Cyclin A regulates kinetochore microtubules to promote faithful chromosome segregation

Abstract

The most conspicuous event in the cell cycle is the alignment of chromosomes in metaphase. Chromosome alignment fosters faithful segregation through the formation of bi-oriented attachments of kinetochores to spindle microtubules. Notably, numerous kinetochore-microtubule (k-MT) attachment errors are present in early mitosis (prometaphase), and the persistence of those errors is the leading cause of chromosome mis-segregation in aneuploid human tumour cells that continually mis-segregate whole chromosomes and display chromosomal instability. How robust error correction is achieved in prometaphase to ensure error-free mitosis remains unknown. Here we show that k-MT attachments in prometaphase cells are considerably less stable than in metaphase cells. The switch to more stable k-MT attachments in metaphase requires the proteasome-dependent destruction of cyclin A in prometaphase. Persistent cyclin A expression prevents k-MT stabilization even in cells with aligned chromosomes. By contrast, k-MTs are prematurely stabilized in cyclin-A-deficient cells. Consequently, cells lacking cyclin A display higher rates of chromosome mis-segregation. Thus, the stability of k-MT attachments increases decisively in a coordinated fashion among all chromosomes as cells transit from prometaphase to metaphase. Cyclin A creates a cellular environment that promotes microtubule detachment from kinetochores in prometaphase to ensure efficient error correction and faithful chromosome segregation.

Figure 1. The stability of k-MT attachments in prometaphase and metaphase

a, Box and whisker plot of k-MT half-lives of RPE-1, U2OS, and PtK1 cells in prometaphase and metaphase calculated from the fluorescence intensity decay curves (r² > 0.99); n = 40 cells for RPE-1 and U2OS, and 20 cells for PTK1 per condition. Black circle represents the cell from panel f. b, Normalized fluorescence intensity of prometaphase (filled circles) and metaphase (white circles) spindles. c, DIC and background subtracted fluorescence images (pseudo-colored heatmaps) of U2OS cells in prometaphase and metaphase. Asterisks mark spindle poles. Scale bar, 5 μm. d, DIC and fluorescence images of an RPE-1 cell in prometaphase and metaphase. Scale bar, 5 μm. e, k-MT half-life of individual RPE-1 cells photoactivated serially in prometaphase (left) or in prometaphase and then in metaphase (right). f, DIC and fluorescence images of a PtK1 in prometaphase. Arrow indicates unaligned chromosome. Scale bar, 5 μm.

Figure 2. K-MT stability relies on cyclin A

a, k-MT half-life of RPE-1 cells treated with 20μM MG132 or 5μM Epoxomicin in prometaphase and metaphase; n = 10 cells per condition. b, DIC and fluorescence images of metaphase spindles in untreated (control) and cyclin A overexpressing (CycA ΔD OX) U2OS cells with cyclin A ΔD visualized by mCherry fluorescence. Scale bar, 5 μm. c, k-MT half-life of untreated (control), cyclin A overexpressing (CycA ΔD OX), and cyclin A depleted (CycA KD) U2OS cells; n = 13 cells for control, n = 10 cells for CycA ΔD OX and CycA KD per condition. d, top: k-MT half-life of RPE-1 cells untreated (control) or released from 12 hour nocodazole treatment (Nocodazole Washout); n = 10 cells per condition. Bottom: cyclin A and actin immunoblot of untreated (control) or Nocodazole arrested (Noc) cells. Graphs show mean ± s.e.m. *P ≤ 0.01, two-tailed t-test.

Figure 3. Cyclin A deficiency increases chromosome mis-segregation

a, anaphase spindles of untreated (control) or cyclin A depleted (CycA KD) U2OS cells. White arrow highlights merotelic kinetochore. Scale bar, 5 μm. b, percent of anaphase cells with lagging chromosomes; n = 300 cells per condition from three independent experiments. c, fluorescence intensities of U2OS cells stained for cyclin A; n = 100 cells per condition from three independent experiments. d, Box and whisker plot of k-MT half-lives of U2OS cells incubated in monastrol for 1 hour (1hr) and 6 hours (6hr); n = 10 cells per condition. e, percent of anaphase cells with lagging chromosomes that were untreated (control), or after recovery from monastrol incubation for 1 hour (1hr) or 6 hours (6hr); n = 100 cells for 1hr and 123 cells for 6 hr. Graphs show mean ± s.e.m. *P ≤ 0.01, two-tailed t-test.

Figure 4. Cyclin A promotes faithful chromosome segregation

a, Immunofluorescence of endogenous cyclin A, tubulin, centromeres (ACA) and DNA of U2OS cells in prometaphase, metaphase, and anaphase. Scale bar, 5 μm. b, fluorescence intensities of U2OS cells stained for cyclin A; n = 150 cells per condition from three independent experiments. c, k-MT half-life of untreated (control) and cyclin A WT overexpressing (CycA WT OX) U2OS cells; n = 10 cells per condition. d, immunofluorescence of endogenous cyclin A, tubulin, centromeres (ACA) and DNA of control or Cyclin A WT overexpressing (CycA WT OX) U2OS cells. Scale bar, 5 μm. e, fluorescence intensities of cyclin A of control U2OS cells or cells expressing cyclin A WT. f, number of anaphases in U2OS cells with lagging chromosomes; n=150 for control cells and 37 for cyclin A WT cells. g, Cells enter prometaphase with high cyclin A and cyclin B. The proteosome-dependent reduction of cyclin A levels below a critical threshold induces a coordinated increase in k-MT attachment stability at the prometaphase to metaphase transition. Graphs show mean ± s.e.m. *P ≤ 0.01, two-tailed t-test.