Lactating Porcine Mammary Tissue Catabolizes Branched-Chain Amino Acids for Glutamine and Aspartate Synthesis

Abstract

The uptake of branched-chain amino acids (BCAA) from plasma by lactating porcine mammary gland substantially exceeds their output in milk, whereas glutamine output is 125% greater than its uptake from plasma. In this study, we tested the hypothesis that BCAA are catabolized for glutamine synthesis in mammary tissue. Mammary tissue slices from sows on d 28 of lactation were incubated at 37°C for 1 h in Krebs buffer containing 0.5 or 2 mmol/L l-[1-¹⁴C]– or l-[U-¹⁴C]–labeled leucine, isoleucine, or valine. Rates of BCAA transport and degradation in mammary tissue were high, with ∼60% of transaminated BCAA undergoing oxidative decarboxylation and the remainder being released as branched-chain α-ketoacids (BCKA). Most (∼70%) of the decarboxylated BCAA were oxidized to CO₂. Rates of net BCAA transamination were similar to rates of glutamate, glutamine, aspartate, asparagine, and alanine synthesis. Consistent with the metabolic data, mammary tissue expressed BCAA aminotransferase (BCAT), BCKA decarboxylase, glutamine synthetase (GS), glutamate-oxaloacetate aminotransferase, glutamate-pyruvate aminotransferase, and asparagine synthetase, but no phosphate-activated glutaminase, activity. Western blot analysis indicated relatively high levels of mitochondrial and cytosolic isoforms of BCAT, as well as BCKA dehydrogenase and GS proteins in mammary tissue. Our results demonstrate that glutamine and aspartate (abundant amino acids in milk protein) were the major nitrogenous products of BCAA catabolism in lactating porcine mammary tissue and provide a biochemical basis to explain an enrichment of glutamine and aspartate in sow milk.

Introduction

Lactating sows have high requirements for branched-chain amino acids (BCAA),⁸ leucine, isoleucine, and valine to support milk production (1). Interestingly, their uptake by porcine mammary glands (76 g/d on d 13–20 of lactation) is much greater than their secretion in milk protein (46 g/d) (2). Thus, the lactating porcine mammary gland may catabolize 30 g BCAA/d (40% of the BCAA taken up from arterial plasma). The possibility that BCAA are degraded by the lactating tissue is supported by the report that BCAA aminotransferase (BCAT) and branched-chain α-ketoacid (BCKA) dehydrogenase (BCKAD) are present in rat mammary tissue, with BCAT being induced ∼10-fold during lactation (3). However, little is known about the metabolic fate of BCAA or their nutritional significance in the lactating mammary gland of any species.

Sow milk contains high concentrations of free and peptide-bound glutamine plus glutamate (4,5), which are crucial for the growth, development, and function of the neonatal small intestine (6–8). The available evidence shows that on d 10 of lactation, the lactating porcine mammary gland takes up 16 g/d glutamine from the arterial circulation (2) but produces 36 g/d glutamine in milk (9). Therefore, extraction of glutamine by the lactating mammary gland is much less than its output in milk and a large amount of glutamine (20 g/d) may be synthesized by mammary tissue to support the production of milk proteins.

Because of the absence of pyrroline-5-carboxylate dehydrogenase and proline oxidase (10), lactating porcine mammary tissue cannot convert arginine, ornithine, or proline into glutamine. Alternatively, BCAA nitrogen might be utilized for glutamine synthesis in lactating mammary gland through the sequential reactions of BCAT and glutamine synthetase (GS), as previously reported for skeletal muscle (11,12) and placenta (13). We tested this hypothesis in the present study using both radiochemical and chromatographic techniques. To compare capacity for BCAA catabolism between mammary and other tissues, we also determined BCAT and BCKAD proteins in skeletal muscle, small intestine, and liver of lactating sows.

Materials and Methods

Materials.

HPLC-grade water and methanol were procured from Fisher Scientific. l-[1-¹⁴C]Leucine, l-[1-¹⁴C]isoleucine, l-[1-¹⁴C]valine, l-[U-¹⁴C]leucine, l-[U-¹⁴C]isoleucine, l-[U-¹⁴C]valine, and l-[U-¹⁴C]glutamate were purchased from American Radioactive Chemicals. [1-¹⁴C]α-Ketoisocaproate (KIC) was obtained from Amersham. Immediately before use, ¹⁴C-labeled BCAA were purified using AG 1-X8 (acetate form, 200–400 mesh) as resin bed (0.6 × 6 cm) and deionized water (2 mL) as eluting solvent (14). [1-¹⁴C]α-KIC was purified by incubation at 25°C with 100 μL of 1.5 mol/L HClO₄ for 30 min, followed by neutralization with 50 μL of 2 mol/L K₂CO₃ (15). Soluene-350, a strong organic base formulated for compatibility with liquid scintillation cocktails and wet tissue solubilization, was obtained from Perkin Elmer. AG 1-X8, Tris, glycine, SDS, Triton X-100, Tween-20, and nitrocellulose membranes were obtained from Bio-Rad. MOPS, SDS running buffer (20×), and NuPage 10% Bis-Tris gel (15-lane) were purchased from Invitrogen. The BCA Protein assay kit and SuperSignal West Dura Extended Duration Substrate were purchased from Pierce. Rabbit anti-mitochondrial and cytosolic BCAT were prepared as described by Hutson et al. (16). Rat anti-BCKAD E1α antibody and rat anti-phosphorylated form of BCKAD E1α antibody were prepared as described by Lynch et al. (17). Mouse anti-GS was obtained from BD Biosciences, Parmingen. All other chemicals were purchased from Sigma Chemicals.

Lactating sows and dissection of tissues.

This study was approved by the Texas A&M University Animal Care and Use Committee. Seven multiparous sows obtained from the Texas A&M University Swine Center and were offspring of Yorkshire × Landrace sows and Duroc × Hampshire boars. During gestation, sows had free access to water and a grain sorghum-soybean meal–based diet (18); dietary contents of metabolizable energy, crude protein, total lysine, total calcium, and total phosphorus were 13,240 kJ/kg, 13.9%, 0.61%, 0.84%, and 0.62%, respectively. During lactation, the sows were fed a grain sorghum-soybean meal–based diet (5) providing 13.24 MJ metabolizable energy/kg, 14.1% crude protein, 0.60% lysine, 0.8% calcium, and 0.60% phosphorus, and nursed 8.86 ± 0.14 piglets/litter (mean ± SEM). On d 28 of lactation, sows were transported from the Swine Center to the Texas A&M University Rosenthal Meat Science and Technology Center (∼5 min driving) and stunned via captive bolt. Immediately before exsanguination, suckled mammary glands were split down the mid-line and mammary tissue was excised from the center portion of 4 glands. Obvious pieces of connective tissue and fat were removed. Mammary tissue was cut into small pieces (0.5 mm in thickness) using curved scissors and placed in oxygenated (95% O₂/5% CO₂) Krebs buffer (pH 7.4) containing 20 mmol/L HEPES and 5 mmol/L d-glucose (12). Immediately after collection of mammary tissue, samples of skeletal muscle (semitendinosus muscle), liver, and small intestine (with luminal content being removed) were obtained and placed in liquid nitrogen. The porcine mammary tissue contains primarily secretory epithelial cells and, to a lesser extent, myoepithelial cells, endothelial cells, adipocytes, fibroblasts, and immune cells (19).

Incubation of mammary tissue.

Incubations were performed in 25-mL polypropylene conical flasks placed in a shaking water bath (70 oscillations/min). Mammary tissue slices (∼60 mg) were incubated at 37°C for 0 or 60 min in 2 mL Krebs buffer (pH 7.4, saturated with 95% O₂/5% CO₂) containing 20 mmol/L HEPES, 5 mmol/L d-glucose, 0.3 mmol/L NH₄Cl, 0.5 or 2 mmol/L each of BCAA, and l-[1-¹⁴C]- or l-[U-¹⁴C]-labeled each of BCAA (4.5 Bq/nmol). In additional experiments, mammary tissue was incubated with 2 mmol/L l-[1-¹⁴C]leucine for 0, 20, 40, and 60 min in the presence of 0 or 5 mmol/L d-glucose and 0 or 1.5 mmol/L l-cycloserine [an inhibitor of BCAA transamination in tissue (12)]. The concentrations of BCAA were chosen because they were close to physiological concentrations in plasma of lactating sows (20) and were sufficient to maintain linear rates of BCAA catabolism during the incubation period. The incubation medium contained other amino acids (except for alanine, aspartate, asparagine, aspartate, glutamate, and glutamine) at concentrations found in the plasma of lactating sows (20). Ammonia was included as a substrate for glutamine synthesis. Incubations were initiated by addition of tissue slices. After 1-h incubation at 37°C, 0.2 mL Soluene-350 was injected through the rubber cap into suspended center-wells and 0.2 mL of 1.5 mol/L HClO₄ acid was injected into the incubation medium to liberate ¹⁴CO₂ that was produced from enzymatic decarboxylation of the corresponding BCKA. After 60-min incubation, suspended wells were transferred to scintillation vials containing 15 mL of cocktail for measurement of ¹⁴CO₂ (14). To measure the net release of [¹⁴C]BCKA by tissue, replacement center-wells each received 0.2 mL Soluene-350, and 0.7 mL of 30% H₂O₂ was introduced into the acidified incubation medium to chemically decarboxylate BCKA. The center-wells were collected following 1-h incubation at 37°C for measurement of ¹⁴CO₂ (14). Rates of net BCAA transamination and oxidative decarboxylation were calculated on the basis of intracellular specific activities of ¹⁴C-labeled BCAA, as previously described (14). Synthesis of amino acids was calculated on the basis of differences in concentrations of amino acids in medium plus cells between 0 and 60 min of incubation. Net synthesis of amino acids as a result of BCAA addition was calculated after correction for the values obtained in the absence of BCAA.

Transport of BCAA by mammary tissue.

BCAA transport by mammary tissue was determined using l-[U-¹⁴C]–labeled leucine, isoleucine, or valine, as described for other tissues (13). Briefly, samples of tissue (∼60 mg) were washed 3 times in oxygenated (95% O₂/5% CO₂; v:v) Krebs buffer containing 20 mmol/L HEPES (pH 7.4) and 5 mmol/L glucose. Samples were then incubated at 37°C for 5 min in 1 mL of oxygenated Krebs-Henseleit bicarbonate buffer consisting of 20 mmol/L HEPES, 5 mmol/L glucose, 0.5 or 2 mmol/L each of BCAA, 0.05 μCi each of l-[U-¹⁴C]BCAA, 0.05 μCi [³H]inulin (an extracellular marker that does not enter cells), and other amino acids (except for alanine, asparagine, aspartate, glutamate, and glutamine) at concentrations found in plasma of lactating sows (20). After the 5-min incubation, the tissues were rinsed thoroughly with fresh KHB buffer and then solubilized in 0.5 mL Soluene-350. The solution was measured for ¹⁴C and ³H radioactivities using a dual-channel counting program (13). The specific activity of each [¹⁴C]-labeled BCAA in the medium was used to calculate BCAA uptake by mammary tissue. Results from preliminary experiments established that BCAA uptake was linear over a 5-min period.

Analysis of amino acids in mammary and other tissues.

Amino acids were determined using HPLC methods involving precolumn derivatization with ophthaldialdehyde (18). Briefly, 100 mg of mammary gland, skeletal muscle, liver, small intestine, kidney, and adipose tissue was homogenized in 2 mL of 1.5 mol/L HClO₄ and then neutralized with 1 mL of 2 mmol/L K₂CO₃. The extract was diluted 10 times with HPLC-grade water and an aliquot (50 μL) was directly used for analysis. Amino acids in the samples were quantified on the basis of authentic standards using the Millennium-32 workstation (Waters).

Determination of enzyme activities.

The activities of BCAT, BCKAD, glutamate-pyruvate aminotransferase (GPT), glutamate-oxaloacetate aminotransferase (GOT), asparagine synthetase (AS), GS, and glutaminase in mammary gland, muscle, liver, and small intestine were determined as we previously described (13,21). For assays of BCAT, BCKAD, GOT, and GPT, samples (∼200 mg) were homogenized in 1 mL of freshly prepared buffer consisting of 50 mmol/L HEPES (pH 7.5), 3 mmol/L EDTA, 5 mmol/L dithiothreitol, 2% (v:v) Triton X-100, and 0.1% (wt:v) protease inhibitor (aprotinin, chymostatin, pepstatin A, and phenylmethylsulfonyl fluoride). For BCKAD assay, the homogenization buffer also contained 1 mmol/L potassium fluoride. Homogenates were centrifuged at 600 × g; 10 min and the supernatant fluid subjected to 3 cycles of freezing in liquid nitrogen and thawing in a 30°C water bath. The BCAT assay mixture consisted of 0.5 mL of 50 mmol/L Tris/HCl buffer (pH 8.6), 0.1 mL of 1.6 mmol/L pyridoxal phosphate, 0.2 mL of 50 mmol/L α-ketoglutarate, 1 mL of 10 mmol/L leucine plus 0.1 μCi purified l-[1-¹⁴C]leucine, and 0.2 mL of tissue extract. The BCKAD assay mixture consisted of 0.1 mL of 20 mmol/L MgCl₂, 0.1 mL of 10 mmol/L dithiothreitol, 0.1 mL of 4 mmol/L thiamine pyrophosphate plus 4 mmol/L coenzyme A plus 10 mmol/L NAD⁺, 0.4 mL of 50 mmol/L potassium phosphate buffer (pH 7.5), 0.1 mL of tissue extract, and 0.1 mL of 1 mmol/L potassium fluoride. Potassium fluoride (an inhibitor of phosphoprotein phosphatase) was used to assess BCKAD activity in its active state. All samples were preincubated for 10 min in a 30°C water bath, after which 0.1 mL of 10 mmol/L KIC plus 0.1 μCi [1-¹⁴C]KIC was added into all tubes. Exogenous dihydrolipoamide dehydrogenase was not supplemented to the BCKAD assay mixture.

The assay mixture (3 mL) for GPT consisted of 80 mmol/L alanine, 7 mmol/L α-ketoglutarate, 0.16 mmol/L NADH, 25 μg l-lactate dehydrogenase, cell extracts (0.2–0.5 mg protein), and 80 mmol/L sodium phosphate buffer (pH 7.6). The assay mixture (3 mL) for GOT consisted of 33 mmol/L aspartate, 7 mmol/L α-ketoglutarate, 0.16 mmol/L NADH, 50 μg l-malate dehydrogenase, cell extracts (0.2–0.5 mg protein), and 80 mmol/L sodium phosphate buffer (pH 7.6). The assay mixture for AS consisted of 20 mmol/L glutamine, 10 mmol/L aspartate, 10 mmol/L MgCl₂, 10 mmol/L ATP, cell extracts (0.5–1 mg protein) and 100 mmol/L Tris-HCl buffer (pH 7.5).

For GS and glutaminase assays, ∼200 mg of samples of various tissues were homogenized in 1 mL buffer consisting of 50 mmol/L Tris (pH 7.9), 2 mmol/L EDTA, and 2 mmol/L dithiothreitol. Homogenates were centrifuged at 10,000 × g; 10 min at 4°C. Then 100 μL of supernatant was mixed with 100 μL of 100 mmol/L MgCl₂, plus 75 mmol/L ATP plus 50 mmol/L phosphocreatine, 100 μL of 100 mmol/L NH₄Cl, 100 μL creatine kinase (100 U/mL), and 100 μL 100 mmol/L l-glutamate plus 0.05 μCi l-[1-¹⁴C]glutamate. The glutaminase assay mixture (0.2 mL) consisted of 20 mmol/L glutamine, 150 mmol/L potassium phosphate buffer (pH 8.2), and cell extracts (0.5–1 mg protein). Protein concentrations in tissue homogenates were measured using the BCA Protein Assay kit (Pierce). Enzyme activities are expressed on the basis of protein content.

Western blot analysis of BCAT, BCKAD, and GS.

Mammary gland, skeletal muscle, liver, and small intestine were pulverized in liquid nitrogen and homogenized lysis buffer containing 20 mmol/L Tris-HCl (pH 7.4), 50 mmol/L NaCl, 50 mmol/L NaF, 50 mmol/L of EDTA, 1% Triton X-100, 1× protease inhibitor cocktail, and 1× phosphatase inhibitor cocktail (Calbiochem). Proteins in homogenates were analyzed using the BCA Protein Assay kit (Pierce). The samples were subsequently diluted with 2× Laemmli buffer (125 mmol/L Tris-HCl, pH 6.8, 4% wt:v SDS, 10% 2-mercaptoethanol, 12% glycerol, and 0.004% wt:v bromphenol blue) and heated in boiling water for 5 min. NuPage 10% Bis-Tris gel from Invitrogen was used for SDS-PAGE separation of proteins. Proteins were then transferred to a nitrocellulose membrane (Bio-Rad) under 12V overnight, using the Bio-Rad Transblot apparatus. Membranes were blocked in 5% fat-free dry milk in Tris-Tween buffered saline (TTBS) (20 mmol/L Tris/150 mmol/L NaCl, pH 7.5, and 0.1% Tween-20) for 3 h and then were incubated with the following primary antibodies overnight at 4°C with gentle shaking: antibodies for mitochondrial BCAT (1:10,000), cytosolic BCAT (1:10,000), total BCKAD E1α (1:10,000), phosphorylated BCKAD E1α (1:50,000), and GS (1:5000). After being washed 3 times with TTBS, the membranes were incubated at room temperature for 2–3 h with secondary antibodies (peroxidase-labeled donkey anti-rat, anti-rabbit or anti-mouse IgG, Jackson Immuno Research, 1:50,000). Finally, the membranes were washed with TTBS, followed by development using Supersignal West Dura Extended Duration Substrate according to the manufacturer's instructions (Pierce). The signals were detected on Fujifilm LAS-3000. Results were expressed as relative pixel intensities (arbitrary unit). Equal numbers of samples from various tissues were always analyzed on the same gel to ensure consistency. Values from replicate gels were normalized to an arbitrary value for a pooled sample included on every gel (22).

Statistical analysis.

Values expressed as mean ± SEM, n = 7. Within each concentration of BCAA, among 3 BCAA groups or among 4 amino acid products, data were analyzed by 1-way ANOVA. For data with unequal variance, as assessed using the Levene's test, log transformation of data was performed before ANOVA. Differences among means were determined by the Tukey's multiple comparison test. Between 0.5 and 2 mmol/L BCAA or between 0 (control) and 1.5 mmol/L l-cycloserine, paired t test was employed, because tissue slices used for incubation were obtained from the same sow. All statistical analyses were performed using SAS (SAS Institute). P-values ≤ 0.05 were considered significant.

Results

BCAA transport and catabolism in mammary tissue.

The porcine mammary tissue had a high capacity to take up BCAA from the incubation medium and rates of leucine transport were higher (P < 0.05) than those for isoleucine and valine (Fig. 1). Increasing extracellular concentrations of BCAA from 0.5 to 2 mmol/L dose dependently increased (P < 0.05) their uptake by mammary tissue (Fig. 1).

FIGURE 1 .

Concentration-dependent enhancement of BCAA transport in lactating porcine mammary tissue. Incubation medium contained 0.5 or 2 mmol/L each of BCAA and l-[U-¹⁴C] isoleucine, leucine, or valine. Data are means ± SEM, n = 7. Within the 0.5 or 2 mmol/L BCAA group, means without a common letter differ, P < 0.05. *Different from corresponding 0.5 mmol/L BCAA mean, P < 0.01.

Using leucine as a prototype BCAA, we found that its transamination, oxidative decarboxylation, and KIC release were linear (P < 0.01) during a 60-min period of incubation in the presence of 5 mmol/L d-glucose (Supplemental Fig. 1). In the absence of glucose from the medium, net leucine transamination was reduced (P < 0.01) by 72% (data not shown).

All BCAA were actively transaminated in this tissue, resulting in the production of BCKA (Table 1). Approximately 60% of transaminated BCAA underwent oxidative decarboxylation, with the remainder (∼40%) being released as BCKA into the incubation medium. Most (∼70%) of the decarboxylated BCAA were oxidized to CO₂. Rates of BCAA catabolism increased (P < 0.05) with their increasing extracellular concentrations from 0.5 to 2 mmol/L. Nitrogenous products of BCAA transamination included alanine, aspartate, asparagine, glutamate, and glutamine, with glutamine and aspartate being the predominant ones (Table 2). Inhibition of BCAA transamination by l-cycloserine reduced (P < 0.05) the production of glutamate, glutamine, alanine, and aspartate in mammary tissue (Fig. 2). Likewise, the absence of glucose from incubation medium reduced (P < 0.01) the synthesis of glutamate, glutamine, alanine, and aspartate by 65–70% (data not shown).

TABLE 1.

Concentration-dependent increase in BCAA catabolism in lactating porcine mammary tissue¹

BCAA	CO2 from all carbons (A)	CO2 from carbon-1 (B)	Net release of BCKA (C)	Net transamination (D = B + C)	Transaminated BCAA released as BCKA (C/D × 100)	Decarboxylated BCAA oxidized as CO2 (A-B)/nB × 100
	nmol/(mg tissue h)				%
0.5 mmol/L BCAA
Isoleucine	1.67 ± 0.11b	0.38 ± 0.02b	0.26 ± 0.02b	0.64 ± 0.04b	40.3 ± 0.62	67.6 ± 2.7
Leucine	2.26 ± 0.14a	0.51 ± 0.03a	0.36 ± 0.02a	0.87 ± 0.05a	40.8 ± 0.57	68.2 ± 3.0
Valine	1.31 ± 0.08c	0.35 ± 0.02b	0.24 ± 0.02b	0.59 ± 0.04b	40.5 ± 0.49	68.8 ± 3.4
2 mmol/L BCAA
Isoleucine	4.02 ± 0.24b*	0.89 ± 0.05b*	0.61 ± 0.04b*	1.50 ± 0.07b*	40.7 ± 0.52	69.7 ± 3.6
Leucine	5.14 ± 0.32a*	1.16 ± 0.08a*	0.75 ± 0.06a*	1.91 ± 0.13a*	39.5 ± 0.69	68.5 ± 2.5
Valine2	3.19 ± 0.21c*	0.85 ± 0.05b*	0.58 ± 0.04b*	1.44 ± 0.08b*	40.1 ± 0.66	69.2 ± 3.1

¹Values are means ± SEM, n = 7. Within each concentration of BCAA, means in a row without a common letter differ, P < 0.05.

²In the equation, n represents the number of noncarboxyl carbons (n = 5 for leucine and isoleucine; n = 4 for valine). *Different from corresponding 0.5 mmol/L BCAA, P < 0.01.

TABLE 2.

Concentration-dependent increase in synthesis of amino acids from BCAA in lactating porcine mammary tissue¹

	Concentrations of BCAA in incubation medium2
Amino acid	0	0.5	2
	nmol/(mg tissue h)
Alanine	0.21 ± 0.02c	0.39 ± 0.03b	0.64 ± 0.05a
Asparagine	0.06 ± 0.01c	0.08 ± 0.01b	0.15 ± 0.02a
Aspartate	0.22 ± 0.02c	0.58 ± 0.04b	1.26 ± 0.07a
Glutamate	0.35 ± 0.02c	0.79 ± 0.06b	1.82 ± 0.12a
Glutamine	0.84 ± 0.07c	1.48 ± 0.10b	2.66 ± 0.24a

¹Values are means ± SEM, n = 7. Means in a row without a common letter differ, P < 0.05.

²Incubation medium contained 0, 0.5, or 2 mmol/L each of BCAA.

FIGURE 2 .

l-Cycloserine inhibited net transamination of BCAA (A) and net synthesis of amino acids (B) in lactating porcine mammary tissue. Incubation medium contained 2 mmol/L each of BCAA and 0 or 1.5 mmol/L l-cycloserine. Data are means ± SEM, n = 7. Within the control or 1.5 mmol/L l-cycloserine group, means without a common letter differ, P < 0.05. *Different from corresponding control mean, P < 0.01. Iso, isoleucine; Leu, leucine; Val, valine.

Activities of BCAA-metabolic enzymes in mammary and other tissues.

All the tissues, including mammary tissue, skeletal muscle, liver, and small intestine, expressed BCAT, BCKAD, GS, GOP, GPT, and AS (Table 3). The BCAT activity in mammary tissue was 258, 538, and 219% higher (P < 0.01) than that in skeletal muscle, liver, and small intestine, respectively. In contrast, the activities of BCKAD and GS were higher (P < 0.01) in the liver than in other tissues. Notably, GS activity in mammary tissue was ∼2.3 times (P < 0.01) that in skeletal muscle and small intestine. The activities of GOT and GPT were high, but those of AS were low in mammary gland and other tissues. Glutaminase activity was not detectable in mammary tissue [< 0.1 nmol/(mg protein·15 min)].

TABLE 3.

Activities of enzymes related to BCAA metabolism and amino acid synthesis in tissues of lactating sows¹

Enzyme	Mammary tissue	Skeletal muscle	Liver	Small intestine
	nmol/(mg protein min)
AS	0.62 ± 0.05b	0.36 ± 0.02c	5.31 ± 0.46a	0.67 ± 0.06b
BCAT	2.36 ± 0.15a	0.66 ± 0.03b	0.37 ± 0.02c	0.74 ± 0.04b
BCKAD	0.21 ± 0.03c	0.18 ± 0.02c	0.63 ± 0.06a	0.42 ± 0.03b
GPT	15.4 ± 1.68c	22.6 ± 2.01b	31.8 ± 2.56a	20.5 ± 1.87b
GOT	32.8 ± 2.51b	10.7 ± 1.06c	36.0 ± 3.22a	12.9 ± 1.59c
Glutaminase	ND2	2.15 ± 0.18c	28.3 ± 2.07a	18.6 ± 1.72b
GS	1.83 ± 0.14b	0.76 ± 0.59c	12.4 ± 0.95a	0.81 ± 0.09c

¹Values are means ± SEM, n = 7. Means in a row without a common letter differ, P < 0.05.

²ND, Not detected [< 0.1 nmol/(mg protein·15 min)].

Isoforms of BCAT in mammary and other tissues.

The mitochondrial and cytosolic isoforms of BCAT were detected in the mammary tissue, skeletal muscle, liver, and small intestine of lactating sows (Fig. 3). The protein levels of both mitochondrial and cytosolic BCAT in mammary gland were higher (P < 0.01) than those in skeletal muscle, liver, and the small intestine (Table 4). The level of mitochondrial BCAT was higher (P < 0.01) in skeletal muscle than in the small intestine, but the opposite was observed for the cytosolic BCAT. The protein levels of both BCAT isoforms were low in the liver.

FIGURE 3 .

Representative western bots of mitochondrial BCAT protein (A), cytosolic BCAT protein (B), total BCKAD protein (C), phosphorylated BCKAD protein (D), and GS protein (E) in mammary tissue, skeletal muscle, liver, and small intestine of lactating sows. Antibodies against BCKAD E1α subunit and phosphorylated form of E1α were used for western blot analysis. The intensity of the images was determined under the same conditions.

TABLE 4.

Relative levels of proteins related to BCAA metabolism and amino acid synthesis in tissues of lactating sows¹

Protein	Mammary tissue	Skeletal muscle	Liver	Small intestine
	arbitrary units
BCATm	488 ± 53a	157 ± 5c	270 ± 12b	53 ± 3d
BCATc	112 ± 15a	8 ± 1c	11 ± 1c	68 ± 7b
Total BCKAD	360 ± 40b	13 ± 3d	52 ± 6c	887 ± 102a
P+BCKAD	699 ± 66b	70 ± 17c	118 ± 13c	3710 ± 407a
GS	177 ± 13b	11 ± 1c	2033 ± 326a	15 ± 1c

¹Values are means ± SEM, n = 7. Means in a row without a common letter differ, P < 0.05.

Total BCKAD and phosphorylated BCKAD proteins in mammary and other tissues.

BCKAD protein was detected in the mammary tissue, skeletal muscle, liver, and small intestine of lactating sows (Fig. 3). The protein level of total BCKAD in mammary gland was higher (P < 0.05) than that in skeletal muscle and liver, but was lower (P < 0.05) than that in the small intestine (Table 4). Phosphorylated BCKAD E1α was found in mammary gland and small intestine but was not detectable in skeletal muscle and liver. The level of phosphorylated BCKAD was relatively high in the small intestine but nondetectable in skeletal muscle and liver (Table 4). The ratio of phosphorylated BCKAD:total BCKAD in mammary gland (0.205 ± 0.02) differed from that in small intestine (0.408 ± 0.03) (P < 0.01).

GS protein levels in mammary and other tissues.

GS protein was detected in the mammary tissue, skeletal muscle, liver, and small intestine of lactating sows (Fig. 3). The protein level of GS in mammary tissue was higher (P < 0.05) than that in skeletal muscle and the small intestine but was lower (P < 0.05) than that in liver. Based on quantification of western blots, GS protein in liver was 11.5-, 183-, and 135-fold higher than that in mammary tissue, skeletal muscle, and small intestine, respectively (Table 4).

Concentrations of amino acids in mammary and other tissues.

In mammary tissue, concentrations of BCAA and all other essential amino acids, as well as serine, cysteine, tyrosine, β-alanine, asparagine, ornithine, and citrulline, were relatively low (<0.25 nmol/mg tissue), but concentrations of taurine, glutamate, aspartate, glycine, and alanine were particularly high (2.2–4.9 nmol/mg tissue), with arginine, glutamine, and proline being in between [0.6–0.86 nmol/mg tissue (Supplemental Table 1)]. For comparison, concentrations of glutamine, glutamate, aspartate, alanine, and taurine in semitendinosus muscle of nonlactating sows was 3.87 ± 0.35, 2.94 ± 0.31, 0.96 ± 0.15, 4.52 ± 0.43, and 4.05 ± 0.37 nmol/mg tissue, respectively. Consistent with a low activity of hepatic BCAT and a high activity of hepatic GS, concentrations of BCAA and glutamine were higher (P < 0.05) in the liver than in mammary tissue, skeletal muscle, and small intestine. Interestingly, concentrations of aspartate were much lower (P < 0.05) in skeletal muscle compared with the other tissues (Supplemental Table 1).

Discussion

The mammary gland of lactating sows produces 125% more glutamine in milk than its uptake from arterial plasma (2,9), whereas the uptake of arterial glutamate by the mammary gland is fairly matched by the amount of glutamate in milk (23). It is unlikely that the extra glutamine is derived from proline, because porcine mammary tissue lacks the proline oxidase enzyme that would be needed for this activity (10). There was increased synthesis of glutamine when BCAA were provided to porcine mammary tissue, suggesting a de novo synthesis from α-ketoglutarate and BCAA nitrogen. Rates of transamination of BCAA with α-ketoglutarate [primarily from glucose via pyruvate carboxylase and possibly other anapleurotic reactions (24,25)] in porcine mammary tissue (Table 1) were high relative to perfused skeletal muscle of rats (26). This reaction generated glutamate, which was subsequently amidated to form glutamine (Fig. 4). The gland appeared to contain sufficient BCAT activity to support glutamine synthesis when ammonia was provided in the medium (Table 1). Interestingly, phosphate-activated glutaminase, the major enzyme initiating glutamine catabolism in the mitochondria, was absent from the mammary tissue of lactating sows (Table 3), as previously reported for lactating rats (27). This result was not due to an artifact, because we readily detected glutaminase activity in other tissues (Table 3). The lack of glutamine degradation via the glutaminase pathway maximizes the availability of newly synthesized glutamine for the production of milk proteins and indicates that ammonia needed for the amidation of glutamate to glutamine is likely to be obtained from the circulation or nonglutaminase pathways. Glutamine made in the alveolar cells is secreted into the mammary alveolar lumen and duct system and then released to piglets after an oxytocin surge (7). The presence of the glutamine-synthetic pathway in the absence of glutamine degradation is consistent with the high abundance of glutamine in the milk of mammals, including pigs (e.g. 2–4 mmol/L free glutamine vs. 0.4–1 mmol/L free glutamate) (5,9) and humans (e.g. 1.4 mmol/L free glutamine vs. 0.3 mmol/L free glutamate) (28), despite insufficient uptake of glutamine from the arterial circulation.

FIGURE 4 .

Pathways of BCAA catabolism and amino acid synthesis in lactating porcine mammary tissue. Enzymes that catalyze the indicated reactions are: 1) BCAT; 2) BCKAD; 3) GS; 4) GOT; 5) GPT; 6) AS; 7) glucose metabolism via glycolysis and the Krebs cycle; and 8) protein synthesis. The corresponding α-ketoacids of leucine, isoleucine, and valine are α-KIC, α-keto-β-methylvalerate, and α-ketoisovalerate, respectively. Mammary tissue takes up BCAA and releases glutamine through specific transporters on the plasma membrane.

Glutamate transaminates with pyruvate and oxaloacetate to form alanine and aspartate, respectively, in mammary gland (Table 2). Aspartate is then converted into asparagine (Fig. 4), as reported for porcine enterocytes (9). The synthesis of these amino acids is consistent with the presence of GPT, GOT, and AS in the mammary gland of lactating sows (Table 3). Interestingly, although the activities of GPT and GOT were greater than those of GS, glutamate and glutamine syntheses were higher than formation of alanine or aspartate in mammary tissue (Table 2), as previously reported for the porcine placenta (13). Of particular note, the generation of products of GOT appears to predominate in lactating mammary tissue (Table 3). In contrast to the skeletal muscle of nonlactating mammals (11) and chickens (12), GPT activity exceeds GOT activity and alanine is a major product of skeletal muscle metabolism.

The synthesis of aspartate and asparagine is also of nutritional importance, because the uptake of these 2 amino acids by the mammary gland of lactating sows accounts for only 50% of their output in milk (2). Aspartate plus asparagine are the 3rd most abundant nonessential amino acids in porcine milk protein (4,15). Thus, by stimulating their synthesis, BCAA play a role in supporting protein synthesis in lactating mammary tissue. Modulation of BCAA catabolism and, therefore, glutamine and aspartate synthesis may play a hitherto unrecognized important role in regulating milk synthesis.

There is a distinct pattern of distribution of BCAT activity in the mammary gland, skeletal muscle, liver, and small intestine of lactating sows (Table 3) compared with that in nonlactating rats, humans, and monkeys (3,29,30). The lactating porcine mammary gland had a particularly high BCAT activity (Table 3), which explains our findings that BCAA were transaminated extensively in this tissue and promoted glutamine synthesis (Table 3). Consistent with this notion, concentrations of free BCAA were low, but concentrations of free glutamate and glutamine were high in mammary tissue (Supplemental Table 1). The mammary tissue also had a relatively high activity of BCKAD that oxidatively decarboxylated BCKA to yield acetyl-CoA (Table 3), which was further oxidized via the Krebs cycle (31). Consistent with a high BCKAD activity, 60% of transaminated BCAA were decarboxylated in lactating porcine mammary tissue (Table 1). Furthermore, the use of [1-¹⁴C]- and [U-¹⁴C]-labeled BCAA reveals that most (70%) of the decarboxylated BCKA were oxidized to CO₂ via the Krebs cycle in the lactating porcine mammary tissue. Thus, during lactation, the mammary gland is a physiologically significant site of BCAA catabolism. Acetyl-CoA derived from leucine and isoleucine may be utilized for the synthesis of fatty acids in mammary tissue, whereas succinyl-CoA produced from isoleucine and valine catabolism may provide carbon skeletons for the formation of nonessential amino acids such as glutamine (32).

The activity of BCAT in the small intestine of lactating sows (Table 3) was lower than that of postweaning pigs (15), enterally-fed infant pigs (33), and adult humans (29). Interestingly, the activity of BCKAD in the small intestine of lactating sows (Table 3) was lower than that of postweaning pigs (15) and adult humans (29) but higher than that of enterally-fed infant pigs (33). A large proportion of BCKAD in the small intestine of lactating sows was in the phosphorylated form (inactive form) (Table 4) as reported for other species (32), which would further limit the irreversible catabolism of transaminated BCAA (34). These results suggest that the entry of dietary BCAA and their carbons into the portal circulation may be higher in lactating sows compared with nonlactating counterparts. Because of a very low activity of hepatic BCAT (Table 3), it is likely that BCAA catabolism is limited or negligible in the liver of lactating sows, as reported for rats (31) and humans (29). This would maximize the escape of BCAA or their carbon skeletons from the liver into the systemic circulation for uptake by extrahepatic tissues. In nonlactating mammals (including growing pigs), skeletal muscle is a major site for initiating BCAA catabolism (15,31). However, under the same assay conditions, BCAT and BCKAD activities in the skeletal muscle of sows during lactation (Table 3) were <15 and 10%, respectively, of those in growing pigs (15). Thus, it can be surmised that in lactating sows, circulating BCAA are preferentially channeled into mammary gland rather than skeletal muscle to support glutamine and aspartate synthesis. This also raises a possibility that glutamine synthesis by the muscle of lactating sows may be reduced. In this regard, it is noteworthy that a recent study reported that concentrations of glutamine in skeletal muscle of lactating sows were <50% of that in nonlactating counterparts (35). Because glutamine has a potential to stimulate net protein synthesis in skeletal muscle in vitro (36), increasing intramuscular levels of glutamine in vivo may be an effective strategy to improve nitrogen balance in lactating sows and other mammals. Collectively, our findings suggest coordinate tissue-specific metabolism of BCAA in lactating mammals to maximize the provision of BCAA for utilization by their mammary glands at the expense of glutamine synthesis in skeletal muscle.

Our findings have important implications for lactation and neonatal nutrition. Milk-derived glutamine is a major substrate for the intestinal synthesis of citrulline and arginine (37), which is a nutritionally essential amino acid for neonates (38) but is remarkably deficient in the milk of most mammals, including pigs and humans (4,5,39). Enterally administered BCAA may effectively increase their concentrations in plasma of lactating sows. There is evidence that supplementing BCAA to the diet of lactating sows can promote piglet growth performance. For example, increasing the content of l-valine from 0.80 to 1.2% in the diet (containing 1.57% l-leucine and 0.68% l-isoleucine) for lactating sows augmented (P < 0.04) litter weaning weight by 5.5% (40). Additionally, increasing the content of total BCAA from 2.92 to 3.62% in the diet of lactating sows enhanced (P < 0.02) litter weight gain between farrowing and weaning by 5.7% (41). The underlying mechanisms may include in part enhancements of mammary syntheses of glutamine and aspartate, whose uptake from arterial plasma by the lactating gland meets <50% of their requirements for the production of milk proteins (2,9). It should be borne in mind that effects of dietary supplementation with any amino acid on physiological responses (including production performance of lactating sows) depend on supplemental doses and the contents of other amino acids in the diet (42,43). Future studies involving rat and pig models are warranted to determine the effect of BCAA supplementation on concentrations of protein and glutamine in milk, as well as the volume of milk production. Because impaired lactogenesis occurs in women in response to various stress conditions (44), BCAA may provide a solution to this nutritional problem.

In conclusion, the results of this study indicate that BCAA were catabolized extensively and stimulated the synthesis of glutamine and aspartate in lactating porcine mammary tissue in vitro. There was no detectable glutaminase in mammary tissue, suggesting that the lack of glutamine catabolism in the porcine gland maximizes the availability of newly synthesized glutamine for the production of milk proteins and for secretion into the mammary alveolar lumen. Our novel findings provide one biochemical explanation for the high abundance of glutamine in sow milk, as well as valuable insight into the role of BCAA in lactation and neonatal nutrition.