Gut microflora, coeliac disease and introducing gluten

A new paper by Sellitto and colleagues* (full-text) has been causing quite a bit of interest in certain circles. The paper as the name suggests is [partially] a 'proof of concept' study which includes several topics of interest for this blog with its focus on gut microflora (and dysbiosis), coeliac (celiac) disease (CD) and some interesting metabolomics science. The added value comes with the name Alessio Fasano as part of the authorship list.

There is quite a bit of details to this paper but in essence the aims were: (i) to characterise the changes from birth to 24 months in terms of gut bacteria to genetically at-risk of coeliac disease children as a function of early or delayed introduction of gluten to the diet, and (ii) to undertake a range of analytical methods to map such bacterial populations with the hope of further informing on any relationship between gut bacteria and coeliac disease.

The paper is full-text but a quick summary of proceedings and findings:

• Forty-seven infants who had one parent with biopsy-proven CD were initially recruited before weaning had commenced. All were breastfed from birth to at least 6 months of age. From 6-12 months of age, 30 infants positive for either the HLA DQ2 and/or HLA DQ8 genotypes were randomly allocated to either a gluten-free - delayed gluten exposure - group (n=13) or a gluten load - early exposure - group (n=17).
• A smaller number of children from each group (n=8 each) were selected randomly for the analytical side of the study (which is what this paper in essence reports) where stool samples were collected at various points over the study period ranging from 7 days in to 24 months.
• The results: none of the 8 infants from the delayed gluten exposure group developed CD over the course of the study. One of the 8 infants in the early gluten introduction group did go on to develop CD at 2 years of age as measured by various serological panels and went on to a gluten-free diet with a remission of serology at follow-up.
• When levels of anti-gliadin antibodies (AGA) (IgG) were examined and corrected for gluten exposure time, the early gluten introduction group showed a higher number of IgG-AGA positive results than the delayed exposure group. The authors discuss how AGA is not necessarily a particularly good measure of CD but could indicate greater levels of intestinal hyperpermeability (leaky gut) as a result of exposure to the gliadin fraction of the gluten protein in the same way that IgG levels have been interpreted in other studies (see Sutterella and autism post). 
• Pyrosequencing of the various species and families of bacteria present across the groups at different time frames suggested some interesting goings-on. To quote: "the GI tract microbiota in DQ2+/DQ8+ infants appears to be lacking significant numbers of member of the phylum Bacteroidetes". That and a higher abundance of Firmicutes, implies that maturationally, the gut microflora of children at elevated risk of CD is different from lower risk groups as determined by comparison with an external dataset** (full-text).
• The application of 1H-Nuclear Magnetic Resonance Spectroscopy (NMR) to proceedings added that metabolomic touch as "..SCFA succinate, acetate, propionate and butyrate are found in the feces" following the introduction of solid foods. Having said that little distinguishing data was found to categorise the two groups in any significant, universal way.

It's taken me a while to get my head around all the findings from this recent paper because there was a lot of data produced bearing in mind the small participant group and preliminary status of the paper. One of the first things that did strike me is the overlap in these findings and some fairly recent data published looking at carbohydrate metabolism and autism included in this post. Decreasing Bacteriodetes and increasing Firmicutes was the preliminary finding from Williams and colleagues*** bearing in mind the difference in samples being analysed and the lack of data on HLA DQ2 / HLA DQ8 genotype provided in the group with autism being studied. I'm not going to say too much more of this 'similarity' aside from the fact that screening for coeliac disease and/or excessive intestinal permeability perhaps ought to be much more commonly undertaken in cases of autism spectrum conditions just to rule them out.

Quite a few mentions of the word 'dysbiosis' are also recorded in the current paper which reaffirms the possibility of a connection between the various populations of bacteria that reside within us and our potential risk of disease. I like the idea that this study looked at both the metabolomic and genomic side of things even if it was just based on the HLA DQ geneotypes.

So from the starting point of a genetic susceptibility to gluten, we have preliminary data on functional changes to the microbiome in susceptible people and some interesting tools for looking at how this might be expressed functionally. I look forward to more studies of this type with greater participant numbers, and in particular how such findings might extend into other autoimmune conditions and even beyond just somatic presentation.

* Sellitto M. et al. Proof of concept of microbiome-metabolome analysis and delayed gluten exposure on celiac disease autoimmunity in genetically at-risk infants. PLoS ONE. March 2012
DOI: 10.1371/journal.pone.0033387

** Palmer C. et al. Development of the human infant intestinal microbiota. PLoS Biology. 2007; 5:e177
DOI: 10.1371/journal.pbio.0050177

*** Williams B. et al. Impaired carbohydrate digestion and transport and mucosal dysbiosis in the intestines of children with autism and gastrointestinal disturbances. PLoS ONE. September 2011.
DOI: 10.1371/journal.pone.0024585