Mutational analysis of the liver, colon and kidney of Big Blue rats treated with 2-amino-3-methylimidazo[4,5-f]quinoline.

The heterocyclic aromatic amine (HAA) 2-amino-3-methylimidazo[4, 5-f]quinoline (IQ) induces intestinal tumours and hepatocellular carcinomas in rats, but no tumourigenic effects have been identified in the kidney. The tissue-specific mutagenicity of IQ was studied at the lacI locus in the liver, colon and kidney of Big Blue transgenic rats. At the highest dosing regime of IQ (20 mg/kg for 5 consecutive days) the mean mutant frequencies were significantly increased above background (P < 0.05) and were highest in the liver (12.9 +/- 6.2 x 10(-5)), followed by colon (7.4 +/- 1.4 x 10(-5)) and kidney (5.9 +/- 0.8 x 10(-5)). The mutational spectra from the livers of IQ-treated rats was statistically significantly different to that from the livers of control rats (P < 0.01). The lacI mutation spectra of the liver, colon and kidney from IQ-treated rats were similar. These were characterized by an increase in GC transversions in the liver and colon and an increase in the proportion of 1 bp G:C deletions in the liver and kidney. A single G deletion in the sequence 5'-CGGGA-3', characteristic of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine exposure, was detected in the liver and colon. A 2 bp GC deletion was identified at an identical position in the liver, colon and kidney. The colon was the only organ to contain two larger deletions of 13 and 33 bp. A preference was observed for base substitution mutations at guanine in the sequence 5'-CGC/T-3' and for 1 bp deletions at the guanine doublet in the sequence 5'-CGGA-3', especially in the liver and colon. Using the lacI gene as marker in the Big Blue rat model, the mutations identified in the IQ spectra have similarities to those identified for other HAAs studied in the same experimental system, but not to mutations identified in IQ-induced tumours.

Nucleotide excision repair modulates the cytotoxic and mutagenic effects of N-n-butyl-N-nitrosourea in cultured mammalian cells as well as in mouse splenocytes in vivo.

The butylating agent N-n-butyl-N-nitrosourea (BNU) was employed to study the role of nucleotide excision repair (NER) in protecting mammalian cells against the genotoxic effects of monofunctional alkylating agents. The direct acting agent BNU was found to be mutagenic in normal and XPA mouse splenocytes after a single i.p. treatment in vivo. After 25 and 35 mg/kg BNU, but not after 75 mg/ kg, 2- to 3-fold more hprt mutants were detected in splenocytes from XPA mice than from normal mice. Using O6-alkylguanine-DNA alkyltransferase (AGT)-deficient hamster cells, it was found that NER-deficient CHO UV5 cells carrying a mutation in the ERCC-2 gene were 40% more mutable towards lesions induced by BNU when compared with parental NER-proficient CHO AA8 cells. UV5 cells were 1.4-fold more sensitive to the cytotoxic effects of BNU compared with AA8 cells. To investigate whether this increased sensitivity of NER-deficient cells is modulated by AGT activity, cell survival studies were performed in human and mouse primary fibroblasts as well. BNU was 2.7-fold more toxic for mouse XPA fibroblasts compared with normal mouse fibroblasts. Comparable results were found for human fibroblasts. Taken together these data indicate that the role of NER in protecting rodent cells against the mutagenic and cytotoxic effects of the alkylating agent BNU depends on AGT.

Elevated frequencies of benzo(a)pyrene-induced Hprt mutations in internal tissue of XPA-deficient mice.

Xeroderma pigmentosum (XP) patients are hypersensitive to sunlight and have a high predisposition to developing cancer. At the cellular level, XP patients are defective in nucleotide excision repair (NER). Recently, mice have been generated via gene targeting that are deficient in the expression of the XPA gene [A. de Vries et al., Nature (Lond.), 377: 169-173, 1995]. We have assessed the consequences of defective NER for mutagenesis in normal and XPA mice exposed to benzo(a)pyrene and 2-acetylaminofluorene. To study mutagenesis, mature T lymphocytes were isolated from the spleen and stimulated to proliferate in vitro to select for mutants at the endogenous Hprt locus. Background mutant frequencies in normal and XPA mice were very similar and not influenced by age. Single doses of benzo(a)pyrene administered i.p. resulted in a dose-dependent increase of the Hprt mutant frequency in normal mice. In addition, after chronic exposure to benzo(a)pyrene, Hprt mutants were readily detectable in XPA mice at an early onset of treatment but only at a later stage in normal mice. In contrast, chronic treatment of either normal or XPA mice with 2-acetylaminofluorene did not increase Hprt mutant frequency above the background frequency. This absence of significant induction of Hprt mutants can be entirely attributed to the low frequency of 2-acetylaminofluorene-induced DNA adducts in lymphoid tissue. These results provide the first direct evidence in mammals that deficient NER leads to enhanced mutagenesis in endogenous genes in internal tissue after exposure to relevant environmental mutagens, such as benzo(a)pyrene.

Electrochemical detection and quantification of the acetylated and deacetylated C8-deoxyguanosine DNA adducts induced by 2-acetylaminofluorene.

The genotoxic agent 2-acetylaminofluorene induces, upon metabolic activation, two main types of DNA adducts in animal tissue, i.e., (deoxyguanine-8-yl)-aminofluorene (dG-C8-AF) and N-(deoxyguanine-8-yl)-acetylaminofluorene (dG-C8-AAF). Quantification of the frequency of these adducts usually relies on the use of radioactively labeled 2-acetylaminofluorene. Here, we report the development of a sensitive, non-radioactive method for the quantification of dG-C8-AF and dG-C8-AAF. Essentially, the modified DNA bases are separated by high-performance liquid chromatography (HPLC) and quantified by electrochemical detection. We established that both modified bases guanine-C8-aminofluorene and guanine-C8-acetylaminofluorene are electrochemically active. Subsequently, a procedure was developed to quantify dG-C8-AF and dG-C8-AAF in genomic DNA. Following DNA hydrolysis the adducted bases were extracted by ethyl acetate, separated by HPLC, and detected electrochemically. This procedure has been applied in the analysis of dG-C8-AAF in N-acetoxy-2-acetylaminofluorene-modified calf thymus DNA and in the detection of dG-C8-AAF and dG-C8-AF in liver DNA of mice injected intraperitoneally with 150-450 mg N-hydroxy-2-acetylaminofluorene/kg. The quantification of relatively low dG-C8-AF and dG-C8-AAF adduct levels (i.e., 0.1-1 adduct/10(6) nucleotides) in mouse liver DNA demonstrates the sensitivity of this electrochemical detection procedure. The detection limit of the method is 1 adduct per 10(6) nucleotides for both adducts using 20 micrograms of DNA and 4 adducts per 10(8) nucleotides using 500 micrograms DNA.

Cloning and characterization of the Drosophila homolog of the xeroderma pigmentosum complementation-group B correcting gene, ERCC3.

Previously the human nucleotide excision repair gene ERCC3 was shown to be responsible for a rare combination of the autosomal recessive DNA repair disorders xeroderma pigmentosum (complementation group B) and Cockayne's syndrome (complementation group C). The human and mouse ERCC3 proteins contain several sequence motifs suggesting that it is a nucleic acid or chromatin binding helicase. To study the significance of these domains and the overall evolutionary conservation of the gene, the homolog from Drosophila melanogaster was isolated by low stringency hybridizations using two flanking probes of the human ERCC3 cDNA. The flanking probe strategy selects for long stretches of nucleotide sequence homology, and avoids isolation of small regions with fortuitous homology. In situ hybridization localized the gene onto chromosome III 67E3/4, a region devoid of known D.melanogaster mutagen sensitive mutants. Northern blot analysis showed that the gene is continuously expressed in all stages of fly development. A slight increase (2-3 times) of ERCC3Dm transcript was observed in the later stages. Two almost full length cDNAs were isolated, which have different 5' untranslated regions (UTR). The SD4 cDNA harbours only one long open reading frame (ORF) coding for ERCC3Dm. Another clone (SD2), however, has the potential to encode two proteins: a 170 amino acids polypeptide starting at the optimal first ATG has no detectable homology with any other proteins currently in the data bases, and another ORF beginning at the suboptimal second startcodon which is identical to that of SD4. Comparison of the encoded ERCC3Dm protein with the homologous proteins of mouse and man shows a strong amino acid conservation (71% identity), especially in the postulated DNA binding region and seven 'helicase' domains. The ERCC3Dm sequence is fully consistent with the presumed functions and the high conservation of these regions strengthens their functional significance. Microinjection and DNA transfection of ERCC3Dm into human xeroderma pigmentosum (c.g. B) fibroblasts and group 3 rodent mutants did not yield detectable correction. One of the possibilities to explain these negative findings is that the D.melanogaster protein may be unable to function in a mammalian repair context.