Sustained and rapid chromosome movements are critical for chromosome pairing and meiotic progression in budding yeast

Abstract

Telomere-led chromosome movements are a conserved feature of meiosis I (MI) prophase. Several roles have been proposed for such chromosome motion, including promoting homolog pairing and removing inappropriate chromosomal interactions. Here, we provide evidence in budding yeast that rapid chromosome movements affect homolog pairing and recombination. We found that csm4Δ strains, which are defective for telomere-led chromosome movements, show defects in homolog pairing as measured in a "one-dot/two-dot tetR-GFP" assay; however, pairing in csm4Δ eventually reaches near wild-type (WT) levels. Charged-to-alanine scanning mutagenesis of CSM4 yielded one allele, csm4-3, that confers a csm4Δ-like delay in meiotic prophase but promotes high spore viability. The meiotic delay in csm4-3 strains is essential for spore viability because a null mutation (rad17Δ) in the Rad17 checkpoint protein suppresses the delay but confers a severe spore viability defect. csm4-3 mutants show a general defect in chromosome motion but an intermediate defect in chromosome pairing. Chromosome velocity analysis in live cells showed that while average chromosome velocity was strongly reduced in csm4-3, chromosomes in this mutant displayed occasional rapid movements. Lastly, we observed that spo11 mutants displaying lower levels of meiosis-induced double-strand breaks showed higher spore viability in the presence of the csm4-3 mutation compared to csm4Δ. On the basis of these observations, we propose that during meiotic prophase the presence of occasional fast moving chromosomes over an extended period of time is sufficient to promote WT levels of recombination and high spore viability; however, sustained and rapid chromosome movements are required to prevent a checkpoint response and promote efficient meiotic progression.

Figure 1.—

Chromosome pairing is defective in csm4Δ. Chromosome pairing was assessed in diploid SK1 strains ectopically expressing tetR-GFP and bearing tet0 arrays at homologous positions (Brar et al. 2009; materials and methods). Homologs were considered paired when only one GFP dot was observed and were considered unpaired when two clear GFP dots were seen (Figure S1). Cells were analyzed using z-stacks to visualize the entire cell volume. (A, B, and C) Representative time courses (chosen from four to eight independent experiments) with tetO arrays at LYS2 (A), CENV (B), and TELV (C). Pairing was considered maximum in these ndt80Δ strains at T = 24 hr. (D and E) Representative time courses demonstrating nonhomologous pairing, with one tetO array at LEU2 and another at CENV (D) and one at LYS2 and another at URA3 (E).

Figure 2.—

csm4-3 is a separation of function allele. (A) Amino acids in the Csm4 polypeptide were mutated to alanine residues as indicated in boldface type. The allele designation is shown above each set of amino acid substitutions. The putative transmembrane domain is underlined near the C terminus. Alleles that conferred a WT phenotype as measured by spore viability and timing of the MI division (materials and methods) are shown in blue; alleles that displayed a null phenotype in these assays are shown in red. Intermediate alleles are shown in black. (B) The phenotypes of csm4 alleles (Table 1) are presented in a graph in which spore viability is plotted vs. MI delay, relative to WT, in completing the MI division. Cells with two, three, or four nuclei were counted as having completed M1 (MI ± MII). The phenotypes of WT (red circle) and csm4Δ (red square) are indicated. csm4-3, a separation-of-function allele, is highlighted (red triangle). (C) Representative time course showing the completion of the MI division (MI ± MII) in WT, csm4Δ, and csm4-3 strains.

Figure 3.—

Analysis of csm4-3 in Nup49-GFP and Zip1-GFP motion assays. (A) Representative time-lapse images of WT, csm4Δ, and csm4-3 cells expressing Nup49-GFP, which marks the nuclear envelope. Images were taken at 1-sec intervals for 30–45 sec. Every other frame is shown for WT; every fourth frame for csm4Δ and csm4-3. Elapsed time is shown in the upper left corner for each frame. (B) Representative time-lapse images of WT, csm4Δ, and csm4-3 strains expressing Zip1-GFP, which localizes to synapsed chromosomes. Images were taken at 1-sec intervals for 45 sec. Every other frame for a portion of the time lapse is shown (see materials and methods). Elapsed time is shown in the upper left corner for each frame. Maximum and mean chromosome velocity for all chromosomes analyzed is shown below each genotype in micrometers per second, on the basis of 30 chromosome measurements from 30 different cells across four time courses.

Figure 4.—

csm4-3 shows a defect in chromosome motion in cells expressing Zip1-GFP. (A) Distributions of average velocities for each chromosome measured (30 each for WT, csm4Δ, and csm4-3). Average values for each chromosome were obtained from 15 to 25 velocity measurements determined in 1-sec intervals. (B) Chromosome velocity measurements in 1-sec intervals presented as a percentage of the total number recorded (678, 654, and 738 measurements for WT, csm4Δ, and csm4-3, respectively).

Figure 5.—

csm4-3 strains display occasional rapid chromosome movements. Chromosome velocity was measured in consecutive 1-sec intervals for three representative chromosomes in WT, csm4Δ, and csm4-3 strains expressing Zip1-GFP.

Figure 6.—

Chromosome pairing is defective in csm4Δ and to a lesser extent csm4-3. Chromosome pairing was assayed as described in Figure 1 for csm4-3 strains. (A, B, and C) Representative time courses with tetO arrays at LYS2 (A), CENV (B), and TELV (C). Pairing was considered maximum in these ndt80Δ strains at T = 24 hr. (D and E) Representative time courses demonstrating nonhomologous pairing, with one tetO array at LEU2 and another at CENV (D) and one at LYS2 and another at URA3 (E). Data from Figure 1 for WT and csm4Δ are shown for comparison purposes.

Figure 7.—

The meiotic delays observed in csm4Δ and csm4-3 are fully rescued by the rad17Δ mutation. Meiosis I completion (MI ± MII) time course experiments are shown for the indicated mutant strains. The spore viability of each strain is shown in parentheses following the genotype. (A and B) Representative time courses showing that a null mutation in RAD17, a DNA-damage checkpoint protein, rescues the meiotic delay of csm4Δ (A) and csm4-3 (B). See materials and methods for details.