The Role of Centromere Alignment in Meiosis I Segregation of Homologous Chromosomes in Saccharomyces cerevisiae

Corresponding author: David B. Kaback, Department of Microbiology and Molecular Genetics, UMDNJ, New Jersey Medical School, 185 South Orange Ave., Newark, NJ 07103., kaback{at}umdnj.edu (E-mail)

Communicating editor: M. LICHTEN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

During meiosis, homologous chromosomes pair and then segregate from each other at the first meiotic division. Homologous centromeres appear to be aligned when chromosomes are paired. The role of centromere alignment in meiotic chromosome segregation was investigated in Saccharomyces cerevisiae diploids that contained one intact copy of chromosome I and one copy bisected into two functional centromere-containing fragments. The centromere on one fragment was aligned with the centromere on the intact chromosome while the centromere on the other fragment was either aligned or misaligned. Fragments containing aligned centromeres segregated efficiently from the intact chromosome, while fragments containing misaligned centromeres segregated much less efficiently from the intact chromosome. Less efficient segregation was correlated with crossing over in the region between the misaligned centromeres. Models that suggest that these crossovers impede proper segregation by preventing either a segregation-promoting chromosome alignment on the meiotic spindle or some physical interaction between homologous centromeres are proposed.

DURING meiosis I prophase, homologous chromosomes pair and undergo crossing over to form bivalents. At later stages, the bivalents attach to the spindle and the homologues segregate from each other to reduce the number of chromosomes by half. Following pairing, homologous centromeres appear to be aligned (SCHERTHAN et al. 1992 ). In some higher organisms, centromeres on each homologue appear to be both aligned and oriented in opposite directions. This orientation is believed to be important for spindle attachment and the subsequent segregation of the homologues (NICKLAS 1997 ). The role of centromere alignment in meiosis I is not known. However, this alignment could be important in facilitating meiosis I segregation.

There have been several attempts to investigate the role of centromere alignment in meiosis of Saccharomyces cerevisiae. KOLLER et al. 1996 found that an ~15-kbp insertion adjacent to CEN15 had a small, but noticeable, effect on that chromosome's ability to segregate from two homologous normal copies of chromosome XV in a trisomic strain. Their results suggested that perturbing centromere alignment might affect segregation. In other experiments, centromeres were misaligned by translocation, and meiotic segregation was examined in heterozygous diploids containing one pair of homologues with one translocated and one normal centromere. The results indicated that misaligning centromeres by 20–25 kbp had no observable effect on segregation (SUROSKY and TYE 1988 ; V. GUACCI and D. KABACK, unpublished results). Unfortunately, the effect of larger misalignments could not be studied because crossovers in the region between the misaligned centromeres produced dicentric and acentric chromosomes that could not be followed during meiosis.

Chromosome bisections may be useful for examining the effect of larger centromere misalignments on meiotic chromosome segregation. S. cerevisiae chromosomes can be bisected into functional chromosome fragments using homologous recombination with small linear centromere-containing plasmids (ZAKIAN et al. 1986 ; GUACCI and KABACK 1991 ). This method places a new centromere on one of the two fragments close to the point of bisection. Bisection at the centromere produces fragments with centromeres that are both aligned with the centromere on an intact copy of that chromosome. Bisection in the middle of a chromosome arm produces fragments where one centromere is misaligned and one is aligned with the centromere on an intact copy of the chromosome. Bisections of chromosome I at its centromere and at two sites on the left arm have been described (GUACCI and KABACK 1991 ; KABACK et al. 1999 ). Studies on haploids containing each of these bisections and on diploids homozygous for each of these bisections indicated that the fragments behaved as functional chromosomes and segregated with relatively high efficiency during mitosis and meiosis I and II (GUACCI and KABACK 1991 ; KABACK et al. 1992 , KABACK et al. 1999 ). Studies on diploids that were heterozygous for another centromere bisection chromosome, where both centromeres were aligned with the centromere on the intact chromosome, indicated that both fragments efficiently segregated away from the intact chromosome during meiosis I (ZAKIAN et al. 1986 ).

To investigate the role of centromere alignment in meiosis I, segregation was studied in heterozygous diploid strains of S. cerevisiae containing one copy of chromosome I bisected into two functional fragments and one intact copy of chromosome I. The centromere of one fragment was aligned with the centromere on the intact homologous chromosome, while the location of the centromere on the other fragment was either aligned or misaligned by 50 or 100 kbp. In these constructs, meiotic crossing over in the region between misaligned centromeres does not produce dicentric and acentric chromosomes, making it possible to follow segregation and analyze the role of centromere alignment in meiosis I segregation. The results indicated that misligned centromeres did not promote normal segregation.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Growth and genetic analysis:
Strains used in this study are listed in Table 1. Growth and sporulation media were described previously (ROSE et al. 1990 ). Ura3^- colonies were selected using synthetic medium containing 5-fluoro-orotic acid (5-FOA; BOEKE et al. 1984 ). Tetrad analysis of chromosome I markers was carried out as described previously (ROSE et al. 1990 ; KABACK et al. 1992 , KABACK et al. 1999 ). Partial disomy was analyzed by plating strains on 5-FOA medium to select for colonies that lost the URA3-marked chromosome fragment and scoring the resultant colonies for uncovered recessive markers. Statistical significance was determined by chi-square tests. Unless mentioned otherwise, P is the probability that results were due to chance.

Recombinant plasmids and yeast transformation:
Recombinant plasmids were prepared by standard protocols (SAMBROOK et al. 1989 ). Plasmid YCp70 was described previously (AGUILERA and KLEIN 1990 ) and is a CEN plasmid derived from plasmid YCp50 by substituting LEU2 for URA3. Plasmids pLF237 (BARTON et al. 1997 ), pLF251, and pLF278 (SU 1998 ) carried chromosome I DNA in plasmid pBS(KS)⁺. Plasmid pCG106 (GUERRA 1995 ) was constructed by introducing a NotI site between the NruI and AatII sites on pBR322 and introducing restriction fragments containing a Y' fragment (0.6-kbp NotI-EcoRI fragment), ARS1 (0.8-kbp EcoRI-HindIII fragment), ARG4 (3.0-kbp HindIII fragment), CEN15 (4.3-kbp HindIII fragment), and another Y' fragment (1.6-kbp HindIII-NotI fragment). Yeast cells were transformed using the lithium acetate/DMSO method (HILL et al. 1991 ). Blot hybridization was carried out as described by SOUTHERN 1975 utilizing Hybond nylon membranes (Amersham Life Science, Arlington Heights, IL) and ³²P-labeled DNA probes produced using Multiprime (Amersham Life Science). Blots were analyzed on a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Electrophoretic karyotyping:
Intact chromosomal DNA was prepared as described (ROSE et al. 1990 ) and separated by pulsed field gel electrophoresis (PFGE) on 1.0% (w/v) agarose in buffer containing 45 mM Tris-borate, pH 8, 1 mM EDTA using a CHEF-DR III system (Bio-Rad Laboratories, Hercules, CA). "Lambda ladder" (New England Biolabs, Beverly, MA) was used as a molecular weight standard. Electrophoresis was performed at 14° for 22 hr using 25-sec pulses at 6 V/cm at an included angle of 120°. Under these conditions, only chromosomes <500 kbp are resolved. Gels were stained for 30 min in 1.0 µg/ml (w/v) ethidium bromide. Prior to blotting, gels were irradiated with 18 mJ ultraviolet light using a Stratalinker (Stratagene, La Jolla, CA).

Translocation of the leftmost 50 kbp of chromosome I next to CEN1:
5-FOA-resistant colonies from strain JL48-1A that contains chromosome I bisected at YAL049 into fragments IA-60 and IB-180 were screened by PFGE and Southern blot hybridization. Homologous mitotic exchange between the CEN1 regions of fragments IA-60 and IB-180 generated a 150-kbp translocation chromosome and a URA3-containing fragment that was lost. A colony (JL48-1A-F3) that contained a copy of the 150-kbp translocation and a copy of fragment IB-180 was identified and crossed with strain CG362-2D, which contains a copy of chromosome I bisected at CEN1 into fragments IL and IR (Table 1). Spores from this cross were screened by PFGE and Southern blot hybridization to identify one containing a copy of the 150-kbp translocation chromosome (Tx) and a copy of a 110-kbp chromosome fragment (IL'). Fragment IL' was produced by meiotic recombination between fragments IB-180 and IL and contained the YAL049-CEN1 interval and URA3 (strain CG366-11C).

Construction of chromosome fragment IA[CEN4 LEU2]-60:
Strain CG325-114B, which contains a copy of chromosome I bisected at YAL049 into fragments IA-60 and IB-180, was transformed with SphI-HindIII-digested YCp70. Ura^- Leu⁺ transformants were screened by conventional agarose gel electrophoresis and Southern blot hybridization to identify one in which CEN1 URA3 on fragment IA-60 was replaced by CEN4 LEU2.

Chromosome I trisection at YAL049 and CEN1:
Diploid strain CG365, which contains chromosome fragments IA [CEN4 LEU2]-60, IB-180, IL, and IR, was sporulated and Ura⁺ Leu⁺ spores were screened by PFGE and Southern blot hybridization to identify those that carried a copy of fragments IA [CEN4 LEU2]-60, IL', and IR. Fragment IL' was produced by meiotic recombination between fragments IB-180 and IL as described above. Strains containing reconstituted copies of fragment IB-180 from fragments IL' and IR were isolated as described previously (GUACCI and KABACK 1991 ).

Construction of a diploid hemizygous for the YAL049-CEN1 interval:
5-FOA-resistant colonies of diploid strain CG410 were screened by PFGE to identify one that lost fragment IL'.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Meiotic segregation of homologous chromosomes containing aligned centromeres:
The meiotic behavior of a heterozygous chromosome I bisection where the centromeres of both fragments were aligned with the centromere of the intact chromosome was investigated. Diploid strain CG403 contains one copy of chromosome I bisected at its centromere into functional fragments IL and IR and one intact copy of chromosome I (Figure 1A). Meiosis I segregation of both fragments from the full-length chromosome produces asci that almost always contain four viable spores. Nondisjunction of one of the fragments produces asci with two viable partially disomic spores and two inviable partially nullisomic spores. The results indicated that both fragments IL and IR segregated efficiently from the intact chromosome (94.8 and 87.9%, respectively; Figure 1B). These results are similar to those obtained using other chromosomes bisected at their centromeres (ZAKIAN et al. 1986 ; SUROSKY and TYE 1988 ; GUERRA 1995 ).

View larger version (37K): In this window In a new window Download PPT slide   Figure 1. Meiotic segregation of aligned centromeres in a diploid heterozygous for a chromosome I bisection at CEN1. (A) Physical maps of chromosome I and PFGE analysis of the haploid parents of strain CG403. CEN1 is indicated by the solid circle, pURA3 denotes the integrated linear bisection plasmid YLpVG47 (GUACCI and KABACK 1991 ), iARG4 and iHIS3 denote insertions produced by one-step gene replacement (KABACK et al. 1999 ). PFGE used conditions that only resolved chromosomes <500 kbp. (B) Meiotic segregation of chromosome fragments IL and IR from the intact copy of chromosome I. Segregation was scored using spore viability. Nullisomic inviable spores are shaded. Nondisjunction of chromosome fragments IL and IR was scored using URA3 (U) and ADE1 (A), respectively. All aneuploidy was confirmed by PFGE. PFGE of all viable spores from a representative ascus is shown below each tetrad class. Disjunction is the percentage of asci segregating each chromosome fragment from the intact copy of chromosome I. Totals are the combined percentage disjunction from each tetrad class showing segregation of that fragment from the intact chromosome. Chromosome I segregation could not be classified in a total of 55 asci: 6 exhibited aberrant (4:0, 3:1, or 1:3) URA3 segregation, 20 contained one viable spore with an intact copy of chromosome I and one viable spore with a bisected copy of chromosome I, 6 contained two viable euploid spores with either bisected or intact copies of chromosome I, and 23 contained either one or no viable spores. () Includes 55 asci with three viable spores.

Meiotic recombination levels were analyzed along the length of chromosome I and were found to be slightly greater than the published values (data not shown). Chromosome I fragments that contained a crossover segregated from the full-length chromosome I in 93–96% of the asci (Table 2A). Fragments containing no detectable crossovers segregated from the intact chromosome in 70–77% of the asci (Table 2A). Segregation of the nonrecombinant fragments was significantly less efficient than recombinant fragments (P < 0.0001) but was not random (P < 0.01). This nonrandom segregation could be due to either distributive disjunction or the segregation of chromosomes containing undetected two-strand double crossovers, possibilities that were not investigated. These results demonstrate that bisection fragments that contain centromeres that are both aligned with the centromere on the full-length chromosome segregate relatively efficiently from the full-length chromosome and that the most efficient disjunction is associated with crossing over.

Meiotic segregation of homologous chromosomes containing misaligned centromeres:
To determine whether homologous centromere alignment is important in meiosis I disjunction, segregation was analyzed in a strain (CG391) that contains one copy of chromosome I bisected near the MAK16 gene and one full-length copy of chromosome I. In this strain, the centromere of the left bisection fragment (IA-110) is located ~50 kbp from the centromere on the intact copy of chromosome I, while the centromeres on the right fragment (IB-140) and the intact chromosome are aligned (Figure 2A). Since a crossover between the left arm of fragment IB-140 and the intact chromosome leads to a reconstituted chromatid, these analyses followed the centromeres of each fragment and the intact chromosome (Figure 2B). The results indicated that the aligned centromere of fragment IB-140 segregated from the centromere of the intact copy of chromosome I in 95.4% of the asci and underwent nondisjunction in only 4.5% of the asci. In contrast, the misaligned centromere of chromosome IA-110 segregated from the centromere on the intact chromosome in only 64.6% of the asci and underwent nondisjunction in 28.0% (= 6.4% + 21.6%) of the asci. Segregation of the IA-110 centromere could not be determined in 7.3% of the asci because they exhibited a crossover adjacent to CEN1 on the right arm and segregated ADE1, the chromosome I centromere marker, at the second meiotic division. As there is no a priori reason to believe that this crossover would influence the segregation of fragment IA-110, exclusion of this class should not affect the conclusions. Thus, segregation of the misaligned IA-110 centromere from the centromere on the intact chromosome was significantly less efficient (P < 0.0001) than that observed for the aligned centromeres, but, nevertheless, also was not random (P < 0.0001).

View larger version (45K): In this window In a new window Download PPT slide   Figure 2. Meiotic segregation of misaligned centromeres in a diploid heterozygous for a chromosome I bisection at MAK16. (A) Physical maps of chromosome I and PFGE analysis of haploid parents of strain CG391. The pURA3 denotes the integrated linear bisection plasmid YLpVG59 (GUACCI and KABACK 1991 ). Other symbols are explained in the legend to Figure 1. (B) Meiotic segregation of centromeres on chromosome fragments IA-110 and IB-140 from the centromere on the intact copy of chromosome I. Segregation was determined using spore viability and the centromere-linked URA3 (U) marker. Inviable spores due to nullisomy are shaded. Crossing over between the misaligned centromeres on the left arm of chromosome IB-140 was indicated by second-division segregation (SDS) of iARG4 with respect to URA3. First-division segregation (FDS) of ADE1 (A) was used for scoring nondisjunction of the fragment IB-140 centromere in asci that did not show a crossover between misaligned centromeres and for scoring disjunction of the fragment IB-140 centromere from the centromere on the intact chromosome in asci that showed a crossover between the misaligned centromeres. Segregation of the chromosome IA-110 centromere with respect to the centromere on the intact chromosome was ambiguous and could not be determined (ND) when ADE1 showed SDS and there was a crossover between the misaligned centromeres. Four-strand double crossovers (NPD for iARG4-URA3) were detected between the misaligned centromeres in 6 asci. PFGE indicated that 2 fell into the chromosome fragment cosegregation class, while 4 showed nondisjunction of fragment IA-110. These asci were included in the totals showing recombination between misaligned centromeres. Disjunction is the percentage of asci segregating each chromosome fragment centromere from the centromere of the intact copy of chromosome I. Chromosome I segregation could not be classified in 52 asci due to spore inviability not caused by nondisjunction and 6 asci due to aberrant, 4:0, 3:1, or 1:3 URA3 segregation. () Includes 15 asci with three viable spores.

Crossing over disrupts meiosis I segregation of chromosomes containing misaligned centromeres:
To investigate why the misaligned centromere of fragment IA-110 segregated less efficiently than the aligned centromeres, the asci from strain CG391 were analyzed for crossovers between all chromosome I markers. Surprisingly, the analysis revealed that crossovers in the region between the misaligned centromeres (iARG4-CEN1) on the left arm of fragment IB-140 were correlated with random segregation of the misaligned centromere on fragment IA-110 with respect to the centromere on the intact chromosome (18.9% segregation vs. 21.6% nondisjunction; Figure 2B). Furthermore, both recombinant and nonrecombinant copies of fragment IA-110 segregated randomly in these asci (Table 2B). In contrast, asci with no crossovers in the iARG4-CEN1 interval exhibited relatively efficient segregation of the fragment IA-110 centromere with respect to the centromere on the intact copy of chromosome I (45.7% segregation vs. 6.4% nondisjunction; Figure 2B). In this group, recombinant copies of fragment IA-110 segregated from the centromere on the intact chromosome in 90% of the asci, while fragments with no detectable crossovers segregated in only 71% of the asci (Table 2B). Segregation of the apparent nonrecombinant copies of fragment IA-110 was not random (P < 0.025). As above, this behavior could be due to either distributive disjunction or the segregation of chromosomes containing undetected two-strand double crossovers, possibilities that were not investigated.

In these studies, nondisjunction or recombination caused a lethal nullisomy that could not be assayed directly. Accordingly, these experiments were repeated on a chromosome I bisection that had a functional fragment with no essential genes where all four spores from most asci were viable and karyotypes could be directly analyzed. Diploid strain CG347 contains a full-length and a bisected copy of chromosome I composed of fragment IA-60, which contains no genes essential for vegetative growth, and fragment IB-180 (Figure 3A; Y. SU, J. LAMB and D. B. KABACK, unpublished results). In strains homozygous for this bisection, homologous copies of fragment IA-60 efficiently recombine and segregate from each other (KABACK et al. 1999 ). In strain CG347, the centromere of fragment IA-60 is located ~100 kbp from the centromere on the full-length chromosome, whereas the centromeres of fragment IB-180 and the full-length chromosome are aligned. Analysis of chromosome I segregation was carried out as described in Figure 3B. On the basis of spore viability, which was comparable to isogenic diploids containing two intact copies of chromosome I (data not shown), the aligned fragment IB-180 centromere segregated from the centromere on the intact copy of chromosome I in virtually all nuclei. In contrast, the misaligned IA-60 centromere appeared to segregate randomly with respect to the centromere on the intact chromosome I (54.0% segregation vs. 46.0% nondisjunction; Figure 3B). Segregation of the fragment IA-60 centromere could not be analyzed in 25 asci (6.1% of the total analyzed) where ADE1 segregated at the second meiotic division (not shown). As above, omission of this class should not affect the conclusions.

View larger version (44K): In this window In a new window Download PPT slide   Figure 3. Meiotic segregation of misaligned centromeres in a diploid heterozygous for a chromosome I bisection at YAL049. (A) Physical maps of chromosome I and PFGE of haploid parents of strain CG347. The pURA3 denotes the integrated linear bisection plasmid YLpLF273, iTRP1 denotes an insertion, and fun30 denotes fun30::LEU2, both produced by one-step gene replacement (KABACK et al. 1992 , KABACK et al. 1999 ). See the legend to Figure 1 for other symbols. (B) Meiotic segregation of centromeres on chromosome fragments IA-60 and IB-180 from the centromere on the intact copy of chromosome I. Segregation of fragment IA-60 centromere was determined using the pURA3 (U) marker. Segregation of the fragment IB-180 centromere was determined using ADE1 (A) and spore viability. Crossing over between the misaligned centromeres on the left arm of chromosome IB-180 was scored using segregation of fun30, cdc24, or iTRP1 with respect to URA3. Segregation could not be classified in 75 asci due to either SDS of ADE1 with respect to URA3 (25 asci), aberrant URA3 segregation (29 asci), or spore inviability (2:2 and 1:3 viable:inviable; 21 asci). () Includes 64 asci with three viable spores. () Includes 47 asci with three viable spores.

Meiotic reciprocal recombination was monitored over the intervals shown in Figure 3A. The frequency of crossing over was approximately equal to that found for two intact chromosomes (data not shown). Both recombinant and nonrecombinant copies of fragment IA-60 segregated randomly with respect to the centromere on the full-length chromosome. The left arm of fragment IB-180 underwent an observable exchange with the intact homologue in 88% of the asci. Correcting for the assumption that two-strand double crossovers produce asci categorized in the nonrecombinant class, the left arm of fragment IB-180 probably underwent crossing over in >90% of the nuclei. Accordingly, random segregation of the fragment IA-60 centromere must also be coincident with the occurrence of crossovers in the region separating the misaligned centromeres in most asci. The fact that the segregation class for fragment IA-60 was slightly greater than the nondisjunction class may be the result of the small percentage of asci that did not contain a crossover on the left arm of fragment IB-180. Thus, these results are identical to those obtained with the larger fragment IA-110. Therefore, misaligned centromeres segregate randomly when there is a crossover in the region separating them.

Centromere realignment restores meiotic segregation of homologous chromosomes:
Fragment IA-60 is located near the end of chromosome I and crossovers that occur near telomeres do not promote meiosis I disjunction (ROSS et al. 1996 ). To rule out the possibility that the fragment's proximity to the telomere and not centromere misalignment was responsible for its random segregation, the fragment IA-60 centromere was realigned with the centromere on its homologue. If telomere proximity was responsible for random segregation, the realigned fragment should still segregate randomly. Alternatively, if centromere misalignment caused random segregation, the realigned centromeres should now segregate more efficiently. Realignment was accomplished by translocating the sequences contained on fragment IA-60 to the region adjacent to the normal chromosome I centromere (Figure 4) and constructing heterozygous bisection strain CG387 (Figure 5A). In this strain, the bisected chromosome fragments consist of a modified version of fragment IA-60 (fragment IA[CEN4 LEU2]-60) and fragment IR, while the translocation (Tx) is equivalent to the intact chromosome in the other experiments. The rest of the left arm of chromosome I is contained on a separate functional chromosome fragment IL' present in two copies. Analysis of segregation in this strain (Figure 5B) indicated that fragment IA[CEN4 LEU2]-60 mostly segregated from the translocation chromosome (67.5% disjunction vs. 32.5% nondisjunction, P = 0.0001). Most of the nondisjunction could be attributed to the failure of this fragment to undergo crossing over in approximately half of the asci. The recombinant copies of fragment IA[CEN4 LEU2]-60 segregated from the translocation chromosome in 89% of the asci (Table 3A). Accordingly, fragment IA[CEN4 LEU2]-60 is capable of relatively efficient segregation and there appear to be no major problems with the ability of chiasmata to promote its disjunction from the intact chromosome. In this strain the modified fragment IA-60 behaved like other recombinant bisection fragments containing aligned centromeres. Thus, random segregation of the fragment IA-60 in strain CG347 was due to its misalignment.

View larger version (45K): In this window In a new window Download PPT slide   Figure 4. Physical maps of chromosome I rearrangements. (A) Wild-type strain CG403-3C. (B) Bisection at YAL049 in strain JL48-1A. (C) Translocation of the leftmost 60 kbp to CEN1 in strain CG366-11C. (D) Trisection at YAL049 and CEN1 in strain CG386-66C. CEN1 is indicated by the solid circle. CEN4 is indicated by the open circle. Rearranged regions are differentially shaded. Other symbols are explained in the legend to Figure 1. Solid boxes (1–3) in A indicate approximate location of hybridization probes. Probe 1 is a 1.0-kbp XbaI fragment from plasmid pLF251, probe 2 is a 1.2-kbp HindIII fragment from plasmid pLF278, and probe 3 is a 3.2-kbp BamHI-HindIII fragment from plasmid pLF237. (E) PFGE karyotypes of strains shown in A–D (lanes A–D). (F) Blot hybridization of strains shown in A–D using probes 1–3. Autoradiograms demonstrate that probes hybridize to the appropriate size chromosomes and chromosomal fragments depicted in A–D (lanes A–D).

View larger version (50K): In this window In a new window Download PPT slide   Figure 5. Meiotic segregation of realigned centromeres in a diploid heterozygous for the chromosome I translocation. (A) Physical maps of chromosome I and PFGE analysis of haploid parents of strain CG387. Symbols are explained in the legends to Figure 1 and Figure 4. (B) Meiotic segregation of chromosome fragment IA-60 from the translocation chromosome (Tx). Segregation was determined using the centromere markers LEU2 (L) and iTRP1 (T) and spore viability. PFGE from a representative ascus is shown below each tetrad class. Segregation could not be analyzed in 235 asci (not shown) because they contained two or fewer viable spores, probably due to nondisjunction of either chromosome IR or IL' (105 asci), did not exhibit 2:2 segregation for LEU2 or iTRP1 (74 asci), or exhibited SDS of LEU2 with respect to iTRP1 (56 asci), due to precocious separation of sister chromatids or ectopic recombination. () Includes 37 asci with three viable spores. () Includes 33 asci with three viable spores.

Eliminating crossovers in the region between misaligned centromeres restores segregation:
Random segregation of chromosome fragments with misaligned centromeres was coincident with crossovers in the region between the misaligned and aligned centromeres. To determine the effect of eliminating crossovers in this region, the left arm of bisection fragment IB-180 was deleted, and meiotic segregation of the chromosome I fragments was studied. Strain CG410-F3 contains fragment IA[CEN4 LEU2]-60, fragment IR, and the full-length chromosome I (Figure 6A). Accordingly, it is hemizygous for the region between YAL049 and CEN1 on the left arm and is expected to exhibit 2:2 segregation for viability. The centromere of fragment IR was aligned with the centromere of the full-length copy of chromosome I, while the centromere of fragment IA[CEN4 LEU2]-60 was misaligned. Meiosis I disjunction of the fragment IA[CEN4 LEU2]-60 from the intact chromosome causes it to segregate into the two inviable spores, whereas nondisjunction causes it to cosegregate with the intact chromosome into the two viable spores (Figure 7). Tetrad analysis indicated that fragment IA[CEN4 LEU2]-60 segregated from the full-length copy of chromosome I in a total of 74.9% of the asci. Recombinant copies disjoined 88% of the time, while nonrecombinant copies disjoined 62% of the time (Table 3B). The observed segregation was almost assuredly not due to chance (P < 0.0001). Therefore, the absence of homology and crossing over in the region separating aligned and misaligned centromeres enables the fragment with the misaligned centromere to behave identically to chromosome fragments containing aligned centromeres.

View larger version (32K): In this window In a new window Download PPT slide   Figure 6. Deletion of homologous sequences between the misaligned centromeres. Physical maps of chromosome I. (A) Deletion strain CG410-F3. Solid boxes (1–3) indicate approximate location of hybridization probes (see legend to Figure 4 for description). Other symbols are explained in the legend to Figure 1. (B) Control, strain CG414. (C) PFGE karyotypes of normal euploid control (wt, strain CG403-3C) and strains CG410-F3 (lane A) and CG414 (lane B). (D) Blot hybridization of wild type and strains shown in A and B using probes 1–3. Autoradiograms demonstrate that probes hybridize to the appropriate size chromosomes and chromosomal fragments.

View larger version (37K): In this window In a new window Download PPT slide   Figure 7. Meiotic segregation of misaligned centromeres in a diploid partially hemizygous for chromosome I. (A) Physical map of chromosome I and PFGE analysis of diploid strain CG410-F3. All symbols are described in the legends to Figure 1 and Figure 4. (B) Meiotic segregation of chromosome fragments IA-60 and IR from an intact copy of chromosome I was determined from segregation of their respective centromere markers LEU2 (L) and ADE1 (A). Inviable spores due to partial nullisomy are shaded. PFGE of all viable spores from a representative ascus is shown below each tetrad class. Only asci showing FDS of the TRP1 gene with respect to spore viability were used. Approximately 25% of all the asci showed SDS of ADE1 with respect to TRP1. Segregation of chromosome IR was determined in these asci by PFGE analysis where necessary. Disjunction is the percentage of asci segregating each chromosome fragment from the intact copy of chromosome I. A total of 47 asci could not be analyzed due to aberrant segregation of LEU2 (18 asci), TRP1 (5 asci), or spore inviability (24 asci with one or no viable spores).

As a control, strain CG414 was analyzed (Figure 6B). This strain was isogenic to strain CG410-F3 except it was not a hemizygote and contained a reconstituted copy of fragment IB-180 in place of fragment IR. For all purposes, it was identical to the heterozygous bisection strain CG347, except for the presence of fragment IA[CEN4 LEU2]-60 in place of IA-60. The results indicate that chromosomes in strain CG414 behaved identically to those in strain CG347. Most asci contained four viable spores, crossing over on the left arm of fragment IB-180 occurred in >90% of the asci, and the amount of recombination on fragment IA[CEN4 LEU2]-60 was very similar to that found at fragment IA-60 (data not shown). In contrast to strain CG410-F3, fragment IA[CEN4 LEU2]-60 segregated randomly with respect to the centromere on the full-length chromosome I (52% disjunction vs. 48% nondisjunction; Table 3C). These results confirm that misaligned centromeres segregate randomly when there is a crossover in the region separating them. They also demonstrate the equivalence of fragments IA[CEN4 LEU2]-60 and IA-60.

Recombinant chromosome fragments containing misaligned centromeres do not behave as univalents:
Random segregation of a misaligned centromere may be the result of the chromosome fragments behaving as if they were univalents. Univalent chromosomes will segregate from either an unpaired nonhomologous chromosome or a CEN plasmid by distributive disjunction (DAWSON et al. 1986 ; MANN and DAVIS 1986 ; KABACK 1989 ; GUACCI and KABACK 1991 ). To determine whether fragment IA-60 behaved as a univalent and underwent distributive disjunction, CEN plasmid pCG106 was introduced into the chromosome I bisection heterozygote, CG347. The plasmid had no detectable effect on spore viability or meiotic recombination on chromosome I (data not shown). Tetrad analysis showed that when there was a crossover on fragment IA-60, the fragment did not exhibit distributive disjunction but segregated randomly with respect to both the CEN plasmid (Table 4) and the centromere on the intact chromosome (data not shown). These results indicate that recombinant fragments with a misaligned centromere did not behave as univalents. In contrast, nonrecombinant copies of fragment IA-60 segregated from the plasmid in 88% of the analyzable tetrads (Table 4; P < 0.0001), but randomly with respect to the centromere on the intact copy of chromosome I (data not shown). Therefore, the nonrecombinant fragments appeared to behave as univalents, suggesting they segregate randomly due to a failure to properly pair with their homologues.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The role of centromere alignment in meiotic chromosome segregation in S. cerevisiae was investigated using diploids that contained one intact copy of chromosome I and one copy bisected into two functional centromere-containing fragments. The centromere on one fragment was aligned with the centromere on the intact chromosome, while the centromere on the other fragment was either aligned or misaligned by 50 or 100 kbp. Aligned centromeres segregated from each other efficiently, while misaligned centromeres segregated from each other much less efficiently. In fact, the centromere that was misaligned by ~100 kbp segregated randomly. Random segregation of all misaligned centromeres was correlated with crossovers between the other chromosome fragment and the intact chromosome in the region separating the centromeres. When there were no crossovers in the region between the misaligned centromeres or when the DNA in this region was deleted on one homologue to prevent recombination, segregation appeared to be as efficient as that observed for aligned centromeres. Therefore, these crossovers are preventing disjunction of a normal chromosome from its homologous fragment.

Segregation of chromosome fragments was scored by analysis of viability in both the heterozygous CEN1 and MAK16 bisections. Tetrad classes were assigned on the basis of assuming normal meiosis I and II segregation of marker pairs on chromosomes and demonstrating disomy in the viable spores when there was nondisjunction. While there are no reasons to believe these assumptions are incorrect, the experiments using fragment IA-60 did not use spore viability as a marker and the results clearly demonstrated that the fragment segregated randomly when its centromere was misaligned and segregated correctly when its centromere was realigned. Therefore, the fragment IA-60 results are in complete agreement with those obtained using the other bisections.

In the experiments where the CEN1-linked ADE1 gene was used to follow first-division segregation, 6–7% of the asci were excluded because there was a crossover between ADE1 and CEN1. This fraction of asci is close to that expected for a ~5-cM genetic interval. There is no a priori reason to believe that exclusion of this class affected any of the results presented. Furthermore, recombination on the same arm distal to ADE1 did not appear to affect segregation of the fragment containing the misaligned centromere. Nevertheless, we cannot completely eliminate the possibility that the excluded asci might quantitatively affect the results. Even if the excluded class fell entirely into the disjunction class for the misaligned centromere-containing fragment, the results would still show that misaligned centromeres segregate more poorly than aligned centromeres when there was a crossover in the region separating them.

When centromeres were aligned or there was no crossover in the region between misaligned centromeres, fragment disjunction appeared to depend on recombination with the intact chromosome. Most if not all of the nondisjunction was associated with the failure to cross over. The number of chromosomes that failed to cross over was consistent with the smaller size of the bisection chromosomes. In contrast, when centromeres were misaligned and a crossover occurred in the region between the misaligned centromeres, recombination between the intact chromosome and the misaligned centromere-containing fragment did not appear to affect its disjunction from the intact chromosome. The nonrecombinant fragments with misaligned centromeres segregated randomly but underwent distributive disjunction with a CEN plasmid, suggesting that these fragments might not be paired with their homologues. Recombinant fragments with misaligned centromeres segregated randomly but did not undergo distributive disjunction with a plasmid, suggesting that these fragments were paired with their homologues. These results demonstrate that pairing and recombination per se are not sufficient to guarantee correct disjunction.

The results demonstrate that misaligned homologous centromeres will segregate randomly when there is a crossover in the region between the centromeres. Therefore, such a crossover appears to prevent normal segregation. How segregation is prevented is not known but two mechanisms may explain the behavior. It is possible that misaligned centromeres segregate randomly because they may be equally capable of stably attaching to either pole of the meiotic spindle (Figure 8A). Stable attachment of kinetochores to microtubules is believed to require the tension generated by pulling against the chiasmata that hold homologues together (NICKLAS 1997 ). A crossover in the region between the misaligned centromeres might prevent normal segregation of the fragment containing the misaligned centromere because its attachment to either side of the spindle generates approximately the same amount of tension. This tension must be generated by pulling on the chiasma between the other fragment and the intact chromosome, since there is no opposing force generated by the chiasma between the fragment containing the misaligned centromere and the intact chromosome. In contrast, when there is no crossover in the region between the misaligned centromeres (or when centromeres on bisected chromosomes are both aligned with the centromere on the intact chromosome), the most efficient way to generate tension between each kinetochore and chiasmata holding the homologues together causes the two fragments to cosegregate from the intact chromosome. Other spindle attachment arrays would generate less or unequal tension, and microtubule attachments to the spindle would not be stable enough to promote segregation.

View larger version (16K): In this window In a new window Download PPT slide   Figure 8. Mechanisms for segregating chromosome fragments containing misaligned centromeres. (A) Tension model: (1) Crossovers between misaligned centromeres produce random segregation of the misaligned centromere-containing fragment. Attachment of the misaligned centromere to either spindle pole produces approximately equal tension because the tension is generated from the chiasmata between the intact chromosome and the other chromosome fragment (). (2) The absence of crossovers between the misaligned centromeres produces proper disjunction of both fragments from the intact chromosome because tension () can be generated only when the centromeres of the two fragments attach to the same pole of the spindle while the centromere on the intact chromosome attaches to the opposite pole. Attachment of the misaligned centromere to the same pole as the intact chromosome would not generate sufficient tension to stabilize the microtubules because there would be no nearby opposing force. (B) The centromere interaction model: (1) The absence of crossovers between the misaligned centromeres produces proper disjunction of both fragments from the intact chromosome because centromeres on both fragments can interact either with each other or with the centromere on the intact chromosome (*). These interactions enable both fragments either to cosegregate as a single functional chromosome or to interact with and segregate from the centromere on the intact chromosome. (2) Crossovers between the misaligned centromeres produce random segregation of the misaligned centromere-containing fragment because its centromere is physically occluded from interacting either with the centromere on the other fragment or on the intact chromosome. Solid circles denote the centromere on the intact chromosome, and open circles denote centromeres on the chromosome fragments. Large open arrows indicate kinetochore movement toward the spindle poles. Chiasmata are denoted by where chromosome arms cross. For simplicity, only one chromatid is shown for each chromosome.

This model is consistent with the behavior of paired multivalent translocation chromosomes in higher organisms (RICKARDS 1983 ). Reciprocal translocation heterozygotes produce balanced or unbalanced meiotic products depending on the orientation of the quadrivalent chromosome complex formed by chiasmata between the two translocation and two normal chromosomes. This orientation appears to be dependent on the relative location and orientation of chiasmata and the kinetochores and their presumed ability to generate microtubule tension from both poles of the spindle (RICKARDS 1964 ; NICKLAS 1997 ). Similarly, Robertsonian translocation heterozygotes resemble yeast centromere bisection heterozygotes and heterozygotes containing misaligned centromeres where there are no crossovers in the region separating the centromeres. The Robertsonian translocation chromosome segregates from the two acrocentric chromosomes following the formation of a trivalent where both acrocentric kinetochores are oriented to the same pole, while the kinetochore of the translocation is oriented toward the opposite pole (BAUER et al. 1961 ; SMITH 1965 ; MARKS 1978 ; HAYS et al. 1982 ).

If the ability to generate a stable bipolar orientation on the spindle is the only factor that dictates homologue segregation, then the physical act of aligning centromeres would play little or no role in promoting proper chromosome segregation. Indeed, misaligned centromeres were capable of segregating efficiently as long as their presumed ability to maintain only a single orientation was not perturbed by crossovers in the region separating them. Nevertheless, in the absence of these crossovers, it is still possible that misaligned centromeres were segregating somewhat less efficiently than two aligned centromeres. However, any effects would have been below our detection limits.

Another possibility is that homologous centromeres physically interact with each other, and this interaction is required for segregation (Figure 8B). For example, such an interaction could be involved in organizing kinetochores so they face in opposite directions (COUNCE and MEYER 1973 ; GOLDSTEIN 1981 ). Alternatively, the centromeres on the two bisection fragments could interact with each other and form a complex that behaves as a single kinetochore that promotes cosegregation of the two fragments (MOSES et al. 1979 ; ELDER and PATHAK 1980 ; PATHAK and LIN 1981 ; MAHADEVAIAH et al. 1990 ; DAVISSON and AKESON 1993 ). In either case, a crossover could physically hinder any interaction involving misaligned centromeres. Irrespective of mechanism, chromosome fragments with misaligned centromeres segregate from their homologues much less efficiently than fragments with aligned centromeres. The inability of chromosomes with misaligned centromeres to segregate properly may serve as a control that prevents ectopically recombined chromosomes from segregating. This control may favor the production of balanced viable gametes and increase reproductive fitness.

	ACKNOWLEDGMENTS

We are extremely grateful to Jerome You for skilled technical assistance and to Jacque Lamb, Yuping Su, Eric White, and Arnold Barton for their valued contributions. We also thank Breck Byers for his thoughtful suggestions and Hannah Klein and Carol Newlon for their valuable comments on the manuscript. This work was supported by a grant from the National Science Foundation.

Manuscript received February 2, 1999; Accepted for publication August 5, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AGUILERA, A. and H. L. KLEIN, 1990  HPR1, a novel yeast gene that prevents intrachromosomal excision recombination, shows carboxy-terminal homology to the Saccharomyces cerevisiae TOP1 gene. Mol. Cell. Biol. 10:1439-1451[Abstract/Free Full Text].

BARTON, A. B., H. BUSSEY, R. K. STORMS, and D. B. KABACK, 1997  Molecular cloning of chromosome I DNA from Saccharomyces cerevisiae: characterization of the 54 kb right terminal CDC15-FLO1-PHO11 region. Yeast 13:1251-1263[Medline].

BAUER, H., R. DIETZ, and C. ROBBELEN, 1961  Die spermatocyteneilungen der tipuliden. III Mitteilung. Das bewegungsverhalten der chromosomen in tranlocationsheterozygoten von Tipula oleracea. Chromosoma 12:116-189[Medline].

BOEKE, J. D., F. LACROUTE, and G. R. FINK, 1984  A positive selection for mutants lacking orotidine-5'phosphate decarboxylase activity in yeast: 5-fluoro-orotic acid resistance. Mol. Gen. Genet. 197:345-346[Medline].

COUNCE, S. J. and G. F. MEYER, 1973  Differentiation of the synaptonemal complex and the kinetochores in Locusta spermatocytes studied by whole mount electron microscopy. Chromosoma 44:231-253[Medline].

DAVISSON, M. T. and E. C. AKESON, 1993  Recombination suppression by heterozygous Robertsonian chromosomes in the mouse. Genetics 133:649-667[Abstract].

DAWSON, D. S., A. W. MURRAY, and J. W. SZOSTAK, 1986  An alternate pathway for meiotic chromosome segregation in yeast. Science 234:713-717[Abstract/Free Full Text].

ELDER, F. F. B. and S. PATHAK, 1980  Light microscopic observations on the behavior of silver-stained trivalents in pachytene cells of Sigmodon fulviventer (Rodentia, Muridae) heterozygous for centric fusion. Cytogenet. Cell Genet. 27:31-38[Medline].

GOLDSTEIN, L. S. B., 1981  Kinetochore structure and its role in chromosome orientation during the first meiotic division in male D. melanogaster.. Cell 25:591-602[Medline].

GUACCI, V. and D. B. KABACK, 1991  Distributive disjunction of authentic chromosomes in Saccharomyces cerevisiae.. Genetics 127:475-488[Abstract].

GUERRA, C. E., 1995 Increasing the chromosome number of the budding yeast and the synthetic lethal phenotype of an acc1 hpr1 double mutant. Ph.D. Thesis, New York University Graduate School of Arts and Science, New York.

HAYS, T. S., D. WISE, and E. D. SALMON, 1982  Traction force on a kinetochore at metaphase acts as a linear function of kinetochore fiber length. J. Cell Biol. 93:374-382[Abstract/Free Full Text].

HILL, J., K. A. IAN, and D. E. GRIFFITHS, 1991  DMSO-enhanced whole cell yeast transformation. Nucleic Acids Res. 19:5791[Free Full Text].

KABACK, D. B., 1989  Meiotic segregation of circular plasmid-minichromosomes from intact chromosomes in Saccharomyces cerevisiae.. Curr. Genet. 15:385-392[Medline].

KABACK, D. B., V. GUACCI, D. BARBER, and J. W. MAHON, 1992  Chromosome size-dependent control of meiotic recombination. Science 256:228-232[Abstract/Free Full Text].

KABACK, D. B., D. BARBER, J. MAHON, J. LAMB, and J. YOU, 1999  Chromosome size-dependent control of meiotic reciprocal recombination in Saccharomyces cerevisiae: the role of crossover interference. Genetics 152:1475-1486[Abstract/Free Full Text].

KOLLER, A., J. HEITMAN, and M. N. HALL, 1996  Regional bivalent-univalent pairing versus trivalent pairing of a trisomic chromosome in Saccharomyces cerevisiae.. Genetics 144:957-966[Abstract].

MAHADEVAIAH, S. K., L. A. SETTERFIELD, and U. MITTWOCH, 1990  Pachytene pairing and sperm counts in mice with single Robertsonian translocations and monobrachial compounds. Cytogenet. Cell Genet. 53:26-31[Medline].

MANN, C. and R. W. DAVIS, 1986  Meiotic disjunction of circular minichromosomes in yeast does not require DNA homology. Proc. Natl. Acad. Sci. USA 83:6017-6019[Abstract/Free Full Text].

MARKS, G. E., 1978  The consequences of an unusual Robertsonian translocation in celery (Apium graveolens var dulce). Chromosoma 70:141-160.

MOSES, M. J., P. A. KARATSIS, and A. E. HAMILTON, 1979  Synaptonemal complex analysis of heteromorphic trivalents in Lemur hybrids. Chromosoma 70:141-160.

NICKLAS, R. B., 1997  How cells get the right chromosomes. Science 275:632-637[Abstract/Free Full Text].

PATHAK, S. and C. C. LIN, 1981  Synaptonemal complex of the sex-autosome trivalent in a male indian Muntjac. Chromosoma 82:367-376[Medline].

RICKARDS, G. K., 1964  Some theoretical aspects of selective segregation in interchange complexes. Chromosoma 15:140-155.

RICKARDS, G. K., 1983  Orientation behavior of chromosome multiples of interchange (reciprocal translocation) heterozygotes. Annu. Rev. Genet. 17:443-498[Medline].

ROSE, M. D., F. WINSTON and P. HIETER, 1990 Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

ROSS, L. O., R. MAXFIELD, and D. DAWSON, 1996  Exchanges are not equally able to enhance meiotic chromosome segregation in yeast. Proc. Natl. Acad. Sci. USA 93:4979-4983[Abstract/Free Full Text].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual, Ed. 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SCHERTHAN, H., J. LOIDL, T. SCHUSTER, and D. SCHWEIZER, 1992  Meiotic chromosome condensation and pairing in Saccharomyces cerevisiae studied by chromosome painting. Chromosoma 101:590-595[Medline].

SMITH, E. B., 1965  Preferential segregation from a trivalent in Haplopappus validus.. Can. J. Genet. Cytol. 7:484-490.

SOUTHERN, E., 1975  Detection of specific sequences among the DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517[Medline].

SU, Y., 1998 Gene expression and meiotic recombination in the left subtelomeric region of chromosome I in Saccharomyces cerevisiae. Ph.D. Thesis, University of Medicine and Dentistry of New Jersey, New Jersey Medical School.

SUROSKY, R. T. and B. TYE, 1988  Meiotic disjunction of homologs in Saccharomyces cerevisiae is directed by pairing and recombination of the chromosome arms but not by pairing of the centromeres. Genetics 119:273-287[Abstract/Free Full Text].

ZAKIAN, V. A., H. M. BLANTON, L. WETZEL, and G. M. DANI, 1986  Size threshold for Saccharomyces cerevisiae chromosomes: generation of telocentric chromosomes from an unstable minichromosome. Mol. Cell. Biol. 6:925-932[Abstract/Free Full Text].

This article has been cited by other articles:

				B. Kemp, R. M. Boumil, M. N. Stewart, and D. S. Dawson A role for centromere pairing in meiotic chromosome segregation Genes & Dev., August 15, 2004; 18(16): 1946 - 1951. [Abstract] [Full Text] [PDF]