a replication origin of K. Iactis (KARS). In addition,. KARS plasmids containing these fragments have a copy number of approximately one, and each of the five ...
Curr
Current Genetics
Genet (1990)18:517-522
9 Springer-Verlag 1990
Centromeric D N A of Kluyveromyces lactis Joris J. Heus, Ben J. M. Zonneveld, H. Yde Steensma, and Johan A. Van den Berg Department of Cell Biology and Genetics, Leiden University, Wassenaarseweg 64, NL-2333 AL Leiden, The Netherlands Received August 20, 1990
Summary. A direct selection method was used to isolate centromeres from a genomic library of the yeast Kluyverornyces lactis. The method is based on the lethality at high copy number of the ochre-suppressing t R N A gene SUPI 1. Five different chromosomal fragments were found that confer mitotic stability to plasmids containing a replication origin of K. Iactis (KARS). In addition, K A R S plasmids containing these fragments have a copy number of approximately one, and each of the five fragments hybridizes to a different chromosome of K. lactis. F r o m these results we conclude that five of the six centromeres of K. lactis have been isolated. These centromeres do not function in S. cerevisiae. Key words: Isolation - Centromeres - Kluyveromyces lactis
Introduction Centromeres are specialized regions of eukaryotic chromosomes that enable replicated chromosomes to segregate properly during cell division. In mammalian cells the centromere can be seen on the mitotic chromosome as a constriction to which many microtubules attach. In contrast, only one microtubule attaches to the centromeres of Saccharomyces cerevisiae (Peterson and Ris 1976) and Aspergillus nidulans (Boylan et al. 1986). Mammalian centromeres are very large structures containing several highly repetitive e-satellite sequences which are tandemly arranged into arrays comprising millions of base pairs (Wevrick and Willard 1989; Joseph etal. 1989). To-date, only the centromeres of Schizosaccharornycespombe and S. cerevisiae have been studied in detail. S. pornbe has three chromosomes with cen-
Offprint requests" to: J. Heus
tromeres ranging in length from 40000 to 100000 base pairs (bp) (Chikasige et al. 1989). In the centromere regions repetitive, 4 kbp long, elements have been found (Nakaseko et al. 1986; Clarke et al. 1986). In S. cerevisiae 13 out of 16 centromeres have been isolated using three different methods. The first centromeres were isolated by chromosome walking (Clarke and Carbon 1980). Subsequently, a method was used that is based on the fact that centromeres increase the stability of ARS plasmids. By reiterant selection of stable transformants Maine et al. (1984) cloned CEN5. Finally a direct selection scheme, devised by Hieter et al. (1985 b) and based on the lethality at high copy number of the S U P l l gene product, was successfully used to isolate ten centromeres. The S. cerevisiae centromere can be divided into three domains: centromeric D N A elements (CDE) I, II and III. CDE I is a highly conserved 8 bp sequence (PuTCACPuTG); CDE II is a 78 86 bp stretch containing 9 0 - 9 5 % AT; and CDE III is a highly conserved 25 bp stretch (TGT.T.TG..TTCCGAA ..... AAA) (Hieter et al. 1985b). Together, these elements can form a functional centromere of only 125 bp (Cottarel et al. 1989). Cloned into a plasmid containing an autonomously replicating sequence (ARS), these 125 bp confer stability upon the plasmid and reduce its copy number to one per cell. Centromeres from the related species Saccharomyces uvarum, which probably has the same number of chromosomes as S. cerevisiae, also match the S. cerevisiae consensus sequence (Huberman et al. 1986; our unpublished results). The large difference in centromere size and structure between S. pombe and S. cerevisiae/uvarum prompted us to study the centromeres of Kluyveromyces lactis, an industrially important yeast. Two other features indicating that the centromeres of K. lactis might differ from those of the other two yeast species are the finding that S. cerevisiae centromeres do not function in K. lactis (Sreekrishna et al. 1984) or in S. pombe (Carbon 1984), and the fact that the chromosome number of K. lactis (six) is intermediate between that of S. cerevisiae (sixteen) and S. pornbe (three).
518 Five fragments containing centromeric DNA of K. lactis w e r e i s o l a t e d u s i n g a d i r e c t s e l e c t i o n s c h e m e ( H i e t e r et al. 1985 b). T h e s e c e n t r o m e r i c D N A f r a g m e n t s d o n o t f u n c t i o n as c e n t r o m e r e s in S. cerevisiae.
ARS1 amp
KARS2 OR[
YRpl 4/KARS2 8.6 kbp
Materials and methods Strains and media. Kluyveromyces laetis strain S D l l (trpl lac4-1, Das and Hollenberg 1982) was mutagenized with EMS (ethylmethanesulfonate) as described by Fink (1970). Using 5-fiuoroorotic acid, a mutant was selected that could be complemented by the URA3 gene of Saccharomyces cerevisiae. This strain, JBDI00 (trpl lac4-1 ura3-100), was used both to construct the genomic library and for transformations. Another EMS mutant, BZL100 (trpl lac4-1 lysA-52), was used in hybridization experiments to correlate the centromeric D N A fragments with the chromosomes of K. lactis. Strain YNN214 (MATa ura3-52 lys2-801 amber ade2-101 ~176 Johnston and Davis 1984) was used for the transformation of Saccharomyces cerevisiae. Non-selective medium (YPD) was as described by Sherman et al. (1979), minimal medium (MY) as described by Zonneveld (1986). When required, amino acids and uracil were added to a final concentration of 20 p,g/ml. Media were solidified by the addition of 1.5% Difco Bacto agar. Plasmids were amplified in Escherichia coli MC1061 [araD 139 (ara leu) 7697 lacX74 galU gall( hsr- hsm + strA, Casadaban and Cohen 1980], E. coli 490 (r[ m~ reeA-, Hollenberg et al. 1976) and E. coli JM83 [ara (lac pro A, B) rps L phi80 laeZ M15 (r+ m + ), Vieira and Messing 1982]. Bacterial media were prepared as described by Sambrook et al. (1989). Transformations, recombinant DNA methods and enzymes. E. coli strains were transformed using either a CaC1 z (Sambrook et al. 1989) or an electroporation protocol (Dower 1988). Yeast strains were transformed as described by Ito et al. (1983). Plasmid D N A was isolated from E. eoli as described by Sambrook et al. 1989. Plasmid D N A was isolated from K. lactis using glassbeads (Hoffman and Winston 1987) with the following modifications. After vortexing with glassbeads (0.4-0.52 ram) in gnanidinium-buffer (4.5M guanidinium-HC1, 0.1 M EDTA, 0.15 M NaC1, 0.05% sarkosyl, pH 8.0), the supernatant was extensively extracted with phenol-chloroform-isoamylalcohol (25:24:1). Total D N A was isolated from K. lactis as described by Holm et al. (1986). Restriction nucleases, Klenow D N A polymerase and T4 D N A ligase were obtained from Pharmacia (Uppsala, Sweden) and used according to the suppliers' specifications. Calf intestine alkaline phosphatase (CIAP), Proteinase K and RNase A were obtained from Boehringer Mannheim GmbH (FRG); Zymolyase 100 T from the Seikagaku Kogyo Co. Tokyo, Japan). DNA-probes were radioactively labeled using a random-prime labeling kit (Amersham) with [e.3/p] dCTP (3000 Ci/mmol, Radiochemical Centre, Amersham, UK). The 1 kb D N A size marker was obtained from Bethesda Research Laboratories Inc. (Gaithersburg, Md., USA). Plasmid constructions. A BglII-SphI fragment from p U R K 528-02 (a gift from N. Overbeeke), containing an ARS from K. lactis (KARS2), was inserted into pUC19B, p u c I g B (a gift from W. Musters) is pUC19 with a 8 bp BglII linker inserted into the Sinai-site of the polylinker. Subsequently the BglII-HindIII KARS2 fragment from pUCI9B/KARS2 was inserted into YRpI4/ARS1 digested with BglII and HindIII. The resulting plasmid was named YRpI4/KARS2 (Fig. 1). Plasmid YRpI4/VSUPKARS2 was constructed by inserting the BglII-HindIII KARS2 fragment from pUC19B/KARS2 into YRp14/ARS1/VSUP11. YRp14/ARS1/ VSUP11 (a gift from A. Barbeiro) has a 4 bp insertion in the S U P l l gene, made by cleaving Y R p I 4 / A R S I with AvaI, treating the linear molecules with Klenow D N A polymerase I and re-ligating. This insertion renders the S U P l l gene inactive. Construction of the genomic library. Genomic D N A was isolated from K. laetis strain JBD100. The D N A was partially digested with
SUP11
URA3
Fig. l. YRp14/KARS2. A, AvaL B, BamHI; Bg, BglII; H, HindIII; s, SphI
Sau3A to obtain 0 - 1 0 kb D N A fragments and separated using a 10-40% sucrose gradient. Thirty-five ng of the 0 - 5 kb and 100 ng of the 5-10 kb fraction were ligated separately in the presence of 200 ng of phosphatase-treated YRp14/KARS2 linearized at its unique BamHI site. The ligation mixtures were treated with phenol/ chloroform, ethanol precipitated and the pellets resuspended in 30 gl HzO; 4 x 5 gl were used to transform E. coli by electroporation. Both libraries yielded approximately 35000 transformants. CHEF conditions and hybridization experiments. Chromosomal D N A molecules from strain BZL100 were separated using a contourdamped homogeneous electric field (CHEF) apparatus (Bio-Rad, Richmond, USA). Agarose plugs were prepared in Bio-Rad Low Gel Temperature Agarose as described (De Jonge et al. 1986). The C H E F conditions used were: 0.6% agarose (Seakem, F M C Bioproducts, Rockland, USA) in 0 . 5 x T B E (0.089M Tris-HC1, 0.089 M boric acid, 0.002 M EDTA, pH 8.3), runtime 160 h, pulse time 50-1800 sec, 60 V. The gel was blotted onto a GeneScreen Plus filter (New England Nuclear Boston, Mass., USA) using a Vacugene 2061 blot apparatus (LKB, Bromma, Sweden). Strips from this filter were hybridized to probes derived from the cloned genomic inserts using standard protocols (Sambrook et al. 1989). The Southern blot for the determination of the copy number was made by capillary blotting (Sambrook et al. 1989) and hybridized to the 1.95 kb BglII-HindIII KARS2-fragment from YRpI4/KARS2. Stability assays. The transformants that survived the SUP11 selection method were initially tested in a simple stability assay. Single colonies were picked and grown on YPD for approximately ten generations. Subsequently, a dilution (+_300 cells per plate) was spread onto a selective (MY+tryptophan) and a non-selective (MY + tryptophan § uracil) plate. The ratio between the number of colonies on the two plates was taken as a measure of stability. Only those transformants showing 20-100% stability were analyzed further using a more accurate stability assay based on a method described by Sikorski and Hieter (1989). A single colony was grown in 50 ml M Y + t r p to late exponential phase (OD26o = 1.6-2.4). Subsequently 10-12 generations of non-selective growth were initiated by inoculating 50 ml YPD with 50 ~tl of the selective culture. The percentage of ceils containing a plasmid at the start (to) and after the ten generations of non-selective growth (ty) was determined in the following way. Appropriate dilutions were plated in triplicate on YPD. After 2 - 3 days incubation at 30~ these plates were replica-plated onto MY + trp and MY + trp + ura. The ratio between these plates was taken as a measure of stability. The number of generations grown was calculated from the numbers on the t o and ty YPD plates. Loss per generation was obtained using the formula: % loss = {1 - e t 100 copies), probably due to a very high segregation bias. This high copy number corresponded to the low stability of this plasmid on selective medium. Since the isolated K. lactis fragments met b o t h requirements for a centromere, namely conferring stability to and reducing the copy number of KARS-plasmids, it
521
Fig. 4. Copy number of pK1CEN plasmids. Hybridization of KARS2 to chromosomal DNA of K. lactis, _+plasmid, and to YRp14/KARS2, all digested with EeoRI and HindIII. Lane a, +pK1CENI-II.9; lane b, +pK1CEN2-II.8; lane c, +pK1CEN31.34; lane d, +pK1CEN4-I.31; lane e, +pK1CEN6-I.2; lane f, -plasmid; lane g, + YRpl4/VSUPKARS2; lane h, YRp14/KARS2 (1 rig) was concluded that the isolated DNA fragments contained centromeres of K. lactis.
Centromeres o f K. lactis do not function in S. cerevisiae
By transforming K. lactis with a KARS plasmid containing a S. cerevisiae centromere Sreekrishna et al. (1984) showed that CEN4 of S. cerevisiae is not capable of increasing the stability of a KARS plasmid. To test whether centromeres of K. lactis are functional in S. cerevisiae, strain YNN214 was transformed with equal amounts of pK1CEN plasmids, YRp14/KARS2 and YRp14/ VSUPKARS2. All these plasmids also contain ARSI for replication in S. cerevisiae. Three days after transformation only a few (0-5) colonies, plus a number of probably abortive transformants (colony diameter < 0.5 mm), were obtained with the pK1CEN plasmids as well as with YRp14/KARS2. Control plasmids YRpl4/VSUPKARS2 and pRS316, which contains a centromere ofS. cerevisiae (Sikorski and Hieter 1989), gave at least 100-fold more transformants. Therefore centromeres of K. lactis do not function in S. cerevisiae.
Discussion
To-date centromeres have only been isolated from the evolutionary distantly related yeasts Saccharomyces cerevisiae and Schizosaccharomycespombe. Both yeasts differ in the way they divide (budding versus fission), their number of chromosomes (16 versus 3) and their centromere size (120 versus 40000-100000 bp). A direct selection scheme devised by Hieter et al. (1985 b) was used to isolate the centromere sequences of Kluyveromyces lactis, which is a budding yeast with six chromosomes. This method, based on the lethal combination of a high copy number plasmid and the S U P l l gene of S. cerevisiae, was shown to be functional in K. lactis.
Plasmid YRpI4/VSUPKARS2, in which the S U P l l gene has been inactivated by a 4 bp insertion, gave 50-200 times more transformants than vector YRpl4/KARS2. The insert size of the genomic libraries of K. lactis was restricted to 10kbp, because we suspected the centromeres ofK. lactis to be more closely related to those of S. cerevisiae than to those of S. pombe. After transformation of the libraries to K. lactis the resulting transformants could be divided into three classes with respect to the stability of the Ura marker; i.e., unstable (< 10%), intermediate stable (23-78%) and stable (100%). The 100% stability class probably arose from integration of plasmids, since isolation of plasmids from these transformants failed in most cases. The rare plasmids that were obtained could not re-transform K. lactis although most of them consisted of YRpl4/ KARS2 plus an insert. The fact that plasmids containing YRpl4/KARS2 plus a genomic insert could be isolated from the 100% stability class indicates that the looping out of a plasmid after integration is possible. One of the inserts was a 7.2 kb fragment derived from the rDNA cluster. The frequency of integration into the repetitive rDNA cluster is expected to be relatively high, since there are about 115 copies of the rDNA unit in K. lactis (Maleszka and Clark-Walker 1989). Transformants escaping the S U P l l lethality without being an integrant or containing a centromere were unstable. They behaved like control plasmid YRpI4/ VSUPKARS2. A conclusive explanation for these "escapers" (Hieter et al. 1985b) is not available, but they could, for example, be caused by mutations in the plasmid, i.e., in the S U P I 1 gene or in the KARS, or by mutations in the host. The plasmids in the 17 transformants from the intermediate stability class represent five different genomic fragments as shown by the restriction maps and hybridization experiments. The mitotic stability of the transformants from the intermediate stability class was approximately 90% on selective media. This value is comparable to those found for centromeres of S. cerevisiae, which vary from 70-90% (Sikorski and Hieter 1989; Murray and Szostak 1983; Bloom etal. 1982; Fitzgerald-Hayes et al. 1982). Control plasmid YRp14/ VSUPKARS2 was retained in only about 7% of the cells, probably due to a high segregation bias. After about ten generations of non-selective growth the pK1CEN plasmids were found in 55 78% of the cells (Table 1). These values are similar to those obtained with CEN plasmids in S. cerevisiae (Cottarel et al. 1989; Hegemann et al. 1986; Hieter et al. 1985a; Sikorski and Hieter 1989) although some groups have reported higher stabilities ranging from 80-97% (Panzeri et al. 1985; Cumberledge and Carbon 1987). Control plasmid YRpI4/VSUPKARS2 had a stability of approximately 5%. The loss rate of YRpl 4/VSUPKARS2 was not determined, because of the low stability at t o (Murray and Szostak 1983). Four of the five isolated centromeric DNA fragments reduced the high copy number of KARS plasmids to approximately one. The fifth fragment, containing CEN6, gave rise to a copy number of about four. A possible explanation for this elevated copy number is that this DNA fragment does not contain a complete centromere.
522 However, the ability to stabilize plasmids, which is thought to be inseparable from the ability to reduce the copy numer, is not impaired. To address these and other questions, sequence analysis of the centromeres is currently being performed. Another implication of this deviant centromere is that CEN5 may not have been cloned using the SUP11 method, because it gives rise to a copy number that is intolerable in combination with the SUP11 gene. We have not shown that the five D N A fragments isolated from K. lactis are indeed functioning as centromeres in the chromosomes of K. lactis. If these sequences are centromeres, not more than one of them should hybridize to a single chromosome, since dicentric chromosomes are usually unstable (McClintock 1939; J/iger and Philippsen 1989). Therefore, the finding that the five isolated fragments each hybridize to a different chromosome conforms with this expectation. We firmly believe that the isolated fragments contain genuine centromeres of K. lactis. On plasmids they behave similar to centromeres of S. cerevisiae, i.e., they stabilize plasmids and reduce their high copy number to approximately one.
Acknowledgements. The authors thank Hanneke van Deventer for her valuable contribution. References
Bloom KS, Fitzgerald-Hayes M, Carbon J (1982) Cold Spring Harbor Syrup Quant Biol 47:1175-1185 Boylan MT, Holland MJ, Timberlake WE (1986) Mol Cell Biol 6: 362] -3625 Carbon J (1984) Cell 37:351-353 Casadaban MJ, Cohen SN (1980) J Mol Biol 138:179-207 Chikashige Y, Kinoshita N, Nakaseko Y, Matsumoto T, Murakami S, Niwa O, Yanagida M (1989) Cell 57:739-751 Clarke L, Carbon J (1980) Nature 287:504 509 Clarke L, Amstutz H, Fishel B, Carbon J (1986) Proc Natl Acad Sci USA 83:8253-8257 Cottarel G, Shero JH, Hieter P, Hegemann JH (1989) Mol Cell Biol 9:3342-3349 Cumberledge S, Carbon J (1987) Genetics 117:203-212 Das S, Hollenberg CP (1982) Curr Genet 6:123-128
De Jonge P, De Jongh FCM, Meijers R, Steensma HY, Scheffers WA (1986) Yeast 2:193-204 Dower WJ (1988) Mol Biol Reports 6:3-4 Fink GR (1970) Methods Enzymol 17A:59-78 Fitzgerald-Hayes M, Clarke L, Carbon J (1982) Cell 29:235-244 Hegemann JH, Pridmore RD, Schneider R, Philippsen P (1986) Mol Gen Genet 205:305-311 Hieter P, Mann C, Snyder M, Davis RW (1985a) Cell 40:381-392 Hieter P, Pridmore D, Hegemann JH, Thomas M, Davis RW, Philippsen P (1985b) Cell 42:913-921 Hoffman CS, Winston F (1987) Gene 57:267-272 Hollenberg CP, Degelmann A, Kustermann-Kuhn B, Royer HD (1976) Proc Natl Acad Sci USA 73:2072-2076 Holm C, Meeks-Wagner DW, Fangman WL, Botstein D (1986) Gene 42:169-173 Huberman JA, Pridmore RD, Jfiger D, Zonneveld B, Philippsen P (1986) Chromosoma 94:162-168 Ito H, Fukuda Y, Murata K, Kimura A (1983) J Bacterio1153:163 168 J/iger D, Philippsen P (1989) EMBO J 8:247-254 Johnston HM, Davis RW (1984) Mol Cell Biol 4:1440-1448 Joseph A, Mitchell AR, Miller OJ (1989) Exp Cell Res 183:494-500 Maine GT, Surosky RT, Tye B-K (1984) Mol Cell Biol 4:86-91 Maleszka R, Clark-Walker GD (1989) Curr Genet 16:429-432 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: a laboratory manual 2na ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York McClintock B (1939) Proc Natl Acad Sci USA 25:405-416 Murray AW, Szostak JW (1983) Cell 34:961-970 Nakaseko Y, Adachi Y, Funahashi S, Niwa O, Yanagida M (1986) EMBO J 5:1011-1021 Panzeri L, Landonio L, Stotz A, Philippsen P (1985) EMBO J 4:1867-1874 Peterson JB, Ris H (1976) J Cell Sci 22:219-242 Sherman F, Fink G, Lawrence C (1979) Methods in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Sikorski RS, Hieter P (1989) Genetics 122:19-27 Sreekrishna K, Webster TD, Dickson RC (1984) Gene 28:73-81 Vieira J, Messing J (1982) Gene 19:259-268 Wevrick R, Willard HF (1989) Proc Natl Acad Sci USA 86: 93949398 Zonneveld BJM (1986) J Microbiol Methods 4:287-291
Communicated by C. P. Hollenberg
