Patterns of Nutrient Uptake by the Mammary Glands of Lactating Dairy Cows’ P. S. MILLER? B. L. REIS,3 C. C. CALVERT, E. J. DePETERS, and R. L. BALDWIN4 Department of Animal Saence University of California Davis 95616

trations of acetate, NEFA, D-phydroxybutyrate, and triacylglyceride are major determinants of uptake by the mammary glands. Factors other than plasma concentration, such as mammary gland biosynthetic capacity, availability of other nutrients, and blood flow, determine uptakes of glucose, lactate, and total and free cholesterol (r2 I .03). (Key words: nutrient uptake, mammary glands, dairy cow)

ABSTRACT

Twenty-one

multiparous

lactating

d;ury cows with previous 305-d milk production records varying from 5900 to 13,600 kg were used to examine patterns of nutrient uptake by the mammary glands. On d 71 and continuing until d 126 of lactation, animals were injected daily with 40 mg of sometribove (bST group) or bicarbonate buffer (control group). Arterial and venous blood plasma samples were collected over a 12-h period on d 35,70, 105, and 126 of lactation. Regression equations developed to evaluate linear effects of plasma arterial concentrations on net arterialvenous difference across the mammary glands demonstrated that, for acetate, NEFA, and D-p-hydroxybutyrate, plasma arterial concentration accounted for over 50% of variation in uptake by the mammary glands. Additionally, a sigmoidal equation fitted the relationship between D-phydroxybutyrate plasma arterial concentration and mammary gland uptake (3= .70). Triacylglyceride concentration was less effective in predicting uptake (r2 = .25). Administration of bST did not alter patterns of nutrient uptake, but a fourfold increase in NEFA uptake was predicted for bST-treated cows from this study, using NEFA concentrations from the literature. These observations indicate that plasma concen-

Abbreviation key: AVD = arterial-venous difference, PHBA = D-fl-hydroxybutyrate, CoA = coenzyme A, FCHOL = free cholesterol, K, = PHBA PAC at which one-half V, is obtained, PAC = plasma arterial concentration, T = sigmoidicity parameter, TAG = triacyl= glyceride, TCHOL = total cholesterol,, V theoretical estimate for maximum AVD at infinite PHBA PAC. INTRODUCTION

Received February 7, 1991. Accepted June 17, 1991. ‘Supported by US Department of Agriculture Biotechnolop Competitive Grant Number 85-CRCR-1-1889. Present address: Department of Animal Sciences, University of Nebraska, Lincoln 68583. 3 ~ address: ~ K~XXMI ~ MC-W t Laboratories. IIIC., 2225 McGaw Avenue, Irvine, CA. ‘%o whom correspondence should be sent. 1991 J Dairy Sci 74:3791-3799

The metabolic processes involved in coordinating the supply of nutrients to the mammary glands in lactating dairy cows are complex and remain. in many instances, poorly defied. The development of a dynamic mechanistic model describing mammary gland metabolism was one approach undertaken to clarify and evaluate underlying determinant mechanisms of nutrient use in lactating dairy cows (2, 31). Representation of regulatory mechanisms and parameterization of equations within the model required improved quantification of nutrient uptakes by the mammary glands. The supply of substrates to the mammary glands is dependent on substrate concentration in blood and mammary blood flow. Specifically, examining the pattern of nutrient uptake by the glands or quantifying the dependency of nutrient uptake on arterial nutrient concentration should provide a basis for determining the

3791

3792

MIULER ET AL.

contribution of blood metabolites to and en- santo Co., St. Louis, MO) in bicarbonate ergy costs associated with milk synthesis. Ex- buffer (bST group) or bicarbonate buffer only periments measunhg arterial-venous differ- (control group). ences (AVD) across the mammary glands have provided estimates of net mammary gland u p Sample Collection and takes for many blood metabolites (7, 23, 24). Blood Plasma Preparation The majority of these AVD investigations On d 34, 69, 104, and 125 of lactation, were restricted to a small number of animals, a cows were placed in open-sided stanchions (3 n m w variation in milk production, or a small m x 2.5 m) in a thermoneutral mom (22'C) interval of lactation. with free access to feed and water. Indwelling One means of increasing the variation in Catheters (i.d. = .86 mm, 0.d = 1.27 mm; milk production is with exogenous somato- Becton Dickinson & Co; Parsippany, NJ) were tropin. Substantial increases of 16 to 40% in inserted into cocyggeal (arterial blood supply) milk production have been documented for and subcutaneous abdominal (venous blood lactating dairy cows administered recombinant removal) veins. Emery et al. (12) validated the bST (26). Nutrients needed to provide addiuse of cocyggeal vein samples as indices of tional milk precursors and energy substrates arterial blood metabolite concentrations. for the augmented production associated with Catheters were flushed and fiiled with heparin exogenous bST administration appear to arise (loo0 W/W, Elkins-Sinn Inc., Cherry Hill, from mobilization of peripheral tissues and NJ) overnight. The following morning (0300 from increased feed intake (8, 13, 28). Quanti- h), cows were milked and catheters reflushed. fication and examination of how key blood Blood samples (8 ml) were collected simultametabolites are delivered to and extracted by neously from both cocyggeal and mammary the mammary glands in bST-treated cows vein catheters three times per hour for 12 h. should provide a basis for evaluating the sig- Plasma was separated from red blood cells by nificance of changes in mammary gland bicentrifugation (3000 x g for 3 min), harvested, osynthetic capacity and blood flow. pooled hourly, and stored at -2o'C. Hourly The experiment was conducted 1) to estabarterial and venous plasma samples were analish patterns of mammary gland nutrient u p lyzed for glucose to check for errors in labeltake for multiparous lactating dairy cows with ing, etc. Approximately 1.5% of samples was a wide range of productivities over the first discarded on the grounds that the samples were 126 d of lactation and 2) to investigate possi- mislabeled (e.g., arterial and venous labels ble changes in patterns of nutrient uptake by were apparently switched). Remaining hourly the mammary glands in cows administered ex- arterial and venous plasma samples for each ogenous bST. cow were pooled over the 12-h period. MATERIALS AND METHODS Anlmals, Houslng, Dlet, and Treatments

Twenty-one multiparous Holstein cows with previous 305-d milk production records varying from 5900 to 13,600 kg were used. Diet, feed analyses, housing, and milking schedules were described by Brown et al. (8). Two animals were removed from the study because of i n j q and health complications. Cows were paired according to previous production records and randomly assigned to either the recombinant bST (n = 10) or conml group (n = 9). On d 71 of lactation and continuing until d 126, cows were injected intramuscularly daily with either 40 mg of sometribove (MonJournal of Dairy Science Vol. 74, No. 11, 1991

Plasma Metabollte Analysls

Twelve-hour pooled arterial and venous plasma samples were analyzed for acetate, Dfbhydroxybutyrate (PHBA), triacylglycerol (TAG), NEFA, glucose, lactate, total cholesterol (TCHOL), and free cholesterol (FCHOL). Plasma acetate was butylated and quantified by GLC (Hewlett Packard model HP5700A; Hewlett Packard, Palo Alto, CA) according to the following procedure as modified from Salanim and Muimead-(29). One milliliter of plasma and .1 ml of internal standard [.1% (voVvol) isobutyric acid, pH > 7] were mixed and 5 ml of denatured 95% ethyl alcohol added in order to precipitate protein. The mix-

NUTRIENT UPTAKE BY THE MAMMARY GLANDS

ture was c e n m g e d for 10 min at 2800 x g and supernatants collected. Excess ethyl alcohol was evaporated under a gentle stream of N2 gas (55'C). One milliliter of chloroform (Optima grade; Fisher Scientific, San Francisco, CA) and .2 ml of BF3-n-butanol were mixed with each sample and then placed into an 80'C heating block for 2 h (samples were tightly capped). Samples were cooled on ice for 10 min before adding .4 ml of trifluoroacetic anhydride (reagent grade; Fisher Scientific, Pittsburgh, PA) to remove excess BF3-nbutanol. Samples were kept at room temperature for 1 h and subsequently washed with 2 ml of H20. Washed preparations were centrifuged for 8 min at 300 x g, and the aqueous phase was removed and discarded. Washing and centrifugation were repeated. Sodium sulfate (.4 g) was added to each sample to remove excess H20. After centrifugation (8 min at 1500 x g), clear Supernatants were transferred to injection vials for analysis. A 182.9cm x 2.5-mm column packed with chromosorb WHP (80 to 100 mesh) and coated with 15% Dexsil 300 GC (Supelco, Inc., Bellefonte, PA) was used to separate butyl esters. Peak areas were measured using a Hewlett Packard integrator (model 3393A. Hewlett Packard, Palo Alto, CA). Temperature settings used were as follows: detector, 3WC; injection port, 250'C; initial column temperature, 65'C and 4'Clmin increase to 150'C; 150'C for 8 min. Gas flow rates used were nitrogen (99.99%), 50 d m h ; hydrogen, 80 ml/min, and air, 250 d m i n . Triacylglyceride and NEFA were analyzed colorimetricalIy using a Technicon Auto , h a lyzer (Technicon Corp., Ardsley, NY). Triacylglyceride was extracted from plasma and quantitied by a modification of the procedure outlined by Levy and Keyloun (20). Briefly, TAG was extracted from 1 ml of plasma with the following mixture: 4 ml of heptane, 7 ml of isopropanol, and 2 ml of HzS04 (.8N). The entire heptane layer was collected, evaporated to dryness under N2 gas (55°C). and resuspended in .7 ml of heptane for analysis. Potassium hydroxide (.25 M ) was used for saponification instead of sodium methylate. A modification of the procedure of Lorch and Gey (22) was used to measure plasma NEPA. Then, NEFA were extracted from 1 ml of plasma with a mixture of 2 ml of isopropanol, 2 ml of heptane, 1 ml of H2SO4 (lN), and 1.5 ml of H20. The heptane layer was collected and evaporated to dryness and resuspended in

3793

3 ml of heptane for analysis on the auto analyzer. Glucose and TCHOL in plasma were measured using a Kodak Ectachem DT60 Analyzer (Eastman Kodak Co.,Rochester, NY). Lactate and PHBA were measured using commercially available kits (Lactate, Sigma number 826; JHBA, Sigma number 310; Sigma Chemical Co.. St. Louis, MO). Plasma FCHOL was measured using a modification of Sigma Diagnostic kit number 352 in which cholesterol esterase was removed. statistics

Linear regressions of plasma arterial concentrations (PAC) on AVD were parameterized with the general linear models procedure of SAS (30) given the following model:

where

Yi is the AVD predicted for the cow i; Po and PI are intercept and slope parameters, respectively; Xi is the PAC for cow i; and ei is the random residual error with a mean of zero and variance o2 (-N (0,02)). The NLIN procedure (Marquadt method) of SAS (30) was used to parameterize the sigmoidal equation describing AVD of PHBA given the model:

where

Yi is PHBA AVD for cow i, is theoretical estimate for the maximum AVD at infinite PHBA PAC, K, (apparent) is the PHBA PAC at which one-half,,V is obtained, Xi is the j3HBA PAC of cow

,V

1,

T is the sigmoidicity parameter, and ei is the random residual error (-N (0,a2)). Journal of Dairy Science Vol. 74, No. 11, 1991

3794

MILLER ET AL.

TABLE 1. Parameta

estimates for equations predicting BlteTiBI-ven0u.s differences from arterial comxntratiom for plasma metabolites rcprescnted in Figures 1 through 9.

Intercept

Item

-

Glucose Lactate Acetate $HBA3 NEFA TAG4 TCHOL~ FCHOL~

Slope

X

SE

.52 -.07 -.36 .04 -35.57

.18 .07 .14 .04 8.08 .46 102 .82

.79

-22 -.13

X

SE

.04

.003

.09

.08

.71

.07 .03

24

21 29

.U2

.OM

.03 .03

.04

Pl

PAC Range2

P

.3327 2597 5.ooO1

2.2 S x 5 4.7 mM .3 S X 5 2.4 m M 1.3 5 X 5 3 mM .7 S x 5 4.7 mM 99 S X 5 1353 pM 2.8 S X 5 12.4 mg/d 50 S X 5 475 mg/dl 17 S X 5 56 mgldl

.01

5.OOO1

S.OOO1 S.OOO1 .8537

.06

.1304

.M .60 .54 .56 .25

.oo .03

hobability that slope = 0. 2Hasma arterial concentratioq independent variable range. 3D-t%Hydroxybutyrate. %acylglyceride. 5 ~ ~ t choiatmi. al 6Ree cholesterol.

The validity of the sigmoidal model with deduced parameters V-, K,, and T was determined by Monte Carlo analysis. Briefly, deduced parameter estimates (mean f SE) and PHBA PAC were used to generate a new data set (AVD). The sigmoidal equation was reparameterized using the generated data set, and deduced parameter estimates were compared with original values for all parameters (V-, K,, and T). Initial and final estimates

GLUCOSE 1.3 i

> I

k

"1

.1 2

3

4

6

ARTERlALGwcosE ImM) Figure 1. Relation between glucose arterial concentration and glucose arterial-vcnous (A-V) diffa-ence. The bST--treated COWS. d 105 and 126 (0);dl COWS, d 35 and 70 and placebo cows. d 105 and 126 (0).Parameter estimates for regression line givm in Table 1. Journal of Dairy Science Voi. 74, No. 11. 1991

differed by less than 2%; therefore, the initial parameter estimates were accepted. RESULTS AND DISCUSSION

Relationships between AVD and arterial concentrations for plasma metabolites are represented in Figures 1 through 9. Parameter estimates for regression relationships depicted in Figures 1 through 9 are given in Table 1. Patterns of nutrient uptake by mammary glands represented in Figures 1 through 9 did not change with bST treatment. Comparison of intercept and slope estimates were not different (P > .l) with b!3T administration. Glucose AVD was independent of plasma arterial concentration (Figure 1, Table 1). Mean glucose AVD was .69 mM across a range in plasma arterial glucose concentration of 2.2 to 4.7 mM. This observation is in accordance with in vitro K, estimates of Forsberg et al. (14), which also suggest that physiological glucose concentrations are not limiting for glucose utilization by the mammary glands. Several workers have shown glucose AVD to be linearly correlated to PAC (16, 21). A review published by Davis and Collier (lo), which extrapolated mean glucose AVD and arterial concentrations fiom a series of separate experiments, also demonstrated a linear relationship between these two variables. Data from these experiments fall within the range of AVD in

3795

NUTRIENT UPTAKE BY THE MAMMARY GLANDS

A -V GLUCOSE VS. A - V ACETATE

LACTATE

1.41

0

9

.6

9

12

15

1.8

A - V A E r A l E (mM)

Figure 2. Relation between @lucose and acetate arterial-venous (A-V) differe~ce.The bST-trealed cows, d 105 and 126 (0);all cows, d 35 and 70 and placebo cow, d 105 and 126 (0).Regression equation: Glucose AVD = 6.93 + (Acetate AVD).5.15, ? = .32, P S .ooO1 where AVD = arterial-venous difference.

Figure 1. In these earlier studies, the apparent dependency of AVD on PAC may have resulted from the limited number of observations considered. The independence of glucose uptake from arterial concentrations suggests that mammary gland metabolism is a key determinant of glucose AVD. Intramammary utilization of glucose is a highly regulated process and dependent on the availability of other milk precursors and energy substrates (14).Figure 1 illustrated the nondependency of glucose AVD on glucose PAC. In contrast, Figure 2 reveals that glucose uptake is partially dependent on acetate uptake (9= .32). This corresponds to earlier work reported by Forsberg et al. (14). demonstrating that, as in vitro acetate concentration increased, the conversion of glucose to lactose, C@, and glycerol-3-phosphate increased. Therefore, interanimal variation in glucose uptake can be potentially explained by the contribution of alternative energy sources and milk precursors and by the requirement of glucose for lactose synthesis. Lactate AVD was not correlated with PAC (Figure 3, Table 1). Mean AVD values over the range of PAC observed in the present study (.3 to 2.4 mM) indicate that lactate uptake from plasma was not dependent on udder metabolism. Forsberg et al. (15) demonstrated in vivo that the mammary gland possesses a high capacity for oxidation of lactate and conversion to fatty acids. The utilization of lactate for

-

1

2

2

3 2

0

ARTERlAL LACTATE (mM) Figure 3. Relation between lactate arterial concentration and lactate artaid-~enous(A-V) difference. The bST-mtd COWS, d 105 a d 126 (0);all COWS, d 35 and 70 and placebo cows, d 105 and 126 (0). Parameter estimates for regression line given in Table 1.

these processes was inhibited by acetate concentrations above 1 mM, apparently because of regulation at pyruvate dehydrogenase and citrate lyase. Therefore, acetate concentrations observed in this study may inhibit lactate conversion to C02 and fatty acids and may potent i d y explain the low AVD for lactate. Acetate PAC accounted for 60% of variation in net acetate uptake CFigUre 4, Table 1). The strong linear relationship between acetate uptake and arterial concentration suggests that acetate uptake and subsequent utilization are not rate-limiting at physiological concentrations and, possibly, that dietary and endoge nous sources of acetate may limit acetate oxidation and de novo fatty acid synthesis in the gland. Pethick and Lindsey (27) showed a positive correlation between acetate entry rate and acetate PAC for lactating ewes, but this relationship was not established for lactating cows (7). hitially, the BHBA AVD response to plasma arterial concentration was represented using simple linear regression (Figure 5 , Table 1). Although this equation accounted for 54% of variation in net PHBA uptake, closer examination warranted the application of a nonlinear equation form. A sigmoidal equation was subsequently fitted and increased the $ by 30% (Figure 5). Baldwin and Smith (3) used a Michaelis-Menten equation to express the relationship between BHBA AVD and PAC for values extrapolated from the literature and calJ o d of Dairy Science Vol. 74. No. 11, 1991

3796

MILLER ET AL.

BHBA 1.1 j

-

/

.l

0

Figure 4. Relation between acetate arterial concentration and acetate arterial-venous (A-V) difference. The bST-treated cows,d 105 and 126 (0);all cows, d 35 and 70 and placebo cows, d 105 and 126 (0). Parameter estimates for regression line given in Table 1.

1

3

2

4

5

ARlERW BHBA ImM) Figure 5. Relation between D-khydroxybutyrate (BHBA) arterial concentration and BHBA arterial-venous (A-V) difference. The bST-trealed cows, d 105 and 126 (0);all cows,d 35 and 70 and placebo cows, d 105 and 126 (0).Parameter estimates for regression line given in Table 1. Si oidal equation: AVD = .82/(1 + (PAC/ 1.29)3.29); .70. P 5 .OOO1 where AVD = arterialvenous difference and PAC = plasma arterial concentration

y=

culated the apparent Ks and V,, to be .39 mM and .32 mmoW of plasma, respectively. Values for these parameters estimated from data in this study are considerably higher (Ks [apparent] = 1.29 mM, V , = .82 mmoW of plasma). The higher K, estimates are a result of the greater number of observations in this study, which supported use of a sigmoidal rather than the simple Michaelis-Menten relationship. Ketosis was not clinically diagnosed in any cows during the study. However, plasma arterial PHBA concentrations above 2 mM were observed in animals sampled during the first two periods (d 35 and 70) of the study, suggesting elevated rates of liver ketogenesis during early lactation. Net uptake of PHBA by the udder approached saturation at plasma concentrations of 2 mM. This nondependency of PHBA uptake on arterial concentration at high concentrations could reflect saturation of FHBAcoenzyme A (CoA) synthetase or competition for CoA, which also is used for uptake and activation of plasma NEFA (also elevated in these cows). Elevated plasma ketone body concentrations are often associated with a reduction in plasma arterial glucose and increased arterial NEFA (19, 33). Plasma arterial glucose was not correlated with plasma arterial PHBA (Y = 4.12 - .11X r2 = .03, P = .12). The NEFA concentrations in plasma were linearly correlated with plasma @IBA (Y = .18 + .lox, r2 = .12, P = .0032). Journal of Dairy Science Vol. 74, No. 11, 1991

Examination of Figure 4 indicates a narrow range for plasma PHBA concentrations in bST-treated cows (.6 to 1.7 mM). This range was not different from that for control cows at the same stage of lactation. Changes in NEFA PAC accounted for 54% of variation in NEFA AVD (Figure 6, Table 1). The NEFA AVD tended to become nega-

NEFA

-im\ 0

,

Po

,

,

,

,

em B a ) l a ARTERW NEFA (HM) 400

,

,

X , l a a , l ~

Figure 6. Relation between NEFA arterial concentration and NEPA arterial-venous (A-V) difference. The bST-Wmted COWS, d 105 and 126 (0);dl COWS, d 35 and 70 and placebo cows, d 105 and 126 (0). Parameter estimates for regression Line given in Table 1.

3797

"T UPTAKE BY THE MAMMARY GLANDS

FREE CHOLESTEROL

TRIACYLGLYCEROL

>

I -6

>

I 4

O

2

4

8

8

10

12

I4

4

T

10

20

90

40

50

m

AAlERlAL TRlACYLGLYcERoL(lng/dl)

ARTDUAL CHOLESTEROL ( W d 9 Figure 7. Relation between triacylglycaide (TAG) arFigure 9. Relation between free cholesterol (FCHOL) tcrial concenhation and TAG Mcrial-vcnous (A-V) difference. me bST-treatad cows, d 105 and 126 (0);all cows, arterial CoIlCmtcation and FCHOL arterial-venous (A-v) d 35 and 70 and placebo cows, d 105 and 126 (0). diffamce. The bST-treated COWS, d 105 and 126 (0);all Parameter estimates for regression lint given in Table 1. COWS. d 35 and 70 and placebo cows,d 105 and 126 (0). Parameter eatimatcs for regression line given in Table 1.

tive at plasma concentrations below 350 pM and, as previously reported (32),was probably a consequence of lipoprotein-lyase hydrolysis of plasma lipoprotein TAG. Previous studies of the relationship between NEFA AVD and arterial concentration have been limited, and results were variable (7, 34). It has been difficult to detect variation in arterial NEFA PAC utilizing animals with the same genetic background, sampled at the same stage of lactation with similar diets and intakes. Results from the present study suggest uptakes ranging from 50

TOTAL CHOLESTEROL 70'

P

E ,"I

g-I: -

Y -=-

0

3

>

-501

I -70 4 100

Po

m

400

500

(Wdr) Figure 8. Relation between total cholesterol (TCHOL) TCHOL arterial-venoas (A-V) difference.The bST-treated cows, d 105 and 126 (0);all cows,d 35 and 70 aml placebo cows, d 105 and 126 (0). Parameter estimates for regression line given in Table 1. arterial concentration and

to 100 pA4 over a range of plasma arterial NEFA concentrations of 375 to 625 pA4. Studies documenting NEFA PAC of 250 to 350 pA4 have shown NEFA AVD to be negative or not different from zero (1, 5). Negative uptakes have been reported in d a b cows supplemented with protected lipid maintaining plasma arterial NEFA concentration of 570 to 800 (34). In these studies, the negative uptakes may reflect hydrolysis of TAG, which was likely elevated by these high fat diets. Twenty-five percent of variation in TAG AVD was accounted for by TAG PAC (Figure 7, Table 1). The low $ for this relationship likely reflects the small range in TAG umcentration observed in this study (4 to 12 mg/dl) rather than the lack of a strong relationship between concentration and uptake, because Baldwin et al. (4) showed a strong hyperbolic relationship in cows fed protected fat and exhibiting TAG arterial concentrations ranging from 5 to 200 mg/dl (apparent K, for TAG uptake was 32 mg/dl). Neither TCHOL (Figure 8) nor FCHOL (Figure 9) AVD was correlated with PAC (Table l). Mean AVD values for TCHOL and FCHOL were not different from zero (P> .1). Bovine milk fat contains approximately .4% cholesterol as FCHOL and cholesterol esters (17). which would correspond to a range in daily cholesterol secretions in milk of 2.2 to 6.7 g in cows from this study. Assuming a Journal of Dairy Science Vol. 74. No. 11, 1991

3798

MILLER ET At.

plasma ffow to milk production ratio of 500: 1, AVD for TCHOL would need to range from .41 to 1.3 mgdl in order to a m m t for the quantity of cholesterol secreted in milk. An additional source of FCHOL and cholesterol esters is de novo synthesis (9, 18). De novo synthesis would decrease the requirement for cholesterol uptake. The relative contribution of these two sources remains unclear. Exogenous administration of bST did not alter patterns of nutrient uptake per unit volume of blood flow for the metabolites examined in this study. Absolute availabilities of nutrients from blood require consideration of blood flow rates, which are elevated in bSTtreated cows (11). The results of this study indicate that bST does not alter extraction rates of nutrients and, thus, that the primary action of bST in increasing milk production must result from increased blood flow rates to the udder, increases in the metabolic efficiency of the udder, or a combination of both. No in vitro experiments investigating bST actions on nutrient utilization at the intracellular levels have been reported. A preliminary report (25) indicates that rates of mammary gland oxidation of glucose and acetate iut lower in bSTtreated versus control cows. Another consideration is that PAC for bST-txeated and control cows may differ. Mean PAC of glucose, lactate, acetate, and BHBA were not different (P > .05) in bST-treated and control COWS during the 105- and 1 2 6 4 sampling periods, whereas NEFA PAC increased from 224 pM in control cows to 361 pM in bST-mated animals (P = .03). Calculation of NEFA AVD from these NEFA PAC using the regression equation in Table 1 estimates an increase in NEFA AVD of 29 mddl in bST-treated cows. In a shortterm study (14-d bST administration), Bauman et al. (6) reported a 51% increase in plasma NEFA concentration with bST administration, which if extrapolated to this study would correspond to a fourfold increase in NEFA AVD. Additionally, these workers showed plasma NEFA concentration to be correlated (9= .64) to NEFA irreversible loss, indicating that the uptake of NEFA is dependent not only on plasma concentration and blood flow but also on the rate of substrate oxidation and conversion to milk components. Therefore, increased blood flow rate and, to a lesser extent, increased udder metabolic efficiency may conJ o d of Dairy Science Vol. 74, No. 11, 1991

tribute to increased milk production, inasmuch as the concentration of the majority of plasma nutrients remains unaltered. Overall, observations from this investigation indicate that PAC of acetate, =A, BHBA, and TAG are major determinants of AVD across the mammary gland, whereas factors other than PAC, such as mammary gland biosynthetic capacity and blood flow, determine uptakes of glucose, lactate, TCHOL,and FCHOL. 'Ihe development of a rigorous model of bST actions on nutrient availability and uptake by the mammary glands q u i r e s quantification of patterns of net nutrient uptake, blood flow, and mammary intracellular metabolism and nutrient interactions. The development of a mechanistic model describing mammary gland metabolism is dependent on the description and measurement of these processes. ACKNOWLEDGMENTS

The authors gratefully acknowledge the assistance of Harvey Freetley, Scott Middleton, Joanne Knapp, Randy Baldwin, Tom Famula, Nada Khalaf, Chris Clothier, Jenny Talbot, Mike Bowers, Karen Charvet, John Argyle, Karla Black, Diane Harris, Gary Hamell, Steve Hannah, and Ted Mori in sample collection and laboratory and data analyses. REFERENCES lAnnison,E.F., J.L.Linzell, S.FazaLnley,andB. W. Nichols. 1967. The oxidation and utilization of palmitate, stearate, oleate and acetate by the """"y gland of the fed goat in relation to thew overall

m e t a b o w and the role of plasma phospholipids and neutral lipids in milk-fat synthesis. Biochem. J. 102: 637. 2 Baldwin, R. L., M. D. W g a u , and S. C. Middleton. 1989. Development of a mechanistic model of mammary gland metabolism in the lactating COW. J. airy Sci. 72(Suppl. 1):315.(Abstr.) 3 Baldwin, R L., and N. E. Smith. 1983. Adaptation of metabolism to various conditions: mi& production. Page 359 in Dynamic biochemistry of animal production.P. M. Riis, ed.Elsevier Sci. hbl ., B. V., Amsterdam, Neth. 4Baldwin, R. L.. N.E. Smith, J. Taylor, and IN Sharp. 1980. Manipnlatine; metabolic parameters to improve growth rate and milk secretion. J. h i m . Sci. 51:1416. 5Barry. J. M,W. Bartlq, J. L. Linzell, and D. S. Robinsoa 1963. Thc uptake from blood of triglyceride fatty acids of chylomicrons and low deusity lipoproteins by the m ~ m m o v yglaad of tJx goat. Bio-

"T UPTAK@BY THE MAMMARY GLANDS

&em. J. 896. 6Bnnmnn. D. E., C. J. Peel, W. D. Steinhour, P. J. Reynolds, H. F. Tyrrell, A.C.G. Brown, and G. L. Waalnnd. 1988. Effect of bovine somatotropin on metabolism of laaating dairy Cows: influence on rates of irreversible loss and oxidation of glucose and nonestaified fatty acids. J. Nu&. 1181031. 7Bickerstaffe. R., E. F. Annison, and J. L. Linzell. 1974. he metabow of gh~cose.acetate. lipids and amino acids in lactating dairy cows. J. Agric. Sci. (Camb.) 82:71. 8 Brown, D. L., S. I. Taylor, E. J. DePeters, and R L. Baldwin. 1989. Influence of sometribove, USAN (recombinant mcthionyl bovine Somatotropin) on the body composition of lactating cattle. J. Nutr. 119:633. BCIarenbmg, R., and I. L. Chaikoff. 1966. Origin of milk cholesterol in the rat dietary versus endogenous xnuces. J. Lipid Res. 727. 10Davis. S.R.,and R. J. Collier. 1985. Mammary blood flow and regulation of substrate supply for milk synthesis. J. Dairy Sci. W1041. 11 Davis, S. R., R J. Collier, J. P. McNamarq H. H. Head and W. Sussman, 1988. Effects of thymxine and growth hormone trealmcnt of dairy cows on milk yield, cardiac output and marrrrrmry blood flow. J. Anim. Sci. 66:70. 12Emcay. R S.. L. D. Bmwn, and J. W.Bell. 1965. Correlation of milk fat with dietary and metabolic factors in cows fed restricted roughage rations supplemented with magnesium oxide or sodium bicarbonate. J. Dahy Sci. 48:1647. 13 Eppard, P. A.. D. E. Banman, and S. N.Mcclltcheon. 1985. Effect of dose of bovine growth hormone on lactation of dairy cows. J. Dairy Sci. W1109. 14Forsberg. N. E., R L. Bald* and N. E. Smith 1985. Roles of glucose and its interaction with acetate in maiatenanct and biosynthesis in bovinc mammary tissue. I. Dairy Sci. 68:2544. 15Forsberg. N. E., R L. Baldwin, and N. E. Smith 1985. Roles of lactate and its htaaction with acetate in maintcnaoce and biosynthesis in bovine mammary tissue. J. Dairy Sci. 68:2550. 16 Graham, W.R., H. D. Kay, and R A. McIntosh. 1936. A convenient mabod for obtaining bovine lnterial blood. Roc. R. SOC. B 124319. 17 Jenness, R. 1985. Biochemical and nutritional aspacts of milk and colostrum. Page 164 in Ladation. B. L. anon, ed. IO= state univ. press, a. 18 Kinsella, J. E., and R D. McCarth. 1968. Lipid c ~ m position and secrdory activity of b o w maumuuy cells in vitro. Biochim. Biophys. Acta 164530. 19 Kronfeld, D. S. 1971. Hypo%yccmia in ketotic cows.

3799

J. Dairy Sci. 54949.

u)LCvy, A. L., and C. Keyloun. 1970. An automated colorimetric (non-fluommeaic)assay for triglycerides.

Wgc 497 in Advances in =tomated analysis. Technicon Int. Congr., New Yo& NY. 21 Linzell, J. L. 1960. Mammary glad blood flow and oxygen, glucose and volatile fatty acid uptake in the conscious goat J. -01. (Lond.)153:481.

E., and K. F. Gey. 1966. Photometric "titration" of free fatty ncids with the techuicon autoanalpa.Anal. Biochaa 16244. 23 McDowe4 0 . H., J. M.Gooden, D. Leumumksa, M. Jois, and A. W. &@ish 1987. Effects of exogenous growth hormone on milk prodaction and nutrient u p take by muscle and memnpary tissues of dairy cows in 22Lorch,

mid-lactation. AWL J. BIO~.S C 40.295. ~

24Maca,J.R.J.M.Goodtn,E.TeleniG.M.Hough, G. H. McDowell, and E. F. Annison. 1980. Uptake and oxidation of glucose by the hind limb muscle and mammary glaud of the lactating ewe. Proc. [email protected]. Allst Anau. Cod. 5:170. 25 Middleton, S. C., R L. Baldwin, and C. C. Calvert. 1988. pattans of glucose metabolism in lactatiog cow :mammary glands. J. Dairy Sci. 71(Suppl. 1): 247.(Absfr.) %Peel, C. I.. and D. E. Bauman 1987. Somatotropin and lactation. J. Dairy Sci. 74474. 27Petitick, D. W.,aod D. B. Lindsay. 1982. Acetate metabolism in lactating sheep. Br. J. Nu@. 48:319. 28Pocms, P. A., and J. H. Herbein. 1986. Effects of in vivo& * . - h'on of growth hormone on milk production and in vitro hepatic metabolism in dairy cows. J. Dairy Sci. 69:713. 29salaoitro, 1. P.,and P. A. M a . 1975. Quantitative metbod for gas chromatographic maIysis of shortchain mOllOCBTbOXylic and dicarboxylic acids in fermentation media Appl. Mimbiol. 29374. ~ O S A S @user's m e : statistics, version 5 tion on. 1985. SAS Inst., Inc., Cary, NC. 31 Waghorn, G. C., and R L. Baldwin. 1984. Model of metabolite ff tu within mammarygland of the lactating cow. J. Dairy Sci. 67531. 32 West,C. E.,R.Bickerstaffe, E.P. Anuison, and I. L. Linzell. 1972. Stadies on the model of uptake of blood triglycerides by the mammary gland of the lactating goat. Biochun. J. 126:477. 33Ysmdagni S., and L. H. &hulk. 1970. F a 9 acid composition of blood plasma lipids of normal and ketotic cows. J. Dairy Sci. 53:lW. 34Yang. Y.T., J. M.W e . and R. L. Baldwin. 1978. Dietary lipid mtabolism in hating dairy COWS. J. Dairy Sci. 61:1400.

Journal of Dairy Science Vol. 74, No. 11, 1991