Editor: Julian Marchesi
Rates of production and utilization of lactate by microbial communities from the human colon
Version of Record online: 6 APR 2011
© 2011 Federation of European Microbiological Societies. Published by Blackwell Publishing Ltd. All rights reserved
FEMS Microbiology Ecology
Volume 77, Issue 1, pages 107–119, July 2011
How to Cite
Belenguer, A., Holtrop, G., Duncan, S. H., Anderson, S. E., Calder, A. G., Flint, H. J. and Lobley, G. E. (2011), Rates of production and utilization of lactate by microbial communities from the human colon. FEMS Microbiology Ecology, 77: 107–119. doi: 10.1111/j.1574-6941.2011.01086.x
- Issue online: 6 JUN 2011
- Version of Record online: 6 APR 2011
- Accepted manuscript online: 11 MAR 2011 01:52PM EST
- Received 23 December 2010; revised 4 March 2011; accepted 7 March 2011., Final version published online 6 April 2011.
- colonic bacteria;
- human health;
- lactate metabolism;
- stable isotope
Lactate metabolism was studied in mixed bacterial communities using single-stage continuous flow fermentors inoculated with faecal slurries from four different volunteers and run for 6 days at pH 5.5 and 6.0, using carbohydrates, mainly starch, as substrates. A continuous infusion of [U-13C]starch and l-[3-13C]lactate was performed on day 5 and a bolus injection of l-[3-13C]lactate plus dl-lactate on day 6. Short-chain fatty acids and lactate concentrations plus enrichments and numbers of lactate-producing and -utilizing bacteria on day 5 were measured. Faecal samples were also collected weekly over a 3-month period to inoculate 24-h batch culture incubation at pH 5.9 and 6.5 with carbohydrates alone or with 35 mmol L−1 lactate. In the fermentors, the potential lactate disposal rates were more than double the formation rates, and lactate concentrations usually remained below detection. Lactate formation was greater (P<0.05) at the lower pH, with a similar tendency for utilization. Up to 20% of butyrate production was derived from lactate. In batch cultures, lactate was also efficiently used at both pH values, especially at 6.5, although volunteer and temporal variability existed. Under healthy gut environmental conditions, bacterial lactate disposal seems to exceed production markedly.
The metabolic activities of gut bacteria have a considerable influence on human health and disease (Guarner & Malagelada, 2003). Dietary carbohydrate substrates, including starch (Jacobasch et al., 1999), that escape digestion by host enzymes may be fermented by microorganisms to short-chain fatty acids (SCFA) in the colon. Acetate is the predominant product of such fermentation, but may also be converted to butyrate by several bacterial species in the colon by the action of butyryl CoA:acetate CoA transferase (Pryde et al., 2002; Duncan et al., 2004a; Louis et al., 2004). Butyrate, which is not further metabolized by microorganisms in the colon (Belenguer et al., 2008), is the preferred energy source for the colonocytes (Gill & Rowland, 2002; Pryde et al., 2002) and may help ameliorate inflammation and prevent colorectal cancer (McIntyre et al., 1993; Hamer et al., 2008; Tazoe et al., 2008; Louis & Flint, 2009).
Propionate is the other major fermentation product detected in the colon while lactate is an intermediate product usually found in low concentrations in faecal samples from healthy subjects (<5 mmol L−1) due to further microbial utilization and conversion to butyrate, propionate or acetate (Belenguer et al., 2007). Lactate is a product of several bacterial groups, including bifidobacteria (Florent et al., 1985) and certain anaerobes (Macfarlane & Gibson, 1991; Duncan et al., 2002). At low concentrations, lactate is considered beneficial in the colon as the low pKa makes it inhibitory to pathogens. Lactate, however, may accumulate to high concentrations (up to 90 mmol L−1) in the colonic lumen of ulcerative colitis sufferers (Vernia et al., 1988), with detrimental effects, including neurotoxic responses (Ewaschuk et al., 2005).
Among the factors that affect the gut microbial ecosystem, pH impacts markedly on the composition and metabolism of the colonic microbiota (Walker et al., 2005; Duncan et al., 2009). This is also the case for lactate metabolism and previous studies have shown that lactate production and utilization are maintained in balance by mixed human faecal bacteria (Bourriaud et al., 2005; Morrison et al., 2006), within the normal physiological pH range (Belenguer et al., 2007). At pH 5.2, however, lactate utilization was curtailed and this metabolite accumulated (Belenguer et al., 2007). This may explain the high lactate concentrations in severe colitis (Vernia et al., 1988), where the colonic pH can approach that of the stomach (Fallingborg et al., 1993). The contribution of various bacterial species to lactate utilization remains ill-defined, but several are known to convert lactate to propionate or butyrate (Duncan et al., 2004b; Falony et al., 2006; Morrison et al., 2006). These include Eubacterium hallii, Anaerostipes caccae and an unnamed species (Duncan et al., 2004b) that are butyrate-producing bacteria and belong to the dominant core group of species in the human intestinal microbiota (Tap et al., 2009; Walker et al., 2011).
Maintenance of low amounts of lactate within the colon represents a balance between utilization and production and imbalances in either can cause lactate accumulation. The current study uses two approaches, long-term (6 days) continuous fermentors and short-term (24 h) batch cultures, to estimate the rates of lactate production and utilization and determine whether these link to certain bacterial groups. The pH of the culture media was shown to modify the rates of lactate metabolism and stable isotope approaches were used to allow the quantification of flow from starches to lactate and to end-product metabolites.
Materials and methods
Collection of faecal samples
These were provided by four adult volunteers (two male, aged 32 and 62 years, and two females, aged 38 and 46 years), all consuming a Western-style diet. The volunteers (referred to as donors A, B, C and D) did not take any antibiotics or drugs known to influence faecal microbiota for the last 6 months before the start of the studies.
Continuous flow fermentor incubations
Single-stage continuous fermentor systems were operated as described previously (Duncan et al., 2003) using a medium based on that of Macfarlane et al. (1989) as modified by Walker et al. (2005). The carbon sources present in the mixed substrate medium were potato starch (0.5% w/v) in addition to xylan, pectin, amylopectin and arabinogalactan, each at 0.06% w/v. The total peptide concentrations (comprising equal amounts of casein hydrolysate and peptone water) were 0.2%. The fermentor growth medium was maintained under a stream of CO2 with a flow rate of fresh medium equating to one pool per day, yielding a dilution rate of 0.042 h−1. Prime doses of SCFA were added to yield initial concentrations of approximately 35 mmol L−1 acetate, 9 mmol L−1 propionate, 5 mmol L−1 butyrate and 1 mmol L−1 each of valerate, isovalerate and isobutyrate, but were not included in the supplied medium. A pH controller delivered sterile solutions of 0.1 mmol L−1 HCl or 0.5 mmol L−1 NaOH to maintain the pH at either 5.5±0.1 (vessel 1) or 6.0±0.1 (vessel 2). The temperature was maintained at 37 °C using a thermal jacket. Faecal suspensions (20%) were prepared by suspending fresh faecal samples in 50 mmol L−1 phosphate buffer (pH 6.5) containing 0.05% cysteine under O2-free CO2 to yield a faecal inoculum of 2% w/v in the vessel. Substrate (mixture of carbohydrates) was infused continuously, with potato soluble starch being supplied at approximately 16 and 22 mg h−1 (which would be equivalent to approximately 92 and 124 μmol glucose h−1) in vessels 1 (110 mL) and 2 (135 mL), respectively, yielding a similar starch supply per unit volume in both vessels.
For the lactate metabolism studies, a continuous infusion of [U-13C]starch and l-[3-13C]lactate was performed for 10 h on day 5, with a prime injection of [1-13C]acetate, [1,2-13C]acetate and [1-13C]propionate given approximately 4 h after the start of the infusion of labelled starch and lactate. The amounts of those labelled metabolites infused or injected are presented in Table 1. The following day (day 6), both vessels received a bolus injection of l-[3-13C]lactate (approximately 23 and 28 μmol in vessels 1 and 2, respectively) plus dl-lactate (approximately 230 and 275 μmol in vessels 1 and 2, respectively).
|Volunteer A||Volunteer B||Volunteer C||Volunteer D|
|pH 5.5||pH 6.0||pH 5.5||pH 6.0||pH 5.5||pH 6.0||pH 5.5||pH 6.0|
|Infusion (μmol h−1)|
|[U-13C]starch (μmol glucose)||8.5||13.3||7.4||11.0||8.4||15.6||8.8||17.0|
Daily samples were taken from each vessel to monitor SCFA and lactate concentrations. On the day of the infusion (day 5), samples were taken every 30 min for the first 3 h of the infusion and hourly thereafter until 10 h to measure SCFA and lactate concentrations and metabolite 13C enrichments. On the day of the injection (day 6), samples were taken at 30-min interval from just before until 4 h after the bolus injection, with lactate concentrations and metabolite 13C enrichments measured.
Batch culture incubations
Fresh faecal samples from the same four volunteers (A, B, C and D) were collected weekly on 12 occasions over a 3-month period. Slurries of this material were used for batch culture incubations with an anaerobic medium similar to that used for the continuous flow fermentor incubations, based on Macfarlane et al. (1989) as modified by Walker et al. (2005). The carbohydrate sources present in the mixed substrate medium were potato starch (0.14% w/v) in addition to xylan, pectin, amylopectin and arabinogalactan, each at 0.015% w/v. The total peptide concentrations (comprising equal amounts of casein hydrolysate and peptone water) were 0.2%. Samples were inoculated at two different pH values (mean±SD 5.9±0.2 and 6.5±0.2) and with either a carbohydrate mixture alone or with dl-lactate (approximately 35 mmol L−1 initial concentration) also present. SCFA were also added to the medium to yield initial concentrations of approximately 33 mmol L−1 acetate, 9 mmol L−1 propionate, 5 mmol L−1 butyrate and 1 mmol L−1 each of valerate, isovalerate and isobutyrate. The fermentor medium was dispensed into Hungate tubes under a stream of CO2 (Miyazaki et al., 1997) and heat sterilized at 121 °C (15 min). After cooling, heat-labile vitamins were added and the medium was inoculated with the faecal slurry under CO2 and incubated at 37 °C. Faecal slurries (20%) were prepared within 2 h of collection in an anaerobic phosphate buffer saline to yield a final concentration of approximately 0.2%. Tubes were inoculated in duplicate and samples were taken at 24 h to measure SCFA and lactate concentrations. Samples of uninoculated medium were also taken to measure the initial concentrations and the initial pH values.
Quantification of bacteria in faecal and continuous fermentor samples by FISH analysis
Samples were taken from faeces (0.5 g) and the fermentor incubations on day 5 (1 mL) for FISH analysis. Faecal samples were diluted with phosphate buffer (1 : 10), and all samples were fixed by mixing 1 : 3 in 4% w/v paraformaldehyde at 4 °C for 16 h and stored at −20 °C. FISH analysis was performed as described by Harmsen et al. (2002). Diluted cell suspensions were applied to gelatin-coated slides and the slides were hybridized overnight with the appropriate probes. Vectashield (50 μL) (Vector Laboratories, Burlingame, CA) was applied to each slide to prevent fading. Cells were counted automatically using image analysis software cellf (Olympus Soft Imaging Solutions GmbH, Germany) with an Olympus microscope, except when the number of cells was <10 per field of view, in which case the cells were counted manually. For each sample, 30 microscopic fields were counted and the data were averaged. All samples were assessed with the following probes: total bacteria (Eub338, Amann et al., 1990), Bifidobacterium spp. (Bif164, Langendijk et al., 1995), as lactate-producing bacteria, and the E. hallii (Ehal1469, Harmsen et al., 2002) and A. caccae (Acac194, Hold et al., 2002) groups, as potential lactate utilizers.
Determination of concentrations and 13C enrichments in SCFA and lactate
Daily samples from the single-stage continuous fermentors were derivatized in duplicate for the estimation of the concentrations of SCFA and lactate by capillary GC (Richardson et al., 1989). Similar analyses were performed for blank and 24-h sample from the batch culture incubations to measure lactate concentrations. Samples from the fermentors on the day of the infusion (day 5) were analysed for lactate and SCFA concentrations and enrichments, but only lactate concentrations and enrichments were determined in the samples collected during the injection day (day 6). For samples from both days 5 and 6, concentrations were quantified by isotope dilution, while enrichments were measured by GC–MS analysis of the tert-butyldimethylsilyl derivatives, as described previously (Duncan et al., 2004a; Belenguer et al., 2006). Analyses were performed under electron impact ionization conditions; for acetate, the ions M+, M+1 and M+2 at mass/charge (m/z) 117, 118 and 119 were monitored; for butyrate, M+, M+1, M+2 and M+4 (i.e. m/z 145, 146, 147 and 149) were determined, the latter to quantify butyrate formation from two [1,2-13C]acetate molecules; for propionate, M+, M+1, M+2 and M+3 (i.e. m/z 131, 132, 133 and 134) were measured; and for lactate, M+, M+1, M+2 and M+3 ion fragments were analysed (m/z 261, 262, 263 and 264). For the concentration determinations, appropriate corrections were applied for the enrichments of the samples.
Metabolic activities of pure cultures of bacteria
Earlier data suggested that Bifidobacterium adolescentis (Macfarlane & Englyst, 1986) and E. hallii (Duncan et al., 2004b) are major contributors to lactate production and utilization, respectively, in the colon. The maximal contribution of these specific bacteria to lactate metabolism in the mixed faecal bacterial population can be estimated based on knowledge of abundance and activity. The latter was determined from pure culture data. From the values reported previously (Belenguer et al., 2006), the rates of conversion of 0.2% w/v starch substrate to lactate in batch cultures of B. adolescentis were calculated at pH 5.7 and 6.7 between 4 and 8 h of incubation. Similarly, the activity of E. hallii to utilize lactate was determined in the presence of 45 mmol L−1 lactate at pH 5.7 and 6.7 between 8 and 24 h of incubation. For both species, the number of bacteria in the respective incubations was determined by OD (1 OD600 nm=109 cells; Lech & Brent, 1987).
The model structure and fates of the various isotopes are shown in Fig. 1. Let q, Q and E denote the labelled amount (μmol), the total (labelled plus unlabelled) amount (μmol) and enrichment [0.01 molar % excess (MPE)] of either acetate, butyrate, propionate, lactate or starch, denoted by subscripts ‘a’, ‘b’, ‘p’, ‘l’ and ‘s’, respectively. Let i denote the interval between any two times t0 and t1, with t1>t0, and let F(i) denote the flow of a metabolite (labelled plus unlabelled) during i. Eff(i) denotes the loss to the effluent during interval i and E(i) denotes the average enrichment during i. Subscript ‘in’ refers to inflow (production) and subscript ‘out’ refers to use in further metabolic processes (e.g. acetate used to produce butyrate). For example, Fa.in refers to acetate production, while Fa.out stands for acetate outflow, etc. Flows to pool y from pool x are denoted by Fyx. Q, q, E and Eff were measured, while the Fyx, Fin and Fout were unknown.
Data are expressed in terms of two carbon (C2) units, to allow for ‘molar equivalent’ transfers. To achieve this, the concentration of butyrate is multiplied by 2 and the enrichment is divided by 2. The enrichments of propionate and lactate, in terms of C2 units, are given as 0.01 [MPE(M+2)+MPE(M+3)]. The concentrations of acetate, propionate and lactate and their M+1 enrichments are as measured directly.
Infusion day (day 5)
Calculations are based on time points during the continuous infusion of labelled starch and lactate between 4 and 10 h, after the prime doses of labelled acetate and propionate. Lactate formation (Fl.in) and utilization (Fl.out) were obtained from the changes in labelled (M+1) and total (labelled plus unlabelled) lactate as observed during the continuous infusion of [3-13C]lactate:
It was assumed that butyrate was formed (Fb.in) via two pathways, either through the extracellular acetate pool (Fba) or directly from lactate (Fbl). First, from the changes in the total butyrate concentration, Fb.in was obtained:
Then, changes in the M+1 enriched butyrate were modelled as
Writing Fba=Fb.in−Fbl and substituting into Eqn. (4) then yields Fbl and Fba follows.
The total production of propionate (Fp.in) was obtained from
where it was assumed that propionate has no further metabolic fates, i.e. Fp.out=0. Changes in labelled propionate derived from [3-13C]lactate were modelled as:
which then provided an estimate for Fpl.
The incorporation of lactate into acetate (Fal) is obtained from assuming that lactate may be utilized only to produce acetate, butyrate and propionate: Fl.out=Fal+Fbl+Fpl. Furthermore, the M+1 acetate movements yield an estimate for Fa.out based on:
Fa.in follows from
Incorporation of starch into acetate (Fas), lactate (Fls) and propionate (Fps) was obtained from the changes in labelled metabolites that were produced from the infused [U-13C]starch. For Fls:
Here, it is assumed that both M+2 and M+3 lactate were formed from [U-13C]starch [i.e. El.m+2=0.01 MPElactate(M+2)+0.01 MPElactate(M+3)]. Making similar assumptions for propionate, Fps follows from:
Incorporation of starch into acetate follows from:
Bolus injection of labelled lactate (day 6)
Except in the vessel at the lower pH (5.5) inoculated with a faecal suspension from volunteer A, no lactate was detected so that lactate enrichments El.m+1 and El.m+2 could not be determined. Estimates of lactate formation were based, instead, on samples collected following the bolus injection of [3-13C]lactate on day 6. Lactate utilization (Fl.out) on day 6 was obtained from the changes in labelled (M+1) lactate [based on Eqn. (1), with ‘Infusion’ set equal to zero]. This was then used to obtain Fl.in, based on changes in total (labelled plus unlabelled) lactate [Eqn. (2), with ‘Infusion’ set to zero]. The remaining calculations are based on the day 5 measurements, as follows: it was assumed that Fl.in was the same on days 5 and 6 and this was substituted in Eqn. (2) to derive Fl.out on day 5. Subsequently, an estimate of the lactate M+1 enrichment on day 5, denoted by E*l.m+1, was obtained from E*l.m+1=Infusion rate/(Fl.in+Infusion rate), assuming that the infusate was fully labelled. In subsequent calculations, E*l.m+1 replaced El.m+1, so that Fbl, Fba, Fpl, Fal, Fa.out and Fa.in could be calculated as before, based on Eqns (4), (6)–(8). To estimate the fates of starch, it was assumed that 40% of the lactate came from starch, so that the lactate M+2 enrichment was assumed to be 0.4 of the starch enrichment. This estimate of the lactate M+2 enrichments, denoted as E*l.m+2, then replaced El.m+2 in Eqns (9)–(11) to provide estimates for Fls, Fps and Fas. Note that Fb.in and Fp.in [Eqns (3) and (5), respectively] are unaffected by any of the assumptions.
The assumption of Fl.in on day 5 being equal to that on day 6 was supported by data from donor A at pH 5.5 (0.29 and 0.28 μmol mL−1 h−1, respectively). Data from the same volunteer showed that approximately 40% of the lactate came from starch and this value was applied to the other volunteers. Based on the model adopted, this assumption should only influence the calculation of Fls, Fps and Fas. In practice, sensitivity analysis with the proportion of lactate from starch varied from 1% to 99% only impacted strongly on Fas, Fps and Fpx, with the coefficient of variation lower than 36%, except for Fpx, which showed flows lower than 0.15 μmol mL−1 h−1.
Where SCFA data were replicated, the average values were used. The daily SCFA data from the continuous fermentors were analysed as repeated measures, with volunteer and time point nested within volunteer as random effects, and time point, pH and their interaction were taken as fixed effects. The weekly lactate data from the batch culture incubations were analysed using the same random structure, but with fixed effects now consisting of time point, pH, substrate and their interactions. To account for dependency between time points, a suitable covariance structure (compound symmetry) was fitted on the basis of Schwarz's Bayesian information model fit criterion. Quantities (such as carbon flows, numbers of bacteria) obtained from the day 5 (or day 6) data in the continuous fermentors were analysed as one-way anova with volunteer as the random effect and pH as the fixed effect. Pure culture data on lactate formation and utilization were also analysed as one-way anova with pH as the fixed effect. P<0.05 was regarded as statistically significant. All data were analysed using the MIXED procedure of the sas software package, version 9.1 (SAS Institute Inc., Cary, NC). In addition, the linear relationships between variables of interest were analysed using the REG procedure of the sas software.
Concentrations of SCFA and lactate over time in continuous flow fermentors
The daily concentrations of SCFA (acetate, propionate, butyrate) and lactate in the single-stage fermentors are presented in Fig. 2. SCFA and lactate concentrations required 3–4 days to stabilize in all vessels. Lactate was usually detectable only during the first 2 days and reduced to negligible amounts by day 3. For other SCFA (data not shown), succinate was occasionally detected, albeit at low concentrations, whereas formate had variable initial values (up to 4 mmol L−1) on day 1, but these decreased to zero by day 3. Volunteer A at pH 5.5 showed a pattern different from the other volunteers, with butyrate nearly undetectable (<0.4 mmol L−1) by 3 days, while lactate was detectable throughout and formate was present at approximately 11 mmol L−1 from day 3 onwards.
Rates of lactate formation and utilization
Originally, based on previous results (Walker et al., 2005; Belenguer et al., 2007), it was anticipated that lactate concentrations would be above the limits of detection in most vessels, but, in practice, this only occurred at day 5 for volunteer A at pH 5.5. This volunteer provided the only direct comparison of metabolism on days 5 and 6, with endogenous lactate formation similar on both days (0.29 and 0.28 μmol mL−1 h−1, respectively). For the other samples, therefore, the various rates of lactate metabolism were calculated based on formation determined on day 6 (Table 2) plus metabolite masses and enrichments from day 5. These parameters of endogenous lactate metabolism are presented in Table 3.
|Bolus size* (μmol mL−1)||Fl.in (μmol mL−1 h−1)|
|P-value for pH||0.043|
|pH 5.5 (n=4)||pH 6.0 (n=4)||SED||P-value for pH|
|Lactate production (Fl.in)||0.23||0.08||0.044||0.044|
|From starch (Fls)||0.09||0.03||0.017||0.043|
|From other sources (Flx)||0.14||0.05||0.027||0.044|
|Lactate utilization (Fl.out)||0.25||0.13||0.038||0.053|
|Acetate production (Fa.in)||2.13||1.72||0.417||0.401|
|From starch (Fas)||0.60||0.38||0.207||0.358|
|From lactate (Fal)||0.18||0.06||0.027||0.023|
|From other sources (Fax)||1.34||1.28||0.261||0.843|
|Acetate utilization (Fa.out)||1.09||1.08||0.475||0.987|
|Propionate production (Fp.in)||0.48||0.50||0.040||0.657|
|From starch (Fps)||0.24||0.33||0.067||0.285|
|From lactate (Fpl)||0.07||0.06||0.012||0.669|
|From other sources (Fpx)||0.17||0.11||0.061||0.438|
|Butyrate production (Fb.in)||0.58||0.35||0.200||0.346|
|From acetate (Fba)||0.58||0.35||0.202||0.347|
|From lactate (via acetate)||0.071||0.024||0.0220||0.120|
|From lactate (direct; Fbl)||<0.002||<0.003||0.0025||0.677|
Lactate production (Fl.in) was consistently greater at the lower pH (0.23 and 0.08 for pH 5.5 and 6.0, respectively; P<0.05; Tables 2 and 3) and this pattern was similar for both starch and nonstarch substrates (Table 3). Endogenous lactate utilization (Fl.out) also tended to be greater at the lower pH (P=0.053; Table 3). As expected, the rates of production and utilization were closely matched to maintain constant lactate concentrations, even below the limit of detection.
Carbon flows between starch, lactate and SCFA in continuous fermentors
The continuous infusion of [U-13C]starch and [3-13C]lactate, together with the bolus injection of [1,2-13C]acetate, [1-13C]acetate and [1-13C]propionate, allowed the estimation of flows (expressed as C2 units) between lactate and the main SCFA (acetate, propionate, butyrate; Table 3). Labelled starch also allowed the quantification of the flow to lactate.
Carbon flow through the acetate pool (Fa.in) was considerable (1.7–2.1 μmol mL−1 h−1), with most (>53%) derived from sources other than starch (contribution 16–42%) or lactate (contribution <14%). Flows from starch to lactate (Fls) and from lactate to acetate (Fal) were greater at the lower pH (P<0.05). Propionate formation (approximately 0.5 μmol mL−1 h−1) was independent of pH (P>0.10) and with the majority derived from starch (>51%Fps : Fp.in). Butyrate formation (Fb.in) was similar to propionate formation and was at least twofold greater for volunteers C and D than A or B at the lower pH. Most butyrate derived from lactate (estimated as Fal×Fba/Fa.out) was via the external acetate pool (>78%), and involved the action of acetyl-CoA transferase. The exception was volunteer A at the lower pH, where no acetate utilization or butyrate formation was observed, and most propionate derived from sources (63%) other than starch (21%) or lactate (16%).
The proportions of lactate carbon converted to acetate, propionate and butyrate were also estimated (Table 4). The proportion of lactate converted to propionate was always greater at pH 6 (P<0.05). In contrast, the proportion of lactate metabolized to acetate and butyrate was independent of pH. The fate of lactate also appeared to be volunteer dependent. For one subject (C), butyrate was the main end product whereas for two other volunteers (A and D) a substantial amount of the lactate (37–68%) was converted to propionate. The proportion of butyrate formed from lactate, either directly or via the external acetate pool, varied between 0% and 20%.
|Volunteer||A||B||C||D||Means (n=4)||SED||P-value for pH|
|Fates of lactate (%)|
|To butyrate (direct and via acetate)||0.0||13.1||35.2||11.0||66.9||38.6||28.8||19.6||32.7||20.6||9.36||0.285|
|Sources of butyrate (%)|
|From lactate (direct and via acetate)||0.0||6.1||20.0||3.8||8.9||12.9||13.4||7.0||10.6||7.5||4.62||0.549|
FISH quantification of bacteria that produce or utilize lactate
Bifidobacterium spp. accounted for 3.8–6.1% of the total bacteria present in the faecal inocula, whereas the populations of the E. hallii group were low and more variable (0.04–0.61%) and A. caccae was below the limit of detection (<0.01%). By day 5 of inoculation, the total bacterial numbers had increased at least fourfold (Table 5). By this time, for three volunteers, the bifidobacteria accounted for only 0.3–4.2% of total bacteria whereas for volunteer A the Bifidobacterium spp. contribution was 47% at the lower pH (an increase of 8.5 × 107 g−1). Overall, the log10 numbers of Bifidobacterium spp. only tended to show a weak relationship with the rate of lactate formation (adjusted r2=0.41, P=0.05). The populations of the E. hallii group increased over time by 160-fold, but these still accounted for <0.7% of the total bacteria and were not affected by pH.
|Total numbers (log10)|
|P-value for pH||0.208||0.112||0.128|
Effect of pH on lactate metabolism in batch cultures
A similar mixture of dietary polysaccharides was used for the batch cultures, in the presence of either 0 or 35 mmol L−1dl-lactate. The two pH studied were similar, but not identical, to the fermentor study (5.9 and 6.5). Over the 24 h of batch culture, the pH remained relatively stable (difference between the initial and the final pH<0.4).
In the absence of added lactate, net lactate formation or utilization was in balance for most cultures at both pH. When lactate was added to the initial medium, net disposal was complete in most incubations at pH 6.5 and was always greater (P<0.001) than that at pH 5.9 (Table 6). In the absence of lactate, acetate was the main end product, whereas butyrate accumulated (P<0.001) when lactate was present. The presence of lactate also decreased net production of acetate (P<0.001), but increased net formation of propionate (P<0.001). Furthermore, net production of all three of these SCFA was enhanced at the higher pH (P<0.005), although for propionate and butyrate, this effect was more pronounced with the mixture plus lactate than the mixture alone (interaction of substrate × pH, P<0.001). Net lactate utilization was also greater at the higher pH, but again, this occurred mainly in the presence of lactate (interaction of substrate × pH, P<0.001). At the lower pH (5.9) and with the mixture plus lactate cultures, a linear relationship was observed between net lactate utilization and butyrate formation (P<0.001; adjusted r2=0.79; Fig. 3).
|Substrate × pH||<0.001||0.49||<0.001||<0.001|
The responses varied between volunteers and weeks. For example, net lactate utilization and butyrate production were lower for volunteers C and D than A and B, at the lower pH, and net formation of propionate and butyrate differed between weeks.
Activity of pure cultures
Estimates of the equicell abilities of B. adolescentis L2-32 to convert starch to lactate and E. hallii L2-7 to metabolize lactate (to butyrate) are given in Table 7. Both types of bacteria were more active (P<0.01) at the lower pH. Nonetheless, at both pH, the ability of E. hallii L2-7 to dispose of lactate exceeded formation by B. adolescentis L2-32 by at least fivefold.
|pH (n=3)||Formation (μmol per 109 cells h−1)||Utilization (μmol per 109 cells h−1)|
|P-value for pH||0.007||0.007|
Kinetics of lactate formation and utilization
Although lactate is a known fermentation product of carbohydrate metabolism within the colon, the concentrations are usually low or undetectable in faecal samples from healthy donors (Vernia et al., 1988; Macfarlane & Cummings, 1991; Duncan et al., 2007). Thus, rapid metabolism must also occur. When lactate accumulates, however, as in patients with severe ulcerative colitis (Vernia et al., 1988; Hove et al., 1994) then, in the absence of altered rates of absorption (Umesaki et al., 1979), this must be due to changes in either the rate of formation and(or) disposal. Of the many factors that influence microbial lactate utilization and production, the most important probably include substrate supply (Cummings et al., 1989; Duncan et al., 2007), pH (Belenguer et al., 2007; Duncan et al., 2009) and abundance of appropriate bacteria (Roberfroid, 2005). In the present study, substrate supply per unit volume was fixed and the effect of pH was tested, both on direct metabolism and via changes in bacterial populations. The use of stable isotopes allowed the direct quantification of lactate production and utilization as well as the conversion of lactate to propionate or butyrate, the latter either via butyrate kinase or via the butyryl CoA:acetate CoA transferase route. Furthermore, cooperative actions between bacteria have been identified (Wolin et al., 1991; Flint et al., 2007), and including those that produce and utilize lactate (Duncan et al., 2004b; Belenguer et al., 2006). Therefore, changes in the activity and population abundances of these bacteria need to be considered alongside the dynamic quantification of inflows and outflows of specific metabolites. This work suggested that up to 20% of butyrate production in the mixed community could be derived from lactate rather than produced directly from carbohydrates.
For the fermentor study, the infusion of carbohydrate was similar in both vessels and equivalent to approximately 1.7 μmol glucose mL−1 h−1 with a theoretical maximal lactate formation >3 μmol mL−1 h−1. In practice, the rates observed were much lower (0.06–0.34 μmol mL−1 h−1), indicating that only a small fraction of the carbohydrate (and peptide) substrates was converted to lactate. In contrast, the capacity to dispose of lactate appears to be higher, as shown from the rates of disposal observed following a bolus injection of lactate (0.36–0.86 μmol mL−1 h−1, data not shown). A high estimated minimal rate of disposal (1.47 μmol mL−1 h−1) was observed for the batch culture incubations with 35 mmol L−1 lactate at both pH 5.9 and 6.5. In both the fermentor and the batch approaches, these values represent a capacity for a rapid response and, therefore, the inherent disposal capacity of the microorganisms involved exceeds the ability to produce lactate under the substrate conditions used with these healthy volunteers. Nonetheless, changes in either process can alter lactate concentrations.
Increased lactate formation has also been observed previously in batch cultures at mild to moderate acidic pH (studied between 5.2 and 6.4; Belenguer et al., 2007). These earlier data (Belenguer et al., 2007) also showed that lactate utilization was strongly inhibited at pH 5.2 and this would help explain lactate accumulation in colitis patients, where a similar low pH occurs (Nugent et al., 2001). At higher pH (5.9), however, the mixed faecal microbiota were able to rapidly utilize lactate (Belenguer et al., 2007) and thus prevent excessive accumulation. The current data show that even at a more acidic pH (5.5), but still within the range reported for the proximal large intestine in healthy people (Bown et al., 1974; Macfarlane et al., 1992), the capacity for lactate utilization still exceeded lactate formation. Nonetheless, changes in the type and supply of fermentable substrate and environmental conditions influence both bacterial populations and the products of their metabolism. For example, for stool samples collected weekly over 3 months from the free-living volunteers in this study, only in 29/41 cases was lactate detected (at >1 mmol kg−1 faecal water). All volunteers had at least five stools with detectable lactate, and the maximum number of stools with lactate for any one volunteer was eight (out of 12 collections). Thus, all the volunteers had lactate producers.
Potential lactate producers and utilizers
Considering lactate producers, Bifidobacterium spp. (Florent et al., 1985) are major starch utilizers within the human colon (Macfarlane & Englyst, 1986; Leitch et al., 2007). Furthermore, in pure culture, lactate production by bifidobacteria is stimulated at a slightly acidic pH (Table 7; Belenguer et al., 2006). Therefore, it was expected, based on earlier observations (Levrat et al., 1991; Silvi et al., 1999; Belenguer et al., 2006), that lactate-producing bacteria, such as bifidobacteria, and lactate formation would both be increased in the fermentors at the lower pH with starch as a substrate. Nonetheless, the increase in bifidobacteria at the lower pH was less, relatively, than the change in lactate production and raises the question of the importance of the bifidobacteria to lactate metabolism. Although 41% of the variance in lactate formation within the fermentors could be explained by the numbers (log10) of Bifidobacterium spp. present, the actual numbers of those bacteria, both in absolute terms and as a percentage of total bacteria, varied between individuals, as observed previously (Flint et al., 2007). Indeed, when these bacterial numbers were combined with the rates of lactate production from a starch substrate for specific Bifidobacterium spp. (Table 7), this would account for between 2.8% and 70% of lactate formation within the fermentors. The largest contribution occurred with volunteer A at pH 5.5, who had the highest abundance of Bifidobacterium spp. (47% at pH 5.5) whereas for this volunteer at the higher pH and the other three volunteers at both pH only a maximum of 21% of lactate formation could be accounted by bifidobacteria. These observations show that other microorganisms make a very important contribution to lactate production. Apart from other lactic acid bacteria such as Lactobacillus spp., additional bacterial groups known to synthesize lactate are Collinsella spp., Eubacterium rectale/Roseburia spp., Faecalibacterium prausnitzii and Bacteroides spp. (Macfarlane & Gibson, 1991; Barcenilla et al., 2000; Duncan et al., 2002). The latter four groups include the most abundant bacterial species found within the human intestinal microbiota (Tap et al., 2009; Walker et al., 2011) and typically account for >50% of total faecal bacteria (e.g. Duncan et al., 2007).
Eubacterium hallii, A. caccae and the new species Anaerostipes coli (Walker et al., 2011) have been identified as lactate utilizers (Duncan et al., 2004b) that form butyrate as the end product in the presence of fermentable polysaccharides (Belenguer et al., 2007). On an equicell basis, the ability of E. hallii to metabolize lactate exceeds considerably the capacity for lactate production by bifidobacteria (Table 7), but there was no relationship between E. hallii abundance and total lactate utilization (P>0.10). Furthermore, the near-maximal rate of lactate disposal by E. hallii (Table 7), when combined with the numbers present in the fermentors, would only account for 1.2–18.0% of lactate total disposal, with <4.8% in most cases. The situation is somewhat different when only lactate converted to butyrate is considered, however, and where 0–47% could be attributed to the action of E. hallii. Thus, other bacteria must play important roles in the utilization of lactate, including conversion to butyrate. Interestingly, recent evidence indicates that A. coli, which may only utilize d-lactate, is of similar abundance to E. hallii in the human colon (Walker et al., 2011). Other candidates not detected by the FISH probes used here include Coprococcus catus (Louis & Flint, 2009) and bacteria related to Megasphaera elsdenii and Eubacterium limosum (Sato et al., 2008). The involvement of these other bacteria would explain why lactate disposal in the fermentors (Table 2) was not pH sensitive and why butyrate was not always the dominant end product.
In summary, lactate was efficiently used at two physiological pH, 5.5 and 6.0, in continuous fermentor systems and, in most cases, exceeded the rates of lactate production by species such as Bifidobacterium. This ability to dispose of lactate in excess of the amounts normally produced should be viewed as a beneficial trait for the human colon where moderate to high accumulation of lactate is usually associated with detrimental responses (Ewaschuk et al., 2005). While some of the key players have been identified, the relative importance of different bacterial species in lactate formation and disposal within the microbial community still remains to be established.
The Rowett Research Institute and Biomathematics and Statistics Scotland are supported by the Scottish Government Rural and Environment Research and Analysis Directorate. A.B. received financial support from the Spanish Ministry of Education and Science.
- 1990) Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl Environ Microb 56: 1919–1925. , , , , & (
- 2000) Phylogenetic relationships of dominant butyrate producing bacteria from the human gut. Appl Environ Microb 66: 1654–1661. , , , , & (
- 2006) Two routes of metabolic cross-feeding between Bifidobacterium adolescentis and butyrate-producing anaerobes from the human gut. Appl Environ Microb 72: 3593–3599. , , , , , & (
- 2007) Impact of pH on lactate formation and utilization by human fecal microbial communities. Appl Environ Microb 73: 6526–6533. , , , , & (
- 2008) Quantitative analysis of microbial metabolism in the human large intestine. Curr Nutr Food Sci 4: 109–126. , , , & (
- 2005) Lactate is mainly fermented to butyrate by human intestinal microfloras but inter-individual variation is evident. J Appl Microbiol 99: 201–212. , , , , , & (
- 1974) Effects of lactulose and other laxatives on ileal and colonic pH as measured by radiotelemetry device. Gut 15: 999–1004. , , , & (
- 1989) Quantitative estimates of fermentation in the hind gut of man. Acta Vet Scand 86: 76–82. , & (
- 2002) Growth requirements and fermentation products of Fusobacterium prausnitzii, and a proposal to reclassify it as Faecalibacterium prausnitzii gen. nov., comb. nov. Int J Syst Evol Micr 52: 2141–2147. , , , & (
- 2003) Effects of alternative dietary substrates on competition between human colonic bacteria an anaerobic fermentor system. Appl Environ Microb 69: 1136–1142. , , , , , & (
- 2004a) Contribution of acetate to butyrate formation by human faecal bacteria. Brit J Nutr 91: 915–923. , , , , & (
- 2004b) Lactate-utilising bacteria, isolated from human faeces, that produce butyrate as a major fermentation product. Appl Environ Microb 70: 5810–5817. , & (
- 2007) Reduced dietary intake of carbohydrate, by obese subjects, results in decreased concentrations of butyrate and butyrate-producing bacteria in feces. Appl Environ Microb 73: 1073–1078. , , , , & (
- 2009) The role of pH in determining the species composition of the human colonic microbiota. Environ Microbiol 11: 2112–2122. , , & (
- 2005) d-Lactate in human and ruminant metabolism. J Nutr 135: 1619–1625. , & (
- 1993) Very low intraluminal colonic pH in patients with active ulcerative colitis. Digest Dis Sci 38: 1989–1993. , , & (
- 2006) Cross-feeding between Bifidobacterium longum BB536 and acetate-converting, butyrate-producing colon bacteria during growth on oligofructose. Appl Environ Microb 72: 7835–7841. , , & (
- 2007) Interactions and competition within the microbial community of the human colon: links between diet and health. Environ Microbiol 9: 1101–1111. , , & (
- 1985) Influence of chronic lactulose ingestion on the colonic metabolism of lactulose in man (an in vivo study). J Clin Invest 75: 608–613. , , , , & (
- 2002) Diet and cancer: assessing the risk. Brit J Nutr 88 (suppl 1): S73–S87. & (
- 2003) Gut flora in health and disease. Lancet 361: 512–519. & (
- 2008) Review article: the role of butyrate on colonic function. Aliment Pharm Therap 27: 104–119. , , , , & (
- 2002) Extensive set of 16S rRNA-based probes for detection of bacteria in human faeces. Appl Environ Microb 68: 2982–2990. , , , & (
- 2002) Assessment of microbial diversity in human colonic samples by 16S rDNA sequence analysis. FEMS Microbiol Ecol 39: 33–39. , , , & (
- 1994) Fecal dl-lactate concentrations in 100 gastrointestinal patients. Scand J Gastroentero 29: 255–259. , & (
- 1999) Dietary resistant starch and chronic inflammatory bowel diseases. Int J Colorectal Dis 14: 201–211. , , & (
- 1995) Quantitative fluorescence in situ hybridization of Bifidobacterium spp. with genus specific 16S rRNA-targeted probes and its application in faecal samples. Appl Environ Microb 61: 3069–3075. , , , , , & (
- 1987) Growth in liquid media. Current Protocols in Molecular Biology (AusubelFM, BrentR, KingstonRE, MooreDD, SmithJA, SeidmanJG & StruhlK, eds), pp. 1.2.1–1.2.2. John Wiley & Sons, New York. & (
- 2007) Selective colonization of insoluble substrates by human faecal bacteria. Environ Microbiol 9: 667–679. , , , & (
- 1991) Very acidic fermentations in the rat cecum during adaptation to a diet rich in amylase-resistant starch (crude potato starch). J Nutr Biochem 2: 31–36. , & (
- 2009) Diversity, metabolism and microbial ecology of butyrate-producing bacteria from the human large intestine. FEMS Microbiol Lett 294: 1–8. & (
- 2004) Restricted distribution of the butyrate kinase pathway among butyrate-producing bacteria from the human colon. J Bacteriol 186: 2099–2106. , , , , & (
- 1991) The colonic flora, fermentation and large bowel digestive function. The Large Intestine: Physiology, Pathophysiology and Disease (PhillipsSF, PembertonJH & ShorterRG, eds), pp. 51–92. Raven Press, New York. & (
- 1986) Starch utilization by the human large intestinal microflora. J Appl Microbiol 60: 195–201. & (
- 1991) Co-utilization of polymerized carbon sources by Bacteroides ovatus grown in a two-stage continuous culture system. Appl Environ Microb 57: 1–6. & (
- 1989) Influence of mucin on glycosidase, protease and arylamidase activities of human gut bacteria grown in a 3-stage continuous culture system. J Appl Bacteriol 66: 407–417. , & (
- 1992) Comparison of fermentation reactions in different regions of the human colon. J Appl Bacteriol 72: 57–64. , & (
- 1993) Butyrate production from dietary fibre and protection against large bowel cancer in a gut model. Gut 34: 386–391. , & (
- 1997) Degradation and utilization of xylans by the rumen anaerobe Prevotella bryantii (formerly P. ruminicola subsp. brevis) B(1)4. Anaerobe 3: 373–381. , , & (
- 2006) Butyrate production from oligofructose fermentation by the human faecal flora: what is the contribution of extracellular acetate and lactate? Brit J Nutr 96: 570–577. , , , , & (
- 2001) Intestinal luminal pH in inflammatory bowel disease: possible determinants and implications for therapy with aminosalicylates and other drugs. Gut 48: 571–577. , , & (
- 2002) The microbiology of butyrate formation in the human colon. FEMS Microbiol Lett 217: 133–139. , , , & (
- 1989) Simultaneous determination of volatile and non-volatile acidic fermentation products of anaerobes by capillary gas chromatography. Lett Appl Microbiol 9: 5–8. , , & (
- 2005) Inulin-type fructans: functional food ingredients. J Nutr 137: 2493S–2502S. (
- 2008) Isolation of lactate-utilizing butyrate-producing bacteria from human feces and in vivo administration of Anaerostipes caccae strain L2 and galacto-oligosaccharides in a rat model. FEMS Microbiol Ecol 66: 528–536. , , , , , , , & (
- 1999) Resistant starch modifies gut microflora and microbial metabolism in human flora-associated rats inoculated with faeces from Italian and UK donors. J Appl Microbiol 86: 521–530. , , & (
- 2009) Towards the human intestinal microbiota phylogenetic core. Environ Microbiol 11: 2574–2584. , , et al. (
- 2008) Roles of short chain fatty acid receptors, GPR41 and GPR43 on colonic functions. J Physiol Pharmacol 59 (suppl 2): 251–262. , , et al. (
- 1979) Effect of organic acid absorption on bicarbonate transport in rat colon. Pflüg Arch Eur J Phy 379: 43–47. , , & (
- 1988) Fecal lactate and ulcerative colitis. Gastroenterology 95: 1564–1568. , , , , & (
- 2005) pH and peptide supply can radically alter bacterial populations and short chain fatty acid ratios within microbial communities from the human colon. Appl Environ Microb 71: 3692–3700. , , , & (
- 2011) Dominant and diet-responsive groups of bacteria within the human colonic microbiota. ISME J 5: 220–230. , , et al. (
- 1991) Changes of fermentation pathways of fecal microbial communities associated with a drug treatment that increases dietary starch in the human colon. Appl Environ Microb 65: 2807–2812. , , , , & (