The Genome Recombination/Regulation Section focuses on three topics: functions that control the stability of genomes, functions that control aspects of gene expression, and the development of a yeast based system to study HIV-1. Studies on genome instability. We are continuing our analysis of how DNA palindromes are generated. These head to head DNA sequences are highly unstable. Some tumor cells undergo gene amplification by unknown mechanisms that generate palindromes. The instability of these sequences contributes to the genome rearrangements that occur in tumors. Because palindromes are unstable in bacteria, it is it nearly impossible to clone them. Similarly, the secondary structures that can be adopted by palindromic DNAs make them very difficult to sequence. We opened the field of research on the origin of DNA palindromes by making progress in three important areas related to the study of palindromes. First, we identified yeast strains that tolerate palindromes. Second, we developed a method that allows us to sequence palindromic DNAs. Third, we developed a recombination substrate that generates palindromes and identified a class of recombinants that is almost exclusively palindromes. We demonstrated that the palindromes are formed in our system by a novel kind of nonhomologous end joining (NHEJ) which is independent of some of the recombination functions that are required for most NHEJ events. We recently demonstrated that we can isolate palindromic sequences from mammalian genomes, opening the door to the analysis of palindromes found in normal and malignant cells. The fidelity of transcription. We developed methods to monitor the fidelity of transcription and the functions that contribute to the accuracy of that process. One such method involves monitoring the fidelity of retrotransposition. We isolated mutations in a subunit of RNA polymerase that reduce the fidelity of retrotransposition and demonstrated that they directly affect the accuracy of transcription. These are the first eukaryotic mutations known to reduce the fidelity of transcription. We also developed a screen for RNA polymerase mutants that increase the frequency of slippage during transcription. This class of transcription error has been shown to occur in bacteria and humans, but the features that avoid such errors have not been determined. With that screen we isolated the first mutations that increase errors of this type. We are investigating the biological consequences of increased transcription error rates. HIV-1 Reverse Transcriptase. We developed hybrid Ty/HIV-1 elements (TyHRT) that have HIV-1 RT substituted for the Ty1 RT . These elements replicate in yeast and are dependent on the polymerase and RNase activities of HIV-1 RT. Replication of the TyHRT elements is sensitive to the non-nucleoside class of HIV-1 RT inhibitors (NNRTI) while TyHRT elements containing HIV-1 RT domains from resistant viruses are themselves resistant. We used TyHRT elements to detect RT domains that confer resistant to the FDA approved NNRTIs efavirenz and nevirapine among the viral sequences recovered from patient blood. These studies allowed increased sensitivity in studies that monitor the persistence of NNRTI resistant viruses in patients that have been treated with single doses of nevirapine. In addition to detecting very low levels of well known NNRTI resistant variants, we have been able to detect novel variants with moderate levels of NNRTI resistance which may have a role in the evolution of resistant HIV. We are also characterizing Ty/HIV-1 variants that have elevated mutation rates as candidates for variants of HIV-RT with decreased fidelity. In parallel, we have carried out a mutagenesis of a canonical RT and looked for RT fidelity variants using a screen that measures both reverse transcriptase activity and mutation frequency.
