翻译The Prostate Specific Membrane
The Prostate Specific Membrane Antigen Regulates the Expression of IL-6and CCL5in Prostate Tumour Cells by Activating the MAPK Pathways1
Marco Colombatti1,Silvia Grasso1.,Alessandra Porzia3.,Giulio Fracasso1,Maria Teresa Scupoli4,Sara Cingarlini1,Ornella Poffe1,Hassan Y.Naim5,Martin Heine5,Giuseppe Tridente1,2,Fabrizio Mainiero3, Dunia Ramarli2*
1Department of Pathology,University of Verona,Verona,Italy,2Clinical Immunology,Giovanni Battista Rossi Hospital,Verona,Italy,3Department of Experimental Medicine,Institute Pasteur-Fondazione Cenci Bolognetti,University of Rome‘‘La Sapienza’’,Rome,Italy,4Interdepartmental Laboratory for Medical Research(LURM), University of Verona,Verona,Italy,5Department of Physiological Chemistry,School of Veterinary Medicine,Hannover,Germany
Introduction
Prostate cancer is the most commonly diagnosed neoplasia in man in developed countries.Death from prostate cancer occurs largely in patients with the aggressive androgen-insensitive metastatic disease.A number of studies have recently demonstrat-ed that a prominent role in tumor survival and progression can be attributed to soluble mediators present in the tumor microenvi-ronment.Among these,Interleukin-6(IL-6)has a fundamental role in the regulation of proliferation,apoptosis,angiogenesis and differentiation in many cell types and it is also implicated in the development and progression of several forms of tumours including that of the prostate[1,2].In fact,the expression of IL-6and its receptor is consistently demonstrated in human prostate cancer cell lines and in freshly isolated cells from human prostate carcinoma and benign prostate hyperplasia[3,4].Clinically,the levels of IL-6in serum are significantly elevated in many men with advanced,hormone-refractory prostate cancer[5,6].Further,IL-6activates androgen receptor-mediated gene expression in LNCaP cells in vitro[7,8],suggesting that IL-6may play a critical role during the progression of prostate cancer.In addition,over-expression of IL-6in androgen-responsive LNCaP cells promotes their androgen-independent growth in vitro and in vivo[9]. Recently,the chemochine CCL5(RANTES)was found to be expressed by human prostate carcinoma cells and reported to stimulate their proliferation and invasion[10].Thus,also CCL5 appears to be directly involved in the behaviour of prostate carcinoma cells.The gene expression of both IL-6and CCL5is mainly regulated at a transcriptional level,by the cooperative activity of NF-k B transcription factor with members of at least five different families of transactivators including AP-1[1].Noticeably, the cooperation between NF-k B and AP-1appears to be essential for the constitutive deregulated production of IL-6observed in the androgen-independent,aggressive prostate cancer cells[11].Gene induction occurs depending on the ability of a variety of cell surface receptors to activate distinct and/or partially overlapping
intracellular signalling pathways eventually targeting the phos-phorylation site(s)of one or more MAP kinases(i.e.p38,ERK1/2, JUNK)committed in turn to activate IL-6and/or CCL5gene transactivators.Cytokines,growth factors receptors,adhesion molecules and many other membrane-generated signals all share the ability to efficiently promote IL-6or CCL5gene expression and consequently also their downstream effects.In addition,under long-term treatment conditions,IL-6can activate its own gene expression and,in prostate cancer,autocrine and paracrine loops involving IL-6and one of its multiple activators,the TGF-beta, have been implicated in the regulation of cell proliferation, survival,and neuroendocrine differentiation[12].
The expression levels of the prostate specific membrane antigen (PSMA)have been proposed as a useful indicator of the severity of the disease in prostate cancer[13–15].PSMA is a type-II integral membrane protein,predominantly localized to the epithelial cells of the prostate gland and endowed with folate-hydrolase and carboxypeptidase activity[13].Its low expression in normal prostate epithelial cells increases several fold in high-grade prostate cancers,in metastatic and in androgen-insensitive prostate carcinoma[14].These features have made it emerging as one of the most promising biomarkers in the diagnosis and treatment of prostate cancer[14,16].The clinical observations suggesting a possible correlation between high levels of IL-6production and PSMA expression in high-grade prostate cancer prompted us to investigate whether a functional relationship may exist between the presence of PSMA at the cell surface and the level of gene expression of https://www.sodocs.net/doc/e614838219.html,L5was also investigated because it shares with IL-6the mechanism of gene induction and the pro-proliferative activity.The hypothesis that PSMA belonged to the ever increasing category of molecules endowed with signalling properties was also suggested by previous observations showing that silencing,inhibiting or knocking-down PSMA modulates the adhesion/de-adhesion processes via activation of focal adhesion kinase(FAK)in LNCaP cells or via activation of p21-activated kinase1(PAK1)in normal endothelial cells(HUVEC)[17,18]. Unlike what previously done by others[18,19],we decided to stimulate rather than to inhibit or silence PSMA at the surface of LNCaP cells and to overcome the lack of defined PSMA ligand(s) by cross-linking the extra-cellular domain of the molecule with specific antibodies.
Here,by using the androgen-independent LNCaP cell line,we present for the first time evidence that the antibody-mediated aggregation of PSMA greatly induce the basal gene expression of both IL-6and CCL5in LNCaP cells.Gene induction occurs due to an activation cascade involving RAS,RAC1,p38and ERK1/2 MAPKs,leading to the phosphorylation of the p65subunit of NF-k B transcription factor.In addition,we demonstrate that both IL-6and CCL5promote the proliferation of LNCaP tumour cells reaching their maximal activity synergistically.Moreover,we present a novel finding related to the ability of CCL5to induced proliferation in LNCaP cells thank to the activation of STAT5and the consequent increase of expression of cyclin D1,a pathway controlling cell growth in normal and prostate cancer cells.This newly discovered functions of PSMA implicate this molecule as an important regulator of the prostate tumour cell growth. Materials and Methods
Chemicals,cell culture reagents and antibodies
PD098059and SB202190inhibitors were from Calbiochem-Novabiochem(San Diego,CA).Poly-D-lysine,trypsin and protease inhibitors were from SIGMA(SIGMA,Milan,Itay), whereas phenylmethylsulfonyl fluoride,antipain and soybean trypsin inhibitor were from Roche Diagnostics(Mannheim, Germany).The Hybond-P PVDF membranes and the ECL plus Western Blotting Detection System were from Amersham Bioscience Inc.(Freiburg,Germany).RPMI1640medium and glutamine were from Societa`Prodotti Antibiotici(Milan,Italy). Fetal bovine serum(FBS)was from Celbio(Pero,Milan, Italy).(Fab)2goat anti-mouse Ig and anti-VCAM-1were from Immunotech(Marseille,France);FITC-labeled(Fab)2goat anti-mouse Ig and anti p65phosphorylated mAb were from Becton-Dickinson(San Jose`,CA);the anti PSMA mAb J591,described by H.Liu et al.[20]was kindly supplied by Dr.N.H.Bander (Medical College of Cornell University,New York);the7E11c mAb,recognizing a cytosolic epitope of PSMA was purified from the supernatant of the hybridoma HB-10494(ATCC,American Type Culture Collection,Rockville Pike,MD);anti-phospho-p38, anti-p38polyclonal Abs and anti-phospho-ERK1/2mAb were from Cell Signaling Technology(Danvers,MA);anti-STAT5and anti-cyclin D1polyclonal antibodies were from Santa Cruz Biotechnology(Santa Cruz,CA);anti-RAS(clone RAS10)and anti-RAC(clone23A8)mAbs were from Upstate Biotechnology (Lake Placid,NY).The anti-PSMA mAb used for this work was produced in our laboratory and used in purified form following affinity chromatography.Briefly,BALB/c mice were immunized with a soluble recombinant form of PSMA and hybridoma cells obtained following standard fusion procedures.Hybrids were screened by flow cytometry using PSMA positive(LNCaP)and PSMA negative(DU145,PC-3)cell lines.
Cells and cell treatments
The PSMA-positive LNCaP and the PSMA-negative DU145 and PC-3human prostate carcinoma cell lines were from ATCC. Cells were maintained in vitro by serially passaging in RPMI-10% FBS at37u C in a humidified atmosphere of5%CO2.In the case of the LNCaP cell line,plating was performed on poly-D-lysine coated tissue culture plasticware(10m g/ml).When required,cells grown to confluence in24well plates were incubated with saturating doses of the appropriate mAb(5m g/ml,30min at room temperature),washed and placed at37u C for different times with(Fab)2-goat anti-mouse(Gam)(10m g/ml)to induce the clustering of PSMA molecules at the cell surface(this procedure will be henceforth designated‘‘PSMA cross-linking’’).In some instances,cells were pre-treated(6h at37u C)with ERK1/2(PD 98059)or p38(SB202190)inhibitors,used alone or in combination,at25or50m M in complete medium.The inhibitors were present throughout the duration of the assays.Cell-free supernatants and cells were separately recovered from each sample of untreated or treated cells24h after the addition of the various compounds.The volume of each cell-free supernatant was measured and the supernatant used to determine IL-6and CCL5production.Cells corresponding to each supernatant were lysed and the amount of proteins assessed by Coomassie protein assay reagent(Pierce,Celbio,Milano,Italy).
Biochemical methods
Cell extracts were prepared from LNCaP cells cultured overnight with2.5%FBS and then starved from serum for2h. Protein amount was measured by Coomassie protein assay reagent.RAS activity was assessed by employing the RAS activation assay kit(Upstate Biotechnology).Briefly,untreated and treated cells were lysed for30min on ice in a buffer containing25mM Hepes pH7.5,NaCl150mM,10%Glicerol, Na orthovanadate1mM,Na pyrophosphate50nM,Na fluoride 25mM,MgCl210mM,1%NP40,phosphatase and protease inhibitors.The cell lysates(250m g)were incubated with10m l of
PSMA Induces IL-6,CCL5
glutathione S-transferase(GST)-Raf-RAS binding domain(BD) (Upstate Biotechnology)coupled to agarose for1h at4u C.The beads were washed with the lysis buffer,resuspended in26SDS sample buffer,and separated by SDS-Page.Western Blot was performed with the anti-RAS mAb and immunoreactive bands were visualized by using a horseradish peroxidase-conjugated secondary antibody and the ECL system(GE Healthcare Bio-Sciences AB,Uppsala,Sweden).
RAC1activity was assessed by employing the RAC activation assay kit(Upstate Biotechnology).Briefly,cell lysates prepared as above(250m g)were incubated with10m l of GST-PAK-1-RAC BD coupled to agarose for1h at4u C.The beads were washed with the lysis buffer,resuspended in26SDS sample buffer and separated by SDS-Page.Western blotting was performed with the anti-RAC antibody(Upstate Biotechnology).
To examine p38and ERK1/2phosphorylation,untreated and treated cells were extracted with the buffer indicated above.Equal amounts of total proteins were boiled in sample buffer and separated by SDS-PAGE.Following immunoblotting with anti-phospho-p38or anti-p38polyclonal Ab,anti-phospho-ERK1/2 mAb or anti-ERK1/2polyclonal Ab,immunoreactive bands were visualized by using horseradish peroxidase-conjugated secondary antibody and ECL.Film densitometry was carried out with a laser densitometer and analyzed using a built-in software(LKB, Bromma,Sweden).
Intracellular phospho-specific flow cytometry
LNCaP cells were treated with ant-PSMA mAb for30min at room temperature,washed and cross-linked with Gam at37u C for 0,15or30min.After this time cells were fixed with Fix buffer (BD)for10min at room temperature to arrest signaling activity. Cells were then washed and permeabilized with50%ice-cold methanol for30min on ice.Fixed and permeabilized cells were washed twice,resuspended in phosphate-buffered saline plus2% FBS and stained for30min at room temperature with monoclonal PE-conjugated antibody against phospho-NF-k B(S529)(Becton-Dickinson).Conjugated isotype-matched,non-reactive monoclo-nal antibody was used as control.Fluorescence signals were detected using a FACSCalibur flow cytometer(Becton-Dickinson) and data elaborated by FlowJo software(Tree Star,Ashland,OR). IL-6and CCL5quantitation
IL-6and CCL5concentration in cell-free supernatants was measured with an ELISA kit used according to the manufacturer’s instructions(Endogen,Woburn,MA).Optical density(OD)values were plotted onto a standard curve and expressed as pg/ml. Having measured the volume of cell-free supernatants of each sample,IL-6or CCL5values could be re-expressed as total IL-6or CCL5/sample.Total IL-6or CCL5of each sample was then normalized versus the amount of total protein of cell lysate of the producing cells and calculated as pg IL-6or CCL5/mg of protein lysate.
Cell proliferation assays
Cell proliferation studies were carried out with LNCaP cells previously cultured with only1%FBS for48h and then treated with the indicated doses of recombinant IL-6or CCL5(Endogen). In the case of treatment with anti-PSMA mAb the binding and the cross-linking of anti-PSMA or7E11c were performed as previously described in the‘‘Cells and cell treatments’’paragraph of the present section.In the case of anti-CCR5treatment,the antibody at the indicated concentrations was preincubated for2h at37u C before adding the CCL5.During the last12h cells were pulsed with1m Ci[methyl-3H]thymidine(3H-TdR)(6.7Ci/mmol)from
NEN Dupont(Boston,MA).Cells were harvested onto glass-fiber filters and radioactivity measured in a beta-spectrometer.Results were expressed as the percentage of incorporation of the untreated control.
Flow cytometry
Cells were incubated with saturating amounts of the indicated antibody for1h on ice and then washed three times with cold RPMI medium.The cells were then incubated with2m l of a goat anti-mouse FITC-labeled antiserum(Becton-Dickinson)for1h on ice.After washings cells were analyzed by a Facs Canto instrument (Becton-Dickinson).
Statistical analysis
Statistical analysis of results was performed applying the parametric unpaired t test.Significance was accepted when p,0.05.
Results
Characteristics of the anti-PSMA mAb used for the
present study
Biochemical and functional studies shown in this paper were performed with an anti PSMA mAb(henceforth designated‘‘anti-PSMA mAb’’)produced by our group.Specificity for PSMA is shown in Fig.1.Staining,competition and immuneprecipitation experiments were performed with LNCaP cells and the anti-PSMA mAb was compared with the J591mAb,a reference mAb recognizing an extra-cellular epitope of the PSMA molecule[20].
Flow cytometry of cells stained with anti-PSMA or J591mAbs demonstrated that both mAbs recognized a surface molecule expressed at comparable levels on LNCaP cells(Fig.1A).No staining was instead observed with the PSMA negative DU145or PC3cell lines(not shown).Competition experiments,performed with biotinilated J591mAb revealed by FITC-streptavidin, demonstrated that the number of positive cells and the mean fluorescence observed following J591staining was progressively decreased when cells were pre-treated with increasing amounts of non-conjugated anti-PSMA mAb(Fig.1,B),thus demonstrating that our anti-PSMA mAb specifically recognizes an epitope of the extra-cellular domain of the molecule identical or spatially close to that identified by the J591mAb.Immunoprecipitations of crude cell lysates revealed the ability of our anti-PSMA mAb to immunoprecipitate a protein of the same apparent molecular weight(110kD)of that immunoprecipitated using J591mAb or 7E11c mAb,this last recognizing an intracellular epitope of PSMA (Fig.1C).Collectively,these results demonstrated that our anti-PSMA mAb specifically recognizes an extra cellular epitope of PSMA.Results obtained at a single cell level were extended by immunohistochemistry to human tissues,demonstrating that the reactivity of anti-PSMA mAb overlapped that of J591(not shown) PSMA recruitment with specific antibodies activates the MAPKs p38and ERK1/2and the small GTPases RAC1and RAS in LNCaP cells
To test whether PSMA was able to transduce intracellular signals leading to p38and/or ERK1/2phosphorylation,LNCaP cells were subjected to PSMA cross-linking by anti-PSMA-specific antibodies as adherent,confluent monolayers.PSMA molecules, recognized by the anti-PSMA mAb were induced to cluster at the cell membrane due to the binding of a goat anti-mouse IgG antiserum(Gam)added to cell cultures at37u C for the indicated times.Controls included treatment with anti VCAM-1,with
PSMA Induces IL-6,CCL5óé Foxit Reader ±à?-
°?è¨?ùóD (C) Foxit Software Company,2005-2006??ó?óú?à1à?￡
7E11c or with Gam alone and the use of untreated cells.Cells were then washed and used as a source of cell lysates.We first assayed the phosphorylation of p38and ERK1/2in crude lysates obtained from the same cell cultures immunoblotted with
antibodies specifically detecting the active,phosphorylated form of the two kinases.As shown in Fig.2A and B,ERK1/2and p38were found activated at low levels in untreated LNCaP cells.Upon PSMA cross-linking,both ERK1/2and p38became
considerably
Figure 1.Reactivity of anti-PSMA mAb.Panel A:representative flow cytometry of LNCaP cells stained with anti-PSMA or with J591mAb.Controls are represented by cells goat anti-mouse-FITC IgG only.Panel B:flow cytometry of LNCaP cells incubated with biotinylated J591mAb (B-J591)with or without pretreatment (45min on ice)with anti-PSMA at the indicated concentrations or with non-biotinylated J591mAb as control.Antibody binding was revealed by FITC-Avidin (F-Avidin).The percentage of positive cells and the Mean Fluorescence Intensity observed under the different conditions are representative of two independent experiments performed with independent cultures.Panel C:Western blot analysis of LNCaP cell lysates immuneprecipitated with the indicated mAbs,and then probed with the anti-PSMA mAb followed by a goat anti-mouse antiserum labelled with horseradish peroxidase.The apparent molecular weight of PSMA band is indicated by the arrow.doi:10.1371/journal.pone.0004608.g001
PSMA Induces IL-6,CCL5
phosphorylated at 10min.Sustained levels of phosphorylation lasted up to 20min,and then returned to basal levels.It is known that the small GTPases RAS and RAC1lie upstream on the MAPKs activation pathways [21].
Thus,to examine if PSMA cross-linking could activate also RAC1in LNCaP cells,we performed an activation assay with a GST-Pak fusion protein which is able to bind and precipitate the active form of RAC1.As shown in Fig.2C,PSMA cross-linking determined endogenous RAC1activation in the same lysates.Similar results were obtained in the case of RAS,whose constitutive activation level was greatly augmented upon PSMA cross-linking (Fig.2D).Cross-linking performed with an isotype matched mAb recognizing VCAM-1failed to induce activation in all assays and at all points of the time-course (Fig.2,all panels).Gel densitometry of independent experiments performed with different cell lysates confirmed the significance of the increased activation of ERK1/2,p38,RAC1and RAS observed following PSMA,but not VCAM-1cross-linking (Table 1).Because the time-course of RAS and RAC1activation paralleled that of p38and ERK1/2,these results overall indicate that PSMA is able to induce a sustained activation of the RAS-and RAC-MAPK pathways.
PSMA-elicited phosphorylation of p38and ERK1/2induces the activation of NF-k B transcription factor
It is well established that p38and ERK1/2have a role in the process of activation of NF-k B transcription factor,and that NF-k B plays a major function in the transcriptional control of various cytokines and chemokines,including IL-6and CCL5[1].To investigate whether NF-k B activation was induced in LNCaP cells by the cross-linking of PSMA,cells were treated as described in Fig.2and stained with a mAb recognizing the phosphorylated Serine 529(pS529)located in the transactivation domain of the p65subunit of the human NF-k B.As the pS529containing domain is unmasked only after detachment of NF-k B inhibitors
Figure 2.Kinetics of ERK1/21/2,p38,RAC 1and RAS phosphorylation induced by PSMA cross-linking.LNCaP cells were left untreated or subjected to PSMA cross-linking for the indicated times at 37u C.An anti-VCAM-1mAb was used as the isotype matched control.Panels A and B:p38and ERK1/2activation was assessed in crude lysates.Equal amounts (20m g)of total proteins were boiled in sample buffer and separated by SDS-PAGE.Following immunoblotting with an anti-phospho-p38or anti-phospho-ERK1/2mAb,immunoreactive bands were visualized by using horseradish peroxidase-conjugated secondary antibody and the ECL system.Panels C and D:RAC1and RAS activation was evaluated by a specific assay as described in the Material and Methods section.Bound active GTP-RAC and GTP-RAS molecules were analyzed by Western blotting using an anti-RAC1or anti-Ras mAb and visualized as above.Total amount of ERK1/2,p38,RAC1and RAS in crude lysates are shown as loading control at the bottom of each gel.Results are representative of one of three independent experiments.doi:10.1371/journal.pone.0004608.g002
PSMA Induces IL-6,CCL5
óé Foxit Reader ±à?-°?è¨?ùóD (C) Foxit Software Company,2005-2006??ó?óú?à1à?￡
induced by stimulation,the exposure of pS529is considered a marker of NF-k B nuclear translocation and activation in normal and neoplastic cells of different origin.
Additionally,phosphory-lated S529further augments NF-k B transactivation potential [22].Controls were represented by cells left untreated,treated with an isotype-matched IgG1antibody or subjected to VCAM-1cross-linking.The results of flow cytometry analysis (Fig.3)demon-strated that the mean fluorescence intensity (MFI)of LNCaP cells (9.260.9SEM)was greatly augmented following PSMA cross-linking,peaking at 15min from stimulation and declining 30min later (230%630SEM and 160%620SEM of control cells,respectively),thus implying that NF-k B activation had occurred.No increase of MFI was instead detected at the various points of the time course in the presence of the control antibody or VCAM-1cross-linking.
PSMA cross-linking enhances the production of IL-6and CCL5in LNCaP cells in a p38and ERK1/2dependent manner
Results shown in Fig.2and 3demonstrated that cross-linking of PSMA recruited a signalling pathway eventually leading to NF-k B activation and nuclear translocation.The contribution of PSMA to the regulation of the production of IL-6and CCL5in LNCaP cells was then investigated in cross-linking experiments carried out as in Fig.2.Further,we analyzed the role of PSMA-activated p38and ERK1/2in a loss-of-function condition achieved by using two specifically p38and blocking reversibly its enzymatic activity or PD 98059(PD),an inhibitor specifically preventing the activation of the ERK1/2-activating kinase MEK-1[23].LNCaP cells were stimulated as confluent monolayer and harvested 24h later.To determine the basal and inducible release of IL-6and CCL5in culture,LNCaP cells and the corresponding cell-free supernatants were separately recovered from each treated or untreated cell culture.Cell lysates and cell-free supernatants were used to determine respectively the total protein amount and IL-6and CCL5accumulation in the supernatant of each sample.The results of IL-6and CCL5measurements were then expressed as pg of IL-6or CCL5/mg of cell protein lysates.As illustrated in Fig.4A,LNCaP cells produced low amounts of IL-6and of CCL5basally (11.362and1463.6pg/mg total cell protein/24h,respectively).The cross-linking of PSMA increased this production by three-fold in the case of IL-6and by two-fold in the case of CCL5(311%648,p ,0.001and 214%638p ,0.001,respective-ly),thus demonstrating that PSMA recruitment at the surface of LNCaP cells powerfully up-regulates the production of both IL-6and CCL5at the same time.The cross-linking of the isotype-
Figure 3.NF-k B activation induced by PSMA crosslinking.LNCaP cells were subjected to stimulation by cross-linking of PSMA or of VCAM-1for 45min at room temperature and then cultured for the indicated times at 37u C.The cells were then harvested,washed,permeabilized and stained with a PE-labelled anti-phosphorylated NF-k B p65antibody.Cells were then analyzed by flow cytometry.The MFI was recorded.An isotype-matched IgG1antibody was used to mock-treat LNCaP cells;the Figure illustrates a representative experiment out of three performed independently.MFI values of controls were:untreated cells 9.2260.9,cells treated with an anti-VCAM-1mAb 5.74+0.09,cells treated with a mouse IgG1mAb 5.0260.3.No substantial differences in the control samples were observed at the various time points of the experiments.MFI of Anti-PSMA cross-linked cells were:22.0362.2and 15.661.3at 15min and 30min,respectively.p values of samples treated with anti-PSMA mAb were ,0.01at 15min and 30min,respectively,with respect to the untreated cells.The first panel from top (mouse IgG1)is the plot corresponding to time 0.doi:10.1371/journal.pone.0004608.g003
PSMA Induces IL-6,CCL5
óé Foxit Reader ±à?-°?è¨?ùóD (C) Foxit Software Company,2005-2006
matched anti VCAM-1mAb lacked instead any significant activity (155%652and 133%611,respectively).
As also illustrated in Fig.4A,the p38inhibitor SB used at 50m M concentration almost abrogated both the basal and the PSMA-induced IL-6production (70%62.5p ,0.01and 87.4%620p ,0.05,respectively),whereas at 25m M the decrease observed did not reach significance.The ERK1/2inhibitor PD showed a poorer activity than SB,as it diminished the IL-6produced basally,(30%65p ,0.05and 50%67p ,0.05,at
25m M and 50m M,respectively)but failed to show any effect on that induced by PSMA cross-linking.Accordingly,IL-6reduction observed in the presence of both inhibitors (both used at 25m M concentration)overlapped that dependent on the sole SB.Similar results were obtained regarding the production of CCL5.As shown in Fig.4B,the basal production of CCL5diminished in a dose-dependent manner in the presence of SB,reaching significance at 50m M (45%612,p ,0.01).The PSMA-induced production of CCL5was significantly reduced already at 25m
M
Figure 4.Inducing activity of PSMA cross-linking on IL-6and CCL5expression and role of PSMA-activated p38and ERK1/2.IL-6(panel A)and CCL5(panel B):production by LNCaP cells left untreated (open bars),stimulated by cross-linking via anti PSMA mAb (filled bars)or anti-VCAM-1mAb (dashed bars)in the presence or in the absence of single or combined inhibitors (SB and PD)at the indicated doses (SB50=50m g/ml,SB25=25m g/ml;PD50=50m g/ml,PD25=25m g/ml;SB/PD25=combinations of SB +PD at 25m g/ml each).Control and treated cell cultures were washed,cultured for further 24h before collecting cells and cell-free supernatants.IL-6and CCL5accumulation in each supernatant was measured by ELISA and expressed as pg/mg of cell lysates/24h.Inhibitors were added to the cells 6h prior to stimulation and maintained throughout the duration of the experiment.Values represent the mean 6SEM of results of four experiments performed in duplicates.Stars on top of the bars indicate statistical significance as follows:p ,0.001,***;p ,0.01,**;p ,0.05,*.doi:10.1371/journal.pone.0004608.g004
PSMA Induces IL-6,CCL5
SB and flattened out to basal levels at50m M(38%64,p,0.01 and72%616,p,0.01,respectively).Also in the case of CCL5the ERK1/2inhibitor PD performed less efficiently.The basal production of CCL5was not significantly affected by the presence of PD and the PSMA-induced CCL5production was only halved although the effect was significant(p,0.001),with no dose-dependency(49%64,p,0.001PD used at50m M;42612, p,0.01,PD used at25m M).No greater activity was obtained by using a mixture of the two inhibitors used at25m M concentration. The possibility that the mix of inhibitors used at higher concentrations may be effective could not be explored because the compounds precipitate when the concentration is brought close to100m M.The effect of SB and PD on IL-6and CCL5 produced by cells stimulated by the cross-linking of the control mAb were similar to those of untreated cells(Fig.4A and B). Collectively,the results shown in Fig.4demonstrate that the activation of p38and ERK1/2elicited by PSMA cross-linking regulates the expression of IL-6and CCL5genes in LNCaP cells like in other pathological and normal epithelial cells[24,25]. Moreover,as also observed in other instances our results underline the fact that p38exerts a wider and greater activity than ERK1/2 on IL-6and CCL5gene expression[24,25].
Proliferation of LNCaP cells is inducible by IL-6,CCL5or by cross-linking of PSMA
Proliferation assays were then carried out to evaluate whether the increased availability of IL-6and CCL5in a neoplastic microenvironment would affect the growth of the LNCaP tumour cell line.
As illustrated in Fig.5A,the rate of unlimited proliferation of LNCaP cells was increased by the addition of recombinant IL-6or CCL5in a dose-dependent manner.IL-6exerted the maximal activity at100ng/ml(175616,p,0.01)whereas CCL5promoted a similar enhancement already at5ng/ml(146%62of increase, p,0.05),although only a subset(8–10%)of strongly positive CCR5cells could be detected in our LNCaP cells population when FBS was lowered in culture to1%(S.Grasso,personal observation).IL-6and CCL5appeared to act synergistically in
proliferation,as demonstrated by the finding that the addition of as little as1ng/ml of IL-6to the most effective dose of CCL5raised the LNCaP cells proliferation up to values which were significantly higher than those of controls (168%62.9,p,0.001),and,perhaps more importantly,greater than those observed in the presence of5ng/ml of CCL5(p,001). As shown in Fig.5B the induction of proliferation induced by CCL5was almost abrogated by blocking CCR5with specific antibodies.No interference was instead observed in the presence of7E11c antibody,used as isotype-matched control,thus putting into evidence a direct correlation between the burst of LNCaP cells proliferation and the CCL5interaction with its receptor CCR5.It has been reported that the triggering of CCR5by CCL5 is rapidly followed by JAK-STAT phosphorylation,more specifically by STAT1and/or STAT3and/or STAT5,depending on the cellular system[26–28].It has been also reported that STAT5phosphorylation regulates the expression of cyclin D1in tumor cells including those of prostate[29,30].To investigate the downstream effects of CCR5engagement in LNCaP cells,the cells were cultured in the presence of5ng/ml of CCL5for the indicated times and then analysed as for the expression of STAT5 and Cyclin D1by immunoblotting with specific antibodies.As illustrated in Fig.5C phosphorylated STAT5appeared as a doublet at30min.The doublet lowered its intensity at1h to return to the almost undetectable level observed in untreated cell at3h.Consistently,the amount of cyclin D1was found more than
doubled at1h and at3h,thus indicating that CCL5can regulate LNCaP cells proliferation in a STAT5-Cyclin D1dependent way.
Taken toghether,these results demonstrate that the antibody-mediated clustering of PSMA at the LNCaP cells surface activates signalling pathways able to directly or indirectly control the LNCaP cells growth.To confirm that the complex series of phenomena reported in the previous paragraphs led to LNCaP cells proliferation we decided to test if the cross-linking of PSMA would indeed induce the LNCaP cells proliferation.As shown in Fig.5D LNCaP cells proliferation was found to be increased following PSMA cross-linking(175%611,p,0.05),whereas the bindings of the sole anti-PSMA,of7E11c or the cross-linking of 7E11c were all devoid of significant activity.
Discussion
In this paper we show for the first time that the tumor biomarker PSMA,a multi-functional cell surface ectopeptidase
[13],is endowed with an efficient signaling activity in prostate
cancer cells.PSMA recruitment at the surface of LNCaP cells elicits a signaling wave involving small GTPases RAS,RAC1,and MAPK p38and ERK1/2activation.Consistent with the p38and ERK1/2phosphorylation,the NF-k B transcription factor was also activated[1]following PSMA stimulation and the minimal basal level of both IL-6and CCL5production was increased two-three fold.The discrete role of PSMA-mediated activation of p38and ERK1/2was examined under loss-of-function conditions secured by the blocking activity of kinase-specific chemical inhibitors,used at concentrations not affecting the adhesion of LNCaP cells nor the expression of PSMA at their surface(not shown).In this setting,p38kinase exerted a wider and greater activity than ERK1/2on both the basal and the PSMA-induced expression of IL-6and CCL5.A possible explanation of the greater function of p38may reside in the different roles played by p38and ERK1/2 in the activation of NF-k B and NF-IL6,major transactivators of both IL-6and CCL5promoters[1].We have previously reported that in normal human thymic epithelial cells p38MAPK induced the post-transcriptional activation of both NF-k B and NF-IL6, whereas ERK1/2was involved in the activation of NF-k B only.
Consistent with the dual activity of p38,its inhibition with SB202190almost production,whereas ERK1/2inhibition with PD98059resulted in a lesser effect [24,25].The distinct functions of the two MAPKs may well be reproduced also in prostate cancer cells,as evidenced by our results with the inhibitors PD98059and SB202190,thus suggesting that also in LNCaP cells p38MAPK exerts a wider role than ERK1/2on IL-6and CCL5transcription factors.
The membrane proximal events leading to RAS and RAC1 activation following PSMA recruitment have not been studied yet and full elucidation of these aspects will require further investigations.
Irrespective of the composition of the signaling platform used by PSMA in transducing external stimuli,our results provide evidence that PMSA can regulate the gene expression of IL-6 and CCL5in the same LNCaP cells.These results were obtained following direct stimulation of the molecule under‘‘in vitro’’conditions closely resembling a physiological setting and without substantial cell manipulations.In addition,because biochemical studies and functional assays were performed with cell lysates and cell supernatants obtained from the same cell cultures,these results directly demonstrated that PSMA-elicited signals flowed efficiently from the membrane to the nucleus.The amount of IL-6and CCL5released in culture by the LNCaP cell line may appear lower than that produced by other cell lines.However,it has to be
PSMA Induces IL-6,CCL5óé Foxit Reader ±à?-
°?è¨?ùóD (C) Foxit Software Company,2005-2006??ó?óú?à1à?￡
considered that a low concentration measured in cell supernatants can be dramatically high for a portion of the cell membrane and that the LNCaP cell line used by us represents only one of the multiple clones composing the prostate cancer cell population,thus emphasizing the importance that the synergic activity of IL-6and CCL5on LNCaP cells proliferation described by us may have in an ‘‘vivo’’microenvironment.IL-6has yielded conflicting results in prostate tumor cell lines including LNCaP.Cell proliferation,anti-apoptotic activity as well as growth arrest and
neuroendocrine differentiation were all observed following treat-ment with IL-6‘‘in vitro’’[12].By contrast,the notion that the level of IL-6and its receptor increase during prostate carcinogen-esis ‘‘in vivo’’is still unchallenged,so that IL-6is nowadays considered a good candidate for targeted therapy in prostate cancer [12].Differences in the cellular and experimental systems employed together with the complexity of the signaling pathway elicited by cytokines may possibly explain the divergence between the clinical and the ‘‘in vitro’’experimental
evidence.
Figure 5.Effects of pro-proliferative stimuli on LNCaP cells.A)LNCaP proliferation induced by treatment with CCL5and/or IL-6.3H-TdR uptake of cells after 24h of culture with CCL5(grey bars),72h of culture with IL-6(hatched bars)or 24h of culture with combinations of CCL5and IL-6(crossed bars)at the indicated concentrations.B)3H-TdR uptake of cells after 24h culture with 5ng/ml of CCL5in the presence or in the absence of anti-CCR5or 7E11c antibodies at the indicated concentrations;C)Western blot analysis of cell extracts (20m g/lane)of LNCaP cells treated with CCL5(5ng/ml)and probed with anti-STAT 5or with anti-cyclin D1antibodies.Equal protein loading in all the lanes is confirmed by not specific bands shown in the middle gel.Arrows indicate phosphorylated STAT5proteins.Western blotting results are representative of two independent experiments.D)3H-TdR uptake of cells 72h after binding or cross-linking of 7E11c or anti-PSMA antibodies.Values represent the mean 6SEM of three independent experiments performed in triplicates.Stars on the bars indicate statistical significance as in Fig.4.doi:10.1371/journal.pone.0004608.g005
PSMA Induces IL-6,CCL5
óé Foxit Reader ±à?-°?è¨?ùóD (C) Foxit Software Company,2005-2006??ó?óú?à1à?￡
Little is known about CCL5-CCR5interaction in LNCaP cells, apart from the pro-proliferative activity previously reported by G. Vaday et al.[10]and confirmed also in the present paper.The novel finding presented herein is the ability of CCL5to activate STAT5thereby regulating the expression of cyclin D1and ultimately controlling the LNCaP cell cycle progression. Overall,it appears that PSMA signaling may be involved at multiple stages of the LNCaP cell deregulation and possibly also of in vivo tumorigenesis,because1)increased ERK1/2activation correlates with increasing Gleason score and advanced tumor stage[31],and2)increased activation of STAT5has been observed‘‘in vivo’’in65%of human prostate cancers,being associated with high histological grade and being a predictor of early disease recurrence[29,30,32].Moreover,STAT5inhibition induces cell death in human prostate cancer cells[33].
It could be hypothesized that following stimulation via PSMA the prostate tumor cells augment IL-6and CCL5production which are in turn used as growth factors in both autocrine and paracrine manner,triggering a cell proliferation/survival loop conferring resistance to apoptosis and an overall definite advantage to tumor cell populations.Within this context,the ability of PSMA to activate RAC1may be highly relevant, inasmuch as RAC1activation decreases the expression of E-cadherins thereby loosening intercellular adhesions and facilitating the cytoskeletal rearrangements required for mitosis[34].If so, activated RAC1could therefore favor the response of LNCaP cells to the mitogenic activity of IL-6and CCL5meanwhile participating to the induction of their expression.
These events may have a direct bearing also on tumor cell proliferation and survival‘‘in vivo’’.The prostate gland is composed mainly of stromal,epithelial,and neuroendocrine cells. The dynamic balance of cell proliferation,differentiation,and apoptosis maintains the cellular and tissue homeostasis.This balance is generated by a continuous cross-talk among these cell populations resulting from cell-cell contacts and the release of growth factors,chemokines and neuropeptides acting in a paracrine and autocrine manner[13].Deregulation in this communication circuit may contribute to the derangement of one or more tissue component(s)of the prostate gland,possibly resulting in benign prostate hyperplasia or prostate carcinoma. The pathway components here described are already the targets of therapeutic interventions with pharmacologic or biotechnologic inhibitors in other types of epithelial tumors[21].Therefore,also in prostate tumors the biochemical intermediates involved in NF-k B activation and augmented IL-6/CCL5gene expression might represent adequate therapeutic targets aiming at interfering with one or more of the MAPKs activation pathway(s),thereby counteracting their effects on tumor cell proliferation and migration
Author Contributions
Conceived and designed the experiments:MC FM DR.Performed the experiments:SG AP GF MTS OP.Analyzed the data:SG AP DR. Contributed reagents/materials/analysis tools:MC AP GF SC HYN MH FM.Wrote the paper:MC DR.Chief of the Section of Immunology: MTS.Chief of the Section:GT.
References
1.Naka T,Nishimoto N,Kishimoto T(2002)The paradigm of IL-6:from basic
science to medicine.Arthritis Res4Suppl3:S233–S242.
2.Trikha M,Corringham R,Klein B,Rossi JF(2003)Targeted anti-interleukin-6
monoclonal antibody therapy for cancer:a review of the rationale and clinical evidence.Clin Cancer Res9:4653–4665.
3.Siegall CB,Schwab G,Nordan RP,FitzGerald DJ,Pastan I(1990)Expression of
the interleukin6receptor and interleukin6in prostate carcinoma cells.Cancer Res50:7786–7788.
4.Hobisch A,Rogatsch H,Hittmair A,Fuchs D,Bartsch G,et al.(2000)
Immunohistochemical localization of interleukin-6and its receptor in benign, premalignant and malignant prostate tissue.J Pathol191:239–244.
5.Adler HL,McCurdy MA,Kattan MW,Timme TL,Scardino PT,et al.(1999)
Elevated levels of circulating interleukin-6and transforming growth factor-beta1 in patients with metastatic prostatic carcinoma.J Urol161:182–187.
6.Drachenberg DE,Elgamal AA,Rowbotham R,Peterson M,Murphy GP(1999)
Circulating levels of interleukin-6in patients with hormone refractory prostate cancer.Prostate41:127–133.
7.Chen T,Wang LH,Farrar WL(2000)Interleukin6activates androgen receptor-
mediated gene expression through a signal transducer and activator of transcription3-dependent pathway in prostate cancer cells.Cancer Res60: 2132–2135.
8.Ueda T,Bruchovsky N,Sadar MD(2002)Activation of the androgen receptor
N-terminal domain by interleukin-6via MAPK and STAT3signal transduction pathways.J Biol Chem277:7076–7085.
9.Lee SO,Lou W,Hou M,de MF M,Gerber L,Gao AC(2003)Interleukin-6
promotes androgen-independent growth in human prostate cancer cells.Clin Cancer Res9:370–376.
10.Vaday GG,Peehl DM,Kadam PA,Lawrence DM(2006)Expression of CCL5
(RANTES)and CCR5in prostate cancer.Prostate66:124–134.
11.Zerbini LF,Wang Y,Cho J,Towia A,Libermann TA(2003)Constitutive
Activation of Nuclear Factor k B p50/p65and Fra-1and JunD Is Essential for Deregulated Interleukin6Expression in Prostate Cancer.Cancer Research63: 2206–2215.
12.Culig Z,Steiner H,Bartsch G,Hobisch A(2005)Interleukin-6Regulation of
Prostate Cancer Cell Growth.Journal of Cellular Biochemistry95:497–505.
13.Rajasekaran AK,Anilkumar G,Christiansen JJ(2005)Is prostate-specific
membrane antigen a multifunctional protein?Am J Physiol Cell Physiol288: C975–C981.
14.Silver DA,Pellicer I,Fair WR,Heston WD,Cordon-Cardo C(1997)Prostate-
specific membrane antigen expression in normal and malignant human tissues.
Clin Cancer Res3:81–85.
15.Chang SS(2004)Overview of prostate-specific membrane antigen.Rev Urol6
Suppl10:S13–S18.
16.Bander NH,Milowsky MI,Nanus DM,Kostakoglu L,Vallabhajosula S,et al.
(2005)Phase I trial of177lutetium-labeled J591,a monoclonal antibody to
prostate-specific membrane antigen,in patients with androgen-independent
prostate cancer.J Clin Oncol23:4591–4601.
17.Barwe SP,Maul RS,Christiansen JJ,Anilkumar G,Cooper CR,et al.(2007)
Preferential association of prostate cancer cells expressing prostate specific
membrane antigen to bone marrow matrix.Int J Oncol30:899–904.
18.Ghosh A,Wang X,Klein E,Heston WD(2005)Novel role of prostate-specific
membrane antigen in suppressing prostate cancer invasiveness.Cancer Res65:
727–731.
19.Lee JC,Laydon JT,McDonnell PC,Gallagher TF,Kumar S,et al.(1994)A
protein kinase involved in the regulation of inflammatory cytokine biosynthesis.
Nature372:739–746.
20.Liu H,Moy P,Kim S,Xia Y,Rajasekaran A,et al.(1997)Monoclonal
antibodies to the extracellular domain of prostate-specific membrane antigen
also react with tumor vascular endothelium.Cancer Res57:3629–3634.
21.Roberts PJ,Der CJ(2007)Targeting the Raf-MEK-ERK mitogen-activated
protein kinase cascade for the treatment of cancer.Oncogene26:3291–3310.
22.Viatour P,Merville MP,Bours V,Chariot A(2005)Phosphorylation of NF-
kappaB and IkappaB proteins:implications in cancer and inflammation.Trends
Biochem Sci30:43–52.
23.Davies SP,Reddy H,Caivano M,Cohen P(2000)Specificity and mechanism of
action of some commonly used protein kinase inhibitors.Biochem J351:
95–105.
24.Mainiero F,Colombara M,Antonini V,Strippoli R,Merola M,et al.(2003)p38
MAPK is a critical regulator of the constitutive and the beta4integrin-regulated
expression of IL-6in human normal thymic epithelial cells.Eur J Immunol33:
3038–3048.
25.Colombara M,Antonini V,Riviera AP,Mainiero F,Strippoli R,et al.(2005)
Constitutive activation of p38and ERK1/2MAPKs in epithelial cells of
myasthenic thymus leads to IL-6and RANTES overexpression:effects on
survival and migration of peripheral T and B cells.J Immunol175:7021–7028.
26.Conway RE,Petrovic N,Li Z,Heston W,Wu D,Shapiro LH(2006)Prostate-
specific membrane antigen regulates angiogenesis by modulating integrin signal
transduction.Mol Cell Biol26:5310–5324.
27.Rodriguez-Frade JM,Vila-Coro AJ,Martin A,Nieto M,Sanchez-Madrid F,et
al.(1999)Similarities and differences in RA.J Cell Biol144:755–765.
28.Wong M,Fish EN(1998)RANTES and MIP-1alpha activate stats in T cells.
J Biol Chem273:309–314.
29.Brockman JL,Schuler LA(2005)Prolactin signals via Stat5and Oct-1to the
proximal cyclin D1promoter.Mol Cell Endocrinol239:45–53.
30.Dagvadorj A,Kirken RA,Leiby B,Karras J,Nevalainen MT(2008)
Transcription factor signal transducer and activator of transcription5promotes
growth of human prostate cancer cells in vivo.Clin Cancer Res14:1317–1324.
PSMA Induces IL-6,CCL5óé Foxit Reader ±à?-
°?è¨?ùóD (C) Foxit Software Company,2005-2006??ó?óú?à1à?￡
31.Carey AM,Pramanik R,Nicholson LJ,Dew TK,Martin FL,et al.(2007)Ras-
MEK-ERK signaling cascade regulates androgen receptor element-inducible gene transcription and DNA synthesis in prostate cancer cells.Int J Cancer121: 520–527.
32.Li H,Ahonen TJ,Alanen K,Xie J,LeBaron MJ,et al.(2004)Activation of
signal transducer and activator of transcription5in human prostate cancer is associated with high histological grade.Cancer Res64:4774–4782.33.Ahonen TJ,Xie J,LeBaron MJ,Zhu J,Nurmi M,et al.(2003)Inhibition of
transcription factor Stat5induces cell death of human prostate cancer cells.J Biol Chem278:27287–27292.
34.Woods A,Wang G,Dupuis H,Shao Z,Beier F(2007)Rac1signaling stimulates
N-cadherin expression,mesenchymal condensation,and chondrogenesis.J Biol Chem282:23500–23508.
PSMA Induces IL-6,CCL5