Stop that phenylalanine now!

I'm going slightly off-piste with this post not strictly related to autism; however remaining true to my interest in all things amino acids, and in particular one of those most interesting aromatic amino acids, phenylalanine.

Phenylalanine @ Wikipedia

A few months back Chemistry World carried a very interesting article by Jon Evans* on how the amino acid phenylalanine might very well have the capacity to form amyloid-like fibrils classically related to conditions like Alzheimers disease. The article was based on this study by Adler-Abramovich and colleagues** which reported on a few pretty important observations.

I've done amyloid or rather amyloid precursor protein (in relation to autism) on this blog before (see here). Fibrils as their name suggests, are fibre-like structures. Without trying to plagiarise the article or study, a few points are worth noting:

• The starting point was the metabolic condition phenylketonuria (PKU), probably the most famous of the inborn errors of metabolism.
• PKU... the problems with metabolising phenylalanine as a result of issues with phenylalanine hydroxylase leads to a build up of the amino acid which has some particularly nasty effects on the developing body and brain. Said phenylalanine levels normally controlled by lifetime dietary restriction of phenylalanine but also potentially another intervention (BH4).
• Adler-Abramovich et al observed that phenylalanine in solution tended to 'clump' together forming something that looked like amyloid-like fibrils. Further when such phenylalanine fibrils were added to cell lines, the fibrils showed some degree of toxicity. Even further, rabbits injected with said fibrils started to generate antibodies against them. Finally, after imaging the phenylalanine fibrils, there was a match between what they saw in the laboratory and what they saw in an engineered mouse model of PKU and indeed in real people diagnosed with PKU.
• Ipso facto: phenylalanine can form amyloid-like fibrils and such fibrils might account for the tissue damage observed in PKU.

I admit to being really quite excited about this work. Modern science knows quite a bit about PKU following the very astute observations made by Dr. Asbjørn Følling. One area however that has always required a little more study was exactly how the build-up of phenylalanine caused the problems that it does; maybe part of that question has been answered with this work.

My mind also goes back to those other aromatic amino acids like tryptophan and tyrosine and whether similar fibrils could be formed from these compounds. The answer, according to the published scientific literature so far, is a very possible yes; as per studies like this one from Cecchini and colleagues*** although quite a lot of the investigations in this area goes well above my state of knowledge.

So, once again amino acids do their stuff. It makes you wonder whether Alzheimer's disease, at least some cases, might also reflect issues with phenylalanine. Logically also whether interventions like a low phenylalanine diet or even BH4 supplementation (see here) might reflect possible areas of intervention requiring much further study?

To finish, I'm sure many of you had the same 'could try harder' reports from school as I did when I was a kid. But to get something like this.... from your Dad?

----------

* Evans J. Anti-social amino acids gang up. Chemistry World. August 2012.

** Adler-Abramovich L. et al. Phenylalanine assembly into toxic fibrils suggests amyloid etiology in phenylketonuria. Nat Chem Biol. 8: 701-706.

*** Cecchini P. et al. The role of tryptophan in protein fibrillogenesis: relevance of Trp7 and Trp14 to the amyloidogenic properties of myoglobin. PEDS. 2012; 25: 199-203.

----------

Adler-Abramovich L, Vaks L, Carny O, Trudler D, Magno A, Caflisch A, Frenkel D, & Gazit E (2012). Phenylalanine assembly into toxic fibrils suggests amyloid etiology in phenylketonuria. Nature chemical biology PMID: 22706200

Caspase the executioner

The very dramatic title to this post pays homage to the caspases and their role in the process of programmed cell death (apoptosis). Their 'executioner' label denotes their role in the cascade effect that ends the life of a cell. If, like me, you watch too many films and documentaries about Medieval England (don't we all?), you might picture the executioner as a hangman, face covered, being paid to go about his duties. Grim stuff indeed.

Enough of all that. Like many things connected to our very complicated human biology, the protease family known as the caspases are also involved in other processes outside of just cell death. One of these processes is inflammation and partially mediating inflammatory functions such as those linked to autoimmunity.

A recent paper by Siniscalco and colleagues* sheds some light on a possible role for the caspases in connection to autism spectrum conditions. There was always going to be some interest for me in this paper given the authorship group. When names like Laura de Magistris (she of the leaky gut) and Alessio Fasano appear, I tend to sit up and take note. Please don't take this as any kind of idol-worshipping or anything like that; merely some admiration for their collected works; sort of winners of an X-Factor for scientists (now there's a novel idea).

Anyway, I mentioned that the caspases are a group of enzymes with quite a few members. Within the family, various caspases have various different duties and show various different effects when things aren't quite as they should be. The crux of the current paper was to demonstrate whether caspase expression (mRNA and protein levels) were different in a small group of children diagnosed with autism compared with controls.

The answer bearing in mind the ever-so-small numbers included for study was yes, levels were different; and different across a few of the caspases. The messenger RNA (mRNA) levels were increased for caspases-1, -2, -4. -5. Caspases-1, -4, and -5 are important because of their connection with inflammation. Important also because the caspases need to work synergistically together in order to have their 'full activity'. When protein levels of the caspases were measured, caspases-3, -7, -12 were found to be elevated in the autistic group. Caspase-3, when activated, for example seems to play a role in neurodegenerative conditions such as Alzheimer's disease (AD) and its connection with the amyloid beta precursor protein (APP). The effect is described as an enhancing one, in that caspase-3 is in the thick of all those tangles and plaques noted in AD. Caspase-12 takes us back to inflammation and caspase-7 back to apoptosis.

Cumulatively, the various caspases examined and reported on, seem to imply inflammation and a role for the immune system in the cases studied. I am a bit baffled by the caspase-3 finding in light of the connection to AD in other work, which perhaps contrasts with what is known about APP and those wretched peptides in relation to autism. Having said that, I do realise that AD is a very complicated condition and the presence of caspase-3 may reflect other processes. So how about a possible link between caspase-3 and type-1 diabetes instead?

This is not the first time that the caspase family have appeared on the autism research landscape. Mady Hornig and colleagues (she of the recent carbohydrate metabolism paper) discussed caspase-1 in relation to apoptosis and neurodevelopmental damage in this paper from a few years back. This paper by Sheikh and colleagues reported caspase-3 as being a more direct marker of apoptosis being elevated in the cerebellum portion of the brains of people with autism. Cerebellum, autism, Eric Courchesne?

The caspases are an interesting family of enzymes, of that there is no doubt. Further large-scale trials are needed to confirm findings and subsequently to start thinking about when, how and why the caspase findings become relevant to autism and what can be done to modify any detrimental effects.

To forget all this talk about executioners, how about Huey and some News?

Siniscalco D. et al. The expression of caspases is enhanced in peripheral blood mononuclear cells of autism spectrum disorder patients. JADD. October 2011.

Amyloid precursor protein and autism

A cobbler should stick to his last. Meaning that people should generally stick to what they know and do best. I think most people would feel comfortable with this phrase. We would after all never (knowingly or willingly) go for surgery at a bakery or have our teeth fixed by a plumber. Likewise I probably wouldn't want my doctor to butcher my pork chops, even if she was handy with a scalpel.

I hope that I am not getting too out of my depth with this quite biochemistry-heavy post on some new findings on the amyloid precursor protein in relation to autism spectrum conditions. I have consulted with a few people who know a little more about this than I, just to check a few facts, but please do not assume I am an expert in this area. We will see if I do the science and research justice.

I start with some definition and description. Those with a medical or scientific background or possibly some family experience will probably already have heard of amyloid precursor protein (APP) in relation to Alzheimer's disease (AD). AD is the most common cause of dementia and the condition's profile is being elevated by people such as Sir Terry Pratchett (he of the Discworld novels). Without going too deeply into AD, the characteristic brain pathology of the condition is noted by the presence of neurofibrillary tangles and amyloid plaques. The plaques are made up of certain accumulated breakdown products of APP.

Nobody really knows how such pathology comes about in terms of genes and environment, but amongst the various investigations being undertaken into AD, one of the more widely suggested hypotheses involves APP and in particular, its peptide fragments, beta-amyloid, which are the main constituent of the plaques found in AD. Like any peptide, beta-amyloid is formed as a result of enzymatic processes involving a protease acting on APP, specifically the beta- and gamma-secretases. There are various other processes involved in this chain of events potentially related to AD but that's as far as I am going for now. I would perhaps also mention about the various pharmacotherapeutic measures currently available to tackle AD, based around two main classes of drug: the acetylcholinesterase inhibitors, which increase acetylcholine (often depleted in AD) and the glutamate-blocker, memantine.  I have talked about acetylcholinesterase inhibitors previously in relation to autism (here). I note that memantine has also been trialled in cases of autism with some indications of positive symptom response.

A recent paper by Ray and colleagues* (open access copy available here) looked at some of the products of amyloid precursor protein (APP) in plasma from people with autism. The paper is a quite complicated one and not the easiest piece to follow, hence my use of bullet-points to break the methods and selected findings down:

• A total of 39 participants were included for study: 15 diagnosed with severe autism (CARS score of 37+), 6 with mild-moderate autism (CARS: 30-36.5) and 18 asymptomatic controls. There was no significant difference across the mean ages of the groups.
• Various measures of the residues and peptides derived from APP in plasma were analysed alongside levels of brain-derived neurotrophic factor (BDNF). The main reason seems to be the study's focus on BDNF and specific APP breakdown products being neurotrophic (related to neuronal growth or survival) on the back of the data looking at head circumference in autism.
• There was no significant difference across the groups in terms of the mean total of total secreted amyloid precursor protein (sAPP) (that is the combined species based on the actions of the various forms of the secretase on APP).
• Levels of sAPP-alpha, the residue of the non-plaque associated form of APP, were elevated in the more severely affected autism group compared with controls.
• Levels of sAPP-beta, the residue associated with plaques from APP, were reduced in the more severely affected autism group compared with controls.
• Levels of the beta-amyloid peptides involved in plaque formation, including the most insoluble and toxic peptide, beta-amyloid-42, were also reduced in the more severely affected autism group compared with controls.

These are interesting findings. Plasma levels of beta-amyloid peptides (particularly 42) are associated with cognitive decline and onset of AD. I would however caution before making too many statements about the study results and risk of AD based on things like the small participant numbers included in the current study, and also their young age.

This is not the first time that this particular research group have published in this area. Indeed in a previous study based on even smaller participant numbers, they reported increased total sAPP in severe autism relative to other groups and controls; contrasting with no overall difference in the current study. Perhaps more interesting however was the confirmation from the recent study of higher levels of non-plaque related sAPP combined with lower levels of plaque-related beta-amyloid-40 peptide in severe autism. Likewise another group has confirmed some of theses findings.

I would like to believe that there may be some good sense in conducting further, large-scale study of levels of sAPP-alpha in autism which some have suggested might be a potential biomarker for autism. There are also a few other potential tie-ins to sAPP-alpha which might also necessitate some further investigation including a peripheral nicotinic effect, the use of statins, and the effect of selected neuropeptides.

As per my initial caveat, I would perhaps advise interested readers to do a little confirmatory reading before taking my observations as any kind of truth. If on the other hand you need someone to name your Star Wars villains, well I'm yer man..

* Ray B. et al. Increased secreted amyloid precursor protein-α (sAPPα) in severe autism: proposal of a specific, anabolic pathway and putative biomarker. PLoS ONE. 2011: 6:6