A novel proteasome inhibitor suppresses tumor growth via targeting both 19S proteasome deubiquitinases and 20S proteolytic peptidases

The successful development of bortezomib-based therapy for treatment of multiple myeloma has established proteasome inhibition as an effective therapeutic strategy, and both 20S proteasome peptidases and 19S deubiquitinases (DUBs) are becoming attractive targets of cancer therapy. It has been reported that metal complexes, such as copper complexes, inhibit tumor proteasome. However, the involved mechanism of action has not been fully characterized. Here we report that (i) copper pyrithione (CuPT), an alternative to tributyltin for antifouling paint biocides, inhibits the ubiquitin-proteasome system (UPS) via targeting both 19S proteasome-specific DUBs and 20S proteolytic peptidases with a mechanism distinct from that of the FDA-approved proteasome inhibitor bortezomib; (ii) CuPT potently inhibits proteasome-specific UCHL5 and USP14 activities; (iii) CuPT inhibits tumor growth in vivo and induces cytotoxicity in vitro and ex vivo. This study uncovers a novel class of dual inhibitors of DUBs and proteasome and suggests a potential clinical strategy for cancer therapy.

hypothesized that copper or zinc complex inhibits the proteasome by a mechanism that is distinct to the clinically used drug bortezomib, targeting the 19S rather than the 20S proteasome 7,11 , but this hypothesis has not been tested. In humans, three deubiquitinases (DUBs) are associated with the 19S regulatory particles. Two of these, UCHL5/Uch37 and USP14/Ubp6, are cysteine proteases and belong to the ubiquitin C-terminal hydrolases (UCH) and ubiquitin specific proteases (USP) families, respectively. The third is POH1 (RPN11), a Zn 21 -dependent protease of the JAB1/MPN/Mov34 (JAMM) family localized within the lid of the 19S proteasome 14 . The physiological roles of the 19S DUBs are not completely clear even though some studies have reported the important roles of proteasome DUBs in protein degradation [15][16][17][18] .
Pyrithione (PT) possesses excellent metal chelating properties [19][20][21] . In this current work, the effects of copper, PT and their complex, copper pyrithione (CuPT), on UPS function and cell viability were investigated. We provide direct evidence that CuPT inhibits the UPS via inhibiting both proteasome-specific DUBs and 20S proteasome peptidases and this inhibition plays an important role in CuPTmediated cytotoxicity, unveiling a novel mechanism for the anticancer effects of metal-containing compounds.

Results
PT and CuCl 2 in combination synergistically enhanced cytotoxicity. We first investigated the cytotoxic effects of PT plus copper on cancer cells. At 24 hours after treatment, cell viability detected by the MTS assay was not discernibly affected by CuCl 2 alone, modestly decreased by PT alone, but dramatically decreased by 251 PT/CuCl 2 combination treatment with IC 50 values of 0.175, 0.125, 0.25, and 0.05 mM in MCF-7, HepG2, U266 and NCI-H929 cancer cell lines, respectively (Figures 1a and b). Also, compared to PT or CuCl 2 alone, the PT/CuCl 2 combination treatment induced cell death more effectively. This is evidenced, for example, by the result from 24 hour treatment of U266 cancer cells, followed by live cell propidium iodide (PI) staining ( Figure 1c) and by Annexin V/PI staining followed by flow cytometry (Figure 1d). Similarly, PT/ CuCl 2 treatment for 24 hours induced high levels of PI-positivity in MCF-7 breast cancer cells, compared to PT or copper alone ( Figure 1e) and such a treatment for 12 hours also induced PARP cleavage and decreases of full-length caspase 8 and caspase 9 ( Figure 1f). These results demonstrate that the combination of PT and CuCl 2 induces cytotoxicity in multiple cancer cell lines much more effectively than PT or CuCl 2 alone.
PT and H 2 O 2 in combination synergistically enhanced cytotoxicity. Since CuCl 2 is a strong oxidant, here we used another oxidant H 2 O 2 instead of CuCl 2 in combination with PT to investigate their cytotoxic effect in cancer cells. U266 cancer cells were treated with PT, H 2 O 2 alone and their combination at the indicated doses for 24 hours. The enhanced decrease of cell viability was observed with the treatment of PT combining with H 2 O 2 at the doses of 25 and 50 mM but not at the low dose of 12.5 mM (Figure 2a); cell death was dramatically accerelated with the combination treatment of PT and H 2 O 2 (50 mM) as detected by recording the PI-positive cells under a fluorescence microscope (Figure 2b) or by flow cytometry with Annexin V/PI staining (Figure 2c). These results clearly show that PT and H 2 O 2 in combination enhanced cytotoxicity. However, whether PT 1 H 2 O 2 uses the identical mechanism of action as that of PT 1 CuCl 2 needs to be further investigated. Indeed, we found that PT 1 CuCl 2 , but not PT 1 H 2 O 2 induced inhibition of the UPS (see below).
CuPT, the chelating product of PT and CuCl 2 , induced cytotoxicity in multiple cancer cells and in primary monocytes from patients with acute myeloid leukemia (AML). CuPT is formed from the combining of one copper with 2 PTs. We then further investigated the effect of CuPT, the complex of PT and CuCl 2 , on cytotoxicity in cancer cell lines. We found, CuPT treatment for 24 hours decreased cell viability (IC 50 : 0.375 mM, Figure 3a) and induced cell death (Figure 3b) in MCF-7 cells; in U266 cells, CuPT treatment for 24 hours decreased cell viability with an 0.130 mM of IC 50 value ( Figure 3c) and induced apoptosis as detected with Annexin V/PI staining by flow cytometry (Figure 3d); similarly, CuPT treatment also decreased cell viability (IC 50 : 0.495 mM, Figure 3e) and induced apoptosis (Figure 3f) in HepG2 cancer cells.
The ex vivo cytotoxic effects of CuPT on bone marrow cells obtained from patients with AML and peripheral blood mononuclear cells from healthy volunteers (CTR) were also evaluated. CuPT and bortezomib decreased cell viability of primary monocytes from AML patients (AML) with an average IC 50 value 57.03 and 20.50 nM, respectively, while in normal controls (CTR) the average IC 50 values were estimated at 101.08 and 74.23 nM, respectively ( Figure 4a). CuPT treatment for 12 hours at doses ranging from 0.25 to 0.75 mM resulted in significant apoptosis in the monocytes from AML patients as detected with Annexin V/PI staining by flow cytometry (Figure 4b) or by fluorescence microscope (Figure 4c). These results demonstrate the in vitro inhibitory effect of CuPT on cancer cells.
PT plus CuCl 2 but not H 2 O 2 synergistically induces the inhibition of the UPS. We next investigated the effect of the CuPT on the UPS. In both MCF-7 and U266 cells, the combination treatment with PT and CuCl 2 (251) for 12 hours induced high levels of ubiquitinated proteins compared with PT and CuCl 2 treatment alone ( Figure 5a). In GFPu-HEK-293 cells stably harboring GFPu, a surrogate proteasome substrate, the CuPT treatment for 12 hours accumulated high levels of GFPu protein as detected by either Western blot (Figure 5b, upper) or by fluorescence microscope (Figure 5b, lower). We next detected the effects of PT plus CuCl 2 on proteasome CT-like peptidase activity in situ. It was found that in U266 and NCI-H929 cancer cells, PT alone could moderately and the combination of PT/CuCl 2 (251) dramatically inhibited the CT-like activity, while CuCl 2 alone could not inhibit the CT-like peptidase activity ( Figure 5c); in MCF-7 and SMMC-7721 cancer cells, PT could cause a dose-dependent inhibition of CT-like activity in the presence of 0.5 mM CuCl 2 ( Figure 5d). The direct effect of PT plus CuCl 2 on a purified 20S proteasome was further measured in vitro. PT, CuCl 2 (less than 0.5 mM) and their combination did not have any inhibitory effect on CT-like activity in vitro ( Figure 5e). 1 mM of CuCl 2 and its combination with PT could inhibit CT-like activity to some extent. Contrary to combining with CuCl 2 , the combination of PT and H 2 O 2 did not induce the accumulation of ubiquitinated proteins (Figure 5f).
CuPT induces UPS malfunction. We next tested whether UPS inhibition induced by the combination of PT and copper relied on the formed product, CuPT. To do so, CuPT was synthesized in our laboratory and used for the below experiments. Endogenous proteasome substrate proteins were first detected in human U266 and MCF-7 cancer cells to assess the effect of the synthetic CuPT on the UPS. We found that 12 h CuPT treatment at 0.5, 0.75 and 1.0 mM and 0.5 mM CuPT treatment for 12, 18 and 24 hours dramatically induced the accumulation of ubiquitinated proteins in MCF-7 and U266 cell lines (Figure 6a and b). In HepG2 cells, CuPT induced the accumulation of both total and K48-linked ubiquitinated proteins ( Figure 6c). Additionally, CuPT markedly increased the accumulation of a surrogate proteasome substrate, GFPu, and ubiquitinated proteins in a stable GFPu-HEK293 cell line (Figure 6d). In primary cells from either AML patients or normal controls, CuPT induced the accumulation of ubiquitinated proteins, which are comparable to 50 nM bortezomib treatment (Vel; Figure 6e). These results support that synthetic CuPT, similar to the mixture of PT and copper, could inhibit the UPS. Further results found that CuPT   could dose-dependently inhibit CT-like activity but not caspase-like and trypsin-like activities, as detected by an in situ peptidase assay in live HepG2 and U266 cells (Figure 6f), and that CuPT only at high doses (.1 mM) inhibited the CT-like but not caspase-like and trypsin-like activities of purified 20S proteasome under cell-free condition (Figure 6g).
CuPT inhibits proteasome deubiquitinase function. To determine whether CuPT could target and inhibit 19S DUBs, we first performed a computational study to predict the docking affinity between CuPT and the 19S DUBs. Previous studies revealed that the catalytic core in the active site of USP14 is formed by Cys113, His434 and Asp450 22 , and that of UCHL5 is formed by Cys88, His164 and Asp179 23 . The chemical structures of CuPT (L 1 ) and its mediated product (L 2 ) were shown in Figure 7a. These docking results suggest that compound L 2 could bind to the active sites of the cysteine-containing DUBs (USP14, UCHL5), with CDOCKER Interaction Energy of 240.63 and 239.11 kcal/mol, respectively. Accordingly, the binding modes of compound L2 at the active sites of the two DUBs were displayed in Figure 7a. It is noted that, in active sites, the Cu 21 of compound L 2 can form networks of coordination bonds with the DUBs. In the binding site of USP14, the side chains of His434, Asp450 and Asp451 coordinate to Cu 21 with distances of 3.385 Å , 2.320 Å and 2.400Å , respectively. Similarly, in the binding site of UCHL5, there are three coordination bonds between the Cu 21 and three key side chains, His164, Phe165 and Asp179, with corresponding bond lengths of 2.460 Å , 2.229 Å and 2.484 Å . At the same time, the S atom of compound L 2 forms two weak hydrogen bonds with Cys88 and Gly180. Because of the strongly binding between compound L 2 and the catalytic cores of the DUBs, it would be expected that the catalysis of the DUBs would be inhibited in the presence of compound L 1 . Since there is no crystal structure of POH1 available, we used AMSH (a JAMM domain-containing protein) instead of POH1 to dock L 2 compound. It was found that L 2 could also bind with the active sites of AMSH, with CDOCKER Interaction Energy of 235.60 kcal/mol (data not shown).
Next we performed biochemical experiments to verify the computational docking results. The activities of DUBs from either cell lysates or 26S proteasome were detected. As shown in Figure 7b, therapeutic dose of CuPT (0.5 mM) yielded inhibitory effect on cell lysate DUB activity to some extent but not as strongly as N-ethylmaleimide (NEM, 2 mM); however, similar to NEM, CuPT (0.5 mM) could completely inhibit the DUB activities of a purified 26S proteasome ( Figure 7c). The cleavage of tetraubiquitin chains (Ub4) mediated by 26S proteasome DUBs was also tested to further confirm this effect. K48-liked Ub chains were cleaved in the presence of 26S proteasome and this was efficiently blocked by CuPT (0.5 and 1 mM) (Figure 7d). DUB labelling test was further performed by using UbVS. UbVS is a potent inhibitor against UCHL5 and USP14. We found that CuPT was able to compete with UbVS's binding with both UCHL5 and USP14 in a dose-dependent manner ( Figure 7e). These results suggest that CuPT could target 19S DUBs, UCHL5 and USP14.
To further investigate this possibility, we knockdowned each DUB subunit in selected cancer cell lines and then determine whether that could affect the efficacy of CuPT. GFPu-HEK-293 cells were first transfected with siRNA against POH1, UCHL5 or USP14, and then GFPu protein level was detected. As shown in Figure 7f, POH1 knockdown highly induced accumulation of GFPu proteins (lanes 3 vs. 1), USP14 knockdown decreased GFPu protein level (lanes 5 vs. 1) and UCHL5 knockdown did not dramatically affect GFPu level (lanes 7 vs. 1). Importantly, when the above cells were treated with CuPT, the induced GFPu protein accumulation was inhibited in USP14-and UCHL5-knockdowned cells (lanes 6 and 8 vs. 2) which was not apparent in POH1 knockdowned cells. Next HepG2 cells were transfected with equal amount of siRNAs of the three DUBs for 48 hours and the effect of CuPT on K48-linked ubiquitinated protein accumulation was investigated in the DUB-silenced cells. As shown in Figure 7g, the DUB proteins including POH1, UCHL5 and USP14 were highly downregulated with transfection of each siRNA. We found that CuPT-induced K48-linked ubiquitinated protein accumulation was more inhibited in POH1 knockdowned cells than in USP14 and UCHL5 knockdowned cells (Figure 7h). It was further confirmed that knockdown of POH1 but not UCHL5 and USP14 lead to the disassembly of the 19S proteasome from the 20S proteasome (data not shown), consistent to previous report 24 . CuPT inhibits tumor growth and the UPS in vivo. We further evaluated the effect of CuPT in vivo using nude mouse HepG2 and NCI-H929 xenograft models. In HepG2 xenograft model, body weight gradually increased in the control group while kept relatively stable in CuPT-treated group (Figure 8a and b); tumor volume gradually decreased and the tumor weight was significantly reduced in the CuPT treatment group compared to the vehicle control (Figure 8c-e); the immunostaining results showed that the proteasome substrate levels such as total or K48-linked ubiquitinated and p21 proteins were significantly increased in the CuPT-treated tumor tissues (Figure 8f). In NCI-H929 xenograft model, tumor weights were reduced and proteasome-related substrates including total or K48-linked ubiquitinated proteins and p21, p27, Bax, and IkB-a proteins were all significantly increased in the CuPT-treated tumor tissues (Figure 8h). Together, the results showed that CuPT inhibited both tumor growth and the UPS function in vivo, consistent to the results in vitro.
It has been found previously that, like other established proteasome inhibitors, copper-binding compounds were only effective in inducing apoptosis in tumor, but not in non-transformed cells, providing great promise in the future clinical trials 6,12 . Like dithiocarbamates (DTCs) or clioquinol, PT spontaneously binds with copper salt thus forming an apoptosis and cytotoxic inducer 12 . The combination of PT and CuCl 2 but not each alone synergistically induced cell death and cytotoxicity with IC 50 values less than 0.5 mM in multiple cancer cell lines, like in NCI-H929 cancer cells as low as 50 nM. It is known that one of the major reactive products of PT and CuCl 2 is the formation of 2,29-dipyridyldisulfide, (PS) 2 , via the oxidation of PT 26 ; the other major product is the formation of a chelating product, CuPT. Here we found that the two reactive products both contributed to cytotoxicity induced by the combination treatment of PT and CuCl 2 . According to a previous report 27 , the oxidative product of PT, (PS) 2 , is relatively less toxic than the chelating product of PT, CuPT. Furthermore, it was found that both CuPT and 2,29-dipyridyldisulfide [(PS) 2 ] are toxic to marine polychaete Perinereis nuntia, but (PS) 2 is much less toxic than CuPT, indicating that CuPT is possibaly the major toxic product from the combination of PT and CuCl 2 .
Similar to the established cancer cell lines, CuPT also induced cytotoxicity in primary cancer cells from AML patients with IC 50 value averaging 57.03 nM, implying its potential applications in clinical conditions. In the current study, bortezomib treatment decreased cell viability with an average of IC 50 value 20.50 nM in AML cancer cells and 74.23 nM in normal peripheral monocytes, indicating that bortezomib more selectively killed cancer cells. Similar to bortezo- mib, CuPT has the same tendency to selectively kill AML cancer cells.
In nude mouse xenograft model, CuPT selectively inhibited tumor growth with relatively low toxic effect on body weight. These results demonstrated that CuPT contributed to PT/CuCl 2 combinationinduced cytotoxicity and CuPT is a promising metal-containing drug candidate as regard to cytotoxicities, providing a great potential in future clinical cancer therapy.
Similar to other copper chelating agents, the combination of PT and CuCl 2 forms a proteasome inhibitor. PT and CuCl 2 combination induced the accumulation of ubiquitinated proteins and GFPu, a surrogate proteasome substrate, but PT and H 2 O 2 combination did not induce the accumulation of ubiquitinated proteins, indicating that the oxidative product of PT did not induce UPS malfunction. Further studies confirmed that CuPT, similar to the combination of PT and CuCl 2 , not only induced the accumulation of endogenous and exogenous proteasome substrates in cultured cell lines but also accumulated ubiquitinated proteins in cancer cells from AML, which is comparable to bortezomib treatment. The effective doses used in the current study for UPS inhibition are mostly less than 1 mM, implying that PT is a very potent metal chelator as most of the reported copper chelating agents (like Schiff base, clioquinol, disul-firam, NCI-109268, dithiocarbamates and 8-hydroxylquinoline) inhibit the UPS at doses higher than CuPT and PT/CuCl 2 complex used in the reported conditions 6,7,9,10,12,13,28,29 .
We and others have reported that copper could directly inhibit 20S proteasome peptidase activities. In a test tube, PT itself did not have any inhibitory effect on proteasome CT-like activity, consistent to other report 30 ; CuCl 2 could directly inhibit CT-like activity with an IC 50 value 3-5.3 mM 6,12 ; the combination of PT/CuCl 2 or CuPT selectively inhibited CT-like activity but not caspase-like and trypsin-like activities with a similar level of IC 50 value , 6.0 mM (data not shown). These results demonstrated that PT/CuCl 2 or CuPT only directly inhibit 20S proteasome activities at much higher doses than their cytotoxic doses in the cell. Therefore, we investigated the effects of CuPT or PT/CuCl 2 combination on 26S proteasome peptidase activities in live cells. As found, CuPT or PT/CuCl 2 combination strongly inhibit proteasome peptidase activities with IC 50 value less than 1 mM, like in U266 and NCI-H929 cells as low as less than 0.25 mM, a 24 , fold decrease compared with the IC 50 value in the test tube, consistent to our previous findings 13 . The mechanism by which CuPT increases more CT-like activity inhibition within the cell than in the test tube remains unknown. We proposed that one of     the possibilities is due to copper binding to proteasome b5 subunit in the cell as previous reported 6,12 and the second possibility is due to some cell death affecting the peptidase activity. The exact mechanisms needs to be further investigated.
In the current study, one of the most important findings is that 19S deubiquitinases are potential molecular targets of CuPT. UCHL5 and USP14 belong to the cysteine-containing DUBs 14 and the Cu atom of the CuPT, theoretically, has the potential to react with SH-containing enzymes 31 . Based on the in silicon model, CuPT has the potential to interact with both UCHL5 and USP14. This has been confirmed by Ub chain disassembly test, active-site-directed labeling experiment and DUB activity assay with CuPT treatment at therapeutic doses like 0.5 mM. The DUB assay system used in this study is mainly suitable for UCH and USP family DUBs. The enzymatic study confirmed that CuPT at 0.5 mM dose, similar to NEM, could almost completely inhibit the proteasome UCHL5 and USP14 DUB activity. These results demonstrated that CuPT targets proteasome UCHL5 and USP14. Even though in our computational model CuPT could bind with active sites of AMSH, a JAMM domain-containing protein, since there is no crystal structure of human POH1 available, whether CuPT could interact and inhibit human POH1 needs to be investigated in the future.
The additional experiments by silencing proteasome DUBs also supply some evidence to support its inhibition on 19S proteasome. It has been reported that silencing POH1, but not UCHL5 and USP14 lead to the disassembly of the 19S from the 26S proteasome 24 and this has also been confirmed in our study (data not shown). POH1 knockdown could dramatically inhibit GFPu protein degradation in HEK-293 cells and induce ubiquitinated protein accumulation in HepG2 cancer cells, and this disassembly of 26S proteasome could mostly blocked the effects of CuPT on protein degradation and ubiquitinated protein accumulation, indicating that CuPT-mediated proteasome malfunction did rely on the existence of intact 26S proteasome. Even though the roles of proteasome UCHL5 and USP14 in the cell are far from being understood, most of the studies up to now suggest the editing function of these two DUBs. In this report, it was found that UCHL5 knockdown itself did not dramatically affect GFPu protein degradation but it could partially block the effect of CuPT on GFPu protein degradation, indicating that UCHL5 plays an important role in CuPT-induced protein inhibition. USP14 inhibition both by either drug like IU1 or siRNA silencing has been confirmed to accelerate protein degradation 24,32 , which has also been confirmed in this report; USP14 silencing also affected CuPTinduced GFPu protein degradation. However, in HepG2 cancer cells, either UCHL5 or USP14 knockdown did not abrogate CuPTmediated ubiquitinated protein accumulation. These results demonstrated that UCHL5 and USP14 possibly play different roles in CuPT-mediated DUB inhibition, rather than in ubiquitinated protein accumulation process. Future studies need to be performed to compare the different roles of proteasome DUBs in protein degradation. Besides targeting proteasome DUBs, CuPT could also inhibit some other non-proteasome DUBs in the cytoplasm as well (Figure 7b), which possibly contributes to some non-proteasomerelated cytotoxicities.
PT has been widely used in some other fields for many years. ZnPT is well-known for its use in treating dandruff and seborrhoeic dermatitis 33 . ZnPT and CuPT were also introduced on the market as two new alternatives to tributyltin (TBT) for antifouling paint biocides (in 1991 and 1996, respectively) by Arch Chemicals 34 . Recently it has been reported that ZnPT exerted significant anticancer effects in vivo and ex vivo 35 , and selectively killed quiescent and rapidly dividing cells in a p53-independent manner 36 . Our study has identified one of the major mechanisms of action of this PT/copper complex on inhibiting tumor cell proteasome.

Materials.
CuPT was synthesized in our lab or purchased from Sigma-Aldrich Inc.  Cell viability assay. MTS assay (CellTiter 96Aqueous One Solution reagent; Promega, Shanghai, China) was used to test cell viability according to previously reported 37 . Briefly, 2 3 10 5 /ml cells in 100 ml were treated with either vehicle or CuPT and other agents for 24 hours. 4 hours before culture termination, 20 ml MTS was added to the wells. The absorbance density was read on a 96-well plate reader at wavelength 490 nm. IC 50 values were calculated.
Cell death assay. Apoptosis was determined by flow cytometry using Annexin Vfluoroisothiocyanate (FITC) /PI double staining 37 . Cells were incubated with GA, then collected and washed with binding buffer (Sigma-Aldrich, St. Louis, MO), then incubated in working solution (100 ml binding buffer with 0.3 ml Annexin V-FITC) for 15 min in dark. Cells were washed and resuspended with binding buffer. PI was added just before flow cytometric analysis. Annexin V/PI staining was also performed as described but in situ. The double stained cells were also imaged with an inverted fluorescence microscope equipped with a digital camera (Axio Obsever Z1, Zeiss, Germany).
To monitor temporal changes in the incidence of cell death in the live culture condition, PI was added to the cell culture medium, and at the desired sequential time points, the cells in the culture dish were imaged with an inverted fluorescence microscope 38 .
Western blot analysis. Whole cell lysates were prepared in RIPA buffer (1 3 PBS, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 10 mM bglycerophosphate, 1 mM sodium orthovanadate, 10 mM NaF, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1 3 Roche Complete Mini Protease Inhibitor Cocktail (Roche, Indianapolis, IN). Western blotting was performed as previously described 39 . In brief, an equal amount of total protein extracts from cultured cells were fractionated by 12% SDS-PAGE and electrically transferred onto polyvinylidene difluoride (PVDF) membranes. Primary antibodies and horseradish peroxidase-conjugated appropriate secondary antibodies were used to detect the designated proteins. The bounded secondary antibodies on the PVDF membrane were reacted to the ECL detection reagents (Amersham Bioscience, Dübendorf, Switzerland) and exposed to X-ray films (Kodak, Japan).
Cell culture and sample collection. Peripheral blood samples of normal controls were obtained from Guangzhou Blood Center and peripheral bone marrow samples of AML patients were obtained from discarded material utilized for routine laboratory tests at the Department of Hematology, Guangzhou First Municipal People's Hospital of Guangzhou Medical University; The use of these materials is approved by the Ethics Committee of these two Institutions with the permission of the patients and volunteers. Totally six patients with AML and six volunteers were recruited in this preclinical study. Mononuclear cells were isolated by Ficoll-Paque (Pharmacia, Uppsala, Sweden) density gradient. Mononuclear cell fraction was cultured in RPMI 1640 culture medium with 15% FBS.
Peptidase activity assay. About 4,000 cells were treated with GA at various concentrations at 37uC for 6 hours. The drug-treated cells were then incubated with the Glo Cell-Based Assay Reagent (Promega Bioscience, Madison, WI) for 10 minutes. The proteasome peptidase activity was detected as the relative light unit (RLU) generated from the cleaved substrate in the reagent following the manual. Luminescence generated from each reaction was detected with luminescence microplate reader (Varioskan Flash 3001, Thermo, USA). In vitro proteolytic peptidase (CT-like peptidase, caspase-like and trypsin-like) activities were detected by using the synthetic fluorogenic peptide substrates as we reported previously 37 .
Computational modeling. In order to obtain valuable binding information of copper pyrithione toward the DUBs mainly including USP14 and UCHL5, molecular docking studies were performed with CDOCKER protocol of Discovery Studio 2.0 [Accelrys Software Inc. (2007)]. The crystallographic structures of USP14 and UCHL5 were directly downloaded from the Protein Data Bank (PDB IDs: 2AYO and 3RIS). After removing irrelevant components, hydrogen atoms were added and their positions were minimized with a 0.01 kcal/mol /Å root mean square gradient by using the all-atom CHARMm forcefield and the Adopted Basis Newton-Raphson (NR) Algorithm. In addition, taking into account the possible hydrolysis of compound PT (L1) in certain physiological conditions, the hydrolysate PT (L2) was selected as the docking ligand. The geometry structure of compound L2 was optimized using the DFT calculations at the B3LYP/LANL2DZ level to obtain NPA charges by using the Gaussian 03 [Revision D.01,Gaussian, Inc., Wallingford CT (2004)]. During the whole docking process, the two proteins were rigid, while the ligand L2 was flexible. The Input Site Spheres of 12Å radius were centered on each active pocket of USP14 and UCHL5, with (x, y, z) 5 (38.12, 84.32, 6.61) and (29.40, 6.57, 61.54), respectively. The conformation corresponding to the lowest CDOCKER Interaction Energy was www.nature.com/scientificreports selected as the most probable binding conformation. All parameters used in calculation were default except for explained.
Deubiquitinase activity assay. This was performed as reported 40 . Briefly, cell lysate (5 mg) or 26S proteasome (25 nM) was solved in ice-cold DUB buffer containing 50 mmol/L Tris-HCl (pH 7.5), 250 mM sucrose, 5 mM MgCl 2 , and 1 mM PMSF and pretreated with Aur (2 mM) or 2 mM NEM for 15 minutes, then incubated with Ub-AMC substrate in a 100 mL reaction volume at 25uC. AMC release generated from the cleaved substrate was temporally recorded with microplate reader (Varioskan Flash 3001, Thermo, USA).
Ubiquitin chain disassembly. In vitro disassembly of purified polyubiquitin chains (K48-linked) was performed as described earlier 41 . Purified 26S proteasome (25 nM) preincubated with either vehicle or CuPT for 10 min in vitro, and then K48-linked chains (1 mg) were added into the reaction DUB buffer for 30 min at 37uC. The extent of chain disassembly was assessed by Western blot.
DUB labeling assays. 26S proteasome (25 nM) was treated with CuPT for 10 minutes and incubated with HA-UbVS for 1 hour at 37uC, followed by boiling in reducing sample buffer and resolving by SDS-PAGE. After protein transfer to PVDF membranes, HA immunoblotting was used to detect DUB labeling 42 .
SiRNA transfection. Three siRNAs against human POH1, UCHL5 and USP14, constructed and ordered from Guangzhou Ribobio Co. Ltd, were used to transfect HepG2 and HEK-293 cells. For each transfection sample, oligomer Lipofectamine TM complexes were prepared. The oligomer Lipofectamine TM complexes were added to each well containing cells and medium, followed by gentle mixing by rocking the plate back and forth. Medium was changed after 6 hours, and the cells were incubated at 37uC in a CO 2 incubator for 48 hours, followed by CuPT treatment as indicated. Cells were collected for Western blot assay. Nude mouse xenograft model. All animal protocols used were approved by the Institutional Animal Care and Use Committee of Guangzhou Medical University. The mice were obtained from Guangdong Laboratory Animal Monitoring Institute (SCXK2008-2002). The nude Balb/c mice were housed in barrier facilities with a 12 h light dark cycle, with food and water available ad libitum. 3 3 10 7 of HepG2 cells was inoculated subcutaneously on the flanks of 5-week-old male nude mice. After 72 h of inoculation, mice were treated with either vehicle (10% DMSO, 30% Cremophor ELand 60% NaCl) or CuPT (2.5 mg/kg/day) for totally 15 days (Day 7 interval), respectively. Tumor volumes were recorded and calculated as previously reported 43 .
Immunohistochemical staining. Formalin-fixed xenografts were embedded in paraffin and sectioned according to standard techniques as we previously reported 43 . Tumor xenograft sections (4 mm) were immunostained using the MaxVision kit (Maixin Biol) according to the manufacturer's instructions. The primary antibodies were used as indicated. 50 ml MaxVisionTM reagent was applied to each slide. Color was developed with 0.05% diaminobenzidine and 0.03% H 2 O 2 in 50 mM Tris-HCl (pH 7.6), and the slides were counterstained with hematoxylin. A negative control for every antibody was also included for each xenograft specimen by substituting the primary antibody with preimmune rabbit serum.
Statistical analysis. All the results were expressed as Mean 6 SD where applicable. GraphPad Prism 4.0 software (GraphPad Software) was used for statistical analysis. Student's t test was used to compare the differences between variables. P value of , 0.05 was considered statistically significant.