by Nick Lane
The report attracted interest, and Lindsay Farrer, at Boston University, wrote an editorial in the same issue of the Journal of the American Medical Association, in which she advocated a “global approach to bad gene hunting”. Both articles discussed a curious finding: in Ibadan, there was no association between ApoE4 alleles and the risk of Alzheimer’s disease. Various possible genetic and environmental reasons were explored, in particular high blood pressure and other vascular risk factors for dementia; but rather surprisingly, neither article mentioned malarial immunosuppression in Africa. This seems to me the most probable explanation. We noted in the last chapter that ApoE4 proteins are easily damaged by oxidative stress, and that this probably explains the link with Alzheimer’s disease. We also saw that people with two ApoE4 alleles gain more from antioxidants and anti-inflammatory drugs. In Ibadan, it is plausible that malarial immunosuppression blunted the severity of cerebral inflammation and so lowered the risk of Alzheimer’s disease.
The Indianapolis–Ibadan Dementia Project also demonstrated the reverse side of the coin, and this too was passed off with little comment: the mortality rate in the African cohort was nearly double that of the American cohort, despite their better cardiovascular health. What they died of is not stated. I imagine many must have died from infections or cancers. Studies in Tanzania, where malaria is also endemic, showed that the death rate fell substantially in regions where malaria had been controlled by draining swamps. The scale of this effect was larger than could be attributed directly to malaria, and prompted research into the ‘hidden morbidity of malaria’. This research has confirmed the suspicions: immunosuppression in areas where malaria is endemic perpetuates opportunistic infections, leading to the spread of diseases such as tuberculosis. In addition, cancers such as Burkitt’s lymphoma (a malignant cancer of B
cells) are common and linked with malarial immunosuppression, probably
330 • LIFE, DEATH AND OXYGEN
through infections in childhood with the Epstein-Barr virus, which are not properly cleared and persist.
Immunosuppression clearly influences health in old age, but at a potentially serious cost. Even so, there are grounds for hope, particularly if we gain the upper hand in the battle against infectious disease: we know that it is possible to lower the probability of age-related diseases by modulating the immune system. Whether or not this is a practical goal depends on exactly how it is done.
The mechanism of malarial immunosuppression is not well understood, and there are various competing theories. Scientists are people, and bring their own expertise, experience and biases to research problems.
This is a far cry from how philosophers tell us that science ‘should’ work, but the idea of the scientific method as an inductive procedure in which facts emerge from accumulating data is as misguided as can be. In reality, experiments are conceived and interpreted in terms of particular hypotheses, so data are generated on particular aspects of a problem rather than the whole problem. I will therefore put forward the ideas I think are the most intriguing and perhaps the most likely to be right.
In terms of malarial immunosuppression, I am struck by a paper in Laboratory Investigation in 2000, by Donatella Taramelli and her colleagues at the University of Milan, in which they studied the behaviour of isolated immune cells. Feeding malarial pigment to the immune cells brought about a rise in oxidative stress, which stimulated a fivefold rise in activity of the stress protein haem oxygenase. When the same immune cells were challenged for a second time, they did not respond normally, by pumping out inflammatory messengers, but instead had a ‘depressing’ effect on neighbouring cells, which became glum and unresponsive. Extrapolating to people, Taramelli argued that frequent malarial infections in childhood might produce a swing in the behaviour of the whole immune system, from activation to depression, by means of a continuous activation of haem oxygenase.
Let me try to put Taramelli’s findings in a broader medical context. If we grow up in an area where malaria is endemic, we can expect to be infected regularly in childhood. For the first few times, we will become ill, and we might die of cerebral malaria. If we survive, however, we adapt, or at least our immune system adapts. Instead of responding vigorously to infection, the immune system is restrained to some degree. The molecular detail is far from clear, but it seems probable that a new balance is estab-
Lessons From Evolution on the Future of Ageing • 331
lished, in which NFB is inhibited, and stress or antioxidant genes such as haem oxygenase are activated (probably via the counterbalancing transcription factor Nrf-2). The continuous activation of haem oxygenase prevents the immune system from causing too much collateral damage to our own body. If true, this scenario is all the more important because the mechanism is not unique to malaria: haem oxygenase and other stress proteins seem to underpin tolerance to common bacterial infections, and can provide almost complete protection against septic shock.5 It is therefore plausible that frequent childhood infections could bring about a persistent immunosuppression later in life. As with malaria, toning down our immune reaction to infections should make us at once more vulnerable to infections, and less vulnerable to autoimmune diseases and the diseases of old age.
There are three considerations that make me think this is true. First, haem oxygenase appears to be necessary for our normal health, even though it is a stress protein and supposedly ‘switched off’ in normal circumstances. Recall from Chapter 10 that an unfortunate six-year-old boy diagnosed with haem oxygenase deficiency suffered from vascular inflammation, severe growth retardation, abnormal blood coagulation, haemolytic anaemia and serious renal injury. He died at the age of seven.
Clearly, a regular dosing with haem oxygenase is necessary to temper inflammation. This idea is corroborated by studies of knock-out mice, in which the genes for haem oxygenase are mutated so that the protein is not produced. We saw that these mice have symptoms similar to those of people with chronic inflammatory diseases like haemochromatosis, including liver fibrosis, joint inflammation, restricted movement, weight loss, shrunken gonads and early death. Thus, haem oxygenase deficiency causes chronic inflammation and a short lifespan in both mice and men, whereas additional haem oxygenase suppresses the immune system and might potentially prolong lifespan.
5 Haem oxygenase has a powerful immunosuppressive effect. Overproduction can block the rejection of hearts transplanted from mice into rats, and even graft-versus-host disease in mice injected with spleen cells from a different species. Whether haem oxygenase directly inhibits NFB activation is not known, but the broader stress response does block the activation of NFB and haem oxygenase is one of the most prominent players in the stress response. Hector Wong and his colleagues at the University of Cincinnati have shown that stress elicits a rise in IB, the natural inhibitor of NF-B, which suppresses the activation of NFB. They also found that previous stress protects against a subsequent septic shock. Incidentally, psychological stress, which is well known to induce immunosuppression, may also operate through a stress response involving haem oxygenase.
332 • LIFE, DEATH AND OXYGEN
Second, the incidence of autoimmune diseases, such as insulin-dependent diabetes, Crohn’s disease and rheumatoid arthritis, is rising throughout the world, especially in Westernized countries. In Europe and the United States, the incidence of insulin-dependent diabetes has risen by an estimated 3–5 per cent a year in the past two decades. Hypersensit-ivity reactions, in which the immune system correctly recognizes foreign antigens, but then over-reacts to them, are also on the increase. The incidence of asthma and allergies has doubled in the last decade. Of the various possible reasons for this rise, one theory is gaining ground — the
‘hygiene hypothesis’. Simply put, too much cleanliness in childhood is bad: we need regular infections for our immune systems to develop properly, just as we must use our eyes to develop a visual understanding of the world.
A number of studies have shown that fre
quent infections in childhood are linked with a lower incidence of allergies and autoimmune diseases later in life, and vice versa. The assumption is that the immune system needs ‘house training’ in infancy, and if deprived of appropriate stimuli behaves like a bull in a china shop at the slightest provocation.
However, it is also plausible that regular infections in childhood could produce a persistent immunosuppressive effect, as in malaria. I am not aware of any clinical data to support this interpretation, but one animal study is intriguing: mice that are deficient in the transcription factor Nrf2, and thus in haem oxygenase and other stress proteins, go on to develop an autoimmune disease similar to lupus, leading to kidney failure. In other words, if the balance skews to inflammation, and away from immunosuppression, there is a higher risk of autoimmune disease.
My final consideration supports the relationship between infections in childhood and lifespan, and is based on the work of the cell biologist Giovanna De Benedictis at the University of Calabria in Italy, and her collaborators, including the demographer Anatoli Yashin at the Max Planck Institute for Demographic Research in Rostock, Germany. Yashin and De Benedictis spent the 1990s searching for ‘longevity’ genes in centenarians. Their basic idea is simple: some genes raise our chances of reaching a ripe old age, while others have a negative or neutral effect. The genes that prolong survival are most likely to be found in the people who did survive, so the best place to look for them is in centenarians. We imagine that the genes which prolong survival should make us ‘more robust’ in some sense. This is certainly true of some genes (which we will come to shortly), but Yashin and De Benedictis found that the situation is in real-
Lessons From Evolution on the Future of Ageing • 333
ity more complex. A surprising number of ‘longevity’ genes turned out to be linked with frailty (or susceptibility to disease) earlier in life. In other words, people who are ill a lot in their youth are more likely than most to survive to a ripe old age, as long as they don’t die first. Yashin and De Benedictis attributed this durability of the weak to adaptation, or as Nietzsche put it, what doesn’t kill us makes us stronger. So long as we avoid really serious illness, a weak disposition might perhaps lend itself to persistent immunosuppression, which reaps its reward in old age.
What can we conclude from infectious diseases? We will probably learn how to modulate the immune system with more subtlety than we know at present, and this should improve our health in old age. I suspect that stress proteins like haem oxygenase will hold the key, and we might even be able to modulate their levels by diet. Plants produce toxins to safeguard them against being eaten. Spices such as curcumin are known to stimulate the activity of haem oxygenase and other stress proteins (and show potential as anti-cancer agents). The trouble with curcumin is its bioavailability: when we eat it, very little is absorbed into the blood stream. How many other plant toxins, with better bioavailability, might stimulate the activity of stress proteins is anybody’s guess. As I suggested in Chapter 10, it is feasible that the benefits of a diet rich in fruit and vegetables go beyond their antioxidant content. Plant toxins, if palatable (and we have adapted to many over evolution), are likely to have beneficial effects on our immune system. I think this may help to explain why plants are clearly beneficial to our health, while antioxidant supplements are much less so.6
Even so, there is a dilemma at the heart of immune modulation, however refined it is: the benefits are always part of a trade-off between susceptibility to infections, on the one hand, and to age-related diseases on the other. Any benefits will depend on a delicate balancing act in which genes, diet, environment, behaviour and luck all have a role. I can see no systematic way of delaying ageing or preventing age-related diseases here. The only way this might be done ‘scientifically’ is to prevent 6 Curiously, a study by Chris Bulpitt and his colleagues at Imperial College, London (published in the Postgraduate Medical Journal in 2001) found that women who looked older than they really were had low levels of bilirubin in their blood stream, and vice versa. Bilirubin is an end-product of haem oxygenase. The implication is that high haem oxygenase activity makes women ‘look’ younger. In men, the strongest connection was with high levels of haemoglobin (which is, of course, broken down by haem oxygenase). Again, the implication is that high haem oxygenase activity makes men ‘look’ younger.
334 • LIFE, DEATH AND OXYGEN
the root cause of inflammation by targeting the mitochondria, so blocking the rise in oxidative stress in the first place.
The question is, how can we be more like birds? Mankind has always envied birds their power of flight, but now it seems we should envy them their mitochondria too. Bird mitochondria hardly leak any free radicals.
Why? The simple answer is that we don’t know, though there are some clues that might point us in the right direction. But before we speculate on these clues, are there people with ‘bird-like’ mitochondria, and do they live for longer? Again, the place to look is the centenarians.
The answer is hidden in a short research letter published in The Lancet in 1998 by Masashi Tanaka and his colleagues at the Gifu International Institute of Biotechnology in Gifu, Japan. In less than two columns, the Japanese group laid out a formidable series of studies on mitochondrial DNA in hundreds of centenarians, healthy volunteers and hospital patients. Their results injected new life into longevity-gene hunting the world over. What Tanaka and his colleagues found was that more people with one particular variant of a mitochondrial gene survived into old age: 62 per cent of centenarians carried the variant, known as Mt5178A, compared with 45 per cent of a random sample of healthy blood donors. Equally important, in a separate group of inpatients and outpatients at Nagoya University Hospital, only one third of patients older than 45 had the variant gene, whereas two thirds had the normal version. In other words, more older people with the ‘normal’ gene ended up in hospital, presumably because they were more susceptible to age-related diseases. This discrepancy did not apply to younger patients, who shared both versions in roughly equal measure. The implication is that the normal gene does not affect health earlier in life, so the spectrum of younger people in hospital reflects the genetic mix of the population as a whole. Taken together, these results suggest that people with the Mt5178A variant are more likely to survive to a hundred and less likely to suffer from age-related diseases than people with the normal version.
There are two points I want to make about this study. First, nearly half the random sample of healthy blood donors in Japan carried the mitochondrial variant Mt5178A. Elsewhere in the world, this variant is much rarer. In one study, for example, only five Asians and one European carried the variant out of 147 samples. Thus, the majority of Japanese
Lessons From Evolution on the Future of Ageing • 335
centenarians who carry the mitochondrial variant are the select survivors of a population in which it is already common. The frequency of the Mt5178A variant in the population as a whole may help to explain the long life expectancy of the Japanese: currently 84 at birth for women, and 77 for men. The less fortunate slight majority of the Japanese population who do not have the variant are nearly twice as likely to end up in hospital with age-related diseases. There could hardly be a clearer link between mitochondrial health and general health in old age.
The second point I want to make concerns the variant itself: it is a single-letter substitution in a mitochondrial gene (a C is replaced with an A). On what a slender thread hangs fate! We have about 35 000 genes, of which a mere 13 protein-coding genes are in the mitochondria instead of the nucleus. Of these 13 genes, a single-letter change in one of them is enough to halve our risk of getting any age-related disease, and virtually double our chance of living to a hundred. What on earth does this change in letter do? Well, at an arcane level, it causes a change of one amino acid in the protein encoded by the gene: a leucine is replaced with methion-ine. Why this should make a difference is not known, but I suspect the real significance lies in the
protein itself. The protein is a component of the respiratory chain, the long chain of proteins responsible for passing electrons to oxygen to generate energy. It is not just any component, but part of the first functional complex of the chain, complex 1 (NADH dehydro-genase). Complex 1 is a notoriously weak point in the chain, and the source of almost all escaping oxygen free radicals. I know of no studies to prove the point, but would be surprised if the single-letter change did not have an inordinately large effect on free-radical leakage from mitochondria. I will go further: this is exactly the kind of evolutionary change we would expect to find in bird mitochondria, making them more leak-proof. The pressure to select such changes in birds is much higher than in people, because flight itself demands very efficient energy production per gram body weight (the flight muscle needs to be lightweight and powerful
— efficient — to enable flight at all).
Mt5178A is not the only mitochondrial variant to be linked with ageing and disease in people. Several others have been identified, although their effects are less pervasive. We get a sense of their overall importance from a looser relationship: the maternal inheritance of longevity. Mitochondria, as we have seen, are only passed on in the egg, so all 13 mitochondrial genes come from our mothers. If these genes really do influence lifespan, and we can only inherit them from our mothers, then our own
336 • LIFE, DEATH AND OXYGEN
lifespan should reflect that of our mothers but not our fathers. This seems to be the case, despite the many other factors that impinge on survival, and was recognized as long ago as the nineteenth century by the American physician, poet and humorist, Oliver Wendell Holmes. In one of his famous ‘breakfast-table’ essays, Holmes wrote that to achieve longevity one should not only choose one’s parents wisely, but “especially let the mother come from a race in which octogenarians and nonagenarians are very common phenomena.”