Identification and Quantitation of C=C Location Isomers of Unsaturated Fatty Acids by Epoxidation Reaction and Tandem Mass Spectrometry

Article abstract of DOI:10.1021/acs.analchem.7b01870

Unsaturated fatty acids (FAs) serve as nutrients, energy sources, and signaling molecules for organisms, which are the major components for a large variety of lipids. However, structural characterization and quantitation of unsaturated FAs by mass spectrometry remain an analytical challenge. Here, we report the coupling of epoxidation reaction of the C=C in unsaturated FAs and tandem mass spectrometry (MS) for rapid and accurate identification and quantitation of C=C isomers of FAs in a shotgun lipidomics approach. Epoxidation of the C=C leads to the production of an epoxide which, upon collision induced dissociation (CID), produces abundant diagnostic ions indicative of the C=C location. The total intensity of the same set of diagnostic ions for one specific FA C=C isomer was also used for its relative and absolute quantitation. The simple experimental setup, rapid reaction kinetics (<2 min), high reaction yield (>90% for monounsaturated FAs), and easy-to-interpret tandem MS spectra enable a promising methodology particularly for the analysis of unsaturated FAs in complex biological samples such as human plasma and animal tissues.

Full text of DOI:10.1021/acs.analchem.7b01870

Subscriber access provided by TULANE UNIVERSITY
Article
Identification and quantitation of C=C location isomers of unsaturated
fatty acids by epoxidation reaction and tandem mass spectrometry
Yaoyao Zhao, Hansen Zhao, Xu Zhao, Jia Jia, Qiang Ma, Sichun
Zhang, Xinrong Zhang, Hitoshi Chiba, Shu-Ping Hui, and Xiaoxiao Ma
Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b01870 • Publication Date (Web): 24 Aug 2017
Downloaded from http://pubs.acs.org on August 24, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 20
Analytical Chemistry
1
2
3
4
5
6
7
8
9
Identification and quantitation of C=C location isomers of
unsaturated fatty acids by epoxidation reaction and tandem
mass spectrometry
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Yaoyao Zhao1, Hansen Zhao2, Xu Zhao2, Jia Jia2, Qiang Ma3, Sichun Zhang2, Xinrong Zhang2,
Hitoshi Chiba1, ShuꢀPing Hui*1, and Xiaoxiao Ma*4
1. Graduate School of Health Science, Hokkaido University, North 12, West 5, Kitaꢀku, Sapporo,
060ꢀ0812, Japan
2. Department of Chemistry, Tsinghua University, Beijing 100084, P.R. China
3. Institute of Industrial and Consumer Product Safety, Chinese Academy of Inspection and
Quarantine (CAIQ), No. 11, Ronghua South Road, Beijing 100176, China
4. Department of Precision Instrument, Tsinghua University, Beijing 100084, P.R. China
To whom correspondence should be addressed:
Prof. ShuꢀPing Hui, keino@hs.hokudai.ac.jp
Prof. Xiaoxiao Ma, maxx@tsinghua.edu.cn
ACS Paragon Plus Environment

Analytical Chemistry
Page 2 of 20
1
2
3
4
5
6
7
8
Abstract
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Unsaturated fatty acids (FAs) serve as nutrients, energy sources and signaling molecules for
organisms, which are the major components for a large variety of lipids. However, structural
characterization and quantitation of unsaturated FAs by mass spectrometry remain an analytical
challenge. Here we report the coupling of epoxidation reaction of the C=C in unsaturated FAs and
tandem mass spectrometry (MS) for rapid and accurate identification and quantitation of C=C
isomers of FAs in a shotgun lipidomics approach. Epoxidation of the C=C leads to the production of
an epoxide which, upon collision induced dissociation (CID), produces abundant diagnostic ions
indicative of the C=C location. The total intensity of the same set of diagnostic ions for one specific
FA C=C isomer was also used for its relative and absolute quantitation. The simple experimental
setup, rapid reaction kinetics (< 2 min), high reaction yield (>90% for monounsaturated FAs), and
easyꢀtoꢀinterpret tandem MS spectra enable a promising methodology particularly for the analysis of
unsaturated FAs in complex biological samples, such as human plasma and animal tissues.
Introduction
As a key component of almost all lipids, including phospholipids and glycolipids, fatty acids
(FAs) are essential for living organisms. Some FAs are important signaling molecules that are
involved in a number of biological processes.1 For instance, the derivatives of arachidonic acid, e.g.
leukotrienes and prostaglandins, mediate inflammation and promote a variety of diseases, including
allergy, obesity and cancer.1 More recently, oxidized arachidonic acid was reported to capable to
navigate cells to ferroptosis.2
The structures of unsaturated FAs are different in terms of the chain length and the number,
location and geometry of the carbonꢀcarbon double bond (C=C). A large number of studies on FAs
have shown that the change in unsaturated FA composition is associated with certain diseases.3,4
Determination of the number and locations of C=C bonds in unsaturated FAs will help understand
ACS Paragon Plus Environment

Page 3 of 20
Analytical Chemistry
1
2
3
4
5
6
7
8
9
their biological functions. More recently, lipid desaturation has been identified as a metabolic marker
and therapeutic target.5 Mass spectrometry (MS) techniques have been widely used for FA analysis,
among which gasꢀchromatography MS (GCꢀMS) is the most traditionally used.6ꢀ8 Both the retention
time and the corresponding mass spectrum are used to characterize FA structures. Currently, it is still
a challenge to identify the location of a C=C in FAs using MS alone. FA C=C isomers cannot be
distinguished by electron ionization (EI) as their mass spectra are almost the same.9 Lowꢀenergy
collisional induced dissociation (CID) is a powerful method to distinguish isomeric structures,
unfortunately the fragment ions of FA C=C isomers cannot be used to locate the C=C bond, because
no diagnostic ions specific to the C=C were produced due to the low acidity/basicity of the C=C
relative to heteroatom positions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to generate structurally informative fragment ions that allows one to accurately
pinpoint the C=C location in a FA, a variety of strategies have been developed. With such strategies,
MS can not only be used for mass measurement, but also allow for structural analysis.
Chargeꢀremote fragmentation (CRF), namely fragmentation of an ion in which the cleaved bond is
isolated from the apparent charge site by a distance of several bonds, has been employed to locate
C=C via highꢀenergy CID of [MꢀH]ꢀ.10ꢀ12 Charge switch derivatization is another wellꢀdeveloped
method for detection and characterization of FAs via conversion of the carboxylic group to a positive
fixedꢀcharge group, offering higher sensitivity of detection than the method reported in this work.13ꢀ15
In intact FAs, by contrast, the charge site is at the distal end of the chain (carboxyl group) and the
C=C is far away from the charge site by several carbonꢀcarbon single bonds. However, abundant
fragment ions arising from fragmentation channels other than that of the C=C are generated as well
in a CRF experiment, especially for polyunsaturated FAs (PUFAs). To increase the relative
abundances of fragment ions indicative of the C=C location, a variety of chemical reactions have
been introduced to covalently modify the C=C,16ꢀ20 making it to be more easily fragmented by CID.
However, these methods were only compatible with FA analysis by gasꢀchromatography MS,
following the conversion of unsaturated FAs to FA methyl esters (FAMEs). With the rapid
development of shotgun FA analysis, new methodologies compatible with electrospray ionizationꢀ
(ESI)ꢀ and nanoESIꢀMS are urgently needed. Blanksby and coꢀworkers have successfully used ozone
to directly cleave the C=C in a spectrum of lipids, directly offering diagnostic ions specific to the
ACS Paragon Plus Environment

Analytical Chemistry
Page 4 of 20
1
2
C=C location without resorting to tandem MS.21 They have integrated ozone reactions into an ESI
source for the structural elucidation of unsaturated lipids, i.e. ozone electrospray ionization (OzESI).
Alternatively, ozone can be introduced into the ion trap of a MS to react with a massꢀisolated lipid,
where the analysis of a complex lipid mixture was made possible (termed ozoneꢀinduced dissociation,
OzID).22ꢀ24 The inconvenience associated with OzID is the need of hardware modification to MS to
allow ozone introduction into the trap. Recently, we have successfully demonstrated the
identification and quantitation of lipid C=C isomers by using Paternò–Büchi (PB) reaction to convert
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a C=C to a fourꢀmembered oxetane ring, which is stable at room temperature but easily fragmented
25ꢀ27
by collisional heating of CID to produce highly abundant diagnostic ions.
Since the C=C is not
directly cleaved, PB reaction products can be massꢀisolated to achieve rapid and accurate assignment
of C=C location(s) in an unsaturated FA from a complex mixture, with a moderate PB reaction yield
(~30%~60%). In comparison with other methods for FA analysis summarized here, PB
reaction/tandem MS has the unique advantage that allows both identification and quantitation of
C=C isomers of FAs, without any chromatographic separation.
In this study, we report a novel method to structurally characterize and quantify C=C isomers of
FAs via epoxidation reaction and tandem mass spectrometry. Transformation of the C=C to a stable
epoxide ring leads to complete fragmentation of the original C=C by CID. The epoxidation reaction
between the C=C bond(s) in FAs and oxygen was facilitated by blowing a lowꢀtemperature plasma
(LTP) into a solution of unsaturated FA(s) dissolved in acetone/water (50/50, v/v). A significant
advantage is that monounsaturated FAs (MUFAs) can be almost quantitatively converted to their
corresponding epoxides within a short reaction time of <1 min, a feature that is highly desirable for
FA quantitation. Following CID, abundant diagnostic ions indicative of the C=C locations are
generated from the epoxides, which not only offer structural information, but also enable rapid
quantitation of unsaturated FAs. Oxygen in air served as the oxidizing reagent. As a practical
application, the identification and quantitation of FA C=C isomers in human plasma were
demonstrated using the developed method.
ACS Paragon Plus Environment

Page 5 of 20
Analytical Chemistry
1
2
3
4
5
Experimental Section
6
7
8
9
Chemicals and reagents. Oleic acid (FA 18:1 (9Z)), palmitoleic acid (FA 16:1 (9Z)),
transꢀ10ꢀheptadecenoic acid (FA 17:1 (10E)), linoleic acid (FA 18:2 (9Z, 12Z)), linolenic acid (FA
18:3 (6Z, 9Z, 12Z)), acetonitrile, and methanol were purchased from Sigma Aldrich (St. Louis, USA).
Vaccenic acid (FA 18:1 (11Z)) and arachidonic acid (FA 20:4 (5Z, 8Z, 11Z, 14Z)) were purchased
from Aladdin (Shanghai, China). Isooctane and ammonium hydroxide were purchased from J&K
Scientific (Beijing, China). Acetone was purchased from Sinopharm Chemical Reagent (Beijing,
China), and Dulbecco's phosphate buffered saline (dPBS) was purchased from Dingguo Changsheng
Biotech. (Beijing, China). Tetradecenoic acid (FA 14:1 (9Z)) was purchased from NuꢀChek Prep, Inc.
All solutions were prepared using deionized water with a resistivity of 18.2 Mꢁ cm, from a MilliꢀQ
water purification system. Helium (purity: 99.999%) was used as the discharge gas for the
lowꢀtemperature plasma (LTP) probe.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fabrication of the LTP probe. Details on the fabrication of a LTP probe can be found
elsewhere.28 Briefly, the probe was fabricated using a quartz tube (o.d. 4 mm and i.d. 2mm) as the
dielectric barrier, with two Cu electrodes surrounding the quartz tube surface, separated by ~2.5 cm.
An alternating current (AC) power with a Vpꢀp of 0.8ꢀ1.02 kV and a frequency of 0.18ꢀ0.27 kHz was
applied to both electrodes to drive plasma generation. The flow rate of the helium supplied to the
LTP probe was 5 L/min.
Experimental setup for the epoxidation of unsaturated fatty acids. The experimental setup used
for the epoxidation of the unsaturated FAs is shown in Figures 1 and S1. Briefly, 0.1 mL unsaturated
fatty acid solution dissolved in 1% NH4OH in acetone/water (50/50, v/v) was transferred to a 1.5 mL
plastic vial. A lowꢀtemperature plasma was then blown into the solution to facilitate the epoxidation
reaction. The resulting solution was then subjected to electrospray ionizationꢀmass spectrometry
(ESIꢀMS) and tandem MS analysis for identification and quantitation of unsaturated FAs.
Mass spectrometry (MS) and nanoESIꢀMS. All experiments were performed on a Thermo LTQ
mass spectrometer (San Jose, CA, USA). The instrument parameters used were as follows: capillary
temperature, 275 °C; capillary voltage, ꢀ10 V; tube lens voltage, ꢀ100 V; maximum injection time,
200 ms; microscans, 2. The commercial ESI source was removed and replaced with a nanoESI
ACS Paragon Plus Environment

Analytical Chemistry
Page 6 of 20
1
2
3
4
5
6
7
8
9
source. Accurate mass measurement experiments were performed on a Thermo QꢀExactive Orbitrap
mass spectrometer (Thermos Scientific, San Jose, CA, USA), operated in the full mass mode. The
QꢀExactive Orbitrap MS instrumental parameters were as follows: capillary temperature: 320 °C,
tube lens voltage: 50 V, mass resolution: 70,000, maximum injection time: 50 ms, and sum of
microscans: 1. Epoxidized FAs were infused into a nanoESI tip of ~10 ꢁm outer diameter pulled
using borosilicate glass capillary tips (1.5 mm o.d. and 0.86 mm i.d.) on a Pꢀ1000 Flaming/Brown
micropipette puller (Sutter Instrument, Novato, CA, USA). FA solution was loaded from the back
opening of the borosilicate glass tip. A stainless steel (SS) wire was inserted to the nanoESIꢀtip to
serve as the electric contact, with the tip aligned with the MS sampling orifice.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Quantitation of fatty acids in human plasma. For absolute quantitation, 10 ꢁM tetradecenoic
acid (FA 14:1 9Z) was used as the internal standard (IS). Taking palmitoleic acid as an example, a
series of palmitoleic acid solutions of 5, 10, 20, 30, 40 ꢁM were prepared respectively, each added
with 10 ꢁM tetradecenoic acid as the IS. After 60 s of epoxidation reaction, the intensity of the
epoxidation products of palmitoleic acid (m/z 269, see Fig 2d) was compared to the intensity of the
products of the IS (m/z 241) to calculate the intensity ratio (I269/I241). A good linear relationship (see
Table S2) was obtained between I269/I241 and the concentration of palmitoleic acid. For the analysis
of fatty acids in human plasma, 10 ꢁM IS was also doped into the FA extract, which was
reconstituted into a volume of solvent that is ten times that of the plasma. FAs were extracted by
following the standard extraction protocol (see SI). Following epoxidation, the concentration of each
fatty acid was calculated from the standard curve prepared for each FA. Other FAs, including
palmitoleic acid (FA 16:1 (9Z)), linoleic acid (FA 18:2 (9Z, 12Z)) and linolenic acid (FA 18:3 (6Z, 9Z,
12Z)), were analyzed in the same way.
Results and Discussion
Epoxidation of unsaturated fatty acids facilitated by a low temperature plasma (LTP)
Oleic acid (one C=C between C9 and C10) was chosen as an example to demonstrate the
methodology developed in this work. The whole analytical process consists of two steps, i.e.
ACS Paragon Plus Environment

Page 7 of 20
Analytical Chemistry
1
2
3
4
5
6
7
8
9
epoxidation of the unsaturated FAs followed by MS analysis (Figure 1a). A centrifuge vial containing
0.1 mL of 70 ꢁM oleic acid (1% NH4OH in 50/50 acetone/water, v/v) was used as the reactor placed
under the LTP. Addition of NH4OH into the reaction system both accelerates the reaction and
improves ESIꢀMS detection of FAs and their epoxidation products. After epoxidation, the FA
solution was analyzed via (ꢀ)nanoESIꢀMS and (ꢀ)nanoESIꢀMS/MS. Before reaction, only oleic acid
([MꢀH]ꢀ, m/z 281.3) can be detected (Figure 1b) in the solution. By contrast, almost no oleic acid was
detectable in the reaction system after a reaction time of 120 s. Instead, the only abundant peak
detected on the mass spectrum was at m/z 297.3, representing a 16 Da mass increase in comparison
with intact deprotonated oleic acid (Figure 1c). Accurate mass measurement using a LTQ QꢀExactive
Orbitrap mass spectrometer (Thermo, San Jose, CA) revealed the mass increase to be 15.99492 Da,
which is exactly the mass of oxygen (15.994915 Da, error: 0.3 ppm, Figure S4). A significant
advantage of this method is the high reaction yield of the epoxidation reaction for unsaturated FAs,
without other major side reactions and is highly desirable for producing abundant C=Cꢀspecific
diagnostic ions and quantitative analysis. CID of the epoxide for oleic acid (m/z 297.3) produces four
major fragments due to the shattering of the epoxide ring, including ions at m/z 155.2, 171.2, 253.3
and 279.3 (Figure 1d). Products at m/z 171.2 and 155.2, each carrying a negative charge at the
carboxyl group, are generated via fragmentation of the threeꢀmembered epoxide ring in two different
mechanisms (Scheme in Figure 1e). For one fragmentation channel, an aldehyde group was
generated at the distal end in one fragment, and in the other complementary fragment, an ethylene
group was formed at the site of cleavage at ꢂ9 C=C via intraꢀmolecular hydrogen transfer. In each
pair of fragments, only the fragment carrying the COOꢀ group can be detected by MS. Therefore, two
ions out of the four fragments can be detected, which are termed diagnostic ions as they can be used
to accurately pinpoint the C=C location. Other product ions at m/z 279.3 and 253.3 were formed
following loss of H2O (18 Da) and CO2 (44 Da) from parent ions. The scheme of the epoxidation
reaction and subsequent decomposition of the epoxide of oleic acid can be found in Figure 1e.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Analytical Chemistry
Page 8 of 20
1
2
3
4
5
6
7
8
9
Helium
AC
a
Low temperature plasma
(LTP)
nano ESI tip
120 s
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-1600 V
DC
MS inlet
Fatty acid(s) in 1% NH3H2O
(acetone/water, 50/50, v/v)
[M+O-H]-
279
b
c
d
-
297
281
[M-H]
100
100
100
m/z 297
171
155
oleic Acid
253
0
0
0
200
300
400
500
500
100
300
200
40
0
200
300
400
500
m/z
m/z
m/z
Not observed
Detected by MS
e
f
CID
oleic acid
m/z 281
epoxide
m/z 297
epoxidation
CID
m/z 155
m/z 171
Figure 1. (a) Experimental setup for the epoxidation of unsaturated FAs catalyzed by a low-temperature
plasma, followed by MS and tandem MS analysis. (b-d) Epoxidation of oleic acid followed by tandem MS
analysis of its epoxidation product. NanoESI mass spectra of 70 ꢀM oleic acid in the negative ion mode:
before (b) and after (c) epoxidation, indicating an almost full conversion of oleic acid without other
reaction channels. (d) MS/MS spectrum of oleic acid epoxide at m/z 297. (e) A general mechanism for the
epoxidation of an unsaturated FA and the CID mechanism of its epoxide to release diagnostic ions. (f)
Scheme for the in situ generation of dioxirane in acetone/water (50/50, v/v) by LTP, and the epoxidation of
unsaturated fatty acids by dioxirane.29,30
To evaluate the versatility of the epoxidation strategy towards other unsaturated FAs,
ACS Paragon Plus Environment

Page 9 of 20
Analytical Chemistry
1
2
3
4
5
6
7
8
9
palmitoleic acid (FA 16:1 (9Z)), transꢀ10ꢀheptadecenoic acid (FA 17:1 (10E)) and transꢀvaccenic
acid (FA 18:1 (11E)) were selected and diagnostic ions similar to those observed for oleic acid were
detected via MS/MS analysis of the epoxides. Regardless of the C=C location and E/Z geometry, all
unsaturated FAs tested were quantitatively converted to their corresponding epoxides (Figure 2aꢀc).
For FA 16:1 (9Z) which contains two less carbons than FA 18:1 (9Z), the massꢀtoꢀcharge ratios of its
diagnostics ions are identical to those of the latter because the C=C locations are identical in both
FAs (Figure 1d). Transꢀvaccenic acid has the same chain length as oleic acid, and both contain one
C=C, albeit at different C=C locations. However, as the C=C in FA 18:1 (11E) is two more carbons
away from the COOꢀ than FA 18:1 (9Z) (oleic acid), the diagnostic ions of FA 18:1 (11E) were 28 Da
higher in mass (m/z 183/199 vs. m/z 155/171) than those of oleic acid (Figure 2f). C=C Locations in
other unsaturated FAs, such as FA 17:1 (10E), can be identified efficiently and accurately in a similar
way (Figure 2e). The fragmentation schemes of the epoxides, which lead to the generation of
diagnostic ions observed in Figure 2dꢀf, were shown in Figure 2g.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Structural analysis of palmitoleic acid (FA 16:1 (9Z)), trans-10-heptadecenoic acid (FA 17:1
(10E)) and trans-vaccenic acid (FA 18:1 (11E)) by epoxidation and MS/MS analysis. (a,b,c) Mass spectra
ACS Paragon Plus Environment

Analytical Chemistry
Page 10 of 20
1
2
3
4
5
6
of FA 16:1 (9Z), FA 17:1 (10E) and FA 18:1 (11E) after epoxidation. (d,e,f) Tandem mass spectra of the
epoxides of FA 16:1 (9Z), FA 17:1 (10E) and FA 18:1 (11E) (collision energy: 30 eV). (g) Proposed
mechanisms of epoxide fragmentation for the release of diagnostic ions upon CID.
7
8
9
Analysis of polyunsaturated FAs (PUFAs) by epoxidation and tandem MS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
When an unsaturated FA contains more than one C=Cs, the epoxidation reaction occurs
nonꢀselectively at each C=C location. This fact was verified through the analysis of linoleic acid (FA
18:2 (9Z, 12Z)) and linolenic acid (FA 18:3 (6Z, 9Z, 12Z)). As linoleic acid and linolenic acid
contain two and three C=Cs, their epoxidation products can contain up to two and three epoxide
rings, respectively (Figure 3aꢀd). Taking linoleic acid as an example, ions at m/z 295.36 correspond
to epoxides generated after addition of one oxygen to either C=C, while ions at m/z 311.36
correspond to the epoxide after addition of two oxygens to both C=Cs. When ions at m/z 295.36 were
massꢀisolated and subjected to CID, two pairs of diagnostic ions were detected, at m/z 155.18/171.18
and m/z 195.18/211.18 (Figure 3b). Identification of each pair of diagnostic ions was straightforward
due to characteristic 16 Da mass difference, as has been discussed previously. Diagnostic ions at m/z
155.18/171.18 were generated after C9ꢀC10 bond cleavage and those at m/z 195.18/211.18 were
generated after C12ꢀC13 bond cleavage (Figure 3c). Together, this set of experimental data showed
that when one oxygen was attached to a PUFA to form an epoxide, it can be attached at any C=C
location. This phenomenon is very useful in identifying the locations of multiple C=Cs in a PUFA as
firstꢀstep epoxidation products containing one oxygen are enough to reveal all C=C locations in a
PUFA. In fact, the MS/MS spectrum of the final reaction product (added with two oxygens, m/z
311.36) is much more difficult to interpret than the products containing only one oxygen (SI, Figure
S2). Similar results were observed for linolenic acid (Figure 3d,e), except that its epoxides contained
up to three oxygens and its epoxides containing one oxygen released three pairs of diagnostic ions
after CID (m/z 113.18/129.18, m/z 153.18/169.18, m/z 193.18/209.18), corresponding exactly to its
three C=Cs at different locations (6Z, 9Z, 12Z). Therefore, for all unsaturated FAs we only use the
epoxides containing one oxygen for MS and tandem MS analysis.
ACS Paragon Plus Environment

Page 11 of 20
Analytical Chemistry
1
2
3
4
5
a)
100
b)
100
279.36, [M-H]-
277.36
6
7
m/z 295.4
8
9
+O
295.36
+2O
267.27
171.18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
195.18
251.27
155.18
311.36
211.18
295.27
300
0
100
0
100
200
300
400
200
400
m/z
m/z
-
c)
C18H31O2 m/z 279.22
Linoleic acid
FA 18:2 (9Z, 12Z)
-
-
C18H31O3 m/z 295.23
C18H31O3 m/z 295.23
Not observed
Not observed
-
-
C9H15O2 m/z 155.11
C12H19O2 m/z 195.13
-
-
C18H31O3 m/z 211.13
C18H31O3 m/z 171.11
d)
e)
100
277.36 , [M-H]-
275.36
100
Linolenic acid
FA 18:3 (6Z, 9Z, 12Z)
m/z 293.4
169.18
+O
293.36
+2O
309.27
265.27
249.27
193.18
153.18
113.18
0
231.27
325.27, +3O
209.18
0
100
200
300
400
100
200
300
400
m/z
m/z
Figure 3. Structural analysis of linoleic acid (FA 18:2 (9Z, 12Z)) by epoxidation and MS/MS analysis. (a)
MS spectrum of linoleic acid after epoxidation reaction. (b) MS/MS spectrum of the epoxide (+16 Da) at
m/z 295. (c) Fragmentation schemes of the +16 Da epoxidation reaction products of linoleic acid. Note
that the oxygen is non-selective for the C=C location when linoleic acid was epoxidized. (d) MS spectrum
of linolenic acid (FA 18:3 (6Z, 9Z, 12Z)) after epoxidation reaction. (e) MS/MS spectrum of epoxide (+16
Da) at m/z 293.
Mechanistic study on the source of oxygen that leads to the production of epoxides
To study the source of oxygen that leads to the production of epoxides, we repeated the
epoxidation of oleic acid by replacing H2O with H218O. If the oxygen in water is responsible for
epoxidation, then the epoxide of oleic acid should have an m/z of 299 rather than 297. However, after
ACS Paragon Plus Environment

Analytical Chemistry
Page 12 of 20
1
2
3
4
5
6
7
8
9
reaction the m/z of the epoxide of oleic acid was still 297 (SI, Figure S3), strongly suggesting that
oxygen in water was not involved in the epoxidation process. Moreover, since the oxygens in
acetone and water can exchange with each other,31 acetone is not involved in reaction either. We thus
hypothesize that the oxygen gas in air may be the O source for epoxidation. To verify this hypothesis,
different amounts of pure oxygen (O2) were mixed with the discharge gas of helium for generating
the LTP. As the amount of oxygen was gradually increased from 0% to 1.5% (by volume), the
relative intensity of the epoxide (m/z 297) increased significantly after the same 20s of reaction
(Figure 4aꢀe). When no oxygen was present in helium, most of the oleic acid was not reacted and the
percentage of oleic acid intensity in the total ion intensity (I281/(I281 + I297 + I313)) was >80% (Figure
4a). By contrast, the introduction of only 0.05% oxygen into helium resulted in ~50% of oleic acid
converted to its epoxide (Figure 4b). We propose that under the influence of LTP, acetone was
converted in situ to 3,3ꢀdimethyldioxirane (DMDO) (see Figure 1e), which was capable to epoxidize
the C=C in FAs.29,30 Conventionally, DMDO is generated in situ by adding peroxymonosulfate
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
(HO2SO3 ) to acetone in buffered aqueous solution.32 In the heliumꢀoxygen plasma, O and HO• are
the primary species formed via well documented pathways.33ꢀ35 We suspect that the reaction between
O and acetone leads to the in situ generation of DMDO for the epoxidation of unsaturated FAs
(Figure 1f). Such a mechanism explains why acetone has to be present in the FA solution, while
other solvents (e.g. methanol, acetonitrile and H2O) did not promote the epoxidation (Figure S7).
Peroxidation, the major sideꢀreaction channel resulting in a net addition of oxygen atom(s) to
unsaturated FAs, competes more favorably for PUFAs than MUFAs. This is likely due to the fact
that the rate constant of hydrogen atom abstraction, the first step for peroxidation, for the bisꢀallylic
36
hydrogens in PUFAs is about 35 times higher than of the allylic hydrogens in MUFAs. Besides,
the direct cleavage of C=C by ozone (O3) generated in LTP are also possible side reaction products,
although they can be barely detected in our experiments.
With the introduction of pure oxygen unwanted overꢀoxidation products of oleic acid emerged
at m/z 313 (addition of two oxygens) and even 329 (addition of three oxygens). Oleic acid was
further consumed as the amount of oxygen was elevated (0.1%~1.5%), producing more epoxides and
sideꢀreaction products (Figure 4cꢀe). Together, these experimental results present strong and direct
evidence that pure oxygen was the oxygen source for epoxidation of unsaturated FAs. Even if no
oxygen is introduced into the discharge gas, O2 in air will inevitably diffuse into the plasma and the
ACS Paragon Plus Environment

Page 13 of 20
Analytical Chemistry
1
2
3
4
5
6
7
8
9
reaction system. However, with the significantly reduced oxygen concentration in helium, reaction
selectivity was greatly improved. Therefore, throughout this work pure helium was used although O2
can drastically accelerate the epoxidation reaction. With an optimal reaction time of 120 s (SI, Figure
S5), the reaction yield can reach ~90% (I281/(I281 + I297)). Other experimental parameters, such as
helium flow rate (SI, Figure S6) and solvent composition (acetone/water, 50/50, v/v, see SI, Figure
S7 and S9), were also studied for optimal reaction selectivity. Addition of 1% NH3•H2O greatly
enhanced the MS detection sensitivity for FAs and their epoxides in negative ion mode (SI, Figure
S8).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Discharge gas (helium) doped with varied amounts of oxygen for the mechanistic study of
epoxidation. Reaction mass spectra collected using (a) pure helium, (b) 0.05% O2 in helium, (c) 0.1% O2
in helium, (d) 0.5% O2 in helium, or (e) 1.5% O2 in helium as the discharge gas for a reaction time of 20 s.
(f) The ratios of the intensity of oleic acid (m/z 281), epoxidation product (m/z 297) and side reaction
products (m/z 313) to the total intensity (I281 + I297 + I313) as the amounts of oxygen introduced into helium
were varied.
ACS Paragon Plus Environment

Analytical Chemistry
Page 14 of 20
1
2
3
4
Analytical performance
5
6
7
8
9
To evaluate the analytical performance of the developed method, the limit of detection (LOD)
was evaluated using oleic acid as the example. When the concentration of oleic acid was reduced to
0.07 ꢁM, the peak of the epoxide (m/z 297) can still be clearly observed (SI, Figure S10). More
importantly, MS/MS CID mass spectrum of m/z 297 clearly shows the peaks of diagnostic ions
indicative of the C=C location (Figure S11). As the absolute amount of oleic acid decreases, the time
needed for its derivatization also decreases. For instance, 10 s of reaction time was enough for the
full conversion of 0.07 ꢁM oleic acid (0.1 mL). The relationship between the reaction yield and
reaction time for 7 ꢁM oleic acid was shown in SI, Figure S12.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Identification and quantitation of unsaturated FAs in human plasma
A variety of FAs exist in the human plasma and a substantial proportion of them are unsaturated.
In this study, as a real application we aim to use the developed method to identify and quantify all FA
C=C isomers in human plasma. For absolute quantitation, tetradecenoic acid (FA 14:1 9Z) was used
as the internal standard (IS). The spectrum of the FA extract from human plasma doped with the IS
before epoxidation was shown in Figure 5a. The mass of unsaturated FAs increased by 16 Da after
60 s of reaction while saturated FAs remained intact (Figure 5b). Tandem MS analysis were
performed to epoxides of unsaturated FAs to determine the C=C locations. It has been found that FA
16:1, FA 18:2 and FA 20:4 were all in pure forms of FA 16:1 (9), FA 18:2 (9, 12) and FA 20:4 (5, 8,
11, 14), respectively. By contrast, FA 18:1 was a mixture of two C=C location (ꢂ9 and ꢂ11) isomers.
A table that summarizes the diagnostic ions for these FAs is listed in Table S1 in SI. For relative
quantitation, a calibration curve (y=1.4466x+0.031) was prepared using a mixture of FA 18:1 ꢂ9 and
ꢂ11 isomers mixed at different molar ratios, at a total concentration of 20 ꢁM. The diagnostic ions
intensity ratio (I11Z/I9Z, y) was plotted against the concentration ratio (x) for preparing the curve. I
represents the total intensity of diagnostic ions. For instance, I11Z is the sum of I199 and I183. A good
linear relationship (R2 = 0.9967) was obtained for the molar ratio (11Z/9Z) ranging from 1:15 to 3:1.
In human plasma, FA 18:1 was a mixture of 92.15% ꢂ9 and 7.85% ꢂ11 isomers, as evidenced from
the two sets of diagnostic ions at m/z 199/183 and m/z 171/155 (Figure 5c). FA 16:1 and FA 18:2
were identified to be FA 16:1 (9) and FA 18:2 (9, 12), respectively. The concentrations of FA 16:1,
ACS Paragon Plus Environment

Page 15 of 20
Analytical Chemistry
1
2
3
4
5
6
7
8
9
FA 18:1 (ꢂ9 and ꢂ11 isomers) and FA 18:2 in the human plasma sample were quantified to be
104±14 ꢁM, 214±23 ꢁM, and 181±35 ꢁM. The amounts of unsaturated FAs obtained by the current
method are in line with those reported in the literature,37 with the concentration of FA 16:1 (9Z) in
the range of 0.9ꢀ93.2 ꢁM, the concentration of FA 18:1 in the range of 36.1ꢀ676 ꢁM and the
concentration of FA 18:2 in the range of 78.7ꢀ586.4 ꢁM (Table S3, SI).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Identification and quantitation of C=C location isomers of FAs from human plasma. (-)NanoESI
mass spectra of the FA extract from human plasma: (a) before and (b) after 60 s of epoxidation reaction.
(c) MS/MS CID of the epoxide (m/z 297) of oleic acid. (d) The calibration curve for the relative quantitation
of FA 18:1 ∆9 and ∆11 isomers.
Conclusion
In conclusion, we have developed a novel method for the structural characterization and
quantitation of FA C=C isomers via epoxidation reaction and tandem MS analysis. The epoxidation
reaction between unsaturated FAs and oxygen was realized by blowing a lowꢀtemperature plasma
into a solution of FAs dissolved in acetone/water. The almost quantitative transformation of MUFAs
ACS Paragon Plus Environment

Analytical Chemistry
Page 16 of 20
1
2
3
4
5
6
7
8
9
makes the developed method especially attractive for the accurate quantitation of simple FA mixtures
by using epoxide intensities. For determination of the C=C location and quantitation of FA C=C
isomers, CID can be performed to FA epoxides of interests to release abundant diagnostic ions. The
total relative intensities of diagnostic ions for a specific C=C isomer allow accurate quantitation.
This analytical capability was successfully demonstrated via the analysis of unsaturated FAs
extracted from human plasma. We are currently extending the developed method to identify and
quantify many other types of unsaturated lipids, with the ultimate goal of using the composition of
lipid C=C isomers for disease diagnosis and pathological studies.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Associated Content
Supporting Information. The Supporting Information is available free of charge on the ACS
Publications website.
Details for fabrication of the LTP probe, mechanistic studies, optimization of experimental
parameters, and calibration curves for quantification (Figures S1−S12 and Table S1ꢀS2) (PDF)
Author Information
Corresponding Authors
*Graduate School of Health Science, Hokkaido University, North 12, West 5, Kitaꢀku, Sapporo,
060ꢀ0812, Japan. Eꢀmail: keino@hs.hokudai.ac.jp
*Department of Precision Instrument, Tsinghua University, Beijing 100084, P.R. China. Eꢀmail:
maxx@tsinghua.edu.cn
Acknowledgement
This work was financially supported by the National Key Research and Development Program of
China (No. 2016YFF0203704), the National Natural Science Foundation of China (Nos. 21390413
ACS Paragon Plus Environment

Page 17 of 20
Analytical Chemistry
1
2
3
4
5
6
and 21621003), and the Ministry of Science and Technology of China (No. 2013CB933800). The
authors acknowledge Dr. Yu Xia for many insightful discussions.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Analytical Chemistry
Page 18 of 20
1
2
3
4
5
6
References
7
8
9
(1) Wymann, M. P.; Schneiter, R. Nat. Rev. Mol. Cell Bio 2008, 9, 162ꢀ176.
(2) Kagan, V. E.; Mao, G.; Qu, F.; Angeli, J. P.; Doll, S.; Croix, C. S.; Dar, H. H.; Liu, B.; Tyurin, V.
A.; Ritov, V. B.; Kapralov, A. A.; Amoscato, A. A.; Jiang, J.; Anthonymuthu, T.; Mohammadyani, D.;
Yang, Q.; Proneth, B.; KleinꢀSeetharaman, J.; Watkins, S.; Bahar, I.; Greenberger, J.; Mallampalli, R.
K.; Stockwell, B. R.; Tyurina, Y. Y.; Conrad, M.; Bayir, H. Nat. Chem. Biol. 2017, 13, 81ꢀ90.
(3) Chavarro, J.; Stampfer, M.; Loda, M.; Campos, H.; Sesso, H.; Ma, J. Cancer Res. 2009, 69.
(4) Zhao, Y.; Wei, Z.; Zhao, H.; Jia, J.; Chen, Z.; Zhang, S.; Ouyang, Z.; Ma, X.; Zhang, X. Anal.
Chem. 2016, 88, 10805ꢀ10810.
(5) Li, J.; Condello, S.; ThomesꢀPepin, J.; Ma, X.; Xia, Y.; Hurley, T. D.; Matei, D.; Cheng, J.ꢀX. Cell
Stem Cell 2017, 20, 303ꢀ314.e305.
(6) Brenna, J. T. Current Opinion in Clinical Nutrition and Metabolic Care 2013, 16, 548ꢀ554.
(7) Christie, W. W. Lipids 1998, 33, 343ꢀ353.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(8) SeppanenꢀLaakso, T.; Laakso, I.; Hiltunen, R. Anal. Chim. Acta 2002, 465, 39ꢀ62.
(9) Vrkoslav, V.; Cvačka, J. J. Chromatogr. A 2012, 1259, 244ꢀ250.
(10) Davoli, E.; Gross, M. L. J. Am. Soc. Mass Spectrom. 1990, 1, 320ꢀ324.
(11) Trimpin, S.; Clemmer, D. E.; McEwen, C. N. J. Am. Soc. Mass Spectrom. 2007, 18, 1967ꢀ1972.
(12) Voinov, V. G.; Claeys, M. Int. J. Mass. Spectrom. 2001, 205, 57ꢀ64.
(13) Bollinger, J. G.; Thompson, W.; Lai, Y.; Oslund, R. C.; Hallstrand, T. S.; Sadilek, M.; Turecek,
F.; Gelb, M. H. Anal Chem 2010, 82, 6790ꢀ6796.
(14) Wang, M.; Han, R. H.; Han, X. Anal. Chem. 2013, 85, 9312ꢀ9320.
(15) Yang, K.; Dilthey, B. G.; Gross, R. W. Anal. Chem. 2013, 85, 9742ꢀ9750.
(16) Kuklev, D. V.; Smith, W. L. J. Lipid. Res. 2003, 44, 1060ꢀ1066.
(17) Michaud, A. L.; Yurawecz, M. P.; Delmonte, P.; Corl, B. A.; Bauman, D. E.; Brenna, J. T. Anal.
Chem. 2003, 75, 4925ꢀ4930.
(18) Moe, M. K.; Anderssen, T.; Strom, M. B.; Jensen, E. J. Am. Soc. Mass Spectrom. 2005, 16,
46ꢀ59.
(19) Moe, M. K.; Jensen, E. Eur. J. Mass. Spectrom. 2004, 10, 47ꢀ55.
(20) Van Pelt, C. K.; Brenna, J. T. Anal. Chem. 1999, 71, 1981ꢀ1989.
(21) Thomas, M. C.; Mitchell, T. W.; Harman, D. G.; Deeley, J. M.; Murphy, R. C.; Blanksby, S. J.
Anal. Chem. 2007, 79, 5013ꢀ5022.
(22) Pham, H. T.; Maccarone, A. T.; Campbell, J. L.; Mitchell, T. W.; Blanksby, S. J. J. Am. Soc.
Mass Spectrom. 2013, 24, 286ꢀ296.
(23) Poad, B. L. J.; Pham, H. T.; Thomas, M. C.; Nealon, J. R.; Campbell, J. L.; Mitchell, T. W.;
Blanksby, S. J. J. Am. Soc. Mass Spectrom. 2010, 21, 1989ꢀ1999.
(24) Thomas, M. C.; Mitchell, T. W.; Harman, D. G.; Deeley, J. M.; Nealon, J. R.; Blanksby, S. J.
Anal. Chem. 2008, 80, 303ꢀ311.
(25) Ma, X.; Chong, L.; Tian, R.; Shi, R.; Hu, T. Y.; Ouyang, Z.; Xia, Y. Proc. Natl. Acad. Sci. U. S.
A. 2016, 113, 2573ꢀ2578.
(26) Ma, X.; Xia, Y. Angew. Chem. Int. Ed. 2014, 53, 2592ꢀ2596.
(27) Ma, X.; Ouyang, Z. TracꢀTrend. Anal. Chem. 2016, 85, 10ꢀ19.
ACS Paragon Plus Environment

Page 19 of 20
Analytical Chemistry
1
2
3
4
5
6
(28) Harper, J. D.; Charipar, N. A.; Mulligan, C. C.; Zhang, X.; Cooks, R. G.; Ouyang, Z. Anal.
Chem. 2008, 80, 9097ꢀ9104.
(29) Wong, O. A.; Shi, Y. Chem. Rev. 2008, 108, 3958ꢀ3987.
(30) Murray, R. W. Chem. Rev. 1989, 89, 1187ꢀ1201.
7
8
9
(31) Roberts, I. Ann. Ny. Acad. Sci. 1939, 39, 375ꢀ394.
(32) Murray, R. W.; Jeyaraman, R. J. Org. Chem. 1985, 50, 2847ꢀ2853.
(33) Sankaranarayanan, R.; Pashaie, B.; Dhali, S. K. Appl Phys Lett 2000, 77, 2970ꢀ2972.
(34) Fehsenfeld, F. C.; Mosesman, M.; Ferguson, E. E. J. Chem. Res. 1971, 55, 2115ꢀ2120.
(35) Laux, C. O.; Spence, T. G.; Kruger, C. H.; Zare, R. N. Plasma Sources Sci. Technol. 2003, 12,
125.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(36) Xu, L.; Porter, N. A. Free Radic. Res. 2015, 49, 835ꢀ849.
(37) Ukolov, A. I.; Orlova, T. I.; Savel’eva, E. I.; Radilov, A. S. J. Anal. Chem. 2015, 70, 1123ꢀ1130.
ACS Paragon Plus Environment

Analytical Chemistry
Page 20 of 20
1
2
3
4
Table of Contents
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment