Betaine, choline and folate content in different cereal genotypes

Journal of Cereal Science 80 (2018) 72e79

Contents lists available at ScienceDirect

Journal of Cereal Science journal homepage: www.elsevier.com/locate/jcs

Betaine, choline and folate content in different cereal genotypes € ft a Mohammed E. Hefni a, b, *, Franziska Schaller a, Cornelia M. Wittho a b

Department of Chemistry and Biomedical Sciences, Linnaeus University, 391 82, Kalmar, Sweden Food Industries Department, Faculty of Agriculture, Mansoura University, 35516, P.O. Box 46, Mansoura, Egypt

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 October 2017 Received in revised form 15 January 2018 Accepted 21 January 2018

The importance of dietary methyl donors, e.g. betaine, choline and folate, is increasingly being recognised. This study examined variations in methyl donor concentrations in different cereals grown in Sweden. Fourteen cereal samples, representing different genera and cultivars, were analysed using HPLCUV/FLD. The content of methyl donors in the cereals varied significantly due to cereal genotype. Betaine content varied most, with 28 mg/100 g DM in oats and 176 mg/100 g DM in rye. Total choline varied less, with 67 mg/100 g DM in rye and 149 mg/100 g DM in naked barley. In wheat, the lowest concentration of folate with 36 mg/100 g DM was found, and the highest of 91 mg/100 g DM in barley. Esterified choline was the major contributor to total choline content (80e95%) in the cereals. Free choline was less abundant, ranging from 3 to 27 mg/100 g DM. 5-CHO-H4folate was the dominant folate form in all cereals, amounting to approx. 35e50% of the sum of folates, as determined after pre-column conversion. Due to the limited number of available cultivars, no interpretation regarding effects from cultivar can be made. In conclusion, the studied cereal genotypes are good sources of methyl donors, but concentrations show considerable variation between different cereals. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Betaine Choline Folate Cereals

1. Introduction Methyl donors are required for normal cell function, DNA methylation, phosphatidylcholine synthesis and protein synthesis as reviewed by Obeid (2013). Insufficient dietary intake of methyl donors is associated with a number of health risks. For example, folate deficiency is linked to development of neural tube defects, macrocytic anaemia and cardiovascular disease (Obeid, 2013). Choline deficiency is assumed to have an effect on liver disease, atherosclerosis and neurological disorders (Zeisel and Da Costa, 2009). The US Institute of Medicine (IOM, 1998) established dietary reference intakes for choline of 550 mg/day 425 mg/day for men and women, respectively. There is no recommendation for betaine. Both methyl donors, folate and betaine, are required for the remethylation of homocysteine to methionine. It was shown that supplemental folic acid, besides lowering homocysteine concentrations, even increased plasma betaine concentrations in healthy adults (Melse-Boonstra et al., 2005). Choline can be metabolised to betaine through a two-step oxidation process in the mitochondria.

* Corresponding author. Department of Chemistry and Biomedical Sciences, Faculty of Health and Life Sciences, Linnaeus University, 391 82 Kalmar, Sweden. E-mail address: [email protected] (M.E. Hefni). https://doi.org/10.1016/j.jcs.2018.01.013 0733-5210/© 2018 Elsevier Ltd. All rights reserved.

Dietary choline could therefore, in principle, supply all the requirements for both choline and betaine, but the converse is not true. However, dietary betaine is important for its choline-saving effect (Dilger et al., 2007). Choline, betaine and folate are mainly obtained from the diet and therefore the relative importance of food sources varies with dietary pattern. Cereal foods are the major source of dietary betaine and folate in the Western diet, because of both relatively high content and high cereal consumption (Ross et al., 2014). Since betaine is also a major plant osmolyte, its content in plant foods depends on growing conditions (Corol et al., 2012). In addition, the content of folate in cereals has been found to be significantly affected by genotype, maturity stage and growing conditions (Giordano et al., 2016; Piironen et al., 2008). The majority of choline in the diet appears in esterified form as phospholipid, e.g. glycerophosphocholine, phosphocholine, phosphatidylcholine and sphingomyelin, and only a small part is present as free choline (Patterson et al., 2008), which makes quantification difficult (Phillips, 2012). While LC-MS/MS is considered the golden standard for measuring individual choline forms, not all laboratories have access to this instrumentation but rather rely on UV/FLD-based methods. To our knowledge, most published choline data, except for data in the USDA database (Patterson et al., 2008), are for free choline only. Therefore, the USDA database is commonly used to

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79

calculate dietary intake of methyl donors, although this can lead to inaccuracies due to lack of data on traditional foods from other regions. Another major challenge in analysis of choline content in foods is extraction and/or hydrolysis. A number of methods have been developed for extraction of choline from foods, including the use of acid, alkali and/or enzymes (Phillips, 2012). Simplification and improvement of a previous extraction method (Hefni et al., 2015) can allow routine quantification of total choline content in a wide range of foods using a simple HPLC-FLD based method. In addition, difficulties in quantification of 5-CHO-H4folate (the dominant folate form in cereal) by HPLC-UV/FLD have been reported, due to co-elution and weak fluorescence detection (FLD) and ultraviolet (UV) absorbance of the molecule (Gujska and € ft, 2012; Kariluoto et al., 2006). Kuncewicz, 2005; Hefni and Wittho The aims of this study were to (1) quantify the content of the methyl donors choline, betaine and folate in different cereal genera and cultivars and (2) improve the methodology for quantification of both 5-CHO-H4folate and total choline in cereal foods.

2. Materials and methods 2.1. Chemicals and reagents All chemicals and enzymes were purchased from SigmaAldrich (St. Louis, USA) except for acetonitrile, which was purchased from VWR International (Stockholm, Sweden). All reagents were of p.a. grade except acetonitrile and methanol, which were of HPLC grade. Water was purified using a Milli-Q system (Merck Millipore, USA). The folate standards folic acid (PteGlu), (6S)5,6,7,8-tetrahydrofolate sodium salt (H4folate), (6S)-5-formyl5,6,7,8-tetrahydrofolate sodium salt (5-HCO-H4folate), 10formylfolic acid sodium salt (10-HCO-PteGlu) and (6S)-5methyl-5,6,7,8-tetrahydrofolate sodium salt (5-CH3-H4folate) were purchased from Merck Eprova AG (Schaffhausen, Switzerland). All were stored at
80  C until use. The actual concentration of folates and folic acid was deter€ ft, mined using the molar extinction coefficients (Hefni and Wittho 2012). Standard stock solutions of folates (z200 mg/mL) were prepared in 0.1 M sodium acetate (pH 4.6) containing 10% sodium chloride, 1% ascorbic acid and 0.1% 2,3-dimercapto-1-propanol under subdued light and stored under nitrogen atmosphere at

73


80  C for a maximum of three months. Choline and betaine stock solutions (1 mg/mL for each) were prepared in Milli-Q water, stored at
20  C and used within 1 month. Thermostable a-amylase suspension (Sigma-Aldrich, St. Louis, USA) was used for folate extraction without additional preparation. Rat serum was obtained from the animal research house at Linnaeus University (Kalmar, Sweden) and used for folate deconjugation as a source of g-glutamyl hydrolase. The serum was prepared by dialysis using 0.05 M phosphate buffer (pH ¼ 6.1) containing 0.1% 2,3-dimercapto-1-propanol at 4  C and stored portioned at
80  C. A phospholipase D solution (400 U/mL) was prepared using 50 mM Tris-HCl buffer (pH ¼ 8), stored at
80  C and used to check the efficiency of acid hydrolysis during choline analysis. 2.2. Cereal samples Grain from two cultivars each of rye (genus Secale), barley (genus Hordeum) and oat (genus Avena) and eight cultivars of wheat (genus Triticum) (Table 1), harvested during 2015 and 2016, €tarna in was obtained from different mills and producers: V€ astgo €taland county, Warbro mill and Saltå mill in So €derman V€ astergo €derman county and Isga €rde farm on county, Nibble farm in So € Oland, Sweden. Grain samples (250e500 g) were vacuum-packed in polyethylene bags and stored at
20  C until analysis. Before analysis, the grain was milled using an ultracentrifuge mill €gana €s, Sweden). (Cyclotec 1093 sample mill, Foss Tecator, Ho 2.3. Analysis 2.3.1. Betaine extraction, derivatisation and quantification Betaine extraction was carried out in duplicate based on the method of Hefni et al. (2016). Briefly, milled cereals (0.15 g) were homogenized in 5 mL Milli-Q water, shaken for 5 min and centrifuged (5 min, 2000g). The supernatant (containing the betaine) was transferred into a plastic tube. The extraction was repeated once more, and the supernatants combined. After extraction, 4 mL dichloromethane were added to the combined extracts and the mixture was shaken for 5 min in order to remove any hydrophobic compounds. After centrifugation (5 min, 2000 g), the aqueous (top) layer was transferred to another tube and stored (
20  C) for derivatisation and analysis by HPLC-UV within one week. Betaine in

Table 1 Betaine and free and total choline content (mg/100 g DM) in cereal samples. Cereal genus

Species

Cultivar (origin)

Harvest year

Dry matter (g/100 g)

Betaine content (mg/100 g)

Free

Total

Rye Secale

S. cereale S. cereale

€stgo €tarna) Schmidt rye (Va Rye (Saltå mill)

2015 2016

87 92

176 A 153 B

14 11

70 67

Wheat Triticum

T. spelta T. spelta T. aestivum

Spelt (Warbro mill) Spelt (Saltå mill) € Oland spring wheat €rde farm) (Isga Wheat (Saltå mill) Jacoby Borst lantvete (Nibble farm) €stgo €tarna) Wheat Dala lantvete (Va €stgo €tarna) Emmer (Va Emmer (Warbro mill)

2016 2016 2016

89 92 88

143 137 115

A

4 4 18

108 103 101

2016 2015

92 88

132 123

AB

3 7

97 A 109 A

2015 2015 2016

89 88 88

98 94 83

9 9 8

106 108 117

A

A

T. aestivum T. aestivum T. aestivum T. dicoccum T. dicoccum

A BC

AB

CD CD D A

Choline content (mg/100 g)

A A A A A

A A

Barley Hordeum

H. vulgare nudum H. vulgare

Naked barley (Warbro mill) Barley (Saltå mill)

2016 2016

86 90

98 46 B

27 5

149 119

Oat Avena

A. sativa nuda A. sativa

Naked oat (Warbro mill) Oat (Saltå mill)

2016 2016

91 93

44 A 28 B

13 ± 2 4

135 A 101 B

Different superscripts within cereal genus represent significant differences (P < 0.05). N ¼ 2-4.

A

74

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79

food extracts and aqueous standard solutions was derivatised using 20 -naphthacyl triflate according to existing methods (Hefni et al., 2016). Briefly, 20 mL of sample (or standard or water for blank) was transferred into a tube with 1 mL of extraction solvent (90% acetonitrile and 10% methanol) and vortexed. Ca 80 mg anhydrous sodium sulfate was added and the vial was shaken (30e60 min) and centrifuged (3 min, 13,000 g). 250 mL of the supernatant was added to 50 mL of 20 -naphthacyl triflate (100 mM in acetonitrile) and 5 mL of an aqueous suspension of base (MgO, 0.1 g/mL). The mixture was vortexed and centrifuged (3 min, 13,000 g). An aliquot of the supernatant was transferred to a capped HPLC vial for analysis. Betaine was quantified by HPLC with UV detection (Agilent 1260, Agilent Technologies, USA) on a strong cation exchange column (150 mm  4.6 mm x 5 mm) (Phenomenex Luna, CA, USA) with a guard column with the same packing material. The separation was performed under isocratic conditions using a mobile phase containing of 5% of an aqueous solution of 15 mmol/L triethylamine, 30 mmol/L succinic acid and 95% acetonitrile. The injection volume was 50 mL, the flow rate 1.2 mL/min, the oven was set to 40  C and the run time was 22 min. Quantification was based on UV detection (249 nm) using a multilevel (n ¼ 5) external calibration curve (R2 ¼ 0.9995). 2.3.2. Choline extraction, derivatisation and quantification The free choline content was analysed in the aqueous extract (section 2.3.1) used for betaine analyses. The quantification of total choline was carried out after hydrolysis of phospholipids using 1 M solution of HCl as follows: 250 mg of the cereal sample were homogenized in 10 mL 1 M HCl and the mixture was incubated at 60  C overnight (18 h). A shorter incubation time (4 h at 70  C) was also tested. Thereafter, samples were cooled to room temperature, neutralised (pH 5-6) using NaOH (10 M and/or 1 M) and the final volume was recorded by weighing. The hydrolysate was centrifuged (2600 g for 15 min) and filtered through 0.45 mm PES syringe filter if required. In order to check the stability of choline during the acid hydrolysis, different concentrations (5e500 mg/mL) of choline standard were treated as sample (overnight incubation with 1 M HCl) and were used for generating the calibration curve. A phospholipase D solution (400 U/mL) was prepared in 50 mM Tris-HCl buffer (pH ¼ 8) and was used to check the efficiency of the acid hydrolysis. An aliquot from the hydrolysate (500 mL) was incubated at 37  C for 15 min with phospholipase D (250 mL, 50 U) according to the method of Fu et al. (2012). Extracted free choline (water extract) and total choline (after acid hydrolysis) and aqueous choline standards were derivatised using 1-naphthyl isocyanate according to a previously described method (Hefni et al., 2015). Briefly, 20 mL of sample (or standard or water “blank”) were transferred into 1 mL of acetonitrile and approximately 80 mg dry magnesium oxide and vortexed. After addition of 60 mL of NaOH (1 M), samples were again vortexed and 20 mL 1-naphthyl isocyanate was added. The samples were shaken at room temperature for 15 min. Sixty microliters of water was added to each sample, the mixture was vortexed and centrifuged at 13,000 g for 5 min. Aliquots of the supernatant were transferred into capped HPLC vials for analysis. Quantification of choline was carried out by HPLC with fluorescence detection (Agilent 1260, Agilent Technologies, USA). The choline derivative was separated using the same column as for betaine quantification (see 2.3.1). The injection volume was 50 mL, flow rate was 1 mL/min, the column temperature was maintained at 40  C and the run time was 15 min. Separation was performed isocratically using a mobile phase containing 10 mmol/L tetramethylammonium hydroxide and 20 mmol/L glycolic acid and 15% water in acetonitrile. Quantification was based on fluorescence

detection (ex/em 220/350 nm) using a multilevel (n ¼ 5) external calibration curve. The calibration curves for free and total choline (treated as sample: overnight incubation with 1 M HCl) were linear from 4 mg/mL to 1000 mg/mL, with correlation coefficient (R2) ¼ 0.9994 and 0.9985 respectively. 2.3.3. Folate extraction, purification and quantification Folate quantification, including extraction, purification and quantification, in cereal grains was performed in duplicate after di€ ft, 2012). Briefly, approxienzyme extraction (Hefni and Wittho mately 0.5e1 g milled sample was extracted in a boiling water bath in 15 mL phosphate buffer after addition of thermostable a-amylase (60 mL). After cooling on ice and centrifugation, folate polyglutamates were deconjugated by incubation at 37  C for 2 h after addition of 40 mL dialysed rat serum per mL sample extract and aamylase (40 mL). Purification of sample extracts was carried out by solid phase extraction on strong anion exchange cartridges (500 mg/3 mL HyperSep column, Thermo Scientific, Rockwood, USA). In brief, the cartridges were conditioned with 5 mL methanol and water each, and 3.5 mL sample extract were loaded. Columns were washed with 5 mL of water. The void volume was eluted with 0.5 mL elution buffer (0.1 M Na acetate containing 10% NaCl, 1% (w/ v) ascorbic acid and 0.1 M (v/v) 2,3-dimercapto-1-propanol). Folate was eluted using 4 mL of elution buffer. In order to quantify 5-CHO-H4folate, a simple rapid method for quantitative pre-column chemical conversion of 5-CHO-H4folate to H4folate was developed. First, 5-CHO-H4folate was converted to 5,10-methenyltetrahydrofolate (5,10-CHþ-H4folate) after acidification using HCl and then reduced to H4folate using sodium borohydride (NaBH4). Different volumes of HCl and concentrations of NaBH4 were tested. The final protocol for the conversion was as follows: 1 mL purified food extract was transferred to a 5 mL volumetric flask, 0.5 mL conc. HCl was added and acidification was allowed to occur for 20 min at room temperature. Thereafter, the volume was made up to 5 mL using NaBH4 (z0.31 mol prepared in 0.1 M sodium acetate containing 10% sodium chloride, 1% ascorbic acid and 0.1% 2,3-dimercapto-1-propanol (elution buffer). The solution was left at room temperature for a further 60 min. One-mL aliquots of folate standard solutions (containing z 100, 75, 175, 250 and 155 ng/mL H4folate, 5-CH3-H4folate, 5-HCOH4folate, 10-HCO-PteGlu and PteGlu, individually or as a mixture) were treated as described above for conversion conditions. Standard dilutions representing a multilevel (n ¼ 8) external calibration were treated using conditions for pre-column conversion (to check for stability and linearity) and were used for quantification of H4folate (deriving from native H4folate and converted 5-HCOH4folate). In order to quantify the contribution of the individual folate forms to the sum of folate, folate in food samples was quantified before and after pre-column conversion by HPLC-FLD/ UV (Agilent 1260, Agilent Technologies, USA). Folates were separated on an C18-PFP column (3 mm, 150  4.6 mm) with a C18 (1 3 mm, 10 x 3 mm) guard column (Chromtech ACE, Scantec Nordic, Aberdeen, Scotland) under linear gradient elution conditions using, for non-converted samples, phosphate buffer (pH 2.3) and acetonitrile at a flow rate of 0.4 mL/min, an injection volume of 20 mL and a run time of 42 min. For converted samples, the injection volume was 100 mL and run time 35 min. The autosampler and column temperature were maintained at 8  C and 23  C, respectively. Quantification was based on fluorescence detection (ex/em 290/ 360 nm for reduced folates and 360/460 nm for 10-HCO-PteGlu), multiwavelength detector (290 nm) for oxidised folate forms and a multilevel (n ¼ 8) external calibration curve. The calibration curves were linear up to 200 ng/mL from: 0.2 ng/mL for H4folate (including native H4folate plus converted 5-HCO-H4folate) (R2 ¼ 0.9997), 0.4 for 5-CH3-H4folate (R2 ¼ 0.9999) and 8 ng/mL for 10-HCOePteGlu

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79

(R2 ¼ 0.999). 2.4. Quality control of analytical methods Flour from milled rye (S. cereale, Saltå mill) was stored in portions at
20  C and used as in house-control sample. CRM 121 wholemeal flour was obtained from the Institute for Reference Material and Measurement (Geel, Belgium) and used for the validation of folate quantification. Recovery tests were performed by addition of choline, betaine and folate standard (in duplicate) at two concentrations (50 and 100% of initial content) to one of the cereal samples (naked oat, Warbro mill), followed by extraction of the respective compound. 2.5. Dry matter determination Dry matter (DM) was determined in triplicate according to AOAC (2000), using 1e2 g of milled cereal grain. 2.6. Calculation and statistics All results are expressed as mean of duplicate analyses. The variation between analytical replicates was always lower than 10%. The sum of individual folate forms was expressed as mg PteGlu/

75

100 g DM, after conversion using a molecular weight of 445.4 for H4folate, 459.5 for 5-CH3-H4folate, 469.4 for 10-HCO-PteGlu, 473.5 for 5-HCO-H4folate and 441.4 for PteGlu. Betaine and free and total choline concentrations were expressed in mg/100 g DM. One-way ANOVA and Tukey's pair-wise comparison, with level of significance set at P < 0.05, were used to analyse differences between cereal genera and within the different species. Statistical analyses were carried out using Minitab (Minitab Ltd, Coventry, United Kingdom). 3. Results 3.1. Method improvement 3.1.1. Acid hydrolysis of esterified choline The comparison of sample pre-treatment methods showed higher measurable total choline content after overnight incubation with HCl (1M, 18 h, 60  C) compared with a short incubation (4 h, 70  C) (data not shown). Subsequent treatment, after overnight incubation, with phospholipase D did not increase the measurable choline content, confirming completeness of the acid hydrolysis. The enzymatic hydrolysis step was therefore excluded from the final method for sample hydrolysis. Choline standards treated as the samples (1 M HCl for 18 h, 60  C) were linear in the range tested

Fig. 1. FLD chromatogram of (A) folate standard mixture, before (solid line) and after pre-column conversion (dotted line), (B) naked barley, before (solid line) and after conversion (dotted line). Peaks numbers refer to (1) H4folate, (2) 5-CH3-H4folate, (3) 5-HCO-H4folate, (4) 10-HCO-PteGlu and (1 þ 3) H4folate plus converted 5-HCO-H4folate. Chromatograms A and B show that 5-HCO-H4folate was completely converted to H4folate after acidification with conc. HCl and reduction using NaBH4 (for conversion conditions refer to section 2.3.3). Concentrations of individual folate forms in standard mixture were 23 ng/mL H4folate, 8 ng/mL 5-CH3-H4folate, 34 ng/mL 10-HCO-PteGlu and 44 ng/mL 5-HCO-H4folate. Concentrations in the naked barley extract were 5 ng/mL H4folate (solid line), 6 ng/mL 5-CH3-H4folate and 9 ng/mL H4folate (dotted line, native H4folate plus converted 5-HCOH4folate).

76

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79

(4e1000 mg/mL; R2 ¼ 0.9985), showing good stability of the compound during hydrolysis. These results demonstrate the necessity for a longer incubation time when lower temperature is used (60  C) during hydrolysis for reliable choline quantification results without affecting the stability of choline content. 3.1.2. Conversion of 5-HCO-H4folate to H4folate 5-HCO-H4folate was efficiently converted to H4folate (>95%), with a minor part (<5%) being converted to 5-CH3-H4folate (Fig. 1). According to the literature (Ndaw et al., 2001), 5-CHO-H4folate is converted to 5,10-CHþ-H4folate after incubation with conc. HCl. Under acidic conditions, we found that the resulting 5,10-CHþH4folate is reduced to H4folate after addition of NaBH4. Increasing the volume of HCl during acidification (from 0.5 to 2 mL/mL sample) resulted in complete conversion of 5-HCO-H4folate to H4folate (without producing any 5-CH3-H4folate), but the stability of other folate vitamers, 10-HCO-PteGlu and PteGlu, was negatively affected and their peak area was decreased by 50%. The final volume of HCl to sample was therefore adjusted to 0.5:1 mL (V/V). All folate forms were stable under these conditions. The calibration curve generated from all folate standards treated under conversion conditions was linear up to 200 ng/mL and resulted in correlation coefficients above 0.999 for H4folate (native H4folate plus converted 5-HCOH4folate), 5-CH3-H4folate and 10-HCO-PteGlu, demonstrating satisfactory stability of the individual folate forms. 3.1.3. Quality control of the analytical methods Between-day precision coefficient of variation (CV) for the inhouse control sample (rye, Saltå mill) was 6% for folate (n ¼ 5), 7% for betaine (n ¼ 8), 10% for total choline (n ¼ 6) and 15% for free choline content (n ¼ 4). The variation between two analytical sample replicates was lower than 10% for all methyl donors analysed. The average relative recovery (two concentrations of 50 and 100% of the initial analyte content before extraction) in a cereal sample (naked oat, n ¼ 4, duplicates at two concentrations) was 85 ± 3%, 94 ± 13%, 102 ± 14% and 101 ± 8% for betaine, free choline, total choline and folate expressed as folic acid, respectively. In CRM 121 (wholemeal wheat flour), H4folate (native H4folate plus converted 5-HCO-H4folate) and 5-CH3-H4folate were quantified and the sum of folates, expressed as folic acid, was found to be 41 mg/ 100 g (n ¼ 2). Similar results for the folate content in CRM 121 (39 mg/100 g), determined using UPLC, have been reported by Edelmann et al. (2013). The certified value of total folate content according to microbiological assay was 50 ± 7 mg/100 g. 3.2. Content of methyl donors in cereal samples Mean betaine content (mg/100 g DM) in cereal samples varied significantly depending on cereal genus and species (Table 1). The highest betaine content (almost 180 mg/100 g) was found in rye. In the other cereals (wheat, barley and oat), the highest and lowest betaine content varied up to 1.5- to 2-fold (Fig. 2). Free choline content was found in small amounts representing 3e20% of the total choline content in the cereals, ranging from 3 mg/100 g in wheat (Saltå mill) to 27 mg/100 g DM in naked barley (Warbro mill) (Table 1). Little or no variation in the free choline content was found among rye samples, while in wheat, rye and barley samples free choline varied up to 5-fold (Table 1). Total choline content ranged from around 70 and 150 mg/100 g DM (Table 1). The lowest content was found in the rye samples (Fig. 2). Interestingly samples with the highest content of total choline (barley and oat) had the lowest content of betaine. On the other hand, rye, with the highest content of betaine, had the lowest amount of total choline (Fig. 2). The highest folate content (expressed as folic acid), above 60 mg/

Fig. 2. Betaine, total choline (mg/100 g DM) and folate (mg/100 g DM) content in Swedish cereals. Grey dots represent data of two cultivars each of rye, barley and oat and eight cultivars of wheat (each analysed in duplicate). Black dots represent the mean for each cereal genotype. Different superscripts represent significant differences (P < 0.05).

100 g DM, was found in barley, rye and one of the oat species (naked oat) (Fig. 2, Table 2). In wheat, the folate content ranged from 35 to 60 mg/100 g DM. The folate vitamer composition (Fig. 3) was similar in different cereal samples. 5-CHO-H4folate was the dominant folate form in all cereal foods, amounting to approximately 35e50%. 10-HCO-PteGlu was below the limit of quantification (0.8 mg/100 g) and folic acid was not detected (Table 2).

4. Discussion 4.1. Methyl donors in cereal samples The high content of methyl donors in studied whole cereal grains of the current investigation is well in line with data reported by others (Giordano et al., 2016; Piironen et al., 2008; Ross et al.,

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79

77

Table 2 Folate content (mg/100 g DM) in cereal samples. Cereal

Cultivar (origin)

H4folatea

5-CH3- H4folate

Sum of folate as folic acid b

Rye

€stgo € tarna) Schmidt rye (Va Rye (Saltå mill)

47 48

25 19

71 66

A

Spelt (Warbro mill) Spelt (Saltå mill) € €rde farm) Oland spring wheat (Isga Wheat (Saltå mill) Jacoby Borst lantvete (Nibble farm) € tarna) Wheat Dala lantvete (V€ astgo € tarna) Emmer (V€ astgo Emmer (Warbro mill)

31 41 40

20 17 16

50 57 55

A

28 29 28 31 37

9 10 12 12 18

36 B 38 B 39 B 42 B 54 A

Barley

Naked barley (Warbro mill) Barley (Saltå mill)

49 75

17 17

65 A 91 B

Oat

Naked oat (Warbro mill) Oat (Saltå mill)

46 28

26 12

70 A 39 B

Wheat

A

A A

a

H4folate represents native H4folate plus converted 5-HCO-H4folate. Sum of folate expressed as folic acid (mg/100 g DM) after conversion using a molecular weight of 445.4 for H4folate, 459.5 for 5-CH3-H4folate and 573.5 for 5-CHO-H4folate. Folic acid was not detected and 10-CHO-PteGlu was below the limit of detection (0.8 mg/100 g) and could not be accurately quantified. Different superscripts within cereal genus represent significant differences (P < 0.05). b

Fig. 3. Folate distribution (%) in different cereal genera and species. 5-HCO-H4folate was calculated by difference from the sum of H4folate and 5-HCO-H4folate quantified as H4folate (after pre-column conversion) minus the content of H4folate (without conversion) using a molecular weight of 445.4 for H4folate and 473.5 for 5-CHO-H4folate.

2014). However, the content of methyl donors in cereals and pseudocereals is generally variable, as e.g. rice, maize, millet, sorghum and buckwheat contain low or very low amount of betaine and choline (Bruce et al., 2010; Ross et al., 2014). By choice of cereal genotypes rich in methyl donors, cereals could be an important dietary source of methyl donors in the European diet (Ross et al., 2014). However, caution is required when evaluating cereal-based foods as a methyl donor source, as the content varies significantly with cereal species and growing conditions (Corol et al., 2012). Moreover, data from the USDA database, which are widely used

when calculating the dietary intake of methyl donors, although a valuable resource, need to be confirmed for European cereals. Data from the HEALTHGRAIN project (Shewry et al., 2013) show that the content of methyl donors in European cereals is on average higher than reported in the USDA database (Patterson et al., 2008). Furthermore, every country has specific cereal wildtypes/landraces for which data are not included in the US database. The comparison of betaine content in different cereals in this study (Fig. 2) showed a 6-fold variation between genera (P < 0.001) and between species within each genus (P < 0.001 in oat to 0.029 in

78

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79

rye). Similarly, it has been reported previously that cereal betaine content is affected by genotype and growing conditions and varies widely in samples of wheat (97e294 mg/100 g DM), rye (176e298 mg/100 g DM), barley (71e136 mg/100 g DM) and oat (29e55 mg/100 g DM) (Corol et al., 2012). It was assumed that betaine, a major osmolyte, is accumulated during osmotic stress conditions in plants for cell volume regulation (Burg et al., 2007). Corol et al. (2012) found highest betaine content in rye followed by wheat, barley and oat. Betaine content in cereal samples of the current study follow the same order (Fig. 2). Total choline content varied only two-fold between the different cereals analysed (Table 1). In general, the total choline concentration after acid hydrolysis determined in this study (Table 1) was 2to 3-fold higher than values on the sum of esterified phospholipids reported in the USDA database (Patterson et al., 2008). This could be partly attributable to lack of standards for individual esterified choline forms when calculating the sum for total choline. The low content of free choline reported in this study is similar to the range of 5e32 mg/100 g reported for various samples of barley, oat, rye, and wheat by others (Corol et al., 2012; Patterson et al., 2008), and follow the same order with barley z rye > wheat > oats. Our data confirm that cereals are not a rich source of free choline. Folate content was also significantly affected by cereal genus and species (Fig. 2), which is in agreement with previous findings (Giordano et al., 2016; Gujska and Kuncewicz, 2005; Hefni and €ft, 2012; Piironen et al., 2008). The average folate content Wittho in the rye cultivars was up to 35% higher than that in wheat (Gujska €ft, 2012). The mean folate and Kuncewicz, 2005; Hefni and Wittho content of the two rye cultivars analysed (66e71 mg/100 g DM) was within the range (57e135 mg/100 g DM) reported for different rye cultivars by others (Gujska and Kuncewicz, 2005; Kariluoto et al., 2001; Shewry et al., 2010). The variation (1.5-fold) observed between the different wheat species is similar to that (2-fold) reported by others (Piironen et al., 2008) and is higher than that reported for rye (z15% variation between the highest and lowest content) (Shewry et al., 2010). The same magnitude of variation in folate content was observed for barley (1.5-fold) and oat (1.8-fold). The distribution of individual folate vitamers in the different cereal samples was similar (Fig. 3). 5-HCO-H4folate was the most abundant vitamer in most cereals, contributing on average 35e50% of quantified vitamers, followed by 5-CH3-H4folate (18e38%) and H4folate (17e35%). A similar folate pattern, with 5-HCO-H4folate as most the abundant folate form, has been reported for wheat (>40%) (Piironen et al., 2008), barley (27e42%) (Edelmann et al., 2013) and rye (25e69%) (Gujska et al., 2009; Kariluoto et al., 2001). However, according to Edelmann et al. (2012), 5-CH3-H4folate represents about 50% of the dominant folate form in oats. A limitation of the present study is that the data obtained for specific cultivars derived from raw material harvested in only one season and a limited number of cultivars. As only two varieties of rye, oat and barely were analysed for methyl donors, therefore, these data should be regarded as indicative. Further systematic studies are warranted to explore both, variation between harvest year and cultivar. 4.2. Analytical issues To our knowledge, no data on total choline content in foods have been reported previously except those in the USDA database, which presents a calculated sum of individual esterified cholines. Direct quantification of different phospholipids (e.g. glycerophosphocholine, phosphocholine, phosphatidylcholine and sphingomyelin) requires advanced equipment such as NMR or LCMS/MS (Graham et al., 2009; Patterson et al., 2008; Zhao et al., 2011). In a

previous study (Hefni et al., 2015), we successfully developed an acid hydrolysis approach where total choline is extracted in a mixture of solvents (chloroform/methanol/water, 1:2:0.8) and then hydrolysed using 1 M HCl in 90% acetonitrile at 115  C. However, this resulted in two immiscible phases (organic and aqueous), which needed to be carefully separated. In the present study, we simplified the sample preparation procedure by incubating the samples overnight at 60  C without prior treatment with 1 M HCl. This simplified method was shown to be suitable for a variety of sample matrices (e.g. different pulses; data not shown), overcame the problem of separating phases, eliminated the need for chloroform and led to similar results with or without phospholipase D treatment. Difficulties in quantification of 5-CHO-H4folate (the dominant folate form in cereals) by HPLC-UV/FLD have been reported (Gujska € ft, 2012; Kariluoto et al., and Kuncewicz, 2005; Hefni and Wittho 2006). The peak for 5-HCO-H4folate is often masked and the compound has a low response in fluorescence and UV detection, which results in inaccurate quantification. In the present study, a simple approach by which 5-HCO-H4folate is converted to H4folate (>90%) after acidification with HCl and subsequent reduction with NaH4B was established. Results from this study show that this approach can be used to obtain more accurate quantitative data for 5-HCO-H4folate in cereal samples. 5. Conclusions The data presented here confirm that studied cereals are good sources of methyl donors, but that concentrations show considerable variation between different cereal genotypes. Hence, information regarding the effects of genotype on methyl donor content in cereals could be helpful in choosing cereals with a high content when aiming to improve dietary intake. At present, no data on methyl donors betaine and choline are available in the food composition database of the Swedish Food Agency. Data from the USDA database are often used instead for estimation of betaine and choline content in Swedish foods, which can give misleading results. This emphasises the importance of national food databases for accurate estimation of dietary intake of the methyl donors betaine and choline in the Swedish population. Improvement of the methodology by using simple acid hydrolysis of phospholipids to estimate total choline enables widespread routine analysis of choline in foods, with no requirement for advanced analytical equipment. Furthermore, by improvement of the methodology for folate quantification, more accurate data for the dominant cereal folate form 5-HCO-H4folate could be obtained using widely available HPLC-UV/FLD. Conflicts of interest The authors declare that they have no conflict of interest. Funding This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors. References AOAC, 2000. Official Methods of Analysis, seventeenth ed. Association of Official Analytical Chemists, Arlington, VA. Bruce, S.J., Guy, P.A., Rezzi, S., Ross, A.B., 2010. Quantitative measurement of betaine and free choline in plasma, cereals and cereal products by isotope dilution LCMS/MS. J. Agric. Food Chem. 58, 2055e2061. Burg, M.B., Ferraris, J.D., Dmitrieva, N.I., 2007. Cellular response to hyperosmotic stresses. Physiol. Rev. 87, 1441e1474.

M.E. Hefni et al. / Journal of Cereal Science 80 (2018) 72e79 Corol, D.-I., Ravel, C., Raksegi, M., Bedo, Z., Charmet, G., Beale, M.H., Shewry, P.R., Ward, J.L., 2012. Effects of genotype and environment on the contents of betaine, choline, and trigonelline in cereal grains. J. Agric. Food Chem. 60, 5471e5481. Dilger, R.N., Garrow, T.A., Baker, D.H., 2007. Betaine can partially spare choline in chicks but only when added to diets containing a minimal level of choline. J. Nutr. 137, 2224e2228. € m, L., Piironen, V., 2012. Folate in oats and its Edelmann, M., Kariluoto, S., Nystro milling fractions. Food Chem. 135, 1938e1947. €m, L., Piironen, V., 2013. Folate in barley grain and Edelmann, M., Kariluoto, S., Nystro fractions. J. Cereal. Sci. 58, 37e44. Fu, S., Tao, B., Lai, S., Zhang, J., Yiping, R., 2012. Determination of total choline by liquid chromatography-electrospray ionization-tandem mass spectrometry in infant formulas. J. AOAC Int. 95, 157e162. Giordano, D., Reyneri, A., Blandino, M., 2016. Folate distribution in barley (Hordeum vulgare L.), common wheat (Triticum aestivum L.) and durum wheat (Triticum turgidum durum Desf.) pearled fractions. J. Sci. Food Agric. 96, 1709e1715. Graham, S.F., Hollis, J.H., Migaud, M., Browne, R.A., 2009. Analysis of betaine and choline contents of aleurone, bran, and flour fractions of wheat (Triticum aestivum L.) using 1H nuclear magnetic resonance (NMR) spectroscopy. J. Agric. Food Chem. 57, 1948e1951. Gujska, E., Kuncewicz, A., 2005. Determination of folate in some cereals and commercial cereal-grain products consumed in Poland using trienzyme extraction and high-performance liquid chromatography methods. Eur. Food Res. Technol. 221, 208e213. Gujska, E., Michalak, J., Klepacka, J., 2009. Folates stability in two types of rye breads during processing and frozen storage. Plant Foods Hum. Nutr. 64, 129e134. Hefni, M., McEntyre, C., Lever, M., Slow, S., 2015. A simple HPLC method with fluorescence detection for choline quantification in foods. Food Analytical Methods 8, 2401e2408. Hefni, M., McEntyre, C., Lever, M., Slow, S., 2016. Validation of HPLC-UV methods for the quantification of betaine in foods by comparison with LC-MS. Food Analytical Methods 9, 292e299. €ft, C.M., 2012. Effect of germination and subsequent oven-drying Hefni, M., Wittho on folate content in different wheat and rye cultivars. J. Cereal. Sci. 56, 374e378. IOM, 1998. Institute of Medicine. Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin, and Choline. The National Academies Press. Kariluoto, S., Liukkonen, K.-H., Myllym€ aki, O., Vahteristo, L., Kaukovirta-Norja, A., Piironen, V., 2006. Effect of germination and thermal treatments on folates in rye. J. Agric. Food Chem. 54, 9522e9528.

79

Kariluoto, S., Vahteristo, L.T., Piironen, V.I., 2001. Applicability of microbiological assay and affinity chromatography purification followed by high-performance liquid chromatography (HPLC) in studying folate contents in rye. J. Sci. Food Agric. 81, 938e942. Melse-Boonstra, A., Holm, P.l.I., Ueland, P.M., Olthof, M., Clarke, R., Verhoef, P., 2005. Betaine concentration as a determinant of fasting total homocysteine concentrations and the effect of folic acid supplementation on betaine concentrations. Am. J. Clin. Nutr. 81, 1378e1382. , M., Aoude -Werner, D., Lahe ly, S., Hasselmann, C., 2001. Ndaw, S., Bergaentzle Determination of folates in foods by high-performance liquid chromatography with fluorescence detection after precolumn conversion to 5-methyltetrahydrofolates. J. Chromatogr. A 928, 77e90. Obeid, R., 2013. The metabolic burden of methyl donor deficiency with focus on the betaine homocysteine methyltransferase pathway. Nutrients 5, 3481e3495. Patterson, K.Y., Bhagwat, S.A., Williams, J.R., Howe, J.C., Holden, J.M., 2008. USDA Database for the Choline Content of Common Foods Release Two. USDA, Agricultural Research Service, Beltsville Maryland, US. Phillips, M., 2012. Analytical approaches to determination of total choline in foods and dietary supplements. Anal. Bioanal. Chem. 403, 2103e2112. } , Z., 2008. Folate in wheat genotypes Piironen, V., Edelmann, M., Kariluoto, S., Bedo in the HEALTHGRAIN diversity screen. J. Agric. Food Chem. 56, 9726e9731. Ross, A.B., Zangger, A., Guiraud, S.P., 2014. Cereal foods are the major source of betaine in the Western diet e analysis of betaine and free choline in cereal foods and updated assessments of betaine intake. Food Chem. 145, 859e865. Shewry, P.R., Hawkesford, M.J., Piironen, V., Lampi, A.-M., Gebruers, K., Boros, D., Andersson, A.A.M., Åman, P., Rakszegi, M., Bedo, Z., Ward, J.L., 2013. Natural variation in grain composition of wheat and related cereals. J. Agric. Food Chem. 61, 8295e8303. Shewry, P.R., Piironen, V., Lampi, A.-M., Edelmann, M., Kariluoto, S., Nurmi, T., Fernandez-Orozco, R., Andersson, A.A.M., Åman, P., Fras, A., Boros, D., Gebruers, K., Dornez, E., Courtin, C.M., Delcour, J.A., Ravel, C., Charmet, G., Rakszegi, M., Bedo, Z., Ward, J.L., 2010. Effects of genotype and environment on the content and composition of phytochemicals and dietary fiber components in rye in the HEALTHGRAIN diversity screen. J. Agric. Food Chem. 58, 9372e9383. Zeisel, S.H., Da Costa, K.-A., 2009. Choline: an essential nutrient for public health. Nutr. Rev. 67, 615e623. Zhao, Y.-Y., Xiong, Y., Curtis, J.M., 2011. Measurement of phospholipids by hydrophilic interaction liquid chromatography coupled to tandem mass spectrometry: the determination of choline containing compounds in foods. J. Chromatogr. A 1218, 5470e5479.