点击显示 收起
1 From the Wageningen Centre for Food Sciences, Nutrition and Health Programme, Wageningen, Netherlands (PV, MRO, and MBK); the Division of Human Nutrition, Wageningen University, Wageningen, Netherlands (PV, MRO, and MBK); and the Department of Nutritional Physiology, TNO Nutrition and Food Research, Zeist, Netherlands (GRS, EB, and TvV)
2 Supported by the Wageningen Centre for Food Sciences, an alliance of Dutch food industry and research institutes (TNO Nutrition and Food Research, Wageningen University and Research Centre, and Maastricht University), that receives funding from the Dutch government. 3 Reprints not available. Address correspondence to P Verhoef, Division of Human Nutrition, Bomenweg 2, 6703 HD Wageningen, Netherlands. E-mail: petra.verhoef{at}wur.nl.
ABSTRACT
Background: A high plasma total homocysteine (tHcy) concentration is a risk factor for cardiovascular disease. The increase in tHcy induced by methionine, the sole dietary precursor of homocysteine, might be modulated by other amino acids present in dietary proteins.
Objectives: Our objectives were to compare the postprandial effect of free and dietary methionine on plasma tHcy concentrations and to investigate whether serine and cystine modify the effect of free methionine on tHcy.
Design: We conducted a randomized crossover trial in 24 healthy men. Each subject ingested 4 meals on separate days, which were separated by 1 wk. tHcy concentrations were measured in the fasting state and at 2, 4, 6, 8, 10, and 24 h after meal ingestion. The meals were 1) a low-protein meal fortified with 30 mg methionine/kg body wt (reference, denoted by "Met"), 2) meal 1 additionally fortified with 60.6 mg serine/kg body wt (MetSer), 3) meal 1 additionally fortified with 12.3 mg cystine/kg body wt (MetCys), and 4) a protein-rich meal containing 30 mg methionine, 60.6 mg serine, and 12.3 mg cystine per kg body wt (Protein).
Results: The mean (±SD) fasting tHcy concentration was 9.1 ± 2.7 µmol/L. Mean peak tHcy concentrations were 17.9 ± 4.5, 14.3 ± 3.3, 14.8 ± 3.9, and 11.2 ± 3.1 µmol/L after Met, MetSer, MetCys, and Protein, respectively. Compared with the mean 24-h area under the tHcy-by-time curve after Met, the mean curves after MetSer, MetCys, and Protein were 37%, 32%, and 77% smaller, respectively (all P < 0.0005).
Conclusions: Dietary methionine increases tHcy much less than does free methionine. Serine and cystine attenuate the tHcy-raising effect of free methionine. Thus, dietary proteins with a high content of serine or cystine relative to methionine may lead to lower postprandial tHcy responses.
Key Words: Dietary protein • methionine • serine • cysteine • homocysteine • crossover study
INTRODUCTION
Dietary methionine, an essential amino acid found in protein-rich foods, is required in protein synthesis and supplies the methyl group for numerous methylation reactions. Demethylation of methionine produces homocysteine, which is subsequently broken down to cysteine and sulfate in the transsulfuration pathway (Figure 1) (1). A high plasma concentration of total homocysteine (tHcy) is a potential risk factor for cardiovascular disease, for pregnancy complications, and for dementia in the elderly (2–5).
FIGURE 1.. Schematic representation of homocysteine metabolism. THF, tetrahydrofolate.
A single, high, oral dose of free methionine forms an acute burden to the transsulfuration pathway, and any excess of intracellular homocysteine is exported to the blood. After oral administration of methionine at a dose of 0.1 g/kg body wt—a classical methionine-loading test, which was developed to diagnose subjects with enzymatic defects in the transsulfuration pathway—plasma tHcy concentrations increase 3-4-fold after 6 h (6). Lower oral doses of free methionine increase plasma tHcy concentrations in a dose-dependent manner (7). However, the increase in plasma tHcy after a meal appears to be lower than expected on the basis of the methionine content of the meal (8, 9). Thus far, no study has compared the acute tHcy-raising effect of free methionine with that of an equal dose of dietary methionine.
Other amino acids derived from dietary proteins probably determine the postprandial tHcy response of dietary methionine. Besides methionine, serine and cysteine are important amino acids in homocysteine metabolism. Serine is the ultimate source of most of the one-carbon units in the folate system; it donates a methyl group to tetrahydrofolate, and the methyl group is eventually donated to homocysteine for remethylation to methionine (Figure 1). Furthermore, in the first step of homocysteine catabolism via transsulfuration to cysteine, serine condenses with homocysteine to form cystathionine (Figure 1). Cellular concentrations of cysteine influence homocysteine metabolism by a feedback mechanism: high concentrations of cysteine decrease homocysteine catabolism to cysteine (10, 11) and may stimulate homocysteine remethylation (11, 12).
Long-term experiments in rats suggest that adding serine to a high-methionine diet leads to lower circulating concentrations of tHcy or methionine and an enhanced ability to catabolize methionine to cysteine (13–15). In cats, a long-term, high dietary intake of cystine—the dimer of 2 cysteine molecules, ie, the usual form in which cysteine is fed—increases tHcy concentrations (16). In humans, high doses of oral N-acetylcysteine lowered plasma tHcy in some studies (17) but not all (18, 19). Furthermore, oral loading of methionine in combination with N-acetylcysteine produces a lower tHcy response than does methionine loading alone (20). Effects of dietary serine and cystine on circulating tHcy have not been studied in single-meal, high-methionine situations or in humans.
We conducted a randomized crossover trial in 24 healthy men. The first aim was to quantitatively compare the effects of free methionine and dietary methionine on circulating tHcy. The second aim was to investigate whether simultaneous ingestion of serine or cystine together with methionine modifies the postprandial effect of free methionine on tHcy.
SUBJECTS AND METHODS
Subjects
The study was conducted according to Good Clinical Practice guidelines at TNO Nutrition and Food Research, Zeist, Netherlands. The protocol was approved by the local medical ethics committee. Twenty-four men with a mean (±SD) age of 40 ± 17 y participated in the study. They were recruited from the pool of volunteers of TNO Nutrition and Food Research and by advertisements in local newspapers. The men had to be healthy as assessed by using a health and lifestyle questionnaire, a physical examination, and the results of several laboratory tests. They had a normal body mass index (30, in kg/m2) and normal dietary habits; no history of metabolic disease, hypertension, or cardiovascular disease; normal blood values for hemoglobin, hematocrit, and white blood cell counts; no glucose or protein in urine; and normal serum folate (>6.5 nmol/L) and tHcy (20 µmol/L) concentrations. All volunteers gave written informed consent.
Design
The study was a crossover study in 24 subjects with 4 treatments; thus, each subject had a unique treatment order. The volunteers came to TNO Nutrition and Food Research for 4 whole-day visits, which were separated by 6-d washout periods. On these 4 occasions, a fasting blood sample was collected in the morning. The subjects subsequently consumed a test meal as breakfast. Blood was sampled at regular intervals after consumption of the test meal. A fasting blood sample was also taken on the morning after the whole-day visit. Throughout the study days, dietary intake was standardized to ensure consistent nutrient intake across treatments.
Test meals
The 4 test meals were as follows: 1) Met, a low-protein meal fortified with 30 mg L-methionine/kg body wt (Methioninum, apyrogenous; BUFA BV Pharmaceutical Products, Uitgeest, Netherlands); 2) MetSer, a low-protein meal fortified with 30 mg L-methionine and 60.6 mg L-serine per kg body wt (Serinum, apyrogenous; BUFA BV Pharmaceutical Products); 3) MetCys, a low-protein meal fortified with 30 mg L-methionine and 12.3 mg L-cystine per kg body weight (Cystinum, apyrogenous; BUFA BV Pharmaceutical Products); and 4) Protein, a protein-rich meal containing, among other amino acids, 30 mg methionine, 60.6 mg serine, and 12.3 mg cystine per kg body weight. The study was double blinded for the first 3 treatments and open for the protein-rich meal. Before the experiment took place, the amounts of methionine, serine, and cystine in the protein-rich meal were measured (see "Blood sampling, handling, and chemical analyses"). On the basis of the outcome of those analyses, the doses of free methionine, serine, and cystine used in the Met, MetSer, and MetCys treatments were determined.
The ratios of methionine to serine (1:2) and of methionine to cystine (1:0.4) are the average ratios as they occur in protein-rich foods. The meals were composed in such a way that for each study participant, the content of energy and the amount of methionine were the same for all 4 treatments. The energy content of the test meals and the individual doses of methionine, serine, and cystine were based on body weight measurements made at screening, which were rounded off to whole kilograms. Thus, the absolute amino acid doses that were ingested per test meal ranged from 1.86 to 3.00 g for methionine, from 3.76 to 6.06 g for serine, and from 0.76 to 1.23 g for cystine. The doses of methionine were in the range of normal daily intake (21, 22).
Composition of the low-protein and high-protein meals
The low-protein meal consisted of low-protein bread, diet margarine, jam, colored sprinkles, apple juice, grape juice, and fruit yogurt. The supplemental amino acids were dissolved in the yogurt. The amounts of bread, margarine, and fillings were dependent on body weight. Bread intake varied between 2 and 5 slices. The amount of additional grape juice varied per person and was provided to obtain a total energy content similar to that of the high-protein meal. The high-protein meal consisted of a boiled egg, 0.5 L semiskim milk, white bread (prepared with milk instead of water, with a protein content of 9.7 g/100 g instead of 8.1 g/100 g for white bread prepared with water), turkey ham, and "Leidse" cheese with 20% fat based on dry weight. The amounts of white-milk bread, ham, and cheese were based on body weight, and these items were consumed in the following proportion: 4 slices of bread with 100 g ham and 80 g cheese. Intake of white bread varied between 3 and 6 slices. The energy content and macronutrient composition of the high-protein and low-protein meals (without supplemental amino acids) for a person weighing 74 kg are shown in Table 1.
View this table:
TABLE 1. Energy content and macronutrient composition of the high-protein and low-protein meals as calculated for a person weighing 74 kg1
Study schedule and assessments
Each of the 4 treatment periods involved 3 consecutive days, ie, an evening, the whole next day, and the following morning. There was a washout period of 6 d after each whole-day visit. On the first day of each treatment period (ie, days 1, 8, 15, and 22 of the study), the subjects were free-living until 1830, when they came to TNO Nutrition and Food Research for consumption of a low-protein frozen dinner. They went home after the dinner, and the subjects were asked to consume nothing but a provided snack (gingerbread) and water during the remainder of the evening. After 2200, the subjects had to fast until the next morning, when blood was drawn at TNO Nutrition and Food Research between 0800 and 0930 (ie, on days 2, 9, 16, and 23 of the study). After the fasting blood sample was collected, the subjects ingested the test meals. Blood was sampled at regular intervals after consumption of the test meal, ie, at 2, 4, 6, 8, and 10 h. Dietary intake was controlled throughout the day to ensure consistent nutrient intake across treatments. Lunch was consumed after blood collection at 4 h, and dinner was consumed after blood collection at 10 h. The lunch consisted of low-protein bread, diet margarine, jam, Heinz sandwich spread, and colored sprinkles. At dinnertime, the subjects ate the same dinner that was consumed on the evening before (ie, the same composition and the same amount). The subjects then went home, consumed the slice of gingerbread before 2200, and fasted thereafter. A fasting blood sample was taken on the next morning, ie, at 24 h (days 3, 10, 17, and 24 of the study).
All test meals were consumed under our supervision. Each subject consumed the complete amount of yogurt supplemented with amino acids and the entire amount of the high-protein meals. Furthermore, the subjects generally reported good adherence to the dietary restrictions.
Blood sampling and handling and chemical analyses
All blood samples were taken from the antecubital vein by using evacuated tubes. On the whole-day visits (ie, days 2, 9, 16, and 23), 10 mL EDTA blood was sampled 6 times via an indwelling cathether (obturator locked) for analyses of plasma tHcy (all time points) and vitamin B-6 (0 h). At 0 h, we additionally collected 6 mL clot blood for analyses of serum vitamin B-12 and folate. On days 3, 10, 17, and 24 (ie, for the fasting sample 24 h after the test meal), we sampled 10 mL EDTA blood for plasma tHcy measurement. Immediately after collection, blood was mixed well and put on ice. Within 30 min, all samples were centrifuged for 10 min at 2000 x g and 4 °C. Aliquots of plasma and serum were stored at <–18 °C. Samples were coded so that the identity and treatment of subjects was hidden from the laboratory technicians. The analyses of samples were carried out after all samples had been collected and with all samples from a given subject in the same run.
tHcy concentrations were measured by using HPLC (23, 24). Within- and between-run CVs were 3.6% and 6.4%, respectively. Vitamin B-6 concentrations were measured by using HPLC (25), and folate and vitamin B-12 concentrations were measured with the use of the SimulTRAC Radioassay Kit (ICN Pharmaceuticals, Orangeburg, NY). Within- and between-run CVs were <8% for all vitamins.
For the amino acid analyses, an oxidized meal sample was hydrolyzed with 6 mol HCl/L. The liberated amino acids were analyzed with the use of an automatic amino acid analyzer, a cation-exchange column, and postcolumn derivatization with ninhydrin. Detection of the amino acids was at 570 nm.
Definitions, calculations, and statistical analysis
The maximum plasma total tHcy concentration after ingestion of the test meals was denoted as Cmax. The time to maximum concentration was denoted as tmax; it indicates actual sampling times because we did not use interpolation. Furthermore, the area under the curve (AUC) for tHcy x time was calculated, after subtraction of the baseline value, by using trapezoidal approximation. As baseline, the line between the value at 0 h and that at 24 h was used.
In all comparisons, the Met treatment was used as the reference treatment. Thus, for each subject, we calculated the differences in values between MetSer and Met, MetCys and Met, and Protein and Met.
The study outcomes of interest were differences in mean tHcy Cmax, in mean tmax, and in mean tHcy AUC. Treatment effects were investigated by using analysis of variance (general linear models procedure in SAS). If the analysis of variance indicated a statistically significant overall treatment effect, comparisons of means were performed by using paired Student’s t tests. Because we made 3 comparisons (ie, Met compared with Protein, Met compared with MetSer, and Met compared with MetCys), we applied Bonferroni correction to evaluate the significance of the differences in means. Thus, instead of considering a difference significant at a P value of 0.05, we took 0.017 (0.05 divided by 3) as the level for statistical significance. In addition, the 95% CIs corresponding to the differences in means between the treatments were calculated. Statistical analyses were carried out by using SAS version 8.2 (SAS Institute Inc, Cary, NC).
RESULTS
The study participants did not have hypertension and had normal folate, tHcy, and cholesterol concentrations (Table 2). The mean fasting (ie, at 0 h, before the test meal was consumed) tHcy concentration over the 4 test days was 9.1 ± 2.7 µmol/L, which was somewhat higher than that measured at selection of the participants (Table 2). As expected in a randomized crossover study, the fasting tHcy concentrations were very similar for all 4 treatment periods (Figure 2). After consumption of each of the test meals, tHcy concentrations gradually increased and reached a peak between 4 and 7 h (Figure 2). The mean tmax was shortest for the Met treatment and longest for the Protein treatment, and the MetSer and MetCys treatments had intermediate tmax values (Table 3). tHcy concentrations decreased thereafter and returned toward pretreatment fasting concentrations after 24 h.
View this table:
TABLE 2. Characteristics of the 24 study subjects1
FIGURE 2.. Mean (± SD) plasma total homocysteine concentrations in 24 men before (0 h) and after (2, 4, 6, 8, 10, and 24 h) consumption of a low-protein meal fortified with methionine (Met), a low-protein meal fortified with methionine and serine (MetSer), a low-protein meal fortified with methionine and cystine (MetCys), and a protein-rich meal (Protein).
View this table:
TABLE 3. Peak homocysteine concentration (Cmax), time to reach peak concentration (tmax), and 24-h area under the curve (AUC) for homocysteine x time in 24 men who received 4 different treatments in a crossover design1
The mean Cmax and AUC for tHcy x time were highest after the Met treatment and lowest after the Protein treatment (Table 3). The MetSer and MetCys treatments led to mean values for Cmax and AUC for tHcy x time that were in between the mean values for the Met and Protein treatments. All differences in mean tmax, Cmax, and AUC for tHcy x time from the values after the Met treatment were significant. Expressed as percentage differences from the mean AUC for tHcy x time after the Met treatment, the mean AUCs were 37%, 32%, and 77% smaller after the MetSer, MetCys, and Protein treatments, respectively. The mean fasting (ie, before the test meal at 0 h) concentrations of serum folate, serum vitamin B-12, and plasma vitamin B-6 did not differ significantly between the 4 treatments (data not shown). Body weight did not change significantly during the study (data not shown). These findings indicate that the subjects did not make rigorous changes to their normal dietary habits while participating in the study.
DISCUSSION
In this experiment in healthy volunteers, we showed that a protein-rich meal providing on average 2.4 g methionine produces a 24-h response in circulating tHcy concentration that is only about one-quarter of that produced by a similar amount of free methionine ingested with a low-protein meal. Furthermore, we showed that ingestion of free serine or cystine together with free methionine reduces the 24-h response by about one-third relative to that produced by free methionine alone. These data suggest that consumption of dietary proteins with a high content of serine or cystine relative to methionine lead to lower postprandial tHcy responses.
To assess the effects of methionine and other amino acids postprandially, we specifically chose to use a methionine dosage in the range of dietary intake (30 mg/kg body wt), instead of the dosage routinely used in methionine-loading tests (100 mg/kg body wt). Of course, the modifying effects of serine and cystine might be different if serine and cystine were combined with a methionine-loading test.
We assessed tHcy concentrations during 24 h after consumption of a high dose of methionine. However, our approach gives no information about intestinal absorption of methionine, first pass effects, or changes in fluxes through transmethylation, remethylation, or transsulfuration. Nevertheless, we would like to speculate on the metabolic explanations for our findings.
Free methionine compared with dietary methionine
In previous experiments, the effects of methionine consumed as part of dietary protein appeared to be quite modest and lower than expected on the basis of the amount of methionine (8, 9). Our study was the first to make a quantitative, controlled comparison between the effects of dietary methionine and those of free methionine. There are several possible explanations for the observation that the tHcy response after dietary methionine was smaller and slower than that after a similar amount of free methionine. First, it may be explained by an attenuating effect on the postmethionine tHcy response caused by serine and cystine present in the proteins (see below). It is striking that adding up the effects of the MetSer and MetCys treatments would yield a modest postprandial tHcy increase that was similar to that observed for the high-protein treatment. Unfortunately, we did not include a treatment of pure methionine, serine, and cystine combined. Second, the methionine from the protein-rich meal may have been absorbed more slowly than was free methionine, because time is needed to break down the proteins. Thus, the rate of tHcy formation from dietary methionine was probably slower than that from free methionine. Third, in contrast with free methionine, part of the methionine from the high-protein meal may have been used for protein synthesis, because other amino acids necessary for protein synthesis were ingested at the same time. This may have reduced the amount of methionine that was directed toward demethylation to homocysteine.
Of course, the protein-rich meal differed from the low-protein meal supplemented with methionine in other respects (Table 2): the former contained more fat, whereas the latter contained more carbohydrates, especially mono- and disaccharides. Although fat does not affect the absorption of amino acids (V Young, personal communication, 2002), the slower gastric emptying may have contributed to the above mentioned slower tHcy kinetics after the meal than after free methionine. Furthermore, the high-carbohydrate meal may have caused an increase in insulin, which has been shown to suppress the activity of cystathionine ß-synthase and hence increase tHcy concentrations (26, 27). However, a high-protein meal may stimulate insulin secretion as well (28), and we consider the insulin effects, if any, to have been small.
Serine and cystine
We observed that, in comparison with methionine alone, a combination of methionine and serine and of methionine and cystine decreased the 24-h tHcy response by about one-third. A likely explanation is the fact that serine may increase fluxes of homocysteine through both the remethylation and the transsulfuration pathways of homocysteine (1). Animal studies (13–15) have shown a tHcy-lowering effect of long-term serine treatment, and we have now shown that this occurs in a single-dose situation in humans as well. The effect of dietary cystine on circulating tHcy has been investigated only in long-term studies, which found contradictory effects. In cats, long-term, high-dose cystine feeding increases tHcy (16), which is in line with the observation that cystine supplementation suppresses homocysteine transsulfuration (ie, homocysteine catabolism) in humans (10). However, long-term supplementation with cystine may also decrease tHcy concentrations by enhancing remethylation of homocysteine to methionine, which is caused by a reduced need for cysteine formation (11, 12). In our single-meal situation, it seems more likely that cysteine may have been a rate-limiting amino acid for protein synthesis and that with the MetCys treatment, more methionine was directed toward protein formation than toward demethylation to homocysteine. Finally, the acute attenuating effect of cystine on postmethionine tHcy concentrations may be explained by competition between cystine and homocystine for carrier protein-binding sites in circulation and subsequently more efficient renal clearance of free homocysteine (17). These 2 latter explanations are supported by the fact that treatment with N-acetylcysteine—at doses similar to the cystine dose in our study—lowers tHcy concentrations after methionine loading in humans (20).
A lower tHcy response in a high-methionine situation was also observed for a combination of methionine and glycine in rats (14). Glycine is another key amino acid in homocysteine metabolism (Figure 1).
Implications for public health
Whether increased tHcy is causally related to cardiovascular disease is still uncertain (29), but the existence of such a relation is becoming increasingly likely (2, 3). At the end of 2004, several clinical trials will report on the effect of tHcy lowering on the risk of cardiovascular disease (30). Although the trials focus on the lowering of fasting tHcy, the importance of high postprandial tHcy concentrations in relation to the risk of cardiovascular disease is indicated by several findings. First, in epidemiologic studies, classical postmethionine-load tHcy concentrations were associated with cardiovascular disease independently of fasting concentrations (31). Second, a strong increase in plasma tHcy concentration in response to free or dietary methionine has been observed to acutely impair the reactivity of the vascular endothelium (32), which is a promising early marker of cardiovascular disease risk. Our findings imply that certain protein-rich foods have a lower postprandial tHcy-raising effect than do others on the basis of their ratio of methionine to serine to cystine. For example, when peanut butter with a 1:4.5:1 ratio of methionine:serine:cystine and Gouda cheese with a ratio of 1:2.1:0.3 are applied to separate pieces of bread in equal amounts, the bread with the peanut butter might induce a lower tHcy response.
Conclusions
The tHcy response is much lower after a high-protein meal than after a similar amount of free methionine. We found that this can partly be explained by the simultaneous ingestion of serine and cystine. The clinical relevance of these observations depends on whether elevated tHcy is proven to be a cause of cardiovascular disease.
ACKNOWLEDGMENTS
We thank the volunteers for their participation, all those involved in the conduct of the experiment at TNO Nutrition and Food Research for their dedication, and the laboratory staff at Wageningen University and TNO for careful analyses. Furthermore, we thank the National Institute of Public Health and the Environment, Department of Chronic Disease Epidemiology, Bilthoven, Netherlands, for data on the amino acid composition of foods.
All authors were responsible for the design of the study. EB and TvV were in charge of the conduct of the experiment, including selection of subjects, planning, implementation, and statistical analyses. PV prepared several drafts of the manuscript. All other authors critically revised the paper. None of the authors had a financial or personal interest or advisory board affiliation in any company or organization sponsoring the research.
REFERENCES