Document Gmy6r2p42B9a395GMEqdekjbN

7-7#, INTERNATIONA!. JOURNAL OP QUANTUM CHEMISTRY, VOL. XXV, 49).j02 (19*4) Some Reactive Properties of Chlorooxirane, a Likely Carcinogenic Metabolite of Vinyl Chloride PATRICIA R. LAURENCE and PETER POLITZER Department of Chemistry, University of New Orleans, New Orleans, Louisiana 70148. U.S.A. . Abstract We have carried out a computational study of the reactive properties of chlorooxirane. the metabolicany produced epoxide of vinyl chloride that is believed to be a direct-acting carcinogenic form of this molecule. An ab Initio scf-mo procedure (GAUSSIAN 70) was used to compute the energy requirements for stretching the C--Cl and both C--O bonds (SN1 reactivity) and to determine the course of the epoxide's possible SN2 reactions with ammonia, taken as a model for nucleophilic sites on DNA- The epoxide was assumed to be protonated; both the oxygen* and chloro-protonated forms were considered. At each step along the various reaction pathways, the structure of tbe system was reoptimized. For the oxygen-protonaled epoxide, the Cj--O bond has a significantly lower energy barrier to stretching than does the C3--O. (The carbon bearing the chlorine is designated Ci'.) However, both are very much higher than that of the C--Cl bond in the ch)oro*protonated form, confirming our earlier finding of the relative weakness of this bond. In the S^2 processes involving ammonia, intermedialc complexes are formed with both carbons of the oxygen-protonated epoxide, the C2-complex being the more stable. However, the most stable ammonia complex occurs pi the chloro-protonated epoxide. Our calculated results, both the energies and also the geometry change*, allow ni to propose two possible mechanisms for the formation of the 7-N-(2vxoethyi) derivative of guanine that has been observed io be the major In vivo DNA alkylation product of vinyl chloride and has been suggested as possibly being responsible for its carcinogenicity. One of these mechanisms is $4 and starts with the chloro-protonated epoxide; the other is SN2 and involves the oxygen-protonaied form. 1, Introduction Numerous studies on animals, cel! cultures, and bacteria have shown vinyl chloride (I) to be both mutagenic and carcinogenic in these systems [1-5]. In addition, an increased incidence of certain types of cancer has been observed among industrial workers who are exposed to vinyl chloride [6, 7], Experimental evidence indicates that the first step in the mutagenic and carcinogenic action of vinyl chloride is its metabolic conversion to the epoxide, a process that is catalyzed by the mixed function oxidase system [2, 8-11]: Cl H \=cy h" vh I toi a-?--C~H HH U This epoxide, chlorooxirane (II), is a strong mutagen [2,10,12], and is regarded as very likely to be a direct-acting carcinogen, an "ultimate" carcinogenic form of vinyl chloride [8,10-13], 1964 John Wiley and Sons, Inc. CCC 0020-7608/64/030493-10*04.00 AP00019836 t i 494 LAURENCE AND POLITZER Chlorooxirane is known to alkylate various nucleic acid bases, as well as DNA, RNA, and protein residues [2, 13-16], forming covalent bonds to nucleophilic sites on these systems, and it has also been found to alkylate DNA in vivo [17]. Such interactions could lead to miscoding [11, 16, 18], and might result in replicational and transcriptional errors [17a, 19], and the development of tumors. We have earlier computed the structures and various properties for chlorooxirane and its oxygen- and chloro-protonated forms (structures 1H and IV) [20].* Protonation of the oxygen was found to weaken both C--O bonds, H* /\ Ci-yCi C*r-H HH m /\ Cl"/C! C*""H HH iv particularly the Q--O.t (The chlorine-bearing carbon shall be designated Ct.) An even more sriking effect was the very marked weakening of the C--Cl bond that resulted from protonation of the chlorine. This suggests that rupture of this bond may conceivably play a significant role in the biological functioning of this molecule. In order to understand better the alkylating properties of chlorooxirane, we have now made a detailed computational study of Us behavior in both S*1 and 5*2 reactions. To simulate the former processes, we simply stretched each of the key bonds in turn, and computed the energy requirements for doing so. For SN2 reactions, the ammonia molecule was used as a model nucleophile. As shall be shown, one of the consequences of this investigation is that it permits us to suggest mechanisms for the formation of the primary in vivo DNA alkylation product that has been observed for vinyl chloride [17]. 2. Methods The oxygen-protonated epoxide was used for investigating the 5*1 reactivities of the C--O bonds. To simulate 5*1 ring opening, each bond separately was stretched in several incremental steps, the entire structure being reoptimized at each step. Essentially the same procedure was used to study the 5*1 reactivity of the C--Cl bond, but involving now the chloro-protonated form of the epoxide. "The extent to which chlorooxirane is protonated in civo is uncertain, However, there are known ta exist subcellulsr regions with high proton activities [21], It is of particular interest that the nuclear membrane contains certain redox enzymes [22], among which are the cytochrome P-450 systems that activate some carcinogens, including vinyl chloride {10, 11, 23] These redox processes are believed to generate high local proton activities [21c, 24] It seems significant that the eynzymatic processes involved in DNA replication are also localized in the nuclear membrane. In addition, there it evidencespecifically forchlorooxirane which suggest that itsbiological activity involves a protonited form [25] t Bond weakening was inferred from increases in bondlengths and decreases In force constants. ^ ft ( f AP00019837 PROPERTIES OF CHLOROOXIRANE 495 Sn2 processes were examined by allowing the ammonia to approach either Ci or C2; at each different C--N distance, all structural parameters were again reoptimized. This was done for both the oxygen- and chloro-protonated species. The computational method used for this work was the GAUSSIAN 70 ab initio self-consistent-field molecular orbital procedure, at the sto-3G level [26]. This has been shown to be an effective technique for structure calculations [27]. 3. Remits and Discussion A SfjJ Processes (a) C--O bonds: The energies of opening the epoxide ring by simply stretch ing either C--O bond from its equilibrium distance of 1.50 A, without any approaching nucleophile, were investigated for the oxygen-protonated form of chlorooxirane. The relative energies of the reoptimized structures at each different C--O distance are given in Table I. Significantly more energy is required to stretch the C2--O bond by the arbitrary amount of 0.7 A than the Ci--O: 44 vs. 32 kcal/mol. This is consistent with our earlier findings concerning the relative strengths of these two bonds [20]. Table I. Calculated relative energies for bond-stretching processes. OxygeB*PratOlitd Chlerooiirins: Cj-0 Bond: Bond Length (A): 1.S0 Ralativa Energy: (ke*l/*l) Cj'O Bond: c.o Bond Length (A); l.SO Rtlttivo Energy, (keol/nolo) 0.0 1. S 2 0.1 I.S2 0.3 Chlorin*-Protonatd Cklorooxiran*: Cj-Cl Bond: Bond LtngEl) (A): 2.07 Rolitiro Enorgy: 0.0 2. IS 0.3 1.60 2.9 l.SO 3.6 2. SO 1.73 13.0 1.7S IS.7 I.7S 2.1 1.00 2J.0 1.90 21.6 3.00 2.7 2.20 S3.4 2.2D 44.0 3.21 3.S As one C--O distance is increased, the other carbon is observed to move toward a more tetrahedral configuration; for instance, its H--C--C angles change from approximately 120 to roughly 112. The C--C and the other C--O bond lengths also change toward their normal single bond values. One notable feature is a significant shortening of the Cj--Cl bond, from 1.77 to 1.71 A, that accom panies the Ci--O stretch. An examination of the calculated atomic charges (population analysis [28]} in these systems shows a pattern very similar to what was observed in a previous AP00019838 496 LAURENCE AND POLITZER study of protonated ethylene oxide [29]. As one of the C--O bonds is stretched, there is a general movement of electronic charge from that carbon and its attached atoms (whether H or Cl) to the remainder of the molecule. Most of this apparent charge shift (the total amount of which is about 0.25 e.u.) goes to the oxygen and the proton. Thus the bond openings can be regarded as having a C*$ * 0"a character. ______________ _____ _ ^____ ___ (b) C--Cl band: Table I also lists the energies involved in stretching the Ci--Cl bond from its equilibrium value in thechloro-protonated epoxide, 2.07 A, to 3.50 A. Only 4 kcal/mol are required, much less than for either C--O bond. This is again in agreement with our earlier work [20], and demonstrates the weakness of the C--CI bond in chlorooxirane. A marked decrease in the Cj--O distance, from 1.37 to 1.28 A, accompanies the stretching of the Q--Cl bond, whereas the C2--O bond lengthens from 1.46 to 1.50 A. These changes can be explained by noting that the large increase in the C--Cl distance is nearly equivalent to removing HC1, which the calculated atomic charges show to come of? as a neutral entity. The resulting positive charge on C, [structure V(a>] can be delocalized by contributions from resonance structures V(b) and V(c), which should concomitantly shorten the Ci--O bond and lengthen the C2--O; ;c-c=c H vu> H //,01v2, / <Ah HH V(b) o // -t+j HH V(c) The fact that the C:--O bondlength is 1.28 A, which is quite close to the standard r=c=0 value of 1.21 A [30], suggests that structures V(b) and V(c) make significant contributions. B. Sn2 Processes (a) Interactions of ammonia with oxygen-protonated chlorooxirane: In attacking either carbon atom, the ammonia molecule was given complete freedom regarding its line and angle of approach. These interactions led to the formation of two stable intermediate complexes, structures VI and VII. H* 1 Cl O \ H- "ClCH NH , H VI H* \ 0.% H r *V~H ay-'--V H NH, VU * t AP00019839 PROPERTIES OP CHLOROOXIRANE 497 Property Dittancti Cj-O cro crcz Cj.Cl CVK C,*H 0-H* Cl-H* C-N Table II. Calculated structures.* 0C-hPlreorxoonxirmd t.0 JUt) CCWhloTrOsoiexnlritafcnd. (IV) AnaonU CO*DlX*S 0-Prateaat*d t Cj (VI) CMree*irn t C2 IVI1) CChl-lPoriootooxnlratn*i (Vltl) 1 .JO 1,36 2.24 1.40 1.42 1.50 1.46 1.42 2.31 1.44 1.S0 1.41 1.36 1.55 1.49 1.77 2.07 1.10 1.14 1. 52 i.io 1.10 1.10 1.10 1.09 1.19 1.19 1.10 1. 09 1.09 1.00 .... 0.99 0. 99 .... -- 1.34 .... .... 1.32 .... .... 1.15 1. 3) 1.54 O-Ct-C* O-Ci-Cj u-crc* !! C,-CrH. v<vV H-CZ-C1 K-C2-Hb H'-O-C, H'.Cl.Cj ' N-e-e 60 60 120 121 120 120 116 117 114 114 ... ... 62 55 140 112 111 11V 99 117 ... ... 103 ... -16 161 120 36 111 U2 113 107 19* 111 110 110 10* 106 10* no 114 106 1S 146 ... 112 in 59 51 126 196 120 111 62 115 ... 164 111 * Distance* are in A; angles are in degrees. *The structures given for the protonated chlorooxinnes are essentially the same as in ref. 20, with some small change* due to further optimization. * Ht is tram to C). Their optimized geometries are given in Table II. In both cases, the nitrogen is is the plane of the ring and forms a roughly tetrahedral N--C--C angle, although its initial approach was approximately perpendicular to the C--C bond. The variation of the total energy during the course of each interaction is shown in Table III. In the formation of VI, a small energy barrier is observed during the early stages of the process, reaching its peak at a Ct--N distance of about 2.2 A. This appears to reflect the pretence of a hydrogen bond between the chlorine and the proton in the equilibrium form of the oxygen-protonated epoxide, structure III. This interpretation is supported by the fact that the Cl--H+ distance in 111 is 2.74 A, a reasonable value for this type of hydrogen bond [31]. AP00019840 49S LAURENCE AND POLITZER Table III. Calculated interaction energies for reactions with ammonia. nvEci-rrotonti Chlorooxlrane: ftr act ion at Z\ ' Cj-M Ciatane* (A): 3.00 Interaction Energy: (kcal/nol*) -10.5 Reaction it Cj' Cj-K Dittmct (A); interaction Enerjy: Ckcal/"ole) 3.00 -10. 6 2.SO --10.3 2.20 -9.J 2.00 -26.0 2.SO 2.20 -12.0 --]2.1 2.00 -21.5 1.80 -53.9 1.55 -71.2 1.00 -ST.7 l.SS -86.9 Chlorine*Protonated ChioreokIran*t Reaction at C} [approach along Cj-Cl line!: Cj-X Diitinca (A); 3.SO S.OO 2.SO Interaction Energy: (kcal/rele) -9.0 -12,2 -11.9 Reaction at Cj [approach along Cj-H list): Cj*K Distance (A)i 3.50 3.00 2. 50 Interaction Energy: --13.5 -23.1 -31.6 2.20 --J7.1 2.20 -4.1 2.00 --S4.6 t.IO --75.1 1.54 -95.9 In addition, the calculated charge on the proton, which was varying quite gradually at large C--N separations, changes very markedly (from +0.33 to +0.25) as the C--N distance decreases from 2.2 to 2.0 A. Table III shows that the intermediate complex formed by S*2 attack at C2 is the more stable of the two possibilities, the computed interaction energies being --87 kcal/mol for VII and -73 kcal/mol for VI. This is in interesting contrast to the observation discussed above, that in the absence of a nucleophile it is the Cj--O bond that opens more readily. The changes in geometry that accompany these two SN2 processes are essen tially similar (Table II). The C--O bond involving the carbon being attacked lengthens to approximately 2.3 A, and there is inversion of the attached hydrogen atoms (or hydrogen and chlorine). The lengths of the C--O and C--C bonds involving the other carbon change in the direction of more typical single-bond values, and the orientation of the bonds around each carbon becomes more tetrahedral. A notable feature of the reaction at C2 is the concomitant lengthening of the C}--Cl bond from 1.76 to 1.84 A. Some possible implications of this will be discussed later. (b) Interaction of ammonia with chloro-protonated chlorooxlrane: In its interaction with the chloro-protonated epoxide, the favored position for the ammonia is essentially in line with the Ci--Cl bond (toward the carbon), and hence not in the plane of the ring (structure VIII). The optimized geometry of the intermediate complex that is formed is given in Table II. The major structural effects are the lengthening of the C|--Cl bond, from 2.07 to 3.52 A, and of the C|--O bond, from 1.37 to 1.42 A, plus an inversion of the hydrogen attached t AP00019841 PROPERTIES OF CHLOROOXIRANE 499 vin to C]. As seen from Table III, the product is more stable (AjE * --96 kcal/mol) than either of the complexes formed with the oxygen-protonated epoxide (VI and VU). On the other hand, we found no significant tendencies for complex formation between ammonia and C3. While examining various possible paths that the NH3 could follow in attacking Ci, it was found that initially approach along the line of the C]--H bond (toward the H) is favored over the Ci--Cl line. (See Table III.) This may reflect another instance of hydrogen bonding, this time involving (rather surprisingly) the ammonia nitrogen and the hydrogen on C,. This route of attack is the more stable one until the C]--N distance is approximately 2.5 A; meanwhile the Cj--H bondlength increases from. 1.1. to 1.35 A, as the hydrogen interacts increasingly strongly with the ammonia. Eventually the approach along the Cj--Cl line does become the preferred one, although there is an energy barrier in going from the Ci---H line to the C(--CI, breaking the hydrogen bond in the process. C. Possible Mechanisms of Interaction with DNA The results that have been presented in this paper suggest several reactive pathways that chlorooxirane can follow. Two of these are of particular interest in the present context, because they represent possible routes to the formation of a DNA-chlorooxirane reaction product that may play a key role in the car cinogenic action of vinyl chloride. Recent studies have concluded [17], in support of earlier work [15], that the 7-N-(2-oxoethyl) derivative of guanine (structure IX) is the major in vivo DNA alkylation product of vinyl chloride. (There is some evidence suggesting that this aldehyde, IX, may be in equilibrium with a cyclic hemiacetal involving 0-6 [17a].) It was proposed that the formation of IX may correspond to the "primary lesion" in DNA that might be responsible for the carcinogenicity of vinyl chloride [17b]. As pointed out earlier, the C--Cl bond in the chloro-protonated form of chlorooxirane is weakened to such an extent that only a small energy input is required for the removal H-C=0 APOOO19842 500 LAURENCE AND PQLITZER of the chlorine. This leaves behind a charged species (V) having considerable aldehyde character. One can then readily envision the highly nucleophilic [32] N-7 position on guanine reacting with C2 to produce IX. The sequence of steps would be: Ah: v<5F>cr- Cir-H l yC` C1'H * " yCr--Ci'-H H H H HH H J> 'H H IX X (sunin) We can also account for the formation of DC by means of an SN2 process involving oxygen-protonated chlorooxirane. Table II shows that the interaction of ammonia with C2 that leads to the intermediate VII is accompanied by a rather surprising increase in the C,--Cl bondlength, presumably indicating a significant weakening of this bond, as well as a shortening of the C|--O distance, from 1.49 to 1.41 A. This suggests the following mechanism, again involving N-7 of guanine: H* A a*/ci cir*H HH H* / '*0 A * a*/'1 H A. H ty" M J! r-S+H A c\` 1 . --a* -1 OH AM__ C^<-H H EX H S .f H H C\a * APOOOI9843 PROPERTIES OF CHLOROOXIRANE 501 Thus, the results obtained in this work allow us to propose reasonable mechanisms, both 5N1 and SN2, for the formation of the key alkylation product, IX. On purely thermodynamic grounds, the 5*2 process should be the favored one, since it starts with the more stable of the two possible protonated epoxides* and the nucleophile goes to the carbon, Ci, at which the more stable product is formed (Table III). 4. Summary This study indicates several possible reactive pathways that protonated chtorooxirane can follow. It supports the suggestion [20] that the departure of the chlorine may play an important role in the biological mode of action of vinyl chloride. Of particular importance with regard to the carcinogenic activity of this molecule is the establishing of two possible routes, one 5*1 and the other 5*2, to the formation of what has been observed experimentally to be the primary in vivo DNA alkylation product of vinyl chloride, structure K [17]. We are currently investigating mispairing possibilities that could result from this modification of guanine. Acknowledgments The authors greatly appreciate several very helpful discussions with Dr. G. Rickey Welch. They thank the U.S. Environmental Protection Agency for partial funding of this work under assistance agreement number CR808866-01-0 to Peter Politzer. The contents do not necessarily reflect the views and policies of the Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. They are also grateful for the financial support provided by the Research and Develop* ment program of the Louisiana Board of Regents and the University of New Orleans Computer Research Center. Bibliography [1] C. Maltoni, Environ. Health Perspec. 21. 1 (1977). [2] U. Rtnnug, R, Gothe, and C. A. W&chtmeister, Chem. Biol. Interact. 12, 251 (1976). [3] C. B. Hong, J. M. Winston, L. P. Thornburg, C. C. Lee, and J. S. Woods, J. ToxicoL Environ. Health 7, 909 (1981). [4] J. M. Rice, Environ. Health Perspec. 41,179 (1981). [5] J. D. Fabricant and M. S. Legator, Environ. Health Perspec. 41,189 (1981). [6] I. J. SelikolT and E. C. Hammond. Eds.. Toxicity of Vinyi Chloride-Polyvinyl Chloride, Ann. N.Y. Acad. Sd. 246 (1975). -- [7] P. F. Infante, Environ. Health Perspec. 41, 89 (1981). [8] B. L. Van Duuren, Ann. N.Y. Acad. Sd. 246, 258 (1975). [9] T. Green and D. E, Hathway, Chem. Biol. Interact. 17, 137 (1977). tlO] H. Bartsch. C. Malaveille, C. Barbin. and G. Planche, Arch. Toxicol. 41, 249 (1979). [11] F. Zajdela, A. Croisy, A. Barbin, C. Malaveille, l~ Tomatis, and H. Bartsch, Cancer Res. 48, 352 (1980). *Our calculated proton affinity for the oxygen in chlorooxirane is 2l6kcal/jnol, compared to 169 kcal/rool for the chlorine [20]. AP00019844 502 LAURENCE AND POLITZER [12] J. D. Elmore, J. L. Wong. A. D. Laumbach, and U. N. Streips, Biochem. Btophys. Acta 442, 405 (1976). [13] T. E. Green and D. E. Hathway. Chem. Biol. Interact. 22. 211 (1978). [14] A. Barbin, H. Bresil, A. Croisy, P. Jacquignon, C. Maiaveille, R. Montesano, and H. Bamch, Biochem. Biophys. Res. Commun. 67, 596 (1975). [15] S. Osterman-Golkar, D. Huitmark, D.Segerback.C. J.Calleman. R. Gothe. L. Ehrenberg, and C. A. Waehtmeister, Biochem- Btophys. Res. Commun. 76, 259 (1977). [16] A. Barbin. H. Baruch, P. Leconte, and M. Redman. Nud. Adds Res. 9, 375 (1981). [17] (a) E. Scherer, C. J. Van Der Laken, L. M. Gwinnei, R. J.Laib.and P. Emmelot, Carcinogenesis 2.671 < 198 l);(b)R, J.Laib,L.M. Gwinner. and H.M. Bolt, Chem. Biol. Interact. 37,219 (1981). [18] D. E. Hathway and G. F. Kolar, Chem. Soc. Rev. 9. 241 (1980). [19] S. Spengler and B. Singer, Nud. Acids Res. 9, 365 (1981). [20] P. Politzer and T. R. Proctor, Int. J. Quantum Chem. 22, 1271 (1982). [21] (a) A. D. McLaren, Enzymologia 21. 356 (1960); (b) A. Sols and R. Marco. Curr. Top. Cell. Regu. 2, 227 (1970); (c) G. R. Welch and M. N. Berry, in Coherent Excitations in Biological Systems. H. Frohlich and F. Kreraer, Eds. (Springer*Verlag, Heidelberg, id press). [22] J. W. DePierre and L. Ernster. Ann. Rev. Biochem. 46, 201 (1977). [23] C. Walsh, Emymatic Reaction Mechanisms (Freeman, San Francisco, 1979), Chap. 15. [24] (a) D. B. Kell, Biochim. Biopbys. Acta 549, 55 (1979); (b) M. N. Berry, A. R. GriveU, and P. G. Wallace, in Comprehensive Treatise on Electrochemistry, VoL 10, Bloeleetrochemistry, S. Srinivasan and Y. A. Bhizmadzhev, Eds. (Plenum, New York, in press). [25] S. Hussain and S. Osterman-Golkar, Chem. Biol. Interact. 12, 265 (1976). [26] W. J. Hehre, W. A. Lathan, R. Ditchfield, M. D. Newton, and J. A. Pople, Quantum Chem. Program Exchange 11. 236 (1973). [27] J. A. Pople, in Applications of Electronic Structure Theory, H. F. Schaefer III, Ed. (Plenum, New York, 1977), Vot. 4, Chap. 1. [28] R. S. Mulliken, J. Chem. Phys. 23, 1833 (1955). [29] P. Politzer and V. M. Estes, in Catalysis in Chemistry and Biochemistry, Theory and Experiment, B. Pullman, Ed. (Reidei, Dordrecht, 1979), p. 305. [30] Tables of Interatomic Distances and Configuration in Molecules and Ions, L. E. Sutton, Ed. (The Chemical Society, London, 1965), Supplement, Spec. Pub. No. 18. [31] P. Kollman. J. McKelvey, A. Johansson, and S. Rothenberg, J. Am. Chem. Soc. 97,955(1975). [32] B. Singer. J. Toxicol. Environ. Health 2. 1279 (1977). Received January 25, 1983 Accepted for publication April 28, 1983 | f t h AP00019845