Laser microsurgery reveals conserved viscoelastic behavior of the kinetochore

Abstract

Accurate chromosome segregation depends on proper kinetochore-microtubule attachment. Upon microtubule interaction, kinetochores are subjected to forces generated by the microtubules. In this work, we used laser ablation to sever microtubules attached to a merotelic kinetochore, which is laterally stretched by opposing pulling forces exerted by microtubules, and inferred the mechanical response of the kinetochore from its length change. In both mammalian PtK1 cells and in the fission yeast Schizosaccharomyces pombe, kinetochores shortened after microtubule severing. Interestingly, the inner kinetochore-centromere relaxed faster than the outer kinetochore. Whereas in fission yeast all kinetochores relaxed to a similar length, in PtK1 cells the more stretched kinetochores remained more stretched. Simple models suggest that these differences arise because the mechanical structure of the mammalian kinetochore is more complex. Our study establishes merotelic kinetochores as an experimental model for studying the mechanical response of the kinetochore in live cells and reveals a viscoelastic behavior of the kinetochore that is conserved in yeast and mammalian cells.

Figure 1.

Merotelic KTs shorten after MT severing. (A) Diagrammatic representation of the in vivo assay in which MTs attached to a merotelic KT are severed by laser. Laser ablation was used to sever one (single ablation) or both (double ablation) MT bundles attached to a merotelic KT in anaphase cells, releasing the forces acting on the KT. Change of the KT length after laser severing reflects its mechanical response. (B) Maximum intensity projection of time-lapse images of a merotelic KT (Hec1-GFP, green) in a PtK1 cell and the severing of one of the two MT bundles (α-tubulin–X-Rhodamine, magenta). Time 0 indicates the frame before ablation. The white lightning sign indicates the ablation site. Images were acquired with a time resolution Δt = 2.5 s. Close-up views show lagging KTs (2×). (C) Time projection of the area containing the merotelic KT from the cell in B. The boxed region indicates the time interval shown in B. (D) Relaxation kinetics of the merotelic KT from B and C. The red dashed line indicates the severing time. (E) Maximum intensity projections of time-lapse images of a merotelic KT (Ndc80-GFP, green) in a S. pombe cell and the severing (lightning sign) of the spindle (α-tubulin–mCherry, magenta) close to one spindle pole body. The time 0 is the frame before ablation. The spindle breakage occurred immediately after ablation. Images were acquired with a time resolution Δt = 2.8 s. Close-up views show lagging KTs (1.35×). (F) Time projection of an area containing the merotelic KT from the cell in E. The boxed region indicates the time interval shown in E. (G) Relaxation kinetics of the merotelic KT from E and F. The red dashed line indicates the severing time.

Figure 2.

Shortening of the inner KT/centromere after MT severing. (A) Maximum intensity projections of a merotelic KT in PtK1 cells expressing GFP-CenpA (green) and injected with α-tubulin–X-rhodamine (magenta). Time 0 indicates the frame before ablation. The white lightning sign indicates the ablation site. Images were acquired with a time resolution Δt = 2.4 s. Close-up views show lagging KTs (1.85×). (B) Time projection of the area containing the merotelic KT from the cell in A. The boxed region indicates the time interval shown in A. (C) Relaxation kinetics of the merotelic KT from A and B. The red dashed line indicates the severing time. (D) Maximum intensity projections of a merotelic KT in fission yeast cell expressing Cnp1-GFP, shown in green, and α-tubulin–mCherry, shown in magenta. The white lightning sign indicates the ablation site. The time 0 is the frame before ablation. The spindle breakage occurred immediately after ablation. Images were acquired with a time resolution Δt = 3.2 s. Close-up views show lagging KTs (1.5× magnification). (E) Time projection of an area containing the merotelic KT from the cell in D. The boxed region indicates the time interval shown in D. (F) Relaxation kinetics of the merotelic KT from D and E. The red dashed line indicates the severing time.

Figure 3.

Outer KT relaxation upon double ablation and 9 µM NOC treatment in PtK1 and fission yeast cells. (A) Maximum intensity projections of a merotelic KT in PtK1 cell expressing Hec1-GFP, shown in green, and injected with α-tubulin–X-rhodamine, shown in magenta. Time 0 indicates the frame before ablation. The white lightning signs indicate ablation sites. Images were acquired with a time resolution Δt = 2.2 s. Close-up views show lagging KTs (2.5×). (B) Time projection of the area containing the merotelic KT from the cell in A. The boxed region indicates the time interval shown in A. (D) Maximum intensity projections of a merotelic KT in fission yeast cell expressed Ndc80-GFP, shown in green, and α-tubulin–mCherry, shown in magenta. The white lightning signs indicate ablation sites. The time 0 is the frame before ablation. The spindle breakage occurred immediately after ablation. Images were acquired with a time resolution Δt = 3 s. Close-up views show lagging KTs (1.65×). (E) Time projection of an area containing the merotelic KT from the cell in D. The boxed region indicates the time interval shown in D. (C and F) Normalized KT length as a function of time, for the inner (magenta) and outer (green) KT in PtK1 (C) and S. pombe cells (F) after single ablation. Normalized outer KT length as a function of time after double ablation is shown in light blue. Data for each KT were normalized setting the value before ablation to 1 and the minimum value reached by the individual KT to 0, binned, and averaged. Error bars represent standard deviation, n = 25/18 for Hec1/Ndc80 single ablation, n = 15/24 for CenpA/Cnp1 single ablation, and n = 9/10 for Hec1/Ndc80 double ablation. All the data were fitted with a single exponential equation, L(t) = A × exp(−t × ln(2)/t1/2). Half-lives are shown in the insets. (G) Maximum intensity projections of a merotelic KT in PtK1 cell expressing Hec1-GFP, shown in green, and injected with α-tubulin–X-rhodamine, shown in magenta. Time 0 indicates the frame before addition of 9 µM NOC. Images were acquired with a time resolution Δt = 20s. Close-up views show lagging KTs (1.5×). (H) Time projection of an area containing the merotelic KT from the cell in panel G. The boxed region indicates the time interval shown in G. (I) Normalized KT length (outer KT) as a function of time. PtK1 cells were treated with 9 µM NOC. Comparison between ablation experiments and 9 µM NOC treatment for the first 200 s is shown in I (inset). The length change was initially slow (I, inset), most likely because of the slow kinetics of NOC-dependent MT depolymerization as compared with a sudden MT severing.

Figure 4.

KT relaxed length depends on the stretched length in PtK1 cells, but not in S. pombe. (A–D) Comparison of the relaxed KT length after laser ablation, NOC treatment, and control unstretched KTs. As control KTs, we measured PtK1 anaphase KTs that were attached to a K-fiber emanating from only one spindle pole and S. pombe KTs that were not lagging during anaphase. Length of control KTs was determined by using automated software to detect the KT and to count pixels with KT-specific signal. These measurements are influenced by the imaging conditions and they may not represent the absolute length of KTs, but they can be used to compare merotelic and control KTs. Relaxed length after severing one K-fiber for the outer KT, Hec1/Ndc80, is shown in green and for inner KT, CenpA/Cnp1, in magenta. The relaxed length of the outer KT after severing both K-fibers is shown in blue. The relaxed length of KTs in PtK1 cells after 9-µM NOC treatment is shown in white. The lengths are calculated as mean of the values measured in the last 30 s and 15 s of the experiment for PtK1 cells and fission yeast, respectively. The tops and bottoms of each box are the 25% and 75% of the sample, respectively. The line in the middle of each box is the sample median. Hec1-single ablation, n = 25; Hec1-double ablation, n = 9; Hec1-NOC treatment, n = 9; Hec1-control KTs, n = 43; CenpA-ablation, n = 15; CenpA-control KTs, n = 67; Ndc80-single ablation, n = 18; Ndc80-double ablation, n = 10; Ndc80-control KTs, n = 11; Cnp1-ablation, n = 24; and Cnp1-control KTs, n = 11. There was a statistically significant difference between the relaxed and control KT length in all cases in A–C (**, P < 0.001; ***, P < 0.0001), but no significant difference (NS, P > 0.05) in D. NOC was added at 9-µM concentration for 15–25 min. (E and F) Correlation between initial stretched KT length and relaxed length. The initial stretched length was calculated as a mean of the values measured before ablation. Solid colored lines represent linear fit for each specific condition.

Figure 5.

Two simple models of a KT as a viscoelastic object consisting of a spring and a dashpot. (A and B) Schematic representation of the models. A spring of a stiffness κ and a dashpot of a viscous drag coefficient γ are connected in series (A) or in parallel (B). The KT (pink ellipse) is under tension F and has a length x. In A, the length is the sum of the length of the spring, x_s, and the length of the dashpot, x_d. The equations of the models are given under the schemes. (C and D) KT length as a function of time for the models in A and B, respectively, for a constant tension F that starts to act at time t = 0 and stops at t = T₁. The length of a control KT is denoted x_c. (E and F) Relaxed length as a function of the stretched length (dark green) for the models in A and B, respectively. In E, γ = 5 × 10⁵ pNs/µm, κ = 1,000 pN/µm, and x_c = 0.7 µm, whereas F and t were varied simultaneously in the intervals F = 400–1,000 pN and T1 = 7–12 min, as shown in the inset. In F, κ = 100 pN/µm, x_c = 0.4 µm, and F was varied in the interval 50–200 pN. Horizontal and oblique dashed lines correspond to complete relaxation to the control length x_c and no relaxation, respectively.