How precisely can data from transgenic mouse mutation-detection systems be extrapolated to humans?: lesions from the human factor IX gene

Sequence variation in the human uracil-DNA glycosylase (UNG) gene

Spontaneous deamination of cytosine results in a premutagenic G:U mismatch that may result in a GC-->AT transition during replication. The human UNG-gene encodes the major uracil-DNA glycosylase (UDG or UNG) which releases uracil from DNA, thus, initiating base excision repair to restore the correct DNA sequence. Bacterial and yeast mutants lacking the homologous UDG exhibit elevated spontaneous mutation frequencies. Hence, mutations in the human UNG gene could presumably result in a mutator phenotype. We screened all seven exons including exon-intron boundaries, both promoters, and one intron of the UNG gene and identified considerable sequence variation in cell lines derived from normal fibroblasts and tumour tissue. None of the sequence variants was accompanied by significantly reduced UDG activity. In the UNG gene from 62 sources, we identified 12 different variant alleles, with allele frequencies ranging from 0.01 to 0.23. We identified one variant allele per 3.8kb in non-coding regions, but none in the coding region of the gene. In promoter B we identified four different variants. A substitution within an AP2 element was observed in tumour cell lines only and had an allele frequency of 0.10. Introduction of this substitution into chimaeric promoter-luciferase constructs affected transcription from the promoter. UDG-activity varied little in fibroblasts, but widely between tumour cell lines. This variation did not however correlate with the presence of any of the variant alleles. In conclusion, mutations affecting the function of human UNG gene are seemingly infrequent in human tumour cell lines.

Spontaneous mutations in the Big Blue transgenic system are primarily mouse derived

The Big Blue transgenic mouse mutation detection system provides a powerful approach for measuring spontaneous and induced mutations in vivo. The observed mutations may contain a fraction of ex vivo or prokaryotic mutational events. Indeed, a modified, selectable form of the Big Blue assay seem to generate artifactual mutants under certain circumstances. Herein we review the evidence that circular mutants (i.e., the plaque circumference is at least 50% blue) collected in the standard Big Blue assay are derived primarily from the mouse. The most direct evidence is the similarity in the types of mutations found in jackpot and nonjackpot mutations. In addition, about half of the spontaneous mutations in the lacI transgene are transitions and transversions at CpG dinucleotides, a mammalian-specific feature. The mutation pattern observed at lacI is consistent with AT mutation pressure operating in a GC rich DNA and approaches that reported for observed germline human factor IX mutations. Furthermore, the spontaneous mutation pattern of circular Big Blue mutants differs significantly from that of an endogenous lacI gene in E. coli. Pinpoint mutants (a dot of blue color peripherally located in a wild type plaque), which a priori were not expected to be mouse-derived, have a mutation pattern consistent with the mutation pattern of an endogenous E. coli lacI gene. Analysis of induced mutagenesis studies reveals mutation frequencies and patterns for the Big Blue circular mutants which are comparable to endogenous genes. In reconstruction experiments, blue plaques derived from a superinfection with wild type and mutant phage produced approximately 50% blue and 50% clear plaques on replating. This phenomenon has not been seen when plaques derived from mouse were replated in the Big Blue assay. Collectively, the evidence strongly supports a murine origin for circular mutants recovered in the standard Big Blue assay. Validation of current assays is an essential step in determining the frequency and pattern of spontaneous murine-specific mutations. Defining this benchmark will be helpful in evaluating the next generation of transgenic mutation detection systems.

Mutation analysis using the restriction site mutation (RSM) assay

The restriction site mutation (RSM) assay (see Steingrimsdottir et al. [H. Steingrimsdottir, D. Beare, J. Cole, J.F.M. Leal, T. Kostic, J. Lopez-Barea, G. Dorado, A.R. Lehmann, Development of new molecular procedures for the detection of genetic alteration in man, Mutat. Res. 353 (1996) pp. 109-121] for a review) has been developed as a genotypic mutation detection system capable of identifying mutations occurring in restriction enzyme sites of genomic DNA. Here we will report the steps taken to overcome some of the initial problems of the assay, namely the lack of quantitative data and limited sensitivity, the aim being to achieve a methodology suitable for the study of low dose chemical exposures. Quantitative data was achieved in the RSM assay by the inclusion of an internal standard molecule in the PCR amplification stage, thus allowing the calculation of both spontaneous and induced mutation frequencies. The sensitivity of the assay was increased through the discovery that intron sequences of genomic DNA accumulated more mutations in vivo compared to the exons, presumably due to differential selective pressure within genes [G.J.S. Jenkins, I.deG. Mitchell, J.M. Parry, Enhanced restriction site mutation (RSM) analysis of 1, 2-dimethylhydrazine-induced mutations, using endogenous p53 intron sequences, Mutagenesis 12 (1997) pp. 117-123]. This increased sensitivity was examined by applying the RSM assay to analyse the persistence of N-ethyl-N-nitrosourea (ENU)-induced mutations in mice testes. Germ line mutations were sought in testes DNA 3, 10 and 100 days after ENU treatment. Mutations were detected in exons and especially intron regions, the intron mutations were more persistent, still being detected 100 days post-chemical treatment. Assignment of these mutations as ENU induced was complicated in some cases where the spontaneous mutation level was high. This theme of mutation persistence was further investigated by studying the presence of 4-nitroquinoline-1-oxide (4-NQO)-induced DNA mutations in vitro. This study also analysed the relationship between DNA adduct formation and DNA mutation induction by the concurrent RSM analysis and 32P post-labelling analysis of 4-NQO treated human fibroblasts. The results demonstrated that early DNA mutations detected 4 days post-treatment by the RSM assay were probably ex vivo mutations induced by Taq polymerase misincorporation of 4-NQO adducted DNA, due to the maximum levels of 4-NQO adducts being present at this time point. A later mutational peak, after the adduct level had declined, was assumed to be due to DNA sequence changes produced in the fibroblasts by the in vivo processing of DNA adducts.

Strategies and applications of DNA level diagnosis in genetic diseases: past experiences and future directions

The development of technologies towards the detection of mutations represents one of the most important areas of molecular biology. It has played a pivotal role in the tremendous success of the elucidation of complex biological problems, including genetic diseases. Today, these proven and emerging technologies have become the basis of successful biological investigations. More importantly, they are expected to play a central role in medicine, particularly the diagnosis and prognosis of genetic diseases including genetic predispositions, the assessment of treatments including transplants and decisions on reproductive choices. In addition, these technologies hold the key to future breakthroughs. This review provides an up-to-date examination of the principles of genetic diseases, the theories behind current methods of genetic diagnosis and detection of mutations including strategies for modification and the development of future technologies as they impact on the practice of medicine and on society as a whole.

Spontaneous mutation frequencies and spectra in p53 (+/+) and p53 (-/-) mice: a test of the 'guardian of the genome' hypothesis in the Big Blue transgenic mouse mutation detection system

TSG-p53/Big Blue double transgenic mice offer a powerful tool for examining the effect of a p53 germline mutation on spontaneous somatic mutation in vivo. After sequencing the DNA-binding domain of the lacI gene, we previously reported no differences in mutant frequency between p53 nullizygous (-/-) and p53 wild-type (+/+) mice in liver, spleen and brain. However, jackpot mutations elsewhere in the gene may have obscured a real difference in mutation frequency and the small sample size of mutations not at CpG dinucleotides (n = 23) may have been insufficient to reveal differences in mutation spectra. Herein we have sequenced the entire lacI gene, including the promoter and lacZ operator regions. 123 additional independent mutations have been found including 70 mutations not at CpG sites. The mutation frequency was determined by correcting for jackpot mutations. There were no statistically significant differences in mutation frequency or spectrum between the p53 (+/+) and p53 (-/-) genotypes in any of the three tissues.

The factor IX gene as a model for analysis of human germline mutations: an update

The variation generated by germline mutation is essential for evolution, but individuals pay a steep price in the form of Mendelian disease and genetic predisposition to complex disease. Indeed, the health of a species is determined ultimately by the rate of germline mutation. Analysis of the factor IX gene in patients with hemophilia B has provided insights into the human germline mutational process. Herein, seven topics will be reviewed with emphasis on recent advances: (i) proposed mechanisms of deletions, inversions, and insertions; (ii) discordant sex ratios of mutation and associated age effects; (iii) somatic mosaicism; (iv) founder effects; (v) mutation rates; (vi) the factor IX gene as a germline mutagen test; and (vii) cancer as a possible mechanism for maintaining a constant rate of germline mutation.

A selectable system for mutation detection in the Big Blue lacI transgenic mouse system: what happens to the mutational spectra over time

Transgenic animals offer a powerful tool to study the mechanisms of spontaneous and induced mutagenesis in vivo. Herein we used a test version of a growth selectable assay to obtain spontaneous mutants in a lacI target transgene recovered from lacI transgenic B6C3F1 mice (Big Blue). This selection system may have certain advantages relative to the more established plaque screening system for mutation detection because: (1) the plating density of the phage is up to 60 times higher in the selectable assay, reducing the number of plates needed to be screened for a comparable amount of mutants; and (2) the mutant frequency obtained from the selectable assay is higher compared to the plaque assay, possibly due to a higher sensitivity for weaker mutants. However, the longer incubation time of the growth selectable assay might allow E. coli host derived mutants to appear. To address this issue, we investigated the sequence changes in the amino-terminal domain of the lacI gene of 405 mutants derived from the liver, spleen, brain, germ cells and skin of five untreated 6-week-old mice. The mutant colonies were isolated after 60, 84, 108 and 150 h of incubation under growth selectable conditions. Tissue-specific differences in the mutational pattern obtained after 60 and 84 h disappear after a longer time of incubation, possibly due to an increasing contribution of E. coli derived mutants. The evolving selectable systems offer the potential to increase screening efficiency, but the results suggest caution in interpreting data from this system because repair by E. coli of DNA lesions or mismatched heteroduplexes either originating in mouse in vivo or produced by ex vivo manipulation as well as de novo mutations in E. coli might contribute significantly to the observed mutational spectra at each timepoint.

Spontaneous mutation in Big Blue transgenic mice: analysis of age, gender, and tissue type

A lacI-containing transgenic mouse mutation detection system (Big Blue) was used to determine the frequency and spectrum of spontaneous mutation in two rapidly dividing tissues (male germ cells and thymus) and one slowly dividing tissue (brain) at 3 and 10 months of age. By screening 9.4 million lambda plaques, a total of 343 circular mutant plaques were recovered from the three tissues. The mutation frequencies and spectra were determined by sequencing the lacI gene and associated lacZ operator in all samples and correcting for "jackpot" mutations. The mutation frequencies and spectra were similar in all three tissues and there were no age-dependent or gender-dependent changes. When the mutation spectrum in each tissue was compared by utilizing large numbers of independent mutations (average: 75 per tissue), there was evidence for small tissue-specific differences. The spectrum of "jackpot" mutations, which clearly represents in vivo mouse-derived mutations, was similar to that of nonjackpot mutations, providing additional evidence that observed mutations occur in mouse. In the aggregate, the results suggest that there is: (i) a core mutation frequency and spectrum that is modified weakly by tissue-specific metabolism, and (ii) a steady-state level of spontaneous mutation in adult mice reflecting the balance between the accumulation of new mutations and the elimination of mutated cells by either selection against suboptimal cellular function or apoptosis triggered by accumulated DNA damage.

Factors affecting somatic mutation frequencies in vivo

The factors that influence the spontaneous mutant frequencies in mammalian tissues have been ranked on the basis of data from our laboratory together with published data. Some of the data come from the endogenous hprt and Dlb-1 loci, but most come from transgenic mice carrying the bacterial lacI and lacZ genes in recoverable lambda phage vectors. Since there is evidence that these bacterial loci are selectively neutral, the mutant frequency observed is the integral of the mutation rates from the formation of the zygote. The factors that affect the inferred mutation rate, in decreasing order of importance are: site of integration of the transgene, age, tissue, and strain. Insufficient data exist to determine the influence of gender (probably small) and inter-laboratory variables (probably at least as important as age). The two most surprising results are (1) that about half of all mutations arise during development (and half of these in utero) and (2) that most somatic tissues, whether queiscent or actively proliferating, have similar mutant frequencies and similar increases during adult life.

Recent human germ-line mutation: inferences from patients with hemophilia B

The gene encoding factor IX has a unique number of advantages for studying human germ-line mutations. Detailed analyses of the observed mutations of this gene, with special attention to the biases in the data, have provided information on mutational hotspots (including 'cryptic' dinucleotide repeats), mutation rates per base pair per generation, and the sex ratios of mutation. The evidence strongly suggests that the great majority of germ-line mutations result from endogenous processes, rather than exogenous mutagens. Perhaps nature does not permit environmental control of such an important process. Instead, the rate of germ-line mutation is placed under selective pressure, of which early-onset cancer may be an important mediator.