NEW ANALYTICAL METHODS FOR THE DETERMINATION OF AFLATOXINS IN FOOD
NOR SHIFA BIN SHUIB
UNIVERSITI SAINS MALAYSIA
2017
NEW ANALYTICAL METHODS FOR THE DETERMINATION OF AFLATOXINS IN FOOD
by
NOR SHIFA BIN SHUIB
Thesis submitted in fulfilment for the requirements for the degree of
Doctor of Philosophy
August 2017
iii
TABLE OF CONTENTS
Acknowledgement.……….….. ii
Table of contents………... iii
List of Tables………. vii
List of Figures……… ix
List of Abbreviations………. xii
Abstrak……….. xiv
Abstract………. xvii
CHAPTER 1: INTRODUCTION……….…. 1
1.1 Introduction to Mycotoxins………... 1
1.1.2 Introduction to Aflatoxins……….…….………... 5
1.2 Health Effects of Aflatoxin Exposure………..……….…… 9
1.3 Aflatoxins Regulation……….………... 3
1.4 Analysis of AFs …….……….….. 15
1.4.1 Sampling………..……….… 16
1.4.2 Extraction……….. 16
1.4.3 Clean-up………..………….. 17
1.4.3(a) Liquid-Liquid Extraction (LLE)……….………... 17
1.4.3(b) Solid Liquid Extraction …….………... 18
1.4.3(b)(i) Solid Phase Extraction (SPE)…..…... 18
1.4.3 (b)(ii) Multifunctional Clean-up Column (MFC)………..………..….. 18
1.4.3(b)(iii) Immunoaffinity Column (IAC)..…… 20
iv
1.4.4 New Sample Preparation Methods for the
Determination of AFs ……… 22
1.4.4(a) Matrix Solid Phase Dispersion (MSPD)...………... 22
1.4.4(b) QuEChERS……….………..………….….. 23
1.4.5 Analytical Determination of AFs………..….. 24
1.4.5(a) Immunochemical methods……….……….. 24
1.4.5(b) Chromatographic Methods……….………. 24
1.4.5(b)(i) Thin Layer Chromatography (TLC)…... 25
1.4.5(b)(ii) Gas Chromatography (GC)……..…….. 26
1.4.5(b)(iii) High Performance Liquid Chromatography (HPLC)………. 27
1.4.5(b)(iv) LC-MS………..…………. 30
1.5 Problem statement………..……….….. 31
1.6 Objectives……….. 32
CHAPTER TWO: SURVEY OF AFLATOXIN M1 CONTAMINATION IN FRESH COW MILK AND HUMAN MILK IN PENANG, MALAYSIA……… 33
2.1 Introduction………. 33
2.2 Materials and Methods….………..…. 35
2.2.1 Samples……….…… 35
2.2.2 Reagents and Standard……….… 35
2.2.3 Determination of AFM1 and Tolerable Daily Intake (TDI) Calculation………..…… 36
2.2.4 Extraction and Clean-up……….….. 36
2.2.5 HPLC Conditions………. 36
2.2.6 Validation Studies……….………… 37
2.3 Results and Discussion……….…………. 38
v
2.4. Conclusions……….………….. 47
CHAPTER THREE: HIGHLY SENSITIVE METHOD FOR THE DETERMINATION OF AFLATOXIN M1 IN MILK AND DAIRY PRODUCTS USING HPLC FLUORESCENCE WITH POST COLUMN PHOTOCHEMICAL DERIVATIZATION ……….………..…… 47
3.1 Introduction………. 47
3.2 Experimental………. 49
3.2.1 Chemical and Materials……….… 49
3.2.2 Sample Preparation………..……….. 49
3.2.1(a) Liquid Milk……….. 49
3.2.1(b) Infant Formula………. 50
3.2.3 HPLC conditions……….. 50
3.2.4 Method Validation………. 51
3.2.5 Statistical Analysis………. 51
3.2.6 Application to Real Samples……….……. 52
3.3 Results and Discussion……….….. 52
3.3.1 Optimization of the Photochemical Derivatization Procedure……… 53
3.3.2 Method Validation………. 56
3.3.3 Comparison With Other Methods……….….. 58
3.3.4 Real Sample Analysis………..… 61
3.3.5 Comparison with Other Studies……….…. 63
3.4 Conclusion……….….… 64
vi
CHAPTER FOUR: IN-SYRINGE DISPERSIVE MICRO-SOLID PHASE EXTRACTION (IN-SYRINGE-D-µSPE) METHOD FOR THE HPLC-FLUORESCENCE DETERMINATION OF AFLATOXINS IN MILK,
PEANUT AND RICE………... 65
4.1 Introduction……….…. 65
4.2 Experimental ………....…….….. 70
4.2.1 Chemicals and Materials………..….… 70
4.2.2 HPLC Conditions………..… 71
4.2.3 Standard Solution Preparation for the Analysis of AFB1, AFB2, AFG1 and AFG2 in peanut and rice……… 72
4.2.4 Sample Pre-treatment……….... 72
4.2.5 Preparation of the In-Syringe Dispersive Micro Solid Phase Extraction Device..………..……….. 73
4.2.6 In-syringe Dispersive Micro-Solid Phase Extraction Method…. 73
4.2.7 Method Validation……… 76
4.2.8 Application to Real Samples……….……….... 76
4.3 Results and Discussion for Milk Samples………. 77
4.3.1 Sample Pre-treatment……….…….….. 77
4.3.2 Optimisation of In-Syringe Dispersive Micro-Solid Phase Extraction Method………... 78
4.3.2(a) Type of Syringe Filter…….………..………... 79
4.3.2(b) Type and Mass of Sorbent………..…....… 80
4.3.2(c) Vortex Speed and Time…….…………..…………...…. 82
4.3.2(d) Washing Solution and Volume…….……..………. 84
4.3.2(e) Type of elution solvent and volume…….……… 86
4.3.3 Method Validation....……….…………..……….…….….... 88
4.3.4 Analysis of Real Samples………..,………..…….... 91
4.3.5 Comparison to Previous Methods……….….……… 92
4.4 Results and Discussion for Peanut and Rice……….…..…. 97
4.4.1 Optimization of the In-Syringe Dispersive Micro Solid Phase Extraction……….…..… 97
vii
4.4.1(a) Optimization of Mobile Phase………..………. 97
4.4.2 Method Validation……….……... 99
4.4.3 Application to Real Samples………..………..………. 101
4.5 Conclusions……….……….. 104
CHAPTER FIVE: CONCLUDING REMARKS AND SUGGESTIONS FOR FUTURE WORK………..……….………… 105
5.1 Concluding Remarks………..…. 105
5.2 Suggestions for Future Studies…….……….. 109
REFERENCES………..………... 111
APPENDIX 1……….…... 131
APPENDIX 2……… 132
LIST OF PUBLICATION AND PRESENTATIONS AT CONFERENCES………..………..….……..… 133
viii
LIST OF TABLES
Page Table 1.1 Major mycotoxins and crops frequently affected by 3
mycotoxins (Reddy et al., 2010).
Table 1.2 Number of data points for the different mycotoxins in the developed mycotoxin concentration database (Perre et al., 2015). 7 Table 1.3 Legal limit range of mycotoxins according to the
Commission regulation (EC) (EC 2006) 14 Table 2.1 Performance of analytical method in raw cow milk
samples for AFM1. 39 Table 2.2 Occurrence and levels of AFM1 in fresh milk in
Penang, Malaysia. 41 Table 2.3 AFM1 contamination in raw milk in different countries. 41 Table 2.4 AFM1 contamination in human milk in different countries. 42
Table 3.1 Effect of mobile phase composition with and without post column photochemical derivatization of AFM1 (0.5 µg L-1). 55
Table 3.2 Limits of detection (LOD) and quantitation (LOQ) when subjected and not subjected to
evaporation/reconstitution before post column photochemical derivatization. 57
Table 3.3 Repeatibility of the proposed method for different milk and milk products. 58
Table 3.4 Comparison of detection methods for the determination of AFM1. 59
Table 3.5 Results for the analysis of AFM1 in milk samples using the proposed method. 62
Table 3.6 AFM1 contamination in milk and milk products in different
countries. 63
Table 4.1 Adopted parameters for the ISDµSPE in milk. 79 Table 4.2 Analytical characteristics of the proposed method. 88
ix
Table 4.3 Recoveries and precision (% RSD) for different types of milk. 90 Table 4.4 Results of samples analysed using the proposed method. 92
Table 4.5 Comparison of microextraction and QuEChERS methods for the determination of AFs. 94
Table 4.6 Analytical characteristics of the proposed method when applied to the analysis of AFs in peanuts and rice. 99
Table 4.7 Recoveries and precision (% RSD) in peanut and rice. 100
Table 4.8 Results for the analysis of AFs using the proposed method in peanut and rice samples. 102
x
LIST OF FIGURES
Page
Figure 1.1 Information notifications distributed to the relevant network members of the EU rapid alert system for food
and feed in 2005 (European Commission, 2006). 2
Figure 1.2 Factors affecting mycotoxin occurrence in the human food and animal feed chains (Patterson and Lima, 2010). 4
Figure 1.3 Structures of the aflatoxins studied. 6 Figure 1.4 Biotransformation of AFM1 and AFM2. 8
Figure 1.5 Biotransformation pathways of AFB1 leading to reactive metabolisme and biomarkers 12
Figure 1.6 General steps involved in the analysis of AFs 15
Figure 1.7 Scheme of aflatoxin sample pretreatment (clean-up and enrichment) using MFC. 19
Figure 1.8 Immunoaffinity column for the sample pretreatment (clean-up and enrichment of AFs). 21 Figure 1.9 Derivatization reactions of AFB1 (Kok, 1994). 29 Figure 2.1 Typical HPLC calibration curve for AFM1 analysis. 39 Figure 2.2 HPLC chromatograms of cow milk sample spiked
with AFM1 (0.5 ug L-1). 40
Figure 3.1 Proposed reaction for the post column photochemical derivatization of AFB1 and AFM1. 53 Figure 3.2 Typical chromatograms of AFM1 standard (0.05 µg L-1)
using methanol:water (35:65, v/v) as mobile phase (a) without post column photochemical derivatization, and (b) after post column photochemical derivatization. 55 Figure 3.3 Typical chromatograms of naturally contaminated goat
milk after post column photochemical derivatization using methanol:water (35:65, v/v) as mobile phase. 62
xi
Figure 4.1 Overview of the percentage of different food groups in the developed mycotoxin concentration database
(Perre et al., 2015). 66 Figure 4.2 Process flow for aflatoxins analysis using ISDµSPE. 75
Figure 4.3 Effects of use of precipitating solvents (0.1 M HCl, acetonitrile (ACN), acetone) and different
dilution of milk with water (1:1, 1:2, 1:3, 1:4) to precipitate protein on the recoveries of the aflatoxin standards spiked with 0.5 ng mL-1 in milk sample. 78
Figure 4.4 Extraction of AFB1, AFB2, AFM1 and AFM2 standards (0.5 ng mL-1) using different type of sorbents. Sorbent
mass; 50 mg, vortex time; 60 s, vortex speed; 2500 rpm. 81 Figure 4.5 Extraction of AFB1, AFB2, AFM1 and AFM2 standards
(0.5 ng mL-1) using different mass of sorbent. Sorbent type; C18, vortex time; 60 s, vortex speed; 2500 rpm. 82
Figure 4.6 Extraction of AFB1, AFB2, AFM1 and AFM2 standards (0.5 ng mL-1) using different vortex speeds. Sorbent
type; C18, sorbent weight; 60 mg, vortex time; 60 s. 83 Figure 4.7 Extraction of AFB1, AFB2, AFM1 and AFM2 standards
(0.5 ng mL-1) using different vortex times. Sorbent
type; C18, sorbent weight; 60 mg, vortex speed; 1200 rpm. 83 Figure 4.8 Extraction of AFB1,AFB2, AFM1 and AFM2 standards
at 0.5 ng mL-1 using different washing solution. Sorbent type; C18, sorbent weight; 60 mg, vortex time; 180 s, vortex speed; 1200 rpm. 85
Figure 4.9. Extraction of AFB1,AFB2, AFM1 and AFM2 standards at 0.5 ng mL-1 using different washing volume. Sorbent
type; C18, sorbent weight; 60 mg, vortex time; 60 s, vortex speed; 1200 rpm, washing solution; water. 85 Figure 4.10 Extraction of AFB1, AFB2, AFM1 and AFM2 (0.5 ng mL-1)
using different type of elution solvents. Sorbent type;
C18, sorbent weight; 60 mg, vortex time; 60 s, vortex speed;
1200 rpm, washing solution; water, water volume; 6 mL. 86 Figure 4.11 Extraction of AFB1,AFB2, AFM1 and AFM2 standards
(0.5 ng mL-1) using different volume of acetone. Sorbent type; C18, sorbent weight; 60 mg, vortex time; 60 s, vortex speed; 1200 rpm, washing solution; water, water volume; 6 mL. 87
xii
Figure 4.12 Chromatograms of (a) blank cow milk, and (b) cow milk spiked with AFB1, AFB2, AFM1 and AFM2 (0.5 ng mL-1). 91 Figure 4.13 Chromatograms of peanut samples in different mobile
phase (a) methanol:water (35:65,v/v, (b) water:methanol:acetonitrile (70:20:10 v/v/v) and (c) water:methanol:acetonitrile (74:13:13 v/v/v) spiked
with AFB1, AFB2, AFG1 and AFG2 (0.5 µg kg-1). 98 Figure 4.14 HPLC chromatograms from extracts of (a) peanut,
and (b) rice sample spiked with aflatoxins (5.0 µg kg-1) after ISDµSPE procedure. 101 Figure 4.15 Chromatograms of blank (a) shelled peanut, (b)
unshelled peanut, (c) powdered peanut and (d) rice after ISDµSPE procedure 103
xiii
LIST OF ABBREVIATIONS
ACN Acetonitrile AF Aflatoxin AFB1 Aflatoxin B1
AFB2 Aflatoxin B2 AFG1 Aflatoxin G1 AFG2 Aflatoxin G2
AFM1 Aflatoxin M1
AFM2 Aflatoxin M2
aw Water activity C18 Octadecysilane
DLLME Dispersive liquid liquid microextraction DON Deoxynivalenol
EU European Union FUMs Fumonisins
FLD Fluorescence detection GC Gas chromatography
HPLC High performance liquid chromatography h Hour
IARC International Agency for Research on Cancer IAC Immunoaffinity columns
LLE Liquid-liquid extraction LC Liquid chromatography L Liter
LOD Limit of detection
xiv LOQ Limit of quantification MeOH Methanol
MFC Multifunctional clean-up column MS Mass spectrometry
MSPD Matrix solid phase dispersion mL Milliliter
min Minute ng Nano gram OTA Ochratoxin A Path. Pathology ppb Parts per billion
RSD Relative standard deviation
QuEChERS Quick, easy, cheap, effective, rugged and safe s Second
SPE Solid phase extraction SPME Solid phase micro extraction ST Saturated
TFA Trifluoroacetic acid TDI Tolerable daily intake TLC Thin-layer chromatography µg Microgram
v Volume Vet Veterinary ZON Zearalenone
xv
KAEDAH ANALISIS BAHARU BAGI PENENTUAN AFLATOKSIN DI DALAM MAKANAN
ABSTRAK
Tesis ini memfokuskan kepada kaedah pembangunan dan validasi penentuan aflatoksin (AFs). Untuk permulaan, satu penyiasatan dijalankan terhadap paras aflatoksin M1 (AFM1) di dalam susu, dengan 102 sampel susu segar dan 45 sampel susu ibu telah dianalisis menggunakan kaedah piawai. Hasil analisis telah menunjukkan empat sampel telah tercemar dengan AFM1, dengan tiga sampel telah melanggar had yang dibenarkan oleh Kesatuan Eropah. Walau bagaimanapun, tiada sampel susu ibu yang telah dicemari dengan AFM1. Disamping itu, satu kaedah penerbitan yang baharu bagi penentuan AFM1 menggunakan pasca turus fotokimia telah dibangunkan. Unit terbitan fotokimia diletakkan di antara penyuntik dan alat pengesan pendarfluor (FLD). Sampel pada mulanya diekstrak dan dimurnikan dengan menggunakan turus immunoafiniti AFLATEST yang pada asalnya disasarkan untuk aflatoksin B1, B2, G1 and G2. Setelah langkah pengeringan/konstitusi dijalankan, sampel yang disuntik akan melalui unit penerbitan fotokimia dan AFM1 disinari oleh lampu cahaya lembayung (λ = 254 nm). Isyarat yang digandakan dikesan oleh FLD masing-masing pada 365 nm (pengujaan) and 440 nm (pancaran). Peningkatan gerakbalas kromatografi yang ketara (66 % peningkatan) AFM1 setelah penerbitan fotokimia telah dicatatkan. Suatu kaedah analisis bagi penentuan aflatoksin B1, B2, M1 and M2 secara serentak di dalam susu juga telah dibangunkan menggunakan teknik penyediaan yang baharu, pengekstrakan fasa pepejal mikro serakan dalam picagari (ISDµSPE) serta
xvi
kromatografi cecair prestasi tinggi (HPLC) dan FLD telah diterangkan. Kaedah penerbitan fotokimia dalam talian telah digunakan untuk meningkatkan pengesanan aflatoksin B1. Beberapa parameter seperti jenis (XAD-2, Oasis HLB dan C-18), jisim penjerap, halaju dan masa vorteks dan pelarut elusi telah dikaji. Teknik ini seterusnya divalidasi bagi sampel beras dan kacang tanah. Nilai perolehan semula yang memuaskan telah diprolehi pada julat 89.6 – 103.3%. LOQ of 0.003, 0.001, 0.1 and 0.004 µg L-1 for aflatoxins B1, B2, M1 and M2 respectively were obtained.
Tatacara di atas yang telah dibangunkan telah diaplikasikan kepada sampel sebenar.
Tiga puluh tiga sampel telah dianalisis menggunakan kaedah penerbitan fotokimia.
AFM1 telah dikesan dalam sampel susu kambing dan lembu. Walau bagaimanapun sampel yang tercemar adalah rendah berbanding had penguatkuasaan EU (0.05 µg L-
1). Daripada dua puluh sampel yang dirawat menggunakan ISDµSPE, dua sampel telah tercemar dengan AF yang di bawah had perundangan Malaysia dan Eropah.
Bagi kes kacang tanah, dua sampel telah dikesan dengan AFB1 and AFB2. Walau bagaimanapun, AF tidak dikesan dalam beras dan susu ibu.
xvii
NEW ANALYTICAL METHODS FOR THE DETERMINATION OF AFLATOXINS IN FOOD
ABSTRACT
This thesis focuses on the development and validation of new analytical methods for the determination of aflatoxins (AFs). Initially, a survey on aflatoxin M1
(AFM1) levels in milk was carried out where 102 fresh milk and 45 human milk samples were analyzed using a standard method. Four samples were contaminated with AFM1, three exceeded the European Community regulatory limit. None of the human milk samples were found to be contaminated with AFM1. Additionally, a new derivatization method for the determination of AFM1 in milk using post column photochemical derivatization was developed. The photochemical derivatization unit was placed between the injector and fluorescence detector (FLD). The samples were first extracted and clean–up using the immunoaffinity AFLATEST column originally targeted for aflatoxins B1, B2, G1 and G2. Then after evaporation/reconstitution step, the injected sample (25 µL) was passed through the photochemical derivatization unit and AFM1 was irradiated by a UV lamp (λ = 254 nm). The amplified signal was detected by the FLD at 365 nm (excitation) and 440 nm (emission), respectively.
Significant enhancement of chromatographic responses (66 % area enhancement) of AFM1 after the photochemical derivatization was found. An analytical method for the simultaneous determination of aflatoxins (B1, B2, M1 and M2) in milk using a new sample pretreatment technique, the in-syringe dispersive micro-solid phase extraction (ISDµSPE), coupled with high performance liquid chromatography (HPLC) with FLD was also described. On-line photo-chemical derivatization was
xviii
used to enhance the detection of aflatoxin B1. Several parameters such as type (XAD- 2, Oasis HLB and C-18) and mass of sorbent, vortex speed, vortex time and elution solvent were evaluated. The technique was further validated for rice and peanut samples. Satisfactory results for recovery were obtained within the range of 89.6 – 103.3%. LOQ of 0.003, 0.001, 0.1 and 0.004 µg L-1 for aflatoxins B1, B2, M1 and M2
respectively were obtained. The above developed procedures were applied to real samples. Thirty three samples were analyzed using photo chemical derivatization method. AFM1 were found in goat and cow milk. However, the contaminated samples were lower than the EU regulatory limit (0.05 µg L-1). Of the twenty samples of milk that were treated using ISDµSPE, two were contaminated with AF that were below the Malaysian and European legislation limits. In the case of peanut, two samples were detected with AFB1 and AFB2. However, AFs were not detected in rice and human milk.
1
CHAPTER ONE INTRODUCTION
1.1 Introduction to Mycotoxins
Mycotoxins are secondary metabolites produced by some fungal species that are potentially hazardous to humans, animals and crops causing illness and considerable economic losses. The word mycotoxin is derived from the Greek word,
‘mykes’ meaning mould and ‘toxicum’ meaning poison (Turner et al., 2009). Fungi and fungal spores are able to colonize and penetrate deep into the matrices of agricultural crops and produce mycotoxins during preharvest, postharvest, processing and storage stages (Bhat et al., 2010). According to the Food and Agriculture Organization of the United Nations, 25 % of the world grain supply is contaminated with mycotoxins each year (CAST, 2003). In 2005, mycotoxin recorded the highest information notification related to human health risk in the EU countries with 906 cases which constitutes 40 % from overall findings (Figure 1.1), (European Commission, 2006).
Although there are more than 300 mycotoxins that have been identified, only 30 are of much interest due to human and animal health impacts. These include aflatoxins (AFs), ochratoxin A (OTA), trichothecenes, fumonisins (FUMs), zearalenone (ZON) and deoxynivalenol (DON). Aspergillus, Penicillium, and Fusarium species are the most important genera of mycotoxin producing fungi (e.g., DON and T-2 is commonly produced by Fusarium, OTA by Aspergillus and
2
Penicillium, AFs produced by Aspergillus). Different fungal species can invade foods and feedstuffs, depending on the geographical and climate conditions.
Penicillium and Aspergillus species can grow at higher temperature and lower water activity (aw) than Fusarium. Fusarium species grow well at higher aw and lower temperature (Bhat et al., 2010).
Figure 1.1 Information notifications distributed to the relevant network members of the EU rapid alert system for food and feed in 2005 (European Commission, 2006).
Mycotoxins occur particularly in regions or countries with climates of high temperature and humidity or where there are poor crop harvesting and storage conditions, which encourage mould growth and mycotoxin development (Patterson and Lima, 2010). Major food commodities affected are cereals, nuts, coffee, spices, beans and fruits such as apple, grapes and apricots. They may also be found in beer
Others 27 %
Mycotoxins 40 %
Pathology microorganism 14 %
Veterinary medicine products 6%
Components 7%
Food
additives 6%
3
and wine that result from the use of contaminated barley and grapes during productions (Table 1.1).
Table 1.1 Major mycotoxins and crops frequently affected by mycotoxins (Reddy et al., 2010).
Mycotoxins Producing Fungi
Food
commodities
Health effects
Aflatoxins (AFs)
Aspergillus flavus and A.
parasiticus
Maize, ground nuts, rice, sorghum, tree nuts, figs
Liver lesions, cirrhosis, primary hepatocellular, carcinoma, Kwashiokor, Reye’s syndrome Ochratoxin A
(OTA)
A. Ochraceus, A.
carbonarius Penicillium verrucosum
Cereals, dried vine fruit, wine, coffee, oats
Endemic nephrophaty, urothelial tumours
Fumonisins (FUMs)
F. verticillioides, F. proliferatum
Maize, maize products, sorghum
Esophageal carcinoma
Deoxynivalenol (DON)
Fusarium graminearum, F.culmorum
Cereals, cereal products
Nausea, vomiting, abdominal pain, diarrhea, dizziness, headache
Zearalenone (ZON)
Fusarium graminearum, F.culmorum
Cereals, cereal products
Premature puberty in girls, cervical cancer
Patulin Penicillium expansum
Apples, apple juice
Damage of gastrointestinal, respiratory systems, DNA, many enzymes
Exposure of mycotoxins to humans and animals is mostly by ingestion of food products from these commodities. They can also enter the human food chain via meat or other animal products such as eggs, milk and cheese, as the result of
4
livestocks eating contaminated feed (Figure 1.2). Mycotoxins are stable although under high temperature, therefore food preparation procedures cannot be expected to remove mycotoxins safely (Sherif et al., 2009).
1 3 2
Biological factors Harvesting Factors Environmental Factors Susceptible Crop Temperature Crop Maturity + Moisture Temperature Compatible Toxigenic Mechanical injury Moisture
Fungus Insect/Bird Damage Detection/Diversion Fungus
4 Storage Temperature Moisture
Detection/Diversion
Distribution-Processing Detection/Diversion
Humans Animals Animal Products
Figure 1.2 Factors affecting mycotoxin occurrence in the human food and animal feed chains (Patterson and Lima, 2010).
5
5
1.1.2 Introduction to Aflatoxins
AFs is a group of secondary metabolites produced by fungi Aspergillus species, such as A. flavus and A. parasiticus in particular. In particular, A. flavus is common in agriculture. A. bombycis, A. ochraceoroseus, A. nomius, and A.
pseudotamari are also AF-producing species, but they are encountered much less frequently (Bennett and Klich, 2003). There are 18 AFs which have been identified (Bakirdere et al., 2012). However only a few are toxic, including aflatoxin B1
(AFB1), aflatoxin B2 (AFB2), aflatoxin G1 (AFG1) and aflatoxin G2 (AFG2), (Figure 1.3). Aspergillus parasiticus, can produce AFB1, AFB2, AFG1 and AFG2 but Aspergillus flavus can produce AFG1 and AFG2 only (Bakirdere et al., 2012).
AFs were named after their generic origin (A. flavus toxins). The word
“aflatoxin” is derived from three words: (i) the “a” that represents the Aspergillus genus; (ii) the “fla” that represents the species flavus; and (iii) the “toxin” that means poison (Bakirdere et al, 2012, Bhat et al., 2010). AFs are normally refered to the group of difuranocoumarins and classified in two broad groups according to their chemical structure; the difurocoumarocyclopentenone series (AFB1, AFB2, AFM1 and AFM2) and the difurocoumarolactone series (AFG1, AFG2). These AF groups exhibit molecular differences. For example, the B group AFs (AFB1 and AFB2) have a cyclopentane ring (pentanone derivatives) while the G group (AFG1 and AFG2) contains the six-membered lactone ring (Afsah-Hejri et al., 2011; Bakirdere et al., 2012).
6
Figure 1.3 Structures of the studied aflatoxins.
AFs contamination can occur very widely. They can be found in peanuts, corn, rice, pistachio nuts, copra, cottonseeds, egg, milk, meat and feeds (Leong et al., 2010; Ali et al., 2015; Herzallah, 2009). They may be present in any foodstuff or animal feed which can support fungal growth, although the main production has been reported in maize, peanuts, Brazil or pistachio nuts, copra and cottonseeds (Leong et
Aflatoxin B1
Aflatoxin G1
Aflatoxin B2 Aflatoxin G2
Aflatoxin M1 Aflatoxin M2
7
al., 2010; Ali et al., 2015). Perre et al. (2015) reported aflatoxins were the most analyzed mycotoxins based on database from the European community (Table 1.2).
Table 1.2 Number of data points for the different mycotoxins in the developed mycotoxin concentration database (Perre et al., 2015).
Mycotoxin No. of available data points Aflatoxin
Ochratoxin A Patulin
Deoxynivalenol Zearalenone Fumonisin T-2
HT-2 Citrinin
Ergot alkaloids Cyclopiazonic acid Alternariol
4754 612 183 40 43
9 40
0 54
0 2 0
AFs are possible to contaminate foodstuffs in the tropics and sub-tropics where high temperature and humidity are optimal for the growth of molds and production of toxins when food is growing, harvested and finally stored (Amate et al., 2010);
Bakirdere et al., 2012; Bhat et al., 2010). The main food products susceptible to aflatoxin contamination are peanuts, maize, pistachio nut, cottonseed, copra and spices (Shephard, 2009). These AFs producing fungi favour hot and humid environments where mean temperatures are about 27°C, and relative humidity ranges between 80 % and 90% (Turner et al., 2005).
8
Lactating animals and humans can produce AFM1 and AFM2 through cytochrome P450-associated enzymes in the liver. They are the metabolites (monohydroxylated derivatives) of AFB1 and AFB2 respectively resulted from consumption of contaminated feed and foods by lactating animals and humans (Figure 1.4). Then, AFM1 and AFM2 were secreted through urine and milk (Prandini et al. 2009; Zinedine et al., 2007).
Figure 1.4 Biotransformation of AFM1 and AFM2 (Mykkanen et al., 2005).
Cyctochrome
Cyctochrome
AFB1 AFM1
AFB2 AFM2
P 450 P 450
9
1.2 Health Effects of Aflatoxins Exposure
The outbreak of ‘Turkey X’ disease in England in 1960 led to the discovery of AF. The disease was traced to a mouldy groundnut (peanut) cake which was traced to Aspergillus flavus. The effect of AF is known as aflatoxicosis. Aflatoxicosis can cause fatality due to AF being consumed at higher dosage in humans and animals (Zain, 2011). The outbreak of AFs in various countries have been reported including Malaysia (Lye et al., 1995). In Malaysia, the first aflatoxicosis case in human was reported in 1988. It happened in Perak causing 13 fatality to children. It was found that the fatality was due to the consumption of noodles contaminated with up to 3 mg of AFs. Since then, extensive studies on AFs have been conducted in Malaysia. AFs have been included in the Malaysian Food Regulation Act (1985) with maximum permissible limit 35 µg kg-1 (AFB1, AFB2, AFG1 and AFG2) and had been revised to 15 µg kg-1 for commodities such as peanut and walnut which require further processing. Food commodities such as peanuts, cereals, spices, and their products are the main commodities commonly found to be contaminated with AFs in Malaysia (Leong et al., 2011, Ali et al., 2015).
Severe aflatoxicosis was reported in Kenya in 2004 (Centers for Disease Control and Prevention. 2004) causing 125 deaths. The outbreak was associated with aflatoxin contaminated maize containing more than 20 ppb of AFB1. In some areas, 3
% to 12 % of the contaminated samples had AFB1 levels reaching up to 8000 ppb.
10
AFs are not only toxic, but also mutagenic, carcinogenic and teratogenic compounds (Aycicek et al., 2005; Bakirdere et al., 2012; Bhat et al., 2010). The terminal furan moiety of AFs is crucial for determining the biological activity. Thus, the order of toxicity follows the order: AFB1, AFG1, AFB2 and AFG2 (Agag, 2004;
Espinosa-Calderón et al., 2011). According to the International Agency for Research on Cancer (IARC), AFs are classified as carcinogens. Among them, AFB1 is the most toxic, causing damage such as hepatitis, hemorrhage, edema, immunosuppression and hepatic carcinoma (Nakai et al., 2008). AFB1 has been reported to be the most powerful source of natural carcinogen known in mammals (Fallah, 2010; Hussain et al., 2008). Therefore AFB1 has been classified by IARC as Group 1 human carcinogen. AFB1 constituted approximately 90 % of the aflatoxin contaminated foodstuffs (Amate et al., 2010). The main target organ of AFB1 is the liver (Kamika and Takoy, 2011). Synergistic effects of AFB1 with viral infections of hepatitis B or C can cause liver cancer (Bakirdere et al., 2012). Endemic diseases, such as Kwarshiorkor and Reye's syndrome (damage to the liver and kidney), are also caused by AFs (Zöllner and Mayer-Helm, 2006).
AFB1 is possible to form metabolites as it possesses unsaturated bond at the 8,9 position on the terminal furan ring allowing reactive form at that position. The metabolite produced by AFB1, AFB1-8,9-epoxide (AFB1-epoxide) was produced by cytochrome P450 (CYP) enzymes at that reactive site and can bind to DNA strands to produce DNA adduct known as AFB1-N7-Gua and AFB1-FAPy adduct. Figure 1.5 shows the biotransformation pathways for AFB1.
11
AFB1 is carcinogenic, as the metabolite produce by AFB1 could cause DNA damage (Wild and Turner, 2002). AFB1 metabolite could binds to DNA at the guanine base in liver cells, causing mutation by corrupting the genetic code that regulates cell growth. For instance, AFB1 could induce the mutation through the G→T transversion at the p53 gene, which is the gene responsible for DNA repair (Levy et al., 1992; Aguilar et al., 1993). Then, out of control cells grow into tumours that could cause cancer (Heidtmann-Bemvenutti et al., 2011). In addition, AFB1-8,9- exo-epoxide could also react with ribonucleic acid (RNA) and proteins causing disruption of cells function.
12
AFB1-8,9-Epoxide
Figure 1.5 Biotransformation pathways of AFB1 leading to reactive metabolisme and biomarkers
Mercapturic acid AFB1-Diol
GSH Transferase Phase 2
H2O Albumin adduct
AFB1 N7 Gua adduct AFB1- FAPy aduct
Phase 1 P450 1A2,3A4
Fungi Corn, peanut
Human AFM1
P450 IA2
13
1.3 Aflatoxins Regulation
In order to minimize human health risks, maximum levels were set in food and feed for many countries (FAO, 2004). The hazardous nature of AF to humans and animals has necessitated the need for the establishment of control measures and tolerance levels by national and international authorities. In general, different countries have different regulations for AFs. The general trend is that industrialized countries usually set lower tolerance levels than the developing countries, where most of the susceptible commodities are produced (Abrar et al., 2012). In some African and Asian countries, no maximum levels are set until present. However, there are some attempts to harmonize the maximum levels (Codex, 2007). Factors influencing mycotoxin regulations include availability of toxicity data, availability of data on the occurrence in different commodities, survey analytical data, methods of sampling and analysis, and trade contacts with other countries (Egmond, 2004).
In 1973, the European Economic Community (EEC) established legislation on maximum permitted levels of AFBl in different types of feedstuffs. The
legislation has been frequently amended since then (Abrar et al., 2015). The European Community established 4 µg kg-1 total AFs in food for human
consumption which is the maximum acceptable limits in the EU, the strictest in standard worldwide. The Food and Drug Administration (FDA) of the USA, had proposed a tolerance level of 20 µg kg-1 which was lowered from the first tolerance level of 30 µg kg-1 of total AFs (AFBl, AFB2, AFG1 and AFG2) (Abrar et al., 2015).
14
Most of the research that has been recently carried out has been focused upon the study of AFB1, as it is the most carcinogenic mycotoxin for human beings. For instance, the maximum residue limit fixed by the EU for AFB1 in rice and its products is 2 μg kg-1. Table 1.3 shows the legal limit for common mycotoxins.
Table 1.3 Legal limit range of mycotoxins according to the Commission regulation (EC) (EC 2006).
Mycotoxin Legal limit range (µg kg-1) AFs
AFB1
AFM1
4.0 (dried fruit and nuts), 15.0 (groundnuts)
0.1 (dietary, processed cereal based and infants food), 8.0 (groundnuts)
0.025 (infant and dietary foods), 0.05 (Milk) Fumonisins 200 (processed mixed based foods)
400 (unprocessed maize)
Ochratoxin A 0.5 (processed cereal based and infant foods) 10 (grape juice and instant coffee)
Deoxynivalenol 0.5 (processed cereal based and infant foods) 1,750 (Unprocessed maize)
Zearalenone 20 (processed cereal based and infant foods) 350 (unprocessed maize)
Patulin 10 (apple juice and solid apple products and baby foods) 50 (spirit drinks and fruit juice)
15
1.4 Analysis of AFs
As the regulation of AFs is at very low level (ppb), the determination of AFs is very challenging. A typical analytical procedure for the determination of AFs generally follow several steps as shown in Figure 1.6.
Figure 1.6 General steps involved in the analysis of AFs
Sampling
Extraction
Report Clean-up
Detection Lot
Derivatization
16
1.4.1 Sampling
AFs are normally not produced homogeneously in crops, but are the results of fungal growth at specific units, such as maize kernels and groundnut pods. A single lot of the product will contain hot spots of contamination (Reiter et al., 2009). The proper selection of a sample from the lot under study and the subsequent steps undertaken to produce a portion for analysis is crucial for the determination of AFs.
The number of contaminated samples may be very low, but the contamination level within a particle can be very high (Egmond et al., 2007). To obtain the same representativeness for batches of food products with large particle sizes, the weight of the incremental sample taken has to be larger than in cases of batches with smaller particle size. Commission Regulation No. 401/2006 regulates the number of incremental samples to be taken from different places of a lot depending on the weight of the entire lot. This may result in rather large aggregate samples, up to 30 kg in the case of AFs.
1.4.2 Extraction
The detection and quantification of AFs in food samples require an efficient extraction step. AFs are generally soluble in polar protic solvents such as methanol, acetone, chloroform, and acetonitrile. Thus, the extraction of AFs involve the use of
17
these organic solvents such as either methanol or acetonitrile or acetone mixed in different proportions with small amounts of water (Bertuzzi et al., 2012).
Several studies exploring the extraction efficiency of different organic- aqueous solvents have been carried out on the commonly contaminated matrices (Stroka et al., 1999; Arranz et al., 2006; Gallo et al., 2010) and different results have been reported.
1.4.3 Clean-up
As food is a complex matrix, therefore clean-up step is necessary. Various clean-up methods are available for AFs analysis. The details are next discussed.
1.4.3 (a) Liquid-Liquid Extraction (LLE)
LLE is a well known and well established clean-up technique. It is based on the partition between two immiscible phases, one of which contains the analyte. The analyte then migrates into the other phase until an equilibrium has been reached. This step can be performed several times with fresh solvent in order to extract the analyte quantitatively. In this way, the analytes can be concentrated in a solvent (e.g., by rotary evaporation) and interferences can be removed. However, the method is used less frequently nowadays, because it is labour intensive and large volumes of (sometimes chlorinated) solvents are needed (Reiter et al., 2009). Contamination and sample losses are also common due to adsorption to glasswares. LLE has several
18
drawbacksincluding contamination, sample losses due to adsorption to glasswares and difficult to be automated. Therefore, the method is now often replaced by alternative techniques such as solid phase extraction (SPE).
1.4.3 (b) Solid Liquid Extraction
1.4.3 (b) (i) Solid Phase Extraction (SPE)
SPE is a variation of liquid chromatographic technique that uses a solid phase and a liquid phase to isolate analyte from a solution. The column contains different packing materials (e.g., silica gel, C18 (octadecylsilane), florisil, phenyl, aminopropyl, ion exchange materials, both anionic and cationic, and molecular imprinted polymers), (Abdel-Azeem et al., 2015; Krska et al., 2008). In general, the procedure involving several steps including sample loading into the column, retaining of the analyte, washing away impurities and finally analyte elution with organic solvents. The SPE technique has several advantages such as reduction in the amount of hazardous solvent used per analysis, shorter analysis time, and can be automated.
1.4.3 (b) (ii) Multifunctional Clean-up Column (MFC)
MFC, a variant of SPE, has been developed for one step clean-up of AF. As an example, the MycoSep column is pushed into a test tube (containing the sample), forcing the sample to filter upwards through the packing material of the column. The
19
interferent will be retained in the column. On the other hand, the purified extract containing the AF of interest passes through a membrane (frit) to the surface of the column (Khayoon et al., 2010). The column has a long shelf-life because it contains no biological reagents, and can be stored at room temperature. However, unlike immunoaffinity columns, the MFC cannot concentrate the analyte during the clean- up procedure, and also the recovery may vary depending on the complexity of the sample. The extracts are subjected to a clean-up step to further purify the sample prior to the analytical determination.
MFC offers several advantages over other clean-up procedures (e.g., speed, solvent efficiency, and in some cases, increased recovery). The MFC also allows quick sample purification. Another advantage is that time-consuming rinsing steps are not required as in SPE (Khayoon et al., 2010). In addition, almost all analytical interfering substances are retained on the column, while the AFs are not adsorbed on the packing material. Sorbents used in MFC include styrene-dibenzene polymer. The mechanism of retention of MFC is shown in Figure 1.7.
Sample loading Elution of aflatoxins
Figure 1.7 Scheme of aflatoxin sample pretreatment (clean-up and enrichment) using MFC.
Matrix interference Sample
Sorbent
Aflatoxins
20
1.4.3 (b) (iii) Immunoaffinity Column (IAC)
IAC is the current recommended method for the purification and concentration of AFs (Scott and Trucksess, 1997; Ma et al., 2013) before their determination using high-performance liquid chromatography (HPLC). IAC employs the high specificity and reversibility of binding between an antibody and antigen to separate and purify target analytes from matrices (Shelver et al., 1998).
The principles of the IAC is based on an antibody (polyclonal or monoclonal) that is able to recognize the analyte and the sorbent is immobilized onto a solid support such as agarose or silica in phosphate buffer, contained in a column. There are several steps involved in the clean-up procedures (Figure 1.8):
(a) Condition. The column is initially conditioned with phosphate buffered saline (PBS) at room temperature.
(b) Loading of the sample. Extract samples are applied to the IAC at slow flow rate (2-3 mL min-1) either by gravity or vacuum system. Then, AF contained in the sample will binds to the antibody and retained in the IAC. In order to avoid damage by organic solvents and reduce interference to the antibody, the applied sample must be in aqueous conditions. AFs determination based on IAC technique requires extraction using a mixture of methanol-water (8:2 v/v) (Stroka et al., 1999; Lee et al., 2004) as methanol has less negative effect on antibodies compared to other organic solvents such as acetone and acetonitrile.
21
(c) Washing. The column is washed with water or phosphate buffered saline solution to remove impurities.
(d) Elution. Elution by organic solvents such as acetonitrile and methanol will break the AF- antibody bond. Thus releasing AF from the antibody and eluted from the column. Then, the eluate containing AF is then derivatized or can be directly measured by instruments such as HPLC.
The advantage of immunoaffinity column is the high selectivity due to the antigen- antibody interaction that is made possible by the use of the coating of the support by the antibody. The antibody is obtained from serum of animals that had been injected with the antibody. However this sorbent is expensive, single use and have limited shelf life.
Loading Washing Elution
Figure 1.8 Immunoaffinity column for the sample pretreatment (clean-up and enrichment of AFs).
Aflatoxins
22
1.4.4 New Sample Preparation Methods for the Determination of AFs
As AFs are normally present at low concentration levels in foods, a suitable sample preparation procedure is required to extract and concentrate them prior to instrumental analysis. Conventionally, LLE and SPE were used, as previously described in section 1.4.3. Lately, there is an increasing effort to reduce sample volume, analysis time and cost, eliminating hazardous (including halogenated) solvents and chemicals and to automate the analysis to reduce the workload. In this section, emerging techniques for the determination of aflatoxins such as microextraction methods (e.g., Matrix solid phase dispersion, QuEChERS) are discussed.
1.4.4 (a) Matrix Solid Phase Dispersion (MSPD)
The MSPD technique was introduced in 1989 by Barker et al. (1989). It is a simple and cheap sample preparation method involving simultaneous disruption and extraction of solids, semi-solids and highly viscous samples. Sample and sorbent material are mixed homogenously, then packed in a catridge and elution process is performed (Rubert et al., 2011). The main advantages of the MSPD technique are that it requires only small amounts of sample and solvents, is rapid, inexpensive, and can be carried out under mild extraction conditions (room temperature and atmospheric pressure), providing acceptable yields and selectivity. However Capriotti et al. (2010) reported this technique is difficult to be automated.
23
Determination of AFs using MSPD is limited to a few reports which include peanuts (Blesa et al., 2003), olive oil (Cavaliere et al., 2007), hazelnuts (Bacaloni et al., 2008) and pistachio nuts (Manoochehri et al., 2013).
1.4.4 (b) QuEChERS
Anastassiades et al. (2003) developed an original analytical method combining the extraction/isolation of pesticides from food matrices and clean-up of the extract. This technique involves micro-scale extraction using acetonitrile and purifying the extract using dispersive SPE. The term QuEChERS refers to quick, easy, cheap, effective, rugged, and safe. QuEChERS-based methodologies have been applied for the extraction of pesticides with a wide range of physico-chemical properties from different samples (Anastassiades et al., 2003; Lehotay et al., 2010).
In addition, it has been employed for other compounds including aflatoxins from various food matrices (Desmarchelier et al., 2014; Cunha et al., 2010., Romero- Gonzalez et al., 2011).
However, the QuEChERS method is not suitable for the extraction of AFs at very low levels (ppb) and requires additional clean-up methods such as SPE in order to pre concentrate the AFs (Aguilera-Luiz et al., 2011; Desmarchelier et al., 2014).
Furthermore, AFs analysis by using QuEChERS in various matrices frequently employs LC-MS/MS (Michlig et al., 2015 ; Sartori et al., 2014).
24
1.4.5 Analytical Determination of AFs
1.4.5 (a) Immunochemical Methods
The availability of reproducible and sensitive methods for the screening of foodstuffs including AF is essential. Specific antibody tests based on the affinities of monoclonal or polyclonal antibodies for immunoassay methods such as ELISA may be well suited for the rapid, routine diagnostic application for the detection of AFs (Krska et al., 2008). However, ELISA also could provide false positive results possibly due to cross reactions and interferences from the sample matrices (Blesa et al., 2003).
Lateral-flow immunochromatographic assay combines chromatography with immunoassay. It offers the advantages of simple, cheap and time-saving, requiring only a simple extraction step (Krska and Molinelli, 2009). However this technique does not enjoyed widespread acceptance as it is more suitable for screening.
1.4.5 (b) Chromatographic Methods
To date, chromatographic methods are the most widely used analytical method for AFs determination. In general, chromatography involves interaction between mobile and stationaty phases. Separation of AFs are based on differential partitioning and distribution of AFs between the mobile and stationary phases. The chromatography methods available for the analysis of AFs are thin-layer
25
chromatography (TLC), high-performance liquid chromatography (HPLC), liquid chromatography mass spectrometry (LC-MS) and gas chromatography (GC).
1.4.5 (b) (i) Thin-Layer Chromatography (TLC)
TLC is one of the important chromatographic technique and became the method of choice by the Association of Official Analytical Chemists (AOAC) since 1990. The stationary phase is made from plastic or glass coated with adsorbent material such as silica, alumina or cellulose. The mobile phase is a mixture of organic solvents, carrying the sample through the stationary phase. The retention of AFs are based on the differences in solubility between the two phases. The molecular structures and interaction plays an important role in determining the distribution of AFs either attract to the stationary or retain in the mobile phase. TLC allows fast and effective separation of AFs.
Quantification at levels required for naturally contaminated food samples by TLC is made possible as AFs are naturally fluorescent compounds and are easily observed under long wavelength UV light (Otta et al., 2000). Quantification is done by visual comparison of the intensities of sample spots to that of standards. TLC has been widely used in the determination of AFs in different foods. The advantage of using the TLC method is that it can detect several types of mycotoxins in a single test sample. Several disadvantages limit the usage of TLC such as the need for pretreatment of sample, and the long analysis time. In addition, TLC lacks precision
26
due to accumulated errors during sample application, plate development, and plate interpretation.
A number of improvements over the conventional TLC analysis have been introduced and applied to AFs analysis such as over pressured-layer chromatography, two dimensional TLC and high-performance thin-layer chromatography (HPTLC) (Nawaz et al., 1995; Otta et al., 2000).
HPTLC is an automated form of TLC and has since overcome the problems associated with the conventional TLC techniques through automation of sample application, development, and plate interpretation. It is not surprising that currently HPTLC is one of the most efficient and precise methods in aflatoxins analysis.
Nevertheless, the requirement for skilled operators, the costs of the equipment coupled with its bulkiness, and the extensive sample pretreatment limit the HPTLC to the laboratory and thus it is not suitable in field situations.
1.4.5 (b) (ii) Gas Chromatography (GC)
In GC analysis, the stationary phase normally consists of inert particles coated with a layer of liquid and is normally confined to a long stainless steel or capillary column, which is maintained at appropriate temperature. The injected sample is vaporized into the gaseous phase and carried through the stationary phase by a carrier gas (mobile phase). The different chemical constituents in the sample will distribute themselves between the mobile and the stationary phases. Analytes of
27
the samples mixture with higher affinity for the stationary phase are retarded in their movement through the column, while those of low affinity pass through the column less impeded. Each analyte have a specific partition coefficient, which, in turn, will govern its rate of passage through the column. Detection is carried out either by using a flame ionization detector (FID) or an electron capture detector (ECD) or mass spectrometer (MS). However, in order to be detected, AFs need to be derivatized due to their non volatility in nature. Therefore, GC is less common for the determination of AFs due to the existence of other superior chromatographic methods.
1.4.5 (b) (iii) High-Performance Liquid Chromatography (HPLC)
HPLC is widely used for the determination of AFs and suitable for routine analysis. It consists of a stationary and mobile phases. The stationary phase are usually made of stainless steel containing packing material such as C18, C8 and etc.
The mobile phase contains mixture of aquous and organic solvents passing through the column. Distribution of the analytes in the column are based on different affinities between mobile and stationary phase resulting analyte separation.
The separated analytes can be detected by using detectors such as ultra violet (UV), diode array detector (DAD) and fluorescence (FL) detector being the most common for AFs analysis. Reversed phase HPLC (RP-HPLC) is by far the most commonly use chromatographic technique compared to the normal phase due to its
28
green features. Organic solvents used in RP-HPLC such as methanol and acetonitrile are less toxic compared to dichloromethane, chloroform which are used in normal phase HPLC. Moreover, both methanol and acetonitrile are compatible with water as mobile phase in RP-HPLC. Therefore RP-HPLC is the preferred choice for AFs analysis. However due to its low fluorescence properties, AFB1 and AFG1 suffers low sensitivity compare to AFB2 and AFG2 when using RP-HPLC. In order to enhance the sensitivity of AFB1 and AFG1, chemical derivatization become necessary.
AFB1 and AFG1 can be derivatized by using trifluoroacetic acid (TFA) by hydrolyzation to produce AFB2a and AFG2a respectively, a kind of hemiacetals which have strong fluorescent properties. In the case of derivatization of AFB1 and AFG1 by using halogens such as bromine and iodine, highly fluorescent AFB1 and AFG1 derivatives of these halogens were formed respectively. As an illustration, the derivatization reactions of AFB1 with trifluoroacetic acid (TFA) and halogens are presented in Figure 1.9.
Papadopoulou-Bouraoui et al. (2002) reported postcolumn derivatization by comparing two postcolumn derivatization methods for the determination of AFs, by fluorescence detection after liquid chromatographic separation. The results showed that both bromination and irradiation by UV light were suitable for the determination of AFs in various foods and animal feed matrices and both produced comparable results for fluorescence (FLD) amplification and repeatability. The fluorescence of
29
AFB1 and AFG1 were significantly enhanced after derivatization reaction either by bromination or by irradiation by UV light.
HPLC provides fast and accurate AFs determination in a relatively short time and amenable to automation. Detection as low as 0.1 ng kg-1 using FLD has been reported. However, the disadvantage of using HPLC is the requirement of vigorous sample purification using immunoaffinity columns. In addition, HPLC requires tedious either pre- or postcolumn derivatization procedure to improve the sensitivity of AFB1 and AFG1.
TFA derived AFB1
TFA
30 min, 50°C
I2
40 s, 60°C
I2 derived AFB1
Br2 derived AFB1
Br2 4s, 20°C (1)
(2)
(3)
Figure 1.9 Derivatization reactions of AFB1 (Kok, 1994).
30
1.4.5 (b) (iv) LC-MS
To overcome the challenges associated with derivatization procedure in AF analysis, a modification of the HPLC method, whereby the HPLC is coupled to mass spectrometry, has been made. Unlike HPLC, the need for chemical derivatization in LC-MS is eliminated.
The LC-MS uses small amounts of sample to generate structural information and exhibits low detection limits. A typical LC-MS system is equipped with an autosampler, the LC system, the ionization source and a mass spectrometer (MS).
Electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) are the two most common ionization sources that can be used as the interface between LC and MS. The ionization for both ESI and APCI occur at atmospheric pressure. So these sources are often referred to as atmospheric ionization (API). LC- MS has become a very valuable analytical tool for the determination of AFs and their metabolites (Rubert et al., 2011; Sartori et al., 2015). LC-MS provides decisive advantages in performing identification as well as determination of analytes at low levels.
However, the analyte signals are susceptible to signal suppression or enhancement known as matrix effect due to the co-eluting matrix components with the analyte of interest (Sartori et al, 2015).
31
1.5 Problem Statement
Sample preparation remains the bottle neck in AF analysis. Thus, new approaches in sample preparation should be given top priority. Current analytical methods also require elaborate sample clean-up to remove possible interferents before the analysis. Clean-up by using immunoaffinity column (IAC) is commonly used (Guo et al., 2016) due to the good selectivity provided by the antigen-antibody interaction between the analyte and the immobilised reagent on this type of sorbent.
However, IAC has inevitable drawbacks of being relatively expensive, is designed mostly for single analyte, single use and have a rather short shelf-life. Matrix contaminants can also mask the antibody and analyte binding site, thus severely reducing the extraction efficiency (Castegnaro et al., 2006). All these points have been discussed on page 21. As mentioned earlier, another common sample preparation technique, SPE, requires multiple-steps and is laborious. Furthermore, there is very little data on the levels of AFM1 in Malaysian food, especially in milk.
32
1.6 Objectives
Thus, the main objectives of the studies were:
i. To obtain baseline data on the levels of AFM1 in milk products consumed in Malaysia using the standard method and define the total dietary intake of the Malaysian population.
ii. To explore new strategies such as online derivatization and sample preparation methods such as In-syringe Dispersive Micro SPE to meet the demands in mycotoxin analysis.
iii. To apply the developed and validated methods to the analysis of AFs in various matrices such as milk, peanut and rice.
