Figure 1.
Early RPMs increase the range without increasing the average speed of chromosome movement.
Movements were assessed for telomere-adjacent 4R lacO256/lacI-GFP spots in 50 cells at each time-point, in time-lapse datasets acquired at 1 frame per second for 120 seconds total. Only measurements made on cells with two GFP spots, representing unpaired telomeres 4R, are included. Maximum speed, bias and area are higher at t = 3 (3 hours following the induction of meiosis by transfer into sporulation medium) than at t = 0 (immediately after shift into sporulation medium) but average speed has not yet begun to increase. (A) Histogram of the maximum distances moved in 1 second, for the 119 movements measured for each spot. (B) Histogram of the bias (the average cosine of the angles between each successive movement) for each spot. (C) Histogram of the average speeds for each spot. (D) The median area for the smallest bounding boxes required to enclose all positions of each spot at t3. Areas are measured for 60 rather than 120 second intervals to facilitate comparisons with area measures in Figure 6, Figure S1 and [27]. By this measure, RPMs in mps3-dCC are not significantly different from wild-type at t3 (*) while RPMs are significantly decreased in mps3-dAR and csm4Δ. Statistical analysis is in the legend to Table 1. Strains used: wild-type (MDY1560XMDY1567), mps3-dCC (MDY2580XMDY2759), mps3-dAR (MDY2523XMDY2756), csm4Δ (MDY2609XMDY2778).
Table 1.
Quantified mutant phenotype values.
Figure 2.
An assay for trapping chromosome collisions.
(A) Diagram of an assay designed to trap chromosome collisions. A pair of homologs is represented as blue lines. lacO256 concatemers are decorated with dimerizing lacI-GFP in a control strain and with tetramerizing lacI-GFP in the experimental strain (represented as yellow/green spots in the diagrams). Collision between sites with the concatemers leads to stable formation of a single spot with tetramerizing lacI-GFP; dimerizing lacI-GFP allows subsequent separation of the sites so that two spots are again visible. (B–D) Examples of nuclei (DNA labeled with DAPI, blue) at different stages of meiotic prophase as determined by the pattern of signal from immunolocalized Zip1 (red). The yellow-green spots are lacI-GFP-decorated lacO concatemers at the middle of chromosome arms 5L. Shown are examples of the majority of nuclei at the different stages, i. e., where pairing has not yet occurred in B and C but is evident in D. Nuclei scored as positive in the collision trap assay have Zip1 signal distribution as in B, i. e., have spots but no distinct lines of Zip1, but in addition have only a single GFP spot, as in D.
Figure 3.
RPMs cause chromosome collisions coincident with pairing.
All collision trap assays were performed with pairs of lacO concatemers at interstitial homotopic loci (5L or 7L) or ectopic loci (5L/7L) and with dimerizing (“d”) or tetramerizing (“t”) lacI-GFP. At least 200 nuclei were scored for each data point for the presence of 1 GFP spot, indicating a pairing or collision event, or 2 separate GFP spots, indicating a separation of ∼0.2 µm or more (see Figure 2C–2E). In wild-type and in spo11Δ, where RPMs are robust, tetramerizing lacI-GFP traps collisions that are lost in dimerizing lacI-GFP (A, B). RPM mutants csm4Δ and mps3-dAR reduce the numbers of collisions whether homotopic (C, D) or ectopic (F), and in the absence of meiotic recombination (E). Statistical analyses are described in the text. Strains used are listed in Table S2. (A–D, F) Results for homotopic collisions in wild-type and mutant backgrounds, as labeled. (E) Results for ectopic collisions in wild-type (w. t.) vs. csm4Δ. The levels of single spots in dimerizing lacI-GFP in E set an upper limit on the background from chance colocalization of homotopic spots, and in F for heterotopic spots.
Figure 4.
Delays of pairing/synapsis in RPM mutants vary among chromosome loci.
(A) Diagram of lacO concatemer positions on chromosomes 1, 5 and 7 scored for pairing frequency. The 4R concatemer (not shown) is adjacent to the telomere. (B) Comparison of kinetics of close pairing at different sites on chromosomes 1, 4, 5 and 7. Error bars are 1 standard deviation. Pairing is most delayed in ndj1Δ with mps3-dCC and mps3-dAR appearing to have no or intermediate delays, depending on the site. With only two exceptions, mps3-dAR at 5L telomere and ndj1Δ at 7 centromere, all sites are minimally paired at t3. Statistical analysis of summary estimates of pairing rates (see text) is in the legend to Table 1. Strains used are listed in Table S3.
Figure 5.
RPMs in the mps3-dAR and mps3-dCC mutants are intermediate between ndj1Δ and wild-type early in prophase and approach wild-type levels following pairing when prophase exit is blocked.
Histograms are shown to display the ranges and changes in RPMs with time by measures of maximum speed, average speed and bias. All measurements are for GFP-tagged spots adjacent to the 4R telomere and are accumulated separately for nuclei where there are 2 spots (unpaired, left columns) or 1 spot (paired, right columns). All populations are in early to middle meiotic prophase at 4 hours. To compare with the later 7 hour time-point of ndj1Δ, when many cells in the other strains would already be undergoing the meiotic divisions, ndt80Δ is added to the wild-type, mps3-dCC and msp3-dAR backgrounds to hold the strains in meiotic prophase and allow the development of the fastest possible RPMs. Note that the “bias” measure, the average of the cosines of the angles made between successive movements, is unitless. Statistical analyses of median values from t4 are in the legend to Table 1. Strains used: wild-type (MDY1560XMDY1567), mps3-dCC (MDY2580XMDY2759), mps3-dAR (MDY2523XMDY2756), ndj1Δ (MDY2294XMDY1560), ndt80Δ (MDY2984XMDY3021), ndt80Δ mps3-dCC (MDY3020XMDY3022), ndt80Δ mps3-dAR (MDY3047XMDY3049).
Figure 6.
Pairing rates are positively correlated with RPMs.
Pairing rates (Figure 4, Table 1; change in the percent of the population with 1 (paired) spot rather than 2 (unpaired) spots) and RPM area measures (Table 1; square microns of bounding box that encloses all spot positions in 60 consecutive time-lapse frames) for telomere 4R are graphed to display the behavior of individual loci (A) and of individual genotypes (B). Paired telomeres tend to move faster and further than unpaired telomeres in wild-type, mps3-dCC and mps3-dAR (but the reverse is true for average and maximum speeds in ndj1Δ; see [27], Figure 5 and Figure S1). Error bars in B are average absolute deviation from the median. Strains used are the same as for Figure 4 and are listed in Table S3. (A) Pairing rates for each locus (key at right) graphed against RPM area for unpaired (left) and paired (right) telomeres 4R. Chromosome 1 pairing was determined only for wild-type and mps3-dAR. Lines connect values for the same locus in different genetic backgrounds. (B) Median pairing rates for each genotype (thick black lines) are shown adjacent to the individual values, graphed against RPM area for unpaired (left) and paired (right) telomeres 4R.
Figure 7.
Defects in production of viable spores, proper chromosome segregation, and time to enter the first meiotic division increase with defects in RPMs.
(A) Percent of cells with 3-4, 2-1 or 0 spores (the latter mainly representing cells that failed to enter or complete sporulation) after 2 days in sporulation medium. Final sporulation levels for wild-type (MCY506XMCY507), mps3-dCC (MCY1378XMCY1379), mps3-dAR (MCY1512XMCY1513) and ndj1Δ (MCY422XMCY423) are 83%, 71%, 63% and 74%, respectively. Statistical analysis is in the legend to Table 1. (B) Percent of 4-spored asci with 4, 3, 2, 1 or 0 viable spores. Total spore viabilities for wild-type, mps3-dCC, mps3-dAR and ndj1Δ are 92%, 75%, 87% and 62%, respectively. Statistical analysis is in the legend to Table 1. (C) Meiotic missegregation measured by the presence of an extra chromosome 3 in viable spores. Strains used: wild-type (MDY493XMDY494), mps3-dCC (MCY1370XMCY1380), mps3-dAR (MCY1584XMCY1588) and ndj1Δ (MCY420XMCY421). (D) Premature separation of sister chromatids of chromosome 7 in the first meiotic division assayed cytologically using a GFP-tagged spot adjacent to the centromere. Strains used: wild-type (MDY2828XMDY2798), mps3-dCC (MDY2843XMDY2825), mps3-dAR (MDY2846XMDY2823), ndj1Δ (MDY2826XMDY2820), csm4Δ (MDY3042XMCY1539), mps3-dNT (MDY2834XMDY2821). (E) Entry into the first meiotic division was assayed by DAPI staining of DNA in fixed whole cells as the appearance of stretching apart of the dividing DNA mass during the first meiotic division or the presence of two or more nuclear DNA masses in a single cell. Strains used: wild-type (MCY506XMCY507), mps3-dCC (MCY1378XMCY1379), mps3-dAR (MCY1512XMCY1513), ndj1Δ (MCY422XMCY423), csm4Δ (MCY1536XMCY1539), mps3-dNT (MCY1401XMCY1407). All error bars represent 1 standard deviation.
Figure 8.
In all RPM mutants, short chromosomes frequently remain unsynapsed when long chromosomes have finished synapsis.
Electron micrographs of silver-stained, spread, flattened meiotic prophase nuclei from (A) mps3-dAR (MCY1512XMCY1513) at t6, (B) mps3-dNT (MCY1401XMCY1407) at t8, (C) ndj1Δ mps3-dAR (MCY1510XMCY1511) at t8 and (D) csm4Δ (MCY1536XMCY1539) at t8. Electron-dense lines are the silver-stained chromosome axes which, when aligned in pairs at uniform spacing, mark completed synapsis to form synaptonemal complexes. Long arrows indicate synapsed long chromosomes and short arrows indicate unsynapsed short chromosome axes. Nucleoli are indicated by “nll.” Polycomplexes, which are commonly found in nuclei that are delayed in synapsis, are indicated by “pc.”
Figure 9.
Telomere association with Ndj1 and Csm4, but not with the nuclear envelope, is diminished in mps3-dAR.
(A–D) Immunolocalization of Ndj1 and Csm4 at telomeres appears wild-type in a spread meiotic prophase nucleus from mps3-dCC (MDY2865XMDY2867). (E–H) Immunolocalization in mps3-dAR (MDY2868XMDY2870) reveals that colocalization of Ndj1 and Csm4 at telomeres is infrequent and that spots of Ndj1 frequently are found away from the telomeres. (I–M) Nuclear pores, marked by immunolocalization of Nup49-GFP, remain associated with telomeres in spread preparations of nuclei at high frequencies in wild-type (MCY1438XMCY1439) (I) and csm4Δ (MDY3449XMDY3450) (J) but not in ndj1Δ (MDY2936XMDY2937) (K). Telomere-pore association is frequent in mps3-dAR (MDY2952XMDY2953) (L) unless combined with ndj1Δ (MDY3445XMDY3447) (M).
Figure 10.
mps3-dCC and mps3-dAR are defective in bouquet formation and display altered patterns of telomere clustering and distribution relative to the SPB.
(A) Bouquet assay strains marked with Spc42-dsRed to visualize the spindle pole body, and with Rap1-CFP to visualize telomeres, are stained with DAPI to visualize the DNA and imaged in 3-dimensional, high-resolution, deconvolved stacks. Results are from single experiments where 200 cells per time-point were scored visually by merging the 3 individual images for each nucleus to generate a 3-color image stack that is then rotated in software to put the spindle pole body at the periphery of the nucleus as viewed in a 2-dimensional projection [9]. Cells are scored as positive when they have a single telomere cluster, within ∼1/5 the apparent nuclear volume of the spindle pole, where the cluster is tight (top panel), loose (middle panel) or either (bottom panel). Strains used: wild-type (MDY2455XMDY2513), mps3-dAR (MDY2509XMDY2511), mps3-dCC (MDY2558XMDY2560), ndj1Δ (MCY1570XMCY1571). (B) Telomere distribution and proximity to the spindle pole are quantified by software in the rec8Δ background, where cells blocked in meiotic prophase are found to have a single telomere cluster [48], unless the cells also are mutant for the ability to cluster the telomeres as in ndj1Δ and mps3-dNT [9]. Each point represents the measurements from a single nucleus, marked and imaged as in (A), above. Dashed lines lie at 0.6 on the respective axes; the associated numbers are the fractions of each population between 0.0 and 0.6. The radius of each nucleus in microns is estimated by the distance in microns from the centroid of the spindle-pole body signal to the centroid of the DAPI signal. Telomere distribution (units of microns) is estimated from the 3-dimensional variance of the Rap1-CFP signal intensity for each nucleus (units of squared microns), normalized by nucleus radius. SPB-telomere proximity (a unitless ratio) is the distance in microns between the centroid of the spindle pole body signal and the centroid of the Rap1-CFP signal, normalized by nucleus radius. Thus, a tight cluster of telomeres adjacent to the spindle pole, as in tight bouquets, would generate a SPB-telomere distance of ∼0, while a tight cluster of all telomeres at the edge of the DAPI signal but opposite the spindle pole would generate a SPB-telomere distance of ∼2. Statistical analysis in the legend to Table 1. Strains used: rec8Δ (MDY2517XMDY2534), ndj1Δ rec8Δ (MCY1533X1535), mps3dCC rec8Δ (MDY2553XMDY2555), mps3-dAR rec8Δ (MDY2557XMDY2539).
Figure 11.
Model for RPM contribution to pairing.
Two pairs of homologs are diagrammed but only one chromatid is shown for each chromosome. Centromeres remain in close proximity following dissolution of the Rabl orientation (the first change diagrammed) due to a centromere-specific mechanism that joins pairs of nonhomologous centromeres but telomere anchoring to the nuclear envelope can hold chromosome arms apart. Short chromosomes depend more than long chromosomes on long range telomere movements that allow and generate collisions that in turn promote homology assessment, stabilization of association and synapsis.