Nitrate, Nitrite and Nitrosamine: Contents and Analyses In

Home , Nitrosamine

Nitrate, nitrite and nitrosamine: contents and analyses in selected foods; effect of vitamin C supplementation on N-nitrosodimethylamine formation in humans; and an investigation of natural alternatives to nitrites as preservatives in cured meat products

James Hsu, M. Sc

Supervisor: Dr. Jayashree Arcot Co-supervisors: Dr. N. Alice Lee and Dr. Julian Cox

A Thesis Submitted to the School of Chemical Sciences and Engineering In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

The University of New South Wales October 2009 Acknowledgement

Research requires a thirst for knowledge and the determination and dedication to match. In addition, a network of people including supervisors, family, friends, colleagues and laboratory technicians all works as a team to give advice and support when need it. That being said, I would like to give my gratitude to my main supervisor Dr. Jayashree Arcot for her consistent and helpful supervision throughout my PhD candidature. I would also like to extend my gratitude to Dr. Alice Lee and Associate Professor Dr. Julian Cox for their expertise in specific parts of my research. Big thanks goes to my close friends and colleagues for lending an ear or a shoulder whatever the occasion calls for, or simply by hanging out with me to save my sanity at times. The laboratory staffs are the glue that holds everything together, without their experience and knowledge the lives of any PhD candidates would be made a lot harder. Since I started my PhD in 2004, Eileen and Evyone had helped me in many ways, but special thanks go to Camillo for his dedication, professionalism and support that just made it all worthwhile. All this could only happen with the financial support from Bruce Veness Chandler Award, and to my mum Stephanie and my siblings Patricia and Frank for showing their love in their own distinct ways. Finally but not least, my gratitude to my late father Michael and my grandmother for their early upbringing that made me the man I am today.

2 Abstract Dietary nitrite intake has been implicated in numerous gastrointestinal cancers in humans due to the formation of a group of carcinogens called N-nitroso compounds. The need to estimate their intake is vital in establishing at risk population and to monitor and perhaps one day manage their dietary intake. This is the first study to estimate nitrate and nitrite in selected vegetables, cured and fresh meat in Australian food supply using ion-paired reversed-phased HPLC. Nitrite content in meat products analysed ranged from 0 to 83.9 mg/kg in medallion beef and Frankfurt, respectively; nitrate content ranged from 18.7 mg/kg in minced beef to 142.5 mg/kg in salami. The nitrite content was below the maximum limit set by the Food Standards Australia and New Zealand. Nitrate content in selected vegetables ranged from 123 to 4850 mg/kg in Iceberg lettuce and English spinach, respectively; only minimal nitrite at 20 mg/kg was present in Gai choy, which was most likely due to bacterial contamination during storage. Based on the food consumption pattern of Australians, the dietary nitrite and nitrate intake from bacon were 1.51 and 3.42 mg per capita per day, which was below the Adequate Daily Intake set by the European Union Scientific Committee for food in 1995. Taking into considerations of oral nitrate reduction to nitrite and the endogenous nitrate formation, the upper extreme of dietary nitrite and nitrate intake in Australians were 44 and 2.4 times over the ADI, respectively. However, this does not take into effect of other dietary promoters and inhibitors. Eighteen healthy human volunteers were put on a low nitrate, nitrite and antioxidant diet for three days during which they were fed one serving of cured meat with and without 500 mg of vitamin C. Using GC- MS, N-nitrosodimethylamine was not detected in the urine before or after vitamin C supplementation, suggested that a diet low on nitrate and nitrite cannot produce NDMA and thus may reduce the risk of developing gastrointestinal cancers. Different extraction methods and combination of herbs and spices were demonstrated in vitro to show inhibition against B. cereus, Escherichia coli, Listeria monocytogenes, Salmonella Enteritidis and Staphylococcus aureus. In addition, autoclaved turmeric powder at 0.3 % (w/v), hot water extracted turmeric with ginger at 0.5 % and rosemary at 1.0 % showed growth inhibition against Clostridium sporogenes, which was used as a surrogate for Clostridium botulinum. The use of these combinations of herbs and spices may replace or at least reduce the use of nitrite as a preservative in cured meat products to prevent botulism and reduce dietary nitrite intake.

3 Table of Contents

Acknowledgement ...... 2 Abstract ...... 3 Table of Contents ...... 4 List of Figures ...... 8 List of Tables ...... 10 List of Abbreviations ...... 12 Chapter 1 ...... 15 Introduction ...... 15 1.1 Experimental aims ...... 16 1.1.1 Nitrate and nitrite in Australian sourced vegetables and cured meat and their estimated dietary intake by Australians ...... 16 1.1.2 Effects of vitamin C supplementation on N‐nitrosodimethylamine formation in healthy human volunteers ...... 16 1.1.3 Anti‐Clostridia properties of herbs and spices and possible substitution for sodium nitrite ...... 16 Chapter 2 ...... 18 Nitrate and nitrite ...... 18 2.1 Dietary sources of nitrates and nitrites ...... 18 2.1.1 Plant food ...... 18 2.1.2 Cured meat ...... 20 2.1.3 Water ...... 21 2.2 Nitrate, nitrite and its derivative ...... 22 2.2.1 Mechanisms of nitrate reduction to nitrite ...... 22 2.2.2 Mechanisms of nitrite reduction to nitric oxide ...... 22 2.2.3 Metabolism of nitrate, nitrite and nitric oxide ...... 23 2.2.4 Physiological roles of nitrate, nitrite and nitric oxide ...... 26 2.2.5 Immunological effects and disease ...... 29 2.3 Toxicity and Metabolism ...... 30 2.3.1 Nitrate and nitrite ...... 30 2.4 Epidemiology studies ...... 34 2.4.1 Gastric cancer incidence ...... 34 2.4.2 Gastric cancer risk factors ...... 38 2.4.2.1 Carbohydrate...... 39 2.4.2.2 Saturated fat ...... 39 2.4.2.3 Alcohol ...... 40 2.4.2.4 Salt, pickled and smoked food ...... 40 2.4.2.5 Cooking method ...... 40 2.4.2.6 Inappropriate food storage ...... 40 2.4.2.7 Helicobacter pylori infection ...... 41 2.4.2.8 Cigarette smoking ...... 41 2.4.2.9 Low consumption of fruits and vegetables ...... 42 2.4.2.10 Others...... 43 2.4.3 Estimated dietary nitrate and nitrite intake ...... 43 2.4.3.1 Asian‐pacific regions ...... 43

4 2.4.3.2 United States and United Kingdom ...... 44 2.4.3.3 Rest of Europe ...... 45 2.4.3.4 Summary ...... 45 2.4.4 Recommended dietary nitrate and nitrite intake ...... 47 2.4.4.1 International ...... 47 2.4.4.2 Food Standards Australia and New Zealand (FSANZ) ...... 47 2.5 Determination of nitrate and nitrite ...... 49 2.5.1 Colourimetric and spectrophotometric ...... 49 2.5.2 Gas chromatography ...... 51 2.5.3 High performance liquid chromatography ...... 51 2.5.4 Ion chromatography ...... 53 2.5.5 Capillary electrophoresis ...... 55 2.5.6 Electronic sensors ...... 56 2.5.7 Others ...... 56 2.5.8 Food ...... 57 2.5.9 Biological fluids ...... 60 2.6 Food Industry and Regulations ...... 65 2.6.1 Functions of nitrate and nitrite in meat‐based products ...... 65 2.6.2 Functions of ascorbate and erythorbate in cured meat ...... 67 2.6.3 International food safety ...... 67 2.6.4 Regulatory agencies and the government ...... 68 2.7 Determination of nitrate and nitrite in food and their estimated dietary intake ...... 69 2.7.1 Introduction ...... 69 2.7.2 Materials and Methods ...... 70 2.7.2.1 Reagents ...... 70 2.7.2.2 Food samples ...... 70 2.7.2.3 Apparatus ...... 71 2.7.2.4 Standards ...... 71 2.7.2.5 Sampling and extraction ...... 71 2.7.2.6 Statistical test ...... 72 2.7.3 Results and discussion ...... 72 2.7.4 Conclusion ...... 82 2.8 References ...... 84 Chapter 3 ...... 96 Nnitroso compounds and nitrosation inhibitors ...... 96 3.1 Dietary sources of Nnitroso compounds ...... 96 3.2 Nnitroso compounds ...... 96 3.2.1 N‐nitrosamines ...... 98 3.2.2 Nitrosation of amines and amides ...... 100 3.3 Estimated dietary nitrosamine intake ...... 102 3.4 Epidemiology of cancer risks ...... 104 3.5 Nnitrosation inhibitors ...... 105 3.5.1 Ascorbic acid ...... 105 3.5.1.1 Ascorbic acid chemistry and physiological roles ...... 107 3.5.1.2 Ascorbic acid interactions with nitrate, nitrite and its derivatives ...... 108 3.5.1.3 Kinetics and mass transfer of ascorbic acid and nitrosation ...... 110 3.6.2 α–tocopherol ...... 112 3.6.3 Fibre ...... 113 3.6.4 Polyphenols and phytochemicals ...... 113 3.6.5 Others ...... 114 3.7 Determination of Nnitroso compounds ...... 116 3.7.1 Volatile nitrosamines ...... 116 3.7.2 Non‐volatile nitrosamines ...... 121

5 3.7.3 Non‐volatile nitrosamides ...... 122 3.7.4 Nitrosamine extraction from food ...... 123 3.7.5 Nitrosamines in biological fluids ...... 124 3.8 Determination of vitamin C ...... 126 3.8.1 Food ...... 128 3.8.2 Biological fluids ...... 130 3.8.3 Extraction ...... 130 3.8.4 Detection ...... 131 3.9 Effects of vitamin C supplementation on Nnitrosodimethylamine formation in healthy human volunteers on a nitrate restricted diet with cured meat ...... 133 3.9.1 Introduction ...... 133 3.9.2 Materials and methods ...... 134 3.9.2.1 Vitamin C Analysis ...... 134 3.9.2.3 Form of Vitamin C used in the experiment ...... 134 3.9.2.4 Vitamin C in urine ...... 135 3.9.2.5 Nitrate and nitrite analysis ...... 135 3.9.2.6 Analysis of meat ...... 135 3.9.2.7 Analysis of saliva ...... 135 3.9.2.8 Analysis of urine ...... 136 3.9.2.9 Quantification of N‐nitrosodimethylamine in meat and urine ...... 136 3.9.2.10 Meat ...... 136 3.9.2.11 Urine ...... 136 3.9.2.12 Human trial ...... 136 3.9.2.13 Recruitment of Volunteers: ...... 137 3.9.2.14 Experimental Protocol: ...... 137 3.10 Results and discussion ...... 139 3.10.1 Results on nitrate conversion rate in the oral cavity ...... 139 3.10.2 Discussion on nitrate conversion rate in the oral cavity ...... 142 3.10.3 Results for nitrate and nitrite excretion in urine ...... 144 3.10.4 Discussion on nitrate and nitrite concentrations in the urine ...... 147 3.10.5 Results on the effects of vitamin C supplement on NDMA formation in vivo . 15 0 3.10.6 Discusssion on the effects of vitamin C supplement on NDMA formation in vivo ...... 152 3.10.11 Conclusion ...... 160 3.11 References ...... 161 Chapter 4 ...... 171 Alternatives to sodium nitrite ...... 171 4.1 Introduction ...... 171 4.2 Foodborne pathogens and food poisoning ...... 172 4.2.1 Clostridium species ...... 172 4.2.2 Common bacterial food inhibitors ...... 174 4.3 Food preservation ...... 174 4.3.1 Traditional methods ...... 174 4.3.2 Recent methods ...... 175 4.3.2.1 Alliaceae: Garlic and Onions ...... 175 4.3.2.2 Thymol ...... 176 4.3.2.3 Turmeric and curcuminoids ...... 176 4.3.2.4 Other herbs and spices ...... 177 4.3.2.5 Other methods ...... 181 4.4. Antibacterial properties of herbs and spices ...... 182 4.5 Combinations of commercially available herbs and spices to prevent food poisoning and botulism in vitro and possible commercial application in cured meat products ...... 187 4.5.1 Introduction ...... 187

6 4.5.2 Materials and Methods ...... 188 4.5.2.1 Bacterial cultures ...... 189 4.5.2.2 Reagents and media ...... 189 4.5.2.3 Herbs and spices ...... 189 4.5.2.4 MIC to common food borne pathogenic bacteria ...... 189 4.5.3 Results and discussion ...... 190 4.5.4 Conclusion ...... 195 4.6 References ...... 196 Chapter 5 ...... 200 Conclusions ...... 200 5.1 Overview to experimental aim one and major findings ...... 200 5.1 Overview to experimental aim two and major findings...... 200 5.3 Overview to experimental aim three and major findings ...... 200 Appendix A ...... 201 Appendix B ...... 208 Appendix C ...... 209 Appendix D ...... 210

7 List of Figures

Figure 2.1 Names and general formulae of NOC precursors……………………….…25

Figure 2.2 General structures of NOCs……………………………………….……….25

Figure 2.3 Enterosalivary recirculation of ingested nitrate and its interaction with gastric vitamin C in the acid stomach………………………………………………….31

Figure 2.4 Equilibrium reactions of nitrite in aqueous media……………………...….33

Figure 2.5 Environmental and dietary factors that influence gastric carcinogenesis….37

Figure 3.1 Nitrosamines found in foods and their averaged concentration……………99

Figure 3.2 Degradation of NOCs forming electrophilic products…………………....100

Figure 3.3 Mechanisms of nitrosation of a) amines and b) amides…………………..101

Figure 3.4 Reaction of ascorbic acid with nitrite……………………………………..106

Figure 3.5 Ascorbic acid and its common oxidation products……………………….109

Figure 3.6. Reaction of α–tocopherol with nitrite…………………………………....113

Figure 3.7 Mass spectral peaks for the identification and confirmation of nitrosamines…………………………………………………………………………..117

Figure 3.8 Human trial study protocol………………………………………………..138

Figure 3.9 The mean rate (%) of nitrate conversion to nitrite in the oral cavity of 18 healthy human volunteers…………………………………………………………….140

Figure 3.10 Mean nitrate and nitrite content (mg/mL) in human saliva after fasting overnight and after chewing cured meat (salami, hotdog and ham)……………….....142

Figure 3.11 Mean nitrate, nitrite and NDMA contents (mg/kg) of popular cured meats……………………………………………………………………………….…144

Figure 3.12 Mean urinary nitrate (mg/mL) in healthy human volunteers consuming cured meat and on a diet low on nitrate and nitrite over three days………………….146

8 Figure 3.13 Mean urinary nitrite in healthy human volunteers consuming cured meat (mg/L) and on a diet low on nitrate and nitrite over three days……………………...147

Figure 3.14 Ascorbic acid content in 500 mg vitamin C tablets. Mean of five replicates from ten tablets……………………………………………………………………….150

9 List of Tables

Table 2.1 Different types and examples of some NOCs……………………………....26

Table 2.2 Mean recoveries for nitrate and nitrite in selected vegetables and meat……72

Table 2.3. Mean nitrate and nitrite contents and their recoveries in fresh vegetables after five minutes boiling…………………………………………………………...….72

Table 2.4 Mean raw nitrate and nitrite contents and recoveries in cured and fresh meat from Sydney supermarkets after pH adjustment………………………………………73

Table 3.1 Maximal wavelength of ascorbic acid at different pH…………………….107

Table 3.2 Quantification parameters on the effects of saliva matrix on nitrate and nitrite recoveries……………………………………………………………………………..140

Table 3.3 Individual Mean nitrate and nitrite contents in the saliva of healthy human volunteers after rinsing with 350mg nitrate water as the control and chewed ten grams of meat as the experiment…………………………………………………………….141

Table 3.4 Quantification parameters on the effects of urine matrix on nitrate and nitrite recoveries……………………………………………………………………………..145

Table 3.5 Individual nitrate and nitrite contents in the urine of healthy human volunteers before and after treatments………………………………………………..145

Table 3.6 Quantification parameters on the effects of urine matrix on ascorbic acid recovery………………………………………………………………………………150

Table 3.7 Individual ascorbic acid and NDMA in the urine of healthy human volunteers before and after treatments…………………………………………………………...151

Table 4.1 MIC range for various bacteria in vitro using curcumin or turmeric oil…..182

Table 4.2 MIC on selected food pathogenic bacteria using sumac spice…………….183

Table 4.3 MIC and MBC on selected food pathogenic bacteria using Viola tricolor herb…………………………………………………………………………………...183

10 Table 4.4 MIC of various spices including allspice, clove, thyme, nutmeg and star anise on selected Escherichia coli strains…………………………………………………..183

Table 4.5 Averaged MIC of classically or high-intensity ultrasound-extracted spices including ginger, fingerroot and turmeric on various strains of Listeria and Salmonella……………………………………………………………………………184

Table 4.6 Growth and survival of Escherichia coli O157 and Salmonella enterica serovar Enteritidis in broth model systems and mayonnaise…………………………184

Table 4.7 Antibacterial activity of clove oil on Listeria monocytogenes in chicken frankfurters…………………………………………………………………………...184

Table 4.8 Antimicrobial activities of garlic and clove on selected foodborne pathogenic bacteria………………………………………………………………………………..185

Table 4.9 MIC of Australian native herb extracts against common food-related bacteria using microtitre broth microdilution assay…………………………………………...185

Table 4.10 MIZ of Turkish spice extracts to the growth of E. coli O157:H7……..…185

Table 4.11 MIZ for oregano, garlic, and rosemary essential oil against common foodborne pathogenic bacteria…………………………………………………………….186

Table 4.12 MIC for nitrite and its common derivatives or complexes on the growth of C. sporogenes and L. monocytogenes……………………………………………...…186

Table 4.13 MIC (w/v) of common herbs and spices individually or in combination (50:50) against common food borne pathogenic bacteria in vitro up to 72 hr……….191

Table 4.14 MIC (w/v) of common herbs and spices individually or in combination (50:50) against Clostridium sporogenes at 24 hr...... 194

11 List of Abbreviations

AA Ascorbic Acid ABS Australian Bureau of Statistics ADI Acceptable Daily Intake AOAC Association of Official Analytical Chemists ASC Ascorbate ion ATP Adenosine triphophate BHP Basic Hydrogen Peroxide cNOS Constitutive Nitric Oxide Synthase cGMP Cyclic Guanosine Monophosphate COX Cyclooxygenase cpe Enterotoxin gene CYP Cytochrome DAN 2,3-diaminonaphthalene DENA Diethylnitrosamine DHAA Dehydroascorbic Acid DHIAA Dehydroisoascorbic acid DMNA N-dimethylnitrosoamine DNA Deoxyribonucleic Acid DTT Dithiothreitol EDTA Ethylenediaminetetraacetic acid EGCG Epigallocatechin gallate eNOS Endothelial Nitric Oxide Synthase EU European Union FAO Food and Agriculture Organization FSANZ Food Standards Australia New Zealand GATT Gernal Agreement on Tariffs and Trade GC-MS Gas chromatography Mass Spectrometer

12 GMP Good Manufacturing Practice GRAS Generally Recognized As Safe GS-NO S-nitroglutathione HACCP Hazard Analysis Critical Control Points HCA Heterocyclic Amine

HNO2 Nitrous Acid HPLC High Performance Liquid Chromatography HS Head Space HTHQ 1-O-hexyl-2,3,5-trimethylhydroquinone IAA Isoascorbic acid IFN Interferon IL Interleukin iNOS Inducible Nitric Oxide Synthase IR Infrared LOD Limit of Detection LOQ Limit of Quantification MBC Minimum Bactericidal Concentration MIC Minimum Inhibitory Concentration MIZ Minimum Inhibitory Zone NA Nutrient Agar

NDEA N-nitrosodiethanolamine

NDMA N-nitrosodimethylamine NMEA N-nitrosomethylethylamine NMR Nuclear Magnetic Resonance NOAEL No Observable Adverse Effect Level NK Natural Killer (Cells) NO Nitric Oxide

NO2 Nitrogen dioxide

N2O3 Dinitrogen trioxide

13 N3O4 Trinitrogen tetraoxide NOC N-nitroso Compound(s) NPIP N-nitrosopiperidine NPRO N-nitrosoproline NPYR N-nitrosopyrrolidine NTCA N-nitrosocarboxylic acid NTHZ N-nitrosothiazolidine ONOOH Peroxynitrous acid OONO- Peroxynitrite PIC-A Tetrabutylammonium phosphate ppb Parts Per Billion ppm Parts Per Million rpm Revolutions Per Minute RNOS Reactive Nitrogrn Oxide Species RP Reversed Phase SCN- Thiocyanate SPME Solid Phase Microextraction SPS Sanitary and Phytosanitory (measures) TBDMS Tert-butyldimethylsilyl TBT Technical Barriers to Trade TCEP Tris(2-carboxyethyl/phosphate) TEA Thermal Energy Analyzer TNF Tumor Necrosis Factor TVC Total Vitamin C TYG Trypticase-yeast extract glucose UV Ultraviolet WHO World Health Organization

14 Chapter 1 Introduction

N-nitroso compounds (NOCs) have received plenty of attention in the past five decades as evidence suggests increasing risk of cancer in the gastrointestinal tract as dietary nitrate and nitrite intake increases. Both anions are commonly found in food and due to their reactive nature, together with their byproducts, can affect different systems in the body. Nitrate and nitrite content in Australian food supply and their respective intakes are not available for Australians, and hence the purpose of this thesis is to determine their content in selected vegetables and cured meat products and their estimated intake. In addition, ascorbic acid is known to reduce NOCs such as N-nitrosodimethylamine (NDMA) in vivo, and its effectiveness will be determined in human subjects based on restricted diet on cured meat with and without vitamin C supplement. Lastly, there is an increasing interest in alternatives to nitrite as a food preservative in cured meat products since they contribute to the majority of dietary nitrite intake. The antibacterial properties of commercially available herbs and spices will be examined as well as potential anti-Clostridia properties will be tested in vitro.

This thesis comprises of five chapters. Chapter One gives an introduction and introduces specific experimental aims; Chapter Two provides literature review on nitrate and nitrite followed by an analytical section on their content in selected foods and their estimated intake; Chapter Three provides literature review on N-nitroso compounds and their nitrosation inhibitors, followed by an analytical section on the human trial; Chapter Four provides background on common food borne bacteria including Clostridium species and antibacterial properties of herbs and spices, followed by an analytical section on their potential anti-botulinal effects; and finally Chapter Five summarizes the aims and the major findings.

15 1.1 Experimental aims

1.1.1 Nitrate and nitrite in Australian sourced vegetables and cured meat and their estimated dietary intake by Australians

High dietary intake of nitrite can increase the production of N-nitroso compounds (NOCs) in the stomach, which can increase the rate of gastrointestinal cancers as shown in numerouns epidemiological studies. The preservative sodium nirite is added to cured meat products to prevent botulism and the concentration varies from countries. In addition, nitrate from green leafy vegetables can be converted to nitrite in the oral cavity thus increases one’s dietary nitrite intake. The amount of nitrate contained in these vegetables is a result of agricultural and environmental factors, which again varies from countries. Therefore, in order to estimate the dietary intake of nitrite in the Australian population, the current nitrate and nitrite concentrations in vegetables and cured meat products need to be evaluated.

1.1.2 Effects of vitamin C supplementation on N-nitrosodimethylamine formation in healthy human volunteers

N-nitrosodimethylamine (NMDA) is a potent carcinogen and is the most common N- nitroso compound found in food and some beverages. NDMA formation is dependent on the dietary intake of nitrate and nitrite, which is expected to increase with increasing dietary intake of these anions. Asocrbic acid is known to reduce the formation of NOCs by competing with nitrite ions for precursors present in the stomach. Based on 18 healthy human volunteers, this chapter will determine the concentration of NDMA in their urine after a serving of cured meat followed by supplementation of 500 mg vitamin C to determine its effective in reducing the formation of NDMA in vivo.

1.1.3 Anti-Clostridia properties of herbs and spices and possible substitution for sodium nitrite

Sodium nitrite is added to cured meat products for public health measures to prevent botulism. However, due to increasing evidence of high nitrite intake with numerous gastrointestinal cancers, its addition is regulated in most developed countries. Many herbs and spices have antibacterial properties that can inhibit many food poisoning bacteria. Their antibacterial action against Clostridium botulinum has been tested and shows great promise in replacing or at least minimizing the use of sodium nitrite as a

16 preservative. However, the combination of these herbs and spices will be looked at in this part of the thesis since many of the active ingredients may have synergistic actions that is worth exploring and may have commercial applications.

17 Chapter 2 Nitrate and nitrite

Nitrate and nitrite are simple anions present mainly in green leafy vegetables and cured meat products, respectively. Their dietary sources, biochemistry and metabolism will be revealed in this chapter.

2.1 Dietary sources of nitrates and nitrites

2.1.1 Plant food

Diet contributes to nitrosamine-related cancer in three different ways: 1) by modulating the in vivo synthesis of nitrosamines; 2) as a source of exposure to preformed nitrosamines, and 3) as a source of amines and nitrite that can react to form a variety of nitrosamines in the mouth, stomach or intestine (Craddock, 1990).

Nitrates are essential plant nutrients as they are the principle source of nitrogen; the presents of both constitutively and inducible nitrate transport systems in plants allow the assimilation of inorganic nitrate from soil and is subject to negative feedback regulation (Forde, 2000). Thus nitrates are naturally present in all vegetables, cereals and fruits, although at different concentrations between plant species and plant parts. Furthermore, nitrate concentration increases in plants grown at low light levels, therefore northern hemisphere produces plants with higher nitrate content (Duncan et al., 1997) than similar horticultural practices in the southern hemisphere.

Plant materials, water and soil are natural sources of nitrates as a consequence of nitrogen fixation. The amount of nitrate present in these sources will vary from region to region as well as dependent on its agricultural practices. For example, the use of nitrogen-based fertilizers will greatly enhance the nitrate content of soil and hence its crops. This may have explained the inter- and intra-variation in nitrate levels observed between retail vegetables (Meah et al., 1994). Nitrite level in vegetables may increase during post-harvest storage by the action of indigenous bacteria and/or the presence of nitrate reductase (Hunt, 1994). However, under certain conditions, nitrite can be oxidized back to nitrate (Cassens, 1995). Selecting the right cultivar and by

18 manipulating the horticultural practices can therefore reduce nitrate concentrations in plants, as was demonstrated for English spinach (National Academy of Science, 1981).

Vegetables, particularly green leafy vegetables, contribute 80-90 % of dietary nitrate according to Duncan et al. (1997). Thus vegetarians consume three times more dietary nitrate than non-vegetarians at 189 mg/person/day and 61 mg/person/day, respectively (Duncan et al., 1997). Therefore vegetables are the main dietary sources of nitrates (Briggs and Lennard, 2002). However, the intake may vary significantly between individuals and from day to day due to variation in consumption pattern and horticultural practices. For example, Amr and Hadidi (2001) demonstrated that cultivar and harvest date had some significant effect on the nitrate and nitrite levels of selected vegetables. Although some vegetables contain low levels of nitrite, including potatoes, tomatoes and beets, their contribution to the total dietary nitrite burden is minimal (Walters, 1980).

Huarte-Mendicoa et al. (1997) demonstrated that fresh broccoli had traces of nitrites and low amounts of nitrates in Spain produce, and that industrial freezing increased the nitrate level perhaps as a result of nitrates in processing water. In addition, they showed that cooking decreased nitrate levels in both fresh and frozen broccoli by 22 to 79 %, and that nitrite levels were not significantly affected by either freezing or cooking. This suggests that nitrites are more stable than nitrates in vegetables, probably due to the lack of acidic condition in vegetables that is required to promote nitrite conversions.

Chen et al. (2004) demonstrated that at 0.30 gN/kg in the form of potassium nitrate resulted in the optimal plant yield in three leafy vegetables. They also demonstrated that increasing nitrate supply increased nitrate concentration in the whole plant due to accumulation in various plant parts. Furthermore they had shown that there was a threshold of nitrate concentration in the metabolic pool to induce nitrate reductase activity.

White Jr. (1975) estimated that in US populations up to 80 % dietary nitrate came from vegetables, with meat contributing only 15 %. Fruits, dairy, water and bread were considered insignificant sources of dietary nitrate. In addition, dietary intake of nitrate generates salivary nitrite that may contribute up to 65 % of dietary nitrite once swallowed, followed by cured meats at 35 %. Other dietary sources of nitrite were considered insignificant. Similarly, the estimated dietary intake of nitrate for UK was

19 95 mg/day and for nitrite was 1.4 mg/day, where vegetables contributed to over 90 % of the nitrate intake and cured meat contributed 65 % of the nitrite intake (Knight et al., 1987). In addition, nitrate is formed endogenously in humans at approximately 1 mg/kg body weight per day (Gangolli et al., 1994), thus an average adult makes around 70 mg of nitrate per day.

Australia’s food composition tables (FSANZ, 2009) were mostly based on overseas data especially those from the United Kingdom and the United States. However, in the revised Australian composition tables based on food analysis performed in Australia, the edible portion of fruit increased by 4 % whereas in meat it decreased by 16 % (Cashel and Greenfield, 1995). Thus dietary contribution of nitrates and nitrites may be over-estimated, whereas dietary intake of antioxidants such as vitamin C and vitamin E may have been under-estimated.

2.1.2 Cured meat

Nitrite used in meat preservation performs three diverse functions. Firstly, nitrite imparts the desirable red color in cured meat at an initial concentration of approximately 20 mg/kg of NaNO2 through the formation of nitrosylmyoglobin as a result of nitric oxide reacting with myoglobin. Secondly, at concentrations of 50 mg/kg of NaNO2, nitrite in combination with salt imparts the characteristic flavor of cured meat. Finally, sodium nitrite is used to protect consumers from botulism by inhibiting toxin production by the causative microbe Clostridium botulinum, and it was estimated at least 100 mg/kg of residual NaNO2 is required for the protection under commercial conditions (Walters, 1980). In addition, sodium nitrite can retard lipid oxidation thus prolonging the shelf life of cured meat products (Pennington, 1998). It is apparent that two out of three known functions of sodium nitrite added to cured meat is related to consumer appeal and only one is important in food poisoning prevention, but the other two functions obviously provide a vital economic incentive in the sale of cured meat products.

Furthermore, since nitrate is much more stable than nitrite in cured meat, it is therefore often added to provide a reservoir of nitrite during storage thorough microbial reduction (Dennis et al., 1990). Nitrite exerts its antimicrobial effect in combination with sodium chloride at low pH present in cured meat over time as the result of acid- producing micro-organisms naturally present. In contrast, sodium nitrate has no direct

20 antimicrobial activity until it is reduced to nitrite by bacterial nitrate reduction (Duncan et al., 1997).

Different authors attributed different percentage of dietary nitrate and nitrite to the major food groups, but the consensus are that vegetables contributed to the majority of dietary nitrate and that cured meat products contributed to the majority of dietary nitrite. However, the nitrate and nitrite concentrations in meat samples can be significantly reduced by boiling and roasting with boiling being more effective. Thus it was recommended to boil meat and poultry before consumption (Ologhobo et al., 1996).

2.1.3 Water

According to Walters (1980), drinking water supplies may contribute to the total dietary nitrate burden. It can also vary dramatically from one country’s water supply to the next, or even within a country since the nitrate present can vary significantly. So besides food, tap water may contribute 10-20 mg nitrate per person per day (MAFF, 1987 in Meah et al., 1994).

The guideline for the maximum concentration of nitrate and nitrite allowed in Sydney water supply was set at 50 and 3 mg/L, respectively (Sydney Water, 2007). This guideline was similar to those of the European Communities for nitrate but nitrite was set at 0.1 mg/L. These limits were introduced to prevent infantile methaemoglobinaemia (Massey, 1991).

As part of Sydney water quality monitorty program, Sydney drinking water satisfied the requirement at 0.02 to 2.89 mg/L of nitrate and 0.003 to 0.234 mg/L of nitrite (Sydney Water, 2007). Assuming that each adult drinks two liters of water a day, tap water in Australia is generally not a significant source of dietary nitrate and nitrite. However, nitrate concentrations in natural waters had increased in many countries due to increased use of artificial fertilizers, changes in land use and disposal of waste from intensive farming (WHO, 1985, in Gangolli et al., 1994), which should be carefully monitored over time and drastic changes to modern farming and land use may be required to prevent nitrate from accumulating in the natural waterways.

21 2.2 Nitrate, nitrite and its derivative

2.2.1 Mechanisms of nitrate reduction to nitrite

In addition to the ingestion of dietary nitrite, the ingested nitrate can contribute to the body’s total burden of nitrite by the enzyme nitrate reductase. In eukaryotic organisms, - the following equation was proposed for its nitrate reductase activity: NO3 + NAD(P)H - + - → NO2 + NAD(P) + OH . The NAD(P)H-nitrate reductase is the first enzyme in the nitrate assimilatory pathway, which catalyses the reduction of nitrate to nitrite (Barbier et al., 2004). Prokaryotic organisms, especially facultative anaerobic bacteria, can synthesize nitrate reductase enzyme when exposed to low oxygen tension, such as the micro-organisms residing in the oral cavities of humans (Duncan et al., 1995).

The total nitrite level in humans may not all come from dietary sources. For example, Shiotani et al. (2004) demonstrated that Helicobacter pylori positive patients had significantly higher nitrite concentrations in their gastric juice compared to H. pylori negative patients. In addition, they had shown that the nitrite contents in H. pylori positive patients decreased after the eradication of H. pylori. Hence the bacterium may have a role in nitrate-reduction in the stomach cavity.

It was suggested that H. pylori infection reduces stomach acid secretion thereby inhibiting the conversion from nitrite to NO in the stomach. Therefore more nitrite is accumulated in the stomachs of H. pylori positive patients. It was proposed by Shiotani et al. (2004) that H. pylori increases gastric nitrite concentration by promoting the conversion of nitric oxide (NO) to nitrite due to more alkaline stomach pH and inflammation with increased formation of superoxide anion. In addition, Shiotani et al. (2004) had shown that the higher stability of nitrite at neutral pH reduced its reactivity with other components present in the gastric content. This should decrease the formation of NOCs and decrease the risk of gastric cancer. However, to the contrary, the incidence of gastric cancer is higher in people infected with H. pylori, which may be due to its irritation effects on the lining of stomach cells causing them to divide uncontrollably.

2.2.2 Mechanisms of nitrite reduction to nitric oxide

Nitrite under acidic condition is converted to nitrous acid and other reactive nitrogen species (see equations below), which can cause nitration of aromatic compounds such

22 as tyrosine, de-amination of DNA bases, and nitrosation of amines (Pannala et al., 2003).

- + NO2 + H → HNO2

2HNO2 → H2O + N2O3

N2O3 → NO + NO2

3HNO2 → HNO3 + 2NO + H2O

+ - + HNO2 + 2HNO3 → H3O + 2NO3 + NO

In addition, haemoglobin had been shown to act as nitrite reductase, where haemoglobin becomes deoxygenated, vacant haem become nitrite reductase to produce methemoglobin and nitric oxide (see equations below). As the pH decreases, the reaction of nitrite reductase by deoxyhaemoglobin increases. However, under oxygenated conditions, nitrite is oxidized to nitrate by oxyhaemoglobin. This process is thought to perform a vasodilation function by the nitric oxide generated (Gladwin et al., 2004).

- II + III NO2 + haemoglobin-Fe (deoxyhaemoglobin) + H → haem-Fe (methemoglobin) + NO + OH-: NO + haem-FeII → haem-FeII-NO (Doyle et al., 1981, in Gladwin et al., 2004).

Interestingly, in the presence of nitrite and under aerobic condition, plant nitrate reductase produces NO and its toxic derivative peroxynitrite (Yamasaki and Sakihama, 2000). How much this process contributes to the body’s total level of nitrite and reactive nitrogen derivatives is not known. However, it is commonly suggested that plant food do not contribute much to the total dietary nitrite intake.

2.2.3 Metabolism of nitrate, nitrite and nitric oxide

The major source of dietary nitrate and nitrite are vegetables and cured meat products, respectively. Once ingested, approximately 20 % of the nitrates can be reduced to nitrite by nitrate-reducing bacteria situated on the posterior surface of the tongue (Mowat and McColl, 2001). The salivary glands can concentrate dietary nitrates and secrete them into the saliva. In addition to nitrate-reducing bacteria, mammalian

23 possesses nitrate reductase, which can also reduce part of the ingested nitrate to nitrite (Duncan et al., 1995).

Once swallowed, nitrate and nitrite is absorbed in the stomach and proximal small intestine. It was estimated that approximately 25 % of ingested nitrate is re-circulated in the plasma and saliva, and approximately 75 % of ingested nitrate is excreted mainly in urine, with little excreted in sweat and faeces (Pannala et al., 2003). Mitsui and Kondo (2000) demonstrated that at least 1 % of dietary nitrate was excreted as breath

N2O. In addition, older subjects gave higher N2O than younger subjects indicated that dietetary nitrate was reduced rapidly in the upper inestine with more bacteria inhabiting their gut. Nitrite on the other hand, can react with hydrochloric acid in the stomach to form nitrous acid, nitric oxide, as well as other reactive nitrogen species (Pannala et al., 2003).

It was demonstrated by Pannala et al. (2003) that high nitrate intake from food lead to a significant increase in nitrate and nitrite concentrations in the urine and saliva. Plasma nitrate level also increased correspondingly, but there were no changes in the plasma nitrite concentration. They also demonstrated that the maximum urinary nitrate excretion occurred 4 to 6 hours after the consumption of high-nitrate meal, and that the absorption of nitrate from organic nitrate source such as food was slower than from inorganic nitrate salt. It was proposed that organic dietary sources must undergo the additional extraction step and hence takes longer to be absorbed.

Some ingested nitrite can bind with numerous precursors (Figure 2.1) or nitrosatable amines to form powerful carcinogens known as N-nitroso compounds (NOCs)(International Agency for Research on Caner, 1987). Two major groups of NOCs include N-nitrosamine and N-nitrosamide (Figure 2.2) with some examples shown in Table 2.1. The former requires metabolic transformation by cytochrome P450-dependent hydroxylation to form an alkylating agent α-hydroxynitrosamine (Tricker and Preussmann, 1991). In contrast, N-nitrosamide is chemically reactive and is not stable at physiological pH and decomposes to form an alkylating agent (Shank, 1975).

Figure 2. 1 Names and general formulae of NOC precursors (Shephard et al., 1987).

Figure 2. 2 General structures of NOCs (Shank, 1975).

25 Table 2.1 Different types and examples of some NOCs (Biaudet et al., 1996 in Nollet, 1996).

N-nitroso compounds Subgroups Examples Volatile nitrosamines Alkyl nitrosamines N-nitrosodimethlyamine N-nitrosodiethylamine N-nitrosodibutylamine Cyclic nitrosamines N-nitrosopyrrolidine N-nitrosopiperidine N-nitrosomorpholine N-nitrosothiazolidine Aryl nitrosamines na N-nitroso-N-methyl- phenylamine N-nitrosodibenzylamine Nitrosated amino acids na N-nitrososarcosine N-nitrosoproline N-nitroso-4- hydroxyproline Hydroxylated na N-nitrosodiethanolamine nitrosamines N- nitrosohydroxypyrrolidine Nitrosamides Alkyl nitrosamides N-nitroso-N- ethylacetamide Nitrosourea N-nitroso-N-methylurea Nitrosocarbamate N-methyl-N-nitrosoethane Nitrosoguanidine N-methyl-N-nitro-N- nitrosoguanidine Nitrosated dipeptides na N-nitrosoprolylglycine Nitroso sugar amino acids na N-nitro-D-fructose-L- histidine na-not available

2.2.4 Physiological roles of nitrate, nitrite and nitric oxide

NO is a small compound with a size of 30 Da. It has no charge with a single unpaired electron making it a radical and chemically active, but it is relatively stable with a physiological half-life of seconds to minutes, which is inversely proportional to its concentration and it is dependent on its immediate environment. Its solubility and diffusion properties is very similar to oxygen; NO is a gas under atmospheric conditions and a solute that is readily diffusible in body fluids, tissues and across cell membranes (Coleman, 2001).

NOo is a reactive free radical, it can exist in tissues in several forms such as, nitroxyl - - ion (NO ), nitrous acid (HNO2), nitrogen dioxide radical (NO2), peroxynitrite (ONOO :

26 combination of superoxide and NO) and peroxynitrous acid (ONOOH). NO may exert a pro-inflammatory reaction, but it may also have immunoregulatory roles (Bryan, 2006).

NO has numerous functions in the human body that includes vasodilation, inflammation, antimicrobial, and is believed to be involved in nerve signaling (Ellis et al., 1998). NO plays a role in defense against infectious organisms, as well as regulating the activity, growth, and death of many immune and inflammatory cells including macrophages, T lymphocytes, mast cells, neutrophils and natural killer cells. The principal enzyme that produces high-level sustained NO is the inducible type-2 isoform of NO synthase (iNOS-2), where the NO is rapidly oxidized to reactive NO species that can S-nitrosate thiols to modify key signaling molecules such as kinases, transcription factors and enzymes in mitochondrial respiration resulting in depletion of ATP and cellular energy (Coleman, 2001).

NO is produced by endothelial cells and diffuses into smooth muscle causing vasodilation and into vessel lumen where most of the gas is rapidly inactivated by dioxygenation reaction with oxyhaemoglobin to form nitrate. As a result of the diffusional barrier of NO around the erythrocyte and along the endothelium in laminar flowing blood, the inactivation reaction of NO by haemogloin is reduced, allowing sufficient NO can escape for vasodilation and to react in plasma and tissues to form nitrite anions and NO-modified peptides and proteins (RX-NO) (Dejam et al., 2004).

Since NO has no cell surface receptor it enters cells indiscriminately. However, its selectivity is dependent on three factors: Firstly, its concentrations and reactivity with surrounding molecules. Secondly, the proximity of target cells, and thirdly the way in which the target cells is programmed to respond (Coleman, 2001).

NO is produced by:

1. Constitutively expressed enzymes NOS-1 and NOS-3 that produce NO rapidly and transiently at low concentrations and are activated by physiological stimuli that trigger intracellular calcium signal such as an action potential or activation of endothelial cell receptors by acetylcholine. They act directly with the iron atom in the haem group of guanylyl cyclase to activate the enzyme cGMP to trigger a rapid transient cellular response (Coleman, 2001).

27 2. Unlike NOS-1 and NOS-3, the induced enzyme NOS-2 is not expressed in resting cells and produces NO at sustained high concentrations when it is induced by immunological stimuli such as bacterial lipopolysaccharide or cytokines such as IL-1, TNF-α or IFN-γ. They have indirect effects because they are unstable at high concentrations and are rapidly oxidized under aerobic conditions to reactive nitrogen oxide species (RNOS) (Coleman, 2001).

Under gaseous conditions the RNOS formed are nitrogen dioxide (NO2), dinitrogen trioxide (N2O3) and trinitrogen tetraoxide (N3O4), but under aqueous biological conditions the major oxidative product is dinitrogen trioxide. RNOS can be hydrolysed and excreted as nitrite, or nitrosate the thiol group in glutathione to produce S- nitrosoglutathione (GS-NO) or thiol groups in proteins to generate protein-S-NO, which inhibit the activity of many proteins including mitochondrial enzymes and transcription factors producing long term cellular effects. Furthermore, when high - concentrations of NO is combined with high oxidative stress, superoxide (O2 ) interacts with NO to produce a highly toxic compound known as peroxynitrite (OONO-), which irreversibly damages mitochondrial complexes I, II, IV and V, as well as acetonitase, creatine kinase, mitrochondrial DNA and superoxide dismutase. At higher concentrations, RNOS can induce cell toxicity by nitrosating DNA and tyrosine residues as well as inducing lipid peroxidation (Coleman, 2001).

In addition, molecular oxygen and NOS converts L-arginine to NO and L-citrulline via the intermediate N-hydroxgy-L-arginine involving an enzymatic process that utilizes electrons donated by NADPH. The nitrogen atom in NO is derived from a terminal guanidine group of the arginine side chain (Coleman, 2001). Furthermore, it was demonstrated by Gladwin et al. (2004) that deoxyhaemoglobin can act as a nitrite reductase that reduces nitrite to give nitric oxide.

Vetrovsky et al. (1996) proposed an alternate pathway to the well-established synthesis of NO by NO-synthase via Hydroxy-L-arginine (OH-L-Arg), which is an intermediate in NO production from L-arginine. They concluded that after the production of OH-L- Arg by NO-synthase, it could decompose to NO by the action of superoxide ion. Vetrovsky et al. (1996) also provided another possible mechanism of NO production by hydrolysis of OH-L-Arg to hydroxylamine, which can then be oxidized to NO by superoxide ion. Once NO is formed, it can be further oxidized to form nitrite ions. It

28 was suggested that these non-enzymic NOx formation might explain tissue formation of nitrite from OH-L-Arg. This process may therefore increase the availability of NO for different functions, where OH-L-Arg and hydroxylamine may serve as NO- precursor for cells that do not possess the L-Arg/NO pathway, for example, human monocytes.

Nitrite exists in all tissues capable of nitric oxide synthesis form L-arginine, and Kozlov et al. (1999) suggested that the mammalian mitochondria may be a site for nitric oxide synthase-independent NO formation from the major metabolic degradation product nitrite.

2.2.5 Immunological effects and disease

NO have several roles in immunity: as a toxic agent towards infectious organisms, an inducer or suppressor of apoptosis, or as an immunoregulator. Coleman (2001) suggested that the inducible NOS-2 is likely involved in the immune response for two reasons. Firstly, NOS-2 produces NO over days or weeks, similar to an immune response. Secondly, NOS-2 generates high levels of NO over sustained period of time, which is necessary to be effective as a toxic or immune regulatory mediator (Coleman, 2001).

Many immune and inflammatory cells express NOS-2 and produce NO, which includes: fibroblasts, endothelial and epithelial cells, macrophages, antigen-presenting cells, natural killer (NK) cells, T lymphocytes, mast cells, and neutrophils. In most of these cells, the effect of NO is inhibitory on cell function or growth. For examples: NO mediates NK cell destroying target cells and regulates NK cell function, it inhibits activation of mast cells, and can enhance or inhibit neutrophil activation depending on its concentration (Coleman, 2001).

Based on animal research on Balb/c mice, Abuharfeil et al. (2001) demonstrated the maximum suppression of immune cell activities including the proliferation of B and T cells, antibody production and reduction in NK (natural killer) cell occurred in the first 24 h at 100 mg/kg sodium nitrite. In addition, they had shown the immunosuppressive effect of sodium nitrite was reversible after stopping the exposure.

The effects of NO on inflammation are concentration dependent. At low concentrations, NO is pro-inflammatory by inducing vasodilation and the recruitment

29 of neutrophils. In contrast, at high concentrations, NO is anti-inflammatory where it down-regulates adhesion molecules, suppresses activation and induces apoptosis of inflammatory cells. The inhibitory effects of NO on immune cell function and growth may be due to its inhibition of mitochondrial respiratory pathways leading to the depletion of energy resources (Coleman, 2001).

NO regulates the death of immune cells, either by induction or inhibition of apoptosis, or by necrosis. NO can also mediate the killing of tissue-specific cells in immunologically mediated diseases (Coleman, 2001).

2.3 Toxicity and Metabolism

2.3.1 Nitrate and nitrite

Dietary intake of nitrate generates salivary nitrite by activities of microorganisms residing in the mouth, which can react with secondary amines in the acidic gastric condition and forming potentially carcinogenic N-nitrosamines amongst other reactive intermediates of nitrogen (Dykhuizen et al., 1996). However, despite its possible harmful effect, one of the intermediate products such as nitric oxide may provide a significant defense against swallowed pathogens in the mouth and lower gut of humans as the result of symbiotic actions of nitrate-reducing bacteria on the surfaces of tongue (Duncan et al., 1995; Dykhuizen et al., 1996). Furthermore, McKnight et al. (1997) had demonstrated that majority of dietary nitrite entering the stomach were rapidly converted to nitric oxide therefore only small amount of nitrite can participate in nitrosation to form NOCs. In addition, it was suggested that nitric oxides might play a role in the normal gastric functioning.

Human oral microflora including Veillonella species, Staphylococcus aureus, S. epidermidis and Nocardia species are known to reduce salivary nitrate to nitrite, which was shown to protect against gastrointestinal diseases most susceptible in people treated with antibiotics that inhibit nitrite-producing bacteria in the mouth. More alarmingly, antibiotic misuse had been implicated in increased antibiotic resistant human gastrointestinal pathogens (Duncan et al., 1997).

Although dietary nitrate is eliminated rapidly in the human body, some is transported to the salivary glands and secreted in the oral cavity, where it is then microbially reduced to nitrite and ends up in the stomach through the action of swallowing saliva (Figure

30 2.3) (Cassens, 1995). Since nitrate-reducing bacteria are more active at higher pH values (Ruddell et al., 1976), majority of nitrate reduction occurs in the oral cavity due to its alkaline condition.

Figure 2.3 Enterosalivary recirculation of ingested nitrate and its interaction with gastric vitamin C in the acid stomach (Mowat and McColl, 2001).

Nitrate is absorbed into the bloodstream by the stomach and the small intestine, whereby 25 and 75 % of dietary nitrate is excreted by the salivary glands and kidneys, respectively (Figure 2.3). Nitrate is concentrated in the salivary glands that are 10 times the concentration in plasma, and salivary nitrate is concerted to nitrite by symbiotic bacteria present on the surfaces of the tongue. Nitrite when swallowed is converted to nitric oxide NO by the acidic stomach, and NO are converted back to their stable end product nitrate, which is then absorbed in the stomach and small intestine, thus

31 completing the cycle of enterosalivary circulation of nitrate. Vitamin C secreted by the stomach lining can increase the conversion of nitrite to NO (Duncan et al., 1997).

The amount of nitrite formed in vivo is dependent on the nitrate reductase activity of microorganisms (Gangolli et al., 1994), which accounts for approximately 20 % of the 25 % ingested nitrate being converted to nitrite in the oral cavity that is then secreted and concentrated in the saliva. This entero-salivary recirculation pathway is the result of 5-7 % of the total absorbed nitrate intake in healthy adults (Bottex et al., 2008). Hence about 5 % of dietary nitrate is converted to nitrite that is dependent on the availability and amount of nitrate consumed, which can then potentially be converted to NOCs (Massey, 1997).

Approximately 80 % of gastric nitrite arises from this endogenous reduction of nitrate. The other 20 % of gastric nitrite arises from ingested nitrite from preserved and smoked meat or fish. Thus most nitrosation occurs 1 to 2 hours after a meal within the human stomach. Gastric nitrosation may be catalysed by halides and thiocyanate in addition to bacteria (Mirvish, 1995). Craddock (1990) suggested that the rate limiting factor was the rate of nitrite formation by oral bacteria, and the extent of endogenous synthesis of nitrite from sources other than vegetables, which its nitrate contents was readily converted to nitrite.

- From Figure 2.4, nitrosating agent is produced between nitrite ions (NO2 ) with protons + + (H or H3O ) to give nitrous acid (HNO2) (Tricker and Kubacki, 1992), which dimerizes with the loss of water to form N2O3 that can react with amines forming NOCs (Mirvish, 1995). Thus neither nitrite nor nitrous acid are nitrosating agents, but are intermediates in the formation of nitrosating agents that includes dinitrogen trioxide + (N2O3), dinitrogen tetraoxide (N2O4) and nitrous acidium ion (H2O NO)(Tricker and + Kubacki, 1992). Furthermore, HNO2 can also be protonated to form H2NO2 , which has an affinity with amides to form nitrosamides (Mirvish, 1995). Since these reactions occur in acid conditions, the stomach with its low pH provides the right kind of condition for the formation of NOCs in vivo.

Figure 2.4 Equilibrium reactions of nitrite in aqueous media (Tricker and Kubacki, 1992). Nitrates once ingested are readily absorbed from the proximal small intestine, where it will be rapidly distributed throughout the body followed by excretion in the urine. Blood nitrate on the other hand, is selectively transported to the saliva where mammalian nitrate reductase activity may account for half the reduction of nitrates to nitrite, with the other half utilised by microorganisms enzymic action (Gangolli et al., 1994). Not many data are available on the absorption of nitrite in humans. However, nitrite could be utilised by gut flora as a source of nitrogen (Gangolli et al., 1994), therefore suggests absorption into the small intestines.

The bioactivity of nitric oxide (NO) is terminated by oxidation to nitrite and nitrate. In addition, nitrite can be recycled to bioactive NO in cells and tissues, and nitrate is a substrate for systemic generation of nitrite, thus a reverse pathway for generation of NO from nitrate is completed (Lundberg and Govoni, 2004). Furthermore, NO can rapidly react with superoxide to form the highly reactive free radical peroxynitrite (ONOO-), which was demonstrated to cause tissue damage and animal mortality (Chow and Hong, 2002).

The no observable adverse effect level (NOAEL) for nitrate based on two-years rat

studies was 2500 mg NaNO3/kg body weight per day, whereas the NOAEL for nitrite

was 10 mg NaNO2/kg body weight per day, or 6.7 mg NO2 ions/kg body weight per day (Gangolli et al., 1994). Unlike nitrate, nitrite is genotoxic and can readily induce methaemoglobinaemia in infants (Fan and Steinberg, 1996). In addition, the lethal dose

for nitrite in adults was estimated to be between 2 and 9 g NaNO2 per day, or 33-250 mg/kg body weight (Corre and Breimer, 1979, in Gangolli et al., 1994), whereas the

33 lethal dose for nitrate ions was estimated at 20 g per day, or 330 mg nitrate ions/kg body weight (Leu et al., 1986, in Gangolli et al., 1994).

2.4 Epidemiology studies

2.4.1 Gastric cancer incidence

It is well established that diet contributes significantly to the development and control of numerous diseases including atherosclerosis, diabetes mellitus, renal failure, osteoporosis and cancer of major body organs. The incidence of cancer was significantly higher in a typical modern Western diet compared to a diet based on vegetables, legumes, grains, nuts and fruits (Hubbard et al., 1994). It is well known that the combination of vitamins, minerals, antioxidants, phytochemical, low saturated fat and fiber makes vegetarians and vegans less susceptible to a range of cancers including cancer of the digestive systems.

It was estimated that 80 % of human cancers are caused by environmental factors associated with food, water and air (Lathia, 1989, in Ologhobo et al., 1996). In addition, malnutrition, dietary habits and lifestyle may be directly or indirectly related to 40 % of the human cancers (Palmer, 1985, in Ologhobo et al., 1996).

Stomach is most at risk from endogenous NOC synthesis since acid catalyses nitrosation reactions. High nitrate intake was associated with gastric cancer in countries such as England, Colombia, Chile, Japan, Denmark, Hungary, Korea and Italy (Forman et al., 1985). However, some findings were contradictory. For example, Forman et al. (1985) concluded in their study that high-risk population had lower intake of nitrate and nitrite than those at low risk. In addition, they found that although smokers had a higher risk of stomach cancer than non-smokers their salivary nitrate/nitrite levels were lower, which suggest possible protective effect of nitrates and/or nitrites. This inverse relationship between nitrate/nitrite intake and gastric cancer incidence may be partly explained by the inherent flaws of the food frequency method adopted by Forman et al. (1985) to derive the crude estimate of the mean population used in their study. For instance, food frequency method may not cover all dietary sources of nitrate and nitrite. Furthermore, since no analyses of samples were required, it was assumed that there are no inter- or intra-variations within food groups (Massey, 1997).

34 South Korea, Japan and China had the highest stomach cancer mortality for men, in decreasing order. Canada, Denmark and US had the lowest stomach cancer mortality for men, in decreasing order. For women, the highest stomach cancer mortality in decreasing order were South Korea, China and Colombia, whereas the least gastric cancer mortality rate for women were Denmark, Canada, and US, in decreasing order (Joossens et al., 1996). The high gastric cancer incidence in the Far East may be due to the consumption of specific foods high in nitrates such as Korean Kimchi, or high in salt as in many traditional Japanese dishes, or particular food preparation method such as broiling of meats (Duncan et al., 1997). Regions of high risk to gastric cancer often coincide with a low intake of foods containing vitamin C. Other risk factors for human gastric cancer include residence in areas with high nitrate-containing soil and foods pickled with salt (Weisburger, 1981). For example, Weisburger (1981) demonstrated the mutagenicity of Salmonella typhimurium TA 1535 found in pickled fish commonly eaten in high risk Japan.

Chen et al. (1992) had associated certain salted fermented fish products including fish sauce to the high gastric cancer mortality in Fujian province of China. Similarly, Seel et al. (1994) had identified potential link of the high gastric cancer rate in Southwest Korea, where salted pickled cabbage and salted seafood sauce were consumed in high amount in that region. The former, also contained high levels of total NOC precursors, and cabbages are known to contain high levels of nitrate than other vegetables.

Exposure to endogenously formed NOC had been associated with increased risks of cancer of the stomach, oesophagus and bladder (Bartsch et al., 1990). Japan has seven times the rate of gastric cancer than U.S., and when immigrated to US their children showed the risk as Caucasian Americans. Thus this epidemiology study showed that environmental factors such as diet played an important part in gastric cancer aetiology. In addition, it was estimated that up to 30 % of cancers are diet related (Dervan, 1999). High bacon consumption in the Netherlands in the 1960s and the consumption of nitrite-cured meat suggested linkage to the high incidence of gastric and other gastrointestinal cancer (Mirvish, 1995).

Joossens et al. (1996) studied dietary salt, nitrate and gastric cancer mortality in 24 countries and demonstrated that nitrate intake became an increased risk factor for gastric cancer when salt intake was also high, but the correlation between gastric cancer

35 and sodium was stronger than with nitrate. They concluded that salt intake was the rate- limiting factor in gastric cancer mortality at the population level. However, one drawback from their study was the relatively small sample size from each country; therefore the representative nature of their data and conclusion is questionable. Populations with a high incidence of gastric cancer also exhibited a high incidence of hypertension since salt exerted a promoting effect in both the etiology of gastric cancer and heart diseases (Weisburger, 1981). Thus a lowered salt intake should reduce the occurrence of hypertension and stroke as well as the risk of gastric cancer in susceptible individuals.

It was suggested by Walters (1980) that gastric ulcers and cancer was due to the resultant pH in the stomach rather than the pathological lesion per se, as an inverse relationship between nitrite and hydrogen ions concentrations was shown in cases with stomach ulcers and cancer.

There was no epidemiological link between populations with high dietary vegetable or nitrate intake and gastric or intestinal cancer (Duncan et al., 1997). This may be due to the increased intake of fruits and the protective role of vitamin C commonly practiced amongst vegetarians and vegans. Figure 2.5 summarizes environmental and dietary factors that either promote or inhibit gastric carcinogenesis.

Figure 2.5 Environmental and dietary factors that influence gastric carcinogenesis (+, possible promoters of carcinogenesis and -, possible inhibitors of carcinogenesis)(Hwang et al., 1994).

Based on cohort and case-control studies in Swedish women between 1987 and 1997, (Larsson et al., 2006a) showed that high consumption of processed meat (bacon, sausage, hotdog, ham and salami) was asscociated with a statistically significant increased risk of stomach cancer. They suggested that dietary nitrosamines might be responsible for the positive association. Similarly, based on meta-analysis between 1966 and 2006 involving six prospective cohort studies and nine case-control studies, Larsson et al. (2006b) concluded that increased consumption of cured meat was associated with an increased risk of stomach cancer. Jakszyn and Gonzalez (2006) reviewd 61 studies (cohorts and case-control) between 1985 and 2005 and showed a positive and significant association between nitrite and nitrosamine intake and gastric cancer, between meat and processed meat intake and gastric and oesophageal caners, and finally between preserved fish, vegeable and smoked food intake and gastric cancers. Suffice to say that the general agreement in the literature is that processed meat and nitrosamines can increase the risk of gastrointestinal cancers. This was

37 discussed by Linseisen et al. (2006) where factors such as preservation methods, cooking methods, and nutrient content were considered in a prospective cohort study in 27 centres across 10 European countries. They concluded that highest intake of cured meats were in Germany, Dutch and northern European centers, but due to manufacturing differences, the highest intake of sodium nitrite was found in Spanish centers that also had the highest salt intake from cured meat. Central and northern European centers consumed the most cholesterol and iron-rich processed meat, and possible hazardous cooking methods such as deep-frying were more common for processed meat preparation in those regions.

2.4.2 Gastric cancer risk factors

By examining the expression of endothelial nitric oxide synthase (eNOS) and inducible nitric oxide synthase (iNOS) in samples from normal gastric muscosa and gastric cancer, Doi et al. (1999) suggested that iNOS and not eNOS plays a role in gastric cancer tumour metastases. Furthermore, they demonstrated that iNOS mRNA may not be induced by either alpha-tumour necrosis factor (α-TNF) or interleukin-6 (IL-6).

The molecular markers to tumour characteristics and metastatic potential including NOS, cyclooxygenase (COX) expression and p53 status in patients with gastric adenocarcinoma were assessed using immunohistochemical technique. It was shown that the expression of inducible enzymes, iNOS and COX-2, increased significantly with increasing tumour stage, size, and the presence of metastases. In contrast, the expression of constitutive enzymes, cNOS and COX-1 showed the opposite trend. It was concluded that turmour-associated nitric oxide production and COX-2 over expression might promote gastric cancer progression by providing a selective growth advantage to tumour cells with non-functioning p53 (Rajnakova et al. 2001).

Hirose et al. (1999) demonstrated strong chemoprevention properties of the synthetic antioxidant 1-O-hexyl-2,3,5-trimethylhydroquinone (HTHQ) in heterocyclic amine (HCA)-induced carcinogenesis. In addition, they have shown that curcumin exerted significant enhancing effects. HCA found in cooked meats was shown to be genotoxic from Ames and other mutation assays. It was suggested that HCA are metabolically activated to N-hydroxy amino forms by CYP1A/1A2 and then further metabolised by N-acetyltransferase-mediated O-acetylation to esters in the liver. These metabolites can then react with DNA and form adducts in the C-8 position.

38 Kamataki et al. (1999) concluded that CYP2A6 gene polymorphism is respnosible for the variability on cancer susceptibility of individuals. Kushida et al. (2000) demonstrated that CYP2A6 was reponsible for the metabolic activation of N- nitrosomethylphenylamine in recombinant Salmonella strain with human cytochrome P450 (CYP2A6) and cytochrome P450 reductase. This established strain may be used to predict human activation of N-nitrosamine promutagens. Furthermore, Fujita and Kamataki (2001) showed that N-nitrosamines with relatively short alkyl chains such as NDMA and NMEA were primarily activated by CYP2E1 in recombinant Salmonella strain by Ames test.

2.4.2.1 Carbohydrate

Carbohydrate consumption was associated with increasing risk of gastric cancer. However, the mechanism was not known and that different types of carbohydrates were not tested (Risch et al., 1985). More recently, it was suggested that the consumption of various types of starchy foods may contribute to the risk of gastric cancer by acting as a physical irritant of the gastric mucosa and combined with a low protein diet may reduce the production of gastric mucus thus facilitating carcinogen absorption (Kato et al., 1992).

2.4.2.2 Saturated fat

Unsaturated fatty acids had been studied for their role in peroxidative reactions possibly resulting in tissue damage and gastric carcinogenesis (Risch et al., 1985). Fat intake was the only macronutrient positively associated with the risk of gastric cancer in a population-based case-control study in Sweden (Hansson et al., 1994a). It was proposed that a fatty meal delayed emptying of the stomach thus prolonged the exposure to carcinogens ingested and/or those already present (Heddle et al., 1988 in Hansson et al., 1994a).

González et al. (1994) concluded that high consumption of exogenous nitrosamines, nitrites, fat and cholesterol increased the risk of gastric cancer in case-control study of the Spanish population.

39 2.4.2.3 Alcohol

Certain alcoholic beverages such as beer, scotch and whiskey contains preformed nitrosamines, and studies had shown that even moderate alcohol consumption is associated with gastric cancer incidence (Hwang et al., 1994).

2.4.2.4 Salt, pickled and smoked food

Excess salt (NaCl) is known to cause mucosal damage in the stomach lining followed by inflammatory changes during repair. Furthermore, salt promotes intragastric nitrosation and cell replication thus increases the rate of carcinogen formation and mutation (Hwang et al., 1994). Lee et al. (1995) showed a positive association between high salt intake, broiling method of cooking and stomach cancer in case-control population in Korea.

Skrökki (1995) had demonstrated that 65 % of the Finnish cured meat products tested contained more salt averaging 2.3 % NaCl (w/w) than was recommended (1.2 - 1.8 % w/w). This can pose increased risk of gastric cancer as well as hypertension in the susceptible populations.

Dietary exposure to NOCs can be either the ingestion of preformed NOC such as nitrosamines, in smoked food, or the ingestion of nitrite and nitrosatable substances (Wogan and Tannenbaum, 1975).

2.4.2.5 Cooking method

High temperature cooking such as broiling or frying yielded the highest nitrosamine levels (Cassens, 1995). In addition, Skrypec et al. (1985) had demonstrated that the fatty acid composition of bacon and the frying atmosphere influenced its N-nitrosamine formation. Furthermore, they showed that wood smoke was linked to the formation of N-nitrosothiazolidine (NTHZ) and N-nitroso carboxylic acid (NTCA) in smoked cured meat.

2.4.2.6 Inappropriate food storage

The introduction of refrigeration as well as other improved food storage technology or processes may be responsible for the universal decline of gastric cancer rate worldwide. This may be due to the reduction of exogenous nitrite production by bacterial

40 conversion of nitrates and a decline in other possible carcinogenic metabolite production by natural food microflora (Hwang et al., 1994; Lee et al., 1995).

2.4.2.7 Helicobacter pylori infection

Peptic ulcers and infection with Helicobacter pylori were common in people diagnosed with gastric cancer. It was suggested that H. pylori alters the integrity of the stomach lining, therefore promotes cancer formation (Hwang et al., 1994). Furthermore, Shapiro and Hotchkiss (1996) demonstrated that viable H. pylori cells stimulated the most NO synthesis in in vitro murine macrophage system by the L-arginine-NO pathway and exhibited dose-dependent relationship. Thus chronic H. pylori may increase endogenous NO production leading to chronic inflammation in the gastric epithelium, which may initiate carcinogenesis associated with gastric cancer. The decrease in vitamin C secretion during H. pylori infection also increased the nitrosation potential in vivo due to decreased antioxidant scavenging of nitrosating agents by vitamin C. Elevated exposure to NO is also potentially genotoxic and mutagenic in bacterial and mammalian cell systems, and can form NOC in the presence of secondary amines (Shapiro and Hotchkiss, 1996).

Chronic H. pylori infection leads to constant nitric oxide production that may lead to tissue and DNA damage thus increases the risk for developing cancer. Rieder et al. (2003) investigated the association between chronic H. pylori infection and iNOS expression in samples from stomach carcinoma patients including biopsies from patients with H. pylori associated gastritis. Their results supported the hypothesis that CagA-positive H. pylori strains were associated with the expression and activity of iNOS, hence may contribute to the development of intestinal metaplasia leading to gastric cancer of the intestinal type.

2.4.2.8 Cigarette smoking

Like alcohol, cigarettes contain preformed nitrosamines therefore smokers are at more risk of gastric cancer compared to non-smokers or ex-smokers (Hwang et al., 1994). Similarly, Lee et al. (1995) showed a positive association between cigarette smoking and the risk of stomach cancer in case-control population in Korea.

Kato et al. (1992) suggested genetic link of stomach cancer between family members with history of the disease. They also concluded that smokers were more likely to die

41 from stomach cancers than those never smoked, but there was no dose-response to the amount of cigarettes smoked. There was a positive association between high fruit intake and reduced risk of stomach cancer. Furthermore, Kato et al. (1992) did not find any association between different cooking methods for meats and fish and the risk of developing stomach cancer. However, in their study only a specific population of rural Japanese living in mountainous area were surveyed, therefore their conclusion may be biased and is not representative of the general population.

Hwang et al. (1994) concluded that the consumption of nitrites, nitrates, alcohol and highly salted, pickled, fermented, or smoked foods increased the risk of gastric cancer. In addition, environmental factors such as H. pylori infection and cigarette smoking also increase the risk. They suggested that high intake of fruits and vegetables or antioxidants such as β–carotene, vitamin E and vitamin C may decrease the risk.

The duration of cigarette smoking was shown to be positively associated with gastric cancer risk among population-based case-control study in Sweden (Hansson et al., 1994b). Furthermore, they have shown that high fruit consumption was more protective among smokers than non-smokers. Hence smokers should be encouraged to increase their fruit and vegetable intake in order to lower their risk of gastric cancer.

Stomachs of smokers had significantly higher thiocynate levels than non-smokers, which were shown to be a potent catalyst of the nitrosation reaction (Ruddell et al., 1976). Ruddell et al. (1976) had shown that patients with gastric cancer and hypochlorhydria (under production of HCl by the stomach) had the highest level of stomach pH and similarly high nitrite concentrations. However, that latter group did not develop gastric cancer suggested that high dietary nitrite may not be linked to gastric cancer, but indicate other environmental or genetic factors are involved.

2.4.2.9 Low consumption of fruits and vegetables

Low intake of fruits and vegetables was common amongst people diagnosed with gastric cancers, since nutrient factors including vitamin C and E, fibre, low saturated fat are known to protect against numerous cancers including gastric cancer (Bartsch et al., 1990).

42 2.4.2.10 Others

Risch et al. (1985) concluded that chocolate or cocoa consumption was strongly associated with the increased trends in the risk of gastric cancer, but did not indicate any possible mechanisms.

Thiocynate is present in relatively large amounts in vegetables of the Brassica family, such as cauliflower, and is known to stimulate the rate of nitrosation of nitrite and secondary amine at low pH (Walters, 1980).

Indoles are present in several plant species and are present in high concentrations in the Brassica family. This is significant as naturally occurring indoles can be readily nitrosated under mild conditions (Tricker and Kubacki, 1992).

In summary, high consumption of red or processed meat might increase colorectal cancer risk, and the consumption of salted fish actually increased risk (Marques-Vidal et al., 2006). The consumption of fruit and vegetables may have beneficial and protective effects against gastrointestinal cancers, whereas alcohol consumption and overweight may increase the risk (van den Brandt and Goldbohm, 2006). The heterogeneity between prospective studies makes it hard in analysing and drawing conclusions from them (Marques-Vidal et al., 2006). Therefore in order to overcome this flaw and faciliate communucations between epidemiologists a standardized method need to be developed and implemented.

2.4.3 Estimated dietary nitrate and nitrite intake

Many studies have been done to estimate the dietary nitrate and nitrite intake in specific regions, countries and/or continents. This will be summarized below.

2.4.3.1 Asian-pacific regions

The average daily consumption of dietary nitrate in Singapore was estimated to be around 215 mg, with Chinese flowing cabbage and kale contributing to half this intake (Dutt et al., 1987). This was significant, as they have identified the two main sources of dietary nitrate amongst Singapore population, which may explain the prevalence of gastric cancer in that country.

43 Pennington (1998) mentioned several flaws in dietary surveys used in compiling national dietary data, such as the lack of honesty or poor food recall in the participants, which will contribute to the accuracy of the population dietary surveys. In addition, the author pointed out a major flaw in national dietary survey that was the lack of data on unusual food or food groups in certain populations. Australia being a multi-cultural country with big ethnic and religious diversity makes its national nutrition survey possibly an underestimate of the true food consumption pattern and hence the data collected may not be representative for the whole nation.

Prospective studies can reduce several kinds of biases associated with data collection, but requires reliable and complete quantitative estimates of nutrient and food consumption in the population of interest (Kato et al., 1992). Australia is currently renewing its national food consumption data in accordance with the most recent findings. This will enhance the accuracy of existing data made eight years ago since dietary preferences had changed dramatically in recent years as a result of improved food technology and analysis.

2.4.3.2 United States and United Kingdom

Gangolli et al. (1994) estimated the mean daily intake of nitrate and nitrites in the US were 106 and 1.5 mg/kg, respectively. In the UK, the estimated nitrite intake in 1979 was between 0.32 and 0.87 mg/day. However, in 1985 the estimate was risen to between 2.4 and 4.2 mg/day with potatoes contributing the major source of dietary nitrite of up to 43 %. It was suggested that the increase was not due to a real increase in dietary intake, but due to the restructuring of the total diet study since 1981. Nevertheless, between 1979 and 1985, it was demonstrated that vegetable food group contributed 75 % of the total nitrate intake from the diet. (Meah et al., 1994). More recently, Gangolli et al. (1994) estimated the mean daily intake of nitrate and nitrites in the UK were 104 and 1.5 mg/kg, respectively.

The differences observed may also be due to underestimation from interfering matrix naturally occurring in plants, such that Hunt (1994) measured nitrite in fresh vegetables and reduced interference from ascorbic acid, tannins and nitrate reductase by extracting the nitrite with zinc acetate at approximately pH 7.3 with activated carbon, since these compounds can react with nitrite and result in underestimation. Hunt and Turner (1994) demonstrated no detectable nitrite in 93.4 % of edible fresh retail vegetables, with

44 potatoes containing less than 0.4 mg/kg nitrite. Thus it was concluded that potatoes would contribute to dietary nitrite of the vegetable food group.

2.4.3.3 Rest of Europe

In 1994 using 24-hour duplicate diet and HPLC-UV method, van Vliet et al. (1997) estimated the mean intake of nitrate in the Dutch population was 80 mg/day per person, and the median intake of nitrite was 0.1 mg/day per person. They showed that the daily intake of nitrate and nitrite was higher than found in the 1984/1985 studies, which may be due to underestimation in the early study, or a better detection method employed in the latter study.

Italy had a mean daily intake of nitrate of 245 mg/day, whereas Poland and Switzerland mean daily nitrate intake were 178 and 125 mg/day, respectively. France’s mean daily intake of nitrate and nitrite were 150.7 and > 3 mg/day, respectively, followed by Netherlands, Germany and Norway mean daily nitrate intake of 71, 68 and 43 mg/day, respectively. The mean daily nitrite consumption in those countries was 0.6, 2.6 and 1.8 mg/day, respectively (Gangolli et al., 1994).

The acceptable daily intake (ADI) for nitrate was set at 3.7 mg/kg body weight by the European Union Scientific committee for Food (1995) and since nitrite has higher acute toxicity than nitrate its ADI was set at 0.06 mg/kg body weight (Reinik et al., 2005). Reinik et al. (2005) estimated the mean daily intake of nitrate was 1.7 mg/kg body weight for Estonian children between 2000 and 2005 and for nitrite and N-nitrosamines were 0.83 mg and 0.073 µg, respectively. They found that the nitrite intake exceeded the acceptable daily intake by up to 140 % for 1 to 6 years-old children in 2000-2001 and up to 105 % in 2003-2004. However, the concentrations of nitrate, nitrite and N- nitrosamines showed a decrease every year between 2000 and 2005.

2.4.3.4 Summary

Walters (1980) reported the estimated total intake of nitrite ion was 12.15 mg/day. Of that amount, 0.02 mg was present in foods, 1.20 mg added to foods, 10.0 and 0.06 mg from nitrate in vegetables and water respectively, and 0.33 mg formed endogenously in the stomach. According to Walters (1980), vegetables contributed up to 82 % of dietary nitrite, whereas other food sources (< 0.002 %) and added nitrate or nitrite salts (0.1 %)

45 were insignificant. However, the author did not account for the remaining nitrite sources of 0.54 mg, but may likely to come from salivary nitrate.

Pennington (1998) estimated that the daily nitrate intake ranges between 53 and 350 mg/day depending on the type and quantity of the vegetable consumed and the level of nitrate in drinking water. Whereas daily nitrite consumption was between 0 and 20 mg/day depending on the levels of nitrite present in cured meat and of much of it was consumed.

According to Cornee et al. (1992), the average daily nitrate intake per person per day was 121 mg (85 % from vegetables, 5 % from preserved and cured meat, and 5 % from cereal products). For the average daily nitrite intake per person per day, it was found to be 1.88 mg (43 % from vegetables, 28 % from cured meat, and 16 % from cereals). The remaining 13% of nitrite must come from non-dietary sources of nitrite such as atmospheric contamination.

A simple, rapid and accurate spectrophotometric method is available for determining the concentration of nitrates in human biological fluids using NADPH oxidation by Aspergillus nitrate reductase (Gilliam et al., 1993), or for the simultaneous detection of nitrate and nitrite in a variety of fluids using nitrate reducing vanadium III and detection by acidic Griess reaction (Miranda et al., 2001). These methods are necessary for long-term monitoring of nitrate and nitrite concentrations in susceptible populations. In addition, it may be useful as an alternate method to estimate nitrate and nitrite intake in populations where dietary history and food composition data are not available, or provide a biochemical confirmation to those reported in dietary history or national nutrition surveys.

Pennington (1998) advocated the Total Diet Survey (TDS) in the United States that focuses on core foods and allows development as more foods were analysed from groceries and restaurants rather than using data from a database. This then provides a foundation for estimating variances for the food components in each food.

46 2.4.4 Recommended dietary nitrate and nitrite intake

2.4.4.1 International

Many countries have restricted the amount of nitrate and nitrite to be added in food and drinking water demonstrate the epidemiological significance of these chemicals in the etiology of human gastric cancer.

ADI set by the Scientific Committee for Food for nitrite is 0 - 4.2 mg/60 kg person/day and 0 – 219 mg/60 kg person/day for nitrate. On average, 0.1 to 4.2 mg of nitrite from a mixed diet was consumed per person per day. The variation observed was dependent on the method of estimation. On average, 54 to 157 mg of nitrate from a mixed diet was consumed per person per day. The variations observed were dependent on the method of estimation (total diet, food frequency, duplicate diet)(Massey, 1997). Similarly, ADI set by the EU Scientific Committee for Food for nitrate was 3.7 mg/kg body weight (Reinik et al., 2005).

This has led to the restriction of nitrate and nitrite intake in 1973 by the Joint FAO/WHO Expert Committee on Food Additives (JECFA), which set an acceptable daily intake of sodium nitrate from 0 to 300 mg for a 60 kg person, and 0 to 12 mg per 60 kg person of sodium nitrite per day (WHO, 1973).

Different countries have set their maximum limits for the addition of nitrate and/or nitrite salts in cured meat. For example, in US the maximum allowed sodium nitrite in cured meat is 156 mg/kg, but in practice, often less than 120 mg/kg is used. The residual nitrite is the amount analytically detectable in the cured meat, and is much lower than the actual amount added (Cassens, 1995). In UK, the maximum allowed sodium nitrate and sodium nitrite for uncooked bacon and ham is 500 mg/kg combined with the latter not exceeding 200 mg/kg (Dennis et al., 1990). In Spain, the maximum allowed addition of potassium nitrate in cooked meat products was 200 mg/kg and less for sodium nitrite at 125 mg/kg. The combined used of nitrate and nitrite salts was allowed at 250 mg/kg (Pérez-Rodríguez et al., 1996).

2.4.4.2 Food Standards Australia and New Zealand (FSANZ)

Under the Australian Food Standard Code 1.3.1 schedule 1, 125 mg/kg of nitrite (potassium or sodium salts) are permitted in cured, dried, and slow dried cured meat;

47 whereas in commercially sterile and canned cured meat, the maximum nitrite (potassium or sodium salts) permitted is 50 mg/kg. For slow dried cured meat, the maximum allowed nitrate (potassium or sodium salts) is 500 mg/kg (FSANZ, 2009). Given the established antimicrobial effect of nitrite salts, particularly in reference to Clostridium botulinum in cured meat, its level should remain sufficient to prevent the occurrence of foodborne illnesses (Cassens, 1995), since public health and safety is the number one priority in the Australian and New Zealand food standards code (FSANZ, 2009).

Since nitrate and nitrite can cause methaemoglobinaemia in infants and young children, as well as contributing to the body’s total burden of NOCs, the National Academy of Science (1981) report recommended the elimination of nitrate salts in curing except for some products, and suggested reducing the nitrate level of vegetables and drinking water. In addition, the report recommended the further study of nitrosation inhibitors.

N-nitrosoproline (NPRO) test had been used to quantify human exposure to endogenous NOCs and/or methylating agents. The test is based on the excretion of NPRO and other N-nitrosoamino acids in the urine, which are measured as an index of endogenous nitrosation. However, the determination of nitrate and nitrite alone in the body fluids did not indicate the extent of in vivo nitrosation process (Bartsch et al., 1990).

Australia’s food composition table set by Food Standards Australia and New Zealand (FSANZ, 2009) did not include nitrate and nitrite. From recent improvements in food analysis and policies, Australia had finally produced a national food composition table using data mostly gained within Australia. Cashel and Greenfield (1996) had shown that these new Australia-based data gave different national dietary references and hence led to new dietary guidelines and core food groups. For example, the 2000 target for total fat of less than 30 % energy had almost been met for 1990-91 data (Cashel and Greenfield, 1997). This finding is significant since many dietary components such as high saturated fat and salt intake had been associated with gastric cancer, whereas high dietary fibre intake had been associated with lower risk of gastric cancer (Bartsch et al., 1990).

Since nitrite can be more active in synthesising NOCs when ingested in one application over a short period of time, than ingestion of low concentrations of nitrite continuously

48 over a long period (Walters, 1980). If these restrictions are enforced and monitored, the average dietary consumption of low nitrite concentrations over time may not pose a greater risk to developing gastric cancer.

2.5 Determination of nitrate and nitrite

With advancing qualitative and quantitative methods: selectivity, sensitivity, time and costs are some important factors for the determinant of the method used for analysis. However, few methods exist that has all the above quality, thus the choice of method is dependent on the specific goals of the researcher. Nevertheless, several potential methods will be described below as well as procedures to improve detection accuracy.

Homogenization of food is difficult but often necessary prior to chromatographic analysis. Factors such as variable texture, structure and the presence of immiscible phases may hinder the homogenization process (Lichon, 1996). Because these properties are inherent properties of the food product, it cannot be changed during the manufacturing process. The sampling of food and their preparation must therefore be considered carefully prior to analysis to ensure representative and accurate results. This means using traditional cooking practices so that factors that affect nitrate and nitrite determination and their recovery are consistent and applicable to dietary exposure of these anions.

2.5.1 Colourimetric and spectrophotometric

Torró et al. (1998) reported good agreement between flow-injection biamperometric method with the reference method of AOAC Griess reaction in the determination of nitrate and nitrite in meat products. Major advantages in their proposed method were its high selectivity, simple design, low cost and the avoidance of carcinogenic reagents and toxic metals as required by the Griess reaction.

It was demonstrated by Wang et al. (1998) that precolumn concentration using diazotization-coupling reaction for nitrite and cadmium-copper column for nitrate before spectrophotometric determination of both nitrates and nitrites in water and some vegetables provided high sensitivity, unlike standard spectrophotometry method, high selectivity and accuracy, unlike flow injection methods, and simplicity and low cost, unlike ion chromatography.

49 Similar to the reference spectrophotometric method, Zatar et al. (1999) determined nitrite and nitrate in water, meat and vegetable samples by using the oxidation of nitrite to the phosphomolybdenum blue complex resulting in a reduction in intensity of the blue color, whereby its rate of reduction was directly proportional to the amount of nitrite added.

A completely automated method for the determination of nitrite, nitrate and chloride in cured meat was developed by Velasco-Arjona et al. (1998), which was based on the coupling of discontinuous (robotic) and continuous (flow injection) systems. This was shown to be accurate when compared with the traditional methods, as well as reducing significant amount of analysis time. However, the set-up seemed to be much elaborated and complex, and requires extra computing knowledge in order to set-up and interpret the results. Nevertheless, once set-up with fully trained technicians, this automated system should save time and will allow large sampling size to be tested.

Clark et al. (2003) measured the nitrate intake in vegetarian duplicate diets in the UK using HPLC with UV detection at 214 nm. They concluded that vegetarians’ nitrate intake was similar to other consumers’ and were both under the acceptable daily nitrate intake. However, their extraction method of nitrate involved a lot more chemical reactions. Firstly, they added hot alkali solution containing borax solution and water. After mixing and cooling, Carrez solutions I and II were added followed by further shaking. Samples were then cleaned up using solid phase extraction on endcapped cyclohexyl cartridges prior to analysis. Separation occurred on a Dionex AS11 column with the same material for the guard column, and was eluted with water and sodium hydroxide mixture (80:20) at 1 ml/min.

Rincón et al. (2003) reviewed the Association of the Official Analytical Chemists (AOAC) method for the detection of nitrite in meat products, and recommended that the reagents Carrez II and I should be eliminated because they do not provide efficient clean up and interferes with nitrite extraction. In addition, they recommended the use of borax reagent, as it is highly efficient in extracting nitrite during heat treatment. More recently, Stalikas et al. (2003) demonstrated a relatively sensitive, fast and accurate method for the simultaneous determination of nitrate and nitrite in cured meat by using ion chromatography method with post-column indirect fluorescence detection.

50 Determination of nitrate and nitrite in complex sample matrices including biological fluids and food products by vanadium (III) reduction with chemiluminescence detection was shown to be rapid in terms of analysis time, less than four minutes per sample, and simple without the need for sample pretreatment (Hendrix and Braman, 1995).

Miranda et al. (2001) developed a sensitive, rapid, inexpensive and simple method for the simultaneous detection of nitrate and nitrite based on nitrate reduction by vanadium (III) combined with detection by the Griess reaction. Furthermore, it was applicable in a range of media including buffer, media and biological fluids.

A flow-injection spectrophotometric method was used for the simultaneous determination of nitrate and nitrite in meat and vegetable samples by the reduction of nitrate to nitrite in a cadmium copper reductor column, and the reduction of nitrite to nitric oxide in a sulphuric acid medium. The range of 0.30-3.00 and 1.00-10.00 mg/L was the detection range for nitrite and nitrate respectively, and the precision and accuracy of this method were comparable to the reference spectrophotometric method developed by the AOAC (Andrade et al., 2003). Similarly, flow injection catalytic spectrophotometric method was used to simultaneously determine nitrates and nitrites in water with rapid speed and sensitivity at 0.3 ng/ml for nitrite and 1.0 ng/ml for nitrate as the detection limits (Yue et al., 2004).

2.5.2 Gas chromatography

Gas chromatography has been used for the measurement of nitrate and nitrite in foods and water. It is based on the formation of a volatile derivative, followed by extraction into organic solvent, which is then measured by gas chromatography using a selective detector. Under good conditions, the detection limit for this method is as low as 5 μg/L of aqueous samples. This method also showed good agreement with the traditional colorimetry assay for the analysis of cured meat, and it is less time-consuming (Massey, 1991).

2.5.3 High performance liquid chromatography

Wootton et al. (1985) concluded that HPLC method for the determination of nitrate and nitrite in meat, fresh and canned vegetable products was satisfactory. However, some meat products were subject to matrix interference. For example, sodium chloride

51 increased the apparent levels of nitrate and nitrite, whereas ascorbic and erythorbic acids decreased and increased the apparent levels of nitrite and nitrate, respectively. Despite the apparent shortcomings of HPLC, it offered a simple, rapid and practical means of determining both nitrate and nitrite simultaneously.

Dennis et al. (1990) used a rapid strong anion exchange HPLC-UV method for the determination of nitrate and nitrite in a wide range of cured meat. They have confirmed the accuracy of this technique by comparing to the existing British Standard colorimetric method. The traditional colorimetric method for the determination of nitrate and nitrite is time consuming and requires maintenance of the efficiency of its cadmium reduction column. In contrast, HPLC was demonstrated by Dennis et al. (1990) to be a better alternative with its speed and elimination of using the toxic metal cadmium. However, the additional costs of HPLC equipment, column and disposable Bond Elut cartridges may represent a significant disadvantage of this technique. Based on their method of nitrate and nitrite determination in cured meat, Reece and Hird (2000) were able to modify it for a more accurate nitrate and nitrite determination in dairy products that was comparable with the AOAC method.

Most of the chromatographic methods mentioned above used UV detection for the determination of nitrate and nitrite ions. However, Di Mattel and Esposito (1997) concluded that HPLC with electrochemical detection (ED) was successfully used for the determination of nitrate and nitrite ions in food, biological and environmental samples and was shown to be more sensitive, selective and faster than methods based on UV absorption, photometry, fluorometry or chemiluminescence. The detection limits for nitrite in fish and cured meat ranged from 1 to 50 ppb.

Reversed-phase (RP) C18 columns are commonly used for separating nitrates and nitrites due to their reproducibility in wide range of pH, relatively low cost and versatility. The separation is based on hydrophobicity thus polar compounds are eluted faster then non-polar compounds (Hanai, 1999).

Butt et al. (2001) had introduced an application of normal phase ion-pair liquid chromatography for the analysis of inorganic anions with tetraethylammonium as the ion-pairing agent. They had developed a relatively simple, sensitive, selective and rapid method for the simultaneous determination of nitrite and nitrate in spinach and lettuce samples. For nitrate and nitrite extraction, they blended 50 g of each sample in a

52 blender with 50 ml of double deionized distilled water for 5 min at 60 oC. They then further homogenized the samples ultrasonically using Bransonic-52. Followed centrifugation, the supernatant was passed through a 0.25 µm Millipore syringe filter and then through a C18 sample clean-up column for direct injection into the liquid chromatograph.

Siu and Henshall (1998) used a similar extraction method as Stalikas et al. (2003). But in addition to salami they also tested ham. They also used a higher centrifugation speed at 4960 g (6000 rpm Beckman GA-10 rotor) for 15 min. Siu and Henshall (1998) then filtered the supernatant through Whatman No.2 and GF/A filters, and then through the 1.2 µm and 0.2 µm Acrodisc filters. The filtrate was then analyzed for nitrate and nitrite using pellicular anion-exchange columns with UV detection at 225 nm. This method was developed to remove most of the potentially matrix interferences such as protein and fat-based substances, and the column chosen provided good selectivity for the separation of nitrate and nitrite, as well as for the separation of other residual matrix components. Nitrite and nitrate content in salami were 108.0 and 98.5 mg/kg, respectively, and 11.6 and 5.4 mg/kg for ham, respectively.

2.5.4 Ion chromatography

Due to interferences from meat samples, Jackson et al. (1984) used low capacity anion- exchange column for the determination of nitrate and nitrite in cured meats. It was shown to be rapid, sensitive and precise with satisfactory recovery of added nitrate and

53 nitrite. In addition, it was shown to be applicable to vegetables and cheese. However, their results did not agree with spectrophotometric method, but they attributed the differences to the deficiencies of the conventional method.

Siu and Henshall (1998) minimized interferences from meat products by using ion chromatography with specific UV absorbance detection. The recovery of nitrate and nitrite were greater than 90 % with detection limits of 50 and 30 μg/l for nitrate and nitrite, respectively. They also used extraction procedures to remove majority of the interfering components, which used then be analyzed within 12 hours to avoid changes in nitrate and nitrite concentration in the meat samples.

Butt et al. (2001) demonstrated the applicability of ion-paired liquid chromatography for the simultaneous determination of nitrite and nitrate in complex food matrices, and was shown to be relatively simple, sensitive, selective and fast. Although ion chromatography and HPLC methods have been shown to be faster, more accurate and sensitive than spectrophotometric methods, these relatively newer methods are often time consuming, require pre-treatment and can be expensive (Badea et al., 2001). Kazemzadeh and Ensafi (2001) used a flow-injection spectrophotometric method for the simultaneous determination of nitrite and nitrate in food samples including meat and cheese and showed good reproducibility and accuracy in comparison to the standard method. The detection limit for nitrite and nitrate were 0.001 and 0.010 μg/ml, respectively. However, this method required the use of specific columns and reagents, including cadmium, but was fast processing about 20 samples per hour.

Stalikas et al. (2003) homogenized ten grams of salami using a blender for 1 min with 70 ml of warmed double-distilled water. Then the mixed sample was heated for 15 min and maintained at 70 oC. After cooling to room temperature, the sample was centrifuged at 1370 g for 10 min, then the supernatant was removed and filtered successively using a Whatman filter paper No. 40 and a GF/A filter. The filtrate was then diluted and collected in a volumetric flask and analyzed using ion chromatography with post-column indirect fluorescence detection. It was shown that this coupling technique was relatively sensitive, fast and accurate for the determination of nitrate and nitrite in meat samples. Matrix interferences were eliminated with fluorescence detection since components that may decrease the background fluorescence of tryptophan are rarely found in real samples.

54 Overall, the theoretical column efficiency of ion-pair liquid chromatography is much better than that of an ion-exchange column. Hence ion-pair liquid chromatography is much more commonly used. In addition, the regeneration of ion-pair column is much faster therefore provides a faster turn over (Hanai, 1999).

2.5.5 Capillary electrophoresis

Using AOAC as reference method, Jimidar et al. (1995) demonstrated the precision and accuracy of capillary electrophoresis with indirect detection for the determination of nitrate and nitrite in vegetables. The major advantage of the capillary electrophoresis method compared to the traditional spectrophotometric method is that the former is extremely fast (around five minutes) as oppose to the latter method (about one hour).

Marshall and Trenerry (1996) used capillary ion electrophoresis to determine the nitrite and nitrate content of a variety of foods including vegetables, water and meat products. They concluded that the method was rapid, simple, and reliable and gave good recoveries of nitrite and nitrate in a variety of foods when using thiocyanate as the internal standard.

Öztekin et al. (2002) used capillary electrophoresis with direct UV detection for the simultaneous determination of nitrite and nitrate in meat products and vegetables, which was demonstrated to be rapid with sufficient detection limit for food samples. In addition, capillary electrophoresis uses much less electrolyte and sample consumption than colorimetric method, and is less expensive and more accurate than chromatographic columns due to minimal interfering matrix.

Öztekin et al. (2002) used capillary coated with a cationic polymer polyethyleneimine in capillary electrophoresis to simultaneously detect nitrite and nitrate in meat products and vegetables. Their method showed very short analysis time, good reproducibility, less expansive than chromatographic columns, low electrolyte and sample consumption and good recovery for the anions. The detection limit for nitrite ions was 4 mg/kg.

The simultaneous determination of nitrite and nitrate by ultra-rapid capillary electrophoresis was demonstrated to be much more rapid and environmentally friendly with excellent detection limits for both anions compared to standard electrophoresis, HPLC, ion chromatography and flow injection analysis (Melanson and Lucy, 2000).

55 However, it was not demonstrated on samples with complex matrices such as food and biological fluids. Furthermore, no recovery was performed to ensure accuracy.

2.5.6 Electronic sensors

New electrochemical sensors were used by Badea et al. (2001) for rapid amperometric detection of nitrites and nitrates in water, and was suggested that this method could be applied to food samples for the rapid and simple determination of nitrates and nitrites.

2.5.7 Others

Basing their method on the generation of nitric oxide and polarography, Ximenes et al. (2000) were able to recover 85.4 to 107.4% of the nitrate in various vegetable samples, with precision and accuracy comparable to those of the standard AOAC spectrophometric method.

Escherichia coli cell method in the measurement of nitrate was based on nitrate reduction by the wild-type E. coli DSM 498k cell under aerobic conditions, followed by nitrite determination with the Griess reaction. The nitrate contained in meat was found to be 217 mg/kg. This method was simpler than those requiring inorganic reductants and purified enzymes, and can be adapted for nitrate measurement in other matrices (Xu et al. 2000).

Chemiluminescence techniques are 600 times more sensitive than the colorimetric methods. Furthermore, traditional colorimetric methods can underestimate the amount of nitrite present in the presence of reductants such as ascorbate. However, this limitation in the colorimetric method can be prevented by the addition of charcoal that removes these reductants (National Academy of Science, 1981). Thus the chemiluminescence method for nitrite determination was recommended by the National Academy of Science (1981) for cured meat and biological samples because of its accuracy, sensitivity, and lack of interferences from the food matrix.

56 2.5.8 Food

There are many methods available for the determination of nitrates and nitrites in food samples including cured meat and vegetables. These methods mainly consist of high- performance liquid chromatography, capillary electrophoresis and ion chromatography. The major advantages with capillary electrophoresis are the speed of separation of nitrate and nitrite anions (approximately 10 seconds), and that less waste is generated due to the absence of continuous mobile phase. However, one major drawback with capillary electrophoresis is its lowered sensitivity when compared with high- performance liquid chromatography. Ion chromatography is perhaps the most suitable method for anion separations, but in terms of sensitivity, accuracy, limits of detection, and speed of analysis, high-performance liquid chromatography is very similar, but more robust with wider applications. However, all the methods mentioned above would require some form of sample pretreatment, which may take up to 1 h, hence is the most time consuming step of the total analysis time. Following are the recent methods used for extracting and detecting nitrate and nitrite in food.

Since both nitrates and nitrite are readily soluble in water, they can be easily extracted by hot water. However, because of nitrite’s high reactivity at acidic pH, alkaline or near neutral conditions are preferred for nitrite extraction. Aqueous extraction of nitrite from meat gave 97 % recovery at pH 5.5 to 6.5, compared to 44 % recovery at pH 3.5 and 92 % recovery at pH 9.5. Because some nitrites are bound to meat compunds such as nitrosothiols and nitrosylmyoglobin, most hot water extraction technique only measures the free nitrite ions. Heavy metal ions such as mercury and silver can cause cleavage of the nitrosyl ion from NOCs hence releases some bound nitrite. Whereas an alkaline condition was shown to degrade nitrosothiols into NO+ that occurs rapidly, which also measures some bound nitrite (Usher and Telling, 1975).

Oxidizing and reducing agents such as ascorbic acid and ascorbate is known to interfere with nitrate and nitrite determination. However, the addition of activated charcoal such as Darco 60 to the extraction solutions can remove any ascorbate. It was also shown that the two-hour water bath extraction technique in the AOAC colorimetric method could eliminate ascorbate. Other interferences such as sulfur dioxide often added to cured meat can interfere with nitrite determination. However, it was demonstrated that recovery increases as the meat solids decreases (Usher and Telling,

57 1975). Hence by using smaller meat samples should minimize the effects of interfering substance.

Butt et al. (2001) had introduced an application of normal phase ion-pair liquid chromatography for the analysis of inorganic anions. They had developed a relatively simple, sensitive, selective and rapid method for the simultaneous determination of nitrite and nitrate in spinach and lettuce samples. For nitrate and nitrite extraction, they blended 50 g of each sample in a blender with 50 ml of double deionized distilled water for 5 min at 60 oC. They then further homogenized the samples ultrasonically using Bransonic-52. Followed centrifugation, the supernatant was passed through a 0.25 µm

Millipore syringe filter and then through a C18 sample clean-up column for direct injection into the liquid chromatograph.

58 Butt et al. (2001) demonstrated that the presence of 50-fold sulfate and chloride did not affect the resolution and percentage recovery of nitrite, but did reduced the resolution and recovery of nitrate. Similarly, 1-2 % salt caused more losses of nitrate than for nitrite. In addition, the presence of magnesium, iron and calcium significantly reduced the percentage recovery of both anions. However, excess iron was suppressed by the addition of ethylenediaminetetraacetic acid (EDTA). Furthermore, they had also demonstrated that under optimized conditions, both nitrate and nitrite peaks began to merge when the concentration of nitrite was above six-fold of nitrate concentration.

Although most interference can be eliminated by using UV detection, large quantities of nitrate and chloride ions can still interfere with nitrite detection in HPLC. UV may detect chloride ions as positive or negative peaks in the wavelength used for nitrate and nitrite, but are eluted before nitrite (Di Mattel and Esposito, 1997). Hence, dilution is used before nitrate and nitrite extraction to minize the effects of interferences.

Ionic compounds such as nitrate and nitrite can be extracted from food at above 57 oC, at lower temperatures cell membranes are impermeable to solutes, whereas at higher

59 temperatures between 80 and 90 oC the maximum leaching rate of ions can be achieved (Gaiser et al., 1996).

Spinach contains varied amount of nitrate depending on the growth conditions and the agricultural practices. However, it is generally in the range from 50 to 5600 mg/kg with a mean nitrate concentration of approximately 2000 mg/kg (Gaiser et al., 1996). The cooking or blanching of spinach prior to preservation significantly reduced the content of nitrates and nitrites. In addition, it was found that preservation by canning resulted in reduced nitrate and nitrite content over a 12 month period compared to preservation by freezing (Jaworska, 2005).

Due to its reactive nature, nitrite analysis from food does not give a true representation of the total nitrite added. In addition, ascorbate accelerates the depletion of nitrite, but can be removed by reduction with iodine. Furthermore, nitrite added to meat is usually present as NO bound with other food components such as myoglobin (5-15 %), sulphydryl groups (5-15 %), lipids (1-5 %), proteins (20-30 %), as nitrate (< 10 %), and as free nitrite (10-15 %)(Zanardi et al., 2002). Because only free nitrite can participate in nitrosation, current methods of food extraction usually estimates the total nitrite present by releasing food-bound nitrite. This may over emphasize the significance of dietary nitrite and the aetiology of gastric cancer.

Based on literature comparisons, HPLC-UV is used for quantifiying nitrate and nitrite in foods for this experiment because it is sensitive, fast, accurate and applicable for both cured meat and vegetables.

2.5.9 Biological fluids

Li et al. (2000) used a simple, highly sensitive and specific and rapid method for the determination of nitrite in biological samples by reducing nitrate to nitrite using nitrate reductase and then derivatizing total nitrite with 2,3-diaminonaphthalene (DAN) to form a highly fluorescence compound 2,3-naphthotriazole (NAT) followed by detection using HPLC coupled with fluorescence detector.

Since the activation of the immune system is associated with an increased production of NO, by measuring the nitrate and nitrite levels in the plasma and urine, it may allow for the monitoring of NO production and possible disease activity in vivo. Approximately 30% of the total body water is found in the extracellular fluid such as

60 plasma, urine and lymph, and the rapid transport and exchange of metabolites to the extracellular fluid provides a good indicator for those metabolic processes that occur at the cellular and tissue level. So the measurement of the excreted extracellular fluids such as urine and saliva provides a simple and non-invasive method for quantifying systemic NO production (Grisham et al., 1995).

However, even though nitrate and nitrite are commonly used as markers for NO production in vivo, its quantification from patient’s urine and sera does not represent the total NO production, since nitrate and nitrite may also be derived from gut- or airway-associated bacteria in healthy individuals (Marzinzig et al., 1997).

It was demonstrated that the enzymatic reduction of nitrate to nitrite by commercially available nitrate reductase was an inexpensive, simple and sensitive method for the determination of total body nitrate and nitrite (Grisham et al., 1995).

The only stable product from the spontaneous decomposition of NO is oxygenated solutions is nitrite. However, nitrite derived from NO is rapidly converted to nitrate via oxidation by certain oxyhemoproteins such as oxyhemoglobin or oxymyoglobin. This explains the low levels of nitrite found in the excreted extracellular fluids (Grisham et al., 1995).

Griess reaction, a colorimetric assay for nitrite, combined with enzymatic conversion of nitrate to nitrite, was suggested to be a cost effective method for measuring urinary nitrate and nitrite levels. In addition, solid phase extraction and pH adjustment of the urine minimized interferences and improved activities of the nitrate reductase to obtain adequate sensitivity and accuracy (Ohkawa et al., 1998).

Bacterial nitrate reductase from E. coli coupled to the Griess reaction was shown to be a sensitive, accurate and inexpensive method for the determination of nitrate and nitrite in biological fluids and in tissue cultures with detection limit at µM level (Granger et al., 1995). Furthermore, Marzinzig et al. (1997) modified the conventional Griess reaction by replacing sulfanilamide with dapsone (4,4’-diamion-diphenylsulfone) coupled with bacterial nitrate reductase, which gave improved detection limit of nitrate and nitrite ions at 0.5 µM in sera, urine or culture media. In addition, they had demonstrated that bacterial nitrate reductase was superior in terms of sensitivity and linear range to reducing metal cadmium to convert nitrate to nitrite. However, in their

61 modified method, the retention time for nitrite and nitrite were 5 and 10 minutes, respectively, together with a costly experimental set up, may not be practical for routine analysis.

Because NO has a short half-life, nitrate and nitrite is measured as a marker for NO production in biological samples as they are the major metabolite of NO in blood (Bories et al., 1999). The measurement of plasma nitrite was shown to be a sensitive index of constitutive NO synthesis, which may be used as a marker for endothelial function. Furthermore, recent research suggested that nitrite represents a circulating reserve of NO and may selectively donate NO to hypoxic vascular beds (Dejam et al., 2004).

Similar to the methods used to determine nitrate and nitrite in food, the detection of these anions in biological fluids such as blood and urine are performed using a UV detector, where traditional methods measures the reduction of nitrate to nitrite either by a metal catalyst in HPLC, or by nitrate reductase using spectrophotometry after using the Griess reaction (Bories et al., 1999).

Bories et al. (1999) developed a method for the determination of nitrate and nitrite in biological fluids using capillary electrophoresis coupled with UV detector. In addition to the low cost, small sample and buffer requirements and rapidity (about 4 minutes per run), the recovery for nitrate and nitrite were 92-106 % and 93-115 %, respectively.

Nitrate and nitrite are usually present in low concentrations in biological samples whereas matrix interferences such as chloride are present at much higher levels and can often mask nitrite peak and proceeds immediately before nitrite peak in chromatographic elution. Therefore sample clean up is important prior to HPLC analysis in order to achieve good recoveries. For example, the use of silver resin to the sample can effectively reduce the chloride level (Jobgen et al., 2007).

Major direct dietary sources of nitrite include cured meat and cereals, however, 90 % of ingested nitrate, from green-leafy vegetable, can be reduced to nitrite by the action of nitrate reductase expressed by natural microflora on the surface of the mouth (Duncan et al., 1995).

Tsikas (2004) reported a simple, rapid GC-MS validated anion-pairing HPLC method with UV detection at 205nm for the accurate determination of urinary nitrate without

62 the need for sample pretreatment. However, this method does not allow for the simultaneous determination of both nitrate and nitrite, which are both commonly occurring anions in many foodstuffs.

Based on similar HPLC conditions as this study, Connolly et al. (2002) demonstrated that urinary sample pre-treatment involving centrifugation and filtering with solid- phase extraction on C18 cartridge was sufficient to remove organic interferences, resulting in both precise and sensitive method for the simultaneous determination of nitrite and nitrate in urine samples.

Traditional sample preparation involves the use of organic solvents followed by clean- up and pre-concentration steps, which are time consuming, labour intensive, costly and is bad for the environment. Hence SPME is gaining popularity as “green chemistry” which is solvent-less, economical and is less time consuming and includes preconcentration step allowing a lower limit of detection (Wardencki et al., 2007).

Nitric oxide synthases (NOS) in humans produces nitric oxide from L-arginine and the metabolic fate of nitric oxide includes oxidation to nitrate by oxyhaemoglobin and autoxidation in haemoglobin-free media to nitrite. Nitrate and nitrite circulate in the blood and are excreted in the urine. Circulating nitrite reflects constitutive endothelial NOS activity, whereas circulatory nitrate indicates systemic NO production. The concentrations of these NO metabolites in biological fluids are often used to measure NO synthesis in vivo under standardized low-nitrate diet (Tsikas, 2005).

It is well established that the major urinary metabolite of NO is nitrate. In addition, it was shown that endogenous NO synthesis follows a circadian variation (Tsikas, 2005); hence the time of urine collection may affect the concentration of nitrate in the urine. For example, the maximum and minimum excretion rate was between 5 and 8 pm and 5 to 8 am, respectively (Tsikas, 2005).

In healthy humans with uncontrolled nitrate diet, urinary excretion rate of nitrate was 1200 µmol/24h and for nitrite was 6 µmol/24h. Urinary nitrate and nitrite excretions were less in the same subjects on low-nitrate diet indicating that diet is a major contribution factor to urinary nitrate (Tsikas, 2005).

The main metabolic products of endogenously formed NO include oxidation to nitrate by oxyhemoglobin in red blood cells and autoxidation to nitrite. Both nitrate and nitrite

63 circulate in blood and are eliminated via renal excretion as urine (Wennmalm et al., 1993). It is known that circulating nitrite reflects constitutive endothelial NOS activity and that excretory nitrate indicates systemic NO production (Tsikas et al., 2006).

Nitrite concentrations in urine of healthy adults usually range in between 100 and 3000 µM, whereas nitrate concentration is usually in the range of 300 to 1400 µmol/day. Humans consistently excreted more nitrate than ingested from food. Nitrate in human urine is not entirely derived from nitrate in the food, but is endogenously produced at a relatively constant rate and independently of nitrate ingestion. Thus differences in urinary nitrate excretion are due to differences in the intensity of nitrate production (Tsikas, 2007).

GC-MS can be used to simultaneously determine nitrate and nitrite in human biological fluids with high sensitivity, selectivity and a low detection limit. However, it requires time consuming derivatisation step and sample pretreatment and cleanup (Helmke and Duncan, 2007).

Nitrite can be reduced to nitric oxide by dietary reductants such as anthocyanin and catechol (caffeic acid) from red wine as was demonstrated in healthy volunteers by measuring NO in the air expelled from the stomach following consumption of wine (Gago et al., 2007). Although NO is now recognized as an important signal molecule, the physiological significance of the amount generated from the consumption of red wine is not known. Furthermore, ascorbic acid (AA) can also reduce nitrite to NO and since AA is commonly occurring in foodstuffs the significance of red wine as a modulator for NO production is unknown.

NO function as a messenger molecule within the cardiovascular system and is a highly unstable free radical in circulating blood and is oxidized rapidly to nitrite and nitrate. However, under hypoxic conditions nitrite is reduced to bioactive NO by deoxyhemoglobin, which represent NO cycle that adapt to demand and supply by the vascular system. Endothelium-derived NO is synthesized from the amino acid L- arginine by the constitutive calcium/calmodulin-dependent endothelial NO synthase. NO autoxidizes exclusively to nitrite in acqueous solutions free of biological materials, however, in whole blood NO is oxidized by oxyhemoglobin almost completely to nitrate. Although HPLC can offer simultaneous and high-throughput detection of nitrate and nitrite without derivatization, it can suffer from interference by chloride

64 with UV detection due to insufficient resolution and column saturation. (Grau et al., 2007).

Urinary and plasma nitrate and nitrite levels not only reflects endogenous NO production, but also total dietary nitrate and nitrite as well as conversions made by bacteria found in the oral cavity and in the gastro-intestinal tract. Various detection methods are available for the simultaneous determination of nitrate and nitrite ions. However, major problems involving biological samples include matrix interferences, low sensitivity and lack of selectivity. Therefore sample preparation may be the crucial step in accurately quantifying these anions in biological fluids (Bryan and Grisham, 2007). Ellis et al. (1998) gave a concise review on nitrite and nitrate analyses from a clinical biochemistry perspective.

2.6 Food Industry and Regulations

2.6.1 Functions of nitrate and nitrite in meat-based products

Nitrite is used in meat curing for colour development, flavour enhancement and as a preservative. Nitric oxide (NO) is formed from nitrite under acidic conditions, and when it combines with the muscle pigment myoglobin, NO myoglobin is formed. This complex is relatively resistant to light and oxygen, but is very resistant to heat, which allows cured meat to maintain its red colour even after cooking. Doses of NaNO2 between 3 and 50 mg /kg are usually enough for this purpose. Similarly, flavour development could be achieved with 20 – 40 mg/kg NaNO2. Lastly, as a preservative,

80 to 150 mg/kg NaNO2 can inhibit the growth of numerous food-poisoning bacteria including Clostridium botulinum, Salmonella and Staphylococci (FAO, 1991).

The inhibitory effect of nitrite on bacterial spore formers such as Clostridium is due to inhibition of outgrowth during cell division. Nitrite acts in threes ways on bacterial cell metabolic processes: Firstly, nitrite interferes with energy conservation by inhibiting oxygen uptake, oxidation phosphorylation, and proton-dependent active transport; secondly, nitrite acts as an uncoupler, causing a collapase of the proton gradient; and finally, nitrite inhibits certain metabolic enzymes. Factors such as pH, oxygen, heating and other food components can affect the effectiveness of nitrite as an anti-Clostridial agent. Nitrites have found to be most inhibitory at acidic pH under anaerobic and heated conditions (Shafiur Rahman, 1999). However, the preservative effect of sodium

65 nitrite decreases with increasing storage temperature, therefore instructions for refrigeration is necessary to prolong shelf life and prevent food poisoning (FAO, 1991).

Due to the toxic effects of nitrite, it is often mixed with table salt at approximately 0.6 % of nitrite when used for curing. This is to ensure public health so that if excessive levels of nitrite are accidentally added to the final product, the accompanying salty taste should be rejected by the consumer, thereby preventing nitrite poisoning. In addition to the prevention of nitrite poisoning, salt is added to cured meat for flavor enhancement, is essential for the functional properties of meat proteins, and also act as a preservative (FAO, 1991).

Alternative preservation method such as gamma irradiation had been demonstrated to inactivate many microbes including the pathogenic C. botulinum in cured meat. However, because irradiation does not impart the characteristic flavor or color typical of cured meat, sodium nitrite will still be needed but in much less quantity (National

Academy of Sciences, 1981) from about 150 to 40 mg/kg NaNO2, which should significantly reduce the body’s burden of dietary nitrite intake.

In Mediterranean countries, the consumption of non-fermented sausages is increasing due to the dislake of intense acid flavour of fermented sausages, thus Sanz et al. (1998) examined the effects of nitrate and nitrite curing salt on microbial changes and sensory quality of non-fermented sausages. It was shown that by the end of the ripening process (26 days), the levels of microorganisms were similar in both batches of sausages except psychrotrophs such as lactic acid bacteria and yeasts were higher in those made from nitrite. However, nitrate-made sausages showed higher aroma and taste intensity. In contrast, nitrite discourages the growth of psychrotrophs in fermented sausages. It was concluded that nitrate-made sausages produced better sensory qualities without compromising microbiological safety.

Pegg and Shahidi (1996) elucidated the structure of pre-formed cooked cured-meat pigment (CCMP) using visible spectroscopy and concluded that CCMP is a mononitrosol, not a dinitrosyl. This was further supported as a pentacoordinated paramagnetic complex using electron paramagnetic resonance study.

It was shown that curing ingredients especially nitrate increased the production of important branched-chain flavor compounds in sausages (Olesen et al., 2004). Cured

66 color development during dry cured sausage processing was studied by Chasco et al. (1996) and it was shown that during fermentation nitrites reacted with myoglobin to form nitrosomyoglobin and metmyoglobin, which was reduced to nitrosomyoglobin during the drying process.

The development of foods with an attractive color is paramount since consumers are more likely to reject a food product based on its color. Furthermore, the additions of naturally occurring colorants are preferred due to their perception of healthy and better quality then those foods with artificial coloring (Bloukas et al., 1999). Bloukas et al. (1999) tested various natural colorants including curcumin in frankfurters but concluded that betanin followed by carminic acid and nitrite gave the highest score for overall acceptability. Furthermore, they had demonstrated that increasing the level of betanin could reduce the amount of nitrite added from 150 to 100 mg/kg.

2.6.2 Functions of ascorbate and erythorbate in cured meat

The salts of ascorbic acid and erythorbic acid in the forms of sodium ascorbate and sodium erythorbate are used in cured meat since they do not react with nitrite to form nitrous oxide. The roles of these antioxidants are four-fold. Firstly, ascorbate accelerates the rates of curing by reducing metmyoglobin to myoglobin; Secondly, ascorbate reacts with nitrite to increase the yield of nitric oxide from nitrous acid; Thirdly, as an antioxidant, excess ascorbate help stabilize both color and flavor by preventing heme-catalyzed lipid oxidation resulting in reduced pigment degradation and delayed rancidity; Finally, ascorbate have been shown to reduce nitrosamine formation under certain conditions. For example, at the ratio of 1:4.5 nitrite and ascorbate, respectively, has been shown to reduce or eliminate nitrosamine formation (Pearson and Gillett, 1996).

2.6.3 International food safety

The HACCP is a food safety management system that ensures the overall management of food quality and safety by identifying and managing critical points in food processing as part of the Good Manufacturing practices (GMP). Risk analysis forms a major part of developing standards, guidelines and other recommendations regarding food safety (FAO, 1995). However, HACCP is not applicable to all food industries, usually only to those food industry sectors with scientifically proven or good

67 epidemiological link of a particular food to public health risks. It is in these industries that are required to develop and implement HACCP plans (FAO, 1995).

The Uruguay Round of the General Agreement on Tariffs and Trade (GATT), the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS) and the agreement on Technical Barriers to Trade (TBT) are all administered by the Codex Alimentarius Commission to ensure fair trading in international food trade as well as to ensure public health by establishing international food standards to minimize food related illnesses (FAO, 1995).

2.6.4 Regulatory agencies and the government

The regulatory agencies help to establish the principles of HACCP and to verify that the plans are working by regular reviews and inspections. In addition, the regulatory agencies monitor epidemiological studies to identify possible food safety hazards and to conduct risk evaluations to improve and update HACCP plans (FAO, 1995). HACCP-based operating practices for the transportation and warehousing sector requires general educational programs for those companies responsible for transporting and/or storing sensitive foods or foods that are potentially hazardous to health if not handled correctly (FAO, 1995).

Food service and food retail represent a diverse work force and a wide range of foods, as well as having an extremely high turnover of employees, up to 400 % per year in some cases has been reported (FAO, 1995). This represented a major problem with HACCP compliance due to the high employee turnover rate and the varying level of education and language competency. It is therefore essential to ensure that educational programs with reference to proper food handling and personal hygiene are taught efficiently in these food sectors, especially when high-risk foods are involved. For example, the use of thermometer in supermarket refrigerators will minimize NOC formation in cured meat products, whereas constant training is required to maintain employee awareness (FAO, 1995).

68 2.7 Determination of nitrate and nitrite in food and their estimated dietary intake

2.7.1 Introduction

High dietary intake of nitrate and nitrite has been implicated in the aetiology of human gastric cancer. Nitrate is naturally present in green leafy vegetables and nitrite is usually added in meat as a preservative in the form of sodium or potassium nitrite (Mirvish, 1995). However, Campos and Gerschenson (1996) demonstrated that in the absence of sodium nitrite, ascorbic acid exhibited pro-oxidant effect. In contrast, in the presence of sodium nitrite, asocorbic acid exhibited a protective role. The combined use of the preservative sorbates and nitrites in meat can result in the formation of a mutagen ethylnitrolic acid during processing and storage (Binstok et al., 1998).

Nitrate can be reduced to nitrite in the oral cavity and in the stomach. Once in the stomach, nitrite can react with amines to form a group of carcinogens known as NOCs. However, there are dietary factors that can inhibit or reduce the formation of these carcinogens. For example, ascorbic acid and α-tocopherol has been shown to compete with nitrite for the precursors thus reducing the amount of precursors available to react with nitrite (Hwang et al., 1994).

It was estimated that 80 % of human cancers were caused by environmental factors associated with food, water and air (Walters, 1980). In addition, malnutrition, dietary habits and lifestyle may be directly or indirectly related to 40 % of the human cancers (Ologhobo et al., 1996). High dietary intakes of nitrate and nitrite have been implicated in the etiology of human gastric cancer based on epidemiology and clinical studies (Bartsch et al., 1990 and Joossens et al., 1996).

Nitrate is naturally present in leafy vegetables and nitrite is usually added to meat as a preservative in the form of sodium or potassium salt (Cammack et al., 1999). In addition nitrate can be reduced to nitrite in the oral cavity and in the stomach (Duncan et al., 1997). Once in the stomach, nitrite can react with amines and amides, which are organics containing nitrogen such as amino acids, to form a group of carcinogens known as NOCs (Archer, 1989). Stomach is most at risk from endogenous NOC synthesis since stomach acid catalyses nitrosation reactions. High nitrate intake was associated with gastric cancer in England, Colombia, Chile, Japan, Denmark, Hungary and Italy (Forman et al., 1997). Exposure to endogenously formed NOCs had been

69 associated with increased risks of cancer of the stomach, oesophagus and bladder (Bartsch et al., 1990).

Australia’s food composition data were mostly based on overseas data especially those from the United Kingdom and the United States until recently. However, in the revised Australian composition tables based on food analysis performed in Australia, the edible portion of fruit increased by 4 % whereas in meat it decreased by 16 % (Cashel and Greenfield, 1995). Thus dietary contribution of nitrate and nitrite may be over- estimated, whereas dietary intake of antioxidants such as vitamin C and vitamin E may have been underestimated.

The dietary intake of nitrates and nitrites in foods can vary greatly from region to region depending on factors such as farming practices, climate, soil quality, manufacturing processes and legislation. Nitrate and nitrite contents of foods are not available in Australia; hence values from overseas are commonly used. Due to the growing concern of NOCs, accurate and robust methods are necessary for long-term monitoring of nitrate and nitrite concentrations in foods for susceptible populations.

It is therefore the aim of this study to develop an accurate, simple and cost-effective method for quantifying the nitrate and nitrite contents in commonly consumed vegetables, cured meat and fresh meat produced in Australia.

2.7.2 Materials and Methods

2.7.2.1 Reagents

Analytical grade sodium nitrite and potassium nitrate from Univar (Ajax Finechem) were used as standards and for recovery studies. HPLC grade methanol from Lab-Scan was used and ion-pairing agent tetrabutylammonium phosphate was purchased from Waters.

2.7.2.2 Food samples

All vegetables were purchased at local supermarkets, produce shops or wholesale and kept at refrigeration temperature and analyzed within 24 hr. All cured and fresh meat products were also purchased at supermarkets and kept at refrigeration temperature (4 oC) and analyzed within 48 hr.

70 2.7.2.3 Apparatus

Waters HPLC controller model number 600 with photo array detector model number

996 and autosampler model number 717 plus were used. Phenomenex C18 110A Gemini column (250 mm x 4.6 mm x 5 µm) was used for the separation. Injection volume was 10 µl with flow rate set at 1 mL/min and wavelength set at 214 nm. Mobile phase consisted of methanol: water (75:25) with 0.075M of tetrabutylammonium phosphate (PIC-A).

2.7.2.4 Standards

Potassium nitrate (KNO3) and sodium nitrite (NaNO2) were mixed in MilliQ water in volumetric flasks to give a range between 5.0 and 100.0 mg/L for nitrite ions and 2.5 to 50.0 mg/L for nitrate ions.

2.7.2.5 Sampling and extraction

Weighed 10.0 to 50.0 g of meat samples including salami, hot dogs, ham, bacon, Frankfurt and beef, which were purchased from the local supermarkets (at least two packets each) of two different brands, were blended with 300 mL distilled water for 1 min, then made up to 500 mL in volumetric flasks. The pH was measured and 1 mL was taken out for measuring nitrate and nitrite content before cooking using HPLC. Ten mL of mixture of each sample was transferred into 100 mL volumetric flasks and heated in a water bath at 75, 80, 90 and 100 oC for 5, 10 and 15 min. The mixture was made up to 100 mL with distilled water and was shaken. The mixture was allowed to settle and cool; then measured pH and nitrite and nitrate contents. The pH was adjusted with 0.1 M NaOH to neutral pH. Then the mixture was centrifuged at 10,000 rpm for 10 min; then supernatant was removed for ultra-filtration. The filtrate was used for further analysis including quality control such as recovery studies.

Fresh vegetables including English spinach, buk choy, choy sum, Chinese cabbage, gai choy and iceberg lettuce were purchased from the local supermarkets and produce stores. Three bunches each from two different locations with at least three replicates were used for the analysis including recoveries. To examine the effects of sample preparation and extraction conditions on nitrate and nitrite determination samples were chopped in thirds or blended or both and weighed between 25 and 100 g in 500 mL

71 beakers. Spiking with standards was done before cooking in water bath between 60 and 100 oC for 5 to 30 min.

2.7.2.6 Statistical test

Multivariant ANOVA by SPSS Data Editor was used to determine any significant differences in recoveries of nitrate and nitrite based on different sampling and cooking parameters.

2.7.3 Results and discussion

Table 2.2 Mean recoveries for nitrate and nitrite in selected vegetables and meat. ______Food Nitrite recovery (%) Nitrate recovery (%) ______Vegetables 97.0 ± 10.1 96.5 ± 11.0 Meat 92.1 ± 10.2 92.2 ± 11.5 ______

All values are means of replicate analyses. Mean recoveries from all vegetables and meat tested. Meat includes cured and fresh meat.

Table 2.3. Mean nitrate and nitrite contents and their recoveries in fresh vegetables after five minutes boiling.

Vegetables Nitrite Nitrate (mg/kg) (mg/kg) English spinach 0 4849.6 ± 573.6 Recovery (%) 89 74 Buk choy 0 1841.1 ± 84.0 Recovery (%) 97 97 Choy sum 0 1376.9 ± 56.0 Recovery (%) 111 102 Chinese cabbage 0 236.2 ± 27.4 Recovery (%) 91 97 Gai choy 19.6 ± 10.8 1642.3 ± 126.0 Recovery (%) 102 100 Iceberg lettuce 0 48.0 ± 30.2 Recovery (%) 92 110 Detection by UV at 214 nm using C18 Phenomenex column with 0.075M PIC-A and methanol and water (25:75) as the mobile phase. Values were means of at least four replicates from two sources and up to fifteen determinations.

Table 2.4 Mean raw nitrate and nitrite contents and recoveries in cured and fresh meat from Sydney supermarkets after pH adjustment.

Meat Nitrite Nitrate (mg/kg) (mg/kg) Hot dog 78.6 ± 16.4 69.9 ± 11.3 Recovery (%) 109 103 Ham 34.2 ± 5.5 19.0 ± 8.1 Recovery (%) 97 87 Salami 0 142.5 ± 36.3 Recovery (%) 91 102 Bacon 15.7 ± 14.5 23.3 ± 8.2 Recovery (%) 91 82 Frankfurt 83.9 ± 10.1 54.9 ± 8.7 Recovery (%) 96 94 Minced Beef 0 18.7 ± 6.2 Recovery (%) 80 104 Beef medallion 0 38.5±14.9 recovery (%) 80 75

Detection by UV 214 nm using C18 Phenomenex column with 0.075M PIC-A and methanol: water (25:75) as the mobile phase. Values were means of at least four replicates from two to four brands with up to five determinations.

Nitrate and nitrite can be unstable and different sampling methods and extraction procedures can influence their recoveries (Usher and Telling, 1975). Hence various sampling and extraction procedures were tested to find the optimal extraction conditions for nitrate and nitrite determinations in fresh vegetables, cured meat and fresh meat. The optimal extraction conditions were determined by analyzing nitrate and nitrite concentration at various temperatures, cooking time and sample size since these can affect their detection. Vegetables were cooked at 100 oC for 5 min with sample size of 100 g to correspond common cooking practises. Cured and freshed meat were extracted at 75 oC for 15 min with a sample size of 50 g. Since cured meat is not normally cooked before consumption, the lowest temperature was used and gave a good recovery (Table 2.2).

Sample size was chosen so that a representative sample can be ensured that also takes into account of natural heterogeneity of the samples tested. Mean recoveries were > 92 % for both nitrate and nitrite in two food matrices tested (Table 2.2). Taking into

73 account of factors affecting nitrate and nitrite recovery in foods include 1) temperature since nitrate and nitrite are not stable at high temperatures, 2) cooking conditions, which can affect pH of the sample water and exposure to atmospheric oxygen, 3) pH of the sample water, since nitrite is readily converted to nitric acid or nitric oxide at acidic pH, and 4) sources of food samples can vary greatly and may contain interfering substances such as iron and magnesium (Usher and Telling, 1975). In addition, plants posesses nitric oxide synthase (del Río et al., 2004) that can further reduce its nitrate and nitrite content to nitric oxide. Boiling was used to extract nitrate and nitrite in vegetables in this study (Table 2.3) because it is the commonly used method to cook vegetables. In addition, pH was monitored and maintained close to neutral pH to minimize conversion of nitrite to nitrous acid or nitrous oxide.

Nitrite levels in vegetables may increase during post-harvest storage by the action of indigenous bacteria and/or the presence of nitrate reductase (Hunt, 1994), especially when they are left at room temperature or higher. This may explain the small amount of nitrite (19.6 mg/kg) present in Gai choy during the preparation at room temperature (Figure 2.6). Likewise, it was demonstrated that there was no detectable nitrite in 94 % of edible fresh retail vegetables (Hunt and Turner, 1994). Cultivar and harvest date can affect the nitrate and nitrite levels of selected vegetables (Amr and Hadidi, 2001). This may explain the high variability and hence the high standard deviation observed in Table 2.3.

English spinach had the highest nitrate content (4850 mg/kg) compared to other vegetables (Table 2.3). This finding correlated well with the literature (Öztekin et al., 2002). However, according to Gaiser et al (1996), spinach blanched for 3 min can contain in the range of 50 to 5600 mg/kg nitrates with a mean nitrate concentration of approximately 2000 mg/kg and a large standard deviation of 1411.4 mg/kg. This demonstrated the high variability of nitrate content in spinach and other green leafy vegetables. Excluding spinach, other vegetables tested had nitrate ranging 48 - 2121 mg/kg, which is half that of spinach (Table 2.3). Thus it can be concluded that spinach contributed to the highest dietary nitrate intake from leafy green vegetables. Lettuce contained lowest amount of nitrate in this study (48.0 mg/kg, Table 2.3), which was significantly lower to earlier studies that demonstrated high nitrate content in lettuce at 2500 mg/kg (Marshall and Trenerry, 1996). The dissimilarity may be due to horticultural practices such as the use of nitrate-based fertilizers.

74 Hot water extraction was used in this study to mimic common household cooking method. However, the nitrate content was shown to vary depending on the cooking method. For example, Prasad and Chetty (2008) demonstrated that frying in soya bean oil increased nitrate content of leafy vegetables from 159 to 307 %. In contrast, boiling reduced nitrate content by 47-56 %, most likely due to nitrate ions leaching into the cooking water, whereas baking had no significant effect on the nitrate content. In addition, Pham et al. (2008) shown that frozen spinach had half the nitrate content of fresh spinach. This may be caused by the leaching of nitrate ions upon defrosting due to the damage to plant cell integrity during freezing.

Based on plasma nitrate and nitrite from a small human trial, van Velzen et al. (2008) demonstrated that the absolute bioavailability of nitrate from vegetables was almost 100 % regardless of whether it was cooked or raw. This suggests that vegetable matrix components may affect extraction efficiency, but once it is ingested it releases all the nitrate content for circulation.

Different countries have set their maximum limits for the addition of nitrate and/or nitrite salts in cured meat. Under the Australian Food Standard Code 1.3.1 schedule 1, 125 mg/kg of nitrite in a form of potassium or sodium salt is permitted in cured, dried, and slow dried cured meat; whereas in commercially sterile and canned cured meat, the maximum nitrite (potassium or sodium salts) permitted is 50 mg/kg. For slow dried cured meat, the maximum allowed nitrate (potassium or sodium salts) is 500 mg/kg (FSANZ, 2009). Given the established antimicrobial effect of nitrite salts, particularly in reference to Clostridium botulinum in cured meat, its level should remain sufficient to prevent the occurrence of food borne illnesses, but also kept to the minimum to minimize dietary nitrite intake in light of its potential adverse health effects based on epidemiological and clinical studies. Due to growing popularity of “natural” and “organic” foods in the United States, alternative processes that use high nitrate vegetable-based ingredients and nitrate-reducing starter culture have produced processed meat products with very similar cured meat properties (Sebranek and Bacus, 2007). However, the nitrate and nitrite content and residual content were hard to standardize in these natural products because of the “live” reactions that occurs. Therefore food safety especially botulism is a key issue as well as product consistency and quality may also be compromised.

75 Seven types of meat tested in this study had at least four replicates each from at least two brands. Nitrate and nitrite contents in various cured meat products and fresh meat were below the maximum allowable limit set by Food Standards Australia and New Zealand (FSANZ, 2009)(Table 2.4). Continuous monitoring of nitrite used in cured meat products is important to ensure that the dietary intake of nitrite is kept to below the limit set by FSANZ. Nitrate and nitrite concentration in these cured meat products were tested before cooking (hot water extraction), after cooking and after pH adjustment. Nitrate and nitrite concentration increased after extraction since these anions were released into the water, however, the final concentration was presented after pH adjustment once they were stabilized.

Interferences naturally present or added additives in cured meat products may account for differences in nitrate and nitrite recovery. For example, Butt et al. (2001) demonstrated that the presence of 50-fold sulphate and chloride did not affect the resolution and percent recovery of nitrite, but did reduce the resolution and recovery of nitrate. In addition, the presence of magnesium, iron and calcium significantly reduced the percentage recovery of both anions, which should be removed to ensure accurate determination of nitrate and nitrite. Furthermore, they also demonstrated that under optimized HPLC conditions, both nitrate and nitrite peaks began to merge when the concentration of nitrite was above six-fold of nitrate concentration, hence nitrite used in calibration curve and for recovery were half the concentration of nitrate to minimize the merging of nitrite and nitrate peaks.

Using similar detection method as Reinik et al. (2005), they found the mean sodium nitrite and nitrate concentrations in ham were 20.8 and 68.0 mg/kg, respectively. However in this study, the nitrite concentration in ham averaged at 34.2 ± 5.6 mg/kg and nitrate concentration was lower at 19.0 mg/kg (Table 2.4). Some manufacturers add less nitrite but more nitrate as a nitrite reserve. This may also explain the differences in the findings by Öztekin et al. (2002), where the nitrite and nitrate contents in ham were 4.0 and 35.6 mg/kg, respectively.

Dionex Corporation (1998) found the nitrite and nitrate contents in ham to be 11.6 and 5.4 mg/kg, respectively, whereas salami contained 108.0 mg/kg nitrite and 98.5 mg/kg nitrate. Using capillary electrophoresis, the nitrite and nitrate content in salami detected were 24.3 and 43.6 mg/kg, respectively (Öztekin et al., 2002). Compared to their

76 findings, the current study showed that the salami contained much less nitrite at 3.7 mg/kg and much more nitrate at 139.5 mg/kg (Table 2.4). Although the extraction methods were similar the temperature used in our study was higher, apart from the differences that may be attributed to the manufacturing practices. Stalikas et al. (2003) used similar extraction temperature and reported that nitrate and nitrite contents in salami were 54.0 and 84.0 mg/kg, respectively. Thus differences are more likely to be due to the manufacturing processes.

It was reported by Dennis et al. (1990) that the mean nitrite content in bacon was 24.0 mg/kg and for nitrate was 43.0 mg/kg, whereas nitrite and nitrate in ham were 56.0 and 22.0 mg/kg, respectively. They used similar extraction and detection methods but with an anion exchange column. Both bacon and ham products in this study contained less nitrate and nitrite (Table 2.4) in comparison. Siu and Henshall (1998) who found that nitrite and nitrate contents in salami were 108.0 and 98.5 mg/kg, respectively, and 11.6 and 5.4 mg/kg for ham, respectively. Sample extraction procedures used in the current study were similar to Marshall and Trenerry (1996), but they omitted the heating step. This may explain the low nitrite content of less than 10 mg/kg in salami, leg ham and bacon. However the nitrate contents were higher at 141.5, 132.5 and 48.0 mg/kg, respectively. Different cured meat products may require different ratio of nitrite and nitrate as preservatives. Since fresh meat does not naturally contain nitrite (Table 2.4), its nitrite and nitrate contents have not been extensively tested. However, based on this study, the nitrate content in minced beef and medallion beef were within the range found in cured meat products (Table 2.4).

It was demonstrated that recovery increases as the meat solids decreases (Usher and Telling, 1975). Hence using smaller meat samples should reduce the effects of interfering substances, which was demonstrated in this study with acceptable recoveries (Table 2.2). Furthermore, most interference can be eliminated by UV detection. However, UV may detect chloride ions as positive or negative peaks in the wavelength used for nitrate and nitrite and are eluted before nitrite (Di Matteo and Esposito, 1997). Chloride peaks were not present at 214 nm in this study, which suggests that chloride ions did not interfere with nitrite quantification since nitrite recovery was above 92 % for both meat and vegetable samples (Table 2.2).

77 Due to its reactive nature, nitrite analysis from food does not give a true representation of the total nitrite added. Furthermore, nitrite added to meat is usually present as nitric oxide bound with other food components such as myoglobin (5-15 %), sulphydryl groups (5-15 %), lipids (1-5 %), proteins (20-30 %), as nitrate (< 10 %), and as free nitrite (10-15 %) (Zanardi et al., 2002). Therefore recovery range may be quite large as a result of nitrite’s reactive nature and its attachment to other food components. Other methods of food extraction estimate the total nitrite present by releasing food-bound nitrite. This may over estimate the significance of dietary nitrite and the etiology of gastric cancer because only free nitrite can participate in nitrosation. Hot water extraction to quantify free nitrite available to participate in nitrosation was used in this study.

Regarding relevance to incidence of gastric cancer, based on age-standardized statistics, diagnosed gastric cancer rate per 100 000 in males and females worldwide is 22 and 10.3 %, respectively, with mortality rate of 14.3 and 8.3 %, respectively. Gastric cancer is the third leading cause of death in men after lung and prostate cancer, and is the fourth leading cause of death in women worldwide (Forman and Burley, 2006). Overall gastric cancer rate is declining, especial in more developed countries, with the exception of Miyagi prefecture of Japan still having the highest gastric cancer rate. Korea, East Asia, South America and Eastern Europe also sustained a high gastric cancer rate. However, Bombay in India always maintained a low gastric cancer rate between 1953 and 1997 (Forman and Burley, 2006). This maybe due to higher consumption of antioxidant rich fruits and vegetables and herbs and spices, which have been shown to reduce the risk of gastric cancer. Joossens et al. (1996) studied dietary salt, nitrate and gastric cancer mortality in 24 countries and demonstrated that nitrate intake became an increased risk factor for gastric cancer when salt intake was also high.

It was predicted that increasing population numbers and longevity could cause a net increase in gastric cancer rate worldwide. Since diagnosis often occurs between the ages of 60 and 80, with up to 30 % mortality rate after five years diagnosis (Forman and Burley, 2006), it is vital to make dietary and lifestyle changes to decrease gastric cancer rate and to increase survival rate with better diagnostic facility and education. Risk factors to be avoided include H. pylori infection, smoking, high consumption of cured meat and salt, and low consumption of fruits and vegetables.

78 Once the nitrate and nitrite content in food is established, one can estimate the intake of these anions based on national dietary surveys. However, the availability for nitrate certified reference materials is still poor, but is necessary for comparing values (Castanheira et al., 2004). Gangolli et al. (1994) estimated that the mean daily intake of nitrate and nitrite in the US were 106 and 1.5 mg/kg, respectively, and in the UK were 104 and 1.5 mg/kg, respectively. In 1994 van Vliet et al. (1997) estimated the mean intake of nitrate in the Dutch population to be 80 mg/day per person, and the median intake of nitrite to be 0.1 mg/day per person. In comparison, Italy had a mean daily intake of nitrate of 245 mg/day, whereas Poland and Switzerland recorded mean daily nitrate intakes of 178 and 125 mg/day, respectively. France’s mean daily intake of nitrate and nitrite were 150.7 and > 3 mg/day, respectively, followed by Netherlands, Germany and Norway where mean daily nitrate intakes were 71, 68 and 43 mg/day, respectively. The mean daily nitrite consumption in those countries was 0.6, 2.6 and 1.8 mg/day, respectively (Gangolli et al., 1994). According to Cornee et al. (1992), the average daily nitrate intake per person per day was 121 mg (85 % from vegetables, 5 % from preserved and cured meat, and 5% from cereal products). For the average daily nitrite intake per person per day, it was found to be 1.88 mg (43 % from vegetables, 28 % from cured meat, and 16 % from cereals). The remaining 13 % of nitrite must come from non-dietary sources of nitrite such as atmospheric contamination.

In summary, Pennington (1998) estimated that the daily nitrate intake ranges between 53 and 350 mg/day depending on the type and quantity of the vegetable consumed and the level of nitrate in drinking water. Whereas daily nitrite consumption was between 0 and 20 mg/day depending on the levels of nitrite present in cured meat and much of it was consumed. In 1995, the ADI for nitrate was set at 3.7 mg/kg body weight by the European Union Scientific Committee for Food. Since nitrite has higher acute toxicity than nitrate its ADI was set at 0.06 mg/kg body weight (Reinik et al., 2005).

Based on the Australian Bureau of Statistics (ABS, 1998-1999), Australians consumed 8.7 kg of bacon and ham combined per capita per year in 1998-1999. Assuming half of each product was consumed at 4.35 kg per capita per year (or 12 g per capita per day) based on the finding in this study, nitrite from bacon per capita per day was 0.19 mg, and for nitrate was 0.41 mg. For ham (12 g per capita per day) nitrite consumed per capita per day was 0.28 mg and for nitrate was 0.23 mg. Thus combined nitrite and nitrate intake from bacon and ham per capita per day were 0.47 and 0.64 mg,

79 respectively. At the upper extreme, assuming 100 g of bacon or ham was consumed every day, the nitrite and nitrate intake from bacon would be 1.57 mg and 3.42 mg per capita per day, respectively. Similarly for ham the nitrite and nitrate intake would be 2.33 and 1.90 mg nitrite and nitrate per capita per day, respectively, giving a total of 3.9 mg of nitrite and 5.32 mg nitrate per capita per day. This is significantly lower than the ADI set by the European Union Scientific Committee for Food in 1995.

However, taking endogenous formation of nitrate into account, this means additional 70 mg of nitrate for an average 70 kg adult (Gangolli et al. 1994). Furthermore, it was estimated approximately 25 % of dietary nitrate is converted to nitrite by bacteria and nitrate reductase in the oral cavity (Gangolli et al. 1994). Thus assuming two servings (150 g) of vegetables comes from green leafy vegetables, again taking the analytical data from this study, this means approximately 727.5 mg of nitrate from English spinach (Figure 2.7) is ingested, of which 181.9 mg of nitrite can participate in nitrosation in the stomach. Based on the above assumptions, the total nitrite and nitrate burden for an Australian adult of 70 kg body weight, the intakes per day is approximately 185.8 and 620.9 mg, respectively, sourced from cured meat, spinach and endogenous nitrate formation. This exceeds the ADI of 4.2 mg for nitrite by 44 times per 70 kg adult per day, and by 2.4 times of ADI of 259 mg nitrate per 70 kg adult per day. However, it must be noted that the above prediction assumed that nitrite only came from bacon and ham, and that nitrate intake only came from two servings of English spinach. Since English spinach had the highest nitrate content, this predicts the upper extreme of dietary nitrate intake. Furthermore, the ADI do not include the 25 % conversion of dietary nitrate to nitrite in the oral cavity, which underestimate the total ingested dietary nitrite. Du et al. (2007) stated that at least 15 % of the adult Dutch population exceeded the acceptable daily nitrate intake, with children this figure went up to 45 % due to dietary practices. Although it is possible to exceed the ADI of nitrate based on the recommended daily intake of 400 g fruit and vegetables, this would not pose a significant health risk in the context of a balanced diet (Heppner et al., 2008). However, the source of dietary nitrate and nitrite is an important risk factor since fruit and vegetable also offers other protections in the form of antioxidants.

80 (Davies and Sano, 2001). Thus this epidemiological study suggests that diet and lifestyle may play an important role in gastric cancer aetiology besides effective screening and management. Countries such as South Korea, Japan and China had the highest stomach cancer mortality for men, whereas countries with the highest stomach cancer mortality for women were South Korea, China and Columbia. Canada and Denmark had the lowest stomach cancer mortality for men and women, respectively (Joossens et al., 1996).

The high gastric cancer incidence in the Far East may be due to the consumption of specific foods that are high in nitrates such as Korean Kimchi or high in salt as in many traditional Japanese dishes, or particular food preparation methods such as broiling of meats (Duncan et al., 1997). Regions of high risk to gastric cancer often coincide with a low intake of foods containing vitamin C. Other risk factors for human gastric cancer include residence in areas with high nitrate-containing soil due to many factors such as the addition of fertilizers and foods pickled with salt (Weisburger, 1981). Certain salted fermented fish products including fish sauce were associated with the high gastric cancer mortality in Fujian province of China (Chen et al., 1992). Similarly, a potential link for the high gastric cancer rate in Southwest Korea was associated with regular high consumption of salted pickled cabbage and salted seafood sauce (Seel et al., 1994). The former also contained high levels of total NOC precursors, and cabbages, which are known to contain high levels of nitrate than any other vegetables. However, it was suggested by Lundberg et al. (2006) that green leafy vegetables might have cardioprotective benefits because of the high nitrate content and its role in vasodilation once it is reduced to nitric oxide in vivo. This can be tested in clinical trials by inorganic nitrate supplementation, but green leafy vegetables may contain other cardioprotective components that work synergistically to prevent a range of human diseases.

There are no nitrate standards for vegetables to date, except in 2000 Chinese government had established a maximum level of nitrate in vegetables of 3100 mg/kg (Du et al., 2007). It was suggested that since there may be health risks associated with dietary nitrate intake, it was very reasonable to limite the nitrate content in vegetables and drinking water by legislations (Du et al., 2007). This is perhaps the first step in achieving a “safety” standard for nitrates in vegetables. However, considering the cost of the produce, production variability and the cost of analysis, whether this can be

81 sustained and monitored requires further considerations. Another key step is the development and compilation of vegetable composition databases based on available data (Hoefkens et al., 2009). This can then be used for estimating nutrient and contaminant including nitrate intake in susceptible populations, thus providing a useful tool in the epidemiology studies of nitrate intake and gastrointestinal cancers.

In addition to forming carcinogens in the stomach, nitrite is also genotoxic and can readily induce methaemoglobinaemia especially in babies (Gangolli et al., 1994). Furthermore, the lethal dose for nitrite in adults was estimated to be between 2 and 9 g

NaNO2 per day, or 33-250 mg/kg body weight (Gangolli et al., 1994), whereas the lethal dose for nitrate ions was estimated at 20 g per day, or 330 mg nitrate ions/kg body weight (Gangolli et al., 1994). Although it is unlikely to reach these toxic levels from dietary intake alone, the long-term effects may be detrimental based on epidemiological and clinical studies.

The European Union have regulated the maximum nitrate levels for vegetables ranging from 2000 mg/kg frozen spinach to 4500 mg/kg fresh lettuce (Bottex et al., 2008). Given the potential adverse health effect of dietary nitrate and nitrite intake, it would be beneficial for Australia to set up similar regulatory and monitoring programs to ensure public health in the long term.

2.7.4 Conclusion

Different authors attributed different percentage of dietary nitrate and nitrite to the major food groups, but the consensus is that vegetables contributed to the majority of dietary nitrate and that cured meat products contributed to the majority of dietary nitrite. This study agreed with the literature that vegetables are the major dietary nitrate contributor and that meat and especially cured meat provided the majority of dietary nitrite. The extraction and detection method used in this study was demonstrated to be simple, fast, sensitive and applicable to both meat and vegetable samples.

Nitrate and nitrite content of foodstuff should be monitored in the long term to estimate the dietary intake and to provide insights into the effects of new horticultural and manufacturing technologies on the levels of nitrates and nitrite in foods and in the aetiology of gastro-intestinal cancers. It was demonstrated that nitrate and nitrite content tested were below the guideline set by FSANZ and their intake were within the

82 range reported in the Western countries and were below the acceptable daily intake. This study will be the first to quantify the levels of nitrate and nitrite in foods in the Sydney supermarket and will provide useful data to industry and health professionals.

83 2.8 References

Abuharfeil, N., Sarsour, E., and Hassuneh, M. 2001. The effect of sodium nitrite on some parameters of the immune system. Food and Chemical Toxicology, 39:119-124.

Amr, A., and Hadidi, N. 2001. Effect of cultivar and harvest date on nitrate (NO3) and nitrite (NO2) content of selected vegetables grown under open field and greenhouse conditions in Jordan. Journal of Food Composition and Analysis, 14:59-67. Andrade, R., Viana, C.O., Guadagnin, S.G., Reyes, F.G.R., and Rath, S. 2003. A flow- injection spectrophotometric method for nitrate and nitrite determination through nitric oxide generation. Food Chemistry, 80:597-602. Archer, M.C. 1989. Mechanisms of action of N-nitroso compounds. Cancer Surveys, 8:241-250. Australia Bureau of Statistics. 1998-1999. 16.59 Apparent per capita consumption of foodstuffs. Available on: www.abs.com.au [25 Feb 2008] Badea, M., Amine, A., Palleschi, G., Moscone, D., Volpe, G., Curulli, A. 2001. New electrochemical sensors for detection of nitrites and nitrates. Journal of Electroanalytical Chemistry, 509:66-72.

Barbier, G.G., Joshi, R.C., Campbell, E.R., and Campbell, W.H. 2004. Purification and biochemical characterization of simplified eukaryotic nitrate reductase express in Pichia pastoris. Protein Expression and Purification, 37:61-71. Bartsch, H., Ohshima, H., Shuker, D., Pignatelli, B., and Calmels, S. 1990. Exposure of humans to endogenous N-nitroso compounds: implications in cancer etiology. Mutation Research, 238:255-267. Binstok, G., Campos, C., Varela, O., and Gerschenson, L.N. 1998. Sorbate-nitrite reactions in meat products. Food Research International, 31:581-585. Bloukas, J.G., Arvanitoyannis, I.S., and Siopi, A.A. 1999. Effect of natural colourants and nitrites on colour attributes of frankfurters. Meat Science, 52:257-265. Bories, P.N., Scherman, E., and Dziedzic, L. 1999. Analysis of nitrite and nitrate in biological fluids by capillary electrophoresis. Clinical Biochemistry, 32:9-14. Bottex, B., Dorne, J.J., Carlander, D., Benford, D., Przyrembel, H., Heppner, C., Kleiner, J., and Cockburn, A. 2008. Risk-benefit health assessment of food-Food fortification and nitrate in vegetables. Trends in Food Science and Technology, 19:S113-119. van den Brandt, P.A., and Goldbohm, R.A. 2006. Nutrition in the prevention of gastrointestinal cancer. Best Practice and Research Clinical Gastroenterology, 20: 589-603. Briggs, D., and Lennard, L. 2002. ‘Risks, additives, contaminates and natural toxicants’, In Food and Nutrition: Australasia, Asia and the Pacific, 2nd ed. ed. M. Wahlqvist, Allen and Unwin Pty Ltd, Sydney.

84 Bryan, N.S. 2006. Nitrite in nitric oxide biology: Cause or consequence? A systems- based review. Free Radical Biology and Medicine, 41:691-701. Bryan, N.S., and Grisham, M.B. 2007. Methods to detect nitric oxide and its metabolites in biological samples. Free Radical Biology and Medicine, 43:645-657. Butt, S.B., Riaz, M., and Iqbal, M.Z. 2001. Simultaneous determination of nitrite and nitrate by normal phase ion-pair liquid chromatography. Talanta, 55:789-797. Cammack, R., Joannou, C.L., Cui, X-Y., Martinez, C.T., Maraj, S.R., and Hughes, M.N. 1999. Nitrite and nitrosyl compounds in food preservation. Biochimica et Biophysica Acta, 1411:475-488. Campos, C.A., and Gerschenson, L.N. 1996. Effect of certain additives on sorbate stability. Food Research International, 29:147-154. Cashel, K.M., and Greenfield, H. 1995. The effect of revised Australian food composition tables on estimates of foods and nutrients available for national consumption, 1983-1984. Journal of Food Composition and Analysis, 8:45-61. Cashel, K.M., and Greenfield, H. 1996. The impact of new Australian food composition data on national dietary references. Food Chemistry, 57:63-66.

Cashel, K.M., and Greenfield, H. 1997. Population Nutrition Goals and Targets for Australia: Influences of New Australian Food Composition Data. Journal of Food Composition and Analysis, 10:176-189.

Cassens, R.G. 1995. Use of nitrite in cured meats today. Food Technology, 49:72-80. Castanheira, I., Oliveira, L., Valente, A., Alvito, P., Costa, H.S. and Alink, A. 2004. The need for reference materials when monitoring nitrate intake. Analytical and Bioanalytical Chemistry, 378:12332-1238. Chasco, J., Lizaso, G., and Beriain, M.J. 1996. Cured colour development during sausage processing. Meat Science, 44:203-211. Cassens, R.G. 1995. Use of nitrite in cured meats today. Food Technology, 49:72-80. Chen, C.S., Pignatelli, B., Malaveille, C., Bouvier, G., Shuker, D., Hautefeuille, A., Zhang, R.F., Bartsch, H. 1992. Levels of direct-acting mutagens, total N-nitroso compounds in nitrosated fermented fish products, consumed in a high-risk area for gastric cancer in southern China. Mutation Research, 265:211-221. Chen, B-M., Wang, Z-H., Li, S-X., Wang, G-X., Song, H-X., and Wang, X-N. 2004. Effects of nitrate supply on plant growth, nitrate accumulation, metabolic nitrate concentration and nitrate reductase activity in three leafy vegetables. Plant Science, 167:635-643. Chow, C.K., and Hong, C.B. 2002. Dietary vitamin E and selenium and toxicity of nitrite and nitrate. Toxicology, 180:195-207. Clark, D.B., Barnes, K.A., Castle, L., Rose, M., Wilson, L.A., Baxter, M.J., Price, K.R., and DuPont, M.S. 2003. Levels of phytoestrogen, inorganic trace-elements, natural toxicants and nitrate in vegetarian duplicate diets. Food Chemistry, 81:287-300.

85 Coleman, J.W. 2001. Nitric oxide in immunity and inflammation. International Immunopharmacology, 1:1397-1406. Connolly, D., Barron, L., and Paull, B. 2002. Determination of urinary thiocyanate and nitrate using fast ion-interaction chromatography. Journal of Chromatography B, 767:175-180. Cornée, J., Lairon, D., Velema, J., Guyader, M., and Berthezene, P. 1992. An estimate of nitrate, nitrite, and N-nitrosodimethylamine concentrations in French food products or food groups. Sciences Des Aliments, 12:155-197. Craddock, V.M. 1990. Nitrosamines, food and cancer: Assessment in Lyon. Food and Chemical Toxicology, 28:63-65. Davies, P.A., and Sano, T. 2001. The difference in gastric cancer between Japan, USA and Europe: what are the facts? what are the suggestions? Critical Reviews in Oncology/Hematology, 40:77-94. Dejam, A., Hunter, C.J., Schechter, A.N., and Gladwin, M.T. 2004. Emerging role of nitrite in human biology. Blood Cells, Molecules, and Diseases, 32:423-429. Dennis, M.J., Key, P.E., Papworth, T., Pointer, M., and Massey, R.C. 1990. The determination of nitrate and nitrite in cured meat by HPLC/UV. Food Additives and Contaminants, 7:455-461. Dionex Corporation. 1998. Determination of nitrate and nitrite in meat using high- performance anion-exchange chromatography. Application Note 112. Doi, C., Noguchi, Y., Marat, D., Saito, A., Fukuzawa, K., Yoshikawa, T., Tsuburaya, A., and Ito, T. 1999. Expression of nitric oxide synthase in gastric cancer. Cancer Letters, 144:161-167. Duncan, C., Dougall, H., Johnston, P., Green, S., Brogan, R., Leifert, C., Smith, L., Golden, M., and Benjamin, N. 1995. Chemical generation of nitric oxide in the mouth from the enterosalivary circulation of dietary nitrate. Nature Medicine, 1:546-551. Duncan, C., Li, H., Dykhuizen, R., Frazer, R., Johnston, P., MacKnight, G., Smith, L., Lamza, K., McKenzie, H., Batt, L., Kelly, D., Golden, M., Benjamin, N., and Leifert, C. 1997. Protection against oral and gastrointestinal diseases: importance of dietary nitrate intake, oral nitrate reduction and enterosalivary nitrate circulation. Comparative Biochemistry and Physiology, 118A:939-948. Du, S-T., Zhang, Y-S., and Lin, X-Y. 2007. Accumulation of nitrate in vegetables and its possible implications to human health. Agricultural Sciences in China, 6:1246-1255. Dutt, M.C., Lim, H.Y., and Chew, R.K. 1987. Nitrate consumption and incidence of gastric cancer in Singapore. Food and Chemical Toxicology, 25:515-520. Dykhuizen, R.S., Frazer, R., Duncan, C., Smith, C., Golden, M., Benjamin, N., Leifert, C. 1996. Antimicrobial effect of acidified nitrite on gut pathogens: importance of dietary nitrate in host defense. Antimicrobial Agents and Chemotherapy, 40:1422-1425. Dykhuizen, R.S., Fraser, A., McKenzie, H., Golden, M., Leifert, C., and Benjamin, N. 1998. Helicobacter pylori is killed by nitrite under acidic conditions. Gut, 42:334-337.

86 Ellis, G., Adatia, I., Yazdanpanah, M., and Makela, S. K. 1998. Nitrite and nitrate analyses: a clinical biochemistry perspective. Clinical Biochemistry, 31:195-220. Fan, A.M., and Steinberg, V.E. 1996. Health implications of nitrate and nitrite in drinking water: An update on methemoglobinemia occurrence and reproductive and developmental toxicity. Regulatory Toxicology and Pharmacology, 23:35-43. Food and Agriculture Organization. 1991. Guidelines for slaughtering, meat cutting and further processing. FAO Animal Production and Health Paper No. 91. FAO of the United Nations, Rome. Food and Agriculture Organization. 1995. The use of hazard analysis critical control point (HACCP) principles in food control. FAO Food and Nutrition Paper No. 58, FAO of the United Nations, Rome. Food Standards Australia and New Zealand. (2009). Food Standards Code. Available online: http://www.foodstandards.gov.au Forde, B.G. 2000. Nitrate transportersin plants: structure, function and regulation. Biochemica et Biophysica Acta, 1465:219-235. Forman, D., Al-Dabbagh, S., and Doll, R. 1985. Nitrates, nitrites and gastric cancer in Great Britain. Nature, 313:620-625. Forman, D., and Shuker, D. 1997. Helicobacter pylori and gastric cancer - A case study in molecular epidemiology. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, 379: S159. Forman, D., and Burley, V.J. 2006. Gastric cancer: global pattern of the disease and an overview of the environmental risk factors. Best Practice and Research Clinical Gastroenterology, 20:633-649. Fujita, K-I., and Kamataki, T. 2001. Role of human cytochrome P450 (CYP) in the metabolic activation of N-alkylnitrosamines: application of genetically engineered Salmonella typhimurium YG7108 expressing each form of CYP together with human NADPH-cytochrome P450 reductase. Mutation Research, 483:35-41. Gaiser, M., Rathjen, A., and Spiess, W. E. L. 1996. Nitrate extraction during blanching of spinach. Lebensmittel-Wissenschaft und Technologie (Food Science and Technology), 30:432-435. Gangolli, S.D., van den Brandt, P., Feron, V.J., Janzowsky, C., Koeman, J.H., Speijers, G.J., Spiegelhalder, B., Walker, R., and Wishnok, J.S. 1994. Nitrate, nitrite and N- nitroso compounds, European Journal of Pharmacology, Environmental Toxicology and Pharmacology Section, 292:1-38. Gago, B., Lundberg, J.O., Barbosa, R.M., Laranjinha, J. 2007. Red wine-dependent reduction of nitrite to nitric oxide in the stomach. Free Radical Biology and Medicine, 43:1233-1242.

Gilliam, M.B., Sherman M.P., Griscavage J.M., Ignarro L.J. 1993. A Spectrophotometric Assay for Nitrate Using NADPH Oxidation by Aspergillus Nitrate Reductase. Analytical Biochemistry, 212:359-365.

Gladwin, M.T., Crawford, J.H., and Patel, R. 2004. The biochemistry of nitric oxide, nitrite, and hemoglobin: role in blood flow regulation. Free Radical Biology and Medicine, 36:707-717. Gonzalez M.J., Valle A.I., Marina M.L. 1994. Determination of nitrate and nitrite anions in natural waters by reversed-phase ion pair liquid chromatography. Microchemical Journal, 50: 1-5.

Granger, D.L., Taintor, R.R., Boockvar, K.S., and Hibbs, Jr., J.B. 1995. Determination of nitrate and nitrite in biological samples using bacterial nitrate reductase coupled with the Griess reaction. Methods: A Companion to Methods in Enzymology, 7:78-83. Grau, M., Hendgen-Cotta, U.B.., Brouzos, P., Drexhage, C., Rassaf, T., Lauer, T., Dejam, A., Kelm, M., and Kleinbongard, P. 2007. Recent methodological advances in the analysis of nitrite in the human circulation: Nitrite as a biochemical parameter of the l-arginine/NO pathway. Journal of Chromatography B, 851:106-123.

Grisham, M.B., Johnson, G.G., Gautreaux, M.D. and Berg, R.D. 1995. Measurement of nitrate and nitrite in extracellular fluids: A window to systemic nitric oxide metabolism. Methods: A Comparison to Methods in Enzymology, 7:84-90.

Hanai, T. 1999. HPLC A Practical Guide. RSC Chromatography Monographs. Royal Society of Chemistry, United Kingdom.

Hansson, L-E., Nyrén, O., Bergström, R., Wolk, A., Lindgren, A., Baron, J., and Adami, H-O. 1994a. Nutrients and gastric cancer risk. A population-based case-control study in Sweden. International Journal of Cancer, 57:638-644. Hansson, L-E., Baron, J., Nyrén, O., Bergström, R., Wolk, A., and Adami, H-O. 1994b. Tobacco, alcohol and the risk of gastric cancer. A population-based case-control study in Sweden. International Journal of Cancer, 57:26-31. Helmke, S.M., and Ducan, M.W. 2007. Measurement of the NO metabolites, nitrite and nitrate, in human biological fluids by GC-MS. Journal of Chromatograpy B, 851:83- 92. Hendrix, S.A., and Braman, R.S. 1995. Determination of nitrite and nitrate by vanadium (III) reduction with chemiluminescence detection. Methods: A Companion to Methods in Enzymology, 7:91-97. Heppner, C., Dorne, J.L., Fabiansson, S., Verger, P., Fuerst, P., Fernandez-Cruz, M.L., Brandy, P.V., Darnerud, P.O., Speijers, G., Addiscot, T., and Cockburn, A. 2008. The first risk benefit assessment of nitrate in vegetables: A European perspective. Toxicology Letter, 180(S1):S65. Hirose, M., Takahashi, S., Ogawa, K., Futakuchi, M., and Shirai, T. 1999. Phenolics: Blocking agents for heterocyclic amine-induced carcinogenesis. Food and Chemical Toxicology, 37:985-992. Hoefkens, C., Sioen, I., De Henauw, S., Vandekinderen, I., Baert, K., De Meulenaer, B., Devlieghere, F., and Van Camp, J. 2009. Development of vegetable composition

88 databases based on available data for probabilistic nutrient and contaminant intake assessments. Food Chemistry, 113:799-803.

Huarte-Mendicoa, J.C., Astiasarán, I., and Bello, J. 1997. Nitrate and nitrite levels in fresh and frozen broccoli: effect of freezing and cooking. Food Chemistry, 58:39-42. Hubbar, R.W., Mejia, A., and Horning, M. 1994. The potential of diet to alter disease processes. Nutrition Research, 14:1853-1896.

Hunt, J. 1994. A method for measuring nitrite in fresh vegetables. Food Additives and Contaminants, 11:317-325. Hunt, J., and Turner, M.K. 1994. A survey of nitrite concentrations in retail fresh vegetables. Food Additives and Contaminants, 11:327-332. Hwang, H., Dwyer, J., and Russell, R.M. 1994. Diet, Helicobacter pylori infection, food preservation and gastric cancer risk: are there new roles for preventative factors? Nutrition Reviews, 52:75-83. International Agency for Research on Cancer (IARC). 1987. Monographs on the evaluation of carcinogenic risks to humans. Volumes 17, World Health Organization. Jackson, P.E., Haddad, P.R., and Dilli, S. 1984. Determination of nitrate and nitrite in cured meats using high-performance liquid chromatography. Journal of Chromatography A, 295:471-478. Jakszyn, P., and Gonzalez, C.A. 2006. Nitrosamine and related food intake and gastric and oesophageal cancer risk: a systematic review of the epidemiological evidence. Carcinogenesis, 27:1497-501. Jaworska, G. 2005. Nitrates, nitrites, and oxalates in products of spinach and New Zealand spinach: Effect of technological measures and storage time on the level of nitrates, nitrites, and oxalates in frozen and canned products of spinach and New Zealand spinach. Food Chemistry, 93:395-401. Jimidar, M., Hartmann, C., Cousement, N., and Massart, D.L. 1995. Determination of nitrate and nitrite in vegetables by capillary electrophoresis with indirect detection. 1995. Journal of Chromatography A, 706:479-492.

Jobgen, W.S., Jobgen, S.C., Li, H., Meininger, C.J., and Wu, G-Y. 2007. Analysis of nitrite and nitrate in biological samples using high-performance liquid chromatography. Journal of Chromatography B, 851:71-82.

Joossens, J.V., Hill, M.J., Elliott, P., Stamler, R., Stamler, J., Lesaffre, E., Dyer, A., Nichols, R., and Kesteloot, H. 1996. Dietary salt, nitrate and stomach cancer mortality in 24 countries. International Journal of Epidemiology, 25:494-504. Kamataki, T., Nunoya, K-I., Sakai, Y., Kushida, H., and Fujita, K-I. 1999. Genetic polymorphism of CYP2A6 in relation to cancer. Mutation Research, 428:125-130. Kato, I., Tominaga, S., and Matsumoto, K. 1992. A prospective study of stomach cancer among a rural Japanese population: a 6-year survey. Japanese Journal of Cancer Research, 83:568-575.

89 Kazemzadeh, A., and Ensafi, A.A. 2001. Simultaneous determination of nitrite and nitrate in various samples using flow-injection spectrophotometric detection. Microchemical Journal, 69:159-166. Knight, T.M., Forman, D., Al-Dabbagh, S.A., and Doll, R. 1987. Estimation of dietary intake of nitrate and nitrate in Great Britain. Food and Chemical Toxicology, 25:277- 285.

Kozlov, A.V., Staniek, K., and Nohl, H. 1999. Nitrite reductase activity is a novel function of mammalian mitochondria. FEBS Letters, 454:127-130. Kushida, H., Fujita, K-I., Suzuki, A., Yamada, M., Nohmi, T., and Kamataki, T. 2000. Development of a Salmonella tester strain sensitive to promutagenic N-nitrosamines: expression of recombinant CYP2A6 and human NADPH-cytochrome P450 reductase in S. typhimurium YG7108. Mutation Research, 471:135-143. Larsson S.C., Bergkvist, L., and Wolk, A. 2006a. Processed meat consumption, dietary nitrosamines and stomach cancer risk in a cohort of Swedish women. World Journal of Gastroenterology, 12:4296-303. Larsson, S.C., Orsini, N., and Wolk, A. 2006b. Processed meat consumption and stomach cancer risk: a meta-analysis. Public Health Nutrition, 9:449-64. Lee, J-K., Park, B-J., Yoo, K-Y., and Ahn, Y-O. 1995. Dietary factors and stomach cancer: a case-control study in Korea. International Journal of Epidemiology, 24:33- 41. Li, H., Meininger, C.J., and Wu, G. 2000. Rapid determination of nitrite by reversed- phase high-performance liquid chromatography with fluorescence detection. Journal of Chromatography B, 746:199-207. Lichon, M. J. 1996. ‘Sample preparation’ in Handbook of Food Analysis: Physical Characterization and Nutrient Analysis, vol. 1. ed L. M. L. Nollet, Marcel Dekker Inc., New York. Linseisen, J., Rohrmann, S., Norat, T. Gonzalez, C.A., Dorronsoro, I.M., Skeie, G., Slimani, N., and Riboli, E. 2006. Dietary intake of different types and characteristics of processed meat which might be associated with cancer risk-results from the 24-hour diet recalls in the European Prospective Investigation into Cancer and Nutrition (EPIC). International Journal of Cancer, 119:915-9. Lundberg, J.O., and Govoni, M. 2004. Inorganic nitrate is a possible source for systemic generation of nitric oxide. Free Radical Biology and Medicine, 37:395-400. Lundberg, J.O., Feelisch, M., Björne, H., Jansson, E.Å., and Weitzberg. 2006. Cardioprotective effects of vegetables: Is nitrate the answer? Nitric Oxide, 15:359-362. Marques-Vidal, P., Ravasco, P., and Camillo, M.E. 2006. Foodstuffs and colorectal cancer risks: A review. Clinical Nutrition, 25:14-36. Marshall, P.A., and Trenerry, V.C. 1996. The determination of nitrite and nitrate in foods by capillary ion electrophoresis. Food Chemistry, 57:339-345.

90 Marzinzig, M., Nussler, A.K., Stadler, J., Marzinzig, E., Barthlen, W., Nussler, N.C., Beger, H.G., Morris, Jr., S.M., and Brückner, U.B. 1997. Improved methods to measure end products of nitric oxide in biological fluids: Nitrite, nitrate, and S-nitrosothiols. Nitric Oxide: Biology and Chemistry, 1:177-189. Massey, R.C. 1991. ‘Methods for the analysis of nitrate and nitrite in food and water’ in Nitrates and Nitrites in Food and Water, ed M. Hill. Ellis Horwood, New York. Massey, R.C. 1997. Estimation of daily intake of food preservatives. Food Chemistry, 60:177-185. Di Matteo, V., and Esposito, E. 1997. Methods for the determination of nitrite by high- performance liquid chromatography with electrochemical detection. Journal of Chromatography A, 789:213-219. McKnight, G.M., Smith, L.M., Drummond, R.S., Duncan, C.W., Golden, M., and Benjamin, N. 1997. Chemical synthesis of nitric oxide in the stomach from dietary nitrate in man. Gut, 40:211-214. Meah, M.N., Harrison, N., and Davies, A. 1994. Nitrate and nitrite in foods and the diet. Food Additives and Contaminants, 11:519-532. Melanson, J.E., and Lucy, C.A. 2000. Ultra-rapid analysis of nitrate and nitrite by capillary electrophoresis. Journal of Chromatography A, 884:311-316. Miranda, K.M., Espey, M.G., and Wink, D.A. 2001. A rapid, simple spectrophotometric method for simultaneous detection of nitrate and nitrite. Nitric Oxide: Biology and Chemistry, 5:62-71. Mirvish, S.S. 1995. Role of N-nitroso compounds (NOC) and N-nitrosation in etiology of gastric, esophageal, nasopharyngeal and bladder cancer and contribution to cancer of known exposures to NOC. Cancer Letter, 93:17-48. Mitsui, T., and Kondo, T. 2002. Assessing nitrate metabolism in the intestinal tract by measuring breath nitric oxide and nitrous oxide, and its clinical significance. Clinica Chimica Acta, 319:57-62. Mowat, C., and McColl, K.E. 2001. Alternations in intragastric nitrite and vitamin C levels during acid inhibitory therapy. Best Practice and Research Clinical Gastroenterology, 15:523-537. National Academy of Sciences. 1981. The Health Effects of Nitrate, Nitrite and N- Nitroso Compounds. National Academy Press, Washington. Nollet, L.M.L. 1996. ‘Handbook of Food Analysis:physical characterization and nutrient analysis. Ohkawa, S., Ohishi, K., Yoneyama, T., Hishida, A., and Kumagai, H. 1998. The effect of eliminating interfering substances by solid-phase extraction and sample-pH adjustment in the colorimetric assay of urinary nitrite and nitrate. Analytical Biochemistry, 258:379-381.

91 Olesen, P.T., Meyer, A.S., Stahnke, L.H. 2004. Generation of flavour compounds in fermented sausages-the influence of curing ingredients, Staphylococcus starter culture and ripening time. Meat Science, 66:675-687. Ologhobo, A.D., Adegede, H.I., and Maduagiwu, E.N. 1996. Occurrence of nitrate, nitrite and volatile nitrosamines in certain feedstuffs and animal products. Nutrition and Health, 11:109-114. Öztekin, N., Nutku, M.S., and Erim, F.B. 2002. Simultaneous determination of nitrite and nitrate in meat products and vegetables by capillary electrophoresis. Food Chemistry, 76:103-106. Pannala, A., Mani, A.R., Spencer, J.P.E., Skinner, V., Bruckdorfer, K.R., Moore, K.P., and Rice-Evans, C.A. 2003. The effect of dietary nitrate on salivary, plasma, and urinary nitrate metabolism in humans. Free Radical Biology and Medicine, 34:576- 584. Pearson, A.M., and Gillett, T.A. 1996. Processed Meats, 3rd Edition. Chapman and Hall, New York. Pegg, R.B., and Shahidi, F. 1996. A novel titration methodology for elucidation of the structure of preformed cooked cured-meat pigment by visible spectroscopy. Food Chemistry, 56:105-110.

Pennington, J.A. 1998. Dietary exposure models for nitrates and nitrites. Food Control, 9:389-395. Pérez-Rodríguez, M.L., Bosch-Bosch, N., and Garciá-Mata, M. 1996. Monitoring nitrite and nitrate residues in frankfurters during processing and storage. Meat Science, 44:65-73. Pham, H.N., Benitez, A., Hommet, F., Bombe, D., Schoefs, O., and Pauss, A. 2008. A new quantitative and low-cost determination method of nitrate in vegetables, based on deconvolution of UV spectra. Talanta, 76:936-940. Prasad, S., and Chetty, A.A. 2008. Nitrate-N determination in leafy vegetables: study of the effects of cooking and freezing. Food Chemistry, 106:772-780. Rajnakova, A., Moochhala, S., Goh, P.M.Y., and Ngoi, S.-S. 2001. Expression of nitric oxide synthase, cyclooxygenase, and p53 in different stages of human gastric cancer. Cancer Letters, 172:177-185. Reece, P., and Hird, H. 2000. Modification of the ion exchange HPLC procedure for the detection of nitrate and nitrite in dairy products. Food Additives and Contaminants, 17:219-222. Rieder, G., Hofmann, J.A., Hatz, R.A., Stolte, M., and Enders, G.A. 2003. Up- regulation of inducible nitric oxide synthase in Helicobacter pylori-associated gastritis may represent an increased risk factor to develop gastric carcinoma of the intestinal type. International Journal of Medical Microbiology, 293:403-412. del Río, L.A., Corpas, F.J., and Barroso, J.B. 2004. Nitric oxide and nitric oxide synthase activity in plants. Phytochemistry, 65:783-792.

92 Rincón, F., Martínez, B., and Delgado, J.M. 2003. Detection of factors influencing nitrite determination in meat. Meat Science, 65:1421-1427.

Reinik, M., Tamme, T., Roasto, M., Juhkam, K., Jurtšenko, S., Tenńo, T., and Kiis, A. 2005. Nitrites, nitrates and N-nitrosamines in Estonian cured meat products: Intake by Estonia children and adolescents. Food Additives and Contaminants, 22:1098-1105. Risch, H.A., Jain, M., Choi, N.W., Fodor, J.G., Pfeiffer, C.J., Howe, G.R., Harrison, L.W., Craib, K.J., and Miller, A.B. 1985. Dietary factors and the incidence of cancer of the stomach. American Journal of Epidemiology, 122:947-959. Ruddell, W.S.J., Bone, E.S., Hill, M. J., Blendis, L.M., and Walters, C.L. 1976. Gastric-juice nitrite: A risk factor for cancer in the hypochlorhydric stomach? The Lancet, 308:1037-1039.

Sanz, Y., Vila, R., Toldra, F., and Flores, J. 1998. Effect of nitrate and nitrite curing salt on microbial changes and sensory quality of non-fermented sausages. International Journal of Food Microbiology, 42:213-217. Sebranek, J.G., and Bacus, J.N. 2007. Cured meat products without direct addition of nitrate or nitrite: what are the issues? Meat Science, 77:136-147. Seel, D.J., Kawabata, T., Nakamura, M., Ishibashi, T., Hamano, M., Mashimo, M., Shin, S.H., Sakamoto, K., Jhee, E.C., and Watanabe, S. 1994. N-nitroso compounds in two nitrosated food products in southwest Korea. Food and Chemical Toxicology, 32:1117-1123. Shafiur Rahman, M. 1999. “Chapter 14: Nitrites in Food Preservation”. In, Handbook of Food Preservation, ed. M. Shafiur Rahman. Marcel Dekker, Inc. New York. Shapiro, K.B., and Hotchkiss, J.H. 1996. Induction of nitric oxide synthesis in murine macrophages by Helicobacter pylori. Cancer Letters, 102:49-56. Shank, R.C. 1975. Toxicology of N-nitroso compounds. Toxicology and Applied Pharmacology, 31:361-368.

Shephard, S.E., Schlatter, C., and Lutz, W.K. 1987. Assessment of the risk of formation of carcinogenic N-nitroso compounds from dietary precursors in the stomach. Food and Chemical Toxicology, 25:91-108.

Shiotani, A., Iishi, H., Kumamoto, M., and Nakae, Y. 2004. Helicobacter pylori infection and increased nitrite synthesis in the stomach. Digestive and Liver Disease, 36:327-332. Siu, D.C., and Henshall, A. 1998. Ion chromatography determination of nitrate and nitrite in meat products. Journal of Chromatography A, 804:157-160. Skrökki, A. 1995. Additives in Finnish sausages and other meat products. Meat Science, 39:311-315.

Skrypec, D.J., Gray, J.I., Mandagere, A.K., Booren, A.M., Pearson, A.M., and Cuppett, S.L. 1985. Effect of bacon composition and processing on N-nitrosamine formation. Food Technology, 39:74-79.

93 Stalikas, C.D., Konidari, C.N., and Nanos, C.G. 2003. Ion chromatographic method for the simultaneous determination of nitrite and nitrate by post-column indirect fluorescence detection. Journal of Chromatography A, 1002:237-241. Sydney Water. 2007. Sydney water quality monitorting program, Typical drinking water analysis. Available online at www.sydneywater.com.au

Torró, I.G., García Mateo, J.V., and Martínez Calatayud, J. 1998. Flow-injection - - biamperometric determination of nitrate (by photoreduction) and nitrite with the NO2 /I reaction. Analytica Chimica Acta, 366:241-249. Tricker, A.R., and Preussmann, R. 1991. Carcinogenic N-nitrosamines in the diet: occurrence, formation, mechanisms and carcinogenic potential. Mutation Research/Genetic Toxicology, 259:277-289.

Tricker, A.R., and Kubacki, S.J. 1992. Review of the occurrence and formation of non- volatile N-nitroso compounds in foods. Food Additives and Contaminants, 9:39-69. Tsikas, D. 2005. Methods of quantitative analysis of the nitric oxide metabolites nitrite and nitrate in human biological fluids. Free Radical Research, 39:797-815.

Tsikas, D., Gutzki, F-H., and Stichtenoth, D.O. 2006. Circulating and excretory nitrite and nitrate as indicators of nitroc oxide synthesis in humans: methods of analysis. European Journal of Clinical Pharmacology, 62:51-59.

Tsikas, D. 2007. Analysis of nitrite and nitrate in biological fluids by assays based on the Griess reaction: Appraisal of the Griess reaction in the L-arginine/nitric oxide area of research. Journal of Chromatography B, 851:51-70.

Usher, C.D., and Telling, G.M. 1975. Analysis of nitrates and nitrite in foodstuffs: A critical review. Journal of the Science of Food and Agriculture, 2:1793-1805. Velasco-Arjona, A., García-Garrido, J.A., Quiles-Zafra, R., and Luque de Castro, M.D. 1998. Full automated robotic method for the determination of chloride, nitrite and nitrate in cured meat products. Talanta, 46:969-976. van Velzen, A.G., Sips, A.J., Schothorst, R.C., Lambers, A.C., and Meulenbelt, J. 2008. The oral bioavailability of nitrate from nitrate-rich vegetables in human. Toxicology Letters, 181:177-181. van Vliet, J.J., Vaessen, H.A., van de Burg, G., and Schothorst, R.C. 1997. Twenty- four-hour duplicate diet study 1994; nitrate and nitrite: method development and intake per person per day. Cancer Letters, 114:305-307. Vetrovsky, P., Stoclet, J-C., and Entlicher, G. 1996. Possible mechanism of nitric oxide production from NG-hydroxy-L-arginine or hydroxylamine by superoxide ion. International Journal of Biochemistry and Cell Biology, 28:1311-1318. Walters, C.L. 1980. The exposure of humans to nitrite. Oncology, 37:289-296. Wang, G.F., Horita, K., Satake, M. 1998. Simultaneous spectrophotometric determination of nitrate and nitrite in water and some vegetable samples by column preconcentration. Microchemical Journal, 58:162-174.

Wardencki, W., Curylo, J., and Namiesnik, J. 2007. Trends in solventless sample preparation techiniques for environmental analysis. Journal of Biochemical and Biophysical Methods, 70:275-288.

Weisburger, J.H. 1981. ‘N-nitroso compounds: diet and cancer trends’, in N-nitroso Compounds. eds. R.A. Scanlan and S.R. Tannenbaum. American Chemical Society Symposium Series 174, United States.

Wennmalm, A., Benthin, G., Edlund, A, Jungersten, L., Kieler-Jensen, N., Lundin, S., Westfelt, U.N., Petersson, A.S., and Waagstein, F. 1993. Metabolism and excretion of nitric oxide in humans. An experimental and clinical study. Journal of the American Heart Association, 73:1121-1127.

White Jr., J.W. 1975. Relative significance of dietary sources of nitrate and nitrite. Journal of Agricultural and Food Chemistry, 23:886-891. Wogan, G.N., and Tannenbaum, S.R. 1975. Environmental N-nitroso compounds: Implications for public health. Toxicology and Applied Pharmacology, 31:375-383.

Wootton, M., Kok, S.H., and Buckle, K.A. 1985. Determination of nitrite and nitrate levels in meat and vegetable products by high performance liquid chromatography. Journal of the Science of Food and Agriculture, 36:297-304. World Health Organization 1973. Toxicological evaluation of certain food additives with a review of general principles and of specifications. 17th Report of the Joint FAO/WHO Expert Committee on Food Additives. FAO Nutrition Report Series No. 53 (Geneva: WHO). Ximenes, M.I.N., Rath, S., and Reyes, F.G.R. 2000. Polarographic determination of nitrate in vegetables. Talanta, 51:49-56.

Xu, J., Xu, X., and Verstraete, W. 2000. Adaptation of E. coli cell method for micro- scale nitrate measurement wit the Griess reaction in culture media. Journal of Microbiological Methods, 41:23-33. Yamasaki, H., and Sakihama, Y. 2000. Simultaneous production of nitric oxide and peroxynitrite by plant nitrate reductase: in vitro evidence for the NR-dependent formation of active nitrogen species. FEBS Letter, 468:89-92.

Yue, X-F., Zhang, Z-Q, and Yan, H-T. 2004. Flow injection catalytic spectrophotometric simultaneous determination of nitrite and nitrate. Talanta, 62:97- 101. Zanardi, E., Dazzi, G., Madarena, G., and Chizzolini, R. 2002. Comparative study on nitrite and nitrate ions determination. Ann. Fac. Medic. Vet. Di Parma, 22:79-86. Zatar, N.A., Abu-Eid, M.A., and Eid, A.F. 1999. Spectrophotometric determination of nitrite and nitrate using phosphomolybdenum blue complex. Talanta, 50:819-826.

95 Chapter 3 N-nitroso compounds and nitrosation inhibitors

Literature review on N-nitroso compounds including nitrosamines and nitrosamides will be presented, followed by their nitrosation promoters, inhibitors and their interactions.

3.1 Dietary sources of N-nitroso compounds

NOCs are almost exclusively found in foods and beverages containing nitrite such as cured meat or those that have been exposed to nitrogen oxides for example beer. In the cured meat group, bacon has the highest NOC content of NDMA ranging from 10 to 100 µg/kg when cooked, which corresponds to a dietary intake of approximately 1 µg of NDMA per 100 g portion. In beer, the NDMA can range from 5 to 70 µg/L and is higher in German beer varieties (Lijinsky, 1999).

There are several hundred NOCs that exist; however, only a few are encountered by humans outside the laboratory mainly in food, industrial plants and the environment. Nitrosamines are the most common occurring NOC encountered, which are stable and not directly acting. Nitrosamines on the whole were shown to induce tumors in a variety of organs including liver, lung, kidney, bladder pancreas, esophagus and tongue in animal experiments. In contrast, nitrosamides are unstable and direct acting; for examples alkylnitrosoureas and alkylnitrosocarbamates, and can induce tumors of the nervous system, stomach, GIT and bone in animals (Lijinsky, 1999).

3.2 N-nitroso compounds

NOCs are carcinogens formed in food or in the digestive tract of mammals. They are carcinogenic because they form potent electrophilic alkylating agents. The electrophiles will then react with DNA of target tissue and alter the bases thus initiating the start of carcinogenesis (Archer, 1989). There are approximately 21 identified NOCs in foods and beverages. Non-volatile nitrosamines, such as N-nitrosoproline and N- nitrosothiazolidine-4-carboxylic acid, account for almost half of them (Tricker and Kubacki, 1992).

96 NOCs are powerful carcinogens and were demonstrated to induce tumours in numerous organs of all tested animal species (Gangolli et al., 1994), and can be divided into three broad groups which are: 1) N-nitrosamines derived from secondary amines containing dialkyl, alkylaryl and diaryl substituents, 2) N-nitrosamides derived from N-alkylureas, N-alkylcarbamates or cyanamides, and 3) simple N-alkylamides (Tricker and Kubacki, 1992). However, more recently, Pourazrang et al. (2002) had categorized NOCs into six categories, which are: 1) volatile N-nitrosamine, 2) non-volatile N-nitrosamine, 3) N-nitrosamide, 4) N-nitrosated glycosylamines and amadori compounds, 5) N- nitrosated heterocyclic carboxylic acids, and 6) products formed when amino sugar reacted with nitrite. Figure 3 listed the names and formulae of NOC precursors. Since nitrosation reactions occur via diaotization and nucleophilic replacement of the amino group, primary amines are generally not considered as precursors of N-nitrosamines (Tricker and Kubacki, 1992).

Shephard et al. (1987) grouped dietary precursors (Figure 2.1) of NOCs in classes depending on its probability of endogenous nitrosation. They concluded that ureas, aromatic amines were potentially important risk factors in gastric cancer. On the other hand, amides guanidines, and primary amino acids had uncertain risk status. Finally, alkylamines and secondary amino acids probably had a negligible role in gastric cancer.

Nitrosation reactions can be inhibited by redox compounds such as ascorbate and vitamin E, or accelerated by catalysts such as metal ions, carbonyl compounds and nucelophilic anions including chlorine and iodine (Tricker and Preussmann, 1991).

Tricker and Preussmann (1991) had grouped foods that are most commonly contaminated with NOCs into six broad groups, and these are: 1) foods preserved with either nitrate and/or nitrite that act as nitrosating agents, 2) foods such as fish and meat preserved by smoking which generates nitrogen oxide that can participate in nitrosation, 3) foods subjected to drying by combustion gases, again includes nitrogen oxide, which include foods such as malt in beer production and other dried products, 4) foods preserved by pickling or salting especially plant-based foods in which microbial reduction occurs, 5) foods that are stored under humid conditions and are therefore susceptible to fungal metabolites from Fusarium species that can produce nitrosamines

97 (Tricker and Kubacki, 1992), and finally 6) food packaging materials that are in contact with the food and allows the migration and formation of nitrosamines.

3.2.1 N-nitrosamines

Nitrosamines as a class are potent carcinogens. However, the specific action and hence potency depends on numerous factors including its structure (Figure 2.2), frequency and route of administration, species, strain and other variables (Weisburger, 1981). N- nitrosamines are stable at neutral pH and require metabolic activation by mammalian enzyme systems before they can exert any mutagenic effects (Shank, 1975), whereas nitrosamides are direct acting alkylating agents (Gangolli et al., 1994). Thus nitrosamines can produce tumours in tissues remote from the site of administration (Archer, 1989). For example, the conversion of nitrates to nitrites in the oral cavities followed by the formation of nitrosamine in the stomach under acidic pH.

The proposed mechanisms by which bacteria contribute to the formation of NOCs both in vitro and in vivo include: 1) decrease pH of the system, 2) reduction of nitrate to nitrite, 3) adsorption of amine onto cell surface or cytoplasm membrane and 4) enzymatic transformation (Ralt and Tannenbaum, 1981). Creating an acidic environment that catalyses the formation of nitrosamines. However, nitrosation is possible above neutral pH in the presence of micro-organisms, thiocyanate, metals, phenol and other nitroso compounds, which acts as catalysts (Ralt and Tannenbaum, 1981). The reduction of nitrate to nitrite and amine adsorption increases the formation of nitrosamine of the two essential component of nitrosamine synthesis pathway. Direct enzymatic conversion of nitrite to nitrosamine is also possible. Examples of NOCs and their associated food product and the relative carcinogenicity for each common nitrosamine found in food are listed in Figure 3.1.

Figure 3.1 Nitrosamines found in foods and their averaged concentration. Foods high in nitrosamines mostly belongs to the cured meat and fermented fish group, with N- nitrosodimethylamine, diethylamine and methylbenzylamine being the most carcinogenic (Lijinsky, 1999).

The optimal pH for nitrosamine formation is at 3.4, which is the pKa for nitrous acid. However, other factors have been identified that interferes with this reaction. For example, not all nitrosatable substances follow the same kinetic pattern and that nitrosation are multi-step processes. A number of chemical compounds can also inhibit, slow down or accelerate the reaction. In addition, foods and stomach contents are non- homogenous thus partitioning of reactants between phases may pose further changes in the rate of nitrosamine formation (Wogan and Tannenbaum, 1975).

Evidence suggests that volatile N-nitrosamines are rapidly absorbed from the duodenum, which is subsequently carried by the portal circulation to the liver (Hashimoto et al., 1976, in Gangolli et al., 1994), possibly for breaking down N- nitrosamines to less toxic forms.

N-nitrosamines can be measured in human blood, urine and fecal matter after the consumption of nitrite or nitrate containing foods such as bacon and spinach, which demonstrated the in vivo production of NOCs in humans (Wagner and Tannenbaum, 1985).

Most drugs are known to cross the placenta but recently it was demonstrated that food mutagen such as NDMA can also cross the placenta from maternal to fetal circulation (Annola et al., 2007). The implication of this finding is enormous since it may involve

99 dietary changes by pregnant mothers such as the avoidance of cured meat products and to increase vitamin C intake in order to minimize in vivo NDMA formation that may harm the baby.

3.2.2 Nitrosation of amines and amides

The biotransformation of NOC is dependent on the cytochrome P450 hydroxylation of the carbon atom adjacent to the N-nitroso group. Spontaneous cleavage of the carbon- nitrogen bond then follows in the α-hydroxynitrsamine that gives rise to an aldehyde and the alkyldiazohyroxide. Finally, the diazohydroxide can produce the potent electrophilic alkyl diazonium ion (Figure 3.2), which either forms an alcohol with water or react with nucleophiles such as DNA. In addition to α-oxidation, β-oxidation and w-oxidation are also involved in the metabolic transformation of numerous nitrosamines (Archer, 1989).

Figure 3.2 Degradation of NOCs forming electrophilic products (Shephard et al., 1987).

100 The nitrosation of amines is based on a nucleophilic substitution reaction (Figure 3.3a), whereby the electron-rich amine nitrogen interacts with the nitrogen of nitrous anhydride, which replaces the nitrite group, causing the latter to separate. The reactants formed are the unprotonated amine, R1R2NH, and two molecules of nitrous acid. Thus the limiting step for determining the rate of amine nitrosation involves the acid-base equilibria of both amine and nitrous acid (Shephard et al., 1987).

Amide nitrosation involves nucleophilic attack of the nitrogen on protonated nitrous acid, with water acting as the leaving group (Figure 3.3b). Unlike amine nitrosation, the nitrosation of amide is first order reaction with respect to nitrous acid, thus the latter is less sensitive to the concentration of nitrite in the stomach. Amide nitrogen is much less basic than an amine, and is usually unprotonated at acid pH. Thus amide nitrosation is dependent only on the nitrous acid-nitrite equilibrium, where decreasing pH lead to increasing reaction rate (Shephard et al., 1987).

Figure 3.3 Mechanisms of nitrosation of a) amines and b) amides (Shephard et al., 1987).

101 Gerbaux et al. (2004) demonstrated that nitrosation occurs through the actions of nitrosonium cation (NO+) carriers such as nitrosothiols or nitrosamines under physiological pH. It was also shown that peroxynitrite can nitrosate phenol by a CO2- dependent pathway, and that nucleophiles such as amines and thiols with pKa values between 7 and 9 could be nitrosated at physiological pH (Uppu et al., 1998). In addition, certain drugs contain secondary and tertiary amines that may contribute to the overall burden of nitrosating agents (Mirvish et al., 1972).

Oxygenated nitrogen species including protonated nitrous acid, dinitrogentrioxide, dinitrogentetroxide and peroxynitrite can react with amines to form molecular nitrogen in a spontaneous reaction with primary amines or via cytochrome P450 catalysed reaction with secondary amines. In addition, secondary amines can be nitrosated by either nitrite or NO. Thus the quantification of molecular nitrogen can provide an indirect estimate of the level of oxygenated nitrogen species acting as nitrosating agents. Molecular nitrogen production is thought to play a protective role-taking place mainly in the gastrointestinal tract to control the body’s NO and nitrite levels (Junghans et al., 1999).

It was suggested that carbon dioxide reacts with dimethylamine, which is a precursor to NDMA, then reacts with nitrite anion to form NDMA. Hence carbon dioxide appears to be a catalyst to promote the formation of a common volatile nitrosamine (Lv et al., 2006).

3.3 Estimated dietary nitrosamine intake

The largest known source of human exposure to exogenous nitrosamines is not foods but the environment such as in the workplace. However, some food and beverage can contain greater than 10 µg/kg of NDMA (Biaudet et al., 1994).

Yamamoto et al., (1984) estimated the average daily intake of endogenous volatile nitrosamines in Japanese population was 0.5 µg per person and 88 % of it came from fish products, which were significant part of their diet. The estimation was based on food consumption data and their analysis of volatile nitrosamine content of commonly eaten food including fresh fish, fish products, meat, cheese, and beer. NDMA was the most common occurring volatile nitrosamine in Chinese foods with up to 6 µg/kg in

102 beer, up to 7.4 µg/kg in meat products, and up to 131.5 µg/kg in some seafood. Other volatile nitrosamines detected were NDEA, NPIP and NPYR (Song and Hu, 1988).

Österdahl (1988) estimated the average dietary intake of volatile nitrosamines in Sweden populations was 0.29 μg per person, and over 80 % of that came from meat and malt products, which was highest in fried bacon and pork. Beer and other malt beverages contributed 14 % of the dietary intake of volatile N-nitrosamines in Sweden, with dried products, cheese and smoked with represent 7 %. Österdahl (1988) reported that NDMA was the most common volatile N-nitrosamine in foods such as cocoa and chocolate products (up to1.2 μg /kg), tea (up to1.2 μg /kg), coffee (0.1-0.8 μg /kg), and cereal products (0.2-0.9 μg /kg). In West Germany, Tricker et al. (1991) estimated the total volatile nitrosamine intake from food for men and women were 0.3 and 0.2 μg/day, respectively. Furthermore, 31 % of the daily intake of NDMA for men came from the consumption of beer, thus represented a significant source of dietary volatile nitrosamine.

In West Germany, the mean daily intake of volatile nitrosamines between 1989 and 1990 were: 0.28 and 0.17 µg NDMA/day for men and women, respectively. A third of the dietary NDMA for men resulted from the consumption of beer. For NPYR and NPIP, the averaged daily intake for both men and women was the same at 0.011 and 0.015 µg NPYR and NPIP/day, respectively (Tricker et al., 1991).

The NDMA range found in sausage products was 0.5 to 1.8 µg/kg with a mean of 0.84 µg/kg, and for bacon and ham it ranged from 0.5 to 1.6 µg/kg with a mean of 1.01 µg/kg. NDMA was highest in fresh fish and fish products with a range of 0.5 to 8.0 µg/kg and a mean of 2.18 µg/kg. Despite this, meat and meat products contributed more than double the daily dietary intake of NDMA for both men and women compared to fish products. There was no detectable NDMA in fresh vegetables and only negligible levels were found in preserved vegetables. Other volatile nitrosamines including NPIP and NPYR were not present in vegetables and only negligible levels were found in sausage products (Tricker et al., 1991).

Tricker et al. (1991) compiled the mean daily dietary intake of NDMA from different countries ranged from 0.08 to 1.10 µg/day with a mean of 0.5 µg/day based on published dietary surveys between 1978 and 1990. It was concluded that beer, cured meat and fish products contributed to most of their dietary NDMA intake.

103 The mean daily intake of NDMA from French foods and beverages between 1987 and 1992 was 0.19µg/day, and one-third of this came from the consumption of alcoholic beverages. Vegetables represented 22 % of NDMA intake followed by meat and meat products at 12.5 % (Biaudet et al., 1994).

Mitacek et al. (1999) concluded that based on case-control studies in Thai population that exposure to exogenous and possible endogenous nitrosamines in food or tobacco may contribute to the development of liver cancer. With over 1800 food samples tested, relatively high levels of NDMA, NPIP and NPYR were detected in traditional Thai foods such as fermented, salted and dried fish. Levels of these volatile nitrosamines ranged from trace amount to 146 µg/kg.

Mavelle et al. (1991) measured NDMA in French food in 1987-88 and found that 89 % of samples tested contained NDMA with a maximum level of 16 µg/kg. Stone fruit spirits were shown to contain the highest level of NDMA followed by nitrite-cured meats, smoked meats and smoked fishes. Beer contained a mean NDMA level of 0.28 µg/kg, which was lower than in the past due to modification of the brewing processes. Other volatile nitrosamines found were rare and was below 0.5 µg/kg.

N-nitrosamines in food and beverages are the result of the manufacturing processes such as smoking, curing and so on. Therefore changes or modification of these processes should eliminate or at least minimize their presence in food (Mavelle et al., 1991).

3.4 Epidemiology of cancer risks

It was demonstrated that NOC concentration and stomach pH increased significantly with age, and there was a positive correlation between pH and NOCs concentration, and between pH and an increase concentration of nitrites. In addition, it was demonstrated that the population of nitrate reductase-positive microorganisms rose as NOCs and nitrite levels increased (Reed et al., 1981). The rise in gastric pH with age maybe due to decreased gastric acid secretion as a result of natural aging process. This rise in pH then allows microorganisms to survive in the stomach-converting nitrate present in the stomach to nitrite. Since nitrite is more stable at higher pH, this may allow more nitrosation to occur thus resulting in higher N-nitrosamine levels with increasing age.

104 N-nitrosamine levels ranged from 0.01 to 40 µmol/L in healthy volunteers and patients aged between 18 and 87 years. A significant rise in N-nitrosamine concentrations and gastric pH was demonstrated with increasing age. Furthermore, N-nitrosamine levels were significantly higher in males than females, taking other factors into account. Cigarette smoking had no significant effect on either N-nitrosamine or nitrite concentrations (Reed et al., 1981).

Case-control study of colorectal cancer in the United States shown that higher exposure to heterocyclic amines was strongly associated with colorectal cancer risk taking into account of meat consumption and cooking method (Nowell et al., 2002).

3.5 N-nitrosation inhibitors

3.5.1 Ascorbic acid

Ascorbic acid can prevent nitrosation by converting nitrite precursors to NO (Figure 3.4) thus preventing or at least reducing the formation of NOCs including nitrosamine (Hwang et al., 1994). Hence a strong negative association between vitamin C supplementation and gastric cancer support the protective role of ascorbic acid (Hansson et al. 1994).

Ascorbate and erythorbate are added in cured meat to inhibit formation of nitrosamines (Cassens, 1995), as well as inhibiting nitrosation from occurring within the stomach from residual nitrite (Walters, 1980). In the United States, reducers such as ascorbate or erythrobates are often added in cured meat to promote the reduction of nitrous acid to nitric acid, which reduces residual nitrite and retards the formation of N-nitrosamines (Pennington, 1998). The maximum addition of ascorbate in the United States was 550 mg/kg, of which 40 % or 209 mg/kg was detected in cured meat samples (Cassens, 1997). Thus the inhibitory effect of ascorbic acid against nitrosation reactions is well recognized.

105

Figure 3.4 Reaction of ascorbic acid with nitrite where nitrite is reduced to NO and ascorbic acid is oxidized to dehydroascorbic acid. The latter competes with amines for nitrosating species such as N2O3 (National Academy of Sciences, 1981).

Contrary to other findings, Risch et al. (1985) found that vitamin C intake was only slightly protective against gastric cancer, and that vitamin E had not effects at all. Similarly, Kato et al. (1992) revealed an adverse effect of fruit consumption, which may be due to the lack of detailed information on the types of fruits consumed in the rural Japanese population surveyed.

The addition of ascorbic acid at a concentration of 2,000 mg/kg to cured meat did not reduce the antimicrobial effect of sodium nitrite (Walters, 1980). However, 500 mg/kg ascorbic acid was shown to be more efficient at reducing the mutagenicity of NOC by 61 % than at 2,000 mg/kg ascorbic acid at only 54 % (Pourazrang et al., 2002).

Since the discovery in 1956 that NOC are carcinogenic to more than 40 animal species, human exposure to NOCs has been related to increased risk of gastro-intestinal cancers. The endogenous nitrosation increases when precursors are ingested in foods, which commonly contains these precursors. Under acidic conditions such as the stomach, nitrite can react with secondary amines to form N-nitrosamines. Ascorbic acid and ascorbate can inhibit nitrosation because they react faster than the amine with the nitrosating agents. AA can reduce nitrous acid to nitric oxide that is not a nitrosating agent, and is itself oxidized to dehydroascorbic acid. Whole strawberries, garlic juice or kale juice decreased NDMA formation in humans fed an amine-rich diet with nitrate by 70, 71 and 44 %, respectively. Increased dietary nitrate intake also significantly increased salivary and urinary nitrate. It was suggested that phenolic compounds derived from plants and fruits are potent nitrosation inhibitors. In addition, strawberries contain the highest vitamin C amongst fruits at 99 mg/100g. Cysteine and other

106 sulfhydryl compounds and allyl sulfur compounds were thought to act as nitrite scavengers (Chung et al., 2002).

Data is not available for the formation of NDMA in vivo with a diet consisting of a mixture of precursors and inhibitors. However, Krul et al. (2004) demonstrated in a dynamic in vitro gastrointestinal model under human physiological conditions that orange juice and tea were effective at reducing the amount of NDMA formed under gastric condition. Furthermore, this model may be used to evaluate human cancer risk in foods containing nitrate and nitrite. Based on experiments on mice, Ohsawa et al. (2003) concluded that although ascorbic acid can reduce liver DNA damage, it could damage stomach DNA in the presence of sodium nitrite by reducing nitrite to NO, which can then be converted to other reactive nitrogen oxide species in the stomach. However, fresh juices did not affect stomach DNA in the presence of sodium nitrite, since food mixtures contains not only ascorbic acid, other components in the juice might preven the stomach DNA from NO genotoxicity.

3.5.1.1 Ascorbic acid chemistry and physiological roles

Ascorbic acid consists of an acidic hydroxyl group with pK1 = 4.04 and pK2 = 11.4 at 25 oC. The UV absorption wavelength is dependent on the pH (Table 3.1). Ascorbic acid can undergo several reactions giving rise to related compounds, but the reaction rate is dependent on other parameters including oxygen partial pressure, pH, temperature and the presence of heavy metal ions particularly copper and iron, which result in high losses. Firstly, ascorbic acid is readily and reversibly oxidized to form dehydroascorbic acid, which in aqueous media forms hydrated hemiketal. Once the lactone ring in dehydroascorbic acid is irreversibly opened, it forms 2,3-diketogulonic acid, where its biological activity is lost (Belitz and Grosch, 1987).

Table 3.1 Maximal wavelength of ascorbic acid at different pH (Belitz and Grosch, 1987).

pH λ maximum (nm) 2 244 6-10 266 >10 294

107 Plants and all mammals except humans, monkey and guinea pig can synthesize ascorbic acid in vivo. Therefore for the minority that cannot synthesize ascorbic acid they must acquire them through dietary sources (Aurand et al., 1987). Fortunately fruits and vegetables are abundant in vitamin C and are a regular part of the diet.

As an antioxidant, AA performs many physiological and biochemical functions. Firstly, it participates as a cofactor in enzymatic reactions including: the synthesis of collagen, carnitine and norepinephrine; the metabolism of tryptophan, tyrosine, histamine and cholesterol; amidation of neuropeptides; and is involved in the detoxification reactions in the liver. Secondly, the antioxidant activity can transfer two electrons between ascorbate or dehydroascorbate redox couple reactions, or it can donate one electron to inactivate highly reactive free radicals such as tocopheryl radical to protect vitamin E. Thirdly, vitamin C can competitively inhibit substrate binding reactions such as the inhibition of carcinogenic nitrosamines formation. Fourthly, it can modulate mineral absorption in the gastrointestinal tract, and lastly, Vitamin C can prevent and treat the deficiency disease known as scurvy (Russell, 2000).

3.5.1.2 Ascorbic acid interactions with nitrate, nitrite and its derivatives

Vitamin C can exist as two forms: ascorbic acid (AA) and dehydroascorbic acid (DHAA)(Figure 3.5), which when measured together represent the total vitamin C (TVC). In fasting gastric juice, vitamin C is mainly present as ascorbic acid. Once it is absorbed, it is actively secreted into and concentrated within the gastric juice of healthy stomachs. Ascorbic acid is the active form of vitamin C, and is a strong reducing agent, which inhibits nitrosation reactions at acidic pH, since nitrosating agents react preferentially with ascorbic acid to form non-nitrosating nitric oxide (NO). In the process, ascorbic acid is oxidized to the inactive DHAA (Sobala et al., 1989). DHAA has an important role in many cell types because it can be used to regenerate AA in metabolically competent cells and be utilized in the same cells or is released to the extracellular fluid. Since DHA recyclying mechanisms exist it may decrease the amount of dietary AA that humans require to ingest daily (Wilson, 2002).

Correa (1992) identified three stages and factors involved in gastric carcinogenesis. The first stages of gastritis and atrophy have been linked to H. pylori infection and excess salt intake; the latter can cause excess cell replication thus potentially increases the risk of endogenous mutation. The intermediate stages have been associated with the

108 ascorbic acid and nitrate/nitrite ratio, which determines the degree of endogenous N- nitrosation. The last stages of gastric carcinogenesis had been linked to β-carotene and excessive salt intake.

Once nitrite is swallowed, it is reduced to nitric oxide by ascorbic acid normally present in the stomach of healthy individuals. The nitric oxide is then absorbed through the mucosa, and the ascorbic acid is oxidized to dehydroascorbic acid. However, in the absence of ascorbic acid, nitrosating compounds such as nitrite can nitrosate precursors to form NOC (Mowat and McColl, 2001).

Figure 3.5 Ascorbic acid and its common oxidation products (Washko et al., 1992).

As the enterosalivary recirculation of nitrate occurs, a constant supply of salivary nitrite to the stomach is ensured. Since ingested nitrite continuously reacts in vivo with gastric juice contents including ascorbic acid, their content is usually undetectable in the acid stomach (Mowat and McColl, 2001). However, once anti-secretary agents are released in the stomach, the pH increases to neutral, where nitrite remains stable and do not react with ascorbic acid, hence accumulates in the stomach (Mowat and McColl, 2001).

109 Nitrite and its nitrogen oxide derivatives can react with other nitrogenous organic compounds present in food and gastric juice to form NOCs. The level of NOC formed is dependent on the amount of available nitrite and ascorbic acid. This nitrosation reaction occurs spontaneously in acidic gastric juice. However, in achlorhydric stomach of pH > 4, bacteria present on the stomach lining can catalyse the nitrosation reaction (Mirvish, 1995).

Correa (1992) identified several risk factors for mid and distal gastric cancer based on epidemiological research. The risk factors were: achlorhydria or hypochlorhydria, diets high in nitrate or nitrite and low in vitamin C and Helicobacter pylori infection. It was proposed that bacteria colonized in the stomach of hypochlorhydria patients might convert ingested nitrate to nitrite and then catalyze the nitrosation process. It was shown that bacteria could colonize the stomach when intragastric pH is above 4 (Gray and Shiner, 1967).

3.5.1.3 Kinetics and mass transfer of ascorbic acid and nitrosation

Both ascorbic acid and ascorbate ion, or ASC, can indirectly inhibit nitrosation by competing for nitrosating agents in the stomach formed from reactions with nitrite and gastric juice contents. Subsequently, for each mole of ascorbic acid and ascorbate ion oxidized, which is an irreversible process, two moles of non-nitrosating nitric oxide is produced from nitrite. Although this reaction is fast, in the presence of oxygen, NO may be reconverted to nitrosating agents, thus counteracting the inhibition effects of ascorbic acid and ascorbate ion (Licht et al., 1988).

Nitrite can be lost from the gastric juice by either oxidation to nitrate (Equation 5) and/or by forming volatile species such as nitric oxide and nitrogen dioxide (Equation 3). It was demonstrated by Licht et al. (1988) that loss of nitrite in vitro was more rapid at low pH and with high rates of agitation. It was suggested that at lower pH more nitrous acid is formed (Equation 1), which shifts the equilibria toward increased production of volatile species, namely NO, NO2 and N2O4 (Equations 2-4).

Furthermore, increasing agitation rate promoted further removal of NO and NO2 from the liquid phase.

110 - + NO2 + H ↔ HNO2…………….....………………………………………...... ………(1)

2HNO2 ↔ N2O3 + H2O……...... ……………………………………………….....…..(2)

N2O3 ↔ NO + NO2…………………………..…………………………….....……….(3)

2NO2 ↔ N2O4…………………………...……...…………………………….…….…(4)

+ - - N2O4 + H2O → 2H + NO2 + NO3 …………………...…………………...………….(5) Equation 7 accounted for all the ASC lost. In addition, it also increases the loss of nitrite via the formation of NO after reacting with NOX. The reaction between ASC and nitrite depends on the competing reactions between the recycling of NO back into nitrosating agent, and the removal of NO via mass transfer. The recycle process is second order in NO while mass transfer is first order. This means that as the NO concentration increases, the recycle reaction will increase faster than physical removal. Thus anything that catalyzes the reaction between ASC and nitrite (Equation 7), for examples SCN-, Cl- and lower pH, tends to favour recycle relative to removal by mass transfer (Licht et al., 1988).

In addition, increasing temperature was shown to increase reaction (7) and the formation of NO, also favouring the recycle of ASC, thus minimizes the effectiveness of nitrosation inhibition by ASC. However, decreasing oxygen concentration was shown to decrease the rate of recycle, thus maximizes nitrosation inhibition by ASC. But since oxygen concentration has an inverse relationship with temperature, so the reaction between ASC and nitrite was dependent on other conditions. For example, more ASC was consumed at 37oC than at 23oC at pH 3.0, and the opposite trend was observed at pH 2.0 (Licht et al., 1988).

ASC- + H+ ↔ ASCo…………………………………………………………....……...(6)

+ - ASC + NOX → DHA + 2NO + H + X + H2O……………………………...…….(7)

+ + HNO2 + H ↔ NO + H2O……………………………………………...... ……...….(8) NO+ + SCN- ↔ NOSCN…………………………………………………………...... (9) NO+ + Cl- ↔ NOCl………………………………………………………………...... (10)

- o + (ASC -ascorbate ion, ASC -ascorbic acid, NOX-N2O3, NO , NOSCN, or NOCl).

111 closely related to intragastric pH, such that at pH < 5.0, the nitrite concentrations are very low. It was suggested that at normal gastric pH the nitrite content is very low because nitrite has a very short half-life (less than 10 minutes), and that it is very reactive at low pH with high absorption rate.

Although ascorbic acid shows little reactivity at neutral pH, as in the stomach of achlorhydric patients, it was shown to be a potent inhibitor of bacterial mediated N- nitrosamine formation by Pseudomonas aeruginosa, which is capable of rapid rates of N-nitrosation (Mackerness et al., 1989). Furthermore, by using in vitro studies of bacterial suspension on nitrite, Mackerness et al. (1989) had confirmed the effectiveness of ascorbate in inhibiting bacterial N-nitrosation by competing with amine for the nitrosating agents derived from the action of bacteria.

People with chronic gastritis, hypochlorhydria or increasing gastric pH resulting in an increase in the colonization of nitrite-reducing bacteria as well as a significant decline in gastric ascorbic acid concentration (Sobala et al., 1989), hence people with these conditions may be more at risk of developing gastric cancer due to elevated nitrite-to- ascorbic acid ratio, which was demonstrated by Mowat et al. (1999) that used omeprazole to induce hypochlorhydria in healthy volunteers in order to study the effects it had on gastric juice ascorbate/nitrite ratio. They concluded that pharmaceutically induced hypochlorhydria caused a decrease in intragastric ascorbic acid concentration and an increase in intragastric nitrite, thus effectively changing the ascorbate/nitrite ratio, and potentially increase the risk of gastric cancer. Ascorbic acid dereviative palmitoyl ascorbate was shown to inhibit the survival and growth of H. plyori in vitro and invivo and showed similar inhibitory effects against other pathogens including B. cereus and B. subtilis, both common foodborne outbreak bacteria (Tabak et al., 2003).

3.6.2 α–tocopherol

Skrypec et al. (1985) demonstrated that α-tocopherol-coated salts retained their anti-N- nitrosamine activity during prolonged storage at 21 oC, this may be use to reduce endogenous nitrite (Figure 3.6) thus decrease nitrosamine formation in cured meats such as bacon.

112

Figure 3.6. Reaction of α–tocopherol with nitrite where α–tocopherol function as a blocking agent by reducing nitrite and is itself oxidized to a quinone (National Academy of Sciences, 1981).

Pourazrang et al. (2002) demonstrated the synergistic effect between vitamin C and vitamin E in inhibiting NOC formation in sausages. It was proposed that ascorbic acid was capable of reducing the tocopherol radical back to α–tocopherol (Lathia, 1989, in Pourazrang et al., 2002). Hence increases the duration of the scavenging property of the antioxidants. In addition, since ascorbate has a low solubility in lipids whereas α– tocopherol has a higher solubility in lipid as a lipid-soluble vitamin, α–tocopherol may be more efficient at inhibiting the formation of nitrosamines in fatty meat products (National Academy of Sciences, 1981).

Vitamin E was shown to be protective against the effects of mutagenic by-products of nitrates and nitrites by limiting the production and availability of superoxide and nitric oxide. In addition, selenium was shown to exert its protective effects by enhancing the seleno-enzymes compounds, which reduce or scavenge peroxynitrite (ONOO-) already formed (Chow and Hong, 2002).

3.6.3 Fibre

In a case-controlled study in Toronto between 1979 and 1982, Risch et al. (1985) found that high fibre intake was associated with decreased risk of gastric cancer. This inverse relationship was also observed in case-control population of Spain, which also included vitamin C, folate, β-carotene and nitrates (González et al., 1994).

3.6.4 Polyphenols and phytochemicals

Weisburger and Chung (2002) reported the findings that tea and its polyphenols (tannic acid) and phytochemicals are bacteristatic and bacterialcidal that can lower the titer of

113 H. pylori and the associated gastric cancer incidence. In addition, tea was shown to react with nitrosating agents thereby reducing the body’s burden of carcinogens. For example, Choi et al. (2002) demonstrated a significant increase in urinary dimethylamine and trimethylamine in subjects fed with diets rich in amine, nitrate and Korean green tea. Others such as α-tocopherol were demonstrated to be an effective N- nitrosamine inhibitor in fried bacon (Skrypec et al., 1985).

Other NOC inhibitors including polyphenols in tea are scavengers for nitrite, however, some will react with the nitrite to form NOC, which can effectively nitrosate amines. Factors such as tea type, processing, preparation will determine whether tea is causative or protective (Bartsch et al., 1990).

Recently Lee et al. (2006) demonstrated that flavanols epicatechin and its dimer B2 can inhibit nitrous acid induced tyrosine nitration and the formation of carcinogenic volatile N-nitrosamine such as NDMA. Furthermore, the resulting nitroso-flavanols may inhibit the proliferation of and/or trigger apoptosis in large intestinal cancer cells.

Turmeric’s roles in cancer prevention at the cellular level was extensively studied. However, its role in inhibiting NOC formation is not yet understood. Turmeric and/or curcumin can prevent the activation of carcinogens and acts as an anti-oxidant and anti- promoter. It can also retard the conversion of pre-neoplasia and repairs damage to DNA (Krishnaswamy and Raghuramulu, 1998).

A recent review by Duvoix et al. (2005) described the properties and mode of action of curcumin on carcinogenesis, gene expression and drug metabolism with reference to their anti-tumor, antioxidant and anti-inflammatory properties. They concluded that curcumin might have potential as a cancer therapeutic agent.

3.6.5 Others

Caffeic acid can be found in fruits, vegetables, wine, olive oil, teas, and coffee beans, and along with its related compounds have been demonstrated to inhibit N-nitrosation reactions by reacting with nitrite side chain at pH 2.0 (Cotelle and Vezin, 2001). Cotelle and Vezin (2001) had shown that under acidic conditions caffeates could react with nitrite to give nitroderivatives, nitrogen heterocycles or benzofuran dimers. However, the formation of these products is dependent on the structure of caffeates as

114 either ester or acid. In addition, sodium azide had also been shown to be an inhibitor of NO production (Yamasaki and Sakihama, 2000).

Ohshima and Bartsch (1981) suggested the potential of measuring N-nitrosoproline excretion in the urine within 24 hours for the estimation of endogenous formation of NOCs in humans after the ingestion of nitrate and proline together. In addition, they had demonstrated the inhibitory effect on the nitrosation of proline in vivo by simultaneous intake of ascorbic acid and α-tocopherol.

Different fractions of curcumin showed different levels of nitrosation inhibition demonstrated by checking the mutagenicity in S. typhimurium strains with methylurea and sodium nitrite at pH 3.6 and 30 oC. Since nitrosomethlurea is a direct-acting mutagen in S. typhimurium TA1535 and TA100, it permits an accurate estimation of the increase of his+ revertants. Both turmeric extract and curcumin exhibited dose- dependent decreases of nitrosation. Under the same condition, curcumin I showed dose-dependent depletion of nitrite ions thus making nitrite unavailable to participate in nitrosation reactions. Furthermore, curcumin III showed higher affinity for nitrite ions than curcumin I (Nagabhushan et al., 1988).

Cooking methods such as boiling, frying and baking of food proteins and carbohydrates pyrolyses form Maillard reaction products, some of which are mutagenic and/or carcinogenic and may contribute to the development and progression of diabetes and age-related degenerative diseases, as well as destroying important essential amino acids. Turmeric and curcumin was shown to block the formation of hazardous Maillard reaction products and its mutagenic activity in histidine mutant strains of S. typhimurium TA98 and TA100. Curcumin is currently undergoing clinical trials for the prevention of cancer at National Cancer Institute in the United States (Kolpe et al., 2002).

Kato et al. (1992) did not find any association between different cooking methods for meats and fish and the risk of developing stomach cancer. However, in their study only a specific population of rural Japanese living in mountainous area were surveyed, therefore their conclusion may be biased and is not representative of the general population.

115 3.7 Determination of N-nitroso compounds

Nitrosamines are present in very small quantities in food in units of µg/kg. Therefore detection and analysis must be highly sensitive. In addition, due to the inherent complexity of food matrices the methods used must also be selective. Nitrosamines are divided into two major groups: volatile and non-volatile nitrosamines. Volatile nitrosamines include simple dialkyl compounds such as N-nitrosodimethylamine and N-nitrosodiethylamine that are relatively non-polar with low molecular weight that possess a sufficiently high vapour pressure for extraction from food matrix by distillation. In contrast, non-volatile nitrosamines such as NPRO and NDELA generally have a higher molecular weight, or are more polar, and hence have a relatively low vapour pressure. This makes their detection and analysis much more difficult. Due to differences in their vapour pressure, volatile nitrosamines can be analysed by GC-TEA, and non-volatile nitrosamines can be analysed using HPLC-TEA. However, due to sample matrices, identification by retention time alone from either instrument requires independent confirmation techniques including spectroscopic analysis (IR, NMR, UV and MS), identification of derivatives or parallel GC-TEA and HPLC-TEA methods. Although many methods and detectors are available for trace analysis of volatile nitrosamines in food samples, GC-TEA is the standard method due to its highly sensitive and selective detector (Sung, 2004).

Distillation is the most common extraction method for determining volatile nitrosamines in food samples since they are relatively non-polar and low molecular weight compounds (Andrade et al., 2005).

3.7.1 Volatile nitrosamines

Volatile nitrosamines are relatively non-polar, low molecular weight compounds that possess sufficient vapor pressure to be removed by distillation in a food matrix (Scanlan and Reyes, 1985), followed by isolation and separation procedures such as solvent partitioning and thin-layer electrophoresis. Gas chromatography with thermal energy analyzer (GC-TEA) as detector is commonly used for the separation and detection of volatile nitrosamines due to its relatively specific and very sensitive detector, with detection level of 1 ppb or less (Fiddler, 1975; Lijinsky, 1999). However, low or high-resolution mass spectrometry is often required for the confirmation of

116 nitrosamines found in food by comparing the peak intensity to reference peaks that have already been established (Figure 3.7)(Scanlan and Reyes, 1985).

Figure 3.7 Mass spectral peaks for the identification and confirmation of nitrosamines (Issenberg, 1981).

Volatile nitrosamines at concentrations of 0.1 to 10 ppb can be accurately determined using reliable analytical methods. Most methods will require distillation, extraction, an optional cleanup step, concentration and final separation by gas chromatography (GC). More recently, the use of thermal energy analyser as a GC detector simplified sample preparation and cleanup without sacrificing the selectivity and sensitivity of the analytical methods. For confirmation, mass spectrometry (MS) is often used to identify the different classes of nitrosamines (Issenberg, 1981).

Yamamoto et al., (1984) had demonstrated that cooking by broiling increased the volatile nitrosamines NDMA and NDEA in cooked fish compared to raw fish. It was suggested that increased temperature increased the formation of secondary amines that can act as precursors for the formation of nitrosamines. In addition, nitrosation with atmospheric nitrogen oxide may occur by broiling on a gas stove. NPIP was the most common volatile nitrosamine in salami due to the presence of NPIP precursor piperidine found in pepper, which is a common ingredient used in salami.

Meat contains the necessary starting materials for nitrosamine formation, which includes primary, secondary and tertiary amines, amides, proteins, peptides, amino

117 acids, nitrates and nitrites. In addition, microorganism can take part by nitrate reduction to nitrite and the degradation of proteins to amines and amino acids. Furthermore, the formation of higher nitrogen oxides after cooking can act as nitrosating agents. Thus the concentration of nitrosamines in meat is dependent on the method of cooking, cooking temperature and time, residual and added nitrite concentrations, concentration of nitrosamine precursors, presence of nitrosating catalysts and inhibitors, and the storage factors (Yurchenko and Mölder, 2005).

NDMA, N-nitrosopyrrolidine and N-nitrosopiperidine are the three most commonly volatile nitrosamines found in six broad food groups. These include cured meats, smoked fish/meat, foods subjected to drying by combustion gases such as malt used in beer, pickled and salt preserved foods that favors microbial nitrate reduction, foods stored under humid conditions that favors fungal growth, and finally interactions with food packaging materials (Cassens, 1995). Furthermore, volatile nitrosamine was generally not detected in other meat samples, and if present was only present at low concentration of 0.001 g/kg (Ologhobo et al., 1996).

Volatile nitrosamines were detected in raw and various food process Estonian meat samples using two-step solid-phase extraction with separation by GC and detection by positive-ion chemical ionization using ammonia as a reagent gas. Limit of detection and limit of quantification of nitrosamines were 0.09 and 0.29 µg/kg, respectively, with about 85 % recovery. The estimated total volatile nitrosamines were 3.97 µg/kg (Yurchenko and Mölder, 2005).

Ventanas and Ruiz (2006) successfully determined nine nitrosamines in gelatine based food model system using SPME coupled to a direct extraction device and subsequent GC-MS in selected ion monitoring mode. In addition, all nitrosamines were extracted without any sample manipulation making it a rapid and non-destructive method for preliminary screening of these carcinogens in solid foods.

Meat contains natural source of precursors for nitrosamine production. For example, the amino acid glycine is a precursor of N-nitrosodimethylamine, which can be derived from amino salts choline, acetylocholine and betaine, which are common in meat (Rywotycki, 2003). In addition, in meat free amino acids praline, glycine, alanine valine and biologically active ones such as putrescine and cadaverine are susceptible to nitrosation (Rywotycki, 2007). Thus it is important to know the initial nitrosamine

118 concentrations in fresh meat before processing. Rywotycki (2003) had shown that animal species, breeding factors and season of the year affected the nitrosamine contents in raw meat. This is most likely due to the variations in nitrate and nitrite content in the feed and water, as well as other environmental factors. It was also demonstrated that in winter the nitrosamine level is at its lowest compared to the highest nitrosamine content occurs in autumn. In addition, the overall NDMA concentration is highest in pork and beef and lowest in ram, goat and veal.

Andrade et al. (2005) developed a method that is simple, rapid with adequate accuracy, selectivity, sensitivity and precision for the determination of N-nitrosamines including N-nitrosodimethylamine, N-nitrosodiethylamine, N-nitrosopiperidine and N- nitrosopyrrolidine in sausages using headspace sampling by solid-phase microextraction and gas chromatography with thermal energy analyzer detection (HS-SPME- GC-TEA). The clean-up and extraction method used is less time-consuming and less labor intensive compared to traditional extraction methods such as low-temperature vacuum distillation, supercritical fluid extraction and solid-phase extraction. In addition, it eliminated the use of potentially toxic and expensive in the traditional solvent-bases extraction methods. Although GC-TEA is a common detection method for volatile nitrosamines, HPLC-TEA can also be used for the analysis of volatile nitrosamines with similar retention time and peak height as the traditional GC-TEA method (Khalaf and Steinert, 2000).

The rate of nitrosamine formation is proportional to the square of the nitrite concentration; therefore reduction of nitrite has a profound effect in reducing the amount of nitrosamines formed in foods. There is a steady decline in the use of nitrite in foods worldwide either by voluntary actions of manufacturers or by government regulations (Lijinsky, 1999).

Food processing technology could be useful besides function in food production and preservation. For example, Jo et al. (2003) demonstrated that residual nitrite content in cooked pork sausage was lowest with carbon dioxide packaging compared to aerobic and vacuum packaging. In addition, irradiation at 5 kGy significantly decreased volatile N-nitrosamines in that product. Similarly, high dose irradiation at more than 10 kGy was shown to reduce nitrosamines content including NDMA and NPYR in fermented sausage during storage. In addition, vacuum packaging gave significant nitrosamine

119 reduction than aerobic packaging (Byun et al., 2004). The reduction of nitrite to nitrogen oxides may explain the higher levels of nitrosamines in aerobic packaging since nitrogen oxides reacts with amines to form nitrosamines (Yurchenko and Mölder, 2006).

Rywotycki (2007) demonstrated that functional additives including sodium chloride and sodium ascorbate caused a decrease in NDMA formation in raw meat. In addition, baking process increased the level of NDMA in meat without the additives compared to meat containing the additives.

Fish especially marine fish contains abundant source of precursors of NDMA including dimethylamine, diethylamine, trimethylamine and trimethylamine oxide. The total concentrations of five nitrosamines (NDMA, NDEA, NPYR, NPIP, and NDBA) in fish products ranged from non-detectable to 10.34 µg/kg. In addition, baking temperature and time was shown to affect the nitrosamine content: increasing temperature and cooking time increased the nitrosamines content of the fish products (Yurchenko and Mölder, 2006).

Miller et al. (1989) demonstrated that cooking method affected the content of N- nitrosamines in bacon. It was shown that fried bacon fat contained up to 20.8 and 23.3 ng/g of NPYR and NDMA, respectively. In contrast, no NPYR and NDMA were detected in bacon or its fat from microwave cooking for 45 sec per slice. Therefore cooking method drastically affected the level of exposure to N-nitrosamines in humans.

The elevation in NPRO excretion in urine after the consumption of nitrite cured meat was not diminished by the concurrent consumption of the nitrosation inhibitor ascorbic acid suggests the presence of pre-formed NPRO or NPRO pre-cursor in the meat. Furthermore, the amount of NPRO in cured meat increased significantly after digestion with proteolytic enzymes, accounting for 3 to 93 % of protein-associated N- nitrosamines and only detectable after proteolytic digestion (Dunn and Stich, 1984).

The vigorous proteolysis of meat in vitro may release greater amount of N-nitrosamines then are released by natural digestive processes in the body. Thus the best way to measure the amount of biologically available N-nitrosamine in foods is to measure what is excreted than subjecting them to any laboratory treatments (Dunn and Stich, 1984).

120 Rywotycki (2002) concluded that meat curing with polyphosphates and sodium nitrite increased the level of short chain nitrosamines such as dimethylonitrosamine (DMNA) and diethylonitrosamine (DENA). However, the content of these nitrosamines can be reduced by the addition of sodium ascorbate.

NDMA is the most common volatile N-nitrosamine. It is found in beer and other beverages based on malted barley. In contrast, other grain- or grape- derived alcoholic beverages only contain trace amount of NDMA (McWeeny, 1983).

Many aliphatic and nitrogen-containing heterocyclic compounds can be nitrosated to form carcinogenic substances. Nitrite in food is the most important precursor and nitrosating agent, as well as nitrate, which can be reduced to nitrite in vivo. The crucial nitrosating agent for secondary amines is another nitroso compound and not nitrous acid or other oxides of nitrogen. For example, S-Nitrosocysteine can transnitrosate secondary amines at pH up to 9.75 therefore may contribute to preformed NOC in foods that are not acidic (Lin, 1990).

Shahidi and Pegg (1994) demonstrated that nitrite-free curing of meat and fish products containing preformed cooked cured-meat pigment did not contain volatile nitrosamines. This is not surprising since nitrite is a strong nitrosating agent normally added to cured meat products. Although they have eliminated problems associated with color and nitrosamine formation, they did not address the potential growth of C. botulinum spores that is normally controlled with nitrite salts.

It was shown that certain nitrosamines are formed from precursors during cooking of salt-preserved fish. Furthermore, different cooking practices such as steaming and frying may result in the presence of nitrosamines in the aqueous phase or oil in the case of frying. It was also suggested that most of the nitrosamines content in the frying oil maybe lost to the atmosphere during frying (Huang et al., 1981).

3.7.2 Non-volatile nitrosamines

Non-volatile nitrosamines such as N-nitrosoproline and N-nitrosodiethanolamine (NDELA) are more polar and have higher molecular weight than volatile nitrosamine, which means they possess relatively low vapor pressure, thus generally cannot be removed from a food matrix by distillation technique. Instead, nonvolatile nitrosamines can either be converted to volatile nitrosamine followed by separation and detection

121 GC-TEA, or it can be separated by HPLC followed by detection by HPLC-TEA. Furthermore, the total NOCs contained in the food can be detected by converting them to nitric oxide followed by separation and detection using GC-TEA. Thus the size of nitric oxide peak should be proportional to the total N-nitroso content in the food (Scanlan and Reyes, 1985). Ologhobo et al. (1996) used ultraviolet light to split nitrosamines to nitrite and amines. The nitrite was then determined using the method proposed by Montgomery and Dymock (1961), in Ologhobo et al. (1996).

The nitrosation-TEA system is based on selective chemical denitrosation of the compounds to nitric oxide followed by detection of the liberated NO by TEA. This system can be coupled to HPLC for the analysis of a variety of polar and non-polar non-volatile nitrosamides that are particularly unstable for determination by other methods (Sen et al., 1985). Sen et al. (1985) had used this system and confirmed its sensitivity for the determination of N-nitroso-N-methylurea (NMU), a nitrosamide in fried bacon.

Liquid-liquid partition is used as the initial isolation step in the determination of non- volatile nitrosamines by extracting volatile nitrosamines in the samples. Subsequently, non-volatile nitrosamines can be extracted by using vacuum distillation or column extraction. Although the latter method offers numerous advantages over the traditional distillation method, its use in determining nitrosamines in foods and beverages must be carefully validated (Issenberg, 1981).

3.7.3 Non-volatile nitrosamides

The distribution of NOC in food lacks information on non-volatile N-nitroso compounds including alkylnitrosamides (alkylnitrosoureas) and hydroxylated nitrosamines (nitrosodiethanolamine and nitrosobis-2-hydroxypropylamine). These are highly potent NOC with both mutagenic and carcinogenic activities, and can only be determined by high-performance liquid chromatography. However, accurate determination of alkylnitrosamide is challenging since they are unstable and difficult to identify at low concentrations (Lijinsky, 1999).

Massey et al. (1982) described a simple clean-up procedure for the routine analysis of N-nitrosamines in cured meats and beverages using reverse phase microbore HPLC with an organic ion-pairing agent and a post-column acetone make-up solvent prior to a

122 TEA detector. This allowed the determination of polar N-nitrosamines in cured meats and beverages that are traditional analyzed using GC-TEA.

3.7.4 Nitrosamine extraction from food

SPME utilizes a fiber support coated in a thin layer of diverse range of polymer material to extract and concentrate analytes for analysis. Application of heat or the addition of salt can increase sorption onto the SPME fiber by increasing the volatility and solubility of the analyte, respectivly. Major advantages includes: no solvents required thus reducing costs and generating less waste, extraction and concentration occur simultaneously, and SPME can be automated thus further reducing extraction time (Grebel et al., 2006).

The optimum headspace to total volume ratio is 0.6 for most nitrosamines including NDMA, which corresponds to 7 ml of sample in a 17 ml vial including HS. The optimal extraction temperature for NDMA is 65 oC with optimal salt saturation at 100 % corresponds to 2.4 g NaCl for a 7 ml sample volume (Grebel et al., 2006). Although NDMA requires about 4 h to reach equilibrium, it was suggested that 45min was close to the point of equilibrium without compromising extraction efficiency. These conditions resulted in a detection limit of 30 ng/L and making it an inexpensive and simple method for the routine analysis of NDMA using HS-SPME method. Considerable reduction in extraction and analysis time is around 1.25 h compared to other methods taking 3-20 h (Grebel et al., 2006).

By using solid-phase microextraction (SPME) coupled to a direct extract device (DED), Ventanas et al. (2006) managed to extract standard mixture containing nine volatile nitrosamines (NDMA, NMEA, NDEA, NPYR, NMOR, NDPA, NPIP, NDBA, and NDPheA) at 1 to 5 ng/ml from a solid food model system composed of gelatine at refrigeration (4 oC) and room temperatures (25 oC) with different extraction time analysed by GC-MS in selected ion monitoring mode. It was concluded that this method is rapid and suitable for extracting nitrosamines from model food systems.

Solid-phase microextraction is a relatively new, sensitive and convenient device for extracting compounds from samples. There are two techniques for the use of SPME: headspace (HS)-SPME and direct immersion (DI)-SPME. In HS-SPME, the fiber is exposed to the vapor phase above a gaseous, liquid or solid sample, whereas in DI-

123 SPME, the fiber is immersed in liquid samples. The choice of technique depends on the analyte. For example, for volatile nitrosamines, the HS-SPME is most suitable and since the fiber is not in direct contact of the sample its shelf life is extended. Following extraction, desorption is achieved by exposing the fiber in the injection port of GC or GC-MS to heat or by running solvent in the SPME-HPLC interface (Kataoka et al., 2000).

One major advantage of SPME is that full equilibrium is not necessary for quantitative analysis since there is a proportional linear relationship between the amount of analyte adsorbed by the SPME fiber and its initial concentration in the sample matrix before reaching partition equilibrium. This means time spent on sample clean up may be avoided or at least minimized (Kataoka et al., 2000).

Mixed coating fibers such as carboxen divinylbenzene (CAR-DVB) is useful for the extraction of low molecular weight volatile and polar analytes such as NDMA. Optimization of extraction using SPME depends on the concentration of analyte extracted, the polarity and thickness of the stationary phase, the extraction time, pH, temperature and amount of agitation. (Kataoka et al., 2000).

3.7.5 Nitrosamines in biological fluids

Traces of NDMA, although at low levels of 0.1 µg/kg, is only found in six of 23 urine samples from subjects ingesting daily large doses of ammonium nitrate (Ellen et al., 1982). There are numerous flaws in their experimental design, specifically the lack of controls. The compositions of the two most recent meals were recorded, but the meals were not controlled so that the amount of nitrosamines precursors varied from subject to subject. In addition, all subjects were patients at a hospital therefore might not have a normal metabolism compared to healthy individuals.

Lakritz et al. (1982) measured nitrosamine levels in human blood, urine and gastric content after feeding 21 healthy volunteers different types of meals including fish, beef or bacon in combination with spinach and vegetable juice. They concluded that gastric formation of nitrosamine did not seem to be a health risk for healthy people since the levels of nitrosamine found in physiological fluids did not markedly increased after eating.

124 Salivary nitrite concentration showed no correlation with blood nitrosamine levels. However, after feeding 23 non-smoking volunteers food including bacon, spinach, bread and beer, the concentration of NDMA increased in the blood with a mean of 0.5 µg/L. It was also demonstrated that the highest level of NDMA in the blood occurred after 1hr eating the meal after which the concentration started to decline (Gough et al., 1983).

Vermeer et al. (1998) tested the effects of amine and nitrate rich diet in human volunteers to determine its impact on volatile N-nitrosamine production in vivo. Salivary nitrate and nitrite, urinary nitrate and urinary NDMA and NPIP increased significantly after the introduction of amine and nitrate rich diet for one week, whereas a significant reduction was shown one week after the diet stopped. Furthermore, it was demonstrated that nitrate and NDMA excretion were significantly correlated, as well as salivary nitrate and nitrite concentration and NDMA excretion. However, NPIP excretion was not directly related to the nitrate intake and composition of the diet. This finding is remarkable as it is perhaps the first human in vivo experiment that showed a correlation between diet and the formation of carcinogenic N-nitrosamines. In addition, they have demonstrated human volunteers that a daily dose of 250 mg ascorbic acid was sufficient to inhibit NDMA excretion to the maximum extent, and that two grams (four cups) of green tea a day also decreased NDMA excretion. However, at more than four grams (eight cups) of green tea a day increased the formation of NDMA via a yet unknown mechanism (Vermeer et al., 1999).

Al-Mamary et al. (2006) investigated nitrosamine formation in aqueous and simulated normal fasting stomach conditions (37 oC at pH 2.0 for 1 h) with constant concentration of nitrite (14.4 mM) with Catha edulis leave extracts as a source of primary (cathinone, cathine, and norephedrine) and secondary amines (ephedrine and pseudoephedrine). It was demonstrated that various levels of nitrite yielded a dose-dependent amount of total nitrosamine compounds. In addition, these were not detected in aqueous solution ≤ 0.5 mM sodium nitrite or ≤ 1.0 mM in stimulated gastric condition.

By using a novel 15N breath test on healthy volunteers and patients infected with H. pylori, Junghans et al. (1999) had demonstrated that molecular nitrogen is produced from the body from oxygenated nitrogen species including protonated form of nitrous + acid (H2ONO ), dinitrogentrioxide (N2O3), dinitrogentetroxide (N2O4) and

125 peroxynitrite (ONOO-). It was suggested that molecular nitrogen can be used as a non- invasive technique for assessing body’s NO and nitrite levels, the body’s nitrosation potential, and possibly oxidative stress status associated with particular diseases such as cancer, which highlights the importance of developing an accurate method for measuring or estimating the body’s NO level. However, only seven subjects participated in this study, therefore more numbers are required to support the validity of the method.

Helaleh and Korenaga (2000) reported 4.85 and 15.48 µg/ml of nitrite and nitrate, respectively, in human saliva based on ion chromatography with suppressed conductivity detection. In addition, they had demonstrated that the saliva sample only needed centrifugation to remove the insoluble matter and that no sample preparation or clean-up process was required.

Levallois et al., (2000) concluded that dietary nitrate intake and urinary nitrate excretion was not related to nitrosamine N-nitrosopiperidine excretion in a small-scale rural population in Québec, Canada. However, they showed a correlation between urinary nitrate excretion and total nitrate intake. Dietary nitrate and vitamins C and E intake were estimated by means of a 24-hr recall and by using a validated Canadian food database.

3.8 Determination of vitamin C

Vitamin C in food exists as two vitamers. Ascorbic acid (AA) can be oxidized to dehydroascorbic acid (DHAA) under conditions such as high temperature, presence of oxygen, degradative enzymes and/or metal ions especially copper and iron, alkaline pH and light. This process is reversible so DHAA can be reduced back to AA (Russell, 2000) in the presence of reducing agents such as hydrogen sulfide, homocysteine and dithiothreitol (Aurand et al., 1987). However, DHAA can be further oxidized to 2, 3- diketogulonic acid that is biological inactive and is an irreversible process. In addition, DHAA can undergo spontaneous hydrolysis to form 2, 3-diketogulonic acid (Washko et al., 1992). AA and DHAA have equivalent biological activities, whereas D-ascorbic acid including isoascorbic acid (IAA) and erythorbic acid have a biological activity about 5 % of L-ascorbic acid (Russell, 2000), and is usually added to food as an antioxidant since it does not occur naturally in food products (van Niekerk, 1982). D- ascorbic acid can also be oxidized to give dehydroisoascorbic acid (DHIAA). Both

126 IAA and DHIAA can interfere with the determination of AA and DHAA (Russell, 2000).

Total ascorbic acid is defined as the sum of both ascorbic acid and its oxidized form dehydroascorbic acid (DHAA). Although DHAA is present naturally in food, its instability makes quantification difficult. Thus its quantification is usually performed after its conversion into AA after the addition of reducing agents. Reducing agents such as cystein, homocystein and dithiothreitol (DTT) are often used to convert DHA into its reduced form and to stabilize AA. However, TCEP was demonstrated to offer better reduction efficiency at low pH compared to DTT. Isoascorbic acid (IAA) also called D- ascorbic acid or erythorbic acid is often used as an antioxidant in food but with only 5 % of the antiscorbutic effect of AA. The traditional AOAC method for determining vitaminc C only estimate the total content without distinguishing between AA and IAA. In recent years reverse phase HPLC with either ion exchange column or ion- pairing agent is used to distinguish between AA and DHA. However, this method required a pre- or post derivatization step, which is time consuming and may result in the degradation of AA in the process (Fontannaz et al., 2006).

Kall and Andersen (1999) used a dual detection system using HPLC for separation with direct detection of ascorbic acid and indirect fluorimetric detection of dehydroascorbic acid after a post-column O-phenyldiamine derivatisation. Total analysis time was only 10 minutes with good reproducibility and recovery. Since dehydroascorbic acid has a weak UV absorption and no response to electrochemical detection, it needs to be derivatised prior to or after chromatographic separation to increase its detection sensitivity. DHAA can be reduced to AA by L-cysteine or dithiothreitol or derivatised with O-phenyldiamine to form a fluorophore.

Metaphosphoric acid may provide the most efficient ascorbic acid extraction by preventing oxidation compared to citric acid, acetic acid, perchloric acid and orthophosphoric acid. However, compared to other acids mentioned above, metaphosphoric acid may cause analytical interferences by reacting with silica-based column materials such as C18 or NH2 bonded-phases thus affecting baseline and retention time. It was suggested that a combination of 1 % (w/v) metaphosphoric acid and 0.5 % (w/v) oxalic acid at pH 2 as extraction buffer provides high stability of AA

127 and DHAA and have minimal interactions with the chromatographic system (Kall and Andersen, 1999).

Furusawa (2001) developed a rapid and simple HPLC-UV method for the identification and quantification of L-ascorbic acid in fruit drinks. DHAA was reduced by dithiothreitol before HPLC analysis using 2 % (v/v) acetic acid as the mobile phase. The limit of detection was 10 µg/ml with analysis time approximately three minutes.

A new micro-calorimetric method based on the oxidation of vitamin C using the enzyme ascorbate oxidase was comparable to the HPLC-UV method in terms of sensitivity, specificity and short analysis time of less than five minutes. But the major advantage of the micro-calorimetric method over the HPLC method is that no pretreatment of the food or pharmaceutical samples is required which significantly reduces overall analysis time (Antonelli et al., 2002).

It was demonstrated that HPLC-UV (243 nm) offered high accuracy, good repeatability and reproducibility in relatively short analysis time as well as good separation of AA and IAA. Recovery was more than 99 % with LOD of 0.1 µg/ml, which corresponds to 0.1 mg/100 g samples (Fontannaz et al., 2006).

3.8.1 Food

Due to the increase practice of adding vitamins to food as a marketing tool, a more reliable and standardized method is required for their accurate determination in food samples. Because vitamin C is a non-volatile and hydrophilic compound, reversed- phase HPLC analysis has been very popular choice for its determination in foods. HPLC affords great advantage over other analytical systems for its ability to separate the vitamins from interferences inherent in complex matrices such as biological fluids and food. In addition, HPLC is much faster compared to spectrophotometic methods or microbiological assays with increased precision, reasonable accuracy, increased specificity by using UV detection and can be easily automated (Russell, 2000).

Because AA is heat sensitive, HPLC is routinely used for its determination and quantification in food samples since minimal heat is generated by using HPLC compared to GC-MS. However, recently GC was shown to be accurate in AA determination by Silva (2005) after simple clean up techniques and a freeze-drying step. However, GC gave slightly less recovery (2 %) and slightly longer analysis time

128 of seven minutes instead of approximately four minutes when compared to HPLC. Furthermore, GC does not distinguish between ascorbic acid and isoascorbic acid.

Based on spectrophotometric, conductometric and traditional titrimetry methods, Grundpan et al. (1999) demonstrated that commercially available vitamin C tablets from Thailand were between 73 and 107 % of the labeled vitamin C content with an average of 96 % detected from the labeled content. Commercial vitamin C tablets in the United States were shown to be between 92 and 111 % of the labeled content based on enzymatic spectrophotometric determinations (Esteban and Ho, 1997).

Nisperos-Carriedo et al. (1992) simultaneously determined dehydroascorbic acid, ascorbic acid and other organic acids in fruits and vegetables by reversed-phase HPLC with UV detection at 215 and 260 nm using a mobile phase consisted of 2 % potassium dihydrogen phosphate (KH2PO4) at 0.4 ml/min. Although ascorbic acid has a maximum absorbance at 245 nm, 260 nm was used to avoid interferences from the food matrix. However, since no food is used for this experiment, interferences are expected to be low and hence the maximal absorbance will be used for increased sensitivity and selectivity.

Similarly, Zapata and Dufour (1992) simultaneously determined ascorbic, dehydroascorbic and isoascorbic acid in fruit juices, beer and vitamin C tablets using reversed-phase ion interaction HPLC with UV detection at 261 nm. Detection was under 11 minutes with limit of detection of 10 ppm. The peak separation and resolution was much better compared to Nisperos-Carriedo et al. (1992).

Total ascorbic acid and isoascorbic acid can be quantified based on acidic extraction of ascorbic acid in the presence of reducing agent Tris [2-carboxyethyl/ phosphate] (TCEP) that maintains ascorbic acid in its reduced form. The separation was achieved using HPLC with C18 column and sodium acetate as mobile phase containing TCEP and decylamine as ion pairing agent. (Fontannaz et al., 2006).

Odriozola-Serrano et al. (2007) compared different HPLC-UV methods and reducing agents for the routine determination of vitamin C in fruits. It was demonstrated that C18 column was more reliable than the NH2 column, and that DL-1,4-dithiotretol (DTT) was a better reducing agent than 2,3-dimercapto-1-propanol.

129 Tai and Gohda (2007) developed a new, simple and sensitive hydrophilic interaction liquid chromatography method for the simultaneous determination of ascorbic acid and its related compounds in foods and beverages with commercial applications. Dithiothreitol (50 mg/L) was used as a stabilizer resulted in good linearity and reproducibility. Detection with UV was set at 260 nm with retention time for ascorbic acid at approximately 12 minutes.

A new method involving radical oxidation of L-ascorbic acid to dehydro-L-ascorbic acid with HPLC detection by fluorometric on a Nova-Pak C18 column was demonstrated to have good reproducibility, sensitivity and accuracy (Burini, 2007).

3.8.2 Biological fluids

A comprehensive review for vitamin C analyses in biological samples were summarized by Washko et al. (1992). Most of the colorimetric and chromatography methods reviewed had limitations involving sensitivity, specificity, matrix interference and stability. In addition, methods involving spectophotometric assays were time consuming, as well as GC that requires lengthy derivatization process. Overall, the HPLC coupled with a UV detector seemed to have more advantages over other methods in reference to the limitations mentioned above. In addition, HPLC-UV has the ability to measure both AA and DHAA using the same system set up either directly using an amine column or indirectly through derivatization or reduction of the vitamers.

In summary, vitamin C analysis suffers from the four ‘S’ syndromes: stability, sensitivity, specificity and substance interference (Washko et al., 1992). Therefore, an accurate determination of vitamin C requires an extraction method that minimizes losses due to the factors mentioned above, as well as being sensitive and selective with short analysis time for large sample runs.

3.8.3 Extraction

Extraction method used to extract vitamins from food is a crucial step, since the extraction method itself may contribute to losses or incomplete extraction of the desired vitamin as they are commonly bound to food components such as carbohydrates and proteins. In addition, due to the trace quantity of naturally occurring vitamins in foods, the detection sensitivity needs to be sufficient to detect minute amounts (Russell,

130 2000). UV detector set at 254 or 264 nm is selective for vitamin C thus increases detection sensitivity and accuracy. However, electrochemical detector offers greater sensitivity and selectivity for ascorbic acid than UV detector, especially for samples with very low vitamin C content (van Niekerk, 1982).

The aim of the extraction method for vitamins analysis is to prevent the degradation and loss of vitamers in order to ensure an accurate determination. Acids such as metaphosphoric acid, trichloroacetic acid and oxalic acid can protect vitamin C vitamers from oxidation and hydrolysis of the lactone ring, at the same precipitating proteins often associated with endogenous vitamin C in foods (Washko et al., 1992). Extraction solutions containing 3 to 6 % of oxalic or metaphosphoric acid are commonly used to extract vitamin C from foods. These acids can prevent catalysis of the oxidation of ascorbic acid by cupric or ferric ions, however, metaphosphoric acid is more useful for its ability to precipitate proteins and deactivate ascorbic acid oxidase. The sample is usually homogenized with the extraction solution followed by centrifugation or filter before chromatography analysis (van Niekerk, 1982).

Metal chelators such as EDTA and diethylenetriaminepentaacetic acid prevents oxidation of AA by binding to metal cations such as iron and copper. By extracting at low temperature and minimizing exposure to white light using amber vials can minimize the oxidation of AA. In addition, purging the mobile phase used in HPLC with inert gases such as helium can minimize oxidation by removing dissolved oxygen in the solution (Russell, 2000). Vitamin C content in foods can decrease considerably during storage and processing (Aurand et al., 1987), so storing of samples at -70 oC is recommended if it is not been analyzed immediately (Washko et al., 1992).

3.8.4 Detection

There are many non-spectrophotometric methods available for the determination of vitamin C in various samples including titrimetry, voltammetry, fluorometry, potentiometry, kinetic-based chemiluminescence, flow-injection analyses and chromatography. Except methods based on chromatography, other non- spectrophotometric methods have numerous disadvantages including the lack of sensitivity and/or selectivity, time of analysis can be long and often requires chemical derivatization. In addition, they can suffer from matrix interferences and have limited sample range (Arya et al., 2000).

131 The two most common detectors used with HPLC for vitamin C analysis are UV and electrochemical detectors. UV detector for AA detection has maximum absorbance at 245 to 265 nm that is pH dependent (Table 3.1). For the detection of DHAA the maximum absorbance is at 210 to 230 nm, but is susceptible to interferences by naturally occurring food components, and has limited choice of reagents and solvents. Electrochemical detector and be used to determine AA due to its reducing capacity, but since DHAA is electrochemically inactive; it is therefore not used for the determination of the latter. Although fluorescence detector can be used to determine AA and DHAA, they need to be derivatized with o-phenylenediamine in order to produce a fluorescent quinoxaline complex. However, chemical dramatization is often required to achieve the sensitivity needed for detecting minute amounts of vitamin C present naturally in food (Russell, 2000). By reducing DHAA to AA with homocysteine at pH 6.8, the total vitamin C content of a sample can be determined (van Niekerk, 1982).

Anion-exchange columns with pellicular or microparticulate packing can be used to determine ascorbic acid, but reversed phase columns are often preferred for its better reproducibility and robustness (van Niekerk, 1982). The major drawback with using reversed-phase column in HPLC is that AA is not readily retained therefore requires the addition of a cationic ion pairing reagent such as tetrabutylammounium phosphate (Arya et al., 2000), which can be very costly. However, the right ion-pairing reagents can minimize food matrix interferences by altering the retention time of ascorbic acid and other components in the sample (van Niekerk, 1982).

Chromatography conditions are important for the accurate determination of ascorbic acid. For example, metal ions especially iron and copper can catalyse the oxidation of ascorbic acid during chromatography therefore should be removed from the food sample and mobile phase. Proper and continuous degassing of the mobile phase during analysis is vital to minimize oxidation by oxygen. In addition, the mobile phase pH should be adjusted to pH 5 or lower since vitamin C is only stable in acid solutions (van Niekerk, 1982).

132 3.9 Effects of vitamin C supplementation on N-nitrosodimethylamine formation in healthy human volunteers on a nitrate restricted diet with cured meat

3.9.1 Introduction

Since the discovery in 1956 that NOCs are carcinogenic to more than 40 animal species, and human exposure to NOCs has been related to increased risk of gastrointestinal cancers. The endogenous nitrosation increases when precursors are ingested in foods, which commonly contain these precursors. Under acidic conditions such as the stomach, nitrite can react with secondary amines to form N-nitrosamines. Ascorbic acid and ascorbate can inhibit nitrosation because they react faster than the amine with the nitrosating agents. AA can reduce nitrous acid to nitric oxide that is not a nitrosating agent, and is itself oxidized to dehydroascorbic acid. Whole strawberries, garlic juice or kale juice decreased NDMA formation in humans fed an amine-rich diet with nitrate by 70, 71 and 44 %, respectively. Increased dietary nitrate intake also significantly increased salivary and urinary nitrate. It was suggested that phenolic compounds derived from plants and fruits are potent nitrosation inhibitors. In addition, strawberries contain the highest vitamin C amongst fruits at 99 mg/100g. Cysteine and other sulfhydryl compounds and allyl sulfur compounds were thought to act as nitrite scavengers (Chung et al., 2002).

NOCs are carcinogens almost exclusively found in foods and beverages containing nitrite such as cured meat or those that have been exposed to nitrogen oxides, for example beer. Nitrosamines are the most commonly occurring NOC encountered and on the whole is shown to induce tumors in a variety of organs including liver, lung, kidney, bladder pancreas, esophagus and tongue in animal experiments (Lijinsky, 1999).

The optimal pH for nitrosamine formation is at 3.4, which is the pKa for nitrous acid. Hence the stomach is the most vulnerable organ that is exposed to nitrosamines for the longest amount of time. However, other factors have been identified that interfere with this reaction. For example, not all nitrosatable substances follow the same kinetic pattern and that nitrosation is a multi-step process. A number of chemical compounds can also inhibit, slow down or accelerate the reaction (Hansson et al. 1994). For example, ascorbic acid can prevent nitrosation by converting nitrite precursors to NO thus preventing or at least reducing the formation of NOCs including nitrosamine

133 (Hwang et al., 1994). In addition, foods and stomach contents are non-homogenous thus partitioning of reactants between phases may pose further changes in the rate of nitrosamine formation (Wogan and Tannenbaum, 1975).

The aim of this study was to determine the effects of 500 mg vitamin C supplement on NDMA formation in vivo in subjects on a low nitrate and a cured meat diet.

3.9.2 Materials and methods

A modified HPLC method was used for the determination of ascorbic acid nitrate and nitrite in human saliva and urine; a modified GC-MS method was used for the determination of NDMA in cured meat and human urine based on Ventanas and Ruiz, 2006 and Grebel et al. 2006.

3.9.2.1 Vitamin C Analysis

Extraction from samples was based on dilution with milliQ water and then purification using Sep-Pak C18 cartridge. All samples food, vitamin C tablets and urine were extracted and purified using the same methods and then analyzed for ascorbic acid using HPLC-UV.

3.9.2.3 Form of Vitamin C used in the experiment

Ten vitamin C tablets claimed to contain 500 mg equivalent of ascorbic acid were crushed with a pestle and mortar and homogenized into powder form as a composite sample. Five replicates were performed by diluting 0.25 g into 100 mL distilled water followed by filtration using Sep-Pak C18 cartridge. Injection volume of 7.5 µL was then injected into Waters HPLC-UV using autosampler and reversed-phase Gemini C18 column. Flow rate was set at 1.6 mL/min with mobile phase containing 0.22 % (w/v)

KH2PO4 and 0.68 % (w/v) cetrimide (hexadecyltrimethylammonium bromide) in methanol and milliQ water (10:90). UV absorbance at 261 nm was used for detection. Calibration curve using ascorbic acid standard diluted with milliQ water to give 0.1 to 0.5 % (w/v).

134 3.9.2.4 Vitamin C in urine

One mL of urine sample in duplicate was diluted to 10 mL with MilliQ water in a volumetric flask. Samples were then centrifuged and filtered through Sep-Pak C18 cartridge before injection into Waters HPLC-UV as for vitamin C tablets analysis.

3.9.2.5 Nitrate and nitrite analysis

Sample extraction was based on hot water extraction followed by centrifugation and ultra-filtration. Filtrate was then injected into Waters HPLC-UV for quantifying nitrate and nitrite in meat, human saliva and urine as was developed in this study based on Sui et al. (1983) (see Chapter 2).

3.9.2.6 Analysis of meat

Two packets of salami, ham and hotdogs at 100 g per packet were purchased from local supermarkets (two different brands) and a composite sample of each type of meat was prepared by homogenizing in a commercial blender. 25 g of each type of meat were taken in duplicates and re-homogenized in 300 mL distilled water for one minute and then made up to 500 mL in a volumetric flask.

Ten mL of the mixture of each sample was transferred into 100 mL volumetric flasks and heated in a water bath at 75 oC for 15 min. On cooling the mixture was centrifuged at 10,000 rpm for 10 min; and then the supernatant was removed for ultra-filtration. Duplicate analysis was performed. The filtrate was also used for recovery studies by adding known quantities of nitrate and nitrite standards.

3.9.2.7 Analysis of saliva

Saliva samples were collected in 50 ml specimen jars containing 50 mg solid NaOH. They were stored in the freezer at -20 oC until analysis. One ml of saliva sample in duplicates was diluted with distilled water to 10 ml in a volumetric flask and was then filtered through conditioned Sep-Pak C18 syringe with duplicates. Similarly recovery was performed by adding nitrate and nitrite standards to give 50 mg NO2 and 25 mg

NO3 ions per 10 ml. Nitrate water for nitrate conversion contained 350 mg of nitrate ions.

135 3.9.2.8 Analysis of urine

Urine samples were collected in 1 L urine bottles containing 300 mg solid NaOH. They were stored in the freezer until analysis. One ml of urine was diluted in 10 ml vol. flask then ultra-filtered using Sep-Pak C18 cartridge with duplicates. Similarly recovery study was performed by adding nitrate and nitrite standards to give 100 mg NO2 and 50 mg

NO3 ions per 10 ml.

3.9.2.9 Quantification of N-nitrosodimethylamine in meat and urine

Extraction of NDMA from meat and urine were based on headspace solid phase microextraction method followed by quantification and confirmation using Agilant GC-MS (5975 inert mass selective detector coupled with 7683B series injector and processed using 6890N network GC system (Ventanas and Ruiz, 2006; Grebel et al., 2006).

3.9.2.10 Meat

Samples were weighed and 150 g of each salami, ham and hotdog were homogenized in a commercial blender for 3 min. Then 1.0 g of the sample was added along with 2.4 g of NaCl to 15 mL headspace vial and distilled water was added to make-up volume to 7 ml then sealed with Teflon-silicone septa. The solution was magnetically stirred and was placed on a hot plate at 65 oC for 10 min. The SPME-CAR/PDMS fibre was then inserted into the vapour phase above the liquid for 15 min and was then withdrawn into the needle, which was subsequently introduced into the injection port of the GC-MS. Desorption of the fibre coating was performed at 200 oC for 8 min.

3.9.2.11 Urine

Similar extraction method as in meat samples except 1 mL of urine sample was used in the HS-SPME.

3.9.2.12 Human trial

This experiment was granted ethics approval by the University of New South Wales, Human Research Ethics Committee (HREC) and the project number was 07129.

136 3.9.2.13 Recruitment of Volunteers:

Volunteers were recruited via posters placed on campus at the University of New South Wales. Total of eighteen healthy volunteers were recruited. Subject consent to participate in the study was obtained.

3.9.2.14 Experimental Protocol:

Day 1: Restricted diet (Figure 3.8): This diet consisted of no source of Vitamin C, nitrate and nitrite. The background information and participant consent form (Appendix A), details of the diet (Appendix B) and questionnaire form (Apendix C) are provided in details in the appendices.

Day 2 and 3: A restricted diet similar to Day 1 was followed.

137 Human Trials:

Nitrate reduction in the oral cavity; and the effects of Vitamin C on NDMA formation in vivo

DAY 1 DAY 2 DAY 3

Take home On Waking provide: On Waking:

3 x urine bottles 3 x saliva specimen jars 3 x meat samples 1 x urine sample (1) 1. Discard urine 1 x nitrate water 1 x saliva sample (A) 2. Brush teeth 1 x vitamin C tablet 1 x chewing gum

Avoid foods Tasks (morning): Tasks (morning): (Appendix A):

1. Nitrate water rinse,

provide saliva (B) 1. Eat meat (B2) Vitamin C rich e.g. 2. Chew meat (A), provide 2. Take one (1) vitamin C vegetables and fruits. saliva (C) tablet provided Also avoid all cured 3. Eat meat (B1) meat except ones given.

3 hours fasting: 3 hours fasting:

1 x urine sample (3)(or 1 x urine sample (2)(or collect urine up to 3 hr collect urine up to 3 hr from meat consumption) after meat consumption) Resume usual diet after Return samples to me providing urine sample.

Figure 3.8 Human trial study protocol. All foods were prepared fresh and stored in refrigerator at 4 oC and all biological samples were stored in containers containing sodium hydroxide and stored in freezer at -20 oC until analysis.

138 3.10 Results and discussion

Firstly, this section discussed the conversion of nitrate to nitrite in the oral cavity and the effects of food matrix and gender on its conversion rate. Secondly, after a nitrate and antioxidant restricted diet, healthy human volunteers were given cured meat sandwich for determining NDMA formation in vivo from analyzing NDMA content extracted in the urine. Finally, the effect of vitamin C supplement on NDMA formation in vivo is discussed.

3.10.1 Results on nitrate conversion rate in the oral cavity

The mean nitrate conversion rate in the oral cavity of 18 healthy volunteers was presented in Figure 3.9. The mean nitrate conversion rate in males was 17 % compared to 24 % in females. The conversion was assumed to be due to nitrate-reducing bacteria residing in the mouth, and the result (Figure 3.9) suggested that more nitrate-reducing bacteria resided in the mouths of females than males. The differences may also be due to other factors that may affect nitrate-reducing bacteria population such as age, race, diet and daily variations.

139

Figure 3.9 The mean rate (%) of nitrate conversion to nitrite in the oral cavity of 18 healthy human volunteers. Water containing 350 mg of nitrate ions were rinsed in the mouth for 30 s and then collected in 50 mL specimen jar for quantifying both nitrate and nitrite ions using ion-paired RP-HPLC-UV. The mean nitrate conversion rate in the oral cavity was derived by averaging individual conversion rate for males and females. Based on one-tail t test, there was no significant difference (p ≥ 0.05) in the nitrate conversion rate between the sexes. However, the error bars showed that there was a larger range in male than in female

Table 3.2 Quantification parameters on the effects of saliva matrix on nitrate and nitrite recoveries.

Parameters LOD (mg/L) LOQ (mg/L) Recovery (%) Linearity (R2) Nitrate 1.0 2.5 96.6 ± 6.68 0.99995 Nitrite 1.0 5.0 100.9 ± 7.77 0.99950 Six replicates.

From Table 3.2. The recoveries for nitrate in nitrite in saliva were excellent with little interence in nitrate recovery resulting less than 100 % recovery when compared to nitrite.

140 Table 3.3 Individual Mean nitrate and nitrite contents in the saliva of healthy human volunteers after rinsing with 350 mg nitrate water as the control and chewing ten grams of meat as the experiment.

Volunteer Fastinga (µg/ml) Experimentalb (µg/ml) number Nitrate Nitrite Nitrate Nitrite 1 9.1 ± 2.3 2.1 ± 0.4 9.4 ± 2.1 2.6 ± 0.1 2 7.8 ± 1.8 5.4 ± 1.1 8.5 ± 1.2 5.5 ± 0.6 3 15.6 ± 3.2 11.2 ± 2.3 14.9 ± 3.1 13.2 ± 1.4 4 26.3 ± 5.4 6.8 ± 1.6 26.4 ± 1.3 7.1 ± 2.1 5 12.7 ± 1.9 7.3 ± 1.8 15.2 ± 1.9 7.6 ± 1.1 6 5.5 ± 0.7 12.8 ± 0.9 4.5 ± 0.1 13.4 ± 0.4 7 11.8 ± 1.1 2.9 ± 0.2 12.0 ± 0.7 2.7 ± 0.3 8 23.2 ± 2.1 5.8 ± 2.1 23.8 ± 1.2 6.5 ± 0.8 9 17.0 ± 3.5 3.6 ± 0.5 15.3 ± 1.6 4.3 ± 1.0 10 9.5 ± 1.6 4.0 ± 0.4 8.2 ± 0.6 4.5 ± 0.2 11 16.4 ± 2.2 3.9 ± 0.2 16.6 ± 1.3 3.4 ± 1.1 12 8.8 ± 1.7 15.5 ± 2.1 8.5 ± 1.1 15.3 ± 1.7 13 24.1 ± 3.9 7.2 ± 1.3 26.1 ± 3.2 7.5 ± 0.9 14 13.5 ± 4.2 3.9 ± 0.8 13.4 ± 1.0 4.2 ± 0.3 15 20.0 ± 3.3 8.8 ± 1.1 19.9 ± 2.2 9.8 ± 0.6 16 14.6 ± 1.8 10.4 ± 2.2 15.2 ± 2.4 10.8 ± 1.3 17 26.7 ± 6.9 21.3 ± 3.9 28.8 ± 0.9 20.9 ± 1.4 18 21.7 ± 5.4 16.9 ± 2.7 15.2 ± 0.3 12.5 ± 2.6

aFasting-after 24h restricted diet, saliva sample was collected first on the second day morning in a jar containing sodium hydroxide as a stabilizer after overnight fasting. bExperimental- chewed ten grams of meat (ham, salami or hotdog) for 30 secs. All samples were mean of triplicates.

141

Figure 3.10 Mean nitrate and nitrite content (mg/mL) in human saliva after fasting overnight and after chewing cured meat (salami, hotdog and ham). Volunteers fasted overnight and then gave saliva samples first thing in the morning. Before brushing their teeth they were asked to chew 15 g of cured meat for 30 sec. Samples were collected in 50 mL specimen jar for quantifying both nitrate and nitrite ions using ion-paired RP- HPLC-UV. Using student t test, there was no significant difference (p < 0.05) between fasting and chewed meat nitrate or between fasting and chewed meat nitrite.

The fasting nitrate and nitrite content in human saliva were very similar to that after chewing cured meat (Figure 3.10). The slightly higher salivary nitrite content compared to fasting nitrite (Figure 3.10) was expected since sodium nitrite was used in the cured meat products.

3.10.2 Discussion on nitrate conversion rate in the oral cavity

Human saliva contains less than 1 % solid and the main ions present are Na+, K+, Cl-, 2+ 2+ 3- - Ca , Mg , PO4 , and HCO3 (Dodds et al., 2005). Sodium and potassium ions may form ionic bond with ingested nitrate and nitrite, however, its determination and quantification should not be affected by using RP-HPLC-UV. However, there are more than 50 different proteins in the saliva (Smith and Morton, 2001) with up to 70 % of the proteins are proline based (Dodds et al., 2005). The remainder of the proteins consists of α-amylase and mucins amongst other enzymes in lesser amounts such as lysozyme, lactoferrin and various immunoglobulins. These proteins may react with ingested nitrate and nitrite and thus may interfere with its analyses. Sodium hydroxide pellets

142 were added to specimen jars in order to stabilize nitrate and nitrite during storage until further analysis as suggested by Vermeer et al. (1998).

Furthermore, the pH of the saliva can vary from 6.2 to 8.0 depending on the rate of flow. At rest the human saliva is slightly acidic whereas stimulated salivary secretion as a result of eating increases the alkalinity of the saliva (Smith and Morton, 2001). Since nitrite is unstable at acid pH, as in fasting, the nitrite content was expected to decrease compared to control because chewing can result in a more alkaline condition and thus stabilize nitrite (Table 3.3). On the other hand, vitamin C supplement given during the experiment also decreased nitrite content (Table 3.3) possibly due to the increased acidity in the mouth and caused nitrite to react to form other derivatives that were not analyzed such as nitric oxide.

The concentration of human salivary nitrite, which is dependent on the dietary intake of nitrate, varies from 0.05 to 1 mmol/L, Gastric nitrite concentrations are significantly lower than that of salivary nitrite because the former is converted to nitrous acid at acidic pH followed by the formation of nitric oxide and nitrogen dioxide that escapes into the gaseous phase (Dykhuizen et al., 1998).

Nitrate and nitrite concentrations in the saliva were not significantly related to those in the gastric juice as demonstrated by Dallinga et al. (1998). In addition, they concluded that there was no association with pH, nitrosamines content, age and sex of the subjects. However, they established a positive correlation between gastric nitrite contents and pH, whereby higher gastric pH resulted in higher gastric nitrite content due to enhanced bacterial growth and thus higher nitrate to nitrite conversion.

It was suggested that the formation of salivary nitrite in symbiosis with facultative anaerobic bacteria in the pharynx enhances the antimicrobial effects of gastric juice in humans by the formation of free radical nitrogen species that have bactericidal effects (Dykhuizen et al., 1998). For example, it was demonstrated in vitro that at pH 2 with 1 mM nitrite resulted in the complete kill of H. pylori within 30 minutes of exposure. Furthermore, the antimicrobial effect of nitrite at pH 2 against H. pylori was dose dependent and complete kill of the organism occurred at ≥ 500 µmol/L. It was concluded that the antimicrobial effect was not due to nitrite per se, but due to the generation of reactive oxides of nitrogen (Dykhuizen et al., 1998). The data presented was limited to in vitro conditions, thus it was suggested that the effect of salivary nitrite

143 on the survival of H. pylori in the human stomach should be examined, since H. pylori was a risk factor in developing stomach cancer (Hwang et al., 1994).

This may explain why vegetarians consuming more nitrates from vegetables than omnivores and yet are less likely to develop stomach cancer might be due to the conversion of ingested nitrate to salivary nitrite followed by nitric oxide formation in the stomach where it has antibacterial properties. Furthermore, Dykhuizen et al. (1998) suggested that swallowing saliva rich in nitrite after a meal high in nitrate might enhance the host defense against ingested pathogens.

3.10.3 Results for nitrate and nitrite excretion in urine

Nitrate, nitrite and NDMA concentrations in cured meat are assessed and presented in Figure 3.11. Individual (Table 3.5) and mean nitrate and nitrite contents in urine (Figures 3.12 and 3.13) are summarized below.

Figure 3.11 Mean nitrate, nitrite and NDMA contents (mg/kg) of popular cured meats. The error bars showed that there was a wider range of nitrate in salami samples compared to hot dog and ham. The lack of nitrite in salami showed that excess nitrate salts were added as reservoir for nitrite which may be formed during storage by bacterial nitrate reductase.

144 Table 3.4 Quantification parameters on the effects of urine matrix on nitrate and nitrite recoveries.

Parameters LOD (mg/L) LOQ (mg/L) Recovery (%) Linearity (R2) Nitrate 1.0 2.5 92.8 ± 9.69 0.99997 Nitrite 1.0 5.0 92.9 ± 4.04 0.99400 Triplicates.

From Table 3.4, urine matrix interfered with nitrate and nitrite recoveries but was still considered reasonable at almost 93 % for both anions.

Table 3.5 Individual nitrate and nitrite contents in the urine of healthy human volunteers before and after treatments.

Volunteer Fastinga (mg/l) Control (mg/l) Experimentc (mg/l) number Nitrate Nitrite Nitrate Nitrite Nitrate Nitrite 1 15.3 ± 2.1 19.0 ± 2.3 0.6 ± 0.2 2.6 ± 0.3 1.0 ± 0.2 11.0 ± 0.6 2 8.1 ± 0.6 7.5 ± 1.1 30.5 ± 3.6 8.2 ± 0.5 41.6 ± 5.3 7.1 ± 0.3 3 2.1 ± 0.3 9.7 ± 0.9 11.2 ± 1.2 10.2 ± 0.6 7.2 ± 1.2 0.4 ± 0.1 4 3.0 ± 0.5 11.3 ± 0.5 6.5 ± 0.7 9.3 ± 1.0 7.7 ± 1.1 14.4 ± 1.3 5 3.2 ± 0.4 13.5 ± 1.7 4.5 ± 0.4 3.8 ± 0.1 2.0 ± 0.1 13.7 ± 0.9 6 8.6 ± 1.6 3.4 ± 0.3 4.4 ± 0.4 6.0 ± 0.2 11.0 ± 0.6 26.3 ± 2.7 7 1.4 ± 0.2 8.2 ± 1.2 3.8 ± 0.6 12.9 ± 0.9 2.3 ± 0.4 19.3 ± 1.8 8 61.7 ± 7.3 25.8 ± 4.1 54.8 ± 1.9 36.3 ± 1.3 0.8 ± 0.1 1.4 ± 0.6 9 3.5 ± 1.1 15.7 ± 1.8 4.6 ± 0.8 14.1 ± 1.1 9.2 ± 1.3 28.6 ± 2.3 10 38.4 ± 3.1 21.2 ± 2.1 12.4 ± 1.7 6.9 ± 0.6 25.5 ± 2.3 11.4 ± 1.6 11 11.9 ± 0.9 13.4 ± 1.6 8.1 ± 0.6 14.1 ± 1.3 3.9 ± 0.2 0.9 ± 0.1 12 7.5 ± 0.8 4.9 ± 0.5 20.9 ± 1.6 13.7 ± 0.9 28.2 ± 1.5 19.3 ± 2.7 13 72.8 ± 6.4 31.1 ± 3.4 37.8 ± 3.8 9.2 ± 0.8 15.1 ± 2.1 21.8 ± 1.8 14 29.0 ± 2.1 18.2 ± 2.3 9.9 ± 1.2 14.2 ± 1.0 38.6 ± 1.9 19.2 ± 1.6 15 37.1 ± 3.9 21.7 ± 2.8 13.5 ± 2.1 19.4 ± 2.2 9.5 ± 1.1 3.9 ± 0.4 16 6.9 ± 0.3 12.6 ± 0.6 15.6 ± 2.3 1.8 ± 0.3 22.3 ± 1.3 12.1 ± 0.7 17 12.3 ± 1.1 7.1 ± 0.3 23.8 ± 3.0 16.3 ± 0.7 16.8 ± 0.8 27.9 ± 1.0 18 36.1 ± 4.4 10.3 ± 0.3 15.3 ± 0.9 18.6 ± 1.6 21.3 ± 3.5 8.8 ± 0.3 aFasting- after 24h restricted diet urine sample was collected first in the morning on the second day in a 2L urine bottle containing sodium hydroxide as a stabilizer after overnight fasting. bControl- on the second day urine was also collected 3 hrs after eating a cured meat sandwich. cExperimental- on the third day after overnight fasting urine was collected 3 hrs after eating another cured meat sandwich with 500 mg vitamin C tablet. a-c mean of duplicates.

145

Figure 3.12 Mean urinary nitrate (mg/mL) in healthy human volunteers consuming cured meat and on a diet low on nitrate and nitrite over three days. Based on t test there was no significant difference (p ≥ 0.05) between fasting, control and experiment nitrate on ham, salami and hotdog.

After fasting overnight, urinary nitrate showed an increase after the consumption of cured meat (ham, salami or hotdog) in the control group, urinary nitrate then decreased after taking vitamin C supplementation consumed with the same cured meat with the exception of hotdog (Figure 3.12). Based on student t test there were no significant difference (p ≥ 0.05) between fasting, control and experiment nitrate in ham salami and hotdog.

146

Figure 3.13 Mean urinary nitrite in healthy human volunteers consuming cured meat (mg/L) and on a diet low on nitrate and nitrite over three days. Based on student t test there were no significant difference (p ≥ 0.05) between fasting, control and experiment nitrite in ham, salami and hotdog.

After fasting overnight, a decrease in urinary nitrite was shown after consuming cured meat (ham, salami or hotdog) with the exception of volunteers consuming salami (Figure 3.13). After consuming the same cured meat with vitamin C supplementation, the urinary nitrite increased with the exception of volunteers consuming salami (Figure 3.13).

3.10.4 Discussion on nitrate and nitrite concentrations in the urine

It was demonstrated by Pannala et al. (2003) that high nitrate intake from food lead to a significant increase in nitrate and nitrite concentrations in the urine and saliva. They also demonstrated that the maximum urinary nitrate excretion occurred 4 to 6 hours after the consumption of high-nitrate meal. However, in this experiment a low nitrate and nitrite diet (Appendix A) was used thus most nitrate and nitrite excretion resulted from that of cured meat and endogenous formation. In healthy humans on a diet low in nitrite and nitrate, approximately 50 % of the urine nitrate originates from the in vivo synthesis of NO from L-arginine and gastrointestinal nitrogen oxides (Jobgen et al., 2007). In addition, Pannala et al. (2003) showed that the absorption of nitrate from organic nitrate source such as food was slower than from inorganic nitrate salt. It was proposed that organic dietary sources must undergo the additional extraction step and

147 hence takes longer to be absorbed. Sodium hydroxide pellets were added to urine bottles in order to stabilize nitrate and nitrite during storage until further analysis as suggested by Vermeer et al. (1998).

The decrease in urinary nitrite after consuming cured meat could be attributed to the reactivity of nitrite in the stomach as the pH in the stomach increased as a result of food and may have facilitated the conversion to NO, nitrous acid or enabled nitrosation (Figure 3.13). However, after consuming the same cured meat with vitamin C supplementation, the urinary nitrite increased with the exception of volunteers consuming salami (Figure 3.13). Urinary nitrite increased after consuming vitamin C supplement since ascorbic acid is known to compete with nitrite for nitrosatable compounds in the stomach

Nitrite can also be lost from the gastric juice by either oxidation to nitrate and/or by forming volatile species such as nitric oxide and nitrogen dioxide that escapes through the mouth and nose (Licht et al., 1988). Therefore only nitrite that remained in the body was excreted as urine. It was also demonstrated in vitro and in rats that nitric oxide can be generated by the anaerobic gut flora in the presence of nitrate or nitrite (Sobko et al., 2005). This may decrease the amount of nitrate and nitrite available to participate in nitrosation.

Air and soil quality is affected by climate change and other human impact, which explains the wide range of levels of nitrosamines of individual game species (Rywotycki, 2003). The NDMA content in cured meat was very low but contained nitrosatable amines and nitrite (Figure 3.11). Using GC-TEA, the mean NDMA in ham and sausage from France was 0.14 and 1.5 µg/kg, respectively (Mavelle et al., 1991). Using GC-MS, the mean NDMA in ham and sausage from China was 0.8 and 0.9 µg/kg, respectively (Song and Hu, 1988). Based on GC-MS method, Yurchenko and Mölder (2005) found the mean NDMA content in Estonian cured meats were: salami 0.84 µg/kg, frankfurters 0.16 µg/kg, ham 2.08 µg/kg and cooked sausage 1.3 µg/kg.

Different dietary sources of NOCs can be contributed to cultural differences in dietary traditions and practices. For examples, cured meats, especially bacon, are the major sources of nitrosamines in Western diets. On the other hand, fish and fish products contribute significant dietary nitrosamines in the Oriental diet. Cured meat is known to

148 contain preformed nitrosamine and nitrite and seafood contains plenty of amines such as methylamine, ethylamine, dimethylamine and diethylamine (Lin, 1990).

Sample preparation is used to increase the accuracy and precision of the analysis and consists of three major steps: sampling, homogenization and extraction. Since food such as cured meats consists of a range of compounds that may interact to form more complex compounds. This creates an inhomogeneous mixture of number of chemical substances, which makes sample preparation of food a crucial step before instrumental analysis. Unfortunately each sample preparation step introduces inherent errors that are irreversible and cumulative, especially the sampling step that is often the greatest source of error in chemical analysis (Lichon, 1996).

Homogenization is used to reduce particle size and for mixing. This is essential for increasing the number of particles in a test portion to increase the probability of sampling parts of the original (Lichon, 1996). The particle size is therefore important to food analyst because the size of surface area may influence the accuracy of the analytical procedure by the degree of exposure of the sample to reagents and other experimental conditions. Hence, recovery data may be inconsistent due to the particle size effects. Spiking just prior to HPLC analysis usually gives favorable recoveries since the only variable is HPLC separation and detection. However, spiking at the beginning of the extraction method gives a better indication of loss or degradation of the analytes during the extraction procedure as well as during HPLC separation and quantification (Russell, 2000).

149 3.10.5 Results on the effects of vitamin C supplement on NDMA formation in vivo

Ascorbic acid content of the vitamin C supplement given to volunteers were assessed (Figure 3.14).

Figure 3.14 Ascorbic acid content in 500 mg vitamin C tablets. Mean of five replicates from ten tablets. Mean recovery was 98.4 % with S.D. 3.96. Based on one-tail t test, ascorbic acid content was significantly different (p < 0.05) from the amount labeled. Very small variation between vitamin C tablets as indicated by small error bars.

Table 3.6 Quantification parameters on the effects of urine matrix on ascorbic acid recovery.

Parameters LOD (mg/L) LOQ (mg/L) Recovery (%) Linearity (R2) L-AA 5.0 5.0 92.5 ± 40.57 0.98768 Fifteen replicates.

From Table 3.6, AA recovery was reasonable at almost 93 % suggested some interference in the urine matrix on AA recovery. This may also be caused by the degradation of AA when exposed to sunlight.

150 Table 3.7 Individual ascorbic acid and NDMA in the urine of healthy human volunteers before and after treatments.

Volunteer Fastinga Controlb Experimentc number AA* NDMA+ ng/day AA NDMA AA NDMA 1 65 ± 12.2 0 17 ± 4.6 0 229 ± 23.2 0 2 14 ± 2.7 0 10 ± 1.1 0 267 ± 31.3 0 3 71 ± 16.8 0 16 ± 2.3 0 311 ± 28.7 0 4 25 ± 8.9 0 19 ± 3.1 0 289 ± 16.8 0 5 48 ± 10.2 0 21 ± 3.6 0 193 ± 10.9 0 6 92 ± 16.7 0 12 ± 1.4 0 267 ± 18.8 0 7 23 ± 2.4 0 8 ± 0.3 0 177 ± 13.6 0 8 31 ± 9.9 0 6 ± 1.0 0 161 ± 12.9 0 9 12 ± 1.3 0 9 ± 0.8 0 174 ± 18.7 0 10 29 ± 1.8 0 5 ± 0.3 0 226 ± 26.1 0 11 16 ± 2.1 0 13 ± 0.4 0 281 ± 16.3 0 12 87 ± 4.3 0 18 ± 1.3 0 199 ± 19.0 0 13 55 ± 6.8 0 15 ± 2.0 0 327 ± 26.1 0 14 69 ± 10. 0 3 ± 0.5 0 262 ± 15.8 0 15 39 ± 5.5 0 10 ± 2.9 0 193 ± 11.6 0 16 36 ± 6.1 0 16 ± 1.7 0 245 ± 13.3 0 17 62 ± 7.9 0 39 ± 5.8 0 388 ± 17.6 0 18 41 ± 3.8 0 29 ± 6.6 0 367 ± 23.5 0 Triplicates. aFasting-after 24h food restricted diet urine sample was collected first thing in the morning on the second day in a 2L urine bottle containing sodium hydroxide as a stabilizer after overnight fasting. bControl-also on the second day urine were collected 3 hrs after eating a cured meat sandwich.. cExperiment-on the third day after overnight fasting urine were collected 3 hrs after eating another cured meat sandwich with 500 mg vitamin C tablet. *AA-ascorbic acid, mg/L, +NDMA-N-nitrosodimethylamine, µg/L.

151

Figure 3.15 Effects of 500 mg vitamin C supplement on NDMA formation in healthy human volunteers consuming different types of cured meat and on a nitrate, nitrite and antioxidant restricited diet. For baseline comparison, volunteers fasted overnight had less than 50 mg/L AA in their urine and no NDMA. For control, volunteers were not given vitamin C supplement and their AA levels dropped to below 25 mg/L and had no NDMA after consuming a cured meat sandwich. The next day volunteers were given vitamin C supplment with a cured meat sandwich and had spiked AA levels but still had no NDMA in their urine. Based on one-way ANOVA ascorbic acid concentrations had no effect on NDMA formation in vivo.

3.10.6 Discusssion on the effects of vitamin C supplement on NDMA formation in vivo

Nitrosamines can be measured in human blood, urine and fecal matter after the consumption of nitrite or nitrate containing foods such as bacon and spinach, which demonstrated the in vivo production of NOCs in humans (Wagner and Tannenbaum, 1985). Once nitrite is swallowed, it is reduced to NO by ascorbic acid normally present in the stomach of healthy individuals. The NO is then absorbed through the mucosa, and the ascorbic acid is oxidized to dehydroascorbic acid. However, in the absence of ascorbic acid, nitrosating compounds such as nitrite can nitrosate precursors to form NOC (Mowat and McColl, 2001).

The NDMA content in human urine was measured before cured meat consumption as baseline level after overnight fasting and a nitrate and antioxidant restricted diet

152 (Appendix A). The NDMA was then measured again 3 h after cured meat consumption. The following day (3), each volunteer was asked to eat the same cured meat with the addition to 500 mg vitamin C tablet. Similarly, after 3 h the urine was collected for NDMA extracted using HS-SPME followed by quantification and confirmation with GC-MS.

It is important to estimate body’s ascorbic acid level since elevated nitrite-to-ascorbic acid ratio increases the formation of NOC in vivo and hence may increase the risks of developing gastric cancer (Mowat and McColl, 2001). Fasting ascorbic acid content was less than 50 mg/L (Figure 3.15) after antioxidant restricted diet and fasting overnight. No NDMA was found in the urine after nitrate and nitrite restricted diet (Table 3.7). Urinary ascorbic acid further dropped on day 2 as a result of antioxidant restricted diet (Table 3.7). No NDMA was formed 3 h after the consumption of cured meat with or without vitamin C supplementation (Figure 3.15). This suggested that on a diet low in nitrate, nitrite and antioxidant, one serving of cured meat was not sufficient on its own to produce detectable NDMA in vivo. However, if volunteers were on a higher or unrestricted nitrate and nitrite diet, some NDMA would be expected to form and be present in the urine. For example, it was shown that vitamin C in strawberries and garlic juice was effective in reducing NDMA formation in a typical Korean diet rich in amine and nitrate in stimulated salivary and gastric conditions (Choi et al., 2006). Vitamin C absorption rate varies depending on the concentration ingested and the vitamin C status. At 100 mg 80-95 % is absorbed, whereas at 1 g the absorption rate falls to 50 %, at 6 g to 25 % and at 12 g to 16 % (Bender, 2005). This may affect the true available vitamin C on NDMA inhibition in subjects (Figure 3.15)

Dallinga et al. (1998) measured the mean total volatile nitrosamines in the gastric juice of patients and were found to be 4.84 nmol/L and ranged between 0 and 17.7 nmol/L. The main nitrosamines found were NDEA at 3.1 nmol/L (approximately 70 % of the total volatile nitrosamnines found) followed by NDMA and NPYR at 0.90 and 0.38 nmol/L, respectively. The high levels found in their subjects do not reflect the major populations since their subjects already had various conditions of the gastrointestinal tract.

Since ascorbic acid is a free radical scavenger, it is suggested that it may prevent carcinogenesis by inhibiting nitrosation from occurring. In addition, it may also inhibit

153 carcinogens by blocking the conversion of precursors into carcinogens and carcinogenic metabolites (Radcliffe et al., 2003). In this experiment, low nitrate and nitrite intake did not result in the formation of NDMA with or without vitamin C supplementation (Fig. 3.15). Hence nitrosation inhibition by vitamin C was not demonstrated. However, this experiment did not look at other volatile and non-volatile nitrosamines and nitrosamides, respectively. Both ascorbic acid and ascorbate ion, or ASC, can indirectly inhibit nitrosation by competing for nitrosating agents in the stomach formed from reactions with nitrite and gastric juice contents. Subsequently, for each mole of ascorbic acid and ascorbate ion oxidized, which is an irreversible process, two moles of non-nitrosating nitric oxide is produced from nitrite. Although this reaction is fast, in the presence of oxygen, NO may be reconverted to nitrosating agents, thus counteracting the inhibition effects of ascorbic acid and ascorbate ion (Licht et al., 1988).

Thus ascorbate and erythorbate are added in cured meat to inhibit formation of nitrosamines (Cassens, 1995), as well as inhibiting nitrosation from occurring within the stomach from residual nitrite (Walters, 1980). In the United States, reducers such as ascorbate or erythrobates are often added in cured meat to promote the oxidation of nitrous acid to nitric acid, which reduces residual nitrite and retards the formation of N- nitrosamines (Pennington, 1998). The maximum addition of ascorbate in the United States was 550 mg/kg, of which 40 % or 209 mg/kg was detected in cured meat samples tested (Cassens, 1997).

As the enterosalivary recirculation of nitrate occurs, a constant supply of salivary nitrite to the stomach is ensured. Since ingested nitrite continuously reacts in vivo with gastric juice contents including ascorbic acid, its level was usually undetectable in the acid stomach (Mowat and McColl, 2001). However, once anti-secretary agents are released in the stomach, the pH increases to neutral, where nitrite remains stable and do not react with ascorbic acid, hence accumulates in the stomach and is excreted mainly in the urine (Mowat and McColl, 2001).

Acidification of urine, for example by taking vitamin C, has been used to protect against lower urinary tract infection, and it was suggested by Lundberg et al., (1997) that nitrite-producing bacteria induce their own death by supplying the substrate for the production of bacteriostatic NO under acidic conditions. In addition, infected urine may

154 contain large amounts of nitrite as a result of bacterial nitrate reductase activity. Thus in any human trials for measuring NO level, the health status of the subjects must be carefully evaluated in order to accurately interpret the findings related to possible oxidative stress and nitrosation potential as suggested by Junghans et al. (1999).

At acid pH, nitrite is converted to nitrous acid (pKa 3.2) and then to nitric oxide and nitrogen dioxide radicals. It is suggested that these radicals are harmful to bacteria but the conversion requires an acid environment (Radcliffe et al., 2003). Radcliffe et al. (2003) demonstrated that ascorbic acid could increase the effectiveness of sodium nitrite (200 mM) in reducing the viability of Streptococcus mutans by reducing the pH of the system thus increasing the conversion of nitrite to its free radical nitrogen species. Since ascorbic acid is not a toxic substance, up to 2000 mg of ascorbic acid showed no adverse effects, hence ascorbic acid may be safely added to food with high nitrite content such as cured meat, to increase the effectiveness of nitrite and possibly inhibit nitrosation reactions.

Österdahl (1988) estimated the average dietary intake of volatile nitrosamines in Sweden populations was 0.29 μg per person, and over 80 % of that came from meat and malt products, which was highest in fried bacon and pork. Beer and other malt beverages contributed 14 % of the dietary intake of volatile N-nitrosamines in Sweden, with dried products, cheese and smoked with represent 7 %. Österdahl (1988) reported that NDMA was the most common volatile N-nitrosamine in foods such as cocoa and chocolate products (up to 1.2 μg /kg), tea (up to1.2 μg /kg), coffee (0.1-0.8 μg /kg), and cereal products (0.2-0.9 μg /kg). In West Germany, Tricker et al. (1991) estimated the total volatile nitrosamine intake from food for men and women were 0.3 and 0.2 μg/day, respectively. Furthermore, 31 % of the daily intake of NDMA for men came from the consumption of beer, thus represented a significant source of dietary volatile nitrosamine.

Yamamoto et al., (1984) estimated the average daily intake of volatile nitrosamines in Japanese population was 0.5 µg per person and 88 % of it came from fish products, which were significant part of their diet. The estimation was based on food consumption data and their analysis of volatile nitrosamine content of commonly eaten food including fresh fish, fish products, meat, cheese, and beer.

155 Different countries and regions within a country may have a distinctive and different dietary practice compared to other countries and regions, therefore the major food groups contributing to dietary NDMA intake will vary depending on their culture and hence dietary lifestyles.

According to Tricker et al. (1991), the NDMA range found in sausage products was 0.5 to 1.8 µg/kg with a mean of 0.84 µg/kg, and for bacon and ham it ranged from 0.5 to 1.6 µg/kg with a mean of 1.01 µg/kg. This studied showed that the NDMA in ham, hotdog and salami were 0.42, 1.43 1.83 µg/kg, respectively (Figure 18), which were slightly outside their range with a mean of 1.23 µg/kg. NDMA was highest in fresh fish and fish products with a range of 0.5 to 8.0µg/kg and a mean of 2.18 µg/kg. Despite this, meat and meat products contributed more than double the daily dietary intake of NDMA for both men and women compared to fish products. There was no detectable NDMA in fresh vegetables and only negligible levels were found in preserved vegetables. Other volatile nitrosamines including NPIP and NPYR were not present in vegetables and only negligible levels were found in sausage products (Tricker et al., 1991).

The mean daily intake of NDMA from French foods and beverages between 1987 and 1992 was 0.19 µg/day, and one-third of this came from the consumption of alcoholic beverages. Vegetables represented 22 % of NDMA intake followed by meat and meat products at 12.5 % (Biaudet et al., 1994).

Mitacek et al. (1999) concluded that based on case-control studies in Thai population that exposure to exogenous and possible endogenous nitrosamines in food or tobacco

156 may contribute to the development of liver cancer. With over 1800 food samples tested, relatively high levels of NDMA, NPIP and NPYR were detected in traditional Thai foods such as fermented, salted and dried fish. Levels of these volatile nitrosamines ranged from trace amount to 146 µg/kg.

NDMA was the most common occurring volatile nitrosamine in Chinese foods with up to 6 µg/kg in beer, up to 7.4 µg/kg in meat products, and up to 131.5 µg/kg in some seafood. Other volatile nitrosamines detected were NDEA, NPIP and NPYR (Song and Hu, 1988).

It was demonstrated that NOC concentration and stomach pH increased significantly with age, and there was a positive correlation between pH and NOC concentration, and between pH and an increase concentration of nitrites. In addition, it was demonstrated that the population of nitrate reductase-positive microorganisms rose as NOC and nitrite levels increased (Reed et al., 1981). The rise in gastric pH with age maybe due to decreased gastric acid secretion as a result of natural aging process. This rise in pH then allows microorganisms to survive in the stomach-converting nitrate present in the stomach to nitrite. Since nitrite is more stable at higher pH, this may allow more nitrosation to occur thus resulting in higher N-nitrosamine levels with increasing age.

N-nitrosamine levels ranged from 0.01 to 40 µmol/L in healthy volunteers and patients aged between 18 and 87 years. A significant rise in N-nitrosamine concentrations and gastric pH was demonstrated with increasing age. Furthermore, N-nitrosamine levels were significantly higher in males than females, taking other factors into account. Cigarette smoking had no significant effect on either N-nitrosamine or nitrite concentrations (Reed et al., 1981).

Contrary to other findings, Risch et al. (1985) found that vitamin C intake was only slightly protective against gastric cancer, and that vitamin E had not effects at all.

157 Similarly, Kato et al. (1992) revealed an adverse effect of fruit consumption, which may be due to the lack of detailed information on the types of fruits consumed in the rural Japanese population surveyed.

Xu and Read (1993) had demonstrated that significantly higher concentration of NOC was produced at lower pH ranges of 1.1 to 3.0 and at higher pH ranges of 6.0 to 8.4 compared to a pH range of 3.0 to 6.0. They had also shown that nitrite concentration is closely related to intragastric pH, such that at pH < 5.0, the nitrite concentrations are very low. It was suggested that at normal gastric pH the nitrite content is very low because nitrite has a very short half-life (less than 10 minutes), and that it is very reactive at low pH with high absorption rate.

Although ascorbic acid shown little reactivity at neutral pH, as in the stomach of achlorhydric patients, it was shown to be a potent inhibitor of bacterial mediated N- nitrosamine formation by Pseudomonas aeruginosa, which is capable of rapid rates of N-nitrosation (Mackerness et al., 1989). Furthermore, by using in vitro studies of bacterial suspension on nitrite, Mackerness et al. (1989) had confirmed the effectiveness of ascorbate in inhibiting bacterial N-nitrosation by competing with amine for the nitrosating agents derived from the action of bacteria.

Mirvish et al. (1972) demonstrated that ascorbate inhibits N-nitrosation by competing + for available nitrite in the form of N2O3 and H2NO2 . It was also shown that ascorbate anion is 240 times more rapidly nitrosated than ascorbic acid. At pH 3.0 to 5.0, ascorbate anion is abundant and that the reaction with nitrite is very rapid, where the

158 formation of N2O3 becomes the rate limiting step. Thus the optimal pH for the inhibition of N-nitrosation by ascorbate is at pH 3.0 to 4.0.

Japan has seven times the rate of gastric cancer than the United States and is also significantly higher compared to the United Kingdom and Germany. There is little evidence that genetic differences contributed to the different gastric cancer rates (Davies and Sano, 2001). Thus this epidemiological study suggests that diet and lifestyle may play an important role in gastric cancer aetiology besides effective screening and management. Countries such as South Korea, Japan and China had the highest stomach cancer mortality for men, whereas countries with the highest stomach cancer mortality for women were South Korea, China and Columbia. Canada and Denmark had the lowest stomach cancer mortality for men and women, respectively (Joossens et al., 1996).

159 high consumption of salted pickled cabbage and salted seafood sauce (Seel et al., 1994). The former also contained high levels of total NOC precursors, and cabbages, which are known to contain high levels of nitrate than any other vegetables.

Comparison with other countries dietary nitrate and nitrite intake can be difficult since different countries and even regions within a country may use different processing technologies to achieve the end product with particular sensory characteristics (Flores, 1997).

3.10.11 Conclusion

N-nitrosamines in food and beverages are the result of the manufacturing processes such as smoking and curing. Therefore changes or modification of this process should eliminate or at least minimize their presence in food (Mavelle et al., 1991).

Diet and lifestyle contributes to significant number of preventable cancers. Nitrite is used as a preservative in cured meat and is known to produce carcinogens in the stomach. Vitamin C was known to reduce the formation of these carcinogens in vitro and in vivo. Based on the human trial presented, the intake of one serve of cured meat was not enough to cause the formation of NDMA in vivo, thus vitamin C supplementation had no effect on NDMA formation. However, vitamin C supplementation may have decreased nitrite ions available for nitrosation by competing for nitrosatable compounds in the stomach, therefore may have prevented the formation of NDMA.

160 3.11 References

Al-Mamary, M., Al-Habori, M., Al-Shoaibi, Z., and Shamsan, B. 2006. Nitrosamine formation from different Catha edulis leaves extracts under stimulated gastric condition. Food Chemistry, 97:586-590. Andrade, R., Reyes, F. G., and Rath, S. 2005. A method for the determination of volatile N-nitrosamines in food by HS-SPME-GC-TEA. Food Chemistry, 91:173-179. Annola, K., Karttunen, V., Luukkainen, P., Segerbäck, D., and Vähäkangas, K. 2007. Transplacental transfer of NDMA in human placental perfusion and DNA adduct formation in the placenta. Toxicology Letters, 172S:S176. Antonelli, M.L., D’Ascenzo, G., Laganà, A., Pusceddu, P. 2002. Food analyses: a new calorimetric method for ascorbic acid (vitamin C) determination. Talanta, 58:961-967. Archer, M.C. 1989. Mechanisms of action of N-nitroso compounds. Cancer Surveys, 8:241-250. Arya, S. P., Mahajan, M., and Jain, P. 2000. Non-spectrophotometric methods for the determination of vitamin C. Analytica Chimica Acta, 417:1-14. Aurand, L. W., Woods, A. E., and Wells, M. R. 1987. Food Composition and Analysis. Van Nostrand Reinhold Company, New York. Bartsch, H., Ohshima, H., Shuker, D., Pignatelli, B., and Calmels, S. 1990. Exposure of humans to endogenous N-nitroso compounds: implications in cancer etiology. Mutation Research, 238:255-267. Belitz, H.-D., and Grosch, W. 1987. Food Chemistry. Springer Verlag, Berlin. Bender, D.A. 2005. Ascorbic acid: physiology, dietary sources and requirements. Encyclopedia of Human Nutrition, 169-176. Biaudet, H., Mavelle, T., and Debry, G. 1994. Mean daily intake of N- nitrosodimethylamine from foods and beverages in France in 1987-1992. Food and Chemical Toxicology, 32:417-421. Burini, G. 2007. Development of a quantitative method for the analysis of total L- ascorbic acid in foods by high-performance liquid chromatography. Journal of Chromatography A, 1154:97-102. Byun, M-W., Ahn, H-J., Kim, J-H., Lee, J-W., Yook, H-S., and Han, S-B. 2004. Determination of volatile N-nitrosamines in irradiated fermented sausage by gas chromatography coupled to a thermal energy analyzer. Journal of Chromatography A, 1054:403-407. Cassens, R.G. 1995. Use of nitrite in cured meats today. Food Technology, 49:72-80. Cassens, R.G. 1997. Residual nitrite in cured meat. Food Technology, 51:53-55. Chen, C.S., Pignatelli, B., Malaveille, C., Bouvier, G., Shuker, D., Hautefeuille, A., Zhang, R.F., Bartsch, H. 1992. Levels of direct-acting mutagens, total N-nitroso

161 compounds in nitrosated fermented fish products, consumed in a high-risk area for gastric cancer in southern China. Mutation Research, 265:211-221. Choi, S.Y., Chung, M.J., and Sung, N.J. 2002. Volatile N-nitrosamine inhibition after intake Korean green tea and Maesil (Prunus mume SIEB. et ZACC.) extracts with an amine-rich diet in subjects ingesting nitrate. Food and Chemical Toxicology, 40:949- 957. Choi, S.Y., Chung, M.J., Lee, S.-J., Shin, J.H., and Sung, N.J. 2006. N-nitrosamine inhibition by strawberry, garlic, kale, and the effects of nitrite-scavenging and N- nitrosamine formation by functional compounds in strawberry and garlic. Food Control, 18:485-491. Chow, C.K., and Hong, C.B. 2002. Dietary vitamin E and selenium and toxicity of nitrite and nitrate. Toxicology, 180:195-207. Chung, M.J., Lee, S.H., Sung, N.J. 2002. Inhibitory effect of whole strawberries, garlic juice or kale juice on endogenous formation of N-nitrosodimethylamine in humans. Cancer Letters, 182:1-10. Correa, P. 1992. Human gastric carcinogenesis: A multistep and multifactorial process- First American cancer society award lecture on cancer epidemiology and prevention. Cancer Research, 52:6735-6740. Cotelle, P., and Vezin, H. 2001. Reaction of caffeic acid derivatives with acidic nitrite. Tetrahedron Letters, 42:3303-3305. Dallinga, J.W., Pachen, D.M., Lousberg, A.H., van Geel, J.A., Houben, G.M., Stockbrugger, R.W., van Maanen, J.M., and Kleinjans, J.C. 1998. Volatile N- nitrosamines in gastric juice of patients with various conditions of the gastrointestinal tract determined by gas chromatography-mass spectrometry and related to intragastic pH and nitrate and nitrite levels. Cancer Letters, 124:119-125. Davis, P.A., and Sano, T. 2001. The difference in gastric cancer between Japan, USA and Europe: What are the facts? What are the suggestions? Critical Reviews in Oncology/Hematology, 40:77-94.

Dodds, M.W., Johnson, D.A., and Yeh, C.-K. 2005. Health benefits of saliva: a review. Journal of Dentistry, 33:223-233.

Duncan, C., Li, H., Dykhuizen, R., Frazer, R., Johnston, P., MacKnight, G., Smith, L., Lamza, K., McKenzie, H., Batt, L., Kelly, D., Golden, M., Benjamin, N., and Leifert, C. 1997. Protection against oral and gastrointestinal diseases: importance of dietary nitrate intake, oral nitrate reduction and enterosalivary nitrate circulation. Comparative Biochemistry and Physiology, 118A:939:948. Dunn, B.P., and Stich, H.F. 1984. Determination of free and protein-bound N- nitrosoproline in nitrite-cured meat products. Food and Chemical Toxicology, 22:609- 613. Duvoix, A., Blasius, R., Delhalle, S., Schnekenburger, M., Morceau, F., Henry, E., Dicato, M., and Diederich, M. 2005. Chemopreventive and therapeutic effects of curcumin. Cancer Letters, 223:181-190.

162 Dykhuizen, R.S., Fraser, A., McKenzie, H., Golden, M., Leifert, C., and Benjamin, N. 1998. Helicobacter pylori is killed by nitrite under acidic conditions. Gut, 42:334-337. Ellen, G., Schuller, P.L., Froeling, P.G., and Bruijns, E. 1982. Non volatile N- nitrosamines detected in blood and urine from patients ingesting daily large amounts of ammonium nitrate. Food and Chemical Toxicology, 20:879-882. Esteban, M.R., and Ho, C-N. 1997. Enzymatic spectrophotometric determination of ascorbic acid in commercial vitamin C tablets. Microchemical Journal, 56:122-129. European Union Scientific Commission on Food. 1995. Fiddler, W. 1975. The occurrence and determination of N-nitroso compounds. Toxicology and Applied Pharmacology, 31:352-360. Flores, J. 1997. Mediterranean vs northern European meat products. Processing technologies and main differences. Food Chemistry, 59:505-510. Fontannaz, P., Kilinç, T., and Heudi, O. 2006. HPLC-UV determination of total vitamin C in a wide range of fortified food products. Food Chemistry, 94:626-631. Furusawa, N. 2001. Rapid high-performance liquid chromatographic identification/quantification of total vitamin C in fruit drinks. Food Control, 12:27-29. Gangolli, S.D., van den Brandt, P., Feron, V.J., Janzowsky, C., Koeman, J.H., Speijers, G.J., Spiegelhalder, B., Walker, R., and Wishnok, J.S. 1994. Nitrate, nitrite and N- nitroso compounds, European Journal of Pharmacology, Environmental Toxicology and Pharmacology Section, 292:1-38. Gerbaux, P., Wantier, P., and Flammang, R. 2004. Nitrosation of thiols and thioethers in the gas phase: A combined theoretical and experimental study. Journal of American Society for Mass Spectrometry, 15:344-355. González, C., Riboli, E., Badosa, J., Batiste, E., Cardona, T., Pita, S., Sanz, J.M., Torrent, M., and Agudo, A. 1994. Nutritional factors and gastric cancer in Spain. American Journal of Epidemiology, 139:466-473. González, P.J., Correia, C., Moura, I., Brondino, C.D., and Moura, J.J.G. 2006. Bacterial nitrate reductases: Molecular and biological aspects of nitrate reduction. Journal of Inorganice Biochemistry, 100:1015-1023. Gough, T.A., Webb, K.S., and Swann, P.F. 1983. An examination of human blood for the presence of volatile nitrosamines. Food and Chemical Toxicology, 21:151-156. Gray, J. D., and Shiner, M. 1967. Influence of gastric pH on gastric and jejunal flora. Gut, 8:574-581. Grebel, J.E., Young, C.C., and Suffet, I.H. 2006. Solid-phase microextraction of N- nitrosamines. Journal of Chromatography A, 1117:11-18. Grudpan, K., Kamfoo, K., and Jakmunee, J. 1999. Flow injection spectrophotometric or conductometric determination of ascorbic acid in a vitamin C tablet using permanganate or ammonia. Talanta, 49:1023-1026.

163 Hansson, L-E., Nyrén, O., Bergström, R., Wolk, A., Lindgren, A., Baron, J., and Adami, H-O. 1994. Nutrients and gastric cancer risk. A population-based case-control study in Sweden. International Journal of Cancer, 57:638-644. Helaleh, M.I.H., and Korenaga, T. 2000. Ion chromatographic method for simultaneous determination of nitrate and nitrite in human saliva. Journal of Chromatography B, 744:433-437. Huang, D.P., Ho, J.H.C., Webb, K.S., Wood, B.J., and Gough, T.A. 1981. Volatile nitrosamines in salt-preserved fish before and after cooking. Fd Cosmet. Toxicology, 19:167-171. Hwang, H., Dwyer, J., and Russell, R.M. 1994. Diet, Helicobacter pylori infection, food preservation and gastric cancer risk: are there new roles for preventative factors? Nutrition Reviews, 52:75-83. Issenberg, P. 1981. ‘Analytical methods for nitrosamines’, in N-nitroso Compounds. eds R.A. Scanlan and S.R. Tannenbaum. American Chemical Society Symposium Series 174, United States. Jo, C., Ahn, H.J., Son, J.H., Lee, J.W., and Byun, M.W. 2003. Packaging and irradiation effect on lipid oxidation, color, residual nitrite content, and nitrosamine formation in cooked pork sausage. Food Control, 14:7-12. Jobgen, W.S., Jobgen, S.C. Li, H., Meininger, C.J., and Wu, G. 2007. Analysis of nitrite and nitrate in biological samples using high-performance liquid chromatography. Journal of Chromatography B, 851:71-82. Joossens, J.V., Hill, M.J., Elliott, P., Stamler, R., Stamler, J., Lesaffre, E., Dyer, A., Nichols, R., and Kesteloot, H. 1996. Dietary salt, nitrate and stomach cancer mortality in 24 countries. International Journal of Epidemiology, 25:494-504. Junghans, P., Fischer, U., and Kuklinski, B. 1999. Body nitrosation potential measured by a novel 15N breath test. Free Radical Biology and Medicine, 27:1302-1307. Kall, M.A., and Andersen, C. 1999. Improved method for simultaneous determination of ascorbic acid and dehydroascorbic acid, isoascorbic acid and dehydroascorbic acid in food and biological samples. Journal of Chromatography B, 730:101-111. Kataoka, H., Lord, H.L., Pawliszyn, J. 2000. Applications of solid-phase microextraction in food analysis. Journal of Chromatography A, 880:35-62. Kato, I., Tominaga, S., and Matsumoto, K. 1992. A prospective study of stomach cancer among a rural Japanese population: a 6-year survey. Japanese Journal of Cancer Research, 83:568-575. Khalaf, H., and Steinert, J. 2000. Chapter 22, ‘Determination of N-nitoso compounds in foods and beverages using HPLC’, In Food Analysis by HPLC, 2nd ed. Ed L. M. L. Nollet. Marcel Dekker, New York. Krishnaswamy, K., and Raghuramulu, N. 1998. Bioactive phytochemicals with emphasis on dietary practices. Indian Journal of Medical Research, 108:167-181.

164 Kolpe, U., Ramaswamy, V., Satish Rao, B. S., and Nagabhushan, M. 2002. Turmeric and curcumin prevents the formation of mutagenic Maillard reaction products. International Congress Series, 1245:327-334. Krul, C.A.M., Zeilmaker, M.J., Schothorst, R.C., and Havenaar, R. 2004. Intragastric formation and modulation of N-nitrosodimethylamine in a dynamic in vitro gastrointestinal model under human physiological conditions. Food and Chemical Toxicology, 42:51-63. Lakritz, L., Gates, R.A., Gugger, A.M., and Wasserman, A.E. 1982. Nitrosamine levels in human blood, urine and gastric aspirate following ingestion of foods containing potential nitrosamine precursors or preformed nitrosamines. Food and Chemical Toxicology, 20:455-459. Lee, S.Y.H., Munerol, B., Pollard, S., Youdim, K.A., Pannala, A.S., Kuhnle, G.G., Debnam, E.S., Rice-Evans, C., and Spencer, J.P. 2006. The reaction of flavanols with nitrous acid protects against N-nitrosamine formation and leads to the formation of nitroso derivatives which inhibit cancer cell growth. Free Radical Biology and Medicine, 40:323-334. Levallois, P., Ayotte, P., van Maanen, J.M.S., Desrosiers, T., Gingras, S., Dallinga, J.W., Vermeer, I.T.M., Lee, J., and Poirier, G. 2000. Excretion of volatile nitrosamines in a rural population in relation to food and drinking water consumption. Food and Chemical Toxicology, 38:1013-1019. Lichon, M. J. 1996. ‘Sample preparation’ in Handbook of Food Analysis: Physical Characterization and Nutrient Analysis, vol. 1. ed L. M. L. Nollet, Marcel Dekker Inc., New York. Licht, W. R., Tannenbaum, S. R., and Deen, W. M. 1988. Use of ascorbic acid to inhibit nitrosation: kinetic and mass transfer considerations for an in vitro system. Carcinogenesis, 9:365-372. Lijinsky, W. 1999. N-Nitroso compounds in the diet. Mutation Research, 443:129-138. Lin, J.K. 1990. Nitrosamines as potential environmental carcinogens in man. Clinical Biochemistry, 23:67-71. Lundberg, J. O., Carlsson, S., Engstrand, L., Morcos, E., Wiklund, N. P., and Weitzberg, E. 1997. Urinary nitrite: More than a marker of infection. Urology, 50:189- 191. Lv, C.L., Liu, Y.D., Zhong, R., and Wang, Y. 2006. Theoretical studies on the formation of N-nitrosodimethylamine. Journal of Molecular Structure: THEOCHEM, 802:1-6. Mackerness, C. W., Leach, S. A., Thompson, M. H., and Hill, M. J. 1989. The inhibition of bacterially mediated N-nitrosation by vitamin C: relevance to the inhibition of endogenous N-nitrosation in the achlorhydric stomach. Carcinogenesis, 10:397-399.

165 Massey, R.C., Crews, C., and McWeeny, D.J. 1982. Method for high-performance liquid chromatographic measurement of N-nitrosamines in food and beverages. Journal of Chromatography, 241:423-427. Mavelle, T., Bouchikhi, B., and Debry, G. 1991. The occurence of volatile N- nitrosamines in French foodstuffs. Food Chemistry, 42: 321-338. McWeeny, D.J. 1983. Nitrosamines in beverages. Food Chemistry, 11:273-287. Miller, B.J., Billedeau, S.M., and Miller, D.W. 1989. Formation of N-nitrosamines in microwaved versus skillet-fried bacon containing nitrite. Food and Chemical Toxicology, 27:295-299. Mirvish, S. S., Wallcave, L., Eagen, M., and Shubik, P. 1972. Ascorbate-nitrite reaction: Possible means of blocking the formation of carcinogenic N-nitroso compounds. Science, 177:65-68. Mirvish, S.S. 1995. Role of N-nitroso compounds (NOC) and N-nitrosation in etiology of gastric, esophageal, nasopharyngeal and bladder cancer and contribution to cancer of known exposures to NOC. Cancer Letters, 93:17-48. Mitacek, E.J., Brunnemann, K.D., Suttajit, M., Martin, N., Limsila, T., Ohshima, H., and Caplan, L.S. 1999. Exposure to N-Nitroso compounds in a population of high liver cancer regions in Thailand: Volatile nitrosamine (VNA) levels in Thai food. Food and Chemical Toxicology, 37:297-305. Mowat, C., Carswell, A., Wirz, A., and McColl, K. E. 1999. Omeprazole and dietary nitrate independently affect levels of vitamin C and nitrite in gastric juice. Gastroenterology, 116:813-822. Mowat, C., and McColl, K. E. 2001. Alternations in intragastric nitrite and vitamin C levels during acid inhibitory therapy. Best Practice and Research Clinical Gastroenterology, 15:523-537. Nagabhushan, M., Nair, U. J., Amonkar, A. J., D’Souza, A. V., and Bhide, S. V. 1988. Curcumins as inhibitors of nitrosation in vitro. Mutation Research, 202:163-169. National Academy of Sciences. 1981. The Health Effects of Nitrate, Nitrite and N- Nitroso Compounds. National Academy Press, Washington. van Niekerk, P. J. 1982. ‘Determination of vitamins’, In HPLC in Food Analysis. Ed. R. Macrae. Academic Press, London. Nisperos-Carriedo, M.O., Buslig, B.S., and Shaw, P.E. 1992. Simultaneous detection of dehydroascorbic, ascorbic, and some organic acids in fruits and vegetable by HPLC. Journal of Agriculture and Food Chemistry, 40:1127-1130. Nowell, S., Coles, B., Sinha, R., MacLeod, S., Ratnasinghe, D.L., Stotts, C., Kadlubar, F.F., Ambrosone, C.B., and Lang, N.P. 2002. Analysis of total meat intake and exposure to individual heterocyclic amines in a case-control study of colorectal cancer: contribution of metabolic variation to risk. Mutation Research, 506-507:175-185.

166 Odriozola-Serrano, I., Hernández-Jover, T., Martín-Belloso, O. 2007. Comparative evaluation of UV-HPLC methods and reducing agents to determine vitamin C in fruits. Food Chemistry, 105:1151-1158. Ohsawa, K-I., Nakagawa, S-Y., Kimura, M., Shimada, C., Tsuda, S., Kabasawa, K., Kawaguchi, S., and Sasaki, Y. F. 2003. Detection of in vivo genotoxicity of endogenously formed N-nitroso compounds and suppression by ascorbic acid, teas and fruit juices. Mutation Research, 539:65-76. Ohshima, H., and Bartsch, H. 1981. Quantitative estimation of endogenous nitrosation in humans by monitoring N-nitrosoproline excreted in the urine. Cancer Research, 41:3658-3662. Ologhobo, A.D., Adegede, H.I., and Maduagiwu, E.N. 1996. Occurrence of nitrate, nitrite and volatile nitrosamines in certain feedstuffs and animal products. Nutrition and Health, 11:109-114. Österdahl, B.-G. 1988. Volatile nitrosamines in foods on the Swedish market and estimation of their daily intake. Food Additives and Contaminants, 5:587-595. Pannala, A., Mani, A. R., Spencer, J. P. E., Skinner, V., Bruckdorfer, K. R., Moore, K. P., and Rice-Evans, C. A. 2003. The effect of dietary nitrate on salivary, plasma, and urinary nitrate metabolism in humans. Free Radical Biology and Medicine, 34:576- 584. Pennington, J.A. 1998. Dietary exposure models for nitrates and nitrites. Food Control, 9:389-395. Pourazrang, H., Moazzami, A.A., and Fazky Bazzaz, B.S. 2002. Inhibition of mutagenic N-nitroso compound formation in sausage samples using L-ascorbic acid and α-tocopherol. Meat Science, 62:479-483. Ralt, D., and Tannenbaum, S.R. 1981. ‘The role of bacteria in nitrosamine formation’, in N-nitroso Compounds. eds R.A. Scanlan and S.R. Tannenbaum. American Chemical Society Symposium Series 174, United States. Radcliffe, C.E., Lamb, R., Blinkhorn, A.S., Drucker, D.B. 2003. Effect of sodium nitrite and ascorbic acid on the growth and acid production of Streptococcus mutans. Journal of Dentistry, 31:367-370. Reed, P.I., Haines, K., Smith, P.L., House, F.R., and Walters, C.L. 1981. Gastric juice N-Nitrosamines in health and gastroduodenal disease. The Lancet, 318:550-552. Risch, H.A., Jain, M., Choi, N.W., Fodor, J.G., Pfeiffer, C.J., Howe, G.R., Harrison, L.W., Craib, K.J., and Miller, A.B. 1985. Dietary factors and the incidence of cancer of the stomach. American Journal of Epidemiology, 122:947-959. Russell, L. F. 2000. Chapter 10, ‘Quantitative determination of water-soluble vitamins’, In Food Analysis by HPLC, 2nd ed. Ed L. M. L. Nollet. Marcel Dekker, New York. Rywotycki, R. 2002. The effect of selected functional additives and heat treatment on nitrosamine content in pasteurized pork ham. Meat Science, 60:335-339.

167 Rywotycki, R. 2003. Meat nitrosamine contamination level depending on animal breeding factors. Meat Science, 65:669-676. Rywotycki, R. 2007. The effect of baking of various kinds of raw meat from different animal species and meat with functional additives on nitrosamine contamination level. Food Chemistry, 101:540-548. Scanlan, R.A., and Reyes, F.G. 1985. An update on analytical techniques for N- nitrosamine. Food Technology, 39:95-99. Seel, D.J., Kawabata, T., Nakamura, M., Ishibashi, T., Hamano, M., Mashimo, M., Shin, S.H., Sakamoto, K., Jhee, E.C., Watanabe, S. 1994. N-nitroso compounds in two nitrosated food products in Southwest Korea. Food and Chemical Toxicology, 32:1117- 1123.

Sen, N.P., Seaman, S.W., and Baddoo, P.A. 1985. N-nitrosothiazolidine and nonvolatile N-nitroso compounds in foods. Food Technology, 39:84-88. Shahidi, F., and Pegg, R.B. 1994. Absence of volatile N-nitrosamines in cooked nitrite- free cured muscle foods. Meat Science, 37:327-336. Shank, R.C. 1975. Toxicology of N-nitroso compounds. Toxicology and Applied Pharmacology, 31:361-368. Shephard, S., Schlatter, C., and Lutz, W. 1987. Assessment of the risk of formation of carcinogenic N-nitroso compounds from dietary precursors in the stomach, Food and Chemistry Toxicology, 25:91-108. Skrypec, D.J., Gray, J.I., Mandagere, A.K., Booren, A.M., Pearson, A.M., and Cuppett, S.L. 1985. Effect of bacon composition and processing on N-nitrosamine formation. Food Technology, 39:74-79. Silva, F.O. 2005. Total ascorbic acid determination in fresh squeezed orange juice by gas chromatography. Food Control, 16:55-58. Sobala, G. M., Schorah, C. J., Sanderson, M., Dixon, M. F., Tompkins, D. S., Godwin, P., and Axon, A. T. 1989. Ascorbic acid in the human stomach. Gastroenterology, 97:357-363. Sobko, T., Reinders, C.I., E.Å., Jansson, Norin, E., Midtvedt, T., and Lundberg, J.O. 2005. Gastrointestinal bacteria generate nitric oxide from nitrate and nitrite. Nitric Oxide, 13:272-278. Song, P.J., and Hu, J.F. 1988. N-nitrosamines in Chinese foods. Food and Chemical Toxicology, 26:205-208. Smith, M.E., and Morton, D.G. 2001. The Digestive System. Churchill Livingstone, Edinburgh. Sung, N.-J. 2004. ‘N-Nitroso compounds in food’, In Handbook of Food Analysis: Residues and Other Food component Analysis, Volume 2, 2nd ed. L.M. Nollet (ed). Marcel Dekker, New York.

168 Tabak, M., Armon, R., Rosenblat, G., Stermer, E., and Neeman, I. 2003. Diverse effects of ascorbic acid and palmitoyl ascorbate on Helicobater pylori survival and growth. FEMS Microbiology Letters, 224:247-253. Tai, A., and Gohda, E. 2007. Determination of ascorbic acid and its related compounds in foods and beverages by hydrophilic interaction liquid chromatography. Journal of Chromatography B, 853:214-220. Tricker, A.R., and Kubacki, S.J. 1992. Review of the occurrence and formation of non- volatile N-nitroso compounds in foods. Food Additives and Contaminants, 9:39-69. Tricker, A.R., Pfundstein, B., Theobald, E., Preussmann, R., and Spiegelhalder, B. 1991. Mean daily intake of volatile N-nitrosamines from foods and beverages in West Germany in 1989-1990. Food and Chemical Toxicology, 29:729-732. Tricker, A.R., and Preussmann, R. 1991. Carcinogenic N-nitrosamines in the diet: occurrence, formation, mechanisms and carcinogenic potential. Mutation Research, 259:277-289. Uppu, R. M., Lemercier, J-N., Squadrito, G. L., Zhang, H., Bolzan, R. M., and Pryor, W. A. 1998. Nitrosation by peroxynitrite: Use of phenol as a probe. Archives of Biochemistry and Biophysics, 358:1-16. Ventanas, S., Martín, D., Estévez, M., and Ruiz, J. 2006. Analysis of volatile nitrosamines from a model system using SPME-DED at different temperatures and times of extraction. Food Chemistry, 99:842-850. Ventanas, S., and Ruiz, J. 2006. On-site analysis of volatile nitrosamines in food model systems by solid-phase microextraction coupled to a direct extraction device. Talanta, 70:1017-1023. Vermeer, I.T., Pachen, D.M., Dallinga, J.W., Kleinjans, J.C., van Maanen, J.M. 1998. Volatile N-nitrosamine formation after intake of nitrate at the ADI level in combination with an amine-rich diet. Environmental Health Perspective, 106:459-463. Vermeer, I.T., Moonen, E.J., Dallinga, J.W., Kleinjans, J.C., and van Maanen, J.M. 1999. Effect of ascorbic acid and green tea on endogenous formation of N- nitrosodimethylamine and N-nitrosopiperidine in humans. Mutation Research, 428:353-361. Wagner, D.A. and Tannenbaum, S.R. 1985. In-vivo formation of N-nitroso compounds. Food Technology, 39:89-90. Walters, C.L. 1980. The exposure of humans to nitrite. Oncology, 37:289-296. Washko, P. W., Welch, R. W., Dhariwal, K., R., Wang, Y., and Levine, M. 1992. Ascorbic acid and dehydroascorbic acid analyses in biological samples. Analytical Biochemistry, 204:1-14. Weisburger, J.H. 1981. ‘N-nitroso compounds: diet and cancer trends’, in N-nitroso Compounds. eds. R.A. Scanlan and S.R. Tannenbaum. American Chemical Society Symposium Series 174, United States.

169 Weisburger, J.H., and Chung, F.-L. 2002. Mechanisms of chronic disease causation by nutritional factors and tobacco products and their prevention by tea polyphenols. Food and Chemical Toxicology, 40:1145-1154. Wilson, J.X. 2002. The physiological role of dehydroascorbic acid. FEBS Letters, 527:5-9. Wogan, G.N., and Tannenbaum, S.R. 1975. Environmental N-Nitroso compounds: implications for public health. Toxicology and Applied Pharmacology, 31:375-383. Xu, G. P., and Reed, P. J. 1993. N-Nitroso compounds in fresh gastric juice and their relation to intragastric pH and nitrite employing an improved analytical method. Carcinogenesis, 14:2547-2551. Yamamoto, M., Iwata, R., Ishiwata, H., Yamada, T., and Tanimura, A. 1984. Determination of volatile nitrosamine levels in foods and estimation of their daily intake in Japan. Food and Chemical Toxicology, 22:61-64. Yamasaki, H., and Sakihama, Y. 2000. Simultaneous production of nitric oxide and peroxynitrite by plant nitrate reductase: in vitro evidence for the NR-dependent formation of active nitrogen species. FEBS Letter, 468:89-92. Yurchenko, S., and Mölder, U. 2005. The occurrence of volatile N-nitrosamines in Estonaian meat products. Food Chemistry, 100:1713-1721. Yurchenko, S., and Mölder, U. 2006. Volatile N-nitrosamines in various fish products. Food Chemistry, 96:325-333. Zapata, S., and Dufour, J.-P. 1992. Ascorbic, dehydroascorbic and isoascorbic acid simultaneous determinations by reverse phase ion interaction HPLC. Journal of Food Science, 57:506-511.

170 Chapter 4 Alternatives to sodium nitrite

Due to increasing concerns over the long term harmful effects of sodium nitrite in developing gastrointestinal cancers, their use in cured meat products are stringently regulated in most developed countries and the use of other technologies and natural preservatives to prevent botulism is been explored to date.

4.1 Introduction

Antimicrobial properties of many spices have been firmly established in the literature, ironically they are often vectors for various pathogenic microorganisms specially spore-formers such as B. cereus and C. perfringens. Banerjee and Sarkar (2004) demonstrated that contaminated spices, including turmeric powder, when added to foods might allow the growth of pathogens and the subsequent toxin production.

Food poisoning by B. cereus and C. perfringens is much more common than one realizes, despite the vast knowledge available on them. Causes of food poisoning by these spore formers are usually due to insufficient cooling and reheating. Food processing does not always apply enough heat to destroy the spores. In addition, pasteurization actually activates the spores, triggering germination and enhances further vegetative cell multiplication (Andersson et al., 1995). Thus other methods for killing the spores are required to ensure food safety at home where temperature abuse is common.

171 Food poisoning outbreaks were estimated at 2 million cases in Australia and the cost was estimated at between $487 and $1900 million per year. The causes of the rises of foodborne diseases in Australia was attributed to a greater diversity of food available and changing consumer demand, as well as changes in food manufacturing, retail, distribution and storage. In addition, new food pathogens have emerged coupled with increasing numbers of susceptible individuals. Common food borne outbreaks in Australia involved B. cereus, Campylobacter species, C. perfringens, E. coli O111, L. monocytogenes and S. aureus. Typical food types involved in those outbreaks included meat, seafood, salad and vegetables and inadequate cooling and inappropriate holding temperature was to blame in most cases (Foodborne Disease Working Party, 1997).

4.2 Foodborne pathogens and food poisoning

4.2.1 Clostridium species

C. perfringens type A food poisoning is very common in the industrialized world producing enterotoxin (cpe) that breaks down normal plasma membrane permeability properties. The toxin producing gene cpe can be either chromosomal or plasmid-borne but only a small minority of global C. perfringens is cpe positive. C. perfringens can cause human gas gangrene and two types of food poisoning: type A diarrhea which is common and mild in industrialized countries and the very serious but rare type C human necrotic enteritis (Brynestad and Granum, 2002). C. perfringens is mainly found in meat products because of their requirement for 13 amino acids that they cannot synthesis and are found naturally in meat (Andersson et al., 1995). Hence spices and their essential oils have been evaluated for their anti-clostridia activities.

The temperature range for C. perfringens is between 33 and 49oC with the optimal temperature at 45oC. They can produce over 13 different toxins with four major lethal toxins used to type isolates (A-E). C. perfringens can grow from pH 5 to 9 with optimal pH between 6 and 7. Food poisoning outbreaks caused by C. perfringens often involves events where large quantity of food is cooked in advance followed by temperature abuse. However, since most illness is mild with short duration time, many cases are not been reported to healthcare professionals thus may underestimate the true number of C. perfringens outbreaks worldwide (Brynestad and Granum, 2002). Nevertheless, food poisoning caused by C. perfringens in the United States is estimated at around 250,000 cases annually, including about 41 hospitalization and seven deaths. Furthermore, the

172 economic impact associated with its outbreak was estimated to be around US$12.5 billion annually (Juneja et al., 2006b).

Juneja et al. (2006b) developed a mathematical model that predicts the growth pattern of C. perfringens spores in cooked cured ham at various temperatures and concluded that it can be used to design microbiologically safe cooling regimes for cured pork ham and similar products. This model may be modified for elucidating C. botulinum growth pattern and may indicate when it is best to add the antimicrobial agent during the fermentation or processing stage to achieve a maximum reduction of their spores and vegetative cells.

C. botulinum poisoning is known as botulism where the toxin binds to presynaptic nerve endings blocking acetylcholine secretion that inhibits muscle contraction. This results in double vision, uncontrolled salivation, blurred speech, and difficulty in swallowing. Urgent medical attention is necessary since it may lead to death as a result of respiratory musculature failure. Therapy consists of intravenous delivery of trivalent (types A, B and E) anti-toxin and intensive care. Recovery ranges from a few weeks to several months (Lindström et al., 2006). However, antibiotic resistance of Clostridium species in hospitals is increasing (Munro, 1989), which may become a probably in major food poisoning outbreaks especially in susceptible populations such as the elderly, immunocompromised patients and young children.

Since C. botulinum spores are widely distributed, they may enter food products through raw materials or by post-processing contamination. They are in abundance in the soil, hence fruits and vegetables can easily carry C. botulinum spores if they are not adequately washed. Most foodborne botulism outbreaks are actually the result of home- processed or home-canned foods. Outbreaks due to commercial products are rare but do occur sporadically. Different types of C. botulinum have different optimal growth conditions so can grow and produce toxins in different product groups that meet their growth requirements (Rhodehamel et al., 1992).

Strains of C. botulinum that can cause human botulisms belong to group I and II. Group I consists of proteolytic mesophilic organisms producing neurotoxins A, B and F that grows optimally at 35 -37 oC but not below 10 oC with growth limiting pH of around 4.4 and can tolerate up to 10 % NaCl concentrations. Group II consists of non- proteolytic psychrotrophic organisms producing type B, E and F toxins with an

173 optimum temperature of 26 -30 oC, but can grow from as low as 3 oC. Growth limiting pH for group II C. botulinum is 5.0 and inhibitory NaCl concentration is 5 %. Spores from group II are not as heat-resistant compared to group I, but they can still survive the heat processes encountered in the food industry, especially in ready-to-eat food (Lindström et al., 2006).

Due to the toxicity of C. botulinum, often C. sporogenes is used to evaluate the anti- Clostridia properties of herbs and spices due to their similar growth requirements. Mead (1985) concluded that tryptone-sulphite-cycloserine media with or without the addition of egg yolk are currently the best media for the purpose of enumerating Clostridium perfringens. In addition, this media can also be used to grow other Clostridia species such as C. sporogenes. However, the colonies produced are indistinguishable from C. perfringens, hence confirmatory tests are necessary.

4.2.2 Common bacterial food inhibitors

Various herbs and spices have been tested for their antibacterial properties. For example, the active antimicrobial compound in garlic is allicin, a diallyl thiosulfinate (2-propenyl-2-propenethiol sulfinate) that is colorless oil and provide the pungent odor and taste to garlic and onion. An intact garlic bulb contains a precursor to allicin that is alliin (S-allyl-L-cysteine-S-oxide). However, once the bulb is disrupted the enzyme allinases, a phosphopyridoxal enzyme, hydrolyses alliin to give allicin, pyruvate and ammonium. Fresh garlic extract at 10 % (w/v) was shown to inhibit the growth of Bacillus cereus on nutrient agar plates, and was germicidal at 5 % to Staphylococcus aureus and inhibitory at more than 2 % in liquid culture (Beuchat, 1994).

4.3 Food preservation

4.3.1 Traditional methods

The bactericidal effects of nitrite are due to its derivatives such as nitric oxide that is toxic to Clostridium and Listeria species. There are numerous possible targets for inhibition including respiratory chains, iron-sulfur proteins and other metalloproteins, membranes and gene expression (Cammack et al., 1999). However, increasing clinical and epidemiological evidence suggest that nitrite can increase the risk of developing gastrointestinal cancers, therefore other alternatives are been researched to date.

174 4.3.2 Recent methods

Most essential oils from aromatic plants and spices have GRAS status and some are routinely used in food industry mainly as flavoring or coloring compounds. However, recent studies have shown that many of these compounds also possess antimicrobial properties against a range of spoilage and pathogenic bacteria, yeasts, and moulds (Chaibi et al., 1997).

Jo et al. (2003) demonstrated that ionizing irradiation at 5 kGy significantly decreased volatile N-nitrosamines in cooked pork sausage as well as destroyed pathogenic and spoilage microorganisms without altering the nutritional and sensory qualities of the food. Similarly, Byun et al. (2002) demonstrated that gamma irradiation at 5 kGy was just effective as the use of 200 mg/kg sodium nitrite in ham without adversely affecting its color, texture or flavor. Furthermore, residual nitrite was at the lowest with carbon dioxide packaging. Therefore irradiation may be used commercially to reduce residual nitrite and N-nitrosamines during storage and increase the shelf life of cured meat products (Ahn et al., 2004). However, the use of irradiation technology is not accepted in all countries, or at least is used on case-by-case basis or for specific food groups. Therefore other methods are still necessary.

Using hurdle technology, which is the combination of intrinsic hurdles such as water activity and acidity, along with extrinsic hurdles such as processing temperature and storage temperature, to reduce the nitrite content in hot dogs without compromising product safety and quality (Jafari and Emam-Djomeh, 2007).

4.3.2.1 Alliaceae: Garlic and Onions

Freshly reconstituted dehydrated onion and garlic at 1 and 5 % (w/v), respectively, were shown to be bactericidal to S. typhimurium and E. coli 9002. However, E. coli was more resistant to onion than S. typhimurium (Johnson and Vaughn, 1969). Garlic and onion oils at 1500 µg/g of meat slurry partially inhibited toxin production by C. botulinum type A, but have minimal effect on type B and E (De Wit et al., 1979). It was concluded that garlic and onion oils should not be used for toxin inhibition in meat since meat can naturally contain several types of Clostridium species. Higher concentrations of garlic and onion oils should also be tested to evaluate their toxin

175 inhibition activity in meat. From Table 4.8, garlic had antimicrobial effects against a range of food borne pathogenic bacteria (Arora and Kaur, 1999).

4.3.2.2 Thymol

The major antimicrobial compound thymol (5-methyl-2-(1-methylethyl)phenol) from thyme, oregano, savory and sage is effective against a broad range of bacteria especially thyme oil. Alcoholic extracts of these spices including rosemary and turmeric can inhibit the growth and toxin production by C. botulinum at 500 µg/ml. Rosemary extract at 1000 µg/ml significantly inhibited the growth of S. typhimurium and S. aureus, and at 3000 µg/ml of either sage or rosemary inhibited the growth of 20 food borne Gram-positive bacteria, and was considered bactericidal at 5000 µg/ml (Beuchat, 1994). Thyme also showed the most inhibition against E. coli O157:H7 followed by myrtle and sage, whereas laurel showed no inhibition (Sağdiç et al., 2002). Similarly, from Table 11, oregano essential oil showed inhibition against E. coli O157:H7, S. aureus, S. enteritidis and L. monocytogenes at 4 % (w/v)(Seydim and Sarikus, 2006).

It was shown that the antimicrobial activity of sage increased with increasing water and salt content in food, possibly due to increasing sensitivity of bacteria and increasing essential oil partitioning in the liquid phase. On the contrary, the efficacy of sage as an antimicrobial agent decreased with increasing fat and protein content in foods, possibly due to the absorption of essential oils in the solid phase (Shelef et al., 1984). Since salt is often added to cured meat products, the combination of salt and other spices may provide synergistic action against pathogenic and spoilage bacteria in foods. However, because cured meat products are often high in fat and protein, sage alone might not be useful as an antibacterial agent for that specific product group.

4.3.2.3 Turmeric and curcuminoids

The yellow pigment in turmeric is composed of three types of curcumin and these are: curcumin (CUR I), mono-demethoxycurcumin (CUR II) and bis-demethoxycurcumin (CUR III)(Pfeiffer et al., 2003). These curcuminoids has been reported to possess antioxidant, anti-inflammatory, anti-angiogenic and anti-carcinogenic properties (Maheshwari et al., 2006). However, the stability of the curcuminoids and the formation of decomposition products and metabolites should be taken into account

176 when determining their physiological roles. The stability of different curcuminoid was demonstrated by Pfeiffer et al. (2003) using cell culture medium with and without fetal calf serum. It was concluded that all three curcuminoids decomposed very rapidly (more than 90 % within 12 h) when serum was omitted compared to with serum. Furthermore, CUR III was the most stable curcuminoid and CUR I was the least stable.

Turmeric is the powdered rhizome of a perennial herb called Curcuma longa L. often cultivated in tropical regions of Asia. It was shown in vitro and animal studies that turmeric possesses anti-inflammatory and anti-spasmodic activities as well as exerting protective effect on the liver and stimulates bile secretion aiding digestion. Since most of the yellow pigment curcumin (diferuloylmethane) in turmeric is excreted in the faeces, it may not be readily digestible and hence absorbed in the gastrointestinal tract (Ammon and Wahl, 1991). Hence other health benefits of turmeric may not be significant if it cannot be absorbed readily.

Furthermore, the doses required to achieve adequate tissue concentrations far exceeds the normal daily intake (Gescher et al., 2001). Other two major constituents in turmeric are derivatives of curcumin [p-hydroxycinnamoyl(feruloyl) methane and p,p’- dihydroxydicinnamoylmethane]. Depending on the origin and cultivation practices, turmeric can vary in essential oil (4.2 to 14 %), fatty oil (4.4 to 12.7 %) and moisture (10 to 12 %). The antibacterial effect of turmeric was contributed to curcumin or the essential oil (Ammon and Wahl, 1991). In addition, Shishu et al. (2003) demonstrated that dibenzoylmethane, a structural analogue of curcumin, was inhibitory against seven cooked food heterocyclic amines.

4.3.2.4 Other herbs and spices

Other less popular or non-culinary herbs and spices has been tested for their anti- botulinal properties, including cedar, Eucalyptus, (Chaibi et al., 1997) and mastic resin (Daifas et al., 2004). It was suggested that the inhibitory effect of spice oil was similar to phenols by affecting the permeability, transport system, electron transport and energy production of bacteria (Ismaiel and Pierson, 1990b). There may be commercial applications for these spice oils in certain foods since traditional preservative sodium nitrite requires more than 20,000 ppm in order to inhibit the germination of C. botulinum.

177 When nitrite was heated in a bacteriological medium, more inhibitory effect was shown against C. sporogenes than nitrite added aseptically to the medium after autoclaving. The unidentified inhibitory substance became known as Perigo factors. It was shown that reducing agents such as ascorbate was necessary to produce the effect. In a meat- less medium at 110 oC heated for 20 minutes containing 20 mg/kg nitrite showed enhanced inhibition of C. botulinum, whereas in a meat-based medium containing 150 mg/kg nitrite the inhibition effect was lessened. Therefore meat was found to neutralize the inhibitory factor (Cammack et al., 1999).

Huhtanen (1980) tested 33 spices for anti-botulinal properties and concluded that mace and achiote were the most effective with MIC of 31 µg/ml. However, achiote is not used in food due to its suspected carcinogenicity. Other effective spices with MIC of 125 µg/ml were: black and white pepper, bay leaf and nutmeg. Spices with MIC of 500 µg/ml were: paprika, rosemary, cloves, oregano, thyme and turmeric. The justification of testing these spices was because many cured meat and sausage products use one or more of the spices tested. However, due to trade confidentiality, the exact quantity used is not known and varies from manufacturers and the type of product. One major flaw in their experimental design is that inhibition was concluded from test-tube turbidity only without selective plate count, thus any contamination might give false negative results.

Chaibi et al. (1997) tested nine essentials oils for their antibacterial activity against B. cereus and C. botulinum, both spore formers. The most effective essential oil against the spores of B. cereus was cedar with MIC of 100 ppm, followed by rosemary at 200 ppm, then eucalyptus and orange at 250 ppm. For inhibiting C. botulinum spores, cedar was again the most effective but at 300 ppm, followed by Savage carrots and Vervain at 350 ppm. Eucalyptus, chamomile and orange also showed inhibition at 400 ppm. The oil fraction of turmeric was shown to be inhibitory towards B. cereus, S. aureus, and E. coli (Beuchat, 1994). These essential oils may have application in minimally processed food to extend shelf life and improve safety. However, the antibacterial activity of essential oils against the spores of B. cereus and C. botulinum need to be tested in food systems taking into account of the interaction with food matrices.

Daifas et al. (2004) demonstrated that mastic resin in ethanol (8 % w/w) and mastic oil (0.3% v/v) inhibited the growth of C. botulinum type A and B. In addition, inhibition was strain specific with C. botulinum type A strains more sensitive to mastic resin and

178 its essential oil than type B strains. However, mastic resin and mastic oil did not inhibit neurotoxin production in English-style crumpets. Even though bakery products are a suitable substrate for the growth of C. botulinum, there is yet botulism outbreak associated with bakery products to date. Hence further testing of mastic resin and its essential oil on neurotoxin production by C. botulinum should be carried out in meat systems instead.

The addition of grapefruit extract Citricidal at 200 ppm significantly reduces the growth of C. perfringens in marinated, vacuum-packed chicken. However, since 1.0 % NaCl was also added to the chicken, the antibacterial activity of Citricidal may be contributed to salt that works synergistically perhaps by increasing the sensitivity of C. perfringens to Citricidal. From the consumer perspective, Citricidal did not have any impact on the sensory aspect of the product (Juneja et al., 2006a).

It was demonstrated by Ismaiel and Pierson (1990a) that the oils of black pepper and clove had significantly higher inhibition of vegetative growth by C. botulinum 67B than any other oils such as garlic, onion, cinnamon, thyme and origanum. However, at concentrations between 100 and 150 ppm all the oils prevented germination of C. botulinum 67B. It was suggested that the inhibitory effect of spice oil was similar to phenols by affecting the permeability, transport system, electron transport and energy production of bacteria. There may be commercial applications for these spice oils in certain foods since traditional preservative sodium nitrite requires more than 20,000 ppm in order to inhibit germination. Similarly, Ismaiel and Pierson (1990c) shown that clove, thyme, black pepper, pimento, origanum, garlic, onion and cinnamon oil at 200 ppm inhibited the growth and germination of C. botulinum 33A, 40B and 1623E.

Ismaiel and Pierson (1990b) demonstrated that at 200 ppm origanum oil was effective at inhibiting C. botulinum growth in TYG broth. However, in meat system the effective concentration increased to 400 ppm in combination with sodium nitrite at 50 to 100 ppm.

Clove oil at 1 to 2 % reduced the initial population of L. monocytogenes in chicken Frankfurts (Table 4.7). However, the inhibition was strain dependent (Mytle et al., 2006). By combining with other herbs and spices such as garlic, it may lead to complete growth inhibition. Clove powder was very bacteriocidal at 0.5 to 1.0 % against S. enteritidis and E. coli O157. Garlic at 0.5 to 1.0 % showed greater inhibition

179 to E. coli O157 than S. enteritidis. In addition, garlic at 0.5 to 1.0 % showed bacteriostatic and bacteriocidal effects against S. enteritidis in broth system. Ground dried ginger and mustard powder showed little antimicrobial activity against either bacteria (Leuschner and Zamparini, 2002).

From Table 4.2, Gram-positive bacteria were more sensitive to sumac spice compared to Gram-negative bacteria. Of the Gram-positive strains, L. monocytogenes showed highest resistance whereas Bacillus species were the least resistant. It was demonstrated that 1 % sumac extract resulted in 4-5 log reduction in Bacillus species and 2-3 log reduction in other bacteria tested including S. aureus, L. monocytogenes, E. coli O157:H7 and S. enteritidis (Nasar-Abbas and Halkman, 2004).

From Table 4.3, using agar dilution assay method, Viola tricolor herb gave a significant log reduction in S. aureus, B. cereus, S. epidermidis, with moderate inhibition against P. aeruginosa, E. faecalis, E. coli and K. pneumoniae (Witkowska-Banaszczak et al., 2005).

From Table 4.4, at 0.1 %, nutmeg extract showed significant reduction of E. coli O157 strains compared to non-pathogenic E. coli strains. At 0.1 %, allspice completely inhibited the growth of both pathogenic and non-pathogenic E. coli strains, whereas at the same concentration, thyme completely inhibited growth for all the pathogenic E. coli strains and inhibited growth for most of the non-pathogenic E. coli strains. Star anise showed complete inhibition for most of the pathogenic and non-pathogenic E. coli strains at 1 %, whereas cinnamon did not show antibacterial properties (Takikawa et al., 2002).

From Table 4.9, all five Australian native herb extracts exhibited antibacterial properties in the concentration range from 7.8 to 500 µg/ml. S. aureus was the most sensitive whereas E. coli and L. monocytogenes were the most resistant (Dupont et al., 2005). These native herbs are often used in traditional Australian Aboriginal tucker or bush foods. Their consumption at low levels appears to be safe for humans, and therefore should be exploited in commercial foods such as cured meat products. From Table 4.10, every Turkish spice extracts tested showed some inhibition against E. coli O157:H7 with the exception of laurel extracts. Thyme gave the largest inhibition zone whereas cumin gave the smallest inhibition zone (Sağdiç et al., 2002).

180 4.3.2.5 Other methods

Sprong et al. (2002) demonstrated in vitro and in vivo using male Wistar rats that bovine milk fat tryglycerides containing C10:0, C12:0 and C18 together with digestion products of sphingolipids were effective bactericidal agents against L. monocytogenes and C. jejuni, whereas E. coli O157:H7 and S. enteritidis were less susceptible. It was suggested that Gram-positive bacteria are more susceptible to fatty acids and monoglycerides due to the lack of protective lipopolysaccharide-rich outer membrane found in Gram-negative bacteria. However, moderate to high dairy consumption in humans does not necessarily result in less foodborne illnesses, but may be added as a natural food preservative. Furthermore, excess consumption of dairy food may result in other harmful effects such as weight gain.

Using media with inoculated C. botulinum spores, it was recently suggested that fat may reduce the efficacy of some antimicrobials added to or found naturally in foods including free fatty acids, sorbic acid and potassium sorbate. It was suggested that many food preservatives are lipophilic acids, which interacts with the phospholipids within the bacterial cell membrane thus interfering with substrate transport across the membrane. Since fats naturally occur in meat products it may compete for the llipophilic molecule thereby reducing the concentration available for antimicrobial activities. Fat had less effect on the antimicrobial activity of amphiphilic compounds such as phosphates, or water-soluble antimicrobials such as sodium nitrite, thus its anti- botulinal activity is only slightly affected (Glass and Johnson, 2004).

Østerlie and Lerfall (2005) shown that lycopene from tomato products can increase shelf life and decrease rancidity in minced meat products as well as providing a stable reddish colour during storage. It was suggested that the antibacterial effect of tomato products was due to the lowering of pH and consequently reducing microbial growth and thus increases shelf life. However, the minced meat mixtures also contained salt and various spices such as white pepper, allspice, ginger and nutmeg, which may contribute to increased shelf life.

Pediococcus damnosus and Pediococcus pentosaceus was shown to inhibit the growth of C. perfringens, L. monocytogenes, Salmonella infantis, and Yersinia enterocolitica in minced meat juice. The antibacterial effect was attributed to the production of a bacteriocin called pediocin A. It was suggested that Pediococcus species may be used

181 as a new preservative systems and starter culture for food fermentation processes (Mattila-Sandholm et al., 1991). It would be necessary to clone the gene for pediocin A to ensure a consistent production of the bacteriocin at various and sometimes extreme conditions encountered in the food fermentation and processing stage. The effects of Pediococcus and pediocin A to the food sensory aspects as well as safety for human consumption are major issues that need to be evaluated.

Depending on the sensory aspect of using combined herbs and spices as an anti- botulinal agent in cured meat products, its concentration may be reduced if combined with hurdle technologies as demonstrated by Jafari and Emam-Djomeh (2007).

4.4. Antibacterial properties of herbs and spices

Summary of recent evaluation of various herbs, spices and essential oils on their antimicrobial properties against common food borne pathogenic bacteria.

Table 4.1 MIC range for various bacteria in vitro using curcumin or turmeric oil.

Bacteria Turmeric/curcumin MIC References

Streptococcus Curcumin 0.5 – 5.0 mg/ml Ammon and Wahl, Essential oil 5.0 – 100 µg/ml 1991 Bacillus Curcumin 0.5 – 5.0 mg/ml Ammon and Wahl, Essential oil 5.0 – 100 µg/ml 1991 Staphylococcus Essential oil 5.0 – 100 µg/ml Ammon and Wahl, 1991 Staphylococcus Curcumin 2.5 -50 mg/ml Bhavani Shankar and Sreenivasa Murphy, 1979 Staphylococcus Turmeric oil 0.45 – 0.9 µl/ml Bhavani Shankar and suppresses and Sreenivasa growth of Murphy, 1979 Lactobacilli Bacillus Turmeric oil 0.05 – 0.2 mg/ml Negi et al., 1999 Staphylococcus E. coli Turmeric oil > 0.2 mg/ml Negi et al., 1999 Pseudomonas Turmeric oil > 0.2 mg/ml Negi et al., 1999 aeruginosa

182 Table 4.2 MIC on selected food pathogenic bacteria using sumac spice (Nasar-Abbas and Halkman, 2004).

Gram test Bacteria MIC (%)

Positive Bacillus spp. 0.25-0.32 S. aureus 0.49 L. monocytogenes 0.67 Negative C. freundii 0.42 H. alvei 0.45 P. vulgaris 0.55 E. coli O157:H7 0.60 E. coli type I 0.63 S. enteritidis 0.67

Table 4.3 MIC and MBC on selected food pathogenic bacteria using Viola tricolor herb (Witkowska-Banaszczak et al., 2005).

Gram test Bacteria MIC (mg/ml) MBC (mg/ml)

Positive S. aureus 0.15 0.15 S. epidermidis 0.15 0.31 B. cereus 0.15 0.15 E. faecalis 5.0 5.0 Negative E. coli 2.5 5.0 P. aeruginosa 1.25 5.0 K. pneumoniae 1.25 1.25

Table 4.4 MIC of various spices including allspice, clove, thyme, nutmeg and star anise on selected Escherichia coli strains (Takikawa et al., 2002).

Gram test Bacteria MIC (%) Spices

Positive E. coli O157 P74 0.1 Allspice, thyme, nutmeg, star anise E. coli O157 P98 0.1 Allspice, thyme, nutmeg, E. coli O157 P100 0.1 Allspice, thyme, nutmeg, star anise E. coli N1 0.1 Allspice, thyme, clove E. coli N2 0.1 Allspice, thyme, star anise E. coli N4 0.1 Allspice, thyme, star anise

183 Table 4.5 Averaged MIC of classically or high-intensity ultrasound-extracted spices including ginger, fingerroot and turmeric on various strains of Listeria and Salmonella (Thongson et al., 2004).

Gram test Bacteria MIC (%) v/v Spices

Positive L. monocytogenes 2.8 Ginger Solvent extraction 0.048 Turmeric 0.076 Fingerroot Negative Salmonella 6 Ginger Typhimurium HI-US extraction 7 Turmeric HI-US extraction 6.25 Fingerroot

Table 4.6 Growth and survival of Escherichia coli O157 and Salmonella enterica serovar Enteritidis in broth model systems and mayonnaise (Leuschner and Zamparini, 2002).

Bacteria Spice Concentration of Significant log spice (%) reduction

S. enterica Clove 1 Yes (broth) Ginger 1 No Mustard 1 No Garlic 0.5-1 Yes (mayonnaise) E. coli O157 Clove 0.5-1 Yes (broth) Ginger 1 No Mustard 1 No Garlic 0.25-1 No

Table 4.7 Antibacterial activity of clove oil on Listeria monocytogenes in chicken frankfurters (Mytle et al., 2006).

Bacteria Clove oils concentration Significant log reduction (% v/w)

L. monocytogenes (7 1 to 2 Yes (dependent on strains, strains) initial population and storage temperature)

184 Table 4.8 Antimicrobial activities of garlic and clove on selected food borne pathogenic bacteria (Arora and Kaur, 1999).

Bacteria Zone of inhibition on NA plates (mm) Garlic Clove B. sphaericus 19.3 0 E. aerogenes 15.6 0 E. coli 20 0 P. aeruginosa 20 0 S. aureus 20 0 S. epidermidis 20.3 0 S. flexneri 30 17.5 S. typhi 21.3 0

Table 4.9 MIC of Australian native herb extracts against common food-related bacteria using microtitre broth microdilution assay (Dupont et al., 2005).

Herbs and Anetholea Backhousia Eucalyptus Eucalyptus Prostanthera MIC (µg/ml) anisata citriodora olida staigerana incisa

Solvent Ethanol Hexane Ethanol Hexane Ethanol Bacteria E. faecalis 250 125 125 62.5 62.5 E. coli 125 >500 62.5 250 500 L. 500 250 500 62.5 125 monocytogenes P. aeruginosa 125 250 250 31.3 62.5 S. Enteritidis 250 250 125 31.3 125 S. Typhimurium 250 250 125 125 125 S. aureus 7.8 125 15.6 62.5 15.6

Table 4.10 MIZ of Turkish spice extracts to the growth of E. coli O157:H7 (Sağdiç et al., 2002).

Spice extracts at 1:1 dilution MIZ (mm) Control (absolute alcohol) 0 Cumin 17 Helichrysum compactum Boiss 28 Laurel 0 Myrtel 35 Oregano 30 Sage 32 Thyme 42

185 Table 4.11 MIZ for oregano, garlic, and rosemary essential oil against common foodborne pathogenic bacteria (Seydim and Sarikus, 2006).

Inhibitory zone (mm) Bacteria Concentration Oregano Garlic Rosemary % (w/v) E. coli 4 37.1 11.4 0 O157:H7 S. aureus 4 43.1 13.45 0 S. enteritidis 4 40.6 10.5 0 L. 4 41.7 12.0 0 monocytogenes L. plantarum 4 13.5 9.2 0

Table 4.12 MIC for nitrite and its common derivatives or complexes on the growth of C. sporogenes and L. monocytogenes (Cammack et al., 1999).

Nitrite and Ki (µmol/L) derivatives/complexes L. monocytogenes C. sporogenes NO2 NA 5000 NO NA 40 NH4[Fe4S3(NO)7] RBS 3 1.3 [Fe2(SCH2CH2OH)2(NO)4] 45 5 RRSE Fe4S4(NO)4 NA 10 Na2[Fe(CN)5NO] SNP Non-toxic 28

Concentrations expressed as amount required to achieving a 50 % inhibition of growth in liquid culture under anaerobic conditions (RBS- RRSE-Roussin’s red salt ester, SNP. NA-not available).

The molecular mechanism of phytochemicals as chemopreventive agent was elucidated by Surh et al. (2001). Cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase (iNOS) are important enzymes that mediate inflammatory processes and improper up- regulation of these enzymes has been associated with pathophysiology of certain types of human cancers and associated inflammatory disorders. Since inflammation is closely linked to tumour promotion, substances that possess anti-inflammatory activities such as curcumin, green tea EGCG and grapes resveratrol are suggested to delay or prevent carcinogenesis by suppressing the activation of eukaryotic

186 transcription factor nuclear factor-kappa B (NF-κB) that is involved in the regulation of COX-2 and iNOS expression.

Hahm et al. (2002) synthesized 12 symmetrical curcuminoids with nearly 90 times higher efficiency than curcumin in inhibiting the formation of Fos-Jun-DNA complex, which are cellular-oncogene products. The anti-inflammatory and anti-carcinogenic properties of curcumin may be contributed to their ability to induce the enzymes involved in the detoxification of the electrophilic products of lipid peroxidation (Piper et al., 1998).

In addition to their anti-microbial activities, many spices also provide other health benefits. For example, it was shown in animal experiments that turmeric, garlic, onion and pepper have one or more of the following benefits: anti-hypercholesterolemic effect, anti-lithogenic effect, anti-diabetic effect, antioxidant and anti-inflammatory activities and stimulate the digestive system (Srinivasan, 2005). Although above average daily intake is required to show some of these benefits, small doses over a long-term period may also be effective especially with some of the synergistic effect shown by the spices

4.5 Combinations of commercially available herbs and spices to prevent food poisoning and botulism in vitro and possible commercial application in cured meat products

4.5.1 Introduction

Food preservation has been practiced for centuries and the methods used have improved shelf life and safety significantly especially in recent years due to advancing knowledge and technology. Currently, numerous natural antimicrobial agents have been used or tested including lactoperoxidase, lactoferrin, lysozyme, ovotransferrin, avidin, nisin, bacteriocins, and plant and spice essential oils. The growing interests in using natural antimicrobials as food preservatives is due to the short and/or long term side-effects as a result of consuming chemical food preservatives over a period of time (Naidu, 2000). For example, sodium nitrite is used in cured meat products to prevent botulism food poisoning. However, it was shown that nitrite in the stomach can form a group of potent carcinogens known as NOCs (Bartsch, 1990). Epidemiological studies, clinical and animal experiments had demonstrated the increased risk of gastrointestinal

187 cancers such as stomach cancer in countries where high intake of nitrite is common in the diet (Jossens et al., 1996).

Clostridium species thrive in meat products because of their requirement for 13 amino acids that they cannot synthesize and their spores can be found in meat products and multiply when the conditions are optimal for their growth (Andersson et al., 1995). The economic impact associated with Clostridia related outbreak was estimated to be around US$12.5 billion annually (Juneja et al., 2006a). Furthermore, many cases of food poisoning outbreaks caused by C. perfringens are not reported due to its mild and short incubation period (Brynestad and Granum, 2002) and was estimated at around 250,000 cases annually in the United States, including about 41 hospitalization and seven deaths. It is therefore in the best interests of the food provider and the public to ensure the safety of the food due to the financial burden as a result of a food poisoning outbreak and the possible lethal outcome in case of botulism food poisoning.

Most food borne botulism outbreaks are actually the result of home-processed or home- canned foods. Outbreaks due to commercial products are rare but do occur sporadically. C. botulinum and their sub-types have different optimal growth conditions and their spores can contaminate food and grow and produce toxins when conditions are right (Rhodehamel et al., 1992). Most essential oils from aromatic plants and spices have GRAS status and some are routinely used in food industry mainly as flavoring or coloring compounds. However, recent studies have shown that many of these compounds also possess antimicrobial properties against a range of spoilage and pathogenic bacteria, yeasts, and moulds (Chaibi et al., 1997).

Therefore it was the aim of this study to determine, using an in vitro method, the minimum inhibitory concentrations of readily available herbs and spices, singly and in combination, on the growth inhibition against Clostridium sporogenes (as a surrogate for C. botulinum) as well as other common food borne pathogenic bacteria.

4.5.2 Materials and Methods

The minimum inhibitory concentration was demonstrated in vitro using herbs and spices either singly or in different combinations against common food borne bacteria and C. sporogenes.

188 4.5.2.1 Bacterial cultures

All bacterial strains tested were provided by the Microbiological Culture Collection at the School of Biotechnoloy and Biomolecular Sciences at University of New South Wales. Detail of strains: B. cereus UNSW011100T (CSIRO Food Research), C. sporogenes UNSW005600 (School of Medicine, Universitry of Sydney), E. coli UNSW002500, L. monocytogenes UNSW 028700, S. enteritidis UNSW031901, and S. aureus UNSW002100 (Prince of Wales Hospital).

4.5.2.2 Reagents and media

Chemical grade sodium nitrite from Univar was used. Nutrient agar and broths, peptone water, re-enforced Clostridial agar and medium were supplied by Oxoid and were made up according to instructions provided by the manufacturer.

4.5.2.3 Herbs and spices

Different brands of dried packaged herb and spices as well as fresh raw garlic were purchased from supermarkets. The garlic was peeled then blended prior to use. Herb included rosemary, sage, thyme and spices included grounded chili, ginger, and turmeric and analytical grade curcumin from Sigma catalogue number C1386. Herbs and spices were extracted at 60 oC in nutrient broths.

4.5.2.4 MIC to common food borne pathogenic bacteria

Herbs and spices were made up to 1.0, 3.0 and 5.0 % (w/v) either individually or in combination at 50:50 in 9.0 ml nutrient broth tubes. These were then UV sterilized for 20 minutes then either incubated at 60 oC for 30 minutes or autoclaved at 121 oC for 15 minutes. Aliquots of 0.1 ml were taken out and streaked onto nutrient agar plates and incubated at 37 oC for 24 hr to ensure no contaminations. One ml of each test bacteria was transferred into four nutrient broth tubes with three different concentrations and one as control without the test substance. All nutrient broth tubes were then incubated at 37 oC for 24 hours then 0.1 ml of the aliquot were taken out and serially diluted with sterilized peptone water and 0.1 ml of each was added to duplicate nutrient agar plates and spread with glass rod and incubated at 37 oC for 24 hr for total live bacteria count. At 48 hrs further samples were taken for dilution and spread plate. Pure bacterial

189 cultures were re-streaked every week and grown in nutrient broths with the exception of C. sporogenes.

Similarly for C. sporogenes, herb and spices were made up to three concentrations as above except in 9.0 ml reinforced Clostridial broth tubes. Incubation and extraction conditions were the same as above except Clostridial broth tubes were sealed and Clostridial plates were incubated in anaerobic jars. In addition, 5.0 and 10.0 % (w/v) of sodium nitrite was made up for assessing the growth inhibition efficiency of the test substances. All reinforced Clostridial agar plates were incubated at 37 oC for 24 hours in anaerobic jars with nutrient agar plates incubated aerobically under the same condition. Pure cultures were re-streaked every week and grown on reinforced Clostridial agar plate in anaerobic jars.

4.5.3 Results and discussion

Various herbs and spices tested showed total growth inhibition between 0.1 and 5.0 % (w/v) against common food borne bacteria within 72 hr (Table 4.13). Table 4.13 showed the MIC for all the bacteria tested and positive herbs and spices used. Results for sage, thyme and curcumin were not show since no inhibition was demonstrated.

190 Table 4.13 MIC (w/v) of common herbs and spices individually or in combination (50:50) against common food borne pathogenic bacteria in vitro up to 72 hr.

Bacteria Treatment Herb/Spice MIC (w/v) Bacillus cereus Autoclave Turmeric 0 Turmeric:garlic 3.0 Hot water Curcumin 0 Raw garlic 0 Turmeric:chili 0 Turmeric 0.1 Turmeric:garlic 1.0 Turmeric:ginger 1.0 Rosemary 1.0 Escherichia coli Hot water Curcumin 0 Turmeric 0 Turmeric:ginger 0 Turmeric:chili 0 Turmeric:garlic 3.0 Raw garlic 5.0 Listeria monocytogene Autoclave Turmeric 0 Turmeric:garlic 1.0 (48hr) Hot water Raw garlic 5.0 Turmeric:garlic 3.0 Turmeric:ginger 1.0 (48hr) Rosemary 1.0 Salmonella enteritidis Autoclave Turmeric 2.5 (72hr) Hot water Curcumin 0 Raw garlic 5.0 Turmeric:garlic 5.0 Staphylococcus aureus Autoclave Turmeric:garlic 1.0 (48hr) Hot water Turmeric:chili 0 Raw garlic 5.0 Turmeric:garlic 3.0 Turmeric:ginger 3.0 Rosemary 1.0

The combination of herbs and spices may enhance their antimicrobial properties by synergistic actions. For example, autoclaved turmeric and hot water extracted raw garlic alone did not inhibit the growth of B. cereus (Table 4.13), but when combined with garlic it showed growth inhibition at 3.0 % (Table 4.13). Furthermore, hot water extracted turmeric and garlic showed growth inhibition at lower concentration of 1.0 % (Table 4.13), suggesting synergistic antibacterial activities. However, high temperature can result in loss of active agents in some common spices (Srinivasan et al., 1992),

191 which may explain the lack of antibacterial properties in some of the herbs and spices tested (Table 4.13). Turmeric combined with ginger was also effective at 1.0 %, as well as rosemary alone (Table 4.13). In contrast Beuchat (1994) demonstrated that raw garlic extract at 10.0 % (w/v) inhibited the growth of B. cereus on nutrient agar plates, and was germicidal at 5.0 % against S. aureus. In this study fresh garlic at 5.0 % only reduced the growth of B. cereus (result not shown), which maybe due to different sub- species, but at 5.0 % it did inhibit the growth of S. aureus (Table 4.13).

From Table 4.13, MIC for curcumin against tested bacteria was less than 0.5 % (w/v) and the required concentration was much less for its essential oils. In general higher concentration of turmeric oil was needed may be because curcumin was the active antibacterial agent. However, turmeric oil must contain other antibacterial agents to display inhibition activities. Autoclaved and hot water extracted curcumin did not show any inhibition (Table 4.13), this is most likely because the active agent can only be extracted in ethanol and may be destroyed at higher temperatures.

Hot water extracted turmeric and garlic required higher MIC against L. monocytogenes than autoclaved turmeric and garlic. Suggesting that higher temperature was required to release the active antibacterial agents. However the latter required longer time before growth inhibition was demonstrated (Table 4.13). Rosemary was also effective at just 1.0 % and raw garlic at 5.0 % (Table 4.13). Extraction was not required in the case of fresh garlic where the antibacterial agent allicin is released upon crushing.

From Table 4.5, it was demonstrated that fingerroot or Chinese ginger extracted with isopropanol-hexane and without HI-US had the highest anti-Listerial effect whereas HI-US-isopropanol fingerroot extract had the greatest inhibition against Salmonella Typhimurium (Thongson et al., 2004). Inhibition against L. monocytogenes was also demonstrated with turmeric combined with ginger at 1.0 % (Table 4.13).

Only hot water extracted raw garlic and turmeric with garlic show growth inhibition against E. coli at 5.0 and 3.0 %, respectively (Table 4.13). Similarly, only hot water extracted raw garlic and turmeric with garlic at 5.0 % was effective at inhibiting the growth of S. enteritidis (Table 1). MIC for S. aureus in hot water extracted raw garlic and rosemary were 5.0 and 1.0 %, respectively (Table 4.13), whereas the MIC for turmeric with garlic and turmeric with ginger were both at 3.0 % (Table 4.13).

192 Farbood et al. (1976) demonstrated bactericidal effect of rosemary extract to S. aureus at 5 % concentration in various meat media. However, there were no effects against other bacteria tested such as E. coli and S. typhimurium. The lack of antimicrobial activity in meat systems was suggested to be due to the lack of rosemary extract in the aqueous phase because of absorption into the solid phase by lipid and protein that are naturally occurring in meat.

Most antibacterial properties of herbs and spices tested here were lost when autoclaved suggesting that the active antibacterial agent is heat sensitive (Table 4.13). However, the antibacterial agent in turmeric and garlic was able to work after 48 to 72 hr against L. monocytogenes, S. enteritidis, and S. aureus (Table 4.13).

Hot water extracted turmeric with garlic required higher MIC against L. monocytogenes at 3.0 % than autoclaved turmeric with garlic at only 1.0 % (Table 4.13). However the latter required a longer time (48 hr) before inhibition of growth was demonstrated (Table 4.13). It could mean that heat may activate or increase the activity of the antibacterial agent in turmeric and garlic. Rosemary was effective at 1.0 % against L. monocytogenes and S. aureus (Table 4.13). In contrast, Seydim and Sarikus (2006) concluded that rosemary essential oil exhibited no inhibitory effect against S. aureus, S. enteritidis, L. monocytogenes, and E. coli. This suggests that the antibacterial activity in rosemary is not just the essential oil; the activity may be associated with other compounds present in the leaves, perhaps in combination with the oil.

In was demonstrated here that hot water extracted raw garlic at 5.0 % was effective at inhibiting the growth of S. enteritidis (Table 4.13). The higher concentration used here resulted in complete inhibition, whereas at 1.0 % used in mayonnaise only gave 10 log growth reduction (Leuschner and Zamparini, 2002). The delayed growth inhibition may also be due to other ingredients present in mayonnaise such as salt. In addition, garlic concentration was reduced to 2.5 % when combined with turmeric at 2.5 % (Table 4.13), suggesting synergistic actions between the two spices.

193 Table 4.14 MIC (w/v) of common herbs and spices individually or in combination (50:50) against Clostridium sporogenes at 24 hr.

Treatment Herb/Spice MIC (w/v) Control Sodium nitrite 5.0 Autoclave Turmeric 0.3 Water bathed Turmeric:ginger 0.5 Water bathed Rosemary 1.0 Reinforced Clostridial broth media and anaerobic jar was used.

From Table 4.14, all nitrite controls showed no growth at 5 %. From Table 4.12, the MIC of nitrite and its derivatives against C. sporogenes was demonstrated (Cammack et al., 1999). Autoclaved turmeric at 0.3 % inhibited the growth of C. sporogenes. The MIC of hot water treated turmeric with ginger and rosemary were 0.5 and 1.0 %, respectively.

Meats and meat products can naturally contain several Clostridium species (De Wit et al., 1979); therefore one must ensure that all are controlled, which is the major gap in all current literature. The effectiveness of herbs and spices as anti-clostridial agents will vary. Hence it is important to find the right combination to prevent botulism that is also acceptable from sensory aspects.

Autoclaved turmeric at 0.3 % inhibited the growth of C. sporogenes. The MIC of hot water extracted turmeric with ginger and rosemary alone were 0.5 and 1.0 %, respectively (Table 4.14). The anti-botulinal properties of turmeric and rosemary were demonstrated in alcoholic extracts of the spices at only 0.05 % (w/v) (Beuchat, 1994), and Ismaiel and Pierson (1990b) shown that garlic oil at 0.2 % (w/w) inhibited the growth and germination of three specific strains of C. botulinum. In all cases the concentration may increase once applied in meat models due to distribution and interactions with food matrices.

Australia has a diverse range of flora including herbs and spices used in traditional bush foods; some have been shown to possess antibacterial properties (Dupont et al., 2005), which should be explored for anti-Clostridial activity. However, the use of these novel antimicrobial agents in food requires further research in terms of their safety and acceptability in the final food product and their effects on nutrients and other food components needs further assessment.

194 The justification of testing these spices was because many cured meat products uses one or more of the spices tested. Furthermore, the synergistic effects of combined herbs and spices have not been explored, specifically as an anti-Clostridia agent in cured meat products. The availability and inexpensiveness of herbs and spices may also be useful for preventing food poisoning in home made cured products, where botulism is common.

4.5.4 Conclusion

Out of the herbs and spices tested alone or in combination showed some degree of in vitro antibacterial activity against food borne pathogenic bacteria including C. sporogenes. The use of moderate heat as the extraction step generally minimized the loss of the active compounds in herbs and spices. In addition, the inexpensiveness and wide availability of the herbs and spices may be used to prevent food poisoning in home made food where botulism is very common. However, the antibacterial properties of combined herbs and spices should be tested in meat models and evaluated for their suitability in the final product in order to replace nitrite and minimize dietary nitrite intake and indirectly reduce the risk of developing gastrointestinal cancers in susceptible populations.

195 4.6 References

Ahn, H.J., Kim, J.H., Jo, C., Lee, J.W., Yook, H.S., and Byun, M.W. 2004. Effects of gamma irradiation on residual nitrite, residual ascorbate, color, and N-nitrosamines of cooked sausage during storage. Food Control, 15:197-203. Ammon, H.P.T., and Wahl, M.A. 1991. Pharmacology of Curcuma longa. Planta Medica, 57:1-7. Andersson, A., Rönner, U., and Granum, P. E. 1995. What problems does the food industry have with the spore-forming pathogens Bacillus cereus and Clostridium perfringens?. International Journal of Food Microbiology, 28:145-155. Arora, D.S., and Kaur, J. 1999. Antimicrobial activity of spices. International Journal of Antimicrobial Agents, 12:257-262. Bartsch, H., Ohshima, H., Shuker, D., Pignatelli, B., and Calmels, S. 1990. Exposure of humans to endogenous N-nitroso compounds: implications in cancer etiology. Mutation Research, 238:255-267. Banerjee, M., and Sarkar, P. K. 2004. Growth and enterotoxin production by sporeforming bacterial pathogens from spices. Food Control, 15:491-496. Beuchat, L.R. 1994. ‘Antimicrobial properties of spices and their essential oils’ in Natural Antimicrobial Systems and Food Preservation. eds. V.M. Dillon and R.G. Board. CAB International, United Kingdom. Bhavani Shankar, T.N., and Sreenivasa Murphy, V. 1979. Effect of turmeric (Curcuma longa) fractions on the growth of some intestinal and pathogenic bacteria in vitro. Indian Journal of Experimental Biology, 17:1363-1366. Brynestad, S., and Granum, P.E. 2002. Clostridium perfringens and foodborne infections. International Journal of Food Microbiology, 74:195-202. Byun, M.W., Lee, J.W., Yook, H.S., Lee, K.H., and Kim, H.Y. 2002. Improvement of shelf stability and processing properties of meat products by gamma irradiation. Radiation Physics and Chemistry, 63:361-364. Cammack, R., Joannou, C.L., Cui, X-Y., Martinez, C.T., Maraj, S.R., and Hughes, M.N. 1999. Nitrite and nitrosyl compounds in food preservation. Biochimica et Biophysica Acta, 1411:475-488. Chaibi, A., Ababouch, L.H., Belasri, K., Boucetta, S., and Busta, F.F. 1997. Inhibition of germination and vegetative growth of Bacillus cereus T and Clostridium botulinum 62A spores by essential oils. Food Microbiology, 14:161-174. Daifas, D. P., Smith, J.P., Blanchfield, B., Sanders, G., Austin, J. W. and Koukoutisis, J. 2004. Effects of mastic resin and its essential oil on the growth of proteolytic Clostridium botulinum. International Journal of Food Microbiology, 94:313-322. De Wit, J.C., Notermans, S., Gorin, N., and Kampelmacher, E.H. 1979. Effect of garlic oil or onion oil on toxin production by Clostridium botulinum in meat slurry. Journal of Food Protection, 42:222-224.

196 Dupont, S., Caffin, N., Bhandari, B., and Dykes, G.A. 2005. In vitro antibacterial activity of Australian native herb extracts against food-related bacteria. Food Control, 17:929-932. Farbood, M.I., MacNeil, J.H., and Ostovar, K. 1976/. Effect of rosemary spice extractive on growth of microorganisms in meats. Journal of Milk and Food Technology, 39:675-679. Foodborne Disease Working Party. 1997. Foodborne disease: Towards reducing foodborne illness in Australia. Technical Report Series No. 2, Communicable Diseases Network Australia and New Zealand, New South Wales. Gescher, A. J., Sharma, R., A., and Steward, W. P. 2001. Cancer chemoprevention by dietary constituents: a tale of failure and promise. The Lancet Oncology, 2:371-379. Glass, K. A., and Eric A. Johnson, E. A. 2004. Antagonistic effect of fat on the antibotulinal activity of food preservatives and fatty acids. Food Microbiology, 21: 675-682.

Hahm, E-R., Cheon, G., Lee, J., Kim, B., Park, C., and Yang, C-H. 2002. New and known symmetrical curcumin derivatives inhibit the formation of Fos-Jun-DNA complex. Cancer Letters, 184:89-96. Huhtamen, C.N. 1980. Inhibition of Clostridium botulinum by spice extracts and aliphatic alcohols. Journal of Food Protection, 43:195-196.

Ismaiel, A.A., and Pierson, M.D. 1990a. Effect of sodium nitrite and origanum oil on growth and toxin production of Clostridium botulinum in TYG broth and ground pork. Journal of Food Protection, 53:958-960.

Ismaiel, A.A., and Pierson, M.D. 1990b. Inhibition of germination, outgrowth, and vegetative growth of Clostridium botulinum 67B by spice oils. Journal of Food Protection, 53:755-758. Ismaiel, A.A., and Pierson, M.D. 1990c. Inhibition of growth and germination of C. botulinum 33A, 40B, and 1623E by essential oil of spices. Journal of Food Science, 55:1676-1678. Jafari, M., and Emam-Djomeh, Z. 2007. Reducing nitrite content in hot dogs by hurdle technology. Food Control, 18:1488-1493. Jo., C., Ahn, H.J., Son, J.H., Lee, J.W., and Byun, M.W. 2003. Packaging and irradiation effect on lipid oxidation, color, residual nitrite content, and nitrosamine formation in cooked pork sausages. Food Control, 14:7-12. Johnson, M.G., and Vaughn, R.H. 1969. Death of Salmonella typhimurium and Escherichia coli in the presence of freshly reconstituted dehydrated garlic and onin. Applied Microbiology, 17:903-905. Joossens, J.V., Hill, M.J., Elliott, P., Stamler, R., Stamler, J., Lesaffre, E., Dyer, A., Nichols, R., and Kesteloot, H. 1996. Dietary salt, nitrate and stomach cancer mortality in 24 countries. International Journal of Epidemiology, 25:494-504.

197 Juneja, V.K., Huang, L-H., and Thippareddi, H.H. 2006a. Predictive model growth of Clostridium perfringens in cooked cured pork. International Journal of Food Microbiology, 110:85-92. Juneja, V.K., Fan, X., Pena-Ramos, A., Diaz-Cinco, M., and Pacheco-Aguilar, R. 2006b. The effect of grapefruit extract and temperature abuse on growth of Clostridium perfringens from spore inocula in marinated, sous-vide chicken products. Innovative Food Science and Emerging Technologies, 7:100-106. Leuschner, R.G.K., and Zamparini, J. 2002. Effects of spices on growth and survival of Escherichia coli O157 and Salmonella enterica serovar Enteritidis in broth model systems and mayonnaise. Food Control, 13:399-404. Lindström, M., Kiviniemi, K., and Korkeala, H. 2006. Hazard and control of group II (non-proteolytic) Clostridium botulinum in modern food processing. International Journal of Food Microbiology, 108:92-104. Maheshwari, R.K., Singh, A.K., Gaddipati, J. and Srimal, R.C. 2006. Multiple biological activities of curcumin: A short review. Life Sciences, 78:2081-2087. Mattila-Sandholm, T., Haikara, A., and E. Skyttä, E. 1991. The effect of Pediococcus damnosus and Pediococcus pentosaceus on the growth of pathogens in minced meat. International Journal of Food Microbiology, 13:87-94. Mead, G.C. 1985. Selective and differential media for Clostridium perfringens. International Journal of Food Microbiology, 2:89-98. Munro, R. 1989. Patterns of resistance to anaerobic organisms in Australia. Diagnostic Microbiology and Infectious Disease, 12:159-163.

Mytle, N., Anderson, G.L., Doyle, M.P. and Smith, M.A. 2006. Antimicrobial activity of clove (Syzgium aromaticum) oil in inhibiting Listeria monocytogenes on chicken frankfurters. Food Control, 17:102-107. Naidu, A. S. ed. 2000. Natural Food Antimicrobial Systems. CRC Press, United States. Nasar-Abbas, S.M., and Halkman, A.K. 2004. Antimicrobial effect of water extract of sumac (Rhus coriaria L.) on the growth of some food borne bacteria including pathogens. International Journal of Food Microbiology, 97:63-69. Negi, P. S., Jayaprakasha, G. K., Jagan Mohan Rao, L., and Sakariah, K. K. 1999. Antibacterial activity of turmeric oil: A byproduct from curcumin manufacture. Journal of Agricultural Food Chemistry, 47:4297-4300. Østerlie, M., and Lerfall, J. 2005. Lycopene from tomato products added minced meat: Effect on storage quality and colour. Food Research International, 38:925-929. Pfeiffer, E., Höhle, S., Solyom, A. M., and Metzler, M. 2003. Studies on the stability of turmeric constituents. Journal of Food Engineering, 56:257-259. Piper, J.T., Singhal, S.S., Salameh, M.S., Torman, R.T., Awasthi, Y.C., and Awasthi, S. 1998. Mechanisms of anticarcinogenic properties of curcumin: the effect of curcumin on glutathione linked detoxication enzymes in rat liver. The International Journal of Biochemistry and Cell Biology, 30:445-456.

198 Rhodehamel, E.J., Reddy, N.R., and Pierson, M.D. 1992. Botulism: the causative agent and its control in foods. Food Control, 3:125-143. Sağdiç, O., Kuşçu, A., Özcan, M., and Özçelik, S. 2002. Effects of Turkish spice extracts at various concentrations on the growth of Escherichia coli O157:H7. Food Microbiology, 19:473-480. Seydim, A.C., and Sarikus, G. 2006. Antimicrobial activity of whey protein based edible films incorporated with oregano, rosemary and garlic essential oils. Food Research International, 39:639-644. Shelef, L.A., Jyothi, E.K., and Bulgarelli, M.A. 1984. Growth of enteropathogenic and spoilage bacteria in sage-containing broth and foods. Journal of Food Science, 49:737- 740. Shishu, Singla, A.K., and Kaur, I.P. 2003. Inhibitory effect of dibenzoylmethane on mutagenicity of food-derived heterocyclic amine mutagens. Phytomedicine, 10:575- 582. Sprong, R. C., Hulstein, M. F. E., and van der Meer, R. 2002. Bovine milk fat components inhibit food-borne pathogens. International Dairy Journal, 12:209-215. Srinivasan, K., Sambaiah, K., and Chandrasekhara, N. 1992. Loss of active principles of common spices during domestic cooking. Food Chemistry, 43:271-274. Srinivasan, K. 2005. Spices as influencers of body metabolism: an overview of three decades of research. Food Research International, 38:77-86. Surh, U-J., Chun, K-S., Cha, H-H., Han, S. S., Keum, Y-S., Park, K-K., and Lee, S. S. 2001. Molecular mechanisms underlying chemopreventive activities of anti- inflammatory phytochemicals: Down-regulation of COX-2 and iNOS through suppression of NF-κB activation. Mutation Research, 480-481:243-268. Takikawa, A., Abe, K., Yamamoto, M., Ishimaru, S., Yasui, M., Okubo, Y., and Yokoigawa, K. 2002. Antimicrobial activity of nutmeg against Escherichia coli O157. Journal of Bioscience and Bioengineering, 94:315-320. Thongson, C., Davidson, P.M., Mahakarnchanakul, W., and Weiss, J. 2004. Antimicrobial activity of ultrasound-assisted solvent-extracted spices. Letters in Applied Microbiology, 39:401-406. Witkowska-Banaszczak, E., Bylka, W., Matlawska, I., Goślińska, O., and Muszyński, Z. 2005. Antimicrobial activity of Viola tricolor herb. Fitoterapia, 76:458-46.

199 Chapter 5 Conclusions

5.1 Overview to experimental aim one and major findings

Nitrate and nitrite in Australian food supply was determined and their estimated dietary intake was shown to be higher than the recommended dietary intake. However, these intakes were based on the upper extreme on high dietary sources of nitrate and nitrite. Thus depending on the dietary practices of individuals, the dietary intake of these anions may not pose a significant public health issue due to the formation of N-nitroso compounds in vivo.

5.1 Overview to experimental aim two and major findings

The consumption of one serving of cured meat with a low nitrate and nitrite diet did not produce detectable N-nitrosodimethylamine in vivo. Therefore the supplementation of 500 mg vitamin C did not show a reduction in N-nitrosodimethylamine in vivo. However, if the diet consisted of high nitrate containing vegetables and other high nitrite containing meat then a reduction of N-nitrosomethylamine would be expected.

5.3 Overview to experimental aim three and major findings

Turmeric, ginger and rosemary leaves showed inhibition against C. sporogenes in vitro. In addition, these spices along with garlic displayed growth inhibition against other common foodborne bacteria including E. coli, L. monocytogenes, S. enteritidis and S. aureus. There is a potential to add these spices in the formulation of cured meat products to reduce the use of sodium nitrite thus minimising their dietary intake.

200 Appendix A Participation information statement and consent form

THE UNIVERSITY OF NEW SOUTH WALES

Fate of nitrates and nitrites in food and body

[Participant selection and purpose of study] You (i.e. the research participant) are invited to participate in a study of nitrates and nitrites in the body. We James Hsu and Dr. Jayashree Arcot hope to learn the rate of nitrite reduction to nitrate in the saliva, and to minimize potential carcinogen formation in vivo using vitamin C. You were selected as a possible participant in this study because you are over 18 years of age and are healthy.

[Description of study and risks] If you decide to participate, we will ask you to provide saliva and urine samples first thing in the morning, rinse your mouth with nitrate containing water and collect saliva, then chew cured meat (salami, ham, sausage) and collect saliva. Same meat will then be provided with bread for consumption followed by a urine sample after 3 hours. The next day after fasting overnight, you will be asked to consume the same meat with bread then provide a urine sample after 3 hours.

Discomforts may be to follow a diet without cured meat, green leafy vegetables and fruits for 48 hours except cured meat provided during the study. Providing saliva and urine samples is not invasive but may cause embarrassment which is minimised if done at the privacy of the participant’s home. Participant is required to bring the bodily fluid samples to UNSW for storage twice. Total participation time is about 48 hours but actual experimental time for collecting bodily fluids and eating food provided is only about 20 minutes in total during the 48 hours period.

Risks are minimal if participant is healthy and eats a variety of other fresh meats and vegetables. Alternatively, experiment can be done on university campus, within the

201 School of Chemical Sciences and Engineering, if participant is unsure or if it is more convenient.

[Confidentiality and disclosure of information] Any information that is obtained in connection with this study and that can be identified with you will remain confidential and will be disclosed only with your permission, except as required by law. If you give us your permission by signing this document, we plan to discuss/publish the results in scientific journals in the form of averaged numbers and statistics reported in the result section. In any publication, information will be provided in such a way that you cannot be identified.

[Recompense to participants] One movie voucher per participant will be given to compensate for their time, and transport costs (cheapest option) will be covered if student needs to come to the University for the sole purpose of the experiment on the day.

Complaints may be directed to the Ethics Secretariat, The University of New South Wales, SYDNEY 2052 AUSTRALIA (phone 9385 4234, fax 9385 6648, email [email protected]). Any complaint you make will be investigated promptly and you will be informed of the outcome.

[Feedback to participants] If requested, a summary of research findings will be sent to the participants as email, mail or fax based on their preferred method of contact.

[Your consent] Your decision whether or not to participate will not prejudice your future relations with the University of New South Wales. If you decide to participate, you are free to withdraw your consent and to discontinue participation at any time without prejudice.

If you have any questions, please feel free to ask us. If you have any additional questions later, Mr. James Hsu, 0410 465 130 or Dr. Jayashree Arcot 9385 5360 will be happy to answer them.

202 EXPERIMENTAL BACKGROUND AND PROCEDURES

INTRODUCTION

Nitrite salt is used as a chemical preservative in cured meat to prevent a type of food poisoning called botulism. However, nitrite can react with food in the stomach to form a group of potent carcinogens known collectively as N-nitroso compounds (NOCs), which increases the risk of developing stomach cancer in the long term. Nitrate naturally found in vegetables can be reduced to nitrite by oral bacteria in the mouth, which also contributes to the daily intake of nitrite. Vitamin C (ascorbic acid) is known to reduce the formation of NOCs such as Nitrosodimethylamine (NDMA).

This study will estimate the amount of nitrate reduction to nitrite in the oral cavity of humans, and the effect of vitamin C on NDMA formation in vivo (in the body) after the consumption of cured meat and vitamin C supplement.

SCIENTIFIC MERITS

Diet consisting of vegetables and cured meat contributes to a person’s daily intake of nitrates and nitrites, the latter can increase the risk of stomach cancer. However, oral bacteria in the mouth can also increase one’s daily dietary intake of nitrite. Therefore by measuring the amount of nitrate and nitrite in foods and understanding factors involved in the conversion rate of nitrate to nitrite in the oral cavity, one can estimate the total dietary intake of nitrates and nitrites. This study will gain useful insights on nitrate and nitrite conversion in vivo, and study the potential of vitamin C to greatly reduce the formation of NDMA. This may also encourage meat manufacturers to find and use healthier alternatives to sodium nitrite.

EXPERIMENTAL PROCEDURE

AIMS: To determine the rate of nitrate reduction to nitrite in human oral cavities; and to understand the effects of ascorbic acid (vitamin C) on the production of NDMA (carcinogen) in human urine.

203

HYPOTHESES: 1. The rate of nitrate reduction in the oral cavity should vary between individuals due to differences in the oral microflora, diet, lifestyle and medications. 2. Diet without vitamin C supplement should produce more NDMA in their urine compared to the same diet with vitamin C supplement.

MATERIALS AND METHODS:

Day 1

1. Collect materials OR participate at UNSW and sign a consent form. 2. Follow a restricted diet (Appendix A) for today and tomorrow-resume normal diet after Day 3. Two (2) breakfast provided for Days 2 and 3. 3. Store all meat samples and one water jar in the fridge.

Day 2 (store all biological samples in fridge then bring them to me on completion of each day’s tasks, OR come to university and do it).

On waking: (after overnight fasting and don’t brush teeth yet): 1. Provide urine in 2 L urine bottle marked (1). Contains sodium hydroxide (corrosive: DO NOT TOUCH OR EAT!) 2. Chew the chewing gum periodically and spit out saliva into 50 mL yellow cap specimen jar marked A to red marked line (may take up to 5 mins, discard chewing gum). Contains sodium hydroxide (corrosive: DO NOT TOUCH OR EAT!) Tasks: 3. Rinse mouth thoroughly with tap water. 4. Rinse mouth with nitrate-water marked D for 30 sec then spit all into a yellow cap specimen jar marked B. (DO NOT DRINK). 5. Rinse mouth thoroughly with tap water. 6. Chew meat (Snap lock A) provided for 30 sec then spit into yellow cap specimen jar marked C (DO NOT EAT MEAT). 7. Brush your teeth as usual then rinse with tap water thoroughly. 8. Eat the meat sandwich (at least eat all the meat, snap lock B) provided then do not eat or drink anything for 3 hrs (water ok).

204 9. After 3 hrs provide urine sample in the second urine bottle marked (2)(If necessary collect urine for up to 3 hrs in that bottle). 10. Bring all samples collected today to me (contact details below).

Day 3

On waking : 1. Discard urine and brush teeth then rinse thoroughly with tap water. Tasks: 2. Eat the meat sandwich (Snap lock C) provided and the vitamin C tablet (in foil wrap) then do not eat or drink anything for 3 hrs (water ok). 3. After 3 hrs provide urine sample in the third urine bottle marked (3)(If necessary collect urine for up to 3 hrs in that bottle). 4. Bring all samples collected today to me. 5. Collect your movie voucher and thank you for your participation (resume normal diet).

205 Human Trials:

FLOW CHART Nitrate reduction in the oral cavity; and the effect of Vitamin C on NDMA formation in vivo

DAY 1 DAY 2 DAY 3 Take home On Waking provide: On Waking: 3 x urine bottles 3 x saliva specimen jars 3 x meat samples 1 x urine sample (1) 1. Discard urine 1 x nitrate water 1 x vitamin C tablet 1 x saliva sample (A) 2. Brush teeth 1 x chewing gum

Avoid foods Tasks (morning): Tasks (morning): (Appendix A):

Vitamin C rich e.g. 1. Nitrate water rinse, 1. Eat meat (B2) vegetables and fruits. provide saliva (B) 2. Take one (1) vitamin C Also avoid all cured 2. Chew meat (A), provide tablet provided meat except ones saliva (C) given. 3. Eat meat (B1)

3 hours fasting: 3 hours fasting:

1 x urine sample (2)(or 1 x urine sample (3)(or collect urine up to 3 hr collect urine up to 3 hr after meat consumption) from meat consumption)

Return samples to me Resume usual diet after providing urine sample.

206 CONCLUSION

Diet and lifestyle contributes to significant number of preventable cancers. Nitrite is used as a preservative in cured meat and is known to produce carcinogens in the stomach. Vitamin C is shown to reduce the formation of these carcinogens in vitro and in animals. By knowing the intake of nitrate and nitrite and the effect of vitamin C in vivo (in the body), it may be possible to modify ones lifestyle and diet to minimize the risk of developing digestive cancers.

THANK YOU for your participation, it means a lot to me and my research for your time and effort.

CONTACT DETAILS

James Hsu Level 7, room 714 Applied Science Building (F10) Enter gate 2, High St. Kensington, UNSW 2052 Office: 9385 5053 Mobile: 0410 465 130, Email: [email protected]

Dr. Jayashree Arcot (supervisor) Level 7, Applied Science Building (F10) Enter gate 2, High St. Kensington, UNSW 2052 Office: 9385 5360 Email: [email protected]

END

207 Appendix B Food to avoid and food allowed during the experiment

Food/drinks to Avoid* Food/drinks Allowed* Fruits and vegetables Fruits and vegetables All vitamin C rich fruits >10mg per serve Fruits with low vitamin C < 10 mg per (strawberries, blueberries, raspberries, serve (raw apple, raw cherries, raw blackberries, kiwi, mango, pineapple, nectarine, any figs, raw grapes, canned tomatoes, orange, mandarin, lemon, lime) apricot, any peaches, any pears, passion All vitamin C rich vegetables >10mg per fruit, any plum, any prunes, raisins, dried serve (green leafy, yellow, orange, or red dates) coloured) Vegetables with low vitamin C < 10 mg Green peas and sprouts per serve (iceberg lettuce, raw celery, Potatoes boiled leek, peeled cucumber, eggplant, Can take fiber supplement (e.g. zucchini, beetroot, carrot, corn, metamucil) mushroom) Garlic and onions (small quantity only) Legumes and beans (chickpea, lentil, broad bean, kidney bean, lima bean, water chestnuts) Alfalfa sprouts Cured/processed/smoked meat Fresh meat Salami, hotdog, ham, cabanosi, bacon, Chicken, pork, beef, meat-based soup and smoked salmon, smoked ham, organ meat pies (kidneys, liver etc). Seafood Bread and cereals Fresh fish (salmon, dory, barramundi, Plain bread and muffin (white, brown, canned fish (tuna, anchovies), bivalves wholemeal), plain cereals (barley, bran, (oysters, mussels, clams) and prawns oatmeal, wheat, rice ), pasta,

Vitamin supplements Dairy and alternatives Vitamin C (ascorbic acid) and Vitamin E Milk, yoghurt, eggs, soy milk, tofu/soy (tocopherols) products All cheeses Fats and oils Swiss, cheddar, parmesan, blue-vein, Margarine, butter, all oils mozzarella, ricotta All nut and seeds Drinks Cashew, peanut, almond, pine, Water, soft drinks, coffee, sports drink macadamia, pecan, pumpkin seeds, (gatorade, powerade) flaxseeds, safflower seeds. Drinks Confectionary Alcohol, tea, fruit and vegetable juices, Milk chocolate, cakes, pastries, hard energy drinks (red bull, mother, V) candies, ice-cream, biscuits, honey, sugar Confectionary Others Dark chocolate Potato chips, popcorn, sauces/gravies, salt, pepper *Please consult your physician if you are concerned about this restricted diet.

208 Appendix C Questionnaire for volunteers

Volunteer No.______Starting Date______

Name (or anonymous):

Age

Gender

Nationality (birth)

Estimated servings of cured/smoked meat* per week Estimated servings of leafy green vegetable+ per week Vitamin/mineral supplements (if any) and frequency General health status

Preferred contacts

*salami, hotdog, sausage, ham, bacon, cabanossi, smoked fish etc. +spinach, bok choy, choy sum, salad leaves etc.

Collection Day 1 Day 2 Day 3 Saliva control NA 1 0 Nitrate water NA 1 0 Chewed meat NA 1 0 Urine fasting NA 1 0 Urine control NA 1 0 Urine experiment NA 0 1

209 Appendix D Publication in the Journal of Food Chemistry

Nitrate and nitrite quantification from cured meat and vegetables and their estimated dietary intake in Australians

James Hsu, Jayashree Arcot *, N. Alice Lee

Food Science and Technology, School of Chemical Sciences and Engineering, University of New South Wales, Sydney, NSW 2052, Australia

210 This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy

Food Chemistry 115 (2009) 334–339

Contents lists available at ScienceDirect

Food Chemistry

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

Analytical Methods Nitrate and nitrite quantiﬁcation from cured meat and vegetables and their estimated dietary intake in Australians

James Hsu, Jayashree Arcot *, N. Alice Lee

Food Science and Technology, School of Chemical Sciences and Engineering, University of New South Wales, Sydney, NSW 2052, Australia article info abstract

Article history: High dietary nitrate and nitrite intake may increase the risk of gastro-intestinal cancers due to the in vivo Received 7 May 2008 formation of carcinogenic chemicals known as N-nitroso compounds. Water and leafy vegetables are nat- Received in revised form 19 November 2008 ural sources of dietary nitrate, whereas cured meats are the major sources of dietary nitrite. This paper Accepted 21 November 2008 describes a simple and fast analytical method for determining nitrate and nitrite contents in vegetables and meat, using reversed-phase HPLC-UV. The linearity R2 value was >0.998 for the anions. The limits of quantiﬁcation for nitrite and nitrate were 5.0 and 2.5 mg/kg, respectively. This method is applicable for Keywords: both leafy vegetable and meat samples. A range of vegetables was tested, which contained <23 mg/kg Anions nitrite, but as much as 5000 mg/kg of nitrate. In cured and fresh meat samples, nitrate content ranged Extraction Carcinogen from 3.7 to 139.5 mg/kg, and nitrite content ranged from 3.7 to 86.7 mg/kg. These were below the regu- Cured meat latory limits set by food standards Australia and New Zealand (FSANZ). Based on the average consump- Nitrates tion of these vegetables and cured meat in Australia, the estimated dietary intake for nitrate and nitrite Nitrites for Australians were 267 and 5.3 mg/adult/day, respectively. Tetrabutylammonium phosphate Ó 2008 Elsevier Ltd. All rights reserved. Matrix interference

1. Introduction Italy (Forman & Shuker, 1997). Exposure to endogenously formed N-nitroso compounds had been associated with increased risks of It was estimated that 80% of human cancers were caused by cancer of the stomach, oesophagus and bladder (Bartsch et al., environmental factors associated with food, water and air (Walt- 1990). ers, 1980). In addition, malnutrition, dietary habits and lifestyle Australia’s food composition data were mostly based on over- may be directly or indirectly related to 40% of the human cancers seas data especially those from the United Kingdom (UK) and the (Ologhobo, Adegede, & Maduagiwu, 1996). High dietary intakes of United States (US) till recently. However, in the revised Australian nitrate and nitrite have been implicated in the etiology of human composition tables based on food analysis performed in Australia, gastric cancer based on epidemiology and clinical studies (Bartsch, the edible portion of fruit increased by 4% whereas in meat it de- Ohshima, Shuker, Pignatelli, & Calmels, 1990; Joossens et al., creased by 16% (Cashel & Greenﬁeld, 1995). Thus dietary contribu- 1996). tion of nitrate and nitrite may be over-estimated, whereas dietary Nitrate is naturally present in leafy vegetables and nitrite is intake of antioxidants such as vitamin C and vitamin E may have usually added to meat as a preservative in the form of sodium or been underestimated. potassium salt (Cammack et al., 1999). In addition nitrate can be The dietary intake of nitrates and nitrites in foods can vary reduced to nitrite in the oral cavity and in the stomach (Duncan greatly from region to region depending on factors such as farming et al., 1997). Once in the stomach, nitrite can react with amines practices, climate, soil quality, manufacturing processes and and amides, which are organics containing nitrogen such as amino legislation. Nitrate and nitrite contents of foods are not available acids, to form a group of carcinogens known as N-nitroso com- in Australia; hence values from overseas are commonly used. pounds (Archer, 1989). Stomach is most at risk from endogenous Due to the growing concern of N-nitroso compounds, accurate N-nitroso compound synthesis since stomach acid catalyses nitro- and robust methods are necessary for long-term monitoring of sation reactions. High nitrate intake was associated with gastric nitrate and nitrite concentrations in foods for susceptible cancer in England, Colombia, Chile, Japan, Denmark, Hungary and populations. It is therefore the aim of this study was to develop an accurate, simple and cost-effective method for quantifying the nitrate and * Corresponding author. Tel.: +61 2 9385 5360; fax: +61 2 9385 5966. nitrite contents in commonly consumed vegetables, cured meat E-mail address: [email protected] (J. Arcot). and fresh meat produced in Australia.

J. Hsu et al. / Food Chemistry 115 (2009) 334–339 335

2. Materials and methods (Usher & Telling, 1975). Hence, the optimal extraction conditions were used for nitrate and nitrite determinations in fresh vegeta- 2.1. Reagents bles, cured meat and fresh meat. Mean recoveries were >92% for both nitrate and nitrite in all three food matrices tested (Tables 1 Analytical grade sodium nitrite and potassium nitrate from Uni- and 2). Factors affecting nitrate and nitrite recovery in foods in- var (Ajax Finechem) were used as standards and for recovery stud- clude (1) temperature, since nitrate and nitrite are not stable at ies. HPLC grade methanol from lab-scan was used and ion-pairing high temperatures, (2) cooking conditions, which can affect pH of agent tetrabutylammonium phosphate was purchased from Waters. the sample water and exposure to atmospheric oxygen, (3) pH of the sample water, since nitrite is readily converted to nitric acid 2.2. Food samples or nitric oxide at acidic pH, and (4) sources of food samples can vary greatly and may contain interfering substances such as iron All vegetables were purchased at local supermarkets, produce and magnesium (Usher & Telling, 1975). shops or wholesale and kept at refrigeration temperature and ana- The extraction and detection method would affect nitrite and lysed within 24 h. All cured and fresh meat products were also pur- nitrate quantiﬁcation in meat and vegetables. This method was chased at supermarkets and kept at refrigeration temperature and chosen because it was fast, sensitive and accurate. In both cases analysed within 48 h. heat (hot water) was used to extract nitrite and nitrate, and in the case of vegetables blanching was a common cooking practice 2.3. Apparatus that was chosen in the present study. In addition, pH was monitored and maintained close to neutral pH to minimize conversion Waters HPLC controller model number 600 with photo array of nitrite to nitrous acid or nitrous oxide. detector model number 996 and autosampler model number Nitrite levels in vegetables may increase during post-harvest storage by the action of indigenous bacteria and/or the presence 717 plus were used. Phenomenex C18 110A Gemini column (250 mm Â 4.6 mm Â 5 lm) was used for the separation. Injection of nitrate reductase (Hunt, 1994), especially when they are left at volume was 10 ll with ﬂow rate set at 1 mL/min and wavelength room temperature or higher. This may explain the small amount set at 214 nm. Mobile phase consisted of methanol: water of nitrite (20 mg/kg) present in Gai choy during the preparation (75:25) with 0.075 M of tetrabutylammonium phosphate (PIC-A). at room temperature (Table 1). Likewise, it was demonstrated that there was no detectable nitrite in 94% of edible fresh retail vegeta- 2.4. Methods bles (Hunt & Turner, 1994).

2.4.1. Standards Table 1 Potassium nitrate (KNO3) and sodium nitrite (NaNO2) were mixed in MilliQ water in volumetric flasks to give a range between Mean nitrate and nitrite contents and their recoveries in cooked fresh vegetables. 5.0 and 100 mg/L for nitrite ions and 2.5–50 mg/L for nitrate ions. Vegetables Nitrite (mg/kg) Nitrate (mg/kg) English spinach 0 4849.6 ± 573.6 2.4.2. Samples Recovery (%) 89 74 Weighed 10–50 g of meat samples including salami, hot dogs, Buk choy 0 1841.1 ± 84.0 ham, bacon, Frankfurt and beef, which were purchased from the lo- Recovery (%) 97 97 Choy sum 0 1376.9 ± 56.0 cal supermarkets (at least two packets each) of two different brands, Recovery (%) 111 102 were blended with 300 mL distilled water for 1 min, then made up to Chinese cabbage 0 236.2 ± 27.4 500 mL in volumetric flasks. The pH was measured and 1 mL was ta- Recovery (%) 91 97 ken out for measuring nitrate and nitrite content before cooking Gai choy 19.6 ± 10.8 1642.3 ± 126.0 using HPLC. Ten millilitre of mixture of each sample was transferred Recovery (%) 102 100 Iceberg lettuce 0 48.0 ± 30.2 into 100 mL volumetric flasks and heated in a water bath at 75, 80, Recovery (%) 92 110 90 and 100 °C for 5, 10 and 15 min. The mixture was made up to 100 mL with distilled water and was shaken. The mixture was al- Values are means of at least four replicate determinations from two sources and up to 15 determinations. lowed to settle and cool; then measured pH and nitrite and nitrate contents. The pH was adjusted with 0.1 M NaOH to neutral pH. Then the mixture was centrifuged at 10,000 rpm for 10 min; then supernatant was removed for ultra-filtration. The filtrate was used for fur- Table 2 Mean raw nitrate and nitrite contents and their recoveries in cured and fresh meat ther analysis including quality control such as recovery studies. from Sydney supermarkets after pH adjustment. Fresh vegetables including English spinach, buk choy, choy sum, Chinese cabbage, gai choy and iceberg lettuce were purchased from Meat Nitrite (mg/kg) Nitrate (mg/kg) the local supermarkets and produce stores. Three bunches each Hot dog 78.6 ± 16.4 69.9 ± 11.3 from two different locations with at least three replicates were Recovery (%) 109 103 Ham 34.2 ± 5.5 19.0 ± 8.1 used for the analysis including recoveries. To examine the effects Recovery (%) 97 87 of sample preparation and extraction conditions on nitrate and ni- Salami 0 142.5 ± 36.3 trite determination samples were chopped in thirds or blended or Recovery (%) 91 102 both and weighed between 25 and 100 g in 500 mL beakers. Spik- Bacon 15.7 ± 14.5 23.3 ± 8.2 ing with standards was done before cooking in water bath between Recovery (%) 91 82 Frankfurt 83.9 ± 10.1 54.9 ± 8.7 60 and 100 °C for 5–30 min. Recovery (%) 96 94 Minced beef 0 18.7 ± 6.2 Recovery (%) 80 104 3. Results and discussion Beef medallion 0 38.5 ± 14.9 Recovery (%) 80 75

Nitrate and nitrite can be unstable and different sampling Values are means of at least four replicates from two to four brands with up to ﬁve methods and extraction procedures can inﬂuence their recoveries determinations. Author's personal copy

336 J. Hsu et al. / Food Chemistry 115 (2009) 334–339

Cultivar and harvest date can affect the nitrate and nitrite levels nitrate concentration was lower at 19.0 mg/kg (Table 2). Some of selected vegetables (Amr & Hadidi, 2001). This may explain the manufacturers add less nitrite but more nitrate as a nitrite reserve. high variability between findings presented in this study also con- This may also explain the differences in the findings by Öztekin tributing to high standard deviation particularly in English spinach et al. (2002), where the nitrite and nitrate contents in ham were as observed in Table 1. However, meat samples with low levels of 4.0 and 35.6 mg/kg, respectively. nitrates had smaller standard deviation (Table 2), most likely be- Dionex Corporation (1998) found the nitrite and nitrate con- cause the low levels of nitrates in general meant less chance of tents in ham to be 11.6 and 5.4 mg/kg, respectively, whereas sala- reacting to the conditions they were exposed to. mi contained 108.0 mg/kg nitrite and 98.5 mg/kg nitrate. Using English spinach had the highest nitrate content (4850 mg/kg) capillary electrophoresis, the nitrite and nitrate content in salami compared to other vegetables (Table 1). This finding correlated detected were 24.3 and 43.6 mg/kg, respectively (Öztekin et al., well with the literature (Öztekin, Nutku, & Erim, 2002). However, 2002). Compared to their findings, the current study showed that according to Gaiser, Rathjen, and Spiess (1996), spinach blanched the salami contained no nitrite but much more nitrate at for 3 min can contain in the range of 50–5600 mg/kg nitrates with 142.5 mg/kg (Table 2). Although the extraction methods were sim- a mean nitrate concentration of approximately 2000 mg/kg and a ilar the temperature used in our study was higher, apart from the large standard deviation of 1411.4 mg/kg. This demonstrated the differences that may be attributed to the manufacturing practices. high variability of nitrate content in spinach and other green leafy Stalikas, Konidari, and Nanos (2003) used similar extraction tem- vegetables. Excluding spinach, other vegetables tested had nitrate perature and reported that nitrate and nitrite contents in salami ranging 48–1841 mg/kg, which was less than half that of spinach were 54 and 84 mg/kg, respectively. Thus differences are more (Table 1). Thus it can be concluded that spinach contributed to likely to be due to the manufacturing processes. the highest dietary nitrate intake from leafy green vegetables. Let- It was reported by Dennis, Key, Papworth, Pointer, and Massey tuce contained lowest amount of nitrate in this study (48.0 mg/kg, (1990) that the mean nitrite content in bacon was 24.0 mg/kg and Table 1), which was significantly lower to earlier studies that dem- for nitrate was 43.0 mg/kg, whereas nitrite and nitrate in ham were onstrated high nitrate content in lettuce at 2500 mg/kg (Marshall & 56.0 and 22.0 mg/kg, respectively. They used similar extraction and Trenerry, 1996). This dissimilarity may be due to horticultural detection methods but with an anion exchange column. Both ba- practices such as the use of nitrate-based fertilizers. con and ham products in this study contained less nitrate and ni- Different countries have set their maximum limits for the addi- trite (Table 2) in comparison. Siu and Henshall (1998) who found tion of nitrate and/or nitrite salts in cured meat. Under the Austra- that nitrite and nitrate contents in salami were 108.0 and lian Food Standard Code 1.3.1 schedule 1, 125 mg/kg of nitrite in a 98.5 mg/kg, respectively, and 11.6 and 5.4 mg/kg for ham, respec- form of potassium or sodium salt is permitted in cured, dried, and tively. Sample extraction procedures used in the current study slow dried cured meat; whereas in commercially sterile and were similar to Marshall and Trenerry (1996), but they omitted canned cured meat, the maximum nitrite (potassium or sodium the heating step. This may explain the low nitrite content of less salts) permitted is 50 mg/kg. For slow dried cured meat, the max- than 10 mg/kg in salami, leg ham and bacon. However the nitrate imum allowed nitrate (potassium or sodium salts) is 500 mg/kg contents were higher at 141.5, 132.5 and 48.0 mg/kg, respectively. (FSANZ, 2007–2008). Given the established antimicrobial effect of Different cured meat products may require different ratio of nitrite nitrite salts, particularly in reference to Clostridium botulinum in and nitrate as preservatives. Since fresh meat does not naturally cured meat, its level should remain sufficient enough to prevent contain nitrite (Table 2), its nitrite and nitrate contents have not the occurrence of foodborne illnesses, but also kept to the mini- been extensively tested. However, based on this study, the nitrate mum to minimize dietary nitrite intake in light of its potential ad- content in minced beef and medallion beef were within the range verse health effects based on epidemiological and clinical studies. found in cured meat products (Table 2). Seven types of meat tested in this study had at least four repli- It was demonstrated that recovery increases as the meat solids cates each from at least two brands. Nitrate and nitrite contents in decreases (Usher & Telling, 1975). Hence using smaller meat sam- various cured meat products were below the maximum allowable ples should reduce the effects of interfering substances, which was limit set by Food Standards Australia and New Zealand (FSANZ) at demonstrated in this study (Table 2). Furthermore, most interfer- 125 mg/kg (Table 2). However, there was no limit set for fresh ence can be eliminated by UV detection. However, chloride ions meat. Continuous monitoring of nitrite used in cured meat prod- maybe detected by UV as positive or negative peaks in the wave- ucts is important to ensure that the dietary intake of nitrite is kept length used for nitrate and nitrite and are eluted before nitrite to below the limit set by FSANZ. (Di Matteo & Esposito, 1997). Chloride peaks were not present at Interferences naturally present or added additives in cured 214 nm in this study, which suggests that chloride ions did not meat products may account for differences in nitrate and nitrite interfere with nitrite quantification since nitrite recovery was recovery. For example, Butt, Riaz, and Iqbal (2001) demonstrated above 92% for both meat and vegetable samples (Tables 1 and 2). that the presence of 50-fold sulphate and chloride did not affect Due to its reactive nature, nitrite analysis from food does not the resolution and percent recovery of nitrite, but did reduce the give a true representation of the total nitrite added. Furthermore, resolution and recovery of nitrate. In addition, the presence of nitrite added to meat is usually present as nitric oxide bound with magnesium, iron and calcium significantly reduced the percentage other food components such as myoglobin (5–15%), sulphydryl recovery of both anions, which should be removed to ensure accu- groups (5–15%), lipids (1–5%), proteins (20–30%), as nitrate rate determination of nitrate and nitrite. Furthermore, Butt et al. (<10%), and as free nitrite (10–15%) (Zanardi, Dazzi, Madarena, & (2001) also demonstrated that under optimized HPLC conditions, Chizzolini, 2002). Therefore recovery range may be quite large as both nitrate and nitrite peaks began to merge when the concentra- a result of nitrite’s reactive nature and its attachment to other food tion of nitrite was above six-fold of nitrate concentration, hence ni- components. However, because only free nitrite can participate in trite used in calibration curve and for recovery were half the nitrosation, other methods of food extraction estimate the total ni- concentration of nitrate to minimize the merging of nitrite and nitrite present by releasing food-bound nitrite. This may over esti- trate peaks. mate the significance of dietary nitrite and the etiology of gastric Using similar detection method as Reinik et al. (2005), they cancer. Hot water extraction to quantify free nitrite available to found the mean sodium nitrite and nitrate concentrations in ham participate in nitrosation was used in this study. were 20.8 and 68 mg/kg, respectively. However in this study, the Regarding relevance to incidence of gastric cancer, based on nitrite concentration in ham averaged at 34.2 ± 5.6 mg/kg and age-standardized statistics, diagnosed gastric cancer rate per 100, Author's personal copy

J. Hsu et al. / Food Chemistry 115 (2009) 334–339 337

000 in males and females worldwide is 22% and 10.3%, respec- trite from bacon per capita per day was 0.19 mg, and for nitrate tively, with mortality rate of 14.3% and 8.3%, respectively. Gastric was 0.41 mg. For ham (12 g per capita per day) nitrite consumed cancer is the third leading cause of death in men after lung and per capita per day was 0.28 mg and for nitrate was 0.23 mg. Thus prostate cancer, and is the fourth leading cause of death in women combined nitrite and nitrate intake from bacon and ham per capita worldwide (Forman & Burley, 2006). Overall gastric cancer rate is per day were 0.47 and 0.64 mg, respectively. At the upper extreme, declining, especial in more developed countries, with the exception assuming 100 g of bacon or ham was consumed every day, the ni- of Miyagi prefecture of Japan still having the highest gastric cancer trite and nitrate intake from bacon would be 1.57 mg and 3.42 mg rate. Korea, East Asia, South America and Eastern Europe also sus- per capita per day, respectively. Similarly for ham the nitrite and tained a high gastric cancer rate. However, Bombay in India always nitrate intake would be 2.33 and 1.90 mg nitrite and nitrate per ca- maintained a low gastric cancer rate between 1953 and 1997 (For- pita per day, respectively, giving a total of 3.9 mg of nitrite and man & Burley, 2006). This may be due to higher consumption of 5.32 mg nitrate per capita per day. This is significantly lower than antioxidant rich fruits and vegetables and herbs and spices, which the ADI set by the European Union Scientific Committee for Food in have been shown to reduce the risk of gastric cancer. Joossens et al. 1995. (1996) studied dietary salt, nitrate and gastric cancer mortality in However, taking endogenous formation of nitrate into account, 24 countries and demonstrated that nitrate intake became an in- this means additional 70 mg of nitrate for an average 70 kg adult creased risk factor for gastric cancer when salt intake was also (Gangolli et al., 1994). Furthermore, it was estimated approxi- high. mately 25% of dietary nitrate is converted to nitrite by bacteria It was predicted that with increasing population numbers and and nitrate reductase in the oral cavity (Gangolli et al., 1994). Thus increasing longevity, it would cause a net increase in gastric cancer assuming two servings (150 g) of vegetables comes from green lea- rate worldwide. Since diagnosis often occurs between the ages of fy vegetables, again taking the analytical data from this study (Ta- 60 and 80, with up to 30% mortality rate after five years diagnosis ble 1), this means approximately 727.5 mg of nitrate from English (Forman & Burley, 2006), it is vital to make dietary and lifestyle spinach (Fig. 1) is ingested, of which 181.9 mg of nitrite can partic- changes to decrease gastric cancer rate and to increase survival ipate in nitrosation in the stomach. Based on the above assump- rate with better diagnostic facility and education. Risk factors to tions, the total nitrite and nitrate burden for an Australian adult be avoided include Helicobacter pylori infection, smoking, high con- of 70 kg body weight, the intakes per day is approximately 184.4 sumption of cured meat and salt, and low consumption of fruits and 617.7 mg, respectively, sourced from cured meat, spinach and vegetables. and endogenous nitrate formation. This exceeds the ADI of Once the nitrate and nitrite contents in food were established, 4.2 mg for nitrite by 44 times per 70 kg adult per day, and by 2.4 one can estimate the intake of these anions based on national die- times of ADI of 259 mg nitrate per 70 kg adult per day. However, tary surveys. Gangolli et al. (1994) estimated that the mean daily it must be noted that the above prediction assumed that nitrite intake of nitrate and nitrite in the US were 106 and 1.5 mg/kg, only came from one serving (50 g) each of bacon and ham respectively, and in the UK were 104 and 1.5 mg/kg, respectively. (Fig. 2), and that nitrate intake only came from two servings of In 1994, van Vliet, Vaessen, van de Burg, and Schothorst (1997) English spinach. Since English spinach had the highest nitrate con- estimated the mean intake of nitrate in the Dutch population to tent, this predicts the upper extreme of dietary nitrate intake. Fur- be 80 mg/day per person, and the median intake of nitrite to be thermore, the ADI do not include the 25% conversion of dietary 0.1 mg/day per person. In comparison, Italy had a mean daily in- nitrate to nitrite in the oral cavity, which underestimate the total take of nitrate of 245 mg/day, whereas Poland and Switzerland re- ingested dietary nitrite. corded mean daily nitrate intakes of 178 and 125 mg/day, Japan has seven times the rate of gastric cancer than the United respectively. France’s mean daily intake of nitrate and nitrite were States and is also significantly higher compared to the United King- 150.7 and >3 mg/day, respectively, followed by Netherlands, Ger- dom and Germany. There is little evidence that genetic differences many and Norway where mean daily nitrate intakes were 71, 68 contributed to the different gastric cancer rates (Davies & Sano, and 43 mg/day, respectively. The mean daily nitrite consumption 2001). Thus this epidemiological study suggests that diet and life- in those countries was 0.6, 2.6 and 1.8 mg/day, respectively (Gang- style may play an important role in gastric cancer etiology besides olli et al., 1994). According to Cornée, Lairon, Velema, Guyader, and effective screening and management. Countries such as South Kor- Berthezene (1992), the average daily nitrate intake per person per ea, Japan and China had the highest stomach cancer mortality for day was 121 mg (85% from vegetables, 5% from preserved and men, whereas countries with the highest stomach cancer mortality cured meat, and 5% from cereal products). For the average daily ni- for women were South Korea, China and Columbia. Canada and trite intake per person per day, it was found to be 1.88 mg (43% Denmark had the lowest stomach cancer mortality for men and from vegetables, 28% from cured meat, and 16% from cereals). women, respectively (Joossens et al., 1996). The remaining 13% of nitrite must come from non-dietary sources of nitrite such as atmospheric contamination. In summary, Pennington (1998) estimated that the daily nitrate intake ranges between 53 and 350 mg/day depending on the type and quantity of the vegetable consumed and the level of nitrate in drinking water. Whereas daily nitrite consumption was between 0 and 20 mg/day depending on the levels of nitrite present in cured meat and much of it was consumed. The acceptable daily intake (ADI) for nitrate was set at 3.7 mg/kg body weight by the European Union Scientific Committee for Food (1995) and since nitrite has higher acute toxicity than nitrate its ADI was set at 0.06 mg/kg body weight (Reinik et al., 2005). Based on the Australian Bureau of Statistics (Australian Bureau of Statistics, 1998–1999), Australians consumed 8.7 kg of bacon and ham combined per capita per year in 1998–1999. Assuming half of each product was consumed at 4.35 kg per capita per year Fig. 1. Mean nitrate and nitrite contents and recoveries in fresh vegetables after (or 12 g per capita per day) based on the finding in this study, ni- 5 min boiling. Values are means of at least four replicate determinations. Author's personal copy

338 J. Hsu et al. / Food Chemistry 115 (2009) 334–339

vegetables contributed to the majority of dietary nitrate and that cured meat products contributed to the majority of dietary nitrite. This study agreed with the literature that vegetables are the major dietary nitrate contributor and that meat and especially cured meat provided the majority of dietary nitrite. The extraction and detection method used in this study was demonstrated to be simple, fast, sensitive and applicable to both meat and vegetable samples. Nitrate and nitrite content of foodstuff should be monitored in the long-term to estimate the dietary intake and to provide insights into the effects of new horticultural and manufacturing technologies on the levels of nitrates and nitrite in foods and in the etiology of gastro-intestinal cancers. It was demonstrated that nitrate and nitrite content tested were below the guideline set by FSANZ and their intake were within the range reported in the Wes- tern countries and were below the ADI. This study will be the ﬁrst to quantify the levels of nitrate and nitrite in foods in the Sydney supermarket and will provide useful data to industry and health professionals. Fig. 2. Mean nitrate and nitrite contents in cured and fresh meat from Sydney supermarkets. Values are means of at least four replicate determinations. References

Amr, A., & Hadidi, N. (2001). Effect of cultivar and harvest date on nitrate (NO3) and The high gastric cancer incidence in the Far East may be due to nitrite (NO2) content of selected vegetables grown under open field and greenhouse conditions in Jordan. Journal of Food Composition and Analysis, 14, the consumption of specific foods that are high in nitrates such as 59–67. Korean Kimchi or high in salt as in many traditional Japanese Archer, M. C. (1989). Mechanisms of action of N-nitroso compounds. Cancer Surveys, dishes, or particular food preparation methods such as broiling of 8, 241–250. Australia Bureau of Statistics. (1998–1999). 16.59 Apparent per capita consumption meats (Duncan et al., 1997). Regions of high risk to gastric cancer of foodstuffs. Available from http://www.abs.com.au Accessed 25.02.08. often coincide with a low intake of foods containing vitamin C. Bartsch, H., Ohshima, H., Shuker, D., Pignatelli, B., & Calmels, S. (1990). Exposure of Other risk factors for human gastric cancer include residence in humans to endogenous N-nitroso compounds: Implications in cancer etiology. Mutation Research, 238, 255–267. areas with high nitrate-containing soil due to many factors such Butt, S. B., Riaz, M., & Iqbal, M. Z. (2001). Simultaneous determination of nitrite and as the addition of fertilizers and foods pickled with salt (Weisburg- nitrate by normal phase ion-pair liquid chromatography. Talanta, 55, 789–797. er, 1981). Certain salted fermented fish products including fish Cammack, R., Joannou, C. L., Cui, X.-Y., Martinez, C. T., Maraj, S. R., & Hughes, M. N. (1999). Nitrite and nitrosyl compounds in food preservation. Biochimica et sauce were associated with the high gastric cancer mortality in Biophysica Acta, 1411, 475–488. Fujian province of China (Chen et al., 1992). Similarly, a potential Cashel, K. M., & Greenfield, H. (1995). The effect of revised Australian food link for the high gastric cancer rate in Southwest Korea was asso- composition tables on estimates of foods and nutrients available for national ciated with regular high consumption of salted pickled cabbage consumption, 1983–1984. Journal of Food Composition and Analysis, 8, 45–61. Chen, C. S., Pignatelli, B., Malaveille, C., Bouvier, G., Shuker, D., Hautefeuille, A., et al. and salted seafood sauce (Seel et al., 1994). The former, also con- (1992). Levels of direct-acting mutagens, total N-nitroso compounds in tained high levels of total N-nitroso compound precursors, and nitrosated fermented fish products, consumed in a high-risk area for gastric cabbages, which are known to contain high levels of nitrate than cancer in southern China. Mutation Research, 265, 211–221. Cornée, J., Lairon, D., Velema, J., Guyader, M., & Berthezene, P. (1992). An estimate of any other vegetables. nitrate, nitrite, and N-nitrosodimethylamine concentrations in French food In addition of forming carcinogens in the stomach, nitrite is also products or food groups. Sciences Des Aliments, 12, 155–197. genotoxic and can readily induce methaemoglobinaemia especially Davies, P. A., & Sano, T. (2001). The difference in gastric cancer between Japan, USA and Europe: What are the facts? What are the suggestions? Critical Reviews in in babies (Gangolli et al., 1994). Furthermore, the lethal dose for ni- Oncology/Hematology, 40, 77–94. trite in adults was estimated to be between 2 and 9 g NaNO2 per Dennis, M. J., Key, P. E., Papworth, T., Pointer, M., & Massey, R. C. (1990). The day, or 33–250 mg/kg body weight (Gangolli et al., 1994), whereas determination of nitrate and nitrite in cured meat by HPLC/UV. Food Additives and Contaminants, 7, 455–461. the lethal dose for nitrate ions was estimated at 20 g per day, or Di Matteo, V., & Esposito, E. (1997). Methods for the determination of nitrite by 330 mg nitrate ions/kg body weight (Gangolli et al., 1994). high-performance liquid chromatography with electrochemical detection. Although it is unlikely to reach these toxic levels from dietary in- Journal of Chromatography A, 789, 213–219. Dionex Corporation. (1998). Determination of nitrate and nitrite in meat using take alone, the long-term effects may be detrimental based on epi- high-performance anion-exchange chromatography. Application Note, 112. demiological and clinical studies. Duncan, C., Li, H., Dykhuizen, R., Frazer, R., Johnston, P., MacKnight, G., et al. (1997). Homogenization of food is difficult but often necessary prior to Protection against oral and gastrointestinal diseases: Importance of dietary chromatographic analysis. Factors such as variable texture, struc- nitrate intake, oral nitrate reduction and enterosalivary nitrate circulation. Comparative Biochemistry and Physiology, 118A, 939–948. ture and the presence of immiscible phases may hinder the Food Standards Australia and New Zealand. (2007–2008). Food standards code. homogenization process (Lichon, 1996). Because these properties http://www.foodstandards.gov.au Accessed 10.11.08. are inherent properties of the food product, it cannot be changed Forman, D., & Shuker, D. (1997). Helicobacter pylori and gastric cancer – A case study in molecular epidemiology. Mutation Research/Fundamental and Molecular during the manufacturing process. The sampling of food and their Mechanisms of Mutagenesis, 379, S159. preparation must therefore be considered carefully prior to analy- Forman, D., & Burley, V. J. (2006). Gastric cancer: Global pattern of the disease and sis to ensure representative and accurate results. This means using an overview of the environmental risk factors. Best Practice and Research Clinical Gastroenterology, 20, 633–649. traditional cooking practices so that factors that affect nitrate and Gaiser, M., Rathjen, A., & Spiess, W. E. L. (1996). Nitrate extraction during blanching nitrite determination and their recovery are consistent and appli- of spinach. Lebensmittel-Wissenschaft und Technologie (Food Science and cable to dietary exposure of these anions. Technology), 30, 432–435. Gangolli, S. D., van den Brandt, P., Feron, V. J., Janzowsky, C., Koeman, J. H., Speijers, G. J., et al. (1994). Nitrate, nitrite and N-nitroso compounds. European Journal of 4. Conclusion Pharmacology, Environmental Toxicology and Pharmacology Section, 292, 1–38. Hunt, J. (1994). A method for measuring nitrite in fresh vegetables. Food Additives and Contaminants, 11, 317–325. Different authors attributed different percentage of dietary ni- Hunt, J., & Turner, M. K. (1994). A survey of nitrite concentrations in retail fresh trate and nitrite to the major food groups, but the consensus is that vegetables. Food Additives and Contaminants, 11, 327–332. Author's personal copy

J. Hsu et al. / Food Chemistry 115 (2009) 334–339 339

Joossens, J. V., Hill, M. J., Elliott, P., Stamler, R., Stamler, J., Lesaffre, E., et al. (1996). Seel, D. J., Kawabata, T., Nakamura, M., Ishibashi, T., Hamano, M., Mashimo, M., et al. Dietary salt, nitrate and stomach cancer mortality in 24 countries. International (1994). N-nitroso compounds in two nitrosated food products in southwest Journal of Epidemiology, 25, 494–504. Korea. Food and Chemical Toxicology, 32, 1117–1123. Lichon, M. J. (1996). ‘Sample preparation’. In L. M. L. Nollet (Ed.). Handbook of food Siu, D. C., & Henshall, A. (1998). Ion chromatography determination of nitrate and analysis: Physical characterization and nutrient analysis (Vol. 1). New York: nitrite in meat products. Journal of Chromatography A, 804, 157–160. Marcel Dekker Inc. Stalikas, C. D., Konidari, C. N., & Nanos, C. G. (2003). Ion chromatographic method for Marshall, P. A., & Trenerry, V. C. (1996). The determination of nitrite and nitrate in the simultaneous determination of nitrite and nitrate by post-column indirect foods by capillary ion electrophoresis. Food Chemistry, 57, 339–345. ﬂuorescence detection. Journal of Chromatography A, 1002, 237–241. Ologhobo, A. D., Adegede, H. I., & Maduagiwu, E. N. (1996). Occurrence of nitrate, Usher, C. D., & Telling, G. M. (1975). Analysis of nitrates and nitrite in foodstuffs: A nitrite and volatile nitrosamines in certain feedstuffs and animal products. critical review. Journal of the Science of Food and Agriculture, 2, 1793–1805. Nutrition and Health, 11, 109–114. van Vliet, J. J., Vaessen, H. A., van de Burg, G., & Schothorst, R. C. (1997). Twenty- Öztekin, N., Nutku, M. S., & Erim, F. B. (2002). Simultaneous determination of nitrite four-hour duplicate diet study 1994; nitrate and nitrite: Method development and nitrate in meat products and vegetables by capillary electrophoresis. Food and intake per person per day. Cancer Letters, 114, 305–307. Chemistry, 76, 103–106. Walters, C. L. (1980). The exposure of humans to nitrite. Oncology, 37, 289–296. Pennington, J. A. (1998). Dietary exposure models for nitrates and nitrites. Food Weisburger, J. H. (1981). ‘N-nitroso compounds: Diet and cancer trends’, in N- Control, 9, 389–395. nitroso Compounds. In R.A. Scanlan, S.R. Tannenbaum (Eds.), American chemical Reinik, M., Tamme, T., Roasto, M., Juhkam, K., Jurtšenko, S., Tenn´ o, T., et al. (2005). society symposium series 174, United States. Nitrites, nitrates and N-nitrosamines in Estonian cured meat products: Intake Zanardi, E., Dazzi, G., Madarena, G., & Chizzolini, R. (2002). Comparative study on by Estonia children and adolescents. Food Additives and Contaminants, 22, nitrite and nitrate ions determination. Annali Facolta di medicina veterinaria Di 1098–1105. Parma, 22, 79–86.