Analytica Chimica Acta 662 (2010) 6268

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: www.elsevier.com/locate/aca

Simultaneous determination of aatoxins B1, B2, G1, G2, M1 and M2 in peanuts and their derivative products by ultra-high-performance liquid chromatographytandem mass spectrometry
Baifen Huang a , Zheng Han b , Zengxuan Cai a , Yongjiang Wu b , Yiping Ren a,
a b

Zhejiang Provincial Center for Disease Prevention and Control, Hangzhou 310051, Zhejiang Province, China Zhejiang University, Hangzhou 310058, Zhejiang Province, China

a r t i c l e

i n f o

a b s t r a c t
A reliable ultra-high-performance liquid chromatographytandem mass spectrometry (UHPLCMS/MS) method for simultaneous determination of aatoxins B1, B2, G1, G2, M1 and M2 in peanuts and their derivative products was developed. The sample was extracted by 84% of acetonitrile aqueous solution and the extract was puried by a reliable solid phase extraction-based clean-up method. Then, the analytes were separated on Acquity UPLC HSS T3 column (100 mm 2.1 mm, 1.8 m particle size), and eluted with a mobile phase consisting of (A) water containing 0.1% formic acid and (B) acetonitrile/methanol (50/50, v/v). The separated compounds were detected with a Waters Micromass Quattro Ultima Pt tandem quadrupole mass spectrometer operating in positive electro-spray ionization using multiple reaction monitoring mode. The established method was extensively validated by determining the linearity (R2 0.9990), average recovery (74.786.8%) and precision (relative standard deviation 10.9%). It was shown to be a suitable method for simultaneous determination of the six aatoxins in peanuts and their derivative products. Finally, a total of 73 samples randomly collected from different areas in Zhejiang province were screened for aatoxins with the proposed method. The results showed that 31 samples of peanut butter, 14 samples of fresh peanut and 5 samples of musty peanut were contaminated with aatoxins. Meanwhile, this was the rst report on aatoxins M1 and M2, which were found in unprocessed peanuts and their derivative products. 2010 Elsevier B.V. All rights reserved.

Article history: Received 3 September 2009 Received in revised form 24 December 2009 Accepted 4 January 2010 Available online 11 January 2010 Keywords: Ultra-high-performance liquid chromatographytandem mass spectrometry Aatoxins Peanuts

1. Introduction Aatoxins (AFs) are a group of mycotoxins produced as secondary metabolites by the spoilage of fungi Aspergillus, particularly Aspergillus avus and Aspergillus parasiticus [1]. The most important members are AFB1, AFB2, AFG1, AFG2, AFM1 and AFM2. They are highly toxic and carcinogenic compounds that cause disease in livestock and humans [24]. The International Agency for Research on Cancer (IARC) has claried AFB1, AFB2, AFG1 and AFG2 in the group I as human carcinogen. Although the toxicity of AFM1 is lower than AFB1, it is known for its hepatotoxic and carcinogenic effects, and has also been claried in the group I by IARC [5]. In previous studies, AFM1 and AFM2 were always recognized as the hydroxylated metabolites of AFB1 and AFB2, as a consequence, the researches mainly focused on the study of milk and milk products [69]. However, in a recent report, AFM1 was also detected in some peanut butter samples with the maximal contents even nearly 1.30 g kg1

Corresponding author. E-mail address: zjryp@yahoo.cn (Y. Ren). 0003-2670/$ see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2010.01.002

[10]. Nevertheless, this study did not nd out whether AFM1 existed in natural peanut or was just introduced by any milk derivative or by-product into the peanut butter samples. Hitherto, there are no methods available for the determination of AFM1 and AFM2 in unprocessed peanuts. Therefore, development and validation of a robust method for simultaneous determination of the six AFs in peanut samples are urgently requested. Some analytical techniques such as thin-layer chromatography (TLC) [11,12], high-performance liquid chromatography (HPLC) [13,14], two-dimensional thin-layer chromatography [15] and enzyme-linked immunosorbent assay (ELISA) [1618] have been available for the qualitative and quantitative analysis of AFs. Poor separation, unsatised accuracy and low sensitivity limit the application of TLC. Although ELISA is a fast and sensitive method, false positive result might be obtained. The most frequently used method for AFs analysis is liquid chromatography combined with uorescence detection, which has been extensively studied in various food matrices [13,19,20]. However, conventional approach by HPLC in a gradient reversed phase mode typically using columns with 5 m particles often costs a lot of time to get a complete separation of the target compounds [21] and, additionally, in order

B. Huang et al. / Analytica Chimica Acta 662 (2010) 6268

63

Fig. 1. The chemical structures of the six AFs in the present study.

Table 1 The parameters and collision energy of parent ions, primary daughter ions and secondary daughter ions, concentration range ( g kg1 ), limit of detection (LOD), limit of quantication (LOQ) and linearity (R2 ) for AFs. AFs Parent ion (m/z) 313.00 (+H) 315.00 (+H) 329.06 (+H) 331.01 (+H) 329.00 (+H) 331.07 (+H) Primary daughter ion (m/z) 285.00 287.10 243.03 245.10 273.10 273.10 Collision energy (eV) 23 26 22 24 22 20 Secondary daughter ion (m/z) 241.00 259.10 283.06 217.10 259.10 285.07 Collision energy (eV) 32 26 24 26 22 20 R2 a Range ( g kg1 ) 0.110.0 0.420.0 0.420.0 0.420.0 0.420.0 0.420.0 LODb ( g kg1 ) LOQc ( g kg1 )

AFB1 AFB2 AFG1 AFG2 AFM1 AFM2

a b

0.9990 0.9993 0.9991 0.9990 0.9993 0.9994

0.009 0.056 0.085 0.212 0.017 0.106

0.012 0.084 0.182 0.273 0.021 0.138

Regression coefcient. Limit of detection (S/N = 3, transition: 313.00 > 241.00 for AFB1, 315.00 > 259.10 for AFB2, 329.06 > 283.06 for AFG1, 331.01 > 217.10 for AFG2, 329.06 > 259.10 for AFM1 and 331.07 > 285.07 for AFM2); c Limit of quantitation (S/N = 10, transition: 313.00 > 285.00 for AFB1, 315.00 > 287.10 for AFB2, 329.06 > 243.03 for AFG1, 331.01 > 245.01 for AFG2, 329.00 > 273.10 for AFM1 and 331.07 > 273.10 for AFM2).

64

B. Huang et al. / Analytica Chimica Acta 662 (2010) 6268

to improve detection limits of AFB1 and AFG1 a tedious pre- or post-column derivatization must be done. The present study will cope with this drawback using the ultra-high-performance liquid chromatography (UHPLC) method, the particle diameter of stationary phase reduced from 5 m (HPLC) to sub-2 m (UHPLC) results in increased speed, better chromatographic resolution and improved sensitivity, selectivity and specicity [10]. The introduction of MS/MS detection also leads to a signicant improvement for the analysis of AFs because of its high sensitivity and ability to quantify AFs without derivatization [22]. Despite the high sensitivity and selectivity of LCMS/MS method, the quantity control is more time consuming and the signal suppression (enhancement) is variable. The aim of this study is to develop a rapid and reliable UHPLCMS/MS method for simultaneous determination of the six AFs in peanuts and their derivative products, mainly considering the actual contaminant situations in Zhejiang province in China. 2. Experimental 2.1. Chemicals, reagents and materials The standards of AFB1, AFB2, AFG1, AFG2, AFM1 and AFM2 were purchased from SigmaAldrich (St. Louis, MO, USA). The chemical structures of the six AFs are shown in Fig. 1. Water was puried using a Milli-Q Gradient A 10 system (Millipore, Billerica, MA, USA). Acetonitrile and methanol were of HPLC grade (Merck, Darmstadt, Germany). Formic acid and other solvents were of analytical grade. Silica, neutral alumina, base alumina and diatomite were purchased from SigmaAldrich (St. Louis, MO, USA). Polypropylene 6 mL SPE empty tubes were purchased from Shenzhen Biocomma Biotech CO, LTD. Oasis HLB SPE cartridges (200 mg, 6 cm3 ) were purchased from Waters Corporation (Milford, MA, USA). Mycosep 226 Aazon+ Multifunctional cartridges were purchased from Romer Labs (Washington, MO, USA). High quality poly(9,9diethyluorene) (PDEF) syringe lters (0.22 m pore size, 13 mm diameter) were by Millipore. 2.2. UHPLCMS/MS UHPLC analysis was performed using a Waters Acquity Ultra Performance LC system (Waters, Milford, MA, USA). Chromatographic separation was achieved using a Waters Acquity UPLC HSS T3 column (100 mm 2.1 mm, 1.8 m particle size; Waters, Milford, MA, USA) with a mobile phase ow rate of 0.3 mL min1 . The mobile phase consisted of (A) water containing 0.1% formic acid and (B) acetonitrile/methanol (50/50, v/v). A gradient elution program was applied as follows: 04.2 min linearly increased from 32% to 45% B, 4.25.0 min linearly increased from 45% to 100% B, 5.05.7 min hold on 100% B, 5.76.0 min linearly decreased from 100% to 32% B. A subsequent re-equilibration time (3 min) should be performed before next injection, giving a total run time of 9 min. The injection volume was 5 L. Moreover, the column and sample temperature were maintained at 40 and 20 C, respectively. MS/MS was performed on a Waters Micromass Quattro Ultima Pt tandem quadrupole mass spectrometer (Micromass, Manchester, UK). The parameters used for the mass spectrometer under the ESI+ mode were set as follows: capillary voltage of 3.50 kV, cone voltage of 45 V, source temperature of 120 C and desolvation temperature of 350 C. The cone and desolvation gas ow were 50 and 570 L h1 , respectively. Quantitation was performed in multiple reaction monitoring (MRM) mode and the conditions were

Fig. 2. Recovery tests of three candidate cartridges.

optimized for each AF during infusion. The parameters on the m/z and collision energy of parent ions, primary daughter ions and secondary daughter ions selected for the analysis of the AFs are shown in Table 1. Data acquisition and processing were performed using Mass Lynx v4.1. 2.3. Preparation of standard solutions Accurately weighed solid portions of each AF standard were dissolved in acetonitrile to prepare 0.1 mg mL1 of stock solutions and stored at 20 C in the darkness. All working solutions were prepared immediately before use by diluting the stock solutions with the blank matrix. 2.4. Preparation of home-made mixed cartridge 1.35 g of silica gel was accurately weighed into a polypropylene 6 mL SPE empty tube and the tube was shaken severely to compact the silica gel. Then, neutral alumina (0.9 g) and kieselguhr (0.25 g) were added on the above of the silica gel. Finally, a cribriform plate was put on the top layer to ensure the supine surface smooth and at. 2.5. Samples A total of 73 samples were randomly collected from local stores in three cities in Zhejiang province, including Hangzhou (33 fresh peanuts, 5 musty peanuts and 25 peanut butters), Ningbo (1 fresh peanut and 6 peanut butters) and Quzhou (1 fresh peanut and 2 peanut butters). Samples were delivered to the laboratory within 24 h of collection and immediately dispensed into plastic bags. All samples were stored at 4 C until analyzed. 2.6. Sample preparation A 2.5 g sample was macerated with 10 mL of 84% acetonitrile aqueous solution for 20 min, and then homogenized for 3 min with IKA T25 high speed homogenizer (Ika-Werke Gmbh, Staufen, Germany) at 13,500 rpm. After extraction, the solution was centrifuged 21,130 g for 15 min with Beckman Coulter Allegra 64R centrifuge (Brea, CA, USA). The supernatant was ltrated and an aliquot of 5 mL ltration was passed through the home-made mixed cartridge, and eluted with 3 mL methanol. The efuents were combined and

B. Huang et al. / Analytica Chimica Acta 662 (2010) 6268

65

dried under nitrogen gas stream at 40 C, the residue was redissolved with 1 mL combined solution (acetonitrile/water, 10/90, v/v), passed through a 0.22- m lter and ready for injection. 3. Results and discussion 3.1. Selection of UHPLCMS/MS columns In the previous study, AFM1 and AFG2 could not be separated [10]. To get a complete separation of the six AFs, three LC columns with different lengths and particle sizes, i.e. (1) Acquity UPLC HSS T3 column (2.1 100 mm, 1.8 m particle size), (2) Acquity UPLC BEH C18 column (2.1 mm 100 mm, 1.7 m particle size), (3) Acquity UPLC BEH SHIELD RP18 column (2.1 mm 100 mm, 1.7 m particle size), were tested. The separation efciencies of column 1 and 3 were obviously better than column 2. When column 3 was chosen, the peaks observed were very asymmetry. At last, column 1 was selected. Under such situation, symmetrical and sharp peaks were got and the elution time was also greatly shortened. The run time per sample in our test was only 9 min with a ow rate of 0.3 mL min1 . 3.2. Selection of mobile phase The choice of mobile phase should be concerned based on the consideration of ionization and separation efciencies. In the previous study, methanol was employed as the strong elution mobile phase. However, under such situation, AFG2 and AFM1 could not be separated [10]. Therefore, the rates of methanol/acetonitrile varied from 90/10 (v/v) to 50/50 (v/v) were compared. The results indicated that all analytes could be completely separated in the shortest time with methanol/acetonitrile (50/50, v/v) as the strong elution mobile phase. For the choice of the weak elution mobile phase, water containing 0.1% formic acid and 10 mM ammonium acetate were compared. The results of multiple injections indicated that nice peak shape and higher sensitivity could be achieved when 0.1% formic acid was added. Finally, the mobile phase was selected as (A) water containing 0.1% formic acid and (B) acetonitrile/methanol (50/50, v/v). 3.3. Optimization of sample pretreatment Four SPE materials (silica, neutral alumina, base alumina and diatomite) were tested for their suitability for the clean-up of sample extracts. Firstly, we evaluated the recovery performance by passing a mixed standard solution through the column lled with one material. All materials except base alumina showed satisfactory recoveries of 8088%. Secondly, the purication efciency was tested. Spiked peanut extracts were cleaned up with the columns lled with two or three materials with different ratios and the sensitivities were compared. Based on these, a simple and reliable home-made mixed cartridge was developed. To characterize and clarify the established method, home-made mixed cartridge, Oasis HLB SPE cartridge (200 mg, 6 cm3 ) and Mycosep 226 Aazon+ Multifunctional cartridge were compared for the sample pretreatment. 5 mL of mixed standard solution (100 ng mL1 ) was diluted with blank matrix to prepare working solutions (5 ng mL1 ). The test was performed as follows: (1) 5 mL aliquots were puried by home-made mixed cartridges as described in the Section 2.6, (2) 5 mL aliquots were concentrated to about 1 mL by nitrogen gas at 40 C, then diluted with water and xed the volume to 5 mL. The solution was passed through the reconditioned Oasis HLB SPE cartridges at a rate of about 12 drops/s, and then 5 mL of 5% methanol-water solution was passed through the cartridges at a rate of about 12 drops/s. All targets were eluted with 6 mL methanol at a rate of about 12 drops/s, and

Table 2 Recovery tests of UHPLCMS/MS method (n = 3). AFs High levela (%) X SD AFB1 AFB2 AFG1 AFG2 AFM1 AFM2 82.7 81.7 84.0 82.3 83.6 82.8 4.9 3.2 2.0 1.8 3.6 1.5 RSD% 6.0 4.0 2.4 2.2 4.4 1.9 Intermediate levelb (%) Low level c (%) X SD 81.2 84.0 86.8 80.1 86.0 85.3 5.3 1.9 3.7 2.1 3.8 1.9 RSD% 6.6 2.3 4.3 2.7 4.5 2.3 X SD 82.5 85.5 80.8 74.7 79.7 80.1 7.7 3.1 7.0 4.0 6.1 4.2 RSD% 9.4 3.7 8.7 5.4 7.7 5.3

a High levels were designed as 20.0 g kg1 for AFB2, AFG1, AFG2, AFM1, AFM2 and 10.0 g kg1 for AFB1. b Intermediate levels were designed as 10.0 g kg1 for AFB2, AFG1, AFG2, AFM1, AFM2 and 5.0 g kg1 for AFB1. c Low level was designed as 1.0 g kg1 for all AFs.

the elute was evaporated under a stream of nitrogen gas at 40 C, (3) 10 mL aliquots were passed through Mycosep 226 Aazon+ Multifunctional cartridges at a rate of about 12 mL min1 and an aliquot of 5 mL of clean-up elution was collected, and dried by nitrogen gas. All of the above residues were re-dissolved in 1 mL acetonitrilewater mixed solution (10/90, v/v). The results showed that the recoveries of the six AFs were lowest when using Oasis HLB SPE cartridges, ranged from 69.8% to 76.7%. Compared to the HLB cartridges, the recoveries were greatly improved when the homemade mixed cartridges and Mycosep 226 Aazon+ Multifunctional cartridges were used (Fig. 2). Alternatively, home-made mixed cartridge was selected because the performance of this cartridge was similar or even better than Mycosep 226 Aazon+ Multifunctional cartridge and it was much cheaper. 3.4. Evaluation of matrix effects The stock solution was diluted with blank matrix or acetonitrilewater (10/90, v/v) mixed solution to yield analyte concentration. The signal suppression/enhancement (SSE) was calculated according to Eq. (1). SSE = 100 Peak areablank matrix Peak arealiquid standard (1)

The results showed that every AF was inuenced by the matrix and the extent was a little different. AFB1 exhibited the major matrix effect (78.2%) while AFB2 exhibited a minor effect (85.3%). To guarTable 3 Inter- and intra-day precision of the UHPLCMS/MS method (n = 3, AFs Intra-day X SD 7.78 0.41 3.36 0.16 0.77 0.07 14.58 0.72 7.82 0.26 0.84 0.02 15.30 0.24 7.92 0.22 0.82 0.04 15.36 0.24 7.58 0.22 0.79 0.07 15.36 0.34 8.24 0.28 0.78 0.07 15.02 0.12 8.04 0.16 0.78 0.05 Precision RSD (%) 5.3 4.8 9.1 5.0 3.4 2.4 1.6 2.8 4.9 1.6 3.0 8.9 2.3 3.4 9.0 0.8 2.0 6.5 Inter-day X SD 7.76 0.61 3.45 0.15 0.77 0.08 14.76 0.76 7.90 0.36 0.83 0.02 15.42 0.34 7.95 0.36 0.78 0.04 15.14 0.30 7.56 0.34 0.76 0.05 15.96 0.26 8.12 0.38 0.74 0.08 14.60 0.38 7.81 0.16 0.78 0.07

g kg1 ). Precision RSD (%) 7.9 4.4 10.4 5.2 4.6 2.5 2.3 4.6 5.2 2.0 4.5 6.6 1.7 4.7 10.9 2.7 2.1 9.0

AFB1

AFB2

AFG1

AFG2

AFM1

AFM2

66

B. Huang et al. / Analytica Chimica Acta 662 (2010) 6268

Fig. 3. MRM chromatograms of the six studied AFs in a standard solution (a) and in a musty peanut sample (b). Experimental conditions: Acquity UPLC HSS T3 column was used; the concentrations in chromatography (a) are 20 ng mL1 for AFB2, AFG1, AFG2, AFM1, AFM2 and 10 ng mL1 for the AFB1, respectively; injection volume 5 L. The linear gradient elution of water containing 0.1% formic acid and acetonitrile/methanol (50/50, v/v) proles were applied.

B. Huang et al. / Analytica Chimica Acta 662 (2010) 6268 Table 4 Presence of AFs in peanuts and their derivative products by UHPLCMS/MS analysis. Sample category Analyzed sample Positive sample AFs ( g kg1 ) Totala Fresh peanuts Musty peanuts Peanut butters
a b c d

67

AFB1 1.0 (ndd 6.4) 245.1 (0.21112.8) 5.8 (nd51.4)

AFB2 0.7 (nd3.1) 63.5 (nd301.2) 1.5 (nd14.0)

AFG1 0.2 (nd1.3) 0.03 (nd0.15) 4.5 (nd20.8)

AFG2 0.6 (nd4.1) 0.4 (nd1.3) 1.3 (nd4.5)

AFM1 nd 13.6 (nd64.7) 0.5 (nd4.2)

AFM2 nd 0.7 (nd3.6) 0.2 (nd1.8)

35 5 33

14 5 31

2.5b (0.37.4)c 323.3 (1.21482.3) 13.8 (0.795.9)

The summation of AFB1, AFB2, AFG1, AFG2, AFM1 and AFM2 levels. The mean of total AFs and individual AF levels. The values in parenthesis represent the range of the AFs detected. Not detected (below the LOD); the mean of individual AF was calculated by assuming that the level of each AF in samples, below the LOQ was equal to zero.

antee a reliable quantitation of the six AFs, matrix calibration was utilized in the present study. 3.5. Method validation 3.5.1. Linearity The calibration curves were performed in blank matrix with the concentration sequence of 0.1, 0.2, 0.5, 0.8, 1, 2, 4, 8, 12, 16, 20, 25 ng mL1 . All measurements were done in duplicate. The results showed that the calibration curves were linear over the ranges of 0.420 g kg1 for AFB2, AFG1, AFG2, AFM1, AFM2 and 0.110 g kg1 for AFB1 in crude peanut samples with the correlation coefcients (R2 ) 0.999 (Table 1). 3.5.2. Limit of detection and limit of quantitation When using a low resolution mass spectrometry, the rst condition to be satised for ascertaining the target compound presence is that the two MRM transitions produce signals distinguishable from the background ion current [22]. According to that, the limit of detection (LOD) was adopted as the concentration of a compound giving S/N = 3:1 for the second intense transition and the limit of quantitation (LOQ) was estimated considering

the most intense transition and dening it as the amount of a compound giving S/N = 10:1. They were determined by serial dilution of spiked extracts with blank matrix under the described UHPLCMS/MS conditions. Data are shown in Table 1, the LOD and LOQ levels of each AF satised the European regulations. When S/N for the most intense transition was remarkably higher than that for the second intense one, LOQ might be very similar to LOD. 3.5.3. Recovery and precision Recovery was performed employing the method of standard addition. Nine portions of the sample (2.5 g each) were spiked with the high, intermediate and low levels of each AF standard (10, 5, 1 g kg1 for AFB1, 20, 10, 1 g kg1 for AFB2, AFG1, AFG2, AFM1 and AFM2) (three samples per concentration level) while three additional portions were selected as the controls. All samples were pretreated as described in Section 2.6 and the results are summarized in Table 2. All recoveries were major than 74.7%, indicating the good accuracy of the method. Intra-day precision is listed in Table 3. All measurements were done in triplicate. The RSDs of the intra-day were in the range of 0.89.1%. Inter-day precision for each compound was also inves-

Fig. 4. The (+) ESI/MS/MS spectrum of AFM1 in standard solutions (a) and in musty peanuts (b).

68

B. Huang et al. / Analytica Chimica Acta 662 (2010) 6268

tigated with RSDs in the range of 1.710.9% (Table 3). All these data revealed that the established method had an acceptable precision. 3.6. Method application

formed in the musty peanuts due to various factors such as cultivation, harvest, storage and so on. Further study covering a large number of samples is needed to nd out the creation mechanism of AFM1 and AFM2. 4. Conclusions

Following the optimization and validation of the analytical method, it was applied to determine the contamination level of AFs in peanuts and their derivative products for human consumption. The sample solutions with high contents of AFs (the concentration level out of the linear range) were diluted with blank matrix to the concentration within the linear range. MRM chromatograms of the six AFs in a standard solution (a) and in a musty peanut sample (b) are shown in Fig. 3. The levels of total and individual AFs were determined by the established method. Among 33 peanut butters, 31 were contaminated with AFs (93.9% of incidence), ranged from 0.3 to 95.9 g kg1 (Table 4). The levels of AFB1, AFB2, AFG1 and AFG2 were detected in the range of 051.4, 014.0, 020.8, 04.5 g kg1 , respectively. 29 out of 33 peanut butters were contaminated with AFB1 (87.9% of incidence), indicating AFB1 was the most common contaminant. AFB2, AFG1 and AFG2 also frequently existed: the occurrence of AFB2 was above 60%, then was AFG1 and AFG2 (>30%). Our results were supported by previous reports that peanut was one of the most susceptible foodstuff contaminated by toxicogenic fungi producing AFs [21,23,24]. In contrast to the low levels of AFs in peanuts and their derivative products reported by other researchers [1,21], the relatively high contamination levels reported in this study might be attributed to the favorable conditions for the Aspergillus species producing AFs, emphasizing the need for regular monitoring and improved control the levels of AFs. The mean levels of AFs in fresh peanuts were obviously lower (2.5 g kg1 ) than that in peanut butters. Among 35 fresh peanuts, only 14 were contaminated with AFs, usually in low levels, indicating that AFs might be formed during the improper storage or preparation. Interestingly, AFM1 and AFM2 were found in peanut butters for the rst time. The mean levels (incidence) of AFM1 and AFM2 were 0.5 g kg1 (30.3%) and 0.2 g kg1 (6.1%), respectively. After investigating the compositions shown by the labels of the tested peanut butters, we were sure that there was not any milk derivative or by-product into the samples. However, in fresh peanuts, no sample was contaminated with AFM1 and AFM2 (Table 4). In order to nd out if AFM1 and AFM2 in peanut butters were introduced by the native state of peanuts, 5 musty peanuts were collected and tested. AFM1 and AFM2 were clearly identied by comparing the abundance ratios between the most intense transition and the less intense one in the standard solution (Fig. 3a) with those in the sample solutions (Fig. 3b). As AFM1 was always considered a major excretion product in milk of lactating animals and women exposed to dietary AFB1, daughter ion scan was also employed for a further study to ensure the existence of AFM1 in natural peanuts. The (+) ESI-MS/MS spectrum of AFM1 in a standard solution (a) and in a musty peanut (b) are shown in Fig. 4. The fact that it was possible to obtain clear fragmentation spectra in the (+) ESI-MS/MS analysis, even though it was of a relatively low abundance in the matrix, also supported the presence of the AFM1 in the sample solutions. Quantitation was performed in MRM mode. The results are shown in Table 4, 2 out of 5 musty peanut samples were contaminated with AFM1 and AFM2, with the maximal contents of 64.7 g kg1 and 3.6 g kg1 , respectively. All these data indicated that AFM1 and AFM2 might not just be carcinogenic metabolites of AFB1 and AFB2 in milk and milk products; they might also be

A reliable UHPLCMS/MS method for simultaneous determination of AFB1, AFB2, AFG1, AFG2, AFM1 and AFM2 in peanuts and their derivative products was described. After optimization of the chromatographic conditions, all analytes could be completely separated in less than 9 min, providing narrow peaks with good peak symmetry. Compared with the previous study [10], a more reliable and cost-efcient solid phase extraction-based clean-up method was developed. In combination with a higher polarity of the extraction solvent, home-made mixed cartridges showed equal or, even higher recoveries in comparison with some other columns, i.e. Oasis HLB SPE cartridges (200 mg, 6 cm3 ), Mycosep 226 Aazon+ Multifunctional cartridges. As the performance of this cartridge was similar or even better than others and it was much more economy, the developed clean-up procedure was a better choice than the standardized Mycosep method commonly used. Finally, the method was successfully applied to analyze a total of 73 peanuts and their derivative products. The results showed that most of the samples were contaminated with AFs. After identication by comparing the mass spectral data (MRM and daughter ion scan) from the sample solutions with those from standard solutions, AFM1 and AFM2 were found in the musty peanuts and their derivative products for the rst time. References
[1] J. Blesa, J.M. Soriano, J.C. Molt, R. Marn, J. Manes, J. Chromatogr. A 1011 (2003) 4954. [2] T.A. Bean, D.M. Yourtee, B. Akande, J. Ogunlewe, J. Toxicol. Tox. Rev. 8 (1989) 4352. [3] J.D. Groopman, K.F. Donahue, J. Assoc. Off. Anal. Chem. 71 (1988) 861867. [4] P. Simon, P. Delsaut, M. Lafontaine, Y. Morele, T. Nicot, J. Chromatogr. B 712 (1998) 95104. [5] IARC Monograph on the Evaluation of Carcinogenic Risk to Humans, vol. 82, IARC, Lyon, 2002, pp. 171300. [6] E.E. Creppy, Toxicol. Lett. 127 (2002) 1928. [7] M.R. Oveisi, B. Jannat, N. Sadeghi, M. Hajimahmoodi, A. Nikzad, Food Control 18 (2007) 12161218. [8] N. Sadeghi, M.R. Oveisi, B. Jannat, M. Hajimahmoodi, H. Bonyani, F. Jannat, Food Control 20 (2009) 7578. [9] M. Badea, L. Micheli, M.C. Messia, T. Candigliota, E. Marconi, T. Mottram, M. Velasco-Garcia, D. Moscone, G. Palleschi, Anal. Chim. Acta 520 (2004) 141148. [10] Y. Ren, Y. Zhang, S. Shao, Z. Cai, L. Feng, H. Pan, Z. Wang, J. Chromatogr. A 1143 (2007) 4864. [11] M.W. Trucksess, W.C. Brumley, S. Nesheim, J. Assoc. Off. Anal. Chem. 67 (1984) 973975. [12] J. Stroka, R.V. Otterdijk, E. Anklam, J. Chromatogr. A 904 (2000) 251256. [13] A.C. Manetta, L. Di Giuseppe, M. Giammarco, I. Fusaro, A. Simonella, A. Gramenzi, A. Formigoni, J. Chromatogr. A 1083 (2005) 219222. [14] B. Giray, G. Girgin, A.B. Engin, S. AydIn, G. Sahin, Food Control 18 (2007) 2329. [15] H.P. Van Egmond, S.H. Heisterkamp, W.E. Paulsch, Food Addit. Contam. 8 (1991) 1729. [16] D. Saha, D. Acharya, D. Roy, D. Shrestha, T.K. Dhar, Anal. Chim. Acta 584 (2007) 343349. [17] S.V. Reddy, D.K. Mayi, M.U. Reddy, K. Thirumala-Devi, D.V. Reddy, Food Addit. Contam. 18 (2001) 553558. [18] B. Sarimehmetoglu, O. Kuplulu, T. Haluk Celik, Food Control 15 (2004) 4549. [19] D.N. Mortimer, J. Gilbert, M.J. Shepherd, J. Chromatogr. A 407 (1987) 393398. [20] R.D. Stubbleeld, J. Assoc. Off. Anal. Chem. 70 (1987) 10471049. [21] H.S. Chun, H.J. Kim, H.E. Ok, J.-B. Hwang, D.-H. Chung, Food Chem. 102 (2007) 385391. [22] C. Cavaliere, P. Foglia, C. Guarino, M. Nazzari, R. Samperi, A. Lagan, Anal. Chim. Acta 596 (2007) 141148. [23] A. Escobar, O.S. Regueiro, J. Food Prot. 65 (2002) 219221. [24] F.A. Mphande, B.A. Siame, J.E. Taylor, J. Food Prot. 67 (2004) 96102.