[CANCER RESEARCH 41, 1469-1482, April 1981) 0008-5472/81 /0041-OOOOS02.00 DMA-damaging Activity in Vivo and Bacterial Mutagenicity of Sixteen Hydrazine Derivatives as Related Quantitatively to their Carcinogenicity1
Silvio Parodi, Silvio De Flora, Marco Cavanna, Albiana Pino, Luigi Robbiano, Carlo Bennicelli, and Giovanni Brambilla2
Departments of Pharmacology [M. C.. A. P., L. P., G. B.¡,Hygiene [S. D. F., C. B.]. and Oncology [S. P.¡,Genoa University. 16132 Genoa, Italy
ABSTRACT enees existing in the carcinogenic potency of different chemi cals, qualitative statements of positivity or negativity in a short- Sixteen hydrazine derivatives (hydrazine, 1,1-dimethylhydra- term test are of very limited value for an assessment of the zine, 1,2-dimethylhydrazine, phenylhydrazine, procarbazine, carcinogenic risk. isoniazid, isocarboxazid, nialamide, 2,4-dinitrophenylhydra- In this work, we attempted a quantitative comparison of the zine, phenelzine, hydralazine, dihydralazine, carbamylhydra- potencies of 16 hydrazine derivatives in: (a) inducing DNA zine, mebanazine, iproniazid, and 1-carbamyl-2-phenylhydra- fragmentation in liver and lung of male Swiss mice, as evaluated zine) were tested for DNA-damaging activity by the alkaline by alkaline elution technique; (b) inducing mutations in the elution technique and for mutagenic activity in the Salmonella- Sa/mone/te-microsome test, both in the presence and in the microsome (Ames) test. The first nine compounds listed (56%) absence of S-9 mix; and (c) inducing pulmonary tumors in were found to induce a significant DNA fragmentation in the chronically treated mice. The significance of a quantitative liver and/or in the lung of i.p.-treated male Swiss mice. The comparison is probably strongly dependent on the internal DNA-damaging potency varied over an ~30-fold range. Thir homogeneity of the sets of experiments that are compared. In teen of the first 14 compounds listed (81% of the total), this respect, the hydrazine derivatives appeared a rather con isocarboxazid being inactive, were positive in the Ames test, venient family of chemical compounds because their carcino with a broad range of activity towards the five bacterial strains genicity was studied mainly by a single group of investigators of Salmonella typhimurium used (TA1535, TA100, TA1537, according to a fixed protocol (42). Moreover, the carcinogenic TA1538, and TA98) and of metabolic behavior in the presence potencies of the hydrazine derivatives tested vary over a 1900- of S-9 mix containing rat liver, mouse liver, or mouse lung fold range, which seems to be a sufficiently large interval to postmitochondrial preparations from Aroclor-treated animals. allow for quantitative comparisons. The internal homogeneity The mutagenic potency varied over an almost 7000-fold range. of DNA damage and of mutagenicity data should be assured For 11 of the 16 hydrazine derivatives tested, homogeneous by the fact they were both directly obtained by our group in carcinogenicity data (induction of pulmonary tumors in mice experiments performed always according to the same experi chronically treated p.o.) were available from literature. Elabo mental protocol. Moreover, the 2 types of short-term tests allow ration of these data showed that carcinogenic potency varied for an additional interesting comparison: differences related to over an ~ 1900-fold range. The five most potent carcinogens in vitro versus in vivo metabolic activation. were all positive in the DNA damage test. Their carcinogenic potency varied over a 130-fold range and their DNA-damaging potency varied over a 22-fold range. DNA-damaging potency MATERIALS AND METHODS seemed to vary on a more compressed scale, but regression Chemicals analysis indicated the existence of a strong positive correlation between in vivo DNA-damaging and carcinogenic potencies, The name, structural formula, molecular weight, LD503 for while a lack of correlation was found between mutagenic and single i.p. administration in mice, source, and chemical purity carcinogenic potencies. There was no correlation between of the 16 hydrazine derivatives investigated are reported in DNA-damaging and mutagenic potencies. Table 1. Tetraethylammonium hydroxide was purchased from E. Merck, Darmstadt, West Germany; 3,5-diaminobenzoic acid INTRODUCTION dihydrochloride was obtained from Fluka AG, Buchs, Switzer land; dimethyl sulfoxide was purchased from BDH Chemicals, It has been repeatedly suggested that both induction of DNA Ltd., Poole, England. All other chemicals were reagent grade. damage in mammalian cells, measured as single-strand breaks (4, 10, 27, 30, 31) and mutagenicity in the Salmonella-micro- DNA Damage in 1/jVo-Alkaline Elution Assay some test (23, 24, 32, 33, 38) are, at least qualitatively, The in vivo assay of DNA damage was performed in non- correlated with carcinogenic activity. In recent years, however, it has become evident that, because of the tremendous differ- inbred male Swiss albino mice (aged 2 to 3 months) according to the following standardized protocol. Each hydrazine deriva 1 This work was supported by grants from the Consiglio Nazionale delle tive was tested at the following dosage schedules: (a) single Richerche (Progetto Finalizzato "Controllo della Crescita, Neoplastica"; Con i.p. administration of LD60 twice, reduced to LD60 (or one-half tracts 78.02856.96, 78.02803.96, and 79.00621.96). 2 To whom requests for reprints should be addressed, at Istituto di Farmacol of the LD50) if necessary because of lethal effect within 6 hr (or ogia dell'Università , Viale Benedetto XV, 2, 1-16132 Genoa, Italy. Received March 27. 1980; accepted October 16, 1980. 3 The abbreviation used is LDso, median lethal dose.
APRIL 1981 1469
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. 1Hydrazines testedStructural
NameHydrazine M.W.H2N-NH2-H20formula purity98% hydrate 50.1 AG, Buchs, Switzer pure land 1,1-Dimethylhydrazine1 60.1CH3-NH-NH-CH3CH3x 12595' Fluka AG, Buchs, Switzer 95-98%pure98% landE.
,2-Dimethylhydrazine-2HCI -2HCI 133.0 Merck, Darmstadt, West pure Germany Carbamylhydrazine H2N-NH-CO-NH2 75.1 123170'SourceFlukaICN-K & K Laboratories, 95-99% pure Plainview, N. Y. PhenylhydrazineTable /=\ 108.1LO»"(mg/kg)156'Carlo Erba, Milan. ItalyChemical Reagent grade
1-Carbamyl-2-phenylhydrazine 151.2 198 ' Carlo Erba, Milan, Italy Reagent grade ¿ ^-NH-NH —CO-NH2
2,4-Oinitrophenylhydrazine 198.1 450' J. T. Baker Chemicals N. V., Reagent grade Deventer, Holland
o-Phenylethylhydrazine oxalate 226.2 138' Imperial Chemical Indus 98% pure (mebanazine) CH-CH3 OXALATE tries, Cheshire, Great Britain
/3-Phenylethylhydrazine •H2SO4 234.3 168 W. R. Warner & Co., Pharmaceutical grade (phenelzine) //-2- CH-NH-NH2 •H2S Eastleigh, Hampshire, Great Britain
W-lsopropyl-a-<2- CO-NH-CH(CH3)2 257.8 800' Hoffmann-La Roche & Co., 98.5% pure methylhydrazino)-p-toluamide • Basel, Switzerland HCI (procarbazine) •HCI
CH2- NH-NH—CH3
Isonicotinic acid hydrazide (isonia- 137.1 132 Vecchi & C. Piam. Genoa, 98% pure zid) Italy
0CO-NH—NH
Isonicotinic acid 2-isopropylhydra- 179.2 640 Hoffmann-La Roche & Co., 99% pure zide (iproniazid) Basel, Switzerland
0CO-NH-NH-CHCCH3)2
Isonicotinic acid 2- 298.4 435 Pfizer Italiana, Rome, Italy Pharmaceutical grade [2(benzylcarbamoyl)ethyl]hydra- zide (nialamide)
CO—NH-NH - (CH)- CO—NH-CH
5-Methyl-3-isoxazolecarboxylic 231.2 49 Hoffmann-La Roche & Co., 98.5% pure acid 2-benzylhydrazide (¡socar- Basel, Switzerland boxazid) 'CO—NH-NH —
1-Hydrazinophthalazine-HCI (hy- NH - NH, 196.7 83 Ciba-Geigy, Basel, Switzer- Pharmaceutical grade dralazine) land
1,4-Dihydrazinophthalazine •H2SO4 288.3 159 Ciba-Geigy, Basel, Switzer- Pharmaceutical grade (dihydralazine) land
NH - NH,
3 LDso after single i.p. administration in mice. LDso's followed by asterisk were determined as described in "Materials and Methods" in male Swiss mice. All other LDso's were obtained from Ref. 34.
1470
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA Damage, Mutagenicity, and Carcinogenicity of Hydrazines
24 hr) from treatment; and (b) daily i.p. administration for 5 where V is the eluted volume in ml. As a first approximation, K (and sometimes 15) successive days of one-third of the LD50, is directly proportional to the number of single-strand breaks reduced to one-sixth of the LD50 or one-twelfth of the LD50 if (19). Controls were grouped into subsets of 20 experiments necessary because of lethal effects. each, which were carried out in a time interval of 3 to 4 weeks, The LDso's in mice for a single i.p. injection were as reported using the same batch of solutions for alkaline elution. The in the literature (34). When the LD50 was unknown, it was elution procedure remained unvaried for the whole period in determined at 10 days by administering i.p. to groups of 5 male which all the experiments were carried out; notwithstanding, mice for each dose a series of doses increasing serially by a some variability was observed among the various subsets of factor of 2; the LD50 was calculated (22) only after the doses controls. Consequently, the increase in DNA elution rate in given were sufficient to obtain both 0% and 100% mortality. duced by a given treatment and its statistical significance was Immediately before use, compounds soluble in water were calculated versus the subset of controls carried out approxi dissolved in 0.9% NaCI solution with small amounts of acid or mately in the same time interval. Because the distribution of alkali, if needed, to bring pH to ~7; compounds insoluble in the fractions of eluted DNA was not normal, the nonparametric water were suspended in 0.9% NaCI solution with 1% carbox- Wilcoxon 2-sample test was used for this statistical analysis ymethylcellulose as suspending agent. All compounds were (36). administered i.p. in 0.01 ml of vehicle per g of body weight. Tissue samples were taken for histológica! examination from Controls were given i.p. injections of the same amount of the at least 3 animals of each treatment schedule for evaluation of same vehicle. Mice receiving a single dose were killed 1 and 6 whether or not cells (or nuclei) applied to the filters for elution hr after administration if treated with hydrazines soluble in were prepared from tissues undergoing degeneration or necro water and 4 and 24 hr after administration if treated with sis. hydrazines insoluble in water. Mice treated with multiple suc cessive daily doses were always killed 6 hr after last adminis Evaluation of Mutagenic Activity tration. DMA damage was evaluated by the technique of alkaline For mutagenesis assays, the compounds were dissolved and elution (19) as reported previously (5, 27). In brief, liver and diluted either in distilled water (isoniazid, phenelzine, procar- lungs were dissected, washed, and minced in cold Merchant's bazine, phenylhydrazine, hydrazine, carbamylhydrazine, 1,2- solution (0.14 M NaCI, 2.7 mw KCI, 8.1 mM Na2HPO4, 1.47mM dimethylhydrazine, and 1,1-dimethylhydrazine) or in dimethyl KHjPO4, and 0.53 mw disodium EDTA, pH 7.4). To obtain a sulfoxide (nialamide, isocarboxazid, iproniazid, hydralazine, suspension of lung cells, small fragments of this tissue were dihydralazine, 1-carbamyl-2-phenylhydrazine, 2,4-dinitro- forced through a stainless steel screen (60 mesh) in the pres phenylhydrazine, and mebanazine) and tested at subtoxic con ence of Merchant's solution. Liver fragments were incubated centrations, as reported under "Results." for an additional 5 min in cold Merchant's solution; nuclear Bacterial Tester Strains. The strains used, TA1535, elements were obtained by washing them through a stainless TA1537, TA1538, TA98, and TA100 of Salmonella typhimu- steel screen (60 mesh). rium, were kindly supplied by Dr. Bruce N. Ames (Department The technique of DNA alkaline elution was as reported pre of Biochemistry, University of California, Berkeley, Calif.). viously (5, 27). In brief, an aliquot of 1 x 106 cells (or nuclei) Metabolic System. A mixture (S-9 mix) of subcellular prep was loaded onto a Millipore filter composed of mixed esters of arations (S-9 fraction) and of suitable cofactors was used as cellulose (25 mm in diameter; 5 (im pore size), washed with the metabolic activating system in bacterial mutagenesis as cold Merchant's solution, and lysed at room temperature says. The S-9 mix contains 5 mw glucose 6-phosphate, 4 mw (-20°) with 0.2% sodium lauroylsarcosinate:2 M NaCI:0.02 M NADP, 8 mw MgCI2, 33 mM KCI, and 100 mM sodium phos disodium EDTA (pH 10). After the lysing solution was washed phate (pH 7.4). Postmitochondrial (9000 x g supernatants) off with 0.02 M disodium EDTA, single-strand DNA was eluted preparations of liver homogenates from Sprague-Dawley rats from the filter in the dark with 0.06 M tetraethylammonium (protein concentrations, 33.5 mg/ml) and from liver (30 mg/ hydroxide:0.02 M disodium EDTA (pH 12.3) at a pump speed ml) and lung (19 mg/ml) homogenates from male Swiss albino of 0.13 ml/min by collecting 10 fractions of 1.3 ml each. The mice were used as sources of microsomes (S-9 fractions). DNA remaining on the filter was recovered by breaking up the Microsomal enzymes were induced by means of a single i.p. filter in 2.6 ml of eluting solution. DNA determination was injection of a polychlorinated biphenyl mixture (Aroclor 1254; performed as described previously (6, 27) by an adaptation of Monsanto Co., St. Louis, Mo.) 5 days before sacrificing the the microfluorometric method of Kissane and Robins (18). The animals. The S-9 fractions were adjusted at a concentration of recovery of applied DNA was 80% or greater. Data are ex 30 mg protein per ml and were used at 3 different concentra pressed both as the percentage of increase of DNA elution tions, i.e., 10, 30, and 50 /il/plate, in 0.5 ml of the buffer- over controls and as elution rate over controls. Increased cofactor solution. elution is equal to the percentage of DNA eluted from the filters Bacterial Mutagenesis Assays. Mutagenicity was checked in treated mice minus the percentage of DNA eluted from the by means both of the plate incorporation test and of the spot filters in corresponding control mice and gives direct informa test, basically as described in detail by Ames e/ al. (1 ). Briefly, tion about the experimental results obtained. K, —Kc is the the plate incorporation test was performed by mixing 50 p.\ of average elution rate over controls, where K, is the elution rate overnight broth cultures (2.5% Oxoid Nutrient Broth No. 2) of of DNA from treated mice and Kc is the elution rate of DNA each bacterial tester strain, 100 ¡uloftest hydrazine, 0.5 ml of from the corresponding control mice. K is given by the following S-9 mix (when needed), and 2.5 ml of molten top agar (0.6% formula Difco agar:0.5% NaCI solution) supplemented with 10% of a -In (fraction of DNA retained on filter) solution of 0.5 mM i_-histidine-HCI:0.5 mM biotin. The mixture K — was immediately poured on a minimal-glucose agar medium
APRIL 1981 1471
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. S. Parodi et al.
(1.5% Difco agar in Vogel-Bonner Medium E with 2% glucose) solidified layer in Pétriplates (25 ml of medium in 90-mm- diameter plastic plates). A number of assays was also carried out by means of a liquid preincubation test, as described in "Results." The spot test was performed by placing 10 jul of hydrazine solution in sterile 7-mm filter paper discs at the center of plates over the top agar layer incorporating bacteria and when needed S-9 mix. All mutagenicity assays were per formed in triplicate, except controls (100 /J of distilled water or dimethyl sulfoxide), which were run on 5 test plates. Revertant colonies were scored after 48 hr at 37°in the dark.
Evaluation of Carcinogenic Potency
The main reason why we chose the class of hydrazine derivatives to compare their carcinogenic potency with their 0 20 20 40 40 60 60 80 80 100 potency in inducing DNA fragmentation in vivo and mutagen Weeks after starting treatment icity in the Sa/mone//a-microsome test was that a large number Chart 1. Relationship between weeks from starting of treatment and lung of these compounds were assayed for their carcinogenicity by tumor incidence In male and female Swiss mice given 0.1% isoniazid in drinking Toth (42) according to a fixed experimental protocol. In all water for life span (/ f) and in male and female control mice «f). Data obtained from the work of Toth and Shubik (47). cases (except procarbazine), male and female Swiss mice were treated for their entire life span with the tested hydrazines dissolved in drinking water. In these experiments, the incidence which can be rewritten as of various types of tumors was investigated; however, for our (-ln< purpose of ranking the compounds tested in terms of molar f- 5 potency, we considered only pulmonary tumors. The large majority of these tumors were represented by lung adenomas. The line of reasoning followed to obtain this equation is pre sented in the "Appendix." The main reason for this choice was that lung adenomas are the most frequent tumors induced by most hydrazine deriva Since the mice received daily doses p.o. of all hydrazine tives. In consequence, they offered a piece of datum significant derivatives tested (except procarbazine) until spontaneous from a statistical point of view, in that we were dealing with death or sacrifice, the daily dose reflects the concentration of reasonable tumor frequencies. Lung adenomas are essentially the carcinogenic agent to which target cells were exposed. In "incidental tumors" (28) (discovered at necropsy of an animal the case of procarbazine (16), a single dose was administered; calculation was adapted for this special case (see "Results"). which died of some other cause). They can perhaps be best assessed by serial killing at specific intervals. An acceptable way of analyzing the experiments (15) can be to estimate the RESULTS proportions of animals with pulmonary tumors at subsequent intervals, for deaths occurred for other reasons. In a case DNA Damage in V/Vo-Alkaline Elution Assay. Control ex where the results of Toth and Shubik (47) could be analyzed periments were grouped in subsets of 20 elements each, as explained in "Materials and Methods." The reproducibility of for the frequency of animals with at least one adenoma at subsequent intervals of 20 weeks duration, we obtained the alkaline elution patterns of DNA from livers and lungs of control graph shown in Chart 1, which is typical of lung adenomas mice is shown in Chart 2. The largest difference in the per (15). This graph suggests that the probability of having at least centage of eluted DNA was found in liver between Subsets 4 one lung adenoma/animal increases exponentially with time, and 7 (~9%) and in lung between Subsets 3 and 6 (~9%). which is typical of most tumors. Unfortunately, in most cases, Most commonly, the differences were on the order of 1 to 5%. only tumor frequency and average latency linked to that fre It seems evident that the results obtained with the various quency could be obtained from Toth's experiments. Conse subsets of controls, even if experiments were performed always quently, we needed a relatively simple strategy that could make with the same methodology, were sometimes statistically dif use of this last type of less detailed data, taking into account ferent when assays were carried out at different times and with the essential effect of time on tumor frequencies, as suggested different batches of solutions. Consequently, even if treated by Chart 1. A simple "local multistage model" of tumor occur subsets were always referred to control subsets carried out rence could be imaged according to criteria similar to those approximately in the same time interval and mainly with the already reported in other studies (20, 29). The formula used to same batches of solutions, we estimated that the possibility of calculate the carcinogenic potency was as follows: a hidden trend, small but reaching statistical significance, could not be disregarded. Therefore, we selected the following rela Carcinogenic potency = [(K, - Kc) x 100/mmol/kg/day] tively severe threshold of positivity: percentage of increase of where K is an index of carcinogenic potency (K, refers to DNA elution rate over controls >10%; and p < 0.05. We treated mice and Kc to control mice). The equation actually considered this a sufficiently high threshold level because used to calculate K from the frequency of animals with at least differences found among control subsets were always less than one tumor (1 - P0) and exposure duration in weeks (f ) was 10%. The results obtained with the 16 hydrazine derivatives ex P0 . e-'*" - w amined are shown in detail in Table 2. In order to classify the
1472 CANCER RESEARCH VOL. 41
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA Damage, Mutagenicity, and Carcinogenicity of Hydrazines
drazine (TA100); hydralazine (TA1535, TA100, and TA98); 40 phenylhydrazine (TA1537); and dihydralazine and 2,4-dinitro- phenylhydrazine (TA1537, TA1538, TA98, and TA100). As shown in the 4 examples reported in Fig. 1, the patterns of the toxic and mutagenic effects in the spot test were considerably 30 variable among the different hydrazines. The same 7 compounds, plus 6 additional hydrazine deriva tives (i.e., hydrazine hydrate, mebanazine oxalate, carbamylhy- drazine, 1,1-dimethylhydrazine, procarbazine, and isoniazid), S20 were found to be positive (at least a 2-fold increase of revert- K ants over controls) in the plate incorporation test. The 3 re O maining negative compounds (i.e., isocarboxazid, iproniazid, ni and 1-carbamyl-2-phenylhydrazine) could be tested without IQ toxic effects for bacteria at maximum doses of 56, 172, and 132 /¿mol,respectively. z Chart 3 shows dose-response curves drawn with each of the o 13 positive hydrazines towards each of the sensitive bacterial 23456 23456 7 tester strains, both in the absence and in the presence of S-9 mix containing rat liver S-9 fraction. LIVER SUBSETS LUNG SUBSETS The levels of spontaneous revertants of each strain, with or 23456 23456 7 without S-9 mix, as well as the spectrum of activity of each
.052.10.OS3>.10.OO2>.104.10.05.10>.K)5.02.10.02.001.026.002.10.001.001.OÒ1>.102.01.023.10.10^14>.10.10.001,105.10.10.001.10^.106.05.05.OO1.05.05..05compound, can be inferred from these curves. The results reported are indicative of representative experiments. How ever, all the compounds were separately tested at least 5 times, and the weakest ones (i.e., 1,1-dimethylhydrazine, mebana zine, and procarbazine) were tested at least 10 times. The mutagenic potency of the positive hydrazines was either L.1O unaffected (hydralazine, dihydralazine, 1,2-dimethylhydrazine, Chart 2. Elution patterns of liver and lung DNA from control mice. Each subset and mebanazine) or decreased more or less markedly (2,4- includes 20 experiments. Bars, S.D. The statistical significance (p) of the differ ences among subsets, evaluated according to the Wilcoxon 2-sample (2-tailed) dinitrophenylhydrazine, nialamide, phenelzine, phenylhydra test (36), is reported in the lower part of the chart. zine, hydrazine, carbamylhydrazine, and isoniazid) in the pres ence of S-9 mix containing rat liver homogenates. The only 2 compounds examined according to their capability of inducing compounds with mutagenicity that was slightly increased by DNA fragmentation, the following criterion was adopted. We this metabolic system were 1,1-dimethylhydrazine and procar considered as a reasonable potency coefficient the ratio be bazine. Liquid preincubation of these compounds with bacteria tween elution rate over controls (roughly proportional to the and S-9 mix for 30 min at 37° did not further enhance their number of single-strand breaks) and administered doses, ex mutagenic activity. All the observed metabolic effects (either pressed as mmol/kg of body weight. For each hydrazine increase or decrease of mutagenicity) were less evident by derivative, computation of DMA-damaging potency was based lowering the amount of S-9 fraction (10 or 30 instead of 50 id/ only on the most effective dosage schedule. In terms of their plate). ability to cause fragmentation of liver and lung DNA (potency Table 4 shows the number of spontaneous revertants of each coefficient), the hydrazine derivatives ranked as shown in Table of the 5 bacterial tester strains, either in the absence or in the 3. presence of S-9 mix containing 3 different S-9 fractions from Light microscopy of histological preparations revealed areas Aroclor-treated animals, i.e., rat liver, mouse liver, and mouse of necrosis (clearly evident but of limited size) only in the livers lung. The mutagenic response induced by the positive com dissected 6 hr after treatment from mice given iproniazid (7.14 pounds in the most sensitive Salmonella strains, with or without mmol/kg), hydrazine (3.12 mmol/kg), phenelzine(0.72 mmol/ metabolic systems, is reported in Table 5. As assessed on the kg), and procarbazine (0.52 mmol/kg/day given for 5 succes basis of experiments performed at least at 3 dose levels for sive days). A comparison with the amount of DNA fragmentation each compound, the metabolic effects produced by mouse observed after treatment with the 16 hydrazine derivatives liver S-9 were generally somewhat less pronounced than those (Table 2) shows that any relationship between DNA damage produced by rat liver S-9 but within the same order of magni and cell necrosis can be excluded. For example, liver necrosis tude. The only exception was nialamide, the mutagenicity of was more severe and constantly present in mice treated with which was enhanced by mouse liver S-9 instead of being iproniazid that did not induce DNA damage. decreased, as was the case for both rat liver and mouse lung Mutagenesis Assays. All the mutagenicity assays were per S-9. Such an effect required the presence of NADP in S-9 mix. formed on blind samples. Preliminary experiments were carried Mouse lung S-9 was even more active than rat or mouse liver out by means of the spot test by assaying 10 /il of a 10% in lowering the mutagenicity of phenelzine, but this effect was solution of each compound, i.e., 1 mg/plate. Under these NADP independent. Conversely, mouse lung S-9 was less conditions, a mutagenic response could be detected for the efficient both in increasing or in decreasing the mutagenic following 7 hydrazines of the 16 tested: nialamide and phenel- potency of all the other positive compounds. No mutagenic zine (sensitive strains, TA1535 and TA100); 1,2-dimethylhy- response was elicited by the 3 negative hydrazines also in the
APRIL 1981 1473
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. Table 2 DNA fragmentation induced by hydrazine derivatives administered i.p. in male Swiss mice controlsCompoundHydrazine % of increase of DMA elution over controlsa/elution rate over
inter val after (mmol/kg X treatment of no. oftreatments)3.12 (hr)166616616616616642464246424616616166166616642466424661664246No.ofmice8959666966118971266666669666776799669791114129588991215695886LiverAv.values7.6/0.00817.4/0.0206.7/0.00712.7/0.0144.5/0.0044.1/0.0045.7/0.00630.2/0.03753.9/0.10841.1/0.0565.5/0.0065.4/0.0066.1/0.00717.0/0.02022.8/0.02717.4/0.0224.1/0.004-2.2/-0.0032.2/0.00210.8/0.011-2.2/-0.0033.2/0.0033.1/0.0030.3/mice995866699e678911666666696666766886e69911131295989999e66776LungAv.values3.9/0.00417.2/0.0214.8/0.00511.9/0.0145.6/0.0064.7/0.00510.4/0.0126.8/0.00810.1/0.0118.9/0.0100.7/0.001-2.6/-0.0037.3/0.00822.6/0.0336.5/0.00815.9/0.0204.7/0.0052.0/0.0025.8/0.0061.2/0.0023.2/0.00415.3/0.0184.5/0.0052.2/0.0032.5/0.0037.2/0.0083.7/0.0046.3/0.00714.3/0.01635.3/0.0491.0/0.00133.7/0.05526.6/0.0396.9/0.008-0.3/ hydrate1 X11
.56 x11 .04 x54.16
,1-Dimethylhydrazine1 X10.7
X51
,2-Dimethylhydrazine-2HCICarbamylhydrazinePhenylhydrazine1-Carbamyl-2-phenylhydrazine2,4-DinitrophenylhydrazineMebanazine.43 X10.24
X53.28
X10.55
X51
.57 X10.79 X10.07 X52.62
X11.31 X10.44 X54.54
X12.27 X10.38 X51.22 oxalatePhenelzine-H2SO4Procarbazine-HCIIsoniazidIpronlazidNialamideIsocarboxazidHydralazineX10.61 X10.20 0-0.2/ X50.72 0-0.2/
X10.24 04.0/0.0043.8/0.00412.8/0.01334.6/0.04419.5/0.02538.7/0.06318.9/0.0248.3/0.0101.9/0.0027.0/0.008-1.7/-0.0022.0/0.0023.0/0.003-0.1/
X53.10
X11
.55 X10.52
X50.96
X10.48 X10.32 012.4/0.0162.8/0.003-2.8/-0.0031.2/0.001-1.6/-0.0019.3/0.0123.6/0.0044.9/0.0062.8/0.00315.6/0.021-1.5/-0.001-6.4/-0.0071.8/0.0022.9/0.0048.4/0.0111.4/0.0027.7/0.0093.9/0.0045.2/0.006P6-0.10<0.05>0.10<0.02>0.10>0.10<0.002<0.02<0.002<0.05>0.10>0.10<0.10<0.002>0.10<0.02<0.10>0.10<0.10>0.10>0.10<0.00?>0.10>0.10>0.10>0.10>0.10<0.05<0.002<0.002>0.10<0.002<0.002<0.05>0.10<0.002>0.10>0.10>0.10>0.10<0.002>0.10>0.10>0.10<0.002>0.10<0.02>0.10>0.10-0.10>0.10<0.05>0.10<0.02 X50.32 X157.14
X11.19
X52.92 013.8/0.0167.0/0.0075.2/0.006-1.2/-0.0016.6/0.0072.1/0.0028.8/0.00924.6/0.0312.4/0.0039.2/0.0111.3/0.0015.1/0.0053.2/0.003-2.S/
X10.49
X50.49 X150.42
X10.07
X50.07 X150.42
HCIDihydralazine-H2SO4Dose X10.14
X51.10
X10.18
X 5Time -0.003p"<0.05<0.002<0.05<0.002-0.05<0.05<0.02<0.002<0.002<0.002>0.10>0.10-0.10<0.020<0.002<0.002>0.10>0.10>0.10<0.02XJ.10>0.10<0.10>0.10>0.10>0.10>0.10>0.10<0.002<0.002<0.002<0.002<0.002-0.10>0.10<0.05>0.10-0.33-0.31>0.10<0.002<0.05<0.02>0.10<0.10>0.10<0.002<0.002>0.10<0.02>0.10>0.10>0.10>0.10No. " Percentage of increase of DNA elution over controls = (percentage of DNA eluted in treated mice) —(percentage of DNA eluted in control mice of the corresponding subset). Elution rate over controls = K, - Kc, where K, is the elution rate of DNA from treated mice and Kc is the elution rate of DNA from control mice of the corresponding subset. K is given by the formula:
-Infraction of DNA retained on filter)
where V is the eluted volume in ml (13 ml). " The statistical significance of the increase in DNA elution rate induced by a given treatment was calculated versus the subset of controls carried out in the same time interval by means of the Wilcoxon 2-sample test (2-tailed) (36). 1474
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA Damage, Mutagenicity, and Carcinogenicity of Hydrazines normalized for latency period (see "Materials and Methods"), Table 3 DNA-damaging potency of hydrazine derivatives was even lower than in male Swiss mice (~3% at 60 days). S.E.)aCompounds1 DNA-damaging potency (±2 The potency values obtained for the hydrazine derivatives tested for carcinogenicity are shown in Table 6. The range of these values is 1 to 1900. Of these 11 hydrazines, the carci ,2-Dimethylhydrazine-2HCIProcarbazine (62.9-90.9)40.6 (4.2-11.9)35.5(23.3-51.0)20.9(13.4-29.3)50.0 -HCIPhenylhydrazineIsocarboxazidHydrazine(29.7-55.5)34.3 nogenic activity of Carbamylhydrazine was borderline (0.10 > (20.3-50.6)29.8 p > 0.05) and that of hydralazine was statistically not significant (33.3-66.7)6.7 (20.0-40.0)6.4 (p > 0.10). It is important to underline that the classification hydrateNialamide2,4-DinitrophenylhydrazineIsoniazid1(4.2-9.0)5.5 (2.2-11.9)Inactive9.5 (3.4-7.5)2.4 remained essentially unchanged in terms of both its ranking (0.7-4.4)InactiveInactiveInactiveInactiveInactiveInactiveInactiveInactiveInactiveLung7.8(3.2-16.3)10.0 and relative potency values even if the less sophisticated (5.6-15.0)3.4 ,1-DimethylhydrazineHydralazine4.6)InactiveInactiveInactiveInactiveInactiveInactiveInactivea (2.3- formula HCIPhenelzine• (1 - PO)x 100 -HzSO,,Dihydralazine H2SO»CarbamylhydrazineMebanazine• (mmol/kg/day) x (f - 5)
oxalate1 was adopted for evaluating carcinogenic potency. The main -Carbamyl-2-phenylhydrazineIproniazidLiver75.5 reason for this is probably related to the fact that there are not DNA-damaging potency was calculated by the formula: large differences in average latent periods in all cases except procarbazine. (K, - Ke) ±2 S.E. x 1000 mmol/kg DISCUSSION where (K, —KJ is the average DNA elution rate over controls. For these experiments, the number of samples was between 5 and 15. In this case (Ref. 2, Firstly, it is important to emphasize, even if obvious, that no p. 50), assuming for potency a normal distribution, for 2 subsequent ranks of potency, the difference is statistically significant (p < 0.05) when the (ith —2 data (ours or from literature) were purposely discarded in order S.E.) potency value is not overlapping with the [(;' + 1)th + 2 S.E.] value. to somehow "improve" a given pattern of the results. In the 6 Inactive, the compound did not meet the threshold of positivity chosen: experimental work, only rare (<5%) and grossly evident tech increase of DNA elution rate over controls >10%; and difference from controls statistically significant (p < 0.05). nical failures were discarded and always before any elaboration of the data. Moreover, data were organized and a pattern presence of mouse liver or lung preparations. discovered only after almost all data had been collected. Carcinogenic Potency. Of the 16 hydrazine derivatives The use of short-term tests for carcinogenic risk assessment tested for DNA-damaging and mutagenic activity, to our knowl requires, at least for broad ranges, a satisfactory correlation edge 11 were already tested for their ability in inducing pul (not only qualitative but also quantitative) between potency in monary tumors in mice. Eight of these 11 compounds (hydra short-term assays and carcinogenic potency. Of course, a zine, 1,1-dimethylhydrazine, 1,2-dimethylhydrazine, carbam- relationship within a 2- to 3-fold range cannot be realistically ylhydrazine, 1-carbamyl-2-phenylhydrazine, phenelzine, iso- expected; however, if no quantitative relationship exists even niazid, and hydralazine) were tested by Toth et al. (40, 41, 43- within a 10- to 100-fold range, the predictive value of the short- 48) in male and female Swiss mice by lifetime administration in term assay is limited. drinking water. We utilized only data on male mice because The capacity of mono- and disubstituted hydrazines to inter our experiments on DNA damage were carried out in this sex. act with DNA had been already reported by Freese et al. (13), However, the differences in carcinogenic potency between the who showed that hydrazine, iproniazid, phenelzine, phenipra- 2 sexes never exceeded a ratio of 1:2. Phenylhydrazine and zine, benzylhydrazine, nialamide, isocarboxazid, mebanazine, iproniazid (8) were administered chronically (36 to 42 weeks) isoniazid, and pivalylbenzhydrazine cause, with efficiency in by daily gavage in BALB/c mice of both sexes (results of males the decrease, inactivation of transforming DNA (as measured and females recorded together). When normalized for the same by the decreased frequency of transformants for the tryptophan latency period (see "Materials and Methods"), tumor fre marker of Bacillus subtilis). In our experimental model, 9 of the quency in control Swiss and BALB/c mice was of the same 16 hydrazine derivatives tested (56%) were found active in order (7 to 17% at 60 weeks). Because of the similar back inducing liver and/or lung DNA fragmentation, according to ground level in the 2 strains, we assumed that they had a the positivity threshold established (Table 3). Of the 9 hydra similar susceptibility to pulmonary tumor induction. Procarba- zines shown to be clearly tumorigenic in mice, 6 (67%) were zine was tested in BALB/c x DBA/2 F, (hereafter called found to be positive and these included all 5 of the most potent CD2Fi) mice and in noninbred albino mice of both sexes (16). carcinogens (Table 6). These 5 spanned a potency range of In both strains, it appeared to be by far the most potent 130 for carcinogenicity and of 22 for DNA damage. Carbam carcinogenic hydrazine. This fact is also confirmed (at least for ylhydrazine and hydralazine, respectively, borderline and neg the small number of mice tested) from a direct comparison with ative in terms of carcinogenicity, were both found to be negative 1,2-dimethylhydrazine (17). Because of the carcinogenic po in terms of DNA damage. The correlation between carcinogenic tency of procarbazine, subsets of mice with a tumor frequency potency and DNA-damaging potency is shown in Chart 4. A less than 100% could be found only after a single p.o. admin regression analysis between the 2 activities (logarithmic values) istration. In this special case, the amount (mmol/kg/day) was was carried out. For liver DNA damage, the slope of the calculated by dividing the single dose for the average latent regression line was 0.77, r = 0.77, and p < 0.005 (one tailed). period. In order to build a larger subset, we pooled together For lung DNA damage, the slope of the regression was 0.57, data obtained in CD2F, mice with 2 different doses (15 and 36 r = 0.73, and p < 0.005 (one tailed). In both cases, the mg/mouse). Tumor frequency in control CD2F! mice, when correlation seems statistically very significant. This suggests
APRIL 1981 1475
Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. 640 razin«
JL 0 9.4 19 38 75 150 0 0.5 8 o"~~~~.» -4—'6400 Hydnlazine N, alam.de /'.*"'' i- 3200 <1> 1600 > 800 ^ 200 r 100 o _J 0 0.2 O.B 3.2 13 51 0 0.13 0.5 2.1 8.4 33 Amount per piate (umol) Bacterial strains O TA 1535 0) • TA 100 with S-9mix -M o TA 1537 A TA 1538 without S-9mix CD A TA 98 U. Carbamylhydrazinc \ TOXIC \EFFECT 0) —^o/ i Q. 0 a3 17 34 67 133 Hydrazine hydrate co 4-» C TOXIC V (D EFFECT— 0 1.25 25 5 10 20 -i— —.— —i— —i— —i—;»—• CD > (D 0 72 9.7 19 39 77 116 155 0 18 36 73 146 292 Amount per plate (jjmol) Chart 3. Dose-response curves obtained with 13 mutagenic hydrazines, with or without S-9 mix, towards 5 his strains of S. typhimurium. For sake of visual clarity, only the curves showing a mutagenic response are drawn in graphs, and the confidence limits are not represented. 1476 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA Damage, Mutagenicity, and Carcinogenicity of Hydrazines Table 4 Mean values and standard deviations of the background revertants for the Salmonella tester strains Revertants/plate With S-9 mix saimoneiia strainTA1535TA1537TA S-9mix12 liver15 liver14 lung11 ± G" ± 5 ±316 ± 7 H20 11 ± 2 14 ± 4 ± 4 13 ±215 DMSO0None 14 ±57 12 ±76 18 ±89 ±58 ±36 ±310 ± 5 ± 6 H2O ± 3 ± 6 7 ± 3 5 ± 3 DMSONone 5 ±419 7 ±425 8 ±629 6 ±221 1538TA98TA100" ± 8 ± 9 ±12 ± 5 H2O 16 ± 6 23 ±10 24 ± 9 19 ± 8 DMSONone 15 ±932 22 ±643 25 ±845 20 ±939 ±11 ± 8 ±18 ±12 HjO 28 ± 8 42 ±11 36 ± 9 35 ±14 DMSONone 25 ±14191 38 ±151 40 ±12166 32 ±7182 ±20 64 ±11 ±19 ±23 HjO 194 ±13 161 ±14 172 ±26 175 ±16 DMSO2 183 ±23Rat 172 ±21Mouse 169 ±15Mouse 173 ±20 Mean ± S.D. 6 DMSO,SolventNone dimethyl sulfoxide.Without Table 5 Mutagenic response (number of revertants in excess of controls) elicited by 13 hydrazine derivatives positive in the Ames test, in the absence or in the presence of S-9 mix containing S-9 fractions from Aroclor-treated animals Revertants/plateCompound2,4-DinitrophenylhydrazineNialamidePhenelzine S-9mixMouse plate2.54.24.3514.61419104.46742116146SalmonellastrainTA1 mix2416S-9 liver2131 liver2197 lung2286 ±219a2586 538TA1535TA ±1371613 ±2414684 ±1191691 ±203789 ±108354 ±376375 ±1871 -H2SO4Hydralazine-HCIPhenylhydrazineDihydralazine 1535TA100TA1537TA±1162487 ±872543 ±942705 68 ±422592 ±181153 ±23675 ±28068 ±183119± ±24390 ±21378 ±16397 32366 H2SO41 • 100TA100TA1535TA1538TA1535TA98TA98TA1535Without±68176± ±511 ±83157 ±74181 ,2-Dimethylhydrazine-2HCIHydrazine 1974 68 ±2944 ±1441 ±3167 hydrateMebanazine ±1926 ±1431 ±829 ±1225 oxalateCarbamylhydrazine1 ±761 ±612 ±225 ±852 ±2824 ±537 ±740 ±1928 ,1-DimethylhydrazineProcarbazine.HCIIsoniazid/imol/ ±1135 ±1259 ±952 ±537 ±961 ±1523 ±1144 ±952 ± 15Rat ± 6With ± 12Mouse ± 11 " Mean ±2 S.D. that the "DMA damage in w'vo-alkaline elution assay" has some served among control subsets, the predictive ability would predictive value. At the same time, we must underline that decrease, even if qualitatively 3 additional compounds iproniazid, phenelzine, and 1-carbamyl-2-phenylhydrazine (phenelzine, hydralazine, and dihydralazine) would have to be were all clearly identified as carcinogens in mice, although they considered positive (i.e., a total of 75% of the 16 hydrazines all had a carcinogenic potency <1.5 (Table 6). On the contrary, tested). We are aware that both different DNA adducts and they escaped any detection in the DMA damage assay. This their poorly known (except for methylatmg-ethylating agents) seems to suggest that the assay, albeit predictive, has a different efficiencies in producing DNA single-strand breaks somehow limited sensitivity. Concerning the 5 hydrazine deriv and alkali-labile sites could have obscured the relationship atives not yet tested for their carcinogenic activity, isocarbox- between DNA fragmentation and carcinogenicity (3, 20). Ap azid, nialamide, and 2,4-dinitrophenylhydrazine were found to parently, this was not the case for hydrazine derivatives. Only be positive in terms of DNA damage (Table 3). Isocarboxazid similar works will tell us what other classes of chemicals will showed a potency level of 29.8 for liver and 50 for lung; the offer a similar favorable situation. It is also evident that the potency level of nialamide was 5.5 for liver, and that of 2,4- parameter chosen to evaluate carcinogenic potency, i.e., inci dinitrophenylhydrazine was 2.4 for liver and 9.5 for lung. dence of pulmonary tumors in mice (mainly adenomas), cannot Considering the correlation existing between DNA-damaging be considered a "universal" parameter of carcinogenicity. potency and carcinogenic potency, it seems likely that at least The mutagenic potency in the Ames test was expressed as isocarboxazid could be a relatively potent carcinogen. If we number of induced revertants per/imol, as calculated from the had established a positivity threshold at a percentage of DNA data reported in Table 5. It varied from 0.3 (procarbazine) to elution over controls smaller than the internal variability ob 966 (2,4-dinitrophenylhydrazine) in assays performed without APRIL 1981 1477 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. S. Parodi et al. Table 6 Carcinogenic potency of hydrazine derivatives Pulmonary tumor Av. latent - Kc) X potency837961.43.533.002.891.501.000.520.340.200.11Ref.15443783683943424140a CompoundProcarbazine1 (wk)0.0420.0220.490.230.190.280.331.340.995.51.2585day incidence (%) period X 10216.82.272.651.951.471.681.221.651.231.971.30KcX1020.900.920.921.260.921.260.890.950.890.891.16(K,10215.91.351.730.690.550.420.330.700.34°1.080.14°Carcinogenic ±10624 HCI1,2-Dimethylhydrazine- ±678 ,1-DimethylhydrazinePhenylhydrazine- ±653 HCIHydrazine ±950 sulfateIproniazidPhenelzineIsoniazidCarbamylhydrazine ±735 ±836 ±742 ±730 HCI1 • ±667 -Carbamyl-2-phenylhydrazineHydralazine-HCImmol/kg/ ±746 ± 71237475367537257685773K, Carcinogenic potency was calculated by the formula:(K, -Kc)x100mmol/kg/day where , (-In Po)1" " f-5 ' K, is K for treated mice, and Kc is K for corresponding control mice. The number of treated mice used for these experiments was between 30 and 55 (except for procarbazine which was 13 mice). The number of control mice was between 40 and 110. When, in a sample, n a 30, the binomial distribution tends to normality, and S.E, was calculated accordingly (Ref. 2, p. 24). Mean ±S.E. c From K, and Kc, the percentage of tumor incidence was normalized for the same latent period (e.g., 50 weeks). The comparison of the 2 percentages (Ref. 2, p. 38) showed that the difference in tumor incidence between treated and control mice was always statistically significant (p < 0.05), except for hydralazine (p > 0.10) and Carbamylhydrazine (0.10 > p > 0.05). 1OOOoZ100.u1-210uzUl Ç8l.2 DMHIHM*: 1spi .V ¡DMH inactivS.i ^MIUcu» 1 1O 100 1000 INHCPHU 01ui0PEHèHPH CARCINOGENIC POTENCY SIPNH ÕnaCtÕV0.1 I 10 100 1000 CARCINOGENIC POTENCY Chart 5. Relationship between mutagenic potency in the Salmonella-micro- some test in the presence (•)and in the absence (O) of S-9 mix (rat liver S-9 fraction) and carcinogenic potency (induction of pulmonary tumors in mice) for 11 hydrazine derivatives. For abbreviations, see legend to Chart 4. tionship between mutagenic potency (in the presence and in the absence of rat liver S-9 mix) and carcinogenic potency is shown in Chart 5. A regression analysis of the relationship Q .£ between the log of carcinogenic potency and the log of muta 10 100 1000 genic potency (expressed in terms of number of revertants per CARCINOGENIC POTENCY fimol) was carried out. For the assays performed without S-9 mix, the slope of the regression line was —0.03, r = —0.03, Chart 4. Relationship between DNA-damaging potency in liver (A) and in lung (B) and carcinogenic potency (induction of pulmonary tumors in mice) for 11 and p » 0.10 (one tailed). Similarly, with S-9 mix containing hydrazine derivatives. HH. hydrazine; 1.1DMH, 1,1-dimethylhydrazine; 1.2DMH. liver homogenates from Aroclor-treated rats, the slope of the 1,2-dimethylhydrazine; CH, Carbamylhydrazine; PH, phenylhydrazine; CPH, 1- carbamyl-2-phenylhydrazine; PEH, phenelzine; MIH, procarbazine; INH, isonia- regression line was 0.09, r = 0.08, and p s> 0.10. Correlation zid; IPNH, ¡proniazid; HPH, hydralazine. with the carcinogenic potency was not improved by considering the mutagenicity data obtained in the presence of liver (r = S-9 mix and from 0.16 (isoniazid) to 1115 (nialamide) in assays 0.03) or lung (r = 7 x 10~5) homogenates from Aroclor-treated performed with rat or mouse S-9 mix, i.e., over an almost 7000- mice. For what concerns this lack of quantitative correlation fold range. For 11 hydrazine derivatives, the quantitative reÃa- between mutagenic potency and carcinogenic potency, it is 1478 CANCER RESEARCH VOL. 41 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA Damage, Mutagenicity, and Carcinogenicity of Hydrazines evident that no automatic extrapolation is allowed to other had a similar activity in modifying the mutagenicity of positive classes of chemicals. For example, Meselson and Rüssel(25) compounds. However, rat liver was generally somewhat more found a positive correlation between mutagenic potency and efficient than mouse liver S-9 in decreasing the mutagenicity carcinogenic potency for 10 compounds of heterogeneous of some hydrazine derivatives. An example is isoniazid, thus chemical structure, and at the same time, they observed that confirming the data reported by other investigators (49). This nitroso compounds were an exception to this relationship. The finding might account for the conflicting results of carcino last conclusion was also drawn by Nagao ef al. (26), who did genicity tests with isoniazid in rats and mice (49). An exception not find any quantitative correlation between the carcinogenic was nialamide, the mutagenic potency of which was lowered and mutagenic potency of various nitrosamines, despite the by both rat liver and mouse lung preparations, while mouse high sensitivity of the Ames test in detecting carcinogenic liver S-9 had a NADP-requiring potentiating effect. Interest compounds in this chemical class. Perhaps, the lack of quan ingly, nialamide was also the only hydrazine derivative inducing titative correlation between carcinogenicity in mice and muta- DNA fragmentation in mouse liver and not in the lung. Mouse genicity of hydrazines in the Ames test is not so surprising. In lung preparations were less efficient than the liver preparations fact, in respect to the DMA damage assays, we are dealing with obtained from the same animals both in decreasing and in the comparison of 2 biological effects which have been ex creasing the mutagenic activity of the compounds tested. How plored under quite different experimental conditions. Moreover, ever, mouse lung S-9 was particularly active in lowering phe the induction of pulmonary tumors in mice (mainly adenomas) nelzine mutagenicity, but the same effect was observed when cannot be considered as automatically representative of all NADP was omitted in S-9 mix composition. types of carcinogenicity. Another extensive study on the mutagenic activity of hydra It is noteworthy that, with the exception of the hydralazine, zine derivatives in the Ames test had been reported previously the 2 compounds showing a marked prevalence of mutagenic by Shimizu ef al. (37), who tested 26 compounds (9 of which over carcinogenic potency (phenelzine and phenylhydrazine) were also included in our experimental protocol) towards tended to decrease their activity in the presence of S-9 mix. strains TA1537, TA98, and TA100 of S. typhimurium, with and Similarly, the only compound showing a marked prevalence of without S-9 mix containing liver S-9 fractions from rats pre- carcinogenic over mutagenic potency (procarbazine) tended to treated with Kanechlor 500. Four of the 9 hydrazines reported increase its activity in the presence of S-9 mix. These findings as positive by Shimizu ef al. (37) (/'.e., 2-hydroxyethylhydrazine, suggest that the observed discrepancies between mutagenic butylhydrazine, p-tolhylhydrazine-HCI, and p,p'-oxybisben- and carcinogenic potencies should rely on quantitative rather zendisulfonylhydrazine) have not been tested in our laboratory. than on qualitative metabolic differences in vivo and in vitro. In both studies, 1-carbamyl-2-phenylhydrazine was found to In the assessment of the correlation between mutagenicity be negative, while phenelzine, 2,4-dinitrophenylhydrazine, and carcinogenicity, it is also important to consider that mice phenylhydrazine (with a different spectrum of positive strains), were treated by lifetime administration of hydrazine derivatives and hydrazine were mutagenic. However, the last compound in drinking water. This might also involve a transformation step (either as hydrate or as sulfate) required metabolic activation in gastric environment. In fact, the mutagenicity of some pure in the Japanese study, while in the present study hydrazine chemicals can be affected in the presence of gastric juice (11 ). appears to be a weak direct mutagen, the activity of which is Moreover, the carcinogenicity of some hydrazine nitroso deriv further lowered by both rat and mouse liver S-9. The other 4 atives has already been documented (12). However, these hydrazines that we have classified as weak mutagens, i.e., 1,1- differences in experimental conditions should have also af and 1,2-dimethylhydrazine, carbamylhydrazine, and isoniazid, fected the predictive ability of the DMA damage assay. were reported as negative by Shimizu ef al., presumably be Despite the lack of quantitative correlation between muta cause they tested submutagenic concentrations. genic and carcinogenic potencies, at least with the in vivo Our positive results are in agreement with other literature model considered, the Ames test was very efficient in revealing data for hydrazine (14, 24, 33, 49), 1,1-dimethylhydrazine (7, the mutagenicity of hydrazine derivatives. In fact, on the whole, 39), and isoniazid (14, 49), while they disagree with those the 80% (4 of 5) of the hydrazines of unknown carcinogenicity reported for 1,2-dimethylhydrazine and procarbazine (24, 32). and the 81.8% (9 of 11) of hydrazine derivatives reported to The latter had been reported as negative also by Bruce and be carcinogenic, accounting for a total of 81 % (13 of 16), were Heddle (7) but only below 0.5 mg/plate. It must be noted that found to be positive in the Sa/mone//a-microsome test. Carba- various hydrazine derivatives, including 1,2-dimethylhydrazine, mylhydrazine, borderline in terms of carcinogenicity, was also had been found to be mutagenic also in other prokaryotic (21) found to be weak without S-9 mix and almost inactive in the and eukaryotic (50) test systems. The 2 carcinogenic hydrazine presence of S-9 mix in terms of mutagenic potency. Hydrala positives in the Ames test, which were not assayed in our zine, while negative in terms of carcinogenicity, was instead study, are2-hydrazino-4-(4-aminophenyl)thiazoleand 2-hydra- found to be rather potent in terms of mutagenicity. The muta zino-4-(4-nitrophenyl)thiazole (38). On the whole, all these data genicity patterns showed a widely variable range of activity, seem to disprove the view (35) that hydrazine derivatives may concerning not only the mutagenic potency but also the influ represent a class of chemicals poorly detected in the Ames ence of the metabolic system used in vitro and the molecular test, although many of them appear to be very weak mutagens. basis for reversions, as inferred from the spectrum of activity Such a circumstance might explain some discrepant result« towards the 5 bacterial tester strains. This indicates that the obtained in different laboratories. Another factor of variability structural variations introduced in the hydrazine molecule affect might be represented by the difficulty of reproducing in vitro both the mutagenic potency and the relevant genetic mecha some metabolic steps required for the activation of some nism, as well as their metabolic behavior. The liver S-9 fractions hydrazines. It is noteworthy, as a support to our findings, that prepared from the 2 animal species considered (rat and mouse) both procarbazine and 1,2-dimethylhydrazine had been found APRIL 1981 1479 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. S. Parod/ ef al. to be mutagenic for S. typhimurium following an i.p. host- ments with procarbazine (16), the most potent carcinogen, 5 mediated assay (32) towards bacterial strains carrying the weeks could be considered to be a reasonable estimate of the same histidine mutations as those reverted in the present study. minimum latent period. We found that when n = 4 is used in The analysis of the possible correlation between DNA-dam- the above equation the curves obtained fit the experimental aging potency in vivo and mutagenic potency in the Salmonella- graph of Chart 1 best. Therefore, for / in weeks, the equation microsome test was also carried out. No correlation was found actually used in our calculations was: in any of the possible combinations (liver DNA-S-9 mix, liver DNA-no S-9 mix, lung DNA-S-9 mix, and lung DNA-no S-9 mix). With the input of frequency of animals with at least one lung In conclusion, the results obtained seem to suggest that a tumor (1 - P0) and exposure duration (f), the output K reflects similar work of quantitative comparison between carcinogenic the potency of the chronic carcinogenic treatment. In Toth's potency and potency in short-term tests should be carried out experiments, (1 - P0) corresponding to the average tumor for other families of chemicals. Besides hydrazine derivatives, latency time, is never extreme but is always between 24 and there are other sizable groups of chemicals adequately tested 78%. Moreover, the majority of animals died at average tumor for their carcinogenicity and with acceptable homogeneity of latency time, that is in the central more linear part of the curve experimental conditions. In this way, it might be possible to profile shown in Chart 1. For this part of the curve, dy/dt can build up a general picture of the capability of quantitatively be considered approximately constant. Under this assumption, predicting the carcinogenic potential of the most common from: short-term tests against the most important chemical families of compounds. An approach of this type was followed by Clive y,= a + dy/dt •f,. ef al. (9) in an important paper published recently. These we obtain authors investigated 25 different chemicals for the relationship between oncogenic potency and mutagenic potency in the y = a + dy/dt •t. mammalian L5178Y/TK+/~ —»TK~/- mutagenic assay (plus S-9 mix) and showed the existence of a good quantitative This is equivalent to saying that (1 —P0) for the total time correlation between the 2 activities. interval of the experiment can be assumed to be reasonably close to (7 —PO) when considering a shorter time interval around the average latency time. In this way, Toth's data can APPENDIX always be normalized with respect to time. Determination of Carcinogenic Potency. Because of the chronic carcinogenic treatment, the "transforming" potency of REFERENCES the cell environment can be considered (as a first approxima 1. Ames, B. N.. McCann. J., and Yamasaki, E. Methods for detecting carcino gens and mutagens with the Sa/mone//a/mammalian-microsome mutagen- tion) constant with time (at a level much higher than background icity test. Mutât.Res., 31. 347-364. 1975. level). If p is the probability that a single cell reaches a specific 2. Bailey. N. T. J. Statistical Methods in Biology, pp. 24, 38. and 50. London: stage, p = rt, where r is the measure of the level of "transform The English University Press Ltd, 1959. ing1' potency of the cell environment and t is exposure duration. 3. Bradley. M. O., Hsu, I. C., and Harris, C. C. Relationships between sister chromatid exchange and mutagenicity, toxicity and DNA damage. Nature It is assumed that, for a single cell, rt 1480 CANCER RESEARCH VOL. 41 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA Damage, Mutagenicity, and Carcinogenicity of Hydrazines 1976. Nati. Cancer Inst.. 62. 919-926. 1979. 15. Hoel, 0. G., and Walburg, H. E., Jr. Statistical analysis of survival experi 33. Purchase. I. F. H.. Longstaff, E.. Ashby, J.. Styles, A., Anderson, D., Lefevre, ments. J. Nati. Cancer Inst., 49: 361-372. 1972. P. A., and Westwood. F. R. An evaluation of 6 short-term tests for detecting 16. Kelly. M. G.. O'Gara, R. W., Gadekar, K., Yancey. S. T., and Oliverio, V. T. organic chemical carcinogens. Br. J. Cancer. 37: 873-959. 1978. Carcinogenic activity of a new antitumor agent, W-isopropyl-a-(2-methylhy- 34. Registry of Toxic Effects of Chemical Substances. Rockville, Md.: United drazino)-p-toluamide hydrochloride (NSC-77213). Cancer Chemother. Rep., States Department of Health, Education, and Welfare, National Institute for 39. 77-80, 1964. Occupational Safety and Health, 1976. 17. Kelly, M. G., O'Gara, R. W., Yancey, S. T., Gadekar. K., Botkin, C., and 35. Rinkus, S. J., and Legator. M. S. Chemical characterization of 465 known Oliverio, V.T. Comparative Carcinogenicity of W-isopropyl-a-(2-methylhydra- or suspected carcinogens and their correlation with mutagenic activity in zine)-p-toluamide HCI (procarbazine hydrochloride). its degradation prod the Salmonella typhimurium system. Cancer Res.. 39. 3289-3318, 1979. ucts, other hydrazines, and isonicotinic acid hydrazide. J. Nati. Cancer Inst., 36. Rumke. C. L., and De Jonge. H. Design, statistical analysis and interpreta 42.: 337-344. 1969. tion. In: D. R. Laurence and A. L. Bacharach (eds.). Evaluation of Drug 18. Kissane, J. M., and Robins. E. The fluorometric measurement of deoxyri- Activities: Pharmacometrics, Vol. 1. pp. 80-82. New York: Academic Press. bonucleic acid in animal tissue with special reference to the central nervous Inc., 1964. system. J. Biol. Chem.. 233. 184-188. 1958. 37. Shimizu, H., Hayashi, K., and Takemura. N. Relationships between the 19. Kohn. K. W., Erickson, L. C.. Ewig, R. A. G., and Friedman, C. A. Fraction- mutagenic and carcinogenic effects of hydrazine derivatives. Jpn. J. Hyg., ation of DNA from mammalian cells by alkaline elution. Biochemistry, 15: 33:474-485, 1978. 4629-4637, 1976. 38. Sugimura. T., Sato, S., Nagao. M.. Yahagy, T.. Matsushima, T., Seino, Y., 20. Lawley. P. 0. DNA as a target of alkylating carcinogens. Br. Med. Bull.. 36. Takeuchi. M., and Kawachi, T. Overlapping of carcinogens and mutagens. 19-24, 1980. In: P. N. Magee. S. Takayama. T. Sugimura. and T. Matsushima (eds.). 21. Lingens, F. Erzeugung biochemischer Mangelmutanten von Escherichia coli Fundamentals in Cancer Prevention, pp. 191-215. Tokyo: University of mit Hilfe von Hydrazin und Hydrazinderivaten. Z. Naturforsch. Teil B Anorg. Tokyo Press. 1976. Chem. Org. Chem., 19: 151-156, 1964. 39. Tosk, C.. Schmeltz, I., and Hoffman, D. Hydrazines as mutagens in a 22. Litchfield. J. T., Jr.. and Wilcoxon, F. A. A simplified method of evaluating histidine-requiring auxotroph of Salmonella typhimurium. Mutât. Res.. 66: dose-effects experiments. J. Pharmacol. Exp. Ther., 96. 99-113, 1949. 308-315, 1979. 23. McCann, J., and Ames. B. N. Detection of carcinogens as mutagens in the 40. Toth, B. Lung tumor induction and inhibition of breast adenocarcinomas by Salmonella/microsome test; assay of 300 chemicals: discussion. Proc. Nati. hydrazine sulfate in mice. J. Nati. Cancer Inst., 42: 469-475, 1969. Acad. Sei. U. S. A., 73. 950-954. 1976. 41. Toth. B. 1,1-Dimethylhydrazine (unsymmetrical) carcinogenesis in mice. 24. McCann, J.. Choi. E.. Yamasaki. E.. and Ames. B. N. Detection of carcino Light microscopic and ultrastructural studies on neoplastic blood vessels. J. gens as mutagens in the Sa/moneWa/microsome test: assay of 300 chemi Nati. Cancer Inst., 50: 181-194, 1973. cals. Proc. Nati. Acad. Sei. U. S. A., 72. 5135-5139, 1975. 42. Toth. B. Synthetic and naturally occurring hydrazines as possible cancer 25. Meselson. M., and Rüssel.K Comparison of carcinogenic and mutagenic causative agents. Cancer Res.. 35: 3693-3697. 1975. potency. In: H. H. Hiatt. J. D. Watson, and J. A. Winsten (eds.). Origins of 43. Toth. B. Tumorigenicity of /?-phenylethylhydrazine sulfate in mice. Cancer Human Cancer, Vol. C, pp. 1473-1481. Cold Spring Harbor, N. Y.: Cold Res.. 36: 917-921, 1976. Spring Harbor Laboratory. 1977. 44. Toth, B. Tumorigenic effect of 1-hydrazinophthalazine hydrochloride in mice. 26. Nagao, M., Suzuki, E., Yasuo. K., Yahagi, T., Seino, Y.. Sugimura, T., and J. Nati. Cancer Inst.. 6ÃŽ:1363-1365. 1978. Okada. M. Mutagenicity of N-butyl-N-(4-hydroxybutyl)nitrosamine, a bladder 45. Toth. B., and Shimizu. H. 1-Carbamyl-2-phenylhydrazine tumorigenesis in carcinogen, and related compounds. Cancer Res., 37. 399-407, 1977. Swiss mice. Morphology of lung adenomas. J. Nati. Cancer Inst., 52: 241- 27. Parodi, S.. Taningher, M., Santi, L., Cavanna, M., Sciabà , L., Maura. A., and 251. 1974. Brambilla, G. A practical procedure for testing DNA damage in vivo proposed 46. Toth, B., Shimizu, H., and Erickson. J. Carbamylhydrazine hydrochloride as for a prescreening of chemical carcinogens. Mutât.Res., 54: 39-46, 1978. a lung and blood vessel tumor inducer in Swiss mice. Eur. J. Cancer. J /. 28. Peto. R. Guidelines on the analysis of tumor rates and death rates in 17-22, 1975. experimental animals. Br. J. Cancer, 29; 101-105, 1974. 47. Toth, B., and Shubik, P. Carcinogenesis in Swiss mice by isonicotinic acid 29. Peto, R., Roe, J. F. C., Lee. P. N., Levy, L., and Clack, J. Cancer and ageing hydrazide. Cancer Res., 26: 1473-1475, 1966. in mice and men. Br. J. Cancer, 32. 411-426, 1975. 48. Toth, B., and Wilson, R. B. Blood vessel tumorigenesis by 1,2-dimethylhy- 30. Petzold. G., Harbach, P., Bodell. M.. Swenberg. J. A. In vitro DNA damage/ drazine dihydrochloride (symmetrical). Gross, light and electron microscopic alkaline elution assay for predicting carcinogenic potential. Mutât.Res.. 53. descriptions. I. Am. J. Pathol., 64: 585-600, 1971. 110, 1978. 49. Wade, D. R., and Lohman, P. H. M. The mutagenicity of isoniazid in 31. Petzold, G. L., and Swenberg. J. A. Detection of DNA damage induced in Salmonella and its effects on DNA repair and synthesis in human fibroblasts. vivo following exposure of rats to carcinogens. Cancer Res., 38 1589- Mutât.Res., 74: 204, 1980. 1598. 1978. 50. Zimmermann, F. K., and Schwaier. R. A genetic effect of symmetric dimeth- 32. Poirier, L. A., and De Serres, F. J. Initial National Cancer Institute studies on ylhydrazine: induction of mitotic recombination. Naturwissenschaften, 54: mutagenesis as a prescreening for chemical carcinogens: an appraisal. J. 251, 1967. APRIL 1981 1481 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. S. Parod/ eÃal. Fig. 1. Four examples of the mutagenic response elicited in the spot test by hydrazine derivatives (1 mg imbued in the paper disc) without S-9 mix. A. nialamide- TA1535; numerous revertants are directly clustered around the paper disc, without any toxic effect; 6, phenelzine-TA1535; revertants are distributed according to a density gradient around an area of inhibition of bacterial growth; C, phenylhydrazine-TA1537; the gradient distribution of revertants is less evident; D, hydralazine- TA100; revertants are crowded in a ring surrounding the bacterial killing. Background colonies are spontaneous revertants, which are typically numerous for this strain. 1482 CANCER RESEARCH VOL. 41 Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research. DNA-damaging Activity in Vivo and Bacterial Mutagenicity of Sixteen Hydrazine Derivatives as Related Quantitatively to their Carcinogenicity Silvio Parodi, Silvio De Flora, Marco Cavanna, et al. Cancer Res 1981;41:1469-1482. Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/41/4/1469 E-mail alerts Sign up to receive free email-alerts related to this article or journal. Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected]. Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/41/4/1469. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site. Downloaded from cancerres.aacrjournals.org on October 4, 2021. © 1981 American Association for Cancer Research.