Proc. Natl. Acad. Sci. USA Vol. 82, pp. 4167-4171, June 1985 Genetics Caenorhabditis elegans DNA that directs segregation in yeast cells (centromere/nematode/Saccharomyces cerevisiae/mitosis/meiosis) DAN T. STINCHCOMB*, CRAIG MELLO*, AND DAVID HIRSH Department of Molecular, Cellular, and Developmental Biology, University of Colorado, Boulder, CO 80309 Communicated by William B. Wood, January 14, 1985 ABSTRACT We have isolated seven DNA fragments from different DNA fragments from C. elegans DNA that stabilize Caenorhabditis elegans that enhance the mitotic segregation of autonomously replicating hybrid molecules. autonomously replicating plasmids in the yeast Saccharomyces cerevisiae. These segregators, designated SEGI-SEG7, behave MATERIALS AND METHODS like isolated yeast chromosomes: they increase the stability and simultaneously lower the copy number of circular plasmids Strains and Materials. BNN6 (also known as HB101) was during mitotic growth in yeast. During meiosis, plasmids used for all Escherichia coli transformation experiments. The containing the C. elegans segregators show higher levels of plasmid YRp17 consists ofpBR322 sequences (12), the URA3 precocious or aberrant disjunction than do plasmids bearing and TRPJ genes, and the putative origin of replication ARSI isolated yeast centromeres. Yet one ofthe segregators improved (6). YNN140 (a his3-1 trpl-289 ura3-1 ura3-2 ade2) was used the meiotic segregation of the parental plasmid. We estimate in all segregator enrichment experiments and stability tests. that there may be as many as 30 segregator sequences in the C. YNN152 (a leul met14 adel ura3 trpl his2 gall, obtained ekgans genome, a value that is consistent with the polycentric from the Yeast Genetic Stock Center as X2928-3D-1C) was nature of C. elegans chromosomes. Five of the seven segrega- mated to YNN140 for meiotic analyses. Restriction tors are linked to sequences that are repeated in the worm endonucleases, E. coli DNA polymerase I, T4 DNA ligase, genome, and four of these five segregators cross-hybridize. DNase, RNase, and calf alkaline phosphatase were all Other members of this family of repetitive DNA do not contain purchased from the usual vendors. Protocols and media for segregator function. Segregator sequences may prove useful for growth of E. coli, plasmids, and bacteriophage X are de- probing the structure of centromeres of both C. elegans and S. scribed by Davis et al. (13). Yeast media were as described cerevisiae chromosomes. by Mortimer and Hawthorne (14). DNA Preparations. Plasmid DNA was prepared by tech- DNA transformation of the yeast Saccharomyces cerevisiae niques described by Maniatis et al. (15). Phage DNA was has been used to identify DNA sequences that direct chromo- prepared as described by Davis et al. (13). C. elegans var. some behavior. Upon introduction into a yeast cell, hybrid Bristol DNA was provided by Steve Carr. Rapid yeast DNA molecules carrying only a yeast structural gene are not preparations were performed as described (16). DNA frag- capable of autonomous replication or segregation. Addition ments were purified out of agarose gels by using DEAE ofARS DNA (Autonomously Replicating Sequence) to such membranes (17) and were inserted into plasmid vectors by a chimera permits its replication (1). In spite of their high standard techniques (13). copy number and their ability to replicate during each cell Construction of Hybrid Pools. Two collections of hybrid division (2, 3), ARS-containing hybrid molecules are not molecules were used to enrich for segregators. For the first propagated efficiently during mitotic growth. Even under screen, 1 ,g of YRp17 and 2.5 jig of C. elegans DNA were selective conditions, only 2-10% ofthe cells in a transformed cut to completion with the restriction endonuclease EcoRI. culture bear the ARS hybrid molecule (1, 4). Under nonselec- The DNA samples were mixed after extraction with phenol tive conditions, the hybrid plasmid is rapidly lost. Centro- and precipitation with ethanol and were resuspended in 50 ,u1 meric sequences, designated CEN, alleviate this instability; of 100 mM NaCl/50 mM Tris HCl, pH 7.4/10 mM MgSO4/ by allowing proper segregation, hybrids are efficiently trans- 1 mM ATP/10 mM dithiothreitol. The EcoRI cohesive ends ferred to daughter cells (5-8). The result is a significant were ligated with 60 units of T4 DNA ligase at 4°C for 16 hr. increase in the proportion of cells that stably maintain the Then this collection of chimeric DNA molecules was used transformed phenotype. This stability can be used to enrich directly to transform the yeast strain YNN140 from Ura- to for hybrids bearing DNA sequences that aid segregation of Ura+ by modification (16) of the original transformation ARS-containing molecules; we term such a sequence SEG for procedure (18). The second pool was constructed as follows. segregator. For instance, when hybrids bearing random YRp17 (0.5 ,ug) was cleaved with BamHI, treated with fragments ofyeast DNA were screened or enriched for stable alkaline phosphatase, extracted with phenol, and mixed with segregation, the only DNA carrying segregator functions 2.5 jig of BamHI-cleaved C. elegans DNA. After precipita- were yeast centromeres or fragments ofthe endogenous yeast tion with ethanol, the DNA samples were ligated in 50-,ul plasmid (6, 8, 9). Herein, we describe the isolation of volumes as described above. The hybrid DNAs were used to segregator functions from DNA of the nematode Caeno- transform E. coli (19) to ampicillin resistance (AmpR). The rhabditis elegans. C. elegans was chosen as a source for such resistant clones were pooled into four separate batches and functions because its chromosomes are poly- or holocentric amplified for plasmid DNA purification. Plasmid DNA also (10, 11). If the lack of a defined centromere is due to multiple was prepared from several individual ampR E. coli colonies sites directing chromosome segregation (and if some of these isolated after amplification. The preparations were digested sites function in yeast), then segregator functions would be with restriction nucleases and sized by agarose gel electro- highly prevalent in worm DNA. Indeed, we isolated seven Abbreviation: kb, kilobase pair. The publication costs of this article were defrayed in part by page charge *Present address: Department of Cellular and Developmental Biol- payment. This article must therefore be hereby marked "advertisement" ogy, Harvard University, 16 Divinity Avenue, Cambridge, MA in accordance with 18 U.S.C. §1734 solely to indicate this fact. 02138. 4167 Downloaded by guest on September 26, 2021 4168 Genetics: Stinchcomb et al. Proc. Natl. Acad. Sci. USA 82 (1985) phoresis, permitting accurate determination ofthe quantity of Mitotic Stability Tests. Cultures grown in minimal media worm DNA present in the pools of hybrid plasmids. This lacking uracil were diluted and plated onto minimal agar plasmid DNA was then used to transform yeast as described. plates containing uracil; these were then replica-plated to Enrichment for Segregators. In the first screen (diagramed selective and nonselective plates to determine the percentage in Fig. 1), 1200 individual yeast transformants were pooled Ura+ cells. In addition, each culture was diluted 1:1000 into into four independent batches; 105 cells from each of the minimal medium with uracil and again grown until saturated. batch cultures were grown to saturation (3 x 107 cells per ml) The percentage of cells that retained the Ura+ phenotype in 5 ml of nonselective medium containing uracil. Each after these 10 generations of nonselective growth was as- culture was then diluted 1:1000 into the same medium lacking sayed by dilution and replica-plating. uracil and again grown until saturated. A second round of Copy Number Determinations. Copy number measure- growth in alternating nonselective and selective media was ments were performed as described (2). Briefly, Ura+ were grown medium. An aliquot of performed. At this time, the culture was diluted and spread transformants in selective each culture was diluted, plated, and replica-plated to assess on plates lacking uracil so that single Ura' colonies could be how many cells carried the plasmid. Each culture was then isolated. In subsequent screens, transformants were grown harvested, and sheared DNA was prepared, spotted on in alternating selective and nonselective media in microtiter nitrocellulose strips, and hybridized with [32P]DNA. The were had disap- wells. Transfers continued until YRp17 hybridization to each spot was quantitated by scintillation peared, while the previously isolated YRp17-SEG2 plasmid counting. We compared the hybridization of single-copy was maintained in control wells. Surviving Ura' cells were probes (fragments of the yeast HIS3 or ADE8 genes) to the plated and analyzed as described in Results. hybridization to pBR322. We adjusted for the differential hybridization efficiencies of the probes by comparing the Segregator Screen hybridization to known amounts of a hybrid plasmid carrying one copy of pBR322 sequences and one copy of yeast ARS I sequence. Thus, we could calculate the number of plasmid sequences per cell in the harvested populations. Dividing by gave copy YRp1 + the fraction of Ura+ cells in each culture the numbers of pBR322 sequences in each transformed cell URA3 shown in Table 1. Meiotic Analysis. Diploids were constructed by mating the Ligate haploid Ura+ transformants of YNN140 with a haploid of ARS i opposite mating type (YNN152). The resulting diploid