Specific Role of Cathepsin L

Home , Cathepsin K

[CANCER RESEARCH 64, 1773–1780, March 1, 2004] Senescence-initiated Reversal of Drug Resistance: Specific Role of Cathepsin L

Xin Zheng,1 Pauline M. Chou,2 Bernard L. Mirkin,1,3 and Abdelhadi Rebbaa1 Children’s Memorial Institute for Education and Research, Children’s Memorial Hospital, 1Department of Pediatrics, 2Pathology, and 1,3Molecular Pharmacology and Biological Chemistry, Feinberg School of Medicine, Northwestern University, Chicago, Illinois

ABSTRACT centration and duration of exposure, doxorubicin was able to induce senescence, apoptosis, and a necrotic type of cell death. In addition, The present study was undertaken to verify whether induction of inhibition of doxorubicin-induced apoptosis by the pan-caspase inhib- senescence could be sufficient to reverse drug resistance and, if so, to itor (Q-VD-OPH) resulted in a switch to a cell death with features of determine the underlying mechanism(s). Our findings indicated that cotreatment of drug-resistant neuroblastoma cells with doxorubicin, at sub- senescence. This suggested that whether or not apoptosis is inhibited, lethal concentrations, in combination with the pan-caspase inhibitor, senescence must be inhibited for cancer cells to survive chemotherapy Q-VD-OPH, elicited a strong reduction of cell viability that occurred in a toxicity and become drug resistant. In light of this, forcing cancer cells caspase-independent manner. This was accompanied by the appearance of to undergo senescence may be sufficient to sensitize them to chem- a senescence phenotype, as evidenced by increased p21/WAF1 expression otherapy and reverse drug resistance. and senescence-associated ␤-galactosidase activity. Experiments using The present study, using the fact that Q-VD-OPH was able to specific inhibitors of major cellular proteases other than caspases have accelerate doxorubicin-induced senescence in cancer cells, was un- shown that inhibition of cathepsin L, but not proteasome or cathepsin B, dertaken to determine the importance of senescence in initiating drug was responsible for the senescence-initiated reversal of drug resistance. resistance reversal. Validation of this hypothesis as well as identifi- This phenomenon appeared to be general because it was valid for other cation of the underlying mechanism(s) were also addressed. drugs and drug-resistant cell lines. A nonchemical approach, through cell transfection with cathepsin L small interfering RNA, also strongly reversed drug resistance. Further investigation of the underlying mecha- MATERIALS AND METHODS nism revealed that cathepsin L inhibition resulted in the alteration of intracellular drug distribution. In addition, in vitro experiments have Reagents. Cell lines including human neuroblastoma SKN-SH, murine demonstrated that p21/WAF1 is a substrate for cathepsin L, suggesting neuroblastoma Neuro2A, osteosarcoma OSA (SaOS), leukemia HL-60, and the that inhibition of this enzyme may result in p21/WAF1 stabilization and p21/WAF1 plasmid were purchased from American Type Culture Collection its increased accumulation. All together, these findings suggest that ca- (Rockville, MD). DMEM and fetal bovine serum were obtained from Bio- thepsin L inhibition in drug-resistant cells facilitates induction of senes- Whittaker (Walkersville, MD). Doxorubicin, 3-(4,5-dimethyl-2-thiazolyl)2,5- cence and reversal of drug resistance. This may represent the basis for a diphenyl tetrazolium bromide (MTT), cisplatin, vinblastine, and 5-bromo-4- ␤ novel function of cathepsin L as a cell survival molecule responsible for chloro-3-indolyl -D-galactopyranoside were purchased from Sigma (St. initiation of resistance to chemotherapy. Louis, MO). Q-VD-OPH was from Enzyme Systems Products, Inc. (Liver- more, CA). Lactacystin, cathepsin B inhibitor [N-(L-3-trans-(propylcarba- moyl)oxirane-2-carbonyl)-L-isoleucyl-L-proline], cathepsin L inhibitor #1 INTRODUCTION [Z-Phe-Tyr(t-Bu)-diazomethylketone], cathepsin K inhibitor [1,3-Bis(N- CBZ-Leu-NH)-2-propanone1,3-di(N-carbobenzoyloxy-L-leucyl)amino acetone], ⅐ Apoptosis has been shown to be a major pathway leading to and cathepsin S inhibitor (Z-Phe-Leu-COCHO H2O) were from Calbiochem- neoplastic cell death upon exposure to chemotherapeutic agents and Novabiochem (San Diego, CA). Cathepsin L substrate (Z-Phe-Arg)2-R110 was has become the focus of many experimental therapeutic investigations purchased from Molecular Probes (Eugene, OR). Purified cathepsin L was from Biomol (Plymouth Meeting, PA). Antibodies to p21/WAF1 and to ␤-actin were (reviewed in Refs. 1 and 2). The common belief is that cytotoxic drugs obtained from Santa Cruz Biotechnology (Santa Cruz, CA); antibodies to cleaved induce DNA damage, leading to activation of the p53/Fas/caspase-8 caspase-3 and secondary antibodies conjugated to horseradish peroxidase were system and initiation of a mitochondria-mediated apoptotic death. The from Cell Signaling Technologies (Beverly, MA). Antibody to cleaved (active occurrence of this pathway has been confirmed for a variety of drugs form of cathepsin L) was purchased from Novus Biologicals Inc. (Littleton, CO). and types of cancer (3–6), leading to the assumption that impairment Enhanced chemiluminescence reagents (ECL) were from Amersham (Arlington of apoptotic pathways could be sufficient to explain the development Heights, IL). Immobilon-P transfer membrane for Western blot was purchased of resistance to chemotherapy. Although this was demonstrated in from Millipore (Bedford, MA). various hematopoietic malignancies (7, 8), it was not necessarily valid Cell Culture, Drug Treatment, and Cytotoxicity Assay. SKN-SH, for most solid tumors (8–11). More importantly, studies designed Neuro2A, and OSA were cultured in DMEM and HL-60 in RPMI 1640 supplemented with 10% fetal bovine serum at 37°C in a 95% air/5% CO specifically to inhibit apoptosis often did not lead to increased cell 2 atmosphere. Doxorubicin-resistant variants of these cells were obtained by survival (12–16); unexpectedly, they resulted in a switch to a non- continuous incubation of the parental cell lines with incremental concentra- apoptotic cell death. The nature of this “default death” has been the Ϫ9 Ϫ6 tions of doxorubicin, varying from 10 to 10 M, over a period of 3 months focus of recent investigations, some of which attribute this type of cell (19). Cytotoxic activity of doxorubicin and cysteine protease inhibitors were death to necrosis (12, 17) or some undefined causes (15, 16). In a quantitatively determined by a colorimetric assay using MTT, as described recent study (18), we have shown that depending on the drug con- previously (19). Western Blot Analysis. Cells were seeded in 25-cm2 flasks in DMEM Received 3/28/03; revised 12/29/03; accepted 1/7/04. containing 10% fetal bovine serum and cultivated for 24 h before the addition Grant support: Illinois Department of Public Aid (to A. Rebbaa), the North Suburban of doxorubicin in the presence or absence of protease inhibitors. After incu- Medical Research Junior Board (to A. Rebbaa and B. L. Mirkin), the Anderson Founda- bation for an additional 24 h, proteins were solubilized with lysis buffer [50 tion (to B. L. Mirkin), the Medical Research Institute Council (to B. L. Mirkin), and the R. mM HEPES (pH 7.4), 150 mM NaCl, 100 mM NaF, 1 mM MgCl2, 1.5 mM Wile Neuroblastoma and Chemotherapy Fund (to B. L. Mirkin). ␮ The costs of publication of this article were defrayed in part by the payment of page EGTA, 10% glycerol, 1% Triton X-100, 1 g/ml leupeptin, and 1 mM phen- charges. This article must therefore be hereby marked advertisement in accordance with ylmethylsulfonyl fluoride]. Equal quantities of protein were separated by 18 U.S.C. Section 1734 solely to indicate this fact. electrophoresis on a 12% SDS-PAGE gel and transferred to Immobilon-P Requests for reprints: Abdelhadi Rebbaa, Children’s Memorial Institute for Educa- membranes. Proteins of interest were identified by reaction with specific tion and Research, M/C 224, Children’s Memorial Hospital, 2300 Children’s Plaza, Chicago, IL 60614. Phone: (773) 880-8302; Fax: (773)-880-8333; E-mail: arebbaa@ primary and secondary antibodies linked to horseradish peroxidase. Reactive childrensmemorial.org. bands were detected by chemiluminescence. 1773

Senescence-associated ␤-Galactosidase (SA-␤-Gal) Staining. Cells were Production of Recombinant p21/WAF1 and Assay of Its Cleavage by seeded into 24-well plates in DMEM culture medium. After 24 h, drugs were Purified Cathepsin L. Plasmid pGEX-2TKcs containing the GST-CIP1 gene added, and the cells were incubated for 5 days. SA-␤-Gal staining was coding for p21/WAF1 was expressed in BL21 Star (DE3) bacteria (Invitrogen, performed as described previously (20). In brief, cells were fixed for 5 min in Carlsbad, CA). Expression of recombinant p21/WAF1 protein was induced by 3% formaldehyde, washed, and incubated at 37°C with 5-bromo-4-chloro- isopropyl-1-thio-␤-D-galactopyranoside (1 mM) and purified from the bacterial 3-indolyl ␤-D-galactopyranoside (1 mg/ml), dissolved in a solution containing culture with the MicroSpin GST Purification Module (Amersham Pharmacia 40 mM citric acid (pH 6.5), 5 mM potassium ferrocyanide, 5 mM potassium Biotech, Piscataway, NJ). The quantity of expressed protein was determined by

ferricyanide, 150 mM NaCl, and 2 mM MgCl2. After 24 h incubation, photo- Western blot using specific antibodies either to p21/WAF1 (C-19; Santa Cruz graphs were taken with a phase microscope. Biotechnology) or to glutathione S-transferase (GST; Amersham Pharmacia Fluorescence Microscopy. Cells were seeded on coverslips and incubated Biotech, Piscataway, NJ). in DMEM containing 10% fetal bovine serum for 24 h. Doxorubicin was added Cleavage of recombinant p21/WAF1 by purified cathepsin L was performed Ϫ at 10 7 M in the presence or absence of cysteine protease inhibitors and by incubating various amounts of the enzyme (1–5 ng) with 100 ng of incubated for 30 min at 37°C. The cells were then washed twice with cold PBS; p21/WAF1 and cathepsin substrate in reaction buffer [100 mM sodium acetate the intracellular localization of doxorubicin was examined by fluorescence (pH 5), 1 mM EDTA, and 4 mM DTT]. After 20 min at room temperature, microscopy (excitation, 480 nm; emission, 560 nm), and photographs were samples were boiled and separated by SDS-PAGE. p21/WAF1 was detected by taken. specific antibodies either to p21/WAF1 (C-19) or to GST. Rhodamine-123 uptake was evaluated by incubating the cells with (10 ␮ g/ml rhodamine) for 30 min at 37°C in the presence or absence of cathepsin RESULTS L inhibitor (10 ␮M), after which the culture medium was removed and cells were washed twice with PBS before determining the intracellular localization Effect of Q-VD-OPH on Doxorubicin-induced Senescence and of rhodamine-123. For the efflux experiment, cells were washed with PBS Reversal of Drug Resistance after rhodamine uptake and incubated in fresh medium without dye for an additional 30 min, either in the presence or absence of cathepsin L inhibitor. Previous work from this laboratory has demonstrated that the Cells were then visualized under fluorescence microscopy (excitation, 480 nm; cysteine protease inhibitor, Q-VD-OPH, switches doxorubicin- emission, 560 nm) and photographed. induced apoptosis to senescence (18). The human neuroblastoma cell Measure of Purified Cathepsin L Activity. Cathepsin L activity was line resistant to doxorubicin SKN-SH/R (19) was used to determine measured according to the manufacturer’s procedure. Briefly, the purified whether Q-VD-OPH could accelerate doxorubicin-induced senes- ␮ enzyme (200 ng) was incubated with the inhibitor (10 M) in a 96-well plate cence in these cells and initiate the reversal of drug resistance. As for 15 min at room temperature in 100 ␮l of reaction buffer [100 mM sodium shown in Fig. 1A, when this inhibitor (at 20 ␮M) was added to the cell acetate (pH 5), 1 mM EDTA, and 4 mM DTT]. One hundred ␮l of substrate were added and incubated for an additional 30 min at room temperature. culture medium in the presence of doxorubicin, decrease in cell Fluorescence was measured in a plate reader (Victor Multilabel Counter; viability was observed after 4 days of incubation, as measured by the Perkin-Elmer) at 380-nm excitation and 450-nm emission wavelengths. MTT assay. Q-VD-OPH alone did not affect cellular viability. Cell siRNA Design and Transfection. The Human Cathepsin L siRNA was cultivation in the presence of the drug combination for an additional designed in our laboratory from the human cathepsin L cDNA sequence 4 days resulted in more extensive cell death (Fig. 1A) indicating that AAGTGGAAGGCGATGCACAAC (91–111) and was synthesized by Dhar- the protease inhibitor was able to sensitize SKN-SH/R cells to doxo- macon (Lafayette, CO). On the day before transfection, 3 ϫ 105 drug-resistant rubicin and to reverse drug resistance. osteosarcoma cells (OSA-R) were seeded in 6-well plates and grown in 2.5 ml Microscopic analyses revealed that cells treated for 8 days with the of DMEM supplemented with 10% fetal bovine serum. After 24 h in culture, Ϫ7 doxorubicin (10 M) in combination with Q-VD-OPH (20 ␮M) 25 ␮lof20␮M stock solution of siRNA duplexes were transfected into cells acquired morphological features typical of senescence (Fig. 1B). This with a GeneSilencer siRNA Transfection Reagent kit according to the manufacturer’s protocol (Gentherapysystem). After 48 h of incubation, the cells finding was supported by an increased expression of p21/WAF1 (Fig. Ϫ ␤ were treated with 10 5 M doxorubicin and maintained in culture for an 1C) and SA- -gal staining (Fig. 1D), both of which are known to additional 48 h before analysis. The cells were counted, and protein lysates accompany development of a senescence phenotype (20, 21). At the Ϫ7 were used to detect cathepsin L and p21/WAF1 expression by Western blot as concentration of doxorubicin (10 M) used, caspase-3 was not acti- described above. vated by doxorubicin, Q-VD-OPH, or the combination of both (Fig.

Fig. 1. Effect of Q-VD-OPH on SKN-SH/R cell response to doxorubicin. A, cells were incubated with doxorubicin (Dox.), Q-VD-OPH (20 ␮M), or the combination of both for 4 or 8 days. Cell viability (percentage related to control nontreated cells) was then measured by MTT assay as described in “Materials and Methods.” Values are the means of three determinations; bars, SE. Conc, concentration. B, morphological analysis of cells Ϫ7 treated with 10 M doxorubicin (Dox.) and/or Q- VD-OPH (20 ␮M) as described above. After 8 days of incubation, photographs were taken under light microscopy. C, Western blot analysis of p21/ WAF1 and cleaved caspase-3 (Cl-Casp-3) expres- Ϫ7 sion in response to treatment with 10 M and/or Q-VD-OPH (Q-VD.;20␮M) for 24 h. Ctl., a pos- itive control for expression of cleaved caspase-3. D, cells treated for 5 days with doxorubicin alone or in combination with Q-VD-OPH (Q-VD) were stained for SA-␤-Gal activity as described in “Ma- terials and Methods.”

1774

Fig. 2. Effect of protease inhibitors on doxorubicin-induced senescence in SKN-SH/R cells. A, morphological analysis of cells treated with the indicated inhibitors with (upper panel) or without (lower panel) doxorubicin for 5 days. Photographs were taken under light microscopy. Ctl, control; Lac, lactacystin; Q-VD, Q-VD-OPH; Cli, cathepsin L inhibitor; Cbi, cathepsin B inhibitor; Cki, cathepsin K inhibitor; Csi, cathepsin S inhibitor. B, expression of p21/WAF1 in response to treatment with doxorubicin with or without protease inhibitors for 24 h. p21/WAF1 expression was determined by West- ern blot using the specific antibody C19. Antibody to ␤-actin was used as a gel loading control. Lac, lactacystin; Qvd, Q-VD-OPH; Cli, cathepsin L inhibitor; Cbi, cathepsin B inhibitor; Cki, cathepsin K inhibitor; Csi, cathepsin S inhibitor. C, comparative effects of Q-VD- OPH and cathepsin L inhibitor on activation of SA-␤-Gal. Cells were treated with drugs for 5 days, followed by SA-␤-Gal staining as described in “Materials and Methods.” Inhib., inhibitor; Dox., doxorubicin; Q-VD, Q-VD-OPH; Cli, cathepsin L inhibitor.

1C). This suggested that activation of caspase-3 was not responsible Differential Effect of Cathepsin L Inhibition on Doxorubicin- for the diminished cellular proliferation and reversal of drug resist- induced Expression of p21/WAF1 and Activation of Caspase-3 ance by Q-VD-OPH. Because this agent is also a general cysteine in Drug-sensitive and Drug-resistant Cells protease inhibitor, it is likely that inhibition of other cysteine proteases than caspases could be implicated in initiating senescence. Expression of p21/WAF1 was used as a marker to assess the ability of wild-type and doxorubicin-resistant cells to undergo senescence, Identification of Putative Cysteine Proteases Associated with and cleaved caspase-3 was used to identify apoptosis in these cells Doxorubicin-induced Senescence (18). Expression of these molecules was followed in the presence of increasing drug concentrations both in wild-type and drug-resistant The ubiquitin-proteasome complex and lysosomal cathepsins rep- cells (Fig. 3A). Although doxorubicin-induced expression of p21/ resent two major families of cellular proteases known to be respon- Ϫ8 WAF1 was initiated at 10 M for wild-type cells, it was elicited only sible for digesting and eliminating undesirable proteins (22, 23). The Ϫ6 at 10 M in their resistant counterparts. This observation is in role of proteasome in cell survival is well described, particularly in agreement with our previous finding on the relative toxicity of doxo- relation to the fact that this molecule is able to degrade certain tumor rubicin (ϳ100 times difference) between these two cell lines (19). suppressors and transcription factors responsible for the inhibition of Activation of caspase-3 followed a similar profile with a maximum at cell proliferation (24). In addition, decreased activity of proteasome Ϫ6 Ϫ4 10 M for wild-type and 10 M for the resistant cells (Fig. 3A). The has been documented to be associated with cellular aging (25), sug- decrease in expression of p21/WAF1 and its relationship to increased gesting a role in senescence. Although the possible implication of cathepsins in apoptotic death has been documented (26–28), their caspase-3 activity was also documented (18) and could be explained putative role in cell survival remains to be confirmed. by the fact that p21/WAF1 has been reported to be a substrate for We have tested a series of specific inhibitors against the proteasome active caspases (30). Therefore, diminished expression of p21/WAF1 and lysosomal cathepsins to determine whether they affect the ability may be attributable to its degradation by active caspase-3. of a drug-resistant cell to undergo senescence in the presence of The profiles for expression of p21/WAF1 and activated caspase-3 doxorubicin. The results shown in Fig. 2 indicate that although the were both changed when cathepsin L inhibitor was added to the cells Ϫ8 proteasome inhibitor (lactacystin) and the cathepsin B inhibitor had no (Fig. 3B). At the doxorubicin concentration (10 M), cotreatment of effect on the action of doxorubicin in SKN-SH/R cells, cathepsin L drug-sensitive cells with this drug and the cathepsin L inhibitor did inhibitor was effective in inducing both morphological features of not affect significantly expression of p21/WAF1 when compared with senescence (Fig. 2A) as well as expression of the corresponding cells treated with doxorubicin alone. However, in doxorubicin-resist- molecular markers p21/WAF1 (Fig. 2B) and SA-␤ gal (Fig. 2C). Cell ant cells (Fig. 3B), cathepsin L inhibition enhanced expression of treatment with cathepsin L inhibitor or with doxorubicin alone had no p21/WAF1 when combined with doxorubicin at a sublethal concen- Ϫ7 effect on cell morphology (Fig. 2A) or expression of these markers tration of 10 M. These data suggested that the enhancement of (data not shown). Inhibition of cathepsins K and S, which are mem- doxorubicin sensitivity attributable to cathepsin L inhibition selec- bers of the cathepsin L family (29), also resulted in induction of tively affects drug-resistant cells. Ϫ7 senescence features, both at the morphological level and expression of Interestingly, treatment of drug-sensitive cells with 10 M doxo- p21/WAF1. Cathepsin L inhibitor appeared to be the most effective, rubicin in the presence of cathepsin L inhibitor led to decreased suggesting that specific targeting of this enzyme may produce a expression of p21/WAF1, and this was not associated with caspase-3 greater effect in terms of drug resistance reversal. activation (Fig. 3B). In contrast, drug-resistant cells treated with 1775

Fig. 3. Effect of cathepsin L inhibition on doxorubicin (Dox.)-induced expression of p21/WAF1 and cleaved caspase-3 (Cl. casp-3) in wild-type (W) and drug-resistant (R) cells. A, cells were treated with doxorubicin at the indicated concentrations (␮M) for 24 h. Expression p21/WAF1, cleaved caspase-3, and ␤-actin were analyzed by Western blot using specific antibodies. B, W and R cells were treated with cathepsin L inhibitor (10 ␮M)in the presence or absence of doxorubicin at the indicated concentrations, and expressions of the molecules described in A were determined. Cli., cathepsin L inhibitor.

Ϫ6 doxorubicin at 10 M in the presence of cathepsin L inhibitor line Neuro2A/R, the osteosarcoma cells OSA/R, and the leukemia cell displayed a decreased expression of p21/WAF1 that was associated line HL60/R. No effect of the drug combination was noticed on with increased caspase-3 activation, suggesting a switch of cell death doxorubicin toxicity in all of the four drug-sensitive cell lines, in from senescence to apoptosis. These data indicate that enhancement of concurrence with the data shown in Fig. 3B demonstrating a lack of doxorubicin toxicity in response to cathepsin L inhibition affects drug-resistant cells only and highlight the specific role of senescence induction in initiating reversal of drug resistance.

Potency of Cathepsin L Inhibition on Reversal of Drug Resistance MTT assay was performed to compare the effect of cathepsin L inhibition to other protease inhibitors on reversal of doxorubicin resistance in SKN-SH/R cells. As expected, molecules that where not able to induce senescence, such as lactacystin or the cathepsin B inhibitor (Fig. 2), did not affect cell response to doxorubicin after 4 days of incubation (Fig. 4A). However, Q-VD-OPH and the cathepsin L family inhibitors reversed resistance to doxorubicin in various degrees. Cathepsin L inhibitor was the most effective, suggesting that the corresponding enzyme could be the lead candidate in this process. The limited effects of cathepsin K and S inhibitors on drug toxicity may be explained by the fact that the corresponding enzymes may play a role in drug resistance, or that these inhibitors are nonspecific and can also inhibit cathepsin L. An in vitro experiment in which the activity of purified cathepsin L was assayed in the presence of protease inhibitors was carried out to verify whether the effect of these inhibitors on the reversal of drug resistance correlates with their effects on the activity of this enzyme. Data shown in Fig. 4B indicate that cathepsin L inhibitor was the most effective, and that Q-VD-OPH and cathepsin K were also able to reduce strongly the activity of this enzyme. Their limited effect on the reversal of drug resistance may be explained by the lack of specificity toward cathepsin L. All together, these findings indicate that cathepsin L may represent a primary target in reversal of resistance to doxorubicin. The role of cathepsins K in this process, although not excluded, seems to be less important than cathepsin L.

Validation for Other Drugs and Cancer Cell Lines Fig. 4. Effect of various protease inhibitors on cell response to doxorubicin and on the activity of purified cathepsin L. A, doxorubicin-resistant SKN-SH/R cells were incubated The effect of cathepsin L inhibitor was tested on drug-sensitive and with the indicated inhibitors in the absence or in the presence of increasing concentrations of doxorubicin. After 96 h of incubation, cell viability was determined by MTT assay. B, -resistant cell lines corresponding to various cancer types and also in purified cathepsin L was incubated with the indicated inhibitors, and its activity was relation to unrelated drugs. As shown in Fig. 5, this molecule was able measured as described in “Materials and Methods” and reported in the graph as arbitrary units (A.U.). Data in A and B represent the means of three determinations; bars, SE. La, to reverse doxorubicin resistance in cell lines such as the human lactacystin; Q, Q-VD-OPH; Li, cathepsin L inhibitor; Bi, cathepsin B inhibitor; Si, neuroblastoma cell line SKN-SH/R, the murine neuroblastoma cell cathepsin S inhibitor; K, cathepsin K inhibitor. 1776

Fig. 5. Cell-type dependent effect of cathepsin L inhibition on response to doxorubicin. Wild-type (W) and doxorubicin-resistant (R) cells representing human neuroblastoma (SKN-SH), murine neuroblastoma (Neuro2A), osteosarcoma (OSA), and leukemia (HL-60) cells were incubated for 96 h in the absence or in the presence of cathepsin L inhibitor (Cli) at 10 ␮M and various concentrations of doxorubicin. MTT staining was performed at the end of the assay, and cell viability was calculated as the percentage of nontreated cells. Data represent the means of six determinations; bars, SE. effect of the drug combination on induction of p21/WAF1 expression or caspase-3 activation in wild-type SKN-SH cells. The cellular response to non-anthracyclin agents, such as cisplatin and vinblastine, was also investigated (Fig. 6). Doxorubicin-resistant SKN-SH/R cells were not resistant to cisplatin. Interestingly, the cellular response to cisplatin was not significantly affected by cathepsin L inhibition in both doxorubicin-sensitive and -resistant cells. This represents an additional argument in favor of the observation made earlier (Fig. 5), indicating that cathepsin L inhibition enhances cytotoxic drug response only in drug-resistant cells. In this case, because there was no resistance to cisplatin, no resistance reversal should be expected. However, in the case of vinblastine, doxorubicin-resistant cells were also resistant to this drug (Fig. 6). More importantly, cathepsin L inhibition reversed this resistance. Overall, the data presented in Figs. 5 and 6 suggest that alteration of cathepsin L function or its related pathway could be associated with development of drug resistance to more than one chemotherapeutic agent and in various cancer types.

Cathepsin L Small Interfering RNA (siRNA) Reverses Drug Resistance To determine the specificity of cathepsin L inhibition in drug resistance reversal, we have transfected doxorubicin-resistant cells with cathepsin L siRNA as an independent approach to chemical inhibition. The data presented in Fig. 7 indicate that transfection of OSA/R cells with cathepsin L siRNA results in almost complete Fig. 6. Effect of cathepsin L inhibition on SKN-SH/R cell response to cisplatin and suppression of the enzyme expression as shown by Western blot (Fig. vinblastine. Doxorubicin-resistant SKN-SH/R cells were incubated in 96-well plates in the 7, upper panel). Interestingly, expression of p21/WAF1 in response to presence of the indicated concentration (Conc) of drugs with or without cathepsin L inhibitor (Cli)at10␮M. An MTT assay was performed after 4 days of incubation, and cell doxorubicin was up-regulated in samples treated with cathepsin L viability was calculated as the percentage of nontreated cells. Data represent the means of inhibitor or with siRNA to cathepsin L, suggesting a potential rela- six determinations; bars, SE. The data presented in this figure are representative of three independent experiments in which we have consistently found that response to vinblastine tionship between this enzyme and p21/WAF1. Although cell treat- in the drug-resistant cells was enhanced in the presence of cathepsin L inhibitor. W, wild ment with doxorubicin or siRNA alone inhibited proliferation to a type; R, resistant. 1777

Fig. 7. Effect of cathepsin L (Cat. L) inhibition by siRNA on cellular response to doxorubicin (Dox). Doxorubicin-resistant OSA/R cells were transfected with transfection reagent alone (TR), or TR plus siRNA, Ϫ5 followed by treatment with doxorubicin at 10 M alone or in combination with siRNA or with the chemical inhibitor of cathepsin L (Cli) for 48 h. Expression of cathepsin L and p/21WAF1 were measured by Western blot (upper panel), and viable cells were counted after doxorubicin treatment (lower panel). Bars, SE.

certain extent, the combination of both inhibited cell proliferation by tion could be of lysosomal nature because anthracyclins, including Ͼ80% (Fig. 7, lower panel). Similar results were obtained when doxorubicin, are basic drugs that have been shown to accumulate doxorubicin was combined with the chemical cathepsin inhibitor. The preferentially in acidic organelles such as lysosomes (31). In the data are in support of the previous findings (Fig. 4) and constitute an presence of cathepsin L inhibitor, a remarkable increase of doxoru- independent method to demonstrate the specificity of cathepsin L bicin content, mainly in the nucleus, was observed (Fig. 8). inhibition and its role in reversing drug resistance in cancer cells. Because these cells also express the drug efflux P-glycoprotein as part of their resistance mechanism (19), drug efflux experiments were Putative Mechanisms Leading to Reversal of Drug Resistance conducted to determine whether cathepsin L inhibition could affect Through Cathepsin L Inhibition the function of this pump. Fig. 8 indicated that cathepsin L inhibition Effect on Cellular Localization of Doxorubicin. To determine did not prevent doxorubicin efflux. Similar experiments conducted whether cathepsin L inhibition affects intracellular drug distribution, with rhodamine-123, a known substrate for P-glycoprotein that accu- we have measured cellular uptake and localization of doxorubicin by mulates mainly in the mitochondria (Fig. 8), indicated that cathepsin fluorescence microscopy in the presence and in the absence of the L inhibitor did not alter cytoplasmic localization of this dye and has enzyme inhibitor. As shown in Fig. 8, in the absence of cathepsin L no effect on its efflux; therefore, it may have no effect on P-glyco- inhibitor, a relatively small accumulation of doxorubicin was ob- protein function. The fact that specific siRNA to cathepsin L also served in the cytoplasm of drug-resistant cells. The precise localiza- reversed drug resistance constitutes an additional argument against the

Fig. 8. Uptake and efflux of doxorubicin and rhodamine-123 in response to cathepsin L inhibition. Doxorubicin-resistant SKN-SH/R cells were incubated on coverslips in the absence (Ϫ) or the presence (ϩ) of cathepsin L inhibitor and doxorubicin (Dox.) or rhodamine-123 (R123). After 30 min of incubation, cells were washed twice with PBS, and photographs were taken under the fluorescence microscope before (Uptake) or after (Efflux) incubation in fresh medium without doxorubicin or rhodamine for an additional 30 min.

1778

tures of senescence (Fig. 1). This effect cannot be attributed to inhibition of caspases because the doxorubicin concentration used did not activate caspase-3 (Fig. 1C), suggesting that senescence-initiated reversal of drug resistance could occur independently of cellular apoptotic status. The key finding derived from this study (Fig. 1C)is that inhibition of a cysteine protease(s) results in induction of p21/ WAF1 expression and cell growth arrest, suggesting that the putative Fig. 9. In vitro cleavage of p21/WAF1 by the purified enzyme. Recombinant p21/ cysteine protease might be considered as a survival molecule. The WAF1 (100 ng) was incubated in reaction buffer with increasing amounts (ng) of purified cathepsin L. After 20 min, the reaction was stopped, and proteins were separated on SDS proteasome and lysosomal cathepsins were considered as potential gel. p21/WAF1 was detected with either the COOH-terminal specific antibody (upper candidates, because these two families of proteases can be active, panel) or with anti-GST antibody (lower panel). even in the absence of any death stimulus. Specific inhibitors were tested to discriminate between major lysosomal proteases and the proteasome in terms of their ability to protect against doxorubicin implication of P-glycoprotein in this process. All together, these data toxicity. Although the proteasome inhibitor lactacystin did not alter suggest that cathepsin L could play a role in drug resistance by cellular response to doxorubicin, cathepsin L inhibitor was found to be facilitating doxorubicin incorporation into lysosomes, thereby pre- very effective in enhancing drug toxicity and forcing cells to undergo venting it from accumulating in the nucleus. senescence (Figs. 2 and 3). Increased expression of the cell cycle Direct Effect of Cathepsin L on p21/WAF1. To address the inhibitor p21/WAF1 and the SA-␤-Gal in cells treated with doxoru- possibility that increased expression of p21/WAF1 in the presence of bicin and cathepsin L inhibitor favor a senescence-initiated reversal of the combination doxorubicin/cathepsin L inhibitor could be attribut- drug resistance (Figs. 1–3). able to its protection from degradation by this enzyme, we have The effect of cathepsin L inhibition on drug toxicity appeared to be performed in vitro experiments to determine whether p21/WAF1 a general phenomenon because it was valid for other murine and could be a substrate for cathepsin L. In these experiments, we have human drug-resistant cell lines (Fig. 5). The potency of drug resist- studied the cleavage of recombinant GST-tagged p21/WAF1 by pu- ance reversal was, however, cell type dependent, suggesting that rified cathepsin L. Data shown in Fig. 9 indicated that cathepsin L either the level of cathepsin L or other associated factors may be readily cleaves p21/WAF1, suggesting that this enzyme may partici- implicated. Interestingly, this effect on drug toxicity was valid only pate in drug resistance by cleaving and inactivating cell cycle inhib- for drug-resistant cells (Fig. 6). We have found that these two cell itors that were induced by cytotoxic stress. This possibility may types express the same amount of cathepsin L (data not shown), represent a putative mechanism by which cathepsin L participates in suggesting that differences in activation of this enzyme, its cellular the prevention of drug toxicity. localization, its endogenous inhibitors, or its downstream targets may explain the relative susceptibility of drug-resistant and -sensitive cells DISCUSSION to cathepsin L inhibition. To determine the specificity of cathepsin L as a target to reverse In the present study, we describe a new pathway that may influence drug resistance, siRNAs for this enzyme were used to reduce its the decision of cancer cells to undergo cell death or survive the initial expression in drug-resistant cells (Fig. 7). The results indicated that treatment with chemotherapeutic agents and ultimately become drug siRNA to cathepsin L were potent inhibitors of the enzyme expres- resistant. This pathway appeared to be dependent on changes in intracellular drug distribution or the ability of cells to destabilize sion, and their action resulted in up-regulation of p21/WAF1 expres- tumor suppressor gene products that are usually induced in the pres- sion and reversal of drug resistance. In light of these results, we ence of cytotoxic drugs. The cysteine protease cathepsin L in partic- conclude that the observed effect on drug toxicity of the chemical ular and probably other members of its close family appeared to play cathepsin L inhibitor was mainly attributable to its action on this an important role in this process. enzyme. Cathepsins were, until recently, considered as housekeeping mol- It has been reported that basic drugs such as anthracyclins prefer- ecules. However, with the aid of mice knock-out technology, specific entially accumulate in acidic organelles such as lysosomes (31, 37). functions have been attributed to each one of these lysosomal pro- This accumulation was found to be greater in drug-resistant cells than teases (23). Relative to other cathepsins, cathepsin L is implicated in their drug-sensitive counterparts (31), suggesting that lysosomal drug many key cellular functions. For instance, it has been shown that sequestration may be responsible, at least in part, for development of transgenic mice lacking cathepsin L suffered from dilated cardiomy- drug resistance. Along these lines, our results (Fig. 8) indicated that opathy (32), suggesting a critical role of this enzyme for cardiac cathepsin L could be the key element responsible for facilitating morphologenesis and function. Neural loss and brain atrophy accom- doxorubicin accumulation in lysosomes, and therefore, its inhibition panied by early death were also reported in double knock-out mice might result in free drug movement from lysosomes to nucleus and lacking both cathepsins B and L (33). Interestingly, cardiomyocytes increased drug toxicity. Another possibility is that inhibition by ca- and neurons from these mice contained multiple, large, and apparently thepsin L could result in stabilization of p21/WAF1 (Fig. 9) and fused lysosomes characterized by storage of electron-dense nonde- induction of senescence. graded material. Cathepsin L deficiency was also associated with hair In conclusion, this investigation has introduced the concept that loss in mice (34, 35). In Caenorhabditis elegans, the loss of this forcing cells to undergo senescence, even without induction of apo- enzyme was found to be associated with embryonic lethality (36). In ptosis, may be sufficient to mediate the reversal of drug resistance. light of these findings, it appeared that cathepsin L could be a key Inhibition of cathepsin L appears to play a key role in acceleration of determinant in cellular survival and conceivably implicated in the drug-induced senescence, particularly in drug-resistant cells, suggest- modulation of cancer cell response to chemotherapy. ing that this protease may act as a survival molecule associated with Treatment of doxorubicin-resistant cells SKN-SH/R with the com- development of drug resistance. Control of intracellular drug distri- bination Q-VD-OPH/doxorubicin (each drug at sublethal concentra- bution and/or cleavage and inactivation of p21/WAF1 represent po- tions) forced cells to undergo cell cycle arrest with phenotypic fea- tential explanations for the protective action of cathepsin L. This 1779

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2004 American Association for Cancer Research. CATHEPSIN L INHIBITION REVERSES DRUG RESISTANCE novel property could validate cathepsin L as a potential therapeutic 20. Dimri, G. P., Lee, X., Basile, G., Acosta, M., Scott, G., Roskelley, C., Medrano, E. E., target to enhance chemotherapy efficacy. Linskens, M., Rubelj, I., and Pereira-Smith, O. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc. Natl. Acad. Sci. USA, 92: 9363–9367, 1995. ACKNOWLEDGMENTS 21. Chang, B. D., Swift, M. E., Shen, M., Fang, J., Broude, E. V., and Roninson, I. B. Molecular determinants of terminal growth arrest induced in tumor cells by a We thank Roberta Gerard and Sandra Clark for assistance in the preparation chemotherapeutic agent. Proc. Natl. Acad. Sci. USA, 99: 389–394, 2002. 22. Miller, R. J., and Wilson, S. M. Neurological disease: UPS stops delivering! Trends of the manuscript. Pharmacol. Sci., 24: 18–23, 2003. 23. Turk, V., Turk, B., and Turk, D. Lysosomal cysteine proteases: facts and opportuni- REFERENCES ties. EMBO J., 20: 4629–4633, 2001. 24. Adams, J. Proteasome inhibitors as new anticancer drugs. Curr. Opin. Oncol., 14: 1. Kroemer, G., and Reed, J. C. Mitochondrial control of cell death. Nat. Med., 6: 628–634, 2002. 513–519, 2000. 25. McNaught, K. S., Olanow, C. W., Halliwell, B., Isacson, O., and Jenner, P. Failure of 2. Reed, J. C. Apoptosis-targeted therapies for cancer. Cancer Cell, 3: 17–22, 2003. the ubiquitin-proteasome system in Parkinson’s disease. Nat. Rev. Neurosci., 2: 3. Friesen, C., Herr, I., Krammer, P. H., and Debatin, K. M. Involvement of the CD95 589–594, 2001. (APO-1/FAS) receptor/ligand system in drug-induced apoptosis in leukemia cells. 26. Isahara, K., Ohsawa, Y., Kanamori, S., Shibata, M., Waguri, S., Sato, N., Gotow, T., Nat. Med., 2: 574–577, 1996. Watanabe, T., Momoi, T., Urase, K., Kominami, E., and Uchiyama, Y. Regulation of 4. Friesen, C., Fulda, S., and Debatin, K. M. Cytotoxic drugs and the CD95 pathway. a novel pathway for cell death by lysosomal aspartic and cysteine proteinases. Leukemia (Baltimore), 13: 1854–1858, 1999. Neuroscience, 91: 233–249, 1999. 5. Debatin, K. M., Poncet, D., and Kroemer, G. Chemotherapy: targeting the mitochon- 27. Hishita, T., Tada-Oikawa, S., Tohyama, K., Miura, Y., Nishihara, T., Tohyama, Y., drial cell death pathway. Oncogene, 21: 8786–8803, 2002. Yoshida, Y., Uchiyama, T., and Kawanishi, S. Caspase-3 activation by lysosomal 6. Ashkenazi, A. Targeting death and decoy receptors of the tumour-necrosis factor enzymes in cytochrome c-independent apoptosis in myelodysplastic syndrome- superfamily. Nat. Rev. Cancer, 2: 420–430, 2002. derived cell line P39. Cancer Res., 61: 2878–2884, 2001. 7. Lowe, S. W., Ruley, H. E., Jacks, T., and Housman, D. E. p53-dependent apoptosis 28. Stoka, V., Turk, B., Schendel, S. L., Kim, T. H., Cirman, T., Snipas, S. J., Ellerby, modulates the cytotoxicity of anticancer agents. Cell, 74: 957–967, 1993. L. M., Bredesen, D., Freeze, H., Abrahamson, M., Bromme, D., Krajewski, S., Reed, 8. Brown, J. M., and Wouters, B. G. Apoptosis, p53, and tumor cell sensitivity to J. C., Yin, X. M., Turk, V., and Salvesen, G. S. Lysosomal protease pathways to anticancer agents. Cancer Res., 59: 1391–1399, 1999. apoptosis. Cleavage of bid, not pro-caspases, is the most likely route. J. Biol. Chem., 9. Borst, P., Borst, J., and Smets, L. A. Does resistance to apoptosis affect clinical 276: 3149–3157, 2001. response to antitumor drugs? Drug Resist. Update, 4: 129–131, 2001. 29. Guay, J., Falgueyret, J. P., Ducret, A., Percival, M. D., and Mancini, J. A. Potency and 10. Roninson, I. B., Broude, E. V., and Chang, B. D. If not apoptosis, then what? selectivity of inhibition of cathepsin K, L and S by their respective propeptides. Eur. Treatment-induced senescence and mitotic catastrophe in tumor cells. Drug Resist. J. Biochem., 267: 6311–6318, 2000. Update, 4: 303–313, 2001. 11. Roninson, I. B. Oncogenic functions of tumour suppressor p21(Waf1/Cip1/Sdi1): 30. Jin, Y. H., Yoo, K. J., Lee, Y. H., and Lee, S. K. Caspase 3-mediated cleavage of association with cell senescence and tumour-promoting activities of stromal fibro- p21WAF1/CIP1 associated with the cyclin A-cyclin-dependent kinase 2 complex is a blasts. Cancer Lett., 179: 1–14, 2002. prerequisite for apoptosis in SK-HEP-1 cells. J. Biol. Chem., 275: 30256–30263, 12. Hartmann, A., Troadec, J. D., Hunot, S., Kikly, K., Faucheux, B. A., Mouatt-Prigent, 2000. A., Ruberg, M., Agid, Y., and Hirsch, E. C. Caspase-8 is an effector in apoptotic death 31. Hurwitz, S. J., Terashima, M., Mizunuma, N., and Slapak, C. A. Vesicular anthra- of dopaminergic neurons in Parkinson’s disease, but pathway inhibition results in cycline accumulation in doxorubicin-selected U-937 cells: participation of lysosomes. neuronal necrosis. J. Neurosci., 21: 2247–2255, 2001. Blood, 89: 3745–3754, 1997. 13. Kim, S. O., and Han, J. Pan-caspase inhibitor zVAD enhances cell death in 32. Stypmann, J., Glaser, K., Roth, W., Tobin, D. J., Petermann, I., Matthias, R., Monnig, RAW246.7 macrophages. J. Endotoxin Res., 7: 292–296, 2001. G., Haverkamp, W., Breithardt, G., Schmahl, W., Peters, C., and Reinheckel, T. 14. Luschen, S., Ussat, S., Scherer, G., Kabelitz, D., and Adam-Klages, S. Sensitization Dilated cardiomyopathy in mice deficient for the lysosomal cysteine peptidase to death receptor cytotoxicity by inhibition of fas-associated death domain protein cathepsin L. Proc. Natl. Acad. Sci. USA, 99: 6234–6239, 2002. (FADD)/caspase signaling. Requirement of cell cycle progression. J. Biol. Chem., 33. Felbor, U., Kessler, B., Mothes, W., Goebel, H. H., Ploegh, H. L., Bronson, R. T., and 275: 24670–24678, 2000. Olsen, B. R. Neuronal loss and brain atrophy in mice lacking cathepsins B and L. 15. Martinsson, P., Liminga, G., Nygren, P., and Larsson, R. Characteristics of etoposide- Proc. Natl. Acad. Sci. USA, 99: 7883–7888, 2002. induced apoptotic cell death in the U-937 human lymphoma cell line. Anticancer 34. Roth, W., Deussing, J., Botchkarev, V. A., Pauly-Evers, M., Saftig, P., Hafner, A., Drugs, 12: 699–705, 2001. Schmidt, P., Schmahl, W., Scherer, J., Anton-Lamprecht, I., Von Figura, K., Paus, R., 16. Yaginuma, H., Shiraiwa, N., Shimada, T., Nishiyama, K., Hong, J., Wang, S., Momoi, and Peters, C. Cathepsin L deficiency as molecular defect of furless: hyperprolifera- T., Uchiyama, Y., and Oppenheim, R. W. Caspase activity is involved in, but is tion of keratinocytes and perturbation of hair follicle cycling. FASEB J., 14: 2075– dispensable for, early motoneuron death in the chick embryo cervical spinal cord. 2086, 2000. Mol. Cell Neurosci., 18: 168–182, 2001. 35. Benavides, F., Starost, M. F., Flores, M., Gimenez-Conti, I. B., Guenet, J. L., and 17. Coelho, D., Holl, V., Weltin, D., Lacornerie, T., Magnenet, P., Dufour, P., and Conti, C. J. Impaired hair follicle morphogenesis and cycling with abnormal epider- Bischoff, P. Caspase-3-like activity determines the type of cell death following mal differentiation in nackt mice, a cathepsin L-deficient mutation. Am. J. Pathol., ionizing radiation in MOLT-4 human leukaemia cells. Br. J. Cancer, 83: 642–649, 161: 693–703, 2002. 2000. 36. Hashmi, S., Britton, C., Liu, J., Guiliano, D. B., Oksov, Y., and Lustigman, S. 18. Rebbaa, A., Zheng, X., Chou, P., and Mirkin, B. L. Caspase inhibition switches Cathepsin L is essential for embryogenesis and development of Caenorhabditis doxorubicin-induced apoptosis to senescence. Oncogene, 22: 2805–2811, 2003. elegans. J. Biol. Chem., 277: 3477–3486, 2002. 19. Rebbaa, A., Chou, P. M., and Mirkin, B. L. Factors secreted by human neuroblastoma 37. Wang, E., Lee, M. D., and Dunn, K. W. Lysosomal accumulation of drugs in mediated doxorubicin resistance by activating STAT3 and inhibiting apoptosis. Mol. drug-sensitive MES-SA but not multidrug-resistant MES-SA/Dx5 uterine sarcoma Med., 7: 393–400, 2001. cells. J. Cell Physiol., 184: 263–274, 2000.

1780

Xin Zheng, Pauline M. Chou, Bernard L. Mirkin, et al.

Cancer Res 2004;64:1773-1780.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/5/1773

Cited articles This article cites 36 articles, 13 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/5/1773.full#ref-list-1

Citing articles This article has been cited by 10 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/5/1773.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/64/5/1773. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.