oxygen toxicity and microbial evolution

Biosystems, 24 (1991) 305-312 Elsevier Scientific Publishers Ireland Ltd.

305

Oxygen toxicity and microbial evolution

Tomasz Bilinski

(Received July lst, 1990) (Revision received February 4th, 1991)

It is postulated that the role of oxygen toxicity in the evolution of life strongly depends on the origin of molecular oxygen, due to the strong redox buffering capacity of Precambrian waters containing large amounts of ferrous and manganese cations. The critical selective pressure could be observed only after aerobic photosynthesis had been developed, due to the high local concentration of oxygen in close vicinity of photosynthesizing cells. It is also postulated that early oxygen-evolving organisms excreted a substantial part of this element in the form of hydrogen peroxide. As a consequence of the high reactivity of this compound with ferrous and manganese cations, an important percentage of iron deposits were produced with HzO,as a major oxidant after the development of aerobic photosynthesis. It is postulated that negatively charged extracellular polymers of sim- ple pro- and eukaryotic organisms function as sacrificial targets of hydroxyl radicals and at the same time as extracellular equivalents of superoxide dismutases, in these two ways protecting cellular membranes against oxidative damage. The role of oxygen toxicity in developing aerobic mechanisms of iron uptake is also discussed.

Keywords: Microbial evolution; Oxygen toxicity; Precambrian; (manganese); (iron); (copper).

Introduction

Although the influence of the toxic properties of oxygen on the evolution of life has been wide- ly recognized, it has recently been postulated (Towe, 1985) that the toxicity of oxygen to early forms of life would appear to be of little significance at an oxygen level close to 0.1% and may be ignored in dealing with early prokaryotic evolution.

It seems worthwhile to re-evaluate the opinions about the role of the toxic properties of oxygen in evolution, on the basis of recent developments in the field of oxygen toxicity.

Although the role of superoxide radicals as an important intermediate of oxygen toxicity and, in consequence, the protective rob of superoxide dismutases were ostulated two ago, the superoxide theory of oxygen toxicity was the center of a heated debate until 1985 (Baum, 1984). The theory obtained its first direct experimental support in vivo only very

recently (Bilinski et al., 1985; Carlioz and Touati, 1986; Van Loon et al., 1986; Phillips et al., 1989). However, the theory of oxygen toxicity in its recent version differs from the original one. During that period the crucial role of transition metal cations in the toxicity of oxygen has been well documented (Samuni et al., 1983). Ac- cording to recent opinions, oxidative damage ascribed to molecular oxygen O2 (Fridovich, 1986) (dioxygen) can be caused by the superoxide radical itself 0, or its protonated form perhydroxyl radical HO;1 (Bielski et al., 1933) and by the hydroxyl radical l OH generated in the so- called metal mediated Haber-Weiss reaction (Halliwell and Gutteridge, 1986):

Fe,Cu (1) 0, + H202 - > l OH + OH- + O2

which consists of two subsequent reactions

(2) Fe3+ + 0; - > Fe2+ + 02

(3) Fe2+ + Hz02 - >Fe3+ + * OH + OH-

0303-2647/91/$03.50 0 1991 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland

Superoxide and perhydroxyl radicals are much less reactive (Sawyer and Valentine, 1981) than the hydroxyl radical and damage to DNA, for example, can be done only by the ‘OH or its sec- ondary radicals (Slater, 1984).

It is worth noting that the superoxide radical can be replaced in reaction (2) by ascorbate (Fee, 1982).

It is important to stress that Precambrian waters may have been rich in ferrous salts (Cloud, 1976) and as a consequence the rate of the reaction (8) depended only on the presence of hydrogen peroxide. The role of copper salts, also involved in the Haber-Weiss reaction can probably be ignored during this period because very likely (Osterberg, 1974) this metal was not available for living organisms due to its in- solubility in the reduced form in which it existed in Precambrian environments.

Oxygen toxicity in the early precambrian

The opinion of Towe (1985), suggesting that the toxicity of atmospheric oxygen may be ignored in dealing with early prokaryotic evolution, is based on the sensitivity to oxygen of present day prokaryotes including strict anaerobes, which can tolerate the presence of oxygen and even grow at 5 x 10e3 PAL (the present-day atmospheric level).

However, it seems a risky procedure to draw conclusions concerning the sensitivity of early forms of life to oxygen from the behavior of present day anaerobic organisms. Some of them can be descendants of aerobic bacteria (Gottschalk, 1981) and can still possess some protective mechanisms inhibiting the toxic effects of oxygen.

Even the others cannot be considered as direct descendants of early prokaryotes because vegetative forms or spores of strict anaerobes had to be transferred to new habitats through aerobic environments, which inevitably eliminated forms not adapted to oxygen. Such tolerance to low concentrations of oxygen does not require a formation of specialized enzymes like superoxide dismutases or enzymes destroying hydrogen peroxide. It could be achieved by

the formation of barriers limiting the access of gases to cell interior, like heterocysts of cyanobacteria (Postgate, 1982), and/or by changes in the metabolism of the cells. Such a process was followed (Bilinski and Litwinska, 1987; Bilinski et al., 1988) in yeast mutants deficient in the activity of superoxide dismutases. The mutation leading to a deficiency in copper and zinc containing superoxide dismutase (CuZnSOD) can phenotypically be suppressed by at least five mutations which permit the survival of the cells under pure oxygen for 24 h in the absence of this enzyme, whereas initial mutants deficient in CuZnSOD die within this period of time. The presence of two such mutations within the cell even permits growth in oxygen atmosphere. A similar finding has also been reported in&J. coli (Fee et al., 1988). It is obvious that the acquirement of tolerance to oxygen in the mutants of these facultative anaerobes can be much easier than in the early organisms, which did not develop specific antioxidant systems still present in these mutant cells deficient only in the activity of superoxide dismutases. The mutations leading to deficiency of superoxide dismutase in yeast can be partially phenotypically suppressed by an addition of manganese or copper cations to the medium (Bilinski and Liczmanski, 1988; Chang and Kosman, 1989). It is known (Halliwell, 1974) that various complexes of tbese metal cations possess superoxide-dismutase-like activity. Similarly microaerophilic Luctobacillus plan- tamm is able to accumulate manganese cations instead of forming superoxide dismutases (Ar- chibald and Fridovich, 19811). The salts of this metal may have been also present in large amounts in Precambrian waters.

Thus, in the course of evolution some genetic modifications of metabolism could have been accumulated and now they permit the survival of the present day strict anaerobes at 1 PAE of oxygen, although these organisms are apparently deficient in typical enzymes preventing the toxic effects of oxygen,

Recent studies on mutagenic effects of oxygen on prokaryotes show that the frequency of the mutations rises by two orders of magnitude in

307

cells deficient in superoxide dismutases exposed to oxygen (Farr et al., 1986). Thus, the inter- pretation of the fact that strict anaerobes can grow in the presence of small amounts of oxygen in the atmosphere does not take into account the increase in the frequency of mutations which for the most part have negative consequences for their bearers.

However, Towe’s general conclusion can be partly accepted due to the fact that Precambrian waters possessed an extremely high redox buffering capacity (Cloud, 19’76) which partly prevented the toxic effects of oxygen of atmospheric origin. Precambrian waters contain- ed large amounts of ferrous and manganese cations which could be oxidized by molecular oxygen to insoluble oxidized forms. The formation of small amounts of molecular oxygen in the atmosphere could not influence early Precambrian organisms living in large aqueous environments, because due to this fact dioxygen molecules could not deeply penetrate the ocean waters of this period.

However the amount of reducing metal salts in some freshwater or shallow water habitats could be low. Therefore, the organisms living in such environments could be exposed to oxidative damage even before the development of aerobic photosynthesis.

The development of oxygen releasing photosynthesis

In contrast to the dioxygen of atmospheric origin, the dioxygen released by aerobic photosynthesizers could create a strong selective pressure toward the formation of numerous mechanisms preventing the toxic effects of oxygen. This was caused by the fact that the formation of oxygen within cyanobacterial cells caused a substantial increase in the concentration of oxygen in close vicinity of oxygen evolving cells. In this case the diffusion of reducing iron and manganese salts to the cell environment could not be high enough to neutralize the negative consequences of generated 02 for unadapted members of microbial communities, especially during intensive insolation. In this

way, waters inhabited by cyanobacteria could be completely depleted of ferrous and manganese salts during the day.

One can assume that oxygen was first dangerous for the cells which generated it. It seems important to consider the situation which occurred at the time when aerobic photosynthesis became a reality.

HOW did the first photosynthesizing organisms prevent the toxic effects of their dangerous metabolite? Cyanobacteria did not live alone. How did the other members of Precambrian communities respond to the chang- ing redox potential of the environment?

Molecular oxygen by itself is rather poorly reactive and could be responsible for the inactivation of only some metabolic systems like the nitrogenase complex (Halliwell, 1981). Most dangerous are its excited (singlet oxygen) and semi-reduced forms (superoxide radical and hydroxyl radical). It is surprising that hydrogen peroxide is rather unreactive in the absence of reduced iron and copper salts, but it becomes dangerous if they are present (Halliwell and Gutteridge, 1986).

Numerous strategies could be applied to prevent the toxic effects of oxygen. The strategy of avoidance cannot be used successfully even by strict anaerobes which are always in danger of contact with this molecule. However, there exist some other types of avoidance strategy which could assure at least a partial protection against oxygen. One of the solutions could be the al- ready mentioned structural changes hindering the access of oxygen to the cell interior. Some of these extracellular structures could have played a previously protective role against drying off. However, even cells exposed to dioxygen can strongly diminish the negative consequences of this fact by eliminating the most important sources of danger. For example, the lack of light absorbing molecules can prevent the formation of singlet oxygen. Another way of preventing oxidative damage could be an elimination or strong repression of dominant sources of superoxide radicals within the cells (Fridovich, 1984).

These strategies of avoidance can be applied

only by the organisms which cannot be considered true aerobes. The aerobic style of life is connected with the formation of superoxide radicals (Fridovich, 1984) and, therefore, it re- quires an active formation of various protective mechanisms. These mechanisms are not restricted to superoxide dismutases (Towe, 1988) and catalases usually discussed in evolutionary papers.

Taking into account the fact that oxidative damage results from the formation of singlet oxygen, superoxide radicals and hydroxyl radicals, the production of these factors has to be prevented.

Metabolically active aerobes cannot prevent the formation of superoxide radicals. Therefore, the only strategy of preventing the toxic effects of this molecule is to develop appropriate enzymes (McCord et al., 1971) or their mimics (Ar- chibald and Fridovich, 1981) which can destroy it. In fact the omnipresence of superoxide dismutases among aerobes usually occurring in one cell in at least two molecular forms is proof of it. These enzymes are indispensable for aerobes (McCord, 1971).

On the contrary, the formation of hydroxyl radicals can be prevented by three strategies. According to reactions (1) and (3), the hydroxyl radical is directly formed in the reaction between hydrogen peroxide and reduced iron and/or copper salts (Halliwell and Gutteridge, 1986). Therefore, the strict control of both reac- tants can prevent ‘OH formation. In addition, the control of the superoxide level can partly prevent the formation of the hydroxyl radical by lowering the concentration of 0, as one of the reducing agents potentially involved in the reduc- tion of these metal cations (reaction 2). The formation of hydroxyl radicals can also be prevented by the oxidation of reduced iron salts to the ferric form (Gutteridge, 1985).

The list of the enzymes controlling the level of hydrogen peroxide is long. Catalases, often discussed in early evolutionary papers, con- stitute only one example of such enzymes. These types of enzymes are also common among aerobes (Chance et al., 1974).

It is also well known that the cells can control the availability of transition metal cations for

reactions (l)-(3). Most of the organisms have evolved highly sophisticated systems of se- questering these metal cations (Fee, 1982) and there are some indications (Bilinski et al., 1985) that this strategy of preventing the formation of hydroxyl radicals is the most successful. There also exist low molecular substances preventing oxidative damage like urate or ascorbate (Halliwell, 1981). However, when this damage is done, there have evolved numerous mechanisms able to repair biomolecules like DNA repair systems of hydroperoxide reductases (Storz et al., 1987).

Hence, the appearance of molecular oxygen had to result in the formation of at least some of the mechanisms listed above. However, these mechanisms can protect only the cell interior, with the exception of the organisms excreting superoxide dismutases and other antioxidant proteins.

The problem of extracellular damage

Oxidative damage is not restricted to the cell interior. How can the cells prevent oxidative damage to the cell membranes, caused by extracellular factors?

As has been mentioned before the presence of oxygen or of its reduced forms in the environments rich in ferrous cations leads to the formation of the extremely reactive hydroxyl radical. The Precambrian oceans were rich in ferrous cations (Cloud, 1976) and the excretion of oxygen by early photosynthesizers into such environment could have led to severe damage to the cell membranes, assuming they survived the internal damage.

Let us consider the possible reactions which could take place when dioxygen was excreted to the environment containing ferrous cations.

(4) 2Fe2+ + 202 ->2Fe3+ + 20,

(5) 0, + H+ -> HO;1

(6) 0, + HO*2 + H+- > Oz + H202

(7) Fe2+ + H202 ->Fe3+ + *OH + OH-

(8) Fe2+ + ‘OH -> Fe3+ + OH-

It is interesting that there are good reasons to believe that early Precambrian cyanobacteria, similarly to those living now, excreted a substantial fraction of dioxygen in the form of hydrogen peroxide (Paterson and Myers, 1973). Similarly, heterotrophic members of early aerobic Precambrian communities could also ex- crete large amounts of hydrogen peroxide into the environment as do present day organisms (Halliwell, 1981). Hydrogen peroxide excreted into environment immediately reacts with ferrous cations thus giving rise to the hydroxyl radical (reactions (7) and (3), called the Fenton reaction (Walling, 1982)). It is worth noting that the reaction between hydrogen peroxide and ferrous cations is very fast and independent of pH, in contrast to the reaction of molecular oxygen.

In consequence, it seems possible that at least the part of Banded Iron Formations deposited at the time when aerobic photosynthesis became a reality, could be deposited with the use of hydrogen peroxide as a major oxidant.

Negatively charged extracellular polymers: sacrificial targets of hydroxyl radicals

How did these hydrogen peroxide releasing organisms prevent their cellular membranes against oxidative damage?

It was postulated (Bilinski, 1988) that negatively charged polymers of microbial cellular envelopes play very important roles in preventing oxidative damage. It is known that all free living aerobic microorganism are covered with thick extracellular structures of diverse chemical structures. However, irrespec- tive of the type of these structures they possess negatively charged groups. The mucilage of various bacteria as well as capsules of cyanobacteria are good examples of these structures. The negatively charged groups of these polymers show a very strong cation exchange capacity, and this ability is exploited in waste disposers to remove heavy metal contamina- tions from the industrial waste tiaters. These polymers should also have bound large amounts of ferrous cations, which dominated in Precam- brian waters.

One can imagine that molecular oxygen excreted by cyanobacteria, as well as hydrogen peroxide released by both groups of aerobic prokaryotes, reacted with ferrous cations predominantly in the outermost parts of cellular envelopes and formed ferric deposits. It is important to stress that the hydroxyl radical is extremely unstable and reacts with practically any molecule in close vicinity of the place of its formation (Halliwell and Gutteridge, 1986). In this way, the spatial separation of the place where hydrogen peroxide reacts with ferrous cations from cellular membranes could ensure a satisfactory protection of the latter. In this case some elements of the external layers of cellular envelopes instead of the cell membranes are sub- ject to oxidative damage. It is possible that the well known stripping of these structures could result from their depolymerization by hydroxyl radicals.

Extracellular equivalents of srrgeroxide dismutases

Both manganese and iron (and later on copper) salts accumulated within cellular envelopes can also play the role of extracellular mimics of superoxide dismutases. Both metals are important parts of the active centers of these enzymes (McCord et al., 1971). Manganese cations are known (Archibald and Fridovich, 1981) to mimic the superoxide dismutase activity when accumulated within the cell. Therefore’their accu- mulation within cell envelopes could also be protective against extracellularly generated superoxide radicals, for example in the case of phagocytosis. Activated phagocytic white blood cells kill invading cells with the use of oxy radicals generated during the so-called respira- tory burst. Under these circumstances enor- mous amounts of superoxide radicals are formed at the expense of NADPH (Badwey and Karnov- sky, 1980).

True extracellular Cu-Zn superoxide dismutases were developed much later in the course of evolution in higher vertebrates, where other methods of protecting cellular membranes against oxidative damage through the postulated mechanisms were impossible. How-

310

ever, also in these organisms gastrointestinal and tracheobronchial mucus is considered a sacrificial target for hydroxyl radicals (Cross et al., 1984).

Paradoxically, these envelopes, rich in organically bound iron and copper salts could also, at least partly, protect microbial cells against phage infection. The presence of these thick extracellular structures slows down the penetration of phage particles to their receptors making them longer exposed to oxidative damage in which transition metal cations play the crucial role. In fact, the site-specific metal- mediated Haber-Weiss reaction version of the theory of oxygen toxicity is based on studies of phage inactivation in the presence of transition metals (Samuni et al., 1983).

These suggestions are in agreement with the observation that there is a strict correlation between the thickness of the mucilage and pathogenicity of some bacterial species. This could result from the dual role of the mucilage, preventing oxidative damage of cellular membranes exerted by 0-z and *OH during phagocytosis on the one hand, and by destroying phase particles penetrating cellular envelopes by ‘OH, on the other.

The oxidation of ferrous cations by neutrophilic organisms can be considered a part of mechanisms preventing the toxic effects of oxygen

The observations of present day neutrophilic, iron oxidizing organisms have revealed that these organisms are unable to extract usable en- ergy from this process and that their mucilage or cellular capsules are often heavily encrusted with ferric deposits. It has been found that organisms could be actively involved in iron oxidation, not only by excreting hydrogen peroxide but also by accelerating iron oxidation by molecular oxygen through creating favourable conditions for this process (Blaszczynski et al., unpublished).

The existence of various mechanisms of iron oxidation can be considered a consequence of the fact that a close contact of cellular mem-

branes with ferrous cations is dangerous in aerobic environments.

Thus, the evolution of high and low affinity systems of trivalent iron assimilation cannot be considered solely (Lewin, 1984) a consequence of the unavailability of divalent iron salts in the environment. This could be the safest mechanism of this metal uptake in aerobic environments (Bilinski, 1988).

The formation of highly oxidizing atmosphere, and the appearance of eukaryotic cells

Much later, when the environment became more oxidizing due to the photosynthetic activity of organisms, there appeared a new problem connected with the release of copper from insoluble minerals. This metal can also be involved in the Haber-Weiss reaction giving rise to the hydroxyl radical. On the other hand it took a part of numerous enzymes involved in oxygen- related enzymes like cytochrome c oxidases, various oxygenases and important elements of aerobic photosynthetic systems. At the same time, copper cations were also introduced into modern superoxide dismutases (CuZnSOD) present mainly in eukaryotic organisms and not found in primitive prokaryotes (Towe, 1988).

The appearance of eukaryotic cells created new mechanisms preventing the toxic effects of oxygen. In eukaryotic cells some processes generating semi-reduced forms of oxygen are localized in separate organelles, like perox- isomes and mitochondria, together with appropriate scavenging enzymes. On the other hand, a complicated transport system of inner mitochondrial membrane made possible a better control of the availability of transition metal cations for the Haber-Weiss reaction.

The development of chromatin resulted in nu- clear DNA being better shielded because it can no longer bind large amounts of transition metal cations involved in the ‘OH formation. In consequence, the mutagenic effects of oxidative stress’ became less dangerous for eukaryotic organisms. The increased frequency of r&a- tions observed under aerobic conditions in pro-

311

karyotes deficient in superoxide dismutases (F’arr et al., 1986) was not observed in yeast mutants deficient in cytosolic SOD (Bilinski et al., 1985).

What is interesting, is the fact that the com- plete deficiency of the mitochondrial enzyme (MnSOD) does not lead to mutagenic effects in an oxygenic environment (Van Loon et al., 1986). Experimental data concerning phenotypic suppression of superoxide dismutase deficiencies (Bilinski and Liczmanski, 1988) by copper and manganese cations in yeast See- charomyces cerewisiae have shown that these metals can mimic superoxide dismutase deficiency only in the cytosol, whereas the deficiency in mitochondrial MnSOD cannot be suppressed by the addition of these cations in the medium (Liczmanski and Bilinski, unpublished), This suggests that the mitochondrial interi- or does not contain sufficient amounts of these cations (or their chelators) to replace the missing SOD activity. Thus, mitochondrial DNA, nlthough similar to prokaryotic DNA, is probably less exposed to oxidative stress, because the availability of transition metal cations in mitochondrial matrix is strongly limited due to the highly selective transport system of internal mitochondrial membrane.

Acknowledgments

Experimental research germane to the ideas presented in the paper was supported by CPBR witbin the Project 3.13. The author is indebted to Professor W.J. Kunicki-Goldfinger for in- valuable comments and suggestions.

References

Archibald, F.S. and Fridovich. I., 1981, Manganese and defenses against oxygen toxicity in Laetobacillus plun- town. J. Bacterial. 145,442-451.

Badwey, J.A. and Karnovsky, M.L., 1980, Active oxygen species and functions of phagocytic leukocytes. Annu. Rev. Biochem. 49,695-726.

Baum, R.M., 1984, Superoxide theory of oxygen toxicity is center of heated debate, Chemical and Engineering News April 9, 20-26.

Bislski, B.H. J., Arudi, R.L. and Sutherland, M.W., 1983, The reactivity of HO’z/O-zwith unsaturated fatty acids. J. Biol. Chem. 258, 4759-4763.

Bilinski, T., 1988, The impacts of toxic properties of oxygen on the evolution of life. III. Selected oxygen-dependent functions of negatively charged polymers in microbial cell envelopes. Bull. Pol. Acad Sci. Biol. Ser. 66, 153-157

Bilinski, T., 1988, The impacts of toxic properties of oxygen on the evolution of life. IV. Assimilation of trivalent iron: the safest mechanism of the metal uptake by microorganisms. Bull. Pol. Acad. Sci. Biol. Ser. 36,

159-162. Bilinski, T. and Liczmanski, A., 1988, Simdation of SOD ac-

tivity by manganese and copper salts in vivo. J. Cell. Biol. Supplement 12A, 43.

Bilinski, T. and Litwinska, J., 1987, On the ideas alternative to the theory of superoxide-mediated oxygen toxicity. Bull. Pol. Acad. Sci. Biol. Sci. 35, 25-31.

Bilinski, T., Krawiec, Z., Liczmanski, A. and Litwinska, J., 1985, Is hydroxyl radical generated by the Fenton reaction in vivo? Biochem. Biophys. Res. Commun. 130, 533-539.

Bilinski, T., Krawiec, Z., Litwinska, J. and Blaszczynski, M., 1988, Mechanisms of oxygen toxicity as revealed by studies of yeast mutants with changed response to oxidative stress, in: Oxy-Radicals in Molecular Biology, P.A. Cerutti, I. Fridovich and J.M. McCord @is.) (Alan R. Liss, New York) pp. 109-123.

Carlioz, A. and Touati, D., 1986, Isolation of superoxide dismutase mutants in E. coli. Is superoxide dismutase necessary for aerobic life? EMBO J. 5, 623-630.

Chance, B., Sies, H. and Boveris, A., 1974, Hydroperoxide metabolism in mammalian organs. Physiol. Rev. 59, 527-605.

Chang, E.C. and Kosman, D.J., 1989, Intracellular Mn(II)- associated superoxide scavenging activity protects Cu,Zn superoxide dismutase-deficient Saccharomyces cerevisiae against dioxygen stress. J. Biol. Chem. 264, 12172-12178.

Cloud, P.E., 1976, Beginnings of biospheric evolution and their biogeochemical consequences. Palaeobiology 2, 351-377.

Cross, C.E., Halliwell, B. and Allen, A., 1984, Antioxidant protection. A function of tracheobronchial and gastrointestinal mucus. Lancet i, 1328-1336.

Farr, S.B., D’Ari, R. and Touati, D., 1986, Oxygen- dependent mutagenesis in E. coli lacking superoxide dismutase. Proc. Natl. Acad. Sci. USA 83, 8268-8272.

Fee, J.A., 1982, Is superoxide important in oxygen poison- ing? Trends Biochem. Sci. 7, 34-35.

Fee, J.A.. Niederhoffer, E.C. and Naranjo, C., 1988, First description of a variant of E ccli lacking superoxide dismutase activity yet able to grow efficiently on minimal oxygenated medium, in: Trace Metals in Man and Animals, L. Hurley, C. Keen, B. Lonnerdal and R. Rucker (eds.) (Plenum Press, New York).

Fridovkh, I., 1984, Overview: Biological sources of 0, in:

312

Methods in EnsymoIogy, S.P. Colowick and N.O. Kaplan (eds.) (Academic Press, Orlando, FL) Vol. 105, pp. 59-61.

Fridovich, I., 1986, Biological effects of superoxide radical. Arch. Biochem. Biophys. 247, l-11.

Got&ha&, G., 1981, The anaerobic way of lie of pro- karyote, in: Prokaryotes, M.P. Starr, II. Stolp, H.G. Trueper, A. Balows and H.G. Schlegel H.G. (eds.) (Springer-VerIag, New York) Vol. 2, pp. 1415-1424.

Gutteridge, J.M.C., 1985, Superoxide dismutase inhibits the super&de-driven Fenton reaction at two different levels. FEBS Lett. 185,19-23.

Hahiwell, B., 1974, The superoxide dismutazi: activity of iron complexes. FEBS Lett. 56,34-38.

Halliwell, B., 1981, Free radicals, oxygen toxicity and aging, in: Age Pigments, E.S. Sohai (ea.) (Elsevier/North- Holland, Amsterdam) pp. l-62.

HaMwell, B. and Gut&ridge, J.M.C., 1986, Oxygen free radicals and iron in reIation to biology and medicine: some problems and concepts. Arch. Biochem. Biopbys. 246, 501-514.

Lewin, R., 1954, How microorganisms transport iron. Science 225, 401-402.

McCord, J.M., Keele, B.B. and Fridovich, I., 1971, An enzyme based theory of obligate anaerobiosis, the physiological functions of superoxide dismutase. Proc. Natl. Acad. Sci. USA 68, 1024-1027.

Gsterberg, R., 1974, Origins of metal ions in biology. Nature 249,382-383.

Paterson, C.O.P. and Myers, J., 1973, Photosynthetic pro. ducticm of hydrogen peroxide by An&y&is n~dulans, Pl. Physio). Lancaster 51.169-114.

Phillips, J.P., Campbell, S.D., Michaud, D., Charbonneau, M. and Hihiker. A.J., 1989, NulI mutation of copper/zinc superoxide dismutase in DmsophiZa confers hypemen-

sitivity to paraquat and reduced longevity. Proc. Natl. Acad. Sci. USA 86,2761-2765.

Postgate, J.R., 1982, Biologic31 nitrogen fixation: fun- damentals. Philos. Trans. R. Sot. London 296,375-385.

Samuni, A., Aronovitch, J., Godinger, D., Chevion, M. and Czapski G., 1983, On the cytotoxicity of vitamin C and metal ions. A site-specific Fenton mechanism. Eur. J. Biochem. 137, 119-124.

Sawyer, D.T. and Valentine, J.S., 1981, How super is superoxide? Act. Chem. Res. 14,393-400.

Slater, T.F., 1984, Free-radical mechanisms in tissue injury. Biochem. J. 222, l-15.

Stars, G., Christman, M.F., Sies, H. and Ames, B., 1987, Spontaneous mutagenesis and oxidative damage to DNA in Salmolasll& tpphimurium. Proc. Nat]. Acad. Sci. USA 84,8917-8921.

Towe, K.M., 1985, A comparison of procaryotic oxygen and UV tolerances: significance to early life. European Joint Committee for Scientific Cooperation Research Group on Chemical Evolution, Early Biological Evolution and Ex- obiology. Newsletter 11, 12.

Towe, K.M., 1988, Early biochemical innovations, oxygen and earth history, in: Molecular Evolution and the Fossil Record, T.W. Broadhead (ed.) (Paleontological Society USA Short Courses in Paleontology) Vol. 1, pp. 114-133.

Walling, C., 1982, The nature of the primary oxidants mediated by metal ions, in: Oxidases and Related Redox Systems, T.E. King, H.S. Mason and M. Morrison (eds.) (Pergamon Press, Oxford) pp. 85-97.

Van Loon, A.P.G.M., Persold-Hurt, B. and &hats, G.A., 1986, A yeast mutant lacking mitochondrial manganese superoxide dismutase. Proc. Natl. Acad. Sci. USA 83, 3820-3824.

oxygen toxicity and microbial evolution

Documents