Screening for Ocular Phototoxicity
Joan E. Roberts
Au: Where is opening parenth?Q2:
Au: Where is opening parenth?Q3:
Au: Roberts et al. 1992a Ok as edited?Q4:
Au: Rodgers 1985. Publisher is North Holland or Elsevier? In Amsterdam?
Screening for Ocular Phototoxicity
Joan E. Roberts
Department of Natural Sciences, Fordham University, New York, New York, USA
radiation. This light serves to direct vision and circadian rhythm
Normally light transmission through the eye is benign and serves
(Roberts 2000a) and, therefore, under normal circumstances
to direct vision and circadian rhythm. However, with very intense
must be benign. However, chronic exposure to intense light
light exposure, or with ambient light exposure to the aged eye
and/or age-related changes can lead to light-induced damage
and/or young or adult eye in the presence of light-activated (photo-
to the eye. Roberts 2001a; Zigman 1993; Dillon 1991; Andley
sensitizing) drugs or dietary supplements, cosmetics, or diagnostic
dyes, light can be hazardous, leading to blinding disorders. Light
damage to the human eye is avoided because the eye is protected by
There are many drugs, dietary supplements, and diagnostic
a very efficient antioxidant system and the chromophores present
dyes, both topical and systemic, that absorb in the ultraviolet
absorb light and dissipate its energy. After middle age, there is a de-
(UV) or visible range and can be excited by wavelengths of
crease in the production of antioxidants and antioxidant enzymes
light transmitted to the lens and retina. This absorption can lead
and an accumulation of endogenous chromophores that are photo-
toxic. The extent to which a particular exogenous photosensitizing
to dramatically enhanced ocular damage through the phototoxic
substance is capable of producing phototoxic side effects in the eye
side effects of those dyes and drugs (Fraunfelder 1982; Roberts
depends on several parameters, including (1) the chemical struc-
1996). The extent to which a particular photosensitizing drug
ture; (2) the absorption spectra of the drug; (3) binding of the drug
will affect the human lens or retina in vivo depends upon (1) its
to ocular tissue (lens proteins, melanin, DNA); and (4) the abil-
residence time in the lens or retina, which is determined by
ity to cross blood-ocular barriers (amphiphilic or lipophilic). For
instance, compounds that have either a tricyclic, heterocyclic, or
its structure; (2) the photoefficiency of the particular sensitizer,
porphyrin ring structure and are incorporated into ocular tissues
due to its ability to absorb the appropriate wavelengths of light
are potentially phototoxic agents in the eye. The extent to which
that have been transmitted to the lenticular or retinal substrates;
these substances might damage the eye (photoefficiency) can be
(3) the mechanism by which it causes that damage, including
predicted using in vitro and photophysical techniques. With sim-
possible binding of the dye to ocular constituents. Binding can
ple, inexpensive testing, compounds can be screened for their po-
tential ocular phototoxicity at the developmental stage. It may be
alter the photochemical mechanism and also increase the re-
that a portion of the molecule can be modified to reduce phototoxi-
tention time of the sensitizer in the eye); and (4) the presence
city while leaving the primary drug effect intact. Preclinical safety
of endogenous quenchers or free-radical scavengers that could
testing may prevent ocular side effects that can range from mild,
stop or retard these photochemical reactions. All of these pa-
reversible blurred vision to permanent blindness.
rameters play a role in the photosensitized oxidation of humanocular tissue and its importance as an underlying mechanism in
cataractogenesis and retinal damage.
Toxicology, Photosensitization, Phototoxic Drugs
Presented here are several screens that can be used to de-
termine the potential of drugs to cause light damage to the
The eye is the organ that is most vulnerable, after the skin,
eye. These include a simple screen that takes into account the
to light damage because it is constantly subjected to ambient
optical properties of the eye and the structure and absorptionspectra of the various drugs that can be used to eliminatevarious drugs as potential photo-oxidants in the eye. In addi-
Received 20 March 2002; accepted 19 June 2002.
tion, a second, more detailed screen is presented that can be
This article is based on a presentation at the symposium entitled
"Ocular Phototoxicity," held during the 22nd Annual Conference of the
used to determine (a) the quantitative potential for a drug to
American College of Toxicology, November 4–8, 2001, in Washington,
cause photo-oxidative damage in the eye, (b) the mechanism by
which it occurs, and (c) the in vivo verification of phototoxicity.
The author wishes to thank the Hugoton Foundation for its finan-
cial support and Dr. Ann Motten, NIEHS, North Carolina, for help inpreparing this manuscript.
Structure of the Eye
Address correspondence to Joan E. Roberts, PhD, Professor of
Chemistry, Fordham University, 113 West 60th Street, New York, NY
The human eye is composed of several layers (Figure 1).
10003, USA. E-mail: [email protected]
The outermost layer contains the sclera, whose function is to
International Journal of Toxicology, 21:1–10, 2002Copyright c
2002 American College of Toxicology
1091-5818/02 $12.00 + .00DOI:
The structure of the human eye.
protect the eyeball, and the cornea, which focuses incoming
diation from the sun can contain varying amounts of UV-C (100
light onto the lens. Beneath this layer is the choroid contain-
to 280 nm), UV-B (280 to 320 nm), UV-A (320 to 400 nm), and
ing the iris, which is known as the uvea. This region contains
visible light (400 to 760 nm). Most UV-C and some UV-B are fil-
melanocytes containing the pigment melanin, whose function
tered by the ozone layer. UV light contains shorter wavelengths
is to prevent light scattering. The opening in the iris, the pupil,
of light than visible; the shorter the wavelength, the greater the
expands and contracts to control the amount of incoming light.
energy and the greater the potential for biological damage.
The iris and the lens are bathed in the aqueous humor, a fluid
The primate/human eye has unique filtering characteristics
that maintains intraocular pressure; this fluid also contains var-
(Bachem 1956). The human cornea cuts off all light below
ious antioxidants and supports transport to the lens. The lens is
295 nm, so that all UV-C and some UV-B are filtered from
positioned behind the iris; its function is to focus light onto the
reaching the lens. The adult human lens absorbs the remaining
UV-B and all UV-A (295 to 400 nm). Therefore, only visible
Behind the lens is the vitreous humor, a fluid that supports
light reaches the adult human retina. However, the young hu-
the lens and retina and that also contains antioxidants. The retina
man lens transmits a small window of UV-B light (320 nm)
is composed of the photoreceptor cells (rods and cones) that re-
to the retina and the elderly lens filters out much of the short
ceive light and the neural portion (ganglion, amacrine, horizon-
blue visible light (Barker, Brainard, and Dayhaw-Barker 1991).
tal and bipolar cells) that transduces light signals through the
Aphakia (removal of the lens) and certain forms of blindness
retina to the optic nerve. Behind the photoreceptor cells are the
may also change the wavelength characteristics of light reach-
retinal pigment epithelial cells, Bruch's membrane, and the pos-
ing the retina.
terior choroid. The photoreceptor cells are avascular and theirnutrient support (ions, fluid, and metabolites) is provided by theretinal pigment epithelial cells. There is transport to the retinal
Direct Light Damage in the Eye
pigment epithelial cells across the Bruch's membrane by the
Short UV light exposure to the cornea leads to an inflamma-
tion reaction (Offord et al. 1999; Pitts, Cullen, and Parr 1976).
This is very painful and similar to a sunburn and is known as
keratitis. However, these corneal wounds usually heal and do
Light Absorption by the Human Eye
not cause permanent damage. On the other hand, UV damage
For light to damage the eye, it must be absorbed. Each wave-
to the lens and visible light damage to the retina are painless,
length of light will affect different areas of the eye. Ambient ra-
accumulative, and permanent. It can lead to the formation of
cataracts (clouding of the lens) (Dillon 1991; Balasubramanian
Drug-Induced Ocular Phototoxicity
2000), which can only be corrected by surgery, and to macular
Most of the damage to the eye caused by direct irradiation
and retinal degeneration (Ham et al. 1982), leading to permanent
from the sun or artificial sources is from UV radiation. However,
blindness for which there is no treatment.
in the presence of a light-activated (photosensitized) diagnostic
A number of factors suppress or enhance light damage to the
dye or drug, patients are in danger of enhanced ocular injury
eye, including oxygen, antioxidants, repair mechanisms, and
from both UV and visible light. The extent to which a particular
photosensitizer will affect the human cornea, lens, and/or retinain vivo depends upon the following factors: (1) Its residence
Oxygen Tension in the Eye
time in the eye, which is determined by its structure and binding
The cornea is highly oxygenated. The retina is supplied by
of the dye to ocular constituents. Binding can not only alter the
the blood so it has varying but high oxygen content in different
photochemical mechanism but also can increase the retention
portions of the retinal tissues. The aqueous humor and the lens
time of the sensitizer in the eye. (2) The photoefficiency of the
have low oxygen content but it is sufficient for photo-oxidation
particular sensitizer (i.e., its ability to absorb the appropriate
to occur. (McLaren et al. 1998; Roberts et al. 1992a; Kwan,
wavelengths of light that have been transmitted to the lenticular
Niinikoske, and Hunt 1971). The higher the oxygen content of
or retinal substrates). These transmission characteristics change
the tissue, the easier it is for light damage to occur.
with age. (3) The oxygen content of the ocular tissue. (4) Thepresence of endogenous quenchers or free-radical scavengers
and enzyme detoxification systems that could stop or retard these
Because the eye is constantly subjected to ambient radia-
tion, each portion of the eye contains very efficient defensesystems. There are antioxidant enzymes (superoxide dismu-tase [SOD] and catalase) and antioxidants (e.g., vitamins E and
PREDICTING OCULAR PHOTOTOXICITY
C, lutein, zeaxanthin, lycopene, glutathione, and melanin) that
Based on the theoretical considerations stated above, it is
serve to protect against oxidative and photo-induced damage
relatively easy to predict that a drug cannot cause ocular dam-
(Handelman and Dratz 1986; Giblin 2000; Seth and Kharb 1999;
age through a photo-induced event. The short screen given in
Edge et al. 1998; Khachik, Bernstein, and Garland 1997). Un-
Table 1 will dramatically reduce the number of potential sub-
fortunately, most of these antioxidants and protective enzymes
stances needed to be considered for ocular phototoxicity.
decrease beginning at 40 years of age (Sarna 1992; Khachik,Bernstein, and Garland 1997; Samiec et al. 1998; Sethna,Holleschau, and Rathbun 1983). With the protective systems
Short Screen for Potential Ocular Phototoxicity
diminish with age, there is loss of protection against all light-
Measure the Absorbance Spectrum
induced damage to the eye and the induction of age related
In order for a chemical compound (diagnostic dye, drug, en-
dogenous sensitizer) to induce a phototoxic response in anybiological tissue, it must first absorb light. This absorption is
limited by the filtering characteristics of the biological tissues
Even if the eye is damaged, the damage does not have to
involved. A comparison of the transmission characteristics of
be permanent. The cornea and retina have very efficient repair
the eye with the absorbance spectrum of the drug may be used
systems. However, damage to the lens is cumulative and not
as a quick screen for phototoxicity. To have the potential to dam-
repairable (Andley 1987).
age the aqueous or the lens, a drug needs a UV spectrum that
consists of absorption wavelengths longer than 295 nm (Bachem
1956). This includes drugs such as chlorpromazine (maximum:
Ocular damage from light can occur through either an inflam-
310 nm), tetracycline (maximum: 365 nm), and the porphyrins
matory response or a photo-oxidation reaction. In an inflamma-
(maxima: 392 nm, 500–650 nm) (Roberts 1984).
tory response, an initial insult to the tissue provokes a cascade
of events that eventually results in wider damage to the tissue
(Busch et al. 1999; Wang et al. 1999). In photo-oxidation reac-
Short screen for potential phototoxicity in the lens/retina
tions, a sensitizing compound in the eye absorbs light, is excitedto a singlet, then a triplet, state, and from the triplet produces
1. Chemical structure
free radicals and reactive oxygen species that in turn damage the
a. Heterocyclic, tricyclic, porphyrin
ocular tissues (Straight and Spikes 1985).
The photo-oxidation reaction is outlined below:
2. Absorbance spectra
a. Longer than 295 nm (lens); 400 nm (retina)
Drug + light → singlet → triplet → free radicals/
b. Binds to DNA, lens protein, melanin
3. Skin phototoxicity
reactive oxygen → ocular damage
In the older human, drugs/dietary supplements, dyes (i.e., hy-
Measure Binding of the Drug to Ocular Tissues
pericin), and diagnostic dyes (photodynamic therapy porphyrins)
Binding of a drug to ocular tissues (DNA—cornea, lens;
that absorb in the 400 to 600 nm region could produce phototoxic
proteins—lens; melanin—retina) (Roberts and Dillon 1987;
damage to both the lens (Schey et al. 2000) and the retina. In very
Dayhaw-Baker and Barker 1986; Steiner and Buhring 1990;
young humans all drugs/dyes absorbing at wavelengths longer
Sarna 1992) would increase its retention time in the eye. Fur-
than 295 nm are potential photosensitizers in both the retina as
thermore, binding of a photosensitizing substance to macro-
well as the lens. Drugs that do not absorb in these regions cannot
molecules increases the lifetime of its triplet state. It is the triplet
state of the dye or drug that leads to further oxidative and free-radical reactions. Therefore, drugs/dyes that bind to ocular tis-sues are very likely to induce phototoxic damage (Roberts et al.
Examine the Chemical Structure
1991a) in that tissue. Binding is determined by measuring the
Most drugs must have a tricyclic, heterocyclic, or porphyrin
absorption spectrum of the drug in the presence and absence of
ring to fulfill the energy requirements to produce a stable, long-
lens proteins, DNA, and/or melanin (Roberts and Dillon 1987;
lived (triplet) reactive molecule that will go on to produce free
Roberts, Atherton, and Dillon 1990; Steiner and Buhring 1990).
radicals and reactive oxygen species. The addition of a halide
A red shift in the absorbance spectrum of the drug in the pres-
group can enhance the amount of triplet produced (Turro 1978).
ence of any of these biomolecules is an indication of binding of
As seen in Figure 2, fluorescein and rose bengal have similar
the sensitizer to the biomolecule.
structures except for the attached iodine groups. Fluorescein is
highly fluorescent (singlet state) but is not very efficient at reach-
Note Reports of Skin Phototoxicity
ing the triplet state (poor quantum yield, 0.03, for the triplet).
Finally, any reports of skin phototoxicity for a particular drug
The singlet state of rose bengal would easily go through inter-
should provide a clear warning of potential ocular phototox-
system crossing to the triplet (good quantum yield for the triplet
icity. Skin phototoxicity is more readily apparent than ocular
(0.60). By simple inspection of the structures of these two di-
phototoxicity, although it is induced by compounds with similar
agnostic dyes, we would conclude that rose bengal has a much
chemical features (Oppenlander 1988).
greater potential to produce phototoxic damage to the eye thanfluorescein.
Detailed Screen for Predicting Ocular Phototoxicity
The short screen can determine a drug's potential for pho-
Test the Solubility Properties
totoxicity. Once it has been determined that a drug/dye is a
Dyes or drugs to be examined should be tested for their par-
possible photosensitizing agent, the more detailed screen given
titioning in protic and aprotic solvents. Their hydrophobicity
below can determine the potential site of damage (in situ fluo-
will indicate potential for crossing blood-ocular barriers. More
rescence techniques), predict the type and efficiency of damage
hydrophilic substances are less likely to cross blood-ocular bar-
(in vitro assays), and determine the mechanism of damage (pho-
riers. Compounds that are amphiphilic or lipophilic cross all
tophysical studies of the short lived excited state intermediates).
blood-retinal and/or blood-lenticular barriers. The probable site
Once the excited state intermediates produced by a particular
of damage may also be determined by the hydrophobicity (mem-
photosensitizing agent have been determined, the introduction
branes) or hydrophilicity (cytosol) of the photosensitizing dye
of specific quenching agents may stop the unwanted reactions.
(Roberts et al. 1991a).
Determine the Potential Sites for Ocular Damage
The site of damage is determined by the penetration of the
drug and the transmission of the appropriate wavelengths oflight to that site. At these sites there are numerous substrates forphototoxic damage in the eye.
Corneal epithelial and endothelial cells may be
easily damaged leading to keratitis (Pitts, Cullen, and Parr 1976;
Hull, Csukas, and Green 1983). However, these cells have a veryefficient repair mechanism and the damage is rarely permanent.
Uveal cells are highly melanized and are ordinarily
protected against light induced damage. However, melanogen-esis may be modified with phototoxic reactions leading to a
greater risk from UV radiation (Hu et al. 2000).
Epithelial cells of the lens have direct contact with
The structures of fluorescein and rose bengal. They have
the aqueous humor. Their function is to control transport to
heterocyclic ring systems.
the lens. They are most vulnerable to phototoxic damage.
Damage to these cells would readily compromise the viability
of the lens (Roberts et al. 1994). The lens fiber membrane can be
Detailed screen for ocular technique
photochemically damaged through damage to the lipids and/orthe main intrinsic membrane protein (Roberts, Roy, and Dillon
1. In vitro studies
1985). This will result in a change in the refractive index, causing
DNA, RNA, protein
an opacification. Phototoxic reactions can lead to a modification
of certain amino acids (histadine, tryptophan, cysteine) (Roberts
1984; McDermott et al. 1991) and/or a covalent attachment of
photoreceptor cell damage
sensitizer to cytosol lens proteins. In either case, this changes
b. Gel electrophoresis
the physical properties of the protein, leading to aggregation and
Amino acid analysis
finally opacification (cataractogenesis). The covalently bound
c. Mass spectrometry
chromophore may now act as an endogenous sensitizer, produc-
ing prolonged sensitivity to light. Because there is little turnover
of lens proteins, this damage is cumulative.
Phototoxic damage can occur in retinal pigment
epithelial tissues, the choroid, and the rod outer segments, which
contain the photoreceptors. If the damage is not extensive, there
f. Normalization for
are repair mechanisms to allow for recovery of retinal tissues.
However, extensive phototoxic damage to the retina can lead to
2. Biophysical techniques
permanent blindness (Dayhaw-Barker and Barker 1986; Ham
et al. 1982).
b. Laser flash photolysis
Determine the Location/Uptake of the Dye/Drug
The traditional method of determining up-
take into ocular tissues is in vivo radiolabeling. This method is
time consuming and expensive, although it is effective in de-
termining which ocular tissues have accumulated the drug in
d. Pulse radiolysis
Radical and oxyradical
An alternative method to de-
Radical and oxyradical
termine uptake of a drug into ocular tissues is ocular fluorome-
try. After a dye or drug has absorbed light and is excited to thesinglet state, it can decay to the ground state and is accompanied
by the emission of light. The is known as fluorescence. Because
The in vitro assays used in determining the phototoxic effi-
most photosensitizers are fluorescent, transmitted or reflective
ciency include the following:
fluorescence provides an accurate means of measuring uptake ofa sensitizer into ocular tissue that is simpler, less expensive, and
1. Cell Culture/Whole Tissues.
The first reported assay for
less arduous than using radiolabeled materials. This technique
phototoxicity in human ocular cells (Roberts l981) measured
may also be used noninvasively, in vivo, for instance using a slit
changes in macromolecular synthesis in the presence and
lamp, to detect uptake of sensitizers into the human, or scanning
absence of a light-activated drug. Other studies have assessed
or reflective fluorometry to determine the presence of endoge-
damage to corneal, lenticular, and retinal cells by measuring
nous and exogenous fluorescent materials in the retina (Docchio
pump function and enzyme activities both in vitro and in
1989; Docchio et al. 1991; Cubeddu et al. 1999; Sgarbossa et al.
situ (Andley et al. 1994; Roberts et al. l994; Organisiak and
Winkler 1994; Rao and Zigler 1992; Dayhaw-Barker 1987).
2. Gel Electrophoresis, Amino Acid Analysis.
Determine the Phototoxic Efficiency
phoresis has been used to monitor polymerization of ocular
The targets of photo-oxidative reactions may be proteins,
proteins (Kristensen et al. 1995; Roberts et al. 1992a; Roberts
lipids, DNA, RNA, and/or cell membranes (Straight and Spikes
l984; Zigler et al. 1982). Photopolymerization is one of the
l985). In vitro tests can be designed to determine the specific
most apparent changes in ocular protein induced by photo-
site(s) of damage to the various ocular compartments (i.e., lens
sensitizing dyes and drugs. Quantitative changes can be mea-
and retinal epithelial cells and photoreceptor cells) and the prod-
sured by scanning the gel and determining relative reaction
ucts of those reactions.
rates. Specific amino acid modifications can be determined
Table 2 presents a summary of additional biochemical and
using amino acid analysis (Roberts 1984, 1996). Zhu and
photophysical techniques that can be performed to more accu-
Crouch (1992) have illustrated the wide variety of classi-
rately predict the potential for and extent of in vivo phototoxicity.
cal protein analysis techniques (gel electrophoresis, amino
acid analysis, sequencing, isoelectric point determination,
= the intensity of the lamp at various wavelengths ad-
Western blot, enzyme-linked immunosorbent assay [ELISA])
justed for the transmission characteristics of the cornea or lens,
that can be used to investigate phototoxic damage induced by
AB is the absorbance of the dye/drug, and λ
is the number of
dyes and drugs.
photons at those wavelengths. The rates of each photo-oxidative
3. Mass Spectrometry.
Recent innovations in the field of
event are then adjusted accordingly for each sensitizer. This in
mass spectrometry (liquid secondary ion mass spectrome-
turn can be corrected for the actual transmission characteristics
try [LSIMS] and electrospray ionization [ESI]) have allowed
of the cornea and/or lens and the output of the sun to predict in
for the identification of specific amino acid modifications
within large proteins through molecular weight mapping.
In summary, in vitro techniques determine the potential dam-
These techniques have been applied to determine the specific
age done to an ocular substrate, which gives information about
sites of photooxidative damage in corneal and lenticular pro-
the photoefficiency of a drug should it be taken up into the
teins (Schey et al. 2000; Roberts et al. 2001). These studies
various compartments of the eye. Additional information about
can serve as a model for defining damage from any potential
the site of potential damage can be predicted based on which
phototoxic agent in the eye.
ocular substrate (DNA, RNA, protein, lipid) is affected.
4. Thin-Layer Chromatography.
This technique is particu-
larly effective at separating triacyclycerol, free fatty acid,
Determine the Excited State Intermediates That Cause
and phospholipids from lens (Fleschner l995) and retinal
the Phototoxic Damage
(Organisiak et al. 1992) membranes. Thin-layer chromo-
Some photochemical reactions in tissues are summarized
tography/gas chromotography/mass spectrometry (TLC/GC/
MS) may be used to measure lenticular or retinal lipid mod-
EXCITED STATES → INTERMEDIATES
ifications (Handelman and Dratz l986). Specific lipids maybe modified in the presence of photosensitizing agents and
separated on TLC plates. The plates can then be scanned for
quantitative analysis of these specific changes.
5. High-Performance Liquid Chromatography (HPLC).
→ TARGET and DAMAGE
HPLC is particularly effective at separating and identify-ing lipid peroxides from the retina (Akasaka, Ohrui, and
proteins = polymers
Meguro 1993). It has also been used to identify adducts
lipid = peroxides
formed between DNA nucleotides and phototoxic agents
RNA = cross-links
(Oroskar et al. 1994). HPLC has also been used to assess
The biophysical techniques used in determining the photo-
the rates of photo-oxidation of lens proteins in the presence
toxic efficiency include the following:
of a sensitizer. Using this technique, it is possible to deter-mine the induced amino acid modification within the protein
1. Quantum Yields.
In predicting the phototoxicity of a dye
and their location and to detect possible binding of sensitizing
or drug, it is important to determine what proportion of pho-
drugs to specific lens crystallins (McDermott et al. 1991).
tons lead to a benign (fluorescence, singlet state) event and
6. Normalization for Photons Absorbed.
Whatever the target
what proportion lead to a potentially destructive event (phos-
tissue or extent of damage, the toxic effects of these dyes
phorescence, triplet state). The efficiency of a photoinduced
and drugs are the result of photochemical reactions. As such,
process may be expressed as its quantum yield (Q
their rate of efficiency is dependant on the number of photonsabsorbed by the sensitizer in the biological tissue. Therefore,in order to get an accurate comparison of the photosensitizing
= the number of photons used to produce an event
the number of photons originally absorbed
potency of various dyes and drugs with different structuresand absorptive characteristics, it is essential to normalize for
It gives a measure of how likely a photochemical event
the number of photons absorbed by each drug in a particular
is to occur. The quantum yield is often expressed as a per-
centage. For instance, the quantum yield for fluorescence of
This can be done with a simple computer-generated math-
fluorescein is 0.92 and that of rose bengal is 0.08 (Figure 2).
ematical formula (Roberts 1996; Kristensen et al. 1995),
This means that most of the light absorbed by fluorescein
which takes into account the absorption spectrum of the drug,
is given off in the form of fluorescence energy, making it a
the output of the lamp source used in the experiments, and
relative safe diagnostic dye for the eye. On the other hand,
the optical properties of the eye. The total relative number
the triplet quantum yield for fluorescein is 0.03 and for rose
of photons absorbed by a drug under particular experimental
bengal it is 0.60. This indicates that very little of the singlet
conditions is the area under the product curve:
fluorescein energy is transformed into a triplet, whereas mostof the energy of rose bengal will be available for intersystem
Photons absorbed = I
× AB × λ
crossing and will reach the triplet state from whence it can
produce ocular damage. Therefore, flourescein is appropriate
transfer (electron exchange) between the sensitizing drug in
but rose bengal would be an inappropriate diagnostic dye for
an excited state and a substrate from the ocular tissues. Al-
though these radicals are very short lived, they can be ob-
2. Laser Flash Photolysis.
This method uses a pulse of mono-
served with ESR in situ during their photogeneration. For
chromatic light to promote a specific dye or drug to an excited
instance, illumination of rose bengal (RB), an ophthalmic
state (Rodgers 1985).
diagnostic dye, in the presence of an electron donor such as
NADH, affords a radical anion of the dye RB·−
to be directly
i. Triplet Detection.
Time resolved techniques (either ab-
measured using ESR (Sarna et al. 1991).
sorption spectroscopy or diffuse reflectance) allow for the
Radicals that are too reactive to accumulate in detectable
detection of the triplet state of the excited chromophore
quantities can frequently be detected by ESR using spin
even in intact tissues. This technique has been used to de-
trapping techniques. In this approach, an agent called a spin
termine the absence (Dillon and Atherton l990) of triplet
trap reacts with a short-lived radical R·
to give a spin adduct
formation by the endogenous 3-hydroxykynurenine, as
, which has a much longer lifetime. The original rad-
well as the presence of triplets from sensitizing drugs in
is identified by the characteristic ESR spectrum of
intact lenses (Roberts, Atherton, and Dillon l991b).
. Carbon-, nitrogen-, and sulfur-centered radicals, as
All ocular damage from photooxida-
well as all of the important oxygen-centered radicals, (hy-
tion reactions occurs through the triplet state of the drug.
droxyl, superoxide, alkoxyl and peroxyl) can be identified
The longer the lifetime, the greater the potential for dam-
using ESR either directly or in combination with the spin
age. The lifetimes of the triplets of sensitizing drugs were
found to be greater when bound to macromolecules or
Using these techniques, the photosensitized generation
in an intact organ than when free in solution (Roberts,
of superoxide in protic (Reszka, Lown, and Chignell l992)
Atherton, and Dillion 1991b). The presence of a triplet
and aprotic (Reszka et al. 1993) media can be monitored.
and an increase in lifetime when bound to intact ocular
These are model systems for the hydrophilic (aqueous) and
tissue is predictive of a drug's causing photo-oxidative
hydrophobic (rod outer segment membrane) portions of the
damage to the eye in vivo (Roberts et al. 1991a).
eye. ESR has recently been used to define the photochemi-cal mechanisms involved in the light activation of endoge-
3. Luminescence—Singlet Oxygen.
The presence and lifetime
nous pigments in the lens (Reszka et al. 1996) and the retina
of singlet oxygen can be determined using time resolved in-
(Reszka et al. 1995). With these systems defined, ESR can be
frared luminescence measurements at 1270 nm (Rodgers and
used to predict potential phototoxic events induced by exoge-
Snowden l982). Using this technique, it can be determined
nous photosensitizing dyes and drugs and natural pigments
whether and how efficiently a dye or drug can produce singlet
(Roberts et al. 2000; Motten et al. 1999).
oxygen (Roberts et al. 2000; Motten et al. 1999). Because sin-glet oxygen is the most powerful oxidant in a photooxidationreaction, a dye or drug that is an efficient producer of singlet
In summary, the molecular mechanism involved in the pho-
oxygen could be predicted to induce phototoxic damage if
totoxic damage induced in the eye is photosensitized oxidation
present in the eye.
reactions. This mechanism begins with the absorption of light by
4. Pulse Radiolysis—Hydroxyl, Peroxy Radicals and Super-
the sensitizing compound (endogenous pigment, dye, or drug),
Pulse radiolysis consists of the delivery of a very short
which promotes the compound to an excited singlet state (short
intense pulse of ionizing radiation to a sample, the resultant
lived) and then, through intersystem crossing, goes to the triplet
changes in light absorption of the sample being followed by a
state (long lived). The excited triplet state of the drug/dye then
very fast spectrophotometer (Land 1985). The technique may
proceeds either via a type I (free radical) or type II (singlet
be used to detect the formation of short-lived radical species
oxygen) mechanism, causing the eventual biological damage
of a dye or drug (Roberts, Hu, and Wishart 1998; Roberts
(Straight and Spikes 1985). Therefore, information about the
et al. 2000). In addition, the interaction of a dye or drug with
efficiency and excited state intermediates for a phototoxic re-
excited oxygen intermediates (hydroxyl radical, superoxide,
action in the eye obtained by using photophysical techniques
peroxy radicals) (Land et al. 1983), which are clearly gener-
(fluorescence, flash photolysis, pulse radiolysis, esr) can be pre-
ated in this system, allows for an understanding of a possible
dictive of phototoxicity in vivo. We have confirmed that pho-
mechanism of in vivo photo-oxidative ocular damage.
tophysical studies collate well with in vivo data (Roberts et al.
1991a). For instance, tetrasulphonatophenylporphyrin (TPPS),
Carbon-Centered Radicals and Superoxide.
which binds to lens proteins, shows a long-lived triplet in the
resonance (ESR or EPR) spectroscopy detects and character-
intact calf and human lens, and produces singlet oxygen effi-
izes species containing an odd number of electrons, namely
ciently, causes photo-oxidative damage in vivo in pigmented
free radicals and paramagnetic metal ions. The photo-
mouse eyes, whereas uroporphyrin (URO), which produces an
oxidation reactions responsible for the phototoxic responses
efficient triplet but does not bind to ocular tissues, does not cause
in the eye involve free radicals that are formed via electron
photooxidative damage in vivo.
In Vivo Testing
tion of exogenous photosensitizing dyes, drugs, and herbal sup-
The use of either the short or more detailed (Tables 1 and
plements in the eye, which may potentially harm the eye. Also,
2) screens for ocular phototoxicity will not totally eliminate the
for those drugs that must be continued, in spite of their phototox-
need for accurate in vivo experiments. The function of these stud-
icity (i.e., antimalarial drugs; Motten et al. 1999), appropriate
ies is to limit the need for in vivo testing for ocular phototoxicity
protection against light (sunglasses) and specific supplementary
of large numbers of drugs. Those drugs found in screening to be
antioxidants may be prescribed to retard or eliminate the most
highly likely to produce phototoxic side effects in the eye should
severe blinding side effects (Roberts and Mathews-Roth 1993).
be tested further with animal studies to determine the exact site
and extent of damage to be expected in humans (Roberts et al.
Prolonged use of a phototoxic drug is most probably of greater
Akasaka, K., H. Ohrui, and H. Meguro. 1993. Normal-phase high-performance
long-term risk to the eye than short-term dosage because of cu-
liquid chromatography with a fluorimetric postcolumn detection system forlipid hydroperoxides. J. Chromatog.
mulative damage. Because there is an active repair system in the
Andley, U. P. 1987. Photodamage to the eye. Photochem. Photobiol.
cornea, there should be little or no long-term side effects of pho-
totoxicity there. However, because there is no turnover in the lens
Andley, U. P., J. S. Hebert, A. R. Marrison, J. R. Reddan, and A. P. Pentland.
constituents, any modification in that tissue will tend to stay and
1994. Modulation of lens epithelial cell proliferation by enhanced
accumulate with age. Thus cataractogenesis may not develop
prostaglandin synthesis after UVB exposure. Invest. Ophthalmol. Vis. Sci.
until much later than the initial insult. In addition, phototoxic
Bachem, A. 1956. Ophthalmic action spectra. Am. J. Ophthal.
damage to the lens may not only cause direct damage to cell via-
Balasubramanian, D. 2000. Ultraviolet radiation and cataract. J. Ocul. Pharma-
bility but may undermine its defense system, so that gross mor-
phological effects may appear much later than the original insult.
Barker, F. M., G. C. Brainard, and P. Dayhaw-Barker. 1991. Transmission of the
The susceptibility of both the lens and retina to light-induced
human lens as a function of age. Invest. Ophthalmol. Vis. Sci.
Busch, E. M. ,T. G. Gorgels, J. E. Roberts, and D. van Norren. 1999. The effects
damage increases with the age-related changes in chromophores
of two stereoisomers of N
-acetylcysteine on photochemical damage by UVA
(Roberts et al. 2000, 2001, 2002), with concurrent decrease in
and blue light in rat retina. Photochem. Photobiol.
antioxidant status (Giblin 2000; Samiec et al. 1998; Sarna 1992).
Cubeddu, R., P. Taroni, D. N. Hu, N. Sakai, K. Nakanishi, and J. E. Roberts.
Therefore the elderly may be particularly sensitive to drugs and
1999. Photophysical studies of A2-E, putative precursor of lipofuscin, in
other agents that induce phototoxic side effects. Sight may be re-
human retinal pigment epithelial cells. Photochem. Photobiol.
gained after cataract surgery; however, damage to the retina that
Dayhaw-Barker, P. 1987. Ocular photosensitization. Photochem. Photobiol.
is not repaired leads to permanent blindness. The environmental
lighting, particularly the constant presence of intense ambient
Dayhaw-Barker, P., and F. M. Barker. l986. Photoeffects on the eye. In Photo-
light, must always be taken into account when assessing the
biology of the skin and the eye
, ed. E. M. Jackson, 117–147. New York, NY:
potential in vivo ocular toxicity of a drug.
Dillon, J. 1991. Photophysics and photobiology of the eye. J. Photochem.
Photobiol. B Biol.
Dillon, J., and S. J. Atherton. 1990. Time resolved spectroscopic studies on the
Even if a drug has the potential to produce phototoxic side
intact human lens. Photochem. Photobiol.
effects in the eye, no damage will be done if the specific wave-
Docchio, F. 1989. Ocular fluorometry: Principles, fluorophores, instrumentation
and clinical applications. Lasers Surg. Med.
lengths of optical radiation absorbed by the drug are blocked
Docchio, F., M. Boulton, R. Cubeddu, R. Ramponi, and P. Dayhaw-Barker. 1991.
from transmittance to the eye. This can be easily done with
Age-related changes in the flourescence of melanin and lipofuscin granules
wrap-around eyeglasses (Gallas and Eisner 2001; Sliney 1999;
of the retinal pigment epithelium: A time-resolved fluorescence spectroscopy
Merriam 1996) that incorporate specific filters. Furthermore,
study. Photochem. Photobiol.
nontoxic quenchers and scavengers could be given in conjunc-
Edge, R., E. J. Land, D. McGarvey, L. Mulroy, and T. G. Truscott. 1998. Rel-
ative one-electron reduction potentials of carotenoid radical cations and the
tion with the phototoxic drug to negate its ocular side effects
interactions of carotenoids with the vitamin E radical cation. J. Am. Chem.
while allowing for the primary effect of the drug (Roberts 1981;
Roberts et al. 1991a; Roberts and Mathews-Roth 1993).
Fleschner, C. R. 1995. Fatty acid composition of triacylglycerols, free fatty acid
and phospholipids from bovine lens membrane fractions. Invest. Ophthal. Vis.
Fraunfelder, F. T. 1982. Drug-induced ocular side effects and drug interactions
2nd ed., Philadelphia: Lea & Febiger.
With simple, inexpensive in vitro testing, compounds can be
Gallas, J., and M. Eisner. 2001. Eye protection from sunlight damage. In Sun-
monitored at their developmental stage for potential ocular pho-
light protection in man
, ed. P. U Giacomoni, 437–455. Amsterdam, Holland:
totoxicity. It may be that a portion of the molecule can be modi-
fied to reduce phototoxicity while leaving the primary drug effect
Giblin, F. J. 2000. Glutathione: A vital lens antioxidant. J. Ocul. Pharmacol.
intact. This may reduce the necessity of later, more costly, drug
Ham, W. T., H. A. Mueller, J. J. Ruffolo, D. Guerry, and R. K. Guerry. 1982.
recalls. In the future, the more effective use of ocular fluorometry
Action spectrum for retinal injury from near ultraviolet radiation in the aphakic
will allow for a more accurate assessment of the uptake and loca-
monkey. Am. J. Ophthal.
Handelman, G. J., and E. A. Dratz. l986. The role of antioxidants in the retina
Roberts, J. E. 1981. The effects of photooxidation by proflavin in HeLa cells I.
and retinal pigment epithelium and the nature of prooxidant induced damage.
The molecular mechanisms. Photochem. Photobiol.
Adv. Free Radic. Biol. Med.
Roberts, J. E. 1984. The photodynamic effect of chlorpromazine, promazine
Hu, D. N. 2000. Regulation of growth and melanogenesis of uveal melanocytes.
and hematoporphyrin on lens protein. Invest. Ophthalmol. Vis. Sci.
Pigment Cell Res.
Hull, D. S., S. Csukas, and K. Green. l983. Trifluoperazine: Corneal endothelial
Roberts, J. E. 1996. Ocular phototoxicity. In Dermatotoxicology
, 5th ed., eds. F.
phototoxicity. Photochem. Photobiol.
Marzulli and M. Maiback, 307–313. Washington, DC: Taylor and Francis.
Khachik, F., P. S. Bernstein, and D. L. Garland. 1997. Identification of lutein and
Roberts, J. E. 2000. Light and immunomodulation. Ann. N.Y. Acad. Sci.
zeaxanthin oxidation products in human and monkey retinas. Invest. Ophthal.
Roberts, J. E. 2001. Ocular phototoxicity. J. Photochem. Photobiol. B Biol.
Kristensen, S., R. H. Wang, H. Tonnesen, J. Dillon, and J. E. Roberts. 1995. Pho-
toreactivity of biologically active compounds. VII. Photosensitized polymer-
Roberts, J. E., S. J. Atherton, and J. Dillon. 1990. Photophysical studies
ization of lens proteins by antimalarial drugs in vitro
. Photochem. Photobiol
on the binding of tetrasulphonatophenylporphyrin (TPPS) to lens proteins.
Kwan, M., J. Niinikoske, and T. K. Hunt. l971. Oxygen tension in the aqueous
Roberts, J. E., S. J. Atherton, and J. Dillon. 1991b. Detection of porphyrin excited
and the lens. Invest. Ophthal.
states in the intact bovine lens. Photochem. Photobiol.
Land, E. J. 1985. Pulse radiolysis. In Primary photoprocesses in biology and
Roberts, J. E., and J. Dillon. 1987. In vitro
studies on the photosensitized
, eds. R. V. Bensasson, G. Jori, E. J. Land, and T. J. Truscott, 35–44.
oxidation of lens proteins by porphyrins. Photochem. Photobiol.
Amsterdam, Holland: Elsevier.
Land, E. J., T. Mukherjee, A. J. Swallow, and J. M. Bruce. 1983. One-electron
Roberts, J. E., E. L. Finley, S. A. Patat, and K. L. Schey. 2001. Photooxidation
reduction of Adriamycin: Properties of the semiquinone. Arch. Biochem.
of lens proteins with xanthurenic acid: A putative chromophore for catarac-
togenesis. Photochem. Photobiol.
McLaren, J. W., S. Dinslage, J. P. Dillon, J. E. Roberts, and R. F. Brubaker.
Roberts, J. E., A. Harriman, S. J. Atherton, and J. Dillon. 1992a. A noninvasive
1998. Measuring oxygen tension in the anterior chamber of rabbits. Invest.
method to detect oxygen tensions and other environmental factors in the lens.
Ophthalmol. Vis. Sci.
In Proceedings of the International Society of Ocular Phototoxicity
McDermott, M., R. Chiesa, J. E. Roberts, and J. Dillon. 1991. Photooxida-
64, Sedona, Arizona.
tion of specific residues in a-crystallin polypeptides. Biochemistry
Roberts, J. E., D. N. Hu, and J. F. Wishart. 1998. Pulse radiolysis studies of
melatonin and chloromelatonin. J. Photochem. Photobiol. B. Biol.
Merriam, J. C. 1996. The concentration of light in the human lens. Trans. Am.
Roberts, J. E, J. Kinley, A. Young, G. Jenkins, S. J. Atherton, and J. Dillon. 1991a.
Motten, A.G., L. J. Martinez, N. Holt, R. H. Sik, K. Reszka, C. F. Chignell,
and photophysical studies on photooxidative damage to lens proteins
H. H. Tonnesen, and J. E. Roberts. 1999. Photophysical studies on antimalarial
and their protection by radioprotectors. Photochem. Photobiol.
drugs. Photochem. Photobiol.
Roberts, J. E., B. M. Kukielczak, D.-N. Hu, D. S. Miller, P. Bilski, R. H. Sik, A. G.
Offord, W. A., N. A. Sharif, K. Mace, Y. Tromvoukis, E. A. Spillare, O. Avanti,
Motten, and C. F. Chignell. 2002. The role of A2E in prevention or enhance-
W. E. Howe, and A. M. A. Pfeifer. 1999. Immortalized human corneal ep-
ment of light damage in human retinal pigment epithelial cells. Photochem.
ithelial cells for ocular toxicity and inflammation studies. Invest. Ophthalmol.
Roberts, J. E., and M. Mathews-Roth. 1993. Cysteine ameliorates photosensi-
Oppenlander, T. 1988. A comprehensive photochemical and photophysical assay
tivity in erythropoietic protoporphyria. Arch. Dermatol.
exploring the photoreactivity of drugs. Chimia
Roberts, J. E., C. Reme, M. Terman, and J. Dillon. 1992b. Exposure to bright
Organisciak, D. T., M. A. Darrow, Y.-L. Jiang, G. E. Marak, and J. C. Blanks.
light and the concurrent use of photosensitizing drugs. New. Engl. J. Med.
1992. Protection by dimethylthiourea against retinal light damage in rats.
Invest. Ophthalmol. Vis. Sci.
Roberts, J. E., D. Roy, and J. Dillon. 1985. The photosensitized oxidation of the
Organisiak, D. T., and B. S. Winkler. 1994. Retinal light damage: Practical and
calf lens main intrinsic protein (MP26) with hematoporphyrin. Curr. Eye Res.
theoretical considerations. Prog. Retinal Res.
Oroskar, A., G. Olack, M. J. Peak, and F. P. Gasparro. 1994. 4-Aminomethyl-
Roberts, J. E., S. Schieb, W. H. Garner, and M. Lou. 1994. Development of a
4,5,8-trimethylpsoralen photochemistry: The effect of concentration and
new photo-oxidative induced cataract model using TPPS in an intact lens.
UVA fluence on photoadduct formation in poly(dA-dT) and calf thymus DNA.
Invest. Ophthal. Vis. Sci.
Roberts, J. E., J. F. Wishart, L. Martinez, and C. F. Chignell. 2000. Pho-
Pitts, D. G., A. P. Cullen, and W. H. Parr. 1976. Ocular ultraviolet effects in the
tochemical studies on xanthurenic acid. Photochem. Photobiol.
rabbit eye. DHEW (NIOSH) Publication
Rao, D. M., and J. S. Zigler, Jr. 1992. Levels of reduced pyridine nucleotides
Rodgers, M. A. J. 1985. Instrumentation for the generation and detection of
and lens photodamage. Photochem. Photobiol.
transient species. In Primary photoprocesses in biology and medicine
Reszka, K., P. Bilski, C. F. Chignell, and J. Dillon. 1996. Photosensitization by
R. V. Bensasson, G. Jori, E. J. Land, and T. J. Truscott, 1–24. North Holland:
the human lens component kynurenine: An EPR and spin trapping investiga-
tion. Free Radic. Biol. Med.
Rodgers, M. A. J., and P. T. Snowden. 1982. Lifetime of O2 (singlet delta) in
Reszka, K., P. Bilski, R. H. Sik, and C. F. Chignell. 1993. Photosensitized gen-
liquid water as determined by time resolved infra-red luminescence measure-
eration of superoxide radical in aprotic solvents: An EPR and spin trapping
ments. J. Am. Chem. Soc.
study. Free Radic. Res. Commun.
Samiec, P. S., C. Drews-Botsch, E. W. Flagge, J. C. Kurtz, P. Sternberg, R. L.
Reszka, K., G. Eldred, R. H. Wang, C. F. Chignell, and J. Dillon. 1995. Pho-
Reed, and D. P. Jones. 1998. Glutathione in human plasma decline in associ-
tochemistry of human lipofuscin as studied by EPR. Photochem. Photobiol.
ation with aging, age-related macular degeneration and diabetes. Free Radic.
Reszka, K., J. W. Lown, and C. F. Chignell. 1992. Photosensitization by ani-
Sarna, T. 1992. Ocular melanins. J. Photochem. Photobiol.
cancer agents. 10. Ortho-semiquinone and superoxide radicals produced dur-
Sarna, R., J. Zajac, M. K. Bowman, and T. G. Truscott. 1991. Photoinduced
ing anthapyrazole-sensitized oxidation of catechols. Photochem. Photobiol.
electron transfer reactions of rose bengal and selected electron donors.
J. Photochem. Photobiol. A Chem.
Schey, K. L., S. Patat, C. F. Chignell, M. Datillo, R. H. Wang, and J. E. Roberts.
Straight, R., and J. D. Spikes. 1985. Photosensitized oxidation of biomolecules.
2000. Photooxidation of lens proteins by hypericin (active ingredient in
In Singlet O, Vol IV. Polymers and Biopolymers
, ed. A. Frimer, 91–143.
St. John's wort). Photochem. Photobiol.
Boca Raton, FL: CRC Press.
Sethna, S. S., A. M. Holleschau, and W. B. Rathbun. 1983. Activity of glutathione
Turro, N. 1978. Modern molecular photochemistry
. Menlo Park, CA:
synthesis enzymes in human lens related to age. Curr. Eye Res.
Seth, R. K., and S. Kharb. 1999. Protective function of alpha-tocopherol against
Wang, X. C., C. Jobin, J. B. Allen, W. Roberts, and G. J. Jaffe. 1999. Suppres-
process of cataractogenesis in humans. Ann. Nutr. Metab.
sion of NF-kappaB-dependent proinflammatory gene expression in human
Sgarbossa, A., N. Angelini, D. Gioffre, T. Youssef, F. Lenci, and J. E. Roberts.
RPE cells by a proteasome inhibitor. Invest. Ophthalmol. Vis. Sci.
2001. The uptake, location and fluorescence of hypericin in bovine intact lens.
Curr. Eye Res.
Zhu, L., and R. K. Crouch. 1992. Albumin in the cornea is oxidized by hydrogen
Sliney, D. H. 1999. Geometrical assessment of ocular exposure to environmen-
tal UV radiation—implications for ophthalmic epidemiology. J. Epidemiol.
Zigler, J. S. Jr., H. M. Jernigan, Jr., N. S. Perlmutter, and J. H. Kinoshita. 1982.
Photodynamic cross-linking of polypeptides in intact rat lens. Exp. Eye Res.
Steiner, K., and K. U. Buhring. l990. The melanin binding of bisoprolol and its
toxicological relevance. Lens Eye Toxicol. Res.
Zigman, S. l993. Ocular light damage. Photochem. Photobiol.
ARIMIDEX® (anastrozole) Tablets The primary endpoint of the trial was disease-free survival (ie, time to occurrence of a distant or local recurrence, or contralateral breast cancer ordeath from any cause). Secondary endpoints of the trial included distant disease-free survival, the incidence of contralateral breast cancer and overall survival. At a median follow-up of 33 months, the combination of ARIMIDEX and tamoxifen did not demonstrate any efficacy benefit when comparedwith tamoxifen in all patients as well as in the hormone receptor positive subpopulation. This treatment arm was discontinued from the trial.
Summer 2007 Turn-around for trees Promotional Pointers Garden MessengerPlant Promotions for 2007 Plant SpotlightHelping you pick the winners! For beautiful blossom in spring and attractive fruits and foliage in autumn,Malus or ornamental apple, is hard to beat. Add to this their compact size andthe availability of varieties that are resistant to scab and mildew and you've gotthe ideal tree for most gardens. This fact is reflected in the way that many gar-den writers list Malus among their top small garden trees.