Electrophoretic mobility shift assay (EMSA)

The native gel for EMSA was prepared and electrophoresed by Mini- PROTEAN®

Tetra (Bio-Rad). A native gel was made by that 1X TBE contained 6% of acrylamide and 2.5% of glycerol, followed by adding appropriate 10% APS and TEMED in order to polymerize the gel. The κB site probe (κB site: 5’-GGGAAATTCC-3’ (Liang et al., 2006) labeled with biotin was ordered from Protech Technology Enterprise Co., LTD.

And the binding buffer and biotin detection system were from LightShift^® Chemiluminescent EMSA Kit (Thermo).

For the binding reaction, 15 μg of nuclear extract and 25 nM probe for final con-centration were mixed in binding buffer in total volume of 20 μl, and the mixtures were incubate at room temperature for 20 min. Then, the sample mixtures were loaded with 5X loading dye contained in the kit and electrophoresed in 0.5X TBE at 80 volts until the loading dye migrated approximately 3/4 down the length of the gel. Before transfer, the NC membrane was soaked in 0.5X TBE for at least 10 min. The transfer was carried out by transfer system (Mini Trans-Blot® Cell, Bio-Rad) in ice-cold 0.5X TBE buffer at 380 mA for 1 h. Subsequently, the transferred DNA was cross-linked to membrane for 20 min with the membrane face down on a transilluminator equipped with 312 nm bulbs (UV box of UNIVERSAL HOOD, Bio-Rad)

Before the detection of biotin-labeled DNA, the blocking buffer and 4X wash buff-er have to pre-warm until all particulate is dissolved. To block the membrane, mem-brane was soaked in blocking buffer for 15 min with gentle shaking, followed by transferring the membrane to anti-biotin antibody which is prepared with blocking buffer in 1:300 dilution and incubating for 15 min. After the incubation of antibody, the membrane was washed 4 times with 1X wash buffer which was diluted by ddH2O

28

each time for 5 min. And then, membrane was removed to appropriate volume of Substrate Equilibration Buffer for 5 min with gentle shaking. Finally, the membrane was transferred to substrate mixture which is prepared before used for 3-5 min, and detected by exposure to film.

29

Results

3.1. Transcription factor NF-κB inhibitor screening from marine natural prod-ucts

In a collaborative effort to discover potent NF-κB inhibitor, we screened more than 70 extracts (supplementary data 1) from a variety of marine resources provided by Dr.

Chung-Kuang Lu’s lab at National Research Institute of Chinese Medicine. Using a previously established NIH3T3 cell system (Liang et al., 2003; Liang et al., 2006), we were able to utilize TNFα as an activator of the NF-κB signaling for detecting inhibi-tors of this pathway.

Among all extracts tested, we found that WYT1-33-6, an extract from the marine bacterium Zooshikella sp., was a potent NF-κB inhibitor. At a concentration of 20 µg/ml, WYT1-33-6 effectively protected IκBα from TNFα-induced degradation in the NIH3T3 cells (supplementary data 1). Furthermore, we treated cells with increasing concentrations (0-20 µg/ml) of WYT-33-6 and found that TNFα induced IκBα degra-dation was inhibited in a dose-dependent manner (Fig. 3A). This result suggested that WYT1-33-6 can against TNFα induced signal and inhibit the degradation of IκBα.

However, this inhibition ability seems to reach the top at 10 μg/ml of WYT1-33-6 since the data showed that there was no apparent difference between IκBα ratio in cells treated with either 10 μg/ml or 20 μg/ml of WYT1-33-6 (Fig. 3B).

3.2. WYT1-33-6 inhibits NF-κB activation by blocking IκBα phosphorylation and NF-κB nuclear translocation in TNFα-induced NIH3T3 cells

Previous studies revealed that phosphorylation of IκBα N-terminal Ser residues is a prerequisite event for proteasome-mediated IκBα degradation (Hayden and Ghosh, 2004). Therefore, we were interested in determining whether WYT1-33-6 could block

30

the phosphorylation of IκBα. Results from Figure 4 showed that the amount of phos-phorylated IκBα and total IκBα in different TNFα induction time from 0 to 20 min.

The IκBα protein of the solvent control was phosphorylated within 5 min after TNFα induction (Fig. 4A, Fig. 4B), and then the phosphorylated IκBα protein was degraded by proteasome at 10 and 20 min after TNFα induction (Fig. 4A, Fig. 4C). In contrast, 20 μg/ml WYT1-33-6 treated cells presented low level of phosphorylated IκBα and constant amount of IκBα (Fig 4). However, the 20-min TNFα induction in the solvent control (Figure 4A) showed a rising IκBα expression which is contrary to the previous data (Fig. 3). Indeed, the IκBα upregulation supports the general understanding that IκBα is resynthesized once NF-κB signaling pathway was activated and can be ex-plained by delay of harvesting. Overall, it was proved that WYT1-33-6 blocks the degradation of IκBα through repressing the phosphorylation of IκBα.

NF-κB nuclear translocation is a downstream event following IκBα degradation in the classical TNFα-induced NF-κB activation (Hayden and Ghosh, 2004). Because our data showed that TNFα-induced IκBα phosphorylation and degradation were blocked by WYT1-33-6 (Figures 3 and 4), we then tested whether WYT1-33-6 could inhibit the NF-κB subunit p65 nuclear translocation in the 3T3 cells. The result showed that the amount of the nuclear p65 protein in the 20 μg/ml WYT-33-6 treated cells was significantly less than that of the solvent control in the TNFα-activated NIH3T3 cells (Fig. 5). In summary, this result is consistent with our previous data and suggests that WYT-33-6 is a potent inhibitor of the NF-κB activation by blocking IκBα phosphorylation and degradation as well as NF-κB nuclear translocation.

3.3. WYT1-33-6 inhibits cell viability and induces apoptosis in the human cer-vical cancer HeLa cells

Our research data suggested that WYT-33-6, the extract from the marine bacterium

31

Zooshikella sp., is an effective inhibitor of the NF-κB activation in TNFα-induced

mouse fibroblast 3T3 cells. We then examined whether WYT-33-6 could also inhibit NF-κB activation in human cancer cell lines. We first tested the frequently used hu-man cervical cancer cell line HeLa and found that WYT-33-6 did not block TNFα-induced NF-κB activation in this particular cell line (data not shown). This re-sult indicates that the NF--κB inhibition effect of WYT-33-6 is cell line-dependent.

We next studied whether WYT1-33-6 could exert cell toxicity in HeLa cells. Cells were treated with increasing concentrations of WYT1-33-6 for 24 h and the cellular morphological changes were examined using microscopy. Cycloheximide, the protein biosynthesis inhibitor in eukaryotic organisms, was used as a control of cell toxicity.

After 24 h incubation, HeLa cells treated with 10 μg/ml of WYT1-33-6, compared with the solvent control, had a dramatic decrease in cell number and showed cell shrinkage in size (Fig. 6A). On the other hand, cycloheximide treated control also had cell shrinkage morphology. Interestingly, WYT1-33-6 treated cells showed a deep red color, the same color of the WYT1-33-6 (6A).

In addition to cellular morphological study, we assessed cell viability using MTT assay in WYT1-33-6 treated HeLa cells. After 24 h of drug incubation, the cell growth inhibition effect of WYT1-33-6 was determined by the MTT assay. The 50%

of cell growth inhibition (IC₅₀) of WYT1-33-6 was approximately 15μg/ml in HeLa cells (Fig. 6B).

Because WYT1-33-6 inhibited cell growth as judged by MTT assay, we wanted to determine whether this agent could also induce apoptosis in HeLa cells.

Cycloheximide was used as a positive apoptosis control. Poly (adenosine 5’-diphosphoateribose) polymerase (PARP) cleavage was used as an apoptosis marker.

Same amount of HeLa cells were seeded and treated with different concentrations of WYT1-33-6 for 24 h and cell lysate was subjected to Western blots for PARP

cleav-32

age. The results revealed that WYT1-33-6 dose-dependently induced PARP cleav-age.(Fig. 7). In summary, our studies showed that WYT1-33-6 is a potent agent in in-hibiting cell growth and inducing apoptosis in HeLa cells.

3.4. Major bioactive components of WYT1-33-6, WYT170-6 (prodigiosin) and WYT1-70-10 (a prodigiosin analog), are cytotoxins in multiple myeloma cells To take a further study in WYT1-33-6, our collaborator Dr. Chung-Kuang Lu and his lab members separated WYT1-33-6 into 11 fractions (WYT1-49-1 to WYT1-49-11). In order to examine which fraction contributed most to the bioactivity found in WYT1-33-6, we took the NF-κB inhibitor screening procedures previously mentioned in the TNFα induced 3T3 cells. The results suggested that two fractions, WYT1-49-6 and WYT1-49-8, could inhibit IκBα degradation (Fig. 8A). Also, results from cell growth inhibition effect assessed by MTT and MTS assay in several cancer cell lines supported the suggestion that these two were the most bioactive fractions in WYT1-33-6 (supplementary data 2). Dr. Lu and his students further purified WYT1-49-8 and obtained 2 compounds, prodigiosin (WYT1-70-6) and its analog 2-methyl-3-heptyl prodigiosin (WYT1-70-10) (Fig. 8B). Moreover, WYT1-49-6 was also separated into 6 fractions (WYT1-96-1 to WYT1-96-6). However, only one frac-tion, WYT1-96-4, had NFκB inhibitory effect in the TNFα induced 3T3 cells (data not shown). Our studies later confirmed that the main effective component of WYT1-96-4 was also prodigiosin (Appendix 2.). For this reason, the following studies were focused on two bioactive compounds WYT1-70-6 (prodigiosin) and WYT1-70-10 (a prodigiosin analog).

Not only the high connection between NF-κB signaling pathway and multiple my-eloma but the incurable property and low 5 years survival rate, we selected the multi-ple myeloma cell line, RPMI 8226, as the target for our research. As precious

proce-33

dure described, we first examined whether WYT1-70-6 and WYT1-70-10 could in-hibit TNFα induced IκBα degradation in 3T3 cells. Our results showed that WYT1-70-6 mediated NF-κB inhibition in TNFα-induced RPMI 8226 cells in a dose-depend manner (Fig. 9), whereas the WYT1-70-10 displayed no NF-κB inhibi-tion (Fig. 10). Interestingly, these two compounds were both effective in inducing apoptosis and suppressing cell growth in the multiple myeloma RPMI 8226 cell line (Fig. 13). Both compounds began to induce PARP cleavage at the concentration of 2.5 μM (Fig. 11, Fig. 12). Moreover, the IC50 value for both compounds were approxi-mately 0.5 μM as judged by MTS assay (Fig. 13). Although WYT1-70-10 could in-duce apoptosis and inhibiting cell viability in RPMI 8226 cells, we were unable to determine its molecular target yet. Therefore, we decided to focus our studies on the NF-κB inhibitor WYT1-70-6 (prodigiosin).

3.5. The DNA binding ability of NF-κB is inhibited by WYT1-70-6 through blocking NF-κB translocation in human multiple myeloma RPMI 8226 cells Our previous study showed that WYT1-70-6 (prodigiosin) blocked the degradation of IκBα in TNFα induced human multiple myeloma RPMI 8226 cells (Fig. 9). We next examined whether NF-κB nuclear translocation could be blocked by WYT1-70-6 as judged by Western blot for the NF-κB subunit p65 using cell nuclear extracts. The result showed that the level of p65 protein in the nuclear extracts of WYT1-70-6 treated RPMI 8226 cells was significantly decreased compared with that of the sol-vent control (Fig. 14). The result suggested that WYT1-70-6 effectively blocks the nuclear translocation of the NF-κB subunit p65.

Although our studies demonstrated that p65 translocation was inhibited by WYT1-70-6 in RPMI 8226 cells, the function of NF-κB which binds to specific se-quence of DNA (κB site) and turns on the downstream genes had not been examined.

34

Therefore, we used electrophoretic mobility shift assay (EMSA), a widely applied to determine the interaction between the NF-κB complex and DNA (κB site), to confirm whether this binding ability could be repressed due to a decrease in the level of NF-κB complex in the nucleus. As shown in Fig. 15, The NF-κB DNA binding activ-ity of TNFα induced RPMI 8226 cells were inhibited by WYT1-70-6. In contrast, the TNFα induced solvent control displayed a strong NF-κB DNA binding activity. Taken together, these results suggested that WYT1-70-6 specifically inhibits the classical NF-κB signaling pathway through blocking the phosphorylation of IκBα in the human multiple myeloma RPMI 8226 cells.

3.6. Evaluating the cell killing effect of WYT1-70-6 in combination with clinical agents for treating multiple myeloma

Doxorubicin is a clinically used chemo drug for treating multiple myeloma. In ad-dition, doxorubicin is also widely used for treating an array of cancer types, including bladder, breast, and the lung. We thus assessed the cell killing effect of combination therapy using WYT1-70-6 and doxorubicin in the multiple myeloma RPMI 8226 cells.

The combination of these two agents displayed a more potent cytotoxicity than either of them used alone (Fig. 16A). However, these two agents did not exhibit significant additive interaction according to the isobologram analysis (Steel and Peckham, 1979).

In contrast, the proteasome inhibitor bortezomib, is a FDA-approved drug for treating multiple myeloma. The recent research revealed that bortezomib actually ac-tivates the NF-κB signaling pathway (Hideshima et al., 2009). The studies indicated the fact that bortezomib-induced cytotoxicity is not associated with NF-κB inhibition in multiple myeloma. Therefore, we assumed a combination therapy using bortezomib and WYT1-70-6 would have a synergistic cell killing effect. The result showed that this combination treatment displayed an additive interaction as judged by MTS assay

35

(Fig. 16B). To examine this additive interaction between bortezomib and WYT1-70-6, we harvested cell lysates treated for 8 h with bortezomib alone or in combination in the RPMI 8226 cells. As shown in Figure 17, cells treated with bortezomib (lane 2) exhibited lower protein level of IκBα than that of the solvent control (lane 1). Moreo-ver, results from the Western blots of the nuclear p65 were consistent with this obser-vation (Fig. 17B). Our data suggested that bortezomib induced the degradation of IκBα and consequently the nuclear translocation of p65 in RPMI 8226 cells. On the other hand, bortezomib in combination with WYT1-70-6 inhibited NF-κB activation (Fig 17A, lane 3) in the same cell line. Taken together, this immunoblotting result supports the additive interaction between bortezomib and WYT1-70-6 in the MTS assay and suggests a more effective therapeutic option by combination therapy using bortezomib and WYT1-70-6 in multiple myeloma.

36

Discussion

4.1. Marine microorganism Zooshikella sp., a new source of prodiginine family In this study, we demonstrated a natural product, prodigiosin, extracted from the marine microorganisim Zooshikella sp. and exhibiting NF-κB inhibitor and anticancer property through a series of experiments. This new species of the genus Zooshikella is gram-negative marine bacterium which was identified as a new source of prodiginine family, including two structures which were unknown before (Lee, Kim et al. 2011).

4.2. Prodigiosin has gotten attention again as an anticancer agent in recent years Although the discovery of prodigiosin was very early, it was regarded as antibiotics at that time (Lichstein and Van De Sand 1946). During these two decades, scientists have perceived that prodigiosin may have other applications. The new characteristics was identified that prodigiosin represses the growth of lymphocytes, especially T-cells, which displays a potential to be an immunosuppressant (Nakamura, Nagai et al. 1986; Magae, Yamashita et al. 1993; Han, Kim et al. 1998). Furthermore, the studies from 2000s revealed that prodigiosin can induce apoptosis in cancer cell lines (Montaner and Perez-Tomas 2001; Montaner and Perez-Tomas 2002; Llagostera, Soto-Cerrato et al. 2003). However, most of these studies only indicated that prodigiosin results in apoptosis through caspase-dependent pathway. In recent re-searches, more characteristics of prodigiosin have been clarified, including antiproliferation, inducing DNA damage and inhibiting the function of topoisomerase I and II (Montaner, Castillo-Avila et al. 2005; Hsieh, Shieh et al. 2012). Thus, those recent studies suggest that prodigiosin induces apoptosis through modulating several pathways as well as damaging DNA directly. And results from this thesis study also support the previous researches about the characteristics of prodigiosin.

37

4.3. Prodigiosin inhibits NF-κB signaling pathway in cancer cells

Previous studies on prodigiosin associated with NF-κB signaling pathway were fo-cused on primary culture of immunocytes for demonstrating its immunosuppressive property (Mortellaro, Songia et al. 1999; Huh, Yim et al. 2007). However, there were no reports that prodigiosin inhibits the activation of NF-κB in cancer cells. The dose of prodigiosin in those immunosuppression studies was usually low to nanomolar scale. Compared to those studies, the dose used in this study was relatively high for demonstrating the NF-κB inhibitory effect (Fig. 9, Fig. 15). This different dose re-quired for NF-κB inhibitoin could be resulted from the differences in cell type since the inhibitory concentrations 50 (IC₅₀) of MTT and MTS are consisted with other published results. Moreover, our results demonstrated that prodigiosin blocks the phosphorylation of IκBα, resulting in the inactivation of NF-κB signaling pathway (Fig. 4, Fig. 9, and Fig. 15). In addition, the EMSA data suggested that prodigiosin specifically inhibited the classical NF-κB signaling pathway. All these characteristics suggested that prodigiosin might be an IKKβ inhibitor.

4.4. Prodigiosin in combination with bortezomib exert synergistic cell toxicity in human multiple myeloma cells

Current chemotherapy seldom uses one drug alone. In most cases, doctor usually gives two or three chemodrugs for combination since most of cancers display drug resistance after a period of time. In this study, we demonstrated that prodigiosin is an effective agent as a chemotherapy candidate in multiple myeloma. In fact, our results suggested that prodigiosin induces growth inhibition and apoptosis in several other cancer types, including small cell lung cancer (SCLC) and none-small cell lung can-cer (NSCLC) (data not shown). Nevertheless, the TNFα induction system did not

38

work well in both of SCLC and NSCLC cell lines. Thus, we were unable to determine whether prodigiosin is an NF-κB inhibitor in the lung cancer cell line studies. The 26S proteasome inhibitor bortezomib, has exhibited good curative effect on multiple mye-loma. However, there were cases reported that some patients were resistance to bortezomib due to activation of the NF-κB pathway (Markovina, Callander et al.

2010). Additionally, the research suggested that bortezomib-induced apoptosis is not related to the repression of NF-κB activity. On the contrary, the studies revealed that bortezomib would activate NF-κB through classical pathway (Hideshima, Ikeda et al.

2009). The same group reported that an IKKβ inhibitor in combination with bortezomib overcame bortezomib resistance (Hideshima, Ikeda et al. 2009). In this thesis study, we demonstrated that prodigiosin suppressed the NF-κB activity induced by bortezomib in a human multiple myeloma cell line. This result supports our hy-pothesis that prodigiosin might be an IKKβ inhibitor.

For chemotherapy, doxorubicin is a common drug used in many types of cancers, including bladder, breast, lung, and multiple myeloma. It interacts with DNA by in-tercalation and inhibition of molecular biosynthesis; thus, it usually accompanies cardiotoxicity in higher dosage. Therefore, a combination with other chemodrugs is a good solution to reduce the dosage of doxorubicin. Our data showed that prodigiosin in combination with doxorubicin were not effective in treating multiple myeloma cells.

On the other hand, doxorubicin in combination with histone deacetylase inhibitor dis-played a synergistic effect (Sanchez, Shen et al. 2011).

39

Reference

Adachi, O., Kawai, T., Takeda, K., Matsumoto, M., Tsutsui, H., Sakagami, M., Nakanishi, K., and Akira, S. (1998). Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function. Immunity 9, 143-150.

Aggarwal, B. B. (2003). Signalling pathways of the TNF superfamily: a double-edged sword. Nature reviews Immunology 3, 745-756.

Aggarwal, B. B., and Gehlot, P. (2009). Inflammation and cancer: how friendly is the relationship for cancer patients? Current opinion in pharmacology 9, 351-369.

Aggarwal, B. B., Vijayalekshmi, R. V., and Sung, B. (2009). Targeting inflammatory pathways for prevention and therapy of cancer: short-term friend, long-term foe.

Clinical cancer research : an official journal of the American Association for Cancer Research 15, 425-430.

Ahn, K. S., Sethi, G., and Aggarwal, B. B. (2007). Nuclear factor-kappa B: from clone to clinic. Current molecular medicine 7, 619-637.

Amado, N. G., Fonseca, B. F., Cerqueira, D. M., Neto, V. M., and Abreu, J. G. (2011).

Flavonoids: potential Wnt/beta-catenin signaling modulators in cancer. Life sciences 89, 545-554.

Anand, P., Kunnumakkara, A. B., Sundaram, C., Harikumar, K. B., Tharakan, S. T., Lai, O. S., Sung, B., and Aggarwal, B. B. (2008). Cancer is a preventable disease that requires major lifestyle changes. Pharmaceutical research 25, 2097-2116.

Annunziata, C. M., Davis, R. E., Demchenko, Y., Bellamy, W., Gabrea, A., Zhan, F., Lenz, G., Hanamura, I., Wright, G., Xiao, W., et al. (2007). Frequent engagement of the classical and alternative NF-kappaB pathways by diverse genetic abnormalities in multiple myeloma. Cancer cell 12, 115-130.

Basseres, D. S., and Baldwin, A. S. (2006). Nuclear factor-kappaB and inhibitor of kappaB kinase pathways in oncogenic initiation and progression. Oncogene 25, 6817-6830.

Beg, A. A., Sha, W. C., Bronson, R. T., Ghosh, S., and Baltimore, D. (1995).

Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature 376, 167-170.

Bonizzi, G., and Karin, M. (2004). The two NF-kappaB activation pathways and their role in innate and adaptive immunity. Trends in immunology 25, 280-288.

Cheng, G., and Xie, L. (2011). Parthenolide induces apoptosis and cell cycle arrest of

40

human 5637 bladder cancer cells in vitro. Molecules 16, 6758-6768.

Coussens, L. M., and Werb, Z. (2001). Inflammatory cells and cancer: think different!

The Journal of experimental medicine 193, F23-26.

D'Incalci, M., and Galmarini, C. M. (2010). A review of trabectedin (ET-743): a unique mechanism of action. Molecular cancer therapeutics 9, 2157-2163.

Dejardin, E., Droin, N. M., Delhase, M., Haas, E., Cao, Y., Makris, C., Li, Z. W., Karin, M., Ware, C. F., and Green, D. R. (2002). The lymphotoxin-beta receptor induces different patterns of gene expression via two NF-kappaB pathways.

Dejardin, E., Droin, N. M., Delhase, M., Haas, E., Cao, Y., Makris, C., Li, Z. W., Karin, M., Ware, C. F., and Green, D. R. (2002). The lymphotoxin-beta receptor induces different patterns of gene expression via two NF-kappaB pathways.