Pure exogenous singlet oxygen: Nonmutagenicity in bacteria

Mutation Research, 201 (1988) 127-136 Elsevier

127

MTR 04619

Pure exogenous singlet oxygen: N o n m u t a g e n i c i t y in bacteria Thomas A. Dahl a, W. Robert Midden b,, and Philip E. Hartman

a

a Departments of Biology and b Environmental Health Sciences, The Johns Hopkins University, Baltimore, MD 21218 (U.S.A.) (Received 25 September 1987) (Revision received 28 December 1987) (Accepted 6 January 1988)

Keywords: Singlet oxygen; Salmonella; Histidine reversion; Forward mutation; Separated-surface-sensitizer; Phage induction; Ames strains

Summary Singlet oxygen (1AGO2) is the lowest energy-excited state of molecular oxygen, and more reactive than the triplet ground-state molecule. Although singlet oxygen has been implicated in a variety of biological effects, including reactions with DNA or some of its components, evidence for mutagenesis by singlet oxygen has remained unclear. We have previously described a system for bacterial exposure to pure exogenous singlet oxygen that eliminates ambiguity regarding the identity of the reactive species responsible for observed results. Despite the potent toxicity of pure singlet oxygen for several different strains of bacteria, we have found no evidence for mutagenicity of singlet oxygen in 26 Salmonella typhimurium histidine-auxotrophic strains killed to 35% survival. These strains included a variety of base-pair substitution or frameshift target sequences for reversion, including targets responsive to oxidative damage and targets rich in GC base pairs. Some strains combined histidine mutations with one or more mutations affecting DNA-repair capacity. 4 strains possessing the hisG46 mutation also were not mutated when exposed to dose ranges killing less than 28% and up to 99% of the bacteria. The relative frequency of small inphase deletions was assayed in hisG428 bacteria exposed to single oxygen and found to be the same as the spontaneous level. In addition to lack of induction of mutation in these strains, the 8-azaguanine forward mutation assay yielded no evidence of mutagenesis by singlet oxygen in strains killed to 15% survival. No induction of genetic changes by singlet oxygen was seen in an assay for duplication of - 1/3 of the bacterial chromosome. Tests for the ability of singlet oxygen to induce lambda prophage in E. coli K12 also proved negative. These studies collectively indicate that pure singlet oxygen generated outside the bacterial cell does not react significantly with the bacterial chromosome in ways leading to base-pair substitutions, frameshift mutations, small or large deletions, large duplications, or damage that interferes with DNA replication and induces the SOS system.

Correspondence: P.E. Hartman, Department of Biology, The Johns Hopkins University, 34th and Charles Sts., Baltimore, MD 21218 (U.S.A.). * Current address: Center for Photochemical Sciences, De-

partment of Chemistry, Bowling Green State University, Bowling Green, OH 43403 (U.S,A.). Abbreviations: 8-AG, 8-azaguanine; TA, 2-thiazole-DL-alanine; TA r, resistant to 2-thiazole-DL-alanine.

0027-5107/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

128 For more than 20 years singlet oxygen has been reported to be one important intermediate in photosensitization (Spikes and Livingston, 1969). It has been difficult, however, to obtain conclusive evidence for the biological significance of singlet oxygen. This ambiguity stems from the methods widely used for generating singlet oxygen. Most singlet oxygen-generating systems that are compatible with biological materials also generate or involve other reactive species whose effects are difficult or impossible to distinguish from those of singlet oxygen. Evidence for the toxic effects of singlet oxygen has been provided for bacteriophage ~X174 (Houba-Herin et al., 1982), the yeast Saccharomyces cereoisiae (Kobayashi and Ito, 1976), mouse mammary carcinoma cells (Weishaupt et al., 1976) and several species of bacteria (e.g. Bezman et al., 1978; Banks et al., 1985). In addition to toxic effects on bacteriophages and cells, singlet oxygen has been reported to oxidize free guanine and deoxyguanosine (Rosenthal and Pitts, 1971; Cadet and Teoule, 1978; Cadet et al., 1983; Houba-Herin et al., 1982; Midden and Wang, 1983; Wang and Midden, 1983) as well as deoxyguanosine in single-stranded DNA (Kawanishi et al., 1986). Singlet oxygen has been purported to be a base-substitution mutagen for bacteria (Gutter et al., 1977) and to induce gene conversion events in yeast (Kobayashi and Ito, 1976). The mutagenic capability of singlet oxygen is of great interest because singlet oxygen is very likely present in numerous biological systems (cf. Krinsky, 1984; Naqui et al., 1986) and has a relatively long lifetime in aqueous solvents (Pryor, 1986). Furthermore, singlet oxygen may be involved in Fanconi's anemia, an inherited cancer-prone chromosomal instability disease (Joenje et al., 1987). Finally, singlet oxygen may be a participant in a number of clinical therapies, in particular the photochemotherapy of tumors (e.g. Weishaumpt et al., 1976; Spikes, 1984; Spikes and Straight, 1985; Spikes and Jori, 1986; Oseroff et al., 1986; Dougherty, 1987). We recently described a system for the efficient exposure to pure singlet oxygen of bacteria collected on membrane filters. Elsewhere we have provided evidence that the system allows exposure to singlet oxygen in the absence of other reactive

oxygen species and excited-state sensitizer (Midden and Wang, 1983; Dahl et al., 1987). This novel system allowed the demonstration of the potent cytotoxicity of singlet oxygen when generated outside the bacterial cells (Dahl et al., 1987). The unequivocal assignment of singlet oxygen as the intermediate responsible for biological effects is one of the strongest assets this system offers. In the present study we employ tiffs bacterial exposure system to assay directly the potential mutagenicity of pure exogenous singlet oxygen. In contrast to initial expectations, exogenous singlet oxygen appears nonmutagenic for bacteria. Materials and methods

Growth and collection of bacteria. Bacteria were grown to stationary phase in liquid cultures, sedimented by centrifugation, washed, resuspended, serially diluted and filtered with membrane filters as previously described (Dahl et al., 1987). Singlet oxygen exposure. The exposure set-up has been described previously (Dahl et al., 1987). In brirf, bacteria are collected on membrane filters placed a short distance ( < 1 mm) from an illuminated photosensitfzer, rose bengal, which serves as a source of singlet oxygen. This physical separation ensures that singlet oxygen is the only reactive species available to interact with the bacteria (see Fig. 1). No significant reduction in viability was observed in this system for bacteria on filters illuminated for up to 2 h without sensitizer or with the sensitizer plate turned upsidedown, whereby singlet oxygen exposure was prevented but all other conditions (light, temperature, etc.) were kept constant. Reversion assays. Following illumination, filters were placed directly on minimal E medium (Vogel and Bonner, 1956) plus 0.3% (v/v) liquid Difco nutrient broth to allow some divisions of all cells but not heavy backgrounds of unreverted his- auxotrophs. Plates were incubated at 37°C until microcolonies appeared large enough to score under a dissecting microscope (7 x magnification, oblique light source). Typical incubation times for

129

~hv A I

B rm

~

filters either placed on the surfaces of the media directly or exposed to a dose of singlet oxygen capable of killing about 85% of the cells and then plated. Plates were incubated at 37 ° C for up to 100 h and scored as described above. Strains TA1975 and TA2410 in a parallel assay were exposed to an 85% killing dose of short-wave ultraviolet light and scored for induction of mutations.

I

C Fig. 1. Diagram of bacterial exposure to pure singlet oxygen (not drawn to scale). (A) Glass plate coated on lower side with photosensitizer (rose bengal). (B) Spacer. (C) Membrane filter on which bacteria have been deposited. Physicalseparation of the bacteria from the photosensitizerprevents exposure of the bacteria to any reactivespecies other than singletoxygen,while allowing singlet oxygen to diffuse across the separation distance.

supplemented E medium were 48-72 h. All assays were performed using at least duplicate filters.

Test for small in-phase deletions. Revertant colonies of hisG428 were picked from filters with a thin platinum wire and streaked radially onto E plates spread with triphenyl tetrazolium chloride (0.05 m g / m l final concentration). A sterile 6-mm filter paper disk was placed in the center of the plate, and 25/~1 of 20 m g / m l 2-thiazole-DL-alanine (TA) was added to the disk as previously described (Levin et al., 1984). Plates were incubated 24 h and scored for resistance to 2-thiazole-DLalanine (TAr). Forward mutation assays. Assays with 8azaguanine (8-AG) were performed after Skopek et al. (1978) with the following modifications: 2.5-ml aliquots of overlay medium were made in a Constantemp hot block (model 1A, Roeco mfg. service, Monterey Park, CA) set at 45 ° C. Final concentrations of additions were: agar, 0.6%; E medium, single strength; glucose, 0.2%; proline, 20 /~g/ml; histidine, 0.1 mM; biotin, 2.5/~M; 8-AG, 50/~g/ml; dimethyl sulfoxide, 1.7%. Mixtures were poured over minimal E plates enriched with 1.25% (v/v) nutrient broth and incubated overnight. Bacteria were collected at 107 cells/filter and the

Reconstruction assays for survival. We have earlier demonstrated that singlet oxygen exposure of populations of bacterial cells is independent of cell concentrations between 102 and 107 cells/filter (Dahl et al., 1987). In order to obtain countable numbers of colonies while exposing large numbers of bacteria on filters, strain TA4125 (arg-) and the strain to be scored were mixed at final concentrations of 107 TA4125/ml and 102/ml of the strain to be scored. Filters were plated, with or without exposure, on the medium described for forward mutation assays, less the 8-AG. Preliminary experiments demonstrated that this medium was equivalent to nutrient agar plates for survival assays when 102-103 cells/filter (countable numbers of colonies) were plated. Tests for chromosomal duplication. Minimal malate and selective media for aroC duplications were made as previously described (Hoffmann et al., 1978, 1983). Cultures of TA1975 and TA2410 (for malate) and TSl121 (aroC) were prepared as described above, and deposited on filters at densities of 2 × 106 cells/filter. Bacteria were exposed to singlet oxygen for various lengths of time and plated on either malate medium (strains TA1975, TA2410) or aroC duplication-selective medium (strain TSll21). Plates were incubated at 37°C until numbers of visible colonies remained constant, as long as 96 h. Induction of lambda prophage. Escherichia coli strain K12 lysogenic for wild-type bacteriophage lambda was grown overnight in tryptone broth (Arber et al., 1983) plus 1 /~g/ml thiamine HCL The bacteria were sedimented by centrifugation, washed, diluted and collected on filters by a procedure previously described (Dahl et al., 1987). The final concentration of bacteria w a s 1 0 7

130

cells/filter. 4 filters were prepared simultaneously: one unexposed, one exposed to singlet oxygen for 20 rain (a 70% killing dose for this strain), and one each exposed to 240 or 480 # J / c m 2 short-wave ultraviolet light from a Westinghouse Sterilamp as measured with a General Electric Haynes germicidal ultraviolet intensity meter. Each filter was then placed in 3 ml tryptone broth pre-warmed to 37°C and incubated 1 h in the dark with gentle aeration. Following incubation cells were lysed with chloroform. Lysates were assayed for phage particles using E. coli C600 as the indicator strain on tryptone agar plates (Arber et al., 1983) supplemented with thiamine.

Results

Results of the histidine reversion assays are shown in Table 1. The primary targets for reversion include base substitution and frameshift mutations, both alone and in combination with various DNA-repair capacities. Some strains also possess rfa mutations which render the strains more sensitive to the killing effects of exogenous singlet oxygen (Dahl et al., 1987). The two base substitution targets used (hisG46, hisG428) together detect all possible base substitution changes (P.E. Hartman et al., 1986). Since singlet oxygen has been shown to preferentially react with free

TABLE 1 RESULTS OF SALMONELLA his REVERSION ASSAYS FOR SINGLET OXYGEN-INDUCED MUTATIONS his target

Strain

Pertinent genetic markers

Average his + revertants/filter a Unexposed

35% survival

4 6 100 9 13 122 9 4

2 4 53 5 14 84 8 2

his G46

hisG46 TA1975 TA1530 TA1950 TS24 TA2410 GW19 SB4278

hisG46 hisG46 rfa hisG46 A uorB-gal hisG46 AuvrB (gal +) hisG46 recA l hisG46 A uvrB pKM101 hisG46 recA A uvrB hisG46 recA A uvrB rfa

hisG428

hisG428 TA102 TA2659 TA2898

hisG428 hisG428 rfa pKM101 hisG428 A uvrB hisG428 recA

2 53 4 2

2 46 4 1

TA88 TA89 TA90

hisD6610 hisO1242 hisD6610 hisO1242 AuvrB hisD6610 his01242 A uvrB rfa hisD6610 his01242 A uvrB rfa pKM101

8 13

7 12

7

5

17

15

hisD6610 01242

TA97

hisD3052

hisD3052 TA94 TA98 TA1534 TA1538 TA1964 TA2641 UTH8413

hisD3052 hisD3052 pKM101 hisD3052 AuorB rfa pKM101 hisD3052 A uvrB hisD3052 A uvrB rfa hisD3052 A uvrB hisD3052/t uvrB pKM101 hisD3052 A uorB rfa pKM101

5 3 5 5 1 2 2 8

2 3 7 5 2 2 3 6

hisD6580

TA95 TA96

hisD6580 his01242 pKM101 hisD6580 his01242 A uvrB pKM101

8

5

11

12

a Values averaged from 2 Expts., each performed with duplicate filters (4 filters total).

131 TABLE 2 ANALYSIS O F T H I A Z O L E A L A N I N E - R E S I S T A N T (TA r) S M A L L IN-PHASE D E L E T I O N S A M O N G R E V E R T A N T S O F hisG428 N u m b e r of TAr/total his + colonies screened % Unexposed (spontaneous) 35% survival

30/54 33/56

56 59

guanine or guanine residues in single-stranded DNA (see Introduction), both base substitution and frameshift targets rich in GC base pairs ( hisG46, his D3052, hisD6610) have been included for scrutiny. The strains used also include two sets that have been reported to respond to agents capable of causing oxidative damage, namely hisG428 and the pKM101 derivatives of hisD6580 (P.E. Hartman et al., 1986). In addition, hisD3052 detects - 2 , - 5 , - 8 , -11, +1, +4, +7 and +10 deletion and addition frameshift mutations (O'Hara and Marnett, 1987; P.E. Hartman et al., 1986). In no case did we observe an increase in revertant frequency for a dose of singlet oxygen capable of kilting approximately 65% of the exposed bacteria. Besides base-pair substitutions, the hisG428 target can revert by small in-phase (3 or 6 bp) deletions (Levin et al., 1984; P.E. Hartman et al., 1986). Anoxic treatment of hisG428 depresses this

class of events, perhaps implicating reactive oxygen species in the mechanism (Z. Hartman et al., 1984). In addition to assaying the total frequency of reversion for this strain, we assayed the relative frequency of this particular type of revertant among bacteria exposed to singlet oxygen. These results, shown in Table 2, indicate no difference in the frequency of small in-phase deletions relative to total reversion frequency between exposed cells and spontaneous events. In our tests for forward mutations to 8azaguanine resistance (gpt-), we included proline in the plating medium to allow detection of any large deletions extending into proline genes that are closely linked to the gpt locus (Kirsh et al., 1978). These results are given in Table 3. Analyses of the means and variances for the mutation frequencies indicate that there is no significant increase in mutation frequency following singlet oxygen exposure for 4 strains of bacteria killed to 15% survival. Exposure of the two recA ÷ strains (TA1975, TA2410) to short-wave ultraviolet light at similar killing doses yielded mutation frequencies of 1.8 × 10 -4 (strain TA1975) and 2.1 × 1 0 - 4 (strain TA2410), which are significant increases over the spontaneous mutation frequencies of 44fold and 38-fold, respectively. It has been argued that for a weak mutagen that is also a potent toxicant, toxicity may mask mutagenicity at all but a narrow range of doses (Thilly, 1985). Since we have previously demonstrated the potent cytotoxicity of pure singlet

TABLE 3 RESULTS OF S A L M O N E L L A 8 - A Z A G U A N I N E R E S I S T A N C E ASSAYS F O R S I N G L E T O X Y G E N - I N D U C E D TIONS Strain

TA1975 SIM278 GW19 TA2410

Relevant genotype

rfa uvrB recA rfa uorBrecA uvrBpKM101

Spontaneous mutation Singlet oxygen frequency exposure ( x l 0 - 6 viable cells) min moles b ~ +S.D.

n a

4.15:4.8 2.45:2.5 12.3+6.1 5.55:3.4

20 21 20 20

15 15 25 25

1.0xl0 1.0X10 1.7×10 1.7×10

-5 -5 5 .5

MUTA-

% survival 2 5: S.D.

n

Mutation frequency ( x 1 0 - 6 viable cells) c ~ 5: S.D. n

15+ 9 155:14 165:11 13:t: 6

10 10 10 8

5.3+ 6.1 3.15:4.7 10.8+12.3 4.5+ 3.0

12 16 8 12

a Mean + standard deviation; n, number of filters tested, b As determined from dosimetry experiments in Dahl et al. (1987). c A t test indicates no significant differences between the spontaneous mutation frequencies and the mutation frequencies in bacteria exposed to singlet oxygen.

132 TABLE 4 DOSE RESPONSES F O R his- REVERSION ASSAYS IN SALMONELLA Exposure time (rain): 0 Singlet oxygen (moles): 0

2 1.3)<10 -6

5 3.3x10 -6

10 6 . 5 x 1 0 -6

15 20 1 . 0 x l 0 -5 1.3x10

25 5 1.7x10

5

30 2.0x10

5

Strain TA1975

SB4278

GW19

TA2410

% survival Revertants/107 bacteria filtered a

100

% survival Revertants/107 bacteria filtered

100

% survival Revertants/107 bacteria filtered

100

% survival Revertants/107 bacteria filtered

100

45: 2(8)

3 ± 2(8)

95: 4(8)

1205:13(12)

-

3___ 1(4) -

50

4±

24

1(4)

53

2 ± 2(4)

1 ± 1(4)

-

-

9±13(12)

6 + 4(4)

-

115±18(12)

72

104+15(12)

15

3 ± 1(4) 26

48

2±

l(4)

37

5 ± 2(4)

106±14(8)

1(4)

15

1 + 1(4)

51

1±

55: 3(8) -

90+16(8)

3

-

-

1 ± 2(4)

-

-

1

-

-

-

-

16

4

2±

1(4)

-

4+

2(4)

3±

1(12)

21

13

835:20(8)

55+13(8)

2+1(8) 2

48+8(4)

a Numbers shown for revertants are mean ± standard deviation (n), where n = total number of filters tested.

oxygen for bacteria (Dahl et al., 1987) and have yet to find any evidence for mutagenicity in the same system, we considered it possible that singlet oxygen may fall into this category of agents. We have therefore performed dose dependence analyses for 4 hisG46 strains (G/C-rich target sequence for mutation) to address this issue. 2 of the 4 strains carried additional rfa mutations blocking extensive cell wall biosynthesis and thus sensitizing the bacteria to the lethal action of singlet oxygen (Dahl et al., 1987). These same mutations may also allow greater penetration into the cell of exogenous singlet oxygen. Table 4 shows the results of various doses yielding up to 99% toxicity for bacteria exposed to singlet oxygen. At no dose did we see an increase in mutagenic response. In fact, the number of revertants appears to drop at the higher doses, indicating that we may not be recovering spontaneous levels of reversion at low survival values. Another system used to examine the ability of singlet oxygen generated outside the cells to cause genetic damage was the frequency of duplication of a large portion of the Salmonella chromosome (Hoffmann et al., 1978, 1983, 1985). The results

using the aroC locus marker (strain TSl121) are shown in Table 5. No significant differences in the frequency of duplication between unexposed cells and those exposed to as much as an 85% killing dose of singlet oxygen were observed. Exposure of filtered bacteria to 240 # J / c m 2 far ultraviolet light increased the duplication frequency approximately 20-fold (Table 5), which is consistent with data earlier reported by Hoffmann and coworkers (1985). Similar results were obtained using growth on malate medium (Hoffmann et al., 1978)

TABLE 5 TESTS F O R T H E I N D U C T I O N OF G E N E T I C DUPLICAT I O N BY S I N G L E T O X Y G E N IN AN aroC321 STRAIN OF SALMONELLA Strain

Exposure

% survival

Duplication frequency

TSll21

Singlet oxygen

100 50 30 15

6.4 x 1 0 - 5 3.6 X 10- 5 1.8 x 10-5 1.2x10 -5

10

1.3 X 10- 4

(aroC321 hisG46)

Ultraviolet light

133 TABLE 6 TESTS F O R S I N G L E T O X Y G E N I N D U C T I O N O F L A M B D A P R O P H A G E IN A N Escherichia coli L A M B D A LYSOGEN Sample

Plaque-forming u n i t s / 10 7 cells exposed

Unexposed

9.1×10 2

1,3 × 10 -5 moles singlet oxygen

3.9×10 2

2 4 0 / x J / c m 2 ultraviolet fight 480 t t J / c m 2 ultraviolet light

2.4 × 10 6 2.8×10 6

as an alternate indicator of induction of gene duplications (data not shown). Induction of prophage lambda provides a sensitive test for agents that affect DNA replication either by producing lesions in the DNA or by interfering with the replication fork of the DNA (Roberts and Devoret, 1983; Elespuru, 1987). Table 6 shows our results with singlet oxygen exposure of a lambda lysogen. A singlet oxygen exposure of 20 min that killed 70% of the bacteria did not induce lambda in this strain. Exposure to short-wave ultraviolet light demonstrated that the lysogen was inducible. Discussion

The most commonly used systems for the laboratory investigation of the biological importance of singlet oxygen employ photodynamic sensitizers. Previous reports by others (see Introduction) have suggested that singlet oxygen is toxic, and also may be genotoxic and mutagenic. Participation of singlet oxygen in photodynamic systems is usually ambiguous, however, since singlet oxygen is only one possible reactive molecular species amid a plethora of reactive species in complex biological systems (Krinsky, 1979; Foote, 1981). Assignment of reactive intermediacy to singlet oxygen is therefore ordinarily a difficult problem in photosensitized oxidations, even under controlled laboratory conditions. The difficulties in being able to separate the effects of singlet oxygen from other products is underscored by the matter of frank strand breaks in exposed DNA. While the commonly used singlet oxygen generator, rose bengal, when il-

luminated causes single-strand breaks in DNA, these strand breaks are not likely due to singlet oxygen since they can arise even in the absence of oxygen (Peak et al., 1984; CiuUa et al., 1986). Thus, a pure source of singlet oxygen is needed which eliminates the possibility of non-singlet oxygen-induced DNA damage which can arise from treatment with illuminated rose bengal in solution. For example, strand breaks were n o t found even in single-stranded DNA when a larger amount of singlet oxygen was generated chemically by the breakdown of an endoperoxide (Nieuwint et al., 1985; Lafleur et al., 1986). Neither is chemical generation of singlet oxygen ideal for these studies, since the chemicals used are often potent oxidants themselves, capable in some cases of direct interaction with the substrate, and the presence of these chemicals and their decomposition products complicates interpretation of the results. Because of the relatively long lifetime of singlet oxygen (diffusion radius = 100-200 nm in pure water. Lindig and Rodgers, 1981) and its ability to penetrate membranes (Gorman et al., 1976), singlet oxygen generated in the cytoplasm of an eukaryotic cell - or even outside the cell (Schiff et al., 1985) - conceivably could interact with the DNA in the nucleus (Gruener and Lockwood, 1979; Schiff et al., 1985). Several in vitro studies using nucleic acids and their components as models for singlet oxygen reactivity with cellular DNA indicate that singlet oxygen might be a direct-acting mutagen (see Introduction). Wefers and co-authors (1987) inactivated pBR322 with microwave-discharge-generated singlet oxygen. Decupyer-Debergh et al. (1987) have observed mutagenesis in the lacZ gene cloned into M13 bacteriophage, using the separated-surface-sensitizer system of Midden and Wang (1983). However, both of these groups (Wefers et al., 1987 and Decuyper-Debergh et al., 1987) have used Tris buffer, which can itself form DNA adducts when illuminated in the presence of methylene blue (Van Vunakis et al., 1966). Kawanishi et al. (1986), Hildebrand et al. (1987) and Blazek et al. (1987) have demonstrated that DNA can be altered by singlet oxygen. Hildebrand et al. (1987) have compared the effect in singlestranded DNA with the degradation of free

134

guanine and deoxyguanosine in the separatedsurface-sensitizer system; they have found that singlet oxygen is approximately 103 times less reactive with guanine residues in single-stranded DNA than with the free base or deoxynucleoside. Further, singlet oxygen may be even less reactive toward double-stranded DNA than single-stranded DNA, except in special cases such as when generated by dyes which tightly bind to DNA, probably by intercalation between DNA bases (Kawanishi et al., 1986). Kobayashi and Ito (1976) earlier reported that only those photosensitizers which both penetrate yeast cells and bind to D N A induce genetic changes in yeast, an observation which correlates nicely with the findings of Kawanishi and co-workers (1986). Thus, when DNA is the direct target of singlet oxygen damage, the conformational structures of both the DNA and the sensitizer used may be extremely important (e.g. Kobayashi and Ito, 1976; Nieuwint et al., 1985; Lafleur et al., 1986). However, genotoxicity does not necessarily require a direct reaction of singlet oxygen with DNA. Singlet oxygen is reactive toward cellular components other than DNA, for example proteins and fatty acids (e.g. Straight and Spikes, 1985). An initial product formed from reaction with singlet oxygen and other cellular components might be capable of reacting with DNA. This two-step mode of action seems unlikely for bacteria, however, based on the results reported here and in an earlier study in which we exposed bacteria to pure exogenous singlet oxygen. We previously demonstrated that toxicity was independent of a variety of DNA-repair capacities, and concluded that DNA damage was not a predominant pathway in singlet oxygen toxicity (Dahl et al., 1987). We have tested now the ability of pure singlet oxygen to damage bacterial DNA in a variety of bacterial mutagenesis and gene duplication assays. Our findings show that pure singlet oxygen generated outside the bacterial cells exerts tremendous cytotoxic damage without interacting with the bacterial chromosome in ways leading to a significant increase in base-pair substitutions, frameshift mutations, small or large deletions, large duplications, or interference with DNA replication (SOS induction). To offer some indication of the disparity between singlet oxygen-induced

toxicity and DNA damage, Hildebrand et al. (1987) have observed a small increase in singlestranded DNA damage over a time course of 9-12 h, under very nearly the same conditions in which we have reduced bacterial viability 5 orders of magnitude in 40 min (Dahl et al., 1987). It seems unlikely, therefore, that singlet oxygen is highly mutagenic, except possibly under certain conditions such as when sensitizer is bound directly to DNA. The substantial majority of biologically relevant sources of singlet oxygen do not appear to perform in this manner. The biological roles of various reactive oxygen species in the antibacterial action of mammalian phagocytic cells remains unclear, but singlet oxygen has been one component seriously incriminated (Krinsky, 1979, 1984). In view of the potent lethal but absent or negligible mutagenic action of exogenous singlet oxygen, it seems that singlet oxygen might serve as a very powerful, but relatively innocuous (to the host), mammalian defense mechanism. Its relatively clean, apparently nonmutagenic (at least to bacteria), cytotoxic properties suggest singlet oxygen might also be an excellent choice as a therapeutic tool for the destruction of unwanted tissue (cf. Spikes, 1984; Spikes and Jori, 1986; Dougherty, 1987).

Acknowledgements We thank Drs. B.N. Ames, D. Maron and G.R. Hoffmann for supplying bacterial strains and Ms. Geraldine Chester for expert technical assistance. This work was supported in part by grants HD07103 (T.A.D.), ES02300 (W.R.M.) and ES03217 and BRSG RR07041 (P.E.H.) from the National Institutes of Health. Contribution No. 1381 of the Department of Biology, The Johns Hopkins University.

References Arber, W., L. Enquist, B. Hohn, M.E. Murray and K. Murray (1983) in: R.W. Hendrix, J.W. Roberts, F.W. Stahl and R.A. Weisberg (Eds.), Lambda II, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 433-466. Banks, J.G., R.G. Board, J. Carter and A.D. Dodge (1985) The cytotoxic and photodynamic inactivation of microorganisms by rose bengal, J. Appl. Bacteriol., 58, 391-400. Bezman, S.A., P.A. Burtis, T.P.J. Izod and M.A. Thayer (1978) Photodynamic inactivation of E. coli by rose bengal im-

135 mobilized on polystyrene beads, Photochem. Photobiol., 28, 325-329. Blazek, E.R., W.R. Midden and M.J. Peak (1987) Singlet oxygen induces piperidine-labile sites in isolated plasmid DNA, Photochem. Photobiol., 45S, 9S. Cadet, J., and R. Teoule (1978) Comparative study of the oxidation of nucleic acid components by hydroxyl radicals, singlet oxygen, and superoxide anion radicals, Photochem. Photobiol., 28, 661-667. Cadet, J., C. Decarroz, S.Y. Wang and W.R. Midden (1983) Mechanisms and products of photosensitized degradation of nucleic acids and related model compounds, Israel J. Chem., 23, 420-429. Ciulla, T.A., J.R. Van Camp and I.E. Kochevar (1986) Oxygen independent photosensitization of DNA by rose bengal, Photochem. Photobiol., 43S, 30S. Dahl, T.A., W.R. Midden and P.E. Hartman (1987) Pure singlet oxygen cytotoxicity for bacteria, Photochem. Photobiol., 46, 345-352. Decupyer-Debergh, D., J. Piette and A. Van de Vorst (1987) Singlet oxygen-induced mutations in M12 lacZ phage DNA, EMBO J., 6, 3155-3161. Dougherty, T.J. (1987) Photosensitizers: Therapy and detection of malignant tumors, Photochem. Photobiol., 45, 879-889. Elespuru, R.K. (1987) Inducible responses to DNA damage in bacteria and mammalian cells, Environ. Mol. Mutagen., 10, 97-116. Foote, C.S. (1981) Photooxidation of biological model compounds, in: M.A.J. Rodgers and E.L. Powers (Eds.), Oxygen and Oxy-Radicals in Chemistry and Biology, Academic Press, New York, pp. 425-439. Gorman, A.A., G. Lovering and M.A.J. Rodgers (1976) The photosensitized formation and reaction of singlet oxygen, O2*(aA), in aqueous micellar systems, Photochem. Photobiol., 23, 399-403. Gruener, N., and M.P. Lockwood (1979) Photodynamic mutagenicity in mammalian cells, Biochem. Biophys. Res. Commun., 90, 460-465. Gutter, B., W.T. Speck and H.S. Rosenkranz (1977) A study of the photoinduced mutagenicity of methylene blue, Mutation Res., 44, 177-182. Hartman, P.E., B.N. Ames, J.R. Roth, W.M. Barnes and D.E. Levin (1986) Target sequences for mutagenesis in Salmonella histidine-requiring mutants, Environ. Mutagen., 8, 631-641. Hartman, Z., P.E. Hartman, W.M. Barnes and E. Tuley (1984) Spontaneous mutation frequencies in Salmonella: Enhancement of G / C to A / T transitions and depression of deletion and frameshift mutation frequencies afforded by anoxic incubation, Environ. Mutagen., 6, 633-650. Hildebrand, E., W.R. Midden and B.L. Murr (1987) DNA damage induced by singlet oxygen, Photochem. Photobiol., 45S, 8S. Hoffmann, G.R., M.J. Walkowicz, J.M. Mason and J.F. Atkins (1983) Genetic instability associated with the aroC321 allele in Salmonella typhimurium involves genetic duplication, Mol. Gen. Genet., 190, 183-188. Hoffmann, G.R., L.S. Catuogno, J.F. Linnane and L.A. Parente (1985) Effects of DNA-repair processes on the induction of

genetic duplications in bacteria by ultraviolet light, Mutation Res., 151, 25-33. Hoffmann, G.R., R.W. Morgan and R.C. Harvey (1978) Effects of chemical and physical mutagens on the frequency of a large genetic duplication in Salmonella typhimurium, Mutation Res., 52, 73-80. Houba-Herin, N., C.M. Calberg-Bacq, J. Piette and A. Van de Vorst (1982) Mechanisms for dye-mediated photodynamic action: singlet oxygen production, deoxyguanosine oxidation, and phage inactivating efficiencies, Photochem. Photobiol., 36, 297-306. Joenje, H., A.W.M. Niewint, A.B. Oostra, F. Arwert, H. de Koning and K.J. Roozendaal (1987) Cytogenetic toxicity of paraquat and streptonigrin in Fanconi's anemia, Cancer Genet. Cytogenet., 25, 37-45. Kawanishi, S., S. Inoue, S. Sano and H. Aiba (1986) Photodynamic guanine modification by hematoporphyrin is specific for single-stranded DNA with singlet oxygen as mediator, J. Biol. Chem., 261, 6090-6095. Kirsh, M., D.R. Dembinski, P.E. Hartman and C.G. Miller (1978) Salmonella typhimurium peptidase active on carnosine, J. Bacteriol., 134, 361-374. Kobayashi, K., and T. Ito (1976) Further in vivo studies on the participation of singlet oxygen in the photodynamic inactivation and induction of genetic changes in Saccharomyces cerevisiae, Photochem. Photobiol., 23, 21-28. Krinsky, N. (1979) Biological roles of singlet oxygen, in: H.H. Wasserman and R.W. Murray (Eds.), Singlet Oxygen, Academic Press, New York, pp. 597-641. Krinsky, N. (1984) Biology and photobiology of singlet oxygen, in: W. Bors, M. Saran and D. Tait (Eds.), Oxygen Radicals in Chemistry and Biology, Waiter de Gruyter, New York, pp. 473-478. Lafleur, M.V.M., A.W.M. Niewint, J.M. Aubry, H. Kortbeek, F. Arwert and H. Joenje (1986) DNA damage by chemically generated singlet oxygen, Free Radic. Res. Commun. 2, 343-350. Levin, D.E., L.J. Marnett and B.N. Ames (1984) Spontaneous and mutagen-induced deletions: Mechanistic studies in Salmonella tester strain TA102, Proc. Natl. Acad. Sci. (U.S.A.), 81, 4457-4461. Lindig, B.A., and M.A.J. Rodgers (1981) Rate parameters for the quenching of singlet oxygen by water-soluble and lipid-soluble substrates in aqueous and micellar systems, Photochem. Photobiol., 33, 627-634. Midden, W.R., and S.Y. Wang (1983) Singlet oxygen generation for solution kinetics: Clean and simple, J. Am. Chem. Soc., 105, 4129-4135. Naqui, A., B. Chance and E. Cadenas (1986) Reactive oxygen intermediates in biochemistry, Annu. Rev. Biochem., 55, 137-166. Niewint, A.W.M., J.M. Aubry, F. Arwert, H. Kortbeek, S. Herzberg and H. Joenje (1985) Inability of chemically generated singlet oxygen to break the DNA backbone, Free Radic. Res. Commun., 1, 1-9. O'Hara, S.M., and L.J. Marnett (1987) DNA sequence analysis of spontaneous and /~-methoxyacrolein induced frameshift mutations in Salmonella typhimurium hisD3052, Fed. Proc., 46, abstr. 64, 1936.

136 Oseroff, A.R., D. Ohuoha, T. Hasan, J.C. Bommer and M.L. Yarmush (1986) Antibody-targeted photolysis: Selective photodestruction of human T-cell leukemia cells using monoclonal antibody-chlorin e6 conjugates, Proc. Natl. Acad. Sci. (U.S.A.), 83, 8744-8748. Peak, M.J., J.G. Peak, C.S. Foote and N.I. Krinsky (1984) Oxygen-independent direct deoxyribonucleic acid backbone breakage caused by rose bengal and visible light, J. Photochem., 25, 309-315. Pryor, W.A. (1986) Oxy-radicals and related species: Their formation, lifetimes and reactions, Annu. Rev. Physiol., 48, 657-667. Roberts, J.W., and R. Devoret (1983) Lysogenic induction, in: R.W. Hendrix, J.W. Roberts, F.W. Stahl and R.A. Weisberg (Eds.), Lambda II, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 123-144. Rosenthal, I., and J.N. Pitts (1971) Reactivity of purine and pyrimidine bases toward singlet oxygen, Biophys. J., 11, 963-966. Schiff, L.J., W.C. Eisenberg and K. Taylor (1985) Evidence of DNA repair in organ cultures of hamster tracheal epithelium following exposure to gas phase singlet oxygen, Mutation Res., 142, 41-44. Skopek, T.R., H.L. Liber, J.J. Krolewski and W.G. Thilly (1978) Quantitative forward mutation assay in Salmonella typhimurium using 8-azaguanine resistance as a genetic marker, Proc. Natl. Acad. Sci. (U.S.A.), 75, 410-414. Spikes, J.D. (1984) Photobiology of porphyrins, in: D.R. Doiron and C.J. Gomer (Eds.), Porphyrin Localization and Treatment of Tumors, Liss, New York, pp. 19-39. Spikes, J.D., and G. Jori (1986) Photodynamic therapy of turnouts and other diseases using porphyrins, Lasers Med. Sci., 2, 3-15.

Spikes, J.D., and R. Livingston (1969) The molecular biology of photodynamic action: Sensitized photoautoxidations in biological systems, Adv. Radiat. Biol., 3, 29-121. Spikes, J.D., and R.C. Straight (1985) Photodynamic behavior of porphyrins in model cell, tissue and tumor systems, in: G. Jori and C. Perria (Eds.), Photodynamic Therapy of Tumors and Other Disease, Libreria Progetto, Padova, pp. 45-53. Straight, R.C., and J.D. Spikes (1985) Photosensitized oxidation of biomolecules, in: A.A. Frimer (Ed.), Singlet Oxygen, Vol. IV, CRC, Boca Raton, FL, pp. 91-143. Thilly, W.G. (1985) Dead cells don't form colonies: a serious source of bias in mutation assays, Environ. Mutagen., 7, 255-258. Van Vunakis, H., E. Seaman, L. Kahan, J.W. Kappler and L. Levine (1966) Formation of an adduct with tris(hydroxymethyl)aminomethane during the photooxidation of deoxyribonucleic acid and guanine derivatives, Biochemistry, 5, 3986-3991. Vogel, H.J., and D.M. Bonner (1956) Acetylornithinase of Escherichia coli: Partial purification and some properties, J. Biol. Chem., 218, 97-106. Wang, S.Y., and W.R. Midden (1983) Photodynamic action on nucleic acids and their components, Stud. Biophys., 94, 7-12. Wefers, H., D. Schulte-Frohlinde and H. Sies (1987) Loss of transforming activity of plasmid DNA (pBR322) in E. coli caused by singlet molecular oxygen, FEBS Lett., 211, 49-52. Weishaupt, K.R., C.J. Gomer and T.J. Dougherty (1976) Identification of singlet oxygen as the cytotoxic agent in photo inactivation of a murine tumor, Cancer Res., 36, 2326-2329.