INTRODUCTION AND LITERATURE REVIEW
Oxygen toxicity
Oxygen has been known since the eighteenth century to be poisonous.1 All
organisms that use oxygen and are therefore exposed to it and to its various
reduction products must protect themselves against oxygen toxicity. It is hardly
necessary to point out that oxygen is toxic to anaerobic microorganisms, though its
effect on these varies from immediate cell death to bacteriostasis. It is less obvious,
however, that oxygen is poisonous to aerobic organisms, as well. High
concentrations of oxygen, above the normal atmospheric level of 21%, slow plant
growth2, kill cultured animal cells3, and increase the rate of occurrence of blindness
in premature infants4. Conversely, lowering ambient oxygen levels below 21%
speeds the growth rate of both plants and cultured animal cells. Hyperbaric oxygen
causes chromosome breakage in grain5 and mutations in bacteria6. Breathing pure
oxygen is so toxic that there has been at least one human fatality resulting from
medical treatment with hyperbaric oxygen7.
Figure 1. The partial reduction products of oxygen.8
_____________________________________________________________
Free radical theory of oxygen toxicity
The causes of the poisonous properties of oxygen were obscure before the
publication of Gershman's free radical theory of oxygen toxicity, which states that the
toxicity of oxygen is due to partially reduced forms of oxygen9. The unusually high
concentrations of oxygen found in tissues during exposure to pure oxygen
encourages the formation of these reduced oxygen species. The toxic species are
formed by one-electron reductions of oxygen by compounds and enzymatic
reactions occurring in vivo8(fig. 1). These reduction products include superoxide
(O2-), hydrogen peroxide (H2O2), and the hydroxyl radical (OH.). These are
commonly referred to as active oxygen species. Ordinary dioxygen is a biradical,
since its two free electrons exist in two different orbitals and have parallel spins8.
Dioxygen's reactivity, however, is greatly lowered by the spin restriction: both of
these electrons are of the same spin, so that if oxygen is to oxidize another molecule
by accepting a pair of electrons from it, that pair must be of antiparallel spin, which
cannot occur if both electrons occupy the same orbital. The spin restriction is
removed when dioxygen is excited to the state known as singlet oxygen. In the cell,
the spin restriction is removed enzymatically by complexing the dioxygen with
transition metals, which results in a delocalization of the electrons. However, the
most common biological reductions of oxygen occur by the transfer of single
electrons.10 All of these forms of oxygen are reactive as oxidants and reductants to a
greater or lesser extent with a wide variety of compounds, both in vitro and in vivo.
Superoxide
Sources of superoxide
Superoxide is one of the species that has been implicated in the toxicity of
oxygen. Generation of O2- in vitro has been observed in a number of different
experimental systems. These systems include the enzyme xanthine oxidase acting
upon the substrates xanthine or hypoxanthine11,12, and the reduction of flavin
mononucleotide and flavodoxin by ferredoxin-TPN+ oxidoreductase13. Aqueous
O2- may also be generated in vitro by dissolving tetramethyl-ammonium superoxide
or potassium superoxide in water. Subcellular components such as cell nuclei14, 15,
chloroplasts16,17, mitochondria18, human monocytes and neutrophils19,
macrophages20, and microsomes21 have been demonstrated to produce O2- in
vitro. The generation of O2- by phagocytic cells has a microbicidal action essential to
normal phagocytosis22. The autoxidation of many biological compounds results in
the production of O2-; among these compounds are oxyhemoglobin, which
becomes methemoglobin23, oxymyoglobin, which becomes metmyoglobin24,
clostridial and spinach ferredoxins25, ascorbate, catecholamines, phenols, and
reduced flavins26. Other circumstances in which O2- is generated include the
photolysis of water27, and the excitation of carboquone, mitomycin c, and
streptonigrin by visible light in the presence of oxygen28. Superoxide is also
produced by redox-cycling compounds in vivo. As an example, bipyridylium
herbicides, such as paraquat, are reduced by an NADPH-dependent cytoplasmic
diaphorase in E. coli, and the monocation radicals so produced reduce dioxygen to
O2- 29.
Reactions of superoxide
Superoxide has been shown to be reactive to a number of compounds in
vitro, as has its conjugate acid, the perhydroxyl radical, HO2. While less than 1% of
O2- will be present as HO2 at the physiological pH of 7.4, the concentration increases
dramatically as pH decreases. Since the pH is lower in the microenvironment
adjacent to a membrane surface or in lysosomes, both the protonated and the
unprotonated forms of O2- should be considered. Among the reactions of O2-
observed in vitro are reactions with small molecules: ascorbate reacts with a rate
constant of 1.5 x 105 M-1s-1 at pH 9.030. a-tocopherol is oxidized by O2- to
8a-hydroxy-a-tocopherone, which spontaneously forms a-tocopherol quinone31.
NADH is oxidized by HO2 and O2-, when bound to lactate dehydrogenase, with rate
constants at 235C of 1.2 x 106 M-1s-1 and 3.6 x 106 M-1s-1, respectively32.
Superoxide also oxidizes and reduces many transition metal ions and their
complexes33. For example, it oxidizes Mn2+ with a rate constant of 6 x 106 M-1s-1 34,
and it reduces Fe3+ complexed to EDTA with a rate constant of 1.3 x 106 M-1s-1; it
reduces Fe3+ complexed to DTPA very slowly, if at all35. The reduction of cytochrome
c by xanthine and xanthine oxidase is mediated by O2- 36, and this reduction is used
as the basis of assays for both O2- and superoxide dismutase (SOD). The rate
constant for the reaction of O2- with ferricytochrome c is 2.6 x 105 M-1s-1, at pH
9.0.30 Other assays for O2- exploit the fact that it reacts with nitroblue tetrazolium, at
a rate constant of 5.9 x 104 M-1s-1.30,37. The perhydroxyl radical has been shown to
react in vitro with fatty acids; the rate constant for the reaction with linoleic acid is 1.2
x 103 M-1s-1; with linolenic acid, the rate constant is 1.7 x 103 M-1s-1; and with
arachidonic acid, it is 3.0 x 103 M-1s-1.38 Sulfur compounds such as dithiothreitol,
2-mercaptoethanol, reduced glutathione, and ethyl mercaptan are oxidized in air;
superoxide has been shown to play a part in these reactions39 and in the oxidation of
phenols40 and catecholamines.41 Superoxide also reacts with certain proteins in
vitro, inactivating them; among these are the enzyme catalase42, and the
selenium-containing glutathione peroxidase43. Sinovial fluid, the breakdown of
which is part of the pathology of arthritis44, has been shown in vitro to be degraded
by a O2--generating system45. In addition, O2- may react with the lipids of intact cell
membrane46, 47. The O2--generating reaction of xanthine oxidase oxidizing
acetaldehyde damages resealed erythrocyte ghosts containing lipid
hydroperoxides, while the enzyme superoxide dismutase protects against this
damage.48 DNA strand scission is yet another form of damage caused by O2-
generation in vitro49, 50, 51.
In vivo effects of superoxide
Further, O2- has been demonstrated to cause some deleterious effects in
vivo. Superoxide generated extracellularly by a xanthine oxidase/purine system has
been shown to kill the bacterium Staphylococcus epiderimidis52. Potassium
superoxide, dissolved in water to produce 2 nmoles/ml O2-, caused 50% cell
mortality in cultured Chinese hamster ovary cells, and caused an increase of
mutants in a dose-related fashion53. O2--generating systems have also been
shown to cause mutations and cancerous tumors. C3H mouse fibroblast 10T1/2
cells, exposed either to human neutrophils which had been stimulated to synthesize
reactive oxygen intermediates, or to hypoxanthine and xanthine oxidase, then
placed in tissue culture, were malignantly transformed in vitro.54 When injected into
nude mice, the cells resulted in the production of malignant and benign tumors in
vivo 55.
Arguments against the toxicity of superoxide
Despite the aforementioned data concerning the reactivity of superoxide,
certain investigators contend that O2- is not toxic.56,57,58 Their conclusions are
based largely on the limited lifespan of O2- in aqueous media59, and its unreactivity
in aprotic solvents due to its relative insolubility in them33. Furthermore, the
reactivities of a number of specific compounds with O2- in aqueous solution are low.
The buffer Tris, the chelator EDTA, and certain metabolites such as a-ketoglutarate,
pyruvate, and succinate are all fairly unreactive, with second-order rate constants
ranging from .001 to 0.3 M-1s-1.30 Superoxide is not very reactive toward amino
acids, and neither is its protonated form, HO260. For HO2, the rate constants with
amino acids range from 10 M-1s-1 for the aliphatic amino acids to about 60 M-1s-1 for
the aromatic amino acids. For O2-, the range is from about 0.1 to 20 M-1s-1.60 Thus, it
is evident that O2- is not a universally reactive species.
Hydrogen peroxide
While it is possible that O2- is not itself reactive enough to explain all of the
biological damage resulting from systems that produce O2-, it is certainly evident
that the damage does occur. Some reaction rates of O2- given above are high
enough to indicate that the species is potentially damaging in itself; additional
damage may be caused by secondary species formed from O2-. One such
potentially damaging species is H2O2. Hydrogen peroxide will be present whenever
O2- is formed, because the dismutation of O2- generates molecular oxygen plus
H2O2, reaction (1):
O2- + O2- + 2 H+ ---> H2O2 + O2 (1)
The spontaneous rate for this reaction at physiological pH is 5 x 105 M-1s-1, and the
enzyme superoxide dismutase, which catalyzes this reaction, increases it by a
factor of 104.44 Thus, much of the damage that has been attributed to O2- may
actually have been caused by H2O2 that was formed from the O2- that was initially
generated.
Biological sources of hydrogen peroxide
Biological sources of H2O2 include lung mitochondria61, whose release of
H2O2 is increased ten-fold as the ambient oxygen concentration rises from 21% to
100% at one atmosphere pressure62. Activation of phagocytic cells also results in
H2O2 production through the dismutation of O2- generated by a membrane-bound
NADPH oxidase.22 Hydrogen peroxide is also formed during the one and two
electron oxidations of numerous phenols, thiols, and catecholamines.63 Insulin
stimulates intracellular H2O2 production in rat epidydimal fat cells; H2O2 has been
proposed to act as a second messenger for this hormone64.
Damage produced by hydrogen peroxide
A number of forms of biological damage have been shown to be caused by
H2O2. One of several sites at which damage has been clearly shown to be due to
H2O2, and not to O2- or OH., is the lens of the eye in organ culture65. The autoxidation
of oxymyoglobin to metmyoglobin, mentioned above as a source of O2-, has been
shown to be inhibited by catalase, the enzyme which breaks down H2O2, suggesting
that peroxide may initiate this oxidation.24 DNA is damaged by H2O2 in the presence
of metals66, 67, 68, 69; low concentrations of H2O2 induces resistance of procaryotic
cells to degradation of DNA by a future challenge with H2O2.67 It also causes cell
death in fibroblasts22 and bacteria70, mutagenicity in bacteria,70 and it has been
shown to cause tumors in Drosophila embryos71. Hydrogen peroxide also slowly but
irreversibly inactivates the enzyme superoxide dismutase72.
Metal catalyzed generation of the hydroxyl radical
Hydrogen peroxide is not the only reactive oxygen species that may be
formed from O2-. Haber and Weiss mentioned in a 1934 paper73 the earlier proposal
by Haber and Willstätter that O2- could react with hydrogen peroxide to form OH.:
O2- + H2O2 ---> H2O + OH- + OH. (2)
This reaction has become known as the Haber-Weiss reaction. Hydroxyl radical,
too, could be the source of part of the damage that has been attributed to O2-.
However, the Haber-Weiss reaction as written is exceedingly slow, with a reaction
rate of 3.0 M-1s-1 74 at pH 7.3. Chelated iron, on the other hand, which is frequently
present in biological fluids, catalyzes this reaction, increasing the reaction rate to a
significant degree8,75. Iron is chelated in vivo with phosphate esters such as ADP,
ATP, GTP, and pyrophosphate76, and other compounds such as citrate and
oxalate77. In vitro chelators such as EDTA and nitrilotriacetic acid (NTA) also allow
iron to catalyze the reaction. In contrast, chelators that prevent iron catalysis of
reaction (2) are diethylenetriaminepentaacetic acid (DTPA),78 bathophenanthroline
sulphonate (BPS),79 phenanthroline,79 and bipyridine.80 The reaction for the iron
catalyzed Haber-Weiss reaction is as shown in equations (3) and (4).44
Fe+++ + O2- --> Fe++ + O2 (3)
Fe++ + H2O2 --> Fe+++ + OH- + OH. (4)
The rate constant for equation (3) when the iron is chelated with EDTA has been
determined to be k2 = 1.3 x 106 M-1s-1 at pH 7.0.35 The rate constant for equation
(4), which is known as the Fenton reaction, for EDTA-chelated iron is given as
k3 = 104 M-1s-1 at pH 7.081. Thus, the catalytic properties of iron make the
Haber-Weiss reaction quite possible in vivo and the hydroxyl free radical (OH.) is
probably formed in vitro and in vivo mainly through the iron-catalyzed Haber-Weiss
reaction (reactions 3 and 4).
Evidence for the Haber-Weiss reaction
Evidence for the occurrence of the iron-catalyzed Haber-Weiss reaction in
vitro includes the observation that more OH. is detected from a xanthine/xanthine
oxidase system when more iron or iron-EDTA is present, but that the iron-chelator
DTPA, in suppressing the formation of the OH. signal, maintains the presence of the
O2- signal. This was shown by using electron spin resonance to detect products of
the reaction of OH. with compounds known as spin traps, such as
5,5'-dimethylpyrroline-N-oxide (DMPO).82 The hydroxylation of salicylate was
used as an assay for OH. formed by the xanthine/xanthine oxidase system, relying
on the ability of OH. to hydroxylate salicylate and form o-diphenols. Hyaluronic
acid's depolymerization was found to be inhibited by inhibitors of the production or
continued presence of OH., namely the iron chelators BPS and DTPA, the enzymes
catalase and SOD, and the OH. scavengers mannitol and formate.79 The in vivo
formation of single strand breaks in DNA by H2O2 was concluded to be mediated by
the Haber-Weiss reaction when phenanthroline and bipyridine, strong chelators of
iron, were found to protect the DNA from breakage by H2O2 in human fibroblasts.80
DNA binds iron (II) in a manner such that OH. formation from H2O2 is enhanced as
compared to the amount formed without DNA83. DNA strand scission in a system
containing xanthine and xanthine oxidase increased as a function of iron
concentration. Strand scission was stimulated by EDTA and inhibited by DTPA, the
OH. scavengers mannitol and formate and the enzymes SOD and catalase.49 The
iron carrying enzymes lactoferrin and transferrin have been found to catalyze the
Haber-Weiss reaction when fully loaded with iron84, but lactoferrin that is not fully
loaded with iron was found to inhibit the reaction by inhibiting iron catalysis85.
Reduction of paraquat has been shown by gas chromatography to result in the
production of OH.86.
Other chemical sources of hydroxyl free radical
Non-Haber-Weiss production of OH. is also possible. The antitumor drug
tallysomycin produces OH. without the production of intermediate O2- or H2O2.87
The cardiotoxic antitumor drug adriamycin semiquinone reacts with H2O2, requiring
neither metal catalyst nor O2- for the production of OH.88. The Fenton reaction
produces OH. whenever Fe++ and H2O2 are present together.
Biological sources of hydroxyl free radical
Biological sources of OH. include liver microsomes89, human neutrophils,90
and human monocytes during the phagocytosis of opsonized zymosan.90
Interestingly, monocytes from a patient with chronic granulomatosis disease, a
hereditary disorder in which the patient suffers persistent and multiple infections,
failed during phagocytosis to generate the ethylene gas from methional used as an
assay for OH. in this experiment.91
Reactivity of hydroxyl free radical
The hydroxyl radical is extremely reactive with a wide variety of compounds.
Second-order reaction rates are commonly on the order of 108 to 109 M-1s-1 92; the
reaction rate of OH. with the enzyme catalase is 2.6 x 1011 M-1s-1. As would be
expected from its extreme reactivity, the OH. is capable of causing extensive and
indiscriminate biological damage. It has been found to degrade DNA,67 form
crosslinks between DNA molecules and between DNA and proteins,93 oxidize
thymidine,94 oxidize and hydroxylate amino acids,92 kill bacterial cells,95 and
mediate the induction of diabetes in mice by alloxan.96,97 Hydroxyl radical also
mediates the degradation of the wood polymer lignin by the fungus Phanerochaete
chrysosporium.98 Toxic oxygen species such as OH. may be involved in the normal
senescence of leaf tissue.99
Hydroxyl radical scavengers
Scavengers of OH. are compounds that react with OH. as rapidly as or even
more rapidly than other compounds, and which thus, when added to a model
systemor biological system in high enough concentrations, prevent other OH.
-mediated reactions from occurring. To be suitable for this purpose, compounds
must not be especially reactive with substances other than the OH.; furthermore, the
products of their reaction with OH. must not be particularly damaging in themselves.
Substances commonly used as OH. scavengers include ethanol, mannitol,
formate, benzoate, thiourea, and dimethylsulfoxide
Table I. Bimolecular Rate Constants for Reactions Between
Scavengers
and the Hydroxyl Radical and Singlet Oxygen.
_____________________________________________________________
Scavenger Rate Constant (M-1sec-1)
_____________________________________________________________
Hydroxyl Radical92 Singlet Oxygen100,101
_____________________________________________________________
Sodium Azide 7.5 x 109 5 x 108
Dimethylsulfoxide 7.0 x 109 ~ 103
Sodium Benzoate 6.0 x 109 3 x 103
Thiourea 5.0 x 109 8 x 105
Sodium Formate 2.8 x 109 ~103
Ethanol 1.5 x 109 ~103
Urea ~106 ~103
_____________________________________________________________
(DMSO). The latter two scavengers are particularly useful for in vivo experiments,
as they are capable of crossing cell membranes. Table I gives a more complete list
of OH. scavengers, along with the rate constants of their reaction with OH.. Urea, a
poor OH. scavenger, is frequently added as a control; the rate of its reaction with OH.
is less than 7.0 X 105 M-1s-1. Some OH. scavengers, such as azide, also scavenge
1O2, but DMSO and ethanol only react with OH..
Oxygen toxicity summary
In conclusion, the poisonous properties of oxygen to all organisms is due to
the reactivity of oxygen species with many cellular components. Oxygen in the
molecular state is not very reactive. However, partially reduced forms of oxygen,
known as active oxygen species, are considerably more reactive. The toxicity of
oxygen is mediated by these species, which comprise O2-, H2O2, and OH.. The
active oxygen species are generated as byproducts of cellular metabolism, as well
as by the activity of toxic agents. The reduction of molecular oxygen to form the more
active species is effected by escaped electrons from the electron transport chain in
respiring cells, by the action of enzymes such as the microbicidal oxidases of
phagocytic cells, by the autoxidation of compounds such as hemoglobin,
catecholamines, and reduced flavins, and by the redox cycling properties of certain
herbicides and antibiotics.
Defenses against Oxygen toxicity
In vivo methods of studying the photodynamic effect can make great use of
manipulations of the cell's natural defenses. Discovering the necessary natural
defenses also yields information on what forms of damage and damaging agents
contribute to lethality.
Cells maintain a variety of defenses against oxygen toxicity. Among these are
an array of enzymes that have evolved to deal with oxidative stress, including
superoxide dismutase and catalase (described below).44 In animals, additional
defensive enzymes include methionine sulfoxide reductase, which repairs
methionine residues in proteins that have been damaged by OH., glutathione
peroxidase, and the glutathione reductase which regenerates the cofactor for the
glutathione peroxidase. Still other defensive enzymes are endonucleases,
exonucleases, and DNA polymerase, which repair single-stranded breaks and
modified bases in DNA caused by oxidative and other stresses. In eucaryotes, uric
acid, a-tocopherol, ascorbic acid, and b-carotene are among the compounds that
function in vivo to scavenge active oxygen species and other organic radicals
formed in the cell by reactions with the oxygen radicals; both procaryotes and
eucaryotes contain high levels of glutathione, a scavenger of OH. and 1O2.44
Superoxide dismutases
Superoxide dismutase (SOD) catalyzes the reaction between two molecules
of superoxide to form hydrogen peroxide and molecular oxygen (reaction 1,
repeated below):
O2- + O2- + 2 H+ ---> H2O2 + O2 (1)
Superoxide dismutases are ubiquitous in aerobic organisms.102 Eucaryotes
contain, typically, both a copper and zinc-containing superoxide dismutase
(CuZnSOD) and another SOD that contains manganese(MnSOD); the two are
unrelated, and the latter is compartmentalized in the mitochondria, while the former
is in the cytoplasm103. Bovine copper-zinc SOD has a molecular weight of
32,500104 and contains two atoms of copper and two atoms of zinc per molecule.105
It is composed of two identical subunits joined by at least one disulfide bond.105 The
human CuZnSOD is similar to the bovine CuZnSOD.106 The mechanism of action of
CuZnSOD involves alternate reduction and reoxidation of the Cu2+ at the active site
during successive interactions with O2-.102 The mitochondrial SOD, MnSOD,
contains four subunits, each with a molecular weight of 20,000; it is located in the
matrix of the chicken liver mitochondrion.102
Procaryotes such as E. coli contain both an MnSOD and an iron-containing
superoxide dismutase (FeSOD). The procaryotic MnSOD shows homology to the
eucaryotic enzyme but contains two subunits rather than four and is consequently
half as large, with a molecular weight of 40,000. The FeSOD is partially homologous
to the MnSOD.107 It has two subunits, one Fe3+ ion per molecule of enzyme, and a
molecular weight of 39,000. The mechanism of the Mn and FeSODs is probably
similar to that of the eucaryotic CuZnSOD.102 Superoxide dismutases increase the
spontaneous rate of O2- removal according to equation (1) from about 5 x 105
M-1s-1 at physiological pH44 to about 1.8 X 109 M-1s-1 for the MnSOD at pH 7.8 and
about 1.6 x 109 M-1s-1 for the CuZnSOD between pH 5.3 and 9.5.44
Regulation of SOD levels
Of the two E. coli superoxide dismutases, FeSOD is constitutively
expressed, while MnSOD is inducible. Studies on the activity of E. coli SODs
indicate that levels of the FeSOD can be increased somewhat with iron
supplementation, and decreased by iron chelators.108 Levels of the MnSOD are
increased more dramatically by chelating agents, Mn(II), paraquat108 or
oxygenation, and decreased by anaerobic conditions. It has been proposed that
oxygenation and intracellular O2- production induce MnSOD because O2- oxidizes
Mn(II) to Mn(III). Mn(II) has a higher affinity for the manganese-free apoMnSOD,
and successfully competes with Fe(II) that would otherwise combine with the
apoMnSOD to produce an inactive enzyme. This results in increasing MnSOD
levels.108 Maximum levels of total SOD in E. coli are induced by growth in minimal
media that has been supplemented by a combination of 100mM Fe(II), 100 mM
Mn(II), 10 mM paraquat, and 100 mM 8-hydroxyquinoline; under these conditions,
as much as ten percent of the total cellular protein is superoxide dismutase.
Compounds that are capable of entering cells and inducing superoxide
dismutase include paraquat (methyl viologen), pyocyanine, phenazine
methosulfate, streptonigrin, juglone, menadione, methylene blue, and azure c, all
known redox-cycling compounds.109
Low levels of the MnSOD are characteristically seen when cells have been
grown in a glucose minimal medium. High levels of SOD are produced by growth in
TSY, or by growth in minimal medium to which 100 mm iron(II), 100 mm
manganese(II), 100 mm 8-hydroxyquinoline, and 10 mm paraquat have been added;
levels are enhanced thirty-fold by the latter treatment.108 Similar induction levels are
observed by the addition of paraquat and manganese alone at similar levels.
Catalase
The catalases belong to the family of enzymes which contains the
hydroperoxidases and peroxidases. Catalase catalyzes the reaction:
2 H2O2 ---> 2 H2O + O2, (5)
whereas peroxidase catalyzes the reaction:
AH2 + H2O2 ---> A + 2 H2O (6)
where A is an electron donor such as dianisidine or guiacol.110
There are two main enzymes with catalase activity in E. coli; one, hydroperoxidase
II (HPII), possesses only catalatic activity. The other enzyme, which migrates more
slowly than HPII during electrophoresis, hydroperoxidase I (HPI), is both a catalase
and a broad spectrum peroxidase. When electrophoresed on a polyacrylamide gel,
HPI can be visualized as two isozymes, HPI-A and HPI-B.111 Peroxidase (HPI) has
a molecular weight of 337,000, is composed of four subunits of equal size, and
contains two molecules of protoheme IX per tetramer.110 Catalase (HPII) has a
molecular weight of 240,000 and consists of four polypeptides, each of which is
associated with a ferric protoporphyrin IX.112
Factors that affect expression of hydroperoxidases
Catalase is induced by peroxide and ascorbate.113 It is repressed by glucose
in a classical example of catabolite repression.114 Catalase is important in
protecting the cell against ultraviolet lethality; in fact, it was some time after the
discovery of nur, a gene named for its conferring resistance to near-ultraviolet light,
that it was found to code for a hydroperoxidase.115 Exogenous catalase provides
protection against lethality due to acridine orange as well as to phenazine
methosulfate.109 Catalase is inducible by treatment of log-phase cells for thirty
minutes with 500 mM H2O2, or with 5 mM ascorbate, which gives rise to H2O2;
catalase levels are enhanced as much as six-fold by this treatment116 and twenty-
to thirty-fold by treatment with phenazine methosulfate and pyocyanine.109
Low levels of catalase are observed in cells that have been grown in glucose
minimal medium; levels four- to five-fold higher are seen when cells have been
exposed for thirty minutes to 0.5 mm hydrogen peroxide or 5 mm ascorbic acid.111
The increase in catalase level is due to the induction of hydroperoxidase I, the broad
spectrum peroxidase of E. coli. Enhancement of intracellular catalase is also seen
when cells are pretreated with paraquat and manganese117.
The oxyR regulon
Christman et al.118 have discovered a regulatory element, oxyR, which
appears to regulate a group of enzymes that defend against oxidative stress,
particularly HPI. Constitutive mutants in oxyR overproduce nine of the proteins that
are induced by oxidative stress, including HPI-A and HPI-B, MnSOD, and
glutathione reductase.
Cloning of superoxide dismutase and catalase genes
The genes for three of the E. coli enzymes of defense against oxidative
stress, HPI (katG), MnSOD (sodA), and FeSOD (sodB) have been cloned onto
multicopy plasmids. Loewen et al.116 screened the Clarke and Carbon E. coli colony
bank by plating out the colonies and looking for high catalase levels, as evidenced
by an increase in the colony's bubbling in response to H2O2. One strain,
JA200/pLC36-19, proved to have a basal catalase level about twice the normal
basal level.116 Loewen and Triggs have subcloned this gene to form plasmids BT22
and BT28, which lead to an increase in catalase and peroxidase levels of twenty to
thirty-fold, compared to the parent strains.119 Nettleton et al.120 identified two
clones in the Clarke and Carbon bank that contained an extra copy of the iron SOD
gene, using a synthetic oligonucleotide probe constructed from information on the
amino acid sequence of the protein. They found that the probe binds exclusively
to a fragment contained in common by clones JA200/pLC18-11 and
JA200/pLC13-47.120 Sakamoto and Touati also cloned the FeSOD gene, using
immunoprecipitation of crude extracts to screen a cosmid bank they had
constructed121; they found that their cosmid contained a 6.6 kb Pst I fragment which
carried sodB. Touati similarly cloned and mapped the MnSOD gene from the same
cosmid bank and subcloned it into plasmid DT1-5; crude extracts of a strain
containing this plasmid contain four to five times as much superoxide dismutase as
the parent strain when grown in rich media to late log phase.122
DNA repair
The defensive enzymes superoxide dismutase, hydroperoxidase, and
catalase act to remove toxic oxygen species before they can do intracellular
damage. In the instance that these defenses fail, other systems are ready to repair
damage, particularly DNA damage, that is done by the toxic oxygen species.
Kinds of DNA damage
The existence of stress agents such as ultraviolet light, chemical mutagens,
intracellular oxygen or organic radicals that damage DNA requires cells to maintain
enzymes to repair damaged DNA. DNA damage may include not only single- and
double-stranded breaks, but also the production of damaged bases, such as
thymine dimers or guanine residues that have been depurinated or hydroxylated,
resulting in alkali-labile phosphodiester bonds.
Kinds of DNA damage that happens under the influence of light include
altered or missing bases, single-strand breaks, double-strand breaks, and
cross-linking. At least three kinds of damage are commonly defined: far ultraviolet
kill, which seems mainly to result from the formation of thymine dimers;
chemically-assisted ultraviolet kill (with psoralens for example), which results in
DNA cross-linking; and the kind of kill seen as a result of exposure to near ultraviolet
light or visible light in the presence of dyes, photodynamic kill, which includes
modified bases, apurinic/apyrimidinic sites, and single stranded breaks. The
differences in kinds of damage produced by these different forms of light were
suggested by their being protected against by different repair systems.
Apurinic/apyrimidinic (AP) sites are generated spontaneously and by
treatment with acid or as a result of chemical alkylation of deoxyguanosine, which
weakens the N-glycosidic bond that attaches the base to the sugar-phosphate
backbone of the DNA.123 They are also generated by DNA glycosylases, which
recognize and cleave N-glycosidic bonds of damaged nucleotide residues in
DNA.124
When the DNA backbone is broken by ionizing radiation or hydrogen
peroxide, the ends contain 3' terminal phosphoglycolaldehydes125 or thymine
glycols,126 or 3' terminal phosphates,125 which cannot act as substrates for DNA
polymerase I or DNA ligase. These blocked 3' ends must be removed before repair
of the DNA can take place, starting with the 3'-OH group of the primer DNA.127
DNA repair enzymes
As E. coli is the best studied of all organisms with respect to DNA repair,128
this discussion will concentrate on the repair systems in this organism. There is
some redundancy of the DNA repair systems, which allow mutants in any one of the
enzymes to function reasonably well under normal physiological conditons; this
indicates their importance to the survival of the organism. Some strains of E. coli are
more sensitive to photodynamic kill than others; E. coli B/s is ten times more
sensitive than E. coli K12 to photodynamic and paraquat toxicity.129 It is possible
that the responsible difference between the two strains involves differential DNA
repair ability.
DNA repair systems in E. coli include photoreactivation repair, excision
repair, recombination repair, and SOS repair. Photoreactivation involves the
cleavage of thymine dimers into a pair of normal thymines by photolyase, an enzyme
activated by visible light.130
Excision repair is a process in which two cuts are made in the
sugar-phosphate backbone of the DNA on either side of a distortion caused either
by a thymine dimer or by a base with which no nucleoside can form a base-pair,
producing a 3'-OH group on the 5' side. This allows DNA polymerase I to synthesize
a new strand, displacing the defective strand; DNA ligase joins the new strand to the
original strand. The original incisions are made in E. coli by endonuclease I. 130
Recombination repair, also called daughter-strand gap repair, involves
by-passing a block to DNA polymerase III, allowing replication to continue, leaving a
gap opposite the block (e.g., a thymine dimer). The unpaired gaps are filled by
excising the homologous piece of undamaged sister strand and inserting it into the
gap. DNA polymerase I and DNA ligase then join the inserted piece to adjacent
regions and fill in the gap left in the donor segment.130
In SOS repair, binding of the recombination A (RecA) gene product to
single-stranded DNA in the region of a thymine dimer or other distortion allows the
insertion of adenines or mismatched bases by DNA polymerase III. The RecA
protein also induces a series of other proteins, called din for damage inducible,
which are kept repressed by the LexA protein in the absence of single-stranded
DNA.130 A model of the SOS regulatory system is shown in figure 2. A fusion of one
of the genes for a din protein and the coding region for the ß-galactosidase gene has
been made, allowing the induction of the SOS response to be easily studied by
assaying for ß-galactosidase levels after submitting the cells to an appropriate
inducing stress agent, such as mitomycin c.131 It was thought at one point that the
protein rec A protected against both far and near ultraviolet and photodynamic
killing in E. coli;132 later studies showed that rec A did not protect at all against
photodynamic killing, while nur showed great protection.115
Figure 2. Model of the SOS regulatory system. The open circles represent
proteolytically inactive RecA molecules and closed circles represent proteolytically
active RecA molecules. The semicircles represent LexA molecules. (From G.
Walker, Microbiol. Rev. 48: 60-93.)
Endonucleases
The phosphodiester bonds at apurinic/apyrimidinic sites are cleaved by AP
endonucleases as the first step in excision repair. These AP endonucleases have
been found to be ubiquitous in both procaryotes and eucaryotes.133 At least four AP
endonucleases have been identified in E. coli.123 These include exonuclease III ,
endonuclease III, endonuclease IV, and endonuclease V; little is known about endo
V at this time.134 AP sites are not the only kinds of lesions that require repair by
these enzymes. Single-stranded breaks must have a free terminal 3' -OH group in
order for DNA polymerase I to initiate repair synthesis, so any ends which contain a
phosphoglycolaldehyde end, a terminal phosphate, or otherwise lack a 3'-OH
group, must be cut out using an exonuclease, before repair synthesis can occur.130
Those enzymes which have this ability include endonuclease III, endonuclease IV,
and exonuclease III.134
Exonuclease III
Exonuclease III is the major AP endonuclease in E. coli, with about 85% of E.
coli's measurable AP endonuclease activity.133 Its functions include cutting at AP
sites, acting as a 3' to 5' exonuclease of double stranded DNA, cutting at the 5' side
of a urea residue in DNA, removing 3' phosphates and phosphoglycolate esters, and
hydrolyzing the RNA from a hybrid duplex in the 3' to 5' direction.134 Exonuclease III
mutants are hypersensitive to H2O2 and to near ultraviolet light, both of which
produce AP sites.115 Both exo III and endo IV can efficiently remove 3' terminal
phosphoglycolaldehyde from DNA.127
Endonuclease III
Endonuclease III cleaves 3' to AP sites and ring-damaged pyrimidines,
especially hydroxylated thymine residues (thymine glycols), generating 3'-OH and
5'-P termini.135 It has no cofactor, so it is active in the presence of EDTA.134 Endo III
does not remove terminal PO or phosphoglycoaldehyde. The gene for
endonuclease III, nth, has been cloned onto plasmid pRPC53.135
Endonuclease IV
In E. coli, endonuclease IV (endo IV) is a nuclease for AP sites; it catalyzes
the cleavage of a phosphodiester bond 5' to an AP site, and has no other known
activity.134 It is resistant to EDTA. Both exo III and endo IV can efficiently remove 3'
terminal phosphoglycoaldehyde from DNA.127 Endo IV is induced in E. coli by
paraquat, which is also known to induce SOD.133 Endo IV was identified as an
EDTA-resistant endonuclease activity. Less than 10% of measurable AP
endonuclease activity is due to endo IV in E. coli grown aerobically in rich medium, in
the absence of an inducing agent such as paraquat.135 Strains mutant in endo IV
have an increased sensitivity to alkylating agents and oxidants such as peroxides
and bleomycin.134 The gene for endo IV has been cloned into plasmids pLC38-27
and pWB21.135
Uses of Defense and Repair Enzymes
Defense and repair enzymes are useful in the study of damage produced by
active oxygen in two ways. First, the existence of the enzymes shows what the cell
finds damaging; if a substance is not harmful, the enzyme is unlikely to exist solely to
remove it. The discovery of SOD was what led to the proposal that superoxide is
harmful.136 Secondly, levels of the enzymes can be manipulated, either by induction
under appropriate stress conditions, or by subcloning of their genes; if increased
concentrations of an enzyme are protective, then the enzyme's substrate must
somehow mediate damage, and this substrate is likely to be produced during a
stress episode.
The gene for the E. coli HPI has been placed onto the plasmids pBT22 and
pBT28 by Triggs-Raine and Loewen.119 The gene for the E. coli MnSOD was placed
onto the plasmid pDT1-5 by Danielle Touati.122 Bernard Weiss's group placed the
endonuclease IV (nfo) gene on a plasmid, pWB21.135 All three of these plasmids
were used in this study.
The Photodynamic Effect
Oxygen-dependent toxicity that also requires the presence of light and a
sensitizing compound is known as the photodynamic effect.137 It was first described
by Raab in 1896.137 It affects all aerobic eucaryotic and procaryotic organisms and
has been the subject of thousands of studies. The mechanism of the photodynamic
effect remains mysterious138. Visible light is not absorbed by most cellular
components, and therefore does not usually have the toxic effects associated with
X-rays or ultraviolet light. However, a sensitizer that absorbs visible light-a colored
substance, or dye-may serve to transfer the energy of the light absorbed, usually
through oxidation or reduction reactions, to a species which will then cause damage
to cellular components. The wavelengths of light and their activities, as well as the
spectra of representative photosensitizing dyes, are shown in figure 3.
Photosensitizers are molecules which can absorb light to produce a chemical
reaction which would not occur in their absence.139
Biological examples of photosensitization
There are many instances of reactions for which this effect is responsible. For
example, skin diseases of phototoxicity in humans,140 livestock, and experimental
animals; light-activated pesticides;141 and medical treatments including
photochemotherapy142 and phototherapy of jaundice,143 herpes simplex144 and
psoriasis.145 Studies have shown that photosensitizers can cause cell death,146
DNA damage,147,148,149 protein damage,150,151 membrane damage,148
mutagenesis,146 and both tumor destruction152 and possibly carcinogenesis.153 It
is thought that photodynamic action is responsible for the
Figure 3. The phototoxic spectrum.
conditions that bring about the end of the toxic red tide, as photosensitizing
compounds are released by
the deaths of creatures killed by the red tide bloom.154
Examples of photosensitizers
Among the chemicals known or suspected to induce skin phototoxicity are
sweeteners, both cyclamates and saccharin,155 cosmetics including perfumes156
and lipsticks,157 germicidal soaps,157 industrial products including pitch, liquor picis
detergents, and crude coal tar,140 agricultural products such as celery,158 and
hypericin, which is found in a toxic agricultural pest eaten by livestock, St. John's
wort.159 Other inadvertent inducers of skin phototoxicity are drugs such as the
phenothiazines, which are prescribed as antipsychotics;160,161
tetracyclines;162,163,164 sulfonamides, which were the first drugs recognized as
having phototoxic side effects;165,166,167 psoralens;168,169 chlordiazepoxide, a
tranquilizer that can cause liver damage due to photoproducts formed in the skin;170
phenazines, drugs used for the treatment of psychotic disorders;161 nalidixic acid171,
an antibiotic used in the treatment of urinary tract infections;145 methotrexate, an
anti-cancer drug;172 and benoxaprofen, an antiinflammatory and analgesic
drug.173 Phototoxicity is also observed in diseases that cause the production or
accumulation of photosensitizing metabolites, such as hyperbilrubinemia.174
Despite all of these known phototoxic activities, the underlying mechanisms remain
poorly understood.138
Phototherapy
The photodynamic effect is not restricted to deleterious effects, but may be
used therapeutically. Niels Finsen was awarded a Nobel prize in 1903 for the
treatment of lupus vulgaris, facial lesions seen in tubercular patients, with a carbon
arc lamp.175 Phototherapy has long been used in the treatment of psoriasis, with the
application of photosensitizers such as coal tar followed by ultraviolet light
exposure.145 Partially refined versions of coal tar are available on prescription under
a variety of proprietary names.145 Phototherapy with neutral red144 and
proflavin176,177 has been used in the treatment of herpes simplex, although not
without harmful side effects.153,178 Phototherapy has also been used in the
treatment of idiopathic vitiligo, seborrhea, eczema, and chronic lichen simplex.145
Phototherapy of cancers
A major recent impetus to the study of photodynamic effect has been the
potential use of dye photosensitizers in phototherapy of cancers.142 Leukemic cells
show greater sensitivity to phototoxicity than normal cells. Tumors in mice treated
with photochemotherapy have shown regression of greater than 90%.179
Derivatives of hematoporphorin are at present undergoing extensive clinical trials
as sensitizers for photodynamic therapy of neoplasms180. At the present time one of
the major problems with photochemotherapy is the current poor understanding of
the basic mechanisms involved.152
Photosensitizers in this study
The classes of photosensitizers considered in this study include: acridines;
other synthetic dyes including thiazines, xanthenes, and phenazines; porphyrins,
and a naphthalimide (see figure 4, structures of dyes). The extinction coefficients,
given here parenthetically where known, are from Houba-
Figure 4. Structure of dyes.
Herrin (1982).181 The acridines studied include acridine orange (wavelength
maximum 492 nm, extinction coefficient 41,600 M-1cm-1), acridine yellow
(wavelength maximum 442 nm, extinction coefficient 29,000 M-1cm-1), and
proflavin (wavelength maximum, 444; extinction coefficient, 34,300 M-1cm-1).
Xanthenes included were fluorescein (wavelength maximum 496 nm) and rose
bengal (wavelength maximum 550 nm, extinction coefficient 99,800 M-1cm-1),
which possesses one of the highest absorption coefficients known.182 The
phenazine studied was neutral red (wavelength maximum 540 nm). Thiazines
included methylene blue (wavelength maximum 665 nm, extinction coefficient
78,000 M-1cm-1), azure c (wavelength maximum 616 nm), thionin (wavelength
maximum 598 nm), and toluidine blue (wavelength maximum 626 nm).
Many, but not all, of these dyes are capable of penetrating into the procaryotic
cell. Those dyes which cannot enter the cell will be able to damage only the cell
membrane, which can itself cause cell death; however, in the absence of a
reductant, dyes that cannot enter the cell are not lethal.183 The porin channels in the
outer membrane of E. coli freely allow the passage of small positively charged
molecules, but keep out large, hydrophobic or negatively charged compounds184
in order to exclude the hydrophobic and anionic bile salts that are a threat to E.
coli in its natural habitat.185 In the drawing in figure 4, the thiazines and the
phenazine have positive charges, while the other dyes are neutral. In experiments
by Martin and Logsdon (1987), the dyes toluidine blue and acridine orange
penetrated the cell, while fluorescein and lucifer yellow did not.183 Neutral red and
methylene blue are accumulated in E. coli at neutral pH, but rose bengal only at
acidic pH.186 In general, the
acridines, thiazines, and phenazine enter the cell, whereas the xanthenes and the
napthalimide lucifer yellow do not.
Most of the dyes used in this study intercalate in between the bases of DNA.
The acridines, thiazines, and phenazines all have planar, tricyclic structures, as
seen in figure 3; however, the xanthenes are bulkier and cannot fit in between the
bases of the DNA. In addition, the xanthines are negatively charged and are
probably repelled by the sugar-phosphate backbone of the DNA. Xanthenes, unlike
acridines, thiazines, and phenazines, do not intercalate into DNA. Since
intercalating dyes will be largely intercalated into the DNA once they enter the cell, it
is important to study dye reactions when the dyes are in the intercalated state, as
well as alone.
Why these dyes are included in this study
These organic dyes have been chosen for this study because they are the
most widely studied compounds in photochemistry and the physiology of
photodynamic action. Although they vary widely in their structures they are all
capable of oxidation reduction reactions, and, because of this fact, of giving rise to
oxygen radicals.183 Most of them are known to generate singlet oxygen (1O2) as
well.146 Extensive studies have been done illustrating the effects of these particular
dyes in E. coli, before the significance of O2- was known. Many of these dyes are
structurally related to known redox active compounds such as pyocyanine and
phenazine methosulfate, which are known to give rise to O2-by redox-cycling within
E. coli.187 Reasons why so many structural classes of dyes were considered in this
study were in order to determine whether their individual toxicities could be
explained by a common mechanism; to determine the relative importance of cell
localization in determining their toxic effects; and to illustrate that all forms of light,
from near ultraviolet through the red, are sufficiently energetic to promote phototoxic
effects in the presence of an appropriate photosensitizing agent and to do so
through the generation of reduced oxygen species.183
Uses of photosensitizers
The dyes used in this study have a variety of functions and origins. Some
acridines occur in coal tar.175 Acridine orange and acridine yellow are used as
adjuncts to laser surgery of cells, and as stains for nucleic acids,188,189 while
proflavin is used as a topical antiseptic.190 Porphyrins occur as metabolic
byproducts, and are used in photochemotherapeutic agents.180 Fluorescein (D & C
Yellow No. 7) is commonly used to test for corneal abrasion.190 Rose bengal has
become one of the most widely used of all photodynamic sensitizers.191 Rose
bengal is used as a biological stain, as a dye for straw and wood chips, as an
ingredient in inks, for coloring edible products, and in cosmetics.190 Related
xanthenes make up a number of dyes approved by the FDA for use in drugs and
foods.192 Neutral red is used as a pH indicator and as a biological stain in Golgi
apparatus studies190 and is related structurally to phenazine methosulfate and
pyocyanine, potent, known redox cycling compounds.187 Methylene blue is used as
a stain in bacteriology, as an antimethemoglobinemic and cyanide antidote in
humans, and as an antiseptic and disinfectant in veterinary use. Eye drops
containing methylene blue are available over the counter in some countries.193
Thionine is used in general nuclear staining, in the counting of bacteria in milk, and
as an antioxidant for linseed oil.190 Azure c and toluidine blue o are used as biological
stains.
Organisms affected by the photodynamic effect
The toxicity of photosensitizers is nearly universal; among those organisms
sensitive to inactivation by the photodynamic effect are viruses, bacteriophage,
bacteria, protista, yeasts, insects, and cultured mammalian cells.
Damage to viruses and bacteriophage by the photodynamic effect
Viruses and bacteriophage are susceptible to inactivation by the
photodynamic effect. Phage particles are inactivated by light plus methylene
blue,194 toluidine blue,195 and proflavin.196 Other isolated viruses, including
adenovirus, vaccinia virus, SV40 papovavirus, influenza A, parainfluenza-3,
measles virus, reovirus 1, and herpesvirus are all inactivated by toluidine blue o.
Neutral red inactivated measles virus, and herpesvirus, while proflavin inactivated
reovirus 1, influenza A, measles virus, and herpesvirus.197
Damage to bacteria by the photodynamic effect
Experiments involving the exposure of bacteria to photosensitizing dyes
have shown them to be killed as a result. Rose bengal immobilized on polystyrene
beads was able to kill E. coli upon exposure to light.198 Toluidine blue killed E. coli
cells.199 Thiazines, xanthenes, phenazines, acridines, and naphthalimides that kill
E. coli include azure c and thionin;200 rose bengal200 and fluorescein;183 neutral
red;200 proflavin,200 acridine yellow,183 and acridine orange;183 and lucifer yellow.183
Other bacteria suffer similarly; for example, methylene blue inactivated
Acholeplasma laidlawii201 and toluidine blue inactivated Micrococcus roseus.202
Damage to yeast cells by the photodynamic effect
Yeasts have been used extensively to study photodynamic lethality. Acridine
orange in combination with 410-600 nm light inactivates yeast. The inactivation
curves are the same shape as for X-ray and FUV killing of yeast, suggesting that
there is at least some overlap between the targets and sites of these three forms of
radiation. Ascorbate, a biological reductant, caused an oxygen-dependent
enhanced lethality in yeast cells of the thiazines toluidine blue, methylene blue, and
thionin, but not the xanthenes rose bengal and eosin yellow, though it had no
lethality in absence of dye.203 Proflavin led to inactivation of yeast.146 Rose bengal,
thionin, methylene blue, toluidine blue, acridine orange, proflavin, and acriflavin all
inactivated yeast cells.146
Damage to single-celled eucaryotes by the photodynamic effect
Other eucaryotic cells have been studied for sensitivity to photodynamic
effects, as well. Paramecia are killed by acridine in the light, but not in the dark;
Raab's discovery of this effect led to the first published example of the photodynamic
effect, in 1900.137 Hematoporphorin derivative photosensitizes DNA strand
breakage in Chinese hamster ovary cells.147 The crustacean Artemia salina was
immobilized by the photodynamic effect of polycyclic aromatic hydrocarbons.204
The activity of nerve cells is altered by acridine orange plus light,205 as well as by
fluorescein206 and lucifer yellow207 while erythrocytes are lysed when exposed to
dyes such as rose bengal, methylene blue, and neutral red plus light of the
appropriate wavelength.137
Other systems susceptible to the photodynamic effect
Other systems in which damage due to photosensitization occurs include
multicellular organisms, both plants and animals. For example, strong illumination
with blue and ultraviolet light photolyses rhodopsin, leading to the ultrastructural
degeneration of the photoreceptors in the Drosophila eye.208,209 Rose bengal
caused breakdown of chlorophyll and carotenoids and a decrease in CO2 uptake in
leaves.210 Chlorophyll is among the most effective of photosensitizers for the
oxygenation of organic substances.211
Human photoreactions
Some of the dyes used in this studied have been observed for their effects on
human skin. Neutral red has been reported to cause acute contact dermatitis in
humans.177 Rose bengal, when injected into the skin before light exposure, causes
erythema and edema.166 Fluorescein and eosin, applied to the skin and then rubbed
off before sun exposure, resulted in whealing.137 Not much is known about the
mechanisms of action of photodermatoses.212
Cellular targets of photodynamic damage
The targets of photodynamic damage in the cell have been examined, both in
vivo and in vitro, using isolated components. Subcellular components including
mitochondria and cell membranes are degraded by photosensitizers plus light.
Mitochondria that are exposed to light plus hematoporphyrin or hematoporphyrin
derivative (HpD) successively lose coupling, calcium ion transport and
respiration.213 Isolated chloroplasts lose their ability to carry out the Hill reaction
(photochemical water-splitting process) following photodynamic treatment.214
Acridine plus ultraviolet light damages cell membranes.148 The photodynamic
treatment with rose bengal of ribosomes isolated from E. coli causes a rapid loss of
their ability to incorporate amino acids into polypeptides.215 Photodynamic damage
to stained membranes, which occurs with many dyes in the presence of oxygen and
intense illumination, is not well understood and is thought to involve production of
free radicals by 1O2.216
Damage to proteins by the photodynamic effect
Among those compounds which have been found to be degraded in vitro are
amino acids, such as tryptophan,which is destroyed by methylene blue in the
presence of light and oxygen.217 Photosensitization has been shown to degrade
proteins,150,151 denaturing proteins such as egg albumin137 and inactivating enzymes
such as catalase and invertase.137
Damage to DNA by the photodynamic effect
DNA has been determined to be a major target of the photodynamic damage.
From an analysis of yeast survival curves, it was concluded that the principle
damage is probably mediated through the DNA.218 Direct evidence of DNA damage
in vitro has included loss of viscosity, single stranded breaks, apurination,
dimerization, and loss of ability to function as a template for DNA polymerase. The
viscosity of calf thymus DNA is reduced by rose bengal and methylene blue in the
light.219 Acridine plus ultraviolet light damages DNA., causing it to sediment more
slowly.148 Acridine orange plus light causes lowered viscosity, a decreased
sedimentation coefficient, and a decrease in the thermal denaturation temperature
of salmon sperm DNA. This is due to depolymerization, as shown by CsCl banding
experiments; depolymerization is due mainly to single-stranded scission. The
damage is maximized when the acridine is bound to DNA; this implies the formation
of a short-lived intermediate.218 Selective degradation of the guanine residues
might be the primary damage responsible for the formation of both alkali-labile
bonds and strand scissions.220 Acridine produces not only thymine dimers but also
other lesions.221 Acridine orange and methylene blue activate phage lambda in
lysogenic Escherichia coli, which indicates that DNA damage is occurring.222 DNA
synthesis is inhibited by acridine orange or methylene blue in the presence of light.
Dark reactivation repair cannot repair after damage by acridine orange plus light.222
In the light, methylene blue, rose bengal, eosin Y, thionin, azure c, toluidine blue o,
acridine orange, and neutral red all decreased melting temperature of DNA,
suggesting that the DNA had been degraded.149 Photosensitization has been
shown to destroy deoxyguanosine and deoxyadenosine.223
Proflavin plus light caused single stranded breaks in the bacteriophage
FX174, where any single-stranded break is a lethal lesion. The packaging of DNA
inside the phage head increases damage over isolated DNA.224 Strand breakage
has been demonstrated to be caused by acridine orange,218 methylene blue,219 and
ethidium bromide.225
Proflavin nicks inhibited template activity of DNA polymerase in FX174,223
indicating that the broken ends of the DNA do not contain a free 3'-OH group, which
is required for the initiation of repair synthesis of the broken strand. The loss of
infectivity of FX174 following photodynamic treatment with proflavin can be
explained by a block in DNA polymerase reaction; termination occurs one base
before a damaged guanine residue.226 The photodynamic reaction of methylene
blue with deoxyribonucleic acid led to the rapid loss of ultraviolet absorbance
accompanied by the uptake of one mole O2/mole derivative; the reaction occurred
with the guanine compounds, while thymine compounds reacted very slowly227.
Photodynamic reaction of methylene blue with DNA was more rapid with denatured
than with native DNA.228 Strand scission of DNA in vivo has been shown for
acridines with bacteria 229 and cultured mammalian cells.230 There is a peak in DNA
damage for human cells but not B. subtilis when irradiated at 450 nm even without
the addition of exogenous sensitizers, indicating that endogenous riboflavin
probably acts as a sensitizer in the human cells.223
Mutations caused by the photodynamic effect
Unsurprisingly, considering the evidence for photodynamic DNA damage,
mutations have been detected in photosensitizer-treated bacteriophage, bacteria,
yeast, and isolated DNA used in transformations.
Mutations in phage caused by the photodynamic effect
The addition and deletion frameshifts induced by proflavin in phage T4 were
used by Crick et al. to determine the nature of the genetic coding unit.231 Proflavin
caused a 2-fold increase in rate of mutation in phage T4; the mutations were of the
base-substitution type.232 The visible light treatment of bacteriophages sensitized
by acridines has been found to induce transitions and transversions.233
Mutations in bacteria caused by the photodynamic effect
Acridine orange causes mutations in bacteria.233,234 Neutral red plus light
results in mutations of the base-substitution type in Salmonella typhimurium.;153
methylene blue has the same effect in the same system.235 S. typhimurium Ames
strains TA 1535 (base substitution-sensitive) and TA1538 (frameshift
mutation-sensitive) were used to show the presence of base-substitution type
mutations caused by acridine orange; no frameshift mutations were seen in this
study.
Mutation in E. coli to resistance to bacteriophage T5 was induced by visible
light (>408 nm) and black light (300-400 nm), causing mutation rates to increase
more than 18-fold. In Escherichia coli B strain S, mutagenesis was produced by
acridine orange and proflavin, both of which are radiomimetic, i.e., they induce
cross-resistance to ultraviolet radiation and to each other; this suggests a similarity
in the mechanisms used to protect the cells against the two forms of damage. All
radiomimetics are mutagens in E. coli B.236 Acridines induced mutation in E. coli
tenfold greater than the normal rate in the absence of dyes.237
Mutation in yeast caused by the photodynamic effect
Rose bengal caused inactivation and gene conversion of yeast;238 as did
thionin; proflavin also led to gene conversion of yeast. Acridine orange, proflavin,
and acriflavin mutated yeast cells.146 Strong mutagenic action was noted for
acridine orange, acridine yellow, methylene blue, and toluidine blue in yeast, as well
as in phage and isolated transforming DNA.
Reactions of photosensitizing dyes
The toxic and damaging effects produced by the photosensitizers described
above require the presence of light and oxygen. The effect of the interaction of light
with a photosensitizer is to add energy to the sensitizer molecule. The energy taken
up by the sensitizer changes the energy level of one of its electrons,239 resulting in
the formation of an excited singlet state. The singlet state rapidly decays to a triplet
state, which has a longer lifespan than the singlet state.240 The primary
photochemical reactions of acridine, xanthene and thiazine dyes are the formation
of the singlet and then a longer-lived triplet state. For example, the triplet excited
state of fluorescein lasts 10-4 second, which is very long compared to the lifetime of
the initial singlet excited state, around 10-9 second.241 Secondary reactions include
a triplet-triplet electron-transfer process, leading to the semi-oxidized and
semi-reduced forms of the dye,242 or the interaction of triplet state dye with
oxidants, reductants, or energy transfer compounds (such as the interaction with
ground state oxygen (O2) to form singlet oxygen(1O2)). The ability of various dyes to
sensitize the photo-dynamic degradation of nucleic acids is correlated with their
ability to form electronically excited dye molecules in a metastable state, in which
case the nucleic acids act as reductants.149 The excited triplet-state sensitizer is
capable of reactions which the ground-state sensitizer would not take part in, or
which would occur much more slowly with the ground-state. Photoreduction of
excited dye molecules is frequently observed, and the reduced dye donates
electrons to other substances, returning to the oxidized state, from which the
process may be repeated.239
Type I and type II reactions of photosensitizers
The triplet state photosensitizer may react directly with a target molecule,
causing damage in an oxygen-independent oxidation or energy transfer reaction, or
react with an oxidizable substrate, gaining an electron (equation 8). Subsequently,
the semi-reduced dye reacts with dioxygen to yield O2- (eq. 9),
_____________________________________________________________
0dye + hv --> 3dye (7)
3dye + AH2 --> dye.- + AH. + H+ (8)
Type I
dye.- + O2 --> dye + O2- (9)
3dye + O2 --> dye + 1O2 (10)
Type II
3dye + O2 --> dye.+ + O2- (11)
1O2 + NADH --> O2- + NAD. (12)
_____________________________________________________________
Figure 5. Dye reactions. "AH2" represents a reductant; "3dye" represents a triplet
state dye.
which secondarily damages the target; in either case, the initial reaction with the
sensitizer is known as a Type I reaction (fig. 5). Alternatively, the triplet state
photosensitizer may be quenched directly by oxygen, yielding either 1O2 or the O2-
radical by electron transfer, in what is known as a Type II reaction (equations 10-11).
For most sensitizer triplets, even in the absence of a reaction with oxidizable
substrate, a small but significant percentage of interactions between the sensitizer
and oxygen result in electron transfer, which produces O2-243. Either of the two
dioxygen radicals, the singlet oxygen or the O2- anion, may then go on to react with a
target molecule. Superoxide, as previously indicated, may dismute to form H2O2 and
may ultimately give rise to the OH.. Both species exert deleterious effects in
biological systems.
Factors that influence whether type I or type II reactions occur
Whether the type I or the type II pathway will be followed depends on the
relative rates of reaction between the sensitizer and substrate244 and the sensitizer
and oxygen, since these reactants compete for triplet state dye molecules, which
are the common substrates of the two reaction pathways (fig. 5.). Also important are
the relative concentrations of oxygen and substrate. In aerobic cells, the
concentration of physiological reductants are quite high. For example, the
concentration of NAD(P)H is about 1.0 mM, and that of glutathione about 6.0 mM.
However, the level of oxygen is limited to 0.24 mM by solubility constraints and is
much lower than this in respiring cells. The rates of reaction between dye and
physiological substrates have not been determined for many of the dyes being
investigated in this study, but for those that have been, reactions proceed with rate
constants similar to that of the triplet state dye with O2. For example, while excited
triplet thionine reacts with dioxygen at a rate constant of 4.5 x 108 M-1s-1, it reacts with
tryptophan at a rate constant of 3.9 x 109 M-1s-1.245 The rate constant of the reaction
between excited methylene blue and oxygen is 3 x 109 M-1sec-1.246 The
semiquinone form of anthraquinone reacts with oxygen at rate constant of
108M-1sec-1.247
Substrate concentrations in vivo
If the rate constants of reaction between the triplet dye and singlet oxygen are
not very different from that of the reaction of triplet dye with an oxidizable substrate,
the factor that will determine which reaction predominates will be the concentration
of substrate versus that of oxygen. Oxygen is soluble in water to a concentration of
0.24 mM at 375C and atmospheric pressure; in a respiring cell, the level will be much
lower, as the oxygen is continually being used up by the electron transport chain.
Substrates under investigation in this study include NAD(P)H, whose in vivo level is
around 0.4 to 1 mM, and reduced glutathione, whose in vivo level is around 6 mM.248
Clearly, the relative concentration of substrates and oxygen in vivo would tend to
encourage type I reactions. Cadet et al. wrote in a recent review (1986) that "the
exact role and the relative contribution of these two competitive mechanisms [type I
versus type II] in the photodynamic effects remains to be determined at the cellular
level."223
Production of singlet oxygen
For many years attention has centered on the type II production of singlet
oxygen, which has frequently been suggested to be the major intermediate for
photodynamic actions both in vitro and in vivo. Each of the photosensitizing dyes
involved in this study is almost certainly capable of the production of singlet oxygen
under certain circumstances, especially in organic solvents where the solubility of
O2 is high and in which many of the photochemical studies have been carried out.
According to Ito,249 while the theory that singlet oxygen is a major intermediate in the
photodynamic effect has served as a valuable stimulant to investigations in the field
of photobiology, no photodynamic action so far investigated in vivo is solely
explained by the singlet O2 mechanism. Toluidine blue,250,251,252 methylene
blue,253 thionin,146 acridine yellow,254 proflavin,254 neutral red, and rose
bengal255,256 have all been claimed to produce singlet oxygen, while the damage
they produce has been claimed to be mediated entirely by this singlet oxygen.
Questionable arguments for singlet oxygen's mediation of photodynamic damage
In the general enthusiasm to determine that singlet oxygen is the major
intermediate for photodynamic action, some rather questionable evidence has been
pressed into service. Frequently, the shakiness of the argument is produced by the
apparent belief that there are only two possibilities to be selected between: the type
II reaction in which the only oxygen species produced is singlet oxygen, and a type I
reaction in which any damage that occurs is caused only by a direct reaction
between the triplet state dye and the target of damage. This scheme ignores
altogether the possibility of other oxygen intermediates, such as O2-, H2O2, or OH.
being involved. For example, one of the most widely used methods to demonstrate
that singlet oxygen is the sole mediator of damage consists of demonstrating that
the reaction is inhibited by the presence of azide. Unfortunately, the efficacy of azide
as an indicator of singlet oxygen mediated damage is compromised by the even
greater rate of reaction to be found between azide and OH.. While azide reacts with
singlet oxygen at a rate constant of k=2.2 x108 M-1sec-1,246 azide reacts with OH. fifty
times faster, at a rate constant of 1.08 x 1010 M-1sec-1.92 Moreover, none of the
techniques used to assay for singlet oxygen involvement in photodynamic effects,
including detection of luminescence, spin-trapping, and increasing reaction rate in
D2O, is completely specific for singlet oxygen.257
Reactions of azide
The mere inhibition of reaction by azide has frequently been produced as
`proof' that singlet oxygen alone is the mediator of phototoxicity. When azide
protected against killing by hematoporphorin, neutral red, methylene blue, rose
bengal, and fluorescein, it was taken as evidence of singlet oxygen involvement.154
Because azide reacts with singlet oxygen faster than it reacts directly to quench
many triplet-state dyes, such as methylene blue, inhibition by azide is explicitly used
as proof that the reaction proceeds by the type II pathway.246 Without confirmation
from other methods, azide inhibition was used as evidence that singlet oxygen
participates as a major intermediate in the photodynamic induction of genetic
changes in yeast by acridine orange.258 Even in a recent (1986) paper, Kraljic used
inhibition by sodium azide as sole proof that photooxidation occurs "exclusively or
predominantly via the singlet oxygen mechanism", "since the triplet state of HP does
not react measurably with azide in aerated solution."259
The deuterium oxide effect
The exacerbation of damaging effects of the dyes by the substitution of
deuterium oxide for water is often used as evidence that singlet oxygen mediates the
damage in question. Singlet oxygen in aqueous solution will be quenched (and
removed) by water molecules at a fast rate, 106 M-1sec-1 at a water concentration of
about 55 M. This quenching reaction is in direct competition with other reactions
involving singlet oxygen. However, deuterium oxide (D2O) reacts with singlet
oxygen much more slowly than water does. When deuterium oxide replaces water,
the rate of singlet removal by the quenching reaction is very low relative to the rate of
singlet reaction with other available targets. The solvent lifetime of singlet oxygen is
10 to 17 times longer in D2O than in H2O, thus allowing it more time to exert any
deleterious effects on targets.44, 260, 261
Lack of evidence for singlet oxygen's mediation of photodynamic damage
Frequently, other data has failed to confirm azide's demonstration of singlet
oxygen involvement in photodynamic action. As exacerbation of damage by D2O
was looked for and not conclusively seen, in 1977 Ito and Kobayashi were unable to
conclude that singlet oxygen was solely responsible for the toxic effects of
xanthenes, thiazines, and acridines in yeast cells, although azide protected.146
When the deuterium effect was not as large as predicted from in vitro experiments
with singlet oxygen, it was hypothesized that natural quenchers in the cell masked
the deuterium effect, and that therefore it could still be concluded that singlet oxygen
must be the major intermediate for photodynamic actions in acridine
orange-sensitized yeast cells.260 Deuterium oxide did not enhance the gene
conversion rate caused by acridine orange in yeast cells.260 It has been stated that it
is unlikely that the attack of singlet oxygen on DNA will give rise to either single
stranded breaks or alkali-labile bonds.223 However, both type I and type II reactions
act in the photodynamic destruction of guanine.262
When Nilsson et al. (1978) observed that "until now there has been no
conclusive evidence for the participation of singlet oxygen in any biological
photooxidation in solution," they claimed that now they had unambiguous evidence
for the participation of singlet oxygen in photodynamic oxidation of amino acids, but
their unambiguous evidence turned out to rely on the observation of the destruction
of histidine, since there is a low rate constant for reaction of histidine with triplet
dye-without considering rate constants for alternative reactions of histidine with
O2-, peroxide, or OH..263 Singlet oxygen has been assayed as the decrease in
tryptophan absorbance at 280 nm, without confirming assays264, when clearly there
are more species than singlet oxygen capable of degrading tryptophan, and some of
them, such as OH., are very likely to be produced by singlet oxygen-generating
systems. In fact, tryptophan degradation has been used as a direct assay for the
production of OH.. Deuterium oxide enhancement is very small in toxicity in yeast
with acridine orange, proflavin, and acriflavin, though somewhat larger for rose
bengal.265 The role of singlet oxygen was investigated but could not be established
in the methylene blue-photosensitized strand cleavage of DNA, because there was
little increase in reaction in the presence of deuterium oxide.266
Production of oxygen radicals and peroxide by the photodynamic effect
Evidence does exist for the production of O2-, H2O2, and OH. by the
photodynamic effect. For example, hemolysis in vitro by light-activated
benoxaprofen results in production of free radicals, singlet oxygen and O2- anion.267
Other evidence concerning the production of O2-, H2O2, or OH is given below.
Production of superoxide by the photodynamic effect
Superoxide is difficult to observe directly in photosensitized irradiations since
most sensitizers absorb light strongly in the 250-300 nm region where O2-
absorbs;262 nevertheless, indirect assays have demonstrated production of O2- by
excited, semi-reduced dye molecules. The superoxide dismutase-inhibitable
reduction of cytochrome c was used to demonstrate that O2- is produced by acridine
yellow, acridine orange, fluorescein, and lucifer yellow, when illuminated.183 A
reduced oxygen radical formed from the fluorescein-photosensitized autoxidation
of tyrosine was mentioned by Kasche in 1967.268 Kasche and Lindqvist suggested
that the reaction between the triplet state of fluorescein and oxygen most likely
produces O2-.269 This observation was made before the discovery that oxygen
radicals are commonly produced in biological systems. Photochemical generation
of O2- by rose bengal in the presence of sulfites was suggested in 1978270. Both
singlet oxygen and O2- are produced by polymer-bound rose bengal through direct
oxidation of rose bengal by dioxygen.271 Superoxide is produced by the methylene
blue-sensitized photooxidation of epinephrine; it was hypothesized to be produced
from singlet oxygen,272 but others maintain that O2- is most likely produced by a
different path than singlet oxygen.273 However, in the presence of appropriate
reductants, singlet oxygen actually proves to be a source of O2-; NADH reacts with
singlet oxygen to produce O2- and a semi-reduced NAD radical (fig. 5).274 Though
rose bengal and its derivatives have been used extensively as a singlet oxygen
source (type II), their strong absorption at ~550 nm suggests a number of other
possible photochemical processes, none of which have been fully exploited or
explored.182 The photooxidation of rose bengal in non-polar solvents is
predominantly type-I.182 Using benzoquinone to detect O2-, Lee and Rodgers
(1987) found that about 20% of activated oxygen molecules resulting from reaction
with rose bengal in the absence of an oxidizable substrate was O2-.262 The
light-activated phenothiazine drug chlorpromazine gave a DMPO-OH adduct
indicative of OH. whose intensity was decreased 50% by SOD.275 Psoralens have
been observed to produce O2- directly.276,277,168 Both light-activated linear
psoralens and isopsoralens produced both singlet oxygen and O2-, in varying
proportions. When riboflavin, psoralen, benzoyl peroxide, and hematoporphyrin
derivative were irradiated with ultraviolet light, they each produced superoxide.278
When water was replaced with deuterium oxide, the enhancement of guanosine
oxidation by the dyes acridine orange, rose bengal, thionin, and methylene blue was
less than expected, suggesting a mixture of singlet oxygen and O2- involvement.279
The production of single stranded breaks in FX174 DNA is thought to occur by a two
photon process-absorption at 440 nm and between 320 & 360 nm, followed by an
electron transfer from the excited dye molecule to the substrate, DNA.220
Production of hydrogen peroxide by the photodynamic effect
Specific evidence for H2O2 production in photodynamic effects has also been
found. The toxicity of ultraviolet-irradiated tryptophan has been shown to be largely
due to H2O2.280 The formation of H2O2 by methylene blue in the presence of
ascorbate has been observed.281 Hydrogen peroxide is produced by the reaction
between hematoporphyrin derivative and ascorbate in the presence of light.282 This
is of physiological significance since some of the tissues targeted for
photochemotherapy contain ascorbate at levels of 2-3 mM-such as the
mammalian lens. Hydrogen peroxide is produced by xanthene and thiazine dyes in
the presence of oxidizable substrates.283 It has been suggested that the same
mechanism is responsible for near ultraviolet inactivation and photodynamic
inactivation of E. coli;115 the H2O2 scavenger catalase, when incorporated into
plating medium protects against near ultraviolet lethality284 and ammeliorates
photodynamic killing by acridines and thiazines.183 The proflavin-sensitized
breakdown of DNA may occur by a `radical pathway', which requires a two-photon
absorption and leads to the formation of peroxide free radicals in the DNA moiety of
the DNA-proflavin complexes, leading to strand breakages.285
Production of hydroxyl free radical by the photodynamic effect
The production of OH. has also been observed in dye photooxidation.
Butylated hydroxytoluene's observed ability to protect against photodermo-toxicity
had been explained on the basis of a hypothesized ability to change the structure of
the skin, but when this was disproved its antioxidant effects were suggested to be
responsible.286 This suggests that free radicals are responsible for the phototoxic
effect in vivo. Vitamin E is changed to its radical form in the presence of
hematoporphyrin derivative and light, indicating that it has been attacked by a free
radical.287 Both singlet oxygen and OH. have been shown to be produced by HpD in
vitro.147 The OH. scavengers DMSO, sodium benzoate, and thiourea protected E.
coli B against the toxic effects of illuminated acridine orange.183 Both singlet oxygen
and hydrogen abstraction mechanisms are involved in the photosensitization of
substituted phenylalanines and tyrosines, but the hydrogen abstraction
mechanism, a typical OH. reaction, predominates at pH 8.288
The formation of OH. by methylene blue in the presence of ascorbate has
been observed.289 When exposed to the spin trap compound DMPO, photoexcited
psoralens were found to produce a DMPO-OH EPR signal that is abolished when
the reaction is carried out in the presence of OH. scavengers.168 Hydroxyl free
radicals are produced by hematoporphyrin derivative, ascorbate, and light.290 In the
dye-sensitized photooxidation in methanol of polyunsaturated fatty acids, using the
dyes methylene blue, erythrosin, hematoporphyrin, and riboflavin, the isomer
product distribution was interpreted in terms of a dual singlet oxygen and radical
mechanism.142 In the photoactivated reaction of tartrazine, which is mutagenic only
in the presence of light, there is no singlet oxygen participation, according to the
cholesterol reaction assay, in which singlet oxygen is detected by its reaction with
cholesterol, which results in reaction products specific to this reaction. However,
superoxide dismutase & catalase inhibited OH. production, as measured by EPR
using the spin-trap DMPO. 291
Other photodynamic reactions that produce the hydroxyl radical
Assorted other reactions appear to be dependent on
photodynamically-produced OH., whether or not the photosensitizer has been
identified. When hematoporphyrin derivative was illuminated in the presence of
methionine or tryptophan and Fe-EDTA, OH. were detected using both the
salicylate and thiobarbituric acid assays292. Tryptophan yielded a photoproduct that
could substitute for H2O2 in the iron-catalyzed Haber-Weiss reaction, while
porphyrin radical could replace O2-.292 Light-induced chromatid damage in cells
being cultured for other studies was found to be prevented by the H2O2 scavenger
catalase or the OH. scavenger mannitol; after filters were installed to protect cells
and medium from light of wavelengths less than 500 nm, the unwanted damage to
the cultured cells was eliminated.293 DNA stand-breaking activity of lipid peroxides
is dependant on metal ions and partly inhibited by catalase294. The formation of
fluorescent products from DNA is increased by metal ions and ascorbate.295,296
Dye photosensitization reactions in vitro may produce singlet oxygen or O2-,
or both species, depending on the prevailing reaction conditions. In order to
determine the identity of oxidizing species in vivo, it is important to 1) use in vitro
model systems that mimic in vivo conditions as closely as possible, with regard to
presence and concentration of targets, substrates, chelators, and trace metals; and
2)utilize in vivo methods of study as much as possible.
I would like to emphasize why we need to know, in as
much detail as possible, about the identity of the oxidizing
species and how they are produced. Without such
information, our attempts to understand the biological role of
these species and then to modify them are likely to fail. This
is particularly important for strongly oxidizing species that
act like the hydroxyl radical.
-H.M.Swartz297
Choice of organism
The reasons why Escherichia coli is the organism of choice for these
studies have to do with the ease of manipulating it and its extensive use in previous
work on oxygen toxicity. The bacterial cell is a single compartment (or at most a dual
compartment) organism whose intracellular targets are all approximately equally
accessible by oxygen radicals. E. coli is the best studied of all systems with respect
to the regulation of SOD and catalase, as well as the role of SOD and catalase in
preventing oxygen toxicity, and has been a very useful organism for previous studies
because the cellular levels of the enzymes superoxide dismutase and catalase are
easily manipulated by the choice of media or additions to the media. E. coli is also the
best-studied system as far as the role of redox-cycling compounds in generating
oxygen radicals is concerned, as well as systems of defensive enzymes. The genes
for one of the catalases, both of the superoxide dismutases, and nucleases involved
in DNA repair have already been cloned onto multicopy plasmids. E. coli is an
organism that is easily grown and counted, one that can be easily manipulated, both
physiologically and genetically. Finally, E. coli has a long history of study in
connection with the phototoxicity of dyes used in this work.298,299,300 As all aerobic
organisms are subject to similar problems as a result of oxygen toxicity, and their
defense mechanisms have much in common, E. coli should provide a useful model
for all aerobic cells.
Summary
Studies on the photodynamic effect have concentrated on singlet oxygen for
historical reasons. The observation that oxygen was necessary for the toxicity of the
photodynamic effect led naturally to the conclusion that an activated form of oxygen
was the mediator.301 The study of singlet oxygen chemistry antedates the discovery
of oxygen radicals and their effects on biological systems. McCord and Fridovich's
discovery of superoxide dismutase in 1969 led to the recognition of the toxic effects
of the superoxide radical.302 These toxic effects were subsequently found to be at
least in part due to the production of hydrogen peroxide and hydroxyl radicals via the
iron-catalyzed Haber-Weiss reaction.80 These two bodies of knowledge,
dye-mediated photosensitization and oxygen radical toxicity, have grown up in
parallel without much cross-fertilization. This study was undertaken in order to show
that many biological effects previously attributed to singlet oxygen can be explained
by free radical mechanisms. The work described in this thesis follows up on
observations already made by our lab in this respect183 and also extends previous
observations on photosensitization.
Lethality due to oxygen radicals is an interesting phenomenon worthy of
study; so is the production of these oxygen radicals by the photodynamic effect. This
investigation is intended to shed light on the identity of cellular targets attacked by
toxic oxygen species, on which of the oxygen species are involved in toxicity, and on
factors involved in the photodynamic effect's production of oxygen species.
A major cellular target attacked by toxic oxygen species is likely to be DNA.
Single-stranded nicking of DNA will be examined in this study in order to determine
which factors are involved in this damage. Among the factors to be examined are the
effects of scavengers of superoxide, hydrogen peroxide, and of hydroxyl free radical
on the DNA damage; effectiveness of a wide range of dyes in classes including the
xanthenes, acridines, thiazines, and a phenazine; effect of different amounts of iron
and copper and of a wide range of biologically significant potential iron chelators;
and the effects of a series of biological reductants on the production of DNA damage
by the photodynamic effect.
The production of toxic oxygen species will be examined in this study by
means of a series of assays. This study will use assays for oxidation of NADH and
reduction of cytochrome c. It will examine the spectral shifts caused by intercalation
into DNA of individual dyes to study intercalation, and the dyes' ability to reduce
cytochrome c and oxidize NADH even when intercalated will be investigated.
Production of hydroxyl radicals as assayed by two different assays will be examined
in the presence of various scavengers of O2-, H2O2, and OH., as well as in the
presence of different metal concentrations, different dyes, different chelators, and
different reductants. Nicking of DNA will be examined similarly. Cellular levels of one
biologically important reductant, glutathione, will be measured after exposure of the
cells to dyes. Amelioration of kill levels will be determined for the presence of
plasmids coding for catalase, SOD, and endonuclease IV, as well as preinduction of
protective enzymes by PQ and Mn and the presence of thiourea. Finally, in vivo
breakage of plasmid DNA will be looked at.
on to Chapter 2
References for Chapter 1
(select the number to go back to
the first citation in the text)
1. Bert, P. 1943. Barometric pressure: Researches in Experimental Physiology. Translated by M.A. Hitchcock and F. A. Hitchcock. College Book Company, Columbus, Ohio, p. 565-570.
2. Quebedeaux, B., U. D. Havelka, K. L. Livak, and R. W. F. Hardy. 1975. Effect of altered pO2 in the aerial part of soybean on symbiotic nitrogen fixation. Plant Physiol. 56:761-764.
3. Cooper, P. D., A. M. Burt, and J. N. Wilson.1958. Critical effect of oxygen tension on rate of
growth of animal cells in continuous suspended culture. Nature 182:1508-1509.
4. Patz, A., A. Eastman, D. H. Higginbotham, and T. Klett. 1953. Oxygen studies in retrolental fibroplasia. Amer. J. Opthamol.36: 1511-1522.
5. Conger, A. D., and L. M. Fairchild. 1952. Breakage of chromosomes by oxygen. Proc. Natl. Acad. Sci. USA 38: 289-299.
6. Fenn, W. C., R. Gershman, and D. L. Gilbert. 1957. Mutagenic effects of high oxygen tensions
on Escherichia coli. Proc. Natl. Acad. Sci. USA 43: 1027-1032.
7. Fuson, R.L., H.A. Saltzman, W.W.Smith, R.E. Whelan, S. Osterhout, R.T. Parker. 1965.
Clinical hyperbaric oxygenation with severe oxygen toxicity. N. Engl. J. Med. 273:415-419.
8. Green, M. J., and H. A. O. Hill. 1984. Chemistry of Dioxygen. Meth. Enzymol. 105:3-22.
9. Gerschman, R., D.L. Gilbert, S.W. Nye, P. Dwyer, W.O. Fenn. 1954. Oxygen poisoning and
X-irradiation: a mechanism in common. Science 119:623-626.
10. Fridovich, I. 1970. Quantitative aspects of the production of superoxide
anion radical by xanthine oxidase. J. Biol. Chem. 245:4053-4057.
11. Nishikimi, M. 1975. Oxidation of ascorbic acid with superoxide anion generated by the
xanthine/XO system. Biochem. Biophys. Res. Com. 63:463-468.
12. Knowles, P. F., J. F. Gibson, F. M. Pick, and R. C. Bray. 1969. Electron-spin-resonance
evidence for enzymic reduction of oxygen to a free radical, the superoxide ion. Biochem. J.
111: 53-58.
13. Misra, H.P., and I. Fridovich. 1972. The univalent reduction of oxygen by reduced flavins and
quinones. J. Biol. Chem. 247: 188-192.
14. Bartoli, G., T. Galeotti, and A. Azzi. 1977. Production of superoxide anions and hydrogen
peroxide in ehrlich ascites tumour cell nuclei. Biochim. Biophys. Acta 497: 622-626.
15. Patton, S.E., G.M. Rosen, and E.J. Raukman. 1980. Superoxide production by purified
hamster hepatic nuclei. Mol. Pharmacol. 18: 588-593.
16. Asada, K., and K. Kiso. 1973.The photo-oxidation of epinephrine by spinich chloroplasts and
its inhibition by SOD: evidence for the formation of superoxide radicals in chloroplasts. Agr.
Biol. Chem. 27: 453-457.
17. Kono, Y., M.-A. Takahashi, and K. Asada. 1976. Oxidation of manganous pyrophosphate by
superoxide radicals and illuminated spinich chloroplasts. Arch. Biochem. Biophys. 174:
454-462.
18. Bovaris, A., R.A.Sanchez, and M.T. Beconi. 1978. Antimycin and cyanide-resistant respiration
and superoxide anion production in fresh and aged potato tuber mitochondria. FEBS Letters
92: 333-338.
19. Johnston, R.B., Jr., J.E. Lehmeyer, and L.A. Guthrie. 1976. Generation of superoxide anion
and chemiluminescence by human monocytes during phagocytosis and on contact with
surface-bound immunoglobulin G. J. Exp. Med. 143: 1551-1556.
20. Johnston, R.B., Jr., C.A.Godzik and Z.A. Cohn. 1978. Increased superoxide anion production
by immunologically activated and chemically elicited macrophages. J. Exp. Med. 148:
115-127.
21. Aust, S.D., D.L. Roerig, and T.C. Pederson. 1972. Evidence for superoxide generation by
NADPH-cytochrome c reductase of rat liver microsomes. Biochem. Biophys. Res. Com. 47:
1133-1136.
22. Simon, R.H., C.H. Scoggin, and D. Patterson. 1981. Hydrogen peroxide causes the fatal injury
to human fibroblasts exposed to oxygen radicals. J. Biol. Chem. 256: 7181-7186.
23. Misra, H.P., and I. Fridovich. 1972. The generation of superoxide radical during the
autoxidation of hemoglobin. J. Biol. Chem. 247: 6960-6962.
24. Gotoh, T., and K. Shikama. 1976. Generation of the superoxide radical during the autoxidation
of oxymyoglobin. J. Biochem. 80: 397-399.
25. Misra, H.P., and I. Fridovich. 1971. The generation of superoxide radical during the
autoxidation of ferredoxins. J. Biol. Chem. 246: 6886-6890.
26. Ballou, D., G.Palmer, and V. Massey. 1969. Direct demonstration of superoxide anion
production during the oxidation of reduced flavin and its catalytic decomposition by
erythrocuprein. Biochem. Biophys. Res. Com. 36: 898-904.
27. McCord, J.M., and I. Fridovich. 1973. Production of superoxide anion in photolyzed water
demonstrated through the use of superoxide dismutase. Photochem. Photobiol. 17: 115-121.
28. Carmichael, A.J., A. Samuni, and P. Rietsz. 1985. Photogeneration of superoxide and
decarboxylated peptide radicals by carboquone, mitomycin C and streptonigrin. Photochem.
Photobiol. 41: 635-641.
29. Hassan, H.M., and I. Fridovich. 1978. Superoxide radical and the oxygen enhancement of the
toxicity of paraquat in Escherichia coli. J. Biol. Chem. 253: 8143-8148.
30. Bielski, B.H.J., and H.W. Richter. 1977. A study of the superoxide radical chemistry by
stopped-flow radiolysis and radiation induced oxygen consumption. J. Amer. Chem. Soc. 99:
3019-3023.
31. Nishikimi, M., H. Yamada, and K. Yagi. 1980. Oxidation by superoxide of tocopherols
dispersed in aqueous media with deoxycholate. Biochim. Biophys. Acta 627: 101-108.
32. Bielski, B.H.J., and P.C.Chan. 1975. Kinetic study by pulse radiolysis of the
lactate dehydrogenase-catalyzed chain oxidation of nicotinamide adenine
dinucleotide by HO2. and O2- radicals. J. Biol. Chem. 250: 318-321.
33. Sawyer, D.T., and J.S. Valentine. 1981. How super is superoxide?Acc. Chem. Res. 14:
393-400.
34. Kono, Y., M.-A. Takahashi, and K. Asada. 1976. Oxidation of manganous
pyrophosphate by superoxide radicals and illuminated spinich chloroplasts.
Arch. Biochim. Biophys. 174: 454-462.
35. Butler, J., and B. Halliwell. 1982. Reaction of iron-EDTA chelates with the
superoxide radical. Arch. Biochem. Biophys. 218:174-178.
36. McCord, J. M., and I. Fridovich. 1968. The reduction of cytochrome c by milk xanthine oxidase.
J. Biol. Chem. 243: 5753-5760.
37. Bielski, B.H.J., G.G. Shiue, and S. Bajuk. 1980. Reduction of Nitro Blue
Tetrazolium by CO2- and O2- Radicals. J. Phys. Chem. 84: 830-832.
38. Bielski, B.H.J., R.L.Arudi, and M.W. Sutherland. 1983. A study of the reac
tivity of HO2/O2- with unsaturated fatty acids. J. Biol. Chem. 258:
4759-4761.
39. Misra, H.P. 1974. Generation of superoxide free radical during the autoxidation of thiols. J.
Biol. Chem. 249: 2151-2155.
40. Marklund, S., and G. Marklund. 1974. Involvement of the superoxicde anion
radical in the autoxidation of pyrogallol and a convenient assay for superoxide
dismutase. Eur. J. Biochem. 47:469-474.
41. Misra, H.P., and I. Fridovich. 1972. The role of superoxide anion in the auto
xidation of epinephrine and a simple assay for superoxide dismutases. J. Biol.
Chem. 247:3170-3175.
42. Kono, Y., and I. Fridovich. 1982. Superoxide radical inhibits catalase. J. Biol. Chem. 257:
5751-5754.
43. Blum, J., and I. Fridovich. 1984. Inactivation of the selenium-glutathione peroxidase by
superoxide. Fed. Proc. 43: 2011-2015.
44. Halliwell, B. and J. M. C. Gutteridge. 1985. Free Radicals in Biology and
Medicine. Clarendon Press, Oxford, p. 280.
45. McCord, J.M. 1974. Free radicals and inflammation: protection of sinovial fluid by superoxide
dismutase. Science 185: 529-531.
46. Lynch, R.E., and I. Fridovich. 1978. Effects of superoxide on the erythrocyte membrane. J.
Biol. Chem. 253:1838-1845.
47. Rosen, G. M., M. J. Barber, and E. J. Raukman. 1983. Disruption of erythrocyte membranal
organization by superoxide dismutase. J. Biol. Chem. 258: 2225-2228.
48. Girotti, A.W. 1985. Lipid photoxidation in erythrocyte ghosts-sensitization of the membranes
toward ascorbate-induced and superoxide-induced peroxidation and lysis. Arch. Biochem.
236: 238-251.
49. Brawn, K., and I. Fridovich. 1981. DNA Strand Scission by enzymically generated oxygen
radicals. Arch. Biochem. Biophys. 206: 414-419.
50. Lesko, S.A., R.J. Lorentzen, and P.O.P. Ts'o. 1980. Role of superoxide in DNA strand scission.
Biochemistry 19: 3023-3028.
51. Ross, W.E., E.R. Block, and R.Y. Chang. 1979. Paraquat-induced DNA damage in mammalian
cells. Biochem. Biophys. Res. Comm. 91:1302-1308.
52. Babior, B., J.T. Curnutte, and R.S. Kipnes. 1975. Biological defense mechanisms: evidence for
the participation of superoxide in bacterial killing by xanthine oxidase. J. Lab. Clin. Med. 85:
235-244.
53. Cunningham, M.L., and B.R. Lokesh. 1983. Superoxide anion generated by KO2 is cytotoxic
and mutagenic to Chinese hamster ovary cells. Mutat. Res. 121: 299-304.
54. Weitzman, S.A., A.B. Weitberg, E. Clark, and T.Stossel. 1985. Phagocytes as carcinogens:
malignant transformation produced by human neutrophils. Science 227: 1231-1233.
55. Weitzman, S.A., A.B. Weitberg, E. Clark, and T.Stossel. 1985. Phagocytes as carcinogens:
malignant transformation produced by human neutrophils. Science 227: 1231-1233.
56. Fee, J.A. 1980. Is superoxide toxic?, p.41-48. In W.H. Bannister and J.V. Bannister (eds.),
Biological and Clinical Aspects of Superoxide and Superoxide Dismutase. Elsevier/North
Holland, New York.
57. Baum, R.M. 1984. Superoxide theory of oxygen toxicity is center of heated debate. Chemical
and Engineering News April 9, 1984: p. 20-26.
58. Fee, J.A. 1981. Is superoxide toxic and are superoxide dismutases essential for aerobic life? In
M.A.J. Rodgers and E.L. Powers (eds.), Oxygen and Oxyradicals in Chemistry and Biology.
Academic Press, New York.
59. J.S. Valentine. 1979. The Chemical Reactivity of Superoxide Anion. In W.S. Caughey (ed.),
Biochemical and Clinical Aspects of Oxygen. Academic Press, New York.
60. Bielski, B.H.J., and G. Schive. 1979. Reaction rates of superoxide radical with the essential
amino acids, p. 43-65. In D. W. Fitzsimons (ed.), Oxygen Free Radicals in Tissue damage.
Elsevier/North-Holland, Amsterdam.
61. Loschen, G., A. Azzi, C. Richler, and L. Flohe. 1974. Superoxide radicals as precursors of
mitochondrial hydrogen peroxide. FEBS Letters 42: 68-72.
62. Turrens, J.F., Freeman, B.A., and Crapo, J.D. 1982. Hyperoxia increases hydrogen peroxide
release by lung mitochondria and microsomes. Arch. Biochem. Biophys. 217: 411-421.
63. Cohen, G., and Heikkila, R. 1974. The generation of hydrogen peroxide, su
peroxide radical and hydroxyl radical by 6-hydroxydopamine, dialuric acid
and related cytotoxic agents. J. Biol. Chem. 249:2447-2452.
64. May, J.M., and C. de Haën. 1979. Insulin-stimulated intracellular hydrogen peroxide
production in rat epididymal fat cells. J. Biol. Chem. 254: 2214-2220.
65. Zigler, J.S., Jr., H.M. Jernigan, Jr., D. Garland, and V.N. Reddy. 1985. The effects of "oxygen
radicals" generated in the medium on lenses in organ culture: inhibition of damage by chelated
iron. Arch. Biochem. Biophys. 241: 163-172.
66. Uchida, Y., H. Shigematu, and K. Yamafuji. 1965. The mode of action of hydrogen peroxide on
DNA. Enzymologia 6: 369-376.
67. Keller, K.M., and E.C.Pollard. 1977. Action of hydrogen peroxide on degradation of DNA after
irradiation in E.coli. Int. J. Radiat. Biol. 31: 407-413.
68. Massie, Y., M. Kitahara, K. Maeda, and H. Umezawa. 1972. The kinetics of degradation of
DNA and RNA by hydrogen peroxide. Bioc. Biop. Acta 272: 539-548.
69. Meneghini, R., and M.E. Hoffmann. 1980. The damaging action of hydrogen peroxide on DNA
of human fibroblasts is mediated by a non-dialyzable compound. Biochim. Biophys. Acta 608:
167-173.
70. McCormick, J.P., J.R. Fischer, J.P. Pachlatki, and A.A. Eisenstark. 1976. Characterization of a
cell-lethal product from the photooxidation of tryptophan: hydrogen peroxide. Science 191:
468-469.
71. Plaine, H.L. 1955. The effect of oxygen and of hydrogen peroxide on the action of a specific
gene and on tumor induction in Drosophila melanogaster. Genetics 40: 268-280.
72. Bray, R.C., and L. Calabres. 1974. Reduction and inactivation of superoxide dismutase by
hydrogen peroxide. Biochem.J. 139: 43-47.
73. F. Haber and J. Weiss. 1934. The catalytic decomposition of hydrogen peroxide by iron salts.
Proc. Roy. Soc. London (A) 147:332-342.
74. Koppenol, W.H., J. Butler, and J.W. van Leeuwen. 1978. The Haber-Weiss
cycle. Photochem. Photobiol. 28:655-660.
75. Czapski, G. 1984. Reaction of hydroxyl radical. Meth. Enzymol. 105:
209-215.
76. Halliwell, B., and J.M.C. Gutteridge. 1984. Role of iron in oxygen radical reactions. Meth.
Enzymol. 105: 47-56.
77. Anonymous. 1970. USDA handbook of nutritional values in foods. Dover Pub
lications, New York.
78. Buettner, G.R., L.W. Oberley, and S.W. Leuthaus. 1978. The effect of iron
on the distribution of superoxide and hydroxyl radicals as seen by spin
trapping and on the superoxide dismutase assay. Photochem. Photobiol.
28:693-695.
79. Halliwell, B. 1978.Superoxide-dependent formation of hydroxyl radicals in the presence of iron
salts-its role in degradation of hyaluronic acid by a superoxide-generating system. FEBS
Letters 96: 238-242.
80. Mello-Filho, A.C., and R. Meneghini. 1984. In vivo formation of single-strand breaks in DNA
by hydrogen peroxide is mediated by the Haber-Weiss reaction. Biochem. Biophys. Acta 781:
56-63.
81. Sutton, H.C., and C.C. Winterbourne. 1984. Role of Metal Chelating Agents as Catalysts in a
Hydroxyl Radical Forming Process- a comparison with the Haber-Weiss reaction,
p.177-183. In W. Bors (ed.), Oxygen Radicals in Chemistry and Biology. Walter de Gruyter &
Co, Berlin.
82. Buettner, G.R., L.W.Oberley, and S.Chan Leuthauser. 1978. The effect of iron on the
distribution of superoxide and hydroxyl radicals as seen by spin trapping and on the
superoxide dismutase assay. Photochem. Photobiol. 28: 693-697.
83. Floyd, R.A. 1981. DNA-ferrous iron catalyzed hydroxyl free radical formation from hydrogen
peroxide. Biochem. Biophys. Res. Com. 99: 1209-1215.
84. Motohashi, N., and I. Mori. 1983. Superoxide-dependent formation of the hydroxyl radical
catalyzed by transferrin. FEBS Letters 157: 197-199.
85. Gutteridge, J.M.C., S.K. Paterson, A.W. Segal, and B. Halliwell. 1981. Inhibition of lipid
peroxidation by the iron-binding protein lactoferrin. Biochem. J. 199: 259-261.
86. Richmond, R. and B. Halliwell. 1982. Formation of hydroxyl free radical from the paraquat radical
cation, demonstrated by a highly specific GC technique. The role of superoxide, hydrogen
peroxide, and glutamine reductase. J. Inorg. Biochem. 17: 95-107.
87. Buettner, G.R., and L.W. Oberley. 1979. The production of hydroxyl radical by tallysomycin and
copper(II). FEBS Letters 101: 333-335.
88. Winterbourn, C.C. 1981. Evidence for the production of hydroxyl radicals from the adriamycin
semiquinone and hydrogen peroxide. FEBS Letters 136: 89-94.
89. Lai, C.-S., and L.H. Piette. 1977. Hydroxyl radical production involved in lipid peroxidation of rat
liver microsomes. Biochem. Biophys. Res. Com. 78: 51-59.
90. Tauber, A.I., and B.M. Babior. 1977. Evidence for hydroxyl radical production by human
neutrophils. J. Clin. Invest. 60: 374-379.
91. Weiss, S.J., G.W. King, and A. F. LoBuglio. 1977. Evidence for hydroxyl radical generation by
human monocytes. J. Clin. Invest. 60: 370-373.
92. Dorfman, L.M., and G.E. Adams. 1973. Reactivity of the Hydroxyl Radical in
Aqueous Solutions. National Standard Reference Data Service- National
Bureau of Standards No. C13.48:46.
93. Lesko, S.A., J.-L. Drocourt, and S.-U. Yang. 1982. DNA-protein and DNA-interstrand
cross-links induced in isolated chromatin by hydrogen peroxide and ferrous EDTA chelates.
Biochemistry 21: 5010-5015.
94. Schellenberg, K.A. 1979. Reaction of hydroxyl radical with thymidine. Scavenging by a
dihydropyridine. Fed. Proc. Abstr. 38: 501.
95. Rosen, H., and S. J. Klebanoff. 1981. Role of iron and EDTA in the bactericidal activity of a
superoxide anion-generating system. Arch. Biochem. Biophys. 208: 512-519.
96. Heikkila, R.E., B. Winston, G. Cohen, and H. Barden. 1976. Alloxan-induced diabetes. Evidence
for hydroxyl radical as a cytotoxic intermediate. Biochem. Pharm. 25: 1085-1092.
97. Grankvist, K., S. Marklund, J. Sehlin, and I. Taljedal. 1979. Influence of trace metals on alloxan
cytotoxicity in pancreatic islets. FEBS Letters 105: 15-18.
98. Kutsuki, H., and M.H. Gold. 1982.Generation of hydroxyl radical and its involvement in lignin
degradation by Phanerochaete chrysosporium. Biochem. Biophys. Res. Com.109: 320-325.
99. Percival, M.P., and A.D.Dodge. 1983. Photodynamic effects of rose bengal
on senescent flax cotyledons. J. Exp. Biol. 34(138)47-54.
100. Wasserman, H.H., and R.W. Murray, eds. 1978. Singlet Oxygen. New York:
Academic Press.
101. Ranby, B., and J.F. Rabek, eds. 1978. Singlet oxygen reactions with organ
ic compounds and polymers. New York: Wiley.
102. Fridovich, I. 1974. Superoxide dismutases. Advances in Enzymology
41:35-97.
103. R.A. Weisiger and I. Fridovich, J. Biol. Chem.248:3582-3592(1973) Superoxide dismutase:
organelle specificity.
104. McCord, J.M. and I. Fridovich. 1969. Superoxide dismutase: an enzymic
function for erythrocuprein (hemocuprein). J. Biol. Chem. 244:6049-6055.
105. Keele, B.B., Jr., J.M. McCord, and I. Fridovich. 1971. Further characteriza
tion of bovine superoxide dismutase and its isolation from bovine heart. J.
Biol. Chem. 246:2875-2880.
106. Carrico, R.J., and H.F. Deutsch. 1970. The presence of zinc in human cy
tocuprein and some properties of the apoprotein. J. Biol. Chem.
245:723-727.
107. Harris, J.I., and H.M. Steinman. 1977. Amino acid sequence homologies
among superoxide dismutases. In "Superoxide and Superoxide Dismu
tases." A.M.Mitchellson, J.M.McCord, and I. Fridovich, eds., Academic
Press, London. pp225-230.
108. Pugh, S.Y.R., and I. Fridovich. 1985. Induction of superoxide dismutases in
Escherichia coli B by metal chelators. J. Bacteriol. 162:196-202.
109. Hassan, H.M., and I. Fridovich. 1979. Intracellular production of superoxide
radical and of hydrogen peroxide by redox active compounds. Arch. Bio
chem. Biophys. 196:385-395.
110. Claibourne, A., and Fridovich, I. 1979. Purification of the o-dianisidine peroxidase of Escherichia
coli B. J. Biol. Chem. 254:4245-4252.
111. Loewen, P.C., B.L. Triggs, C.S. George, and B.E. Hrabarchuk. 1985. Genetic mapping of katG, a
locus that affects synthesis of the bifunctional catalase-peroxidase hydroperoxidase I in
Escherichia coli. J. Bacteriol. 162:661-667.
112. Schonbaum, G.R., and B. Chance. 1976. Catalase. In The Enzymes, vol.
XIII, P.D. Boyer, ed. Academic Press, New York.
113. Richter, H.E., and P.C. Loewen. 1981. Induction of catalase in Escherichia
coli by ascorbic acid involves hydrogen peroxide. Biochem. Biophys. Res.
Comm. 100:1039-1046.
114. Hassan, H.M. and I Fridovich. 1978. Regulation of the synthesis of cata
lase and peroxidase in Escherichia coli. J. Biol. Chem. 253:6445-6450.
115. Tuveson, R.W., and M.E.March. 1980. Photodynamic and sunlight inactiva
tion of Escherichia coli strains differing in near-uv sensitivity and recom
bination proficiency. Photochem. Photobiol.31:287-289.
116. Loewen, P.C., B.L. Triggs, G.R. Klassen, and J.H. Weiner. 1983. Identification and physical
characterization of a col E1 hybrid plasmid containing a catalase gene of Escherichia coli .
Can. J. Biochem. Cell Biol. 61:1315-1321.
117. Martin, J.P., Jr. Unpublished observation, 1986.
118. Christman, M.F., R.W. Morgan, F.S. Jacobson, and B.N. Ames. 1985. Positive control of a
regulon for defenses against oxidative stress and some heat-shock proteins in Salmonella
typhimurium. Cell 41: 753-762.
119. Triggs-Raine, B.L., and P.C. Loewen. 1987. Physical characterization of
katG, encoding HPI of Escherichia coli. Gene 52:121-128.
120. Nettleton, C.J., C. Bull, T.O. Baldwin, and J.A.Fee. 1984. Isolation of th