Identification and Quantification of the Depurinating DNA Adducts

Apr 30, 2005 - John C. Lodge Freeway, Detroit, Michigan 48202-3398, Department of Analytical Chemistry and. Applied Spectroscopy, Laser Centre, Vrije ...
0 downloads 0 Views 148KB Size
976

Chem. Res. Toxicol. 2005, 18, 976-983

Identification and Quantification of the Depurinating DNA Adducts Formed in Mouse Skin Treated with Dibenzo[a,l]pyrene (DB[a,l]P) or Its Metabolites and in Rat Mammary Gland Treated with DB[a,l]P Ercole L. Cavalieri,*,† Eleanor G. Rogan,† Kai-Ming Li,‡ Rosa Todorovic,§ Freek Ariese,| Ryszard Jankowiak,⊥ Nenad Grubor,⊥ and Gerald J. Small⊥ Eppley Institute for Research in Cancer and Allied Diseases, University of Nebraska Medical Center, 986805 Nebraska Medical Center, Omaha, Nebraska 68198-6805, Ash Stevens, Inc., 5861 John C. Lodge Freeway, Detroit, Michigan 48202-3398, Department of Analytical Chemistry and Applied Spectroscopy, Laser Centre, Vrije Universiteit, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands, and Department of Chemistry, Iowa State University, Ames, Iowa 50011-3111 Received November 15, 2004

Dibenzo[a,l]pyrene (DB[a,l]P) is the most potent carcinogenic polycyclic aromatic hydrocarbon and has been identified in the environment. Comparative tumorigenicity studies in mouse skin and rat mammary gland indicate that DB[a,l]P is slightly more potent than DB[a,l]P11,12-dihydrodiol and much more potent than (()-syn-DB[a,l]P-11,12-dihydrodiol-13,14-epoxide {(()-syn-DB[a,l]PDE} and (()-anti-DB[a,l]PDE. We report here the identification and quantification of the depurinating adducts formed in mouse skin treated with DB[a,l]P, DB[a,l]P11,12-dihydrodiol, (()-syn-DB[a,l]PDE, or (()-anti-DB[a,l]PDE and rat mammary gland treated with DB[a,l]P. The biologically formed adducts were compared with standard adducts by their retention times on HPLC and their spectra obtained by fluorescence line-narrowing spectroscopy at low temperature. In mouse skin treated with DB[a,l]P, depurinating adducts comprised 99% of the total adducts. Most of the depurinating adducts were formed by one-electron oxidation, with 63% at Ade and 12% at Gua. The remainder were formed by the diol epoxide, with 18% at Ade and 6% at Gua. When mouse skin was treated with DB[a,l]P-11,12-dihydrodiol, depurinating adducts comprised 80% of the total, and the predominant one was with Ade (69%). Treatment of skin with (()-syn-DB[a,l]PDE resulted in 32% depurinating adducts, primarily at Ade (25%), whereas treatment with (()-anti-DB[a,l]PDE produced 97% stable adducts. The formation of depurinating adducts following treatment of rat mammary gland with DB[a,l]P resulted in approximately 98% depurinating adducts, with the major adducts formed by oneelectron oxidation. Only one depurinating diol epoxide adduct was formed. Tumorigenicity, mutations, and DNA adduct data suggest that depurinating Ade adducts play a major role in the initiation of tumors by DB[a,l]P.

Introduction Dibenzo[a,l]pyrene (DB[a,l]P) is the most potent carcinogen among the various dibenzo[a]pyrenes (1) and is the most potent carcinogenic polycyclic aromatic hydrocarbon (PAH) (1-5). Comparative tumorigenicity studies of DB[a,l]P, DB[a,l]P-11,12-dihydrodiol, (()-syn-DB[a,l]P11,12-dihydrodiol-13,14-epoxide {(()-syn-DB[a,l]PDE}, and (()-anti-DB[a,l]PDE in mouse skin and rat mammary gland indicate that DB[a,l]P is slightly more potent than the DB[a,l]P-11,12-dihydrodiol and much more potent than the two diol epoxides (2, 3, 5). DB[a,l]P has been tentatively identified in the biologically active fraction of cigarette smoke condensate (6). This PAH is also thought to be a component of other * To whom correspondence should be addressed. Tel: 402-559-7237. Fax: 402-559-8068. E-mail: [email protected]. † University of Nebraska Medical Center. ‡ Ash Stevens, Inc.. § E-mail: [email protected]. | Vrije Universiteit. ⊥ Iowa State University.

pollutants (7). Definite identification of this compound in the environment has been obtained in particulates formed by combustion of smoky coal (8, 9), coal tar extract, air particulate matter, and marine sediment (10). The identification and quantification of DB[a,l]P-DNA adducts formed in vitro when DB[a,l]P is activated by 3-methylcholanthrene-induced rat liver microsomes reveal two major features: (i) the preponderant adducts are the depurinating adducts (84%), and the stable adducts comprise 16% of the total adducts; and (ii) the depurinating adducts are formed 60% by DB[a,l]P radical cation and 40% by the fjord region DB[a,l]PDE (11). In those studies, all depurinating adducts and approximately 85% of the stable adducts were identified and quantified (11, 12). Identification and quantification of the adducts formed by DB[a,l]P in vivo provide information on the DNA damage that leads to cancer-initiating mutations. In this article, we report the identification and quantification of the depurinating adducts formed in mouse skin treated with DB[a,l]P, DB[a,l]P-11,12-dihydrodiol, (()-syn-DB[a,l]PDE, or (()-anti-DB[a,l]PDE and rat mammary

10.1021/tx049682k CCC: $30.25 © 2005 American Chemical Society Published on Web 04/30/2005

Depurinating DB[a,l]P-DNA Adducts Formed in Vivo

Chem. Res. Toxicol., Vol. 18, No. 6, 2005 977

Figure 1. Structures of synthesized and biologically formed DB[a,l]P radical cation and diol epoxide depurinating adducts.

gland treated with DB[a,l]P. Identification and quantification of the stable adducts formed in these target organs are presented in an accompanying article (13). These analyses were possible because of the synthesis and unequivocal structural determination of authentic standard adducts formed by DB[a,l]P radical cation (1416) or the (()-syn-DB[a,l]PDE and (()-anti-DB[a,l]PDE (17-19). Standard adducts arising from the radical cation were synthesized by anodic or iodine oxidation of DB[a,l]P in the presence of dG or dA (14, 15). The depurinating adduct at the N-3 of Ade was obtained by reaction of DB[a,l]P radical cation with Ade and not with dA, because the adjacent deoxyribose moiety at N-9 in dA hinders the approach of the DB[a,l]P radical cation to the N-3 of dA (15, 16). This hindrance also occurs with other PAH (20, 21) and catechol estrogen quinones (22). The reaction of the (()-syn-DB[a,l]PDE and (()-anti-DB[a,l]PDE with dA or dG provided the authentic standard depurinating adducts formed by the diol epoxides (17-19). Identification and quantification of the depurinating adducts were achieved by comparison of the biologically formed adducts with authentic standard adducts in terms of their retention times on HPLC and their spectra obtained by fluorescence line-narrowing spectroscopy (FLNS) at low temperature.

Experimental Procedures Caution: DB[a,l]P, DB[a,l]P-11,12-dihydrodiol, (()-syn-DB[a,l]PDE, and (()-anti-DB[a,l]PDE are hazardous chemicals and were handled according to NIH guidelines (23). Materials. DB[a,l]P was obtained from the National Cancer Institute Chemical Carcinogen Repository (Bethesda, MD). DB[a,l]P-11,12-dihydrodiol, (()-syn-DB[a,l]PDE, and (()-anti-DB[a,l]PDE were synthesized according to the method of Gill et al. (5). The standard depurinating adducts, DB[a,l]P-10-N7Ade, DB[a,l]P-10-N7Gua, and DB[a,l]P-10-C8Gua (Figure 1), were synthesized by electrochemical or iodine oxidation of DB[a,l]P in the presence of dA or dG, respectively (14, 15). DB[a,l]P-10N3Ade (Figure 1) was prepared by electrochemical or iodine oxidation of DB[a,l]P in the presence of Ade (15, 16). (()-synDB[a,l]PDE-14-N7Ade and (()-anti-DB[a,l]PDE-14-N7Ade (Figure 1) were synthesized by reacting the diol epoxide with dA (18), and the analogous (()-syn-DB[a,l]PDE-14-N7Gua and (()-

anti-DB[a,l]PDE-14-N7Gua (Figure 1) were synthesized by reacting the diol epoxide with dG (19). Treatment of Mouse Skin. Eight week old female Swiss mice (Eppley Colony) were treated in groups of eight on a shaved area of dorsal skin with DB[a,l]P, DB[a,l]P-11,12-dihydrodiol, (()-syn-DB[a,l]PDE, or (()-anti-DB[a,l]PDE (200 nmol in 50 µL of acetone). After 4 h, the mice were sacrificed, the treated area of skin was excised, and the epidermis was isolated, minced, and ground in liquid nitrogen. The pooled, ground epidermis was divided into two aliquots weighing approximately 7 and 1 g each. The smaller aliquot was used to purify DNA and analyze stable adducts by the 32P-postlabeling procedure (13), while the larger aliquot was Soxhlet extracted with a mixture of chloroform-methanol (1:1) for 48 h and analyzed for depurinating adducts. Treatment of Rat Mammary Gland. Groups of 20 seven week old female Sprague-Dawley rats (Harlan, Indianapolis, IN) were lightly anesthetized with ether, and the mammary region was shaved. They were then treated by intramammillary injection (with a 27-gauge needle) under the nipple region of the 4th and 5th mammary glands on both the right and the left sides with DB[a,l]P at a dose of 200 nmol per gland (in 20 µL of Me2SO). After 24 h, the rats were sacrificed, and the mammary gland areas were excised. The mammary tissue was minced, ground in liquid nitrogen, and split into two samples weighing approximately 25 and 1 g each. The smaller aliquot was used to purify DNA and analyze stable adducts by the 32P-postlabeling procedure (13), while the larger aliquot was Soxhlet extracted with a mixture of chloroform-methanol (1:1) for 48 h and analyzed for depurinating adducts. HPLC. HPLC was conducted on a Waters 600E solvent delivery system equipped with a Waters 700 WISP autoinjector. Effluents were monitored for UV absorbance at 254 nm with a Waters 996 photodiode array detector, and data were collected on a PowerMate computer. Analytical runs were conducted by using a YMC (Wilmington, NC) ODS-AQ 5 µm, 120 Å column (6.0 mm × 250 mm). After the column was eluted for 5 min with 20% CH3CN in H2O, an 80 min linear gradient to 100% CH3CN was run at 1 mL/min. Both CH3CN/H2O and CH3OH/ H2O gradients were used for adduct purification (see below). Extracts obtained from each mouse skin or rat mammary gland experiment, which contained depurinating DNA adducts and metabolites, were evaporated to dryness under vacuum. The residue was dissolved in 0.2 mL of Me2SO, followed by an equal volume of CH3OH. After sonication to enhance solubilization, the undissolved residue was removed by centrifugation. The sample was then analyzed with the above CH3CN gradient, and

978

Chem. Res. Toxicol., Vol. 18, No. 6, 2005

fractions were collected at the retention times of the standards, DB[a,l]P-10-N7Ade, DB[a,l]P-10-N3Ade, DB[a,l]P-10-N7Gua, DB[a,l]P-10-C8Gua, (()-syn-DB[a,l]PDE-14-N7Ade, (()-antiDB[a,l]PDE-14-N7Ade, (()-syn-DB[a,l]PDE-14-N7Gua, and (()anti-DB[a,l]PDE-14-N7Gua (11). The collected fractions were evaporated to dryness under vacuum, and the residues were dissolved in 0.1 mL of Me2SO/CH3OH (1:1). The redissolved fractions were then reinjected on HPLC eluted with CH3OH/ H2O (30:70) for 5 min, followed by a concave (CV5) 70 min gradient to 100% CH3OH at a flow rate of 1 mL/min. On-line detection of the adducts was provided by a Jasco FP-920 (Jasco, Easton, MD) fluorescence detector. If the adduct was not pure, the reinjected fraction was further purified by a third HPLC eluted isocratically with CH3OH/H2O (70:30). Analysis of Depurinating Adducts by FLNS. The adducts formed in mouse skin or rat mammary gland were analyzed by FLN spectroscopy, as previously described (24, 25). Nonlinenarrowing (NLN) fluorescence spectra at T ) 77 K and FLN spectra (S1 r S0 excitation) at T ) 4.2 K were obtained using a Lambda Physik FL-2002 dye laser pumped by a Lamba Physik Lextra 100 XeCl excimer laser as the excitation source. For FLNS, several excitation wavelengths were used, each revealing a portion of the S1 excited state vibrational frequencies of the analyte. Samples were cooled in a glass cryostat with quartz optical windows. The fluorescence was dispersed by a McPherson 2061 1 m focal length monochromator and detected by a Princeton Instruments IRY 1024/G/B intensified diode array. For time-resolved detection, a Princeton Instruments FG-100 pulse generator was employed. The detector delay time and gate width were set to 25 and 200 ns, respectively. For FLN measurements, the monochromator was equipped with a 2400 G/mm grating, providing a 9 nm spectral window at 0.05 nm resolution. Ethanol was used as the solvent matrix; samples were dissolved in approximately 20 µL of ethanol by using sonication. Samples were subsequently transferred to quartz tubes (2 mm i.d.) and sealed with a rubber septum. Concentrations of adduct standards were in the 10 µM range. This analytical method provides unequivocal identification of the adducts (14-19). Quantification of Adduct Levels. The depurinating adducts were quantified by HPLC by integrating the adduct peaks determined by the on-line fluorescence detector and comparing the total peak area with a standard curve made by a series of known concentrations of synthesized authentic adducts. The total amount of each adduct was normalized to the amount of DNA calculated from the weight of the mouse epidermis or rat mammary gland sample. The amounts of the stable adducts (13) were calculated by the 32P-postlabeling method previously described (26, 27). Recovery of adducts is limited to those adducts present in tissue at the end of the experiment, 4 h in mouse skin and 24 h in rat mammary gland (26, 27).

Results Identification of the depurinating DNA adducts formed in mouse skin treated with DB[a,l]P, DB[a,l]P-11,12dihydrodiol, (()-syn-DB[a,l]PDE, or (()-anti-DB[a,l]PDE and in rat mammary gland treated with DB[a,l]P was accomplished by comparison of the retention time on HPLC of the biologically formed adducts with authentic synthesized adducts (Figure 1). Confirmation of the structure of the major biologically formed adducts was obtained by using FLN and NLN spectroscopy, including comparison of spectra with those of the standard adducts. Quantification of the depurinating adducts in mouse skin was obtained from the total peak areas of the adducts separated by HPLC and detected by fluorescence. Mouse Skin. 1. Identification of Depurinating Adducts. The depurinating adducts formed in mouse skin were initially identified by comparison of their retention times on HPLC with those of the authentic

Cavalieri et al.

Figure 2. FLN spectra of the DB[a,l]P-10-N7Ade adduct standard (spectrum a), the biological fractions from the rat mammary tissue (spectrum b) and mouse skin (spectrum c); λex ) 416.0 nm; T ) 4.2 K. The FLN peaks are labeled with their excited state vibrational frequencies in cm-1.

Figure 3. FLN spectra of the DB[a,l]P-10-C8Gua adduct standard (spectrum a) and the biological fraction from rat mammary tissue (spectrum b) and mouse skin (spectrum c). λex ) 416.0 nm; T ) 4.2 K. The FLN peaks are labeled with their excited state vibrational frequencies in cm-1.

synthesized adducts. For most of the abundant depurinating adducts, validation of their structure was obtained by FLN spectroscopy. The presence of the adduct DB[a,l]P-10-N7Ade was readily confirmed by comparison of the standard adduct (Figure 2, spectrum a) with that of the adduct detected in mouse skin treated with DB[a,l]P (Figure 2, spectrum c), which are virtually indistinguishable. Similarly, the presence of DB[a,l]P-10-C8Gua found in mouse skin was validated by FLN spectroscopy, comparing the standard (Figure 3, spectrum a) with the mouse skin adduct (Figure 3, spectrum c) at an excitation wavelength of 416 nm and a temperature of 4.2 K. FLN spectra acquired with the use of other excitation frequencies showed a similar match (data not shown), confirming that DB[a,l]P-10-C8Gua is formed in mouse skin. Identification of DB[a,l]P-10-N3Ade in mouse skin was

Depurinating DB[a,l]P-DNA Adducts Formed in Vivo

Figure 4. FLN spectra of the DB[a,l]P-10-N3Ade adduct standard (spectrum a) and the biological fraction from mouse skin (spectrum b). λex ) 416.0 nm; T ) 4.2 K. The FLN peaks are labeled with their excited state vibrational frequencies in cm-1.

Figure 5. (A) FLN spectra of the (()-syn-trans-DB[a,l]PDE14-N7Ade adduct standard (spectrum a) and the biological fraction from rat mammary tissue (spectrum b). λex ) 371.0 nm; T ) 4.2 K. (B) FLN spectra of the (()-syn-trans-DB[a,l]PDE14-N7Ade adduct standard (spectrum a) and the biological fraction from mouse skin treated with (()-syn-DB[a,l]PDE (spectrum b). λex ) 372.0 nm; T ) 4.2 K; resolution ) 5 cm-1. The FLN peaks are labeled with their excited state vibrational frequencies in cm-1.

straightforward after comparison of spectra from the standard adduct (Figure 4, spectrum a) and the adduct formed in mouse skin (Figure 4, spectrum b) at an excitation wavelength of 416 nm. The spectra of the standard and mouse skin adducts at other excitation wavelengths were also indistinguishable (data not shown), rendering unequivocally the structure elucidation of this compound. The results of fluorescence studies of fractions corresponding to the depurinating diol epoxide adducts, such as (()-syn-DB[a,l]PDE-14-N7Ade, confirmed that the structure of this compound is the same as the one identified by HPLC after comparison with the standard adduct (data not shown). The FLN spectrum (λex ) 372 nm) obtained for the (()-syn-trans-DB[a,l]PDE-14-N7Ade adduct standard (Figure 5B, spectrum a) was compared with the FLN spectrum obtained for the HPLC fraction

Chem. Res. Toxicol., Vol. 18, No. 6, 2005 979

Figure 6. Vibronically excited FLN spectra of the (()-syntrans-DB[a,l]PDE-14-N7Ade. Spectra a correspond to the synthesized reference adduct, and spectra b correspond to the adduct isolated from mouse skin treated with DB[a,l]P11,12-dihydrodiol. Panels A and B were obtained for λex ) 372.0 and 374.0 nm, respectively; T ) 4.2 K. Gate delay, 20 ns; gate width, 200 ns. The FLN peaks are labeled with their excited state vibrational frequencies in cm-1 (see text for details).

from mouse skin treated with (()-syn-DB[a,l]PDE (Figure 5B, spectrum b). Spectra a and b of Figure 5B are identical, indicating that the (()-syn-trans-DB[a,l]PDE14-N7Ade adduct is formed in mouse skin. On the basis of previous conformational studies of DB[a,l]PDE-derived adducts (28) and metabolites (29), we conclude that the (()-syn-trans-DB[a,l]PDE-14-N7Ade adduct, with a characteristic (0,0) band at ∼382 nm, exists in conformation I, in which the cyclohexenyl ring adopts a half-boat structure. The presence of this adduct in mouse skin treated with DB[a,l]P was also confirmed by FLN spectroscopy (data not shown). Both (()-syn-trans-DB[a,l]PDE-14-N7Ade (minor) and (()-syn-cis-DB[a,l]PDE-14-N7Ade (major) adducts are formed in mouse skin exposed directly to DB[a,l]P11,12-dihydrodiol. Frames A and B of Figure 6 display FLN spectra obtained for excitation wavelengths of 372 and 374 nm, respectively. Spectra a and b in Figure 6 correspond to the (()-syn-trans-DB[a,l]PDE-14-N7Ade adduct standard and to the biological HPLC fraction, respectively. The peaks marked by asterisks in spectra b match the standard, indicating that a small amount of the (()-syn-trans-DB[a,l]PDE-14-N7Ade is formed in mouse skin exposed to DB[a,l]P-11,12-dihydrodiol. However, the major adduct is presumably formed by cis addition of the diol epoxide to Ade. The position of the fluorescence origin band (at ∼385 nm) and the similarity of the FLN spectra to those for syn-cis-DB[a,l]P tetrol (29) permit a tentative assignment of this analyte as the (()syn-cis-DB[a,l]PDE-14-N7Ade adduct. This indicates that biologically the (()-syn-cis-DB[a,l]PDE-14-N7Ade adduct is mainly formed. The presence of the minor diol epoxide adducts could not be confirmed by FLNS because of their low abundance. 2. Quantification of Depurinating Adducts. In mouse skin treated with DB[a,l]P, the depurinating adducts comprised 99% of the total adducts, whereas the stable adducts were 1% (Table 1). Most of the depurinating adducts (75%) were formed by one-electron oxidation, DB[a,l]P-10-N7Ade (14%), DB[a,l]P-10-N3Ade (49%), DB[a,l]P-10-N7Gua (2%), and DB[a,l]P-10-C8Gua (10%). The remainder of the depurinating adducts (24%) were

980

Chem. Res. Toxicol., Vol. 18, No. 6, 2005

Cavalieri et al.

Table 1. Quantification of Biologically Formed DB[a,l]P-DNA Adductsa depurinating DNA adducts (µmol adducts/mol DNA-P)

µmol/mol DNA-P in vivo experiments mouse skin treated with DB[a,l]P DB[a,l]P-11,12dihydrodiol (()-syn-DB[a,l]PDE (()-anti-DB[a,l]PDE rat mammary gland treated with DB[a,l]P

(()-syn-DB- (()-anti-DB- (()-syn-DB- (()-anti-DBtotal stable DB[a,l]P- DB[a,l]P- DB[a,l]P- DB[a,l]P- [a,l]-PDE- [a,l]-PDE- [a,l]-PDE[a,l]PDEadducts adducts 10-N7Ade 10-N3Ade 10-N7Gua 10-C8Gua 14-N7Ade 14-N7Ade 14-N7Gua 14-N7Gua

20.1 23.8

0.2 (1)b 2.8 (14)c 4.7 (20)

40.5 42.9

27.6 (68) 41.7 (97)

8.6

0.2 (2)

9.8 (49)c

0.5 (2)

1.9 (10)c

2.4 (12)c 16.5 (69)c

1.3 (6) 1.4 (6)

10.1 (25)c

0.5 (2) 0.5 (2) 2.8 (7)

1.0 (2) 3.5 (40)c

1.5 (17)d

0.5 (6)

1.0 (12)c

0.7 (4) 0.7 (3)

2.0 (23)c,d