The protease activity of yeast separase (esp1) is required for anaphase spindle elongation independently of its role in cleavage of cohesin

Abstract

Separase is a caspase-family protease required for the metaphase-anaphase transition in eukaryotes. In budding yeast, the separase ortholog, Esp1, has been shown to cleave a subunit of cohesin, Mcd1 (Scc1), thereby releasing sister chromatids from cohesion and allowing anaphase. However, whether Esp1 has other substrates required for anaphase has been controversial. Whereas it has been reported that cleavage of Mcd1 is sufficient to trigger anaphase in the absence of Esp1 activation, another study using a temperature-sensitive esp1 mutant concluded that depletion of Mcd1 was not sufficient for anaphase in the absence of Esp1 function. Here we revisit the issue and demonstrate that neither depletion of Mcd1 nor ectopic cleavage of Mcd1 by Tev1 protease is sufficient to support anaphase in an esp1 temperature-sensitive mutant. Furthermore, we demonstrate that the catalytic activity of the Esp1 protease is required for this Mcd1-independent anaphase function. These data suggest that another protein, possibly a spindle-associated protein, is cleaved by Esp1 to allow anaphase. Such a function is consistent with the previous observation that Esp1 localizes to the mitotic spindle during anaphase.

Figure 1.—

Mitotic spindles fail to elongate in esp1-C113 cells even after loss of cohesion. (A) Kinetics of budding in ESP1 and esp1-C113 cells released from α-factor arrest. (B) Average spindle lengths for the same time courses (n ≥ 120 for each point). Error bars represent one standard deviation. (C) Representative images of those used to determine spindle elongation kinetics in B. Top, wild-type cells; bottom, esp1-C113 mutant cells. Left, chromosome IV dots; right, spindles. Bar (top left), 2 μm.

Figure 2.—

Time-lapse analysis of spindle elongation following loss of sister chromatid cohesion in cells depleted of Mcd1. Selected frames of representative time-lapse series for spindle elongation in ESP1 (A) or esp1-C113 cells (B). (C) Plots depicting the overall kinetics of spindle elongation in ESP1 and esp1-C113 cells. Cells were selected by direct observation to verify loss of sister chromatid cohesion and the presence of a short spindle at the beginning of the recording. Numbers indicate time elapsed in minutes. Bar, 2 μm.

Figure 3.—

Mitotic spindles fail to elongate in esp1 cells depleted of Mcd1 by Tev-protease cleavage. (A) Kinetics of budding in ESP1 and esp1-C113 cells containing a Tev-protease cleavable Mcd1 released from α-factor arrest under conditions of Tev protease induction (galactose) or repression (dextrose). (B) Average spindle length for the same Tev protease-induced time courses (n ≥ 120 for each point) described above. Error bars represent one standard deviation. (C) Proteolysis of Tev protease-cleavable Mcd1-6xMyc. The 105-kDa species (solid arrow) corresponds to full-length Mcd1-Myc6. The 77-kDa species (thin arrow) corresponds to the C-terminal fragment of Mcd1-Myc6 (aa 181–621) after cleavage at amino acid 180 (aa180) by Tev protease. An uncharacterized 42-kDa fragment of Mcd1 (*) is also present in ESP1 lysates.

Figure 4.—

Proteolytic activity of ESP1 is required for anaphase elongation in the absence of sister chromatid cohesion. (A) Kinetics of budding in esp1-C113 mutant cells containing plasmids expressing either wild-type ESP1 or a protease active site mutant esp1-C1531S released from α-factor arrest. (B) Average spindle lengths for the same time courses (n ≥ 120 for each point). Error bars represent one standard deviation.

Figure 5.—

Analysis of temperature sensitivity of various esp1 mutant alleles. (A) Analysis of growth. Strains harboring the indicated mutant esp1 alleles were serially diluted and spotted on plates at the indicated temperatures. (B) Analysis of Esp1 protease function. Strains harboring the indicated esp1 alleles and containing 6xMyc-tagged Mcd1 were synchronized by α-factor block and released at 34°. Cells were harvested at the indicated times and extracts analyzed for Mcd1 levels by Western blot (left). Mcd1 levels were quantitated and normalized to α-tubulin levels (right). Note that α-factor was added 60 min after the first block to prevent cells that had completed mitosis from progressing past G1.