Autism, aromatic amino acids and gut bacteria: a hypothesis

Mass-ive @ Paul Whiteley

The publication of an interesting article by T. Andrew Clayton* in FEBS Letters (full-text) with a speculative hypothesis regarding the amino acid phenylalanine, gut bacteria and autism has my full attention. A quick search for the author suggests quite an accomplished research career with a scientific publication track-record to boot particularly in the area of metabolomics and with some very distinguished company. That and mention of one Prof. Glenn Gibson in the acknowledgements leads me to believe that this chap knows what he is talking about.

This paper was always going to attract my attention because it made reference to the compound trans-indolylacryloylglycine or IAG and autism. What on earth is IAG I hear you ask? Well it's a compound, a peculiar metabolite of the amino acid tryptophan, that has surrounded the research I have been involved in for many years both from a detection and possible biological marker point of view. As it happens, the initial speculation on IAG potentially being a more general biomarker for autism has subsided as per this article and this article but that's not to say that it may not be relevant to some groups on the autism spectrum, particularly those who also present with comorbid gastrointestinal issues. Indeed there still remains the possibility that there may be some 'relationship' between IAG, gluten and the so-called leaky gut (more on this later).

The crux of the current paper is that another amino acid, phenylalanine, which shares more than a passing relationship with other aromatic amino acids particularly tryptophan and tyrosine in terms of the 'give-and-take' ability of the respective hydroxylase enzymes and the shared use of cofactors such as BH4, is influenced by certain types of gastrointestinal bacteria. This in turn might also explain some of the origins of IAG, derived from tryptophan.

I will hopefully not get too technical here but there are some important points raised in the paper which are perfectly testable in an autism research and other context.

So:

• The proposed model is based on two different rat urinary phenotypes distinguished by a favourite metabolomics method, NMR spectroscopy - one of the gold standards. One rat phenotype - the 'HIP' phenotype - produces rather a lot of urinary hippurate, whilst the other phenotype - the chlorogenic acid phenotype - is characterised by low levels of urinary hippurate and elevated levels of 3-(3-hydroxyphenyl)propionic acid (3-HPPA). The differences between the biochemistry of the two phenotypes is gut bacterial composition not readily attributed to factors like different diets.
• Proposals for the bacterial agents and biochemical pathways pertinent to these rat phenotypes is given with specific focus on the origin of the benzoic acid that consequently forms the hippurate levels seen in the HIP phenotype. Words like cinnamic acid and PAL are also included.
• Transposing the various biological reactions noted in phenylalanine and tyrosine on to tryptophan, Clayton moves through the production of IAG from tryptophan via a most interesting intermediary compound called indole-3-acrylic acid (IAA). IAA is thought to be quite a reactive molecule which subsequently conjugates with glycine to form IAG, also able to potentially do various things because of its flat planar chemical geometry. There is for example, some prior speculation that IAA may have some 'membrane-busting' potential in relation to autism as a result of its dose-dependent effect on prostaglandin (E) production, but this observation still remains preliminary. 

Acknowledging that this is a hypothesis paper and so still subject to some degree of testing, I am very curious about some of the concepts detailed. Elevated levels of 3-(3-hydroxyphenyl)-3-hydroxypropionic acid (HPHPA) have been reported in cases of autism, a compound which fits rather nicely into the whole phenylalanine-bacteria connection with particular focus on an old bacterial friend, Clostridia. This accompanied by various other related findings in cases of autism such as elevated levels of urinary 4-hydroxyhippuric acid detailed in this paper adds to the curiosity.

Speculation on some role for IAG and precursors in relation to autism and particularly some reported [non-significant] differences to levels according to use of a gluten-free diet, has again been something of interest to me for a while. Allowing for interfering variables based on urine sample concentration noted in some cases of autism, the question of why a gluten-free diet in particular, might affect urinary IAG levels has always been a puzzle. Does a gluten-free diet affect gut bacterial composition? Likely, according to this paper. Does the gluten-free diet affect gut permeability? Yes it probably does to some extent and the de Magistris results seem to back that up in cases of autism. Could gut bacteria contribute to gut permeability either directly or peripherally through the formation of IAG or related mechanisms? Mmm... I dunno.

There is a lot to take in with this recent paper and even after a few reads I still find myself not fully understanding all of the chemistry. Nevertheless an important connection has been posited between our gut bacterial masters and their potential effect on important aspects of our biochemistry. I would like to think that I will post further about some of the implications of this hypothesis as more research starts to move into this area.

To end, an oldie but goodie which was the soundtrack to a favourite film of mine. Cue Richard Dreyfuss.. "I never had friends later on like the ones I had when I was 12...".

* Clayton TA. Metabolic differences underlying two distinct rat urinary phenotypes, a putative role for gut microbial metabolism of phenylalanine and a possible connection to autism. FEBS Letters. February 2012.
DOI: 10.1016/j.febslet.2012.01.04