Introduction
Because of ruminal biohydrogenation (BH) of fatty-acids (FA), milk has a low concentration of polyunsaturated FA (PUFA), particularly omega-3 FA, which are thought to have beneficial effects on human health (Simopoulos, Reference Simopoulos2002). Milk can contain FA originating in ruminal PUFA BH, which can also affect human health. The main isomer of conjugated linoleic acid (CLA), cis-9, trans-11 18:2, is the first intermediate of cis-9, cis-12 18:2 BH, and has been shown to inhibit carcinogenesis (Parodi, Reference Parodi1999). In the rumen, CLA is reduced to trans-11 18:1, further hydrogenated to 18:0 (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988). The first step of cis-9, cis-12, cis-15 18:3 ruminal BH is an isomerisation of the cis-12 double bond to a trans-11 double bond, resulting in an isomer of conjugated linolenic acid (CLnA), cis-9, trans-11, cis-15 18:3 (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988); chemically prepared isomers of this natural CLnA exhibit stronger cytotoxic effects than CLA on human tumour cells (Igarashi and Miyazawa, Reference Igarashi and Miyazawa2000). In the rumen, CLnA is reduced to trans-11, cis-15 18:2, which can be reduced to either trans-11 C18:1, cis-15 18:1 or trans-15 18:1; the latter two are not further hydrogenated in the rumen, but trans-11 18:1 can be reduced to 18:0 (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988). Besides these well-known pathways, which mainly result in trans-11 isomers, rumen BH can result in many other cis or trans positional isomers of 18:2 or 18:1 FA, specially trans-13 and/or trans-14 isomers when linseed is added to the diet (Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004).
Among oilseeds, linseed has the highest proportion of cis-9, cis-12, cis-15 18:3, and also contains cis-9, cis-12 18:2, so that the addition of linseed to the diet of dairy cows could improve PUFA and conjugated FA contents in milk fat; higher proportions of cis-9, cis-12, cis-15 18:3 and CLA have been observed in plasma and milk fat after the dietary addition of linseeds or linseed oil (Gonthier et al., Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005; Loor et al., Reference Loor, Ferlay, Ollier, Doreau and Chilliard2005). Higher proportions of other BH intermediates, particularly trans-11, cis-15 18:2 and trans-6–trans-11 18:1, were found in plasma and milk after linseed oil supplementation (Loor et al., Reference Loor, Ferlay, Ollier, Doreau and Chilliard2005). On the contrary, little is known about the concentration of CLnA in milk (Destaillats et al., Reference Destaillats, Trottier, Garro Galvez and Angers2005), and the effects of diet on this FA have not been studied. Extrusion of oilseeds has been shown to increase the proportions of main BH intermediates in milk fat, with canola (Bayourthe et al., Reference Bayourthe, Enjabert and Moncoulon2000), soya beans (Chouinard et al., Reference Chouinard, Corneau, Bulter, Chilliard, Drackley and Bauman2001) and linseed (Gonthier et al., Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005), and increased (Chapoutot and Sauvant, Reference Chapoutot and Sauvant1997) or decreased (Chouinard et al., Reference Chouinard, Corneau, Bulter, Chilliard, Drackley and Bauman2001; Gonthier et al., Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005) milk PUFA have been reported. The effects of linseed extrusion on CLnA or minor trans-18:1 intermediates have not been studied.
The objectives of this study were to investigate the effects of raw or extruded linseed on plasma and milk FA profiles in lactating dairy cows, focusing on PUFA and BH intermediates.
Material and methods
Experimental design and diets
All procedures for this study complied with the Guide for the Care and Use of Agriculture Animals in Agricultural Research and Teaching (Federation of Animal Sciences Societies, 1999). Thirty lactating Holstein cows, 116 ± 64 days in milk, were assigned to three groups of 10 cows according to parity, milk production, milk fat and protein contents, and days in milk. One cow was excluded during the second period because of an acute mastitis. Groups of cows were housed in three adjacent pens. In a 3 × 3 Latin-square design, the groups of cows were assigned to one of the following three diets: (1) control diet (C), (2) diet with raw linseed (RL) and (3) diet with extruded linseed (EL). Diets were based on maize silage, and contained similar amounts of crude protein (CP) and fibre (Table 1). Supplemented linseed was a blend of 70% linseed and 30% wheat bran to avoid oil losses during extrusion, and the word linseed will designate this blend throughout the text. Linseed was crushed through a 3-mm screen. For EL, the mixture was preconditioned at 50°C before extrusion at 120°C. Cows were fed by pen, and had free access to the diet all day long. Diet ingredients were mixed and distributed once per day in the morning, at the rate of 23 kg dry matter (DM) per cow, corresponding to the actual ad libitum DM intake measured before experiment. Water was available ad libitum. Each period lasted 21 days: the cows were adapted to diets for 19 days and sampled during the last 2 days.
Table 1 Ingredients and chemical composition of diets
†Details of ingredients are as follows. Concentrate 1: composed of protected soya-bean meal, protected sunflower meal, maize germ meal, urea, palm kernel meal and cane vinasse; 44.7% CP (dry-matter basis). Fat supplement: calcium salts of palm and soya-bean oil. Concentrate 2: composed of maize, barley, wheat, sorghum, maize cob, wheat bran and cane molasses; 24.6% CP (dry-matter basis). Raw and extruded linseed: blend linseed/wheat bran (70:30). Vitamin–mineral mix: contained on a per kg basis: 70 g P, 210 g Ca, 60 g Mg, 10 g Na, 300 000 IU vitamin A, 60 000 IU vitamin D3, 700 mg vitamin E, 4500 mg Zn, 3500 mg Mg, 800 mg Cu, 70 mg I, 20 mg Co, 15 mg Se.
Samples and chemical analysis
Samples of diet ingredients were taken and frozen until analysis for DM and CP (Association of Official Analytical Chemists, 1996), NDF and ADF according to the method of Van Soest et al. (Reference Van Soest, Robertson and Lewis1991) and FA content and profile as explained here-under. Blood samples were taken at 1400 h on day 20 and at 0800 h on day 21 from the coccygeal vessels using heparinised tubes, and immediately centrifuged. Plasma was kept at –20°C until analysis. At the evening milking of day 20 and the morning milking of day 21, a 50 ml milk sample was immediately frozen and kept for FA analysis, and a 10 ml sample was used for determination of fat and protein content. The samples from the two milkings were mixed before analysis. Milk fat and true protein contents were determined by IR analysis (Milkoscan 605, Foss Electric, F-75001 Paris).
Milk samples for FA analysis were freeze-dried (Vitris Freezemobile 25; Vitris Gardiner, NY). Plasma total lipids were extracted as described by Folch et al. (Reference Folch, Lees and Stanley1957) and 19:0 was used as an internal standard. FAs in plasma lipids extracts and in unextracted diet ingredients and milk samples were methylated using sodium methoxide followed by boron trifluoride as described by Park and Goins (Reference Park and Goins1994). This method, which successively uses basic and acid transmethylations, allows methylation of all lipid classes, including non-esterified fatty acids, and does not alter the stereochemistry of CLA double-bonds (Duckett et al., Reference Duckett, Andrae and Owens2002). One part of FA methyl esters from each sample was fractionated by argentation thin-layer chromatography (Ag-TLC) (plates 20 × 20 cm, Silica gel 60, Merk KGaA, Germany) to separate the trans-18:1 FA, as described by LeDoux et al. (Reference LeDoux, Rouzeau, Bas and Sauvant2002). Total and trans-18:1 FA profiles were analysed by GLC (Agilent 6890N, equipped with a model 7683 auto injector, Network GC System, Palo Alto, CA, USA). The column was a fused silica capillary (CPSil88, 100 m × 0.25 mm ID, 0.2 μm film thickness, Chrompack-Varian, Middleburg, The Netherlands).
For plasma analysis, flame ionisation detector temperature was maintained at 260°C and the injector at 255°C, and a splitless injection with an automatic injector was used. Helium was the carrier gas with a constant pressure (24.6 p.s.i.). The samples were injected in 0.5 μl of hexane. Initial temperature of the oven was 70°C, held for 1 min, increased by 5°C/min to 100°C, held at 100°C for 2 min, increased by 10°C/min to 175°C, held at 175°C for 40 min, increased by 5°C/min to a final temperature of 225°C and maintained at 225°C for 15 min, as described by Loor et al. (Reference Loor, Bandara and Herbein2002). Trans-10 18:1 and trans-11 18:1 were not completely separated with this method, and were considered together and designated as trans-10 + 11 18:1. Plasma analysis was performed before the publication of Loor et al. (Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004), indicating coelution of cis-15 18:1 and 19:0. Because we utilised 19:0 as an internal standard, we used a different method for determination of milk FA, with the same initial temperatures of the detector and injector, and with a split ratio of 1:50. Hydrogen was the carrier gas with a constant pressure (23.2 p.s.i.). The samples were injected with an automatic injector, in 0.1 μl of hexane. Initial temperature of the oven was 60°C, held for 1 min, increased by 20°C/min to 150°C, held at 150°C for 10 min, increased by 2°C/min to 175°C, held at 175°C for 20 min, increased by 10°C/min to a final temperature of 225°C and maintained at 225°C for 10 min. In addition to the separation of cis-15 18:1 from 19:0, this method allows a better separation of trans-10 and trans-11 18:1 when their proportions are very different.
Identification and quantification of peaks were made by comparison with commercial standards when available (Sigma, St. Louis, USA). Identification of trans-4 to trans-8 18:1, trans-12 to trans-16 18:1, and trans-11, cis-15 18:2 was made by comparison with published chromatograms (Precht and Molkentin, Reference Precht and Molkentin1999). Trans-10 18:1 and trans-11 18:1, measured together in plasma or their sum in milk samples, were used as an internal standard to quantify trans-18:1 FA determined by Ag-TLC. After quantification of the amounts of trans-18:1 isomers, cis-9 18:1 was corrected by subtracting the overlapping trans isomers (trans-15 18:1 in plasma samples and trans-13 and trans-14 18:1 in milk samples) as suggested by Precht and Molkentin (Reference Precht and Molkentin1999).
CLnA was identified by gas chromatography–mass spectroscopy (GC–MS). The FA methyl esters were converted into 2-alkenyl-4,4-dimethyloxazoline (DMOX) derivatives according to Yurawecz et al. (Reference Yurawecz, Hood, Roach, Mossoba, Daniels, Ku, Pariza and Chin1994). Briefly, 100 μl of 2-amino-2-methyl-1-propanol were added to the fatty-acid methyl ester (FAME). The reaction mixture was maintained at 170°C for 8 h under nitrogen atmosphere. The analysis of 4,4-DMOX derivatives was achieved using a GC-2010 coupled with a QP-2010 mass spectrometer (Shimadzu, Champs-sur-Marne, France). The column described above was utilised. Helium was used as the carrier gas at a constant velocity of 24.3 cm/s. The oven was programmed from 60 to 210°C at 20°C/min and the final temperature was maintained for 50 min. The injector in splitless mode was maintained at 250°C. The electron impact mass spectra were recorded at 70 eV between 100 and 450 amu. The identified CLnA was 9,11,15 18:3, the most probable configuration being cis-9, trans-11, cis-15 18:3 (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988; Destaillats et al., Reference Destaillats, Trottier, Garro Galvez and Angers2005).
Low area peaks were rejected from quantification, the rejection threshold corresponding to around 0.02% of injected FA. Trans-10, cis-12 18:2 was under area rejection threshold in nearly all plasma and milk samples, so that the only measured CLA isomer was cis-9, trans-11 18:2, and the term CLA will refer to this isomer in this paper. Figure 1 presents chromatograms of a whole milk sample and from an Ag-TLC extract of milk.
Figure 1 GLC chromatograms of fatty-acid methyl esters (FAMEs) obtained from milk fat of a cow receiving extruded linseed: total FAMEs (upper graph) and trans-18:1 fraction obtained by argentation thin-layer chromatography (on lower graph).
Calculations and statistical analysis
Desaturase ratios between cis-9 unsaturated FA and their precursors, which can serve as a proxy for Δ-9 desaturase activity (Bauman et al., Reference Bauman, Corl, Baumgard and Griinari2001), were calculated as described by Kelsey et al. (Reference Kelsey, Corl, Collier and Bauman2003).
Milk production, protein and fat content, proportions of each FA and desaturase ratios were compared with SYSTAT (Statistical Packages for the Social Sciences, 1998), using the model
Differences among treatments were assessed using contrasts between control and mean values of linseed diets, and between raw and extruded linseeds. Significance was declared at P < 0.05, and tendencies at 0.05 < P < 0.10.
Results
Milk yield and composition
Milk production was lower for cows fed linseed diets than for those fed diet C (Table 2). Linseed had no significant (P = 0.109) effect on milk fat percentage but decreased protein percentage. Yields of milk protein and milk fat were lower for cows fed linseed. Compared with diet RL, diet EL tended to decrease yields of milk, milk fat and milk protein, without affecting milk fat and protein percentages.
Table 2 Least-square means of milk yield and composition of lactating cows fed control diet or diets supplemented with raw or extruded linseed
Plasma fatty acids profile
Plasma FA profiles are presented in Tables 3 and 4. The proportions of 12:0, 14:0, 15:0, 16:0, cis-9 16:1, 17:0, cis-9 18:1, cis-9, cis-12 18:2, and cis-11 18:1 were decreased, but the proportions of 18:0, cis-9, cis-12, cis-15 18:3, trans-5 to trans-16 18:1, CLA, and CLnA were increased in the plasma of lactating cows fed linseed diets compared with those fed C diet.
Table 3 Least-square means of proportions of fatty acids (other than biohydrogenation intermediates) in the plasma of lactating cows fed control diet or diets supplemented with raw or extruded linseed
Table 4 Least-square means of proportions of biohydrogenation intermediates in the plasma of lactating cows fed control diet or diets supplemented with raw or extruded linseed
Feeding diet EL decreased the proportions of 12:0, 14:0, 16:0, 17:0 and 18:0, and increased the proportions of cis-9, cis-12, cis-15 18:3, trans-10 + 11 to trans-16 18:1 FA, CLA, and CLnA, compared with diet RL. The proportion of trans-11, cis-15 18:2 was not affected by treatments.
Milk fatty acids profile
Milk FA profiles are presented in Tables 5 and 6. Feeding linseed did not affect the milk fat proportions of 4:0 but decreased the proportions of 18:2, cis-11 18:1 and FA from 6:0 to 17:0 except 13:0. The proportions of 18:0, cis-9 18:1, cis-9, cis-12, cis-15 18:3, all individual trans-18:1, cis-15 18:1, CLA and trans-11, cis-15 18:2 were increased with linseed-supplemented diets. CLnA was under the rejection threshold in the milk from cows fed diet C. Linseed diets resulted in a lower desaturase ratio (Table 7) of 18:0, but did not affect desaturase ratios of 14:0, 16:0 and trans-11 18:1.
Table 5 Least-square means of proportions of fatty-acids (other than biohydrogenation intermediates) in the milk of lactating cows fed control diet or diets supplemented with raw or extruded linseed
Table 6 Least-square means of proportions of biohydrogenation intermediates in the milk of lactating cows fed control diet or diets supplemented with raw or extruded linseed
Table 7 Least-square means of desaturase ratios in the milk of lactating cows fed control diet or diets supplemented with raw or extruded linseed
Compared with diet RL, diet EL increased the proportions of trans-6 + 7 + 8 to trans-15 18:1, CLA, trans-11, cis-15 18:2 and CLnA, tended to increase the proportion of cis-9, cis-12, cis-15 18:3, reduced the proportions of all even-chain FA from 6:0 to 16:0, but did not affect the proportions of 18:0, cis-9 18:1 and cis-9, cis-12 18:2. Extrusion did not affect the desaturase ratios of milk fat.
Discussion
Milk production and composition, plasma and milk fatty acids other than biohydrogenation intermediates
Feeding linseed reduced milk production in this experiment. The Literature on the response of milk production to diets supplemented with 10–15% of linseed (DM basis) report either reduction (Kennelly (Reference Kennelly1996) with 10% linseed; Petit et al. (Reference Petit, Ivan and Mir2005)) or little effect (Kennelly (Reference Kennelly1996) with 15% linseed; Gonthier et al. (Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005)). Petit et al. (Reference Petit, Ivan and Mir2005) explained their observed decreased production by a lower DM intake, but this parameter was not measured in our experiment.
Linseed feeding decreased milk protein percentage in our experiment, which is consistent with the results of Kennelly (Reference Kennelly1996) but contrasts with the observations of Gonthier et al. (Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005) on late lactation dairy cows. This lowered milk protein percentage with high fat diets has been related to a decreased mammary blood flow (Cant et al., Reference Cant, DePeters and Baldwin1993).
Both linseed forms reduced proportions of all FA with less than 18 carbons in plasma and milk, except for the proportions of 4:0 and 13:0 in milk. Because of their lower proportion in milk fat and because of the lower milk fat production, the daily amount of short-chain FA was decreased with linseed diets: the total output of 8:0–14:0-saturated even-chain FA was 315, 258 and 211 g/day with diets C, RL and EL, respectively (P < 0.001 for both contrasts C v. linseed and RL v. EL), which represented a 18% and 33% decrease, respectively, and explained the reduction of milk fat yield. With raw linseed, Gonthier et al. (Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005) reported a 38% decrease of 8:0–14:0 daily output, and a 34% decrease was observed due to linseed oil (Loor et al., Reference Loor, Ferlay, Ollier, Doreau and Chilliard2005). Similar effect was observed after soybean (Chouinard et al., Reference Chouinard, Lévesque, Girard and Brisson1997) or sunflower (Schingoethe et al., Reference Schingoethe, Brouk, Lightfield and Baer1996) dietary addition.
Milk FA from 6:0 to 14:0 and part of 16:0 are synthesised by the mammary gland, and increasing amounts of unsaturated fat supplement are known to inhibit the synthesis of these FA (Clapperton and Banks, Reference Clapperton and Banks1985). Among possible reasons for this inhibition, Grummer (Reference Grummer1991) cited a direct inhibition of mammary acetyl-CoA carboxylase, which is mediated by trans-10, cis-12 18:2 (Baumgard et al., Reference Baumgard, Corl, Dwyer, Saebo and Bauman2000), but in our experiment, milk proportion of trans-10, cis-12 18:2 was very low with all diets. Loor et al. (Reference Loor, Ferlay, Ollier, Doreau and Chilliard2005) suggested that milk fat depression could also be explained by other FA, including trans-10 18:1, which exhibited high proportions in milk with low proportions of 6:0–14:0 in our experiment. However, Lock et al. (Reference Lock, Tyburczy, Dwyer, Harvatine, Destaillats, Mouloungi, Candy and Bauman2007) recently demonstrated that this FA does not reduce milk fat synthesis in dairy cows.
Compared with diet C, linseed feeding increased 18:0 and decreased cis-9 18:1 proportions in plasma lipids by 3.5% and 4.5%, respectively. Consequently, the lowered 18:0 desaturase ratio with linseed diets could have been due to this lowered cis-9 18:1/18:0 ratio in plasma rather than to a decreased activity of mammary Δ9-desaturase. Moreover, 14:0, 16:0 and trans-11 18:1 desaturase ratios were not affected by linseed addition.
In spite of a 35% increase in cis-9, cis-12 18:2 dietary concentration, the linseed diets decreased cis-9, cis-12 18:2 proportion in both plasma lipids and milk fat: the cis-9, cis-12 18:2 milk excretion decreased from 30.3 g/day with diet C to 25.8 and 22.4 g/day with diets RL and EL, respectively, and the lower milk yield with linseed diets could only account for a little part of these differences. This demonstrates a lower transfer from diet to milk, due to a higher ruminal disappearance of cis-9, cis-12 18:2, a lower mammary uptake, a competition between cis-9, cis-12 18:2 and cis-9, cis-12, cis-15 18:3 in the mammary gland, or both.
Linseed feeding increased the plasma proportion of cis-9, cis-12, cis-15 18:3 by only three times in our experiment, whereas the dietary amount was nine times higher than in diet C. A similar three-fold increase in plasma lipids has been observed after linseed addition by Gonthier et al. (Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005). This low transfer from diet to plasma is due to the extensive BH of cis-9, cis-12, cis-15 18:3 in the rumen, and suggests that ruminal cis-9, cis-12, cis-15 18:3 BH extent is increased when dietary supply is high. The mammary gland can efficiently uptake blood cis-9, cis-12, cis-15 18:3 because its proportion in milk FA reaches 13.9% after a duodenal daily infusion of 500 g of linseed oil (Petit et al., Reference Petit, Dewhurst, Scollan, Proulx, Khalid, Haresign, Twagiramungu and Mann2002). In our experiment, the transfer rate of cis-9, cis-12, cis-15 18:3 from diet to milk decreased from 7.45% with diet C to 3.10% with linseed diets.
The proportion of cis-9, cis-12, cis-15 18:3 was increased in the plasma, and tended to be increased in the milk with diet EL compared with diet RL, suggesting a lower ruminal BH after extrusion. On the contrary, Gonthier et al. (Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005) reported that the extrusion of linseed decreased milk cis-9, cis-12, cis-15 18:3 proportion. The discrepancy between our results and theirs could be attributed to the difference of the extrusion temperature of linseed, which was 155°C in their experiment compared with 120°C in our experiment. However, Chouinard et al. (Reference Chouinard, Lévesque, Girard and Brisson1997) demonstrated using soya beans that increasing extrusion temperature from 120 to 140°C has only minor effects on milk PUFA.
Plasma and milk biohydrogenation intermediates
Linseed supplement was the only source of CLnA in the milk because there was no detectable CLnA in the milk of cows fed diet C. Destaillats et al. (Reference Destaillats, Trottier, Garro Galvez and Angers2005) found 0.03% of CLnA in the fat of Canadian summer milk. Our rejection threshold was under this value, but lack of CLnA with C diet in our experiment could be related to the low amount of dietary cis-9, cis-12, cis-15 18:3, because dietary cis-9, cis-12, cis-15 18:3 is mainly provided by grass or linseed supplements, which were not ingredients of our diet C.
Proportions of CLnA and CLA were increased in both plasma lipids and milk fat by linseed supplementation. The same effect was observed for trans-11, cis-15 18:2 in milk fat, but not in plasma lipids where a high variability was noticed. Plasma and milk CLnA and trans-11, cis-15 18:2 originate from the ruminal BH of cis-9, cis-12, cis-15 18:3 (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988), so that increased values could be expected with linseed addition. Milk CLnA possibly could also originate from mammary Δ9-desaturation of trans-11, cis-15 18:2, but the ratio of CLnA to trans-11, cis-15 18:2 was higher in plasma lipids than in milk fat, which does not support this hypothesis. However, plasma non-esterified FA and triacylglycerols are the primary FA sources for the mammary gland, so that the interpretation of this difference of ratio can be biased if CLnA and trans-11, cis-15 18:2 have different distributions among plasma lipid classes.
Milk CLA can originate from ruminal CLA, the first intermediate of cis-9, cis-12 18:2 BH, but most milk CLA originates in mammary Δ9-desaturation of trans-11 18:1 (Griinari et al., Reference Griinari, Corl, Lacy, Chouinard, Nurmela and Bauman2000; Lock and Garnsworthy, Reference Lock and Garnsworthy2002). Because trans-11 18:1 is only a minor isomerisation product of cis-9 18:1 in the rumen (Mosley et al., Reference Mosley, Powell, Riley and Jenkins2002), most ruminal trans-11 18:1 originates in cis-9, cis-12 18:2 and cis-9, cis-12, cis-15 18:3 BH (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988), which can explain the higher proportion of this FA after linseed addition. In plasma lipids, linseed feeding decreased cis-9, cis-12 18:2, but increased CLA. This higher ratio of CLA to cis-9, cis-12 18:2 could be due to an inhibition by cis-9, cis-12, cis-15 18:3 of the ruminal reductase, which converts CLA to trans-11 18:2 (Troegeler-Meynadier et al., Reference Troegeler-Meynadier, Nicot, Bayourthe, Moncoulon and Enjalbert2003).
Total trans-18:1 percentage in the milk fat of cows fed diet C in our experiment was similar to the percentage reported in French and German milks from cows under various nutritional and management conditions: 3.3–3.8% (Wolff et al., Reference Wolff, Precht and Molkentin1998). Linseed feeding in our experiment increased the proportions of trans-18:1 FA, in agreement with the results reported by Loor et al. (Reference Loor, Ferlay, Ollier, Doreau and Chilliard2005) with a linseed oil supplement.
From the BH pathways of cis-9, cis-12 18:2 and cis-9, cis-12, cis-15 18:3 described by Harfoot and Hazlewood (Reference Harfoot and Hazlewood1988), we could have expected very different patterns of positional distribution of trans-18:1 isomers between C and linseed diets, because cis-9, cis-12 18:2 was the major PUFA in our diet C, and cis-9, cis-12, cis-15 18:3 was the major PUFA with linseed diets. Trans-11 18:1 should have largely been the most important isomer with diet C, whereas cis-15 18:1, trans-11 18:1 and trans-15 18:1 should have been the major isomers with linseed diets, and important proportions of trans-13 + 14 18:1 could also be expected (Ward et al., Reference Ward, Scott and Dawson1964; Loor et al., Reference Loor, Ueda, Ferlay, Chilliard and Doreau2004). In our experiment, with diet C, the trans-13 + 14 18:1 proportion was 66% and 55% of that of trans-10 + 11 18:1 in plasma and milk, respectively, and proportions of trans-12 18:1, trans-15 18:1, and trans-16 18:1 were all around 10% of the total trans-18:1 in plasma and milk. Trans-12 18:1, trans-15 18:1, and trans-16 18:1 have already been suggested to be produced during cis-9, cis-12 18:2 BH (Loor et al., Reference Loor, Bandara and Herbein2002), in addition to their production during cis-9 18:1 (Mosley et al., Reference Mosley, Powell, Riley and Jenkins2002) and cis-9, cis-12, cis-15 18:3 BH, and Piperova et al. (Reference Piperova, Sampugna, Teter, Kalscheur, Yurawecz, Ku, Morehouse and Erdman2002) published a distribution of trans-18:1 isomers close to ours in the milk from cows fed a diet without added fat, and where cis-9, cis-12, cis-15 18:3 represented 10.8% of the total dietary FA. On the contrary, Loor et al. (Reference Loor, Ferlay, Ollier, Doreau and Chilliard2005) found, with a diet without added fat but where cis-9, cis-12, cis-15 18:3 represented 25.8% of the total dietary FA, that trans-13 + 14 18:1 proportion was only 24% of that of trans-10 + 11 18:1 in milk, but 61% in plasma. After linseed oil supplementation, these authors found a distribution of major trans-18:1 isomers that was closer to our values with both added linseed forms, where trans-10 + 11 18:1 averaged 29.2% and 33.5% of plasma and milk total trans-18:1, respectively, and trans-13 + 14 18:1 averaged 33.5% and 29.3% of plasma and milk trans-18:1, respectively. In our experiment, relative to total trans-18:1, trans-12 18:1 and trans-16 18:1 were not changed by linseed addition. Trans-15 18:1 and cis-15 18:1, which are considered to be important final products of cis-9, cis-12, cis-15 18:3 BH (Harfoot and Hazlewood, Reference Harfoot and Hazlewood1988), increased by 244% and 677% in milk fat after linseed addition, but in spite of these large increases, these isomers were minor compared with trans-10 18:1, trans-11 18:1, or trans-13 + 14 18:1. The relative increase was higher for cis-15 than for trans-15 18:1, which was consistent with the large increase of trans-11, cis-15 18:2, the direct precursor of cis-15 18:1 in the rumen.
Most studies on the relationship between trans-18:1 and cardiovascular risk in human have compared cis-9 18:1 to trans-9 18:1, and the biological properties of this isomer have often been extrapolated to all trans-18:1 isomers. However, the position of the double bond can strongly affect the biological effect of trans-18:1 isomers, making this extrapolation suspect (Bauman et al., Reference Baumgard, Corl, Dwyer, Saebo and Bauman2004). Trans-9 18:1 represents about 6.9% of trans-18:1 FA in the fat of German milks produced under various feeding conditions (Wolff et al., Reference Wolff, Precht and Molkentin1998). In our experiment, this proportion was 5.3% with diet C and was decreased by linseed feeding. However, linseed increased the proportion of this isomer relative to total FA by 35% and 106% with diets RL and EL, respectively.
Compared with diet RL, diet EL resulted in higher proportions of CLnA and CLA in plasma and milk fat, and increased trans-11, cis-15 18:2 proportion in milk fat. Higher CLA proportion with extruded than raw oilseeds has already been reported (Chouinard et al., Reference Chouinard, Corneau, Bulter, Chilliard, Drackley and Bauman2001; Gonthier et al., Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005). In agreement with Gonthier et al. (Reference Gonthier, Mustafa, Ouellet, Chouinard, Berthiaume and Petit2005), feeding EL in our experiment also increased proportions of total and most individual 18:1 BH intermediates percentages in plasma and milk, compared with RL. The relative increase in milk was over 40% for trans-6 to trans-11 18:1 and cis-15 18:1, but was only 18% for trans-15 18:1, and around 10% for trans-12 and trans-16 18:1. Extrusion of oilseeds is known to result in increased proportions of trans intermediates in milk fat (Chouinard et al., Reference Chouinard, Lévesque, Girard and Brisson1997; Chouinard et al., Reference Chouinard, Corneau, Bulter, Chilliard, Drackley and Bauman2001) but our results show that the effects differ according to the trans-18:1 isomer, suggesting that extrusion affects the pathways of ruminal BH.
Conclusion
Including linseed in the diet of dairy cows resulted in a lower milk production and protein percentage. Milk proportions of FA from 6:0 to 16:0 decreased, but the proportions of cis-9 18:1, cis-15 18:1, all trans-18:1 isomers, CLA, trans-11, cis-15 18:2, CLnA and cis-9, cis-12, cis-15 18:3 increased when the diet contained raw or extruded linseed. With these diets, plasma and milk proportions of trans-13 + 14 18:1 were in the same range than that of trans-10 + 11 18:1, and were more than twice that of trans-15 18:1. EL, when compared with RL, tended to lower milk, milk fat and milk protein daily productions, further decreased the milk fat proportions of FA from 6:0 to 16:0, and increased the milk fat proportions of trans-9 18:1, trans-10 18:1, trans-11 18:1, trans-13 + 14 18:1, trans-15 18:1, cis-15 18:1, CLA, trans-11, cis-15 18:2 and CLnA. The present experiment provides the first evidence that milk CLnA can be manipulated via dietary linseed.
Acknowledgement
The authors thank Valorex (7 La Messayais, 35210 Combourtillé, France) for supply of linseed sources.
