Reactions of stereoisomeric and structurally related bay region diol epoxide derivatives of benz[a]anthracene with DNA. Conformations of non-covalent complexes and covalent carcinogen-DNA adducts

Chem.-BioL Interactions, 66 (1988) 121--145 Elsevier Scientific Publishers Ireland Ltd.

121

REACTIONS OF STEREOISOMERIC AND STRUCTURALLY RELATED BAY REGION DIOL EPOXIDE DERIVATIVES OF BENZ[a]ANTHRACENE WITH DNA. CONFORMATIONS OF NONCOVALENT COMPLEXES AND COVALENT CARCINOGEN-DNA ADDUCTS

SUSAN E. CARBERRY°, MANOUCHEHR SHAHBAZ ~*, NICHOLAS E. GEACINTOVa'** and RONALD G. HARVEY b *Chemistry Department and Radiation and Solid State Laboratory, New York University, New York, N Y 10003 and ~The Ben May Institute, The University of Chicago, Chicago, IL 60637 [U.S.A.J (Received May 12th, 1987) (Revision received November 23rd, 1987) (Accepted November 23rd, 1987)

SUMMARY

The modes of reaction of the tumorigenic bay region diol epoxide antiBADE (( ± )-trans-3,4-diol~nti-l,2-epoxy-l,2,3,4-tetrahydrobenz[a]anthracene) and the less potent tumor initiating diastereomer syn-BADE ((±)-trans-3,4diol-syn-l,2-epoxy-l,2,3,4-tetrahydrobenz[a]anthracene) with native, doublestranded DNA were compared. The bay-region diol epoxide derived from 3methylcholanthrene (3-MCDE, racemic trans-9,10-diol-anti-7,8-epoxy-7,8,9,10tetrahydromethylcholanthrene) was included in this study in order to assess the effects of the methyl and methylene substituents on the reactivity with DNA. Utilizing linear dichroism and other spectroscopic methods, it is shown that all three diol epoxides form non-covalent complexes with DNA. The diastereomers anti-BADE and syn-BADE form intercalative physical complexes, but the association constant K of the syn-diastereomer is about 6--7 times smaller than for anti-BADE; this effect is ascribed to the bulky quasi-diaxial conformation of the diol epoxide ring in the syn diastereomer. The value of K (4000 M-1) is similar for anti-BADE and 3-MCDE, although the latter is not intercalated in the classical sense since the short axis of the Abbreviations: BA, benz[a]anthracene; BP, benzo[a]pyrene; anti-BADE, anti-BADE, ( ± )-trans-3,4dihydroxy~nti-l,2-epoxy-l,2,3,4-tetrahydrobenz[a]anthracene; syn-BADE, ( ± )-trans-3,4-dihydroxysyn-l,2-epoxy-l,2,3,4-tetrahydrobenz[a]anthraeene; anti-BPDE, (±)-trans-7,8-dihydroxy-anti-9,10epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene; syn-BPDE, ( ± )-trans-7,8-dihydroxy-syn-9,10-epoxy7,8,9,10-tetrahydrobenzo{a]pyrene; 3-MCDE, ( ± )-trans-9,10-dihydroxy~nti-7,8-epoxy-7,8,9,10tetrahydro-3-methylchplanthrene. *Present address: Scripps Research Clinic, Division of Biochemistry, 1~56~ North Torrey Pines Road, La Jolla, CA 92037, U.S.A. **To whom all correspondence should be addressed.

0009-2797/88/$08.50 © 1988 Elsevier Scientific Publishers Ireland Ltd.

Printed and Published in Ireland

122 molecule is tilted closer to the axis of the DNA double helix. The conformations of the covalent DNA adducts are interpreted in terms of a quasi-intercalative conformation (site I), and a conformation in which the long axes of the polycyclic molecules are tilted closer to the axis of the helix (site II). Both tumorigens, anti-BADE and 3-MCDE, undergo a marked reorientation from a non-covalent site I to a covalent site II conformation upon binding chemically with the DNA bases, although a small fraction of the covalent anti-BADE adducts remains quasi-intercalated; in contrast, the alkyl substituents in 3-MCDE not only prevent the formation of intercalative physical complexes, but also the formation of site I covalent adducts. In the case of the less tumorigenic syn-BADE, both the non-covalent complexes and the covalent adducts are of the site I-type. The bay-region diol epoxide of benz[a]anthracene and of 3-methylcholanthrene display a similar pattern of reactivities and covalent adduct conformations as the bay region diol epoxide derivatives of benz[a]pyrene, suggesting that adduct conformation might be an important h c t o r in determining the levels of mutagenic and tumorigenic activities of this class of compounds.

Key words: Polycyclic aromatic hydrocarbon -- Carcinogenesis - DNAcarcinogen adducts - - Linear dichroism -- Intercalation -- Benzo[a]pyrene -Benz[a]anthracene

INTRODUCTION Benz[a]anthracene (BA) is 50--100 times less potent as a tumor initiator than benzo[a]pyrene (BP) [1,2] and is considered to be, at least, a weak carcinogen [3]. Like other polycyclic aromatic hydrocarbons, BA itself is chemically unreactive and biologically inert, but can be metabolically activated to highly reactive oxygenated derivatives [4]. The proximate and ultimate carcinogenic metabolites of BA have been identified as the 3,4dihydrodiol derivative of BA and the bay-region diol epoxide anti-BADE (Fig. 1), respectively. Possible differences in the efficiencies of metabolic activation of BA and BP may partially [2,6-9] contribute to the differences in the tumorigenic activities of BA and BP [7--9]. However, differences in the activities of the bay region diol epoxide metabolites [10], the ultimate tumorigenic forms of BP and BA, and might also account for the differences in the biological activities of these compounds. The diastereomeric bay region diol epoxides (_+)ant/-BPDE and (_+)syn-BPDE, are several-fold more mutagenic in bacterial and mammalian test systems than the analogous antiBADE and syn-BADE compounds [11]. The relative activities of the different BADE stereoisomers resemble those of the analogous BPDE stereoisomers; thus, (+)~nti-BADE is approximately 30 times as effective as (_+)-syn-BADE in inducing pulmonary tumors in new born mice [6], with the (+)-enantiomer of anti-BADE displaying by h r the highest activity [12]. On mouse skin [12,13], and in the

123 0

: ....

11

0 H

~

O

0

H

12

~o,

OH 8

,

1'

~ I ~

i ~

I ~

i

_---OH

6

anti-BADE

syn- BADE

0 -/~*

4

H3 C

~ 0H

H

.~11 2

I

3- MCDE

Fig. 1. Structures of anti-BADE, syn-BADE and 3-MODE showing the orientations of the shortaxis transition moment (absorption band above 320 nm), and the long-axis transition moment (255 -265 nm) absorption bands.

Chinese hamster V79 cell mutagenicity test system [11] on the other hand, this stereoselectivity is less pronounced. The covalent binding of aromatic diol epoxides to cellular DNA is believed to be a critical event in mutagenesis and the initiation of the tumorigenic process [14--19]. The striking differences in the biological activities [4,12,18 --20] of the different BPDE stereoisomers have been related to the spatial conformations of the pyrenyl residues in the covalent DNA adducts [15]. Spectroscopic studies of adducts derived from the binding of various epoxides and diol epoxides to DNA have shown that two classes of binding sites can be distinguished [21,22]. Type I sites are characterized by significant base-stacking interactions with a quasi-intercalative conformation (possibly a wedge-shaped intercalation complex [23,24]); type II sites are characterized by a lower degree of carcinogen-base stacking, and correspond to pyrenyl residues located in solvent-accessible regions of the DNA, or at external binding sites [25,26]. The highly tumorigenic ( + )-enantiomer of antiBPDE forms predominantly site II adducts [27,28], while the biologically less active ( - }~enantiomer of anti-BPDE and racemic syn-BPDE form predominantly site I type of adducts [27-30]. The highly tumorigenic bay region diol epoxides [10] of 3-methylcholanthrene [31] and 5-methyl chrysene [32] also form predominantly type II adducts, while other less active epoxide derivatives form mostly site I adducts [23]. The stereoisomers of the bay region diol epoxides of BA are thus characterized by a pattern of tumorigenic activities [12] similar to those of benzo[a]pyrene [4] and of chrysene [33]. The aim of this work was to investigate the patterns of chemical reactivities and DNA adduct conformations of the diastereomers (_+)anti-BADE and (_+)syn-BADE with DNA in vitro, the conformations of the covalent DNA adducts formed, and to determine if the stereoisomeric diol epoxides of BA also fit the apparent site

124 I/site II correlation between biological activity and adduct structure [23]. Other factors such as differences in (1) the formation of physical complexes with DNA prior to covalent binding [23,24,35] and (2) the fraction of molecules which are detoxified by hydrolysis to tetraols or other products, rather than binding covalently to DNA [23], were also investigated. The highly tumorigenic diol epoxide derived from 3-methylcholanthrene (3MCDE, Fig. 1), was also included in this study; while the stereochemistry of the epoxide and OH groups are the same in (_+)anti-BADE and in (_+)anti-3MCDE, the additional methyl group and aliphatic ring system in 3-MCDE might be expected to modulate the interaction of this diol epoxide derivative with DNA. MATERIALS AND METHODS

Preparation of DNA and epoxide stock solutions, hydrolysis of epoxides Calf thymus DNA (Worthington Biochemicals, Freehold, N J) was dissolved in 5 mM sodium cacodylate buffer solution (pH 7.0), 0.1 M NaCl and 3 mM sodium ethylenediaminetetraacetate (EDTA) and then exhaustively dialyzed against 5 mM cacodylate buffer (pH 7.0). The DNA was then centrifuged at 10000 rev./min prior to use; the hypochromicity of the DNA was in the range of 38 _+ 1%. Racemic anti-BADE (Lot No. CR53-8-2) and syn-BADE (Lot CSRR15-7-2) were obtained from the National Cancer Institute Chemical Carcinogen Reference Standard Repository, while the racemic 3-MCDE was synthesized as described elsewhere [36]. Stock solutions of the epoxides ( 2 - 5 raM) were prepared by dissolving small amounts of the solids in tetrahydrofuran (THF). The concentrations of the epoxides were determined spectrophotometrically utilizing the extinction coefficients of anti and synBADE provided with the compounds by the National Cancer Institute Chemical Carcinogen Reference Standard Repository. In aqueous solutions the values of the molar extinction coefficients are about 5000 and 3000 M -1 cm -1 at the 366 and 386 nm absorption maxima, respectively. Upon the hydrolysis of both anti-BADE and syn-BADE in the aqueous buffer solutions, the absorption maxima of the hydrolysis products are shifted uniformly to the blue by about 6 nm {data not shown). In the case of 3-MCDE in buffer solution, the absorption maxima at 397 and 417 nm a r e blue-shifted to 379 and 398 nm, respectively, upon hydrolysis of the diol epoxide in buffer solutions [31].

Preparation of covalent adducts Small aliquots of the diol epoxide stock solutions were added to buffer solutions containing from 0.53 to 1.5 mM DNA flin concentration of nucleotides) and incubated at 23 _+ 1 °C for 4 - 6 h, allowing the diol epoxide reactions to proceed to completion. The aqueous reaction mixtures were then extracted with buffer-saturated ether solution up to 20 times to remove the hydrolysis products of BADE or 3-MCDE not covalently bound to the DNA. Further purification procedures of the adducts involving precipitation of the modified DNA with cold ethanol were eliminated since some of the adducts,

125 especially those derived from the covalent binding of syn-BADE and 3MCDE to DNA were not sufficiently stable over longer periods of time. The level of covalent binding denoted by the parameter fcov (moles diol epoxide molecules covalently bound to DNA/moles initially present in the reaction mixture) was estimated utilizing two different methods. In the first, the concentration of covalently bound adducts in the aqueous, ether-extracted reaction mixtures was determined by measuring the absorbance due to the aromatic chromophores in the 3 5 0 - 4 0 0 nm region of the spectrum, as previously described [31]. In the second method, the quantity of tetraols recovered in the ether phase was determined spectrophotometrically utilizing reaction mixtures with and without DNA; the difference in absorbance between these two measurements is due to molecules which are covalently bound to the DNA and thus cannot be extracted with ether. The agreement in the fcov values determined by these two methods was within about _+15%, which is approximately equal to the reproducibility in these experiments. Hydrolysis kinetics in D N A solutions and determinations of physical association constants Two types of kinetic measurements (at 23 _+ l°C) were performed involving absorption spectrophotometry and fluorescence techniques. (1) Absorbance measurements. When BADE or 3-MCDE is added to an aqueous DNA solution, an immediate change in the absorption spectra of the diol epoxide molecules due to the formation of non-covalent complexes is observed (Fig. 2). As the diol epoxide molecules undergo chemical reaction (either covalent binding or formation of hydrolysis products), marked changes in the absorption spectra are observed (Fig. 3). The absorption spectra at various time intervals after mixing were determined with a Hewlett-Packard diode-array spectrophotometer (Model HP-4581). The pseudo-first order rate constant k of the reactions was determined by means of plotting the absorbance A at a given wavelength according to the equation: A(t) - A(oo) A(0) - A(oo)

= e x p [ - kt]

(1)

where A(t) is the absorbance at time t, and A(0) and A(oo) are the absorbances measured at the same wavelength initially (within 5 s of adding the diol epoxide molecules to the DNA solutions), and at the end of the reaction, respectively. The values of the apparent physical association constants K (M-I) [35] can be determined by measurements of the absorbance extrapolated to t = 0, the time of mixing, at different DNA concentrations, utilizing procedures described in detail elsewhere [29]. (2) Fluorescence. The fluorescence yields of the diol epoxides are generally lower than those of the hydrolysis products (mainly the tetraols). Consequently, the progress of the reaction can also be conveniently

126

(A)

.02 0

A

...~J

"... ~

I

I

320

~

.oe

O, OH

340

I

360

I

I

380

I

420 440

~,y~',

(a)

0H

~m .04 ~ .~z m n.O u~

.02

....

0

rn
320

340

360

.06 (C)

380

420 440

O.

~,OH H3C~

OH

.04

.02 0

I 340

i 360

i I 380 400

WAVELENGTH

i I 420 440

460

(nm)

Fig. 2. Comparison of the absorption spectra of anti-BADE (A), syn-BADE (B), and 3-MCDE (C) in 5 mM sodium cacodylate buffer solutions (pH 7.0) under various conditions: ( ): buffer solution, 4 s after addition of the diol epoxides; ( - -- --): buffer solution containing 2.25 mM DNA (A), 1.5 mM DNA (B), and 2.7 mM DNA (C), measured 4 s after addition of the diol epoxides (concentration 0.013 mM in A and mM in . . . . . . . ): same diol epoxide-DNA mixtures as above, but 60 min after mixing (equilibrated reaction mixture in which all diol epoxide molecules have been converted to hydrolysis products or covalent DNA adducts).

B,0.009

C);(

m o n i t o r e d b y f o l l o w i n g t h e f l u o r e s c e n c e y i e l d of t h e a n t h r a c e n e - l i k e c h r o m o p h o r e of t h e B A D E d i a s t e r e o m e r s a n d 3 - M C D E [31], a n d t h e e q u a t i o n F(t) -

F=~ -

F(O)

= (1 -

exp[

-

kt])

(2)

F(0)

w h e r e F(t) is t h e f l u o r e s c e n c e i n t e n s i t y of t h e r e a c t i o n m i x t u r e m e a s u r e d a t t i m e t. w h i l e F(0) a n d Fro. x a r e t h e i n i t i a l a n d f i n a l f l u o r e s c e n c e y i e l d s , respectively. P r o v i d e d t h a t t h e r a t e of f o r m a t i o n a n d d i s s o c i a t i o n of t h e n o n - c o v a l e n t diol e p o x i d e - D N A c o m p l e x e s [23] is m u c h f a s t e r t h a n t h e r a t e of t h e c h e m i c a l

127 '

.06

I

I

I

I

I

I

I

~

I

0 ,0 H

-

2"" "";-

H

.04

m 0

.02

0 I

I

I

I

I

I

:320

:340

:360

380

400

420

WAVELENGTH

I

44(

(rim)

Fig. 3. Absorption spectra of anti-BADE (0.0125 mM)-DNA (0.60 mM) buffer solution (pH 7.0) measured at various time intervals after addition of the diol epoxide to the DNA solution. ( ): absorption spectrum of non~ovalent diol epoxide-DNA complexes, measured 4 s after mixing (maxima at 388, 366, and 348 nm); (. . . . . . . ): absorption spectrum of equilibrated reaction mixture (40 m after mixing) consisting of tetraols and covalent DNA adducts; the absorption maxima occur at 380, 360 and 342 nm. Some intermediate absorption spectra obtained at successive 8-s time intervals after measuring the first (4 s) spectrum are also shown (the successive spectra are characterized by a decreasing absorbance at 388 nm.

reaction, the a p p a r e n t non-covalent association c o n s t a n t K can also be d e t e r m i n e d from t h e s e fluorescence m e a s u r e m e n t s by m e a s u r i n g k at different D N A c o n c e n t r a t i o n s and utilizing t h e equation [37]: k = k h (1 -

X b) -~- k 3 X b

(3)

w h e r e X b is t h e fraction of diol epoxide molecules bound non-covalently to DNA: Xb =

K[DNA]

(4)

1 + K[DNA] The r a t e c o n s t a n t k h p e r t a i n s to the h y d r o l y s i s of the diol epoxide molecules in buffer solutions in the a b s e n c e of DNA, while k 3 is the overall a p p a r e n t pseudo-first o r d e r r a t e c o n s t a n t [35] for the reaction of the diol epoxide molecules b o u n d non-covalently to the D N A to form both covalent a d d u c t s and h y d r o l y s i s p r o d u c t s (tetraols).

128 The fluorescence studies were performed utilizing a photon counting spectrofluorometer (Model 1902, Spex Industries, Edison, N J). The excitation wavelength was fixed at 357 _+ 2 nm, and the emission wavelength was fixed at 403 _+ 3 nm.

Measurements of orientations of chromophores bound to DNA by linear dichroism methods In this technique, the DNA molecules are oriented in a flow gradient utilizing a Couette cell [38,39]. The axes of the DNA double helices tend to align themselves parallel to the flow direction. The linear dichroism is defined as hA = All - A±, where All and A± are the absorbances measured with the polarization vector of the light oriented either in a parallel or perpendicular direction with respect to the flow direction. If the transition moment of a chromophore bound to the DNA is oriented at an angle (9 with respect to the flow direction, the relationship between this angle, hA, and the isotropic absorbance A, given by [29,31,32,38,39]: AA/A = (3/2)[3cos28 - 1IF(G)

(5)

where F(G) is an orientation factor (0 < F(G) < 1.0) which defines the degree of alignment of the DNA molecules in the flow gradient, while G is the velocity gradient. The transition moments are oriented within the planes of the DNA bases, and e is thus 90 ° within the DNA absorption band at 260 nm. The maximum possible value of AA/A at 260 nm is - 1.5. Because there is only a partial alignment of the bases with respect to the flow direction, we observe maximum values of AA/A of only - 0 . 1 0 when G = 1300 s-l; this result is in good agreement with previously published values [38,39]. In estimating e for the transition moments of polycyclic aromatic (PA) chromophores bound to DNA, the factor F(G) is usually eliminated by evaluating the ratio A' = {AA(PA)/A(PA)}

(6)

(AA/A)2eo and the equation [27]: cos2e = (1/3)(1 - A')

(7)

where the numerator in Eqn. 6 describes the net contribution of the polycyclic aromatic moiety to the total linear dichroism and absorbance at a given wavelength, while (AA/A)2eo is the reduced dichroism of the DNA molecule at 260 nm. When there are overlapping absorption bands at a given wavelength, it is difficult to evaluate AA(PA) and A(PA) accurately, particularly if unoriented molecules, as well as oriented molecules, contribute to the overall absorbance; this situation arises in solutions of noncovalent PA-DNA complexes, since only those molecules which are bound to

129 the DNA molecules contribute to the linear dichroism, but both free and bound PA chromophores contribute to the absorbance A. In our experiments, the AA values are obtained in relative units [29,31,32]; considerable information about the chromophore orientations can be gleaned from the wavelength dependence and the sign of AA. For 0 ~> 55 °, AA is negative in sign, while AA ~> 0 when O is less than 55 °. More exact values of O can, in principle, be evaluated from the values of A ' and Eqn. 7; the values of A' at the most important wavelengths of observation are thus provided in the appropriate figure legends. RESULTS

Non-covalent complex formation and reaction rate constants The absorption spectra of syn-BADE, anti-BADE, and 3-MCDE in buffer solutions, and in the presence of an excess of DNA measured within 5 s of mixing, are shown in Fig. 2. The formation of non-covalent complexes is indicated by the red shift in the vibronic maxima of the first electronic transition and a decrease in the absorbance in the 340--440 nm range. As in the case of anti-BPDE, the formation of these complexes occurs on time scales much faster than 5 s, and probably on time scales faster than a few ms [40]. In the absence of DNA, the hydrolysis rate constant, k h, of the syn diastereomer of BADE is about 6 _+ 2 times faster than that of antiBADE, while 3-MCDE exhibits an intermediate value (Table I). The reasons for the higher reactivity of the syn-BADE diastereomer are discussed by Sayer et al. [41].

TABLE I C O M P A R I S O N OF R E A C T I O N R A T E C O N S T A N T S (k h, ks), P H Y S I C A L B I N D I N G C O N S T A N T S (K) A N D P E R C E N T C O V A L E N T B I N D I N G 0f~,) F O R anti-BADE, sFn-BADE, A N D 3-MCDE

f~(%)

kl~($ -1)

anti-BADE

syn-BADE

3-MCDE

23 - 3 (1.7 ± 0.3) x 10 -~

8 + 1 (1.1 + 0.3) x 10 -8

22 - 3 (7 + 1) x 10 -4"

(2.0 + 0.3) x 10 -2 (1.9 ± 0.2) x 10 -2

(9 ± 2) x 10 -s (11 ± 3 ) × 10 -s

(2.4 -+ 0.2) X 10 -e (3.9 ± 0.8) x 10 -2

k 8 (s-1)

By fluorescence By a b s o r p t i o n

k2/kb

By fluorescence By a b s o r b a n c e K (M -1) By fluorescence By a b s o r b a n c e •F r o m Ref. 31.

120 ± 25 112 ± 23

8 ± 3 10 ± 3

35 ± 6" 56 ± 14

4000 ± 300 3600 ± 400

600 + 56 610 ± 60

4400 ± 600" 4600 ± 200

130 I

!

I

I

J

I

I

i

|

I

I

!

I

I

J

i

I

I

I

J

I

I

I

I

~

I

I

i

I

2O

~,X

5

-

0

5

10

15 I

20

25

30

x 10 - 3 , M -1

[ONA]

Fig. 4. Plot of data according to Eqn. 8 from which values of the parameters ks and K are obtained utilizing either the fluorescence technique (e, anti-BADE; i , syn-BADE, both 0.023 mM), or the absorption technique (A, 3-MCDE, 0.009 mM); all solutions at pH 7.0, 5 mM sodium cacodylate.

The reaction r a t e c o n s t a n t k of all t h r e e diol epoxides increases with increasing DNA c o n c e n t r a t i o n , t h u s following the p a t t e r n established by B P D E and o t h e r a r o m a t i c epoxides and diol epoxides [29,32,37,42]. T h e values of k m e a s u r e d b y e i t h e r t h e UV a b s o r b a n c e or the fluorescence m e t h o d w e r e found to be t h e same within e x p e r i m e n t a l e r r o r . Upon i n v e r t i n g Eqn. 3, the following linearized (with r e s p e c t to the DNA concentration) equation is obtained: 1

1 =

k - kh

-

-

k8 - kh

1 +

(8)

(k 8 - kh)K[DNA]

Typical plots of the r a t e c o n s t a n t s k as a function of the DNA concentration, according to this equation, are shown in Fig. 4. T h e noncovalent association c o n s t a n t K and k 3, t h e reaction r a t e constant of diol epoxide molecules complexed with DNA, can be obtained from t h e relationship b e t w e e n t h e slope and the i n t e r c e p t . T h e values of t h e s e p a r a m e t e r s are s u m m a r i z e d in Table I. T h e non-covalent association c o n s t a n t K can be obtained m o r e d i r e c t l y by

131

- (A) I

I

i

I

i

i

i

i

i

i

I

'

'

'

'

0.05 ! {3 0

0,04

hi (.3 Z

0.03

nn nO

0.02

~r

o') m
• anti-BADE

• syn- BADE

O. Ol I

,

,

,

,

0

I

,

,

~

0.5

,

I

,

,

J

,

I

1.5

1.0

[DNA] ,M

- 500

(B) /a~//~

--'

I

-

I

I

I

I

|

I

I

I

I

I

-400

'i" r~ - 3 0 0 0 <] -

200

-100 I

o

I

1ooo

I

I

2000

3000

4000

[ D N A ] -1 , M - ]

Fig. 5. (A) Absorbance of anti-BADE ( o ) and syn-BADE ( I ) , both 0.013 mM, at 388 nm as a function of DNA concentration (in 5 mM sodium cacodylate buffer solution, pH 7.0), measured 5 s after mixing. (B) Benesi-Hildebrand plot of the data in A.

monitoring the decrease in the absorbance at 388 nm as a function of increasing DNA concentration, measured within several seconds after mixing the diol epoxide into the DNA solutions (Fig. 5A). The slopes and intercepts of Benesi-Hildebrand plots [43] of this data (Fig. 5B) provide values of K [29]. The values of K obtained by the absorbance and fluorescence methods are in reasonable agreement with one another (Table I). The association constant K is five to seven times smaller in the case of the syn-diastereomer of BADE than in the case of the anti-diastereomer. A similar, though smaller difference in the association constants was found in the cases of syn- and anti-BPDE [29]. The catalytic effect of DNA on the reactivities of the diol epoxides can be estimated from the ratio k3/k h. For syn-BADE this ratio is 9 _+ 4, while for anti-BADE this ratio is 115 _+ 25. For 3-MCDE this ratio is 50 _+ 20.

132

Covalent binding The fraction for of diol epoxide molecules which bind covalenty to DNA, is in the range of 19--26O/o for 3-MCDE and anti-BADE, and about 80/0 in the case of syn-BADE. These levels of covalent binding are considerably higher than those previously reported for anti-BPDE (8--10o/0) and syn-BPDE (2-30/0) under similar experimental conditions [29]. Thus, the BADE and BPDE diastereomers are characterized by similar relative abilities to bind covalently to DNA, except that the BA diol epoxide diastereomers are approx. 2--3 times more reactive with respect to binding covalently to DNA than the corresponding BP diol epoxide diastereomers.

Conformations of non-covalent complexes The anthracene-like aromatic chromophore of BADE and 3-MCDE is characterized by a strong electronic absorption band near 260 nm (in aqueous solutions) and a weaker electronic-vibronic absorption band system in the 300--400 nm region of the spectrum. The 260 nm absorption band is polarized along the long-axis of the anthracene ring system (Fig. 1), while the weak near-UV absorption band is polarized along the short axis [44]. The polarization properties of these two transitions, in conjunction with the linear dichroism technique, can be utilized to obtain information about the tilt of the plane of the aromatic ring system of the diol epoxide molecules with respect to the average orientations of the planes of the DNA bases. The determination of the linear dichroism properties of the short axis transition is not difficult since the corresponding absorption band is located above 340 rim, and therefore does not overlap with the absorption band of the DNA. However, the 260 nm absorption band of the anthracene ring system does overlap with the DNA absorption band, thus complicating the measurement of the dichroism in this wavelength region. Fortunately, the molar extinction coefficients of the anthracenyl chromophores of BADE and 3-MCDE near 260 nm are over ten times larger than those of the DNA bases. Utilizing signal averaging and computer manipulation of the dichroism spectra, it is possible to determine the dichroism of the chromophores of such small molecules bound to the DNA even below 300 nm, when there is approximately one bound residue per 100 bases [31]. The orientation of the aromatic ring system relative to the orientation of the planes of the DNA bases can thus be determined more accurately. More specifically, difference absorption spectra and difference linear dichroism spectra were obtained by subtracting the absorption and hA spectra of unmodified DNA from those of the non-covalent carcinogen-DNA complexes or covalent adducts, the DNA concentration being identical in both samples. (1) Anti-BADE. The linear dichroism spectrum of the non-covalent DNA complex of anti-BADE is shown in Fig. 6A. Both the short and the long axes of the anthracenyl chromophore are characterized by negative hA signals. In contrast to the absorption spectra (Fig. 2A), the linear dichroism spectra of these non-covalent complexes are considerably sharper and are characterized by minima at 355, 372 and 392 nm (Fig. 6A). In part, the diffuseness of the

133

O-(A)

-0.5

<~ (/3 0 nT

-I

.0 0

~

-

-

(B)

:3 a Y ~

392 _

-

.I....~

a LIJ Z _i~

-1.0

-2.0 240

•

~

3

7

280

0 320

WAVELENGTH

370 I 360

t 400

(nm)

Fig. 0. Linear dichroism spectra (vertical scale in arbitrary units) of non-covalent anti~BADEDNA (A) and syn-BADE-DNA (B) complexes. The spectral scans were initiated 20 s after mixing (0.06 mM diol epoxide and 0,13 mM DNA in 5 mM sodium cacodylate buffer solution adjusted to pH 8.5 in order to decrease the rate of reaction), and were carried out at a scanning rate of 120 nm/min. The spectra shown are due to the non~ovalently bound diol epoxide molecules, and were obtained by subtracting the linear dichroism spectra of 0.13 mM DNA solutions from those of the complexes. For clarity, the AA spectra in the 300--420 nm region are shown on an amplified scale as well (amplified by a factor of 12 in (A) and a factor of 25 in (B)). Typical values of A' (see Eqn. 6) are +0.12 (260 nm) and +0.022 (392 nm) in (A), and +0,12 (260 nm) and + 0.028 nm (391 nm) in (B),

a b s o r p t i o n s p e c t r u m in F i g . 2 is d u e t o t h e p r e s e n c e of a s i g n i f i c a n t f r a c t i o n of f r e e m o l e c u l e s ( a b o u t 340/0 c a l c u l a t e d a c c o r d i n g t o E q n . 4, t h e v a l u e of K g i v e n in T a b l e I, a n d u t i l i z i n g t h e i n f o r m a t i o n p r o v i d e d in t h e l e g e n d of F i g . 6). S i n c e t h e f r e e m o l e c u l e s d o n o t c o n t r i b u t e t o t h e l i n e a r d i c h r o i s m s p e c t r u m , i t is r e a s o n a b l e t h a t t h e AA s p e c t r u m s h o u l d e x h i b i t a b e t t e r resolution than the absorption spectrum. In general, orientation angles of the different transition moments of the c h r o m o p h o r e s b o u n d t o D N A c a n b e e s t i m a t e d f r o m v a l u e s of A ' . H o w e v e r ,

134 it is difficult to determine A ' for physical PA diol epoxide-DNA complexes because both free and complexed chromophores are present in the solutions; only the DNA-bound molecules contribute to the linear dichroism, but both free and bound molecules contribute to the absorbance. Since only the contribution of the bound molecules is of interest in evaluating the quantity AA(PA)/A(PA), from which, in turn, A ' is calculated, the magnitudes of A' provided in the legends of Figs. 6 and 7 are lower bounds. Depending on the sign of A', the values of e estimated from these A ' values and Eqn. 7 are thus either lower or upper bound values of these orientation angles. In the case of the non-covalent anti-BADE-DNA complexes it is therefore possible to conclude only that e ~, 55 ° for both the short and the long axes of the anthracenyl chromophore, since the sign of A ' is positive in both cases. The conformations of the non-covalent complexes are thus consistent with those of intercalation complexes; however, conformations which are similar, but not identical to those of classical intercalation complexes, cannot be excluded. (2) syn-BADE. The linear dichroism due to the short and long-axis polarized absorption bands are also both negative in sign (Fig. 6B). The A ' values are consistent with (9 ~ 55 ° for both the long and the short axes transition moments, as expected for intercalation complexes. However, in contrast to the AA spectrum of the anti-diastereomer, the AA spectrum in the 320--400 nm region is diffuse; this suggests a considerably greater heterogeneity of conformations of syn-BADE complexes than of anti-BADE complexes. The greater physical bulkiness of the quasi-diaxial substituted cyclohexenyl ring [45--47] in the syn diastereomer might prevent a good intercalative fit between adjacent base pairs, thus accounting for the 5--7fold lower overall association constant K (Table I), and a less specific, more heterogeneous set of conformations, as suggested by the diffuseness of the AA spectrum (Fig. 6B). (3) 3-MCDE. A linear dichroism spectrum of an aqueous 3-MCDE-DNA reaction mixture is shown in Fig. 7. When the linear dichroism scan is initiated 20 s after mixing, a prominent negative 5.4 signal due to the long axis transition is observed at 268--270 nm (A' ~ + 0.27 and thus O ~, 61°); however, for the short axis transition, the dichroism is positive in sign in the 3 5 0 - 4 5 0 nm wavelength range (A' < -0.97, O < 36 o). The short axis of the physically bound 3-MCDE molecules is therefore tilted away from the planes of the DNA bases and closer towards the axis of the helix. It should be noted that the orientation of the short axis of physically complexed 3-MCDE molecules is thus quite different from the orientation of the short axis of the two BADE stereoisomers. The linear dichroism of the physical 3-MCDE-DNA complexes is therefore not consistent with classical intercalation, since the planar aromatic portion of the 3oMCDE molecules is clearly not parallel to the planes of the DNA bases. This behavior may be attributed to the bulkiness of the methyl and methylene substituents, which prevent the 3-MCDE molecules from adopting an intercalative conformation. If the linear dichroism scan is initiated 80 s after mixing, the linear

135 i

i

i

0-

I -

i -"

[t

-

' N

ii

~-..,,,OH

~ " ~ 0

H ~ +0.6

-

-lo

+0.4

:2o +0.2

- 2 .(3

o

_J

-0.2 -3.0

[ 240

280

i 320

:.360

WAVELENGTH

400

l 440

I 480

(nm)

Fig. 7. Linear dichroism spectrum of non-covalent 3-MCDE (0.015 mM)-DNA (0.075 mM) complexes. All conditions as in the legend of Fig. 6, except that the amplification vertical scale factor in the 320--480 nm region of the spectrum is 3.3. Two linear dichroism difference spectra were determined with the scans initiated 20 s and 80 s after mixing. Note that the AA signals are opposite in sign for the short and long wavelength transitions. In the case of the 20-s spectrum, the values of A' are + 0.27 (260 nm) and -0.97 (390 nm).

dichroism signals due to b o t h t h e s h o r t and the long axes are significantly r e d u c e d (Fig. 7). Analogous kinetic effects are also o b s e r v e d in the ease of t h e B P D E s t e r e o i s o m e r s [29,48]. With increasing time a f t e r mixing, the m a g n i t u d e of t h e AA signal diminishes as t h e non-covalently bound diol epoxide molecules u n d e r g o chemical r e a c t i o n to covalent adduets and hydrolysis products. While t h e t e t r a o l h y d r o l y s i s p r o d u c t s of 3-MCDE also form physical c o m p l e x e s with DNA, t h e i r binding c o n s t a n t s are lower [31], t h u s accounting for t h e diminishing linear dichroism signals as the diol epoxide molecules u n d e r g o reaction.

Conformations of the covalent adducts T h e c h a r a c t e r i s t i c s of t h e covalent a d d u c t s are b e s t d e s c r i b e d in t e r m s of the site I/site II classification of binding sites [21,23]. Syn-BADE. T h e a b s o r p t i o n and linear dichroism c h a r a c t e r i s t i c s d e r i v e d from t h e covalent binding of syn-BADE to D N A are shown in Figs. 8 and 9. T h e dichroism is n e g a t i v e in sign within the 330--400 nm band s y s t e m (Fig. 8B) as well as within t h e 240--270 nm a b s o r p t i o n band (Fig. 9F). The linear dichroism minima coincide with t h e m a x i m a in t h e a b s o r p t i o n s p e c t r u m of t h e a d d u c t s which are s i t u a t e d at 347, 367 and 386 nm (Figs. 8A and 8B). T h e s e a b s o r p t i o n m a x i m a a r e red-shifted b y a b o u t 6 nm with r e s p e c t to

136

(A)

A

(c)

0

-.06

~~OH

,~L)HH w f,.) Z

.04

.02

m

o

.01

"~

0

fJ) m

0

0,,

.02 0

I

e

)

I(D'

~

'

'

I

,

,

,

,

I

-

5

3 I .j

0 -l

"1c.) c) i ]

310

,

l

350

,

,

I

400

-I

,

,

,

310

I

350

,

,

,

,

I

,--3

400

WAVELENGTH (nrn) Fig. 8. Absorption spectra (A) and (C), and linear dichroism spectra (B) and (D) of covalent anti. BADE-DNA and syn-BADE-DNA adduets within the short-axis polarized absorption band. The covalent adducts were prepared by reacting the diol epoxide molecules (0.059 raM) with DNA (0.50 raM) and isolated as described in the text. The values of A ' are + 0.13 (367 nm) for the synBA-DNA adduct, and - 0.17 (362 nm) and + 0.41 (390 nm) for the anti-BA-DNA adduet.

those of the tetraols (derived from the hydrolysis of the BADE stereoisomers in aqueous solutions), which is less than the 11--12-nm redshift characterizing the formation of the physical BADE-DNA complexes (Fig. 6). The negative AA peak at 250--253 nm (Fig. 9F) is slightly blueshifted with respect to the absorption maximum which occurs at 256 nm (Figs. 9D and 9E). Since the sign of the linear dichroism within both the short and long axes absorption bands is negative, these results suggest that the covalent syn-BADE adducts are of the site I type, the anthracenyl chromophore tending to be tilted towards the planes of the DNA bases. Since the linear dichroism spectrum in the 330--400 nm wavelength range resembles an inverted absorption spectrum, there is no evidence of the presence of site II species which would have given rise to positive contributions to the AA spectrum. The covalent binding of the syn-BADE diastereomer to DNA can thus be classified as predominantly of the site I type. Utilizing the A ' values given in the legends of Figs. 8 and 9, the following orientation angles are estimated: 0 = 57 ° (short axis), 0 = 60 ° (long axis). Anti-BADE. In the case of the covalent anti-BADE-DNA adducts, the linear dichroism spectrum is dominated by positive maxima at 343, 360 and 380 nm, which coincide approximately with the corresponding absorption

137

2.o(A)

o

I-

12.oi (D)

""~",,.

I

0

o.o.

---ADDUCT O-

0

I

h /\ ~<1

-2

240

o

J

,

,

0

a

J

,

/

-\ ~

n

"-~'~,.

5 p ~ .I ,

260

,

,

280

,

~

,

I

~

,

:300 320 240 260

,

280

. 3 0 0 320

WAVELENGTH (n m )

Fig. 9. Absorption spectra of covalent adducts and unmodified DNA (A and D), difference absorption spectra (B and E), and difference linear dichroism spectra (C and F) of covalent antiBADE-DNA and sya-BADE-DNA adducts. The difference spectra were obtained by subtracting either the absorption or linear dichroism spectra of unmodified DNA from those of the adduets by normalizing these spectra to one another in the 300--320 nm region (minimal absorbance of the anthraeenyl chromophore). The covalent adducts were prepared as described in the text by mixing either 0.06 mM anti-BADE or 0.12 mM syn-BADE into a 0.13 mM DNA solution. The estimated values of A' are -0.42 (260 nm, anti-BA-DNA adduct) and +0.25 (252 nm, syn-BADNA addnct).

m a x i m a a t 347, 362 and 380 n m (Figs. 8C and 8D); t h e s e m a x i m a r e p r e s e n t at m o s t a 2-nm r e d shift w i t h r e s p e c t to t h o s e of t h e t e t r a o l h y d r o l y s i s p r o d u c t s . T h e s e a b s o r p t i o n and p o s i t i v e linear d i c h r o i s m c h a r a c t e r i s t i c s a r e due to site I I a d d u c t s . H o w e v e r , n e g a t i v e hA m i n i m a a r e also o b s e r v e d at 370 and 390 nm. T h e s e m i n i m a in t h e hA s p e c t r u m a l m o s t coincide (within 2 nm) w i t h t h o s e e x h i b i t e d b y t h e i n t e r c a l a t e d non-covalent a n t i - B A D E c o m p l e x e s (see Fig. 6A), and a r e t h u s due to t h e p r e s e n c e of t y p e I adducts; it should b e n o t e d t h a t t h e r e a r e no visible a b s o r p t i o n m a x i m a or s h o u l d e r s (Fig. 8C) c o r r e s p o n d i n g to t h e s e n e g a t i v e linear dichroism bands. T h e s t r e n g t h of a linear d i c h r o i s m signal d e p e n d s not only on t h e c o n t r i b u t i o n of a p a r t i c u l a r species to t h e o v e r a l l a b s o r b a n c e a t a g i v e n w a v e l e n g t h , b u t also on its o r i e n t a t i o n [31]; t h u s , a m i n o r species can s t r o n g l y c o n t r i b u t e to

138

the linear dichroism spectrum, but not to the absorption spectrum. Since there is no visible structure in the absorption spectrum in the vicinity of 390 nm, and since the overall absorption spectrum with maxima at 346, 360 and 380 nm is unshifted (relative to tetraols), it is clear that site II adducts dominate the absorption spectrum in this region of the spectrum, and thus dominate the binding of anti-BADE to DNA. A small shoulder at 250 nm in the absorption spectrum (Fig. 9B) is observed, however, this small shoulder, which coincides with a negative linear dichroism minimum at 250 nm (Fig. 9C), is due to site I adducts. The mixed positive/negative linear dichroism in the 320--400 nm region of the spectrum is also evident in the absorption and linear dichroism spectra below 300 nm which are due to the long-axis transition moment of the anthracenyl chromophore. The difference linear dichroism spectrum exhibits a positive peak at 261 nm, and a minimum at 252 nm (Fig. 9C); corresponding small peaks superimposed on the broad DNA absorption band are observed in the absorption spectrum of the adducts (Fig. 9A, see also the difference absorption spectrum in Fig. 9B). The positive dichroism at 260 nm is characterized by a value of 0 = 47 o for the long axis of the anthracenyl chromophore; this is an upper bound for O, since the magnitude of A ' evaluated at 260 nm is probably underestimated due to the contribution of a negative site I dichroism; this diminishes the positive dichroism at 260 nm, and thus the magnitude of A ' attributable to site II adducts at 260 nm. The negative dichroism contribution of site I adducts is most pronounced near 250 nm (Fig. 9C). Because of the overlap of AA spectra due to site I and site II in the 3 5 0 400 nm region as well, only an upper bound of O = 51 ° (A' = - 0 . 1 7 at 362 nm) and a lower bound of O = 64 ° (t1' = + 0.40 at 390 rim) can be estimated for the orientations of the short axes of the two types of different adducts. 3-MCDE. In the case of the adducts derived from the covalent binding reaction of 3-MCDE with DNA, both the absorbance and the linear dichroism signals are very weak in the 360--420 nm region of the spectrum (Fig. 10). While the linear dichroism signal in the 360--400 nm range is small, it is definitely (and reproducibly) negative in sign, indicating that for the short axis O ~ 55 °. The difference absorbance and difference linear dichroism spectra due to the long axis transition in the 240--300 nm region are shown in Fig. 10A; the linear dichroism is positive in sign, and thus of the site II type. In contrast to results obtained with the covalent adducts derived from anti-BADE, there is no evidence of the presence of site I adducts in the linear dichroism spectra of the 3-MCDE°DNA covalent adducts. Since the stereochemistry of the diol and epoxide groups are the same in both antiBADE and 3-MCDE, it is evident that the methyl group at the 3-position, and the bridged methylene groups in 3-MCDE, prevent the formation of quasi-intercalative site I covalent adducts. The linear dichroism signals due to the long and the short axis of 3-MCDE are opposite in sign in the non-covalent complexes (Fig. 7) and in the covalent adducts (Fig. 10). It is evident that both the short and the long axis

139 0.16 (A) 0.14

A
(B)

265

~

O

H

0.12 O.lO

H 3 C ~

6

o.o8

375

-1-

0.06

¢D

0.04 nr

0,5 OH

0.4 ~ W Z 0,2 m '~ n," 0.1 0

0.3

\~xlO

0.02 0 -

--

0

m '~

o o2

- 0.04 I

240

I

260

I

I

280

I

I

300

360

380

400

420

WAVELENGTH (nm) Fig. 10. Absorption ( ) and difference linear dichroism spectrum (noisy traces) of the longaxis (A) and short-axis (B) polarized transitions of the anthracenyl chromophore in covalent 3MCDE-DNA adducts (the covalent adduct was obtained by reacting 0.015 mM 3-MCDE with 0.075 mM DNA in buffer solution). In the case of the difference linear dichroism spectrum, the hA curve of the adduct and of the unmodified DNA were normalized with respect to one another at 310--320 nm before subtraction. A t 265 nm, the value of A ' is - 0 . 2 0 .

of the anthracenyl chromophore of 3-MCDE undergo a significant reorientation when physically bound 3-MCDE molecules bind covalently to DNA. Since A' = - 0 . 2 0 at 267 nm, O = 51 ° for the long axis, which is consistent with a previous estimate [31]. DISCUSSION

Relative reactivities of the different diol epoxide derivatives with D N A In buffer solution, the reaction rate constant k h is largest in the case of s y ~ B A D E {Table I). When complexed non-covalently with DNA, the pseudofirst order reaction rate constant of all three diol epoxides increase significantly. The rate constants k 3, defining the rates of reactions of the non-covalently bound diol epoxide molecules, presumably via a ratedetermining step of formation of intermediate carbonium ions [49], are about the same in magnitude for anti-BADE and 3-MCDE. In the case of synBADE, k s is two to three times smaller than for the other two compounds. Particularly striking are the differences in the non-covalent association constants K of anti and syn-BADE. Both types of complexes are characterized by linear dichroism spectra which are consistent with those of intercalation complexes, thus paralleling the characteristics of syn-BPDE and anti-BPDE [29]; in the case of the benzo[a]pyrene diol epoxide diastereoisomers, however, the value of K is only 50% smaller for the syn

140 than for the anti diastereomer, while in the case of BADE the association constant is 6--7 times smaller in the case of the s y n stereoisomers. The observed lower intercalative binding of syn diastereomers relative to the anti isomers was predicted earlier [45--47] on the basis of the greater bulkiness of the cyclohexene ring in the syn diastereomers; syn diol epoxides prefer a more bulky quasi-diaxial conformation, while the anti-diastereomers favor the more compact quasi-diequatorial conformation [41,50--53]. Jerina et al. [45] speculated that the more bulky quasi-diaxial conformations of s y n diol epoxide stereoisomers might account for the generally observed lower tumorigenic activities of these isomers. In a cellular environment, the ability of a given polycyclic aromatic diol epoxide compound to adopt a quasidiequatorial conformation might indeed be an important factor in determining its biological activity [35]. However, other factors such as the relative chemical reactivities with DNA, the yield of covalent DNA adducts formed versus the yield of hydrolysis products, and the characteristics of the covalent DNA adducts, should also play an important role in biological activity. The fraction of diol epoxide molecules which bind covalently to DNA is also smaller for syn-BADE than for the other two diol epoxide compounds studied; f v is 8 _+ 1%, which is about three times smaller than in the case of a n t i - B A D E and 3-MCDE (Table I). Thus, considering the physical association constants K, the chemical reactivities of the complexes, k 3, and the covalent binding efficiencies, the reactivity of syn-BADE with DNA in vitro is significantly lower than that of the anti diastereomer. Comparing now the relative reactivities of a n t i - B A D E and 3-MCDE, the values of K, k 8, and fcov are remarkably similar for these two compounds (Table I). Thus, the alkyl substituents in 3-MCDE do not seem to affect significantly the chemical reactivities, and levels of non-covalent and covalent DNA binding of these related diol epoxide molecules. However, the alkyl substituents in 3-MCDE prevent the formation of physical intercalation complexes, even though the magnitude of the physical association constant K is unaffected. While the A/I spectra of physical anti. BADE-DNA complexes suggest an intercalative conformation, in the case of 3-MCDE the long axis tends to be aligned parallel to the planes of the DNA bases, while the short axis is tilted closer to the axis of the double helix. Among the various non-covalent polycyclic aromatic epoxide-DNA complexes which have been studied up till now [29,31,32,42], 3-MCDE is the first compound which forms physical complexes with DNA with linear dichroism properties different from those expected for intercalation complexes. Apparently, the bulkiness of the aliphatic substituents in 3-MCDE prevent an intercalative fit between adjacent base pairs. The relevance of physical, intercaiative complex formation to the formation of the biologically more significant covalent adducts has been discussed in the literature [34,35]. The behavior of 3-MCDE suggests that the nature of the pre-reaction, physical complexes do not have to be intercalative in nature for extensive covalent binding to DNA to occur. This is evident

141 from the relative behavior of anti-BADE and 3-MCDE. The former gives rise to intercalation complexes before covalent binding, while 3-MCDE does not; nevertheless, the level of covalent binding is similar in both cases (Table I).

Conformations of the covalent adducts The dominance of site II in the anti-BADE-DNA adducts, and the dominance of site I conformations in the syn-BADE-DNA adducts, parallels the behavior of the corresponding adducts derived from the covalent binding of racemic anti- and syn-BPDE to DNA [29,30,54]. However, site I adducts are not evident in the covalent adducts derived from the covalent binding of 3-MCDE to DNA. Thus, the methyl substituents in 3-MCDE not only interfere with the formation of intercalative physical complexes, but also with the formation of quasi-intercalative site I covalent adducts. The findings reported here are in accord with the apparent correlations between tumorigenic activities and mutagenicities in mammalian cell systems, and the preponderance of site II adducts observed with a number of different polycyclic aromatic diol epoxides [21,23,27,28,30]. According to the trends observed, covalent syn-BADE-DNA adducts are of the site I type, and are expected to be less tumorigenic than the corresponding adducts derived from anti-BPDE and 3-MCDE. In both of these two latter cases, site II covalent adducts are dominant. The tumorigenic bay region diol epoxide stereoisomers derived from BP [48], 3-methylcholanthrene [31], and 5-methyl chrysene [32], undergo striking reorientations as a result of the formation of the covalent products, forming adducts with apparently external, or solvent accessible conformations. The less active isomers, e.g. (-)~znti-BPDE [25,48] and racemic syn-BPDE [29,30] are predominantly of the site I type, and thus appear to suffer little reorientation upon covalent binding. Thus, the racemic BADE stereoisomers studied in this work also seem to follow this pattern. The site I/site II heterogeneity observed in the case of the anti-BADE-adducts may, as in the case of the (+)- and ( - ) - B P D E enantiomers, be due to adducts derived from ( - ~ and (+)-BADE enantiomers, respectively. This hypothesis will be verified when the optically pure enantiomers of anti-BADE become available to us. Finally, studies of the conformations of carcinogen-DNA adducts and correlations with the biological effects produced by these adducts, may ultimately provide a better understanding of the mechanisms of mutagenesis and the initiation events in tumorigenesis on a molecular level. Some possible mechanisms explaining the differences in the biological effects of base-base stacked (site I) and carcinogen-base stacked (site II) adducts have been advanced [55]. ACKNOWLEDGEMENTS

This work was supported by the U.S. Public Health Service, Grant CA 20851 awarded by the National Cancer Institute, Department of Health and

142

Human Services, and in part by the Department of Energy (Contract DEAC02-78EV04959 and Grant FG-0286ER60405) at New York University. The portion of the work carried out at the University of Chicago was supported by the American Cancer Society (Grant BC-132). We gratefully acknowledge the assistance of Dr. Y. Mnyukh in performing the linear dichroism measurements. REFERENCES 1

2

3

4 5

6

7

8

9

10

11

12

13 14

T.J. Slaga, G.T. Boroden, J.D. Scribner and R.K. Boutweli, Dose-response studies on the ability of 7,12~limethylbenz[a]anthracene and benz[a]anthracene to initiate skin tumors, J. Natl. Cancer Inst., 53 (1974) 1337-1340. T.J. Slaga, E. Huberman, J.K. Selkirk, R.G. Harvey and W.M. Bracken, Carcinogenicity and mutagenicity of benz[a]anthracene diols and diol epoxides, Cancer Res., 38 (1978~ 16991704. A. Dipple, R.C. Moschel and C.A.H. Bigger, Polynuclear aromatic carcinogens, in: C.E. Searle (Ed.), Chemical Carcinogens, ACS Monograph 182, American Chemical Society, Washington, D.C., 2nd edition, Vol. 2, 1984, pp. 41-163. A.H. Conney, Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic aromatic hydrocarbons, Cancer Res., 42 (1982) 4875-4917. S~K. Yang, M. Mushtaq and P.-L. Chiu, Stereoselective metabolism and activation of polycyclic aromatic hydrocarbons, in: R.G. Harvey (Ed.), Polycyclic Hydrocarbons and Carcinogenesis, ACS Symposium Series No. 283, The American Chemical Society, Washington, D.C., 1985, pp. 19-34. P.G. Wislocki, M.K. Buening, W. Levin, R.E. Lehr, D.R. Thakker, D.M. Jerina and A.H. Conney, Tumorigenicity of the diastereomeric benz[a]anthracene 3,4-dihydrodiol in newborn mice, J. Natl. Cancer Inst., 63 (1979~201-204. J. Jacob, G. Grimmer and A. Schmoldt, The influence of polycyclic aromatic hydrocarbons as inducers of monooxygenases on the metabolite profile of benz[a]anthracene in rat liver microsomes, Cancer Lett., 14 (1981) 175--185. D.R. Thakker, W. Levin, H. Yagi, D. Ryan, P.E. Thomas, J.M. Karle, R.E. Lehr, D.M. Jerina and A.H. Conney, Metabolism of benzo[a]anthracene to its tumorigenic 3,4-dihydrodiol, Mol. Pharmacol., 15 (1978} 138-- 153. A,D. MacNicoli, C.S. Cooper, O. Ribiero, K. Pal, A. Hewer, P.L. Grover and P. Sims, The metabolic activation of benz[a]anthracene in three biological systems, Cancer Lett., 11 (1981) 243 -- 249. R.E. Lehr, S. Kumar, W. Levin, A.W. Wood, R.L. Chang, A.H. Conney, H. Yagi, J.M. Sayer and D.M. Jerina, The bay region theory of polycyclic aromatic hydrocarbon carcinogenesis, in: R.G. Harvey (Ed.), Polycyclic Hydrocarbons and Carcinogenesis, ACS Symposium Series No. 283, The American Chemical Society, Washington, D.C., 1985, pp. 63-38. A.W. Wood, R.L. Chang, W. Levin, R.E. Lehr, M. Schaefer-Ridder, J.M. Karle, D.M. Jerina and A.H. Conney, Mutagenicity and cytotoxicity of benz[a]anthracene diol epoxides and tetrahydro-epoxides: exceptional activity of the bay region 1,2~poxides, Proc. Natl. Acad. Sci. U.S.A., 74 (1977) 2446-2750. W. Levin, R.L. Chang, A.W. Wood, H. Yagi, D.R. Thakker, D.M. Jerina and A.H. Conney, High stereoselectivity among the optical isomers of the diastereomeric bay region diol epoxides of benz[a]anthracenes in the expression of tumorigenic activity in murine tumor models, Cancer Res., 44 (1984)929--933. W. Levin, D.R. Thakker, A.W. Wood, R.L. Chang, R.E. Lehr, D.M. Jerina and A.H. Conney, Evidence that benzo[a]anthracene 3,4-diol-l,2-epoxide is an ultimate carcinogen on mouse skin, Cancer Res., 38 (1978) 1705--1710. R.G. Harvey, Activated metabolites of carcinogenic hydrocarbons, Acc. Chem. Res., 14 (1981) 218- 226.

143 15

16

17

18 19

20

21

22

23

24 25 26

27

28

29

30

31

32

33

P. Brookes and M.R. Osborne, Mutation in mammalian cells by stereoisomers of antibenzo[u]pyrene diol epoxide in relation to the extent and nature of the DNA reaction products, Carcinogenesis, 3 (1982) 1223-1226. L.L. Yang, V.M. Maher and J.J. McCormick, Relationship between excision repair and the cytotoxic and mutagenic effect of the "antz~' 7,8-diol-9,10-epoxide of benzo{a]pyrene in human cells, Mutat. Res., 94 (1982) 435--447. S.W. Ashurst, G.W. Cohen, S. Nesnow, J. DiGiovanni and T.J. Slaga, Formation of bonzo[a]pyrene-DNA adducts and their relationship to tumor initiation in mouse epidermis, Cancer Res., 43 (1983) 1024-1029. C.W. Stevens, N. Bouck, J.A. Burgess and W.E. Fahl, Benzo[a]pyrene diol epoxides: different mutagenic efficiency in human and bacterial cells, Mutat. Res., 152 (1985) 5--14. J.A. Burgess, C.W. Stevens and W.E. Fahl, Mutation at separate gene loci in Salmonella typhimurium TA100 related to DNA nucleotide modification by stereoisomeric benz[a]pyrene-7,8-diol-9,10-epoxides, Cancer Res., 45 (1985) 4257-4262. RA~.Newbold, P. Brookes and R.G. Harvey, A quantitative comparison of the mutagenicity of carcinogenic polycyclic hydrocarbon derivatives in cultured mammalian cells, Int. J. Cancer, 24 (1979) 203-209. N.E. Geacintov, A.G. Gagliano, V. Ibanez and R.G. Harvey, Spectroscopic characterizations and comparisons of the covalent adducts derived from the reactions of 7,8-dihydroxy7,8,9,10-tetrahydrobenzo[a]pyrene and 9,10,11,12-tetrahydrobenzo[elpyrene with DNA, Carcinogenesis, 3 (1982) 247-253. A.G. Gagiiano, N.E. Geacintov, V. Ibanez, R.G. Harvey and H.M. Lee, Application of fluorescence and linear dichroism techniques to the characterization of the covalent adducts derived from the interaction of ( +-)-trans-9,10-dihydroxy-anti-11,12-epoxy-9,10,11,12tetrahydrobenzo[a]pyrene with DNA, Carcinogenesis, 3 (1982) 969--976. N.E. Geacintov, Mechanisms of interaction of polycyclic aromatic diol epoxides with DNA and structures of the adducts, in: R.G. Harvey (Ed.), Polycyclic Hydrocarbons and Carcinogenesis, ACS Symposium Series No. 283, Washington, D.C., 1985, pp. 107--124. M.E. Hogan, N. Dattagupta and J.P. Whitlock, Jr., Carcinogen-induced alteration of DNA structure, J. Biol. Chem., 256 (1981) 4504--4513. D. Zinger, N.E. Geacintov and R.G. Harvey, Selective photodissociation of heterogeneous benzo[a]pyrene diol epoxide enantiomer-DNA adducts, Biophys. Chem., 27 (1987) 131-138. V. Kolubayev, H.C. Brenner and N.E. Geacintov, Stereoselective covalent binding of enantiomers of anti~benzo[a]pyrene diol epoxide to DNA as probed by optical detection of magnetic resonance, Biochemistry (U.S.A.), 26 (1987) 2638--2641. N.E. Geacintov, V. Ibanez, A.G. Gagiiano, S.A. Jacobs and R.G. Harvey, Stereoselective covalent binding of anti-benzo[a]pyrene diol epoxide to DNA. Conformation of enantiomer adducts, J. Biomol. Struct. Dyn., 1 (1984) 1473-1484. B. JernstrSm, P.O. Lyckseli, Gr~islund and B. Nord~n, Spectroscopic studies of DNA complexes found after reaction with anti-benzo[a]pyrene-7,8-dihydrodiol-9,10-oxide enantiomers of different carcinogenic potency, Carcinogenesis, 5 (1984) 1129-1135. M. Shahbaz, N.E. Geacintov and R.G. Harvey, Non-covalent intercalative complex formation and kinetic flow linear dichroism of racemic syn- and anti~benzo[a]pyrene diol epoxide-DNA solutions, Biochemistry (U.S.A.), 25 (1986) 3290--3296. O. Undeman, P.O. Lyckseli, A. Gr~islund, T. Astlund, A. Ehrenberg, B. Jerstrom, F. Tjerneld and B. Norden, Covalent complexes of DNA and two stereoisomers of benzo[a]pyrene-7,8-dihydrodiol-9,10-epoxide studied by fluorescence and linear dichroism, Cancer Res., 43 (1983) 1851-1860. M.-H. Kim, N.E. Geacintov, D.G. McQuillen, M. Pope, R.G. Harvey, Reaction mechanism of trans-9,10-dihydroxy-anti-7,8-epoxy-7,8,9,10-tetrahydro-3-methylcholanthrene with DNA in aqueous solutions. Conformation of adducts, Carcinogenesis, 7 (1986) 4 1 - 47. M.-H. Kim, C.J. Roche, N.E. Geacintov, M. Pope, J. Pataki and R.G. Harvey, Conformations of complexes derived from the interactions of two stereoisomeric bay-region 5methylchrysene diol epoxides with DNA, J. Biomol. Struct. Dyn., 3 (1986) 949--985. R.L. Chang, W. Levin, A.W. Wood, H. Yagi, M. Tada, K.P. Vyas, D.M. Jerina and A.H.

144

34

35 36 37

38 39

40

41

42

43 44 45

46

47

48

49 50 51

Conney, Tumorigenicity of enantiomers of chrysene 1,2-dihydrodiol and of the diastereomeric bay region chrysene 1,2-diol-3,4~poxides on mouse skin and in newborn mice, Cancer Res., 43 (1983) 192-196. P. LeBreton, The intercalation of bonzo[a]pyrene and 7,12-dimethyl benz[a]anthraeene metabolites and metabolite model compounds to DNA, in: R.G. Harvey (Ed.), Polyeyclic Hydrocarbons and Carcinogenesis ACS Symposium Series No. 283, The American Chemical Society, Washington, D.C., 1985, pp. 209--238. N.E. Geacintov, Is intercalation a critical factor in the covalent binding of mutagenie and tumorigenic polycyelic aromatic diol epoxides to DNA, Carcinogenesis, 7 (1986) 759-768. S.A. Jaeobs, C. Cortez and R.G. Harvey, Synthesis of potential proximate and ultimate carcinogenic metabolites of 3-methylcholanthrene, Carcinogenesis, 4 (1983) 519- 522. N.E. Geacintov, H. Yoshida, V. Ibanez and R.G. Harvey, Non-covalent binding of 7,8dihydroxy-9,10~poxy-tetrahydrobenzo[a]pyrene to deoxyribonucleic acid and its catalytic effect on the hydrolysis of the diol epoxide to tetraol, Biochemistry (U.S.A.), 21 (1982) 1864 - - 1869. A. Wada and S. Kozawa, Instrument for the differential flow diehroism for polymer solutions, J. Polym. Sci., 2 (1984) 853--864. B. Nord~n and F. Tjern61d, High sensitivity linear diehroism as a tool for equilibrium analysis in biochemistry. Stability constant of DNA~thidium bromide complex, Biophys. Chem., 4 (1976) 1 9 1 - 1 9 8 . N.E. Geacintov, H. Yoshida, V. Ibanez and R.G. Harvey, Non-covalent intercalative binding of 7,8-dihydroxy-9,10-epoxy benzo[a]pyrene to DNA, Biochem. Biophys. Res. Commun., 100 (1981) 1569-1577. J.M. Sayer, D.L. Whalen, S.L. Friedman, A. Park, H. Yagi, K.P. Vyas and D.M. Jerina, Conformational effects in the hydrolyses of benzo-ring diol epoxides that have bay region diol groups, J. Am. Chem. Soc., 106 (1984) 226-233. M.-H. Kim, N.E. Geadntov, M. Pope and R.G. Harvey, Structural effects in reactivity and adduct formation of polycyclic aromatic epoxides and diol epoxide derivatives with DNA: comparison between l~xiranylpyrene and benzo[a]pyrene diol epoxide, Biochemistry (U.S.A.), 23 (1984) 5433-5439. H.A. Benesi and J.H. Hildebrand, (1949). A spectrophotometrie investigation of the interaction of iodine with aromatic hydrocarbons, J. Am. Chem. Soc., 71 (1949) 2703--2707. H.B. Klevens and J.R. Platt, Spectral resemblance of catacondensed hydrocarbons, J. Chem. Phys., 17 (1949) 470--481. D.M. Jerina, J.M. Sayer, J. Yagi, M. Croisy-Delcey, Y. Ittah, D.M. Thakker, A.W. Wood, R.N. Chang, W. Levin and A.H. Conney, Highly tumorigenic diol epoxides from the weak carcinogen benzo[c~henanthrene in: Advances in Experimental Medicine and Biology: Biological Reactive Intermediates IIA, R. Snyder, D.V. Parke, J.J. Kocsis, D.J. Jollow, C.G. Gibson and C.M. Witmer (Eds.), Plenum Press, New York, 1982, pp. 501--523. G. Klopman, H. Grinberg and A.J. Hopfinger, MINDO/3 calculations of the conformation and carcinogenidty of epoxy-metabolites of aromatic hydrocarbons; 7,8-dihydroxy-9,10~)xy7,8,9,10-tetrahydrobonzo[a]pyrene, J. Theor. Biol., 79 (1979) 355--366. K.J. Miller, E~t. Taylor, J. Dommen and J.J. Burbaum, Stereoselectivity of benzo~a]pyrene diol epoxides by DNA for adduet formation with N2 on guanine, in: M. Rein (Ed.), Molecular Basis of Cancer, Part A, A.R. Liss, Inc., New York, 1985, pp. 187-197. N.E. Geaeintov, H. Yoshida, V. Ibanez, S.A. Jaeobs and R.G. Harvey, Conformations of adduets and kinetics of binding to DNA of the optically pure enantiomers of antibenzo[a]pyrene diol epoxide. Biochem. Biophys. Res. Commun., 122 (1984) 33-39. N.E. Geaeintov, H. Hibshoosh, V. Ibanez, M.J. Benjamin and R.G. Harvey, Mechanisms of reaction of benzo[a]pyrene-7,8-diol-9,10~poxide with DNA in aqueous solution. Biophys. Chem., 20 (1984) 121-133. F.A. Beland and R.G. Harvey, The isomeric 9,10~)xides of trans-7,8-dihydroxy-7,8dihydrobenzo[a]pyrene. J. Chem. Soc., Chem. Commun., (1976) 84-85. H. Yagi, D.R. Thakker, O. Hernandez, M. Koreeda and D.M. Jerina, Synthesis and reactions

145 of the highly mutagenic 7,8-diol-9,1O-epoxides of the carcinogen benzo[a]pyrene, J. Am. Chem. Soc., 99 (1977) 1604--1611. 52 H. Yagi, O. Hernandez and D.M. Jerina, Synthesis of (±)-7~,8a-dihydroxy-9~,10a-epoxy7,8,9,10-tetrahydrobenzo[a]pyrene, a potential metabolite of the carcinogen benzo[a]pyrene with stereochemistry related to the anti-leukemic triptolides, J. Am. Chem. Soc., 97 (1975) 6881 - 6883. 53 J.M. Sayer, H. Yagi, J.V. Silverton, S.L. Friedman, D.L. Whalen and D.M. Jerina, Conformational effects in the hydrolysis of rigid benzylic epoxides: implications for diol epoxides of polycyclic hydrocarbons, J. Am. Chem. Soc., 104 (1982) 1972-1978. 54 N.E. Geacintov, A.G. Gagliano, V. Ivanovic and I.B. Weinstein, Electric linear dichroism study of the orientation of benzo[a]pyrene-7,8-dihydrodiol-9,10~xide covalently bound to DNA, Biochemistry (U.S.A.), 17 (1978) 5256-5262. 55 B.E. Hingerty and S. Broyde, Carcinogen-base stacking and base-base stacking in dCpdG modified by (+) and ( - ) anti-BPDE, Biopolymers, 24 (1985) 2279-2299.