molecular structures of the chemical carcinogens 7 ... · molecular structures of the chemical...

[CANCER RESEARCH 36, 2428-2435, July 1976]

Molecular Structures of the Chemical Carcinogens 7- Chloromethylbenz[a]anthracene and 7-Chloromethyl-12- methylbenz[a]anthracene

Jenny P. Glusker, David E. Zacharias, and H. L. Carrell

The Institute for Cancer Research, The Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111

SUMMARY

The three-dimensional structures of two carcinogens, 7- chloromethyl-12-methylbenz[a]anthracene and 7-chloromethylbenz[a]anthracene, have been determined by X-ray crystal lographic techniques. Both compounds are carcinogenic and are believed to act by alkylating DNA. However, the first has a nonplanar ring system, whereas the second has a planar ring system. The nonplanarity of 7-chloromethyl-12-methylbenz[a]anthracene results from steric hindrance between a hydrogen atom of the 12-methyl group and a hydrogen atom on the [a] ring. This molecule cannot be made planar unless the 12-methyl group or the [a] ring is removed.

It is concluded that the carcinogenic activity of these compounds does not correlate with planarity of the ring system. This implies that, if DNA is the critical target of attack by these carcinogens, complete intercalation of the aromatic ring system of the carcinogen between the bases of DNA is not a likely mechanism of carcinogenic action in this system of compounds. The results presented here and those of others are more consistent with a model for a common interaction of the carcinogens 7-chloromethyl-12- methylbenz[a]anthracene and 7-chloromethylbenz[a]anthracene with DNA, in which they alkylate the bases of DNA and then lie with their long axes approximately parallel to the helix axis, probably in the major groove.

INTRODUCTION

An ultimate aim of those who study chemical carcinogenesis is to be able to describe, in 3-dimensional terms, the details of the interaction of the carcinogen with the critical target. In this way we will be able to characterize those features that make a chemical carcinogenic. There is not much detailed physical information available to date on the structures of chemical carcinogens or on the correlation of 3-dimensional structure with activity. Since the mode of interaction of these carcinogens with the critical target may depend on chemical reactivities at certain sites in the molecules, or on the geometrical configurations of the carcinogens, or both, it is important to study the 3-dimensional

' Research suppoted by Grants CA-10925, CA-06927, and RR-05539 from the NIH, USPHS; by Grant AG-370 from the National Science Foundation; and by an appropriation from the Commonwealth of Pennsylvania.

Received December 29, 1975; accepted March 19, 1976.

structures of such compounds in detail. We report here the crystal structures of 2 carcinogenic chloromethyl alkylating agents related to DMBA, 2 I. They are 7-CICH2-MBA, II, and 7- CICH_,-BA, III; the formulae are shown in Chart 1. The study was undertaken in order to investigate whether there is a conformational relationship among these 3 analogous carcinogenic compounds. Their 3-dimensional structures, so derived, were examined in the light of hypotheses that their carcinogenic activities result from alkylation of DNA, from intercalation in DNA, or from both types of effects.

The mechanism of chemical carcinogenesis by the polycyclic aromatic hydrocarbon DMBA (I) has been extensively investigated. Dipple et al. (7) suggested that a carbonium ion, Ar-CH._, ' , might be a reactive intermediate in such carcinogenesis. 7-Hydroxymethyl-12-methylbenz[a]anthracene, a metabolite of DMBA (4, 5, 10), was found to induce cancer in rats and mice. Because this metabolite might have re- sulted from a carbonium ion from DMBA, some bromo- and chloromethyl derivatives of benz[a]anthracene were prepared and studied (5, 8, 9) since, it was reasoned, these compounds might readily yield carbonium ions. The compounds, when prepared, were found to show carcinogenic (8, 9, 10, 22), antitumor (21), and mutagenic activities (18). They are inhibitors of RNA as well as of DNA synthesis (5), and their binding to DNA in v i t ro and in vivo has been investigated (2, 25, 27). The general result from these studies is that 7-BrCH~-BA is less carcinogenic than the compound with an additional 12-methyl group (8, 9), whereas both chloromethyl compounds (11 and III) are almost equally carcinogenic (22).

MATERIALS AND METHODS

Crystals of 7-CICH2-MBA (ll) 3 and 7-CICH2-BA (Ill) 4 were prepared by Dr. R. M. Peck of this Institute. The experimental X-ray work, with pertinent results, is reported in Table 1. Details wil l be published elsewhere 2.4 The structures were refined to conventional R values of 0.045 and 0.063 for II and III, respectively, for which R value = ~IIF,,I - IF,.II/~IF,,I

= The abbreviations used are: DMBA, 7,12-dimethylbenz[a]anthracene; 7- BrCH2-MBA, 7-bromomethyl-12-methylbenz[a]anthracene; 7-BrCH=-BA, 7- bromomethylbenz[a]anthracene; 7-CICH=-MBA, 7-chloromethyl-12-methylbenz[a]anthracene; 7-CICH2-BA, 7-chloromethylbenz[a]anthracene; ApU, adenylyl-3',5'-uridine; UpA, uridylyl-3',5'-adenine.

3 H. L. Carrell, manuscript in preparation. " D. E. Zacharias, manuscript in preparation.

2428 CANCER RESEARCH VOL. 36

Research. on April 23, 2020. © 1976 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

CH 3

Chloromethylbenz[a]anthracene Structures

CH2Ce CH2C~,

DMBA I 7- C~CH2-rvlBA E 7- C~.CH 2- 8A m

Chart 1. Formulae of compounds discussed or studied here. I, DMBA; II, 7-CICH2-MBA (studied here); III, 7-CICH2-BA (studied here).

where F,, is an observed structure factor and F,. is a calculated structure factor. All atoms, including hydrogen atoms, were located, and their positions were refined. We chose to work on the chloromethyl compounds because chlorine, with a lower atomic number than bromine, does not domi- nate the X-ray scattering to as great a degree, so that atomic positions, including those of hydrogen atoms, could therefore be determined with higher precision.

each bond in the compounds. These were derived assuming a ;-r bond order of zero for a bond of length 1.54 /~, (diamond), 0.667 for a bond of length 1.397 ,~ (benzene), and 1.00 for a bond of length 1.334 A (ethylene) (26). In both II and III, the highest 7T bond order is, as expected, that in the K-region.

Free valences or unsaturation indices for each atom in the molecules were then calculated from the formula:

RESULTS

Comparison of Bond Lengths and Bond Angles. A comparison of interatomic distances in molecules of 7-CICH.~- MBA (11), and 7-CICH,-BA (111), determined in these crystallographic studies, is given in Chart 2. Differences in bond lengths that are possibly significant (greater than 5o-, where ()- is the estimated standard deviation of a measurement) are indicated by heavy lines in this chart.

As shown in Table 1, II crystallizes with 2 nonidentically packed molecules (designated IIA and liB) in the crystal. Therefore, molecular parameters for each IIA and l ib are determined independently. The differences in bond lengths between Molecules IIA and liB all lie within 2.7o-, so that averaged values are given in Chart 2. The exception is 2 bonds in the [a] ring where each value is given in Chart 2 (marked by asterisks). This is a region in each molecule where the thermal motion is high (that is, much vibration is possible); therefore, positions of atoms are less accurately determined. It is also the area where Molecules IIA and liB interact with each other (see below).

Large differences in bond lengths between Molecules II (average) and III (Chart 2) occur mainly at the junction of the C and D rings, an area affected by distortion due to the 12- methyl group. The shortest bond, a double bond, lies in the K-region for each compound, i.e., the area in the aromatic molecule that behaves most like an olefinic double bond. However, it can be seen in Chart 2 that bond lengths around the K-region are similar in both II and III, so that 12-methyla- tion does not have a significant effect on the K-region.

Since accurate bond lengths, which are dependent on the multiplicit ies of the bonds, have been determined for both Molecules II and III, an attempt was then made to describe the areas of highest ~--electron density and, hence, of reactivity. This was done by computing "~- bond orders" (26) for

[1.732 - Y, (~- bond orders) = free valence]

as described by Pullman and Pullman (26). The summation (E) was made over all bonds connected to the atom for which the free valence is being determined. Values of free valences are very sensitive to the curve chosen to describe bond order. Such curves vary principally in the value chosen for a "single bond" in an aromatic ring system [that for diamond (1.54 A) was used here]. In general, atoms in such conjugated systems with high free valences would be expected to be most susceptible to attack [electrophil ic, nu- cleophil ic, or free radical (26)].

Values for free valences calculated in this way are listed in Table 2. As shown in this table by asterisks, the change in free valence between Molecules II and III occurs mainly in the area of nonplanarity (Positions a to j).

Newman et al. (19, 20), who have studied this system extensively, suggest that the reactive intermediate in the case of DMBA may be a compound with a keto group at position 5 and a methylene group at position 4, possibly formed from an epoxide intermediate. In each of the compounds I, II, and III, it can be seen in Chart 3 that the free valence at Position 5 is less than that at Position 4. This suggests that all 3 compounds may behave in a similar manner on reaction in this area of the molecule. In all 3 compounds (I, II, and III) the highest free valence is at Position 12.

It is concluded from a study of interatomic distances in II and III that the K regions of these 2 compounds are similar and that the major differences occur near the junct ion of Rings C and D (Chart 4). This is the area that might be expected to be most affected by the overcrowding caused by the 12-methyl group in II.

Planarity of the Ring Systems. The aromatic ring system of II, like that of DMBA (I) (12, 29), and the K-region oxide of DMBA (11), is markedly nonplanar. There are angles of 21 ~

JULY 1976 2429



J. P. G lusker et al.

Table 1 Crystal and activity data on II and III

7-CICH.~-MBA (11) 7-CICH.~-BA (111)

Formula Crystal system Space group Cell dimensions

a b c # V

Cell contents

Observed density Calculated density No. of independent intensity

data collected No. of reflections below

threshhold value Radiation used Structure solution Final R value All hydrogen atoms located

and refined Estimated standard devia-

tions in C ~ or C---C bond lengths

Angle between outermost rings

Relative biological activity of the corresponding bromo compounds (8, 9)

Conclusions by Rayman and Dipple (27) for the bromo compounds

C2,,H,sCI C,,H,:~CI Monoclinic Monoclinic P2,/c P2,/n

20.457 (4) ~, 14.381 (7)- ,~ 11.474 (2) 4.228 (2) 13.034 (2) 22.141 (7)

108.91 (1)o 95.87 (5) ~ 2894 (1) 1339 (2) A:' 8 molecules (2 per 4 molecules (1 per

asymmetrical asymmetric unit) unit)

1.33 g cm -:~ 1.37 g cm -:~ 1.34 g cm " 1.38 g cm -:~

5378 2490

1907 773

CuKo~ (1.5418 A) CuKo~ (1.5418 A) Direct methods Direct methods 0.045 0.063 Yes Yes

0.004 A 0.005 A

20.8, 21.8 ~ 1.6 ~

More carcinogenic Less carcinogenic

Products with DNA Products with DNA more light sensi- less light sensitive. tive. Higher spec- Lower ratio ade- ificity for attack nine:guanine sub- of adenine than stituted in DNA guanine residues in DNA

" Values in parentheses are estimated standard deviations with respect to the last digit given.

24 ~ and 35 ~ respectively, between the 2 outermost rings (Chart 4, A and D), of these 3 compounds (Table 3). In contrast, the ring system of III is a lmost completely planar (excluding the --CH2CI group).

The nonplanar i ty of the ring system of II is an inherent property of the molecule that cannot be changed (unless the 12-methyl group or the angular ring is removed), and it is found in other 12-methylbenz[a]anthracene derivatives (13). The nonplanar i ty results f rom steric hindrance between a hydrogen atom of the 12-methyl group and a hydrogen atom on the angular ring. This is i l lustrated diagram- mat ical ly in Chart 3. The nonplanar i ty in DMBA (I), reported in its crystal structure (12, 29), results from an identical interact ion. Nonplanar i ty in I, II, and the oxide of DMBA confers opt ical activity on these compounds (since mirror images of the molecules cannot be super imposed on the or ig inal molecule), and it is possible that one opt ical isomer of a given compound may be b io log ica l ly more active than the other.

If the 2 compounds II and III have s imi lar conformat ions of their ring systems when they interact wi th a biological mac- romolecule, this conformat ion must be nonplanar. This is

possible since III can buckle, whereas it is not possible for II to have a flat ring system.

Packing in the Unit Cell. A str iking di f ference in the structures of these compounds is that III crystal l izes with only 1 molecule in the asymmetr ic unit, whereas II crystallizes with 2 nonident ical ly packed molecules (designated IIA and liB) in the asymmetr ic unit. There seems to be no specif ical ly strong interact ion between these 2 nonidentically packed molecules in the overall crystal unit cell. Ex- c luding interact ions involving chlor ine atoms, the closest interact ions between Molecules IIA and l ib are van der Waals-type contacts between the [a] rings shown in Chart 4. This is a "her r ing-bone"- type packing. That is, the aromatic ring systems do not lie in planes parallel to each other, but incl ined to each other. As a result one (or more) hydrogen atom(s) of one molecule interact with the ~--electron system of the other molecule. This is character ist ic of many aromatic compounds, e.g., benzene.

The tendency for carcinogens to crystal l ize with more than 1 molecule in the asymmetr ic unit has been noted before (17). It may reflect the awkward shapes of such molecules, making eff ic ient packing possible only with 2




independent units. Alternatively, the interaction of a hydrogen atom of one molecule with the =-electron system of another may be a predominant ly stable interaction. Mason (17) called the number of molecules in the asymmetr ic unit an "aggregat ion factor" which he attempted to correlate with carc inogenic activity. However, no direct correlation has been found.

The molecules can also pack in planes parallel to each other. The stacking perpendicular to the general direct ion of the planes of the ring systems is i l lustrated for each compound in Chart 5. The areas of stacking involve only the more planar por t ions of the molecules. Molecules of II (Chart 5a) pack as head-to-tail "d imers" between like molecules, i.e., IIA packs with another IIA molecule, and liB packs with another liB molecule. As noted earlier, IIA packs with liB in a "her r ing-bone" fashion. The more planar por- tion of II (on the side of chloromethyl subst i tut ion) overlaps sl ightly (as shown by the shading) in planes separated by 3.55/~. The chlor ine atoms point away from the center of the dimer. The area of maximum nonplanari ty does not take part in the overlap of the dimers. Molecules of III (Chart 5b) pack in a stepwise manner throughout the crystal in planes 3.55 ,~ apart so that both sides of this flatter molecule are involved in stacking.

In both types of molecules the K-region double bond of one molecule lies above a nonlocal ized ring bond (i.e., a more aromatic type of bond) of another molecule. The area of the molecules involved in vertical stacking (shaded in Chart 5, a and b) is less than that found in crystals of the intercalator 9-aminoacr id ine (Chart 5c) (31) as well as in crystals of the complex-o f 9-aminoacr idine with ApU (a model for intercalat ion of a polycycl ic melecule in DNA (30) (Chart 5d). It can be seen in Chart 5 that, in its interact ion

t.374 12 J~ ~ ) J "

1.511

1.360 ( ~ 1.434 1.419 ( ~ 1.400 1.476

11,4 , B I1,448 A I,.425 1.439

L404 D i! 1:422 . , , j :

1.416

~.4~5 II 1.434

~.378(,r 1, .~ 1.35,~5)J 1.359

1.350 ~ 1.427 1.420 ~11.5011.4941"391 1.450

1.336 1.333 K-REGK)N

t.804(2) 1.811(Z]

Chart 2. Bond lengths (~,). The upper value is for II (i.e., the average for IIA and liB, except where the bond is marked with an asterisk when both values are given, IIA above liB). The lower value is for III. Estimated standard deviations are 0.003 to 0.004 A, except where they are shown in parentheses with respect to the last digit in the bond length. Bond length differences of less than 0.01 ,~ are probably not significant. Changes greater than 0.015 A are indicated by heavy lines for the bond.

Ch lo romethy lbenz [a ]an th racene St ruc tu res

Table 2 Free valences calculated from experimental bond distances

Atomic number- Av. (7-CICH2- ing (see Av. 7-CICH_,- MBA)-(7- diagram 7-CICH2-BA MBA (IIA, CICH2-BA) (ll- below) (111) liB) DMBA (I) III)

a = 10 0.06 0.15 0.33 0.09" b = 11 0.25 0.39 0.36 0.14" c -0.05 0.10 0.14 0.15" d = 12 0.33 0.39 0.39 0.06 e 0.17 0.36 0.32 0.19" f 0.08 0.22 0.21 0.14" g = 1 0.28 0.37 0.32 0.09" h = 2 0.28 0.18 0.38 -0.10" i = 3 0.18 0.15 0.28 -0.03 j = 4 0.22 0.36 0.31 0.14" k 0.02 -0.01 0.02 -0.03 I = 5 0.23 0.22 0.19 -0.01 m = 6 0.29 0.28 0.34 -0.01 n 0.13 0.17 0.19 0.04 o = 7 0.27 0.26 0.34 -0.01 p 0.08 0.05 0.14 - 0.03 q = 8 0.32 0.32 0.35 0.00 r = 9 0.13 0.11 0.23 -0.02

" Large differences.

b

h

g i

d f . J

m

~ c t

with ApU, there is much overlap of 9-aminoacr id ine with the "bases and with the hydrogen bond system between the bases.

Both molecules, then, show some tendency to stack in parallel planes - 3 . 5 A apart, al though only the planar por- t ions of the molecules are involved in this stacking. How- ever, the extent of stacking is less than that found for some known intercalators. ~

Conformations of Molecules. The shapes of molecules of II and of III are shown in Chart 6. These are views (1, 3) in direct ions perpendicular to and parallel to the best plane through the [a] ring (D ring, chosen arbitrari ly). Spheres with van der Waals radii for each atom are stippled over the bond diagrams. Views onto the molecular planes are given in Chart 6, a and b. These indicate the large hydrophobic area that is expected to lie in the conf ines of the bases of the DNA helix if intercalat ion occurs (6). Side views of Molecule II in Chart 6, c and e, show the extent of buckl ing of the ring systems compared with Molecule III in Chart 6, d and f. It is shown in Chart 6e that the buckl ing of molecules of II causes them to be wedge shaped (also shown in Chart 6c). The vertical lines in Chart 6, e and f, indicate the width of the space that would be available for intercalat ion between the bases of DNA.

JULY 1976 2431



J. P. G l u s k e r et al.

[a]- r ing .... . .......... ~.II..

::--~- ..... :z C H

12-methyl, group

II III

Chart 3. Diagram illustrating the steric hindrance between the 12-methyl group and the hydrogen atom of the [a] ring in II. This steric hindrance does not occur for Ill.

Table 3 Angles between planes through individual r ing systems

C

3. ,,3.6o C~/ ~J-

x

"i MoLecute [I B (a) (b)

Molecule E A

Chart 4. The only interactions between molecules IIA and liB. These involve packing of the D rings of each molecule against each other.

ThOs, molecu les of II could not slip easily between base pairs in DNA w i thou t causing cons iderab le d is tor t ion of the helix axis. No such problem exists for molecules of III. This is i l lustrated d iagrammat ica l l y in Chart 7a where it is shown that, it they are inserted between the bases of DNA, II and III have d i f ferent effects on the d i rect ion of the DNA helix axis.

D I S C U S S I O N

It is not known whether the cr i t ical target in carc inogene- sis is a prote in , a nucleic acid, or another type of molecule. However, since II and III are both mutagens and carc inogens, it is of interest to cons ider the effects of the interac- t ion of these c o m p o u n d s with DNA.

I m p o r t a n c e of P lanar i ty in C a r c i n o g e n e s i s in B e n z [ a ] a n t h r a c e n e Der ivat ives. Of the 2 carc inogens stud-

I (a) X = H, Y = CH3 DMBA II (a) X = CI, Y = CH3 7-CICH2-MBA III (a) X = CI, Y = H 7-CICH2-BA IV (b) K-region oxide of DMBA

(Refs. 12 and 28) (this work) (this work) (Ref. 11)

I IIA liB III IV A:B 5.0 3.8 3.9 2.1 6.4 A:C 13.2 11.3 12.3 2.0 20.9 A:D 24.0 20.8 21.8 1.6 34.8 B:C 10.8 8.1 9.0 0.7 15.1 B:D 21.2 17.5 18.4 1.0 28.6 C:D 10.8 9.6 9.6 0.5 14.2

ied here, one has a planar ring system whereas the other does not. In the b romomethy lbenz [a ]an th racenes , which presumably have s imi lar structures, the repor ted ly more potent species has a markedly nonplanar r ing system. We conc lude that p lanar i ty of the aromat ic system in halo- methy lbenz[a ]anthracenes is not a requ i rement for carc inogenic i ty. Further, r ing system planari ty may not be important in an t i tumor or mutagenic activity, s ince both chloro- methy lbenz[a ]anthracenes (11 and III) are active.

I m p o r t a n c e of In terca la t ion in C a r c i n o g e n e s i s by B e n z [ a ] a n t h r a c e n e Der ivat ives. Intercalat ion may be de- f ined as the comple te enve lopment of a f lat hydrophob ic molecu le wi th in the hydrophob ic conf ines of a DNA helix, w i thout d isp lacement or d is tor t ion of the hel ix d i rect ion (6). Detailed studies (15) of the binding of benzo(a)pyrene to DNA have led to a cor robora t ion of the in tercalat ion model , and they showed that the more compac t ring systems bind more strongly. Craig and Isenberg (6) showed that, if a polycyl ic aromat ic hydrocarbon was too large in area to lie wi th in the bases of DNA, protected f rom the aqueous env i ronment , b ind ing and, hence, intercalat ion, could n o t occur. From structural studies on intercalat ion, in the cases of a 9- aminoacr id ine :ApU complex (30) and of an e th id ium:5- iodo




UpA complex (32), it is known that the intercalat ing agent must be reasonably flat. We therefore must conclude that II (7-CICH=-MBA) cannot intercalate well in DNA. The situa- t ions for insert ion of the 2 compounds (11 and III) between the bases of DNA are shown in Chart 7a. The 2 compounds have different effects.

Since both Compounds II and III are carc inogenic and yet one of them, II, cannot intercalate well in DNA, we conclude that these compounds do not exert their carc inogenic activity simply by intercalat ion between the bases of DNA. This must be true unless there is conversion of II in vivo to a

(a)

(c)

~~~--'~~ (b)

Chart 5. Views of molecules perpendicular to the plane through the ring system, showing, boy shading, the overlap with molecules in parallel planes approximately 3.5 Aapart. a, Molecule II, "head-to-tail" dimers. The upper molecule is drawn with heavy lines. The chlorine atom is indicated by an open circle, b, Molecule III, "head-to-head" stacking throughout the crystal. The central molecule in this group of 3 molecules is drawn with heavy lines. c, 9-aminoacridine (31), showing extensive overlap for this intercalator. The central molecule is drawn with heavy lines. Nitrogen atoms are indicated by open circles, d, 9-aminoacridine-ApU complex (30), showing the intercalation of 9-aminoacridi ne between the bases. Note how the polycyclic molecule lies between hydrogen-bonded base pairs.

(a) (b)

(e) (f)

Chart 6. Views of the molecules with respect to the best plane through Ring D. These diagrams were drawn with the program STIPL (1). Rings A, B, C, and D are marked. The thickness of an aromatic ring system is shown (e and f). Molecule II (7-CICH~-MBA) is shown in a, c, and e, and an analogous view of Molecule III (7-CICH2-BA) is shown in b, d, and f, respectively.

C h l o r o m e t h y l b e n z [ a ] a n t h r a c e n e S t ruc tu res

compound lacking either the 12-methyl group or the [a] ring so that the ring system becomes planar. Since DMBA is so much more active than the compound lacking a 12-methyl group and since anthracene derivatives are less carcinogenic than benz[a]anthracene derivatives, we conc lude that this metabol ic demethylat ion or ring opening is unlikely. Otherwise, the simpler compound might be expected to be more active, whereas, in fact, the reverse is the case.

It is not impl ied that intercalat ion of, III, for example, does not occur but simply that this form of interaction is not important in carcinogenesis. An interact ion of a carc inogen with a protein receptor involved in some way with DNA [such as a steroid receptor (16) or a polymerase] probably has a much lower requirement for planarity of the ring system than does intercalat ion in a double helical nucleic acid.

Importance of Alkylation of DNA by Benz[a]anthracene Derivatives. The compounds that have been studied here are alkylating agents, and it is interesting to consider the consequences of the alkylat ion of DNA by II and III, since these compounds have also been shown to be mutagens, in which case they almost certainly interact in some way with DNA. It has been shown that amino groups on the bases of DNA are major sites of attack (14, 27, 28). The C - C I bond of the alkylat ing agents lies approx imate ly at an angle of 109.5 ~ to the plane of the ring system, and there is no f lexibi l i ty in this side chain except for rotat ion about the carbon-carbon bond adjacent to the carbon-chlor ine bond (see Charts 3 and 6, e and f). In this model we assume that epoxidizat ion is un important and that the simple halomethyl group alkylates DNA. When such alkylat ion occurs the chlorine atom is replaced by DNA. As a result, as shown in Chart 7b, the covalent ly bound carc inogen cannot lie in the plane of the bases (for steric reasons) but must also lie at an angle of approximately 109.5 ~ to the planes of the bases.

Studies with an endonuclease II f rom Escher i ch ia co i l (14) showed that this enzyme excises N6-(12-methylbenz[a]an - thracenyl-7-methyl)adenine and N2-(12-methylbenz[a]anthracenyl-7-methyl)guanine, part icular ly the adenine derivative, when DNA is alkylated with 7-BrCH:-MBA. Fluorescence studies of the interact ion of 7-BrCH2-BA with DNA (23, 24) have indicated that alkylat ion of the bases of DNA occurs wi thout distort ion or denaturat ion of the sec- ondary structure of DNA. This suggests that a base dis- placement mechanism, postulated for the interact ion with DNA with certain carcinogens (35), does not occur in the benz[a]anthracene system. The red shift in the absorpt ion spectrum on interaction of the carc inogen with DNA was believed to indicate an arrangement of aromatic rings parallel to the helix axis, i .e., perpendicular to the base pairs. This model is in agreement with our reasoning above.

General Conclusion. Our structural data indicate that intercalation of the 2 compounds between the bases of DNA is not a l ikely mechanism of carc inogenic action. Since alkylation and intercalation are not s imul taneously possible for steric reaons and since partial insertion of the carcinogens between the bases of DNA will cause a large hydrophobic area of the carcinogen to protrude beyond the con- fines of the double helix into the aqueous env i ronment [which Craig and Isenberg thought unl ikely (6)], we conclude that the most plausible common mechanism for ac-

JULY 1976 2433



J. P. G l u s k e r e t a l .

]I rrr (a) (b)

b a s e ~ base

base ~ Ct base ~ CE

base ~ / base

]I ]Tr Ct Ct

//+ HN, base / + H N ~ b a s e

- FtCf., I -HCt

~ N - - base ~ - - N base

(c)

Chart 7. Geometrical argument against intercalation as a mode of carcinogenesis in this system of benz[a]anthracene derivatives. The interaction of II (7- CICH2-MBA, left side for a and b and III (7-CICH=-BA, right side for a and b) with the bases of DNA. The bases, viewed edge-on, are drawn assing/e lines, a, effects of the molecules lying between the bases of DNA. Note the different effects on the directions of the bases for the 2 compounds, i.e., for II the helix direction is changed. The intercalated compound is not in a good position to alkylate as well. b, effects of alkylation of the DNA bases. The aromatic group projects out from the bases of DNA but is covalently linked to 1 base in each case. The shapes of the projecting groups differ, i.e., one is wedge-shaped whereas the other is planar, c, diagram of the carcinogen lying in the major groove of DNA after alkylation of a base of DNA.

t ion of the 2 c o m p o u n d s on DNA is a l ky la t i on of the bases . As a resu l t , the c a r c i n o g e n wi l l l ie in a g roove of DNA,

p r e s u m a b l y the m a j o r g roove , as s h o w n d i a g r a m m a t i c a l l y in Cha r t 7, b and c. Th is mode l , a l t h o u g h re levant to an

i n t e r a c t i o n w i th DNA and , poss ib l y , re levant to m u t a g e n e - sis, is, of c o u r s e , re levant to c a r c i n o g e n e s i s on ly if it can be

s h o w n tha t DNA is the c r i t i ca l t a rge t of a t tack of the c a r c i n o - gen . S im i l a r m o d e l s are be ing c o n s i d e r e d fo r the in te rac -

t i on of s t e r o i d s w i th DNA (33, 34). In the mode l j us t de- s c r i b e d ( w h i c h is s im i l a r to tha t p r o p o s e d by P o c h o n e t a l .

(23, 24), p lanar i t y of the a r o m a t i c r ing sys tem is no t essen - t ia l , a l t h o u g h the p o s i t i o n i n g of the 12-methy l g r o u p may be

i m p o r t a n t . The c a r c i n o g e n , in th is mode l , is i nc l i ned at an

a n g l e to the bases of DNA, bu t it does no t p e r t u r b the

d o u b l e he l ica l DNA s t r u c t u r e . It may, howeve r , i n te r fe re s u b s t a n t i a l l y w i t h s o m e b a s e - r e c o g n i t i o n sys tem.

A C K N O W L E D G M E N T S

We particularly wish to thank Dr. Richard M. Peck of this Institute for providing us with the crystals, and we thank Dr. Richard M. Peck, Dr. Sam Sorof, and Dr. Ronald Harvey for many helpful discussions.

R E F E R E N C E S

1. Andrews, L. C., and Carrell, H. L. Program STIPL, Modified. Part of the CRYSNET System (1974).

2. Baird, W. M., Dipple, A., Grover, P. L., Sims, P., and Brookes, P. Studies on the Formation of Hydrocarbon-Deoxyribonucleoside Products by the Binding of Derivatives of 7-Methylbenz[a]anthracene to DNA in Aqueous

Solution and in Mouse Embryo Cells in Culture. Cancer Res., 33: 2386- 2392, 1973.

3. Bernstein, H. J., Andrews, L. C., Berman, H. M., Bernstein, F. C., Campbell, G. H., Carrell, H. L., Chiang, H. B., Hamilton, W. C., Jones, D. D., Klunk, D.. Koetzle. T. F., Meyer, E. F., Morimoto, C. N., Sevian, S. S., Stodola, R. K., Strongson, M. M., and Willoughby, T. V. Crysnet-A Network of Intelligent Remote Graphics Terminals. In: Second Annual AEC Scientific Computer Information Exchange Meetings, Proceedings of the Technical Program, pp. 148-158. Report BNL 18803. Upton, Long Island, N. Y.: Brookhaven National Laboratory, 1974.

4. Boyland, E., Sims, P., and Higgins, C. Induction of Adrenal Damage and Cancer with Metabolites of 7,12-Dimethylbenz(a)anthracene. Nature, 207: 816-817, 1965.

5. Buu-Hoi, N. P., Croisy-Delcey, M., Jacquignon, P., Chayet, C., Daudel, P., Gachelin, F., Jacquier, A., Lataillade, G., and Moreau, J. Action of Halomethyl Hydrocarbon Derivatives on in Vitro DNA and RNA Synthesis, and on tRNA Properties. Relationship with Carcinogenic Potency of Methyl Parent Hydrocarbons. In: P. O. P. Ts'o and J. A. DiPaolo (eds.), Chemical Carcinogenesis, Part A, Chap. 5, pp. 159-175. New York: Marcel Dekker, Inc., 1974.

6. Craig, M., and Isenberg, I. Testing of a Size Criterion for DNA-Hydrocar- bon Binding_ Biopolymers, 9: 689-696, 1970.

7. Dipple, A., Lawley, P. D., and Brookes, P. Theory of Tumour Initiation by Chemical Carcinogens: Dependence of Activity on Structure of Ultimate Carcinogen. European J. Cancer, 4: 493-506, 1968.

8. Dipple, A., and Slade, T. A. Structure and Activity in Chemical Carcino- genesis: Reactivity and Carcinogenicity of 7-Bromomethylbenz[a]anthracene and 7-Bromomethyl-12-methylbenz[a]anthracene. European J. Cancer, 6: 417-423, 1970.

9. Dipple, A., and Slade, T. A. Structure and Activity in Chemical Carcino- genesis. Studies of Variously Substituted 7-Bromomethylbenz[a]anthracenes. European J. Cancer, 7. 473-476, 1971.

10. Flesher, J. W., and Sydnor, K. L. Carcinogenicity of Derivatives of 7,12- Dimethylbenz(a)anthracene. Cancer Res., 31: 1951-1954, 1971.

11. Glusker, J. P., Carrell, H. L., Zacharias, D. E., and Harvey, R. J. Crystallo- graphic Studies of K-Region Arene Oxides: 7,12-Dimethyl- benz[a]anthracene-5,6-oxide and Phenanthrene-9,1O-oxide. Cancer Bio- chem. Biophys., 1: 43-52, 1974.

12. Iball, J. A Refinement of the Structure of 9:10-Dimethyl-1:2-benzanthra-

2 4 3 4 CANCER RESEARCH VOL. 36



cene. Nature, 201: 916-917, 1964. 13. Jones, D. W., and Sowden, J. M. Overcrowding in Structures of

Benz[a]anthracenes Methyl-substituted at 1 and 12 Positions. Acta Cryst., (A)31: $126, 1975.

14. Kirtikar, D. M., Dipple, A., and Goldthwait, D. A. Endonuclease II of Escherichia co~i: DNA Reacted with 7-Bromomethyl-12-methylbenz[a]anthracene as a Substrate. Biochemistry, 14: 5548-5553, 1975.

15. Lesko, S. A., Jr., Smith, A., Ts'o, P. O. P., and Umans, R. S. Interaction of Nucleic Acids. IV. The Physical Binding of 3,4-Benzpyrene to Nucleo- sides, Nucleotides, Nucleic Acids, and Nucleoprotein. Biochemistry, 7: 434-447, 1968.

16. Litwack, G., Morey, K. S., and Ketterer, B. Corticosteroid Binding Pro- teins of Liver Cytosol and Interactions with Carcinogens. In: B. R. Rabin and R. B. Freeman (eds.), A Symposium on Effects of Drugs on Cellular Control Mechanisms, pp. 105-130. Baltimore: University Park Press, 1972.

17. Mason, R. Structure and Carcinogenic Acitvity of the his-Cyclic Dibenz- acridine and Methyl Benzanthracenes. Naturwissenschaften, 11: 252, 1956.

18. McCann, J., Choi, E., Yamasaki, E., and Ames, B. N. The Detection of Carcinogens as Mutagens in the Salmonella/Microsome Mutagenicity Test. Proc. Natl. Acad. Sci. U. S., 72: 5135-5139, 1975.

19. Newman, M. S., and Cunico, R. F. The Synthesis of 5-Hydroxymethyl-, 5- Acetoxymethyl-, and 5-Methylmercapto-7,12-dimethylbenz[a]anthracene, and of 5,7,12-trimethylbenz[a]anthracene. J. Med. Chem., 15: 323- 325, 1972.

20. Newman, M. S., and Olson, D. R. A New Hypothesis Concerning the Reactive Species in Carcinogenesis by 7,12-Dimethylbenz[a]anthracene. The 5-Hyd roxy-7,12-d imethylbenz[a]anthracene-7,12-Dimet hylbenz-[a]<z- anthracen-5(6H)-one Equilibrium. J. Am. Chem. Soc., 96: 6207-6208, 1974.

21. Peck, R. M., and O'Connell, A. P. Mixed Bifunctionality. IV. Antitumor Activity of Alkylating Derivatives of Polycyclic Aromatic Hydrocarbons as a Function of Structure and of Vehicle. J. Med. Chem., 15: 68-70, 1972.

22. Peck, R. M., Tan, T. K. and Peck, E. B. Pulmonary Carcinogenesis by Derivatives of Polynuclear Aromatic Alkylating Agents. Cancer Res., 36: 2423-2427, 1976.

23. Pochon, F., Brookes, P., and Michelson, A. M. Action of the Carcinogen 7-Bromomethylbenz[a]anthracene on DNA. European J. Biochem., 21: 154-160, 1971.

C h l o r o m e t h y l b e n z [ a ] a n t h r a c e n e S t r u c t u r e s

24. Pochon, F., and Michelson, M. A. Action of the Carcinogen 7-Bromo- methylbenz[a]anthracene on Synthetic Polynucleotides. European J. Biochem., 21: 144-153, 1971.

25. Pullman, B. Summary of the Chemical Aspects of Carcinogenesis. In: P. O. P. Ts'o and J. A. DiPaolo (eds.), Chemical Carcinogenesis, Part B, Chap. 41, pp. 713-727. New York: Marcel Dekker, Inc., 1974.

26. Pullman, B., and Pullman, A. Quantum Biochemistry, Chap. 3, p. 84; Chap. 4, pp. 155-181. New York: Interscience Publishers, 1963.

27. Rayman, M. P., and Dipple, A. Structure and Activity in Chemical Carci- nogenesis. Comparison of the Reactions of 7-Bromomethyl- benz[a]anthracene and 7-Bromomethyl-12-methylbenz[a]anthracene with Deoxyribonucleic Acid in Vitro. Biochemistry, 12: 1202-1207, 1973.

28. Rayman, M. P., and Dipple, A. Structure and Activity in Chemical Carci- nogenesis. Comparison of the Reactions of 7-Bromomethyl- benz[a]anthracene and 7-Bromomethyl-12-methylbenz[a]anthracene with Mouse Skin Deoxyribonucleic Acid in Vivo. Biochemistry, 12: 1538- 1542, 1973.

29. Sayre, D., and Friedlander, P. H. Crystal Structure of 9:10-Dimethyl-1:2- benzanthracene. Nature, 187: 139-140, 1960.

30. Seeman, N. C., Day, R. O., and Rich, A. The Molecular Structure of an Intercalative Complex: Adenylyl-3',5'-uridine Plus 9-Aminoacridine. Na- ture, 253: 324-326, 1975.

31. Talacki, R., Carrelt, H. L., and Glusker, J. P. 9-Aminoacridine Hydrochlo- ride Monohydrate. Acta Cryst., (B) 30: 1044-1047, 1974.

32. Tsai, C-C, Jain, S. C., and Sobell, H. M. X-ray Crystallographic Visualiza- tion of Drug-Nucleic Acid Intercalative Binding: Structure of an Ethid- ium-Dinucleoside Monophosphate Crystalline Complex, Ethidium:5-10- douridylyl(3'-5')adenosine. Proc. Natl. Acad. Sci. U. S., 72: 628-632, 1975.

33. Weeks, C. M., Erman,M. G., and Duax, W. L. Deoxycorticosterone Ade- nine Monohydrate, C2~H3oO.~'CsHsNs"H20. Cryst. Struct. Commun., 4: 637-640, 1975.

34. Weeks, C. M., Rohrer, D. C., and Duax, W. L. Deoxycorticosterone- Adenine Interactions in a Crystalline Complex. Science, 190: 1096-1097, 1975.

35. Weinstein, I. B., Grunberger, D., and Blobstein, S. Conformational and Functional Changes in Nucleic Acids Induced by Chemical Carcinogene- sis. In: P. Bucalossi, U. Veronesi, and N. Cascinelli (eds.), Proceedings of the 1 lth International Cancer Congress, Florence Italy, October 20 to 26, 1974, Vol. 2, pp. 47-53. Amsterdam: Excerpta Medica, 1975.

JULY 1976 2 4 3 5



1976;36:2428-2435. Cancer Res Jenny P. Glusker, David E. Zacharias and H. L. Carrell

]anthracenea7-Chloromethyl-12-methylbenz[]anthracene anda7-Chloromethylbenz[

Molecular Structures of the Chemical Carcinogens

Updated version

http://cancerres.aacrjournals.org/content/36/7_Part_1/2428

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/36/7_Part_1/2428To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



molecular structures of the chemical carcinogens 7 ... · molecular structures of the chemical...

Documents