Nutrition is the main environmental factor modifying milk fat secretion in cows. Dietary modifications can alter milk fat content, yield and composition, sometimes even leading to milk fat depression. Indeed, several nutrients produced during digestive processes are known to modify milk fat content and yield, as these nutrients are precursors or inhibitors of milk fat synthesis in the mammary gland. Acetic acid (Rook et al. Reference Rook, Balch and Johnson1965) and butyric acid (Huhtanen et al. Reference Huhtanen, Miettinen and Ylinen1993) increase milk fat content whereas propionic acid (C3) (Rigout et al. Reference Rigout, Hurtaud, Lemosquet, Bach and Rulquin2003), glucose (Hurtaud et al. Reference Hurtaud, Rulquin and Verite1998a) and trans-10, cis-12 conjugated linoleic acid (CLA) (Shingfield & Griinari, Reference Shingfield and Griinari2007) decrease milk fat content. Variations in all these nutrients explain the variations in milk fat following dietary modifications.
Infusion experiments have characterized and quantified the individual effects of these nutrients on milk fat synthesis (reviewed in Rulquin et al. Reference Rulquin, Hurtaud, Lemosquet and Peyraud2007). However, these nutrients vary simultaneously following dietary modifications, and there are scarce reports on their effects on milk fat synthesis when they are supplied together to dairy cows. It is still not known whether these nutrients have an additive effect or an interaction effect on milk fat secretion. Moreover, experimental conditions (diet, breed or period duration) differ among previously reported infusion experiments, thus preventing a comparison of effects between nutrients. A better knowledge of the respective effects of these nutrients under similar conditions and their putative interactions could also contribute to the prediction of the effects of dietary modifications on milk fat content and yield, which is not taken into consideration by current models used to estimate dairy cow requirements and formulate diets (NRC, 2001; INRA, 2007).
The objectives of this experiment were (i) to compare the individual effects on milk fat secretion of two of these nutrients when supplied alone under identical experimental conditions and (ii) to study their effects when supplied together, in order to test the possible interaction on milk fat secretion between the two nutrients. The two nutrients chosen were trans-10, cis-12 CLA and C3, both of which are known to reduce milk fat secretion, but by different mechanisms.
Materials and Methods
Animals and diet
This study used four Holstein cows fitted with cannulas in the rumen and duodenum (207±65 days in milk). Procedures related to surgery, care and use of animals for the experiment were approved by a French Ministry of Agriculture animal care committee in accordance with French regulations (Decree-law 2001-464, 29 May 2001). Cows were housed in tie stalls and milked at 06.30 and 17.00. They were individually fed a controlled amount of a total mixed ration (TMR) formulated to meet their energy, protein, mineral and vitamin requirements (INRA, 2007). TMR consisted of 198 g/kg dry matter (DM) orchard grass hay, 396 g/kg DM dehydrated corn pellets and 406 g/kg DM concentrate feedstuffs (Table 1), and was offered in equal amounts at 08·00 and 18·00.
Table 1. Ingredients and chemical compositionFootnote † (g/kg dry matter unless stated otherwise) of the total mixed ration (TMR) and feedstuffs used
† OM, organic matter; CP, crude protein; NDF, neutral detergent fibre; ADF, acid detergent fibre; EE, ether extract
‡ Wheat 200 g/kg, corn 200 g/kg, barley 200 g/kg, beet pulp 200 g/kg, wheat bran 150 g/kg, cane molasses 30 g/kg, vegetable oil 10 g/kg and salt 10 g/kg
§ Net energy for lactation values of feedstuffs were calculated from predictive equations (INRA, 2007)
Treatments and experimental design
The experiment was conducted according to a 4×4 Latin square design with 14-d periods. The four experimental treatments were in a 2×2 factorial arrangement: CON: control; CLA: trans-10, cis-12 CLA infused into the duodenum; C3: propionic acid infused into the rumen; and CLA+C3: trans-10, cis-12 CLA infused into the duodenum and propionic acid infused into the rumen. The trans-10, cis-12 CLA was provided by a mixture of fatty acids (Lutalin®, BASF AG, Ludwigshafen, D-67056, Germany) in which trans-10, cis-12 CLA represented 30% of total fatty acids (FA) (Table 2). The FA mixture was emulsified in skimmed milk (Lactalis, Bourgbarré, F-35230, France) with glycerol (Sigma-Aldrich Chimie, Saint Quentin, F-38297, France) and lecithin (Acros Organics, Geel, B-2440, Belgium) in order to obtain an adequate volume (2 l/d) for uniform and continuous infusion of trans-10, cis-12 CLA (A Ferlay, personal communication). The FA mixture was infused at 6·1 g/d, corresponding to 1·85 g/d of trans-10, cis-12 CLA. During CON and C3 treatments, only skimmed milk was infused into the duodenum. The daily dose of propionic acid (VWR International, Strasbourg, F-67023, France) was 500 g/d diluted in 40 l of tap water. C3 and CLA+C3 treatments included buffers (250 g/d NaHCO3 and 100 g/d KHCO3 dissolved in 10 l of tap water) infused into the rumen to limit the decrease in ruminal pH following propionic acid infusion. CON and CLA treatments included a solution of NaCl and KCl (volume 50 l) infused into the rumen in order to equalize Na and K supply across treatments. All solutions were prepared every morning prior to infusions. Peristaltic pumps delivered the infusates continuously over 24 h into the duodenum and the rumen via polyvinyl chloride tubes (0·5 cm i.d).
Table 2. Composition of the fatty acid (FA) mixture used for infusion into the duodenum
Measurements and sampling
Daily intake was determined by weighing TMR offered and refused. TMR and refusals were sampled each day and oven-dried for DM determination. Hay and concentrates were sampled at the end of each period. The chemical composition of ingredients and TMR is in Table 1.
At each milking, milk yield was recorded and fat and protein contents were determined by infrared analysis (Milkoscan, Foss Electric, Hillerød, Denmark). Lactose was measured by infrared analysis on day 12. On day 14, 100 ml of milk was taken from each cow at the morning and evening milkings. Morning and evening milk samples were pooled according to milk yields and stored at −20°C until freeze-drying for analysis of FA composition. Fatty acids in freeze-dried milk were directly methylated and analysed according to Loor et al. (Reference Loor, Doreau, Chardigny, Ollier, Sebedio and Chilliard2005).
Ruminal fluid was assayed for pH, volatile fatty acids (VFA) and ammonia on day 13 of each period. Individual subsamples (50 ml) were collected from the ventral sac of the rumen via a polyvinyl chloride tube at −0·5, +3 and +6 h after the morning meal. Ruminal pH was measured immediately. Then, following filtration through six layers of cheesecloth, individual subsamples were pooled before measuring VFA as described by Rigout et al. (Reference Rigout, Hurtaud, Lemosquet, Bach and Rulquin2003).
Statistical analysis
DM intake (DMI), milk yield and protein and fat contents were averaged over the last 5 d of each period. Data were analysed as a 4×4 Latin square using the MIXED procedure in SAS software (SAS Institute, Cary NC, USA). The statistical model included cow (random effect), period, CLA infusion, C3 infusion, interaction between CLA and C3 infusions (fixed effects) and residual error. Multiple comparisons of means were performed with the Tukey test, and differences (P<0·05) are indicated by different superscript letters in the tables. Milk Δ9-desaturase ratios were calculated according to Kelsey et al. (Reference Kelsey, Corl, Collier and Bauman2003). Total milk FA yields were computed from milk fat yields×0·933 (Glasser et al. Reference Glasser, Doreau, Ferlay and Chilliard2007).
Results
Intake and rumen parameters
Infusions had no effect on dietary DM or energy intakes (Table 3). Total net energy for lactation (NEL) (diet+infusions) was higher for C3 and CLA+C3 treatments than for CON and CLA treatments (+7·5 MJ/d on average). Infusions of C3 increased ruminal C3 percentage [from 21·6% of total VFA with CON to 25·1% and 27·2% with C3 and CLA+C3 treatments, respectively] at the expense of acetic acid. Following C3 infusions, there were no significant variations in the percentages of butyric, isobutyric, valeric and isovaleric acids, whereas caproic acid decreased. Compared with CON treatment, trans-10, cis-12 CLA infusions had no effect on rumen VFA percentages, confirming that there was no major modification of rumen fermentation. There was no significant between-treatment variation in ruminal pH and total VFA concentration.
Table 3. Feed intake, rumen pH and volatile fatty acids (VFA) in dairy cows infused with trans-10, cis-12 CLA into the duodenum or with propionic acid into the rumen, or both. Values are means with sem for 4 animals.
† CON=control, CLA=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA, C3=ruminal infusion of 500 g/d propionic acid, and CLA+C3=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA plus ruminal infusion of 500 g/d propionic acid
‡ Net energy for lactation, Intake+infusion
§ CLA: main effect of CLA infusions, C3: main effect of C3 infusions, CLA×C3: interaction between CLA and C3 infusions
¶ Values within a row without a common superscript letter are significantly different (P<0·05)
Milk yield and composition
There were no significant treatment effects on milk yield (Table 4). Trans-10, cis-12 CLA infusions decreased milk fat yield and content by on average 18% but did not affect milk protein or lactose. C3 infusions did not modify milk fat or protein, but tended to increase the percentage of lactose in milk (P⩽0·1). Interactions between CLA and C3 infusions were never significant.
Table 4. Milk yield and composition in dairy cows infused with trans-10, cis-12 CLA into the duodenum, or with propionic acid into the rumen, or both. Values are means with sem for 4 animals
† CON=control, CLA=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA, C3=ruminal infusion of 500 g/d propionic acid, and CLA+C3=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA plus ruminal infusion of 500 g/d propionic acid
§ CLA: main effect of CLA infusions, C3: main effect of C3 infusions, CLA×C3: interaction between CLA and C3 infusions
¶ Values within a row without a common superscript letter are significantly different (P<0·05)
Milk fatty acid composition
Interactions between trans-10, cis-12 CLA and C3 infusions were never significant for milk FA composition (Table 5). Infusions of trans-10, cis-12 CLA decreased the percentage of short- and medium-chain FA (6:0–12:0 and 16:0) and increased the percentage of total C18 (Table 5). Among the C18 FA, only cis-9 18:1, cis-11 18:1 and trans-10, cis-12 CLA were significantly increased with CLA infusions. Milk Δ9-desaturase ratios were not altered by CLA infusions. C3 infusions decreased the percentage of 4:0 but had no effect on the other even-chain FA synthesized de novo (6:0–16:0). C3 infusions decreased the percentage of 18:0 and increased the percentages of all the odd-chain FA (except 5:0 and 19:0). Milk Δ9-desaturase ratios tended to increase with C3 infusions (P⩽0·1): the increases in 14:1/14+14:1 and 16:1/16+16:1 were the consequence of 14:1 and 16:1 percentage increases, whereas the increase in 18:1, cis9/18:0+18:1 was related to a reduction in 18:0 percentage.
Table 5. Milk fatty acid (FA) profiles in dairy cows infused with trans-10, cis-12 CLA into the duodenum, or with propionic acid into the rumen, or both. Values are means with sem for 4 animals
† CON=control, CLA=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA, C3=ruminal infusion of 500 g/d propionic acid, and CLA+C3=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA plus ruminal infusion of 500 g/d propionic acid
‡ CLA: main effect of CLA infusions, C3: main effect of C3 infusions, CLA×C3: interaction between CLA and C3 infusions
§ Possibly collected with some 12:1
¶ Δ9-desaturase ratios were defined as [product of Δ9-desaturase]/[product of Δ9-desaturase+substrate of Δ9-desaturase]
†† Values within a row without a common superscript letter are significantly different (P<0·05)
Milk fatty acid yields
Infusions of trans-10, cis-12 CLA decreased the milk yield of all FA, i.e. short- and medium-chain FA (4:0–16:0), long-chain FA (total C18) and odd-chain FA (Table 6). C3 infusions decreased the yields of 4:0 and 18:0 and increased the yield of odd-chain FA (sum 5:0 to 19:0). Concerning milk FA yield, there was no significant interaction between trans-10, cis-12 CLA and C3 infusions. Milk trans-10, cis-12 CLA was undetectable in the CON and C3 treatments. Milk yield of trans-10, cis-12 CLA in the CLA and CLA+C3 treatments increased to 0·38 g/d and 0·29 g/d.
Table 6. Milk fatty acid yields in dairy cows infused with trans-10, cis-12 CLA into the duodenum, or with propionic acid into the rumen, or both. Values are means with sem for 4 animals
† CON=control, CLA=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA, C3=ruminal infusion of 500 g/d propionic acid, and CLA+C3=duodenal infusion of 1·85 g/d trans-10, cis-12 CLA plus ruminal infusion of 500 g/d propionic acid
‡ CLA: main effect of CLA infusions, C3: main effect of C3 infusions, CLA×C3: interaction between CLA and C3 infusions
§ Values within a row without a common superscript letter are significantly different (P<0·05)
Discussion
The first objective of this experiment was to compare individual effects on milk fat secretion of trans-10, cis-12 CLA and propionic acid in dairy cows. These two nutrients are produced during digestive processes and are known to decrease milk fat content and yield. The amounts of trans-10, cis-12 CLA and C3 infused were chosen so as to induce a similar and moderate (−3·0 to −5·0 g/kg) decrease in milk fat content, based on data from published experiments (Rulquin et al. Reference Rulquin, Hurtaud, Lemosquet and Peyraud2007). These amounts chosen (1·85 g/d of trans-10, cis-12 CLA and 500 g/d of C3) were, however, lower than the quantities infused in most experiments, i.e. 4 g/d for trans-10, cis-12 CLA (reviewed in Shingfield & Griinari, Reference Shingfield and Griinari2007) and 1000 g/d for C3 (reviewed in Hurtaud et al. Reference Hurtaud, Rulquin and Verite1998b). The moderate amounts chosen in the present experiment were designed so that the sum of the decreases induced by the two products infused together (−6·0 to −10·0 g/kg) would still be physiological.
In the present study, trans-10, cis-12 CLA reduced milk fat yield and content by 168 g/d and 6·5 g/kg, respectively, which was slightly higher than the reduction predicted from other experiments using similar infused amounts (Peterson et al. Reference Peterson, Baumgard and Bauman2002; de Veth et al. Reference De Veth, Griinari, Pfeiffer and Bauman2004). The other FA present in the FA mixture are neutral towards milk fat secretion (Christensen et al. Reference Christensen, Drackley, LaCount and Clark1994; Enjalbert et al. Reference Enjalbert, Nicot, Bayourthe and Moncoulon1998; Baumgard et al. Reference Baumgard, Corl, Dwyer, Saebø and Bauman2000) and cannot be responsible for the observed effects of trans-10, cis-12 CLA. The late lactation stage and the longer duration of experimental periods in the present experiment (14 d v. <5 d in most infusion experiments) could explain this higher than expected decrease: these two factors are suspected to increase the response to CLA infusions (Griinari & Bauman, 2006). Trans-10, cis-12 CLA infusion decreased the secretion of all FA, including the odd-chain FA. This decrease was proportionally greater for those FA synthesized de novo, resulting in a decrease in short and medium-chain FA percentages and an increase in preformed long-chain FA percentages. This result is consistent with the patterns generally observed with trans-10, cis-12 CLA infusions and the inhibiting role of this isomer on milk FA synthesis de novo (Shingfield & Griinari, Reference Shingfield and Griinari2007; Harvatine et al. Reference Harvatine, Boisclair and Bauman2009). As previously reported with small-amount infusions (Peterson et al. Reference Peterson, Baumgard and Bauman2002; Saebo et al. Reference Saebø, Perfield, Delmonte, Yurawecz, Lawrence, Brenna and Bauman2005), trans-10, cis-12 CLA did not alter Δ9-desaturase ratios, even if milk fat depression occurred.
Infusion of C3 did not modify milk fat content or yield. This result is in contrast with two published papers that reported a significant milk fat reduction with similar or smaller amounts infused (Rigout et al. Reference Rigout, Hurtaud, Lemosquet, Bach and Rulquin2003; Vanhatalo et al. Reference Vanhatalo, Varvikko and Huhtanen2003). Two hypotheses can be raised to explain this result: first, the basal percentage of C3 in the rumen was higher than in the two other studies (on average 22% v. 17%) which could explain why the increase in rumen C3 provoked by the infusions was lower. Second, comparison with other experiments suggests that the decrease in milk fat could be more pronounced when the initial milk fat content is high (i.e. above 40 g/kg) for a given amount infused and under similar basal diets (Rook et al. Reference Rook, Balch and Johnson1965; Ueyama et al. Reference Ueyama, Tanaka and Hirose1972). The initial milk fat content in the present experiment was lower than in the two experiments using similar amounts (Rigout et al. Reference Rigout, Hurtaud, Lemosquet, Bach and Rulquin2003; Vanhatalo et al. Reference Vanhatalo, Varvikko and Huhtanen2003), i.e. only 33 g/kg here v. more than 40 g/kg. With the C3 infusion, although there was no decrease in milk fat content or yield, milk FA profiles and yields were modified: 4:0 and 18:0 yields and percentages decreased while all the odd-chain FA from 7:0 to 17:0 increased. This result is consistent with several published reports (Hurtaud et al. Reference Hurtaud, Rulquin and Verite1998b; Rigout et al. Reference Rigout, Hurtaud, Lemosquet, Bach and Rulquin2003). In the present experiment, however, we did not observe the decrease in short- and medium-chain even-chain FA reported by these authors. The drop in even-chain FA is the consequence of a decrease in plasma concentrations of the following precursors: acetate and β-hydroxybutyrate for de-novo synthesized FA, non-esterified FA and total glycerides for preformed FA in all experiments with (Hurtaud et al. Reference Hurtaud, Rulquin and Verite1993; Huhtanen et al. Reference Huhtanen, Blauwiekel and Saastamoinen1998; Rigout et al. Reference Rigout, Hurtaud, Lemosquet, Bach and Rulquin2003) or without isoenergetic supplementation (this study: data not shown; Lemosquet et al. Reference Lemosquet, Delamaire, Lapierre, Blum and Peyraud2009). However, the precise mechanisms by which C3 acts on milk fat synthesis remain unknown. The increase in odd-chain FA can be explained by the increase in propionate available to the udder as a precursor of de-novo synthesis (Massart-Leen et al. Reference Massart-Leen, Roets, Peeters and Verbeke1983; Emmanuel & Kennelly, Reference Emmanuel and Kennelly1985). The increase in Δ9-desaturase ratios has already been observed with C3 infusions (Hurtaud et al. Reference Hurtaud, Rulquin and Verite1993; Miettinen & Huhtanen, Reference Miettinen and Huhtanen1996) suggesting increased desaturase activity by the mammary gland due to the C3 supply.
The second objective of this experiment was to study the combined effects of trans-10, cis-12 CLA and C3 on milk fat secretion when supplied together to dairy cows, and to determine whether their effects were additive. Under our conditions, interactions between CLA and C3 were never significant, meaning that the effects of the two nutrients on milk FA yields were additive, whatever their individual effects: (i) when both nutrients had a negative effect, these effects were additive when they were provided together (e.g. on 4:0 and 18:0 yields); (ii) when C3 had no significant effect and trans-10, cis-12 CLA had a negative effect, the effect of CLA+C3 was similar to that of CLA alone (e.g. 16:0 yield); (iii) when C3 had a positive effect and trans-10, cis-12 CLA had a negative effect, the result of CLA+C3 was similar to that of the control treatment (e.g. odd-chain FA yield). These three situations confirm the additive effects of these two nutrients on mammary lipogenesis. The observed additivity could result from the different mechanisms by which trans-10, cis-12 CLA and C3 act on milk fat synthesis.
In conclusion, this study describes the individual effects on milk fat secretion of trans-10, cis-12 CLA and C3 supplied alone or together, under the same experimental conditions. It shows that trans-10, cis-12 CLA infusions had a greater effect than C3 infusions, and induced different patterns of change in the milk FA profiles and yields. Moreover, under our conditions, their effects on milk FA yields were additive when they were supplied together. If these findings are validated for other nutrients, this could be used to estimate milk fat and FA changes based on variations of nutrient supplies.
The authors would like to thank P Lamberton and his team: D Chevrel, A Cozien, JL Harel, M Lemarchand and B Grehal for their helpful assistance and for animal care and feeding, and M Texier, T Le Mouel, N Huchet and P Capitan for technical assistance. We also thank Arrivé, BNA Nutrition Animale, CCPA, Evialis, Inzo, Glon-Sanders, Primex and Provimi France for funding this study.
