How important are our differences in compatibility genes? The brief answer is that these genes influence if and how fast you recover from almost any disease you may get. This chapter will explain this fully, but first, to gain perspective, it helps to consider the impact of our genetic inheritance on our well-being in general.
There’s no clearer example of how dramatic an effect genes can have on one’s health and behaviour than in the life of Woody Guthrie, the American folk singer, left-wing observer of the Great Depression, and inspiration to Bob Dylan and Bruce Springsteen. Guthrie played a guitar scrawled with the words ‘This machine kills fascists’, reputedly wrote over a thousand songs and published an intoxicating autobiography, Bound for Glory, an ode to living free, riding freight trains and personal rebelliousness.1 But Guthrie suffered gravely from a genetic mutation; his daughter Nora never knew him as healthy.2 While he was in his thirties, when he should have been at the peak of success, his behaviour became erratic and even violent.
On the evening of 15 May 1952 he attacked his wife, Nora’s mother, Marjorie. She had come into the house to find him strangely glazed over and holding scissors. He’d cut the telephone wire; the attack was premeditated. Marjorie got upstairs, onto the bed, but he followed and began pummelling her with his fists. Something was very strange, and he seemed to be frothing at the mouth. This attack was crazier than ever before, and Marjorie distinctly noticed that he was sober. She escaped to get the neighbours’ attention, and the police arrived not a moment too soon. When an officer said he knew Guthrie’s songs, things quietened. Marjorie, having just experienced hell, told the truth: ‘Woody, you’re sick . . . I don’t know what it is and you don’t know what it is, but you’re sick.’3
The next day Guthrie checked into hospital on a three-week programme for alcoholics and, once out of hospital, he wrote to Marjorie to try to convince her that it was just liquor that turned him into a senseless raving idiot.4 Finally, on 3 September 1952, a neurologist realized that Guthrie had Huntington’s disease. Until that point, his tics, shakes and slurred speech had just been taken as eccentric mannerisms of a great artist. Lost in illness, one day in May 1956 he was arrested for not having a bus ticket, and police took him to Greystone Psychiatric Hospital, New Jersey, where he would stay for many years. Bob Dylan visited him there to bring cigarettes and play songs – first in January 19615 – and remembers how ‘it was a strange environment to meet anybody, least of all the true voice of the American spirit . . . The experience was sobering and psychologically draining.’6
After Guthrie’s death – on 3 October 1967, when he was fifty-five – Marjorie took out an ad in the New York Times to bring afflicted families together. She persuaded US President Jimmy Carter to set up a commission for studying neurological diseases. This led to the founding of the Huntington’s Disease Society of America, which raised money for the research that led to the first genetic markers of the disease. Marjorie died in 1983 but the Woody Guthrie Folk Festival, held in Oklahoma each July, raises money for more research. It includes an annual pancake breakfast run by Woody’s younger sister Mary Jo, who by chance didn’t inherit her brother’s illness.7 Eventually, in 1993, the single gene that underlies Huntington’s disease was discovered.8
The normal version of this gene has a section in which a small stretch is repetitive. In the disease-causing mutant, this repeated part of the gene is longer than normal, so cells then produce a version of the protein that’s bigger than normal, and this leads to cognitive problems and dementia – although exactly why this is remains unclear. Speech and many higher brain functions commonly deteriorate, although long-term memory is spared, not necessarily making life any happier. As many as one in four patients attempt suicide; depression is common.9
Because a single copy of the mutant gene is enough to cause the disease, then, if one parent has it, their child simply has a 50 per cent chance of inheriting it. Children often witness one of their parents suffering, knowing there’s a chance their fate is similar. Some choose to know for certain by taking a genetic test; some don’t. The question – to know or not to know – eats away at many for much of their lives. You can only gain certain knowledge whether you will inevitably get the disease or not; there’s no cure.
Cystic fibrosis is another well-known genetic disease, one of the most common potentially lethal genetic disorders for European Caucasians. The difference from Huntington’s is that for cystic fibrosis people can carry one dysfunctional copy of the gene without any symptoms because they still have another, normal, version that compensates. Disease arises only when a person inherits two copies of the impaired gene. That means that, if each parent has a copy of the mutant gene, there’s a one in four chance their child will inherit cystic fibrosis. Inheriting two aberrant copies of the cystic fibrosis gene leaves people unable to correctly make a particular protein which normally helps salt move in and out of cells. Loss or inefficiency of this protein has an effect on many organs which leads to a range of symptoms such as difficulty breathing following a lung infection. Shortly before he became the UK’s prime minister in 2007, Gordon Brown’s new-born son Fraser was diagnosed with cystic fibrosis. Brown rarely discusses it in public, but in one interview he said what anyone would think at some point in the face of such difficulty: ‘We sometimes say: well, why, why, why, why us? You know, why did this happen to us?’10 Science can’t give an answer as to why any one particular person is affected – but one of the great triumphs of human biology is that we do have an explanation as to why some disease-causing mutations are kept in the population at all.
Sickle cell anaemia is a powerful example where this is clear. This disease comes from inheriting a mutant version of haemoglobin – a protein in red blood cells that binds oxygen to release it where appropriate around the body. A change in one of the haemoglobin genes gives rise to an altered protein, haemoglobin S. The presence of haemoglobin S changes red blood cells from their normal disc-like shape into an abnormally curved ‘sickle’ shape, which makes them fragile and interferes with their circulation through small blood vessels. Those who inherit one copy of haemoglobin S produce both normal and abnormal haemoglobin. This doesn’t cause a life-threatening problem, as there is still enough normal haemoglobin to transport oxygen around the body. People with one copy of this gene simply have to take care when oxygen might be in short supply, at high altitude for example. However, an inheritance of two copies leads to sickle cell anaemia, which can cause damage to several organs and early death. Despite this danger, the mutation turns out to be surprisingly common in sub-Saharan Africa. For example, in Nigeria about one in four people carry a copy of the haemoglobin S gene. There are tribal variations – 45 per cent of people in the Baamba tribe in the west of Uganda, for example, carry it.11 Why is this potentially lethal mutation found in so many people?
The first clue to answering this came from mapping where people have haemoglobin S; the area where people have this gene has a striking resemblance to the distribution of malaria. This fits with haemoglobin S providing protection against malaria. Even within the same geographical region of Africa, haemoglobin S is common in people living in lowlands but relatively rare in people indigenous to the highlands, where malaria is not endemic, because mosquitoes don’t live that high up. In fact, we now know that the sickle cell trait caused by haemoglobin S can be 90 per cent protective against severe and complicated forms of malaria, and that’s why the mutant gene is retained in the population.
So this specific genetic mutation protects us against one major disease, malaria, at a cost of some people inheriting two copies and getting sickle cell anaemia. That is, there’s one reason to keep the gene – it helps against malaria – and one reason to get rid of it – as it makes people susceptible to anaemia. So it’s kept in the population at a frequency dependent on balancing the level of threat from these two diseases. Where malaria is endemic, such as in West African lowlands, the mutation is retained, because having it makes you more likely to have and raise healthy children. In populations living without malaria, such as in the African highlands, the mutation is very rare. In short, the outcome is human diversity.
So just how diverse are all the genes we each have? Roughly speaking, humans are 99.9 per cent the same genetically – just 0.1 per cent of genes varies from person to person. You might guess that the most variable genes among us would be those that influence hair type, eye colour, or skin pigmentation. But, in fact, the genes that vary most between people don’t affect anything superficially obvious at all. Those that vary the most are our immune-system genes – especially our compatibility genes. Why are these genes so enormously diverse – and why does it matter which ones you have?
The early pioneers of human compatibility genes – Dausset, van Rood, Payne and the others we met in Chapter 3 – wondered whether or not our variation in these genes linked with disease. They were encouraged by studies in animals, where data was easier to interpret because of inbred strains of mice – one influential study, for example, reported in 1964 that different strains of mice varied in their susceptibility to a type of leukaemia caused by a virus.12 In humans, the first link between HLA and disease was established for a type of white blood cell cancer, Hodgkin’s lymphoma, in 1972. But it is only a small effect; people with Hodgkin’s lymphoma have a slightly increased chance of having certain HLA genes.13
Unlike the genes that underlie Huntington’s, cystic fibrosis or sickle cell anaemia, HLA is not an all-or-nothing mark of disease. Single genes that underlie disease are very rare; most illnesses are complex and involve both genetic and environmental factors. Virtually all cancers, for example, have been linked to specific genetic variants but having certain mutations doesn’t usually make cancer inevitable. Your chance of getting the associated cancer can be increased by a particular genetic inheritance, but usually cancer still remains a small risk overall. Likewise, the compatibility genes you have inherited make you more susceptible to some diseases and more resistant to others – but usually to a small extent.
This in itself causes a problem in testing for connections between HLA and disease. Because HLA is so diverse – and its effects are subtle – one version of an HLA gene can easily show up just by chance as being unusually frequent in a particular set of people with a specific disease. That is, any single study can quite easily give a false indication that one HLA type is associated with a particular disease or trait.14 This needs to be taken into account by using a statistical analysis more complex than used in everyday lab experiments. Another problem is that any small effects detected could be due to unknown genes nearby or connected to HLA genes, rather than HLA itself.
That’s why, historically, consensus about which HLA types genuinely associate with a disease has been hard to reach. The early pioneers struggled to convince everyone about the importance of HLA in disease. Eventually, in 1973, a clear link between HLA and disease was established by two groups – one in the US and one in the UK – who reported their discoveries almost at the same time. Although other links between HLA and disease had been indicated, this one was dramatic.
In the US, Rodney Bluestone, Paul Terasaki and their colleagues in the University of California Los Angeles (UCLA) discovered that 88 per cent of people with ankylosing spondylitis had HLA-B*27, whereas its frequency in the general population was just 8 per cent.15 Ankylosing spondylitis is an autoimmune disease that leads to inflammation within joints, especially in the spine. It had been studied for over a century but little progress had been made in deciphering its causes. Strangely, the UCLA team’s breakthrough finding was rejected for publication by the top medical journal the Lancet.16 Instead, they published their work in the New England Journal of Medicine, only to then see that another group, led by Derrick Brewerton at Westminster Hospital, London, had the same conclusion published in the Lancet – the very same journal which had just rejected their paper.17 Why the Lancet rejected one study and published another with the same conclusion is unfortunately not known. Terasaki has suggested that anyway he was first in the discovery, evidenced by his team’s abstract from a conference in November 1972 being the first publication of the finding.18 In contrast, Brewerton regards the outcome of this race as a dead heat.19 For science, the real value of having the result confirmed in two independent studies was that it immediately gave everyone confidence in the conclusion; together, these two studies unequivocally established for everyone that HLA does influence disease.
The UCLA team had been fully engaged in HLA research for over a decade – Terasaki was one of the HLA pioneers involved in the early international meetings, and he developed a widely used method for testing HLA types. Brewerton’s paper, on the other hand, seemed to come out of the blue; he was a clinical rheumatologist, entirely unknown to HLA scientists. His only previous contribution to immunology was indirect: he had helped treat Peter Medawar after his stroke and was the target of Medawar’s anger when told that he should learn to write with his left hand as his right side may not recover.20
Brewerton’s contribution to the story of our compatibility genes began one summer’s day in 1971, when he happened to sit down for lunch opposite the head of his hospital’s tissue-typing facility, who had just returned from a conference where there was a lot of discussion about the possibility that HLA could associate with disease. From that discussion Brewerton realized that he should test the HLA types of patients with ankylosing spondylitis, because it was known that an inherited factor was important in this disease, and nobody knew what it was. A grant application on the idea was rejected, but some hospital funds became available to employ someone to help in tissue-typing, which gave them the opportunity for research around their clinical duties. ‘As a result,’ Brewerton remembers, ‘I often started work in my office at 4am and stopped only when the library closed at night. Family life went on hold.’21
Why was this so pioneering? Because it was a time when thinking about genes didn’t permeate our culture as deeply as it does now. So, although it was clear that certain diseases congregated in families, many clinicians didn’t think it a top priority to find out which genes associate with arthritic diseases. Brewerton tested eight patients and found that all of them had HLA-B*27, a probability of happening by chance of less than one in a million. He felt that he was on to the most important discovery ever made in understanding arthritis.22
Brewerton decided that he next needed to plan a military-style operation to test a hundred patients and several control groups. It was before computers were widely used, so there weren’t databases of patients to call upon. He recruited patients by asking, across thirty-seven hospitals, if doctors could recall anyone appropriate for his study – ‘all done in a very primitive way’, he later recalled.23 The results were astonishing – 24 out of 25 patients had HLA-B*27, then 48 out of 50. Brewerton was elated, but he was also very afraid; it just seemed too good to be true.24
Losing sleep at night, he kept thinking of problems: what if something unknown, a new virus perhaps, caused the disease, and this new thing somehow interfered with the HLA test to give a false reading of B*27 being present? Or what if some of the drugs used for treatment somehow influenced the HLA tests to give false readings? In hindsight, these worries seem overly cautious; theoretically feasible but beyond reasonable doubt. Yet the anxiety is almost inevitable for any conscientious scientist going over every possible alternative interpretation of their data before being sure their discovery is right. Eventually, Brewerton found a way out.
He decided to test the HLA type of each patient’s close relatives – because some of these relatives should also have inherited B*27 even when they don’t have the disease. That way he would know his tests were giving the correct HLA type, and not a result of something strange in people with ankylosing spondylitis. Brewerton travelled all over London, visiting families to collect blood samples, usually very early each morning.25 It took a few unnerving months, but he finally confirmed that family members also had B*27, even when they did not show signs of illness. Brewerton’s fog of fear cleared, and he accepted that his team had proven that someone who had inherited HLA-B*27 was about 300 times more likely to develop ankylosing spondylitis. After publishing, Brewerton learned of the UCLA results, and his reaction was disappointment that he had not made the discovery alone.26
The US team – Bluestone and Terasaki – had arrived at the same discovery by a very different route. Born to a poor immigrant family in Los Angeles in 1929, Terasaki spent most of his career at UCLA, where, in 1964, he established a test for HLA types that could use very small samples of blood. Over the following few years, he and his team automated this method so that hundreds of HLA tests could be done in a day. This enabled him to set up an ambitious programme to study huge numbers of patients, searching for unusual frequencies of HLA types in all kinds of diseases. One of the diseases he tested, with clinician Rodney Bluestone, was gout.
For something to compare with gout, they decided to use patients with another rheumatic disease, ankylosing spondylitis. It was these control patients that turned out to be more interesting than they planned – not a blank control at all – and the link between B*27 and ankylosing spondylitis was revealed. Serendipity played a big role for both teams: Terasaki’s team chose interesting patients as a comparison to the disease they actually set out to study, and Brewerton had sat down to lunch in just the right place.
From here, Terasaki’s and Brewerton’s careers couldn’t have diverged any further. Brewerton remained focused on clinical duties, wrote a book about arthritis,27 and later chaired a local campaign group set up to protect a small beach near his Sussex home.28 Terasaki, on the other hand, founded, in 1984, a company called One Lambda to sell the tools needed for tissue-typing; as a result, in 2010, he could donate $50 million to UCLA. When I met Brewerton at age eighty-seven, in 2011, he said that he no longer thinks about his work on HLA-B*27, that it all seems very distant – like it happened to someone else.29
The fact that so many people with this disease have B*27 is striking, but it’s also important to realize that if you have B*27 you are still very unlikely to get the disease. Other genes and environmental factors are also important, so that, overall, it’s not very useful to have a genetic test for B*27, because it’s not strongly predictive of disease on its own. It can perhaps have a role in diagnosing ankylosing spondylitis if symptoms are mild or ambiguous, but the main impact of the B*27 discovery has been scientific; a stepping stone to understanding our HLA system.
Later, it was discovered that people with HLA-B*27 are also more susceptible to the skin disease psoriasis and an inflammation of the eye known as uveitis. This meant that the same compatibility gene could make someone susceptible to a range of diseases that are clinically very different. Then, around thirty-five years after the ankylosing spondylitis story broke, B*27 hit the limelight again. A monumental project involving over 200 different research centres found that B*27 was linked to Acquired Immunodeficiency Syndrome (AIDS) – but this time it protected against the illness.
People infected with HIV don’t progress to AIDS at the same rate. The general course of events following infection is that the virus multiplies dramatically in the first few weeks and ’flu-like symptoms develop. By about four weeks, the immune system starts to lower the amount of virus in the blood, until it reaches a steady level about two months later. The number of copies of the virus in a person’s blood then stays level over a period called the chronic phase of infection, and during this time patients don’t seem ill at all, although they can still infect others. Later, virus numbers increase and AIDS develops – but how long this takes varies considerably. Some succumb quickly, whereas others infected with the virus can stay free of the disease for many years.
A fortunate few – about 1 in 10 – stay disease-free for a very long time – more than seven years. They are known as the Long Term Non-Progressors. But about 1 in 300 is even luckier than that. Their immune system can attack the virus to such an extent that it becomes almost entirely undetectable. They are the HIV Controllers or the Elite Controllers. In the battle against AIDS, these few individuals have superpowers endowed by their genes.
HLA genes were first linked with resistance to AIDS in 1996 – although this first study involved a relatively small number of patients.30 HLA-B*57 was found to occur far more frequently in patients that do not rapidly progress to AIDS.31 But one criticism of this research in the 1990s – as with all research into HLA and disease at that time – was that these data didn’t establish how important the link to HLA was in comparison to other genes. Nowadays, all of the genes can be compared across many individuals with or without a particular trait or disease. This approach – called a genome-wide association study, or GWAS – has found genetic variations that affect a range of human traits, including our height, body mass index and blood lipid levels. A first GWAS for HIV indicated that regions of the genome at or near HLA were associated with lower levels of virus in a patient’s blood.32 Then the International HIV Controllers Study – a study led by Bruce Walker at Massachusetts General Hospital and Paul de Bakker at Harvard University – aimed to pin down exactly which genes were important.
Walker had been building up to this moment for twenty-five years, having first submitted an application to study our immune response to HIV in 1985. At that time, his application was rejected because – as the committee told him – HIV suppresses our immune response, so what could be the point in studying our immune response to it? But, after being inspired by reading Alain Townsend’s research – the work we discussed in Chapter 4, that HLA proteins present bits of proteins or peptides from viruses – Walker applied for funding again. His application was rejected a second time, along with the clarification: Dr Walker, you really don’t get it – this is an immune suppressive illness and we told you before.33
Two decades later, however, one couple’s donation of $100 million freed Walker to pursue his dreams.34 He had recently accompanied philanthropist Terry Ragon to Africa to visit AIDS patients with a view to raising money for HIV research, the kind of trip he’d done several times before. In 1978, Ragon had founded a company to provide database software which became the source of his fortune. After their trip to Africa together, Walker mentioned that he thought progress was hampered by research being fragmented in silos: that teams studying different aspects of HIV didn’t mix. Ragon reckoned out loud that an institute to bring everyone together might require $10 million a year over ten years, and then casually added: and my wife and I would like to give that to you.35 The remark triggered an out-of-body experience for Walker.
He knew that this kind of flexible money would be entirely transformative. But his elation was quickly followed by sheer terror – with such a huge amount of money comes great responsibility.36 In fact, Walker proved adept at raising money and he obtained several charitable donations – in effect establishing his own funding agency. One of the first studies to come from the Ragon Institute, with its many collaborating centres, was the HIV Controllers Study.37
In this huge study, 3,622 HIV-infected people of different ethnicities were divided up according to whether or not they were HIV Controllers – that is, according to whether or not they had maintained very low levels of virus in their blood after being infected, on average, ten years earlier. Their genomes were scanned, and the only statistically strong characteristic of HIV Controllers was in their compatibility genes. Carefully analysing that region of the genome revealed precise details; that having HLA-B*57, -27 or -14 is protective against AIDS, whereas having HLA-B*35 or HLA-Cw*07 associates with fast progression to AIDS.38
There were earlier studies showing that HLA genes influence the progression to AIDS,39 including the identification of B*57 as being important,40 but the scale of this study was new. By looking across the whole human genome, this study clarified that our HLA genes have the biggest influence in the human genome. Indeed, only the variations in our compatibility genes were significant when the whole genome was assessed.41 This establishes that the very same genes that control the success of, say, a kidney transplant, influence how long we live after infection with HIV – two aspects of human biology that, at a glance, seem entirely unrelated.
While it might be a stretch to consider these genes as endowing superpowers, on the other hand, what is a superpower if it’s not a rare ability to survive something that destroys others unfailingly? Genetic variation can’t endow anyone an ability to fly like in the X-men comic books, but a real-life genetic superpower is that some of us can survive a deadly virus. How is this possible?
The answer lies in the details of how our HLA proteins work. Recall Townsend’s experiments, the ones initially doubted by Zinkernagel. They showed that HLA proteins present samples of what’s being made inside our cells – peptides – that can be scrutinized by our immune cells – T cells. The picture of the HLA protein – hard-won by Bjorkman, Strominger and Wiley – showed that peptides were clasped in a groove at the top of the HLA protein. In fact, there are about 100,000 of these HLA proteins on a cell’s surface, so collectively they present a good sampling of what’s currently being made inside. And detection of diseased cells occurs because a T cell can react to a peptide present in the groove of an HLA protein that hasn’t been in your body before.
Bjorkman’s pictures of HLA-A*02 – and the pictures of other HLA proteins that other teams obtained later – revealed another killer fact. Our differences in compatibility genes encode for slight variations in the HLA proteins we have – and pictures of HLA proteins showed that our differences are not positioned randomly throughout the protein’s shape.42 The bits of, say, HLA-A*02 and HLA-B*27 that are different are at the top of the molecule – in and around the groove where the peptides sit. This is a revelatory finding: it means that each type of HLA gene makes a protein with a slightly different-shaped groove on top.
Why is this so critical? Because this means that each type of HLA protein is best at clasping certain peptides – not all the ones that a cell might have. Put another way, each type of HLA protein presents a different sampling of what’s being made inside a cell. And, most important of all, this means that for any particular peptide – say one that’s produced in large amounts by a specific virus – only some HLA types of all those present in the population will be good at clasping it. So each person is better or worse at detecting any one particular peptide – according to which type of HLA proteins they have inherited.
HLA proteins that can’t hold on to one particular peptide will have the right-shaped groove for others – perhaps one from another virus or an alternative peptide made by the same virus. Any one particular virus will encode many peptides, each of which can be used by many types of HLA proteins, but another factor to consider here is that certain combinations of peptides and HLA protein are particularly good at triggering a T-cell response. For example, a particular peptide may bind especially well to one type of HLA protein, which in turn strongly activates the appropriate T cells. The upshot of all this is that some of us will be inherently better than others at defending against a particular infection, such as one type of virus.
Back to our battle with HIV: this virus is especially hard for the immune system to attack, mainly because there’s actually an enormous variation in the virus itself. The number of different versions of HIV in one infected person is greater than the variety in influenza virus seen worldwide in any one season.43 When one version of HIV causes an infection, it takes only a few days for there to be immense variety – because, as the virus multiplies, it makes copies that are each slightly different. Variation in the virus helps it escape any particular attack from human T cells.
Each T cell, as we’ve seen, detects the virus by recognizing a particular peptide that it makes. So that means that each T cell will attack all forms of the virus that make a particular peptide – but other versions of the virus survive. As one HIV peptide provokes an attack by T cells, the pool of virus in that patient will be replaced by versions of the virus that don’t make that particular peptide. In other words, the virus evolves inside us – changing its parts – to avoid attacks by our immune system. And that’s why it’s really hard for us to get rid of it.
In that case, why are some types of HLA proteins better than others in fighting HIV? Our HLA proteins grasp different peptides – chopped-up pieces of protein made by the HIV virus – but, crucially, some of these peptides are better than others for our immune system to detect. There are two reasons for this. The first is that it helps if the peptide comes from a protein that the virus makes a lot of. That way, there’s an abundance of it available for T cells to see, triggering a strong reaction. The second reason is subtle – but actually more important.
Although the virus can change and produce many different versions, some parts of the virus are just so critical that they can’t be altered: changing these parts would render the virus unable to work properly. This means that there are some bits of the virus that are the same across all the variants of HIV. So the second thing about a peptide that makes it a really good target for our immune system is if it is made from one of these conserved parts of the virus. If the immune system can detect one of these peptides, there aren’t versions of the virus that can escape detection.
So the HLA types that are best at fighting a virus are those that can hold peptides made in large amounts from conserved parts of the virus. It turns out that HLA-B*57 is good at binding a particular peptide from a conserved part of an abundant HIV protein (called Gag). And so it makes sense that HLA-B*57 is found in around 30–50 per cent of HIV Controllers, five to ten times more frequently than it’s found generally.
A warning: this is research in progress. And presentation of particular peptides by B*57, for example, may not be the sole way in which HIV Controllers avoid AIDS but it is (almost certainly) a major factor. B*27 is also one of the HLA types frequent in HIV Controllers, and yet this same genetic variation is part of the problem in ankylosing spondylitis – discovered by Brewerton and Terasaki. That is, the very same HLA that is good for us in dealing with one disease, HIV, is bad for us by contributing to another disease, ankylosing spondylitis. The role that B*27 has in the auto-immune disease ankylosing spondylitis is (almost certainly) also a result of its job of presenting peptides to T cells. But in this case, normal self-peptides – chopped-up pieces of protein found in healthy cells – are wrongly identified as non-self and trigger an immune response. This gives rise to the auto-immune disease – because our immune system makes a mistake and attacks healthy tissue.
But why would one particular type of HLA gene underlie this kind of disease? The short answer is that nobody really knows. It could be that B*27 binds an abundant self-peptide extremely well and this leads to problems. Or alternatively, there could be some real non-self peptide, say from a virus, presented by B*27 to activate T cells, which then accidentally attack uninfected cells by responding to a self-peptide that just happens to be similar to the one from the virus.44 Even though we don’t understand exactly how B*27 is involved in ankylosing spondylitis, there’s plenty of evidence that, in general, T cells and HLA genes are important in many auto-immune diseases. Class II compatibility genes designated HLA-DR*03 and -DR*04 are found in the vast majority of people with type I diabetes, for example. In fact, virtually the whole gamut of possible illnesses that could ever affect us are known to be influenced by our compatibility genes – including cancer, infections, auto-immune diseases and even some neurological disorders. HLA-B*53 can protect against severe malaria, for example, by binding a particular peptide made by the parasite.45 And other compatibility genes are linked to multiple sclerosis, Parkinson’s, Hodgkin’s lymphoma, inflammatory bowel disease, leprosy, narcolepsy and so on.
By controlling our response to disease, our compatibility genes influence how we live, when we die – and from what we die. Even so, it’s important to keep in mind that, even though many Elite Controllers of HIV have HLA-B*27, having this type of HLA is not sufficient to survive AIDS. There is still much to be learned about our immune response to HIV and other infections. And, whatever HLA type you have inherited, by far the best way to stay safe is to avoid being infected in the first place.
The compatibility gene system doesn’t only explain how any one of us might fare better or worse against a particular disease like AIDS. It is a far bigger system than that. It also works between us all, across the whole human species, because our immune system has evolved in defence of humanity as a whole – to protect all of us as a species from anything dangerous that could arise.
As we’ve seen, these genes vary from person to person, and across the human population there are billions of combinations of HLA types possible. In fact, it’s theoretically possible for everyone on the planet to have a different set of HLA genes. This doesn’t quite happen because some HLA types are more common than others, but there is still enormous diversity in the population. This means that a virus escaping detection by the HLA proteins in one individual will face different HLA types in someone else. If we were all infected with one especially deadly virus, for example, some of us may survive by having a particularly potent HLA type for dealing with that virus. An outcome of this is that the distribution of HLA types among us evolves over time, as waves of different infections influence who reproduces.
But if some HLA genes are better at fighting a particular infection – because they can pick up appropriate peptides – then why do we each have only six class I HLA genes, for example? Why don’t we each have hundreds or thousands of HLAs like the whole population does? If a virus has managed to avoid detection via one HLA type, it’s useful to have another that may work. So hundreds or thousands of HLA genes in each person would surely be better at catching all possible infections, wouldn’t they?
It’s hard to do an experiment to answer this, but the generally accepted theory is that the system has a limit because of the way our body discriminates self and non-self. Recall that any T cell capable of reacting to ourselves is killed off in the thymus. That implies that, for each HLA type, all the T cells reacting to self peptides clasped by that type of HLA protein have to be killed off. Too many HLA types and it would be hard to have a big enough pool of T cells left. So there is a balance between maximizing the number of non-self peptides that can be grabbed by HLA, while still allowing enough T cells to exist for the detection of all possible non-self peptides. Or, in broader terms, there is delicate balance in making sure the immune system cannot attack our own bodies yet can respond to all kinds of potential infections. The outcome of this balancing act is that our HLA type makes each of us more susceptible or resistant to different diseases – not like the individual genetic mutations which cause cystic fibrosis or Huntington’s; compatibility genes influence our response to all kinds of diseases.
From this depth of understanding, the next question is pragmatic: how do we get to new cures?