SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS (addendum) First draft prepared by

Transcription

1 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS (addendum) First draft prepared by Dr A. Mattia Division of Biotechnology and GRAS Notice Review, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, Food & Drug Administration, College Park, Maryland, USA Evaluation Introduction Estimated daily intake Absorption, distribution, metabolism and elimination Application of the Procedure for the Safety Evaluation of Flavouring Agents Consideration of combined intakes from use as flavouring agents Consideration of secondary components Conclusion Relevant background information Additional considerations on intake Biological data Biochemical data: absorption, distribution, metabolism and excretion Toxicological studies Acute toxicity Short-term studies of toxicity Genotoxicity References EVALUATION 1.1 Introduction The Committee evaluated 12 flavouring agents that included simple aliphatic and aromatic sulfides and thiols (see Table 1) by the Procedure for the Safety Evaluation of Flavouring Agents (see Figure 1, Introduction). At its fifty-third meeting, the Committee evaluated 137 other members of this chemical group of flavouring agents (Annex 1, reference 143). The group was divided into 12 subgroups on the basis of the position of the sulfur atom, in order to facilitate the assessment of the relevant data on metabolism and toxicity. All 137 substances in that group were concluded to be of no safety concern on the basis of currently estimated levels of intake. Of the 12 additional flavouring agents considered in this monograph addendum, six agents are thiols with oxidized side chains (subgroup v) (Nos ) and contain an additional alcohol, aldehyde, ketone, or ester functional group. Two agents are acyclic sulfides with oxidized side-chains (subgroup ii) (Nos 1297 and 1298) in which an alcohol or ester functional group is present. The remaining four 419

2 Table 1. Summary of safety evaluations of simple aliphatic and aromatic sulfides and thiols used as flavouring agents a,b,c Flavouring agent No. CAS No. and Step B3 d Step B4 Comments Conclusion structure Does intake Adequate margin of safety on predicted based exceed the for substances or related metabolism on current threshold for substances? intake human intake? Subgroup ii Acyclic sulfides with oxidized side chains Structural class I 2-(Methylthio)ethanol No Yes. The NOEL of 1.4 mg/kg See notes 7 No safety S Europe: 1 bw per day (Cox et al., 1979) and 2 concern USA: 0.9 for the related substance 2-(methylthiomethyl)-3- phenylpropenal (No. 505) is > times the estimated daily intake of 2-(methylthio)ethanol when used as a flavouring agent Ethyl 5-(methylthio)valerate No Yes. The NOEL of 1.4 mg/kg bw See notes 5 No safety O Europe: 2 per day (Cox et al., 1979) for the and 7 concern S USA: 2 related substance 2- (methylthiomethyl)-3- phenylpropenal (No. 505) is > times the estimated daily intake of ethyl 5- (methylthio)valerate when used as a flavouring agent Subgroup iii Cyclic sulfides Structural class III spiro(2,4-dithia-1-methyl No Yes. The NOEL of 25 mg/kg See notes No safety oxabicyclo(3.3.0)octane-3,3 - Europe: ND bw per day for spiro(2,4-dithia- 10, 1 and 3 concern (1 -oxa-2 -methyl)- USA: 2 1-methyl-8-oxabicyclo(3.3.0)octane- S O cyclopentane) 3,3 -(1 -oxa-2 -methyl)- S O OH O cyclopentane) is > times the estimated daily intake when used as a flavouring agent

3 Table 1. (contd) Flavouring agent No. CAS No. and Step B3 d Step B4 Comments Conclusion structure Does intake Adequate margin of safety on predicted based exceed the for substances or related metabolism on current threshold for substances? intake human intake? Subgroup v Thiols with oxidized side chains Structural class I erythro- and threo-3-mercapto SH No Yes. The NOEL of 0.7 mg/kg See notes 1 No safety 2-methylbutanol Europe: 1 bw per day (Cox et al., 1974) for and 2 concern OH USA: 2 the related substance 2-mercapto- 3-butanol (No. 546) is > times the estimated daily intake of erythro- and threo-mercapto-2- methylbutan-1-ol when used as a flavouring agent (±)2-Mercapto No Yes. The NOEL of 0.7 mg/kg See notes 1 No safety methylpentan-1-ol HS Europe: 3 bw per day (Cox et al., 1974) for and 2 concern OH USA: 4 the related substance 2-mercapto- 3-butanol (No. 546) is > times the estimated daily intake of (±)2- mercapto-2-methylpentan-1-ol when used as a flavouring agent 3-Mercapto-2-methylpentan No Yes. The NOEL of 0.7 mg/kg See notes 1 No safety 1-ol (racemic) Europe: 1 bw per day (Cox et al., 1974) for and 2 concern OH USA: 0.7 the related substance 2-mercapto- SH 3-butanol (No. 546) is > times the estimated daily intake of 3-mercapto-2-methylpentan-1-ol (racemic) when used as a flavouring agent

4 Table 1. (contd) Flavouring agent No. CAS No. and Step B3 d Step B4 Comments Conclusion structure Does intake Adequate margin of safety on predicted based exceed the for substances or related metabolism on current threshold for substances? intake human intake? 3-Mercapto-2-methylpentanal No Yes. The NOEL of 0.7 mg/kg See notes 1 No safety Europe: 3 bw per day (Cox et al., 1974) for and 4 concern H USA: 4 the related substance 2-mercapto- SH 3-butanol (No. 546) is > times the estimated daily intake of 3-mercapto-2-methylpentenal when used as a flavouring agent 4-Mercapto-4-methyl No Yes. The NOEL of 1.9 mg/kg bw See notes 1 No safety pentanone Europe: 0.01 per day (Morgareidge, 1971) for the and 3 concern SH O USA: 0.02 related substance 3-mercapto-2- pentanone (No. 560) is > times the estimated daily intake of 4-mercapto-4-methyl-2-pentanone when used as a flavouring agent (±)Ethyl 3-mercaptobutyrate No Yes. The NOEL of 0.7 mg/kg bw See notes 1 No safety O SH Europe: 4 per day (Cox et al., 1974) for the and 5 concern O O USA: 4 related substance 2-mercapto-3- butanol (No. 546) is > times the estimated daily intake of (±)ethyl 3-mercaptobutyrate when used as aflavouring agent

5 Table 1. (contd) Flavouring agent No. CAS No. and Step B3 d Step B4 Comments Conclusion structure Does intake Adequate margin of safety on predicted based exceed the for substances or related metabolism on current threshold for substances? intake human intake? Subgroup vii Simple disulfides Structural class I 2,3,5-Trithiahexane No Yes. The NOEL of 0.3 mg/kg bw See notes 7, No safety Europe: 0.03 per day for the related substance 8 and 9 concern USA: methyl-1,2,4-trithiane (No. 574) S S S (Mondino, 1981) is > times the estimated daily intake of 2,3,5- trithiahexane when used as a flavouring agent Subgroup ix Trisulfides and polysulfides Structural class I Diisopropyl trisulfide No Yes. The NOEL of 4.8 mg/kg bw See notes 7, No safety Europe: per day (Morgareidge & Oser, 8 and 9 concern USA: ) for the related substance S S S dipropyltrisulfide (No. 585) is > times the estimated daily intake of diisopropyl trisulfide when used as a flavouring agent Subgroup xi Thioesters Structural class I Ethyl 4-(acetylthio)butyrate No Yes. The NOEL of 6.5 mg/kg bw See notes 1, No safety O Europe: 4 per day (Shellenberger, 1970) 5 and 6 concern O S O USA: 4 for the related substance ethylthioacetate (No. 483) is > times the estimated daily intake of ethyl 4-(acetylthio)butyrate when used as a flavouring agent

6 Notes to Table 1 CAS, Chemical Abstracts Service; ND, no data on intake reported a One hundred and thirty-seven (137) flavouring agents in this group were previously evaluated by the Committee (Annex 1, reference 144). To facilitate the evaluations, the group was divided into 12 subgroups based on the position of the sulfur atom. The subgroup designations are indicated in the table b Step 1: Eleven flavouring agents are in structural class I and one (No. 1296) is in structural class III c Step 2: All of the agents in this group cannot be predicted to be metabolized to innocuous products d The threshold for human intake for structural class I, II and III are 1800, 540 and 90mg/day, respectively. All intake values are expressed in mg/day. The combined per capita intake of the 11 flavouring agents in structure class I is approximately 21mg per day in Europe and 24mg per day in the USA. The combined per capita intake of the remaining flavouring agent in structural class III is 2mg per day in the USA. The cumulative per capita intake for the amended group as a whole including the 137 substances in the original evaluation and the 12 additional substances is 1181 and 1034mg/person per day in Europe and the USA, respectively Notes: 1. Sulfur is expected to be oxidized to sulfonic acid, undergo alkylation and conjugation followed by excretion 2. The hydroxy group is expected to undergo oxidation to the carboxylic acid 3. The ketone group is expected to be reduced to the alcohol, conjugated and subsequently excreted 4. The aldehyde group is expected to be oxidized to the corresponding carboxylic acid, conjugated and subsequently excreted 5. The ester is expected to undergo hydrolysis to the corresponding carboxylic acid and alcohol 6. The thioester is expected to undergo hydrolysis to acetate and the corresponding thiol, which will be further oxidized 7. The sulfur is expected to be oxidized to the sulfoxide and sulfone 8. The di- or trisulfides are expected to be reduced to free thiols 9. Free thiols may form mixed disulfides with glutathione or cysteine 10. Thioketal will hydrolyse to liberate the corresponding ketone and dithiol

7 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS 425 substances are a thioester (subgroup xi) (No. 1295), a disulfide (subgroup vii) (No. 1299), a trisulfide (subgroup ix) (No. 1300) and a cyclic sulfide (subgroup iii) (No. 1296). None of these agents has been evaluated previously. Seven of the 12 flavouring agents in this group are naturally occurring components of food (Nos , 1297, 1299, 1300) and have been detected in onions, fruits, broccoli, cabbage, cauliflower, hop oil, wine, fish and cheese (Darriet et al., 1995; Private communication to FEMA, ; Maarse et al., 1999; Kendrick, 2000). 1.2 Estimated daily intake The total annual volume of production of the 12 simple aliphatic and aromatic sulfides and thiols is approximately 150kg in Europe and in the USA (Private communication to FEMA, ; Lucas et al., 1999). The daily per capita intake of each agent is reported in Table 1. Annual volumes of production of this group of flavouring agents are summarized in Table Absorption, distribution, metabolism and elimination All of the sulfur-containing flavouring agents considered in this addendum are of low relative molecular mass and are sufficiently lipophilic to be absorbed. These flavouring agents are expected to be metabolized through various pathways, described below and in the previous evaluation by the Committee (Annex 1, reference 143). Thiols with oxidized side-chains (Nos ) The metabolism of thiols with oxidized side-chains is predicted to involve a combination of pathways for simple thiols together with further oxidation or conjugation of the oxidized side-chain. Metabolic options for simple thiols include oxidation to form unstable sulfenic acids (RSOH) which are oxidized to sulfinic acids (RSO 2 H), undergo methylation to yield methyl sulfides which then form sulfoxides and sulfones, react with endogenous thiols to form mixed disulfides, are conjugated with glucuronic acid, or undergo oxidation of the a-carbon which results in desulfuration and the formation of an aldehyde (Dutton & Illing, 1972; McBain & Menn, 1969; Maiorino et al., 1988; Richardson et al., 1991). Acyclic sulfides with oxidized side-chains (Nos 1297 and 1298) The presence of oxygenated functional groups, such as an alcohol (No. 1297) or ester (No. 1298), provides additional sites for biotransformation of sulfides (thioethers), and the presence of these polar sites would result in increased renal excretion of these agents. The biotransformation of such oxygenated groups is well characterized and has been described for groups of flavouring agents evaluated previously by the Committee (Annex 1, references 131, 132, 138, 144). Simultaneous metabolism of sulfur and oxygenated functional groups has been reported for various substrates (Fatih et al., 1988; Gachon et al., 1988; Feng & Solsten,

8 426 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS Table 2. Annual volumes of production of simple aliphatic and aromatic sulfides and thiols used as flavouring agents Agent (No.) Most recent Intake b Intake of Annual Consumption annual ( eaters only ) alcohol volume in ratio e volume equivalents naturally (kg) a mg/day mg/kg bw mg/kg bw occurring per day per day c foods (kg) d erythro and threo-mercapto-2-methylbutan-1-ol (1289) Europe USA NA (±)2-Mercapto-2-methylpentanol (1290) Europe USA NA 3-Mercapto-2-methylpentan-1-ol (racemic) (1291) Europe USA f NA 3-Mercapto-2-methylpentanal (1292) Europe USA f NA 4-Mercapto-4-methyl-2-pentanone (1293) Europe j USA j g NA (±)Ethyl 3-mercaptobutyrate (1294) Europe USA h NA Ethyl 4-(acetylthio)butyrate (1295) Europe USA NA spiro(2,4-dithia-1-methyl-8-oxabicyclo(3.3.0)octane-3,3 -(1 -oxa-2 -methyl)cyclopentane (1296) Europe ND ND ND USA i NA 2-(Methylthio)ethanol (1297) Europe USA NA Ethyl 5-(methylthio)valerate (1298) Europe USA NA 2,3,5-Trithiahexane (1299) Europe k USA k NA Diisopropyl trisulfide (1300) Europe l USA l NA

9 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS 427 Table 2. (contd) Agent (No.) Most recent Intake b Intake of Annual Consumption annual ( eaters only ) alcohol volume in ratio e volume equivalents naturally (kg) a mg/day mg/kg bw mg/kg bw occurring per day per day c foods (kg) d Total Europe USA NA, not available; ND, no intake data reported; +, reported to occur naturally in foods (Maarse et al., 1999), but no quantitative data; -, not reported to occur naturally in foods a The volumes cited, unless otherwise indicated, are anticipated annual volumes (Private communication to FEMA ), which were the maximum amount of the flavouring agent estimated to be used annually in both Europe and the USA by the manufacturer at the time the material was proposed for use as a flavouring agent b Intake (mg/person per day) was calculated as follows: [(annual volume, kg) ( mg/kg)]/[population survey correction factor 365 days], where population (10%, eaters only ) = for Europe and for the USA; where correction factor = 0.6 for Europe and USA anticipated volumes and 0.8 for the Lucas et al. survey in the USA, representing the assumption that only 60% and 80% of the annual flavour volume, respectively, was reported in the poundage surveys (International Organization of the Flavor Industry, 1995; Lucas et al., 1999; Private communication to FEMA ) Intake (mg/kg bw/d) calculated as follows: [(mg/person per day)/body weight], where body weight = 60 kg. Slight variations may occur from rounding c Calculated as follows: (molecular weight of alcohol/molecular weight of ester) daily per capita intake ( eaters only ) of ester d Quantitative data for the USA reported by Stofberg & Grundschober (1987) e The consumption ratio was calculated as follows: (annual consumption via food, kg)/(most recent reported volume as a flavouring agent, kg) f Natural occurrence data reported in a private communication to FEMA ( ) g Darriet et al., 1995 h Kendrick, 2000 i Annual volume reported in the 1995 FEMA Poundage Survey (Lucas et al., 1999) j Intake calculated based on 1% solution of 4-mercapto-4-methyl-2-pentanone in propylene glycol k Intake calculated based on a 10% solution of 2,3,5-trithiahexane in triacetin l Intake calculated based on a 2% solution of diisopropyl trisulfide in triacetin 1991; Wilson et al., 1991; Black et al., 1993). Sulfoxide formation is usually the predominant metabolic detoxication pathway. Cyclic sulfides (No. 1296) Cyclic sulfides can be expected to undergo extensive S-oxidation by the cytochrome P450 superfamily to produce the corresponding sulfoxides.

10 428 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS Simple disulfides (No. 1299) The reduction of xenobiotic disulfides is believed to be extensive and can be catalysed enzymatically, by glutathione reductase (Waring, 1996) or thioltransferases (Wells et al., 1993), as well as chemically, by exchange with glutathione, thioredoxin, cysteine or other endogenous thiols. Reduction of non-cyclic disulfides (No. 1299) would result in the formation of thiols of low molecular mass, which are metabolized via the various pathways described above for simple thiols. Trisulfides (No. 1300) The trisulfide of glutathione is labile and readily converted to the disulfide, with the release of sulfur as hydrogen sulfide (Moutiez et al., 1994). Trisulfides are predicted to be converted rapidly to the corresponding disulfides with subsequent reduction to thiols, which would then be metabolized via the various pathways described above for simple thiols. Thioesters (No. 1295) Thioesters are hydrolysed by lipase and esterases (Kurooka et al., 1976); the rate of hydrolysis increases as the length of the carbon chain increases and decreases as the oxygenation of the carbon chain in the thiol moiety increases (Greenzaid & Jenks, 1971). After hydrolysis, the resulting alcohol and carboxylic acid would participate in the metabolic pathways described above for sulfides containing oxygenated functional groups. 1.4 Application of the procedure for the safety evaluation of flavouring agents Step B1. In applying the Procedure for the Safety Evaluation of Flavouring Agents to these 12 flavouring agents, the Committee assigned 11 agents (Nos , ) to structural class I. The remaining flavouring agent (No. 1296) was assigned to class III (Cramer et al., 1978). Step B2. At currently estimated levels of intake, none of the flavouring agents in this group is predicted to be metabolized to innocuous products. The evaluation of these substances therefore proceeded via the B- side of the decision-tree. Step B3. The estimated daily per capita intakes of the 11 flavouring agents in this group in structural class I are below the threshold of concern (i.e mg). The estimated daily per capita intake for the one flavouring agent in structural class III is below the threshold of concern (i.e. 90 mg). Accordingly, the evaluation of all 12 agents in the group proceeded to step B4. Step B4. For erythro- and threo-3-mercapto-2-methylbutanol (No. 1289), the NOEL of 0.7mg/kg body weight per day for the structurally related substance 2-mercapto-3-butanol (No. 546) from a 92-day study in rats fed by gavage (Cox et al., 1974) provides an adequate margin of safety

11 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS 429 (>10000) in relation to known levels of intake of this agent. This NOEL is also appropriate for the structurally related agents (±)-2-mercapto-2- methylpentan-1-ol (No. 1290), 3-mercapto-2-methylpentan-1-ol (racemic) (No. 1291), 3-mercapto-2-methylpentanal (No. 1292), and (±)- ethyl 3-mercaptobutyrate (No. 1294), because they are all acyclic thiols with oxidized side-chains that are anticipated to undergo oxidation or hydrolysis and subsequent metabolism via similar metabolic pathways. For 4-mercapto-4-methyl-2-pentanone (No. 1293), the NOEL of 1.9 mg/kg bw per day for the structurally related substance 3-mercapto-2- pentanone (No. 560) administered to rats by gavage in a 92-day study (Morgareidge, 1971) provides an adequate margin of safety (>10 000) in relation to known levels of intake of this agent. For ethyl 4-(acetylthio)butyrate (1295), the NOEL of 6.5 mg/kgbw per day reported in a 13-week study in rats (Shellenberger, 1970) fed with the structurally related substance ethylthioacetate (No. 483) provides an adequate margin of safety (>10 000) in relation to known levels of intake of this agent. For ethyl 2-(methylthio)ethanol (No. 1297), the NOEL of 1.4mg/kgbw per day reported in a 13-week study in rats (Cox et al., 1979) fed by gavage with the structurally related substance 2-(methylthiomethyl)-3- phenylpropenal (No. 505) provides an adequate margin of safety (>10 000) in relation to known levels of intake of this agent. This NOEL is also appropriate for the structurally related agent ethyl 5-(methylthio)valerate (No. 1298), which is also an acyclic sulfide with an oxidized side-chain that is anticipated to undergo oxidation and subsequent metabolism via similar pathways. For 2,3,5-trithiahexane (No. 1299), the NOEL of 0.3mg/kgbw per day reported in a 13-week study (Mondino, 1981) in rats fed with the structurally related substance 3-methyl-1,2,4-trithiane (No. 574) provides an adequate margin of safety (>10 000) in relation to known levels of intake of this agent. For diisopropyl trisulfide (No. 1300), the NOEL of 4.8mg/kgbw per day reported in a 13-week study (Morgareidge & Oser, 1970) in rats fed by gavage with the structurally related substance dipropyltrisulfide (No. 585) provides an adequate margin of safety (> ) in relation to known levels of intake of this agent. For spiro(2,4-dithia-1-methyl-8-oxabicyclo(3.3.0)octane-3,3 -(1 -oxa- 2 -methyl)-cyclopentane) (No. 1296), the NOEL of 25 mg/kg bw per day in the diet reported in a 13-week study in rats (Wheldon et al., 1970) provides an adequate margin of safety (>100000) in relation to known levels of intake of this agent. Table 1 summarizes the evaluations of the 12 simple aliphatic and aromatic sulfides and thiols in this group.

12 430 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS 1.5 Consideration of combined intakes from use as flavouring agents In the unlikely event that the 11 agents considered in this evaluation and the 97 agents considered previously in structural class I were to be consumed concurrently on a daily basis, the estimated combined intake would not exceed the daily per capita human intake threshold for class I (1800mg). In the unlikely event that the one agent considered in this evaluation and the six agents considered previously in structural class III were to be consumed concurrently on a daily basis, the estimated combined daily per capita intake would not exceed the human intake threshold for class III (90 mg). 1.6 Consideration of secondary components One member of this group of flavouring agents (No. 1293, 4-mercapto-4- methyl-2-pentanone) has a minimum assay value of <95%. Information on the safety of the secondary component of this compound is summarized in Annex 6 (Summary of the safety evaluation of secondary components of flavouring agents with minimum assay values of less than 95%). The secondary component (4- methyl-3-penten-2-one) was evaluated by the Committee at its fifty-ninth meeting, and was considered not to present a safety concern at current levels of intake. 1.7 Conclusion The Committee concluded that these 12 flavouring agents, which are additions to the group of simple aliphatic and aromatic sulfides and thiols evaluated previously, would not give rise to safety concerns at the currently estimated levels of intake. 2. RELEVANT BACKGROUND INFORMATION 2.1 Additional considerations on intake There is no additional information on intake. 2.2 Biological data Biochemical data: absorption, distribution, metabolism and excretion No significant changes in the absorption, distribution, metabolism, and excretion of these agents have been reported since the submission of the original monograph (Annex 1, reference 143). Some references not included in the original group monograph are cited in section 1.3.

13 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS 431 Table 3. Results of short-term studies of toxicity with simple aliphatic and aromatic sulfides and thiols No. Flavouring agent Species; No. of test Route Duration NOEL Reference sex groups a /no. (days) (mg/kg per group b bw per day) 1296 spiro(2,4-dithia-1- Rat; M 3/15 Diet Wheldon methyl-8-oxabicyclo et al. (3.3.0)octane-3,3 - (1970) (1 -oxa-2 -methyl) cyclopentane) M, male a Total number of test groups does not include control animals b Total number per test group includes both male and female animals Toxicological studies (a) Acute toxicity There are no additional studies of acute toxicity. No significant changes in the acute toxicity of flavouring agents in this group have been reported since the submission of the original monograph (Annex 1, reference 143). (b) Short-term studies of toxicity The results of one additional 90-day study are described below and are summarized in Table 3. (i) Spiro(2,4-Dithia-1-methyl-8-oxabicyclo(3.3.0)octane-3,3 -(1 oxa-2 -methyl)cyclopentane) (No. 1296) In a 90-day study, groups of 15 male Cfy Wistar rats were fed diets containing spiro(2,4 - dithia-1-methyl-8-oxabicyclo(3.3.0)octane-3,3 -(1 -oxa-2 methyl)cyclopentane) at a concentration of 0, 500, 5000 or 5000 ppm (raised to ppm after week 1 and to 20000ppm at week 6) (Wheldon et al., 1970). The three lower concentrations correspond to estimated intakes of 0, 25 and 250mg/kg bw per day, respectively (Food and Drug Administration, 1993). Rats in the group receiving the highest dose were initially given an estimated intake of 250 mg/kg bw per day, which was increased to 500mg/kgbw per day after week 1, and finally to 1000 mg/kg bw per day at week 6. Clinical observations were performed daily. Animals were housed five per cage and had access to food and water ad libitum. A decrease in food consumption, which was presumably a result of the unpalatability of the diet containing the malodorous sulfur-containing test agents, was reported in all treated groups throughout the study. Weekly measurement of body weights showed a decrease in body-weight gain at the two higher doses throughout the study, and at the lowest dose only after 8 weeks. The calculated

14 432 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS Table 4. Results of studies of genotoxicity with simple aliphatic and aromatic sulfides and thiols No. Flavouring agent End-point Test system Concentration Results Reference In vitro 1289 Erythro- and Reverse S. typhimurium mg/ Negative a Gocke threo-3-mercapto- mutation TA1535, TA97, plate (1997) 2-methylbutanol TA98, TA100, TA102 a With and without metabolic activation from S9 food conversion efficiency was significantly reduced only at the two higher doses. Haematological examination revealed decreases in erythrocyte volume fraction (p < 0.01), concentration of haemoglobin (p < 0.001) and erythrocyte count (p < 0.001), and increases in corpuscular volume (p < 0.001), neutrophil (p < 0.001) and lymphocyte counts (p < 0.01) at week 13 in the group given the highest dose. Similar changes were also reported for the group given the intermediate dose, but the decreases in erythrocyte volume fraction and differential leukocyte counts were not statistically significant. The animals in the group given the highest dose also showed abnormal blood pigmentation. Haematological measurements for the group given the low dose and for the control group were comparable throughout the study. At necropsy, measurement of organ weights (adrenals, heart, kidneys, liver, lungs, spleen, testes, and thyroid) revealed increased absolute and relative spleen weights at the highest dose. Increases in relative (to body weights) but not absolute organ weights at the two higher doses were associated with the significant decrease in body weights reported for these two groups. Histopathological examination of 13 different tissues revealed pigmented macrophages in the liver and spleen of the animals in the group receiving the highest dose. The authors reported a NOEL of 250 mg/kgbw per day (Wheldon et al., 1970). Owing to the decreased body-weight gain and significantly reduced food conversion efficiency reported at this dose, however, a NOEL of 25 mg/kgbw per day is more conservative. The results of this study are summarized in Table 3. (b) Genotoxicity The results of one additional test for genotoxicity in vitro are described below and summarized in Table 4 (Annex 1, reference 144). Erythro- and threo-3-mercapto-2-methylbutanol (No. 1289) ( mg/plate) was evaluated for mutagenic activity in the modified Ames test with preincubation in the presence and absence of metabolic activation in Salmonella typhimurium strains TA1535, TA97, TA98, TA100 and TA102. No genotoxic effects were observed (Gocke, 1997).

15 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS REFERENCES Black, R.M., Brewster, K., Clarke, R.J., Hambrook, J.L., Harrison, J.M. & Howells, D.J. (1993) Metabolism of thioglycol (2,2 -thiobis-ethanol): Isolation and identification of urinary metabolites following intraperitoneal administration to rats. Xenobiotica, 23, Cox, G.E., Bailey, D.E. & Morgareidge, K. (1974) 90-Day feeding studies in rats with compound (2-mecapto-3-butanol). Unpublished report by Food & Drug Research Laboratories, Inc. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States of America, Washington, DC, USA. Cox, G.E., Rucci, G. & Babish, J.G. (1979) 90-Day subacute dietary toxicity study of in Sprague-Dawley rats. Unpublished report by Food & Drug Research Laboratories, Inc. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States of America, Washington, DC, USA. Cramer, G.M., Ford, R.A. & Hall, R.L. (1978) Estimation of toxic hazard a decision tree approach. Food. Cosmet. Toxicol., 16, Darriet, P., Tominaga, T., Lavigne, V., Boidron, J.-N. & Dubourdieu, D. (1995) Identification of a powerful aromatic component of Vitis vinifera L. var. Sauvignon wines: 4-mercapto- 4-methylpentan-2-one. Flavour and Fragrance Journal, 10, Dutton, G.J. & Illing, H.P.A. (1972) Mechanism of biosynthesis of thio-beta-d-glucosides. Biochem. J., 129, Fatih, E., Karim, I.A., Millership, J.S., Temple, D.J. & Woolfson, A.D. (1988) An investigation of the metabolism of S-carboxymethyl-L-cysteine in man using a novel HPLC-ECD method. Eur. J. Drug Metab. Pharmacokin., 13, 253. Feng, P.C.C. & Solsten, R.T. (1991) In vitro transformation of dithiopyr by rat liver enzymes: Conversion of methylthioesters to acids by oxygenases. Xenobiotica, 21, Gachon, F., Nicolas, C., Maurizis, C., Verny, M., Chabard, J.L., Faurie, M. & Gaillard, G. (1988) Disposition and metabolism of letosteine in rats. Drug Metab. Dispos., 253. Gocke, E. (1997) Evaluation of the mutagenic potential of Ro /000 in the Ames test (Study No.: 134M97). Private communication to the Flavor and Extract Manufacturers Association. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States of America, Washington, DC, USA. Greenzaid, P. & Jenks, W.P. (1971) Pig liver esterase. Reactions with alcohols, structurereactivity correlations, and the acyl-enzyme intermediate. Biochemistry, 10, International Organization of the Flavor Industry (1995). European inquiry on volume use. Unpublished report. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington DC, USA. Kendrick, L. (2000) Identification of ethyl 3-mercaptobutyrate in mango. Unpublished report. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington, DC, USA. Kurooka, S., Hashimoto, M., Tomita, M., Maki, A. & Yoshimura, Y. (1976) Relationship between the structure of S-acyl thiol compounds and their rates of hydrolysis by pancreatic lipase and hepatic carboxylic esterase. J. Biochem., 79, Lucas, C.D., Putnam, J.M. & Hallagan, J.B. (1999) Flavor and Extract Manufacturers Association of the United States 1995 Poundage and Technical Effects Update Survey, Washington DC: Flavor and Extract Manufacturers Association of the United States. Maarse, H., Visscher, C.A., Willemsens, L.C. & Boelens, M.H. (1999) Volatile Components in Food Qualitative and Quantitative Data. Centraal Instituut Voor Voedingssonderzioek TNO., Zeist, The Netherlands.

16 434 SIMPLE ALIPHATIC AND AROMATIC SULFIDES AND THIOLS Maiorino, R.M., Bruce D.C., & Aposhian H.V. (1988) Determination and metabolism of dithiol chelating agents VI. Isolation and identification of the mixed disulfides of meso-2,3- dimercaptosuccinic acid with l-cysteine in human urine. Toxicol. Appl. Pharmacol., 97, 338. McBain, D.A. & Menn, J.J. (1969) S-methylation, oxidation, hydroxylation, and conjugation of thiophenol in the rat. Biochem. Pharmacol., 18, Mondino, A. (1981) Thirteen week repeated dose study of the test article TT191 (3-methyl- 1,2,4-trithiane) orally administered to Sprague-Dawley Charles River CD(SD)BR rats. Unpublished report from Instituto di richierche biomedicine Antoine Marxer SpA. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington, DC, USA. Morgareidge, K. (1971) 90-Day feeding study with 2-keto-3-pentanthiol in rats. Unpublished report. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington, DC, USA. Morgareidge, K. & Oser, B.L. (1970) 90-Day feeding studies in rats with dipropyltrisulfide (30204). Unpublished report by Food & Drug Research Laboratories, Inc. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington, DC, USA. Moutiez, M., Aumercier, M., Tessier, E., Parmentier, B., Tartar, A. & Sergheraert, C. (1994) Reduction of a trisulfide derivative of glutathione by glutathione reductases. Biochem. Biophys. Res. Commun., 202, Private communication to the Flavor and Extract Manufacturers Association (FEMA). ( ) Submitted to WHO by FEMA, Washington, DC, USA. Richardson K.A., Edward V.T., Jones B.C., & Hutson D.H. (1991) Metabolism in the rat of a model xenobiotic plant metabolite S-benzyl-N-malonyl-l-cysteine. Xenobiotica, 21, 371. Shellenberger, T.E. (1970) Subacute toxicity evaluation of ethyl thioacetate in rats: Final report. Unpublished report GSRI Project NC-373 from Gulf South Research Institute Association. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington, DC, USA. Stofberg, J. & Grundschober, F. (1987) Consumption ratio and food predominance of flavoring materials. Perfumer Flavorist, 12, 27. Waring, R.H. (1996) Sulfur sulfur compounds. In: Mitchell, S.C., ed., Biological Interactions of Sulfur Compounds. London: Taylor & Francis, pp Wells, W.W., Yang, Y., Diets, T.L. & Gan, Z.R. (1993) Thioltransferases. Adv. Enzymol. Relat. Areas Mol. Biol., 66, Wheldon, G.H., Amyes, S.J., Street, A.E., Hague, P.H. & Mawdesley-Thomas, L.E. (1970) Toxicity of Wa4295 [spiro(2,4-dithia-1-methyl-8-oxabicyclo(3.3.0)octane-3,3 -(1 -oxa-2 methyl)-cyclopentane)], Sa927, Stl3048 and Wa3328. Unpublished report from Huntingdon Research Centre. Submitted to WHO by the Flavor and Extract Manufacturers Association of the United States, Washington, DC. Wilson, J.E., Chissick, H., Fowler, A.M., Frearson, F.J., Gittins, M. & Swinbourne, F.J. (1991) Metabolism of benzothiazole I. Identification of ring-cleavage products. Xenobiotica, 21, 1179.

17 THE SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES First draft prepared by Dr I. C. Munro 1 and Dr A. Mattia 2 1 CANTOX Health Sciences International, Mississauga, Ontario, Canada 2 Division of Biotechnology and GRAS Notice Review, Office of Food Additive Safety, Center for Food Safety and Applied Nutrition, United States Food & Drug Administration, Maryland, USA Introduction Safety evaluation of natural flavouring complexes Specifications for natural flavouring complexes Estimates of daily intake Proposed modification to the current Procedure for the Safety Evaluation of Flavouring Agents Conclusions and recommendations References INTRODUCTION For nearly 40 years, the Joint FAO/WHO Expert Committee on Food Additives (JECFA) has played a leading role in the development of principles and procedures for the safety evaluation of flavouring agents. In 1996, the Committee began an ongoing programme to conduct safety evaluations and establish specifications for individual flavouring agents. On the basis of this experience, the Committee recognizes that the safety evaluation of flavouring agents presents unique challenges, principally owing to the fact that there are more than 2000 such agents used in commerce, the vast majority of which are added at extremely low levels in food. In view of the fact that the majority of individual flavouring agents occur naturally in food, the Committee has concluded that to evaluate each of these substances by traditional toxicological testing is simply not warranted in most cases (WHO, 1987). In 1987, WHO classified flavouring agents into four groups: (a) Artificial substances that are unlikely to occur naturally in food; (b) Natural substances that are not normally consumed as food, their derived products, and the equivalent nature-identical flavourings; (c) Herbs and spices, their derived products, and the equivalent natureidentical flavourings, and (d) Natural flavouring substances obtained from vegetable and animal products and normally consumed as food whether processed or not, and their synthetic equivalents. 435

18 436 THE SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES Since 1996, the Committee has evaluated the safety of approximately 1150 individual flavouring agents using the Procedure for the Safety Evaluation of Flavouring Agents (Annex 1, references 116 and 117 ). The Committe organized these agents into a number of well-defined chemical groups (i.e. congeneric groups) in order to effectively evaluate such a large number of substances. The evaluations conducted to date have demonstrated that flavouring agents within a congeneric group have a similar biochemical fate and toxicological potential. Most individual flavouring agents have been shown to be efficiently detoxicated to yield innocuous metabolites. Knowledge of their metabolic fate, coupled with low intake levels that are typically below thresholds of toxicological concern, has formed the basis for the majority of the Committee s evaluations of individual flavouring agents. In instances where the metabolic fate of an individual flavouring agent was not well known or readily predictable, or where intake exceeded the threshold of toxicological concern for the relevant structural class, the Committee has relied on data on the toxicity of the agent itself or of a structurally related substance in order to perform a rigorous safety evaluation. The results of these evaluations comprise an extensive database, which has been published since 1996 in the WHO Food Additive Series as a series of monographs on congeneric groups. A key development in the process by which the Committee evaluates the safety of flavouring agents has been the establishment of specifications, consistent with JECFA practice, for all flavouring agents. The criteria underpinning these specifications require that the chemical assay for individual flavouring agents be specified. The Committee has adopted the criterion of a minimum assay of 95% purity of the named flavouring agent. In some cases, flavouring agents have been reported to be of less than 95% purity. In such instances, the Committee has required the identification of the secondary components such that at least 95% of the chemical composition of the named flavouring agent can be accounted for. Key features of secondary components are that many are themselves flavouring agents that are likely have been evaluated previously by the Committee, as a member of a different congeneric group, or they are structurally related substances that belong to the same congeneric group as the named flavouring agent. In order to provide for a rigorous safety evaluation when there are many secondary components, the Committee has collected and evaluated data on each of these components. This process has been facilitated by the fact that, since 1996, the Committee has reviewed and published monographs on most of the chemical groups of flavouring agents. Using these monographs on congeneric groups, the Committee now has the capacity to efficiently evaluate additional flavouring agents and numerous secondary components. An obvious extension of this process is that the Committee can begin to evaluate mixtures of flavouring agents, as found in natural flavouring complexes.

19 THE SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES Natural flavouring complexes fall into the following categories: (1) Essential oils: the volatile flavouring constituents of plant sources (leaves, fruits, buds, bark, etc.) usually obtained by steam distillation, by expression or extraction, or some combination of these processes. (2) Extracts: the volatile and non-volatile flavouring constituents of plant sources, as described in (1) above, obtained by extraction with a permitted polar or non-polar solvent, the choice of which depends on the botanical source and the constituents desired. (3) Oleoresins, often called solid extracts: obtained by extraction as in (2) above, followed by removal of the solvent. These categories fall into groups (b), (c), and (d) listed above (WHO, 1987). The agents that comprise these categories range from those that consist almost entirely of a single chemical entity, such as bitter almond oil (benzaldehyde), to those with a highly complex composition (e.g. rosemary oil). In some instances, dozens of constituents are essential to the technical flavour characteristics of the natural flavouring complex. Although the chemical composition of these complexes is variable, their constituents can be assigned to relevant congeneric groups. Table 1 gives examples of natural flavouring complexes and individual flavouring agents and provides an indication of the number of these substances that are used in commerce in the United States. Tables 2 and 3 list the constituents of two natural flavouring complexes, bois de rose oil and lemongrass oil, organized by congeneric group. While each of these complexes has numerous constituents, these constituents fit readily into congeneric groups, most of which have already been evaluated by the Committee. This monograph proposes that the existing Procedure for the Safety Evaluation of Flavouring Agents (Annex 1, references 116 and 117), as refined by the Committee (Annex 1, references 122, 131), be further modified to accommodate the safety evaluation of natural flavouring complexes. After consideration of an approach to determining specifications that could be used to specify a chemical assay for a natural flavouring complex, details of the suggested modifications to the existing Procedure in order to provide for the evaluation of natural flavouring complexes are outlined herein. 2.1 Specifications for natural flavouring complexes The Committee has always insisted upon specifications that define the substance being considered and, as previously noted, has typically required that a chemical assay for individual flavouring agents be at least 95%. In cases in which the named substance could not practically be freed from secondary components, the Committee has evaluated each secondary component individually. The emphasis throughout has been on assuring the identity and safety of the substance being evaluated.

20 438 THE SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES In dealing with natural flavouring complexes, it may be useful to invoke compendia of specifications published by organizations that have previously considered these agents. The Food Chemicals Codex is one such organization, recognized by regulation in the United States and elsewhere. Food Chemicals Codex specifications typically list physiochemical properties, contaminants (including heavy metals and polynuclear aromatics) and microbiological parameters, to ensure that food ingredients, including flavouring agents, are safe for human consumption. For many natural flavouring complexes, Food Chemicals Codex also specifies a minimum chemical assay for key chemical constituents. For example, the chemical assay for caraway oil specifies that it should contain not less than 50%, by volume, of ketones calculated as carvone. The assay for peppermint oil is not less than 5% esters, calculated as menthyl acetate, and not less than 50% menthol. Clearly, flexibility is maintained in a chemical assay specified by Food Chemicals Codex in that a minimum level is specified for a chemical group that may be measured or calculated as a key chemical constituent in the natural flavouring complex. This approach could be used effectively to specify a chemical assay for natural flavouring complexes evaluated by the Committee. To perform an evaluation, the Committee would organize the constituents of the natural flavouring complex into congeneric groups (e.g. phenols, tertiary alcohols, etc.) and perform a series of evaluations for the congeneric group. Within key congeneric groups (e.g. alicyclic secondary alcohols and ketones) selected constituents (e.g. menthol) will be associated with the technical flavouring characteristics of the natural flavouring complex (e.g. peppermint oil). In a manner entirely consistent with Food Chemicals Codex, the evaluation of congeneric groups within the complex could be linked by chemical assay to constituents that are fundamental to the flavouring characteristics of the natural flavouring complex. An essential general principle is that the specification should be no more complex than is essential to assure the most critical aspects of safety, identity, and technical function. Determination of these key characteristics requires the use of methods shown to be reliable through collaborative studies. Existing relevant specifications from Food Chemicals Codex for bois de rose oil and lemongrass oil are provided in Tables 4 and 5 (Food Chemicals Codex, 1996). 2.2 Estimates of daily intake The principles and procedures used by the Committee to estimate intake in the safety evaluation of individual flavouring agents are equally valid and appropriate for natural flavouring complexes, provided that the intended conditions of use are similar to those for individual flavouring agents. One obvious difference between individual flavouring agents and natural flavouring complexes is that some of the latter have much broader patterns of use in the food supply and much higher volumes of disappearance into the marketplace than do individual agents. Natural flavouring complexes such as vanilla extract and lemon oil are used in a wide variety of food categories that have high rates of consumption from products such as baked goods, beverages, soft and hard candy, and dairy products. Annual volumes of production for twelve natural flavouring complexes in use in the United States exceed kg.

21 THE SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES 439 Table 1. Examples of individual flavouring agents and natural flavouring complexes Type of flavouring agents Examples (No.) Chemical assay (number of agents of this type) Individual flavouring agents with minimum Cinnamyl alcohol (No. 647) Minimum assay of >95% (JECFA) assay value of >95% (1400) Individual flavouring agents with minimum (E,R)-3,7-Dimethyl-1,5,7-octatrien-3-ol Minimum assay of 93%; 3 5% linalool, and assay value of <95% (230) (No. 1154) lesser quantities of linalool oxide and nerol oxide Natural flavouring complexes Wintergreen oil (Group (b)) a FCC chemical assay: not less than 98% and not Essential oils (190) more than 100.5% methyl salicylate Lemongrass oil (Group (c)) a FCC chemical assay: not less than 75%, by volume, of aldehyde as citral Lemon oil, distilled (Group (d)) a FCC chemical assay: between 1% and 3.5% aldehydes, calculated as citral Extracts (100) Vanilla extract (Group (c)) a Oleoresins (30) Black pepper oleoresin (Group (c)) a FCC chemical assay: piperine: not less than 36%; volatile oil content: between 15ml and 35 ml/100 g. a Group as classified by WHO (1987) (see Introduction) FCC, Food Chemical Codex

22 440 THE SAFETY EVALUATION OF NATURAL FLAVOURING COMPLEXES Table 2. Constituents of bois de rose oil, organized by congeneric group a Congeneric DT FEMA CAS No. Constituent Composition of the group (date of class No. JECFA review) Data from industry c Ind. 1 Ind. 2 (2001) (2001) Target Target analysis analysis 3 I Geranial 3 I Geraniol I Geranyl acetate 3 I Neral 3 I Nerol (1997, 2003) I a,b-unsaturated aliphatic primary alcohol/aldehyde/ acid/acetal/ester 6 I trans-dehydrolinalool 6 I ,7-Dimethyl-1,5,7- octatrien-3-ol 6 I 2,6-Dimethyl-3,7- octadien-2,6-diol 6 I Linalool I trans-nerolidol I Ocimenol 6 I Spathulenol 6 I Terpinen-4-ol 6 I a-terpineol I g-terpineol (1998) I Aliphatic, alicyclic and aromatic saturated and unsaturated tertiary alcohol and related ester 16 II ,8-Cineole II cis-linalool oxide 16 II trans-linalool oxide 16 II Linalool oxide (2 peaks) II ,2,6-Trimethyl-6- vinyltetrahydropyran 16 (2003) II Aliphatic and alicyclic ethers 31 I Cadinene 31 I delta-cadinene 31 I g-cadinene 31 I Camphene 31 I b-caryophyllene 31 I a-copaene I Cyclosativene 31 I para-cymene I b-elemene I Eremophilene 31 I Limonene I Myrcene