Mycotoxins are naturally occurring secondary metabolites of fungi that can be produced on a wide range of agricultural commodities. Cereals used in beer production, particularly barley, wheat, and sometimes maize, can be contaminated with different mycotoxins. Under favourable ecological conditions, cereal contamination can occur both in field and during storage, and also during malting. Previous studies, investigating the fate of naturally occurring or spiked mycotoxins at various stages of the brewing process, have demonstrated that mycotoxins can be transmitted from contaminated grains into beer [1
]. Considering the most known mycotoxins, numerous surveys have reported the occurrence of ochratoxin A (OTA) and deoxynivalenol (DON) in several beer samples, while zearalenone and aflatoxins were rarely detected. Fumonisins were found only when maize was used in the brewing process. In addition to known mycotoxins, for which maximum levels in food are enforced, also, the so-called, currently unregulated, “emerging mycotoxins”, can frequently occur in agricultural products. Among these toxins, EFSA (European Food Safety Authority) requested data on the presence of sterigmatocystin (STC) and citrinin (CIT) in food [5
]. All these mycotoxins have been associated with human and animal diseases; further, the International Agency for Research on Cancer (IARC) has classified OTA and STC as possible human carcinogens (class 2B). As a result of too limited or inadequate data, CIT and DON were not classifiable with regard to their carcinogenicity to humans (class 3) [7
In Italy, the consumption of beer has increased over the past few years (about 31.5 L per person/year in 2016), in particular, of the small-scale brewed beer (called craft beer); an opening of about 500 microbreweries was registered in Italy from 2007 to 2017. In 2016, the Italian Parliament defined as craft beer “the unpasteurized and unfiltered beer produced by micro-breweries, whose production is less than 200,000 hectoliters/year” [10
]. Craft beer group is a heterogeneous class of beers, sometimes brewed from unusual ingredients; they were scarcely investigated for the content of both healthy and undesirable components, as mycotoxins.
The objective of this study was to evaluate and compare the incidence and contamination level of some mycotoxins, OTA, DON, STC, and CIT, in small- and large-scale brewed beer samples; for both types of beers, the most-sold beers in Italy were collected.
2. Materials and Methods
A total of 83 samples of small- (SS) and large-scale (LS) brewed beer (42 and 41, respectively) were sampled; for both types of beer, the most-sold beers in Italy were collected (data obtained from IRi reports; Information Resources), representing about 75% of the total volume of beers sold in Italy. The percentage of beer samples produced in Italy was 47.6% (n = 20) and 41.5% (n = 17) for SS and LS brewed beers, respectively. The majority of LS brewed beers were regular beer (36); the others were wheat beer (3), red (1), and double malt beer (1). Among SS brewed beers, 25 were regular, 6 wheat beer, 5 red, 4 double malt, and 2 dark beer. A representative sampling, putting together at least 3 samples, collected in different retail shops or pubs within a three-month period (June–August 2017), was carried out; the samples were mixed and two subsamples (about 100 mL each) were kept at +4 °C. Before the analysis, each sample was gently shaken and degassed by ultrasonication.
2.2. Reagents and Standards
The chemicals and solvents used for the extraction and clean-up solutions were ACS grade or equivalent (Carlo Erba, Milan, Italy); deionised water was purified through a Milli-Q treatment system (Millipore, Bedford, MA, USA). For LC-MS/MS analysis, water, methanol, acetonitrile, and formic acid were HPLC grade (Merck, Darmstadt, Germany). Phosphate saline buffer (PBS) was prepared as follows: NaCl 8 g/L, KCl 0.2 g/L, Na2
1.15 g/L, KH2
0.2 g/L; pH 7.4. OTA and trichothecene standard solutions were prepared as described by Bertuzzi et al. [11
The STC analytical standard was obtained from Sigma-Aldrich (Milan, Italy), while the internal standard [13
]-sterigmatocystin (96.4% 13
C) was purchased from Biopure (Tulln, Austria) as standard solution in acetonitrile (1.2 mL, 25.7 μg/mL, uncertainty 1.02 μg/mL). Stock and working STC standard solutions were prepared and calibrated spectrophotometrically as reported by Bertuzzi et al. [12
The CIT standard (5 mg), purchased as powder from Sigma-Aldrich (Milan, Italy), was dissolved in ethanol (25 mL), and the solution, after dilution, was calibrated spectrophotometrically at 319 nm using the value 4710 L/(mol·cm) for the absorption coefficient (Neely et al., 1972) and stored at −20 °C when not in use. 13C13-citrinin internal standard solution (in acetonitrile) was purchased from Romer Labs (10.6 µg/mL; Romer Labs Division Holding GmbH, Getzersdorf, Austria). Working standard solutions, mixed with isotopically-labelled CIT standard solution (10 µg/mL; 90:10 v/v), were prepared by dilution with methanol–water (30:70 v/v) in the range between 0.10 and 5.0 µg/L.
2.3. Analysis for Mycotoxin Determination
Analysis for OTA and DON quantification was carried out using HPLC-FLD for OTA and GC-MS for DON, according to a previously published method (Bertuzzi et al. [11
For STC, an aliquot of degassed sample (15 mL) was diluted with 15 mL of PBS, and was then passed through an immunoaffinity column (Easi-extract Sterigmatocystin, R-Biopharm-Rhône, Glasgow, UK). The column was washed with 2 mL water, then STC was eluted using 6 mL acetonitrile in a graduated test tube. After concentration under a gentle flow of nitrogen, the purified extract was brought to 1 mL with acetonitrile–water 40 + 60 v/v; then, 900 µL was mixed with 100 µL of isotopically labelled STC (12 µg/L) in an autosampler vial and injected (20 µL) into an LC-MS/MS system consisting of a LC 1.4 Surveyor pump, a Quantum Discovery Max triple-quadrupole mass spectrometer (Thermo-Fisher Scientific, San Jose, CA, USA), and a PAL 1.3.1 sampling system (CTC Analytics AG, Zwingen, Switzerland). The system was controlled by Xcalibur 1.4 software (Thermo-Fisher Scientific, Waltham, MA, USA). After separation on a Betasil RP-18 column (5 µm particle size, 150 × 2.1 mm, Thermo-Fisher) with a gradient of acetonitrile–water (both acidified with 0.2% formic acid; flow rate 0.2 mL/min) from 40:60 to 75:25 in 3 min, then, isocratic 40:60 for 6 min; the flow rate was 0.2 mL/min. The ionisation was performed using positive atmospheric pressure chemical ionisation (APCI) as follows: voltage 4.0 kV, sheath and auxiliary gas 29 and 5 psi, respectively, temperature of the heated capillary 270 °C. The mass spectrometric analysis was performed in selected reaction monitoring (SRM). For fragmentation of the [M + H]+ ions (325 and 343 m/z for STC and [13C18]-STC, respectively), argon was used as collision gas at the pressure of 1.5 mTorr. For STC, three transitions were measured: 310 m/z (24 V) (quantifier), 281 and 253 m/z (35 V) (qualifiers). For the isotopic label, the transitions were: 327 m/z (24 V) (quantifier), 297 and 268 m/z (35 V) (qualifiers).
For CIT, a 15 mL degassed sample was mixed with 15 mL PBS and the solution was passed through an immunoaffinity column (Easi-extract Citrinin, R-Biopharm-Rhône, Glasgow, UK). The column was washed with PBS (2 mL) and CIT was slowly eluted (0.5 mL/min) with methanol (4 mL) into a graduated test tube; the eluate was concentrated under a gentle stream of nitrogen, brought to 1 mL with methanol–water (30:70 v/v) and vortex-mixed for a few seconds. The extract was filtered (Millex HV 0.45 µm, Merck Millipore, Darmstadt, Germany), then 100 µL of internal standard solution (13C-CIT13, 10.6 µg/L) was added to 400 µL of purified extract. Analysis was carried out using a HPLC-MS/MS system (see above). CIT was separated on a XBridge BEH RP-18 column (2.5 µm particle size, 100 × 3.0 mm, Waters Corporation, Milford, MA, USA) with a mobile-phase gradient methanol–water (both acidified with 0.2% formic acid) from 45:55 to 75:25 in 3 min, then isocratic for 4 min; gradient to 45:55 in 1 min and re-equilibration for 7 min. The flow rate was 0.2 mL/min; the injection volume was 20 µL. The ionisation was carried out with an ESI interface (Thermo-Fisher Scientific) in positive mode as follows: spray capillary voltage 4.0 kV, sheath and auxiliary gas 35 and 8 psi, respectively, temperature of the heated capillary 270 °C. For fragmentation of [M + H]+ ion (251 m/z), the argon collision pressure was set to 1.5 mTorr; the selected fragment ions were: 233 m/z (20 V), 205 and 191 m/z (28 V). For the isotopic label (13C13), 264 m/z was used as precursor ion with 246 m/z (18 V), 217 and 202 m/z (26 V) as product ions. Quantitative determination was performed using a LC-Quan software (2.0, Waltham, MA, USA).
2.4. Performances of the Methods
The matrix effect in LC-MS/MS is due to the presence of compounds that can co-elute, affecting the ionisation of the analyte. This effect was avoided by the use of isotopically labelled [13C18]-STC and [13C13]-CIT as internal standard; therefore, the internal standard calibration was applied, adding the isotope-labelled standard to all sample extracts and calibration standards at a fixed concentration.
The linearity of the instrumental measurement was established through five calibration standards in solvent, at different concentrations for each mycotoxin, covering a wide analytical range.
The limit of detection (LOD) was defined as the level corresponding to a signal-to-noise ratio (S/N) of three in the purified extract for qualifier transition. The limit of quantification (LOQ) was defined in the purified extract as the lowest level for which the repeatability of the analysis was below 10%. The LOD and LOQ were 0.001 and 0.003 µg/L for OTA; 1 and 3 µg/L for DON; 0.001 and 0.003 µg/L for STC; 0.005 and 0.015 µg/L for CIT.
The accuracy of the used methods was evaluated by recovery experiments, performed by spiking a regular, a double malt, and a red beer sample in triplicate at two levels (2× LOQ and 5× LOQ), for each mycotoxin. The samples were also analysed without spiking, as well as a reagent blank. The spiked sample was allowed to stand for two hours at ambient temperature under a fume hood to allow any residual solvent to evaporate. For each mycotoxin, the average recovery ranged between 86.4% and 95.6%, with low standard deviations, indicating a satisfactory accuracy and precision for each method. The results of the analyses were corrected for recovery.
2.5. Statistical Analysis
Statistical analysis of mycotoxin contamination data was carried out after common logarithm transformation. This is suggested when the variance of data is higher than the mean. Statistical analysis was run using the package IBM SPSS statistics (ver. 24, 2015 Inc., Chicago, IL, USA). One-way ANOVA was applied to evaluate significant differences between SS and LS beers.