Determination of selected water-soluble vitamins (thiamine, riboflavin, nicotinamide and pyridoxine) from a food matrix using hydrophilic interaction liquid chromatography coupled with mass spectroscopy

Highlights

• Simultaneous analysis of water-soluble vitamins utilising LC-MS.

• HILIC chromatography employed.

• Quick and simple vitamin extraction from a complex food matrix.

• Method suitable for high throughput analysis, offering good sensitivity.

Abstract

Water-soluble vitamins are essential dietary components with a multitude of important functions that require quantification from food sources to characterise the nutritional status of food. In this study, we have developed a hydrophilic interaction chromatography (HILIC) based method coupled to single-quadrupole mass spectrometry (MS) for the analysis of selected water-soluble vitamins. Due to their involvement in energy release from macronutrients, the quantification of thiamine (B₁), riboflavin (B₂), nicotinamide (B₃) and pyridoxine (B₆) offers significant value in food analysis.

A commercially available vegetable soup was selected as the food matrix for this study and utilised to develop an efficient extraction procedure for the vitamins of interest. Vitamins were extracted using meta-phosphoric acid coupled with a reducing agent, DL-dithiothreitol (DTT) to produce the parent compound. The extracted vitamins were then analysed using an LC-MS system with electrospray – atmospheric pressure ionization (ES-API) source, operated in positive single ion monitoring (SIM) mode. The MS provided good linearity within the investigated range from 5 to 400 ng/mL with coefficient of determination (r²) ranging from 0.98 to 0.99. Retention times (0.65–9.04 min) were reproducible and no coelution between vitamins was observed. Limit of detection (LOD) varied from 2.4 to 9.0 ng/mL and limit of quantification (LOQ) was from 8 to 30 ng/mL, comparable to previously published studies. The extraction method provided good intra-day (%CV 1.56–6.56) and inter-day precision (%CV 8.07–10.97). Standard injections were used as part of quality control measures and provided excellent reproducibility (%CV 0.9–3.4). The overall runtime of this method was 19 min, including column reconditioning. Using this method, the quantity of thiamine (67 ± 7 ng/g), riboflavin (423 ± 39 ng/g), nicotinamide (856 ± 77 ng/g) and pyridoxine (133 ± 11 ng/g) was determined from a complex food matrix.

In conclusion, we have developed a rapid and reliable, HILIC-single quad MS method utilising SIM for the low-level quantification of four B vitamins in a vegetable soup matrix in under 20 min. This method has shown excellent linearity, intra- and inter-day reproducibility and is directly applicable to other plant-based food matrices.

Introduction

Water-soluble vitamins and their derivatives are important dietary components essential for health due to their wide range of functions within central metabolism. The B group vitamins, for example, function to release energy from macronutrients, play crucial roles in redox reactions and act as enzyme co-factors [1]. The accurate quantification of water-soluble vitamins in foodstuffs is of significant value and importance. However, simultaneous determination and quantification of these vitamins from various food matrices represents a substantial challenge. In part, this is due to the diverse structural forms and different chemical compounds that make up these vitamins, many of which occur side by side in various foods. Naturally occurring vitamins often occur covalently bound to enzymes as cofactors, or as phosphorylated forms bound non-covalently to proteins and carbohydrates [2]. Thiamine (B₁) occurs in plant-derived foods chiefly in its free form but is present in animal foods as mono-, di-, or triphosphates and their esters, protein complexes, and as thiamine disulfides and their pyrophosphoric acid esters [1], [3]. This renders precise quantification of vitamin levels in food a difficult process, as conventional methods require the parent compound in its free form to compare against pure standards. Another limitation is related to the extraction process employed to release bound vitamins, as B group vitamins can be labile and degrade in the presence of light, heat and pH conditions [2], [4], [5], [6]. Due to the reasons presented above highlights the difficultly to develop one method to analyse all vitamins in a given sample and the reason several methodologies exist for each vitamin [7], [8], [9], [10].

Analysis of individual vitamins in foods were traditionally performed using microbial assays, however technological advances in liquid chromatography (LC) has provided methods offering increased accuracy and precision [11], [12]. Reversed-phase (RP) or hydrophilic interaction chromatography (HILIC) has been the separation method of choice for several years, but with different methods for detection and quantitation. For example, UV/vis detection [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], fluorescence detection [18], [24], [25], or electrochemical/coulometric [8], [26], [27], [28] have been employed, and we have previously reported a comparison of such methodologies [1]. However, the increased sensitivity offered by mass spectroscopy (MS) is of particular value for vitamin quantification when present in trace amounts and when complex food matrices are required for analysis. MS detection has been utilised for analysis of water-soluble vitamins in various food matrices including infant/nutritional formula, supplements, beverages and fortified produce [2], [10], [12], [21], [29], [30], [31], [32], [33], [34], [35]. Santos et al. (2012) [21] validated a LC-MS/MS RP method simultaneously analysing seven free form water soluble vitamins in 12 different green leafy vegetables. Gentili et al. (2008) [2] developed a LC/ESI-MS/MS RP method to determine B vitamins in four plant-based matrices. The authors utilised a solid–liquid extraction to determine the quantity of B vitamins, from which the vitamin content in maize flour (3.27–0.020 µg/g), green kiwi (2531–0.012 µg/g), golden kiwi (7758–0.19 µg/g) and tomato pulp (116.3–0.10 µg/g) were determined. However, the majority of these methods for quantifying B vitamins in food matrices utilise MS with RP separation, with limited papers utilising HILIC.

Chromatographic separation of water-soluble vitamins can be challenging using standard RP; they are highly polar compounds with different extents of hydrophilicity and due to this can be poorly retained using RP separation mechanisms [30], [31]. HILIC is an alternative to RP separation, receiving popularity due to it being best suited for analysis of polar analytes improving selectivity and compatible with MS detection [15]. Goldschmidt and Wolf et al. (2010) [30] study into different chromatographic methods involving RP and HILIC for analysis of B vitamins found RP to be problematic for retention of thiamine, nicotinamide and pyridoxine. Due to these vitamins being more hydrophilic they were more effectively retained using HILIC separation. However, as highlighted by Fatima et al. (2019) [36], a very limited number of validated methods utilising HILIC separation coupled to MS detection have been developed to analyse water soluble vitamins in complex food matrices. A study by Chatterjee et al. (2017) [37] developed a LC-MS/MS method targeting nine water-soluble vitamins in fish using HILIC separation for increased sensitivity. The authors optimised an enzymatic hydrolysis method to subsequently extract B vitamins; including nicotinamide, nicotinic acid, pyridoxine, pantothenic acid, biotin, thiamine, riboflavin and cyanocobalamin from fish flesh. However, this method primarily utilised the advantages offered by Multiple Reaction Monitoring (MRM) of a triple-quadrupole MS for identification and quantification of each analyte. This method may not translate to reliable quantification on a single-quadrupole MS without access to MRM. Single-quad MS instruments are widespread and comparatively cheaper than triple-quad MS systems; therefore, a method specifically developed for quantification on a single-quad MS may be more widely applicable for routine vitamin analysis.

Alongside, extraction procedures employed to extract vitamins from food matrices are often time consuming, involving many steps and chemical preparations. For example, Nurit et al. (2015) [12], employed enzymatic hydrolysis consisting of 6 different solutions and an 14 h incubation stage to extract vitamins. Therefore, in this study it was important to develop a simple analytical method suitable for high throughput analysis. We report here a method that combines a relatively simple extraction and combination of HILIC chromatography and single-quad MS detection that offers a practical and affordable approach to the quantification of B vitamins from food matrices.