J. Dairy Sci. 2009. 92:3861-3873. doi:10.3168/jds.2008-1437
© 2009 American Dairy Science Association ®
Investigating unsaturated fat, monensin, or bromoethanesulfonate in continuous cultures retaining ruminal protozoa. II. Interaction of treatment and presence of protozoa on prokaryotic communities1
S. K. R. Karnati*,
,2,
Z. Yu and
J. L. Firkins*,
* Ohio State University Interdisciplinary Nutrition Program (OSUN), and
Department of Animal Sciences, The Ohio State University, Columbus 43210
2 Corresponding author: karnati.1{at}osu.edu
 |
ABSTRACT
|
|---|
Increasing the consistency of responses to reduce emissions of ruminal methane and nitrogenous wastes into the environment using microbial inhibitors requires an accurate assessment of microbial community profiles. In addition to direct inhibition of methanogens by feed additives, protozoa are often targeted for inhibition because their close physical association with endo- and ectosymbionts stimulates methanogenesis in the rumen. In this study, we first modified a continuous culture system to maintain a diverse protozoal population (faunated subperiod) and then selectively effluxed them without using any chemical agents (defaunated subperiod). In both subperiods, unsaturated fat (potentially inhibitory to ciliate protozoa, methanogens, and gram-positive bacteria), monensin (assumed to inhibit gram-positive bacteria), and bromoethanesulfonate (BES; a potent inhibitor of methanogens) were used to suppress the respective functional groups of microorganisms. Changes in microbial populations were determined using denaturing gradient gel electrophoresis, followed by cloning and DNA sequencing of the excised bands . Neither monensin nor unsaturated fat consistently affected methanogen populations under our conditions in either the faunated or defaunated subperiods. When BES was administered, bands presumptively linked to protozoa-associated methanogens in the faunated subperiod disappeared in the defaunated subperiod. However, there was no noticeable adaptation of the sensitive methanogens to BES. The effect of dietary treatments on bacterial populations in the fermenters was harder to ascertain because of the overriding period effect caused by a different inoculum in each period. Defaunation selectively decreased the intensity of bands associated with ruminococci and clostridia but seemed to increase some Butyrivibrio and related populations. Presence of protozoa influenced both bacterial and archaeal populations, probably by selective predation, competition for substrate, or through symbiotic interactions.
Key Words: bacteria • methanogens • monensin • bromoethanesulfonate
|
INTRODUCTION
|
|---|
The ruminal microbial ecosystem is complex and interactive (Dehority, 2003). A universal driving force in fermentation is to maximize ATP yield while disposing of hydrogen equivalents (Wolin et al., 1997). Thus, by maintaining a low partial pressure of H2, the methanogenic archaea indirectly change the flux of fermentation by bacteria, protozoa, and fungi that express hydrogenase to produce H2 while stoichiometrically increasing acetate and butyrate production. If this interspecies H2 transfer increases the ATP yield for some bacteria, we reasoned that, in reverse, inhibition of methanogenic archaea by compounds such as bromoethanesulfonate (BES) should thereby modify bacterial populations. Because of their symbiosis with endo- and ectosymbiotic archaea, protozoa depend on methanogenesis to maximize ATP yield; however, the specific interactions between specific methanogens and protozoa have not been characterized adequately (Janssen and Kirs, 2008).
As explained in the companion manuscript (Karnati et al., 2009), we modified our continuous culture system to retain or efflux protozoa to characterize interactions between protozoa and populations of archaea and bacteria, with respect to changes in methane production, microbial protein synthesis, and biohydrogenation (BH). Our previous studies with continuous culture fermenters (which were defaunated) have identified changes in BH through specific manipulation of fermenter conditions (Qiu et al., 2004) or diets (Ribeiro et al., 2005). Ziemer et al. (2000) noted that a fermenter system similar to ours supported a core prokaryotic functional community structure and had fermentation characteristics and substrate degradation consistent with in vivo conditions. However, a rapid washout of protozoa was associated with a shift in archaeal families, probably resulting from loss of some archaeal symbionts. Consequently, the modified culture system used in the current report offers a unique evaluation of how protozoa influence prokaryotic populations and ecology or vice versa.
To overcome limitations in cultivability of the majority of prokaryotes, most metagenomics analyses have used the small subunit rRNA gene for phylogenetic analyses. All of these procedures have assumptions or limitations that must be considered relative to their advantages (Firkins and Yu, 2006). As a compromise between the amount of information provided and the large number of samples to be analyzed, we chose to use denaturing gradient gel electrophoresis (DGGE). Banding pattern analysis will objectively compare lanes loaded by treatment to overcome the profound period (different inoculation) effects in continuous culture (Noftsger et al., 2003). We combined this approach with sequence analysis of excised bands to profile potential differences in abundant populations of prokaryotes (Mackie et al., 2007). Our previous work has documented the robustness of the V3 hypervariable region of the 16S rDNA for bacteria and archaea for DGGE analyses (Firkins et al., 2008; Yu et al., 2008). However, because of potential co-migration of different sequences recovered from the same band (Firkins et al., 2008), one current aim was to perform a rigorous screening analysis of within- and among-band sequences.
Our hypotheses for this experiment were that a) unsaturated fat should decrease methanogenic populations by inhibiting protozoa and should interact with faunation status on the populations of the lipid-metabolizing butyrivibrios; b) monensin should decrease methane production indirectly by inhibiting gram-positive, H2-producing bacteria; and c) BES should directly inhibit sensitive methanogens and, indirectly, protozoa but with little effect on bacteria. Thus, our objectives were to evaluate shifts in prokaryotic populations, as assessed by DGGE combined with sequence analysis, in the presence or absence of protozoa in continuous culture.
|
MATERIALS AND METHODS
|
|---|
Experimental Design
Specific details of fermenter design and conditions are described in Karnati et al. (2009). Four modified dual-flow continuous culture fermenters were provided a single meal of 40 g/d of a 70:30 forage:concentrate diet (34.5% NDF, 14% CP) containing either no additive, 5% animal-vegetable fat, BES (250 µM final concentration), or monensin (2.5 µM final concentration) in a 4 x 4 Latin square design with 4 periods of 23 d each. At the beginning of each experimental period, special filters (50-µm pore size) were used on the filtrate pumps to retain protozoa in the fermenters. After 7 d of adaptation, the fermenters were sampled for 3 d (subperiod 1). At the start of subperiod 2, conventional filters (300-µm pore size) replaced the smaller filters to allow the protozoa to efflux via filtrate pumps over 3 d; after a further 7 d of adaptation, the fermenters were sampled for 3 d as done in the first subperiod.
Sample Collection
Generic distribution of protozoa was determined in an aliquot of the effluent sample collected as described in Karnati et al. (2009). On the last day of the collection period in each subperiod, the fermenter contents were sampled at 0, 3, 6, 9, and 12 h after feeding. Before sample collection, the fermenter contents were briefly agitated at 150 rpm to ensure uniform mixing of contents. A calibrated rigid plastic tube with a wide diameter was used to aspirate the fermenter contents twice to obtain a 10-mL sample. These samples were fixed in formalin (1 g/100 mL final concentration of HCHO) and stored at –20°C until DNA extraction (Sylvester et al., 2004).
Extraction of DNA and DGGE Analyses
The frozen formalin-fixed samples of fermenter contents were thawed and 2-mL subsamples were taken from each sample and composited by fermenter and subperiod. Total community DNA was extracted from 1-mL aliquots of each composited sample for bacterial and archaeal populations (Yu and Morrison, 2004) and for protozoal populations (Sylvester et al., 2004).
For DGGE analysis of protozoa, the ciliate-specific PCR primer set 316f and 539r-GC (Sylvester et al., 2005) was used with the same running conditions for PCR, but reaction volumes were 50 µL. The PCR products were electrophoresed on an 8% polyacrylamide gel (37.5:1) with a denaturing gradient from 30 to 36% [100% denaturant being 40% (vol/vol) formamide and 7 M urea]. Electrophoresis was performed for 1,350 volt-hours at a constant temperature of 60°C. After the run, gels were stained for 30 min in 0.5 x Tris-borate-EDTA buffer containing SYBR Green nucleic acid stain (Molecular Probes, Eugene, OR), and banding patterns were documented using a FluorChem Imaging System (Alpha Innotech Corp., San Leandro, CA).
For DGGE analysis of bacteria, the V3 region of 16S rRNA genes was amplified using primers 357f-GC and 519r (Larue et al., 2005). The PCR amplifications were conducted in a total volume of 50 µL containing 100 pmol of each primer, 125 µM of each deoxyribonucleotide triphosphate, 2 mM magnesium chloride, 0.05% BSA, 1x PCR buffer, and 5.0 units of Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA). The thermoprofiles of PCR were the same as those reported previously (Yu and Morrison, 2004). Negative controls without DNA templates were included in parallel. The PCR products were first verified on a 1% agarose gel and then resolved on a 7.5% polyacrylamide gel (37.5:1) with a 40 to 70% denaturing gradient for 1,600 volt-hours (Yu and Morrison, 2004). Staining and visualization of gels were performed as described previously. For DGGE analysis of methanogens, a methanogen-specific forward primer (344f-GC; Raskin et al., 1994) was used with the primer 519r. The PCR and DGGE conditions were the same as those described previously for bacteria.
Cluster Analysis of Banding Patterns
The TIFF images of the previously documented DGGE banding patterns were imported into the database in the BioNumerics software (BioSystematica, Devon, UK). Automated and manual alignment of the banding patterns to adjust for any possible migration differences was performed by pattern recognition using external reference markers and internal reference bands. The DGGE bands in the bacterial and methanogen gels were detected using the band-searching algorithm. The calling of bands was further manually refined to find and mark uncertain bands. After normalization of the gels, only those bands with a peak height intensity exceeding 5% of the strongest band in each lane were included in further analyses. Comparisons were accomplished by cluster analysis using the Jaccard method, which accounts for number of bands shared among samples. The position tolerance was set at 1% to allow for migratory differences, and uncertain bands were ignored. To illustrate pairwise similarities between all lanes in each gel, the dendrogram was constructed using the unweighted-pair group method with arithmetic mean using average linkages.
Excision of Bands and DNA Sequencing
Representative single bands within each subperiod in each gel were cored from DGGE gels using sterile pipette tips and transferred to sterile 0.5-mL tubes containing 30 µL of Tris-EDTA buffer. Tubes were subjected to 3 cycles of freezing (–80°C for 10 min) and thawing (55°C for 15 min) to enhance diffusion of DNA fragments into the buffer (Yu et al., 2008). The amplicons were reamplified in PCR with respective primers and conditions described previously, purified using a QIAquick PCR purification kit (Qiagen, Valencia, CA), and sequenced at the Plant-Microbe Genomics Facility at The Ohio State University. From the methanogen DGGE gels, 23 reamplified PCR products were selected and cloned using the TA cloning kit (Invitrogen). Four recombinant clones from each cloning reaction were selected and screened by colony PCR, and plasmid DNA was extracted from these transformants using the QIAprep Spin Miniprep kit (Qiagen) for DNA sequencing. Similarly, 19 of the bands cut from the bacterial DGGE gels were cloned as described above, but 5 recombinant clones were selected for sequencing. The DNA sequences were checked for errors by visual evaluation of their respective electropherograms and tested for chimera formation using the Chimera-Check program (Cole et al., 2003). Sequence similarity searches were performed using BLASTn (Altschul et al., 1997), and the highest identity to a known sequence is reported.
|
RESULTS AND DISCUSSION
|
|---|
Generic Counts and DGGE Analysis of Protozoa
No protozoa were observed microscopically during subperiod 2. In the faunated subperiod, total counts of protozoa were highest (P = 0.05) for the 5% added fat diet (Karnati et al., 2009), but there was a trend (P = 0.12) for monensin to decrease the counts of Epidinium relative to control (Table 1). Monensin did not inhibit Epidinium in one study with dairy cows (Oelker et al., 2009), but a similar selective toxicity of monensin to Epidinium was observed in another study (Reveneau, 2008). In the latter study, though, Epidinium was one of the few genera not inhibited by coconut oil. One possible explanation for the selective toxicity of monensin to Epidinium in the current study could be differences in the cell membrane structure between the different genera of protozoa. Alternatively, Epidinium is fibrolytic (Williams and Coleman, 1992), and plant particles probably have high concentrations of adsorbed monensin (Chow et al., 1994), perhaps increasing the toxicity of monensin toward Epidinium.
View this table:
[in this window]
[in a new window]
|
Table 1. Generic distribution of protozoa in continuous culture fermenters provided control, fat, monensin, or bromoethanesulfonate (BES) treatments1
|
|
The protozoa-specific DGGE analysis revealed similar banding patterns for all treatments. Two prominent bands and 2 faint bands were amplified from all of the samples from the faunated subperiod in all 4 experimental periods (data not shown).
Analysis of Methanogens Using DGGE
Multiple bands were resolved on the methanogen-specific DGGE gels from each sample (Figure 1). These gels were loaded by faunation status, and the banding patterns were then merged using the BioNumerics software for cluster analysis, which provides a method of objectively determining groups of similar patterns based on the normalized gels using external (molecular weight markers) and internal reference bands. Also, DNA sequencing revealed that co-migrating bands (for example bands 2 and 3, versus bands 14 and 15 in the defaunated and faunated lanes, respectively, in Figure 1) were from the same methanogens (99% identity with uncultured Methanobrevibacter spp. CSIRO2.04 in Table 2) (Wright et al., 2004). Similarly, bands 5 and 17 had the same sequence matches (99% identity with uncultured archaeon clone CN10–3 from ovine rumen; Ohene-Adjei et al., 2008).
View larger version (93K):
[in this window]
[in a new window]
|
Figure 1. Denaturing gradient gel electrophoresis profiles of methanogens produced from community DNA samples from continuous culture fermenters provided control (Con), fat, monensin (Mon), or bromoethanesulfonate (BES) treatments. The lanes are labeled with the period (P) number, fermenter (F) number, defaunated (Def) or faunated (Faun) sub-period, and treatment. The calculated similarity coefficients determined using Bionumerics software are shown on the left-hand side. Sequence characterization of the labeled bands is shown in Table 2.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Characterization of bands excised from denaturing gradient gel electrophoresis gel of methanogens in continuous culture fermenters provided control, fat, monensin, or bromoethanesulfonate treatments
|
|
Of the 23 methanogen bands that were reamplified and cloned, 4 bands (bands 3, 6, 11, and 13 in Figure 1) seemed to have more than one co-migrating sequence (Table 2). For these 4 bands, none of the 4 cloned sequences from each band were identical. However, for bands 3, 6, and 11, the 4 clone sequences in each respective band identified most closely with 2 sets of sequences retrieved from the database; for band 13, the 4 clone sequences identified with 3 retrieved sequences representing characterized methanogens. The profiles of the samples clustered by faunation status except for one lane from the BES treatment in the defaunated subperiod, which clustered with the lanes from the BES treatment in the faunated subperiod. One BES treatment (period 2, fermenter 3) did not cluster with the other 3 BES-faunated treatments. The same BES treatment in the defaunated subperiod also did not cluster with other BES treatments. The other treatments did not appear to have any effect on the clustering pattern. Repetition of DGGE loading by faunation status or by period for methanogens provided similar banding patterns. Because the original inoculum was changed every period, the general grouping by faunation status to overcome the predominant period grouping seems to support the importance of protozoal predation or symbiosis to influence archaeal community structure.
Based on DNA sequencing of the bands, most of the abundant methanogens belonged to the family Methanobacteriaceae. However, some of the bands (4, 5, 16, and 17 in Figure 1) were most closely represented by a novel group of rumen archaea (Table 2) that is considered atypical for the rumen environment and were suggested to belong to a putative new order (Wright et al., 2004). Defaunation caused decreased intensity of bands present in the faunated subperiod (bands 13, 15, and 21 in Figure 1), and these bands produced sequences of methanogens associated with protozoa (99% identity with clones CAw563_TA, CAw273_ILP, CAw464_ID, and CAw448_ID, Table 2). These matching clone sequences were previously obtained from DGGE gels from rumen samples of sheep developed by inoculating unfaunated sheep with specified protozoal species from other donor sheep (Ohene-Adjei et al., 2007).
Two DGGE bands present in other treatments (bands 1, 10, and 11 in Figure 1) were not visualized from the BES treatment in either faunated or defaunated subperiods. In addition, BES also decreased the intensity of the bands in the faunated period (bands 12 and 13 in Figure 1), which disappeared from other treatments after subsequent defaunation. These 5 bands all revealed 98 to 100% sequence identity with clones CAw464_ID, CAw563_TA, and CAw273_ILP (Table 2), which represent the protozoa-associated methanogens from the rumen of sheep (Ohene-Adjei et al., 2007). Because BES does not directly affect protozoa and bacteria (Nagaraja et al., 1997), the increased generation time of protozoa from 43.2 to 55.6 h by BES (Karnati et al., 2009) indicates a close positive association of these endosymbionts. This explanation is corroborated by the premise that methanogens that cluster near the hydrogenosomes can keep the hydrogen partial pressure low to stimulate protozoal fermentation pathways toward higher ATP yields, partially compensating for the limited oxidative capacity of the hydrogenosomes compared with mitochondria in aerobic protozoa (Lange et al., 2005).
No new bands appeared, but the intensified bands (numbered 18, 19, 20, 22, and 23 in Figure 1) suggest that BES did partially select for nonsensitive methanogens. Sequences recovered from these bands were most closely associated with Methanobrevibacter smithii ATCC 35061 (99% identity) and other Methanobrevibacter spp. (97 to 99% identities; Table 2). In contrast, Ungerfeld et al. (2004) noted that BES (250 µM) inhibited methane production of a strain of Methanobrevibacter ruminantium for 6 d without any adaptation. The presence of feed or other populations present in our mixed cultures might have decreased BES concentrations below effective thresholds needed to suppress methanogenesis (Karnati et al., 2009) but not enough to allow certain shifts in methanogen populations, particularly those that might be adapted to living in symbiosis with protozoa. Denman et al. (2007) noted that bromochloromethane increased the diversity index of methanogens. Although those authors used the methyl coenzyme-M reductase A gene for archaeal community analyses, they noted that phylogenetic distributions in other published studies were very similar when comparing this gene to those using 16S rDNA. An alternative explanation to increased diversity that should be considered is that, with a limited number of sequences (n = 50), inhibition of dominant representatives might allow compensatory recovery of a greater number of clone sequences representing phylotypes that were less statistically likely to be recovered in the control (Firkins and Yu, 2006). Our DGGE banding analysis only represents dominant members of the community (Mackie et al., 2007) but did not support an increase in recovery of bands with BES.
Monensin also did not suppress methane production in the presence or absence of protozoa (Karnati et al., 2009), supporting a lack of response in DGGE banding analysis in the current report. Compared with BES, dietary fat elicited the opposite responses; there was no major response on methanogen populations even though it tended (P = 0.07) to increase methane production (Karnati et al., 2009). Ruminal methanogens are much more diverse than previously thought, whether assessed by a procedure similar to our approach (Nicholson et al., 2007) or by a more complete analysis based on clone libraries from community DNA from several steers (Wright et al., 2007). Even after a prolonged adaptation, the daily suppression of methanogenesis by bromochloromethane required a 6-h lag before a concomitant suppression of methanogenic abundance (Denman et al., 2007). Firkins and Yu (2006) cited other instances in which methane production was uncoupled from abundance of methanogens when inhibitors were used. Inhibitors can be overcome by selection for less sensitive populations or by an accumulating concentration of H2 eventually pushing the metabolic flux to increase substrate concentration relative to the concentration of a competitive inhibitor. Therefore, attempts to decrease methane production might be more successful only when this diversity is better characterized and related to the total microbial ecosystem.
Analysis of Bacteria Using DGGE
In contrast with the methanogens, when all the faunated samples were loaded on one gel and the defaunated samples loaded on another (data not shown) for analysis of bacterial communities, the gels tended to cluster more strictly by gel than when they were loaded by period, perhaps because of minor variations in migrating conditions and gel composition. The banding patterns were more complex than with the archaeal DGGE analyses. Consequently, for DGGE analysis of bacteria, the PCR products were loaded within gels by period to allow a comparison between faunated and defaunated subperiods (Figures 2 and 3). When loaded by period, our banding analyses agreed with a previous DGGE cluster analysis documenting shifts in bacterial populations in control versus protozoa-free sheep (Yáñez-Ruiz et al., 2007).
View larger version (90K):
[in this window]
[in a new window]
|
Figure 2. Denaturing gradient gel electrophoresis profiles of bacteria produced from community DNA samples from continuous culture fermenters provided control (Con), fat, monensin (Mon), or bromoethanesulfonate (BES) treatments. The lanes are labeled with the period (P) number, fermenter (F) number, defaunated (Def) or faunated (Faun) sub-period, and treatment. The calculated similarity coefficients are shown on the left-hand side of the figure. The labeled bands in the ovals represent successfully characterized bands, whereas sequence characterization was not successful for bands enclosed in boxes. Sequence characterization of the labeled bands is shown in Table 3.
|
|
View larger version (84K):
[in this window]
[in a new window]
|
Figure 3. Denaturing gradient gel electrophoresis profiles of bacteria produced from community DNA samples from continuous culture fermenters provided control (Con), fat, monensin (Mon), or bromoethanesulfonate (BES) treatments. The lanes are labeled with the period (P) number, fermenter (F) number, defaunated (Def) or faunated (Faun) sub-period, and treatment. The calculated similarity coefficients are shown on the left-hand side of the figure. The labeled bands in the ovals represent successfully characterized bands, whereas sequence characterization was not successful for bands enclosed in boxes. Sequence characterization of the labeled bands is shown in Table 4.
|
|
Because dendrograms were grouped by the software based on objective criteria established a priori, consistent grouping of treatments to overcome effects of different inoculum in each period provides strong evidence for shifts in bacterial populations. Approximately 97% sequence identity with the full-length 16S rRNA gene is typically accepted as being within the same taxonomic unit (Firkins and Yu, 2006; Wallace, 2008). Our comparisons of sequence identities of all deposited sequences among various strains of 5 representative bacterial species were comparable in range for the full-length 16S rDNA versus those for the corresponding V3 region (Firkins et al., 2008). Even if this result supports a relatively similar cutoff for this V3 region, the highest sequence identities shown in Tables 3 and 4 should not be misconstrued as an attempt to specifically identify the corresponding band’s taxonomic unit and its niche, particularly for those bands from which recovered sequences declined progressively below 97% identity. However, even for those bands that were best-matched to sequences of characterized species below this 97% taxonomic cutoff, if multiple bands were strengthened or weakened consistently by treatment, overcoming the effect of a different inoculum in each period, this shift provides a useful inference for the treatment.
View this table:
[in this window]
[in a new window]
|
Table 3. Characterization of bands excised from denaturing gradient gel electrophoresis gel of bacteria from experimental periods 1 and 2 in continuous culture fermenters provided control, fat, monensin, or bromoethanesulfonate treatments
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Characterization of bands excised from denaturing gradient gel electrophoresis gel of bacteria from experimental periods 3 and 4 in continuous culture fermenters provided control, fat, monensin, or bromoethanesulfonate treatments
|
|
As reported previously (Kocherginskaya et al., 2001; Klieve et al., 2007), the sequences obtained from direct sequencing of DNA recovered from the bands might not be closely identified with clone sequences recovered from the same band. However, the presentation of the best-matched sequences from the database again helped document that this result was uncommon in our study. It also demonstrates the general consistency in analysis of co-migrating sequences recovered after cloning. Thus, our results support the conclusion by Zoetendal et al. (2004), who ranked DGGE as a useful tool for community analysis when sequence identity could be related to the bands representing the predominant bacterial populations (Mackie et al., 2007). With consideration to alternative approaches, Firkins and Yu (2006) argued that random cloning and sequencing approaches have potential biases but are especially limited by sheer statistical probability when inadequate numbers of sequences are used; therefore, random clone libraries from each of our 32 samples would have exceeded the scope of our study’s aims to compare treatments rather than to assess diversity for just a few samples. Using a battery of real time PCR targets, Weimer et al. (2008) recently performed an extensive analysis of bacterial targets after the addition or removal of monensin. The analysis was limited to 2 cows, and the sum of their 16S rDNA copies targeting species assumed to be predominant from culture-based studies included only a modest proportion of the copies using primers targeting total bacteria. Although revealing interesting treatment differences in the targeted populations chosen, such an approach also would have limited our coverage of potential shifts in populations.
About 56 of the 95 bands excised could not be resolved by direct sequencing, probably because of the presence of multiple sequences. Even when there were multiple sequences among clones associated within the same bands, we noted previously that these sequences were more homogeneous as they migrated through the gel. Most of the sequences obtained from both bacterial-specific DGGE gels matched with gram-positive bacteria, although some bands (13 and 17 in Figure 2) were most closely related to prevotellas. When using universal primers, clone libraries from full-length 16S rDNA can be predominantly represented by either Firmicutes or Bacteroidetes (Firkins and Yu, 2006). These approaches typically bias against Fibrobacter succinogenes (Larue et al., 2005), although it is not known if a similar bias exists for the smaller fragment sizes targeted with PCR-DGGE.
Comparison of sequences resolved from co-migrating bands from neighboring lanes revealed that these bands (i.e., bands 5 and 12, bands 14 and 16, and bands 15 and 17 in Figure 2) represented the same or very similar phylogenetic groups of bacteria between different subperiods (Table 3). We accepted results from clone sequences as more definitive except in some cases in which the direct sequence was associated with bands that had a low ability to explain treatment differences or when the sequence only had a very few ambiguous bases (i.e., when there was not a dominating peak corresponding with a base position in the electropherogram).
In periods 1 and 2 (Figure 2), the defaunated lanes clustered together, whereas the faunated lanes from period 1 formed a separate cluster. In periods 3 and 4 (Figure 3), the lanes clustered by faunation status. Within faunation status in both gels, the bands formed period-specific clusters. Because each period was initiated by a new inoculum, the period effect might also hide potential treatment effects (Noftsger et al., 2003). One of the samples (Period 4, fermenter 1) exhibited very few bands and branched off from other lanes (Figure 3), perhaps because of a mistake in loading the gel. When the experiment was repeated, this sample clustered within period 4 in the faunated subperiod (data not shown).
In periods 1 and 2, bands that were most closely related with Ruminococcus gnavus (band 2), clostridia (bands 1, 3, and 6), and lipolytic clostridia (bands 8 and 9) seemed to diminish in the same treatments during the defaunated subperiod (Figure 2). However, compared with the respective faunation subperiods, defaunation intensified bands that best identified with Pseudobutyrivibrio spp. in the fat-supplemented diet (bands 15, 17, and 20), which represent a group of bacteria that is diverse in substrate degradation (Krause et al., 2003; Cotta and Forster, 2006) and lipid metabolism (Paillard et al., 2007). Defaunation also strengthened the bands that matched with Ruminococcus spp. (band 18) in period 2.
As expected, the co-migrating bands near the top of the gel (i.e., bands 7 and 10 as well as bands 21 and 22 in Figure 2) produced different sequences, suggesting they are probably heteroduplexes. Similarly, co-migrating bands that did not migrate far into the gel (bands 32, 36, and 43) were not uniformly related to a single group in periods 3 and 4 (Figure 3). On the other hand, bands that were excised from further in the gel (bands 27 and 31; bands 34, 35, and 38) usually revealed the same or very similar phylogenetic groups of bacteria between different subperiods (Table 4). Bands representing Pseudobutyrivibrio ruminis (bands 27 and 31) were present in all the lanes. Defaunation in both periods, except for fermenter 2, strengthened bands that most closely related with the lipolytic Clostridium tetanomorphum (bands 34, 35, and 38) and Butyrivibrio fibrisolvens H17c. The latter has been described as a xylanolytic and lipolytic eubacterium in the rumen that clustered with stearic acid-producing bacteria related to Clostridium proteoclasticum (Paillard et al., 2007). In period 4, except for the BES treatment, defaunation also strengthened the band (band 42) that best matched with C. tetanomorphum.
In general, defaunation decreased the recovery of sequences most closely identifying with Clostridium while increasing those representing Prevotella and Eubacterium spp. In the faunated subperiod, there were multiple (
3 each) bands that most closely related to Clostridium symbiosum, Clostridium coccoides, R. gnavus, and Anaerovibrio lipolytica that were not recovered in the defaunated subperiod, which had multiple bands best matching with Pseudo. ruminis, C. tetanomorphum, and Ruminococcus bromii. Although not definitive, these results suggest that a shift in bacterial populations resulting from defaunation influenced fatty acid metabolism. Using real-time PCR, Ozutsumi et al. (2006) reported a greater number of clone sequences most closely related with Clostridium, Ruminococcus, and Prevotella in unfaunated (separated at birth) than in faunated cattle. In the same study, the numbers of Fibrobacter sequences were greater in faunated than in unfaunated cattle. Removal of protozoa is usually associated with a decrease in cellulolytic activity (Williams and Coleman, 1997), but defaunation of our fermenters increased NDF and OM digestibility (Karnati et al., 2009). The banding patterns combined with sequence analysis supports the likelihood that the absence of protozoal predation of cellulolytic bacteria (Koenig et al., 2000) without any benefit needed by protozoa to help maintain pH or quench O2 under our conditions explains the increased digestibility in the defaunated subperiod (Karnati et al., 2009).
Defaunation decreased the ratio of vaccenic acid:unsaturated fatty acid (Karnati et al., 2009), suggesting that lipolysis of triglycerides or BH was perhaps limiting in the defaunated fermenters because of loss of protozoal lipolytic activity and reservoirs of unsaturated fatty acids (Devillard et al., 2006). Defaunation also decreased the number of clones most closely associated with A. lipolytica (bands 10 and 11, Figure 2) while increasing those associated with cluster XIVa of the Clostridium subphylum, which includes Butyrivibrio and Pseudobutyrivibrio as well as other genera such as Clostridium, Lachnospira, Eubacterium, Ruminococcus, and Roseburia (Cotta and Forster, 2006). These gram-positive bacteria have variable cellulolytic, xylanolytic, amylolytic, pectinolytic, proteolytic, and lipolytic activities in the rumen. Fatty acid BH in this group is linked with mechanism of butyrate formation, lipid metabolism, and sensitivity to growth by linoleic acid in the presence of lactate; classification based on these criteria into 3 broad subgroups might help explain variable accumulation of BH intermediates among studies (Jenkins et al., 2008). The consistent representation of butyrivibrios in the presence of monensin under our conditions and with our dosage rate contradicts strong inhibition of pure cultures of B. fibrisolvens, apparently through bacterial adaptation mechanisms (Weimer et al., 2008). In the latter report, the 16S rDNA copies of B. fibrisolvens were consistently low but still were decreased (P < 0.05) by monensin, whereas among those not affected was Eubacterium ruminantium, which also can contribute to BH and is taxonomically related to B. fibrisolvens. The current data also add the dimension that susceptibility to protozoal ingestion could influence the relative representation by specific butyrivibrios. For example, it is not well known if clumping and adherence of bacteria to starch and fiber particles hides them from predation or, conversely, provides the vehicle for their uptake.
|
CONCLUSIONS
|
|---|
The use of inhibitors (fat, monensin, and BES) against mixed ruminal microbes in faunated and defaunated continuous cultures provided a deeper insight into the complexity of microbial interactions than would be available with batch culture or in vivo procedures. The apparent inhibition of protozoa-associated methanogens by BES corresponds with our previous finding that BES increased protozoal generation time, documenting the importance of inter-species hydrogen transfer. Again consistent with our previous findings on the passage of FA to the effluent and with prior conclusions that protozoa incorporate BH intermediates into membranes, the current results extend their role by documenting how the absence of protozoa changes populations of lipid-metabolizing bacteria. Even so, the dominant effects of a different inoculum in each period demonstrate the challenges of community profiling among multiple animals without rapid profiling procedures such as DGGE that can be combined with more exhaustive or emerging technologies to better integrate the influence of dietary conditions on the complex rumen microbial ecology.
|
FOOTNOTES
|
|---|
1 Research was supported by state and federal funds appropriated to the Ohio Agricultural and Development Center, The Ohio State University. Manuscript number 10/08AS. Additional support was also provided by the USDA Cooperative State Research, Education, and Extension Service USDA/NRICGP Grant 2003-35206-12872 and Elanco Animal Health, Greenfield, Indiana. 
Received for publication June 9, 2008.
Accepted for publication March 8, 2009.
|
REFERENCES
|
|---|
Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. and Lipman, D. J.. 1997. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 25:3389–3402.[Abstract/Free Full Text]
Anderson, K. L. 1995. Biochemical analysis of starch degradation by Ruminobacter amylophilus 70. Appl. Environ. Microbiol. 61:1488–1491.[Abstract/Free Full Text]
Attwood, G., Li, D., Pacheco, D. and Tavendale, M.. 2006. Production of indolic compounds by rumen bacteria isolated from grazing ruminants. J. Appl. Microbiol. 100:1261–1271.[CrossRef][Medline]
Avgustin, G., Wallace, R. J. and Flint, H. J.. 1997. Phenotypic diversity among ruminal isolates of Prevotella ruminicola: Proposal of Prevotella brevis sp. nov., Prevotella bryantii sp. nov., and Prevotella albensis sp. nov. and redefinition of Prevotella ruminicola. Int. J. Syst. Bacteriol. 47:284–288.[Abstract/Free Full Text]
Chow, J. M., Van Kessel, J. S. and Russell, J. B.. 1994. Binding of radiolabeled monensin and lasalocid to ruminal microorganisms and feed. J. Anim. Sci. 72:1630–1635.[Abstract]
Cirne, D. G., Delgado, O. D., Marichamy, S. and Mattiasson, B.. 2006. Clostridium lundense sp. nov., a novel anaerobic lipolytic bacterium isolated from bovine rumen. Int. J. Syst. Evol. Microbiol. 56:625–628.[Abstract/Free Full Text]
Cole, J. R., Chai, B., Marsh, T. L., Farris, R. J., Wang, Q., Kulam, S. A., Chandra, S., McGarrell, D. M., Schmidt, T. M., Garrity, G. M. and Tiedje, J. M.. 2003. The Ribosomal Database Project (RDP-II): Previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy. Nucleic Acids Res. 31:442–443.[Abstract/Free Full Text]
Cotta, M., and R. Forster. 2006.The family Lachnospiraceae, including the genera Butyrivibrio, Lachnospira and Roseburia. Pages 1002–1021 in Prokaryotes. Vol. 4. 3rd ed. M. Dworkin, S. Falkow, E. Rosenberg, K.-H. Schleifer, and E. Stackebrandt, ed. Springer New York, NY.
Dehority, B. A. 2003. Rumen Microbiology. Nottingham University Press, Nottingham, UK.
Denman, S. E., Tomkins, N. W. and McSweeney, C. S.. 2007. Quantitation and diversity analysis of ruminal methanogenic populations in response to the antimethanogenic compound bromochloromethane. FEMS Microbiol. Ecol. 62:313–322.[CrossRef][Medline]
Devillard, E., McIntosh, F. M., Newbold, C. J. and Wallace, R. J.. 2006. Rumen ciliate protozoa contain high concentrations of conjugated linoleic acids and vaccenic acid, yet do not hydrogenate linoleic acid or desaturate stearic acid. Br. J. Nutr. 96:697–704.[Medline]
Firkins, J. L., Karnati, S. K. R. and Yu, Z.. 2008. Linking rumen function to animal response by application of metagenomics techniques. Aust. J. Exp. Agric. 48:711–721.[CrossRef]
Firkins, J. L., and Z. Yu. 2006. Characterisation and quantification of the microbial populations in the rumen. Pages 19–54 in Ruminant Physiology, Digestion, Metabolism and Impact of Nutrition on Gene Expression, Immunology and Stress. K. Sejrsen, T. Hvelplund, and M. O. Nielsen, ed. Wageningen Academic Publishers, Wageningen, the Netherlands.
Fonty, G., Joblin, K., Chavarot, M., Roux, R., Naylor, G. and Michallon, F.. 2007. Establishment and development of ruminal hydrogenotrophs in methanogen-free lambs. Appl. Environ. Microbiol. 73:6391–6403.[Abstract/Free Full Text]
Hudson, J. A., Cai, Y., Corner, R. J., Morvan, B. and Joblin, K. N.. 2000. Identification and enumeration of oleic acid and linoleic acid hydrating bacteria in the rumen of sheep and cows. J. Appl. Microbiol. 88:286–292.[CrossRef][Medline]
Janssen, P. H. and Kirs, M.. 2008. Structure of the archaeal community of the rumen. Appl. Environ. Microbiol. 74:3619–3625.[Free Full Text]
Jenkins, T. C., Wallace, R. J., Moate, P. J. and Mosley, E. E.. 2008. Board-Invited Review: Recent advances in biohydrogenation of unsaturated fatty acids within the rumen microbial ecosystem. J. Anim. Sci. 86:397–412.[Abstract/Free Full Text]
Klieve, A. V., O'Leary, M. N., McMillen, L. and Ouwerkerk, D.. 2007. Ruminococcus bromii, identification and isolation as a dominant community member in the rumen of cattle fed a barley diet. J. Appl. Microbiol. 103:2065–2073.[Medline]
Kocherginskaya, S., Aminov, R. I. and White, B. A.. 2001. Analysis of the rumen bacterial diversity under two different diet conditions using denaturing gradient gel electrophoresis, random sequencing, and statistical ecology approaches. Anaerobe 7:119–134.[CrossRef]
Koenig, K. M., Newbold, C. J., McIntosh, F. M. and Rode, L. M.. 2000. Effects of protozoa on bacterial nitrogen recycling in the rumen. J. Anim. Sci. 78:2431–2445.[Abstract/Free Full Text]
Kopecny, J., Zorec, M., Mrázek, J., Kobayashi, Y. and Marin
ek-Logar, R.. 2003. Butyrivibrio hungatei sp. nov. and Pseudobutyrivibrio xylanivorans sp. nov., butyrate-producing bacteria from the rumen. Int. J. Syst. Evol. Microbiol. 53:201–209.[Abstract/Free Full Text]
Krause, D. O., Denman, S. E., Mackie, R. I., Morrison, M., Rae, A. L., Attwood, G. T. and McSweeney, C. S.. 2003. Opportunities to improve fiber degradation in the rumen: Microbiology, ecology, and genomics. FEMS Microbiol. Rev. 27:663–693.[CrossRef][Medline]
Lange, M., Westermann, P. and Ahring, B. K.. 2005. Archaea in protozoa and metazoa. Appl. Microbiol. Biotechnol. 66:465–474.[CrossRef][Medline]
Larue, R., Yu, Z., Parisi, V. A., Egan, A. R. and Morrison, M.. 2005. Novel microbial diversity adherent to plant biomass in the herbivore gastrointestinal tract, as revealed by ribosomal intergenic spacer analysis and rrs gene sequencing. Environ. Microbiol. 7:530–543.[CrossRef][Medline]
Mackie, R. I., I. K. O. Cann, E. Zoetendal, and E. Forano. 2007. Molecular approaches to study bacterial diversity and function in the intestinal tract. Pages 75–107 in Proc. 7th Int. Symp. Nutrition of Herbivores. China Agricultural University Press, Beijing.
Maia, M. R. G., Chaudhary, L. C., Figueres, L. and Wallace, R. J.. 2007. Metabolism of polyunsaturated fatty acids and their toxicity to the microflora of the rumen. Antonie Van Leeuwenhoek 91:303–314.[CrossRef][Medline]
Nagaraja, T. G., C. J. Newbold, C. J. Van Nevel, and D. I. Demeyer. 1997. Manipulation of rumen fermentation. Pages 523–632 in The Rumen Microbial Ecosystem. P. N. Hobson and C. S. Stewart, ed. Chapman and Hall, London, UK.
Nicholson, M. J., Evans, P. N. and Joblin, K. N.. 2007. Analysis of methanogen diversity in the rumen using temporal temperature gradient gel electrophoresis: Identification of uncultured methanogens. Microb. Ecol. 54:141–150.[CrossRef][Medline]
Noftsger, S. M., St-Pierre, N. R., Karnati, S. K. R. and Firkins, J. L.. 2003. Effects of 2-hydroxy-4-(methylthio) butanoic acid (HMB) on microbial growth in continuous culture. J. Dairy Sci. 86:2629–2636.[Abstract/Free Full Text]
Oelker, E. R., Reveneau, C. and Firkins, J. L.. 2009. Interaction of molasses and monensin in alfalfa hay- or corn silage-based diets on rumen fermentation, total tract digestibility, and milk production by Holstein cows. J. Dairy Sci. 92:270–285.[Abstract/Free Full Text]
Ohene-Adjei, S., Chaves, A. V., McAllister, T. A., Benchaar, C., Teather, R. M. and Forster, R. J.. 2008. Evidence of increased diversity of methanogenic archaea with plant extract supplementation. Microb. Ecol. 56:234–242.[CrossRef][Medline]
Ohene-Adjei, S., Teather, R. M., Ivan, M. and Forster, R. J.. 2007. Postinoculation protozoan establishment and association patterns of methanogenic archaea in the ovine rumen. Appl. Environ. Microbiol. 73:4609–4618.[Abstract/Free Full Text]
Ozutsumi, Y., Tajima, K., Takenaka, A. and Itabashi, H.. 2006. Real-time PCR detection of the effects of protozoa on rumen bacteria in cattle. Curr. Microbiol. 52:158–162.[CrossRef][Medline]
Paillard, D., McKain, N., Chaudhary, L. C., Walker, N. D., Pizette, F., Koppova, I., McEwan, N. R., Kopecny, J., Vercoe, P. E., Louis, P. and Wallace, R. J.. 2007. Relation between phylogenetic position, lipid metabolism and butyrate production by different Butyrivibrio-like bacteria from the rumen. Antonie Van Leeuwenhoek 91:417–422.[CrossRef][Medline]
Qiu, X., Eastridge, M. L., Griswold, K. E. and Firkins, J. L.. 2004. Effects of substrate, passage rate, and pH in continuous culture on flows of conjugated linoleic acid and trans C18:1. J. Dairy Sci. 87:3473–3479.[Free Full Text]
Raskin, L., Stromley, J. M., Rittmann, B. E. and Stahl, D. A.. 1994. Group-specific 16S rRNA hybridization probes to describe natural communities of methanogens. Appl. Environ. Microbiol. 60:1232–1240.[Abstract/Free Full Text]
Reilly, K., Carruthers, V. R. and Attwood, G. T.. 2002. Design and use of 16S ribosomal DNA-directed primers in competitive PCRs to enumerate proteolytic bacteria in the rumen. Microb. Ecol. 43:259–270.[CrossRef][Medline]
Reveneau, C. 2008. Dietary source and availability of fatty acids to manipulate ruminal protozoa, metabolism of fat, and milk fatty acid profile in lactating dairy cows. PhD Dissertation. The Ohio State University, Columbus.
Ribeiro, C. V. D. M., Karnati, S. K. R. and Eastridge, M. L.. 2005. Biohydrogenation of fatty acids and digestibility of fresh alfalfa or alfalfa hay plus sucrose in continuous culture. J. Dairy Sci. 88:4007–4017.[Abstract/Free Full Text]
Rieu-Lesme, F., Dauga, C., Fonty, G. and Dore, J.. 1998. Isolation from the rumen of a new acetogenic bacterium phylogenetically closely related to Clostridium difficile. Anaerobe 4:89–94.[CrossRef][Medline]
Rieu-Lesme, F., Morvan, B., Collins, M. D., Fonty, G. and Willems, A.. 1996. A new H2/CO2-using acetogenic bacterium from the rumen: Description of Ruminococcus schinkii sp. nov. FEMS Microbiol. Lett. 140:281–286.[Medline]
Stewart, C. S., Duncan, S. H. and Cave, D. R.. 2004. Oxalabacter formigenes and its role in oxalate metabolism in the human gut. FEMS Microbiol. Lett. 230:1–7.[CrossRef][Medline]
Stewart, C. S., H. J. Flint, and M. P. Bryant. 1997. The rumen bacteria. Pages 10–72 in The Rumen Microbial Ecosystem. P. N. Hobson and C. S. Stewart, ed. Blackie Academic & Professional, New York, NY.
Sylvester, J. T., Karnati, S. K. R., Yu, Z., Morrison, M. and Firkins, J. L.. 2004. Development of an assay to quantify rumen ciliate protozoal biomass in cows using real-time PCR. J. Nutr. 134:3378–3384.[Abstract/Free Full Text]
Sylvester, J. T., Karnati, S. K. R., Yu, Z., Newbold, C. J. and Firkins, J. L.. 2005. Evaluation of a real-time PCR assay for measuring the ruminal pool and duodenal flow of protozoal nitrogen. J. Dairy Sci. 88:2083–2095.[Abstract/Free Full Text]
Ungerfeld, E. M., Rust, S. R., Boone, D. R. and Liu, Y.. 2004. Effects of several inhibitors on pure cultures of ruminal methanogens. J. Appl. Microbiol. 97:520–526.[Medline]
van de Vossenberg, J. L. C. M. and Joblin, K. N.. 2003. Biohydrogenation of C18 unsaturated fatty acids to stearic acid by a strain of Butyrivibrio hungatei from the bovine rumen. Lett. Appl. Microbiol. 37:424–428.[CrossRef][Medline]
Varel, V. H., Tanner, R. S. and Woese, C. R.. 1995. Clostridium herbivorans sp. nov., a cellulolytic anaerobe from the pig intestine. Int. J. Syst. Bacteriol. 45:490–494.[Abstract/Free Full Text]
Wallace, R. J. 2008. Gut microbiology—Broad genetic diversity, yet specific metabolic niches. Animal 2:661–668.
Weimer, P. J., Stevenson, D. M., Mertens, D. R. and Thomas, E. E.. 2008. Effect of monensin feeding and withdrawal on populations of individual bacterial species in the rumen of lactating dairy cows fed high-starch diets. Appl. Microbiol. Biotechnol. 80:135–145.[CrossRef][Medline]
Williams, A. G., and G. S. Coleman. 1992. The Rumen Protozoa. Springer-Verlag, New York, NY.
Williams, A. G., and G. S. Coleman. 1997. The rumen protozoa. Pages 73–139 in The Rumen Microbial Ecosystem. P. N. Hobson and C. S. Stewart, ed. Blackie Academic & Professional, New York, NY.
Wolin, M. J., T. L. Miller, and C. S. Stewart. 1997. Microbe-microbe interactions. Pages 467–491 in The Rumen Microbial Ecosystem. P. N. Hobson and C. S. Stewart, ed. Blackie Academic & Professional, New York, NY.
Wright, A.-D. G., Auckland, C. H. and Lynn, D. H.. 2007. Molecular diversity of methanogens in feedlot cattle from Ontario and Prince Edward Island, Canada. Appl. Environ. Microbiol. 73:4206–4210.[Abstract/Free Full Text]
Wright, A.-D. G., Williams, A. J., Winder, B., Christopherson, C. T., Rodgers, S. L. and Smith, K. D.. 2004. Molecular diversity of rumen methanogens from sheep in Western Australia. Appl. Environ. Microbiol. 70:1263–1270.[Abstract/Free Full Text]
Yáñez-Ruiz, D. R., Williams, S. and Newbold, C. J.. 2007. The effect of absence of protozoa on rumen biohydrogenation and the fatty acid composition of lamb muscle. Br. J. Nutr. 97:938–948.[CrossRef][Medline]
Yu, Z., García-González, R., Schanbacher, F. L. and Morrison, M.. 2008. Evaluations of different hypervariable regions of archaeal 16S rRNA genes in profiling of methanogens by Archaea-specific PCR and denaturing gradient gel electrophoresis. Appl. Environ. Microbiol. 74:889–893.[Abstract/Free Full Text]
Yu, Z. and Morrison, M.. 2004. Comparisons of different hypervariable regions of rrs genes for use in fingerprinting of microbial communities by PCR-denaturing gradient gel electrophoresis. Appl. Environ. Microbiol. 70:4800–4806.[Abstract/Free Full Text]
Ziemer, C. J., Sharp, R., Stern, M. D., Cotta, M. A., Whitehead, T. R. and Stahl, D. A.. 2000. Comparison of microbial populations in model and natural rumens using 16S ribosomal RNA-targeted probes. Environ. Microbiol. 2:632–643.[CrossRef][Medline]
Zoetendal, E. G., Collier, C. T., Koike, S., Mackie, R. I. and Gaskins, H. R.. 2004. Molecular ecological analysis of the gastrointestinal microbiota: A review. J. Nutr. 134:465–472.[Abstract/Free Full Text]
Related articles in JDS:
- Investigating unsaturated fat, monensin, or bromoethanesulfonate in continuous cultures retaining ruminal protozoa. I. Fermentation, biohydrogenation, and microbial protein synthesis
- S. K. R. Karnati, J. T. Sylvester, C. V. D. M. Ribeiro, L. E. Gilligan, and J. L. Firkins
JDS 2009 92: 3849-3860.
[Abstract]
[Full Text]
This article has been cited by other articles:
|
|
|
|
 
A. N. Hristov, M. Vander Pol, M. Agle, S. Zaman, C. Schneider, P. Ndegwa, V. K. Vaddella, K. Johnson, K. J. Shingfield, and S. K. R. Karnati
Effect of lauric acid and coconut oil on ruminal fermentation, digestion, ammonia losses from manure, and milk fatty acid composition in lactating cows
J Dairy Sci,
November 1, 2009;
92(11):
5561 - 5582.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
