OUP user menu

H2 and acetate transfers during xylan fermentation between a butyrate‐producing xylanolytic species and hydrogenotrophic microorganisms from the human gut

Christophe Chassard, Annick Bernalier‐Donadille
DOI: http://dx.doi.org/10.1111/j.1574-6968.2005.00016.x 116-122 First published online: 1 January 2006

abstract

The aim of this work was to investigate in vitro interrelationships during xylan fermentation between an H2 and butyrate‐producing xylanolytic species recently isolated in our laboratory from human faeces and identified as Roseburia intestinalis and the H2‐utilizing acetogen Ruminococcus hydrogenotrophicus or the methanogen Methanobrevibacter smithii. H2 transfer between M. smithii or Ru. hydrogenotrophicus and the xylanolytic species was evidenced, confirming the great potential of these H2‐consuming microorganisms to reutilize fermentative H2 during fibre fermentation in the gut. In addition, acetate transfer was demonstrated between the xylanolytic Roseburia sp. and the acetogenic species, both metabolites transfers leading to butyric fermentation of oat xylan without production of H2.

Keywords
  • human colon
  • xylan fermentation
  • microbial interaction
  • hydrogen
  • butyrate

Introduction

In the human colon, the anaerobic microbial community degrades and ferments dietary fibres that have escaped digestion in the upper intestinal tract (Macfarlane & Cummings, 1991). The fibres comprise many types of macromolecules from plant sources, including insoluble plant cell wall polysaccharides (mainly cellulose and hemicelluloses) that represent 20–30% of the dietary fibres ingested daily. The degradation and fermentation of these substrates in the colon requires the contribution of different groups of microorganisms linked in a trophic chain transforming dietary polysaccharides into smaller fragments that are further fermented into short chain fatty acids represented by acetate, propionate and butyrate and gases, mainly H2 and CO2 (Macfarlane & Cummings, 1991). Fermentative H2 is a central metabolite in the overall organic matter degradation that must be removed from the ecosystem to allow a more complete oxidation of substrates (Wolin & Miller, 1983). The main pathway of H2 disposal in the colon remains its utilization in situ by hydrogenotrophic microorganisms (mainly methanogenic archaea and acetogenic bacteria) through interspecies H2 transfer (Christl , 1992). Interactions between microorganisms from the human gut have not been extensively explored, in contrast to the rumen microbiota (Wolin & Miller, 1988; Morvan , 1996). Interspecies H2 transfer between H2‐producing cellulolytic species and hydrogenotrophic microoganisms (methanogen and acetogen) from the human gut was however reported recently (Robert , 2001).

Methane, formed from H2/CO2 by methanogenic archaea, represents a stable end product that is not further metabolized in the gut. By contrast, acetate, produced from both oxidative decarboxylation of pyruvate and reductive acetogenesis from H2/CO2 and/or organic substrate, is used as energy source by different types of epithelial cells (Macfarlane & Gibson, 1994). Acetate could also be efficiently utilized by certain groups of anaerobic bacteria, in particular by the butyrate‐producing species belonging to Faecalibacterium prausnitzii and the Roseburia intestinalis/Eubacterium rectale group (Duncan , 2002a; Hold , 2003). These bacterial species are indeed net consumers of acetate in pure culture (Barcenilla , 2000) and contribution of external acetate to butyrate formation could reportedly ranged from 60% to 90% (Duncan , 2004). This contribution of exogenous acetate to butyrate synthesis by human gut bacteria was also shown to be dependant on the carbohydrate‐derived energy sources, xylan being one of the most important substrates for this biosynthesis (Duncan , 2004). Thus, exogenous acetate could be an important metabolite for butyrate formation in the gut. Butyrate formed in the gut is a fuel for epithelial cells (Csordas, 1996; Cavaglieri , 2003) and is also recognized for providing health benefits (protective role against colorectal cancer and colitis) (Scheppach , 2001).

The aim of the present work was to investigate in vitro H2 and acetate transfers during xylan fermentation between an H2‐ and butyrate‐producing xylanolytic species recently isolated in our laboratory from human faeces and two H2‐utilizing microorganisms from the human gut, the hydrogenotrophic acetogen Ruminococcus hydrogenotrophicus (Bernalier , 1996) and the predominant methanogen Methanobrevibacter smithii.

Materials and methods

Organisms and growth conditions

The xylanolytic strain XB6B4 studied was isolated in our laboratory from human faeces. Strain XB6B4 is a strictly anaerobic Gram‐variable rod. Sequencing of the 16S rRNA gene (1411 pb) showed that strain XB6B4 could be assigned to Roseburia genus. Furthermore, the 16S rRNA gene sequence of this strain (EMBL Deposit No.: AM055815) is 99% similar to that of Roseburia intestinalis (Duncan , 2002b). In addition, there is no major phenotypic difference between strain XB6B4 and Ro. intestinalis (Duncan , 2002b), our isolate being however more efficient at degrading xylan than the type strain. Strain XB6B4 can therefore be considered as a new strain of the species Ro. intestinalis.

Methanobrevibacter smithii (DSM 861) came from the Deutsche Sammlung von Bakterien und Zellkulturen and Ruminococcus hydrogenotrophicus (DSM 10507) was previously isolated and identified in our laboratory (Bernalier , 1996).

All microorganisms were grown under strictly anaerobic conditions according to Hungate (1969). The basal medium used to grow xylanolytic and methanogenic strains was the semi‐synthetic BC medium previously described (Robert & Bernalier‐Donadille, 2003). Oat spelts® xylan (10 mg mL−1) was added to the medium for cultivation of the xylanolytic strain while H2/CO2 [60/40, volume in volume (v/v), 202 kPa] represented the sole energy source for culture of the methanogen. The acetogenic bacterium was grown in AC21' semisynthetic medium (Leclerc , 1997) with H2/CO2 (60/40, v/v, 202 kPa) as sole energy source.

Mixed culture experiments

Mixed culture experiments were performed in the basal semisynthetic BC medium (Robert & Bernalier‐Donadille, 2003). Xylan oat spelts® (100 mg) was added to each 16 mL screw‐cap tube (Bellco Glass Inc., Vineland, NJ) before the addition of the prereduced medium (10 mL per tube) kept under O2‐free CO2. Inoculum of the xylanolytic strain was composed of 0.4 mL of the supernatant of a 2‐day‐old oat xylan culture while inocula of H2‐utilizing strains were composed of 0.4 mL of 5‐day‐old H2/CO2 grown cultures. Two different types of coculture inoculations were carried out. In a first experiment, the xylanolytic species and the hydrogenotrophic one (M. smithii or Ru. hydrogenotrophicus) were inoculated simultaneously. Monoculture of the xylanolytic strain was performed as control. In the second experiment, the xylanolytic species was inoculated in the coculture 4 days after the hydrogenotrophic microorganisms (M. smithii or Ru. hydrogenotrophicus). The H2‐utilizing microorganisms were first grown with H2/CO2 (60/40, v/v, 101 kPa) before addition of the xylanolytic bacteria in the coculture medium. The gas phase was not replaced before inoculation of the xylanolytic bacteria. Monoculture of the xylanolytic strain and monocultures of each hydrogenotrophic species were used as controls. For both coculture experiments, mono‐ and cocultures were carried out in triplicate for each incubation time studied. All experiments were also performed in triplicate.

Analytical methods

Gases in the headspace of the cultures were analysed by gas chromatography (Jouany, 1982). Short‐chain fatty acids were analysed by gas chromatography in 5‐day‐old mono‐ and cocultures for simultaneous inoculation experiment and in 7‐day‐old cocultures for sequential inoculation experiment. Lactate, formate, ethanol and succinate were determined by enzymatic methods (Roche kit, R‐biopharm, St Didier au Mont d'Or, France).

Fluorescence in situ hybridization (FISH) analysis

Fluorescence in situ hybridization (FISH) analysis of the bacterial population present in each preculture and in mono‐ and co‐cultures incubated for 1, 2 and 3 days was performed as previously described (Chassard , 2005). Briefly, aliquots (500 μL) of each culture were dispensed into sterile plastic tubes, immediately fixed by addition of formaldehyde to a final concentration of 4% and stored in the dark for 24 h. After dilution, aliquots were filtered on white 0.2 μm pore size polycarbonate membrane filters [Millipore (St Quentin, France) GTTP, 48 mm diameter]. The filters were rinsed with phosphate‐buffered saline (PBS) (pH 7.2, 137 mM NaCl) and increasing concentrations of ethanol (50%–80%–100%). These solutions were removed by applying vacuum. The filters were then dried and stored at −20°C until processing.

The oligonucleotide probe EUB 338 (5′‐CGTGCCTCCCGTAGGAGT‐3′) targeting Eubacteria (Amann , 1990) was used for FISH analysis. The probe was labelled with indocarbocyanine fluorescent dye CY3 (MWG, Ebersberg, Germany) and stored at −20°C. Filters were cut into quarters, the sections placed on coverclips, and then covered with 40 μL of hybridization buffer (40 mM Tris‐HCl, 0.9 M NaCl, 35% formamide, pH 7.4) containing 2 μL (50 ng mL−1) of fluorescent probe. Cells were hybridized for 4 h. Afterwards, the filter sections were incubated in 20 mL of prewarmed washing buffer (20 mM Tris HCl, 0.4 M NaCl, 40 mM ethylenediamine tetra‐acetic acid, 100 μL−1 sodium dodecyl sulfate, pH 7.4) at 48°C for 40 min. In order to determine total bacteria abundance, the filter sections were stained for 20 min with 4′6‐diamino‐2‐phenyllidone (DAPI, 1 μg mL−1 final concentration) prior to microscopic examination. Preparations were put on glass slides, mounted in glycerol medium (Citifluor, Canterbury, England) and examined using a fluorescence microscope (Leica, Marseilles, France) (nd=1.516). The UV excitation was used to illuminate DAPI stained and green excitation to count the probed cells. For each hybridized filter, 400–600 bacteria stained with fluorescent probe were counted (magnification × 1000).

Data and statistical analysis

Data are expressed as mean±standard error (SE). Statistical analyses were performed with Instat 2.01 GraphPad software using Student's t‐test. Tests were two‐tailed and the level used to establish significance was P<0.05.

Results

Kinetics of H2 production

The xylanolytic species, Roseburiaintestinalis, produced large amounts of H2 from xylan fermentation (Figs 1a and b).

Figure 1

  H2 evolution in the monoculture of the xylanolytic species Roseburia intestinalis strain XB6B4 (♦), in the coculture of the xylanolytic species (strain XB6B4) and Ruminococcus hydrogenotrophicus (▪) and in the coculture of xylanolytic species (strain XB6B4) and Methanobrevibacter smithii (▴). (a) Microorganisms were inoculated simultaneously in the coculture medium. (b) Hydrogenotrophic microorganisms were inoculated 4 days before the xylanolytic species in the coculture medium. Each point represents the mean of three determinations±standard error.

The simultaneous association of this xylanolytic bacterium with the H2‐utilizing Methanobrevibacter smithii or Ruminococus hydrogenotrophicus led to a rapid decrease in H2 concentration in the gas phase of the coculture. The rate of H2 disappearance was faster in the coculture with M. smithii than in that with Ru. hydrogenotrophicus (Fig. 1a).

In the second experiment, cocultures were made by inoculating the two microorganisms sequentially: the hydrogenotrophic microorganisms were inoculated 4 days before the xylanolytic strain. The H2, added to the gas phase of the cultures as energy source for hydrogenotrophic microorganisms, quite totally disappeared within this 4‐day‐period (10 and 20 μmol of H2 remaining in M. smithii and Ru. hydrogenotrophicus pregrown cultures, respectively). After addition of the H2‐producing xylanolytic strain in the cocultures, H2 remained at a very low concentration in presence of either M. smithii or Ru. hydrogenotrophicus by contrast to the large quantity of this gas detected in the monoculture of the xylanolytic bacteria (Fig. 1b).

End products of fermentation

Roseburia intestinalis XB6B4 mainly fermented oat xylan to butyrate, with formate and lactate being also produced in smaller quantities (Tables 1 and 2).

View this table:
Table 1

  End products of xylan fermentation by the xylanolytic species Roseburia intestinalis (strain XB6B4) in monoculture and in cocultures with H2‐utilizing microorganisms

Fermentation products (mM)Ro. intestinalisRo. intestinalis + Ruminococcus hydrogenotrophicusRo. intestinalis + Methanobrevibacter smithii
Acetate06.9±1.90
Butyrate15.3±3.116.6±3.614.9±1.2
Lactate3.3±0.83.8±0.53.6±0.6
Formate5.5±1.36.9±15.6±0.9
H24.3±0.30.2±0.10.1±0.05
CH4001.0±0.2
  • * Analyses of end products of xylan fermentation were performed in 5‐day‐old cultures.

  • Microorganisms were inoculated simultaneously in cocultures. Cultures were incubated for 5 days at 37°C. Each value represents the mean of three cultures±SE.

View this table:
Table 2

  End products of xylan fermentation by the xylanolytic species Roseburia intestinalis (strain XB6B4) in monoculture and in cocultures with H2‐utilizing microorganisms

Fermentation products (mM)Ro. intestinalisRuminococcus
hydrogenotrophicusRu. hydrogenotrophicus + 
 Ro. intestinalisMethanobrevibacter smithii + 
 Ro. intestinalis
Acetate019.9±3.410.9±1.90
Butyrate14.1±1.9019.8±2.113.2±2.3
Lactate2.8±0.803.8±0.72.7±0.4
Formate4.6±0.92.8±1.27.3±1.94.3±1.1
H24.2±0.40.5±0.30.4±0.20.6±0.2
CH40003.5±0.6
  • * Analyses of end products of xylan fermentation were performed in 3‐day‐old cocultures, in 3‐day‐old monoculture of Ro. intestinalis and 7‐day‐old monocultures of Ru. hydrogenotrophicus.

  • Coculture experiments were performed by inoculating the xylanolytic species 4 days after the hydrogenotrophic microorganisms. Monoculture of Ro. intestinalis and cocultures were incubated for 3 days at 37°C. Each value represents the mean of three cultures±SE.

In the cocultures obtained by simultaneous inoculation of the xylanolytic and hydrogenotrophic microorganisms, butyrate and lactate, produced only by Ro. intestinalis from xylan fermentation, were found in similar concentrations as in the monoculture of the xylanolytic strain (Table 1). Butyrate represented the main end product of xylan fermentation in these two cocultures. In addition, CH4 was detected in the coculture with M. smithii while acetate was produced in coculture with Ru. hydrogenotrophicus (Table 1).

In the coculture obtained by sequential inoculation of the methanogen and the xylanolytic bacteria, the xylan fermentation pattern was similar to that observed in the monoculture of Ro. intestinalis. Indeed, the butyrate concentration observed in the coculture with M. smithii was not significantly different (P>0.05) from that measured in the monoculture of the xylanolytic strain. In this coculture, most of the H2 produced by the xylanolytic bacteria was metabolized into CH4 by M. smithii (Table 2).

In the coculture obtained by sequential inoculation of the acetogenic and the xylanolytic species, the end products of xylan fermentation were mainly butyrate and acetate, with formate and lactate being also detected in smaller concentrations (Table 2). Butyrate concentration increased significantly (P<0.05) in coculture with Ru. hydrogenotrophicus compared with the monoculture of Ro. intestinalis (Fig. 2 and Table 2). By contrast, acetate concentration largely decreased after addition of the xylanolytic strain to the Ru. hydrogenotrophicus culture (Fig. 2). Acetate concentration was reduced in the coculture to half of that measured in the monoculture of Ru. hydrogenotrophicus (Table 2).

Figure 2

  (a) Kinetics of acetate metabolism in the monoculture of the acetogenic bacteria Ruminococcus hydrogenotrophicus (•) and in the coculture of the xylanolytic Roseburia intestinalis (strain XB6B4) and Ru. hydrogenotrophicus (○). (b) Kinetics of butyrate production in the monoculture of the xylanolytic Ro. intestinalis (▵) and in the coculture of the xylanolytic Ro. intestinalis and Ru. hydrogenotrophicus (▴). The acetogenic bacterium was inoculated 4 days before the xylanolytic species Ro. intestinalis. Each point represents the mean of three determinations±SE.

FISH analysis

The population level of the two H2/CO2‐grown precultures of hydrogenotrophic microorganisms was similar and composed of 106 cells mL−1. The population level of the xylan‐grown preculture of the xylanolytic strain was composed of 2 × 108 bacteria mL−1. The population level of acetogenic and xylanolytizc bacteria in each preculture, evaluated with the EUB338 probe, represented more than 90% of total bacterial abundance determined with DAPI labelling.

In the monoculture of Ro. intestinalis as well as in the two cocultures obtained by simultaneous inoculation of microorganisms (Fig. 3), cell populations increased progressively to reach a maximum after 3 days. Total bacterial populations in the monoculture and in these cocultures were similar (4.6 × 108 bacteria mL−1 in monoculture of the xylanolytic strain vs. 4.9 × 108 and 5 × 108 cells mL−1 in the two cocultures after 3‐day‐incubation) (Fig. 3).

Figure 3

  Quantification of bacterial populations, by fluorescence in situ hybridization using Eubacteria targeting probe (EUB 338), in monoculture of the xylanolytic species Roseburia intestinalis (strain XB6B4) (♦), in the coculture of xylanolytic Ro. intestinalis and Ruminococcus hydrogenotrophicus (▪) and in the coculture of Ro. intestinalis and Methanobrevibacter smithii. (▴). (a) Microorganisms were inoculated simultaneously in the coculture medium. (b) Hydrogenotrophic microorganisms were inoculated 4 days before the xylanolytic species in the coculture medium. Each point represents the mean of three determinations±SE. *Significantly different (P<0.05).

By contrast, the total bacterial population in the coculture obtained by sequential inoculation of the acetogenic and the xylanolytic species was significantly increased (P<0.05) compared with the monoculture of Ro. intestinalis and to the coculture with the methanogen (5.6 × 108 bacteria mL−1 in coculture with the acetogenic strain vs. 4.5 × 108 bacteria mL−1 in monoculture and 4.6 × 108 bacteria mL−1 in coculture with methanogens after 3‐day‐incubation) (Fig. 3). The bacterial population of the two sequential cocultures represented 85% of total bacterial abundance. At the end of incubation, the total microbial abundance in monocultures of hydrogenotrophic species, determined with DAPI labelling was similar (9.7 × 105 acetogenic bacteria mL−1 and 1 × 106 methanogens mL−1).

Discussion

The present work constitutes the first demonstration of interspecies H2 and acetate transfers during xylan degradation between microorganisms from the human gut. This microbial interaction was investigated with a newly isolated H2‐producing xylanolytic species and two H2‐utilizing microorganisms, the predominant colonic methanogen M. smithii and the human acetogenic species Ruminococcus hydrogenotrophicus (Bernalier , 1996). Based on phylogeny analysis and phenotypic characteristics, the new xylanolytic isolate was assigned to the species Ro. intestinalis (C. Chassard et al., unpublished data). Strain XB6B4 is characterized by an important ability to hydrolyse xylan and to produce butyrate and H2 as main fermentative metabolites. Similarly, the Roseburia species previously isolated from the human gut showed ability to degrade polysaccharides such as starch and xylan, and are recognized as an important group of butyrate‐producing bacteria in the gut (Barcenilla , 2000; Duncan , 2002a). As Ro. intestinalis XB6B4 showed higher xylanolytic activity than the other Roseburia species (data not shown) and produced large amount of H2 from xylan fermentation, this strain was selected for investigating H2 transfer during hemicellulose fermentation.

The association of the H2‐utilizing Methanobrevibacter smithii or Ru. hydrogenotrophicus with the xylanolytic species led to a rapid decrease in H2 concentration confirming the great potential of these H2‐consuming microorganisms to efficiently reutilize fermentative H2 during fibre fermentation in the human gut (Robert , 2001). In the 5‐day‐period of incubation of cocultures, formate produced by Ro. intestinalis was not used by the hydrogenotrophic microorganisms, utilization of this metabolite by M. smithii and Ru. hydrogenotrophicus being only observed in cocultures incubated for 7 or 8 days (data not shown). H2/CO2 was metabolized into CH4 by M. smithii whereas acetate was formed from these gases by the acetogen Ru. hydrogenotrophicus. This interspecies H2 transfer between Ro. intestinalis and M. smithii or Ru. hydrogenotrophicus did not influence the metabolism of the xylanolytic species when microorganisms were simultaneously inoculated. Similarly, the H2 transfer observed in sequential coculture of Ro. intestinalis and M. smithii, did not induce any shift in the metabolism of the xylanolytic strain. The absence of any effect of these H2 transfers on the metabolism of the xylanolytic bacterium suggests that the pathway of H2 synthesis in Ro. intestinalis may be independent of the partial pressure of this gas.

By contrast, the presence of the acetogen Ru. hydrogenotrophicus, 4 days before addition of the xylanolytic bacteria, led to a large increase in butyrate concentration in this coculture compared with the Ro. intestinalis monoculture. Concomitantly, acetate concentration decreased in the coculture compared with the monoculture of the acetogen. Large amount of acetate was indeed produced by Ru. hydrogenotrophicus in the coculture medium (more than 10 mM after 4 days of incubation) before inoculation of the xylanolytic strain. The addition of the xylanolytic bacteria led to reduction of acetate concentration while butyrate production increased. It could therefore be concluded that R. intestinalis was able to use acetate produced by the acetogenic bacteria to form butyrate. As previously reported for other Roseburia species (Barcenilla , 2000), Ro. intestinalis XB6B4 was shown to be able to consume external acetate in presence of xylan, in pure culture (data not shown). The acetate transfer between Ro. intestinalis and the acetogenic species was however not observed when the two bacteria were inoculated simultaneously. In this coculture condition, butyrate production from xylan fermentation by Ro. intestinalis was maximal within the first 2 days of incubation while acetate synthesis by the acetogenic bacterium only occurred after this 2‐day‐incubation period (data not shown). Growth of Ro. intestinalis supported by rapid fermentation of xylan, was probably faster than that of the acetogenic bacterium, which was dependant on fermentative H2 produced by the xylanolytic species. It could thus be hypothesized that contribution of acetate to butyrate synthesis by Ro. intestinalis should have been delayed in this coculture, H2 and acetate transfers not occurring simultaneously.

Increase in total bacterial cells was also observed in the sequential coculture between Ru. hydrogenotrophicus and the xylanolytic species, as estimated by both in situ hybridization with EUB 338 probe and DAPI staining, compared with the monoculture of the xylanolytic strain and the other cocultures. This enhancement in bacterial concentration is probably mainly because of an increase in the number of Roseburia cells in the coculture. Indeed, this increase in bacterial number seems difficult to attribute to the acetogenic species, this bacterium growing only autotrophically in cocultures, using H2 produced by Ro. intestinalis. Furthermore, stimulation of Ro. intestinalis population could be explained by utilisation of the acetate produced in the medium by the acetogen before inoculation of the xylanolytic species. In the presence of carbohydrate, acetate utilization was indeed shown to increase the Ro. intestinalis XB6B4 cell concentration (from 20% to 30%– data not shown), as previously demonstrated for other Roseburia species (Duncan , 2002b, 2004). Acetate produced by the acetogen could have thus stimulated growth of Ro. intestinalis in the coculture.

In conclusion, an efficient co‐operation between H2‐producers, butyrogenic bacteria and H2‐utilizing acetogenic bacteria was evidenced, both H2 and acetate transfers between the xylanolytic Ro. intestinalis and the acetogenic species, leading to butyric fermentation of oat xylan without production of H2. The contribution of H2‐utilizing acetogenic bacteria to acetate production that could be rerouted into butyrate by butyrogenic species may therefore be important in the human gut. Furthermore, this cross utilization of metabolites could also play an important role in both H2 elimination and butyrate production in the colon. Indeed, H2 constitutes one of the main gases that may be responsible for digestive troubles (flatus, abdominal distension, Irritable Bowel Syndrome …) whereas butyrate is generally thought to have protective effect against colonic pathologies such as colitis and cancer (Scheppach , 2001). This interspecies metabolite transfer should further be investigated in vitro, using continuous culture of these microorganisms or in vivo, with gnotobiotic animal models.

Acknowledgement

C. Chassard is supported by a fellowship from the French Ministère de la Recherche et de l'Enseignement Supérieur. We express many thanks to A. Ameilbonne and R. Roux for their skilled technical assistance.

References

View Abstract