Você está na página 1de 7

Carcinogenesis vol.29 no.3 pp.466472, 2008 doi:10.

1093/carcin/bgm212 Advance Access publication October 4, 2007

The alarm anti-protease, secretory leukocyte protease inhibitor, is a proliferation and survival factor for ovarian cancer cells
Fiona A.Simpkins1, Nick M.Devoogdt1,2, Nabila Rasool1, Nana E.Tchabo1, Emilyn U.Alejandro1, Mitchell M.R.N.Kamrava1 and Elise C.Kohn1,
1 Molecular Signaling Section, Laboratory of Pathology, Center for Cancer Research, National Cancer Institute, Bethesda, MD 20892 and 2Fonds voor Wetenschappelijk Onderzoek Vlaanderen Fellow, Department Molecular and Cellular Interactions, Vlaams interuniversitair Insituut voor Biotechnologie, Vrije Universiteit Brussel, Brussels, Belgium To whom correspondence should be addressed. Tel: 1 301 402 2726; Fax: 1 301 480 5142; Email: ek1b@nih.gov

Alarm anti-proteases are secreted locally in response to inammation and have been shown to be elevated in cancers. Secretory leukocyte protease inhibitor (SLPI), an alarm anti-protease, is amplied in ovarian carcinoma and is induced and binds to and protects progranulin (prgn) in inammation. We reported prgn is a survival protein in ovarian cancer and now hypothesize that SLPI/prgn would promote proliferation and survival. Neutralizing anti-SLPI antibody treatment of HEY-A8 and OVCAR3 ovarian cancer cells decreased cell number (P < 0.001), induced apoptosis and reduced prgn quantity. This was conrmed using SLPI small interfering RNA. Prgn and SLPI were co-immunoprecipitated and co-localized by confocal microscopy. Prgn is a substrate of the serine protease elastase and SLPI is an inhibitor of elastase. Elastase reduced prgn expression, inhibited proliferation in a dosedependent manner (P 0.01) and was pro-apoptotic. SLPI protected prgn from elastase-mediated degradation and restored its survival and proliferative function (P 0.04). SLPI also reversed elastases pro-apoptotic effects (P 0.03), yielding recovery of Sphase fraction (P 0.001) and increased cyclin D1. Treatment with a general serine protease inhibitor increased prgn, but did not reverse elastase-mediated prgn loss or apoptosis. These data demonstrate that inappropriate over-expression of the alarm antiprotease, SLPI, creates a pro-survival milieu for ovarian cancer.

Ovarian cancer is the fth most common cause of death from cancer in women (7). SLPI has been described as amplied and over-expressed in ovarian carcinoma (1,811). Its messenger RNA and protein have been reported up-regulated in ovarian tumors compared with normal surface epithelium (12). SLPI and other whey acidic protein alarm anti-proteases, such as elan and HE-4, have been proposed as putative serum biomarkers for malignant ovarian masses (1316). HE-4 has been shown to be expressed in ovarian inclusion cysts lined by metaplastic epithelium and to be expressed highly in serous and endometriod ovarian cancers (13). Together these data suggest that SLPI, as an alarm anti-protease, may have contextdependent function. We hypothesize that SLPI has a pro-cancer function in ovarian cancer, independent of its anti-protease activity. We identied progranulin (prgn), a 68 kDa protein with multiple polyglycosylated higher molecular weight isoforms, to be up-regulated in invasive serous ovarian carcinoma (1719). It was present selectively in stage III epithelial ovarian cancer compared with serous borderline tumor. Blocking prgn expression by transfection of OVCAR3 human ovarian cancer cells with anti-sense prgn reduced monolayer and density-independent cell growth (19) and caused apoptosis (20). Prgn has also been shown to be induced in broblasts and endothelial cells after injury to promote proliferation, migration and formation of capillary-like structures, indicating a role for prgn in angiogenesis (21). It was also shown to be present in activated blood vessels in ovarian cancers, co-localizing with perlecan (22). The results indicate a role for prgn in growth and survival of ovarian cancer. Zhu et al. (3) describes a balanced interaction between prgn and SLPI necessary for wound healing, where SLPI, prgn and elastase elaborated by activated neutrophils come together to regulate the proinammatory microenvironment. That both SLPI and prgn are overexpressed in ovarian cancer led us to propose that these two proteins may function to support the tumorigenic and malignant activity of this cancer. We now report SLPI promotes proliferation and survival of ovairan cancer cells. It furthers their survival through partnering with and protection of prgn. Materials and methods

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

Introduction Anti-proteases are produced and secreted to function as systemic monitors, or acute alarm responders, increasing in response to local cytokine production (1,2). They have been described as being altered in expression in multiple pathologies including cancers. Secretory leukocyte protease inhibitor (SLPI), a whey acidic protein, is an alarm anti-protease, demonstrated to be synthesized and secreted by inammatory and stromal cells in the local microenvironment. It has been shown in a wound-healing model to be a pro-inammatory protein, in response to and inducing cytokine production and local inammation (3,4). Alarm anti-proteases have also been described as increased in a fashion not directly related to anti-protease activity in malignancy and metastasis (1). SLPI over-expression in Lewis lung cancer cells was shown to promote tumorigenic and metastatic potential in a syngeneic mouse model; this activity was dependent upon its antiprotease activity (5). However, SLPI was shown also to reduce the production of pro-inammatory cytokines, E-selectin and results in reduced Lewis lung carcinoma metastasis formation in the liver in a different model, arguing it is a metastasis inhibitor (6).
Abbreviations: CM, conditioned medium; prgn, progranulin; GAPDH, glyceraldehyde dehydrogenase; siRNA, small interfering RNA; SLPI, secretory leukocyte protease inhibitor; TAME, p-toluene sulfonyl)-L-arginine methyl ester; TNF-a, tumor necrosis factor-a; XTT, -3#-[1-(phenylamino-carbonyl)3-4-tetrazolium]-bis (4-methoxy-6-nitro) benzene sulfonic acid hydrate. Published by Oxford University Press 2007.

Reagents Recombinant human SLPI (rhSLPI) and goat anti-rhSLPI antibody were purchased from R&D Systems (Minneapolis, MN). Human neutrophil elastase was acquired from Calbiochem (La Jolla, CA). ELISA kits measuring SLPI and elastase were from Cell Sciences (Canton, MA). Precast PAGE gels were from Invitrogen (San Diego, CA). XTT (-3#-[1-(phenylamino-carbonyl)-3-4tetrazolium]-bis (4-methoxy-6-nitro) benzene sulfonic acid hydrate) labeling agent was obtained from Roche Applied Science (Indianapolis, IN). DAPI vectashield mounting medium was purchased from Vector Laboratories (Burlingame, CA) and propidium iodine, donkey anti-goat alexauor-546 and donkey anti-rabbit alexauor 488 were obtained from Molecular Probes (Eugene, OR). Small interfering RNA (siRNA) against SLPI, control siRNA and transfection reagents were obtained from Dharmacon (Lafayette, CO). Antibody to glyceraldehyde dehydrogenase (GAPDH) was from Abcam (Cambridge, MA). All other reagents were of molecular or imaging grade. A peptidespecic anti-SLPI antibody was generated against the peptide DTPNPTRRKPGKC in the A2 segment of SLPI (19,23). This antibody was peptide-puried, titered for optimal immunoprecipitation and immunoblot and peptide competed for validation prior to use. The BAG3 antibody has been reported (2426). Cell culture Ovarian cancer cells expressing a range of SLPI were selected for study. The OVCAR3 human ovarian cancer cell line was obtained from the American Type Culture Collection (Manassas, VA). HEY-A8 and SKOV3 lines were a generous gift of Dr G.Mills (M.D. Anderson Cancer Center, Houston, TX). Human ovarian surface epithelium cultures were a generous gift of Dr B.Karlan (Cedars Sinai Cancer Center, Los Angeles, CA). All ovarian cancer cell lines were cultured in RPMI 1640 with 10% fetal bovine serum unless otherwise

466

SLPI is a survival factor in ovarian cancer

indicated. Cells were cultured in serum-free medium for at least 6 h, washed and treated with indicated reagents, including neutralizing antibodies, in serum-free medium for 24 h unless otherwise indicated. Sodium acetate 50 mM, NaCl 200 mM (pH 5) and Tris 100 mM, CaCl2 10 mM, 0.1% fetal bovine serum (pH 7.5) were used as vehicle controls for elastase and rhSLPI, respectively. A rabbit polyclonal IgG control was included when cells were treated with peptide-puried anti-peptide antibodies; control IgG and anti-SLPI concentrations added were 30 lg and up to 22 lg, respectively, based on an antiSLPI concentration of 1.1 lg/ll. Conditioned medium (CM) was claried by centrifugation and then protease inhibitors (PI: aprotinin 10 lg/ml, leupeptin 10 lg/ml and phenylmethylsulfonyl uoride 1 mM) were added. CM was then frozen until used. Where indicated, SLPI concentration in CM was measured using Quantikine Human SLPI Immunoassay kit (R&D Systems) according to the manufacturers instructions. Incubations with p-toluene sulfonyl)-L-arginine methyl ester (TAME) mirrored those for the neutralizing antibodies or elastase. Gene silencing SKOV3 cells, expressing the highest amount of SLPI, were selected for these experiments. In some experiments, cells were plated on gelatin-coated circular cover glass slides for later immunouorescence. Dharmafect 1 transfection reagent was added to a nal concentration of 1.0 ll/ml, with a non-targeting siRNA control or anti-SLPI siGenome SMARTpool siRNA at concentrations of 20 nM in antibiotic-free medium. Cells were incubated at 37C for 96 h and subjected to analysis. Antibody production, immunoblot and immunoprecipitation SLPI rabbit anti-peptide antibody was generated, puried and validated as described for anti-prgn antibodies (19,23). Peptide competition conrmed recognition of the 14 kDa SLPI and a dimer of 28 kDa (27). Total cell lysates were prepared using modied RIPA buffer, sonicated and aliquoted for use as reported (20). Proteins and CM underwent no more than one freeze/thaw cycle. Unconcentrated CM (1 ml) or lysate (500 lg) was immunoprecipitated overnight with anti-prgn Ab737 as described (20,23) at 4C and then captured with protein A/G beads. Beads were washed three times with NP-40 lysis buffer (PI 20 mM TrisHCL, pH 8, 137 mM NaCl, 10% glycerol, 1% NP-40, 2 mM ethylenediaminetetraacetic acid) and twice with phosphate-buffered saline PI. Immunoprecipitates were subjected to reducing gel electrophoresis followed by immunoblot with anti-SLPI. The protein bands of interest were visualized using Super Signal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL). All blots are representative of at least three replicate experiments. Where shown, parallel gels were immunoblotted with antibody to GAPDH as a putative housekeeping protein to show balance for loading. Fluorescence microscopy Cells were grown overnight on 0.1% gelatin-coated cover slips, serum starved for 6 h, and then treated with indicated reagents in a serum-free background for up to 24 h or transfected with SLPI or control siRNA as described. Cells were then xed in 3.7% formaldehyde and permeabilized with 0.1% Triton X-100. Cover slips were stained with anti-prgn and/or anti-SLPI antibodies where indicated and mounted with DAPI stain mounting medium. Immunouorescent images were acquired on a Zeiss 510 LSM confocal microscope at 63/ 1.4 NA with an oil Differential interference contrast-Nomarski objectives and scan zoom of 1. The results presented are representative of at least three independent experiments. Flow cytometric analyses of apoptosis Cells were cultured and treated as described for microscopy. Floating and adherent cells were collected, pooled and xed with 70% ethanol overnight, then RNAase 1000 U/ml nal volume was added. Fixed cells were stained with propidium iodide 50 lg/ml and subjected to ow cytometric analysis using CellQuest (BD Biosciences, San Jose, CA). FlowJo (Tree Star, San Carlos, CA) and ModFit software (Verity Software House, Topsham, ME) were used for cell cycle analysis. Statistical analysis Unpaired Students t-tests were applied to replicate experiments (Microsoft Excel, Seattle, WA). Two-sided P values are reported.

highest in SKOV3 (Figure 1A). The treatment of HEY-A8 and OVCAR3 cells with rhSLPI showed insignicant increase in proliferation (not shown). However, immunoneutralization of SLPI resulted in a net loss of both cell types in a dose-dependent fashion (P , 0.001; Figure 1B). SKOV3 cells, which secrete larger concentrations SLPI, are inhibited by neutralizing antibody but to a lesser extent (20% reduction; P 0.03; data not shown). This doseresponse effect is consistent with the competitive nature of neutralizing antibodies. Anti-SLPI antibody induced nuclear degeneration, blebbing and apoptosis in a 16 h exposure (Figure 1C; blank IgG negative control not included). Apoptosis was conrmed by ow cytometry demonstrating a 10-fold and 23-fold increase in DNA fragmentation in HEY-A8 (0.2 versus 2.2% subG0G1 fraction; P 5 0.005) and OVCAR3 cells (2 versus 5% subG0G1 fraction; P 5 0.02). We have reported previously that treatment with anti-sense prgn and anti-prgn antibody caused decreased proliferation and induced apoptosis, indicating that prgn is a survival factor in ovarian cancer (19,20). Both anti-prgn and anti-SLPI antibodies cause loss of prgn protein, with accumulation of a previously conrmed 50 kDa fragment (Figure 1D) (20). Concomitant exposure to SLPI protects cells from anti-prgn mediated loss of prgn. The mechanism through which SLPI protects prgn from loss due to immunoneutralization with anti-SLPI or anti-prgn is not clear. It may be related to binding site competition between the antibody and either prgn or SLPI. RNA interference of SLPI reduces ovarian cancer cell survival and prgn quantity RNAi was used to conrm the role of SLPI on ovarian cancer cell prgn and survival. Figure 2 shows marked reduction of SLPI (red) and prgn (green) in cells exposed to SLPI siRNA with some relocalization. Co-localization of SLPI and prgn is overlapping in the control siRNA-treated cells, but altered and reduced in cells treated with SLPI siRNA, also shown by the merge and the higher power insets. The two siRNA sets of panels show different elds of view from replicate experiments. These cells have reduced quantity and globular organization of SLPI, and appear to have more nuclear disruption and apoptotic changes by DAPI stain. The general co-localization of SLPI and prgn also is reduced in late apoptotic cells (arrow heads). These cells are shown to have initiated DNA fragmentation, nuclear blebbing and chromatin condensation. This conrms SLPI-mediated protection of prgn and cell survival. The specicity of silencing of SLPI to SLPI and prgn were tested by examination of expression of BAG3, a stress co-chaperone protein under investigation in our laboratory (2426) and of GAPDH (in part as a loading control; Figure 2B). Neither BAG3 nor GAPDH were reduced under conditions in which both SLPI and prgn were diminished. SLPI and prgn are binding partners We next asked if both SLPI and prgn interact in ovarian cancer cells and their local microenvironment. A SLPI/prgn complex was shown by co-immunoprecipitation of both lysates and CM from OVCAR3 and SKOV3 cells, those containing the greatest quantity of SLPI (Figure 3A). More SLPI/prgn complex was present in CM consistent with the secretion of both proteins. Binding was conrmed by reverse order immunoprecipitation (data not shown). Co-localization was further demonstrated using uorescence confocal microscopy (Figure 3B). Prgn and SLPI are both cytosolic, found in the region of the golgi, typical of secreted proteins. SLPI reverses both anti-proliferation and pro-apoptotic effects of elastase Prgn has been reported to be susceptible to cleavage by serine proteases such as elastase, which was shown to be necessary for the balance of prgn and SLPI in a wound-healing model (3). The ovarian cancer cell lines produce no detectable elastase measurable by ELISA (data not shown). Therefore, puried elastase was introduced to cultures for 24 h causing a marked and dose-dependent decrease in prgn protein in both HEY-A8 and OVCAR3 cells (Figure 4A, left panel). Addition of SLPI protected prgn from elastase shown both by

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

Results SLPI protein is essential for ovarian cancer cell growth SLPI protein was present in concentrations of 27 ng/ml in OVCAR3 and SKOV3 cell lysates and CM; concentrations were below the level of detection (,0.5 ng/ml) in HEY-A8 cell lysates and CM. Immunoblot shows undetectable SLPI protein in human ovarian surface epithelium primary culture, minimal presence in HEY-A8 cells and

467

F.A.Simpkins et al.

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

Fig. 1. Neutralizing anti-SLPI antibody decreases ovarian cancer cell proliferation and prgn production and induces apoptosis. (a) SLPI is present in ovarian cancer cell lines. Primary culture HOSE cells do not produce SLPI, while it is detectable in lysate of malignant cells. Lanes 1: recombinant huSLPI; 2: HOSE; 3: HEY-A8; 4: OVCAR3; 5: SKOV3. (b) Neutralizing polyclonal anti-SLPI antibody inhibits proliferation of Hey-A8 and OVCAR3. Cell viability was measured using the XTT assay for cells starved for 24 h then exposed to control IgG or anti-SLPI antibody for 48 h. Data shown represents the mean SEM of at least three independent experiments; P value is comparison to control. (c) Apoptosis is demonstrated by DAPI immunouorescence. Reduced cell number, nuclear degradation and apoptotic bodies are present when cells were exposed to neutralizing anti-SLPI antibody for 16 h. A shorter time was selected due to cell loss. Inset: apoptotic bodies. (d) Prgn is protected by SLPI. The left panel shows loss of prgn in HEY-A8 cells in the presence of anti-prgn antibody and recovery when SLPI is added. Prgn is lost when the anti-SLPI antibody is used in HEY-A8 cells (right panel). The 50 kDa prgn band is a previously documented prgn fragment (20).

Fig. 2. SLPI siRNA selectively reduces SLPI and prgn protein and ovarian cancer cell survival. (a) siSLPI reduces expression of both SLPI (red) and prgn (green). Control or SLPI siRNA were introduced and cells incubated for 96 h prior to xation and staining. The right most panel shows the confocal overlay with yellow indicating presence of both proteins. Apoptosis is shown by loss of cells and nuclear changes (DAPI, blue). There is a dissociation of SLPI from prgn seen in apoptotic cells shown by arrow heads. Two independent elds of siRNA-treated cells are shown (rows 2 and 3), representative of at least three replicate transfection experiments. (b) Silencing of SLPI does not reduce BAG3 quantity. Cells were exposed to siSLPI as in A, lysed and subjected to immunoblot for SLPI, prgn and BAG3 as indicated. No loss of BAG3 or GAPDH (as loading control) is seen under conditions where SLPI and prgn are markedly reduced.

immunoblot (Figure 4A, right panel) and confocal imaging (Figure 4B). These data demonstrate the ability of SLPI to protect prgn from serine protease-mediated degradation. The functional consequence of SLPI protection of prgn was investigated next. Figure 4B shows loss of apoptosis in cells exposed to elastase concomitantly with SLPI. This was further measured in a dose-dependent fashion in all three cell

lines using an XTT assay (P 0.01; Figure 4C). Cell cycle analysis was done to determine whether this was due to the anti-proliferative and/or pro-apoptotic effects of elastase. Treatment with elastase signicantly decreased the S-phase fraction in HEY-A8 and OVCAR3 cells (Table I). Addition of SLPI to elastase partially reversed the loss of DNA synthesis, with little proliferative activity of its own.

468

SLPI is a survival factor in ovarian cancer

This effect was further evaluated by assessing cyclin D1 (Figure 4D). In parallel with the other ndings, there is a reduction in cyclin D1 expression in elastase-exposed cells that is restored when SLPI is added with elastase. Thus, SLPI reversed both the negative effects of elastase on proliferation and its pro-apoptotic activity. SLPI reverses elastase effects through prgn interaction and not by protease inhibition SLPI has been shown to bind both prgn and elastase, providing two opposing mechanisms for its protection of prgn (3). TAME, a general serine protease inhibitor, was used to address the protease inhibition function. Exposure of HEY-A8 and OVCAR3 cells to increasing concentrations of TAME under the same experimental conditions in which the elastase or SLPI were included caused small but reproducible increases in prgn in a dose-dependent manner, suggesting the involvement of a cellular serine protease in local prgn regulation (Figure 5A). However, TAME treatment did not protect prgn from elastase-mediated degradation (Figure 5B), strongly suggesting that SLPI protects prgn through a mechanism unrelated to its protease inhibitory activity. TAME also did not protect cells from elastasemediated inhibition of proliferation although SLPI signicantly overcame elastase-mediated injury (Figure 5C). This demonstrates that the

Fig. 3. SLPI and prgn are binding partners. (a) SLPI and prgn coimmunoprecipitate. Both cell lysates and CM were subjected to immunoprecipitation with anti-prgn and probed for SLPI. Similar results were seen with anti-SLPI immunoprecipitation followed by immunoblot with anti-prgn (not shown). (b) Confocal imaging demonstrates cytosolic colocalization of prgn and SLPI.

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

Fig. 4. SLPI protects cells from elastase-mediated apoptosis. (a) Prgn is markedly reduced when cells are exposed to elastase in a dose-dependent fashion. This loss is reversed upon concomitant exposure to SLPI (right panel). (b) Elastase-mediated injury is reversed by SLPI. The left panel conrms the loss and recovery of prgn upon elastase treatment in the absence and presence of SLPI. Elastase apoptotic injury and SLPI-associated protection from injury is seen under the same conditions. (c) SLPI reverses elastase injury in a dose-dependent fashion in HEY-A8 cells. XTT assay is used to quantitate cell injury with elastase and recovery with concomitant SLPI (mean and SEM, n 5 3; similar results are shown for OVCAR3 and SKOV3). (d) Cyclin D1 expression is reduced when cells are exposed to elastase and restored with co-treatment with SLPI.

469

F.A.Simpkins et al.

protective role of SLPI on prgn and on cell proliferation is independent of its protease inhibitory activity. Thus, both SLPI and prgn have key roles in ovarian cancer but may synergize to make a locally permissive microenvironment. Discussion Genomic analysis of ovarian cancer identied amplication of SLPI and the whey acidic protein locus in a large proportion of epithelial ovarian cancers (28,29). Multiple functions of SLPI have been reported previously, including function as a protease inhibitor and as a regulator of local inammatory responses downstream of chemokines and tumor necrosis factor-a (TNF-a) (5,30). One mechanism of its anti-inammatory activity is through inhibition of cleavage of prgn into component pro-inammatory granulins (3). We identied over-

Table I. Elastase alters proliferation and apoptosis in a SLPI-sensitive fashiona % S phase HEY-A8 Control SLPI 10 lg/ml Elastase1 lg/ml SLPI elastase OVCAR3 Control SLPI 10 lg/ml Elastase1 lg/ml SLPI elastase
a c b

P value

% SubG0G1 0.28 0.23 1.7 0.95 2.05 2.51 4.6 2.5

P value

33 34 21 37 16 26 13 18

0.002b ,0.001c

0.04b 0.026c

0.02b 0.003c

0.008b 0.007c

Average of three independent experiments. Control versus elastase. Elastase versus SLPI elastase.

expression of prgn in epithelial ovarian cancer and demonstrated that it is both a proliferation and survival factor for ovarian cancer cell lines (19,20). The ndings linking prgn and SLPI in inammation led to the hypothesis that they would interact in malignancy to promote stabilization of the pro-survival prgn form. We now show that SLPI itself has pro-survival function, that by direct interaction it stabilizes prgn, and that SLPI protects prgn from serine protease-mediated degradation. This interaction makes a more potent anti-apoptotic and progrowth environment for the ovarian cancer cells. Further, SLPI is an independent pro-survival factor for ovarian cancer, as shown by the stimulation of apoptosis in response to its neutralization and silencing. Thus, these results identify prgn and SLPI as putative molecular therapeutic targets in ovarian cancer. Zhu et al. (3) proposed a balance necessary between SLPI-prgn and elastase for normal tissue growth. Williams et al. (1) describe SLPI as an alarm anti-protease secreted in response to inammatory cytokines, such as the interleukins and TNF-a. Ovarian cancer is known to produce large amounts of and a large variety of pro-inammatory cytokines that also behave as pro-invasive and pro-angiogenic factors (3133). Thus, we proposed in our ovarian cancer model that there is an imbalance in SLPI caused by the cancer and not the alarm response to local inammation. SLPI over-expression is caused both by amplication and over-expression (11) as well as its production in response to changes in the local microenvironment (1). The behavior of SLPI in this setting would be expected to be independent of its anti-protease activity. This is conrmed by the pro-survival activity of SLPI by itself and its serine protease-independent protection of prgn. These results suggest that interruption of the prgn/SLPI interaction may be targeted for therapeutic gain without altering its immunomodulatory role. This theme is further supported by the observation that introduction of SLPI can markedly inhibit apoptosis caused by elastase, also independent of elastase protease activity. Immunoneutralization of SLPI caused apoptosis of the ovarian cancer cells similar to that seen

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

Fig. 5. The protective effects of SLPI are not caused by its protease inhibitory activity. (a) The general serine protease inhibitor, TAME, increases prgn quantity in a dose-dependent manner in both HEY-A8 and OVCAR3 cells. (b) Immunouorescence demonstrates protection of prgn only when SLPI is added to elastase. (c) SLPI but not TAME protects cells from elastase-mediated inhibition of proliferation. Elastase versus SLPI elastase: P 5 0.01.

470

SLPI is a survival factor in ovarian cancer

with elastase and immunoneutralization of prgn (20). One mechanism of this may be the small but dose-dependent increase in prgn observed when cells are incubated with SLPI. It may be due to the protective role of SLPI on prgn, such that loss of SLPI may leave prgn open to cleavage by elastase or other, as yet undened serine proteases. The ndings with recombinant SLPI also may be limited by the nature of the bacterial recombinant protein (11 kDa) that may not be recognized as physiologically as the endogenously produced glycoslylated form (14 kDa). It is possible that there may be differential activity of the modied forms of SLPI; we found more expression of the higher molecular weight SLPI in the HEY-A8 cells which also seemed more susceptible to neutralization of SLPI than would be expected by their relatively lower secreted concentrations. These ndings are supported by the increase in prgn quantity in the presence of TAME. SLPI appears to have its pro-malignant effects more through inhibition of cell death than promotion of cell proliferation in ovarian cancer, contrary to reports assessing its role in other cancers (6,8,30,34,35). Preliminary data indicates that SLPI has a limited direct proliferative drive on these ovarian cancer cells in the absence of elastase or neutralizing antibody, despite the marked protective effects observed when the injuring agent is present. Findings in the literature indicate that the anti-protease-independent role for SLPI in cancer may be either context or cancer dependent. SLPI has been implicated in HIV disease in an anti-proteaseindependent fashion (3638). Induction of SLPI in response to inammatory cytokines has been shown in several systems (39,40). TNF-a, produced by many types of cancer stimulates SLPI production (41). Devoogdt et al. (5) showed previously that TNF-a producing macrophages enhanced resistance to TNF-a mediated lysis and promoted malignancy of 3LL cells associated with up-regulation of SLPI. Further studies have revealed that although induction of SLPI occurred during TNF-a treatment, the promotion of tumor progression by SLPI abrogated tumor advance in a TNF-a-independent fashion (5,30,41,42). Thus, induction of SLPI in the microenvironment whether from tumor or local inammatory cells can shift the local balance to favor tumor progression. SLPI and prgn have both been shown to be over-expressed in ovarian cancer. This over-expression is now reduced to function showing that these proteins have proliferative and anti-apoptotic activity independently, but that they work in concert to yield a stronger effect as a partnered complex to promote survival and proliferation of the malignant cells. Interruption of this complex is a logical molecular therapeutic target and one that may also be monitored as a biomarker due to the secreted nature of both proteins. Funding Intramural Research Program of the National Institutes of Health; National Cancer Institute; Center for Cancer Research; Howard Hughes Medical Institute/National Institutes of Health Research Scholar to M.K. Acknowledgements
Conict of Interest Statement: None declared.

References
1. Williams,S.E. et al. (2006) SLPI and elan: one glove, many ngers. Clin. Sci., 110, 2135. 2. King,A.E. et al. (2003) Differential regulation of secretory leukocyte protease inhibitor and elan by progesterone. Biochem. Biophys. Res. Commun., 310, 594599. 3. Zhu,J. et al. (2002) Conversion of proepithelin to epithelins: roles of SLPI and elastase in host defense and wound repair. Cell, 111, 867878. 4. Ding,A. et al. (2005) Induction of macrophage-derived SLPI by Mycobacterium tuberculosis depends on TLR2 but not MyD88. Immunology, 116, 381389.

5. Devoogdt,N. et al. (2003) Secretory leukocyte protease inhibitor promotes the tumorigenic and metastatic potential of cancer cells. Proc. Natl Acad. Sci. USA, 100, 57785782. 6. Wang,N. et al. (2006) The secretory leukocyte protease inhibitor is a type 1 insulin-like growth factor receptor-regulated protein that protects against liver metastasis by attenuating the host proinammatory response. Cancer Res., 66, 30623070. 7. Jemal,A. et al. (2006) Cancer statistics. CA Cancer J. Clin., 56, 106130. 8. Bouchard,D. et al. (2006) Proteins with whey-acidic-protein motifs and cancer. Lancet Oncol., 7, 167174. 9. Hough,C.D. et al. (2001) Coordinately up-regulated genes in ovarian cancer. Cancer Res., 61, 38693876. 10. Clauss,A. et al. (2005) The evolution of a genetic locus encoding small serine proteinase inhibitors. Biochem. Biophys. Res. Commun., 333, 383389. 11. Israeli,O. et al. (2005) In silico chromosomal clustering of genes displaying altered expression patterns in ovarian cancer. Cancer Genet. Cytogenet., 160, 3542. 12. Shigemasa,K. et al. (2001) Expression of the protease inhibitor antileukoprotease and the serine protease stratum corneum chymotryptic enzyme (SCCE) is coordinated in ovarian tumors. Int. J. Gynecol. Cancer, 11, 454461. 13. Drapkin,R. et al. (2005) Human epididymis protein 4 (HE4) is a secreted glycoprotein that is overexpressed by serous and endometrioid ovarian carcinomas. Cancer Res., 65, 21622169. 14. Tsukishiro,S. et al. (2005) Use of serum secretory leukocyte protease inhibitor levels in patients to improve specicity of ovarian cancer diagnosis. Gynecol. Oncol., 96, 516519. 15. Bingle,L. et al. (2002) The putative ovarian tumour marker gene HE4 (WFDC2), is expressed in normal tissues and undergoes complex alternative splicing to yield multiple protein isoforms. Oncogene, 21, 27682773. 16. Hellstrom,I. et al. (2003) The HE4 (WFDC2) protein is a biomarker for ovarian carcinoma. Cancer Res., 63, 36953700. 17. Bateman,A. et al. (1990) Granulins, a novel class of peptide from leukocytes. Biochem. Biophys. Res. Commun., 173, 11611168. 18. Bateman,A. et al. (1998) Granulins: the structure and function of an emerging family of growth factors. J. Endocrinol., 158, 145151. 19. Jones,M.B. et al. (2003) The granulin-epithelin precursor/PC-cell-derived growth factor is a growth factor for epithelial ovarian cancer. Clin. Cancer Res., 9, 4451. 20. Kamrava,M. et al. (2005) Lysophosphatidic acid and endothelin-induced proliferation of ovarian cancer cell lines is mitigated by neutralization of granulin-epithelin precursor (GEP), a prosurvival factor for ovarian cancer. Oncogene, 24, 70847093. 21. He,Z. et al. (2003) Progranulin is a mediator of the wound response. Nat. Med., 9, 225229. 22. Gonzalez,E.M. et al. (2003) A novel interaction between perlecan protein core and progranulin: potential effects on tumor growth. J. Biol. Chem., 278, 3811338116. 23. Davidson,B. et al. (2004) Granulin-epithelin precursor is a novel prognostic marker in epithelial ovarian carcinoma. Cancer, 100, 21392147. 24. Doong,H. et al. (2000) CAIR-1/BAG-3 forms an EGF-regulated ternary complex with phospholipase C-gamma and Hsp70/Hsc70. Oncogene, 19, 43854395. 25. Doong,H. et al. (2003) CAIR-1/BAG-3 abrogates heat shock protein-70 chaperone complex-mediated protein degradation: accumulation of polyubiquitinated Hsp90 client proteins. J. Biol. Chem., 278, 2849028500. 26. Kassis,J.N. et al. (2006) CAIR-1/BAG-3 modulates cell adhesion and migration by downregulating activity of focal adhesion proteins. Exp. Cell Res., 312, 29622971. 27. Seemuller,U. et al. (1986) The acid-stable proteinase inhibitor of human mucous secretions (HUSI-I, antileukoprotease). Complete amino acid sequence as revealed by protein and cDNA sequencing and structural homology to whey proteins and Red Sea turtle proteinase inhibitor. FEBS Lett., 199, 4348. 28. Tanner,M.M. et al. (2000) Frequent amplication of chromosomal region 20q12-q13 in ovarian cancer. Clin. Cancer Res., 6, 18331839. 29. Courjal,F. et al. (1996) DNA amplications at 20q13 and MDM2 dene distinct subsets of evolved breast and ovarian tumours. Br. J. Cancer, 74, 19841989. 30. Devoogdt,N. et al. (2004) Secretory leukocyte protease inhibitor in cancer development. Ann. N. Y. Acad. Sci., 1028, 380389. 31. Wang,Y. et al. (2005) Regulatory effect of e2, IL-6 and IL-8 on the growth of epithelial ovarian cancer cells. Cell Mol. Immunol., 2, 365372. 32. Xu,L. et al. (2000) Inhibition of malignant ascites and growth of human ovarian carcinoma by oral administration of a potent inhibitor of the

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

471

F.A.Simpkins et al.

vascular endothelial growth factor receptor tyrosine kinases. Int. J. Oncol., 16, 445454. 33. Yoneda,J. et al. (1998) Expression of angiogenesis-related genes and progression of human ovarian carcinomas in nude mice. J. Natl Cancer Inst., 90, 447454. 34. Velarde,M.C. et al. (2005) The secretory leukocyte protease inhibitor gene is a target of epidermal growth factor receptor action in endometrial epithelial cells. J. Endocrinol., 184, 141151. 35. Westin,U. et al. (2002) The presence of elan, SLPI, IL1-RA and STNFalpha RI in head and neck squamous cell carcinomas and their relation to the degree of tumour differentiation. Mediators Inamm., 11, 712. 36. Jana,N.K. et al. (2005) Human immunodeciency virus type 1 stimulates the expression and production of secretory leukocyte protease inhibitor (SLPI) in oral epithelial cells: a role for SLPI in innate mucosal immunity. J. Virol., 79, 64326440. 37. Farquhar,C. et al. (2002) Salivary secretory leukocyte protease inhibitor is associated with reduced transmission of human immunodeciency virus type 1 through breast milk. J. Infect. Dis., 186, 11731176.

38. Pillay,K. et al. (2001) Secretory leukocyte protease inhibitor in vaginal uids and perinatal human immunodeciency virus type 1 transmission. J. Infect. Dis., 183, 653656. 39. Ashcroft,G.S. et al. (2000) Secretory leukocyte protease inhibitor mediates non-redundant functions necessary for normal wound healing. Nat. Med., 6, 11471153. 40. Lentsch,A.B. et al. (1999) Secretory leukocyte protease inhibitor in mice regulates local and remote organ inammatory injury induced by hepatic ischemia/reperfusion. Gastroenterology, 117, 953961. 41. Devoogdt,N. et al. (2006) The tumor-promoting effect of TNF-alpha involves the induction of secretory leukocyte protease inhibitor. J. Immunol., 177, 80468052. 42. Aarbiou,J. et al. (2004) Human neutrophil defensins and secretory leukocyte proteinase inhibitor in squamous metaplastic epithelium of bronchial airways. Inamm. Res., 53, 230238. Received January 31, 2007; revised September 14, 2007; accepted September 15, 2007

Downloaded from carcin.oxfordjournals.org by guest on February 16, 2011

472

Você também pode gostar