Determination of adulterants in whey protein food supplements by liquid chromatography coupled to Orbitrap high resolution mass spectrometry

Determinação de adulterantes em proteína de soro de leite por cromatografia líquida acoplada à espectrometria de massas de alta resolução do tipo Orbitrap

Abstract

Liquid chromatography coupled to Orbitrap high resolution mass spectrometry was shown to be an adequate technique to control the adulteration of whey protein food supplements with prohibited substances, not declared on the labels. An extraction method combined with an instrumental analysis that allowed for the determination of 105 substances in whey protein food supplements, was established. The pre-treatment of the samples consisted of protein precipitation and solid-phase extraction using weak cation exchange functionalized polymeric sorbent cartridges. The samples were directly analyzed by LC-Orbitrap-HRMS. The selectivity, limit of detection, repeatability, recovery, carryover and matrix effect were estimated as the validation parameters. The repeatability obtained was 96.19% and the recovery 83.80%, but carryover and the matrix effect were not observed. The present method was successfully applied to the analysis of commercial samples, verifying adulteration by diuretics (conivaptan and politiazide) and a stimulant (benfluorex) in seven of the eleven brands evaluated.

Keywords:
Whey protein food supplement; Adulterants; Pharmacological action; Stimulants; Diuretics; Orbitrap; Method validation

Resumo

A cromatografia líquida acoplada à espectrometria de massas de alta resolução do tipo Orbitrap demonstrou ser uma técnica adequada para o controle de adulteração de substâncias proibidas não declaradas nos rótulos de suplementos proteicos derivados de soro de leite. Foi estabelecido um método de extração combinado a uma análise instrumental que permitiu a determinação de 105 substâncias em suplemento proteico derivado de soro de leite. O pré-tratamento das amostras consistiu na precipitação de proteínas e na extração em fase sólida, utilizando-se cartuchos com sorventes poliméricos baseados em troca catiônica. As amostras foram diretamente analisadas por CL-EMAR-Orbitrap. Foram estimados, como parâmetros de validação, seletividade, limite de detecção, repetitividade, recuperação, arraste e efeito de matriz. A repetitividade obtida foi de 96,19% e a recuperação foi de 83,80%. Arraste e efeito de matriz não foram observados. O presente método foi aplicado com sucesso na análise de amostras comerciais, nas quais foram verificadas adulterações, em sete das 11 marcas avaliadas, em diuréticos (conivaptan e politiazida) e estimulante (benfluorex).

Palavras-chave:
Suplemento alimentar de proteína de soro de leite; Adulterantes; Ação farmacológica; Estimulantes; Diuréticos; Orbitrap; Validação de método

1 Introduction

Over the years, technological and scientific advancement for improving human performance have been studied in different areas (Andrade et al., 2019; Roco & Bainbridge, 2002; Thomas et al., 2015). However, nutrition is still considered the most relevant aspect regarding muscle building, endurance and strength (Bagchi et al., 2013; McClung & Murray-Kolb, 2013; Pritchard-Peschek et al., 2013). Due to progress in this area, many athletes, non-athletes and patients with different diseases have been using functional foods to improve their health (Fayh et al., 2013; Horikawa et al., 2013; Mathews, 2018; Maughan et al., 2018; Rondanelli et al., 2016).

The class of food supplements most widely used in the world is that of milk constituents named Whey Protein Food Supplement (WPFS) (Chen et al., 2014; Fayh et al., 2013). The WPFS is obtained from the preparation of cheese, specifically during the casein precipitation step (milk protein) in which it forms a supernatant, the milk serum (Aquino et al., 2017; Chen et al., 2014; Garrido et al., 2016). WPFS shows the following properties: increase in resistance, muscle hypertrophy and decreased body fat (Andrade et al., 2019; Chen et al., 2014; Frestedt et al., 2008; Garrido et al., 2016).

There is no compatible regulation for WPFS between countries (Neves & Caldas, 2015), and the absence of a specific regulation and better monitoring in the manufacturing process of food supplements, may result in incompatibilities related to the label and content (Andrade et al., 2019; Parra et al., 2011). These mismatches could be related to the quantities of nutrients or other components described on the label or to the presence of substances (intentionally added) that are not reported, resulting in adulteration issues (Andrade et al., 2019; Marcus, 2016).

There are many studies on the adulteration of supplements, and more attention has been focused on adulteration by substances that have pharmacological properties (Garrido et al., 2016; Lu et al., 2010; Martínez-Sanz et al., 2017; Müller et al., 2018; Woo et al., 2013). The main cases of food supplement adulteration are related to the following classes: anabolic agents (provide increases in muscle mass and decreases in body fat); diuretics (decrease body liquids and mask the presence of other substances in the sample); and stimulants (weight loss, increase alertness and reduce fatigue) (Hernandez & Nahas, 2009; Müller et al., 2018; Neves & Caldas, 2015; Martínez-Sanz et al., 2017). Martello et al. (2007), described a qualitative liquid chromatography tandem mass spectrometry (LC-MS/MS) method used to detect the following anabolic androgenic steroids (4-androsten-3,17-dion, 4-oestren-3,17-dion, 5α-androsten-17β-ol-3-one, boldenone, nandrolone, nandrolone decanoate, testosterone and testosterone decanoate) and ephedrine in food supplements. The LC-MS/MS analysis was carried out using selected reaction monitoring (SRM) in an ion-trap system equipped with an atmospheric pressure chemical ionization (APCI) probe operating in the positive-ion mode. However, this is a target method for a limited number of substances. The method was applied to 64 nutritional supplements and a total of 12.5% of the nutritional supplements analyzed contained banned substances not declared on the label (anabolic steroids and ephedrine) (Martello et al., 2007). However, some relevant classes of substances such as diuretics and anorectic agents were not evaluated, probably because analysis by SRM only in the positive ionization mode does not allow for the scanning of a comprehensive number of substances. Moreover, for the LC-MS/MS analysis, the liquid-liquid extraction sample preparation using n-pentane and diethyl ether, limited the extraction of acid analytes. In 2010, Lu et al., described a sensitive and specific liquid chromatography-electrospray ionization mass spectrometry (LC/ESI-MS) method for the analysis of 18 drugs used in the treatment of hypertension, including diuretics, as adulterants in dietary supplements (Lu et al., 2010). However, once again it was a very limited procedure regarding the number of substances analyzed.

Multi-target procedures need a more elaborate analytical method, and usually include an ESI interface operating in both positive and negative ionization modes. Moreover, extraction and matrix effects are also a vulnerable point for the routine inspection of WPFS by a single and comprehensive approach. Although, MS/MS experiments allow for the enhancement of sensitivity by applying the selected reaction monitoring (SRM) mode for the determination of selected compounds, the high-resolution mass spectrometry (HRMS) approach enables the specific identification of analytes from the full scan data, making every measurement accessible to subsequent analysis and the search for new, previously not encountered compounds.

Therefore, to determine the presence of adulterants in WPFS, a liquid chromatography method coupled to Orbitrap high resolution mass spectrometry (LC-Orbitrap-HRMS) after solid phase extraction, was optimized and validated to detect the following different classes of substances: anabolic agents, beta-agonists, hormone and metabolic modulators, diuretics and stimulants.

2 Experimental

2.1 Quality assurance

All analytical and managerial procedures were carried out in an ISO/IEC 17025 standard environment, accredited by the Brazilian National Metrological Institute (BNMI - INMETRO) (Associação Brasileira de Normas Técnicas, 2005).

2.2 Chemicals and materials

All solvents used were HPLC grade: methanol, formic acid, ammonium formate and acetic acid (Tedia; Fairfield, USA), and the distilled water was purified by the milli-Q purification system (Millipore, Massachusetts, USA). Reference compounds were purchased mainly from NMI (Sydney, Australia), Sigma-Aldrich (St. Louis, USA) and Logical (Luckenwalde, Germany) or were kindly donated by other anti-doping laboratories. The Strata-X-CW, weak cation mixed mode polymeric sorbent (30 mg, 3 mL) SPE cartridges (São Paulo, Brazil) were purchased from Phenomenex. The internal standards mefruside, methyltestosterone, n-methylhexanamine, buspirone hydrochloride, 4-methylefedrine-D3 HCl, and 7-propyltheophylline were purchased from Sigma-Aldrich (St. Louis, USA).

2.3 Standard solutions

The fortification solution containing the reference standards consisted of a mixture of the following classes of substances: anabolic agents, beta-agonists, hormone and metabolic modulators, diuretics and masking agents and stimulants, at different concentrations (2 to 8 ng µL^-1), prepared in methanol.

The internal standard (IS) solution consisted of a mixture of the substances mefruside, methyltestosterone, n-methylhexanamine, propyl-theophylline and 7-propyl-theophylline at different concentrations (6 to 20 ng µL^-1), all dissolved in methanol.

2.4 Sample preparation

Eleven brands of WPFS were purchased from local Brazilian markets and 20 mg of each sample dissolved in 2 mL of water. Three controls were prepared: reagent blank (water), negative control (matrix without the analytes) and positive control (matrix spiked with 50 µL of fortification solution). The samples and controls were homogenized during 20 s and centrifuged for 20 min at 1.5 x G. One aliquot of the 10 µL in each test tube was transferred to a conical Eppendorf tube and 50 µL of the 2% (v/v) acetic acid solution added. The tubes were centrifuged for 5 min at 1.5 x G and stored at 2 °C to 6 °C for the subsequent reconstitution of the extract. The second aliquot taken from the supernatant obtained from each initial sample was transferred to a new test tube, 20 µL of the internal standard solution added and the contents homogenized by vortex for 20 s. In sequence, the solid-phase extraction (SPE) step was carried out. The SPE cartridges were conditioned with 1 mL of methanol and 1 mL of Milli-Q water; the sample was applied and the cartridges washed with 1 mL of Milli-Q water and 1 mL of 50% (v/v) methanol in water. The analytes were eluted with 1 mL of 5% (v/v) formic acid in methanol. All samples were evaporated under a nitrogen flow at 45 °C. The first aliquot was added to the dried residue and the mixture homogenized by vortex for 20 s, transferred to vials with the inserts and refrigerated at 4 °C for 4 h. The supernatants were transferred to new vials with inserts and injected into the chromatographic system.

2.5 Instrumentation

The liquid chromatography system was an Accela LC liquid chromatography (Thermo Scientific, Bremen, Germany), with an Accela 1250 pump and auto sampler fixed at 10 °C. The column was a Zorbax SB-C₁₈ one, 3.0 mm × 50 mm, 1.8 µm (Agilent, Böblingen, Germany). The mobile phases were 0.1% ammonium formate/0.1% formic acid in water (A) and 0.1% formic acid in methanol (B). The gradient program was as follows: 95% A for 0.5 min, then decreasing linearly to 90% at 10.0 min, then to 0% at 11.1 min, followed by an increase to the initial concentration of 95% A at 14.0 min. The total run time was thus 14 min. The column was maintained at 40 °C, the flow rate constant at 600 µL min^-1 and the injection volume was 0.5 µL.

The LC effluent was pumped to a Q Exactive Orbitrap-based high resolution mass spectrometer (Thermo Scientific, Bremen Germany) operating in the positive-negative polarity switching mode and equipped with an electrospray ionization (ESI) source. The nitrogen gas flow and auxiliary gas were set to 60 and 20 (arbitrary units), respectively. The capillary temperature was 380 °C, the spray voltage 3900 kV - 2900 kV and the capillary voltage 3.9 V or -2.9 V, in the positive or negative modes, respectively. The instrument was operated in the full scan mode from m/z 100 to 620 and from 70 to 630, in the positive and negative modes, respectively, at 70,000 resolution power. The automatic gain control (AGC) was 10⁶. The performance of the Orbitrap in both the positive and negative ionization modes was evaluated daily.

The data obtained after the LC-Orbitrap-HRMS analyses were processed using the Qual Browser (Thermo Electron, San Jose, CA), applying a 5 ppm tolerance error and FULL-MS acquisition. In addition, the formula calculator used included carbon, hydrogen and oxygen atoms to provide chemical formula and saturation values (ring double-bonds equivalent - RDBE) for the precursor ions [M+H]⁺ and [M+H]^-. A comparison between the theoretical and experimental precursor molecular mass values was evaluated in the identification of molecule structures. The instrumental conditions used were established according to anti-doping control (Sardela et al., 2018).

2.6 Method validation

The method was validated by qualitative screening, according to the internal validation protocol of the Laboratório Brasileiro de Controle de Dopagem, based on the DOC-CGCRE 008 protocol of the Instituto Nacional de Metrologia, Qualidade e Tecnologia. The selectivity evaluated ten different WPFS samples, declared as negative, and verified the absence of interfering substances in the retention times of each analyte. The limit of detection was evaluated by preparing 10 different samples spiked with 10% of the usual concentration and 10 different samples spiked with 50% of the usual concentration. The repeatability was determined using 7 replicates of WPFS, and the peak area values were expressed according to the relative standard deviation of each substance monitored. The recovery was evaluated by preparing 7 replicates spiked before the solid phase extraction step (repeatability) and 7 replicates spiked after the solid phase extraction step. The carryover was verified by an analysis of the spiked WPFS sample with twice the fortification solution concentration between 2 blank WPFS samples. The matrix effect evaluated 10 different WPFS samples, declared negative, which were spiked with the monitored analytes.

2.7. Application to real samples

The optimized and validated method by LC-HRMS was applied to the analysis of eleven commercial samples of WPFS products.

3 Results and discussion

3.1 Sample preparation

The WPFS samples contain a high concentration of the proteins β-lactoglobulin, α-lactalbumin, serum albumin, immunoglobulins and glycomacropeptides, as well as other minor proteins such as lactoperoxidase, lactoferrin, β-microglobulin, lysozyme, insulin-like growth factor and others (Haraguchi et al., 2006). These proteins should be removed before injecting the sample into the chromatographic system. The presence of those compounds may clog the SPE column or the analytical column, due to precipitation, reducing the life of the column and competing for the electrospray ionization source, thus interfering in substance detection. Hence, it is indispensable to carry out sample pre-treatment procedures to remove these proteins, without loss of the target analytes. Thus a solvent precipitation step before the solid phase extraction was tested with different solvents, in order to reduce the amount of protein in the matrix, but not remove the target substances. Cold water, acetonitrile and a combination of acetonitrile/water (50% - v/v) were tested and the final recoveries of the substances compared based on their peak areas.

Acetonitrile and acetonitrile/water were not effective, because they also removed some target substances, especially the diuretics. Only the use of the cold water dissolved the WPFS and, at the same time, eliminated the excess of proteins and allowed for the detection of all the target substances with good recovery.

3.2 Method validation

Selectivity verifies the absence of interference in the retention times of the monitored substances. No significant interference in the retention times was observed for the target substances.

Table 1 summarizes the results observed for all the substances validated. The detection limit (LOD) represents the lowest concentration of the substance that is detectable but not necessarily quantified using an experimental procedure. The LOD for anabolic agents was 1.25 ng g^-1, beta-agonists from 5 to 10 ng g^-1, hormone and metabolic modulators from 5 to 12.5 ng g^-1, diuretics from 6.25 to 62.5 ng g^-1, and stimulants and anorectic agents from 3.16 to 25 ng g^-1. All compounds were detected with more than 8 point-acquisitions at the LOD. This low limit of detection is associated with the screening method sensitivity and it is important for the screening of adulterants in commercial WPFS samples.

Table 1
Chemical formula, polarity, retention times (t_R), theoretical masses (m/z), sample concentration (SC), repeatability, recovery, limit of detection (LOD) and matrix effect of the monitored compounds.

Compound	Chemical Formula	Polarity	tR (min)	m/z	SC	Repeatability	Extraction	LOD	Matrix
					(ng g-1)	(%)	yield (%)	(ng.g-1)	interference (%)
Andarine	C19H18F3N3O6	-	6.07	440.10749	12.5	14.34	99.0	1.25	0.30
Gestrinone	C21H24O2	+	7.00	309.18491	12.5	18.00	96.8	1.25	0.15
Methyldienolone	C19H26O2	+	6.96	287.20056	12.5	10.06	95.8	1.25	0.11
Methyltrienolone	C19H24O2	+	6.90	285.18491	12.5	6.87	95.1	1.25	0.14
Ostarine	C19H14F3N3O3	-	6.46	388.09145	12.5	21.58	96.2	1.25	0.12
Oxandrolone	C19H30O3	+	6.60	307.22677	12.5	8.26	95.7	1.25	0.13
Tetrahydrogestrinone	C21H28O2	+	7.87	313.21621	12.5	18.49	96.3	1.25	0.14
Bamethan	C12H19NO2	+	1.80	210.14886	50.0	1.94	91.6	5.00	1.17
Formoterol	C19H24N2O4	+	3.29	345.18088	100.0	5.38	92.0	10.00	1.12
Isoxsuprine	C18H23NO3	+	3.71	302.17507	50.0	5.69	94.1	5.00	0.90
Metaproterenol	C11H17NO3	+	0.49	212.12812	50.0	15.86	96.7	5.00	2.63
Procaterol	C16H22N2O3	+	1.96	291.17032	50.0	8.86	95.2	5.00	1.17
Ritodrine	C17H21NO3	+	2.02	288.15942	50.0	5.53	97.8	5.00	1.24
Salmeterol	C25H37NO4	+	6.33	416.27954	50.0	10.16	90.0	5.00	0.12
Aminoglutethimide	C10H12N	+	2.31	146.09642	50.0	10.80	92.8	5.00	0.94
Anastrazole	C17H19N5	+	4.80	294.17132	50.0	7.44	92.8	5.00	0.14
Androstatrienedione	C19H24O2	+	6.20	283.16926	50.0	9.49	95.7	5.00	0.13
Exemestane	C20H24O2	+	6.77	297.18490	50.0	13.45	98.5	5.00	0.12
Flutamide	C11H11F3N2O3	-	6.39	275.06381	50.0	13.42	97.4	5.00	0.29
Fulvestrant	C32H47F5O3S	+	8.79	607.32388	50.0	32.86	75.4	5.00	0.08
Gw501516	C21H18F3NO3S2	+	8.70	454.07530	50.0	37.94	75.3	5.00	0.10
Raloxifene	C28H27NO4S	+	5.00	474.17336	50.0	17.68	89.6	5.00	0.43
Bendroflumethiazide	C15H14F3N3O4S2	-	4.85	420.03051	125.0	10.21	92.7	12.50	0.26
Benzbromarone	C17H12Br2O3	+	8.48	422.92260	125.0	25.83	77.4	62.50	0.14
Benzthiazide	C15H14CIN3O4S3	-	4.75	429.97622	62.5	10.73	94.9	6.25	0.11
Bumetanide	C17H20N2O5S	-	6.26	363.10202	125.0	34.93	29.5	62.50	0.13
Chlorothiazide	C7H8CIN3O4S2	-	0.95	293.94155	125.0	22.76	5.6	12.50	1.21
Chlorthalidone	C14H11CIN2O4S	-	3.61	337.00553	125.0	13.94	9.8	62.50	0.34
Clopamide	C14H20ClN3O3S	-	4.10	344.08411	62.5	12.04	40.8	6.25	0.10
Conivaptan	C32H26N4O2	+	5.84	499.21285	125.0	12.49	88.7	12.50	0.21
Cyclopenthiazide	C13H18ClN3O4S2	-	5.30	378.03545	125.0	11.98	97.1	12.50	0.13
Cyclothiazide	C4H16ClN3O4S2	-	4.88	388.01979	125.0	11.34	95.4	12.50	0.25
Diclofenamide	C6H6Cl2N2O4S2	-	2.52	302.90733	125.0	16.94	6.8	12.50	0.51
Etacrynic acid	C13H12Cl2O4	-	6.47	301.00399	125.0	35.81	15.3	12.50	0.38
Hydrochlorothiazide	C7H8CIN3O4S2	-	1.18	295.95720	250.0	26.58	5.3	25.00	0.00
Hydroflumethiazide	C8H8F3N3O4S2	-	1.75	329.98356	62.5	24.89	4.1	6.25	0.24
Lixivaptan	C27H21CIFN3O2	+	7.52	474.13791	125.0	38.07	81.4	62.50	0.11
Methazolamide	C5H8N4O3S2	-	2.12	234.99650	125.0	31.49	4.4	12.50	0.24
Methyclothiazide	C9H11Cl2N3O4S2	+	3.37	359.96408	500.0	5.36	117.7	250.00	0.38
Piretanide	C17H18N2O5S	-	5.84	361.08637	125.0	30.09	12.4	62.50	0.31
Polythiazide	C11H13ClF3N3O4S3	-	4.81	437.96360	62.5	8.62	95.9	6.25	0.14
Probenecid	C13H19NO4S	-	6.24	284.09620	62.5	23.93	27.1	6.25	0.13
Spironolactone	C24H32O4S	+	6.80	341.21112	62.5	13.94	99.3	6.25	0.17
Torasemide	C16H20N4O3S	-	4.77	347.11833	62.5	8.25	89.4	6.25	0.23
Triamterene	C12H11N7	+	2.96	254.11487	62.5	5.24	94.2	6.25	1.49
Trichlormethiazide	C8H8Cl3N3O4S2	-	3.06	377.89490	125.0	16.54	21.5	12.50	0.44
Xipamide	C15H15CIN2O4S	-	5.61	353.03683	62.5	19.16	24.6	6.25	0.15
(s)-2-aminooctane	C8H19N	+	3.90	130.15902	125.0	2.94	91.5	12.50	1.19
3,3-diphenylpropylamine	C15H17N	+	4.68	212.14338	125.0	3.53	90.1	12.50	0.69
4-fluoroamphetamine	C9H12FN	+	2.00	154.10265	125.0	4.50	89.1	12.50	1.10
Amiphenazole	C9H9N3S	+	1.59	192.05899	125.0	14.42	101.8	62.50	1.37
Benfluorex	C19H20F3NO2	+	5.80	352.15189	125.0	9.96	86.5	12.50	0.12
Benzphetamine	C17H21N	+	4.18	240.17468	125.0	4.24	91.2	12.50	0.83
Benzylpiperazine	C11H16N2	+	0.90	177.13863	31.25	11.67	87.9	3.16	5.66
Carphedon	C12H14N2O2	+	3.18	219.11280	125.0	12.71	36.1	12.50	0.13
Cathine	C9H13NO	+	1.36	134.09640	250.0	5.69	88.3	25.00	1.70
Chlorphentermine	C10H14ClN	+	3.62	184.08875	125.0	3.76	91.3	12.50	1.45
Clobenxorex	C16H18ClN	+	4.93	260.12005	62.5	4.55	92.1	6.25	0.57
Cocaine	C17H21NO4	+	3.24	304.15433	125.0	4.45	96.2	12.50	1.26
Cropropamide	C13H24N2O2	+	5.10	241.19105	125.0	5.10	78.0	12.50	0.13
Crotethamide	C12H22N2O2	+	4.25	227.17540	125.0	15.34	70.6	12.50	2.00
Cyclazodone	C12H12N2O2	+	4.04	217.09715	125.0	7.37	87.3	12.50	0.25
Dobutamine	C18H23NO3	+	2.73	302.17507	125.0	7.38	95.6	12.50	1.56
Etamivan	C12H17NO3	+	4.16	224.12812	62.5	11.03	52.6	6.25	0.20
Etilefrine	C10H15NO2	+	0.55	182.11756	125.0	7.64	96.6	12.50	3.12
Famprofazone	C24H31N3O	+	6.40	378.25399	62.5	6.85	88.5	6.25	0.28
Fenbrutazate	C23H29NO3	+	6.30	368.22202	62.5	8.59	87.8	6.25	0.24
Fencamine	C20H28N6O2	+	3.26	385.23465	62.5	7.79	92.9	6.25	0.89
Fenethyline	C18H23N5O2	+	3.53	342.19245	125.0	3.43	95.3	12.50	0.82
Fenfluramine	C12H16F3N	+	3.93	232.13076	125.0	3.25	90.4	12.50	1.19
Fenproporex	C12H16N2	+	1.92	189.13863	62.5	1.72	89.8	6.25	1.13
Flephedrone	C10H12FNO	+	4.33	182.09757	125.0	23.19	108.7	12.50	1.58
Furfenorex	C15H19NO	+	3.46	230.15394	125.0	4.24	88.8	12.50	1.17
Heptaminol	C8H19NO	+	1.30	146.15394	250.0	6.18	87.6	25.00	2.00
Isometheptene	C24H48N2O8	+	3.02	142.15903	125.0	2.95	90.4	12.50	1.55
Mefenorex	C12H18CIN	+	3.37	212.12005	62.5	2.62	90.3	6.25	1.40
Mephedrone	C11H15NO	+	2.50	178.12264	125.0	3.24	92.9	12.50	1.46
Mesocarb	C18H18N4O2	+	6.20	323.15025	125.0	14.89	93.0	12.50	0.13
Methoxyphenamine	C11H17NO	+	2.77	180.13829	62.5	3.15	92.6	6.25	1.48
Methylenodioxyamfetamine	C10H13NO2	+	2.09	163.07540	125.0	5.92	38.8	12.50	1.23
Methylenodioxymethamfetamine	C11H15NO2	+	2.14	194.11756	62.5	4.69	91.1	6.25	1.62
Methylenodioxy-n-ethylamfetamine	C12H17NO2	+	2.46	208.13321	125.0	2.87	92.9	12.50	1.47
Methylephedrine	C11H17NO	+	1.61	180.13829	125.0	3.68	94.4	12.50	1.30
Methylphenidate	C14H19NO2	+	3.28	234.14885	125.0	3.62	92.7	12.50	1.12
Mitragyne	C23H30N2O4	+	4.75	399.22783	125.0	6.59	91.7	12.50	0.41
Modafinil	C15H15NO2S	+	4.85	296.07157	125.0	4.87	92.9	12.50	0.14
Nikethamine	C10H14N2O	+	2.84	179.11789	62.5	15.14	8.3	6.25	0.00
Norfenfluramine	C10H12F3N	+	3.60	204.09946	125.0	3.05	91.5	12.50	1.47
Octhylamine	C8H18N	+	4.35	130.15903	125.0	2.91	88.8	12.50	0.98
Oxilofrine	C10H15NO2	+	0.39	133.06479	125.0	7.63	94.8	12.50	3.80
Pemoline	C9H8N2O	+	2.09	177.06585	125.0	32.90	4.7	12.50	0.25
Pentetrazol	C6H10N4	+	2.03	139.09782	125.0	23.38	6.2	12.50	0.25
Phendimetrazine	C12H17NO	+	1.88	192.13829	125.0	2.68	91.3	12.50	3.65
Phenmetrazinha	C11H16CINO	+	1.92	178.12264	125.0	3.20	91.4	12.50	1.73
Pholedrine	C10H15NO	+	0.89	166.12264	125.0	9.77	94.6	12.50	4.30
p-hydroxy amphetamine	C9H13NO	+	0.85	135.08044	125.0	8.73	99.4	12.50	4.30
Pipradol	C18H21NO	+	4.07	268.16959	125.0	3.06	93.5	12.50	1.08
Prenilamyne	C24H27N	+	6.52	330.22163	125.0	11.48	90.2	12.50	0.12
Prolintane	C15H23N	+	3.82	218.19033	125.0	2.18	91.5	12.50	0.98
Propylhexedrine	C10H21N	+	3.79	156.17468	125.0	2.69	93.0	12.50	1.21
s(+)-methamphetamine	C10H15N	+	2.05	150.12773	62.5	3.90	91.5	6.25	1.42
Selegine	C13H17N	+	2.54	188.14338	125.0	5.28	88.2	12.50	1.24
Sibutramine	C17H26CIN	+	6.06	280.18265	125.0	7.17	86.4	12.50	0.13
Strychnine	C21H22N2O2	+	2.53	335.17540	125.0	3.97	90.4	12.50	1.29
Trimetazidine	C14H22N2O3	+	1.86	267.17032	125.0	7.17	82.2	12.50	1.20

The repeatability was verified using the peak areas of each analyte from 7 WPFS replicates and Table 1 shows this variation as the relative standard deviation (RSD). The RSD values obtained were compared with the maximum relative standard deviation for each level of concentration calculated by the Horwitz equation (Horwitz et al., 1980). According to the comparison between the RSD and the values obtained using the Horwitz equation, the majority of the substances monitored (96.19%) showed an adequate relative standard deviation, indicating a low result dispersion. Only ethacrynic acid, benzbromarone, bumetanide and fulvestrant showed RSD values above the values preconized by the Horwitz equation.

The recovery can also be observed in Table 1 and it was lower than 50% for 15% of the targeted substances including ethacrynic acid, benzbromarone, bumetanide and fulvestrant again, and also clopamide, chlorothiazide, chlorthalidone, diclofenamide, hydrochlorothiazide, hydroflumethiazide, methazolamide, piretanide, probenecid, carphedon, methylenedioxyamfetamine, nikethamide, pemoline, pentetrazol, trichlormethiazide and xipamide. Almost all of these substances are acids. However, despite the recovery below 50%, all the substances were properly detected in all the replicates, mainly because by HRMS, low noise and clear signals were obtained, allowing the presence of the substances to be detected.

As a consequence of this comprehensive sample preparation procedure, a high influence of the matrix in the instrumental conditions was expected. When complex matrices such as WPFS are analyzed, signal suppression of an analyte can occur and/or a shift in the retention time can be observed, probably because of the punctual modification of the stationary phase or due to overlay of the WPFS matrix components. The ion suppression was evaluated in the matrix effect experiments and the variation in the retention time (t_R) was evaluated by monitoring the t_R peak for all targeted substances for their respective retention times in a window of 1 minute. After injecting 10 different replicates, the highest RSD observed was 2% for the t_R of the substances that elute before 1 minute (Table 1).

Finally, the existence of carryover was tested, but none was observed. Carryover verifies the existence of significant variations amongst sequential injections.

3.3 Application to the commercial samples

The eleven brands of WPFS (identified by the numbers 1 to 11) were analyzed and compared with the positive control to check for the presence of adulterants. According to the LC-Orbitrap-HRMS analysis, peaks with retention times and m/z ratios equal or similar (error below 5 ppm) to the substances conivaptan, polythiazide and benfluorex, were found in some commercial samples. The samples showing a suspicion of adulteration were extracted and analysed twice more to confirm the presence of the adulterants.

Suspicious samples were confirmed based mainly on the parameters of the mass/charge ratios of the precursor ions, mass accuracy calculation, the RDBE values of the samples, and comparison with the positive control. Variations in the t_R and m/z ratios of the positive control and the commercial samples were observed in the suspect peaks. Subsequently the mass accuracy was calculated (maximum limit of 5 ppm for confirmation of the identities of the compounds) and the RDBE values. Figure 1 shows the retention times (t_R, min), molecular ions, calculation of mass accuracy, and the RDBE values relevant to the presence of the following diuretics: conivaptan, polythiazide and/or the stimulant: benfluorex.

Figure 1
Chemical structures, m/z, ppm error, retention time and RDBE of conivaptan, polythiazide and benfluorex in the reference material and in the samples.

According to the comparison of the parameters, the values obtained for the samples suspected of adulteration were compatible with those of the positive control, confirming the presence of adulterants in those whey protein food supplements. After applying the method, seven of the eleven brands analysed (63.64%) showed adulteration by at least one of the above-mentioned substances.

The administration of these compounds can cause health risks (depending on the associated factors) and another aggravating fact is that many individuals are consuming products classified as foods without knowing that they may contain substances with pharmacological properties. The adulteration by diuretics and/or anorectic stimulants is related to the effects they may cause, and an effective weight loss is amongst the effects common to these classes. Conivaptan and polythiazide act by increasing diuresis, masking the other substances present, and benfluorex is a stimulant with an anorectic effect that induces a loss of appetite (Docherty, 2008; Woo et al., 2013).

4 Conclusions

Whey protein food supplement samples contain a high concentration of proteins which can be removed by solvent precipitation and solid-phase extraction clean-up before sample injection into the chromatographic system. Cold water was the best solvent option to remove these proteins and maintain the procedure comprehensive.

The LC-Orbitrap-HRMS method allowed for better separation, detection and identification of the analytes, due to the high sensitivity and resolution of the mass analyser.

The parameters selectivity, LOD, repeatability, extraction yield, carryover and matrix effect were duly validated. All these parameters showed satisfactory results in the detection of the substances with pharmacological action in the WPFS matrix.

After applying the method, it was shown that four commercial samples were not adulterated and seven were. One showed the presence of both conivaptan and politiazide (a combination of diuretic agents), two showed the presence of politiazide and benfluorex, and the others only showed the presence of either politiazide or benfluorex.

Acknowledgements

This work was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Laboratório Brasileiro de Controle de Dopagem (LBCD) and by the Instituto Militar de Engenharia (IME).

• Cite as: Roiffé, R. R., Sardela, V. F., Lima, A. L. S., Oliveira, D. S., Aquino Neto, F. R., Lima, K. S. C., & de la Cruz, M. N. S. (2019). Determination of adulterants in whey protein food supplements by liquid chromatography coupled to Orbitrap high resolution mass spectrometry. Brazilian Journal of Food Technology, 22, e2018206. https://doi.org/10.1590/1981-6723.20618

References

• Andrade, J., Pereira, C. G., Almeida Junior, J. C., Viana, C. C. R., Neves, L. N. O., Silva, P. H. F., Bella, M. J. V., & Anjos, V. C. (2019). FTIR-ATR determination of protein content to evaluate whey protein concentrate adulteration. Food Science and Technology, 99, 166-172.
• Aquino, L. F. M. C., Ribeiro, R. O. R., Simoes, J. S., Mano, S. B., Mársico, E. T., & Conte Junior, C. A. (2017). Mercury content in whey protein and potential risk for human health. Journal of Food Composition and Analysis, 59, 141-144. http://dx.doi.org/10.1016/j.jfca.2017.02.014
  » http://dx.doi.org/10.1016/j.jfca.2017.02.014
• Associação Brasileira de Normas Técnicas – ABNT. (2005). NBR ISO/IEC 17025: General requirements for the competence of testing and calibration laboratories Rio de Janeiro: ABNT.
• Bagchi, D., Nair, S., & Sen, C. K. (2013). Nutrition and enhanced sports performance: Muscle building, endurance, and strength (1st ed). United States: Academic Press.
• Chen, W., Huang, W., Chiu, C., Chang, W., & Huang, C. (2014). Whey protein improves exercise performance and biochemical profiles in trained mice. Medicine and Science in Sports and Exercise, 46(8), 1517-1524. PMid:24504433. http://dx.doi.org/10.1249/MSS.0000000000000272
  » http://dx.doi.org/10.1249/MSS.0000000000000272
• Docherty, J. R. (2008). Pharmacology of stimulants prohibited by the World Anti-Doping Agency (WADA). British Journal of Pharmacology, 154(3), 606-622. PMid:18500382. http://dx.doi.org/10.1038/bjp.2008.124
  » http://dx.doi.org/10.1038/bjp.2008.124
• Fayh, A. P. T., Silva, C. V., Jesus, F. R. D., & Costa, G. K. (2013). Consumo de suplementos nutricionais por freqüentadores de academias da cidade de Porto Alegre. Revista Brasileira de Ciências do Esporte, 35(1), 27-37. http://dx.doi.org/10.1590/S0101-32892013000100004
  » http://dx.doi.org/10.1590/S0101-32892013000100004
• Frestedt, J. L., Zenk, J. L., Kuskowski, M. A., Ward, L. S., & Bastian, E. D. (2008). A whey-protein supplement increases fat loss and spares lean muscle in obese subjects: A randomized human clinical study. Nutrition & Metabolism, 5(8), 1-7. PMid:18371214. http://dx.doi.org/10.1186/1743-7075-5-8
  » http://dx.doi.org/10.1186/1743-7075-5-8
• Garrido, B. C., Souza, G. H. M. F., Lourenço, D. C., & Fasciotti, M. (2016). Proteomics in quality control: Whey protein-based supplements. Journal of Proteomics, 147, 48-55. PMid:27072112. http://dx.doi.org/10.1016/j.jprot.2016.03.044
  » http://dx.doi.org/10.1016/j.jprot.2016.03.044
• Haraguchi, F. K., Abreu, W. C., & Paula, H. (2006). Proteínas do soro do leite: Composição, propriedades nutricionais, aplicações no esporte e benefícios para a saúde humana. Revista de Nutrição, 19(4), 479-488. http://dx.doi.org/10.1590/S1415-52732006000400007
  » http://dx.doi.org/10.1590/S1415-52732006000400007
• Hernandez, A. J., & Nahas, R. M. (2009). Modificações dietéticas, reposição hídrica, suplementos alimentares e drogas: Comprovação de ação ergogênica e potenciais riscos para a saúde. Revista Brasileira de Medicina do Esporte, 15(3), 3-12.
• Horikawa, Y., Tsutsumi, R., & Tsutsumi, Y. (2013). Whey protein diets can limit inflammation and oxidative stress in the critically ill. Critical Care Medicine, 41(12), A181-A182. http://dx.doi.org/10.1097/01.ccm.0000439972.21054.05
  » http://dx.doi.org/10.1097/01.ccm.0000439972.21054.05
• Horwitz, W., Kamps, L. R., & Boyer, R. W. (1980). Quality assurance in the analysis of foods and trace constituents. Journal - Association of Official Analytical Chemists, 63(6), 1344-1354. PMid:7451398.
• Lu, Y. L., Zhou, N. L., Liao, S. Y., Su, N., He, D. X., Tian, Q. Q., Chen, B., & Yao, S. Z. (2010). Detection of adulteration of anti-hypertension dietary supplements and traditional Chinese medicines with synthetic drugs using LC/MS. Food Additive and Contaminants: Part A, 27(7), 893-902. PMid:20544454. http://dx.doi.org/10.1080/19440040903426710
  » http://dx.doi.org/10.1080/19440040903426710
• Marcus, D. M. (2016). Dietary supplements: What’s in a name? What’s in the bottle? Drug Testing and Analysis, 8(3-4), 410-412. PMid:27072845. http://dx.doi.org/10.1002/dta.1855
  » http://dx.doi.org/10.1002/dta.1855
• Martello, S., Felli, M., & Chiarotti, M. (2007). Survey of nutritional supplements for selected illegal anabolic steroids and ephedrine using LC-MS/MS and GC-MS methods, respectively. Food Additives and Contaminants, 24(3), 258-265. PMid:17364927. http://dx.doi.org/10.1080/02652030601013729
  » http://dx.doi.org/10.1080/02652030601013729
• Martínez-Sanz, J. M., Sospedra, I., Ortiz, C. M., Baladía, E., Gil-Izquierdo, A., & Ortiz-Moncada, R. (2017). Intended or unintended doping? A review of the presence of doping substances in dietary supplements used in sports. Nutrients, 9(10), 1093. PMid:28976928. http://dx.doi.org/10.3390/nu9101093
  » http://dx.doi.org/10.3390/nu9101093
• Mathews, N. M. (2018). Prohibited contaminants in dietary supplements. Sports Health, 10(1), 19-30. PMid:28850291. http://dx.doi.org/10.1177/1941738117727736
  » http://dx.doi.org/10.1177/1941738117727736
• Maughan, R. J., Burke, L. M., Dvorak, J., Larson-Meyer, D. E., Peeling, P., Phillips, S. M., Rawson, E. S., Walsh, N. P., Garthe, I., Geyer, H., Meeusen, R., van Loon, L., Shirreffs, S. M., Spriet, L. L., Stuart, M., Vernec, A., Currell, K., Ali, V. M., Budgett, R. G. M., Ljungqvist, A., Mountjoy, M., Pitsiladis, Y., Soligard, T., Erdener, U., & Engebretsen, L. (2018). IOC consensus statement: Dietary supplements and the high-performance athlete. International Journal of Sport Nutrition and Exercise Metabolism, 28(2), 104-125. PMid:29589768. http://dx.doi.org/10.1123/ijsnem.2018-0020
  » http://dx.doi.org/10.1123/ijsnem.2018-0020
• McClung, J. P., & Murray-Kolb, L. E. (2013). Iron nutrition and premenopausal women: Effects of poor iron status on physical and neuropsychological performance. Annual Review of Nutrition, 33(1), 271-288. PMid:23642204. http://dx.doi.org/10.1146/annurev-nutr-071812-161205
  » http://dx.doi.org/10.1146/annurev-nutr-071812-161205
• Müller, L. S., Muratt, D. T., Molin, T. R. D., Urquhart, C. G., Viana, C., & Carvalho, L. M. (2018). Analysis of pharmacologic adulteration in dietary supplements by capillary zone electrophoresis using simultaneous contactless conductivity and UV detection. Chromatographia, 81(4), 689-698. http://dx.doi.org/10.1007/s10337-018-3496-2
  » http://dx.doi.org/10.1007/s10337-018-3496-2
• Neves, D. B. J., & Caldas, E. D. (2015). Dietary supplements: International legal framework and adulteration profiles, and characteristics of products on the Brazilian clandestine market. Regulatory Toxicology and Pharmacology, 73(1), 93-104. PMid:26107294. http://dx.doi.org/10.1016/j.yrtph.2015.06.013
  » http://dx.doi.org/10.1016/j.yrtph.2015.06.013
• Parra, R. M. T., Palma, A., & Pierucci, A. P. T. R. (2011). Contaminação de suplementos dietéticos usados para prática esportiva. Revista Brasileira de Ciências do Esporte, 33(4), 1071-1084.
• Pritchard-Peschek, K. R., Osborne, M. A., Slater, G. J., Taaffe, D. R., & Jenkins, D. G. (2013). Pseudoephedrine and preexercise feeding: Influence on performance. Medicine and Science in Sports and Exercise, 45(6), 1152-1157. PMid:23274597. http://dx.doi.org/10.1249/MSS.0b013e3182808e23
  » http://dx.doi.org/10.1249/MSS.0b013e3182808e23
• Roco, M. C., & Bainbridge, W. S. (2002). Converging technologies for improving human performance. Journal of Nanoparticle Research, 4(4), 281-295. http://dx.doi.org/10.1023/A:1021152023349
  » http://dx.doi.org/10.1023/A:1021152023349
• Rondanelli, M., Klersy, C., Terracol, G., Talluri, J., Maugeri, R., Guido, D., Faliva, M. A., Solerte, B. S., Fioravanti, M., Lukaski, H., & Perna, S. (2016). Whey protein, amino acids, and vitamin D supplementation with physical activity increases fat-free mass and strength, functionality, and quality of life and decreases inflammation in sarcopenic elderly. The American Journal of Clinical Nutrition, 103(3), 830-840. PMid:26864356. http://dx.doi.org/10.3945/ajcn.115.113357
  » http://dx.doi.org/10.3945/ajcn.115.113357
• Sardela, V. F., Martucci, M. E. P., de Araújo, A. L. D., Leal, E. C., Oliveira, D. S., Carneiro, G. R. A., Deventer, K., Van Eenoo, P., Pereira, H. M. G., & Aquino Neto, F. R. (2018). Comprehensive analysis by liquid chromatography Q‐Orbitrap mass spectrometry: Fast screening of peptides and organic molecules. Journal of Mass Spectrometry, 53(6), 476-503. PMid:29524299. http://dx.doi.org/10.1002/jms.4077
  » http://dx.doi.org/10.1002/jms.4077
• Thomas, J. R., Silverman, S., & Nelson, J. (2015). Research methods in physical activity (7th ed.). United States: Human Kinetics.
• Woo, H., Kim, J. W., Han, K. M., Lee, J. H., Hwang, I. S., Lee, J. H., Kim, J., Kweon, S. J., Cho, S., Chae, K. R., Han, S. Y., & Kim, J. (2013). Simultaneous analysis of 17 diuretics in dietary supplements by HPLC and LC-MS/MS. Food Additive and Contaminants: Part A, 30(2), 209-217. PMid:23116261. http://dx.doi.org/10.1080/19440049.2012.738939
  » http://dx.doi.org/10.1080/19440049.2012.738939

Publication Dates

• Publication in this collection
  27 June 2019
• Date of issue
  2019

History

• Received
  28 Aug 2018
• Accepted
  03 Jan 2019