By Rob O’Brien with contribution from Brandon Tavshanjian

Most people nowadays are familiar with the advice that “correlation does not imply causation.” However, this is difficult advice to take when considering human disease where we are unable to ethically experiment on humans without exceptionally good justification. To get that justification we have to rely on studying the effects of a disease in order to make an informed guess about what causes it. The problem with this is that distinguishing the effects of a disease from its causes can get murky.

Comparing a healthy brain to a brain with advanced Alzheimer's disease (AD) reveals severe degeneration in the AD brain.

Patients with Alzheimer’s disease (AD) experience memory loss, cognitive decline, and severe brain atrophy.

A good example of this confusion between effects and causes is Alzheimer’s disease (AD). AD is defined by the presence of “amyloid plaques” (sticky globs of a protein fragment called amyloid beta or Aβ) in the brains of people with dementia. There is lots of good evidence that Aβ can cause AD. For example, there are families that inherit early-onset forms of AD, and have mutations in the protein itself or in enzymes that process the protein into sticky fragment forms. When these mutants are expressed in mice, the mice develop some (but not all) of the symptoms of AD. Additionally, there are people who naturally produce 50% more of the precursor to Aβ in their cells due to a completely different genetic disease: individuals with trisomy-21, or Down’s Syndrome. Down’s syndrome is cause by an extra copy of chromosome 21, which adds a whole extra copy of the Aβ precursor. An unfortunate side effect of this disease is that the majority of individuals with trisomy 21 go on to develop dementia and AD in their 30’s and 40’s.

So stopping Aβ formation is a good way to stop AD, right? That has been the main approach that pharmaceutical companies have taken to treat the disease in the past. Many drugs that aim to reduce Aβ levels in a variety of different ways have been tested, and have all more or less failed to show significant effects in people with mild cognitive impairment (mild memory loss that can develop into AD) or AD. These failures have been costly, surprising, and a bit disheartening.

So why does targeting Aβ fail? Two possibilities:

The first is that the drugs tested may have failed to get enough of the drug to the right place (the brain), or have had off target effects that were worse than the disease itself.

Postmortem tissue sample from an AD patient brain reveals AD pathology including amyloid-beta plaques and Tau tangles. (Photo credit Dr. Dale Bredesen).

Postmortem tissue sample from an AD patient brain reveals AD pathology including amyloid-beta plaques and Tau tangles. (Photo credit Dr. Dale Bredesen).

The second is a bit more controversial—maybe amyloid plaques are a red herring in  some people? AD comes in two general forms: genetic and sporadic. All of the cases previously laid out (and which were used as the basis for targeting Aβ) can be grouped into the genetic category, however these individuals make up about ~10% of the total population with AD. The sporadic form of the disease occurs in the majority of AD patients and displays the same buildup of amyloid plaques. However buildup of Aβ in sporadic AD may not be only the cause of the disease in these individuals. Instead, the disease may also be an effect of dysregulation of another protein called microtubule-associated protein tau (or tau, for short), whose function is also disrupted in AD. Tau is a fascinating protein that deserves it’s own article, but for now I’ll just say that it also forms aggregates of it’s own in the brains of AD patients, and these aggregates display characteristics that may go a long way toward explaining how and why the disease can affect specific circuits in the brain, rather than geographical regions.

In short, Aβ targeting strategies may have failed because there are two forms of Alzheimer’s that might have different root causes. If this is the case, these two diseases are being confused with each other because of the similarity of their symptoms and effects in the brain. In one (the early-onset genetic variety), problems with Aβ or its processing cause plaques that lead to neurological dysfunction. In the other, problems with tau cause defects in neurological function and form plaques of Aβ as a side effect.

So then is Aβ dead as a target for Alzheimer’s? Certainly not; in fact, Dr. Dale Bredesen from the Buck Institute has begun a clinical trial testing the small molecule F03 in patients with mild cognitive impairment (a precursor to AD). Dr. Bredesen has shown previously that F03 alters the processing of Aβ and improves cognitive function and reduces memory loss in animal models of the disease. To read more about this clinical trial, check out our blog post written by Postdoc Matt Laye. Additionally, there has been interest in specifically treating individuals with familial AD caused by mutations in Aβ with these anti-Aβ drugs. At the same time, the failures of anti-Aβ drugs has generated interest in targeting Tau as the true culprit in some forms of AD, with at least one molecule in phase III clinical trials and others in earlier stages of therapeutic development. Will any of these approaches work? Time will tell, but no matter the results of these trials, we will be closer to understanding the underlying causes of the two main forms of AD.