Background Dietary intervention with prebiotics can cause changes in the colonic microbiota and their metabolic activities.
Aim To investigate whether the response to prebiotic dosing is influenced by the baseline metabolic activity of the colonic flora and bifidobacteria counts.
Methods The 4-week effect of lactulose (10 g bid.; n = 29) and oligofructose-enriched inulin (10 g bid.; n = 19) was evaluated in healthy human volunteers. Lactose-[15N, 15N]-ureide was used to study the colonic NH3-metabolism. Urine (48 h) and faeces (72 h) were collected and analysed for p-cresol and 15N-content by gas chromatography–mass spectrometry and isotope ratio mass spectrometer, respectively. Faecal bifidobacteria were quantified by real-time polymerase chain reaction.
Results After the 4-week prebiotic administration period, the urinary excretion of p-cresol and 15N was significantly decreased in both groups (P < 0.05) corresponding to a significantly higher faecal excretion of 15N (P < 0.05). The decrease in urinary 15N and p-cresol excretion significantly correlated with baseline 15N and p-cresol levels (P < 0.05), indicating that subjects with higher baseline levels showed a higher response to prebiotic dosing. Furthermore, a significant correlation was seen between baseline bifidobacteria counts and the effect of prebiotic intake (P < 0.05).
Conclusion The response to prebiotic dosing, as indicated by the fate of NH3, p-cresol and bifidobacteria, is determined by the initial colonic conditions.
A major metabolic function of the colonic microflora consists of the fermentation of substrates of dietary origin, which have not been absorbed in the upper gastrointestinal tract (resistant starches, dietary fibre, sugars, oligosaccharides, proteins, peptides and amino acids) and endogenous sources such as pancreatic enzymes and mucus produced by the epithelia. Overall outcomes of this complex metabolic activity are recovery of metabolic energy and absorbable substrates for the host, and supply of energy and nutritive products for bacterial growth and proliferation.1 The fact that diet is a major factor controlling the intestinal balance has triggered the development of prebiotic substrates.
Dietary intervention with prebiotics has been shown to stimulate selectively growth and/or activity of one or a restricted number of bacteria in the colon, such as bifidobacteria and lactobacilli.2–4 Several nondigestible food ingredients have been studied as potential prebiotics including lactulose, galacto-oligosaccharides, fructo-oligosaccharides (oligofructose and inulin) and xylo-oligosaccharides.5 The prebiotic properties of the carbohydrates are likely to be influenced by the monosaccharide composition (prebiotics are primarily built up from glucose, fructose, galactose and xylose), the glycosidic linkage between the monosaccharide residues (crucial factor in determining both selectivity of fermentation and digestibility in the small intestine) and the degree of polymerization (DP) of the prebiotic.6 The bifidogenic capacities, i.e. the ability to stimulate the growth of bifidobacteria, of short-chain carbohydrates (DP between 2 and 10) appear to be almost an order of magnitude higher than that of high polymer substrates (DP > 10).7, 8 Besides the structure–function relationship of the prebiotics, the effect of these substrates on the relative populations of colonic bacteria has also been shown to depend on the initial colonic microbial condition of the host. Upon modulation of the colonic microbiota, it has been demonstrated that a relative increase in faecal bifidobacteria in response to a prebiotic was highly dependent on the initial bacterial count, i.e. a more pronounced effect of prebiotic dosing was observed with low starting counts.9, 10
In recent years, a number of studies have investigated the ability of prebiotics to increase the saccharolytic activity and to decrease the proteolytic activity of the colonic microbiota.2, 11–14 Fermentation of carbohydrates results in the formation of short-chain fatty acids (SCFA), which are considered to be beneficial, whereas fermentation of proteinaceous material is considered to be rather detrimental to the host since potentially toxic metabolites such as phenols, indoles, amines and ammonia can be formed.15, 16 The ratio between carbohydrates and proteins has been reported to be an important factor influencing the fermentation processes in the colon.
The extent of urinary and/or faecal excretion of ammonia and p-cresol has been shown to give an indication of the degree of proteolytic colonic fermentation, which is considered to be relevant to colonic health.17 Previously, we developed two biomarkers, respectively, lactose-[15N, 15N]-ureide and p-cresol, to estimate in vivo the impact of prebiotic administration on the colonic metabolism by quantification the formation of the potentially toxic metabolites, ammonia and p-cresol, in the colon.18, 19 Lactose-[15N, 15N]-ureide can be considered as a precursor of labelled NH3 and the ratio of faecal vs. urinary excretion of the label reflects the stimulation of colonic bacteria. Indeed, stimulation of bacterial growth and/or activity caused by prebiotic administration results in a higher uptake of ammonia by the bacteria (and hence higher faecal excretion) corresponding to a lower excretion of 15N in urine.20–22 p-Cresol, on the other hand, is a unique bacterial metabolite of tyrosine, which is not formed by human enzymes and is either excreted in the faeces or absorbed through the colonic wall and, after sulphate or glucuronide conjugation in the mucosa or liver, urinary excreted. It was demonstrated that the urinary excretion of p-cresol adequately reflects the influence of dietary intervention on amino acid fermentation in the colon.18, 23
In this study, it was investigated whether the effect of prebiotic dosing on colonic metabolism is influenced by the baseline activity of the microbiota and whether changes in colonic metabolism correlate with changes in the bifidobacteria numbers. For this purpose, two commercially available prebiotic substrates [lactulose and oligofructose-enriched inulin (OF-IN)] were administered for 4 weeks to healthy volunteers.
Materials and methods
Fifty healthy volunteers (26 females and 24 males, age range: 19–26) participated in the study. None of the subjects had a history of gastrointestinal or metabolic disease or previous surgery (apart from appendectomy) nor had used antibiotics or any other medical treatment influencing gut transit or intestinal flora during the preceding 3 months. The Ethics Committee of the University of Leuven approved the study protocol and written informed consent was obtained from all participants.
Two independent studies were performed in which two independent groups of volunteers received a prebiotic substrate for 4 weeks. The subjects of group 1 (n = 30) received 10 g lactulose bis in diem, whereas the subjects of group 2 (n = 20) received 10 g OF-IN bis in diem. The intervention periods were performed in a single-blind design. Throughout the study, the subjects were asked to keep their usual dietary habits, taking care that the diet remained as stable as possible over the study period. In addition, they were advised to avoid intake of fermented milk products and food components containing high quantities of fermentable carbohydrates. Immediately before the start of the study (baseline) and after the intervention period, the volunteers consumed a test meal containing lactose-[15N, 15N]-ureide as biomarker.
Lactulose (Duphalac; Solvay Pharma & Cie, Brussels, Belgium) and OF-IN (Synergy1; Orafti, Tienen, Belgium) were chosen as prebiotic substrates.24–27 OF-IN is a 1:1 (w/w) mixture of oligofructose (DP 2–8; DPav = 4) and long-chain inulin known as inulin HP (DP 10–60; DPav = 25).
The test meal consisted of a pancake [8.4 g proteins, 11.2 g fat and 26.7 g carbohydrates (244 kcal)], which contained 75 mg lactose-[15N, 15N]-ureide. The latter substrate was synthesized according to the method of Schoorl28 as modified by Hofmann29 with [15N, 15N]-urea obtained from Euriso-top (St Aubin, Cédex, France). Absence of remaining [15N, 15N]-urea or lactose was confirmed using thin layer chromatography.30 185 kBq of [3H]-polyethylene glycol ([3H]-PEG; NEN Life Science Products Inc., Boston, MA, USA) was added to the test meal as an inert radio-labelled transit marker.
Urine samples were always collected in dedicated receptacles to which neomycin was added for prevention of bacterial growth. After an overnight fast, a basal urine sample was collected before consumption of the test meal. After intake of the test meal, a 48-h urine collection was performed. After measurement of the volume, samples were stored at −20 °C until analysis.
All stool samples were collected for 72 h by 21 of 30 volunteers of group 1 and by all volunteers of group 2. Upon delivery of the faecal samples at the lab, a 5-g (wet weight) aliquot was taken for DNA extraction and immediately frozen at −20 °C. All stools collected on the same day were combined, weighed and homogenized before further analysis. Samples of known weight were removed and freeze-dried. Aliquots of the dried material were used for subsequent analysis of total nitrogen and 15N, and 3H-PEG content.
Determination of urinary p-cresol content. The p-cresol content in urine was determined by gas chromatography–mass spectrometry (GC–MS) technology, as described earlier.31 Urine samples at a volume of 950 μL were taken and the pH of the samples was adjusted to pH 1 with concentrated H2SO4 (Merck KgaA, Darmstadt, Germany). This solution was heated for 30 min at 90 °C to deproteinize and hydrolyse the conjugated phenols. After cooling down to room temperature, 50 μL of 2,6-dimethylphenol (20 mg/ 100 mL; Sigma-Aldrich Chemie GmbH, Steinheim, Germany) was added as internal standard. p-Cresol was extracted with 1 mL of ethyl acetate (Merck KgaA). The ethyl acetate layer was dried and 0.5 μL of this solution was analysed on a GC–MS (Trace GC–MS, Thermofinnigan, San José, CA, USA). The analytical column was a 30 m × 0.32 mm i.d., 1 μm AT5-MS (Alltech Associates, Deerfield, IL, USA). Helium GC grade was used as carrier gas with a constant flow of 1.3 mL/min. The oven was programmed from 75 °C (isothermal for 5 min), with 10 °C/min to 160 °C and to 280 °C with 20 °C/min. Mass spectrometric detection (single quadrupole) was performed in E.I. full scan mode from m/z 59 to m/z 590 at two scans per second. The results were expressed as total p-cresol content (mg) excreted in urine.
Determination of total nitrogen content and 15N-enrichment in urine and faeces. Total N content and 15N-enrichment were determined by a continuous flow elemental analyser isotope ratio mass spectrometer (ANCA-2020; Europa Scientific, Crewe, UK). Therefore, a known amount of urine (15 μL, adsorbed on chromosorb; Elemental Microanalysis Limited, Devon, UK), faeces (freeze-dried, 5–7 mg) were introduced in the oxidation–reduction module coupled to the isotope ratio mass spectrometer. In this module, the samples were oxidized using copper oxide, oxygen and chromium oxide to nitrous oxides (NOx) at 1000 °C and subsequently reduced to nitrogen directed (N2) using copper at 600 °C. This gas was directed to the ion source of the mass spectrometer where the total nitrogen content and the 15N-enrichment were measured. The 15N to 14N isotope ratio of N2 was measured with reference to a calibrated laboratory standard (i.e. a standard ammonium sulphate solution). The results for the urinary and faecal samples were expressed as total nitrogen content and as percentage of the administered dose of 15N.18
Determination of faecal [3H]-PEG. The [3H]-PEG content in faecal samples was measured with liquid-scintillation counting (Packard Tricarb Liquid Scintillation Spectrometer, model 3375; Packard Instruments Inc., Downers Grove, IL, USA) after oxidation to [3H]-H2O (Packard Sample Oxidiser, model 306; Packard Instruments Inc.). The tritium content in the faecal samples was expressed as percentage of administered dose recovered over 72 h and was used to correct the 15N-excretion data for gastrointestinal transit by dividing the cumulative percentage of administered dose of 15N recovered over 72 h by the cumulative percentage of administered dose of 3H recovered over 72 h.
DNA extraction from faecal specimens. Total bacterial DNA was extracted from faecal samples using a modified version of the method of Pitcher et al.32 as previously described.33 A culture-independent approach using PCR-denaturing gradient gel electrophoresis (DGGE) population fingerprinting was applied to determine the composition of the predominant faecal microbiota using a universal primer targeting the V3 region of the 16S rRNA gene.10
Real-time PCR analysis. Quantification of the faecal Bifidobacterium population was performed with the LightCycler system I (Roche, Mannheim, Germany) using the FastStart DNA Master SYBR Green I kit and a Bifidobacterium genus-specific PCR primer set.34 Data were expressed as log10 bifidobacteria/g wet weight.
Results are expressed as mean values and standard deviations. The statistical analysis was performed with spss software (spss 14.0 for Windows; SPSS Inc., Chicago, IL, USA). Statistical evaluation of the data within a group was performed by applying a paired Student’s t-test. To evaluate changes between the different groups, an unpaired t-test was applied. The level for statistical significance was set at P < 0.05. Pearson correlation coefficients were calculated in the different groups to assess possible associations between different variables.
The characteristics of the subjects in both groups are summarized in Table 1. One male and one female subject withdrew from the study because of the necessity of antibiotic intake, respectively, in groups 1 and 2. Data from these subjects were excluded from analysis. All other subjects completed the study. No statistically significant differences were observed between the groups.
Influence of prebiotic administration on urinary and faecal parameters
The results of the influence of prebiotic intervention are summarized in Table 2. After the 4-week prebiotic administration, the urinary excretion of p-cresol and 15N was significantly decreased in both groups corresponding to a significantly higher 72 h cumulative excretion of 15N corrected for transit. The individual excretion patterns are shown in Figures S1–S3 (published online as Supplementary material). Total nitrogen excretion in urine and faeces, as well as oro-anal transit, were not influenced by the prebiotic intervention in both groups.
Table 2. Effect of prebiotic administration on urinary and faecal parameters (15N, total N, p-cresol and transit) in groups 1 and 2 (mean ± s.d.; Student’s t-test; P < 0.05)
Urinary 15N-excretion (%dose/48 h)
44.1 ± 13.1
36.5 ± 12.6
Urinary N excretion (g/48 h)
16.3 ± 6.7
16.3 ± 6.5
Urinary p-cresol (mg/day)
20.7 ± 11.6
12.7 ± 8.9
Corrected faecal 15N-excretion (72 h)
19.9 ± 6.5
26.1 ± 8.2
Faecal N excretion (g/72 h)
3.9 ± 1.9
3.9 ± 2.1
Transit (cum %3H-PEG/72 h)
59.3 ± 25.7
60.4 ± 25.6
Urinary 15N (%dose/48 h)
49.5 ± 11.9
34.8 ± 12.4
Urinary N excretion (g/48 h)
17.7 ± 4.1
14.9 ± 5.1
Urinary p-cresol (mg/day)
27.7 ± 15.3
17.8 ± 10.8
Corr. faecal 15N-excretion (72 h)
17.9 ± 5.1
22.2 ± 5.2
Faecal N excretion (g/72 h)
3.0 ± 0.88
2.8 ± 1.3
Transit (cum %3H-PEG/72 h)
59.3 ± 22.4
57.5 ± 22.8
Furthermore, a significant difference in change of the urinary 15N-excretion between lactulose (−8 ± 15%) and OF-IN dosing (−16 ± 8%) was found (P = 0.034), which indicated that the change observed after OF-IN intake was significantly larger than for lactulose intake. No significant difference was observed for the difference in change of the other urinary and faecal parameters between both groups.
Quantification of faecal bifidobacteria levels using real-time PCR
DGGE band profiles of V3 16S rRNA gene amplicons were relatively stable within each subject before and after prebiotic intake. However, one band fragment at a specific position in the V3 profiles appeared or intensified after prebiotic intake. Based on the relative positions in the DGGE profiles, this band fragment was tentatively assigned to bifidobacterial organisms. Subsequently, total faecal bifidobacteria were quantified using real-time PCR. Total faecal bifidobacteria significantly increased after lactulose intake: from 8.4 ± 0.81 log10 bifidobacteria/g wet weight at baseline conditions to 9.1 ± 0.65 (P = 0.017). In group 2, OF-IN administration resulted also in a significant increase in the number of bifidobacteria from 7.6 ± 0.99 at baseline to 8.4 ± 0.77 (P < 0.001). Comparison of the change in bifidobacteria between groups 1 and 2 did not result in a significant difference. The individual numbers of bifidobacteria are shown in Figure S4 (published online as Supplementary material).
Correlations between the baseline activity of the colonic microbiota and the effect of prebiotic dosing on the colonic metabolism. Figures 1 and 2 show the correlation between the baseline values of the urinary 15N- and p-cresol excretion, and the influence of, respectively, lactulose and OF-IN administration. Significant correlations were found between the baseline urinary 15N and p-cresol levels and the influence of lactulose and OF-IN administration on the urinary 15N- and p-cresol excretion. This correlation suggests that a higher amount of 15N or p-cresol in urine at baseline results in a more pronounced decrease in urinary 15N- and p-cresol excretion compared to those persons who had a lower baseline value of these markers. The opposite effect was noticed when correlating the initial faecal 15N-content with the influence of prebiotic intake (Figure 3). A more pronounced increase in faecal 15N recovery after lactulose and OF-IN dosing was observed in those subjects who had a lower baseline 15N-excretion. Also, a significant correlation was observed between baseline bifidobacteria counts and the effect of prebiotic dosing (Figure 4; P < 0.05). A higher increase in the number of bifidobacteria was found in the subjects with lower baseline bifidobacteria numbers.
Correlations between changes in colonic metabolism and changes in the composition of the microbiota. No significant correlation was found between the effect of lactulose and OF-IN administration on the bifidobacteria counts and, respectively, the urinary 15N and p-cresol excretion, suggesting that the increase in bifidobacteria after prebiotic intake is not causal to reduced urinary 15N and p-cresol levels (Figure S5, published online as Supplementary material).
The aim of this study was to investigate whether the response to prebiotic dosing is influenced by the baseline activity of the microbiota and the initial bifidobacteria counts, and whether changes in colonic metabolism correlate with changes in the number of bifidobacteria.
The prebiotic potential of a nondigestible carbohydrate is usually assessed trough its ability to stimulate health-promoting bacteria such as bifidobacteria and lactobacilli. In this way, the prebiotic nature of lactulose and OF-IN has been demonstrated in various human volunteer studies.24–27 Also in this study, the bifidogenicity of both substrates was clearly observed. Furthermore, it was found that the relative increase in faecal bifidobacteria depended on the baseline concentration of bifidobacteria. In other words, subjects with lower initial bifidobacterial counts showed a higher response to lactulose or OF-IN administration than those exhibiting higher numbers under baseline conditions. Similar observations have been made previously in other in vivo studies with lactulose or OF-IN.7, 10, 24, 35 These results suggest that prebiotic intake may be particularly effective for subjects exhibiting low intrinsic numbers of bifidobacteria. A decreased number of bifidobacteria has been reported in elderly subjects and in chronic intestinal disorders such as inflammatory bowel disease in which the intestinal microbiota is believed to play a role in the pathogenesis of the disease.36, 37 Prebiotic intake might counteract these changes; however, further longer term investigations are required to substantiate their use.
Contrary to the clearly established inverse correlation between the initial level of bifidobacteria and the increase in bifidobacteria after the intake of both prebiotics, there is no information available on the correlation between the initial colonic metabolic conditions and the response to prebiotic dosing. Therefore, it was investigated in this study whether the baseline metabolic conditions of the volunteers influenced the effects induced by lactulose and OF-IN administration. We focused on the evaluation of the bacterial protein fermentation metabolites ammonia and p-cresol, because of their documented toxicity and relevance for colon cancer risk.13, 14, 38
The results have clearly demonstrated an inverse correlation between the effect of both lactulose and OF-IN intake on the ammonia and p-cresol excretion and their baseline metabolic levels, indicating that the responses to prebiotic dosing also depend on the initial metabolic state of the microbiota. The effect of prebiotic intake on protein metabolites ammonia and p-cresol is the highest in persons with higher baseline levels of these metabolites. It may be assumed that similar favourable effects may be induced on other potentially toxic metabolites from putrefactive metabolism such as indoxylsulphate and phenylacetic acid.
The observed association between the reduced proteolytic activity in the colon and the increase in bifidobacterial counts after administration of lactulose and OF-IN urged us to investigate whether a significant correlation could be found between the increase in bifidobacterial counts and the influence on the 15N and p-cresol excretion. In other words, is the effect on the 15N and p-cresol excretion the highest in volunteers showing the highest increase in bifidobacteria? However, the lack of significant correlations between these parameters suggests that the increase in bifidobacteria is not the cause of the observed effects on the 15N and p-cresol levels. A further exploitation of the colonic microbiota in future studies will be necessary to elucidate the observed effects on metabolic activity.
The effect of prebiotic intake on 15N-excretion can be ascribed to a lower generation of colonic ammonia and to an enhanced uptake of NH3 by the microbiota, which is reflected by a shift from urinary to faecal 15N-excretion.39 Indeed, previous experiments have demonstrated that the increased faecal 15N-excretion was because of an increased incorporation of the biomarker into the bacteria.20, 21 However, it can be assumed that the biomarker is not exclusively taken up by bifidobacteria, as also other colonic bacteria may use ammonia for their metabolism and growth. This might explain the lack of significant correlation between the effect of prebiotic intake on the 15N-excretion and the bifidobacteria counts.
The decreased excretion of p-cresol after lactulose and OF-IN intake was probably because of either an enhanced uptake of tyrosine or metabolic products of tyrosine for bacterial biosynthesis or because of a decrease in protein degradation. Protein degradation is impaired because of acidification of the colonic content and consequent inactivation of proteases, as a result of prebiotic fermentation to SCFA. In addition, a process of so-called catabolite repression results in an inhibition of the deamination of the amino acids and hence, less fermentation metabolites.13, 40, 41 As these mechanisms are not related to a stimulation of bifidobacteria, no correlation could be observed between the effect on the bifidobacteria and on p-cresol excretion. Furthermore, the effect on the proteolytic activity may also be attributed to a downregulation of bacteria involved in protein putrefaction.12
A larger effect of OF-IN on the urinary 15N-excretion was found compared with lactulose. However, no differential effect of both prebiotics was observed on the bifidobacteria counts. Possibly, this differential effect on the 15N-excretion may be linked to the fact that OF-IN does not only have a stimulating effect on the microbiota in the lumen of the colon, but also promotes the mucosa-associated microbiota.42, 43
The results of this study may not only be of benefit for healthy individuals, but also in relation to certain pathological conditions in which elevated baseline levels of ammonia and/or p-cresol have been demonstrated. Hyperammonaemia is generally accepted as an important agent in the development of hepatic encephalopathy in patients with liver cirrhosis who have an impaired capacity to convert ammonia into urea in the liver. Lactulose has already been used as a treatment in these patients because it reduces the level of ammonia in the blood by decreasing its absorption from the colon.44, 45 Elevated levels of p-cresol have recently been demonstrated to be correlated to a higher mortality in uraemic syndrome.46 The exact pathogenic mechanisms occurring in case of renal failure are not completely understood today. Therefore, strategies that counteract the accumulation of p-cresol and other protein fermentation metabolites might constitute a significant improvement in the management of those patients.
In conclusion, the results of this study have shown that the response to prebiotic dosing as indicated by the fate of NH3, p-cresol and bifidobacteria, is determined by the initial colonic conditions. In addition, it has been shown that prebiotic intake induces favourable changes in the colonic bacterial metabolism that are not related to a stimulation of bifidobacteria. Hence, the definition of a prebiotic focusing only on the effect of beneficial bacteria does not include all favourable effects related to prebiotics.
Declaration of personal interests: None. Declaration of funding interests: This work was supported by IWT-Vlaanderen, Brussels, Belgium (GBOU project nr. 010054; ‘Development of a fast, non-invasive technological tool to investigate the functionality and effectivity of pro- and prebiotics in normal healthy volunteers: the use of a labelled biomarker’), the Fund for Scientific Research–Flanders, the University Research Councils and several companies. G. Huys is a postdoctoral fellow of the Fund for Scientific Research–Flanders (F.W.O. Vlaanderen, Belgium).