Synthesis of all-trans arachidonic acid and its effect on rabbit platelet aggregation

Article abstract of DOI:10.1016/j.bmcl.2005.03.109

A simple and high-yielding method to convert natural all-cis PUFA derivatives to the corresponding all-trans geometrical isomers is described. The method is based on the thiyl radical-catalyzed cis-trans isomerization. The all-trans isomer of arachidonic acid was found to cause rabbit platelet aggregation at concentrations higher than 0.1 mM and inhibition of PAF-induced platelet aggregation in a concentration dependent manner with an IC₅₀ in the micromolar range.

Full text of DOI:10.1016/j.bmcl.2005.03.109

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2766–2770
Synthesis of all-trans arachidonic acid and its effect on
rabbit platelet aggregation
Dimitris Anagnostopoulos,a Chryssostomos Chatgilialoglu,b Carla Ferreri,b,*
Abdelouahid Samadib and Athanassia Siafaka-Kapadaia
aDepartment of Chemistry, University of Athens, Panepistimiopolis, 15771 Athens, Greece
bISOF, Consiglio Nazionale delle Ricerche, Via P. Gobetti 101, 40129 Bologna, Italy
Received 3 February 2005; revised 21 March 2005; accepted 29 March 2005
Available online 3 May 2005
Abstract—A simple and high-yielding method to convert natural all-cis PUFA derivatives to the corresponding all-trans geometrical
isomers is described. The method is based on the thiyl radical-catalyzed cis–trans isomerization. The all-trans isomer of arachidonic
acid was found to cause rabbit platelet aggregation at concentrations higher than 0.1 mM and inhibition of PAF-induced platelet
aggregation in a concentration dependent manner with an IC50 in the micromolar range.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
conditions, has given a multidisciplinary meaning to
the formation of trans lipids.4–6
Lipids are central to the cell membrane composition due
to their amphiphilic nature, and to regulation and con-
trol of cellular functions through their signaling activ-
ity.1,2 There is currently much research interest in lipid
signaling cascades, which mediate diverse physiological
and pathological responses. In this context, polyunsatu-
rated fatty acid (PUFA) structures with methylene-inter-
rupted double bonds in the cis configuration are known
to be crucial for enzyme binding.3 In recent years, the
role of the naturally occurring cis lipid geometry has at-
tracted attention, in particular for the cis–trans isomeri-
zation process catalyzed by free radical species.4
It is worth recalling that trans lipids are studied as die-
tary components, which can exert harmful effects on hu-
man health.7 In this view, trans lipid metabolism and
biological activity have been examined and the incorpo-
ration of these isomers in different lipid structures,
including membrane phospholipids, is well docu-
mented.8,9 The total trans content in a sample of PUFA
can be easily determined from its infrared spectrum, be-
cause the unconjugated trans double bonds show a spe-
cific band.10 On the other hand, the access to each trans
isomer is a difficult task. Eicosenoates that have a 20
carbon atom chain with a high number of unsaturations
(up to five double bonds) are one of the most important
classes of lipid mediators, and each isomer can have a
biological activity. In fact, two isomers of arachidonic
acid (AA), namely the mono-trans isomers in positions
5 and 14 prepared by total synthesis, were tested in lipid
enzymatic cascades.11,12
Scheme 1 shows the isomerization mechanism catalyzed
by thiyl radicals attacking an isolated double bond. The
fact that the conversion of a cis to the more stable trans
configuration is very efficient also under biomimetic
The 14-trans isomer decreased platelet aggregation and
inhibited prostaglandin biosynthesis,13,14 whereas the
5-trans isomer did not cause inhibition and could be
transformed to 5-trans-PGE2.11 Each of the four
possible mono-trans isomers of arachidonic acid has
Scheme 1. Mechanism of cis–trans isomerization catalyzed by thiyl
radicals.
Keywords: Radical isomerization; trans PUFA; trans Lipids; Platelet aggregation.
*
Corresponding author. Tel.: +39 051 6398289; fax: +39 051 6398349; e-mail: cferreri@isof.cnr.it
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.109

D. Anagnostopoulos et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2766–2770
2767
been recently prepared by a multistep stereospecific
protocol, and their interaction with enzymes was
demonstrated.14,15
Based on our studies of thiyl radical-catalyzed isomeri-
zation process with formation of mono- and polyunsat-
urated trans fatty acid derivatives,4–6 we thought that
the synthesis of all-trans PUFA could be of interest,
since a single geometrical isomer is produced, and is ex-
pected to have pharmacological applications. From the
known reaction profile of arachidonic acid isomeriza-
tion,6 it is derived that several cis/trans isomeric compo-
sitions can be obtained at different reaction times and a
straightforward synthesis of all-trans isomers could be
envisaged, by an exhaustive isomerization process.
In this letter, we on report the preparation of all-trans
isomer of arachidonic acid (t-AA) and its corresponding
methyl ester (t-AAME) and the results of biological
activity assays for rabbit platelet aggregation
(Scheme 2).
2. Synthesis of t-AA and t-AAME
Figure 1. GC runs: (A) before and (B) after isomerization. Peak labels:
(1) methyl stearate (18:0) as reference; (2) AAME; (3) t-AAME.
Among the possible isomerization protocols, the radi-
cal-catalyzed methodology was found to convert specifi-
cally the cis double bonds into their geometrical trans
counterparts.16 A 2.2 mL solution of 15 mM AA in i-
PrOH was placed in a quartz photochemical reactor
and bubbled with nitrogen for 20 min. The thiol (2-
mercaptoethanol) in a 50% mole equivalent with respect
to AA was added and the mixture was exposed to low-
pressure mercury lamp (5.5 W) at 25 ꢁC. A TLC moni-
toring of the reaction under appropriate conditions,
showed that t-AA was formed as the major product in
40 min, by a step-by-step conversion of the mono-, di-,
and tri-trans isomers progressively formed in the reac-
tion mixture.6b,17 After workup and purification, t-AA
was obtained in an 85% yield (purity >97%). The acid
was also converted to the corresponding methyl ester
(t-AAME) by diazomethane for GC/MS analysis. Simi-
larly, arachidonic acid methyl ester (AAME) was trans-
formed to all-trans isomer (t-AAME). Figure 1 shows
the different retention times of AAME (run A) and t-
AAME (run B) in GC, the reference being methyl stea-
rate (peak 1).
it is known that NMR spectra of cis and trans isomers
have significant differences in the ethylenic carbon atom
resonances, and bisallylic and allylic hydrogen chemical
shifts.19,20
In Figure 2, the comparison of 13C NMR regions be-
tween AAME (spectrum A) and t-AAME (spectrum
B) is given and all ethylenic carbon atom resonances
of the trans isomer appeared at lower fields than the cor-
responding cis isomer.19 In Figure 3, an enlargement of
NMR analysis provides significant information on the
differences between the two isomeric structures. Indeed,
Scheme 2. Structure and abbreviation of arachidonic acid (AA) and its
all-trans isomer (t-AA).
Figure 2. Ethylenic carbon atom region of the 13C NMR spectra of
AAME (A) and t-AAME (B).

2768
D. Anagnostopoulos et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2766–2770
radical in turn reacted with the thiol to give thiyl radicals
(Eq. 3). It is worth underlining that reaction (3) is
reversible with an equilibrium constant of K = 1 · 104.
hm
Å
Å
HOCH2CH2SH ! HOCH2CH2S þ H
ð1Þ
Å
Å
H þ ðCH3Þ CHOH ! H2 þ ðCH3Þ Cð ÞOH ð2Þ
2
2
Å
ðCH3Þ Cð ÞOH þ HOCH2CH2SH
2
Å
ꢀ ðCH3Þ CHOH þ HOCH2CH2S
ð3Þ
2
Under our experimental conditions, that is, [i-PrOH] =
13.06 M and [HOCH2CH2SH] = 0.007 M, the equilib-
rium is still shifted to the right but the forward reaction
is only 5–6 times faster than the reverse reaction.6a
Figure 3. 1H NMR spectra of AAME (A) and t-AAME (B) in the
region 1.5–3.0 ppm. Bisallylic hydrogens are indicated with an asterisk.
1H NMR spectrum in the region of 1.5–3.0 ppm of
AAME (A) and t-AAME (B) is shown, and the
difference for the two isomers is considerable, being
the trans bisallylic hydrogen signal at higher fields, as
expected.20
3. Biological activity of t-AA and t-AAME
The physiological role of AA in the aggregation cascade
is well known and some of its mono-trans isomers were
found to be inhibitors of this process.21,22 Although
mono-trans arachidonic acid isomers have been detected
in human plasma,23 there is no evidence and it is highly
improbable that all-trans isomer could be generated in
the blood. We tested the effect of t-AA and t-AAME
on rabbit platelet aggregation. We also focused on
PAF-induced platelet aggregation since cis fatty acids,
but not saturated and trans fatty acids, were reported
From a mechanistic point of view, thiyl radicals were
generated by means of direct photolysis of 2-mercap-
toethanol in deoxygenated i-PrOH solutions.6a In this
condition, the generation of thiyl radicals and H atoms
occurred as shown in Eq. 1, and the H atoms were
efficiently quenched by the solvent. The (CH3)2C(Å)OH
Figure 4. Aggregation effect of t-AA on rabbit platelets. t-AA (in ethanol) at concentrations indicated, was added to cuvettes containing 0.5 mL of
prewarmed (37 ꢁC) platelet suspension (2.5 · 108 cells/mL). Control aggregation curves with PAF and thrombin were obtained after preincubation
with ethanol.

D. Anagnostopoulos et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2766–2770
2769
as inhibitors of this path.24 t-AA was added at
several concentrations to the platelet suspension
(2.5 · 108 cells/mL) and the results are shown in Figure
4. Control aggregation curves with platelet aggregation
factor PAF (4.0 · 10ꢀ10 M) and thrombin (0.5 U/mL)
were also obtained after preincubation with ethanol
for 1 min. t-AA induced rabbit platelet aggregation in
a concentration higher than 2.5 · 10ꢀ4 M. At high,
non-physiological concentrations, cell viability was al-
ways checked without detecting any toxicity or lysis of
platelets. In Figure 4, runs b, c, and d correspond
to 18%, 63%, and 78% aggregation, respectively. At
concentration of 1.5 · 10ꢀ4 M (run a), it could be also
observed that t-AA induces a shape change on rabbit
platelets. Aggregation was not 100% reversible even at
concentrations that induce a small aggregation (e.g.,
2.5 · 10ꢀ4 M). When thrombin (0.5 U/mL) was added
after desaggregation, platelets were activated again and
aggregated irreversibly, suggesting that the high concen-
trations of t-AA were not toxic. For comparison, under
the same experimental conditions, 4.0 · 10ꢀ10 M
PAF and 0.5 U/mL thrombin induced 78% and 67%
platelet aggregation, respectively, whereas AA induced
platelet aggregation at a 10 times lower concentration
(not shown). t-AA activity was also assayed on the
PAF-induced aggregation of rabbit platelets and Table
1 shows the results of different concentrations. t-AA
was found to be inactive under the same experimental
conditions (not shown).
In conclusion, we have described a straightforward and
simple access to all-trans PUFA molecules and in partic-
ular, t-AA was prepared and tested for rabbit platelet
aggregation showing to be active only at high concentra-
tions (>10ꢀ4 M). More interestingly, at micromolar con-
centrations it caused inhibition of aggregation induced
by the strong platelet agonist PAF, which was not pre-
dictable from the known behavior of other trans-mono-
unsaturated fatty acids, and apparently it could be due
to the unique geometry of the polyunsaturated molecule.
Indeed, fatty acid structural modifications can lead to a
change in the biological activity and the presence of
trans double bonds is known to confer a more linear
molecular shape, compared to the cis configuration,
which usually forms a kink in the carbon atom chain.
All-trans fatty acids appear to be interesting candidates
for a chemical biology approach investigating the molec-
ular mechanism of PAF-induced platelet aggregation,
which is currently under study.
Acknowledgements
This work was supported in part by the European Com-
munityÕs Human Potential Program under contract
HPRN-CT-2002-00184 [SULFRAD], by HPMF-CT-
2001-01313 (Marie Curie individual fellowship to A. Sa-
madi), and by the University of Athens Special Account
for Research Grants (70/4/3351).
caused inhibition of
a
4 · 10ꢀ10 M PAF-induced
aggregation in a concentration dependent manner,
with IC50 ꢁ 6 · 10ꢀ5 M. t-AA at 2.1 · 10ꢀ4 M concen-
tration caused 100% inhibition of PAF 8 · 10ꢀ10 M.
It should be recalled that AA at micromolar concentra-
tions caused platelet aggregation. CP/CPK caused inhi-
bition (100%) of the aggregation induced by t-AA
(2.8 · 10ꢀ4 M), suggesting the involvement of ADP re-
lease. This hypothesis needs further investigation.
References and notes
1. Lipoproteins and Membranes; Vance, D. E., Vance, J. E.,
Eds.; Elsevier: Amsterdam, 2002.
2. Funk, C. D. Science 2001, 294, 1871–1875.
3. Malkowski, M. G.; Ginell, S. L.; Smith, W. L.; Garavito,
R. M. Science 2000, 289, 1933–1937.
4. Chatgilialoglu, C.; Ferreri, C. Acc. Chem. Res. 2005, Web
Release Date: Mar 09, 2005.
5. (a) Chatgilialoglu, C.; Altieri, A.; Fischer, H. J. Am.
Chem. Soc. 2002, 124, 12816–12823; (b) Ferreri, C.;
Costantino, C.; Perrotta, L.; Landi, L.; Mulazzani, Q.
G.; Chatgilialoglu, C. J. Am. Chem. Soc. 2001, 123, 4459–
4468.
6. (a) Ferreri, C.; Sassatelli, F.; Samadi, A.; Landi, L.;
Chatgilialoglu, C. J. Am. Chem. Soc. 2004, 126, 1063–
1072; (b) Ferreri, C.; Faraone-Mennella, M. R.; Formi-
sano, C.; Landi, L.; Chatgilialoglu, C. Free Radical Biol.
Med. 2002, 33, 1516–1526; (c) Chatgilialoglu, C.; Ferreri,
C.; Ballestri, M.; Mulazzani, Q. G.; Landi, L. J. Am.
Chem. Soc. 2000, 122, 4593–4601.
The effect of t-AAME on rabbit platelet aggregation was
not found to be significant. In fact, it did not cause
rabbit platelet aggregation at concentration between
0.2 · 10ꢀ4 M and 6.0 · 10ꢀ4 M indicating that
a
free form of acid is necessary for the aggregation
effect.
Furthermore, it did not cause inhibition of PAF–in-
duced aggregation at any of the concentrations tested
in the range 2.0–4.0 · 10ꢀ4 M. It should be also added
that the respective saturated fatty acid (arachidic acid)
Table 1. Effect of t-AA on rabbit platelet aggregation induced by
4 · 10ꢀ10 M PAF
´ ´
7. Trans Fatty Acids in Human Nutrition; Sebedio, J. L.,
Christie, W. W., Eds.; The Oily: Dundee, 1998.
8. When trans isomers of linoleic acid derived from altered
fats are ingested and metabolized,9 arachidonate moieties
found in tissues contain the 11 and 14 double bonds in the
trans configuration.
t-AA M
Inhibition %
1.5 · 10ꢀ6
1.5 · 10ꢀ5
3.0 · 10ꢀ5
7.5 · 10ꢀ5
2.1 · 10ꢀ4
0
6
24 ( 5)
61 ( 8)
100
9. (a) Beyers, E. C.; Emken, E. A. Biochim. Biophys. Acta
´ ´
1991, 1082, 275–284; (b) Sebedio, J. L.; Vermunt, S. H. F.;
Chardigny, J. M.; Beaufrere, B.; Mensink, R. P.; Arm-
`
strong, R. A.; Christie, W. W.; Niemela¨, J.; Henon, G.;
Riemersma, R. A. Eur. J. Clin. Nutr. 2000, 54, 104–113; (c)
Values are means of three experiments. Standard deviation is given in
parentheses. Aggregation given by 4 · 10ꢀ10 M PAF is considered
100%.

2770
D. Anagnostopoulos et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2766–2770
´ ´
Chardigny, J. M.; Sebedio, J. L.; Juaneda, P.; Vatele, J.
M.; Grandgirard, A. Nutr. Res. 1993, 13, 1105–1111.
`
with cerium ammonium sulfate/ammonium molybdate
reagent. The t-AA fraction was separated and silica gel
was washed using chloroform:methanol (2:1 v/v). After
filtration the organic phase was evaporated to afford a
complex of Ag/arachidonic acid as a white powder
insoluble in n-hexane. The complex was dissolved with
5% aq NH4OH (3 mL) and extracted with chloroform.
The organic phase was dried over anhydrous sodium
sulfate and concentrated under vacuum. t-AA (oil; 85%
yield) was transformed into its corresponding methyl ester
(t-AAME) by diazomethane. 1H and 13C NMR spectra
were recorded on a Varian VXR (400 and 100.6 MHz)
instrument using CDCl3 as the solvent and the reference
peak and they are shown in Figures 2 and 3. GC/MS (m/z),
318 (M+), 275, 262, 250, 217, 203, 177, 161, 147, 133, 105,
93, 79, 55.
10. McDonald, R. E.; Mossoba, M. M. In Food Lipids
Chemistry, Nutrition and Biotechnology; Akoh, C. C., Min,
D. B., Eds.; Marcel Dekker: New York, 2002, pp 169–204.
11. Egmond, G. J. N.; Pabon, H. J. J.; van Dorp, D. A. Rec.
Trav. Chim. Pays-Bas 1977, 96, 172–175.
`
12. Berdeaux, O.; Vatele, J.-M.; Eynard, T.; Nour, M.;
Poullain, D.; Noe¨l, J. P.; Sebedio, J. L. Chem. Phys.
´ ´
Lipids 1995, 78, 71–80.
13. Berdeaux, O.; Chardigny, J. M.; Sebedio, J. L.; Mairot, T.;
´ ´
Poullain, D.; Vatele, J. M.; Noe¨l, J. P. J. Lipid Res. 1996,
37, 2244–2250.
`
14. Krishna, U. M.; Reddy, M. M.; Xia, J.; Falck, J. R.;
Balazy, M. Bioorg. Med. Chem. Lett. 2001, 11, 2415–2418.
15. Roy, U.; Loreau, O.; Balazy, M. Bioorg. Med. Chem. Lett.
2004, 14, 1019–1022.
18. Williams, C. M.; Mander, L. N. Tetrahedron 2001, 57,
425–447.
19. Bus, J.; Sies, I.; Lie Ken Jie, M. S. F. Chem. Phys. Lipids
1997, 18, 130–144.
20. Frost, D. J.; Barzilay, J. Rec. Trav. Chim. 1971, 90, 705–
712.
21. Siess, W. Physiol. Rev. 1989, 69, 58–178.
22. Samuelsson, M.; Goldfine, M.; Granstro¨m, E. Annu. Rev.
Biochem. 1978, 997–1029.
23. Zghibeh, C. M.; Gopal, V. R.; Poff, C. D.; Falck, J. R.;
Balazy, M. Anal. Biochem. 2004, 332, 137–144.
24. Siafaka-Kapadai, A.; Hanahan, D. Biochim. Biophys. Acta
1993, 1166, 217–221.
16. Dugave, C.; Demange, L. Chem. Rev. 2003, 103, 2475–2532.
17. The formation of t-AA acid was monitored by Ag/TLC18
using n-pentane–dichloromethane–acetone (2:1:1, v/v/v) as
the eluent, and by GC analysis after transesterification by
diazomethane.6a,b After 40 min photolysis, TLC showed
the formation of t-AA as the major product, and the
reaction was stopped and concentrated under vacuum.
The residue was treated three times with absolute ethanol
and evaporated under vacuum in order to be thiol-free.
The crude t-AA was then purified by preparative Ag/TLC
with the eluent described above. The fractions were
detected by spraying the plate (or a portion of the plate)