Discussion
The present study realized a specific blockade of the AR and its target gene IGF-1, which induces both apoptosis and cell cycle arrest by a novel antiandrogen SBF-1. This unique compound was used for the treatment of prostate cancer in a xenograft model. As we know, AR is a hormone inducible transcription factor, which drives expression of tumor promoting genes and represents an important therapeutic target in prostate cancer (Dalal et al., 2018). Higher expression of AR in prostate cancer with shorter overall survival was found in patients (Fig. S3). Currently, small molecule drugs used in the treatment of prostate cancer mainly interfere with steroid recruitment to prevent AR-driven tumor growth (Caboni & Lloyd, 2013; Lallous, Dalal, Cherkasov & Rennie, 2013). However, those kinds of small molecules are rendered ineffective in the advanced prostate cancer by emergence of LBD mutations or expression of constitutively active variants such as AR-V7 that lack the LBD. As the cell lines of prostate cancer, LNCaP cells express AR with a novel mutation T878A in the AR-LBD, which is similar to human prostatic adenocarcinoma, and their growth can be inhibited by androgen withdrawal. Another cell line PC3/AR+ cells have a stable expression of ARWT, which is originally a sub-line from the parental cells AR negative PC3 cells. By using these two cell lines, we aimed at finding a novel inhibitor against prostate cancer in both ARWT and ARmutant levels through targeting AR.
As the result, SBF-1 strongly inhibited proliferation of both LNCaP and PC3/AR+ cells, suggesting a possibility that this compound can be used for the treatment of prostate cancer with both ARWT and ARmutant. For investigating the mechanisms underlying its effect, we next found that SBF-1 showed a significant inhibition on the adhesive ability of both LNCaP and PC3/AR+ cells to fibronectin and laminin, increased the percentage of apoptotic cells, and caused the cell cycle arrest in G1 and G2/M phase.
AR can be activated by the binding of endogenous androgens, including testosterone and DHT (Gao, Bohl & Dalton, 2005). The AR mediates the growth of benign and malignant prostate in response to DHT. In patients undergoing androgen deprivation therapy for prostate cancer, AR drives prostate cancer growth despite low circulating levels of testicular androgen and normal levels of adrenal androgen (Mohler et al., 2011). Usually, prostate cancer cells gradually lose their dependence on androgen signaling through the AR and become resistant to hormonal therapy. It is now established that throughout various hormonal manipulations, castrate-resistant prostate cancers continue to express the AR (Hobisch, Culig, Radmayr, Bartsch, Klocker & Hittmair, 1995; Hobisch, Culig, Radmayr, Bartsch, Klocker & Hittmair, 1996; Sadi & Barrack, 1991; Tilley, Lim-Tio, Horsfall, Aspinall, Marshall & Skinner, 1994; van der Kwast et al., 1991) and IGF-1 as the transcription product of AR has been shown to independently activate the AR in the absence of DHT, with a mechanism that involves downstream phosphorylation of either the AR or its associated proteins (Culig et al., 1994; Gregory et al., 2004). It has also been reported that inhibition of IGF-1 could suppress prostate cancer cell growth (Zheng et al., 2012). On the other hand, IGF-1 will also activate its downstream proteins such as AKT kinase, which regulates the cell proliferation and survival. AKT is usually activated by hormones, growth factors, and chemical drugs (Zheng et al., 2015; Zheng et al., 2012; Zhu et al., 2015). The downstream forkhead transcription factor family FOXO plays a vital function in cell apoptosis and survival in variety of cell types, which could be phosphorylated by AKT kinase (Tzivion, Dobson & Ramakrishnan, 2011). Those findings inspired us to test the effect of SBF-1 on the AR signaling while stimulated with DHT. As shown in Fig. 2A, the expressions of IGF-1, PCNA, Bcl-2, pARS515, pAKTS473 and pFOXO1S256 were greatly decreased in both LNCaP and PC3/AR+ cells by SBF-1 without effects on the AR, AKT1, and FOXO1 expressions. This result suggests that SBF-1 may downregulate both AR/IGF-1 and IGF1/AKT/FOXO1/PCNA pathways. When we used DHT to stimulate AR, DHT greatly increased the AR expression as well as the AR phosphorylation and the expressions of downstream proteins IGF-1 and PCNA. Against this, SBF-1 down-regulated DHT-increased expressions of AR, pARS515, IGF-1, and PCNA in both cell lines (Fig. 2B). This result indicates a dual effect of SBF-1 on AR/IGF1 axis and their down-stream signaling. SBF-1 also strongly suppressed the mRNA expression of IGF-1 and PCNA even in the case of DHT-doubled expression (Fig. 2C). These findings suggest a possibility of SBF-1 to directly block the gene transcription mediated by AR that is known to bind the promoter of its target genes.
To examine how SBF-1 affects AR and its subsequent signaling pathways, we hypothesized that SBF-1 might directly bind to AR since it is a steroidal glycoside. Both MST and ITC assays were used to examine the binding affinity between SBF-1 and purified ARWT. After a strong binding was concluded (Fig. 2D, and E), we performed the polarity shift assay to check whether the AR-LBD is the binding site of SBF-1 because AR-LBD is recognized as the vital domain to be targeted in prostate cancer therapy (Wang et al., 2006). However, despite the significant binding to AR-LBD by DHT as a positive control, what we hypothesized for the binding of SBF-1 to AR-LBD is absolutely negative (Fig. 2F). Above findings suggest that SBF-1 does bind to the AR protein but AR-LBD is not the binding site.
AR signaling in CRPC tumor epithelial cells could be caused by activating AR point mutations. Such mutations are rare in untreated PC, but detected in 15–20% of CRPC patients (Grasso et al., 2012; Robinson et al., 2015; Taylor et al., 2010), and in up to 40% of CRPC patients treated with AR antagonists (Balk, 2002). Activating AR point mutations generally affects the c-terminal LBD, while about one-third occur in the transactivating NTD (Gottlieb, Beitel, Nadarajah, Paliouras & Trifiro, 2012; Steinkamp et al., 2009), resulting in broaden ligand specificity. The first and most frequently identified AR point mutation is the flutamide-driven T878A mutation (Fenton et al., 1997; Taplin et al., 1999; Taplin et al., 1995; Veldscholte et al., 1990), while W742C also has been reported after treatment with first-generation AR antagonists (Lallous et al., 2016; Steketee, Timmerman, Ziel-van der Made, Doesburg, Brinkmann & Trapman, 2002; Tan et al., 1997; Taplin et al., 2003; Watson, Arora & Sawyers, 2015; Yoshida et al., 2005). The T878A and L702H mutations have been observed in CRPC patients during abiraterone treatment (Lallous et al., 2016; Steketee, Timmerman, Ziel-van der Made, Doesburg, Brinkmann & Trapman, 2002; Tan et al., 1997; Taplin et al., 2003). Also, F876L mutation confers an antagonist-to-agonist switch that drives phenotypic resistance (Korpal et al., 2013). Taken together, we decided to cover these frequently occurring mutations in CRPC through the construction of AR mutants as shown in (Fig. 3) and determined the effect of SBF-1 on those mutants. As the result, SBF-1 showed a strong binding to all the mutant constructs, L702H, W742C, F876L and T878A (Fig. 3A). It should be emphasized that SBF-1 showed an almost complete inhibition against the DHT-increased risen activity of ARWT and the mutants (Fig. 3B). Above results suggest that SBF-1 may have a novel binding site of SBF-1 on AR.
Growing evidence suggests that castration resistance of prostate cancer may be partially mediated by AR splice variants lacking the LBD coding sequence, which leaves only the NTD and DBD as viable domains that are targetable by small molecules (Dehm, Schmidt, Heemers, Vessella & Tindall, 2008; Guo et al., 2009; Hu et al., 2009; Sun et al., 2010a; Watson et al., 2010). Inhibition of splice variant transcriptional activity would be a significant breakthrough in the development of a new class of anti-AR drugs (Dalal et al., 2014). To find the binding site of SBF-1 in AR, we constructed an ARE-1 sequence, which known to be consensus recognition site for the AR (Denayer, Helsen, Thorrez, Haelens & Claessens, 2010). This constructed sequence was incubated with each purified ARWT or AR∆DBD (Androgen receptor lacking DBD). As the result, SBF-1 did show a binding to ARWT but not AR∆DBD (Fig. 4A, and B), suggesting that AR-DBD might be the target domain for SBF-1. The ITC technique confirmed that SBF-1 totally failed to show any binding signal with the purified AR∆DBD (Fig. 4C). To understand the result of SBF-1 binding to AR-DBD, furthermore, the AR-induced gene expression enrichment assay demonstrated that DHT induced a significant increase in the AR target gene IGF-1 enrichment while SBF-1 diminished this enrichment (Fig. 4E). These results suggest that SBF-1 might block AR from binding to its target genes by binding to its DBD.
The binding between SBF-1 and AR or its mutants leads to further insight into where could SBF-1 binds to AR, and especially all the found mutants are located in AR-LBD. This hints us that the current efforts for the treatment of prostate cancer to target AR may become resistant after the mutation accompanying with the malignant progression, and a novel inhibitor targeting the new domain of AR is needed. To our knowledge, a small molecule EPI-001 was reported to block transactivation of the NTD and was specific for inhibition of AR without attenuating transcriptional activities of related steroid receptors (Andersen et al., 2010). On the other hand, targeting AR-DBD may be a new strategy for CRPC treatment (Dalal et al., 2014). However, there is still little development of inhibitors that specifically target the NTD or DBD of the AR (Caboni & Lloyd, 2013; Lallous, Dalal, Cherkasov & Rennie, 2013). Here we propose a solid proof of strong anti-tumor small molecule SBF-1 that inhibit the growth of advanced prostate cancer through the binding to AR-DBD. That is why SBF-1 can bind both AR and its multiple mutants for the inhibition of both ARWT and ARmutant cells. In fact, IGF-1/AKT/FOXO1 axis has been known to be of importance in prostate cancer castration resistance. Stimulation of AR through androgens, i.e. DHT has shown to activate the antiapoptotic IGF-1/AKT/FOXO1/PCNA pathway in LNCaP and PC3/AR+ cells (Zhao, Tindall & Huang, 2014). The activation of IGF-1/AKT/FOXO1/PCNA signaling is critical for mediating cell survival and involved in the castration resistance through the modulation of AR expression and its down-stream signaling. In this study, we used LNCaP cells that express AR, including the common AR mutation, ETV1 overexpression, and PTEN deletion, which serves as a good model to examine late-stage prostate cancer with metastatic potential, and also, PC3/AR+ cells that express AR but WT and has normal PTEN expression (Kim, Park & Dong, 2006). These two cell lines will give a broad spectrum of how mutant prostate cancer cases could be affected by the treatment of SBF-1.
Currently, targeting AR-DBD has become increasingly in-focus as the AR-DBD exists in all forms of AR and its mutants, also helps in directly treating castration resistance in prostate cancer (Dalal et al., 2018). Blocking AR from regulating its target gene IGF-1 could help overcome castration resistance, and stimulation of IGF-1 using glucose will lead to the over-expression of IGF-1, which could explain how SBF-1 works. SBF-1 almost completely overcame the activated IGF-1 signaling through glucose intake or DHT stimulation (Fig. 5). This characteristic suggests that SBF-1 may have different mode of actions from the current agents against prostate cancer despite the endogenous levels of androgens and AR mutation, which is quite beneficial to various situations of PCa patients. Finally, such unique mechanism of SBF-1 was tested on prostate cancer growth in nude mice, where SBF-1 significantly reduced the tumor size in mice bearing either ARWT or ARmutant cells, and prolonged the survival rate at very low doses in a dose-dependent manner, with a strong decrease in the IGF-1 protein and its downstream signaling but without loss of body weights of tumor-bearing mice (Fig. 6). These findings provided great advantages for the targeting of AR-DBD against AR wild type and mutant prostate cancer by SBF-1.
In summary, we present a novel antiandrogen targeting AR-DBD, for the first time, which attenuates AR in a wide spectrum of different variants for the treatment of prostate cancer especially the castration resistant prostate cancer (summarized in Fig. 7). Our findings suggest a better strategy in dealing with the development of advanced prostate cancer by targeting AR-DBD rather than the conventional methods.