calreticulin expression is associated with androgen ...€¦ · inc....

[CANCER RESEARCH 59, 1896–1902, April 15, 1999]

Calreticulin Expression Is Associated with Androgen Regulation of the Sensitivity toCalcium Ionophore-induced Apoptosis in LNCaP Prostate Cancer Cells1

Ning Zhu and Zhou Wang2

Departments of Urology [N. Z., Z. W.] and Molecular Pharmacology and Biological Chemistry [Z. W.] and The Robert H. Lurie Cancer Center [Z. W.], Northwestern UniversityMedical School, Chicago, Illinois 60611

ABSTRACT

Calreticulin has been identified previously as one of the androgen-response genes in the prostate. The role of calreticulin in androgen actionwas studied using androgen-sensitive LNCaP and androgen-insensitivePC-3 human prostate cancer cell lines. Calreticulin appears to be aprimary androgen-response gene in cultured LNCaP cells because andro-gen induction of calreticulin mRNA resists protein synthesis inhibition.Calreticulin is a high capacity intracellular Ca21 binding protein, sug-gesting that calreticulin expression is likely to be associated with theintracellular Ca 21 buffering capacity that could regulate the sensitivity tocytotoxic intracellular Ca21 overload. As expected, androgen protectsandrogen-sensitive LNCaP but not androgen-insensitive PC-3 cells fromcytotoxic intracellular Ca21 overload induced by Ca21 ionophore A23187.To provide evidence for the role of calreticulin in reducing cytotoxic effectof Ca21 influx in prostatic cells, we have shown that calreticulin antisenseoligonucleotide down-regulates calreticulin protein level and significantlyincreases the sensitivity to A23187-induced apoptosis in both LNCaP andPC-3 cells. Furthermore, calreticulin antisense oligonucleotide reversesthe androgen-induced resistance to A23187 in LNCaP cells. The aboveobservations collectively suggest that calreticulin mediates androgen reg-ulation of the sensitivity to Ca21 ionophore-induced apoptosis in LNCaPcells.

INTRODUCTION

Androgen plays an important role in the progression of prostatecancer. Androgen ablation is still the front line treatment of metastaticprostate cancer. Unfortunately, androgen ablation therapy is onlypalliative, and patients eventually relapse with androgen-independentprostate cancer, which is responsible for the death of about 40,000prostate cancer patients annually. There is a desperate need for newapproaches for prevention and/or treatment of androgen-independentprostate cancer. Understanding the mechanism of androgen action isimportant for investigating the role of androgen in prostate cancerprogression and may lead to the development of novel prostate cancertherapies.

Recently, we began to study the mechanism of androgen action inthe prostate by identification and characterization of androgen-response genes in the rat ventral prostate model. Using a PCR-basedcDNA subtraction method (1), we have searched extensively forandrogen-response genes (2). One of the identified androgen-responsegenes encodes calreticulin (2). Both Northern and Western blot anal-yses showed that castration dramatically down-regulates the expres-sion of calreticulin in the prostate, whereas androgen replacementrapidly up-regulates calreticulin mRNA and protein. The expressionof calreticulin in the prostate is much more abundant than that in any

other organs, and it appears that androgen regulates calreticulin ex-pression in male sex accessory organs only.In situ hybridization andimmunohistochemistry showed that calreticulin is an intracellularprotein abundantly expressed in the glandular prostatic epithelial cells(3). These observations suggest that calreticulin could play an impor-tant role in androgen action in the prostate.

Calreticulin is a highly conserved multifunctional protein found ina wide range of species (4, 5). Calreticulin was initially identified asa major lumenal Ca21 binding protein of endoplasmic reticula innonmuscle cells (6). The human mature calreticulin consists of 400amino acids and has a calculated molecular mass of 46.6 kDa (7). Itcontains one high-affinity and 25 low-affinity Ca21-binding sites (6,7). Calreticulin contains a KDEL endoplasmic reticulum retrievalsequence at its extreme COOH terminus and a putative nuclearlocalization signal in the middle of the protein (6, 7). As a highcapacity intracellular Ca21-binding protein, calreticulin can protectcells from cytotoxic intracellular Ca21 overload (6). Down-regulationof calreticulin expression by antisense oligonucleotide treatment low-ers the Ca21 response to bradykinin and increases the sensitivity toionomycin-induced cell death in neuroblastoma3 glioma NG-108-15cells (8). Conversely, overexpression of calreticulin increases theCa21 buffering capacity and protects HeLa cells against apoptosis (9).Also, marked reduction of calreticulin expression was observed priorto apoptosis in human leukemia HL-60 cells (10). These observationssuggest that calreticulin has the potential to influence apoptosis viamodulation of the intracellular Ca21 levels.

The importance of intracellular Ca21 elevation in apoptosis hasbeen demonstrated in aggressive prostate cancer cell lines (11, 12).Agents such as ionomycin and thapsigargin cause sustained intracel-lular Ca21 increase and lead to apoptosis. Overexpression of a Ca21

binding protein, calbindin D, protects prostate cancer cells againstintracellular Ca21 elevation (12). Because calreticulin is a high-capacity Ca21-binding protein and its expression is androgen depend-ent and highly abundant in prostate epithelial cells, calreticulin islikely to have a role in buffering intracellular Ca21 during androgenmanipulation in both normal and cancerous prostatic epithelial cells.

We have chosen androgen-sensitive human prostate cancer cell lineLNCaP as a model to investigate the role of calreticulin in androgenaction, particularly the role in the regulation of the intracellular Ca21

buffering capacity during androgen manipulation. The LNCaP iswidely used for studying the mechanism of androgen action becauseit retains critical characteristics of a normal prostatic epithelial cell,including androgen-sensitive cell proliferation and production ofPSA3 (13, 14). Thus, LNCaP appears to be an excellent model forinvestigating the effect of androgen on intracellular Ca21 bufferingand the role of calreticulin in androgen regulation of intracellularCa21 buffering. This report describes the interactions between andro-gen, calreticulin, and the sensitivity to Ca21 ionophore-induced apo-ptosis in LNCaP cells.

Received 10/14/98; accepted 2/18/99.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by Boehringer Ingelheim International GmbH, a CaPCURE award, Lester G. Wood Foundation, and NIH Grant R01 DK51193. N. Z. is arecipient of the AACR-Glaxo Wellcome Research Scholar Travel Award and the AACR-AFLAC Scholar Travel Award. Z. W. is a recipient of a Junior Faculty Research Awardfrom the American Cancer Society and an Alexander I. Newman Award.

2 To whom requests for reprints should be addressed, at Department of Urology, Tarry11-715, Northwestern University Medical School, Chicago, IL 60611. Phone: (312) 908-2264; Fax: (312) 908-7275; E-mail: [email protected].

3 The abbreviations used are: PSA, prostate-specific antigen; FBS, fetal bovine serum;GST, glutathioneS-transferase; CHX, cycloheximidine.

1896

Research. on March 25, 2021. © 1999 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

MATERIALS AND METHODS

Materials. Agarose, chloroform, CsCl, DMSO, ethanol, EDTA, formam-ide, guanidine thiocyanate, NaCl, Permount, SDS, Tris, xylene, and nylonmembrane were from Fisher Biotech. FBS, penicillin/streptomycin, PBS so-lution, RPMI 1640, and trypsin/EDTA solution were from Life Technologies,Inc. 4-(2-aminoethyl)benzenesulfonyl fluoride, anisomycin, calcium iono-phore A23187, cycloheximide, hydrogen peroxide, isopropyl-1-thio-b-D-gal-actoside, leupeptin, pepstatin, phenol, phenylmethylsulfonyl fluoride, andtrypan blue were from Sigma Chemical Co. ECL (enhanced chemilumines-cence) kit and goat anti-rabbit secondary antibody were from Amersham.Synthetic androgen analogue mibolerone, [32P]dCTP, and [35S]methioninewere from DuPont NEN.

Oligonucleotide Synthesis.Calreticulin sense and antisense phosphoro-thioate oligonucleotides (18-mers and 21-mers) were synthesized by the Bio-technology Facility at Northwestern University Medical School. The 18-merand 21-mer sequences were chosen according Liuet al. (8) and Leung-Hagesterijaet al. (15), respectively. The 18-mer pair is localized at the regioncoding for amino acids 83–88 (8), and the 21-mer pair is flanking thetranslation initiation region of the mRNA (15). The sequences of the oligonu-cleotides were as follows: 18-mer sense, 59-GAGCAGAACATCGACTGT -39;18-mer antisense, 59-ACAGTCGATGTTCTGCTC-39; 21-mer sense, 59-CG-GCCCGCCATGCTGCTATCC-39; and 21-mer antisense, 59-GGATAGCAG-CATGGCGGGCCG-39.

RNA Isolation and Northern Analysis. Total RNA was isolated using theguanidinium/CsCl gradient method (16). Purified RNA samples were fraction-ated in a 1% agarose-formaldehyde gel. Tenmg of total RNA were loaded ineach lane. After electrophoresis, RNA was transferred to a nylon membrane bycapillary blotting and then cross-linked to the membrane by UV irradiation.Northern hybridization of the membrane was carried out at 42°C overnight ina buffer containing 53 SSPE, 23Denhart’s solution, 0.1% SDS, 0.1 mg/mldenatured salmon sperm DNA, and 50% formamide in the presence of humancalreticulin DNA probes (IMAGE Consortium, EST: ze97a10.s1) labeled byrandom priming. The membrane was then washed at room temperature with13 SSC and 0.1% SDS for 20 min, followed by three 20-min washes at 65°Cwith 0.23SSC and 0.1% SDS. The blot was exposed to X-ray film at280°C.

Cell Culture. The human prostate cancer cell line LNCaP and PC3 werepurchased from American Type Culture Collection. LNCaP and PC3 cellswere cultured in RPMI 1640 with 10% FBS and 1% penicillin-streptomycin in12-well plates at 37°C in 5% CO2 and 95% air to 60–70% confluency. FBSwas replaced with charcoal-stripped FBS in the androgen treatment experi-ment. To study the effect of androgen on the sensitivity to Ca21 ionophoreA23187, LNCaP cells were pretreated for 2 days with androgen analoguemibolerone with ethanol as the vehicle prior to the addition of A23187. Tostudy the effect of calreticulin down-regulation on the cell sensitivity to Ca21

ionophore, LNCaP cells were pretreated for 18 h with 10mM calreticulinantisense or sense oligonucleotides with H2O as the vehicle prior to theaddition of A23187. After the pretreatment, LNCaP cells were treated for 2days with Ca21 ionophore A23187 dissolved in DMSO at various concentra-tions. Control groups received vehicle(s) only. The volume of each vehicle is10 ml in 1 ml of culture medium per well in 12-well plates.

After the Ca21 ionophore treatment, cell viability was determined usingtrypan blue staining. LNCaP cells were harvested by treatment with 0.05%trypsin/0.53 mM EDTA. Cells were centrifuged at 2003 g for 5 min to removetrypsin/EDTA. After the cell pellets were resuspended in 500ml of PBS, 500ml of 0.4% trypan blue solution were added and mixed. Trypan blue cellsuspension was left at room temperature for 10 min before the cells werecounted with a hemocytometer. Cell viability is defined as the percentage ofthe unstained alive cells in the whole-cell population. Each experiment wasrepeated three times.

For protein synthesis inhibition experiments, cells were treated for 2 h withcycloheximide (50mg/ml) and anisomycin (80mg/ml) in ethanol vehiclebefore the addition of mibolerone. After 36 h of mibolerone treatment, totalRNA was extracted from the cells. The efficiency of protein synthesis inhibi-tion was measured by [35S]methionine incorporation (17).

DNA Fragmentation Assay. A simplified DNA fragmentation assaymethod (18) has been used. LNCaP cells were cultured in 12-well plates andtreated with antisense oligo and Ca21 ionophore as described previously.About 1 3 106 cells were collected from each group separately. Cells were

washed once with ice-cold PBS and centrifuged, and the supernatants wereremoved carefully. Then, cells were dispersed in 30ml of lysis buffer (10 mM

Tris, 100 mM NaCl, 25 mM EDTA, and 1% sarcosyl/pH 8.0) by gentlevortexing, and 4ml of proteinase K (10mg/ml) were added. The cell lysateswere incubated at 45°C for 1–2 h. Twoml of RNase (10mg/ml) were added,and the cell lysates were continuously incubated for 1 h at room temperature.Fourml of 63 DNA loading dye was mixed in the lysate (the final volume ofeach sample was about 40ml), and the samples were analyzed on a 2% agarosegel.

Protein Preparation and Western Blot. Protein extract was preparedfrom LNCaP and PC3 cells by sonication in a lysis buffer consisting of 13PBS containing 1% SDS, 10 mM EDTA, 100mM phenylmethylsulfonyl fluo-ride, 10mM leupeptin, 200mM 4-(2-aminoethyl)benzenesulfonyl fluoride, and1 mM pepstatin. Protein concentration was determined by the Lowry methodusing the Bio-Rad DC protein assay kit. Samples were separated by 10% SDSpolyacrylamide gel electrophoresis. A standard Western blot procedure wasused (19), which is based on the anti-calreticulin antibody and the secondaryantibody linked with horseradish peroxidase. The polyclonal anti-calreticulinantibody was generated using a GST fusion protein technique. The rat calre-ticulin cDNA was inserted into a pGEX expression vector and transformed intoBL21 Escherichia coli. The GST-calreticulin fusion protein was induced byisopropyl-1-thio-b-D-galactoside and purified using glutathione/agarose beads.The polyclonal antibody was generated by injection of calreticulin fusionprotein into rabbit. An anti-calreticulin polyclonal antibody from Stress-Genwas also tried and showed a similar result. The enhanced chemiluminescence(ECL) method was used to detect the conjugated horseradish peroxidase.

Immunocytochemistry. The LNCaP cells were plated on sterile 18-mm2

coverslips and incubated for 2 days to 60–70% confluency as describedpreviously. Cells were fixed by 4% paraformaldehyde in PBS at 4°C for 1 h.Then cells were rinsed three times in PBS for 5 min each. The endogenousperoxidase activity of the cells was quenched by incubating with 3% hydrogenperoxide in methanol for 30 min, followed by rinsing twice in PBS with 5 mineach. Nonspecific binding sites were blocked by incubating with 2% goatserum in PBS at room temperature for 1 h. Cells were incubated with anti-calreticulin antibody (1:1500 dilution in the blocking solution) in a humidifiedbox at 4°C overnight. The control group was incubated with a mixture of theprimary antibody and the GST-calreticulin fusion protein. Cells were rinsedtwice in PBS with 5 min each. Then the cells were processed as described inthe Vectastain ABC kit from Vector Laboratories and developed in 3,39-diaminobenzidine substrate (Vector Laboratories) for 5 min. After stopping thereaction in distilled water, the coverslips were dehydrated in an ethanol series,cleared in xylenes, and dried overnight at room temperature. Finally, thecoverslips were mounted with Permount. Pictures were taken using an Olym-pus VANOX-S camera and an Olympus AH-2 microscope.

RESULTS

Calreticulin Is a Primary Androgen-Response Gene in LNCaPCells. To use the LNCaP cell line to study calreticulin function inandrogen action, it is important to first characterize the expression ofcalreticulin in response to androgen in this androgen-sensitive cellline. Fig. 1A shows that calreticulin mRNA is regulated by androgenin LNCaP cells and androgen induction of calreticulin mRNA resistsprotein synthesis inhibition. PSA (20–22) was probed as a control forearly androgen-response genes, and its induction partially resistsprotein synthesis inhibition (Fig. 1A). Glyceraldehyde-3-phosphatedehydrogenase was probed as a control for androgen-insensitive gene.The inhibition of protein synthesis by cycloheximide and anisomycinin cultured LNCaP cells was.97%, as assayed by measuring[35S]methionine incorporation (data not shown). Our observation in-dicates that calreticulin is a primary androgen-response gene in cul-tured LNCaP cells.

The androgen treatment time for inducing calreticulin mRNA inLNCaP cells was 36 h in the presence of CHX and anisomycin. Thisandrogen induction time was chosen because the total RNA yielddrops dramatically in LNCaP cells after 36 h in the presence of CHX.In fact, partial RNA degradation occurs within 36 h of the CHX

1897

CALRETICULIN IN ANDROGEN ACTION AND APOPTOSIS IN LNCaP



treatment, as indicated by the smear of 18S and 28S rRNAs inmethylene blue staining (Fig. 1A).

We have also studied the effect of androgen on calreticulin at theprotein level. As expected, calreticulin protein is regulated by andro-gen in LNCaP cells (Fig. 1B).

Calreticulin Is Localized in an Endoplasmic Reticulum-likeStructure in LNCaP Cells. Previous studies have showed that cal-reticulin is often colocalized with the membrane system of the endo-plasmic reticulum (23). Fig. 2 shows the immunocytochemistry studyof the intracellular distribution of calreticulin in LNCaP cells. Thedistribution in LNCaP is virtually identical to that in other cell types,such as a rat pigment epithelial cell line and dedifferentiated chickembryo cardiac myocytes (23), where calreticulin is predominantlylocalized in the endoplasmic reticulum. Thus, it appears that calreti-culin is associated with the endoplasmic reticulum in LNCaP cells.However, we cannot rule out the possibilities that calreticulin may beassociated with specialized areas of the endoplasmic reticulum and

that a small amount of calreticulin may localize in areas other thanendoplasmic reticulum.

Ca21 Ionophore A23187 Induces Apoptosis in LNCaP andPC-3 Cells. It has been demonstrated previously that Ca21 ionophoreionomycin induces apoptosis in prostatic cancer cells (11). It isexpected, but has not been proven, that other Ca21 ionophores such asA23187 also induce apoptosis. Although Ca21 ionophore A23187 hasbeen used frequently to induce cytotoxic intracellular Ca21 influx, noexperiment has been carried out to test whether A23187-induced celldeath is apoptotic. Fig. 3Ashows that A23187 induces internucleo-somal DNA fragmentation, a hallmark of apoptosis, in both androgen-sensitive LNCaP and androgen-insensitive PC-3 prostate cancer celllines. Thus, A23187 induces cell death via apoptosis. Because it isdifficult to quantitatively determine cell death using DNA fragmen-tation assay, we also used the trypan blue staining method to measurecell death of LNCaP and PC-3 in response to A23187. Fig. 3Bshowsthe dose-response curves of cell viabilityversusA23187 in LNCaP

Fig. 1. Androgen regulation of calreticulin expression in cultured LNCaP cells.A,Northern blot analysis. The effect of protein synthesis inhibition on androgen-induced up-regulation of calreticulin (CRT), PSA, and glyceraldehyde-3-phosphatedehydrogenase (GAPDH) genes was characterized. LNCaP cells were treated with orwithout 10 nM mibolerone (Mib) for 36 h in the presence or absence of proteinsynthesis inhibitors CHX plus anisomycin (CHX) that were added 2 h before theaddition of mibolerone. Total RNA was extracted from the harvested cells andelectrophoresed for Northern blot analysis. The total RNA loading was showed bymethylene blue staining.B, Western blot analysis. Protein extract was prepared asdescribed in “Materials and Methods” from LNCaP cells in the absence or presenceof 10 nM mibolerone. Calreticulin protein migrates at 60 kDa in the Western blot.The amount of protein extracts loaded on the gels was examined by stainingthe transferred nitrocellulose membrane with Ponceau-S.

Fig. 2. Immunocytochemistry analysis of calreticulin inLNCaP cells. Cultured LNCaP cells were probed with anti-calreticulin polyclonal antibody in the absence (A) or presence(B) of GST-calreticulin fusion protein.

1898




and PC-3 cells using trypan blue assay. More than 50% of the cells aredead in the presence of 5mM or higher concentrations of A23187.Virtually all of the cells are dead in the presence of 20mM A23187.

Calreticulin Antisense Oligonucleotides Down-Regulate Calre-ticulin and Enhances the Sensitivity to Ca21 Ionophore in LNCaPand PC-3 Cells.To examine whether calreticulin is involved in theregulation of the sensitivity to Ca21 ionophore in prostate cancercells, we studied the effect of calreticulin antisense oligonucleotides

on Ca21 ionophore-induced cell death in LNCaP and PC-3 cells.Treatment with calreticulin antisense oligonucleotides significantlyincreases the sensitivity of LNCaP cells to Ca21 ionophore A23187-induced cell death (Fig. 4A). The 21-mer pair of antisense and sensephosphorothioate oligonucleotides (see “Materials and Methods”) wasused in the experiment. None of the oligonucleotides were toxic toLNCaP cells because treatment with oligonucleotides (10mM) alonehad no effect on LNCaP viability, as determined using trypan blue

Fig. 4. The effect of antisense oligonucleotide on the sensi-tivity to Ca21 ionophore A23187 and calreticulin expression inLNCaP (AandC) and PC-3 (BandD). A andB, the effect ofA23187 at indicated concentrations on cell viability in theabsence of oligonucleotides (control,M) or in the presence ofeither the antisense (f) or sense (p]) 21-mer oligonucleotides.Values are the means of three independent measurements;bars,SD. C andD, Western blot analysis of calreticulin with mocktreatment (M) in the absence oligonucleotides and in the pres-ence of sense (S) or antisense (A/S) oligonucleotides for theindicated number of days (d). An equal amount of proteinextracts was loaded in each lane. The amount of protein extractsloaded in the gels was examined by staining the transferrednitrocellulose membrane with Ponceau-S.

Fig. 3. The effect of Ca21 ionophore A23187 onLNCaP and PC-3 cells.A, induction of DNA frag-mentation by A23187 in LNCaP and PC-3 cells.Cultured cells were treated with (1) or without (2)20 mM A23187 for 2 days before the DNA fragmen-tation assay.B, dose-response curves of A23187concentrationversuspercentage of cell viability ofLNCaP (solid line) and PC-3 (dotted line) as deter-mined by trypan blue assay. Cultured cells weretreated with various concentrations of A23187 for 2days before the trypan blue assay. Values are themeans of three independent measurements;bars,SD.

1899




staining. The treatment of sense oligonucleotides did not change thesensitivity of LNCaP cells to A23187. In contrast, the treatment withantisense oligonucleotides significantly enhanced the sensitivity ofLNCaP cells to A23187. Antisense oligonucleotide treatment causes a20–45% decrease in cell viability in the presence of A23187. Theeffect of calreticulin antisense oligonucleotides on PC-3 cells is verysimilar to the effect on LNCaP cells (Fig. 4B).

Antisense oligonucleotide treatment reduced calreticulin proteinlevels in both LNCaP and PC-3 cells (Fig. 4,C andD). Densitometric

analysis shows that antisense treatment for 1 day caused a 2-folddecrease in calreticulin protein level, and antisense treatment for 2days caused a 5-fold decrease (data not shown). The down-regulationof calreticulin protein by an antisense oligonucleotide in this experi-ment is comparable with that in other cell lines (8, 15).

It is important to point out that another pair of antisense and senseoligonucleotides, the 18-mer pair (see “Materials and Methods”), wasalso used in the experiment, and the results were virtually identical tothat described above with the 21-mer pair. The fact that both pairs ofcalreticulin antisense and sense oligonucleotides had identical effectson calreticulin protein expression and the sensitivity to A23187 incultured LNCaP cells indicates that antisense oligonucleotides arespecific to calreticulin mRNA.

Androgen Regulates the Sensitivity of LNCaP but not PC-3Cells to Ca21 Ionophore A23187.To study the effect of androgenon intracellular Ca21 buffering capacity, LNCaP cells were culturedin charcoal-stripped FBS for 1 day and then treated with a syntheticandrogen analogue, mibolerone, at 0 or 10 nM for an additional 2 daysbefore the addition of A23187. Up-regulation of calreticulin by an-drogen is expected to protect LNCaP cells from Ca21 ionophoreA23187-induced cytotoxic intracellular Ca21 overload via enhancingthe Ca21 buffering capacity. Fig. 5A shows that mibolerone had asignificant protective effect against A23187 in LNCaP cells. There areabout 20–30% more cells alive in the whole population in the pres-ence of 10 nM mibolerone. The effect of mibolerone is most strikingin the presence of 20mM A23187. In the absence of mibolerone,,5%of cells are alive when treated with 20mM A23187 for 2 days. Incontrast, in the presence of 10 nM mibolerone, 35% of LNCaP cellsare still alive after 2 days of treatment with 20mM A23187. Thetreatment of 10 nM mibolerone caused a 7-fold enhancement in cellviability in the presence of 20mM A23187.

As a comparison, we also studied the effect of androgen on thesensitivity of Ca21 ionophore-induced apoptosis in PC3, an androgen-insensitive human prostate cancer cell line. Fig. 5B shows that andro-gen had little or no effect on the sensitivity of PC-3 cells to A23187.Also, as expected, calreticulin expression is not influenced by mibo-lerone in PC-3 cells (result not shown).

Antisense Oligonucleotide Reverses the Androgen-induced Re-sistance to Ca21 Ionophore in LNCaP Cells. To further investigatethe relationship between androgen, calreticulin, and the sensitivity ofLNCaP to Ca21 ionophore A23187-induced cell death, we studied theeffect of calreticulin antisense oligonucleotide on the sensitivity ofLNCaP to A23187 either in the presence or absence of mibolerone

Fig. 5. The effect of androgen on the sensitivity of LNCaP (A) and PC3 (B) cells toCa21 ionophore A23187. The GraphPad PRISM software was used to generate the bargraph of cell viability at the indicated A23187 concentrations in the presence (■) orabsence (h) of 10 nM mibolerone. Values are the means of three independent measure-ments;bars,SD.

Fig. 6. The effect of calreticulin antisense oli-gonucleotide on the sensitivity of LNCaP to Ca21

ionophore A23187 in the presence or absence ofmibolerone. A, cell viability at the indicatedA23187 concentrations in LNCaP cells culturedwith or without 10 nM mibolerone (Mib) in thepresence of either antisense or sense 21-mer oligo-nucleotides.Bars,SD. B, Western blot analysis ofcalreticulin in LNCaP cells cultured with or with-out 10 nM mibolerone in the presence of eithersense (S) or antisense (AS) oligonucleotides. Theamount of protein extracts loaded in the gels wasexamined by staining the transferred nitrocellulosemembrane with Ponceau-S.

1900




(Fig. 6A). The protective effect of mibolerone on LNCaP cells wasabrogated in the presence of calreticulin antisense oligonucleotide. Asa control, the down-regulation of calreticulin protein by antisenseoligonucleotide was demonstrated by Western blot (Fig. 6B). The cellviability of LNCaP to A23187 in the presence of both antisenseoligonucleotide and mibolerone is higher than that in the presence ofantisense oligonucleotide but absence of mibolerone. This differencein cell viability correlates with the difference in calreticulin proteinlevel under these two treatment conditions. It appears that the morecalreticulin protein is expressed, the higher percentage of viable cellscan be detected.

DISCUSSION

The present experiments have studied the interplay between andro-gen, calreticulin, and the sensitivity to Ca21 ionophore in LNCaPcells. Our results showed that androgen can regulate the sensitivity toCa21 influx-induced apoptosis, and calreticulin is involved in thisandrogen-dependent process in LNCaP cells. This observation sug-gests that calreticulin is likely to play an important role in modulatingintracellular Ca21 in prostatic epithelial cellsin vivo during androgenmanipulation.

The involvement of intracellular Ca21 level in androgen action inprostatic epithelial cells was established previously by several elegantstudies (24–26): (a) cell death induced by Ca21 ionophore is indis-tinguishable from the cell death induced by castration in the prostate(26); (b) Ca21 channel blockers can inhibit as much as 70% of thecastration-induced increase in the rate of cell death in the prostate (24,26); and (c) the channel blocker also inhibits the induction of Ca21-response genes in prostatic epithelial cells after castration (27). Theseobservations suggest that castration induces Ca21 influx throughCa21 channels in prostatic epithelial cells because intracellular Ca21

elevation could be inhibited by channel blocker. We have demon-strated that androgen regulates the sensitivity of LNCaP cells to Ca21

ionophore A23187-induced apoptosis, suggesting that androgen couldalso modulate intracellular Ca21 buffering capacity in prostatic epi-thelial cells. A combination of Ca21 influx and reduction of Ca21

buffering capacity could cause apoptosis very effectively in prostaticepithelial cells upon castration.

Androgen regulation of the sensitivity of cytotoxic Ca21 overloadis likely to be mediated through the expression of an androgen-response gene(s) because the androgen receptor is a ligand-dependenttranscription factor. Calreticulin has the potential to mediate androgenregulation of the sensitivity to A23187 because it is an intracellularCa21-binding protein encoded by a primary androgen-response geneand is abundantly expressed in prostatic epithelial cells in the ratmodel. The present study showed that in LNCaP cells calreticulin isalso a primary androgen-response gene, and both calreticulin mRNAand protein are regulated by androgen, which agrees with our previousobservations in the rat model (3).

The role of calreticulin in buffering intracellular Ca21 wasdemonstrated previously by the observation that down-regulationof calreticulin by antisense oligonucleotides significantly increasesthe sensitivity of NG-108-15, a neuroblastoma3 glioma cell line,to A23187-induced cytotoxic intracellular Ca21 overload (8). Therole of calreticulin in prostatic epithelial cells is likely to be similarto its role in other types of cells because calreticulin intracellularlocalization in LNCaP cells is virtually the same as its intracellularlocalization in other types of cells. The role of calreticulin inbuffering intracellular Ca21 in prostatic epithelial cells is furthersupported by the observation that down-regulation of calreticulinby antisense oligonucleotides significantly increases the sensitivityto A23187 in both LNCaP and PC-3 cells. Two different pairs of

calreticulin sense and antisense oligonucleotides caused the sameeffect, indicating the antisense oligonucleotides are specific tocalreticulin. Our study has also extended previous studies bydemonstrating that A23187 induces cell death in LNCaP and PC-3cells via apoptosis.

To further support the importance of calreticulin in androgen reg-ulation in the sensitivity to A23187-induced apoptosis, we have dem-onstrated that calreticulin antisense oligonucleotide significantly in-hibits the protective effect of androgen. This observation stronglysuggests that calreticulin plays a major role in regulating the intracel-lular Ca21 buffering capacity in LNCaP cells during androgen ma-nipulation.

Although androgen induction of calreticulin in LNCaP cells is notas dramatic as the induction in the rat ventral prostate model, modestchange in calreticulin level could have a profound effect on cells. Forexample, a 1.6-fold increase of calreticulin expression in mouse Lfibroblast cell line increases intracellular Ca21 storage and decreasesstore-operated Ca21 influx (28). It is possible that androgen couldhave a more profound effect on intracellular Ca21 buffering capacityin vivo because the response of calreticulin to androgen is verydramatic in the prostatein vivo.

Down-regulation of calreticulin does not appear to be sufficientto cause apoptosis in prostatic epithelial cells: (a) most epithelialcells in castrated prostate are still alive, although they express verylow levels of calreticulin (3); (b) down-regulation of calreticulin inLNCaP cells by androgen deprivation or antisense oligonucleotidesdoes not influence the viability of the cells; and (c) calreticulin alsoappears to be nonessential to the survival of other types of cells inculture. For example, homozygous knockout of the calreticulingene in mouse embryonic stem cells does not influence the cellviability (29).

Although the calreticulin expression level has been shown to beassociated with the sensitivity of LNCaP cells to Ca21 ionophore, themechanism of such an association remains unclear. Further charac-terization of calreticulin in LNCaP model may shed more light on themechanism by which calreticulin regulates intracellular Ca21 buffer-ing and/or signaling in prostatic cells.

ACKNOWLEDGMENTS

We thank Karen L. Kaul for providing the PSA cDNA, and we thank ourcolleagues for critical reading of the manuscript.

REFERENCES

1. Wang, Z., and Brown, D. A gene expression screen. Proc. Natl. Acad. Sci. USA,88:11505–11509, 1991.

2. Wang, Z., Tufts, R., Haleem, R., and Cai, X. Genes regulated by androgen in the ratventral prostate. Proc. Natl. Acad. Sci. USA,94: 12999–13004, 1997.

3. Zhu, N., Pewitt, E. B., Cai, X., Cohn, E. B., Lang, S., Chen, R., and Wang, Z.Calreticulin: an intracellular Ca11-binding protein abundantly expressed andregulated by androgen in prostatic epithelial cells. Endocrinology,139: 4337–4344, 1998.

4. Michalak, M., Milner, R., Burns, K., and Opas, M. Calreticulin. Biochem. J.,285:681–692, 1992.

5. Sontheimer, R. D., Nguyen, T. Q., Cheng, S. T., Lieu, T. S., and Capra, J. D. Theunveiling of calreticulin–a clinically relevant tour of modern cell biology. J. Investig.Med., 43: 362–370, 1995.

6. Krause, K. H., and Michalak, M. Calreticulin. Cell,88: 439–443, 1997.7. Nash, P., Opas, M., and Michalak, M. Calreticulin: not just another calcium-binding

protein. Mol. Cell. Biochem.,135: 71–78, 1994.8. Liu, N., Fine, R. E., Simons, E., and Johnson, R. J. Decreasing calreticulin expression

lowers the Ca21 response to bradykinin and increases sensitivity to ionomycin inNG-108–15 cells. J. Biol. Chem.,269: 28635–28639, 1994.

9. Bastianutto, C., Clementi, E., Codazzi, F., Podini, P., De Giorgi, F., Rizzuto, R.,Meldolesi, J., and Pozzan, T. Overexpression of calreticulin increases the Ca21

capacity of rapidly exchanging Ca21 stores and reveals aspects of their lumenalmicroenvironment and function. J. Cell Biol.,130: 847–855, 1995.

10. Nakajo, S., Okamoto, M., Masuda, Y., Sakai, I., Ohsawa, S., and Nakaya, K.Geranylgeraniol causes a decrease in levels of calreticulin and tyrosine phosphoryl-

1901




ation of a 36-kDa protein prior to the appearance of apoptotic features in HL-60 cells.Biochem. Biophys. Res. Commun.,226: 741–745, 1996.

11. Martikainen, P., Kyprianou, N., Tucker, R., and Isaacs, J. Programmed death ofnonproliferating androgen-independent prostatic cancer cells. Cancer Res.,51: 4693–4700, 1991.

12. Furuya, Y., Lundmo, P., Short, A., Gill, D., and Isaacs, J. The role of calcium, pH,and cell proliferation in the programmed (apoptotic) death of androgen-independentprostatic cancer cells induced by thapsigargin. Cancer Res.,54: 6167–6175, 1994.

13. Horoszewicz, J. S., Leong, S. S., Kawinski, E., Karr, J. P., Rosenthal, H., Chu, T. M.,Mirand, E. A., and Murphy, G. P. LNCaP model of human prostatic carcinoma.Cancer Res.,43: 1809–1818, 1983.

14. Lee, C., Sutkowski, D., Sensibar, J., Zelner, D., Kim, I., Amsel, I., Shaw, N., Prins,G., and Kozlowski, J. Regulation of proliferation and production of prostate-specificantigen in androgen-sensitive prostatic cancer cells, LNCaP, by dihydrotestosterone.Endocrinology,136: 796–803, 1995.

15. Leung-Hagesteijn, C. Y., Milankov, K., Michalak, M., Wilkins, J., and Dedhar, S.Cell attachment to extracellular matrix substrates is inhibited upon downregulation ofexpression of calreticulin, an intracellular integrina-subunit-binding protein. J. CellSci., 107: 589–600, 1994.

16. Chirgwin, J., Przbyla, A., MacDonald, R., and Rutter, W. Isolation of biologicallyactive ribonucleic acid from sources enriched in ribonuclease. Biochemistry,18:5294–5299, 1979.

17. Hu, G., Gura, T., Sabsay, B., Sauk, J., Dixit, S. N., and Veis, A. Endoplasmic reticulumprotein Hsp47 binds specifically to the N-terminal globular domain of the amino-propeptide of the procollagen Ia1 (I)-chain. J. Cell. Biochem.,59: 350–367, 1995.

18. Zhu, N., and Wang, Z. An assay for DNA fragmentation in apoptosis withoutphenol/chloroform extraction and ethanol precipitation. Anal. Biochem.,246: 155–158, 1997.

19. Bollag, D. M., and Edelsten, S. J (eds.). Protein Methods, pp. 181–208. New York:Wiley-Liss, 1991.

20. Young, C., Montgomery, B., Andrews, P., Qui, S., Bilhartz, D., and Tindall, D.Hormonal regulation of prostate-specific antigen messenger RNA in human prostaticadenocarcinoma cell line LNCaP. Cancer Res.,51: 3748–3752, 1991.

21. Henttu, P., Liao, S., and Vihko, P. Androgens up-regulate the human prostate-specificantigen messenger ribonucleic acid (mRNA), but down-regulate the prostatic acidphosphatase mRNA in the LNCaP cell line. Endocrinology,130: 766–772, 1992.

22. Schuur, E., Henderson, G., Kmetec, L., Miller, J., Lamparski, H., and Henderson, D.Prostate-specific antigen expression is regulated by an upstream enhancer. J. Biol.Chem.,271: 7043–7051, 1996.

23. Opas, M. The intracellular distribution and expression of calreticulin.In: M. Michalak(ed.), Calreticulin, pp. 31–36. Austin: R. G. Landes Company, 1996.

24. Connor, J., Sawczuk, I., Benson, M., Tomashefsky, P., O’Toole, K., Olsson, C., andButtyan, R. Calcium channel antagonists delay regression of androgen-dependenttissues and suppress gene activity associated with cell death. Prostate,13: 119–130,1988.

25. Kyprianou, N., English, H., and Isaacs, J. Activation of a Ca21-Mg21-dependentendonuclease as an early event in castration-induced prostatic cell death. Prostate,13:103–117, 1988.

26. Martikainen, P., and Isaacs, J. Role of calcium in the programmed death of ratprostatic glandular cells. Prostate,17: 175–187, 1990.

27. Sells, S., Wood, D., Jr., Joshi-Barve, S., Muthukumar, S., Jacob, R., Crist, S.,Humphreys, S., and Rangnekar, V. Commonality of the gene programs induced byeffectors of apoptosis in androgen-dependent and -independent prostate cells. CellGrowth Differ., 5: 457–466, 1994.

28. Mery, L., Mesaeli, N., Michalak, M., Opas, M., Lew, D. P., and Krause, K. H.Overexpression of calreticulin increases intracellular Ca21 storage and decreasesstore-operated Ca21 influx. J. Biol. Chem.,271: 9332–9339, 1996.

29. Coppolino, M. G., Woodside, M. J., Demaurex, N., Grinstein, S., St. Arnaud, R., andDedhar, S. Calreticulin is essential for integrin-mediated calcium signalling and celladhesion. Nature (Lond.),386: 843–847, 1997.

1902




1999;59:1896-1902. Cancer Res Ning Zhu and Zhou Wang Apoptosis in LNCaP Prostate Cancer CellsRegulation of the Sensitivity to Calcium Ionophore-induced Calreticulin Expression Is Associated with Androgen

Updated version

http://cancerres.aacrjournals.org/content/59/8/1896

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/59/8/1896.full#ref-list-1

This article cites 26 articles, 12 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/59/8/1896.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/59/8/1896To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/59/8/1896.full#ref-list-1

http://cancerres.aacrjournals.org/content/59/8/1896.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



calreticulin expression is associated with androgen ...€¦ · inc....

Documents