Microbes involved in dissimilatory nitrate reduction in the human large ...

FEMS Microbiology Ecology 31 (2000) 21^28

www.fems-microbiology.org

Microbes involved in dissimilatory nitrate reduction in the human large intestine Nick J. Parham *, Glenn R. Gibson

1

Microbiology Department, Institute of Food Research, Earley Gate, Reading RG6 6BZ UK Received 14 June 1999; received in revised form 20 September 1999; accepted 30 September 1999

Abstract Nitrate-limited batch cultures, incorporating 20 different fermentation substrates and inoculated with human faeces, mainly selected for the growth of enterobacteria. The microbial diversity involved was determined by a combination of phenotypic and genotypic procedures. Continuous culture with lactate as the sole electron donor selected for similar micro-organisms, but when antibiotics were incorporated to inhibit Escherichia coli and lactate was replaced with choline, there was a wider microbial diversity recovered. Clostridium ramosum and Bacteroides vulgatus were then isolated as well as enterobacteriaceae. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Colonic bacterium ; Nitrate reduction ; Enterobacteria

1. Introduction There is considerable potential for gas production, by anaerobic bacterial fermentation, in the human large intestine. A proportion of this gas is excreted in the breath or £atus, but absorption and di¡usion rates across the colonic epithelium are limited [1], thus alternative disposal mechanisms occur in the normal bowel, preventing problems such as abdominal discomfort and excessive £atulence [2^5]. Christl et al. [6,7] showed, using full-body calorimetry, that the excretion of only 2.5^14% of the predicted H2 production occurred in humans. Thus, a large proportion of the gas produced by fermentation may be disposed of by other means. There are at least four possible mechanisms of H2 utilisation by prokaryotes in the gut. These are dissimilatory sulfate reduction, methanogenesis, dissimilatory nitrate reduction and acetogenesis [8]. Dissimilatory sulfate and nitrate reduction both produce metabolites which are potentially harmful to the host. These include hydrogen sul¢de, nitrite and ammonia. Methane is generally considered to

* Corresponding author. Tel. : +44 (1189) 357000; Fax: +44 (1189) 267917; E-mail : [email protected] 1

Present address: Food Microbial Sciences Unit, Department of Food Science and Technology, The University of Reading, Reading, UK.

be inert to both bacteria and the host, whereas acetate may be absorbed and metabolised in peripheral tissues, thereby contributing positively towards host energy requirements [1,9^12]. Each of these processes involves the metabolism of four H2 molecules per reaction, thereby reducing the overall gas volume in the gut. In many anaerobic microbial ecosystems, nitrate (NO3 3 ) is the most commonly used alternative electron acceptor to oxygen. It is a strong oxidising agent which can be used to drive oxidative phosphorylation, regenerate oxidised coenzymes for anaerobic intermediate metabolism and provide a source of nitrogen for growth [13,14]. Dissimilatory nitrate reduction is a two-step, energy-yielding process 3 where NO3 3 is ¢rst reduced to nitrite (NO2 ) and then H2 is involved in a further reduction to ammonia (NH 4 ). The product of dissimilatory nitrate reduction in NO3 3 -limited conditions is NH4 , but during carbon limi3 3 tation (NO3 excess) NO2 predominates. The observation that bacteria from the human large intestine are capable of nitrate reduction [9,15^17], coupled with the fact that this anaerobic environment is highly fermentative, suggests that the process may have signi¢cance in humans. Allison and Macfarlane [9] showed that the addition of 10 mM KNO3 to human faecal slurries, supplemented with either starch or casein, signi¢cantly reduced H2 and CH4 production. Although nitrate reduction is the most energetically favourable bacterial process and the most e¤cient mechanism for scavenging

0168-6496 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 6 4 9 6 ( 9 9 ) 0 0 0 7 7 - X

FEMSEC 1083 27-12-99

22

N.J. Parham, G.R. Gibson / FEMS Microbiology Ecology 31 (2000) 21^28

of H2 , its occurrence in the human large intestine may be limited by electron acceptor (NO3 3 ) availability. Vegetables, water and preservatives added to meat are the main sources of nitrate and nitrite. It has been estimated that a typical Western diet provides in excess of 100 mg NO3 3 per day [18,19]. Vega and Bontoux [20] noted that estimates may vary depending on the type of vegetables consumed, but that vegetarians have a high intake of about 200 mg NO3 3 per day. The type of foodstu¡ ingested can therefore markedly a¡ect levels of 3 NO3 3 and NO2 in the distal ileum [10] and may be a factor in determining input into the proximal colon. It is not known how much NO3 3 enters the colon and is therefore available for nitrate-reducing bacteria (NRB) [9]. However, Bartholomew and Hill [21] and Florin et al. [22] have suggested that some NO3 3 enters the large intestine via ileal e¥uent. 3 Since mammals on a low NO3 3 diet excrete more NO3 than they ingest [23^25], endogenous sources must also be present. In the colon this may be via transmural exchange from the blood [26]; from synthesis by in£ammatory cells in the lamina propria [27^29]; or by endogenous formation in tissues and secretion into the gastro-intestinal tract [30,31]. At present, there are few data on the nature of species involved in nitrate metabolism among the human colonic bacteria. This study has used batch and chemostat enrichment culture coupled with molecular characterisation procedures to establish the microbial diversity involved.

age of six replicate samples. Pure cultures of NRB were 3 tested for nitrate reduction to NO3 2 and for NO2 reduction to ammonia prior to genotypic analysis. Short chain fatty acid analysis of spent culture media was performed in triplicate by HPLC. 2.2. Dissimilatory NRB growth in continuous culture A reaction vessel (working volume 0.28 l) containing basal nitrate broth, with lactate (5 g l31 ) as the carbon source, was inoculated with faecal slurry to give a ¢nal concentration of 3% (w/v). The vessel was continuously sparged with oxygen-free nitrogen at a rate of 50 ml min31 , with pH controlled at 6.8 and temperature maintained at 37³C. After inoculation, the chemostat was left for 24 h to equilibrate before the medium pump was switched on to give a dilution rate of 0.1 h31 . Six replicate samples were taken at time of inoculation (T0 ), after the equilibration period (T24 ) and after steady-state conditions had been reached (at least seven culture turnovers) (Tss ). Population changes of major colonic bacterial groups were enumerated. This involved the plating out of chemostat £uid onto agar media designed to select for total anaerobes, total aerobes, bacteroides, bi¢dobacteria, clostridia, coliforms, Gram-positive cocci, lactobacilli and nitrate reducers. The growth media and phenotypic methods of bacterial identi¢cation were as described by Macfarlane et al. [32], Wang and Gibson [33] and Holdeman et al. [34]. 2.3. Dissimilatory NRB growth in chemostat culture incorporating antibiotics

2. Materials and methods 2.1. Dissimilatory NRB growth with various substrates Anaerobically prepared faecal slurries (¢nal concentration 5% w/v) were used to inoculate 75-ml serum bottles containing 50 ml basal nitrate broth incorporating 0.5% (w/v) of an organic substrate (acetate, arabinogalactan, benzoate, butyrate, cellulose, choline, cystine, formate, fructooligosaccharide (FOS), guar gum, inulin, mucin, palmitate, pectin, phenol, propionate, pyruvate, starch or xylan). The basal medium contained per litre : KH2 PO4 , 0.5 g; NaCl, 4.5 g; CaCl2 W2H2 O, 0.06 g; MgCl2 (anhydrous), 0.06 g; FeCl3 , 0.04 g; NaNO3 , 1.0 g; sodium citrate, 0.3 g; ferric citrate (1% w/v), 0.6 ml ; trace elements, 10 ml. The trace element solution contained per litre : H3 BO3 , 23 mg; CoCl2 W7H2 O, 93 mg; CuCl2 W2H2 O, 80 mg; MnCl2 W4H2 O, 30 mg; NaMoO4 W2H2 O, 300 mg; ZnCl2 , 1.74 g; Na2 WO4 W2H2 O, 50 mg. Culture medium pH was adjusted to 6.8 prior to inoculation. Basal broth containing no added organic substrate was inoculated with the same faecal slurry and used as a control. The fermenter headspace was ¢lled with oxygen-free nitrogen and the batch cultures incubated at 37³C for 48 h. Bacterial counts of total anaerobes and NRB were determined as an aver-

To facilitate a greater diversity of organisms being isolated from chemostat culture, two chemostats (À' and `B') were set up in parallel, but with lactate being replaced by choline (as batch culture results indicated this to be a good substrate for nitrate reduction). Chemostat `B' was fed medium containing a ¢nal concentration of 0.5 Wg ml31 gentamicin. At each sampling time (T0 , T24 and Tss ), two identical samples were taken from each chemostat. The samples were processed as described above, but the second one was ¢rst treated with 200 Wg ml31 nalidixic acid for 1 h to inhibit growth of Escherichia coli. 2.4. Assay for the bacterial reduction of nitrate to nitrite Nitrate reduction to nitrite was tested using an adaptation of the method described by Mills et al. [35]. A loopful of bacterial cells was transferred anaerobically from an actively growing pure culture into 300 Wl of nitrate solution (10 mM NaNO3 prepared in 0.85% (w/v) NaCl solution) in a 0.5-ml microcentrifuge tube. The mixture was then vortexed for 10 s and then incubated anaerobically at 37³C for 2 h. The mixture was again vortexed for 10 s and then tested for nitrite production with a Merckoquant

FEMSEC 1083 27-12-99


nitrite test strip (BDH Chemicals Ltd., Poole, UK) according to the manufacturer's instructions. 2.5. Assay for the bacterial reduction of nitrite to ammonia A loopful of bacterial cells was anaerobically transferred from an actively growing pure culture into 300 Wl of nitrite solution (1 mM NaNO2 prepared in 0.85% (w/v) NaCl solution) in a 0.5-ml microcentrifuge tube. The mixture was vortexed for 10 s and then incubated anaerobically at 37³C for 2 h. The mixture was again vortexed for 10 s and then tested for ammonia production by addition of 1 drop of universal indicator solution. The solution was mixed and colour compared visually to negative (uninoculated nitrite solution) and positive controls (0.1, 0.5 and 1.0 mM NH3 in a 0.85% w/v NaCl solution). 2.6. Determination of short-chain fatty acid (SCFA) production As an index of the degree of fermentation in the enrichments, SCFA concentrations were determined using an adaptation of the method described by Masson et al. [36]. An Aminex HPX-87H ion-exclusion column (300 mmU7.8 mm I.D., 9 Wm particle size, Bio-Rad) was used, maintained at 50³C and protected by a cation-H guard column cartridge (30 mmU4.6 mm I.D., Bio-Rad). Detection was achieved using a variable-wavelength detector and di¡erential refractometer connected in series. Data were acquired and chromatograms integrated on both channels (UV at 210 nm and refractive index) using the Bio-Rad PC integration software package ValueChrom1. One-millilitre volumes of culture media were centrifuged (13 000Ug, 5 min) and 20-Wl aliquots of the resulting supernatants injected onto the column. The mobile phase consisted of 0.005 M sulfuric acid, diluted from ÀnalaR' reagent (BDH) in HPLC-grade water (Fisher Scienti¢c), delivered at a £ow rate of 0.6 ml min31 . SCFA (lactate, formate, acetate, propionate and butyrate) were identi¢ed by UV and RI detection and quanti¢ed by comparison to standard curves (0^165 Wmol ml31 ). 2.7. Genotypic identi¢cation of bacterial colonies A presumptive identity of bacterial colonies grown from the batch and chemostat enrichments was derived using phenotypic criteria. However, di¡erent colony morphotypes were also identi¢ed using a genotypic approach, which would facilitate a more reliable characterisation. This involved the extraction of DNA from pure cultures using the InstaGene (Bio-Rad) method and 16S rRNA genes ampli¢ed by PCR. 16S rRNA genes were directly sequenced with a dye-deoxy terminator cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and an automatic DNA sequencer (model 373A; Applied Biosystems) as previously described [37]. The closest relatives of

23

the isolates were identi¢ed using the FASTA program and con¢rmed by pairwise analysis using the GAP program [38]. 2.8. Reagents Unless otherwise stated, all chemicals and reagents were obtained from Sigma Chemical Co. 3. Results 3.1. Batch culture enrichments Experiments in anaerobic batch culture showed that the substrates arabinogalactan, benzoate, butyrate, cellulose, cystine, pectin, phenol, propionate and starch did not have any large stimulatory e¡ect on nitrate reduction. In contrast, acetate, choline, formate, FOS, guar gum, inulin, mucin, palmitate, pyruvate and xylan were found to stimulate nitrate reduction (Table 1). This was determined by the counts of NRB (compared to the negative control, i.e. no added substrate) as well as their reductive ability in pure culture. A combination of phenotype and 16S rRNA gene sequencing identi¢ed the predominant nitrate-reducing species isolates as Escherichia coli, Shigella dysenteriae, Shigella £exneri and Shigella sonnei. 3.2. Dissimilatory NRB growth in chemostat culture Data presented in Table 2 show that while most groups maintained steady populations, numbers of lactobacilli fell by two log values over the period T0 to Tss and nitrate reducers increased by around four logs during the same time period. Analysis of 16S rRNA sequences identi¢ed the main nitrate-reducing species to be the same as those found in the batch enrichments. Predominantly, these were E. coli. A major aim of these experiments was to, as fully as possible, determine NRB diversity in the colon. As such, it was desirable to inhibit the growth of E. coli and thus, possibly allow growth of less competitive organisms. Gentamicin and nalidixic acid were chosen as the test antibiotics, because they have been utilised as selective agents in mating experiments between E. coli and Desulfovibrio desulfuricans [39]. D. desulfuricans strain ATCC 27774 (which is able to reduce NO3 3 to NH4 [40]) has been shown [41] to have minimum inhibitory concentrations for gentamicin (100 Wg ml31 ) and nalidixic acid (50 Wg ml31 ), which are higher than those of E. coli at 100 Wg ml31 [41] and 20 Wg ml31 [39] respectively. It is noteworthy that in the experiments of Argyle et al. [39], some E. coli survived nalidixic acid concentrations of 200 Wg ml31 . The authors attributed this to `the probable protective e¡ect of D. desulfuricans'. The use of these antibiotics

FEMSEC 1083 27-12-99

24


Table 1 Bacterial countsa from nitrate-limited batch culture Substrate

Total anaerobes (CFU ml31 )

NRB (CFU ml31 )

Acetate Arabinogalactan Benzoate Butyrate Cellulose Choline Cystine Formate FOS Guar gum Inulin Mucin None Palmitate Pectin Phenol Propionate Pyruvate Starch Xylan

1.13U109 ( þ 2.25U108 ) 6 1.67U108 4.75U108 ( þ 8.46U107 ) 5.25U108 ( þ 9.31U107 ) 3.50U108 ( þ 4.34U107 ) 1.25U1011 ( þ 2.07U1010 ) 2.00U108 ( þ 2.37U107 ) 4.25U108 ( þ 7.18U107 ) 6 1.67U108 6 1.67U108 6 1.67U108 6 1.67U108 8.50U108 ( þ 3.02U108 ) 2.75U108 ( þ 5.75U107 ) 6 1.67U108 6 1.67U108 1.45U109 ( þ 6.06U108 ) 9.00U108 ( þ 3.03U108 ) 6 1.67U108 6 1.67U108

8.00U106 ( þ 1.26U106 ) 2.50U103 ( þ 1.64U103 ) 6 1.67U102 6 1.67U102 1.25U104 ( þ 3.39U103 ) 3.00U105 ( þ 7.75U104 ) 1.67U102 ( þ 4.08U102 ) 4.00U105 ( þ 3.16U105 ) 2.35U105 ( þ 4.18U104 ) 3.25U105 ( þ 6.47U104 ) 3.00U106 ( þ 7.95U105 ) 1.35U107 ( þ 4.89U106 ) 5.50U104 ( þ 1.64U104 ) 8.00U106 ( þ 2.45U106 ) 6 1.67U102 6 1.67U102 3.25U104 ( þ 4.51U103 ) 1.00U107 ( þ 2.10U106 ) 2.35U104 ( þ 9.73U103 ) 1.00U105 ( þ 1.00U104 )

Anaerobically prepared faecal slurry (¢nal concentration 5% w/v) was used to inoculate 75-ml serum bottles containing 50 ml basal nitrate broth incorporating 0.5% (w/v) of an organic substrate; the pH was adjusted to 6.8 prior to inoculation. The fermenter headspace was ¢lled with oxygen-free nitrogen gas and the batch cultures incubated at 37³C for 48 h. a All values are means of six determinations þ S.D. (shown in parentheses).

would probably a¡ect the growth of other bacterial species in the culture, thus the true diversity would not be evident. However, the aim was merely to inhibit E. coli and thus allow growth of less competitive NRB. Data in Table 3 show a one-log reduction in numbers of total anaerobes and nitrate reducers in chemostat À' between T0 and Tss (with or without nalidixic acid treatment) and a two-log reduction in chemostat `B'. However, this reduction in NRB count may be misleading, because the count did not truly re£ect numbers of nitrate-reducing species as determined through pure culture assays. Di¡erences in counts between T48 (11/19) and Tss (9/9) showed an increase in the ratio of NRB to non-NRB, thus indicating an enrichment for NRB in the culture vessels.

Nitrate-reducing species were identi¢ed, by 16S rRNA sequence analysis, from each of the four samples and are shown in Table 4. Essentially, these were composed of Enterobacteriaceae, but with the use of antibiotics more bacterial diversity involved in nitrate reduction was recovered, including both facultatively and strictly anaerobic species of the genera Enterobacter, Klebsiella, Clostridium and Bacteroides. 3.3. SCFA production Changes in SCFA concentrations over the 48-h fermentation period were determined and are shown in Table 5. It is interesting to note that when formate was used as

Table 2 Bacterial countsa from nitrate-limited chemostat culture

Total anaerobes Total aerobes Bi¢dobacteria Bacteroides Clostridia Lactobacilli Gram-positive cocci Coliforms NRB

T0 (CFU ml31 )

T24 (CFU ml31 )

Tss (CFU ml31 )

6.17U109 1.67U107 7.00U108 2.33U108 3.33U106 1.97U106 5.67U108 8.33U107 7.33U104

8.33U108 (3.33U108 ) 8.17U108 (1.94U108 ) 7.83U108 (2.99U108 ) 1.50U108 (4.38U107 ) 6 1.67U106 5.67U106 (1.86U106 ) 1.70U108 (9.49U107 ) 1.15U109 (4.51U108 ) 8.17U108 (3.31U108 )

1.03U109 (1.63U108 ) 9.33U108 (3.78U108 ) 4.67U108 (1.97U108 ) 3.33U107 (5.99U106 ) 6 1.67U106 6 1.67U104 no datab 7.17U108 (2.40U108 ) 9.50U108 (4.09U108 )

(1.94U109 ) (5.79U106 ) (3.03U108 ) (1.36U108 ) (4.72U105 ) (6.12U105 ) (2.34U108 ) (3.08U107 ) (3.98U104 )

Growth medium containing sodium lactate (5 g l31 ) was fed into the culture vessel (£ow rate of 0.1 h31 ) which contained a faecal inoculum (initial concentration 5% w/v). Samples were taken at time of inoculation (T0 ), after the equilibration period (T24 ) and after steady-state conditions were achieved (at least seven culture turnovers) (Tss ). a All values are means of six determinations þ S.D. (shown in parentheses). b Due to a contamination problem.

FEMSEC 1083 27-12-99


25

Table 3 Bacterial countsa from chemostat cultures incorporating antibiotics T24

Tss

Vessel

T0 AnO2

NRB

AnO2

NRB

AnO2

NRB

A

2.08U109 ( þ 4.92U108 ) 1.40U109 ( þ 3.03U108 ) 1.60U109 ( þ 2.00U108 ) 1.05U109 ( þ 2.66U108 )

7.83U108 ( þ 2.32U108 ) 5.00U108 ( þ 1.41U108 ) 8.33U108 ( þ 2.88U108 ) 9.67U108 ( þ 3.93U108 )

1.75U109 ( þ 3.94U108 ) 1.15U109 ( þ 3.51U108 ) 1.02U109 ( þ 2.56U108 ) 3.67U108 ( þ 1.37U108 )

8.00U108 ( þ 2.45U108 ) 6.00U108 ( þ 1.10U108 ) 1.00U109 ( þ 3.90U108 ) 2.92U108 ( þ 7.94U107 )

6.50U107 ( þ 1.05U107 ) 4.00U107 ( þ 1.79U107 ) 9.50U106 ( þ 2.17U106 ) 1.48U107 ( þ 2.79U106 )

6.83U107 ( þ 2.32U107 ) 3.58U107 ( þ 5.78U106 ) 9.67U106 ( þ 3.27U106 ) 1.23U107 ( þ 4.80U106 )

AN B BN

Two chemostats (À' and `B') were run in parallel, each with choline as the carbon source and operated under nitrate-limited, anaerobic conditions at a £ow rate of 0.1 h31 . Chemostat `B' was fed medium containing a ¢nal concentration of 0.5 Wg ml31 gentamicin. At each sampling time (T0 , T24 and Tss ), two identical samples were taken from each chemostat. Bacteria were isolated from each sample using selective growth media, but the second sample from each vessel was ¢rst treated with 200 Wg ml31 nalidixic acid for 1 h. a All values are given in CFU ml31 as means of six determinations þ S.D. (shown in parentheses).

the growth substrate, more than 90% was fermented. It has been previously recognised that formate is an important electron donor for nitrate reductase [42]. The four growth substrates which produced the most formate were : inulin, guar gum, mucin and xylan respectively. These were all stimulatory to NRB (Table 1); this may be due to the formate produced from their fermentation by other faecal microbes. There were no other clear trends for production of lactate, acetate, propionate or butyrate, with e¡ects on NRB growth. However, the higher molecular mass, polymeric compounds (such as mucin, guar gum, FOS and arabinogalactan) produced the most SCFA. Pectin and arabinogalactan produced high levels of acetate, propionate and butyrate, but were non-stimulatory towards growth of NRB. Therefore, there may be some potential for employing these compounds to control dissimilatory nitrate reduction and produce SCFA in the gut. 4. Discussion Dissimilatory nitrate reduction may have a role in re-

ducing colonic gas volume by utilising H2 as a reducing agent. There are currently few available data on the extent of nitrate reduction in the human colon or of its physiological and clinical relevance. Although it is unclear how much nitrate is available for metabolism in the human hindgut, high NO3 3 -containing diets may provide the colon with substantial quantities whilst there are also potential endogenous sources. It is possible that nitrate reduction in the gut has a certain toxic potential. For example, concern has arisen regarding the carcinogenicity of nitrosamines [43] and the association of high NO3 3 exposure with incidence of tumorigenesis. In the acidic environment of the stomach there exists the potential for endogenous formation of nitrosamines from NO3 2 produced by the bacterial reduction of NO3 3 [44^46] with endogenous or exogenous amines [27]. The extent of nitrosamine formation in the colon is not known, but may have an involvement in the induction of colon cancer [47]. Allison and Macfarlane [9] demonstrated that in the presence of NO3 3 , enteric, fermentative bacteria produced more acetate and less butyrate and this was not an e¡ect of methanogenesis inhibition. There have been similar

Table 4 Nitrate-reducing bacterial isolates were identi¢ed by 16S rRNA sequencing, from chemostat cultures incorporating antibiotic selection Culture vessel

T48

Tss

À' (no treatment)

Escherichia coli Clostridium ramosum Bacteroides vulgatus Escherichia coli

Escherichia coli Klebsiella pneumoniae

À' (nalidixic acid treatment) `B' (no treatment) `B' (nalidixic acid treatment) Two £ow Tss ), each

Escherichia coli Bacteroides vulgatus Escherichia coli

Klebsiella species Enterobacter cloacae Escherichia coli Enterobacter dissolvens Enterobacter dissolvens Shigella dysenteriae

chemostats (À' and `B') were run in parallel, each with choline as the carbon source and operated under nitrate-limited, anaerobic conditions at a rate of 0.1 h31 . Chemostat `B' was fed medium containing a ¢nal concentration of 0.5 Wg ml31 gentamicin. At each sampling time (T0 , T24 and two identical samples were taken from each chemostat. Bacteria were isolated from each sample using selective media, but the second sample from vessel was ¢rst treated with 200 Wg ml31 nalidixic acid for 1 h.

FEMSEC 1083 27-12-99

26


Table 5 SCFA concentrationsa produced from 48-h batch culture fermentations Substrate

Acetate Arabinogalactan Benzoate Butyrate Cellulose Choline Cystine Formate FOS Guar gum Inulin Mucin No substrate Palmitate Pectin Phenol Propionate Pyruvate Starch Xylan

SCFA concentration (mM) Lactate

Formate

Acetate

Propionate

Butyrate

ND ND ND 1.2 ( þ 1.7) ND ND ND ND 11.7 ( þ 0.4) 1.8 ( þ 0.1) 1.1 ( þ 0.2) 0.2 ( þ 0.0) ND ND ND ND ND ND 6.7 ( þ 0.7) ND

ND ND 0.3 ( þ 0.1) ND ND ND ND 5.8 ( þ 1.0) ND 0.9 ( þ 0.1) 3.3 ( þ 0.5) 0.8 ( þ 0.2) ND ND 0.3 ( þ 0.1) ND ND ND ND 0.8 ( þ 0.1)

32.4 ( þ 0.8) 21.1 ( þ 0.0) 6.1 ( þ 0.8) 2.0 ( þ 0.0) 1.1 ( þ 1.8) 9.4 ( þ 1.1) 15.590 ( þ 0.9) 3.8 ( þ 0.1) 17.0 ( þ 1.0) 8.8 ( þ 0.8) 9.2 ( þ 1.4) 16.3 ( þ 2.2) 3.8 ( þ 0.0) 2.8 ( þ 0.8) 13.6 ( þ 1.3) 4.3 ( þ 0.2) 4.1 ( þ 0.2) 13.4 ( þ 1.3) 12.3 ( þ 0.1) 8.2 ( þ 0.6)

2.8 ( þ 0.1) 3.2 ( þ 0.6) 2.3 ( þ 0.5) ND 0.0 ( þ 1.0) 0.6 ( þ 0.2) 2.5 ( þ 0.9) 0.7 ( þ 0.0) 1.1 ( þ 0.2) 3.1 ( þ 0.5) 1.7 ( þ 0.2) 5.4 ( þ 1.3) 2.8 ( þ 0.0) 1.5 ( þ 0.4) 1.7 ( þ 0.1) 0.3 ( þ 0.1) 26.4 ( þ 1.6) 2.1 ( þ 0.8) 1.2 ( þ 0.1) 1.5 ( þ 0.3)

2.2 ( þ 0.1) 5.9 ( þ 0.3) 1.0 ( þ 0.1) 41.9 ( þ 0.5) 0.8 ( þ 0.7) 0.9 ( þ 0.1) 0.8 ( þ 0.1) 0.4 ( þ 0.1) 2.3 ( þ 0.2) 3.5 ( þ 0.1) 10.8 ( þ 1.6) 9.8 ( þ 0.7) 1.7 ( þ 0.0) 1.2 ( þ 0.2) 8.0 ( þ 0.5) 1.0 ( þ 0.1) 0.2 ( þ 0.1) 5.8 ( þ 0.7) 1.2 ( þ 0.1) 2.7 ( þ 0.5)

The starting concentration of each substrate was 0.5% (w/v). ND = not detectable. a All values are means of three determinations þ S.D. (shown in parentheses).

¢ndings with pure cultures of E. coli [48], Clostridium perfringens [49], C. butyricum [50] and Propionibacterium acnes [51]. The e¡ect on butyrate production may have relevance for colonic function as this SCFA is a major energy-yielding substrate for epithelial cells [52]. Butyrate promotes the di¡erentiation of malignant colonocytes [53] and, in rats, de¢ciency results in the onset of clinical and biochemical lesions characteristic of ulcerative colitis [54]. Given the ostensibly `protective' nature of butyrate, reduced levels in the gut may contribute towards predisposing factors for onset of bowel cancer [55]. Ammonia is thought to be toxic to mammalian cells [56]. This may further implicate a pathogenic nature for NO3 3 reduction. However, the process is likely to be quantitatively unimportant when compared to NH3 formed from amino acid fermentation [57]. On the other hand, nitrate reduction has the potential to reduce gas volume in the large intestine and may therefore be of some bene¢t to the host. Since it would appear that nitrate reduction has potentially positive as well as harmful e¡ects for the host, it is pertinent to determine the extent of this process in the human colon. This study has used batch and continuous culture systems to identify the microbial diversity involved in dissimilatory nitrate reduction. The bacteria generated were tested for their NO3 3 -reducing capacity in pure culture. Predominant species involved were identi¢ed through phenotypic and genotypic criteria. Essentially, the microbial species involved were faculta-

tively anaerobic enterobacteria, which are normally present in the gut in low numbers [58]. However, exposure of the in vitro systems to commonly used antibiotics allowed a more widespread nitrate-reducing diversity to be isolated. These data do indicate the possible e¡ects of antimicrobial use on normal gut £ora function, but only a limited number of bacterial species were involved in nitrate reduction. Whilst the biological impact of dissimilatory nitrate reduction in the human gut remains to be determined, our data show that the representatives of only six bacterial genera may be involved. Of these, two (Bacteroides vulgatus and Clostridium ramosum) are recognised as strictly anaerobic and were evident only after antimicrobial exposure. References [1] Levitt, M.D., Gibson, G.R. and Christl, S.U. (1995) Gas metabolism in the large intestine. In: Human Colonic Bacteria: Role in Nutrition, Physiology and Pathology (Gibson, G.R. and Macfarlane, G.T., Eds.), pp. 131^154. CRC Press, Boca Raton, FL. [2] Gibson, G.R. (1994) Gas metabolism by human colonic bacteria and consequences for the host. In: The W.H. Pierce Memorial Symposium (Sussman, M., Ed.), pp. 25^29. J. Appl. Bacteriol./Unipath, London. [3] Kirk, E. (1949) The quantity and composition of human colonic £atus. Gastroenterology 12, 782^794. [4] Levitt, M.D. (1969) Production and excretion of hydrogen gas in man. New Engl. J. Med. 281, 122^127.

FEMSEC 1083 27-12-99

N.J. Parham, G.R. Gibson / FEMS Microbiology Ecology 31 (2000) 21^28 [5] Levitt, M.D. (1971) Volume and composition of human intestinal gas determined by means of an intestinal washout technique. New Engl. J. Med. 284, 1394^1398. [6] Christl, S.U., Gibson, G.R. and Cummings, J.H. (1992) Role of dietary sulphate in the regulation of methanogenesis in the human large intestine. Gut 33, 1234^1238. [7] Christl, S.U., Murgatroyd, P.R., Gibson, G.R. and Cummings, J.H. (1992) Quantitative measurement of hydrogen and methane from fermentation using a whole body calorimeter. Gastroenterology 102, 1269^1277. [8] Macfarlane, G.T. and Gibson, G.R. (1997) Carbohydrate fermentation, energy transduction and gas metabolism in the human large intestine. In: Gastrointestinal Fermentation and Ecosystems (Mackie, R.I. and White, B.A., Eds.), pp. 269^318. Chapman and Hall, London. [9] Allison, C. and Macfarlane, G.T. (1988) E¡ect of nitrate on methane production and fermentation by slurries of human faecal bacteria. J. Gen. Microbiol. 134, 1397^1405. [10] Radcli¡e, B.C., Hall, C. and Roediger, W.E.W. (1989) Nitrite and nitrate levels in ileostomy e¥uent : e¡ect of dietary change. Br. J. Nutr. 61, 323^330. [11] Cummings, J.H. and Macfarlane, G.T. (1991) The control and consequences of bacterial fermentation in the human colon. J. Appl. Bacteriol. 70, 443^459. [12] Gibson, G.R. and Macfarlane, G.T. (1995) Mechanisms and clinical consequences of H2 disposal in the human colon. In: Medical and Dental Aspects of Anaerobes (Duerden, B.I., Wade, W.G., Brazier, J.S., Eley, A., Wren, B. and Hudson, M.J., Eds.), pp. 121^135. Science Reviews, Northwood. [13] Cole, J.A. (1988) Assimilatory and dissimilatory reduction of nitrate to ammonia. In: Symposium 42: The Nitrogen and Sulphur Cycles (Cole, J.A. and Ferguson, S.J., Eds.), pp. 281^329. Cambridge University Press, Cambridge. [14] Coleman, K.J., Newman, B.M., Cornish-Bowden, A.J. and Cole, J.A. (1978) Nitrite reduction by bacteria. In: Microbiology (Schilessinger, D., Ed.), pp. 334^338. American Society for Microbiology, Washington, DC. [15] Archer, M.C., Saul, R.L., Lee, L.J. and Bruce, W.R. (1981) Analysis of nitrate, nitrite and nitrosamines in human feces. In: Banbury Report 7, Gastrointestinal Cancer: Endogenous Factors (Bruce, W.R., Correa, P., Lipkin, M.R., Tannenbaun, S.R. and Wilkins, T.D., Eds.), pp. 321^330. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [16] Forsythe, S.J. and Cole, J.A. (1987) Nitrite accumulation during anaerobic nitrate reduction by binary suspensions of bacteria isolated from the achlorhydric stomach. J. Gen. Microbiol. 133, 1845^ 1849. [17] Darwin, A., Tormay, P., Page, L., Gri¤ths, L. and Cole, J. (1993) Identi¢cation of the formate dehydrogenases and genetic determinants of formate-dependent nitrite reduction by Escherichia coli K12. J. Gen. Microbiol. 139, 1829^1840. [18] Walker, R. (1975) Naturally occurring nitrate/nitrite in foods. J. Sci. Food Agric. 26, 1735^1742. [19] Knight, T.M., Forman, D., Al-Dabbagh, S.A. and Doll, R. (1987) Estimation of dietary intake of nitrate and nitrite in Great Britain. Food Chem. Toxicol. 25, 277^285. [20] Vega, M. and Bontoux, L. (1997) Nitrates in foodstu¡s: A foodsafety issue. IPTS Rep. 19, 5^11. [21] Bartholomew, B. and Hill, M.J. (1984) The pharmacology of dietary nitrate and the origin of urinary nitrate. Food Chem. Toxicol. 22, 789^795. [22] Florin, T.H.J., Neale, G. and Cummings, J.H. (1990) The e¡ect of dietary nitrate on nitrate and nitrite excretion in ileal e¥uent and urine in man. Br. J. Nutr. 64, 387^397. [23] Mitchell, H.H., Shonle, H.A. and Grindley, H.S. (1916) The origin of the nitrates in the urine. J. Biol. Chem. 24, 461^490. [24] Tannenbaum, S.R., Fett, D., Young, V.R., Land, P.D. and Bruce,

[25]

[26]

[27]

[28] [29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

27

W.R. (1978) Nitrate and nitrite are formed by endogenous synthesis in the human intestine. Science 200, 1487^1489. Green, L.C., Ruiz De Luzuriaga, K., Wagner, D.A., Rand, W., Istfan, N., Young, V.R. and Tannenbaum, S.R. (1981) Nitrate biosynthesis in man. Proc. Natl. Acad. Sci. USA 78, 7764^7768. Thayer, J.R., Chasko, J.H., Swartz, L.A. and Parks, N.J. (1982) Gut reactions of radioactive nitrite intratracheal administration in mice. Science 217, 151^153. Stuehr, D.J. and Marletta, M.A. (1985) Mammalian nitrate biosynthesis: Mouse macrophages produce nitrite and nitrate in response to Escherichia coli lipopolysaccharide. Proc. Natl. Acad. Sci. USA 82, 7738^7742. Beeken, W., Northwood, I., Beliveau, C. and Gump, D. (1987) Phagocytes in cell suspensions of human colon mucosa. Gut 28, 976^980. Iyengar, R., Stuehr, D.J. and Marletta, M.A. (1987) Macrophage synthesis of nitrite, nitrate, and N-nitrosamines: Precursors and role of the respiratory burst. Proc. Natl. Acad. Sci. USA 84, 6369^ 6373. Witter, J.P., Balish, E. and Gatley, S.J. (1979) Distribution of nitrogen-13 from labelled nitrate and nitrite in germ-free and conventional £ora rats. Appl. Environ. Microbiol. 38, 870^878. Lee, K., Greger, J.L., Consaul, M.S., Graham, M.S. and Chinn, M.S. (1986) Nitrate, nitrite balance, and de novo synthesis of nitrate in humans consuming cured meats. Am. J. Clin. Nutr. 44, 188^194. Macfarlane, G.T., Hay, S. and Gibson, G.R. (1989) In£uence 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. Wang, X. and Gibson, G.R. (1993) E¡ects of the in vitro fermentation of oligofructose and inulin by bacteria growing in the human large intestine. J. Appl. Bacteriol. 75, 373^380. Holdeman, L.V., Cato, C.P. and Moore, W.E.C. (Eds.) (1977) Anaerobic Laboratory Manual, 4th edn. VPI Anaerobe Laboratory, Blacksburg, VA. Mills, C.K., Afable, C., Candanoza, C., Ferreira, V., Landon, S., Repaske, A., Scully, J. and Gherna, R. (1989) Comparison of the Merckoquant 10007 nitrite test strip with conventional reagents in the detection of nitrate reduction by bacteria. J. Microbiol. Methods 9, 233^237. Masson, S., Sciaky, M., Desmoulin, F., Fontanarava, E. and Cozzone, P.J. (1991) Simple cation-exchange high-performance liquid chromatography optimized to the measurement of metabolites in the e¥uents from perfused rat livers using refractive index and ultraviolet detectors. J. Chromatogr. 563, 231^242. Hutson, R.A., Thompson, D.E. and Collins, M.D. (1993) Genetic interrelationships of saccharolytic Clostridium botulinum types B, E and F and related clostridia as revealed by small-subunit rRNA gene sequences. FEMS Microbiol. Lett. 108, 103^110. Devereux, J., Haeberli, P. and Smithies, D. (1984) A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12, 387^395. Argyle, J.L., Rapp-Giles, B. and Wall, J.D. (1992) Plasmid transfer by conjugation in Desulfovibrio desulfuricans. FEMS Microbiol. Lett. 94, 255^262. Steenkamp, D.J. and Peck, H.D. Jr. (1980) The association of hydrogenase and dithionite reductase activities with the nitrite reductase of Desulfovibrio desulfuricans. Biochem. Biophys. Res. Commun. 94, 41^48. van Dongen, W.M.A.M., Stokkermans, J.P.W.G. and van der Berg, W.A.M. (1994) Genetic manipulation of Desulfovibrio. In: Inorganic Sulfur Metabolism. Peck Jr., H.D. and Legall, J. (Eds) Methods in Enzymology, Vol. 243. Academic Press, New York, pp. 319^330. Cole, J. (1996) Nitrate reduction to ammonia by enteric bacteria : redundancy, or a strategy for survival during oxygen starvation? FEMS Microbiol. Lett. 136, 1^11. Bogovski, P. and Bogovski, S. (1981) Animal species in which Nnitroso compounds induce cancer. Int. J. Cancer 27, 471^474.

FEMSEC 1083 27-12-99

28


[44] Tannenbaum, S.R., Sinskey, S.R., Weisman, M. and Bishop, W. (1974) Nitrite in human saliva: Its possible relationship to nitrosamine formation. J. Natl. Cancer Inst. 53, 79^84. [45] Tannenbaum, S.R., Mora¨n, D., Rand, W., Cuello, C. and Correa, P. (1979) Gastric cancer in Colombia : IV. Nitrite and other ions in gastric contents of residents from a high risk region. J. Natl. Cancer Inst. 62, 9^12. [46] Saul, R.L., Kabir, S.H., Cohen, Z., Bruce, W.R. and Archer, M.C. (1981) Re-evaluation of nitrate and nitrite levels in the human intestine. Cancer Res. 41, 2280^2283. [47] Suzuki, K. and Mitsuoka, T. (1984) N-Nitrosamine formation by intestinal bacteria. In: N-Nitroso compounds: Occurrence, Biological E¡ects and Relevance to Human Cancer (O'Neill, I.K., Von Borstel, R.C., Miller, C.T., Long, J. and Bartsch, H., Eds.), pp. 275^281. IARC Scienti¢c Publications, Lyon. [48] Cole, J.A. and Brown, C.M. (1980) Nitrite reduction to ammonia by fermentative bacteria : a short circuit in the biological nitrogen cycle. FEMS Microbiol. Lett. 7, 65^72. [49] Ishimoto, M., Umeyama, M. and Chiba, S. (1974) Alteration of fermentation products from butyrate to acetate by nitrate reduction in Clostridium perfringens. Z. Mikrobiol. 17, 501^506. [50] Keith, S.M., Macfarlane, G.T. and Herbert, R.A. (1982) Dissimilatory nitrate reduction by a strain of Clostridium butyricum isolated from estuarine sediments. Arch. Microbiol. 132, 62^66.

[51] Allison, C. and Macfarlane, G.T. (1987) Studies of dissimilatory nitrate reduction by a colonic strain of Propionibacterium acnes. In: Proceedings of the 107th Ordinary Meeting of the Society for General Microbiology, pp. 19, St. Andrews. [52] Roediger, W.E.W. (1980) The role of anaerobic bacteria in the metabolic welfare of the colonic mucosa in man. Gut 21, 793^798. [53] Augeron, C. and Laboisse, C.L. (1984) Emergence of permanently di¡erentiated cell clones in a human colonic cancer cell line in culture after treatment with sodium butyrate. Cancer Res. 44, 3961^3966. [54] Roediger, W.E.W. and Nance, S. (1986) Metabolic induction of experimental ulcerative colitis by inhibition of fatty acid oxidation. Br. J. Exp. Pathol. 67, 773^782. [55] Forman, D., Al-Dabbagh, S. and Doll, R. (1985) Nitrates, nitrites and gastric cancer in Great Britain. Nature 313, 620^625. [56] Visek, W.J. (1978) Diet and cell growth modulation by ammonia. Am. J. Clin. Nutr. 31, 216^220. [57] Macfarlane, G.T., Cummings, J.H. and Allison, C. (1986) Protein degradation by human intestinal bacteria. J. Gen. Microbiol. 132, 1647^1656. [58] Fuller, R. and Gibson, G.R. (1997) Modi¢cation of the intestinal micro£ora using probiotics and prebiotics. Scand. J. Gastroenterol. 32 (Suppl. 222), 28^31.

FEMSEC 1083 27-12-99