Structure-activity relationships of tricyclic antidepressants and related compounds in the wing somatic mutation and recombination test of Drosophila melanogaster

Mutation Research, 286 (1993) 155-163 © 1993 Elsevier Science Publishers B.V. All rights reserved 0027-5107/93/$06.00

155

MUT 05212

Structure-activity relationships of tricyclic antidepressants and related compounds in the wing somatic mutation and recombination test of Drosophila melanogaster Nancy van Schaik a,b and Ulrich Graf a a

Institute of Toxicology, Swiss Federal Institute of Technology and University of Zurich, Schwerzenbach, Switzerland and b Department of Genetics, University of the Witwatersrand, Johannesburg, South Africa (Received 9 July 1992) (Revision received 15 September 1992) (Accepted 16 September 1992)

Keywords: Tricyclic antidepressants; Genotoxicity; Somatic cells; Drosophila; Chlorpromazine; Clomipramine; Lofepramine; Maprotiline; Mianserin

Summary Four antidepressants and one neuroleptic drug were tested for genotoxicity using the somatic mutation and recombination test (SMART) in wing cells of Drosophila melanogaster. Three-day-old larvae trans-heterozygous for two linked recessive wing hair mutations (multiple wing hairs and flare) were fed the test compounds in water or solvents mixed with a standard dry food for 48 h. Wings of the emerging adult flies were scored for the presence of spots of mutant cells which can result from either somatic mutation or mitotic recombination. The tricyclic antidepressant clomipramine, which is closely related to imipramine (previously shown to be genotoxic in somatic cells of Drosophila), was clearly genotoxic at concentrations above 10 mM. The structurally related antidepressants lofepramine and mianserin were positive only at 100 mM which is the maximum tolerated dose. The antidepressant maprotiline and the antipsychotic chlorpromazine, which are distinguished from the other compounds by a 6-membered central ring instead of a 7-membered one, were not genotoxic in the same dose range. These results lend further support for the hypothesis that an N atom in the heterocyclic 7-membered ring of the tricyclic molecule is responsible for the genotoxic property of the compounds in Drosophila.

The wing spot test SMART, which detects somatic mutation and recombination events on the third chromosome of Drosophila melanogas-

Correspondence: Dr. U. Graf, Institute of Toxicology, ETH and University of Zurich, P.O. Box 550, CH-8603 Schwerzenbach, Switzerland.

ter, has proven useful as a test for the detection of a wide variety of both direct acting mutagens and those requiring bioactivation (Graf et al., 1984, 1989; Graf and van Schaik, 1992). This test is easy, quick and relatively inexpensive for an assay using a eukaryotic organism. SMART is also appropriate for the study of structure-activity relationships (Alonso Moraga and Graf, 1989;

156 Frei et al., 1992b; Graf et al., 1992; Negishi et al., 1991). We are using the test in this way to study a group of compounds widely used in the treatment of psychiatric patients, the tricyclic antidepressants and other psychoanaleptics. These studies should add to the further validation of the wing spot test SMART and provide information about the possible genotoxic properties of a group of compounds administered to large numbers of patients worldwide. We recently reported that imipramine and desipramine are clearly mutagenic at concentrations of 10 and 1 mM, respectively, in the standard Drosophila wing spot test while amitriptyline, protriptyline and nortriptyline are negative at concentrations of 100 mM (van Schaik and Graf, 1991). A comparison of the structures of these closely related tricyclic antidepressants led us to propose that the nitrogen atom at position 5 in the central 7-membered ring was the feature most likely to be responsible for genotoxicity in this test. We report here the results in the SMART test of two additional tricyclics, clomipramine and lofepramine, and the tetracyclic mianserin, as well as of maprotiline and the structurally similar phenothiazine, chlorpromazine. Clomipramine was chosen because of its similarity to imipramine except for the extra chlorine atom and lofepramine because of the bulky side chain. According to our hypothesis, both of these should prove mutagenic. Mianserin is interesting since it contains no side chain and is a tetracyclic. Although it lacks the N atom at position 5, the 7-membered ring contains an N at position 10. Maprotiline differs from the typical tricyclics in that the central ring is a bicyclo-octadiene. Its aliphatic side chain is identical to that of protriptyline and desipramine. It contains no N atom in the main ring structures. Chlorpromazine is similar to clomipramine because it also contains the N atom at position 5, a CI attached to C7 and an identical side chain but differs in the 6-membered central ring with sulfur at position 10. There is little information in the literature on genotoxicity studies with the four antidepressants reported on here. Ziebarth (1982) reported that clomipramine, like imipramine and desipramine, is nitrosatable at pH 2 which may imply that these compounds could be converted to N-nitroso

compounds known to be mutagenic and carcinogenic but found that the nitrosation of this group of substances was markedly decreased under conditions simulating those of the human stomach. Clomipramine has also been reported to have a cytostatic effect on a renal cancer cell line but the mechanism is unclear (Sauter and Ernst, 1991). Maprotiline was found by Hess et al. (1973) to have no dominant lethal effect in mice but, like clomipramine, had a cytostatic effect in the study of Sauter and Ernst (1991). Chaplin (1986) in a review of reports of blood dyscrasias concludes that mianserin appears to be more toxic to bone marrow than related compounds but the mechanism is unclear. There are many conflicting reports about the genotoxicity of chlorpromazine. In different laboratories using the Salmonella/microsome test the compound was found to be negative with or without $9 addition (Balbi et al., 1980), positive when photoactivated (De Mol et al., 1986), and positive but requiring microsomal activation (ObaseikiEbor and Akerele, 1988). The evidence on cytogenetically detectable damage is also confusing both for in vitro and for in vivo studies. For example, for sister chromatid exchanges (SCE) Speit (1982) found no effect in an analysis of Chinese hamster bone marrow after exposure in vivo while Yu et al. (1988) found a positive effect in human peripheral lymphocyte cultures. Lialiaris et al. (1988) showed chlorpromazine alone to have no significant effect but to enhance the effect of alkylating antineoplastics and caffeine. No doubt some of the variation in results is due to presumably genetic differences in drug metabolism in different individuals (Crossen and Morgan, 1982). The drug has also been reported to be teratogenic (Yu et al., 1988) and to reduce the mitotic index and cell number, to inhibit formation of blastocysts, and to delay implantation of embryos in the mouse (Kola and Folb, 1986). There is also evidence that chlorpromazine reduces UV-induced squamous cell carcinogenesis in hairless mice presumably by the induction of DNA damage (Peak et al., 1989) and has photoclastogenic but not photomutagenic properties in Chinese hamster ovary cells exposed to no-effect doses of UV light in the presence of the drug (Dean et al., 1991).

157 Materials and methods

Chemical compounds The compound chlorpromazine HC1 (CAS No. 69-09-0) was purchased from Fluka, Buchs, Switzerland. The antidepressants clomipramine HCI (17321-77-6) and maprotiline HCI (10347-816) were obtained from Ciba-Geigy, lofepramine HCI (26786-32-3) was obtained from Merck and mianserin HC1 (21535-47-7) from Beecham. These four compounds were gifts received from the subsidiaries of the respective companies in South Africa. The structural formulas of the five compounds are shown in Fig. 1. For the feeding of the larvae the compounds were dissolved either in distilled water or in a mixture of 5 % Tween-80 (Serva, Heidelberg, Germany) plus 5% ethanol (Merck, Darmstadt, Germany). Larval feeding Eggs were collected for 8 h in culture bottles containing a solid agar base (5% (w/v) agar agar in water) covered completely with an approx. 5 mm layer of live baker's yeast supplemented with sucrose. Three days later, the larvae were washed out of these bottles with tap water through a fine-meshed stainless steel strainer. They were then put into plastic vials containing medium with the test compounds. The solutions of the various

compounds were used to prepare Drosophila Instant Medium (Formula 4-24, Carolina Biological Supply Co., Burlington, NC, USA) by adding 5 ml solution to 1.5 g of dry Instant Medium. Negative water or solvent controls were included. The larvae were fed on this medium for the rest of their development (approx. 48 h until pupation).

Somatic mutation and recombination test The following cross of flies carrying markers on the left arm of chromosome 3 was used: fir3~ In(3LR)TM3, rip p sep bx 34e e s Ser females mated to mwh males. Detailed information on the genetic markers is given by Lindsley and Zimm (1985, 1990). This is the reciprocal cross of the standard cross previously used (Alonso Moraga and Graf, 1989; van Schaik and Graf, 1991). The f l r 3 / T M 3 females are much more fertile than mwh females; therefore it is easier to obtain ample numbers of larvae for treatment. The surviving flies were collected from the treatment vials, and flies of the trans-heterozygous (mwh fir +/mwh + fir 3) genotype were stored in 70% ethanol. The mounting of the wings and the classification of the spots were done as described previously (Graf et al., 1984, 1989; van Schaik and Graf, 1991). All experiments were conducted at 25°C and 60% relative humidity.

CH2CH2CH2N(CH3)2 Clomipramine

r.c I CH2CH2CH2~---CH2--C"~'~.,/ CH3

Lofepramine C~H3

CH2CH2CH2NHCH3 Maprotiline

CH2CH2CH2N(CH3)2 Mianserin

Chlorpromazine Fig. 1. Structural formulas of the five test compounds.

158

Data evaluation and statistical analysis The wing spot data were evaluated with the computer program SMART (Wiirgler, unpublished). The statistical analysis was performed in the same way as previously described (van Schaik and Graf, 1991). Details on the statistical proce-

dure are given in Frei and Wiirgler (1988). For the calculations, the Kastenbaum-Bowman test was used with P = 0.05. Based on the number of mwh clones, the number of wings analyzed, and the number of cells scored in each wing (ca. 24,400), the clone formation frequency per cell

TABLE 1 SUMMARY OF RESULTS OBTAINED IN THE DROSOPHILA WING SPOT TEST WITH FIVE ANTIDEPRESSANTS Compound

Number of wings

Conc. (mM)

Spots per wing (Number of spots) Stat. Diagn. a Small Large Twin Total

mwh

Mean clone size

clone

class

(24) (28)(44)+ (53)+ (23)+ (53)+

24 28 44 52 23 53

2.00 2.18 1.70 1.56 1.70 1.62

1.5 1.6 3.6 3.7 3.4 4.3

0.1 2.1 2.2 1.9 2.9

0.01 (2) 0.03 (1)i 0.01 (1)i 0.04 (5)i 0.02 (4)i 0.04 (3)i

0.44 (83) 0.42 (16) 0.41 (57)0.55 (68)0.50 (103)0.70 (57)+

83 16 57 68 103 57

1.64 1.38 1.68 2.01 1.72 1.81

1.8 1.7 1.7 2.2 2.1 2.9

- 0.1 -0.1 0.5 0.3 1.1

0.03 (3) 0.05 (2)i 0.10 (15)+ 0.05 (9)i 0.05 (5)i 0.18 (12)+

0.02 (2) 0.00 (0)i 0.04 (6)i 0.04(7)i 0.01 (1)i 0.06 (4)i

0.38 0.40 0.51 0.51 0.41 0.93

(43) (16)(77)(83)(38)(63)+

43 16 77 81 37 63

1.74 1.69 1.97 1.73 1.78 2.06

1.5 1.6 2.1 2.0 1.6 3.8

0.1 0.5 0.5 0.1 2.3

0.31 (37) 0.28 (11)0.34 (27)0.37 (46)0.43 (63)i 0.47 (27)i

0.07 0.10 0.01 0.01 0.04 0.02

(8) (4)(1) (1)(6)(1)-

0.03 (3) 0.00 (0)i 0.00 ( 0 ) 0.01 ( 1 ) 0.03 (4)i 0.00 ( 0 ) -

0.40 0.38 0.35 0.39 0.50 0.48

(48) (15)(28)(48)(73)(28)-

47 15 28 48 72 28

2.13 1.93 1.36 1.40 1.74 1.29

1.6 1.5 1.4 1.6 2.0 2.0

-0.1 - 0.2 0.0 0.4 0.4

0.25 (69) 0.23 (61)0.30(33)0.32 (36)0.23 ( 5 ) -

0.04 (10) 0.03 ( 8 ) 0.05 ( 6 ) 0.03 ( 3 ) 0.14 (3)i

0.02 (6) 0.02 ( 5 ) 0.01 ( 1 ) 0.02 ( 2 ) 0.05(1)i

0.31 0.28 0.36 0.36 0.41

(85) (74)(40)(41)(9)i

85 73 40 41 9

1.85 1.81 1.73 1.56 2.67

1.3 1.1 1.5 1.5 1.7

-0.1 0.2 0.2 0.4

spots

Spots with

single spots (1-2 cells) [m = 2.0]

single spots ( > 2 cells) [m= 5.0]

spots

[m = 5.0]

[m = 2.0]

0.30 (20) 0.26 (19)0.72(36)+ 0.83 (48)+ 0.71 (20)+ 0.94 (47)+

0.03 0.08 0.12 0.09 0.04 0.10

0.03 (2) 0.04(3)i 0.04 (2)i 0.00 ( 0 ) 0.07 (2)i 0.02 (1)i

0.36 0.38 0.88 0.91 0.82 1.06

0.05 (9) 0.00 ( 0 ) 0.04 ( 5 ) 0.09 (11)0.07 (14) 0.05 ( 4 ) -

Frequency of clone formation per 105 cells b observed

control corrected

Clomipramine (water) 0 1 10 25 50 100

66 74 50 58 28 50

(2) (6)i (6)i (5)i (1)i (5)i

Lofepramine (5% Tween + 5% ethanol) 0 1 10 25 50 100

190 38 138 124 204 82

0.38(72) 0.39 (15)0.37 (51)0.42 (52)0.42 (85)0.61 (50)+

Mianserin (5% Tween + 5% ethanol) 0 1 10 25 50 100

114 40 152 164 92 68

0.33 (38) 0.35 (14)0.37 (56)0.41(67)0.35 (32)0.69 (47)+

Maprotiline (5% Tween + 5% ethanol) 0 1 10 25 50 100

120 40 80 124 146 58

Chlorpromazine (water) 0 10 25 50 75

276 264 110 114 22

a Statistical diagnoses according to Frei and Wiirgler (1988) for comparisons with corresponding controls: +, positive; - , negative; i, inconclusive, m, multiplication factor. Kastenbaum-Bowman tests, one-sided. Probability levels: a = fl = 0.05. u Frequency of clone formation: mwh clones/wings/24,400 cells (without clone size correction).

[]

12

•~

I ~

159

CLOMIPRQMINE

CLOM

CONC. : 100 mM [ T R E Q T E D : SO w i n g s

LOFE

0.8

----~--

MIKN

--~--

MAPR

~

CHLO

0 HZ ~o -.¢

o

0.6

~ ffl 0.4

SINGLE SPOTS

o O)

0.2 0

TREATMENT: 48 h [-]CONTROL: 66 w i n g s

o

25

go

7'~

lOO

Concentration {mM} Fig. 2. Dose-response relationships for clomipramine (CLOM), lofepramine (LOFE), mianserin (MIAN), maprotiline (MAPR) and chlorpromazine (CHLO) for the total number of spots per wing.

1

2

3-q

5-8 9-16 -32 -6q -128 SPOT SIZE

-256 >512

TWIN SPOTS ¢n_o

cycle and 105 ceils was calculated. No clone size correction was made (for details see Frei et al., 1992a). Results and discussion

All five compounds were tested in two or three independent experiments with concurrent negative controls. The combined wing spot test data are given in Table 1. The dose-response relationships for the total number of spots per wing are plotted in Fig. 2. Spot size distributions for single and twin spots obtained with the highest concentration of each compound are shown in Figs. 3-7. For the evaluation of the effects, the results obtained in the various treated series were always compared to the corresponding concurrent controis. Two compounds were tested in water, the other three in a solvent mixture of 5% Tween-80 plus 5% ethanol. The frequencies of spontaneous spots obtained in the five control series are all similar, although there seems to be a tendency for slightly higher, but nonsignificant, values in the solvent series. Of the five compounds tested, clomipramine was clearly genotoxic in the concentration range between 10 and 100 mM (see Fig. 2). The highest frequency of the total of spots per wing recorded is 1.06 which is comparable to the frequency of 1.08 obtained with imipramine (van Schaik and Graf, 1991). This is also reflected in the similar frequencies of clone formation of 2.9 and 3.1 × 10 -5, respectively. The two compounds clomipra-

Z

1

2

3-q

__ f~ 5"-i_ 5-8 9-16 -32 -6q -128 -256 >512 SPOT SIZE

Fig. 3. Spot size distributions for single and twin spots obtained with 100 mM clomipramine.

mine and imipramine are structurally very closely related; clomipramine has an additional CI atom attached to the C7 atom (see Fig. 1). From this it

LOFEPRQMINE CONC. : 100 mH []TREATED: 82 wings

TREATMENT: 48 h []CONTROL:190 wings

¢;

SINGLE SPOTS

~,. 1

2 3-q 5-9 9-16 -32 -~ -128 -256 >512 SPOT SIZE TWIN SPOTS

1

2

3-Y 5-8 9-16 -32 -6q -128 -256 >512 SPOT SIZE

Fig. 4. Spot size distributions for single and twin spots obtained with 100 mM lofepramine.

160

MIANSERIN

i-! CONC.: 100 mM

TREATMENT: 48 h

~TREATED: 68 u l n g s

[-7CONTROL:114 u i n g s

0

CHLORPROMAZINE CONC.: SO mM []TREATED: 114 wings

TREATMENT: q8 h {-'l CONTROL: 276 wings

(.9

SINGLE

SPOTS

o~ O9,o

7~

1

2

3-q 5-8 9-16 -32 -64 -128 -256 >512

00 ,::; I

SPOT SIZE

2 .3-q

5-8 B-16 -32 -64 -128 -256 >512 SPOT SIZE

co~

z¢;

THIN SPOTS

~_o.

TNIN SPOTS

o0_o.

7) Z

Z

1

2

3-q 5-8 9-16 -32 -6q -128 -256 >512 SPOT SIZE

Fig. 5. Spot size distributions for single and twin spots obtained with 100 mM mianserin.

can be concluded that the additional C1 atom does not alter the genotoxicity of the tricyclic compound. Furthermore, the result obtained with

MAPROTI L I NE CONC. : i00 mM

TREATMENT: 48 h

~TREATED: 58 u i n g s

[-]CONTROL: 1 2 O u i n g s

(.o

z=o SINGLE

SPOTS

oo

1

2

3-q 5-8 9-16 -32 -64 -128 -256 >512 SPOT SIZE

~o TWIN SPOTS o0o O. 09 Z

I

2

341 5-8 9-16 -32 -64 -128 -256 >512 SPOT SIZE

Fig. 6. Spot size distributions for single and twin spots obtained with 100 mM maprotiline.

I

2

3-q

5-8 9-16 -32 -64 -128 -256 >512 SPOT SIZE

Fig. 7. Spot size distributions for single and twin spots obtained with 50 mM chlorpromazine.

clomipramine is also in agreement with the hypothesis that the presence of an N atom at position 5 in the central 7-membered ring is crucial for the genotoxic properties of the tricyclic antidepressants (van Schaik and Graf, 1991). Lofepramine is another compound which is similar to imipramine and clomipramine, the difference being the longer and more complex side chain attached to N5. This compound has the highest molecular weight of all the compounds tested and is not soluble in water. The addition of 5% Tween-80 plus 5% ethanol as solvents did not yield a clear solution at the highest concentration of 100 mM. Only at this concentration, which produced very high toxicity in the treated larvae (approx. 90% as estimated from the number of surviving flies compared to control flies), did we obtain a positive result for the frequency of the total of spots per wing. However, this frequency of 0.70 is slightly lower than the highest one obtained with clomipramine. This weaker effect is also reflected in a lower frequency of clone formation of 1.1 × 10 -5 compared to 2.9 × 10 -5. We conclude that this compound is weakly genotoxic at the highest testable concentration. Since

161

its major metabolite, desipramine, is mutagenic in this test, the weak response is most probably due to the solubility problem and the large size of the molecule. We propose that the genotoxic activity of this compound can be attributed to the presence of an N atom at position 5 in the central ring of the molecule. Mianserin is a piperazino-azepine tetracyclic with a structure different from those previously discussed. It has an additional ring attached to the central ring at C10 and C l l and has no side chain at position 5. Furthermore, the N atom is at position 10 and not at position 5. It is completely soluble in the solvent mixture at 100 mM and was clearly genotoxic at this concentration. The frequencies of both small single spots and large single spots were increased. The total spot frequency of 0.93 is comparable to that obtained with clomipramine. Therefore, this result seems to indicate that an N atom at position 10 in the central ring also renders a compound genotoxic. The next two compounds to be discussed, maprotiline and chlorpromazine, differ basically from the other tricyclic compounds analyzed so far by the presence of a 6-membered central ring instead of a 7-membered one. In addition, maprotiline has a transverse ethylene group bridging C5 and C10, whereas chlorpromazine has an N atom at position 5 and an S atom at position 10 as well as a CI attached to C7. As can be seen from the data in Table 1, both compounds were negative over the whole dose range tested. Chlorpromazine was clearly more toxic than the other tricyclic compounds. There were no surviving flies at 100 mM, and only a very few flies survived at 75 mM. The negative result obtained with these two compounds may be due to the presence of a 6-membered central ring which may render the molecule more stable and therefore less accessible for metabolic transformation. This would especially apply to maprotiline with its bicyclo central ring. Trush et al. (1979) found that, in general, tricyclics with a heterocyclic nitrogen at position 5 (e.g., imipramine and desipramine) generated chemiluminescence in human polymorphonuclear leukocytes, whereas those with a C at position 5 (e.g., amitriptyline) did not. This property has been attributed to the molecule's ability to inter-

TABLE 2 LINEAR INTERPOLATION OF THE DOSE-RESPONSE CURVES FOR 10 ANTIDEPRESSANTS (SPOTS PER WING PER mM UNIT OF EXPOSURE CONCENTRATION) Compound

b ( × 10 2)

a

Desipramine * Imipramine * Clomipramine Mianserin Lofepramine Chlorpromazine Maprotiline Nortriptyline * Protriptyline * Amitriptyline *

3.87 1.47 0.57 0.46 0.26 0.14 0.12 0.08 0.00 - 0.02

0.70 0.40 0.56 0.38 0.42 0.30 0.38 0.35 0.32 0,29

b, slope; a, Y-axis intercept. * Data from van Schaik and Graf (1991).

act with reactive oxygen species. Clomipramine, however, which has an N at position 5, failed to produce chemiluminescence, implying an inhibitory effect of the chlorine atom on this reaction. It would seem that this property is not involved with mutagenicity in Drosophila since clomipramine, like imipramine and desipramine, is mutagenic. To obtain a quantitative comparison of the genotoxic effects of the compounds presented here with those previously reported (van Schaik and Graf, 1991), linear interpolations of the dose-response relationships were calculated and plotted as given in Table 2 and Fig. 8. As we have already pointed out in the previous paper, the two positive compounds, desipramine and imipramine, show a drop in the dose-response curve at the highest concentration (100 mM). This is most probably due to the high toxicity of these treatments. The same phenomenon has also been observed with other compounds that are highly toxic (Graf et al., 1984). For this reason, the 100 mM data points were not included for the linear interpolation of the desipramine and imipramine curves. Desipramine is over twofold more genotoxic than imipramine which in turn is two- to threefold more active than clomipramine and mianserin. On the other hand, lofepramine is only half as genotoxic as mianserin. In general, it seems that the later generations of antidepres-

162

•

DESI

O

IMI

a

2.5 2

•

NOR

V

AMI

[]

PRO

1.5: O

1

a3 0.5

the nonmutagenic chlorpromazine has the same side chain as the mutagenic clomipramine and imipramine. It also seems safe to assume that a CI atom attached to C7 is unimportant for genotoxicity since this is present in mutagenic clomipramine and nonmutagenic chlorpromazine.

Acknowledgements ¸

ol0

v ....

f ....

25

i ....

50

i

, r ,

75

100

C o n c e n t r a t i o n (raM)

b

2.5

•

CLOM

o

LOFE

•

MIAN

v

MAPR

[3

CHLO

We thank Dr. H. Frei for valuable discussions and critical reading of the manuscript. Thanks are also due to Doris Singer, Eveline Holliger and Ottavina Lutz for skillful technical assistance. The work of N.v.S. in Schwerzenbach was supported by a grant from the Swiss Federal Institute of Technology, Zurich, Switzerland.

References

g 1.5ff?

0.5 e ~ 0

25

50

75

100

C o n c e n t r a t i o n (raM) Fig. 8. Linear interpolation of the dose-response curves for 10 antidepressants tested in the wing spot test. (a) Data from van Schaik and Graf (1991) for desipramine (DESI), imipramine (IMI), nortriptyline (NOR), amitriptyline (AMI) and protriptyline (PRO). (b) Data from Table 1; abbreviations as in Fig. 2.

sants are less genotoxic in Drosophila than the first generation drugs. The conclusion from the results presented here combined with those of our previous studies on tricyclics is that the mutagenicity of these substances in the wing SMART test seems to be dependent on the presence of a 7-membered ring containing an N atom. The aliphatic side chains attached to position 5 which are thought to be responsible for the anticholinergic effects of the tricyclics cannot be involved in the reactions leading to genotoxicity since the nonmutagenic maprotiline, like protriptyline, has the same side chain as the mutagenic desipramine. Likewise,

Alonso Moraga, A., and U. Graf (1989) Genotoxicity testing of antiparasitic nitrofurans in the Drosophila somatic mutation and recombination test, Mutagenesis, 4, 105-110. Balbi, A., G. Muscettola, N. Staiano, G. Martire and F. De Lorenzo (1980) Psychotropic drugs: Evaluation of mutagenic effect, Pharmacol. Res. Commun., 12, 423-431. Chaplin, S. (1986) Bone marrow depression due to mianserin, phenylbutazone, oxyphenbutazone, and chloramphenicol Part I, Adv. Drug React. Ac. Pois. Rev., 2, 97-136. Crossen, P.E., and W.F. Morgan (1982) The effect of chlorpromazine on SCE frequency in human chromosomes. An in vitro and in vivo study, Mutation Res., 96, 225-232. Dean, S.W., M. Lane, R.H. Dunmore, S.P. Ruddock, C.N. Martin, D.J. Kirkland and N. Loprieno (1991) Development of assays for the detection of photomutagenicity of chemicals during exposure to UV light. I. Assay development, Mutagenesis, 6, 335-341. De Mol, N.J., A.B.C. Becht, J. Koenen and G. Lodder (1986) Irreversible binding with biological macromolecules and effects in bacterial mutagenicity tests of the radical cation of promethazine and photoactivated promethazine. Comparison with chlorpromazine, Chem.-Biol. Interact., 57. 73-83. Frei, H., and F.E. Wiirgler (1988) Statistical methods to decide whether mutagenicity test data from Drosophila assays indicate a positive, negative or inconclusive result, Mutation Res., 203, 297-308. Frei, H., J. Clements, D. Howe and F.E. Wiirgler (1992a) The genotoxicity of the anti-cancer drug mitoxantrone in somatic and germ cells of Drosophila melanogaster, Mutation Res., 279, 21-33. Frei, H., J. Liithy, J. Brauchli, U. Zweifel, F.E. Wiirgler and C. Schlatter (1992b) Structure/activity relationships of the genotoxic potencies of sixteen pyrrolizidine alkaloids assayed for the induction of somatic mutation and mitotic

163 recombination in wing cells of Drosophila melanogaster, Chem. Biol. Interact., 83, 1-22. Graf, U., and N. van Schaik (1992) Improved high bioactivation cross for the wing somatic mutation and recombination test in Drosophila melanogaster, Mutation Res., 271, 59-67. Graf, U., F.E. Wiirgler, A.J. Katz, H. Frei, H. Juon, C.B. Hall and P.G. Kale (1984) Somatic mutation and recombination test in Drosophila melanogaster, Environ. Mutagen., 6, 153-188. Graf, U., H. Frei, A. K~igi,A.J. Katz and F.E. Wiirgler (1989) Thirty compounds tested in the Drosophila wing spot test, Mutation Res., 222, 359-372. Graf, U., D. Wild and F.E. Wiirgler (1992) Genotoxicity of 2-amino-3-methyl-imidazo[4,5-f]quinoline (IQ) and related compounds in Drosophila, Mutagenesis, 7, 145-149. Hess, R., R.M. Diener, H. Fritz, C. Pericin, R. Heywood, A.J. Newman and K.T. Rivett (1973) Tossicita ed effetti sulla riproduzione nell'animale del nuovo farmaco antidepressivo maprotilina (Ludiomil), Boll. Chim. Farm., 122, 782. Kola, I., and P.I. Folb (1986) Chlorpromazine inhibits the mitotic index, cell number, and formation of mouse blastocysts, and delays implantation of CBA mouse embryos, J. Reprod. Fertil., 76, 527-536. Lialiaris, T., D. Mourelatos and J. Dozi-Vassiliades (1988) Enhancement of cytogenetic damage by chlorpromazine in human lymphocytes treated with alkylating antineoplastics and caffeine, Mutation Res., 206, 361-365. Lindsley, D.L., and G. Zimm (1985) The genome of Drosophila melanogaster, Part 1: genes A-K, Drosophila Inf. Serv., 63, 227 pp. Lindsley, D.L., and G. Zimm (1990) The genome of Drosophila melanogaster, Part 4: genes L-Z, balancers, transposable elements, Drosophila Inf. Serv., 68, 382 pp.

Negishi, T., T. Shiotani, K. Fujikawa and H. Hayatsu (1991) The genotoxicities of N-nitrosamines in Drosophila melanogaster in vivo: the correlation of mutagenicity in the wing spot test with the DNA damages detected by the DNA-repair test, Mutation Res., 252, 119-128. Obaseiki-Ebor, E.E., and J.O. Akerele (1988) The mutagenic activity of chlorpromazine, Mutation Res., 208, 33-38. Peak, M.J., M. Pfaff and C. Peraino (1989) Chlorpromazine reduces UV-induced squamous cell carcinogenesis in hairless mice and enhances UV-induced DNA damage in cultured cells, Br. J. Cancer, 60, 220-222. Sauter, C., and H. Ernst (1991) Cytostatic activity of tricyclic antidepressants as measured by a tetrazolium-based colorimetric assay, Cancer J., 4, 332-334. Speit, G. (1982) Intercalating substances do not induce sisterchromatid exchanges (SCEs) in vivo, Mutation Res., 104, 261-266. Trush, M.A., M.J. Reasor, M.E. Wilson and K. van Dyke (1979) Comparison of the interaction of tricyclic antidepressants with human polymorphonuclear leukocytes as monitored by the generation of chemiluminescence, Chem.-Biol. Interact., 28, 71-81. Van Schaik, N., and U. Graf (1991) Genotoxicity evaluation of five tricyclic antidepressants in the wing somatic mutation and recombination test in Drosophila melanogaster, Mutation Res., 260, 99-104. Yu, J.-F., Y.-S. Yang, W.-Y. Wang, G.-X. Xiong and M.-S. Chen (1988) Mutagenicity and teratogenicity of chlorpromazine and scopolamine, Chinese Med. J., 101, 339-345. Ziebarth, D. (1982) Vergleichende Untersuchungen des Nitrosierungsverhaltens von Arzneimitteln bei konstantem pH-Wert und unter simulierten Magenbedingungen, Arch. Geschwulstforsch., 52, 429-442.