Eicosapentaenoic-acid-derived isoprostanes: Synthesis and discovery of two major isoprostanes

Article abstract of DOI:10.1016/j.bmcl.2008.09.008

The stereospecific synthesis of two all-syn-EPA-derived isoprostanes (iPs), 5-epi-8,12-iso-iPF_3α-VI 17 and 8,12-iso-iPF_3α-VI 18, has been accomplished. These two synthetic probes have been used to discover and identify their presence in human urine. The eventual quantitative measurement of these two iPs may be a valuable index of oxidative stress in people with eicosapentaenoic acid- (EPA) and docosahexaenoic acid- (DHA) enriched phospholipids.

Full text of DOI:10.1016/j.bmcl.2008.09.008

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5523–5527
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Eicosapentaenoic-acid-derived isoprostanes: Synthesis and discovery
of two major isoprostanes
Chih-Tsung Chang a, Pranav Patel a, Namin Kang a, John A. Lawson b, Wen-Liang Song b, William S. Powell c,
Garret A. FitzGerald b, Joshua Rokach a,
*
a Claude Pepper Institute and Department of Chemistry, Florida Institute of Technology, 150 West University Boulevard, Melbourne, FL 32901, USA
b Institute for Translational Medicine and Therapeutics and Department of Pharmacology, University of Pennsylvania, Philadelphia, PA 19104, USA
c Meakins-Christie Laboratories, Department of Medicine, McGill University, 3626 St. Urbain Street, Montreal, Que., Canada H2X 2P2
a r t i c l e i n f o
a b s t r a c t
Article history:
The stereospecific synthesis of two all-syn-EPA-derived isoprostanes (iPs), 5-epi-8,12-iso-iPF3 -VI 17 and
a
Received 7 August 2008
Revised 2 September 2008
Accepted 3 September 2008
Available online 6 September 2008
8,12-iso-iPF3 -VI 18, has been accomplished. These two synthetic probes have been used to discover and
a
identify their presence in human urine. The eventual quantitative measurement of these two iPs may be a
valuable index of oxidative stress in people with eicosapentaenoic acid- (EPA) and docosahexaenoic acid-
(DHA) enriched phospholipids.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Eicosapentaenoic acid
EPA
Docosahexaenoic acid
DHA
Arachidonic acid
AA
Isoprostane
iP
Isoprostanes (iPs) are natural products, isomeric with prosta-
glandins (PGs). Unlike PGs they are produced non-enzymatically
by a free-radical oxygenation of arachidonic acid (AA), a polyunsat-
urated fatty acid1,2 (Scheme 1), mostly esterified to phospholipids.
The mechanism of iP formation has been discussed in some
detail.3–5
The formation of these oxygenation products has been related
to oxidant stress and inflammatory diseases such as atherosclero-
sis6 and Alzheimer’s.2,7 Whereas a substantial amount of work has
been focused on AA-generated iPs, little attention has been given
to eicosapentaenoic-acid- (EPA) derived iPs.8,9 We are reporting
here on the synthesis and identification of two major all-syn-
population with high EPA and DHA intake through their high fatty
fish diets, for example, Inuits, etc. The rationale for their use is to
enrich the phospholipids with EPA and DHA and reduce the per-
centage of other polyunsaturated fatty acids, especially arachi-
donic acid. EPA and DHA are somewhat poorer substrates for
cyclooxygenase than AA and are also partial enzyme inhibitors,
with one result being the diminution of TXA2 production and re-
duced platelet aggregation. This may lead to reduced cardiovascu-
lar incidents such as heart attacks.
The AA-derived iPs we and others have measured so far as an
index of oxidant stress10,11 may not be representative for people
in which the major polyunsaturated fatty acids in the phospholip-
ids are EPA and DHA. A more appropriate index in those cases
might be the measurement of EPA-derived iPs, for example, 17
and 18.
EPA-derived iPs, namely 5-epi-8,12-iso-iPF3 -VI 17 and 8,12-iso-
a
iPF3 -VI 18 (Scheme 2).
a
The selection of EPA-derived iPs as our target focus stems from
the following: EPA, as well as docosahexaenoic acid (DHA), are
widely used as dietary supplements under the label of omega-3
fish oil. Various foods containing increased ratios of omega-3 to
omega-6 polyunsaturated fatty acids, such as milk and eggs, are
also widely available. In addition, there are some sectors of the
Target selection: The reason that we were particularly interested
in 17 and 18 stems from the hypothesis that 5-hydroxy-isopros-
tanes are resistant to b-oxidation.8 As shown in Scheme 1, of the
six iP groups, only Group VI 9 has a hydroxyl group in the 5-posi-
tion. We and others have proposed in the past that such a relation-
ship is responsible for the lack of metabolism by b-oxidation at C-1
8,12
of other eicosanoids including LTE4 and LTB4
possibly due to
complexation and inactivation of the b-oxidation enzymes. We ex-
* Corresponding author. Tel.: +1 321 674 7329.
E-mail address: jrokach@fit.edu (J. Rokach).
tended that concept to the iP field and hypothesized that if such
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.09.008

5524
C.-T. Chang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5523–5527
HO
CO2H
CO2H
O2
CO2H
HO
HO
1
7
EPA
Group I
OH
OH
2
2O2
ROS
O2
CO2H
O
O
CO2H
8
Group II
3
HO
HO
HO
OH
OH
HO
HO
HO
HO
HO
CO2H
CO2H
CO2H
CO2H
HO
OH
5
OH
9
4
6
Group VI
Group IV
Group V
Group III
OH
HO
HO
OH
HO
HO
HO
HO
CO2H
CO2H
CO2H
CO2H
HO
HO
OH
OH
10
11
12
13
Group IV
Group V
Group VI
Group III
2O2
ROS
CO2H
14
AA
Scheme 1. Isoprostanes derived from AA and EPA.
OH
OH
OH
OH
HO
HO
HO
HO
HO
HO
CO2H
CO2H
CO H
2
CO2H
CO2H
HO
HO
1
EPA
iPF3α -VI
15
5-epi-8,12-iso-iPF3α -VI
17
5-epi-iPF3α-VI
16
8,12-iso-iPF3α -VI
18
β-oxidation
OH
OH
HO
OH
OH
HO
HO
CO2H
HO
CO2H
CO2H
CO2H
CO H
2
DHA
19
HO
HO
HO
HO
7-epi-10,14-iso-nPF4α -VI
nPF4α -VI
20
10,14-iso-nPF4α -VI
23
7-epi-nPF4α -VI
21
22
Scheme 2. Isoprostanes derived from Neuroprostanes by b-oxidation.
were the case we would stand a much higher chance of finding
increased amounts of intact Group VI iPs in urine. As a result, we
have discovered that AA-derived Group VI iPs 13, Scheme 1, were
the most abundant AA-derived iPs in urine.13
Fig. 1 panel A). This also explains why attempts to detect intact
neuroprostanes in urine have been unsuccessful.14
Having selected Group VI 9 as a higher-probability target for the
non-invasive discovery of EPA-derived iPs in urine, we still needed
to reduce the number of iPs to focus upon within this group to a
more manageable number. Group VI, as is the case for all other
groups, is composed of 16 isomers, eight of which are enantiomers.
Since our analytical tool is the mass spectrometer, which cannot
In addition, we have carried out very recently a detailed study8
in which we have been able to demonstrate that 15 is resistant to
b-oxidation and that 20, a member of the neuroprostane (nP)
family derived from DHA, is metabolized to 15 (Scheme 2 and

C.-T. Chang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5523–5527
5525
Figure 1. Identification of 5-epi-8,12-iso-iPF3 -VI 17 and 8,12-iso-iPF3 -VI 18 in human urine using LC/MS/MS LC/MS/MS conditions: isoprostanes were evaluated by liquid
a
a
chromatography/tandem mass spectrometry (LC/MS/MS), using reverse phase chromatography, negative ion electrospray introduction and selected reaction monitoring
(SRM) techniques. A 200 ꢀ 2.1 mm Hypersil Gold C18 1.9 m particle size UHPLC column was used. The mobile phase was generated from (A), HPLC-grade water and (B) 5%
methanol/95% acetonitrile, both containing 0.005% acetic acid and adjusted to pH 5.7 with ammonium hydroxide. The flow rate was 350 L/min using a linear gradient from
20% B to 38% B in 20 min. The transitions monitored were m/z 351 and m/z 115 for both 5-epi-8,12-iso-iPF3 -VI 17 and 8,12-iso-iPF3 -VI 18.
l
l
a
a
distinguish between enantiomers, it is safe to ignore these com-
pounds for the present application. Of the eight remaining isomers,
four have the side chains trans to each other. Since radical cycliza-
tion precedents indicate that cis arrangements are largely pre-
ferred,15,16 we decided to ignore the trans isomers and focus on
the four cis isomers, 15, 16, 17, and 18.
Synthesis: Scheme 3 describes the stereospecific synthesis of 5-
epi-8,12-iso-iPF3 -VI 17 and 8,12-iso-iPF3 -VI 18. Scheme 3(a)
a
a
(a)
N
N
HO
TESO
N
N
O
TESO
TESO
O
TESO
N
N
OH
OTES
S
OH
S
N
d
a
N
b
c
O
O
O
O
O
O
O
O
O
O
24
25
26
27
28
OTBDMS
TESO
OTBDMS
CO2CH3
OTBDMS
TESO
HO
e
TESO
TESO
CO2CH3
CO2CH3
g
f
OTBDMS
CO2CH3
O
OH
OHC
h,i
29
OTES
30
31
O
45
OH
OTBDMS
HO
TESO
CO2H
CO2CH3
j,k
HO
TESO
32
5-epi-8,12-iso-iPF -VI
3α
17
n
(b)
TESO
O
TESO
TESO
TESO
CHO
OTES
CO2CH3
CO2CH3
m
O
l
O
O
O
O
O
O
34
35
O
O
P
CO2CH3
O
O
25
33
MeO
44
OMe
OTBDMS
OH
OH
TESO
HO
HO
TESO
CO2CH3
CO2H
CO2CH3
o
f
k
p
O
36
O
37
8,12-iso-iPF3α -VI
18
O
O
Scheme 3. Stereospecific synthesis of 5-epi-8,12-iso-iPF3 -VI 17 and 8,12-iso-iPF3 -VI 18. Reagents and conditions: (a) TESCl, Et3N, DMAP, CH2Cl2, rt, 20 h, 98%; (b)
a
a
Wilkinson’s catalyst, catechol borane, THF, 0 °C–rt, overnight, 91%; (c) PT-SH, Ph3P, DIAD, THF, 0 °C–rt, 4.5 h, 97%; (d) m-CPBA, NaHCO3, CH2Cl2, rt, overnight, 96%; (e) KHMDS,
45,21 DME, ꢁ60 °C–rt, 8 h, 52% (E: Z = 63: 37); (f) NaBH4, diethyl ether/MeOH (2:1), rt, 35 min, 70%; (g) TESCl, Et3N, DMAP, CH2Cl2, 20 h, 95%; (h) DMSO, (COCl)2, Et3N, CH2Cl2,
ꢁ70 to ꢁ20 °C, 1.5 h; (i) LiHMDS, THF, ꢁ78 °C to rt, I-Ph3P+(CH2)2CH@CHCH2CH3, 2 h (two steps 73%); (j) TBAF, THF, rt, overnight; (k) LiOH, iPrOH/H2O (1:1), rt, 4 h (two steps
93%); (l) DMSO, (COCl)2, Et3N, CH2Cl2, ꢁ70 to ꢁ20 °C, 1.5 h; (m) LiHMDS, 44, THF, ꢁ78 °C to rt, 2 h (two steps 34, 8.4%, 35, 73%); (n) I2, CH2Cl2, rt, 15 h, 87%; (o) EtOH, LiAlH4,
(S)-Binal-H, THF, ꢁ100 °C, 45 min, 92%; (p) TBDMSCl, Et3N, DMAP, CH2Cl2, rt, overnight, 85%.

5526
C.-T. Chang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5523–5527
43
PPh3
H2C
TESO
-78oC, 82%
OHC
H
-78oC, 57%
H
OTES
CHO
OTES
O
O
O
BrPh3PCH3
BrPh3PCH3
O
O
O
O
33
39
38
39
41
O
40
TESO
TESO
CHO
O
O
40
42
O
O
Scheme 4. Elimination of TES-OH during Wittig reaction.
shows the first total synthesis of iP 17. The starting dihydroxy lac-
tone 24 has been described by us previously.4 The key feature in
this approach is the Kocienski-modified Julia olefination17 reaction
between 27 and 45. This guarantees the R-stereochemistry of the
OTBDMS at position C-5 in 29. However, as is usually the case in
this coupling, a mixture of cis and trans (2:1) is formed which is
separated by silica column chromatography and readily identified
by 1H NMR (trans and cis coupling of the olefins). The selective con-
version of bis-TES 25 to the mono-TES 28 was performed in excel-
lent yield using the Rh-catalyzed catechol borane procedure we
reported recently.18 Other methods tried were not as selective.
The reaction of DIBAL on 29 produces a mixture of the lactol and
the diol. We elected instead to convert the lactone to the diol 30
and proceed via the tris-TES 31. Some reduction of the ester func-
tion also occurs during the reduction step and is separated by silica
gel column chromatography.19
We have previously reported on a stereospecific synthesis of
18.19 The preparation of 18 as described in Scheme 3(b) is shorter
and may be better suited for the preparation of some analogs and
derivatives. To obtain the aldehyde 33 we used a very convenient
one-step procedure20 which avoids the need for protection/depro-
tection. The S-Binal reduction of 35 to 36 proceeded with high e.e.
as judged by the LC/MS analysis of the final compound 18 (not
shown).
A final comment on the origin of these two peaks: we have pre-
viously shown that 20 (Scheme 2) is metabolized to 15, suggesting
that urinary 15 is derived from a combination of direct formation
from EPA, as shown in Scheme 1, and b-oxidation of the DHA-de-
rived nPs 20.8 It is tempting to assume that in this case, too, urinary
1724 and 1824 may have arisen as a result of these two pathways,
and may be good in vivo indicators of peroxidation of both
EPA- and DHA-containing lipids.
Acknowledgments
We acknowledge the National Institutes of Health for support
under Grants HL-81873 (J.R.) and HL-62250 (G.A.F.). J.R. acknowl-
edges the National Science Foundation for the AMX-360 (CHE-
90-13145) and Bruker 400 MHz (CHE-03-42251) NMR instru-
ments. G.A.F. is the McNeil Professor of Translational Medicine
and Therapeutics. W.S.P. wishes to acknowledge the Canadian
Institutes of Health Research, grant number MOP-6254, the Heart
and Stroke Foundation of Quebec, and the J.T. Costello Memorial
Research Fund.
References and notes
1. Morrow, J. D.; Hill, K. E.; Burk, R. F.; Nammour, T. M.; Badr, K. F.; Roberts, L. J., II
Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 9383.
The reactivity of aldehyde 33 merits comment. The Wittig reac-
tion with the stabilized ketophosphonate 4422 proceeded unevent-
fully in high yield, as shown in Scheme 3. Attempts to perform a
similar reaction with the more nucleophilic and basic 43 resulted
mostly in the elimination of TES-OH to afford 38 (Scheme 4). Var-
iable amounts of the target compound 42 are produced. As can be
2. Lawson, J. A.; Rokach, J.; FitzGerald, G. A. J. Biol. Chem. 1999, 274, 24441.
3. Hwang, S. W.; Adiyaman, M.; Khanapure, S.; Schio, L.; Rokach, J. J. Am. Chem.
Soc. 1994, 116, 10829.
4. Rokach, J.; Khanapure, S. P.; Hwang, S. W.; Adiyaman, M.; Schio, L.; FitzGerald,
G. A. Synth. First Special Issue, Nat products Synt. 1998, 569.
5. Yin, H.; Havrilla, C. M.; Gao, L.; Morrow, J. D.; Porter, N. A. J. Biol. Chem. 2003,
278, 16720.
6. Praticò, D.; Tangirala, R. K.; Rader, D. J.; Rokach, J.; FitzGerald, G. A. Nature Med.
1998, 4, 1189.
seen in the three-dimensional structure 33, the hydrogen
a to the
aldehyde is exposed with little hindrance, resulting in a preferen-
tial attack of the ylide 43 on this hydrogen instead of the desired
attack on the aldehyde. To prove the point we prepared 40 from
commercial ent-Corey lactone. In this case the hydrogen is hin-
dered and no elimination product is formed and a high yield of
the olefin 41 obtained.
7. Montine, T. J.; Neely, D. M.; Quinn, J. F.; Beal, F. M.; Markesbery, W. R.; Roberts,
L. J., II; Morrow, J. D. Free Radic. Biol. Med. 2002, 33, 620.
8. Lawson, J. A.; Kim, S.; Powell, W. S.; FitzGerald, G. A.; Rokach, J. J. Lipid Res.
2006, 47, 2515.
9. Gao, L.; Yin, H.; Milne, G. L.; Porter, N. A.; Morrow, J. D. J. Biol. Chem. 2006, 281,
14092.
10. Kadiiska, M. B.; Gladen, B. C.; Baird, D. D.; Germolec, D.; Graham, L. B.; Parker,
C. E.; Nyska, A.; Wachsman, J. T.; Ames, B. N.; Basu, S.; Brot, N.; FitzGerald, G. A.;
Floyd, R. A.; George, M.; Heinecke, J. W.; Hatch, G. E.; Hensley, K.; Lawson, J. A.;
Marnett, L. J.; Morrow, J. D.; Murray, D. M.; Plastaras, J.; Roberts, L. J., II; Rokach,
J.; Shigenaga, M. K.; Sohal, R. S.; Sun, J.; Tice, R. R.; VanThiel, D. H.; Wellner, D.;
Walter, P. B.; Tomer, K. B.; Mason, R. P.; Barrett, J. C. Free Radic. Biol. Med. 2005,
38, 698.
11. Kadiiska, M. B.; Gladen, B. C.; Baird, D. D.; Graham, L. B.; Parker, C. E.; Ames, B.
N.; Basu, S.; FitzGerald, G. A.; Lawson, J. A.; Marnett, L. J.; Morrow, J. D.; Murray,
D. M.; Plastaras, J.; Roberts, L. J., II; Rokach, J.; Shigenaga, M. K.; Sun, J.; Walter,
P. B.; Tomer, K. B.; Barrett, J. C.; Mason, R. P. Free Radic. Biol. Med. 2005, 38, 711.
12. Murphy, R. C.; Wheelan, P. Metabolism of leukotrienes in vitro and in vivo. In
SRS-A to Leukotrienes: The Dawning of a New Treatment; Holgate, S., Dahlen, S. E.,
Eds.; Blackwell Science Ltd.: London, 1997; p 101.
Discovery of 8,12-iso-iPF3 -VI and 5-epi-8,12-iso-iPF3 -VI in hu-
a
a
man urine: We have used the two synthetic probes 17 and 18 to
identify these iPs in urine. It is interesting to note that the two
peaks that we identified as 17 and 18 are by far the most promi-
nent peaks in the mass chromatogram representing Group VI iPs
(Fig. 1A). This is not unlike the case we encountered previously
in the 14,15-dihydro series of AA-derived iPs, in which the two
all-syn iPs in Group VI are the most prevalent.13,23 Panels B and C
of Figure 1 represent urine supplemented with 1 ng of either 17
or 18, respectively. Various co-injections with 1–5 ng produced
similar results (not shown).
13. Li, H.; Lawson, J. A.; Reilly, M.; Adiyaman, M.; Hwang, S. W.; Rokach, J.;
FitzGerald, G. A. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 13381.

C.-T. Chang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5523–5527
5527
14. Montine, T. J.; Quinn, J. F.; Milatovic, D.; Silbert, L. C.; Dang, T.; Sanchez,
S.; Terry, E.; Roberts, L. J., II; Kaye, J. A.; Morrow, J. D. Ann. Neurol. 2002,
52, 175.
15. Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical Reactions; VCH:
Weinheim, Germany, 1995.
23. Praticò, D.; Barry, O. P.; Lawson, J. A.; Adiyaman, M.; Hwang, S. W.; Khanapure,
S. P.; Iuliano, L.; Rokach, J.; FitzGerald, G. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95,
3449.
24. Spectral data of 5-epi-8,12-iso-iPF3 -VI (17) 1H NMR (Methanol-d4, 400 MHz)
a
d 5.712(1H, dd, J = 15.3, 10.6), 5.439-5.320 (2H, m), 5.281-5.155 (3H, m), 4.094–
3.925 (3H, m), 2.768–2.641 (2H, m), 2.589–2.482 (1H, m), 2.391–2.285 (1H, m),
2.268–1.921 (6H, m), 1.780–1.681 (1H, m), 1.635–1.371 (5H, m), 0.866 (3H, t,
J = 7.5); 13C NMR (methanol-d4, 100 MHz) d 177.59, 137.48, 132.59, 130.12,
129.58, 129.40, 128.66, 74.88, 73.08, 72.86, 52.03, 48.65, 43.53, 37.86, 35.30,
26.65, 25.11, 22.40, 21.48, 14.67; HRMS m/z calcd for C20H31Oþ4 335.2222 found
16. Beckwith, A. L. J. Chem. Soc. Rev. 1993, 22, 143.
17. Blakemore, P. R.; Cole, W. J.; Kocienski, P. J.; Morley, A. Synlett 1998, 26–28.
18. Patel, P.; Chang, C. T.; Kang, N.; Lee, G. J.; Powell, W. S.; Rokach, J. Tetrahedron
Lett. 2007, 48, 5289.
19. Jacobo, S. H.; Chang, C. T.; Lee, G. J.; Lawson, J. A.; Powell, W. S.; Pratico, D.;
FitzGerald, G. A.; Rokach, J. J. Org. Chem. 2006, 71, 1370.
20. Rodriguez, A.; Nomen, M.; Spur, B. W.; Godfroid, J. J. Tetrahedron Lett. 1999, 40,
5161.
21. Leblanc, Y.; Fitzsimmons, B. J.; Charleson, S.; Alexander, P.; Evans, J. F.; Rokach,
J. Prostaglandins 1987, 33, 617.
22. Adiyaman, M.; Lawson, J. A.; Hwang, S. W.; Khanapure, S. P.; FitzGerald, G. A.;
Rokach, J. Tetrahedron Lett. 1996, 37, 4849.
335.2221.8,12-iso-iPF3 -VI (18) 1H NMR (methanol-d4, 400 MHz) d 5.771 (1H,
a
dd, J = 15.3, 10.7), 5.380–5.149 (5H, m), 4.098–3.924 (3H, m), 2.771–2.643 (2H,
m), 2.581–2.495 (1H, m), 2.393–2.291 (1H, m), 2.249–1.942 (6H, m), 1.779–
1.661 (1H, m), 1.623–1.369 (5H, m), 0.868 (3H, t, J = 7.5); 13C NMR (Methanol-
d4, 100 MHz) d 177.58, 137.48, 132.60, 130.98, 129.98, 129.44, 128.58, 74.83,
73.82, 72.79, 51.95, 48.49, 43.56, 37.61, 34.95, 26.66, 25.12, 22.29, 21.50, 14.68;
HRMS m/z calcd for C20H31Oþ4 335.2222 found 335.2221.