Oxygen
Page 20
140 • GREEN PLANET
Catalase is responsible for getting rid of hydrogen peroxide. This is a potential killer for bacteria as we saw in Chapter 6. Virtually all aerobic organisms possess a form of this enzyme, and even some anaerobic bacteria, which try to avoid oxygen like the plague, retain some catalase just in case. It works extraordinarily quickly. Without catalase, and in the absence of iron, hydrogen peroxide takes several weeks to break down into water and oxygen. Dissolved iron, of course, catalyses the breakdown of hydrogen peroxide into hydroxyl radicals, and eventually water, via the Fenton reaction (see Chapter 6, page 118). If iron is incorporated into a pigment molecule such as haem (the pigment in haemoglobin) the rate of decomposition is increased 1000-fold. If the haem pigment is embedded in a protein, as is the case with catalase, then hydrogen peroxide is broken down directly and safely into oxygen and water, at a rate that is estimated to be 100 million times faster than the rate in the presence of iron alone.
There are several different types of catalase. Most animal cells have a form that has four haem molecules embedded in its core. In contrast, some microbes have a different sort of catalase, which contains manganese instead of haem at its core. Despite their different structures, both enzymes are equally fast, and are correctly called catalase, in the sense that they work in the same way — they both catalyse the reaction of two molecules of hydrogen peroxide with each other to form oxygen and water:
2H
씮
2O2
2H2O + O2
This simple reaction mechanism reveals a great deal about conditions on the Earth 3.5 billion years ago. It is the exact equivalent of the natural reaction between two molecules of hydrogen peroxide, but is speeded up 100 million times by the enzyme. The need for two molecules of hydrogen peroxide means that catalase is extremely effective at removing hydrogen peroxide when concentrations are high, when it is easy to bring two molecules together. It works less well at low concentrations of hydrogen peroxide, when it is harder to find two molecules close together. Catalase is thus swift to remove high concentrations of hydrogen peroxide, but is poor at mopping up trace amounts or at maintaining a stable low-level equilibrium.
Today, most aerobic organisms have a second group of enzymes, known as the peroxidases, which can dispose of trace amounts of hydrogen peroxide. These enzymes work better at low levels of hydrogen perox-
Radiation and the Evolution of Photosynthesis • 141
ide because they act in a fundamentally different way. Rather than bringing two molecules of hydrogen peroxide together, they use antioxidants such as vitamin C to convert a single molecule of hydrogen peroxide into two molecules of water, without generating any oxygen. Most aerobic cells have both sets of enzymes, and break down hydrogen peroxide using both mechanisms. Catalase is used for bulk removal, peroxidase for subtle adjustments.
We might infer that any cell using catalase would need to cope with large fluxes of hydrogen peroxide, at least occasionally. Catalase is highly specialized: it has no other known target and works at an extraordinary speed. Such tremendous efficiency does not appear out of the blue by chance: we might as well believe that the eighteenth-century theologian Tom Paley stumbled across a nuclear reactor, rather than his celebrated watch, and instead of inferring the hand of a designer, ascribed it to an accidental arrangement of the elements.
There is nothing accidental about catalase. If it was present on the early Earth, before photosynthesis, then there must have been hydrogen peroxide too, and in abundance. This is counter-intuitive, to say the least.
Is it really credible that the early Earth could have been so rich in hydrogen peroxide that there was a selective pressure for the evolution of catalase?
As we have seen (Chapter 6), Mars is rich in iron peroxides; but their abundance in Martian soils tells us nothing about how quickly they were formed on the early Earth. While they were almost certainly formed on Earth (which is, after all, closer to the Sun, and so more drenched in ultraviolet rays), the abundance of hydrogen peroxide on Earth would have depended on its rate of formation and destruction — and these in turn are dependent on atmospheric and oceanic conditions. While the existence of catalase implies that hydrogen peroxide was indeed abundant, the story is suggestive but far from conclusive. Luckily, there are other ways to answer the question, and they not only support the notion that photosynthesis evolved in response to oxidative stress, but they also explain a few other long-standing paradoxes.
One of the most respected atmospheric scientists of recent decades is James Kasting, now at Pennsylvania State University, and during the 1980s at the NASA Ames Research Centre in California. In the mid-1980s, Kasting set out to answer the question of just how abundant hydrogen
142 • GREEN PLANET
peroxide might have been on the early Earth. He was not really aiming to answer questions about the evolution of photosynthesis but rather to look at the time-line for oxygen production.
As we saw in Chapter 3, a surrogate measure for the accumulation of oxygen in the air is the extent of iron-leaching from fossil soils. Because iron is soluble in the absence of oxygen, it can be washed out of soils by rainfall on an oxygen-free planet. As oxygen builds up in the atmosphere, it reacts with iron in the soil to produce insoluble rusty iron deposits, which cannot be leached out by rainfall in the same way. In theory, then, fossil soils preserve a record of atmospheric oxygen levels in their iron content — the more oxygen there is in the air, the more iron is left in the soil. The trouble is that the fossil-soil record can be read to imply that oxygen began to accumulate in the air well over 3 billion years ago (long before the major rise 2 billion years ago). This early date does not tally with the sulphur-isotope measurements discussed in Chapter 3, or with the larger-scale deposits, such as banded iron formations, red-beds and uranium ores. Kasting was interested in the discrepancy.
Earlier studies of fossil soils had tacitly assumed that the most important oxidant dissolved in rainwater had always been oxygen itself. Kasting queried this assumption and set out to compute the possibility that hydrogen peroxide had been the most important oxidant in rainwater before the advent of atmospheric oxygen. In a detailed theoretical paper, Kasting, working with Heinrich Holland and Joseph Pinto at Harvard, calculated the rate at which water is split by ultraviolet rays under a variety of conditions. They then took into consideration the solubility of the degradation products (such as hydrogen peroxide) in rain droplets, to calculate their likely steady-state concentrations in rainwater and in lakes on the early Earth.
Under the most likely conditions 3.5 billion years ago — high carbon dioxide levels, a trace of oxygen (less than 0.1 per cent of present atmospheric levels) and virtually no ozone screen — Kasting calculated that there should have been a continuous flux (based on the rate of formation and removal by reaction or rainfall) of about 100 billion molecules of hydrogen peroxide per second per square centimetre.
Although this number sounds fantastically big, we should bear in mind the inconceivably large number of molecules that make up matter. There are said to be more molecules in a single glass of water than there are glasses of water in all the oceans. We should not be too surprised to discover, then, that 100 billion molecules of hydrogen peroxide weigh about
Radiation and the Evolution of Photosynthesis • 143
56 thousand billionths of a gram.5 To put these numbers into some sort of perspective, Kasting calculates that dissolved hydrogen peroxide, which is much more soluble than oxygen, accounts for between 1 and 6 per cent of the total oxidant concentration in rainwater today. There is no reason why the amount of hydrogen peroxide in rainwater should have been any less 3 billion years ago, and it may well have been higher, as the intensity of ultraviolet radiation was more than 30 times greater.
Such a large flux of hydrogen peroxide must have placed the first cells under oxidative stress. The level of stress would have been exacerbated by the reactivity of hydrogen peroxide in comparison with ox
ygen. In particular, hydrogen peroxide reacts quickly with dissolved iron, to produce hydroxyl radicals, whereas oxygen reacts much more slowly. In today’s well-oxygenated oceans, the reactivity of hydrogen peroxide is limited by the low availability of dissolved iron (which long ago reacted with oxygen and precipitated out as banded iron formations), but during the early Precambrian, the oceans were so full of dissolved iron that hydrogen peroxide must have been continuously reacting with iron to produce hydroxyl radicals via the Fenton reaction. Thus, not only was there more hydrogen peroxide on the early Earth, it was also more likely to react to produce oxidative stress.
The effect that hydrogen peroxide had on the environment must have depended on the amount of iron available. In the deep oceans there was such a vast amount of dissolved iron that any hydrogen peroxide dissolved in rainwater could never have altered the overall chemical balance.
In the shallow seas and freshwater lakes, however, there was much less iron. These low levels of iron could plausibly have been depleted, or exhausted, by a steady drizzle of hydrogen peroxide. With the loss of iron and hydrogen sulphide, such secluded environments would have grown steadily more oxidized. According to the mathematical models of Hyman Hartman and his colleague Chris McKay, at the NASA Ames Research Center, the sheltered lakes and sea basins may well have become oxidizing enough to stimulate the evolution of antioxidant enzymes such as catalase. Once this had happened, bacteria living in shallow-water environments would have been pre-conditioned to the appearance of free oxygen.
5 This can be calculated from Avogadro’s number, the number of molecules in one mole of any substance, which is 6.023 ⫻ 1023. One mole of hydrogen peroxide weighs 34 grams.
One gram of hydrogen peroxide therefore contains 1/34 ⫻ 6.023 ⫻ 1023 molecules, or about 177 ⫻ 1021. A hundred billion molecules of hydrogen peroxide weighs 100 billion/177 ⫻
1021 grams, or 56 ⫻ 10–12. This is one 56 thousand billionths of a gram.
144 • GREEN PLANET
Thus, there are good grounds for thinking that there was indeed plentiful hydrogen peroxide on the early Earth, and that it built up in sheltered environments. The oxidation of such environments by hydrogen peroxide was probably a strong enough selective pressure to stimulate the evolution of the antioxidant enzyme catalase. Catalase itself seems to have been the basis of the oxygen-evolving complex, enabling the evolution of oxygenic photosynthesis. So far the story makes sense, but we are left with one difficult question. Why would oxygenic photosynthesis evolve from catalase?
Catalase would presumably have been present in the photosynthetic bacteria that generated energy by splitting hydrogen sulphide or iron salts in the era before oxygenic photosynthesis. In fact, hydrogen peroxide has some parallels with these early photosynthetic fuels. To remove electrons from hydrogen peroxide requires a similar input of energy to that required to remove electrons from hydrogen sulphide, and so could have been achieved using the same bacteriochlorophyll. Hydrogen peroxide would therefore have been a good source of hydrogen for photosynthesis. And, while far less plentiful than hydrogen sulphide and iron salts, it was nonetheless formed most readily in the surface waters, closest to the full power of the Sun. If this scenario is true, then catalase could have doubled as a photosynthetic enzyme. Because splitting hydrogen peroxide generates oxygen, this recruitment of catalase to photosynthesis also bridges the evolutionary gap between anoxygenic and oxygenic photosynthesis.
If catalase was acting as a photosynthetic enzyme, then it would be natural for a number of catalase molecules to cluster around the photosynthetic apparatus. In these circumstances, it would be simple for two catalase molecules to became associated as a complex: the prototype oxygen-evolving complex. At first it would have continued to use hydrogen peroxide as an electron donor, but given the right energy input, this complex could split water. We know that three small changes in the structure of bacteriochlorophyll can transform its properties, enabling it to absorb high-energy light at a wavelength of 680 nm. We now have a prototype oxygen-evolving complex (the nut-cracker that can physically split water) and a chlorophyll that can provide enough energy for it to do so (or the hand that presses the nut-cracker). Thus, with no foresight and no disadvantageous steps, we have taken a path leading from anoxygenic photosynthesis to oxygenic photosynthesis.
The evolution of oxygenic photosynthesis, then, seems practically inevitable, as long as three conditions are met: a selective pressure to use
Radiation and the Evolution of Photosynthesis • 145
water; a mechanism for splitting water; and a tolerance to the oxygen waste. The selective pressure to use water was the loss of iron and hydrogen sulphide from sheltered environments. The mechanism for splitting water was a simple binding together of two catalase molecules. Tolerance for oxygen was imparted by catalase, and probably several other antioxidant enzymes which had evolved in response to oxidative stress from ultraviolet radiation.
These conditions could never have been fulfilled in the deeper oceans. They were full of iron and hydrogen sulphide, and shielded from the effects of ultraviolet radiation. Life there would have had no need to tolerate oxygen. In these places, even if given enough light, any mutations that produced chlorophyll from bacteriochlorophyll would have been eliminated by natural selection as worse than useless. They would have slashed the light-capturing capacity of bacteria without any gainful return.
The explanatory power of an ‘oxidative stress before free oxygen’ hypothesis is strong. If true this reverses received wisdom. The hypothesis implies that photosynthesis would not have been possible without the oxidative stress generated by ultraviolet radiation. Far from cowering away at the bottom of the oceans, in sulphurous hydrothermal vents (or black smokers), life embraced the surface oceans very early, and dealt with the conditions there through the evolution of potent antioxidant enzymes such as catalase. Without these radiation-scorched conditions, water-splitting photosynthesis could never have evolved. Even more significantly, the evolution of oxygenic photosynthesis hangs by a single thread: the accidental association of two catalase molecules.
If this hypothesis seems to be over-reliant on a single lucky chance, it is worth remembering that, unlike flight or vision, which evolved independently many times, oxygenic photosynthesis only ever evolved once. All algae, all plants, the entire green planet, use exactly the same system. All of them inherited it from the cyanobacteria, which invented it once, perhaps 3.5 billion years ago. No other cells on Earth ever learnt to split water. All known water-splitting complexes are related in structure, and all are similar to catalase. Perhaps life once existed on Mars, but found another way of dealing with the less-intense solar radiation. Catalase never evolved. Without catalase, oxygenic photosynthesis never
146 • GREEN PLANET
evolved. Without photosynthesis, free oxygen never accumulated in the air. And without oxygen, there was no multicellular life, no little green Martians, no war of the worlds.
Convinced? Perhaps not, but there is more. To my mind, the most conclusive evidence comes not from atmospheric modelling, or structural and functional similarities, but from comparative genetics. Not the genetics of photosynthesis, which are still rather murky, but the genetics of respiration: how life came to use free oxygen as a means of extracting energy from food in the first place. Again, intuition is turned on its head.
It seems impossible for oxygen-respiring organisms to have evolved before there was any free oxygen in the air. Surely they could not have evolved before oxygenic photosynthesis! We may have to think again.
According to another iconoclastic viewpoint, put forward and backed with increasingly strong evidence by José Castresana and Matti Saraste of the European Molecular Biology Laboratory in Heidelberg, this is exactly what happened.6 Respiration using oxygen evolved before photosynthesis, oxygen breathers before there was any free oxygen in the air. The arguments of Castresana and Saraste hinge on the identity of a single-celled creature named LUCA, the Last
Universal Common Ancestor. We will find out who she was in the next chapter.
6 Sadly, Matti Saraste died on the 20 May 2001 at the age of 52. His colleagues in Heidelberg let his own words speak as his testimonial, and there is no better tribute to his memory, or celebration of the fun and magic of biochemistry than these words. I hope they will inspire students in future generations to become biochemists. “The nicest aspect of biochemistry is the possibility to combine mental and practical work. One can even do these simultaneously. For me, controlling the practical work, the experiment, is extremely fun. An experiment is to approach the current scientific problem at the border of the known and unknown. It is as much fun to try to grasp the location of this magic border in your mind.
To be a good biochemist, you do not have to be an egg-head or an absolute genius in maths or physics, but understanding the problems requires thinking, reading, experience and planning. On the other hand, you can keep your hands busy with minimal thinking: the bottleneck in research is often the experimental work.” Matti Saraste, 1985.