Oxygen

by Nick Lane

  198 • THE ANTIOXIDANT MACHINE

  surprisingly, mucus has some big advantages over calcite shells. James Lovelock was confounded by mucus while trying to sterilize hospital wards with high-intensity ultraviolet rays in the early 1950s. He succeeded in killing bacteria that had been denuded of mucus, but had no impact at all on their intact cousins, even at hundreds of times the normal atmospheric levels of ultraviolet. We saw in Chapter 6 that the damage wrought by radiation on living cells is mediated by the formation of free radicals from water. By protecting against free-radical damage, mucus goes some way towards explaining the remarkable ability of bacteria to survive in outer space and in other highly irradiated environments. Survival in outer space! This is a startling property for a substance we tend to associate with a blocked nose. The horror-movie concept of repulsive slimy aliens may be more soundly based in biology than their creators imagined. Perhaps we should not be too surprised to learn that the way in which mucus deals with free radicals is far more clever than just a high slime index.

  Bacterial mucus is a mixture of long-chain polymers, analogous to plastics, which have one thing in common: they all carry a negative electrical charge. This means that mucus binds tightly to positively charged atoms such as iron and manganese, removing them from the surroundings. The affinity is so strong that some bacteria are exploited industrially to trap heavy-metal contaminants in waste water.

  What advantage does a full metal jacket offer to bacteria? The answer is somewhat counterintuitive. We saw in Chapter 6 that hydrogen peroxide and superoxide radicals are not very reactive, and so are likely to diffuse some distance before reacting. They are only dangerous in the presence of metals such as iron, which can catalyse the formation of ferociously reactive hydroxyl radicals. Given that metals catalyse dangerous free-radical reactions, a metal jacket might seem to be a liability — it must be fizzing with menace all the time. Yet by stockpiling iron in their jackets, bacteria make sure that any dangerous free-radical reactions take place at arms length, rather than inside the cell itself. The mucus acts as a sacrificial target and the iron is converted into biologically inactive rust. The effect is equivalent to disposing of a bomb by exploding it at a safe distance. Not only this, but the controlled explosions detonate invaders such as bac-teriophages (viruses that infect bacteria) and possibly even immune cells that try to engulf them. There is, certainly, a direct correlation between the thickness of the jacket and both the survival and the infectiousness of many bacteria.

  A Hundred and One Ways of Living with Oxygen • 199

  As the bacteria become encrusted with metals, their jackets grow ever more cumbersome. In the end, they may be obliterated by their accumulated wealth. The presence of hollow microscopic pits in the structure of some banded iron formations suggests that these rocks might have been formed from the buried corpses of innumerable iron-encrusted bacteria.

  As the bacteria themselves dissolved away, only the metal jackets remained to testify to the existence of a mass graveyard.

  We should not think of these rudimentary tricks as fit only for microbes. Each of them has an equivalent in ourselves. We too hide behind a layer of dead cells, otherwise known as the skin. Like the ciliates, we too use haem proteins as sensors to maintain internal oxygen at a constant low level. Like the sulphate-reducing bacteria, we too use sulphur as a buffer against oxygen (a trick we shall look into later in this chapter).

  We too produce mucus to protect against oxygen or bacterial infection in our nasal passages, airways and lungs. In the same way that anaerobic bacteria shelter within the intestine to avoid oxygen, so our own cells might be said to shelter within the body itself, where the oxygen level is kept at a fraction of that in the hostile outside world.

  In this regard, we can reasonably view the gigantism discussed in Chapter 5 as an antioxidant response. The increase in body size compensates for the higher external oxygen levels, especially in diffusion-limited creatures such as the Bolsover dragonfly. An increase in size presumably lowers the oxygen levels within the end-users, the mitochondria. As we saw in Chapter 8, mitochondria operate best at oxygen levels not far above those tolerated by sulphate-reducing bacteria. If external oxygen levels rise, an increase in size will counteract the internal rise, and maintain oxygen at the right level.

  The mitochondria themselves contribute to this balance. Mitochondria were once free-living bacteria that found shelter and protection within larger cells. The bargain was not one-sided: the internalized bacteria gained protection, but also lowered oxygen levels for the host cell by actively respiring. The relationship is now far more complex, but we should not forget that mitochondria do still lower the oxygen level inside cells. If mitochondria are defective, but the blood stream continues to deliver oxygen at the same rate, then our cells would presumably become oxidized. As we grow older, our mitochondria do begin to fail and our cells do become oxidized. Such oxidation is often attributed to free radicals escaping from defective mitochondria, but it might also be the result

  200 • THE ANTIOXIDANT MACHINE

  of higher oxygen levels in the rest of the cell, due to lower mitochondrial uptake of oxygen.

  Oxygen-dependent organisms cannot rely on shielding as the only solution to oxygen toxicity. In all these organisms, oxygen must be consumed continuously to stay alive, because it is an essential ingredient in their only, or main, means of producing energy. To hide is not just impossible, but also metabolically detrimental. A different balance must be found, in which the dangerous effects of free radicals are prevented or contained by antioxidant enzymes and free-radical scavengers — the second and third defence mechanisms on our list. I will deal with the enzymes first.

  Two of the most important antioxidant enzymes are superoxide dismutase and catalase. Virtually all organisms that spend any time in the presence of oxygen possess copies of these two enzymes. Their near ubiquity in aerobic organisms sharpens the paradox of cyanobacteria. These bacteria were the first photosynthetic organisms to split water and produce oxygen. If they had evolved in a world without oxygen, they should have been poisoned by their own waste. The fallacy of this textbook argument is discussed in Chapter 7, and we saw that the cyanobacteria were, in all likelihood, already protected against their poisonous product by superoxide dismutase and catalase. These enzymes, and probably others, evolved in response to the formation of reactive oxygen intermediates by ultraviolet radiation very early in the history of life. Indeed, we saw that there is good evidence that antioxidant enzymes were present in LUCA.

  Superoxide dismutase, with the memorable acronym SOD, holds a special place in the affections of free-radical biochemists. It is, symbolic-ally, the proof of the pudding. Ever since the early 1950s, when researchers first started to speculate about the role of oxygen free radicals in ageing and disease, proof of their importance has been hard to come by.

  Free radicals are fleeting. For years, their existence was supported only by tell-tale traces of the damage caused — evidence as flawed and controversial as deducing the existence of a yeti from outsized footprints in the snow. Then, in 1968, Joe McCord and Irwin Fridovich, at Duke University in North Carolina, showed that an abundant blue-green protein called haemocuprein, long thought of as an inert depository for copper, did actually have catalytic activity. It converted superoxide radicals (O •–

  2

  )

  into hydrogen peroxide (H2O2) and oxygen. Despite an intensive search,

  A Hundred and One Ways of Living with Oxygen • 201

  no other target for this enzyme could be found. Moreover, the rate at which it disposed of superoxide almost beggared belief. Superoxide radicals are unstable and react with each other within seconds to produce hydrogen peroxide, but haemocuprein sped up this natural reaction by a factor of a billion. Surely this could not be accidental!1 McCord and Fridovich renamed the protein superoxide dismutase (SOD) in their seminal 1969 paper in the Journal of Biological Chemistry — in the opinion of many, the most imp
ortant discovery in modern biology never to win the Nobel Prize.

  The implications transformed the field. If an enzyme as efficient as SOD evolved specifically to eliminate superoxide radicals, then superoxide radicals must be biologically important. Seen the other way round, free radicals are a normal feature of biology and life has evolved a remarkably effective mechanism for dealing with them. Given the tendency of mutations to corrupt the superfluous, this mechanism can only be that good because it needs to be that good. So what would happen if SOD

  failed in some way, swamping the cell with free radicals? Ageing? Death?

  The myriad possibilities had a galvanizing effect that has persisted to this day.

  It was not long before other forms of SOD were discovered. The first, reported in 1970 by McCord and Fridovich again, was isolated from the bacterium Escherichia coli. This was a startlingly different kettle of fish: a pink, manganese-containing enzyme, but with a similar capacity for eliminating superoxide radicals. Even more intriguingly, it turned out that many eukaryotic cells actually possess both forms of SOD. As I write, 30 years on, it has become a commonplace that many eukaryotes produce several different types of SOD — usually one in the mitochondria, another in the cytosol, and a third that is secreted from the cell. Although their detailed structures differ, all contain metal atoms at their catalytic centres: copper and zinc, manganese, and iron or nickel.

  1 Fridovich, now venerated as the godfather of the free-radical field, points out that the speed of this reaction is close to the maximum rate of diffusion and on the face of it seems impossibly fast. The copper site at the centre of the enzyme takes up less than 1 per cent of the surface of the enzyme. On the basis of random diffusion, one would expect more than 99 per cent of collisions between enzyme and substrate to be fruitless. In fact, superoxide radicals are guided into the active site down an electrostatic gradient, like an aeroplane being guided into an airport by runway lights. Thus the reason that the enzyme can match the speed of diffusion is because it streamlines the chaotic natural process. Substituting amino acids in the enzyme alters the electrostatic fields and affects the speed of reaction, without necessarily having any impact on the active site itself. It is actually possible to produce even faster versions of superoxide dismutase.

  202 • THE ANTIOXIDANT MACHINE

  The importance of these enzymes is illustrated by the fate of so-called

  ‘knock-out’ mice, which lack part of a gene for one of the SOD enzymes but are in other respects genetically ‘normal’. In a 1996 study, Russell Lebovitz and his colleagues at the Baylor College of Medicine in Houston reported that mice born with a defective mitochondrial form of SOD died within three weeks of birth. The mice were pathetic runts, severely anaemic and with a form of motor neuron disease that led to weakness, rapid fatigue, and what Lebovitz described as ‘circling behaviour’ — presumably a disoriented tendency to stumble sideways in search of their tails. At autopsy, they were also found to have cardiac abnormalities and fatty deposits in the liver. The mitochondria of mice that had survived beyond seven days were shot to pieces, especially in tissues with a high metabolic rate, such as heart muscle and brain. Another knock-out strain of mice, with a different mutation in the same enzyme, failed to survive beyond five days. In people, smaller defects in mitochondrial SOD have been linked with ovarian cancer and insulin-dependent diabetes, among other conditions. In contrast, loss of the cytosolic form of SOD is less devastating, although this, too, causes problems in later life, notably infertility, neurological damage and cancer.

  There could hardly be more convincing evidence of the importance of SOD. Yet the enzyme demonstrates once again the need for networking between antioxidants. Far from disposing of a toxic chemical, SOD merely defers the problem. The product of SOD is hydrogen peroxide, a potentially dangerous chemical in its own right. We may well wonder if it is necessarily beneficial to generate hydrogen peroxide at a billion times the natural rate. There are, in fact, circumstances in which too much SOD

  may be detrimental. For example, people with Down syndrome have an additional copy of chromosome 21. Why this should be so damaging is unknown, but researchers have noted that the gene for SOD sits on chromosome 21. People with Down syndrome therefore make too much SOD.

  The syndrome itself is characterized by oxidative stress, leading to neurological degeneration. It may be that people with Down syndrome suffer oxidative stress because they have too much SOD.

  In normal physiological conditions, however, hydrogen peroxide is removed equally quickly by catalase, which converts it into oxygen and water. We saw in Chapter 7 that there are other enzymes that can remove hydrogen peroxide and organic peroxides safely, without producing oxygen, using a variety of electron donors such as glutathione and vitamin C.

  A Hundred and One Ways of Living with Oxygen • 203

  Enzymes that dispose of peroxides are being reported all the time. In 1988, for example, a new family of antioxidant enzymes, now known as peroxiredoxins, were discovered by Sue Goo Rhee and her colleagues at the National Heart, Lung and Blood Institute in the United States. These enzymes do not have metals at their centre, but instead have two adjacent sulphur atoms that accept electrons from a small sulphur-rich protein called thioredoxin. By the mid 1990s, similar peroxiredoxins had been isolated from representatives of all three domains of life, suggesting that they, too, dated all the way back to LUCA. At least 200 related peroxiredoxin genes have now been found, five of which are found in people, and their DNA sequences determined,

  I mention the peroxiredoxins, in part, because they solve a long-standing puzzle relating to human parasites, including Plasmodium falciparum, the protozoan responsible for the most severe form of malaria, and helminth worms such as Fasciola hepatica. When invading the body, these parasites are subjected to a barrage of oxygen free radicals from neutrophils and other immune cells. The ferocity of this attack can trigger inflammation and fever serious enough to kill the host, let alone the parasite. Most parasites defend themselves using antioxidant enzymes such as SOD, but, curiously, few possess catalase to remove hydrogen peroxide. To researchers in the 1980s, this seemed a contradiction: the action of SOD

  in the absence of catalase should have flooded the parasites in hydrogen peroxide and so exacerbated the effects of the immune attack. It did not.

  The parasites held their own. Unless the theory was full of holes, the parasites must have been using an unknown enzyme that could detoxify hydrogen peroxide. The search for this ‘missing link’ was finally rewarded by the discovery of the peroxiredoxins, which have since been found in virtually all parasites lacking catalase. In fact, representatives of all three domains of life produce related peroxiredoxins, implying that they too may have been present in LUCA.

  Our understanding of the peroxiredoxins is beginning to open up new targets in the struggle against parasitic diseases. One possibility, for instance, is to vaccinate against the parts of the parasite enzymes that are appreciably different from the human versions, thus gearing up the immune system to attack one of the critical bastions of the parasite antioxidant defences.

  The combination of SOD with some sort of enzyme to remove hydrogen peroxide is almost mandatory if cells are to defuse free radicals before too

  204 • THE ANTIOXIDANT MACHINE

  much damage is done. Because there are so many enzymes that can dispose of hydrogen peroxide, catalase deficiency is less injurious than SOD

  deficiency. Moreover, as we have seen, hydrogen peroxide is only really toxic in the presence of iron or copper, either of which can catalyse the formation of hydroxyl radicals. Under normal circumstances, these metals are kept locked away in their respective protein cages, ferritin and caeruloplasmin. Tomasz Bilinski, a free-thinking evolutionary microbiologist at the University of Lublin in Poland, has argued that stowing away metals might be the single most successful strategy to prevent formation of hydroxyl radicals. Even so, despite the many precautions, some hydroxyl radicals are
undoubtedly still formed. We noted in Chapter 6 that the rate at which oxidized fragments of DNA are excreted in the urine suggests that each cell in the body sustains thousands of ‘hits’ to its DNA every day. Even allowing for experimental error, we must conclude that the enzyme networks are not fool-proof. This conclusion is supported by our need for dietary antioxidants such as vitamin C and vitamin E. They are technically referred to as ‘chain-breaking’ antioxidants, because they quench free-radical chain reactions already started by hydroxyl radicals.

  Chain-breaking antioxidants are the third defence mechanism on our list.

  Most chain-breaking antioxidants work in essentially the same way as vitamin C, by donating electrons. Many of the best-known antioxidants, including carotenoids, flavenoids, phenols and tannins, must be obtained in the diet from plants. Their relative importance in the body’s antioxidant balance is tricky to determine, but their overall contribution is generally held to be responsible for the benefits of fruit and vegetables.

  On the other hand, we should not think of chain-breaking antioxidants as merely dietary components. A few are products of our own metabolism, such as uric acid, bilirubin (a bile pigment and breakdown product of haem) and lipoic acid. These are powerful chain-breaking antioxidants in their own right, at least as potent as vitamins C and E. Some conditions that we tend to think of as pathological, such as jaundice in newborn babies, may really be an evolved physiological adaptation. In the case of jaundice, evidence suggests that bilirubin is stockpiled in the skin to protect against the oxidative stress of birth. The baby emerges from the cloistered security of the womb into the high-oxygen world outside, but has not yet built up protection from dietary antioxidants, hence the need for bilirubin. Similarly, the disfiguring colours of a bruise — courtesy, again,

  A Hundred and One Ways of Living with Oxygen • 205