Tyl Transposition in Saccharomyces cerevisiae Is

A large collection of T y l insertions in the URA3 and LYS2 loci was generated using a GALl-Tyl fusion to augment the transposition frequency. The sites of insertion of most of these T y elements were sequenced. There appears to be a gradient of frequency of insertion from the 5’ end (highest frequency) to the 3’ end (lowest frequency) of both loci. In addition we observed hotspots for transposition. Twelve of the 82 T y l insertions in the URA3 locus were inserted in exactly the same site. Hotspots were also observed in the LYS2 locus. All hotspots were in the transcribed part of the genes. Alignment of the sites of insertion and of the neighboring sequences only reveals very weak sequence similarities. T RANSPOSABLE elements are characterized by their ability to insert copies of themselves into nonhomologous target sequences. In most cases, transposons duplicate a small number of base pairs of the target sequence as part of the transposition process; the number of base pairs duplicated is characteristic of that type of transposon. This duplication strongly implies that a specific transposase (or in the case of retrotransposons and retroviruses, integrase) encoded by a given type of element mediates the breaking and joining reactions that are the crux of a transposition event. Thus a study of the target sequences recognized during transposition reactions should yield information on the properties of transposition intermediates. The genome of the yeast, Saccharomyces cerevisiae, contains three families of transposable lements, called T y elements (reviewed by BOEKE 1989). Two of these families, consisting of the T y l and Ty2 elements, are about 6 kb long and are flanked by long terminal repeats (LTRs, also called 6) of approximately 340 bp. The third family, Ty3, has the same overall structure but is very different from the other Tys in sequence. T y 3 elements and their solo LTR sequences are always found inserted 15 to 19 bp from the 5’ end of tRNA coding regions (HANSEN, CHALKER and SANDMEYER 1988; SANDMEYER et al. 1988). Transcription of T y elements initiates in the 5’ LTR and terminates in the 3’ LTR; 45 nucleotides of the major T y transcript are terminally reduntant (ELDER, LOH and DAVIS 1983). These structural features, as well as sequence comparisons between T y ’ Present address: Department of Cellular and Developmental Biology, Harvard University, Cambridge, Massachusetts 02138. The publication costs of this article were partly defrayed by the payment ment” in accordance with 18 U.S.C. $1734 solely to indicate this fact. of page charges. This articles must therefore be hereby marked “advertiseGenetics 123: 269-279 (October, 1989) elements and retrotransposons from other eucaryotes, suggested an evolutionary link with the retroviruses of higher eucaryotes. Functional evidence for this relationship was obtained when T y l elements were shown to transpose through an RNA intermediate, to encode a reverse transcriptase activity and to produce Ty-specified virus-like particles (Ty-VLPs) (BOEKE et al. 1985; GARFINKEL, BOEKE and FINK 1985; MELLOR et al. 1985). The full-length T y transcript is packaged into Ty-VLPs, where it is converted into doublestranded DNA by the Ty-encoded reverse transcriptase. The double-stranded reverse transcript is then integrated into the host genome as the final step in the transposition process. The integration reaction results in the duplication of five bp of target DNA at each end of the new copy of the T y element (FARABAUGH and FINK 1980; GAFNER and PHILIPPSEN 1980). A DNA-containing form of the Ty-VLP appears to be an intermediate in the transposition process (EICHINGER and BOEKE 1988). The precise sites of insertion of several T y l and Ty2 elements have previously been determined at several loci. These results pointed to the possibility that Tyl and T y 2 elements insert preferentially in the 5‘ region of genes. For example, eight of nine randomly selected T y insertions in the LYS2 locus were located in the 5’ region (EIBEL and PHILIPPSEN 1984; SIMCHEN et al. 1984). Similarly, the hid -912 and his4 91 7 mutations are caused by the insertion of T y elements upstream of the HIS4 gene (ROEDER et al. 1980). Many examples of T y l and T y 2 insertions that activate the expression of silent genes have also been studied; not surprisingly, these insertions are all [with one exception (BACH, 1984)] in 5’ noncoding regions (ERREDE et al. 1980; reviewed in BOEKE 1989). Thus 270 G. Natsoulis et al. the existing set of sequenced T y insertions is far from being a random sample. In fact, in most cases, these mutations were chosen for their remarkable genetic properties such as the ability to yield extragenic revertants or the ability to alter the regulation of adjacent genes. In this paper we describe the molecular characterisation of a large number of T y l transposition events and related events that occurred at the URA3 and LYS2 loci. In an attempt to select unbiased sets of T y insertions, we made use of the positive selections that exist for the loss or reduction of URA3 and LYS2 gene function. We describe the isolation, cloning and sequencing of the site of insertion of nearly one hundred T y l elements in the URA3 and the LYS2 loci of yeast. MATERIALS AND METHODS Media and strains: Yeast growth media have been described (SHERMAN, FINK and LAWRENCE 1978). The 5fluoroorotic acid (5-FOA) containing medium was described by BOEKE, LACROUTE and FINK 1984). The strains we used are JBXl69-lOB (a lys2, t r p l A l ) , JBX169-11B (a lys2, t r p l A l ) , JB282, a yeast transformant of BWGl-7a (a adel-100, his#-516, ura3-52, leu2-3,112) carrying the plasmid pGTyl-H3, and GNXlO9 (ala leu2Al/ +, ura3-52/ura3-52, his3A200/+, ade2-101/+, lys2-801/+). Plasmid constructions: The plasmid pGTyl-H3 has been described previously (BOEKE et al. 1985). Plasmid pJEFlll4 was constructed by replacing the SmaI-BamHI fragment containing the URA3 gene of plasmid pJEF724 (=pGTylH3) with an EcoRI-BamHI fragment containing the TRPl gene (the EcoRI end had been previously filled in using the Klenow fragment). Plasmid pGN801 was constructed as follows: pSK179 (kindly provided by SAM KUNES, MIT) consists of a 13-kb EcoRI yeast genomic fragment containing the URA3 gene cloned in the EcoRI site of pBR322 [see ROSE, GRISAFI and BOTSTEIN (1984) for a BglII restriction map of this region]. pSK179 was digested with BgII, filled in with Klenow large fragment enzyme, and ligated to NotI linkers. The largest fragment was redigested with NotI , EcoRI and SalI. We isolated a 2-kb SalI-Not1 fragment from the 5' flanking region of URA3 and a 4-kb EcoRI-Not1 fragment from the 3' flanking region of URA3. These two fragments were ligated to the pBM453 vector that had been previously digested with EcoRI and SalI. pBM453 was kindly provided by MARK JOHNSTON (Washington University). It is a pBR322 derivative carrying ARSl, TRPl and CEN3 sequences. Thus the resulting CEN plasmid, pGN8O 1, carries about 7.6 kb of URA3 flanking regions surrounding a unique NotI site in place of 5.4 kb of DNA that contains the URA3 gene itself. DNA procedures: Plasmid isolations, restriction analysis, gel electrophoresis and Southern blot analysis were described by MANIATIS, FRITSCH and SAMBROOK (1982). Radiolabeled probes were prepared using random oligonucleotide primers as described by FEINBERC and VOCELSTEIN (1983). Most of the sites of T y element insertion were sequenced on cloned double stranded plasmid DNA using either Sequenase (US. Biochemicals) or reverse transcriptase, using the method of HUIBRECSTE and ENCELKE (1986). A few of the insertion points could be unambiguously identified by sequencing genomic DNA directly with oligonucleotide primers (labeled with y["P]ATP using polynucleotide kinase) and AMV reverse transcriptase (HUIBREGSTE and ENCELKE 1986). We synthesized three primers complementary to URA3 coding sequences. All URA3 numbering in this paper is according to the sequence of ROSE, GRISAFI and BOTSTEIN (1 984). The sequences of the primers used are as follows: URA3-1: TAACTGTGCCCTCCATGG (432-449) URA3-2: GTCGCTCTTCGCTCCCTG (734-751) URA3-3: AGTTTTGCTGGCCGCATC (1010-1028) Most of the URA3 gene can be sequenced using one of these primers. URA3-3 tended to give poor sequence results, particularly on total genomic DNA. In some cases we used primers reading out from the 5' end or the 3' end of the LTR. All T y l numbering is according to BOEKE et al. (1 988). The primers used were as follows: USOUT: AACACCGTATATGATAATAT (52-72) U50UT: AATGGAATCCCCAACAAT (294-309) When U3OUT and U50UT were used as primers, and when the template plasmid carried a whole Ty, we sequenced a purified restriction fragment carrying a single LTR. Most T y element junction sequences were obtained from only one end, however, in all cases where both ends were sequenced, a target site duplication of 5 bp was observed. Testing the phenotype of a deletion 5' to the URA3 Gene: The diploid strain GNXlO9 (ura3-52/ura3-52) was transformed with a linear piece of DNA containing the wildtype URA3 gene but lacking the HindIII (-650) to HindIII (+1) fragment found just upstream of the URA3 gene. Ura+ diploid transformants were selected and sporulated. Fifteen asci were dissected; all 15 tetrads had four viable spores. The Ura' phenotype segregated 2+/2-. For two tetrads we confirmed by Southern blotting that the Ura+ spores carried the 0.65-kb deletion.