Novel Mps1 Kinase Inhibitors with Potent Antitumor Activity

Cell-cycle deregulation represents one of the classical hallmarks of cancer, and cell-cycle arrest is the predominant mode of action of antimitotic cancer drugs (e.g., taxanes and vinca alkaloids). Targeted disruption of mitotic cell-cycle checkpoints offers a novel approach to cancer treatment: driving tumor cells into cell division despite DNA damage or unattached/misattached chromosomes, resulting in a lethal degree of DNA damage or aneuploidy.

Monopolar spindle 1 (Mps1), also known as TTK, is a serine threonine kinase, which ensures proper biorientation of sister chromatids on the mitotic spindle by the activation of the spindle assembly checkpoint (SAC). The SAC controls attachments between microtubules (MT) and kinetochores (KT) during prometaphase and blocks transition to anaphase until all chromosomes are correctly tensed, attached, and bioriented at the metaphase plate (1). Mps1 therefore plays an important role in securing proper chromosome alignment, orientation, and segregation during mitosis.

Mps1 mediates direct activation of SAC components, including kinetochore null protein 1 (KNL1), mitotic arrest deficient (MAD) 1/2, and budding uninhibited by benzimidazole (Bub) 1, 3 and related 1 (BubR1) and/or their recruitment to unoccupied KTs (2–10). Mps1-mediated direct phosphorylation of KNL1 is required for Bub1 and Bub3 kinetochore localization (9,10). When activated checkpoint proteins are released to the cytosol, they catalyze the assembly of the mitotic checkpoint complex (MCC), which inhibits the anaphase-promoting complex (APC)–mediated polyubiquitination of mitotic proteins, for example, cyclin B and securin, thereby delaying anaphase onset until the checkpoint is satisfied (1,11,12). Furthermore, Mps1 has been reported to be implicated in an error correction mechanism that resolves erroneous KT–MT attachments, which are frequently formed in early mitosis, by phosphorylation of the chromosome passenger complex subunit borealin (13,14). Accordingly, the inhibition of Mps1 abrogates the SAC, thereby leading to shortened mitosis, chromosome missegregation, aneuploidy or polyploidy, and cell death (13,15–17).

Mps1 is expressed during mitosis in proliferating cells. Overexpression of Mps1 is observed in a number of cancer cell lines and tumor types, including breast cancer, anaplastic thyroid carcinoma, and lung cancers (18–21), correlating with high histologic grade, tumor aggressiveness, and aneuploidy/CIN in breast cancer (18). Cancer-associated frameshift mutations, associated with a premature stop of Mps1 biosynthesis, are frequently found in gastric and colorectal tumors with microsatellite instability (22).

On the basis of these findings, Mps1 has been considered as one among the most promising drug targets for cancer therapy (23). In recent years, several Mps1-inhibitory compounds have been identified and explored for their anticancer activity in preclinical assays (12,24), although none have entered clinical development.

Established antimitotic drugs, such as vinca alkaloids, taxanes, or epothilones, activate the SAC either by stabilizing or destabilizing microtubule dynamics, resulting in a mitotic arrest. This arrest prevents separation of sister chromatids to form the two daughter cells. Dependent on the levels of cyclin B1 versus prosurvival proteins prolonged arrest in mitosis forces a cell either into mitotic exit without cytokinesis (“mitotic slippage”) or into mitotic catastrophe, leading to cell death (25). In contrast, inhibitors of Mps1 induce SAC inactivation, which accelerates progression of cells through mitosis, resulting in chromosomal segregation errors and finally cell death. Silencing of Mps1 leads to the failure of cells to arrest in mitosis in response to antimitotic drugs. Remarkably, the combination of microtubule-interfering agents and Mps1 inhibition even increases chromosomal segregation errors and cell death (26,27). Therefore, the combined increase of chromosomal segregation errors induced by combination of antimitotics with SAC inhibition constitutes an efficient strategy for eliminating tumor cells. Altogether, the inhibition of Mps1 in combination with microtubule-targeting agents represents a valuable mechanism that is expected to improve therapeutic efficacy of antimitotic drugs.

In a previous work, we have presented Mps1 kinase inhibitors from two structural distinct chemical classes (28). Derived from these structures, we have developed two novel, potent, and highly selective Mps1 kinase inhibitors, BAY 1161909 and BAY 1217389. Here, we present, for the first time, their chemical structures and pharmacologic in vitro and in vivo profiles. Both compounds are currently in phase I clinical trials (NCT02138812 and NCT02366949).

BAY 1161909, (2R)-2-(4-fluorophenyl)-N-[4-(2-{[2-methoxy-4-(methylsulfonyl)phenyl]amino}[1,2,4]triazolo[1,5-a]pyridin-6-yl)phenyl]propanamide, and BAY 1217389, N-cyclopropyl-4-{6-(2,3-difluoro-4-methoxyphenoxy)-8-[(3,3,3-trifluoropropyl)amino]imidazo[1,2-b]pyridazin-3-yl}-2-methylbenzamide, were synthesized at Bayer Pharma AG. Paclitaxel was purchased from Bristol-Myers Squibb, and cisplatin from Sigma-Aldrich.

The inhibition of recombinant human Mps1 by BAY 1161909 or BAY 1217389 was assessed in TR-FRET–based in vitro kinase assays via phosphorylation of a biotinylated peptide (biotin-Ahx-PWDPDDADITEILG-NH2). Kinase and test compound were preincubated for 15 minutes before enzyme reaction was started by the addition of substrate and ATP upon 10 μmol/L. For further details, see Supplementary Materials and Methods.

BAY 1161909 and BAY 1217389 were counterscreened against a panel of kinases using the Millipore Kinase or DiscoveRx profiler screen (29). BAY 1161909 was initially tested at 1 μmol/L in the DiscoveRx kinase panel, followed by K_D determination for 11 kinases (Supplementary Table S1). BAY 1161909 was tested at 10 μmol/L in the Millipore kinase panel, followed by retesting at 1 and 0.1 μmol/L and IC₅₀ determination for JNK1alpha, JNK2alpha, and JNK3 (Supplementary Table S1). BAY 1217389 was initially tested at 1, 0.1, and 0.01 μmol/L in the DiscoveRx kinase panel (Supplementary Table S2).

Tumor cell lines were obtained either from the ATCC or from the German Collection of Microorganisms and Cell Cultures. HeLa-MaTu and HeLa-MaTu-ADR cells were obtained from Epo GmbH. Authentication of all human cell lines used was performed at the German Collection of Microorganisms and Cell Cultures. Cells were propagated under the suggested growth conditions in a humidified 37°C incubator.

Cells were seeded into 96-well plates at densities ranging from 1,000 to 5,000 cells per well in the appropriate medium supplemented with 10% FCS. After 24 hours, cells were treated in quadruplicates with serial dilutions of compounds. After further 96 hours, adherent cells were fixed with glutaraldehyde and stained with crystal violet. IC₅₀ values were calculated by means of a 4-parameter fit using the company's own software.

HeLa (ATCC CCL-2) cells were plated at a density of 1,000 cells per well in a 1,536-well microtiter plate in 2 μL cell culture medium and incubated overnight at 37°C. G₂–M arrest was initiated by adding 0.1 μg/mL nocodazole for 24 hours. Cells were then treated for 4 hours at 37°C in the presence of test compounds (0 μmol/L, 0.005–10 μmol/L) solubilized in DMSO 0.5% (v/v), fixed with 4% (v/v) paraformaldehyde (PFA), permeabilized with 0.5% (v/v) Triton X-100, and blocked with 1% (v/v) BSA in PBS. After washing with PBS, the phosphorylation of histone H3 at serine 10 was labeled by antibody detection (Upstate Biotechnology, cat# 16-222). Nuclei were marked with Hoechst 33342 (Life Technologies, cat# H1399). After washing, cells were covered with PBS and stored at 4°C until image acquisition. Images for high-throughput microscopy were acquired with a PerkinElmer Opera High-Content Analysis reader. Images were analyzed with image analysis software MetaXpress from Molecular Devices utilizing the Mitotic Index Application Module. Assay data were further analyzed by 4-parameter Hill equation using Genedata's Assay Analyzer and Condoseo software.

U-2 OS (osteosarcoma ATCC: HTB-96) cells were plated at a density of 2,500 cells per well in a 384-well microtiter plate in 20 μL cell culture medium and incubated overnight at 37°C. BAY 1161909 or BAY 1217389 was added at a final concentration of 100 nmol/L in triplicates. Cells were treated for 0, 24, 48, and 72 hours at 37°C in the presence of test compounds. Thereafter, cells were fixed with 4% (v/v) PFA, permeabilized with 0.5% (v/v) Triton X-100, and blocked with 0.5% (v/v) BSA in PBS. α-Tubulin structures were detected by antibody labeling (Abcam, cat# ab7291). After blocking with goat IgG (Jackson ImmunoResearch, cat# 005-000-003), secondary antibodies (Life Technologies, cat# A-11017) were applied in the blocking solution. Cells were washed with PBS, and nuclei were marked with Hoechst 33342 (Life Technologies, cat# H1399). Finally, cells were washed and covered with PBS and stored at 4°C until image acquisition. Images were acquired with a PerkinElmer Opera High-Content Analysis reader.

Pharmacokinetic studies were performed in male Wistar rats and female CD1 or NMRI nu/nu mice. For i.v. studies in rats and mice, BAY 1161909 was solubilized in 1% DMSO, 99% plasma; for p.o. studies in rats in 50% polyethylene glycol (PEG) 400, 10% ethanol, 40% water, and for p.o. studies in mice in 75% PEG 400, 5% ethanol, and 25% solutol. BAY 1217389 was solubilized in 50% PEG 400, 10% ethanol, and 40% water for intravenous and p.o. dosing in rats and mice. In pharmacokinetic studies, plasma samples were collected after 2, 5, 15, 30, 45 minutes, 1, 2, 4, 6, 8, and 24 hours after intravenous application and after 8, 15, 30, 45 minutes, 1, 2, 4, 6, 8, and 24 hours after p.o. administration and precipitated with ice-cold acetonitrile (1:5). Supernatants were analyzed via LC/MS-MS. Pharmacokinetic parameters were estimated from the plasma concentration data, e.g., using the log-linear trapezoidal rule for AUC estimation. Maximal plasma concentrations (C_max) and time thereof (T_max) were taken directly from the concentration time profiles.

Housing and handling of animals was in strict compliance with European and German Guidelines for Laboratory Animal Welfare. For tumor xenograft studies, female athymic NMRI nu/nu mice (Taconic), 50 days old, average body weight 20 to 22 g, were used after an acclimatization period of 14 days. Feeding and drinking was ad libitum 24 hours per day. Human tumor cells derived from exponentially growing cell cultures were resuspended for A2780cis, NCI-H1299, and SUM-149 models in 100% Matrigel (BD Biosciences) to a final concentration of 2 × 10⁷, 3 × 10⁷, or 5 × 10⁷ cells/mL, respectively. Subcutaneous implants of 0.1 mL of 2 × 10⁶ A2780cis, 3 × 10⁶ NCI-H1299, or 5 × 10⁶ SUM-149 cells were inoculated into the inguinal region of athymic mice. Tumor fragments of patient tumor explants MAXF 1384 or LU384, obtained from serial passage in nude mice, were cut into fragments of 4 to 5 mm diameter and transplanted subcutaneously into the flank of athymic mice. Tumor area (product of the longest diameter and its perpendicular), measured with a caliper, and body weight were determined two to three times a week. Tumor growth inhibition is presented as treatment/control ratio (T/C) calculated with tumor areas at the end of the study. Animal body weight was used as a measure for treatment-related toxicity. Body weight loss > 20% was dedicated as toxic. When tumors reached a size of approximately 20 to 40 mm², depending on growth of the tumor model, animals were randomized to treatment and control groups (8–10 mice/group) and treated p.o. with vehicle (70% PEG 400, 5 % ethanol, and 25% Solutol), BAY 1161909, BAY 1217389, and/or paclitaxel, as indicated in Tables and Figure legends. In combination treatment groups, Mps1 inhibitor and paclitaxel were applied at the same day within a time frame of 1 hour. The treatment of each animal was based on individual body weight. Animals were euthanized according to the German Animal Welfare Guidelines. Data were expressed as means ± SD. Statistical analysis included one-way ANOVA, and differences to the control were compared versus control group by pair-wise comparison procedure using the SigmaStat software.

For analysis of polyploidy and multinuclearity induction in vivo, A2780cis tumor–bearing female NMRI nude mice (see above) were treated with paclitaxel (intravenously once with 24 mg/kg), BAY 1161909 (p.o. twice daily for 2 days with 2.5 mg/kg), and in combination with paclitaxel (i.v. once 24 mg/kg) and BAY 1161909 (p.o. twice daily for 2 days 1 mg/kg). Treatment for all groups started at a tumor size of 60 mm² at day 14 after tumor cell inoculation. Tumor samples were prepared 4 and 8 hours after first BAY 1161909 application at treatment day 1, as well as 4, 8, and 24 hours after first application of BAY 1161909 on treatment day 2. At each time point, 3 animals per treatment group were analyzed. Tumors were used for histologic examination after paraffin embedding and hematoxylin and eosin staining.

For determination of phosphorylated KNL1 or total KNL1 in tumor tissue, A2780cis tumor–bearing female NMRI nude mice (see above) were treated with vehicle, BAY 1217389 alone p.o. for 2 days with 1, 2, 4, or 8 mg/kg (day 1 twice daily, day 2 once), paclitaxel alone intravenously once 24 mg/kg, and in combination with paclitaxel (i.v. once 24 mg/kg) and BAY 1217389 (p.o for 2 days with 1, 2, or 4 mg/kg, day 1 twice daily, day 2 once). Treatment for all groups started at a tumor size of 60 mm² at day 15 after tumor cell inoculation. Tumor samples were prepared 36 hours after the first treatment at day 1. Five animals per treatment group were analyzed. Tumors were used for histologic examination after paraffin embedding and pKNL1 staining.

Peripheral blood was collected in K-EDTA tubes for the determination of relative reticulocyte counts (%RET) and micronuclei-containing reticulocytes counts (%MN-RET) using the MicroFlow Plus Kit (mouse blood) from Litron Laboratories and the Flow Cytometer BD FACSCanto II (BD Biosciences). Sample collection, processing, and analysis were performed according to the manufacturer's instructions (version 130712).

An antibody directed against phosphothreonine 875 in KNL1 protein was produced by immunization of rabbits with a phospho-KNL1 peptide as described previously (9). Sera with highest phospho-KNL1 specific titer were used for IHC. Paraffin-embedded tumor sections were stained using the rabbit anti-phosphothreonine 875 KNL1 antibody (1:400) in DAKO antibody diluent (S3022) and anti-rabbit Envision secondary antibody (DAKO K4010). Stained slides were scanned using a Mirax Slide Scanner and evaluated for positive pixels above background in three independent fields per slide. Total KNL1 peptide was detected with the primary rabbit anti-human CASC5 antibody (Sigma-Aldrich HPA026624) in a 1:400 dilution. The development and quantification of KNL1 staining was performed accordingly to pKNL1.

In biochemical assays, BAY 1161909 and BAY 1217389, two novel Mps1 kinase inhibitors from different structural classes (Fig. 1), inhibited Mps1 kinase activity, with IC₅₀ values of 0.34 ± 0.09 nmol/L and 0.63 ± 0.27 nmol/L, respectively.

Both compounds showed high selectivity against other kinases. When assayed against the Millipore panel (230 kinases), BAY 1161909 inhibited only two kinases, JNK2 and JNK3, more than 50% at a concentration of 1 μmol/L and no other kinase at 100 nmol/L (Supplementary Table S1). BAY 1217389 was tested against the DiscoveRx kinase panel (395 kinases) and was found to bind to PDGFRβ (<10 nmol/L), Kit (between 10 and 100 nmol/L), CLK1, CLK2, CLK4, JNK1, JNK2, JNK3, LATS1, MAK, MAPKAP2, MERTK, p38β, PDGFRα, PIP5K1C, PRKD1, and RPS6KA5 (between 100 and 1,000 nmol/L; Supplementary Table S2). Thus, BAY 1161909 and BAY 1217389 both showed higher kinase selectivity as compared with previously described Mps1 inhibitors, such as SP600125 and reversine (Supplementary Table S3; refs. 7,30).

In cellular mechanistic assays, BAY 1161909 and BAY 1217389 abrogated nocodazole-induced metaphase arrest in HeLa cells with IC₅₀ values of 56 ± 21 nmol/L and 0.11 ± 0.006 nmol/L, respectively. In addition, the treatment of osteosarcoma cells U-2 OS with BAY 1161909 led to nuclear enlargement, multilobulation of nuclei, and generation of multinucleated cells, resulting in strong aneuploidy after several cell cycle turns with blocked Mps1 activity (Supplementary Fig. S1). Thr-875 within a MELT repeat of the KNL1 protein has been described as a primary phosphorylation site of Mps1 kinase (9). BAY 1161909 and BAY 1217389 abolished KNL1 Thr-875 phosphorylation in A2780 cells pretreated with 100 nmol/L paclitaxel within 30 minutes, indicating rapid inactivation of intracellular Mps1 kinase (IC₅₀, BAY 1161909 84 nmol/L, IC₅₀, BAY 1217389 12 nmol/L; data not shown).

The antiproliferative activity of BAY 1161909 and BAY 1217389 was evaluated against panels of 86 and 68 tumor cell lines covering various cancer histologies, multidrug-resistant cell lines, and syngeneic pairs of TP53 wild-type (WT) and TP53 knockout cells (Supplementary Table S4). BAY 1161909 inhibited the proliferation with a median IC₅₀ of 160 nmol/L (range 32 to >3,000 nmol/L), whereas BAY 1217389 was found to inhibit cell proliferation with a median IC₅₀ of 6.7 nmol/L (range 3 to >300 nmol/L). Sensitive and rather insensitive cell lines were identified, and overall, both compounds showed a very similar sensitivity pattern based on the comparison of relative IC₅₀s (ratio of cell line IC₅₀ vs. median IC₅₀; Supplementary Table S4). Attempts to correlate cell line sensitivity or insensitivity towards Mps1 inhibition with gene mutations (Sanger) did not lead to the identification of single-gene mutations with statistical significance (data not shown). Even the comparison of the two pairs of isogenic TP53 WT and TP53 knockout HCT116 and RKO cell lines revealed only slightly lower IC₅₀s for the cell lines with a TP53–WT background, which is in line with our previous findings (28).

Taken together, BAY 1161909 and BAY 1217389 represent two potent Mps1 inhibitors with high-kinase selectivity, specific cellular on-target activity, and antiproliferative potency. Although cell lines with high and low sensitivity towards Mps1 inhibition were identified, no single-gene mutation could be correlated, indicating that multiple genes may contribute to cellular sensitivity to SAC inhibition or that other local factors at the kinetochore, such as the presence of phosphatases (10), may determine the cellular response to Mps1 inhibition.

Pharmacokinetic parameters were determined in mouse and rat. Following intravenous administration of BAY 1161909 as bolus of 0.5 mg/kg to male CD1 mouse and 0.5 mg/kg to male Wistar rat, the compound exhibited low blood clearance. The volume of distribution (V_ss) was high in both species and terminal half-lives were long. After oral administration of 1 mg/kg to female NMRI mouse and 0.5 mg/kg to male Wistar rat, peak plasma levels were reached after 4 hours. The oral bioavailability was moderate in mouse and rat (Table 1). BAY 1217389 was administered intravenously as bolus of 1.0 and 0.5 mg/kg to female CD1 mouse and male Wistar rat, respectively. Blood clearance was found to be low in the tested species. V_ss was high and terminal half-lives were long. BAY 1217389 was administered orally to female NMRI mouse (1 mg/kg) and male Wistar rat (0.5 mg/kg). Peak plasma concentrations were observed between 1.5 and 7 hours. Oral bioavailability was high in rat and moderate in mouse (Table 1).

Our data demonstrate that we have identified two novel Mps1 kinase inhibitors with a favorable pharmacokinetic profile, supporting further development for clinical application.

Therapeutic efficacy and tolerability of the Mps1 inhibitors BAY 1161909 or BAY 1217389 was investigated in monotherapy using the A2780cis human cisplatin-resistant ovarian tumor xenograft model. Athymic mice bearing established A2780cis xenograft tumors were treated with MTDs of BAY 1161909 or BAY 1217389 in an intermittent 2 days on/5 days off dosing schedule in comparison with cisplatin. As expected, cisplatin treatment resulted in weak antitumor efficacy in A2780cis tumors (T/C 0.75). BAY 1161909 showed moderate antitumor efficacy (T/C 0.43; Fig. 2A) but achieved statistically significant improvement of tumor growth inhibition compared with vehicle control and cisplatin treatment at overall good tolerability (Fig. 2B). Comparable antitumor efficacy was achieved for BAY 1217389 (T/C of 0.53) at acceptable tolerability (Fig. 2C and D).

Our data demonstrate that monotreatment with Mps1 inhibitors BAY 1161909 or BAY 1217389 can achieve moderate antitumor efficacy at MTDs, as it has been presented for other Mps1 inhibitors before (2).

As antimitotics/antitubulins lead to mitotic checkpoint activation, checkpoint overrun by Mps1 kinase inhibition should impact antitumor activity in a combination treatment setting (Figs. 3, 4, and 5; Supplementary Tables S5 and S6). On the basis of our mechanistic in vitro findings supported by shRNA knockdown experiments in combination with low doses of paclitaxel showing synergistic effects (31), we tested the ability of our Mps1 inhibitors to inhibit tumor growth in combination with paclitaxel.

The response of human triple-negative breast cancer xenograft tumors (no expression of Her2/neu, progesterone receptor, and estrogen receptor) to treatment with Mps1 inhibitor in combination with paclitaxel was tested in tumor models that develop paclitaxel insensitivity during continued paclitaxel treatment, the cell line derived model SUM-149, and the patient-derived tumor model MAXF 1384 (Fig. 3). In the SUM-149 xenograft model, the effect of combination treatment of BAY 1217389 and paclitaxel was studied. BAY 1217389 was applied upon dosing with 30% of MTD in a twice-daily intermittent (2 days on/5 days off) dosing schedule in combination with paclitaxel upon its respective MTD. A clear tumor growth delay was observed upon combination of paclitaxel and BAY 1217389 after 48 treatment days, achieving statistically significant improvement over paclitaxel monotherapy efficacy at acceptable tolerability (Fig. 3A). In the patient-derived triple-negative breast cancer xenograft model MAXF 1384, BAY 1161909 was applied upon dosing with 80% of MTD in a twice-daily intermittent (2 days on/5 days off) dosing schedule in combination with paclitaxel upon its respective MTD. After 42 treatment days, statistically significant improvement over paclitaxel monotherapy efficacy was achieved in the paclitaxel/BAY 1161909 combination treatment group, inducing complete tumor remission (Fig. 3C).

The activity of Mps1 inhibitors in combination with paclitaxel was further tested in human lung carcinoma (NSCLC) xenograft models, the cell line–derived model NCI-H1299, which acquires paclitaxel insensitivity during continued paclitaxel treatment, and the intrinsically taxane resistant patient-derived model LU387 (Fig. 4). In the NCI-H1299 NSCLC xenograft model, the effect of combination treatment of BAY 1217389 and paclitaxel was tested. BAY 1217389 was applied upon dosing with 60% of MTD in the twice-daily intermittent (2 days on/5 days off) dosing schedule. Paclitaxel was applied upon its MTD. After 36 treatment days, statistically significant improvement over paclitaxel monotherapy efficacy was achieved by paclitaxel/BAY 1217389 combination (Fig. 4A). BAY 1217389 was slightly overdosed with 60% MTD in combination treatments as transiently critical body weight loss (>10%) and toxicity occurred (Fig. 4B). The intrinsically taxane-resistant patient-derived lung carcinoma xenograft model LU387 was used to study the effect of BAY 1161909/paclitaxel combination therapy. BAY 1161909 was applied upon dosing with doses 40% of MTD in the intermittent (2 days on/5 days off) dosing schedule in combination with paclitaxel, applied upon its respective MTD. The combination therapy of paclitaxel and BAY 1161909 achieved statistically significant reduction of tumor size compared with vehicle-treated control and monotherapy groups, inducing clear tumor growth delay at good tolerability of treatments (Fig. 4C and D).

Taken together, we show here for the first time that the Mps1 inhibitors BAY 1161909 and BAY 1217389 exhibit strong cooperativity with taxanes in acquired or intrinsically taxane-resistant tumor models of triple-negative breast cancer and lung carcinoma. Paclitaxel monotherapy efficacy upon dosing at the respective MTD could be significantly improved in combination with sub-MTD doses of Mps1 inhibitors at overall good tolerability. Therefore, we could validate the hypothesis that Mps1 inhibitors in combination with antitubulins, such as taxanes, can enhance their efficacy and potentially overcome resistance. This opens the opportunity to improve overall survival of taxane-resistant patients in indications where taxanes are used as standard of care, such as breast, lung or ovarian cancer.

To demonstrate that in vivo efficacy is mediated by the anticipated mode of action of Mps1 inhibition, different in vivo mechanistic assays were applied.

The induction of cellular polyploidy and multinuclearity was analyzed in A2780cis ovarian xenograft tissue after treatment with Mps1 inhibitor in monotherapy or combination with paclitaxel at doses that have shown in vivo antitumor efficacy. BAY 1161909 monotreated A2780cis tumors showed induction of a pleomorphic phenotype when compared with vehicle-treated control tumors, including multinuclearity, whereas paclitaxel treatment alone induced atypical mitoses, as well as increased necrosis and apoptosis. Combination treatment with paclitaxel and BAY 1161909 induced atypical mitoses as well as increased pleomorphism and multinucleated tumor cells in A2780cis ovarian tumors, demonstrating the expected mode of action of Mps1 inhibition (Supplementary Fig. S2). Incorrect distribution of chromosomal material leads in some cell types, e.g., erythrocytes or reticulocytes, to the formation of cytoplasmic bodies of chromosome fragments or whole chromosomes, so called micronuclei. Therefore, peripheral blood reticulocytes of Mps1 inhibitor–treated mice were analyzed. A clear increase of the fraction of micronuclei-harboring reticulocytes 3 days after BAY 1161909 treatment at doses ≥1 mg/kg twice daily for 2 days corresponding to ≥40% monotherapy MTD could be determined. Total reticulocyte counts dropped at doses ≥2 mg/kg twice daily for 2 days (≥80% monotherapy MTD; Supplementary Fig. S3). In addition, the effect of Mps1 inhibitor alone or in combination with paclitaxel on KNL1 Thr-875 phosphorylation has been determined in A2780cis ovarian tumor xenografts. Monotherapy of Mps1 inhibitor BAY 1217389 dosed upon 4 mg/kg (80% of efficacious dose in monotherapy) led to 80% reduction of basal pKNL1; 2 mg/kg led to 60% of basal pKNL1 level in A2780Cis tumors (Fig. 5A and B). Treatment with paclitaxel strongly induced the pKNL1 signal in A2780cis tumor tissue (Fig. 5C and D). Remarkably, cotreatment with Mps1 inhibitor BAY 1217389 was able to completely suppress paclitaxel-induced pKNL1 at an efficacious combination dose of BAY 1217389 (Fig. 5C and D). No changes of total KNL1 level were detectable (Supplementary Fig. S4), indicating a specific effect of Mps1 inhibitors on KNL1 phosphorylation.

These results demonstrate that Mps1 inhibitor treatment prevents correct distribution of chromosomal material during cell division in tumors as well as peripheral blood reticulocytes. Furthermore, KNL1 phosphorylation might serve as a direct read out of target engagement in tumors. Overall, our data suggest that an early pharmacodynamic effect of Mps1 inhibition in monotherapy or paclitaxel combination could be determined by micronuclei induction in peripheral blood reticulocytes or by monitoring the phosphorylation status of KNL1.

Antitubulin–based chemotherapy is used as standard of care in many cancer patients across a variety of tumor types, such as breast, lung, or ovarian cancers. Resistance to antitubulins is a common drawback in cancer chemotherapy (32). Alternative treatment options are often not available, whereby antimitotics/antitubulins are still the only treatment option for patients with aggressive cancer types. Therefore, we have evaluated whether efficacy of antitubulin reagents (taxanes) can be improved by combination treatment with Mps1 inhibitors in line with their mechanisms of action.

Our findings validate the innovative concept of SAC abrogation and justify clinical proof-of-concept studies evaluating Mps1 inhibitors BAY 1161909 and BAY 1217389 in combination with antimitotic cancer drugs to enhance their efficacy and potentially overcome resistance. BAY 1161909 and BAY 1217389 are currently in phase I clinical trials (NCT02138812 and NCT02366949). To our knowledge, these are the first clinical trials of Mps1 kinase inhibitors.

All authors are employees of Bayer Pharma AG. A.M. Wenger, P. Lienau, B. Bader, G. Siemeister, D. Kosemund, and U. Klar have ownership interests (including patents) in Bayer Pharma AG. No potential conflicts of interest were disclosed by the other authors.

Conception and design: A.M. Wengner, G. Siemeister, M. Koppitz, V. Schulze, D. Kosemund, U. Klar, B. Bader, M. Michels, B. Kreft, F. von Nussbaum, M. Brands, D. Mumberg

Development of methodology: A.M. Wengner, G. Siemeister, V. Schulze, D. Kosemund, U. Klar, B. Bader, S. Prechtl, O. von Ahsen

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.M. Wengner, G. Siemeister, D. Stoeckigt, R. Neuhaus, P. Lienau, B. Bader, S. Prechtl, M. Raschke, A.-L. Frisk, O. von Ahsen

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.M. Wengner, G. Siemeister, D. Kosemund, U. Klar, D. Stoeckigt, R. Neuhaus, P. Lienau, B. Bader, S. Prechtl, M. Raschke, A.-L. Frisk, O. von Ahsen, M. Michels, F. von Nussbaum

Writing, review, and/or revision of the manuscript: A.M. Wengner, G. Siemeister, M. Koppitz, V. Schulze, D. Kosemund, U. Klar, D. Stoeckigt, R. Neuhaus, P. Lienau, B. Bader, S. Prechtl, M. Raschke, A.-L. Frisk, O. von Ahsen, M. Michels, B. Kreft, F. von Nussbaum, M. Brands, D. Mumberg, K. Ziegelbauer

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.M. Wengner, G. Siemeister, M. Koppitz, V. Schulze, D. Kosemund, U. Klar, D. Stoeckigt, R. Neuhaus, P. Lienau, B. Bader, S. Prechtl, M. Raschke, A.-L. Frisk, O. von Ahsen, M. Michels, B. Kreft, D. Mumberg, K. Ziegelbauer

Study supervision: A.M. Wengner, P. Lienau, M. Brands, K. Ziegelbauer

Other (invention and synthesis of candidate compound): M. Koppitz

Other (design and chemical synthesis of novel Mps1 kinase inhibitors): V. Schulze

Other (medicinal chemistry): F. von Nussbaum

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.