[Gann, 75, 245-252; March, 1984]

FORMATION OF VOLATILE NITROSAMINES BY DRUG-NITRITE

INTERACTIONS UNDER PHYSIOLOGICAL CONDITIONS

Ayako SAKAI, Takiko INOUE and Akio TANIMURA Division of Food Additives, National Institute of Hygienic Sciences, Kamiyoga 1-18-1, Setagaya-ku, Tokyo 158

Twenty-eight drugs, most of which are tertiary amines, were tested for the forma

tionof volatile nitrosamines by reaction with nitrite under physiological conditions;

the drugs (10m•ôNH•ôM•ôNS•ô)were incubated with nitrite (40m•ôNH•ôM•ôNS•ô)at pH 3.0, 37•‹ for 1 and 4hr. The volatile nitrosamines formed were determined by gas chromatography-thermal

energy analysis. Of the 28 drugs, 24 formed measurable amounts of volatile nitrosa minesthat are known carcinogens. The yields of nitrosodimethylamine (NDMA)

from aminopyrine (55-65%) and minocycline (11%) were higher than that from

dimethylamine under the same conditions. This result suggests that there may be a pathway not involving the secondary amine (dimethylamine) as an intermediate

in the formation of NDMA from minocycline as well as from aminopyrine, Tolaza midegave rise to nitrosopiperidine (NPIP) in addition to nitrosohexamethylenei

mine(NHXI), formation of which was expected from the chemical structure of tol azamide,and the yield of NPIP (2-7%) was higher than that of NHXI (0.2-1.2%). Ascor

bicacid (40m•ôNH•ôM•ôNS•ô)was effective in decreasing the formation of nitrosamines from drugs by reaction with nitrite, although the blocking effects varied between 88 and

100% depending on the drugs tested or on the nitrosamines formed.

Key words: Formation of nitrosamines-Drug-nitrite interactions-Ascorbic

acid-Nitrosopiperidine from tolazamide-Gas chromatography-thermal energy

analysis

Nitrosamines, many of which are car tainedin many foods,2) as well as being used cinogens,are now known not to occur fre as a food additive. Not only secondary amines quentlyin the environment in significant but also many tertiary amines are potential quantities, contrary to initial expectations.2, precursors of nitrosamines,10,20) and such 3,15) Nevertheless , there are risks that man compounds are widely distributed in the may be exposed to nitrosamines, since in environment. gestedamines can interact with nitrite in Drugs have also been discussed as possible vivo to form nitrosamines.14,19) Nitrite is a precursors of nitrosamines,16,18) and some usual constituent of saliva,6,12) and is con drugs containing tertiary amino groups were found to give rise to carcinogenic volatile

* Abbreviations used: NDMA nitrosamines by reaction with nitrite under , nitrosodimethyl amine;NDEA, nitrosodiethylamine; NDPA, ni the mildly acid conditions of the mamma trosodipropylamine;NHPI, nitrosoheptamethyl lianstomach.8,9) However, ease of forma eneimine;NHXI, nitrosohexamethyleneimine; tionof nitrosamines from such drugs varies NMP, N-nitroso-N'-methylpiperazine; NPIP, ni trosopiperidine;NPYR, nitrosopyrrolidine; GC quite widely. Aminopyrine reacts with nitrite very easily to form NDMA* in high yield, - TEA, gas chromatography-thermal energy anal ysis;GC-MS, gas chromatography-mass spec but chlorpromazine gives only a very low trometry. yield of NDMA.8)

75(3) 1984 245 A. SAKAI, ET AL.

Fig. 1. Structures of the drugs used

246 Gann NITROSAMINES FROM DRUGS AND NITRITE

In the present study, volatile nitrosamines Pure drugs were gifts from the manufacturers formed from the reaction of nitrite with 28 and were of the same grades as used for commercial drugs (Fig. 1), mainly tertiary amines, were drug formulations. NMP and NHXI were synthesized from N- determined by using a gas chromatograph methylpiperazine and hexamethyleneimine, re equipped with a thermal energy analyzer, spectively,according to the procedure of Lijinsky which is a sensitive and specific detector of and Reuber.11) The other nitrosamines used were nitrosamines.5) The drugs examined were purchased from Wako Pure Chemical Ind., Ltd., selected because they are commonly admin Osaka (NDMA, NDEA), Tokyo Chemical Ind. Co., Ltd., Tokyo (NDPA, NHPI), Aldrich Chem isteredby the oral route and represent a wide icalCo., Inc., Milwaukee (NPYR) and Sigma variety of chemical structures. Aminopyrine Chemical Co., St. Louis (NPIP). was selected as a reference compound known Reaction of Drugs with Nitrite Reactions of to form a volatile nitrosamine by reaction drugs with nitrite in vitro were carried out under the conditions recommended by the WHO Study with nitrite (its oral administration has been Group on the Potential Carcinogenicity of Nitro prohibited).8,12) satableDrugs (12-16 June 1978, in Geneva)22): Nitrosation of amines can be inhibited NaNO2 dissolved in 0.5M CH3COONa-HCl buffer solution (pH3.0) was added to a solution by ascorbate, which competes with amines or suspension (pH3.0) of a pure drug or formula for nitrite,1) and Mirvish et al.12,13)have re tion,and the mixture was immediately readjusted commendedsimultaneous administration of to pH3.0 with diluted HCl, if necessary. The final ascorbate with nitrosatable drugs. Thus, the concentrations of NaNO2 and drugs were 40mM and 10mM, respectively. The reaction mixture inhibitory effect of ascorbic acid was exam was incubated at 37•‹ for 1 and 4hr. inedon the formation of nitrosamines from The effect of ascorbic acid was examined by the test drugs. adding ascorbic acid to the solution or suspension of a pure drug before the addition of NaNO2. MATERIALSAND METHODS The final concentration of ascorbic acid was 40mM, and the incubation time was 4hr.

Chemicals Formulations of the drugs tested Commercial formulations (tablets, dragees, werepurchased at retail outlets in the Tokyo area. capsules or powder) were used for nitrosation re Each formulation contained only the intended actionsafter being ground in a mortar without drug and additivesand did not contain other me removing any additives, and the drug contents dicinallyactive components. The contents of claimed by the manufacturers were adopted in drugsin the formulations, according to the manu the preparation of reaction solutions. facturers,are as follows (the weight of one tablet, Determination of Volatile Nitrosamines At dragee or capsule is given in parentheses): chlor the end of the allotted time the reaction was ter promazinehydrochloride, 25mg/dragee (222 minatedby the addition of a large excess of am mg); oxytetracycline hydrochloride, 250mg/ moniumsulfamate to remove nitrite, and the re

capsule (526mg); diphenyldimethylaminoethane actionmixture was extracted 3 times with an hydrochloride,25mg/dragee (490mg); erythro equal volume of CH2Cl2. When the extraction of mycin,200mg/dragee(708mg); acetylspiramycin, NMP was intended, the reaction mixture was

200mg/dragee (463mg); metformin hydrochlo neutralized with NaHCO3 prior to the extraction

ride,250mg/tablet (352mg); thioproperazine procedure. The CH2Cl2 extract was rinsed with dimethanesulfonate,7.15mg/dragee (120mg); a small volume of 1% NaHCO3 solution, dried etamsylate,500mg/1g of powder; procainamide over anhydrous Na2SO4, and concentrated to 10 hydrochloride, 125mg/tablet (203mg); proglu ml in a rotary evaporator under reduced pres mide,200mg/tablet (250mg); probenecid, 250 sure.Nitrosamines in the concentrates were deter mg/tablet (388mg); clemizole hydrochloride, minedby GC-TEA. Identification of the nitrosa 20mg/dragee (188mg); glyclopyramide, 250 mineformed was based on the retention time in a mg/tablet (329mg); triprolidine hydrochloride, GC-TEA chromatogram. Recoveries of nitrosa

10mg/1g of powder; piromidic acid, 250mg/ mineswere 83 to 90%. The values presented in tablet (300mg); dipyridamole, 25mg/dragee (120 tables are not corrected for the recoveries.

mg); thiethylperazinedimaleate, 10.276mg/dragee Chromatography (179mg); perlapine, 25mg/tablet (132mg); a) GC-TEA: Gas chromatograph; Shimadzu tolazamide,250mg/tablet (369mg); guanethidine GC-6AM. Column; 1.5m•~3mm glass, 10% sulfate,100mg/1g of powder. Carbowax 20M on 60/80 mesh Gaschrom P.

75(3) 1984 247 A. SAKAI, ET AL.

Column temperature; 110•‹ for NDMA and NDEA, or 170•‹ for other nitrosamines. Injection port tem RESULTS perature;160 or 190•‹. Carrier gas; argon, 60 The yields of volatile nitrosamines ob ml/min. Thermal energy analyzer; Thermo Elec

tronCorp., TEA-502. Furnace temperature; 400•‹. tainedby the reaction of drugs with nitrite Cold trap; liquid N2-CH2Cl2. Reaction chamber are shown in Table I. The yields were based pressure; 1mmHg. on the concentration of drugs. b) GC-MS: Gas chromatograph; Hewlett- Volatile nitrosamines were formed from Packard 5710A. Column; 25m•~0.3mm capil

larycolumn, SE-54, 0.17ƒÊm film. Column tem all of the drugs incubated with nitrite except

perature;programmed at the rate of 30•‹/min from diphenhydramine, procainamide, probenecid 70 to 170•‹. Injection port temperature; 200•‹. and piromidic acid. The yields of NDMA Carrier gas; helium, 25cm/sec. Mass spectrome from aminopyrine (55-65%), NDMA from ter;Hewlett-Packard 5985B. Ionization source; minocycline (11%) and NPIP from tolaza electron impact. Source temperature; 200•‹. Elec tronvoltage; 70 eV. mide(2-7%) by reaction with nitrite were

Table I. Yields of Volatile Nitrosamines from Drugs and Nitritea)

a) The concentrations of drugs and nitrite were 10 and 40mM, respectively. Reaction mixtures were incubated at 37•‹, pH3.0 for 1 and 4hr. b) Percentage with respect to the concentration of drugs (10mM) c) Not detected, <0.001%.

248 Gann NITROSAMINES FROM DRUGS AND NITRITE high under the mild conditions used for the NDMA or other nitroso compounds as con present experiment, and the yields of NDMA taminants.4,7)Pure aminopyrine contained a from oxytetracycline (0.4-1.3%), NMP from trace of NDMA (0.002%) but no nitrosamine thioproperazine (0.3-1.2%) and NHXI from was detected in the other drugs. tolazamide (0.2-1.2%) were also noteworthy. Typical secondary amines, dimethylamine However, the other drugs only gave volatile and diethylamine, were incubated with nitrosamines in the yields of less than 1% nitrite under the same conditions as used for after incubation for 4hr. the drugs, and the yields of nitrosamines No other notable peak in addition to those (Table II) were compared with those from of the nitrosamines listed in Table I was de several drugs. The yields of NDEA from tectedby GC-TEA. However, no attempt etamsylate (Table I), which is a salt of di was made to identify possible non-volatile ethylamine,were far lower than those from nitrosamines formed by cleavage of one of diethylamine. The yields of NDMA from the small alkyl groups. aminopyrine and minocycline exceeded In every case (except NPIP from tolaza those from dimethylamine. Since minocy mide),the volatile nitrosamine formed was clinehas two dimethylamino groups in the that which was expected from each tertiary molecule (Fig. 1), the yields of NDMA from or secondary (etamsylate) amine (Fig. 1). minocycline shown in Table I might have The formation of NHXI from tolazamide to be recalculated for this comparison. The was expected on the basis of the chemical yields in Table I corresponded to 5.4 and structure, and has been reported.8) However, 5.7% yields based on the dimethylamino tolazamide gave rise to NPIP in addition to groups (corresponding to 20mM) for 1 and NHXI, and the yield of NPIP was higher 4hr, respectively. Even if the latter yields than that of NHXI. GC-MS was carried out were adopted, the yields of NDMA from on dichloromethane extracts of the reaction minocycline were higher than those from mixtures of tolazamide with nitrite, and the dimethylamine. mass spectra of the two products coincided The inhibitory effect of ascorbic acid on with those of authentic NPIP and NHXI, nitrosation was examined with 9 drugs, respectively. The concentrations of NPIP promethazine, oxytetracycline, minocycline, and NHXI determined by selected ion mon thioproperazine, etamsylate, glyclopyramide, itoring(M+; 114.4 for NPIP, 128.5 for thiethylperazine, tolazamide and aminopy NHXI) agreed with those obtained by rine,which gave volatile nitrosamines in GC-TEA. relatively high yields. Pure drugs were used Drugs, both pure substances and formula for this purpose. The results are shown in tions,used for the present experiments were Table III. Ascorbic acid was effective in analyzed for volatile nitrosamines, since it reducing nitrosamine formation from all 9 was reported that some drugs contained drugs, although differences in the extent of

Table II. Yields of Nitrosamines from Typical Secondary Amines, Dimethylamine and Diethylamine, and Nitritea)

a) Amines (10 or 20mM) were incubated with nitrite (40mM) at 37•‹, pH3.0, for 1 and 4hr. b) Percentage with respect to the concentration of amines (10 or 20mM).

75(3) 1984 249 A. SAKAI, ET AL.

Table III. Effect of Ascorbic Acid on the Formation of Nitrosaminesa)

a) Pure drugs were used. The concentrations of drugs, ascorbic acid and nitrite were 10, 40 and 40mM, respectively. Reaction mixtures were incubated at 37•‹, pH3.0 for 4hr.b

) Percentage with respect to the concentration of drugs (10mM). blocking were noticed depending on the drugs thioproperazine, etamsylate and thiethylper nitrosated or on the nitrosamines formed. azine.However, the possibility that the ad ditivesin the formulations contain inhibitors DISCUSSION or accelerators of nitrosation must be con sideredwhenever formulations are used for There was no significant difference be nitrosation reactions in place of pure drugs.13, tweena pure drug and its formulation in the 22) yield of volatile nitrosamine by reaction with Although etamsylate is a salt of diethyl nitrite except in the cases of tolazamide and amineand 2,5-dihydroxybenzenesulfonic acid glyclopyramide (Table I). The formulations (Fig. 1), the yields of NDEA from etamsylate of tolazamide and glyclopyramide gave ni and nitrite (Table I) were very much lower trosaminesin higher yields than the pure than those from diethylamine hydrochloride drugs did. This may be because tolazamide and nitrite (Table II) under the same con and glyclopyramide are scarcely soluble in ditions.2,5-Dihydroxybenzenesulfonate moi water (the pure substances were hydro etymay suppressively affect the formation of phobicand floated on the surface of reac NDEA from diethylamine moiety and ni tionsolutions), while the formulation ren trite. deredthese substances readily dispersable in Rapid NDMA formation from amino the reaction solutions thanks to the use of pyrineand nitrite is considered to proceed additives. In the cases of both drugs, the re not through dimethylamine but through a actionsusing the formulations may reflect pathway specific to aminopyrine,12) though the situation in the stomach. Therefore, it many other tertiary amines10,20) including seems to be possible to use drug formulations tetracycline antibiotics17) have been regarded in order to estimate the yields of volatile as giving rise to volatile nitrosamines by nitrosamines from the drugs and nitrite reaction with nitrite via the corresponding when pure substances are unavailable. Drug secondary amines which are produced by additives appear not to affect markedly the nitrosative dealkylation of tertiary amines nitrosation of drugs, at least in the formula in mildly acid medium. In the present ex tionsof chlorpromazine, oxytetracycline, periment,however, it was found that the

250 Gann NITROSAMINES FROM DRUGS AND NITRITE yields of NDMA from minocycline were This finding is significant in the safety more than those from dimethylamine, even evaluation of drugs, especially in relation to if the presence of two dimethylamino groups possible long-term administration. However, in the molecule of minocycline were con the yields of volatile nitrosamines from the sidered.This result suggests that there is drugs tested were lower (those from most of another pathway not involving the secondary the drugs were much lower) than that from amine as an intermediate in the formation aminopyrine. Further studies are required of NDMA from minocycline. If dimethyl to establish clearly whether or not these amineis an intermediate, the yield of NDMA drugs contribute significantly to cancer risk from minocycline could never exceed that in humans through the formation of volatile from dimethylamine, even if the rate of nitrosamines. Further clinical and epide nitrosative dealkylation of minocycline were - miologicalevidence should aid the objec infinite for both its dimethylamino groups. tiveassessment of the drugs. Not only the formulation of tolazamide but also the pure drug gave NPIP by reaction ACKNOWLEDGMENTS with nitrite, showing that NPIP formation The authors wish to thank Dr. I. Tomita of was not owing to a contaminant in the for Shizuoka College of Pharmacy for encouragement mulation.However, the mechanism of NPIP and useful discussions. This work was supported in part by a Grant-in-Aid for Cancer Research formation from tolazamide remains unclear (55-4) from the Ministry of Health and Welfare. so far. There is no evidence that the tolaza (Received Sept. 17, 1983/Accepted Jan. 26, 1984) mideused for the present experiment con tainedan analogue having a piperidino group in place of the hexamethyleneimino group, though 1H and 13C NMR analyses were carried out. Since tolazamide is an REFERENCES oral hypoglycemic and is chronically taken by diabetics at quite high dose levels (100- 1) Archer, M. C., Tannenbaum, S. R., Fan, T.-Y. and Weisman, M. Reaction of ni

300mg/day), further studies are considered tritewith ascorbate and its relation to ni to be essential. The ingestion of ascorbic trosamineformation. J. Natl. Cancer Inst., acid together with tolazamide might reduce 54, 1203-1205 (1975). the potential hazard of nitrosamine forma 2) Archer, M. C. and Wishnok, J. S. Quan titativeaspects of human exposure to ni tionin vivo, because ascorbic acid can effec trosamines.Food Cosmet Toxicol., 15, 233- tivelyinhibit the formation of NPIP, which 235 (1977). is the predominant product in the nitrosa 3) Crosby, N. T., Foreman, J. K., Palframan, tionof tolazamide (Table III). J. F. and Sawyer, R. Estimation of steam - volatile N-nitrosamines in foods at the 1 The possible risk arising from long-term ƒÊ g/kg level. Nature (London), 238, 342-343

administration of other readily nitrosatable (1972). drugs might also be decreased by the simul 4) Eisenbrand, G., Spiegelhalder, B., Kann, taneousingestion of ascorbic acid. As shown J., Klein, R. and Preussmann, R. Carci in Table III, ascorbic acid inhibited the nogenicN-nitrosodimethylamine as a con taminationin drugs containing 4-dimethyl formation of nitrosamines from all the drugs - amino-2,3-dimethyl-1-phenyl-3-pyrazolin- tested, though the blocking effects varied 5-one (amidopyrine, aminophenazone). Arz between 88 and 100%. neim.-Forsch. 29 (II), 867-869 (1979). In the present study it was found that 24 5) Fine, D. H. and Rounbehler, D. P. Trace analysis of volatile N-nitroso compounds by of the 28 drugs examined gave rise to carci combined gas chromatography and thermal nogenicvolatile nitrosamines by reaction energy analysis. J. Chromatogr., 109, 271- with nitrite under physiological conditions. 279 (1975).

75(3) 1984 251 A. SAKAI, ET AL.

6) Harada, M., Ishiwata, H., Nakamura, Y., 15) Panalaks, T., Iyengar, J. R., Donaldson, Tanimura, A. and Ishidate, M. Studies on B. A., Miles, W. F. and Sen, N. P. Further in vivo formation of nitroso compounds (I). survey of cured meat products for volatile Changes of nitrite and nitrate concentra N-nitrosamines. J. Assoc. Offic. Anal. Chem., tionsin human saliva after ingestion of salted 57, 806-812 (1974). Chinese cabbage. J. Food Hyg. Soc. Jpn., 16, 16) Rao, G. S. and Krishna, G. Drug-nitrite 11-18 (1975). interactions: formation of N-nitroso, C-ni 7) Krull, I. S., Goff, U., Silvergleid, A. and troso,and nitro compounds from sodium Fine, D. H. N-Nitroso compound contam nitrite and various drugs under physiological inantsin prescription and nonprescription conditions. J. Pharm. Sci., 64, 1579-1581 drugs. Arzneim.-Forsch. 29 (II), 870-874 (1975). (1979). 17) Roper, H. and Heyns, K. Possible nitro 8) Lijinsky, W. Reaction of drugs with ni sodimethylamineformation in comparative trousacid as a source of carcinogenic ni in vitro nitrosation experiments with six dif trosamines.Cancer Res., 34, 255-258 (1974). ferenttetracycline antibiotics. In "Environ 9) Lijinsky, W., Conrad, E and Van de Bogart, mentalAspects of N-Nitroso Compounds," R. Carcinogenic nitrosamines formed by ed. E. A. Walker, M. Castegnaro, L. Griciute drug/nitrite interactions. Nature (London), and R. E. Lyle, IARC Scientific Publications 239, 165-167 (1972). No 19, pp. 219-237 (1978). International 10) Lijinsky, W., Keefer, L., Conrad, E. and Agency for Research on Cancer, Lyon. Van de Bogart, R. Nitrosation of tertiary 18) Sakai, A. and Tanimura, A. N-Nitroso amines and some biologic implications. J. compounds formed by the reaction of sul Natl. CancerInst., 49, 1239-1249 (1972). pyrinewith nitrite. Chem. Pharm. Bull. 28, 11) Lijinsky, W. and Reuber, M. D. Compara 1820-1826 (1980). tivecarcinogenesis by some aliphatic ni 19) Sander, J., Schweinsberg, F. and Menz, trosaminesin Fischer rats. Cancer Lett., 14, H. -P. Untersuchungen uber die Entste 297-302 (1981). - hung cancerogener Nitrosamine im Magen. 12) Mirvish, S. S., Gold, B., Eagen, M. and Ar Hoppe-Seyler's Z. Physiol. Chem., 349, 1691- nold,S. Kinetics of the nitrosation of ami 1697 (1968). nopyrineto give dimethylnitrosamine. Z. 20) Smith, P. A. S. and Loeppky, R. N. Ni Krebsforsch.,82, 259-268 (1974). trosativecleavage of tertiary amines. J. 13) Mirvish, S. S., Wallcave, L., Eagen, M. and Am. Chem. Soc., 89, 1147-1157 (1967). Shubik, P. Ascorbate-nitrite reaction: pos 21) Tannenbaum, S. R., Sinskey, A. J., Weis siblemeans of blocking the formation of man,M. and Bishop, W. Nitrite in human carcinogenic N-nitroso compounds. Science, saliva. Its possible relationship to nitrosa 177, 65-68 (1972). mineformation. J. Natl. Cancer Inst., 53, 14) Ohshima, H. and Bartsch, H. Quantita 79-84 (1974). tiveestimation of endogenous nitrosation 22) World Health Organization. Nitrosatable in humans by monitoring N-nitroso-proline drugs-an assessment of the risks. Drug excreted in the urine. Cancer Res., 41, 3658- Inf., 1978, No. 2, 4-7 (1978). 3662 (1981).

252 Gann