Furan is not genotoxic in the micronucleus assay in vivo or in vitro

Toxicology Letters 169 (2007) 43–50

Furan is not genotoxic in the micronucleus assay in vivo or in vitro Louise J.K. Durling, Kettil Svensson, Lilianne Abramsson-Zetterberg ∗ Livsmedelsverket, National Food Administration, Toxicology Division, Box 622, SE-751 26 Uppsala, Sweden Received 13 July 2006; received in revised form 23 August 2006; accepted 24 August 2006 Available online 20 December 2006

Abstract Furan, a potential human carcinogen, is formed during heat-treatment of food. Previous studies of the genotoxicity of furan have given disparate results. Hence, there is a need for complementary data to clarify the mechanism behind the carcinogenicity of furan. In this study, we have used the flow cytometer-based micronucleus assay in mice and the cytokinesis-block micronucleus assay in human lymphocytes to investigate the genotoxic potential of furan. Three in vivo experiments were performed: intraperitoneal or subcutaneous injection of furan in male Balb/C mice (0–300 and 0–275 mg/kg body weight, respectively) and intraperitoneal injection of male CBA mice (0 and 225 mg/kg body weight). No increased level of micronucleated erythrocytes was detected in any of the in vivo experiments. In the in vitro setup, human lymphocytes from two donors were treated with furan in concentrations from 0 to 100 mM, either with or without metabolic activation (liver homogenate from rat). In parity with the in vivo results there was no significant increase in the frequency of micronucleated cells here either. As neither the in vivo nor the in vitro studies disclose any significant increase in the micronucleus frequency after treatment with furan, our results support that the carcinogenicity of furan is caused by a non-genotoxic mechanism. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Furan; Micronucleus assay; Flow cytometer; Chromosome aberration

1. Introduction Heat-induced food toxicants have during the last years reemerged as a topic of interest. This is especially true after it was reported that acrylamide is formed during heat treatment of carbohydrate-rich foods (Tareke et al., 2002; Svensson et al., 2003). However, a great variety of other, possibly harmful, compounds are formed during the heating process. One such compound is furan that has been classified as a possible human carcinogen (group

∗ Corresponding author. Tel.: +46 18 17 57 63; fax: +46 18 17 14 33. E-mail address: [email protected] (L. Abramsson-Zetterberg).

2B) by the International Agency for Research on Cancer (IARC, 1995). It has since long been known that the heterocyclic compound furan is present in some kinds of food as a flavour volatile (Maga, 1979). In 2004, the US Food and Drug Administration (FDA) performed a survey and found that furan was present in numerous kinds of heat-treated foods (FDA, 2005). The highest levels were found in canned and jarred products like baby food, soups, sauces, vegetables, and fruits, but furan is also found in other products, for example in coffee (FDA, 2005). The mechanism behind furan formation in food is still not fully understood. One proposed pathway is thermal degradation or rearrangement of organic compounds, particularly carbohydrates (Maga, 1979), but

0378-4274/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.toxlet.2006.08.020

44

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

there are probably multiple routes of formation. Besides the occurrence of furan in food it also has a wide use as an intermediate and solvent in industrial processes. Despite this extensive use, information on the toxicology of furan is still incomplete. After the discovery that furan is present in a broad range of food items, it is even more important to clarify the toxicological properties of it. The mechanism behind the carcinogenic effect of furan is not well understood. Various studies suggest both genotoxic (Stich et al., 1981; McGregor et al., 1988; NTP, 1993) and non-genotoxic mechanisms (Mortelmans et al., 1986; Wilson et al., 1992; Fransson-Steen et al., 1997). Furan is metabolised into the key metabolite, cis-2-butene-1,4-dial, by CYP2E1. This metabolite is probably responsible for the carcinogenic properties of furan (Byrns et al., 2006). It is both cytotoxic and binds to proteins (Burka et al., 1991) and nucleosides (Byrns et al., 2002). Genotoxic activity has an impact on the risk assessment of a chemical. Therefore, it is important to further investigate these properties of furan, as the results from bacterial culture and whole animal genotoxicity studies are disparate. European Food Safety Authority (EFSA) (2004) and FDA (2004) have asked for complementary studies to clarify this matter, to be able to perform a thorough risk assessment of furan. Furthermore, to determine if furan in food constitutes a health risk, additional toxicological intake data are needed. A commonly used method to assess the potential of a chemical to induce chromosomal aberrations in vivo is the micronucleus assay in mice (e.g. semicarbazide (Abramsson-Zetterberg and Svensson, 2005) and acrylamide (Abramsson-Zetterberg, 2003)). This method allows the detection of both clastogenic (chromatide/chromosome breakage) and aneugenic (chromosome maldistribution) effects. To increase the sensitivity of this method the cells can be scored with a flow cytometer, which gives the opportunity to score a large number of cells. It has been reported that furan is rapidly transported from the peripheral blood system to the liver (Burka et al., 1991). Thus, it is especially important to use sensitive methods when analysing effects of furan, as the bone marrow dose might be low. Another common method is the in vitro cytokinesisblock micronucleus assay in human lymphocytes. This method is a way to ascertain that cells are exposed to the test substance. In the present study, we investigated whether furan increases the micronucleus frequency, both in vivo and in vitro. This is an important step on the way to discriminate between genotoxic and non-genotoxic mechanisms behind the cancerogenicity of furan.

2. Materials and methods 2.1. Micronucleus assay in vivo in mice (experiment 1) 2.1.1. Animals Male Balb/C and male CBA mice, 6–7 weeks old, weighing approximately 25 g, were obtained from Scanbur AB, Sollentuna, Sweden. The animals were housed at the National Food Administration in Sweden in a 12 h light/12 h dark cycle with free access to solid food and tap water. All mice were acclimatised 1 week before treatment. The experiment was reviewed and approved by Uppsala Ethical Committee on Animal Experiments, application C228/3. 2.1.2. Chemicals Furan (CAS no. 110-00-9), Colchicine (CAS no. 64-868), and Hoechst 33342 (HO342) was purchased from Sigma– Aldrich, Sweden; Percoll from Amersham Biosciences, Sweden; Fluothane from Astra, Sweden; Thiazole Orange (TO) from Molecular Probes, OR, USA. 2.1.3. Experimental design and sampling The genotoxic effect of furan was determined using the flow cytometer-based micronucleus assay in mice. The animals were randomly divided into different dose-groups. Three separate experiments were performed (1a–1c). In order to disclose any methodological differences between species two strains of mice, Balb/C and CBA, were used. In the first (1a) and second experiment (1b) Balb/C mice were injected with furan intraperitoneally (i.p.) and subcutaneously (s.c.), respectively. In the third experiment (1c) CBA mice were injected i.p. In all experiments the mice were injected with a single dose of 10 ␮l/g b.w. Furan was diluted in corn oil just prior to the injection. The positive control mice received injections of 1 mg/kg b.w. colchicine dissolved in PBS. Forty-two hours after injection, the animals were anaesthetised with Fluothane and blood samples were drawn from the orbital vein into heparinised tubes. Directly after blood sampling the animals were killed by cervical dislocation. The sampling time was based on the results of Cao et al. (1993) and on the knowledge of time between appearance of polychromatic erythrocytes (PCE) in the bone marrow and in peripheral blood (Abramsson-Zetterberg et al., 1996). In experiment 1a, i.p. administration of furan to male Balb/C mice, a total of 27 mice were used. They were given the following doses of furan: 0, 50, 75, 90, 110, 125, 150, 175, 200, 250, and 300 mg/kg b.w. All groups consisted of two mice except the ones given a dose of 0 and 300 mg/kg b.w., and the positive control (colchicine), which consisted of three mice each. Experiment 1b, s.c. administration of furan to male Balb/C mice, involved 16 mice, which were given three different doses of furan: 0, 150, and 275 mg/kg b.w. Five mice were given 150 mg/kg b.w. and six mice 275 mg/kg b.w. Three mice were given 0 mg/kg b.w. furan and two mice the positive control colchicine.

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

In experiment 1c, i.p. injection of male CBA mice, eight mice were involved. Four mice were given one dose of furan: 225 mg/kg b.w. and four mice served as control. 2.1.4. Purification, fixation and staining of erythrocytes The methods for purification, fixation and staining of erythrocytes have been presented previously (Graw´e et al., 1992, 1993b; Abramsson-Zetterberg et al., 1995). In brief, 5 ␮l of whole blood was layered on a 65% Percoll gradient and centrifuged for 20 min at 600 × g. Platelets, and the majority of nucleated cells present in the supernatant were carefully removed. The pellet was resuspended in PBS and fixed in glutaraldehyde according to a method described by Hayashi et al. (1992). The samples were coded and stored at 4 ◦ C for a few days. A staining solution, containing the fluorescent dyes HO342 (DNA-dye) and TO (RNA-dye) in PBS, was added to the fixed cells followed by incubation for 35 min at 37 ◦ C and continuing over night at 4 ◦ C. The staining was performed 1 day prior to analysis. 2.1.5. Flow cytometric analysis The analysis was done using a FACSVantage SE flow cytometer (Becton-Dickinson Immunocytometry Systems, Sunnyvale, CA) according to a method described elsewhere (Graw´e et al., 1992; Abramsson-Zetterberg et al., 1995). The cells were automatically analysed when they, one by one, passed through two laser beams (350 and 488 nm). Information about size and structure was collected and was used to exclude remaining nucleated cells, as well as debris in the sample from further analysis. DNA, and RNA content were detected as fluorescence from the two dyes HO342 and TO, respectively. Threshold values were set for all parameters using the control sample. RNA content, measured as the signal from the TO dye, is an indication of erythrocyte age. This information was used to distinguish between young and old erythrocytes, PCE and NCE (normochromatic erythrocytes), respectively. The ratio of young to old cells gives an indication of cell proliferation (%PCE). This analysis was based on the information from about 20,000 cells/sample. Furthermore, since erythrocytes normally do not contain any DNA, a signal from the HO342 dye implies that it contains a micronucleus. Limiting the analysis to only young cells enabled us to calculate the frequency of micronucleated PCE formed during the experiment. About 140,000 PCE were scored per animal. All analyses were performed using CellQuest software (Becton-Dickinson). 2.1.6. Determination of micronucleus frequency Scatter plots of the information about DNA content versus RNA content were displayed for each analysed sample. Regions were defined for NCE, PCE, and MPCE (micronucleated PCE), respectively. On the basis of this, the number of events in each region was determined. Frequencies of PCE and MPCE were calculated. Furthermore, to obtain information about the mechanism behind the formation of micronuclei, i.e. aneugenic or clastogenic effect, we divided the region defined

45

as MPCE in two parts, MPCE and MPCE(II). MPCE(II) was defined as the part of MPCE with a high DNA content, since a high DNA content is an indication of aneugenic effects (Graw´e et al., 1994). 2.2. Micronucleus assay in vitro in human lymphocytes (experiment 2) 2.2.1. Chemicals Furan (CAS no. 110-00-9) was obtained from Sigma; dimethylsulfoxide (DMSO) from Riedel-de-Ha¨en AG, Germany; F-10 HAMs medium and fetal calf serum from HyClone USA; penicillin/streptomycin (PEST) and l-glutamine from Bio Whittaker Europe; phytohemaglutinine (PHA), cytochalasin B, and benzo(a)pyrene from Sigma–Aldrich, Sweden. 2.2.2. Experimental design and treatment Two experiments (2a and 2b) were performed with blood from two healthy non-smoking females, 27 and 53 years old, respectively. Whole blood cultures were treated with furan dissolved in DMSO. Concentrations of 2 mM, 5 mM, 7.5 mM, 10 mM, 15 mM, 20 mM or 100 mM and 5 mM, 10 mM, 17.5 mM or 25 mM were used for the two experiments, respectively. Treatments were performed in duplicates each including both with and without metabolic activation (S9-mix from Arochlor 1254 induced rats according to the protocol of Ames et al. (1975)). Benzo(a)pyrene, final concentration 70 nM, was used as a positive control in cultures with metabolic activation and colchicine 75 nM in cultures without metabolic activation. The experiments were carried out according to the method described by Fenech and Morley (1985) and refined by Migliore et al. (1995, with changes of the cytochalasin B concentration to (6 ␮g/ml)). In brief, whole blood was cultured in F-10 HAMs medium supplemented with 12% fetal calf serum, 0.5% l-glutamine, 2% PEST, and 1.2% PHA for 72 h. The cultures were kept in a humidified environment with 5% CO2 at 37 ◦ C. After 24 h of incubation, the different treatments, furan, DMSO and the positive controls were added, as well as S9-mix according to the experimental setup. Cytochalasin B was added to the cultures after 44 h, to a final concentration of 6 ␮g/ml. It prevents the cells from completing cytokinesis and causes formation of multinucleated cells. The incubation proceeded until 72 h after initiation. The cultures were then centrifuged and the supernatant was removed. Erythrocytes were lysed by adding KCl (0.075 M), 37 ◦ C and, after 3 min, prefixative (methanol/acetic acid 3:1 v/v) was added. The cells were rinsed and fixed in methanol and methanol/acetic acid (5:1 v/v) and spread on clean, cold, and wet glasses. The glasses were air dried before staining with 3% Giemsa’s stain. 2.2.3. Scoring All slides were randomised and coded prior to analysis. The scoring was performed in accordance with Fenech (2000) using a Leica Ortholux II microscope. About 2000 binucleated lymphocytes (cells that have undergone one mitotic division

46

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

during the exposure) were scored from each culture for the presence of micronuclei. As a measure of cell division the proportion of binucleated cells (%BN) among total viable cells was calculated. About 600 cells/culture were scored for this purpose. 2.3. Statistics A linear regression model was fitted to the data to test for a dose–response relationship in both the in vivo and the in vitro tests. In experiments 1a and 1b, the frequency of MPCE (fMPCE) was used as response variable, while the dose of furan was used as predictor variable in the model. In experiment 1c, where a single dose was compared to the control, a t-test was used. Similarly to experiments 1a and 1b, the frequency of micronucleated lymphocytes was used as response variable in experiments 2a and 2b, while the furan concentration was used as predictor variable. The effect of the positive controls was tested using Fisher’s exact test. All statistical analyses were carried out using the R statistics package (R Development Core Team, 2006).

3. Results 3.1. In vivo micronucleus test (experiments 1a–1c) The results from the flow cytometric analyses after i.p. and s.c. injection of furan is presented in Table 1 and Fig. 1. From each furan-treated mouse about 140,000

Fig. 1. The frequency of micronucleated polychromatic erythrocytes (fMPCE) in peripheral blood of mice given single doses of furan intraperitoneally in Balb/C mice (+), subcutaneously in Balb/C mice () or intraperitoneally in CBA mice () (experiments 1a–1c). Blood was collected 42 h after treatment. The frequency of micronucleated cells was determined using flow cytometeric analyses. Each data point represents the mean value from the animals in each dose group. Standard deviation for fMPCE is given for groups with at least three animals as vertical bars (experiments 1b and 1c). Colchicine was used as positive control (fMPCE = 12.4 and 9.3 per mille for experiments 1a and 1b, respectively).

PCEs were analysed. No significant depression of cell proliferation, %PCE, was seen in any of the experiments 1a–1c. Despite this, it would not be informative to further increase the dose due to the general toxicity of furan. One mouse given 300 mg/kg b.w. of furan i.p. and two given 275 mg/kg b.w. of furan s.c. died after treatment. Signs of decreased physical activity were observed in mice treated with the highest doses of furan, both i.p. and s.c. The positive control colchicine yielded a depression of cell proliferation and also a positive response in micronucleus induction. No significant dose–response relationship was found neither after i.p. nor s.c. administration of furan (R2 = 0.03, slope = −0.0006, n.s. and R2 = 0.005, slope = 0.0002, n.s.; t-test, n.s. for experiments 1a–1c, respectively) (Table 1 and Fig. 1). The flow cytometer-based method also allows for restriction of the analysis to micronuclei with a high DNA content. This gives a possibility to further distinguish between aneugenic and clastogenic properties of a genotoxic compound. However, not even with this approach we could detect any significant increase in the frequency of micronucleated erythrocytes compared with the control mice. 3.2. In vitro micronucleus test (experiments 2a–2b) We also set up two experiments (2a and 2b) using the micronucleus assay in vitro with two different blood donors. Micronucleus frequencies in lymphocytes treated with furan (0–100 and 0–25 mM) are shown in Table 2 and Fig. 2. No significant increase in micronucleus frequency was observed, neither with nor without metabolic activation (experiment 2a: R2 = 0.002, slope = −0.004, n.s. and R2 = 0.04, slope = −0.015, n.s., respectively. Experiment 2b: R2 = −0.10, slope = 0.03, n.s. and R2 = −0.17, slope = 0.003, n.s., respectively). Duplicate cultures were performed for each concentration tested and about 2000 binucleated cells were scored from each culture. Due to cytotoxicity fewer cells were scored at higher concentrations of furan and no cells could be scored at 25 and 100 mM concentration. The positive controls yielded a significant increase in micronucleus frequency (p < 0.05 in all experiments). 4. Discussion In the present study, we report results from the testing of genotoxicity of furan, using the micronucleus assay in vivo and in vitro. Our results did not disclose any increase in micronucleus frequency in any of the assays. The flow cytometry-based micronucleus assay in vivo enables the analysis of a large number

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

47

Table 1 Results of the flow cytometric analyses of peripheral blood from male Balb/C or CBA mice given furan intraperitoneally or subcutaneously Animal code

Dose (mg/kg b.w.)

PCE (%)

Number of PCE

Number of MPCE

fMPCE (per mille)

fMPCE(II) (per mille)

Experiment 1a 1–3 4–5 6–7 8–9 10–11 12–13 14–15 16–17 18–19 20–21 22–23 24–26

0 50 75 90 110 125 150 175 200 250 300 Colchicine

6.0 2.1 2.8 2.8 2.4 2.6 2.6 1.9 2.4 2.7 4.3 1.0

360,307 228,270 241,131 207,010 226,370 243,249 265,252 174,449 265,190 251,912 381,665 36,726

809 413 459 337 403 458 419 265 546 464 699 462

2.2 1.8 1.9 1.6 1.8 1.9 1.6 1.5 2.1 1.8 1.8 12.4

0.9 0.8 0.8 0.7 0.7 0.8 0.6 0.7 1.0 0.9 0.7 9.7

Experiment 1b 1–3 4–8 9–12 13–14

0 150 275 Colchicine

4.1 ± 0.6 2.6 ± 0.8 3.1 ± 1.3 1.2

586,719 965,649 767,673 74,042

1290 2067 1720 698

2.2 ± 0.1 2.1 ± 0.3 2.2 ± 0.2 9.3

1.1 ± 0.5 1.1 ± 0.2 1.1 ± 0.0 9.1

Experiment 1c 1–4 5–8

0 225

2.4 ± 0.2 1.8 ± 0.1

838,278 854,025

754 747

0.9 ± 0.5 0.9 ± 0.7

0.42 ± 0.1 0.39 ± 0.1

Blood samples were collected 42 h after administration of furan. Colchicine (1 mg/kg b.w.) was used as positive control. PCE is polychromatic erythrocytes; MPCE is micronucleated polychromatic erythrocytes; fMPCE is frequency of micronucleated PCE averaged over mice in each dose group; fMPCE(II) is frequency of micronucleated PCE restricted to PCE with a high DNA content, averaged over mice in each dose group. Experiment 1a: Balb/C mice i.p. injected with furan, experiment 1b: Balb/C mice s.c. injected with furan, and experiment 1c: CBA mice i.p. injected with furan. Standard deviation for %PCE, fMPCE, and fMPCE(II) is given for groups with at least three animals.

Fig. 2. Micronucleus frequency in human binucleated lymphocytes (fMNBN) exposed to furan 0–100 mM with (+) or without () metabolic activation (only data from experiment 2a). Treatments were performed in duplicates. For each culture about 2000 binucleated cells were scored for the presence of micronuclei. At the concentrations 15 and 20 mM, fewer cells were scored due to cytotoxicity. The highest concentration tested, 100 mM, resulted in total cytotoxicity. Each data point represents one culture. Benzo(a)pyrene was used as positive control (fMNBN = 4.5 per mille).

of erythrocytes and thus the sensitivity of the test is increased compared to manual scoring. This method also enables the calculation of relative DNA content of the micronuclei and thereby allows a refined evaluation of the aneugenic (maldistribution of chromosomes) and clastogenic (chromatide/chromosome breakage) properties of a compound (Graw´e et al., 1993a). The results from our study show that furan is neither clastogenic nor aneugenic. There was no significant increase in micronucleus frequency after i.p. or s.c. administration of furan in either Balb/C or CBA mice. In the micronucleus assay in vivo it is common to give the test substance i.p. However, since furan is accumulated in the liver (Burka et al., 1991), it can be argued that the target organ, bone marrow, might be exposed to a lower dose than anticipated. To address this issue and increase the dose reaching the bone marrow we performed a second experiment where furan was injected s.c. (experiment 1b). An agent given this way is usually absorbed at a lower rate and enters directly into the general circulation, instead of through the portal circulation (Lukas et al., 1971). Hence, this way of administration may give an elongated exposure of the bone marrow. Furthermore, to elucidate whether different strains of mice

48

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

Table 2 Results from the micronucleus test in human lymphocyte cultures (experiments 2a–2b) treated with furan 0–100 or 0–25 mM with and without metabolic activation Furan (mM)

Experiment 2a 0 0 2 2 5 5 7.5 7.5 10 10 15 15 20 20 100 100 B(a)P B(a)P Experiment 2b 0 0 5 5 10 10 17.5 17.5 25 25 B(a)P B(a)P Colchicine Colchicine

With metabolic activation

Without metabolic activation

fBN (%)

BN

MNBN

fMNBN (per mille)

fBN (%)

BN

MNBN

fMNBN (per mille)

30 29 26 22 21 21 20 26 20 23 18 19 13 14 – – 17 17

2039 2025 2015 2017 1980 1961 2011 2014 2012 2012 2014 1699 758 1103 – – 1263 861

1 2 2 2 2 3 3 1 2 1 4 1 1 0 – – 7 3

0.5 1.0 1.0 1.0 1.0 1.5 1.5 0.5 1.0 0.5 2.0 0.6 1.3 0.0 – – 6.0 3.0

31 28 35 30 25 29 26 26 25 24 20 22 13 15 – – 29 28

2013 2001 2000 2000 2020 2000 2006 2035 1962 2033 1369 1551 682 589 – – 2019 1364

3 3 0 2 2 3 1 1 2 1 1 2 1 0 – – 1 1

1.5 1.5 0.0 1.0 1.0 1.5 0.5 0.5 1.0 0.5 0.7 1.3 1.5 0.0 – – 0.0 1.0

17.5 13.8 14.3 16.5 9.4 12.7 9.6 9.9 – – 17.3 16.0

2307 1962 1952 2077 1834 1561 1041 774 – – 2027 1242

2 4 4 6 6 1 2 2 – – 8 7

0.9 2.0 2.0 2.9 3.3 0.6 1.9 2.6 – – 3.9 5.6

19.9 18.6 23.6 18.8 20.1 16.0 10.6 9.3 – –

1978 1416 1933 2035 2057 2048 611 617 – –

0 2 4 1 3 3 1 0 – –

0.0 1.4 2.1 0.5 1.5 1.5 1.6 0.0 – –

11.4 10.2

1203 1302

7 4

5.8 3.1

Human lymphocytes, blood from two healthy, non-smoking females, 27 and 53 years old (experiments 2a and 2b), were treated with different doses of furan with and without metabolic activation. All treatments were performed in duplicates. After 24 h, cells were treated with cytochalasin B to stop cell division and form binucleate cells. Benzo(a)pyrene (B(a)P) and colchicine were used as positive controls. fBN is the frequency of binucleate cells in the culture, BN number of binucleate cells scored for presence of micronuclei, MNBN micronucleated binucleate cells, fMNBN is frequency of micronucleated binucleate cells. Cultures which could not be scored due to cytotoxicity are indicated with ‘–’.

react differently due to different metabolic activity we performed a third in vivo experiment with another strain of mice (experiment 1c). Despite these additional experiments, no significant increase in the frequency of MPCE could be detected. Furan is a volatile substance with a boiling point of 32 ◦ C and a large proportion of furan is rapidly lost through exhalation (Burka et al., 1991). Still, the highest applicable dose in our experiments was limited by acute toxicity, both death and decreased physical activity was observed among mice treated with the highest doses of furan.

We also performed the micronucleus assay in vitro in human lymphocytes, as this method assures that the target cells are exposed to the test substance and also represents the effect in human cells. The results from this assay concur with the results of the in vivo tests, no significant increase in micronucleus frequency was observed either with or without metabolic activation. It should be acknowledged that PCB-induced (Arocholor 1254) rat liver may cause a low CYP2E1 activity. This enzyme is critical for the formation of the genotoxic metabolite cis-2-butene-1,4-dial (Kedderis et al., 1993). It should

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

also be noticed that the high volatility and low boiling point of furan, mentioned above, could lead to a more transient exposure and an overall lower mean concentration than expected in the cultures (Wilson et al., 1992). However, since cytotoxicity was reached at the highest concentration tested, this should not be a major problem in our study. The mechanism of furan-induced carcinogenesis has so far not been successfully determined. There are several studies proposing that furan acts through a nongenotoxic mechanism, e.g. Fransson-Steen et al. (1997) and Wilson et al. (1992) have demonstrated an induced cell proliferation after exposure to furan. Moreover, furan is not mutagenic in Ames test (Mortelmans et al., 1986) and does not induce DNA repair in liver (Wilson et al., 1992) or SCE in mice (NTP, 1993). On the other hand, it is clastogenic in CHO cells with S9-mix (Stich et al., 1981), causes chromosomal aberrations and SCE in CHO cells (NTP, 1993), and induces SCE in modified V79 cells (Glatt et al., 2005). It is also shown to be positive in the mouse lymphoma assay without S9-mix (McGregor et al., 1988) and induces structural chromosomal aberrations in mice after i.p. administration (NTP, 1993). Furthermore, some studies suggest that the metabolite of furan, cis-2-butene-1,4-dial, is responsible for the carcinogenic effect. For example, the metabolite is shown to be mutagenic in Ames test (Peterson et al., 2000), it induces DNA single-strand breaks and DNA cross-links in CHO cells (Marinari et al., 1984) and also forms DNA adducts (Byrns et al., 2006) and binds to proteins (Burka et al., 1991). However, it is unclear to what extent furan is metabolised both in vivo and in vitro. If the metabolism is not extensive enough it implies that the metabolite does not reach levels where genotoxic effects are detectable even with the most sensitive methods available today. Our study corroborates the above-mentioned studies that furan is not genotoxic. This indicates that the carcinogenic effects shown in mice (NTP, 1993) are not caused by chromosomal aberrations. The results presented in this study indicate that if the metabolite cis-2-butene-1,4-dial is responsible for the toxic activity, the metabolism of furan is not high enough to induce a detectable number of micronuclei, neither in mice nor in our in vitro system. Another plausible explanation for the outcome of the in vivo study is that the distribution to the target cells, the erythroblasts, is too low to cause any detectable increase in micronucleus frequency. It is assumed that the metabolite of furan, cis-2butene-1,4-dial, is responsible for the genotoxic activity. Concerning the non-genotoxic mechanism behind cancerogenicity it has been proposed that the metabolite

49

stimulates cell proliferation, which in turn increases the likelihood of tumour formation (Fransson-Steen et al., 1997). To clarify whether cancerogenicity in the liver is caused by a genotoxic effect or solely by an increased cell proliferation, further studies are advisable. A desirable method to study this would be a micronucleus assay restricted to liver cells in mice. Such studies, however, implies numerous methodological difficulties. In conclusion, we have used state of the art methods for detection of clastogenic and aneugenic effects. This has revealed that furan is not clastogenic or aneugenic, neither in vivo nor in vitro. Our results together with some previously published results do not support that the carcinogenic effects of furan are caused by a genotoxic mechanism. The information presented here is a contribution to the risk assessment of furan formed in heat-treated foods. Acknowledgements We thank Esmeralda Hadzic, Ingalill Gadhasson, and Elvy Netzel for technical assistance. This work has been carried out with support from the European Commission, Priority 5 on Food Quality and Safety (contract no. FOOD-CT-2003-506820, Specific Targeted Research Project), ‘Heat-generated food toxicants—identification, characterisation and risk minimisation’. References Abramsson-Zetterberg, L., 2003. The dose–response relationship at very low doses of acrylamide is linear in the flow cytometer-based mouse micronucleus assay. Mutat. Res. 535, 215–222. Abramsson-Zetterberg, L., Svensson, K., 2005. Semicarbazide is not genotoxic in the flow cytometry-based micronucleus assay in vivo. Toxicol. Lett. 155, 211–217. Abramsson-Zetterberg, L., Graw´e, J., Zetterberg, G., 1995. Flow cytometric analysis of micronucleus induction in mice by internal exposure to 137Cs at very low dose rates. Int. J. Radiat. Biol. 67, 29–36. Abramsson-Zetterberg, L., Zetterberg, G., Graw´e, J., 1996. The time–course of micronucleated polychromatic erythrocytes in mouse bone marrow and peripheral blood. Mutat. Res. 350, 349–358. Ames, B.N., McCann, J., Yamasaki, E., 1975. Methods for detecting carcinogens and mutagens with the salmonella/mammalianmicrosome mutagenicity test. Mutat. Res. 31, 347–364. Burka, L.T., Washburn, K.D., Irwin, R.D., 1991. Disposition of [14C]furan in the male F344 rat. J. Toxicol. Environ. Health 34, 245–257. Byrns, M.C., Predecki, D.P., Peterson, L.A., 2002. Characterization of nucleoside adducts of cis-2-butene-1,4-dial, a reactive metabolite of furan. Chem. Res. Toxicol. 15, 373–379. Byrns, M.C., Vu, C.C., Neidigh, J.W., Abad, J.L., Jones, R.A., Peterson, L.A., 2006. Detection of DNA adducts derived from the

50

L.J.K. Durling et al. / Toxicology Letters 169 (2007) 43–50

reactive metabolite of furan, cis-2-butene-1,4-dial. Chem. Res. Toxicol. 19, 414–420. Cao, J., Beisker, W., Nusse, M., Adler, I.D., 1993. Flow cytometric detection of micronuclei induced by chemicals in poly- and normochromatic erythrocytes of mouse peripheral blood. Mutagenesis 8, 533–541. EFSA (European Food Safety Authority), 2004. Report of the CONTAM Panel on provisional findings on furan in food. http:// www.efsa.europa.eu/science/contam/contam documents/760/ contam furan report7-11-051.pdf.2006-06-20. FDA (Food and Drug Administration), 2004. Furan in Food, Thermal Treatment; Request for Data and Information. http://www. cfsan.fda.gov/∼lrd/fr040510.html.2006-06-20. FDA (Food and Drug Administration), 2005. Exploratory Data on Furan in Food. http://www.cfsan.fda.gov/∼dms/furandat.html. 2006-06-20. Fenech, M., 2000. The in vitro micronucleus technique. Mutat. Res. 455, 81–95. Fenech, M., Morley, A.A., 1985. Measurement of micronuclei in lymphocytes. Mutat. Res. 147, 29–36. Fransson-Steen, R., Goldsworthy, T.L., Kedderis, G.L., Maronpot, R.R., 1997. Furan-induced liver cell proliferation and apoptosis in female B6C3F1 mice. Toxicology 118, 195– 204. Glatt, H., Schneider, H., Liu, Y., 2005. V79-hCYP2E1-hSULT1A1, a cell line for the sensitive detection of genotoxic effects induced by carbohydrate pyrolysis products and other food-borne chemicals. Mutat. Res. 580, 41–52. Graw´e, J., Zetterberg, G., Amneus, H., 1992. Flow-cytometric enumeration of micronucleated polychromatic erythrocytes in mouse peripheral blood. Cytometry 13, 750–758. Graw´e, J., Zetterberg, G., Amneus, H., 1993a. DNA content determination of micronucleated polychromatic erythrocytes induced by clastogens and spindle poisons in mouse bone marrow and peripheral blood. Mutagenesis 8, 249–255. Graw´e, J., Zetterberg, G., Amneus, H., 1993b. Effects of extended low-dose-rate exposure to 137Cs detected by flow-cytometric enumeration of micronucleated erythrocytes in mouse peripheral blood. Int. J. Radiat. Biol. 63, 339–347. Graw´e, J., Abramsson-Zetterberg, L., Eriksson, L., Zetterberg, G., 1994. The relationship between DNA content and centromere content in micronucleated mouse bone marrow erythrocytes analysed by flow cytometry and fluorescent in situ hybridization. Mutagenesis 9, 31–38. Hayashi, M., Norppa, H., Sofuni, T., Ishidate Jr., M., 1992. Flow cytometric micronucleus test with mouse peripheral erythrocytes. Mutagenesis 7, 257–264.

IARC (International Agency for Research on Cancer), 1995. Monographs, vol. 63, pp. 393–407. Kedderis, G.L., Carfagna, M.A., Held, S.D., Batra, R., Murphy, J.E., Gargas, M.L., 1993. Kinetic analysis of furan biotransformation by F-344 rats in vivo and in vitro. Toxicol. Appl. Pharmacol. 123, 274–282. Lukas, G., Brindle, S.D., Greengard, P., 1971. The route of absorption of intraperitoneally administered compounds. J. Pharmacol. Exp. Ther. 178, 562–564. Maga, J.A., 1979. Furans in foods. CRC Crit. Rev. Food Sci. Nutr. 11, 355–400. Marinari, U.M., Ferro, M., Sciaba, L., Finollo, R., Bassi, A.M., Brambilla, G., 1984. DNA-damaging activity of biotic and xenobiotic aldehydes in Chinese hamster ovary cells. Cell Biochem. Funct. 2, 243–248. McGregor, D.B., Brown, A., Cattanach, P., Edwards, I., McBride, D., Riach, C., Caspary, W.J., 1988. Responses of the L5178Y tk+/tk− mouse lymphoma cell forward mutation assay. III. 72 coded chemicals. Environ. Mol. Mutagen. 12, 85–154. Migliore, L., Scarpato, R., Falco, P., 1995. The use of fluorescence in situ hybridization with a beta-satellite DNA probe for the detection of acrocentric chromosomes in vanadium-induced micronuclei. Cytogenet. Cell Genet. 69, 215–219. Mortelmans, K., Haworth, S., Lawlor, T., Speck, W., Tainer, B., Zeiger, E., 1986. Salmonella mutagenicity tests. II. Results from the testing of 270 chemicals. Environ. Mutagen. 8 (Suppl. 7), 1–119. NTP (National Toxicology Program), 1993. Toxicology and carcinogenesis studies of furan (CAS no. 110-00-9) in F344/N rats and B6C3F1 mice (gavage studies). Technical report series. Peterson, L.A., Naruko, K.C., Predecki, D.P., 2000. A reactive metabolite of furan, cis-2-butene-1,4-dial, is mutagenic in the Ames assay. Chem. Res. Toxicol. 13, 531–534. R Development Core Team, 2006. R: A Language and Environment for Statistical Computing. Vienna, Austria. http://www.Rproject.org.2006-07-05. Stich, H.F., Rosin, M.P., Wu, C.H., Powrie, W.D., 1981. Clastogenicity of furans found in food. Cancer Lett. 13, 89–95. Svensson, K., Abramsson, L., Becker, W., Glynn, A., Hellen¨as, K.E., Lind, Y., Ros´en, J., 2003. Dietary intake of acrylamide in Sweden. Food Chem. Toxicol. 41, 1581–1586. Tareke, E., Rydberg, P., Karlsson, P., Eriksson, S., T¨ornqvist, M., 2002. Analysis of acrylamide, a carcinogen formed in heated foodstuffs. J. Agric. Food Chem. 50, 4998–5006. Wilson, D.M., Goldsworthy, T.L., Popp, J.A., Butterworth, B.E., 1992. Evaluation of genotoxicity, pathological lesions, and cell proliferation in livers of rats and mice treated with furan. Environ. Mol. Mutagen. 19, 209–222.