Mutation Research,

244 (1990) 15-20

15

Elsevier MUTLET 0335

Genotoxicity of two metabolites of benzene: phenol and hydroquinone show strong synergistic effects in vivo R. B a r a l e , A. M a r r a z z i n i l, C. B e t t i I, V. V a n g e l i s t i l, N . L o p r i c n o I a n d I. B a r r a i Istituto di Zoologia, Universit& di Ferrara (Italy) and IDipartimento di Scien=e dell'Ambiente e de/Territorio. Universita di Pisa (Italy)

(Accepted 5 December 1989)

Keyword~." Benzene;Hydroquinone;Phenol; Interaction;Synergism;Micronuclei

Summary Possible interactions between hydroquinone (HQ) and phenol (PHE), 2 known benzene metabolites, in inducing micronuclei in mouse bone marrow cells were investigated. HQ and P H E administered alone gave weak and negative results, respectively, at the doses tested. However, simultaneous administration of both compounds caused a considerable increase in the induction of micronuclei as well as an increase in bone marrow toxicity. Using 3 different statistical methods, it was shown that the observed joint effect was significantly higher than additive interaction, and was close to multiplicative interaction. These findings bring further support to the hypothesis that the toxic and genotoxic effects of benzene are produced by several metabolites acting synergistically.

The acute toxicity, carcinogenic and mutagenic effects of benzene have been widely investigated and reported in several reviews (see Aksoy, 1989; Dean, 1985; IARC, 1982; NTP, 1986). These effects are likely to be produced by mechanisms which involve benzene metabolites (Glatt et al., 1989). To date, none of the known benzene metabolites can explain, assayed alone, the amount of the genotoxic effects produced upon administration of the parent compound. This failure of known metabolites to reproduce benzene-induced genotoxicity led some authors to formulate the hypothesis that such toxicity may be the result of Correspondence: Prof. R. Barale, lstituto di Zoologia, Via L. Borsari 46, 46100 Ferrara (Italy).

the joint effects of metabolites rather than the result of any individual metabolite. In particular, it was reported that a positive interaction exists between 2 major benzene metabolites, phenol and hydroquinone, in producing myelotoxicity in mice (Eastmond et al., 1987). We studied the induction of micronuclei in mouse bone marrow after administration of several derivatives of benzene, including phenol and hydroquinone. We found that the action of individual metabolites cannot account for the benzene genotoxicity, even upon addition of the effects of all the metabolites assayed (Ciranni et al., 1988a,b). In the present work, we analyze the interaction of phenol (PHE) and hydroquinone (HQ) in induc-

0165-7992/90/$ 03.50 (c) 1990 ElsevierScience Publishers B.V. (BiomedicalDivision)

16 ing micronuclei in mouse polychromatic erythrocytes (PCEs), applying several models through which the interactions between compounds in causing genotoxicity can be assessed. We analyzed our data with the model proposed by Shaeffer et al. (1982), with a model derived from Shaeffer which we present here, and with a third model proposed by Hass et al. (1987) which tests for multiplicative effects of 2 compounds. Materials and methods

Micronucleus assay Swiss CD-1 male mice about 6-8 weeks old (Charles River, Calco, Italy) were used. Animals were randomized and treated in groups of 3 with the solvent only (water) or test chemicals (single substances or mixtures): hydroquinone (CAS No. 123-31-9, Sigma) 40, 60, 80 mg/kg, and phenol (CAS No. 108-95-2, Ega-Chemie, Steinheir/ Albuch) 40, 80, 160 mg/kg. Nine mixtures of HQ and P H E were prepared in all possible combinations of the single doses. Compounds and their mixtures were administered by intraperitoneal (i.p.) injection. Animals were killed by cervical dislocation 18 h after treatment and bone marrow cells were obtained and processed according to Schmid (1975). Bone marrow analysis was performed on at least 3 animals for each experimental point and 3000 PCEs were scored for each animal. During PCE counting, the number of normochromatic erythrocytes (NCEs) was also counted, to evaluate the P C E / N C E ratio as an index of bone marrow toxicity. Statistical analysis Shaeffer et al. (1982) proposed a methodology for the analysis of interactive response in net events (observed events minus spontaneous events). The model is, in its simplest formulation for 2 substances only: Y = b l X l + b2X2 which is equivalent to a regression in 2 variables

forced through the origin, i.e., without the constant term of the regression. The interaction is investigated comparing the effect of the mixture with the sum of the separate effects of individual substances, called the standard. The expected values of the mixture are obtained by the multiple regressions given above, and the expected values of the standard by adding the values of the individual linear regressions for each compound. We have also slightly modified the technique of Shaeffer et al., in the sense that we compute the expected values of the standard from the multiple regression forced through zero, fitted to the sum of net individual effects. A third method was proposed by Hass et al. (1987), which specifically investigates multiplicative effects of the doses of the 2 substances. This method fits to the net revertants the equation: Y = bl×Xl

+ b2×X2 + bl2xXlxX2

A third independent variable is fitted to the regression, which is the product of the 2 primitive variables. We have prepared a computer program which allows the analysis of the interaction between 2 compounds according to (i) the technique of Shaeffer et al., (ii) our modification, and (iii) the method of Hass et al. The software (MUTAPAK) is available upon request. Results and discussion

The effects of the individual metabolites are reported in Table 1. In the Table, we give the induced micronuclei at different doses in polychromatic erythrocytes (MNPCEs), in normochromatic erythrocytes (MNNCEs), and also the ratio between normochromatic and polychromatic erythrocytes (NCE/PCE). We harvested cells 18 h after treatment, since in previous studies this was the time of maximum expression of micronuclei at the highest dose of hydroquinone (Ciranni et at., 1988b). A significant increase in micronuclei in PCEs is observed as a function of dose for hydroquinone, whereas phenol does not show any

17

TABLE I MICRONUCLEUS INDUCTION AFTER SINGLE i.p. ADMINISTRATIONS OF H Y D R O Q U I N O N E AND PHENOL Dose (mg/kg)

MNPCEs (% ± SD)

MNNCEs (07o _ SD)

NCE/PCE ( ± SD)

Control

0.109 _+ 0.049 0.183 ± 0.026 1.066 _+ 0.036

Hydroquinone 40 0.174 ± 0.020 0.150 ± 0.040 1.107 +_ 0.081 60 0.230 ± 0.002 0.163 _ 0.062 1.017 + 0.451 0.537 ± 0.015 0.210 ± 0.059 1.041 ± 0.037 80 Phenol 0 40 80 160

0.109 0.111 0.118 0.098

± ± ± ±

0.049 0.018 0.016 0.001

0.183 0.127 0.120 0.118

_+. 0.026 _+ 0.049 _ 0.078 + 0.002

1.066 1.058 1.077 1.104

± ± +_ ±

0.036 0.040 0.021 0.012

For each experimental point 3 animals have been analyzed.

genotoxic effect. No appreciable myelotoxicity is produced by either compound at the doses tested, as visible from the N C E / P C E ratio. In Table 2, the effects of the mixture of the 2 metabolites are given at different combinations of doses. A more than 2-fold increase in the number of micronuclei is observed in PCEs as compared to TABLE 2 MICRONUCLEUS INDUCTION AFTER SINGLE i.p. ADMINISTRATIONS OF VARIABLE MIXTURES OF HYDROQ U I N O N E (HQ) AND P H E N O L (PHE) Dose (mg/kg)

MNPCEs (°70 + SD)

MNNCEs (% ± SD)

NCE/PCE (_+ SD)

HQ

PHE

0

0

0.109 + 0.049 0.183 _+ 0.026 1.066 _ 0.036

40

40 80 160

0.368 ± 0.024 0.127 ~ 0.045 1.109 ± 0.038 0.461 ± 0.032 0.149 + 0.026 1.099 _+ 0.032 0.549 _+ 0.017 0.108 + 0.017 1.115 +_ 0.008

60

40 80 160

0.372 :t: 0.019 0.100 ± 0.010 1.321 + 0.145 0.406 + 0.018 0.101 ± 0.027 1.313 ± 0.031 0.578 ± 0.021 0.189 ± 0.064 1.333 + 0.019

80

40 80 160

0.790 ± 0.140 0.100 ± 0.024 1.316 _ 0.018 0.944 ± 0.103 0.106 ± 0.036 1.544 + 0.043 1.204 ± 0.057 0.133 ± 0.061 1.550 ± 0.060

For each experimental point 3 animals have been analyzed.

O

i

Fig. I. Observed interaction between phenol and hydroquinone in inducing micronuclei in bone marrow cells. Plane coordinates are the doses, given in mg/kg of body weight. The vertical coordinate represents the response measured as the number of micronuclei per 100 polyehromatic erythrocytes.

the number of micronuclei observed with the same dose of hydroquinone administered alone. Greater than additive effects are apparent even at a simple inspection of the data (Fig. 1). A similar, although less intense phenomenon, is observed for myelotoxicity expressed as an increase of the N C E / P C E ratio. These results are in agreement with the myelotoxic effects observed by Eastmond et al. (1987). They observed that phenol is not toxic, hydroquinone is weakly toxic, but the mixture induces a dramatic and significant decrease in bone marrow cellularity. In the present case, the interaction between phenol and hydroquinone is better described by the appropriate statistical analysis. We have used 3 methods aimed at the same purpose, namely of detecting significant additive, synergistic, or antagonistic effects. The method of Shaeffer et al. (Table 3) is based on the comparison between the means of the expected values of the mixture and of the standard. The means, with their standard deviations, and the Student t's are given in the last 3 lines of Table 3. With this method, positive values of the expected for the mixture, minus the expected for the standard, give a positive signifi-

18 TABLE 3 ANALYSIS OF INTERACTION BETWEEN HQ AND PHE ACCORDING TO SHAEFFER ET AL. (1982) Dose (mg/kg)

Net micronuclei

HQ

PHE

M

EM

S

ES

M - EM

M - S

EM - ES

40 40 40 60 60 60 80 80 80

40 80 160 40 80 160 40 80 160

25.9 35.2 44.0 26.3 29.7 46.9 68.1 83.5 109.5

33.7 38.6 48.6 48.0 53.0 63.0 62.4 67.4 77.3

6.7 7.4 5.4 12.3 13.0 11.0 42.9 43.6 41.7

15.0 14.9 14.7 22.6 22.5 22.4 30.2 30.1 29.9

-

19.2 27.8 38.6 13.9 16.7 35.9 25.1 39.9 67.8

18.5 23.7 33.9 25.3 30.4 40.6 32.1 37.2 47.4

52.0 27.5

54.6 13.2

20.5 15.9

22.5 6.1

- 2.7

31.5

32.0 8.4

7.9 3.5 4.7 21.8 23.4 16.2 5.6 16.0 32.1

R 2= 0.914

Mean SD Student's t

17.2 - 0.46

15.6 6.09

11.4

M, mixture; EM, expected mixture, S, standard; ES, expected standard.

cant d i f f e r e n c e when the 2 substances act synergistically, no significance when they act a d d i t i v e l y , a n d a negative significant d i f f e r e n c e when there is a n t a g o n i s m . In the present case, the a v e r a g e difference between the m i x t u r e a n d the s t a n d a r d is highly significant, with t = 6 . 0 9 , P < 0 . 0 0 1 ; the a v e r a g e difference between the expected for the m i x t u r e and the expected for the s t a n d a r d is even m o r e significant ( t = 11.40, P < 0 . 0 0 0 1 ) . The difference is on average 3 times the s p o n t a n e o u s level o f micronuclei. Using o u r m o d i f i c a t i o n o f the S h a e f f e r m e t h o d ( T a b l e 4), we o b t a i n very much the s a m e results a n d the s a m e levels o f significance; however, the values o f the t ' s are smaller, r e n d e r i n g o u r test less sensitive in the d e t e c t i o n o f i n t e r a c t i o n s than the o r i g i n a l m e t h o d . O n the o t h e r h a n d , the m e t h o d is m o r e conservative, in the sense that at b o r d e r l i n e values o f significance, it w o u l d reject the conclusion o f either synergistic or a n t a g o n i s t i c effects in f a v o r o f a d d i t i v i t y . W h e n the synergism is as d r a m a t i c as in the present case, it m a k e s little difference w h e t h e r the original m e t h o d o f S h a e f f e r or o u r m o d i f i c a t i o n is used.

T h e third m e t h o d we used is specific for m u l t i p l i c a t i v e effects between doses. In T a b l e 5, we give the results o f the analysis a c c o r d i n g to Hass et al.; here, m u l t i p l i c a t i v e effects are revealed by the significant t for the regression coefficient o f the p r o d u c t o f the 2 doses ( t = 1 . 9 7 , b o r d e r l i n e significance). T h e r e f o r e , the effect o f h y d r o q u i n o n e in c o m b i n a t i o n with p h e n o l is a l m o s t m u l t i p l i c a t i v e , which is a special case o f high synergism. It is possible, however, that the m e t h o d o f Hass et al. might give a n o n - s i g n i f i c a n t result, while the o t h e r 2 m e t h o d s w o u l d e q u a l l y detect a synergism less t h a n multiplicative a n d m o r e t h a n additive. W e believe that the analysis o f the effects o f mixtures with the 3 m e t h o d s which we have used here gives a b e t t e r r e s o l u t i o n o f the several elements involved in the i n t e r a c t i o n between the 2 substances. Conclusions

T h e o b s e r v e d p o w e r f u l synergism between h y d r o q u i n o n e a n d p h e n o l in inducing m y e l o t o x i c i ty a n d g e n o t o x i c i t y can explain, at least partially,

19

TABLE 4 ANALYSIS OF INTERACTION BETWEEN HQ AND PHE EXPECTED FOR THE STANDARD CALCULATED WITH MULTIPLE REGRESSION OF STANDARD

Dose

Net micronuclei

(mg/kg) EM

S

ES

M - EM

M - S

EM - ES

33.7 38.6 48.6 48.0 53.0 63.0 62.4 67.4 77.3

6.7 7.4 5.4 12.3 13.0 I 1.0 42.9 43.6 41.7

16.6 13.4 6.9 26.5 23.3 16.8 36.5 33.2 26.8

-7.9 - 3.5 -4.7 -21.8 -23.4 - 16.2 5.6 16.0 32. I

19.2 27.8 38.6 13.9 16.7 35.9 25.1 39.9 67.8

17.0 25.2 41.7 21.4 29.7 46.1 25.9 34.1 50.5

SD

- 2.6 18.1

31.6 16.5

32.4 11.5

Student's t

- 0.431

5.7

HQ 40 40 40 60 60 60 80 80 80

PHE

M

40 80 160 40 80 160 40 80 160

25.9 35.2 44.0 26.3 29.7 46.9 68.1 83.5 109.5

Mean

8.45

M, mixture; EM, expected mixture; S, standard; ES, expected standard.

the benzene effects. This does not exclude the possibility that other benzene metabolites, including trans, trans-muconic acid (Rossman et al., 1989), catechol, p-benzoquinone and 1,2,4-benzenetriol, which have been shown to be active in

vitro (Erexson et al., 1985), may play an important role in the interaction also in vivo. The nature of the observed synergism is not clearly understood, although Eastmond et al. (1986, 1987) and Smith et al. (1989) presented find-

TABLE 5 ANALYSIS OF INTERACTION A C C O R D I N G TO HASS ET AL. (1987) FOR MULTIPLICATIVE EFFECTS

Metabolite

Regression coefficient

Standard error

Student's t

HQ PHE HQ × PHE

0.539 - 0.232 0.0074

0.199 0.210 0.0038

2.71 - 1.10 1.97

HQ 40 40 40 60 60 60 80 80 80

PHE 40 80 160 40 80 160 40 80 160

HQ x PHE 1600 3200 6400 2400 4800 9600 3200 6400 12800

Observed

Expected

25.9 35.2 44.0 26.3 29.7 46.9 68.1 83.5 109.5

24.3 26.8 31.9 41.1 49.6 66.4 57.9 72.3 100.9

Coefficient of correlation between model and data r = 0.951, df= 5, P