Proc. Natl. Acad. Sci. USA Vol. 91, pp. 12711-12715, December 1994 Genetics Transcriptional induction of Ty recombination in yeast (repeated sequences/ectopic gene conversion/DNA repair/RAD genes) YAEL NEVO-CASPI AND MARTIN KuPIEC* Department of Molecular Microbiology and Biotechnology, Tel Aviv University, Ramat Aviv 69978, Israel Communicated by Roger D. Kornberg, September 6, 1994 (receivedfor review February 18, 1994) ABSTRACT Families of repeated sequences are present in of transcription (6). Evidence for induced recombination the genomes of all eukaryotes. Little is known about the upon transcription was also obtained in Schizosaccharomy- mechanism(s) that prevents recombination between repeated ces pombe (7) and in mammalian cells (8-11). sequences. In the yeast Saccharomyces cerevisiae, recombina- A close relationship between transcription and DNA repair tion between homologous sequences placed at nonhomologous has been found: Lesions in actively transcribed genes are locations in the genome (ectopic recombination) has been repaired faster than those in inactive ones and the template shown to occur at high frequencies for artificially created strand of actively transcribed genes is preferentially repaired repeats, but at relatively low frequencies for a natural family (12-15). In addition, it was also recently found that the DNA of repeated sequences, the Ty family. We have previously repair genes RAD3, RAD25, and SSLI encode proteins that shown that a high level ofTy cDNA in the cell causes an increase play a role also in transcription (16, 17). in the rate of nonreciprocal recombination (gene conversion) of In this paper we present evidence that elevated transcrip- a marked Ty element. In the present study, we show that it is tion of a Ty element causes an increase in its rate of also possible to elevate the rate of recombination of a marked recombination, and we investigate the role of different RAD Ty by increasing its transcription. This induction is different genes in this process. from, and acts synergistically to, the one seen upon increased levels of donor Ty cDNA. We show that the induction by transcription does not require the products of the RADSO, MATERIALS AND METHODS RADS5, and RAD57 genes. In contrast, cDNA-mediated re- Yeast strains. All the Saccharomyces cerevisiae strains combination is dependent on the product ofthe RADSI gene but used in the present study are isogenic and were derived from not on products of the genes RAD5O or RAD57. strain MK89 (4) (MATa ura3-Nco- his3-11,15 leu2-3,112 trpl-Xba- can.1-101) by transformation. They carry, at the Ty retrotransposons constitute the main family of naturally L YS2 locus, a Ty element marked by the insertion ofa URA3 occurring repeated sequences in the yeast genome, repre- gene (TyUra). Strains MK87 and MK104 have been de- senting 1-2% of the genome (for review, see ref. 1). Ty scribed (4); MK87 carries a Ty2Ura, and MK104 carries a elements are retrovirus-like transposons that code for an TylUra. MK95, MK116, and YN2 were created by trans- end-to-end transcript that is reverse-transcribed to cDNA by forming MK89 with plasmids pM106, pM128, and pYN10, Ty-encoded proteins. They transpose at very low frequencies respectively. In MK95, the first 237 nt of the Tyl have been (10-9 to 10-7 per locus per cell division). The two main replaced by 741 bp from the GAL] promoter (GalTylUra) classes of Ty elements, Tyl and Ty2, share extensive ho- (18); in MK116, a frameshift mutation was introduced in the mology. There are -30-40 copies ofTy elements per haploid GalTylUra by filling-in the Asp718 site of the Ty; in YN2, a genome (1). linker containing amber stop mutations in all three frames Recombination between homologous sequences located at was inserted at nt 475 ofthe Ty. YN3 carries a deletion ofthe nonhomologous positions in the genome (ectopic recombi- Ty promoter and was created by transformation with plasmid nation) occurs readily between artificially duplicated genes, pYN9. Strains MK172 (rad50:: TRPl), UB1, (rad51::LEU2), in mitotic and meiotic yeast cells. Both reciprocal and and UB2 (rad57: :LEU2) were created by a one-step replace- nonreciprocal recombination (gene conversion) have been ment transformation (19) using plasmids pME305 (20), observed. Ectopic reciprocal recombination causes chromo- pAM28 (21), and pSM51 [a Pvu II-Sal I fragment of the somal aberrations, such as translocations, deletions, and RAD57 gene replaced by the LEU2 marker (22)], respec- inversions (2). Since repetitive sequences are present in the tively. After transformation, all the relevant chromosomal genomes of all eukaryotes, the karyotypic stability might be configurations were confirmed by Southern blot analysis. maintained by (a) mechanism(s) that prevent(s) recombina- Plasmids. pM106, carrying the lys2::GalTyl Ura allele was tion between such repeats. We have been using the Ty family constructed in several steps: First, the 1.2-kb URA3 fragment of retrotransposons as a model to study recombination be- was cloned into the 3'Bgl II site ofa GalTy (18), thus creating tween naturally occurring repeated sequences. The rate at plasmid pRJ2. Then the BamHI fragment containing the which Ty elements engage in homologous ectopic recombi- whole GalTylUra was inserted in the unique Bgl II site ofthe nation is relatively low and is mainly nonreciprocal in nature LYS2 gene in plasmid pDP6 (23). Plasmid pM128 was con- (3). Recent experiments in our laboratory have shown that Ty structed by filling-in the Asp718 site in the Ty of pM106 with cDNA participates in gene conversion events involving a the Klenow fragment ofEscherichia coli DNA polymerase I, marked Ty element (4). followed by ligation; this process generates a SnaBI site in its Transcription by DNA-dependent RNA polymerase I has place. pYN10 was constructed in several steps: First, a been shown to cause an increase in recombination in yeast 6.2-kb BamHI-Cla I fragment of pM106 was replaced by a (5). Elevated levels of recombination between directly re- BamHI-Cla I 380-bp fragment of pBR322, thus creating peated GALIO genes have also been observed upon induction pYN6. Then, a Xho I-Asp718 fragment of the Ty was replaced by a similar one from pGTyl-H3His3 (Tya-475Am) The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviation: LTR, long terminal repeat. in accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed. 12711 Downloaded by guest on September 27, 2021 12712 Genetics: Nevo-Caspi and Kupiec Proc. Natl. Acad. Sci. USA 91 (1994) (24) containing a triple amber linker insertion. Finally, the and calculated by the method of the median (29). Twelve to 6.2-kb BamHI-Cla I fragment deleted in the first step was 24 cultures grown in SD or SGal medium (4, 27) were used in restored (instead of the pBR322 fragment), recreating a each experiment. DNA was prepared from one Ura- Leu- full-length Ty element. This plasmid was called pYN10. colony (strains carrying pM97) or one Ura- His- colony pYN9 was created by first deleting a Nhe I-Xho I fragment (strains carrying pJefl678) from each culture and subjected to (containing the Ty promoter) from pYN6 and then recon- Southern blot analysis using LYS2, neomycin-resistance structing the whole Ty with the BamHI-Cla I fragment, as gene (neo), and URA3 sequences as probes (4). described above. Plasmids pM97, pM102, and pM109 have been described (4). pYN8, carrying the same triple-amber linker as pYN10 was constructed by replacing the Xho I-Cla RESULTS I fragment of pM97 with the corresponding fragment from Increased transcription of a marked Ty (GalTylneo), result- plasmid BJC28 (24). pJefl678 (a gift from Jef Boeke, Johns ing in high levels of Ty cDNA, causes elevated rates of Hopkins University, Baltimore) carries a GalTylneo on a recombination ofanother differently marked Ty (TyUra), due 2-,um plasmid-containing vector with a HIS3 selectable to gene conversion events in which cDNA copies of the marker. pM43 was used as a probe in Southern blot hybrid- GalTylneo act as donors of information (4). The level of ization experiments; it carries a Pvu II fragment of the LYS2 conversion of the marked TyUra can be estimated by select- gene (25). ing for Ura- cells on 5-fluoroorotic acid medium (28). Media, Growth Conditions, and General Procedures. Stan- To analyze the effect of transcription of the recipient dard molecular biology procedures such as cloning, restric- molecule on the level ofrecombination, we constructed yeast tion enzyme analysis, and Southern blot analysis were car- strain MK95 (Fig. 1). This strain carries a TylUra element on ried out as described in ref. 26. Yeast media and molecular chromosome II in which the natural promoter has b6en biology procedures (transformations, DNA preparations, replaced by the GAL] promoter (GalTylUra) (4). This Ty is etc.) were carried out as in ref. 27. Ura- colonies were not expressed in a glucose-containing medium, but high selected on SD complete medium with uracil (50 mg/liter) levels of transcription are seen in a medium containing and 5-fluoroorotic acid (0.85 mg/ml) (28). galactose (results from Northern blot analysis not shown). As Measurement of Ty Recombination. Recombination rates controls, isogenic strains with different TyUra elements were were measured by fluctuation test analysis as described (4) used: MK104, bearing a Tyl with its natural promoter, and A A URA3 B MK104 > . TyA neo TyB _PR N RT RH A MK95 -1 Ll A A YN3 pbs- pM109 - --{ 7 SnB MK116 SnB pMl1 02 - t _>~~~~~-- YN2 1am A am A pYN8 - _= A B MK87 (Ty2) FIG. 1. Schematic representation ofTy elements used in the present study.