DEMENTIA AND ALZHEIMER’S LEADING CAUSE OF DEATH IN ENGLAND AND WALES
Guardian headline, November 2016
AND THERE IT was. Confirmation of what many had feared. The headline reached numerous broadcasts and, thanks to colleagues, appeared numerous times in my inbox, like an alarm I couldn’t switch off. Alzheimer’s had now overtaken heart disease as a leading cause of death in my country and one of its closest neighbours. Citing a report by the Office for National Statistics, the author quoted an increase from 13.4 per cent of all recorded deaths in 2014 to 15.2 per cent in 2015.1
Reading the article, I was torn: as a patient relative, I felt sorrow but also relief: That’s terrible, though maybe now people will do more about it. As a scientist, both emotions yielded to anger: It should not have got this bad! Clearly an effective treatment was now urgent. Of course, not a week goes by when the news doesn’t declare a breakthrough with the spectre of a cure. The irony is that the ceaseless headlines reflect how much we don’t know, rather than how much we do.
People often ask me, ‘What is Alzheimer’s disease?’ I respond with an explanation of plaques and tangles, cell death and memory loss. But by the end of the twentieth century, the real question was where did Alzheimer’s come from? How did it start?
Three remarkable theories emerged and neuroscience adopted a new mantra: Alzheimer’s disease is a process.
As a science obsessed with small changes occurring over long stretches of time, it seems apt that genetics would be the field to usher in this concept. The marriage of disciplines hatched a hybrid called neurogenetics: a special branch of genetics focused solely on the brain. It also ushered in a new armada of neurogeneticists. Among them is John Hardy, the most cited Alzheimer’s researcher in the UK. Hardy has been working on dementia since the days of Kidd and Terry, when only a handful of people on the planet were focusing their research efforts on the disease. An avuncular, plain-speaking man, he has an almost celebrity status among his peers, and is often seen wandering the corridors of his laboratory in University College London in shorts and flip-flops, a stack of papers under his arm, poking his head over the shoulders of young academics, eager to see what the new generation of researchers are up to.
In 1992 Hardy put forward a bold new theory on the cause of Alzheimer’s, one that was to prove so alluring, so self-evident and so impressive in its scope that since then an explosion of work has emanated from it. ‘In all cases of Alzheimer’s disease,’ he told me during my visit to his office, ‘we have amyloid plaques scattered throughout the brain. In all cases we have tangles inside neurons. In all cases we have nerve cell loss. And in all cases we have a dementia. As scientists we have to work out which of those things is first. We have to put an order to them.’
And that’s exactly what he did. Hardy argued that the formation of beta-amyloid plaques in the brain is the primary event in the disease. Tangles, neurotransmitter loss, cell death, memory loss and dementia, he said, are all secondary events–brain flotsam and jetsam left by a harrowing and crippling storm of amyloid. He called his theory the amyloid cascade hypothesis,2 a hypothesis he confidently asserts is ‘no question, the best idea’. He and his supporters are known as the Baptists (from Beta Amyloid Protein)–a highly appropriate name given the fervour of their belief in it.
Though beta-amyloid’s function is still unknown, biochemists agree that it looks like a protein with responsibilities at the cell surface. Cell surface proteins often act like molecular drawbridges for the cell, permitting the entry and exit of other molecules. Alternatively, they can act as molecular antennae for communication with neighbouring cells. If one of these proteins malfunctions–due to a genetic mutation, say–the cell might self-destruct to stop the damage leaking into the cellular circuitry protecting us from cancer. At this level, life is ruthlessly totalitarian.
To play out Hardy’s hypothesis, then, malfunctioning fragments of beta-amyloid first drift away from the neuron and accumulate as plaques. Over time, these plaques grow in size to the point where normal neuronal communication is no longer possible, like islands of waste preventing maritime trade. Starved of biochemical support, conditions inside the neuron start to break down–cue the tangles–and the neuron soon does what evolution has instructed it to do. It kills itself.
Exactly how beta-amyloid triggers such a neural catastrophe is, Hardy admits, a complete mystery. ‘We don’t know. I mean, we really don’t know. And I would say that’s the biggest hole. We just don’t understand how plaques kill neurons.’
The theory had two key advantages. First, it put Alzheimer’s disease on a temporal plane. Framing it this way meant scientists could make testable predictions about the disease’s trajectory and evolution. Second, it gave drug companies another target to complement acetylcholine. The meagre effects of acetylcholine-based drugs created a dire need to try something new and left a gaping (and lucrative) hole for pharmaceutical companies to fill with new, amyloid-based therapeutics.
Families like the Jennings were a source of major support. ‘Because we’d found families which had amyloid mutations,’ Hardy explained, ‘that told us that, in those families anyway, amyloid is where the disease starts. So the simplest thing to assume is that amyloid always starts it; that it’s always the first event.’
Proof of Hardy’s theory arrived when a group of researchers at Athena Neurosciences, a San Francisco-based biotechnology company, did what was long considered impossible. On 9 February 1995 Athena’s scientists injected mouse embryos with a human APP gene mutation.3 The idea of a mouse possessing human DNA is too strange to contemplate. Suffice it to say, the invention broke new ground by providing a means to actually breed the disease ad infinitum.
So did these animals really get Alzheimer’s? They certainly developed plaques in their brain, and showed cognitive impairments in memory-related tasks such as navigating a maze. But strangely, they didn’t show any signs of tangles–nor, for that matter, a great deal of cell death. It was as if they had partial Alzheimer’s. But the fact that Carol Jennings’s mutation–a mutation leading to excessive beta-amyloid production–caused the animals’ downfall provided strong evidence for Hardy’s amyloid cascade hypothesis. It didn’t prove the theory, but it might as well have, his supporters proclaimed. After all, no model is perfect. In an editorial in Nature the same year, Hardy boldly stated that the generation of amyloid plaques in these mice ‘settles this argument, perhaps for good’.4
Working in Hardy’s laboratory is an invigorating experience. In the years following my grandfather’s diagnosis, reading scientific literature had left me awash with ifs, buts and maybes. Nearly everyone I told–family and friends alike–returned looks of soft commiseration: ‘That’s what happens when you get old,’ they’d say over and over again. Working alongside Hardy, I felt grounded in my quest for a better answer.
But not everyone shares Hardy’s conviction that beta-amyloid marks the start of Alzheimer’s disease.
As dusk descended on a cool and clear April evening in 1984, Allen Roses, a neurologist at Duke University, North Carolina, waited anxiously at a railroad crossing, watching the cars of a passing coal train gently rattle by. The train was a rare sight. But of all the days it could appear, on that particular day it was most unwelcome. Standing next to him was a colleague, and in between them, lying on a wheeled stretcher, was an elderly woman with Alzheimer’s, who had been pronounced dead only thirty minutes earlier. They were taking her from the hospital, down a narrow concrete track, to a post-mortem facility less than 300 yards away.
Roses didn’t usually have to run an urban obstacle course as part of his day; his normal routine involved sitting at a lab bench. But that day was different. Roses’ boss wanted him to head a new Alzheimer’s research programme and so Roses submitted a grant application to the National Institute on Aging and was swiftly rejected. If Roses wanted the money, the NIA said, he would have to prove that he could get patient brain tissue from the hospital to the lab in less than one hour.
So as the train sluggishly rolled along the track, Roses and his colleague could do nothing but wait. When it finally passed, the pair dragged the stretcher the rest of the way as fast as they could. As they pushed through the doors to the post-mortem suite, they immediately checked their watches: forty-one minutes. They had done it.
Over the next few years Roses joined the hunt for an Alzheimer’s gene, and by 1990 he had identified a genetic variant for late-onset Alzheimer’s.5 (Variants are not the same as mutations: whereas mutations often directly cause disease, variants simply increase a person’s risk of disease. They are commonly dubbed ‘genetic risk factors’.)
Meanwhile, one of Roses’ colleagues, a neurologist named Warren Strittmatter, was busy trying to wrap his head around a perplexing technical issue with his experiments. Strittmatter, like George Glenner before him, was an amyloid expert. Whenever he purified amyloid from patient brains, however, he kept fishing out another protein stuck to the plaques. It must be a contaminant, he thought. But Roses wasn’t so sure, and asked him to pursue the lead and identify the substance. Four months later Strittmatter discovered that it was apolipoprotein (APOE), a decidedly uninteresting liver protein that carries fat and cholesterol in the blood and can be found throughout the body. It probably had nothing to do with Alzheimer’s.
But in that moment Roses had an epiphany. He knew the APOE gene was located on chromosome 19, the same chromosome in which his latest work had flagged up a new Alzheimer’s gene. Was the connection mere coincidence? Roses didn’t think so. His team, on the other hand, almost certainly did. They refused to do any more experiments on it, convinced, as Roses later put it, that ‘the chief was off on one of his crazy ideas’.6
But Roses kept digging. He learned that the APOE gene exists in three versions–APOE2, APOE3 and APOE4–and became struck by the possibility that one of these versions might increase the risk of late-onset Alzheimer’s. The best way a version could be distinguished from the others was by a burgeoning technology called polymerase chain reaction (PCR). Invented in 1983 by the American biochemist Kary Mullis, PCR is essentially a DNA photocopier. It allows scientists to amplify tiny amounts of DNA for the purposes of paternity testing, forensics and medical diagnostics.
Although PCR is easy to perform, it helps to have an experienced hand. And so, ostracised by his team, Roses turned to his wife, Ann Saunders, a mouse geneticist well versed at PCR. By spring of 1992 the couple had unearthed a truly startling finding. APOE4 carriers have a high risk of developing both early- and late-onset Alzheimer’s–fourfold higher if one copy of the gene is inherited, twelvefold with two copies. The gene is present in 30 per cent of the population and, astonishingly, in 50 per cent of all Alzheimer’s patients, making it the leading genetic risk factor for the disease.7
But how could a liver protein be involved in Alzheimer’s? At scientific conferences, where passions run high, the criticisms launched at Roses ‘went from nasty to vicious’, he recalled. But despite the scepticism, Roses persevered.
On 14 November 1995, at a debate over the motion ‘This house believes beta-amyloid deposition causes Alzheimer’s disease’, he fired back. He presented a slideshow displaying three photographs. The first photograph was of a Japanese Shinto grave, an elaborate and complex work of art; the second was of an old tombstone found in a Catholic cemetery; the third was his father’s grave, a bronze plaque in a Jewish cemetery. He pointed at them and said, ‘Every one of these is absolutely diagnostic of what’s underneath it. But nobody would say that the tombstone caused the death.’
But the analogy did little to dissuade the disbelievers. His intention, though, was not so much to dispel interest in beta-amyloid as it was to receive acknowledgement of APOE4. ‘I have no doubt that there is plaque formation,’ he assured me. ‘I just don’t think it’s the cause. But everybody in the Alzheimer’s community thought APOE4 was a big joke. They just didn’t want to hear about it. I couldn’t even get another grant to pursue it.’
Listening to Roses, I felt both sympathetic and inured to his plight. Scientists are not the paragons of mutual camaraderie we might imagine them to be–all hell-bent on uniting under one banner to seek the truth. They are human. Big intellects bring big egos, which partly explains why Roses was dismissed by so many. Although a string of European studies soon confirmed his discovery, by 1997 the Alzheimer’s community was firmly focused on the amyloid cascade hypothesis. Unable to fund further work on APOE4, Roses was forced to leave academia and pursue the lead in the pharmaceutical industry. There, he pioneered a new theory on the cause of Alzheimer’s, one that put APOE4 in the spotlight and which came to be called the mitochondrial-impairment hypothesis. Or more simply, ‘type three diabetes’.
In type one diabetes, problems arise when the death of insulin-producing cells in the pancreas depletes the body of insulin. Type two diabetes, on the other hand, results from insulin resistance when cells stop responding to it following an excessive dietary intake of glucose (although genetic and lifestyle influences are also thought to be involved). In type three diabetes, the theory goes, the APOE4 gene somehow interferes with normal blood sugar uptake in the brain, thereby depriving the brain of the energy it needs to fuel cellular activities. Proof for it came in late 2000, when two psychiatrists–Eric Reiman at the University of Arizona, Tucson, and Gary Small at the University of California, Los Angeles–used brain imaging to show that people who have the APOE4 gene metabolise glucose at lower rates than people carrying the APOE2 or APOE3 versions.8 Roses’ followers don’t have a nickname, but let’s call them the ‘E4ists’.
So, let’s play out Roses’ hypothesis. Over time, a brain starved of energy causes the functions of its neurons to decay, and like a city undergoing economic collapse, its service sector, or proteins, will start to malfunction–hence both the plaques and the tangles–until eventually the neuron gives up and self-destructs. But unlike other cell types, neurons can’t replace themselves. So with each loss the energy burden on neighbouring neurons increases. And as an organ that constitutes 2 per cent of the body’s weight yet requires 25 per cent of its energy, this is bad news for the brain. Unable to cope, more neurons self-destruct, a pathological cascade is initiated, and Alzheimer’s takes hold.
For the goal of treating Alzheimer’s effectively, both theories had their merits and pitfalls. Hardy’s theory drew a straight line between amyloid and the disease, offering pharmaceutical tycoons an easy target. But it also imposed a narrative that was arguably too simplistic. And as ‘seductive as this narrative might be,’ one critic recently wrote, ‘the dementing illness that we recognise as Alzheimer’s disease is associated with a complex biology and biochemistry, as well as a pattern of brain disintegration that cannot easily be explained by a simple linear disease model.’9
Holding APOE4 responsible, on the other hand, would help plumb the genetic depths of old-age dementia, and perhaps help explain why amyloid formed plaques in the first place. But a gene that merely increased the chances of disease left drug discovery empty-handed. As one pharmaceutical magnate told me, ‘Yes, APOE4 is by far the most prominent genetic link. But it doesn’t mean anything. You can’t offer someone who is APOE4-positive any therapy.’10
But there’s a third option: what if neither theory could explain the disorder?
The fundamental precept of good science is that there is no place for beliefs. It’s why scientists are repeatedly taught to replace ‘believe’ with ‘think’. Enter our third group: the Tauists.
Tau stands for ‘tubulin-associated unit’. It’s the name of the protein that forms Alzheimer’s tangles, the twisted knots of debris that seem to strangle neurons from within (the ‘paired helical filaments’ that Kidd and Terry spent years arguing about). Discovered in 1986 by three separate groups of researchers, tau normally acts as a kind of sealant for ropelike structures called microtubules, which stretch out along axons to create an internal transport system for every neuron. Scientists discovered that tau becomes tangled when it mixes with too much phosphorus inside the neuron. This tangled, hyperphosphorylated tau causes microtubules to fall apart. And that fact led the Tauists to assert their own hypothesis on the cause of Alzheimer’s, which has come to be called, fittingly, the tau hypothesis.
Picture a zip-line whizzing sacks of crops between high-altitude villages in the tropics (farmers actually do this in Bolivia and other parts of South America). If the line comes apart, the crops won’t reach their destination. For the neuron, those crops are neuro-transmitters and biochemical nutrients, whizzing along the axon en route to synapses and other neurons. The failed deliveries have lethal consequences. Millions of synapses, the wellsprings of memory, will collapse and vanish. Then the axon itself begins to deteriorate, dying back until nothing but a limbless cell body remains. With all lines of transport and communication effectively terminated, internal chaos ensues and cell death becomes inevitable. When the neuron finally dies, all that’s left are eerie coils of tau–what neuropathologists call ‘ghost tangles’.
For Tauists, then, Hardy and Roses are both wrong. What’s more, they have misunderstood the nature of Alzheimer’s. There is no ‘primary event’ in the disease. Amyloid and APOE4 are just two triggers and there are probably many more. The crucial point was this: whatever triggered the disease, it all converged on a common end point–tau. It represented the ‘how’ behind neuronal death, when all else only represented the ‘why’. It was, in short, all that truly mattered.
The idea had hefty support. Alois Alzheimer himself would have been more Tauist than Baptist. In 1911 he wrote: ‘We have to conclude that the plaques are not the cause of [old-age] dementia but only an accompanying feature.’11 And by the mid-1990s it had emerged that there are more than twenty brain diseases caused solely by tau malfunction–known collectively as tauopathies. If that wasn’t enough to give tau the respect it deserved, argued the Tauists, nothing was.
At scientific conferences at the time the debate was raging. But the Baptists stood their ground. ‘Tangles aren’t particularly important,’ Hardy told the New York Times’s Gina Kolata in 1995. He remains convinced that tangles are merely the result of plaques and therefore less worthy of our attention. It is inescapable that we should wonder how all this infighting actually helped patients. But such quarrels were and still are compulsory, because there are many theoretical paths for developing an effective drug. If we ceased thrashing out the best possible path, then the solution might take even longer to arrive.
I myself lean towards the amyloid cascade hypothesis, for the simple reason that amyloid appears to be where Alzheimer’s begins. But I also think it unwise to assume that this fact somehow diminishes the relevance of APOE4 and tau tangles, because our understanding of causation remains unsophisticated. We still don’t know, for instance, whether biological causes produce their effects by guaranteeing them: the inescapable fact that many people develop plaques yet remain cognitively healthy exemplifies that point. Thus, the cause of Alzheimer’s is unlikely to be one single thing.
The most conspicuous flaw in the Tauists’ argument was a lack of genetic evidence. Unlike amyloid, no mutations in tau were found in Alzheimer’s patients, and without genetics as a guiding light, the idea of building a case against tau was like telling detectives to find a murderer by asking random people on the street. It was a messy lead, in other words. And at the start of a new century, the Baptists kept the upper hand, while the Tauists fast became renegades.
Meanwhile, steadily ticking over in the background, the world’s largest, most ambitious biology project was under way: the Human Genome Project. Initiated in 1990 by the US Congress, the goal was to map every gene that made up a human being. Costing $3 billion and involving scientists in some twenty different countries, it was the biggest collaboration of biologists the world has ever seen.
On 14 April 2003, when the final draft was unveiled, it was heralded as the most valuable information humanity has ever known: ‘More significant than splitting the atom or going to the moon,’ declared Francis Collins, the project’s US lead scientist; ‘The most wondrous map ever produced by human kind,’ announced US President Bill Clinton; ‘The foundation of biology for decades, centuries or millennia to come,’ said the UK lead John Sulston, who was to win a Nobel Prize for his work.
For Alzheimer’s research it was game-changing. As the technology improved, thousands of patients rallied to have their genome sequenced, giving rise to genome-wide association studies (GWAS), in which small, previously hidden genetic variants could be uncovered. To date, more than twenty genetic variants have been identified–and the list will grow.