Spindle checkpoint protein Xmad1 recruits Xmad2 to unattached kinetochores

Abstract

The spindle checkpoint prevents the metaphase to anaphase transition in cells containing defects in the mitotic spindle or in chromosome attachment to the spindle. When the checkpoint protein Xmad2 is depleted from Xenopus egg extracts, adding Xmad2 to its endogenous concentration fails to restore the checkpoint, suggesting that other checkpoint component(s) were depleted from the extract through their association with Xmad2. Mass spectrometry provided peptide sequences from an 85-kD protein that coimmunoprecipitates with Xmad2 from egg extracts. This information was used to clone XMAD1, which encodes a homologue of the budding yeast (Saccharomyces cerevisiae) checkpoint protein Mad1. Xmad1 is essential for establishing and maintaining the spindle checkpoint in egg extracts. Like Xmad2, Xmad1 localizes to the nuclear envelope and the nucleus during interphase, and to those kinetochores that are not bound to spindle microtubules during mitosis. Adding an anti-Xmad1 antibody to egg extracts inactivates the checkpoint and prevents Xmad2 from localizing to unbound kinetochores. In the presence of excess Xmad2, neither chromosomes nor Xmad1 are required to activate the spindle checkpoint, suggesting that the physiological role of Xmad1 is to recruit Xmad2 to kinetochores that have not bound microtubules.

Figure 5

p85 is Xmad1. (A) Alignment of amino acid sequences of Xenopus p85 (X. laevis), human (H. sapiens), budding yeast (S. cerevisiae), and fission yeast (S. pombe) Mad1. The peptide sequences obtained by mass spectrometry analysis of p85 are indicated as dots on top of the Xenopus sequence. The common residues shared by two or more sequences are shaded. The alignment was performed with DNASTAR “MegAlign” program. (B) Predicted coiled-coil regions in p85 (X.l.), human (H.s.), budding yeast (S.c.), and fission yeast (S.p.) Mad1. The scale of the amino acid sequence is indicated on top. The prediction was performed with DNASTAR “PROTEAN” program.

Figure 5

p85 is Xmad1. (A) Alignment of amino acid sequences of Xenopus p85 (X. laevis), human (H. sapiens), budding yeast (S. cerevisiae), and fission yeast (S. pombe) Mad1. The peptide sequences obtained by mass spectrometry analysis of p85 are indicated as dots on top of the Xenopus sequence. The common residues shared by two or more sequences are shaded. The alignment was performed with DNASTAR “MegAlign” program. (B) Predicted coiled-coil regions in p85 (X.l.), human (H.s.), budding yeast (S.c.), and fission yeast (S.p.) Mad1. The scale of the amino acid sequence is indicated on top. The prediction was performed with DNASTAR “PROTEAN” program.

Figure 3

Excess Xmad2 inhibits sister chromatid segregation. Metaphase spindle and chromosomes were assembled in egg extracts treated with buffer (top, Control) or Xmad2 protein (100 ng/μl extract; bottom, Xmad2) as described in Materials and Methods. To initiate anaphase, calcium chloride was added and samples were taken immediately before (0′) or at the indicated times after calcium addition to visualize the nuclear (DNA) and spindle morphology (Spindle). The reduced microtubule intensity in the control experiment at 40 and 60 min reflects the reduced microtubule density typical of interphase extracts.

Figure 9

Anti-Xmad1 antibodies interfere with Xmad1 and Xmad2 binding to kinetochores. Metaphase chromosomes were assembled in egg extracts and treated with (B–D) or without (A) nocodazole. C and D, anti-Xmad1 and anti-Xmad2 antibodies, respectively, were added to the extracts 1 h before the addition of nocodazole. The chromosomes were isolated through a sucrose cushion and stained for Xmad1 and Xmad2. For D, Xmad2 staining is essentially the same if the anti-Xmad2 antibody was omitted during the immunofluorescent staining procedures (data not shown). Aggregates of chromosomes from several sperm nuclei are shown.

Figure 8

Xmad1 localized to unattached kinetochores during mitosis. Asynchronously growing XTC cells (A–E) or cells treated with nocodazole (F) were fixed in 3% paraformaldehyde and stained with affinity-purified mouse anti-Xmad1 and rabbit anti-Xmad2 antibodies as indicated on top. Texas red– conjugated anti–mouse and fluorescein-conjugated anti– rabbit IgG antibodies were used as secondary antibodies. The nuclei were viewed by staining with the DNA-binding dye Hoechst 33258 (DNA). The merges of all three fluorochromes are also shown (All). The cell cycle stages were determined by the morphology of the cell and its chromosomes. A, interphase; B, prophase; C, prometaphase; D, metaphase; E, anaphase; F, a nocodazole-arrested cell. Representative pictures of each cell type are shown.

Figure 1

Excess Xmad2 protein inhibits exit from mitosis in CSF-arrested extracts. (A) Immunoblot of Xmad2 in egg extracts (lane 1), in mock-depleted extracts (lane 2), and in Xmad2-depleted extracts (lane 3). Extracts shown in lanes 2 and 3 were used in the experiment shown in B. The migration of molecular size standards and Xmad2 is indicated on the left and right, respectively. (B) Autoradiogram of histone H1 kinase assay. Mock- or Xmad2-depleted CSF-arrested egg extracts were treated as indicated on the left: low 6H-Xmad2, recombinant 6H-Xmad2 protein supplemented to the level of endogenous protein; high 6H-Xmad2, 6H-Xmad2 added to 17-fold the concentration of endogenous protein. After 1 h of incubation on ice, sperm nuclei (10,000/μl) were added to all samples and incubated for 10 min at 23°C, followed by incubation with nocodazole (lower three panels) for another 10 min. The metaphase arrest was then released by the addition of calcium chloride to inactivate CSF activity. Aliquots were taken immediately before calcium was added (time 0) and every 15 min thereafter to determine Cdc2-associated histone H1 kinase activity. (C) Dose-dependent inhibition of Cdc2 inactivation by Xmad2. Xmad2 protein at the indicated concentrations (□, 0; ⋄, 25; ○, 50; ▵, 100 ng/μl of extract, respectively) was preincubated with CSF-arrested extracts, followed by incubation with sperm nuclei (500/μl) in the absence of nocodazole. H1 kinase assays were performed as described in B and incorporation of ³²P into H1 was quantified using a PhosphorImager.

Figure 2

(A) Excess Xmad2 sustains Cdc2 activity by maintaining cyclin B levels in the absence of any nuclei and nocodazole. The addition of Xmad2 (100 ng/μl of extract), sperm nuclei (500/μl of extract), and/or nocodazole to CSF-arrested extracts is indicated on the left. Samples were taken for H1 kinase measurement and for cyclin B Western blots at the indicated times after calcium addition. (B) Excess Xmad2 arrests cycling extracts at the first mitosis. Eggs were activated in vitro to start the embryonic cell cycle and extracts were prepared. This cycling extract entered mitosis at 30 and 90 min upon incubation at 23°C (top) and became arrested at first mitosis in the presence of excess 6H-Xmad2 (bottom). Autoradiograms of H1 kinase assays are shown.

Figure 4

(A) Xmad2 associates with an 85-kD protein in CSF-arrested extracts. Immunoprecipitation was performed by using a control antibody (lanes 1 and 3) or an anti-Xmad2 antibody (lanes 2 and 4) as indicated. The immunoprecipitates were washed with a buffer containing 1% Triton X-100 (lanes 1 and 2) or 0.1% SDS (lanes 3 and 4). The proteins were resolved by SDS-PAGE and stained with Coomassie blue. The migration of p85, IgG, and Xmad2 is indicated on the left, and that of molecular weight standards on the right. (B) Sequencing of Xmad1. The unseparated pool of tryptic peptides recovered from the gel matrix was analyzed using a novel quadrupole TOF mass spectrometer. (a) Mass spectrum of the unseparated tryptic digest of the Xmad1 band. The peptide ions designated with T were in turn isolated by a quadrupole mass analyzer, fragmented in the collision cell, and then their tandem mass spectra were acquired with a reflector time-of-flight module. Peaks designated with asterisks are trypsin autolysis products. The peptide ion designated with A belongs to the antibody used in the purification of Xmad1. (b) Tandem mass spectrum of the doubly charged peptide ion T₈ (marked with an arrow in a). Upon collisional fragmentation, tryptic peptides tend to produce a continuous series of fragment ions containing the COOH terminus (Y″ ions; Roepstorff and Fohlman, 1984). The peptide sequence (shown above) is deduced by considering precise mass differences between adjacent Y″ ions. M*, N-acetylated methionine sulfoxide amino acid residue. The spectrum also contains minor series of the NH₂-terminal fragment ions (B ions) as well as internal fragment ions. Therefore to deduce the sequence with high confidence it is necessary to distinguish Y″ ions from the other ions in the spectrum. This has been achieved by selective isotopic labeling of the COOH-terminal carboxyl group of the peptides. (c) A zoom of the region in the tandem mass spectrum of T₈ shows the principal of sequence readout of isotopically labeled peptides. Upon tryptic cleavage of the protein in a buffer containing H₂ ¹⁶O/H₂ ¹⁸O 1:1 (vol/vol) COOH-terminal carboxyl groups of peptide molecules incorporate ¹⁶O and ¹⁸O atoms in a 1:1 ratio. Thus the fragment ions containing the COOH terminus of the peptide (mostly Y″ ions) appear in the fragment spectrum as a characteristic isotopic pattern—a doublet split by 2 D. B ions or other fragment ions not containing COOH-terminal carboxyl group are observed as ions having normal isotopic pattern.

Figure 4

(A) Xmad2 associates with an 85-kD protein in CSF-arrested extracts. Immunoprecipitation was performed by using a control antibody (lanes 1 and 3) or an anti-Xmad2 antibody (lanes 2 and 4) as indicated. The immunoprecipitates were washed with a buffer containing 1% Triton X-100 (lanes 1 and 2) or 0.1% SDS (lanes 3 and 4). The proteins were resolved by SDS-PAGE and stained with Coomassie blue. The migration of p85, IgG, and Xmad2 is indicated on the left, and that of molecular weight standards on the right. (B) Sequencing of Xmad1. The unseparated pool of tryptic peptides recovered from the gel matrix was analyzed using a novel quadrupole TOF mass spectrometer. (a) Mass spectrum of the unseparated tryptic digest of the Xmad1 band. The peptide ions designated with T were in turn isolated by a quadrupole mass analyzer, fragmented in the collision cell, and then their tandem mass spectra were acquired with a reflector time-of-flight module. Peaks designated with asterisks are trypsin autolysis products. The peptide ion designated with A belongs to the antibody used in the purification of Xmad1. (b) Tandem mass spectrum of the doubly charged peptide ion T₈ (marked with an arrow in a). Upon collisional fragmentation, tryptic peptides tend to produce a continuous series of fragment ions containing the COOH terminus (Y″ ions; Roepstorff and Fohlman, 1984). The peptide sequence (shown above) is deduced by considering precise mass differences between adjacent Y″ ions. M*, N-acetylated methionine sulfoxide amino acid residue. The spectrum also contains minor series of the NH₂-terminal fragment ions (B ions) as well as internal fragment ions. Therefore to deduce the sequence with high confidence it is necessary to distinguish Y″ ions from the other ions in the spectrum. This has been achieved by selective isotopic labeling of the COOH-terminal carboxyl group of the peptides. (c) A zoom of the region in the tandem mass spectrum of T₈ shows the principal of sequence readout of isotopically labeled peptides. Upon tryptic cleavage of the protein in a buffer containing H₂ ¹⁶O/H₂ ¹⁸O 1:1 (vol/vol) COOH-terminal carboxyl groups of peptide molecules incorporate ¹⁶O and ¹⁸O atoms in a 1:1 ratio. Thus the fragment ions containing the COOH terminus of the peptide (mostly Y″ ions) appear in the fragment spectrum as a characteristic isotopic pattern—a doublet split by 2 D. B ions or other fragment ions not containing COOH-terminal carboxyl group are observed as ions having normal isotopic pattern.

Figure 6

Xmad1 abundance is constant throughout the cell cycle. (A) Specificity of anti-Xmad1 antibodies in immunoblots of frog egg extracts (lanes E), frog cultured cells XTC (lanes X), and HeLa cells (lanes H). An affinity-purified anti-Xmad1 antibody (lanes 4–9) or the same antibody preblocked with recombinant Xmad1 protein (lanes 1–3) were used. Lanes 1–6, egg extracts or cell lysates. Lanes 7–12, anti-Xmad2 immunoprecipitates prepared from egg extracts or cell lysates were immunoblotted with anti-Xmad1 (lanes 7–9) or anti-Xmad2 (lanes 10–12) antibodies. (B) CSF-arrested extract (time 0) was incubated with calcium for the time indicated on top, and immunoblotted for Xmad1 (middle) or Xmad2 (bottom). Top, histone H1 kinase assay. (C) Immunoblots of cycling extracts using antibodies specific to Xmad1 (middle) or Xmad2 (bottom). Mitosis occurs at ∼50 and 120 min as determined by histone H1 kinase assay (top) and by the nuclear morphology (data not shown).

Figure 7

Xmad1 is important for establishing and maintaining the spindle checkpoint in frog egg extracts. (A) Xmad1 is important for spindle checkpoint function. CSF-arrested extracts were preincubated with a control antibody (IgG), an anti-Xmad1 antibody, or an anti-Xmad1 antibody preblocked with recombinant Xmad1 protein as indicated. Nocodazole and sperm nuclei were added to activate the spindle checkpoint for the lower three panels, whereas nocodazole was omitted for the top panel. H1 kinase activities were determined immediately before and at the indicated times after calcium addition as described in Fig. 1 B. (B) Xmad1 is required for maintaining the spindle checkpoint. The spindle checkpoint was first activated in CSF-arrested extracts with nocodazole and sperm nuclei followed by incubation with control (top), anti-Xmad1 (middle), or anti-Xmad2 (bottom) antibodies. H1 kinase activities were determined as in A.

Figure 10

Mitotic arrest induced by excess Xmad2 is independent of Xmad1. (A) Immunoblots of Xmad1 (top) and Xmad2 (bottom) in CSF-arrested extracts (lane 1), mock- (lane 2), Xmad1- (lane 3), or Xmad2-depleted (lane 4) extracts. (B) Excess Xmad2 can still induce a mitotic arrest in the absence of Xmad1. The extracts shown in lanes 2–4 of A were used. Top panels, extracts were incubated with sperm nuclei alone. Middle panels, nocodazole and sperm nuclei were added to activate the spindle checkpoint. Bottom panels, excess 6H-Xmad2 was added to the extracts in the absence of sperm nuclei and nocodazole. H1 kinase activities were determined before and at various times after calcium addition as described for Fig. 1 B.

Figure 11

A model for how Xmad1 and Xmad2 work to activate the spindle checkpoint. Binding of Xmad1 to unattached kinetochores enables its associated Xmad2 to interact with a downstream checkpoint component X. This interaction converts X into a form that is capable of directly or indirectly inhibiting the anaphase-promoting complex (APC). The interaction between Xmad2 and X is unstable when Xmad2 is not associated with kinetochores, so that the checkpoint is not activated without unattached kinetochores. Increasing the global Xmad2 concentration drives the complex formation by mass action even in the absence of kinetochores. At a high enough concentration of Xmad2, the level of Xmad2–X complex becomes comparable to that induced through kinetochore-associated Xmad2 and results in constitutive activation of the spindle checkpoint. A likely candidate for X is Cdc20.