| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Dairy and Swine Research and Development Centre, Agriculture and Agri-Food Canada, Lennoxville, Quebec, Canada J1M 1Z3
University of New Hampshire, Durham 03824-3542
Université de Montréal, St-Hyacinthe, Québec, Canada J2S 3B7
BIOSS, Rowett Research Institute, Aberdeen, UK AB21 9SB
# Rowett Research Institute, Aberdeen, UK AB21 9SB
2 Corresponding author: lapierreh{at}agr.gc.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
Key Words: dairy cow • amino acid • gut • portal absorption
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
In nonruminants, intake represents supply and any deficiency can be corrected with the simple addition of any individual AA directly to the diet. Ruminants are different: their capacity to use forages that are indigestible for the nonruminant relies on the presence of microorganisms in the rumen, reticulum, and omasum to digest such feedstuffs. Unfortunately, this ability incurs a toll. First, during passage through the reticulorumen, dietary ingredients are partially digested and reused for microbial growth. Therefore, nutrients delivered for absorption differ from those present in the diet making prediction of nutrient supply to the animal difficult. Second, because free AA are rapidly degraded in the rumen environment (Velle et al., 1997; Volden et al., 1998), simple addition of an AA to the diet is not an efficient option to increase AA flow at the duodenum. Instead, the AA must either be coated with a form of protection against the microorganisms (but still allow digestibility in the small intestine) or be presented in chemical forms that are more resistant to ruminal degradation or more readily absorbed through the rumen wall (e.g., hydroxy analogues).
Over the last few years, there has been renewed interest in strategies to improve the efficiency of N use in dairy cows without reducing productivity. Increased efficiency of N use is best achieved by reducing the total amount of CP fed to the animal. Such a reduction, however, must be made cautiously but can be accomplished if appropriate supplies of the critical AA are provided. This raises a number of crucial questions. How do we measure the supply of individual AA? What is the true supply of AA? In ruminants, it is clear that AA intake is a meaningless way to express the real supply of AA to the animal. Should we then look at duodenal supply or perhaps small intestinal disappearance? Is either really net supply? Would it be better to measure AA appearing in the portal circulation that drains the whole gut? Answering these challenging questions requires knowledge of AA metabolism between intake and absorption. Recognizing and understanding the complex interactions of gut AA metabolism will enable the development of better predictive models to formulate diets that contain less total CP but are better balanced to meet AA requirements. Such an approach will improve the efficiency of conversion of diet N into milk protein.
This review will focus on specific aspects of present knowledge and suggest improvements to current prediction models taking into account gut metabolism of AA. For simplicity and to reflect the paucity of data on nonessential AA flows in dairy cows, this review will focus on essential AA (EAA), even though data for these are also scarce. Nonetheless, future models will also need to consider metabolism of nonessential AA as these are extensively used by gut tissues to maintain integrity and functionality.
|
DUODENAL FLOW |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
Therefore, determination of the true net supply of protein (and of AA) flowing to the duodenum requires that the contribution from preduodenum endogenous proteins be subtracted from measured total duodenal flow. In practice, however, a number of simplifications are usually adopted. Rumen microbial protein flow, due to its potentially large contribution and associated variation, has received much attention and different methods have been developed for measurement (see review in Broderick and Merchen, 1992). Consequently, in most studies where protein flow at the duodenum is measured, MCP is determined and RUP is calculated as the difference between total CP and MCP flows, with any contribution from endogenous protein ignored. Indeed, although it is critical to determine what fraction of duodenal flow represents recycled material, and thus accurately determine the real net supply, the endogenous fraction has received little attention in ruminants. This has not only been because of a lack of recognition of the relative importance of the endogenous contribution but also because of the technical challenges associated with its measurement. Even when allowance is made for endogenous inputs, these are often based on data from studies conducted in rather artificial conditions having little relevance to the feeding regimen of high-yielding dairy cows. Thus, NRC (2001) adopted an average value of 1.9 g of N/kg of DMI for the endogenous contribution to duodenal flow based on data obtained in studies using animals receiving either only intragastric infusions of volatile fatty acids (Ørskov et al., 1986) or fed diets with very low protein supply and degradability (Hart and Leibholz, 1990; Hannah et al., 1991; Lintzenich et al., 1995). Similarly, the French system uses a value equivalent to 1.7 g of N/kg of DMI (Vérité and Peyraud, 1989). More appropriate estimates of endogenous protein flows, pertinent to "real" situations, require use of tracer techniques but, even with these, caution is needed. Endogenous secretions estimated from 15N-urea infusion (Brandt et al., 1984; Leng and Nolan, 1984) include use of urea for microbial protein synthesis. Although this can be considered "endogenous," the impact on AA availability will differ from endogenous protein arising from previously absorbed AA. In this review, only endogenous proteins that arise from direct use of AA will be considered and discussed.
Recently, Ouellet et al. (2002) used an 8-d infusion of 15N-leucine to develop a model in dairy cows to estimate the endogenous contribution to the duodenum, including free endogenous secretions and endogenous proteins incorporated into the MCP flow. This latter fraction averaged 2.1 g of N/kg of DMI, approximately equal to the free endogenous flow of 2.3 g of N/kg of DMI. Together (4.4 g of N/kg of DMI), these 2 fractions represented 15% of the total duodenal N flow, irrespective of the fiber content of the diet. In sheep, using a continuous abomasal infusion of 15N-labeled, grass meal-beer yeast suspension, the endogenous contribution averaged 12% of duodenal N flow (Van Bruchem et al., 1997). Another study in sheep, based on 15N-labeled digesta exchange between animals, reported endogenous N contributing between 3 and 12% of total N duodenal flow for diets ranging from 15 to 25% crude fiber content, respectively (Sandek et al., 2001). Furthermore, in a second study in lactating cows using an infusion of 15N-leucine (Ouellet et al., 2005), the endogenous contribution to duodenal N flow averaged 18% when silage was fed, and 20% when hay was offered, representing in this latter case, 5.9 g of N/kg of DMI. In dairy cows, using a mathematical approach including estimates of the composition of AA in each fraction, N from endogenous origin varied between <1% (Shabi et al., 2000) and 32% (Larsen et al., 2001) of total duodenal N flow.
With such a high potential contribution, the endogenous fraction can no longer be ignored and needs to be subtracted from measured duodenal flow to determine the true net supply at the duodenum. However, with such a large imprecision in these evaluations, more research is required to determine what factors affect endogenous duodenal flows to obtain better estimation of the contribution in different feeding situations. In addition, these endogenous secretions will not only have an impact on the evaluation of the total duodenal N flow, but will also alter the pattern of the net supply of individual AA, due to the differences in AA composition between endogenous proteins, MCP, and RUP (Table 1
). Although such differences are acknowledged, more information is needed on the AA profile of endogenous secretions at the duodenum. The best estimation available relies on just one study in which AA composition of abomasal isolates was measured in animals receiving only intragastric infusions of volatile fatty acids (Ørskov et al., 1986).
|
and 3, respectively. In the text, we will follow the fate of one AA, lysine, because it is considered one of the limiting AA in typical North American diets. From measurements reported by Berthiaume et al. (2001), duodenal flow of lysine averaged 144 g/d, with 30 g estimated to be of endogenous origin. This means that the net supply of lysine arriving at the duodenum was, in fact, 114 g/d. So, right at the start, acknowledgment of the contribution of endogenous secretions changes the picture of what is supplied to the cow! Furthermore, because the AA profile of the different duodenal fractions is not homogeneous, not only does the true duodenum flow decrease when the endogenous contribution is removed, but the proportion of each AA to total EAA may increase (e.g., leucine) or decrease (e.g., threonine).
|
|
|
DISAPPEARANCE IN THE SMALL INTESTINE |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
Although ileal digestibility might appear a more reliable measurement of AA availability, a closer examination reveals that several correction factors need to be applied to determine the true net supply to the animal. A proportion of the ileal digesta are of endogenous origin and this is not a negligible amount. For example, in dairy cows infused with 15N-leucine, Ouellet and colleagues (D. R. Ouellet, R. Berthiaume, G. Holtrop, G. E. Lobley, and H. Lapierre; Agriculture and Agri-Food Canada and Rowett Research Institute, UK; unpublished data) observed that endogenous N losses at the ileum represented 28% of total ileal N flow. In sheep, the endogenous contribution to the ileum N flow comprised 32% (Sandek et al., 2001) and 48% of total ileal N flow (Van Bruchem et al., 1997). Therefore, estimation of true digestibility of protein and AA available to the animal varies greatly depending on whether, and how, these endogenous secretions are accounted for.
The presence of endogenous protein in ileal digesta is recognized in pig nutrition. Rather than apparent digestibility, standardized ileal digestibility, which accounts for endogenous secretion (Jansman et al., 2002), is successfully used to balance pig diets, because the relative endogenous contribution to total flow differs with the protein supply. However, in pigs, the relevant endogenous proteins are secreted between the input (intake) and the output (ileal cannula) of the digestibility measurement; that is, there is no endogenous contribution to the input. In contrast, for dairy cows, we have already seen that the input (i.e., duodenal flow) already contains endogenous secretions. As a result, part of the endogenous loss at the ileal cannula will originate from secretions already present at the input (duodenal cannula) that were not digested, as well as secretions that occurred between the input and the final measurement (i.e., of small intestinal origin). This means that in dairy cows, the ileal endogenous protein present at the ileum needs to be partitioned into that derived from undigested preduodenal endogenous secretions and that which originates from endogenous proteins secreted into the small intestine but not reabsorbed. A recent study in dairy cows indicated that both fractions were approximately equivalent (D. R. Ouellet, R. Berthiaume, G. Holtrop, G. E. Lobley, and H. Lapierre; Agriculture and Agri-Food Canada and Rowett Research Institute, UK; unpublished data). Therefore, in addition to the definition of the true digestibility calculated to account for secretions from the small intestine, as adopted in pigs, this introduces a consideration unique to the ruminants: the amount of AA truly digested in the small intestine does not represent the net supply to the animal. Digestion and absorption of the endogenous secretions present at the entrance of the duodenum will not provide a new supply to the animal. Instead, their absorption into blood circulation simply replaces AA previously extracted from arterial blood supply to support their synthesis. To determine the amount of AA absorbed from digested preduodenal endogenous secretions, both the endogenous contribution to the duodenal flow (see above) and the digestibility of this fraction need to be determined. In sheep, where labeled digesta were exchanged between animals, the digestibility of these secretions averaged 60% (Sandek et al., 2001), which is close to the estimation of 68% in dairy cows (Ouellet et al., 2002).
To better integrate these concepts, we will continue to follow the fate of lysine across the gut, based on the study of Berthiaume et al. (2001; Table 3
). The difference between duodenal (144 g/d) and ileal (37 g/d) flow yields an apparent SID of 107 g/d. Does this number represent the amount of lysine that really disappears from the small intestine? If we want to know the total amount of lysine that was truly digested in the small intestine (equivalent to true ileal digestibility in the pig), the endogenous contribution arising from intestinal non-reabsorbed endogenous secretion must be removed from the ileal flow; this is 7 g/d in our example. Therefore, the total amount of lysine digested increases to 114 g/d (107 + 7). Although this true ileal digestibility is used in pigs to determine AA availability, this concept cannot be directly applied in dairy cows. As previously discussed, the true ileal digestible supply includes a portion of endogenous proteins that was already present at the duodenum and digested in the small intestine; this fraction does not constitute a net supply to the animal. We have already determined that endogenous contribution to the duodenal N flow is 30 g/d of lysine. From Ouellet and colleagues (D. R. Ouellet, R. Berthiaume, G. Holtrop, G. E. Lobley, and H. Lapierre; Agriculture and Agri-Food Canada and Rowett Research Institute, UK; unpublished data), it was determined that 9 g of these secretions was still present at the ileum and therefore not digested. This means that 21 (30 – 9) g/d of the lysine that was digested was not a true net supply, but rather represents a recycling of endogenous proteins. Therefore, the true amount of lysine digested from net supply is now 93 (114 – 21) g/d. At this point, bear in mind the 16 g of lysine from endogenous origin flowing at the ileum; this will be discussed in the next section.
It must also be remembered that endogenous secretions have an impact not only on the "total" AA but also on the profile of AA available for the cow. Analogous to the situation already discussed for the duodenal AA profile, the proportion of an individual AA relative to total EAA can be markedly altered when corrections for endogenous inputs are applied. In our example, this results in a relative increase for leucine while threonine decreased.
Synopsis
The small intestinal disappearance, either apparent or "true" (i.e., corrected for small intestinal endogenous contribution at the ileum) is not yet "the" measurement that will give the real net amount of AA supplied from MCP and RUP. The endogenous contribution to the duodenal flow digested in the small intestine needs to be removed from the "true" amount digested to indicate the correct net supply to the animal
|
PORTAL ABSORPTION |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
One option to quantify how much AA are lost across the gut is to compare apparent SID with portal absorption. The first attempt to do this in sheep reported large variations in the ratio of portal absorption on SID between the various AA, from 19% for histidine to 69% for lysine, at a high protein intake (Tagari and Bergman, 1978). These findings suggest a substantial loss of AA across the gut. More recent data obtained in sheep and dairy cows report ratios of portal absorption to SID smaller than 1, but the "losses" for essential AA were of smaller magnitude, with an average recovery of apparent SID as portal absorption varying between 43% for threonine and 76% for phenylalanine (Table 4; Mac-Rae et al., 1997; Berthiaume et al., 2001). But does this ratio really indicate AA catabolism/loss by gut tissues?
|
The real loss of AA across the gut originates from 2 fractions: endogenous proteins still present at the ileum that will not be reabsorbed, and direct AA oxidation. As previously determined in the lysine example, at the ileum, 16 g/d were of endogenous origin, not reabsorbed in the small intestine, and therefore, lost in the hindgut or in the feces. Therefore, the maximal quantity of lysine that could reach the portal vein is the net quantity apparently digested corrected for the digested preduodenal endogenous secretions (114 – 21 = 93 g/d) minus these endogenous losses (16 g/d). This would give a portal absorption:SID ratio of 72% [(93 – 16)/107]. Accounting for losses of endogenous protein would then yield a potential maximal availability of 77 (93 – 16) g/d, higher than the reported portal absorption of 60 g/d. Ratios for portal absorption on apparent SID, on estimated net absorption, and on net availability are given in Table 4. For some AA, the ratio of portal absorption on net availability, including endogenous, is close to unity (even higher), but for other essential AA, the loss of endogenous proteins does not explain all the lack of recovery between estimated availability and portal flow. Other losses must occur; indeed oxidation of some EAA by gut tissues has been reported.
Impact of AA Oxidation by Gut Tissues
Obviously, many assumptions have been made for these calculations, particularly as relates to the quantities and profile of AA as endogenous secretions. Nonetheless, the exercise suggests that there is minimal oxidation of some AA, such as histidine and phenylalanine, and more extensive catabolism of the branched-chain AA, lysine and methionine. Unfortunately, available data on AA oxidation across the gut in lactating dairy cows are as limited as the data on endogenous secretions. Nonetheless, in ruminants, leucine oxidation across the gut has been reported, including observations in dairy cows (see review, Lobley and Lapierre, 2003). In sheep, oxidation of methionine averaged 10% of portal absorption, but no oxidation was observed for lysine and phenylalanine (Lobley et al., 2003). A recent study in sheep reported that net appearance into the portal circulation in comparison with abomasal infusions of incremental amounts of casein had a slope different to unity for the branched-chained AA and lysine (El-Kadi et al., 2004), but not for other EAA. This would support substantial oxidation (between 43 and 51% of the incremental supply) of these AA by gut tissues.
The few data available do not allow identification of all factors that would affect the magnitude of AA oxidation but it becomes increasingly evident that this is not a fixed absolute amount. For example, in dairy cows, leucine oxidation across the gut increased with MP supply and averaged 20 and 30% of portal absorption at low vs. high MP supply, representing 31 vs. 38% of whole-body leucine oxidation (Lapierre et al., 2002). In sheep, diet quality, quantity, and the presence of parasites affected leucine oxidation across the gut (see review: Lobley and Lapierre, 2003). All these measurements were made with labeled leucine infused into the systemic circulation; that is, the leucine used for oxidation was from an arterial source. In pigs, leucine used for oxidation by the gut originated from both luminal and arterial supply (Van der Schoor et al., 2001). Oxidation from the lumen source includes oxidation of the AA by the gut cells during passage to the blood circulation but may also involve oxidation by gut microorganisms. On the other hand, in pigs, lysine oxidation across the gut was sensitive to lysine supply but there was no oxidation of the AA from the systemic circulation (van Goudoever et al., 2000), as also observed in sheep, but enteral lysine could be catabolized by the porcine digestive tract. Indirect evidence also indicates that oxidation might be related to supply to the gut. In dairy cows in which AA were infused into a jugular vein, resulting in a substantial increase in arterial concentrations of AA, net portal appearance decreased for most of the EAA except phenylalanine and threonine (Berthiaume et al., 2002). This result indicates greater uptake by gut tissues, probably related to the incremental total inflow. Similarly, the linear relationship between portal recovery of the branched-chained AA and lysine in sheep receiving abomasal infusion of incremental amount of casein, with slopes lower than unity, strongly indicates that these AA were oxidized in proportion to supply at the gut (El-Kadi et al., 2004).
Examination of our "case study" for lysine flow in dairy cows presents a mixed message about oxidation of lysine by the gut. In the example we used, portal absorption (60 g/d) was lower than the net available after endogenous losses were accounted (77 g/d), suggesting a potential lysine oxidation of 17 g/d. This would represent 15% of the total amount digested, in agreement with studies in pigs (van Goudoever et al., 2000) and sheep (El-Kadi et al., 2004) in which oxidation from the lumen side was reported. On the other hand, there was a substantial difference between mesenteric absorption and SID of lysine found by Berthiaume et al. (2001), suggesting a measured SID higher than expected; more recent data (R. Berthiaume and H. Lapierre; unpublished data) yielded close agreement between net available (after endogenous losses were accounted) and portal absorption. This result would suggest limited oxidation of lysine by the gut, as observed by Lobley et al. (2003).
To understand why the gut may utilize AA, we have to consider the many separate functions that the different sections of the digestive tract perform. The gut is not a homogeneous tissue and AA are absorbed only from one section—the small intestine—and this is the only part in positive AA balance. The other sections (fore-stomachs and hindgut) are in negative AA balance (Lobley et al., 2003; Rémond et al., 2003) and, to maintain their integrity and function, utilize AA from the arterial supply; that is, those that have already been absorbed and circulated around the body. Indeed, use of AA by the whole gut is dominated by arterial sources; 80% of gross use was from this source in sheep (MacRae et al., 1997b).
It is not clear to what extent this removal by the fore and hind regions of the gut is obligate or regulated. Although the need to support endogenous secretion is relatively easy to understand, why should the gut directly oxidize some AA, particularly methionine and perhaps lysine, that may be limiting to support other body processes, including lactation? Certain AA such as the branched-chain AA and lysine are catabolized by nonhepatic tissues (Goodwin et al., 1987; DeSantiago et al., 1998; Pink et al., 2003); perhaps the digestive tract needs to contribute to whole body oxidation. Alternatively, these AA may perform metabolic functions linked to their oxidation. For example, lysine degradative pathways produce glutamate, an important energy source for the gut cells (Windmueller and Spaeth, 1980) and a precursor for de novo synthesis of arginine and proline (Ball, 2002). Methionine metabolism is linked to provision of 1-carbon units, as S-adenosylmethionine, and nucleic acid biosynthesis, important functions for proliferative cells, such as those of the gut epithelium. Leucine is the best EAA for transamination reactions.
Absorption of peptide-bound AA into the blood circulation has remained a controversial issue over recent years. This issue will not be addressed in depth in the current review, because for several studies in which peptides have been measured, the inclusion of absorption as peptides yields results that do not fit well with the integration of N metabolism across the gut. For example, N absorption into blood circulation (ammonia and AA) should not exceed entry of N in the gut [urea recycling plus digested N; in cow, see for example, Delgado-Elorduy et al. (2002) and Tagari et al. (2004)]. Even accurately measuring veno-arterial differences of individual AA across the gut presents a severe technical challenge and we believe that the greater complexity of direct peptide veno-arterial measurement with actual techniques is at the limit, or beyond, the sensitivity of the current methods available. In addition, in the following section, it will be shown that predicted digestible AA does not appear to account for all free AA absorbed into portal circulation, so the requirement for extra amino acid absorption in peptide form looks relatively small.
Synopsis
Gut metabolism substantially and differentially alters the availability of AA supplied from MCP and RUP. Based on the ratio of portal absorption to true net digestibility, endogenous loss and oxidation of EAA averaged 35% of the net digestible amount, with losses varying from 25% for phenylalanine to 55% for threonine. This is an important new area that requires further investigations, because losses of AA through the 2 routes of endogenous proteins and gut oxidation will need to be incorporated into future predictive models to define the AA demands of the dairy cow.
|
LINK BETWEEN PREDICTED DIGESTIBLE AA AND PORTAL ABSORPTION |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
Prediction of Digestible AA
Due to the scarcity of studies in which both SID and portal absorption of AA were measured simultaneously, we decided to use an existing subrumen model predicting digestible flow of EAA and link these estimations with measurements of portal absorption of AA.
Duodenal flow and intestinal digestibility, as discussed in this review, apply for total digesta and do not partition between individual feed ingredients. As such, they cannot be used to predict AA supply from different diets. Instead, to predict digestible AA to the dairy cow, sophisticated subrumen models have been developed to estimate duodenal protein flow based on characteristics of feed ingredients appropriate to each model [NRC, 2001; Cornell Net Carbohydrate and Protein System (CNCPS) in Fox et al., 2004; Vérité and Peyraud, 1989]. Despite the acknowledged endogenous contribution to the duodenal protein flow, for modeling purposes, certain models do not include this fraction into their estimation of the duodenal protein flow (CNCPS, Fox et al., 2004). Duodenal flow of individual AA is then determined with either a factorial approach attributing an AA composition to each of the fractions (Vérité and Peyraud, 1989; CNCPS in Fox et al., 2004) or through the development of regression equations yielding the proportion of the AA relative to total EAA in the duodenal flow (NRC, 2001). In such models, conversion from protein duodenal flow to MP or from AA duodenal flow to digestible AA is achieved by multiplying the flow presented at the duodenum by the digestibility of each fraction.
As discussed previously, the difficulty of direct measurement of true intestinal digestibility of feed ingredients has led to the development of an indirect technique, the mobile bag technique. Feed ingredients, after incubation in porous bags suspended in the rumen, are transferred into other small porous bags preincubated, or not, with pepsin to mimic gastric digestion and then inserted into the small intestine via a duodenal cannula and eventually recovered from an ileal cannula or the feces. Using this mobile bag technique, there was no difference in recovery at the ileum or the feces, as long as the feed was preincubated in the rumen (Prestløkken and Rise, 2003), except for mature forages (Hvelplund and Weisbjerg, 2000). Feces could thus be used as the end-point, alleviating the need to maintain a patent ileal cannula (NRC, 2001). After washing the bags, it is assumed that there are no residual endogenous proteins, and therefore, disappearance from the bag is assumed to be the true digestibility. Using these feed digestibilities, coupled with a value for microbial digestibility plus another for endogenous secretions (when considered), models now predict the supply of digestible or metabolizable AA to the cow (French system: Vérité and Peyraud, 1989; CNCPS: Fox et al., 2004; NRC, 2001).
Link with Portal Absorption
The first step in our approach was to decide which model predicting digestible AA should be used. Bateman et al. (2001) had already tested several models designed for North American conditions [NRC, 1989; CNCPS, version 3; Cornell Penn Miner Dairy (version 1.0); and Mepron Dairy Ration Evaluator (version 1.1 from the Degussa Corporation, Ridgefield Park, NJ)] and reported that no single model could be identified as better in predicting flow of individual EAA to the duodenum. To complete the picture in terms of models currently available, we simulated the same publications included in Bateman et al. (2001) using the latest version of NRC (2001), and concluded that this model underestimates the mean flow of EAA to the duodenum (Table 5). These results are similar to those presented by the authors of the original study, who also reported significant proportions of the prediction errors were associated with systematic bias (30 to 40%) for all the models tested.
|
Based on concepts previously explained in this review, the digestible flow of AA predicted by NRC (2001) was corrected by subtracting the contribution of endogenous preduodenal secretion, using the estimation of the NRC model and AA composition of abomasal isolate (Table 1), whereas the CNCPS model did not include any endogenous contribution to the duodenal flow. All subsequent references to NRC (2001) predicted duodenal digestible AA flow are corrected for this endogenous contribution; that is, the values presented would now represent only the contribution from RUP and MCP to duodenal flow. Results of this exercise showed that digestible supply of EAA estimated by CNCPS were numerically higher (by 5 to 40%) than those obtained from the NRC model with the exception of leucine, which represented only 92% of the NRC value (Table 6). Analysis of correspondence indicated that AA profiles; that is, the relative amount of each EAA to the total EAA, were not significantly different between portal measurement and NRC or CNCPS estimations (P > 0.05, 
2 test, 16 df), although there appeared to be a trend for the NRC model to give a higher relative proportion for leucine than CNCPS. In contrast, CNCPS estimates for the relative proportions of arginine and histidine tended to be higher than those from NRC.
|
).
Although the CNCPS yielded higher estimates of digestible AA flows than NRC, this did not indicate which model gave the better fit against our portal measurements. With the assumption that digestible supply of individual AA is directly related to the corresponding net portal absorption, we applied the same assessment protocol as that used by Bateman et al. (2001), but here we used the portal absorption as the value modeled from the digestible supply data. Although neither NRC nor CNCPS were designed for estimation of net portal absorption, we would expect a degree of association between the model estimates (particularly as all the ingredient data and feed intake were known) and our measurements of portal absorption. We also assumed that errors in determining the true digestible flow would be equivalent between models. From this comparison, NRC gave a better prediction of net portal absorption from our experiments, as based on a lower root mean prediction error (Table 7). However, the partitioning of the prediction error indicated a systematic underprediction (mean bias). This then raised the obvious question: are NRC estimates underpredictions or are our net portal absorptions overpredictions? We have no clear answer to this question, but the close agreement in our studies between the post-liver supply of AA from group I (i.e., histidine, methionine, and phenylalanine; Mepham, 1982) and milk AA output would suggest that our blood measurements did not suffer from a systematic overprediction.
|
, 6, and 7, NRC (2001) appeared to underpredict the flows, certainly for the AA where the slope was greater than unity (histidine, methionine, and phenylalanine) but also for other AA with a slope close to, or lower than, unity (isoleucine, leucine, lysine, threonine, and valine), recognizing the use of these AA across the gut (see above). A revisit of the NRC (2001) equations was then undertaken in an effort to improve the predictive ability of the model taking into account criticisms and suggested refinements of the model.
First, the use of combined data sets from lactating and growing cattle to generate equations for duodenal flows of protein and AA has been criticized (Alderman et al., 2001). Indeed, 22% of the observations used to generate the equations for duodenal flow of total EAA and the proportional of individual EAA are from growing cattle. Because the steer data cluster at low levels of EAA from RUP and MCP, and knowing the impact of intake on rumen fermentation, this point needed to be taken into account. Second, St-Pierre (2001) suggested that the data should be weighted based on the error associated with the measurements in each study. Therefore, the equations were reformulated using only the cow data and data were weighted as recommended. This yielded equations that increased the duodenal flow only by 2% (Tables 8 and 9). The equation to determine duodenal flux of total essential AA (Table 8), however, has a better fit with biological events: a coefficient of 0.95 (vs. 0.86) with the RUP fraction indicates almost a total recovery of EAA from the RUP fraction. A coefficient of 0.40 (vs. 0.43) with MCP is in agreement with MCP containing 80% of true protein and an EAA ratio on total AA lower than 50%. Finally, the intercept 60 g/d of EAA could be considered an estimate of endogenous protein and would scale to 120 g/d of total AA or 240 g/d of total endogenous CP (true protein, or AA, content of endogenous protein estimated at 50% - NRC, 200). This latter value would yield 28 g/d of N or, at an estimated intake of 20 kg of DMI, 1.9 g of N/kg of DMI, in line with the actual value used in the NRC model (2001). However, for individual AA, these improvements were still insufficient to account for AA loss across the gut, when comparing digestible AA flow with portal net absorption.
|
|
), and the slope of the portal measurements vs. these estimates also decreased (Table 10), indicating that the recovery of digestible AA flow into the portal circulation was not complete. Although this result would allow some potential loss of AA through gut metabolism, in agreement with the concepts presented earlier in this review, the magnitude of the losses still appear insufficient (e.g., losses of only 16% for leucine, and effectively no loss for lysine). However, the ranking of the AA in term of potential loss follows the same trend as previously discussed: minimal oxidation of phenylalanine and histidine, limited oxidation of lysine, substantial oxidation of the branched-chain AA, and a loss of threonine larger than for other AA due to its relatively high proportion in endogenous protein (Table 1). The slope was still greater than unity for methionine and phenylalanine. At this point there is no clear explanation of this observation, but we are currently investigating the preparation of the digesta samples before hydrolysis in the studies used to develop the equations.
|
Synopsis
Integrating knowledge on AA metabolism will help build predictive models that should acknowledge a decreasing conversion of feed protein to milk protein as protein dietary supply is increased. Although all the needed information is not yet available, current knowledge can already help us to assess the validity of current models and direct us toward critical improvements.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
Although knowledge of factors affecting gut AA metabolism in dairy cows is slowly accumulating, there is already enough information available to signify the importance of this area. Therefore, future predictive models will have to integrate concepts and approaches relevant to this area improve their accuracy.
|
FOOTNOTES |
|---|
3 Current address: AgResearch Ltd., Private Bag 11008, Palmerston North, New Zealand.
Received for publication September 21, 2005. Accepted for publication November 18, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONDUODENAL FLOWDISAPPEARANCE IN THE SMALL...PORTAL ABSORPTIONLINK BETWEEN PREDICTED...CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. P. Weiss and N. R. St-Pierre A method to quantify changes in supply of metabolizable methionine to dairy cows using concentrations of selenium in milk J Dairy Sci, June 1, 2009; 92(6): 2835 - 2842. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. El-Kadi, K. R. McLeod, N. A. Elam, S. E. Kitts, C. C. Taylor, D. L. Harmon, B. J. Bequette, and E. S. Vanzant Nutrient net absorption across the portal-drained viscera of forage-fed beef steers: Quantitative assessment and application to a nutritional prediction model J Anim Sci, September 1, 2008; 86(9): 2277 - 2287. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Marini, D. G. Fox, and M. R. Murphy Nitrogen transactions along the gastrointestinal tract of cattle: A meta-analytical approach J Anim Sci, March 1, 2008; 86(3): 660 - 679. [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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
