Chemistry of trans-resveratrol with singlet oxygen: [2 + 2] addition, [4 + 2] addition, and formation of the phytoalexin moracin M

Article abstract of DOI:10.1021/ol201922u

Resveratrol (1) reacts with singlet oxygen by two major pathways: A [2 + 2] cycloaddition forming a transient dioxetane that cleaves into the corresponding aldehydes and a [4 + 2] cycloaddition forming an endoperoxide that, upon heating, undergoes a rearrangement to moracin M. The rate constant by which singlet oxygen is removed by 1 (k_T) was determined by time-resolved infrared luminescence spectroscopy to be 1.5 ×10⁶ M ^-1 sec^-1 in CD₃OD, smaller than previously reported values. Chemical reaction accounts for ca. 25% of k_T.

Full text of DOI:10.1021/ol201922u

ORGANIC
LETTERS
2011
Vol. 13, No. 18
4846–4849
Chemistry of trans-Resveratrol with
Singlet Oxygen: [2 þ 2] Addition,
[4 þ 2] Addition, and Formation of
the Phytoalexin Moracin M
Jeff A. Celaje, Dong Zhang, Angela M. Guerrero, and Matthias Selke*
Department of Chemistry and Biochemistry, California State University, Los Angeles,
Los Angeles, California 90032, United States
mselke@calstatela.edu
Received July 15, 2011
ABSTRACT
Resveratrol (1) reacts with singlet oxygen by two major pathways: A [2 þ 2] cycloaddition forming a transient dioxetane that cleaves into the
corresponding aldehydes and a [4 þ 2] cycloaddition forming an endoperoxide that, upon heating, undergoes a rearrangement to moracin M. The
rate constant by which singlet oxygen is removed by 1 (kT) was determined by time-resolved infrared luminescence spectroscopy to be 1.5 ꢀ
106 Mꢁ1 secꢁ1 in CD3OD, smaller than previously reported values. Chemical reaction accounts for ca. 25% of kT.
A variety of beneficial health effects1 (cardiovascular
modulating effects,1,2 antiaging effects,1,3 and even anti-
tumor effects1,4) have been attributed to the phenolic
antioxidant trans-resveratrol (1, trans-5-(para-hydroxystyryl)-
resorcinol), which is found in numerous plants, including
Vitis vinifera (wine grapes). It is believed that so-called
reactive oxygen species react with 1, leading to their removal
and formation of several trans-resveratrol oligomers.5
These oligomers may also have a wide range of health
benefits, including cyclooxygenase inhibitory activity.5 Itis
therefore not surprising that in recent years a very large
amount of research has been undertaken to elucidate the
nature of the antioxidative properties of trans-resveratrol.
There have been several recent reports that trans-resver-
atrol is a selective quencher of singlet oxygen and thus may
protect tissues and cells from photooxidative damage.6
This includes inhibition of photooxidation of the fluoro-
phore A2E, a pyridinium bisretinoid that acts as a singlet
(1) (a) Sinclair, D. A.; Baur, J. A. Nat. Rev. Drug Discovery 2006, 5,
493–506. (b) Das, M.; Das, D. K. Mol. Aspects Med. 2010, 31, 503–512.
(2) (a) Baur, J. A.; Pearson, K. J.; Price, N. L.; Jamieson, H. A.;
Lerin, C.; Kalra, A.; Prabhu, V. V.; Allard, J. S.; Lopez-Lluch, G.;
Lewis, K.; Pistell, P. J.; Poosala, S.; Becker, K. G.; Boss, O.; Gwinn, D.;
Wang, M.; Ramaswamy, S.; Fishbein, K. W.; Spencer, R. G.; Lakatta,
E. G.; Le Couteur, D.; Shaw, R. J.; Navas, P.; Puigserver, P.; Ingram,
D. K.; de Cabo, R.; Sinclair, D. A. Nature 2006, 444, 337–342. (b) Soleas,
G. J.; Diamandis, E. P.; Goldberg, D. M. Clin. Biochem. 1997, 30, 91–
113. (c) Lee, S. K.; Zhang, W.; Sanderson, B. J. S. J. Agric. Food Chem.
2008, 56, 7572–7577.
(3) (a) Howitz, K. T.; Bitterman, K. J.; Cohen, H. Y.; Lamming,
D. W.; Lavu, S.; Wood, J. G.; Zipkin, R. E.; Chung, P.; Kisielewski, A.;
Zhang, L.-L.; Scherer, B.; Sinclair, D. A. Nature 2003, 425, 191–196.
(b) Cohen, H. Y.; Miller, C.; Bitterman, K. J.; Wall, N. R.; Hekking, B.;
Kessler, B.; Howitz, K. T.; Gorospe, M.; de Cabo, R.; Sinclair, D. A.
Science 2004, 305, 390–392. (c) Westerheide, S. D.; Anckar, J.; Stevens,
S. M., Jr.; Sistonen, L.; Morimoto, R. I. Science 2009, 323, 1063–1066.
(4) (a) Jang, M.; Cai, L.; Udeani, G. O.; Slowing, K. V.; Thomas,
C. F.; Beecher, C. W. W.; Fong, H. H. S.; Farnsworth, N. R.; Kinghorn,
A. D.; Mehta, R. G.; Moon, R. C.; Pezzuto, J. M. Science 1997, 275,
218–220. (b) Lin, J.-N.; Lin, V. C.-H.; Rau, K.-M.; Shieh, P.-C.; Kuo,
D.-H.; Shieh, J.-C.; Chen, W.-J.; Tsai, S.-C.; Way, T.-D. J. Agric. Food
Chem. 2010, 58, 1584–1592.
(5) Cichewicz, R. H.; Kouzi, S. A. In Studies in Natural Products
Chemistry; Rahman, A. U., Ed.; Elsevier Science: New York, 2002; Vol. 26,
pp 506ꢁ579.
(6) (a) Sparro, J. R.; Vollmer-Snarr, H. R.; Zhou, J.; Jang, Y. P.;
Jockush, S.; Itagaki, Y.; Nakanishi, K. J. Biol. Chem. 2003, 278, 18207–
18213. (b) Rimando, A. M.; Cuendet, M.; Desmarchelier, C.; Mehta,
R. G.; Pezzuto, J. M.; Duke, S. O. J. Agric. Food Chem. 2002, 50, 3453–
3457.
r
10.1021/ol201922u
Published on Web 08/22/2011
2011 American Chemical Society

oxygen sensitizer.6a The self-sensitized photooxidation of
A2E leads to macular degeneration, which is inhibited by
trans-resveratrol.6a Pan et al.7 reported a very large rate
constant for the total singlet oxygen removal (kT; this rate
constant is the sum of all processes by which singlet oxygen
is removed, namely physical quenching and chemical
reaction) by 1 in an aqueous system, that is, kT = 3.92 ꢀ
109 Mꢁ1 secꢁ1, while Jung et al.8 measured a kT value of
2.55 ꢀ 107 Mꢁ1 secꢁ1 in methanol. The latter paper also
suggested that there is a small amount (ca. 5% of kT) of
chemical reaction of 1 with singlet oxygen, although no
products were reported. On the other hand, Pan et al.
suggested, based on GC-MS analyses, that the resorcinol
ring is oxidized by singlet oxygen to its corresponding
endoperoxide, followed by rearrangement to quinones.7
However, no NMR analysis of the products was per-
formed, and reactivity of the electron-rich trans double
bond was not considered by any of the authors. Given the
very large discrepancy of the reported rate constants for
singlet oxygen removal (over 2 orders of magnitude), the
resulting uncertainty over the rate constant ratio of che-
mical reaction vs physical quenching, and the lack of
conclusive assignment of any reaction products, a detailed
study of the chemistry of singlet oxygen with trans-resver-
atrol is overdue. We now report that trans-resveratrol
reacts with singlet oxygen via two entirely different path-
ways, both of which involve the trans double bond, namely
[2 þ 2] addition and [4 þ 2] addition. The latter pathway
leads to an endoperoxide that undergoes a surprising
rearrangement to the phytoalexin moracin M.
oxygen removal by other phenolic antioxidants under
basic conditions.9
Table 1. Rate Constants of Singlet Oxygen Removal by
trans-Resveratrol
kr (1) ꢀ 10ꢁ5
kT (1) ꢀ 10ꢁ6
solvent
CD3CN
temp (°C)
(Mꢁ1secsꢁ1 a
)
(Mꢁ1secsꢁ1 b
)
22
0
1.6 ( 0.1
1.2 ( 0.2
1.5 ( 0.1
9.2 ( 0.2
370 ( 20
3.0 ( 0.2
CD3OD
22
22
22
CD3OD/D2O c
D2O, pH = 10 d
a kr = rate constant for chemical reaction of 1 with singlet oxygen;
average of three runs, error is one standard deviation. Sensitizer:
methylene blue, cutoff filter at 493 nm. b Average of three runs, error
is one standard deviation. Sensitizer: methylene blue, λexc = 532 nm.
Rate constants were determined from pseudofirst order decay of the
singlet oxygen luminescence. c 3:2 ratio of CD3OD/D2O. d 0.05 M
Na2CO3/0.05 M NaHCO3 buffer in D2O.
We have determined the total rate by which singlet
oxygen is removed by 1 using time-resolved near-infrared
(NIR) spectroscopy (Figure 1). This was accomplished by
flash generation of singlet oxygen followed by monitoring
its near-infrared luminescence decay in a variety of sol-
vents. Rate constants for singlet oxygen removal are
summarizedinTable1. OurvaluesinCD3CNand CD3OD
are somewhat smaller than those obtained by Jung et al.8
(about 1 order of magnitude) in methanol and over 3
orders of magnitude smaller than those reported by Pan
et al.7 Neither one of these groups used time-resolved
methods, which are considered far more accurate than
any other methodology for the determination of singlet
oxygen quenching rate constants. There is no significant
variation of our rate constants in protic vs aprotic solvents
(i.e., methanol vs acetonitrile), indicating that hydrogen
bonding by solvent molecules to the hydroxy groups of 1
does not affect the value of kT and cannot account for the
differences between our kT values and those previously
reported in the literature. At high pH in D2O, the value of
kT for 1 increases considerably (more than 2 orders of
magnitude). Likewise, in a 3:2 CD3OD/D2O mixture
(where ionization of the phenol groups is higher) the kT
value is larger (by a factor of 6) thanin neat CD3OD. These
increases are similar to what has been observed for singlet
Figure 1. Singlet oxygen removal by trans-resveratrol in CD3CN.
In contrast to previous reports,7,8 we have found that
trans-resveratrol reacts with singlet oxygen leading to a
rather complex product mixture. Careful NMR analyses
indicated that there are three major products. For all cases,
the trans double bond of 1 must be involved in the reaction,
since the signals for the trans vinyl protons in the 1H NMR
completely disappear when the progress of the reaction is
directly monitored by carrying out the photooxidation in
an NMR tube. Thus, the previous suggestion;based solely
on mass spectral analyses;that singlet oxygen adds across
the resorcinol ring cannot be correct. Indeed, we did not
find any quinone products. Two products were separated
from the crude product mixture and purified by flash column
chromatography on silica gel. NMR analyses, including
comparison with authentic samples, clearly demonstrated
that these products were 4-hydroxybenzaldehyde (2) and
3,5-dihydroxybenzaldehyde (3). Their formation can eas-
ily be rationalized by [2 þ 2] addition of singlet oxygen to
the electron-rich trans double bond of 1 followed by clea-
vage of the resulting dioxetane (Scheme 1).10 The [2 þ 2]
(9) Bisby, R. H.; Morgan, C. G.; Hamblett, I.; Gorman, A. A.
J. Phys. Chem. A 1999, 103, 7457–7459.
(10) (a) Richardson, W. H.; Batinica, G.; Janota-Perret, K.; Miller,
T.; Shen, D. J. Org. Chem. 1991, 56, 6140–6144. (b) Ohkubo, K.; Nanjo,
T.; Fukuzumi, S. Org. Lett. 2005, 7, 4265–4268.
(7) Jiang, L.-Y.; He, S.; Jiang, K.-Z.; Sun, C.-R.; Pan, Y.-J. J. Agric.
Food Chem. 2010, 58, 9020–9027.
(8) Jung, M. Y.; Choi, D. S. J. Agric. Food Chem. 2010, 58, 11888–
11895.
Org. Lett., Vol. 13, No. 18, 2011
4847

addition channel accounts for nearly 40% of the products
for the reaction of 1 with singlet oxygen. Detailed NMR
analyses revealed that the major product was an endoper-
oxide formed by a [4 þ 2] addition of singlet oxygen to the
trans double bond and the adjacent CdC double bond
of the para-phenol ring of 1, followed by tautomerization
of the primary cycloaddition product to the corre-
sponding ketone 4. All attempts to isolate endoperoxide
4 as a pure compound by column chromatography led to
its decomposition.
derivative, namely the known phytoalexin moracin M
(5); this assignment was confirmed by comparison with
the NMR data published in the literature.13 Moracin M
has been isolated from Morus alba L. (Moraceae) in
China; the root bark of Morus alba L. is extensively used
in traditional Chinese medicine.13b Moracin M has been
reported to have antimicrobial activity, and is also
believed to be a precursor to other arylbenzofurans with
potent antioxidant activity.14 The formation of 5 from the
[4 þ 2] addition product from the reaction of 1 with singlet
oxygen represents a very simple and convenient synthesis
of moracin M.15 A possible mechanism for the formation
of 5 from 4 is depicted in Scheme 2. The rearrangement
should initially involve homolytic cleavage of the OꢁO
bond of 4 followed by a 1,5-hydrogen shift to a γ-hydroxy
enone. Cyclization to a hemiketal followed by loss of
water concomitant with aromatization would then lead
to moracin M.
Scheme 1. Reactivity Pathways Leading to the Observed Pro-
duct Mixture of the Reaction of Singlet Oxygen and trans-
Resveratrol
Scheme 2. Proposed Mechanism of Moracin M Formation
Endoperoxide 4 accounts for nearly 60% of the products
for the reaction of 1 with singlet oxygen. In addition to the
three products mentioned above, we also noted a very
small amount (<5%) of an as-of-yet unidentified com-
pound that is a secondary photoproduct formed upon
prolonged irradiation of product 4. While formation of 4
was unexpected, a similar [4 þ 2] cycloaddition of singlet
oxygen and trans-4-propenyl anisole has been reported by
Greer et al.11 However, since the anisole studied by Greer
et al. did not possess the para-hydroxy group present in
compound 1, their primary product did not undergo the
tautomerization that led to formation of 4 from the [4 þ 2]
adduct of resveratrol.
A substrate may either react with singlet oxygen and/or
physicallyremoveitwithoutchemicalchange(i.e., physical
quenching of singlet oxygen). Previous reports have claimed
that physical quenching of singlet oxygen may be an
important aspect of the biological function of 1.7,8 To
evaluate whether there is significant physical quenching by
1, we have conducted competition experiments with sev-
eral singlet oxygen acceptors that remove singlet oxygen
only by chemical reaction. We have found that chemical
reaction of 1 with singlet oxygen is rather slow, and con-
sequently many singlet oxygen coacceptors commonly
used for this type of experiment (for example, 9,10-
dimethylanthracene) were unsuitable as they were comple-
tely consumed before an appreciable amount (>5%)
of 1 was photooxidized. We eventually found that 1,4-
Endoperoxide 4 is stable at room temperature, but upon
warming, it undergoes further rearrangement.12 Detailed
1
NMR analyses (including H, 13C, and COSY) indicate
that the product of this rearrangement is an arylbenzofuran
(11) Greer, A.; Vassilikogiannakis, G.; Lee, K.-C.; Koffas, T. S.;
Nahm, K.; Foote, C. S. J. Org. Chem. 2000, 65, 6876–6878.
(12) Our rearrangement leading to Moracin M involves a six-mem-
bered endoperoxide, in contrast with the bicyclic furan-derived endo-
peroxide rearrangements studied by Vassilikogiannakis et al.:
Montagnon, T.; Tofi, M.; Vassilikogiannakis, G. Acc. Chem. Res.
2008, 41, 1001–1011.
(13) (a) Fukai, T.; Hano, Y.; Hirakura, K.; Nomura, T.; Uzawa, J.
Chem. Pharm. Bull. 1980, 32, 808–811. (b) Piao, S.-J.; Qiu, F.; Chen, L.-
X.; Pan, Y.; Dou, D.-Q. Helv. Chim. Acta 2009, 92, 579–587.
(14) Kapche, G. D. W. F.; Fozing, C. D.; Donfack, J. H.; Fotso,
G. W.; Amadou, D.; Tchana, A. N.; Bezabih, M.; Moundipa, P. F.;
Ngadjui, B. T.; Abegaz, B. M. Phytochemistry 2009, 70, 216–221.
(15) A transition metal-directed total synthesis of 5 has previously
been reported: Widdowson, D. A.; Mann, I. S. Tetrahedron 1991, 47,
7981–7990.
4848
Org. Lett., Vol. 13, No. 18, 2011

dimethylnaphthalene (1,4-DMN) was a suitable coaccep-
tor, and measured the rate constant for its reaction with
singlet oxygen in CD3CN to be 2.9 ꢀ 105 Mꢁ1 secꢁ1 at
room temperature and 2.7 ꢀ 105 Mꢁ1 secꢁ1 at 0 °C. For
this reference compound, all of the singlet oxygen removal
is due to endoperoxide formation, that is, kT = kr.16 The
endoperoxide formed from reaction of singlet oxygen and
1,4-DMN slowly reverts back to 1,4-DMN (ca. 50% over
5 h) at room temperature17 but is indefinitely stable at 0 °C.
We therefore conducted the competition experiments be-
tween 1 and 1,4-DMN at 0 °C. Using the methodology
developed by Higgins et al.,18 we obtained a value for the
chemical rate constant kr for the reaction of singlet oxygen
with trans-resveratrol of 3.0 ꢀ 105 Mꢁ1 secꢁ1. Since the
reported in this paper, that is, in the 106ꢁ107 Mꢁ1 secꢁ1
range.20,21 Given the rather small value of kT by 1,
especially compared with tocopherols, we do not think
that singlet oxygen scavenging is the main pathway from
which the health benefits of 1 originate. On the other hand,
our data demonstrates that chemical reaction of singlet
oxygen with 1 is a significant pathway in the interaction of
1 with singlet oxygen. The reaction of 1 with singlet oxygen
does not lead to formation of quinones, but instead
involves an intriguing [4 þ 2] cycloaddition channel lead-
ing to an endoperoxide that can rearrange to the phytoa-
lexin moracin M. Further experiments to elucidate the
chemistry of singlet oxygen with glycosylated and oligo-
meric derivatives of 1 are in progress.
value of kT for 1 in CD3CN at 0 °C is 1.2 ꢀ 106 Mꢁ1 secꢁ1
,
kr is approximately 25% of kT. Hence, physical quenching
of singlet oxygen by 1 accounts for about 75% of the total
removal of singlet oxygen.
The total rate constant for singlet oxygen removal by 1 is
considerably smaller than previously reported, and is in
fact somewhat smaller than those of other phenolic anti-
oxidants such as vitamin E.19 However, other resorcinol
derivativesthatdo not possessthe aromatic ethermoietyof
tocopherols have kT values very similar to the ones
Acknowledgment. We gratefully acknowledge support
from the NIH-NIGMS 5SC1GM084776. J.A.C. is also
acknowledging support from the NIH-RISE program
(GM6133), and A.M.G. is acknowledging support
from the NIH MARC program (GM08228). We thank
Timothy Dong (UCLA), Dr. Errol Mathias (CSULA),
and Dr. Ali Jabalameli (CSULA) for valuable experimen-
tal assistance.
Supporting Information Available. Experimental pro-
cedures, full spectroscopic data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(16) Gollnick, K.; Griesbeck, A. G. Tetrahedron Lett. 1984, 25, 725–
728.
(17) Adam, W.; Prein, M. Acc. Chem. Res. 1996, 29, 275–283.
(18) Higgins, R.; Foote, C. S.; Cheng, H. Adv. Chem. Ser. 1968, 77,
102–117.
(19) (a) Clough, R. L.; Yee, B. G.; Foote, C. S. J. Am. Chem. Soc.
1979, 101, 683–686. (b) Foote, C. S.; Ching, T.-Y.; Geller, G. G. Photo-
chem. Photobiol. 1974, 20, 511–513. (c) Fahrenholtz, S. R.; Doleiden,
F. H.; Trozollo, A. M.; Lamola, A. A. Photochem. Photobiol. 1974, 20,
505–509.
(20) Mukai, K.; Daifuku, K.; Okabe, K.; Tanigaki, T.; Inoue, K.
J. Org. Chem. 1991, 56, 4188–4192.
ꢀ
(21) Martire, D. O.; Braslavsky, S. E.; Garcı
´
a, N. A. J. Photochem.
Photobiol., A 1991, 61, 113–124.
Org. Lett., Vol. 13, No. 18, 2011
4849