- Email: [email protected]
Secretion of Oleic Acid in Milk Fat in Response to Abomasal Infusions of Canola or High Oleic Sunflower Fatty Acids1
Secretion of Oleic Acid in Milk Fat in Response to Abomasal Infusions of Canols or High Oleic Sunflower Fatty Acids 1 D. W. LaCOUNT, J. K. DRACKLEY,2 S. O. LAESCH,3,4 and J. H. CLARK Department of Animal Sciences University of Illinois Urbana 61801 ABSTRACT
between yield of C18:1 in milk fat and the amount of C18:1 infused into the abomasum was linear, and transfer efficiency was 54.1 %. Increased concentration and yield of C 18:1 in milk were attributable mostly to the increased exogenous supply of C I8:1' (Key words: fatty acid, milk fat, oleic acid, postruminal infusion)
The feasibility of dietary strategies to increase oleic acid content of milk fat is unclear. Four Holstein cows were infused abomasally with free long-chain fatty acids from canola (62.5% C I8:1) or high oleic sunflowers (86% CI8:1)' Each fatty acid mixture was infused for 3 d at 0, 133,267,400,267, 133, and 0 gld for a total of 21 d; cows then were changed to the opposite fatty acid mixture, and the infusion sequence was repeated. The DMI and percentages of casein and whey N in milk were decreased by infusion of fatty acids, but milk yield and percentages of fat and NPN in milk were increased. Contents of short- and mediumchain fatty acids and C16:0 in milk fat decreased, and contents of CI8:), CI8:2, and C18:3 increased, as fatty acid infusion increased. Contents of C16:0 and C18:0 in plasma triglyceride were decreased, and content of C18:1 was increased, by increasing infusion of fatty acids. All changes reversed when the amount of fatty acid infused was decreased. Within the range of amounts of fatty acids infused, the relationship
Abbreviation key: FA = fatty acids, HOSF = high oleic sunflower, TGLP = triglyceride-rich lipoproteins. INTRODUCTION
Received September I, 1993. Accepted January 7, 1994. ISupported by Hatch funds appropriated to the lllinois Agricultural Experiment Station (Projects 35-0352 and 35-0359) and by a gift from Milk Specialties Co., Dundee, lL. 2 Address correspondence and reprint requests to James K. Drackley, Department of Animal Sciences, 260 Animal Sciences Lab, University of Illinois, 1207 West Gregory Drive, Urbana, lL 61801. 3Supported by the Jonathan Baldwin Turner Undergraduate Research/Scholarship program, College of Agriculture, University of llIinois. 4Present address: Rolling Lawn Dairy Fana, RR 3, Box 135, Greenville, lL 62246.
1994 J Dairy Sci 77:1372-1385
During the past decade, Americans have become increasingly health conscious. Reports in the media have highlighted recommendations from the medical community and government agencies that people decrease their intake of saturated fatty acids (FA) because of the purported relationship of some saturated FA with atherosclerosis. In particular, the FA C14:0 and C 16:0 may increase plasma cholesterol (9); however, C 18:0 may not affect cholesterol (4). Dairy products have been a target for human health advocates because milk fat contains, on average, about 70% saturated FA, 25% monounsaturated FA, and 5% polyunsaturated FA (15). Over one-half of the saturated FA in milk are C 14:0 and C 16:0 (15). In 1988, the Wisconsin Milk Marketing Board assembled a group of 15 researchers from industry and academia to discuss the potential future uses of milk fat. These researchers concluded that the "ideal" milk fat for use in the human diet would contain <10% polyunsaturated FA, ~8% saturated FA, and ~82% monounsaturated FA (23). Grummer (15) concluded that such a large change in the FA composition of milk is not possible by alteration of the diet of dairy cows. Although manipulation of the diet of dairy cows may not yield the ideal FA profile in
1372
INCREASING OLEIC ACID IN MILK FAT milk, the composition of milk fat potentially can be altered by addition of fat to the diet. Inclusion of canola (1, 3, 19, 22, 33) or regular sunflower (5, 25) seeds in diets of lactating dairy cows decreased the concentrations of short- and medium-chain FA and increased the concentration of C I 8:1 in milk fat. Canola oil is 55 to 60% C 18: 1; regular sunflower oil is 65 to 70% CI8:2' When either of these oil seeds is fed, similar increases occur in the C I8 :1 content of milk; however, when regular sunflower seeds are fed, much of the resulting increase OfCI8:1 in milk fat is in the trans11 configuration because of incomplete biohydrogenation of CI8:2 (5, 20). In contrast, when high oleic sunflower (HOSF) seeds were fed, increases of trans-CI8:1 were smaller (5). Increased trans monounsaturated FA in the human diet may not be desirable because high intakes of trans FA increase total and low density lipoprotein cholesterol and decrease high density lipoprotein cholesterol compared with a diet high in cis FA (21). Technology to increase bypass of FA from sources such as canola and HOSF that are high in cis monounsaturated FA may more effectively alter the FA composition of milk than when oil seeds are fed. Milk ciS-C I8 :1 also can be derived from desaturation of C18:0 in the intestine or mammary gland (15). Few data are available on the responsiveness of C18:1 in milk to different amounts of CI8:1 absorbed from the intestine. Our laboratory has developed a technique to infuse free FA into the abomasum of dairy cows (11). The primary objective of this study was to determine the potential to alter the content of C I 8:1 and other FA in milk by abomasal infusion of different amounts of two FA mixtures (derived from canola or HOSp) rich in CI8:I' Previous experiments in our laboratory (7, 11) demonstrated that abomasal infusion of unsaturated FA decreased DMI and milk yield compared with saturated FA and that milk protein percentage tended to be decreased more by unsaturated than by saturated FA (7). Potential changes in content or composition of milk N fractions related to specific postruminal FA would be important for nutritional value and processing characteristics of milk (10). Therefore, a secondary objective was to determine the effects of FA infusions on yield and composition of N fractions in milk.
1373
TABLE 1. Ingredient and nutrient composition of TMR. Composition
(%
Ingredient Alfalfa haylage Corn silage Soybean hulls Ground shelled corn Soybean meal (48% CP) Sodium chloride Mineral and vitamin mixture l Limestone Dicalcium phosphate Magnesium oxide Sodium bicarbonate Sodium sulfate Nutrient2 DM OM CP
of
DM)
30.00 20.00 6.00 29.50 11.50 .30 .15
.80 .60 .20 .75 .20 70.6 90.9 17.0
21.0 33.2
ADF NDF
Total FA3
3.1
lContained .004% Co, .5% Cu, .025% I, 2.0% Fe, 5.0% Mg, 3.0% Mn, 7.5% K, .015% Se, 10.0% S, 3.0% 20, 2200 IU/g of vitamin A, 660 IUlg of vitamin D3' and 8 IU/g of vitamin E. 2As analyzed. 3Fatty acids.
MATERIALS AND METHODS Cows, Experimental Design, end Diet
Four multiparous Holstein cows, fitted with ruminal cannulas and averaging 166 DIM (range 129 to 182 d), were used in a singlereversal design with 21-d periods. Cows were housed in individual stanchions, milked twice daily, and fed for ad libitum intake a TMR (Table 1) twice daily. The diet contained 17% CP, 21% ADF, 33% NDF, and 3.1% total FA (Table 1). Treatments were continuous abomasal infusions of free FA prepared from HOSF or canola oils (Henkel Corp., Emery Division, Cincinnati, OH). Two cows were infused with each FA mixture during each 21-d period. The FA were infused for 3 d at each of four amounts in a sequence of 0, 133, 267,400,267, 133, and 0 gld (i.e., a total of 21 d for each FA mixture); cows then were changed to the opposite FA mixture, and the infusion sequence was repeated. The largest amount of FA infused was 400 gld because previous work in our laboratory (11) indicated that postruminal infusion of >450 gld of unJournal of
Dairy
Science Vol. 77, No.5. 1994
1374
LaCOUNT ET AL.
TABLE 2. Composition of fatty acid (FA) mixtures and TMR. FA
High oleic sunflower
Canola
TMR
(glIOO g of FA) CI2:0 C 14 :0 C!6:0 C 16 :1 C 18 :0 C I8:! C18:2 C18:3
NDI .1 4.5 .4
1.2 62.5 24.1 7.2
ND ND 5.0 .2 3.5 86.0 5.3 ND
2.1 .3 22.1 2.0 3.4 17.0 44.3 8.8
INot detected.
saturated FA caused cows to go off feed, to decrease milk yield, and to develop diarrhea. The composition of the two FA mixtures that were infused and of the TMR is given in Table 2. The major FA in canola were C18:1 (62.5%) and C 18:2 (24.1%). The HOSF FA contained 86.0% C I8 :! and only small quantities of other FA.
An infusion tube was placed into the abomasum through the ruminal cannula (II). Daily infusions consisted of the FA complexed with 180 g of meat solubles (Milk Specialties Co., Dundee, IL) or 180 g of meat solubles alone in 9 L of water. To prepare the solutions for infusion, the meat solubles were dissolved in hot (50 to 60°C) tap water. The FA were heated in an oven at 150°C until they were fully melted and then were added slowly to the solution of meat solubles with continuous stirring. Suspensions were stirred continuously on magnetic stir plates while being infused into the abomasum over 20 to 22 hid with peristaltic pumps (Harvard Apparatus, South Natick, MA). Feed intake was measured daily. Feeds and orts were sampled every 3 d and composited within each period. Samples of feeds were dried at 55°C, ground, and analyzed for contents of OM, ash (600°C for 8 h), CP (2), NDF by using a-amylase (32), ADF (32), and FA (31). Milk Yield and Composition
Milk yield was measured and recorded at each milking. Milk samples were taken from two consecutive milkings every 3 d so that a Journal of Dairy Science Vol. 77, No.5, 1994
sample was obtained during infusion of each amount of FA. Milk samples were preserved with 2-bromo-2-nitropropane-I,3-diol, composited into daily samples in proportion to milk production, and analyzed for contents of CP, true protein, and fat by mid-infrared spectrophotometric (28) analysis (New York DHIA Laboratory, Ithaca, NY). Fat was measured using the A filter (28). Noncasein N was determined by Kjeldahl analysis of the filtrate after precipitation with 10% acetic acid and IN sodium acetate (17). Casein N was calculated as the difference between total Nand noncasein N, and NPN was calculated as the difference between total N and true protein N. A portion of the daily composite milk sample was placed into a test tube and allowed to stand overnight in a cold room at 4°C. The cream layer was transferred into a clean test tube, and FA were methylated and quantified by GLC using the procedures of Sukhija and Palmquist (31), except that an external standard rather than an internal standard was used to quantify the proportions and yields of FA in milk (27). Glycerol in milk was calculated as described by Schauff et al. (27). Blood Sampling and Analyses
On d 3 of infusion of each amount of FA, blood v.as sampled from a coccygeal vein at 1300 h, which was 5 h after the infusions began in the morning. Concentrations of NEFA (27) and total cholesterol (Sigma kit number 352; Sigma Chemical Co., St. Louis, MO) were measured in plasma. An additional 60 ml of blood were obtained for separation of triglyceride-rich lipoproteins (fGLP) by ultracentrifugation (7, 16), and the concentration of triglyceride in TGLP was determined (Sigma kit number 339; Sigma Chemic~l Co). Fatty acid composition of triglycendes In the TGLP fraction was determined after TLC. One milliliter of the TGLP fraction was extracted with 20 ml of chloroform-methanol (2: 1, voVvol), which contained an internal standard (triheptadecanoin, .01 mg/ml) and an antioxidant (butylated hydroxytoluene, .2 mg/ml). Four milliliters of .74% (wtJvol) KCl were added to the tube, and the tube was mixed and centrifuged (100 x g) for 10 min. The ~ueous layer was aspirated and discarded. The Sides of the tube and the surface of the chloroform
1375
INCREASING OLEIC ACID IN MILK FAT
layer were washed three times with 2 ml of .74% (wt/vol) KCl. The chloroform fraction was evaporated under air, and the dry extract was redissolved immediately in .4 ml of 2: I (volJvol) chloroform-methanol. Aliquots (100 1-'1) of lipid extract were spotted on TI...C plates (silica gel G with preadsorbent zone, channeled Analtech Uniplates™; Alltech Associates, Deerfield, ll...). Plates were developed twice in methanol to 1.5 cm above the preadsorbent line, air dried, and then developed to 1.5 cm above the preadsorbent line in chloroform: methanol (1:1, volJvol). After air drying, plates were fully developed using hexane:diethyl ether:acetic acid (80:20:1, volJvolJvol). Lipid bands were visualized under UV light after the plates were sprayed with rhodamine-6G (.02% in 95% ethanol). The triglyceride band was identified by comparison with standards in adjacent lanes and scraped into screw-capped tubes. The FA were methylated and quantified by GLC using the procedure of Sukhija and Palmquist (31).
26 25 24
,..... -0
23
......... Ol
.:,{. ......,.. 22
:::::!: 0
\" . /. . ~ '>\/~ •
21 20 19 18
L-----'_----'-_-L_-l-_-'--_L-----'_-'
o
'33 267 400 267
133
0
Amount of FA infused (g/d)
Figure I. The DMI for cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers ~) (SE of the difference between treatment means at any amount = .75). Probabilities of greater F for effects in model: treatment, P = .88; amount, P = .03; and treatment by amount, P = .05.
Statistical Analyses
Data were subjected to ANOVA for a splitplot, single-reversal design using the general linear models procedure of SAS (26). The model contained effects of sequence (1 df), cow within sequence (2 df), period (1 df), FA source (i.e., treatment, I df), whole-plot error (2 df), FA amount (6 df), and the interaction of FA source and amount (6 df). Effects of FA source and period were tested using the wholeplot error term, and effects of FA amount and the interaction of source of FA by amount of FA infused were tested using the residual error. Linear and quadratic effects of increasing the amount of FA infused from 0 to 400 gld were tested (within treatments, if effects of treatment or the interaction of treatment by amount were significant). The second period was decreased to 18 d, and FA were infused in the sequence of 0, 133, 267, 400, 115, and 0 gI d for both FA sources because of a limited amount of FA from HOSF. Missing values were assumed for the second infusion of 267 gld, and the amount of 115 gld replaced the amount of 133 gld used in the first period. Least squares means are presented throughout. Significance was declared at P :s; .05.
RESULTS AND DISCUSSION
OMI, Milk Yield, and Milk Composition
The interaction of source of FA by amount of FA infused for OMI was significant because of a linear decrease of DMI when cows were infused with increasing amounts of FA from canola (Figure 1). The DMI increased as the amount of canola FA in the infusate was decreased, but OMI tended to decrease as the amount of FA from HOSF infused was decreased. Milk yield increased linearly when cows were infused with FA from HOSF (Figure 2). Peak milk yield seemed to occur 3 d after the infusion of 400 gld of FA, which may have been a consequence of the short infusions at each amount of FA. The design and number of cows used did not allow accurate determination of effects of infused FA on OMI and milk yield; longer term studies with more cows are needed to evaluate these responses. Duodenal infusion of 1100 gld of rapeseed oil decreased OMI (14), and abomasal infusions of 450 gld of canola or regular sunflower FA (7) or 450 gld of soybean FA (11) decreased OMI and milk yield. Journal of Dairy Science Vol. 77. No.5. 1994
1376
LaCOUNT ET AL.
.36
Similar to milk yield, the N fractions of milk seemed to have a delayed response to the FA infusions (i.e., changes were not maximal until 3 d after the infusion of 400 gld of FA). True protein N and whey protein N expressed as percentages of total N in milk decreased quadratically, and NPN as a percentage of total N increased quadratically, as increasing amounts of FA were infused (Table 3). These changes
.35 ~
, , , , ,. ,, , ,
.34
.---.. .33 "U
"(J) .:c£ ......,
32
.:c£
:::;;;
.31 30 29
4.2
A 28
'------'---'------'-_-'--_--L_---'---_....L-----.J
o
133
267
400
267
133
4.0
0
Amount of FA infused (g/d)
........ 3.8 Figure 2. Milk yield for cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers (II) (SE of the difference between treatment means at any amount = .99). Probabilities of greater F for effects in model: treatment, P = .67; amount, P = .01; and treatment by amount, P = .14.
~
'-'
....0 ....
3.6
.::£
:::;;;
3.4 3.2
The percentage of fat in milk was affected significantly by the amount of FA infused (Figure 3A), primarily because of a decrease when the amount of infused FA was decreased. Yield of milk fat increased when more FA were infused (Figure 3B). When cows were infused with canola FA, the increased fat yield was attributable mostly to a tendency (P = .15) for linearly increased fat percentage, but for HOSF infusion the increased fat yield was from greater milk yield. Similar to this study, postruminal infusion of 1100 gld of rapeseed oil increased milk fat percentage slightly in midlactation dairy cows (13). Abomasal infusion of highly saturated or soybean FA did not alter milk fat percentage (11). Percentages and yields of total N, true protein N, casein N, and whey protein N in milk were unaffected (P > .15) by source of FA (data not shown). Percentage of whey protein N in milk decreased quadratically, and the percentage of NPN increased quadratically, as increasing amounts of FA were infused (Table 3). Yields of total N, true protein N, and casein N were unaffected by the amount of FA infused. The yield of NPN increased quadratically as the amount of FA infused increased. Journal of Dairy Science Vol. 77, No.5, 1994
3.0 1300
B
1250
........
1200
"U
"(J)
1150
'-'
....0 .... .::£
1100 1050
,. ,·~·/\ . " .d •v· "
,,
,, ,
:::;;;
•
1000 950 900
'------'-_-'-_-'-_"-----'_---'-_....L----O
o
133 267 400 267 133
0
Amount of FA infused (g/d) Figure 3. Percentage and yield of milk fat from cows infused postruminally with fatty acids (FA) from canola Ce ) or high oleic sunflowers (II). A) Milk fat percentage CSE of the difference between treatment means at any amount = .13). Probabilities of greater F for effects in model: treatment, P = .06; amount, P = .0001; and treatment by amount, P = .66. B) Milk fat yield; standard error of the difference = .05. Treatment, P = .86; amount, P = .0004; and treatment by amount, P = .43.
1377
INCREASING OLEIC ACID IN MILK FAT TABLE 3. Effect of amount of fatty acid (FA) infused on N components of milk. l Amount of FA infused Item
o g/d
133 g/d
267 g/d
400 g/d
267 g/d
133 g/d
o g/d
SEM
.519 156.4
.517 153.8
.522 161.4
.509 158.0
.496 166.1
.509 160.8
.515 156.5
2,83
.0453 .2169
.486 146.4 93.7
.485 143.8 93.7
.488 150.9 93.5
.471 146.5 92.5
.459 153.4 92.5
.472 149.1 92.8
.481 145.8 93.2
.0045 2.63 .17
.0067 .3351 ,00012
.401 120.8 77.2
398 118.5 76.8
.403 124.5 77.2
.394 122.1 77.2
.380 127.0 76.4
.391 123.4 76.6
.395 119.8 76.5
.0039 2.28 .20
.0357 .3102 .0395
.085 25.7 16.5
.087 25.7 16.9
.085 26.2 16.4
.078 24.1 15.3
.080 26.6 16.2
.082 25.8 16.2
086 26.0 16.8
.0016 .53 .27
.0026 2 .1155 2 .0085 2
.033 9.9 6.3
.033 9.8 6.3
.034 10.6 6.5
.038 11.8 7.5
.037 12.5 7.5
.036 11.6 7.2
.034 10.6 6.8
.0009 .31 .17
.00102 .00012 .00012
P
Total N %
gld True protein N %
g/d % of Total N Casein N %
g/d % of Total N Whey protein N %
g/d % of Total N NPN %
gld % of Total N
.0044
'Means across both treatments for each amount of FA infused; effects of treatment and the interaction of treatment and amount were nonsignificant. 2Quadratic effect of increasing FA infusion from 0 to 400 gld (J> < .05).
are similar to those often observed when supplemental fats are fed to cows (10). In other studies, milk CP was decreased (7, 13) or unaffected (11) by postruminal infusion of FA. FA Composition and Yields in Milk
Infusion of increasing amounts of FA from canola or HOSF into the abomasum linearly decreased contents of CIO:O, CI4:0' C 14 : b CI5:0' CI6:0, CI6:J, and C 17:o in milk fat (Table 4). Infusion of FA from HOSF, but not canola, linearly decreased content of C8:0. The decrease of content of C12:0 in milk was linear for canola and quadratic for HOSF FA. The decreases in the contents of total short- and medium-chain FA (y,:o to CI4:0) and C16:0 in milk fat are shown in Figure 4, A and B, respectively. Infusion of FA from canola or HOSF did not affect the contents of C4:0' C6:0, or C 18:0 in milk fat. The content of glycerol decreased linearly as the amount of FA from HOSF infused increased, as a consequence of the increase in long-chain FA and decrease in short- and medium-chain FA attached to the glycerol moiety.
The interaction between source of FA and amount of FA infused was significant for content of C18:1 in milk because it increased more when cows were infused with FA from HOSF than from canola (Figure 4C). Significant interactions also occurred for contents of C18:2 (Figure 4D) and C18:3 because concentrations of these FA in milk increased more when cows were infused with FA from canola than from HOSF. Differences in contents of CI8:J, CI8:2, and C18:3 in milk between FA sources correspond to the differences in composition of the FA sources (Table 2). The interaction between source of FA and amount of FA infused was not significant for the contents of other FA or glycerol in milk. The changes in contents of FA in milk that accompanied the infusion of increasing amounts of FA reversed when the amount of FA infused was decreased. These data indicate that changes in milk fat composition can be made that might be beneficial for human nutrition (4, 9, 23), including decreased contents of C14:0 and CI6:0' increased content of C18:b and unchanged content of C I 8:0' The yields of FA and glycerol in milk (data not shown) generally followed patterns similar Journal of Dairy Science Vol. 77, No.5, 1994
1378
LaCOUNT ET AL.
TABLE 4. Effects of amount of canola or high oleic sunflower (HOSF) fatty acids (FA) infused on weight percentages of FA and glycerol in milk. Amount of FA infused Item
o gld
133 gld
267 gld
400 gld
267 gld
133 gld
o gld
SEM
Significant effects l
(gil()() g of fat) C4:0 Canola HOSF
3.84 4.00
3.81 4.14
4.16 4.05
4.32 3.97
3.15 3.95
3.82 3.60
4.04 3.69
.21 .21
C6:0 Canola HOSF
2.54 266
2.59 2.70
2.96 2.61
2.70 2.44
2.27 2.58
2.58 2.39
2.82 2.65
.12 .12
C8:0 Canola HOSF
1.31 1.42
1.38 1.36
1.43 1.31
1.32 1.25
1.28 1.26
1.33 1.29
1.39 1.42
.04 .04
L
C1O:0 Canola HOSF
3.00 3.15
3.04 2.88
2.95 2.71
2.68 2.56
2.74 2.55
2.98 2.80
3.10 3.25
.08 .08
A L L
C12:0 Canala HOSF
3.78 3.98
3.72 3.50
3.38 3.16
3.10 3.00
3.27 3.07
3.67 3.46
3.88 4.09
.11 .11
C14:0 Canala HOSF
A
115 11.7
11.1 10.8
10.1 10.2
9.5 9.8
10.2 9.9
11.0 10.9
11.6 11.8
A L Q
.16 .16
A L L
C14:1 Canala HOSF
1.73 1.77
153 157
1.29 1.47
1.10 1.36
1.35 1.42
1.63 1.67
1.80 1.88
.04 .04
A L L
C15:0 Canala HOSF
1.26 1.15
1.12 1.11
.96 1.05
.87 .95
.99 1.05
1.14 1.19
1.24 1.26
.04 .04
A L L
C16:0 Canola HOSF
28.52 28.70
25.54 25.93
21.93 22.71
20.07 20.68
22.96 21.40
24.59 24.52
27.43 27.46
.42 .42
A L L
C16:1 Canola HOSF
1.85 1.86
1.61 1.70
1.39 1.52
1.27 1.42
1.45 1.44
1.63 1.66
1.81 1.87
.04 .04
A L L
C17:0 Canola HOSF
.70 .70
.68 .65
.61 .60
.58 .56
.63 60
.66 .64
.68 .68
.01 .01
A L L
C18:0 Canola HOSF
7.10 6.81
7.12 6.85
6.91 6.84
6.83 6.22
7.47 7.10
7.25 7.16
7.03 6.81
.25 .25
C18:1 Canola HOSF
17.54 16.59
20.07 21.35
22.94 26.60
25.42 3Q.42
2446 28.29
21.03 23.63
17.52 17.87
.49 .49
C18:2 Canala HOSF
2.33 2.27
3.33 2.28
4.92 2.28
6.13 2.52
4.74 2.30
3.50 2.12
2.27 2.22
.12 .12
C18:3 Canola HOSF
.30 .41
.69 .38
1.33 .27
1.38 .32
.71
.43
.51 .40
.47 .31
.14 .14
Glycerol Canala HOSF
12.77 12.87
12.72 12.81
12.79 12.68
12.71 12.57
12.38 12.60
12.71 12.59
12.89 12.79
.08 .08
A, T, TxA L L A, T x A L A, T x A L A
=
=
IFor overall effects of model: A amount; T treatment; and T x A treatments: L linear effect of increasing FA infusion from 0 to 400 gld; Q from 0 to 400 g/d.
=
Journal of Dairy Science Vol. 77. No.5. 1994
L
= treatment by amount interaction. Within =quadratic effect of increasing FA infusion
1379
INCREASING OLEIC ACID IN MILK FAT
to changes in concentrations, except that yields of Cs: o and C IO :O were unaffected by amounts of FA infused. De novo synthesis of these FA was unaffected by infusion of FA despite decreases of concentrations of Cs:o and C10:0 in milk. As more FA were infused, greater decreases of concentrations and yields of FA that are synthesized de novo occurred as the chain length increased from C6:0 to CI4:0' This pattern of changes is similar to the effect of mobilization of long-chain FA from adipose
tissue on de novo FA synthesis in the mammary gland during early lactation (24). In other experiments, postruminal infusion of FA or supplementation of FA to the diet also changed the FA profile of milk in a manner similar to this study. Abomasal infusion of 450 gld of saturated or unsaturated FA decreased contents of short- and medium-chain FA and C 16:0 in milk fat but increased contents of unsaturated CIS FA in milk (11). Chilliard et al. (6) reported decreased contents of CI4:0, C I4:], C I5 :0, C I6 :0, and C16:1 and increased contents
34
A o
24
~
28
:ii
22
fI
16
I
I
I
I
I
26
to
u
24
18
,;
\,~
,'.
e,\
V
18 16
t
14
I
\"
~
"
8
"\'\
D _
7 6
.
/-
"~~:/
22 20
f-
~
B
.30 '-
o
,
'
20
I
."/------- " •
I
.y.
32
28
24
c
,~ . .
I
u... 22
20 .... 18
30
......... 26
to
U
~
.....
~
r-
:.j.
o .....o
,......
32
5 N lC
4
3
2
-
I-
o
I
o
133 267 400 267 133
0
Amount of FA infused (g/d)
o
133267400267 133
0
Amount of FA infused (g/d)
Figure 4. Weight percentages (grams per 100 g) of C6 to C I4 :O' C 16:0' CI8:1. and C 18:2 in the fat of milk from cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers (II). A) Weight percentage of C6 to C14:0 in milk fat (SE of the difference between treatment means at any amount = .5). Probabilities of greater F for effects in model: treatment, P .47; amount, P .0001; and treatment by amount. P .40. B) Weight percentage of C 16 :0 in milk fat (SE of the difference .6). Treatment, P .S7; amount, P = .0001; and treatment by amount, P .47. C) Weight percentage of C18:1 in milk fat (SE of the difference .7). Treatment, P .04; amount, P = .0001; treatment by amount, P .0001. D) Weight percentage of C18:2 in milk fat (SE of the difference .2). Treatment, P .001; amount, P = .0001; and treatment by amount, P .0001.
=
=
=
=
=
=
=
=
= =
=
=
Journal of Dairy Science Vol. 77, No.5,
1994
1380
LaCOUNT ET AL.
of C I 8:!> C18:2, and C18:3 in milk fat when midlactation cows were infused with 1100 g1d of rapeseed oil into the duodenum. Supplementing canola (I, 19, 20) or sunflower seeds (5, 25) to the diet of dairy cows decreased the contents of short- and mediumchain FA and increased the content of C 18:1 in milk. Inclusion of regular sunflower seeds in the diet (5) increased the C18:1 content of milk fat by 16 percentage units. Because regular sunflower seeds contain predominantly C I 8:2, the increased content of C18:1 was attributable in part to incomplete biohydrogenation of C I8 :2; 8.1% of the milk FA were in the transC 18:1 configuration (5). In contrast, the same amount of HOSF seeds (5) increased C 18:1 by 12 percentage units, but only 4.3% of the FA in milk was in the trans-C I 8:1 configuration. Decreases in contents of short- and medium-chain FA and C 16:0 in milk when more FA were infused occurred more rapidly than the increases when less FA were infused (Figure 4, A and B; Table 4). De novo synthesis of short- and medium-chain FA by the mammary gland seems to have been inhibited almost immediately by infusion of increasing amounts of FA. Alleviation of this inhibition evidently was not as rapid when the amount of FA infused was decreased. The rapid decrease of short- and medium-chain FA and C 16:0 in milk fat caused by infusion of more FA and the subsequent slower increase of these FA in milk as FA were removed from the infusate may explain the similar pattern of changes in milk fat percentage (Figure 3A). The decrease in milk fat percentage was similar when Ca salts of long-chain FA were removed from the diet (12); the authors speculated that de novo synthesis of FA was slow to recover after inhibition by dietary fat (12). Other researchers (29, 30) proposed that the decrease in short- and medium-chain FA in milk fat when cows were fed supplemental fats was a ruminal response because intravenous infusion of FA did not depress those milk FA. In our study, contents of short- and mediumchain FA and C16:0 in milk fat decreased linearly when more FA were infused into the abomasum, which suggests that this response is not rumina\. The decreased yields of shortand medium-chain FA cannot be fully attributed to a decrease in the availability of VFA for de novo synthesis of FA by the Journal of Dairy Science Vol. 77. No.5, 1994
mammary gland. Infusion of 400 g/d of FA into the abomasum decreased DMI by only 6.8% compared with the initial FA infusion of o g/d, but the yield of short- and mediumchain FA decreased by 15.5%. Because total tract digestibility of OM is not decreased by postruminal infusion of FA (7, 11), a decrease in fermentable OM and VFA produced would not likely account fully for the decreased yields of short- and medium-chain FA. Clapperton and Banks (8) attributed 20% of the inhibition of de novo FA synthesis when fat is fed to decreased formation of VFA in the rumen. Our data indicate that postruminal administration of FA probably inhibits the enzymatic pathway for de novo FA synthesis by the mammary gland, as proposed by others (15). Figure 5A illustrates the linear relationship between C 18:1 yield in milk and the amount of C 18:1 infused into the abomasum. Regression analysis gave the equation Y
=
184.5 (SE = .70)
(~
= 7.8) + .541X (SE = .050)
where Y = yield (grams per day) of C 18:1 in milk fat and X = amount (grams per day) of C18:1 infused. Yield of C18:2 in milk fat also increased linearly with increasing infusion of C 18:2 (Figure 5B). Regression analysis resulted in the equation Y
=
22.2 (SE = .9) + .527X (SE (~ = .88)
= .027)
where Y = yield (grams per day) of C 18:2 in milk fat, and X = amount (grams per day) of C 18:2 infused. Within the range of C18:1 and C 18:2 infused in this experiment, yields of C 18 :1 and C 18:2 in milk fat responded linearly to increasing or decreasing amounts of C 18:1 and C 18 :2 infused into the abomasum. The transfer efficiencies of C 18 :1 and C 18:2 from the abomasum into milk fat, as indicated by the regression coefficients (54.1 and 52.7% for C 18:1 and C I 8:2), are within the range of transfer efficiencies reported by others (11, 24). These transfer efficiencies are underestimated to the extent that infrared analysis of milk underestimates fat content in milk enriched with unsaturated FA (28) and to the extent that
INCREASING OLEIC ACID IN MILK FAT
400
BO 70
•
B
••
,..... "'0
"0>
60 50
lC
U
•
•
'-' N
•
40
..>t. ~
30 20 10 0
20
40
60
BO
1381
TGLP fraction of plasma as percentages of the sum of these five FA. The FA C16:1 and C 18:3 were not detected in all samples; each was <2% of the total. The contents of C16:0 and C18:0 in plasma TGLP decreased linearly (Figure 6, A and B) as the amount of FA infused into the abomasum increased. Content of C14:0 decreased with infusion of HOSF (Table 5). The interaction between source of FA and amount of FA infused was significant for the content of C18:1 in TGLP (Figure 6C). Cows infused with FA from HOSF had more C18:1 in TGLP than cows infused with canola FA, most likely because HOSF contained 25 percentage units more C 18:1 than canola FA. The interaction between source of FA and amount of FA infused was significant for the content of C18:2 in TGLP (Figure 6D). The percentage of C 18:2 in TGLP was greater when cows were infused with canola FA than with HOSF FA, probably because the canola FA contained about 20 percentage units more C 18:2 than HOSF FA. The large alterations in the proportions of triglyceride FA in TGLP indicate that the epithelium of the small intestine was capable of absorbing the infused FA and incorporating them into TGLP. The large changes of FA in TGLP corresponded to changes in milk FA, indicating that increased C18:1 in milk probably was not attributable to desaturase activity in the intestine or mammary gland.
100
C 18 : 2 infused (g/d)
Figure 5. Relationship between amount of C 18 :1 or Cl8:2 infused postruminally and yields of C18:1 and Cl8:2 in milk. A) Transfer of postruminally infused C18:l into milk fat [Y = 184.5 + .541X (r2 = .70) where Y = yield (grams per day) of C 18:1 in milk fat and X = amount (grams per day) of C 18:1 infused into the abomasum]. B) Transfer of postruminally infused C 18:2 into milk fat [Y = 22.2 + .527X (r2 = .88) where Y = yield (grams per day) of C18:2 in milk fat, and X =amount (grams per day) of C18:2 infused into the abomasum].
DMI and, consequently, FA intake were decreased by increasing FA infusion (Figure 1). Triglyceride FA In Blood
Table 5 shows the contents of CI4:0, CI6:0, CI8:0, CI8:" and C18:2 in triglyceride of the
Metabolites in Blood
The effects of source of FA and the interaction of source of FA by amount of FA infused were not significant for concentrations of metabolites in plasma (data not shown). The concentration of TGLP in plasma increased linearly as increasing FA from canola or HOSF were infused and decreased when the amount of FA infused was decreased (Table 6). Coupled with the changes of FA composition of triglyceride in TGLP, the increased concentration of TGLP provides additional evidence that the supply of exogenous C 18:! was responsible for the changes in milk CI8:1' Concentration of NEFA in plasma tended (P = .08) to increase as more FA were infused postrurninally (Table 6); however, NEFA concentration was highly variable. Plasma cholesterol was affected significantly by the Journal of Dairy Science Vol. 77, No.5, 1994
1382
LaCOUNT ET AL.
amount of FA infused; however, the pattern for the effect of amount of FA infused on cholesterol concentration was not clear. The short time of infusion for each amount of FA probably increased the variability in concentrations of NEFA and cholesterol in plasma; therefore, the statisticalIy significant changes of cholesterol and NEFA may be of little biological significance. Unlike data from this study, data from other researchers have shown more consistent changes in concentrations of metabolites in blood when various fat sources were infused or fed. Postruminal infusions of rapeseed oil (14) or saturated or unsaturated FA (11) increased concentrations of NEFA and cholesterol in plasma. Plasma cholesterol and NEFA increased when cows were fed increasing amounts of canola seed (18). Rafalowski and Park (25) reported that blood cholesterol increased when cows were fed 7.3 or 11.7%, but not 4.5%, regular sunflower seeds. Differences in experimental design or number of cows may
explain differences between the results of this study and others (11, 14, 18, 25). CONCLUSIONS
The FA composition of milk can be altered by providing specific FA postruminally to midlactation dairy cows. Our data indicate that contents of short- and medium-chain FA (including C 14:0 and C I6:0) decreased, and the content of C18:1 increased, when 400 gld of FA from HOSF or canola were infused into the abomasum. Within the range (0 to 400 gld) of FA infused in our study, yields of C18:1 and C18:2 in milk were increased linearly by increasing the amount of C 18 :1 and C 18 :2 infused. Additional research will be necessary to determine whether increases in the oleic acid content of milk can be even greater if more oleic acid is provided postruminalIy and whether the response is dependent on stage of lactation (24). Content of whey protein N in milk decreased, and that of NPN increased, as
TABLE 5. Effects of amount of canola or high oleic sunflower (HOSP) fatty acids (FA) infused on percentages of FA in TGLPLtriglyceride. Amount of FA infused Item C14:0 Canola HOSF C16:0 Canola HOSF C18:0 Canola HOSF CI8:! Canola HOSF C18:2 Canola HOSF
o gld
133 gld
267 gld
3.42 3.28
3,70 3.26
2.44 2.48
3.26 1.87
26.53 26.95
23.70 22.26
20,84 20.32
45.60 44.63
37.22 35.64
16,53 16.62 7.93 8,53
133 gld
o gld
SEM
2.62 1.96
2.94 4.45
5.10 4.36
.77
20,30 17.18
19.67 19.23
23.43 26.08
31.83 27.24
2.03 2.03
33.99 34.11
29.59 29.45
33.47 35.17
40.48 35,98
34.59 43.84
2.40 2.40
25.56 30.65
29.80 35.55
34.12 45.17
31.86 35.30
22.70 25.58
19.77 17.91
1.35 1.35
9.83 8.19
12.94 7.55
12.72 6.33
12.37 8.36
10.45 7.92
8.71 6.67
.88 .88
400 gld
267 gld (% of identified FA3)
Significant effects 2
,77 L A L L A L L A, T, AxT Q Q
AxT Q
L
ITriglyceride-rich lipoprotein, 2For overall effects of model: A = amount; T = treatment; T x A = treatment by amount interaction. Within treatments: L = linear effect of increasing FA infusion from 0 to 400 gld; Q = quadratic effect of increasing FA infusion from 0 to 400 gld. 3Percentages of the sum of C14:0, C16:0, C18:0' C18:I' and CI8:2' The FA C16:1 and C18:3 were not detected in all samples and were <2% of the total FA when detected, Journal of
Dairy
Science Vol. 77, No.5. 1994
1383
INCREASING OLEIC ACID IN MILK FAT TABLE 6. Effect of amount of fatty acid (FA) infused on concentrations of metabolites in plasma. I Amount of FA infused Item TGLP-TG mgldl NEFA, /LeqIL Cholesterol, mgldl
o gld
133 gld
267 gld
400 gld
267 gld
133 gld
o gld
SEM
2.6 259
4.4 300
38 274
5.6 272
4.8 244
3.2 224
2.4 196
23
.00013 .0800
232
262
222
215
223
215
208
9
.0241
.4
P
IMeans across both treatments for each amount of FA infused; effects of treatment and the interaction of treatment and amount were nonsignificant. 2Triglyceride-rich lipoprotein and triglyceride. 3Linear effect of increasing FA infusion from 0 to 400 gld (P < ,05).
.34 .32 .30 o 28 co 26 24 22 20
48
",
I
u ~
l-
I (L
18
-.J ~
16 14
/
.32 28 24 20 16
I-
~ ~
~
:· I
/
•
\ '.
iI
/-
I~/·
\
'
\
,I
I
12
48 46
:, "
;e
•
I
18
D
16
44
14
42 o 40 ex> .38 .36 .34 .32
~
ex>
u ~
l-
I (L
30
28 26
I
.36 ex>
c
~
44
40
--I ~
I-
o
133 267 400 267 133
0
Amount of FA infused (g/d)
12
f-
10
~
8
f-
6
~
4
r-
2
o
I
o
133 267 400 267 133
0
Amount of FA infused (g/d)
Figure 6. Weight percentages (grams per 100 g) of C16:0> C18:0> CIS:I' and C 18:2 in triglyceride-rich lipoproteins (TGLP) in plasma from cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers (II). A) Weight percentage of C 16:0 in TGLP-triglyceride (TG) (SE of the difference between treatment means at any amount 2.7). Probabilities of greater F for effects in model: treatment, P = .34; amount, P = .0001; and treatment by amount, P = .67. B) Weight percentage of ClS:O in TGLP-TG (SE of the difference = 3.2). Treatment, P = .34; amount, P = .0001; and treatment by amount, P = .19. C) Weight percentage of CIS: I in TGLP-TG (SE of the difference 9.5). Treatment, P = .01; amount, P = .0001; and treatment by amount, P = .001. D) Weight percentage of C 18:2 in rGLP-TG (SE of the difference = 1.2). Treatment, P = .04; amount, P = .07; and treatment by amount, P = .007,
=
=
Journal of Dairy Science Vol. 77, No,S, 1994
1384
LaCOUNT ET AL.
FA infusion increased. Attention should be given to flavor and processing qualities of milk resulting from efforts to modify milk FA composition. ACKNOWLEDGMENTS
The authors extend their appreciation to J. P. Elliott for assistance with blood collection and to R. S. Younker, L. E. Williams, and C. R. Millsap for cow care. Special thanks are extended to HiDee Ekstrom for manuscript preparation. REFERENCES I Ashes, J. R., P. St. Vincent Welch, S. K. Gulati, T. W, Scott, and G. H. Brown. 1992. Manipulation of the fatty acid composition of milk by feeding protected canola seeds. J. Dairy Sci. 75:1090. 2 Association of Official Analytical Chemists. 1984. Official Methods of Analysis. 14th ed. AOAC, Arlington, VA. 3 Atwal, A. S., M. Hidiroglou, and 1.K.G. Kramer. 1991. Effects of feeding Protec~ and a-tocopherol on fatty acid composition and oxidative stability of cows milk. J. Dairy Sci. 74:140. 4 Bonanome, A., and S. M. Grundy. 1988. Effect of dietary stearic acid on plasma cholesterol and lipoprotein levels. New England J. Med. 318:1244. 5 Casper, D. P., D. J. Schingoethe, R. P. Middaugh, and R. J. Baer. 1988. Lactational responses of dairy cows to diets containing regular and high oleic acid sunflower seeds. 1. Dairy Sci. 71 :1267. 6 Chilliard, Y., G. Gagiiostro, 1. Flechet, 1. LeFaivre, and I. Sebastian. 1991. Duodenal rapeseed oil infusion in early and midlactation cows. 5. Milk fatty acids and adipose tissue lipogenic activities. J. Dairy Sci. 74: 1844. 7 Christensen, R. A., J. K. Drackley, D. W. LaCount, and 1. H. Clark. 1994. Infusion of four long-chain fatty acid mixtures into the abomasum of lactating dairy cows. 1. Dairy Sci. 77:1052. 8 Clapperton, J. L., and W. Banks. 1985. Factors affecting the yield of milk and its constituents, particularly fatty acids, when dairy cows consume diets containing added fat. 1. Sci. Food Agric. 36:1205. 9 Denke, M. A, and S. M. Grundy. 1992. Comparison of effects of lauric acid and palmitic acid on plasma lipids and lipoproteins. Am. 1. Clin. Nutr. 56:895. 10 DePeters, E. J., and J. P. Cant. 1992. Nutritional factors influencing the nitrogen composition of bovine milk: a review. 1. Dairy Sci. 75:2043. 11 Drackley, 1. K., T. H. Klusmeyer, A. M. Trusk, and J. H. Clark. 1992. Infusion of long-chain fatty acids varying in saturation and chain length into the abomasum of lactating dairy cows. J. Dairy Sci. 75:1517. 12 Erickson, P. S., M. R. Murphy, and 1. H. Clark. 1992. Supplementation of dairy cow diets with calcium salts Journal of Dairy Science Vol. 77, No.5, 1994
of long-chain fatty acids and nicotinic acid in early lactation. 1. Dairy Sci. 75: 1078. 13 Gagiiostro, G., and Y. Chilliard. 1991. Duodenal rapeseed oil infusion in early and midlactation cows. II. Voluntary intake, milk production, and composition. 1. Dairy Sci. 74:499. 14 Gagliostro, G., Y. Chilliard, and M. J. Davicco. 1991. Duodenal rapeseed oil infusion in early and midlactation cows. 3. Plasma hormones and mammary apparent uptake of metabolites. 1. Dairy Sci. 74: 1893. 15 Grummer, R. R. 1991. Effects of feed on the composition of milk fat. 1. Dairy Sci. 74:3244. 16 Grummer, R. R., W. L. Hurley, C. L. Davis, and C. A. Meacham. 1986. Effect of isolation temperature on the determination of bovine plasma very low density lipoprotein concentrations. 1. Dairy Sci. 69:2083. 17 International Dairy Federation. 1964. Determination of the casein content of milk. Int. Dairy Fed. Stand. No. 29, Int. Dairy Fed., Brussels, Belgium. 18 Khorasani, G. R., G. DeBoer, P. H. Robinson, and J. J. Kennelly. 1992. Effect of canola fat on ruminal and total tract digestion, plasma hormones, and metabolites in lactating dairy cows. J. Dairy Sci. 75:492. 19 Khorasani, G. R., P. H. Robinson, G. DeBoer, and 1. J. Kennelly. 1991. Influence of canola fat on yield, fat percentage, fatty acid profIle, and nitrogen fractions in Holstein milk. J, Dairy Sci. 74:1904. 20 Mackie, R. I., B. A. White, and M. P. Bryant. 1991. Lipid metabolism in anaerobic ecosystems. Crit. Rev. Microbiol. 17:449. 21 Mensink, R. P., and M. B. Kattan. 1990. Effect of dietary trans fatty acids on high-density and lowdensity lipoprotein cholesterol levels in healthy subjects. New England J. Med. 323:439. 22 Murphy, J. 1., G. P. McNeill, J. F. Connolly, and P. A. Gleeson. 1990. Effect on cow performance and milk fat composition of including full fat soyabeans and rapeseeds in the concentrate mixture for lactating dairy cows. 1. Dairy Res. 57:295. 23 O'Donnell, J. A. 1989. Milk fat technologies and markets: summary of the Wisconsin Milk Marketing Board 1988 Milk Fat Roundtable. 1. Dairy Sci. 72: 3109. 24 Palmquist, D. L., A. D. Beaulieu, and D. M. Barbano. 1993, Feed and animal factors influencing milk fat composition. 1. Dairy Sci. 76:1753. 25 Rafalowski, W., and C. S. Park. 1982. Whole sunflower seed as a fat supplement for lactating cows. J. Dairy Sci. 65:1484. 26 SAS~ User's Guide: Statistics, Version 5 Edition. 1985. SAS Inst., Inc., Cary, NC. 27 Schauff, D. 1., J. P. Elliott, J. H. Clark, and J. K. Drackley. 1992. Effects of feeding lactating dairy cows diets containing whole soybeans and tallow. 1. Dairy Sci. 75:1923. 28 Stegeman, G. A, R. 1. Baer, D. 1. Schingoethe, and D. P. Casper. 1991. Influence of milk fat higher in unsaturated fatty acids on the accuracy of milk fat analyses by the mid-infrared spectroscopic method. J. Food Prot. 54:890. 29 Storry, J. E., A. J. Hall, B. Tuckley, and D. Millard. 1969. The effects of intravenous infusions of cod-liver and soya-bean oils on the secretion of milk fat in the cow. Br. J. Nutr. 23:173. 30 Storry, 1. E., B. Tuckley, and A J. Hall. 1969. The
INCREASING OLEIC ACID IN MILK FAT effects of intravenous infusions of triglycerides on the secretion of milk fat in the cow. Br. J. Nutr. 23:157. 31 Sukhija, P. S., and D. L Palmquist. 1988. Rapid method for determination of total fatty acid content and composition of feedstuffs and feces. J. Agric. Food Chern. 36:1202. 32 Van Soest, P. J., J. B. Robertson, and B. A. Lewis.
1385
1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74:3583. 33 Wiesen. B., R. L. Kincaid, J. K. Hillers. and J. H. Harrison. 1990. The use of rapeseed screenings in diets for lactating cows and subsequent effects on milk yield and composition. 1. Dairy Sci. 73:3555.
Journal of Dairy Science Vol. 77, No.5, 1994
between yield of C18:1 in milk fat and the amount of C18:1 infused into the abomasum was linear, and transfer efficiency was 54.1 %. Increased concentration and yield of C 18:1 in milk were attributable mostly to the increased exogenous supply of C I8:1' (Key words: fatty acid, milk fat, oleic acid, postruminal infusion)
The feasibility of dietary strategies to increase oleic acid content of milk fat is unclear. Four Holstein cows were infused abomasally with free long-chain fatty acids from canola (62.5% C I8:1) or high oleic sunflowers (86% CI8:1)' Each fatty acid mixture was infused for 3 d at 0, 133,267,400,267, 133, and 0 gld for a total of 21 d; cows then were changed to the opposite fatty acid mixture, and the infusion sequence was repeated. The DMI and percentages of casein and whey N in milk were decreased by infusion of fatty acids, but milk yield and percentages of fat and NPN in milk were increased. Contents of short- and mediumchain fatty acids and C16:0 in milk fat decreased, and contents of CI8:), CI8:2, and C18:3 increased, as fatty acid infusion increased. Contents of C16:0 and C18:0 in plasma triglyceride were decreased, and content of C18:1 was increased, by increasing infusion of fatty acids. All changes reversed when the amount of fatty acid infused was decreased. Within the range of amounts of fatty acids infused, the relationship
Abbreviation key: FA = fatty acids, HOSF = high oleic sunflower, TGLP = triglyceride-rich lipoproteins. INTRODUCTION
Received September I, 1993. Accepted January 7, 1994. ISupported by Hatch funds appropriated to the lllinois Agricultural Experiment Station (Projects 35-0352 and 35-0359) and by a gift from Milk Specialties Co., Dundee, lL. 2 Address correspondence and reprint requests to James K. Drackley, Department of Animal Sciences, 260 Animal Sciences Lab, University of Illinois, 1207 West Gregory Drive, Urbana, lL 61801. 3Supported by the Jonathan Baldwin Turner Undergraduate Research/Scholarship program, College of Agriculture, University of llIinois. 4Present address: Rolling Lawn Dairy Fana, RR 3, Box 135, Greenville, lL 62246.
1994 J Dairy Sci 77:1372-1385
During the past decade, Americans have become increasingly health conscious. Reports in the media have highlighted recommendations from the medical community and government agencies that people decrease their intake of saturated fatty acids (FA) because of the purported relationship of some saturated FA with atherosclerosis. In particular, the FA C14:0 and C 16:0 may increase plasma cholesterol (9); however, C 18:0 may not affect cholesterol (4). Dairy products have been a target for human health advocates because milk fat contains, on average, about 70% saturated FA, 25% monounsaturated FA, and 5% polyunsaturated FA (15). Over one-half of the saturated FA in milk are C 14:0 and C 16:0 (15). In 1988, the Wisconsin Milk Marketing Board assembled a group of 15 researchers from industry and academia to discuss the potential future uses of milk fat. These researchers concluded that the "ideal" milk fat for use in the human diet would contain <10% polyunsaturated FA, ~8% saturated FA, and ~82% monounsaturated FA (23). Grummer (15) concluded that such a large change in the FA composition of milk is not possible by alteration of the diet of dairy cows. Although manipulation of the diet of dairy cows may not yield the ideal FA profile in
1372
INCREASING OLEIC ACID IN MILK FAT milk, the composition of milk fat potentially can be altered by addition of fat to the diet. Inclusion of canola (1, 3, 19, 22, 33) or regular sunflower (5, 25) seeds in diets of lactating dairy cows decreased the concentrations of short- and medium-chain FA and increased the concentration of C I 8:1 in milk fat. Canola oil is 55 to 60% C 18: 1; regular sunflower oil is 65 to 70% CI8:2' When either of these oil seeds is fed, similar increases occur in the C I8 :1 content of milk; however, when regular sunflower seeds are fed, much of the resulting increase OfCI8:1 in milk fat is in the trans11 configuration because of incomplete biohydrogenation of CI8:2 (5, 20). In contrast, when high oleic sunflower (HOSF) seeds were fed, increases of trans-CI8:1 were smaller (5). Increased trans monounsaturated FA in the human diet may not be desirable because high intakes of trans FA increase total and low density lipoprotein cholesterol and decrease high density lipoprotein cholesterol compared with a diet high in cis FA (21). Technology to increase bypass of FA from sources such as canola and HOSF that are high in cis monounsaturated FA may more effectively alter the FA composition of milk than when oil seeds are fed. Milk ciS-C I8 :1 also can be derived from desaturation of C18:0 in the intestine or mammary gland (15). Few data are available on the responsiveness of C18:1 in milk to different amounts of CI8:1 absorbed from the intestine. Our laboratory has developed a technique to infuse free FA into the abomasum of dairy cows (11). The primary objective of this study was to determine the potential to alter the content of C I 8:1 and other FA in milk by abomasal infusion of different amounts of two FA mixtures (derived from canola or HOSp) rich in CI8:I' Previous experiments in our laboratory (7, 11) demonstrated that abomasal infusion of unsaturated FA decreased DMI and milk yield compared with saturated FA and that milk protein percentage tended to be decreased more by unsaturated than by saturated FA (7). Potential changes in content or composition of milk N fractions related to specific postruminal FA would be important for nutritional value and processing characteristics of milk (10). Therefore, a secondary objective was to determine the effects of FA infusions on yield and composition of N fractions in milk.
1373
TABLE 1. Ingredient and nutrient composition of TMR. Composition
(%
Ingredient Alfalfa haylage Corn silage Soybean hulls Ground shelled corn Soybean meal (48% CP) Sodium chloride Mineral and vitamin mixture l Limestone Dicalcium phosphate Magnesium oxide Sodium bicarbonate Sodium sulfate Nutrient2 DM OM CP
of
DM)
30.00 20.00 6.00 29.50 11.50 .30 .15
.80 .60 .20 .75 .20 70.6 90.9 17.0
21.0 33.2
ADF NDF
Total FA3
3.1
lContained .004% Co, .5% Cu, .025% I, 2.0% Fe, 5.0% Mg, 3.0% Mn, 7.5% K, .015% Se, 10.0% S, 3.0% 20, 2200 IU/g of vitamin A, 660 IUlg of vitamin D3' and 8 IU/g of vitamin E. 2As analyzed. 3Fatty acids.
MATERIALS AND METHODS Cows, Experimental Design, end Diet
Four multiparous Holstein cows, fitted with ruminal cannulas and averaging 166 DIM (range 129 to 182 d), were used in a singlereversal design with 21-d periods. Cows were housed in individual stanchions, milked twice daily, and fed for ad libitum intake a TMR (Table 1) twice daily. The diet contained 17% CP, 21% ADF, 33% NDF, and 3.1% total FA (Table 1). Treatments were continuous abomasal infusions of free FA prepared from HOSF or canola oils (Henkel Corp., Emery Division, Cincinnati, OH). Two cows were infused with each FA mixture during each 21-d period. The FA were infused for 3 d at each of four amounts in a sequence of 0, 133, 267,400,267, 133, and 0 gld (i.e., a total of 21 d for each FA mixture); cows then were changed to the opposite FA mixture, and the infusion sequence was repeated. The largest amount of FA infused was 400 gld because previous work in our laboratory (11) indicated that postruminal infusion of >450 gld of unJournal of
Dairy
Science Vol. 77, No.5. 1994
1374
LaCOUNT ET AL.
TABLE 2. Composition of fatty acid (FA) mixtures and TMR. FA
High oleic sunflower
Canola
TMR
(glIOO g of FA) CI2:0 C 14 :0 C!6:0 C 16 :1 C 18 :0 C I8:! C18:2 C18:3
NDI .1 4.5 .4
1.2 62.5 24.1 7.2
ND ND 5.0 .2 3.5 86.0 5.3 ND
2.1 .3 22.1 2.0 3.4 17.0 44.3 8.8
INot detected.
saturated FA caused cows to go off feed, to decrease milk yield, and to develop diarrhea. The composition of the two FA mixtures that were infused and of the TMR is given in Table 2. The major FA in canola were C18:1 (62.5%) and C 18:2 (24.1%). The HOSF FA contained 86.0% C I8 :! and only small quantities of other FA.
An infusion tube was placed into the abomasum through the ruminal cannula (II). Daily infusions consisted of the FA complexed with 180 g of meat solubles (Milk Specialties Co., Dundee, IL) or 180 g of meat solubles alone in 9 L of water. To prepare the solutions for infusion, the meat solubles were dissolved in hot (50 to 60°C) tap water. The FA were heated in an oven at 150°C until they were fully melted and then were added slowly to the solution of meat solubles with continuous stirring. Suspensions were stirred continuously on magnetic stir plates while being infused into the abomasum over 20 to 22 hid with peristaltic pumps (Harvard Apparatus, South Natick, MA). Feed intake was measured daily. Feeds and orts were sampled every 3 d and composited within each period. Samples of feeds were dried at 55°C, ground, and analyzed for contents of OM, ash (600°C for 8 h), CP (2), NDF by using a-amylase (32), ADF (32), and FA (31). Milk Yield and Composition
Milk yield was measured and recorded at each milking. Milk samples were taken from two consecutive milkings every 3 d so that a Journal of Dairy Science Vol. 77, No.5, 1994
sample was obtained during infusion of each amount of FA. Milk samples were preserved with 2-bromo-2-nitropropane-I,3-diol, composited into daily samples in proportion to milk production, and analyzed for contents of CP, true protein, and fat by mid-infrared spectrophotometric (28) analysis (New York DHIA Laboratory, Ithaca, NY). Fat was measured using the A filter (28). Noncasein N was determined by Kjeldahl analysis of the filtrate after precipitation with 10% acetic acid and IN sodium acetate (17). Casein N was calculated as the difference between total Nand noncasein N, and NPN was calculated as the difference between total N and true protein N. A portion of the daily composite milk sample was placed into a test tube and allowed to stand overnight in a cold room at 4°C. The cream layer was transferred into a clean test tube, and FA were methylated and quantified by GLC using the procedures of Sukhija and Palmquist (31), except that an external standard rather than an internal standard was used to quantify the proportions and yields of FA in milk (27). Glycerol in milk was calculated as described by Schauff et al. (27). Blood Sampling and Analyses
On d 3 of infusion of each amount of FA, blood v.as sampled from a coccygeal vein at 1300 h, which was 5 h after the infusions began in the morning. Concentrations of NEFA (27) and total cholesterol (Sigma kit number 352; Sigma Chemical Co., St. Louis, MO) were measured in plasma. An additional 60 ml of blood were obtained for separation of triglyceride-rich lipoproteins (fGLP) by ultracentrifugation (7, 16), and the concentration of triglyceride in TGLP was determined (Sigma kit number 339; Sigma Chemic~l Co). Fatty acid composition of triglycendes In the TGLP fraction was determined after TLC. One milliliter of the TGLP fraction was extracted with 20 ml of chloroform-methanol (2: 1, voVvol), which contained an internal standard (triheptadecanoin, .01 mg/ml) and an antioxidant (butylated hydroxytoluene, .2 mg/ml). Four milliliters of .74% (wtJvol) KCl were added to the tube, and the tube was mixed and centrifuged (100 x g) for 10 min. The ~ueous layer was aspirated and discarded. The Sides of the tube and the surface of the chloroform
1375
INCREASING OLEIC ACID IN MILK FAT
layer were washed three times with 2 ml of .74% (wt/vol) KCl. The chloroform fraction was evaporated under air, and the dry extract was redissolved immediately in .4 ml of 2: I (volJvol) chloroform-methanol. Aliquots (100 1-'1) of lipid extract were spotted on TI...C plates (silica gel G with preadsorbent zone, channeled Analtech Uniplates™; Alltech Associates, Deerfield, ll...). Plates were developed twice in methanol to 1.5 cm above the preadsorbent line, air dried, and then developed to 1.5 cm above the preadsorbent line in chloroform: methanol (1:1, volJvol). After air drying, plates were fully developed using hexane:diethyl ether:acetic acid (80:20:1, volJvolJvol). Lipid bands were visualized under UV light after the plates were sprayed with rhodamine-6G (.02% in 95% ethanol). The triglyceride band was identified by comparison with standards in adjacent lanes and scraped into screw-capped tubes. The FA were methylated and quantified by GLC using the procedure of Sukhija and Palmquist (31).
26 25 24
,..... -0
23
......... Ol
.:,{. ......,.. 22
:::::!: 0
\" . /. . ~ '>\/~ •
21 20 19 18
L-----'_----'-_-L_-l-_-'--_L-----'_-'
o
'33 267 400 267
133
0
Amount of FA infused (g/d)
Figure I. The DMI for cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers ~) (SE of the difference between treatment means at any amount = .75). Probabilities of greater F for effects in model: treatment, P = .88; amount, P = .03; and treatment by amount, P = .05.
Statistical Analyses
Data were subjected to ANOVA for a splitplot, single-reversal design using the general linear models procedure of SAS (26). The model contained effects of sequence (1 df), cow within sequence (2 df), period (1 df), FA source (i.e., treatment, I df), whole-plot error (2 df), FA amount (6 df), and the interaction of FA source and amount (6 df). Effects of FA source and period were tested using the wholeplot error term, and effects of FA amount and the interaction of source of FA by amount of FA infused were tested using the residual error. Linear and quadratic effects of increasing the amount of FA infused from 0 to 400 gld were tested (within treatments, if effects of treatment or the interaction of treatment by amount were significant). The second period was decreased to 18 d, and FA were infused in the sequence of 0, 133, 267, 400, 115, and 0 gI d for both FA sources because of a limited amount of FA from HOSF. Missing values were assumed for the second infusion of 267 gld, and the amount of 115 gld replaced the amount of 133 gld used in the first period. Least squares means are presented throughout. Significance was declared at P :s; .05.
RESULTS AND DISCUSSION
OMI, Milk Yield, and Milk Composition
The interaction of source of FA by amount of FA infused for OMI was significant because of a linear decrease of DMI when cows were infused with increasing amounts of FA from canola (Figure 1). The DMI increased as the amount of canola FA in the infusate was decreased, but OMI tended to decrease as the amount of FA from HOSF infused was decreased. Milk yield increased linearly when cows were infused with FA from HOSF (Figure 2). Peak milk yield seemed to occur 3 d after the infusion of 400 gld of FA, which may have been a consequence of the short infusions at each amount of FA. The design and number of cows used did not allow accurate determination of effects of infused FA on OMI and milk yield; longer term studies with more cows are needed to evaluate these responses. Duodenal infusion of 1100 gld of rapeseed oil decreased OMI (14), and abomasal infusions of 450 gld of canola or regular sunflower FA (7) or 450 gld of soybean FA (11) decreased OMI and milk yield. Journal of Dairy Science Vol. 77. No.5. 1994
1376
LaCOUNT ET AL.
.36
Similar to milk yield, the N fractions of milk seemed to have a delayed response to the FA infusions (i.e., changes were not maximal until 3 d after the infusion of 400 gld of FA). True protein N and whey protein N expressed as percentages of total N in milk decreased quadratically, and NPN as a percentage of total N increased quadratically, as increasing amounts of FA were infused (Table 3). These changes
.35 ~
, , , , ,. ,, , ,
.34
.---.. .33 "U
"(J) .:c£ ......,
32
.:c£
:::;;;
.31 30 29
4.2
A 28
'------'---'------'-_-'--_--L_---'---_....L-----.J
o
133
267
400
267
133
4.0
0
Amount of FA infused (g/d)
........ 3.8 Figure 2. Milk yield for cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers (II) (SE of the difference between treatment means at any amount = .99). Probabilities of greater F for effects in model: treatment, P = .67; amount, P = .01; and treatment by amount, P = .14.
~
'-'
....0 ....
3.6
.::£
:::;;;
3.4 3.2
The percentage of fat in milk was affected significantly by the amount of FA infused (Figure 3A), primarily because of a decrease when the amount of infused FA was decreased. Yield of milk fat increased when more FA were infused (Figure 3B). When cows were infused with canola FA, the increased fat yield was attributable mostly to a tendency (P = .15) for linearly increased fat percentage, but for HOSF infusion the increased fat yield was from greater milk yield. Similar to this study, postruminal infusion of 1100 gld of rapeseed oil increased milk fat percentage slightly in midlactation dairy cows (13). Abomasal infusion of highly saturated or soybean FA did not alter milk fat percentage (11). Percentages and yields of total N, true protein N, casein N, and whey protein N in milk were unaffected (P > .15) by source of FA (data not shown). Percentage of whey protein N in milk decreased quadratically, and the percentage of NPN increased quadratically, as increasing amounts of FA were infused (Table 3). Yields of total N, true protein N, and casein N were unaffected by the amount of FA infused. The yield of NPN increased quadratically as the amount of FA infused increased. Journal of Dairy Science Vol. 77, No.5, 1994
3.0 1300
B
1250
........
1200
"U
"(J)
1150
'-'
....0 .... .::£
1100 1050
,. ,·~·/\ . " .d •v· "
,,
,, ,
:::;;;
•
1000 950 900
'------'-_-'-_-'-_"-----'_---'-_....L----O
o
133 267 400 267 133
0
Amount of FA infused (g/d) Figure 3. Percentage and yield of milk fat from cows infused postruminally with fatty acids (FA) from canola Ce ) or high oleic sunflowers (II). A) Milk fat percentage CSE of the difference between treatment means at any amount = .13). Probabilities of greater F for effects in model: treatment, P = .06; amount, P = .0001; and treatment by amount, P = .66. B) Milk fat yield; standard error of the difference = .05. Treatment, P = .86; amount, P = .0004; and treatment by amount, P = .43.
1377
INCREASING OLEIC ACID IN MILK FAT TABLE 3. Effect of amount of fatty acid (FA) infused on N components of milk. l Amount of FA infused Item
o g/d
133 g/d
267 g/d
400 g/d
267 g/d
133 g/d
o g/d
SEM
.519 156.4
.517 153.8
.522 161.4
.509 158.0
.496 166.1
.509 160.8
.515 156.5
2,83
.0453 .2169
.486 146.4 93.7
.485 143.8 93.7
.488 150.9 93.5
.471 146.5 92.5
.459 153.4 92.5
.472 149.1 92.8
.481 145.8 93.2
.0045 2.63 .17
.0067 .3351 ,00012
.401 120.8 77.2
398 118.5 76.8
.403 124.5 77.2
.394 122.1 77.2
.380 127.0 76.4
.391 123.4 76.6
.395 119.8 76.5
.0039 2.28 .20
.0357 .3102 .0395
.085 25.7 16.5
.087 25.7 16.9
.085 26.2 16.4
.078 24.1 15.3
.080 26.6 16.2
.082 25.8 16.2
086 26.0 16.8
.0016 .53 .27
.0026 2 .1155 2 .0085 2
.033 9.9 6.3
.033 9.8 6.3
.034 10.6 6.5
.038 11.8 7.5
.037 12.5 7.5
.036 11.6 7.2
.034 10.6 6.8
.0009 .31 .17
.00102 .00012 .00012
P
Total N %
gld True protein N %
g/d % of Total N Casein N %
g/d % of Total N Whey protein N %
g/d % of Total N NPN %
gld % of Total N
.0044
'Means across both treatments for each amount of FA infused; effects of treatment and the interaction of treatment and amount were nonsignificant. 2Quadratic effect of increasing FA infusion from 0 to 400 gld (J> < .05).
are similar to those often observed when supplemental fats are fed to cows (10). In other studies, milk CP was decreased (7, 13) or unaffected (11) by postruminal infusion of FA. FA Composition and Yields in Milk
Infusion of increasing amounts of FA from canola or HOSF into the abomasum linearly decreased contents of CIO:O, CI4:0' C 14 : b CI5:0' CI6:0, CI6:J, and C 17:o in milk fat (Table 4). Infusion of FA from HOSF, but not canola, linearly decreased content of C8:0. The decrease of content of C12:0 in milk was linear for canola and quadratic for HOSF FA. The decreases in the contents of total short- and medium-chain FA (y,:o to CI4:0) and C16:0 in milk fat are shown in Figure 4, A and B, respectively. Infusion of FA from canola or HOSF did not affect the contents of C4:0' C6:0, or C 18:0 in milk fat. The content of glycerol decreased linearly as the amount of FA from HOSF infused increased, as a consequence of the increase in long-chain FA and decrease in short- and medium-chain FA attached to the glycerol moiety.
The interaction between source of FA and amount of FA infused was significant for content of C18:1 in milk because it increased more when cows were infused with FA from HOSF than from canola (Figure 4C). Significant interactions also occurred for contents of C18:2 (Figure 4D) and C18:3 because concentrations of these FA in milk increased more when cows were infused with FA from canola than from HOSF. Differences in contents of CI8:J, CI8:2, and C18:3 in milk between FA sources correspond to the differences in composition of the FA sources (Table 2). The interaction between source of FA and amount of FA infused was not significant for the contents of other FA or glycerol in milk. The changes in contents of FA in milk that accompanied the infusion of increasing amounts of FA reversed when the amount of FA infused was decreased. These data indicate that changes in milk fat composition can be made that might be beneficial for human nutrition (4, 9, 23), including decreased contents of C14:0 and CI6:0' increased content of C18:b and unchanged content of C I 8:0' The yields of FA and glycerol in milk (data not shown) generally followed patterns similar Journal of Dairy Science Vol. 77, No.5, 1994
1378
LaCOUNT ET AL.
TABLE 4. Effects of amount of canola or high oleic sunflower (HOSF) fatty acids (FA) infused on weight percentages of FA and glycerol in milk. Amount of FA infused Item
o gld
133 gld
267 gld
400 gld
267 gld
133 gld
o gld
SEM
Significant effects l
(gil()() g of fat) C4:0 Canola HOSF
3.84 4.00
3.81 4.14
4.16 4.05
4.32 3.97
3.15 3.95
3.82 3.60
4.04 3.69
.21 .21
C6:0 Canola HOSF
2.54 266
2.59 2.70
2.96 2.61
2.70 2.44
2.27 2.58
2.58 2.39
2.82 2.65
.12 .12
C8:0 Canola HOSF
1.31 1.42
1.38 1.36
1.43 1.31
1.32 1.25
1.28 1.26
1.33 1.29
1.39 1.42
.04 .04
L
C1O:0 Canola HOSF
3.00 3.15
3.04 2.88
2.95 2.71
2.68 2.56
2.74 2.55
2.98 2.80
3.10 3.25
.08 .08
A L L
C12:0 Canala HOSF
3.78 3.98
3.72 3.50
3.38 3.16
3.10 3.00
3.27 3.07
3.67 3.46
3.88 4.09
.11 .11
C14:0 Canala HOSF
A
115 11.7
11.1 10.8
10.1 10.2
9.5 9.8
10.2 9.9
11.0 10.9
11.6 11.8
A L Q
.16 .16
A L L
C14:1 Canala HOSF
1.73 1.77
153 157
1.29 1.47
1.10 1.36
1.35 1.42
1.63 1.67
1.80 1.88
.04 .04
A L L
C15:0 Canala HOSF
1.26 1.15
1.12 1.11
.96 1.05
.87 .95
.99 1.05
1.14 1.19
1.24 1.26
.04 .04
A L L
C16:0 Canola HOSF
28.52 28.70
25.54 25.93
21.93 22.71
20.07 20.68
22.96 21.40
24.59 24.52
27.43 27.46
.42 .42
A L L
C16:1 Canola HOSF
1.85 1.86
1.61 1.70
1.39 1.52
1.27 1.42
1.45 1.44
1.63 1.66
1.81 1.87
.04 .04
A L L
C17:0 Canola HOSF
.70 .70
.68 .65
.61 .60
.58 .56
.63 60
.66 .64
.68 .68
.01 .01
A L L
C18:0 Canola HOSF
7.10 6.81
7.12 6.85
6.91 6.84
6.83 6.22
7.47 7.10
7.25 7.16
7.03 6.81
.25 .25
C18:1 Canola HOSF
17.54 16.59
20.07 21.35
22.94 26.60
25.42 3Q.42
2446 28.29
21.03 23.63
17.52 17.87
.49 .49
C18:2 Canala HOSF
2.33 2.27
3.33 2.28
4.92 2.28
6.13 2.52
4.74 2.30
3.50 2.12
2.27 2.22
.12 .12
C18:3 Canola HOSF
.30 .41
.69 .38
1.33 .27
1.38 .32
.71
.43
.51 .40
.47 .31
.14 .14
Glycerol Canala HOSF
12.77 12.87
12.72 12.81
12.79 12.68
12.71 12.57
12.38 12.60
12.71 12.59
12.89 12.79
.08 .08
A, T, TxA L L A, T x A L A, T x A L A
=
=
IFor overall effects of model: A amount; T treatment; and T x A treatments: L linear effect of increasing FA infusion from 0 to 400 gld; Q from 0 to 400 g/d.
=
Journal of Dairy Science Vol. 77. No.5. 1994
L
= treatment by amount interaction. Within =quadratic effect of increasing FA infusion
1379
INCREASING OLEIC ACID IN MILK FAT
to changes in concentrations, except that yields of Cs: o and C IO :O were unaffected by amounts of FA infused. De novo synthesis of these FA was unaffected by infusion of FA despite decreases of concentrations of Cs:o and C10:0 in milk. As more FA were infused, greater decreases of concentrations and yields of FA that are synthesized de novo occurred as the chain length increased from C6:0 to CI4:0' This pattern of changes is similar to the effect of mobilization of long-chain FA from adipose
tissue on de novo FA synthesis in the mammary gland during early lactation (24). In other experiments, postruminal infusion of FA or supplementation of FA to the diet also changed the FA profile of milk in a manner similar to this study. Abomasal infusion of 450 gld of saturated or unsaturated FA decreased contents of short- and medium-chain FA and C 16:0 in milk fat but increased contents of unsaturated CIS FA in milk (11). Chilliard et al. (6) reported decreased contents of CI4:0, C I4:], C I5 :0, C I6 :0, and C16:1 and increased contents
34
A o
24
~
28
:ii
22
fI
16
I
I
I
I
I
26
to
u
24
18
,;
\,~
,'.
e,\
V
18 16
t
14
I
\"
~
"
8
"\'\
D _
7 6
.
/-
"~~:/
22 20
f-
~
B
.30 '-
o
,
'
20
I
."/------- " •
I
.y.
32
28
24
c
,~ . .
I
u... 22
20 .... 18
30
......... 26
to
U
~
.....
~
r-
:.j.
o .....o
,......
32
5 N lC
4
3
2
-
I-
o
I
o
133 267 400 267 133
0
Amount of FA infused (g/d)
o
133267400267 133
0
Amount of FA infused (g/d)
Figure 4. Weight percentages (grams per 100 g) of C6 to C I4 :O' C 16:0' CI8:1. and C 18:2 in the fat of milk from cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers (II). A) Weight percentage of C6 to C14:0 in milk fat (SE of the difference between treatment means at any amount = .5). Probabilities of greater F for effects in model: treatment, P .47; amount, P .0001; and treatment by amount. P .40. B) Weight percentage of C 16 :0 in milk fat (SE of the difference .6). Treatment, P .S7; amount, P = .0001; and treatment by amount, P .47. C) Weight percentage of C18:1 in milk fat (SE of the difference .7). Treatment, P .04; amount, P = .0001; treatment by amount, P .0001. D) Weight percentage of C18:2 in milk fat (SE of the difference .2). Treatment, P .001; amount, P = .0001; and treatment by amount, P .0001.
=
=
=
=
=
=
=
=
= =
=
=
Journal of Dairy Science Vol. 77, No.5,
1994
1380
LaCOUNT ET AL.
of C I 8:!> C18:2, and C18:3 in milk fat when midlactation cows were infused with 1100 g1d of rapeseed oil into the duodenum. Supplementing canola (I, 19, 20) or sunflower seeds (5, 25) to the diet of dairy cows decreased the contents of short- and mediumchain FA and increased the content of C 18:1 in milk. Inclusion of regular sunflower seeds in the diet (5) increased the C18:1 content of milk fat by 16 percentage units. Because regular sunflower seeds contain predominantly C I 8:2, the increased content of C18:1 was attributable in part to incomplete biohydrogenation of C I8 :2; 8.1% of the milk FA were in the transC 18:1 configuration (5). In contrast, the same amount of HOSF seeds (5) increased C 18:1 by 12 percentage units, but only 4.3% of the FA in milk was in the trans-C I 8:1 configuration. Decreases in contents of short- and medium-chain FA and C 16:0 in milk when more FA were infused occurred more rapidly than the increases when less FA were infused (Figure 4, A and B; Table 4). De novo synthesis of short- and medium-chain FA by the mammary gland seems to have been inhibited almost immediately by infusion of increasing amounts of FA. Alleviation of this inhibition evidently was not as rapid when the amount of FA infused was decreased. The rapid decrease of short- and medium-chain FA and C 16:0 in milk fat caused by infusion of more FA and the subsequent slower increase of these FA in milk as FA were removed from the infusate may explain the similar pattern of changes in milk fat percentage (Figure 3A). The decrease in milk fat percentage was similar when Ca salts of long-chain FA were removed from the diet (12); the authors speculated that de novo synthesis of FA was slow to recover after inhibition by dietary fat (12). Other researchers (29, 30) proposed that the decrease in short- and medium-chain FA in milk fat when cows were fed supplemental fats was a ruminal response because intravenous infusion of FA did not depress those milk FA. In our study, contents of short- and mediumchain FA and C16:0 in milk fat decreased linearly when more FA were infused into the abomasum, which suggests that this response is not rumina\. The decreased yields of shortand medium-chain FA cannot be fully attributed to a decrease in the availability of VFA for de novo synthesis of FA by the Journal of Dairy Science Vol. 77. No.5, 1994
mammary gland. Infusion of 400 g/d of FA into the abomasum decreased DMI by only 6.8% compared with the initial FA infusion of o g/d, but the yield of short- and mediumchain FA decreased by 15.5%. Because total tract digestibility of OM is not decreased by postruminal infusion of FA (7, 11), a decrease in fermentable OM and VFA produced would not likely account fully for the decreased yields of short- and medium-chain FA. Clapperton and Banks (8) attributed 20% of the inhibition of de novo FA synthesis when fat is fed to decreased formation of VFA in the rumen. Our data indicate that postruminal administration of FA probably inhibits the enzymatic pathway for de novo FA synthesis by the mammary gland, as proposed by others (15). Figure 5A illustrates the linear relationship between C 18:1 yield in milk and the amount of C 18:1 infused into the abomasum. Regression analysis gave the equation Y
=
184.5 (SE = .70)
(~
= 7.8) + .541X (SE = .050)
where Y = yield (grams per day) of C 18:1 in milk fat and X = amount (grams per day) of C18:1 infused. Yield of C18:2 in milk fat also increased linearly with increasing infusion of C 18:2 (Figure 5B). Regression analysis resulted in the equation Y
=
22.2 (SE = .9) + .527X (SE (~ = .88)
= .027)
where Y = yield (grams per day) of C 18:2 in milk fat, and X = amount (grams per day) of C 18:2 infused. Within the range of C18:1 and C 18:2 infused in this experiment, yields of C 18 :1 and C 18:2 in milk fat responded linearly to increasing or decreasing amounts of C 18:1 and C 18 :2 infused into the abomasum. The transfer efficiencies of C 18 :1 and C 18:2 from the abomasum into milk fat, as indicated by the regression coefficients (54.1 and 52.7% for C 18:1 and C I 8:2), are within the range of transfer efficiencies reported by others (11, 24). These transfer efficiencies are underestimated to the extent that infrared analysis of milk underestimates fat content in milk enriched with unsaturated FA (28) and to the extent that
INCREASING OLEIC ACID IN MILK FAT
400
BO 70
•
B
••
,..... "'0
"0>
60 50
lC
U
•
•
'-' N
•
40
..>t. ~
30 20 10 0
20
40
60
BO
1381
TGLP fraction of plasma as percentages of the sum of these five FA. The FA C16:1 and C 18:3 were not detected in all samples; each was <2% of the total. The contents of C16:0 and C18:0 in plasma TGLP decreased linearly (Figure 6, A and B) as the amount of FA infused into the abomasum increased. Content of C14:0 decreased with infusion of HOSF (Table 5). The interaction between source of FA and amount of FA infused was significant for the content of C18:1 in TGLP (Figure 6C). Cows infused with FA from HOSF had more C18:1 in TGLP than cows infused with canola FA, most likely because HOSF contained 25 percentage units more C 18:1 than canola FA. The interaction between source of FA and amount of FA infused was significant for the content of C18:2 in TGLP (Figure 6D). The percentage of C 18:2 in TGLP was greater when cows were infused with canola FA than with HOSF FA, probably because the canola FA contained about 20 percentage units more C 18:2 than HOSF FA. The large alterations in the proportions of triglyceride FA in TGLP indicate that the epithelium of the small intestine was capable of absorbing the infused FA and incorporating them into TGLP. The large changes of FA in TGLP corresponded to changes in milk FA, indicating that increased C18:1 in milk probably was not attributable to desaturase activity in the intestine or mammary gland.
100
C 18 : 2 infused (g/d)
Figure 5. Relationship between amount of C 18 :1 or Cl8:2 infused postruminally and yields of C18:1 and Cl8:2 in milk. A) Transfer of postruminally infused C18:l into milk fat [Y = 184.5 + .541X (r2 = .70) where Y = yield (grams per day) of C 18:1 in milk fat and X = amount (grams per day) of C 18:1 infused into the abomasum]. B) Transfer of postruminally infused C 18:2 into milk fat [Y = 22.2 + .527X (r2 = .88) where Y = yield (grams per day) of C18:2 in milk fat, and X =amount (grams per day) of C18:2 infused into the abomasum].
DMI and, consequently, FA intake were decreased by increasing FA infusion (Figure 1). Triglyceride FA In Blood
Table 5 shows the contents of CI4:0, CI6:0, CI8:0, CI8:" and C18:2 in triglyceride of the
Metabolites in Blood
The effects of source of FA and the interaction of source of FA by amount of FA infused were not significant for concentrations of metabolites in plasma (data not shown). The concentration of TGLP in plasma increased linearly as increasing FA from canola or HOSF were infused and decreased when the amount of FA infused was decreased (Table 6). Coupled with the changes of FA composition of triglyceride in TGLP, the increased concentration of TGLP provides additional evidence that the supply of exogenous C 18:! was responsible for the changes in milk CI8:1' Concentration of NEFA in plasma tended (P = .08) to increase as more FA were infused postrurninally (Table 6); however, NEFA concentration was highly variable. Plasma cholesterol was affected significantly by the Journal of Dairy Science Vol. 77, No.5, 1994
1382
LaCOUNT ET AL.
amount of FA infused; however, the pattern for the effect of amount of FA infused on cholesterol concentration was not clear. The short time of infusion for each amount of FA probably increased the variability in concentrations of NEFA and cholesterol in plasma; therefore, the statisticalIy significant changes of cholesterol and NEFA may be of little biological significance. Unlike data from this study, data from other researchers have shown more consistent changes in concentrations of metabolites in blood when various fat sources were infused or fed. Postruminal infusions of rapeseed oil (14) or saturated or unsaturated FA (11) increased concentrations of NEFA and cholesterol in plasma. Plasma cholesterol and NEFA increased when cows were fed increasing amounts of canola seed (18). Rafalowski and Park (25) reported that blood cholesterol increased when cows were fed 7.3 or 11.7%, but not 4.5%, regular sunflower seeds. Differences in experimental design or number of cows may
explain differences between the results of this study and others (11, 14, 18, 25). CONCLUSIONS
The FA composition of milk can be altered by providing specific FA postruminally to midlactation dairy cows. Our data indicate that contents of short- and medium-chain FA (including C 14:0 and C I6:0) decreased, and the content of C18:1 increased, when 400 gld of FA from HOSF or canola were infused into the abomasum. Within the range (0 to 400 gld) of FA infused in our study, yields of C18:1 and C18:2 in milk were increased linearly by increasing the amount of C 18 :1 and C 18 :2 infused. Additional research will be necessary to determine whether increases in the oleic acid content of milk can be even greater if more oleic acid is provided postruminalIy and whether the response is dependent on stage of lactation (24). Content of whey protein N in milk decreased, and that of NPN increased, as
TABLE 5. Effects of amount of canola or high oleic sunflower (HOSP) fatty acids (FA) infused on percentages of FA in TGLPLtriglyceride. Amount of FA infused Item C14:0 Canola HOSF C16:0 Canola HOSF C18:0 Canola HOSF CI8:! Canola HOSF C18:2 Canola HOSF
o gld
133 gld
267 gld
3.42 3.28
3,70 3.26
2.44 2.48
3.26 1.87
26.53 26.95
23.70 22.26
20,84 20.32
45.60 44.63
37.22 35.64
16,53 16.62 7.93 8,53
133 gld
o gld
SEM
2.62 1.96
2.94 4.45
5.10 4.36
.77
20,30 17.18
19.67 19.23
23.43 26.08
31.83 27.24
2.03 2.03
33.99 34.11
29.59 29.45
33.47 35.17
40.48 35,98
34.59 43.84
2.40 2.40
25.56 30.65
29.80 35.55
34.12 45.17
31.86 35.30
22.70 25.58
19.77 17.91
1.35 1.35
9.83 8.19
12.94 7.55
12.72 6.33
12.37 8.36
10.45 7.92
8.71 6.67
.88 .88
400 gld
267 gld (% of identified FA3)
Significant effects 2
,77 L A L L A L L A, T, AxT Q Q
AxT Q
L
ITriglyceride-rich lipoprotein, 2For overall effects of model: A = amount; T = treatment; T x A = treatment by amount interaction. Within treatments: L = linear effect of increasing FA infusion from 0 to 400 gld; Q = quadratic effect of increasing FA infusion from 0 to 400 gld. 3Percentages of the sum of C14:0, C16:0, C18:0' C18:I' and CI8:2' The FA C16:1 and C18:3 were not detected in all samples and were <2% of the total FA when detected, Journal of
Dairy
Science Vol. 77, No.5. 1994
1383
INCREASING OLEIC ACID IN MILK FAT TABLE 6. Effect of amount of fatty acid (FA) infused on concentrations of metabolites in plasma. I Amount of FA infused Item TGLP-TG mgldl NEFA, /LeqIL Cholesterol, mgldl
o gld
133 gld
267 gld
400 gld
267 gld
133 gld
o gld
SEM
2.6 259
4.4 300
38 274
5.6 272
4.8 244
3.2 224
2.4 196
23
.00013 .0800
232
262
222
215
223
215
208
9
.0241
.4
P
IMeans across both treatments for each amount of FA infused; effects of treatment and the interaction of treatment and amount were nonsignificant. 2Triglyceride-rich lipoprotein and triglyceride. 3Linear effect of increasing FA infusion from 0 to 400 gld (P < ,05).
.34 .32 .30 o 28 co 26 24 22 20
48
",
I
u ~
l-
I (L
18
-.J ~
16 14
/
.32 28 24 20 16
I-
~ ~
~
:· I
/
•
\ '.
iI
/-
I~/·
\
'
\
,I
I
12
48 46
:, "
;e
•
I
18
D
16
44
14
42 o 40 ex> .38 .36 .34 .32
~
ex>
u ~
l-
I (L
30
28 26
I
.36 ex>
c
~
44
40
--I ~
I-
o
133 267 400 267 133
0
Amount of FA infused (g/d)
12
f-
10
~
8
f-
6
~
4
r-
2
o
I
o
133 267 400 267 133
0
Amount of FA infused (g/d)
Figure 6. Weight percentages (grams per 100 g) of C16:0> C18:0> CIS:I' and C 18:2 in triglyceride-rich lipoproteins (TGLP) in plasma from cows infused postruminally with fatty acids (FA) from canola (e) or high oleic sunflowers (II). A) Weight percentage of C 16:0 in TGLP-triglyceride (TG) (SE of the difference between treatment means at any amount 2.7). Probabilities of greater F for effects in model: treatment, P = .34; amount, P = .0001; and treatment by amount, P = .67. B) Weight percentage of ClS:O in TGLP-TG (SE of the difference = 3.2). Treatment, P = .34; amount, P = .0001; and treatment by amount, P = .19. C) Weight percentage of CIS: I in TGLP-TG (SE of the difference 9.5). Treatment, P = .01; amount, P = .0001; and treatment by amount, P = .001. D) Weight percentage of C 18:2 in rGLP-TG (SE of the difference = 1.2). Treatment, P = .04; amount, P = .07; and treatment by amount, P = .007,
=
=
Journal of Dairy Science Vol. 77, No,S, 1994
1384
LaCOUNT ET AL.
FA infusion increased. Attention should be given to flavor and processing qualities of milk resulting from efforts to modify milk FA composition. ACKNOWLEDGMENTS
The authors extend their appreciation to J. P. Elliott for assistance with blood collection and to R. S. Younker, L. E. Williams, and C. R. Millsap for cow care. Special thanks are extended to HiDee Ekstrom for manuscript preparation. REFERENCES I Ashes, J. R., P. St. Vincent Welch, S. K. Gulati, T. W, Scott, and G. H. Brown. 1992. Manipulation of the fatty acid composition of milk by feeding protected canola seeds. J. Dairy Sci. 75:1090. 2 Association of Official Analytical Chemists. 1984. Official Methods of Analysis. 14th ed. AOAC, Arlington, VA. 3 Atwal, A. S., M. Hidiroglou, and 1.K.G. Kramer. 1991. Effects of feeding Protec~ and a-tocopherol on fatty acid composition and oxidative stability of cows milk. J. Dairy Sci. 74:140. 4 Bonanome, A., and S. M. Grundy. 1988. Effect of dietary stearic acid on plasma cholesterol and lipoprotein levels. New England J. Med. 318:1244. 5 Casper, D. P., D. J. Schingoethe, R. P. Middaugh, and R. J. Baer. 1988. Lactational responses of dairy cows to diets containing regular and high oleic acid sunflower seeds. 1. Dairy Sci. 71 :1267. 6 Chilliard, Y., G. Gagiiostro, 1. Flechet, 1. LeFaivre, and I. Sebastian. 1991. Duodenal rapeseed oil infusion in early and midlactation cows. 5. Milk fatty acids and adipose tissue lipogenic activities. J. Dairy Sci. 74: 1844. 7 Christensen, R. A., J. K. Drackley, D. W. LaCount, and 1. H. Clark. 1994. Infusion of four long-chain fatty acid mixtures into the abomasum of lactating dairy cows. 1. Dairy Sci. 77:1052. 8 Clapperton, J. L., and W. Banks. 1985. Factors affecting the yield of milk and its constituents, particularly fatty acids, when dairy cows consume diets containing added fat. 1. Sci. Food Agric. 36:1205. 9 Denke, M. A, and S. M. Grundy. 1992. Comparison of effects of lauric acid and palmitic acid on plasma lipids and lipoproteins. Am. 1. Clin. Nutr. 56:895. 10 DePeters, E. J., and J. P. Cant. 1992. Nutritional factors influencing the nitrogen composition of bovine milk: a review. 1. Dairy Sci. 75:2043. 11 Drackley, 1. K., T. H. Klusmeyer, A. M. Trusk, and J. H. Clark. 1992. Infusion of long-chain fatty acids varying in saturation and chain length into the abomasum of lactating dairy cows. J. Dairy Sci. 75:1517. 12 Erickson, P. S., M. R. Murphy, and 1. H. Clark. 1992. Supplementation of dairy cow diets with calcium salts Journal of Dairy Science Vol. 77, No.5, 1994
of long-chain fatty acids and nicotinic acid in early lactation. 1. Dairy Sci. 75: 1078. 13 Gagiiostro, G., and Y. Chilliard. 1991. Duodenal rapeseed oil infusion in early and midlactation cows. II. Voluntary intake, milk production, and composition. 1. Dairy Sci. 74:499. 14 Gagliostro, G., Y. Chilliard, and M. J. Davicco. 1991. Duodenal rapeseed oil infusion in early and midlactation cows. 3. Plasma hormones and mammary apparent uptake of metabolites. 1. Dairy Sci. 74: 1893. 15 Grummer, R. R. 1991. Effects of feed on the composition of milk fat. 1. Dairy Sci. 74:3244. 16 Grummer, R. R., W. L. Hurley, C. L. Davis, and C. A. Meacham. 1986. Effect of isolation temperature on the determination of bovine plasma very low density lipoprotein concentrations. 1. Dairy Sci. 69:2083. 17 International Dairy Federation. 1964. Determination of the casein content of milk. Int. Dairy Fed. Stand. No. 29, Int. Dairy Fed., Brussels, Belgium. 18 Khorasani, G. R., G. DeBoer, P. H. Robinson, and J. J. Kennelly. 1992. Effect of canola fat on ruminal and total tract digestion, plasma hormones, and metabolites in lactating dairy cows. J. Dairy Sci. 75:492. 19 Khorasani, G. R., P. H. Robinson, G. DeBoer, and 1. J. Kennelly. 1991. Influence of canola fat on yield, fat percentage, fatty acid profIle, and nitrogen fractions in Holstein milk. J, Dairy Sci. 74:1904. 20 Mackie, R. I., B. A. White, and M. P. Bryant. 1991. Lipid metabolism in anaerobic ecosystems. Crit. Rev. Microbiol. 17:449. 21 Mensink, R. P., and M. B. Kattan. 1990. Effect of dietary trans fatty acids on high-density and lowdensity lipoprotein cholesterol levels in healthy subjects. New England J. Med. 323:439. 22 Murphy, J. 1., G. P. McNeill, J. F. Connolly, and P. A. Gleeson. 1990. Effect on cow performance and milk fat composition of including full fat soyabeans and rapeseeds in the concentrate mixture for lactating dairy cows. 1. Dairy Res. 57:295. 23 O'Donnell, J. A. 1989. Milk fat technologies and markets: summary of the Wisconsin Milk Marketing Board 1988 Milk Fat Roundtable. 1. Dairy Sci. 72: 3109. 24 Palmquist, D. L., A. D. Beaulieu, and D. M. Barbano. 1993, Feed and animal factors influencing milk fat composition. 1. Dairy Sci. 76:1753. 25 Rafalowski, W., and C. S. Park. 1982. Whole sunflower seed as a fat supplement for lactating cows. J. Dairy Sci. 65:1484. 26 SAS~ User's Guide: Statistics, Version 5 Edition. 1985. SAS Inst., Inc., Cary, NC. 27 Schauff, D. 1., J. P. Elliott, J. H. Clark, and J. K. Drackley. 1992. Effects of feeding lactating dairy cows diets containing whole soybeans and tallow. 1. Dairy Sci. 75:1923. 28 Stegeman, G. A, R. 1. Baer, D. 1. Schingoethe, and D. P. Casper. 1991. Influence of milk fat higher in unsaturated fatty acids on the accuracy of milk fat analyses by the mid-infrared spectroscopic method. J. Food Prot. 54:890. 29 Storry, J. E., A. J. Hall, B. Tuckley, and D. Millard. 1969. The effects of intravenous infusions of cod-liver and soya-bean oils on the secretion of milk fat in the cow. Br. J. Nutr. 23:173. 30 Storry, 1. E., B. Tuckley, and A J. Hall. 1969. The
INCREASING OLEIC ACID IN MILK FAT effects of intravenous infusions of triglycerides on the secretion of milk fat in the cow. Br. J. Nutr. 23:157. 31 Sukhija, P. S., and D. L Palmquist. 1988. Rapid method for determination of total fatty acid content and composition of feedstuffs and feces. J. Agric. Food Chern. 36:1202. 32 Van Soest, P. J., J. B. Robertson, and B. A. Lewis.
1385
1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74:3583. 33 Wiesen. B., R. L. Kincaid, J. K. Hillers. and J. H. Harrison. 1990. The use of rapeseed screenings in diets for lactating cows and subsequent effects on milk yield and composition. 1. Dairy Sci. 73:3555.
Journal of Dairy Science Vol. 77, No.5, 1994