Characterization of novel MPS1 inhibitors with preclinical anticancer activity.

Monopolar spindle 1 (MPS1), a mitotic kinase that is overexpressed in several human cancers, contributes to the alignment of chromosomes to the metaphase plate as well as to the execution of the spindle assembly checkpoint (SAC). Here, we report the identification and functional characterization of three novel inhibitors of MPS1 of two independent structural classes, N-(4-{2-[(2-cyanophenyl)amino][1,2,4]triazolo[1,5-a]pyridin-6-yl}phenyl)-2-phenylacetamide (Mps-BAY1) (a triazolopyridine), N-cyclopropyl-4-{8-[(2-methylpropyl)amino]-6-(quinolin-5-yl)imidazo[1,2-a]pyrazin-3-yl}benzamide (Mps-BAY2a) and N-cyclopropyl-4-{8-(isobutylamino)imidazo[1,2-a]pyrazin-3-yl}benzamide (Mps-BAY2b) (two imidazopyrazines). By selectively inactivating MPS1, these small inhibitors can arrest the proliferation of cancer cells, causing their polyploidization and/or their demise. Cancer cells treated with Mps-BAY1 or Mps-BAY2a manifested multiple signs of mitotic perturbation including inefficient chromosomal congression during metaphase, unscheduled SAC inactivation and severe anaphase defects. Videomicroscopic cell fate profiling of histone 2B-green fluorescent protein-expressing cells revealed the capacity of MPS1 inhibitors to subvert the correct timing of mitosis as they induce a premature anaphase entry in the context of misaligned metaphase plates. Hence, in the presence of MPS1 inhibitors, cells either divided in a bipolar (but often asymmetric) manner or entered one or more rounds of abortive mitoses, generating gross aneuploidy and polyploidy, respectively. In both cases, cells ultimately succumbed to the mitotic catastrophe-induced activation of the mitochondrial pathway of apoptosis. Of note, low doses of MPS1 inhibitors and paclitaxel (a microtubular poison) synergized at increasing the frequency of chromosome misalignments and missegregations in the context of SAC inactivation. This resulted in massive polyploidization followed by the activation of mitotic catastrophe. A synergistic interaction between paclitaxel and MPS1 inhibitors could also be demonstrated in vivo, as the combination of these agents efficiently reduced the growth of tumor xenografts and exerted superior antineoplastic effects compared with either compound employed alone. Altogether, these results suggest that MPS1 inhibitors may exert robust anticancer activity, either as standalone therapeutic interventions or combined with microtubule-targeting chemicals.

P21-activated kinase 4 (PAK4) is required for metaphase spindle positioning and anchoring.

The oncogenic kinase PAK4 was recently found to be involved in the regulation of the G1 phase and the G2/M transition of the cell cycle. We have also identified that PAK4 regulates Ran GTPase activity during mitosis. Here, we show that after entering mitosis, PAK4-depleted cells maintain a prolonged metaphase-like state. In these cells, chromosome congression to the metaphase plate occurs with normal kinetics but is followed by an extended period during which membrane blebbing and spindle rotation are observed. These bipolar PAK4-depleted metaphase-like spindles have a defective astral microtubule (MT) network and are not centered in the cell but are in close contact with the cell cortex. As the metaphase-like state persists, centrosome fragmentation occurs, chromosomes scatter from the metaphase plate and move toward the spindle poles with an active spindle assembly checkpoint, a phenotype that is reminiscent of cohesion fatigue. PAK4 also regulates the acto-myosin cytoskeleton and we report that PAK4 depletion results in the induction of cortical membrane blebbing during prometaphase arrest. However, we show that membrane blebs, which are strongly enriched in phospho-cofilin, are not responsible for the poor anchoring of the spindle. As PAK4 depletion interferes with the localization of components of the dynein/dynactin complexes at the kinetochores and on the astral MTs, we propose that loss of PAK4 could induce a change in the activities of motor proteins.

Mps1 is a kinetochore-associated kinase essential for the vertebrate mitotic checkpoint.

The mitotic checkpoint acts to inhibit entry into anaphase until all chromosomes have successfully attached to spindle microtubules. Unattached kinetochores are believed to release an activated form of Mad2 that inhibits APC/C-dependent ubiquitination and subsequent proteolysis of components needed for anaphase onset. Using Xenopus egg extracts, a vertebrate homolog of yeast Mps1p is shown here to be a kinetochore-associated kinase, whose activity is necessary to establish and maintain the checkpoint. Since high levels of Mad2 overcome checkpoint loss in Mps1-depleted extracts, Mps1 acts upstream of Mad2-mediated inhibition of APC/C. Mps1 is essential for the checkpoint because it is required for recruitment and retention of active CENP-E at kinetochores, which in turn is necessary for kinetochore association of Mad1 and Mad2.

The interplay between cyclin-B-Cdc2 kinase (MPF) and MAP kinase during maturation of oocytes.

Throughout oocyte maturation, and subsequently during the first mitotic cell cycle, the MAP kinase cascade and cyclin-B-Cdc2 kinase are associated with the control of cell cycle progression. Many roles have been directly or indirectly attributed to MAP kinase and its influence on cyclin-B-Cdc2 kinase in different model systems; yet a principle theme does not emerge from the published literature, some of which is apparently contradictory. Interplay between these two kinases affects the major events of meiotic maturation throughout the animal kingdom, including the suppression of DNA replication, the segregation of meiotic chromosomes, and the prevention of parthenogenetic activation. Central to many of these events appears to be the control by MAP kinase of cyclin translation and degradation.

CENP-E as an essential component of the mitotic checkpoint in vitro.

Accurate chromatid separation is monitored by a checkpoint mechanism that delays anaphase onset until all centromeres are correctly attached to the mitotic spindle. Using Xenopus egg extracts, the kinetochore-associated microtubule motor protein CENP-E is now found to be required for establishing and maintaining this checkpoint. When CENP-E function is disrupted by immunodepletion or antibody addition, extracts fail to arrest in response to spindle damage. Mitotic arrest can be restored by addition of high levels of soluble MAD2, demonstrating that the absence of CENP-E eliminates kinetochore-dependent signaling but not the downstream steps in checkpoint signal transduction. Because it directly binds both to spindle microtubules and to the kinetochore-associated checkpoint kinase BUBR1, CENP-E is a central component in the vertebrate checkpoint that modulates signaling activity in a microtubule-dependent manner.

CENP-E forms a link between attachment of spindle microtubules to kinetochores and the mitotic checkpoint.

Here we show that suppression of synthesis of the microtubule motor CENP-E (centromere-associated protein E), a component of the kinetochore corona fibres of mammalian centromeres, yields chromosomes that are chronically mono-orientated, with spindles that are flattened along the plane of the substrate. Despite apparently normal microtubule numbers and the continued presence at kinetochores of other microtubule motors, spindle poles fragment in the absence of CENP-E, which implicates this protein in delivery of components from kinetochores to poles. CENP-E represents a link between attachment of spindle microtubules and the mitotic checkpoint signalling cascade, as depletion of this motor leads to profound checkpoint activation, whereas immunoprecipitation reveals a nearly stoichiometric association of CENP-E with the checkpoint kinase BubR1 during mitosis.

MAP kinase inactivation is required only for G2-M phase transition in early embryogenesis cell cycles of the starfishes Marthasterias glacialis and Astropecten aranciacus.

Downregulation of MAP kinase is a universal consequence of fertilization in the animal kingdom. Here we show that oocytes of the starfishes Astropecten aranciacus and Marthasterias glacialis complete meiotic maturation and form a pronucleus when treated with 1-methyladenine and then complete DNA replication and arrest at G2 if not fertilized. Release of G2 by fertilization or a variety of parthenogenetic treatments is associated with inactivation of MAP kinase. Prevention of MAP kinase inactivation by microinjection of Ste11-DeltaN, a constitutively active budding yeast MAP kinase kinase kinase, arrests fertilized eggs at G2 in either the first or the second mitotic cell cycle, in a dose-dependent manner. G1 arrest is never observed. Conversely, inactivation of MAP kinase by microinjection of the MAP kinase-specific phosphatase Pyst-1 releases mature starfish oocytes from G2 arrest. The role of MAP kinase in arresting cell cycle at various stages in oocytes of different animal species is discussed.

Stepwise reconstitution of interphase microtubule dynamics in permeabilized cells and comparison to dynamic mechanisms in intact cells.

Microtubules in permeabilized cells are devoid of dynamic activity and are insensitive to depolymerizing drugs such as nocodazole. Using this model system we have established conditions for stepwise reconstitution of microtubule dynamics in permeabilized interphase cells when supplemented with various cell extracts. When permeabilized cells are supplemented with mammalian cell extracts in the presence of protein phosphatase inhibitors, microtubules become sensitive to nocodazole. Depolymerization induced by nocodazole proceeds from microtubule plus ends, whereas microtubule minus ends remain inactive. Such nocodazole-sensitive microtubules do not exhibit subunit turnover. By contrast, when permeabilized cells are supplemented with Xenopus egg extracts, microtubules actively turn over. This involves continuous creation of free microtubule minus ends through microtubule fragmentation. Newly created minus ends apparently serve as sites of microtubule depolymerization, while net microtubule polymerization occurs at microtubule plus ends. We provide evidence that similar microtubule fragmentation and minus end-directed disassembly occur at the whole-cell level in intact cells. These data suggest that microtubule dynamics resembling dynamics observed in vivo can be reconstituted in permeabilized cells. This model system should provide means for in vitro assays to identify molecules important in regulating microtubule dynamics. Furthermore, our data support recent work suggesting that microtubule treadmilling is an important mechanism of microtubule turnover.

The Polo-like kinase Plx1 is a component of the MPF amplification loop at the G2/M-phase transition of the cell cycle in Xenopus eggs.

We have investigated whether Plx1, a kinase recently shown to phosphorylate cdc25c in vitro, is required for activation of cdc25c at the G2/M-phase transition of the cell cycle in Xenopus. Using immunodepletion or the mere addition of an antibody against the C terminus of Plx1, which suppressed its activation (not its activity) at G2/M, we show that Plx1 activity is required for activation of cyclin B-cdc2 kinase in both interphase egg extracts receiving recombinant cyclin B, and cycling extracts that spontaneously oscillate between interphase and mitosis. Furthermore, a positive feedback loop allows cyclin B-cdc2 kinase to activate Plx1 at the G2/M-phase transition. In contrast, activation of cyclin A-cdc2 kinase does not require Plx1 activity, and cyclin A-cdc2 kinase fails to activate Plx1 and its consequence, cdc25c activation in cycling extracts.

MAPK inactivation is required for the G2 to M-phase transition of the first mitotic cell cycle.

Down-regulation of MAP kinase (MAPK) is a universal consequence of fertilization in the animal kingdom, although its role is not known. Here we show that MAPK inactivation is essential for embryos, both vertebrate and invertebrate, to enter first mitosis. Suppressing down-regulation of MAPK at fertilization, for example by constitutively activating the upstream MAPK cascade, specifically suppresses cyclin B-cdc2 kinase activation and its consequence, entry into first mitosis. It thus appears that MAPK functions in meiotic maturation by preventing unfertilized eggs from proceeding into parthenogenetic development. The most general effect of artificially maintaining MAPK activity after fertilization is prevention of the G2 to M-phase transition in the first mitotic cell cycle, even though inappropriate reactivation of MAPK after fertilization may lead to metaphase arrest in vertebrates. Advancing the time of MAPK inactivation in fertilized eggs does not, however, speed up their entry into first mitosis. Thus, sustained activity of MAPK during part of the first mitotic cell cycle is not responsible for late entry of fertilized eggs into first mitosis.

Mitogen-activated protein kinase activation down-regulates a mechanism that inactivates cyclin B-cdc2 kinase in G2-arrested oocytes.

The G2 arrest of oocytes from frogs, clams, and starfish requires that preformed cyclin B-cdc2 complexes [prematuration-promoting factor (MPF)] be kept in an inactive form that is largely due to inhibitory phosphorylation of this pre-MPF. We have investigated the role of mitogen-activated protein (MAP) kinase in the activation of this pre-MPF. The cytoplasm of both frog and starfish oocytes contains an activity that can rapidly inactivate injected MPF. When the MAP kinase of G2-arrested starfish or Xenopus oocytes was prematurely activated by microinjection of c-mos or Ste-11 delta N fusion proteins, the rate and extent of MPF inactivation was much reduced. Both effects were suppressed by expression of the specific MAP kinase phosphatase Pyst 1. These results show that MAP kinase down-regulates a mechanism that inactivates cyclin B-cdc2 kinase in Xenopus oocytes. In starfish oocytes, however, MAP kinase activation occurs only after germinal vesicle breakdown, much after MPF activation. In this case, down-regulation of the cyclin B-cdc2 inhibiting pathway is a sensitive response to hormonal stimulation that does not require MAP kinase activation.

MAP kinase does not inactivate, but rather prevents the cyclin degradation pathway from being turned on in Xenopus egg extracts.

Unfertilized frog eggs arrest at the second meiotic metaphase, due to cytostatic activity of the c-mos proto-oncogene (CSF). MAP kinase has been proposed to mediate CSF activity in suppressing cyclin degradation. Using an in vitro assay to generate CSF activity, and recombinant CL 100 phosphatase to inactivate MAP kinase, we confirm that the c-mos proto-oncogene blocks cyclin degradation through MAP kinase activation. We further show that for MAP kinase to suppress cyclin degradation, it must be activated before cyclin B-cdc2 kinase has effectively promoted cyclin degradation. Thus MAP kinase does not inactivate, but rather prevents the cyclin degradation pathway from being turned on. Using a constitutively active mutant of Ca2+/calmodulin dependent protein kinase II, which mediates the effects of Ca2+ at fertilization, we further show that the kinase can activate cyclin degradation in the presence of both MPF and the c-mos proto-oncogene without inactivating MAP kinase.

The proteolysis-dependent metaphase to anaphase transition: calcium/calmodulin-dependent protein kinase II mediates onset of anaphase in extracts prepared from unfertilized Xenopus eggs.

It has been shown, using spindles assembled in vitro in extracts containing CSF (the cytostatic factor responsible for arresting unfertilized vertebrate eggs at metaphase), that onset of anaphase requires Ca(2+)-dependent activation of the ubiquitin-dependent proteolytic pathway that destroys both mitotic cyclins and an unknown protein responsible for metaphase arrest (Holloway et al., 1993, Cell, 73, 1382-1402). We showed recently that Ca2+/calmodulin-dependent protein kinase II (CaM KII) activates the ubiquitin-dependent cyclin degradation pathway in CSF extracts (Lorca et al., 1993, Nature, 366, 270-273), but did not investigate its possible effect on sister chromatid segregation. In this work we identify CaM KII as the only target of Ca2+ in inducing anaphase in CSF extracts, and further show that transition to anaphase does not require the direct phosphorylation of metaphase spindle components by CaM KII. A possible interpretation of the above results could have been that the ubiquitin-dependent degradation pathway is required for onset of anaphase only when spindles are clamped at metaphase due to CSF activity, and not in the regular cell cycle that occurs in the absence of CSF activity. We ruled out this possibility by showing that competitive inhibition of the ubiquitin-dependent degradation pathway still prevents the onset of anaphase in cycling extracts that lack CSF and do not require Ca2+ for sister chromatid separation.