Marginal deficiency of vitamin B-6 is common among segments of the population worldwide. Because pyridoxal 5′-phosphate (PLP) serves as a coenzyme in the metabolism of amino acids, carbohydrates, organic acids, and neurotransmitters, as well as in aspects of one-carbon metabolism, vitamin B-6 deficiency could have many effects. Healthy men and women (age: 20-40 y; n = 23) were fed a 2-day controlled, nutritionally adequate diet followed by a 28-day low-vitamin B-6 diet (<0.5 mg/d) to induce marginal deficiency, as reflected by a decline of plasma PLP from 52.6±14.1 (mean ± SD) to 21.5±4.6 nmol/L (P<0.0001) and increased cystathionine from 131±65 to 199±56 nmol/L (P<0.001). Fasting plasma samples obtained before and after vitamin B6 restriction were analyzed by 1H-NMR with and without filtration and by targeted quantitative analysis by mass spectrometry (MS). Multilevel partial least squares-discriminant analysis and S-plots of NMR spectra showed that NMR is effective in classifying samples according to vitamin B-6 status and identified discriminating features. NMR spectral features of selected metabolites indicated that vitamin B-6 restriction significantly increased the ratios of glutamine/glutamate and 2-oxoglutarate/glutamate (P<0.001) and tended to increase concentrations of acetate, pyruvate, and trimethylamine-N-oxide (adjusted P<0.05). Tandem MS showed significantly greater plasma proline after vitamin B-6 restriction (adjusted P<0.05), but there were no effects on the profile of 14 other amino acids and 45 acylcarnitines. These findings demonstrate that marginal vitamin B-6 deficiency has widespread metabolic perturbations and illustrate the utility of metabolomics in evaluating complex effects of altered vitamin B-6 intake.
Citation: Gregory JF III, Park Y, Lamers Y, Bandyopadhyay N, Chi Y-Y, Lee K, et al. (2013) Metabolomic Analysis Reveals Extended Metabolic Consequences of Marginal Vitamin B-6 Deficiency in Healthy Human Subjects. PLoS ONE 8(6): e63544. https://doi.org/10.1371/journal.pone.0063544
Editor: Daniel Tomé, Paris Institute of Technology for Life, Food and Environmental Sciences, France
Received: March 2, 2012; Accepted: April 5, 2013; Published: June 11, 2013
Copyright: © 2013 Gregory et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by National Institutes of Health (NIH) Grant DK072398 (to Jesse F. Gregory) and NIH National Center for Research Resources CTSA Grant 1UL1RR029890. Support of the Clinical Biomarkers Laboratory, Emory University was from NIH Grant ES016731 to Dean P. Jones, and support of the Sarah W. Stedman Nutrition and Metabolism Center, Duke University, was from NIH Grant PO1-DK-58398 to Christopher B. Newgard. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Vitamin B-6 exists in many dietary sources, yet an individual's particular food consumption pattern and certain drug-nutrient interactions can lead to low vitamin B6 status. The vitamin B-6 status of much of the United States population is adequate. However, the lower percentiles of intake are associated with low vitamin B-6 status ,  that is more prevalent in smokers, women and the elderly , , . Many inflammatory conditions also are associated with lower vitamin B-6 status regardless of intake , , but the mechanism is unknown. The use of certain common drugs such as theophylline  and oral contraceptive agents ,  also is associated with reduced vitamin B-6 status.
The coenzymic form of vitamin B-6, pyridoxal phosphate (PLP), serves as a coenzyme for over 140 enzymes in human metabolism. PLP is thus involved in a wide array of functions  including: the catabolism and interconversion of most amino acids; the formation of various organic acids, including species involved in the TCA cycle and gluconeogenesis; heme synthesis; and several key steps in pathways associated with one-carbon metabolism. Vitamin B-6 deficiency also is associated with interconversions of long-chain polyunsaturated fatty acids.
Plasma PLP concentration of <20 nmol/L reflects vitamin B-6 deficiency , while 20–30 nmol/L indicates marginal status , . The consequences of marginal deficiency are unclear, but chronically low vitamin B6 status is associated with increased risk of cardiovascular disease –, deep-vein thrombosis –, stroke  and certain cancers , . The mechanisms responsible for these disease connections are unknown but do not appear to be associated with hyperhomocysteinemia . In view of the many coenzymic roles of PLP, further investigation of the in vivo metabolic consequences of inadequate vitamin B6 status may provide better insight into the effects of marginal vitamin B-6 deficiency.
We have investigated the consequences of inadequate vitamin B6 status using a series of protocols that involve the use of controlled low-vitamin B-6 diets in healthy volunteers –. In these studies, we employed targeted metabolite profiling and in vivo stable isotope tracer kinetic protocols to derive functional information about specific vitamin-dependent processes in one-carbon metabolism and related pathways while the participants were in adequate and marginal vitamin B6 status. These studies led to the following major observations concerning the effects of vitamin B-6 restriction: (a) surprising resiliency of one-carbon metabolism to effects of vitamin B6 deficiency, (b) changes glycine kinetics and concentration, (c) the resiliency of transsulfuration flux concurrent with an expansion of the cystathionine pool, (d) individual variability in the kinetics of glutathione synthesis, and (e) altered patterns of circulating n-3 and n-6 polyunsaturated fatty acids –. This work has led to new insights into PLP-dependent metabolic processes and the influence of vitamin B6 nutritional status.
Advances in both NMR and mass spectral aspects of metabolomics have impacted many facets of biology including the nutritional sciences , . The nutritional applications of NMR metabolomics to date have tended to focus on dietary effects on macronutrient metabolism and intermediary metabolites (for example, , , with few applications of these powerful tools in characterizing the metabolic effects varying levels of micronutrient status (for example, ). The direct analysis of plasma or urine by NMR provides a useful approach that complements mass spectrometry for evaluating metabolic phenotypes associated with nutritional adequacy and deficiency and for evaluating nutrient-gene and nutrient-disease interactions.
The study reported here was conducted to investigate the impact of controlled vitamin B-6 depletion through the use of 1H-NMR analysis of plasma from 23 healthy participants from two recent vitamin B-6 restriction studies , . We examined NMR spectra of intact plasma with and without deproteination by filtration as an untargeted means of evaluating vitamin B6-dependent changes in plasma constituents. The results were evaluated using multivariate analysis accounting for the paired structure of the data to assess differences in spectral patterns and metabolite profiles. On a subset of these samples (n = 18) we also performed provisional quantification on selected metabolites by spectral curve fitting (Chenomx). Because of the critical role of vitamin B6 in the metabolism of amino acids and organic acids, we also conducted targeted metabolite profile analysis of amino acids, selected organic acids and acylcarnitines using quantitative analyses of NMR signals, and mass spectrometry.
Materials and Methods
Participants and nutritional protocols
The blood samples for this study were obtained as described previously ,  from healthy men (n = 12) and women (n = 11), aged 20–40 y, who participated in two identical dietary vitamin B6 restriction protocols designed to assess the metabolic effects of marginal vitamin B6 status. Health was determined by physical examination and routine tests of hepatic, renal, thyroid and hematological function , 28. No participant had a history of gastrointestinal surgery, chronic disease, smoking or chronic drug use or alcoholism, body mass index not greater than 28 kg/m2, vitamin, amino acid, or protein supplementation, or chronic consumption of a high-protein diet. Nutritional adequacy was indicated normal concentrations of serum folate (greater than 7 nmol/L), serum vitamin B-12 (greater than 200 pmol/L), plasma PLP (greater than 30 nmol/L), and plasma total homocysteine (less than 12 µmol/L). All participants gave written informed consent. The study was approved by the University of Florida Institutional Review Board and the UF Clinical Research Center (CRC) Scientific Advisory Committee. This study was registered at clinicaltrials.gov as NCT00877812.
The experimental protocol consisted of two dietary periods with a metabolic assessment of tracer kinetics and blood sampling for metabolite profiling at the end of each , . The first consisted of a 2-d nutritionally adequate controlled diet (total vitamin B-6 = 1.02±0.11 mg/d), followed by the first metabolic assessment of participants in documented adequate vitamin B6 status. The second period involved a 28-d low vitamin B-6 diet to induce a state of marginal vitamin B-6 deficiency (PLP ∼20 nmol/L), followed immediately by a metabolic assessment identical to the first. The vitamin B-6 restriction diet (total vitamin B-6 = 0.37±0.04 mg/d) provided total protein intake of 1 g kg−1 d−1 and mean methionine and cystine intake of 21 and 17 mg kg−1 d−1, respectively. During both metabolic assessments, blood samples analyzed in this metabolomic analysis were obtained in the morning after withholding food overnight and were followed by primed, constant infusion protocols reported previously , . The analyses reported here were performed on plasma obtained before and after the 28-d vitamin B-6 restriction period. The samples taken for NMR and mass spectral analysis were drawn by syringe from an indwelling catheter in an antecubital vein, mixed gently in a sodium EDTA tube, cooled on ice, centrifuged within 15 min of collection and then stored at −80°C.
Targeted quantitative analysis
Measurement of plasma PLP ,  and aminothiols (total glutathione, cysteine, and homocysteine; ) was performed by fluorometric HPLC. Plasma cystathionine, a sensitive functional biomarker of vitamin B-6 deficiency , , ), was determined by gas chromatography-mass spectrometry . Plasma concentrations of free amino acids and acylcarnitines were determined by electrospray tandem mass spectrometry using isotope dilution methods . We determined plasma glucose concentration using a commercial hexokinase-based assay kit (Sigma GAHK-20).
1H-NMR spectra were obtained at 600 MHz on a Varian INOVA 600 spectrometer with water presaturation at 25°C at the Emory University Clinical Biomarkers Laboratory, as described previously . Plasma samples were analyzed in intact form (i.e., unfiltered) and following filtration through a 3 kD nominal cut-off membrane that removes proteins, lipids and lipoproteins as well as protein-bound forms of certain small molecules.
Unfiltered Plasma – 1H-NMR spectroscopy.
Plasma samples were thawed and 600 μL was mixed with 66 μL of deuterium oxide (D2O) with TMS [3-(trimethylsilyl)-1-propanesulfonic acid sodium salt (C6H15NaO3SSi, 1% w/w)] as internal standard, pH 7.4–7.6. 1H-NMR spectra were measured at 600 MHz on a Varian INOVA 600 spectrometer with water presaturation at 25°C essentially as described , . Preprocessing of spectra included baseline correction, spectral alignment, elimination of uninformative spectral regions and normalization , , with baseline correction and peak alignment , . The water signal region (4.5–5 ppm) was eliminated because of variable suppression of the water signal, as were the regions between 5.4–6.7 ppm and 0–0.1 ppm that contained no significant metabolite signal , . The spectral regions containing signals from EDTA and EDTA complexes with calcium and magnesium  also were excluded.
Filtered Plasma – 1H-NMR spectroscopy.
We examined the spectra of plasma samples following ultrafiltration, which removed most proteins, lipoproteins and lipids, in order to facilitate evaluation of small molecules. Centrifugal ultrafiltration devices (Microcon YM-3, 3,000 nominal molecular weight cutoff, Millipore Corp.) were pre-washed by centrifuging three times for 15 min each with 500 μl of portions of distilled water to remove the glycerol preservative. Each plasma sample was transferred to a pre-washed device and centrifuged at 13,800×g for 60 min. Filtered plasma samples were mixed with TMS and adjusted pH 7.0–7.4. 1H-NMR spectra were acquired at 600 MHz on a Varian INOVA 600 spectrometer with water presaturation at 25°C as described above. As described for previous 1H-NMR studies of human plasma conducted by the Emory Clinical Biomarkers Laboratory , the identity of major signals was supported by the addition of standards (amino acids, organic acids, energy intermediates), comparisons to spectra in chemical databases, reference to a rigorous description of NMR spectra of human plasma , application of the Chenomx database, and confirmation with 2-dimensional (2-D) NMR techniques.
Quantitative 1H-NMR analysis of metabolites.
The analysis of NMR spectra from filtered samples utilizing spectral fitting techniques (Chenomx, Edmonton, Alberta, Canada) allowed estimation of the concentrations of major plasma constituent small molecules , . For 5 of the 23 participants, at least one of the NMR spectra did not meet the criteria needed for the Chenomx analysis. Therefore, quantitative analysis is reported only on the 18 participants for which pre- and post-vitamin B-6 restriction data were obtained. We report from this analysis estimates of the concentrations of 15 plasma constituents for these 18 participants before and after vitamin B-6 restriction.
Principal Component Analysis (PCA) and Orthogonal Signal Correction-Partial Least Squares Discriminant Analysis (OPLS-DA) of processed spectra were performed using Pirouette software (Infometrix, Bothell, WA). PCA and OPLS-DA score plots and loading plots of NMR spectral data were used to visualize relational patterns and discriminatory factors to classify the groups. The OPLS-DA procedure, being a supervised technique (i.e., treatment groupings included in the analysis), removes variation not correlated to classification after centering the mean value for each frequency , , . Data were evaluated for unfiltered and filtered plasma using score plots in OPLS-DA analysis to evaluate the classification according to vitamin B-6 status. For each discriminating spectral region, we assigned the compounds that possibly contributed on the basis of previous literature , experiments involving the addition of selected standards, comparison to spectra in databases or applications with 2-dimensional NMR procedures , .
To account for the paired data structure within our study (i.e., observations before and after vitamin B-6 restriction), we then conducted multilevel PLS-DA as described by Westerhuis et al.  as follows. First, the paired structure was created by averaging each ppm for all individual spectra including B6 baseline and restricted. From each individual, the average spectrum were created with B6 baseline and restricted. Secondly, for each subject, the table of correction factors for each ppm was extracted by dividing averaged intensity at each ppm from all individual spectra with averaged intensity at each ppm from each individual averaged spectrum. Thirdly, using correction factors for each ppm, the intensities for each B6 baseline and restricted were adjusted to generate paired data structure. Lastly OPLS-DA was run on the paired data structure. Additonal analysis to identify and visualize discriminatory spectral regions was conducted using S-plots of covariance versus correlation for spectral data  followed by loading plots of covariance versus chemical shift on mean spectra. After calculation of the top 5% of discriminatory loadings in this analysis (647 and 381 for unfiltered and filtered plasma, respectively), we visually selected points (65 and 26 for unfiltered and filtered, respectively) in regions of the S-plots that were clearly separated from the central risk zone. These represented the most highly discriminating features.
Further statistical analysis of metabolite concentrations was performed after log base 2 transformation using a recently developed method for global hypothesis testing in repeated measures high-dimensional data . The log base 2 transformation was a means to meet the Gaussian requirement. For subsequent local hypothesis testing, we evaluated differences in the transformed concentration of individual plasma constituents using paired t-test and the positive false discovery rate method  to adjust p-values in anticipation of the inflation of false positive rate as a result of multiple testing. The effects of change in plasma PLP and change in plasma cystathionine were evaluated using the test for Pearson correlation coefficient in conjunction with the positive false discovery rate method to account for multiple testing.
Ratios of selected metabolites derived from NMR analysis were evaluated to assess product/precursor relationships before and after vitamin B-6 restriction. Asparagine/aspartic acid and pyruvate/alanine ratios served as static indicators of the proportions of amide versus acid forms of glutamate and aspartate. The ratios of pyruvate/alanine and 2-oxoglutarate/glutamate provided a measure of the product/precursor ratio for these aminotransferases.
A level of statistical significance of 0.05 was employed in all procedures.
Efficacy of the vitamin B-6 restriction protocol
The dietary protocol provided a controlled, nutritionally adequate 2-day stabilization period followed by very low vitamin B-6 intake for 28 days, and this protocol is effective for selectively inducing a state of marginal vitamin B-6 deficiency in healthy adults , , . Plasma PLP concentration declined from 52.6±14.1 (mean ± SD) to 21.5±4.6 nmol/L (P<0.0001), which indicated a change from adequate vitamin B-6 status to marginal deficiency. Additional descriptive data regarding the 23 participants in this protocol have been reported in part previously – and are summarized (Table S1).
Metabolic effects of vitamin B-6 restriction
Global statistical analysis of 1H NMR spectral data.
1H-NMR analysis of plasma samples yielded spectra that were qualitatively similar to those originally reported and annotated by Nicholson and colleagues , . Filtration of plasma attenuated the regions of the spectrum attributable to lipids and proteins, as expected. Representative spectra of unfiltered and filtered plasma in vitamin B-6 adequate and restricted states are presented in Figure S1.
Statistical evaluation of spectra first involved assessing the change in overall spectral data (before versus after vitamin B-6 restriction) using a recent method for global hypothesis testing for high-dimensional repeated measures outcomes of pairwise data . By this method, the overall change across the entire spectrum was not significant for unfiltered plasma (p-value = 0.432) and filtered plasma (p-value = 0.549). Gender also had no significant effect on the overall change in spectra of unfiltered plasma (p-value = 0.683) or filtered plasma (p-value = 0.275) in this analysis. Because this n = 23 data set was derived from two equivalent vitamin B-6 restriction trials, we also tested for effects of trial 1 versus trial 2. The effect of trial (1 versus 2) had no significant effect on the overall change for unfiltered plasma (p-value = 0.103) or filtered plasma (p-value = 0.346).
Multilevel PLS-DA evaluation of 1H-NMR spectra to assess effects of vitamin B-6 restriction in paired data.
Principal Component Analysis that did not account for paired structure of the spectral data obtained before and after vitamin B-6 restriction showed no separation or grouping of data according to vitamin B-6 status for both unfiltered and filtered plasma (data not shown). OPLS-DA, which is a supervised technique in which treatment groupings are identified, yielded unambiguous classification of data patterns according to nutritional status (i.e., pre- and post-vitamin B-6 restriction) for both unfiltered and filtered plasma (not shown). This provided evidence of extended compositional differences associated with vitamin B-6 restriction, although the OPLS-DA method also did not account for the paired data structure.
Multilevel PLS-DA , which is an approach to multivariate analysis that incorporates provisions for paired data, showed distinct separation according to vitamin B-6 status in both unfiltered and filtered plasma (Figure 1). For unfiltered plasma, the first orthogonal component (40% of total variation) and the first principal components explained 40% and 30% of the total variation, respectively. For filtered plasma the first orthogonal component and the first principal component explained 29% and 60% of total variation, respectively. S-plots derived from this analysis of1H-NMR data in the multilevel PLS-DA allowed detection of the spectral regions that discriminated between vitamin B-6 adequacy (baseline) and restricted conditions for both filtered and unfiltered plasma (Figure 2A and 2B). When discriminating features were superimposed on mean spectra (Figure 2C and 2D), the discriminating regions appeared to be primarily attributable to lipids, organic acids and amino acids for unfiltered plasma and organic acids, and amino acids for filtered plasma (Figure 2C and 2D).
Plasma samples from 23 healthy, young adults collected at baseline and after 4 weeks consuming a vitamin B6-restricted diet were examined by 1H-NMR spectroscopy followed by multilevel partial least square-discriminant analysis (multilevel PLS-DA). A) Score plot for unfiltered plasma, which contains relatively large signal from lipoproteins. B) Score plot for plasma filtered through a 3 μm pore size filter to remove most of the lipoprotein before 1H-NMR spectroscopy analysis.
Panels A and B: S-plots that respectively correspond with the score plots in Figure 1A (Not Filtered) and Figure 1B (Filtered) plasma. In these panels, the top 5% of metabolites that contribute to 95% of the separation of baseline and restricted samples are highlighted in red squares. Red-framed points within the ovals represent the most highly discriminating signals. Panels C and D: loading plots illustrating discriminating spectral features from S-plots. Covariance of each discriminating feature is superimposed on the corresponding NMR chemical shift on mean spectra.
Vitamin B-6 restriction affects the plasma metabolite profile.
Analysis using NMR spectral fitting provided an estimation of the concentration of selected plasma constituents for which methods of targeted quantitative analysis were not readily available in this study. We report here this analysis using Chenomx software for acetate, acetoacetate, aspartate, asparagine, choline, formate, fumarate, glutamate, glutamine, lactate, myo-inositol, 2-oxoglutarate, pyruvate, succinate taurine and trimethylamine oxide in a subset (n = 18) of the participants (Table 1). Local hypothesis testing of changes in the estimated plasma concentrations derived from spectral features attributed to acetate, asparagine, glutamine, myo-inositol, lactate, 2-oxoglutarate, pyruvate, taurine and trimethylamine-N-oxide significantly increased (adjusted P<0.05), while that of glutamate significantly decreased (adjusted P<0.05) (Table 1).
The evaluation of the ratios of several of these plasma constituents allowed us to probe possible biochemical effects of vitamin B-6 restriction. Whereas spectral features attributable to asparagine and glutamine concentrations increased (Table 1), the ratios reflecting the relative extent of amidation differed (Table 2). Vitamin B-6 restriction did not significantly change the asparagine/aspartate ratio, but the ratio of glutamine/glutamate more than doubled due to the restriction (P<0.001). In addition, the ratio of the 2-ketoacids pyruvate and 2-oxoglutarate and the corresponding amino acids alanine and glutamate reflected the balance of respective aminotransferase reactions. The pyruvate/alanine ratio did not change significantly, while the 2-ketoglutarate/glutamate ratio increased 76% due to vitamin B-6 restriction (P<0.001). The ratio of plasma lactate/pyruvate, which is indicative of cytoplasmic redox state and cellular respiration, did not change due to vitamin B-6 restriction (8.3±2.0 versus 7.6±1.5; P = 0.242).
The concentration of amino acids and glutathione determined by targeted quantitative methods (tandem mass spectrometry, GC-MS and HPLC) revealed significant effects of vitamin B-6 restriction (Table 3). Plasma cystathionine, which is a functional biomarker of vitamin B-6 deficiency indicating functional impairment of the transsulfuration pathway , , , increased markedly (adjusted P = 0.0002). Plasma proline exhibited an unexpected increase (adjusted P<0.05), but there was no significant change in most of the other amino acids evaluated (alanine, arginine, asx, citrulline, glx, glycine, histidine, leucine + isoleucine, methionine, ornithine, phenylalanine, serine, tyrosine, and valine). In this context, asx and glx refer to the concentrations of aspartate and glutamate plus a contribution from partial hydrolysis of their amides, asparagine and glutamine, under the conditions of the tandem mass spectral analysis. The concentration of plasma glucose measured enzymatically was not changed by vitamin B-6 restriction (4.71±0.676 versus 4.52±0.543 mmol/L; adjusted P = 0.122). We found no change in the concentrations of 45 acylcarnitines or in total acylcarnitines (Table S2).
We also evaluated whether the change in the various quantified plasma constituents was related to the change in plasma PLP concentration. No significant relationship with PLP was found for the amino acids and glutathione (adjusted P = 0.764) or for the acylcarnitines (adjusted P = 0.747).
Effectiveness of the vitamin B-6 restriction protocol
The controlled vitamin B-6 restriction protocol employed in this study provided a novel and reproducible tool for investigating metabolic effects of marginally deficient vitamin B-6 status –. We recognize, however, that the short term metabolic effects of marginal vitamin B-6 status may not reflect the mechanisms responsible for the association of chronic low vitamin B-6 status with greater long-term risk of various forms of chronic disease. We also recognize that subtle differences between the controlled diet and the previous self-selected diet of the participants could contribute to observed metabolite patterns. This protocol constituted an opportunity in which to investigate the utility of metabolomic approaches in the context of controlled low vitamin B-6 status over a relatively brief period. To our knowledge, this is the first report of the use of combined global and targeted metabolomic approaches for probing micronutrient deficiency of any type.
Metabolomic tools in vitamin B-6 research
The combined NMR and targeted mass spectral analysis in this study yielded new insights into the metabolic effects of vitamin B-6 restriction. They also indicate the merits of concurrently using NMR and mass spectrometry in nutritional metabolomics. 1H-NMR has several analytical strengths when investigating micronutrient status. First, NMR can detect differences in overall plasma composition in both whole (unfiltered) plasma and following filtration to remove most proteins, lipids and lipoproteins. Minimal sample preparation is required, and aspects of the procedure are suitable for automation. Second, because of the comparatively low sensitivity of 1H-NMR (relative to mass spectrometry-based procedures), the spectra reflect the metabolic phenotype at the level of macronutrients and their metabolites. This provides a particularly useful tool in evaluating the widespread effects of its micronutrient restriction on human metabolism. The focus of 1H-NMR on relatively high-concentration plasma constituents, such as amino acids and organic acids, yields a broad assessment of the direct and indirect effects of vitamin B-6 restriction on human intermediary metabolism. Third, our results show that 1H-NMR has potential as a diagnostic technique that compliments conventional analysis of a single biomarker (e.g., plasma PLP) in characterizing marginal vitamin B-6 deficiency. Further investigation will be needed to validate this diagnostic application more fully.
Based on the current data, the sensitivity and specificity of multilevel PLS-DA of NMR spectra in reflecting overall metabolic effects of vitamin B-6 restriction are excellent. The minimal sample preparation needed for NMR analysis preserves the patterns of labile compounds such as asparagine and glutamine. However, the estimation of metabolite concentrations by spectral fitting using techniques such as illustrated here with Chenomx software should not be viewed as quantitatively conclusive. Confirmatory targeted analysis should be conducted wherever possible, especially for these metabolites in which changes are subtle. In the case of free amino acids such as asparagine, aspartate, glutamate and glutamine, differences between NMR-based and targeted mass spectrometry-based analyses may partially be attributable to NMR response to amino acid residues in the plasma peptides  that would be present in the ultrafiltrates used in NMR-Chenomx analysis. Other sources of variability in metabolite quantification by spectral fitting have been reported .
Metabolic consequences of vitamin B-6 restriction
Because pyridoxal phosphate serves a coenzymatic function in over 140 reactions, the potential for metabolic effects of vitamin B-6 deficiency extends broadly through many phases of human metabolism. As illustrated by our previous studies involving the transsulfuration enzymes cystathionine β-synthase and cystathionine γ-lyase , , PLP binding affinity is not fully informative of impairment caused by vitamin B-6 restriction. Broadly focused metabolite profiling methods allow one to assess more fully the extended effects of a nutritional condition, such as marginal vitamin B-6 deficiency.
Many reactions of the interconversion and catabolism of amino acids require PLP. PLP-dependent aspects of one-carbon metabolism (cytoplasmic and mitochondrial serine hydroxymethyltransferases and the mitochondrial glycine cleavage system) that regulate these amino acids play an important role in the supply and balance of 1C units in metabolism , . Applications of in vivo stable isotopic tracer kinetics and mathematical modeling have shown situations in which in vivo fluxes of PLP-dependent processes and substrate/product concentrations are not totally predicted by vitamin B-6 status. For example, although the in vitro activities of cytoplasmic serine hydroxymethyltransferase (SHMT) and mitochondrial SHMT are strongly influenced by vitamin B-6 status , , in vivo SHMT flux is relatively resilient to marginal deficiency , . Likewise, overall transsulfuration flux and cysteine production are maintained during marginal vitamin B-6 deficiency ,  despite the sensitivity of the cystathionine γ-lyase reaction to cellular PLP deficiency and resulting buildup of cystathionine in tissues and plasma , , . The PLP-dependent mitochondrial glycine cleavage system accounts for approximately 20 times more 5,10-methylenetetrahydrofolate production than needed for cellular methyl synthesis demands . The increased glycine concentration frequently observed in marginal vitamin B-6 deficiency presumably reflects substrate accumulation due to reduced activity of the glycine cleavage system , , . However, in vivo tracer kinetic results indicate no significant reduction in glycine cleavage system flux from vitamin B-6 restriction .
The spectral features reflecting substantial increases in asparagine and glutamine and the marked increase in the glutamine/glutamate ratio (Table 2) observed here were unexpected. Neither glutamine synthetase, which catalyzes ATP-dependent condensation of glutamate and ammonia, nor asparagine synthetase, which catalyzes the ATP-dependent transfer of the terminal amide group of glutamine to aspartate, requires PLP, nor do the catabolic enzymes glutaminase and asparaginase. In view of the importance of inter-organ transport of glutamine and the regulatory role of glutaminases , , our findings suggest fundamental perturbations in the inter-organ trafficking of these amino acids may be responsible for the changes in their plasma concentrations associated with vitamin B-6 restriction. The unexpected increase in plasma proline suggests that it may be a secondary consequence of the altered glutamate and glutamine concentrations because proline synthesis occurs by way of pyrroline 5-carboxylate . It is likely that the changes in plasma myo-inositol and trimethylamine oxide were mainly attributable to differences in diet composition from the participants' pre-study diets, because their metabolism does not directly involve PLP-dependent processes.
The activity and extent of coenzyme saturation of PLP-dependent aminotransferases has long been recognized and used diagnostically in assessing vitamin B-6 status . However, the impact of vitamin B-6 insufficiency on the profile of tissue and plasma amino acids and, particularly, the sensitivity of such reactions in maintaining anaplerotic reactions remains unclear. Increases in plasma concentrations of pyruvate and 2-oxoglutarate may reflect reduced utilization of 2-ketoacids in aminotransferase reactions or in other phases of metabolism. The lack of a significant change in formate concentration is consistent with findings from an NMR study in rats, in which deficiency of vitamin B-12, but not B-6 deficiency, caused an increase in plasma and urinary formate . Changes in the profile of tryptophan catabolic products in the kynurenine pathway are associated with vitamin B-6 deficiency , , . Targeted analysis has revealed such changes in tryptophan metabolites in the present study (reported separately).
We have previously shown that vitamin B-6 deficiency is associated with markedly increased total glutathione concentration in rat liver in direct proportion to the severity of the deficiency  and in altered glutathione synthesis in a subset of vitamin B-6 restricted healthy humans . These responses may be due to oxidative stress associated with vitamin B-6 deficiency – independent of substrate availability, a hypothesis that has been supported by mathematical modeling . The role of such oxidative stress and its impact on the plasma metabolome during vitamin B-6 deficiency is presently unknown.
This study demonstrates that our vitamin B-6 restriction protocol induces fundamental changes in the plasma metabolite profile, strongly suggesting that marginal vitamin B-6 deficiency exerts far-reaching effects on human metabolism. These observations extend our understanding of the impact and interpretation of the marginal nutritional status and will aid in hypothesis development in the design of future research regarding the mechanistic relationship between low vitamin B-6 status and chronic disease.
Representative 1H-NMR spectra of unfiltered and filtered plasma from a single participant shown before and after vitamin B-6 restriction, with signals from EDTA designated. Small arrows designate visually apparent differences in certain spectral features of these spectra.
Baseline characteristics of 23 healthy men and women participating in the study.
Conceived and designed the experiments: JFG YL PWS. Performed the experiments: YL VdS RDS CBN YP DPJ. Analyzed the data: YP KL SK NB SR TK YYC NH KEM. Wrote the paper: JFG. Clinical oversight: PWS.
- 1. Morris MS, Picciano MF, Jacques PF, Selhub J (2008) Plasma pyridoxal 5′-phosphate in the US population: the National Health and Nutrition Examination Survey, 2003–2004. Am J Clin Nutr 87: 1446–1454.
- 2. Fulgoni VL 3rd, Keast DR, Bailey RL, Dwyer J (2011) Foods, fortificants, and supplements: Where do Americans get their nutrients? J Nutr 141: 1847–1854.
- 3. Institute of Medicine, Food and Nutrition Board (1998) Dietary Reference Intakes: Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline. Washington, DC: National Academy Press.
- 4. Centers for Disease Control and Prevention (2012) Second National Report on Biochemical Indicators of Diet and Nutrition in the U.S. Population 2012. Atlanta (GA): National Center for Environmental Health. 1–484 p.
- 5. Lotto V, Choi SW, Friso S (2011) Vitamin B6: a challenging link between nutrition and inflammation in CVD. Br J Nutr: 1–13.
- 6. Morris MS, Sakakeeny L, Jacques PF, Picciano MF, Selhub J (2010) Vitamin B-6 intake is inversely related to, and the requirement is affected by, inflammation status. J Nutr 140: 103–110.
- 7. Ubbink JB, van der Merwe A, Delport R, Allen RH, Stabler SP, et al. (1996) The effect of a subnormal vitamin B-6 status on homocysteine metabolism. J Clin Invest 98: 177–184.
- 8. Lussana F, Zighetti ML, Bucciarelli P, Cugno M, Cattaneo M (2003) Blood levels of homocysteine, folate, vitamin B6 and B12 in women using oral contraceptives compared to non-users. Thromb Res 112: 37–41.
- 9. da Silva VR, Russell KA, Gregory JF (2012) Vitamin B-6. In: Erdman JW, McDonald IA, Zeisel SH, editors. Present Knowledge in Nutrition, 10th Edition. 307–320.
- 10. Leklem J (1998) Vitamin B6. In: Shils ME OJ, Shihe M, Ross AC,, editor. Modern Nutrition in Health and Disease, Ninth Edition. Blatimore, MD: Williams & Wilkins. 1413–1421.
- 11. Mackey A, Davis S, Gregory J (2005) Vitamin B6. In: Shils M, editor. Modern Nutrition in Health and Disease, 9th Ed. Baltimore, MD: Wilkins and Wilkins. 452–461.
- 12. Rimm EB, Willett WC, Hu FB, Sampson L, Colditz GA, et al. (1998) Folate and vitamin B6 from diet and supplements in relation to risk of coronary heart disease among women. JAMA 279: 359–364.
- 13. Robinson K, Arheart K, Refsum H, Brattstrom L, Boers G, et al. (1998) Low circulating folate and vitamin B6 concentrations: risk factors for stroke, peripheral vascular disease, and coronary artery disease. European COMAC Group. Circulation 97: 437–443.
- 14. Robinson K, Gupta A, Dennis V, Arheart K, Chaudhary D, et al. (1996) Hyperhomocysteinemia confers an independent increased risk of atherosclerosis in end-stage renal disease and is closely linked to plasma folate and pyridoxine concentrations. Circulation 94: 2743–2748.
- 15. Dalery K, Lussier-Cacan S, Selhub J, Davignon J, Latour Y, et al. (1995) Homocysteine and coronary artery disease in French Canadian subjects: relation with vitamins B12, B6, pyridoxal phosphate, and folate. Am J Cardiol 75: 1107–1111.
- 16. Verhoef P, Stampfer MJ, Buring JE, Gaziano JM, Allen RH, et al. (1996) Homocysteine metabolism and risk of myocardial infarction: relation with vitamins B6, B12, and folate. Am J Epidemiol 143: 845–859.
- 17. Miller JW, Green R, Mungas DM, Reed BR, Jagust WJ (2002) Homocysteine, vitamin B6, and vascular disease in AD patients. Neurology 58: 1471–1475.
- 18. Hron G, Lombardi R, Eichinger S, Lecchi A, Kyrle PA, et al. (2007) Low vitamin B6 levels and the risk of recurrent venous thromboembolism. Haematologica 92: 1250–1253.
- 19. Cattaneo M, Lombardi R, Lecchi A, Bucciarelli P, Mannucci PM (2001) Low plasma levels of vitamin B(6) are independently associated with a heightened risk of deep-vein thrombosis. Circulation 104: 2442–2446.
- 20. Vanuzzo D, Pilotto L, Lombardi R, Lazzerini G, Carluccio M, et al. (2007) Both vitamin B6 and total homocysteine plasma levels predict long-term atherothrombotic events in healthy subjects. Eur Heart J 28: 484–491.
- 21. Kelly PJ, Shih VE, Kistler JP, Barron M, Lee H, et al. (2003) Low vitamin B6 but not homocyst(e)ine is associated with increased risk of stroke and transient ischemic attack in the era of folic acid grain fortification. Stroke 34: e51–54.
- 22. Larsson SC, Orsini N, Wolk A (2010) Vitamin B6 and risk of colorectal cancer: a meta-analysis of prospective studies. JAMA 303: 1077–1083.
- 23. Johansson M, Relton C, Ueland PM, Vollset SE, Midttun O, et al. (2010) Serum B vitamin levels and risk of lung cancer. JAMA 303: 2377–2385.
- 24. Cuskelly G, Stacpoole P, Williamson J, Baumgartner T, Gregory JF (2001) Deficiencies of folate and vitamin B(6) exert distinct effects on homocysteine, serine, and methionine kinetics. Am J Physiol Endocrinol Metab 281: E1182–1190.
- 25. Davis SR, Scheer JB, Quinlivan EP, Coats BS, Stacpoole PW, et al. (2005) Dietary vitamin B-6 restriction does not alter rates of homocysteine remethylation or synthesis in healthy young women and men. Am J Clin Nutr 81: 648–655.
- 26. Davis S, Quinlivan E, Stacpoole P, Gregory JF (2006) Plasma glutathione and cystathionine concentrations are elevated but cysteine flux is unchanged by dietary vitamin B-6 restriction in young men and women. J Nutr 136: 373–378.
- 27. Lamers Y, Williamson J, Ralat M, Quinlivan EP, Gilbert LR, et al. (2009) Moderate dietary vitamin B-6 restriction raises plasma glycine and cystathionine concentrations while minimally affecting the rates of glycine turnover and glycine cleavage in healthy men and women. J Nutr 139: 452–460.
- 28. Lamers Y, Coats B, Ralat M, Quinlivan EP, Stacpoole PW, et al. (2011) Moderate vitamin B-6 restriction does not alter postprandial methionine cycle rates of remethylation, transmethylation, and total transsulfuration but increases the fractional synthesis rate of cystathionine in healthy young men and women. J Nutr 141: 835–842.
- 29. Zhao M, Lamers Y, Ralat MA, Coats BS, Chi YY, et al. (2012) Marginal vitamin B-6 deficiency decreases plasma (n-3) and (n-6) PUFA concentrations in healthy men and women. J Nutr 142: 1791–1797.
- 30. German JB, Bauman DE, Burrin DG, Failla ML, Freake HC, et al. (2004) Metabolomics in the opening decade of the 21st century: building the roads to individualized health. J Nutr 134: 2729–2732.
- 31. Zeisel SH, Freake HC, Bauman DE, Bier DM, Burrin DG, et al. (2005) The nutritional phenotype in the age of metabolomics. J Nutr 135: 1613–1616.
- 32. Park Y, Le NA, Yu T, Strobel F, Gletsu-Miller N, et al. (2011) A sulfur amino acid-free meal increases plasma lipids in humans. J Nutr 141: 1424–1431.
- 33. Gu H, Chen H, Pan Z, Jackson AU, Talaty N, et al. (2007) Monitoring diet effects via biofluids and their implications for metabolomics studies. Anal Chem 79: 89–97.
- 34. Duggan GE, Joan Miller B, Jirik FR, Vogel HJ (2011) Metabolic profiling of vitamin C deficiency in Gulo−/− mice using proton NMR spectroscopy. J Biomol NMR 49: 165–173.
- 35. Ubbink JB, Serfontein WJ, de Villiers LS (1985) Stability of pyridoxal-5-phosphate semicarbazone: applications in plasma vitamin B6 analysis and population surveys of vitamin B6 nutritional status. J Chromatogr 342: 277–284.
- 36. Pfeiffer CM, Huff DL, Gunter EW (1999) Rapid and accurate HPLC assay for plasma total homocysteine and cysteine in a clinical laboratory setting. Clin Chem 45: 290–292.
- 37. Park YK, Linkswiler H (1970) Effect of vitamin B6 depletion in adult man on the excretion of cystathionine and other methionine metabolites. J Nutr 100: 110–116.
- 38. Davis SR, Quinlivan EP, Stacpoole PW, Gregory JF 3rd (2006) Plasma glutathione and cystathionine concentrations are elevated but cysteine flux is unchanged by dietary vitamin B-6 restriction in young men and women. J Nutr 136: 373–378.
- 39. Ferrara CT, Wang P, Neto EC, Stevens RD, Bain JR, et al. (2008) Genetic networks of liver metabolism revealed by integration of metabolic and transcriptional profiling. PLoS Genet 4: e1000034.
- 40. Park Y, Kim SB, Wang B, Blanco RA, Le NA, et al. (2009) Individual variation in macronutrient regulation measured by proton magnetic resonance spectroscopy of human plasma. Am J Physiol Regul Integr Comp Physiol 297: R202–209.
- 41. Forshed J, Schuppe-Koistinen I, Jacobsson SP (2003) Peak alignment of NMR signals by means of a genetic algorithm.
- 42. Lee GC, Woodruff DL (2004) Beam search for peak alignment of NMR signals. Analytica Chimica Acta 513: 413–416.
- 43. Nicholson JK, Foxall PJ, Spraul M, Farrant RD, Lindon JC (1995) 750 MHz 1H and 1H-13C NMR spectroscopy of human blood plasma. Anal Chem 67: 793–811.
- 44. Barton RH, Waterman D, Bonner FW, Holmes E, Clarke R, et al. (2010) The influence of EDTA and citrate anticoagulant addition to human plasma on information recovery from NMR-based metabolic profiling studies. Mol Biosyst 6: 215–224.
- 45. Weljie AM, Newton J, Mercier P, Carlson E, Slupsky CM (2006) Targeted profiling: quantitative analysis of 1H NMR metabolomics data. Anal Chem 78: 4430–4442.
- 46. Weljie AM, Newton J, Jirik FR, Vogel HJ (2008) Evaluating low-intensity unknown signals in quantitative proton NMR mixture analysis. Anal Chem 80: 8956–8965.
- 47. Walsh MC, Brennan L, Pujos-Guillot E, Sebedio JL, Scalbert A, et al. (2007) Influence of acute phytochemical intake on human urinary metabolomic profiles. Am J Clin Nutr 86: 1687–1693.
- 48. Wiklund S, Johansson E, Sjostrom L, Mellerowicz EJ, Edlund U, et al. (2008) Visualization of GC/TOF-MS-based metabolomics data for identification of. Anal Chem 80: 115–122.
- 49. Chi YY, Gribbin M, Lamers Y, Gregory JF, Muller KE (2012) Global hypothesis testing for high-dimensional repeated measures outcomes. Stat Med 31: 724–742.
- 50. Storey JD (2012) A direct approach to false discovery rates. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 64: 479–498.
- 51. Lamers Y, Williamson J, Gilbert L, Stacpoole P, Gregory JF (2007) Glycine turnover and decarboxylation rate quantified in healthy men and women using primed, constant infusions of [1,2-(13)C2]glycine and [(2)H3]leucine. J Nutr 137: 2647–2652.
- 52. Lamers Y, Coats B, Ralat M, Quinlivan EP, Stacpoole PW, et al. (2011) Moderate vitamin B-6 restriction does not alter postprandial methionine cycle rates of remethylation, transmethylation, and total transsulfuration but increases the fractional synthesis rate of cystathionine in healthy young men and women. J Nutr 141: 835–842.
- 53. Solanky KS, Bailey NJ, Beckwith-Hall BM, Davis A, Bingham S, et al. (2003) Application of biofluid 1H nuclear magnetic resonance-based metabonomic techniques for the analysis of the biochemical effects of dietary isoflavones on human plasma profile. Anal Biochem 323: 197–204.
- 54. Westerhuis JA, van Velzen EJ, Hoefsloot HC, Smilde AK (2010) Multivariate paired data analysis: multilevel PLSDA versus OPLSDA. Metabolomics 6: 119–128.
- 55. Aristoteli LP, Molloy MP, Baker MS (2007) Evaluation of endogenous plasma peptide extraction methods for mass spectrometric biomarker discovery. J Proteome Res 6: 571–581.
- 56. Tredwell GD, Behrends V, Geier FM, Liebeke M, Bundy JG (2011) Between-person comparison of metabolite fitting for NMR-based quantitative metabolomics. Anal Chem 83: 8683–8687.
- 57. Scheer J, Mackey A, Gregory JF (2005) Activities of hepatic cytosolic and mitochondrial forms of serine hydroxymethyltransferase and hepatic glycine concentration are affected by vitamin B-6 intake in rats. J Nutr 135: 233–238.
- 58. Nijhout H, Gregory J, Fitzpatrick C, Cho E, Lamers K, et al. (2009) A mathematical model gives insights into the effects of vitamin B-6 deficiency on 1-carbon and glutathione metabolism. J Nutr 139: 784–791.
- 59. Lamers Y, Williamson J, Theriaque D, Shuster J, Gilbert L, et al. (2009) Production of 1-carbon units from glycine is extensive in healthy men and women. J Nutr 139: 666–671.
- 60. Curthoys NP, Watford M (1995) Regulation of glutaminase activity and glutamine metabolism. Annu Rev Nutr 15: 133–159.
- 61. Brosnan JT (2003) Interorgan amino acid transport and its regulation. J Nutr 133: 2068S–2072S.
- 62. Sauberlich HE, Canham JE, Baker EM, Raica N Jr, Herman YF (1972) Biochemical assessment of the nutritional status of vitamin B 6 in the human. Am J Clin Nutr 25: 629–642.
- 63. Lamarre SG, Molloy AM, Reinke SN, Sykes BD, Brosnan ME, et al. (2012) Formate can differentiate between hyperhomocysteinemia due to impaired remethylation and impaired transsulfuration. Am J Physiol Endocrinol Metab 302: E61–67.
- 64. Midttun O, Ulvik A, Ringdal Pedersen E, Ebbing M, Bleie O, et al. (2011) Low plasma vitamin B-6 status affects metabolism through the kynurenine pathway in cardiovascular patients with systemic inflammation. J Nutr 141: 611–617.
- 65. Vilter RW, Mueller JF, Glazer HS, Jarrold T, Abraham J, et al. (1953) The effect of vitamin B6 deficiency induced by desoxypyridoxine in human beings. J Lab Clin Med 42: 335–357.
- 66. Lima CP, Davis SR, Mackey AD, Scheer JB, Williamson J, et al. (2006) Vitamin B-6 deficiency suppresses the hepatic transsulfuration pathway but increases glutathione concentration in rats fed AIN-76A or AIN-93G diets. J Nutr 136: 2141–2147.
- 67. Lamers Y, O'Rourke B, Gilbert L, Keeling C, Matthews D, et al. (2009) Vitamin B-6 restriction tends to reduce the red blood cell glutathione synthesis rate without affecting red blood cell or plasma glutathione concentrations in healthy men and women. Am J Clin Nutr 90: 336–343.
- 68. Benderitter M, Hadj-Saad F, Lhuissier M, Maupoil V, Guilland JC, et al. (1996) Effects of exhaustive exercise and vitamin B6 deficiency on free radical oxidative process in male trained rats. Free Radic Biol Med 21: 541–549.
- 69. Taysi S (2005) Oxidant/antioxidant status in liver tissue of vitamin B6 deficient rats. Clin Nutr 24: 385–389.
- 70. Cabrini L, Bergami R, Fiorentini D, Marchetti M, Landi L, et al. (1998) Vitamin B6 deficiency affects antioxidant defences in rat liver and heart. Biochem Mol Biol Int 46: 689–697.