Serum- and Glucocorticoid-inducible Kinase 1 is Essential for Osteoclastogenesis and Promotes Breast Cancer Bone Metastasis

Bone is a common target of metastatic spread in solid tumors. Bone metastasis is generally incurable in patients because of limited therapeutic options (1, 2). Bone metastasis can be regarded as a result of the crosstalk between bone resorption and tumor growth (3). During the process of bone metastasis, cancer cells undergo epithelial–mesenchymal transition (EMT), get transported as circulating tumor cells (CTC), and home to target sites (4). Presently, bisphosphonates and antibodies against receptor activator of NF-κB ligand (RANKL) are the standard of care in patients with bone metastases (5–7).

Once cancer cells infiltrate bones, the metastatic cells encounter native bone microenvironmental cells. During the formation of bone metastatic lesions, various osteolytic factors aiming at osteoclasts are secreted by tumor cells. Among these factors, macrophage CSF (M-CSF) and RANKL play a central role in this multistep process (8–10). M-CSF upregulates RANK expression on osteoclast precursors and promotes osteoclast differentiation, whereas RANKL binds to RANK and cause the activation of nuclear transcriptional regulators (11). The recruitment of variable adaptor molecules then activates downstream signaling pathways, includes the NF-κB and PI3K/AKT pathways, thereby initiating osteoclast formation and activity (12–15).

In the terminal of osteoclasts differentiation, Ca²⁺ channels also involved. Hwang and colleagues (16) have shown that Orai1-mediated calcium entry is a major inducer and activator of nuclear factor of activated T cells c1 (NFATc1)-mediated RANKL signaling. RANKL evokes Ca²⁺ oscillations, leading to a selective and stable induction of NFATc1, a Ca²⁺-responsive transcription factor that drives osteoclast formation (16, 17).

Serum- and glucocorticoid-inducible kinase 1 (SGK1) belongs to a subfamily of serine/threonine kinases that is transcriptionally stimulated by serum and glucocorticoids (18). SGK1 expression is induced by insulin and growth factors via PI3K, 3-phosphoinositide-dependent kinase 1 (PDK1), or mTOR complex-2 (19–21). In mammals, SGK1 is expressed in all examined tissues and regulates a wide variety of cellular molecules, including signaling factors and ion channels (22–24). Pietro and colleagues (25) showed that SGK1 expressed in white adipose tissues promotes adipogenesis by regulating Foxo1 phosphorylation. In megakaryocytes, as a novel regulator of platelet function, SGK1 upregulates Orai1 through the NF-κB pathway and increases store-operated Ca²⁺ entry (SOCE) in platelets (26). However, the role of SGK1 in osteoclast formation remains unknown.

In this study, we demonstrated that SGK1 triggers osteoclastogenesis through Orai1, and the inhibitor of SGK1 can reduce bone metastasis in vitro. Thereby SGK1 is a promising therapeutic target for bone metastasis.

GSK650394 and BTP2 was purchased from Sigma-Aldrich. Soluble recombinant human M-CSF and mouse RANKL were obtained from PeproTech.

C57BL/6 mice and BALB/c nu/nu mice were purchased from the Experimental Animal Center of Tongji Medical College (Wuhan, China). The mice were housed in animal care facility of Tongji Medical College at 25°C with 12-hour light/dark cycle and free access to standard chow and water. All animal studies were approved by the Institutional Animal Research Committee of Tongji Medical College. All animal experiments were carried out in accordance with the protocols approved by the Institutional Animal Care and Use Committee.

MDA-MB-231 cells were resuspended in sterile PBS at a concentration of 5 × 10⁵ cells per 100 μL. We used a marked location midway between the sternal notch and the top of xiphoid process, and slightly left (anatomic) of the sternum. Four-week-old male (BALB/c nu/nu) mice were used for these experiments. A successful intracardiac injection was determined on the basis of the systemic distribution of bioluminescence throughout the animal at 12 hours postinjection. Only mice demonstrating evidence of a successful injection were used for further analyses. Four weeks after inoculation, bone metastatic cells in hind limbs were isolated and cultured in DMEM supplemented with 10% FBS, penicillin (100 U/mL), and streptomycin (100 μg/mL). Cultured bone metastatic cells were reinoculated into the left ventricle of the heart. This procedure was repeated four times until no metastases were detected in organs other than bones. These bone metastatic cells in hind limbs, designated MBS4-TJGK cells, were isolated and cultured in DMEM.

MDA-MB-231 and RAW264.7 cell lines were purchased from the Shanghai Institute of Life Sciences Cell Resource Center, Shanghai, China. Cell lines authentication were conducted by China Center for Type Culture Collection (CCTCC). STR data were compared with ATCC and DSMZ, all the locations were exactly matched. BMMs were obtained from femoral and tibial bone marrow of 6- to 8-week-old C57BL/6 mice as described earlier (27). All cell cultures were regularly checked for mycoplasma contamination. BMMs and RAW264.7 were cultured and induced to osteoclast as described previously (15). Scrambled, SGK1-knockdown and Orai1-overexpression shRNA were obtained from ViGene Biosciences, Inc.

Antibodies against the following targets were purchased from Cell Signaling Technology: IκB kinase α (IKKα), p-IKKα, IκBα, p-IκBα, p65, p-p65, AKT, and phosphorylated AKT (p-AKT). Antibodies against c-Fos, NFATc1, and SGK1 was purchased from Abcam. Rabbit antibodies against the PI3K catalytic subunit p110 and Orai1, the TRAP staining kit and all other reagents were purchased from Sigma-Aldrich. Mouse anti-β-actin and secondary antibodies were obtained from Boster. Protein bands were visualized using a ChemiDoc XRS+ System with Image Lab Software (Bio-Rad).

For IF analysis of cultured cells, BMMs were fixed with 4% paraformaldehyde for 15 minutes and then blocked with 5% bovine serum albumin and 0.1% Triton X-100 in PBS for 30 minutes at room temperature. Immunostaining was performed using the appropriate primary and secondary antibodies. Nuclei were counterstained with DAPI. Images were captured using a fluorescence microscope.

Total RNA was extracted from cultured cells and cDNA was synthesized with the Easy Script First-Strand cDNA Synthesis Super Mix Kit (TransGen Biotech). Real-time PCR was performed using Power SYBR Green PCR Master Mix (TransGen Biotech). The specific primers used for PCR amplification were as follows (F, forward; R, reverse): SGK1, F 5′-TGTGAAGTCCCTTCTGTGGA-3′ and R 5′-CCATCTTCGTACCCGTTTCT-3′; TRAP, F 5′-GATGCCAGCGACAAGAGGTT-3′ and R 5′-CATACCAGGGGATGTTGCGAA-3′; NFATc1, F 5′-CAACGCCCTGACCACCGATAG-3′ and R 5′-GGGAAGTCAGAAGTGGGTGGA-3′; CK, F 5′-GAAGAAGACTCACCAGAAGCAG-3′ and R 5′-TCCAGGTTATGGGCAGAGATT-3′; MMP9, F 5′-CTGGACAGCCAGACACTAAAG-3′ and R 5′-CTCGCGGCAAGTCTTCAGAG-3′;GAPDH, F 5′-CTCCCACTCTTCCACCTTCG-3′ and R 5′-TTGCTGTAGCCGTATTCATT-3′.

Follow-up data of 3,951 patients from Kaplan–Meier plotter (kmplot.com/analysis/) and the GEO database were included in this study, which was divided into two groups according to the expression of SGK1 (28). Expression data of SGK1 was drawn from the Oncomine dataset, which embodies large quantities of high-throughput data of varies disease.

Results are presented as the means ± SD. Every experiment had a sample size greater than three and was independently performed three times, unless otherwise stated. Multigroup comparisons of the means were carried out by one-way ANOVA test with post hoc contrasts by Dunnett multiple comparisons test. Cumulative survival time was calculated by the Kaplan–Meier method and analyzed by the log-rank test.

We evaluated the expression of SGK1 during RANKL-induced osteoclastogenesis by quantitative real-time RT-PCR and immunoblotting. Upon RANKL treatment, the mRNA and protein levels of SGK1 were significantly increased (Fig. 1A and B; Supplementary Fig. S1A). We treated bone marrow mononuclear cells (BMM) once a day with different concentrations of GSK650394, an SGK1 inhibitor, in the presence of RANKL (100 ng/mL) and M-CSF (30 ng/mL) for 7 days (29). Osteoclast formation were inhibited in a dose-dependent manner (Fig. 1C and D). Subsequently, the potential toxicity of GSK650394 was evaluated by measuring the proliferation of BMMs via the Cell Counting Kit-8 (CCK-8) after treating the cells with different concentrations of GSK650394. However, the proliferation of BMMs was not markedly altered by GSK650394 at concentrations lower than 2 μmol/L (Fig. 1E).

To control the possible off-target effects of GSK650394, we knocked down SGK1 in RAW264.7 cells using small interfering RNAs (siRNAs; Supplementary Fig. S1B). Moreover, three shRNA lentiviruses tested (sh1, sh2, sh3) were selected to create stable SGK1 knockdown in both BMMs and RAW264.7 cells. The protein level of SGK1 was reduced as shown in Fig. 1F and Supplementary Fig. S1C. Upon RANKL treatment, osteoclast formation was significantly inhibited in cells with either siRNA or shRNA-mediated knockdown of SGK1 (Fig. 1G and H; Supplementary Fig. S1D–S1F). Therefore, SGK1 activity is upregulated during RANKL-induced osteoclastogenesis, and the suppression of SGK1 inhibits RANKL-induced osteoclastogenesis in vitro.

Given that osteoclast formation was impaired in SGK1-knockdown osteoclast precursors, we next investigated the effect of GSK650394 on bone resorption. Mature osteoclasts were seeded in calcium phosphate-coated plates in the presence of RANKL and M-CSF and treated with different concentrations of GSK650394 for 3 days. Compared with that in control cells, the formation of resorption pits in vitro was significantly reduced in GSK650394-treated osteoclasts (Fig. 1I and J). The luciferase reporter gene assays also suggested that GSK650394 treatment inhibited NF-κB luciferase reporter in BMMs (Supplementary Fig. S2A). Thus, impaired SGK1 activity inhibits the resorption function of mature osteoclasts.

The NF-κB and PI3K/AKT transcription factor families are both downstream of TRAF6, which play pivotal roles in osteoclastogenesis. SGK1 is known to modulate transcription by upregulating NF-κB activity through the phosphorylation and activation of IKKα, whereas in the PI3K/AKT pathway, SGK1 is highly homologous to AKT and is a critical downstream mediator of PDK1 (24). Specific inhibitors of the PI3K/AKT pathway result in the reciprocal activation of SGK1 (20). To explore the potential mechanisms by which SGK1 modulates osteoclastogenesis, we analyzed the expression of proteins associated with osteoclast formation in BMMs. Western blot analysis showed that the inhibition of SGK1 via GSK650394 dramatically suppressed the RANKL-induced phosphorylation and degradation of IKKα, IκBα, and p65 approximately 60 minutes after stimulation with RANKL (Fig. 2A; Supplementary Fig. S2A). Moreover, regarding to the AKT/PI3K pathway, GSK650394 treatment decreased the levels of p-AKT and the PI3K catalytic subunit p110 (Fig. 2B; Supplementary Fig. S2B). Therefore, the suppression of SGK1 inhibits RANKL-induced NF-κB and PI3K/AKT activation.

To determine whether the suppression of SGK1 inhibits the expression of c-Fos and NFATc1, which are essential transcription factors in osteoclast formation, we analyzed the expression of these two factors by Western blot analysis. As shown in Fig. 2C and Supplementary Fig. S2B, the protein levels of c-Fos and NFATc1 were upregulated during RANKL-induced osteoclast differentiation, which can be prevented efficiently by GSK650394 treatment.

We then tested whether GSK650394 suppresses the expression of osteoclast marker genes such as tartrate-resistant acid phosphatase (TRAP), MMP9, NFATc1, and cathepsin K (CK); notably, TRAP, MMP9, and CK are target genes of NFATc1 (30). As shown in Fig. 2D and E, GSK650394 inhibited the mRNA expression of MMP9 and CK during late stages of osteoclastogenesis and inhibited TRAP and NFATc1 at both early and late stages of osteoclastogenesis, respectively.

Orai1 is a SOCE channel on membrane, which regulates Ca²⁺ oscillations and NFATc1 transactivation during osteoclastogenesis (31). Immunofluorescence (IF) and Western blot analysis showed that the inhibition of SGK1 decreased Orai1 protein expression during osteoclastogenesis (Fig. 3A, C and D). Ca²⁺ imaging analysis confirmed essentially complete inhibition of SOCE in BMMs treated with GSK650394 compared with control (Fig. 3E). We therefore speculated that Orai1 might mediate the effect of SGK1 on osteoclast. To elucidate the role of Orai1 in the SGK1-mediated inhibition of osteoclast formation, we overexpressed Orai1 in BMMs (Fig. 3B). Scrambled or Orai1 overexpressed BMMs were treated with or without GSK650394 in the presence of RANKL (75 ng/mL) for 5 days, respectively. As shown in Fig. 3F and G, the inhibitory effects of GSK650394 on osteoclast differentiation were markedly alleviated in Orai1 overexpression group, indicating that Orai1 is a key downstream mediator of the terminal differentiation of osteoclasts by SGK1 (Supplementary Fig. S2C).

To investigate the effects of SGK1 on bone loss, we analyzed the bone mass of femurs in ovariectomized (OVX) mouse model. The result revealed that the reduction of intrabecular bone volume/tissue volume (BV/TV) and trabecular number (Tb.N) in the OVX mice was dramatically relieved by GSK650394 treatment (50 mg/kg; Fig. 4A and B). Furthermore, we performed TRAP staining in femoral sections. The number and size of osteoclasts in OVX mice were significantly higher than those in sham-operated mice. In response to GSK650394 treatment, the OVX mice demonstrated dramatically reduced ovariectomy-induced osteoclast activity (Fig. 4C). Histomorphometric analyses confirmed that the osteoclast surface/bone surface (Oc.S/BS) ratio was strikingly higher in OVX mice than in sham-operated mice (Fig. 4D). Furthermore, compared with those of untreated OVX mice, these osteoclastic parameters were significantly decreased in GSK650394-treated OVX mice (Fig. 4B).

In addition, compared with OVX group GSK650394 treatment significantly suppress the ovariectomy-induced elevation in type I collagen cross-linked C-terminal telopeptide (CTX-I) and RANKL levels in serum (Fig. 4E). Moreover, we tested the levels of OPG by ELISA. The results are similar (Fig. 4E). Therefore, compared with that in the untreated OVX mice, the RANKL/OPG ratio was markedly suppressed in the GSK650394-treated OVX mice (Fig. 4E).

To further verify the effect of GSK650394 on the formation of bone metastases, we used a model based on the intracardiac inoculation of tumor cells (32). A bone-seeking clone (MBS4-TJGK) of the human breast cancer cell line MDA-MB-231 was established by repeated sequential intracardiac inoculation of the cells into nude mice and further culture of metastatic cells obtained from bone metastases in vitro. After four cycles, MBS4-TJGK metastasize exclusively to bones. Then, MBS4-TJGK cells were inoculated intracardially. Nude mice were intraperitoneally injected with GSK650394 (50 mg/kg, five times per week) for 4 weeks. Bioluminescence imaging showed that compared with the DMSO-treated group, hind limb and spine tumor burden in mice was significantly decreased in the GSK650394-treated group (Fig. 5A and B). Micro-CT imaging confirmed that the mice treated with GSK650394 exhibited a reduction in bone lesions (Fig. 5C).

Because SGK1 is also highly expressed in heart and gastrointestinal tract, we then evaluate the toxicity of GSK650394. Mice were euthanized 3 weeks after injection and plasma biochemistry analyses were performed. The indicators for heart and liver injuries, as well as kidney functions did not change significantly at the concentration of 50 mg/kg (Fig. 5D and E). However, when lifted dose to 75 mg/kg, lactic dehydrogenase (LDH-L) dropped in treated group mice (Fig. 5D). These results suggest that intraperitoneal injection of GSK650394 in our model did not lead to reductions in hepatic and renal function but may have off-target influence on myocardial cells damaging with high dose.

Then, direct effect of GSK650394 treatment on MDA-MB-231 cell lines was tested by flow cytometry and CCK-8 (Fig. 5F and G). Apoptotic and necrosis cells were detected using FITC-Annexin V/propidium iodide (PI). Cells incubated with GSK650394 at the concentration of 5 μmol/L showed higher apoptotic rate among other groups. The proliferation of MDA-MB-231 cells were altered by GSK650394 in a dose-dependent manner.

Comparing the mRNA expression level of SGK1 in normal breast tissues and breast cancer tissues (Fig. 6A), we found SGK1 is significantly increased in breast cancer (33, 34, Perou and colleagues). It validated our results of RT-PCR. To further determine the clinical significance of SGK1, we examined whether SGK1 expression was associated with the clinical outcome of breast cancer patients (Fig. 6B). The survival analysis based on large population included in the KM plotter showed no significant prognostic performance of SGK1 on breast cancer patients (HR = 0.92; 95% CI, 0.82–1.02; P = 0.120). But in other datasets GSE2603 (HR = 2.24; 95% CI, 1.05–4.79; P = 0.033) and GSE12276 (HR = 1.74; 95% CI, 1.30–2.34; P < 0.001) we found high expression of SGK1 could predict poor prognosis, indicating the potential oncogenic effect of SGK1 (35, 36).

Ca²⁺ signaling is essential for diverse biological processes and is known to be involved in osteoclast formation and function (37–43). Ca²⁺ release from intracellular stores and Ca²⁺ entry across the cell membrane both accomplish Ca²⁺ in cytoplasm. SOCE, a major Ca²⁺ influx pathway in most nonexcitable cell types, is mediated by the Ca²⁺ release-activated Ca²⁺ channel, which consist of Orai1, 2, or 3 binding to their regulators STIM 1 or 2 (31). Previous work has identified SGK1 as a powerful stimulator of several ion channels including Orai1. Some clinical studies have been undertaken to corroborate the relevance of SGK1 in diseases such as hypertension and cancer. The mechanism can be related to coordinate with different signaling pathways include FoxO1 and NF-κB and lead to target organ inflammation (44). However, it was unknown whether SGK1 affects osteoclast differentiation and bone resorption via Ca²⁺ channel through Orai1 until now.

In our study, we observed SGK1 expression was upregulated upon the initiation of osteoclastogenesis. Both pharmacologic inhibition and knockdown of SGK1 by shRNA suppressed osteoclast differentiation and function. In addition, Orai1 protein abundance appeared as well as the significant decrease of Ca²⁺ influx after GSK650394 treatment (Fig. 3A, C–E). Moreover, overexpressing Orai1 in BMMs counteracted the suppressive effect of GSK650394 on osteoclast formation. On the basis of these data, we concluded that SGK1 expression is closely related to osteoclastogenesis and has a strong effect on Orai1. To further confirm SGK1 inhibition on ovariectomy-induced bone loss in vivo, we used GSK650394 to prevent bone loss in OVX mouse model. We observed attenuated ovariectomy-induced bone loss in GSK650394 treated group, which was consistent with results in vitro. In the previous work, SGK1 was showed to be evolved in nuclear translocation and activation of NF-κB (24, 45, 46). SGK1 could phosphorylate IKKα, which lead to IKKβ activation. The IKK complex and the NF-κB pathway would be activated similarly, which were confirmed in our study. In addition, we demonstrated that Orai1established an previously unknown link between SGK1 and osteoclast genesis.

SGK1 expressed heavily in several tumor cells, including colon cancer, myeloma, medulloblastoma, prostate cancer, and ovarian tumors (47, 48). Some research has shown that SGK1 contributes to the glucocorticoid-induced resistance of breast cancer cells to chemotherapy, and SGK1 silencing could increase the toxicity of chemotherapeutic drugs. On the basis of the Oncomine database, we found the expression of SGK1 was significantly upregulated in different types of breast cancer compared with normal breast tissue. However, the result of further survival analysis in breast cancer varies from different database. To investigate the exact role of SGK1 in prognosis of breast cancer, further research should be conducted in specific population cohort with confounding factors adjusted.

SGK1-dependent regulation of Orai1 has been shown to play a critical role in regulation of cell proliferation (17). In this study, we investigated the effect of SGK1 on bone metastasis in breast cancer using an intracardiac injection-induced model of bone metastasis in mice. Our results demonstrated the effects of GSK650394 in the inhibition of breast cancer bone metastasis. Besides the effects on the cell proliferation, more application of SGK1 inhibition has been explored.

On the basis of this study, SGK1 is a powerful regulator in osteoclastogenesis and bone metastasis. SGK1 inhibitors might be a promising therapeutic approach target for various pathologic conditions including bone metastasis. Because we are just exploring the effect of SGK1 on osteoclastogenesis, the possible osteoblastic effect also needs to be considered. Previous study has confirmed that SGK1 could mediate osteoinductive effects of phosphate in the vascular tissue and trigger vascular calcification. SGK1 leads to IκBα phosphorylation and NF-κB activation, which results in osteo-/chondrogenic transdifferentiation of VSMCs. Further study needs to be done in the future (49–51).

There are several limitations to our study. Because SOCE is not the only calcium entry pathway, the calcium influx could be through t-type calcium channel, TRP channels, and so on. We used a selective SOCE-specific inhibitor BTP2 and monitored calcium influx after intracellular store depletion (52). The drug abolished the Ca²⁺ transient amplitude but cannot be fully rescued by Orai1 overexpression (Supplementary Fig. S1B). Also, upregulation of Orai1 is not the only way SGK1 modify Ca²⁺ activity. SGK1 can also have effects on activation of K⁺ channel or Ca²⁺-permeable TRP channels which could hyperpolarize the cell membrane and influence the Ca²⁺ channels. Our results only focused on Orai1, not on the whole family of Orai. It is also possible that different levels of other isoforms, Orai2 and Orai3, are present in the cells in both groups (53, 54). It highlights the need for future deeper studies to analyze the differences and connections among the three. Another limitation is that the effects of the inhibitor is not specific for bone metastasis, but also impair the survival of tumor cells. Although in our experiment, GSK650394 did not show a direct effect on breast cancer apoptosis (Fig 5F), high levels of SGK1 have been found as compensation upon PI3K or AKT pathway and contribute to the resistance to AKT inhibitors in breast cancer cells, whereas loss of SGK1 profoundly impacts thyroid cancer cell proliferation and survival despite intact PI3K and AKT activity (50, 55, 56). Therefore, the effect of GSK650394 to bone metastasis may be composed of complex factors.

No potential conflicts of interest were disclosed.

Conception and design: Z. Zhang, C. Song, H. Guan, F. Li

Development of methodology: Z. Zhang, Q. Xu, B. Mi, H. Zhang, H. Liu, Y. Sun, H. Guan, F. Li

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Z. Zhang, Q. Xu, B. Mi, H. Zhang, H. Liu, Y. Sun, F. Li

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Z. Zhang, Q. Xu, C. Song, J. Wang, Z. Lei, H. Guan, F. Li

Writing, review, and/or revision of the manuscript: Z. Zhang, Q. Xu, H. Zhang, H. Guan, F. Li

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Z. Zhang, Q. Xu, B. Mi, H. Kang, J. Wang, Z. Lei, F. Li

Study supervision: H. Guan, F. Li

This work was supported by National Natural Science Foundation of China (no. 81572857, to H. Guan; no. 81401762, to J. Wang; and 81472133 to F. Li)

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.