| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton T6G 2P5
2 Dairy and Swine Research and Development Centre, Agriculture and Agri-Food Canada, Lennoxville, QC, J1M 1Z3
3 Purina Mills Inc., St. Louis, MO 63166-6812
4 Instituto de Producción Animal, Universidad Austral de Chile, Valdivia, Chile
Corresponding author: H. Lapierre; e-mail: lapierreh{at}agr.gc.ca.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Although milk protein yield was positively related with total energy supply, and both digestible duodenal supply and infused AA, in this database there was no relationship between milk protein yield response above control treatments and the nutrient status of the cows (energy or protein). Total milk protein yield was defined as a function of individual AA supply, using a segmented-linear and a logistic model to obtain estimates of the efficiency of conversion of AA into milk protein. Except for Lys and Met supply, the segmented-linear model yielded lower root mean square error and better correlation, but both models were similar in their reliability. For both models, the estimated efficiency of conversion of AA to milk differed among AA. Estimations of the ideal profile of AA for lactating dairy cows were similar between models, with requirements for Lys and Met in line with 2001 National Research Council recommendations. The major difference is that the segmented-linear model yields a constant efficiency of conversion of an AA until requirements are met, with zero efficiency beyond this point. The logistic model allows for an estimation of the decreasing marginal efficiency of conversion of AA as the supply approaches the requirements. The use of variable efficiency factors should improve our ability to predict protein yield in response to supplemental protein.
Key Words: amino acid • requirement • lactation • efficiency
Abbreviation key: AAT = total digestible AA from diet and infusion, CPT = total CP supply from diet and infusion, EAA = essential AA, MP = metabolizable protein, MPT = total MP supply from diet and infusion, NEAA = nonessential AA, NELT = total NEL supply from diet and infusion, PDI = protein truly digested in the small intestine, PY = milk true protein yield, PY
= milk protein response, RMSE = root mean square error
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
) to such supplementation is often variable. Similarly, postruminal infusion of casein has produced inconsistent results (Clark, 1975; Hanigan et al., 1998). Positive responses may not be observed because the supplemental AA may not have been limiting in the basal diet. Additionally, intestinal and hepatic AA metabolism have a major impact on the amount and profile of AA supplied to the mammary gland relative to intestinal disappearance and portal absorption (Seal and Reynolds, 1993). This is evident in the study of Blouin et al. (2002), in which it was demonstrated that the liver removes from 4 (Lys) to 80% (Ser) of the AA absorbed into the portal vein, so clearly the liver is changing the profile of AA available to the mammary gland relative to the amount absorbed from the gut.
The challenge therefore remains to accurately predict PY to protein or AA supplementation. The model from the National Research Council (NRC, 2001) predicts PY using fixed efficiencies of conversion of metabolizable protein (MP) supply for maintenance (67%), gestation (33%), and milk production (67%). The Cornell Net Carbohydrate and Protein System (CNCPS, 2000) also relates milk AA output to AA duodenal flow "devoted" to milk production using fixed efficiencies of conversion of essential AA (EAA) that vary among individual AA from 62 to 100% (excluding Arg). However, the recovery of postruminally infused casein into milk protein averaged only 21% across 7 studies (Hanigan et al., 1998), which contrasts markedly with the efficiencies of conversion used by NRC (2001) or CNCPS (2000). This strongly suggests that the efficiency of conversion of AA to milk protein declines as AA supply approaches estimated requirements.
The current state of knowledge in the area of ruminant AA metabolism apparently does not allow for accurate prediction of PY in response to AA supply. Prediction schemes however do exist that attempt to relate AA supply to demand (O’Connor et al., 1993), and the NRC (2001) provides recommendations for Lys and Met supply relative to total MP based on Rulquin et al. (1993) and Schwab et al. (1992b). Rulquin et al. (2001) has also proposed requirements for other EAA based on a limited number of infusion studies.
Consequently, we integrated data from studies in which AA were infused postruminally in dairy cows with the objective of defining equations to predict PY and PY as functions of AA supply. Our second objective was to use this data set to test our hypothesis that the efficiency of conversion of digestible AA into milk protein is not a constant as is assumed in current prediction schemes (CNCPS, 2000; NRC, 2001). Specifically, our aim was to determine the influence of AA supply on the variation of the efficiency of conversion of AA into milk and to estimate requirements for EAA. Because of the limitations of the data set, the equations and efficiency values reported herein are not intended to be used on a practical feed formulation level, but rather to expand our perspective of how AA recommendations could be generated.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Thirty-three of the trials involved abomasal infusions, 8 involved intravenous infusions, 17 involved duodenal infusions, and 1 involved intravenous and abomasal infusions. In 53 of the trials, infusions were given continuously (>20 h/d). In 4 of the trials, infusions were given over a period of less than 10 h/d, and in 2 trials, both continuous and noncontinuous infusions were used. Infusions were administered on average for 11 d (SD 4.8), with a range of 4 to 28 d. Two of the trials involved hyperinsulinemic-euglycemic clamps, and results obtained during administration of the clamp (4 observations) were excluded from the data set.
Of the 59 trials, Holstein (Friesian) cows were used in 52, Ayrshires were used in 5, and both Holsteins and Ayrshires were used in 1. In one trial, the breed of cow was not reported. The cows were multiparous in 58 of the 59 trials. Twenty-seven trials began in early lactation (<84 DIM), 29 began in midlactation (84 to 210 DIM), and 1 began in late lactation. Days in milk were not reported in 2 trials. Body weight was not given in 19 of the 59 trials. In these cases, BW was estimated using reported information from other papers by the same authors included in the database. Feed intake, BW, and milk production data are shown in Table 1
. Of the 213 treatments analyzed, 57 were control treatments, 41 were casein (sodium or potassium caseinate) infusions, 9 were casein hydrolysate infusions, and 106 were free AA infusions. The number of AA infused ranged from 1 to 20 (Appendix A).
|
Data Analysis
Prediction of milk true protein yield.
Principal component analysis was performed to determine which of the variables explained the majority of the variation in the data and to explore association among the variables (Jobson, 1992; SAS, 1999). The variables analyzed included production traits (PY, PY, DIM); and dietary traits [AA (digestible and infused), MP supply derived from diet, MP supplied from infusion, and total supply of net energy of lactation (NELT = diet + infusion]. Prediction equations for true milk PY (estimated as CP yield x 0.93 (NRC, 2001) when not published) using dietary components as predictors were obtained by regressing PY against NELT (Mcal/d), total CP supply (CPT = CP diet + CP infusion, g/d), total MP supply (MPT = MP diet + MP infusion, g/d), and MPT and NELT together. Experiment within publication was included in the model as a random effect to account for variation between studies not described by other factors considered in the model. Regression analyses were performed with the MIXED procedure of SAS (1999). Intercept, linear, and quadratic terms were tested as fixed effects. A random shift on intercept and slope was modeled using experiment within publication as a subject, and a simple variance-covariance matrix for the random parameters. The PY was calculated as the difference in PY between the infusion treatments and the control treatments within each experiment.
Modeling of PY as a function of AA supply.
Milk true PY was modeled as a function of total supply of AA (AAT: dietary digestible plus infused AA). The NRC (2001) does not predict Trp or nonessential AA (NEAA) duodenal flows, so these terms were not included in any of the models tested. The AA in milk were estimated using the AA concentrations in milk reported by Jensen (1995). Linear and nonlinear functions were examined. The nonlinear models were fitted using a weighting of each observation based on the reported SEM of the milk protein yield in each experiment. A weighting variable was calculated for each observation by squaring the SEM and then dividing by the average of the squared SEM across all experiments (St-Pierre, 2001).
The linear function, a segmented-linear (also referred to as "broken-stick" or "bent-line") model, was fitted to the data using the model:
 | ([1]) |
 | ([2]) |
such that 
is the predicted yield of true protein or AA in milk (g/d), x is AAT (g/d; for example, total supply of His is referred to as HisT), a is the intercept, b is the slope, and xb is the value of x beyond which the marginal efficiency (y/x) is equal to zero. This point, xb, will be referred to as the breakpoint of the segmented-linear model. Estimates of a, b, and xb were obtained using the NLMIXED procedure of SAS (SAS, 1999).
As the majority of the data lay below the breakpoint, a linear model was tested with those data relating PY to all AAT and NELT. Variables with the largest P-values were sequentially removed until the model contained only significant variables (P < 0.10). Residuals (observed-predicted) were then regressed against variables not used in the model to determine if additional terms would contribute to the power of the model. These additional terms included ADF intake, NDF intake, nonfiber carbohydrate intake, DIM, parity, BW, feeding frequency, and site and length of infusion. Only DIM was found to be significant and subsequently included in the model.
Final regression equations were obtained using Proc MIXED as descibed above. Michaelis-Menten, exponential, and logistic nonlinear functions were tested to predict PY as a function of AAT using the NLMIXED procedure in SAS (SAS, 1999). From the 3 models, the logistic model best fitted the observed data, based on the RMSE and Akaike’s information criteria (AIC; SAS, 1999; Table 2). The logistic model selected had the form
|
 | ([3]) |
where and x are the same as in equation [1]
; and A, B, and k are the parameters of the function defining the ceiling of the response, the amplitude and the steepness of the curve, respectively.
For the final models presented, residuals were tested for heteroscedasticity using White’s and Breusch-Pagan tests in the procedure MODEL of SAS (SAS, 1999).
Calculation of efficiency of conversion of AA.
The total cumulative efficiency of conversion of individual AAT to milk protein was calculated as the ratio between the milk AA yield predicted from the fitted models (: equations [1–3]) and the corresponding amount of total supply of an AA (x = AAT). Marginal total efficiency of conversion (/x) was calculated from the ratio between increment of predicted PY and increment between corresponding supply (x). For the linear model, it is equal to the slope b from equation [1], and for the logistic model, it is calculated from its first derivative [4]:
 | ([4]) |
Calculation of the optimal total supply of AA.
For the segmented-linear model, the parameter xb is assumed to represent the value of the optimal total supply of AA, because supply beyond this point would not increase the output of y. In the logistic model, the maximal is attained at an infinite supply of x. Thus, an alternative approach to estimate the optimal total supply of AA is outlined as follows. In the logistic function, the marginal efficiency (equation [4]; Figure 1) first increases exponentially, reaches a maximum, and decreases exponentially, giving the logistic curve its characteristic sigmoid form with point of symmetry equal to the maximum marginal efficiency. In each of these symmetrical halves it is possible to define a "critical point" in which the function either starts a faster increase or decrease (i.e., the points defining the curvature of the "S"). These critical points can be defined as the minimum and maximum values of the second derivative of the logistic function (Figure 1). Algebraically, it is possible to define these two critical points (lower: xL and upper: xU) as
|
 | ([5]) |
 | ([6]) |
such that xL and xU are the values of x (AAT) where the marginal efficiency increases or decreases rapidly, respectively; and B and k are the parameters for the logistic equation fitted to individual AA (Grossman, 1986). The upper critical point (xU) was assumed to represent the optimum for AAT, as supply in excess of this value results in marginal efficiencies rapidly approaching zero. The individual AAT that corresponded to the optimum (xb for the segmented-linear and xU for the logistic models) were summed to give optimum supply of total EAA. The proportion of each EAA in total EAA was then calculated, with the proportion being indicative of the required relative contribution of each AA to an "ideal" pattern of metabolizable EAA. Approximate standard errors for the requirements (absolute amounts and amounts relative to EAA) were calculated using the ESTIMATE option of the NLMIXED procedure in SAS (SAS, 1999). Essential AA requirements were also determined on an MP basis assuming that EAA represent 48% of MP (average for the database).
Calculation of the efficiencies of conversion of AA for lactation.
In addition to total efficiency of conversion, the efficiency for lactation was calculated using the AA supply only available for milk, with the maintenance requirement (see below for estimation of requirement) of each AA being subtracted from AAT. Therefore, the linear and the logistic functions were fitted again, this time with the x being only the AA available for milk. For the segmented-linear model, the ‘no intercept’ option was used with the assumption that body stores are not contributing to milk PY, therefore when AA available for milk are zero, AA yield in milk is assumed to be zero. The maintenance requirement of each AA was calculated from the estimated NRC (2001) maintenance requirement expressed as MP, with the appropriate composition of AA for each component (Table 3). The four components that comprise the maintenance MP requirement are scurf protein, urinary protein, metabolic fecal protein, and endogenous protein (NRC, 2001). To calculate the scurf requirement, the AA composition of keratin (Block and Bolling, 1951) was used. For the urinary requirement, whole empty body composition was used (Williams, 1978; Rohr and Lebzien, 1991; Ainslie et al., 1993). Metabolic fecal protein consists of various compounds including mucous secretions, bile pigments, sloughed epithelial cells, bacterial debris, and keratinized cells (O’Connor et al., 1993; NRC, 2001). According to NRC (2001), the bacterial contribution to metabolic fecal protein was calculated as 0.5 x (bacterial MP/0.8 – bacterial MP). The AA composition of endogenous protein from rumen fluid was used to estimate the AA requirement for this component (Ørskov et al., 1986). The remainder of the metabolic fecal protein requirement was calculated as metabolic fecal protein less the bacterial contribution, and was assumed to be of intestinal origin. The AA composition of porcine intestinal endogenous protein (Stein et al., 1999; de Lange et al., 1989a, 1989b) was used for this requirement. In these studies, the endogenous protein composition was determined in animals fed protein-free diets.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
on digestible nutrients arriving at the duodenum either as a result of the digestion of the diet or supplied by the infusions. The first principal component comprised the basal digestible duodenal AA flow together with NEL intake. The second principal component comprised the supplemental AA from the infusates. The first and second principal components explained 65 and 20% of the total variance, respectively. Addition of a third principal component containing DIM and BW increased this value an additional 5%. Protein yield was positively related to both AA infused and digestible AA in duodenal flow and inversely related to DIM. The PY was negatively correlated with digestible duodenal AA flow (first principal component), and poorly correlated with supplemental AA (second principal component).
Milk Protein Yield
After accounting for a significant (P < 0.01) random effect of experiment on slope, PY and NELT were linearly related (equation 4.1, Table 4). In contrast, the PY relationships with CPT and MPT were quadratic (equations 4.2 and 4.3, Table 4), with a significant (P < 0.01) random effect of experiment on intercepts. The prediction of PY was improved when MPT replaced CPT in the model (cf. equations 4.2 and 4.3). Prediction of PY was only marginally improved by including both the NELT and MPT terms (equation 4.4). From the observations of the principal component analysis, it was concluded that a regression model that included AAT and NELT would account for the majority of the variation in PY. Therefore, for data below the breakpoint determined by the segmented-linear model, PY was regressed against AAT and NELT. Removal of the nonsignificant terms and addition of the DIM term resulted in an equation containing HisT, LysT, MetT, NELT, and DIM (equation 4.5, Table 4). To determine whether any one of these 3 AA had a dominant role in determining PY, PY was regressed against NELT, DIM, and each of these three AA individually. The residual variance after fitting the models for these AA was similar (data not shown), indicating that PY is not dependent on only one AA but that the AA are highly interrelated. Regression and principal component analyses did not identify any significant relationships between PY and AA infused, dietary digestible AA, AAT, or NELT. As a result, further attempts to generate prediction equations for PY were not pursued.
|
). Due to the different way they were defined, the optimum absolute amounts of AA estimated with the logistic model were on average 88% of the estimates from the broken stick model (Table 6). However, when the results are expressed as a percentage of EAA total supply, or as a percentage of MP, the values from the 2 models were similar (Table 6).
|
|
For both the segmented-linear and the logistic models, variation in the efficiency of conversion among AA was shown (Table 7). With both models, and across all trials and milk PY, His was the most efficiently used AA, and Arg the least efficiently used. For comparison, the efficiency values used in version 4 of the CNCPS (2000) prediction model are also presented in Table 7.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
The principal component analysis revealed that total AA supply and NELT are the primary determinants of PY. As previously discussed, energy supply is a major factor regulating MP supply. Thus, it is to be expected that these two factors would be the most influential in determining PY. The regression equation that explained most of the variation in the PY data (equation 4.5, Table 4) also agrees well with the observations of the principal component analysis. The positive coefficients of HisT, LysT, and MetT indicate that as the supply of these AA increases, PY increases. Lys and Met are often considered to be first and second limiting for milk protein synthesis (Schwab et al., 1992a; Guinard and Rulquin, 1994). Vanhatalo et al. (1999) suggested that His was first limiting when grass silage diets were fed. Although this equation suggests that His, Lys, and Met may be limiting, it must be remembered that in the majority of these studies, other AA were infused, so the possibility also exists that several AA are co-limiting.
One of the main objectives of this work was to produce equations to predict the PY to supplementary AA. However, we were not able to detect a significant association between the magnitude and direction of the PY and the changes in AA supply from the diets or the infusion across the studies in our database.
Requirements for Essential AA
In terms of absolute amounts, the estimation of optimum AA supply using the logistic model yielded values that were, on average, 88% of those estimated with the segmented-linear model. This is most likely the result of the different assumptions used to define the optimum supply in the two models. For the segmented-linear model, optimum supply is assumed when the efficiency is equal to zero, whereas the value for the logistic model is estimated assuming that the efficiency approaches zero. However, when the optimum values are represented as the amount of AA relative to total EAA and MP, both the linear and logistic models yielded fairly similar recommendations (Table 6). The recommendations for digestible Lys and Met, expressed as the percentage of total EAA are 15.1 and 5.2% and 15.0 and 5.3% for the segmented-linear and the logistic models, respectively. Both models agree with the current Lys recommendations (NRC 2001, 7.2% of MP; Rulquin et al., 1993, 7.3% of protein truly digested in the small intestine [PDI]) when expressed on an MP basis. Our estimates of the optimal supply of Met (2.5% of MP) are slightly greater than the 2.4% recommended by NRC, but match the 2.5% of PDI suggested by Rulquin et al. (1993).
The recommended allowance of 2.4% of MP for His is similar to that for Met. This value is considerably lower than that reported by Rulquin and Pisulewski (2000b), who suggested that the requirement for metabolizable His was between 3.4 and 5.6% of PDI. Considering the similarities between His and Met in terms of their hepatic uptake (Blouin et al., 2002), mammary gland uptake to output ratio (Guinard and Rulquin, 1995), and abundance in milk protein, a similar value for the recommendations of these AA seems biologically valid.
The recommendation for Leu (9.4 and 8.9% of MP for the linear and the logistic models, respectively) is consistent with Rulquin and Pisulewski (2000a), who recommended that Leu as a percentage of PDI be between 8.9 and 11.1%. These values are both in agreement with CNCPS (2000). The allowances for the other 2 branched-chain AA are also consistent with CNCPS (2000) and Rulquin et al. (2001). The Phe allowance (5.2 and 5.5% of MP for the linear and logistic models, respectively) is also in agreement with Rulquin and Pisulewski (2000c), who recommended a level between 4.6 and 5.8% of PDI.
Efficiency of Amino Acid Use
The use of the segmented-linear model is mathematically advantageous as it allows us to determine simply an estimate of AA requirements. Biologically, this model is somewhat simplistic as it suggests that below the breakpoint the efficiency of conversion of AA is constant, and beyond this breakpoint there is no increase in PY in response to increasing supplies of AA. Both Guinard et al. (1994) and Whitelaw et al. (1986) demonstrated diminishing partial efficiencies of protein use for milk protein synthesis. In the study of Guinard et al. (1994), the efficiency of converting PDI into milk protein decreased from 0.47 to 0.38 as PDI increased from 1403 to 2073 g/d as a result of casein infusions (from 0 to 762 g/d). Whitelaw et al. (1986) reported that a 200 g/d infusion of casein increased PY by 81 g/d, whereas a 600-g infusion only increased PY by 158 g/d. The use of a fixed efficiency factor in current models would at least partially explain why observed responses to supplemental protein are usually less than predicted (Bequette et al., 1998).
With current ration evaluation programs (CNCPS, 2000; NRC, 2001), increased flow of digestible AA at the duodenum is predicted to result in increased milk protein synthesis because of the assumptions that mammary gland AA supply is positively correlated with duodenal supply, that milk protein secretion is directly related to AA supply to the mammary gland, and that the efficiency of AA use is constant. The logistic model in the present study suggests that efficiency of AA use is not constant. The question then is, why would MP be used with diminishing efficiency for milk protein synthesis as its supply is increased? It may be partially attributable to changes in mammary blood flow and inefficiencies within the mammary gland. Guinard and Rulquin (1995) observed a quadratic decrease in mammary blood flow as the amount of infused Met increased from 0 to 32 g/d, with a concomitant decrease in extraction rate such that mammary gland Met uptake was unchanged. Bequette et al. (1996) reported an increased oxidation of Leu across the mammary gland as the supply of Leu increased. Diminished efficiency may also relate to metabolism in nonmammary tissues, which would reduce the AA supply to the mammary gland. Liver extraction of AA relative to portal absorption increases at higher AA supply, such that the increment in post-liver supply is much smaller than the increment in portal absorption (Guerino et al., 1991; Bruckental et al., 1997). It is still not clear if the liver is then acting to remove non-used excess AA or is having a cut on the first pass after absorption (Lobley, 2002). Therefore, as several studies have demonstrated that productive responses to AA supplementation in lactating dairy cows are not linear (Whitelaw et al., 1986; Rulquin et al., 2001), a mathematical model that allows for the changing efficiency of conversion of absorbed AA to milk protein will be a better tool to predict productive responses to dietary manipulation in dairy cows. In nonruminants, the logistic function has been used to describe the diminishing return responses to AA supplementation (Gahl et al., 1994, 1996). In our study, using the parameters from the logistic fit, this efficiency value was substantially different below and above the estimated optimum (0.91 vs. 0.56 at 50% vs. 125% of the optimum intake, respectively). These data, although applicable only to a restricted data set, clearly show that it is possible to mathematically estimate a variable coefficient for the conversion of MP to milk protein. The logistic model allows for the use of different marginal efficiencies of conversion depending on the level of AA supplied. Across AA, the marginal efficiency of conversion of digested AA into milk averaged 28% at the optimum supply and 19% for values 125% of the optimum. This marginal efficiency is in agreement with the 21% marginal efficiency reported by Hanigan et al. (1998) when casein was infused above the estimated MP requirements.
For the linear model, the efficiency of AA conversion for lactation (AA in milk/[AAT – AA for maintenance]) varied considerably among AA, ranging from a low of 0.59 for Arg to a high of 0.95 for His. A similar pattern was observed for the logistic equation. The efficiency values for the 3 branched-chain AA ranged from 0.70 to 0.76, which are in close agreement with those used in the CNCPS (2000) model (0.62 to 0.72). Likewise, the efficiency value for His and Lys are also in agreement with those of CNCPS (2000).
Although Met was the second most efficiently used AA in this study, its calculated efficiency of 80% is substantially below the CNCPS value of 100%. The difference between the 2 values may be a reflection of the methodology used to calculate the efficiency values. The CNCPS value is based on the ratio of mammary gland AA uptake to milk AA output (Overton, personal communication). The uptake of Met by the mammary gland in a 1:1 ratio with its output in milk (Guinard and Rulquin, 1994 and 1995), and its minimal catabolism in the mammary gland contribute to its high efficiency value (Mepham, 1982). The high efficiency of conversion of His can be explained by its metabolism, which in terms of its extraction by the liver (Blouin et al., 2002) and its uptake by the mammary gland relative to its output in milk, parallels that of Met (Guinard and Rulquin, 1995).
The efficiency of conversion of Phe (0.64 for the linear model, 0.53 for the logistic model) for lactation was low relative to that predicted by CNCPS (0.98; CNCPS 2000). Several possible explanations exist to account for this difference. First, it is possible that Phe was not limiting in the basal diets in the studies used in our database, and so additional Phe supplied by the infusions was not incorporated into milk protein but simply catabolized. Second, the maintenance requirement for Phe as calculated in this study may be underestimated because of the assumptions that were made regarding metabolic fecal protein requirements. Because metabolic fecal protein requirements comprise such a large proportion (~64%) of MP maintenance requirements, alterations in intestinal endogenous protein AA composition will have a major impact on the efficiency of AA use for lactation. For example, a 20% increase in Phe efficiency of use was obtained simply by changing the composition of the intestinal endogenous loss from 4.0 (Table 3) to 7.1% (obtained under AA infusion; de Lange et al., 1989b). The fact that a calculated efficiency can be increased by up to 20% so easily raises an important question—how should maintenance requirements be calculated? If the approach taken in the current study is considered reasonable and worth refining, what are the AA profiles of the constituent maintenance components that should be used? Clearly, this is an area that warrants further research.
The efficiency of Arg use, calculated to be 0.59, is considerably higher than the 0.35 used by CNCPS. This discrepancy may be attributable to endogenous synthesis of Arg from citrulline. In human adult males, the conversion of citrulline to Arg accounted for 9% of whole-body Arg flux (Castillo et al., 1993). The whole body irreversible loss rate of Arg is approximately 80% of that of Leu (Lobley et al., 1996). In the dairy cow, this would translate to a whole body irreversible loss rate of Arg of 80 mmol/h (Lapierre et al., 2002). Assuming that 9% of this is derived from de novo synthesis, then daily Arg synthesis is approximately 30 g/d. Addition of this amount of Arg to the total supply (ArgT) reduces the efficiency of conversion to 0.35.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
|
FOOTNOTES |
|---|
Recipient of a scholarship granted by the Natural Sciences and Engineering Research Council of Canada.
Current address: AgResearch Ltd., Private Bag 11008, Palmerston North, New-Zealand; the contribution of the second author is equivalent to the contribution of the first author.
Received for publication March 17, 2003. Accepted for publication September 5, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Vyas and R. A. Erdman Meta-analysis of milk protein yield responses to lysine and methionine supplementation J Dairy Sci, October 1, 2009; 92(10): 5011 - 5018. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. C. Benefield, R. A. Patton, M. J. Stevenson, and T. R. Overton Evaluation of rumen-protected methionine sources and period length on performance of lactating dairy cows within Latin squares J Dairy Sci, September 1, 2009; 92(9): 4448 - 4455. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Doepel, G. E. Lobley, J. F. Bernier, P. Dubreuil, and H. Lapierre Differences in splanchnic metabolism between late gestation and early lactation dairy cows J Dairy Sci, July 1, 2009; 92(7): 3233 - 3243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Bobe, A. R. Hippen, P. She, G. L. Lindberg, J. W. Young, and D. C. Beitz Effects of glucagon infusions on protein and amino acid composition of milk from dairy cows J Dairy Sci, January 1, 2009; 92(1): 130 - 138. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Wang, J. X. Liu, Z. P. Yuan, Y. M. Wu, S. W. Zhai, and H. W. Ye Effect of Level of Metabolizable Protein on Milk Production and Nitrogen Utilization in Lactating Dairy Cows J Dairy Sci, June 1, 2007; 90(6): 2960 - 2965. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Lanzas, L. O. Tedeschi, S. Seo, and D. G. Fox Evaluation of Protein Fractionation Systems Used in Formulating Rations for Dairy Cattle J Dairy Sci, January 1, 2007; 90(1): 507 - 521. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Pacheco, C. G. Schwab, R. Berthiaume, G. Raggio, and H. Lapierre Comparison of Net Portal Absorption with Predicted Flow of Digestible Amino Acids: Scope for Improving Current Models? J Dairy Sci, December 1, 2006; 89(12): 4747 - 4757. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Lapierre, D. Pacheco, R. Berthiaume, D. R. Ouellet, C. G. Schwab, P. Dubreuil, G. Holtrop, and G. E. Lobley What is the True Supply of Amino Acids for a Dairy Cow? J Dairy Sci, March 1, 2006; 89(e_suppl_1): E1 - E14. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. R. Stein, D. T. Allen, E. B. Perry, J. C. Bruner, K. W. Gates, T. G. Rehberger, K. Mertz, D. Jones, and L. J. Spicer Effects of Feeding Propionibacteria to Dairy Cows on Milk Yield, Milk Components, and Reproduction J Dairy Sci, January 1, 2006; 89(1): 111 - 125. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Raggio, D. Pacheco, R. Berthiaume, G. E. Lobley, D. Pellerin, G. Allard, P. Dubreuil, and H. Lapierre Effect of Level of Metabolizable Protein on Splanchnic Flux of Amino Acids in Lactating Dairy Cows J Dairy Sci, October 1, 2004; 87(10): 3461 - 3472. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
