Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/7493450

MolecularPathwaysInvolvedintheAntineoplasticEffectsofCalcitriolonInsulinomaCells

ARTICLEinENDOCRINOLOGY·JUNE2003

ImpactFactor:4.5·DOI:10.1210/en.2002-221014·Source:PubMed

CITATIONS

20

READS

32

8AUTHORS,INCLUDING:

PhilippeDupraz

MerckSerono

18PUBLICATIONS696CITATIONS

SEEPROFILE

PaoloFiorina

OspedalediSanRaffaeleIstitutodiRicover…

168PUBLICATIONS2,965CITATIONS

SEEPROFILE

UgoCavallaro

IEO-IstitutoEuropeodiOncologia

55PUBLICATIONS2,602CITATIONS

SEEPROFILE

AlbertoMDavalli

OspedalediSanRaffaeleIstitutodiRicover…

99PUBLICATIONS3,019CITATIONS

SEEPROFILE

Availablefrom:UgoCavallaro

Retrievedon:03February2016

Molecular Pathways Involved in the AntineoplasticEffects of Calcitriol on Insulinoma Cells

FRANCESCA GALBIATI, LUCA POLASTRI, BERNARD THORENS, PHILIPPE DUPRAZ,PAOLO FIORINA, UGO CAVALLARO, GERHARD CHRISTOFORI, AND ALBERTO M. DAVALLI

Division of General Medicine, Unit of Endocrinology and Metabolic Disease, San Raffaele Scientific Institute (F.G., L.P.,P.F., A.M.D.), 20132 Milan, Italy; Institute of Pharmacology, University of Lausanne (B.T., P.D.), CH-1005 Lausanne,Switzerland; and Institute of Biochemistry and Genetics, University of Basel (U.C., G.C.), CH-4051 Basel, Switzerland

We have previously reported that in tumorigenic pancreatic�-cells, calcitriol exerts a potent antitumorigenic effect byinducing apoptosis, cell growth inhibition, and reduction ofsolid �-cell tumors. Here we have studied the molecular path-ways involved in the antineoplastic activity of calcitriol onmouse insulinoma �TC3 cells, mouse insulinoma �TC express-ing or not expressing the oncogene p53, and �TC-tet cellsoverexpressing or not the antiapoptotic gene Bcl2. Our resultsindicate that calcitriol-induced apoptosis was dependent onthe function of p53 and was associated with a biphasic in-crease in protein levels of transcription factor nuclear factor-�B. Calcitriol decreased cell viability by about 40% in p53-retaining �TC and in �TC3 cells; in contrast, �TC p53�/� cellswere only minimally affected. Calcitriol-induced cell death

was regulated by members of the Bcl-2 family of apoptosisregulatory proteins, as shown by calcitriol-induced up-regu-lation of proapoptotic Bax and Bak and the lack of calcitriol-induced cytotoxicity in Bcl-2-overexpressing insulinomacells. Moreover, calcitriol-mediated arrest of �TC3 cells in theG1 phase of the cell cycle was associated with the abnormalexpression of p21 and G2/M-specific cyclin B2 genes and in-volved the DNA damage-inducible factor GADD45. Finally, in�TC3 cells, calcitriol modulated the expression of IGF-I andIGF-II genes. In conclusion, these findings contribute to theunderstanding of the antitumorigenic effects of calcitriol ontumorigenic pancreatic �-cells and further support the ratio-nale of its utilization in the treatment of patients with malig-nant insulinomas. (Endocrinology 144: 1832–1841, 2003)

TODAY IT IS well recognized that the active form ofvitamin D3 (1,25-dihydroxyvitamin D3 [1,25-

(OH)2D3]), also known as calcitriol, acts as an effective reg-ulator of cell growth and differentiation in a number ofdifferent cell types, including neoplastic cells (1). In the pastdecade several studies have demonstrated the antineoplasticeffect of calcitriol on a variety of cancer cell lines, supportingits potential use in the treatment of diverse types of malig-nancies. Malignant insulinomas are rare endocrine tumorsthat, in contrast to benign �-cell adenomas, are rarely curedby surgery (2). In addition, malignant insulinomas respondpoorly to conventional chemotherapy and treatment withsomatostatin (or somatostatin analogs), which are usuallyeffective on other types of endocrine tumors (3, 4). Moreover,hypoglycemia associated with malignant insulinomas is gen-erally unresponsive to agents that inhibit insulin release,such as diazoxide and certain calcium channel blockers (5–7).

We have recently reported that in murine insulinoma �TC3cells, calcitriol induces growth inhibition and apoptosis (8).In the same study we have shown that the cytotoxic effectsof calcitriol are limited to �-cells with a malignant phenotype(�TC3), whereas benign human insulinoma cells and normalhuman islets are unaffected. We have also demonstrated thata short course of treatment with calcitriol reduces signifi-cantly the mass of �-cell tumors in transgenic RIP1Tag2 mice(8), a well characterized model of �-cell tumorigenesis (9).

The nuclear vitamin D receptor (nVDR) is a member of thenuclear receptor superfamily that functions as a transcrip-tional factor and is also present in pancreatic �-cells (10).Calcitriol acts mainly through the so-called genomic path-way, which involves binding of the hormone to its nVDR.The complex calcitriol/nVDR binds to vitamin D-responsiveelements in the promoter region of target genes and mod-ulates their transcription. Vitamin D-responsive elementshave also been identified in the promoter region of nonclas-sical vitamin D target genes such as calbindin D (11), p21 (12),c-Fos, and TGF�2 (13). It has now become evident that cal-citriol also induces nontranscriptional responses through theactivation of a putative, yet unidentified, membrane receptor(14, 15). Indeed, calcitriol increases the intracellular levels oractivities of several signaling molecules, including proteinkinase C (PKC), Raf, MAPK, and Src kinases (16–18). Werecently reported that in �TC3 cells as well as in human islets,calcitriol activates the MAPK cascade and induces PKC ac-tivation via a nongenomic pathway (8). In �TC3 cells, MAPKactivation contributes to calcitriol-induced cytotoxicity, asMAPK kinase (MEK) inhibition with UO126 significantlyprevents this effect (8). We also reported that calcitriol in-creases caspase-3 activity in �TC3 cells (8), but no particularefforts were posed in that study to explore in more detail themechanisms responsible for the antineoplastic effect of cal-citriol on these cells.

The aim of this study was to elucidate the molecular path-ways involved in the antitumorigenic effects of calcitriol on�TC3 cells. The results show that calcitriol increased thelevels of p53 protein and p53 phosphorylated at serine 15, aneffect that was prevented by staurosporine, but not by the

Abbreviations: cdk, Cyclin-dependent kinase; 1,25-(OH)2D3, 1,25-dihydroxyvitamin D3; MEK, MAPK kinase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NF�B, nuclear factor �B; nVDR,nuclear vitamin D receptor; PDTC, pyrrolidine dithiocarbamate; PKC, pro-tein kinase C.

0013-7227/03/$15.00/0 Endocrinology 144(5):1832–1841Printed in U.S.A. Copyright © 2003 by The Endocrine Society

doi: 10.1210/en.2002-221014

1832

MEK inhibitor (UO126). Moreover, calcitriol-induced �TC3cell apoptosis was associated to a biphasic increase in theprotein levels of nuclear factor-�B (NF�B), which appears tobe a prosurvival cell adaptation. Calcitriol transcriptionallyactivated several p53-regulated genes involved in the regu-lation of cell cycle (p21 and G2/M-specific cyclin B2), DNAdamage (GADD45), and apoptosis (Bak and Bax). Calcitriol-induced insulinoma cell death was significantly reduced inp53 null �TC (�TC p53�/�) cells and was completely pre-vented by Bcl-2 overexpression. Finally, exposure of �TC3cells to calcitriol altered the expression of the survival factors(IGF-I and IGF-II).

Materials and MethodsCell lines and cultures

�TC3 were originally provided by Shimon Efrat (Sackler School ofMedicine, Tel Aviv University, Tel Aviv, Israel). �TC3, �TCtet, and�TCtet/bcl2 were grown in RPMI 1640 supplemented with 10% fetal calfserum, 2 mm l-glutamine, and 100 IU/ml streptomycin/penicillin. �TCp53�/� and �TC p53�/� were cultured in DMEM supplemented with20% fetal calf serum, 2 mm glutamine, and 100 IU/ml streptomycin/penicillin. Cultures were performed under standard humidified condi-tions of 5% CO2 and 95% air at 37 C.

Western blot analysis

��C3 cells were seeded at a density of 2 � 106 onto 10-cm2 tissueculture plates and allowed to attach and grow for 48 h. The medium wasthen replaced with fresh medium containing vehicle (ethanol; final con-centration, 0.04%) or increasing calcitriol concentrations (10, 100, and1000 nm). After 20 min and 4, 8, 24, and 48 h, cells were harvested andlysed in 200 �l lysis buffer (30 mm Tris-HCl, 5 mm EGTA, 5 mm EDTA,250 mm sucrose, 1% Triton X-100, 1 mm sodium fluoride, 2 mm sodiumorthovanadate, 1 mm phenylmethylsulfonylfluoride, and 1 �g/ml apro-tinin). After 1 h at 4 C, lysates were centrifuged at 13,000 rpm for 5 min,and the extracted proteins were analyzed by Western blotting using thefollowing antibodies: anti-p53 (rabbit antigoat; 1:100; Santa Cruz Bio-technology, Inc., Santa Cruz, CA), anti-p21 (rabbit antigoat; 1:100; SantaCruz Biotechnology, Inc.), anti-NF�B (p65, rabbit antigoat; 1:100; SantaCruz Biotechnology, Inc.), anti-I�B-� (goat antirabbit; 1:100; Santa CruzBiotechnology, Inc.), and antiactin (goat antirabbit; 1:100; Sigma-AldrichCorp., St. Louis, MO) commercial antibodies.

The levels of phosphorylated p53 protein were measured in �TC3cells after 20-min exposure to calcitriol (1000 nm) in the presence orabsence of MEK inhibitor (Uo126, 2 nm) and PKC inhibitor (staurospor-ine, 5 nm). Extracted proteins were analyzed by immunoblotting withantisera against anti-phospo-p53 Ser15 (rabbit polyclonal antibody;1:1000; CELBIO, New England Biolabs, Inc., Beverly, MA). Western blotbands were quantitatively analyzed by scanning acquisition (ScanJet5300C, Hewlett-Packard Co., Palo Alto, CA) and analyzed using ScionImage software (NIH).

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) assay

The number of viable cells was determined using the MTT methodaccording to the manufacturer’s recommendation (Sigma-AldrichCorp.). The MTT assay was performed on �TC3, �TC (p53�/� andp53�/�), and �TCtet and �TCtet/bcl2 cells seeded at a density of 2 � 104

cells/well onto 96-well culture plates. Cells were allowed to attach andgrow for 48 h in standard medium. Then cells were washed and refedwith fresh medium containing vehicle or 1000 nm calcitriol. For theexperiments with pyrrolidine dithiocarbamate (PDTC), cells wereplated as described above, and before calcitriol addiction, cells werepretreated with PDTC (20, 50, and 100 �m). One hour later, mediawere changed and replaced with RPMI with or without calcitriol andPDTC. The number of viable cells was measured after 48 h as pre-viously described (19).

cDNA expression array

The different patterns of gene expression in �TC3 cultured for 48 hin presence and absence of 1000 nm calcitriol were compared by the AtlasMouse cDNA Expression Array (CLONTECH Laboratories, Inc., PaloAlto, CA) following the customer’s instruction.

Relative quantitative RT-PCR

Calcitriol-induced expression of Bak and Bax genes in �TC3 wasdetermined using the Gene Specific Relative RT-PCR kit (Ambion, Inc.,Austin, TX). The relative expressions of IGF-I and IGF-II were deter-mined by RT-PCR using primers and amplification conditions previ-ously described (20). IGF-II primers amplify common sequences of 4.8and 6.0 transcription isoforms and do not discriminate between the twoisoforms.

Northern blot analysis

The expressions of IGF-I, IGF-II, and GADD45 genes were measuredin �TC3 cells cultured for 48 h in the presence of increasing concentra-tions of calcitriol (10, 100, and 1000 nm). Total cellular RNA was preparedwith the RNAfast RNA isolation system (M-Medical, Firenze, Italy), and20 �g RNA from each sample were electrophoresed on 1% denaturingagarose gel. Blots were sequentially hybridized with a human IGF-IcDNA probe (a gift from Dr. Antonio Torsello, University of Milan,Milan, Italy), a rat IGF-II cDNA probe (provided by Dr. Steen Gam-meltoft, Bispebjerg Hospital, Copenhagen, Denmark), a humanGADD45 cDNA probe (provided by Dr. Michael O’Reilly, University ofRochester, Rochester, NY), and a 18S cDNA probe. Relative expressionlevels of IGF-I, IGF-II, GADD45, and 18S were determined by densito-metric analyses.

Statistical analysis

In vitro studies consisted of a minimum of three independent exper-iments, each carried out at least in duplicate. Datasets were expressedas the mean � se. Statistical analysis was performed using the unpairedt test for pairwise comparisons or one/two-way ANOVA (Tukey post hoctest), as appropriate. Statistical significance was considered at P � 0.05.

ResultsGenes differentially expressed by �TC3 cells cultured in thepresence of calcitriol

Gene array analysis was performed to identify genes dif-ferentially expressed upon treatment of �TC3 cells with cal-citriol. Exposure to 1000 nm calcitriol induced the differentialexpression of a number of genes; some are known to befunctional in the regulation of the cell cycle and apoptosis,and some are survival factors, all consistent with calcitriol’santineoplastic effect on �TC3 cells (Tables 1 and 2). Seven ofthe 588 genes tested were switched on/off by calcitriol,

TABLE 1. Genes up- and down-regulated by 48 h of exposure to1000 nM calcitriol in �TC3 cells

Fold increase/decrease

MAPK p38 1.33 � 0.03BAK1 1.98 � 0.20BAX 1.66 � 0.12G2/M-specific cyclin B2 0.51 � 0.10Transforming protein p21 1.36 � 0.03Thyroid hormone receptor �-1 1.69 � 0.10HMG-14 1.70 � 0.20HSP86 0.77 � 0.05Glutathione S-transferase 5 0.73 � 0.03

Values are the mean � SE; n � 3 independent experiments; P � 0.05for each gene listed in the table.

Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells Endocrinology, May 2003, 144(5):1832–1841 1833

whereas nine of them were statistically up- or down-regu-lated. The proapoptotic genes Bak and Bax were up-regu-lated by 198% and 166% of control values, respectively. Amodest, but significant, increase in the expression of p21mRNA levels was also detected. Moreover, the expression ofGADD45, undetectable in control �TC3, was induced bycalcitriol. N-Cadherin was another gene with up-regulatedexpression upon calcitriol treatment. Conversely, calcitrioldown-regulated the expression of G2/M-specific cyclin B2and IGF-II genes to 50% and 80% of control values, respec-tively, whereas the only gene that was silenced by calcitriolwas angiogenin. Down-regulation of this potent angiogenicfactor may contribute to the antineoplastic activity of calcit-riol observed during insulinoma development in RIP1Tag2mice in vivo (8).

Calcitriol increases p53 and phosphorylated (Ser15) p53protein levels in �TC3

We have previously shown that the MAPK pathway isnongenomically activated by calcitriol in �TC3 (8). Becauseof the well established link between p53 signaling and theMAPK cascade (21), we investigated whether the p53 path-way was involved in the antineoplastic effect of calcitriol oninsulinoma cells. After 48 h of culture in the presence ofincreasing calcitriol concentrations, whole cell lysates werecollected, and p53 levels were assessed by Western blot. Asshown in Fig. 1A, p53 protein doubled after exposure to 100nm calcitriol. Increasing the calcitriol concentration to 1000mm did not further increase p53 protein levels. At 10 nm,calcitriol induced only a modest, nonsignificant increase inp53 protein.

Calcitriol also induced an increase in the levels of p53phosphorylated at Ser15 to about 200% of control values by20 min after treatment with 1000 mm calcitriol (Fig. 1B). Thiseffect was prevented by the PKC inhibitor staurosporine, butnot by the MEK inhibitor Uo126 (Fig. 1B).

�TC p53�/� cells are less susceptible to calcitriol-inducedcytotoxicity

To further investigate the involvement of p53 in calcitriol-induced insulinoma cell death, we used p53-null �TC cells(�TC p53�/�), which have been derived from tumors inRIP1Tag2/p53�/� mice (22). Forty-eight hours of culture in1000 nm calcitriol induced a significant decrease in the num-ber of viable control �TC p53�/� cells, quantitatively similarto what was observed in �TC3 cells. In contrast, �TC p53�/�

cells were partially resistant to the cytotoxic effects of cal-

FIG. 1. Calcitriol effects on p53 protein levels in �TC3, �TC p53�/�,and �TC p53�/� cells. A, Culture in presence of calcitriol induced adose-dependent increase in p53 protein levels in �TC3 cells. p53 in-creased, albeit not significantly, at 10 nM calcitriol and increasedfurther at 100 nM calcitriol. At 100 nM calcitriol exerted its maximaleffect (*, P � 0.05; #, P � 0.01 vs. control; n � 3 experiments, eachperformed in triplicate). B, Calcitriol (1000 nM) induced an increasein p53 phosphorylation at the level of serine 15 after 20 min of ex-posure (*, P � 0.05; #, P � 0.01 vs. control; n � 5 experiments, eachperformed in duplicate). Phosphorylation of p53 was efficaciouslyprevented by the PKC inhibitor staurosporine, but not by the MEKinhibitor Uo126. C, Forty-eight hours of calcitriol exposure decreased�TC3 and �TC p53�/� cell viability by about 30% and 40%, respec-tively. In contrast, in �TC p53�/� cells calcitriol decreased cell via-bility only modestly (*, P � 0.05 vs. control; **, P � 0.05 vs. controland vs. �TC p53�/� exposed to calcitriol; n � 6 experiments, eachperformed 16 times). Data are the mean � SE; significance was de-termined by one-way ANOVA, post hoc Tukey test.

TABLE 2. Genes turned on and off by 48 h of exposure to 1000nM calcitriol in �TC3 cells

Calcitriol effect

GADD45 OnSRY-box containing gene 4 OnTranscription factor UBF OnIRS-1 OnACE OnAngiogenin OffN-cadherin On

n � 3 independent experiments.

1834 Endocrinology, May 2003, 144(5):1832–1841 Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells

citriol, exhibiting only a modest decrease in the number ofviable cells (Fig. 1C).

Calcitriol increases p21 protein levels in �TC3

Antiproliferative signals, including serum deprivation orDNA damage, result in transcriptional activation of p21, awell known inhibitor of cyclin-dependent kinase (cdk2). p21,a bona fide transcriptional target gene of p53, was significantlyup-regulated by calcitriol, as demonstrated by the gene array(Table 1) and Western blots (Fig. 2A). Calcitriol increased p21protein in �TC3 cells in a dose-dependent manner to 160%,

180%, and 210% of control values after culture in 10, 100, and1000 nm calcitriol, respectively (Fig. 2A).

Calcitriol increases NF�B protein levels

It has been previously shown that induction of p53 causesan activation of NF�B that correlates with the ability of p53to induce apoptosis (23). Therefore, we determined the pro-tein levels of NF�B in �TC3 cells after 48 h of culture in thepresence of increasing concentrations of calcitriol. NF�B pro-tein levels increased to 200% of control values at 10 nmcalcitriol, a level that did not significantly change with higherconcentrations of calcitriol (Fig. 2B).

Time course of the expression levels of p53, NF�B, I�B, andp21 protein upon calcitriol exposure

The protein levels of p53, NF�B, I�B and p21 were mea-sured 20 min and 4, 8, 24, and 48 h after exposure to 100 nmcalcitriol and were compared with control levels. As shownin Fig. 3A, the increase in p53 protein levels was significantafter 8 h and then remained stable for the entire period ofstimulation. Conversely, NF�B protein levels showed a bi-phasic pattern of activation with two peaks, at 8 and 48 h, thatwere separated by a decrease to control levels at 24 h. Asimilar biphasic pattern of NF�B activation was recentlyreported (24). Interestingly, I�B protein levels showed theopposite pattern, with a decrease at 8 h (�70% of control) anda peak at 24 h (�200% of control; Fig. 3C). Finally, p21showed a significant increase in protein levels only after 48 h,suggesting that the transcriptional regulation exerted by cal-citriol on this gene is exclusively mediated by its classicalgenomic effect (Fig. 3D).

NF�B inhibition with PDTC fails to protect �TC3 cells fromcalcitriol-induced apoptosis

To test whether NF�B activation was involved in calcitriol-induced apoptosis, treatment with calcitriol was performed inthe presence or absence of PDTC (25, 26), an inhibitor of NF�BDNA-binding activity (Fig. 4A). PDTC alone was toxic for �TC3cells at the lowest concentration known to have a biologicaleffect (20 �m). Moreover, when calcitriol and PDTC were addedtogether, the number of viable cells decreased further, indicat-ing an additive effect of calcitriol and PDTC on �TC3 cells.

Calcitriol does not increase NF�B protein levels in �TCp53�/� cells

To better understand the relationship between p53 andNF�B, we measured NF�B protein levels in �TC p53�/�

and p53�/� cells, which, as previously shown, are par-tially resistant to calcitriol-induced apoptosis. As shown inFig. 4B, 8 and 48 h after calcitriol exposure, NF�B proteinlevels were increased in �TC3 and �TC p53�/� cells, butnot in p53�/� cells.

Calcitriol increases GADD45 gene expression

DNA damage and environmental stress activate GADD45,a p53-regulated gene that is thought to play a role in growtharrest and cell death. Gene array analysis showed that cal-citriol switched on GADD45 expression in �TC3 cells (Table

FIG. 2. Effects of calcitriol on p21 and NF�B protein levels. A, In�TC3 cells, 48 h of exposure to increasing concentrations of calcitriol(10, 100, and 1000 nM) induced an increase in p21 protein levels,which were maximal already at 10 nM (*, P � 0.01; #, P � 0.05 vs.control, by one-way ANOVA, post hoc Tukey test; mean � SE; n � 4experiments, each performed in triplicate). B, In �TC3 cells, calcitriolsignificantly increased NF�B protein levels at the lowest concentra-tion (10 nM) that was also maximally effective. After 48 h of culture,10 and 100 nM calcitriol increased NF�B protein to 215% and 250%,respectively (*, P � 0.02 vs. control, by one-way ANOVA, post hocTukey test; mean � SE; n � 4 experiments, each performed in trip-licate).

Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells Endocrinology, May 2003, 144(5):1832–1841 1835

2). Even though a low level of expression of GADD 45 wasdetectable by Northern blot analysis, the inducible effect ofcalcitriol on GADD45 expression (Fig. 5) was confirmed.After 48 h of treatment, GADD45 mRNA increased to 206%,

236%, and 217% in the presence of 10, 100, and 1000 nmcalcitriol.

Calcitriol up-regulates Bak and Bax gene expression in�TC3 cells, whereas Bcl-2 overexpression completelyprevents calcitriol-induced cytotoxicity

Gene array analysis revealed an increased expression ofthe proapoptotic genes Bax and Bak (Table 1). We performedrelative quantitative multiplex RT-PCR to confirm these re-sults. We used two primer sets in a single PCR reaction: oneset to amplify the cDNA of interest (Bak and Bax) and asecond to amplify an invariant endogenous control (18S).These experiments revealed that after 48 h of culture in thepresence of increasing calcitriol concentrations Bak and Baxexpression were significantly increased in �TC3 cells, with amaximal effect achieved at the lowest (10 nm) calcitriol con-centration (Fig. 6, A and B).

The �TCtet insulinoma cell line (27) was also sensitive tocalcitriol-induced cytotoxicity; however, forced expressionof Bcl-2 in �TCtet/Bcl-2 cells (28) completely abolished thiseffect (Fig. 6C). These results suggest that calcitriol-inducedapoptosis of tumor �-cells is modulated by members of theBcl-2 family of genes.

Calcitriol induces IGF-I, but down-regulates IGF-II,gene expression

It has been shown that the IGF-I/IGF-II signaling pathwaycontributes to the antineoplastic effect of vitamin D in avariety of cancer cell lines (29–31). Therefore, we studied theexpression of IGF-I and IGF-II by RT-PCR and Northernblotting in �TC3 cells cultured for 48 h in the presence ofincreasing calcitriol concentrations. In �TC3 cells, calcitriolinduced the expression of IGF-I mRNA, which was com-pletely silenced in control condition (Fig. 7, A and B), aneffect that was already evident at 10 nm. In contrast, calcitriolmodestly, yet significantly, down-regulated the expressionof two IGF-II mRNA isoforms (Fig. 7D), thereby confirmingthe data obtained by RT-PCR (Fig. 7, A and C).

Discussion

Besides its central role in the maintenance of calcium ho-meostasis and bone mineralization, calcitriol is an importantmodulator of cell growth and differentiation in a variety ofcell types. We previously described a wide spectrum of neg-ative effects exerted by calcitriol on mouse insulinoma �TC3cells, ranging from the induction of growth arrest and apo-ptosis of cultured tumor �-cell lines in vitro to the reductionof �-cell tumors in vivo in RIP1Tag2 transgenic mice (8).

To learn more about the mechanisms responsible for theantineoplastic effects of calcitriol on insulinoma cells, we setout to identify genes that are differentially expressed by �TC3cells upon treatment with calcitriol. Calcitriol induced thedifferential expression of a number of genes known to beimportant in the regulation of the cell cycle. �TC3 cells ex-posed to calcitriol exhibited a 50% decrease in the expressionof the cyclin B2 gene, which is a central regulator of theprogression from the G2 phase to mitosis (32). Another cellcycle gene differentially expressed by �TC3 exposed to cal-

FIG. 3. Time course of p53, NF�B, I�B, and p21 protein levels. Anal-ysis were performed in �TC3 cells after 20 min and 4, 8, 24, and 48 hof exposure to 100 nM calcitriol. A, p53 protein levels increased sig-nificantly after 8 h of stimulation and reached maximum levels after24 h. B, NF�B showed a biphasic pattern of induction, with a peak at8 h, a decrease to control values at 24 h, and a second peak at 48 h.C, I�B protein levels decreased at 8 h and increased thereafter, reach-ing a peak after 24 h. D, Finally, p21 protein levels increased at 24and 48 h. The data are expressed as a percentage of the control andare the mean � SE of three independent experiments, each performedin duplicate (*, P � 0.05; **, P � 0.02; ***, P � 0.01 vs. control, byone-way ANOVA, post hoc Tukey test).

1836 Endocrinology, May 2003, 144(5):1832–1841 Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells

citriol was p21, whose mRNA expression was modestly, yetsignificantly, increased by calcitriol. Calcitriol-mediated ex-pression of p21 was confirmed by Western blotting, revealinga dose-dependent increase in p21 protein in response toincreasing calcitriol concentrations. Interestingly, time-course experiments showed that p21 protein levels increased24 h after calcitriol exposure, suggesting that this particulareffect of calcitriol is mediated by the classic genomic path-way. p21 is an inhibitor of cdk and inhibits the formation ofcyclin-cdk complexes necessary for transition from the G1 tothe S phase of the cell cycle (33). Together, these findings areconsistent with the previously reported growth inhibitoryeffect of calcitriol on �TC3 cells, characterized by an in-creased number of �TC3 cells found in the G0/G1 fraction anda contemporary decrease in those in S and G2/M phases (8).

p53 is the most important tumor suppressor protein iden-tified to date and can influence the cell cycle in several ways.

It can cause growth arrest and apoptosis by forming com-plexes with other proteins, and it also acts as a transcriptionfactor (34–36). Noteworthy, both cyclin B2 and p21 are tran-scriptional targets of p53 (33, 34). Simian virus 40 large Tantigen, a potent oncoprotein present in �TC3 and �TC cells,exerts its oncogenic effect in part through binding to andinactivating the tumor suppressor gene products p53 andretinoblastoma (37). In �TC3, calcitriol induced a 2-fold in-crease in p53 protein levels, an effect that was already max-imal at 100 nm. Time-course experiments showed that at thisconcentration of calcitriol the increase in p53 protein levelswas already evident after 8 h of stimulation, even though themaximal effect occurred at 48 h. It has been shown that inresponse to DNA damage, p53 protein levels and activityincrease mainly as a result of its phosphorylation (35), whichcan be induced by DNA damage-sensing kinases as well asby MAPKs (39, 40). Therefore, we explored whether calcitriol

FIG. 4. Role of NF�B in calcitriol-induced apoptosis of �TC3 cells.A, The NF�B inhibitor PDTC de-creased �TC3 cell viability even atthe lower concentration known toexert a biological effect (20 �M)and exerted an additive effecton calcitriol-induced decrease in�TC3 cell viability (*, P � 0.05; **,P � 0.01 vs. control; #, P � 0.02vs. 100 nM calcitriol and PDTC,by one-way ANOVA, post hocTukey test; mean � SE; n � 3,experiments, each performedin quadruplicate). B, At 8 and48 h of stimulation, calcitriol-induced increase in NF�B pro-tein levels was completely ab-sent in �TC p53�/�, whereas itwas present in �TC3 and �TCp53�/� cells. Data are themean � SE (n � 3 experiments,each performed in duplicate; *,P � 0.02 vs. control, by one-wayANOVA, post hoc Tukey test).

Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells Endocrinology, May 2003, 144(5):1832–1841 1837

would induce p53 phosphorylation in insulinoma �TC3 cells.We decided to study phosphorylation at serine 15, becausethis site of phosphorylation prevents the interaction of p53with Mdm2, a protein that can down-regulate p53 via ubiq-uitin-mediated proteolysis. After 20 min of calcitriol treat-ment, �TC3 cells showed a 2-fold increase in phospho-p53protein levels. This effect was prevented by staurosporine,but not by UO126, suggesting that MAPKs are not involvedin this specific phosphorylation, whereas PKC may be anupstream activator of p53 kinases. Taken together, these datasuggest that calcitriol, by increasing the p53 level and activityby combining nongenomic and genomic effects, may reducethe oncogenic potential of T antigen in �TC3 and �TC insu-linoma cells.

Previous reports indicated that �TC cell lines derived fromtumors of RIP1Tag2/p53-null mice have an apoptotic inci-dence comparable with normal �TC lines and suggested thatp53-independent apoptotic pathways are used in �TC cells(41, 22). To learn more about the role of p53 in calcitriol-induced insulinoma cell death, we performed a set of ex-periments with �TC p53-null cells (�TC p53�/�). In �TCp53�/� cells, calcitriol induced only a modest decrease in thenumber of viable cells (�10% reduction), significantly lowerthan that observed in control �TC p53�/� cells (�40% re-duction). Therefore, in contrast to the majority of classicchemotherapeutic agents (22), calcitriol needs a functionalp53 to fully exert its antineoplastic activity in insulinomacells. Similar results were observed in glioma cells (42), butnot in breast and colon cancer cells, where the antineoplastic

FIG. 5. Effects of calcitriol exposure on GADD45 gene expression. In�TC3 cells, 48 h of exposure to calcitriol significantly increased themRNA levels of DNA damage-inducible GADD45 gene. The calcitrioleffect was maximal at a concentration of 10 nM (#, P � 0.01 vs. control,by one-way ANOVA, post hoc Tukey test; mean � SE; n � 3 experi-ments, each performed in triplicate).

FIG. 6. Role of the Bcl-2 family in calcitriol-induced insulinoma celldeath, measured after 48 h of calcitriol exposure. A and B, RT-PCRanalysis showed that after 48 h of exposure, calcitriol increased theexpression of Bak and Bax already at the lowest (10 nM) concentrationthat was also maximally effective (A and B; *, P � 0.01 vs. control, byone-way ANOVA, post hoc Tukey test; mean � SE; n � 3 experiements,each performed in duplicate). C, As in �TC3 cells, calcitriol induceda significant decrease in the number of viable �TC-tet cells. In con-trast, lentivirus-mediated Bcl-2 overexpression in �TC-tet cells com-pletely abolished calcitriol-induced cytotoxicity in the resultant �TC-tet/Bcl-2 cell line (*, P � 0.05 vs. control, by one-way ANOVA, post hocTukey test; n � 6 experiments, each performed 16 times). Data aremean � SE.

1838 Endocrinology, May 2003, 144(5):1832–1841 Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells

effects of calcitriol and its analogs do not require the functionof p53 (43, 44).

The role of NF�B in programmed cell death and cell cycleis still controversial. Reportedly, the proapoptotic or anti-apoptotic effect of NF�B depends on the different cell typesand the external stimuli applied (45). In the resting cell, NF�Bis sequestered in the cytoplasm linked to its inhibitor protein,I�B. Upon activation, free NF�B increases as a result of I�Bphosphorylation and degradation. Free NF�B translocates tothe nucleus, binds its consensus sequences, and regulates thetranscription of a variety of genes. We show here that cal-citriol induces in �TC3 cells a biphasic increase in NF�Bprotein, similar to what occurs in skeletal muscle cells (24).The first peak was observed after 8 h of stimulation and wasassociated with a significant decrease in I�B protein levels,which, as previously reported (45), might be consequent toI�B phosphorylation by MAPKs. Perhaps a similar series ofevents may occur in �TC3 cells, as calcitriol can induce ac-tivation of the MAPK pathway (8). A second increase in

NF�B protein levels was detected at 48 h and was associatedwith a decrease in I�B.

It has been previously reported that induction of p53causes an activation of NF�B, which correlates with the abil-ity of p53 to induce apoptosis (23). To exclude such possi-bility, we studied the pattern of NF�B activation in �TCp53�/� cells and the effect of NF�B inhibition with PDTC. Asshown in Fig. 4B, calcitriol-induced increase in NF�B wascompletely absent in �TC p53�/� cells at both 8 and 48 h.Moreover, as shown in other cell lines (46, 47), inhibition ofNF�B decreased �TC3 cell viability and amplified calcitriol-induced cytotoxicity. Taken together these data suggest thatNF�B up-regulation is an antiapoptotic reaction mounted bythese cells against calcitriol-induced cell death.

As previously reported in glioma cells (42), calcitriol up-regulated the expression of GADD45, another importantp53-dependent regulator of cell fate (48), in �TC3 cells.

Calcitriol also increased the expression of Bak and Bax,two proapoptotic members of the Bcl-2 gene family (49).

FIG. 7. Effects of calcitriol on IGF-I andIGF-II expression in �TC3 cells. A andB, After 48 h of culture in the presenceof calcitriol, IGF-I mRNA, undetectableby RT-PCR in control cells, wasswitched on by calcitriol already at thelowest tested concentration (10 nM).The induction of IGF-I gene expressionwas evident by RT-PCR (A) and North-ern blot (B) experiments. IGF-II mRNAlevels were reduced upon calcitriol ex-posure, as shown by in RT-PCR (A andC). Northern blot experiments (B andD) showed a parallel reduction of bothIGF-II isoforms (4.8- and 6.0-kb tran-scripts). This effect was significant at100 nM and increased slightly at 1000nM calcitriol (*, P � 0.05; #, P � 0.01 vs.control, by one-way ANOVA, post hocTukey test). Data are the mean � SE(n � 3 experiments, each performed induplicate).

Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells Endocrinology, May 2003, 144(5):1832–1841 1839

This effect is consistent with the reported capability of p53to transcriptionally activate Bax (50). The involvement ofproapoptotic Bak and Bax may contribute to calcitriol-induced �TC3 cell death by causing mitochondrial mem-brane damage. Overexpression of the antiapoptotic Bcl-2gene completely prevented calcitriol-induced apoptosis.Hence, although multiple pathways may be involved incalcitriol-induced �TC3 cell apoptosis, they can be mod-ulated by the function of Bcl-2, which is known to play apivotal role in the regulation of cell death by acting at themitochondrial and pre- and postmitochondrial levels (51).It is noteworthy that Bcl-2 overexpression in breast cancercells conferred complete protection against apoptosis in-duced by vitamin D compounds, which in these cells is ap53-independent event (43).

IGF-I and IGF-II are the most abundant growth factorsin the body and act as potent survival factors. Exposure of�TC3 to calcitriol induced the expression of IGF-I mRNA,that is undetectable in control cells, as shown by RT-PCRand Northern blotting. The induction of IGF-I expressionwas associated with a modest, but significant, down-regulation of the IGF-II gene in both its isoforms. Calcitriol-mediated IGF-II down-regulation is consistent with itsantineoplastic effects. Similar results have been reportedin prostate cancer cells, where vitamin D-mediated growthinhibition was associated with increased levels of IGF-binding protein-3 mRNA (52, 53). Calcitriol-inducedIGF-II down-regulation may be particularly relevantin vivo in the prevention of insulinomas in transgenicRIP1Tag2 mice. In these mice, which represent a wellcharacterized model of multistage �-cell tumorigenesis, allthe islets of Langerhans express simian virus 40 T antigen,yet only half of them become hyperplastic, and only aminority of these eventually progress to solid �-cell tu-mors (9, 54). Notably, in these mice the onset of prolifer-ation strictly correlates with the expression of IGF-II in the�-cells (55). Perhaps, a precocious treatment of RIP1Tag2mice with calcitriol may prevent this proliferative switchand thus the development of �-cell tumors.

Calcitriol-induced IGF-I expression is less compatiblewith the antineoplastic effect of calcitriol on �TC3 cells.Nevertheless, calcitriol-induced IGF-I up-regulation inpancreatic �-cells is particularly appealing and may berelevant to the reported beneficial effect of calcitriol on thepathogenesis of type 1 diabetes. It has been shown thattreatment with calcitriol prevents diabetes in the NODmouse (56) as well as the recurrence of autoimmune de-struction of syngeneic islet grafts (57). These effects havebeen related to the known immunomodulatory propertiesof vitamin D3 on dendritic cells (58). However, our ob-servation that calcitriol induces IGF-I up-regulation intumor �-cells and the recent finding that �-cell-targetedIGF-I overexpression counteracts insulitis-associated type 1 dia-betes (59) suggest that calcitriol might exert its antidiabetic effectalso by decreasing �-cell susceptibility to proinflammatorycytokines via IGF-I up-regulation.

In conclusion, calcitriol exerts a profound antineoplasticeffect on mouse insulinoma cells, which is mediated by theabnormal expression of a series of genes involved in thecontrol of cell cycle and cell death. The antineoplastic

effect of calcitriol on insulinoma cells appears to dependmainly on the p53 pathway and to involve both thegenomic and nongenomic pathways. These data supportthe rationale for testing calcitriol or its lower calcemicanalogs in the treatment of patients with malignant insu-linomas and possibly other p53 function-retaining endo-crine tumors. Moreover, calcitriol-induced IGF-I over-expression, if confirmed to occur also in normal �-cells,may provide an additional rationale for using vitamin Dcompounds in the prevention of autoimmune diabetes andto promote the engraftment and survival of pancreatic isletallografts.

Acknowledgments

We thank Dr. Antonio Torsello (Department of Pharmacology, Uni-versity of Milan, Milan, Italy) for the human IGF-I cDNA probe, Dr.Steen Gammeltoft (Department of Clinical Chemistry, Bispebjerg Hos-pital, Copenhagen, Denmark) for the rat IGF-II cDNA probe, and Dr.Michael O’Relly (University of Rochester, Rochester, NY) for theGADD45 cDNA probe.

Received September 30, 2002. Accepted January 28, 2003.Address all correspondence and requests for reprints to: Dr. Alberto

M. Davalli, Division of General Medicine, Unit of Endocrinology andMetabolic Disease, San Raffaele Scientific Institute, Via Olgettina 60,20132 Milan, Italy. E-mail: alberto.davalli@hsr.it.

This work was supported by a grant from Ministry of Health of Italy(Projects RF98.52 and RF98.50; to F.G. and L.P.).

References

1. Campbell MJ, Koeffler HP 1997 Toward therapeutic intervention of cancer byvitamin D compounds. J Natl Cancer Inst 89:182–185

2. Ruszniewski P, Malka D 2000 Hepatic arterial chemoembolization in themanagement of advanced digestive endocrine tumors. Digestion 62(Suppl1):79–83

3. Arnold R, Simon B, Wied M 2000 Treatment of neuroendocrine GEP tumorswith somatostatin analogues: a review. Digestion 62(Suppl 1):84–91

4. Schott M, Scherbaum WA, Feldkamp J 2000 Drug therapy of endocrineneoplasms: Part II: malignant gastrinomas, insulinomas, glucagonomas, car-cinoids and other tumors. Med Klin 95:81–84

5. Ulbrecht JS, Schmeltz R, Aarons JH, Green DA 1986 Insulinoma in 94-year-old woman: long-term therapy with verapamil. Diabetes Care 9:186–188

6. Imanaka S, Matsuda S, Ito K, Matsuoka T, Okada Y 1986 Medical treatmentfor inoperable insulinoma: clinical usefulness of diphenylhydantoin and dil-tiazem. Jpn J Clin Oncol 16:65–71

7. Stehouwer CD, Lems WF, Fischer HR, Hackeng WH 1989 Malignant insu-linoma: is combined treatment with verapamil and the long-acting somatosta-tin analogue octreotide (SMS 201–995) more effective than single therapy witheither drug? Neth J Med 35:86–94

8. Galbiati F, Polastri L, Grogori S, Freschi M, Casorati M, Cavallaro U, FiorinaP, Bertuzzi F, Zerbi A, Pozza G, Adorini L, Folli F, Christofori G, Davalli AM2002 Antitumorigenic and antiinsulinogenic effects of calcitriol on insulinomacells and solid � cell tumors. Endocrinology 143:4018–4030

9. Hanahan D 1985 Heritable formation of pancreatic � cell tumours in transgenicmice expressing recombinant insulin/simian virus 40 oncogenes. Nature 315:115–122

10. Ishida H, Norman AW 1988 Demonstration of high affinity receptor for 1,25-dihydroxyvitamin D3 in rat pancreas. Mol Cell Endocrinol 60:109–117

11. Gill RK, Chistakos S 1993 Identification of sequence elements in mousecalbindin D28k gene that confer 1,25-dihydroxyvitamin D3 and butyrate in-ducible responses. Proc Natl Acad Sci USA 90:2984–2988

12. Liu M, Lee MH, Cohen, M, Bommakanti M, Freedman LP 1996 Transcrip-tional activation of the CDK inhibitor p21 by vitamin D3 leads to the induceddifferentiation of the myelomonocytic cell line U937. Genes Dev 10:142–153

13. Wu Y, Craig TA, Lutz WH, Kumar R 1999 Identification of 1�,25-dihydroxy-vitamin D3 response elements in the human transforming growth factor �2gene. Biochemistry 38:2654–2660

14. Nemere Z, Schwartz Z, Pedrozo H, Sylvia VL, Dean DD, Boyan BD 1998Identification of a membrane receptor for 1�,25-dihydroxyvitamin D3 whichmediates rapid activation of protein kinase C. J Bone Miner Res 13:1353–1359

15. Nemere I, Darmanen MC, Hammand MW, Okamura WH, Norman AW 1994Identification of a specific binding protein for 1�,25-dihydroxyvitamin D3 in

1840 Endocrinology, May 2003, 144(5):1832–1841 Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells

basal-lateral membranes of chick intestinal epithelium and relationship totranscaltochia. J Biol Chem 269:23750–23756

16. Capiati DA, Vazquez G, Inon MTT, Boland RL 2000 Role of protein kinaseC in 1, 25(OH)2-vitamin-D3 modulation of intracellular calcium during de-velopment of skeletal muscle cells in culture. J Cell Biochem 77:200–212

17. Hmama Z, Nandan D, Sly L, Knutson KL, Henera-Velit P, Reiner NE 19991�,25-Dihydroxyvitamin D3-induced myeloid cell differentiation is regulatedby a vitamin D receptor-phoshatidylinositol 3-kinase signaling complex. J ExpMed 190:1583–1594

18. Baran DT 1994 Non genomic actions of the steroid hormone 1�,25-dihy-droxyvitamin D3. J Cell Biochem 56:303–306

19. Mosmann T 1983 Rapid colorimetric assay for cellular growth and survival:application to proliferation and cytotoxicity assays. J Immunol Methods 65:55–63

20. Kocamis H, Gahr Scott A, Batelli L, Hubbs AF, Killefer J 2002 IGF-I, IGF-IIand IGF-receptor I and IGF-II protein expression in myostatin knockout micetissues. Muscle Nerve 26:55–63

21. Lee SW, Fang L, Igarashi M, Ouchi T, Lu KP, Aaranson SA 2000 Sustainedactivation of Ras/Raf/mitogen-activated protein kinase cascade by the tumorsuppressor p53. Proc Natl Acad Sci USA 97:8302–8305

22. Herzig M, Novatchkova M, Christofori G 1999 An unexpected role for p53in augmenting SV40 large T antigen-mediated tumorigenesis. J Biol Chem380:203–211

23. Ryan KM, Ernst MK, Rice NR, Vousden KH 2000 Role of NF-�B in p53-mediated programmed cell death. Nature 404:892–896

24. Ladner KJ, Caligiuri MA, Guttridge DC 2003 TNF regulated biphasic acti-vation of NF�B is required for cytokine-induced loss of skeletal muscle geneproducts. J Biol Chem 278:2294–2303

25. Schrek R, Meier B, Mannel DN, Droge W, Baeuerle PA 1992 Dithiocarbam-ates as potent inhibitors of nuclear factor �B activation in intact cells. J Exp Med175:1181–1194

26. Rangan G, Wang Y, Tay YC, Harris D 1999 Inhibition of NF�B activation withantioxidants is correlated cytokine transcription in PTC. Am J Physiol 277:F779–F789

27. Efrat S, Fusco-DeMane D, Lemberg H, Al Emran O, Wang X 1995 Conditionaltransformation of a pancreatic � cell line derived from transgenic mice ex-pressing a tetracycline-regulated oncogene. Proc Natl Acad Sci USA 92:3576–3580

28. Dupraz P, Rinsch C, Pralong WF, Rolland E, Zufferey R, Trono D, ThorensB 1999 Lentivirus-mediated Bcl-2 expression in �TC-tet cells improves resis-tance to hypoxia and cytokine-induced apoptosis while preserving in vitro andin vivo control of insulin secretion. Gene Ther 6:1160–1169

29. Xie SP, Pirianov G, Colston KW 1999 Vitamin D analogues suppress IGF-Isignaling and promote apoptosis in breast cancer cells. Eur J Cancer 35:1717–1723

30. Colston KW, Perks CM, Xie SP, Hally JMP 1998 Growth inhibiting of bothMCF-7 and Hs578T human breast cancer cell lines by vitamin D analogues isassociated with increased expression of insulin-like growth factor bindingprotein-3. J Endocrinol 20:157–162

31. Razen F, Yong XF, Huynh H, Pollak M 1997 Antiproliferative action of vitaminD-related compounds and insulin-like growth factor binding protein 5 accu-mulation. J Natl Cancer Inst 89:652–656

32. Gautier J, Minshull J, Lalika M, Glotzer M, Maller JL 1990 Cyclin is acomponent of maturation-promoting factor from Xenopus. Cell 60:487–494

33. El-Deiry WS, Tokino T, Velculescu VE, Leny DB, Parsans R, Trent JM, LinD, Mircer WE, Kinzler KW, Vogelstein B 1993 WAF 1, a particular mediatorof p53 tumor suppression. Cell 75:817–825

34. Krause, K, Wasner M, Reinhard W, Haugwitz V, Dahna CL, Mossner J,Engeland K 2000 The tumor suppressor protein p53 can repress transcriptionof cyclin B. Nucleic Acids Res 28:4410–4418

35. Levine AJ 1997 p53 the cellular gatekeeper for growth and division. Cell88:323–331

36. Ko LJ, Prives C 1996 P53 puzzle and paradigm. Genes Dev 10:1054–1072

37. Ludlow JW 1993 Interactions between SV40 large-tumor antigen and thegrowth suppressor proteins pRB and p53. FASEB J 7:866–871

38. Shieh SY, Ikeda M, Taya Y, Prives C 1997 DNA-damage-induced phosphor-ylation of p53 alleviates inhibition by MDM2. Cell 91:325–334

39. Huang C, Mai WY, Maxiner A, Sun Y, Dang Z 1999 p38 kinase mediatesUV-induced phosphorylation p53 protein at serine 389. J Biol Chem 276:12229–12235

40. Fuchs SY, Adler V, Pincus MR, Ronai Z 1998 MEKK1/JNK signaling stabilizesand activates p53. Proc Natl Acad Sci USA 18:10541–10546

41. Lamm GL, Christofori G 1998 Impairment of survival factor function poten-tiates chemotherapy-induced apoptosis in tumor cells. Cancer Res 58:801–807

42. Baudet C, Chevalier G, Chassevent A, Canova C, Filmon R, Larra F, BrachetP, Wion D 1996 1,25-Dihydroxyvitamin D3 induces programmed cell death ina rat glioma cell line. J Neurosci Res 46:540–550

43. Mathiasen IS, Lademann U, Jaattela M 1999 Apoptosis induced by vitaminD compounds in breast cancer cells is inhibited by Bcl-2 but does not involveknown caspases or p53. Cancer Res 59:4848–4856

44. Diaz GD, Paraskeva C, Thomas MG, Binderup L, Hague A 2000 Apoptosisis induced by the active metabolite of vitamin D3 and its analog EB1089 incolorectal adenoma and carcinoma cells: possible implications for preventionand therapy. Cancer Res 60:2304–2312

45. Chen F, Catranova F, Shi X 2001 New insight into the role of nuclear factor-�Bin cell growth regulation. Am J Pathol 159:387–397

46. Smirnov AS, Ruzov AS, Budanov AV, Prokhortchouk AV, Ivanov AV,Prokhortchouk EB 2001 High constitutive level of NF�B is crucial for viabilityof adenocarcinoma cells. Cell Death Differ 8:621–630

47. Gunawardena K, Murray DK, Swope RE, Meikle AW 2002 Inhibition ofnuclear factor �-B induced apoptosis following treatment with tumor necrosisfactor � and an antioxidant in human prostate cancer cells. Cancer Detect Prev26:229–237

48. O’Reilly MA, Staversky RJ, Watkins RH, Maniscalco WM, Keng PC 2000P53-independent induction of GADD 45 and GADD 153 in mouse lungsexposed to hyperoxia. Am J Physiol 278:L552–L559

49. Harris MH, Thompson CB 2000 The role of the Bcl-2 family in the regulationof outer mitochondrial membrane permeability. Cell Death Differ 7:1182–1191

50. Kaelin Jr WG 1999 Choosing anticancer drug targets in the postgenomic era.J Cancer Inst 104:1503–1506

51. Liu X, Bhalla K, Naekyung Kim C, Ibrado AM, Cai J, Deng TI, Jones DP,Wong X 1997 Prevention of apoptosis by Bcl-2: release of cytochrome c frommitochondrial. Science 275:1129–1132

52. Huynh H, Pollak M, Zhang JC 1998 Regulation of insulin-like growth factor(IGF) II and IGF binding protein 3 autocrine loop in PC-3 prostate cancer cellsby vitamin D metabolite 1, 25(OH)2D3 and its analog EB 1089. Int J Oncol13:137–143

53. Nickerson T, Huynh H 1999 Vitamin D analogue EB 1089-induced prostateregression is associated with increased gene expression of insulin-like growthfactor binding proteins. J Endocrinol 160:223–229

54. Tietelman G, Alpert S, Hanahan D 1988 Proliferation, senescence, and neo-plastic progression of b cells in hyperplastic pancreatic islets. Cell 52:97–105

55. Christofori G, Naik P, Hanahan D 1994 A second signal supplied by insulin-like growth factor II in oncogene-induced tumorigenesis. Nature 369:414–418

56. Mathieu C, Waer M, Laureys J, Rutgeerts O, Bouillon R 1994 Prevention ofautoimmune diabetes in NOD mice by 1,25-dihydroxyvitamin D3. Diabeto-logia 37:552–558

57. Casteels K, Waer M, Laureys J, Valckx D, Depovere J, Bouillon R, MathieuC 1998 Prevention of autoimmune desctruction of syngeneic islet grafts inspontaneously diabetic nonobese diabetic mice by combination of vitamin D3analog and cyclosporine. Transplantation 65:1225–1232

58. Casteels K, Bouillon R, Waer M, Mathieu C 1995 Immunomodulatory effectsof 1,25-dihydroxyvitamin D3. Curr Opin Nephrol Hypertens 4:313–318

59. George M, Ayuso E, Casellas A, Costa C, Christophe JD, Bosh F 2002 B cellexpression of IGF-I leads to recovery from type 1 diabetes. J Clin Invest109:1153–1163

Galbiati et al. • Calcitriol Effects on Tumorigenic �-Cells Endocrinology, May 2003, 144(5):1832–1841 1841