Competition among three predominant ruminal cellulolytic bacteria in the absence or presence of non-cellulolytic bacteria

Junqin Chen1 and Paul J. Weimer1,2

Department of Bacteriology, University of Wisconsin-Madison, Madison, WI 53706, USA1
United States Department of Agriculture, Agricultural Research Service, US Dairy Forage Research Center, USDA-ARS, 1925 Linden Drive West, Madison, WI 53706, USA2

Author for correspondence: Paul J. Weimer. Tel: +1 608 264 5408. Fax: +1 608 264 5147. e-mail: pjweimer{at}facstaff.wisc.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Competition among three species of ruminal cellulolytic bacteria Fibrobacter succinogenes S85, Ruminococcus flavefaciens FD-1 and Ruminococcus albus 7 – was studied in the presence or absence of the non-cellulolytic ruminal bacteria Selenomonas ruminantium or Streptococcus bovis. Co-cultures were grown under either batch or continuous conditions and populations were estimated using species-specific oligonucleotide probes to 16S rRNA. The three cellulolytic species co-existed in cellobiose batch co-culture, but inclusion of either Sel. ruminantium or Str. bovis yielded nearly a monoculture of the non-cellulolytic competitor. In cellobiose chemostats, R. albus completely dominated the triculture, but R. flavefaciens became predominant over F. succinogenes and R. albus when Sel. ruminantium was co-inoculated into the chemostats. Similar effects on competition were observed in the presence of Str. bovis at a lower (0·021 h-1), but not at a higher (0·045 h-1) dilution rate. In cellulose batch co-cultures, R. albus was more abundant than both F. succinogenes and R. flavefaciens, regardless of the presence of the non-cellulolytic species. Co-existence among the three cellulolytic species was observed in almost all cellulose chemostats, but Sel. ruminantium altered the relative proportions of the cellulolytic species. R. albus and R. flavefaciens were found to produce inhibitors that suppressed growth of R. flavefaciens and F. succinogenes, respectively. These data indicate that interactions among cellulolytic bacteria, while complex, can be modified further by non-cellulolytic species.

Keywords: cellobiose, cellulose, competition, Fibrobacter succinogenes, Ruminococcus


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Numerous cellulolytic microbial species have been identified in the rumen, but it is generally agreed that ruminal cellulolysis is carried out primarily by three species of bacteria: Fibrobacter succinogenes, Ruminococcus flavefaciens and Ruminococcus albus (Hungate, 1966 ; Dehority, 1993 ). Because these species are nutritional specialists that differ in fermentation end products, and because cellulose is a major component of the diets of forage-fed ruminants, the relative populations of these three species can potentially impact on the ratios of volatile fatty acids available to the animal, an important determinant of animal performance. Several studies have examined competition among these cellulolytic species. Odenyo et al. (1994a , b ) used oligonucleotide probes to species-specific segments of 16S rRNA to quantify specific populations in binary (two-membered) and ternary (three-membered) batch cultures grown on cellobiose, cellulose or alkaline hydrogen peroxide-treated wheat straw. Their data indicated that R. albus 8 generally out-competed R. flavefaciens FD-1, due to production by the former of a bacteriocin-like substance. Fondevila & Dehority (1996) presented evidence that cellulose digestion was reduced when strains of F. succinogenes A3c and R. flavefaciens B34b were grown together in batch culture and suggested that an inhibitor was produced by one of the two species. Mosoni et al. (1997) reported that R. flavefaciens FD-1 became detached from cellulose in the presence of R. albus 20.

Within a permissive range of ruminal pH, digestion of cellulose is considered to be a first-order process that is limited by the available surface area of cellulose rather than by the hydrolytic capabilities of the cellulolytic species (Waldo et al., 1972 ; Weimer et al., 1990 ). Studies with paired combinations of these species in chemostats revealed that substrate limitation intensifies competition for cellobiose, resulting in monocultures dominated by strains having a higher affinity for substrate (Shi & Weimer, 1997 ). By contrast, cellulose-limited chemostats inoculated with paired species in different combinations yielded stable co-cultures in which the two species displayed niche specialization (Shi et al., 1997 ).

In the rumen, competition among cellulolytic bacteria is complicated by potential interactions with non-cellulolytic species. While specific interactions among cellulolytic and non-cellulolytic species have been characterized in binary culture (Scheifinger & Wolin, 1973 ; Stanton & Canale-Parola, 1980 ; Pavlostathis et al., 1990 ) the effects of non-cellulolytic species on the outcome of competition among the cellulolytic species have not been characterized. The purpose of this study was to determine the effects of two important ruminal non-cellulolytic species, Selenomonas ruminantium and Streptococcus bovis, on the outcome of competition among the predominant cellulolytic species in defined co-culture under conditions of substrate excess and substrate limitation.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cultures and growth conditions.
Bacteria used in this study were Ruminococcus albus 7, Ruminococcus flavefaciens FD-1, Fibrobacter succinogenes S85, Selenomonas ruminantium D, GA192, HD4 and H18, and Streptococcus bovis JB-1. All cultures were maintained at -80 °C in 50% (v/v) glycerol until required. For competition studies, pure cultures of R. albus, R. flavefaciens and F. succinogenes were combined in a sterile serum vial and then inoculated simultaneously into the culture vessels; where indicated, non-cellulolytic strains were also co-inoculated with the cellulolytic triculture. Individual pure cultures were combined at equal volumes (0·5 ml per strain) from cultures grown to mid-exponential phase on the substrate of interest (cellobiose or cellulose). No attempts were made to provide identical cell numbers in the inocula, because this could not be immediately assessed in cellulose-grown cultures due to the abundance of adherent cells. However, previous experiments (Shi et al., 1997 ; Shi & Weimer, 1997 ) revealed that the outcome of competition was independent of inoculum size.

Competition experiments with Str. bovis used strain JB1 (pregrown on cellobiose) in both batch and continuous culture. Batch culture competition experiments with Sel. ruminantium (also pregrown on cellobiose) used strain D, while in continuous culture all four strains of Sel. ruminantium were used to minimize the chance that a single strain may give an anomalous result. Thus, direct comparisons of batch versus continuous culture could not be made in cultures inoculated with Sel. ruminantium, although comparisons across substrate (cellobiose versus cellulose) within a culture mode (batch or continuous) could be made, as could comparisons of tricultures in the presence or absence of Sel. ruminantium within a culture mode.

Modified Dehority Medium (MDM), described previously (Weimer et al., 1991 ), included either Sigmacell 20 microcrystalline cellulose (4 g l-1; Sigma) or cellobiose (4 g l-1) as carbon and energy source. All incubations were conducted anaerobically under CO2 at 39 °C. Batch cultures were conducted in 158 ml serum bottles (containing 100 ml medium), each fitted with a butyl stopper and an aluminium crimp seal. Continuous cultures were performed in a system described previously (Weimer et al., 1991 ). The cellulose-containing medium in the reservoir was homogenized by stirring and diffusive gas sparging with CO2, and was delivered as a CO2-segmented slurry to a stirring fermenter (875 ml working volume) by a peristaltic pump. Cellobiose continuous cultures were conducted in a stirring reactor (139 ml working volume), continuously fed cellobiose-containing MDM. Both cellulose and cellobiose reactors were continuously sparged with filter-sterilized, humidified CO2.

Substrate and fermentation product assays.
For batch cultures, the entire culture was collected after incubation for 24 h (cellobiose) or 48 h (cellulose). Chemostat samples (28 ml for cellulose, 6 ml for cellobiose) were first collected after the reactor had received three dilutions of feed medium. Subsequent samples were removed from reactors at 6–20 h intervals over a 3 d period. These samples were then analysed and those showing constant fermentation product concentrations were used to define the steady state, under the assumption that this stability in product ratios results from a stabilization of the microbial population (whose individual species vary markedly with respect to product ratios).

Samples were analysed for pH using a Corning model 320 pH meter, residual soluble sugars by a phenol/sulfuric acid method (Dubois et al., 1956 ) and fermentation acids and ethanol by HPLC (Weimer et al., 1991 ). Concentrations of cellulose in the reactor or reservoir were measured by a modified neutral detergent fibre method (Weimer et al., 1990 ).

Quantification of relative populations of three cellulolytic species using oligonucleotide probes.
Three samples were taken from triplicate batch cultures and from each steady-state continuous culture. The procedures used for RNA isolation, hybridization and detection were similar to those of Shi et al. (1997) . Samples from cellulose cultures were separated into adherent and non-adherent bacterial populations by filtering through a 47 mm diameter polycarbonate membrane (3 µm pore size; Poretics). This method is based on microscopic observation that none of the three cellulolytic strains forms chains or clumps when growing in the planktonic mode with cellulose as energy source, although we observed that R. flavefaciens and R. albus form chains and clumps, respectively, when grown in batch culture on cellobiose. The filtrate (containing non-adherent cells) and filter cake (containing adherent cells and rinsed with deionized water into a 50 ml centrifuge tube) were recovered separately by centrifugation at 4000 g for 30 min.

RNA isolation.
RNAs were separated from other cellular components using a low-pH phenol extraction procedure modified from Odenyo et al. (1994a) that minimized extraction of DNA (Wallace, 1987 ). All reagents were prepared in diethylpyrocarbonate (DEPC)-treated water. Cell pellets (4000 g, 30 min) were transferred to 2 ml screw-cap conical tubes (Sarstedt) along with 0·5 g zirconium beads (~0·1 mm diam., heated overnight at 160 °C), 700 µl 50 mM sodium acetate/10 mM EDTA (pH 5·1), 50 µl 20% (w/v) SDS and 700 µl saturated phenol (pH 4·3; Amresco). The tubes were shaken twice for 2 min in a bead beater at 4 °C. These tubes were then heat-shocked in a 65 °C water bath for 10 min, followed by another two cycles (2 min each) in the bead beater. The suspensions were centrifuged at 12000 g at 4 °C for 5 min. The aqueous phases were transferred to new 1·7 ml microcentrifuge tubes and extracted twice with each of the following reagents in order: saturated phenol; phenol (pH 4·3)/chloroform (1:1, v/v) and chloroform. RNAs were precipitated by adding 2 vols absolute ethanol and 0·1 vols 3 M sodium acetate, and frozen at -80 °C overnight. RNA pellets were collected by centrifugation (12000 g, 4 °C) for 20 min, washed with 70% ethanol and centrifuged. Pellets of purified RNA were dissolved in 50–100 µl water.

The purities and concentrations of the RNAs were determined by reading absorbance at 260 and 280 nm in a Beckman DU series 600 spectrophotometer. RNA solutions having A260/A280 ratios >=1·7 were used for slot-blots and hybridizations; otherwise extractions with phenol/chloroform (1:1, v/v) and chloroform were repeated to obtain purer preparations. Purified RNA was diluted to ~5 ng µl-1 and stored in microcentrifuge tubes (50 µl per tube) at -80 °C.

Hybridizations.
RNA samples (4–40 µl, containing 20–200 ng RNA) were diluted to 100 µl with denaturation solution (DEPC-treated water/10x SSC/formaldehyde, 5:3:2, by vol.), then incubated at 65 °C for 15 min. RNA loading buffer (2 µl of 1 mM EDTA, 0·4% bromophenol blue, 50% glycerol) was then added and the entire 102 µl solution was slot-blotted onto a Nytran membrane (Schleicher & Schuell) prewetted with 10x SSC. RNAs were cross-linked onto the membrane by UV radiation (Stratalinker 1800; Stratagene) at 0·12 J cm-2 for both sides of the membrane. Membranes were then air-dried and prehybridized in Rapid-Hyb buffer (Amersham) for 1–2 h at dissociation temperature (48 °C for R. albus probe S-S-R.alb-0196-a-A-18 and F. succinogenes probe S-Ss-F.s.suc-0207-a-A-21, and 41 °C for R. flavefaciens probe S-S-R.fla-0196-a-A-17). These temperatures were also used for hybridization. At the end of prehybridization, the Rapid-Hyb buffer was poured out and 14 ml Rapid-Hyb buffer containing 10 pM 5'-digoxigenin-labelled oligo-DNA probes was poured into the tube and the hybridization reaction continued overnight (usually 16 h).

Detection and quantification.
After hybridization, membranes were washed twice with 2x SSC and 0·1% SDS at room temperature (15 min), and twice with 0·5x SSC and 0·1% SDS at the specific hybridization temperature (15 min). Membranes were then immersed in blocking solution (10-fold dilution of 10%, v/v, blocking stock solution in maleate buffer (150 mM NaCl, 100 mM maleic acid, pH adjusted to 7·5 with NaOH) and shaken for 30 min. Alkaline-phosphatase-conjugated anti-digoxigenin (750 U ml-1; Roche Molecular Biochemicals) was added at 1 µl blocking solution ml-1 and the solution was shaken for 1 h. Membranes were washed twice, 15 min per wash with maleate buffer, followed by a 2 min wash in detection buffer (0·1M Tris/HCl, 0·1M NaCl, pH 9·5). Lumiphos 530 (Roche) was used as chemiluminescent substrate for the hybridized RNA and exposure to Lumi-film (Roche) for 3 h was used to record light emission. After developing, the films were scanned on a Molecular Dynamics laser densitometer and the bands were quantified with ImageQuant software (Molecular Dynamics).

The amounts of RNA in each band from the densitometer image were determined from standard curves prepared with purified RNA from pure cultures and slot-blotted on the same membranes as the RNAs from samples. The relative population size of each cellulolytic species among the total cellulolytic population was expressed as the ratio of the amount of each bacterial RNA detected to the sum of the amounts of RNAs detected for the three cellulolytic species. Across all experiments with tricultures of the three cellulolytic strains (in the absence of added non-cellulolytic strains), the sums of the RNA detected using the three species-specific probes averaged 91·3% of the RNA detected using the bacterial domain probe S-D-Bact-0338-a-A-18 of Amann et al. (1990) . Differences among the relative population sizes for the cellulolytic strains were contrasted by a two-tailed t-test within the general linear model procedure of the software application (SAS Institute, 1985 ).

Although the RNA isolation procedure was conducted under conditions that minimize DNA extraction, the potential cross-reaction of each of the probes with DNA was tested. For these analyses, DNA was isolated from pure cultures using a Wizard Genomic DNA Purification Kit (Promega) according to the manufacturer’s instructions for Gram-positive and Gram-negative bacteria, except that each centrifugation step was extended for an additional 1 min. Although hybridization of the probes to 100 ng purified DNA could not be detected, weak hybridization was sometimes detected when the amount of DNA was increased to 1000 ng. By normalizing the signal to that obtained for the hybridization of 50 ng RNA, the extent of cross-reaction with a probe to the DNA of that species was estimated to range from 0 to 0·59% and the extent of cross-reaction to the DNA of the non-target species was estimated to range from 0 to 0·26%, relative to the extent of hybridization detected using purified RNA and the species-specific, RNA-targeted probe.

Inhibitor screening.
Two methods were used to test the antagonistic activities between the bacterial strains. In the liquid culture assay, 2 ml supernatant of spent culture of the test strain was added to culture tubes containing 10 ml MDM/cellobiose. The indicator strain was then inoculated into tubes at room temperature, the tubes placed in a 39 °C incubator and OD600 was recorded at intervals to allow determination of maximum growth rate. Sterilized water (2 ml) was supplemented to control tubes when determining growth in the absence of inhibitor.

Inhibitory activity was also measured using a plate assay, modified from that of Tagg et al. (1973) . A lawn of indicator strain was first seeded by mixing 0·2 ml of an overnight culture with 15·5 ml melted (~45 °C) solid medium [MDM supplemented with (l-1): 10 g agar, 2 g Trypticase, 1 g yeast extract, 0·25 g each of cysteine/HCl and Na2S.9H2O]. This suspension was plated on sterile 100x15 mm Petri dishes in an anaerobic chamber (5% H2/95% CO2, v/v). A drop (~50 µl) of test strain from either exponential-phase or stationary-phase cultures was spotted onto the lawn as it solidified. The plate was examined for zones of inhibition after overnight incubation in the chamber at 39 °C.

Protease sensitivity.
The sensitivities of the inhibitors were tested with two broad-spectrum proteases, Streptomyces griseus protease (EC 3 . 4 . 24 . 31; Sigma cat. no. P5147) and porcine pancreatin (Sigma cat. no. P7545). A filter-sterilized stock solution (10 mg protease ml-1 in 0·01 M phosphate buffer, pH 7·0) was mixed with the overnight culture or culture supernatant of the test strain to give a final concentration of 100 µg enzyme ml-1 and incubated for 4 h at 40 °C (for pancreatin) or 37 °C (for Streptomyces griseus protease). The mixtures were taken into the anaerobic chamber and allowed to reduce for 1 h before drop testing. Elimination or reduction in diameter of the zones of inhibition was examined by comparison to controls that contained no added enzyme or that contained enzyme but no added test culture or culture supernatant.

Determination of the molecular mass range of the inhibitor.
Supernatants (3·5 ml) of test strain were loaded onto MICROSEP microconcentrators (Pall-Filtron) of 1, 3, 10 or 30 kDa molecular mass cut-off and centrifuged at 7000 g for 3, 2, 1·5 or 1 h, respectively. The inhibitory activity of both the concentrated sample and the filtrate were checked by the plate assay described above.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Relative population sizes of ruminal cellulolytic species in cellobiose co-cultures
Batch cultures. Batch-mode incubation of cultures co-inoculated with the three cellulolytic bacterial species resulted in a triculture that consisted of a similar population size of F. succinogenes and R. albus, and a smaller population of R. flavefaciens (Table 1). Fermentation end products were mainly acetate, formate, succinate (produced primarily by F. succinogenes) and ethanol (produced only by R. albus) (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 1. Population distributions in cellobiose co-cultures of the three cellulolytic species [F. succinogenes S85 (F.s.), R. albus 7 (R.a.) and R. flavefaciens FD-1 (R.f.)] in the presence or absence of the non-cellulolytic bacteria Sel. ruminantium (S.r.) or Str. bovis (S.b.)

 

View this table:
[in this window]
[in a new window]
 
Table 2. Fermentation data for cellobiose co-cultures of the three cellulolytic species [F. succinogenes S85 (F.s.), R. albus 7 (R.a.) and R. flavefaciens FD-1 (R.f.)] in the presence or absence of the non-cellulolytic bacteria Sel. ruminantium (S.r.) or Str. bovis (S.b.)

 
Batch-mode incubation of the three cellulolytic species co-inoculated with Sel. ruminantium D contained Sel. ruminantium as the dominant species. Microscopy revealed a predominance of curved rod-shaped cells and large amounts of lactate were detected in fermentation end products. Sel. ruminantium is nearly a homolactic fermenter at rapid growth rates (Wallace, 1978 ) and lactate is not produced in significant amounts by any of the three cellulolytic species (Hungate, 1966 ). Moreover, succinate (from F. succinogenes and R. flavefaciens) and ethanol (from R. albus) were not detectable. Although the population of Sel. ruminantium was not quantified, its dominance was confirmed by the observation that the proportion of RNA attributable to the three cellulolytic species (determined using species-specific probes) represented only 10·1% of the RNA detected using the bacterial domain probe (data not shown). Among the cellulolytic species, R. albus was more abundant than F. succinogenes and R. flavefaciens, which each accounted for only half the population size of R. albus.

Co-cultures of Str. bovis JB1 with the cellulolytic triculture were dominated by Str. bovis. Lactate was almost the sole fermentation end product (Table 2) and only 11·2% of the RNA detected with the bacterial domain probe was attributable to the three cellulolytic species (data not shown). R. albus again was the most abundant of the cellulolytic species, accounting for two-thirds of the cellulolytic population (Table 1).

Continuous cultures. Table 1 shows that cellobiose-limited chemostats co-inoculated with the three cellulolytic species produced monocultures of R. albus at all dilution rates tested. The population pattern was consistent with the detection of acetate and ethanol as sole non-gaseous fermentation end products (Table 2). Succinate, a major fermentation product of both F. succinogenes and R. flavefaciens (Hungate, 1966 ), was not observed. These results are in agreement with data from two-membered continuous cultures in which R. albus displaced either F. succinogenes or R. flavefaciens (Shi & Weimer, 1997 ).

By comparison, cellobiose-limited continuous cultures initially inoculated with a mixture of four strains of Sel. ruminantium along with the three cellulolytic species displayed a dramatic change in the population distribution pattern in which R. flavefaciens became dominant over the other two cellulolytic species. A large amount of propionate, but no succinate, was observed, suggesting a complete conversion of succinate (produced by R. flavefaciens) to propionate by a substantial population of Sel. ruminantium.

In the presence of Str. bovis, the population distribution of cellulolytic species was affected by dilution rate. At higher D values (0·045 h-1), the pattern was similar to those of the triculture chemostats, with R. albus dominating the competition among the cellulolytic species. At lower D values (0·021 h-1), results were similar to those of chemostats with Sel. ruminantium. These population data were in agreement with the fermentation end product data, which showed higher amounts of succinate but lower amounts of ethanol at D=0·021 h-1. By contrast, more ethanol (but no succinate) was observed at D=0·045 h-1.

Relative population sizes of ruminal cellulolytic species in cellulose co-cultures
Batch cultures. R. albus was much more abundant than F. succinogenes and R. flavefaciens in both cellulose-adherent and planktonic phases of cellulose batch cultures and this outcome was not affected by the inclusion of Sel. ruminantium D or Str. bovis JB1 (Table 3). High amounts of ethanol and very low amounts of succinate in the culture supernatants (Table 4) were in accord with the population data. Clearly, there were no effects on the competition among the three cellulolytic species in co-culture with Str. bovis, although there was a slight enhancement (P<0·05) of R. albus or inhibition of F. succinogenes in the presence of Sel. ruminantium.


View this table:
[in this window]
[in a new window]
 
Table 3. Population distributions of the three cellulolytic species [F. succinogenes S85 (F.s.), R. albus 7 (R.a.) and R. flavefaciens FD-1 (R.f.)] in cellulose batch and in cellulose-limited chemostats under steady-state conditions in the presence or absence of the non-cellulolytic bacteria Sel. ruminantium (S.r.) or Str. bovis (S.b.)

 

View this table:
[in this window]
[in a new window]
 
Table 4. Fermentation data for cellulose co-cultures of the three cellulolytic species [F. succinogenes S85 (F.s.), R. albus 7 (R.a.) and R. flavefaciens FD-1 (R.f.)] in the presence or absence of the non-cellulolytic bacteria Sel. ruminantium (S.r.) or Str. bovis (S.b.)

 
Continuous cultures. Co-existence among the three cellulolytic species was observed in all cellulose-limited continuous co-cultures (Table 3). In tricultures of the three cellulolytic species, a slight dominance of F. succinogenes over R. albus was observed, along with a small population of R. flavefaciens. This was in sharp contrast to batch cultures, which were clearly dominated by R. albus. As in batch culture, the presence of Str. bovis did not alter the competition among the cellulolytic species. There was, however, a shift of the population pattern between R. albus and F. succinogenes in tricultures co-inoculated with a mixture of Sel. ruminantium strains. Appearance of propionate in fermentation products also indicated the presence of Sel. ruminantium, but substantial residual concentrations of succinate suggested that Sel. ruminantium populations were low in the co-cultures.

Identification of inhibitors involved in competition among the cellulolytic species
Table 5 shows the effect of supernatants from pure cultures on the growth of the three cellulolytic species on cellobiose in batch mode. The growth of R. flavefaciens FD-1 was suppressed and its lag time was prolonged by supernatant from the R. albus 7 culture. Although occasional tubes within the replicated set showed growth, the maximum OD600 (0·14) was lower than that of controls not containing this supernatant (0·67). Inhibition was also shown in that R. flavefaciens FD-1 did not grow in the medium that included half of the volume from supernatant of R. albus 7, while the growth of R. albus 7 and F. succinogenes S85 was not affected under the same conditions (data not shown). Inhibition of F. succinogenes S85 by a supernatant of R. flavefaciens FD-1 was also observed in terms of both decreased maximum growth rate and increased lag time.


View this table:
[in this window]
[in a new window]
 
Table 5. Effect of supernatants from pure cultures on growth parameters of pure cultures of R. flavefaciens FD-1 (R.f.), R. albus 7 (R.a.) and F. succinogenes S85 (F.s.) in cellobiose batch culture

 
Additional experiments were performed to further demonstrate the inhibition between R. albus 7 and R. flavefaciens FD-1, and between R. flavefaciens FD-1 and F. succinogenes S85. Overnight cultures of each test strain were spotted onto agar lawn seeded with an indicator strain. After 16 h incubation, zones of inhibition were observed around the colonies of R. albus 7 on R. flavefaciens FD-1 lawns and around the colonies of R. flavefaciens FD-1 on F. succinogenes S85 lawns. Using this method, no other inhibitory activities were observed for any other combination of test culture and culture supernatant from the five strains (three cellulolytic strains, Sel. ruminantium D and Str. bovis JB1).

The inhibitor from R. albus 7 was protease-sensitive. Streptomyces griseus protease completely destroyed the inhibitory activity of a supernatant from an overnight culture of R. albus 7 and pancreatin reduced the size of the inhibition zone relative to culture supernatants not treated with pancreatin. Thus the inhibitor from R. albus 7 is protein in nature, as reported by Odenyo et al. (1994a ) for a bacteriocin-like substance produced by R. albus 8. The molecular mass range of the R. albus 7 agent, estimated by ultrafiltration, fell between 10 and 30 kDa.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A number of studies have described interactions between cellulolytic species and non-cellulolytic species of ruminal bacteria (Scheifinger & Wolin, 1973 ; Stanton & Canale-Parola, 1980 ; Kudo et al., 1987 ; Pavlostathis et al., 1990 ; Debroas & Blanchart, 1993 ; Williams et al., 1994 ; Horvan et al., 1996 ). Several reports have also described interactions between two or among three cellulolytic species of ruminal bacteria (Odenyo et al., 1994a , b ; Fondevila & Dehority, 1996 ; Mosoni et al., 1997 ; Shi et al., 1997 ; Shi & Weimer, 1997 ). Most of these studies have been performed under batch culture conditions. This report is the first to examine interactions of tricultures of ruminal cellulolytic species and the impacts of non-cellulolytic bacteria on these interactions under substrate-limited, continuous culture conditions. Although the experiments involved only a single strain of each cellulolytic species and should be interpreted cautiously, the observed co-existence of several strains at one or more trophic levels suggests a complex pattern of interactions among these strains.

The observed suppression of R. flavefaciens FD-1 in both cellobiose and cellulose batch co-cultures (Tables 1 and 3) is consistent with the data of Odenyo et al. (1994a , b ). The inhibition of R. flavefaciens FD-1 by R. albus 7 more than compensated for the fact that R. flavefaciens FD-1 has a greater µmax than do both R. albus 7 and F. succinogenes S85 and shorter lag time than does F. succinogenes S85 in cellobiose medium (Table 5). Both growth suppression (Table 5) and possible inhibition of adherence to cellulose particles (Mosoni et al., 1997 ) by R. albus 7 could have led to undetectable levels of R. flavefaciens FD-1 in cellulose batch co-cultures. By contrast, the ability of R. albus 7 to outcompete F. succinogenes S85 in cellulose batch co-cultures (Table 3) may have been due to the former’s shorter lag time in cellulose medium (as suggested by visual inspection of pure cultures). During growth on cellulose, the distribution patterns of the adherent and planktonic populations of all three cellulolytic species were similar (P<0·05). This result suggests that many of the factors governing the interactions among the cellulolytic bacteria during growth on cellulose particles may be similar to those governing interactions during growth on soluble substrates.

In continuous culture with a single limiting nutrient, pure and simple competition is expected to yield a monoculture of a single strain based on a combination of maximum growth rate and affinity for substrate (Hansen & Hubbell, 1980 ). Thus, the dominance by R. albus in our cellobiose triculture chemostats (Table 1) is consistent with a previous report that this strain adapted to cellobiose limitation more readily than did the other two species (Shi & Weimer, 1997 ). However, the suppression of R. flavefaciens may also have been due to the production of an inhibitory agent by R. albus. According to the definition of pure and simple competition by Fredrickson & Stephanopoulos (1981) , competition in our chemostat cultures was not pure because competition for nutrients was not the sole interaction. Furthermore, competition for cellulose was not simple due to the complex profile of substrates (soluble cellodextrins of different chain lengths) available as cellulose degradation proceeded. In theory, populations of up to n strains can co-exist in a spatially homogeneous system with constant inputs if n nutrients exert dynamic effects on the system (Fredrickson & Stephanopoulos, 1981 ). As a result, our cellulose-limited chemostats yielded a triculture of cellulolytic species rather than a monoculture.

In continuous culture on cellobiose (Table 1), but not on cellulose (Table 3), addition of a mixture of Sel. ruminantium strains allowed R. flavefaciens FD-1 to outcompete the other two cellulolytic species, although Sel. ruminantium itself appeared to be the dominant species in the co-culture. The prevalence of R. flavefaciens appears to be due to a combination of several factors: its lower Ks for cellobiose (Shi & Weimer, 1996 ), its inhibition of F. succinogenes and reduced inhibition, resulting from a presumably lower concentration of inhibitor produced by a smaller population of R. albus. The predominance of R. albus at the higher dilution rate (0·045 h-1) in the presence of Str. bovis (Table 1) is probably due to a higher concentration of inhibitor produced by a greater population size of R. albus at the higher growth rate. R. flavefaciens outgrew the other two cellulolytic species at D=0·021 h-1, but R. albus outgrew the other two at D=0·045 h-1, suggesting that there may be a critical growth rate between these values that determines the outcome of the competition. Although the rumen is not a truly continuous culture habitat, the dilution at which this putative switch might occur is within the range of mean dilution rates reported for rumen contents (Hungate, 1966 ). In our experiments, the dilution rates on cellobiose were kept low because this permitted direct comparison with the cellulose-grown cultures that were grown at these same low dilution rates (i.e. they permitted comparison of interactions for two different substrates, without the confounding effect of growth rate). All of these organisms certainly are capable of much higher growth rates on high concentrations of cellobiose. However, it is unlikely that they encounter such high concentrations in the rumen, where cellobiose is produced from the slow hydrolysis of cellulose, and its concentration is kept low by intense competition among the many species (cellulolytic and non-cellulolytic) that can utilize this substrate (Russell, 1985 ).

F. succinogenes competed much more effectively against R. albus under cellulose limitation than in cellulose batch culture (Table 3), perhaps because of its lower Ks for cellodextrins (Shi & Weimer, 1996 ). Moreover, under cellulose limitation a small population of R. flavefaciens was detected, presumably due to a reduced concentration of inhibitor produced by the smaller population of R. albus. The addition of several strains of Sel. ruminantium selectively enhanced the population of R. albus. Because F. succinogenes (succinate producer) and Sel. ruminantium (succinate utilizer) have been shown to grow well in binary culture (Scheifinger & Wolin, 1973 ), the selective stimulation of R. albus suggests that Sel. ruminantium may enhance some aspect of cellulose digestion or metabolism by R. albus, rather than by suppressing F. succinogenes. This stimulation of R. albus may be due to enhanced adherence to cellulose, as the enhancement in cellulose chemostats was greater for the adherent population (Table 3) and the effect was not noted in cellobiose chemostats (Table 1).

Sel. ruminantium and Str. bovis display some similarities in physiological characteristics, including the use of soluble sugars and cellodextrins (but not cellulose) as energy sources, a homolactic fermentation at high growth rates and a heterolactic fermentation at low growth rates (Wallace, 1978 ; Russell & Hino, 1985 ). In our study, both species also dominated the cellulolytic species in cellobiose-limited chemostats (Table 1), indicating a superior ability of these non-cellulolytic species to compete for this soluble sugar when it was provided exogenously (rather than it being released during attachment to, and hydrolysis of, cellulose by the cellulolytic species). However, some differences in these two non-cellulolytic species were observed with respect to their effects on competition among the cellulolytic species. The fact that Sel. ruminantium altered this competition over a wide range of dilution rates, while Str. bovis did not, suggests the existence of specific interactions among these species that are not easily predictable from physiological characteristics alone.

The production of a bacteriocin-like inhibitor by a ruminal cellulolytic bacterium was first reported for R. albus 8 by Odenyo et al. (1994a) . Kalmokoff & Teather (1997) have shown that bacteriocin-like activities are common among isolates of another ruminal species, Butyrivibrio fibrisolvens. In our study, R. albus 7 was shown to produce a protein active against R. flavefaciens FD-1 and a second inhibitory activity of R. flavefaciens FD-1 against F. succinogenes S85 was also observed (Table 5). These data support suggestions that bacteriocins or other allelochemicals produced by ruminal microbes may represent an important survival and interaction strategy that confers competitive fitness in the ruminal environment (Odenyo et al., 1994a ; Kalmokoff & Teather, 1997 ), although in situ data supporting this notion is lacking.

Studies on the interactions among single strains of different ruminal bacterial species must be interpreted cautiously to avoid generalizations that may not hold for other strains of these species. However, the trend of dominance by R. albus suggested in this study is consistent with oligonucleotide probe data that showed R. albus was generally far more abundant than either F. succinogenes and R. flavefaciens in the bovine rumen (Weimer et al., 1999 ). This suggests that factors affecting the interactions of these species in substrate-limited continuous cultures are likely to be similar to those in the rumen. The overall success of R. albus in both in vivo and in vitro studies may due to several factors: (i) production of substances that inhibit the growth of R. flavefaciens (Odenyo et al., 1994a , this study); (ii) successful competition for adherence to cellulose (Mosoni et al., 1997 ); (iii) greater adaptability under selective pressure to more rapid growth at low cellobiose concentrations (Shi & Weimer, 1997 ); (iv) improved adherence through interaction with non-cellulolytic species (this study); and (v) greater ability to degrade hemicellulose and ferment pentoses (Dehority, 1973 ). Purification and characterization of the proteinaceous inhibitor from R. albus 7 is in progress. Further studies on the inhibition of F. succinogenes S85 by R. flavefaciens FD-1 and the stimulation of R. albus by Sel. ruminantium during growth on cellulose will provide more information on the mechanisms underlying the interactions between ruminal bacteria.


   ACKNOWLEDGEMENTS
 
We thank C. L. Odt for technical assistance during a portion of this study. We also thank J. B. Russell for supplying a culture of Str. bovis JB-1, M. A. Cotta for supplying cultures of Sel. ruminantium and D. M. Stevenson for helpful suggestions.

Disclaimer – mention of specific commercial products is for informational purposes only and does not imply an endorsement or warranty of these products over other products not mentioned.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Amann, R. I., Binder, B. J., Olson, R. J., Chisolm, S. W., Deveraux, R. & Stahl, D. A.(1990). Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl Environ Microbiol 56, 1919-1925.[Medline]

Debroas, D. & Blanchart, G.(1993). Interaction between proteolytic and cellulolytic rumen bacteria during hydrolysis of plant cell wall protein. Reprod Nutr Dev 33, 283-288.[Medline]

Dehority, B. A.(1973). Hemicellulose degradation by rumen bacteria. Fed Proc 32, 1819-1825.[Medline]

Dehority, B. A.(1993). Microbial ecology of cell wall fermentation. In Forage Cell Wall Structure and Digestibility , pp. 425-453. Edited by H. G. Jung, D. R. Buxton, R. D. Hatfield & J. Ralph. Madison, WI:American Society of Agronomy/Crop Science Society of America/Soil Science Society of America.

Dubois, M., Giles, K. A., Hamilton, J. K., Rebers, P. A. & Smith, F.(1956). Colorimetric method for determination of sugar and related substances. Anal Chem 28, 350-356.

Fondevila, M. & Dehority, B. A.(1996). Interactions between Fibrobacter succinogenes, Prevotella ruminicola, and Ruminococcus flavefaciens in the digestion of cellulose from forages. J Anim Sci 74, 678-684.[Abstract/Free Full Text]

Fredrickson, A. G. & Stephanopoulos, G.(1981). Microbial competition. Science 213, 972-979.[Medline]

Hansen, S. R. & Hubbell, S. P.(1980). Single-nutrient microbial competition: qualitative agreement between experiment and theoretically forecast outcomes. Science 28, 1491-1493.

Horvan, B., Rieu-Lesme, F., Fonty, G. & Gouet, P.(1996). In vitro interaction between rumen H2-producing cellulolytic microorganisms and H2-utilizing acetogenic and sulfate-reducing bacteria. Anaerobe 2, 175-180.

Hungate, R. E. (1966). The Rumen and its Microbes. New York: Academic Press.

Kalmokoff, M. L. & Teather, R. M.(1997). Isolation and characterization of a bacteriocin (butyrivibriocin AR 10) from the ruminal anaerobe Butyrivibrio fibrisolvens AR10: evidence for the widespread occurrence of bacteriocin-like activity among ruminal isolates of B. fibrisolvens. Appl Environ Microbiol 63, 394-402.[Abstract]

Kudo, H., Cheng, K. J. & Costerton, J. W.(1987). Interactions between Treponema bryantii and cellulolytic bacteria in the in vitro digestion of straw cellulose. Can J Microbiol 33, 241-248.

Mosoni, P., Fonty, G. & Gouet, P.(1997). Competition between ruminal cellulolytic bacteria for adhesion to cellulose. Curr Microbiol 35, 44-47.[Medline]

Odenyo, A. A., Mackie, R. I., Stahl, D. A. & White, B. A.(1994a). The use of 16S rRNA-targeted oligonucleotide probes to study competition between ruminal fibrolytic bacteria: development of oligonucleotide probes for Ruminococcus species and evidence for bacteriocin production. Appl Environ Microbiol 60, 3688-3696.[Abstract]

Odenyo, A. A., Mackie, R. I., Stahl, D. A. & White, B. A.(1994b). The use of 16S rRNA-targeted oligonucleotide probes to study competition between ruminal fibrolytic bacteria: pure culture studies with cellulose and alkaline peroxide-treated wheat straw. Appl Environ Microbiol 60, 3697-3703.[Abstract]

Pavlostathis, S. G., Miller, T. L. & Wolin, M. J.(1990). Cellulose fermentation by continuous cultures of Ruminococcus albus and Methanobrevibacter smithii. Appl Microbiol Biotechnol 33, 109-116.

Russell, J. B.(1985). Fermentation of cellodextrins by cellulolytic and noncellulolytic ruminal bacteria. Appl Environ Microbiol 49, 572-576.[Medline]

Russell, J. B. & Hino, T.(1985). Regulation of lactate production in Streptococcus bovis: a spiraling effect that contributes to rumen acidosis. J Dairy Sci 68, 1712-1721.

SAS Institute (1985). SAS User’s Guide: Statistics, Version 5. Cary, NC: SAS Institute.

Scheifinger, C. C. & Wolin, M. J.(1973). Propionate formation from cellulose and soluble sugars by combined cultures of Bacteroides succinogenes and Selenomonas ruminantium. Appl Microbiol 26, 789-795.[Medline]

Shi, Y. & Weimer, P. J.(1996). Utilization of individual cellodextrins by three predominant ruminal cellulolytic bacteria. Appl Environ Microbiol 62, 1084-1088.[Abstract]

Shi, Y. & Weimer, P. J.(1997). Competition for cellobiose among three predominant ruminal cellulolytic bacteria under substrate-excess and substrate-limited conditions. Appl Environ Microbiol 63, 743-748.[Abstract]

Shi, Y., Odt, C. L. & Weimer, P. J.(1997). Competition for cellulose among three predominant ruminal cellulolytic bacteria under substrate-excess and substrate-limited conditions. Appl Environ Microbiol 63, 734-742.[Abstract]

Stanton, T. B. & Canale-Parola, E.(1980). Treponema bryantii sp. nov., a rumen spirochete that interacts with cellulolytic bacteria. Arch Microbiol 127, 145-156.[Medline]

Tagg, J. R., Read, R. S. D. & McGiven, A. R.(1973). Bacteriocin of a group A Streptococcus: partial purification and properties. Antimicrob Agents Chemother 4, 214-221.[Medline]

Waldo, D. R., Smith, L. W. & Cox, E. L.(1972). Model of cellulose disappearance from the rumen. J Dairy Sci 55, 125-129.[Medline]

Wallace, D. M.(1987). Large- and small-scale phenol extractions. Methods Enzymol 152, 33-41.[Medline]

Wallace, R. J.(1978). Control of lactate production by Selenomonas ruminantium: homotropic activation of lactate dehydrogenase by pyruvate. J Gen Microbiol 107, 45-52.[Medline]

Weimer, P. J., Lopez-Guisa, J. M. & French, A. D.(1990). Effect of cellulose fine structure on kinetics of its digestion by mixed ruminal microorganisms in vitro. Appl Environ Microbiol 54, 2421-2429.

Weimer, P. J., Shi, Y. & Odt, C. L.(1991). A segmented gas/liquid delivery system for continuous culture of microorganisms on solid substrates, and its use for growth of Ruminococcus flavefaciens on cellulose. Appl Microbiol Biotechnol 36, 178-183.

Weimer, P. J., Waghorn, G. C., Odt, C. L. & Mertens, D. R.(1999). Effect of diet on populations of three species of ruminal cellulolytic bacteria in lactating dairy cows. J Dairy Sci 82, 122-134.[Abstract/Free Full Text]

Williams, A. G., Withers, S. E. & Joblin, K. N.(1994). The effect of cocultivation with hydrogen-consuming bacteria on xylanolysis by Ruminococcus flavefaciens. Curr Microbiol 29, 133-138.

Received 14 July 2000; revised 10 September 2000; accepted 9 October 2000.