Denham Harman, pioneer of free radicals in biology, first proposed the mitochondrial theory of ageing in 1972. Harman's central point was simple: the mitochondria are the main source of oxygen free radicals in the body. Such free radicals are destructive, and attack the various components of the cell, including the DNA, proteins, lipid membranes, and carbohydrates. Much of this damage could be repaired or replaced in the usual way by the turnover of cell components, but hotspots of damage, most notably the mitochondria themselves, would be harder to protect simply by consuming dietary antioxidants. Thus, spake Harman, the rate of ageing and the onset of degenerative diseases should be determined by the rate of free-radical leakage from mitochondria, combined with the cell's innate ability to protect against, or repair, the damage.
Harman based his argument on the correlation between metabolic rate and lifespan in mammals. He explicitly labelled the mitochondria the 'biological clock'. In essence, he said, the faster the metabolic rate, the greater the oxygen consumption, and so the higher the free-radical production. We shall see that this relationship is often true, but not always. The proviso may seem trivial, yet it has confounded an entire field for a generation. Harman made a perfectly reasonable assumption, which has proved untrue. Unfortunately, his assumption has become entangled with the theory in general. To disprove it does not disprove Harman's theory, but it does overturn his single most important, and best-known, prediction—that antioxidants can prolong life.
Harman's sensible but confounding assumption was that the proportion of free radicals that leak from the mitochondrial respiratory chains is constant. He assumed that the leakage is basically an uncontrolled, unavoidable by-product of the mechanism of cell respiration, which juxtaposes the passage of electrons down the respiratory chains with a requirement for molecular oxygen. Inevitably, the theory goes, a proportion of these electrons escape to react directly with oxygen to form destructive free radicals. If free radicals leak at a fixed rate, let's say 1 per cent of total throughput, then the total leakage depends on the rate of oxygen consumption. The higher the metabolic rate, the faster the flux of electrons and oxygen, and the faster the free-radical leakage, even if the proportion of free radicals actually leaking never changes. So animals with a fast metabolic rate produce free radicals fast and have short lives, whereas animals with a slow metabolic rate produce free radicals slowly, and live a long time.
In Part 4, we saw that the metabolic rate of a species depends on its body mass to the power of 2/3: the larger the mass, the slower the metabolic rate in individual cells. This link is largely independent of the genes, depending instead on the power laws of biology. Now, if free-radical leakage depends only on the metabolic rate, then it follows that the only way to prolong the lifespan of a species relative to metabolic rate, is to bolster the strength of antioxidant (or anti-stress) protection. An implicit prediction of the original mitochondrial theory of ageing is therefore that creatures living a long time must have inherently better antioxidant protection. Birds, then, which live a long time, must stockpile more antioxidants. Accordingly, if we wish to live longer ourselves, we must seek to bolster our own antioxidant protection. Harman supposed that the only reason we have failed to extend our own lifespan by taking antioxidant therapy so far (this back in 1972) is because it is difficult to target antioxidants to the mitochondria. Many people still agree with this position today, despite another thirty-plus years of industrious failure.
These ideas are tenacious but erroneous, in my opinion: they cling to the mitochondrial theory of ageing like burrs. In particular, the idea that antioxi-dants might prolong our lives is the basis of a multibillion-dollar supplement industry, which has remarkably little solid evidence to support its claims; unlike the biblical house that was built on shifting sands, it has somehow remained standing. For over thirty years, medical researchers and gerontolo-gists (including myself) have flung antioxidants at all kinds of failing biological systems and found that they just don't work. They may correct dietary deficiencies, and perhaps protect against certain diseases, but they don't affect maximal lifespan at all.
It is always difficult to interpret negative evidence. We are reminded, in that smug phrase, that 'absence of evidence is not evidence of absence'. The fact, if it is a fact, that antioxidants don't work, may always be related to the difficulties of targeting: the dose is wrong, or the antioxidant is wrong, or the distribution is wrong, or the timing is wrong. At what point are we entitled to walk away saying: 'No, this isn't a pharmacological problem—antioxidants really don't work'? The answer depends on the temperament, and there are some distinguished researchers who have yet to turn away. But the field as a whole did turn away in the 1990s. As two well-known free-radical gurus, John Gutteridge and Barry Halliwell, put it a few years ago: 'By the 1990s it was clear that antioxidants are not a panacea for ageing and disease, and only fringe medicine still peddles this notion.'
There are stronger reasons to challenge the standing of antioxidants too, and these come from comparative studies. I mentioned the prediction that animals with long lives should have high levels of antioxidants. For a time this prediction seemed to be true, but only after subjecting the data to a little innocent statistical jiggery-pokery. In the 1980s, Richard Cutler, at the National Institute of Ageing in Baltimore reported, somewhat misleadingly, that long-lived animals harbour more antioxidants than short-lived animals. The trouble was that he presented his data relative to the metabolic rate, and in so doing, he air-brushed out the far stronger association between metabolic rate and lifespan. In other words, a rat has a lower level of antioxidants than a human being, but only when antioxidant concentration is divided by metabolic rate, which is seven times faster in the rat; no wonder the poor rat seems so bereft of help. This manoeuvre concealed the true relationship between antioxidant levels and lifespan: a rat has actually far more antioxidants in its cells than a human being. A dozen independent studies have since confirmed that there is in fact a negative correlation between antioxidant levels and lifespan In other words, the higher the antioxidant concentration, the shorter the lifespan.
Perhaps the most intriguing aspect of this unexpected relationship is how closely antioxidant levels balance the metabolic rate. If the metabolic rate is high, then antioxidant levels are also high, presumably to prevent oxidation of the cell; yet lifespan is still short. Conversely, if the metabolic rate is low, then antioxidant levels are also low, presumably because there is a lower risk of cell oxidation; yet lifespan is still long. It seems that the body doesn't waste any time and energy in manufacturing more antioxidants than it needs—it uses them simply to maintain a balanced redox state in the cell (which means the dynamic equilibrium between oxidized and reduced molecules is kept optimal for the cell's function).1 The cells of short-lived and long-lived animals maintain a similar, flexible, redox state by counterpoising the antioxidant concentration against the rate of free-radical generation; but lifespan is not affected by antioxidant concentration in any way. We are forced to conclude that antioxidants are virtually irrelevant to ageing.
These ideas are borne out by the birds, which live long lives in relation to their metabolic rate. According to the original version of the mitochondrial theory of ageing, birds ought to have higher antioxidant levels, but again this is not true. The relationship is inconsistent, but in general birds have lower antioxidant levels than mammals, reversing the predictions. Another test-case
1 The status quo is also linked with tissue oxygen concentration. In Part 4, we noted that tissue oxygen levels are poised at 3 or 4 kilopascals across the entire animal kingdom. This means that both oxygen levels and antioxidant levels are balanced in the cell to retain a roughly constant redox state. We'll see why later in Part 7.
is calorie restriction. To date, calorie restriction is the only mechanism proved to extend the lifespan of mammals like rats and mice. Exactly how it works is debated, but the relationship with antioxidant levels in different species is ambiguous. Sometimes antioxidant concentrations go up, sometimes they go down, but there is no clearly consistent relationship. Even a piece of encouraging work from the early 1990s, suggesting that fruit flies live longer when genetically modified to express higher levels of antioxidant enzymes, turned out to be unrepeatable, at least in the hands of the original researchers (who make a distinction between strains that are long-lived and strains that are short-lived: higher antioxidant levels might extend the life of short-lived breeds of flies, in other words, they may correct a genetic deficiency). If any solid conclusion emerges from all this, it is certainly not that high levels of antioxidants prolong lifespan in healthy, well-nourished animals.
We've been confounded by the lure of antioxidants for a simple reason: the proportion of free radicals escaping from the respiratory chains is not constant—Harman's original assumption was wrong. While free-radical leakage often does reflect oxygen consumption, it can also be modulated up or down. In other words, far from being an uncontrolled and unavoidable by-product of cell respiration, the rate of free-radical leakage is controlled and largely avoidable. According to the pioneering work of Gustavo Barja and his colleagues at the Complutense University in Madrid, birds live long lives because they leak fewer free radicals from their respiratory chains in the first place. As a result, they don't need to have so many antioxidants, despite consuming large amounts of oxygen. Importantly, it seems that calorie restriction might work in a similar way. While there are various genetic changes, one of the most significant is a restriction in free-radical leakage from the mitochondria, despite similar oxygen consumption. In other words, in both long-lived birds and mammals the proportion of free radicals that leaks from the respiratory chains decreases.
This answer seems inoffensive enough but it is actually troublesome and hacks a hole in the established evolutionary theory of ageing. The problem is this. Animals that live a long time do so by restricting the free-radical leakage from their mitochondria. Because genes control the rate of ageing, it follows that in birds (and presumably in humans to a lesser degree) there has been selection to lower the rate of free-radical leakage. Fine. But if free radicals were simply damaging, why wouldn't a rat also do better by restricting its free-radical leakage? There seems to be no cost, indeed quite the contrary—there would be no need for the rat to go on manufacturing all those extra antioxidants to prevent itself from being oxidized. And surely it would have everything to gain, because a long-lived rat would have more time, and could leave behind more offspring. So rats, and by the same token humans, could live longer, cost-free, if they simply restricted free-radical leakage.
So why don't they? Is there a hidden cost, or do our ideas of ageing stand in need of radical revision? The cost of a long life is usually said to be a degree of impairment in sexuality. According to the disposable soma theory, first proposed by Tom Kirkwood, at the University of Newcastle, longevity is balanced against fecundity: long-lived species tend to have smaller litters, and rear them rather less frequently, than short-lived species. This is certainly true, at least in most known cases. The reason is less certain. Kirkwood suggested that the reason relates to the balance of resource use in individual cells and tissues: resources diverted towards attaining reproductive maturity, and raising litters, detract from those required to ensure longevity of the cell, such as DNA repair, antioxidant enzymes, and stress resistance—there are only so many ways to divide limited resources. Barja's data challenge this idea. Restricting free-radical leakage should have no cost on fecundity, as cellular damage is restricted without any need for better stress-resistance—the cost imputed in the disposable soma theory is negated. So, if the disposable soma theory is correct, there ought to be a hidden cost to restricting free-radical leakage; we shall see, in the final chapter, that there is indeed a hidden cost, and it holds vital connotations for our own quest to live longer.
To understand why, we need to consider another prediction of Harman's mitochondrial theory, which has also caused trouble. This held that free radicals don't necessarily damage the cell in general very much—they're mopped up by antioxidants—but they do specifically damage the mitochondria, especially their DNA. Harman actually mentioned mitochondrial DNA only in passing, but its involvement later became a fundamental tenet of the theory. Let's see why, for the gap between the predictions and hard prosaic reality reveals a great deal about what's really going on.
Was this article helpful?