Nimbolide-induced oxidative stress abrogates STAT3 ...

Antioxidants & Redox Signaling Nimbolide-induced oxidative stress abrogates STAT3 signaling cascade and inhibits tumor growth in transgenic adenocarcinoma of mouse prostate model (doi: 10.1089/ars.2015.6418) This article has been peer-reviewed and accepted for publication, but has yet to undergo copyediting and proof correction. The final published version may differ from this proof.

1

Original Research Communication Title of the article: Nimbolide-induced oxidative stress abrogates STAT3 signaling cascade and inhibits tumor growth in transgenic adenocarcinoma of mouse prostate model

Authors: Jingwen Zhang1#, Kwang Seok Ahn2#, Chulwon Kim2, Muthu K. Shanmugam1, Kodappully Sivaraman Siveen1, Frank Arfuso3, Ramar Perumal Samy4,5,6, Amudha Deivasigamani7, Lina Hsiu Kim Lim6, Lingzhi Wang1,8, Boon Cher Goh1,8, Alan Prem Kumar1,8,9,10*, Kam Man Hui7*, and Gautam Sethi1,9*

Institutions in which the work was done: Department of Pharmacology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 117597; College of Korean Medicine, Kyung Hee University, #47, Kyungheedae-gil, Dongdaemoon-gu, Seoul 130-701, Republic of Korea; Division of Cellular and Molecular Research, Humphrey Oei Institute of Cancer Research, National Cancer Centre, Singapore 169610; and Cancer Science Institute, National University of Singapore, Centre for Translational Medicine (CeTM), 14 Medical Drive, #11-01M, Singapore 117599.

An abbreviated title: NL modulates GSH/GSSG system and STAT3 activation.


2 Corresponding authors: 1. Dr. Gautam Sethi, Department of Pharmacology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 117597; Phone: +65-65163267; Fax: +65-68737690; E-mail: [email protected] 2. Dr. Alan Prem Kumar, Cancer Science Institute of Singapore, National University of Singapore, Centre for Translational Medicine (CeTM), 14 Medical Drive, #11-01M, Singapore 117599. Phone: +65 6516 5456; Fax: +65- 68739664 ; Email: [email protected] 3. Prof. Kam M Hui, Division of Cellular and Molecular Research, Humphrey Oei Institute of Cancer Research National Cancer Centre, Singapore 169610, Phone: +65-64368337; Fax: +6562263843; Email: [email protected]

Author’s affiliations: 1

Department of Pharmacology, Yong Loo Lin School of Medicine, National University of

Singapore, Singapore 117597 2

College of Korean Medicine, Kyung Hee University, #47, Kyungheedae-gil, Dongdaemoon-gu,

Seoul 130-701, Republic of Korea 3

School of Biomedical Sciences, CHIRI Biosciences Research Precinct, Curtin University,

Western Australia 6009 4

Venom and Toxin Research Programme, Department of Anatomy; Yong Loo Lin school of

Medicine, National University of Singapore, Singapore 5

Infectious Diseases Programme, Department of Microbiology; Yong Loo Lin school of

Medicine, National University of Singapore, Singapore


3 6

Department of Physiology, NUS Immunology Programme, Centre for life sciences, Yong Loo

Lin school of Medicine, National University of Singapore, Singapore 117456 7

Division of Cellular and Molecular Research, Humphrey Oei Institute of Cancer Research,

National Cancer Centre, Singapore 169610 8

Cancer Science Institute, National University of Singapore, Centre for Translational Medicine

(CeTM), 14 Medical Drive, #11-01M, Singapore 117599 9

School of Biomedical Sciences, Curtin University, Australia

10

Department of Biological Sciences, University of North Texas, Denton, Texas, USA, 76203

Word count: 5674 Reference numbers: 49 Number of grey scale illustrations: 8 Number of color illustrations: 1 Number of supplemental figures: 10 Both JZ and KSA contributed equally to this work.


4 ABSTRACT Aims: Prostate cancer (PCa) is one of the most commonly diagnosed cancers worldwide. Currently available therapies for metastatic PCa are only marginally effective; hence novel treatment modalities are urgently required. Considerable evidence(s) suggest that deregulated activation of oncogenic transcription factor, signal transducer and activator of transcription 3 (STAT3) plays a pivotal role in the development and progression of PCa. Thus agents that can abrogate STAT3 activation could form the basis of novel therapy for PCa patients. In the present study, we analyzed whether the potential anticancer effects of nimbolide (NL), a limonoid triterpene derived from Azadirachtaindica, against PCa cell lines and transgenic adenocarcinoma of mouse prostate (TRAMP) model are mediated through the negative regulation of STAT3 pathway. Results: Data from the in vitro studies indicated that NL could significantly inhibit cell viability, induce apoptosis and suppress cellular invasion and migration. Interestingly, NL also abrogated STAT3 activation, and this effect was found to be mediated via an increased production of reactive oxygen species (ROS) due to GSH/GSSG imbalance. Oral administration of NL significantly suppressed the tumor growth and metastasis in TRAMP mouse model without exhibiting any significant adverse effects. Innovation: The present study demonstrates the critical role of GSH/GSSG imbalance-mediated ROS production contributing to the STAT3 inhibitory and tumor suppressive effect of NL in PCa. Conclusion: Overall our findings indicate that NL exhibits significant anticancer effects in PCa that may be primarily mediated through the ROS-regulated inhibition of STAT3 signaling cascade.


ZHANG ET AL.

INTRODUCTION Ranked as the second most common cancer for male worldwide, PCa is the fifth leading cause of cancer-related mortality (34). Current treatments for PCa are dependent on multiple high-risk features including PSA levels >20, biopsy Gleason score 8–10, or stage T3 disease (15). Organ confined cancers are generally cured by prostatectomy or radiation therapy, while for patients with advanced cancer, androgen deprivation therapy (ADT) constitutes the major treatment modality (41). Although ADT is effective in the first few years of treatment, most patients eventually develop resistance to this therapy and progress into incurable castration resistant PCa (CRPC) (31). Currently FDA approved pharmacological agents (e.g. docetaxel and abiraterone) are capable of controlling CRPC (31), limitations still exist such as drug resistance, chemotoxic effects and moderate increase in overall survival (16,22). Thus novel pharmacological strategies are needed to improve therapeutic outcome in PCa patients. A common characteristic of PCa is its dependence on activated STAT3 for proliferation and survival. STAT3 is an oncogenic transcription factor that is constitutively activated in many solid tumors, including PCa (4,14,39,42,44). Normally STAT3 is present in an inactive form as a monomer in the cytoplasm; once the monomer is phosphorylated by the receptor-associated tyrosine kinases, it can form active dimers and then migrate into the nucleus to induce gene transcription (42). The receptor-associated tyrosine kinases are activated through ligand engagement; these ligands include the cytokine such as interlukin-6 (IL-6) and growth factors such as epidermal growth factor (EGF). The persistent activation of STAT3 contributes to the development of PCa through distinct mechanism(s). First, STAT3 is involved in the progression of PCa; it participates in the change from hyperplasia to neoplasia in prostate epithelial cells (3). Second, STAT3 determines the survival of PCa cells; inhibition of STAT3 by

1


ZHANG ET AL.

antisense STAT3 oligonucleotides can induce apoptosis of PCa (1,2,30). Third, STAT3 might be involved in CRPC development, as it not only protects PCa cells from apoptosis due to androgen deprivation (28), but may also enhance the function of androgen receptor even in the absence of androgen (6). These important oncogenic characteristics make STAT3 an excellent molecular target for PCa therapy. NL, a terpenoid limonoid isolated from neem trees and flowers (5,10), has been found to exhibit diverse pharmacological activities such as anti-feedant (45), anti-malarial (36), anti-HIV (46), antimicrobial (37), and also significant anti-cancer responses (21). The diverse anticancer effects of NL have been reported in multiple tumor types including those of breast, colorectal, brain, and liver (11,20,25,26). More interestingly, both supercritical extract of azadirachtaindica (neem) leaves as well as purified NL have also been reported to exert anti-proliferative and proapoptotic effects against PCa (35,48), but the in depth molecular mechanism(s) of its anticancer effects were not completely deciphered in these studies. Considering the critical role of STAT3 in PCa progression as discussed above, we hypothesized that NL may attenuate PCa growth through targeting STAT3 signaling cascade. Our results indeed indicate that NL can induce oxidative stress-mediated blockade of STAT3 activation that contributes to its growth and metastasis inhibitory effects in PCa.

2


ZHANG ET AL.

RESULTS NL inhibits the viability as well as reduces invasion and migration of PCa cells DU145 and LNCaP were treated with NL (0-20μM) for 24, 48 and 72h and thereafter MTT assay was performed. As shown in Fig.1B, there were significant reductions in the viability of two tumor cell lines upon NL treatment, and these decreases were both time and dose dependent. Cell invasion and migration are considered as initial steps during the process of tumor metastasis (47). Fig.1C and Fig.1D present the results obtained from wound healing and boyden chamber assays. These two assays were performed in DU145 cells stimulated with or without chemotactic agent CXCL12. We observed significant decreases in both cellular migration and invasion following NL treatment with and without CXCL12 stimulation, thereby providing evidence(s) that NL may exhibit anti-metastatic effects against PCa cells.

NL induces substantial apoptosis in PCa cells We next analyzed the potential of NL to induce apoptosis in DU145 and LNCaP using diverse molecular techniques. As cells accumulated in the SubG1 phase represent the apoptotic population (24), flow cytometry was first performed to study the pro-apoptotic effect of NL in PCa cells. Fig.2A shows a clear increase of cell population percentage in SubG1 phase following NL treatment, and this increase was time-dependent in both cell lines. DNA fragmentation is another characteristic marker of cellular apoptosis (9). From Fig.2B, it is apparent that NL increased the DNA fragmentation levels in both DU145 and LNCaP cells, in a time dependent manner. The effects of NL on Caspase-3/7 and PARP activation were also determined using luciferase assay and Western blot analysis. Fig.2C and D show that NL induced a significant

3


ZHANG ET AL.

increase in Caspase-3/7 activities, and also elevated cleavage of pro-Caspase-3 and PARP in PCa cells. Overall, these results demonstrate the apoptotic effect of NL against PCa cells.

NL inhibits STAT3 activation cascade in PCa cells We next investigated the potential effect of NL on STAT3 signaling pathway using diverse molecular biology techniques. First, we analyzed its effect on STAT3 phosphorylation by detecting phospho-STAT3 (p-STAT3) (Tyr 705) levels using Western blot analysis. We found that both constitutive p-STAT3 in DU145 and IL-6-stimulated p-STAT3 levels in LNCaP were substantially reduced upon NL treatment (Fig.3A and C). Janus kinases (JAK) and Src families are the major upstream tyrosine kinases that can regulate STAT3 activation (42). In DU145 cells, both JAK1 and JAK2 phosphorylation levels were reduced by NL, while Src phosphorylation was not affected (Fig.3B). Similar trend was also observed in LNCaP cells, in which NL reduced the phosphorylation of both JAK1 and JAK2 stimulated by IL-6 (Fig.3C). These data indicate that the STAT3 inhibition caused by NL treatment may be due to down-regulation of activation of upstream through JAK1 and JAK2 kinases. In addition, as STAT3 binds to DNA and thereby induces gene transcription after nuclear translocation, we next determined the potential effects of NL on both STAT3 DNA binding and transcriptional activities. As shown in Fig.3D and E, both constitutive and IL-6-stimulated STAT3 DNA binding ability as well as constitutive and EGFinduced luciferase reporter activities were significantly suppressed upon NL treatment, thereby demonstrating the negative effect of NL on STAT3 signaling pathway.

Antioxidants reverse NL-induced STAT3 inhibition and cellular apoptosis

4


ZHANG ET AL.

N-acetyl-L-cysteine (NAC) and glutathione (GSH) are two major thiol-related antioxidants (7,33). Interestingly, pre-treatment of these two antioxidants substantially abolished STAT3 abrogation caused by NL treatment (Fig.4A), thereby indicating that oxidative stress may be involved in the inhibitory effect of NL on STAT3 in PCa cells. Furthermore, NAC/GSH pretreatment was also found to significantly reduce cellular apoptosis induced by NL as observed by flow cytometric and Western blot analysis (Fig.4B and 4C), further demonstrating that oxidative stress may also contribute to the observed pro-apoptotic effects of NL.

GSH/GSSG imbalance contributes to NL-induced ROS production To further validate whether NL can induce oxidative stress in PCa cells, ROS levels were measured in DU145 cells by staining with H2-DCFDA. We observed a significant increase of ROS production upon NL treatment (Fig.5A), suggesting that NL can indeed induce oxidative stress in PCa cells. GSH/GSSG system is one of the major intracellular antioxidant systems (27). The ratio of GSH to oxidized glutathione (GSSG) is an indicator of cellular oxidative stress (13). To explore the possible mechanism(s) of increased ROS production, we analyzed the effect of NL on GSH/GSSG system. As shown in Fig.5B and 5C, a significant decrease of GSH and an obvious increase of GSSG were observed in DU145 cells upon treatment, thereby indicating that exposure of the cells to the drug resulted in an imbalance of GSH/GSSG system. In addition, almost two folds elevation of GSSG/GSH ratio was observed (Fig.5D), demonstrating that NL can induce oxidative stress in DU145 cells. To further investigate the role of GSH/GSSG imbalance in mediating NL-induced oxidative stress, both GSH synthesis blocker buthionine sulfoximine (BSO) and GSH prodrug NAC were employed. We found that pretreatment with NAC significantly prevented NL-induced ROS production, while BSO enhanced the ROS

5


ZHANG ET AL.

production (Fig. 5E). To further clarify the type of ROS involved, we measured the H2O2 level in NL-treated cell. As shown in Fig. 5F, a time-dependent increase of H2O2 was observed, and the increase reached to almost 2 fold compared to the control level after 4h (Fig. 5F). Similar to the general ROS, increased H2O2 production was also prevented by NAC and enhanced by BSO (Fig. 5G). These results demonstrate that H2O2 is the major ROS induced by NL in PCa cells. In addition, BSO enhanced NL-induced cellular apoptosis when applied in combination with NL, and the increased apoptosis could also be attenuated by NAC pre-treatment (Fig. 5H), providing strong evidence(s) that GSH/GSSG imbalance primarily contributes to the NL-induced oxidative stress and its observed anticancer effects in PCa cells.

NL inhibits the activity of glutathione reductase (GR) in PCa cells GR catalyzes the reduction of GSSG to GSH to resist oxidative stress (8). As we observed a significant alteration in GSH/GSSG ratio, we next investigated whether this imbalance may be due to the disturbance in GR activity. We treated the cells with NL and collected the cell homogenates to measure cellular GR activity. A time-dependent decrease of GR activity was observed (Fig.6A). However, no change was observed in GR protein expression level (Fig.6B), indicating that GR inhibition by NL is not mediated through negative regulation of its protein expression. To clarify how NL inhibits GR, we next investigated the interaction between GR and NL in a cell-free assay system. GR was incubated with NL in the presence of NADPH for 30mins, and then the substrate GSSG was added to initiate the reaction. As shown in Fig. 6C, 70% of GR activity was lost upon NL treatment. These results demonstrate that NL directly suppresses GR by inhibiting its kinetic activity.

6


ZHANG ET AL.

NL inhibits the expression of STAT3 regulated gene products STAT3 regulates the expression of various oncogenic genes involved in tumor growth, angiogenesis and metastasis (42). To analyze the effect of NL on STAT3-regulated proteins as well as genes, both Western blot analysis and real-time PCR assays were performed. As shown in Fig.7A, clear decreases were observed in the levels of various proteins such as survivin, Cyclin D1 and MMP-9 upon NL treatment, and the similar decreases were also observed in the mRNA levels (Fig.7B). These results provide evidence that NL treatment down-regulates the expression of diverse STAT3-regulated genes. Interestingly, down-regulation of these gene products could be reversed upon NAC pre-treatment (Fig.7C), thereby further demonstrating the role of oxidative stress in STAT3 inhibitory effects of NL.

NL suppresses prostatic intraepithelial neoplasia (PIN) formation in TRAMP mouse model PIN formation by 4-10 week of age has been reported in TRAMP mice model (17). To investigate whether NL suppresses PIN formation, 4-week-old mice were administered with vehicle or 3 mg/kg NL by oral gavage, 5 times a week for 8 weeks (n=8) and then sacrificed. Representative H&E staining in DLP sections from non-transgenic mice, control and NL-treated mice are shown in Fig. 8A (i). DLP from non-transgenic mice exhibited typical acini with abundant eosinophilic intralumenal secretions. Compared with non-transgenic mice, DLP from vehicle-TRAMP exhibit high incidence of PIN, but with treatment of 3 mg/kg NL, PIN was significantly reduced by 35% (Fig. 8A(ii)). In addition, none of the animals showed welldifferentiated carcinoma (WDC) and poorly differentiated carcinoma (PDC) at this stage (Fig. 8A(ii)).

NL prevents the progression from PIN to PCa in TRAMP mouse model 7


ZHANG ET AL.

To study whether NL can prevent progression from PIN to PCa, 12-week-old TRAMP-mice were treated with vehicle or 3 mg/kg NL by oral gavage, 5 times a week for 6 weeks (n=8) and then sacrificed. Representative H&E staining in DLP sections from non-transgenic mice, control and NL-treated TRAMP mice are shown in Fig. 8B (i). Compared with non-transgenic mice, DLP from vehicle-TRAMP exhibit higher incidence of PIN and WDC. NL administration resulted in a 20% decrease of WDC incidence and a modest increase (30%) of PIN incidence compared with vehicle-TRAMP (Fig.8B (ii)). These results indicated that NL significantly suppressed the progression from PIN to WDC in TRAMP mice.

NL suppresses the growth of established PCa in vivo To determine whether NL can suppress the growth of established PCa, 24-week-old TRAMP mice were treated with vehicle or 3 mg/kg NL by oral gavage, 5 times a week for 12 weeks (n=8) and then sacrificed. At this age, H&E staining revealed greater incidences of PIN, WDC and PDC in the vehicle-TRAMP mice than that in non-transgenic mice. Compared to vehicle-TRAMP mice, the DLP of the NL-treated TRAMP mice exhibited significantly lower incidence of PIN, WDC and PDC. For example, the incidence of the PIN in the DLP of NL treated mice was lower by 10% in comparison with vehicle-TRAMP mice, while the incidences of the WDC and PDC were lower by 50% and 20% respectively. Overall, these results indicate that NL administration significantly inhibited PCa progression as well as the tumor growth.

NL inhibits STAT3 phosphorylation and expression of Ki-67 in DLP of TRAMP mice We next examined the potential effect of NL on STAT3 phosphorylation in the DLP of TRAMP mice. As shown in Fig.8D, DLP from vehicle-treated TRAMP mice exhibited higher phosphorylation level of STAT3 compared with non-transgenic prostates, while NL intake 8


ZHANG ET AL.

resulted in a marked reduction of STAT3 phosphorylation. The inhibition on phosphorylation status of STAT3 in Group 3 was also observed in IHC experiments (Fig.8E). We also investigated the effect of NL on the expression of proteins involved in cancer proliferation (Ki67) and apoptosis (cleaved-Caspase-3). As shown in Fig. 8D and E, we observed an increase in cleaved-Caspase-3 level and a decrease in proliferation biomarker Ki-67 expression as compared to the control TRAMP group.

NL suppresses tumor growth and metastasis in TRAMP mouse model To investigate the effect of NL in the tumor growth and metastasis, 24-week-old TRAMP mice were treated with vehicle or 3 mg/kg NL by oral gavage, 5 times a week for 12 weeks (n=8) and then sacrificed at the 36 weeks. A significant (p