Introduction
The different positional and geometric isomers of conjugated linoleic acid (CLA) confer different health effects on mammals (Belury, Reference Belury2002). The c9t11 CLA isomer has been shown to be anti-carcinogenic (Ip et al., Reference Ip, Banni, Angioni, Carta, McGinley, Thompson, Barbano and Bauman1999), while the t10c12 CLA isomer has been shown capable of decreasing body fat and increasing lean body mass (Park et al., Reference Park, Storkson, Albright, Liu and Pariza1999). The t10c12 CLA isomer also decreases fat concentration in dairy cows' milk in a dose-dependent fashion (Peterson et al., Reference Peterson, Baumgard and Bauman2002). These effects on mammals have encouraged research efforts to identify methods of increasing the c9t11 CLA isomer and to understand the mechanisms of t10c12 CLA production so that its synthesis would occur only when desired.
Conjugated linoleic acid and trans C18:1 fatty acids (FA) are produced during biohydrogenation of unsaturated FA in the rumen (AbuGhazaleh et al., Reference AbuGhazaleh, Riley, Thies and Jenkins2005; Harfoot & Hazlewood, Reference Harfoot, Hazlewood and Hobson1988) and are subsequently incorporated into milk and meat of ruminant animals. Formation of trans C18:1 and CLA in the rumen are influenced by dietary supplementation with unsaturated vegetable oils (AbuGhazaleh et al., Reference AbuGhazaleh, Schingoethe, Hippen and Kalscheur2003; Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004a and Reference Loor, Ueda, Ferlay, Chilliard and Doreaub) or changes in rumen pH as a result of alterations in dietary forage to concentrate ratios (F:C) (Piperova et al., Reference Piperova, Sampugna, Teter, Kalscheur, Yurawecz, Ku, Morehouse and Erdman2002; Loor et al., Reference Loor, Hoover, Miller-Webster, Herbein and Polan2003; Sackmann et al., Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003).
Previous studies have demonstrated an increase in the vaccenic acid (VA) and c9t11 CLA content of bovine milk fat when fish oil (FO) and linoleic acid oil sources were added to dairy cattle rations containing 50% forage (AbuGhazaleh et al., Reference AbuGhazaleh, Schingoethe, Hippen and Whitlock2000; Whitlock et al., Reference Whitlock, Schingoethe, Hippen, Kalscheur, Baer, Ramaswamy and Kasperson2002). Studies examining the effect of forage or concentrate feeding on the distribution of trans C18:1 and CLA isomers in the rumen (Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004a) or ruminant fat (Sackmann et al., Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003; Aharoni et al., Reference Aharoni, Orlov and Brosh2004; Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2005) have used only vegetable oils or FO. Feeding high concentrate diets (Piperova et al., Reference Piperova, Sampugna, Teter, Kalscheur, Yurawecz, Ku, Morehouse and Erdman2002) or low forage diets supplemented with vegetable oils rich in linoleic acid or FO (Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004a) altered ruminal biohydrogenation resulting in t10 rather than VA being the predominant trans C18:1 intermediate. Alterations in the profile of trans C18:1 intermediates formed in the rumen have direct consequences on the supply of VA available for endogenous conversion in the mammary gland into c9t11 CLA (Bauman et al., Reference Bauman, Baumgard, Corl and Griinari1999). Previously, AbuGhazaleh et al. (Reference AbuGhazaleh, Schingoethe, Hippen and Whitlock2000, Reference AbuGhazaleh, Schingoethe, Hippen and Kalscheur2003) and Whitlock et al. (Reference Whitlock, Schingoethe, Hippen, Kalscheur, Baer, Ramaswamy and Kasperson2002) demonstrated that the greatest concentrations of VA and c9t11 CLA in milk fat and rumen contents can be obtained by adding FO along with linoleic acid fat source to ruminant animals' diet. The aim of this study was to investigate how changing dietary forage levels in diets containing FO and sunflower oil (SFO) affect trans C18:1 and CLA isomers distribution in the rumen.
Material and methods
Experiment protocol
Three 1700 ± 12 ml dual-flow continuous culture fermenters (Stern & Hoover, Reference Stern, Hoover, Ulyatt, Lee and Corson1990) were used in a 3 × 3 Latin square over three periods of 10 days each. Each experimental period consisted of 7 days for adaptation and followed by 3 days for sample collection. Treatments used in this study were as follows: (1) 75:25 F:C (HF); (2) 50:50 F:C (MF); and (3) 25:75 F:C (LF). Menhaden FO (Omega Protein Inc., Hammond, LA) and SFO (purchased from a local store) were added to each diet at 1 and 2 g/100 g dry matter (DM) basis, respectively. Maize, soya-bean meal, limestone, vitamins and minerals made up the concentrate mix (Table 1). The forage source was alfalfa pellets. A total of 140 g of feed (DM basis) was placed in each fermenter daily in three equal portions at 0800, 1500, and 2100 h.
Table 1 Ingredient and chemical composition of treatment diets
† HF = 75:25 F:C, MF = 50:50 F:C, and LF = 25:75 F:C.
Continuous culture
Whole ruminal contents were taken from two ruminally fisulated Holstein cows fed a 50:50 F:C diet. At each collection time, approximately 4.5 kg of ruminal content were taken from the cow 4 h after feeding, strained through two layers of cheesecloth and transported to the laboratory in a sealed container and used within 20 min. Fermenter canisters (15 cm long and 12.5 cm wide) were filled with approximately 1300 ml of rumen fluid and 400 ml of pre-warmed buffer with urea added (Weller & Pilgrim, Reference Weller and Pilgrim1974). Solids and liquid dilution rates were adjusted, twice daily, to values of 3 and 10% per h, respectively, by regulation of buffer input and filtrate removal rates. Fermenters were constantly mixed at 120 r.p.m. via a magnetic impeller stirrer unit, purged with N2 gas (80 ml/min) and temperature was maintained at 39°C. The pH was measured daily at 0800, 1500, and 2100 h using a portable pH meter (Accumet* AP85 Portable, Fisher Scientific, Pittsburgh. PA).
Sample collection and analysis
Effluent from each fermenter was collected into 5-l plastic jugs submerged approximately three-quarters into a 4°C water bath. The solid and liquid effluent volumes were cataloged daily before the morning feeding and discarded until the final 3 days of each period. On the last 3 days, the solid and liquid portions were combined, homogenised on a stir plate and a 10% volume subsample was collected and stored at − 5°C. Subsamples from each fermenter were composited for days 8, 9 and 10 resulting in one sample per fermenter per period. Samples were thawed in a 50°C waterbath, transferred into 250 ml plastic bottles (10 cm long and 5 cm wide) and centrifuged (Beckman J2-21) at 15 000 r.p.m. for 15 min, after which the supernatant was removed. This process was repeated until all liquid in the thawed sample was removed. The bottle and fiber pellet was stored at − 80°C for 48 h, freeze dried and then ground to 1 mm using a Willey mill (Arthur Thomas Company, Philadelphia, PA).
Samples of alfalfa pellets and concentrate mixes were collected twice each period (days 5 and 10) and stored at − 20 °C until analysis. Samples were freeze dried for at least 48 h, then ground through a 2-mm screen of a standard Wiley mill (model 3; Arthur H. Thomas Co., Philadelphia, PA) and composited by period. Composites were analysed for crude protein (CP), ether extract, and ash according to Association of Official Analytical Chemists methods (1997). Samples were reground (Brinkman ultracentrifuge mill) through a 1-mm screen and analysed for neutral-detergent fibre (NDF, procedure B of Van Soest et al. (Reference Van Soest, Robertson and Lewis1991)) and acid detergent fibre (ADF, Robertson and Van Soest, Reference Robertson, Van Soest, James and Theander1981) using an ANKOM fiber analyser and filter bag technique (ANKOM Technology Corp., Fairport, NY).
Feed and effluent samples were methylated using NaOCH3 and HCl two steps procedure as outlined by Kramer et al. (Reference Kramer, Fellner, Dugan, Sauer, Mosoba and Yurawecz1997) and analysed in triplicate for FA on a Shimadzu GC-2010 gas chromatograph (Shimadzu Scientific Instruments, Inc., Columbia, Maryland, USA) equipped with a flame ionisation detector and a Supelco 100-m SP-2560 fused silica capillary column (0.25 mm i.d. × 0.2 μm film thickness; Supelco Inc., Bellefonte, PA). The helium carrier gas was maintained at a linear velocity of 23 cm/s. The oven temperature was programmed at 170°C for 50 min, then increased at 5°C/min to 249°C and held for 10 min. The injector and detector temperatures were set at 255°C. One milligram of nonadecanoic acid (C19:0; 1 mg/ml benzene) was added to all samples before methylation as an internal standard. Peaks were identified by comparing the retention times with those of corresponding standards (Nu-Chek-Prep., Elysian, MN; Supelco, Bellefonte, PA; and Larodan Fine Chemicals, Malmo, Sweden). The trans C18:1 isomers that were not available commercially (trans-6/8, trans-10, trans-12, trans-13/14) were identified according to the elution sequence reported by Loor et al. (Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004b). Conjugated linoleic acid isomers were identified according to the elution sequence reported by Roach et al. (Reference Roach, Mossoba, Yurawecz and Kramer2002) using Nu-Check-Prep #UC59mx (Elysian, MN; Supelco). Retention times for c9t11, t10c12, and t9t11 were confirmed with pure standards (Matreya LLC, Pleasant Gap, PA).
A summary of dietary ingredients and chemical composition of the experimental diets is shown in Table 1. Dietary CP was similar across diets averaging 17.7 g/100 g DM. Dietary ADF and NDF decreased as dietary forage level decreased (Table 1). Daily input of total FA (g/day) was also affected by dietary treatments, increasing as dietary forage decreased. Supplies of oleic and linoleic acids increased while linolenic acid decreased as dietary forage level decreased (Table 2). All treatments had an equal supply of C18:0, C20:5, and C22:6.
Table 2 Fatty acids intake (g/day)
† HF = 75:25 F:C, MF = 50:50 F:C, and LF = 25:75 F:C.
Statistical analysis
Data were analysed using the general linear model procedure of (Statistical Analysis Systems Institue, Inc., Cary, NC) According to the following statistical model:
where Y ijk = the observation; μ = overall mean; F i = fermenter effect; D j = diet effect; P k = period effect and e ij = residual error associated with Y ijk.
The linear and quadratic effects of treatments were analysed by orthogonal contrasts. Least-square means are reported throughout and significance threshold was set at P ≤ 0.05 and the trend at P ≤ 0.10.
Results
As expected, fermenter pH was affected by dietary forage level (Figure 1), averaging 6.55, 6.15, and 5.65 for diet HF, MF and LF, respectively.
Figure 1 pH in continuous culture fermenters for HF = 75:25 F:C; MF = 50:50 F:C; and 25:75 F:C. Standard errors of the means were 0.06, 0.07, and 0.06, respectively.
Effect of diet on effluent FA concentrations (mg/g of DM) is presented in Table 3. The concentration of trans C18:1 increased (P < 0.03) in a linear manner as dietary forage level decreased. Compared with HF, the concentration of trans C18:1 increased by 55 and 110% as dietary forage level decreased to 50 and 25%, respectively. Dietary treatments also affected the distributions of trans C18:1 isomers (Table 3). The concentration of VA decreased whereas the concentration of t10 C18:1 increased (P < 0.01) in a linear manner as dietary forage level decreased. Vaccenic acid and t10 C18:1 were the predominant trans C18:1 isomers in the HF and LF diets, respectively, accounting for 85 and 93% of total trans C18:1 isomers, respectively. The concentrations of trans-6/8, trans-9, and trans-12 was not affected (P>0.15) by treatment diets (Table 3). Trans-13/14 was detected only in the LF fermenter and was the second major trans C18:1 isomer.
Table 3 Effect of treatment diets on the outflow fatty acids (mg/g of dry matter)
† HF = 75:25 F:C, MF = 50:50 F:C, and LF = 25:75 F:C.
‡ Linear and quadratic effects.
§ CLA = conjugated linoleic acid.
Dietary treatments also affected effluent CLA concentrations and isomers distribution (Table 3). The concentration of CLA decreased linearly (P = 0.03) as dietary forage level decreased. The concentrations of c9t11 CLA decreased (P < 0.02), whereas the concentration of t10c12 CLA tended to increase (P < 0.09) in a linear manner as dietary forage decreased. Cis9 t11 and tt CLA were the major isomers in the HF diet accounting for 70 and 20% of total CLA, respectively, whereas t10c12 and t11t13 CLA were the predominate isomers in the LF diet accounting for 43 and 30% of total CLA (Table 3).
The effect of diet on C18 unsaturated FA biohydrogenation is presented in Table 4. As dietary forage decreased, biohydrogenation of linoleic and linolenic acids decreased linearly (P < 0.03). Biohydrogenation of oleic acid also tended to decrease linearly (P = 0.06) with decreasing forage proportion in the diet.
Table 4 Effect of treatment diets on the biohydrogenation (%) of unsaturated fatty acid
† HF = 75:25 F:C, MF = 50:50 F:C, and LF = 25:75 F:C.
‡ Linear and quadratic effects.
Discussion
Trans C18:1 in effluents accounted for 21, 30, and 33% of total FA for diet HF, MF, and LF, respectively (Table 3). The increase in concentration of trans C18:1 in LF (36.1 mg/g of effluent) compared with HF (17.20 mg/g) indicated that low forage diets promote trans C18:1 accumulation. The linear increase in trans C18:1 concentration as dietary forage level decreased may have resulted, in part from the higher unsaturated C18 FA input for MF and LF diets compared with HF diet (Table 2). Daily input for unsaturated C18 FA was higher by 8 and 25% for MF and LF diets compared with HF diet. Incomplete biohydrogenation of unsaturated C18 FA in the rumen results in trans C18:1 accumulation (Harfoot & Hazlewood, Reference Harfoot, Hazlewood and Hobson1988). Additionally, high trans C18:1 accumulation observed in the LF diet may have resulted from inhibiting the reductase enzyme in ruminal micro-organisms responsible for the terminal hydrogenation of trans C18:1 to C18:0. The low concentration of C18:0 observed in LF effluent supports the suggestion that inhibition occurred at the final reductase step. The linear increase in trans C18:1 concentration indicates that low ruminal pH favours trans C18:1 accumulation. Low ruminal pH, caused by a high-concentrate diet, increased the accumulation of trans C18:1 in other studies (Karlscheur et al., Reference Kalscheur, Teter, Piperova and Erdman1997; Piperova et al., Reference Piperova, Sampugna, Teter, Kalscheur, Yurawecz, Ku, Morehouse and Erdman2002; Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004b). Although this study and others (Karlscheur et al., Reference Kalscheur, Teter, Piperova and Erdman1997; Piperova et al., Reference Piperova, Sampugna, Teter, Kalscheur, Yurawecz, Ku, Morehouse and Erdman2002; Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004b) clearly show that low rumen pH promotes trans C18:1 accumulation, the mechanism is still unknown. The accumulation of trans C18:1 under low rumen pH conditions may be caused by altering the rumen ecosystem and/or inhibiting the reductase activity of ruminal microorganisms, causing the accumulation of trans C18:1. Low rumen pH has been shown to have a negative effect on microbial growth (Russell & Dombrowski, Reference Russell and Dombrowski1980; Martin et al., Reference Martin, Fonty, Michalet-Doreau and Martin2002), particularly on the growth of cellulolytic bacteria, the main rumen biohydrogenating bacteria, (Harfoot & Hazlewood, Reference Harfoot, Hazlewood and Hobson1988).
Six trans C18:1 isomers were identified in effluents (Table 3). As dietary forage level decreased, VA concentration decreased, but t10 C18:1 concentration increased linearly. This shift indicates that incremental increases in concentrate level have the potential to enhance ruminal production of t10 C18:1. To our knowledge, no in vivo or in vitro experiment has evaluated the ruminal distribution of trans C18:1 isomers, primarily VA and t10 C18:1, in response to a graded decrease of dietary forage level at a constant level of supplemental FO and SFO. When cows were switched from a control diet (60:40 F:C) to a high-concentrate diet (25:75 F:C), the proportion of t10 C18:1 increased, concomitant with decreases in VA (Piperova et al., Reference Piperova, Sampugna, Teter, Kalscheur, Yurawecz, Ku, Morehouse and Erdman2002). Additionally, t10 C18:1 replaced VA as the predominant trans C18:1 isomer in the rumen when high concentrate-low fibre diets were fed to cows (Griinari et al., Reference Griinari, Dwyer, McGuire, Bauman, Palmquist and Nurmela1998) and steers (Sackmann et al., Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003). Under such conditions, Bauman et al. (Reference Bauman, Baumgard, Corl and Griinari1999) proposed a putative pathway for the production of t10 C18:1 where the t10c12 CLA-producing bacteria become predominant in the rumen resulting in formation of t10c12 CLA as the first intermediate during linoleic acid biohydrogenation. Hydrogenation of the c12 bond would then result in formation of t10, analogous to the production of VA from c9t11 CLA. High concentrate diets have been shown to promote the growth of Megasphera elsdenii YJ-4 which can convert linoleic acid to t10c12 CLA (Kim et al., Reference Kim, Liu, Rychlik and Russell2002). Furthermore, feeding high grain diets to steers has been shown to stimulate the growth of the YE34 strain of Megasphera elsdenii and cause a rapid decline in Butyrivibrio fibrisolvens YE44 (Klieve et al., Reference Klieve, Hennessy, Quwerkerk, Forster, Mackie and Attwood2003) and its is well established that VA is an intermediate of linoleic acid metabolism by isolates of Butyrivibrio fibrisolvens (Harfoot & Hazlewood, Reference Harfoot, Hazlewood and Hobson1988).
The t10 C18:1 could also arise via the isomerisation of oleic acid (AbuGhazaleh et al., Reference AbuGhazaleh, Riley, Thies and Jenkins2005). The fact that t10 accounted for 93% of total trans C18:1 isomers with the LF diet compared with the 37% and 60% reported by Griinari et al. (Reference Griinari, Dwyer, McGuire, Bauman, Palmquist and Nurmela1998) and Sackmann et al. (Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003), when high-concentrate diets supplemented with linoleic acid oil source were fed, supports our previous finding that docosahexaenoic acid (C22:6; DHA) in FO blocks the final step in the biohydrogenation of unsaturated C18 FA causing trans C18:1 accumulation (AbuGhazaleh and Jenkins, Reference AbuGhazaleh, Schingoethe, Hippen and Kalscheur2004). What seems evident from this experiment and others (Griinari et al., Reference Griinari, Dwyer, McGuire, Bauman, Palmquist and Nurmela1998; Sackmann et al., Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003) is that a supply of high linoleic acid oil along with FO under low pH condition has the greatest potential to enhance t10 C18:1 production in the rumen. These conditions simultaneously eliminated or reduced concentration of t6/8, t9, and t12 C18:1 in effluent, but increased concentration of t13/14 C18:1. Such changes in effluent trans C18:1 isomers profile may suggest a possible alteration in the microbial ecosystem and/or enzyme activities as a result of altering the proportion of forage in the diet. The total absence of t10 and VA in effluents of HF and LF, respectively, should be taken with precaution since the separation of these two isomers can be difficult when their concentrations are very different. Indeed, a small VA shoulder peak was observed in some LF t10 peaks, however, the ratio of VA to t10 was consistently less than 1:30.
The effect of dietary treatments on CLA is presented in Table 3. Total concentrations of CLA were highest in the HF, intermediate in MF, and least in LF indicating that CLA formation is favoured by high ruminal pH. Troegeler-Meynadier et al. (Reference Troegeler-Meynadier, Nicot, Bayourthe, Moncoulon and Enjalbert2003) also reported more CLA accumulation in rumen cultures incubated at high than low pH. The predominate CLA isomer detected in the HF and LF were c9t11 CLA and t10c12 CLA, respectively. In ewes, duodenal flow of t10c12 CLA increased and c9t11 CLA decreased in response to graded increments of dietary concentrate with a constant level (7.4 g/100 g diet DM) of supplemental FA from soya-bean oil (Kucuk and Hess, Reference Kucuk, Hess and Rule2004). Our results support the findings of others (Beaulieu et al., Reference Beaulieu, Drackley and Merchen2002; Duckett et al., Reference Duckett, Andrae and Owens2002; Shingfield et al., Reference Shingfield, Reynold, Lupoli, Toivonen, Yurawecz, Delmonte, Griinari, Grandison and Beever2005) that high concentrate diets support t10c12 CLA formation in the rumen. Replacing incremental portions of red clover with maize grain in dual-flow continuous cultures resulted in a linear increase in the output of t10c12 CLA into effluent (Latham et al., Reference Latham, Storry and Sharpe1972). Shingfield et al., (Reference Shingfield, Reynold, Lupoli, Toivonen, Yurawecz, Delmonte, Griinari, Grandison and Beever2005) reported higher c9t11 CLA concentration in milk when cows were fed a high forage diet (65:35) than low forage diet (35:65) with FO and SFO as lipid supplements. These results indicates that incremental grain or concentrate have the potential to enhance ruminal production of t10c12 CLA. Trans10c12 CLA is a potent inhibitor of fat synthesis (Baumgard et al., Reference Baumgard, Matitashvili, Corl, Dwyer and Bauman2002). What seems evident from this study and others (Beaulieu et al., Reference Beaulieu, Drackley and Merchen2002; Duckett et al., Reference Duckett, Andrae and Owens2002; Kucuk and Hess, Reference Kucuk, Hess and Rule2004) is that a high linoleic acid oil source along with low dietary F:C ratio has the greatest potential to increase t10c12 CLA. The t11t13 CLA was the second predominant isomer in the LF. A similar increase in t11t13 CLA concentration was also reported by Loor et al., (Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004b) when linseed oil was added to a high-concentrate diet. The formation of these different CLA isomers in ruminal cultures provides comparative evidence for the existence of alternative pathways for the biohydrogenation of C18 polyunsaturated FA other than those established by Harfoot & Hazlewood (Reference Harfoot, Hazlewood and Hobson1988).
The biohydrogenation C18 unsaturated FA were affected by treatment diets (Table 4). The biohydrogenation of oleic, linoleic and linolenic acids was greater in HF and MF than LF diet. Wang et al. (Reference Wang, Song, Son and Chang2002) and Troegeler-Meynadier et al. (Reference Troegeler-Meynadier, Nicot, Bayourthe, Moncoulon and Enjalbert2003) also reported decreased biohydrogenation of linoleic and linolenic acids at low pH in vitro. Under in vivo conditions, Sackmann et al. (Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003) and Kucuk and Hess (Reference Kucuk, Hess and Rule2004) have shown that higher dietary forage levels increase biohydrogenation of dietary C18 unsaturated FA. The decreased biohydrogenation of linoleic and linolenic acids with the LF diet suggested a lower biohydrogenation activity by cultures microbes. The low biohydrogenation values for oleic acid in this study compared with others (Duckett et al., Reference Duckett, Andrae and Owens2002; Sackmann et al., Reference Sackmann, Duckett, Gillis, Realimi, Parks and Eggelston2003Kucuk and Hess, Reference Kucuk, Hess and Rule2004) may have resulted from the DHA effect on the reduction step in the biohydrogenation (AbuGhazaleh and Jenkins, Reference AbuGhazaleh and Jenkins2004).
Conclusion
Dietary forage level affected trans C18:1 and CLA concentrations in effluent with more trans C18:1 and less CLA accumulation seen in LF diet. Decreasing dietary forage levels resulted in t10 C18:1 and t10c12 CLA replacing VA and c9t11 CLA, respectively as predominate trans C18:1 and CLA isomers in the rumen. Our results showed that a supply of high linoleic acid oil along with FO under low pH condition has the greatest potential to enhance t10 C18:1 and t10c12 CLA production in the rumen.
