Heavy metal: the chemistry of Alzheimer’s disease

This feature appeared in the May/June 2009 issue of Australian Life Scientist. To subscribe to the magazine, go here.

In the mid-1990s, when the key figures behind the metal-ion hypothesis of Alzheimer’s disease first began making their thoughts known, the idea that copper and zinc had anything to do with neurodegenerative diseases was considered far-fetched, to say the least. But researchers like Australia’s Colin Masters and Ashley Bush and Harvard University’s Rudi Tanzi were on to something, and they knew it.

The metal-ion hypothesis is now almost the orthodoxy, but in that time much has been learned about the complex interactions between the amyloid precursor protein (APP), the soluble beta amyloid peptide (Abeta) it cleaves, the role of amyloid plaques in the brain and how it all turns toxic. Helping to understand the basic chemistry at play is Associate Professor Kevin Barnham, who has used his inorganic chemistry expertise to help elucidate the biology.

Barnham’s lab at the University of Melbourne concentrates on delineating the chemical events associated with neurodegeneration, particularly the metallobiology and membrane biology of diseases like Alzheimer’s and Parkinson’s. Recently, he and his team have developed a mass spectrometry-based technology for identifying the oligomeric form of Abeta that is suspected of being the cytotoxic species.

Barnham originally started out as an inorganic chemist, so metals were there from the start, he says. He trained with Trevor Appleton at the University of Queensland, studying platinum anti-cancer drugs and learning about drug discovery using nuclear magnetic resonance (NMR). He then spent three years at Birkbeck College in London, working with the inorganic chemist and drug design expert Peter Sadler, again on platinum chemistry and its interaction with protein and DNA. When that project had run its course, he returned to Australia to spend a very productive five years with Ray Norton at the now-defunct Biomolecular Research Institute (BRI), where he learned how to do protein structure by NMR.

It may not be something many would admit to, but he rather liked it. “It’s really like doing jigsaw puzzles,” he says.

It was at the BRI that Barnham first met the driving forces behind the nascent biotech company Prana Biotechnology, Colin Masters and Ashley Bush. “Ashley was flogging his metals hypothesis of Alzheimer’s at the time, so when the project landed on the BRI’s desk, they took one look at it and said Kevin’s the perfect guy for this, because of my protein structure and NMR skills,” Barnham says. “So I got involved, basically looking to help drive the Prana chemistry program.”

Barnham was a co-inventor of Prana’s metal-protein attenuating compound (MPAC) technology, which is now in Phase II trials for Alzheimer’s. While he is most certainly not speaking as a representative of Prana, he has a deep understanding of the MPACs and believes they are working as a kind of chaperone.

“Basically our working hypothesis at the moment – and these things are evolving – is that as people age there is a breakdown in homeostatic systems,” Barnham says. “One of the most important homeostatic systems is metal ions. The transport and utilisation of oxygen is dependent on metalloenzymes and that chemistry needs to be very tightly regulated. It doesn’t take much to put it out of balance.

“We believe that as you get older, that balance starts to break down, and what we believe the MPACs are doing effectively is they are helping to restore that balance. If there is a breakdown in metal homeostasis – the amyloid beta peptide binds zinc, it binds copper, these are essential for a variety of processes – and if they are sequestered by the Abeta peptide then they cannot carry out those processes.” ---PB---

Metal ion hypothesis

The basic metal ion hypothesis is that copper ions combined with Abeta catalyses the production of hydrogen peroxide, which generates oxygen free radicals that damage proteins. Of course, it is not as simple as that, Barnham says.

“We were able to make mutant Abeta peptides that can make four times as much hydrogen peroxide as other peptides but they weren’t toxic, so clearly it wasn’t just hydrogen peroxide,” he says. “Our current working hypothesis is that Abeta interaction with metals does lead to redox chemistry, but that redox chemistry leads to modified forms of Abeta, and it is these modified forms of Abeta, these soluble oligomers, which are the problem. They are covalently cross-linked, so they won’t break down and are more proteolytically resistant.

“We believe this is what is driving the toxicity of Abeta. The reactive oxygen species are leading to the modification of Abeta, to oxidatively modified forms of soluble Abeta oligomers, and that’s our current working hypothesis.”

Much of Barnham’s work is informed by colleagues at the Mental Health Research Institute, headed by Colin Masters, where Ashley Bush runs the Oxidation Disorders Laboratory. There are also colleagues at the Austin Hospital, including Chris Rowe and Victor Villemagne, who have been running ground-breaking PiB-PET research comparing Alzheimer’s-affected brains with normal. They and others have shown that the amount of amyloid in the brain does not actually correlate with dementia, and that while all people with Alzheimer’s have amyloid plaques, so do 30 per cent of controls.

“Whether that means that those people would potentially be in trouble later on remains to be seen, but the bottom line is that all the genetic data points to Abeta,” Barnham says. “The amyloid burden does not correlate and therefore people are now starting to focus on the oligomers. The problem with the oligomers, however, is that they are very poorly defined. What is an oligomer? Is it a dimer, a trimer, is it a particular structure? You’ll find in the literature that it is a potpourri of anything people want it to be. And quite frankly the quality of the characterisation of these oligomers is poor, speaking as a physical scientist. It is not just any oligomer that is toxic. There is a lot of work left to do.”

One of the most exciting areas of research at the moment is into the roles metal ions play in the function of the NMDA receptor, Barnham says. “The metal ions modulate the NMDA receptor but if Abeta is trapping them, they can’t play that role, and then you have the Abeta toxicity on top of that.”

He believes that the MPACs developed by Prana are basically taking metals from a place they are doing bad things, and putting them somewhere they do good. “Because these are moderate chelators the metals become bioavailable again and are able to go back to doing their various functions,” he says. “They activate certain signalling pathways that play a role in a variety of processes. It is the concept of a chaperone – a metallo-chaperone.”

Prana’s researchers have recently shown that its lead compound PTB2 is able to prevent the loss of synapses, work also being conducted by Paul Adlard and David Finkelstein at the MHRI. Other work by Barnham’s colleague at the University of Melbourne, Tony White, is informing the understanding of how metals modulate cell signaling events associated with cell survival and cell death pathways.

“There is no doubt that when the metals are released it activates a whole range of things, one of which of course is the up-regulation of growth factors and the promotion of synaptic plasticity,” Barnham says. “It really is a watch this space situation.” ---PB---

Blood biomarkers

In the meantime, Barnham’s own lab is working on developing technology to identify the dimeric form of Abeta in blood. The literature is full of failed attempts to find good biomarkers in blood, and Barnham thinks he knows why.

“The way that everyone does blood tests is you collect blood, you split it into various fractions, most of which is either plasma or serum and that is normally what is analysed. All of the chunky stuff is thrown away. Well, the chunky bit contains all of the cellular membranes. Now, 10 years of Abeta research has taught us that Abeta toxicity is dependent on an interaction between an oligomeric form of Abeta, and a cell membrane. An example of that is the work that (one of Barnham’s postgrad students) Lin Hung published in the Journal of Neuroscience last year, showing the preferential binding of oligomers to membranes.

“So we figured that if Abeta is going to be in blood, where in blood is it going to be? It’s going to be binding to the membrane. Where are the membranes? They’re the bits that people throw away – the pellet fraction.”

Barnham’s lab then went out and purchased Surface Enhanced Laser Desorption Ionization Time of Flight mass spectrometry (SELDI-TOF MS) technology, which combines a chromatographic surface with mass spectrometry, to have a look at the pellet fraction.

“They have chips that we have coated in an antibody and then do mass spec directly off the chip. It’s kind of frowned upon by the main mass spec community because the mass spec component is not as accurate as high-end mass spectrometers, but the advantage is it’s a bit more sensitive, so it’s great for profiling, just not so great for positive identification of individual fragments.

“But we wanted to profile, and it turned out to be beautiful for that. We attached antibodies to our chips and we screened the blood, and low and behold …”

Working with Villemagne and Rowe at the Austin Hospital, Barnham’s team has been able to get blood from patients, fractionate it and get profiles. “We clearly saw that the profiles were different between AD and control and when we looked at some of the key peptides that were different, you could clearly say that this is Abeta 42 and see the peak there.

“Because these people had had a full work-up as part of the PiB studies – they’d had their brain amyloid burden tested, they’d had the Mini Mental State Exam (MMSE), they’d had a whole range of psych tests –we were able to show that not only were these levels elevated in Alzheimer’s but the levels in the blood were correlating with brain amyloid as determined by PiB. They correlate with MMSE and they correlate with memory. It is very encouraging.

“The monomer and the dimer are the most obvious ones because we can look at the molecular weight and go bang, we know what that is. There are other peaks in the profile that allow us to learn more – we know there is a peak that is elevated in control by 55 per cent over AD. While we positively don’t know what it is, we can make a pretty good guess: the profiles look like there is alternate processing of APP in controls versus AD.”

Barnham emphasises that this work is very preliminary, as the team has only been able to look at 117 subjects. “Ideally the field would like us to look at thousands and have it done in multiple sites and all that sort of thing, but if we can identify what these various fragments are, then we may be able to develop cheaper methods to screen and develop diagnostics.”