Biochemical Society Transactions

The Molecular Biology of Inflammatory Bowel Diseases

Nutritional influences on the gut microbiota and the consequences for gastrointestinal health

Karen P. Scott, Sylvia H. Duncan, Petra Louis, Harry J. Flint


The human colonic microbiota degrades dietary substrates that are indigestible in the upper GIT (gastrointestinal tract), releasing bacterial metabolites, some of which are important for gut health. Advances in molecular biology techniques have facilitated detailed analyses of the composition of the bacterial community resident in the lower GIT. Such analyses have indicated that more than 500 different bacterial species colonize an individual, and that, although there is much functional consistency in the resident bacterial groups, there is considerable inter-individual variation at the species/strain level. The bacterial community develops during early childhood until it reaches an adult-like composition. Whereas colonization and host factors influence the species composition, dietary factors also have an important impact, with specific bacterial groups changing in response to specific dietary interventions. Since bacterial species have different metabolic activities, specific diets have various consequences for health, dependent on the effect exerted on the bacterial population.

  • anaerobe
  • dietary carbohydrate
  • gut microbiota
  • human colon
  • resistant starch
  • short-chain fatty acid (SCFA)

Introduction: the gut microbiota

The human GIT (gastrointestinal tract) is the most densely colonized niche in the human body, with approximately 1014 resident bacteria along its length, with the greatest density and diversity present in the large intestine, or colon [1]. Conditions become increasingly anaerobic descending the GIT, and the majority of the bacterial species resident in the large intestine are obligate anaerobes, unable to survive even short exposure to air [1]. The composition of the gut microbiota is partially defined by host genetics and bacteria acquired at birth, although diet has an enormous influence in shaping the subsequent development. The gut microbiota composition changes from infancy when it is dominated by bifidobacteria [2], through childhood and adulthood, and again during old age [3,4]. The most dramatic change occurs around weaning, when the introduction of alternative carbohydrate-based substrates enables the expansion of the microbiota. The adult microbiota is relatively stable until old age, whereupon changes in the digestive process (longer transit time and reduced digestive secretions) affect the microbial composition characterized by an increased incidence of Escherichia coli and a reduced Firmicutes/Bacteroidetes ratio [3].

Gut microbiota in health and disease

More than 500 different bacterial species reside in the human large intestine, and advances in molecular methods to profile the gut microbiota reveal considerable inter- and intra-individual variation. The two most abundant phyla found in most healthy individuals are the Bacteroidetes and Firmicutes [5,6]. Predominant Firmicutes include Faecalibacterium prausnitzii, Eubacterium rectale, Eubacterium hallii and Anaerostipes coli [6], all key butyrate-producing bacteria [7]. Multiple representatives of this functional group, although not always the same individual species, are always found in healthy adults [8]. The existence of a dominant group of bacteria presumably acts to conserve the metabolic functions essential to gut health, and a functional gene core has also been identified in the gut metagenome [9].

The most abundant bacterial groups are readily quantified using high-throughput sequencing approaches, but different methods are required to enumerate low-abundance bacteria that contribute to the genetic and functional diversity. Using a combination of the molecular FISH (fluorescence in situ hybridization) detection method and bacterial cultivation on specific substrates, the major hydrolytic communities and the different groups of hydrogen-utilizing bacteria were enumerated in faecal samples from different individuals [10]. The latter bacteria are essential for gut homoeostasis, but are undetectable in many molecular surveys because of their low abundance.

Many diseases are associated with inflammation, and gut inflammation may be due to a dysregulation of the immune response to the gut microbiota. Consequently, bacterial community analyses have been undertaken to try to establish whether changes in the bacterial communities can be linked to disease states. Crohn's disease is characterized by serious inflammation of the gut epithelium, and studies have shown that patients with Crohn's disease have an altered gut microbiota and a slightly reduced bacterial diversity compared with healthy individuals [11]. Low-grade inflammation of the gut has been observed in elderly individuals [12], and a general reduction in bacterial diversity in elderly individuals has been observed [13], although considerable inter-individual variation has also been reported [4]. IBS (irritable bowel syndrome) is clinically defined as a non-inflammatory bowel disease, and the disease phenotypes are so diverse that it is difficult to correlate information on the composition of the microbiota collected in different studies. However, one study showed that the diversity of the faecal microbiota was reduced in females with IBS [14].

It is difficult to ascertain whether such changes in the gut microbial community actually cause gut diseases, or whether they are a consequence of the disease state (e.g. the increased permeability of the gut epithelial barrier due to gut inflammation). Whereas particular bacterial species have been linked to gut disease phenotypes, e.g. numbers of adherent invasive E. coli are increased [15] and those of F. prausnitzii are decreased [16] in IBD (inflammatory bowel disease), it is not established whether the changes in bacterial profile are a cause or an effect of the disease.

Impact of diet on the composition and metabolism of the gut microbiota

Plant cell wall polysaccharides such as cellulose, xylans and pectins, and storage polysaccharides such as inulin and resistant starch are important for growth of bacteria resident in the large intestine. In addition, certain oligosaccharides, disaccharides and sugar alcohols show limited hydrolysis and absorption in the small intestine. A subset of these NDCs (non-digestible carbohydrates) are classed as prebiotics, defined as “selectively fermented food ingredients that result in specific changes in the composition and/or activity of the gastrointestinal microbiota with consequent benefits for host health” [17].

Although nutritionists often employ the blanket term ‘fibre’, it is increasingly apparent that the effects of NDC on the gut microbiota depend critically on chemical structure and physical state of the carbohydrate entering the colon. In particular, there is evidence that different gut micro-organisms specialize in the degradation of insoluble substrate particles compared with soluble carbohydrates.

Resistant starch

Previous work on cultured bacteria has emphasized the amylolytic ability of Bacteroides spp. and certain Bifidobacterium spp. (reviewed by Flint et al. [17a]), whereas recent work in vivo has established the importance of Firmicutes in starch utilization [1821]. A human dietary intervention study compared the effects of diets supplemented either with a type 3 resistant starch or a source of non-starch polysaccharide (wheat bran) in obese subjects [6]. The impact of the dietary change on the gut microbiota was rapid, with the bacterial profile in human volunteers changing within 3 days. The two bacterial groups showing the most significant responses to resistant starch were Firmicutes related to Ruminococcus bromii and E. rectale (Figure 1), with bifidobacteria responding strongly in only one out of 14 individuals [6]. In a separate study, decreasing carbohydrate intake significantly decreased both the detectable numbers of the E. rectale/Roseburia group and butyrate production [18]. This strong positive correlation between numbers of the E. rectale/Roseburia group, butyrate detection and carbohydrate intake (Figure 2) is supported by other human studies [19] and in vitro studies [20]. Studies in model simulated colonic fermenter systems indicated that 80% of sequences that adhered to insoluble starch particles belonged to R. bromii, E. rectale and Bifidobacterium spp. [20]. The same groups of bacteria were detected using stable isotope probing with [13C]starch [21]. These findings suggest that Bacteroides spp. may not adhere to insoluble starch particles, and are perhaps better equipped to exploit solubilized starch molecules.

Figure 1 Effect of resistant starch consumption on the abundance of ruminococci and E. rectale/Roseburia groups

Fourteen obese male volunteers (body mass index >30) were put on maintenance (M), non-starch polysaccharide (NSP), resistant starch (RS) or weight loss (WL) diets for fixed time periods. Specific bacterial groups within the faecal microbiota were quantified using qRT-PCR (quantitative real-time PCR) with targeted primers. The abundance of the ruminococci and E. rectale/Roseburia groups is expressed as a percentage of the signal obtained with the general bacterial primer set. For full experimental details, see [6]. * indicates values that are significantly different (P<0.05, post-hoc test).

Figure 2 Relationship between concentration of butyrate and E. rectale/Roseburia populations in faeces

Eighteen obese male volunteers (body mass index >30) were put on high-carbohydrate (■), moderate-carbohydrate (△) and low-carbohydrate (◆) diets. Following each dietary period, faecal butyrate concentration was measured and plotted against the E. rectale/Roseburia population within that sample, enumerated using FISH. Correlation r=0.74. Reprinted from [18], Applied and Environmental Microbiology, 2007, vol. 73, pp. 1073–1078, doi:101128/AEM.02340-06 with permission from American Society for Microbiology.

Cellulose and hemicellulose

Although the human gut microbiota do not appear to degrade the more crystalline forms of cellulose, it is estimated that up to 70% of cellulose and hemicellulose present in normal foodstuffs are fermented during passage through the large intestine [22]. The bacteria involved apparently include both Gram-positive Firmicutes and Gram-negative Bacteroides spp. [23]. Interestingly, a Ruminococcus spp. of human origin, related to rumen cellulolytic species such as Ruminococcus albus and Ruminococcus flavefaciens, has been reported to degrade cellulose [24]. The clostridial cluster IV ruminococci were significantly enriched in the fibre fraction compared with the liquid phase of human faecal samples [25]. It has been proposed that the dominant types of fibrolytic bacteria present in an individual may depend upon the nature of the hydrogen-utilizing flora present (e.g. differ between methanogenic and non-methanogenic individuals) [24].


Research into prebiotics so far has concentrated on indigestible oligosaccharides, with particular emphasis on inulin-type FOS (fructo-oligosaccharides) [26]. Initial studies focused on the stimulation of bifidobacteria, which are generally regarded as beneficial microbes in the gut, and there are numerous studies documenting increases in the abundance of bifidobacteria in response to FOS supplementation. However, since it is now clear that bifidobacteria constitute a relatively small proportion of the gut microbiota, it is important to consider the whole community rather than a single targeted species when assessing the effect of prebiotics. Two groups of butyrate-producing bacteria increased following inulin/FOS supplementation in a human intervention study, F. prausnitzii [27] and a group of clostridial cluster XIVa bacteria related to strain SS2/1 (proposed new species A. coli) [8], as well as bifidobacteria. Boosting butyrate production would improve gut health, while F. prausnitzii has additional potential anti-inflammatory effects [28]. Pure culture work indicates that several bacteria belonging to the Bacteroidetes and Firmicutes phyla are also able to degrade inulin or FOSs [29,30]. Inclusion of FOSs in the diet prevented the elimination of bifidobacteria in conventional mice fed on a high-fat diet [31].

The prebiotics GOS (galacto-oligosaccharides) and lactulose have not been investigated so extensively, but limited studies demonstrated stimulation of bifidobacteria, specifically Bifidobacterium adolescentis, with GOS [32], and also a reduction in the numbers of potential pathogenic bacteria including enterobacteria and Clostridium perfringens with GOS and lactulose (reviewed in [33]). New potential prebiotics include malto-oligosaccharides and xylo-oligosaccharides. Studies in a model system indicated that arabinoxylan oligosaccharides were mainly degraded in the transverse colon, with an associated increase in propionate production [34].

Role of bacterial metabolites in the gut

Bacterial fermentation of the dietary fibre, carbohydrates and prebiotics that escape digestion in the upper intestinal tract forms a range of bacterial metabolites, including SCFAs (short-chain fatty acids) and gases. SCFA concentrations detected in faeces can exceed 100 mM, and it has been estimated that 90% of SCFA are absorbed across the colonic wall [35]. The molar proportions of the three most abundant SCFAs routinely range between 3:1:1 and 10:2:1 for acetate, propionate and butyrate respectively. Acetate is mainly metabolized in the peripheral tissues and is lipogenic, whereas propionate is transported to the liver and is gluconeogenic. Butyrate is the preferred energy source for the colonocytes and has a role in regulating host gene expression [7,36]. Extensive metabolic cross-feeding occurs between the primary degraders of complex substrates and other bacterial species that metabolize the first set of products, forming others (Figure 3). Hydrogenotrophic bacteria utilize the CO2, H2 and SO42− released during bacterial fermentation, and the balance of these methanogenic, acetogenic and sulfate-reducing bacteria in an individual determine the end-products CH4, acetate and H2S respectively. In addition, host mucus and secreted enzymes and approximately 10 g of dietary protein per day reaches the colon. Proteins are fermented in the large intestine by Bacteroides spp. and clostridia [37] to a range of products depending on the amino acid composition; for example, branched-chain amino acids yield branched-chain fatty acids, including isobutyrate and isovalerate.

Figure 3 Major fermentation products in the large intestine

Schematic diagram illustrating bacterial fermentation of polysaccharides, showing the intermediate and final metabolites (boxed). The types of bacteria performing the key reactions are shown in italics.

Bacterial fermentation of dietary substrates in the colon influences the health of the host, with dietary changes affecting the proportions of SCFAs detected. High-protein low-carbohydrate diets reduce faecal butyrate levels approximately 4-fold [18], increase the proportion of branched-chain fatty acids and nitrosamines and alter the profile of microbially generated phenolic metabolites [38]. Active fermentation of dietary carbohydrates decreases the pH in the proximal colon, influencing microbial composition since different phylogenetic and functional groups of gut bacteria differ in their pH-tolerance. In the presence of typical colonic concentrations of SCFAs, growth of human colonic Bacteroides spp. was strongly inhibited at pH 5.5, whereas many Firmicutes, including the most abundant butyrate producers, were less affected. In the mixed community, in a continuous-flow fermenter model system held at pH 5.5 compared with pH 6.5, butyrate production was stimulated and Bacteroides populations were curtailed [39]. An inverse correlation between butyrate production and colonic pH is also observed in vivo [40]. Colonic pH may also influence other microbial functions, such as methanogenesis and proteolysis [37].

In healthy individuals, lactate is virtually undetectable in faecal samples (<5 mM), yet lactate is formed by many species in the large intestine, including bifidobacteria. In healthy individuals, lactate serves as a substrate for lactate-utilizing butyrate-producing bacteria such as E. hallii [41], and up to 20% of butyrate formation is estimated to be derived from lactate [42]. However, colonic lactate concentrations can increase to more than 20 mM in the disease state [43]. Sulfate-reducing bacteria, including Desulfovibrio spp., compete effectively for lactate in the colon which is oxidized to acetate concomitant with the formation of H2S [44]. This can inhibit butyrate metabolism by the colonocytes, therefore dietary products that increase lactate formation in the colon may also increase sulfide formation.

Effect of dietary changes on bacterial gene expression

Genomic approaches have identified specific bacterial genes that are required for survival in the human gut, which frequently include those involved in substrate degradation. Many commensal gut bacteria have genomes that contain above average numbers of polysaccharide-degradation genes, presumably permitting growth on a diverse range of substrates. Bacteroides spp., for example, have a vast number of polysaccharide-utilizing genes, frequently arranged in operons called PULs (polysaccharide-utilization loci) [45]. Comparison of the genomes of five Bacteroides spp. indicated that the genetic composition of the PULs involved in fructan utilization determined which fructans the Bacteroides spp. were able to utilize for growth [29]. The fructan PULs in Bacteroides thetaiotaomicron is up-regulated in response to both fructose and fructose-containing polysaccharides [29]. Bifidobacterium longum contains a large number of genes involved in carbohydrate metabolism [46], collectively known as the glycobiome, but also encodes proteins facilitating survival in the GIT. Similar ‘gut-specific’ genes absent from other bifidobacteria were present in unrelated gut commensal bacteria [47].

Survival within the gut depends on both the ability to use the substrates available and the ability to adapt rapidly to changes in substrate availability. Whereas metagenomic analysis can provide information on the substrates that bacteria can metabolize, only gene expression analysis can indicate which they are actually using. Roseburia inulinivorans can utilize starch and inulin for growth, expressing the required set of genes dependent on the growth conditions [30]. During growth on fucose, an operon of genes similar to those found in Salmonella enterica serotype Typhimurium is induced, with the simultaneous repression of genes involved in glucose metabolism [48]. The metabolic end-products are butyrate and propanol, but also significant amounts of propionate, which is not normally produced by this bacterium. The dramatic metabolic switch would enable the bacterium to survive periods of dietary starvation by using the fucose present on epithelial glycoconjugates for growth.

Co-colonization of gnotobiotic mice with B. thetaiotaomicron and E. rectale provides a valuable insight into changes in bacterial gene expression that may occur due to specific bacterial interactions. B. thetaiotaomicron specifically up-regulated the expression of various PULs, expanding its substrate-utilization capabilities to include host-derived glycans [49]. In contrast, E. rectale down-regulated expression of many glycoside hydrolases, and increased the expression of peptide and amino acid transporters to maximize the use of alternative nutrients. Butyrate-production genes were also highly expressed, with the acetate produced by B. thetaiotaomicron converted into butyrate, which was efficiently absorbed by the mice following up-regulation of a specific transporter. When the diet was adjusted to a high-fat diet, detection of E. rectale and butyrate decreased 5-fold [49]. In contrast, B. thetaiotaomicron persisted well, but the expression of PULs involved in plant cell wall polysaccharide degradation was decreased, while genes involved in utilizing host glycans were up-regulated.

General conclusions

The gut microbiota constitutes an important ‘organ’ in the human body, with bacteria degrading dietary components releasing metabolites that can be either beneficial or detrimental to health. Bacterial cross-feeding is a key component of this interaction, defining the true end-products of bacterial metabolism. The gut microbiota has a considerable adaptive capacity, both in terms of composition and function, to respond to both dietary changes and to the presence of other specific bacteria, and to revert to baseline upon the removal of any transient pressure.


The Rowett Institute of Nutrition and Health, University of Aberdeen receives support from the Scottish Government Rural and Environmental Research and Analysis Directorate (SG-RERAD).


  • The Molecular Biology of Inflammatory Bowel Diseases: A Biochemical Society Focused Meeting held at University of Durham, Durham, U.K., 20–22 March 2011. Organized and Edited by Tony Corfield (Bristol, U.K.), Chris Probert (Bristol, U.K.) and Heather M. Wallace (Aberdeen, U.K.).

Abbreviations: FISH, fluorescence in situ hybridization; FOS, fructo-oligosaccharides; GIT, gastrointestinal tract; GOS, galacto-oligosaccharides; IBS, irritable bowel syndrome; NDC, non-digestible carbohydrate; PUL, polysaccharide-utilization locus; SCFA, short-chain fatty acid


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 17a.
  19. 18.
  20. 19.
  21. 20.
  22. 21.
  23. 22.
  24. 23.
  25. 24.
  26. 25.
  27. 26.
  28. 27.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
  34. 33.
  35. 34.
  36. 35.
  37. 36.
  38. 37.
  39. 38.
  40. 39.
  41. 40.
  42. 41.
  43. 42.
  44. 43.
  45. 44.
  46. 45.
  47. 46.
  48. 47.
  49. 48.
  50. 49.
View Abstract