Trans Lipid Library: Synthesis of Docosahexaenoic Acid (DHA) Monotrans Isomers and Regioisomer Identification in DHA-Containing Supplements

Article abstract of DOI:10.1021/acs.chemrestox.8b00021

Docosahexaenoic acid (DHA) is a semiessential polyunsaturated fatty acid (PUFA) for eukaryotic cells that is found in natural sources such as fish and algal oils and widely used as an ingredient for omega-3 containing foods or supplements. DHA effects are

Full text of DOI:10.1021/acs.chemrestox.8b00021

Subscriber access provided by UNIV OF DURHAM
Article
Trans Lipid Library: synthesis of docosahexaenoic acid (DHA) monotrans
isomers and regioisomer identification in DHA-containing supplements
GEORGIA MENOUNOU, GIORGIA GIACOMETTI, ROBERTA SCANFERLATO, Paolo Dambruoso,
anna sansone, Itziar Tueros, JAVIER AMEZAGA, Chryssostomos Chatgilialoglu, and Carla Ferreri
Chem. Res. Toxicol., Just Accepted Manuscript • DOI: 10.1021/acs.chemrestox.8b00021 • Publication Date (Web): 27 Feb 2018
Downloaded from http://pubs.acs.org on February 28, 2018
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 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 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.
Chemical Research in Toxicology 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 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
Trans Lipid Library: synthesis of docosahexaenoic acid
(DHA) monotrans isomers and regioisomer identification
in DHAꢀcontaining supplements
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
Georgia Menounou1‡, Giorgia Giacometti1‡, Roberta Scanferlato1, Paolo Dambruoso1, Anna Sansone1, Itziar
Tueros2, Javier Amézaga2, Chryssostomos Chatgilialoglu1, Carla Ferreri1,*
1ISOF, Consiglio Nazionale delle Ricerche, Via P. Gobetti 101, 40129 Bologna (Italy)
2AZTI, Food and Health, Parque Tecnológico de Bizkaia, Astondo Bidea 609, 48160 Derio (Spain)
‡ contributed equally to the work
KEYWORDS. Omegaꢀ3 fatty acids, trans fatty acid isomers, isomer identification, deodorization,
nutraceuticals.
ABSTRACT: Docosahexaenoic acid (DHA) is a semiꢀessential polyunsaturated fatty acid (PUFA) for
eukaryotic cells, found in natural sources such as fish and algal oils and widely used as ingredient for omegaꢀ3
containing foods or supplements. DHA effects are connected to its natural structure with six cis double bonds,
but geometrical monotrans isomers can be formed during distillation or deodorization processes, as an
unwanted event that alters molecular characteristics and annihilates health benefits. The characterization of the
six monotrans DHA regioisomers is an open issue to address for analytical, biological and nutraceutical
applications. Here we report the preparation, separation and first identification of each isomer by a dual
52 approach consisting of: i) the direct thiyl radicalꢀcatalyzed isomerization of cisꢀDHA methyl ester, and ii) the
53
54
twoꢀsteps synthesis from cisꢀDHA methyl ester via monoꢀepoxides as intermediates, which are separated and
55
56
57
58
59
ACS Paragon Plus Environment
60
1

Chemical Research in Toxicology
Page 2 of 32
1
2
3
4
5
6
7
8
9
identified by bidimensional nuclear magnetic resonance spectroscopy, followed by elimination for the
unequivocal assignment of the double bond position. This monoꢀtrans DHA isomer library with NMR and GC
analytical characterization was also used to examine the products of thiyl radical catalysed isomerization of a
fish oil sample and to evaluate the trans isomer content in omegaꢀ3 containing supplements commercially
available in Italy and Spain.
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
INTRODUCTION
Cis polyunsaturated fatty acids (PUFA) are essential components of human fats, particularly important for
biological functions, such as for example to form membrane phospholipids and signalling molecules. In vivo,
the omegaꢀ3 fatty acid docosaꢀ4Z,7Z,10Z,13Z,16Z,19Zꢀhexaenoic acid (cisꢀDHA) (Figure 1) is produced
through several steps of elongation and desaturation from alphaꢀlinolenic acid (octaꢀ9Z,12Z,15Zꢀdecatrienoic
acid), the essential precursor taken exclusively from the diet 1. This nineꢀsteps biosynthesis is variably efficient
in humans, therefore the semiꢀessential nature of DHA is now worldwide recognized by Health Agencies. Due
to its importance for a correct human growth, adequate daily intakes of 100ꢀ200 mg have been established by
the main international food safety and health agencies from dietary sources, such as algae or fish, especially in
children and pregnant women 2. This led to an increased marketing of DHAꢀrich formulas, either as functional
foods and supplements. On the other hand, the multiple roles of DHA for molecular pathways and signalling
have attracted interest in the last decade, all biological activities being strictly dependent from the structural
requisite of the cis geometry 3. When high temperatures and low pressures are used in deodorization processes,
to eliminate the unpleasant fishy smell of oils, the natural cis structure of the polyunsaturated fatty acids is
affected and the formation of geometrical trans isomers was demonstrated 4,5. The health consequences of trans
PUFA are matters of several studies that evidenced harmful roles for cardiovascular health and pregnancy 6,7
,
whereas it is worth noting that omegaꢀ3 supplementation in persons with cardiovascular diseases and pregnant
women is highly recommended.
ACS Paragon Plus Environment
2

Page 3 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
Figure 1. Structure of DHA (4Z,7Z,10Z,13Z,16Z,19Zꢀdocosahexaenoic acid)
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
In the frame of our research on the free radical modifications of biomolecules, we studied geometrical trans
fatty acids (TFA) obtained from an isomerization process catalysed by sulfurꢀcentered radicals, which occurs
via an additionꢀelimination mechanism without involving the shift of the double bond, with interesting insights
for polyunsaturated fatty acids 8,9. The number of geometrical isomers for the unsaturated fatty acid is
calculated as to 2n, where the n is the number of double bonds, therefore in case of long chain PUFA, such as
DHA, a high number of isomers can be calculated (26=128). However, due to the stepꢀbyꢀstep mechanism of
this reaction, monotrans isomers are the first and major products and they are also the most relevant products of
radical stress in biological systems, where the low radical concentration produced in vivo is able to involve one
double bond, as examined by using the biomimetic model of membranes in the form of unilamellar liposomes 9.
Indeed, monotrans isomers of arachidonic acid in the 5 and 8 positions were found the most relevant products
created during oxidative metabolism with production of thiyl radicals, distinguishable from TFA obtained from
the dietary intake and metabolic transformation of chemically manipulated oils containing trans isomers of
linoleic acid 10ꢀ13. Some attention has been given to the biological effects of specific monotrans isomers,
showing differences from the natural cis analogues 14ꢀ17, however extension of such studies is limited by the
availability of TFA molecular library and by the few commercially available TFAs.
Synthetic routes to PUFA monotrans isomers are useful to build up the trans lipid library and address
analytical protocols and biological studies. We first characterized the monotrans isomers of arachidonic acid (4
isomers)8 and eicosapentaenoic acid (EPA, 5 isomers)18 using gas chromatography (GC) and carbonꢀ13 nuclear
magnetic resonance (13C NMR) in combination for the assignment of each isomer structure. In the case of EPA,
ACS Paragon Plus Environment
3

Chemical Research in Toxicology
Page 4 of 32
1
2
3
4
5
6
7
8
9
we tested the efficiency of a dual synthetic approach to obtain the five monoꢀtrans isomers in comparison with
the isomers coming from the free radicalꢀcatalysed isomerization, as shown in Scheme 1. The strategy of
separation and assignment of the monoꢀepoxide structures, prior to the elimination, successfully worked for the
first unambiguous determination of the trans alkene position, integrating previous data in the literature 19ꢀ21
.
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
Scheme 1. Dual synthetic approach to obtain the geometrical trans isomer: from cisꢀalkene via epoxide
formation and ring opening, or via radicalꢀbased isomerization.
We considered the dual approach shown in Scheme 1 to be useful also for the synthesis of the monoꢀtrans
DHA isomers. Only the 4E regioisomer is commercially available as methyl ester and can be used as GC
reference, whereas in literature monotrans DHA isomers were analysed by several authors after
deodorization/isomerization reactions of DHAꢀrich oils, assaying various conditions of GC analysis but lacking
a specific regioisomer assignment 4,5,16,20,21. As far as the synthesis of the DHA monoꢀepoxides is concerned,
the classical epoxidation reaction can lead to these compounds 22, which are named epoxydocosapentaenoic
acid (EDP). 4,5ꢀEDP and 19,20ꢀEDP were known 19,23ꢀ25 also due to their biological relevance as metabolites of
cytochrome P450 26,27. The strategy of the separation of the intermediate monoꢀepoxide regioisomers was
successful to unambiguously individuate the epoxidation of the double bond position, using monoꢀ and
bidimensional nuclear magnetic resonance (NMR) experiments. Consequently, after ring opening, dibromide
formation and elimination from each monoꢀepoxide regioisomer the double bond was unequivocally
ACS Paragon Plus Environment
4

Page 5 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
established, thus allowing for the assignment of the corresponding peaks in the GC analysis. This double
approach clarified that the direct isomerization by free radicalꢀcatalysed reaction gives the six isomers as
exclusive products within few minutes. The usefulness of such identification was then assayed for analysis of
fish oil and its isomerization products, also in view of applying the analytical route to the purity of fish oil
preparations, which is a hot topic of food research 5, 28, 29. We studied transꢀcontaining triglycerides, and we
showed that NMR represents a powerful analytical methodology to apply directly on the oil without
transformation to fatty acid methyl esters 30. The monotrans DHA library and GC characterizations were
finally applied to the analysis of omegaꢀ3 containing supplements in capsules, commercially available in Italy
and Spain, evidencing trans isomer contamination in the oil ingredient.
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
Experimental Section
Materials and Methods
Chloroform, methanol, nꢀhexane, 2ꢀpropanol and acetonitrile were purchased from Merck (HPLC purity).
Absolute ethanol was purchased from BDH Prolabo VWR (AnalaR NORMAPUR) and 2ꢀmercaptoethanol was
purchased from Aldrich. Silver nitrate, 30% aq NH4OH and anhydrous sodium sulfate (Na2SO4) were
purchased from Carlo Erba. All fatty acid methyl esters (FAME) used as reference standard for GC analyses
were purchased from Aldrich, Fluka or Sigma and used without further purification. 4E, 7Z10Z,13(Z) 16(Z)
19(Z) DHAꢀmethylester was purchase from Lipidox (Lidingö, Sweden). Deuterated benzene and deuterated
CDCl3 were purchased from Eurisotop (France). Silica gel thinꢀlayer chromatography (analytical and
preparative) was performed on Merck silica gel 60 plates (0.25 and 2 mm thickness, respectively).
GC analyses. Fatty acid methyl esters were analyzed by gas chromatograph (Agilent 6850, Milan) equipped
with a 60m × 0.25mm × 0.25ꢁm (50%ꢀcyanopropyl)ꢀmethylpolysiloxane column (DB23, Agilent, USA). The
instrument has a flame ionization detector (FID) that requires air (450 mL / min) and hydrogen (40 mL / min)
ACS Paragon Plus Environment
5

Chemical Research in Toxicology
Page 6 of 32
1
2
3
4
5
6
7
8
9
and is maintained at a temperature of 250 ° C. Two identical equipment conditions were used for the analyses,
applying the same injection temperature (230°C) but different oven conditions and carrier gas. Method A: from
an initial temperature of 165°C held for 3 min, followed by an increase of 1°C/min up to 195°C, held for 40
min. A final ramp, with a temperature increase of 10°C/min up to a maximum temperature of 240°C, was
maintained for 10 min for column purge. A constant pressure mode (29 psi) was chosen with helium as carrier
gas. Methyl esters were identified by comparison with the retention times of commercially available standards
or trans fatty acid references, obtained as described elsewhere. Method B: temperature started from 195°C, held
for 26min, followed by an increase of 10°C/min up to 205°C, held for 13min, followed by a second increase of
30°C/min up to 240°C, held for 10min. A constant pressure mode (29 psi) was chosen with hydrogen as carrier
gas. GC/MS spectra were recorded on a Clarus 500 GC apparatus equipped with a Clarus 560S mass
spectrometer (GC/MS transfer line temperature 230°C) and a 60m x 0.25mm x 0.25µm (50%ꢀ
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
cyanopropylphenyl)ꢀdimethylpolysiloxane column (DB225 ms, Agilent, USA), using injector temperature of
230°C, a split ratio of 50:1 and helium as carrier gas at a constant flow of 1.2 mL/min, with the following oven
conditions: temperature started from 195°C, held for 52min, followed by an increase of 3°C/min up to 205°C,
held for 10min, followed by a second increase of 3°C/min up to 225°C, held for 15min and a final increase of
5°C/min up to 230°C, held for 10min.
Photolysis. Photolysis was carried out in a quartz photochemical reactor (Sigma Aldrich) equipped with a 5.5
W lowꢀpressure mercury lamp. The temperature was maintained at (22 ± 2)°C by means of a thermostat bath.
Nuclear magnetic resonance (NMR). All NMR spectra were collected with a fully automated Agilent NMR
system, consisting of a 54ꢀmm bore, 500 MHz (11.7 T) Premium Shielded superconducting magnet, a DD2
Performa IV NMR console and the Agilent OneNMR probe. All samples were dissolved in C6D6 and
transferred in 3 mm thin wall NMR tubes (Wilmad 335ꢀPPꢀ8). After having calibrated each sample pw90, the
ACS Paragon Plus Environment
6

Page 7 of 32
Chemical Research in Toxicology
1
2
3
4
5
following 7 NMR experiments were collected on each sample (2D experiments were collected with 50% NUS
and F2 sw (1H): 10 ppm, fromꢀ0.5 to 9.5 ppm):
6
7
8
9
•
1H NMR spectra: pw 90 was calibrated for each sample, sw: ꢀ0.5 to 9.5 ppm
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
• one gHSQC: F1 sw (13C) 135 ppm, from 5.0 to 140 ppm; nt=32; ni=512; res: 33 Hz (0.26 ppm)
F2 sw (1H): 10 ppm, from ꢀ0,5 to 9,5 ppm, NUS: 50%
• two bsgHSQCAD: first F1 sw (13C): 60 ppm, from 5.0 to 65 ppm; nt=32; ni=512; res: 15 Hz (0.12 ppm)
F2 sw (1H): 10 ppm, from ꢀ0.5 to 9.5 ppm, NUS: 50%
second: F1 sw (13C): 20 ppm, from 120.0 to 140 ppm; nt=32; ni=512; res: 5 Hz (0.04 ppm)
F2 sw (1H): 10 ppm, from ꢀ0.5 to 9.5 ppm, NUS: 50%
• one gHMBCAD: F1 sw (13C): 175 ppm, from 5.0 to 180 ppm; nt=64; ni=512; res: 43 Hz (0.34 ppm)
F2 sw (1H): 10 ppm, from ꢀ0.5 to 9.5 ppm, NUS: 50%
• two bsgHMBC: first F1 sw (13C): 60 ppm, from 5.0 to 65 ppm; nt=64; ni=512; res: 15 Hz (0.12 ppm)
F2 sw (1H): 10 ppm, from ꢀ0.5 to 9.5 ppm, NUS: 50%
second F1 sw (13C): 60 ppm, from 120.0 to 180 ppm; nt=64; ni=512; res: 15 Hz (0.12 ppm)
F2 sw (1H): 10 ppm, from ꢀ0.5 to 9.5 ppm, NUS: 50%.
Epoxidation of Methyl Allꢀ(Z)ꢀ4,7,10,13,16,19ꢀDocosahexaenoate (DHA–Me, 1). Synthesis of
epoxydocosapentaenoic acid methyl ester (EDPꢀMe) regioisomers. A 5.84 mM solution of metaꢀ
chloroperoxybenzoic acid in dichloromethane (2.5 mL corresponding to 0.0146 mmol; 2.51 mg) was added
dropwise to a solution of DHA methyl ester 1 (50 mg; 0.146 mmol) in dichloromethane (2.5 mL). After the
addition was completed, the mixture was stirred for 10 min at room temperature under argon atmosphere,
ACS Paragon Plus Environment
7

Chemical Research in Toxicology
Page 8 of 32
1
2
3
4
5
6
following the formation of monoꢀepoxides with traces of diꢀepoxides by TLC (see Figure S1 in Supporting
Information). Eluent: 7/3 nꢀhexane /diethyl ether; DHAꢀMe Rf = 0.9, first mixture (5, 6 and traces of 4) Rf =
7
8
9
0.7, second mixture (3 and 7) Rf = 0.6, pure isomer (2) Rf = 0.5, diepoxides Rf ≤ 0.4). Workꢀup was carried out
by addition of 5 mL of iceꢀcold NaHCO3 (25% w/v) stirred for 2 min. Then, the reaction mixture was
transferred in a separating funnel, the aqueous layer was discarded and the organic layer was washed two times
with NaHCO3, deionized H2O and dried over Na2SO4. The crude was purified with flash chromatography
(eluent: 9/1 nꢀhexane/diethyl ether) to give a first fraction containing an inseparable mixture (colourless oil; 8
mg; 0.0223 mmol; 15.3% yield) containing 13,14ꢀEDPꢀMe (5), 16,17ꢀEDPꢀMe (6) with traces of 10,11ꢀEDP 4
and of residual starting material (1). Structures 5 and 6 in Figure 2, were found in a 29:71 ratio, respectively, as
calculated by the integration of protons of Cꢀ22 on 1H NMR spectrum (trace II in Figure 5). EDPꢀMe
regioisomers 5 and 6 were further purified by preparative TLC using 100 mL toluene/500 ꢁL isopropanol as the
eluent: (Rf = 0.5, pure isomer 6, and Rf = 0.4, pure isomer 5) to give pure 6 (3 mg; 0.0084 mmol; 37.5% yield)
with only traces of 4.
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
The second fraction contained an inseparable mixture of 7,8ꢀEDPꢀMe (3) and 19,20ꢀEDPꢀMe (7) (colourless
oil; 7mg; 0.0195 mmol; 13.4% yield) (Figure 2) found in a 43:57 ratio, respectively, as calculated by the
integration of protons of Cꢀ22 on 1H NMR spectrum (trace III in Figure 5). The subsequent fraction contained
pure 4,5ꢀEDPꢀMe (2) regioisomer (Figure 2) (colourless oil; 5 mg; 0.0134 mmol; 9.2% yield) identified as
shown in Figure 5 (trace IV) for the Cꢀ22 triplet at 1H NMR.
The starting material DHAꢀMe (1) recovered was 30 mg (0.0877 mmol; 60% recovered yield).
The structural assignment of each regioisomer in the fractions was performed by dissolving the fractions in
C6D6 as NMR solvent in which they are stable, and carrying out monoꢀ and biꢀdimensional experiments. Table
S1 and Figure S2 in Supporting Information report the detailed 1H and 13C NMR resonances attributed to the
ACS Paragon Plus Environment
8

Page 9 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
regioisomers as well as to cisꢀDHA methyl ester. For the 1H NMR of the monoꢀepoxides see Figures S6 and S7
in the Supporting Information. 4,5ꢀEDP (2) and 19,20ꢀEDP (7) were previously reported in CDCl3 (cfr.,
Table S2) [22,23]. In Supporting Information the 2D NMR experiments for the assignment of the C and H
resonances in the EDP regioisomers are shown (Figures S4 and S5).
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
Here below, a summary of the main NMR spectroscopy for the EDP regiosiomers assignment is provided.
Methyl (Z)-13,14-Epoxy-all-(Z)-4,7,10,16,19-docosahexaenoate (5). 1H NMR (C6D6, 500 MHz): δΗ 5.25 ꢀ 5.55
(m, 10H), 3.31 (s, 3H), 2.72 ꢀ 2.83 (m, 8H), 2.25 ꢀ 2.41 (m, 4H), 2.12 (s, 4H), 1.93 ꢀ 2.00 (m, 2H), 0.88 (t, J =
7.30 Hz, 3H). 13C NMR (C6D6, 126 MHz): δC 172.35 (C1), 131.90 (C20), 130.53 (C17), 130.19 (C10), 129.02
(C5), 128.32 (C8), 128.11 (C4), 127.87 (C7), 126.82 (C19), 124.76 (C11), 124.55 (C16), 55.66 (C14), 55.65
(C13), 50.65 (C1`), 33.63 (C2), 26.35 (C15), 26.34 (C12), 25.78 (C9), 25.68 (C18), 25.57 (C6), 22.75 (C3),
20.53 (C21), 14.03 (C22).
Methyl (Z)-16,17-Epoxy-all-(Z)-4,7,10,13,19-docosahexaenoate (6). 1H NMR (C6D6, 500 MHz): δ 5.25 ꢀ 5.55
(m, 10H), 3.31 (s, 3H), 2.72 ꢀ 2.83 (m, 8H), 2.25 ꢀ 2.41 (m, 4H), 2.12 (s, 4H), 1.87 ꢀ 1.94 (m, 2H), 0.86 (t, J =
7.30 Hz, 3H). 13C NMR (C6D6, 126 MHz): δ 172.35 (C1), 133.81 (C20), 130.19 (C13), 129.02 (C5), 128.40
(C10), 128.32 (C8), 128.11 (C4), 127.87 (C7), 127.70 (C11), 124.76 (C14), 123.66 (C19), 55.79 (C17), 55.65
(C16), 50.65 (C1`), 33.63 (C2), 26.34 (C15), 26.22 (C18), 25.78 (C12), 25.66 (C9), 25.57 (C6), 22.75 (C3),
20.53 (C21), 14.00 (C22).
Methyl (Z)-7,8-Epoxy-all-(Z)-4,10,13,16,19-docosahexaenoate (3). 1H NMR (C6D6, 500 MHz): δ 5.35 ꢀ 5.51
(m, 10H), 3.30 (s, 3H), 2.69 ꢀ 2.85 (m, 8H), 2.19 ꢀ 2.32 (m, 4H), 2.08 ꢀ 2.14 (m, J = 7.30 Hz, 4H), 1.93 ꢀ 2.04
(m, 2H), 0.89 (t, J = 7.50 Hz, 3H). 13C NMR (C6D6, 126 MHz): δ 172.30 (C1), 131.82 (C20), 130.17 (C11),
130.08 (C4), 128.35 – 127.80 (C13, C14, C16 and C17), 127.06 (C19), 125.77 (C5), 124.78 (C10), 55.65 (C8),
ACS Paragon Plus Environment
9

Chemical Research in Toxicology
Page 10 of 32
1
2
3
4
5
6
55.63 (C7), 50.66 (C1`), 33.51 (C2), 26.35 (C9), 26.23 (C6), 25.78 (C12), 25.75 – 25.60 (C15), 25.58 (C18),
22.83 (C3), 20.55 (C21), 14.07 (C22).
7
8
9
Methyl (Z)-19,20-Epoxy-all-(Z)-4,7,10,13,16-docosahexaenoate (7). 1H NMR (C6D6, 500 MHz): δ 5.25 ꢀ 5.52
(m, 10H), 3.31 (s, 3H), 2.73 ꢀ 2.86 (m, 9H), 2.60 (dt, J = 4.20, 6.40 Hz, 1H), 2.25 ꢀ 2.37 (m, 3H), 2.04 ꢀ 2.16
(m, J = 7.30 Hz, 3H), 1.40 (ddq, J = 7.30, 7.30, 14.10 Hz, 1H), 1.28 (ddq, J = 6.10, 7.30, 13.50 Hz, 1H), 0.84 (t,
J = 7.70 Hz, 3H). 13C NMR (C6D6, 126 MHz): δ 172.32 (C1), 130.03 (C16), 129.06 (C5), 128.35 – 127.80 (C7,
C8, C10, C11, C13 and C14), 128.08 (C4), 124.96 (C17), 57.38 (C20), 55.75 (C19), 50.65 (C1`), 33.64 (C2),
26.31 (C18), 25.79 (C15), 25.75 – 25.60 (C9 and C12), 25.58 (C6), 22.76 (C3), 21.10 (C21), 10.45 (C22).
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
Methyl (Z)-4,5-Epoxy-all-(Z)-7,10,13,16,19-docosahexaenoate (2). 1H NMR (C6D6, 500 MHz): δ 5.23 ꢀ 5.55
(m, 10H), 3.29 (s, 3H), 2.65 ꢀ 2.88 (m, 10H), 2.13 ꢀ 2.27 (m, 3H), 1.95 ꢀ 2.09 (m, 3H), 1.68 ꢀ 1.76 (m, 1H), 1.58
ꢀ 1.68 (m, 1H), 0.89 (t, J = 7.63 Hz, 3H). 13C NMR (C6D6, 126 MHz): δ 172.35 (C1), 131.89 (C20), 130.27
(C8), 128.63/128.38/128.11/128.96 (C13, C14, C16 and C17), 128.42 (C11), 127.95 (C10), 127.19 (C19),
124.78 (C7), 56.07 (C5), 55.32 (C4), 50.85 (C1`), 30.76 (C2), 26.38 (C6), 25.86 (C9), 25.75 (C12 and C15),
25.66 (C18), 23.42 (C3), 20.66 (C21), 14.17 (C22).
Methyl All-(Z)-4,7,10,13,16,19-Docosahexaenoate (1). 1H NMR (C6D6, 500 MHz): δ 5.24 ꢀ 5.47 (m, 12H), 3.31
(s, 3H), 2.73 ꢀ 2.88 (m, 10H), 2.29 (dddd, J = 0.80, 1.40, 7.34, 14.70 Hz, 1H), 2.29 (dddd, J = 0.80, 1.40, 7.30,
14.70 Hz, 1H), 2.11 (t, J = 7.40 Hz, 2H), 1.95 ꢀ 2.03 (m, 2H), 0.89 (t, J = 7.61 Hz, 3H). 13C NMR (C6D6, 126
MHz): δ 174.32 (C1), 131.78 (C20), 129.08 (5), 128.50 (C17), 128.22 (C14), 128.15 (C8, C10 and C11),
128.11 (C7, C13 and C19), 128.06 (C4), 127.91 (C16), 50.65 (C1`), 33.65 (C2), 25.67 (C15), 25.66 (C9), 25.65
(C18), 25.57 (C6), 25.56 (C12), 22.75 (C3), 20.54 (C21), 14.08 (C22).
ACS Paragon Plus Environment
10

Page 11 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
Transformation of EDPꢀMe regioisomers to Monotrans DHA–Me Isomers. The reaction is described for
16,17ꢀEDPꢀMe and 13,14ꢀEDPꢀMe (first fraction of the EDP chromatographic separation), and the same steps
were followed for the other isolated EDPꢀMe fractions.
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
Step 1. (Bromination of Epoxide). To a freshly prepared solution of triphenylphosphine dibromide (7 mg;
0.0167 mmol) in dry dichloromethane (50 ꢁL) a solution containing a mixture of EDPꢀregioisomers, 6 and 5, (3
mg; 0.0084 mmol containing traces of 10,11ꢀEDP 4) in dry dichloromethane (100 ꢁL) and pyridine (1 ꢁL) was
added, stirring at 0°C under argon atmosphere. The mixture was left stirring overnight at 5°C, and then was
quenched with a 1 M aqueous solution of hydrochloric acid (0.5 mL) and extracted three times with
chloroform/ethanol (4/1). The reaction was monitored by TLC for the ring opening as dibromide (7/3 nꢀ
hexane/diethyl ether, starting material Rf = 0.7 and dibromide product Rf = 0.4). The organic layers were
collected, dried over Na2SO4, and evaporated under vacuum. The crude containing methyl 16,17ꢀdibromoꢀ
(4Z,7Z,10Z,13Z,19Z)ꢀdocosapentaenoate and 13,14ꢀdibromoꢀ(4Z,7Z,10Z,16Z,19Z)ꢀdocosapentaenoate (with
traces of the 10,11ꢀdibromoderivative) was used for the next step. The same reaction was also performed for the
inseparable mixture of the EDPꢀMe regioisomers 3 and 7 and for the EDPꢀME regioisomer 4. In all cases the
reaction crude was used for the next step.
Step 2. (Elimination). To an iceꢀcooled slurry of activated zinc (4.7 mg; 0.072 mmol), acetic acid (2 ꢁL), and
N,Nꢀdimethylformamide (150 ꢁL)
a
solution of methyl 16,17ꢀdibromoꢀ(4Z,7Z,10Z,13Z,19Z)ꢀ
docosapentaenoate and traces of 13,14ꢀdibromoꢀ(4Z,7Z,10Z,16Z,19Z)ꢀdocosapentaenoate (3 mg; 0.0060 mmol)
in N,Nꢀdimethylformamide (100 ꢁL) was added and the mixture was left stirring at 0°C for 14 hrs. The reaction
was monitored by TLC (eluent: 7/3 nꢀhexane/diethyl ether) evidencing the formation of alkenes (Rf = 0.9) in
the presence of traces of starting material (Rf = 0.4). At this point, the reaction was stopped by addition of a 1
M aqueous hydrochloric acid solution (1 mL), and the precipitate was filtered off. The filtrate was extracted
three times with a mixture of chloroform/methanol (3/1). The organic layers were collected and dried over
ACS Paragon Plus Environment
11

Chemical Research in Toxicology
Page 12 of 32
1
2
3
4
5
6
7
8
9
Na2SO4. The solvent was removed under vacuum to afford a colourless oil (1.5 mg; 0.0044 mmol; 73% yield).
The reaction crude was analysed by AgꢀTLC (eluent: 1/9 nꢀhexane/diethyl ether with 0.4% MeOH) evidencing
the presence of two spots, referred to monotrans DHAꢀMe and cis DHAꢀMe (see Figure S8 (b) in Supporting
Information). The other two crude reaction mixtures of dibromides were treated for elimination as described
above. The debromination reaction occurred also in the second fraction transformed into the dibromide
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
intermediates
affording
methyl
(4Z,7Z,10Z,13Z,16Z,19E)ꢀdocosahexaenoate
and
methyl
(4Z,7E,10Z,13Z,16Z,19Z)ꢀdocosahexaenoate (colourless oil; 3.5 mg; 0.0010 mmol; 86% yield), as well as for
the last EDP fraction affording methyl (4E,7Z,10Z,13Z,16Z,19Z)ꢀdocosahexaenoate (colourless oil;
The elimination products were then analysed by GC taking advantage of the parallel NMR assignments of
regioisomers, thus allowing for the attribution of the trans double bond position. Figures S11ꢀS13 in
Supporting Information are the GC traces of the elimination of the three EDP fractions, in comparison with
the trans isomer mixture obtained by photolysis and with literature data [4,5,16,20,21]. The assignment of the
monotrans isomer peaks in the GC analysis was obtained on the basis of the starting monoꢀepoxide assignment
(Figure 4).
Scheme 2 in the main text summarizes the formation of monoꢀepoxides and the chromatographic separation of
fractions that underwent the brominationꢀelimination step affording the separation of the monotrans
regioisomers.
Synthesis of Monotrans DHAꢀMe Isomers by Photolysis. A 15 mM solution of DHAꢀMe ester (20 mg, 0.058
mmol) in 2ꢀpropanol (3.87 mL) was transferred into a quartz photochemical reactor (Sigma Aldrich) equipped
with a 5.5 W lowꢀpressure mercury lamp. The reaction mixture was degassed with argon for 20 min, then an
aliquot of a previously degassed stock solution of 2ꢀmercaptoethanol in 2ꢀpropanol was added (0.03 mmol) and
the UV lamp was turned on for 5 minutes at the temperature of 22 ± 2 °C, kept by means of a thermostat bath.
The reaction was monitored by AgꢀTLC, spraying the plate with cerium ammonium sulfate/ammonium
ACS Paragon Plus Environment
12

Page 13 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
molybdate reagent (CAM), to evidence the formation of the monotrans fraction together with ditrans, tritrans
products, in order to stop the reaction when monotrans isomers were prevalently formed in the presence of
remaining cisꢀDHA (see Figure S8 (a) in Supporting Information). The solvent was removed under vacuum,
and addition of chloroform with further evaporation under vacuum helped the removal of 2ꢀmercaptoethanol.
The crude of the reaction was subsequently dissolved in 1 mL of nꢀhexane and loaded onto a preparative Agꢀ
TLC plate. A rapid and efficient separation was obtained using 1:9 nꢀhexane/diethyl ether with 0.4% MeOH as
mobile phase. After elution, the plate portion corresponding to the monotrans isomer fraction (Rf = 0.58),
detected by spraying a small portion of the plate with CAM, was scraped off. Silica was washed with
chloroform (3 x 5 mL). The solvent was evaporated to give a solid material, which is the Agꢀfatty acid
complex insoluble in nꢀhexane. This material was dissolved in a 5% aqueous solution of NH4OH, vigorously
stirred (600 rpm) for 10 min and extracted with aliquots of nꢀhexane (3 x 5 mL). Finally the organic phase was
dried over Na2SO4, filtered and evaporated. The unreacted allꢀcis DHAꢀMe was recovered and subjected to a
second isomerization cycle. This procedure leaded to the collection of a colourless oil that corresponded to the
monotrans DHA methyl ester isomer mixture (5.2 mg, 0.0152 mmol, overall yield from two isomerization
cycles 26%) (see Figure 2). GC analysis was carried out under the conditions described in Materials and
Methods (see Supporting Information, see trace II, Figure 4 or trace IV in Figure 6 of the main text). 13C NMR
(Figures S9 and S10 in Supporting Information) of the mixture of monotrans DHAꢀMe isomers obtained from
the isomerization and from the purification were also performed.
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
13C NMR (C6D6) ꢀ 10.94, 13.72, 14.12, 20.44, 20.56, 22.76, 25.26, 25.57, 25.66, 27.12, 28.01, 28.17, 28.57,
29.24, 30.39, 33.67, 33.71, 35.64, 45.29, 50.64, 50.73, 61.17, 126.93, 127.12, 127.48, 127.67, 127.86, 128.14,
128.48, 128.78, 128.81, 129.28, 129.37, 131.66, 131.68, 131.71, 131.75, 131.97, 131.99, 132.00, 132.23,
132.27, 132.49,132.52, 132.55, 172.35, 172.37.
ACS Paragon Plus Environment
13

Chemical Research in Toxicology
Page 14 of 32
1
2
3
4
5
6
7
8
9
After purification the 13C NMR of the pure monotrans fraction was carried out (see Figure S10 in Supporting
Information): 13C NMR (C6D6) ꢀ 14.10, 20.55, 22.75, 30.40, 33.65, 50.65, 127.47, 127.67, 127.86, 131.68,
131.72, 131.77, 132.02, 132.28, 172.33.
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
Isomerization of fish oil by photolysis. A sample of commercially available fish oil, used as omegaꢀ3
containing ingredient for food and nutraceutical preparation, was isomerized following the procedure above
described for the DHAꢀMe isomerization. After 30 minutes, 30 % of total trans content was obtained as
determined after transesterification to the corresponding FAME and GC analysis. In Supporting Information
the 13C NMR of the isomerized oil are reported (Figures S21 and S22). By GC analysis of the FAME obtained
from the triglyceride transesterification the cis and trans fatty acid contents were also determined. In Figure 6
of the main text the GC separation of the DHA isomers is shown, in comparison with the monotrans DHA
methyl ester isomer library. In Figure S23ꢀS26 in Supporting Information two different GC methods were
used with hydrogen or helium as carrier gas in order to evidence the best fatty acid separation.
Analysis of monotrans DHA isomers from commercially available supplements. Nineteen DHAꢀcontaining
supplements in softꢀgel capsules were examined for their cis and trans fatty acid contents, being six products
commercially available in Italy and thirteen in Spain. The oily content was taken from the capsules and the
lipids were extracted according to the Folch method [31]. Briefly, approx. 3 mg of each sample were dissolved
in 2:1 chloroform/methanol (3 mL) and partitioned with brine (1 mL); the extracted organic phases (3 x 3mL)
containing lipids were joined and dried with anhydrous Na2SO4. After filtration, the solvent was eliminated
using a rotary evaporator and the oily residue was left under vacuum for 10 minutes to remove any possible
trace of water. The lipid extract was checked by TLC (mobile phase: nꢀhexane/diethyl ether 7: 3) and then
treated with freshly prepared 0.5 M KOH/MeOH solution for 30 min at room temperature under stirring.
Reaction was quenched using brine and fatty acid methyl esters (FAME) were extracted using nꢀhexane (3x
3mL); the organic phase was collected and dried with anhydrous Na2SO4. After filtration and solvent
ACS Paragon Plus Environment
14

Page 15 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
evaporation, the FAME mixture was dissolved in 10 µL nꢀhexane and 1 µL was injected for GC analysis.
Tables S2 and S3 in Supporting Information report the FAME content: products 1ꢀ6 are omega 3
supplements present in the Italian market and products 7ꢀ19 are present in the Spanish market.
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
RESULTS AND DISCUSSION
The dual synthetic approach for DHA transformation and monotrans DHA isomer identification.
According to Scheme 1 we approached the dual synthetic strategy first by carrying out the direct radicalꢀ
catalysed isomerization of DHA methyl ester, as previously described by us for EPA 17, thus obtaining a
mixture of the six monotrans DHA isomers (Figure 2). AgꢀTLC was used to purify the monotrans DHA
isomers from the starting material (Figure 3, lane C). The recognition of each monotrans isomer was not
realizable at this stage, however satisfactory separation of these isomers could be obtained by gas
chromatography, the GC trace showing five out of six separable peaks, meaning that only two isomers are
superimposed. It is also worth noting that one monotrans DHA isomer elutes similarly to cisꢀDHA (Figure 4,
GC trace II). GC analysis, under the conditions described in Materials and Methods (see Supporting
Information: trace II, Figure 4 or trace IV in Figure 6 of the main text) gave a pattern of peaks similar to those
described in literature 4,5,20,22. At this point the assignment of each monotrans isomer could not be performed.
Only by elution of the commercially available 4EꢀDHAꢀMe, the third eluting peak could be assigned to this
isomer (Figure 4, trace I).
By the second synthetic route DHA methyl ester was transformed in twoꢀsteps to give the corresponding
monotrans alkenes (Scheme 1), via monoepoxide formation, ringꢀopening of the epoxide, using the
stereoselective formation of the dibromide derivatives, followed by elimination reaction, that provide the trans
geometry of each starting double bond. In this strategy, the key step is the separation and characterization of the
ACS Paragon Plus Environment
15

Chemical Research in Toxicology
Page 16 of 32
1
2
3
4
5
6
7
8
9
monoꢀepoxide products, in order to assign the double bond position for the subsequent elimination step. Metaꢀ
chloroperoxybenzoic acid (mꢀCPBA) epoxidation followed by purification by flash chromatography (with 9/1
nꢀhexane/diethyl ether as the eluent; see TLC separation with 7/3 nꢀhexane/diethyl ether in Figure S1,
Supporting Information) provided three fractions of monoepoxide regioisomers (Scheme 2). The structures of
the six regioisomers are shown in Figure 2.
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
Scheme 2. Twoꢀsteps transformation of allꢀcis DHAꢀMe into monotrans isomers (yields of the
products are in parenthesis): (a) mꢀCPBA, DCM, RT, 10 min; (b) dry DCM, pyridine, Ph3PBr2, 5°C,
overnight followed by (c) DMF, activated Zn, AcOH, 0°C, 10 h.
Figure 2. Structures of the six EDP regioisomers (left) and the corresponding monotrans DHA methyl ester
isomers (right).
ACS Paragon Plus Environment
16

Page 17 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
Under our experimental conditions, described in the Experimental Section, the 4ꢀmonoepoxide (compound 2)
was isolated in pure form as third fraction of the chromatography. The other regioisomers were obtained in the
first and second chromatographic fractions, each fraction as a mixture of two compounds. EDPs are
biologically relevant compounds obtained from DHA by cytochrome P450 enzyme, exhibiting a variety of
biological effects in inflammation, pain, angiogenesis and cancer 26. Actually cytochrome P450 is able to
provide the stereoselective epoxidation of DHA, preferably at the last double bond obtaining 19,20ꢀEDP 25.
However, also 16,17ꢀEDP, 13,14ꢀEDP, 10,11ꢀEDP and 7,8ꢀEDP have been detected and separated by HPLC
analysis, whereas 4,5ꢀEDP was reported to be unstable 27. To the best of our knowledge, the full EDP isomer
analysis by NMR spectroscopy has not been reported so far. We were interested to carry out monoꢀ and biꢀ
dimensional NMR experiments in order to assign the epoxide position in each compound of the reaction
mixture. It is worth noting that the use of deuterated benzene (C6D6) as NMR solvent is required, in order to
avoid decomposition during the experiment in the slightly acidic environment of deuterated chloroform. We
performed also the spectrum of DHAꢀMe in deuterated benzene for examining the epoxidation results using the
same experimental conditions (see Figure S2 in Supporting Information). Table S1 in Supporting
Information collects the 13C NMR data of the EDP isomers and the 10,11ꢀEDP regioisomer is not presented
since it was formed only in traces. The reasons for the scarce formation of this stereoisomer in the epoxidation
reaction were not investigated. In Supporting Information Figure S2 shows the structures of the five EDP
isomers with their proton and carbon atom resonances as assigned in this study using deuterated benzene as
solvent. The NMR spectra of all the five EDP isomers are reported in the Experimental Procedures. In order to
assign the monoepoxide resonances we examined 1H and 13C NMR resonances comparing with DHAꢀMe for
each fraction separated by column chromatography as described in the next section dedicated to NMR. The
subsequent transformation of monoꢀepoxides, forming in situ the corresponding dibromides followed by
elimination to the corresponding alkenes, was performed in oneꢀstep for each fraction. The latter reaction
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
17

Chemical Research in Toxicology
Page 18 of 32
1
2
3
4
5
sequence was adapted for DHA methyl ester from a procedure described for arachidonic acid methyl ester 10,14
and successfully applied to eicosapentanoic acid methyl ester (EPA) 18.
6
7
8
9
During the in situ transformation of epoxides to dibromide derivatives and the subsequent elimination to trans
alkenes, the concerted elimination is not the only occurring mechanism for which the double bond can be
formed. Therefore, the presence of cisꢀDHA was detected in all the resulting elimination crude mixtures.
Having recognized in each starting fraction the monoꢀepoxide structures as major and minor isomers, as
described in the NMR section and summarized in Scheme 2, the assignment of the corresponding monotrans
DHA isomers could be satisfactorily carried out examining the peaks present in the GC traces of the DHAꢀMe
isomers after the twoꢀsteps synthesis, as shown in Figure 4 (traces II, IV, V).
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
Figure 3. AgꢀTLC (eluent: 1/9 nꢀhexane/diethyl ether with 0.4% MeOH) of the transformation of one monoꢀ
44 epoxide fraction (second fraction) via bromination and elimination reactions. A: standard allꢀcis DHAꢀMe. B:
45
46
47
48
elimination reaction showing the presence of allꢀcisꢀDHAꢀMe and monotrans isomers (19E and 7E) of DHAꢀ
Me. C: mixture of monotrans DHAꢀMe obtained by thiylꢀradical catalysed isomerization after purification as
49
50
reported in Experimental Procedures.
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
18

Page 19 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
The isomer overlapping with DHAꢀMe is the 16E monotrans DHAꢀMe, and the 7E and 10E monotrans DHAꢀ
Me are also superimposed (lanes (III) and (IV) corresponding to the conversion of these two monoꢀepoxide
fractions according to Scheme 2). It is worth recalling that the 10E isomer was obtained from the traces of the
10,11ꢀEDP, formed with the lowest yield among all EDP isomers, as ascertained by the study of the NMR
spectra reported in the following section. Finally, the following GC order of elution could be assigned: 19E <
16E = allꢀcis < 4E < 13E < 7E = 10E.
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
Figure 4. GC chromatograms of the 39.0ꢀ42.5 min region referred to: (I) all cisꢀDHA methyl ester; (II)
mixture of monotrans DHA methyl ester isomers obtained by photolysis in the presence of thiyl radicals; (III)
16Eꢀisomer superimposed with all cisꢀDHA methyl ester, 13E and traces of 10E; (IV) 19E, all cisꢀDHA methyl
ester and 7E; and (V) 4E (commercially available product).
This is the first time the assignment of monoꢀtrans geometrical isomers of the omegaꢀ3 fatty acid DHA is
54 made, by a combination of AgꢀTLC, NMR spectroscopy, gas chromatography, and this result contributes to the
55
56
57
58
59
ACS Paragon Plus Environment
60
19

Chemical Research in Toxicology
Page 20 of 32
1
2
3
4
5
6
library of nonꢀcommercially available trans isomers developed by us for the longꢀchain polyunsaturated fatty
acids (LCꢀPUFA) omegaꢀ6 and omegaꢀ3.
7
8
9
NMR study of the EDP isomers and the monotrans DHA isomers. As previously mentioned, the 1H and
13C NMR spectra of the EDP and monotrans DHAꢀMe isomers obtained in this work (see Figure 2) were
carried out in C6D6, primarily used to avoid decomposition of the monoꢀepoxide compounds. The 1H NMR
spectra of the EDP regioisomers was very diagnostic for the terminal methyl group (Cꢀ22), that appears as
triplet in the 0.80ꢀ0.91 ppm region and is influenced by the relative position of the epoxide functionality. As
shown in Figure 5, in the starting material DHAꢀMe the triplet is centred at 0.885 ppm (trace I), whereas the
pure 4,5ꢀEDP isomer (isolated by chromatography) shows the triplet centred at 0.888 ppm (trace IV). Two
other fractions obtained from the chromatographic separation contained other four isomers (see Scheme 2): i) a
mixture with the 16,17ꢀEDP isomer with the triplet centred at 0.853 ppm and the 13,14ꢀEDP isomer with the
triplet centred at 0.880 ppm, with traces of the 10,11ꢀEDP isomer (Figure 5, trace II); ii) a mixture with the
19,20ꢀEDP isomer with the triplet at 0.838 ppm (major isomer) and the 7,8ꢀEDP isomer with the triplet at 0.892
ppm. The assignment of the triplet was made by correlation experiments, crossing the 1H and 13C resonances of
the methyl group and epoxide carbon atoms, respectively (see Table S2 in Supporting Information).
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
20

Page 21 of 32
Chemical Research in Toxicology
1
2
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
Figure 5. 1H NMR region (0.81ꢀ0.91 ppm) related to the terminal methyl group (Cꢀ22) of EDP regioisomers
(spectra run in C6D6). The triplet is centred at different ppm in relationship with the position of the epoxide
along the fatty acid chain; (I) DHAꢀMe allꢀcis; (II) mixture of 13,14ꢀEDP 5, 16,17ꢀEDP 6 (traces of 10,11ꢀEDP
4); (III) mixture of 7,8ꢀEDP 3 and 19,20ꢀEDP 7; (IV) 4,5ꢀEDP 2.
Examining the 13C NMR spectrum of the EDP regioisomers and using the assigned resonances of starting
DHAꢀMe, the carbon atom resonances of the epoxide function were determined. In fact, for the 16,17ꢀ
monoepoxide two peaks at 55.79 and 55.65 ppm can be assigned to the carbon atoms of the epoxide function,
since at the same time in the spectrum are absent the chemical shifts of the ethylenic carbon atoms assigned at
the C16ꢀC17 double bond of DHAꢀMe (128.50 and 127.91 ppm, respectively). The 13,14ꢀEDP has two peaks at
55.66 and 55.65 ppm assigned to the epoxide function, since the C13ꢀC14 double bond resonances of DHAꢀMe
at 128.22 and 128.11 ppm are not present. The 19,20ꢀEDP has the epoxide carbon atoms at 57.38 and 55.75
ppm, and correspondently the chemical shifts at 131.78 and 127.11 ppm (C19ꢀC20 of DHAꢀMe) are absent. In
the 19,20ꢀEDP regiosiomer the chemical shift of the Cꢀ22 is noticeably moved at 10.45 ppm (instead of the
ACS Paragon Plus Environment
21

Chemical Research in Toxicology
Page 22 of 32
1
2
3
4
range 14.08 – 14.00 ppm for all the other EDP regioisomers), indicating an upꢀfield shift due to a different
5
6
7
8
electronic distribution influencing the end of the carbon atom chain (i.e., the ω position). The 7,8ꢀEDP showed
the chemical shifts of the epoxy carbon atoms at 55.65 and 55.63 ppm, corresponding to the absence of the
resonances at 128.15 and 128.11 ppm (C7ꢀC8 in the DHAꢀMe). Finally, for the EDP in the position 4,5 the
epoxy function was individuated by the chemical shifts at 56.07 and 55.32 ppm corresponding to the absence of
the peaks at 129.08 and 128.06 ppm (C4ꢀC5 of DHAꢀMe).
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
Bidimensional experiments can also be performed to acquire more data on these assignments. A
representative description of the assignment of resonances to 7,8ꢀ and 19,20ꢀEDP isomers obtained by
HSQC/HMBC experiments is shown in Supporting Information (Figures S4 and S5). By this additional
information we could increase data on the epoxide assignments, however the above described 1H and 13C NMR
characteristics satisfactorily individuated the EDP regioisomers.
As far as the DHA monotrans isomers are concerned, they were obtained from the EDP two steps
transformation as shown in Scheme 2, as well as mixture of the six isomers by radicalꢀcatalyzed isomerization
of the allꢀcis DHAꢀMe. This latter mixture was examined by NMR and Figure 6 shows a trace of the
13
enlargement of the C NMR region corresponding to the Cꢀ20 resonance (traces I and II, panel b), which was
found to be diagnostic for the geometrical isomers. The Cꢀ20 atom is the most deshielded among the olefinic
peaks, and its distance from the position of the trans double bond in the fatty acid chain influences its chemical
shift. Five different chemical shifts for this carbon atom could be individuated in the monotrans isomer mixture
isolated from the isomerization (trace II in Figure 6, panel b), one of them representing two overlapped
resonances, and they are clearly different from the Cꢀ20 of DHAꢀMe appearing at 131.78 ppm (cfr., trace I in
Figure 6, panel b). The importance of the last vinylic carbon atom of the PUFA fatty acid chain was previously
described for arachidonic and eicosapentaenoic acids [10,18]. The present data support the use of NMR
spectroscopy for the analysis of natural and marine oils, as already shown by other groups 32ꢀ34 and recently
ACS Paragon Plus Environment
22

Page 23 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
reviewed 35. The herein described chemical synthesis of monotrans DHA isomers, as well as our previously
published work on trans lipid library 8ꢀ11, certainly increase availability of reference compounds and will help to
gather more information by the powerful NMR methodology.
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 6. Panel a) Partial GC traces of: (I) allꢀcis DHAꢀMe; (II) the six DHAꢀMe monotrans isomers and allꢀ
cis DHAꢀMe; (III) FAME obtained from the transesterification of DHAꢀcontaining fish oil. The presence of
monotrans DHA isomers as contaminants is visible when the trace is expanded (× 10); (IV) DHAꢀMe isomers
obtained after isomerization and transesterification of fish oil (this mixture contains monoꢀ, diꢀ and triꢀtrans
13
isomers). Panel b) Enlargement of the C NMR region (131.5ꢀ132.5 pm) corresponding to the Cꢀ20 ethylenic
carbon atom of: (I) allꢀcis DHAꢀMe; (II) the six DHAꢀMe monotrans isomers and allꢀcis DHAꢀMe; (III) DHAꢀ
containing triglycerides from fish oil; (IV) DHAꢀcontaining triglycerides from fish oil after isomerization (this
mixture contains monoꢀ, diꢀ and triꢀtrans isomers).
Isomerization of DHAꢀcontaining fish oil and determination of the monotrans DHA isomer content in
commercially available supplements. Natural oils are important sources of cis PUFA, and trans isomers of
ACS Paragon Plus Environment
23

Chemical Research in Toxicology
Page 24 of 32
1
2
3
4
EPA and DHA are contaminants after deodorization or distillation procedures that are used in order to eliminate
5
6
the fishy odour. PUFA isomerization in oils has been described by heating procedures 4,21 or pꢀtoluene sulfinic
7
8
9
28
30
acid treatment , whereas we reported the isomerization of some oils by thiyl radical catalysed process , but
not of fish oils. Here we performed the isomerization of a sample of DHAꢀcontaining fish oil, commercially
available as raw material for food or supplements. The reaction was carried out as described in the
Experimental section and GC analysis was performed after transesterification of the triglycerides into the
corresponding FAMEs. Figure 6 (panel a) shows the GC regions corresponding to FAMEs from fish oil in the
starting material (trace III) and it is worth noting that in this sample monoꢀtrans DHA isomers are already
present as contaminants, as it can be seen by comparison with the FAME of the oil after free radicalꢀcatalysed
isomerization (trace IV), and related cis and monotrans DHAꢀMe isomers (traces I and II). In the same Figure 6
(panel b) the enlarged region of the 13C NMR spectra corresponding to the Cꢀ20 (ethylenic carbon atom)
resonance is shown for the DHAꢀcontaining triglycerides of fish oil (trace III) in comparison with the allꢀcis
DHAꢀMe (trace I). It is gratifying to see that the Cꢀ20 chemical shift is similar for FAME and triglycerides that
contains the DHA moieties. Moreover, after fish oil isomerization, transꢀDHAꢀcontaining triglycerides show
strong similarity of the Cꢀ20 resonances for the monotrans DHAꢀMe isomers that were previously described
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
13
(cfr., trace IV with trace II). In Supporting Information the full C NMR spectra of the fish oil triglycerides
before and after the isomerization are shown (Figures S21 and S22).
It is worth mentioning that the GC resolution of monotrans isomers achieved under our analytical conditions,
using a specific oven program with a 60 m × 0.25 mm × 0.25 ꢁm (50%ꢀcyanopropyl)ꢀmethylpolysiloxane
column, were obtained after testing several gas chromatographic conditions, in particular concerning the carrier
gas. Using natural sources of fatty acids as marine oils, it was envisaged that other longꢀchain fatty acids can
elute in the region of the DHAꢀMe and its monotrans isomers. This was the case of C22ꢀPUFAs, such as 22:5
ωꢀ3 (DPA) and 22:6 ωꢀ6, as well as long chain saturated and monounsaturated fatty acids, such as C24:0 and
ACS Paragon Plus Environment
24

Page 25 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
C24:1. In Supporting Information two GC experimental conditions are described in Materials and Methods,
and Figures S23ꢀ26 show the peak resolution connected with these two methods. We were aware that a very
careful study was carried out with a 200 m SLB IL111 ionic liquid column verifying the accuracy of the
proposed method using Standard Reference Material (SRM) 3275 “Omegaꢀ3 and Omegaꢀ6 Fatty Acids in Fish
Oil” from the National Institute of Standards and Technology (NIST) 29. We could compare the reported results
with our method using hydrogen as carrier gas (trace III in Figure 6 and Figure S26 in Supporting
Information) and it was gratifying to see that interferences with other fatty acids can be avoided also under our
conditions. As already shown for EPA isomers 18, the use of hydrogen as carrier gas together with specific oven
temperature program is satisfactory, in terms of no interferences by C22ꢀPUFAs and long chain saturated and
monounsaturated fatty acids.
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
The combination of GC and 13C NMR analyses, the latter using the resonance of the Cꢀ20, can be of diagnostic
value for individuating monotrans DHA isomers, with the advantage that 13C NMR can be performed directly
on the triglycerides without any transformation. In this case, it is also worth noting that by studying the region
between 130.40 and 132.60 ppm, also in different NMR solvents, the overview of omegaꢀ6 and omegaꢀ3
monotrans isomer content can be obtained, as shown by previous and actual work by us. This overall picture
can be relevant for further developments in metabolomics research.
With these data in our hands we performed the analysis of the triglyceride fractions contained in DHAꢀ
containing supplements available on the markets in Italy and Spain. As matter of facts, analytical protocols to
run quality control of supplements are increasingly addressed for the safety of producers and consumers, and
researchers raised problems of oxidized or trans contaminants by evaluating commercial products of different
countries 28, 29. Table 1 shows our results in terms of fatty acids, such as omegaꢀ3 DPA, DHA, the DHA
monotrans isomers and the monounsaturated 24:1, detected in some representative supplements available in
Italian and Spanish markets, presenting low and highꢀcontent of monotrans DHA isomers. The values of these
ACS Paragon Plus Environment
25

Chemical Research in Toxicology
Page 26 of 32
1
2
3
4
fatty acids are shown as relative percentages (% rel) in the whole fatty acid composition determined on the
5
6
7
8
natural oil ingredient of the capsules. The full fatty acid analysis referred to 19 products of Italian and Spanish
markets are available in Supporting Information (Tables S2 and S3 and Figures S23ꢀ26 of representative GC
chromatograms).
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
Table 1. Representative fatty acid contents in terms of DPA, DHA, 24:1 and monotrans DHA isomers eluting
in the GC time window 37ꢀ41.5 min (see Figure 6 and Figures S23ꢀ26) obtained from natural oils of
commercially available supplements in Italy and Spain. The values are expressed as percentages relative to
100% of the fatty acid peak areas identified in the GC analysis. In Tables S2 and S3 (Supporting
Information) the complete analysis for 19 products is reported.
1
2
4
5
8
13
14
15
19
22:6 ω3 transꢀ∆19
22:5 ω3 (DPA)
22:6 ω3 (DHA)
22:6 ω3 transꢀ∆4
22:6 ω3 transꢀ∆13
24:1
0.12±0.01
3.67±0.07
18.51±0.10
0.08±0.00
0.19±0.02
0.34±0.03
0.26±0.04
0.06±0.00
4.31±0.02
20.64±0.37
0.01±0.00
0.02±0.00
0.10±0.00
0.02±0.00
0.13±0.08
3.91±0.18
20.98±0.07
0.07±0.00
0.14±0.03
0.66±0.01
0.29±0.02
0.12±0.04
0.15±0.02
57.35±0.15
0.17±0.00
0.33±0.01
nd
0.09±0.01
2.10±0.03
14.60±0.15
0.08±0.00
0.09±0.01
0.24±0.01
0.07±0.00
0.27±0.03
9.94±0.32
83.68±0.17
0.09±0.03
0.10±0.04
0.14±0.03
0.20±0.03
0.34±0.03
1.17±0.02
86.19±0.11
0.16±0.02
0.46±0.03
nd
0.31±0.02
13.32±0.04
78.78±0.11
0.18±0.03
0.35±0.02
nd
0.66±0.24
2.09±0.33
76.75±0.41
0.30±0.01
0.74±0.02
0.34±0.14
1.00±0.14
22:6 ω3 transꢀ∆7+∆10
0.50±0.02
0.68±0.01
0.49±0.03
TOTAL
DHA
monotrans
0.75±0.04
0.11±0.00
0.63±0.08
1.12±0.04
0.33±0.01
0.66±0.04
1.64±0.03
1.33±0.03
2.70±0.25
In Table 1 is shown that the percentage of monotrans DHA isomers was found between 0.11% and 2.70%, with
the highest trans contents contained in the DHAꢀrichest formulas (cfr., product 19 and product 2 in Table 1).
This would suggest that the procedures in order to enrich DHA content in the oil be responsible of raising the
isomer presence in the resulting material. Our studies on the trans lipid library support previously reported
analyses of PUFAꢀcontaining oils and supplements, discovering contaminations either as oxidation and
isomerization products [4,5,20,21,28,29], addressing the important issue of the precise monotrans isomer
ACS Paragon Plus Environment
26

Page 27 of 32
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
identification in case of EPA and DHA. This is strongly needed especially because health benefits are highly
discussed for omegaꢀ3 supplements, whereas contaminants can certainly annihilate positive effects in a doseꢀ
proportional manner, as shown by us in a murine model 15.
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
In conclusion, here we showed the extension of the trans lipid library to include an important long chain PUFA
such as DHA, with the individuation of its structural changes due to chemical treatment as well as free radical
reactivity, converting the cis geometry into trans. It is expected that this new knowledge will successfully
integrate other analytical methodologies for the quality control of trans fats in health care and food products
needed by regulatory bodies all over the world 36.
Conflict of Interest. The authors declare no conflict of interests.
Supporting Information. The following files are available free of charge: a) EDPꢀregioisomer identification
with full assignment of 1H and 13C NMR resonances and examples of the 2D NMR experiments to corroborate
assignments; b) experimental details of the monotrans DHA isomer preparation, separation and identification;
c) further analytical information on triglyceride isomerization; d) fatty acid contents of commercially available
supplements and representative gas chromatographic analyses in different conditions.
AUTHOR INFORMATION
Corresponding Author
*Carla Ferreri; email: carla.ferreri#@isof.cnr.it
Author Contributions
ACS Paragon Plus Environment
27

Chemical Research in Toxicology
Page 28 of 32
1
2
3
4
5
6
The manuscript was written through contributions of all authors. All authors have given approval to the final
version of the manuscript. ‡These authors contributed equally.
7
8
9
Funding Sources
10
11
12
13
14
15
16
17
18
19
20
21
22
23
The funding of the MSCAꢀITNꢀ2014ꢀETN ꢀ Marie SkłodowskaꢀCurie ITN project CLICKGENE (#642023)
supported research activity as part of PhD training program of GM and GG. RS received a grant from the Di
Bella Foundation. This work was partially funded by the Basque Country Government (Department of the
Environment, Regional Planning, Agriculture and Fisheries). J.A. is recipient of a PhD grant from AZTI.
ACKNOWLEDGMENT
24 C.F. is grateful to Dr. Matxalen Uriarte (AZTI) for her strong commitment to introduce the subjects of
25
26
lipidomic analyses and personalized health in the Basque Country.
27
28
29
30
31
32
33
ABBREVIATIONS
34
35
ACOH, acetic acid; AgꢀTLC, silverꢀthin layer chromatography, CAM, cerium ammonium sulfate/ammonium
36
molibdate reagent; C6D6, deuterated benzene; CDCl3, deuterated chloroform; DCM, dichloromethane; DHA,
docosahexaenoic acid; DPA, docosapentaenoic acid; EDP, epoxydocosapentaenoic acid; EPA,
eicosapentaenoic acid FAME, fatty acid methyl esters; mꢀCPBA, metaꢀchloroperoxybenzoic acid; PUFA,
polyunsaturated fatty acid; TFA, trans fatty acid.
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
REFERENCES
52 (1) Siddiqui, R. A., Harvey, K., and Stillwell, W. (2008) Anticancer properties of oxidation products of
53
54
55
56
57
58
59
60
docosahexaenoic acid. Chem. Phys. Lipids 153, 47–56.
ACS Paragon Plus Environment
28

Page 29 of 32
Chemical Research in Toxicology
1
2
3
4
5
(2) Scientific Opinion of the Panel on Dietetic products, Nutrition and Allergies on a request from European
Commission related to labelling reference intake values for nꢀ3 and nꢀ6 polyunsaturated fatty acids. (2009) The
EFSA Journal 1176, 1–11.
6
7
8
(3) Sanders, TA. (2009) DHA status of vegetarians. Prostaglandins Leukot. Essent. Fatty Acids 81, 137–141.
9
(4) Fournier, V., Destaillats, F., Hug, B., Golay, P.A., Joffre, F., Juanéda, P., Sémon, E., Dionisi, F., Lambelet,
P., Sébédio, J.L., and Berdeaux, O. (2007) Quantification of eicosapentaenoic and docosahexaenoic acid
geometrical isomers formed during fish oil deodorization by gasꢀliquid chromatography. J. Chromatogr. A
1154, 353–359.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
(5) Sciotto, C., and Mjøs, S.A. (2012) Trans isomers of EPA and DHA in omegaꢀ3 products on the European
market. Lipids 47, 659–667.
(6) Liu, X.ꢀR., Deng, Z.ꢀY. Deng, Hu, J.ꢀNꢀ, and Fan, Y.ꢀW. (2013) Erythrocyte membrane transꢀfatty acid
index is positively associated with a 10 year CHD risk probability. Br. J. Nutr. 109, 1695–1703.
(7) Anderson, A. K., McDougald, D. M., and SteinerꢀAsiedu, M. (2010) Dietary trans fatty acid intake and
maternal and infant adiposity. Eur. J. Clin. Nutr. 64, 1308–1315.
(8) Chatgilialoglu, C., Ferreri, C., Melchiorre, M., Sansone, A., and Torreggiani, A. (2014) Lipid geometrical
isomerism: from chemistry to biology and diagnostics. Chem Rev. 114, 255–284.
(9) Chatgilialoglu, C., Ferreri, C., Guerra, M., Samadi, A., Bowry, V. (2017) The Reaction of Thiyl Radical
with Methyl Linoleate: Completing the Picture. J. Am. Chem. Soc. 139, 4704–4714.
(10) Ferreri, C., Sassatelli, F., Samadi, A., Landi, L., and Chatgilialoglu, C. (2004) Regioselective cisꢀtrans
isomerization of arachidonic double bonds by thiyl radicals: the influence of phospholipid supramolecular
33 organization. J. Am. Chem. Soc. 126, 1063−1072.
34
35 (11) Ferreri, C., Faraone Mennella, M. R., Formisano, C., Landi, L., and Chatgilialoglu, C. (2002) Arachidonate
36
37
38
39
40
41
42
43
44
45
46
47
geometrical isomers generated by thiyl radicals: the relationship with trans lipids detected in biological
samples. Free Radic. Biol. Med. 33, 1516−1526.
(12) Zambonin, L., Ferreri, C., Cabrini, L., Prata, C., Chatgilialoglu, C., and Landi, L. (2006) Occurrence of
trans fatty acids in rats fed a transꢀfree diet: a free radicalꢀmediated formation? Free Radic. Biol. Med. 40,
1549–1556.
(13) Ferreri, C., Kratzsch, S., Brede, O., Marciniak, B., and Chatgilialoglu C. (2006) Trans lipid formation
induced by thiols in human monocytic leukemia cells. Free Radic. Biol. Med. 38, 1180–1187.
(14) Souabni, H., Thoma, V., Bizouarn, T., Chatgilialoglu, C., SiafakaꢀKapadai, A., Baciou, L., Ferreri, C.,
48 HouéeꢀLévin, C., and Ostuni, M.A. (2012) Trans arachidonic acid isomers inhibits NADPHꢀoxidase activity by
49
50
51
52
53
54
55
56
57
58
59
60
direct interaction with enzyme components. BBA Biomembranes 1818, 2314–2324.
(15) Aoun, M., FeilletꢀCoudray, C., Fouret, G., Chabi, B., Crouzier, D., Ferreri, C., Chatgilialoglu, C., Cabello,
C., Cristol, J. P., Carbonneau, M.A., and Coudray, C. (2012) Rat liver mitochondrial membrane characteristics
and mitochondrial functions are more profoundly altered by dietary lipid quantity than by dietary lipid quality:
effect of different nutritional lipid patterns. Br. J. Nutr. 107, 647–659.
ACS Paragon Plus Environment
29