Mutation Research, 261 (1991) 139-147 © 1991 Elsevier Science Publishers B.V. All rights reserved 0165-1218/91/$03.50 ADONIS 0165121891001418

139

MUTGEN 01704

Biological monitoring of bidi rollers with respect to genotoxic hazards of occupational tobacco exposure Rajani A. Bhisey and Rukmini B. Govekar Carcinogenesis Division, Cancer Research Institute, Tata Memorial Centre, Bombay 400012 (India) (Received 2 January 1991) (Revision received 10 April 1991) (Accepted 22 April 1991)

Keywords: Occupational tobacco exposure; Bidi rollers; Urine mutagenicity; Urinary thioethers; Urinary cotinine

Summary Smokeless tobacco habits are associated with a high incidence of oropharyngeal cancer in India. Hence, the biological effects of occupational exposure to smokeless tobacco used for making bidis (the Indian version of cigarettes) were studied in 2 groups of bidi rollers designated BR-K and BR-S and in control subjects with no tobacco habits. Specific tobacco exposure and the electrophilic burden were determined by estimating urinary cotinine and thioethers respectively. Urine mutagenicity was tested with the Ames assay using Salmonella typhimurium strains TA98 and TA100. While cotinine was not detected in control samples, the mean cotinine levels (mmole/mole creatinine) in the BR-K and BR-S groups were 0.79 + 0.30 and 0.09 _+ 0.03 respectively. Urinary thioether excretion (mmole/mole creatinine) was significantly elevated in the BR-S group (4.59 + 0.52; p < 0.001) but it was lower in the BR-K group (0.54 + 0.08; p < 0.001) compared to the control (1.83 + 0.34). Furthermore, /3-glucuronidase-treated samples from both groups of bidi rollers exhibited increased mutagenicity to TA98 compared to the control group; in addition, BR-S samples exhibited direct mutagenicity to TA98. The results show that occupational tobacco exposure modulates the glutathione conjugation pathway and increases the mutagenic burden of bidi rollers.

Epidemiological studies have shown that smokeless tobacco habits are causally related to the high incidence of oral cancer in India (Jayant and Notani, 1991). Besides habit-related exposure, workers engaged in the manufacture of 'bidis', the Indian version of cigarettes, constitute a

Correspondence: Dr. (Mrs.) R.A. Bhisey, Carcinogenesis Division, Cancer Research Institute, Tata Memorial Centre, Dr. E. Borges Marg, Parel, Bombay 400012 (India).

group that is exposed occupationally to large quantities of smokeless tobacco. Bidis are made by placing a small quantity of bidi tobacco (Nicotiana tabacum) on a dry 'temburni' leaf (Diospyros melanoxylon) and rolling it by hand into a bidi. Thus, bidi rollers, mostly women, are exposed to tobacco constituents via the cutaneous route or through inhalation of tobacco dust. On average, each worker makes 500-1000 bidis per day and handles 225-450 g of tobacco. In the present study 2 groups of female bidi rollers were

140 monitored for biological effects of prolonged exposure to bidi tobacco. Bidi rollers and control individuals without any tobacco habit were included in order to delineate the effects of occupational tobacco exposure. Many hazardous chemicals are metabolised in the body to reactive electrophiles that can interact with cellular macromolecules to produce genotoxicity (Chasseaud, 1979). Glutathione and glucuronic acid prevent this hazardous interaction by conjugating non-specifically with many electrophiles which are then excreted in urine as thioethers and glucuronides respectively (Sorsa et al., 1982). Therefore, the electrophilic burden imposed by innumerable chemicals in tobacco was assessed by estimating urinary thioether and glucuronide levels. Cotinine, the major metabolite of the tobacco-specific alkaloid nicotine (Beckett et al., 1971), served as an indicator of specific tobacco exposure. Earlier studies have shown a good correlation between urine mutagenicity and the degree of exposure to tobacco smoke (van Doorn et al., 1979). Hence exposure of bidi rollers to mutagens in smokeless tobacco was analysed by determining urine mutagenicity with the Ames assay. Comparative data on urinary cotinine, thioethers and glucuronide excretion as well as urine mutagenicity in bidi rollers and control subjects are presented in this communication. Material and methods

Chemicals 5,5'-Dithiobis-(2-nitrobenzoic acid) (DTNB), N-acetyl-L-cysteine, chloramine-T, cotinine, nicotinamide adenine dinucleotide phosphate (NADP), histidine, biotin, glucose-6-phosphate (G-6-P) and XAD-2 resin were purchased from Sigma (U.S.A.). Diethylthiobarbituric acid (DETB) and potassium cyanide were obtained from Aldrich (U.S.A.). Agar was obtained from Difco (U.S.A.). All other reagents used were of analytical grade.

Subjects The population under study consisted of 2 groups of female bidi rollers located 11 km apart and designated BR-K (n = 26) and BR-S (n = 18). All the subjects were free of tobacco habits and

belonged to the same community, shared common food habits and purchased tobacco for rolling bidis from the same source. However, the 2 groups of bidi rollers differed in their living conditions. BR-K individuals resided in the most densely populated part of Bombay, while BR-S individuals lived in a less congested area with easy access to open spaces. Age- and sex-matched non-bidi rollers (n = 26) without tobacco habits served as controls. Information regarding age, diet, number of bidis rolled per day, years of exposure and medical history was recorded for all the subjects.

Exposure conditions Workers were engaged in rolling bidis for 4-6 h each day of the week. In the BR-K group 8-10 women rolled bidis in a poorly ventilated room, while BR-S individuals worked singly in open courtyards.

Sample collection 8-h urine samples ( ~ 500 ml) were collected, starting with the early-morning voided sample. Upon arrival in the laboratory, an aliquot was used for routine analysis while remaining samples were frozen at - 20 ° C without preservative until assayed.

Estimation of cotinine Cotinine levels were estimated using the method described by Peach et al. (1986). After mixing with 0.4 ml of 4 M sodium acetate buffer, pH 4.7, the urine sample (1 ml) was treated sequentially with 0.2 ml each of freshly prepared 10% aqueous potassium cyanide, 10% aqueous chloramine-T and 1 ml of a 1% solution of DETB prepared in acetone-water (1:1, v/v). The generation of a pink colour after 20-min incubation at room temperature indicated a positive reaction. The reaction product was extracted into 2 ml ethyl acetate. To 1.5 ml of the extract, 0.2 ml of ethanol was added and the optical density was measured at 532 nm using a Uvicon double beam spectrophotometer. The concentration of nicotine metabolite was calculated using cotinine as standard. All urine samples in which cotinine was not detected were retested after concentration on an XAD-2 column. Recovery of cotinine was checked by passing through the column, a cotinine solution of known concentration. Upon elu-

141 tion, 80% of cotinine was recovered in a reproducible manner and this factor was taken into account for quantitation. Comparison of cotinine estimated in urine samples spiked with cotinine and their XAD-2 concentrates gave similar results indicating that concentration of other urine constituents did not interfere with the assay. The sensitivity of the colorimetric method was found to be 0.5/zg of cotinine per ml of urine per se, so that 100-fold concentration brought the detection level to 5 n g / m l .

Estimation of thioethers A 5-ml urine sample was acidified to pH 2 with 4 N hydrochloric acid and extracted with 20 ml ethyl acetate. The organic fraction was evaporated to dryness and the residue was dissolved in 2 ml distilled water. After addition of 0.5 ml of 4 N sodium hydroxide to 1 ml of the extract, the solution was heated in a boiling water bath for 50 rain. Hydrolysis was carried out under inert conditions by replacing the air space of the screwcapped tube with nitrogen. After cooling to 0 ° C, 0.5 ml of 4 N HC1 (prepared in 0.65 M phosphate buffer, p H 7.1) was added to the hydrolysate. The reaction mixture consisted of 0.5 M phosphate buffer, pH 7.1 (2 ml), 0.3 ml of DTNB solution (0.4 mg D T N B / m l of 1% aqueous sodium citrate) and 0.25 ml of the neutralized hydrolysate. Absorbance was read at 412 nm using a Uvicon double beam spectrophotometer. Correction was done for the colour of urine and concentration of thioethers was derived from the reference curve obtained with N-acetyl-L-cysteine (Heinonen et al., 1983).

Estimation of glucuronides A 5-ml urine sample was reacted with 1.0 ml of a 1 % solution of naphthoresorcinol in 95% alcohol, and 5 ml of concentrated HC1. The temperature of the reaction mixture was raised gradually and maintained at boiling point for 1 min, shaking the tube continuously. After cooling to room temperature the sample was extracted with an equal volume of ether. The appearance of a violet-red colour indicated the presence of glucuronides. Absorbance was read at 570 nm and quantitation was done using sodium glucuronate as standard (Oser, 1964).

Concentration of urine Urine samples were thawed, filtered through Whatman No. 1 filter paper and passed through a glass column (1.5 cm x 10 cm) containing 10 g XAD-2 resin according to the method described by Clonfero et al. (1986). The resin was washed sequentially with acetone, a c e t o n e : w a t e r (1:1), methanol and water. After passing the sample through the column (percolation speed 3-5 ml/min), the column was washed twice with water before elution with 15 ml acetone followed by 20 ml water. The eluates were lyophilised and solubilised in dimethyl sulphoxide (0.4 ml/100 ml urine).

Salmonella/microsome mutagenicity assay Salmonella typhimurium strains TA98 and TA100 were kindly provided by Professor B.N. Ames (University of California, Berkeley, CA, U.S.A.). The modified preincubation procedure of Malaveille et al. (1989) was used to test urine concentrates for mutagenicity. SIX samples each from the control and BR-K groups and 5 from the BR-S group were randomly selected for mutagenicity testing.

Liquid preincubation assay Urine concentrates were tested without prior modification or were pretreated with 200 units of /3-glucuronidase or 0.2 M sodium nitrite (pH 2) at 37 °C for 20 min. The samples were assayed at 3 - 4 concentrations equivalent to 0.3-1.25 ml of urine. The assay mixture, consisting of 100/zl of 2-fold concentrated bacterial culture ( 4 x 108 cells), 100/zl $9 mix and appropriate amounts of urine concentrates, was incubated at 37 ° C for 90 rain. In experiments without $9, an equivalent amount of phosphate buffer (pH 7.4) was added. After mixing with 2 ml of histidine-poor soft agar, the mixture was plated onto minimal glucose agar. 1 ml of $9 mix contained 100 ~zl each of 80 mM MgCI2, 8 mM NADP, 50 mM G-6-P, 300 /zl of 0.25 M Sorensen's phosphate buffer (pH 7.4) and 50 /xl $9 prepared from the livers of Aroclortreated male Sprague-Dawley rats. A sample was considered mutagenic if it induced a reproducible dose-related increase in the number of mutant colonies and caused a 2-fold increase in the number of revertants over the

142 spontaneous reversion rate. Based on these criteria, a revertant number greater than 0.2 × 10 6 per mole creatinine was found to be mutagenic.

Creatinine estimation Creatinine was determined with the alkaline picrate method (Varleg, 1967). Values of urinary cotinine, thioethers, glucuronides and mutagenicity are expressed per mole creatinine in order to compensate for differences in concentration of urine.

TABLE l URINARY COTININE, THIOETHER AND GLUCURONIDE LEVELS (mmole/mole CREATININE) IN BIDI ROLLERS AND CONTROL SUBJECTS (MEAN+ SE) Group Control BR-K BR-S

Cotinine ND 0.79+0.30 0.09_+0.03

Thioethers 1.83 -+0.34 0.54_+0.08* 4.59_+0.52*

Glucuronides 72.07 + 8.46 78.97_+12.92 115.36_+15.46

ND, not detectable. * p < 0.001 for comparisons between BR and control.

Statistical analysis Data from bidi rollers and controls on urinary cotinine, thioethers and glucuronides were statistically evaluated using Student's t-test for comparison of means from 2 independent samples. When the assumption of equality of the compared groups did not hold, Cochran's approximation to the B e h r e n s - F i s h e r test was used. The Wilcoxon test was applied for comparative evaluation of urine mutagenicity. The influence of occupational tobacco exposure on the nature and extent of mutagenic activity was assessed by comparing the mutagenicity of samples from bidi rollers and control subjects by the Wilcoxon rank sum test. In order to ascertain the nature of mutagens, the mutagenic activity of unmodified samples was compared with that of enzymatically or chemically modified samples by the Wilcoxon signed rank test. Correlation between 2 independent parameters such as mutagenic activity and thioethers was determined using the correlation matrix of the computer program 'Statgraphics'. Results

Evaluation of questionnaire and urine analysis The age (mean +_ SE) of the control, B R - K and BR-S groups was 29 +_ 2, 24 + 1 and 26 _+ 2 years respectively. The average number of bidis rolled per day by B R - K and BR-S workers was 717 + 53 and 632 +_ 49 respectively, the range of bidis rolled per day being 250-1000. Routine urine analysis revealed an absence of ketone bodies and sugar indicating a normal metabolism of fat and carbohydrates in bidi rollers and control subjects. Renal pathology was ruled out as none of the samples exhibited the presence

of casts, abnormal types of crystals or cells. Bile salts were not detected while urobilinogen was within normal limits, thereby ensuring normal liver function. The creatinine levels (mean + SE) for control, B R - K and BR-S groups were 63 + 6, 88 + 11 and 57 + 5 mg % respectively.

Urinary cotinine levels Table 1 shows that cotinine was not detected in any of the control samples. However, 22 bidi rollers from the B R - K (84.6%) and 13 from the BR-S group (72.2%) showed the presence of urinary cotinine. The mean cotinine level was significantly higher in the B R - K group ( p < 0.05) than in the BR-S group.

Urinary thioether and glucuronide levels As shown in Table 1, urinary glucuronide excretion in the 2 groups of bidi rollers was similar to that in the control group. However, the m e a n thioether excretion was significantly lower in the B R - K group, while in the BR-S group it was significantly higher than in the controls. In the control group, the mean urinary thioether level was 1.83 + 0.34 with a 95% confidence interval of 1-3 m m o l e / m o l creatinine. Hence, thioether levels of < 1, 1-3, > 3 were referred to as low, intermediate and high respectively. Control subjects were distributed evenly over low (42.3%), intermediate (30.8%) and high (26.9%) excretor categories, while 92% of the B R - K subjects were low excretors and the remainder belonged to the intermediate range. On the other hand, none of the BR-S workers were low thioether excretors, 33% were in the intermediate range and the majority (67%) exhibited high levels. Although

143

the duration of occupational exposure did not influence urinary thioether levels, in each group workers who rolled up to 1000 bidis per day showed higher urinary thioether excretion than those who made up to 500 bidis. The increase was statistically significant in the BR-K group (0.8 _+ 0.11 vs. 0.48 + 0.07; p < 0.05). Urine mutagenicity Tables 2 and 3 show the mutagenic activity exhibited by samples from bidi rollers and controls to strains TA98 and TA100 respectively, while the comparative mean mutagenic activity is depicted in Figs. 1 and 2. TA98 Control Table 2 shows that in the absence or presence of $9, 2/6 and 3/6 samples were mutagenic while /3-glucuronidase treatment rendered 3 samples mutagenic and only 1 sample

elicited a mutagenic response after fl-glucuronidase + $9 treatment. All the samples exhibited mutagenicity after treatment with sodium nitrite, the mutagenic potency decreasing after further incubation with $9. However, the mutagenic potency of samples treated with sodium nitrite was higher than that exhibited by unmodified samples (p < 0.04). BR-K. Only 1 sample exhibited direct mutagenicity while 3 samples were mutagenic in the presence of $9, the mean mutagenic potency (Fig. 1) being similar to that of the control. Upon deconjugation with fl-glucuronidase all the samples were mutagenic, while the mutagenic response was reduced after fl-glucuronidase + $9 treatment. Furthermore, the mean mutagenic potency of fl-glucuronidase-treated samples was higher (p < 0.02) than that of the controls (Fig. 1). Whereas 5 samples treated with sodium nitrite + $9 were mutagenic, the mutagenic response of

TABLE 2 URINE MUTAGENICITY Group

a

IN TA98

Modifier None

$9

fl-gluc

/3-gluc + $9

NaNO 2

NaNO 2 + $9

Control

0.31 0.91 NM NM NM NM

0.69 NM 0.74 1.66 NM NM

NM 0.26 0.41 NM NM 0.21

0.63 NM NM NM NM NM

1.26 3.50 1.02 2.84 1.99 1.61

0.47 1.82 0.78 1.66 0.77 0.63

BR-K

NM 0.31 NM NM NM NM

NM NM NM 1.66 0.41 0.75

0.63 0.21 1.46 1.30 0.85 0.92

NM NM 0.49 0.65 NM NM

NM NM NM 0.24 NM NM

0.21 0.49 1.11 0.72 1.13 NM

BR-S

NM 1.54 0.72 1.90 1.53

NM 1.67 1.78 NM 0.38

0.82 NM 0.48 1.62 NM

NM NM NM NM NM

2.82 6.28 1.20 8.56 2.93

0.77 1.92 2.22 7.23 2.31

a Expressed as number of induced revertants × 1 0 - 6 / m o l e creatinine. The values represent mutagenic activity of individual samples in the respective groups. Spontaneous revertant number: 2 0 _ 2 (18-22). NM, non-mutagenic;/3-gluc,/3-glucuronidase.

144 TABLE 3 URINE MUTAGENICITY Group

a

IN TA100

Modifier None

$9

fl-gluc

/3-gluc +$9

NaNO 2

NaNO 2 +$9

Control

NM NM NM NM NM NM

0.88 0.65 2.42 2.37 1.46 2.40

2.89 NM 1.39 3.32 NM NM

1.82 0.78 1.19 1.25 NM NM

5.67 18.61 5.82 18.01 18.51 6.21

2.33 8.95 3.11 8.77 7.10 2.62

BR-K

0.36 0.21 NM NM NM NM

NM 0.31 1.95 1.16 NM NM

NM NM NM NM NM NM

NM 0.56 0.90 NM NM NM

1.63 1.15 NM NM 0.35 0.29

0.41 0.79 2.57 0.27 0,21 0.27

BR-S

NM NM NM NM NM

NM 7.10 0.24 0.76 0.77

NM NM NM 0.52 NM

1.95 16.02 0.23 2.85 1.66

1.13 NM NM 10.08 2.34

0.59 NM 0.58 15.03 3.61

a Expressed as n u m b e r of induced revertants x 1 0 - 6 / m o l e creatinine. The values represent mutagenic activity of individual samples in the respective groups. Spontaneous revertant number: 1 0 0 + 3 (80-120). NM, non-mutagenic; fl-gluc,/3-glucuronidase.

/,D tO

X .¢

.S

1'4

i

_~

1"2

6

~J

.z

+1

0.8 06

®~ 0.4 ";

0-2 MODIFIERS :

none

Sq I~ICONTROL

J3-g luc ~

BR-K

J'3-g luc+ Sq []

BR-S

nitrite

nitrite+

Sq

4(- P < 0"05

Fig. 1. Comparison of mutagenic activity in samples from bidi rollers and controls to TA98. Data are expressed as m e a n n u m b e r of revertants per mole creatinine_+ SEM: * indicates statistical significance ( p < 0.05) obtained on comparing mutagenic activity in bidi rollers and controls by the Wilcoxon test; fl-gluc =/3-glucuronidase.

145

~o

4.2t_

~

1"8

~

1.6

"

'+7. o

1"4

u

3

+1

I

m

'2IT 1

I

o-8 c X •o -

, °1o

ill,i:

0.6

,++-Io~,

0.4 o 6

I

6

0,:~ °.,

::::::

z

MODIFIERS

:

none

Sq

DCONTROL

~-gluc

~

BR-K

.~-gluc

+

~

Sq

nitrite

BR-S

-X- P