Chem.-Biol. Interactions, 26 (19’?9) 339-347 0 EIsevier/North-Holland Scientific Publishers Ltd.
339
J.L. WOOD, C.L. BARKER and CJ. GRUBBS*
Department of Biochemistry , University Sciences, Memphis, TN 38163 (U.S.A.: 1(Received August 21st, 1973) (Revision received December l&h, (Accepted March 12th, 1979)
of
Tennessee,
Center
for
the Health
19’78)
SUMMARY
The principal products of the photooxidation of 7,12dimethylbenz[o] anthracene @MBA) in aqueom solutions by photooxidation induced by laboratory lighting tlave been characterized by high performance liquid chromatograms jEPLCj, ultraviolet and mass spectrograms and by comparisons with authentic samples. The products identified were the 7,12epidioxy-7,12-dihydro-7,12-dimethyl-, 7,12-dione, ‘I-hydroxymethyl-12methyl-, 12-hydroxymethyl-7-methyl-, ‘I-formyl-f2-methyl-, X2-formyl-7methyl-, and 12-hydroxy-12methyl-7-one derivatives of benz[c]-anthracene. The HPLC profile of products is similar to that obtained from oxidation of DMBA by ‘one-electron’ reagents, singlet oxygen, or liver microsomal
metabolism. The first points of attack are the 7- and 12- positions. The mechanism of photooxidation appears t;s be generation of singlet oxygen by photodynamic effect of DMBA. None of the products is photosensitizing, however.
INTRODUCTION
Photodynamic effects in biological systems and sensitivity of polycyclic hydrocarbons to destruction by light have been known for a long time. Until recently the modification of aromatic hydrocarbons by relatively short exposure to laboratory lighting during experimental manipulations has not been well recognized. Previous work in this laboratory has-shown that photooxidation of polycyclic aromatic hydrocarbons by visible light occurs rea *Present address: Division 60616,
L, i’!T Research
Institute,
10th West 35th Street, Chicago, IL
U.S.A.
Abbreviations: c:hrom tography.
DMBA, 7,12_dimethylbenz[a]anthracene;
HFLC, high performance
liquid
when air and moisture are present [ 11. Some of the products of the oxidation were effective in inactivating the sulfhydryl-dependent enzymes, lactic dehydrogenase and glyceraldehyde3-phosphate dehydrogenase, and also to bind to protein [2]. We have determined the chemical natures of the principal products of the photooxidation of DMBA in water suspension. Comparisons of the compounds formed by one-electron oxidations, by singlet oxygen, and by metabolism of DMBA by microsomes have been made and some similarities noted. MATERIALSANDMETHODS DMBA (Eastman) was recrystallized before use. When chromatographed LC, the compound emerged as a single sharp peak. Benz[a]anthraceneby 7,12-dions was obtained from Eastman. DMBA-7,12-photoperoxide (7,12epidioxy-7,.12,dihydro-7,12dimethylbenz[a] anthracene) was prepared according to the method of Sandin and Fieser [3]. cis-7,12Dihydro-7,12dihydmxy-T,l Bdimethylbenz [ a ] anthracene was prepared by catalytic hydrogenation of the 7,12-peroxide over 10% palladium on charcoal [ 41. The 12hydwoxy-lZmethylbenz[a]anthrone was prepared by the action of methyl magnesium iodide on benz[a]anthracene-7,12-dione [4]. T-Hydroxym&hyl-12-methylbenz [a] anthracene and 7,12-dihydroxymethyl-benz[u] anthracene were supplied by Dr. Charles Irving. ‘I-Formyl-la-methyl- and 12formy l-T-methylbenz[a]anthracene were prepared by the method of Badger and Cook [5]. All of the above compounds were characterized by their melting points, ultraviolet absorption and mass spectra. Singlet oxygen was generated by the method of Foote et al. [6]. To a solution of hydrogen peroxide (36 pmol in 3.4 ml of water cooled in an ice bath) was added 1 pmol of DMBA in 1 ml of cold methanol. The suspension was treated with 3 pmol of cold sodium hypochlorite (Purex 14) dropwise with vigorous stirring. After 5 min the solution w&aevaporated to dryness in vacua and the residue was taken up in 2 ml of methanol and subjected to fractionation by HPLC (Fig. 1C). HPLC chromatograms were obtained with a Waters Model 6000 chromatograph equipped with a quarter inch C II Bondapak column 30 cm long (Waters Associates, Milford, MA). Products of DMBA photooxidation were eluted from the column by a reverse phase gradient system, methanol/water, 60-100% methanol. Gradient rate of change was 5%/min with a flow rate of 2 ml/min. Emergence of fractions from the column was detected by an ultraviolet absorption m&nitor which utilized a 254 nm band. Separation of the photooxidation components required recycling from three to four times. The percentage distrubution of the material in the peaks was determined, through the courtesy of Dr. Byron Leach, on a Hewlett-Packard Liquid ChromatogTaph No. 1084a equipped qrith a CIs Bondapak column, 75 cm long, and fed into a data processor. Low resolution mass spectrograms were obtained with a Finnegan Quadrapole 1015 Spectrometer equipped with a 9500 gas chroma&ograpb input and a No. 6000 data system.
341 The photooxidation pocedure involved exposure of 1 pm01 of DMBA, or other compound, in 5 ml of 10% methanol in a crystallizing dish, for 1.5 h on the desk top in the laboratory. The light was provided by two General Electric Cool White fluorescent tubes that had a spectral o ranging from 340 to 750 nm with a small amount of radiation from the mercury resonance lines. Most of the energy was in the spectrum from 400 to 650 nm. The incident light on the desk top was 1350 lux. After irradiation the solvent was evaporated in vacua and the residue was taken up in methanol for injection into the HPLC apparatus. Oxidations of DMBA by ‘one-electron’ reagents were carried out as described by Schumm [4]. DMBA (10 mg) in aqueous acetone solution was treated with a 1 molar ratio of ceric ammonium nitrate or 4 mol of ferric ferricyanide overnight at room temperature. Similarly, 10 mg of DMBA in benzene were stirred with 50 mg of activated manganese dioxide for 22 h. The reaction products of each mixture were extracted into ethyl acetate, dried and transferred to methanol for HPLC fractionation.
Metabolism studies Male albino rats were injected intraperitoneally with 1 mg of methylcholanthrene 48 h prior to use. Liver microsomes were prepared by the method of Kinoshita et al. [ 71. Microsomal pellets were ob+tained by ultracentrifugation of the liver homogenate at 105 000 g for 60 min. The pellets were mixed with a small volume of sucrose and Tris buffer and stored in liquid nitrogen until used. Protein concentration in the microsomal fraction was estimated by the microbiuret method of Itzhaki and Gill [8]. Incubation mixtures contained in 2 ml in each flask: microsomes, 10.9 mg; N, 0.72 ymol, glucose, 9.4 pmol; glucose 6-phosphate, 10 pmol; MgCl:!, 60 clmoi DMBA, 2 pmol; and Tris buffer (pH 7.5) 100 pmol. Ten flasks were incubated while protected from direct light at 37°C for 30 min. The reaction was stopped by the addition of 2 ml of acetone and the mixture was extracted 5 times with 2 ml of ethyl acetate. The pooled extracts were dried over MgS04, filtered and evaporated in vacua. The residue was dissolved in I ml of methanol for analysis by HPLC. RESULTS
AND DISCUSSION
Figure 1A shows a typical profile of the HPLC chromatogram which resulted from photooxidation of DMBA for l-l.5 h in aqueous sol&ion which contained 10% methanol as a dispersing agent. Identification cf t-he principal peaks involved comparisons with synthetic compounds rega;rding ultraviolet and iqw resolution spectrograms and retention times on the HLPC. Figure 2 shows structures of the compounds as ide_it$ified and their interrelationships. The products listed as X, XI and XII on Fig. 1A were not consistently present in the HPLC profiles and are considered second products. The mixture of compounds in cluster X was collected from the effluent and subjected to repeated recycling. This separated 4 products. None of these could be characterized by comparison to known oxidation products.
342 10. 8.
A
6.
.--.
.
0
4
.
t
.
. .
It
II
.
. (D
.
.
ta
. * tt
I,,.._........ 0
4
RETENTION
L
. 0
. 4
t
RETENTION
12
IC
t
8*
IS
to
t4
TIME (MINI
PD
TIME 0WINl
Fig. 1. HPLC Profiles of DMBA Oxidation Roducte. See Materials and Methods for preparation of samples. (1A) r)W%l. irradiated in 10% methanol for l-l.5 h; (1B) F’hotooxidation products of DMBA alter standing; (1C) Oxidation of DMBA by singlet oxygen; (1D) Oxidation of DMBA by Cew, (1E) Oxidation of DMBA by MnOz ; (1F) Oxidation of DMBA of Fe Fe(CPm5J 6 , (1G) Oxidation of DMBA by rat liver microsones.
343
ml
C% Ix
Fig. 2. Interrelationships between products of oxidation of DMBA. (I) 7,12dimethylbenz[a]anthracene; (II) 7,12~pidioxy-7,12-dihydo-7,12d~ethylber~z~~]ant~ene; (III) 12-hy~~xy-12-methyIbenz[olanthracen-7~~e; (IV) ‘I-hydroxy-‘f-methylbenz[a] anthracen-IO-one; (V) benz[a]authracen-7,12-dione; (VI) 7-hydroxymethyl-12-methylbenz[o]antbzacene; (VII) 12-hydroxymethyl-7-methylbenz[o]anthracene;(VIII) 7-formyl12-methylbenz[o]anthracene; (IX) 12-formyl-7-methylbenz[a]anthracene.
Light in the ultraviolet region of the spectrum or by direct sunlight has been used by a number of investigators for photooxidation of aromatic hydrocarbons but the media were non-p&r solvents [9]. High yields of the photooxide, I, were generally ohtied but, with intense, prolonged irradiation, many decomposition products were observed. Similarly, we observed that exposure of DMBA to light, in dimethyl sulfoxide solution produced the photooxide and only traces of other compounds. The present study utilized irradiation of DMBA in aqueous solution by exposure to visible light at the desk top w characteristic of the modem laboratory. In some ins+ances the light was filtered through a watch glass. This produced no change in the nature of the oxidation products. The photooxidation products could be detected by inhibition of glucose-6-phosphate dehydrogenase after as little as one minute of irradiation [l] but l-l.5 h of exposure were used to produce quantities of products suitable for HPLC. Longer irradiations resulted in aL?pearance of many secondary products on the HPLC profiles. Similarly the initial mixture of irradiation products decomposed or interacted when sitting in the laboratory for long periods (Pig. ll3). It is well-established that poly cyclic aromatic hydrocarbons can produce ‘dye-type’ photosensitizations [6,10,11]. Photooxidation of the derivative of DMBA depended upon the photosensitizing property of the hydrocarbon in the presence of oxygen [ 1]*The photooxidation products of DMBA
were not photosensitizing and not oxidized when dispersed in 10% :nethand in water medium*. Thus irradiation of the monohydroxy derivatives, VI and VII, the 7,12-endoperoxide, II, the 7,12-dione, V, or the bermanthrone, III, under laboratory light in the absence of DMBA was ineffectual. Many endoperoxides dissociate into the hydrocarbon and oxygen [ 121 but it is obvious that the 7,12endoperoxide of DMBA does not under these conditions. Singlet oxygen molecules can be effectively generated by transfer of energy from excited state sensitizer molecules [ 13,141. Although other mechanisms have been proposed [6] a number of investigators have concluded that the preponderance of experimental evidence indicates the addition of singlet oxygen to polycylcic aromatic hydrocarbons to be a uniquely characteristic reaction [14]. This conclusion is readily borne out by comparison of the profiles of the HPLC chromatogram when DMBA was oxidized by chemicallygenerated singlet oxygen (1C with 1A). The short half-life of singlet oxygen requires its generation in situ. With a large excess of hydrogen peroxide and a limiting excess of sodium hypochloride, good yields of product can be expected [ 111. Under the conditions used, compared to photooxidation, a larger proportion of the products, X and XI, were found. When larger quantities of singlet oxygen were generated at room temperature in 40-50% methanol, the only product observed in the HPLC effluent was the endoperoxide, II. Hydrogen peroxide alone had no effect on the hydrocarbon. Comparison of the photooxidation products of DMBA with those reported by Schlumm [4] and Fried [ 151 for one-electron chemical oxidations reveals remarkable similarity in the nature of the products (Fig. lD, lE, 1F). It is apparent that the initial positions of electrophilic attack on DMBA are at the 7- and 1Zmethyls and the 7- and 12-carbons of the benz [alanthracene ring system [15]. This appears true also for the metabolic oxidations [16]. TabIe I shows amounts of products formed from DMBA by photooxidation for 1.5 h in comparison to products from chemical oxidation and metabolism. The numbers have qualitative significance only since variations in the time of irradiation or oxidation will change the relative amounts of products. Oxidation of DMBA by a rat liver microsamal system was done for comparison purposes. The similarity to photooxidation is readily apparent in Fig. 16 although the metabolites involving epoxidation of the 5,6 and 8,9 positions and the formation of 3- and 4-hydroxy benz[a]anthracene derivatives, which were detected by Yang and Dower [ 171, were not found. Boyland and Sims [16] considered the formation of the photooxide, II, in microsomal incubations of DMBA to be an artefact introduced in the isolation procedures. However, Chen and Tu 1.IS] found the cytochrome P-450 system of rat liver microsomes to form the photooxide, the cis-7,12-dihydroxy-7,12dihydro-7,!2-dimethylbenz[a]anthracene and 7-hydroxymethyl-lBmethylbenz[a]anthracene as the principal products. Increased yields of these metabdites and detectian of 7-hydroxymethyl-l%methylbenz[a Janthracene, *me 7-methoxymethyl-l2-methylbenz[a]anthracene
was active as a photosensitizer
[l J.
2.5 4.3 3.5 none 1.2 0.2 5.1 8.3 1.2
60.0 8.0 4 28 4 4 12 15 16 18 -
13 0 0 22 8 14 8 -
33 2 1 7 4 42 23 -
Fe( FeCN)? (%)
25
2
MnOz (%)
One-electron oxidationb
none 9.6 1.9 1.5 -
d
(nmollmg microsomal protein)
Microsomal metabolismC
*The measurement of photooxidation products is described in Materials ar.d Methods. ‘Values for one-electron oxidation products were obtained by Schumm [4] by thin-layer chromatographic separations. ‘Metabolites as measured by Yang and Dower [ 171 in motabolites of DMBA by rat liver micrdsomes. The principal initial oxidation product was 8,9dihydro-8,9dihydroxy-7,12dimethylbenz[c] snthracene (24.2 nmol/mg protein). dChen and Tu [ 181 found approx. 2% conversion of DMBA to compound II, by a rat liver cytochrome P-450 system.
-_ Benz[o,‘anthracene, -7-12.dimethyl (I) -7,12epidioxy-7,12_dihydro-7,12dimethyl( II) -7,312-dione (V) -7.hydroxym;thyl-12-methyl (VI) -12-hydroxymethyl-7-methyi (VII) -7 ,12-dihydroxyme thy1 -7-formyl-12.methyl (VIII) -12.formyl-7-methyl (IX) -12-hydroxy-12-methyl-ll-one (III) -7-hydroxy-7-methyl-l2-one (IV) cluster (X) cluster (XII)
Compound
Photooxidatlon* !,%)
OF OXIDATION PRODUCTS OF DMBA
Compounds are keyed to Fig. 2.
DISTRIBUTION
TABLE I
\qx, was obtained when methylcholanthrene-stimulated microsomes were used. In our experiments, the cis-dihydroxy-dihydro compound would appear m a distinct peak on HBPLC profiles near II. None was observed either in light-irradiation or microsomal oxidations of DMBA. The involvement of photooxidation of polycyclic aromatic hydrocarbons in a carcinogenesis process has long been an attractive hypothesis [ 13,19,20]. Morganand Warshawski [21] concluded from studies on Artemia salinethat there is a significant association between carcinogenesis and photodynamic activities of polycyclic aromatic hydrocarbons. Warshawsti et al. [22] proposed that some of the biological effects attributed to DMBA are due to the photaprodcct, the 7,12-endoperoxide, II, which had developed during synthesis of the hydrocarbon or during manipulations in biological systems. HouRver, some investigators have concluded that the photooxidation products are not in themselves carcinogenic but represent a detoxication process [ 23,241. This is an ove@mplification since in the case of DMBA it is apparent that the formation of the 7- and 12-hydroxymethyl derivatives which are carcinogenic and necrotizing to the adrenals is not a detoxication process [ 251. Light, air and moisture rapidly destroy polycyclic hydrocmbons in the environment but to term this detoxication may be improper until the processes are better understood. Our investigations of the effect of casual, short term illumination of aqueous solutions in the laboratory were undertaken to determinine if photooxidation could be a factor in studies of metabolism of hydrocarbons and their interactions with tissue constituents. While the possibility has been established, it is apparent from the current literature that most investigators are now taking this factor in account in present work by carefully shielding compounds from light. REFERENCES 1 C.J. Grubbs, E.T. Hutcheson and J.L. Wood, The inhibition of glyceraldehyde-3phosphate dehydrogenase by benz[a]anthracene and its derivatives after exposure to laboratory lighting, Chem-Biol. Interact., 10 (1975) 173. 2 C.J. Grubbs and J.L. Wood, The binding of benz[a]anthracene derivatives to glyceraldehyde-t-phosphate dehydrogenase and mammary tissue following exposure to laboratory light, Chcm.-Biol. Interact., 12 (1976) 135. 3 R.B. Sandin and L.F. Fieser, Synthesis of a 9,10-dimethyl-1,2-benzanthracene and of a thiophene isolog, J. Am. Chem. Sot., 62 (1940) 3098. 4 D.L. Schumm, In vitro and in vivo one-electron oxidation of 7,12-dimethylbenz[a 1 anthracene, Dissertation, Univ. Chicago, Dec., 1969. 5 G.M. Badger and J.W. Cook, The synthesis of growth-inhibitory polycyclic compounds, Part II, J. Chem. Sot. (1940) 409. 6 C.S. Foote, S. Wexler, @. Ando and R. Higgins, Chemistry of singlet oxygen. IV. Oxygenations with hypochloritc-hydrogen peroxide, J. Am. Chem. Sot., 90 (1968) 975. 7 N. Kinoshita, B. Shears, and H.V. Gelboin, K-region and non-K-region metabolism of benzo[a]pyrene by rat liver microsomes, Cancer Rep., 33 (1973) 1937. 8 R.F. Itzhaki and D.M. Gill, A micro-biuret method for estimating proteins, Anal. Biochem, 9 (1964) 401. 9 W. Bergmann and M.J. McLean, Transannular peroxides, Chem. Rev. 28 (1941) 367.
347 10 11 12
13 14 15 16 17 18 19 20 21
22
23 24 25
D.R. Ksarns and A.U. Khan, Sensitized photooxygenation reactions and the role of singlet oxygen, Photochem Photobiol., 10 (1969) 193. C.S. Foote, Photosensitized oxygenations and ‘:he role of singlet oxygen, Act. Chem. Res., 1(1968) 104. J. Rigaudy, M.C. Perlat, D. Simon and N.K. Cuong, Transformations thermiques des photooxydes m&o des scenes. 1. Cas de photooxides de phenyE9 et de methyl-9 anthracene, Bull. Sot. Chim. Fr., (1976) 493. A.U. Khan and M. Kasha, An optical residue singlet oxygen theory of photocarcinogenicity, Ann. N. Y. Acad. Sei., 171(1970) 24. D.R. Keams, Physical and chemical properties of singlet oxygen, Chem. Rev., 71 (1971) 395. 5. Fried, One-electron oxidation of polycyclic aromatics as a model for metabolic activation of carcinogenic hydrocarbons, Biochem. Dis., 4 (1974) 197. E. Boyland and P. Sims, Metabolism of polycyclic compounds. The metabolism off 7,12-dimethylbenz[o]anthracene in rat liver homogenates, Biochem. J., 95 (1965) 780.. SK. Yang and W.V. Dower, Metabolic pathways of 7,12-dimethylbenz[o Janthracene in hepatic microsomes, Proc. Natl. Acad. Sci., U.S.A., 72 (1975) 2601. C. Chen and M.H. Tu, Transannular dioxygenation of 9,10-dimethylbenzanthracene by cytochrome P-450 oxygensae of rat liver, Biochem. J., 160 (1976) 805. E. Cavalieri and M. Calvin, Photochemical coupling of benzo[o]pyrene with lmethylcytosine; Photoenhancement of carcinogenicity, Photochem. Photobiol., (1971) 641. I.R. Politzer, G.W. Griffin and J.L. Laseter, Singlet oxygen and lbiobgical systems, Chen-Biol. Interact., 3 (1971) 73. D.D. Morgan and D. Warshawsky, Thz photodynamic immobilization of Artemiu salina Nauplii by polycyclic aromatic hydrocarbons and its relationship to carcinogenic activity, Photochem. Photobiol., 25 (1977) 39. D. Warshawsky, E. Kerns, M.J. Biiell and M. Calvin, Characterization of a photoproduct of 7,12-dimethylbenz[a]anthracene and its effects on chick embryo cells in culture, Biochem. J., 164 (1977) 481. W. Graef and A.G. Haller, The phototoxicity of carcinogenic polycyclic hydrocarbons and their degradation products in biological model. Zentralbl. Bakeriol. Parasitenkd., Infektionskr., Hyg., Abt 1: Orig. Reiche, B., 164 (1977) 250. MI. Gubergrits, A.B. Linik, L.P. Paal’me and L.M. Shabad, Study of the carcinogenicity of products from photoinduced oxidation of benzo[a]pyrene, Vapr. Cnkol., 20 (1974) 77. E. Boyland, P. Sims and C. Huggins, Induction of adrenal damage and cancer with metabolites of 7,12-dimethylbenz[a]anthracene, Nature (London), 207 (1965) 816.
339
J.L. WOOD, C.L. BARKER and CJ. GRUBBS*
Department of Biochemistry , University Sciences, Memphis, TN 38163 (U.S.A.: 1(Received August 21st, 1973) (Revision received December l&h, (Accepted March 12th, 1979)
of
Tennessee,
Center
for
the Health
19’78)
SUMMARY
The principal products of the photooxidation of 7,12dimethylbenz[o] anthracene @MBA) in aqueom solutions by photooxidation induced by laboratory lighting tlave been characterized by high performance liquid chromatograms jEPLCj, ultraviolet and mass spectrograms and by comparisons with authentic samples. The products identified were the 7,12epidioxy-7,12-dihydro-7,12-dimethyl-, 7,12-dione, ‘I-hydroxymethyl-12methyl-, 12-hydroxymethyl-7-methyl-, ‘I-formyl-f2-methyl-, X2-formyl-7methyl-, and 12-hydroxy-12methyl-7-one derivatives of benz[c]-anthracene. The HPLC profile of products is similar to that obtained from oxidation of DMBA by ‘one-electron’ reagents, singlet oxygen, or liver microsomal
metabolism. The first points of attack are the 7- and 12- positions. The mechanism of photooxidation appears t;s be generation of singlet oxygen by photodynamic effect of DMBA. None of the products is photosensitizing, however.
INTRODUCTION
Photodynamic effects in biological systems and sensitivity of polycyclic hydrocarbons to destruction by light have been known for a long time. Until recently the modification of aromatic hydrocarbons by relatively short exposure to laboratory lighting during experimental manipulations has not been well recognized. Previous work in this laboratory has-shown that photooxidation of polycyclic aromatic hydrocarbons by visible light occurs rea *Present address: Division 60616,
L, i’!T Research
Institute,
10th West 35th Street, Chicago, IL
U.S.A.
Abbreviations: c:hrom tography.
DMBA, 7,12_dimethylbenz[a]anthracene;
HFLC, high performance
liquid
when air and moisture are present [ 11. Some of the products of the oxidation were effective in inactivating the sulfhydryl-dependent enzymes, lactic dehydrogenase and glyceraldehyde3-phosphate dehydrogenase, and also to bind to protein [2]. We have determined the chemical natures of the principal products of the photooxidation of DMBA in water suspension. Comparisons of the compounds formed by one-electron oxidations, by singlet oxygen, and by metabolism of DMBA by microsomes have been made and some similarities noted. MATERIALSANDMETHODS DMBA (Eastman) was recrystallized before use. When chromatographed LC, the compound emerged as a single sharp peak. Benz[a]anthraceneby 7,12-dions was obtained from Eastman. DMBA-7,12-photoperoxide (7,12epidioxy-7,.12,dihydro-7,12dimethylbenz[a] anthracene) was prepared according to the method of Sandin and Fieser [3]. cis-7,12Dihydro-7,12dihydmxy-T,l Bdimethylbenz [ a ] anthracene was prepared by catalytic hydrogenation of the 7,12-peroxide over 10% palladium on charcoal [ 41. The 12hydwoxy-lZmethylbenz[a]anthrone was prepared by the action of methyl magnesium iodide on benz[a]anthracene-7,12-dione [4]. T-Hydroxym&hyl-12-methylbenz [a] anthracene and 7,12-dihydroxymethyl-benz[u] anthracene were supplied by Dr. Charles Irving. ‘I-Formyl-la-methyl- and 12formy l-T-methylbenz[a]anthracene were prepared by the method of Badger and Cook [5]. All of the above compounds were characterized by their melting points, ultraviolet absorption and mass spectra. Singlet oxygen was generated by the method of Foote et al. [6]. To a solution of hydrogen peroxide (36 pmol in 3.4 ml of water cooled in an ice bath) was added 1 pmol of DMBA in 1 ml of cold methanol. The suspension was treated with 3 pmol of cold sodium hypochlorite (Purex 14) dropwise with vigorous stirring. After 5 min the solution w&aevaporated to dryness in vacua and the residue was taken up in 2 ml of methanol and subjected to fractionation by HPLC (Fig. 1C). HPLC chromatograms were obtained with a Waters Model 6000 chromatograph equipped with a quarter inch C II Bondapak column 30 cm long (Waters Associates, Milford, MA). Products of DMBA photooxidation were eluted from the column by a reverse phase gradient system, methanol/water, 60-100% methanol. Gradient rate of change was 5%/min with a flow rate of 2 ml/min. Emergence of fractions from the column was detected by an ultraviolet absorption m&nitor which utilized a 254 nm band. Separation of the photooxidation components required recycling from three to four times. The percentage distrubution of the material in the peaks was determined, through the courtesy of Dr. Byron Leach, on a Hewlett-Packard Liquid ChromatogTaph No. 1084a equipped qrith a CIs Bondapak column, 75 cm long, and fed into a data processor. Low resolution mass spectrograms were obtained with a Finnegan Quadrapole 1015 Spectrometer equipped with a 9500 gas chroma&ograpb input and a No. 6000 data system.
341 The photooxidation pocedure involved exposure of 1 pm01 of DMBA, or other compound, in 5 ml of 10% methanol in a crystallizing dish, for 1.5 h on the desk top in the laboratory. The light was provided by two General Electric Cool White fluorescent tubes that had a spectral o ranging from 340 to 750 nm with a small amount of radiation from the mercury resonance lines. Most of the energy was in the spectrum from 400 to 650 nm. The incident light on the desk top was 1350 lux. After irradiation the solvent was evaporated in vacua and the residue was taken up in methanol for injection into the HPLC apparatus. Oxidations of DMBA by ‘one-electron’ reagents were carried out as described by Schumm [4]. DMBA (10 mg) in aqueous acetone solution was treated with a 1 molar ratio of ceric ammonium nitrate or 4 mol of ferric ferricyanide overnight at room temperature. Similarly, 10 mg of DMBA in benzene were stirred with 50 mg of activated manganese dioxide for 22 h. The reaction products of each mixture were extracted into ethyl acetate, dried and transferred to methanol for HPLC fractionation.
Metabolism studies Male albino rats were injected intraperitoneally with 1 mg of methylcholanthrene 48 h prior to use. Liver microsomes were prepared by the method of Kinoshita et al. [ 71. Microsomal pellets were ob+tained by ultracentrifugation of the liver homogenate at 105 000 g for 60 min. The pellets were mixed with a small volume of sucrose and Tris buffer and stored in liquid nitrogen until used. Protein concentration in the microsomal fraction was estimated by the microbiuret method of Itzhaki and Gill [8]. Incubation mixtures contained in 2 ml in each flask: microsomes, 10.9 mg; N, 0.72 ymol, glucose, 9.4 pmol; glucose 6-phosphate, 10 pmol; MgCl:!, 60 clmoi DMBA, 2 pmol; and Tris buffer (pH 7.5) 100 pmol. Ten flasks were incubated while protected from direct light at 37°C for 30 min. The reaction was stopped by the addition of 2 ml of acetone and the mixture was extracted 5 times with 2 ml of ethyl acetate. The pooled extracts were dried over MgS04, filtered and evaporated in vacua. The residue was dissolved in I ml of methanol for analysis by HPLC. RESULTS
AND DISCUSSION
Figure 1A shows a typical profile of the HPLC chromatogram which resulted from photooxidation of DMBA for l-l.5 h in aqueous sol&ion which contained 10% methanol as a dispersing agent. Identification cf t-he principal peaks involved comparisons with synthetic compounds rega;rding ultraviolet and iqw resolution spectrograms and retention times on the HLPC. Figure 2 shows structures of the compounds as ide_it$ified and their interrelationships. The products listed as X, XI and XII on Fig. 1A were not consistently present in the HPLC profiles and are considered second products. The mixture of compounds in cluster X was collected from the effluent and subjected to repeated recycling. This separated 4 products. None of these could be characterized by comparison to known oxidation products.
342 10. 8.
A
6.
.--.
.
0
4
.
t
.
. .
It
II
.
. (D
.
.
ta
. * tt
I,,.._........ 0
4
RETENTION
L
. 0
. 4
t
RETENTION
12
IC
t
8*
IS
to
t4
TIME (MINI
PD
TIME 0WINl
Fig. 1. HPLC Profiles of DMBA Oxidation Roducte. See Materials and Methods for preparation of samples. (1A) r)W%l. irradiated in 10% methanol for l-l.5 h; (1B) F’hotooxidation products of DMBA alter standing; (1C) Oxidation of DMBA by singlet oxygen; (1D) Oxidation of DMBA by Cew, (1E) Oxidation of DMBA by MnOz ; (1F) Oxidation of DMBA of Fe Fe(CPm5J 6 , (1G) Oxidation of DMBA by rat liver microsones.
343
ml
C% Ix
Fig. 2. Interrelationships between products of oxidation of DMBA. (I) 7,12dimethylbenz[a]anthracene; (II) 7,12~pidioxy-7,12-dihydo-7,12d~ethylber~z~~]ant~ene; (III) 12-hy~~xy-12-methyIbenz[olanthracen-7~~e; (IV) ‘I-hydroxy-‘f-methylbenz[a] anthracen-IO-one; (V) benz[a]authracen-7,12-dione; (VI) 7-hydroxymethyl-12-methylbenz[o]antbzacene; (VII) 12-hydroxymethyl-7-methylbenz[o]anthracene;(VIII) 7-formyl12-methylbenz[o]anthracene; (IX) 12-formyl-7-methylbenz[a]anthracene.
Light in the ultraviolet region of the spectrum or by direct sunlight has been used by a number of investigators for photooxidation of aromatic hydrocarbons but the media were non-p&r solvents [9]. High yields of the photooxide, I, were generally ohtied but, with intense, prolonged irradiation, many decomposition products were observed. Similarly, we observed that exposure of DMBA to light, in dimethyl sulfoxide solution produced the photooxide and only traces of other compounds. The present study utilized irradiation of DMBA in aqueous solution by exposure to visible light at the desk top w characteristic of the modem laboratory. In some ins+ances the light was filtered through a watch glass. This produced no change in the nature of the oxidation products. The photooxidation products could be detected by inhibition of glucose-6-phosphate dehydrogenase after as little as one minute of irradiation [l] but l-l.5 h of exposure were used to produce quantities of products suitable for HPLC. Longer irradiations resulted in aL?pearance of many secondary products on the HPLC profiles. Similarly the initial mixture of irradiation products decomposed or interacted when sitting in the laboratory for long periods (Pig. ll3). It is well-established that poly cyclic aromatic hydrocarbons can produce ‘dye-type’ photosensitizations [6,10,11]. Photooxidation of the derivative of DMBA depended upon the photosensitizing property of the hydrocarbon in the presence of oxygen [ 1]*The photooxidation products of DMBA
were not photosensitizing and not oxidized when dispersed in 10% :nethand in water medium*. Thus irradiation of the monohydroxy derivatives, VI and VII, the 7,12-endoperoxide, II, the 7,12-dione, V, or the bermanthrone, III, under laboratory light in the absence of DMBA was ineffectual. Many endoperoxides dissociate into the hydrocarbon and oxygen [ 121 but it is obvious that the 7,12endoperoxide of DMBA does not under these conditions. Singlet oxygen molecules can be effectively generated by transfer of energy from excited state sensitizer molecules [ 13,141. Although other mechanisms have been proposed [6] a number of investigators have concluded that the preponderance of experimental evidence indicates the addition of singlet oxygen to polycylcic aromatic hydrocarbons to be a uniquely characteristic reaction [14]. This conclusion is readily borne out by comparison of the profiles of the HPLC chromatogram when DMBA was oxidized by chemicallygenerated singlet oxygen (1C with 1A). The short half-life of singlet oxygen requires its generation in situ. With a large excess of hydrogen peroxide and a limiting excess of sodium hypochloride, good yields of product can be expected [ 111. Under the conditions used, compared to photooxidation, a larger proportion of the products, X and XI, were found. When larger quantities of singlet oxygen were generated at room temperature in 40-50% methanol, the only product observed in the HPLC effluent was the endoperoxide, II. Hydrogen peroxide alone had no effect on the hydrocarbon. Comparison of the photooxidation products of DMBA with those reported by Schlumm [4] and Fried [ 151 for one-electron chemical oxidations reveals remarkable similarity in the nature of the products (Fig. lD, lE, 1F). It is apparent that the initial positions of electrophilic attack on DMBA are at the 7- and 1Zmethyls and the 7- and 12-carbons of the benz [alanthracene ring system [15]. This appears true also for the metabolic oxidations [16]. TabIe I shows amounts of products formed from DMBA by photooxidation for 1.5 h in comparison to products from chemical oxidation and metabolism. The numbers have qualitative significance only since variations in the time of irradiation or oxidation will change the relative amounts of products. Oxidation of DMBA by a rat liver microsamal system was done for comparison purposes. The similarity to photooxidation is readily apparent in Fig. 16 although the metabolites involving epoxidation of the 5,6 and 8,9 positions and the formation of 3- and 4-hydroxy benz[a]anthracene derivatives, which were detected by Yang and Dower [ 171, were not found. Boyland and Sims [16] considered the formation of the photooxide, II, in microsomal incubations of DMBA to be an artefact introduced in the isolation procedures. However, Chen and Tu 1.IS] found the cytochrome P-450 system of rat liver microsomes to form the photooxide, the cis-7,12-dihydroxy-7,12dihydro-7,!2-dimethylbenz[a]anthracene and 7-hydroxymethyl-lBmethylbenz[a]anthracene as the principal products. Increased yields of these metabdites and detectian of 7-hydroxymethyl-l%methylbenz[a Janthracene, *me 7-methoxymethyl-l2-methylbenz[a]anthracene
was active as a photosensitizer
[l J.
2.5 4.3 3.5 none 1.2 0.2 5.1 8.3 1.2
60.0 8.0 4 28 4 4 12 15 16 18 -
13 0 0 22 8 14 8 -
33 2 1 7 4 42 23 -
Fe( FeCN)? (%)
25
2
MnOz (%)
One-electron oxidationb
none 9.6 1.9 1.5 -
d
(nmollmg microsomal protein)
Microsomal metabolismC
*The measurement of photooxidation products is described in Materials ar.d Methods. ‘Values for one-electron oxidation products were obtained by Schumm [4] by thin-layer chromatographic separations. ‘Metabolites as measured by Yang and Dower [ 171 in motabolites of DMBA by rat liver micrdsomes. The principal initial oxidation product was 8,9dihydro-8,9dihydroxy-7,12dimethylbenz[c] snthracene (24.2 nmol/mg protein). dChen and Tu [ 181 found approx. 2% conversion of DMBA to compound II, by a rat liver cytochrome P-450 system.
-_ Benz[o,‘anthracene, -7-12.dimethyl (I) -7,12epidioxy-7,12_dihydro-7,12dimethyl( II) -7,312-dione (V) -7.hydroxym;thyl-12-methyl (VI) -12-hydroxymethyl-7-methyi (VII) -7 ,12-dihydroxyme thy1 -7-formyl-12.methyl (VIII) -12.formyl-7-methyl (IX) -12-hydroxy-12-methyl-ll-one (III) -7-hydroxy-7-methyl-l2-one (IV) cluster (X) cluster (XII)
Compound
Photooxidatlon* !,%)
OF OXIDATION PRODUCTS OF DMBA
Compounds are keyed to Fig. 2.
DISTRIBUTION
TABLE I
\qx, was obtained when methylcholanthrene-stimulated microsomes were used. In our experiments, the cis-dihydroxy-dihydro compound would appear m a distinct peak on HBPLC profiles near II. None was observed either in light-irradiation or microsomal oxidations of DMBA. The involvement of photooxidation of polycyclic aromatic hydrocarbons in a carcinogenesis process has long been an attractive hypothesis [ 13,19,20]. Morganand Warshawski [21] concluded from studies on Artemia salinethat there is a significant association between carcinogenesis and photodynamic activities of polycyclic aromatic hydrocarbons. Warshawsti et al. [22] proposed that some of the biological effects attributed to DMBA are due to the photaprodcct, the 7,12-endoperoxide, II, which had developed during synthesis of the hydrocarbon or during manipulations in biological systems. HouRver, some investigators have concluded that the photooxidation products are not in themselves carcinogenic but represent a detoxication process [ 23,241. This is an ove@mplification since in the case of DMBA it is apparent that the formation of the 7- and 12-hydroxymethyl derivatives which are carcinogenic and necrotizing to the adrenals is not a detoxication process [ 251. Light, air and moisture rapidly destroy polycyclic hydrocmbons in the environment but to term this detoxication may be improper until the processes are better understood. Our investigations of the effect of casual, short term illumination of aqueous solutions in the laboratory were undertaken to determinine if photooxidation could be a factor in studies of metabolism of hydrocarbons and their interactions with tissue constituents. While the possibility has been established, it is apparent from the current literature that most investigators are now taking this factor in account in present work by carefully shielding compounds from light. REFERENCES 1 C.J. Grubbs, E.T. Hutcheson and J.L. Wood, The inhibition of glyceraldehyde-3phosphate dehydrogenase by benz[a]anthracene and its derivatives after exposure to laboratory lighting, Chem-Biol. Interact., 10 (1975) 173. 2 C.J. Grubbs and J.L. Wood, The binding of benz[a]anthracene derivatives to glyceraldehyde-t-phosphate dehydrogenase and mammary tissue following exposure to laboratory light, Chcm.-Biol. Interact., 12 (1976) 135. 3 R.B. Sandin and L.F. Fieser, Synthesis of a 9,10-dimethyl-1,2-benzanthracene and of a thiophene isolog, J. Am. Chem. Sot., 62 (1940) 3098. 4 D.L. Schumm, In vitro and in vivo one-electron oxidation of 7,12-dimethylbenz[a 1 anthracene, Dissertation, Univ. Chicago, Dec., 1969. 5 G.M. Badger and J.W. Cook, The synthesis of growth-inhibitory polycyclic compounds, Part II, J. Chem. Sot. (1940) 409. 6 C.S. Foote, S. Wexler, @. Ando and R. Higgins, Chemistry of singlet oxygen. IV. Oxygenations with hypochloritc-hydrogen peroxide, J. Am. Chem. Sot., 90 (1968) 975. 7 N. Kinoshita, B. Shears, and H.V. Gelboin, K-region and non-K-region metabolism of benzo[a]pyrene by rat liver microsomes, Cancer Rep., 33 (1973) 1937. 8 R.F. Itzhaki and D.M. Gill, A micro-biuret method for estimating proteins, Anal. Biochem, 9 (1964) 401. 9 W. Bergmann and M.J. McLean, Transannular peroxides, Chem. Rev. 28 (1941) 367.
347 10 11 12
13 14 15 16 17 18 19 20 21
22
23 24 25
D.R. Ksarns and A.U. Khan, Sensitized photooxygenation reactions and the role of singlet oxygen, Photochem Photobiol., 10 (1969) 193. C.S. Foote, Photosensitized oxygenations and ‘:he role of singlet oxygen, Act. Chem. Res., 1(1968) 104. J. Rigaudy, M.C. Perlat, D. Simon and N.K. Cuong, Transformations thermiques des photooxydes m&o des scenes. 1. Cas de photooxides de phenyE9 et de methyl-9 anthracene, Bull. Sot. Chim. Fr., (1976) 493. A.U. Khan and M. Kasha, An optical residue singlet oxygen theory of photocarcinogenicity, Ann. N. Y. Acad. Sei., 171(1970) 24. D.R. Keams, Physical and chemical properties of singlet oxygen, Chem. Rev., 71 (1971) 395. 5. Fried, One-electron oxidation of polycyclic aromatics as a model for metabolic activation of carcinogenic hydrocarbons, Biochem. Dis., 4 (1974) 197. E. Boyland and P. Sims, Metabolism of polycyclic compounds. The metabolism off 7,12-dimethylbenz[o]anthracene in rat liver homogenates, Biochem. J., 95 (1965) 780.. SK. Yang and W.V. Dower, Metabolic pathways of 7,12-dimethylbenz[o Janthracene in hepatic microsomes, Proc. Natl. Acad. Sci., U.S.A., 72 (1975) 2601. C. Chen and M.H. Tu, Transannular dioxygenation of 9,10-dimethylbenzanthracene by cytochrome P-450 oxygensae of rat liver, Biochem. J., 160 (1976) 805. E. Cavalieri and M. Calvin, Photochemical coupling of benzo[o]pyrene with lmethylcytosine; Photoenhancement of carcinogenicity, Photochem. Photobiol., (1971) 641. I.R. Politzer, G.W. Griffin and J.L. Laseter, Singlet oxygen and lbiobgical systems, Chen-Biol. Interact., 3 (1971) 73. D.D. Morgan and D. Warshawsky, Thz photodynamic immobilization of Artemiu salina Nauplii by polycyclic aromatic hydrocarbons and its relationship to carcinogenic activity, Photochem. Photobiol., 25 (1977) 39. D. Warshawsky, E. Kerns, M.J. Biiell and M. Calvin, Characterization of a photoproduct of 7,12-dimethylbenz[a]anthracene and its effects on chick embryo cells in culture, Biochem. J., 164 (1977) 481. W. Graef and A.G. Haller, The phototoxicity of carcinogenic polycyclic hydrocarbons and their degradation products in biological model. Zentralbl. Bakeriol. Parasitenkd., Infektionskr., Hyg., Abt 1: Orig. Reiche, B., 164 (1977) 250. MI. Gubergrits, A.B. Linik, L.P. Paal’me and L.M. Shabad, Study of the carcinogenicity of products from photoinduced oxidation of benzo[a]pyrene, Vapr. Cnkol., 20 (1974) 77. E. Boyland, P. Sims and C. Huggins, Induction of adrenal damage and cancer with metabolites of 7,12-dimethylbenz[a]anthracene, Nature (London), 207 (1965) 816.
