H2O2-mediated one electron oxidation of carcinogenic N-fluorenylacetohydroxamic acids to nitroxyl free radicals

Chem.-Biol. Interactions. Elsevier

Scientific

46 (1983) 317-334

Publishers

Ireland

317

Ltd.

DIATED ONE ELECTRON OXIDATION CYTOCHRO dH&tOF CARCINOGENIC N-FLUORENYLACETOHYDROXAMIC ACIDS TO NITROXYL FREE RADICALS

C.L. RITTER.

Laboratoy Lajomtory (Receivd (Revision (Accepted

D. MALJZJKA.GIGANTI’

and CF.

POLNASZEK”

for Cancer Research, Veterans Administration Medical Center and Department of Medicine and Pathology, University of Minnesota, Minneapolis, MN 55417 (U.S.A.) December 10th. 1982) received March 2&h, 1983) April 4th. 1983)

SUMMARY The oxidation of carcinogenic hydroxamic acids, N-hydroxy-N-2fluorenylacetamide (N-OH-2-FAA) and N-hydroxy-N-3.fluoi enylacetamide (IV-OHd-FAA) catalyzed by horseradish peroxidase (HRP) or cytochrome c in the presence of HzOz was investigated. HRP/H202 was a more efficient system in oxidation of both hydroxamic acids and the standard substrate, guaiacol, then cytochrome c/HzOs. Peroxidative activity of cytochrome c was shown after incubation with Triton X-100 and H202 for 20 min at room temperature in 0.05 M phosphate buffer (pH 7.5) or in 0.1 M sodium acetate (pH6.0) without Triton X-100. Both hydroxamic acids were oxidized to nitroxyl free radicals as shown by electron spin resonance (ESR) spectroscopy. These radicals dismutated to equimolar amounts of 2- or 3-nitrosofluorene and acetate esters of the corresponding hydroxamic acids as shown by thin layer chromatography and spectrophotometric analysis of the products. In addition, large amounts of the N-fluorenylamides were generated in the reactions with cytochrome c/Hz02 system. Of the products, only 2- or 3-nitrosofluorena per se roxamic acids, interacted or when generated from the oxidation of the with lecithin (1 mg/ml) to yield ESR signals of the immobilized nitroxyl free

‘To whom reprint requests should be sent at: Veterans Administration Medical Center, Medical Research Laboratories 151B. 54th Sreet and 48th Avenue South, Minneapolis. MN 55417 [U.S.A.) “Resent address: Department of Chemistry. University of Minnesota, Minneapolis, MN 55455, U.S.A. Abbreviations: N-OH-2.(or -3.)FAA. N-hydroxy-N-P-(or -3.)fluorenylacetamide: 2. or 3-NOF. Zor 3.nitrosofluorene; N-A&.%-(or -3.)FAA, N-acetoxy-N-P-(or -3-)fluorenylacetamide: 2- or 3FAA, N-2.(or -3.)fluorenylacetamide; 2- or 3-NOrF, 2- or 3.nitrofluorene: HRB. horseradish peroxidase; ESR, electron spin resonance; TLC, thin-layer chromatography; UV. ultraviolet; RT, room tempernture (23”-25°C). 00042797/83/$03.00 0 1983 Elsevier Scientific Publishers Printed and Published in Ireland

Ireland

Ltd.

318 radicals. In contrast to HFP/HzOz system, in which the initial velocity of the radical formation was too fast to measure and the maximal concentrations of the nitroxyl free radicals of both hydroxamic acids were similar, in the cytochrome c/Hz02 system the nitroxyl free radical of N=OH-2-FAA formed at a 6-fold faster rate and accumulated at a a-fold higher concentration than the radical of N-OH-&FAA. In both enzyme systems, the persistence of the signal and the length of time before it had decreased to one half its maximum were several-fold longer for the nitroxyl free radical of N-OH=39 FAA than for that of N-OH-B-FAA. These data showed that these nitroxyl free radicals differed in their kinetic properties. One electron oxidation of N-OH-3-FAA by HRFVHZO~ system ax:d of both isomeric hydroxamic acids by cytochrome c/Hz02 system are reporktd for the first time in this work and may be considered an activation reaction in carcinogenesis by these corn. pounds. Key words: Carcinogen oxidation-Hydroxamic Cytochrome c

acids-Nitroxyl

radicals-

WTRODUCTION

One electron oxidation of a carcinogenic N-fluorenylacetohydroxamic acid, N-OH-B-FAA, to a nitroxyl free radical followed by its dismutation to 2-NOF and N-AcO-2-FAA has been shown to occur with chemical oxidants IL21 and with various peroxidaselperoxide systems 13-51. Since both products of dismutation are capable of covalent binding to cellular macromolecules 16-81 this reaction sequence has been suggested as one mechanism of activation of N-OH-Z-FAA in carcinogenesis 191.One electron oxidation of N-OH-P-FAA to the nitroxyl free radical has also been shown with hepatic microsomes [lo,111 and mammary gland cells of the rat 112,131.Both liver and mammary gland of the male and female rat, respectively, are the sites of carcinogenic action of N-OH-B-FAA 1141.Since the isomeric N-fluorenylacetohydroxamic acid, N-OH-8FAA, is also carcinogenic for the rat mammary gland 115,161, we were interested whether it undergoes a similar oxidation to a nitroxyl free radical followed by its dismutation to 3.NOF and N-AcO=3=FAA. We, therefore, compared this reaction sequence for both isomeric fluorenylhydroxamic acids using HRFVHzO~ as an oxidizing system. It has been suggested that in hepatic microsomes one electron oxidation of N-OH-B-FAA is mediated via cYtochromeP-456 system 110,111.However, in the rat mammary gland, the cytochrome P-450 system is much less active than in the liver 1171.Hence, we sought a possibility that other cytochrome-associated peroxidases may be involved in this type of oxidation. In this work, we present evidence that both isomeric fluorenylhydroxamic acids are oxidized to nitroxyl free radicals by a cytochrome c/HzOzsystem.

319 MATERIALS AND METHODS

Comjxww-is 2=FAA and 2-NOZFwere purchased from Aldrich, Milwaukee, WI. 2.FAA was recrystallized from ethanol/water (7:3) and had a m.p. of 196-198°C. 2-NOgF was purified by preparative TLC on silica gel with benzene as a mobile phase and had a m.p. of 159-160°C. N-OH-2.FAA, m.p. 150-151°C E14], 2=NOF, m.p. 79-81°C 181, N-AcO-2.FAA, m.p. 111-112°C [18], 3.FAA, m.p. 194-196°C [19], 3=NOzF, m.p. 104-105°C [20], N-OH-8FAA, m.p. 131132°C [21] and N=AcO=3=FAA, m.p. 104-105°C [22] were prepared by the published procedures. The IR and W spectra of the compounds matched those of the authentic samples. The compounds were found pure by TLC on silica gel GFPMor cellulose MN 300 (Macherey Nagel and Co., Diiren, F.R.G.) with solvent systems described previously [8,23]. A new compound, 3-NOF, was prepared by the procedure used for the a-isomer [8]. N-8Fluorenylhydroxylamine (0.5 mmol) 1241was oxidized in cold (4°C) dimethylformamide with a solution of Nrsaturated ferric ammonium sulfate (50 ml, 0.02 N) in 0.04 N sulfuric acid. The reaction mixture was kept under Nz and stirred for 15min. The green precipitate of 3-NOF was filtered, washed with NY saturated cold water and dried in vacuum. It was then recrystallized from hot hexane. The yield of the compound melting at 75-76°C was 46.6%. Calculated for (&H$IO: C, 79.98%; H, 4.65%; N, 7.17%. Found: C, 79.84%; H, 4.83%; N, 6.81%. W, A$$‘” 266nm (E, 25.2 mIW*cm~ ‘), 314 nm (F. 8.97 mM-’ cm- ‘).

Reagents Cytochrome c (horse heart, type III), HRP (type II), guaiacol, L-a-lecithin (egg yolk, type V-E) and Triton X-100 were from Sigma Chemical Co., St. Louis, MO. OmniSolv@ methanol was from MCB manufacturing Chemists, Inc., Cincinnati, OH. The spin probe, 2,2,5,5-tetramethyl-3-pyrrolin-1-oxyl-3-carboxamide, was from Molecular Probes Inc., Junction City, OR. All other solvents were ACS grade from Fischer Scientific Co., Fairlawn. NJ.

Enzyme assays The concentration of cytochrome c was determined from the absorbance at 550 nm of a dithionite-reduced solution (P, 29.5 m&-l cm-‘) [25]. The concentration of HRP was determin ?d from the absorbance at 403 nm (F, 102 mM_’ cm-‘) [26]. Peroxidase activity was determined at room temperature (RT) using guaiacol as a substrate 1271. The reaction mixture (3 ml) contained 33 mM guaiacol, 0.33 mM HzO, in 0.05M sodium phosphate buffer (pH 7.5) and sufficient enzyme in 0.02ml to give a linear increase in absorbance at 470 nm. In the assay of cytochrome c, 0.015% (v/v) Triton X-100 was added and the mixture incubated for 20 min. The rate of the cytochrome c-catalyzed reaction was also determined in 0.1 M sodium acetate buffer @H 6.0) without Triton X-100. One entzyme unit is defined as the amount of

enzyme oxidizing 1 pm01 of guaiacol/min based on an extinction coefficientof 26.6 rnM_’cm-’ for tetraguaiacol. Incubation of N-OH-24or -30)FAA with HRPIH~OZ or cytochrome clHgOa Incubations were at RT in vials protected from light. The ingredients were a’ddedin the order given to a total volume of 3 ml. Incubations with 0.16 PM HRP were carried out in 0.05 M sodium phosphate buffer (pH 7.5) to which 0.031 to 0.167mM N-OH-2-(or -S)FAA in 0.02ml methanol and 0.6mM HzOz were added. Cytochrome c, 8.9 PM, was incubated for 20 min with 0.015% (v/v) Triton X-100 and 0.6mM Hz02 in 0.05 M sodium phosphate buffer (pH 7.5) before the addition of N-OH-2-(or -3.)FAA. Incubations of the hydroxamic acids with cytochrome c were also carried out in 0.1 M sodium acetate buffer (pH 6.0) and 0.6 mM Hz02 without preincubation with Triton X-100. The reaction was stopped by the addition of an equal volume of ice-cold Nz-preflushed CH&lz and vortexing for 1 min. The aqueous layer was separated by centrifugation and re-extracted. The CHzClz layers were dried over anhydrous Na2S04 and evaporated under Na which had passed through Oxyclear Pierce Chemical Co., Rockford, IL). Interactions of N-substituted fluorenyl compounds with lecithin ESR was used to detect the interaction products of lecithin with the N-substituted fluorenyl compounds per se or formed in the incubation mixtures. Lecithin solution in CHCls was evaporated to dryness under Nz immediately before use. In a total volume of 6.0 ml, 0.015% (v/v) Briton X-100, 8 9 PM cytochrome c, and 0.6 mM Hz02 were incubated 20 min at RT in 0.05 M sodium phosphate buffer (pH 7.5). Immediately before use the fluorenyl compound was dissolved in Nppreflushed methanol to give a 10 mM solution and added to the above with stirring. Three milliliters of the reaction mixture were immediately (-5 s) scanned for free nitroxyl radical (system A]. Scanning was continued for 1 h. The remaining 3.0 ml of the reaction mixture were incubated 1 h at 37°C with fresh lecithin, 1 mglml. and then scanned for nitroxyl radicals (system B). In the case of N-OH-3-F.q..\,it was necessary to use 0.16 PM HRP, omitting Triton X-100 to obtain the ESR spectrum with lecithin. Control incubations consi5Ledof lecithin in 0.05M sodium phosphate buffer (pH 7.5) and the fluorenyl compound in methanol. Chromatography Dried extracts of reaction mixtures were taken up in diethyl ether and spotted on Whatman linear-K high performance thin layer plates (10 x 10 cm) with a fluorescent indicator (Pierce Chemical Co., Rockford, IL). The short bed continuous development TLC apparatus manufactured by Regis Chemical Co., Morton Grove, IL was used and covered with aluminum foil to protect from light. The development of the plates was in CH&l&enzene/acetone = 42.5:5:2.5 (solvent A). To separate 2- or 3-NOzF from 2- or 3=NOF, respectively, portions of the extracts were chromatographed with 50 ml of CH&l~acetone/glacial acetic acidlhexane =

321

17: 1: 1.5:312 (solvent B). Compounds were located with shortwave UV light, the appropriate areas scraped and extracted twice with CH&L. The CH& was evaporated under Nz which had passed through Oxyclear. The dried extracts were dissolved in Nppreflushed methanol and their contents determined spectroscopically. Corrections were made for the recovery of standard compounds subjected to the same incubation, extraction and chromatographic procedures.

Sp?ctlm?copy

UV and visible absorption was recorded with a Hitachi 110 A spectrophotometer (Hitachi Instruments, Inc., Mountain View, CA), using the peak valley sensor program. ESR spectra were obtained with a Varian E-4 band spectrometer (Ovarian Instrument Group, Palo Alto, CA) operating at 100 KHz modulation and using a TM ilo cavity A vacuum was used to draw the sample into the large aqueous cell through stainless steel tubing inserted into its bottom. This apparatus enabled monitoring to begin in less than 5 s after initiation of the reaction. Scanning was at 3365G field strength and with a 40 G or 100 G range at RT. The concentration of nitroxyl free radical as a function of time was assayed by setting the magnetic field at the low field maximum which varied slightly with the different concentrations of the hydroxamic acids and different buffers.

Determination of concentration and kinetic characteristics of the nitroxyl free r-adica.?s To quantitate the free radical we used spectral simulations to relate the concentration of the standard spin probe to that of the nitroxyl free radicals of the hydroxamic acids by comparisons of the low field maxima [28]. Simulations were performed on a Cyber 74 Computer (Control Data, Minneapolis, MN) and included the effects of the modulation amplitude i29] and of the unresolved proton hyperflne splittings [30] upon the signal amplitude. We used the previously reported 1311proton splitting values for the standard. However, these values were unknown for t e nitroxyl free radicals of N-OH-29(or -3.)FAA. An effective hyperline splitting value was determined for the 12 protons [30] on these radicals after correction for the high modulation amplitudes used [29]. The amplitude of the ESR low field peak of the standard probe was 5.3. and 9.6-fold greater, respectively, than that of the signals from N-OH-2,FAA and N-OH-&FAA. The time it took for the amplitude to decrease to 50% of its maximum was defined as tAi and comprised several factors such as rates of radical formation and decay. enzyme inactivation and decreasing substrate concentration. Signal persistence was defined as the length of time during which the signal of the radical species could be detected. Initial velocity was measured directly from the trace of the reaction with respect to time.

RESULTS

The peroxidase activity of HRP is believed to be responsible for the oxidation of N.OH.B.FAA to a nitroxyl free radical [3]. Cytochrome c has, likewise, been reported to have peroxidase activity 1321 and appeared to act as a peroxidase in our system. Thus, we compared their relative peroxidstive activities under the conditions of our incubations using guaiacol as a substrate [27]. The peroxidase activity of HRP was 3000=fold greater than that of cytochrome c (47 000 vs. 16 unitslpmol, respectively). Cytochrome c had the same activity whether assayed in the buffer (pH 7.5) after preincubation with Triton X-100 for 20 min or in the buffer (pH 6.0) without pretreatment with Triton X-100. Products of the interaction of HRP or cytochrome c and I-I& with N-OH-2.FAA and N-OH-3.FAA in various buffer systems were extracted into CHzC12, separated by TLC and analyzed by WMS spectroscopy (Table I). In agreement with previous reports, oxidation of N=OH-2. FAA in the HRP/H202 system yielded equimolar quantities of 2.NOF and N.AcO-2.FAA, i.e., the dismutation products of the nitroxyl free radical of N-OH-2-FAA [3,5]. Oxidation of the isomeric hydroxamic acid, N=OH-3. FAA, by HRFVH202 also yielded equimolar quantities of the corresponding 3-NOF and N-AcO-3-FAA suggesting that the nitroxyl free radical of N-OH3.FAA was likewise, an intermediate. The amounts of 3.NOF and N-AcO-3. FAA were nearly the same as the corresponding products formed from N-OH-BFAA (Table I). In addition, small amounts of 2.FAA and 3.FAA were isolated from the reactions with both substrates. Trace to small quantities of the acetate esters, nitroso compounds and amides were also found in the extracts of the incubatiton mixtures from which Hz02 was omitted. The quantities of these products varied with :-he lot of HRP used. That no products were found when HRP was omitted from the incubation system suggested involvement of a factor inherent to HRP. By swctrally monitoring the reaction to establish conditions for the peroxidative activity of cytochrome c, we found that N-OH-2.FAA was converted to ZNOF only when Triton X-100 was added or an acidic buffer was used. At the acidic pH, the yields of N-AcO-2.FAA and 2-NOF from N-OH-O-FAA approached those obtained witb the HRP/H20~ system, but large quantities of 2-FAA were formed (Table I). The yields of these products in the cytochrorne c/I-I202 system at pH 7.5 in the presence of Triton X-100 were 3-4.fold lower than those at pH6.0. Incubation of the same concentration of N-OH-&FAA 10.167 mM) with cytochrome c/Hz02 at pH 7.5 in the presence of Triton X-100 yielded only trace amounts of N-AcO-3-FAA and 3.NOF unless the in. cub&on time was extended to 30 min. Then, the amounts of these products were similar to those formed at pH 6.0 and to the amounts of N.Ac0.2.FAA and 2.NOF formed from N-OH-B-FAA. The amounts of 3.FAA formed by cytochrome cm202 system at pH 6.0 were U-fold greater than those formed at pH 7.5. When the concentration of N-OH-SFAA in the system at pH 6.0 was decreased from 6.167 to 6.631 mM, the amounts of N-AcO.3.FAA and 3.NOF calculated in moles per mole of the substrate increased 5. and Cfold, respectively, and

323 TABLE I PRODUCTS FROM THE INTERACTION HRP/H& OR CYTOCHROME c/HsOs Substrate 6nM)

N.OH.2.FAA (0.167)

N.OH.3.FAA (0.167)

N-OH-3.FAA (0.031) -

Catalytic system

HRP/H~~

Buffer system@

OF

Incubation time (mitt)

N-OH-P-FAA

AND

N.OH.BFAA

WITH

Products (mollmol substrateJb N-AcO.FAA

NOF

FAA

NW

HzOZ

A A A A

10 10 .~0.08 10

0.36 traced 0.006 0

0.3g trace 0.011 0

o.oosc 0.027 0.002 0

0 0 -e -

Cyt. cf/H202 Cyt. c

B B

10 10

0.085 0

0.108 0

0.056 trace

0 0

Cyt. c/Hz02 Cyt. c

C C

10 10

0.30 0

0.43 0

0.23 trace

0 -

HRP/HzOz

A A

10 10

0.35 trace

0.32 0

0.013 trace

0 0

Cyt. c/Hz02 Cyt. c

B B

10 10

0.007 0

trace 0

0.187 0.01

0 0

Cyt. c/Hz02 Cyt. c

B B

30 30

0.074 0

0.092 0

0.212 0

-

Cyt. c/H202 Cyt. c

C C

10 10

0.0% 0

0.092 0

0.318 0.014

0 0

Cyt. c/HsOs Cyt. c

C C

10 10

0.26 0

0.35 0

0.42 0.05

-

“A: 0.05 M sodium phosphate (pH 7.5); B: 0.05 M sodium phosphate (pH 7.5) 0.015% (v/v) Triton X-100; C: 0.1 M sodium acetate @H 6.0). bSeparated by TLC, identified and quantified by UV spectroscopy. The quantities were corrected for percent recovery of standard compounds. ‘Average of 2 determinations. dIdentiikd by TLC against standard compounds; amount insufficient for spectral quantification. “Not analyzed in this system. ‘Cyt. c, cytochrome c. @Average of 3 determinations.

approached those obtained in the same system from 0.167 mM N-OH-2.FAA. The yield of 3.FAA increased 1.3-fold with the decrease of the substrate. These results suggested that the peroxidative activity of cytochrome c was saturated at the lower concentration of N-OH-&FAA an /or the rate of decay of the intermediate nitroxyl free radical from N-OH-S-FAA was diff erent from that of N=OH-2-FAA. The origin of +.he relatively large amounts of the amides, in particular of 3.FAA, isolated from cytochrome c/H202 systems is presently unknown though the reduction of ;he hydroxamic acids by the reduced cytochrome c is conceivable. The acetate esters, nitroso compounds and amides represented the major spots found on the chromatograms. Several additional spots could be visualized by short wavelength W, but contained insufficient

324

material for spectral identification. None of these comigrated with the nitro compounds indicating that no further oxidation of the nitroso compounds OCCUlTed. ESR spectroscopy was

used to show that nitroxyl free radicals, the presumed precursors of the acetate esters and nitrosofluorenes, were indeed generated from N-OH-B-FAA and N-OH-&FAA by HRI’/HcOc and cytochrome c/&Oz systems. The ESR spectra of the nitroxyl free radicals generated from the interaction of N-OH-O-FAA and N=OH-&FAA with cytochrome C/B&O,in the presence of Triton X-100 (Figs. lB,Cl were similar to each other and to those generated with HRP/H202. The calculated g value of 2.0066 and nitrogen splitting constant of 7.6 Gauss were similar to the values previously reported for the nitroxyl free radical of N-OH-S-FAA 1331. However, the width of the low field line in the spectrum of the nitroxyl free radical of N-OH-SFAA (4.0 G) was broader than that of N-OH-B-FAA (3.4 G). In the cytochrome c&O2 system the triplets typical , _ nhindered nitroxyl free radicals were generated from N-OH-a-FAA within a broader range of the concentration of the compound (0.033-0.33 mM1 than from N-OH-SFAA (0.033~O.XIIIM), suggesting saturation of the enzyme by the latter. In HRP/H& system, an ESR signal was seen with either compound at a concentrationof 0.33 mM. It was shown that nitroxyl free radicals of N-arylhydroxylamines and nitroso compounds interacted with double bonds of lipids yielding ESR signals of immobilized nitroxyl free radicals 110,341.Therefore, we examined the interaction with lecithin of the products generated by cytochrome c/Hz02 from the hydroxamic acids or of the individual N-substituted fluorenyl compoundsshown to arise from this oxidation (Table II). We added lecithin to the reaction mixture when the amplitude of the nitroxyl free radical of N-OH-S-FAA was either maximal or had almost disappeared and then incubated for 1 h at 37°C. In both reaction mixtures, the ESR signals of immobilized free radicals developed and were identical to that produced by the interaction of 2-NOF and lecithin (Fig. 1E; Table II). Attempts to produce this immobilized species from N-OH-3-FAA in the cytochrome c/HzOzsystem failed, even though a triplet of the nitroxyl free radical of N-OH-S-FAA was obtained when a low concentration of the hydroxamic acid (0.08mM) was used (Table IQ. Apparently, the amount, of 3-NOF formed from this radical was insufficientfor the product of its interaction with lecithin to be detected, When HRP&O~ which generated more 3-NOF (Table I) was used, the addition of lecithin led to the development of the spectrum of the immobilized free radical which was identical to that produced by interaction of 3-NOF and lecithin (Fig. IF; Table IIl. When the fluorenyl compounds were incubated with lecithin alone for 1 h at 37”C, only 2-NOF and 3-NOF gave ESR signals of immobilized free radicals (Table II). Neither the nitroso compounds nor lecithin alone yielded a free radical signal after incubation in the presence or absence of @ochrome c/HzOZ, indicating that the ESR spectra observed above were due to the interaction of the nitroso compounds with lecithin. The nitroso comlpouuds,as shown by TLC, were present in the fresh extracts of

--lot---

Fig. 1. ESR spectra of nitroxyl free radicals of fluorenylhydroxamic acids generated by 8.9 PM cytochrome c, preincubated with O.OlS%, v/v, Triton X-100 and 0.6m.M HzO, for 20 min at RT in 0.06 M sodium phosphate buffer, pH 7.6 (A-C). Scans were over a 40 G range and centered at 3353 G field strength with a total scan time of 4 min. The modulation amplitude was 5 G and microwave power of 20 mW. No signsl was produced if one of the components of the incubation mixture was absent (A), 0.12111% N-Oif-2-F&4 (B) and 0.12 mM N-OH-S-FAA (C) were incubated as above. ESR spectra of immobilizd nitroxyl free radicals obtained from the incubation of lecithin (1 n&ml) with h’OF for 1 b at 37°C (D-F). Scans were over a 100 G range centered at 3365 G field strength wit+. a scan time of 16 min. a modulation amplitude of 8 G and a microwave power of 26 mW. No r.ignal v_as produced by NOF or lecithin alone 0). 0.33 n&f 2-NOF (E) and 0.33 mM 3.NOF (F) were incubated with lecithin.

326 TABLE II TYPES OF ESR SIGNALS FROM INTERACTION OF N.SUBSTITUTED FLUORENYL COMPOUNDS OR OF THEIR CYTOCHROME c/HzOz-GENERATED PRODUCTS WITH LECITHIN Fluorenyl compound

Cont. tmM

Type of ESR signal Systemm A

B

C

N-OH&?-FAA N.OH-3-FAA N-OH&FAA

0.33 0.33 0.08

Triplet None Triplet

Immobilized None Immobilizedb

None None None

ZNOF 3.NOF 5NOF

0.33 0.33 0.08

None None None

Immobilized Immobilized _c

Immobilized Immobilized

ZNOfl 5NOP

None None None

None None None

None None

5N02fi

0.33 0.33 0.08

N-A&Z-FAA N-AcG-2-FAA N-AcO-SFAA N-AcO&FAA

0.33 0.17 0.33 0.17

None None Tripletd Tripletd

None

None

None None

None

None’

0

None

*A: Triton X-100. cytochrome c and HzOz, at concentrations described in Materials and Methods, were preincubated for 20 min in 0.05M sodium phosphate buffer @H 7.5). A methanolic solution of the guorenyl compound was added, and one-half the solution was scanned for nitroxyl free radical for up to 1 h; B lmg/ml lecithin was added to the remainder of the solution of system A which was then incubated at 37°C for 1 h. The solution was equilibrated at RT 5min before scanning for nitroxyl free radical; C: the fluorenyl compound was incubated with 1 mg/ml lecithin at 37°C for 1 h, equilibrated at RT and scanned for the radical. bJ.mmobilized spectrum was obtained when the nitroxyl free radical of N-OH-S-FAA was generated by HRP/H&. but not when generated by cytochrome c/HzOz. ‘Not analyzed in this system. dBarely detectable. ‘Lecithin control.

the incubation mixtures (T.able I). After rechromatographing of the nitroso compounds or storage of the extracts, the nitro compounds, 2- and 3=NOzF, were also present. However, the nitro compounds produced no triplet or immobilized free radical signals in the systems examined (‘Table II) suggesting that they did not interact with double bonds of the lecithin. The other dismutation products, N-AcO-2-FAA and IV-AcO-3-FAA, likewise did not react with lecithin since no ESR signals of immobilized nitroxyl free radicals

327 were detected. In contrast to N-AcO-2-F4A, N-AcO-3-FAA yielded a barely detectable ESR signal of nitroxyl free radical in the cytochrome c/H202 system. However, the concentration of the free radical was insufficient for the detection of the interaction with lecithin. This nitroxyl free radical was probably due to the one electron oxidation of N-OH-8FAA. The latter was present in trace amounts on chromatograms of extracts of the incubation mixture of N-AcOa3-FAA with cytochrome c/HzOz. The diff erc”:;1cesin the amounts of the nitroso compounds and acetate esters generat& by HRP/HzOz or cytochrome c/H&z from N-OH-2.FAA and N-OH3-FAA suggested differences in the formation and kinetic properties of their respective nitroxyl free radicals as intermediates. Hence, we compared timecourse of generation and decay of nitroxyl free radical signals from N-OH-2FAA and N-OH-3-FA.4 in the cytochrome c/H202 system (Fig. 2) and the kinetic characteristics of these nitroxyl free radicals in HRP/H& and cytochrome c/Hz02 systems (Table III). Preliminary experiments showed that a 20,min incubation of cytochrome c, Hz02 and Triton X-100 before the addition of the hydroxamic acids prevented a lag phase in the reaction velocity. In this system, higher concentrations of nitroxyl free radicals, as shown by the maximum amplitudes of the signals, accumulated from N-OH2-FAA than from N-OH-8FAA. In the cytochrome c/Hz02 system, in which the initial velocity was measurable, the faster rate, about -&fold, of formation of the nitroxyl free radical of N-OHS-FAA probably contributed to

TABLE III KINETIC CHARACTERISTICS DERIVED FROM ESR TIME-COURSE NITROXYL FREE RADICALS OF IV-OH&FAA AND N.OH.3.FAA

SCANS

OF

The nitroxyl free radicals were generated by HRPIHsOs (A) or cytochrome c/HsOs (B) in 0.95 M sodium phosphate buffer @H 7.5) aa described in Materials and Methods.

Kinetic characteristics~

N=OH.ZFAQ (n&I)

N.OH3.FAA

0.167

0.167

0.033

A Initial velocity (ccM a-‘) x 10’ Maximum cone. (PM Signal persistence (min)d t A! (a)”

B

J

0.3

7.4#

0.492 1L.O

4.7

i.7.9

8.5

A

B 9.4

-

A .-

5.92

0.436

7.1’

3.6

4.1

10.1

10.6

47.2

55.2

(mM) 0.033 A

B 1.0 0.241 30.2 151

-

B 1.5

4.82

0.268

13.6

8.0

78.0

139.0

_--

‘Determined as descr’led in Mate, ials and Methods, All values are the averages of determinatio.Te from tripl; iate scans %eneratio,‘ of nitror yl free radical was too rapid for accurate measurement. Rapid generction r .ld decay of radical suggests that maximum may be greater than indicsted. dTime span whsV:n,he recorded signal was above the baseline. @Time for the sidnal to decrease to 50% of its maximum amplitude.

_

Fig.

TIME IMINI

(4 B) and N-OH-3-FAA (C, D). Triton X-100 (0.015%.v/v), 8.9 @l cytoeh~~e C, ad 0.6mM H& were incubated 20 min at RT. ~~~424~ -3-)FAA in methane!was added with stirring to give final concentrationsof 0.0'33HIM (B and D) or 0.167mM (A and 0. Recording Was begun *thin 5 8. The magneticfield was set at the positivemaxima Gf the low field transitionsand the modulationamplitude was 5 G with a mi_aVe Power Of 20 mw-

I

2. Time courseof generationand decay of nitroxyl free radical signals from cytochromec/H&ycat&zed oxidation of ~-OH_2FAA

31

-

329

the differences in concentrations of nitroxyl free radicals. Accordingly, the length of time that the radical species from N-OH-&FAA could be detected, i.e., its signal persistence, was 2-2.3-fold longer than that from N-OH-2.FAA and the time required for the ESR signal to decrease to half maximal amplitude, i.e., tA& was 34fold longer. The longevity of the ESR signal of the nitroxyl free radical of N-OH-&FAA seemed to reflect slower rate of formation of the radical. Comparison of the reaction at the high and low concentrations of either hydroxamic acid in the cytochrome c/HzOz system indicated that the initial velocities of formation and the maximal concentrations of radical signals were similar indicating that the cytochrome c system was saturated (Table III). The effects of this saturation can be seen in Fig. 2. The long flattened portions of the scans A and C represent the time periods where the rates of generation and disappearance of the radical were in equilibrium. The radical signals formed at the higher concentration persisted 3-3.8fold longer than the signals generated at lower concentrations (Fig. 2, scans B and D), reflecting the greater amount of substrate available. The long flat portions of the scans at th
We showed that a carcinogenic N-tluorenylacetohydroxamic acid, N-OH-3FAA, in the HRPIH202 system underwent one electron oxidation to the nitroxyl free radical followed by its dismutation to 3.NOF and N-AcO-3-FAA. This reaction sequence had been previously reported for oxidized N-OH-SFAA 1351. We also showed that both isoneric hydroxamic acids were oxidized to the nitroxyl free radicals by a cytochrome cl @2 system, from which the same dinmutation products were recovered. Cytochrome c was active as a peroxidase in the presence of Hz02 at pH 6.0 or at pH 7.5 after preincubation with Triton X-100 as tested with either guaiacol or the hydroxamic acids as substrates. O’Brien [36] reported the enhanced peroxidase activity of cytochrome c at low pH or in the presence of long chain fatty acids or ionic detergents and attributed this to an unfolding of the protein moiety around the heme. Interestingly, the cytochrome c had the same peroxidase activity in both buffer systems when guaiacol was the substrate. However, the rates of generation of the nitroxyl free radicals and the yields of their dismutation

330 products were more favorable in 0.1 M sodium acetate, pH 6.0, than in 0.05 M .&ium phosphate (pH 7.5) containing 0.015% (V/V) Triton X-100. This may reflect substrap- specificities of cytochrome c or may indicate that other factors such as pH of the buffer, presence of Triton X-100 and the high concentration of cytochrome c affect the generation and ultimate fate of the nitroxyl free radicals. In addition to the dismutation products, 2- and 3-FAA were also present in the extracts of the reaction mixtures following oxidation of the respective hydroxamic acids by HRP/H202 or cytochrome c/H202 system. The presence of B-FAA among the products of one electron oxidation of IV-OH-2-FAA was reported by other investigators [1,5] and the amount of 2-FAA formed varied with the peroxidase system used [5]. In our experiments, larger amounts of 29 and 3-FAA were recovered from cytochrome c/HsO,, especially at pH6.0, than from HRFVH202 system. In the absence of Hz02 the amounts of the amides, except in one experiment with HlW, were markedly decreased and no nitroxyl free radicals were detected from either N-OH-P-FAA or N-OH-3-

FAA under these conditions. The dependence of the generation of the large amounts of the amides with cytochrome c on the presence of Hz02 might be partially due to the effect of H202 on the unfolding of the heme. This is supported by the fact that preincubation of cytochrome c with both Triton X-100 and H202 was necessary to prevent the lag phase in the generation of the ESR signals. The larger amounts of the amide generated from N-OH-3FAA than from N-OH-ZFAA might have resulted from substrate specificity of cytochrome c. The source of the amides in the reaction mixtures involving one electron oxidation of fluorenylhydroxamic acids is at present unknown, although several possibilities may be considered. For-rester et al. [l] reported equimolar amounts of amides and nitro.compounds among the products isolated from several N-arylhydroxamic acids oxidized with silver oxide in benzene, They postulated an intermolecular oxygen-transfer reaction involving one of the dismutation products, i.e., the nitroso compound, and the nitroxyl free radical. However, in our reaction mixtures we did not detect the nitro compounds that would be expected if this mechanism was in operation. Nakagawa et al. [37] reported an amide as the major product from the oxidation of N-benzoyl-N-phenylhydroxylamine with nickel peroxide in ether. The postulated mechanism involved attack by a hydroxyl radical on the nitroxyl free radical resulting in an intermediate peroxy compound which decomposed to the amide. The amides in both of the above systems were generated in nonpolar organic solvents with non-protein catalysts. A possible source of the amid< in the cytochrome c/Hz02 system could be the reduction of the hydroxamic acid or the nitroxyl free radical. According to the supplier, about 10% of the cytochrome c is in the reduced form. However, generation of the amides is greatly enhanced in the presence of HSOS.This suggests that cytochrome c is reduced du.ring the peroxidative reaction and/or by the nitroxyl free radicals. Reduction of ferricytochrome c by free radicals was reported by Simic et al. [381. Conversely, Butterfield et al. [39] showed that

331 ferrocytochrome c reduced the small nitroxide spin label, 2,2,6,6_tetramethylpiperidin-1-oxyl-4.01 (Tempo]), and that the route of electron transfer probably involved a direct attack by the radical through the heme area. Hence, it is possible that cytochrome c was reduced during the course of radical formation and in turn reduced nitroxyl free radicals of the fluorenylhydroxamic acids to the amides, the formation of which would then depend on the generation of the radicals. The high concentrations of cytochrome c necessary to show peroxidase activity and the longevity of the radicals in this system increase the changes for interaction of the radical with protein which could lead to reactions other than dismutation. It has been postulated that nitroxyl free radicals of the N-arylhydroxylamines and the nitroso compounds interact with carbon-carbon double bonds of lipids and that these interactions are of importance in carcinogenesis by aromatic amines and amides 140,411. Floyd and coworkers [34,42] showed that 29NOF adds to a double bond in unsaturated fatty acids, most likely in the Alder-ene reaction with a shift of double bond to the adjacent position. They assumed that the resulting adduct undergoes oxidation to a nitroxyl free radical that gives an ESR signal corresponding to the spectrum of immobilized free radical. Alternatively, a similar type of ESR spectrum might be obtained on a direct addition of the nitroxyl free radicals of N-arylhydroxylamines to double bonds of lipids with their concurrent saturation [lo]. We showed that both 2. and 3-NOF yielded ESR signals of the nitroxyl free radicals upon their interaction with lecithin without prior activation by a peroxidase system. These signals were also obtained when lecithin was added to the reaction mixtures of the hydroxamic acids, cytochrome c or HRP, and HzOz, after the triplets of their nitroxyl free radicals had nearly disappeared and the nitroso compounds were present. Unless reasonably high concentrations of the nitroxyl free radicals were generated to yield sufficient amounts of the nitroso compounds for interaction with lecithin, the ESR signals of immobilized nitroxyl freti radicals were not detectable. Neither the nitro compounds nor the other dismutstion products, i.e., the acetate esters were activated in this fashion. Hence, our data support the notion that the nitroso compounds rather than the nitroxyl ‘ree radicals of the hydroxamic acids are reactive with lipids. The mechanism(s) by which N-OH-B-FAA and N-OH9-FAA act as carcinogens is as yet unclear. Several mechanisms of activation of N-OH-ZFAA to the electrophilic reactants capable of covalent interactions with nucleic acids, which are the presumeri critical macromolecules in initiation of carcinogenesis, have been demon&rated and reviewed [9]. In contrast to that of N-OH-%FAA, the activation of N-OH-3-FAA via mechanisms such as :;11fation of the hydroxamic acid to the N-O-sulfate 1221or N,O-acyltransieraseintramolecular N,O-acetyltransfer to yield N-acetoxycatalyzed fluorenamine (C.M. King, pers. comm.) failed to occur. In this work la-e showed that one electron oxidation of N-OH-39FAA to its nitroxyl free radical took place in the same systems as the oxidation of M-OH-B-FAA. Likewise, the nitroxyl free radical of N-OH-3.FAA dismutated to 3-NOF and

N.Ac0.3.FAA. Even though N-AcO-3-FAA did not interact with macromolecules in the same fashion as did N-AcO-2-FAA 122,431, it was capable of extensive acetylation of nucleophilic sites in protein and thus could potentially modify critical proteins such as chromatin 1441.Whether the carcinogenesis by N-fluorenylacetohydroxamic acids is associated with the acetate esters and nitroso compounds generated via dismutation of the nitroxyl free radicals and/or the radicals themselves remains to be investigated. The differences in the kinetic properties of the nitroxyl free radicals of N-OH-B-FAA and N-OH-3.FA.A, shown in this study, may determine their reactivities and the extent of subcellular interactions leading to cell damage. Moreover, the different amounts of the nitroxyl free radicals and of their rqective dismutation products generated under different conditions may underlie capacities of various tissues to produce these radicals. It has been shown that nitroxyl free radicals of carcinogenic N-arylhydroxylamines and N-arylhydroxamic acids are readily generated by hepatic microsomes, probably by cytochrome P-450 system [10,11,41,451. However, it must be noted that hepatic microsomes of species and strains both susceptible and nonsusceptible to hepatocarcinogenesis were capable of generating these nitroxyl free radicals. Perhaps, the knowledge of the fate of these radicals in suscep. tible and norrsusceptible liver may more closely determine their role in tumorigenesis. In extrahepatic targets for tumor induction, such as mammary gland, cytochrome P-450 is present in extremely small amounts [17] and probablig other peroxidases are involved in generation of the nitroxyl free radicals as shown with N-OH-2-FAA 112,131. Since cytochrome c is a multifunctional enzyme with oxidizing and reducing activities and is presmt in the rat mammary gland in relatively large amounts 1451, it may play a role in the generation and metabolism of the nitroxyl free radicals of carcinogenic iV-arylacylhydroxamic acids in this target tissue. ACKNOWLEDGEMENT This work was supported by the U.S. Veterans Administration and by PHS Grant CA-28000 awarded by the National Cancer Institute, DHHS. REFERENCES 1 A.R.Forreeter,MM. OgilvyandR.H.Thomson,Modeof action of carcinogenic amines. Part I. Oxidation of N-arylhydroxamic acids, J. Chem. Sot., C (1970) 1081. 2 H. Bartseh, M. Traut and E. Hecker, On the metabolic activation of N-hydroxy-N-2. acetylaminofluorene. II. Simultaneous formation of Znitrosofluorene and N-acetoxv-N-2. acetyhuninofluorene from N-hydroxy-N-2acetylaminofluorene via a free radical inter. mediate, B&him. Biouhvs. Acta. 237 (1971) 586. 3 H. Bartsch and E. H&ker, On the metabolic activation of the carcinogen N-hydroxy-N-2acetylaminofluorene. III. Oxidation with horseradish peroxidase to yield 2.nitrosofluorene and N-acetoxy-N-Zacetylaminofluorene, B&him. Biophys. Acta, 237 (1971) 678.

333 4 R.N. Walker and R.A. Floyd, Free radical activation of N-hydroxy-2-acetylaminofiuorene by methemoglobin and hydrogen peroxide, Cancer Biochem. Biophys.. 4 (1979) 87. 5 R.A. Floyd. L.M. Soong and P.L. Culver, Horseradish peroxidase/hydrogen peroxide. catalyzed oxidation of the carcinogen N-hydroxy-N-acetyl-2-aminofluorene as effected by cyanide and ascorbate, Cancer Res.. 36 (1976) 1610. 6 P.D. Lotlikar, J.D. Scribner, J.A. Miller and E.C. Miller, Reactions of esters of aromatic: N-hydroxy amines and amides with methionine in vitro: a mode1 for in vivo binding of amine carcinogens to protein, Life Sci.. 5 (1966) 1263. 7 E. Kriek, J.A. Miller, U. Juhl and EC. Miller. 8-(N-2-Fluorenylacetaido)guanosine, an awlamidation reaction product of guanosine and the carcinogen N-acetoxy.N&$fluorenyl. acetamide in neutral solution, Biochemistry, 6 (1967) 177. 8 E.J. Barry. D. Malejka-Giganti and H.R. Gutmann. Interaction of aromatic amines with rat-liver proteins in vivo. III. Gn the mechanism of binding of the carcinogens, N-2. fluorenylacetamide and N-hydroxy-24uorenylacetamide. to the soluble proteins, Chem.-Biol. Interact., 1 (1969/70) 139. 9 E.C. Miller and J.A. Miller, The metabolism of chemical carcinogens to reactive electrophiles and their possible mechanisms of action in carcinogenesis, in: C.E. Searle (Ed.), Chemical Carcinogens, ACS Monograph 173, Washington, D.C. 1976, pp. 737-762. 10 A. Stier, R. Clauss, A. Lticke and I. Reitz. Redox cycle of stable mixed nitroxides formed from carcinogenic aromatic amines. Xenobiotica. 10 (1980) 661. 11 D.L. Reigh and R.A. Floyd. Evidence for a cytochrome P-420 catalyzed mechanism of activation of N-hydroxy-2.acetylaminofluorene, Cancer Biochem. Biophys.. 5 (1981) 213. 12 D.L. Reigh. M. Stuart and R.A. Floyd, Activation of the carcinogen N-hydroxy-2.acetylaminotluorene by rat mammary peroxidase, Expsrientia, 34 (1978) 107. 13 P.K. Wang, M.J. Hampton and R.A. Floyd, Evidence for lipoxygenase-peroxidase activation of N-hydmxy-2-acetylaminofluorene by rat mammary pland parenchymal cells, in: Prostaglandins and Cancer: First International Conference, Alan R. Liss, Inc., New York, 1982. pp. 167-179. 14 E.C. Miller, J.A. Miller and H.A. Hartmann. N-Hydroxy-Zacetylaminofluorene: A metabolite of 2-acetylaminoguorene qx.ich increased carcinogenic activity in the rst. Cancer Res.. 21 (1961) 815. 15 H.R. Gutmann. D.S. Leaf, Y. Yost, R.E. Rydell and CC. Chen, Structure-activity relationships of N-acylarylhydroxylamines in the rat, Cancer Res., 30 (1970) 1485. 16 D. Malejka-Giganti, R.E. Rydell and H.R. Gutmann, Mammary cmcinogenesis in the rat by topical application of fluorenylhydroxamic acids and their acetates, Cancer Res.. 37 (1977) 111. 17 CL. Ritter and D. Malejka-Giganti. Mixed function oxidase in the rat mammary gland and liver microsomes of lactating rats. Effects of 3-methylcholanthrene and P-naphthotlavone. Biochem. Pharmacol., 31(1982) 239. 18 H.R. Gutmann and R.R. Erickson, The conversion of the carcinogen N-hydroxy-2-fluorenylacetamide to o-amidophenols by rat liver in vitro, J. Biol. Chem.. 244 (1969) 1729. 19 E.K. Weisburger and J.H. Weisburger. An improved synthesis of 1-aminofluorena. J. Org. Chem., 18 (1963) 864. 20 Y. Yost, Oxidation of fluorenamines and preparation of 2,2’- and 4.4’-azofluorene. J. Med. Chem., 12 (1969) 961. 21 Y. Yost and H.R. Gutmann. Flu,irenylhydroxamic acids isomeric with the carcinogen Nfluoren-2-ylacetohydroxamic acid I. Synthesis of N-fluoren-1-yl-, N-fluoren-3-yl. and Nfluoren-4.ylacetohydroxamic acid J. Chem. Sot.. C (1969) 345. 22 F.J. Zieve and H.R. Gutmann Reactivities of the carcinogens. N-hydroxy-2-guorenylacetamide and N-hydroxy-3.flucrenylacetamide, with tissue nucleophiles. Cancer Res.. 31 (1971) 471. 23 D. Malejka.Giganti, HR. Gutmann. R.E. Rydell and Y. Yost, Activation of the carcinogen. by desulConylation to N-2-fluorenylhydroxylN-hydroxy-2-fluorenylbenzenesulfonamide. amine in vivo. Cancer Res., 31 (1971) 778.

334 24 D. Malejka-Giganti, H.R. Gutmann and R.E. Rydell, Mammarycarcinogeneeia in the rat by topical applic,ation of fluorenylhydroxamic acids, Cancer Res., 33 (1973) 2489. 25 B.F. Van Gelder and E.C. Slater, The extinction coefficient of cytochrome c, Biochim. Biophys. Acta. 58 (1962) 593. 26 G.R. Schonbaum, New complexes of peroxidases with hydroxamic acids, hydraxinee, and amides, J. Biol. Chem., 248 (1973) 562. 27 A.C. Maehly and B. Chance, The assay of catalases and peroxidases, in: D. Glick {Ed.), Methods of Biochemical Analysis, Vol. 1, Inter-science Publishers, Inc., New York, 1954, pp. 357-424. 28 H. Zeldes and R. Livingston, The use of spectral simulations for the assay of radical concentrations by ESR. J. Magn. Reson., 49 (1982) 84. 29 H. Wahlquist, Modula”.ion broadening 0; unsaturated Lorentxian lines, J. Chem. Phys., 36 (1961) 1708. 30 B.L. Bales, Correction Toi-inhomogeneous line broadening in spin labels II. J. Magn. Reson., 48 (1982) 416. 31 J.J. WindIe, Hyperfine coupling constants for nitroxide spin probes in water and carbon tetrachloride, J. Magn. Resort.. 45 (1981) 432. 32 A. Albert and S.E. Falk, The formation of hydrogen carriers by haematin-catalyzed peroxidations. 1. Hydrogen carriers from certain a&dine and quinoline compounds, Biochemietry, 44 (1949) 129. 33 R.A. Floyd and L.M. Soong, Obligatory free radical intermediate in the oxidative activation of the carcinogen N-hydroxy-2_acetylaminofluorene, Biochim. Biophys. Acta, 498 (1977) 244. 34 R.A. Floyd, Free radicals produced in a nitrosofluorene-unsaturated lipid reaction, Experien. tia, 33 (1977) 197. 35 H. Bartsch, J.A. Miller and E.C. Miller, N-Acetoxy-N-acetyl-aminarenes and nitrosoarenes. One-electron non-enzymatic and enzymatic oxidation products of various carcinogenic aromatic acethydroxamic acids, Bicchim. Biophys. Acta, 273 (1972) 46. 36 P.J. O’Brien, Intracellular mechanisms for the decomposition of a lipid peroxide. 1. Decomposition of a lipid peroxide by metal ions, heme compounds, and nucleophiles, Can. J. Biochem.. 47 (1969) 485. 37 K. Nakagawa, H. Onoue and K. Minami, Oxidation with nickel peroxide. VI. Oxidation of N-substituted hydroxylamine derivatives with nickel peroxide, Chem. Pharm. Bull., 17 (1969) 835. 38 M.G. Simic. LA. Taub. J. Tocci and P.A. Hurwitx, Free radical reduction of ferricytochrome c, Biochem. Biophys. Res. Commun., 62 (1975) 161. 39 D.A. Butterfield, A.L. Crumbliss and D.B. Chesnut, Radical decay kinetics in ferrocyto. chrome c model membranes. A spin label study, J. Am. Chem. Sot., 97 (1975) 1388. 40 R.A. Floyd, Observations on nitroxyl free radicals in arylamine carcinogenesis and on spin-trapping hydroxyl free radicals, Can. J. Chem., 66 (1962) 1577. 41 A. Stier, R. Clause, A. Liicke and I. Reitz. Radicals in carcinogenesis by aromatic amines, in: D.C.H. McBrien and T.F. Slater (Ede.), Academic Press, London, 1932, pp. 329-351. 42 R.A. Floyd, L.M. Soong, M.A. Stuart and D.L. Reigh, Free radicals and carcinogenesis. Some properties of the nitroxyl free radicals produced by covalent binding of %nitrosofluorene to unsaturated lipids of membranes, Arch. Biochem. Biophys.. 165 (1978) 456. 43 i. Y*at, H.R. Gutmann and R.E. Rydell. The carcinogenicity of fluorenylhydroxamic acids and N-acetoxy-N-fluorenylacetamides for the rat as related to the reactivity of the esters toward nuclel>philes, Cancer Res., 35 (1975) 447. 44 E.J. Barry and H.R. Gutmann, Protein modifications by activated carcinogens. II. The acetylation of ribonuclease by N-acetoxy&luorenylacetamide, Chem..Biol. Interact., 13 (1976) 47. 45 D. Malejka-Gliganti, C.L. Ritter and CM. Ryzewski, Pathobiologic and metabolic aspects of mammary gland tumorigenesis by N-substituted aryl compounds, Environ. Health Perspect., 49 (1963) 175.