A Novel Compound ARN-3236 Inhibits Salt-Inducible Kinase 2 and Sensitizes Ovarian Cancer Cell Lines and Xenografts to Paclitaxel

This article is featured in Highlights of This Issue, p. 1877

Ovarian cancer is one of the most lethal malignancies. The survival of ovarian cancer patients depends critically on response to platinum- and taxane-based chemotherapy. Whereas 70% of ovarian cancers respond to platinum derivatives, only approximately 40% of patients undergo objective regression following treatment with paclitaxel (1). Improved outcomes might be attained if sensitivity to paclitaxel was enhanced. Several attempts have been made to reverse mechanisms of acquired taxane resistance, but less attention has been given to enhancing sensitivity to taxanes during primary treatment of ovarian cancer, especially high-grade serous ovarian cancer (2–5).

In earlier studies, we had undertaken an siRNA screen to identify kinases that regulate sensitivity to paclitaxel (6). One of the most promising candidates to emerge from that screen was salt-inducible kinase 2 (SIK2), a serine/threonine kinase, which is required for bipolar mitotic spindle formation and normal mitotic progression (7). Consequently, SIK2 presents an attractive therapeutic target for ovarian cancer treatment (7).

Here, we report that a novel 1H-(pyrazol-4-yl)-1H-pyrrolo[2,3-b]pyridine small-molecule inhibitor of SIK2 (ARN-3236) [16] inhibits ovarian cancer cell growth and sensitizes ovarian cancer cells and xenografts to paclitaxel by inhibiting centrosome splitting and AKT/survivin signaling.

A formalin-fixed, paraffin-embedded tissue microarray (TMA) that contained samples of 193 primary serous ovarian cancers (183 cases of high grade and 10 cases of low grade) was obtained from the MD Anderson Pathology Department (MD Anderson Cancer Center, Houston, TX). Additional details are provided in Supplementary Table S1. The protocols for handling paraffin-embedded ovarian cancer specimens and analyzing patient data were approved by the ethical committees of the MDACC Institutional Review Boards. TMA construction was performed as described previously (8).

Cells were seeded in 96-well cell culture plates in triplicate and incubated for 16 hours. Then, cells were treated with DMSO or ARN-3236 for 24 hours, followed by an additional 72-hour incubation with paclitaxel at the concentration indicated. The sulforhodamine B (SRB) assay was used to measure the growth inhibition of each cell line with and without treatment as described previously (9). The concentration producing 50% growth inhibition (IC₅₀) was calculated by the following formula: 100 × (T − T₀)/(C − T₀) = TC₅₀, in which T is the optical density (OD) value after drug treatment, T₀ is the OD value at time 0, and C is the OD value for the diluent treatment (10). GraphPad Prism 5 software was used to generate growth curves. For studies of ARN-3236 and paclitaxel in combination, four groups were evaluated: (i) drug-free control; (ii) ARN-3236 alone; (iii) paclitaxel alone; and (iv) a combination of ARN-3236 and paclitaxel at 9 different concentrations (0, 0.2, 0.39, 0.78, 1.56, 3.13, 6.25, 12.5, and 25 μmol/L) of paclitaxel were used. To evaluate additive or synergistic interactions, a combination index (CI value) was calculated with CalcuSyn software (Biosoft), which was developed on the basis of the median effect principle of Chou–Talalay method (11). Values less than 1 were considered synergistic and those equal to 1 were additive.

HEY and A2780 human ovarian cancer cell lines were purchased from the ATCC. UPN251, OVCAR3, OVCAR5, OVCAR8, ES-2, OC316, SKOv3, and IGROV1 were kindly provided by Dr. Gordon Mills' laboratory (12–15), and all the cell lines were confirmed with STR DNA fingerprinting, which was performed by the MDACC Characterized Cell Line Core (supported by NCI CCSG # CA016672). SKOv3-SIK2 cell line was kindly provided by Dr. Ahmed's laboratory (16). SKOv3 and SKOv3-SIK2 cells were cultured in McCoy's 5A medium; OVCAR3, OVCAR5, OVCAR8, HEY, OC316, A2780, IGROV1, ES-2, and UPN251 cells were cultured in RPMI1640 medium. All media were obtained from the Media Preparation Core Facility at MD Anderson Cancer Center. All cell lines were tested for mycoplasma with a MycoSensor PCR Assay Kit from Stratagene (cat. 302109) and found to be free from contamination.

The siGenome siRNAs (Dharmacon) were used to target individual genes for inhibiting their expression as described previously (7). Cells were transfected with a negative control siRNA or a specific siRNA using DharmaFECT 4 reagent (Dharmacon). A mixture of siRNA (25 nmol/L final concentration) and DharmaFECT 4 reagent was incubated for 20 minutes at room temperature before being applied to the cells.

The distance from the nucleus to the centrosome was measured as described previously (7), and details regarding staining and measurement are provided in Supplementary Methods.

After treatment, cells were fixed and stained using anti-γ-tubulin and anti-phospho-histone H3 antibodies. DAPI was used for nuclear staining. The percentage of mitotic cells was calculated using the following formula: (phosphor-histone H3–positive cells/DAPI-positive cells) × 100%. Mitotic cell polarity was determined as described previously (7). Additional detailed information is provided in Supplementary Methods.

SKOv3 cells were treated with ARN-3236 (2 μmol/L) or DMSO diluent for 48 hours. Reverse-phase protein array (RPPA) analysis was performed at the MD Anderson Cancer Center RPPA Core Facility. Additional information is provided in the Supplementary Methods.

Female athymic nu/nu mice were used to determine the anticancer effect of ARN-3236. Experiments using Nu/Nu mice were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC ID: 00001052-RN00; MD Anderson Cancer Center, Houston, TX). SKOv3ip1 and OVCAR8 cells were injected intraperitoneally. ARN-3236 was administrated orally at a dose of 60 mg/kg/day for 7 days. Paclitaxel was administrated intraperitoneally once a week at a dose of 5 mg/kg in SKOv3ip cells and 1 mg/kg in OVCAR8 cells. On day 7 after cancer cell injection, the mice were randomly assigned to the following treatment groups (n = 6 mice per group): (i) nontreatment control; (ii) vehicle control for ARN-3236; (iii) ARN-3236; (iv) paclitaxel; and (v) combination of paclitaxel and ARN-3236. SKOv3ip-bearing mice were treated for 3 weeks, and OVCAR8-bearing mice were treated for 4 weeks. After treatment, mice were sacrificed by CO₂ at the end of the studies. All tumors were collected immediately after death and weighed.

Data are expressed as mean ± SD. We used SPSS23.0 Software to calculate statistical significance between groups. GraphPad Prism 6.0 was used for multiple comparison tests. P < 0.05 was considered as statistically significant.

More details of experiments are listed in Supplementary Methods.

To determine SIK2 expression in epithelial ovarian cancer, a cohort of 193 serous ovarian cancers and 60 normal ovarian tissues and normal fallopian tube specimens were analyzed for SIK2 protein expression by IHC. Positive SIK2 staining was observed in about 30% of ovarian cancers in comparison with negative staining in normal ovarian surface epithelial cells and normal fallopian epithelial cells (Fig. 1A). In addition, SIK2 protein was elevated in five of 12 high-grade serous ovarian cancer tumor cell lysates compared with normal ovarian epithelium cell lysates (Fig. 1B). Thus, SIK2 is strongly expressed in 30% of human ovarian cancer and may serve as a potential therapeutic target.

Previously, we have shown that SIK2 plays a key role in the progression of mitosis and in regulating paclitaxel sensitivity by knocking down SIK2 protein with siRNA (7). Given the challenges of delivering siRNA to cancer cells in vivo, we had sought orally available small-molecule inhibitors of SIK2 to enhance the chemosensitivity of ovarian cancers. Arrien Pharmaceuticals has developed a novel 1H-(pyrazol-4-yl)-1H-pyrrolo[2,3-b]pyridine inhibitor, ARN-3236 (17), which competes with ATP binding to SIK2 protein and inhibits SIK2 kinase activity (18, 19). Previous publications have shown that ARN-3236 inhibits SIK2 activity with an IC₅₀ <1 nmol/L but does not significantly inhibit the other two SIK family members SIK1 and SIK3 as well as other AMP-activated protein kinase (AMPK) family members (19). We tested the inhibitory effect of ARN-3236 using TORC2 (transducer of regulated CREB protein 2)-derived peptide, a direct substrate of SIK2 with a phosphorylation consensus sequence LX[HKR]S/TXSXXXL (aa166-175; ref. 20) and demonstrated that ARN-3236 diminished TORC2 phosphorylation in a dose-dependent manner (Supplementary Fig. S1A). To examine the pharmacodynamics of ARN-3236, we studied S358 autophosphorylation of SIK2, which has been correlated with SIK2 activity (16). We confirmed that ARN-3236 blocked autophosphorylation of SIK2 and deceased SIK2 activity in ovarian cancer cells in a dose-dependent and a time-dependent manner (Fig. 2A). Thus, ARN-3236 is a potent and specific SIK2 kinase inhibitor.

To determine whether ARN-3236 could inhibit growth of ovarian cancer cells, the effect of ARN-3236 was measured in 10 ovarian cancer cell lines using short-term cell proliferation assays, and the IC₅₀ of ARN-3236 was calculated in each cell line. Significant inhibition was achieved in all cell lines. The IC₅₀ values of ES2, SKOv3, OVCAR3, A2780, HEY, OC316, IGROV1, UPN251, OVCAR8, and OVCAR5 cells were 1.22, 1.23, 2.58, 0.93, 0.8,1.63, 2.51, 2.42, 1.56, and 1.19 μmol/L, respectively. Thus, the IC₅₀ ranged from 0.8 to 2.6 μmol/L (Fig. 2B and Supplementary Fig. S1B–S1J). Endogenous SIK2 protein expression correlated inversely with the IC₅₀ value of ARN-3236 in these cancer cell lines (Pearson r = −0.642, P = 0.03; Fig. 2C and D), indicating the cells with relatively high expression of SIK2 protein are more sensitive to treatment with the SIK2 inhibitor. The ability of ARN-3236 to sensitize ovarian cancer cells to paclitaxel was documented in eight ovarian cancer cell lines, including OC316, OVCAR8, SKOv3, A2780, OVCAR5, HEY, ES2, and UPN251, where treatment with ARN-3236 produced a >1.5-fold change in IC₅₀ (Fig. 3A and B and Supplementary Fig. S2A–S2F). Sensitization to paclitaxel was not observed in OVCAR3 and IGROV1 (Supplementary Fig. S2G and S2H) that exhibited relatively low expression level of SIK2 protein. In addition, ARN-3236 and paclitaxel combination study in SKOv3, OC126, and OVCAR8 cell lines further confirmed that treatment of ARN-3236 with paclitaxel provide synergistic effect (Supplementary Fig. S2J). Thus, ARN-3236 increased sensitivity to paclitaxel in cell lines with relatively high expression of SIK2.

The antitumor activity of ARN-3236 was measured using ovarian cancer xenograft model in nu/nu mice (Fig. 3C). In SKOv3ip xenograft model, ARN-3236 plus paclitaxel produced greater inhibition of tumor growth than any other group and showed statistically significant difference compared with paclitaxel alone (P = 0.028; Fig. 3D). A combination of ARN-3236 with paclitaxel produced a similar trend in growth inhibition in a xenograft model using OVCAR8 high-grade serous ovarian cancer cells, where significant inhibition (P = 0.02) was observed only with the combined treatment (Supplementary Fig. S2I). Together, these data confirm that ARN-3236 inhibits SIK2 activity and enhances paclitaxel sensitivity not only in cultured cells but also in ovarian cancer xenografts, suggesting the potential of ARN-3236 in improving chemotherapy efficiency in ovarian cancers.

The centrosome, located close to the nuclear envelope during interphase, is an organelle that serves as the microtubule-organizing center for cells as well as a regulator of cell-cycle progression. SIK2 knockdown resulted in nuclear–centrosome uncoupling (NCU; nuclear centrosome distance of >2 μm; ref. 7). To test the effect of ARN-3236 on centrosome uncoupling, nuclear centrosome distance was measured in SKOv3 and A2780 cells after treatment of ARN-3236 for 48 hours. ARN-3236 treatment increased the nuclear centrosome distance from 1.790 to 3.198 μm in SKOv3 cells and from 1.410 to 2.230 μm in A2780 cells (P < 0.0001; Fig. 4A and B and Supplementary Fig. S3A), suggesting that the SIK2 inhibitor abolished this centrosome function. In addition, a significant increase in the percentage of NCU-positive cells was also detected after SIK2 inhibitor treatment compared with control (30.0%–15.2% in SKOv3 cells, 37.6%–18.7% in A2780 cells, respectively, P < 0.01; Fig. 4B and Supplementary Fig. S3A). Thus, ARN-3236 interfered with the appropriate centrosome positioning in interphase.

Mitosis is a continuous process that passes through different stages: prophase, prometaphase, metaphase, anaphase, and telophase. In early mitosis, separation of centrosomes is driven by several motor proteins with the onset of prophase to ensure spindle bipolarity (21, 22). Previously, we reported that SIK2 regulates centrosome splitting, and SIK2 knockdown resulted in centrosome splitting failure and mitotic catastrophe (7, 23). To determine whether ARN-3236 blocks centrosome function in mitotic cells, we calculated the percentage of each phase in mitotic cells after ARN-3236 treatment. We have shown that ARN-3236 induced accumulation of cells in prometaphase (Fig. 4C and D and Supplementary Fig. S3B and S3C) and increased the ratio of prometaphase to metaphase cells in mitosis, suggesting that ARN-3236 induces prometaphase arrest in ovarian cancer cells. Importantly, ARN-3236 decreased the percentage of bipolar mitotic cells and increased the percentage of monopolar/multipolar mitotic cells (Fig. 4E and F). These data indicate that ARN-3236 inhibits centrosome splitting in mitotic ovarian cancer cells.

To confirm that the failure of centrosome splitting and the delay in mitotic progression induced by ARN-3236 can result in cell-cycle arrest and apoptosis, we first tested the effect of ARN-3236 on cell-cycle progression in ovarian cancer cells. Treatment with ARN-3236 induced G₂–M arrest (Fig. 5A and Supplementary Fig. S4A). Furthermore, measurement of apoptosis by flow cytometry indicated that ARN-3236 induced apoptosis both alone and in combination with paclitaxel (Fig. 5B and Supplementary Fig. S4B). Previously, we have shown that SIK2 depletion increases the cell ploidy index (7). In our current study, ARN-3236 treatment resulted in the accumulation of tetraploid cells (4n and 8n; Fig. 5C), and sequential treatment of ARN-3236 and paclitaxel led to further accumulation of tetraploid cells (4n and 8n) compared with treatment of paclitaxel or ARN-3236 alone (Fig. 5D). Tetraploid cells are often associated with failing mitosis, which leads to the gross nuclear alterations that constitute the most prominent morphologic traits of mitotic catastrophe. Mitotic catastrophe also has been extensively used as a term to indicate cell death resulting from aberrant mitosis (24, 25). Altogether, the results suggest that ARN-3236 and/or paclitaxel treatment may inhibit ovarian cancer growth and induce cell death through multiple mechanisms, including G₂–M arrest, apoptosis, and mitotic catastrophe.

To identify potential mechanisms underlying regulation of apoptosis by SIK2 inhibition, RPPAs were used to monitor changes in signaling pathways. We found that ARN-3236 significantly inhibited AKT activation (Supplementary Fig. S5A), which is consistent with the results obtained when SIK2 was silenced with siRNA (7). To validate results obtained with RPPA analysis, positive hits were confirmed by Western blot analysis. ARN-3236 inhibits AKT activation by reducing phosphorylation of serine 473 as well as threonine 308 (Fig. 6A and Supplementary Fig. S5B). Furthermore, ARN-3236 also reduces the expression of survivin (Fig. 6A and Supplementary Fig. S5B), a downstream target of AKT, which contributes to paclitaxel resistance in many cancers (26, 27), suggesting that SIK2 regulates ovarian cancer cell growth and paclitaxel resistance, at least in part, through AKT/survivin signaling. Importantly, the ARN-3236–mediated changes in AKT phosphorylation and survivin observed in cultured cells were also confirmed in ovarian cancer xenografts in vivo (Fig. 6A and Supplementary Fig. S5C). In addition, whether survivin overrides the proapoptotic function of ARN-3236 and paclitaxel was investigated in the cells engineered to overexpress survivin (Fig. 6B). As shown in Fig. 6C and D, ARN-3236 moderately but significantly sensitized OC316-control cells to paclitaxel (paclitaxel IC₅₀: 4.732–5.248 nmol/L in DMSO, 3.873–4.252 nmol/L in ARN-3236, IC₅₀ fold change 1.22, P < 0.01) but did not sensitize an OC316 subline with forced expression of survivin (paclitaxel IC₅₀: 4.737–5.254 nmol/L in DMSO, 4.613–5.5230 nmol/L in ARN-3236, P = 0.73). OC316-survivin cells also exhibited attenuated activation of caspase-3 after ARN-3236 and paclitaxel treatment (Fig. 6E). Thus, downregulation of survivin expression and activation of apoptosis is at least partially responsible for ARN-3236–mediated paclitaxel sensitization.

SIK2 is a member of a family of AMPKs, which was originally cloned from the adrenal gland of a salt-treated rat (28). SIK2 has been shown to function in diverse biological processes, including gluconeogenesis, neuronal survival, melanogenesis, hepatic steatosis, and centrosome splitting (7, 29–32). SIK2 is also implicated in the progression of cancer (7, 33–35). In addition, a fraction of ovarian cancers depends on the SIK2 kinase for maintenance of cell proliferation and chemoresistance (7). Some 30% of ovarian cancers overexpress SIK2, associated with a trend toward poor prognosis. Selective inhibition of SIK2 kinase is a potentially important approach for the treatment of ovarian cancer.

ARN-3236 is an orally bioavailable SIK2 small-molecule inhibitor. ARN-3236 inhibits SIK2 activity with subnanomolar potency but does not significantly inhibit the other two SIK family members SIK1 and SIK3 (19). To our knowledge, ARN-3236 is the first and most potent SIK2 inhibitor reported to date. ARN-3236 inhibited the growth of 10 ovarian cancer cell lines with IC₅₀ values of 0.8 to 2.6 μmol/L. Importantly, the IC₅₀ of ARN-3236 was inversely correlated with endogenous SIK2 expression across 10 different cell lines. Consequently, we have identified a potential biomarker for susceptibility to SIK2 inhibition with this compound. In our current study, primary sensitivity to paclitaxel has been enhanced by treatment of ARN-3236. As less than half of ovarian cancer patients exhibit primary sensitivity to paclitaxel, there is considerable room for improvement in primary treatment. ARN-3236 and its congeners could fill this unmet medical need. Inhibition of growth and enhancement of paclitaxel sensitivity was observed in 8 of 10 ovarian cancer cell lines (36), suggesting that the combination of ARN-3236 and paclitaxel might improve efficacy in a significant fraction of patients with this disease. In addition, our data suggest that inhibition of SIK2 enzyme activity can mimic the effects of siRNA-mediated knockdown of SIK2 protein with regards to cell-cycle changes, induction of polyploidy, uncoupling of centrosomes from the nucleus, centrosome splitting, growth inhibition, and enhanced sensitivity to paclitaxel, associated with decreased signaling through AKT.

Regulators of mitosis have been successfully targeted by ARN-3236 to enhance response to taxane chemotherapy. Centrosomes are complex cellular substructures that are dynamically regulated by a series of biochemical and morphologic changes that parallel the progression of cell cycle. In addition to organizing cytoplasmic microtubule arrays during interphase, centrosomes direct formation and positioning of bipolar mitotic spindle and thereby enable equal partitioning of the replicated genome, and impaired centrosome function may lead to chromosome missegregation and accumulation of tetraploid cells. Tetraploid cells are often associated with failing mitosis and mitotic catastrophe, which has been extensively used as a term to indicate cell death resulting from aberrant mitosis (37). Because the functional significance of centrosome has been described in a variety of solid tumors, including breast, prostate, colon, ovarian, and pancreatic cancer (37, 38), centrosome has become an important focus for cancer therapy development (39, 40). Here, we show that ARN-3236 with the known SIK2 inhibition disturbed SIK2-dependent centrosome positioning in interphase and blocked centrosome separation in mitosis, sensitizing ovarian cancers to paclitaxel in culture and in xenografts. We believe these data further confirmed that ARN-3236 has potential for enhancing primary treatment of ovarian cancers.

H. Vankayalapati reports receiving other commercial research support from and holds ownership interest (including patents) in Arrien Pharmaceuticals.

No potential conflicts of interest were disclosed by the other authors.

Conception and design: J. Zhou, A. Alfraidi, H. Vankayalapati, Z. Lu, R.C. Bast

Development of methodology: J. Zhou, J.M. Santiago-O'Farrill, A. Alsaadi, A.A. Ahmed, W. Mao, Y. Wang, H. Vankayalapati

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Zhou, S. Zhang, J.M. Santiago-O'Farrill, A. Alsaadi, H. Yang, J. Liu, W. Mao, Y. Wang, H. Vankayalapati, R.C. Bast

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J. Zhou, A. Alfraidi, S. Zhang, J.M. Santiago-O'Farrill, A.A. Ahmed, H. Yang, W. Mao, Y. Wang, H. Vankayalapati, Z. Lu, R.C. Bast

Writing, review, and/or revision of the manuscript: J. Zhou, A. Alfraidi, H. Vankayalapati, Z. Lu, R.C. Bast

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Alfraidi, S. Zhang, H. Takemori, H. Vankayalapati

Study supervision: A. Alfraidi, A.A. Ahmed, J. Liu, H. Vankayalapati, Z. Lu, R.C. Bast

Other (ARN-3236 inhibitor synthesis): V. Krishna Yerramreddy Reddy

This work was supported by the Cancer Prevention and Research Institute of Texas (RP110-595), the NCI at the NIH (P50 CA83639, U54 CA151668, R01 CA135354, P30 CA16672, and UH2 TR000943-01), the Zarrow Foundation, the Mossy Foundation, the National Foundation for Cancer Research, the RGK Foundation, and the generous support of Stuart and Gay Lynne Zarrow.

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.