The prebiotic galactooligosaccharide (B-GOS) and autism: just add to poo(p)

Yes, it is childish but...

With all the continued chatter on a possible role for the collected gut microbiota - those wee beasties that inhabit our deepest, darkest recesses - in relation to some autism (see here for example), the paper by Roberta Grimaldi and colleagues [1] (open-access available here) provides yet more potentially important information.

So, poo(p) samples were the starring material in the paper - "obtained from three non-autistic children and three autistic child donors"- and specifically what happened when something called B-GOS "a prebiotic galactooligosaccharide" was added to samples following their journey through a "Three stage continuous culture gut model system" otherwise known as an artificial gut. Said gut model based at Reading University has already been the topic of other news (see here).

As well as looking at the initial bacterial profile of those stool samples, researchers plotted the changes to the stool's inhabitants (or what was left of the stool) over the course of B-GOS addition, as well as looking at things like the "production of SCFAs [short-chain fatty acids] in the fermentations" and other metabolites via the gold-standard chemical analytical technique called 1H-NMR (see here for more details).

Results: "Consistent with previous studies, the microbiota of ASD [autism spectrum disorder] children contained a higher number of Clostridium spp. and a lower number of bifidobacteria compared to non-autistic children." With the addition of B-GOS to the 'mixture', researchers reported on a significant increase in bifidobacterial populations at the different stages of their gut model and in samples from both those with autism and those without autism. Such "bifidogenic properties of B-GOS" are not unheard of.

As to the metabolites of those bacteria present in the poo(p) samples, there were some interesting knock-on effects noted in both raw and B-GOS supplemented samples. "Our data show a lower concentration of butyrate and propionate in autistic models, compared to non-autistic models, but no
differences in acetate before adding B-GOS into the system." Propionic acid (propionate) has some research history with autism in mind (see here). Butyric acid (butyrate) is something of a rising star in quite a few domains, having also been mentioned in the context of autism too (see here). Indeed it's interesting to note that B-GOS administration "mediated significant production of... butyrate... simulating the transverse and distal colon respectively. There was no effect on propionate." The findings of lower starting levels of butyrate in samples from children with autism were also substantiated by the NMR analyses undertaken. Increases in butyrate and changes to various other metabolites ("increasing ethanol, lactate, acetate and butyrate and decreasing propionate and trimethylamine") were also noted via this analytical method for this group.

A long quote coming up: "This in vitro study showed promising and positive results in that supplementing the microbiota of ASD children with 65%B-GOS may manipulate the gut bacterial population and alter metabolic activity towards a configuration that might represent a health benefit to the host. However, further work will be required to assess such changes in an in vivo human intervention study."

Just before anyone makes a run on B-GOS or any similar product however, I do need to stress a few important points. First, this was a study of poo(p) samples from 3 autistic children compared with samples from 3 non-autistic children. Aside from the small participant numbers, we don't know anything about participants' various comorbidities (although we know they were "free of any metabolic and gastrointestinal diseases") and only limited information on their dietary habits and medication history. Second, poo(p) was the target material included for analysis and what happened when B-GOS was supplemented during the journey through the artificial gut model. This study said nothing about what happens when real people with autism take B-GOS orally for example, and how it might affect gut bacterial populations and metabolites as it progresses down a real gastrointestinal (GI) tract. This also includes a lack of information on any potential side-effects in a real-world situation. We are also assuming that any supplement survives the stomach. There is quite a bit more to do in this area.

But for now, I stick to the idea that the Grimaldi paper provides some potentially important information and certainly, some new routes/methods for further study of the link between prebiotics, probiotics and synbiotics in the context of the gut microbiota and autism...

----------

[1] Grimaldi R. et al. In vitro fermentation of B-GOS: Impact on faecal bacterial populations and metabolic activity in autistic and non-autistic children. FEMS Microbiol Ecol. 2016 Nov 16. pii: fiw233.

----------

Grimaldi R, Cela D, Swann JR, Vulevic J, Gibson GR, Tzortzis G, & Costabile A (2016). In vitro fermentation of B-GOS: Impact on faecal bacterial populations and metabolic activity in autistic and non-autistic children. FEMS microbiology ecology PMID: 27856622

The continued rise of autism research metabolomics

For anyone that has followed this blog down the years you'll probably have noticed that I'm quite a big fan of the inclusion of the science of metabolomics on to the autism research menu (see here for example).

Looking at the myriad of chemical footprints left behind by an almost incomprehensible number of cellular processes, metabolomics offers some real promise to autism in terms of teasing apart phenotypes and as a valuable partner to other -omics sciences in ascertaining the relevance or not of specific biological pathways. All of this set within the context of the plural autisms and the important role of comorbidity (see here).

It is therefore with metabolomics again in mind that I bring to your attention the paper by Binta Dieme and colleagues [1] who weren't joking when they talk about a "multiplatform analytical methodology" with autism in mind. That multiplatform approach included "1H- and 1 H-13C-NMR-based approaches and LC-HRMS-based approaches (ESI+ and ESI- on a HILIC and C18 chromatography column)." If all that sounds like gibberish, the watchwords are NMR - Nuclear magnetic resonance spectroscopy - and LC-HRMS - Liquid chromatography–high resolution mass spectrometry - two of the gold-standard analytical techniques for detecting and identifying compounds of interest in this realm of biology. Some of the other details such as HILIC columns are all to do with how one goes about separating out the individual components of a complicated biological medium like urine as well as some further details about what the authors did to detect them. I might add that this authorship group have some previous form in this area of the autism research landscape (see here).

Based on the analysis of urine samples initially from 22 children with autism and 24 not-autism controls (a training group), researchers talked about the results they obtained from the various metabolomic approaches employed including processing of results by OPLS-DA (orthogonal partial least squares discriminant analysis). I don't want to bore you with the ins-and-outs of what OPLS-DA means (yeah, as if I know!) but suffice to say its all about how one classifies the multitude of data one generates via such analytical methods. This data and analyses were then used to generate a set of compounds (pattern of compounds) potentially predictive of whether or not it could classify a urine sample from someone with autism from a urine sample from someone without autism. Samples from a separate group of participants - "8 autistic children and 8 controls" - were used to 'test' the predictions generated. The authors report that the OPLS-DA model generated "showed an enhanced performance... compared to each analytical modality model, as well as a better predictive capacity (AUC=0.91, p-value 0.006)." AUC by the way, refers to area under the curve and is a term associated with a ROC (receiver operating characteristic). In this respect, the Dieme paper seemed to do pretty well at classifying samples according to autism or not-autism status bearing in mind the relatively small participant group numbers.

Just in case you're not confused enough, there are a few other details about the Dieme paper and findings that are worthy of comment. So: "Metabolites that are most significantly different between autistic and control children (p<0.05) are indoxyl sulfate, N-〈-Acetyl-L-arginine, methyl guanidine and phenylacetylglutamine." Indoxyl sulfate is a particularly interesting compound for quite a few reasons. Not only is the source material for this compound one of the those oh-so-interesting aromatic amino acids, tryptophan (y'know serotonin, melatonin and all that jazz) but the compound itself is described as a uremic toxin [2]. Without wishing to make connections where none may exist, uremic compounds in relation to autism have been discussed before on this blog as per the Elaine Hsiao findings on bacteria and leaky gut in a mouse model of autism (see here) and some chatter about p-cresol and autism (see here and see here). If there is an overlapping factor potentially uniting these findings, it would have to be a possible role for those trillions of wee beasties that call our gut home - the gut microbiome. I might also briefly mention the arginine finding too in relation to a related tryptophan observation for some autism... BH4 (see here).

As I mentioned at the start of this post I am a fan of this area of research area and its potential for furthering knowledge about autism. Larger datasets and perhaps a focus outside of just zooming in on the label of autism are perhaps elements that are needed to aid investigations in this area, alongside a more general combinatorial -omic initiative with a systems biology slant (see here).

Music and I've played this before but here it is again... Weapon Of Choice by Fatboy Slim (a favourite video of my brood).

----------

[1] Dieme B. et al. Metabolomics study of urine in autism spectrum disorders using a multiplatform analytical methodology. J Proteome Res. 2015 Nov 5.

[2] Vanholder R. et al. The uremic toxicity of indoxyl sulfate and p-cresyl sulfate: a systematic review. J Am Soc Nephrol. 2014 Sep;25(9):1897-907.

----------

Dieme B, Mavel S, Blasco H, Tripi G, Bonnet-Brilhault F, Malvy J, Bocca C, Andres CR, Nadal-Desbarats L, & Emond P (2015). Metabolomics study of urine in autism spectrum disorders using a multiplatform analytical methodology. Journal of proteome research PMID: 26538324

More on urinary metabolomics in autism research

The -omics. Y'know all those new-fangled disciplines which have sprung up to describe how sciences look at genes, bacteria, etc. We used to call it plain old scientific analysis, but now depending on what your sample medium or technology or your target species is, its been rebranded and repackaged as an -omic.

Shepherdess @ Wikipedia  

I've talked about a few of the -omics quite a bit on this blog and their relationship to systems biology; ranging from microbiomics (studying bacteria) to epigenomics (chemical modifications of the genome) to metallomics (metals affecting cellular functions). I've even invented a new -omic: psychobacteriomics.

Indeed in my other life I'm currently helping out on an article talking about one of the 'next big' -omic things: lipidomics. I'm just waiting for the science of Paulomics to emerge and discover just what makes me tick.

Anyhow, the reason for the -omics chatter is due to my stumbling across an interesting couple of papers from Patrick Emond and colleagues* and by the same authorship group, Sylvie Marcel and colleagues** and their description of findings based on the science of metabolomics. Aside from the metabolomics link (which ties into some of my own research interest or at least that of the people I work with) I was always going to be interested in these papers because of some of the authorship group on the papers and their work on a favourite autism assessment schedule of mine***.

So, metabolomics - think low molecular weight metabolites and where we look for them (blood, saliva, urine, CSF) - and how there may be good reason for looking at group differences and similarities across different peoples and different conditions. I might point out that this is not the first time urinary metabolomics - the sample medium described - has been discussed on this blog as per the Ming findings and Yap findings with autism in mind. I want to also direct you to the Yang findings on the appliance of metabolomics to schizophrenia which were really rather interesting and are still crying out for independent replication.

The Emond paper briefly:

• The tools of the trade were gas chromatography - mass spectrometry (GC-MS), which were applied to urine samples received from 26 children diagnosed with an autism spectrum disorder (ASD) and compared with 24 asymptomatic controls.
• Analysis of the samples was followed by some nifty statistics to help identify any potential discriminating metabolites between the groups and bingo, a few compounds of interest were reported on.
• Results: "The relative concentrations of the succinate and glycolate were higher" in the autism group. 
• But, "hippurate, 3-hydroxyphenylacetate, vanillylhydracrylate, 3-hydroxyhippurate, 4-hydroxyphenyl-2-hydroxyacetate, 1H-indole-3-acetate, phosphate, palmitate, stearate, and 3-methyladipate were lower" in the autism group.

The Mavel paper also:

• A slightly different analytical technique based on nuclear magnetic resonance spectroscopy (NMR) applied to urine samples from 30 children with ASD and 28 controls. Indeed quite a specific type of NMR was done (Heteronuclear Single Quantum Coherence, HSQC) which aids in metabolite identification. I'm not sure if the participant groups from this paper overlapped with that of the Emond paper or not.
• Similar statistics and modelling to that in the GC-MS paper were used on the data obtained and differentiation data were presented across the groups.
• Results: findings for a few compounds intersected those reported in the Emond paper (i.e. succinate reported to be present in higher quantities alongside β-alanine, glycine and taurine). Other compounds were reported as being lower: "creatine and 3-methylhistidine concentrations were lower in autistic children than in controls".
• More than that however were the details of the methodology used, and how 2D HSQC NMR might be applied to further larger studies in the autism research field. 

As a bit of a cop-out, I'm not going to go through each compound with a fine-toothed comb. OK, perhaps a little more explanation of some of those metabolites might be in order and in particular the role that gut bacteria may very well have had on them.

Hippurate for example, is an interesting metabolite given its proposed links to all things gut bacterial. Indeed I note on that very interesting paper by Andrew Clayton on gut bacteria and the aromatic amino acids potentially with autism in mind, hippurate and its precursor benzoic acid were mentioned in one of the rat models discussed. That the autism group seemed to show generally lower levels of hippuric acid (hippurate) might potentially imply some involvement of the amino acid glycine (which conjugates with benzoic acid) although I am of course, just speculating. That being said, the glycine finding in the Mavel paper (being higher) might suggest that there is more going on here.

I'll also just mention that the higher levels of urinary taurine were also noted in the Yap paper**** too. Oh and some wondering about whether lower urinary creatine levels might also be tied into the lower urinary creatinine levels that we reported on a few years back (paper is here in case your interested).

I note that in the Emond paper, the authors talk quite a bit about how you treat the urine samples prior to analysis might affect what results you get. Again, it's not something that I really want to get into. Indeed just before you completely switch off, although I know a little bit about sample prep when it comes to things like SPE, I freely admit that the process of oximation (for derivatisation) is a different language for me. I will however say that basing results of ion mass with only one (sometimes no) decimal place, is a short-coming as compared to the power of accurate mass via Time-of-Flight (ToF) spectrometry for example when it comes to authoritative compound assignment, but that's not a criticism.

I am genuinely intrigued over the potential of the -omics when it comes to conditions like autism (sorry, the autisms). The two French papers are another step into that brave new world bearing in mind that chromatographic methods***** also need to be allied with strong detection technology****** (open-access). Assuming science can also start to sort out some of those phenotypes which make up the autism spectrum, this area of work promises so much in terms of insights into pathology and the development of objective diagnostic markers.

----------

* Emond P. et al. GC-MS-based urine metabolic profiling of autism spectrum disorders. Anal Bioanal Chem. April 2013.

** Mavel S. et al. 1H–13C NMR-based urine metabolic profiling in autism spectrum disorders. Talanta. 2013; 114: 95-102.

*** Barthélémy C. et al. Validation of the Revised Behavior Summarized Evaluation Scale. J Autism Dev Disord. 1997; 27: 139-53.

**** Yap IK. et al. Urinary metabolic phenotyping differentiates children with autism from their unaffected siblings and age-matched controls. J Proteome Res. 2010; 9: 2996-3004.

***** Zurawicz E. et al. Chromatographic methods in the study of autism. Biomed Chromatogr. April 2013.

****** Wood AG. et al. Mass spectrometry as a tool for studying autism spectrum disorder. Journal of Molecular Psychiatry 2013; 1: 6.

----------

Emond P, Mavel S, Aïdoud N, Nadal-Desbarats L, Montigny F, Bonnet-Brilhault F, Barthélémy C, Merten M, Sarda P, Laumonnier F, Vourc'h P, Blasco H, & Andres CR (2013). GC-MS-based urine metabolic profiling of autism spectrum disorders. Analytical and bioanalytical chemistry PMID: 23571465



Mavel, S., Nadal-Desbarats, L., Blasco, H., Bonnet-Brilhault, F., Barthélémy, C., Montigny, F., Sarda, P., Laumonnier, F., Vourc′h, P., Andres, C., & Emond, P. (2013). 1H–13C NMR-based urine metabolic profiling in autism spectrum disorders Talanta, 114, 95-102 DOI: 10.1016/j.talanta.2013.03.064