; JAN-10-1995 09:39 FROM.. •»/ offtarRadicals inBiology 4 Mt&unt. Vol :. pp. 3??-5*1, I*S6

.•A u, $•' C'SA, All nghis merved.

TO

t ,11

STABY P.02

074|.$5U/$6 WOO+.00C 1987 fafamOft JoufMl* Ltd

Hypothesis Paper

MECHANISM OF ANTIBACTERIAL ACTION: ELECTRON TRANSFER<S AND OXY RADICALS

A

/^s |^ James R.Iam^J Michael DJiRvAK/t and Peter{Kqvacic**^ DeAanwanf ChemUgy Urivtww^ Witcoiyp^ilirattUc,. Milwaukee, NkUwowi* 53201. USA, tad tDittinawiff

' tXrliwtho•Mirqyitu Unhnwit), Mil4a*lmi Wiwwikfl 63flflii UB*1 --^

Aft^iicir»Mfta.g£^y main eatffgong? of bacjerjcjflftiagents, namely, aliphatic and heterocyclic oitro compounds,metal flenvauves and chelators, quinones, azo dyes, and iminium-type ions, are proposed to exert their actioni?J^ nanism. The toxic effect » believed to rauit generally fom the catalvric mpducdon of reac-«ve oxygen radicals duKjiinaJly ame^olecffcfltransfer. Cyclic vohammetry was performed on anumber oftoese agents. Reductions were for the most part reversible, with potentials in the favorable range of -0.2010 — u>*9 ».

Keywords—Antibacterial action. Nitro heterocycles, a-Haloniiro compounds, Metal derivatives and chelators.Quinones, A20 dyes. Iminium ions

INTRODUCTION

radical generation and electron transfer (ET) phe-sna are being increasingly implicated aTiiiechan-pathways for avariety of xenobiotics and inhumanjses.1"' In our laboratory, application of this gen-iheme has been made to carcinogens,10 anticancers.u-u antimalarials,15 benzodiazepines,1' phen-dine (PCP)," nicotine," spermine.17 l-methyl-4-yl-1.2.3,6-tetrahydropiTidine (MPTP),11 me-.ics,19 and heterocyclic di-.v-oxides.20'2' Therear to be several main classes ofCT agents: quinone. and precursors, metal (for example, Cu and Fe)'lexes, ArNOj, and iminium salts.22 According tonifying hypothesis, the. ultimate form ofthe druglively binds to areceptor, for example, bacteria)..and effects catalytic electron transfer tooxygen,jcing toxic oxy radicals that destroy the cell. Al-tively, there may be interference with normalron transport chains.ie natural defense against invading microorga-»involves the use ofreactive oxygen species gen-d by neutrophilic granulocytes during j>hagocy~a"a The rapid uptake of oxygen results in the

.i correspondence to: Peter Kovaeie. Department of Chem-^Rivcnnyof Wisconsin-Milwaukee. Milwaukee. WI53201,

yn

formation of superoxide, hydrogen peroxide, and hydroxy) radicals. Other species that may be involvedare HOC! and singlet oxygen. HypochJorous acid canreact readily with cellular constituents to generate labile A'-chloro entities. It is quite significant thai thisscenario represents the normal response that has beenfound by evolutionary development to be most effective. Similar operation by various antibacterial drugswould comprise a reasonable working hypothesis.

A number of peroxides are reported to display antibacterial activity. Evidence shows thatexogenous hydrogen peroxide is able to destroy bacteria.23-26 Therewas a marked increase in killing efficiency, presumably because of the Fenton reaction, when the ironcontent of the system increased. Generation of oxyradicals may also account for otherbiological effectsof this peroxide, for example, carcinogenic,10 mutagenic,27 and anticancer." ~H Ethyl hydroperoxide, whichis not decomposed by catalase, was more effective thanthe parent as a bactericide against Escherichia coll*Cyclic peroxides can act as antibacterial,29 antimalarial,18 and antifungal10 agents. E$R studies with somemembers demonstrated the ability to function as precursorsof hydroxyl radicals.31 Peracetic acid has beenconsidered for use as a stetilam.12 The diverse peroxides can serve as precursorsof oxy radicals, similarto the phagocytic process.

•j r n i x «_>

378

Arelated category consists of ^-chloro derivativesof sulfonamides, qyclic imides, and amidines. Familiar examples are1 chloramine T. dichloramine T, andbalazone. Although it is known that chlorine can betransferred to anitrogen-containing cellular receptor,the precise mechanism ofbiorction is unresolved. TOesesubstances represent analogues of the N-chloro denv-atives generated during phagocytosis.

Although individual groups of antibacterial agentshave been rationalized mechanistically,32 there has beenno report of abroad, detailed, systematic approach.Ouite some years ago, suggestions were made that chargetransfer and oxidative stress might well play arole mthe action of some medicinals""38 We now proposeaunifying theme based gn the ET-oxy radical conceptmat"eTWJmpiMe$Tiarge variety of bactericides, including metal derivatives ofmercury (merbromin, mer-thioiate). metal chelators (nalidixic acid and oxine),

' nitroheterocycles (nitrofurazone, nitrofurantoin), m-troaliphatics (chloropicrin. bronopol). qumones (ha -ogenated o-naphthoquinones). azo dyes (scarlet red),aid iminium species (triarylmethane dyes, heterocyclicdi-tf-oxides, and heterocyclic salts). These representalmost all ofthe major categories discussed inBurger sMedicinal Chemistry* Our experimental getlwM«odetermine the reduction potential and reversibility forthe various classes in order to obtain evidence concerning the feasibility of catalytic electron transfer invivo In some cases, literature data provide the requisite base for incorporation within the theoreticalframework. Related considerations are also discussed.

MATERIALS AND METHODS

Materials

Chemicals and antibacterial «^«W|J*™^from the Aldrich Chemical Company CJ*lwaukee, W»or the Sigma Chemical Company (St. Uj». MO),except CuCU • 2H30 (MCB, Norwood. OH). Fed,(Baker and Adamson, Morristown, NJ), .f»W*™(Eastman Kodak, Rochester. NY), and silver sulfadi-azlne (Polysciences, w*^\%£?^ £naphthoquinone was m^W*W*™»9published methods" (m.p. ^f^^JJ*;167°C). The infrared spectrum (Perkm Elmer model735B) displayed the reported peaks. Elemental analysisfor CoHfBrOj was as follows: calculated: C, 50.7; H,2.1; found: C, 50.4; H, 2.2.

The electrode used in the nonbuffered electrochemical studies was tetraethylammonium perchlorate(0 1M) (G. F. Smith. Columbus. OH). Diroethylfor-mamide (DMF) (Aldrich) was obtained in the highestavailable purity. Absolute ethanol (U.S. Industrial,Tuscola, ID was used to prepare the aqueous solutions;

J. R. Ames. M. D. Ryas. and P. KovaCIC

j\a--

pH 70 aqueous buffer (0.1 MKHaPO</0.1 MNaOH)was used for some measurements.

Methods

The cyclic voltammetric measurements were performed at ambient temperature with aWwcioft Applied Research Corporation (PARC) model l^Apo-fa ographic analyzer connected to aHewlett Packardmodel 7035B X-Y recorder. The scan rates generallyranged from 20 to 200 mV/s. Solutions were saturatedfor 15 min with prepurified nitrogen (30 min for metalcomplexes) that was passed through an oxygen-scntb-bing system. The electrodes consisted of either aplatinum flag (Sargent-Welch, Skokie. IL) or ahangingmercury drop (HMDE) working electrode, with aplatinum wire as the counter. The reference was asaturatedcalomel electrode (SCE) (Corning). Observed poten-tials (our work and literature values) were convertedto the normal hydrogen electrode (NHE) by the addition of 0.24 V to the SCE values. The reported dataare the average oftwo or more measurements involving fresh solutions. Thefollowine equations were usedf^th^f-wavc potentials, ^differences in pogfrtials, and current function: £,* » (£* + M,2«

RESULTS AND DISCUSSION

Kitro compounds1 Nitro heterocxcles. This chemotherapeutic cat

egory encompasses mainly nitro derivatives of furanand imidazole.32 Specific examples ofthe furan classinclude nitrofurazone, 1. and nitrofurantoin. 2.

Electrochemical studies were performed on both land 2. The results are presented in Table 1. Compound1undergoes aone-electron reduction (CF ratio 0.88;benzil45 as reference) with £,* - -0.67 V, in agreement with the literature (Table 1). Reduction to aradical anion is in agreement with prior studies ofoiirofurans in aprotic solvent.17 The reduction was quasi-reversible and diffusion controlled, as characterizedbythe A£, values. 70-90 mV, and the constant currentfuoction. Tht nitro anion formed is quite stable withan ULc value of 0.97 at 100 mV/s sweep rate. Nitrofurantoin, 2. reduced in two waves; the differencein cathodic peak potentials was 110 mV at the sweep

JULCH=NNHCONH2

(1)

01/10/95 07:57JAN-10-1995 09:40 FROM TO STABY P.04

Mechanism of antibacienal action 379

,nAoJLch-n-i/~|/i— NH

o*(2)

ate of 0.1 V/s. The first wave. Ept « -0.66 V, wasrTeversible. with E„* of 80 mV. The CF ratio (0.98)ndicatesthe transferofone electron.The current func-on decreased as the scan rate was lowered, suggestingaat reduction was followed by another chemical re-ction. The second wave was quasi-reversible with a0-mV change in EPi upon adecade change in sweep4te (-0.74 V at 20 mV/s). The Ey: was -0.73 Vith A£, of 80mV.These reductions may involve both•tro and conjugated imine. The value for the one-ectron reduction of 2-nitrofuran in DMF is -0.7617 The potentials in our study are more positive be

muse of increased conjugation.Reductions of 1 and 2 have alsobeenexamined in

jffered media. Nitrofurazone exhibited two irrever-Dle waves (-0.23 and -0.97 V) at pH 8.6.4? Theocess involved thereduction of bothnitroand imineoieties. Another investigation*4 resulted in values of0.04 and -0.02 V for 1 and 2, respectively. Anj-lier study reported multiple waves for 2 in aqueousffer.45 The proposed pathway involved reductive fis-

•>n of the N-N bond. The prior data are difficult toropare to ours because of differences in conditions,

.- example, solvent.The mechanism of action for this general class hasen reviewed recently.5 4e"4» A correlation isobserved•ween electron affinity and radio-sensitization, hylic cell toxicity, and chronic aerobic toxicity, sug

gesting that redox processesplay an importantrole. Inan aerobic environment the intermediate nitro radicalanion readily conveys an electron to oxygen. For example, nitrofurantoin stimulated the uptake of oxygenin hepatic incubations, which was partially reversedby superoxide dismutase (SOD) andcatalase; this suggested the presence of superoxide and H202.49 Increased oxygen consumption has also been observedfor nitrofurazone. Activated oxygen may also be re-sponsible for toxic effects, such as pulmonary edemaand fibrosis, that may occur during nitrofurantoin therapy.** In other cases, including anaerobic conditions,the radical anion exerts its effect by another route,presumably interference with normal electron transport. A serious problem with these types is acute toxicity, usually associated with mutagenesis or carcinogenesis. One can speculate that both the favorableeffect and the liabilities might be intimately related tothe ease of electron transfer. Prior reviews have discussed ESR of the intermediate nitro radical anions.549

2. O'Halonitro aliphatics. The common structuraltheme is the presence of the a-halonitro moiety alongwith two other electron withdrawing substituents.32Members are exemplified by bronopol, 3, and chloro-picrin, 4. Compound 4 is a sterilant.

Bronopol underwent irreversible reduction with themost positive peak at £r « -O.ScTV (Table n. Theplot of the peak current versus the square root of thesweep rate was linear and passed through the origin.EPF s values varied from 180 to 260 mV for the scanrates employed. The£, changed 190mV withatenfoldincrease insweep rate. TheCFratio ofbronopol versuspenal as the reference gBve avalue of 0.89. indicatingthat only one electron was transferred. The absence ofan oxidation peak is rationalized by the participationof a chemical reaction after initial reduction, for ex-

Table 1. Electrochemical Characteristics ofNitro Derivatives ofHeterocyclic and Aliphatic Compounds

Substrate

2'Nitxofuttn

1"

V2"2'«

3>3

4»

2rBromo-2-jutioprDpane2'Chloio-2-njffopropan-J ,3'diolNttrocthane

RedactionPotential

(V)

-0.6?'

-0 68•0.66'

-0.73'-0.7o•0.56'•0.10

-0.7I'

-0.4

•0.54

•1.2S

im'L

0.97

CF ratio* Reference

0.88 _

0.98—

— _

0.8937

r*>37

3839

40

....y m* 7?/ CF***-: c?nu. • * *V *>' lC = 16.27 (0.5 mM. h. DMF). 0.185 (0.5 mM HMDE. DMF.. MOO mV/i. fctrKthvfanmoaurcperchjoraie (0-1 Mj. subsist* (0.5 mM,. DMF. vmui SHE. Pi ekatoa*. «fcevmible «Ref 41. •Futi »»ve. Ontvenibte. 'Second wave.

01/10/95

JAN-10-1995

07:58

09:41 FROM TO STfiBY P. 05

380J. R. AMES. M. D. Ryan. *nd P. KovaOC

N02I

HOCH2CCH2OH

Br

(3)

ample, dimerization. The reduction of 3 is similar tothat of 2-bromo-2-nitropropane inCHjCN, Ep of about-0.4 V.11 The reaction inyolved one-electron reduc-tionwith subsequent loss of bromide ion. The ensuingproducts that m%% from hydrogen abstraction or dimerization might function as stable ET catalysts. In aprior study, compound 3 in aqueous solution gave irreversible redaction at - 0.10V,w Electrolysis yielded2*nitropropane-l ,3-diol. which has also been observedas ametabolite of bronopol in rats and dogs.30

Chloropicrin exhibited irreversible reduction withEp of - 0.71 V(Table 1). Alinear plot passing throughthe origin was obtained for the peak current versus thesquare root of the sweep rate. The E^* value at 100mV/i was 230 mV. The reduction potential changed110 mV upon a tenfold increase in sweep rate. Comparison ofthe current function with benzil gave avalueof 1.99, indicating the transfer of two electrons. Thenumber of electrons involved in the reaction of 4 isdifficult to explain on the basis of the discussion for3. However, a possible rationale entails formation ofa carbenc intermediate. Such a process was observedfor the two-electron reduction of gem-dihalides inaproticmedia." Chloropicrin undergoes two-electron reduc*tion in aqueous buffer.52 yielding an £,..2 value of +0.30V that is invariant with pH.

The literature contains an appreciable amount ofwork on the electroreduction of aliphatic nitro compounds. The;£,2 values range from about - 1.33 Vfor nitroethane. 1-nitropropane, and 1-nitrobutane40to -1.40 V for 2-nitropropane*> in DMF. The E,.j(- 0.64 V) for nitromethane isthe same as for 1-nitropropane in pH 7.0 buffer.54 Electron-withdrawinggroups, such as those in 3and 4.should make the reduction potential more positive.

In relation to more detailed mechanistic considerations, the o-halonitro-alkanes readily accept an electron toyield aradical anion that eliminates halide.MMThe resulting delocalized radical could dimerize orcombine with oxygen in a process leading to variousreactive intermediates. Alternatively, uptake of another electron would generate thecorresponding anion.

CI3CNO2

(4)

which might abstract aproton or eliminate aremaininga-halogen.

Significant to our theme is the.cojim*iisoji-Of re-^^"jTrtrnifr'* «"ifr activity for bronopol and itsanalogues. For example, the chloro. counterpart of 3

"exhibits alarjc£injnimumJn|iifritory concentration to-ward bacteria than bronopol;36 it is also more difficulttoreduce,lv-0.MV versus - 0.10V for 3.2-Bromo-2-nitropropene (£„* ~ -0.4 V) and 2-chloro-2-nitro-propane (£,* ~ -0.96 V) show asimilar relationshipbetween £„a* and activity.1* deduction potential mayalsopM nmk iiuhc •fflimtffliM ftctivtafteqb-sttoed S-nit"^^ ^.ffoxanes (analogues of 3). Re-

"placement ofbromine by chlorine or hydrogen resultedin marked reduction in activity.57 Additional evidenceis available that sheds light on the mode ofaction. The \biocirfai effects of bronopol are inhibited by thiols." 1Our ET-oxy_radical hypothesis is in accord with thisobservation. It is significant that prior investigatorshave also proposed involvement of an oxidative mechanism** However, their suggestion for thelethal actionentailed oxidative conversion of enzymatic mercaptogroups to disulfides. Inhalation of chloropicrin pro*duced edema and lesions in the lungs of exposed an-imals."-5' The toxicity toward Sitophilus granarius andTenebroides mauritanicus was potentiated by oxygen.40 Compound 4 is an indirect, as well as aweakdirect-acting, mutagen.61

Metal-containing drugs

Metal species are known to elicit avariety of physiological responses. Specific chemical reactions pertinent toour approach that have been observed includeoxy radical formation*-62 and DNA strand cleavage65The formation of complexes with DNA has been reported in some cases."^ Much of the prior work isrelated to the cancer10 and anticancer areas.11 Of thevarious heavy metals employed over the years, mercury and silver remain the only ones widely used inthetreatment orprevention ofmicrobial disease."Thederivatives of heavy metalsexaminedin this studyinclude thiomersol (5, merthiolate), merbromin (6, mer-curochrome), and silver sulfadiazine. 7.

Compound 5 reduced with E^ « —0.50 V in aprocess that was 44% reversible (tgjig, = 0.44) at 100mV/s in pH 7 aqueous buffer (Table 2). The A£, was

HgC2H5

01/10/95 07:59

JAN-10-1995 09:42 FROM TO STABY P.06

Mechanism of antibacterial action 381

HgOH

C02Na

(6")

75 mV. Irreversibility pertained at slower scan rates.Diffusion control was evidenced by the constant CFvalue. Asecond peak was observed at -1.03 V. Upona change of cathode to platinum, no reduction wasobserved in the buffer. The results for 5are in agree-rnent with prior studies.65-6* The first wave varied withpH, -£,, m 0.I8 + 0.05 pH;" the second wave.Efz" -0.96, was invariant with pH. The first steptesulted in the formation ofthe alkylmercury radical,**a'hich is known to arise from the reduction oforgan-omercuric haJides.6'

Merbromin. 6, reduced in two irreversible steps,he first with Ep = -0.54 V (Table 2). The process*as diffusion-controlled. Irreversibility was supponed3>' M &** value of 90-100 mV. Lower diffusion is•xpected for 6 due to its size, which should result iness current being passed. The maximum current forne first peak was approximately one-half that observedor 5. The second peak occurred at Ep = -0.74 V.>rug 6 also gave no reduction in aqueous buffer withiatinum as the working electrode, in line with otherleetrochemical studies involving mercury commands.6* Apparently, merbromin has not been exten-jvely studied electrochemical!)-. Page and Waller66:ported no reduction in 1.0M HC1. The £, * foriuorescein in pH 6.25 buffer is -0.73 V** (noref-rence electrode given); thequinone methide form ex-;ts in neutral alcohol solution.90

Several lines of evidence are in agreement with theiesis that the toxic action ofmercury compounds mayrjse from oxyradical generation via electron transfer,hiomersol caused hemolysis, which was decreased by

A9

<7j

Table 2. Reduction Potential* of Metal Derivatives'

Reduction

PotentialSubstrate Electrode (V) i+lir

3* HMDE -0.50* 0.445* Pi —.« ^m

e HMDE -0-54* -

e Pi __< ^^

T HMDE r m^

Silver protein* HMDE —* ^^

8 • CuClj* * -0.23* .

8 • FeCtf Pt -0.21' -1

'100mV/s.ttvaedylamooftka pnakmie (0.1 M.eaciatis buffer), nb-an* (0.5 mM). ntsui NHE.

•pH 7.0 buffo- (0.1 M KftPOyO.1 M NaOH).

V© tedueoae beibtc backfrauad.Inevatiblc-'uuolitbk.*so%aouMuieeMDai.

DMSO.71 a well-known oxy radical scavenger.'1 Anumber of fluorescent dyes act as hemolysins whenirradiated with UV light or in darkness when presentin high concentrations, or in combination with H202.73Oxidative stress isbelieved to be responsible forhem*olysis byvarious antimalarial agents.,s Thiomersol wasobserved toincrease oxygen uptake inTrichophyton.1*Organic mercury compounds are known to be metabolized to inorganic derivatives,74 Mercuric chloridecauses extensive DNA breakage, presumably by anoxygen-stress mechanism.'6*" Lipid74 and membrane7*oxidation have been observed. Several common antioxidants, including Se and vitamin E, reduce the ef-

lects of Hg intoxication.7jL ~Prior rationale for the modeof action has involved

the binding of mercury derivatives to the thiols of cellular proteins.31-7* Mercury may form complexes withDNA.64The crystal structure ofan unusual methyl mer-cury-adenine complex has been published recently.10

We were unable to obtain data for compound 7and>Silver proteinbecausc of insolubility, which irnotluTTprising since the drugs may be in a colloidal state. Invivo, metabolism could presumably take place, resulting in a form that can bind to DNA.*4 Ingestion ofsilver ion or colloidal silver produces liver necrosis inrats deficient in vitamin E.7i

Metal chelators

1. Quinolones. Nalidixic acid, 8. the most prominent member in this class, has received considerableattention.32 It is used in the treatment ofurinary tractinfections and is particularly active against gram-neg-ative organisms. Several reviews deal with structureactivity relationship (SAR) and mode of action. }2fi,«Various mechanisms have been proposed; recent at-

[

3s:J. R Ames, M. D Ryan, and P. Kovacic

c2hs

f8)

tention has been focused on coordination with metals.The conclusion was drawn that Cu(II) and/or Fe(II)are involved in the in vivo activity .••

Since the bactericidal activity may be associatedwith the formation of a 1:1 (8:metal) complex, weperformed electrochemical studies on solutions of thesodium salt ofnalidixic acid and CuCfe •2H30 or FtO,at the 1:1 molar ratio in 50% aqueous ethanol. Theresults are summarized in Table 2. Sodium nahdixateproduced no reduction before - 0.8 V. Cupric chloridegave two irreversible reduction peaks, at +0.13 and- 002 V. With equimolar concentrations ofmetal and8(Na salt), abroad irreversible peak with Ep of - 0.23V and asmall peak at -0.01 V were observed. It islikely that the minor peak is due to uncomplexed copper. The EP for the complex was -0.08 Vat asweeprate at 20 mV/s, while at 200 mV/s it was -0.30 VFerric chloride gave irreversible reduction. £,« -0.5?V, with an anodic wave at +0.16 V. Asmall cathodicpeak was present at -^0.06 V with acurrent of about25* ofthat at - 0.59 V. The 1:1 solution ofiron and8 (Na salt) resulted in a quasi-reversible reductionat -0.21 V with A£,of about 190 mV. The peak atabout +0.1 V was also observed. The i^ti* ratio wasabout 1at 100 mV/s, indicating the formation of astable reduction product.

Evidence that ametal chelate of8may act as apotentET agent was obtained from in vitro studies, indicatingan increase in the ET rate from FeSO. to cytochromec •' An intriguing relationship exists between bactericidal effect and concentration." The activity is decreased by increasing levels ofthe drug, similar to theoxine case. A possible rationale is that some uncomplexed positions must be available on the metal to permit coodination with the active site, for example, DNA.Single-strand scission in Dl^A has been observed inthe presence of 8M These breaks are commonly associated with the formation of oxy radicals.3'7 How-ever, the potency toward Plasmodium falciparum, amalarial parasite sensitive to oxygen tension, was associated with low oxygen concentration.6 Alternatively, activity may derive from the free form of 8.

Recently, its one-electron reduction potential (-0.87V at pH 6.4) was ^eported,,*

Other mechanisms proposed for the action of 8include inhibition ofRNA synthesis. DNA intercala-tion. and inhibition of DNA gyrase. However, it hasbeen observed that an analogue of8 does not bind togyrase, but rather to DNA." There is selective bindingin vitro to guanine ofDNA in the presence of astoichiometric equivalent of the metal ion.

Another member is oxalinic acid, 9, which displaysan antibacterial spectrum similar to thatof8, but whichis 2-4 times as potent." Since metabolic studies revealconversion to the corresponding catechol derivative, itis reasonable to expect subsequent facile oxidation tothe o-quinone." Hence, in vivo transformations of9could result in twoelectroactive sites, namely quinoncand metal complex.

2 8-Hydrcxyquinoline (oxine). The 8-hydroxyqui-nolines (antibacterial and antifungal)" can be includedin this division, since they readily form metal com-plexes." Of the various quinolinols, the 8-isomer isuppermost in its ability to chelate biochemically important metals. The physiological effects have beenattributed tothe generation of such coordination compounds. Iron and copper are known to play importantroles with both endogenous and exogenous compounds, entailing oxy radical formation in somecases.3"•" Significantly, data indicate that either theCud) or Cu(ll) form ofrelated complexes might engageinET.90 A ft0

The Fe(W) 1:1 complex with oxine reduces at •*- u.5-V" while the reduction potential for the 1:2 Cu(II)complex varies with pH, £,, - 0.12 - 0.058 pH-Iron oxine complexes catalyze the oxidation ofthiolsin nucleoproteins,* while the copper complex is toxicto algae in seawater.82 The mechanism may be due tothe formation of activated oxy species.

3 Others. Biguanides can be similarly classified-Cblorhexidine, 10, a widely used topical antiseptic,serves to illustrate." Informative mechanistic studieshave been carried out onthe antimalarial counterparts,

C02H

C,H2^5

(9)

JI-I1T-1U- 133J U3'4J

Mechanism of antibacterial action383

IIjp-CIC6H4NHCNH(CH2) 1(10)

or example, chlorguanide (proguanil, paludrine). Al-jough evidence has beep found indicating the impor-^nce of metabolism, the conclusion was drawn from(her studies that the activity against primate malariamst be attributed in part to the parent drug" In priorlecuioreo'uction investigations, the indicated values weresported: -0.85Vfrom extrapolation to pHO,* -1.38(,P?« °5)" "L38 V(pH604)"aftd*rea<erAan1.8 V in acid." too negative to be operative in vivo.

tguanides coordinate readily with certain metal ions.*7*value of -0.24 V for the reduction potential was

stained by other investigators using the Cu(ll) com-exes," Hence, the general mechanistic scenario out-ned for the other classes could also be applied here.ne related amidines" may also operate in this manner.dditional antibacterial drugs that can chelate are"licylic acid, hexachlorophane, antimycin, 2,2'-bi-ndyl, 1,10-pheoanthroline, dithiocarbamates, andoxides ofpyridine, quinoltoe, or benzoquinoline with-oordinating group in the 2-position,

'inones and phenols

Quioone antibiotics have found widespread appli-«on in recent years in the treatment of malig-

•icy.*61-1*™ Extensive studies have principally in-ved anthracyclines. mitomycins, streptonigrin "andramycins. The toxicity, found to be oxygen-*ndent,,fl° apparently results from redox cycling ofquinone. Initial metabolic reduction to the semi-

aone^ intermediate, which can evidently bind toA,102 is an essential step.101 The overall process hasndesignated site-specific free radical generation.101ibition of the rate of DNA scission was observedh added cataiase, SOD. and free radical scaven-s "" However. Adriamycin bound toDNA isunableparticipate inredox reactions.104 Strand scission canur is the absence of binding. Alternatively, the uJ-ate agent may be a metal complex.11 Electronicctures related to quinones are found in some othercricidal agents,1* for example, triarylmethane dyesmercurochrome.

ecently, a number of halogen-substituted 1.2-athoquinones, 11 (ft=»R"=Br; R'«H; R=C1,-=H, Br) were shown topossess antibacterial activ-* To determine the possible involvement in ET,synthesized a congener, namely, 6-bromo-1.2-

naphthoquinone (bonaphthon), an antiviral agent,w fromthe corresponding 2-naphthol by oxidation with Fre-my's salt, and studied its electrochemistry. Two reduction waves were observed for 11a in DMF (Pt andHMDE). With platinum, the first, EV2 = -0.20 V(Table 3), was quasi-reversible and diffusion-con-trolled. A£p increased from 70 to 110 mV upon adecade increase in the sweep rate (20-200 mV/s).The CF ratio of 11a with berwil,4* acompound knownto undergo one-electron reduction, gave a value of0.97 (Table 3), denoting the formation ofthe semiqui-none. This intermediate shows good stability with an' " value of 1.0 at 100 mV/s. Toe second wave

-0.88 V) was broadened, reduced in height,(Bpcand coupled to asmall anodic peak, £^ «s -0.64 v!This behavior may be due to some type of chemicalreaction after electroreduction,"1 such as proton abstraction from solvent. Very similar results were obtained with HMDE (Table 3).

An Etti of -0.27 V is reported for the parent 1,2-naphthoquinone in DMF.1* The effect of the bromosubstituent istomake electroreduction more favorable.This is in accord with data for 2,3-dihydroxynaphtho-quinone and its 6-bromo derivative.1 wThe £,„ for thelatter is 0.03 V more positive. Halogen in the 3-po-sitioa should make the reduction potential more positive because of the inductive effect. Thus. 2-chJoro-1,4-naphthoquinooe reduces 24mVmore positive thanthe parent10* Bromo substitution exhibits similarinfluences.110

The proposed mechanism ofoxidative stress for theo-naphthoquinone agents 11 is supported by severallines ofevidence from closely related systems: (1) theo-naphthoquinone, pMapachone (trypanocidal and antitumor agent),*-"* shows activity linked to superoxideproduction and (2) activated oxygen species are formedby 1,2-naphthoquinone in rat liver microsomes.1"

Phenols are widely used and represent one of theearliest classes of antimicrobial agents. The members

384i. R. Ames. M. D. Rias, and P. Kovacic

Table 3. Redvction Potentials and Electrochemical Characteristics of Naphthoquinones

Substrate

Ila*lit'

1.2-Naphthoquinpnc1^.Naphthoquinone2-CW3TO-1 .^-naphthoquinone2,3-Dihydroxy-1.4-naphthoqoinone©-BftHn©-2.3«dihydjgxy4 ,4-naphthoqujnone

Electrode

PiHMDE

ReductionPotential

(V).

-0.30"-0.2O*-0.2740.48

+0.50••0.29-40.32

CF wtto*

0.97064

iji*

1.010

Reference

108

1091091)0110

•Set feomotceTable I. ^ _ ftTtir•100 mV/s. tsovetbylaaaMBUBB pochlgme (0-1 M). whstratt (0.5 mM). DMF. venm NWfc.*QuatMranib!e-

include mono- and diols that contain awide variety ofsubstituents.13 A commontransformation of phenols invivo is bydroxyiation to o- and p-diols followed byfacile conversion to quinones." Hence, it is conceivable that these types serve as precursors of quinonesthat function as the ultimate form of the drug via chargetransfer.

Azo dyes and aromatic amines

Antibacterial activity hasbeen observed withanumber of azo dyes,32 including scarlet red, 12, diaceiaz-otol, and phenazopyridine.

Cyclic voltammetry was performed on 12 (scarletred; Sudan IV) in DMF. The results are presented inTable 4. Compound 12 gave several reduction waveswith£„ » -062, - 1.13. and - 1.44 V. and subsequent oxidation waves at £M - —1.31, -0.96, -0.67,and -0.33 V. The first wave. £,: = -0-58 V, isquasi-reversible and diffusion-controlled. A£f rangedfrom 60 to 70 mV (scan rates of 50-200 mV/s) andwas 65 mV at 100 mV/S- The i^i^ value increasedfrom 0.54 to 0.67 for the above scan rates. Slower scanrates (20 mV/s) produced no anodic peak. This indicates a reduction process accompanied by kinetic orother complications, such as fast follow-up chemistry.The CF ratio of 0.90 for 12 (Derail as reference)*5denotes the transfer of oneelectron, with the formation

yCH3 /CH3

(12)

of the azo anion radical, in agreement with previousstudies onvarious azo compounds in aprotic media."*The otherwaves were not examined in detail. Analogous results were observed with HMDE (Table 4) except that adsorption was found at scan rates of50 and20mV/s. Upon the addition of acetic acid, reductionwas observed at E, = -0.37 V, accompanied by severe adsorption.

Other data on the reduction of azo compounds inaprotic solvents have been reported (in CH3CN) (Table4). The comparatively more positive value for 12 isexpected because of increased conjugation. Previousstudies have also been carried outin protic systems."4The £|<3 values in buffered aqueous alcohol (C2HjOHor CHjOH) for azobeniene, p-bisazobenzene, and 2-hydroxyazobenzeoe are presented inTable 4. Substituents on the aromatic nuclei; namely, methyl"9 andhydroxyl,"0 have an adverse effect on reduction asobserved for the reported values.

A mechanism involving redox cycling resulting inoxidative pressure for the azo class is reasonable. It ispostulated that the radical anion formed upon the incubation of sulfonazo m. a diazonaphthol, with hepatic microsomes reacts with oxygen in a catalyticmanner, producing superoxide.4* The azo radical anionintermediate was observed by ESR,

Alternatively, theactive agent may be anaromaticamine generated by enzymatic reduction."6 This typeof transformation has been shown with 12 in rats."7A similar process may be responsible for the mutagenicity of Sudan IV, observed after chemical reduction and microsomal activation.111 The most importantmembers of the aromatic amine class contain a sulfonamide (sulfanilamide) or sulfone substituent on thenucleus in addition to the aminogroups.32"9 There ismuch documentation for the generally accepted viewthatthe modeof action involves inhibition ofdibydro-folate synthesis by competition with p-aminobenzoicacid. On the other hand, one should recognize that anappreciable body ofdata from studies on dapsone (p,p*-

Mechanism of antibacterial action365

Table 4, Electrochemical Characteristic* ofAio compounds

Substrate Electrode

ReductionPotential

(V) CF ratio1 L'L Reference12*121

Axe-benzeneAzobcnzeneJ-Pbenylaxohenseoc2-Methylaiobeiuenep-Bitawbenzene2-Hydioxyatooen2ene

. * -0.5S*HMDE -0 57'

— -1.12— +0.352 - 0.079 pH'— -1.02~ -1.14— +0.405 - 0.059 pH*— +0.29 - 0.Q68 pH'

0.900.82

0610.58

t!4114

114114114

114

**cc feoomc a Table 1.

^^^Y)ma^mVtta^*' MJ' «**•* «« «*>• DMF. <aut MB.**H 1.6-9.0.•pH 4.S.I2.7.•pH 2.0-6.0

Jtam nodtphenylsulfone) points to the generation ofreactive states of oxygen. A number of reports deal*itn in vivo or in vitro conversion to the A'-hydroxyJerivative12™' and possibly to the nitroso form/'several investigations have demonstrated involvementAoxidative phenomena on exposure to dapsone,120-122ncluding generation of superoxide and hvdrogen per-*ide by the hydroxylamino form in vitro.12' Oxygeni required for the cytotoxicity observed during redoxycling with different xenobiotics, including aromaticmmes. The same intermediates generated by theanial oxidation of the amino functionality can alsorise from the reduction of the nitro group.'

•ninium species

There are various subdivisions in this category, such•tnarylmethane dyes, heterocyclic di-A-oxides, andrterocychc salts. The first application of our working.pothesis to the mechanism of drug action involvede heterocyclic di-A'-oxides.20-21

I. Triarylmethane dyes. Examples, all ofwhich in-•rporate the iminium moiety, include gentian violet,>, and brilliant green. Compound 13 is used as aPical^ antibacterial, anthelmintic, and antifungal«m. •The synthetic dyes can produce toxic oxygenscies through redox cycling."2 The metabolism ofntian violet yields aone-electron reduction product,nch was detected by ESR spectroscopy* This de-

(p-Me2NC6H4t-C=

(13)

localized radical reacts with oxygen to form a muchmore reactive pcroxyl type, in 1952 the suggestion wasmade that the pharmaceutical effect of these drugsmay be intimately associated with charge transfer ,24The £,-5 value for 13 is -0.55 V. and for brilliantgreen, -0.42 V. Some of the dyes are carcinogens,and gentian violet damages DNA.« This class, structurally related to quinones, represents one of the bestsupported examples of incorporation into our theoret-ical framework.

2. Heterocyclic di-K-oxides. Our initial investigation, experimentally based on the iminium 14 ET-oxyradical theory22 included quinoxaJines**' and phena-zines,21 Representative members are dioxidine. 15,and iodinin, 16, which display £,„ values of -0 87and -0.34 V (DMF), respectively. Reduction was favored by increased conjugation and intramolecular hydrogen bonding involving Moxide. Reasonable correlations were found to exist involving reductionpotential, structure, and drug activity for about 16 qui-noxalines and 8phenazines. The quinoxaline di-A^ox-ide free radical has been detected by ESR.4* Otherheterocyclic A'-oxides may conceivably fit in thiscategory.I25J2*

3. Acridinium ions. Proflavine and acriflavine aretwoof the more importantrepresentatives; occasionallyquat salts are used* Studies have demonstrated acor-relation between base strength and antibacterial po-

+

=NMe. CI'

386 J. R. Ames, M. D. Ryan, and P. KOvaCIC

N~C—I I(14)

tency."5 The acridinium cations that can arise at physiological pH are the active entities. These salts arenecessary for strong binding withnucleic acid, aswellas for bacteriostatic action. DNA intercalation is evidently associated with the biological effect. The Ev%value for 2,7-Jiaminoacridine is -0.59 V at pH 7."*Quite some time ago the proposal was advanced thatthese substances undergo in vivo reduction to radicalanions that exert their antibacterial effect by interference with the respiratory electron transport chain."*Alternatively, we feel that reactive oxygen-containingentities may be generated, which could contribute tothe toxic effect.

For the antimalarial drugs in this class," namely,the 9-aminoacridine derivatives, oxidative stress andDNA strand cleavage are known tooccur. Oxy radicalformation is postulated to result from charge transferby the heterocyclic iminium species. The reductionpotential (£i/j) of the quinacrine cation is -0.72 V(DMF), which is probably considerably more positivein vivo because of steric inhibition of resonance as aresult of site binding. Kier, whosuggested an ETmechanism, pointed out the favorable energetics associatedwith the protonated form during ET.12'

4. Quinolinium ions. Illustrative of this type isquindecamine, 17, which has been used as a topicalantibacterial and antifungal agent.*2 Clues concerningthe possible involvement of an ET mechanism can beobtained by examining the literature15 on closely related antimalarial agents, for example, chloroquine,18a, and amodiaquine, 18b. The bases exist at physiological pH as cations that are involved inbinding toDNA. Marked deviation of the side chain and nucleusfrom coplanarity is indicated, which should result inashiftof reduction potential to more positive values be

ll*)

cause of steric inhibition of resonance in the cation.Free rotation of the aide chain permits delocalizatiOnofthe positive charge on the heterocyclic nitrogen making for greater difficulty inreduction. Electrochemicalinvestigations have been carried out onalarge numberof quinolinium14 and isoquinoliuium" salts of variedstructure that display anticancer properties. Reductionpotentials fall in the range -0.43 to -1,5 V.

There is evidence for participation of reactive intermediates derived from oxygen for the 4-aminoqui-nolines." Organic intercalating agents of the iminiumtype are known to photodamage DNA, presumably viaactivated oxygen.122 Theribose unit was thought tobethe electron donor.

5. Other heterocyclic quatsalts. Bactericidal properties are exhibited by various compounds of this group,for example. benzo[h)naphthyridinium ions.12' Pyo-cyanine, anatural tf-methylphenazinium antibiotic, isquite susceptible toelectroreduction,130 £1/2 • -0.54V. Mason has reviewed the chemistry of the AT-methyl-phenazinium ion relative to the formation of radicalcations and activated oxygen.4' In the anticancer domain," tf-raethylphenanthridinium salts are known toundergo charge transfer."1 The activity of fused iso-quinolinium salts has been correlated with the presenceof the iminium site.m Binding to DNA could well bean important feature related to the activity of thesetypes.

6. From alkylating agents and DNA. A considerable number of the gaseous cbemosterilams are described asalkylating agents,including ethyleneoxide.

NH(CH2>5-^-

(17)

Mechanism of antibacterial action 38?

NHR

a, R=~CH(CH3)(CH2) NEt2

OH*«•-©-CH2NEt2

(28)

glycidaidehyde, aziridine, and &-propiolactone.i: Jt isgenerally accepted that nonspecific alkylation of nu-cleophilic sites in essential metabolites constitutes themode of action.

The alkylating class contains a large group of carcinogenic agents,1" including those displaying ami-oacterial activity. Concomitant production of oxy radicals has been observed with various members"'•7-'*Although the precise role of these reactive interme-jtates has not been ascertained, it appears that DNA.trand cleavage maybe a crucial event.7-'* In a recentinvestigation of the mechanism of carcinogenesis, aiovel proposal was advanced in which the salt formiminJum, 14) of alkylated nucleic acid was assigned.key function as an ET agent.10 The purines (guanine-fld adenine) of DNA are the principal targets of alack.'" For example, the ionic structure 19. aconju-ated form of 14,is generated from 0-6 alkylation ofuanine and could conceivably undergo one-electroneduction. Electrochemical data from the literature1"nd our own studies'0 are in reasonable accord withne current picture relating siteof alkylation and defectrrsistence to oncogenic response. Thus, it appearsiite plausible that the salt foiro is functioning in a

catalytic manneras a generator of toxic oxy radicals.A similar rationale appliedto the sterilant types wouldincorporate them into the generalized ET-oxy radicalapproach.

Naturally occurring antibiotics

A number of naturally occurring antibiotics92 fallinto thestructural categories thathave beendiscussed,for example, albomycin (Fe chelate), ferrimycin A.(Fe chelate), bleomycin (Fe or Cu chelate), lucenso-mycin (epoxide), oleandomycin (epoxide), chloramphenicol (ArNOa), pyrrolnitrin (ArNOj), rifamycin(quinone), geldamycin (quinone), actinomycin D (im-inoquinone), tetracyclines (possible precursor ofiminoquinone15 or metal chelate*9), streptovaricins (p.quinone monomethane), CC-1065 (alkylating agent),136kojic acid (metal chelate),19 and bacitracin (mew)chelate).19

Other considerations

In our prior discussion, much evidence has beencited for the formation of radical intermediates andinvolvement of activated oxygen species. Superoxideis presumed to be the precursor. Increasing evidencefrom studies involving xenobiotics and the initiationand progression ofvarious diseases implicates free radicals derived from oxygen. '-*•»•»?

According to the theory, thedifferent agents act invivo as ET entities in the production of oxy radicalsor disruption of normal ET. However, some of thereductions were found to be irreversible. Reversibilitymay beaffected bythescan rate employed, as observedin thisand other studies."1 Association with the activesite resulting inimmobilization could prevent reactionswhich otherwise occurin vitro, such as dimerization.thus making ET possible. This aspect has been treatedelsewhere. '4'7'2' Several reports indicate that reductionpotential in vivo may well be betterthan in vitro."4"*Evidence points toarelationship between electrochemical characteristics ofdrugs and physiological activity.For example, antibacterial mitomycins (quinones) possessing less negative £„3 values exhibited more powerful activity.140 A study on the antibacterial activityand electrochemical activity of 9-methylaminoacri-dines and their nitro derivatives revealed a symbaticrelationship.'4' Similar correlations have been notedfor heterocyclic di-tf-oxides* and heterocyclic nitrocompounds.44-46142 Other examples are given elsewhere.,0-'*

The present theory is clearly an oversimplificationof a very complex situation, since absolute correlationbetween electrochemical behavior and physiological

388J. R. AMES. M D Ryan, and P. KovaCIC

CCH3)3N(CH2)15CH3

(20)

Br'

activity is not expected because of the many variablesinvolved in vivo, for example, metabolism, stereochemistry, cell permeability, solubility, site binding,and diffusion. It should be emphasized that the toxiceffects of antibacterial agents may be manifested byavariety of routes. One of the most widespread andgenerally accepted mechanistically is interference withDNA replication or synthesis." For aliphatic quaternary ammonium salts, such as cetrimoniuro bromide,20, and extended-chain imidazoliurn'4* and pyridiniumMl'l$t«.u3 cen membranes and closely associated structures are likely targets." Conceivably a number ofpathways, including oxy radical generation, may actin concert for certain drugs.

Examination of the literature reveals that many ofthe classes discussed here also display activity in othermedicinal areas.™ Furthermore, large fturo*?J*^f*elicit carcinogenic,«^-,44-,4J mutagenic,<•>••*"••"••» -ca^diotoxic,,3, and pesticidai" responses. Since we arestressing aunifying theme, it might well be that theother effects are frequently related to ET-oxy radicalinvolvement.

In summary, based on the evidence from pnor contributions and our work, the various categories appearto display the following common characteristics:.bind-ing to DNA, involvement in ET. and production ofactivated oxygen. However, there are gaps for the various categories, which need to be filled by furtherinvestigation.

REFERENCES

1. H. Sies. ed. Oxidative Stress. Academic Press. Orlando. FL2. BjSiiwU»<J J. M. CO^Ti^FrceR^binBMogy

ond Medicine. Oxford University Press. New York "985)3 B Jieluwell nod J. M. C. Guttcridge. The importance of free

radicals «d catalytic metal ions in human diseases. Moke.Aspects Med. 8:89-lW(»985). .

4 BN Ames. Dietary carcinogens and enucarcwogens. $«♦

to' STStomiiic .aica. of dims w! pt»8«y»« «cU«. to.044-288 Academic Press, N*w »OT* (198*).6. JJV.SetSdC. F. Cnigaell.^.^*^gK^Md toxicology-selected topics. Pharmacol. R*. 33.189-211

eardaoieMiU.i. Environ. Path. Toxicol. f^-303(198W.8 HKappu*. Overview of entyme systems involved in bioreLtK drugs and in redox cycling. Btockam. Pnarmacol.33:1-6 (1986).

9 W A Pryor, ed Free Radicals tn Biology and Medtemc.' Academic Press. New York (1982).

10 PK^acic P. WCrawford. M. D. Ryan, and V. C. Neteee."Charge transfer mechanisin for ^^* ^fftKgf

«nd other agent*. Biaelectrockem. Bwenen H*^1*"9™-II. P. Koveci?, J. R. Ames, P. Ummc. H. £»o. O. Coa. H.

Jackson, L. A. Rivera. L.Ramirtt. and M. D. Ryan. Chargetrensfer-oxy mdical mechanism for anticancer agent*. Anti;Center Drug Design. 1:196-21* 0*86).12 KiaR.AiwJ, art M. D. Ryan. Charge transfer-o»y J»S mecbanittn for •Bticahoer agents: 9.AmUno8cn-eUa«<»bBdtiedloTptiblication).

ii p W Crawford. P. Lumme, K. Elo. M. D. Ryan, eno r.

car «enu: mAMSA derivatives, roodamine 123 and mckelMlicylaldoximate. Free Mod. Res. Ccnvmtn. (in press).

14 ? v7 Crawford. W. O. Foyt, M. D. *y», and P. Kovacic.' Cyclic voltammetry of quinoUnium aalu and related com

pounds: Correlation with structure and anticancer activity (sub-mined fof publication). VMUmM.

15 J. R. AmesTM. D. Ryan, D. L. Klayman. and P. Kovacic.Charge transfer and oxy radicals in antimalanel action. Qui*nones, dapsoec metabolites, metal complexes, ^iumions,andpioxWes.^. Pre* Radicals Biol. Med. i:3S3:*6i (1985).

16 PWCiawford. P- Kovacic. K. W. Gilman, and M. D. Ryan-' Charge transfer mechanism for benzodiazepine (B2) action

Contlition ofreduction potential ofBZimioium with structureand drug activity. Biotlecrrochem. Bioenerg. 16:40*-426(1986)

17 J R. Ames. S. Brandinge. B. Rodrigutt.N. Castagnoli. Jr.,M D Ryan, and P. Kovacic. Cyclic voltammetry with cycliciminium ions: Implicaiions for charge transfer with biomole-cvles (metabolites of nicotine. phencycMw and apennine).Bioon Che* 14:228-241 (1986).

18. J. R. Ames, N- Castagnoli. Jr.. M. D. Ryan, and P. KovacicOxidative ionic metabolites of l-methyl-4-phcnyl.l.2.3.6-te-CThvc^pyridine (MPTP): Correlation ofetectrorcduction withphysiological behavior. Free Rod. Res. Commun. 2:107-113

19 I R. Ames. K. T. P«ts. M. D. Ryan, and P- Kovacic Con-jusated and cross-conjugated mesomeric betaines: Correlationof'elccnoreduction *ith sememre and physiological activity.Lift Sci. 39:1085-1091 (1986).

20 M D. Ryan. R.C. Scamehom. and P. Kovacic Charge van*-' for in the mechanism ofdrug action involving qaiaoxalme di-

N-oxide*.;. Pham. $cl. 74:492-495 (1985).21 P. W. Crawford. R.G.Scamehom, U.Hollsteln, M- o. Ryan,

and P. Kovacic Cyclic voltammetry of phenawnes and qui-noxalines Including' mono- and di-A'-oxidts: RelaUon l0™!:~tm and antinucrobial acdvit>-. Cnem.-Biol. Interactions 60:67-

*»"» P Kovacic. Does charge transfer by diimintum play a wide-"" spread role in living systems? Ktm. Ind- 33:473-492 (1984).23 M N Hamers and D. Roos. Oxidative stress in human neu-

' mmhUic cranulocyie*; Host defence and self-defence. In: Oxidative Stress (H. Sies. ed.)T pp. 351-381. Academic Press.Orlando. FL (1985).

24 B Hafflwell and h M. C. Guttcridge. Free radicals as useful' species. In: Free Radicals in Biology and Medicine, pp. 246-

278. Oxford Univenity Press. Oxford, England UW5).•»5 R L. Baehner. L. A. Boxer, and L. M- Ingranam- Reduced

oKveen by-productt and white blood cells. Is: Free Radicalsin Biology Volume 5 (W. A. Pryor. ed). pp. 91-113. Academic Press. New York (1982).26 B HalliweUandJ.M.C.Guneridge.Thechctm$tryofoxygen

' radicals and Other oxygenHlerived species. In: Free RadicalsinBiohgy and Medicine, pp. 20-66,Oxford University Press,Oxford. England (1985). .

2? Y Fuiita. K- Wakabayaihi. M. Nagao. and T. Sugtmura. Im-' plication of hydrogen peroxide in the mutagenicity of coffee.

Mutation Res. 144.227-230 (1985).28. Tl Shenep. D. C Stokes, and W. T. Hughes. Lack of an-

Mechanism of antibacterial action389

bacterial activity after intravenous hydrogen peroxide infu-icn inexperimental Escherichia coh species, infect. Immun.a 607-610 (1985).

B. Suerle and D. I. Faulkner. Metabolites of the marine*»ge Chondrosio collectrix. J. Org. Chen. 44:964-968979).

W. Phillipson and K L. Rinehart. Jr. Antifungal peroxide-jOU£?i!?Idl-?om l*° Caribbean sponges J. Am. Chem.>c. 105.7735-7736(1983).Matsugo. N. KayamoH. Y. Hatano. T Ohia. andT. Konishi.

•egredation mediated OH radical generation from synthetic/che peroxides: ESR studies. FEBSUtt 184:25-29 (1985).

E. Bembury. Synthetic antibacterial agents. In: Burger'sfedicinai Chemistry. Pan 2 (M. E. Wolff, ed). pp. 41-81.iley-Intcrxienee. New York (1979).968)B^Cy*r*yi' ***krtwwM, Academic Press. New YorkN. Kirkman. Glocose-6-phosphaie dehydrogenase varianu

d drug-induced hemolysis, HY Acad. Sci. Ann. 151:753-- (1968).

L. Holuman. Role of reactive oxygen and metabolite bind-' indrug toxicity. Life Set. 30; 1-9 (1982).W. A, Oliver. R. M. Rashman. and A-W. SomerviJle. The.thesis and spectroscopic properties of some b8logeno-l,2-rmthoquinones. Tetrahedron 24*06*-40?2 (196&;.

P. Stradyn'. R. A.Gavar. L.Kh. Baumane. L. M. Baider.A.Medyoya. and M. A. Trushule. Effect of the heteroatomthe one-electron reduction of o-nitro derivatives of fivc-aibered rings. Chem. Hetero. Cmpds. 16:1211-1217 (1980).E. Banak and M. D. Hawley. The electrochemical studvAe fomsiion and decomposition of 2-bBlo.2-nitropropaneob radicals. J. Eitetroanal Chem. 33:13-22 (!9?1).Mullet and L. Schroder, Elftfitrochtmical reduction of 2-^2»n*tn>propant-13^iol$. Z. Chem. 21:286 (1981).B. Gavioli. G. Grand}, and R. Andreoli. Solvent effectsme mechanism of polarographic reduction of nitrealk&ncs-.'/. Csech. Chem. Commun, 36:730-739 (1971).Breccia. G. BerriJU. and S. Roffii. Chemical radiosetui-iuon of hypoxic cells and redox potential;: Correlation oflaametnc results with pulse radioiysis data of nitro-com-.ads tad ranloscnsiuzers. Int. J. Radial Bio!. 3fi:S<-8979).

D. Ryan and D. H. Evans. Effect of metal ions on theirocbcmical reduction of beniil innon-aqueous solvents.

•Jectroanal. Chem. 67-333-257 (1976).K. Mishra and K. D. Gode. Electrochemical reduction ofofuratoae and its deiermination is pharmaceutical dosage•is b> D. C.poJarography. Analyst 1101373-1376 (1985)... Ohve. Correlation between the half-wave reduction po-ulsofmiroheierocycles and their mutzgemcitv in Chinese.iter V79 spheroids. Mutation Res- 82:137-14*5 (198 h.

Burtniw. W. F. Smyth, and R. F. Palmer. Anultraviolettral and polarographic study of nitrofurantoin, a urinarv. antibiotic. Analyst 101:986-991 (1976).: ChigneU. Strueture-*ctivity relationships inthe frce-rad-metabolism of xenobiodes. Environ. Health Persoect33-137(1985). *

Joseph) and R. P. Mssoo. Toxicity of nitroimicazoJes.iioaaivation ofForeign Compounds (M. W. Anders, ed.).451-483, Academic Press, Orlando. FL (1985).». J. Moreno and R. Docampo. Mechanism of toxicity of•compounds used inthe chenotherapv of trichomoniasis.ron. Health Perspect. 64:199-20$ (1985).?. Mason. Free-radical intermediates in the metabolism>xic chemicals. In: Free Radicals in Biology. Volume 5A. Pryor, ed.). pp ,161-222. Academic Press, New York21.

i. Moore, L. F- Chasseaud. J. D. Uwis. P. C. Risdall.i. L.Cramptoa. The meubolism ofthe antibacterial agentjpol(2-bromo-2-niiropropans-1.3-d»ol: given orallvtoratsJOgS. Food. Cosmei. Toxicol. 14:183-18? (19/6/; Chtm.. 85:103727 (1976).awwnek and R. C. Duty. Polarographic studies in ate-

tomtnle and dimethylformamide. VI: The formation of dihal-ocarbenes. J Eiectrochem. Soc. 108:1135-1138 (1961)

52. J. Armand. Controlled-potential polarography and electrolysisof nitro- and nitroso-derivatives with another group inthe gem-inal position. Bull. Soc. Chim. France. 543-551 (1966).

53. M E. Peover and J. S Powell. Dependence of electrode kinetics on molecular structure: Nitre-compounds in dimethyl-formamide. J. Eiectroanal. Chem. 20:427-433(1969).

54. P. E. Stewart end W. A. Bonner. Polarographic study of lowernitroparafflos in oonalkaline media. Anal. Chem. 22:793-795(1950).

55. C A Russell and W. C. Danes. Electron-transfer processes.VIII: Coupling reactions of radicals with chloroform J. Am.Chem. Soc. 90:347-353 0968).

56- R. I. Stretton and T. W. Maoson. Some aspeeu of (he modeof action of theantibacterial compound proeopo) (2-bTOmc-2-nitropropan*l,3-diol). /. Appl. Bacterial, 36:61-76 (1973).

57. L. C Lappas. C A. Hirscb, and C. L. Wisely. Substituted5-nitro-l ,3-dioaenes:Correlation ofchemical structureandantimicrobial activity. J. Pharm. Sci. 63:1301-1305 (1976).

58. L. A. Buckley. X. Z. Jiang, R. A. James. K T- Morgan, andC S. Barrow. Respiratory traei lesions induced by sensoryirritants atthe RD50 concentration. Toxicol. Appl. Pharmacol74:417-429 (1984).

59. M Kawai. Inhalation toxicity of phosgene and orichloronitro-methane (chioropicrio), Sangyo Igaku, 15:406-407 (1973):Chem. Absn. 81:46083 (1974).

60. E.J. Bond. The action of fumiganu oninsects. IV: Theeffectsof oxygen oa the toxicity of fumiganu to insects. Can. J.Biochem. Physiol. 41:993-1004 (1963).

61. M Moriya, T. Onta. K. Watanabe. T. Miyazawa. K. Kato,and Y. Shirasu. Further mutagenicity studies on pesticides inbacterial reversion assay systems. Mutation Res. 116:185-216(1983).

62. A. Stem. Red cell oxidative damage. In: Oxidative Stress (H.Sies. ed.), pp. 340-341. Academic Press. Orlando. FLO985).

63 A. Furst and S. B. Redding. New developments in the studyof metal carcinogenesis. / Environ. Sci. health. C2:103-133(1984).

64. VY. Saenger Metal ion binding to nucleic acids, in: Principleso/S'ucleic Acid Structure (C. R. Cantor, ed.), pp. 201-219.Springer-Verlag. New York (1984).

65- R. Benesch and R. E. Benesch. Polarographic studies of organic mercury compounds. J. Am. Chem. Soc. 73:3391-3396(1951).

66 I E. Page and J. G. Waller. Polarographic determination ofthiomenalate Analyst 74:292-296(1949).

67. N. S. Hush and K B. Oldham. The etectroreduction oforganiciodides and organomercuric betide*. J. Eiectroanal. Chem6:34-45 (1963).

68. W. Strieks and S. K.Chakravarti. Ampoterometric titration ofsulfhydryl and disulfide groups with organic mercury compounds at the rotated dropping mercury indicator electrode.Anal. Chem. 33:194-199 (1961).

69. S. K- Vig and N. R. Bannerjee. Polarographiccummt/volugecurves of fluorescein. IV: Effect of coocemrauoo. Electro-chim. Acta 16:157-165 (1971).

70. W R. Omdorff, R. C. Gibbs. and C. V. Shapiro. The ah-sorptionspectra of fluorescein, fluoran and some related compounds. J. Am. Chm. Soc. 50:819-8280928).

7|. H. C. Ansel and G.£. Cabre. Influence ofdimethyl sulfoxideon the hemolytic activity of antimicrobial preservatives. JPharm. Sci. 59:478-481 (1970).

72. A. J. Wotuisk, B. S. Glisson,K. R. Hande. andW. E Ross.Inhibition ofeioposide-induced DNA damage and cytotoxicityin LI210 cells by dehydrogenase inhibitors andotheragentsCancer Res 44:626-632 (1984).

73. E. W. Baur. Inhibition of hemolysis by tricyclic dyestuffs-fluorescein-phenothiaxine antagonism- Biochem. Pharmacol -22:1509-1516(1983).

74. T. Moroi. The respiration of Trichophyton: II. Effects of antifungal agents on the respiration of Trichophyton. Hffu ToHinyv. 25:221-232 (1963): Chem. Abstr. 61:3450 (1964).

390J. R. Ames, M. D. Ryan, and P. Kovacic

75. B. Venugopal andT. D. Luckcy. Metal Toxicity inMammals:2. Chemical Toxicity of Metals andMetalloid*, Plenum. NewYork (1978).

76. O. Cantoni. S. H Robison. R. M. Evans. N- T. Christie, D.B. Draw, and M. Costa. Possible involvement of superoxidefree radicals in the HgCl.-induced DNA damage in CHO cells.Fed. Pro:. 421135(1983).

77. O. Cantoni, N. T. Christie. S. H. Robison. and M. Costs.Characterization of DNA lesions produced by HgClj is cellculture systems. Ckem.-Biol. Interactions 49.209-224 (1984).

78. S R. Riberov. L C. Beaov, aad I- C Bencbev. HgClj increasesthe methemoglobin prooxidantactivity: Possiblemechanismof Hg-'-induced lipidperoxidation inerythrocytes. Chem.'Biol. Interactions 50:11M19 (1984).

79. S R. Ribarov, L. C. Beoov, V. 1. Marcovi, and I. C. Bencbev.Hemoglobiihcaulyzed lipid peroxidation in the presence ofmercuricchloride. Chem.-Biol. Interactions4&AQS-U2 (1983).

80. ).*P, Chariaad, J. F. Britten, andA. L. Beaucaarap. Structureof an unusual 5:2 methylmercury: adenine complex. Inorg.Chim. Acta 124:161-167 (1986;.

81. G. C Crvmphn. ). M. Midgley, and1. T. Smith. Mechanismsof action of nalidixic acid and its congeners. In: Topics inAntibiotic Chemistry. Volume 3. Pan A (P. G Sammes. ed-),pp. 13-36, Wiley. New York. (1980).

t 82. L. Verbis!. Quinounes-.Pharmacology. Pharm. Weekbl. Sci.Ed. 8:22-25 (1986): Chem. Abstr. 104:179577 (1986).

83. A. Cole, J. Goodficld. D. R Williams, and J. M- Midgiey.The complexation of transition senes metal ions by nalidixicacid- Inorg. Chim. Acta 92:91-97 (1984).

84. W. E. Hill and w. L- Fangmar.. Single-strand breaks in deoxyribonucleic acid and viability loss during deoxyribonucleicacid synthesis inhibition in Escherichia colt. J. Bacteriol.116:1329-1335(1973).

85. A, A. Divo, T. G. Geary,andJ. B. Jensen.Oxygen* andtime*dependent effects of antibiotics and selected mitochondria) inhibitors on Plasmodium falciparum in culture. Antimicrob.Agents Chemother. 27:21-27 (1985).

86. w. J. Van Oon. R. H A. Sore). D. Brussee. S. G. Scbulman,P- Zuroan, and J. Den Hanigh. Poiarographie reduction anddetermination of nalidixic acid. Ana!. Chim. Acta. 149:175-191 (19B3).

87. L. L Shen and A. G. Pemet. Mechanism of inhibition of DNAgyrase by analogues of nalidixic acid: The tareet of the drugsis DNA. Proc. Natl. Acad. Set. USA 82:30-311 (I965i

88. R. D. IronsandT. Sawabata.Phenols, catechols andquinones.In:Bioacrivation of Foreign Compounds (M W. Anders,ed.).pp. 259-281. Academic Press. Orlando, FL (1985).

89. A. Albert. Metal-binding substances, in: Selective Toxicity,pp. 385-412. Chapman and Hal!. New York (1979).

90. Y. H Miriam and G P. Glover. Degradation of DNA bymetalloanthracyclines; Requirement for metal ions. Biochem.Biophys. Res. Commun. 136:1-7 (1986).

91. C. G- Butler and R. C Kayc. The polaiographic reduction ofmetalcomplexes. /. Electroanal. Chem. 8:463-471 (1964).

92. T. M Florence. B- G Luasdea, and J. J. Fardy. Algae asindicators of copper speciation. In: Complexation of TraceMetals In Neutral Waters (C. J. M. Kramer and J. C Duinker.eds). pp. 411-418, Nyhoff, The Hague. Netherlands (1984).

93. T. R. Sweeney and R. E. Strobe. Antimalarials. Io: Burger'sMedicinal Chemistry, Pan II. (M. E. Wolff, ed.), pp. 333-413, Wiley-Interscience. New York (1979)

94. F.Vicente. 3. Tnjueque. and F. Thomas. Correlation betweenphysical-chemical parameters and the pharmacological bchav-iorofsome biguanides.Qum. Ind. (Madrid)28:619-622 (1982).

95. F. T. Ven. F. V. Pedros.J. M. Calatayud.and V. P. Marline*.Poiarographie determination of proguanil and chlorhexidine.Talanta 30:977-979 (1983).

96. F. V. Pedros. C $. Box. and J. T. Monge. AC poiarographiereduction of biguanides. </. Electroanal. Chem. 182:413-426(1985).

97. F. H S. Curd and F. L. Rose A possible mode of action ofpaiudnne. Nature 158.707-708 (1946;.

98- C. R. Gale. A. M. Welch, and J. B. Hynes. Pharmacology ofcertain antimicrobial N',^substituted biguanides. /. Pharmacol. Exptl. Therap. 138:277-284 (1962).

99. B. Datta, K. Singh, and D. Sen Poiarographie investigationof Cu(II) complexes with « donor ligands. Blectrachim. Acta25:533-537 (1980).

100. B. Halliwell and J. M. C. Guttcridge. Free radicals, aging,and disease. In: Free Radicals in Biology and Medicine, pp.279-315. Oxford University Press.Oxford, England (1985).

101. J. W. Lows. Molecular loecbanitmi of action of anticanceragents involving free radical intermediate*. Adv. Frte Had.Biol. Med. 1:225-264 <198S).

102 B. K. Sinha and C, F. Cbigaell. Binding mode of chemicallyactivated semiquiaone free radicals from quinone anticanceragentsto DNA. Chem.-Bbt. Interactions 28:301-308 (1979).

103. N. R Bacbur, M. V. Gee, and R. D. Friedman. Nuclear cat*alyted antibiotic freeradicalformation. CancerRes. 42:1078-1081 (1982).

104. R. i. Yoangman. F. Gottt, tad E. F. Elsmer. Role of oxygenactivation in Adriamycin-mediated DNA strand scission andthe effect of binding on the redox properties of the drug. In:OxygenRadicalsinChemistryandBiology(W. Bors, M. Saranand D- Tait. eds.), pp. 131-136. De Gruytei. Berlin (1984).

105. M. E. Wolff (ed.). Burger's Medicinal Chemistry, Part II,Wiley. New York (1979).

106. A. N. Grinev. 1. K. Anenicbev, 1. S. Nikolaeva. L. A- Go*lovanova, T V. Pushkiaa. O. V. Okinshevicb. A. N. Fomina,and G. N. Pershin. Synthesis and biological activity of 4-substiiutcd-1,2-naphthoquinones, 1.2-aaphibaleaediols and theirderivatives- Khm.-Farm. Zh. 19:699-705 (1985), Chem. Abstr.105:6285 (1986).

107. V. R. Famshyan, S. F. Gerasina. I. A. Poiyakova. N. I. Fa*deeva, andV. M- Zhdanov. Influenza vims replication in thepresenceof booaphthonaadother inhibitorsof macromoieculesynthesis. Vopr. VtrusoL 30:174-180 (1985); Chem. Abstr.103.243 (1985).

108. T. Fujiaaga. K. lzutsu. and T. Nomura. Effect of metal ionson the poiarographie reduction of organic compounds in dipolar aprotic solvents. J. Electroanal Chem. 29:203-209(1971).

109- L. F. Fieser and M- Fieser. The reduction potentials of variousnaphthoquinones. J. Am. Chem. Soc. 57:491-494 (1935).

110 P. Zuman. Substituent Effects in Organic Polarograph\, PlenumPress. New York (1967).

111. M. D. Stallings. M. M. Morrison, and D. T. Sawyer. Redoxchemistry of mctal-catechol complexes in aprotic media: I.Electrochemistry of substituted catechol &and their oxidationproducts. Inorg. Chem. 20:2655-2660 (1981).

112. I. A. Clark and W. B. Cowden. Anumaianali. la: OxidativeStress (H. Sies, ed.), p. 143, Academic Press, Orlando. FL(1985).

113. P. J. Tbomelley. M. d'Arcy Doberty, M. T. Smith. J. V.Bannister, andG. M. Cohen. The formation of active oxygenspecies following activation of 1-aaphtbo). 1.2* tad 1.4-naphtboquinone by rat liver microsomes. Chm.-Biol. Inter-actions 48:195-206 (1984).

114. J. p. StTBdins end V. T Gle2er. Azo, azoxy and diazo compounds. In: Encyclopedia ofElectrochemistry of the Elements.Volume 13 (A. J. Bard and H. Lund, eds.), pp. 163-208,Marcel Dekkcr. New York (1979).

115. A. Albert, SelectiveToxicity, 7th ed.. ChapmanaadHall.NewYork (1985).

116. G. C. Mueller aad J. A, Miller. The reductive cleavage of4.dimethylamiae-azobenzeae by rat liven Reactivation of carbondioxide-treated homogeaates by riboflavin-adenine dinucleo-tide. J. Biol. Chem. 185:145-154 (1950).

117. R. A. Parent aad I. Dressier. Absorption and disttibutioo ofC- I. solvent red 24 ia rats; intra-tracheal administration ofCarbon-I4-Ubeleddye. Drug Chem. Toxicol. 2:409-420 (1979):Chem. Abstr. 92:122635 (1980).

118. J P. Brown, G. W. Roehm, aad R. J. Brown. Mutagenicitytesting of certified food colors and related azo, xanthenc and

i9

4.

Mechanism of antibacterial action 391

triphenyJmethane dyes with the salmonella .microsome system. Mutation Res. 56:249-27) (197$).N. Aaand. Sulfonamides and sulfones. In: Burger's MedicinalChemistry, Pan 2(M. E Wolff. ed). pp. 1-40. Wiley. NewYork (1979). '

10. C. L. Scott and M. R. Rasbridge. In vitro action ofdapsone•nd its derivatives era norma) and 06PD (giucosc-6-pbosphate3l?y(WaT *™* ""*' *''J' Ha9mMl' a4:304"

H. R. M. Weetman. L. A. Bow, M. P. Brown. N. M. Munich.•nd RJL Baphner. In vitro inhibition ofgranulopoiesis by 4-^^ntaSST^^^^ Mltofi Br J' *-""•

l. H. J. Scboler, R, Leimer and R. Richie. In: Handbook of8tt?2.. pharmacology: Antimalarial Drugs II, Volume68/1. W Peters and W. H. G Richard*. ed*.)7pp. 123-206.Spnnger-Vcriag. New York (1984).

3" ?" ^J"4, 0*w*liv* «ni( «>d formation ofexcited species.to Oxidative Stress (H. Sies. ed.). pp. 318-319. AcademicPress, Orlando, FL (1985).R. C. Kayc and H. I. Stoathill. The polarographic reductionof some natural and artificial hydrogen-earners in bacteria)cniyme systems. J. Chem. Soc. 3244-3247 (1932).D. Binder. C. R.Noc. B. C. Pragcr. and F. Turnowsky. Pyr-Mmc-M-dioxides fused to hetcrecydes: 3 Synthesis and antibacterial activity of substituted pieridinc-5.b.dioxide*. Arc.nl?; •"**' M:80^W5 0983); Chem. Abstr. 99 )05047i_ l. Johnson. D. F. Worth. N. L. Colbry. and L. M. WerbeJ.Studies on 8-aminoquiooUne I-oxidc antimalarial agents. J.Heterocyclic Chem. 21:1093-1096 (1984).L^ B Kier. Molecular Orbital Theory in Drug Research. Ac-ademicPress, New York (1971).A. J. Blacker, J. Jaawinski, J.-M. Lehn, and P. X. Wilhelm.Photochemical cleavage of DNA by 2.?.d>axapyrenium cat-tons. Chem. Commun. 1035-1037 (1986).W. Sliwa. G. Maustak. and A. Postawka. The chemisirv ofN-subsuruted pyridinium salts. Heterocvcles 13:1513-1554(1985).

M. M. Morrison. E. T. Sec. J. K Howie, and D. T. SawyerFlavin model systems: I. The electrochemistry of 1-hvdrox.yphttMtne and pyocyaninc inaptotic solvents. J. Am. Chem.Soc. 100:207-211(1978).C. Parkanyi and 0. i. Leu. Charge-transfer spectra of someheterocyclic cations, leix. Naturforseh. 30b:984-986 (1975).

A.L?0,° Md F*•*«*»»tt- Benzophenanthridinium saltequiubru. Heterocyeles 12.11-15 (1979).

133. J a. Miller and E. C Miller. Tbe metabolic activation andnucleic acid adducts of naturally-occurring carcinogens: Recent results with ethyl carbamate and toe spice flavors safroleand estragole. Br. J. Cancer 48:1-15 (1983).

134. R. A- Floyd, ed. Free Radicalsand Cancer, Marcel Dekker,New York (1982).

135. G. Dryhurst. BUctrochemisny of Biological Molecules, pp.71-85. Academic Press, New York (1977).

136. V. L.Reynolds. I.J. Molineux. D.J. Kaplan. D.K. Swensoo.and L. H. Hurley. Reaction of m« antitumor antibiotic CC>1065 with DNA. Location of tbe aite of thermallv inducedstrand breakage and aaalyits of DNA sequence specificity.Biochemistry. 24:6228-6237 <1985).

137. S. D. Nelson. Metabolic actrvtuoo and drug toxicity J. hied.Chem. 25:733-765 (1982).

136. K Mijaxaki. Y. Matsvfaiea. and T. Kubou Cyclic volum-meiry of aromatic •mine N-oxidtsin Doroqoeous aoivenu andthe aubiliry of tbe free radicals produced. Bull. Chem. Soc.Jpn. 44:3850-3853 (1981).

139. P. Neta. M.-C. Ricboox. and A. Hamman. Intramolecularassociation of eovakntly linked vioiogen radicals. J. Chem.Soc. Faraday Trans. 2 §1:1427-1443 (1985).

140. 5.Ktnoshita.K.Ufci, K.'Nalano.M Sbimixu, T.Takanashi.and M. Matsut. Mitomycin derivatives: 1. Preparation of mi-tosane and mitoseoe compounds and tbeir biological activitiesJ. Med. Chem. 14:103-109 (1971).V. A. Shapovelov, a. N. Gatdtikevica, end V. D. Bezuglyi.Polarographic investigation of 6* and 7<njtro derivatives of 9-(methylamino)acrtune, which havean antibacterial actios JGen. Chem. USSR 80:2264-2266 (1980).G. E. Adams. I. J. Stratford. R. G- Wallace, P. Wardmen.and M. £. Warts. Toxicity ofnitrocompoundi toward hypoxicmammalian eetls invitro. Dependence onJtduction potential/. Natl. Cancer Instit. 64:555-560 (1980).1 Pernak. S. Kucbenki, and J. Krysinski. Activity of newquanernary ammonium compounds on strains of bacteria andfungi. Part 5: Synthesis of 3-mrtnyl-B-alkyhhJometfaylpyn-d»ne.. l-n«myl-3-n-allryltWoaietbylirnldaxole- and l-cchvi-3.n-alkyltfaiolimidaxohnc chlorides. Phamazie 38752-754(1983).P. L. Chesis. D. E. Levin. M. T. Smith. L. Ernster. and B.N. Ames. Mutagenicity of quinones: Pathways of metabolicactivation and detoxification, froc. Natl. Acad. Set USA81:1696-1700(1984).R. N. Misra and B. Misra. Genetic toxicologiea] testing ofsome dyes by the micronueleus test. Mutation Res. 170:75-

141

142.

143.

144.

145.

TOTAL P.16