Oxygen

by Nick Lane

  children per couple. The Hayflick limit is just as irrelevant to preventing cancer. The most likely answer is that, like most genes in most cells of the body, telomerase is switched off simply because it is not needed.

  How is it possible, then, that normal cells can be transformed into immortal cells, just by adding the gene for telomerase? How do mitochondria fit into this story? I had my first inkling of this a few years ago, when I started growing kidney tubule cells in culture, and wasted weeks in the

  272 • GENDER BENDER

  lab. I had taken a few lessons from people growing other sorts of cell, and had applied their methods to my own problem. Each time, my culture plates became overgrown with spidery looking cells, which I assumed were fibroblasts. Fibroblasts do very well in cell culture, and even a trifling contamination can lead to them taking over the plate. I threw away my fibroblasts and started again, more conscious this time of good technique.

  The same thing happened again and again. Finally I went to see a fibroblast specialist, who looked at my plates and laughed — “They’re not fibroblasts”, he said, “I don’t know what they are, but they’re absolutely not fibroblasts. They’re probably your kidney cells!”

  I was shocked. I had spent hours looking at kidney sections down the microscope, and I knew what tubule cells looked like: prolific brush borders, providing a massive surface area for reabsorbing solutes, and thousands of mitochondria, packed together like a Roman phalanx. My cultured cells had no brush border, and I could see no more than a handful of mitochondria. There was nothing for it but to turn to the textbooks and the original papers. I was in for another shock. My sad cells were exactly what kidney tubule cells were supposed to look like in culture! I had been planning experiments to see how vulnerable the cells would be to oxygen, and whether they could be protected by antioxidants, but now, after reading the small print, I realized that cultured kidney tubule cells do not require oxygen at all: they live quite happily by anaerobic respiration. In fact, the only way to get them to breathe oxygen is to deprive them of glucose in the culture medium, and to catch them in the act of growing, before they have completely covered the dish. I aban-doned my experiments as irrelevant to real kidneys, chastened but a little wiser.

  This pattern is characteristic of cells grown in culture: they don’t need much energy, so they don’t have many mitochondria. In fact, this is true not just of cells grown in culture, but of any cells that have a low energy expenditure. Perhaps surprisingly, these include actively dividing cells, such as stem cells and cancer cells: their energy requirement is much lower than that required for specialized metabolic work. Just think of the brain: it accounts for 20 per cent of resting oxygen consumption, but only 2 per cent of body weight. If the oxygen supply to the brain is cut off for more than a couple of minutes, we lose consciousness. Neurons do not divide, and the brain’s support network, the glial cells, only divide occasionally — the brain needs all this oxygen for its normal metabolic work. Other metabolically active body tissues have a similarly grasping

  The Rate of Living and the Need for Sexes • 273

  demand for oxygen. Liver, kidney and heart-muscle cells have as many as 2000 mitochondria each, so densely packed that it can be hard to see any cytoplasm at all. In contrast, stem cells, whose job it is to replenish cell populations with a continuous turnover, such as skin cells, have remarkably few mitochondria. Similarly, immune-system cells such as lymphocytes, which again divide frequently once activated, are virtually devoid of mitochondria.

  In general, there is a striking relationship between the degree of cellular differentiation — the cell’s commitment to a metabolic purpose — and the number of mitochondria. Specialized differentiated cells have large numbers of mitochondria and suffer the consequences — serious oxidative stress. Stressed cells gain the most benefit from improved stress-resistance. Recall, for example, that the protective effects of calorie restriction are most marked in the long-lived cells of tissues where oxidative stress is worst, such as the brain, heart and skeletal muscle. This, not telomerase, is what really confers immortality on populations. To survive, get rid of your mitochondria — throw them overboard like so much ballast. Cancer cells do. Cancer cells become less differentiated as they multiply, and lose their mitochondria in the process. They thrive on anaerobic respiration. Most tumours are a dense mass of tissue with a low requirement for oxygen. Indeed, oxygen is toxic to many tumours: radiotherapy is three or four times more effective when the tumour is oxygenated. As is so often the case, the exceptions prove the rule. Some cancer cells are rich in mitochondria. In particular, some glandular tumours (oncocytomas) and liver tumours (Novikoff hepatomas) have cells with huge numbers of mitochondria. In both cases, however, close biochemical inspection suggests that the mitochondria in the tumours are not actually functional. Thus, cells can reproduce indefinitely if they have active telomerase and a small number of relatively inactive mitochondria.

  There is a second factor that helps to preserve rapidly dividing cells: their fast turnover. When a cell divides, it must reproduce its cytoplasm and mitochondria, as well as its DNA. This means that mitochondria replicate faster in rapidly dividing cells than in non-dividing cells, even if the latter are packed with mitochondria. In most cells, the mitochondrial population varies from good condition to completely shredded. Undamaged mitochondria replicate faster than damaged mitochondria. Each time a cell divides, then, the new pool of mitochondria is derived from the least-damaged survivors of the last pool, and this helps to replenish

  274 • GENDER BENDER

  the population. In rapidly dividing cancer cells, then, we can predict that there should be relatively few mitochondria in relatively good condition.

  In the case of non-dividing cells, where the rate of mitochondrial replication is much slower, the rate of mitochondrial breakdown becomes more important. In these cells, mitochondria are normally replaced every few weeks. In non-dividing cells, partially damaged mitochondria may be broken down more slowly than healthy mitochondria, and this discrepancy can lead to a takeover by damaged mitochondria.8 The phenomenon is known as the ‘survival of the slowest’, or SOS, and may contribute to the demise of old differentiated cells.

  We can conclude that the lifespan of a cell depends on the activity of its mitochondria and the efficiency of its damage-prevention and repair systems. Damage-prevention and repair are never 100 per cent efficient, so energetic cells will eventually accumulate defective mitochondria. In the end, these will undermine the integrity of the cell. This situation is exacerbated in non-dividing cells, which cannot replenish their mitochondrial populations by selecting for less-damaged mitochondria. We therefore have a spectrum of potential longevity, ranging from stem cells and cancer cells, which are virtually immortal, through to neurons, body muscle cells and heart muscle cells, which are doomed from the moment that they specialize in tasks that require large amounts of energy. In principle, the lifespan of these metabolically active cells can be extended by building up their ability to resist oxidative stress. However, all the energy directed at cell renewal is subtracted from tasks that the cell would normally perform. To protect neurons from free-radical attack, energy must be diverted away from thinking or coordinating the body — obviously to the detriment of performance. Longevity and biological fitness are therefore incompatible, and we must find an optimal trade-off. Can we live longer? Perhaps: some tortoises can live for 200 years, but their success does not depend on quick movement and sharp wits. Their shells confer a different kind of protection, enabling them to be less metabolically active. They have a different trade-off.

  8 The idea, proposed by Aubrey de Gray at the University of Cambridge, is that damage to mitochondrial DNA may prevent oxygen metabolism, and this would paradoxically lower the burden of free radicals. As a result, the mitochondrial membranes would be less damaged than normal, and so their turnover would be lower. This may sound implausible, but it is supported by empirical data.

  The
Rate of Living and the Need for Sexes • 275

  Two broad conclusions emerge from this analysis. First, Weismann was wrong again: there is no fundamental distinction between germ cells and somatic cells. Some somatic cells, such as cancer cells, achieve immortality by losing mitochondria and replicating quickly. This is how the simple, tentacled Hydra courts immortality: it has a large pool of stem cells, which can develop into any one of the mature cells in its body. It continually replaces its worn-out cells. The price for this is a simple body plan, allowing cells to be replaced without affecting the function of whole organs. Our own stem cells might possess similar powers of regeneration

  — think only of cloning — but our body structure is fundamentally different: as noted earlier, we cannot replace the neurons in our brain while at the same time maintaining a sense of continuity and experience. As one body system begins to wear out, the repercussions are felt by others. If the ageing pituitary gland in the brain starts producing fewer hormones, the vitality of stem cells in the skin will inevitably be affected. Until we find a way around this problem, we will never outlive our neurons.

  Second, the disposable soma theory is not just about sex. Sex steals resources that we would otherwise use for staying alive, but then so does being human. If we are to think, run, create, interact — anything that makes us human — we sell ourselves a short life. Perhaps Raymond Pearl was right after all — perhaps laziness does pay, so long as we don’t eat or drink our way to an early death. The idea is supported by a recent best-selling book, The Okinawa Way, written by a Japanese cardiologist and two American colleagues. Based on a 25-year study, the book argues that the secret of the Okinawans (the inhabitants of a Japanese island with more centenarians than anywhere else in the world) goes beyond genes, diet and exercise to their relaxed lifestyle and low levels of stress. The Okinawans have a word for it, tege, which means ‘half-done’: forget timetables, forget finishing today things that can be done tomorrow. I suspect they are probably right.

  We are left with a final problem, which still has the power to make or break the arguments so far. I have argued that mitochondrial respiration will be the death of us. In killing cells, mitochondria first damage their own integrity. But if all mitochondria damage themselves, how do mitochondria-bearing organisms evade decay over generations? How is it that babies are born young?

  276 • GENDER BENDER

  The situation reminds me of the decline and fall of the Byzantine Empire. If we believe the eighteenth-century historian, Edward Gibbon, the empire was in a state of continuous decline for 1000 years. Some emperors succeeded in reviving its flagging fortunes temporarily, but the

  “corrupting spirit” of the Greeks meant it was only ever a matter of time before the empire fell. Gibbon’s ‘corrupting spirit’ meets its match in mitochondria: it should be only a matter of time before mitochondria corrupt their hosts, even if it takes 1000 generations. Yet the fall of Constantinople has no echo in nature: how have we avoided the corrupting spirit?

  Let me define the problem. Undamaged mitochondrial DNA is necessary for the function of any organism. Sex cells must pass on newly minted mitochondria, so the mitochondrial DNA must somehow be rejuvenated.

  The problem is that mitochondria replicate their DNA asexually. As we have seen, sex is reinvigorating for genes, but without sex it is hard to see how the mitochondrial genome can regenerate itself. How can it reset its biological clock to zero in a newborn infant? Free-living asexual organisms such as bacteria preserve genetic integrity over generations by combining rapid reproduction with heavy natural selection. Bacterial-style selection for mitochondria is out of the question, however. They would have to replicate themselves at the same rate as cancer cells, and we would turn into mitochondrial tumours. This paradox — how mitochondria regenerate themselves without either sexual recombination or heavy selection — is known as Muller’s ratchet and seems insurmountable; but clearly it is not. So how do they do it?

  To understand the solution to this conundrum, we must think about the fate of mitochondria following the act of sex, in particular the fate of sperm mitochondria. Wriggling human sperm are a familiar sight to television audiences: we all know that their power and endurance is extraordinary. Surprisingly, there is some confusion about exactly how they power their performance. There is a common misconception that sperm are too small to contain mitochondria. In fact, sperm contain about 40 to 60 mitochondria, encased in the midpiece. The sperm’s mitochondria enter the fertilized egg, along with the midpiece, but may not survive there for long. What becomes of them is uncertain, but essentially all our mitochondria are inherited from our mothers. This is true not just for us, but also for the great majority of sexual organisms, including plants.

  Why male mitochondria are not passed on to the next generation has taxed the minds of some of the finest biologists. The most widely accepted

  The Rate of Living and the Need for Sexes • 277

  general explanation was articulated well by John Maynard Smith and Eörs Szathmáry in their book The Origins of Life. Essentially, if mitochondria are inherited from both parents, the stage is set for the evolution of

  ‘selfish’ organelles. The argument is as follows. When a cell divides, all its nuclear DNA is replicated, with half going to each daughter cell: the two daughters have identical sets of genes, so there is no unequal competition. This is not true of mitochondria, which have their own DNA and use it to replicate independently. The overall make-up of the mitochondrial population in a cell therefore depends on the speed at which individual mitochondria replicate (or break down), and this makes the cell vulnerable to abuse. Any mutation in mitochondrial DNA that increases the speed of a mitochondrion’s replication will lead to its progeny taking over the entire cell and its descendants, even if the same mutation makes them worse at oxygen respiration (indeed, especially if it makes them worse at respiration, as they will damage themselves less). If the mutant mitochondrion is in a sex cell, then the entire organism will become compromised as it grows. According to the selfishness theory, the proliferation of selfish mitochondria is prevented by the device of uniparental inheritance, in which only one parent provides all the mitochondria. Instead of mixing unrelated mitochondria from the merger of two similar (but otherwise unrelated) sex cells, one sex specializes in providing all the mitochondria, while the other specializes in providing none at all. Thus, gender grew out of an evolutionary trick to exclude selfish mitochondria.

  This theory is almost certainly true in some instances, but there are two objections to it as a general explanation. First, a mutation that causes uniparental inheritance is only advantageous if the selfish mitochondria are waiting in the wings to take advantage. This is improbable: any organism that harbours selfish mitochondria is less likely to survive and reproduce than a robustly fit organism. It is equivalent to pitting a broken old man, with his broken mitochondria, against a heroic youth, to win the favours of a lady. In fact, the chances are that an organism with defective mitochondria would not even get through development. Think of the difficulties many couples experience in getting pregnant. Some evidence suggests that a large proportion of embryos fail to develop past the early stages of pregnancy because of a problem with their mitochondria. Such problems may also account for the high failure rate of cloning experiments.

  Second, a number of species do not rely on uniparental inheritance at all — their mitochondria are inherited from both parents. They

  278 • GENDER BENDER

  circumvent the selfishness problem somehow. As we have seen, the same may even be true of us to a lesser extent — the fate of sperm mitochondria is uncertain. Some researchers put their apparent disappearance down to a simple dilution effect. This has not been ruled out. A sperm has 40 to 60 mitochondria, while a human egg cell is believed to contain more than 100 000. The dilution factor is therefore at least 1000, which is in fact below the detection limit of many techniques for detecting mitochondrial DNA. Some studies in mice, us
ing more sensitive techniques, suggest that male mitochondrial DNA is present at a frequency of between 1 in 1000 and 1 in 10 000, relative to the maternal contribution: very close to what one would expect from dilution alone. The issue is still unresolved, although we shall see in a moment that recent work does suggest a solution.

  These two objections to the selfish mitochondria theory question its validity as an explanation for the evolution of sexes; but the fact remains that some animals go to bizarre lengths to exclude the male mitochondria. A few species of Drosophila apparently sequester the sperm mitochondria in the gut of the larvae during development, and defecate them out soon after hatching. Something peculiar is going on. At the level of the sex cells, mitochondrial transmission is virtually the defining difference between the sexes. Why should this be? The answer, in my view, was spelled out in 1996 by John Allen, a biologist at Lund University in Sweden, in the Journal of Theoretical Biology — a journal always crammed with impressive erudition and lofty ideas, from the sublime to the ridicu-lous. Allen draws on the logic of the mitochondrial theory of ageing to explain the evolution of two sexes. In essence, he argues that male mitochondria are not passed on to the next generation because they are time bombs: they have been fatally damaged by oxygen, and if passed on would cause the birth of prematurely aged babies. Breathing oxygen dictates the need for two sexes. If so, then oxygen is the ultimate gender bender.