Structural biology of the separase-securin complex with crucial roles in chromosome segregation

Abstract

The cysteine protease separase opens the cohesin ring by cleaving its kleisin subunit and is a pivotal cell cycle factor for the transition from metaphase to anaphase. It is inhibited by forming a complex with the chaperone securin, and in vertebrates, also by the Cdk1-cyclin B1 complex. Separase is activated upon the destruction of securin or cyclin B1 by the proteasome, after ubiquitination by the anaphase-promoting complex/cyclosome (APC/C). Here we review recent structures of the active protease segment of Chaetomium thermophilum separase in complex with a substrate-mimic inhibitor and full-length Saccharomyces cerevisiae and Caenorhabditis elegans separase in complex with securin. These structures define the mechanism for substrate recognition and catalysis by separase, and show that securin has extensive contacts with separase, consistent with its chaperone function. They confirm that securin inhibits separase by binding as a pseudo substrate.

Figure 1

Structures of separase-securin complexes. (a). Domain organization of separase. The helical region of yeast separase is divided into four domains (I–IV) and given different colors, which is followed by SD (substrate-binding domain) and CD (catalytic domain). The domain boundaries for the N-terminal region of C. thermophilum and human separase are not known and therefore are not indicated. The unstructured segments (US) in the helical region are shown in gray. For Drosophila separase, only the subunit containing the SD-CD (known as SSE) is shown. Sc: S. cerevisiae (yeast), Ct: Chaetomium thermophilum, Ce: C. elegans, Dm: D. melanogaster, Hs: Homo sapiens. (b). Domain organization of securin. The separase interaction segment (SIS) is shown in magenta. The N-terminal KEN and D-boxes are indicated. (c). Schematic drawing of the structure of the yeast separase-securin complex. The domains of separase are colored as in Fig. 1a, and the securin SIS is in magenta. The catalytic Cys1531 is shown as a sphere model. Two of the phosphorylation sites in securin are indicated with spheres. The ends of the unstructured segment (US2) are indicated by the two gray spheres. (d). Structure of the yeast separase-securin complex, with separase shown as a molecular surface, viewed after 50° rotation around the vertical axis from panel c. The active site of separase is indicated with the red star. (e). Schematic drawing of the structure of the C. elegans separase-securin complex. The catalytic Cys1040 is shown as a sphere model. (f). Structure of the C. elegans separase-securin complex, with separase shown as a molecular surface. All structure figures were produced with PyMOL (www.pymol.org).

Figure 2

Structures of separase domains and their interaction with securin. (a). Schematic drawing of domains I and II of yeast separase. The segment of SIS that interacts with these domains are also shown (magenta). (b). Schematic drawing of domain III of yeast separase, colored from blue at the N-terminus to red at the C-terminus. (c). Schematic drawing of domain III of C. elegans separase. (d). Overlay of the structures of yeast IV-SD-CD (in color) and the SD-CD of C. thermophilum separase (gray). The catalytic Cys residue is indicated with the red asterisk. Blue arrowhead points to the additional β-strand provided by domain IV to the SD in yeast separase. Red arrowhead points to conformational differences in the loop L4 region. (e). Overlay of the structures of yeast IV-SD-CD (in color) and the IV-SD-CD of C. elegans separase (gray). The superposition is based on the CD only, and a 10° difference is observed for the forientation of the SD β sheet (green arrow).

Figure 3

Molecular mechanism for the inhibition of separase by securin. (a). Alignment of the cleavage sites in separase substrates. The two cleavage sites in each protein are named a and b. The equivalent residues in securin are also shown. The asterisks indicate securin mutants (mutations in green) that become substrates of separase. The P and P′ residues are labeled at the top, and the cleavage site is indicated with the vertical line. (b). Binding mode of residues 258-265 of yeast securin SIS (magenta) in the active site of yeast separase. Side chains of residues in the interface are shown as stick models and labeled. The bound position of a substrate-mimic inhibitor to C. thermophilum separase SD-CD is shown in gray. (c). Binding mode of residues 121-127 of C. elegans securin SIS (magenta) in the active site of C. elegans separase. (d). Closeup of the active site region of yeast separase. The catalytic Cys1531 side chain is hydrogen-bonded to the main-chain amide of Val1566 (dashed line in red) and 6 Å from the carbonyl carbon of Pro263 (dashed line in blue). (e). Close-up of the active site region of C. elegans separase.

Figure 4

Structural comparisons with caspase and gingipain. (a). Schematic drawing of SD-CD of yeast separase (green and cyan, respectively), together with the portion of securin SIS (magenta) that is in the CD active site (indicated with the red star). The four-helical bundle in SD is omitted for clarity. (b). Schematic drawing of caspase 7 in a covalent complex with a substrate-mimic inhibitor (magenta). One molecule is colored in light cyan and cyan for its two fragments, while the other in light green and green. The blue arrowhead indicates the loop preceding strand β6 that is in the active site. The orientation of the first molecule is the same as that of separase CD. (c). Schematic drawing of gingipain in a covalent complex with a substrate-mimic inhibitor (magenta). The A and B domains are colored in green and cyan, respectively.