11.7
intelligentsia
The Intellectual
"Intellectual frontier scholar", senior researcher of Shenzhen Bay Laboratory Wang Liming
Introduction
In July, an investigative report in the journal Science revealed that an important paper in the field of Alzheimer's disease research with more than 2,000 citations was suspected of being fraudulent. This survey does not negate the accumulation of scientific research on the pathogenesis of Alzheimer's disease, but it also suggests that it is time to rethink Alzheimer's research.
Wang Liming, a "scholar at the forefront of intellectual knowledge" and senior researcher at Shenzhen Bay Laboratory, reviewed important research and clinical trials on Alzheimer's disease in this issue of "Mountain Patrol Report • Life Science", and pointed out that Alzheimer's disease research has reached a critical point of paradigm shift.
Written by | Wang Liming
● ● ●
Hello, I'm Wang Liming. On November 6, 2022, the 45th issue of the "Mountain Patrol Report" will meet you again.
The theme of this issue of the report is Alzheimer's disease and the Aβ model that has dominated the field for nearly 30 years.
On July 21, 2022, Charles Pillar, an investigative reporter for Science magazine, published a blockbuster investigation report on the field of Alzheimer's disease research. The protagonist of this report, Matthew Schrag, a neuroscientist at Vanderbilt University in the United States, found that dozens of Alzheimer's disease research papers may show signs of reusing images and manipulating data, including several papers that have been widely disseminated in the field and even directly directed several drug development [1].
The report sparked an extremely lively discussion among the global scientific and drug discovery community. Many people understand this as if the basic research and drug development of Alzheimer's disease in the past few decades have been completely misled by these fake studies. Everything has to start all over again.
Is that the case?
A simple answer might be: to understand this way, you are getting a potentially correct answer in the wrong way. The academic misconduct mentioned in this survey report certainly deserves to be taken seriously, but it is far from enough to shake the accumulation of Alzheimer's disease research over the past few decades. But at the same time, even without considering this investigation report, the accumulation of Alzheimer's disease research in the past few decades has really reached the time for serious reflection and review.
The above two sentences may seem contradictory, but behind them are the great challenges and historic opportunities facing the field of Alzheimer's disease research.
Alzheimer's disease got its name from the German physician Alois Alzheimer more than 100 years ago. Beginning in 1901, the doctor tracked and documented an elderly female patient named Auguste Deter. In years-long and continuous interviews, Dr. Alzheimer documented many peculiar behaviors of the female patient: she couldn't remember her name, whether she was married, couldn't remember what she ate at lunch, and couldn't understand the tasks the doctor asked her to complete.
In 1906, Tite died, and Dr. Alzheimer was authorized to dissect her brain. In the patient's brain, Alzheimer observed severe brain atrophy and widespread death of nerve cells. Note, however, that simply seeing brain atrophy is not enough to separate Tite's symptoms from the normal aging process, after all, nerve cells die in the normally aging brain. But at the same time, Dr. Alzheimer also found that Titter's brain region was abundantly distributed with two strange and never-before-seen things: a clump of dark, granular precipitates (amyloid plaques) and equally dark fibrous filaments (neurofibrillary tangles).
Based on these anomalies, Dr. Alzheimer judged that the woman's symptoms during her lifetime should be attributed to an unknown special degenerative disease of the brain's nerve cells, rather than the normal aging of the human brain. From this, Alzheimer's named this disease, commonly known as "Alzheimer's". [2]
Currently, more than 50 million people worldwide suffer from Alzheimer's disease. It is alarming that the incidence of this disease increases with age. 65-year-olds have a 5% chance of getting the disease, while the prevalence of 80-year-olds has risen to 10%, and 90-year-olds have reached a staggering 50%. We know that the world is rapidly aging as global public health and medical services advance—that is, the world of the future is destined to emerge with an astronomical population of Alzheimer's patients.
More importantly, the disease has a long-term lack of effective treatment. In the past 20 years, only five drugs for Alzheimer's disease have been approved in the world's major pharmaceutical markets: Donepezil in 1996, Rivastigmine in 2000, Galantamine in 2001, Memantine in 2003, and donepezil + in 2014 Memantine combination drugs (we discuss several exceptions below). And all these drugs can only briefly improve the patient's condition for half a year to two years, and cannot reverse or even delay the progression of the disease in the brain. Once the drug is stopped or the drug fails, the patient's cognitive status will decline rapidly. [3]
That said, there is a huge unmet clinical need in the field of Alzheimer's disease. However, unlike many chronic diseases with complex causes and a long course (such as diabetes, depression, atherosclerosis), Alzheimer's disease has actually pointed out a very clear path for researchers and drug developers from the first day of its name: to understand or even defeat the disease, characteristic amyloid plaques and nerve fiber tangles are an inevitable starting point. They are densely present in the patient's brain, but are rare in healthy human brains.
In fact, in the past two or three decades, research around Alzheimer's disease has mainly tried to understand the origin of these two abnormal substances. Take the more well-studied amyloid plaque, which is mainly composed of a small protein called Aβ. The predecessor of this protein is a large protein called APP, composed of 6~700 amino acids, and its specific function is not particularly clear. In nerve cells, APP can be cut from the middle by two different molecular "scissors" (beta-/gamma-secretase, beta-/gamma-secretase), and finally a small segment of 30~50 amino acid length small protein is left. This is Aβ. The two main Aβ inside the protein plaque are 40 and 42 amino acids in length, respectively, so they are also called Aβ40 and Aβ42. These Aβ proteins clump together in large quantities to form the protein plaques in the patient's brain that Dr. Alzheimer observed.
In fact, the evidence has accumulated is not only that. If APP or Aβ is overexpressed in nerve cells using genetic engineering, a large amount of protein plaque does accumulate in the mouse brain. Another particularly compelling piece of evidence comes from human genetics. Almost 5% of Alzheimer's patients have a significant family inheritance, and often have a very early age of onset. The researchers found that the vast majority of these familial patients carry characteristic gene variants and are concentrated in three genes closely related to Aβ formation, namely APP, PSEN1 and PSEN2. APP is the predecessor of Aβ, and the PSEN1/2 gene encodes the main component of gamma-secretase. [4]
In this way, the relationship between Aβ, amyloid plaques and Alzheimer's disease is quite clear. The cleavage of APP proteins produces Aβ, which aggregates in large quantities to form protein plaques, which kill nerve cells and lead to Alzheimer's disease. This is known as the "Aβ hypothesis".
This explanation also immediately suggests the idea of treating Alzheimer's disease: designing drugs to reduce Aβ production, or to remove protein plaques that have already appeared. That's exactly what people have done for the last two or three decades. According to incomplete statistics, more than 200 drugs have been pushed to clinical trials, but all have suffered painful failures at different stages. In particular, many drugs have been shown to be effective in removing protein plaques from the patient's brain, but they have not been able to save the patient's cognitive decline, and some have even worsened the patient's symptoms.
A particularly noteworthy example is the drug Aducamumab that was officially approved by the US Food and Drug Administration in 2021. The drug directly targets the Aβ protein, which effectively reduces the level of protein plaques, both in animal models and in patient brains. But the drug's ability to alleviate cognitive decline is highly questionable. In two independent global multicenter phase III clinical trials, it significantly reduced the rate of cognitive decline in only one study (EMERGE), but no similar phenomenon was observed in the other study (ENGAGE). [5]
Fig. 1 Aducamamamab | Image source: medpagetoday
In July 2021, the NMPA approved the drug for marketing based on available data and huge clinical needs. However, this decision immediately attracted great controversy and criticism. Doctors and health insurers refused to pay for the drug, resulting in dismal sales of $3 million after its launch in 2021 (the annual cost of the drug is as high as $56,000 per patient), and then the European and Japanese food and drug administrations rejected its marketing application. This drug, which cost a lot of money to develop and pinned great hopes on drug developers, has in fact withdrawn from the pharmaceutical market.
In September 2022, the developers of aducamamamab, Biogen of the United States and Eisai of Japan, announced phase III clinical data of another similar drug, Lecanemab, suggesting that it can significantly improve the decline of cognitive level in patients [6]. But the drug's improvement in cognitive performance is extremely limited, and whether it can explain the reliability of the Aβ hypothesis is also a big question mark.
These clinical results clearly contradict the Aβ hypothesis we just mentioned. If Aβ aggregates to form protein plaques that can lead to Alzheimer's disease, then targeting Aβ to remove protein plaques should naturally cure the disease.
So, is the Aβ hypothesis wrong?
It is not possible to jump to conclusions so quickly. After all, the relationship between Aβ, protein plaque and Alzheimer's disease is still very solid. They always seem to appear at the same time. Therefore, another idea of patching the Aβ hypothesis naturally emerged: the relationship between the three is fine, but we are wrong about the causal relationship. The real toxic effect is the so-called oligomer formed by soluble Aβ monomers or a few Aβ molecules. In order to protect nerve cells from poisoning, the human brain actively gathers these toxic substances together to form a precipitate that is insoluble in water. In this way, Aβ aggregation certainly produces protein plaques, but the latter is not the cause of disease, but rather a barrier to protect us from disease. If it is removed, it will only increase the risk of disease [7].
Under the guidance of this idea, researchers began to carefully analyze the aggregation morphology of various Aβ molecules, and indeed proved that Aβ molecules can be combined to form oligomers in several to dozens of ways. In 2006, a paper published in the journal Nature claimed to have found a stable, very specific Aβ dodegumer (named Aβ*56 because of its relative molecular mass of 56,000) in the brains of Alzheimer's disease mice [8]. Researchers from the University of Minnesota also purified Aβ*56 aggregates in mouse brains and injected them into the brains of healthy rats, proving that it can indeed lead to cognitive decline.
This paper is an important object of attention in the Science survey report. This popular paper, which has been cited more than 2,000 times, was found to have the problem of reusing some experimental images; Not only is there duplication within the same article, but also the same picture of the paper published in the same laboratory a few years later. In the world of science, this is basically equivalent to indicating that researchers are deliberately fabricating data, and the experimental results of the paper cannot be believed.
This is, of course, a very serious academic misconduct that also needs to be seriously punished. But to say that this crackdown has shaken up the entire Alzheimer's disease study is an exaggeration. In fact, even among peers, Aβ*56 has long been a very suspicious finding. As early as 2008, some researchers published a paper claiming that Aβ*56 could not be extracted in the brain [9]. Subsequently, many fellow scientists claimed that they could not replicate similar studies, and they had not heard of anyone who could replicate [10]. In other words, although the scientific community did not explicitly falsify and disprove this discovery, colleagues have long voted with their feet to dismiss it. Laymen often need to go deep into their small scientific counterparts to realize this.
More critically, in the field of drug development, drugs targeting Aβ oligomers, albeit in smaller numbers, have all failed (including Solanezumab, Crenezumab, and Gantenerumab). Like the logic just discussed, human clinical trials are the evaluation criteria for the conclusion of scientific hypotheses. Since clearing Aβ oligomers does not treat the disease, it means that they may not be the direct cause of Alzheimer's disease either.
Speaking of which, you can understand the meaning of the sentence at the beginning of my article: the problem mentioned in this survey report is serious, because it reveals the falsification of many important academic papers, but it is far from enough to shake the accumulation of Alzheimer's disease research in the past decades, especially the Aβ hypothesis, which is supported by biochemistry, cell biology and human genetics. But at the same time, even without considering this investigation report, all experimental drugs targeting Aβ have all suffered clinical failures, and the accumulation of Alzheimer's disease research in the past few decades has really reached the time for serious reflection and review.
But Alzheimer's disease is still a problem that we must understand and solve. What's next?
To be fair, I think it's a good opportunity for the field as a whole to reflect and start again. The philosopher of science Karl Raimund Popper famously said, "Whenever a theory seems to you to be the only explanation, you need to take it as a warning that you may not understand either the theory nor the problem it is trying to solve." I think it's very appropriate to use it here.
The truth behind this sentence is very real, and people have the psychological urge to grab life-saving straws under heavy pressure. When the Aβ hypothesis provides the only systematic explanation for Alzheimer's disease, when we face great pressure to understand and solve Alzheimer's disease, the Aβ hypothesis becomes a lifesaver, a crutch and lifebuoy that researchers have to rely on when doing basic research and drug development. And this may well indicate that we are already on the road of self-paralysis and deception.
Now is such a time for reflection – especially when an important paper in the field has been found to have been fraudulent.
In fact, if you look at it quietly, there have already been many cracks in the "shell" of the Aβ hypothesis. It is true that the protein plaques in the patient's brain are made of Aβ, which is formed by secretases that cleave APP proteins. But the causal relationship between Aβ and nerve cell death and cognitive decline has never been particularly tight. As we just said, overexpression of Aβ in nerve cells can produce a large number of amyloid-like plaques as expected, but the appearance of these plaques does not kill the cells of the mouse brain, and the cognitive function changes in mice are not obvious. Combined with the results of human clinical trials that even if the protein plaque is removed, it cannot cure the disease, we can at least say that the pathogenesis of Alzheimer's disease is still very vague. Of course, it and Aβ often appear synchronously, but whether there is a causal link between the two, and what kind of causal connection, is a very suspicious question.
To abandon the Aβ hypothesis, we have to start all over again. Re-examine and analyze existing data, look for new evidence and direction, find hope from despair, and look for any bright clue from the pitch black. This is bound to be a very painful process. In fact, there is a special term to describe it - paradigm shift. The term was coined by the famous historian of science, Thomas Kuhn. It means, in layman's terms, that all the basic assumptions and rules in a field must be overturned.
In my opinion, Alzheimer's disease has also reached a critical point in a paradigm shift. No matter how self-justifying the Aβ model may seem, since this model cannot explain the disease or guide drug development, we have to smash it up with our eyes closed and make a new model.
In this process, there will inevitably be a phenomenon that occurs in all paradigm shifts: we will see a large number of various hypotheses and models emerging in a short period of time, each claiming to be able to replace the original model and do better. This phenomenon, at best, is a hundred flowers blooming, and a difficult thing to say, is a group of demons dancing.
Here I will give a few more representative examples.
Some are "revolutionary" and relatively weak. It's more about patching the Aβ hypothesis in the hope that it will continue to work.
For example, there are still many scientists who believe that the toxic effect of protein plaques is real, and various drugs have failed, possibly because the onset of Alzheimer's disease is very long, and when the patient is officially diagnosed, the toxic effect has lasted too long to save [11]. Some scientists still believe that although Aβ*56 is most likely an experimental error, and may even be deliberately falsified, some oligomer forms of Aβ are still the culprit, and these special pathogenic oligomers need to be found to effectively develop drugs. [12]
In addition to Aβ, many people have set their sights on another sign that appears in the brains of Alzheimer's patients: nerve fiber tangles. Just now we deliberately skipped this part of the introduction. Like Aβ proteins that aggregate to form amyloid precipitates, nerve fiber tangles are made up of an abnormal protein, superphosphorylated tau. Mutations in the tau gene can also trigger neurodegenerative manifestations similar to Alzheimer's disease. In other words, the evidence supporting the Aβ hypothesis is mostly true for tau proteins. Tau protein deposition has also been found to predict cognitive decline in the human brain better than Aβ protein plaques [13].
The reason why this new tau model is relatively weak is because everyone feels that the aggregation of phosphorylated tau protein is actually affected by Aβ protein. In other words, it is nothing more than the difference between whether the Aβ protein directly causes the disease or indirectly causes the disease through tau. But this less revolutionary model may also be wrong, as the drugs developed so far for tau protein, albeit in small numbers, have failed in clinical trials [14].
While these less revolutionary models were still struggling, some of the more deviant hypotheses gradually began to surface.
For example, some scientists believe that Alzheimer's disease may be a contagious brain disease similar to mad cow disease (Creutzfeldt–Jakob disease). It has been previously found that in addition to transmission through the consumption of diseased beef products, mad cow disease also has a very rare transmission channel, and some people will also suffer from mad cow disease after receiving growth hormone injections. Later studies found that the growth hormone products received by these infected people came from extracts from the pituitary gland of mad cow disease patients, which carried prions, the pathogen of mad cow disease.
In 2015, an academic paper published in the journal Nature found that some patients who developed mad cow disease as a result also developed typical symptoms of Alzheimer's disease. In their brains, a large number of amyloid precipitates formed by Aβ proteins were also found. The researchers further found that the growth hormone products injected into these people contained not only prions, but also Aβ. Therefore, one explanation is that Aβ itself is also contagious, and once injected into the brain of a normal person, it attracts and gathers more Aβ proteins to form protein plaques [15]. In 2018, researchers demonstrated this possibility with mouse models [16].
If the link between Alzheimer's disease and mad cow disease is wild enough, then the next explanation may be even more maddening. In early 2019, a paper published in the journal Science Advances said that a very strange bacterium - Porphyria gingivalis - has been found in the brains of many Alzheimer's deaths. The reason why it is strange is that this bacterium is supposed to live in the human mouth, and it is the culprit that causes a series of gum diseases. Therefore, these scientists believe that Alzheimer's disease is caused by bad bacteria in the mouth that do not know how to enter the brain. They found that killing these bacteria alleviated the disease manifestations of Alzheimer's disease in mice [17].
On this lively occasion, our Chinese scientists are not absent. In 2019, Geng Meiyu's lab at the Institute of Materia Medica, Chinese Academy of Sciences, found that an oligosaccharide molecule called mannutner, extracted from kelp, was able to alleviate cognitive impairment in mouse models of Alzheimer's disease. More interestingly, they believe that this effect is achieved by regulating the gut flora of mice, thereby affecting the immune function of mice [18]. This molecule has completed a phase 3 clinical trial in human beings and has been officially conditionally approved for marketing by the State Food and Drug Administration, with the trade name "phase nine one".
There is another research direction worth mentioning. In 2016, scientists at the Massachusetts Institute of Technology in the United States found that in a mouse model of Alzheimer's disease, the intensity of a brain wave with a frequency of 20-50 hertz, gamma brain waves, was significantly reduced. If gamma brain waves were intensified in the brain using optogenetic methods, the mice showed signs of alleviating protein plaques and cognitive impairment. This still sounds like a logical discovery, but then it began to develop in an increasingly sci-fi direction. Researchers have found that as long as mice are shown 40 hertz of flashing light, listen to 40 hertz sounds, or sound and light combined stimulation, they can directly activate gamma brain waves in the mouse brain and play a role in the treatment of Alzheimer's disease [19]. Similar studies have even made it to formal clinical trials. At the recent 2022 Alzheimer's Association Annual Meeting (AAIC 2022), the developers of 40 Hz acousto-optic therapy also submitted preliminary human clinical data stating that it does delay brain atrophy and is safe [20].
For the above two studies, some people even joked that it seems that Alzheimer's disease does not need any treatment, eat kelp, and then remove the fluorescent lamp rectifier at home, the disease will be better. This is certainly a joke, but it does reflect the current chaos of Alzheimer's research. One says Alzheimer's disease can be contagious like mad cow disease, one says Alzheimer's disease is caused by oral bacteria, one says that diseases in the brain can be treated by affecting the bacteria in the intestines, and the last says that flashing light can cure the disease. It's all crazy-sounding hypothesis.
But I have to remind you once again that you have to be mentally prepared for strange studies and hypotheses when the scientific paradigm shifts. It is likely that most of these hypotheses are gibberish, or even all of them false. But in the process of scientific paradigm shifting, not only do we have to accept and examine them with an open mind, we also have to expect that some of these bizarre theories can be crazy enough to be right, and even a solution that can actually explain Alzheimer's disease and help us develop drugs. This may be where this fake thesis was exposed, and it is really worth thinking about and looking forward to.
In addition, although the Aβ model is still the mainstream of academia, the pharmaceutical industry has long been honest and actively exploring new drug development ideas. Unlike academia, which seeks to thoroughly understand Alzheimer's disease, the industry's demand is more practical: in any case, it can reduce the patient's symptoms and improve the patient's quality of life. As a result, of the hundreds of investigational drugs currently in the pipeline (about one-third) still directly target Aβ, many more point to processes such as nerve cell protection, nerve signaling, inflammation, and immunity [21].
In other words, perhaps before we really understand Alzheimer's disease, there is already hope of getting a batch of drugs that can at least provide practical help to patients. Even, as has often happened in the history of human drug development, the emergence of effective drugs may help us better understand the nature of disease.
The above is all the content of this issue of the mountain patrol, I am Wang Liming, next month on the 6th, I will continue to patrol the mountain for you.
References: (swipe up and down to browse)
[1] https://www.science.org/content/article/potential-fabrication-research-images-threatens-key-theory-alzheimers-disease
[2] https://pubmed.ncbi.nlm.nih.gov/22034141/
[3] https://www.alz.org/alzheimers-dementia/treatments/medications-for-memory
[4] https://memory.ucsf.edu/genetics/familial-alzheimer-disease
[5] https://investors.biogen.com/static-files/8e58afa4-ba37-4250-9a78-2ecfb63b1dcb
[6] https://investors.biogen.com/news-releases/news-release-details/lecanemab-confirmatory-phase-3-clarity-ad-study-met-primary
[7] https://pubmed.ncbi.nlm.nih.gov/22286176/
[8] https://www.nature.com/articles/nature04533
[9] https://www.nature.com/articles/nm1782
[10] https://www.alzforum.org/news/community-news/sylvain-lesne-who-found-av56-accused-image-manipulation
[11] https://pubmed.ncbi.nlm.nih.gov/28460541/
[12] https://www.nature.com/articles/s41570-021-00254-9
[13] https://pubmed.ncbi.nlm.nih.gov/31894103/
[14] https://pubmed.ncbi.nlm.nih.gov/33303932/
[15] https://www.nature.com/articles/nature15369
[16] https://www.nature.com/articles/s41586-018-0790-y
[17] https://www.science.org/doi/10.1126/sciadv.aau3333
[18] https://www.nature.com/articles/s41422-019-0216-x
[19] https://pubmed.ncbi.nlm.nih.gov/30879788/
[20] https://www.bloomberg.com/press-releases/2022-07-13/cognito-therapeutics-announces-data-presentations-at-the-alzheimer-s-association-international-conference-2022
[21] https://alz-journals.onlinelibrary.wiley.com/doi/10.1002/trc2.12295
Plate Editing | Ginger duck