High-fidelity amplification using a thermostable DNA polymerase isolated from Pyrococcus furiosus.

A thermostable DNA polymerase which possesses an associated 3'-to-5' exonuclease (proofreading) activity has been isolated from the hyperthermophilic archaebacterium, Pyrococcus furiosus (Pfu). To test its fidelity, we have utilized a genetic assay that directly measures DNA polymerase fidelity in vitro during the polymerase chain reaction (PCR). Our results indicate that PCR performed with the DNA polymerase purified from P. furiosus yields amplification products containing less than 10% of the number of mutations obtained from similar amplifications performed with Taq DNA polymerase. The PCR fidelity assay is based on the amplification and cloning of lacI, lacO and lacZ alpha gene sequences (lacIOZ alpha) using either Pfu or Taq DNA polymerase. Certain mutations within the lacI gene inactivate the Lac repressor protein and permit the expression of beta Gal. When plated on a chromogenic substrate, these LacI- mutants exhibit a blue-plaque phenotype. These studies demonstrate that the error rate per nucleotide induced in the 182 known detectable sites of the lacI gene was 1.6 x 10(-6) for Pfu DNA polymerase, a greater than tenfold improvement over the 2.0 x 10(-5) error rate for Taq DNA polymerase, after approx. 10(5)-fold amplification.

The use of selection in recovery of transgenic targets for mutation analysis.

Transgenic animal mutagenesis assays using lambda shuttle vectors have recently been described for isolation and characterization of spontaneous and chemical induced DNA mutations. Extensive information on lambda and E. coli genetics provides a wealth of techniques to allow selection of mutant target genes. Here we describe the modification of an E. coli host which permits two methods for the direct selection of mutant genes. These methods reduce the number of plates needed to be screened for a comparable amount of frequency data by 20-100-fold and thus provide a significant savings of the materials and time required for the screening of mutations. In addition, mutants selected by these approaches described here may alter or broaden the spectrum of mutations that are recoverable. Ultimately, a combination of selective and nonselective techniques may prove valuable for the analysis of mutations produced in vivo in transgenic animals.

Transgenic systems for in vivo mutation analysis.

Transgenic mice carrying shuttle vectors containing the lacI gene as the target permit the in vivo measurement of mutations in multiple tissues and have been used to test the mutagenic effects of several compounds. Tissue-specific and time-dependent responses have been observed, and the spectrum of mutations determined by sequencing allows analysis of the role of expression time in mutagenesis. The results obtained from sequencing analysis have demonstrated spectra paralleling those observed in alternative in vivo assays. In addition to color screening, modifications to this system have permitted direct selection for mutations in the lacI target by a variety of methods. Transgenic rats containing the same lambda/lacI shuttle vector have been developed for inter-species comparison of mutagenesis testing results, which may offer a better understanding of the specific mechanisms involved in mutagenesis at the molecular level in vivo.

Spontaneous mutations in lacI-containing lambda lysogens derived from transgenic mice: the observed patterns differ in liver and spleen.

The pattern of somatic mutation observed in tumor suppressor genes, such as the p53 gene, vary dramatically with tumor type. Some of the observed differences are due to tissue specific effects of mutagens, but it is also possible that some differences may reflect the tissue/cell type specificity of spontaneous mutation. Transgenic mouse models with recombinant shuttle vectors containing the lacI or lacZ target genes may shed light on the extent to which spontaneous mutation displays tissue specificity. Herein we utilize a recently described selectable system to obtain spontaneous mutants for analysis of the molecular lesions. Spontaneous mutations were isolated in the lacI gene recovered from five transgenic mice carrying a lambda shuttle vector. Seventy-three and 67 independent mutations derived from liver and spleen DNA, respectively, were defined in the amino terminal region of lacI. Although technical barriers preclude a direct assessment of the E. coli derived pattern of mutation in this system, five pieces of circumstantial evidence suggest that many of the mutations arose in mouse rather than in E. coli. In DNA from both liver and spleen, mutations at CpG dinucleotides predominate (58% and 51%, respectively). In spleen, most of the mutations at CpG are transitions, while in liver most are transversions. In addition, liver has a higher frequency of GC-->TA transversions at non-CpG dinucleotides while spleen had a higher frequency of deletions and insertions. The data provide evidence that the spontaneous pattern of mutation is tissue specific. In addition, the high frequency of transversions at CpG suggests the need to reevaluate the mechanisms by which mutations occur at this methylated dinucleotide.