The evaluation of ammonia detoxification by pre- and probiotics by means of colonic lactose-[15N2]ureide (15N-LU) degradation is of great interest both scientifically and in terms of nutrition physiology.
Pre- and probiotics were supplemented in healthy adults to evaluate the effect of the ammonia metabolism in the human colon by means of 15N-LU.
A total of 14 participants aged 20–28 years daily received a regular diet either without (no treatment) or with supplementation of 30 g fibre of potatoes (FPs), 30 g wrinkle pea starch (WPS, resistant starch content: 12 and 70%, respectively) and 375 g Lactobacillus acidophilus (LC1) yoghurt, over a 10-day period in a randomised order. After 1 week, 5.7 mg/kg body weight 15N-LU was administered together with breakfast. A venous blood sample was taken after 6 h. Urine and faeces were collected over a period of 48 and 72 h, respectively. The 15N abundances were measured by isotope ratio mass spectrometry.
The mean renal 15N-excretion differed significantly between the supplementation of FP and no treatment (32.5 versus 46.3%, P=0.034), FP and LC1 (32.5 versus 51.6%, P=0.001), and WPS and LC1 (38.5 versus 51.6%, P=0.048). The mean faecal 15N-excretion amounted to 42.7% (no treatment), 59.7% (FP), 41.8% (WPS) and 44.0% (LC1). In comparison with no treatment, the urinary 15NH3-enrichment was significantly decreased at 16 h after FP supplementation.
The prebiotic intake of FP and WPS lowered the colonic generation and the renal excretion of toxic 15NH3, respectively, when using 15N-LU as a xenobiotic marker.
In recent years, 15N-labelled lactose ureide (15N-LU) has been successfully used for studying colonic nitrogen and ammonia metabolism, respectively, to assess the effects of pre- and probiotics on the human intestinal flora (Jackson et al., 2004; De Preter and Verbeke, 2006; De Preter et al., 2006, 2008; Geboes et al., 2006; Cloetens et al., 2008a). The molecular bond between the sugar moiety and urea of 15N-LU has been shown to resist enzymatic degradation in the small bowel, but to be split by Clostridium innocuum in the caecum (Heine et al., 1995; Mohr et al., 1999; Wutzke and Glasenapp, 2004; Wutzke and Schütt, 2007). 15N-LU has the advantage of supplying both the nitrogen and the energy needed for microbial protein synthesis simultaneously and has therefore been used as a carrier for transporting 15NH3 (common abbreviation for the ammonia/ammonium equilibrium 15NH3+H3O+↔15NH4++H2O shifted to the ammonium side under physiological conditions) into the colon. Probiotics are defined as viable microorganisms that have beneficial effects in the prevention and treatment of specific pathologic conditions in the gut (Fuller, 1989). Human in vivo studies have proven that Lactobacillus johnsonii, formerly known as Lactobacillus acidophilus (La1 and LC1, respectively), supports the physiological function of the gastrointestinal tract and stimulates the immune defence system (Pfeiffer and Rosat, 1999). Prebiotics are defined as non-digestible food ingredients, mainly oligosaccharides, non-digestible polysaccharides, resistant starches, and others that support the growth of certain bacteria such as lactobacilli and bifidobacteria that have various beneficial effects on the gut, and, consequently, for the health of the host (Gibson and Roberfroid, 1995). Prebiotics have the ability of increasing the saccharolytic activity, but contrastingly to decrease the proteolytic activity of the microbiota of the gut (Swanson et al., 2002). The colonic microbiota is responsible for degrading proteins partly to potentially toxic metabolites such as ammonia, amines and others which are implicated in the pathogenesis of different diseases (Smith and Macfarlane, 1996). It is assumed that ammonia is assimilated systematically by the bacteria, resulting in reduced urinary excretion and corresponding increased faecal excretion. The scientific literature previously provided evidence that the consumption of probiotics and prebiotics containing resistant starch and other non-digestible carbohydrates may influence the bacterial production of toxic metabolic end products, particularly ammonia, in the human colon (Cummings and Macfarlane, 2002; De Preter et al., 2004; Geboes et al., 2005). Thus, the evaluation of ammonia detoxification by pro- and prebiotics, respectively, by means of bacterial 15N-LU degradation to 15NH3 and its subsequent excretion and absorption is still of great interest scientifically, in terms of nutrition physiology, and, with respect to pre- and probiotics, of commercial interest. Therefore, the aim of this study is to investigate further prebiotics in form of fibre of potatoes (FPs) containing a type-1 resistant starch (RS1) of 12% and a total dietary fibre content of 60%, wrinkle pea starch (WPS) containing a type-2 resistant starch (RS2) of 70% and a total dietary fibre content of 1%, and more probiotics in form of L. acidophilus yoghurt (LC1), respectively, on the renal and faecal nitrogen excretion, the urinary 15NH3-enrichment, the absorption and the potential incorporation in selected blood plasma fractions by using orally administered lactose-[15N2]ureide as a xenobiotic marker.
Tracer substance and supplements
15N-LU was synthesised according to our method as described in detail previously (Wutzke et al., 1997). Lactose was obtained from Merck Darmstadt, Germany, whereas [15N2]urea (99 atom%) was obtained from Campro Scientific Berlin, Germany. FP and WPS, with a RS1 content of 12% and a RS2 content of 70%, respectively, were provided from the manufacturer Emsland Group GmbH Emlichheim, Germany, whereas the LC1 yoghurt pure (colony forming units: 2.6 × 109/100 g, calorific value: 75 kcal/100 g) was used from Nestlé AG, Frankfurt/Main, Germany. Table 1 shows the detailed composition of both starch products as well as the LC1 yoghurt. Both FP and WPS are commercially available and are commonly used as supplements in bakery as well as in meat products.
In all, 14 healthy adults (9 female, 5 male, age: 20–28 years, body weight: 51.5–107.3 kg) volunteered for this study. They were in good health throughout the study. None of the participants were receiving any medication or had a history of gastrointestinal diseases. Furthermore, none of the participants regurgitated after having received the pre- and probiotics and the 15N-LU, respectively. The volunteers were requested to avoid the additional consumption of pre- and probiotics within the study period. Written consent was obtained from all participants. The testing protocol was approved by the Committee on Ethics of the Faculty of Medicine of the University of Rostock (Registration Number: II HV 11/2006).
All volunteers received an individually standardised regular diet (carbohydrate, fat and protein content 55%, 30% and 15%, respectively) over a time span of 54 days. According to the precursor studies of the Leuven group, no standard diets were imposed (De Preter and Verbeke, 2006; De Preter et al., 2004, 2006, 2007, 2008; Cloetens et al., 2008b). Nevertheless, the volunteers were urged to keep their usual dietary habits, with taken care that their diet remained as constant as possible within the four test phases. On the days of 15N-LU administration, the breakfast was identical in all the 14 participants. All volunteers received a continental breakfast made up of 1 wheat flour roll, 50 g jam, 50 g cheese and 1 cup of tea at 07:45 hours. No other food or drink was allowed until 6 h after breakfast. Blood, urine and faeces samples were taken 30 min before the ingestion of 15N-LU to determine the baseline abundance of 15N. 15N-LU was administered in a dosage of 5.7 mg/kg body weight, corresponding to a 15N-excess dose of 0.4 mg/kg, at 8:00 hours as a single oral pulse strewn on a small area of the jam to ingest the label with one bite. A sample of venous blood was taken 6 h after 15N-tracer administration at 14:00 hours. On the first and on the second day, the urine was collected quantitatively in 2- and 4-, and in 6-h intervals, respectively, whereas faeces were collected quantitatively over 72 h. After a 4-day wash-out phase, four participants each received either 10 g FP, 10 g WPS (dissolved in 100 ml tea), or 125 g LC1 (the equivalent of one yoghurt cup) together with breakfast, lunch and dinner at 8:00, 13:00, and 19:00 hours, respectively, in randomised order over a 10-day period. The 15N-LU administration was repeated on the twenty-third day followed by the wash-out phase. On the thirth day, the randomised cross-over procedure was repeated with the next pre- or probiotic supplement followed by the 15N-tracer administration on the thirty-seventh day. After the wash-out period, the test was repeated with the remaining substance on the fourty-fourth day followed by the final 15N-LU administration (Figure 1). The procedure of 15N-LU administration, blood, urine and faeces collection was identical in all four tests. The 10-day supplementation period includes the complete collection of urine and faeces, whereas 15N-LU was administered on the seventh day of the respective pre- or probiotic supplementation (Figure 1).
The sample preparation before the isotope ratio mass spectrometry (IRMS) analysis has been recently explained in detail (Wutzke and Sattinger, 2006). After centrifugation of the citrated blood sample, the plasma was treated with sulphuric Na2WO4. The soluble supernatant was removed by centrifugation and was stored at −20 °C until analysis (Faust et al., 1981). After defrosting, 0.5 ml supernatant was transferred to tin boats and was dried at 60 °C for 2 h. Urine volumes and faecal masses scraped from polyvinyl film were homogenised, recorded and stored at −20 °C until analysis (Wutzke and Oetjens, 2005). After defrosting, 25 μl of urine was transferred to tin boats, whereas 5 mg wet faeces were weighed on tin discs. Urinary ammonia was separated in vessels by microdiffusion, following alkalisation of the urine with NaOH and trapped in boric acid (Faust et al., 1981).
Supernatant, urine and faeces samples were combusted in the elemental analyser SL (SerCon, Crewe, UK) using copper oxide at 900 °C and subsequently reduced to nitrogen gas using copper at 600 °C. Thereafter, the 15N-abundances were measured by IRMS (Tracer mass 20–20, SerCon, Crewe, UK) as detailed previously (Wutzke et al., 1997). A capillary interface is included for connecting the Tracer mass 20–20 with the SL. The precision (mean % relative standard deviation or coefficient of variation for 15N is 0.2‰. The data were expressed either as 15N-enrichment in atom percent excess (atom%exc), or as percentage cumulative urinary and faecal 15N-excretion.
Percentage of cumulative urinary and faecal 15N-excretion
The 15N-excess dose (D) of 15N-LU is the product of the 15N-enrichment in atom%exc (APE15N−LU) times the intake (I) times the number of labelled 15N-atoms (n=2) divided by the molecular weight (MW=403.4 g/mol) and 100.
The 15N-excretion (E) in a specified urine or faeces sample is the product of the measured 15N-enrichment in atom%exc (APEsample) times the urine volume (vol) or the faecal mass (m) in ml or g, respectively, times the N-concentration (cN) of the urine or faeces sample in mg/ml or mg/g, respectively, divided by 100.
The percentage of cumulative 15N-excretion (PCE) is the product of the urinary or faecal 15N-exretion (E) in mmol, respectively, times 100 divided by the 15N-excess dose (D) in mmol, respectively (Wutzke and Oetjens, 2005). 15ND = 0.0267 mmol 15N per kg body weight.
The Kolmogorov–Smirnow test was used for testing the distribution normality. Then, the factorial one- and two-way analysis of variance using a post hoc Bonferroni correction for multiple comparisons on programme SPSS 15.0 for Windows were used for statistical analysis of the total N-, the faecal 15N- and the urinary 15N-excretion, respectively, (Tables 2, 3 and 4) whereas the Kruskal–Wallis and the Mann–Whitney U-test were used for the urinary 15NH3-enrichment (significance P0.01). All results are quoted as means ± s.d.
The maximum 15N-enrichments in urinary ammonia were reached 12 h after 15N-LU administration in all four tests (no treatment, FP, WPS and LC1 supplementation) averaging 0.027±0.019, 0.018±0.024, 0.023±0.019 and 0.029±0.018, atom%exc, respectively (Figure 2). After 48 h, the mean renal percentage cumulative excretion amounted to 46.3±11.1, 32.5±8.0, 38.5±12.9 and 51.6±16.9% (Table 2), respectively, whereas after 72 h, the mean total faecal excretion amounted to 42.7±20.6, 59.7±24.4, 41.8±26.9 and 44.0±21.7%, respectively (Table 3). The urine fractionation in intervals of 0–6 h, 6–24 h and 24–48 h revealed that the major part of 15N was excreted between 6–24 h (Table 2). The resulting total 15N-exretion (urine + faeces) shows that 89.0, 92.2, 80.2 and 95.6%, respectively, of the ingested dose were excreted. The measured 15N-enrichments in the supernatant amounted to 0.036±0.015, 0.040±0.024, 0.041±0.017, and 0.036±0.020 atom%exc, respectively, whereas the respective numbers of the plasma protein ranged below the detection limit of the IRMS of 0.001 atom%. With the exception of the urinary 15NH3-enrichment values, the Kolmogorov–Smirnow test showed that by the majority, the data were normally distributed. When using analysis of variance, between no treatment and the three different supplementation regimes statistically significant differences of the total urinary 15N-excretion (0–48 h) were observed: no treatment versus FP, FP versus LC1 and WPS versus LC1, respectively, (Table 2), whereas no significant differences could be found neither for the respective faecal 15N-excretion nor for the 15N-enrichment of the supernatant of the blood plasma. Furthermore, significant differences of the fractional urinary 15N-excretion were observed between 6–24 h and 24–48 h as well as for the 15NH3-enrichment between no treatment and FP supplementation (P=0.008) 16 h after 15N-LU ingestion. The total urinary N-excretion measured over the 2-day period after 15N-LU administration (25.1, 22.0, 23.9 and 29.4 g N/2 days) yielded no statistically significant differences between the four tests phases (Table 4).
The primary metabolic function of the microbiota is the fermentation of non-digestible food ingredients that have passed the small bowel such as resistant starch, non-digestible polysaccharides, non-digestible oligosaccharides and proteins and others. During this passage, these substrates are exposed to the differing conditions of the gastrointestinal tract and are then found nearly indigested in the faeces of the large bowel (Wutzke and Oetjens, 2005). The outcome of the subsequent complex metabolic activity of the colon is the supply of energy and the retrieval of reusable substrates for the microbiota itself, and, consequently, for the host. The protein degradation in the large bowel leads to the generation of amino acids (which can be further partly deaminated to ammonia) and of other potentially toxic metabolic end products such as phenols, indoles, amines and others (Smith and Macfarlane, 1996). In vivo studies have been shown that the ammonia generation and accumulation, respectively, can be reduced by lowering the protein supply and by the colonic fermentation of suitable non-digestible carbohydrates of the food (Geboes et al., 2005). In contrast to polysaccharides, preferably butyric acid is generated in the colon during the fermentation of resistant starch. However, the resulting generation of both short chain fatty acids and butyric acid lead to a decrease of the pH value in the colon, and, consequently, to an inhibition of the protease activity, whereas the additional energy supply increases the total bacterial mass (Macfarlane and Cummings, 1991). In combination, these effects lead to an increased incorporation of both amino acid nitrogen and ammonia by the microbiota. The Leuven group recently showed that after 15N-LU administration the supplementation of type-3 resistant starch (RS3) in comparison with baseline significantly reduced the renal 15N-excretion (Cloetens et al., 2008b). Furthermore, the study revealed that the major part of 15N was excreted between 6 and 24 h after 15N-LU ingestion suggesting that the 15NH3 had not reached the distal colon, and, therefore, it only justifies the evaluation of the degradation processes in the proximal colon. When considering the results of this study, our data are in the same order of magnitude of those of the Leuven studies and confirm the tendency of lowering the urinary 15N-excretion after supplementation with prebiotics when using 15N-LU (De Preter et al., 2004; Geboes et al., 2005; Cloetens et al., 2008b). In this study, the mean renal 15N-excretion 48 h after 15N-LU ingestion differed significantly between the supplementation of FP and baseline, FP and LC1, and WPS and LC1 (Table 2). Similar to the findings obtained after RS3 supplementation, we could confirm that the major part of 15N was excreted in urine between 6 and 24 h after 15N-LU ingestion, suggesting that the 15NH3 had not reached the distal colon (Cloetens et al., 2008b, Table 2). In contrast, the urinary 15N-excretion between 0 and 6 h mainly reflects the glucose-[15N2]ureide absorption deriving from the enzymatic degradation of 15N-LU by β-galactosidase in the small bowel (Wutzke and Glasenapp, 2004). RS3 is mainly retrograded amylose formed after cooking and cooling, whereas RS2 occurs relatively dehydrated in uncooked natural granular form (Sajilata et al., 2006). When considering the results obtained after RS3 supplementation containing 54% RS3 and the data of our WPS supplementation containing 70% RS2 the outcome becomes obvious. In contrast to the findings of Cloetens et al., the decrease in the urinary 15N-excretion in this study was less pronounced, indicating that, in comparison with RS3, less fermentable RS2 had reached the colon (Heijnen and Beynen, 1997). However, the corresponding findings after the supplementation of the mainly dietary fibre-containing FP were similar to those after WPS supplementation, indicating that the anaerobic bacterial degradation of cellulose to short chain fatty acids in the human colon is likewise responsible for decreasing the urinary 15N-excretion (Figure 2 and Table 2). The observed significant difference of the urinary 15NH3-enrichment between no treatment and FP, 16 h after 15N-LU ingestion confirm the decreased urinary 15N-excretion after FP supplementation in comparison with no treatment conditions and clearly reflect the microbial degradation of the 15N-LU (Birkett et al., 1996, Table 1). The observed significant 15N-enrichments of the mainly urea-containing supernatant fraction of the blood plasma (0.035 atom%exc on average) 6 h after 15N-LU ingestion clearly shows the appearance of 15N-urea and possibly of certain free 15N-amino acids in the metabolic pool of the blood plasma indicating the release of 15NH3 after 15N-LU degradation, its subsequent absorption and the following re-synthesis to 15N-urea and partly to 15N-amino acids. Nevertheless, the supernatant fraction of the blood plasma can also be potentially contaminated by the absorption of glucose-[15N2]ureide deriving from the enzymatic 15N-LU degradation. The urinary 15N-excretion observed within the first 6 h after 15N-LU administration could confirm this assumption. According to the kinetics of the urinary urea and ammonia enrichment after 15N-LU administration described in an earlier published paper, the sample of venous blood was deliberately taken 6 h after 15N-LU administration, as we assumed the highest 15N-enrichment in the blood plasma fraction around this time point (Wutzke et al., 1997). The significant 15N-enrichments of the urinary ammonia fraction and of the supernatant of the blood plasma, respectively, indicate the immediate excretion, and, in particular, the absorption of 15N deriving from the bacterial degradation of 15N-LU in the human colon. However, the respective 15N-abundances of the plasma protein precipitate ranged below the detection limit of the IRMS of 0.001 atom%. The measurement of the 15N-enrichments in urinary ammonia as well as in the supernatant of the blood plasma when using 15N-LU as a xenobiotic marker is a novelty. The acidification of the colon caused by the fermentation to short chain fatty acids could reduce the enzymatic cleavage of 15N-LU and the conversion of [15N2]urea to 15NH3, respectively, resulting in an increased faecal 15N-excretion. After supplementation of FP and LC1, the resulting faecal 15N-excretions (59.7 and 44.0%, respectively) tend to result in an increase in comparison with no treatment (42.7%) but without a statistical significance. The comparatively high standard deviations were presumably caused by an incomplete homogenisation of the faeces before the combustion in the SL. De Preter et al. (2004) observed that the supplementation of Lactobacillus casei to healthy adults in comparison with no treatment revealed no statistical significant differences of the urinary 15N-excretion. Furthermore, a short term intake of Bifidobacterium breve and L. casei as well neither influenced the urinary nor the faecal 15N-excretion, respectively (De Preter et al., 2007). Similar to these findings, we also did not observe a decrease of the urinary 15N-excretion after the supplementation of LC1 yoghurt in comparison to no treatment. We deliberately applied a portion of three yoghurt cups per day (375 g, equivalent to 9.8 × 109 colony forming units) to mimic common usage consumption (Sanders and Huis In’t Veld, 1999). The constant total urinary N-excretion over the 2-day period after 15N-LU administration clearly argues for the ingestion of a standardised diet within the four test phases (Table 4). We assume that the relatively short 10-day period of our LC1 supplementation and the 7-day period before the 15N-LU administration, respectively, are adequate for achieving steady state conditions. Nevertheless, our findings clearly show that commercially available LC1 yoghurt in the chosen dosage led to identical results as under no treatment conditions and has no immediate impact on the colonic ammonia metabolism when using 15N-LU. When summarising our findings, we are convinced that lactose-[15N2]ureide is a valid tool for investigating nitrogen and ammonia metabolism in the colon. Furthermore, we suggest that the prebiotic intake lowered the colonic generation of 15NH3 and its corresponding urinary 15N-excretion and tends to enhance the 15NH3 uptake by the microbiota. Seen in this light, the process of ammonia detoxification in the colon can be interpreted as an interplay of the microbiota with certain supplements establishing appropriate conditions for their growth and proliferation. In this sense, the lowering of the proteolytic generation and accumulation of ammonia in the colon and the resulting noxious effects for the host by suitable supplementations of pre- and probiotics, particularly in participants suffering from certain diseases such as kidney insufficiency, hepatic encephalopathy, inflammatory bowel disease and colon cancer is understandably of great interest for biochemists, physicians, dieticians and in particular the patients as well.
This study was supported by Emsland Group GmbH Emlichheim, Germany.