The Journal of Nutrition Vol. 128 No. 10 October 1998, pp. 1752-1759

Threonine Requirement of Neonatal Piglets Receiving Total Parenteral Nutrition Is Considerably Lower than That of Piglets Receiving an Identical Diet Intragastrically^1,2,3

Robert F. P. Bertolo^*, Cathy Z. L. Chen^*, Garson Law^**, Paul B. Pencharz^{, **, , ,} , and Ronald O. Ball^{*, , **, , , 4}

^* Department of Human Biology and Nutritional Sciences and  Department of Animal and Poultry Science, University of Guelph, Guelph, ON, Canada N1G 2W1; ^** Department of Agricultural, Food, and Nutritional Sciences, University of Alberta, Edmonton, AB, Canada T6G 2P5;  The Research Institute, The Hospital for Sick Children, Toronto; and  Department of Paediatrics and  Department of Nutritional Sciences, University of Toronto, Toronto, ON, Canada M5G 1X8

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Evidence is accumulating that the amino acid requirements for neonates receiving total parenteral nutrition (TPN) are significantly different than those for oral feeding and need to be determined. The parenteral threonine requirement was determined in 3-d-old male Yorkshire piglets (n = 25) by examining the effect of varying dietary threonine intakes [0.05-0.6 g/(kg·d)] on phenylalanine oxidation. The diet included adequate energy, total amino acids and phenylalanine, with excess tyrosine. Phenylalanine kinetics were determined from a primed, continuous intravenous infusion of L-[1-¹⁴C]phenylalanine. Phenylalanine oxidation, estimated from the rate of ¹⁴CO₂ released in expired air during isotope infusion, decreased (P < 0.05) as threonine intake increased from 0.05 to 0.15 g/(kg·d) and was low and constant for threonine intakes >0.15 g/(kg·d). Using breakpoint analysis with 95% confidence interval (CI), mean requirement and safe level of parenteral threonine intake were estimated to be 0.19 and 0.21 g/(kg·d), respectively (equivalent to 13 and 14 mg/g amino acids, respectively). To compare these data with those of orally fed controls, we then repeated the experiment by infusing identical diets intragastrically to piglets (n = 25); the varying dietary threonine intakes were 0.1-1.2 g/(kg·d). Employing identical kinetics and analyses, the mean requirement and safe level of oral threonine intake were estimated to be 0.42 and 0.51 g/(kg·d), respectively (equivalent to 28 and 34 mg/g amino acids, respectively). These data demonstrate that the threonine requirement of neonates during TPN is ~45% of the mean oral requirement.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Part of the nutritional strategy for neonates fed enterally or parenterally is to provide indispensable amino acids at a level sufficient to meet the demands for protein synthesis while avoiding an excessive supply. The accumulation of an amino acid or its degradative products in body pools may stress immature enzymatic systems and lead to further accumulation and possible toxicity. The threonine requirement is affected by the ontogenic development of metabolic enzymes. In both preterm (Schanler and Garza 1987) and term infants (Janas et al. 1985), threonine intake positively correlated with plasma threonine, suggesting that oxidation of excess threonine was not sufficient. Furthermore, in preterm infants fed formula, the highest serum threonine concentrations were observed in the infants with the lowest gestational ages (Rigo and Senterre 1980). In growing rats, hepatic threonine dehydratase (EC 4.2.1.16) activity was low during most of the postnatal period, compared with adult rats (Grogan et al. 1988). High plasma concentrations of threonine are of potential concern in infants because of the possible neurotoxic effects of threonine. Because threonine concentrations in cerebrospinal fluid correlate with increasing plasma concentrations and because of the extensive neurologic development occurring during the early life of infants, high threonine intakes represent a possible danger to the developing brain (Anderson and Raiten 1992). Therefore, excess threonine intakes might be hazardous to the premature infant fed orally or parenterally due to the immature metabolic system of these infants.

The amino acid profile and concentrations in parenteral nutrition solutions generally reflect those in oral diets. However, there is evidence that parenteral amino acid requirements are lower than enteral requirements (House et al. 1998). Parenterally administered threonine, in particular, may be in excess because gastrointestinal requirements are diminished as a result of gut atrophy. In low birth weight infants receiving intravenous infusions containing threonine, increased plasma threonine concentrations were observed even when total nitrogen and energy intakes were low (Anderson et al. 1977). High threonine intakes, such as those provided by intravenous solutions (Anderson et al. 1977) and whey protein-predominant formulas (Janas et al. 1985), appear to exceed the capacity of the low birth weight infant to metabolize this amino acid (Rigo et al. 1994, Scott et al. 1985). Given a reduced capacity to metabolize threonine during early infancy, threonine intakes in excess of requirement should be avoided in both oral and intravenous formulations.

No reports were found in the literature in which the threonine requirement was determined directly in either orally or parenterally fed premature infants. The objective of this study was to determine threonine requirement in a neonatal piglet model receiving total parenteral nutrition (TPN)⁵ by employing the indicator amino acid oxidation technique using L-[1-¹⁴C]phenylalanine. Discovery of a threonine requirement different from current practice would justify direct studies in the human infant. This study contributes to our overall goal of designing an optimal amino acid profile for neonates.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Animals and study protocol.  All procedures used in this study were approved by the Animal Care Committee of the University of Guelph. A total of 50 male Yorkshire piglets, ~3 d of age, were transferred from a specific pathogen-free herd (Arkell Swine Research Unit, University of Guelph, Guelph, Canada) to the animal holding facilities in the Department of Animal and Poultry Science, University of Guelph. On arrival, the piglets were weighed, anesthetized and fitted with catheters (Ed-Art, Don Mills, Canada) by using modified methods of Wykes et al. (1993) for the venous catheters and Rombeau et al. (1984) for the stomach catheter. Briefly, catheters were tunnelled under the skin from the point of exit on the left side of the chest to the points of entry into the blood vessels or stomach. In Experiment 1 (parenteral threonine requirement), an infusion catheter was inserted into the left jugular vein and advanced to the superior vena cava just cranial to the heart. In Experiment 2 (oral threonine requirement), the infusion catheter was inserted into the stomach using a Stamm gastrostomy. A sampling catheter was introduced into the left femoral vein of all pigs and advanced into the inferior vena cava just caudal to the heart. Catheter positions were verified at necropsy. After surgery, the piglets were fitted with an adjustable cotton jacket with an attached anchoring button attached to a tether. The laboratory conditions and piglet housing have been described previously (Wykes et al. 1993).

An elemental and complete diet was infused continuously through a tether-swivel system (Alice King Chatham Medical Arts, Los Angeles, CA) as previously described (Wykes et al. 1993). In Experiment 1, TPN was administered using pressure-sensitive infusion pumps and lipid (Intralipid 20%, Pharmacia-Upjohn, Stockholm, Sweden) was infused simultaneously using syringe pumps. In Experiment 2, water (1.5-fold dilution to lower osmolarity) and the appropriate amount of lipid were added to the final solutions, which were infused continuously by using the same pressure-sensitive infusion pumps. The infusion regimen (continuous, 24 h) was designed to supply all nutrients required by piglets (Wykes et al. 1993); piglets received 15 g amino acids/(kg·d) and 1.1 MJ metabolizable energy/(kg·d) with glucose and lipid each supplying 50% of nonprotein energy intake. Full infusion rates for Experiment 1 [272 mL/(kg·d) TPN, 52 mL/(kg·d) lipid] and Experiment 2 [480 mL/(kg·d)] were adjusted so that energy and nitrogen intakes were identical for all piglets. The base amino acid profile of the TPN consisted of (mg/g total L-amino acids): alanine, 92; arginine, 61; aspartic acid, 61; cysteine, 15; glutamic acid, 105; glycine, 33; histidine, 31; isoleucine, 46; leucine, 104; lysine, 56; methionine, 19; phenylalanine, 32; proline, 83; serine, 56; taurine, 5; threonine, 53; tryptophan, 21; tyrosine, 27 (supplied as the soluble dipeptide glycyl-L-tyrosine); and valine, 53. The amino acid profile was based on human milk protein (Vaminolact: Pharmacia-Upjohn) with phenylalanine and tyrosine provided at their estimated safe levels of intake (House et al. 1997a and 1997b).

All piglets received 150 mL of a TPN/lipid (5:1) solution intravenously at ~10 mL/h until the morning after surgery; this protocol prevented dehydration and provided some nutrients. In Experiment 1, pigs then received TPN and lipid intravenously at 75% of full rate for 12 h and then at full rate. In Experiment 2, pigs then received the oral diet diluted 2.5-fold for 12 h, 1.5-fold for another 12 h and then full strength. After adaptation, the base TPN formulation was provided to all pigs until 1800 h on d 5; at that time, the base TPN solution was replaced with TPN solutions containing one of seven test levels of threonine [Experiment 1: 0.05, 0.10, 0.15, 0.20, 0.40, 0.50 or 0.60 g/(kg·d); Experiment 2: 0.10, 0.20, 0.40, 0.60, 0.80, 1.00 or 1.20 g/(kg·d)]. These threonine intakes corresponded to threonine contents ranging from 3.3 to 40 mg/g of total amino acids (Experiment 1) and 6.7 to 80 mg/g of total amino acids (Experiment 2). Treatment allocation was determined by randomization on study d 0. Isonitrogenous treatment solutions were made by dissolving the appropriate quantities of L-threonine (at test level) and L-alanine (to maintain isonitrogenous treatments) in 75 mL distilled water; these solutions were quantitatively transferred to infusion bags containing 675 mL of concentrated TPN solution. The solutions were filter-sterilized through a 0.22-µm filter (Millipore, Milford, MA) for Experiment 1, but not for Experiment 2.

Tracer infusion, ¹⁴CO₂ collection and analytical procedures.  Phenylalanine flux and oxidation were determined in Experiment 1 by a primed [260 kBq (7 µCi)/kg], constant infusion [130 kBq (3.5 µCi)/(kg·h)] of a tracer solution containing 85.1 MBq (2.3 mCi)/L of L-[1-¹⁴C]phenylalanine [200 MBq (54 mCi)/mmol; Dupont Canada, Mississauga, Canada]. In Experiment 2, the prime infusion was reduced [186 kBq (5 µCi)/kg]; furthermore, because of the uncertainty of our new intragastric model, we extended the constant infusion to 6 h to ensure that adequate plateaux were achieved. Details of the infusion protocol, ¹⁴CO₂ and blood collection procedures have been described previously (House et al. 1997a). Immediately after the infusion, piglets were killed by infusion of 750 mg of sodium pentobarbital into the venous sampling catheter.

Plasma amino acids and the specific radioactivity (SRA) of plasma phenylalanine and tyrosine were measured by reverse-phase HPLC with the use of phenylisothiocyanate derivatives; collection and liquid scintillation counting of radioactive fractions were as previously described (House et al. 1997a).

The calculations for phenylalanine flux, intake, oxidation, nonoxidative disposal, release from protein breakdown, balance and percentage of dose oxidized were as reported previously (House et al. 1997a). SRA for plasma phenylalanine and tyrosine for each time point during the infusion study were plotted. Plateau values were calculated as the mean of the plasma SRA at the time points within the plateau. Plateaux were verified by determining through regression analysis that the slope was not different from zero.

Statistical analyses.  Each experiment was a completely randomized design, with intake of threonine serving as the main treatment effect. Differences among treatments were determined by ANOVA (SAS/STAT, version 6.09, SAS Institute, Cary, NC). If P values were <0.05 for the F-value of the ANOVA model, significant differences between treatment means were assessed by using the Student-Newman-Keuls multiple comparisons procedure. Threonine requirements were established using a two-phase linear regression crossover model, as described previously (Ball and Bayley 1984, Seber 1977). Regression variables included the level of amino acid intake as the independent variable, and phenylalanine oxidation [µmol/(kg·h) and percentage of dose] or phenylalanine balance as the dependent variable. The upper limit of the 95% confidence interval (CI) of the breakpoint was estimated to determine a safe level of intake.

	RESULTS

Abstract Introduction Methods Results Discussion References

The piglets remained healthy and were interested in the environment during the course of both trials. The weight upon arrival (1.75 kg, pooled SD = 0.14) and final weight (2.57 kg, pooled SD = 0.20) were not significantly different among diet levels or between feeding routes. Rates of body weight gain for the 5 d after adaptation and before diet treatment initiation were not different (145 and 146 g/d for parenterally fed and intragastrically fed piglets, respectively).

Experiment 1: Parenteral threonine requirement.  Threonine intake significantly influenced plasma phenylalanine concentration, which declined from 119 to 77 µmol/L as threonine intake increased from 0.05 to 0.10 g/(kg·d), respectively (P < 0.05); no significant changes in phenylalanine concentration were observed with further increases in threonine intake (not shown). Plasma threonine concentration increased significantly from 88 to 647 µmol/L as threonine intake increased from 0.05 to 0.60 g/(kg·d) (P < 0.0001) (not shown).

Values for ¹⁴CO₂ production and plasma SRA are summarized in Table 1. Plateaux in breath ¹⁴CO₂ production, plasma phenylalanine SRA and plasma tyrosine SRA were reached by 2 h after initiation of the primed, constant infusion in all pigs. The CV were <10% for each plateau calculated. The plasma phenylalanine SRA/tyrosine SRA ratio was not different across diet treatments and thus did not affect the breakpoint analysis (Zello et al. 1995).

Phenylalanine flux, intake, nonoxidative disposal and release from protein were not different among threonine treatments (Table 2). Phenylalanine oxidation was significantly influenced by threonine intake, whether expressed as an absolute rate (Fig. 1) or as a percentage of the dose oxidized (Fig. 2). As threonine intake increased from 0.05 to 0.15 g/(kg·d) phenylalanine oxidation declined (P < 0.05). Further increases in threonine intake from 0.20 to 0.60 g/(kg·d) had no significant effect on phenylalanine oxidation (P > 0.05, slope not different from zero). Apparent phenylalanine balance, calculated as the difference between the rates of phenylalanine intake and oxidation, was lowest at the lowest threonine intake and increased (P < 0.05) to a plateau at threonine intakes >0.20 g/(kg·d) (P > 0.05) (Fig. 3).

View larger version (17K): [in this window] [in a new window]   Fig 1. Experiment 1: rate of L-[1-14C]phenylalanine oxidation in piglets receiving total nutrition parenterally with graded intakes of threonine. Pooled SD: 4 µmol/(kg·h).

View larger version (17K): [in this window] [in a new window]   Fig 2. Experiment 1: oxidation of L-[1-14C]phenylalanine as a percentage of dose in piglets receiving total nutrition parenterally with graded intakes of threonine. Pooled SD: 0.5%.

View larger version (18K): [in this window] [in a new window]   Fig 3. Experiment 1: apparent phenylalanine balance in piglets receiving total nutrition parenterally with graded intakes of threonine. Pooled SD: 4 µmol/(kg·h).

To determine threonine requirements, the data points were partitioned between two distinct regression lines (Figs. 1, 2 and 3). This data partitioning was chosen on the basis of models that produced the highest regression coefficients for all of the dependent variables. Other combinations reduced the fit of the model. The breakpoint gave an estimate for the mean requirement of threonine and the corresponding CI. The breakpoint estimate for phenylalanine oxidation rate (Fig. 1) or balance (Fig. 2) (0.20, 95% CI: 0.15-0.25, r² = 0.773), was similar to that for phenylalanine oxidation as a percentage of dose (Fig. 3) (0.19, 95% CI: 0.17-0.21, r² = 0.954). The plasma threonine data did not exhibit a suitable pattern for breakpoint analysis, and thus was not submitted to breakpoint analysis.

Experiment 2: Oral threonine requirement.  As in the parenteral requirement study, plasma phenylalanine concentration declined as threonine intake increased [146 µmol/L at 0.1 g threonine/(kg·d) to 82 µmol/L at 0.2 g threonine/(kg·d); P < 0.05)]; phenylalanine concentration did not change with further increases in threonine intake (not shown). Plasma threonine concentration increased from 27 to 1732 µmol/L with increasing threonine intakes from 0.1 to 1.2 g/(kg·d) (P < 0.001) (not shown).

Values for ¹⁴CO₂ production and plasma SRA in intragastrically fed piglets are summarized in Table 3. As in parenterally fed piglets, plateaux (CV < 10% for each) in breath ¹⁴CO₂ and plasma SRA for phenylalanine and tyrosine were reached by 2 h after initiation of the isotope infusion in all pigs. Furthermore, the plasma phenylalanine SRA/tyrosine SRA ratio was not different across diet treatments.

Qualitatively, results were similar between the studies, except for higher variability among intragastrically fed piglets with respect to kinetics. Phenylalanine flux, intake, nonoxidative disposal and release from protein were not different among oral threonine treatments (Table 4), as in the parenteral requirement study. Phenylalanine oxidation was significantly influenced by threonine intake, whether expressed as an absolute rate or as a percentage of the dose oxidized. As threonine intake increased from 0.10 to 0.40 g/(kg·d), phenylalanine oxidation declined (P < 0.05) (Figs. 4, and 5). Further increases in threonine intake from 0.40 to 1.20 g/(kg·d) had no effect on phenylalanine oxidation (P > 0.05). Apparent phenylalanine balance increased (P < 0.05) to a plateau at threonine intakes >0.40 g/(kg·d) (P > 0.05) (Fig. 6).

View larger version (17K): [in this window] [in a new window]   Fig 4. Experiment 2: rate of L-[1-14C]phenylalanine oxidation in piglets receiving total nutrition intragastrically with graded intakes of threonine. Pooled SD: 12 µmol/(kg·h).

View larger version (17K): [in this window] [in a new window]   Fig 5. Experiment 2: oxidation of L-[1-14C]phenylalanine as a percentage of dose in piglets receiving total nutrition intragastrically with graded intakes of threonine. Pooled SD: 2.1%.

View larger version (18K): [in this window] [in a new window]   Fig 6. Experiment 2: apparent phenylalanine balance in piglets receiving total nutrition intragastrically with graded intakes of threonine. Pooled SD: 12 µmol/(kg·h).

As described above, the data partitioning that produced the highest regression coefficient was used for breakpoint analysis for each set of data (Figs. 4, 5 and 6). The breakpoint estimate for phenylalanine oxidation rate (Fig. 4) or balance (Fig. 5) (0.46, 95% CI: 0.32-0.61, r² = 0.663) was similar to that for phenylalanine oxidation as a percentage of dose (Fig. 6) (0.42, 95% CI: 0.33-0.51, r² = 0.777).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The indicator amino acid oxidation technique has been successfully applied to determine amino acid requirements in piglets and humans (Zello et al. 1995). Because threonine has a complicated metabolic pathway, the direct oxidation technique is inappropriate; if threonine oxidation had been measured by the production of labeled ¹⁴CO₂ from ¹⁴C-threonine, a major underestimate would have been introduced by the sequestration of carbons 1 and 2 of threonine into glycine (Ballevre et al. 1990). Thus, the use of the indicator amino acid oxidation technique is a more suitable method to estimate threonine requirements.

The data reported here indicated that the mean parenteral requirement for threonine, determined by the breakpoint of the two-phase linear regression crossover model, was 0.19 g/(kg·d) when based on phenylalanine oxidation as a percentage of the dose, and 0.20 g/(kg·d) when based on either phenylalanine oxidation rate or balance. Assuming a normal distribution, this mean estimate would meet the needs of only 50% of the population. A safe level of intake would meet the needs of 95% of the population, or the upper 95% CI of the breakpoint estimate. The safe level of parenteral threonine intake was 0.21 g/(kg·d) when percentage of dose oxidized was used as the dependent variable. The safe level was higher [0.25 g/(kg·d)] when phenylalanine oxidation or balance was used as a result of greater variability in the data. Therefore, the least variable data set, that of percentage of dose oxidized, was used to choose the safe level of parenteral threonine intake because it was presumed to reflect the threonine requirement more accurately. Consequently, a threonine intake of 0.21 g/(kg·d) (14 mg/g amino acids) is recommended as a safe level of intake for neonatal piglets receiving TPN.

In a previous experiment, we determined the parenteral lysine requirement in neonatal piglets and attempted to compare it with oral requirements for piglets as described by the NRC (1998) (House et al. 1998). However, NRC requirements cannot be used for direct comparison because they are extrapolated from data that are expressed as a percentage of a corn-soybean meal diet (Kim et al. 1983). Furthermore, within the experiments used to determine the requirements, the protein intakes and digestibilities varied widely. Therefore, to provide a direct comparison for the parenteral threonine requirement, we determined the oral threonine requirement in similar piglets continuously fed identical diets via intragastric catheters. The data for the oral requirement study indicated that the mean oral threonine requirement was 0.46 g/(kg·d) [upper 95% CI: 0.61 g/(kg·d)] when based on either phenylalanine oxidation rate or balance, and 0.42 g/(kg·d) [upper 95% CI: 0.51 g/(kg·d)] when based on phenylalanine oxidation as a percentage of dose. As in the parenteral requirement study, the data set for phenylalanine oxidation as a percentage of dose was the least variable; therefore, an oral threonine intake of 0.51 g/(kg·d) (34 mg/g amino acids) is proposed as a safe intake level for intragastrically fed piglets.

The threonine requirement in orally fed piglets weighing ~2.5 kg has been previously determined by indicator amino acid oxidation using ¹⁴C phenylalanine provided orally (Kim et al. 1983). The "broken-line" regression model analysis produced a mean threonine requirement of 6.0 g/kg diet [~0.58 g/(kg body weight·d)]. Using the same technique, the mean lysine requirement was estimated at 12 g/kg diet, resulting in a threonine/lysine ratio of 1:2 (Kim et al. 1983). According to the NRC (1998), the estimated mean requirement for 3-kg piglets fed a corn-soybean meal diet (growing at 141 g/d) is 8.7 g available threonine/kg diet[~0.68 g/(kg body weight·d)] compared with the mean lysine requirement of 14.0 g/kg diet; thus, the NRC (1998) recommends a ratio of 1:1.61 for 3-kg piglets. In 10-kg piglets, Chung and Baker (1992) proposed an ideal threonine/lysine ratio of 1:1.54. Hence, according to these experiments in orally fed animals, the threonine requirement for young piglets is ~50-65% of the lysine requirement.

On the basis of the ideal amino acid concept (Agricultural Research Council 1983), the threonine requirement in TPN-fed piglets was initially predicted to be 0.4-0.5 g/(kg·d) , which is 50-65% of the lysine requirement [0.79 g/(kg·d)] we had previously determined for TPN-fed piglets (House et al. 1998). A lower estimate of amino acid requirement for TPN-fed piglets compared with that of orally fed piglets might be expected because first-pass metabolism by the splanchnic organs is bypassed during parenteral feeding and the gastrointestinal tract atrophies. Indeed, in the lysine requirement study, a reduced demand (72% of oral requirement) for lysine during parenteral feeding was found [0.79 and 1.09 g/(kg·d) for TPN and oral feeding, respectively, for same age piglets] (House et al. 1998). However, the predicted threonine requirement for TPN-fed piglets of 0.4-0.5 g/(kg·d) was more than twice the mean requirement of 0.19 g/(kg·d) determined in this study. Possible explanations to be considered for this difference are as follows: TPN delivers nutrients intravenously that might affect protein and amino acid metabolism; threonine demand is reduced during TPN because threonine is a critical amino acid for the synthesis of glycoproteins at the surface of the small intestine, which is atrophied.

The effects of parenteral feeding on protein and amino acid metabolism in piglets have been studied. Compared with orally fed and suckled piglets, tissue protein synthesis rates in TPN-fed piglets were altered in muscle, liver and kidney, which would affect whole-body nitrogen, energy and amino acid metabolism (Adeola et al. 1995). In another study, the kinetics of glycine, leucine and phenylalanine, which represent different catabolic sites in the body, have been investigated during both TPN and enteral feeding in human infants (Wykes et al. 1992). That study demonstrated that kinetics of the individual amino acids were influenced by the interaction of several factors, including amino acid intake, route of feeding and route of tracer administration. The effect of the feeding regimen was specific to each amino acid such that glycine flux was lower during parenteral feeding compared with oral feeding, leucine flux was unaffected and phenylalanine flux was higher. Similar to phenylalanine, threonine is catabolized primarily in the liver; because parenteral threonine does not undergo first-pass metabolism by the gut and liver, threonine metabolism might be similarly affected during TPN.

The mucosa of the gastrointestinal tract has one of the most rapid turnover rates of any tissue in the body (Johnson and McCormack 1994). In a recent study in piglets, Stoll et al. (1998) showed that the gastrointestinal tract metabolizes ~35% of dietary leucine, lysine and phenylalanine on first pass, compared with 61% for dietary threonine. Furthermore, they showed that nearly 90% of this metabolized threonine was either secreted as mucosal proteins or catabolized. In humans, gastrointestinal losses of most indispensable amino acids accounted for 14-33% of daily maintenance requirement, but for threonine the contribution was up to 61% (Fuller et al. 1994). These studies suggest that threonine is in high demand by the gastrointestinal tract and that a significant proportion of the whole-body threonine requirement is accounted for by the gastrointestinal tract during oral feeding. The route of nutrient delivery has a dramatic effect on mucosal morphology and function. TPN feeding in piglets leads to a rapid reduction in small intestinal weight, villous height, mucosal depth, disaccharidase activity and the RNA/DNA ratio in spite of the maintenance of normal energy and protein intakes (Goldstein et al. 1985, Shulman 1988). Furthermore, by using piglets and protocol identical to those used in this study, those fed intravenously for 8 d had total mucosal mass that was reduced by 40% compared with piglets fed identical TPN solutions intragastrically (Bertolo et al. 1996). Reduced mucosal mass and function resulting from TPN feeding may account for a large part of the reduced whole-body requirement of threonine.

Although the morphological changes in mucosa during TPN have been well documented, changes to the "protective coat" or mucus of the intestinal mucosa have not been substantially investigated and reported. Structurally, the mucosa is protected by a complex network of glycoproteins (mucus), of which mucin is an important component (Bengmark and Jeppson 1995). The highly glycosylated domains of rat small intestinal mucin have been isolated, and the protein cores contain large amounts of threonine. Indeed, threonine accounted for >40% of the amino acid residues, whereas the majority of the other amino acids were serine and proline (17-24 and 18-19%, respectively) (Carlstedt et al. 1993). Goblet cells throughout the length of the small and large intestine are responsible for the production and maintenance of the protective mucous coating by synthesizing and secreting mucin (Specian and Oliver 1991). Absence of enteral nutrition leads not only to atrophy of mucosal epithelial cells, but also to atrophy of the goblet cells, resulting in a deficient mucous protection layer (Bengmark and Jeppson 1995). Goblet cell atrophy during TPN feeding would reduce mucin production and hence reduce threonine requirements.

Hormonal modulation of nutrient metabolism might also be involved in a decreased threonine requirement. Lack of enteral nutrients decreases secretion of gastrointestinal hormones, which play an important role in normal gut function and growth (Johnson et al. 1975). Decreased antral and serum gastrin in TPN-maintained animals suggests that this hormone might be involved in maintenance of gut mass (Rossi 1986). In addition, growth factors in mammalian breast milk play an important role in the growth and development of the gastrointestinal tract and are missing from TPN solutions and milk formulas (Odle et al. 1996). In a recent study in which threonine was supplied to infants at an identical level via breast milk or formula, a lower plasma threonine concentration and higher threonine oxidation rates were observed in the breast milk-fed infants (Darling et al. 1995). These differences were attributed to the presence of hormones and growth factors in the breast milk. Parenteral nutrition has neither the gastrointestinal nutrients required to trigger hormone release (i.e., gastrin and CCK) nor the dietary trophic factors (i.e., IGF and EGF), that may influence threonine metabolism.

In conclusion, with the use of the indicator amino acid oxidation technique, a safe threonine intake for neonatal piglets receiving all of their nutrients intravenously is 0.21 g/(kg·d) (14 mg/g amino acids), compared with the safe oral intake of 0.51 g/(kg·d) in similar piglets fed identical diets. This requirement was lower than expected, even considering the lower predicted amino acid requirements for parenteral feeding. Our data support the concept that threonine is especially required by the growing gut, which is atrophied during TPN. In addition, our findings suggest that TPN does not affect requirements of all amino acids similarly; the mean lysine requirement in TPN fed piglets [0.79 g/(kg·d)] is 72% of the oral requirement [1.09 g/(kg·d)] (House et al. 1998, NRC 1998), whereas the mean threonine requirement during TPN [0.19 g/(kg·d)] is 45% of the oral requirement [0.42 g/(kg·d)].

	FOOTNOTES

Manuscript received 3 November 1997. Initial reviews completed 5 January 1998. Revision accepted 2 June 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Adeola O., Wykes L. J., Ball R. O., Pencharz P. B. Comparison of oral milk feeding and total parenteral nutrition in neonatal pigs. Nutr. Res. 1995; 15:245-265
• Agricultural Research Council (1981) The Nutrient Requirements of Pigs. Commonwealth Agricultural Bureaux, Slough, UK.
• Anderson G. H., Bryan H., Jeejeebhoy K. N., Corey P. Dose-response relationships between amino acid intake and blood levels in newborn infants. Am. J. Clin. Nutr. 1977; 30:1110-1121[Abstract/Free Full Text]
• Anderson, S. A. & Raiten, D. J. (1992) Safety of amino acids used as dietary supplements. Prepared for Food Safety and Applied Nutrition, Food and Drug Administration, Department of Health and Human Services. Under FDA Contract No. 223-88-2124, Task Order No. 8.
• Ball R. O., Bayley H. S. Tryptophan requirement of the 2.5-kg piglet determined by the oxidation of an indicator amino acid. J. Nutr. 1984; 114:1741-1746
• Ballevre O., Cadenhead A., Calder A. G., Rees W. D., Lobley G. E., Fuller M. F., Garlick P. J. Quantitative partition of threonine oxidation in pigs: effect of dietary threonine. Am. J. Physiol. 1990; 259:E483-E491[Abstract/Free Full Text]
• Bengmark S., Jeppsson B. Gastrointestinal surface protection and mucosa reconditioning. J. Parent. Enteral Nutr. 1995; 19:410-415[Abstract/Free Full Text]
• Bertolo, R.F.P., Pencharz, P. B. & Ball, R. O. (1996) Plasma amino acid and small intestinal changes in piglets fed identical diets enterally or parenterally. FASEB J. 10: A207 (abs.).
• Carlstedt I., Herrmann A., Karlsson H., Sheehan J., Fransson L. A., Hansson G. C. Characterization of two different glycosylated domains from the insoluble mucin complex of rat small intestine. J. Biol. Chem. 1993; 268:18771-18781[Abstract/Free Full Text]
• Chung T. K., Baker D. H. Ideal amino acid pattern for 10-kilogram pigs. J. Anim. Sci. 1992; 70:3102-3111[Abstract]
• Darling, P., Dunn, M., Sarwar, G., Ball, R. O. & Pencharz, P. B. (1995) Threonine kinetics in preterm infants fed formula with varying whey/casein ratios or mother's milk. FASEB J. 9: A730 (abs.).
• Fuller M. F., Milne A., Harris C. I., Reid T. M., Keenan R. Amino acid losses in ileostomy fluid on a protein-free diet. Am. J. Clin. Nutr. 1994; 59:70-73[Abstract/Free Full Text]
• Goldstein R. M., Hebiguchi T., Luk G. D., Taqi F., Guilarte T. R., Franklin F. A., Niemiec P. W., Dudgeon D. L. The effects of total parenteral nutrition on gastrointestinal growth and development. J. Pediatr. Surg. 1985; 20:785-791[Medline]
• Grogan C. K., Janas L. M., Hendrix M. K., Layman D. K., Picciano M. F. Impact of nutrition on postnatal development of serine-threonine dehydratase and branched-chain keto acid dehydrogenase in the rat. Biol. Neonate 1988; 54:224-231[Medline]
• House J. D., Pencharz P. B., Ball R. O. Phenylalanine requirements determined by using L-[1-¹⁴C]phenylalanine in neonatal piglets receiving total parenteral nutrition supplemented with tyrosine. Am. J. Clin. Nutr. 1997a; 65:984-993[Abstract/Free Full Text]
• House J. D., Pencharz P. B., Ball R. O. Tyrosine kinetics and requirements during total parenteral nutrition in the neonatal piglet: the effect of glycyl-L-tyrosine supplementation. Pediatr. Res. 1997b; 41:575-583[Medline]
• House J. D., Pencharz P. B., Ball R. O. Lysine requirement of neonatal piglets receiving total parenteral nutrition as determined by oxidation of the indicator amino acid L-[1-¹⁴C]phenylalanine. Am. J. Clin. Nutr. 1998; 67:67-73[Abstract]
• Janas L. M., Picciano M. F., Hatch T. F. Indices of protein metabolism in term infants fed human milk, whey predominant formula, or cow's milk formula. Pediatrics 1985; 75:775-784[Abstract/Free Full Text]
• Johnson L. R., Lichtenberger L. M., Copeland E. M., Dudrick S. J., Castro G. A. Action of gastrin on gastrointestinal structure and function. Gastroenterology 1975; 68:1184-1192[Medline]
• Johnson, L. R. & McCormack, S. A. (1994) Regulation of gastrointestinal mucosal growth. In: Physiology of the Gastrointestinal Tract (Johnson, L. R., ed.), 3rd ed., vol. 1, pp. 611-642. Raven Press, New York, NY.
• Kim K. I., Elliott J. I., Bayley H. S. Oxidation of an indicator amino acid by young pigs receiving diets with varying levels of lysine or threonine, and an assessment of amino acid requirements. Br. J. Nutr. 1983; 50:391-399[Medline]
• National Research Council (1998) Nutrient Requirements for Swine, 10th ed. National Academy Press, Washington, DC.
• Odle J., Zijlstra R. T., Donovan S. M. Intestinal effects of milkborne growth factors in neonates of agricultural importance. J. Anim. Sci. 1996; 74:2509-2522[Abstract]
• Rigo J., Salle B. L., Putet G., Senterre J. Nutritional evaluation of various protein hydrolysate formulae in term infants during the first month of life. Acta Paediatr. Suppl. 1994; 402:100-104[Medline]
• Rigo J., Senterre J. Optimal threonine intake for preterm infants fed on oral or parenteral nutrition. J. Parent. Enteral Nutr. 1980; 4:15-17[Abstract/Free Full Text]
• Rombeau, J. L., Barot, L. R., Low, D. W. & Twomey, P. L. (1984) Feeding by tube enterostomy. In: Clinical Nutrition, Vol. 1: Enteral and Tube Feeding (Rombeau, J. L. & Caldwell, M. D., eds.), pp. 275-285. W.B. Saunders, Philadelphia, PA.
• Rossi, T. M. (1986) Effects of total parenteral nutrition on the digestive organs. In: Total Parenteral Nutrition: Indications, Utilization, Complications, and Pathophysiological Considerations (Lebenthal, E., ed.), pp. 173-184. Raven Press, New York, NY.
• Schanler R. J., Garza C. Plasma amino acid differences in very low birth weight infants fed either human milk or whey-dominant cow milk formula. Pediatr. Res. 1987; 21:301-305[Medline]
• Scott P. H., Berger H. M., Wharton B. A. Growth velocity and plasma amino acids in the newborn. Pediatr. Res. 1985; 19:446-450[Medline]
• Seber, G.A.F. (1977) Linear Regression Analysis. John Wiley & Sons, New York, NY.
• Shulman R. J. Effect of different total parenteral nutrition fuel mixes on small intestinal growth and differentiation in the infant miniature pig. Gastroenterology 1988; 95:85-92[Medline]
• Specian R. D., Oliver M. G. Functional biology of intestinal goblet cells. Am. J. Physiol. 1991; 260:C183-C193[Abstract/Free Full Text]
• Stoll B., Henry J., Reeds P. J., Yu H., Jahoor F., Burrin D. G. Catabolism dominates the first-pass intestinal metabolism of dietary essential amino acids in milk protein-fed piglets. J. Nutr. 1998; 128:606-614[Abstract/Free Full Text]
• Wykes L. J., Ball R. O., Menendez C. E., Ginther D. M., Pencharz P. B. Glycine, leucine, and phenylalanine flux in low-birth-weight infants during parenteral and enteral feeding. Am. J. Clin. Nutr. 1992; 55:971-975[Abstract/Free Full Text]
• Wykes L. J., Ball R. O., Pencharz P. B. Development and validation of a total parenteral nutrition model in the neonatal piglet. J. Nutr. 1993; 123:1248-1259
• Zello G. A., Wykes L. J., Ball R. O., Pencharz P. B. Recent advances in methods of assessing dietary amino acid requirements for adult humans. J. Nutr. 1995; 125:2907-2915

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences