Inhalation exposure of rats and mice to 1,3-butadiene induces N6-adenine adducts of epoxybutene detected by 32P-postlabeling and HPLC.

In this paper we report DNA binding of butadiene monoepoxide, a first metabolite of 1,3-butadiene catalyzed by monooxygenases. We prepared alkylated purines as marker compounds for 32-P-postlabeling and electrochemical analysis and developed methods to measure the corresponding products. The traditional postlabeling assay was modified by incorporating a solid phase extraction column and high-performance liquid chromatography (HPLC) enrichment steps to the assay prior to labeling. The final analysis of adducted N6 adenines is based on two dimensional thin-layer chromatography (TLC) and an on-line HPLC/radioactivity analysis. The qualitative and quantitative results are based on positively identified marker compounds. Alkylated N7 guanines were released from DNA by neutral thermal hydrolysis, prepurified by HPLC, and analyzed by HPLC with a sensitive electrochemical detection procedure. By using these methods, we found alkylation of calf thymus DNA exposed to butadiene monoepoxide in vitro at adenine N6 and guanine N7 sites. Analysis of lung DNA samples from mice and rats exposed to butadiene through inhalation showed that adenine N6 adducts were formed in vivo in a dose responsive manner.


Introduction
1,3-Butadiene (BD) is not only a high production volume chemical for the polymer industry but is also a common environmental contaminant. It has been estimated that most environmental BD emissions derive from mobile sources. Burning of organic material produces emissions containing low concentrations of BD. Exposure to low concentrations of BD is thus a common characteristic of the entire human population (1).
Long-term inhalation studies indicate that BD is a multisite carcinogen in mice and rats, mice being by far the more sensitive species (2). A recent epidemiological study in the BD industry indicates an elevated mortality of lymphatic and reticular sarcoma among the production workers (3). This paper was developed from a poster that was presented at the 2nd International Conference on Environmental Mutagens in Human Populations held 20-25 August 1995 in Prague, Czech Republic. Manuscript received 22 November 1995; manuscript accepted 28 November 1995.
Financial support for the study was obtained from the projects CT 90-0152 (coordinated by M. Sorsa) and EV5V-CT94-0543 (coordinated by F. Pacchierotti) of DG XIl of the European Communities. We thank the exchange program between the Academies of Finland and the Czech Republic and the Eemil Aaltonen Foundation for personal grants to P. Koivisto.
Small alkylating agents react mainly with the N7 position of guanine, but other alkylation sites in DNA are also detected in vitro (4,5). Goodrow et. al (6) have reported activation of the K-ras protooncogene at codon 13 involving a mutation in guanine. Butadiene monoepoxide (BMO), one of the reactive metabolites of BD, has been reported to alkylate guanine at the N7 position, but poor stability and low labeling efficiency of the product limits its use in postlabeling analyses (7,8). In an inhalation study in which transgenic mice were exposed to BD, an elevated level of A to T transversions were observed in the lad gene (9). Although the alkylation at adenine may be a minor reaction site in DNA, this observation makes adenine an interesting alternative in DNA adduct monitoring.
In this paper we report alkylation of both adenine and guanine derived from BMO in relation to in vitro exposure to BMO or in vivo exposure to BD.

Materials and Methods
All commercial chemicals used in this study were of analytical grade. Adenine N6 adducts of BMO were prepared by reacting deoxyadenosine with 3' and 5'-deoxyadenosine monophosphates in alkaline conditions. The products formed were purified chromatographically and characterized by ultraviolet spectroscopy (UV), nuclear magnetic resonance (NMR), and mass spectrometry (MS) (10,11). Guanine N7 adducts were prepared and analyzed as described by Neagu et al. (7). The structures of the compounds studied are shown in Figure 1.
Calf thymus DNA (1 mg/ml) was exposed to BMO (10 pl/ml) for 1 and 6 hr. BMO was evaporated from the samples and the samples were subjected to enzymemediated digestion or thermal hydrolysis.
DNA for N6 adduct analysis was isolated using a phenol extraction procedure (12). For the postlabeling analysis, an aliquot of DNA was taken and digested to 3' nucleotides with micrococcal nuclease and spleen phosphodiesterase. BMO alkylated 2'deoxyadenosine-3'-monophosphates were first enriched with disposable solid phase extraction cartridges (octadecylsilane phase). Normal nucleotides (NNs) were washed out from the column with 50 mM ammonium formate (AF) containing 3% methanol, pH 4. The amount of DNA Environmental Health Perspectives -Vol 104, Supplement 3 * May 1996 digested was measured from the normal nucleotides by HPLC. N6 alkylated adenine adducts were eluted with 50% water/methanol.
Adduct enrichment was performed with reversed-phase HPLC. A fraction was collected from 17 to 20 min (residual NNs had the following retention times: 2'-deoxycytidine-3'-monophosphate, 4.3 min; 2'-deoxyguanosine-3'-monophosphate, 5.0 min; 2'-deoxythymidine-3'-monophosphate, 6.6 min; 2'-deoxyadenosine-3'monophosphate, 7.7 min). Enriched adducts were postlabeled to 5' position by T4 kinase; the formed bisphosphates were treated with nuclease P1 resulting in 5' monophosphate adducts. The samples were spiked with the corresponding 5' compound and separated by two dimensional TLC chromatography. Dimension 1 was 0.3 M AF and dimension 2 was done with isobutyric acid/ammonia/water eluent in a ratio 69/1/30. The adducts were located on TLC plates with autoradiography ( Figure 2); the spot was scraped and consequently extracted with IM AF. The ultimate analysis was carried out on a reversedphase HPLC equipped with UV and radioactivity detectors. The HPLC gradient was 3% acetonitrile maintained for 5 min followed by a linear gradient to 16% during 26 min. The buffer was 0.75M AF and the ratio of scintillation solvent to HPLC effluent was 4:1.

Results and Discussion
We applied a solid phase extraction (SPE) procedure as a first step in our analytical protocol. The alkylated N6 products were retained in the cartridge and the normal nucleotides were washed out. The SPE step was necessary to avoid column loading caused by NNs in preparative HPLC. If severe column loading occurs, NNs may interfere with the postlabeling analysis.
By carrying out an ion exchange TLC and a reversed-phase HPLC, the analysis used two different retention mechanisms, which improved the qualitative reliability of the analysis. From a quantitative point of view, the use of a HPLC/radioactivity detector decreases the detection limit by the factor of 3 when compared to traditional scintillation counting. However, the standard deviation is much less pronounced if the analysis is carried out using the HPLC/radioactivity detector (SD ± 4%; n= 3) instead of the commonly used Cherenkov counting technique (SD ± 23%; n= 3). N7-Guanine adducts When calf thymus DNA was exposed to BMO for 1 and 6 hr, five times more adducts (both N7 and N6) were observed with the longer time exposure. The amount of N7 alkylated guanines was about 300 times higher than the amount of N6 products; however, the stability of the product was shown to be very different. As reported by Neagu et. al. (7), the N7 products showed a half life of 48 hr, but the N6 Environmental Health Perspectives -Vol 104, Supplement 3 * May 1996 products were stable and showed no decrease in the adduct levels within the 2week period studied. This fact may favor the use of N6 adducts in future biomonitoring applications because the N6 adducts of adenine are able to reflect cumulative effects from long-term exposure.
Two dimensional TLC by itself was not able to resolve all of the interfering products present in the in vivo samples. By combining a subsequent HPLC analysis with a radioactivity detector, the selectivity of the analysis was much improved ( Figure  3). In the lung samples of the rats and mice exposed by inhalation to BD ( Figure  4), a dose-response increase of the N6 adenine alkylation by BMO was detected. Insignificant differences were seen between the species, the lung tissue of mouse being only very weakly more sensitive than the rat lung to the BMO-derived N6 adenine adduct formation. At the lowest exposure level (50 ppm) no differences were seen in the adduct levels when compared to the  the corresponding lung samples from which the N6 adducts of adenine are now reported.