The structure of amyloid-β dimers in Alzheimer’s disease brain: a step forward for oligomers

Abstract

This scientific commentary refers to ‘Identification of neurotoxic cross-linked amyloid-β dimers in the Alzheimer’s brain’, by Brinkmalm et al. (doi:10.1093/brain/awz066).

This scientific commentary refers to ‘Identification of neurotoxic cross-linked amyloid-β dimers in the Alzheimer’s brain’, by Brinkmalm et al. (doi:10.1093/brain/awz066).

Scientists have sought to understand the causes of Alzheimer’s disease by identifying the pathogenic and pathophysiological molecular species that cause cognitive impairment and dementia. The discoveries of autosomal dominant mutations in genes that produce amyloid-β provided insights into amyloid-β’s role in pathogenesis; however, the relationships between amyloid-β plaques and cognitive decline turned out to not be as strong as the evidence associating tau and neuronal pathology with cognitive decline (Terry et al., 1991; Arriagada et al., 1992). This apparent disconnect, with amyloid-β plaques necessary for diagnosing Alzheimer’s disease but not sufficient for clinical symptoms of dementia, led researchers to try to resolve the discrepancy. Soluble amyloid-β assemblies termed oligomers were hypothesized to impair synaptic, neuronal and ultimately cognitive function (Gong et al., 2003; Lesné et al., 2006; Shankar et al., 2008), leading to revision of the amyloid hypothesis (Hardy and Selkoe, 2002). Amyloid-β oligomers have now been implicated more directly in synaptic dysfunction, neuronal loss, and behavioural impairments in vitro and in animal models, and have been detected in brain lysates from patients. Studies have shown a strong correlation between dementia and levels of amyloid-β oligomers (McLean et al., 1999), and endogenous amyloid-β oligomers can be up to ∼1000 times more toxic than synthetic soluble amyloid-β monomers. However, precise structural characterization of amyloid-β oligomers has been lacking, and direct evidence of their involvement in Alzheimer’s disease has proved challenging to obtain (Benilova et al., 2012).

In this issue of Brain, Brinkmalm and co-workers partially address these issues by describing the complex structures of oligomeric amyloid-β dimers (Brinkmalm et al., 2019). After extracting soluble amyloid-β from Alzheimer’s disease brains, they characterized the amyloid-β forms present within the 7 kDa band on SDS-PAGE, and identified a complex heterogeneous mixture of covalently cross-linked dimers by mass spectrometry. The authors then demonstrated that fractions containing these forms impair a measure of synaptic function, and that clearing the mixture with a pan anti-amyloid-β antibody reduces the synaptic toxicity induced by this fraction. These findings support the idea that certain brain lysate fractions containing amyloid-β can be neurotoxic, and provide much needed identification of Alzheimer’s disease brain amyloid-β dimers. The molecular identification of these amyloid-β dimers promises to open up new avenues of research testing the specificity and mechanisms of action of these putative toxic species.

However, matters are complicated by the fact that the oligomers described to date are highly heterogeneous, ranging from dimers (∼7 kDa) reported by Brinkmalm et al. and others to large, high molecular weight species (∼150–600 kDa) (Noguchi et al., 2009; Esparza et al., 2016). This wide range can be attributed to differences in the purification and extraction methods used. There is also the risk of artefactual oligomeric and other species being created in the extraction and purification process. In addition, different sized oligomers may represent discrete pools of amyloid-β conformers that are part of a wide spectrum of dynamically interchangeable species. Soluble amyloid-β appears to exist in a dynamic equilibrium between monomeric and oligomeric states, with amyloid-β dimers representing potentially important building blocks for toxic aggregates.

The findings by Brinkmalm et al. demonstrate that the 7 kDa band on western blot, and the corresponding size exclusion chromatography fractions, are a heterogeneous mixture of many hetero-dimers (e.g. amyloid-β_{1–40×1–37} and amyloid-β_{1–42×1–38}). In other words, there was no single bioactive or toxic dimer. Many in the field believe that multiple species are likely to be toxic, acting via a range of mechanisms. There is also the risk of identifying an in vitro biologically active oligomer that is nevertheless unrelated to the clinical mechanisms of Alzheimer’s disease dementia. It is unknown which, if any, of the dimers identified may be responsible for synaptic or neuronal dysfunction. However, the results of the current study will enable future testing of synthetic dimers to confirm their toxicity.

There is still much to learn about oligomers and their role in Alzheimer’s disease (Benilova et al., 2012). Identifying specific oligomeric structures that exist in vivo in the Alzheimer’s disease brain is difficult owing to experimental artefacts, and poor reproducibility of oligomer biomarker findings has been a challenge for the field. Despite the availability of drugs demonstrating target engagement of oligomeric amyloid-β (Yang et al., 2018), success in the clinic has remained elusive, as highlighted by the recent failure of an anti-oligomeric amyloid-β antibody to demonstrate clinical efficacy in early Alzheimer’s disease.

The oligomer field needs to identify and produce known oligomer structures and demonstrate which individual moieties are toxic, avoiding purification-associated artefacts. The results of Brinkmalm et al. move us a step closer to that goal. Developing models of oligomeric amyloid-β could help accelerate the development of biomarkers and therapeutics. Finally, specific drugs targeting oligomeric species will be needed if we are to demonstrate beneficial effects in patients. The clinical and biomarker data from human trials targeting oligomeric amyloid-β will be the ultimate test of the oligomeric amyloid hypothesis. In parallel with these efforts, ongoing clinical studies to understand and target tau, inflammation and other mechanisms will be required, as well as research into strategies such as removing amyloid plaques, decreasing soluble amyloid-β species, and modulating the production of amyloid-β. A diverse portfolio of basic and clinical research is needed to address the challenge of Alzheimer’s disease.

Funding

Funding for work in this paper is supported by the National Institutes of Health (R01NS095773, R56AG061900, R01NS065667) and an anonymous foundation.

Competing interests

R.J.B. receives laboratory research funding from the National Institutes of Health (R01NS095773, R56AG061900, R01NS065667), Alzheimer's Association, BrightFocus Foundation, Rainwater Foundation Tau Consortium, Association for Frontotemporal Degeneration, the Cure Alzheimer's Fund, and the Tau SILK Consortium (AbbVie, Biogen, and Eli Lilly and Co.) Funding for clinical trials include the National Institutes of Health (U01AG042791, R01AG046179, R01AG053267, U01AG059798, Alzheimer's Association, Eli Lilly and Company, Roche/Genentech, Janssen, Avid Radiopharmaceuticals, GHR Foundation, and an anonymous organization. R.J.B. also receives research funding from the National Institutes of Health (UF1AG032438, R13AG055232, U19AG010483) and the DIAN-TU Pharma Consortium (AbbVie, Amgen, AstraZeneca, Biogen, Eisai, Elan, Eli Lilly and Co./Avid Radiopharmaceuticals, Forum Pharmaceuticals, Roche/Genentech, Janssen, Mithridion, Novartis, Pfizer, and Sanofi). R.J.B. reports consulting fees/honoraria from AC Immune, Johnson & Johnson, Pfizer, Eisai, Roche, and Merck. R.J.B. reports equity ownership interest/advisory board income from C2N Diagnostics.