The stability of the metabolic turnover of arachidonic acid in human unruptured intracranial aneurysmal walls is sustained

Abstract

Objective
Intracranial aneurysm (IA) is considered a chronic inflammatory condition that affects intracranial arteries. Cyclooxygenase 2 (COX2) and prostaglandin E2 (PGE2) are considered potential targets of specific medical treatment for IAs. Previous studies have reported the elevated COX2 expression in the IA wall. However, not much has been studied about the upstream regulation of COX2 and PGE2, and the metabolism of arachidonic acid (AA) in human IAs. In this study, we aimed to elucidate the distribution of fatty acids in human IA walls for the first time.

Methods
Samples from 6 ruptured and 5 unruptured human IAs were surgically resected after the aneurysmal clipping and analyzed using desorption electrospray ionization imaging mass spectrometry.

Results
AA and AA-containing phospholipids were not detected in the unruptured IA walls. On the contrast, significantly larger amounts of AA and AA-containing phospholipids were detected in the ruptured IA walls compared to unruptured IA walls.

Conclusions
This study showed for the first time that AA was not detected in unruptured human IA walls. Our findings suggest that the stability of the turnover of AA in human unruptured IA walls is sustained. In contrast, this study showed that larger amounts of AA and AA-containing phospholipids were detected in the ruptured IA walls. More cases and further analysis are necessary to interpret our present results.

Full text

Clinical Neurology and Neurosurgery 208 (2021) 106881
Available online 8 August 2021
0303-8467/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
The stability of the metabolic turnover of arachidonic acid in human
unruptured intracranial aneurysmal walls is sustained
Ririko Takeda
a
,
b
,
*
,
1
, Ariful Islam
c
,
1
, Tomohito Sato
c
,
d
, Hiroki Kurita
b
, Tomoaki Kahyo
c
,
d
,
Tetsumei Urano
e
, Mitsutoshi Setou
c
,
d
,
f
a
Department of Neurosurgery, Teikyo University Hospital, Mizonokuchi, Kawasaki, Japan
b
Department of Cerebrovascular Surgery, International Medical Center, Saitama Medical University, Hidaka, Japan
c
Department of Cellular and Molecular Anatomy, Hamamatsu University School of Medicine, Hamamatsu, Japan
d
International Mass Imaging Center, Hamamatsu University School of Medicine, Hamamatsu, Japan
e
Department of Physiology, Hamamatsu University School of Medicine, Hamamatsu, Japan
f
Department of Systems Molecular Anatomy, Institute for Medical Photonics Research, Preeminent Medical Photonics Education & Research Center, Hamamatsu, Japan
ARTICLE INFO
Keywords:
Intracranial aneurysm
Imaging mass spectroscopy
Arachidonic acid
Phospholipid
ABSTRACT
Objective: Intracranial aneurysm (IA) is considered a chronic inammatory condition that affects intracranial
arteries. Cyclooxygenase 2 (COX2) and prostaglandin E2 (PGE2) are considered potential targets of specic
medical treatment for IAs. Previous studies have reported the elevated COX2 expression in the IA wall. However,
not much has been studied about the upstream regulation of COX2 and PGE2, and the metabolism of arachidonic
acid (AA) in human IAs. In this study, we aimed to elucidate the distribution of fatty acids in human IA walls for
the rst time.
Methods: Samples from 6 ruptured and 5 unruptured human IAs were surgically resected after the aneurysmal
clipping and analyzed using desorption electrospray ionization imaging mass spectrometry.
Results: AA and AA-containing phospholipids were not detected in the unruptured IA walls. On the contrast,
signicantly larger amounts of AA and AA-containing phospholipids were detected in the ruptured IA walls
compared to unruptured IA walls.
Conclusions: This study showed for the rst time that AA was not detected in unruptured human IA walls. Our
ndings suggest that the stability of the turnover of AA in human unruptured IA walls is sustained. In contrast,
this study showed that larger amounts of AA and AA-containing phospholipids were detected in the ruptured IA
walls. More cases and further analysis are necessary to interpret our present results.
1. Introduction
Intracranial aneurysm (IA) is currently considered a chronic in-
ammatory disease that affects the intracranial arteries [1]. As an un-
derlying mechanism, the prostaglandin E2 (PGE2)-EP2-NF-κB signaling
cascade in macrophages was considered by Aoki et al. as a factor regu-
lating such chronic inammation involved in the progression and
rupture of IAs [1]. PGE2 is a lipid mediator and an arachidonic acid (AA)
metabolite generated from sequential enzymatic reactions, including
cyclooxygenase (COX), and a potential therapeutic target for IAs [1].
Increased levels of free fatty acids (FFAs), including AA, in several brain
regions after subarachnoid hemorrhage were reported in a previous
study using a rat model [2]. The association of AA-containing phos-
pholipids in the development and rupture of IA walls in a rat model were
also reported in a previous study [3]. However, to date, no report has
described the distribution of FFAs in the walls of IA in humans.
Therefore, we investigated the extent of FFA distribution in the walls
of IAs in human subjects by using a recently developed imaging mass
spectrometry (IMS) technique, desorption electrospray ionization IMS
(DESI-IMS).
* Correspondence to: Department of Neurosurgery, Teikyo University Hospital, 5-1-1 futako, Takatsu-ku, Mizonokuchi, Kawasaki, Kanagawa 213-8507, Japan.
E-mail address: rtakeda@med.teikyo-u.ac.jp (R. Takeda).
1
These authors have equal contribution.
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery
journal homepage: www.elsevier.com/locate/clineuro
https://doi.org/10.1016/j.clineuro.2021.106881
Received 8 January 2021; Received in revised form 3 August 2021; Accepted 4 August 2021

Clinical Neurology and Neurosurgery 208 (2021) 106881
2
2. Materials and methods
2.1. Specimen collection and preparation
We included in this study a series of 11 patients with 5 unruptured
and 6 ruptured saccular IAs who underwent surgery in the Department
of Cerebrovascular Surgery of Saitama Medical University International
Medical Center in Hidaka, Japan (supplementary Table 1). The IA
samples were obtained by resecting a small part of the aneurysmal wall
intraoperatively after the clipping of the aneurysmal neck. The speci-
mens were immediately frozen on dry ice and stored at −80 ◦C until
sectioning. All the specimens were sectioned at a 10 µm thickness using a
cryostat for DESI-IMS analysis.
2.2. DESI-IMS data acquisition from the IAs and standard lipids
Sectioned slides were kept at room temperature just before DESI-MS
acquisition. Mass spectra were acquired in the negative ion mode. All
the experiments were performed with the DESI source attached to a
quadrupole time-of-ight (Q-TOF) mass spectrometer (Xevo G2-XS Q-
TOF; Waters, Milford, MA, USA). The mass spectra were calibrated
externally prior to measurement using a sodium formate solution (500
µM) in 90:10 2-propanol-to-water ratio (v/v). The parameters used for
the optimization of DESI and acquisition of data from IA tissues are
given in supplementary Table 2. Tandem mass spectrometry (MS/MS)
using same instrument was performed to conrm the candidate mole-
cule corresponded to m/z 303.23 applying a collision energy of 10 eV,
source temperature 120º C, capillary voltage 4.0 kV, and 98% methanol
(98:2; methanol: water) as spray solvent at a ow rate of 2 µL/min. Two
standard FFAs (arachidonic acid and oleic acid) and those fatty acids
(FAs) containing standard lipids were also measured using the same MS
instrument and parameters (supplementary Table 2) to conrm whether
FAs were detected in IA walls as FFAs or fragment of lipids. For that
purpose, 1 µL solution of standard lipids (10 µg/mL in ethanol) was
applied on glass slide and acquired DESI-MS data.
2.3. Fold changes in FAs and lipid contents between ruptured and
unruptured IA walls
Fold changes (rupture IA vs unruptured IA walls) of detected FAs and
lipids were also analyzed using the average intensity of each FA and
lipid. In mass spectrometric measurement, noise is a common fact which
can hide MS peaks with small intensities. Therefore, intensities of noise
nearby the MS peaks of candidate lipids which were not detected in one
group of IA samples were used as their intensities and calculated their
fold changes [4].
2.4. Data analyses
The MassLynx 4.1 (Waters) software was used for data acquisition
and processing; The HDImaging (Waters) software was used for the
image analysis. The MS Excel and SPSS version 16 software were used
for the statistical analysis. All the values are expressed as mean ±
standard error of mean (SEM). Differences with P values <0.05 (two-
tailed t-test) were considered signicant.
2.5. Chemical and reagents
Liquid chromatography/MS-grade methanol, 2-propanol, and ultra-
pure water were purchased from Wako Pure Chemical Industries (Osaka,
Japan); leucine enkephalin was purchased from Waters (Germany); and
sodium formate, AA, oleic acid (OA), phosphatidylcholine (PC; 18:1/
18:1), and PC (20:4/20:4) were purchased from Sigma-Aldrich (St.
Louis, MO, USA).
2.6. Ethics committee approval
The institutional review board of Saitama Medical University Inter-
national Medical Center (No. 16–223) and Hamamatsu University
School of Medicine (No. 17–288) approved all aspects of the study.
Informed consent was obtained from all the patients.
3. Results
3.1. Distribution of FFAs in the IA walls on DESI-IMS
We rst analyzed the DESI-IMS mass spectra acquired from ruptured
and unruptured IA walls. Five MS peaks with m/z of 255.23, 279.23,
281.25, 283.26, and 303.23 were identied in the samples, corre-
sponding to palmitic acid (PA; C16:0), linoleic acid (LA; C18:2), OA
(C18:2), stearic acid (SA; C18:0), and AA (C20:4), respectively (sup-
plementary Fig. 1 and Supplementary Table 3) according to previous
reports [5,6]. AA was further conrmed by MS/MS analysis (Supple-
mentary Fig. 2). Using standard lipids, we have also conrmed that all
these FAs were detected as FFAs in this study (Supplementary Fig. 3).
The distribution patterns of PA, LA, OA, SA, and AA were analyzed in
both the ruptured and unruptured IA tissues (Fig. 1). AA was detected
only in the ruptured but not in the unruptured IA walls (Fig. 1A and B).
Moreover, accumulations of LA and AA in the same tissue region as the
ruptured IA walls were also observed (Fig. 1A and B). No signicant
change in the distribution of PA and SA was found between ruptured and
unruptured IA walls.
The average intensity of the distribution of FFAs were also analyzed
in this study. The distribution of LA and OA were increased in the
ruptured IA walls by 3.48-fold (P =0.003) and 2.85-fold (P =0.013),
respectively, as compared with the unruptured IA walls (Fig. 2A). In the
unruptured IAs, the amount of AA in the sample was less than the
detection sensitivity of the device, so it was not detected as MS peak
corresponding to AA. Compared with the noise levels in the unruptured
IA walls, an increase in AA level of approximately 300 times was
detected in the ruptured IA walls (Fig. 2B).
3.2. Distribution of AA-containing phospholipids in the IA walls on DESI-
IMS
We next analyzed the distribution of AA-containing phospholipids in
the IA walls using DESI-IMS. Four MS peaks with m/z of 766.53, 794.55,
810.53, and 885.55 were detected in the ruptured IA walls and assigned
to phosphatidylethanolamine (PE; 18:0/20:4), PE (20:0/20:4), phos-
phatidylserine (PS; 18:0/20:4), and phosphatidylinositol (PI; 18:0/
20:4), respectively, on the basis of their mass accuracy (Supplementary
Table 3), biological distributions, and data from previous reports [3,7].
Among these phospholipids, PE (20:0/20:4) and PI (18:0/20:4) were
observed in abundance especially in the walls of the ruptured IAs
(Fig. 3A). In the unruptured IAs, the amounts of AA containing phos-
pholipids in the sample were less than the detection sensitivity of the
device, so it was not detected as a peak corresponding to the phospho-
lipids (Fig. 3B). Compared with the noise levels of DESI-IMS data ac-
quired from unruptured IA walls, the intensity level of the distributions
of PE (18:0/20:4), PE (20:0/20:4), PS (18:0/20:4), and PI (18:0/20:4) in
the ruptured IA walls were increased by approximately 2.0-fold,
3.7-fold, 3.5-fold, and 5.5-fold, respectively (Fig. 3C).
The changes in the distribution of FFAs and phospholipids between
ruptured and unruptured IA walls were given by Supplementary Table 4.
3.3. Inspections of the IA walls by the immunohistology
We tried to examine the IA walls by immunohistochemical staining
method. The sectioned slides of the samples were stained by
hematoxylin-eosin (HE) staining and immunostaining including COX2.
It was difcult to interpret the results of the immunostaining because the
R. Takeda et al.

Clinical Neurology and Neurosurgery 208 (2021) 106881
3
condition of the slides was inappropriate for the immunostaining. Under
the HE staining, many blood cells were detected on ruptured IA walls
different from unruptured IA walls (Fig. 1).
4. Discussion
This study showed for the rst time that AA was not detected in the
unruptured human IA walls. Our ndings suggest that the stability of the
turnover of AA in human unruptured IA walls is sustained. Previous
studies have reported the elevated COX2 expression in the IA wall [7,8].
Additionally, IA is considered a chronic inammatory disease and is
conceptually the same as other chronic inammatory diseases such as
cancer, atherosclerosis, and impaired glucose tolerance [1]. Our nd-
ings may suggest that the inammatory responses in human IAs,
Fig. 1. Distribution of free fatty acids (FFAs) in the walls of ruptured (R) and unruptured (U) human intracranial aneurysms (IAs). A
,
Molecular ion images of FFAs in
the ruptured IA human walls. B
,
Molecular ion images of the FFAs in the unruptured human IA walls. ND indicates not detected. Scale bar: 0.5 mm. H&E:
hematoxylin-eosin. Palmitic acid (PA), linoleic acid (LA), oleic acid (OA) and stearic acid (SA) were detected in both ruptured and unruptured human IA walls, but
arachidonic acid (AA) was only detected in ruptured human IA walls.
Fig. 2. Differences in the intensity level of the distribution of free fatty acids (FFAs) between the ruptured and unruptured human intracranial aneurysm (IA) walls.
A, Average intensity of FFAs in the walls of ruptured and unruptured IAs walls (n =6 for ruptured IA and n =5 for unruptured IA). All data are presented as mean
±SEM. * *P =0.003 and *P =0.013 (two-tailed t-test). Signicantly increased distribution of oleic acid (OA) and linoleic acid (LA) in ruptured human IA walls
compared to that of unruptured human IA walls was noted in this study. No signicance change was found in the distribution of palmitic acid (PA) and stearic acid
(SA) between ruptured and unruptured IA walls. B
,
Ratio of the intensity of palmitic acid (PA) and arachidonic acid (AA) between the ruptured and unruptured IA
walls. The noise level in the unruptured IA walls was used as the intensity of AA in the unruptured IA walls. Compared to the unruptured IA walls, about 300-fold
increased AA was observed in ruptured IA walls.
R. Takeda et al.

Clinical Neurology and Neurosurgery 208 (2021) 106881
4
including the COX2 expression, is not induced from the stimulation that
AA regulates independent of its metabolites. Furthermore, considering
that AA has been detected enough and suggested to play an important
role in other inammatory disease [9–14], our ndings might suggest
that unruptured human IAs have a different inammatory association
from other inammatory diseases regarding the response of AA and
COX2.
Unlike the unruptured IA walls, high levels of AA were detected on
the ruptured IA walls. We also found high level of LA, precursor in AA
synthesis, in the ruptured IA walls, and accumulations of LA and AA in
the same tissue region as the ruptured IA walls were also observed.
Considering that not only AA but also the precursor of AA increased in
the ruptured IA, it may be possible that AA is associated to the pro-
gression and rupture of IA walls. However, according to the result that
many blood cells were also detected on ruptured IA walls by
hematoxylin-eosin staining, it seems reasonable to think that much of
the AA in the ruptured IAs is free AA caused by the platelets activated by
the rupture of the aneurysms. If that is the case, this nding suggests that
DESI-IMS is a useful tool for semi-quantication of the activation of
platelets in excised specimens. In this study, we also found high level of
OA. We can’t have interpreted this result yet. The signicance of OA was
smaller than that of AA and LA, and more cases are necessary to interpret
this result.
We have previously demonstrated that PI (18:0/20.4) accumulated
at high levels in the thickened aneurysmal walls in experimentally
induced IA with synthetic dedifferentiated smooth muscle cells [3]. It is
uncertain why the current study demonstrated AA-containing phos-
pholipids including PI (18:0/20.4) at the noise levels in the human
unruptured IA walls. It might be due to the difference of the wall of the
aneurysms in each sample; the difference of the species and experi-
mentally induced or not. On the other hand, high levels of
AA-containing phospholipids including PI (18:0/20.4) were detected on
ruptured IA walls. PI (18:0/20:4) play key roles in a wide range of
cellular processes, including cell migration, invasion, and proliferation
[3]. PE also play important roles in the regulation of cell proliferation,
metabolism, organelle function, endocytosis, autophagy, stress
response, and apoptosis [15]. Therefore, our results might be a clue for
exploring the mechanism of rupture and hemostasis in IAs.
The sample size in this study was too small, and the human IA
samples, especially those from ruptured IAs, might have been already
modied by several factors, which might have affected the results of the
present study. Therefore, more cases and further analysis are necessary
to interpret our present results. However, our ndings might offer new
insights into the associations of AA metabolites, including COX2, with IA
walls in humans.
Fig. 3. Distribution of arachidonic acid (AA)-containing phospholipids in the walls of human intracranial aneurysm (IA)s. A, Molecular ion images of AA-containing
phospholipids in ruptured human IA walls. B, Molecular ion images of AA-containing phospholipids in unruptured human IA walls. ND indicates not detected. AA
containing phospholipids were only detected I ruptured IA walls. Scale bar: 0.5 mm. C, Ratio of the intensity of the AA-containing phospholipids between the
ruptured and unruptured IA walls. The noise level in the unruptured IA walls was used as the intensity of AA-containing phospholipids in the unruptured IA walls.
About 2–6 folds increased AA-containing phospholipids were detected in ruptured IA walls compared to that of unruptured IA walls. PE: Phosphatidylethanolamine;
PS: Phosphatidylserine; PI: Phosphatidylinositol.
R. Takeda et al.

Clinical Neurology and Neurosurgery 208 (2021) 106881
5
5. Conclusion
In this study, DESI-IMS revealed the changes in the distribution of
FFAs and lipids in the ruptured and unruptured IA walls in human
samples. We found that AA and AA-containing phospholipids were not
detected in the unruptured IA while they were detected in a signicantly
larger amount in the ruptured IA. These ndings may offer new insights
on the association of AA metabolites, including COX2, with IA walls in
humans.
Funding
This work was supported by JSPS KAKENHI Grant Numbers
17K10849 (to Dr. Takeda), 21K09188 (to Dr. Takeda), JP15H05898B1
(to Dr. Setou), and AMED Grant Number JP19gm0910004 (to Dr.
Setou).
CRediT authorship contribution statement
Ririko Takeda: Conceptualization, Methodology, Resources,
Writing – original draft, Writing – review & editing, Project adminis-
tration. Ariful Islam: Investigation, Formal analysis, Data curation,
Writing – review & editing. Tomohito Sato: Validation, Formal anal-
ysis, Investigation, Data curation. Hiroki Kurita: Resources, Supervi-
sion. Tomoaki Kahyo: Validation, Investigation. Tetsumei Urano:
Supervision. Mitsutoshi Setou: Methodology, Project administration.
Declarations of interest
None.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.clineuro.2021.106881.
References
[1] T. Aoki, J. Frȍsen, M. Fukuda, K. Bando, G. Shioi, K. Tsuji, E. Ollikainen, K. Nozaki,
J. Laakkonen, S. Narumiya, Prostaglandin E2-EP2-NF-κB signaling in macrophages
as a potential therapeutic target for intracranial aneurysms, Sci. Signal. 10 (2017)
eaah6037, https://doi.org/10.1126/scisignal.aah6037.
[2] R.J. Gewirtz, H.S. Dhillon, S.E. Goes, S.M. DeAtley, S.W. Scheff, Lactate and free
fatty acids after subarachnoid hemorrhage, Brain Res. 840 (1999) 84–91, https://
doi.org/10.1016/s0006-8993(99)01752-7.
[3] T. Ikedo, M. Minami, H. Kataoka, K. Hayashi, M. Nagata, R. Fujikawa, F. Yamazaki,
M. Setou, M. Yokode, S. Miyamoto, Imaging mass spectroscopy delineates the
thinned and thickened walls of intracranial aneurysms, Biochem Biophys. Res
Commun. 495 (2018) 332–338, https://doi.org/10.1016/j.bbrc.2017.10.133.
[4] J. Kast, M. Gentzel, M. Wilm, K. Richardson, Noise ltering techniques for
electrospray quadrupole time of ight mass spectra, J. Am. Soc. Mass Spectrom. 14
(7) (2003) 766–776, https://doi.org/10.1016/S1044-0305(03)00264-2.
[5] M. Mamun, S. Sato, E. Naru, O. Sakata, E. Hoshikawa, A. Suzuki, A. Islam,
T. Kahyo, T. Sato, T.K. Ito, M. Horikawa, R. Fukui, K. Izumi, M. Setou, Higher
Accumulation of Docosahexaenoic Acid in the Vermilion of the Human Lip than in
the Skin, Int. J. Mol. Sci. 21 (2020) 2807, https://doi.org/10.3390/ijms21082807.
[6] E. Takeyama, A. Islam, N. Watanabe, H. Tsubaki, M. Fukushima, M. Mamun,
S. Sato, T. Sato, F. Eto, I. Yao, T.K. Ito, M. Horikawa, M. Setou, Dietary intake of
green nut oil or DHA ameliorates DHA distribution in the brain of a mouse model of
dementia accompanied by memory recovery, Nutrients 11 (2019) 2371, https://
doi.org/10.3390/nu11102371.
[7] J.S. Hudson, A.J. Marincovich, J.A. Roa, M. Zanaty, E.A. Samaniego, D.M. Hasan,
Aspirin and intracranial aneurysms: a critical review of the current evidence,
Stroke 50 (2019) 2591–2596, https://doi.org/10.1161/STROKEAHA.119.026094.
[8] J. Rodemerk, A. Junker, B. Chen, D. Pierscianek, P. Dammann, M. Darkwah
Oppong, A. Radbruch, M. Forsting, S. Maderwald, H.H. Quick, Y. Zhu, R. Jabbarli,
U. Sure, K.H. Wrede, Pathophysiology of intracranial aneurysms: COX-2
expression, iron deposition in aneurysm wall, and correlation with magnetic
resonance imaging, Stroke 51 (2020) 2505–2513, https://doi.org/10.1161/
STROKEAHA.120.030590.
[9] T. Hiraide, K. Ikegami, T. Sakaguchi, Y. Morita, T. Hayasaka, N. Masaki, M. Waki,
E. Sugiyama, S. Shinriki, M. Takeda, Y. Shibasaki, S. Miyazaki, H. Kikuchi,
H. Okuyama, M. Inoue, M. Setou, H. Konno, Accumulation of arachidonic acid-
containing phosphatidylinositol at the outer edge of colorectal cancer, Sci. Rep. 6
(2016) 29935, https://doi.org/10.1038/srep29935.
[10] M. Hughes-Fulford, C.F. Li, J. Boonyaratanakornkit, S. Sayyah, Arachidonic acid
activates phosphatidylinositol 3-kinase signaling and induces gene expression in
prostate cancer, Cancer Res. 66 (2006) 1427–1433, https://doi.org/10.1158/
0008-5472.CAN-05-0914.
[11] J. Qin, W. Wang, R. Zhang, Novel natural product therapeutics targeting both
inammation and cancer, Chin. J. Nat. Med. 15 (2017) 401–416, https://doi.org/
10.1016/S1875-5364(17)30062-6.
[12] N.K.V. Gundala, V.G.M. Naidu, U.N. Das, Amelioration of streptozotocin-induced
type 2 diabetes mellitus in Wistar rats by arachidonic acid, Biochem Biophys. Res.
Commun. 496 (2018) 105–113, https://doi.org/10.1016/j.bbrc.2018.01.007.
[13] F. Cipollone, M.L. Fazia, COX-2 and atherosclerosis, J. Cardiovasc Pharmacol. 47
(Suppl 1) (2006) S26–S36, https://doi.org/10.1097/00005344-200605001-00006.
[14] T. Sato, M. Horikawa, S. Takei, F. Yamazaki, T.K. Ito, T. Kondo, T. Sakurai,
T. Kahyo, K. Ikegami, S. Sato, R. Sato, Y. Jinno, H. Kawano, S. Naoe, M. Arita,
Y. Kashiwagi, M. Setou, Preferential incorporation of administered
eicosapentaenoic acid into thin-cap atherosclerotic plaques, Arterioscler. Thromb.
Vasc. Biol. 39 (2019) 1802–1816, https://doi.org/10.1161/
ATVBAHA.119.313093.
[15] Y. Chen, Z. Ma, X. Shen, L. Li, J. Zhong, L.S. Min, L. Xu, H. Li, J. Zhang, L. Dai,
Serum lipidomics proling to identify biomarkers for non-small cell lung cancer,
Biomed. Res. Int. 2018 (2018), 5276240, https://doi.org/10.1155/2018/5276240
eCollection 2018.
R. Takeda et al.