Ionic and radical oxidations of DNA by hydrogen peroxide.

Hydrogen peroxide mediated oxidation of 2'-deoxyadenosine and isolated DNA was investigated. Reaction of hydrogen peroxide with 2'-deoxyadenosine under non radical conditions led to the formation of a predominant decomposition product. This was identified as 2'-deoxyadenosine N-1-oxide on the basis of detailed 1H and 13C NMR analysis and further confirmed by photolysis experiments. Quantitative determination of both radical and ionic DNA type damage was based on the use of a 32P-postlabeling method (5-hydroxymethyluracil and adenine N-1-oxide) and of an HPLC-EC assay (8-hydroxyguanine). Adenine N-1-oxide was shown to be the predominant ionic DNA base damage under non radical conditions. The presence of Fe(II)-DTPA complex in the reaction medium led to a reduction in the amount of adenine N-1-oxide by a factor of 4 whereas radical DNA type damages including 8-hydroxyguanine and 5-hydroxymethyluracil was increased by a factor of 2-3.

Chemical and biochemical postlabeling methods for singling out specific oxidative DNA lesions.

A survey of the main available chemical and biochemical postlabeling assays for measuring oxidative DNA damage is reported. Two main approaches, radio and fluorescent postlabeling, have been used in order to reach a high level of sensitivity of detection. This is required for the measurement of DNA damage within cells and tissues upon exposure to agents of oxidative stress. Most of the methods are based on liquid chromatographic separation of defined DNA modifications following either acidic hydrolysis or enzymic digestion of DNA. In a subsequent step, the isolated base or sugar damages are either radiolabeled or made fluorescent by chemical or enzymatic reactions. Emphasis is placed on the recently developed high performance liquid chromatographic 32P-postlabeling assay, which allows the specific and sensitive measurement of various base damages including adenine N-1 oxide and 5-hydroxymethyluracil at the level of one modification per 10(7) normal bases in a sample size of 1 microgram of DNA. Examples of application of radioactive postlabeling to the measurement of DNA base damage following exposure of human cells to oxidizing agents including hydrogen peroxide and UVA radiation are provided.

32P-postlabeling measurement of adenine N-1-oxide in cellular DNA exposed to hydrogen peroxide.

A 32P-postlabeling assay has been developed for monitoring the formation within DNA of adenine N-1-oxide, the specific H2O2-mediated oxidation product under nonradical conditions. This has required the chemical synthesis of both 2'-deoxyadenosine N-1-oxide 3'-monophosphate and 2'-deoxyadenosine N-1-oxide 5'-monophosphate, the substrate and the product of polynucleotide kinase mediated phosphorylation. Isolation of the substrate from the other nucleotides was found to be necessary in order to improve the rate of phosphorylation and to prevent self-radiolysis processes. [32P]-2'-deoxyadenosine N-1-oxide 5'-monophosphate was obtained after successive ion exchange and reverse-phase HPLC and was characterized by a microreaction. The sensitivity of the assay, which is close to 1 modified adenine N-1-oxide/10(6) bases, allowed the determination of this lesion within the DNA of cells exposed to nonlethal levels of H2O2.

Sequence analysis of the CCG polymorphic region adjacent to the CAG triplet repeat of the HD gene in normal and HD chromosomes.

The CAG expansion responsible for Huntington's disease (HD) is followed by an adjacent polymorphic CCG repeat region which may interfere with a PCR based diagnosis. We have sequenced this region in 52 unrelated HD patients, from both normal and HD chromosomes. Fifty percent of the normal alleles were (CCG)7(CCT)2, 48% (CCG)10(CCT)2, and 2% (CCG)7(CCT)3. In contrast (CCG)7(CCT)2 was found in 85% of the HD alleles which represents significant linkage disequilibrium with the HD mutation.

Measurement of oxidative base damage to DNA by using HPLC-32P-postlabelling and GC/MS-selective ion monitoring assays.

A 32P-postlabelling assay has been developed for singling out specific oxidized base lesions. Emphasis was placed on the quantitative aspect and the accuracy of the assay, which require the use of calibration curves and microreactions, respectively. The method was successfully applied to the detection and the measurement of adenine N1-oxide and 5-hydroxymethyluracil in cells exposed to agents inducing oxidative stress including H2O2 and UV-A radiation. The sensitivity of the assay allows the detection of one lesion in 10(6) normal bases in 1 microgram of DNA. The GC/MS method when coupled to the selective ion monitoring (SIM) technique is about twenty times less sensitive, even for suitable substrates such as 5-hydroxymethyluracil and 5-hydroxyuracil, than the 32P-postlabelling assay. However, the former assay is much easier to apply, even though a derivatization step is necessary, and provides unambiguous structural information on the compound to be measured. Accurate quantitative measurements can be obtained when stable, isotopically labelled standards are available.

Huntington's disease in French families: CAG repeat expansion and linkage disequilibrium analysis.

The molecular defect causing Huntington's disease (HD) has been found as an expansion of CAG triplets in the 5' coding region of IT15 gene. In the 29 French families reported, the HD disease is due to the expansion of the CAG triplets region above 38 copies. The complete sequencing of 10 HD alleles PCR products allowed us to confirm that expansion is restricted to the CAG repeat region and does not extend to the adjacent CCG repeat region which is also present in the PCR product. Then, we analysed linkage disequilibrium between the molecular defect and 6 DNA markers mapping to the 4p16.3 region. The most striking finding in this study is the presence of a strong linkage disequilibrium between HD and D4S127 (PvuII), D4S95 (AccI, MboI, TaqI) located in a region of 130 kb distal to IT15 gene. Two major haplotypes, comprising D4S127 (PvuII) and D4S95 (MboI, AccI) polymorphic sites, were found in the normal population as only one was found associated with HD alleles. This result can be interpreted either as an evidence for a rather recent founder effect or as several independent mutations occuring in chromosomes bearing the same haplotype.