidi, A.; Da is, G.; She man, J.H. Cu en ends in he su gical managemen and ea men o adul glioblas-
oma. Ann. T ansl. Med. 2015,3, 121. [PubMed]
2. Gallego, O. Nonsu gical T ea men o Recu en Glioblas oma. Cu . Oncol. 2015,22, 273–281. [C ossRe ] [PubMed]
3.
Oike, T.; Suzuki, Y.; Sugawa a, K.I.; Shi ai, K.; Noda, S.E.; Tamaki, T.; Nagaishi, M.; Yokoo, H.; Nakaza o, Y.; Nakano, T.
Radio he apy plus concomi an adju an emozolomide o glioblas oma: Japanese mono-ins i u ional esul s. PLoS ONE
2013
,8,
e78943. [C ossRe ] [PubMed]
4.
Chambe lain, M.C. Temozolomide: The apeu ic limi a ions in he ea men o adul high-g ade gliomas. Expe Re . Neu o he .
2010,10, 1537–1544. [C ossRe ] [PubMed]
5.
B oekman, M.L.; Maas, S.L.N.; Abels, E.R.; Mempel, T.R.; K iche sky, A.M.; B eake ield, X.O. Mul idimensional communica ion
in he mic oen i ons o glioblas oma. Na . Re . Neu ol. 2018,14, 482–495. [C ossRe ]
6.
Pa el, A.P.; Ti osh, I.; T ombe a, J.J.; Shalek, A.K.; Gillespie, S.M.; Wakimo o, H.; Cahill, D.P.; Nahed, B.V.; Cu y, W.T.; Ma uza,
R.L.; e al. Single-cell RNA-seq highligh s in a umo al he e ogenei y in p ima y glioblas oma. Science
2014
,344, 1396–1401.
[C ossRe ]
7.
Pa sons, D.W.; Jones, S.; Zhang, X.; Lin, J.C.H.; Lea y, R.J.; Angenend , P.; Mankoo, P.; Ca e , H.; Siu, I.M.; Gallia, G.L.; e al. An
in eg a ed genomic analysis o human glioblas oma mul i o me. Science 2008,321, 1807–1812. [C ossRe ]
8.
Louis, D.N.; Pe y, A.; Rei enbe ge , G.; on Deimling, A.; Figa ella-B ange , D.; Ca enee, W.K.; Ohgaki, H.; Wies le , O.D.;
Kleihues, P.; Ellison, D.W. The 2016 Wo ld Heal h O ganiza ion Classi ica ion o Tumo s o he Cen al Ne ous Sys em:
A summa y. Ac a Neu opa hol. 2016,131, 803–820. [C ossRe ]
9.
Wang, Q.; Hu, B.; Hu, X.; Kim, H.; Squa i o, M.; Sca pace, L.; deCa alho, A.C.; Lyu, S.; Li, P.; Li, Y.; e al. Tumo E olu ion o
Glioma-In insic Gene Exp ession Sub ypes Associa es wi h Immunological Changes in he Mic oen i onmen . Cance Cell
2017
,
32, 42–56. [C ossRe ]
10.
Fanelli, G.N.; G assini, D.; O enzi, V.; Pasquale i, F.; Mon emu o, N.; Pe ini, P.; Nacca a o, A.G.; Sca ena, C. Deciphe he
Glioblas oma Mic oen i onmen : The Fi s Miles one o New G oundb eaking The apeu ic S a egies. Genes
2021
,12, 445.
[C ossRe ]
11.
Al a ez, J.I.; Dodele -De ille s, A.; Kebi , H.; I e gan, I.; Fab e, P.J.; Te ouz, S.; Sabbagh, M.; Wosik, K.; Bou bonniè e, L.; Be na d,
M.; e al. The Hedgehog Pa hway P omo es Blood-B ain Ba ie In eg i y and CNS Immune Quiescence. Science
2011
,334,
1727–1731. [C ossRe ]
12.
O’B ien, E.; Howa h, C.; Sibson, N.R. The ole o as ocy es in CNS umou s: P e-clinical models and no el imaging app oaches.
F on . Cell. Neu osci. 2013,7, 40.
13.
Kim, J.-K.; Jin, X.; Sohn, Y.-W.; Jin, X.; Jeon, H.-Y.; Kim, E.-J.; Ham, S.W.; Jeon, H.-M.; Chang, S.-Y.; Oh, S.-Y.; e al. Tumo al RANKL
ac i a es as ocy es ha p omo e glioma cell in asion h ough cy okine signaling. Cance Le . 2014,353, 194–200. [C ossRe ]
14.
B andao, M.; Simon, T.; C i chley, G.; Giamas, G. As ocy es, he ising s a s o he glioblas oma mic oen i onmen . Glia
2019
,67,
779–790. [C ossRe ] [PubMed]
15.
Nikolopoulou, E.; Papacleo oulou, G.; Jean-Alphonse, F.; G imaldi, G.; Pa ke , M.G.; Hanyaloglu, A.C.; Ch is ian, M. A achidonic
acid-dependen gene egula ion du ing p eadipocy e di e en ia ion con ols adipocy e po en ial. J. Lipid Res.
2014
,55, 2479–2490.
[C ossRe ] [PubMed]
In . J. Mol. Sci. 2022,23, 2949 20 o 22
16.
Ponec, M.; Wee heim, A.; Kempenaa , J.; Mommaas, A.M.; Nug e en, D.H. Lipid composi ion o cul u ed human ke a inocy es in
ela ion o hei di e en ia ion. J. Lipid Res. 1988,29, 949–961. [C ossRe ]
17.
Yoshida, K.; Shinoha a, H.; Su yono; Haneji, T.; Naga a, T. A achidonic acid inhibi s os eoblas di e en ia ion h ough cy osolic
phospholipase A2-dependen pa hway. O al Dis. 2007,13, 32–39. [C ossRe ] [PubMed]
18.
Bes a d-Escalas, J.; Ga a e, J.; Maimó-Ba celó, A.; Fe nández, R.; Lopez, D.H.D.H.; Lage, S.; Reigada, R.; Kho ami, S.; Gina d,
D.; Reyes, J.; e al. Lipid inge p in image accu a ely con eys human colon cell pa hophysiologic s a e: A solid candida e as
bioma ke . Biochim. Biophys. Ac a-Mol. Cell Biol. Lipids 2016,1861, 1942–1950. [C ossRe ]
19.
A illa-Gokcumen, G.E.; Mu o, E.; Rela -Gobe na, J.; Sasse, S.; Bedigian, A.; Coughlin, M.L.; Ga cia-Manyes, S.; Egge , U.S.
Di iding cells egula e hei lipid composi ion and localiza ion. Cell 2014,156, 428–439. [C ossRe ]
20.
Sagona, A.P.; Nezis, I.P.; Pede sen, N.M.; Lies øl, K.; Poul on, J.; Rus en, T.E.; Sko heim, R.I.; Raibo g, C.; S enma k, H. P dIns(3)P
con ols cy okinesis h ough KIF13A-media ed ec ui men o FYVE-CENT o he midbody. Na . Cell Biol.
2010
,12, 362–371.
[C ossRe ]
21. Echa d, A. Phosphoinosi ides and cy okinesis: The “PIP” o he icebe g. Cy oskele on 2012,69, 893–912. [C ossRe ] [PubMed]
22.
Lopez, D.H.D.H.; Bes a d-Escalas, J.; Ga a e, J.; Maimó-Ba celó, A.; Fe nández, R.; Reigada, R.; Kho ami, S.; Gina d, D.; Okazaki,
T.; Fe nández, J.A.J.A.; e al. Tissue-selec i e al e a ion o e hanolamine plasmalogen me abolism in dedi e en ia ed colon
mucosa. Biochim. Biophys. Ac a BBA-Mol. Cell Biol. Lipids 2018,1863, 928–938. [C ossRe ] [PubMed]
23.
Lee, S.T.; Lee, J.C.; Kim, J.W.; Cho, S.Y.; Seong, J.K.; Moon, M.H. Global Changes in Lipid P o iles o Mouse Co ex, Hippocampus,
and Hypo halamus Upon p53 Knockou . Sci. Rep. 2016,6, 36510. [C ossRe ] [PubMed]
24.
Baenke, F.; Peck, B.; Miess, H.; Schulze, A. Hooked on a : The ole o lipid syn hesis in cance me abolism and umou
de elopmen . Dis. Model. Mech. 2013,6, 1353–1363. [C ossRe ]
25.
Su phen, R.; Xu, Y.; Wilbanks, G.D.; Fio ica, J.; G endys, E.C.; LaPolla, J.P.; A ango, H.; Ho man, M.S.; Ma ino, M.; Wakeley, K.;
e al. Lysophospholipids a e po en ial bioma ke s o o a ian cance . Cance Epidemiol. Bioma k. P e . 2004,13, 1185–1191.
26.
Zhou, X.; Mao, J.; Ai, J.; Deng, Y.; Ro h, M.R.; Pound, C.; Henega , J.; Wel i, R.; Bigle , S.A. Iden i ica ion o Plasma Lipid
Bioma ke s o P os a e Cance by Lipidomics and Bioin o ma ics. PLoS ONE 2012,7, e48889. [C ossRe ]
27.
Chen, X.; Chen, H.; Dai, M.; Ai, J.; Li, Y.; Mahon, B.; Dai, S.; Deng, Y. Plasma lipidomics p o iling iden i ied lipid bioma ke s in
dis inguishing ea ly-s age b eas cance om benign lesions. Onco a ge 2016,7, 36622. [C ossRe ]
28.
As iga aga, E.; Ba eda-Gómez, G.; Lomba de o, L.; F esnedo, O.; Cas año, F.; Gi al , M.T.; Ochoa, B.; Rod íguez-Pue as, R.;
Fe nández, J.A. P o iling and imaging o lipids on b ain and li e issue by ma ix-assis ed lase deso p ion/ioniza ion mass
spec ome y using 2-me cap obenzo hiazole as a ma ix. Anal. Chem. 2008,80, 9105–9114. [C ossRe ]
29.
Veloso, A.; Fe nández, R.; As iga aga, E.; Ba eda-Gómez, G.; Manuel, I.; Gi al , M.T.; Fe e , I.; Ochoa, B.; Rod íguez-Pue as, R.;
Fe nández, J.A. Dis ibu ion o lipids in human b ain. Anal. Bioanal. Chem. 2011,401, 89–101. [C ossRe ]
30.
Veloso, A.; As iga aga, E.; Ba eda-Gómez, G.; Manuel, I.; Fe e , I.; Te esa Gi al , M.; Ochoa, B.; F esnedo, O.; Rod íguez-Pue as,
R.; Fe nández, J.A.; e al. Ana omical dis ibu ion o lipids in human b ain co ex by imaging mass spec ome y. J. Am. Soc. Mass
Spec om. 2011,22, 329–338. [C ossRe ]
31.
Ebe lin, L.S.; No on, I.; Dill, A.L.; Golby, A.J.; Ligon, K.L.; San aga a, S.; G aham Cooks, R.; Aga , N.Y.R. Classi ying human b ain
umo s by lipid imaging wi h mass spec ome y. Cance Res. 2012,72, 645–654. [C ossRe ] [PubMed]
32.
Pinkham, K.; Pa k, D.J.; Hashemiaghdam, A.; Ki o , A.B.; Adam, I.; Rosiak, K.; da Ho a, C.C.; Teng, J.; Cheah, P.S.; Ca alho,
L.; e al. S ea oyl CoA Desa u ase Is Essen ial o Regula ion o Endoplasmic Re iculum Homeos asis and Tumo G ow h in
Glioblas oma Cance S em Cells. S em Cell Rep. 2019,12, 712–727. [C ossRe ] [PubMed]
33.
She galis, A.; Bankhead, A., 3 d; Luesakul, U.; Muangsin, N.; Neama i, N. Cu en Challenges and Oppo uni ies in T ea ing
Glioblas oma. Pha macol. Re . 2018,70, 412–445. [C ossRe ]
34.
Panagopoulos, A.T.; Gomes, R.N.; Almeida, F.G.; da Cos a Souza, F.; Veiga, J.C.E.; Nicolaou, A.; Colquhoun, A. The p os anoid
pa hway con ains po en ial p ognos ic ma ke s o glioblas oma. P os aglandins O he Lipid Media .
2018
,137, 52–62. [C ossRe ]
[PubMed]
35.
Ga a e, J.; Fe nández, R.; Lage, S.; Bes a d-Escalas, J.; Lopez, D.H.; Reigada, R.; Kho ami, S.; Gina d, D.; Reyes, J.; Amengual,
I.; e al. Imaging mass spec ome y inc eased esolu ion using 2-me cap obenzo hiazole and 2,5-diaminonaph alene ma ices:
Applica ion o lipid dis ibu ion in human colon. Anal. Bioanal. Chem. 2015,407, 4697–4708. [C ossRe ] [PubMed]
36.
Maimó-Ba celó, A.; Ga a e, J.; Bes a d-Escalas, J.; Fe nández, R.; Be hold, L.; Lopez, D.H.D.H.; Fe nández, J.A.J.A.;
Ba celó-Coblijn
, G. Con i ma ion o sub-cellula esolu ion using o e sampling imaging mass spec ome y. Anal. Bioanal. Chem.
2019,411, 7935–7941. [C ossRe ] [PubMed]
37.
Liu, C.; Sage, J.C.; Mille , M.R.; Ve haak, R.G.W.; Hippenmeye , S.; Vogel, H.; Fo eman, O.; B onson, R.T.; Nishiyama, A.; Luo,
L.; e al. Mosaic analysis wi h double ma ke s e eals umo cell o o igin in glioma. Cell 2011,146, 209–221. [C ossRe ]
38.
Dawson, M.R.L.; Poli o, A.; Le ine, J.M.; Reynolds, R. NG2-exp essing glial p ogeni o cells: An abundan and widesp ead
popula ion o cycling cells in he adul a CNS. Mol. Cell. Neu osci. 2003,24, 476–488. [C ossRe ]
39.
Zeng, A.; Hu, Q.; Liu, Y.; Wang, Z.; Cui, X.; Li, R.; Yan, W.; You, Y. IDH1/2 mu a ion s a us combined wi h Ki-67 labeling index
de ines dis inc p ognos ic g oups in glioma. Onco a ge 2015,6, 30232–30238. [C ossRe ]
40.
Ma, J.; Beni ez, J.A.; Li, J.; Miki, S.; Pon e de Albuque que, C.; Gala o, T.; O ellana, L.; Zanca, C.; Reed, R.; Boye , A.; e al.
Inhibi ion o Nuclea PTEN Ty osine Phospho yla ion Enhances Glioma Radia ion Sensi i i y h ough A enua ed DNA Repai .
Cance Cell 2019,35, 504–518.e7. [C ossRe ]
In . J. Mol. Sci. 2022,23, 2949 21 o 22
41.
S oyano , G.S.; Dzhenko , D.L.; Ki ano a, M.; Done , I.S.; Ghene , P. Co ela ion Be ween Ki-67 Index, Wo ld Heal h O ganiza ion
G ade and Pa ien Su i al in Glial Tumo s wi h As ocy ic Di e en ia ion. Cu eus 2017,9, e1396. [C ossRe ] [PubMed]
42. Sun, X.; Kau man, P.D. Ki-67: Mo e han a p oli e a ion ma ke . Ch omosoma 2018,127, 175–186. [C ossRe ] [PubMed]
43.
Ga a e, J.; Lage, S.; Ma ín-Saiz, L.; Pe ez-Valle, A.; Ochoa, B.; Boyano, M.D.; Fe nández, R.; Fe nández, J.A. In luence o Lipid
F agmen a ion in he Da a Analysis o Imaging Mass Spec ome y Expe imen s. J. Am. Soc. Mass Spec om.
2020
,31, 517–526.
[C ossRe ] [PubMed]
44. Ishizuka, I. Chemis y and unc ional dis ibu ion o sul oglycolipids. P og. Lipid Res. 1997,36, 245–319. [C ossRe ]
45. Snyde , F.; Snyde , F. E he Lipids Chemis y and Biology; Else ie : New Yo k, NY, USA, 1972; ISBN 9780126541502.
46.
Russo, P.S.T.; Fe ei a, G.R.; Ca dozo, L.E.; Bü ge , M.C.; A ias-Ca asco, R.; Ma uyama, S.R.; Hi a a, T.D.C.; Lima, D.S.; Passos,
F.M.; Fuku ani, K.F.; e al. CEMiTool: A Bioconduc o package o pe o ming comp ehensi e modula co-exp ession analyses.
BMC Bioin o m. 2018,19, 56. [C ossRe ]
47.
Ve haak, R.G.W.; Hoadley, K.A.; Pu dom, E.; Wang, V.; Qi, Y.; Wilke son, M.D.; Mille , C.R.; Ding, L.; Golub, T.; Mesi o , J.P.;
e al. In eg a ed Genomic Analysis Iden i ies Clinically Rele an Sub ypes o Glioblas oma Cha ac e ized by Abno mali ies in
PDGFRA, IDH1, EGFR, and NF1. Cance Cell 2010,17, 98–110. [C ossRe ]
48.
Goldman, M.J.; C a , B.; Has ie, M.; Repeˇcka, K.; McDade, F.; Kama h, A.; Bane jee, A.; Luo, Y.; Roge s, D.; B ooks, A.N.; e al.
Visualizing and in e p e ing cance genomics da a ia he Xena pla o m. Na . Bio echnol. 2020,38, 675–678. [C ossRe ]
49. Lee, S.Y. Temozolomide esis ance in glioblas oma mul i o me. Genes Dis. 2016,3, 198–210. [C ossRe ]
50.
Ha, S.; Showal e , G.; Cai, S.; Wang, H.; Liu, W.; Cohen-Gadol, A.; Sa ka ia, J.; Rickus, J.; Sp inge , J.; Adamec, J.; e al. Lipidomic
Analysis o Glioblas oma Mul i o me Using Mass Spec ome y. Cu . Me ab. 2014,2, 132–143. [C ossRe ]
51. Gimple, R.C.; Kidwell, R.L.; Kim, L.J.Y.; Sun, T.; G omo sky, A.D.; Wu, Q.; Wol , M.; L , D.; Bha ga a, S.; Jiang, L.; e al. Glioma
S em Cell–Speci ic Supe enhance P omo es Polyunsa u a ed Fa y-Acid Syn hesis o Suppo EGFR Signaling. Cance Disco .
2019,9, 1248–1267. [C ossRe ]
52.
Kambach, D.M.; Halim, A.S.; Gesine Caue , A.; Sun, Q.; T is an, C.A.; Celiku, O.; Kesa wala, A.H.; Shanka a am, U.; Ba chelo , E.;
S ommel, J.M. Disabled cell densi y sensing leads o dys egula ed choles e ol syn hesis in glioblas oma. Onco a ge
2017
,8, 14860.
[C ossRe ] [PubMed]
53.
Pa el, D.; Ahmad, F.; Kambach, D.M.; Sun, Q.; Halim, A.S.; K amp, T.; Camphausen, K.A.; S ommel, J.M. LXR
β
con ols
glioblas oma cell g ow h, lipid balance, and immune modula ion independen ly o ABCA1. Sci. Rep.
2019
,9, 15458. [C ossRe ]
[PubMed]
54.
Leb e o, P.; As udillo, A.M.; Rubio, J.M.; Fe nández-Caballe o, L.; Koko os, G.; Balboa, M.A.; Balsinde, J.; Fe nandez-Caballe o,
L.; Koko os, G.; Balboa, M.A.; e al. Cellula Plasmalogen Con en Does No In luence A achidonic Acid Le els o Dis ibu ion
in Mac ophages: A Role o Cy osolic Phospholipase A2gamma in Phospholipid Remodeling. Cells
2019
,8, 799. [C ossRe ]
[PubMed]
55.
Thomas, S.; Bye s, D.M.; Palme , F.B.S.; Spence, M.W.; Cook, H.W. Inco po a ion o polyunsa u a ed a y acids in o plasmalogens,
compa ed o o he phospholipids o cul u ed glioma cells, is mo e dependen on chain leng h han on selec i i y be ween (n
−
3)
and (n −6) amilies. Biochim. Biophys. Ac a-Lipids Lipid Me ab. 1990,1044, 349–356. [C ossRe ]
56.
Pende -Cudlip, M.C.; K ag, K.J.; Ma ini, D.; Yu, J.; Guidi, A.; Skinne , S.S.; Zhang, Y.; Qu, X.; He, C.; Xu, Y.; e al. Del a-6-
desa u ase ac i i y and a achidonic acid syn hesis a e inc eased in human b eas cance issue. Cance Sci.
2013
,104, 760–764.
[C ossRe ]
57.
Ka suki, H.; Akino, N.; Okuda, S.; Sai o, H. An ioxidan s, bu no cAMP o high K+, p e en a achidonic acid oxici y on neu onal
cul u es. Neu o epo 1995,6, 1101–1104. [C ossRe ]
58.
Palomba, L.; Ce ioni, L.; Can oni, O. A achidonic acid: A key molecule o as ocy e su i al o pe oxyni i e. Glia
2009
,57,
1672–1679. [C ossRe ]
59.
Comba, A.; Almada, L.L.; Tolosa, E.J.; Iguchi, E.; Ma ks, D.L.; Messle , M.V.; Sil a, R.; Fe nandez-Ba ena, M.G.; En iquez-Hesles,
E.; V abel, A.L.; e al. Nuclea ac o o ac i a ed T cells-dependen down egula ion o he ansc ip ion ac o glioma-associa ed
p o ein 1 (GLI1) unde lies he g ow h inhibi o y p ope ies o a achidonic acid. J. Biol. Chem. 2016,291, 1933–1947. [C ossRe ]
60.
Williams, J.R.; Lea e , H.A.; I onside, J.W.; Mille , E.P.; Whi le, I.R.; G ego , A. Apop osis in human p ima y b ain umou s:
Ac ions o a achidonic acid. P os aglandins Leuko . Essen . Fa . Acids 1998,58, 193–200. [C ossRe ]
61.
Bes a d-Escalas, J.; Reigada, R.; Reyes, J.; de la To e, P.; Liebisch, G.; Ba celó-Coblijn, G. Fa y Acid Unsa u a ion Deg ee o
Plasma Exosomes in Colo ec al Cance Pa ien s: A P omising Bioma ke . In . J. Mol. Sci. 2021,22, 5060. [C ossRe ]
62.
Chil on, F.H. Po en ial phospholipid sou ce(s) o a achidona e used o he syn hesis o leuko ienes by he human neu ophil.
Biochem. J. 1989,258, 327–333. [C ossRe ] [PubMed]
63.
Gil-de-Gómez, L.; As udillo, A.M.; Guijas, C.; Mag io i, V.; Koko os, G.; Balboa, M.A.; Balsinde, J. Cy osolic G oup IVA and
Calcium-Independen G oup VIA Phospholipase A 2 s Ac on Dis inc Phospholipid Pools in Zymosan-S imula ed Mouse
Pe i oneal Mac ophages. J. Immunol. 2014,192, 752–762. [C ossRe ] [PubMed]
64.
Yang, L.; Zhang, H. Exp ession o Cy osolic Phospholipase A2 Alpha in Glioblas oma Is Associa ed wi h Resis ance o Chemo he -
apy. Am. J. Med. Sci. 2018,356, 391–398. [C ossRe ] [PubMed]
65.
Höland, K.; Bolle , D.; Hagel, C.; Dolski, S.; T eszl, A.; Pa do, O.E.; ´
Cwiek, P.; Salm, F.; Leni, Z.; Shephe d, P.R.; e al. Ta ge ing class
Ia PI3K iso o ms selec i ely impai s cell g ow h, su i al, and mig a ion in glioblas oma. PLoS ONE
2014
,9, e94132. [C ossRe ]
In . J. Mol. Sci. 2022,23, 2949 22 o 22
66.
Holland, E.C.; Celes ino, J.; Dai, C.; Schae e , L.; Sawaya, R.E.; Fulle , G.N. Combined ac i a ion o Ras and Ak in neu al
p ogeni o s induces glioblas oma o ma ion in mice. Na . Gene . 2000,25, 55–57. [C ossRe ]
67.
Shulga, Y.V.; Ande son, R.A.; Topham, M.K.; Epand, R.M. Phospha idylinosi ol-4-phospha e 5-Kinase Iso o ms Exhibi Acyl
Chain Selec i i y o Bo h Subs a e and Lipid Ac i a o . J. Biol. Chem. 2012,287, 35953–35963. [C ossRe ]
68. Eckha d , M. The ole and me abolism o sul a ide in he ne ous sys em. Mol. Neu obiol. 2008,37, 93–103. [C ossRe ]
69.
He, H.; Nilsson, C.L.; Emme , M.R.; Ji, Y.; Ma shall, A.G.; K oes, R.A.; Moskal, J.R.; Colman, H.; Lang, F.F.; Con ad, C.A.
Pola lipid emodeling and inc eased sul a ide exp ession a e associa ed wi h he glioma he apeu ic candida es, wild ype p53
ele a ion and he opoisome ase-1 inhibi o , i ino ecan. Glycoconj. J. 2010,27, 27–38. [C ossRe ]
70.
Lee, E.; Yong, R.L.; Paddison, P.; Zhu, J. Compa ison o glioblas oma (GBM) molecula classi ica ion me hods. Semin. Cance Biol.
2018,53, 201–211. [C ossRe ]
71.
Zhang, P.; Xia, Q.; Liu, L.; Li, S.; Dong, L. Cu en Opinion on Molecula Cha ac e iza ion o GBM Classi ica ion in Guiding
Clinical Diagnosis, P ognosis, and The apy. F on . Mol. Biosci. 2020,7, 241. [C ossRe ]
72.
Tu can, S.; Rohle, D.; Goenka, A.; Walsh, L.A.; Fang, F.; Yilmaz, E.; Campos, C.; Fabius, A.W.M.; Lu, C.; Wa d, P.S.; e al. IDH1
mu a ion is su icien o es ablish he glioma hype me hyla o pheno ype. Na u e 2012,483, 479–483. [C ossRe ] [PubMed]
73.
B ennan, C.W.; Ve haak, R.G.W.; McKenna, A.; Campos, B.; Noushmeh , H.; Salama, S.R.; Zheng, S.; Chak a a y, D.; Sanbo n,
J.Z.; Be man, S.H.; e al. The soma ic genomic landscape o glioblas oma. Cell 2013,155, 462–477. [C ossRe ] [PubMed]
74.
Cecca elli, M.; Ba hel, F.P.; Mal a, T.M.; Sabedo , T.S.; Salama, S.R.; Mu ay, B.A.; Mo ozo a, O.; New on, Y.; Radenbaugh, A.;
Pagno a, S.M.; e al. Molecula P o iling Re eals Biologically Disc e e Subse s and Pa hways o P og ession in Di use Glioma.
Cell 2016,164, 550–563. [C ossRe ]
75.
Shwe ha, S.D.; Shas y, A.H.; A i azhagan, A.; San osh, V. Manganese supe oxide dismu ase (MnSOD) is a malignan as ocy oma
speci ic bioma ke and associa ed wi h ad e se p ognosis in p53 exp essing glioblas oma. Pa hol.-Res. P ac .
2016
,212, 17–23.
[C ossRe ] [PubMed]
76.
Behnan, J.; Finocchia o, G.; Hanna, G. The landscape o he mesenchymal signa u e in b ain umou s. B ain
2019
,142, 847–866.
[C ossRe ]
77.
Pasquale i, F.; Mon emu o, N.; Deside i, I.; Loi, M.; Giannini, N.; Gadducci, G.; Mal a i, G.; Can a ella, M.; Gonnelli, A.;
Mon one, S.; e al. Impac o ecu ence pa e n in pa ien s unde going a second su ge y o ecu en glioblas oma. Ac a Neu ol.
Belg. 2021,65, 1–6. [C ossRe ]
78.
Kim, Y.; Va n, F.S.; Pa k, S.-H.; Yoon, B.W.; Pa k, H.R.; Lee, C.; Ve haak, R.G.W.; Paek, S.H. Pe spec i e o mesenchymal
ans o ma ion in glioblas oma. Ac a Neu opa hol. Commun. 2021,9, 50. [C ossRe ]
79.
B un, M.; Coles, J.E.; Monck on, E.A.; Glub ech , D.D.; Bisg o e, D.; Godbou , R. Nuclea Fac o I Regula es B ain Fa y Acid-
Binding P o ein and Glial Fib illa y Acidic P o ein Gene Exp ession in Malignan Glioma Cell Lines. J. Mol. Biol.
2009
,391,
282–300. [C ossRe ]
80.
Elshe biny, M.E.; Ema a, M.; Godbou , R. In e ac ion o b ain a y acid-binding p o ein wi h he polyunsa u a ed a y acid
en i onmen as a po en ial de e minan o poo p ognosis in malignan glioma. P og. Lipid Res. 2013,52, 562–570. [C ossRe ]
81.
Elshe biny, M.E.; Chen, H.; Ema a, M.; Godbou , R.
ω
-3 and
ω
-6 a y acids modula e con en ional and a ypical p o ein kinase C
ac i i ies in a b ain a y acid binding p o ein dependen manne in glioblas oma mul i o me. Nu ien s
2018
,10, 454. [C ossRe ]
82.
Bha , K.P.L.; Balasub amaniyan, V.; Vaillan , B.; Ezhila asan, R.; Hummelink, K.; Hollingswo h, F.; Wani, K.; Hea hcock, L.; James,
J.D.; Goodman, L.D.; e al. Mesenchymal Di e en ia ion Media ed by NF-
κ
B P omo es Radia ion Resis ance in Glioblas oma.
Cance Cell 2013,24, 331–346. [C ossRe ] [PubMed]
83.
Ga nie , D.; Renoul , O.; Al es-Gue a, M.-C.; Pa is, F.; Pecqueu , C. Glioblas oma S em-Like Cells, Me abolic S a egy o Kill a
Challenging Ta ge . F on . Oncol. 2019,9, 118. [C ossRe ] [PubMed]
84.
Fe nandez, R.; Ga a e, J.; Ma in-Saiz, L.; Gale ich, I.; Fe nandez, J.A.; Fe nández, R.; Ga a e, J.; Ma ín-Saiz, L.; Gale ich, I.;
Fe nández, J.A. Ma ix Sublima ion De ice o MALDI Mass Spec ome y Imaging. Anal. Chem.
2019
,91, 803–807. [C ossRe ]
[PubMed]
85.
Huang, D.W.; She man, B.T.; Lempicki, R.A. Bioin o ma ics en ichmen ools: Pa hs owa d he comp ehensi e unc ional analysis
o la ge gene lis s. Nucleic Acids Res. 2009,37, 1–13. [C ossRe ] [PubMed]
86.
Demša , J.; Cu k, T.; E ja ec, A.; Go up, ˇ
C.; Hoˇce a , T.; Milu ino iˇc, M.; Možina, M.; Polajna , M.; Toplak, M.; S a iˇc, A.; e al.
O ange: Da a Mining Toolbox in Py hon. J. Mach. Lea n. Res. 2013,14, 2349–2353.
