Animals

Thirty-two fattening lambs were fed their corresponding total mixed ration (TMR) during the experimental period (at least 5 weeks) as described below: eight control lambs (CTRL, 34 g palm oil/kg TMR), eight lambs fed ground FS, (85 g FS/kg of TMR), eight lambs fed control diet plus QCT (Shaanxi Sciphar Biotechnology Co., Ltd, Xi'an, China) (QCT, 34 g palm oil plus 2 g QCT/kg TMR), and eight lambs fed ground FS plus QCT (FS-QCT, 85 g FS plus 2 g QCT/kg TMR). The four TMRs were formulated to be isoenergetic and isoproteic. The ingredients and chemical composition of TMR are shown in Table 1. The TMR were weighed and supplied ad libitum at 09.00 h every day, and fresh drinking water was always available. All handling practises followed the recommendations of the European Council Directive 86/609/EEC (EC 1986) for the protection of animals used for experimental and other scientific purposes and all the animals were able to see and hear other animals. Further details can be found in Andrés et al. (Reference Andrés, Morán, Aldai, Tejido, Prieto, Bodas and Giráldez2014).

Table 1.

Ingredients (g/kg fresh matter) and chemical composition (g/kg dry matter) of the experimental total mixed rations (TMR)

CTRL, control group (34 g palm oil/kg TMR); FS,  flaxseed group (85 g FS/kg TMR); QCT, quercetin group (CTRL plus 2 g QCT/kg TMR); FS-QCT, FS plus QCT group (85 g FS plus 2 g QCT/kg TMR); FAME, fatty acid methyl esters.

Animals were slaughtered on four different days (two animals per group each day), once they reached the intended body weight (c. 25 kg), just before morning feeding of the rest of the lambs. Total rumen contents of each lamb were sampled and immediately frozen at –80 °C for DNA extraction (32 samples in total). Another fraction of rumen contents of 24 lambs slaughtered in three different days was strained through four layers of cheesecloth into an Erlenmeyer flask with an O₂-free headspace. The filtrated rumen fluid from the lambs being fed the same experimental diet (two animals from each group were slaughtered each day) was mixed (four inocula per day, three incubation series on three different days).

In vitro incubation

The control TMR (CTRL) was used as substrate in in vitro incubations. For this purpose, an aliquot of CTRL diet was ground through a 1 mm screen; then samples of 400 mg were accurately weighed into 120 ml serum bottles. Rumen fluid from each diet was mixed with the buffer solution of Goering & Van Soest (Reference Goering and Van Soest1970; no trypticase added) in a proportion of 1 : 4 (vol/vol) at 39 °C under continuous flushing with carbon dioxide (CO₂). Serum bottles were pre-warmed (39 °C) prior to the addition of 40 ml of buffered rumen fluid under CO₂ flushing. Then, bottles were sealed with rubber stoppers and aluminium caps and incubated at 39 °C.

Three incubation series were performed on different days (3 days of slaughter), so that each treatment was conducted in triplicate. On each incubation day a total of 16 bottles were incubated (4 inocula × 1 substrate × (2 repetitions + 2 blanks)), for each inoculum (a mix of rumen fluid from two lambs from the same group) two bottles with substrate (control diet) and two additional bottles without substrate (blanks), which were included to correct the gas production values and methane production for gas release from endogenous substrates. After 20 h of incubation the gas production of each bottle was measured using a pressure transducer (Delta Ohm DTP704-2BGI, Herter Instruments SL, Barcelona, Spain) and a sample of gas (about 15 ml) was removed and stored in a Haemoguard Vacutainer (BD Vacutainer^®, Oxford, UK) for methane analysis. Immediately, bottles were opened, pH was measured with a pH-meter (model 507, CRISON, Alella, Barcelona, Spain), and fermentation was stopped by swirling the bottles in ice. A sample (1 ml) of the supernatant was added to 1 ml of a deproteinizing solution (100 g of metaphosphoric acid and 0·6 g of crotonic acid/l) for volatile fatty acid (VFA) determination, and 2 ml of content were added to 2 ml of hydrochloric acid (HCl), 0·5 m for ammonia-N analysis. The contents of the bottles were transferred to previously weighed filter crucibles (pore size 100–160 µm) and filtered under vacuum to determine dry matter (DM) disappearance.

DNA extraction and quantification of rumen contents (in vivo samples)

Frozen rumen samples were freeze-dried and thoroughly homogenized before DNA extraction by physical disruption (1 min) using a bead beater (Mini-bead Beater, BioSpec Products, Bartlesville, OK, USA) and subsequent DNA purification with the QIAamp DNA Stool Mini Kit (Qiagen Inc., Valencia, CA, USA), with the modification of a greater temperature (95 °C) to improve cell lysis. The DNA samples were used as templates for quantitative real-time polymerase chain reaction (qPCR) amplification and terminal restriction fragment length polymorphism (T-RFLP) analysis.

Cloning of gene fragments of universal bacteria, Butyrivibrio and ciliate protozoa

For qPCR quantification, plasmid DNA used as PCR standards were obtained by PCR cloning of partial 16S rRNA gene fragments of universal bacteria and Butyrivibrio species (Butyrivibrio vaccenic acid-producing bacteria, Butyrivibrio stearic acid-producing bacteria and Butyrivibrio spp.), whereas a partial 18S rRNA gene fragment was cloned for PCR standards of ciliate protozoa. These gene fragments were amplified (see primers in Table 2) using genomic DNA extracted from the rumen contents of a cannulated sheep fed a forage diet. Then PCR products were isolated using a gel extraction kit (QIaquick Gel Extraction Kit, Qiagen Inc., Valencia, CA, USA) after agarose gel electrophoresis, and finally cloned into pGEM-T Easy Vector System II (Promega, Madison, WI, USA) according to the manufacturer's instructions. Escherichia coli strains JM109 were transformed by heat shock at 42 °C. The transformed cells were initially plated onto Luria-Bertani (LB) agar plates supplemented with ampicillin (100 µg/ml) to which 25 µl of 50 mg/ml 5-bromo-4-chloro-3-indolyl-ß-D-galactoside (X-gal) and 40 µl of 100 mmol/l isopropyl-ß-D-thiolgalactopyranoside (IPTG) were applied. Then, the LB/ampicillin/IPTG/X-gal plates were held overnight at 37 °C. Clones were identified from selected white colonies (positive transformants) and plasmid DNA was purified with a Kit QUIagen: QIAprep Miniprep (Qiagen Inc., Valencia, CA, USA) according to the manufacturer's instructions. For preliminary characterization of the PCR products, the inserts were PCR amplified and then sequenced. Their sequence similarities were assessed by Basic Local Alignment Search Tool (BLAST; http://www.ncbi.nlm.nih.gov/BLAST/). Then, plasmids containing respective target DNA fragments were used as standards in qPCR.

Table 2.

Primers used in the current study for quantitative real time polymerase chain reaction (qPCR) and terminal restriction fragment length polymorphism (T-RFLP)

* Molecular beacon probe.

Quantitative real-time polymerase chain reaction analysis

Quantitative real-time PCR was carried out using the Applied Biosystems StepOne Plus™ real time PCR system (Applied Biosystems, Foster City, CA, USA) to investigate the abundance of ciliate protozoa and three bacterial groups potentially able to perform the rumen BH: Butyrivibrio VA-, Butyrivibrio SA- and Butyrivibrio spp. The gene-targeted primer sets used for qPCR analysis are presented in Table 2. All qPCR amplifications were performed in triplicate with SYBR Green Supermix (Takara Bio Inc., Otsu, Shiga, Japan) as previously described (Maeda et al. Reference Maeda, Fujimoto, Haruki, Maeda, Kokeguchi, Petelin, Arai, Tanimoto, Nishimura and Takashiba2003), using a 20 µl reaction mixture and 0·85 µl of each primer (10 µm) for total bacteria and 0·4 µl for Butyrivibrio VA-producing bacteria, Butyrivibrio spp. and ciliate protozoa. For the detection of the Butyrivibrio SA-producing bacteria (B. proteoclasticus group) using qPCR, the molecular beacon approach with the primers and probe designed by Paillard et al. (Reference Paillard, McKain, Rincon, Shingfield, Givens and Wallace2007) was employed.

Quantification of universal bacteria, Butyrivibrio SA-producing bacteria, Butyrivibrio spp. and ciliate protozoa was based on standard curves obtained from the amplification profile of known concentrations of the standard plasmid for the respective target. For this purpose the concentration of total DNA was measured by spectrophotometry (NanoDrop ND-1000 Spectrophotometer, Nanodrop Technologies, Wilmington, DE, USA) and the corresponding copy number was calculated according to Lee et al. (Reference Lee, Kim, Shin and Hwang2006). Then, a 10-fold dilution series of the standard plasmid for the respective target was run along with the samples. The qPCR efficiency ranged between 94 and 100% for the analysis of rumen samples for both biohydrogenating groups and universal bacteria. Negative controls without DNA template were run with every assay to assess the overall specificity.

Quantification of Butyrivibrio VA-producing bacteria was finally based on standard curves obtained from dilutions of purified genomic DNA from a control strain (B. fibrisolvens JW11), since standard plasmid failed to amplify.

Terminal restriction fragment length polymorphism analysis

Terminal RFLP is a technique used to rapidly distinguish bacterial communities based on the size of the terminal fragments of 16S rRNA genes following restriction endonuclease digestion (Liu et al. Reference Liu, Marsh, Cheng and Forney1997).

Two different T-RFLP analyses were performed, firstly using a universal bacteria-specific primer pair set (Hongoh et al. Reference Hongoh, Yuzawa, Ohkuma and Kudo2003) and secondly using Butyrivibrio group-specific primers (Boeckaert et al. Reference Boeckaert, Vlaeminck, Fievez, Maignien, Dijkstra and Boon2008; Table 2). Both forward primers were labelled with 6-carboxy-fluorescein. Fluorescence-labelled PCR products were purified using the GFX PCR DNA and Gel Band Purification kit (GE Healthcare, Roosendaal, Netherlands). Then, the DNA was quantified by spectrophotometry (NanoDrop ND-1000 Spectrophotometer, Nanodrop Technologies, Wilmington, DE, USA) and 100 ng of each sample was subjected to a restriction reaction with enzyme HaeIII for universal bacteria and HhaI for Butyrivibrio group for 12 h at 37 °C. Fluorescence-labelled terminal restriction fragments were analysed by capillary electrophoresis on an automatic sequence analyser Megabace 500 (GE Healthcare, Roosendaal, Netherlands) with internal Et-ROX-labelled DNA size standards (Amersham Biosciences, Fairfield, CT, USA).

Fatty acid analysis

All details regarding lipid extraction and FA determination (gas chromatography conditions, peak separation and identification: 100 m SP2560, fatty acid methyl esters –FAME– standards) have been previously described in Aldai et al. (Reference Aldai, Lavín, Kramer, Jaroso and Mantecón2012). The FAMEs were expressed as percentage (%) of total FAME.

Analytical procedures

Dry matter, ash and nitrogen (N) were determined according to the Association of Official Analytical Chemists (AOAC 1999). The procedure described by Van Soest et al. (Reference Van Soest, Roberston and Lewis1991) was used for the analysis of neutral detergent fibre (NDF), whereas acid detergent fibre (ADF) and acid detergent lignin (ADL) were determined according to the procedures proposed by AOAC (1999). Concentrations of VFA and NH₃-N were determined according to Carro & Miller (Reference Carro and Miller1999).

Methane was analysed by injecting 0·3 ml of gas into a gas chromatograph (Shimadzu GC 14B; Shimadzu Europa GmbH, Duisburg, Germany), heated at 200 °C and equipped with a flame ionization detector and a column packed with Carboxen 1000 (Supelco, Madrid, Spain). Helium was used as the carrier gas and peaks were identified by comparison with a standard of known methane concentration.

Statistical analyses

The qPCR data were subjected to a two way analysis of variance (ANOVA), using the General Linear Model (GLM) procedure of Statistical Analysis Software (SAS) (SAS 2012) according to the following model:

$${y_{ijk}} = \mu + {\rm F}{{\rm S}_i} + {\rm QC}{{\rm T}_j} + {\left( {{\rm FS} \times {\rm QCT}} \right)_{ij}} + {\varepsilon _{ijk}}$$

where y _ijk is the dependent variable, μ is the overall mean, FS is the effect of FS addition, QCT is the effect of QCT addition, FS × QCT is the effect of the interaction between QCT and FS, and ε _ijk is the residual error. Least square means were generated and separated using the PDIFF option of SAS for main or interactive effects, with significance declared at P < 0·05.

The model used to analyse the in vitro incubation data differed slightly. After averaging vial replicates, the 12 observations were subjected to ANOVA, with inocula (CTRL, QCT, FS and FS-QCT) as fixed factor and the batch as a blocking factor (B _k ). Consequently, the model previously described also included B _k for in vitro incubation data.

Finally, the T-RFLP data [matrix showing the presence (1) or absence (0) of the peaks, i.e., the presence of operational taxonomic units] obtained for each ruminal sample were analysed by principal components analysis (PCA), using the Unscrambler software version 8.0.5 (CAMO, Oslo, Norway).