In vitro incubations with ruminal fluid

All experimental procedures were approved by the ethical commission of the Institute of Agricultural and Fisheries Research, Belgium (ILVO, approval number EC2009/114). Batch in vitro incubations were established using rumen contents collected from adult sheep according to Vlaeminck et al. (Reference Vlaeminck, Braeckman and Fievez2014). Rumen contents were obtained from three mature wethers fitted with ruminal cannulae fed grass hay ad libitum and a grain-based concentrate (200 g/day) twice daily at 0900 and 1700 h to meet their maintenance requirements (van der Meer, Reference Van der Meer1985). Approximately 0.5 l of ruminal digesta was collected from each animal just before the morning feeding. The rumen fluid was filtered through a sieve with a pore size of 1 mm under a stream of CO₂ at 39°C. Donor fluid from each animal were kept separate except for experiment 1 where donor rumen fluid was combined. Rumen inocula was diluted (1 : 4, v/v) in bicarbonate/phosphate buffer containing a mixture of fermentable substrates (0.21 g glucose, 0.21 g cellobiose, 0.19 g xylose, 0.19 g arabinose, 1.3 g acid casein hydrolysate, 1.3 g peptone and 1.3 g yeast extract per litre of distilled water (Vlaeminck et al., Reference Vlaeminck, Braeckman and Fievez2014)). Diluted buffered rumen fluid (25 ml for experiment 1, 50 ml for experiment 2 and 500 ml for experiment 3) was added to incubation flasks (120 ml flasks for experiments 1 and 2 and 1 l flasks for experiment 3) containing none (control) or one of four adsorbants and 22:6n-3 (Nu-check-Prep, Elysian, MN, USA) dissolved in ethanol (20 mg/ml) according to the experimental design (c.f. below). On the day before inoculation, adsorbents were added to the incubation flasks. On the day of inoculation, the 22:6n-3 solution was added after which the flasks were closed. After flushing with CO₂ during five cycles of 900 mbar underpressure and 600 mbar overpressure, diluted buffered rumen fluid was added. Diluted buffered rumen fluid was added between 30 and 45 min after adding the 22:6n-3 solution, allowing 22:6n-3 to interact with the adsorbant before inoculum addition. Before incubation flasks were placed at 39°C, overpressure was released. Flasks were maintained under anaerobic conditions at 39°C, with intermittent shaking in a batch culture incubator (Edmund Bühler GmbH, Hechingen, Germany). At the end of each incubation period, the flasks were removed from the incubator and the reaction was stopped by cooling the flasks in an ice bath. Sub-samples were collected for analysis of volatile FA (VFA: 2 ml) and long-chain FA (LCFA: 5 ml). All treatment combinations were performed in duplicate (technical replicates).

Experiment 1 was performed as a 9×3 factorial experiment in a randomized block design to examine the effect of four adsorbants (mucin, gum arabic, bentonite and silicic acid) at two different concentrations (1 and 5 mg/ml) plus a control treatment on the disappearance and metabolism of 22:6n-3 added to incubation flasks at three concentrations (0.02, 0.06 and 0.10 mg/ml corresponding to 25, 75 and 125 µl ethanol solution, respectively). Each flasks received the same amount of ethanol. Statistical replicates were obtained by using a mixed inoculum of three wethers in three different runs on separate days. After incubation for 24 h, composition of the gas phase (Hassim et al., Reference Hassim, Lourenco, Goel, Vlaeminck, Goh and Fievez2010) and pH was measured (Hanna Instruments, Temse, Belgium) and culture contents were sampled for analysis of VFA and LCFA.

Experiment 2 was a 7×3 factorial experiment in randomized block design. Cultures containing 0.06 mg/ml of 22:6n-3 (corresponding to 150 µl ethanol solution) were incubated up to 48 h (6, 24 and 48 h) to evaluate the disappearance of 22:6n-3 in the presence of gum arabic, mucin, cellulose, xylan or starch (5 mg/ml). Two additional treatments were included in the design: a treatment containing no gum arabic, mucin, cellulose, xylan or starch (i.e. only the substrate present in the buffer solution was provided) and a treatment without additional substrate nor carbohydrates in the buffer solution (i.e. removal of glucose, cellobiose, xylose and arabinose from the buffer solution). Statistical replicates were obtained by using inocula from three donor animals. Samples were collected after 6, 24 and 48 h of incubation by removing temporarily the flasks from the incubator and taking 10 ml of incubation fluid using a syringe. After pH was measured, culture contents were sampled for analysis of VFA and LCFA.

In experiment 3, a single-factor repeated-measures design was used to compare rate of disappearance of 18:2n-6 and 22:6n-3. Three fermenters (1 l) were used, all containing gum arabic (5 mg/ml), 18:2n-6 (0.1 mg/ml, corresponding to 1000 µl ethanol solution prepared by dissolving 18:2n-6 (Nu-check-Prep) in ethanol (50 mg/ml)) and 22:6n-3 (0.02 mg/ml, corresponding to 500 µl ethanol solution), and were inoculated with diluted rumen fluid (500 ml) and incubated for 24 h in a water bath at 39°C. The concentration of 18:2n-6 and 22:6n-3 were chosen to allow formation of 18:0 and 22:0, when incubated separately. Inocula for each fermenter was obtained from a different sheep (statistical replicate). Culture content was continuously mixed by means of a magnetic stirrer and the headspace continuously flushed with CO₂ to assure anaerobic conditions. Samples were taken after 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 24 h of inoculation for analysis of VFA and LCFA. Samples were pipetted directly from the fermenter (2 ml for VFA and 5 ml for LCFA) through a closable opening in the lid.

Distribution of 22:6n-3 in incubation contents

Two experiments were performed to determine the effect of adding an adsorbant or fermentable substrate on the distribution of 22:6n-3 between the solid and liquid phase of diluted rumen fluid. Experiments were carried out in polycarbonate tubes containing 10 ml of diluted buffered rumen fluid supplemented with 0.06 mg/ml of 22:6n-3. Diluted buffered rumen fluid was prepared as described before. In experiment 4, the effect of four different adsorbents (mucin, gum arabic, bentonite and silicic acid at 5 mg/ml) on the distribution of 22:6n-3 between liquid and solids was determined. In experiment 5, the effect of gum arabic, mucin, cellulose, xylan and starch (5 mg/ml) was evaluated. All treatment combinations were mixed using a multiple vortex. After 30 min, reactions were stopped followed by centrifugation for 10 min at 3400 g at 4°C. Supernatant and pellet were stored at −20°C before LCFA analysis. Preliminary experiments showed that gum arabic and mucin were recovered in the liquid phase and bentonite and silicic acid in the solid phase.

Effect of gum arabic on growth of B. fibrisolvens and B. proteoclasticus

B. fibrisolvens D1 (DSM 3071) and B. proteoclasticus P18 (Rowett Institute of Nutrition and Health) were grown anaerobically for 48 h at 39°C in medium 704 (DSMZ) or RM02 medium (Kenters et al., Reference Kenters, Henderson, Jeyanathan, Kittelmann and Janssen2011) in 12.5×1.5 cm culture tubes closed with screw caps fitted with butyl rubber septa (Bellco Biotechnology, Vineland, NJ, USA). The medium 704 was modified by omitting the VFA mixture, haemin and glycerol, and by increasing the proportion of rumen fluid to 20% (v/v). Rumen fluid was collected from three mature wethers. Donor animals were the same as those used for in vitro incubations with ruminal fluid and were fed the same diet. Rumen fluid was filtered through a sieve with a pore size of 1 mm and then fine particles were removed from the filtrate by centrifugation at 27 000 g for 20 min at 4°C. The supernatant was sterilized by autoclaving for 20 min at 121°C and stored frozen at −20°C. The stored rumen fluid was thawed before use and any new precipitates formed were removed by centrifugation at 27 000 g for 15 min at 4°C. Incubations were carried out anaerobically with and without gum arabic (5 mg/ml). The 22:6n-3 solution was prepared by emulsifying 30 mg of 22:6n-3 with polyoxyethylene sorbitan and diluting the mixture in 50 ml of distilled water (0.6 mg/ml) and was added to the medium before autoclaving. Several concentrations of 22:6n-3 were evaluated (0, 10, 20, 40, 80 mg/l corresponding to 0, 167, 333, 667 and 1333 µl of the 22:6n-3 solution, respectively). Amounts of polyoxyethylene sorbitan (0.1 mg/ml) were kept the same in all tubes by adding 1333, 1166, 1000, 666 and 0 µl of the polyoxyethylene sorbitan solution without 22:6n-3. At this concentration, polyoxyethylene sorbitan did not negate the growth inhibitory effects of 18:2n-6 on growth of Propionibacterium freudenreichii (Rainio et al., Reference Rainio, Vahvaselkä, Suomalainen and Laakso2001). Inoculum volumes were 5% (v/v) of a fresh culture that was grown in the same media for 12 h. Growth was measured from the increase in optical density at 600 nm (Ultraspec10, Amersham Biosciences Corp., Piscataway, NJ, USA) for duplicate cultures incubated on the same day. Each set of culture experiments was repeated on three different days.

Analysis

VFAs in 2 ml of culture contents were prepared according to Castro-Montoya et al. (Reference Castro-Montoya, De Campeneere, Van Ranst and Fievez2012). Samples were analysed using a gas chromatograph (HP 7890A, Agilent Technologies, Diegem, Belgium) equipped with a flame ionization detector (Agilent Technologies) and a Supelco Nukol capillary column (30 m×0.25 mm×0.25 µm, Sigma-Aldrich, Diegem, Belgium). The temperature program was initially set at 120°C for 0.2 min; increased at 10°C/min up to 180°C and held at 180°C for 3 min; injector temperature: 250°C; detector temperature 255°C. For this temperature program, 0.3 µl was injected with a split/splitless ratio of 25 : 1 using H₂ as carrier gas at 0.8 ml/min. VFA peaks were identified based on their retention times, compared with external standards (acetic acid, propionic acid, isobutyric acid, butyric acid, iso-valeric acid and valeric acid; Sigma-Aldrich). Samples for LCFA (5 ml) were freeze dried and LCFA were extracted and methylated as described by Vlaeminck et al. (Reference Vlaeminck, Braeckman and Fievez2014). Analysis for FA methyl esters was carried out by gas chromatography (GC) (HP 7890A, Agilent Technologies) using a SP-2560 column (75 m×0.18 mm, i.d. ×0.14 µm thickness, Supelco Analytical, Bellefonte, PA, USA) and a flame ionization detector. The temperature program was initially set at 70°C for 2 min, increased at 15°C/min to 150°C, then increased at 1°C/min up to 165°C, held at 165°C for 12 min, increased at 2°C/min to 170°C, held at 170°C for 5 min, increased at 5°C/min to 210°C, held at 210°C for 20 min, increased at 5°C/min to 220°C and held at 220°C for 15 min. Inlet and detector temperatures were 250°C and 255°C, respectively. The split ratio was 50 : 1. Hydrogen was used as the carrier gas at a flow rate of 1 ml/min. Peaks were identified based on retention time comparisons with a mixture of FA methyl esters (FAME) standards (GLC463, Nu-Check-Prep and CLA cis-9, trans-11-18:2 and trans-10, cis-12-18:2, Larodan Fine Chemicals AB, Malmö, Sweden). Moreover, aliquots of methylated samples were pooled and subsequently fractionated by Ag⁺-SPE as described by Kramer et al. (Reference Kramer, Hernandez, Cruz-Hernandez, Kraft and Dugan2008). FAs commercially not available were identified by order of elution based on Shingfield et al. (Reference Shingfield, Bonnet and Scollan2006) and Kramer et al. (Reference Kramer, Hernandez, Cruz-Hernandez, Kraft and Dugan2008). The GC run time was chosen as to allow conjugated isomers of 22:6 to be detected, which would be expected to elute after 22:6n-3. As no standards of conjugated isomers of 22:6 are commercially available, conjugated isomers of 22:6 were prepared by alkaline treatment according to Association of Official Analytical Chemists (Reference Helrich1990) with modifications. A mixture of potassium hydroxide in ethylene glycol 21% (w/w) was prepared and degassed for 5 min with nitrogen. A total of 10 mg of 22:6n-3 was added to a clean 10 ml test tube containing 4 ml of 21% KOH solution. The mixture was degassed with nitrogen gas, capped and allowed to stand for 5 min at 160°C. After cooling to room temperature, the reaction mixture was acidified to below pH 2 with HCl. The conjugated products were extracted twice with 3 ml of hexane. Spectrophotometric readings (Shimadzu UV-1700; Shimadzu Europe, Duisburg, Germany) confirmed the formation of conjugated dienes (at 233 nm), trienes (268 nm), tetraenes (315 nm), pentaenes (345 nm) and hexaenes (375 nm) (Pitt and Morton, Reference Pitt and Morton1957). Initially, the GC run time was extended with the oven temperature at 220°C for 120 min instead of 15 min. As no compounds were observed after the 15 min at 220°C, this temperature program was used in subsequent analysis.

Calculations

The net amount of VFA produced during the 24 h incubation period was calculated by subtracting the amount of VFA in the rumen fluid/buffer mixture present before and after incubation. Biohydrogenation or disappearance of 22:6n-3 was calculated based on the proportional loss of 22:6n-3 during the incubation period as the difference of the amount of 22:6n-3 before and at the end of the incubation divided by the amount before the incubation. Formation of 22:0 was calculated by subtracting the amount of 22:0 in the rumen fluid/buffer before the incubation from the amount at the end of incubations with 22:6n-3.

Rate of disappearance of 18:2n-6 and 22:6n-3 in experiment 3 were estimated according to a first-order exponential model as:

$${\rm Residual}\,{\rm 18\colon \!2}n{\rm {\hbox \-}6}\,\,{\rm or}\,\,{\rm 22\colon6}n{\rm {\hbox \-}3\,{\equals}\,}a{\rm {\plus} }b{\rm {\times} exp}\,{\rm ({\minus}}c{\rm {\times}(time\,{\minus}\, lag))}$$

where a is the amount of FA substrate not hydrogenated, b the amount of FA that can potentially disappear during the incubation with ruminal fluid, c the fractional rate of disappearance of fraction b (1/h) with a lag time, h. Parameters were fitted to the data using the NLIN procedure of the SAS (version 9.2; SAS Institute Inc., Cary, NC, USA).

Statistical analysis

Analytical duplicates were averaged before statistical analysis. All statistical analyses were completed using the MIXED procedure of SAS (version 9.2). Differences among means at P<0.05 were considered significant. Differences among least square means were evaluated by the Tukey–Kramer multiple comparison test.

Data from experiment 1 were analysed with a model that included the fixed effects of treatment, 22:6n-3 concentration and their interaction and random effect of incubation run. Sums of squares were separated into single degree of freedom orthogonal contrasts to evaluate the linear component of adsorbant addition at each 22:6n-3 concentration. For experiments 2 and 3, data were analysed by ANOVA for repeated measures. For experiment 2, the model included the fixed effects of treatment, incubation time and their interaction and random effect of rumen inoculum. The effect of incubation time was evaluated as a repeated measure using the heterogeneous autoregressive covariance structure based on the Akaike’s information criterion. For experiment 3, the model included the fixed effect of incubation time and inoculum source as a random effect. The effect of incubation time was evaluated as a repeated measure using the heterogeneous autoregressive covariance structure based on the Akaike’s information criterion. Data from the experiments evaluating the distribution of 22:6n-3 between pellet and supernatants (experiments 4 and 5) were analysed with a model that included the fixed effect of treatment and the random effect of biological replicate (i.e. sheep inoculum). Data from the pure culture experiment were evaluated with a model that included the fixed effect of treatment (without or with gum arabic) and 22:6n-3 concentration and their interaction. Incubation run was included as a random effect. The analysis was performed separately for B. fibrisolvens and B. proteoclasticus for each growth media.