Introduction

The partitioning of methionine is of developmental importance as both protein synthesis and transmethylation are significant metabolic processes requiring substantial dietary methionine. Free methionine is either partitioned to protein, or it is adenosylated to S-adenosylmethionine, which serves as the methyl precursor for more than 50 transmethylation reactions (Schubert et al. 2003) including creatine and phosphatidylcholine synthesis, as well as DNA and protein methylation. However, while transmethylation and protein synthesis are integral to the requirement of methionine (Bertolo and McBreairty 2013), the separate requirements of these processes for methionine have not been described. The paucity of data regarding the contribution of transmethylation to the methionine requirement is likely due to the complexity of methionine metabolism. After undergoing transmethylation, homocysteine is partitioned either toward transsulfuration or remethylation. Transsulfuration is initiated by cystathionine-β-synthase (CBS) and represents the irreversible oxidation of homocysteine to cysteine; remethylation represents the reformation of methionine by labile methyl sources. Transsulfuration demands have been widely shown to affect the methionine requirements of piglets (Shoveller et al. 2003a), children (Huang et al. 2012) and adults (Storch et al. 1990), and the inclusion of dietary cysteine noticeably spares methionine for protein synthesis. Indeed, the nutritional importance of transsulfuration is underscored as the presence of cysteine reduces the methionine requirement by 40 % [from 2.8 mmol/(kg day) to 1.7 mmol/(kg day)] in piglets (Shoveller et al. 2003a), similar to data in infants (Albanese et al. 1949). However, the ability of remethylation, and its methyl precursors, to spare the methionine requirement has not been studied. Intakes of labile methyl groups do alter methionine metabolism in adults (Mudd and Poole 1975; MacCoss et al. 2001), and thus we rationalized that remethylation could overcome methionine restriction by enhancing methionine availability in the neonate.

Remethylation is a complex process involving two distinct pathways with a common function: to transfer a labile methyl group to homocysteine, and reform methionine. (1) Methionine synthase (MSyn) (with cyanocobalamin as a cofactor) transfers the labile methyl group from 5-methyltetrahydrofolate (folate) to homocysteine. (2) Betaine:homocysteine methyltransferase (BHMT) transfers a methyl group from betaine, an oxidation product of the essential nutrient choline, to form dimethylglycine (DMG). We have recently shown that diet devoid of folate, betaine, and choline affect specific transmethylation reactions in liver (Robinson et al. 2014). However, the capacity of dietary methyl donors (via remethylation) to overcome methionine restriction has yet to be determined.

In the late 1980’s, a method was developed to calculate in vivo rates of whole body remethylation, transmethylation, and transsulfuration using the dilution of two methionine isotopomers (Storch et al. 1988, 1991), which was later modified to correct for intracellular methionine enrichment (MacCoss et al. 2001; Riedijk et al. 2007). The method readily detected changes in methionine metabolism after betaine supplementation in adult men (Storch et al. 1991), sulfur amino acid deficiency in piglets (Riedijk et al. 2007), and vitamin B6 deficiency in adults (Davis et al. 2005). This method is employed in the current study to determine the capacity of dietary methyl donors to contribute to whole body remethylation and transsulfuration, as well as their effects on partitioning between transmethylation and protein incorporation in the piglet. The piglet is an ideal model for the infant with respect to methionine metabolism due to similarities in milk composition (Davis et al. 1994), methionine requirement relative to growth rate (Huang et al. 2012; Shoveller et al. 2003a), cysteine sparing capacity (Albanese et al. 1949; Shoveller et al. 2003a), and homocysteine response to methionine intake (Shoveller et al. 2004). We report the in vivo effects of dietary methyl donors on whole body remethylation, transmethylation, and transsulfuration, as well as on the in vitro capacity of the hepatic enzymes responsible for homocysteine elimination. At last, we investigated the utility of a [2H3-methyl]methionine infusion to trace the whole body catabolism of de novo synthesized phosphatidylcholine to free choline and betaine.

Materials and methods

Chemical reagents and isotopes

All chemicals and reagents were of the highest available purity and were obtained from Sigma, Fisher Scientific, Alfa Aesar, and Ajinomoto. The [14C]methyl-tetrahydrofolic acid ([14C]MTHF), barium salt was from Amersham Biosciences. L-[3-14C]serine ([14C]serine) and [14C-methyl]N,N,N-trimethyl glycine ([14C]betaine) were acquired from Moravek Biochemicals. [3,3,3′,3′,4,4,4′,4′-2H]homocystine ([2H8]homocystine), [13C1]methionine and [2H3-methyl]methionine were obtained from Cambridge Isotope Laboratories.

Piglets and surgical procedures

All procedures performed in this manuscript were in accordance with the ethical standards by the Memorial University of Newfoundland animal care committee in accordance with the Canadian Council on Animal Care. Sixteen Yucatan miniature piglets (eight male, eight female) were weighed and removed from the sow in pairs at 4–9 days of age on study day 0 (1.7 ± 0.3 kg). Piglets were anesthetized with an intramuscular injection of acepromazine (0.5 mg/kg; Atravet; Ayerst Laboratories), ketamine hydrochloride (20 mg/kg; Rogarsetic Rogar STB), and atropine (20 mg/kg; Rafter Dex Canada) prior to endotracheal intubation. Anesthesia was sustained using 1–2 % isoflurane (Abbott Laboratories Ltd), delivered with oxygen via the endotracheal tube. Piglets were fitted with catheters in the jugular and femoral veins, as well as the stomach (for intragastric feeding). Intravenous doses of analgesic (0.03 mg/kg buprenorphine; Temgesic, Schering-Plough Ltd) and antibiotics (trimethoprim and sulfadoxine; Borgal, Intervet) were administered postoperatively. Animals were fitted with mesh jackets and individually housed in metabolic cages equipped with a swivel and tether system that allowed for free movement within wire cages, and facilitated continuous infusion of diet (Lomir Biomedical Inc.). The room was lit from 0800 to 2100 h at 28 °C and supplemental heat was provided with heat lamps. The piglets were closely monitored for daily weight gain, and surgical incision sites were treated daily with topical antibiotics (Hibitane Veterinary Ointment, Ayerst Laboratories). Blood was sampled daily during the 8-day protocol. Plasma was immediately separated and stored at −20 °C for further analyses. Piglet weights were used to adjust for the daily rate of feeding.

Dietary regimen

Diet was provided using medical-grade, pressure-sensitive peristaltic pumps (Baxter Corporation). Intravenous feeding into the jugular vein began immediately after surgery at a rate of 5.7 mL/(kg h), which increased to 8.5 mL/(kg h) on the morning of day 1. Enteral feeding commenced on the evening of day 1 at 11.3 mL/(kg h), for the remainder of the study. Amino acids were provided in crystalline form at 16 g/(kg day) in a profile similar to Vaminolact (Shoveller et al. 2003b; Brunton et al. 2012). From days 0 to 5, the dietary amino acid composition was adequate (National Research Council 2012); on the evening of day 5, dietary methionine was restricted to ~80 % of the estimated piglet requirement in the presence of excess cysteine (Shoveller et al. 2003a) and made isonitrogenous with alanine. Piglets remained on this methionine-restricted diet (0.73 g methionine/kg diet) for the remainder of the study (Online Resource 1). Minerals were provided at >100 % of the NRC requirements for a piglet (National Research Council 2012). Soybean oil (J.M. Smucker Co.) was provided as the lipid source to minimize choline and betaine intakes (Engel 1943), and was delivered intragastrically with the elemental diet at a rate of 0.4 mL/(kg h). Overall, diets provided 1.0 MJ/(kg day) of metabolizable energy with dextrose and lipid each supplying 50 % of the non-protein energy intake.

The piglets were randomly assigned to two dietary groups with one group (n = 8) maintained on a methyl-deficient (MD−) diet (i.e., without folate, choline, and betaine) and the other group (n = 8) was fed the same diet made methyl-sufficient (MS+) by providing folate [38 µg/(kg day)], choline [60 mg/(kg day)], and betaine [238 mg/(kg day)] in the diet. These sufficient dietary concentrations of folate and choline were based on the NRC requirements for pigs of this age (National Research Council 2012). The level of sufficient dietary betaine was chosen to provide an equimolar amount of methyl groups that would have been provided if methionine were fed at the requirement (Shoveller et al. 2003b). Vitamins were provided at >110 % of piglet requirements (National Research Council 2012) to the MS+ piglets with a pediatric commercial solution (Infuvite Pediatric, Baxter Corporation). The MD− piglets received all vitamins from the same commercial solution except folate, biotin, and cyanocobalamin; biotin (20 µg/mL) and cyanocobalamin (1 µg/mL) were provided at the same rates as the MS+ piglets using solutions made in the laboratory.

On day 7, piglets received equimolar infusions of [13C1]methionine and [2H3-methyl]methionine via the gastric catheter with their respective diet. The infusion commenced with a 19.97 µmol/kg priming dose, and continued at 9.97 µmol/(kg h) for 8 h. The stable isotope infusion took place in air-tight metabolic chambers and chamber exhaust (i.e., pig breath) was either bubbled into 1 M NaOH to measure [13C1]methionine oxidation or was diverted a CO2 analyzer (AEII Technologies). After the infusion, piglets were returned to their metabolic cages and continued on their MD− or MS+ diet. The following day, the piglets were euthanized by exsanguination and liver samples were rapidly removed and freeze-clamped. Tissues remained at −80 °C until analysis.

Chromatography

The plasma enrichments of [13C1]methionine and [2H3-methyl]methionine were quantified by GCMS. A 50 µL aliquot of plasma was mixed with 15 µL of 0.5 M phosphate buffer (pH 8), 5 µL of 1 mM [2H8]homocystine and 5 µL of 10 % tri-butylphosphine (in DMSO) in a microcentrifuge tube. Samples were left at room temperature for 30 min, and 135 µL of 133 mM pentafluorobenzyl bromide was added to the samples which were mixed prior to being placed in a 60 °C oven for 30 min during alkylation. Thorough mixing with 335 µL of hexane terminated the reaction, and a 2 µL aliquot of the organic fraction was analyzed using a GCMS (Agilent Technologies) fitted with a DB-5MS column (0.25 mm × 30 m × 0.22 µm) (Agilent Technologies). The GC (6890 N Network GC System) oven was preheated to 50 °C and equipped with helium carrier gas. Oven temperature was maintained initially for 3 min and subsequently ramped at 30 °C/min until it reached 280 °C, where it was held for 4 min. Column effluents were immediately ionized with 70 eV of EI collision energy upon entering a quadrupole MS (5973 inert Mass Selective Detector). The m/z of methionine was 328 [M + 0], which was compared to 329 [M + 1] and 331 [M + 3] for [13C1]methionine and [2H3-methyl]methionine, respectively. The m/z of [13C1]homocysteine was also monitored during the infusion at 494 [M + 0] and 495 [M + 1]. Sample reduction was confirmed by detection of reduced [2H8]homocystine at m/z 498 (i.e., [M + 4] homocysteine).

The incorporation of [2H3-methyl] from [2H3-methyl]methionine into choline, betaine, and dimethylglycine (DMG) was measured as described elsewhere (Holm et al. 2003; Kirsch et al. 2010). Briefly, plasma samples were mixed with 3× volume of acetonitrile and centrifuged at for 5 min at 5000×g. A 1 µL aliquot of supernatant was injected into an HPLC column (Atlantis HILIC Silica 3 µm, 2.1 × 100 mm, Waters Corporation) and the quaternary ammonium compounds were eluted using an isocratic solvent system consisting of ammonium formate (15 mM, pH 3.5; 17.5 %) and acetonitrile (82.5 %) for 6 min at 0.6 mL/min with an HPLC MS/MS (Waters Alliance 2795, Micromass Ultima Triple-Quad MS, Waters Corporation). The compounds were detected in multiple-reaction monitoring mode using the following transitions: [2H3]betaine 121 → 62 [M + 3], betaine 118 → 59, [2H3-methyl]choline 107 → 63 [M + 3], choline 104 → 60, [2H3]DMG 101 → 55, DMG 104 → 58. Choline and betaine concentrations were quantified using internal standards and the following transitions: [2H11]betaine 129 → 68 [M + 11] and [2H9-methyl]choline 113 → 69 [M + 9].

Methionine and other amino acid concentrations were measured by HPLC using phenylisothiocyanate derivatives as described elsewhere (Bidlingmeyer et al. 1984). Homocysteine and cysteine concentrations were measured by the HPLC method of others (Vester and Rasmussen 1991).

Hepatic MSyn, BHMT, and CBS activities

The hepatic capacity of the homocysteine eliminating enzymes were measured in vitro by the methods of others (Finkelstein and Mudd 1967; Koblin et al. 1981; Garrow 1996; Taoka et al. 1998). Briefly, 0.5 g of frozen liver was homogenized in 2.5 mL homogenization buffer (50 mM potassium phosphate dibasic buffer, pH 7.0) and centrifuged at 20,000×g for 30 min at 4 °C. The protein concentration of the supernatant was measured using biuret reagent (Gornall et al. 1949). All of the following assays have been shown to be linear with respect to time and protein concentration in piglet tissues (MacKay et al. 2012). Blank samples containing only homogenization buffer were completed for each activity assay to account for background radioactivity. Samples were analyzed in triplicate.

Methionine synthase (MSyn) activity was adapted from Koblin et al. (1981). Briefly, MSyn activity was initiated by combining 100 µL of liver supernatant with 100 µL of MSyn substrate mixture, which included 5-[14C]methyltetrahydrofolic acid. The reaction was stopped with 400 µL of cold deionized water and placed on ice for the immediate isolation of the reaction product ([14C-methyl]methionine) using ion-exchange (AG 1–x8 resin, 200–400 mesh, chloride form, Biorad). MSyn activity is expressed as nmol methionine/(min mg protein) (MacKay et al. 2012).

BHMT activity was measured by modifying the methods of others (Finkelstein and Mudd 1967; Garrow 1996). Briefly, BHMT activity was initiated in 100 µL of liver supernatant by addition of 100 µL of 10 mM [14C]betaine (3.7 MBq), and was terminated with 2.5 mL of cold deionized water. A 2 mL aliquot of terminated reaction solution was applied to an ion-exchange column (AG 1-x4, 200–400 mesh, hydroxide form, Biorad) and the BHMT products, [14C-methyl]methionine, and [14C]dimethylglycine (DMG), were eluted with 3 mL of 1.5 N HCl after washing unreacted [14C]betaine. BHMT activity is expressed as nmol methionine/(min mg protein) (MacKay et al. 2012).

CBS activity was measured by the method of Taoka et al. (1998). Briefly, CBS activity was measured by mixing 150 µL of liver supernatant with 400 µL of CBS substrate mixture including 300 mM [14C]serine (3.7 MBq). The reaction was terminated by 300 µL of cold 15 % TCA and the reaction mixture was centrifuged for 10 min at 9000×g. A 500 µL aliquot of deproteinized supernatant was immediately applied to an ion-exchange resin (AG 50w-x4 Resin, 200–400 mesh, H+ form, Biorad) and the product, [14C]cystathionine, was eluted with 3 N ammonium hydroxide after washing unreacted [14C]serine. A 1 mL aliquot of the [14C]cystathionine-containing effluent was counted for radioactivity and the CBS activity was expressed as nmol cystathionine/(min mg protein) (MacKay et al. 2012).

Expression of MSyn, BHMT, and CBS genes in liver

RNA was extracted from 50 to 100 mg of frozen liver using RNEasy Mini Kit according to manufacturer’s instructions (Qiagen, RNEasy Mini Kit). RNA concentration and quantity were assessed with a spectrophotometer (OD 260/280) and cDNA was synthesized using reverse transcriptase (Qiagen, QuantiTect ReverseTranscription kit). Quantitative real-time PCR was performed using an Eppendorf Mastercycler® ep realplex and SYBR Green binding chemistry (BioRad). The following primer pairs were used: CBS (GenBank accession No. BP446783) forward primer: CCCCCTTGCTCACTACGACA, CBS reverse primer: TTGCACTGCAGCAGGATCTC; BHMT (GenBank Accession No. NM_001204771.1) forward primer: GTTCGCCAGCTTCATC, BHMT reverse primer: CTCCAGCTTGTCCT; MSyn (GenBank Accession No. AF276463.1) forward primer: CTGGAAGCCTTTCTTTGACG, MSyn reverse primer: TTTGGCCTCTTCACCAGGTTG; β-actin (GenBank Accession No. DQ845171.1) forward primer: CCCAGCACGATGAAGA, β-actin reverse primer: CGATCCACACGGAGTA. Gene expression was normalized to the expression of β-actin (Pfaffl 2001).

Calculations

Methionine kinetics

The flux (Q) of the [13C1]methionine (Q C) and [2H3-methyl]methionine (Q M) tracers were calculated as:

$$ Q \, = \, i(E_{\text{IT}} /E_{\text{Met}} - \, 1), $$

where i is the isotope infusion rate, E IT is the enrichment (MPE) of the infused tracer, and E Met is the methionine enrichment in plasma (MPE). To account for intracellular methionine enrichment using plasma, the MPE of [13C1]homocysteine was used in place of the [13C1]methionine label in plasma to calculate Q C. The loss of the [2H-methyl] moiety during transmethylation prevents the utilization of a homocysteine isotopomer as a surrogate for intracelluar [2H3-methyl]methionine enrichment, which was calculated by multiplying [2H3-methyl]methionine enrichment by the fractional ratio of [13C]homocysteine and [13C]methionine enrichment using the equation:

$$ E^{\prime }_{{[{}^{2}H{\text{-Met}}]}} = \, E_{{[{}^{2}H{\text{-Met}}]}} \left( {E_{{[{}^{13}{\text{C}}{\text{-Hcy}}]}} /E_{{[{}^{13}{\text{C}}{\text{-Met}}]}} } \right), $$

where \( E_{{[{}^{2}H{\text{-Met}}]}} \), \(E_{{[{}^{13}{\text{C}}{\text{-Hcy}}]}} \), \( E_{{[{}^{13}{\text{C}}{\text{-Met}}]}}\) are the plasma MPE enrichments of [2H-methyl]methionine, [13C1]homocysteine, and [13C1]methionine, respectively. The intracellular enrichment of [2H3-methyl]methionine (i.e. E′ 2[ H-Met]) was used to calculate Q M (House et al. 1997; MacCoss et al. 2001).

Remethylation was calculated as: RM = Q M − Q C and transsulfuration, which is related to methionine oxidation (Ox), was calculated by relating the 13CO2 excretion in breath to the plasma enrichment of [13C1]homocysteine as in the equation:

$$ {\text{TS}} = {\text{ Ox }} = \, F^{13}_{{{\text{CO}}_{ 2} }} (1 /E_{{[{}^{13}C{\text{-Hcy}}]]}} {-} \, 1 /E_{\text{IT}} ) \times 100, $$

where \( F^{13}_{{{\text{CO}}_{ 2} }} \) is the rate of 13C output in breath (Storch et al. 1988, 1990) after being adjusted by 0.93 for bicarbonate retention in neonatal piglets (House et al. 1997; Bertolo et al. 2003), and E IT is the enrichment of the [13C1]methionine infusate. Transmethylation was calculated as the sum of transsulfuration and remethylation (i.e,. TM = TS + RM). The rate of whole body protein synthesis (PS) and protein breakdown (PB) were calculated using:

$$ \begin{aligned} {\text{PS }} & = \, Q_{C} {-}{\text{ TS}}, \\ {\text{PB }} & = \, Q_{C} {-} \, I, \\ \end{aligned} $$

where I is the dietary intake of methionine [88 µmol met/(kg h)] (Clarke and Bier 1982; Storch et al. 1988, 1990; Thompson et al. 1989). Protein deposition (PD) was calculated as PS − PB = PD.

Choline and betaine kinetics

To calculate the fraction of choline and betaine flux from [2H3-methyl]methionine, the following equation was first applied to express the percent of product enrichment (Bertolo et al. 2003; Urschel et al. 2007):

$$ {\text{Fractional net conversion }}\left( {\text{FNC}} \right) \, = \, (E_{\text{Product}} /E_{\text{Precursor}} ) \, \times \, 100\;\% , $$

The E Product was the steady state MPE enrichment of either [M + 3] choline or [M + 3] betaine. The E Precursor was E′ 2[ H3-Met].

The term E Product/E Precursor has been applied previously to calculate Q Precursor→Product [µmol/(kg h)] using the formula:

$$ Q_{{_{{{\text{Precursor}} \to {\text{Product}}}} }} = \, E_{\text{Product}} /E_{\text{Precursor}} \times \, Q_{\text{Product}} \times \, \left[ {Q_{\text{Precursor}} /\left( {i_{\text{PC}} + \, Q_{\text{Precursor}} } \right)} \right], $$

where i PC is the precursor infusion rate (i.e., [2H3-methyl] methionine) and the product was either choline or betaine. The term [Q Precursor/(i PC + Q Precursor)] corrects for the contribution of the precursor infusion to the Q Precursor→Product calculation (Clarke and Bier 1982; Thompson et al. 1989; Urschel et al. 2007).

Because the above equation requires a separate infusion of labeled product to calculate Q Product, this was factored out of the equation to generate “fraction of product flux” (FPF), and is considered the stable isotope equivalent to FNC as it corrects for the contribution of the [2H3-methyl] infusion to Q Precursor→Product /Q Product. FPF was defined as follows:

$$ {\text{FPF }} = \, \frac{{Q_{{{\text{Precursor}} \to {\text{Product}}}} }}{{Q_{\text{Product}} }} = {\text{ FNC}} \times \, \frac{{Q_{\text{Precursor}} }}{{\left( {i_{\text{PC}} + \, Q_{\text{Precursor}} } \right)}}. $$

The FPF expresses the relative contribution of [2H-methyl]methionine to choline and betaine flux. We present both FNC and FPF values for the conversion of methionine to choline and betaine.

The relationship of FPF to Q Precursor→Product was evaluated by calculating FPF using previously published data by Urschel et al. (2007). In that study, researchers used a multi-tracer approach to quantify rates of arginine synthesis from proline, ornithine, and citrulline, under conditions of high and low dietary arginine in piglets. We calculated the FPF using the reported means of enrichment, flux, and infusion rate, as well as through substitution of the published Q Precursor→Product and Q Precursor values into our FPF equation (Storch et al. 1988; MacCoss et al. 2001; Riedijk et al. 2007; Urschel et al. 2007). As expected, the two methods of calculating FPF correlated strongly (r = 0.98; P < 0.001), and moreover, FPF values correlated with reported (Shoveller et al. 2003b; Urschel et al. 2007) rates of Q Precursor→Product (r = 0.55; P < 0.02).

Enzyme specific activity

Enzyme activity levels are expressed as nmol product/(min mg protein). In the case of the MSyn and BHMT assays, the product was methionine and the activity was determined by the following equation:

$$ {\text{Activity }} = \, \frac{{\left( {{\text{DPM }} - {\text{ DPM}}_{\text{blank}} } \right)\left( {v_{\text{inc}} } \right)}}{{\left( {\text{SA}} \right)\left( {v_{\text{add}} } \right)\left( t \right)\left( {c_{\text{s}} } \right)\left( {v_{\text{s}} } \right)}}. $$

In the case of CBS, the product was cystathionine, and its activity was calculated by:

$$ {\text{Activity}} = \, \frac{{({\text{DPM}} - {\text{DPM}}_{\text{blank}} )(v_{\text{inc}} )(v_{\text{elu}} )}}{{({\text{SA}})(v_{\text{add}} )(t)({\text{c}}_{\text{s}} )(v_{\text{s}} )(v_{\text{count}} )}}. $$

Collectively, DPM is the radioactivity of the eluted sample; DPMblank is the radioactivity of the blank; v inc is the final volume (mL) of the incubation mixture after stopping the reaction; SA is the specific activity (DPM/nmol) of the radioactive substrate; t is the time in min; v add is the volume of incubation mixture that was applied to the column for elution; v elu is the total elution volume that was collected into scintillation vials; v count is the volume of the eluent counted for radioactivity; c s is the concentration of the supernatant in mg/mL, and v s is the volume of liver supernatant used to catalyze the reaction. All samples were analyzed in triplicate and the entire assay was repeated if the coefficient of variation was >20 %.

Statistics

Unpaired t-test was used to compare data between the MD− and MS+ animals. Presented values are mean ± standard deviation. Published Q Precursor→Product values (Urschel et al. 2007) were correlated to FPF values using linear regression. In all cases, P < 0.05 was considered significant. Statistics were calculated using Prism software 5.0b (La Jolla, CA).

Results

Animals

Piglet weight gain was not different between groups from day 0 to 7 (382 ± 64 g/kg MD− versus 327 ± 77 g/kg MS+; P = 0.3), or between any time points during the study (data not shown). On the day of infusion (day 7) the pooled piglet weights were 2.3 ± 0.3 kg. Plasma concentrations of selected amino acids and metabolites are presented in Online Resource 1.

Methionine kinetics

Prior to calculating tracer flux, steady state was confirmed during the [13C]methionine and [2H-methyl]methionine infusions. A plateau was achieved in breath 13CO2 and plasma for the [13C]methionine and [2H-methyl]methionine tracers, as well as [M + 1] homocysteine by 6 h (Fig. 1a). Furthermore, the plasma appearance of the [2H-methyl] moiety into choline and betaine [M + 3] was also at plateau by 6 h and as expected, the MPE steady state enrichment was lower for betaine than for choline (Fig. 1b). Steady state could not be achieved for [M + 3] DMG in plasma over the 8 h time period examined. There was no effect of treatment on plasma methionine concentrations (P = 0.4), or methionine enrichment at steady-state (P > 0.5). However, while plasma homocysteine concentrations were elevated during MD− feeding (P = 0.001), the homocysteine enrichment was unaffected by treatment.

Fig. 1
figure 1

Typical time course data for plasma and breath enrichments for infusions of a [13C, 2H3-methyl]methionine and b [M + 3] choline and betaine during infusion of [2H3-methyl]methionine. Plasma enrichment is expressed as MPE and breath is expressed as APE

Full size image

Methionine kinetics during the [13C1]methionine and [2H3-methyl]methionine infusion are presented in Table 1. There were no effects of MD− feeding on methionine flux (Q), and there were no statistical differences between Q C and Q M. The rates of whole body remethylation (P = 0.02) and transmethylation (P = 0.01) were 65 and 63 % lower during MD− feeding, respectively. However, the transsulfuration rate (i.e., homocysteine oxidation) was low in all piglets, which was expected given the marginal level of dietary methionine and the inclusion of dietary cysteine (P = 0.3). Also, as expected during methionine restriction, remethylation was 5–8-times greater than transsulfuration (P < 0.0001). Whole body protein synthesis, breakdown, and deposition were unchanged, as measured by this tracer approach (Table 1).

Table 1 Whole body methionine kinetics
Full size table

To estimate the relative metabolic fate of methionine among transsulfuration, transmethylation, and protein synthesis, those values were expressed as a percentage of [2H3-methyl]methionine flux (Q M). The percent of methionine flux partitioned to transmethylation (TM/Q M) was 74 % lower during MD− feeding compared to MS+ piglets (P = 0.007), which is partially accounted for by 6 % more methionine flux being consumed by protein synthesis (i.e., PS/Q M) in MD− pigs (P = 0.009). Indeed, gross methionine partitioning (i.e, TM/PS) was 61 % lower in the MD− animals (P = 0.01) suggesting partitioning to transmethylation is secondary to protein synthesis during methionine restriction; moreover, protein synthesis consumed a greater percentage of whole body methionine flux than transmethylation. Consistent with the rate of transsulfuration in these piglets, the percent of methionine metabolism undergoing transsulfuration (TS/Q M) was unaffected by methyl restriction. Moreover, following transmethylation, the partitioning of homocysteine between transsulfuration (TS/TM) and remethylation (RM/TM) was unaffected by MD− feeding (Table 1).

Fractional incorporation of [2H-methyl] into choline and betaine

The relative contribution of methionine to betaine and choline flux was expressed as fraction of product flux (FPF), fractional net conversion (FNC) and mole percent excess (MPE) (Table 2). Choline and betaine concentrations were reduced during MD− feeding (P < 0.05). For choline, there was no effect of MD− feeding on FPFMet→Choline, FNCMet→Choline and MPE. However, the FPFMet→Betaine (P = 0.0003) and FNCMet→Betaine (P = 0.03) of MD− piglets were significantly greater than MS+ pigs. Additionally, the steady state MPE for betaine was 14-fold greater in the MD− piglets (P = 0.0001).

Table 2 Isotope kinetics of methionine to choline and betaine flux
Full size table

Hepatic MSyn, BHMT, and CBS activities and expression

There were no effects of MD− feeding on in vitro remethylation enzyme activities in the liver (i.e., MSyn and BHMT) (Table 3). However, in vitro transsulfuration activity was sensitive to methyl restriction, as CBS activity was 35 % lower after MD− feeding (Table 3). Expressions of MSyn, BHMT, and CBS genes were not responsive to MD− feeding.

Table 3 In vitro remethylation and transsulfuration enzyme activities and gene expression in liver
Full size table

Discussion

A key feature of methionine metabolism is that remethylation replenishes demethylated methionine to form a cycle; thus, we hypothesized that remethylation contributes significantly to methionine utilization in the neonate. We demonstrated that removal of dietary methyl donors, folate, choline, and betaine, diminishes the methionine available for use in transmethylation in vivo and limits the in vitro capacity for methionine disposal. Furthermore, transmethylation is secondary to protein synthesis when methionine is restricted in neonatal piglets. At the whole body level, remethylation contributes a significant amount of methionine for the neonate, mostly by providing more substrate for transmethylation.

The in vivo flux of whole body remethylation, transsulfuration, and transmethylation were measured using an isotope dilution technique described elsewhere (Storch et al. 1988; MacCoss et al. 2001; Riedijk et al. 2007). As was hypothesized, remethylation was reduced by 60 % during methyl restriction (MD−); indeed, we calculated that the provision of dietary methyl donors contributed 7.2 µmol methionine/(kg h), or 8 % more methionine than that which was provided in the diet (i.e., 10 % of the methionine requirement for piglets).

Dietary methyl donors also contributed 7.7 µmol methyl/(kg h) for transmethylation in these piglets, which equates to 425 µmol methyl/day for methylated products in these 2.3-kg piglets. Indeed, whole body transmethylation was 56 % lower during methyl restriction (MD−), and thus a 56 % reduction in the synthesis of the >50 transmethylation products is also expected. Taken as a percentage of methionine metabolism (Q M), transmethylation (TM/Q M) was reduced while protein synthesis (PS/Q M) was enhanced during methyl restriction (Table 1). Indeed, dietary methionine was preferentially partitioned toward protein synthesis at the expense of transmethylation. Moreover, the relative partitioning of dietary methionine (TM/PS) was 61 % lower due to the reduction in transmethylation. Because the primary site of methionine partitioning is likely liver (Walkey et al. 1998; Vance et al. 2007; Bertolo and McBreairty 2013), this shift in partitioning confirms our hypothesis that hepatic and whole body transmethylation were sacrificed for protein synthesis during methyl restriction, and is in agreement with Bauchart-Thevret et al. (2009) who reported that transmethylation was reduced by ~75 % compared to a 50 % reduction in protein synthesis in piglets fed a sulfur amino acid-free diet.

One of the intricacies of methionine metabolism is that a major transmethylation product, phosphatidylcholine, can be hydrolyzed to free choline and theoretically fulfill choline and betaine demands for remethylation. To gain insight into these conversions, the incorporation of [2H-methyl] into choline, betaine, and DMG was measured. Interestingly, the percent of choline flux derived from methionine (via phosphatidylcholine) was unchanged with methyl deficiency. Because PEMT activity is enhanced during choline restriction (Walkey et al. 1998), it was expected that de novo choline synthesis from PC would also be enhanced. However, because circulating choline concentrations were significantly lower in MS− pigs (Table 2), PEMT was clearly unable to maintain free choline during dietary choline restriction. Indeed, given PC synthesis from choline via the Kennedy pathway was likely diminished in the absence of dietary choline, the synthesized PC from PEMT was likely needed for its other functions so was not readily hydrolyzed to choline. These data suggest that the choline requirement cannot be fulfilled by transmethylation alone and is in agreement with others (Zeisel et al. 1991).

The contribution of methionine to betaine flux was enhanced during methyl restriction. Because the contribution of methionine to choline flux was unchanged, the enhanced contribution of methionine to betaine flux was due to increased conversion of choline to betaine, and not the futile cycling of methyl groups. We predicted that prolonged choline restriction would reduce de novo betaine synthesis, similar to data in rat hepatocytes (Schneider and Vance 1978). Furthermore, in developing rats, betaine supply is primarily from exogenous sources (Clow et al. 2008), and fetal and young rats are limited in their hepatic capacity to synthesize betaine in vitro (Weinhold and Sanders 1973). We demonstrate that betaine is synthesized in neonatal piglets in vivo from PEMT-derived PC, and our data suggest an adaptive mechanism whereby betaine synthesis is enhanced during methyl restriction, potentially to regenerate methionine for protein synthesis. However, MS+ piglets were fed ample betaine and thus diminished betaine enrichment may be an artifact of high plasma betaine concentrations (Table 2). Future studies should investigate the plasticity of de novo betaine synthesis as well as discern if betaine is synthesized for the purposes of remethylation or as an osmolyte.

Transsulfuration rates were low in all pigs, consistent with low methionine and adequate cysteine in our diet design. Although TS flux rates were unaffected by the presence of dietary methyl donors, in vitro CBS specific activity, the rate-limiting enzyme of TS, was reduced during methyl restriction, which is in agreement with rats fed low and adequate methionine diet (Finkelstein and Mudd 1967). However, enzyme specific activity can be considered enzyme ‘capacity’ and is not necessarily reflective of in vivo flux through the enzyme. So CBS capacity was reduced with methyl deficiency, but because TS flux was very low in both groups (Table 1), it is possible that flux through CBS was well below its capacity and thus whole body TS was unchanged in vivo. Caution is warranted when using enzyme specific activity to interpret metabolic fluxes as these outcomes were disconnected in this model. Indeed, we also observed no changes in hepatic remethylation enzyme activities or expressions in vitro, in spite of clear changes to in vivo remethylation flux.

The abundance of dietary methyl donors represents a considerable variable in determining the minimum obligatory requirement of methionine in neonates, as remethylation contributes significantly to critical transmethylation reactions in the methionine-restricted neonate. Moreover, during methyl restriction, dietary methionine is preferentially partitioned to protein synthesis at the expense of transmethylation reactions. Because of this preferential partitioning, it is likely that the TM requirement for methionine is not detectable in typical methionine requirement studies using protein synthesis as a primary outcome. Therefore, current estimates of the methionine requirement may not be adequate to fulfill the transmethylation requirement. Because of methionine’s well established role in oxidative stress (Brosnan et al. 2007) and immune function (Li et al. 2007), it would be interesting to investigate whether antioxidant capacity (Métayer et al. 2008) or susceptibility to infectious disease (Tang et al. 2015) are compromised with methyl deficiency.