Power, Sex, Suicide

by Nick Lane

  Why on earth did such a daft system evolve? That depends on one’s view of evolution. Stephen Jay Gould used to vent his frustrations about what he called the ‘adaptationist program’ in biology—the assumption that all is adaptation, in other words, everything has a reason, however innocent of function it may appear, and is shaped by the processes of natural selection. Even today, biologists can be split into those who are reluctant to believe that nature does anything for nothing, and those who believe that some things are just beyond direct control. Is ‘junk’ DNA really junk or does it have some unknown purpose? We don’t know for sure, and the answer you get will depend on whom you ask. Similarly, the ‘point’ of ageing is disputed. The most widely accepted view is that we are less likely to reproduce as we grow older, so natural selection is less able to weed out the genetic variants that cause damage late in life. Because mutations in mitochondrial DNA build up late in life, natural selection is unable to come up with an efficient mechanism of eliminating them. Only when the expected lifespan increases, as it does in animals isolated on an island without predators, or in birds that can fly away, or humans who use their large brains and social structures, can selection act to prolong life. If we subscribe to this view, then the sheer lunacy of storing badly protected mitochondrial genes in the incinerator is just one of those things: an accident of evolutionary history.

  Is this nihilistic vision correct? I don’t think so. The fault is that the line of reasoning is too stiffly chemical: it doesn’t take into account the dynamism of biology. We’ll see the difference this makes later. Nonetheless, it is a bold theory, and it has the great merit of making some explicit testable predictions. There are two in particular that we’ll look into. First, the theory predicts that mitochondrial mutations are sufficiently corrosive to bring about the whole sorry trajectory of ageing. This, we’ll see, is likely to be true. But the second prediction is probably false, at least in the full diabolical sense in which it was originally put forward—that mitochondrial mutations should accumulate with age. As they do they ought, ultimately, to bring about an ‘error catastrophe’. There’s little strong evidence to say that this happens. And therein lies the secret.

  Mitochondrial diseases

  The first report of a patient suffering from a mitochondrial disease was in 1959, a few years before the discovery of mitochondrial DNA. The patient was a 27-year-old Swedish woman, who had the highest metabolic rate ever recorded in a human being, despite a completely normal hormonal balance. It turned out the problem was a defect in mitochondrial control—her mitochondria respired at full blast even when there was no need for ATP. As a result, she ate prodigiously, yet always remained thin, and she sweated profusely even in winter. Sadly her doctors couldn’t help her and she committed suicide ten years later.

  A number of other patients were diagnosed with mitochondrial diseases over the following two decades, usually on the basis of their clinical records and various specific tests. For example, in many cases when the mitochondria are not functioning properly, lactic acid (a product of anaerobic respiration) accumulates in the blood, even when the patient is at rest. Biopsies of the muscles often show that some muscle fibres—typically not all—are badly damaged. They stain red on histological preparation and are known as ‘ragged red fibres’. When subjected to biochemical tests, the mitochondria in these fibres are found to lack the terminal enzyme in the respiratory chain, cytochrome oxidase, rendering them incapable of respiration.

  From a clinical point of view, such reports were little more than sporadic scientific curiosities. But a tidal change followed after 1981, when the dual Nobel laureate Fred Sanger and his team in Cambridge reported the complete sequence of the human mitochondrial genome. Through the 1980s and 1990s, as sequencing technology improved, it became possible to sequence the mitochondrial genes of many patients with suspected mitochondrial diseases. The results were startling, and showed not only how common mitochondrial diseases are—about 1 in 5000 people are born with mitochondrial disease—but also how bizarre. Mitochondrial diseases flout the normal rules of genetics. Their pattern of inheritance is peculiar, and often does not follow Mendel’s laws.2 The onset of symptoms may vary by decades, and occasionally diseases vanish altogether in individuals who ‘should’ (theoretically) have inherited them. In general, mitochondrial diseases progress with advancing age, so a disease that causes little inconvenience at the age of twenty may become utterly debilitating by the age of forty. Beyond this, however, very few generalizations can be made. Diverse tissues can be affected in individuals who carry the same mutation, while different mutations may well affect the same tissue. If you want to retain your sanity, don’t try to read a textbook on mitochondrial diseases.

  Despite such gross difficulties categorizing mitochondrial disease, a few general principles do help to explain the thrust of what’s happening. These same principles are pertinent to ageing. Recall from Part 6 that mitochondria are normally inherited from the mother, but even so, the variation in mitochondrial DNA in egg cells is surprisingly high. We saw that some degree of heteroplasmy (a mixture of genetically different mitochondria) is found in about half of the egg cells taken from the same ovary of a normal fertile woman. If these variations do not affect the functional performance of the mitochondria too seriously, then they are not eliminated during embryonic development.

  But why would a defect not affect embryonic development? There are various possibilities. One is that the defective mitochondria are inherited in low numbers. All mitochondrial diseases are heteroplasmic, which is to say there are both normal and abnormal mitochondria. Of the 100 000 mitochondria inherited in the egg, if only 15 per cent are abnormal then the healthy majority might mask their shortcomings. Alternatively, the mutation may be less harmful, but present in a greater proportion of mitochondria, perhaps 60 per cent. If so, then the embryo can still develop normally, despite the large number of mutants. And then there is the matter of segregation. When a cell divides, its mitochondria partition themselves at random between the two daughter cells. One cell might inherit all the defective mitochondria, whereas the other gets none, or there could be any combination in between. In the developing embryo, individual cells provide the mitochondria that eventually populate different tissues with different metabolic requirements. If the cells that develop into long-lived, metabolically active tissues, such as muscle, heart, or brain, happen to inherit defective mitochondria in high numbers, then all might be lost; but if instead the defective mitochondria end up in short-lived, or less metabolically active cells, like skin cells or white blood cells, then embryonic development might well be normal. As a result of such differing thresholds, the more serious mitochondrial diseases affect long-lived, energetically active tissues, especially muscle and brain.

  There are parallels with ageing here. We don’t inherit all of our defective mitochondria from the egg cell: some accumulate in adult life, due to free radicals formed by normal metabolism. This generates a mixed population of mitochondria in the cells affected. What happens next depends on the type of cell. If the cell is an adult stem cell (responsible for regenerating tissue), a possible outcome would be the clonal expansion of defective mitochondria. This happens in some muscle fibres, producing the ‘ragged red fibres’ characteristic of mitochondrial diseases, but also found in ‘normal’ ageing. Conversely, if the mutation affected a long-lived cell no longer capable of division, such as a heart-muscle cell or a neurone, then the mutation could not spread beyond the bounds of that single cell. We would then expect to see different mutations in different cells, forming a ‘mosaic’ of disparate mitochondrial function.

  Another aspect of mitochondrial disease is pertinent to normal ageing—their tendency to progress (to become more debilitating) with advancing age. The reason relates to the metabolic performance of tissues and organs. As we’ve seen, each organ has a threshold of function that is required for its normal performance. Symptoms only set in when an organ’s performance falls below its own
particular threshold. So for example we might be able to lose one kidney altogether and still function normally, but if the second kidney begins to fail too we’ll die, unless we receive dialysis or have a transplant. Because all work costs energy, an organ’s threshold depends on its metabolic requirements. Mitochondrial diseases are less serious if they happen to affect tissues with low metabolic requirements, like the skin, and are worse if they affect active cells like muscle cells. A similar process takes place in tissues as their cells age. A youthful muscle cell, in which 85 per cent of its mitochondria are ‘normal’, can handle all the energetic demands imposed on it in youth; but as its mitochondria decline with age, the energy demands placed on the remaining mitochondria rise. We draw closer to the metabolic threshold. As a result, the impairment caused by a mutant population is progressively unmasked as we get older.

  But are mitochondrial mutations sufficiently corrosive to bring about the whole sorry trajectory of ageing? Some of them certainly are. An awful condition, in which apparently normal babies lose their mitochondrial DNA soon after birth, leads swiftly to liver and kidney failure. When the disease is severe, as much as 95 per cent of mitochondrial DNA can be depleted, and the afflicted babies die in weeks or months, despite having appeared completely normal at birth. More common diseases include Kearns Sayre syndrome and Pearson’s syndrome, which cause substantial disability later in life, and premature death. Typical symptoms are similar to those caused by slow poisoning with a metabolic toxin like cyanide, and include loss of coordination (ataxia), seizures, movement disorders, blindness, deafness, stroke-like symptoms, and muscular degeneration. One mitochondrial mutation has even been associated with a condition similar to Syndrome X—that deadly combination of high blood pressure, diabetes, and raised cholesterol and triglycerides, said to affect 47 million Americans. Clearly, mutations in mitochondrial genes can have very serious effects, corrosive enough to underpin ageing. But other mitochondrial conditions are far less serious, and herein lies the problem.

  The severity of any mitochondrial disease depends on the proportion of mutant mitochondria and the tissue in which they find themselves, but also on the type of mutation: on which bit of the genome it affects. If the mutation affects a gene for a particular protein, the effects may or may not be catastrophic; indeed they might even be beneficial. On the other hand, if the mutation affects a gene encoding RNA, the consequences are usually serious. Depending on the type of RNA, the mutation could alter the synthesis of all mitochondrial proteins, or all proteins containing a particular amino acid. Mutations in the control region also potentially have serious consequences, as these might alter the entire dynamic of mitochondrial replication and protein synthesis in response to changing demand.

  Mutations can also occur in the nuclear genes encoding mitochondrial proteins, with similar consequences (except that these mutations follow a typical Mendelian inheritance pattern, with one gene coming from each parent; see footnote, page 281). If the nuclear mutation affects a mitochondrial transcription factor, which controls the synthesis of mitochondrial proteins, then the effects could in principle apply to all the mitochondria in the body. On the other hand, some mitochondrial transcription factors appear to be active only in particular tissues, or in response to particular hormones. A mutation in the gene for these would tend to have tissue-specific effects.

  Taken together, these considerations explain the extreme heterogeneity of mitochondrial diseases. A mutation may affect a single protein, or all proteins containing a particular amino acid, or all mitochondrial proteins altogether, or the rate of protein synthesis in response to changing demands. The mutations may be tissue-specific or global, affecting all the body. They may be inherited in a classic Mendelian fashion, if the mutations are in nuclear genes, or they may be inherited from the mother only if the mutations are in the mitochondrial genes. If the latter, the effects depend on the proportion of mitochondria affected, as well as the way in which they segregate in dividing cells during embryonic development, and the metabolic threshold of the organs involved.

  Given this degree of heterogeneity, the problem is the very spectrum of disease. While we tend to succumb to our own particular mixture of degenerative diseases, the underlying process of ageing is similar in all of us. How is it that we all share such a basic underlying similarity, not just with each other, but also with other animals that age at utterly different rates? If we accumulate mitochondrial mutations at random as we age, why do we not all age in very different ways and at different rates, as random and varied as the mitochondrial diseases themselves? The answer might conceivably lie in the nature of the mutations that accumulate, but this leads straight to a second problem: the magnitude and type of mutations that do accumulate don’t seem sufficient to cause ageing. So what’s going on?

  The paradox of mitochondrial mutations in ageing

  The pursuit of genetic mutations in ageing has proved to be a frustrating occupation. One promising theory held that nuclear mutations, accumulating over a lifetime, were the main culprit behind ageing. While nobody would dispute that mutations in nuclear genes do contribute to ageing, and especially to diseases like cancer, there is no relationship at all between lifespan and the accumulation of mutations in the nucleus, so they can’t be the primary cause.

  The generally accepted evolutionary theory of ageing, first postulated by J. B. S. Haldane and Peter Medawar, is a variation on the theme of mutations: if they don’t accumulate over a lifetime, perhaps they accumulate over many lifetimes. Natural selection has no power to eliminate the genes that defer their detrimental effects until later in life. The classic example is Huntington’s disease, which sets in well after reproductive maturity, enabling the gene to be passed on to children, so it can’t be eliminated by selection. While Hunting-ton’s disease is particularly horrible, how many other genes have similar late effects? Haldane proposed that ageing was little more than a dustbin of late-acting gene mutations, hundreds or thousands of them, which could not be eliminated by natural selection, and thus accumulated over many generations. Again, there must be some truth in this idea, but I don’t think it can be squared with the plasticity of ageing observed in nature. Nearly two decades of genetic studies have shown that lifespan can be extended dramatically, even in some mammals, by single point mutations in critical genes. If ageing really was written into the actual sequence of hundreds or thousands of genes, surely this couldn’t happen. Even if one critical gene controls the activity of many others, the subordinate genes are still mutated—it is their sequence, not their activity, which is the problem. To correct the sequences, there would need to be thousands of simultaneous mutations in all the right genes, to begin to have an effect on lifespan, and this would surely take several generations at the least. For whatever reasons, the fact is that lifespan is regulated, with a surprising degree of control, throughout the animal kingdom.

  Mitochondrial mutations have their own story, and this, too, is difficult to reconcile with the facts. On the face of it, though, mitochondrial genes hold more promise to explain ageing and disease. There are two reasons. First, as we saw in the last chapter, mitochondrial mutations accumulate far more quickly, from one generation to the next, than do nuclear mutations. It follows that mitochondrial mutations are more likely to build up within a single lifetime, and so in principle could tally with the rate of decline in ageing. And second, these mutations do have the corrosive power to undermine lives: they are not at all minor bit-players. The mitochondrial diseases underscore just how devastating such mutations can be.

  So how fast do mitochondrial mutations really accumulate? It’s difficult to say for sure because the rate of change over generations is restrained by natural selection. For most mitochondrial genes, the evolution rate is about 10- to 20-fold the rate found in nuclear genes, but in the control region it can be as much as fifty times faster. Because mutations are only ‘fixed’ in the genome if they don’t cause catastrophic damage (otherwise they are eliminated by select
ion) the actual rate of change must be faster than this. To get an idea of how fast the change might really be, Anthony Linnane, at Monash University in Australia, and his collaborators at the University of Nagoya in Japan, considered yeast in their classic Lancet paper published in 1989. Yeast is revealing because, as any brewer or vintner knows, it doesn’t depend on oxygen: yeast can also ferment to produce alcohol and carbon dioxide. Fermentation takes place outside the mitochondria, so yeast can tolerate serious damage to their mitochondria and still survive. Such damage was first noted in the 1940s with the discovery of ‘petite’ strains of yeast, whose growth is stunted. It turned out that the petite mutation involves the deletion of a large section of mitochondrial DNA, rendering the dispossessed mutant unable to respire. Critically, the petite mutation crops up spontaneously in cell cultures at a rate of about 1 in 10 to 1 in 1000 cells, depending on the yeast strain. In contrast, nuclear mutations occur at an almost infinitesimally slow rate, which is similar in both yeast and higher eukaryotes like animals—about 1 in every 100 million cells. In other words, if yeast is anything to go by, mitochondrial mutations accumulate at least 100 000 times faster than nuclear mutations. If such a fast mutation rate is true of animals too, then it could certainly account for ageing; indeed it would be hard to explain why we don’t drop dead almost immediately.

  The search was on: how quickly do mitochondrial mutations accumulate in the tissues of animals and people? It has to be said that this area is controversial and a consensus is only now emerging. Part of the problem is the technology used to measure the mutations: the techniques used to sequence DNA letters sometimes amplify mutated sequences at the expense of ‘normal’ sequences, making it very difficult to quantify the extent of damage. In consequence, results from different laboratories can be extremely variable, sometimes ranging by as much as 10 000-fold. As so often in all walks of life, those people who hope to find mitochondrial mutations tend to find them, whereas the doubters inevitably find but a few. This is almost certainly not because researchers are deliberately fabricating their data, but rather where and how they look: it’s possible for both sides to be right.