Archives of Microbiology

, Volume 183, Issue 3, pp 203–208

Sensitivity of dark mutants of various strains of luminescent bacteria to reactive oxygen species

Authors

  • Robert Łyżeń
    • Department of Molecular BiologyUniversity of Gdańsk
    • Institute of OceanologyPolish Academy of Sciences
    • Department of Molecular BiologyUniversity of Gdańsk
    • Institute of OceanologyPolish Academy of Sciences
Original Paper

DOI: 10.1007/s00203-005-0764-y

Cite this article as:
Łyżeń, R. & Węgrzyn, G. Arch Microbiol (2005) 183: 203. doi:10.1007/s00203-005-0764-y

Abstract

Recent studies indicated that bioluminescence of the marine bacterium Vibrio harveyi may both stimulate DNA repair and contribute to detoxification of deleterious oxygen derivatives. Therefore, it was also proposed that these reactions can be considered biological roles of bacterial luminescence and might act as evolutionary drives in development of luminous systems. However, experimental evidence for the physiological role of luciferase in protection of cells against oxidative stress has been demonstrated only in one bacterial species, raising the question whether this is a specific or a more general phenomenon. Here we demonstrate that in the presence of various oxidants (hydrogen peroxide, cumene hydroperoxide, t-butyl hydroperoxide and ferrous ions) growth of dark mutants of different strains of Vibrio fischeri and Photobacterium leiognathi is impaired relative to wild-type bacteria, though to various extents. Deleterious effects of oxidants on the mutants could be reduced (with different efficiency) by addition of antioxidants, A-TEMPO or 4OH-TEMPO. These results support the hypotheses that (1) activities of bacterial luciferases may detoxify deleterious oxygen derivatives, and (2) significantly different efficiencies of this reaction are characteristic for various luciferases.

Keywords

Bacterial luciferasesOxidative stressAntioxidantsVibrio fischeriPhotobacterium leiognathi

Introduction

Bioluminescence is a phenomenon of light emission by living organisms. This process occurs in various forms of life including bacteria, fungi and animals. Interestingly, among luminescent organisms, light-producing bacteria are the most abundant and widespread, and most of these bacteria occur in marine environments (for reviews see Wilson and Hastings 1998; Węgrzyn and Czyż 2002, and references therein). Although genetics and biochemistry of bacterial luminescence have been investigated extensively, until recently, the biological role of this process remained unclear.

Recent studies indicated that luminescence of marine bacterium Vibrio harveyi stimulates DNA repair, perhaps due to enhanced photoreactivation, as dark mutants of this bacterium revealed significantly higher sensitivity to UV than wild-type cells (Czyż et al. 2000a, b). On the other hand, many years ago it was proposed that bioluminescence did not first evolve for the production of light, but instead as a mechanism of detoxifying an aerobic atmosphere (McElroy and Seliger 1962). Since bacterial luciferase catalyses oxidation of a long-chain aldehyde and reduced flavin mononucleotide, it was speculated that protection against an oxidative stress may also be a role for bacterial luminescence (Rees et al. 1998). An indirect support for this hypothesis could be provided by the finding that V. harveyi luciferase, when overproduced in Escherichia coli cells, may act as a source of superoxide in the absence of the bioluminescence substrate but not in its presence (Gonzales-Flecha and Demple 1994). Nevertheless, the best argument for the proposal of Rees et al. (1998) was evidence for physiological significance of detoxification of reactive oxygen species, mediated by bacterial luciferase in V. harveyi (Szpilewska et al. 2003).

Results of recent studies, summarized above, suggest that stimulation of DNA repair and detoxification of reactive oxygen species might be evolutionary drives for bacterial luminescence, especially at early stages of evolution, and might play important roles in present-day luminescent bacteria (Czyż and Węgrzyn 2001; Węgrzyn and Czyż 2002; Szpilewska et al. 2003). One of the biggest problems with this hypothesis was that experiments showing both luminescence-stimulated DNA repair and luciferase-dependent detoxification of reactive oxygen species were performed using only one bacterial species, V. harveyi (Czyż et al. 2000b; Szpilewska et al. 2003). Therefore, the question appeared whether these phenomena are common for most (if not all) luminous bacteria or represent a very specific reaction of V. harveyi. Although some suggestions were presented (Czyż et al. 2002) that luminescence-dependent response of bacterial cell to UV light and chemical mutagens may also be characteristic for other species, to our knowledge, no actual studies addressing the problem of potential detoxification of deleterious oxygen species in various bacterial strains and their dark derivatives were reported to date. Therefore, we asked whether luminescent and dark strains of different bacterial species respond to oxidants similarly to V. harveyi lux+ cells and lux mutants.

Materials and methods

Bacterial strains

The following wild-type strains of marine luminescent bacteria were used: Vibrio fischeri MJ1 (Ruby and Nealson 1976), Photobacterium leiognathi LN-1a (Dunlap 1985) and P. leiognathi 721 (Ast and Dunlap 2004). For isolation of dark (lux) mutants of these strains, a chemical mutagenesis was performed using 4-nitroquinolone-N-oxide (NQNO) or 2-aminofluorene (2-AF). To bacterial cultures growing exponentially (A575 between 0.2 and 0.8) in the BOSS medium (1% Bacto-peptone, 0.5 % beef extracts, 0.1% glycerol, 3% NaCl; described by Klein et al. 1998) at 30°C, a mutagen was added to various concentrations (NQNO between 20 ng/ml and 100 ng/ml, or 2-AF between 100 ng/ml and 500 ng/ml). Following 1–2 h incubation with shaking at 30°C, serial dilutions were spread onto BOSS agar plates (BOSS medium containing 1.5% bacteriological agar). After incubation of plates at 30°C for 48 h, colonies were tested for luminescence using Fluor-S MultiImager (Bio-Rad), and candidates for dark mutants were further investigated in a liquid culture assay using a luminometer (Junior, EG&G Berthold). In further studies, two independently isolated mutants of each strain were used. These mutants were named lux-1 and lux-2 in each case. All these mutants revealed drastically decreased levels of luminescence relative to their parental strains (Fig. 1).
Fig. 1

Luminescence of wild-type strains and lux mutants of Vibrio fischeri MJ1, Photobacterium leiognathi LN-1a and P. leiognathi 721. Bacteria were grown in BOOS medium at 30°C to A575 of 0.2. Luminescence of culture samples was measured using a luminometer (Junior, EG&G Berthold). AU Arbitrary units

Reagents

Hydrogen peroxide, 4-O-acetyl 2, 2, 6, 6 tetramethyl-piperidin-1-oxyl (A-TEMPO) and 4 OH 2, 2, 6, 6 tetramethyl-piperidin-1-oxyl (4OH-TEMPO) were purchased from Lancaster Synthesis. Cumene hydroperoxide, t-butyl hydroperoxide and ferrous sulphate were purchased from Sigma. Ferrous sulphate solutions were prepared immediately before use from freshly opened vials of the reagent. In pilot experiments, concentrations of oxidants were chosen, in which the biggest differences between growth of wild-type strains and dark mutants were observed. For further experiments, following concentrations of oxidants were used: hydrogen peroxide, 1 mM; cumene hydroperoxide, 0.15 mM; t-butyl hydroperoxide, 0.3 mM and ferrous sulphate, 0.2 mM. A-TEMPO and 4OH-TEMPO were added to final concentration 0.7 mM.

Estimation of effects of oxidants and antioxidants on bacterial growth

Bacteria were grown in the BOSS medium at 30°C to OD575 of about 0.2. Following addition of oxidants and/or antioxidants to indicated concentrations, bacterial growth was monitored by measurement of OD575 of the cultures.

Results

Reactive oxygen species occur naturally in cells most often due to incomplete reduction of molecular oxygen, but they can also arise from a variety of environmental sources, for instance, ionizing radiation or redox cyclic compounds (Sies 1991; Ozben 1998). Examples of such reactive oxygen species are hydrogen peroxide, the superoxide radical and the hydroxyl radical. Cumene hydroperoxide and t-butyl hydroperoxide are chemicals that form radicals, and thus they can be used to provoke the oxidative stress under laboratory conditions. Moreover, transition metal ions, e.g. ferrous ions, enhance formation of hydroxyl radicals (Tabatabaie and Floyd 1996).

Using the compounds mentioned above, we investigated influence of oxidative stress on growth of three different luminescent strains of marine bacteria and their dark derivatives. Growth of V. fischeri MJ1 strain was similar to that of its dark derivatives (lux-1 and lux-2) in the absence of oxidative stress (Fig. 2a). Addition of hydrogen peroxide, cumene hydroperoxide, t-butyl hydroperoxide or ferrous ions caused impairment of V. fischeri MJ1 growth, but complete inhibition of growth of lux-1 and lux-2 mutants (in fact, cumene hydroperoxide and t-butyl hydroperoxide caused cell death, as indicated by a decrease in optical density of bacterial cultures; Fig. 2a).
Fig. 2

Effects of hydrogen peroxide (HP), cumene hydroperoxide (Cumene), t-butyl hydroperoxide (t-butyl) and ferrous sulphate [Fe(II)] on growth of V. fischeri MJ1 (a), P. leiognathi LN-1a (b) and P. leiognathi 721 (c) strains. Growth of wild-type cells (open circles), and lux-1 (filled circles) and lux-2 (filled squares) mutants is shown. Sub-panels marked as Control show results obtained without addition of oxidants, which were otherwise added at time 0 to bacterial cultures to following final concentrations: hydrogen peroxide, 1 mM; cumene hydroperoxide, 0.15 mM; t-butyl hydroperoxide, 0.3 mM and ferrous sulphate, 0.2 mM. Values shown in the figure are representative results. Each experiment was performed at least three times, and high reproducibility was achieved in each case

The lux-1 and lux-2 mutants of P. leiognathi LN-1a grew somewhat slower than their wild-type counterpart (Fig. 2b). Therefore, to estimate effects of oxidative stress, it was reasonable to use concentrations of different compounds that inhibited growth of the parental strain completely or almost completely. Otherwise, it would be difficult to compare effects of the reactive oxygen species on wild-type and mutant cells. Namely, it would be hard to assess whether a decrease in growth rate of the wild-type stain is more or less significant than that of the mutant when initial values are different and growth inhibition of the control strain is not dramatic. In fact, the chosen concentrations were the same as those causing inhibition of V. fischeri MJ1 growth, described above. Under these conditions, deleterious effects of reactive oxygen species were significantly more pronounced in lux mutants than in wild-type bacteria only in the case of addition of cumene hydroperoxide. Other compounds caused only slightly higher sensitivity of the mutants relative to the luminous strain (Fig. 2b).

Dark mutants of P. leiognathi 721 grew similar to the parental strain, but this parental strain was more sensitive to oxidative stress than P. leiognathi LN-1a. The lux mutants were only slightly more sensitive, except for lux-1, which was very sensitive (Fig. 2c).

To test whether deleterious effects of the compounds used in our experiments on the growth of dark mutants of tested strains resulted from their enhanced sensitivity to oxidative stress or from any other putative effects of the chemicals added to bacterial cultures, we investigated effects of antioxidants. Nitroxides, exemplified by A-TEMPO and 4OH-TEMPO, are antioxidants that neutralize deleterious radicals in a catalytic mode. They mimic superoxide dismutase and pro-catalase activities, oxidize reduced metal ions and detoxify hypervalent metals (Rosantsev 1970; Samuni and Krishna 1997; Cadenas 1998; Skórko-Glonek et al. 1999).

Growth and/or survival of V. fischeri lux-1 and lux-2 mutants in the presence of most examined compounds that cause oxidative stress was significantly improved by addition of A-TEMPO or 4OH-TEMPO to bacterial cultures (Fig. 3). Effects of these antioxidants were less pronounced in experiments with mutants of P. leiognathi LN-1a (Fig. 4) and P. leiognathi 721 (Fig. 5). Generally, the smallest protective effects of A-TEMPO or 4OH-TEMPO were observed in experiments with ferrous ions. This may arise from a mechanism by which these ions cause formation of hydroxyl radicals, which is different than those of other tested compounds (Tabatabaie and Floyd 1996). Nevertheless, in most cases where oxidative stress resulted in significant differences in growth of wild-type bacteria and dark mutants, A-TEMPO and 4OH-TEMPO improved survival of these mutants considerably (Figs. 4, 5).
Fig. 3

Rescue by antioxidants, A-TEMPO or 4OH-TEMPO, of survival or growth of lux-1 (a) and lux-2 (b) mutants of V. fischeri MJ1, inhibited by the following oxidants: hydrogen peroxide (HP), cumene hydroperoxide (Cumene), t-butyl hydroperoxide (t-butyl) and ferrous sulphate [Fe(II)]. Sub-panels marked as Control show results obtained without addition of oxidants, which were otherwise added at time 0 to bacterial cultures to following final concentrations: hydrogen peroxide, 1.25 mM; cumene hydroperoxide, 0.15 mM; t-butyl hydroperoxide, 0.3 mM and ferrous sulphate, 0.2 mM. Bacteria were cultured without antioxidants (closed circles) or in the presence of 0.7 mM A-TEMPO (open triangles) or 0.7 mM 4OH-TEMPO (closed triangles). Values shown in the figure are representative results. Each experiment was performed at least three times, and high reproducibility was achieved in each case

Fig. 4

Rescue by antioxidants, A-TEMPO and/or 4OH-TEMPO, of survival or growth of lux-1 (a) and lux-2 (b) mutants of P. leiognathi LN-1a, inhibited by the following oxidants: hydrogen peroxide (HP), cumene hydroperoxide (Cumene), t-butyl hydroperoxide (t-butyl) and ferrous sulphate [Fe(II)]. See legend to Fig. 3 for symbols and experimental details

Fig. 5

Rescue by antioxidants, A-TEMPO and/or 4OH-TEMPO, of survival or growth of lux-1 (a) and lux-2 (b) mutants of P. leiognathi 721, inhibited by the following oxidants: hydrogen peroxide (HP), cumene hydroperoxide (Cumene), t-butyl hydroperoxide (t-butyl) and ferrous sulphate [Fe(II)]. See legend to Fig. 3 for symbols and experimental details

Discussion

The main question asked in this work was whether luciferase-mediated detoxification of the deleterious oxygen derivatives, predicted theoretically (Rees et al. 1998) and confirmed experimentally in V. harveyi (Szpilewska et al. 2003), occurs also in other luminescent bacteria. We considered this question important, because if the function of the luminous system in decreasing effects of oxidative stress were common for most (if not all) light-emitting bacteria, this role might be described as an evolutionarily drive for bacterial luminescence.

Results of our experiments indicate that, similarly to V. harveyi luxA and luxB strains (Szpilewska et al. 2003), dark mutants of V. fischeri are significantly more sensitive to oxidative stress than their luminescent counterparts. This phenomenon was less pronounced in lux derivatives of two different strains of P. leiognathi, although some compounds that provoke oxidative stress caused considerably more deleterious effects on dark mutants than on the wild-type bacteria. Similar differences were observed in rescuing effects of antioxidants (A-TEMPO and 4OH-TEMPO) on growth and survival of bacteria in cultures treated with various compounds causing oxidative stress. Therefore, we conclude that luciferase-mediated detoxification of the deleterious oxygen derivatives may be a general reaction occurring in bacterial cells; however, efficiency of this process is different in various species and strains.

On the basis of estimation of effects of treatment of bacteria with hydrogen peroxide on luminescence efficiency, and because of demonstration that luciferase and catalase act independently, Katsev et al. (2004) speculated that luciferases with slow kinetics may be more efficient in detoxification of reactive oxygen species than those with fast kinetics. Results of our studies may support this proposal, as varying efficiency of luciferase-mediated protection of cells against oxidative stress in different bacterial strains can be concluded.

This report indicates that protection of cells against oxidative stress, although of varying efficiency in different species and strains, appears to be a general phenomenon occurring in luminescent bacteria. Thus, this process, together with stimulation of DNA repair, could be evolutionary drives of luminescence. During early evolution, the mechanisms of selection based on these functions could operate at stages when first luciferases appeared, at a time when efficiency of luminescent systems was too weak to produce light detected by a naked animal eye but sufficiently high to stimulate photoreactivation and detoxification of reactive oxygen species. After improved luminescent systems, which produced light sensed by animals, were formed, other evolutionary drives could start to operate. This probably led to establishment of ecological benefits of bioluminescence, which resulted in formation of sophisticated processes, reactions and behaviours, including symbiosis between luminescent bacteria and animals. Nevertheless, stimulation of DNA repair and protection of cells against oxidative stress may be still important biological roles of bioluminescence in cells of present-day free-living bacteria.

Finally, it is interesting that activity of luciferase in detoxification of the deleterious oxygen derivatives (Szpilewska et al. 2003, this work) and effects of reactive oxygen species, like hydrogen peroxide, on luminescence (Katsev et al. 2004) are not restricted to bacteria. Apart from previous findings that coelenterazine, a substrate in the bioluminescence reaction in many marine animals, may act as an antioxidant (reviewed and discussed by Rees et al. 1998), it was proposed recently that it is hydrogen peroxide that triggers the flash of firefly (Ghiradella and Schmidt 2004). In such a case, rather than simply detoxifying a reactive oxygen species, an organism would be harnessing and using its reactivity to achieve otherwise impossible ends. This proposal is currently attracting attention and a support is developed for it (Ghiradella and Schmidt 2004).

Acknowledgements

This work was supported by the Institute of Oceanology of the Polish Academy of Sciences (task grant no. IV.3.1).

Copyright information

© Springer-Verlag 2005