SGK1 in Cancer: Biomarker and Drug Target

Serum- and glucocorticoid-regulated kinases (SGKs) are members of the AGC family of serine/threonine kinases, consisting of three isoforms: SGK1, SGK2, and SGK3. SGK1 was initially cloned as a gene transcriptionally stimulated by serum and glucocorticoids in rat mammary tumor cells. It is upregulated in some cancers and downregulated in others. SGK1 increases tumor cell survival, adhesiveness, invasiveness, motility, and epithelial to mesenchymal transition. It stimulates tumor growth by mechanisms such as activation of K+ channels and Ca2+ channels, Na+/H+ exchanger, amino acid and glucose transporters, downregulation of Foxo3a and p53, and upregulation of β-catenin and NFκB. This chapter focuses on major aspects of SGK1 involvement in cancer, its use as biomarker as well as potential therapeutic target.


Introduction
The serum-and glucocorticoid-inducible kinase-1 (SGK1) was first discovered in rat tumor cells as a gene, transcriptionally regulated by serum and glucocorticoids [1] and, later, in human as gene upregulated by cell shrinkage [2]. The human SGK1gene is localized in chromosome 6q23. SGK1 is a member of AGC kinase family, which also includes kinases such as PKA, PKC, AKT, RSK, S6K, PDK1. As most AGC kinases, SGK1 consists of three domains: an N-terminal variable region, a catalytic domain, and the C-terminal tail ( Figure 1). SGK1, SGK2, and SGK3 are closely related to AKT kinases; however, they lack lipid-binding PH domain.
Currently used biomarkers could be grouped into risk assessment markers (also known as diagnostic), prognostic markers, and predictive markers. By testing for mutations or other abnormalities in known genes, it is possible to identify individuals who are at increased risk of developing cancer. Identification of these patients has the potential to result in early diagnosis and possibly prevention. Prognostic markers can be characterized as factors which correlate with patient outcome. They are most useful for identification of either patients with favorable outcome which do not require adjuvant systemic therapy or patients with prognosis too poor for conventional approaches [42]. Predictive markers can be defined as factors which predict response or resistance to the specific therapy.
Because of quite an impressive involvement of SGK1 in the development and growth of various tumors, it would be logical to predict, that this kinase is a potential diagnostic, prognostic, or predictive biomarker in cancers. The potency of SGK1 as a biomarker was already shown in several studies.
Fifteen cortisol-secreting adrenocortical adenomas were analyzed together with matched blood samples using high-resolution single nucleotide polymorphism microarrays in order to detect copy number alterations and copy-neutral losses. Forty six recurrent copy number alterations were identified that each affected a single gene (31 gains and 15 losses), including SGK1. CN losses in SGK1 gene were also confirmed by FISH and DNA qRT-PCR analysis and corresponded to a low mRNA expression level [43]. Another study was performed on 227 adrenocortical tumors (40 adenomas and 187 carcinomas) and 25 normal adrenal tissues using immunohistochemistry and/or tissue microarrays. In addition, 62 frozen tumor samples were used for mRNA analysis. It was found that SGK1 mRNA levels were lower in cortisol-secreting than in nonsecreting tumors (p < 0.005). Nonsecreting cancers displayed a significant correlation between SGK1 and CTNNB1 mRNA levels (p < 0.001; r = 0.57). Overall survival was shorter in patients with low SGK1 protein expression (HR = 2; 95% CI = 1.24-3.24; p = 0.0048). However, disease-free survival was not significant, although with promising tendency (HR = 1.98; 95% CI = 0.9-4.3; p = 0.08). Subgroup of patients with a low SGK1 combined with high nuclear β-catenin protein expression showed poor prognosis (HR = 3.3; 95% CI = 0.5-7.3; p = 0.03) [44]. The cell culture experiments, using 21 breast cancer cell lines had shown that sensitivity of breast cancer cell lines to AKT inhibitors AZD5363 and MK-2206 correlates with SGK1 mRNA levels. Moreover, knockdown of SGK1 impairs proliferation of AKT-inhibitor-resistant but not -sensitive cells. That effect can be rescued by ectopic SGK1 overexpression [45]. In 224 patient samples and 103 matched adjacent of non-small cell lung cancer samples, it was first established that expression of SGK1 mRNA in cancerous tissues was way higher compared to the adjacent non-cancerous tissues (p < 0.001). Interestingly, high SGK1 protein expression was significantly associated with differentiation (χ 2 = 5.279, p = 0.022) and histological type (χ 2 = 4.127, p = 0.042). High expression of SGK1 (HR = 1.926; 95% CI = 1.452-2.903; p < 0.001) was a significant negative prognostic factor for five-year survival. In addition, multivariate Cox regression analysis demonstrated that high expression of SGK1 (HR = 1.726; 95% CI = 1.396-2.865; p < 0.001) is an independent prognostic factor for the five-year survival [46]. (Table 1) Adrenocortical tumor immunohistochemistry, tissue microarrays.

SGK1 as Drug Target
Since abnormalities in SGK1 expression, activity, and regulation have also been found in pathological conditions, search for small-molecule SGK1 inhibitors (Figure 4) for the therapeutic purposes was rigorously initiated and is still ongoing.
SI113 is a selective SGK1 inhibitor, which has an IC50 of 600 nM (Figure 3). Treatment of RKO human colon carcinoma cell line with 12.5 µM SI113 for 24 h showed an impressive delay in cell cycle progression and accumulation in G0-G1, when compared with untreated cells (p = 0.0024). Moreover, combination of SI113 (12.5 µM) with paclitaxel (50 nM) significantly increased apoptosis and necrosis in RKO cells (p = 0.00037) [47]. SI113 was also capable to inhibit cell cycle progression and induce apoptosis in HuH-7 and HepG2 hepatocellular carcinoma cell lines. In addition, NOD/SCID female mice were implanted with HuH-7 cells for in vivo treatment with SI113 and showed a significantly smaller tumor volume than control mice (p = 0.0009). SI113 also potentiated and synergized with radiotherapy in tumor killing by reducing MDM2 phosphorylation on serine 166 by SGK1 [48]. In glioblastoma multiforme cell lines, significant increase in caspase-mediated apoptosis was detected (p < 0.05). It also found that SI113 was co-inducing proliferation inhibition as well as cell cycle block together with radiotherapy, oxidative-stress-mediated cell viability, and autophagy [5]. In endometrial cancer cells, SI113 induced apoptosis, as proven by the cleavage of the apoptotic markers PARP and Caspase-9. It also induced autophagy, which was shown by the increase of the autophagy markers LC3B-II and beclin I. The effects were associated with the induction of endoplasmic reticulum stress markers GRP78 and CHOP [49]. SI113 restores sensitivity to taxanes in tumors resistant or maderesistant to taxanes [59] This inhibitor, in synergy with radionuclides, induces theranostic effects in glioblastoma [60]. GSK650394 is a selective SGK1 inhibitor, which has an IC50 of 62 nM (Figure 3). Since SGK1 expression was increased under hypoxia and regulated unsaturated fatty acid uptake in NCI-H460 lung adenocarcinoma cells, GSK650394 was applied in order to investigate this effect. GSK650394 caused reduction in fatty acid uptake, decreased longterm survival, and sensitized to the cytotoxic effects of ionizing radiation (p ≤ 0.05) [50]. The analog of GSK650394, QGY-5-114-A, inhibited HT29 cell proliferation (p < 0.001) and HCT116 cell migration in vitro (p < 0.001). This inhibitor also obstructed colonic tumor growth and HCT116 cell proliferation in vivo (p < 0.05) [51]. In prostate cancer PC3 and LNCaP cells, GSK650394 impaired migration and invasion in vitro (p < 0.05). In addition, GSK650394 treatment also induced autophagy, which led to the inhibition of cell metastasis (p < 0.05) [52]. Stimulation of non-small cell lung cancer cell with γradiation and GSK650394 induced apoptosis and p53 pathway [46]. Using the subcutaneous xenotransplant mouse model of colorectal cancer (HCT116 and HT29 cells), it was shown that GSK650394 repressed tumor cell proliferation (p < 0.05) and tumor growth (p < 0.001). In addition, GSK650394 reduced SGK1 expression and increased p27 expression levels in xenograft tumors (p < 0.05) [53]. In a study, mice injected with cervical cancer xenograft (ME180 cells) GSK650394 in combination with melatonin, showed a significant tumor size decrease in all cases and even complete tumor remission in 33% of mice (p ≤ 0.001) [54]. Inhibition of SGK1 with GSK650394 reduced the radioresistance of colorectal cancer in xenotransplant mouse of HT29 cells. The combination of inhibitor with radiotherapy resulted in minimal tumor size compared to radiotherapy or inhibitor alone (p < 0.05) or control groups (p < 0.01) [55]. Combined PDGFR inhibitor (CP-673451) and GSK-650394 treatment of xenotransplantation breast cancer models (MDA-MB-231 and BT-549) showed significantly decreased tumor formation compared to control or single agent treatment [61].
EMD638683 is a highly selective inhibitor of SGK1 kinase, which has an IC50 of 3 µM (Figure 3). In vitro, EMD638683 treatment of colon cancer CaCo-2 cells significantly increased the radiation-induced decrease of forward scatter, increase of phosphatidylserine exposure, decrease of mitochondrial potential, increase of caspase 3 activity, increase of DNA fragmentation, and increase of late apoptosis (p < 0.01). In vivo, the number of developing tumors following chemical carcinogenic treatment was significantly blunted by EMD638683 treatment [56]. In breast cancer MCF-7 cells, EMD638683, GSK650394, and testosterone albumin conjugate induced strong apoptotic response and caspase 3 activation (p < 0.01), enhanced radiation-induced cell growth control (p < 0.001) and induced late FAK and AKT dephosphorylation (p < 0.01) [57]. EMD638683 treatment resulted in a statistically significant decline of cell viability in both RD and RH30 rhabdomyosarcoma cells (p < 0.05). In addition, doxorubicin treatment decreased the viability of RD and RH30 cells, which was significantly enhanced by the synchronous administration of EMD638683 (p < 0.05). The migration and clonogenic potential of RD and RH30 cells were significantly decreased in the presence of EMD638683 (p < 0.05) [58].

Conclusions and Future Perspectives
Cancer biomarker research has developed enormously over the last 15 years, promising new possibilities for cancer diagnostics. Nonetheless, most of the discovered biomarkers have some limitations such as deficiency of methodological standardization, quality control, or significant comparison between healthy individuals and cancer patients. At the moment, all of experimentally discovered biomarkers still need careful validation before they could be applied in everyday diagnostics. Same rules also apply for SGK1; therefore, a lot more pre-clinical and clinical research are needed in order to figure out whether SGK1 is suitable for diagnostics. On the other hand, the rise of "omics" techniques such as genomics, transcriptomics, proteomics, phophoproteomics, kinomics, and metabolomics provide more sensitive detection of biomarkers [42] and allows the use of entire sets of genes instead of just several or sometimes even one biomarker. The combination of different "omics" techniques allows further combination of different types of biomarkers (for example proteins + metabolites + mutated DNA) [62].
In the past decade, developments in the small-molecule SGK1 kinase inhibitor field have led to several products and many more are still in development. Pre-clinical studies in cell lines and animal models provide basal information for the arrangement of clinical studies evaluating the efficiency and side effects of these agents. Clinical studies with SGK1 kinase inhibitors will help to determining which of these inhibitors are most effective for anticancer therapy. Many types of cancers can be targeted by SGK1 inhibitors; thus, many more clinical trials are needed based on pre-clinical findings. Recent increase in anti-kinase small-molecule inhibitors or monoclonal antibodies is already approved by the U.S. Food and Drug Administration, suggesting some perspectives and hopes for SGK1 [63].
In conclusion, the future in SGK1 research both as biomarker and drug target is in our hands and should definitely not be disregarded.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.