Authors: Pertti Koivisto, Marja Sorsa, Francesca Pacchierotti and
Kimmo Peltonen.
Carcinogenesis vol. 18 no.2 pp.439-443, 1997.
Presenter: Jack Rosenfeld
Presenter's Summary
Background:
It has long been recognized that oxidative metabolism of exogenous and endogenous organic compounds is capable of producing chemically reactive molecules. Such compounds can react with DNA and initiate mutation and carcinogenisis. The presence and in-vivo concentrations of the reactive metabolites was believed to measure the degree of mutation risk. This hypothesis could not be sustained for several reasons. First, there are several reactions that must occur before the ultimate formation of adduct between the reactive metabolites and nucleic acids. Second, by virtue of their chemistry, reactive oxidative metabolites are unstable, exhibit short half lives and are difficult to isolate. Efforts, therefore, focused on the isolation of the actual adducts between nucleic acids and the reactive metabolites as a true measure of the extent to which genetic material had been chemically modified.
The theoretical elegance of this concept was fully matched by the challenges in separation science. Requirements for sensitivity are very high, sample preparation methods are complex and characterization of the desired products is difficult. Radiolabelling offers the best approach with the most sensitive technique being 32P postlabelling. In this technique, hydrolysate of DNA is labeled by T4 polynucleotide kinase mediated transfer of the 32P radiolabelled terminal phosphate of ATP to the 5'-hydroxyl of a 3'-nucleotide. The adducted or alkylated nucleotides (AN) are then separated from normal nucleotides (NN) and from each other by thin layer chromatography (TLC), high pressure liquid chromatography (HPLC) or capillary zone electrophoresis (CZE).
While it is feasible to isolate and separate AN the identification of the structure presents a more serious difficulty. Typically there are only pico or femtomoles of materials available. With such small amounts the identification of structure is beyond the scope of instrumentation such as NMR that is typically used to characterize structure. Elucidation of reaction pathways and the enzymes that control those pathways is not possible if the structure of the end products is not known.
A compound that typifies the need for understanding the mechanisms of the reaction and its products is 1,3 butadiene (BD). This colourless gas is used in the rubber industry with annual production of 1.5 million tons in the United States and six million tons world wide. In addition, BD is produced gasoline powered vehicles, leaks and waste emission from manufacturing and burning of organic materials. Arguably, the entire human population is constantly exposed to low levels of BD. Understanding the metabolism of this compound is important but it is even more important to be able to assess the extent to which exposure of this compound can alkylate and modify DNA thus producing a mutation. The low levels of exposure put an additional stress on sensitivity in the measurement of these compounds.
1,3 butadiene also provides a unique probe of molecular mechanism of mutation. It is a simple molecule containing only four carbon atoms and two conjugated trans double bonds. The double bonds in BD are oxidized by the cytochrome P450 dependent mono-oxygenases 2E1 and 2A6 to produce an epoxide (BDE) which can exist in the R and S configuration. This three ringed functionality is highly reactive towards any nucleophile and so can alkylate DNA.
The authors had previously prepared the two enantiomers of BDE as well as the possible reaction products between BDE and nucleotides. This provided the basis for the current rigorous investigation of the mechanism for aklylating DNA presented in this paper.
In this study, adult rats weighing 190-230 g were exposed for five days to 200 ppm BD by inhalation for 6hr/day. DNA isolated from liver was hydrolyzed and the nucleotides were isolated by a solid phase sample extraction (SPE) which is essentially low resolution, very high capacity column chromatography. Two different SPE techniques were used. Chromatography on an ion exchange column separated nucleotides from the rest of the hydrolysate on the basis of charge. Absorption chromatography was then used to separate the more lipophilic AN from the NN. Subsequent 32P-postlabelling followed by TLC and HPLC provided the final separation, identification and quantitation of the AN.
This sample preparation was controlled both by external and internal standards. The external standards were the modified 3'-dGMPs and served to control for yield of 32P labeling. Modified but "cold" 5'-dGMPs were added after the labelling step and served as the optical markers and as internal standards. These internal standards controlled for losses during chromatographic separations which included both thin layer chromatography and HPLC. In the case of TLC the identification was carried out by autoradiography. The HPLC analysis was monitored by ultraviolet absorption and radioactivity detection. The modified 5'-dGMP added as an external standard acted as cold carriers for the AN produced in-vivo and provided sufficient mass for detection by ultraviolet absorption. In addition, HPLC analysis of the liver DNA digest provided good separation of NN and allowed determination 3'-UMP which demonstrated a very low (i.e.1%) contamination by RNA of the DNA isolate.
The relative standard deviation of recoveries of the entire procedure was ñ 28%. This is high by conventional standards of separation science but is acceptable given the complexity of the sample preparation (labelling, TLC separation, isolation from the TLC plate, HPLC) and the low concentrations (fmol). It should be noted that the without the use of external and internal standards it would not have been possible to achieve even this apparently high variability.
With this chemistry, the authors demonstrated, that following the exposure regimen described above, adducts were formed to the level of 7.2 fmol/10 ug DNA or 2.4 adducts/10-7 nn. They were able to characterize several features of these reaction pathways. Specifically they showed that:
i) the site of alkylation was the N-7 position of guanine;
ii) the predominant reaction was between DNA and the R-enantiomer of BDE;
iii) reaction at the 2 position of BDE also predominated.
The S enantiomer also formed in-vivo and other reaction at the 1 position also occurred these but these represented reaction pathways a smaller percentage of the overall reaction yields.
The possibility of measuring adduct formation in-vivo has import both for basic understanding of mechanism as well as practical application. This work demonstrates the importance of stereochemistry in understanding the mutagenicity and toxicity of even a relatively simple compound such as BD which is a planar molecule. During the oxidation of BD by cytochrome P450, both R and S stereochemistries should have been, and indeed may have been, formed in equal amounts. Despite this alkylation of DNA by the resulting reactive metabolite exhibit a fair degree of stereoselectivity for the R enantiomer. There are several possible explanations for this biological enantioselectivity. There could be selectivity in the detoxification of the S enantiomer. Stereoselectivity of the reaction between the BDE and DNA is also possible. Finally, there may have been differences in the repair mechanisms for DNA alkylated by the R or S enantiomer. This report cautions that further studies and models of quantitative structure-activity relationships must therefore take into account the steric factors at all stages of metabolism leading to the alkylation of DNA.
This work also demonstrates the extent of sample preparation necessary to obtain valid results. The solid phase extractions each using a different sorption mechanisms is an important technique in the purification method and gives a high degree of bulk separation prior to the labeling and analytical separation and quantitation of the AN. It also demonstrates the degree control that can be obtained with the proper use of external and internal standards.
On a practical level the knowledge of the structure of the adducts presents the possibility of using markers to identify the formation not only of the reactive metabolite but also the alkylation of DNA which is the ultimate mutagenic event. This would be useful in assessing risk both in the work place and in the population as whole.
We have established a protocol that allows qualitative and quantitative determination of butadiene monoepoxide- DNA adducts formed as a result of inhalation exposure to 1,3-butadiene. We observed that in this particular case in vivo samples required extensive sample purification to facilitate a low background. Sample preparation included a solid phase extraction carried out with a strong anion exchange column and one-dimensional ion exchange TLC. The ultimate analysis is based on reverse phase HPLC with on-line radioactivity and UV detectors. The qualitative identification and quantitation is based on characterized markers, which are used as external and internal standards. Modified 3'-dGMP markers were used to control labelling efficiency, which varies, and modified 5'-dGMP markers were used as an optical standard to qualitatively assign the products and to determine recovery of the sample preparation. Using this method we were able to demonstrate, for the first time, specific enantio- and regioisomeric adduct formation at the N7 position of guanine residues in liver DNA of rats inhalation-exposed to 1,3-butadiene. The major adduct formed was the C-2 isomer derived from the R enantiomer of butadiene monoepoxide, contributing 47% of all adducts formed at the N7 position of guanine. The relative proportions of the remaining three other adducts detected were 22 (R C-l), 18 (S C-2) and 14% (S C-l) respectively. Inhalation exposure to 200 p.p.m. for 5 days resulted in an alkylation level of 7.2 fmol/10 ug DNA or 2.4 adducts/10-7 normal nucleotides.