Despite billions of dollars spent and decades of research, the exact cause of Alzheimer’s is still unknown, and more than 400 clinical trials have failed to produce a meaningful treatment.
In 1906, Dr. Alois Alzheimer presented brain autopsy findings on his patient, fifty-one-year-old Auguste Deter, who had died with dementia, including the hallmark plaques and tangles, but also abnormal lipid deposits that are only recently a focus of research. For decades, the “amyloid cascade hypothesis” has dominated Alzheimer’s research and drug discovery. Under certain conditions, amyloids and other proteins can form clumps that are present in many disease processes within and outside the brain (Ezzat 2022). The amyloid cascade hypothesis holds that the hallmark plaques composed of beta amyloid, which begin to appear years before symptoms, are toxic and are the cause of progressive neurodegeneration in Alzheimer’s disease.
Beta amyloid is just one of many bioactive fragments produced by enzymatic cleaving of the much larger amyloid precursor protein (APP). APP is produced in the brain within neurons and astrocytes but also in many other tissues outside the brain. APP and/or its bioactive fragments appear to synthesize, maintain, monitor, and modulate activities in the synapse, influence the growth of axons and dendrites, transport neurotransmitters, and regulate lipid metabolism related to regeneration of the cell and the mitochondrial membranes. In the amyloid cascade hypothesis, overproduction of APP presumably leads to increased secretion of beta amyloid from the neuron, thereby promoting formation of plaques, tangles, and inducing inflammation.
Beta amyloid is produced continuously from APP and crosses the blood brain barrier in both directions. Beta amyloid fragments (or peptides) are strings of amino acids of variable length and Aβ42 and 40 appear to have the greatest propensity to form amyloids. Soluble beta amyloid has normal beneficial functions within and outside of the brain such as activation of certain enzymes, protection against oxidative stress, regulation of cholesterol transport, and antimicrobial activity (Chen 2017). APP and beta amyloid appear to have many other functions that have not yet been fully defined.
Under certain conditions, soluble (dissolvable) beta amyloid fragments (specifically Aß 40 and 42) secreted from brain cells can misfold and become insoluble as they stick together to form sheet-like structures, clumps, and larger plaques. When present in excess, beta amyloid is thought to enter and directly damage or kill neurons, other brain cells (glia), and their organelles, including mitochondria, where ATP and thousands of proteins are produced. Microbes like herpes simplex and fungi as well as calcium carbonate, zinc, iron, copper, and other heavy metals have been found within the beta amyloid plaques.
Outside the cells, the beta amyloid clumps and plaques can stick to and damage small and large blood vessels and impede blood flow, damage cells, nerve fibers, and synapses, and can interfere with transmissions along nerve pathways. Overproduction of APP and/or defective removal from the brain could explain the abnormal accumulations of beta amyloid plaques in people with Alzheimer’s. An important unanswered question is whether the formation of beta amyloid plaques is the cause or a downstream effect of whatever causes Alzheimer’s disease.
Emerging evidence challenges the amyloid cascade hypothesis
About 60 percent of people in their mid-eighties have significant accumulation of beta amyloid plaques in the brain but a much smaller number have Alzheimer’s or other type of dementia. Two biomarkers commonly used to establish Alzheimer’s for clinical trials are a high burden of amyloid plaque on PET imaging or a low level of cerebrospinal fluid soluble beta amyloid (CSF Aβ42). The working idea is that soluble beta amyloid in the brain, as evidenced by low CSF Aβ42, decreases as plaque accumulation increases. However, many people with large amounts of plaque in the brain perform normally on cognitive testing and some people diagnosed with Alzheimer’s have minimal beta amyloid plaque on PET imaging studies, including people with the dominantly inherited Osaka mutation, who also have low levels of soluble CSF Aβ42 (Sturchio 2021).
Two recent studies by a collaborating group of researchers in the U.S. and Europe seriously challenge the amyloid cascade hypothesis. In their first study, 598 people completed amyloid PET imaging, neurocognitive testing, and measurement of CSF Aβ42. Regardless of whether a low, medium, or large amount of amyloid plaque was present, people who had higher levels of CSF Aβ42 had normal cognition. People with mild cognitive impairment had reduced levels and people with Alzheimer’s had even lower levels of CSF Aβ42 (Sturchio 2021). In the second study, 162 Dominantly Inherited Alzheimer’s Network (DIAN) participants, who develop early-onset Alzheimer’s, were followed for more than three years on average including brain imaging, CSF Aβ42 measurement, and neurocognitive testing. Forty-three people showed progression during that time frame on the Clinical Dementia Rating (CDR) scale. The people with higher baseline CSF Aβ42 levels had reduced risk of progression and had longer progression-free survival, as well as higher glucose uptake on FDG-PET, and higher hippocampal volume compared to the people with lower baseline CSF Aβ42 levels. The amount of amyloid plaque accumulation did not carry a similar prediction and did not correlate with CSF Aβ42 levels. It would then appear that reduced availability of soluble beta amyloid rather than excessive beta amyloid production and plaque accumulation is related to progressive cognitive decline in Alzheimer’s (Sturchio 2022).
In another study of human neural stem cells with the “Swedish mutation” of inherited familial Alzheimer’s, there was increased production of soluble beta amyloid by the cells compared to the “wild type” stem cells. Unexpectedly, rather than producing toxic effects, the increased beta amyloid production resulted in higher numbers and density of synapses, a beneficial effect. Previous studies have used external sources of beta amyloid applied directly to the neurons, possibly in excessive amounts or concentrations, or provided in the oligomeric (several peptides stuck together) rather than monomeric form (a single peptide), which is less likely to clump (Zhou 2022).
The unexpected results in these three studies could explain why drugs aimed at reducing beta amyloid have failed to bring about meaningful improvement and sometimes accelerated the disease process. Out of thirty-five clinical trials of drugs that reduce beta amyloid, compared to placebo, thirteen drugs showed accelerated worsening of cognition, and sometimes increased brain atrophy, twenty showed no change in cognition, and two have not published results. While some of the drug companies have abandoned further testing, others concluded that larger doses must be needed or that the drug should be given earlier in the disease. So long as the drug company continues to conduct a large clinical trial, the FDA recently approved a very expensive intravenous drug Aduhelm® (aducanumab) for mild Alzheimer’s based on a miniscule improvement on a cognitive test in just one of three clinical trials in people with mild Alzheimer’s (Espay 2021).
Another study turns the amyloid cascade hypothesis upside down
Experiments in five different Alzheimer’s mouse models characterized by overproduction of APP showed that, rather than forming plaques outside the neuron and causing death of the neuron due to a toxic effect, beta amyloid accumulates and forms plaques within the neuron as it dies and disintegrates. They reported that the death of the neuron begins with defective acidification of cell structures called lysosomes that help maintain the health of the cell by removing obsolete proteins and organelles (called autophagy). Abnormal processing of ATP appears to be involved as well. Beta amyloid is just one of the proteins that builds up while the cell is dying and forms a flower-like network of fibrils around the nucleus of the neuron thereby called a PANTHOS, or poisonous flower. When the neuron dies and disintegrates, the “flower” left behind is the classic beta amyloid plaque. PANTHOS formations were also confirmed in the brains of people who died with Alzheimer’s. In this scenario, the formation of plaque appears to happen as part of the dying process of the cell rather than causing the death of the cell through external toxic effects (Lee 2022).
Is there a treatment that could increase CSF soluble beta amyloid?
A crossover study reported in 2020, makes a solid case for adopting a Mediterranean-style ketogenic diet to help people with mild cognitive impairment and Alzheimer’s. Twenty people with subjective memory complaints or mild cognitive impairment (MCI) consumed either a modified Mediterranean-ketogenic diet (5 to 10 % carbohydrate, 60 to 65 % fat, and 30 % protein) or an American Heart Association (AHA) Diet (55 to 65 % carbohydrate, 15 to 20 % fat, and 20 to 30 % protein) for six weeks. They resumed their usual diet for six weeks then switched to the other diet for six more weeks.
Compared to baseline studies, participants on the Mediterranean keto diet had elevated ketone levels and significantly improved HbA1c, fasting glucose, fasting insulin, triglycerides, and very-low-density lipoprotein (VLDL) cholesterol; there was a trend toward higher HDL cholesterol and total cholesterol was unchanged; brain blood flow and uptake of ketone bodies increased; CSF Aβ42 increased (usually decreased in Alzheimer’s); tau protein decreased in the subgroup of people with mild cognitive impairment (usually increased in Alzheimer’s). The CSF Aβ42/tau ratio increased (usually lower in Alzheimer’s). The low-fat AHA diet did not increase cerebrospinal fluid Aβ42, brain blood flow, or ketone uptake, blood ketone levels, or change fasting glucose, fasting insulin, HbA1c, total cholesterol, VLDL cholesterol, or triglyceride levels, and HDL cholesterol levels were lower. CSF tau decreased in people on the AHA diet. Memory performance improved after both diets (Neth 2020).
Conceivably, other ketogenic strategies like exercise, fasting, and ketone supplements could have similar effects to the Mediterranean keto diet. If confirmed in future studies, adopting a ketogenic lifestyle could provide a treatment that would increase CSF Aβ42, along with many other beneficial metabolic changes, and we will not have to wait many years for a targeted drug that increases CSF Aβ42 to be developed and tested. Experts report that modifiable lifestyle risk factors account for at least 30% of dementia cases, with poor diet at the top of the list. Adopting a ketogenic lifestyle could potentially delay or even prevent Alzheimer’s.
Written by Mary T. Newport, M.D.
Espay AJ, et al. “Soluble amyloid-β consumption in Alzheimer’s disease.” J Alzheimers Dis V. 82 No. 4 (2021):1403-1415.
Ezzat K, et al. “Proteins Do Not Replicate, They Precipitate: Phase Transition and Loss of Function Toxicity in Amyloid Pathologies.” Biology (Basel) V. 11 No. 4 (2022):535.
Lee JH, et al. “Faulty autolysosome acidification in Alzheimer’s disease mouse models induces autophagic build-up of Aβ in neurons, yielding senile plaques.” Nat Neurosci V. 25 No. 6 (2022):688-701.
Sturchio A, et al. “High soluble amyloid-β42 predicts normal cognition in amyloid-positive individuals with Alzheimer’s disease-causing mutations.” J Alzheimers Dis V. 90 No. 1 (2022):333-348.
Sturchio A, et al. “High cerebrospinal amyloid-β 42 is associated with normal cognition in individuals with brain amyloidosis.” E Clinical Medicine V. 38 (2021):100988.
Zhou B, et al. “Synaptogenic effect of APP-Swedish mutation in familial Alzheimer’s disease.” Sci Transl Med V. 14 No. 667 (2022):eabn9380.