Discovery and Characterization of a Novel MASTL Inhibitor MKI-2 Targeting MASTL-PP2A in Breast Cancer Cells and Oocytes

Although microtubule-associated serine/threonine kinase-like (MASTL) is a promising target for selective anticancer treatment, MASTL inhibitors with nano range potency and antitumor efficacy have not been reported. Here, we report a novel potent and selective MASTL inhibitor MASTL kinase inhibitor-2 (MKI-2) identified in silico through a drug discovery program. Our data showed that MKI-2 inhibited recombinant MASTL activity and cellular MASTL activity with IC50 values of 37.44 nM and 142.7 nM, respectively, in breast cancer cells. In addition, MKI-2 inhibited MASTL kinase rather than other AGC kinases, such as ROCK1, AKT1, PKACα, and p70S6K. Furthermore, MKI-2 exerted various antitumor activities by inducing mitotic catastrophe resulting from the modulation of the MASTL-PP2A axis in breast cancer cells. The MKI-2 treatment showed phenocopies with MASTL-null oocyte in mouse oocytes, which were used as a model to validate MKI-2 activity. Therefore, our study provided a new potent and selective MASTL inhibitor MKI-2 targeting the oncogenic MAST-PP2A axis in breast cancer cells.

Recently, several groups reported candidate compounds as potential MASTL inhibitors [22][23][24]. Greatwall kinase inhibitor-1 (GKI-1) was reported as a first-line inhibitor against MASTL, which was designed by using the solved protein structure of the kinase domain of MASTL [23]. GKI-1 can inhibit MASTL in vitro and disrupt mitotic events by decreasing phosphorylated ENSA with µM range potency in HeLa cells [23]. However, GKI-1 did not show the anticancer activity in breast cancer cells [22]. Further, two compounds identified in silico in another study could be potential MASTL inhibitors without any validated biochemical data [24]. We previously reported MASTL kinase inhibitor-1 (MKI-1), a new MASTL inhibitor, showing antitumor activities across in vitro and in vivo tumor models by inhibiting MASTL in breast cancer cells [22]. Nonetheless, MKI-1 also showed µM range potency and efficacy for MASTL inhibition [22]. In the present study, we developed a second-generation MASTL inhibitor MKI-2 using an in silico-based drug discovery program involving a docking model. MKI-2 showed nM range potency and efficacy for MASTL inhibition in breast cancer cells. In addition, we demonstrated that MKI-2 treatment showed a phenocopy with MASTL-null oocyte in mouse oocytes, which were used as a model to validate MKI-2 activity. Thus, our data provided a new potent MASTL inhibitor MKI-2.

Identification and Validation of MASTL Inhibitor MKI-2 In Silico and In Vitro
Since MKI-1, a first-line inhibitor of MASTL, showed µM range potency for the inhibition of MASTL [22], more potent and selective MASTL inhibitors were further screened in an in silico-based drug discovery program using a docking model for the MASTL-MKI-1 interaction. Briefly, 436 compounds were selected by the structure-based in silico analysis from 65,000 active chemical library. A total of 436 compounds were further analyzed using various biochemical assays, including the IN-Cell-based immunofluorescence, homogeneous time-resolved fluorescence (HTRF) KinEASE, ADP kinase, and cell viability assay ( Figure 1A). In these experiments, AT13148, a multiple-AGC kinase inhibitor [25], and MKI-1 were used as the positive controls. As a result, seven compounds were selected by the biochemical screens. As shown in Figure 1B-E, #248 (2-(4-((4-((5-cyclopropyl-1H-pyrazol-3-yl)amino)quinazolin-2-yl)amino)phenyl)acetonitrile) compound was finally selected as a potential candidate for the MASTL inhibitor because of the highest potency for the inhibition of MASTL. Thus, a #248 compound, named MKI-2, emerged as a new candidate compound for inhibiting MASTL (Figure 2A). The two important amino acid residues, LEU-113 and GLU-111, was predicted to interact with the pyrazole group of MKI-2 through strong hydrogen bonding. The terminal cyclopropyl moiety attached to the pyrazole group interact with hydrophobic interaction with VAL-94. These hydrogen bonding and hydrophobic interactions allowed MKI-2 to face the correct position for entry into the binding pocket of the MASTL active site. The "hinge" region of the MASTL existed between ASP-117 and LEU-163 residues, allowing other strong interactions between aminoquinazolin scaffold and ASP-117 key residue. The docking model predicted that the phenylacetonitrile moiety of MKI-2 is faced to the solvent explosion region ( Figure 2B). In vitro kinase assay using recombinant MASTL indicated that in vitro IC 50 of MKI-2 was 37.44 nM ( Figure 2C). The cellular IC 50 was 142.7 nM, as determined by a dose-response curve analysis of MKI-2 using immunofluorescence analysis for phospho-ENSA, a cellular substrate of MASTL ( Figure 2D,E), suggesting that MKI-2 is a potent MASTL inhibitor with nM range potency. We further tested whether MKI-2 inhibited other four AGC kinases, including ROCK1, AKT1, PKA Cα, and p70S6K, since AGC kinases have a similar ATP-binding pocket in the Pharmaceuticals 2021, 14, 647 3 of 12 kinase domain [26]. Notably, the in vitro kinase assay indicated that MKI-1 more selectively inhibited MASTL than the other AGC kinases ( Figure 2F). Collectively, our data indicated that MKI-2 is a potent and selective MASTL inhibitor with nM range potency. substrate of MASTL ( Figure 2D,E), suggesting that MKI-2 is a potent MASTL inhibitor with nM range potency. We further tested whether MKI-2 inhibited other four AGC kinases, including ROCK1, AKT1, PKA Cα, and p70S6K, since AGC kinases have a similar ATP-binding pocket in the kinase domain [26]. Notably, the in vitro kinase assay indicated that MKI-1 more selectively inhibited MASTL than the other AGC kinases ( Figure 2F). Collectively, our data indicated that MKI-2 is a potent and selective MASTL inhibitor with nM range potency.   The data are presented as the mean ± standard deviation of three independent experiments. ** p < 0.01. n.s.: not significant.

MKI-2 Induces Mitotic Catastrophe of Breast Cancer Cells via MASTL-PP2A
To validate that MKI-2 inhibited MASTL in breast cancer cells, we first examined whether MKI-2 inhibits phospho-ENSA in three breast cancer cell lines, including MCF7, BT549, and MDA-MB468 cells. Our data indicated that MKI-2 inhibited phosphorylated-ENSA in the three breast cancer cell lines ( Figure 3A). In addition, we examined whether

MKI-2 Induces Mitotic Catastrophe of Breast Cancer Cells via MASTL-PP2A
To validate that MKI-2 inhibited MASTL in breast cancer cells, we first examined whether MKI-2 inhibits phospho-ENSA in three breast cancer cell lines, including MCF7, BT549, and MDA-MB468 cells. Our data indicated that MKI-2 inhibited phosphorylated-ENSA in the three breast cancer cell lines ( Figure 3A). In addition, we examined whether MKI-2 increases PP2A activity and inhibits c-Myc since we previously observed that MASTL inhibition activates PP2A and decreases c-Myc stability in breast cancer cells [12,22]. Forskolin and OA were used as positive and negative controls in the PP2A activity assay, respectively. Our data indicated that MKI-2 can increase PP2A activity and decrease c-Myc proteins ( Figure 3A,B), suggesting that MKI-2 inhibits the MASTL-PP2A-c-Myc axis in breast cancer cells. Since our previous reports showed that MASTL depletion causes mitotic catastrophe of breast cancer cells [12,22], we next evaluated whether MKI-2 induces mitotic catastrophe. Notably, the microscopic analysis indicated that MKI-2 induced the mitotic arrest of MCF7 cells, which was similar to the mitotic arrest phenotype in MASTL-depleted MCF7 cells ( Figure 3C). In addition, we observed increased levels of cleaved-PARP and γ-H2AX and decreased procaspase-2 and phospho-ENSA in MKI-2-treated MCF7 cells, which was consistent with the result in the MASTL-depleted MCF7 cells ( Figure 3D). Furthermore, MKI-2 increased aberrance in nuclei of MCF7 cells ( Figure 3E), consistent with the observation that MASTL depletion or inhibition increased aberrant nuclei, such as nuclei with irregular shapes or fragmented nuclei in MCF7 cells [12,22]. Moreover, the gene set enrichment (GSEA) analysis from transcript analysis in MKI-2-treated MCF7 cells showed that MKI-2 modulated G2/M arrest, apoptosis, mitotic spindle, and DNA repair of MCF7 cells ( Figure 3F), which are the main functions of MASTL [13]. Taken together, our results suggested that MKI-2 causes mitotic catastrophe in breast cancer cells by modulating the MASTL-PP2A axis. Since our previous reports showed that MASTL depletion causes mitotic catastrophe of breast cancer cells [12,22], we next evaluated whether MKI-2 induces mitotic catastrophe. Notably, the microscopic analysis indicated that MKI-2 induced the mitotic arrest of MCF7 cells, which was similar to the mitotic arrest phenotype in MASTL-depleted MCF7 cells ( Figure 3C). In addition, we observed increased levels of cleaved-PARP and γ-H2AX and decreased procaspase-2 and phospho-ENSA in MKI-2-treated MCF7 cells, which was consistent with the result in the MASTL-depleted MCF7 cells ( Figure 3D). Furthermore, MKI-2 increased aberrance in nuclei of MCF7 cells ( Figure 3E), consistent with the observation that MASTL depletion or inhibition increased aberrant nuclei, such as nuclei with irregular shapes or fragmented nuclei in MCF7 cells [12,22]. Moreover, the gene set enrichment (GSEA) analysis from transcript analysis in MKI-2-treated MCF7 cells showed that MKI-2 modulated G2/M arrest, apoptosis, mitotic spindle, and DNA repair of MCF7 Pharmaceuticals 2021, 14, 647 5 of 12 cells ( Figure 3F), which are the main functions of MASTL [13]. Taken together, our results suggested that MKI-2 causes mitotic catastrophe in breast cancer cells by modulating the MASTL-PP2A axis.

MKI-2 Inhibits the Oncogenic Properties and Enhances the Radiosensitivity of Breast Cancer Cells
MASTL inhibition reduces the oncogenic properties and enhanced the radiosensitivity of breast cancer cells [12]. Thus, we next examined the antitumor activities of MKI-2 in breast cancer cells through various analyses, including cell viability, clonogenic, mammosphere formation, 3D culture, and invasion and migration assays. The results indicated that MKI-2 inhibited the proliferation of various breast cancer cells, including the human breast cancer MCF7, BT549, and MDA-MB468 cells and mouse breast cancer 4T1 cells with nM range efficacy (IC 50 of MKI-2 was 56~124 nM, dependent on cancer cell lines) ( Figure 4A). MKI-2 also clearly inhibited the colony and mammosphere formation of MCF7 cells with nM range efficacy ( Figure 4B-D). Furthermore, MKI-2 inhibited the migration and invasion of BT549 cells ( Figure 4E), indicating that MKI-2 can inhibit the oncogenic properties of breast cancer cells. Next, we found that MKI-2 decreased colony formation in MCF7 cells in response to irradiation ( Figure 4F), with increased cleaved-PARP and phospho-Chk2 and decreased procaspase-2 ( Figure 4G), which was more potent than AT13148 and MKI-1, suggesting that MKI-2 can enhance the radiosensitivity of breast cancer cells. Therefore, these results collectively suggested that MKI-2 has potent antitumor activities in breast cancer cells. MASTL inhibition reduces the oncogenic properties and enhanced the radiosensitivity of breast cancer cells [12]. Thus, we next examined the antitumor activities of MKI-2 in breast cancer cells through various analyses, including cell viability, clonogenic, mammosphere formation, 3D culture, and invasion and migration assays. The results indicated that MKI-2 inhibited the proliferation of various breast cancer cells, including the human breast cancer MCF7, BT549, and MDA-MB468 cells and mouse breast cancer 4T1 cells with nM range efficacy (IC50 of MKI-2 was 56 ~ 124 nM, dependent on cancer cell lines) ( Figure  4A). MKI-2 also clearly inhibited the colony and mammosphere formation of MCF7 cells with nM range efficacy ( Figure 4B-D). Furthermore, MKI-2 inhibited the migration and invasion of BT549 cells ( Figure 4E), indicating that MKI-2 can inhibit the oncogenic properties of breast cancer cells. Next, we found that MKI-2 decreased colony formation in MCF7 cells in response to irradiation ( Figure 4F), with increased cleaved-PARP and phospho-Chk2 and decreased procaspase-2 ( Figure 4G), which was more potent than AT13148 and MKI-1, suggesting that MKI-2 can enhance the radiosensitivity of breast cancer cells. Therefore, these results collectively suggested that MKI-2 has potent antitumor activities in breast cancer cells.

MKI-2 Treatment Phenocopies the MASTL-Null Oocytes
As the function of MASTL has been well characterized in oocytes [8,9,27,28], we validated the MKI-2 activity using mouse oocytes by comparing phenotypes. First, we treated immature oocytes with MKI-2 and examined the cytotoxicity. The survival rate of oocytes treated with various concentrations of MKI-2 was not significantly different among groups ( Figure 5A), suggesting that MKI-2 is not toxic up to oocytes. We then examined the effect of MKI-2 on meiotic cell cycle. Consistent with recent studies that the activation of MASTL is essential for G2-M transition [5,29], MKI-2 treatment effectively inhibited GVBD (indicative of the entry into the meiotic M phase in oocytes) in a dosedependent manner. Moreover, most oocytes remained in the GV stage even after prolonged culture ( Figure 5B,C), implying that the effect of MKI-2 is not transient in oocytes. It has been recently shown that MASTL-null oocytes reassembled a nuclear structure containing decondensed chromatin [28]. Therefore, we investigated whether MKI-2 could induce pronuclear formation in oocytes. Notably, oocytes formed distinct nuclei with decondensed chromatin after MKI-2 treatment ( Figure 5D,E). Taken together, our results reveal that treating oocytes with MKI-2 displayed similar phenotypes, observed as MASTL-depleted or -null oocytes, corroborating that MKI-2 has the capacity to inhibit MASTL.

MKI-2 Treatment Phenocopies the MASTL-null Oocytes
As the function of MASTL has been well characterized in oocytes [8,9,27,28], we validated the MKI-2 activity using mouse oocytes by comparing phenotypes. First, we treated immature oocytes with MKI-2 and examined the cytotoxicity. The survival rate of oocytes treated with various concentrations of MKI-2 was not significantly different among groups ( Figure 5A), suggesting that MKI-2 is not toxic up to oocytes. We then examined the effect of MKI-2 on meiotic cell cycle. Consistent with recent studies that the activation of MASTL is essential for G2-M transition [5,29], MKI-2 treatment effectively inhibited GVBD (indicative of the entry into the meiotic M phase in oocytes) in a dosedependent manner. Moreover, most oocytes remained in the GV stage even after prolonged culture ( Figure 5B, C), implying that the effect of MKI-2 is not transient in oocytes. It has been recently shown that MASTL-null oocytes reassembled a nuclear structure containing decondensed chromatin [28]. Therefore, we investigated whether MKI-2 could induce pronuclear formation in oocytes. Notably, oocytes formed distinct nuclei with decondensed chromatin after MKI-2 treatment ( Figure 5D,E). Taken together, our results reveal that treating oocytes with MKI-2 displayed similar phenotypes, observed as MASTLdepleted or -null oocytes, corroborating that MKI-2 has the capacity to inhibit MASTL.

Discussion
Although recent studies suggest that MASTL is a promising target for selective anticancer treatment [16], MASTL inhibitors with nM range potency and antitumor efficacy have not been reported. Our previous report showed that MKI-1 was an MASTL inhibitor with antitumor activities in breast cancer cells. However, it showed relatively low potency for MASTL inhibition with IC 50 of 9.9 µM. In the current study, our data reported the second-generation MASTL inhibitor MKI-2, a new potent MASTL inhibitor with in vitro IC 50 of 37.44 nM and cellular IC 50 of 142.7 nM in breast cancer cells. In addition, MKI-2 did not modulate other AGC kinases. Furthermore, we found that MKI-2 had potent antitumor activities by disrupting mitosis via the MASTL-PP2A-Myc regulation in breast cancer cells. Moreover, MKI-2 treatment inhibited GVBD and formed distinct nuclei with condensed chromatin in mouse oocytes, which phenotypes were a clear phenocopy with MASTL-null oocytes [28]. Therefore, this study presents a new potent and selective MASTL inhibitor MKI-2 targeting MASTL-PP2A in breast cancer cells.
Two compounds, GKI-1 and MKI-1, were recently reported as MASTL inhibitors [22,23]. GKI-1 inhibited MASTL with in vitro IC 50 of 10 µM and reduced phosphorylated ENSA with mitotic disruption in HeLa cells [23]. However, it showed no anticancer activities in breast cancer cells [22]. Our previous report also showed that MKI-1 inhibited MASTL with in vitro IC 50 of 9.9 µM and reduced phosphorylated ENSA with the induction of aberrant nuclear cells in breast cancer cells [22]. Compared with GKI-1, MKI-1 showed anticancer activities in in vitro and in vivo breast cancer models [22]. However, the in vitro IC 50 of both GKI-1 and MKI-1 for the inhibition of MASTL was approximately 10-fold less than that of the multiple AGC kinase inhibitor AT13148 [22], indicating that both compounds are ineffective as MASTL inhibitors. Compared with GKI-1 and MKI-1, MKI-2 showed the superior in vitro inhibition of MASTL (37.44 nM of in vitro IC 50 and 142.7 nM of cellular IC 50 ). In addition, MKI-2 can induce the mitotic arrest phenotype of breast cancer cells, which were identical to the phenotype of MASTL-depleted breast cancer cells. Moreover, MKI-2 treatment showed the phenocopy of MASTL-null oocytes in mouse oocytes. Furthermore, MKI-2 did not modulate other AGC kinases, including ROCK1, AKT, p70S6K, and PKA Cα. Therefore, our data suggested that MKI-2 is a new, potent, and selective MASTL inhibitor.
Our data showed that MKI-2 had superior anticancer activities with nM range efficacy in breast cancer cells. MKI-2 showed approximately 50-fold more potent anticancer activity than that of AT13148 in clonogenic and 3D culture analysis of breast cancer cells. In addition, MKI-2 can also enhance the radiosensitivity of breast cancer cells with nM range efficacy, indicating the superior anticancer activity of MKI-2. In addition, we also observed that MKI-2 disrupted mitosis of breast cancer cells by modulating MASTL-PP2A-c-Myc regulation, a mechanism which is consistent with the results in the MASTL-depleted cells [12,22]. Thus, our data suggested that MKI-2 demonstrates anticancer activity in breast cancer cells by regulating MASTL-PP2A during mitosis.

In Silico Screening
In silico screening from the chemical library that included 65,000 bio-active compounds (Korea Chemical Bank of Korea Research Institute of Chemical Technology (http://www. chembank.org (accessed on 14 November 2021); Daejeon, Korea) was performed as previously described [22].

Immunofluorescence
Immunofluorescence analysis was performed as previously described [30,31]. Briefly, the cells were fixed with 4% paraformaldehyde, permeabilized, and blocked with 0.1 Triton X-100 and 5% fetal calf serum in PBS. The fixed cells were consecutively incubated with primary antibodies against phospho-ENSA (Ser67)/ARPP19 (Ser62) (Cell Signaling Technology; 1:100), and secondary antibodies, such as anti-mouse Alexa-488 and anti-rabbit Alexa-594 (Molecular Probes, Eugene, OR, USA; 1:200). The slides were mounted in DAPIcontaining medium and images of the bands were then obtained by using IN Cell analyzer 2000 (GE Healthcare, Chicago, IL, USA).

HTRF Assay
The recombinant GST-tagged-MASTL (Thermo Fisher Scientific) was used for a kinase assay based on HTRF. An HTRF KinEASE assay kit (CisBio, Bedford, MA, USA) was used according to the manufacturer's instructions.

In Vitro Kinase Assay
Two recombinant GST-tagged-MASTL (Thermo Fisher Scientific) and His-tagged-ENSA (Sino Biological Inc., Beijing, China) were employed for the in vitro kinase assays. The in vitro kinase assay was performed as previously described [22].

Cell Viability Assays
Breast cancer cell lines, including MCF7, BT549, 4T1, and MDA-MB-468 cells, were treated with serial dilutions of MKI-2 from 1 µM to 15.63 nM for 72 h. Cell viability was measured by WST-8 assay (Cyto X TM cell viability assay kit; LPS solution, Daejeon, Korea) according to the manufacturer's instructions.

Clonogenic and Sphere Formation Assay
A clonogenic assay [32] and sphere formation assay [12,33] were performed as previously described.

Three-Dimensional Culture
A 3D culture was performed as recommended by the manufacturer (TheWell Bioscience, North Brunswick Township, NJ, USA).

Invasion and Migration Assay
The invasion assay was performed using a Transwell chamber (Corning Inc.). The inserting chamber was coated with 0.5 mg/mL Matrigel for 2 h at 37 • C, and BT549 cells (5 × 10 4 ) were treated with DMSO or 250 nM MKI-2. The lower chamber was filled with RPMI containing 10% FBS, and the invaded cells were analyzed after 24 h. The migration assay was performed as described for the invasion assay without Matrigel coating.

PP2A Activity Assay
The PP2A activity assay was performed as previously described [30], and the PP2A phosphatase assay according to the manufacturer's protocol (RediPlate 96 EnzChek serine/threonine phosphatase assay kit; Invitrogen). Okadaic acid (OA) (50 nM; Sigma-Aldrich) or forskolin (40 µM; Sigma-Aldrich) were used for the inhibition or activation of PP2A as negative and positive controls, respectively.

GSEA Analysis
MCF7 cells were treated with DMSO or 250 nM MKI-2 for 24 h. RNA-sequencing was performed at e-Biogen (Seoul, Korea), and the data were analyzed by GSEA software [34].

Oocyte Culture and Treatment
All procedures for mouse care and use were conducted in accordance with the guidelines and approved by the Institutional Animal Care and Use Committees of Sungkyunkwan University. First, 3-4-week-old CD1 female mice were purchased from a local company (Koatech, Pyeongtaek, Korea). To collect fully grown germinal vesicle (GV) stage oocytes, female mice were injected with 5 IU of a pregnant mere's serum gonadotropin (PMSG). After 48 h, oocytes were collected in M2 medium (M7167) fortified with 0.4% bovine serum albumin (BSA; A7906) and 100 µM of 3-isobuthly-1-methylxanthine (IBMX) to prevent meiotic maturation. During in vitro maturation, oocytes were cultured in IBMX-free M2 medium covered with mineral oil (M5310) at 37 • C in a 5% CO2 atmosphere for 16 h. To collect ovulated oocytes, female mice were induced by intraperitoneal injections of 5 IU of a pregnant mare's serum gonadotrophin, followed by 5 IU human chorionic gonadotrophin (hCG) after 48-52 h. Mature oocytes were collected 15 h later from the oviduct into the M2 medium and briefly incubated in 300 µg/mL hyaluronidase (H4272) to remove cumulus cells. For chemical treatment, the indicated concentration of chemicals or an equivalent amount of DMSO was added to the culture medium.

Statistical Analysis
The two-tailed Student's t-test was performed to analyze statistical differences between groups. p-values of less than 0.05 were considered statistically significant. Excel was employed for statistical analyses.

Conclusions
Although MASTL is an attractive target for anticancer treatment, MASTL inhibitors with nM range potency and anticancer efficacy have not been reported. The present study newly identified and characterized a second-generation MASTL inhibitor MKI-2 with nM range potency and anticancer efficacy. In addition, we demonstrated the mode of action of MKI-2 targeting of MASTL-PP2A regulation in breast cancer cells. Thus, our data reports a novel MASTL inhibitor for development as a small molecule inhibitor for therapy in MASTL-overexpressed cancers, such as breast cancer.