The synthesis and aqueous superoxide anion scavenging of water-dispersible lutein esters

Article abstract of DOI:10.1016/j.bmcl.2005.11.024

Xanthophyll carotenoids of the C40 series, which includes commercially important compounds such as lutein, zeaxanthin, and astaxanthin, have poor aqueous solubility in the native state. Hawaii Biotech, Inc. (HBI) and others have shown that the aqueous dispersibility of derivatized carotenoids can be increased by varying the chemical structure of the esterified moieties. In the current study, the published series of novel, highly water-dispersible C40 carotenoid derivatives has been extended to include (3R,3′R,6′R)- lutein (β,ε-carotene-3,3′-diol) derivatives. Two novel derivatives were synthesized by esterification with inorganic phosphate and succinic acid, respectively, and subsequently converted to the sodium salts. Red-orange, clear, aqueous suspensions were obtained after addition of these novel derivatives to USP-purified water. Aqueous dispersibility of the disuccinate sodium salt of lutein was 2.85 mg/mL; the diphosphate salt demonstrated a >10-fold increase in dispersibility at 29.27 mg/mL. As reported previously, these aqueous suspensions were obtained without the addition of heat, detergents, co-solvents, or other additives. The direct aqueous superoxide scavenging abilities of these novel derivatives were subsequently evaluated by electron paramagnetic resonance (EPR) spectroscopy in a well-characterized in vitro isolated human neutrophil assay. The novel derivatives were nearly identical aqueous-phase scavengers, demonstrating dose-dependent suppression of the superoxide anion signal (as detected by spin-trap adducts of DEPMPO) in the millimolar range. These lutein-based soft drugs will likely find utility in those commercial and clinical applications for which aqueous-phase singlet oxygen quenching and direct radical scavenging may be required.

Full text of DOI:10.1016/j.bmcl.2005.11.024

Bioorganic & Medicinal Chemistry Letters 16 (2006) 775–781
The synthesis and aqueous superoxide anion scavenging of
water-dispersible lutein esters
Geoff Nadolski,a Arturo J. Cardounel,b Jay L. Zweierb and Samuel F. Lockwooda,*
aHawaii Biotech, Inc., 99-193 Aiea Heights Drive, Suite 200, Aiea, HI 96701, USA
bDavis Heart and Lung Research Institute, 473 West 12th Avenue, Columbus, OH 43210-1252, USA
Received 8 October 2005; revised 8 November 2005; accepted 8 November 2005
Available online 28 November 2005
Abstract—Xanthophyll carotenoids of the C40 series, which includes commercially important compounds such as lutein, zeaxan-
thin, and astaxanthin, have poor aqueous solubility in the native state. Hawaii Biotech, Inc. (HBI) and others have shown that
the aqueous dispersibility of derivatized carotenoids can be increased by varying the chemical structure of the esterified moieties.
In the current study, the published series of novel, highly water-dispersible C40 carotenoid derivatives has been extended to include
(3R,30R,60R)-lutein (b,e-carotene-3,30-diol) derivatives. Two novel derivatives were synthesized by esterification with inorganic
phosphate and succinic acid, respectively, and subsequently converted to the sodium salts. Red-orange, clear, aqueous suspensions
were obtained after addition of these novel derivatives to USP-purified water. Aqueous dispersibility of the disuccinate sodium salt
of lutein was 2.85 mg/mL; the diphosphate salt demonstrated a >10-fold increase in dispersibility at 29.27 mg/mL. As reported pre-
viously, these aqueous suspensions were obtained without the addition of heat, detergents, co-solvents, or other additives. The direct
aqueous superoxide scavenging abilities of these novel derivatives were subsequently evaluated by electron paramagnetic resonance
(EPR) spectroscopy in a well-characterized in vitro isolated human neutrophil assay. The novel derivatives were nearly identical
aqueous-phase scavengers, demonstrating dose-dependent suppression of the superoxide anion signal (as detected by spin-trap
adducts of DEPMPO) in the millimolar range. These lutein-based soft drugs will likely find utility in those commercial and clinical
applications for which aqueous-phase singlet oxygen quenching and direct radical scavenging may be required.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The synthesis of water-dispersible C40 carotenoids—as
potential parenteral agents for clinical applications1–6
micelles.13,14 The net overall effect is a vast improvement
in potential clinical utility for the lipophilic carotenoid
compounds as therapeutic agents.
—
has proven to be an effective strategy to improve the
injectability of these compounds as therapeutic agents,
a result perhaps not achievable through other formula-
tion methods.7 The methodology has also been extended
to non-symmetric carotenoids with fewer than 40 carbon
atoms in the molecular skeleton and differing ionic char-
acter.8,9 The aqueous dispersibility of these compounds
also facilitates proof-of-concept studies in model systems
(e.g., cell culture),10,11 where the high lipophilicity of
these compounds previously limited their bioavailability
and hence proper evaluation of efficacy.12 In addition,
esterification has proven useful to increase oral bioavail-
ability, a fortuitous side effect of the esterification process
which can increase solubility in gastric mixed
In the current study, the principles of retrometabolic
drug design were again utilized to produce novel soft
drugs15 from the asymmetric parent C40 carotenoid
scaffold (3R,30R,60R)-lutein. Lutein was obtained com-
mercially as purified natural source material (marigold
extract), mainly the (3R,30R,60R)-stereoisomer (one of
eight potential stereoisomers). Lutein (Scheme 1)
possesses key characteristics—similar to those of HBIÕs
previous starting material astaxanthin—which also
made it an ideal starting platform for retrometabolic
syntheses: one, synthetic handles (hydroxyl groups) for
conjugation, and two, an excellent safety profile for
the parent compound.16 As stated previously, lutein is
available commercially from multiple sources in bulk
as predominantly the (3R,30R,60R)-stereoisomer, the pri-
mary isomer in the human diet and human retinal tis-
sue.17 Recent studies have demonstrated the utility of
lutein-based supplementation for the clinical improve-
ment of vision, reduction of ultraviolet (UV)-based
Keywords: Aqueous dispersibility; Carotenoids; Carotenoid deriva-
tives; DEPMPO; Electron paramagnetic resonance; EPR; Lutein;
Lutein diphosphate sodium salt; Lutein disuccinate sodium salt;
Spin traps; Superoxide anion.
*
Corresponding author. Tel.: +1 808 486 5333; fax: +1 808 792
1343; e-mail: slockwood@hibiotech.com
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.024

776
G. Nadolski et al. / Bioorg. Med. Chem. Lett. 16 (2006) 775–781
Scheme 1. Reagents and conditions: (a) succinic anhydride, N,N-diisopropylethylamine, CH2Cl2 (64%); (b) NaOMe, CH2Cl2 (91%).
inflammation, and potentially the inhibition and/or
amelioration of age-related macular degeneration
(ARMD).16–19 The synthesis, characterization, and
aqueous-phase superoxide anion scavenging abilities of
the diphosphate and disuccinate sodium salts of natural
source lutein are herein reported. The potential utility of
lutein-based formulations has been extended for clinical
application by providing compounds with sufficient
aqueous dispersibility for parenteral administration,
thereby capturing the significant human population of
carotenoid oral non-responders as well as acute clinical
application(s) requiring rapid loading of therapeutic
doses. In addition, potent and nearly identical direct
superoxide anion scavenging ability (at millimolar con-
centration) is documented for both novel compounds.
The resulting salt was diluted with water and lyophilized
to provide sodium salt 3 in good yield.
In the pursuit of preparing a water-dispersible phos-
phate derivative of lutein, a three-step protocol was
envisioned: one, phosphorylation of hydroxyls as
protected phosphates; two, mild removal of protecting
groups to yield free phosphates; and three, salt forma-
tion to provide the diphosphate sodium salt.
Successful diphosphorylation of lutein was first achieved
using dimethyl phosphoroiodidate, formed in situ by
reacting trimethyl phosphite with iodine.25 Attempted
deblocking of the phosphates using potassium cya-
nide,26 phenyl thiolate,26 tert-butylamine,27 lithium
hydroxide, or bromotrimethylsilane28 in the presence
of N,O-bis(trimethylsilyl)acetamide failed to provide
target 7 (see Scheme 2). Such efforts resulted in the
decomposition of the polyene, cleavage of phosphates,
or the incomplete deprotection of methyl phosphate
esters.
In HBIÕs continuing efforts to impart water dispersibility
to clinically relevant hydrophobic carotenoids, the prep-
aration of two water-dispersible lutein derivatives, the
sodium salts of lutein disuccinate and lutein diphos-
phate, as seen in Schemes 1 and 2, is reported.20
The synthesis of disuccinate salt 3 was straightforward
and began with succinylation of lutein using succinic
anhydride and Hunig base (Scheme 1). Disuccinylation
¨
of lutein was optimized by running the reaction in a con-
centrated fashion and using modest excesses of anhy-
dride and base. Aqueous acidic workup yielded
disuccinate 2; excess reagents were removed by washing
with dilute HCl. The water-dispersible derivative (3) was
generated by treatment of 2 with sodium methoxide.
As seen in Scheme 2, sodium salt 8 was successfully pre-
pared when using benzyl esters as phosphate-protecting
groups. Lutein was phosphorylated using in situ gener-
ated dibenzyl phosphoroiodidate,25 forming a mixture
of benzyl-protected diphosphates. Workup of the phos-
phate mixture provided a crude red oil that was used in
the deprotection step without further purification.
Successful debenzylation of the protected phosphates
was achieved using bromotrimethylsilane28 in the
Scheme 2. Reagents and conditions: (a) benzyl alcohol, triethylamine, Et2O (83%); (b) I2, CH2Cl2; (c) 1, pyridine, CH2Cl2 then 5; (d)
bromotrimethylsilane, pyridine, CH2Cl2; (e) NaOMe, CH2Cl2 (35% for three steps).

G. Nadolski et al. / Bioorg. Med. Chem. Lett. 16 (2006) 775–781
777
Figure 1. Mean percent inhibition ( SEM) of superoxide anion signal as detected by DEPMPO spin-trap by the sodium disuccinate derivative of
lutein (tested in water). A 100 lM formulation (0.1 mM) was also tested in 40% EtOH, a concentration shown to produce a molecular (i.e., non-
aggregated) solution (data not shown). As the concentration of the derivative increased, inhibition of the superoxide anion signal increased in a dose-
dependent manner. At 5 mM, approximately 3/4 (75%) of the superoxide anion signal was inhibited. No apparent scavenging (0% inhibition) was
observed at 0.1 mM in water, and scavenging was not significantly increased over background scavenging by the EtOH vehicle (5%)1 by the addition
of 40% EtOH. The millimolar concentration scavenging by the novel derivative was accomplished in water alone, without the addition of organic
co-solvent (e.g., acetone, EtOH),1 heat, detergents, or other additives.
presence of pyridine. As shown in Scheme 2, the sodium
salt was generated by treating 8 with sodium methoxide.
The resulting salt was washed with diethyl ether, then
water, and resuspended in aqueous methanol. The solu-
tion was diluted with water and lyophilized to provide
sodium salt 8 in good yield.
icant scavenging (0% inhibition) was observed at
0.1 mM in water. Addition of 40% EtOH to the deriva-
tive solution at 0.1 mM did not significantly increase
scavenging over that provided by the EtOH vehicle
alone (5% inhibition).1 These data suggested that
card-pack (ÔHÕ-type) aggregation for this derivative
was not occurring in aqueous solution (and thus limiting
the interaction of the aggregated carotenoid derivative
with aqueous superoxide anion) (see Tables 1 and
2).29,30
The mean percent inhibition of superoxide anion signal
( SEM) as detected by DEPMPO spin-trap by the sodi-
um disuccinate derivative of lutein (tested in water) is
shown in Figure 1. A 100 lM formulation (0.1 mM)
was also tested in 40% EtOH, a concentration shown
to produce a molecular (i.e., non-aggregated) solution.
As the concentration of the derivative increased, inhibi-
tion of superoxide anion signal increased in a dose-
dependent manner. At 5 mM, approximately 3/4 (75%)
of the superoxide anion signal was inhibited. No signif-
The mean percent inhibition of superoxide anion signal
(
SEM) as detected by DEPMPO spin-trap by the sodi-
um diphosphate derivative of lutein (tested in water) is
shown in Figure 2. A 100 lM formulation (0.1 mM)
was also tested in 40% EtOH, a concentration also
shown to produce a molecular (i.e., non-aggregated)
Table 1. Descriptive statistics of mean % inhibition of superoxide anion signal for aqueous and ethanolic (40%) formulations of sodium disuccinate
derivatives of lutein tested in the current study
Sample
Solvent
Concentration (mM)
N
Mean (% inhibition)
SD
SEM
Min
Max
Range
Lutein disuccinate sodium salt
Lutein disuccinate sodium salt
Lutein disuccinate sodium salt
Lutein disuccinate sodium salt
Lutein disuccinate sodium salt
40% EtOH
Water
Water
0.1
0.1
1.0
3.0
5.0
3
1
3
3
3
5.0
0.0
4.4
ND
5.6
4.0
4.5
2.5
ND
3.2
2.3
2.6
0
0
8
0
8
0
13.0
61.7
74.7
8
19
66
79
11
8
Water
Water
58
70
9
Sample sizes of three were evaluated for each formulation, with the exception of lutein in 40% EtOH stock solution (N = 1). Mean % inhibition did
not increase over background1 levels until sample concentration reached 1 mM in water; likewise, addition of 40% EtOH at the 0.1 mM concen-
tration did not increase scavenging over background levels attributable to the EtOH vehicle (mean = 5% inhibition).1

778
G. Nadolski et al. / Bioorg. Med. Chem. Lett. 16 (2006) 775–781
Table 2. Descriptive statistics of mean % inhibition of superoxide anion signal for aqueous and ethanolic (40%) formulations of sodium diphosphate
derivatives of lutein tested in the current study
Sample
Solvent
Concentration (mM)
N
Mean (% inhibition)
SD
SEM
Min
Max
Range
Lutein diphosphate sodium salt
Lutein diphosphate sodium salt
Lutein diphosphate sodium salt
Lutein diphosphate sodium salt
Lutein diphosphate sodium salt
40% EtOH
Water
Water
0.1
0.1
1.0
3.0
5.0
3
1
3
3
3
18.0
0.0
7.0
ND
3.5
3.1
2.6
4.0
ND
2.0
1.8
1.5
11
0
25
0
14
0
9.3
6
13
75
93
7
Water
Water
72.3
91.0
69
88
6
5
Sample sizes of three were evaluated for each formulation, with the exception of lutein diphosphate in water at 100 lM (0.1 mM) where N = 1. Mean
% inhibition of superoxide anion signal increased in a dose-dependent manner as the concentration of lutein diphosphate was increased in the test
assay. At 100 lM in water, no inhibition of scavenging was seen. The molecular solution in 40% EtOH (mean % inhibition = 18%) was increased
above background scavenging (5%) by the ethanolic vehicle,1 suggesting that disaggregation increased scavenging at that concentration. Slightly
increased scavenging (on a molar basis) may have been obtained with the diphosphate derivative in comparison to disuccinate derivative (see Table 1
and Fig. 1).
Figure 2. Mean percent inhibition ( SEM) of superoxide anion signal as detected by DEPMPO spin-trap by the sodium diphosphate derivative of
lutein (tested in water). A 100 lM formulation (0.1 mM) was also tested in 40% EtOH, a concentration shown to produce a molecular (i.e., non-
aggregated) solution (data not shown). As the concentration of the derivative increased, inhibition of the superoxide anion signal increased in a dose-
dependent manner. At 5 mM, greater than 90% of the superoxide anion signal was inhibited. No significant scavenging (0% inhibition) was observed
at 0.1 mM in water, however a significant increase over background scavenging by the EtOH vehicle (5%)1 was observed after the addition of 40%
EtOH (mean = 18% inhibition). This suggested that the molecular solution of the compound demonstrated increased scavenging over the aggregated
solution, and that the diphosphate derivate may have increased scavenging potential in comparison to the disuccinate derivative. Again, the
millimolar concentration scavenging by the novel derivative was accomplished in water alone, without the addition of organic co-solvent (e.g.,
acetone, EtOH),29 heat, detergents, or other additives.
solution of this derivative. As the concentration of the
derivative increased, inhibition of the superoxide anion
signal increased in a dose-dependent manner. At
5 mM, slightly more than 90% of the superoxide anion
signal was inhibited (versus 75% for the disuccinate lu-
tein sodium salt). As for the disuccinate lutein sodium
salt, no apparent scavenging (0% inhibition) was
observed at 0.1 mM in water. However, a significant
increase over background scavenging by the EtOH vehi-
cle (5%) was observed after the addition of 40% EtOH,
resulting in a mean 18% inhibition of superoxide anion
signal. This suggested that disaggregation of the
compound leads to an increase in scavenging ability by
this novel derivative, pointing to a slightly increased
scavenging ability of molecular solutions of the more
water-dispersible diphosphate derivative relative to the
disuccinate derivative.
In the current study, facile preparation of the sodium
salts of (3R,30R,60R)-lutein disuccinate and diphosphate
is described. These asymmetric C40 carotenoid deriva-
tives exhibited aqueous dispersibility of 2.85 and
29.27 mg/mL, respectively. Increased dispersibility of
the phosphate diesters of lutein versus the succinate
diesters is believed to be primarily a function of the
increased potential for hydrogen bonding by ionized
phosphate groups; in aqueous solvent, the more polar
phosphate moiety will impart greater water dispersibility
to the parent carotenoid scaffold than a disuccinate
functionality. Electronic paramagnetic resonance

G. Nadolski et al. / Bioorg. Med. Chem. Lett. 16 (2006) 775–781
779
16. Richer, S.; Stiles, W.; Statkute, L.; Pulido, J.; Frankowski,
J.; Rudy, D.; Pei, K.; Tsipursky, M.; Nyland, J. Optometry
2004, 75, 216.
17. Landrum, J. T.; Bone, R. A. Arch. Biochem. Biophys.
2001, 385, 28.
18. Olmedilla, B.; Granado, F.; Blanco, I.; Vaquero, M.
Nutrition 2003, 19, 21.
19. Lee, E. H.; Faulhaber, D.; Hanson, K. M.; Ding, W.;
Peters, S.; Kodali, S.; Granstein, R. D. J. Invest. Dermatol.
2004, 122, 510.
spectroscopy of direct aqueous superoxide scavenging
by these derivatives demonstrated nearly identical
dose-dependent scavenging profiles, with slightly in-
creased scavenging noted for the diphosphate derivative.
In each case, scavenging in the millimolar range was ob-
served. These results suggest that as parenteral soft
drugs with aqueous radical scavenging activity, both
compounds are potentially useful in those clinical appli-
cations in which rapid and/or intravenous delivery is de-
sired for the desired therapeutic effect(s). In particular,
parenteral administration of lutein esters should over-
come the problems with dose-proportionality and carot-
enoid non-response observed with oral administration
of lutein.31
20. Chemical syntheses (general). Natural source lutein (90%)
was obtained from ChemPacific, Inc. (Baltimore, MD) as
a red-orange solid and was purified by dissolving in a
minimal volume of CH2Cl2, passed through a 0.45 lm
filter, and concentrated in vacuo. This process was
repeated three times. All other reagents and solvents used
were purchased from Acros (New Jersey, USA) and were
used without further purification. All reactions were
performed under N2 atmosphere. All flash chromato-
graphic purifications were performed on Natland Inter-
national Corporation 230–400 mesh silica gel using the
indicated solvents. LC/MS (APCI) was recorded on an
Agilent 1100 LC/MSD VL system; column: Zorbax
Eclipse XDB-C18 Rapid Resolution (4.6 · 75 mm,
3.5 lm); temperature: 25 ꢁC; starting pressure: 105 bar;
flow rate: 1.0 mL/ min; mobile phase (%A = 0.025%
trifluoroacetic acid in H2O, %B = 0.025% trifluoroacetic
acid in acetonitrile). Gradient program: 70% A/30% B
(start), step gradient to 50% B over 5 min, step gradient to
98% B over 8.30 min, hold at 98% B over 25.20 min, step
gradient to 30% B over 25.40 min; PDA Detector: 470 nm.
The presence of trifluoroacetic acid in the LC eluents acts
to protonate lutein disuccinate and diphosphate salts to
give the free diacid forms, yielding M+ = 768 for the
disuccinate salt and M+ = 728 for the diphosphate salt.
LRMS: + mode; APCI: atmospheric pressure chemical
ionization, ion collection using quadrapole.
Acknowledgments
The authors would like to recognize Dean Frey Ph.D.,
Mark McClaws Ph.D., and Juris Ekmanis Ph.D. of
Albany Molecular Research, Inc. (AMRI) for help with
analytical chemistry considerations and dispersibility
determinations for these novel derivatives. Hans-Rich-
ard Sliwka Ph.D. provided advice and comment on
the biophysical behavior of the novel derivatives in
aqueous solution, and ongoing inspiration in the devel-
opment of water-dispersible carotenoids. Dave M.
Watumull provided spectral analysis of the compounds
prior to testing in the EPR-based human neutrophil
assay.
References and notes
Lutein (b,e-carotene-3,30-diol) (1). LC/MS (APCI):
9.95 min (2.78%), kmax 423 nm (70%), 446 nm (100%),
474 nm (80%); 10.58 min (3.03%), kmax 423 nm (73%),
446 nm (100%), 474 nm (82%); 11.10 min (4.17%), kmax
423 nm (68%), 447 nm (100%), 474 nm (79%); 12.41 min
(90.02%), kmax 423 nm (73%), 447 nm (100%), 474 nm
(83%); m/z 568 M+ (69%), 551 [MꢀH2O+H]+ (100%), 533
(8%).
1. Cardounel, A. J.; Dumitrescu, C.; Zweier, J. L.; Lock-
wood, S. F. Biochem. Biophys. Res. Commun. 2003, 307,
704.
2. Zsila, F.; Simonyi, M.; Lockwood, S. F. Bioorg. Med.
Chem. Lett. 2003, 13, 4093.
3. Frey, D. A.; Kataisto, E. W.; Ekmanis, J. L.; OÕMalley, S.;
Lockwood, S. F. Org. Process Res. Dev. 2004, 8, 796.
4. Gross, G. J.; Lockwood, S. F. Life Sci. 2004, 75, 215.
5. Jackson, H. L.; Cardounel, A. J.; Zweier, J. L.; Lock-
wood, S. F. Bioorg. Med. Chem. Lett. 2004, 14, 3985.
b,e-Carotenyl 3,30-disuccinate (2). To a solution of lutein
(1) (0.50 g, 0.879 mmol) in CH2Cl2 (8 mL) were added
N,N-diisopropylethylamine (3.1 mL, 17.58 mmol) and
succinic anhydride (0.88 g, 8.79 mmol). The solution was
stirred at rt overnight and then diluted with CH2Cl2 and
quenched with cold 5% aqueous HCl. The aqueous layer
was extracted two times with CH2Cl2 and the combined
organic layer was washed three times with 5% aqueous
HCl, dried over Na2SO4, and concentrated to yield
disuccinate 2 (0.433 g, 64%) as a red-orange solid; LC/
MS (APCI):10.37 min (8.42%), kmax 423 nm (74%),
446 nm (100%), 474 nm (83%); m/z 769 [M+H]+ (8%),
668 [MꢀC4O3H4]+ (7%), 650 (100%), 532 (22%); 11.78 min
(90.40%), kmax 269 nm (18%), 423 nm (68%), 446 nm
(100%), 474 nm (80%); m/z 769 [M+H]+ (7%), 668
[MꢀC4O3H4]+ (9%), 650 (100%), 532 (23%).
´
6. Zsila, F.; Fitos, I.; Bikadi, Z.; Simonyi, M.; Jackson, H.
L.; Lockwood, S. F. Bioorg. Med. Chem. Lett. 2004, 14,
5357.
7. Lockwood, S. F.; OÕMalley, S.; Mosher, G. L. J. Pharm.
Sci. 2003, 92, 922.
8. Foss, B. J.; Naess, S. N.; Sliwka, H. R.; Partali, V. Angew.
Chem., Int. Ed. 2003, 42, 5237.
9. Foss, B. J.; Sliwka, H. R.; Partali, V.; Cardounel, A. J.;
Zweier, J. L.; Lockwood, S. F. Bioorg. Med. Chem. Lett.
2004, 14, 2807.
10. Hix, L. M.; Lockwood, S. F.; Bertram, J. S. Cancer Lett.
2004, 211, 25.
11. Hix, L. M.; Lockwood, S. F.; Bertram, J. S. Redox Rep.
2004, 9, 181.
12. Bertram, J. S. Nutr. Rev. 1999, 57, 182.
13. Bowen, P. E.; Herbst-Espinosa, S. M.; Hussain, E. A.;
Stacewicz-Sapuntzakis, M. J. Nutr. 2002, 132, 3668.
14. Showalter, L. A.; Weinman, S. A.; Østerlie, M.; Lock-
wood, S. F. Comp. Biochem. Physiol. C: Toxicol. Phar-
macol. 2004, 137, 227.
b,e-Carotenyl 3,30-disuccinate sodium salt (3). To a solu-
tion of disuccinate 2 (0.32 g, 0.416 mmol) in CH2Cl2
(6 mL) at 0 ꢁC was added dropwise sodium methoxide
(25% wt in methanol; 0.20 mL, 0.874 mmol). The solution
was diluted with water, and the clear, red-orange aqueous
solution was lyophilized to yield 3 (0.278 g, 91%) as a red-
orange, hygroscopic solid; LC/MS (APCI): 11.75 min
(91.32%), kmax 269 nm (16%), 423 nm (70%), 446 nm
15. Buchwald, P.; Bodor, N. Pharmazie 2004, 59, 396.

780
G. Nadolski et al. / Bioorg. Med. Chem. Lett. 16 (2006) 775–781
(100%), 474 nm (82%); m/z 769 [M+H]+ (7%), 668
(100%), 474 nm (86%), m/z 912 (41%), 780 (15%), 692
(5%), 630 (100%), 550 (43%); 9.45 min (12.40%), kmax
268 nm (21%), 335 nm (16%), 423 nm (76%), 446 nm
(100%), 474 nm (80%).
[MꢀC4O3H4]+ (9%), 650 (100%), 532 (23%).
Tribenzyl phosphite (4). To a well-stirred solution of
phosphorus trichloride (1.7 mL, 19.4 mmol) in Et2O
(430 mL) at 0 ꢁC was added dropwise a solution of
triethylamine (8.4 mL, 60.3 mmol) in Et2O (20 mL),
Determination of aqueous dispersibility. 30.13 mg of 3 was
added to 1 mL USP-purified water. The sample was
rotated for 2 h and then centrifuged for 5 min. After
centrifuging, solid was visible in the bottom of the tube. A
125 lL aliquot of the solution was then diluted to 25 mL.
The sample was analyzed by UV/Vis spectroscopy at
436 nm, and the absorbance was compared to a standard
curve compiled from four standards of known concentra-
tion. The concentration of the original supernatant was
calculated to be 2.85 mg/mL and the absorptivity was
36.94 AU mL/cm mg. Slight error may have been intro-
duced by the small size of the original aliquot. Next,
30.80 mg of 8 was added to 1 mL USP-purified water. The
sample was rotated for 2 h and then centrifuged for 5 min.
After centrifuging, solid was visible in the bottom of the
tube. A 125 lL aliquot of the solution was then diluted to
25 mL. The sample was analyzed by UV/Vis spectroscopy
at 411 nm, and the absorbance was compared to a
standard curve compiled from four standards of known
concentration. The concentration of the original superna-
tant was calculated to be 29.27 mg/mL and the absorptiv-
ity was 2.90 AU mL/cm mg. Slight error may have been
introduced by the small size of the original aliquot.
Leukocyte isolation and preparation. Human polymorpho-
nuclear leukocytes (PMNs) were isolated from freshly
sampled venous blood of a single volunteer (S.F.L.) by
Percoll density gradient centrifugation as described previ-
ously.1,21 Briefly, each 10 mL of whole blood was mixed
with 0.8 mL of 0.1 M EDTA and 25 mL saline. The
diluted blood was then layered over 9 mL Percoll at a
specific density of 1.080 g/mL. After centrifugation at 400g
for 20 min at 20 ꢁC, the plasma, mononuclear cell, and
Percoll layers were removed. Erythrocytes were subse-
quently lysed by addition of 18 mL ice-cold water for 30 s,
followed by 2 mL of 10· PIPES buffer (25 mM PIPES,
110 mM NaCl, and 5 mM KCl, titrated to pH 7.4 with
NaOH). Cells were then pelleted at 4 ꢁC, the supernatant
was decanted, and the procedure was repeated. After the
second hypotonic cell lysis, cells were washed twice with
PAG buffer [PIPES buffer containing 0.003% human
serum albumin (HSA) and 0.1% glucose]. Afterward,
PMNs were counted by light microscopy on a hemocy-
tometer. The isolation yielded PMNs with a purity of
>95%. The final pellet was then suspended in PAG-CM
buffer (PAG buffer with 1 mM CaCl2 and 1 mM MgCl2).
EPR measurements. All EPR measurements were per-
formed using a Bruker ER 300 EPR spectrometer oper-
ating at X-band with a TM110 cavity as previously
described.1,22 The microwave frequency was measured
with a Model 575 microwave counter (EIP Microwave,
Inc., San Jose, CA). To measure superoxide anion (O°2 )
generation from phorbol-ester (PMA)-stimulated PMNs,
EPR spin-trapping studies were performed using the spin-
trap DEPMPO (Oxis, Portland, OR) at 10 mM. 1 · 106
PMNs were stimulated with PMA (1 ng/mL) and loaded
into capillary tubes for EPR measurements. To determine
the radical scavenging ability of 3 and 8 in aqueous and
ethanolic formulations, PMNs were pre-incubated for
5 min with test compound, followed by PMA stimulation.
Instrument settings used in the spin-trapping experiments
were as follows: modulation amplitude, 0.32 G; time
constant, 0.16 s; scan time, 60 s; modulation frequency,
100 kHz; microwave power, 20 mW; and microwave
frequency, 9.76 GHz. The samples were placed in a quartz
EPR flat cell, and spectra were recorded. The component
followed by
a solution of benzyl alcohol (8.1 mL,
77.8 mmol) in Et2O (20 mL). The mixture was stirred at
0 ꢁC for 30 min and then at rt overnight. The mixture was
filtered and the filtrate concentrated to give a colorless oil.
Silica chromatography (hexanes/Et2O/triethylamine, 4/1/
1%) of the crude product yielded 4 (5.68 g, 83%) as a clear,
colorless oil that was stored under N2 at ꢀ20 ꢁC; 1H
NMR: d 7.38 (15 H, m), 4.90 (6H, d).Dibenzyl phospho-
roiodidate (5). To a solution of tribenzyl phosphite
(5.43 g, 15.4 mmol) in CH2Cl2 (8 mL) at 0 ꢁC was added
I2 (3.76 g, 14.8 mmol). The mixture was stirred at 0 ꢁC for
10 min or until the solution became clear and colorless.
The solution was then stirred at rt for 10 min and used
directly in the next step.
(Monobenzyl-phosphoryloxy)-(phosphoryloxy)-b,e-caro-
tene (6). To a solution of lutein (1) (0.842 g, 1.48 mmol) in
CH2Cl2 (8 mL) were added pyridine (4.8 mL, 59.2 mmol).
The solution was stirred at 0 ꢁC for 5 min and then freshly
prepared 5 (14.8 mmol) in CH2Cl2 (8 mL) was added
dropwise to the mixture at 0 ꢁC. The solution was stirred
at 0 ꢁC for 1 h and then diluted with CH2Cl2 and
quenched with brine. The aqueous layer was extracted
twice with CH2Cl2 and the combined organic layer was
washed once with NaSSO4, once with brine, then dried
over Na2SO4 and concentrated. Pyridine was removed by
azeotropic distillation using toluene to yield mono benzyl-
protected diphosphate 6, used in the next step without
further purification; LC/MS (APCI): 9.67 min (13.31%),
kmax 268 nm (26%), 423 nm (74%), 446 nm (100%), 476 nm
(84%); m/z 850 (5%), 825 (4%), 810 (100%), 532 (96%);
10.02 min (86.69%), kmax 268 nm (26%), 423 nm (72%),
446 nm (100%), 476 nm (89%); m/z 850 (5%), 825 (4%),
810 (100%), 532 (92%).
3,30-Diphosphoryloxy-b,e-carotene (7). To a solution of 6
(1.48 mmol) in CH2Cl2 (10 mL) at 0 ꢁC were added
pyridine (1.2 mL, 14.8 mmol) and then bromotrimethylsi-
lane (0.97 mL, 7.40 mmol). The solution was stirred at
0 ꢁC for 30 min, quenched with triethylamine, diluted with
CH2Cl2, and then concentrated to yield crude diphosphate
7 as a red-orange oil, used in the next step without further
purification; LC/MS (APCI): 8.90 min (54.88%), kmax
268 nm (20%), 423 nm (70%), 446 nm (100%), 476 nm
(90%); m/z 693 (5%), 639 (48%), 555 (42%), 538 (100%);
9.18 min (43.33%), kmax 423 nm (78%), 446 nm (100%),
476 nm (91%); m/z 693 (7%), 639 (45%), 555 (38%), 538
(100%).
3,30-Diphosphoryloxy-b,e-carotene sodium salt (8). To a
solution of crude 7 (1.48 mmol) in CH2Cl2 (10 mL) at 0 ꢁC
was added dropwise sodium methoxide (25% wt in
methanol; 6.77 mL, 29.6 mmol). The solution was stirred
at rt overnight and then diethyl ether was added to the
salt. The suspension was centrifuged and the supernatant
discarded. Water was added to the salt and the suspension
was centrifuged and the supernatant discarded. The salt
was redissolved in methanol and diluted with water.
Lyophilization of the clear, red-orange aqueous solution
yielded 8 (0.38 g, 35% over three steps) as a red-orange,
hygroscopic solid; LC/MS (APCI): 8.54 min (25.86%),
kmax 268 nm (25%), 423 nm (74%), 446 nm (100%), 474 nm
(68%); 8.85 min (31.13%), kmax 268 nm (20%), 423 nm
(66%), 446 nm (100%), 474 nm (80%), m/z 912 (50%), 780
(18%), 692 (7%), 630 (100%), 550 (45%);9.15 min
(30.62%), kmax 268 nm (23%), 423 nm (75%), 446 nm

G. Nadolski et al. / Bioorg. Med. Chem. Lett. 16 (2006) 775–781
781
signals in the spectra were identified and quantified as
reported previously.23,24
26. Savignac, P.; Lavielle, G. Bull. Soc. Chim. Fr. 1974, 7,
1506.
21. Fraticelli, A.; Serrano, C. V., Jr.; Bochner, B. S.; Capog-
rossi, M. C.; Zweier, J. L. Biochim. Biophys. Acta 1996,
1310, 251.
22. Wang, P.; Zweier, J. L. J. Biol. Chem. 1996, 271, 29223.
23. Lee, C.; Miura, K.; Liu, X.; Zweier, J. L. J. Biol. Chem.
2000, 275, 38965.
24. Roubaud, V.; Sankarapandi, S.; Kuppusamy, P.; Tordo,
P.; Zweier, J. L. Anal. Biochem. 1997, 247, 404.
25. Stowell, J. K.; Widlanski, T. S. Tetrahedron Lett. 1995, 36,
1825.
27. Smith, D. J. H.; Ogilvie, K. K.; Gillen, M. F. Tetrahedron
Lett. 1980, 21, 861.
28. Salomon, C. J.; Breuer, E. Tetrahedron Lett. 1995, 36,
6759.
´
29. Zsila, F.; Bikadi, Z.; Keresztes, Z.; Deli, J.; Simonyi, M.
J. Phys. Chem. B 2001, 105, 9413.
´
30. Zsila, F.; Bikadi, Z.; Deli, J.; Simonyi, M. Chirality 2001,
13, 446.
31. Thurmann, P. A.; Schalch, W.; Aebischer, J. C.; Tenter,
¨
U.; Cohn, W. Am. J. Clin. Nutr. 2005, 82, 88.