Modulation of photochemical damage in normal and malignant cells by naturally occurring compounds

Article abstract of DOI:10.1111/j.1751-1097.2012.01156.x

Certain phytochemicals, such as the stilbene, resveratrol (RES, found in red grapes and berries), and the triterpenoid, ursolic acid (UA, found in waxy berries and herbs such as rosemary and oregano), have antioxidant, anti-inflammatory and antiproliferative effects. Two human-derived cell lines, hTERT-RPE with a nonmalignant phenotype derived from retinal pigment epithelium, and ATCC CRL-11147 derived from a malignant skin melanoma, were used as in vitro models of photooxidative stress produced by exposure to the broadband output of a 150 W Hg vapor arc lamp at an irradiance of 19-26 mW cm ^-2. In untreated cells, UV-VIS broadband light exposure produced a loss of proliferative ability, an activation of NF-κB and an increase in protein carbonyl adducts at 24 h postexposure. Pretreatment of the cells with RES or UA at 1-2 μmsignificantly reduced the amount of phosphorylated NF-κB at 24 h postexposure. RES pretreatment reduced the burden of light-induced protein carbonyl adducts by up to 25% in exposed cells. UA treatment markedly increased the sensitivity of melanoma cells to UV radiation, while conferring some photoprotection to RPE cells. These observations indicate that phytochemicals modulate the cellular response to photochemical stress by interacting with specific cell-signaling pathways. There is heightened interest in naturally occurring compounds as potential therapeutic agents. One such compound is ursolic acid (UA), a triterpenoid found in common foods such as cranberries, apples, as well as rosemary and other herbs. When skin melanoma (SM) cells are treated with UA, they become more susceptible to UV and visible light exposure, in contrast to normal phenotype retinal pigment epithelial (RPE) cells, whose sensitivity to light damage is unchanged by UA treatment. The figure shows the appearance of SM and RPE cells 24 h following light exposure with or without UA pretreatment. 2012 Wiley Periodicals, Inc. Photochemistry and Photobiology

Full text of DOI:10.1111/j.1751-1097.2012.01156.x

Photochemistry and Photobiology, 20**, **: *–*
Modulation of Photochemical Damage in Normal and Malignant Cells by
Naturally Occurring Compounds†
Yuan-Hao Lee1, Neeru C. Kumar2 and Randolph D. Glickman*1,2,3
1Department of Radiological Sciences, University of Texas Health Science Center, San Antonio, TX
2Department of Ophthalmology, University of Texas Health Science Center, San Antonio, TX
3Center for Biomedical Neuroscience, University of Texas Health Science Center, San Antonio, TX
Received 29 December 2011, accepted 28 March 2012, DOI: 10.1111/j.1751-1097.2012.01156.x
retinal pigment epithelial (RPE) lesions. Because of the chronic
light irradiation of retina and RPE, photochemical damage has
ABSTRACT
Certain phytochemicals, such as the stilbene, resveratrol (RES,
found in red grapes and berries), and the triterpenoid, ursolic
acid (UA, found in waxy berries and herbs such as rosemary and
oregano), have antioxidant, anti-inflammatory and antiprolifer-
ative effects. Two human-derived cell lines, hTERT-RPE with a
nonmalignant phenotype derived from retinal pigment epithe-
lium, and ATCC CRL-11147 derived from a malignant skin
melanoma, were used as in vitro models of photooxidative stress
produced by exposure to the broadband output of a 150 W Hg
vapor arc lamp at an irradiance of 19–26 mW cm)2. In
untreated cells, UV–VIS broadband light exposure produced a
loss of proliferative ability, an activation of NF-jB and an
increase in protein carbonyl adducts at 24 h postexposure.
Pretreatment of the cells with RES or UA at 1–2 lM signifi-
cantly reduced the amount of phosphorylated NF-jB at 24 h
postexposure. RES pretreatment reduced the burden of light-
induced protein carbonyl adducts by up to 25% in exposed cells.
UA treatment markedly increased the sensitivity of melanoma
cells to UV radiation, while conferring some photoprotection to
RPE cells. These observations indicate that phytochemicals
modulate the cellular response to photochemical stress by
interacting with specific cell-signaling pathways.
been postulated as a factor contributing to age-related retinal
degeneration (5,6). In addition, the use of exogenous antiox-
idants and free radical scavengers to block or reduce photo-
chemical damage has been extensively studied (7). An
important source of protective compounds are phytochemicals
naturally occurring in the human diet (sometimes called
‘‘nutraceuticals’’) (8). Two compounds, commonly found in
human diets and which have been found to confer some
protective actions against a variety of oxidative stress mech-
anisms, are resveratrol (RES) and ursolic acid (UA) (Fig. 1).
RES (see Fig. 1A) is a stilbene phytoalexin, a compound
produced by plants in response to stress, such as infections or
drought. It is commonly found in grapes, grape extracts and
derivative products, such as wine: RES is present in red wine
at 0.2–7.0 mg L)1 (9,10). Originally thought to be an estrogen
agonist because of its structural similarity to estrogen (11,12),
additional studies have found either ambiguous effects on the
estrogen receptor (13), or variable, concentration-dependent
effects (14). Other reported bioeffects of RES have been more
consistent, in particular antioxidant, anti-inflammatory, anti-
coagulative and vasodilatory actions. Due to the oxidation-
reduction activity of its hydroxyl substituents, RES can exert
direct antioxidant actions. In cardiac tissue, RES quenches
ROS (15–17). RES has also been demonstrated to scavenge
superoxide anion radicals and directly inhibit lipid peroxida-
tion by scavenging reactive oxygen radical intermediates
(15–17). RES may also maintain the concentration of thiol
compounds, such as glutathione (18), which serve as sub-
strates for glutathione peroxidases and other antioxidant
enzyme systems (19). The antioxidant effects of RES may
serve to prevent other chronic conditions, such as cancer.
Oxidative damage to DNA is known to produce gene
mutations through nucleotide base alterations, chromosome
deletions and frame shift errors, thus initiating potentially
tumorigenic processes (20,21). By scavenging ROS in the cell,
oxidative stress-induced DNA damage is reduced or pre-
vented (22).
INTRODUCTION
Light damage to the retina may occur through several
well-studied photophysical mechanisms: thermal, mechanical,
nonlinear (i.e. dielectric breakdown and plasma formation
associated with high peak power light pulses) and photochem-
ical processes (1–4). Photochemical damage, because it is most
often related to the effects of chronic light exposure and
therefore possibly amenable to prevention by external phar-
maceutical agents or nutritional factors, will be considered in
the present work. Excessive production of reactive oxygen
species (ROS) or other free radicals by the absorption of
energetic photons in tissue chromophores damages surround-
ing targets in the immediate cellular vicinity. Accumulation of
sufficient cellular damage ultimately leads to visible retinal and
UA (Fig. 1A) is a triterpene present, along with many other
phytochemicals, in fruits such as privet fruit (Ligustrum) and
hawthorne fruit (Cratagegi sp.), as well as members of various
medicinal plants, such as Rosemarinus officinalis, Ocimum
sanctum (and other members of the family Lamiaceae) and
†This invited paper is part of the Symposium-in-Print ‘‘Retinal Photodamage.’’
*Corresponding author email: glickman@uthscsa.edu (Randolph D. Glickman)
ꢀ 2012 Wiley Periodicals, Inc.
Photochemistry and Photobiology ꢀ 2012 The American Society of Photobiology 0031-8655/12
1

2
Yuan-Hao Lee et al.
Figure 1. Molecular structures of the compounds used in this study. (A) From left to right: trans-RES (the native form), cis-RES (the photoinduced
isomer), fl-RES (a highly fluorescent photoderivative; the structure shown here is that proposed by Galeano-Diaz et al. [42]) and ursolic acid. (B)
The UV absorption spectra of each of the above compounds; from left to right: t-RES, c-RES, fl-RES and ursolic acid. The absorption spectra of
t-RES, c-RES and fl-RES were obtained from the photodiode array detector during the separation of the samples shown in Fig. 3. The absorption
spectrum of ursolic acid, 10 lg mL)1, was obtained by standard spectrophotometry using a Shimadzu Pharmaspec UV-1700. The ursolic acid
sample was prepared by dilution in distilled water of a 1 mg mL)1 ethanolic stock solution (1% residual ethanol).
Eugenia jumbolana. These plants have been used since ancient
times in several traditional medicine pharmacopeias,
e.g. Greek and Chinese, to alleviate inflammatory conditions
and reduce hypertension (23,24). UA is reported to have a
wide range of actions in cells and tissues. In a study utilizing as
an endpoint the metabolism of arachidonic acid by murine
macrophages, human platelets and HL60 leukemic cells, UA
was found to inhibit the activity of lipoxygenase in the cells,
thus reducing the production of leukotrienes (24). In that same
study, cellular toxicity of UA was low; only at a concentration
of 100 lM was a loss of cellular viability observed. UA appears
to be especially active in the cellular environment; although
UA was relatively ineffective in scavenging superoxide anion in
a cell-free system, it was a very effective scavenger in a system
containing human neutrophils stimulated into a respiratory
burst by treatment with zymosan A (23). In contrast, in
different cell-free systems of lipid peroxidation and superoxide
anion production, UA was found to have a similar level of
antioxidant activity as a-tocopherol (25), and was generally
effective in preventing lipid peroxidation in an in vitro model
utilizing isolated rat liver microsomes (26). UA also inhibits
NF-jB activation at multiple sites in this transcription factor’s
pathway, including its phosphorylation and ability of p50 ⁄ p65
to bind to DNA (27). NF-jB is an important transcription
factor responding to a variety of external stressors, such as
toxins, pathogens, ischemia, shear stress and radiation-induced
oxidative stress (28–30). Photooxidative stress induced in cells
and tissues by light (UV and VIS) and laser exposures also
activates NF-jB (31–33); the transcription factor is regarded
as an important biomarker in photobiology (34). Through the
modulation of this transcription factor, UA reduces inflam-
mation not only by lowering the activity of proinflammatory
enzymes such as cyclooxygenase-2 (COX-2) and matrix
metalloproteinase-9, but also by exerting an antiproliferative
effect, thus conceivably conferring an anticancer action.
Indeed, UA has been demonstrated to have antiproliferative
effects on several cervical cancer cell lines (HeLa, CaSki and
SiHa) (35,36). In mouse models of skin cancer, a topical
application of UA reduced the number of skin cancers, and
lowered the activity of several proinflammatory markers,
produced by the application of several tumor promoters and
inflammatory agents (37,38). UA-supplemented chow inhib-
ited the growth of exogenous human mammary tumors in a
mouse model (39). Thus, UA exerts anticancer activity in both
in vitro and in vivo models.
Although UA itself has no strong chromophore in the UV
or VIS bands, RES is highly optically active (the optical
absorption spectra of these compounds are shown in Fig. 1).
UV and VIS light in the range of 254–550 nm will photoiso-
merize trans-RES (t-RES) to cis-RES (c-RES), as reported by
Trela and Waterhouse (40) and confirmed in our own
laboratory (unpublished observations), yet there are few
reports on the bioeffects of the cis isomer (41). Moreover,
the photoisomerization of t-RES to c-RES is not the only
light-sensitive reaction. Continued exposure of c-RES to UVA
produces an irreversible conversion to a tricyclic, phenan-
threne-like compound, fl-RES, that is highly fluorescent (42)
(Fig. 1A). We have confirmed the existence of this photode-
rivitization reaction, as well as the highly fluorescent nature of
the resulting product, and report here that this derivative is

Photochemistry and Photobiology
3
irradiance of 19 mW cm)2 for 7.5 min. Sham controls were placed on
the laboratory bench under normal ambient laboratory light for the
same period of time. Following exposure, the cells were returned to
the incubator for 1 or 24 h, without changing the culture medium. At
the end of the incubation period, NF-jB was assayed by ELISA as
described below.
capable of photosensitizing proteins in RPE and presumably
retinal cells. We have also studied the ability of RES and UA
to modulate the response of normal and cancer cells to
photochemical stress. In our previous work, we demonstrated
that exposing the cells to short-wavelength visible light, e.g. the
488 and 514.5 nm mixed output of an Argon laser, at an
irradiance of 10–20 mW cm)2 and an exposure duration of
5–10 min, produced photooxidative stress in the cells, as
indicated by fatty acid peroxidation (43), protein carbonyl
adducts (3) and NF-jB nuclear translocation (33). In the
present work, two in vitro models, a well-characterized RPE
cell line and a standardized human skin melanoma (SM) cell
line, were used to test the antioxidant effects of RES, its
photoderivatives and UA.
Measurement of p65 and p53 in cell extracts by immunoprecipitation
and Western blot. Following irradiation, the cells were returned to
normal growth medium, incubated for 4 h and harvested for protein
extraction. For the extraction of nuclear proteins, cell pellets were
processed according to published methods (45). For whole protein
extraction, modified RIPA buffer (50 mM Tris, pH 7.2, 150 mM NaCl,
1% NP-40, 1% sodium deoxycholate, 0.1% SDS and 2 mM EDTA)
with protease inhibitor cocktail was used. Protein concentration was
determined using the bicinchoninic acid (BCA; Pierce) assay. Immu-
noprecipitation was conducted with at least 100 lg total protein
(input) to conjugate with an antibody raised against the phosphory-
lated NF-jB p65 (p65-P) component (cat. no. SC-33039; Santa Cruz
Biotechnology, Santa Cruz, CA) and an anti-p53 antibody (cat. no.
P6749; Sigma-Aldrich) followed by incubation using agarose beads
(cat. no. SC-2003; Santa Cruz Biotechnology). Heat denaturation at
95ꢁC for 10 min was performed before the separation of protein
extracts on a 12% SDS gel (0.37 M, pH 8.8 Tris, 12% acrylamide ⁄ Bis
29:1, 0.1% SDS, 0.5 mg mL)1 and 0.05% TEMED) with a 4%
stacking gel (0.12 M, pH 6.8 Tris, 3.9% acrylamide ⁄ Bis 29:1, 0.1%
SDS, 0.5 mg mL)1 and 0.1% TEMED). During electrophoresis, the
current in the SDS running buffer (0.025 M Tris base and 0.19 M
glycine) was set to 80 mA for up to 2 h; for gel transfer, the voltage in
the transfer buffer (0.025 M Tris base, 0.19 M glycine, 0.05% SDS and
4% methanol) was set to 22 V overnight, with cooling applied to the
transfer apparatus. These data are reported in Fig. 10A,B.
MATERIALS AND METHODS
Cell lines. Experiments were conducted with the p53-reactive hTERT-
RPE (human retinal pigment epithelium) cell line (44) and the CRL-
11147 human SM cell line (ATCC, Manassus, VA). Cells were grown
in standard culture conditions in DMEM ⁄ F-12 and DMEM, respec-
tively, with 10% fetal calf serum. Both media were supplemented with
gentamycin (5 lg mL)1), penicillin (100 IU mL)1), streptomycin
(100 lg mL)1) and amphotericin B (250 ng mL)1) (all supplied by
Cellgro, Manassas, VA). The studies were designed to measure
radiation-induced oxidative stress and DNA damage in relation to
the antioxidant effect of UA.
Reagents. RES, >98% pure, from Polygonum root was obtained
from MP Bio (Solon, OH). UA, >90% pure, from cranberries was
obtained from Sigma-Aldrich (St. Louis, MO). The structures of RES
(MW 228.24), its isomer and photoderivative, as well as that of UA
(MW 456.70), are shown in Fig. 1A. The structure of fl-RES is that
proposed by Galeano-Diaz et al. (42).
A stock solution of RES was prepared at a concentration of
1 mg mL)1 in ethanol and stored at )20ꢁC. The UA stock solution
was prepared in ethanol or in DMSO (in different experiments) at a
concentration of 1.7 mg mL)1 and stored at 4ꢁC. Analysis by HPLC
indicated that these stock solutions were stable under these storage
conditions for at least 4 months. For use, the stock solutions were
diluted in cell culture medium to working concentrations of 0.1–
8.0 lM or as otherwise specified. The residual solvent (ethanol or
DMSO) in the culture media was 0.8% or less, depending on the
working concentration of the study compound. For the HPLC
measurements, chromatography grade acetic acid and methanol were
obtained from EMD (Gibbstown, NJ) and acetonitrile from Sigma-
Aldrich.
Measurement of NF-jB in RPE cells by ELISA. In some experi-
ments, the activation of NF-jB was assessed by determining the level
of p65-P in cytosolic lysates using an ELISA, following treatment with
RES and UA prior to light exposure, as described above. In these
experiments, RES and UA were tested over a concentration range of
0.1–2.0 lM. After the light exposure, the cells were incubated at 37ꢁC
for either 1 or 24 h, harvested from the Petri dishes, lysed and probed
for NF-jB activation using a commercial ELISA kit selective for the
phosphorylated p65 component (p65-P) of NF-jB (Cell Signaling
Technologies, Danvers, MA), according to the manufacturer’s proto-
col. These data are reported in Fig. 4A,B.
Photoderivitization of RES. A RES solution (10 lg mL)1 in etha-
nol) was exposed to the output of the 150 W Hg arc lamp at an
irradiance of 19 mW cm)2 for 20 min. This solution was freely open to
the atmosphere during the light exposure. Aliquots from this stock
were taken and diluted for use in the cell culture experiments.
Apoptosis assay. Cellular apoptosis was assayed at 4 h postirradi-
ation with the fluorochrome YO-PRO-1 and red PI (propidium
iodide), contained in the Vybrant Apoptosis Assay Kit #4 (Molecular
Probes ⁄ Invitrogen-Life Technologies). The manufacturer’s protocol
was followed for processing the cells and then the cells were observed
with a fluorescence microscope. Necrotic cells initially admit only PI
(giving red fluorescence), but exclude YO-PRO-1. Apoptotic cells, in
contrast, admit only YO-PRO-1 (giving green fluorescence). Eventu-
ally, dead cells may admit both probes and fluoresce yellow. Apoptotic
cells were identified by the presence of the green fluorescence of
YO-PRO-1 alone, and necrotic cells by the presence of PI alone. Cell
counting was performed in triplicate with randomly chosen fields for
determining apoptosis indices.
High performance liquid chromatography. Separation of RES iso-
mers and its photoderivative: Separations were carried out at RT using
a Waters lBondapak-C18 column (150 · 3.9 mm) and a mobile phase
consisting of 25% methanol ⁄ 10% acetonitrile ⁄ 1% acetic acid in water,
at a flow rate of 0.5 mL min)1. Twenty microliter aliquots were used
for the analysis. The RES isomers t-RES and c-RES are moderately
fluorescent, whereas fl-RES is highly fluorescent; however, of all of
these compounds, t-RES has the strongest absorption in the UV, with
a broad absorption peak in the range of 300–310 nm, as shown in
Fig. 1A (40). Therefore, tandem UV detection at 306 nm and
fluorescence detection with kEx = 315 nm and kEm = 382 nm were
employed to discriminate the RES isomers and the fluorescent
photoderivative.
Treatment and irradiation of cells. Cells were exposed to the
broadband output (containing UVA, UVB and VIS spectra) of a
150 W Schoeffel Instruments Hg vapor arc lamp for a specified
period of time. Two lamp configurations, which produced different
irradiances, were used to accommodate different cell culture formats.
For cells growing in standard 24- or 96-well plates or 100-mm Petri
dishes, the exposure irradiance was 19 mW cm)2. For cells growing
in 6-cm Petri dishes, the exposure irradiance was 26.7 mW cm)2
.
Two exposure protocols were used: (1) For the experiments in which
p65-P and p53 were measured by immunoprecipitation and Western
blot techniques, the cells were grown until they were fully attached,
pretreated with either RES or UA for durations ranging from 1 to
8 h, and then the treatment was terminated by media replacement
(i.e. no phytochemical was present) prior to irradiation. Cells were
exposed for 10 min to the Hg vapor arc lamp at an irradiance of
26.7 mW cm)2, in the complete growth medium, including phenol red
indicator. (2) For the experiments in which NF-jB (p65-P) was
measured by ELISA, cells were plated in 100-mm diameter Petri
dishes and allowed to grow to confluence. Prior to light exposure, the
standard culture medium was replaced with phenol indicator-free
medium containing the phytochemical, either RES or UA at the
desired concentration, and the cells were incubated at 37ꢁC for 1 h.
Following this treatment period, the culture dishes containing the
cells were exposed to the output of the Hg vapor arc lamp at an

4
Yuan-Hao Lee et al.
Assessment of cellular oxidative stress using 2¢,7¢-dichlorodihydro-
fluorescein (DCFH): Esterified DCFH diacetate stock solution
(Molecular Probes ⁄ Invitrogen-Life Technologies) was prepared as
0.5 mg mL)1 in methanol and stored at )20ꢁC. The working dilution
of 10 lM DCFH was made in complete culture media shortly before its
use in the experimental protocol; the media containing the DCFH was
kept in a light-tight container until it was added to the culture plates.
After 24 h of UA treatment, 10 lM DCFH diacetate was administered
to cells 1 h prior to irradiation. Intact suspended and attached cells
were collected immediately after irradiation. Extraction of probes
from the cells and HPLC analysis of oxidized DCFH were performed
as previously described (46). Briefly, under the Phillips F32T8
fluorescent laboratory lighting shielded by yellow (550 nm cut-on
longpass) filters, the cell lysates were extracted on Waters Oasis HLB
solid-phase extraction cartridges using methanol as the elution solvent.
The eluates were dried and resuspended in mobile phase (8 mM
ammonium phosphate, pH 8.0, in 60% methanol). The HPLC
analysis was carried out on a 3.9 mm · 300 mm Phenomenex Bond-
clone-C18 column, at a flow rate of 1 mL min)1. Twenty microliters
aliquots were injected into the column and separations were carried
out at RT. Fluorescence detection parameters were kEx = 488 nm and
kEm = 530 nm.
Protein carbonyl assay. Carbonyl adducts produced by oxidative
stress to proteins were assessed at 24 h following a 10 min exposure of
RPE cells to the Hg arc lamp at 19 mW cm)2, as described above. The
effects of a 1 h pretreatment with t-RES or fl-RES on protein
oxidation were tested by adding these compounds to the growth media.
A dilution of a stock solution of t-RES was added, so the final
concentration in the growth media was 10 lM. The fl-RES was
prepared by photoderivitization of t-RES, as described above. The
precise concentration of fl-RES was not known exactly, but an amount
of the fl-RES material was added to the growth media to achieve a 1:10
dilution of the original stock. Measurement of carbonyl adducts was
carried out with the Oxyblot assay (Chemicon ⁄ Millipore, Billerica,
MA), following the manufacturer’s protocol. In the Oxyblot assay,
RPE cells were lysed and the soluble proteins reacted with dini-
trophenylhydrazine to derivatize the adducts, which were then
separated using polyacrylamide gel electrophoresis, followed by
Western blots using a chemiluminescence-linked antibody (supplied
in the Oxyblot kit) to detect the derivatives. The intensity of the
resulting images of the luminescent transfers was measured and
analyzed using the Image Pro software package (Media Cybernetics,
Bethesda, MD). By taking a linear measurement of the pixel intensity
along the length of the gel image, and then integrating the area under
that curve, a quantitative measure of the relative, total protein
carbonyl content of the retinal proteins was obtained for each
experimental condition. Additional details on this analytical technique
have been previously reported (47).
Figure 2. A family of fluorescence excitation–emission curves, show-
ing the increasing yield of the fluorescent derivative of RES (Fl-RES),
proportional to the duration of light exposure. A 1 mM stock solution
of native RES, i.e. the trans isomer, was made in ethanol. Two
milliliters of this solution was placed in a quartz spectrophotometer
cuvette, and irradiated with a Hg vapor arc lamp at 19 mW cm)2 for
various times up to 7.5 min. After each exposure time was completed,
an aliquot was taken, diluted to a working concentration of 2 lM in
distilled water, and placed in a spectrofluorimeter to measure its
excitation–emission curve. Note the excitation peak at 315 nm, and
emission peaks at 360, 380 and 405 nm. Spectra were recorded in a
Jobin-Yvon-Spex Fluorolog-3 spectrofluorimeter.
BrdU staining. RPE cells were plated to about 50% confluency in
Chamber Slides (Nalge Nunc International). After 24 h of UA
treatment, all cell culture media were replaced by fresh media
containing 10 lM BrdU. Cell fixation was performed with 70%
ethanol followed by washing and incubation with anti-BrdU antibody
and secondary fluorescein antibody, sequentially. Hoechst 33342
(10 lM) was applied for 2 min before microscopic observation, in
order to stain all cell nuclei present in the field.
Statistical analysis. All data were normalized relative to the
corresponding internal control. Error bars were drawn based on the
ranges of disparity (in cases of n = 2) or the standard deviation values
(in cases of n > 2). Other statistical tests, specifically ANOVA
followed by the Bonferroni multiple comparison tests, were carried
out using the ProStat statistical software (Polysoftware, Pearl River,
NY).
Figure 3. HPLC analysis of RES and its photoproducts. Lower trace:
native trans-RES (in the dark). Upper trace: photoproducts formed by
RES exposure to Hg arc lamp at 19 mW cm)2 for 20 min. Detection of
all compounds was by fluorescence detection with kEx = 315 nm and
kEm = 382 nm. Small peaks at 6.5 and 8 min were not identified.
Other chromatographic details are given in the Materials and Methods
section. The UV spectra of the major peaks (t-RES, c-RES and fl-RES)
were obtained with a PDA detector placed in tandem with the
fluorescence detector, and are shown in Fig. 1.
appeared with an excitation peak at 315 nm, and a triphasic
emission curve, with peaks at 360, 380 and 405 nm, likely
corresponding to the three rings of the tricyclic Fl-RES,
which is the proposed structure of this photoderivative (42).
Longer duration light exposures increased the yield of Fl-RES
(Fig. 2).
RESULTS
RES and its photoderivatives
The ability of UVA and short wavelength VIS light to
photoderivatize RES was confirmed. Excitation–emission
scans were made to characterize the fluorescent properties
of the derivatives (Fig. 2). Following exposure to the Hg
vapor arc lamp, a fluorescent derivative, designated fl-RES,
These photoproducts were separated using HPLC (Fig. 3).
After 20 min of exposure to the Hg arc lamp (19 mW cm)2),
a 10 lg mL)1 solution of RES (100% in the form of t-RES)
was reduced to 0.85 lg mL)1 t-RES (>90% loss), due to the
conversion to c-RES (peak with RT ꢀ 14.5 min, upper trace

Photochemistry and Photobiology
5
in Fig. 3) and fl-RES (peak with RT ꢀ 17 min, upper trace in
Fig. 3). Although t-RES and c-RES are somewhat fluores-
cent, the fl-RES derivative is highly fluorescent (the peak
shown in the figure could not be precisely quantitated, due to
lack of an appropriate standard). The fl-RES prepared in this
way was used in the subsequent studies of its biological
activity.
oxidative stress induced by broadband light exposure (condi-
tions given in the Materials and Methods section). Results of
these experiments indicated that both RES and UA treatment,
at a concentration of 1–2 lM, significantly reduced at 24 h
postexposure, but not at 1 h, the levels of activated NF-jB
(P < 0.05 for RES effect and P < 0.002 for the UA effect),
compared with untreated cells (Fig. 4). We speculate that the
RES effect was not observed at 1 h postexposure because the
p65 was derived at that early time point from cytosolic stores,
whereas at the 24 h time point newly synthesized protein was
being utilized, which may have been more sensitive to
upstream effects of RES treatment on cell-signaling pathways.
Modulation of photooxidative stress responses in RPE cells by
RES and UA
The bioeffects of RES isomers and UA were investigated in the
hTERT-RPE cell line, using NF-jB as a biomarker of
Bioeffects of Fl-RES
Considering the strong fluorescence of fl-RES, the possibility
existed that this photoderivative of RES could function as a
photosensitizer. To address this question, cultures of hTERT-
RPE cells were pretreated with RES (10 lM) or fl-RES
(concentration not quantified, but the material was prepared
as described in the Materials and Methods section) for 1 h and
then exposed to light from the Hg arc lamp for 10 min at an
irradiance of 19 mW cm)1. After the light exposure, the cells
were incubated for 24 h, harvested, and the relative levels of
protein carbonyl adducts were determined in the samples using
the Oxyblot assay. Broadband light exposure increased the
overall protein carbonyl content of virtually all RPE cell
proteins by 15.4% (Table 1 and Fig. 5A). RES pretreatment
lowered the light-induced carbonylation of most RPE cell
proteins; in fact, the overall carbonyl load in light-exposed
RPE cells, pretreated with RES, was reduced by 9.7%
compared with the nonexposed cells and was about 25%
lower than in the light-exposed cells (Fig. 5A and Table 1). In
contrast, fl-RES appeared to have a pronounced photosensi-
tizing effect, increasing light-induced protein oxidation in the
RPE cells by 70.5% compared with the nonexposed control
cells (Fig. 5B). These results are summarized in Table 1.
Differential effects of UA on photooxidative stress responses in
normal and tumor cells
The results described above, showing UA modulation of
NF-jB activation, following photooxidative insult, suggested
that this phytochemical may have broader effects, such as on
inflammatory responses and cell cycle propagation. In addition
to the hTERT-RPE cell line, UA effects were also investigated
in an SM cell line. After the administration of a subcytotoxic
dose of UA (<10 lM), the cellular content of ROS, as
measured by DCFH oxidation, was unexpectedly increased in
both RPE and SM cell lines. The elevated DCFH oxidation,
however, was more marked in RPE cells than in the SM cells
(Fig. 6). Nevertheless, the UA-induced elevation in ROS did
not reduce cell viability as assessed by the MTT assay (data not
shown). Although the variability in the data in these measure-
ments was rather high, the overall trend indicated that, when
combined with UV irradiation, the UA-induced ROS was
reduced in RPE cells but increased in SM cells (Fig. 6,
compare ‘‘UA+L’’ with ‘‘L’’). This finding suggests that UA
pretreatment specifically potentiates the radiation effect in SM
cells leading to a disparity of cellular ROS between RPE and
SM cells, at least in the early responses to the radiation. This
Figure 4. Broadband light exposure causes activation of NF-jB in
RPE cells. The phosphorylation of p65 was measured in the cells using
ELISA. (A) Treatment of the RPE cells, starting at 1 h before the light
exposure, with t-RES at 2 lM reduced NF-jB activation in these cells
at 24 h, but not at 1 h, after exposure to the Hg vapor arc lamp for
7.5 min at an irradiance of 19 mW cm)2, compared with untreated
cells. The difference in the NF-jB activation in treated and untreated
RPE cells at 24 h postexposure was significant (n = 4; P < .05,
Bonferroni multiple comparison test). (B) Treatment of the RPE cells
with UA reduced NF-jB activation in light-exposed RPE cells. Arc
lamp exposure conditions were as described above. Cells were treated,
starting at 1 h before light exposure, with 2 lM UA, and p65-P was
measured in the cells a 1 and 24 h postexposure. UA treatment did not
reduce NF-jB activation at 1 h; in fact it was significantly elevated,
but a significant reduction in the phosphorylated transcription factor
was found at 24 h, compared with the untreated cells (n = 4;
P ꢁ 0.002, Bonferroni multiple comparison).

6
Yuan-Hao Lee et al.
Table 1. Integrated values of gel density scans from the Oxyblot assay of protein carbonyl adducts in RPE cells treated with t-RES or fl-RES and
.
then exposed to the broadband output of the Hg vapor arc lamp for 10 min at an irradiance of 19 mW cm)2
Control
(no light or pretreatment)
% Difference from
control
% Difference from
control
Pretreatment
Light only
Light + pretreatment
t-RES
fl-RES
233.3 21.1
278.0 11.6
272.2 4.0
407.1 27.1
15.4
37.7
211.8 12.8
580.7 57.6
)9.7
70.5
Values represent the means and ranges from two assays. Note that these values are integrals and thus are unitless. Percent differences between
values are calculated as [V1 ) V2 ⁄ ((V1 + V2) ⁄ 2)] · 100, where Vn represents an experimental value.
hypothesis received some support by the observation that the
radiation killing effect on SM cells was preferentially enhanced
by UA pretreatment at 24 h post exposure to UV–VIS light
(Fig. 7). These results imply that UA exerts distinct actions in
healthy and tumor cells.
shown in the apoptosis assay, the distinct regulatory effects of
UA on p53 protein expression may contribute to the relative
photoprotection of normal cells, while enhancing the cytotoxic
effect of the UV–VIS radiation on tumor cells.
The time course of apoptosis following UA treatment and
UV exposure was evaluated in SM cells using the Vybrant
Live ⁄ Dead assay. Cells were observed microscopically and
live, necrotic and apoptotic cells were counted in each
microscopic field. The results of these studies indicated that
radiation-induced cell death was facilitated by UA pretreat-
ment, particularly in the SM cells, as shown by the drop in the
number of SM cells per microscope filed as early as 8 h
postirradiation (Fig. 8). The proportion of cells dying through
apoptosis increased in the UA-plus-radiation treated SM cells
throughout the 24 h monitoring period (Fig. 9). This result
suggests differential modulatory functions of UA on cell death
and apoptosis in RPE and SM cells. Studying cell cycle arrest
with BrdU staining indicated that UA blocks the cell cycle
rather than directly initiating apoptosis in RPE cells (data not
shown, but confirmed by flow cytometry, which showed that
UA treatment causes a majority of RPE cells to remain in G1
phase). This finding was in contrast to the effect of UA on SM
cells, in which the treatment of 1 lg mL)1 (2.2 lM) UA
reduced DNA synthetic activity to about 50% of the basal
rate. This finding indicates the potential involvement of p53,
which dominates both cell cycle regulation and apoptosis, in
mediating the mechanism of action of UA. This hypothesis
was supported by the following observation that UA increased
the expression of p53 in SM cells.
By using anti-p53 and anti-p65-P antibodies in both
immunoprecipitation and immunoblotting studies, the specific
action of UA on p53 regulation and NF-jB activation was
evaluated following UV–VIS irradiation (Fig. 10). The level of
p65-P was decreased by UA treatment in SM cells (compare, in
Fig. 10A, the ‘‘UA’’ treatment group with ‘‘Control’’ for SM
cells) whereas the level of p53 was increased in the total protein
fraction (Fig. 10B, compare the ‘‘Control’’ vs ‘‘UA’’ groups).
In contrast, UA treatment had little effect on the basal level of
p65 phosphorylation (Fig. 10A) and p53 expression in RPE
cells (Fig. 10B). With respect to the cellular response to
UV–VIS radiation, pretreatment with UA restored the activa-
tion of NF-jB inhibited by UV radiation in SM cells,
rendering it similar to the NF-jB response in the normal,
RPE cells (Fig. 10A, compare the ‘‘UA-L’’ and ‘‘L’’ groups),
and reduced the radiation-induced p53 upregulation in RPE
cells while maintaining a strong p53 response in SM cells.
These results reveal the ability of UA to modulate cellular
signaling in response to radiation insult. Along with the result
DISCUSSION
Differential effects of RES and UA in normal and malignant
cells exposed to photooxidative stress
Although in the adult, phakic eye, retina and RPE cells are not
normally exposed to UV light, they are irradiated by wave-
lengths as short as 360 nm, which transmitted through the
ocular cornea and lens are capable of inducing photochemical
damage. This topic has been recently reviewed (47). Moreover,
in the juvenile eye, before the physiological lens has begun its
yellowing process, the retina and RPE are subject to UVB light
damage (48). Therefore, although normally situated in very
different bodily locations, both RPE and skin cells—at some
point in or throughout their life—receive UV and VIS
irradiation, and it is reasonable to compare the photodamage
mechanisms in these different cell types. A major observation
in the present work is that two phytochemicals, RES and UA,
exert general antioxidant effects in cells subjected to photoox-
idative stress, i.e. exposure to broadband light containing
radiation in the UVB, UVA and VIS bands. In the case of
RES, this is not a surprising finding, because of the well-
documented antioxidant properties of the compound. The pro-
oxidant property of fl-RES, the photoderivative of RES, is a
new finding, and is discussed in detail below. UA, based on its
structure, does not possess any obvious substituent capable of
supporting redox reactions, and thus its effects may be the
result of interactions with specific cell-signaling pathways.
A surprising observation made in the present study is that UA
may have different actions in normal and malignant cells. In
the nonmalignant RPE cells, UA increased the content of
ROS, as measured by DCFH oxidation, whereas this effect
was less apparent in the SM cancer cells (Fig. 6). In the present
data set, the effect of UA on the cellular oxidative state is not
entirely established; however, the trend in the data suggests
that there is a difference. The import of such a difference in the
intracellular redox state is not entirely clear at this time; we
found that the RPE cells were not adversely affected by the
increased ROS content, as evidenced by their continued
growth and viability. On the other hand, UA treatment
markedly sensitized the SM cells to the UV–VIS exposure; the
early cell mortality was greatly increased in the SM cells
treated with UA (Fig. 8). The response of the RPE cells to the
UV–VIS irradiation was not noticeably affected by UA
treatment. This differential effect of UA appears to be due

Photochemistry and Photobiology
7
A
Figure 6. Reactive oxygen species in RPE and SM cells treated with
UA. HPLC analysis of cells probed with the DCFH fluorescent
indicator was used to assess oxidative stress in the cells under
the various experimental conditions. The ordinate of the graph shows
the proportion of DCFH oxidized in each group (n = 3 for RPE cells,
and n = 2 for SM cells), which was determined by first measuring with
HPLC the amount of oxidized DCFH in the treated cells, and then
determining the total cellular content of DCFH in an aliquot of the cell
lysate after chemically oxidizing all of the DCFH present, using horse-
radish peroxidase. Bars show the means and the error bars represent
the ranges of experimentally determined values. Description of
experimental conditions: Control = cells without any treatment;
UA = cells treated with 2.2 lM UA; UA+L = cells treated with
2.2 lM UA and exposed to broadband light; L = cells treated with
broadband light. In the groups exposed to light, the cells were
irradiated at 26.7 mW cm)2 for 10 min.
B
the sirtuins), whereas UA appears to block cell cycle propa-
gation, increase apoptosis and modulate—through its interac-
tions with NF-jB signaling—some aspects of the cellular
response to oxidative stress. Thus, the selective use of these
phytochemicals may be useful in a variety of therapeutic
applications.
Figure 5. Protein carbonylation in RPE cells exposed to the output of
the Hg vapor arc lamp (10 min at 19 mW cm)2). RES treatment
decreased the number of carbonyl adducts below that found in the
control (untreated and unexposed) cells, whereas fl-RES increased
oxidation of virtually all of the proteins resolved in the gel. The figures
show optical density scans of the images of Western blots resulting
from the Oxyblot assay of protein carbonyl adducts in the cells
following a 24 h incubation at 37ꢁC after the light exposure (further
details are given in the Materials and Methods section). The area under
these traces was integrated and used as a quantitative measure to
compare the carbonyl loads in each treatment group (see Table 1 for a
summary of these data). (A) Effect of RES pretreatment. Explanation
of traces (top to bottom): ‘‘Light exposed, no RES’’ shows the level of
protein carbonyls found in proteins in RPE cells that were exposed to
light without RES pretreatment; ‘‘No light, no RES’’ is the protein
carbonyl level in control cells neither light-exposed nor treated with
RES; and ‘‘Light, +10 lM RES pretreatment’’ shows the protein
carbonyl level in cells exposed to light after pretreatment with 10 lM
RES. (B) Effect of fl-RES pretreatment (fl-RES concentration not
determined, see text). Explanation of traces (top to bottom): ‘‘Light
exposed, +flRES’’ shows protein carbonylation in cells pretreated
with the fl-RES photoderivative and then exposed to light; ‘‘Light
exposed, No flRES’’ shows protein carbonylation in cells exposed to
light but not treated with fl-RES; and ‘‘No light (control)’’ shows
protein carbonylation in control cells that did not receive light
exposure or treatment with fl-RES.
Antioxidant effect of RES
Native RES, i.e. t-RES, at a low micromolar concentration,
exerts a variety of antioxidant actions. This effect is typified by
the reduction of protein carbonyls produced by photochemical
damage (cf. Fig. 5A). The other salient finding is that the
activation of NF-jB is reduced by RES pretreatment, suggesting
a possible anti-inflammatory action of this phytochemical in
ocular tissue. RES has been reported to exert anti-inflammatory
effects in other tissue systems. For example, RES inhibits COX-
2, a member of the family of cyclooxygenase enzymes, which
plays a critical role in mediating inflammatory responses. RES
has demonstrated anti-inflammatory activity in a number of
model systems, such as LPS or other chemical and toxin-induced
inflammatory responses (49–51), and ischemia-reperfusion (17).
RES has also been found to modulate or reduce cell-signaling
pathways, such as NF-jB and AP-1, both of which are involved
in mediating inflammatory responses (52).
to the upregulation of p53 in SM cells (Fig. 10B), which blocks
cell cycling and proliferation, leading to an increase in
apoptosis in SM cells treated with UA and exposed to
UV–VIS. Thus, the effects of RES and UA are somewhat
complementary: RES exerts general antioxidant effects (but
see the next section for a discussion of its other actions, e.g. on
RES, and other plant polyphenols, may also confer protec-
tion through an entirely different pathway, i.e. by enhancing
the activity of the SIRT1 or other sirtuin genes. This gene
family increases gene stability by deacetylating histones and
many other proteins in a variety of cell-signaling pathways,

8
Yuan-Hao Lee et al.
Figure 7. Morphological comparison of skin melanoma (SM) cells and RPE cells, 24 h after treatment with UA and UV–VIS radiation. Cells were
treated with ursolic acid (2.2 lM) for 24 h prior to UV–VIS irradiation. After 24 h or postexposure incubation, the cells were imaged. The
photomicrographs in the upper and the lower rows represent the morphological appearance of SM and RPE cells under four experimental
conditions, respectively. Description of experimental conditions from left to right: Control = cells without any treatment; UA Alone = cells
treated with UA only; UA + Light = cells treated with UA and broadband light exposure; Light Alone = cells exposed to broadband light
without UA treatment. The cells were irradiated at an intensity of 26.7 mW cm)2 for 10 min.
Figure 8. Cell proliferation in SM cells treated with UA and UV
radiation. Cell growth was assessed by the average number of cells
counted in three independent microscope fields over the designated
time. The cells were imaged using phase contrast microscopy and the
number of cells per field was counted manually. Legend: Con-
trol = cells without any treatment; UA = cells treated with 2.2 lM
UA only; UA+L = cells treated with 2.2 lM UA and UV–VIS light
exposure; L = cells exposed to UV–VIS light only. Light exposure
conditions were as described in Fig. 7.
Figure 9. Proportion of apoptotic cells in each treatment group over
time, as determined with the Vybrant YO-PRO-1 assay. This assay was
performed using the Vybrant 4 live ⁄ dead cell assay, 4 h following the
treatment of SM cells with UA (2.2 lM) and ⁄ or light exposure. Dead
cells (i.e. necrotic) are initially permeable to the PI fluorescent probe
only, whereas apoptotic cells only admit YO-PRO into the nuclei.
Apoptosis in the figure is expressed mathematically by the apoptosis
index (AI) as: AI ¼ ½ðh ꢂ YÞ=ðh ꢂ PÞꢃ ꢂ 1, where h is the total number
of cells, Y denotes the number of YO-PRO stained cells and P denotes
the number of PI stained cells in the same field. Error bars in the figure
represent the standard deviation values for each condition (n = 3).
Legend: Control = cells without any treatment; UA = cells treated
with 2.2 lM UA only; UA+L = cells treated with 2.2 lM UA and
UV–VIS light exposure; L = cells exposed to UV–VIS light only.
Light exposure conditions were as described in Fig. 7.
one result of which is increased lifespan (53). The action of
plant polyphenols, such as RES, increases the affinity of
SIRT1 for its substrates (53). In terms of ocular protection,
activation of SIRT1 by intraocular injection of RES or
synthetic pharmaceutical agents reduced the apoptotic loss of
retinal cells in animal models of experimental uveitis (54,55).
Other signaling pathways are beneficially modulated by RES,
including signaling mediated by proinflammatory cytokines in
cells of the trabecular meshwork in a glaucoma model (56),
and CaM kinase II in retinal cells in a diabetes model (57).
Thus, the ability to regulate signaling pathways is likely to be
critical in the ability of this phytochemical to regulate
inflammatory and immune responses in vivo.
oxidative damage to proteins (Fig. 5), is of considerable
relevance to the safe and effective use of RES as a nutritional
supplement. It is well known that photodegradation of RES
readily occurs upon exposure to natural or artificial light
sources, e.g. red wine is packaged in green or brown glass
bottles to shield out the shorter wavelengths that would
degrade the quality of the product; however, if a sufficient
quantity of the fl-RES end product is produced, then a
potential health hazard may exist. This possibility must also be
considered in the design of animal nutritional supplementation
experiments, during which animal chow containing RES might
be exposed for considerable lengths of time to standard
laboratory fluorescent lighting, resulting in the conversion of
Photosensitization by RES photoderivatives
The finding in this study that the RES photoderivative, fl-RES,
can function as a photosensitizer, at least in promoting

Photochemistry and Photobiology
9
measured by the extent of DCFH oxidation. This effect
suggests that UA may be able to modify the underlying cellular
metabolism, which is a potentially useful strategy for cancer
therapy (58). Notably, SM cells maintained better homeostasis
when confronted with elevated oxidative stress than did RPE
cells upon radiation exposure, based on the result of the
intracellular DCFH assay. The capacity of cells to withstand
oxidative stress, however, was reported to be restored by UA,
albeit at a higher concentration (IC50 = 140.76 lg mL)1) (23)
than was used in the present investigation. The importance of
the UA-induced increase in ROS may be that it triggers
markedly different intracellular signaling patterns found
between normal and cancer cells following UV radiation.
Upon treatment with UA at micromolar concentrations, the
resulting levels of p53 expression and NF-jB activation were
very different in the RPE and SM cell lines, strongly
implicating a specific modulatory effect of UA. The stimula-
tory effect on p53 expression and the inhibitory effect on
NF-jB activation exerted by UA preferentially occurred in SM
cells. The differential activation of these two signaling path-
ways, combined with the observed radiation killing effect
through apoptosis, supports the conclusion that UA sensitizes
SM cells to UV radiation while protecting RPE cells from
radiation damage. Preliminary data suggest that similar
actions of UA are also found in these cells after exposure to
gamma radiation. Considering its low toxicity to normal cells,
and common occurrence in the human diet, UA may be a
promising candidate for inclusion in clinical radiotherapeutic
protocols to increase the radiosensitivity of tumor cells while
simultaneously providing some degree of radioprotection to
normal cells.
Figure 10. Protein expression in RPE and SM cells after UA treatment
and UV–VIS exposure. Expression of NF-jB (p65-P) and p53 were
determined by immunoprecipitation and Western blot analysis. (A)
Total p65-P in response to the treatment of RPE and SM cells with UA
and ⁄ or UV–VIS light. (B) Nuclear p53 upregulation in response to the
treatment of UA and ⁄ or UV–VIS light. Bars indicate the mean values,
and the error bars represent the ranges of experimentally determined
values in the groups between each run (n = 2 for all measurements).
Description of experimental conditions: Control = cells without any
treatment; UA = cells treated with 2.2 lM UA only; UA+L = cells
treated with 2.2 lM UA and UV–VIS light; L = cells exposed to
UV–VIS light only. Light exposure conditions were as described in
Fig. 7.
Practical considerations in the use of nutraceuticals to achieve
therapeutic endpoints
In our studies, as well as most of those in the literature, the
observed therapeutic endpoints have been achieved at concen-
trations of the agent in the range of 1–10 lM. The bioavail-
ability of these agents in vivo, however, is generally limited, and
typically much lower systemic levels are achieved by oral
intake. For example, the serum levels of RES in mice required
to obtain some of the pharmacological effects reported have
been as high as 1500 mg kg)1 (9)—practically impossible to
achieve by feeding alone. In humans, the safe intakes of RES
from sources such as wine are limited, so that even including
the circulating amounts of RES metabolites, an upper effective
concentration of about 2 lM might be achieved by dietary
sources (9) (although the sulphonated and glucuronide metab-
olites probably have little bioactivity). Similarly, the few
pharmacokinetic investigations of UA have found low sys-
temic uptake. For example, in rats after gavage feeding of large
amounts of the compound, serum UA reached a peak
concentration of approximately 300 ng mL)1 (0.66 lM) (59).
There is no information on the ocular pharmacokinetics of
these compounds, but it is likely that much lower levels reach
the retina and RPE from the serum. We do not yet have
information on the cellular uptake or retention of these
compounds. Therefore, the obstacle of low ocular bioavail-
ability of RES and UA, along with limited pharmacokinetic
data, must be overcome before these compounds can be used
to achieve predictable therapeutic effects in the eye. This may
some of the compound to fl-RES. Even if the amount of
fl-RES produced is insufficient by itself to produce photosen-
sitized reactions in the animals, the loss of RES due to
photoderivatization will result in a lower RES intake for the
animals. Fortunately, the pellet matrix of some laboratory
animal chows may shield the RES from excessive light
exposure; our group has found that RES formulated in AIN-
93G mouse chow pellets was stable for at least 6 months
(unpublished data). Nevertheless, RES may not be stable in all
pellet formulations, and precautions should be taken to
prevent excessive exposure of RES-supplemented animal
chows to fluorescent or daylight sources.
Differential effects of UA in normal and tumor cells
Although the present data are inconclusive, UA treatment
tended to increase the level of intracellular oxidative stress as

10 Yuan-Hao Lee et al.
matory responses induced by ischemia ⁄ reperfusion, platelet-
activating factor, or oxidants. Free Radic. Biol. Med. 34, 810–817.
18. Losa, G. A. (2003) Resveratrol modulates apoptosis and oxidation
in human blood mononuclear cells. Eur. J. Clin. Invest. 33, 818–823.
19. Yen, G. C., P. D. Duh and C. W. Lin (2003) Effects of resveratrol
and 4-hexylresorcinol on hydrogen peroxide-induced oxidative
DNA damage in human lymphocytes. Free Radic. Res. 37, 509–514.
20. Ahmad, A., F. A. Syed, S. Singh and S. M. Hadi (2005) Prooxi-
dant activity of resveratrol in the presence of copper ions: Muta-
genicity in plasmid DNA. Toxicol. Lett. 159, 1–12.
21. Gatz, S. A. and L. Wiesmuller (2008) Take a break—Resveratrol
in action on DNA. Carcinogenesis 29, 321–332.
22. Roemer, K. and M. Mahyar-Roemer (2002) The basis for the
chemopreventive action of resveratrol. Drugs Today (Barc) 38,
571–580.
23. Ali, M. S., S. A. Ibrahim, S. Jalil and M. I. Choudhary (2007)
Ursolic acid: A potent inhibitor of superoxides produced in the
cellular system. Phytother. Res. 21, 558–561.
24. Najid, A., A. Simon, J. Cook, H. Chable-Rabinovitch, C. Delage,
A. J. Chulia and M. Rigaud (1992) Characterization of ursolic
acid as a lipoxygenase and cyclooxygenase inhibitor using mac-
rophages, platelets and differentiated HL60 leukemic cells. FEBS
Lett. 299, 213–217.
25. Yin, M. C. and K. C. Chan (2007) Nonenzymatic antioxidative
and antiglycative effects of oleanolic acid and ursolic acid.
J. Agric. Food Chem. 55, 7177–7181.
26. Balanehru, S. and B. Nagarajan (1991) Protective effect of ole-
anolic acid and ursolic acid against lipid peroxidation. Biochem.
Int. 24, 981–990.
be achieved by the synthesis of chemical derivatives with
superior systemic uptake and distribution, a route currently
being explored commercially. Until such agents are available,
it is likely that regular nutritional intake of these compounds
will be the most practical way to obtain whatever beneficial
effect on ocular health is conferred by the low systemic
concentrations achieved via dietary sources.
Acknowledgements—This work has been supported in part by the
RMG Research Endowment, a grant from the Julio Palmaz Endow-
ment in the Department of Radiology at the UTHSCSA (Y-HL) and a
grant from the University of Texas System ⁄ San Antonio Life Sciences
Institute (RDG). Y-HL was supported by National Science Founda-
tion Partnerships for Research and Education in Materials (PREM)
Grant No. DMR-0934218 in collaboration with Northwestern
University MRSEC.
REFERENCES
1. Hillenkamp, F. (1989) Laser radiation tissue interaction. Health
Phys. 56, 613–616.
2. Jacques, S. L. (1997) Laser-tissue interactions. In Non-Ionizing
Radiation: An Overview of the Physics and Biology (Edited by K.
Hardy, M. Meltz and R. D. Glickman), pp. 81–98. Medical
Physics Publishing, Madison,WI.
27. Shishodia, S., S. Majumdar, S. Banerjee and B. B. Aggarwal
(2003) Ursolic acid inhibits nuclear factor-kappaB activation
induced by carcinogenic agents through suppression of IkappaB-
alpha kinase and p65 phosphorylation: Correlation with
down-regulation of cyclooxygenase 2, matrix metalloproteinase 9,
and cyclin D1. Cancer Res. 63, 4375–4383.
28. Mohan, S., K. Koyoma, A. Thangasamy, H. Nakano, R. D.
Glickman and N. Mohan (2007) Low shear stress preferentially
enhances IKK activity through selective sources of ROS for per-
sistent activation of NF-kappaB in endothelial cells. Am. J.
Physiol. Cell Physiol. 292, C362–C371.
29. Pahl, H. L. (1999) Activators and target genes of Rel ⁄ NF-kappaB
transcription factors. Oncogene 18, 6853–6866.
30. Raju, U., G. J. Gumin and P. J. Tofilon (2000) Radiation-induced
transcription factor activation in the rat cerebral cortex. Int. J.
Radiat. Biol. 76, 1045–1053.
3. Glickman, R. D. (2002) Phototoxicity to the retina: Mechanisms
of damage. Int. J. Toxicol. 21, 473–490.
4. Sliney, D. H. (2005) Exposure geometry and spectral environment
determine photobiological effects on the human eye. Photochem.
Photobiol. 81, 483–489.
5. Taylor, H. R., B. Munos, S. West, N. M. Bressler, S. B. Bressler
and F. S. Rosenthal (1990) Visible light and risk of age-related
macular degeneration. Trans. Am. Ophthalmol. Soc. 88, 163–178.
6. Tomany, S. C., K. J. Cruickshanks, R. Klein, B. E. Klein and M.
D. Knudtson (2004) Sunlight and the 10-year incidence of
age-related maculopathy: The Beaver Dam Eye Study. Arch.
Ophthalmol. 122, 750–757.
7. Cabrera, M. P. and R. H. Chihuailaf (2011) Antioxidants and the
integrity of ocular tissues. Vet. Med. Int. 2011, 905153 (pp. 1–8).
8. Kelsey, N. A., H. M. Wilkins and D. A. Linseman (2010)
Nutraceutical antioxidants as novel neuroprotective agents.
Molecules 15, 7792–7814.
31. Vile, G. F., A. Tanew-Ilitschew and R. M. Tyrrell (1995) Activation
of NF-kappa B in human skin fibroblasts by the oxidative stress
generated by UVA radiation. Photochem. Photobiol. 62, 463–468.
32. Li, N. and M. Karin (1998) Ionizing radiation and short wave-
length UV activate NF-kappaB through two distinct mechanisms.
Proc. Natl Acad. Sci. USA 95, 13012–13017.
9. Baur, J. A. and D. A. Sinclair (2006) Therapeutic potential of res-
veratrol: The in vivo evidence. Nat. Rev. Drug Discov. 5, 493–506.
10. Das, D. K. and N. Maulik (2006) Resveratrol in cardioprotection:
A therapeutic promise of alternative medicine. Mol. Interv. 6, 36–
47.
33. Glickman, R. D., M. Natarajan, B. A. Rockwell, M. Denton, S.
Maswadi, N. Kumar and F. Nieves-Roldan (2005) Intracellular
signaling mechanisms responsive to laser-induced photochemical
and thermal stress. In Optical Interactions with Tissue and Cells
XVI (Edited by S. L. Jacques and W. P. Roach), Proc. SPIE 5695,
260–269.
34. Legrand-Poels, S., S. Schoonbroodt, J. Y. Matroule and J. Piette
(1998) NF-kappa B: An important transcription factor in photo-
biology. J. Photochem. Photobiol. B, Biol 45, 1–8.
35. Yim, E. K., K. H. Lee, S. E. Namkoong, S. J. Um and J. S. Park
(2006) Proteomic analysis of ursolic acid-induced apoptosis in
cervical carcinoma cells. Cancer Lett. 235, 209–220.
36. Yim, E. K., M. J. Lee, K. H. Lee, S. J. Um and J. S. Park (2006)
Antiproliferative and antiviral mechanisms of ursolic acid and
dexamethasone in cervical carcinoma cell lines. Int. J. Gynecol.
Cancer 16, 2023–2031.
11. Gehm, B. D., J. M. McAndrews, P. Y. Chien and J. L. Jameson
(1997) Resveratrol, a polyphenolic compound found in grapes and
wine, is an agonist for the estrogen receptor. Proc. Natl Acad. Sci.
USA 94, 14138–14143.
12. Klinge, C. M., K. E. Risinger, M. B. Watts, V. Beck, R. Eder and
A. Jungbauer (2003) Estrogenic activity in white and red wine
extracts. J. Agric. Food Chem. 51, 1850–1857.
13. Turner, R. T., G. L. Evans, M. Zhang, A. Maran and J. D. Sib-
onga (1999) Is resveratrol an estrogen agonist in growing rats?
Endocrinology 140, 50–54.
14. Basly, J. P., F. Marre-Fournier, J. C. Le Bail, G. Habrioux and A.
J Chulia (2000) Estrogenic ⁄ antiestrogenic and scavenging prop-
erties of (E)- and (Z)-resveratrol. Life Sci. 66, 769–777.
15. Sato, M., G. Maulik, D. Bagchi and D. K. Das (2000) Myocardial
protection by protykin, a novel extract of trans-resveratrol and
emodin. Free Radic. Res. 32, 135–144.
37. Huang, M. T., C. T. Ho, Z. Y. Wang, T. Ferraro, Y. R. Lou, K.
Stauber, W. Ma, C. Georgiadis, J. D. Laskin and A. H. Conney
(1994) Inhibition of skin tumorigenesis by rosemary and its con-
stituents carnosol and ursolic acid. Cancer Res. 54, 701–708.
38. Tokuda, H., H. Ohigashi, K. Koshimizu and Y. Ito (1986)
Inhibitory effects of ursolic and oleanolic acid on skin tumor
16. Sato, M., P. S. Ray, G. Maulik, N. Maulik, R. M. Engelman, A. A.
Bertelli, A. Bertelliand and D. K. Das (2000) Myocardial protection
with red wine extract. J. Cardiovasc. Pharmacol. 35, 263–268.
17. Shigematsu, S., S. Ishida, M. Hara, N. Takahashi, H. Yoshimatsu,
T. Sakata and R. J. Korthuis (2003) Resveratrol, a red wine
constituent polyphenol, prevents superoxide-dependent inflam-

Photochemistry and Photobiology 11
promotion by 12-O-tetradecanoylphorbol-13-acetate. Cancer Lett.
33, 279–285.
and crystalline lens photoprotection. Am. J. Ophthalmol. 149,
543–549.
39. De Angel, R. E., S. M. Smith, R. D. Glickman, S. N. Perkins and
S. D. Hursting (2010) Antitumor effects of ursolic acid in a mouse
model of postmenopausal breast cancer. Nutr. Cancer 62, 1074–
1086.
40. Trela, B. C. and A. L. Waterhouse (1996) Resveratrol: Isomeric
molar absorptivities and stability. J. Agric. Food Chem. 44, 1253–
1257.
41. Jayatilake, G. S., H. Jayasuriya, E. S. Lee, N. M. Koonchanok, R.
L. Geahlen, C. L. Ashendel, J. L. McLaughlin and C. J. Chang
(1993) Kinase inhibitors from Polygonum cuspidatum. J. Nat.
Prod. 56, 1805–1810.
49. Birrell, M. A., K. McCluskie, S. Wong, L. E. Donnelly, P. J.
Barnes and M. G. Belvisi (2005) Resveratrol, an extract of red
wine, inhibits lipopolysaccharide induced airway neutrophilia and
inflammatory mediators through an NF-kappaB-independent
mechanism. FASEB J. 19, 840–841.
50. Chen, G., W. Shan, Y. Wu, L. Ren, J. Dong and Z. Ji (2005)
Synthesis and anti-inflammatory activity of resveratrol analogs.
Chem. Pharm. Bull. (Tokyo) 53, 1587–1590.
51. Elmali, N., I. Esenkaya, A. Harma, K. Ertem, Y. Turkoz and B.
Mizrak (2005) Effect of resveratrol in experimental osteoarthritis
in rabbits. Inflamm. Res. 54, 158–162.
42. Galeano-Diaz, T., I. Duran-Meras and D. Airado-Rodriguez
(2007) Isocratic chromatography of resveratrol and piceid after
previous generation of fluorescent photoproducts: Wine analysis
without sample preparation. J. Sep. Sci. 30, 3110–3119.
43. Dontsov, A. E., R. D. Glickman and M. A. Ostrovsky (1999)
Retinal pigment epithelium pigment granules stimulate the photo-
oxidation of unsaturated fatty acids. Free Radic. Biol. Med. 26,
1436–1446.
44. Rambhatla, L., C. P. Chiu, R. D. Glickman and C. Rowe-Ren-
dleman (2002) In vitro differentiation capacity of telomerase
immortalized human RPE cells. Invest. Ophthalmol. Visual. Sci.
43, 1622–1630.
45. Natarajan, M., B. K. Nayak, C. Galindo, S. P. Mathur, F. N.
Roldan and M. L. Meltz (2006) Nuclear translocation and DNA-
binding activity of NFKB (NF-kappaB) after exposure of human
monocytes to pulsed ultra-wideband electromagnetic fields
(1 kV ⁄ cm) fails to transactivate kappaB-dependent gene expres-
sion. Radiat. Res. 165, 645–654.
46. Glickman, R. D., M. Vendal, M. A. Gonzales and N. Kumar
(1999) Intracellular photochemical reactions in the RPE cell
exhibit a wavelength dependence that resembles the action spectrum
of melanin. In Laser-Tissue Interaction X: Photochemical, Photo-
thermal, and Photomechanical (Edited by S. L. Jacques), Proc. SPIE
3601, 94–101.
52. Rahman, I., S. K. Biswas and P. A. Kirkham (2006) Regulation of
inflammation and redox signaling by dietary polyphenols.
Biochem. Pharmacol. 72, 1439–1452.
53. Porcu, M. and A. Chiarugi (2005) The emerging therapeutic
potential of sirtuin-interacting drugs: From cell death to lifespan
extension. Trends Pharmacol. Sci. 26, 94–103.
54. Anekonda, T. S. and G. Adamus (2008) Resveratrol prevents
antibody-induced apoptotic death of retinal cells through upreg-
ulation of Sirt1 and Ku70. BMC Res. Notes 1, 122.
55. Shindler, K. S., E. Ventura, T. S. Rex, P. Elliott and A. Rostami
(2007) SIRT1 activation confers neuroprotection in experimental
optic neuritis. Invest. Ophthalmol. Vis. Sci. 48, 3602–3609.
56. Luna, C., G. Li, P. B. Liton, J. Qiu, D. L. Epstein, P. Challa and
P. Gonzalez (2009) Resveratrol prevents the expression of glau-
coma markers induced by chronic oxidative stress in trabecular
meshwork cells. Food Chem. Toxicol. 47, 198–204.
57. Kim, Y. H., Y. S. Kim, S. S. Kang, G. J. Cho and W. S. Choi
(2010) Resveratrol inhibits neuronal apoptosis and elevated
Ca2+ ⁄ calmodulin-dependent protein kinase II activity in diabetic
mouse retina. Diabetes 59, 1825–1835.
58. Baffy, G., Z. Derdak and S. C. Robson (2011) Mitochondrial
recoupling: A novel therapeutic strategy for cancer? Br. J. Cancer
105, 469–474.
59. Liao, Q., W. Yang, Y. Jia, X. Chen, Q. Gao and K. Bi (2005)
LC-MS determination and pharmacokinetic studies of ursolic acid
in rat plasma after administration of the traditional Chinese
medicinal preparation Lu-Ying extract. Yakugaku Zasshi 125,
509–515.
47. Glickman, R. D. (2011) Ultraviolet phototoxicity to the retina.
Eye Contact Lens 37, 196–205.
48. Mainster, M. A. and P. L. Turner (2010) Ultraviolet-B photo-
toxicity and hypothetical photomelanomagenesis: Intraocular