- Email: [email protected]
Most probable number enumeration of H2-utilizing acetogenic bacteria from the digestive tract of animals and man
ELSEVIER
FEMS Microbiology
Letters 130 (I 405) 7- 12
Most probable number enumeration of H ,-utilizing acetogenic bacteria from the digestive tract of animals and man J. Dark aT* , B. Morvan ‘, F. Rieu-Lesme ‘, I. Goderel ‘, P. Gouet b, P. Pochart ’ ” [NRA. CR de Jouy-enJosa.s. Laboratoire de Nutrition et SdcuritC A&nmtoire, 78352 Jouy-enJonas Ceder, France h [NRA. CR de Their, 63122 Saint-Genl,v-Champanelle, France ’ Consewatoire Natlonul des Arts et M&iers, 75003 Paris, France Received 30 December lYY4; revised 20 April IY95; accepted 20 April 1995
Abstract A method is proposed that allows the enrichment and most probable number estimation of H,/CO,-utilizing acetogenic bacteria. It is based on the difference in acetate production for serial dilutions incubated under either a test HZ/CO, (4:1), or a control NJCO, (4:l) headspace atmosphere. A nutritionally non-selective medium was used, containing bromoethanesulfonic acid as inhibitor of methanogenic archaea and 10%~pre-incubated clarified rumen fluid. Acetogenic bacteria were enumerated in rumen and hindgut contents of animals and in human feces. They ranged from below 10’ to above 10’ per gram wet weight gut content and their population levels were the highest in the absence of methanogenesis. The method described therein should prove useful to better understand the diversity and ecological importance of dominant gut acetogens. Keywrdv:
Acetogenic anacrobe; Hydrogen; Gut microbe; Rumcn; Hindgut: fluman colon
1. Introduction Since the isolation of Clostridium aceticum [l], the physiology of hydrogen-oxidizing, carbon-dioxide reducing, acetogenic anaerobes has been well documented [2,3]. They derive energy from the reaction: 4H2 + 2HC0, + CH,COO
+ H _ + 4H,O (reaction
1)
The recognized diversity within this microbial group has been expanding owing to the isolation of several
species from various ecosystems such as sewage sludge and sediments [2]. The contribution of acetogenesis in anoxic sediments has Hz/CO, nevertheless been reported to account for less than 5% of H, uptake [4]. On the other hand, the contribution of HZ/CO2 acetogenesis to H, recycling in animal guts may be much more important and its nutritional significance has been stressed in recent reviews [5-71. Noticeably, while methanogenesis appears to be the predominant electron sink in various
sites of animal guts, Hz/CO2 acetogenesis may contribute to as much as 27% and 33% of total acetate production
* Corresponding author. Tel.: + 3.7 (I ) 34 65 27 09; Fax: + 33 (I) 34 65 23 11 0378-1097/95/$09.50 G lYY.5 Federation of kuropcan Microbiological SSDI 0371(-10’)7(9j)O017h-X
in the human colon and the ter-
mite gut, respectively [6]. In spite of this, the current understanding of the ecology of H,/CO, acetogenic gut microbes is still very poor. i few strains have So&tics.
All rights reserved
8
J. DO& rt al. / FEMS Microbiology
been isolated from the rumen of sheep and cattle. and from the gut of termites and man [6-91. The activity and relative counts of the corresponding populations in vivo have been inferred from these isolations. A plating method designed to selectively enumerate acetogenic bacteria [lOI was used for anaerobic sediments. Its intrinsic sensitivity (one acetogen per 250 total anaerobes) did not permit the investigation of a large range of population levels. Because acetogenie gut bacteria are of definite interest, we have developed a highly sensitive, specific procedure for enumerating and enriching these bacteria from digestive contents of animals and man, which facilitates the comparison of their distribution to that of other HZ-utilizing microorganisms. (A preliminary report of these findings has been presented (J. Dori et al. (1993) Abstr. Annu. Meet. Am. Sot. Microbial.))
2. Materials and methods 2.1. Habitats and sample preparation Rumen samples obtained from fistulated adult animals fed a hay-based diet and from young suckling lambs using oesophageal tubes were processed within 2 h. Cecal and colonic contents of growingfinishing pigs were collected at slaughter using CO,-flushed glass containers, and processed within 3 h. Freshly voided stools were obtained from two female Wistar rats (IFFA CREDO, Orleans, France) fed commercial laboratory chow (A04,VAR, Villemoisson-sur-Orge, France). Stools from only one of the two rats were methanogenic upon incubation (1 h) at 37°C in a 10 ml, Hz/COz-flushed syringe. Fecal specimens were obtained from four healthy human volunteers, aged 22-35, eating unrestricted western meals. Two volunteers were repeatedly tested positive breath methane producers (CH, -t ; range 5-21 ppm [ll]) while the other two consistently had breath methane levels that never exceeded the trace levels present in ambient air controls by more than one part per million (CH, - ). Human volunteers gave informed consent to the work, which was approved by the local Ethics Committee. Fecal samples were processed within 1 h. Sludge samples were collected from the municipal anaerobic digester of
Letters I.30 (1995) 7-12
Lezoux (Clermont-Ferrand). The latter were stored in filled glass containers and processed within 2 h. Samples were homogenized in stoppered vials using magnetic stirring (1000 rpm) for at least 2 min under an anaerobic 100% CO2 atmosphere (CFPO, France) to give a ten-fold dilution (wet weight/ volume) and serially diluted in anaerobic basal medium containing all the components of the enrichment medium described below except yeast extract and rumen fluid. Serial dilutions down to 10-l’ were prepared. 2.2. Media and enumeration procedures All liquid and solid media were prepared using strict anaerobic techniques [ 121.All preparations were carried out under medical grade 100% CO, gas (CFPO, France) or after removal of traces of oxygen of analytical grade CO, by passage over a copper column heated to 350°C. After inoculation, the headspace atmosphere was replaced with HZ/CO, (4:l) at an initial pressure of 202 kPa for Hz-utilizing microorganisms. The medium for enumeration of acetogenic microorganisms was derived from AC-11 medium [a]. It contained per liter of distilled water: 0.1% resazurin, 1 ml; NH,Cl, 0.5 g; KH, PO,, 0.28 g; K,HPO,, 0.94 g; NaCl, 0.14 g; KCl, 0.16 g; MgSO, 7HZ0, 0.02 g; trace metals, 10 ml; vitamins, 10 ml; yeast extract, 0.5 g; incubated clarified rumen fluid, 100 ml [13]; cysteine-sulfide reducing agent 20 ml; NaHCO,, 6 g. The trace metals and vitamins solutions were as described by Greening and Leedle [14]. A sodium tungstate solution (10 PM) was added as 0.1% (v/v) in the case of animal gut contents. The cysteine-sulfide reducing agent was prepared under 100% N, as described by Braun et al. [lo] to contain 1.25% (w/v) each cysteine . HCl H ?O and Na ZS . 9H ,O. Stoppered, crimp-sealed culture tubes (18 X 150 mm; Bellco Glass Co., Vineland, NJ) were each supplemented with 2% (v/v) of a 2.5 M anaerobic filter-sterilized solution of 2bromoethanesulfonic acid, sodium salt (BESA) just prior to inoculation. TWO series of triplicate tubes were inoculated (0.3 ml in 5 ml medium) and incubated for each dilution between 10e3 and lo-“. The headspace of each tube of the control series was brought at 202 kPa initial pressure using N/CO, (4:1), while the test series was treated similarly using HZ/CO, (4:l). Acetate was analysed in the super-
J. Do& ef al. / FEMS Mmwhiology
natant of each tube as detailed below, after 20 days of incubation. The number of positive or negative tubes recorded for Most Probable Number (MPN) estimation was based on the difference in acetate concentration between each tube of a test series and the average acetate concentration in the control tubes of the corresponding dilution. Estimations of MPN were made according to Clarke and Owens [15]. Total anaerobes were enumerated by the roll tube technique using the medium of Leedle and Hespell [13], except for those from rat and human feces which were counted as formerly described [16] using Wilkins-Chalgren agar (Difco Lab., Detroit, MI). Dilutions between lo-’ and lo- ’ ’ were used. Pressurized tubes were incubated horizontally. Incubations were kept in the dark at 37°C for hindgut contents, 39°C for rumen contents and 30°C for sludge samples. 2.3. Analytical methods The acetate concentration in the culture supernatant was measured by gas liquid chromatography, or by an enzymatic method (Boehringer, Mannheim, Germany). Methane was determined by gas chromatography using an IGC-121 DFL-type (Intersmat. Courtry, France) or a Shimadzu GC8A-type (Touzart et Matignon, Vitry, France) gas chromatograph equipped with a molecular sieve 5A-60/80 mesh packed 2 m stainless-steel column (Touzart et Matignon, Vitry, France), flame ionization detector and using 20 ml/min Nz or Ar as carrier. Injector, oven and detector temperatures were set at 120°C. A 60 ppm CH, standard was used for excreted methane quantitation. Where mentioned, gas-phase consumption was assessed by plunger displacement using a 50-ml lubricated glass syringe, or by direct manometric measurement using a Capsuhelic-type manometer (Dwyer Instruments, Inc., Michigan City. MI).
3. Results 3.1. Most probable number series for acetogenic bacteria When measured, gas consumption was found to decrease sharply (2-0.2 bars) within 8-10 days fol-
9
Letters I.30 (1995) 7-12
2
10
" 1
2
3
lo-fold
4
5
6
dilution
7
8
9
10
factor
Fig. 1. Example of MPN enumeration of H2-dependent acetogenic bacteria in a human fecal sample. Dots indicate acetate concentrations in individual H2 /COz-incubated tubes. Bars and lines indicate average acetate concentrations in control N1 /CO?-incubatcd tubes and 95% confidence intervals, respectively. Dilutions 1-6, 7-X and Y-10 gave 3/3, 2/3 and O/3 positive H,-dependent acetogenic incubations, respectively, corresponding to a population estimate for acetogens of 7.0~ lO’/g wet weight; 95% confidence interval: 1.9X lo’-2.5 X 10’. Gas consumption ( kmol) and absorbance at 600 run for H2 /CO, incubations of dilutions 6, 7, 8 and 9 were: 260 + 18 and 0.414 f 0.011. 162 f 59 and 0.355+0.009, 122+50 and 0.319~0.016, 12*5 and 0.297 & 0.002, respectively.
lowing inoculation of the HZ/CO, acetogens enumeration medium with dilutions from all gut sources. Bacterial growth after 20 days, as measured by optical density COD,,,,), was usually 0.300-0.500, except for the lo-’ dilution. Methane was never detected in the headspace after growth. The acetate concentration in the enumeration medium inoculated with the lo-’ dilution was approximately 5 mM prior to incubation. The final acetate concentration in the control NZ/CO, series was below 12 mM. It was markedly influenced by the presence of HZ/CO2 in the gas phase in all the lower dilutions tested from various habitats, and could reach 50 mM or more (Fig. l), approximate values near the maximum theoretical H,-dependent acetate accumulation of 100 mM under our experimental conditions. For each dilution, the concentrations of acetate accumulated in the presence of HZ/CO2 were compared to the statistical distribution of concentrations accumulated under N2/C02. The former decreased with end-point dilutions, and levels within the 95% confidence interval of acetate concentrations accumulated under N2/C02 were eventually observed for all the series analysed (above lo-* dilution in example of Fig. 1).
10
J. Do& et al. / FEMS MicrohioioRy Letters 130 (1995) 7-12
Using this later criterion as an indication of the absence of Hz-dependent acetogenesis, the consistency of observations allowed the MPN estimation of HZ-dependent acetogens in all samples characterized. Neither gas consumption nor OD,,,,, did appear as sufficiently precise criteria for the purpose of enumeration. Gas consumption was nonetheless used as a means of positive selection for subculturing and isolation. Although gas consumption rarely led to the observation of negative pressures in the MPN series, this occasionally occurred within a few days of incubation after the 2nd transfer from positive MPN tubes. 3.2. Counts of H, /CO,
acetogenic microorganisms
Using the procedure described above, H ,-utilizing acetogens have been enumerated from most of the
Table 1 Distribution of Hz/CO:-acetogenic Habitats
Rumen 1-day-old lambs 2-week-old lambs Adult sheep steer Cecum Pigs b Colon Pigs Feces (CH, + 1 ’ Rats Humans Feces (CH, - 1 Rats Humans Digester sludge
gut samples studied. The populations observed are reported in Table 1. In a few instances they were below the detection limit of the dilution series. The highest population levels recorded in this study were in feces of non-methane-excreting human individuals (2.5 X lO’/g wet weight). Among all sites of animal gut contents investigated, the highest population levels were then found in feces of non-methane-excreting rats, in the rumen of l-day-old lambs and pig hindgut contents. They never exceeded total numbers of 8.0 X lob/g wet weight of animal gut contents. The highest ratio of acetogens to total anaerobes were observed in human feces and rumen contents of l-day-old lambs (l/7 and l/31, respectively). As a comparison, populations of 2.0 X lo4 acetogens/g were observed in sludge samples containing 1.3 X 10’ total anaerobes/g (approx. l/650).
bacteria in various anaerobic environments Ratio of acetogens/total
Counts (g wet weight- ’ ) a Total anaerobes
Acetogens
3.‘) x 1.1 x 2.9 x 3.2 x 1.6 x 1.4 x 1.3 x
3.1 x 10s 3.6 x 10’ < 10” < IO3 < 10’ < lo3 < 10’
(7.5 x IO?-1.3 x 106) (7.5 x lo’-1.3 x 106) -
l/1258 l/31 < 1om5 < 10-s < 1om5 < 1o-5
1.6 X 10’
< 3.2 x 10’
(7.5 x 10”-1.3 x 106)
< l/500
4.0 x IO”
1.6 x 10”
(3.4 X lo”-6.0
l/25 000
4.0 x 10” 4.0 x 10”’ 3.2 x 10”’
< 10’ 3.1 x 10’ < 10’
(7.5 x lo?-1.3 x 104) -
< 10-s < 10-s < 10-s
3.2 2.0 6.3 5.4 2.1
8.0 x 3.1 x 7.2 x 1.4 x 3.1 x
(2.2 (7.5 (1.9 (3.4 (7.5
l/400 l/7 l/875 l/386 l/677
x x x x x
IOX 10” lov 10’ 10’ 10”’ IO4
10” 10’ 10”’ 10h 10’
lo6 loK 10’ lo4 lo4
x x x x x
IOh-3.0 lo’-1.3 lo’-2.5 loa-5.9 lo’-1.3
X 10’)
x x x x x
10’) 109) 108) 10”) 10’)
a Values marked ’ < ’ were below the detection limit indicated for the experiment. It was otherwise set at 103. Counts of acetogenic bacteria are indicated as MPN estimates in microorganisms per gram wet content (95% confidence interval) [15]. b Pig gut contents were pools of total caecum or distal colon collected at slaughter from three animals. ’ The methane-excreting status, as indicated, refers to the detection of methane concentrations in expired air at least one part per million in excess of trace levels present in ambiant air (CH, + ) or below (CH, - ) [l 11.
J. DorP CI al. / FEMS Microhiolop
3.3. isolation of H, /CO, isms
acetogenic microorgan-
For isolation of microorganisms utilizing HJCO,, enrichments maintained on the enumeration broth medium by biweekly subcultures were transferred to roll tubes containing the same medium with 20 g/l Bacto agar (Difco Lab., Detroit, MI). Gas phase depletion was used as a positive selection criterion. Colonies picked from roll tubes were either transferred back to broth medium until further purification on roll tubes, or streaked on slants of the same solid medium. We observed that the addition of 10 g/l CaCO,, as used by Balch et al. [2] in the isolation of Acetobacterium woodii, led to a much faster gas phase consumption. However, it did not allow the detection of acidogenic colonies by clearing zone formation. Colonies remained viable for several months on slants, although gas phase consumption was rapid only after the first few pressurizations following transfer. Purity was established based on colony type homogeneity and microscopic examinations of living and Gram-stained cells after autotrophic or heterotrophic growth. Three isolates (FRAl, FRCl, SER8) from the rumen of two lambs were obtained from lo-“-10 -’ dilutions. They produced acetate from H2 and CO, in amounts expected from the Acetyl-CoA pathway of reaction 1. The strains were spore-forming Gram-positive rods resembling Clostridium aceticum, short chain forming Gram-negative cocco-bacilli and Gram-positive cocci, respectively. Gram-positive cocci have also been enriched as the dominant morphology from the highest positive dilution of an enumeration broth series from another l-day-old lamb. but they could not be isolated successfully.
4. Discussion The methodology described in the present paper allowed the enumeration of H,/CO, acetogens in all the habitats investigated, even at low levels. Their abundance in the digester sludge samples was comparable to that formerly reported [lo]. We were able to isolate strains of acetogens from the higher positive dilutions from the enumeration series. Populations within the rumen of adult animals appeared
Letters 130 II 99.5) 7-12
11
close to our lower detection limit. This may explain the impossibility of enumerating the microorganisms involved by plating onto solid media [lo]. This also agrees with an accepted minor contribution of acetogenesis to H, recycling in the rumen where methanogenesis is consistently present in adults [6]. On the contrary, in the hindgut of pigs, rats and man for which acetogenesis has been evidenced by labelled HCO, /CO, incorporations [5,6,17], populations were as high as 105-lOa/g. Levels of acetogenic bacteria were apparently negatively correlated to the methanogenic activity of the sample. This was true in the lumen of lambs, where they were detected prior to the natural establishment of methanogenic archaea [18], as well as in the rat and human colon where variable levels of methanogenic archaea are encountered in different individuals [6,16]. The highest populations were observed in the feces of non-methane-excreting individuals. This supports the potential contribution of acetogens to overall H, recycling in the human colon of non-methane excreting subjects, first observed by Lajoie et al. [5,7]. One of the weaknesses of MPN enumerations performed with low numbers of repeats, imposed by anaerobic methodologies, is the lack of precision in the evaluation of population levels. In spite of this, and considering the wide range of populations encountered between species as well as within a given species (e.g. age-related changes as in lambs or associated with methane excretion in humans), the ability to measure such populations appears of utmost interest. It therefore seems worthwhile to further investigate the colon of monogastrics as well as apply this method to other ecosystems such as the termite gut where acetogenesis may play an important role [6]. The diversity of the dominant H,-utilizing acetogenic populations from the human colon and the rumen or cecum of herbivores is currently under investigation (Bernalier, personal communication; [19]). Despite a large phylogenetic diversity, Hz-utilizing acetogenic bacteria all derive energy from the acetyl-CoA pathway and thereby share a set of specific enzymatic activities. In that respect, it is likely that non-culture based technologies relying on probes targeting genomic sequences specific of the acetogens will, in the near future, extend current means of investigation and overcome limitations of
.I. Do& et al. / FEMS Microbiology
12
classical
microbiological
enumeration
techniques
ml.
References [l] Wieringa, K.T. (1940) The formation of acetic acid from carbon dioxide and hydrogen by anaerobic spore-forming bacteria. Antonie van Leeuwenhoek J. Microbial. Serol. 6. 251-262. [2] Dolfing, J. (1988) Acetogencsis. In: Biology of Anaerobic Microorganisms (Zehnder, A.J.B., Ed.), pp. 417-468. John Wiley and Sons, New York, NY. [3] Wood, H.G. and Ljungdahl. L.G. (1991) Autotrophic character of the acetogenic bacteria. In: Variations in Autotrophic Life (Shively, J.M. and Barton, L.L., Eds.), pp. 201-250. Academic Press, New York, NY. [4] Lovley, D.R. and Klug, M.J. (1983) Methanogenesis from methanol and methylamines and acetogenesis from hydrogen and carbon dioxide in the sediments of a eutrophic lake. Appl. Environ. Microbial. 45, 1310-1315. [5] Lajoie, SF., Shelton, B., Miller, T.L. and Wolin, M.J. (1988) Acetate production from hydrogen and [“C]carbon dioxide by the microflora of human feces. Appl. Environ. Microbial. 54, 2723-2727. [6] Breznak, J.A. and Kane, M.D. (1990) Microbial Hz /CO, acetogenesis in animal guts: nature and nutritional significance. FEMS Microbial. Rev. 87, 309-314. [7] Durand, M. and Bernalier, A. (1993) Reductive acetogenesis in animal and human gut. In: Physiological and Clinical Aspects of Short Chain Fatty Acid Metabolism (Cummings, J.H., Rombeau, J.L. and Sakata, T., Eds.), pp. 107-117. Cambridge University Press, Cambridge. [8] Kane, M.D. and Breznak, J.A. (1991) Acetonema longum gen. nov. sp. nov., an H, /CO, acetogenic bacterium from the termite, Pferolermes occidentis. Arch. Microbial. 1.56, 91-98. [9] Wolin, M.J. and Miller, T.L. (1993) Bacterial strains from human feces that reduce CO> to acetic acid. Appl. Environ. Microbial. 59, 3551-3556.
Lefters 130 (1995) 7-12 [lo] Braun, M., Schoberth, S. and Gottschalk, G. (1979) Enumeration of bacteria forming acetate from H, and CO, in anaerobic habitats. Arch. Microbial. 120, 201-204. [ll] Bond, J.H., Engel, R.R. and Levitt, M.D. (1971) Factors influencing pulmonary methane excretion in man. J. Exp. Med. 133, 572-588. [12] Balch, W.E. and Wolfe, R.S. (1976) A new approach to the cultivation of methanogenic bacteria: 2-mercaptoethanesulfonic acid (HS-CoM)-dependent growth of Merhanobacrerium ruminantium in a pressurized atmosphere. Appl. Environ. Microbial. 32, 781-791. [13] Leedle, J.A.Z. and Hespell, R.B. (1980) Differential carbohydrate media and anaerobic replica plating techniques in delineating carbohydrate-utilizing subgroups in rumen bacterial populations. Appl. Environ. Microbial. 39, 709-719. [14] Greening. R.C. and Leedle, J.A.Z. (1989) Enrichment and isolation of Acetitomaculum ruminis, gen. nov., sp. nov.: acetogenic bacteria from the bovine rumen. Arch. Microbial. 151, 399-406. [15] Clarke, K.R. and Owens, N.J.P. (1983) A simple and versatile micro-computer program for the determination of ‘Most Probable Number’. J. Microbial. Methods 1, 133-137. [16] Pochart, P., Dore, J., Lemann, F., Goderel, I. and Rambaud, J.-C. (1992) Interrelations between populations of methanogenic archaea and sulfate-reducing bacteria in the human colon. FEMS Microbial. Lett. 98, 225-228. [17] Bernalier, A., Doisneau, E., Cordelet, C., Beaumatin, P., Durand, M. and Grivet, J.P. (1993) Competition for hydrogen between methanogenesis and hydrogenotrophic acetogenesis in human colonic flora studied by 13C-NMR. Proc. Nutr. Sot. 52, 118A. [18] Morvan, B., Dore, J., Rieu-Lesme, F., Foucat, L., Fonty, G. and Gouet, P. (1994) Establishment of hydrogen-utilizing bacteria in the rumen of the newborn lamb. FEMS Microbial. Lett. 117, 249-256. [19] Rieu-Lesme, F., Fonty, G. and Dore, J. (1994) Isolation and characterization of a new H,-utilizing bacterium from the rumen. FEMS Microbial. L.&t. 125, 77-82. [20] Lovell, CR. and Hui, Y. (1991) Design and testing of a functional group-specific DNA probe for the study of natural populations of acetogenic bacteria. Appl. Environ. Microbial. 57, 2602-2609.
FEMS Microbiology
Letters 130 (I 405) 7- 12
Most probable number enumeration of H ,-utilizing acetogenic bacteria from the digestive tract of animals and man J. Dark aT* , B. Morvan ‘, F. Rieu-Lesme ‘, I. Goderel ‘, P. Gouet b, P. Pochart ’ ” [NRA. CR de Jouy-enJosa.s. Laboratoire de Nutrition et SdcuritC A&nmtoire, 78352 Jouy-enJonas Ceder, France h [NRA. CR de Their, 63122 Saint-Genl,v-Champanelle, France ’ Consewatoire Natlonul des Arts et M&iers, 75003 Paris, France Received 30 December lYY4; revised 20 April IY95; accepted 20 April 1995
Abstract A method is proposed that allows the enrichment and most probable number estimation of H,/CO,-utilizing acetogenic bacteria. It is based on the difference in acetate production for serial dilutions incubated under either a test HZ/CO, (4:1), or a control NJCO, (4:l) headspace atmosphere. A nutritionally non-selective medium was used, containing bromoethanesulfonic acid as inhibitor of methanogenic archaea and 10%~pre-incubated clarified rumen fluid. Acetogenic bacteria were enumerated in rumen and hindgut contents of animals and in human feces. They ranged from below 10’ to above 10’ per gram wet weight gut content and their population levels were the highest in the absence of methanogenesis. The method described therein should prove useful to better understand the diversity and ecological importance of dominant gut acetogens. Keywrdv:
Acetogenic anacrobe; Hydrogen; Gut microbe; Rumcn; Hindgut: fluman colon
1. Introduction Since the isolation of Clostridium aceticum [l], the physiology of hydrogen-oxidizing, carbon-dioxide reducing, acetogenic anaerobes has been well documented [2,3]. They derive energy from the reaction: 4H2 + 2HC0, + CH,COO
+ H _ + 4H,O (reaction
1)
The recognized diversity within this microbial group has been expanding owing to the isolation of several
species from various ecosystems such as sewage sludge and sediments [2]. The contribution of acetogenesis in anoxic sediments has Hz/CO, nevertheless been reported to account for less than 5% of H, uptake [4]. On the other hand, the contribution of HZ/CO2 acetogenesis to H, recycling in animal guts may be much more important and its nutritional significance has been stressed in recent reviews [5-71. Noticeably, while methanogenesis appears to be the predominant electron sink in various
sites of animal guts, Hz/CO2 acetogenesis may contribute to as much as 27% and 33% of total acetate production
* Corresponding author. Tel.: + 3.7 (I ) 34 65 27 09; Fax: + 33 (I) 34 65 23 11 0378-1097/95/$09.50 G lYY.5 Federation of kuropcan Microbiological SSDI 0371(-10’)7(9j)O017h-X
in the human colon and the ter-
mite gut, respectively [6]. In spite of this, the current understanding of the ecology of H,/CO, acetogenic gut microbes is still very poor. i few strains have So&tics.
All rights reserved
8
J. DO& rt al. / FEMS Microbiology
been isolated from the rumen of sheep and cattle. and from the gut of termites and man [6-91. The activity and relative counts of the corresponding populations in vivo have been inferred from these isolations. A plating method designed to selectively enumerate acetogenic bacteria [lOI was used for anaerobic sediments. Its intrinsic sensitivity (one acetogen per 250 total anaerobes) did not permit the investigation of a large range of population levels. Because acetogenie gut bacteria are of definite interest, we have developed a highly sensitive, specific procedure for enumerating and enriching these bacteria from digestive contents of animals and man, which facilitates the comparison of their distribution to that of other HZ-utilizing microorganisms. (A preliminary report of these findings has been presented (J. Dori et al. (1993) Abstr. Annu. Meet. Am. Sot. Microbial.))
2. Materials and methods 2.1. Habitats and sample preparation Rumen samples obtained from fistulated adult animals fed a hay-based diet and from young suckling lambs using oesophageal tubes were processed within 2 h. Cecal and colonic contents of growingfinishing pigs were collected at slaughter using CO,-flushed glass containers, and processed within 3 h. Freshly voided stools were obtained from two female Wistar rats (IFFA CREDO, Orleans, France) fed commercial laboratory chow (A04,VAR, Villemoisson-sur-Orge, France). Stools from only one of the two rats were methanogenic upon incubation (1 h) at 37°C in a 10 ml, Hz/COz-flushed syringe. Fecal specimens were obtained from four healthy human volunteers, aged 22-35, eating unrestricted western meals. Two volunteers were repeatedly tested positive breath methane producers (CH, -t ; range 5-21 ppm [ll]) while the other two consistently had breath methane levels that never exceeded the trace levels present in ambient air controls by more than one part per million (CH, - ). Human volunteers gave informed consent to the work, which was approved by the local Ethics Committee. Fecal samples were processed within 1 h. Sludge samples were collected from the municipal anaerobic digester of
Letters I.30 (1995) 7-12
Lezoux (Clermont-Ferrand). The latter were stored in filled glass containers and processed within 2 h. Samples were homogenized in stoppered vials using magnetic stirring (1000 rpm) for at least 2 min under an anaerobic 100% CO2 atmosphere (CFPO, France) to give a ten-fold dilution (wet weight/ volume) and serially diluted in anaerobic basal medium containing all the components of the enrichment medium described below except yeast extract and rumen fluid. Serial dilutions down to 10-l’ were prepared. 2.2. Media and enumeration procedures All liquid and solid media were prepared using strict anaerobic techniques [ 121.All preparations were carried out under medical grade 100% CO, gas (CFPO, France) or after removal of traces of oxygen of analytical grade CO, by passage over a copper column heated to 350°C. After inoculation, the headspace atmosphere was replaced with HZ/CO, (4:l) at an initial pressure of 202 kPa for Hz-utilizing microorganisms. The medium for enumeration of acetogenic microorganisms was derived from AC-11 medium [a]. It contained per liter of distilled water: 0.1% resazurin, 1 ml; NH,Cl, 0.5 g; KH, PO,, 0.28 g; K,HPO,, 0.94 g; NaCl, 0.14 g; KCl, 0.16 g; MgSO, 7HZ0, 0.02 g; trace metals, 10 ml; vitamins, 10 ml; yeast extract, 0.5 g; incubated clarified rumen fluid, 100 ml [13]; cysteine-sulfide reducing agent 20 ml; NaHCO,, 6 g. The trace metals and vitamins solutions were as described by Greening and Leedle [14]. A sodium tungstate solution (10 PM) was added as 0.1% (v/v) in the case of animal gut contents. The cysteine-sulfide reducing agent was prepared under 100% N, as described by Braun et al. [lo] to contain 1.25% (w/v) each cysteine . HCl H ?O and Na ZS . 9H ,O. Stoppered, crimp-sealed culture tubes (18 X 150 mm; Bellco Glass Co., Vineland, NJ) were each supplemented with 2% (v/v) of a 2.5 M anaerobic filter-sterilized solution of 2bromoethanesulfonic acid, sodium salt (BESA) just prior to inoculation. TWO series of triplicate tubes were inoculated (0.3 ml in 5 ml medium) and incubated for each dilution between 10e3 and lo-“. The headspace of each tube of the control series was brought at 202 kPa initial pressure using N/CO, (4:1), while the test series was treated similarly using HZ/CO, (4:l). Acetate was analysed in the super-
J. Do& ef al. / FEMS Mmwhiology
natant of each tube as detailed below, after 20 days of incubation. The number of positive or negative tubes recorded for Most Probable Number (MPN) estimation was based on the difference in acetate concentration between each tube of a test series and the average acetate concentration in the control tubes of the corresponding dilution. Estimations of MPN were made according to Clarke and Owens [15]. Total anaerobes were enumerated by the roll tube technique using the medium of Leedle and Hespell [13], except for those from rat and human feces which were counted as formerly described [16] using Wilkins-Chalgren agar (Difco Lab., Detroit, MI). Dilutions between lo-’ and lo- ’ ’ were used. Pressurized tubes were incubated horizontally. Incubations were kept in the dark at 37°C for hindgut contents, 39°C for rumen contents and 30°C for sludge samples. 2.3. Analytical methods The acetate concentration in the culture supernatant was measured by gas liquid chromatography, or by an enzymatic method (Boehringer, Mannheim, Germany). Methane was determined by gas chromatography using an IGC-121 DFL-type (Intersmat. Courtry, France) or a Shimadzu GC8A-type (Touzart et Matignon, Vitry, France) gas chromatograph equipped with a molecular sieve 5A-60/80 mesh packed 2 m stainless-steel column (Touzart et Matignon, Vitry, France), flame ionization detector and using 20 ml/min Nz or Ar as carrier. Injector, oven and detector temperatures were set at 120°C. A 60 ppm CH, standard was used for excreted methane quantitation. Where mentioned, gas-phase consumption was assessed by plunger displacement using a 50-ml lubricated glass syringe, or by direct manometric measurement using a Capsuhelic-type manometer (Dwyer Instruments, Inc., Michigan City. MI).
3. Results 3.1. Most probable number series for acetogenic bacteria When measured, gas consumption was found to decrease sharply (2-0.2 bars) within 8-10 days fol-
9
Letters I.30 (1995) 7-12
2
10
" 1
2
3
lo-fold
4
5
6
dilution
7
8
9
10
factor
Fig. 1. Example of MPN enumeration of H2-dependent acetogenic bacteria in a human fecal sample. Dots indicate acetate concentrations in individual H2 /COz-incubated tubes. Bars and lines indicate average acetate concentrations in control N1 /CO?-incubatcd tubes and 95% confidence intervals, respectively. Dilutions 1-6, 7-X and Y-10 gave 3/3, 2/3 and O/3 positive H,-dependent acetogenic incubations, respectively, corresponding to a population estimate for acetogens of 7.0~ lO’/g wet weight; 95% confidence interval: 1.9X lo’-2.5 X 10’. Gas consumption ( kmol) and absorbance at 600 run for H2 /CO, incubations of dilutions 6, 7, 8 and 9 were: 260 + 18 and 0.414 f 0.011. 162 f 59 and 0.355+0.009, 122+50 and 0.319~0.016, 12*5 and 0.297 & 0.002, respectively.
lowing inoculation of the HZ/CO, acetogens enumeration medium with dilutions from all gut sources. Bacterial growth after 20 days, as measured by optical density COD,,,,), was usually 0.300-0.500, except for the lo-’ dilution. Methane was never detected in the headspace after growth. The acetate concentration in the enumeration medium inoculated with the lo-’ dilution was approximately 5 mM prior to incubation. The final acetate concentration in the control NZ/CO, series was below 12 mM. It was markedly influenced by the presence of HZ/CO2 in the gas phase in all the lower dilutions tested from various habitats, and could reach 50 mM or more (Fig. l), approximate values near the maximum theoretical H,-dependent acetate accumulation of 100 mM under our experimental conditions. For each dilution, the concentrations of acetate accumulated in the presence of HZ/CO2 were compared to the statistical distribution of concentrations accumulated under N2/C02. The former decreased with end-point dilutions, and levels within the 95% confidence interval of acetate concentrations accumulated under N2/C02 were eventually observed for all the series analysed (above lo-* dilution in example of Fig. 1).
10
J. Do& et al. / FEMS MicrohioioRy Letters 130 (1995) 7-12
Using this later criterion as an indication of the absence of Hz-dependent acetogenesis, the consistency of observations allowed the MPN estimation of HZ-dependent acetogens in all samples characterized. Neither gas consumption nor OD,,,,, did appear as sufficiently precise criteria for the purpose of enumeration. Gas consumption was nonetheless used as a means of positive selection for subculturing and isolation. Although gas consumption rarely led to the observation of negative pressures in the MPN series, this occasionally occurred within a few days of incubation after the 2nd transfer from positive MPN tubes. 3.2. Counts of H, /CO,
acetogenic microorganisms
Using the procedure described above, H ,-utilizing acetogens have been enumerated from most of the
Table 1 Distribution of Hz/CO:-acetogenic Habitats
Rumen 1-day-old lambs 2-week-old lambs Adult sheep steer Cecum Pigs b Colon Pigs Feces (CH, + 1 ’ Rats Humans Feces (CH, - 1 Rats Humans Digester sludge
gut samples studied. The populations observed are reported in Table 1. In a few instances they were below the detection limit of the dilution series. The highest population levels recorded in this study were in feces of non-methane-excreting human individuals (2.5 X lO’/g wet weight). Among all sites of animal gut contents investigated, the highest population levels were then found in feces of non-methane-excreting rats, in the rumen of l-day-old lambs and pig hindgut contents. They never exceeded total numbers of 8.0 X lob/g wet weight of animal gut contents. The highest ratio of acetogens to total anaerobes were observed in human feces and rumen contents of l-day-old lambs (l/7 and l/31, respectively). As a comparison, populations of 2.0 X lo4 acetogens/g were observed in sludge samples containing 1.3 X 10’ total anaerobes/g (approx. l/650).
bacteria in various anaerobic environments Ratio of acetogens/total
Counts (g wet weight- ’ ) a Total anaerobes
Acetogens
3.‘) x 1.1 x 2.9 x 3.2 x 1.6 x 1.4 x 1.3 x
3.1 x 10s 3.6 x 10’ < 10” < IO3 < 10’ < lo3 < 10’
(7.5 x IO?-1.3 x 106) (7.5 x lo’-1.3 x 106) -
l/1258 l/31 < 1om5 < 10-s < 1om5 < 1o-5
1.6 X 10’
< 3.2 x 10’
(7.5 x 10”-1.3 x 106)
< l/500
4.0 x IO”
1.6 x 10”
(3.4 X lo”-6.0
l/25 000
4.0 x 10” 4.0 x 10”’ 3.2 x 10”’
< 10’ 3.1 x 10’ < 10’
(7.5 x lo?-1.3 x 104) -
< 10-s < 10-s < 10-s
3.2 2.0 6.3 5.4 2.1
8.0 x 3.1 x 7.2 x 1.4 x 3.1 x
(2.2 (7.5 (1.9 (3.4 (7.5
l/400 l/7 l/875 l/386 l/677
x x x x x
IOX 10” lov 10’ 10’ 10”’ IO4
10” 10’ 10”’ 10h 10’
lo6 loK 10’ lo4 lo4
x x x x x
IOh-3.0 lo’-1.3 lo’-2.5 loa-5.9 lo’-1.3
X 10’)
x x x x x
10’) 109) 108) 10”) 10’)
a Values marked ’ < ’ were below the detection limit indicated for the experiment. It was otherwise set at 103. Counts of acetogenic bacteria are indicated as MPN estimates in microorganisms per gram wet content (95% confidence interval) [15]. b Pig gut contents were pools of total caecum or distal colon collected at slaughter from three animals. ’ The methane-excreting status, as indicated, refers to the detection of methane concentrations in expired air at least one part per million in excess of trace levels present in ambiant air (CH, + ) or below (CH, - ) [l 11.
J. DorP CI al. / FEMS Microhiolop
3.3. isolation of H, /CO, isms
acetogenic microorgan-
For isolation of microorganisms utilizing HJCO,, enrichments maintained on the enumeration broth medium by biweekly subcultures were transferred to roll tubes containing the same medium with 20 g/l Bacto agar (Difco Lab., Detroit, MI). Gas phase depletion was used as a positive selection criterion. Colonies picked from roll tubes were either transferred back to broth medium until further purification on roll tubes, or streaked on slants of the same solid medium. We observed that the addition of 10 g/l CaCO,, as used by Balch et al. [2] in the isolation of Acetobacterium woodii, led to a much faster gas phase consumption. However, it did not allow the detection of acidogenic colonies by clearing zone formation. Colonies remained viable for several months on slants, although gas phase consumption was rapid only after the first few pressurizations following transfer. Purity was established based on colony type homogeneity and microscopic examinations of living and Gram-stained cells after autotrophic or heterotrophic growth. Three isolates (FRAl, FRCl, SER8) from the rumen of two lambs were obtained from lo-“-10 -’ dilutions. They produced acetate from H2 and CO, in amounts expected from the Acetyl-CoA pathway of reaction 1. The strains were spore-forming Gram-positive rods resembling Clostridium aceticum, short chain forming Gram-negative cocco-bacilli and Gram-positive cocci, respectively. Gram-positive cocci have also been enriched as the dominant morphology from the highest positive dilution of an enumeration broth series from another l-day-old lamb. but they could not be isolated successfully.
4. Discussion The methodology described in the present paper allowed the enumeration of H,/CO, acetogens in all the habitats investigated, even at low levels. Their abundance in the digester sludge samples was comparable to that formerly reported [lo]. We were able to isolate strains of acetogens from the higher positive dilutions from the enumeration series. Populations within the rumen of adult animals appeared
Letters 130 II 99.5) 7-12
11
close to our lower detection limit. This may explain the impossibility of enumerating the microorganisms involved by plating onto solid media [lo]. This also agrees with an accepted minor contribution of acetogenesis to H, recycling in the rumen where methanogenesis is consistently present in adults [6]. On the contrary, in the hindgut of pigs, rats and man for which acetogenesis has been evidenced by labelled HCO, /CO, incorporations [5,6,17], populations were as high as 105-lOa/g. Levels of acetogenic bacteria were apparently negatively correlated to the methanogenic activity of the sample. This was true in the lumen of lambs, where they were detected prior to the natural establishment of methanogenic archaea [18], as well as in the rat and human colon where variable levels of methanogenic archaea are encountered in different individuals [6,16]. The highest populations were observed in the feces of non-methane-excreting individuals. This supports the potential contribution of acetogens to overall H, recycling in the human colon of non-methane excreting subjects, first observed by Lajoie et al. [5,7]. One of the weaknesses of MPN enumerations performed with low numbers of repeats, imposed by anaerobic methodologies, is the lack of precision in the evaluation of population levels. In spite of this, and considering the wide range of populations encountered between species as well as within a given species (e.g. age-related changes as in lambs or associated with methane excretion in humans), the ability to measure such populations appears of utmost interest. It therefore seems worthwhile to further investigate the colon of monogastrics as well as apply this method to other ecosystems such as the termite gut where acetogenesis may play an important role [6]. The diversity of the dominant H,-utilizing acetogenic populations from the human colon and the rumen or cecum of herbivores is currently under investigation (Bernalier, personal communication; [19]). Despite a large phylogenetic diversity, Hz-utilizing acetogenic bacteria all derive energy from the acetyl-CoA pathway and thereby share a set of specific enzymatic activities. In that respect, it is likely that non-culture based technologies relying on probes targeting genomic sequences specific of the acetogens will, in the near future, extend current means of investigation and overcome limitations of
.I. Do& et al. / FEMS Microbiology
12
classical
microbiological
enumeration
techniques
ml.
References [l] Wieringa, K.T. (1940) The formation of acetic acid from carbon dioxide and hydrogen by anaerobic spore-forming bacteria. Antonie van Leeuwenhoek J. Microbial. Serol. 6. 251-262. [2] Dolfing, J. (1988) Acetogencsis. In: Biology of Anaerobic Microorganisms (Zehnder, A.J.B., Ed.), pp. 417-468. John Wiley and Sons, New York, NY. [3] Wood, H.G. and Ljungdahl. L.G. (1991) Autotrophic character of the acetogenic bacteria. In: Variations in Autotrophic Life (Shively, J.M. and Barton, L.L., Eds.), pp. 201-250. Academic Press, New York, NY. [4] Lovley, D.R. and Klug, M.J. (1983) Methanogenesis from methanol and methylamines and acetogenesis from hydrogen and carbon dioxide in the sediments of a eutrophic lake. Appl. Environ. Microbial. 45, 1310-1315. [5] Lajoie, SF., Shelton, B., Miller, T.L. and Wolin, M.J. (1988) Acetate production from hydrogen and [“C]carbon dioxide by the microflora of human feces. Appl. Environ. Microbial. 54, 2723-2727. [6] Breznak, J.A. and Kane, M.D. (1990) Microbial Hz /CO, acetogenesis in animal guts: nature and nutritional significance. FEMS Microbial. Rev. 87, 309-314. [7] Durand, M. and Bernalier, A. (1993) Reductive acetogenesis in animal and human gut. In: Physiological and Clinical Aspects of Short Chain Fatty Acid Metabolism (Cummings, J.H., Rombeau, J.L. and Sakata, T., Eds.), pp. 107-117. Cambridge University Press, Cambridge. [8] Kane, M.D. and Breznak, J.A. (1991) Acetonema longum gen. nov. sp. nov., an H, /CO, acetogenic bacterium from the termite, Pferolermes occidentis. Arch. Microbial. 1.56, 91-98. [9] Wolin, M.J. and Miller, T.L. (1993) Bacterial strains from human feces that reduce CO> to acetic acid. Appl. Environ. Microbial. 59, 3551-3556.
Lefters 130 (1995) 7-12 [lo] Braun, M., Schoberth, S. and Gottschalk, G. (1979) Enumeration of bacteria forming acetate from H, and CO, in anaerobic habitats. Arch. Microbial. 120, 201-204. [ll] Bond, J.H., Engel, R.R. and Levitt, M.D. (1971) Factors influencing pulmonary methane excretion in man. J. Exp. Med. 133, 572-588. [12] Balch, W.E. and Wolfe, R.S. (1976) A new approach to the cultivation of methanogenic bacteria: 2-mercaptoethanesulfonic acid (HS-CoM)-dependent growth of Merhanobacrerium ruminantium in a pressurized atmosphere. Appl. Environ. Microbial. 32, 781-791. [13] Leedle, J.A.Z. and Hespell, R.B. (1980) Differential carbohydrate media and anaerobic replica plating techniques in delineating carbohydrate-utilizing subgroups in rumen bacterial populations. Appl. Environ. Microbial. 39, 709-719. [14] Greening. R.C. and Leedle, J.A.Z. (1989) Enrichment and isolation of Acetitomaculum ruminis, gen. nov., sp. nov.: acetogenic bacteria from the bovine rumen. Arch. Microbial. 151, 399-406. [15] Clarke, K.R. and Owens, N.J.P. (1983) A simple and versatile micro-computer program for the determination of ‘Most Probable Number’. J. Microbial. Methods 1, 133-137. [16] Pochart, P., Dore, J., Lemann, F., Goderel, I. and Rambaud, J.-C. (1992) Interrelations between populations of methanogenic archaea and sulfate-reducing bacteria in the human colon. FEMS Microbial. Lett. 98, 225-228. [17] Bernalier, A., Doisneau, E., Cordelet, C., Beaumatin, P., Durand, M. and Grivet, J.P. (1993) Competition for hydrogen between methanogenesis and hydrogenotrophic acetogenesis in human colonic flora studied by 13C-NMR. Proc. Nutr. Sot. 52, 118A. [18] Morvan, B., Dore, J., Rieu-Lesme, F., Foucat, L., Fonty, G. and Gouet, P. (1994) Establishment of hydrogen-utilizing bacteria in the rumen of the newborn lamb. FEMS Microbial. Lett. 117, 249-256. [19] Rieu-Lesme, F., Fonty, G. and Dore, J. (1994) Isolation and characterization of a new H,-utilizing bacterium from the rumen. FEMS Microbial. L.&t. 125, 77-82. [20] Lovell, CR. and Hui, Y. (1991) Design and testing of a functional group-specific DNA probe for the study of natural populations of acetogenic bacteria. Appl. Environ. Microbial. 57, 2602-2609.