Mutation Research, 269 (1992) 301-306 30 i © 1992 Elsevier Science Publishers B.V. All rights reserved 0027-5107/92/$05.00

MUT 05158

A contribution to the study of the structure-mutagenicity relationship for a-dicarbonyl compounds using the Ames test

L. Dorado , M. Ruiz Montoya and J.M. Rodr igucz Mcl l ado Departamento de Qu[mica Ffxica y Termodindmica Aplicada, Facultad de Ciencias, Unicersidad de Cdrdoba, 14004 Cordoba, Spain

(Received 16 January 1992) (Revision received 14 April 1992)

(Accepted 16 April 1992)

Keywords: Glyoxal derivatives; a-Dicarbonyl compounds; Ames test

Summary

The mutagenicity of a series of nine a-dicarbonyl compounds against S. typhimurium strain TA100 was studied using the Ames test (standard plate incorporation assay) without preincubation. Acetylben- zoyl, sodium glyoxylate and camphorquinone were not mutagenic. The following sequence of activities (in revertants per/zmole of free dicarbonyi added) was obtained: glyoxai > methyiglyoxal > phenylglyoxal :~ 1,2-cyclohexanedione ~ diacetyl > 3,4-hexanedione.

These compounds can be grouped in three series: aldehydes, ketones and enolizable ketones (1,2-cyclohexanedione). in each of the two first groups the mutagenic activity decreases when the size of the substituent increases. No relation was found between the mutagenicity and the molecular eJectronic and/or resonance parameters. The low or non-existent activity of some of the chemicals studied is discussed. A relation between the mutagenic activities and the polarographic reduction potentials and, consequently, the structures of the mutagens was found.

In 1979, Bjeldanes and Chew examined the mutagenic activity of several a-dicarbonyl com- pounds with the Ames test using Salmonella ty- phimurium strains TA98 and TA100. These au- thors found that glyoxai, diacetyl and 1,2- cyclohexanedione show dose-related mutagenic activity against TA100 but not against TA98, with glyoxal having the greatest activity. The muta- genie activity of glyoxal and derivatives has been

Correspondence: Dr. J.M. Rodr[guez Mellado, Departamento de Qufmica Ffsica y Termodinfimica Aplicada, Facultad de Ciencias, Universidad de C6rdoba, 14004 Cordoba, Spain.

related to the formation of oxygen radicals (Yamaguchi and Nakagawa, 1983).

The Ames test has been used for comparison of the results obtained with strains TAI00 and TA104 (Marnett et al., 1985). The sensitivity of the latter to the mutagenicity induced by a-di- carbonyl compounds is greater than that of the former, though the sequence of activities is the same in both cases.

The mutagenic activity of a-dicarbonyl com- pounds (mainly methylglyoxal) has also been de- tected using other mutagenic assays (Cajelli et al., 1987; Ariza eta!., !988).

The molecular mechanism of the mutagenic action of these compounds has been related to its

302

ability to form adducts with the guanine residues of nucleic acids. Since the early study of Staehe- lin (1959) it has been proposed that the inactiva- tion of the tobacco mosaic virus (and of its nu- cleic acid) is caused by the formation of cyclic adducts with guanine residuals.

The reaction product between glyoxal and guanosine has been isolated, and the stochiome- try of the adduct is 1 : 1 (Shapiro and Hachmann, 1966). Proton magnetic resonance measurements showed the structure of the adduct to be as follows:

O OH H ~ ~,N~

HOHa , ~ HO OH

In the same way, phenylglyoxal and di-3-pyri- dylglyoxal give adducts with guanosine but the products were not isolated. The adducts between guanine and glyoxal, methylglyoxal and ketoxal (/3-ethoxy-a-keto butyraldehyde) were also pre- pared and structures similar to that given above were found (Shapiro et al., 1969). Moreover, the reactions between glyoxal and nucleic acids, nu- cleotides and their bases were examined by UV- visible spectroscopy measurements (Nakaya et al., 1968). The spectral change due to the reaction of glyoxal with thermally denatured DNA was as- signed to its specific reaction with guanine residues.

As far as we know, there are no systematic studies dealing with the origin of the differences in mutagenic activity between the a-dicarbonyl compounds. Thus, the aim of this work is to contribute to the study of the mutagenic activity of a comprehensive series of a-dicarbonyl com- pounds in relation to their molecular structures.

tABLE 1

STRUCTURES AND CAS NUMBERS OF THE CHEMICALS STUDIED

Compound Structure 0 0 II II

Glyoxal H - C- C- H 0 0 Li , Methylslyoxal CH ~-° C - C - H

O O Ph~nylglyoxal ~ (,.,),/~" C - C - H

" 1 . . . . . . . #

Di~cetyl C H 3 - C - C - C H ~ O 0 II II

3,4-Hexanedione CH 3CH 2 - C - C - CH 2CH 3 O u

,2-Cyclohexanedione L ~ 0 I V

O O Acetylbenzoyl (~(,.,)/~- C - C - CH

3

O O 11 II

Sodium glyoxylate Na + O- - C - C - H

CH3 ,~CH3 I~CH~ Camphorquinone ~ ~ 0 0 0

CAS No.

1(17-22-2

78-98-8

1074-12-0

431-03-8

4437-51-8

765-87-7

579-07-7

2706-75-4

465-29-2

Materials and methods

All the chemicals were of analytical grade and dissolved in sterilized doubly distilled water or dimethyl sulfoxide (DMSO). Purity was checked by means of polarographic measurements. Phen- yiglyoxal, 3,4-hexanedione, diacetyl, acetylbenzoyl and camphorquinone were from Aldrich; 1,2-cy- clohexanedione and methylglyoxal were from Sigma; sodium glyoxylate and glyoxai (30% in water) were from Fluka and Merck, respectively. Chemicals were used without further purification with the exception of 1,2-cyclohexanedione which was recrystallized twice and stored in a vacuum desiccator. Table 1 lists the structures and CAS numbers of all chemicals studied.

Salmonella typhimurium tester strain TA100 (hisG46, AuvrB, pKM101) was received from the Genetics Department of Cordoba University. Cells grown at 37°C in. Luria-Bertani medium (Davis et al., 1980) for 14-16 h without shaking were used in the mutagenicity test. The strain was

rov~tate t f 2SO0. e

2000,

lSOO] I "

1000

500,

m

I

v ~ Vp 0 b S 1"0 1'5 ~ o tat,

Fig. 1, Mutagenicity of glyoxal (o), phenylglyoxal (e) and methylglyoxal (O) against S. typhimurium strain TAI00. Each point is the average of at least four runs and the number of spontaneous revertants (around 85) was subtracted from each

value.

303

tested for appropriate phenotypic markers ac- cording to recommendations of McCann and Ames (1976). All culturing and storage methods were as recommended (Maron and Ames, 1983). All drugs were tested in the standard plate incor- poration assay without preincubation. The com- pounds were administered in at least 8 doses to the limit of solubility or toxicity. No $9 mix was added in any case.

Results and discussion

All the compounds studied showed low or very low activity; as expected from the previous report (Bjeldanes and Chew, 1979). Acetylbenzoyl was toxic at doses above ca. 1 /zmole/plate. Below this dose no mutagenic activity was detected. Sodium glyoxylate and camphorquinone did not present clear mutagenicity below their solubility limits (50 and 25 ~tmole/plate, respectively) though a slight increasing trend in the number of revertants was observed in both cases. This fact is in accord with the observations of Marnett et al. (1985) that glyoxylate is active against TA104, with a low activity.

In Fig. 1 are shown the results obtained for the aldehydes studied. As can be seen, they are clearly mutagenic, but in the cases of methyl- and phenyiglyoxal toxicity was achieved at doses of 3-4 ~mole/plate. In Fig. 2 are shown the results obtained for the ketones. The higher dispersions in the data are due to the lower activities, re- flected in the low values of revertants obtained; thus, small oscillations of those values preduce appreciable relative errors As can be seen for diacetyl and 1,2.cyclohexanedione, the number of revertants obtained at the higher dose duplicates that of spontaneous revertants. This is not the case with 3,4-hexanedione, which shows the low- est activity of the tested compounds (with the obvious exception of those clearly inactive). Nev- ertheless, the results of six independent measure- ments at 10 doses shown in Fig. 2 indicate that the trend in the number of revertants is clearly increasing with the dose.

In Table 2 are gathered the activities in rever- tants per ~mole of mutagen added. "['hough the absolute values are not the same, the relative order of mutagenic activities of glyoxai, methyl-

3O4

TABLE 2

MUTAGENIC ACTIVITIES IN THE AMES TEST AGAINST TA100 FOR THE CHEMICALS STUDIED

Compound A (rev//tmole) A* (rev//tmole)

Glyoxal 51 3.7 x 106 Methylglyoxal 921 1.0 x 106 Phenyiglyoxal 278 1.1 × 10 s Diacetyl 7.7 22.8 3,4-Hexanedione 0.9 2.8 1,2-Cyclohexanedione 12 1450

Activities denoted by A were obtained from the slopes of the graphs given in Figs. 1 and 2. Activities denoted by A* were obtained taking into account the hydration constant of each compound.

glyoxal and diacetyl is the same as that reported in the literature (Kim et al., 1987).

As stated in the introduction, the mutagenic activity of these compounds must be related to their ability to form adducts with guanine residues (Staehelin, 1959; Shapiro and Hachmann, 1966; Shapiro et al., 1969) in addition to other factors such as their uptake by the cell, their stability in solution and their reactivity with other molecules. The adducts are formed between the free a-di-

carbonyl compound and the amine groups of gua- nine. Nevertheless, these compounds are gener- ally hydrated in solution, their degree of hydra- tion depending on the nature of the groups at- tached to the dicarbonylic arrangement (Gutsche, 1967). Since the mutagenic species is the non-hy- drated one, it seems reasonable to assume that the mutagenic activity must be calculated as a function of its actual concentration. It is easy to show that this can be accomplished by multiply- ing the experimental activities (A) by the factor 1 + K H, where K H is the equilibrium constant of the hydration reaction defined as:

[Hydrate] KH ffi [Free carbonyl] "

K . values were obtained from the literature (diacetyl: Greenzaid et al., 1967; glyoxylate: Sere- nsen et al., 1974; glyoxah Wasa and Musha, 1970) or calculated from high-speed linear-sweep voltammetric measurements (Ruiz Montoya and Roddguez Me!!ado, unpublished results). In Table 2 are given the values of the corrected

revlptote 100- J / s// 50

25

0 0 1() 2'0 3() 4'0 pmol/plate

Fig. 2. Mutagenicity of 1,2-cyclohexanedione (o), diacetyl (e) and 3,4-hexanedione (123) against S. typhimurium strain TA100. Each point is the average of at least four runs and the number of spontaneous revertants (around 85) was subtracted from each value.

For 3,4-hexanedione the standard deviations have been plotted with each point.

305

activities A*. From these data it can be con- cluded that the mutagenic activities of aldehydes are much greater (by several orders of magni- tude) than those of the ketones. This fact is obviously related to the higher activity associated with the aldehydes. In these case of the ketones, 1,2-cyclohexanedione is much more mutagenic than the rest. This can be attributed to the facility of enolization of this compound in aqueous solu- tion (Bakule and Long, 1963). The enhanced mu- tagenic activity can be due either to the higher reactivity compared to the rest of ketones or to the appearance of an additional mutagenic species (the enol) which can present higher activity. In the case of camphorquinone enolization is not possible (Leonard and Mader, 1950), this fact explaining its low or absent mutagenic activity.

In each group there are differences in activity. For the aldehydes and ketones (1,2-cyclohexane- dione excluded) the order is: glyoxal > methylgly- oxal > phenylglyoxal; diacetyl > 3,4-hexanedione (> camphorquinone).

It can be seen ~that the relative activities of aldehydes are changed compared to those based on A. This reordering of activities seems reason- able because in this case, in both groups (al- dehydes and ketones) the activity decreases when the size of the substituent increases, while this is not true when the order is based on A.

On the other hand, the very different re- sponses of glyoxal and sodium glyoxylate could be due to the fact that the latter cannot form cyclic (and consequently stable) adducts with guanine.

An attempt has been made to correlate the mutagenic activities of the acyclic compounds with the electronic and resonance parameters of the substituents obtained from the literature (Exner, 1988; Hansch et al., 1991), but no good correla- tion was obtained in any case. The steric factor appears to be the more important. In fact, the inactivity of camphorquinone can be attributed to the high steric hindrance caused by the bicyclic structure. On the other hand, it could be assumed that the ratio of activities between methyl- and phenylglyoxal must be the same as that between diacetyl and acetylbenzoyl (because the sub- stituents are the ~ame on both cases); taking into account the activities given in Table 2, the activity of acetylbenzoyl must he around 2 rev//zmole,

which at doses below the toxic dose (1 /zmole/ plate) is within the experimental error.

The differences in activity between the differ- ent groups of compounds must be related to differences, at a molecular level, between the a-dicarbonyl compounds. As a matter of fact, the a-dicarbonyl arrangement is a conjugated system which is modified by the substituents. This modi- fication implies differences of energy of the HOMO and LEMO molecular orbitals, which are responsible for the reactivity.

The energies of the LEMO are correlated with the polarographic half-wave potentials (Pullman and Pullman, 1963). Thus, we can use this vari- able as a structure-related parameter. The reduc- tion potentials were obtained by polarographic measurements and agree with those previously reported by our laboratory (Avila et al., 1982; Rodrlguez Mellado et al., 1984a,b, 1985; Rodrfguez Mellado and Ruiz, 1986; Ruiz Mon- toya and ~o~rfguez Mellado, 1991) with the ex- ception of that of camphorquinone, which was measured fol this work, yielding a value of -0.74 V. The reduction potentials of aldehydes are around -0.42 V, tl~ose of the ketones around -0 .7 V, that of 1,2.cyclohexanedione is -0.55 V and that of glyoxylate ion is - 1.18 V. As can be seen, the mutagenic activities seem to be related to these potentials: the lower the potential (in absolute value) the higher the activity. Moreover, the reduction potentials of both camphorquinone and glyoxylate ion (specially this latter com- pound) are very negative and, consequently, the activities of these compounds must be very low, as is found experimentally.

Acknowledgements

The authors wish to gratefully acknowledge the gift of the S. typhimurium strain from the Genetics Department of C6rdoba University. The work has been supported by Junta de Andalucla and CICYT (Research Project PB88-0284).

References

Ariza, R.R., G. Dorado, M. Barbancho and C. Pueyo (1988) Study of the causes of direct-acting mutagenicity in coffee and tea using the Ara test in Salmonella typhimurium, Mutation Res., 201, 89-96.

306

Avila, J.L., M. Blazquez and J.J. Ruiz (1982) Reduction of dicarbonyl compounds on the DME. Part I. Reduction mechanism of diacetyl, Electrochim. Acta, 27, 1369-1372.

Bakule, R., and F.A. Long (1963) Keto-enol transformation o[ 1,2-cyclohexanedione, J. Am. Chem. S~., 85, 2309-2313.

Bjeldanes, L.F., and H. Chew (1979) l~,lutagenicity of 1,2.di- carbonyl compounds: maltol, kojic acid, diacetyl and re- lated substances, Mutation Res., 67, 367-371.

Cajelli, E., R. Canonero, A. Martelli and G. Brambilla (1987) Methylglyoxal-induced mutation to 6-thioguanine resis- tance in V79 cells, Mutation Res., 190, 47-50.

Davis, R.W., D. Botstein and J.R. Roth (1980) A Manual for Genetic Engineering. Advanced Bacterial Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 251 PP.

Exner, O. (1988) Correlation Analysis of Chemical Data, Plenum, New York, Chapter 1.

Greenzaid, P., Z. Rappoport and D. Samuel (1967) Limita- tions of ultra-violet spectroscopy for the study of the reversible hydration of carbonyl compounds, Trans. Fara- day Soc., 63, 2131-2139.

Gutsche, C.D. (1967) The Chemistry of Carhonyl Compounds, Prentice-Hall, Englewood Cliffs, N J, p. 35.

Hansch, C., A. Leo and R.W. Taft (1991) A survey of Ham- mett substituent constants and resonance and field param- eters, Chem. Rev., 91, 165-195.

Kim, S.B., F, Hayase and H. Kato (1987) Desmutagenic effect of wdicarbonyi and a-hydroxycarbonyl compounds against mutagenic heterocyclic amines, Mutation Res,, 177, 9-15.

Leonard, N.J., and P.M. Mader (1950) The influence of steric configuration on the ultraviolet absorption of 1,2-diketo- nes, J, Am, Chem, Soc,, 85, 5388-5397,

Marnett, L.J., H.K. Hurd, M,C. Hollstein, D.E. Levin, H. EsterbLuer and B.N, Ames (1985) Naturally occurring carbonyl compounds are mutagens in Salmonella tester strain TAI04, Mutation Res., 148, 25-34.

Maron, D.M,, and B,N, Ames (1983) Revised methods for the Salmonella mutagcnicity test, Mutation Res., 113, 173-215.

McCann, J., and B.N. Ames (1976) Detection of carcinogens as mutagens in the Salmonella/microsome test: assay of 300 chemicals, Proc. Natl. Acad. Sci. (USA), 73, 950-954.

Nakaya, K., O. Takenaka, H. Horinishi and K. Shibata (1968) Reactions of glyoxal with nucleic acids, nucleotides and

their component bases, Biochim. Biophys. Acta, 161, 23- 31.

Pullman, B., and A. Pullman (1963) Quantum Biochemistry, Interscience/Wiley, New York.

Rodr,.'guez Mellado, .I.M., and J.J. Ruiz (1986) Reduction of dicarbonyl compounds on the DME. Part VII. Electro- chemical behaviour of glyoxal, J. Electroanal. Chem., 199, 177-185.

Rodr[guez Mellado, J.M., L. Camacho and J.J. Ruiz (1984a) Reduction of dicarhonyl compounds on the DME. Part III. Electrochemical behaviour of phenylglyoxal: first wave, J. Electroanai. Chem., 177, 39-50.

Rodriguez Mellado, J.M., L. Camacho, R. Rodr[gucz Amaro and J.J. Ruiz (1984b) Reduction of dicarbonyl compounds on the DME. Part V. Experimental verification of the theoretical equations for CEC mechanisms, J. Electroanal. Chem., 177, 69-75.

Rodr[guez Mellado, J.M., J.L. Avila and J.J. Ruiz (1985) Reduction of dicarbonyl components on the DME. Part It. Reduction mechanism of acetylbenzoyl, Can. J. Chem., 63, 891-895.

Ruiz Montoya, M., and J.M. Rodr[guez Mellado (1991) Effect of proton-donors on the electroreduction of 3,4-hexane- dione, Portug. Electrochim. Acta, 9, 211-214.

Shapiro, R., and J. Hachmann (1966) The reaction of guanine derivatives with 1,2-dicarbonyl compounds, Biochemistry, 5, 2799-2807.

Shapiro, R., B.I. Cohen, S.-J. Shiuey and H. Mauer (1969) On the reaction of guanine with glyoxal, pyruvaldehyde, and kethoxal, and the structure of the acylguanines. A new synthesis of N2-alk~lguanines, Biochemistry, 8, 238-245.

S~rensen, P.F.., K. Bruhn and F. Lindl~v (1974) Kinetics and equilibria for the reversible hydration of the aldehyde group in glyoxylic acid, Acta Chem. Scand., 28, 162-168.

Staehelin, M, (1959) Inactivation of virus nucleic acid with glyoxal derivatives, Biochim. Biophys. Acta, 31, 448-454.

Wasa, T., and S. Musha (1970) Polarographic behavior of glyoxal and its related compounds. Bull. Univ. Osaka Prefect. Ser. A, 19, 169-180.

Yamaguchi, T., and K. Nakagawa (1983) Mutagenicity of and formation of oxygen radicals by trioses and glyoxal deriva- tives, Agric. Biol. Chem,, 47, 2461-2465.