Chromosomal instability by inefficient Mps1 auto-activation due to a weakened mitotic checkpoint and lagging chromosomes

Abstract

Background: Chromosomal instability (CIN), a feature widely shared by cells from solid tumors, is caused by occasional chromosome missegregations during cell division. Two of the causes of CIN are weakened mitotic checkpoint signaling and persistent merotelic attachments that result in lagging chromosomes during anaphase.

Principal findings: Here we identify an autophosphorylation event on Mps1 that is required to prevent these two causes of CIN. Mps1 is phosphorylated in mitotic cells on at least 7 residues, 4 of which by autophosphorylation. One of these, T676, resides in the activation loop of the kinase domain and a mutant that cannot be phosphorylated on T676 is less active than wild-type Mps1 but is not kinase-dead. Strikingly, cells in which endogenous Mps1 was replaced with this mutant are viable but missegregate chromosomes frequently. Anaphase is initiated in the presence of misaligned and lagging chromosomes, indicative of a weakened checkpoint and persistent merotelic attachments, respectively.

Conclusions/significance: We propose that full activity of Mps1 is essential for maintaining chromosomal stability by allowing resolution of merotelic attachments and to ensure that single kinetochores achieve the strength of checkpoint signaling sufficient to prevent premature anaphase onset and chromosomal instability. To our knowledge, phosphorylation of T676 on Mps1 is the first post-translational modification in human cells of which the absence causes checkpoint weakening and CIN without affecting cell viability.

Figure 1. Mps1 is phoshorylated during mitosis on at least 7 residues.

(A, B) A monoclononal Hela cell line stably expressing LAP-tagged Mps1 (HeLa-LAP-Mps1) was analyzed for expression level of LAP-Mps1 relative to endogenous Mps1 as shown by immunoblot (anti-Mps1) (A) and localization of LAP-Mps1 (anti-GFP) together with centromeres (ACA) and DNA (DAPI) (B). Scale bar is 5 µm. (C) List of phosphorylation sites identified by mass spectrometry on LAP-Mps1 isolated from mitotic HeLa-LAP-Mps1 cells (left column, in vivo) or on recombinant Mps1 isolated from insect cells (middle column, in vitro). Phosphorylation sites on recombinant Mps1 were designated ‘auto-phosphorylations’ when sites were not found on similarly expressed and isolated recombinant kinase-dead Mps1. (D) Schematic overview of phosphorylated residues on Mps1.

Figure 2. T676 phosphorylation is required for correct chromosome segregation and full checkpoint signaling.

(A) Expression levels, as shown by immunoblot (anti-Mps1), in asynchronously growing cells of LAP-Mps1 mutants (asterisk (*) indicates mutations that confer resistance to Mps1 shRNA) transiently transfected into U2OS cells. (B) Flow cytometric analysis of the fraction of mitotic cells transfected with Mock or Mps1 shRNA together with the indicated plasmids (asterisk (*) indicates mutations that confer resistance to Mps1 shRNA), and treated with nocodazole for 18 hrs. (C) Quantification of (B), average of three experiments (+/−SD). (D) U2OS cells transfected as in 2B were treated for 30 minutes with MG132. Fixed cells were stained with DAPI and the percentage of cells with all chromosomes aligned to the metaphase plate was determined (n = 30 cells). Graph represents the average of three experiments (+/−SD). (E) Mitotic progression of U2OS cells transfected as in 2B along with H2B-EYFP Mps1 was followed by live cell imaging. Stills from two representative cells showing anaphase with misaligned/lagging chromosomes (arrows) are shown (see Movie S1 and Movie S2). Scale bar is 5 µm. (F) Quantification of (E), average of three experiments (+/−SD), n = total amount of cells. (G) Graph showing the fraction of total missegregations as displayed in 2F with misaligned chromosomes, lagging chromosomes, or both in cells expressing Mps1-T676A.

Figure 3. Lack of T676 auto-phosphorylation reduces Mps1 kinase activity.

(A) Kinase activity of LAP-Mps1 mutants (asterisk (*) indicates mutations that confer resistance to Mps1 shRNA) extracted from mitotic U2OS cells was examined by in vitro kinase activity towards recombinant kinase-dead His-Mps1 (D664A). (B) Quantification of (A). Relative values to LAP-Mps1-WT were calculated after subtraction of the background signal derived from ‘mock’ immunoprecipitations (first lane in Fig. 3A). Average of three experiments (+/−SD). (C) Western blot showing phosphorylation specifically on T676 in lysates form U2OS cells transfected with the indicated constructs (asterisk (*) indicates mutations that confer resistance to Mps1 shRNA). (D) Western blot showing in vitro intermolecular auto-phosphorylation on T676. Combinations (as indicated) of recombinant wild-type GST-Mps1(-WT), kinase-dead GST- and His-Mps1(-KD) were subjected to a in vitro kinase reactions, after which proteins were analyzed for reactivity to pT676 antibody (green) and Mps1 antibody (red) by immunoblot.

Figure 4. T676 phosphorylation is required for maintaining chromosomal stability but not cell viability.

(A) U2OS cells were transfected as in 2B along with pBabe-puro, grown under puromycin selection for two weeks after which surviving colonies were stained. Immunoblot (anti-Mps1) shows expression levels in lysates taken 3 days after transfection of the different LAP-Mps1 mutants. (B) Quantification of the number of colonies counted in 4A, average of three experiments (+/−SD). (C) Immunoblot (anti-Mps1) showing expression levels of endogenous Mps1 and LAP-Mps1 in lysates of U2OS cells, and UTRM10 cells continuously expressing Mps1 shRNA through doxycycline treatment that stably express wild-type LAP-Mps1 (UTRM10-WT) or LAP-Mps1-T676A (UTRM10-T676A). (D) Graph shows average percentages of UTRM10-T676A and UTRM10-WT cells that missegregate chromosomes as observed with live cell imaging (as in 2E), (+/−SD), n = total amount of cells.