Introduction

Mycotoxin contamination of foodstuffs has become a top priority issue in human health. In terms of regulatory aspects, since the discovery of the Aflatoxins (AFs) in the 1960s, regulations have been established in many countries to protect consumers from intake of harmful mycotoxin-contaminated foodstuffs, as well as to ensure fair practices in the food trade (López-García 2010).

Mycotoxins, products of secondary metabolism of filamentary micromycetes, represent the most important contaminants of natural origin. Ochratoxins (OTs) are a family of toxic compounds produced mainly by Aspergillus and Penicillium fungi that occur as natural contaminants of different foods. Of these, the most important, due to its toxicity and occurrence, is Ochratoxin A (OTA). Different studies have demonstrated that it can cause nephrotoxic, hepatotoxic, carcinogenic, teratogenic and immunotoxic effects in animals (Pfohl-Leskowicz and Manderville 2007).

In humans, OTA has been associated with urinary tract tumors, and in 1993 the International Agency for Research on Cancer (IARC) evaluated it “as a possibly carcinogenic to humans” (Group 2B) (IARC 1993). It is widely spread in agricultural commodities such as grains, spices, coffee, grapes and fruits.

Wine, after cereals, represents the second largest source of OTA human intake (about 13 %), identified by SCOOP report (Miraglia and Brera 2002) and by Joint FAO/WHO Expert Committee on Food Additives (JECFA 2007). OTs contamination of grapes takes place in the field and it is caused mainly by Aspergillus carbonarius. OTA was first detected in wines by Zimmerli and Dick (1996). Since then, the presence of OTA in wines has been reported from a number of European and other countries wines, with an apparent increase in levels in wines originating from southern areas of Europe including Italy, Greece, Turkey, some parts of Spain and France (Otteneder and Majerus 2000; Varga and Kozakiewicz 2006; Remiro et al. 2013).

OTA is sometimes accompanied by the non-chlorinated analogue, Ochratoxin B (OTB), OTB methyl and ethyl esters (MeOTB and EtOTB, respectively), and OTA ethyl and methyl esters: Ochratoxin C (OTC) and methyl Ochratoxin A (MeOTA), respectively (Remiro et al. 2012). Their chemical structures are shown in Fig. 1.

Fig. 1
figure 1

Chemical structures of Ochratoxin A and its analogues

Full size image

The simultaneous presence of OTA and its analogues in wine samples was recently studied (Remiro et al. 2013; Valero et al. 2008). The co-occurrence of different mycotoxins in one same foodstuff could origin additive or synergic effects on human or animal health (Heussner et al. 2006).

Several toxicological studies are reported in literature about OTA, OTB, OTC and their analogues (O’Brien et al. 2001; Mally et al. 2005; Knasmüller et al. 2004; Dietrich et al. 2001; O’Brien et al. 2005; Xiao et al. 1996; Müller et al. 2003, 2004).

Currently, EU legislation permits a maximum of 2 μg/kg of OTA in wine and grape juice (Commission Regulation [EC] 1881 2006); this limit does not apply to liquor or dessert wines with more than 15 % alcohol content. On the contrary, there are no legal limits for OTB in food and wine.

The presence of OTA in wines has been extensively studied. Taking into account the levels encountered in different Mediterranean countries, the highest OTA concentration was found in samples from North Africa followed by Greece, Italy, Spain, France, Turkey and Croatia, although all below the maximum levels permitted (Remiro et al. 2013).

AFs are a group of an extremely toxic mycotoxins produced mainly by the filamentous fungi Aspergillus flavus and A. parasiticus. Major members are designated as AFB1, AFB2, AFG1 and AFG2. Aflatoxins are acutely toxic, immunosuppressive, mutagenic, teratogenic, and carcinogenic agents (Williams et  al. 2004). AFB1 is the most predominant and the most toxic of the group, classified as a Group I carcinogen (IARC 1993) and thus carcinogenic to human. It is one of the most potent genotoxic agents and strongest hepatocarcinogens identified which cause liver tumor (Hamid et al. 2013). AF's chemical structures are shown in Fig. 2.

Fig. 2
figure 2

Chemical structures of the main Aflatoxins

Full size image

AFs and AF-producing strains (Fredj et al. 2007) have been detected in grape and musts occasionally (El Khoury et al. 2008; Aydogdu and Gucer 2009; Somma et al. 2012). Unfortunately, literature reports few studies regarding AF contamination in wines (Chunmei et al. 2013). The European Commission has established maximum permitted levels for AFs in cereals: 2 μg/kg for AFB1 and 4 μg/kg for the sum of AFB1, AFG1, AFB2 and AFG2 (Commission Regulation (EC) 1881 2006). At present, the EU has not set a maximum allowable limit for AFs in wine, but this does not means that the problem can be ignored.

Italy is the world’s second highest wine-producing country, behind France (OIV 2012), and Sicily is one of Europe's oldest viticultural regions. Today, Sicilian wines account for 17.5 % of the national total of Italian wine, and together with Puglia, Sicily is now the largest producer of Italian wine in Italy. Historically, Sicily has been world famous for the quality of its special sweet wines.

Sicilian sweet wines are made using the traditional method whereby grapes are grown on the vine as for a standard dry wine. Left to ripen for a little longer than is usual, providing weather allows, the grapes are harvested and laid out on mats or in racks and left for few weeks to dry. This process concentrates the sugars within the grapes to the point that the shrivelled, darkened grapes are almost raisin-sweet. It is at this point that fermentation and subsequent bottling takes place. This winemaking process leads to very fine wines, with higher alcohol content and extremely sweet. In Sicily, the geographical areas where sweet wines are traditionally produced are very limited. Noto, Siracusa and Pantelleria are famous for the production of Moscato and Passito, the Aeolian Islands for the production of Malvasia.

The choice of studying this type of wine comes because the traditional treatment of the grapes before fermentation (exposure of grapes to air, water reduction inside the berries and the simultaneous increase in the sugar concentration), makes the grapes more disposed to infection by bacteria and facilitates the development of molds which give rise to the formation of toxic metabolites.

The natural levels of OTA, OTB, AFB1, AFG1, AFB2 and AFG2 in 30 commercial sweet wines were investigated. Moreover, the simultaneous determination of the two classes of toxins is important because the presence of different mycotoxins in the same sample can produce synergistic effects on consumer health.

Materials and Methods

Samples

Sweet wines from five winemaking Sicilian regions were chosen for this study: Moscato of Siracusa, Moscato-Passito of Noto, Passito of Pantelleria, Malvasia of Lipari, Moscato of Partinico; their regional location is shown in Fig. 3.

Fig. 3
figure 3

Location of sweet wines with designation of origin

Full size image

Sweet dessert wines were purchased from local retails (bottled wine). A total of 30 samples belonging to the study areas were chosen. All wines were from 2007 to 2011 vintages. Their different alcoholic grade varied from 12.5 % to 14.5 % (v/v); their measured pH was in the range of 3.0–4.1.

All samples were stored at 4 °C until their analysis. All information on the samples was taken from the bottle labels.

Chemicals and Materials

Water, methanol and acetonitrile were HPLC grade. Glacial acetic acid, hexane, trifluoroacetic acid, sodium hydrogen carbonate and sodium chloride were all ACS grade. PEG 6000 and methanol were supplied by Merck (Darmstadt, Germany). Solvents and water were degassed for 20 min. using ultrasonic bath. The immunoaffinity columns used for OTs and AFs analysis were Ochraprep and Easy-extract, respectively, from r-Biopharm Rhone Ltd (Scotland, UK). OTA, OTB analytical standard solutions 10 μg/ml in acetonitrile, were purchased from Fluka and kept at −15 °C. Mix AFs standard solution at 1 μg/ml in acetonitrile (containing 250 ng of AFB1, 250 ng of AFB2, 250 ng of AFG1 and 250 ng of AFG2) were purchased from r-Biopharm Rhone Ltd and kept at −8 °C. The calibration standard solutions were made in mobile phase according to the concentration established, kept in security conditions, wrapped in aluminium foil, due to that mycotoxins gradually break down under UV light, held for less than 3 months and stored at −15 °C.

Ochratoxins and Aflatoxins Clean-up

We developed a method modified in comparison to that described by Visconti et al. (1999) because our purpose was to identify and quantify OTs and Afs simultaneously.

Sweet wine samples (10 ml) were diluted with 10 ml of water solution containing PEG 6000 (1 %) and NaHCO3 (5 %), in this way the pH of the diluted wine was always within the optimal range (pH 7–8.5). The mixture was shaken vigorously and vortexed for 2 min. The IAC was placed on a SPE vacuum manifold (Visiprep, Supelco) and preconditioned with 3 ml PBS. Then, the mixture of the diluted sample (20 ml) was applied to the Ochraprep column (1–2 drops/s). The column was successively washed with 5 ml water solution containing NaCl (2.5 %) and NaHCO3 (0.5 %) followed by 5 ml distilled water at a flow rate of 1–2 drops/s and dried with air. OTs were then slowly eluted from the IAC with 2 ml methanol/acetic acid 98:2 into a glass vial; the eluate was evaporated completely with nitrogen gentle stream and reconstituted with 500 μl of mobile phase before to HPLC analysis.

The extraction of the AFs was adapted from the OTs cleanup method with modifications. Ten milliliters of wine was taken on 10 ml of water solution containing PEG 6000 (1 %) and NaHCO3 (5 %), shaken vigorously and vortexed for 2 min. The Easy-extract AFs column was preconditioned with 3 ml PBS, and the mixture (20 ml) was applied to the IAC (1–2 drops/s). The column was successively washed with 5 ml water solution containing NaCl (2.5 %) and NaHCO3 (0.5 %) followed by 5 ml distilled water at a flow rate of 1–2 drops/s and dried with air. AFs were then slowly eluted from the IAC with 3 ml of acetonitrile into a glass vial. Also, in this case, the solutions obtained were dried under a gentle stream of nitrogen, but before HPLC/FLD analysis, the sample must be derivatized in order to increase the natural fluorescence of AFG1 and AFB1 (Fig. 4).

Fig. 4
figure 4

Aflatoxins AFB2a and AFG2a by derivatization of AFB1 and AFG1

Full size image

Precolumn Derivatization

The eluate containing AFs was evaporated to dryness with a gentle stream of N2, redissolved with 400 μl of hexane and 100 μl of trifluoroacetic acid. The mixture was stirred for 1 min. and was made to stand for 5 min. Then the sample was added to 2 ml of a solution water/acetonitrile (9:1), stirred vigorously, then left until separation of the two phases and finally 1 ml of the lower phase containing AFs was drawn through the filters PTFE 0.45 μm and placed in a glass vial, ready for HPLC/FLD analysis. The solution containing derivatized AFs (AFB2a and AFG2a), must be readily analyzed in order to avoid deterioration due to their limited stability.

LC-FLD Analysis

OTs and AFs analysis was carried out in an Agilent Technologies 1100 liquid chromatographic system equipped with a Fluorescence detector (model G1321A), controlled by Chemstation software. A selected RP-18 analytical column (5 μm particle size, 150 mm × 4.6 mm i.d.) (Phenomenex Luna C18 (2)100A) fitted with guard column was chosen. The mobile phase consisted of acetonitrile/water/acetic acid (99:99:2, v/v/v) for OTs analysis and acetonitrile/water (70:30 v/v) for AFs analysis. The injection volume was 50 μl, and the flow rate was 1.0 ml/min. Chromatography was performed at 40 °C, and the fluorescence conditions were as follows: excitation at 335 nm and emission at 465 nm for OTs analysis and excitation at 360 nm and emission at 440 nm for AFs ones for the entire analysis. In these chromatographic conditions, the retention times were 4.04 and 6.55 min. for OTB and OTA, respectively, as shown in Fig. 5, and 3.07, 3.53, 5.83 and 7.25 min. for AFG1, AFB1, AFG2 and AFB2, respectively, as shown in Fig. 6. The limit of detection (LOD) was determined based on the lowest quantity of analyte that can be clearly distinguished from background (S/N = 3). LOD was about 0.01 μg/l for OTA, AFB1 and AFB2, 0.02 μg/l for OTB and AFG2 and 0.03 μg/l for AFG1. Quantification of mycotoxins analyzed was performed by measuring peaks areas of AFs and OTs at retention time and comparing them with the appropriate calibration curves.

Fig. 5
figure 5

Chromatogram of OTA (6.55 min.) and OTB (4.04 min.) standard solution 2 μg/l (injected volume 50 μl)

Full size image
Fig. 6
figure 6

Chromatogram of AFs (AFG1, AFB1, AFG2 and AFB2) standard solution 2 μg/l (injected volume 50 μl)

Full size image

Preparation of Standard Calibration Curve

From OTA and OTB standard solutions (1 μg/ml in acetonitrile), and from mix AFs standard solutions (250 μg/l each AFs in acetonitrile) were prepared six different calibration solution at concentrations of 0.5, 1, 1.5, 2, 4, and 8 μg/l using mobile phase for dilution.

Each solution at different concentration was prepared in duplicate and injected in triplicate, the mean was obtained from all measurements. Calibration curve was constructed as the dependence of the peak area on concentration of the standard. Calibration solutions were freshly prepared each day before the measurement. The regression coefficients (r 2) were 0.99 for all toxins analyzed.

Performance of the Analytical Method

Results of recovery experiments of the full analytical procedure (triplicate measurements) carried out on four mycotoxins-free sweet wines spiked with the six tested mycotoxins (OTA, OTB, AFB1, AFG1, AFB2 and AFG2) at two different levels (0.5 and 2 μg/l) are reported in Table 1.

Table 1 Recovery results from the analysis of blank samples of Sicilian sweet wines spiked with OTA, OTB, AFB1, AFB2, AFG1 and AFG2 at two different levels
Full size table

Average recoveries of AFs and OTs ranged from 75 % to 87 % and from 89 % to 95 %, respectively. Values of relative standard deviation (RSD) ranged from 2.7 % to 8.0 %. Results of recovery experiments are in accordance with those reported in the literature (OIV-MA-AS315-10: R2011).

Precision was calculated in terms of intra-day repeatability (n = 5) and inter-day reproducibility (5 different days) on two level of concentration (2 and 4 μg/l). The results of the tests are shown in Table 2.

Table 2 Repeatability (intra-day) and reproducibility (inter-day) results from analysis of blank samples with micotoxins at two different levels
Full size table

Linearity was verified (n = 5) with six concentrations, 0.5, 1, 2, 3, 4 and 8 μg/l. The regression coefficients (r2) were 0.999 for OTA, 0.998 for AFG1, 0.997 for OTB, AFB1 and AFB2 and 0.095 for AFG2. The LOD of the method was about 0.012 μg/l for OTA, AFB1 and AFB2, 0.02 for OTB and AFG2 and 0.04 for AFG1.

The results of the study reflected that the analysis gave good repeatability, rapidity and precision. Recoveries were considered as valid for analyzing residues of OTA in wines according with European signification (Commission Regulation (EC) 401 2006).

The slope values of the regression curves obtained with standard solutions were then compared with the slope values obtained by the method of standard additions using the Student t test. In both studies, the slopes obtained were similar for a 95 % confidence limit, so it was concluded that the matrix effect was absent.

Results and Discussion

The analytical method was applied to 30 samples sweet wines originating from five sweet-wine-making Sicilian regions. The findings are reported in Table 3.

Table 3 Ochratoxins and Aflatoxins analyzed occurrence in Sicilian dessert wines
Full size table

None of the analyzed wines contained OTA and OTB at a concentration in excess of the tolerated limit at 2 μg/kg.

The levels of AFs, if present, were very far from those currently enforced for other regulated foodstuffs, ranging from 0.10 to 8 μg/kg for AFB1 and from 4 to 15 μg/kg for the sum of all four Aflatoxins (Commission Regulation (EC) 401 2006).

The six mycotoxins investigated did not appear simultaneously in studied samples.

Most of these samples were contaminated by OTs and the levels of OTA ranged from <LOD to 1.56 μg/l, while OTB ranged from <LOD to 1.205 μg/l. The highest OTs values were found in a sample of white Moscato of Noto. Only one sweet dessert wine contained OTA below the LOD while five samples contained OTB below the LOD value.

Results revealed the presence, to detectable values, of OTA in 96.6 % and of OTB in 83.3 % of the samples examined (mean 0.246 and 0.168 μg/l, respectively).

AFs showed a very rare presence. The concentration of at least one AFs has been clearly established in 30 % of tested samples, while about 23 % of the samples showed a contamination at values below LOD. AFB1, AFG1 and AFG2 were in 13.3 % of the samples, while AFB2 only in the 10 % of them. In all cases, levels of contamination were very low. Their mean values were 0.0247, 0.043, 0.0146 and 0.0265 μg/l for AFB1, AFG1, AFB2 and AFG2, respectively. The highest AFs value found was 0.068 μg/l for AFG1.

AFs and OTs chromatograms of both natural contaminated are shown in Figs. 7 and 8.

Fig. 7
figure 7

Chromatogram of naturally OTs contaminated sweet wine sample

Full size image
Fig. 8
figure 8

Chromatogram of naturally AFB2 and AFG2 contaminated sweet wine sample

Full size image

Conclusion

This is the first report of OTA, OTB and the main AFs (AFB1, AFB2, AFG1 and AFG2) in dessert Sicilian wines.

OTs and AFs contents in Sicilian sweet wines tested in this survey suggested that a previously existing fungal contamination in the grape could occur, and this contamination, associated with the conditions of the drying process itself, led to the growth of the fungus and, consequently, of toxic metabolites.

This finding is related to the data provided by Valero et al. (2008) which demonstrated that it is after the harvesting, and particularly in the process of raising, when the Aspergillus species becomes predominant over other fungi present on the grape, due principally to greater tolerance of high temperatures and low activity of the water (aw).

The use of IAC for clean-up followed by HPLC has shown to be a technique with good analytical performance for mycotoxins determination in wines. The choice of the method for the identification of tested mycotoxins which provides for their determination by HPLC/FLD is the most used method. The instrumentation is not very expensive and does not require the use of highly specialized staff. So this method can be developed in small laboratory inside wineries, for the control of wines prior to commercialization.

The application of this procedure to analyze 30 sweet wines from Sicily has demonstrated that none of them contained levels above the European MRLs, but the study suggests that there is a real risk of contamination by OTs and AFs in Sicilian sweet wines. In any case, due to their toxicity, a low but continued dietary exposure could contribute to a risk to humans, so the level of these mycotoxins in food should be as low as reasonably achievable.

However the incidence of OTA in wines produced in Sicily is definitively very low, compared with the findings in other wine producing countries. The four AFs were found separately in the different samples analyzed, with values still well below the lowest value reported in the EC Regulation no. 165 (2010) for foods.

The authors consider it necessary to continue the study in order to assess the mycotoxins contamination in the various years of production and to evaluate the effective influence of environmental parameters such as humidity, temperature, latitude and geographical location of the vineyard.

In conclusion, this study strongly confirms the excellent quality and safety of Sicilian sweet wines that are proven to be contaminated at levels far below the limits given in almost all samples analyzed.