at SciVerse ScienceDirect

Clinical Nutrition 33 (2014) 136e142

Contents lists available

Clinical Nutrition

journal homepage: http: / /www.elsevier .com/locate/c lnu

Original article

Calcitriol supplementation effects on Ki67 expression andtranscriptional profile of breast cancer specimens frompost-menopausal patientsq

Yuri Nagamine Urata a,e, Eduardo Carneiro de Lyra a,b,e, Maria Lucia Hirata Katayama a,Ricardo Alves Basso b, Paulo Eduardo Zuccolotto de Assis b, Ana Paula Torres Cardoso c,Rosimeire Aparecida Roela a, Suely Nonogaki d, João Carlos Guedes Sampaio Góes b,M. Mitzi Brentani a, Maria Aparecida Azevedo Koike Folgueira a,*

aDepartamento de Radiologia e Oncologia, LIM24, Faculdade de Medicina, Universidade de São Paulo, São Paulo, SP, Brazilb Instituto Brasileiro de Controle do Câncer, São Paulo, SP, BrazilcConsultoria em Patologia, São Paulo, SP, BrazildDepartamento de Patologia, Instituto Adolfo Lutz, São Paulo, SP, Brazil

a r t i c l e i n f o

Article history:Received 31 October 2012Accepted 1 April 2013

Keywords:Breast cancerCalcitriolProliferative indexGene expression

q This work was partly presented during the 200Annual Symposium, San Antonio, TX, USA.* Corresponding author. Faculdade de Medicina da

Departamento de Radiologia e Oncologia, Avenida Dr4112, CEP: 01246-903 São Paulo, SP, Brazil. Tel./fax: þ

E-mail address: makoike@lim24.fm.usp.br (M.A.Aze Yuri Nagamine Urata and Eduardo Carneiro de Ly

work.

0261-5614/$ e see front matter � 2013 Elsevier Ltd ahttp://dx.doi.org/10.1016/j.clnu.2013.04.001

s u m m a r y

Background & aims: High concentration of 1,25(OH)2D3 (50e100 nM), which cause hypercalcemia in vivo,induce the hormone transcriptional targets and exert antiproliferative effects in cultured breast cancerlineages, however, no studies investigated whether these effects might be reproduced in tumor speci-mens in vivo. Our aim was to evaluate the effects of calcitriol supplementation on the proliferative index(Ki67 expression) and gene expression profile of post-menopausal breast cancer samples.Methods & results: Tumor samples were collected from 33 patients, most of whom (87.5%) presenting25(OH)D3 insufficiency, before and after a short term calcitriol supplementation (0.50 mg/day PO, for 30days). Tumor dimension remained stable in ultrasound evaluations. A slight reduction in Ki67 immu-noexpression was detected, however in only 10/32 post-calcitriol samples an expressively low prolif-erative index [Ln (%Ki67þ) < 1] was achieved. Gene expression from 15 matched pre/post-supplementation samples was analyzed by microarray (U133 Plus 2.0 GeneChip, Affymetrix) and 15genes were over-expressed in post-supplementation tumors, including FOS and EGR1, which were pre-viously shown to be regulated by vitamin D. However, these results were not confirmed in another fourbreast cancer samples.Conclusions: Calcitriol supplementation is neither sufficient to expressively elicit an antiproliferativeresponse nor to induce the hormone transcriptional signaling pathway in breast cancer specimens.

� 2013 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.

1. Introduction

There is some evidence that women affected by breast cancerpresent lower 1,25OH)2D3 or 25(OH)D3 serum concentration thanunaffected ones1,2 and that patients with 25(OH)2D3 present a pooroutcome of the disease.3

9 San Antonio Breast Cancer

Universidade de São Paulo,. Arnaldo, 455, 4� andar, sala55 11 3082 6580.evedoK. Folgueira).ra participated equally in this

nd European Society for Clinical N

Receptors for 1,25(OH)2D3 (VDR) were shown in breast cancercell lines as well as in most breast cancer specimens. In addition,local synthesis of the active metabolite of vitamin D was demon-strated in normal human breast, as well as in breast carcinomasamples.2,4 These observations indicate that both normal andcancerous tissues may be targets of 1,25(OH)2D3 action.

Enough data suggest an antiproliferative effect of vitamin D inbreast cancer cell lines.4,5 However, most studies unravelingcellular processes regulated by 1,25(OH)2D3 were performed incancer lineages exposed to high concentrations of the hormone,which is associated with hypercalcemia in vivo. This approach doesnot take into consideration the tumor microenvironment andphysiological hormone concentrations, that may differentially in-fluence tumor behavior.

utrition and Metabolism. All rights reserved.

Y.N. Urata et al. / Clinical Nutrition 33 (2014) 136e142 137

Vitamin D antitumoral effects in vivo are controversial. Itwas reported that vitamin D may have chemopreventive andanti-tumoral actions in animal studies.6 However, in a great pro-spective trial of calcium and vitamin D supplementation (400 IUvitamin D3 daily for a long period of time) to postmenopausalwomen, invasive breast cancer incidence was not reduced.7 On theother hand, vitamin D topical administration was related with atrend toward clinical benefit in breast cancer lesions.8 None ofthese studies however, have directly analyzed human tumor sam-ples in search for transcriptional effects of vitamin D.

Our previous data suggest that exposure of breast cancer tissueslices in vitro to 1,25(OH)2D3 0.5 nM, a concentration that may beattained in vivo,9 elicits a very modest transcriptional response.10

Our present aim was to analyze the potential effects of calcitriolsupplementation (dose to reduce osteoporosis complications)11 topost-menopausal women in tumor proliferation (as determined byKi67 expression) and gene expression profile.

2. Patients and methods

2.1. Patients

Post-menopausal patients (n ¼ 33) with operable breast cancer,in the absence of distant metastasis, were invited to take part in thestudy (Table 1). Post-menopausal status was defined as amenor-rhea for at least one year in women �50 years old or bilateral oo-phorectomy accompanied by serum estradiol and FSH in post-menopausal level.

This protocol was carried out in compliance with the HelsinkiDeclaration andwas approvedby the Institutional Ethics Committee(Comitê de Ética do Instituto Brasileiro de Controle do Câncer, pro-tocol number 108/2006/7; Comitê de Ética para Análise de Projetosde Pesquisa do Hospital das Clínicas da Faculdade de Medicina daUniversidade de São Paulo, CAPPesq protocol number 626/06). Awritten informed consent was signed by all participants. This pro-tocol was registered at clinicaltrials.gov identifier NCT00926315.

Patients had blood and tumor samples collected during biopsy,and were prescribed with calcitriol supplementation (Rocaltrol�0.50 mg/day PO), as recommended for osteoporosis prevention,11 forapproximately one month. This was an exploratory study aimed toevaluate calcitriol effects in tumor samples. The intervention wasnot meant to interfere with treatment protocol of the institutionand one month was the usual timeframe between diagnosis andbreast surgery.

Physical examination of the breasts and regional lymph nodes,as well as breast ultrasonography (carried out by the same exam-iner, PEZA), were performed before and after calcitriol supple-mentation. In an early phase of the study, ten patients made use ofcalcitriol 0.25 mg/day PO, and were excluded of further analysis.

Table 1Characteristics of patients.

Cases n ¼ 33 Control n ¼ 23 p

Median age 62 (49e67) 60 (48e87) ns (MW)IDC 23/33 (69.7%) 20/33 (87%) nsHG 3 3/23 (13%) 4/18 (22.2%) nsER(þ) 22/33 (66.7%) 17/23 (73.9%) 0.562HER2(þ) 14/33 (42.4%) 11/23 (47.8%) 0.689Ki67 > 10% 20/32 (62.5%) 10/23 (43.5%) 0.162CS I/II 22/33 (66.7%) 21/23 (91.3%) 0.032

Patients received calcitriol supplementation (cases, n ¼ 33) or did not receive(control, n¼ 23). Median age: years (minemax); IDC: invasive ductal carcinoma; HG3: histological grade 3; ER(þ): estrogen receptor immunoexpression positive;Her2(þ): Her2(þ):3þ/3 þ immunoexpression; CS I/II: clinical stage I/II. p: Pearsonchi square or Mann Whitney test (MW), as appropriate. ns: not significant.

A control group of post-menopausal breast cancer patients(n¼ 23), whowere not supplementedwith calcitriol, was identifiedfrom hospital records. This analysis was approved by the Institu-tional Ethics Committee. These historical controls were used forcomparison of Ki67 tumor expression on biopsy and surgery, andmedian time period between both procedures was 37 days. Tumorcharacteristics were similar to the calcitriol supplementationgroup, except for a higher percentage of patients diagnosed withearly stage disease (Table 1).

2.2. Ki67 immuno-expression

Sections of 3 mm thickness were cut from paraffin blocks con-taining invasive cancer, and immunohistochemistry for Ki67 (Dako,Glostrup, Denmark) (clone MIB-1, DAKO 7240, dilution 1:50) wascarried out, with antigen retrieval using pressure cooking in 10 nMcitrate buffer at pH 6.0, and detection via DAKO EnVision�þ Sys-tem, HRP (Dako). Only nuclear staining of invasive carcinoma cellswas used for scoring, under light microscope Nikon e Eclipse E600(Shinagawa-ku, Tokyo, Japan). Images were acquired on NikonACT-1 Software using 20� objective, and 1000 tumor cells werecounted, on each sample.

Ki67 expression was evaluated by two pathologists (RB andAPTC), who were blinded concerning sample origin. A positive cor-relation was found for Ki67 immunoexpression (n ¼ 82, r ¼ 0.850,p< 0.0001; Spearman correlation). Themean value of expression (%of positive cells) was calculated and used for further analysis.

To define a significant lowKi67 proliferative index a cut off valuedetermined by the natural logarithm of percentage positive Ki67cells of less than 1 in tumor samples was adopted, as reported byother authors.12

2.3. Vitamin D serum determination

25(OH)D3 and 1,25(OH)2D3 serum concentration was deter-mined by 125I radioimmunoassay (RIA) Kit (DiaSorin Inc., Stillwater,Minnesota, USA).

2.4. RNA extraction and microarray hybridization

Tumor specimens obtained during biopsy (pre-supplementa-tion) or breast surgery (post-supplementation) were handdissected and samples with at least 70% tumor cells were furtherprocessed. Total RNA was isolated using RNeasy kit (Qiagen,Valencia, CA, USA) and 100 ng (RNA integrity number > 4.8; me-dian value 8.4, verified in a Bioanalyzer 2100, Agilent Technologies,Santa Clara, CA, USA) were processed through a two-round linearamplification (Two Cycle Target Labeling Kit, Affymetrix, SantaClara, CA, USA), followed by biotin-labeled cRNA synthesis fromdouble strand cDNA, using IVT labeling kit (Affymetrix). After-wards, 15 mg of biotinylated fragmented aRNA was hybridized ontoHuman Genome U133 Plus 2.0 GeneChip (Affymetrix), which wasscanned using Affymetrix GeneChip Scanner 3000. The backgroundcorrection, normalization and summarization of raw data (CELfiles) was performed using the Robust Multi-Array Average (RMA)method, available on R package. Filtering was set to select 30% ofgenes with the highest standard deviation. Comparisons ofexpression levels were performed using MeV (MultiExperimentViewer, version 4.5.1) software.13 Differential gene expression wasassessed using two class paired Significance Analysis of Microarray(SAM) for post vs pre-supplementation samples and using twoclass SAM for pre-supplementation samples attaining a low pro-liferative index (ln Ki67 < 1) vs those not attaining a low prolifer-ative index (ln Ki67 � 1) after supplementation. Unsupervisedhierarchical clustering based on Euclidean distance and average

Fig. 1. Ki67 expression in breast cancer samples. Post-menopausal patients weresupplemented with calcitriol for a median period of 30 days. Tumor samples werecollected before (through biopsy, left hand side) or after (during breast surgery, righthand side) supplementation and processed for Ki67 immunodetection. Tumorsincluded in the control group were collected from post-menopausal breast cancerpatients, who were not supplemented with calcitriol. Bar inside figure: 10 mm.

Y.N. Urata et al. / Clinical Nutrition 33 (2014) 136e142138

linkagewas used to verify association patterns. The reliability of theclustering was assessed by the Bootstrap technique usingMEV. Rawdata complying with MIAME format can be achieved at the GeneExpression Omnibus (GEO) data repository (www.ncbi.nlm.nih.gov/projects/geo).

To explore functional enrichment associated with calcitrioltreatment based on Ontologies (GO, Pathway) and other features,differentially expressed genes were subject to subsequent analysisusing ToppFun, available on ToppGene Suite (http://toppgene.cchmc.org/enrichment.jsp)14 and were considered significant ifp < 0.05.

2.5. Real time RT-PCR

At first, cDNA was synthesized from 1 mg aRNA. Reverse tran-scription was performed with random primers and Superscript III(Invitrogen Corporation, Carlsbad, CA, USA). Real-time RT-PCR wascarried out using specific primers and SYBR-green I (Sigma, St.Louis, MO, USA) in a Rotor-gene system (Corbett Research, Mor-tlake, Australia). GAPDH expressionwas used as an internal control.Relative expression of target genes was normalized to the value ofHB4A mammary epithelial cell line (donated by Drs. Mike O’Hareand Alan Mackay, Ludwig Institute for Cancer Research, London,UK) and calculated using the 2�DDCT methodology.

2.6. Statistical analysis

KolmogoroveSmirnov test was applied to check for normality ofthe data, followed by parametric or non-parametric tests, asappropriate. To detect an association between variables, Pearsonchi-square or Fisher exact tests were used. A two-tailed p value�0.05 was considered significant. Analysis was undertaken usingSPSS software (Chicago, IL, USA).

3. Results

3.1. 25(OH)D3 serum concentration and tumor dimension

Thirty two patients had at least one 25(OH)D3 serum concen-tration determination (collected pre- and/or post-supplementation). Most patients (87.5%) presented 25(OH)D3insufficiency (<30 ng/mL), including six patients (18.7%) with25(OH)D3 deficient levels (<20 ng/mL). There was no correlationbetween 25(OH)D3 serum concentration and age. Consideringpaired samples collected before and after calcitriol supplementa-tion, no differences were detected in 25(OH)D3 serum concentra-tion (n ¼ 26; 23.7 � 5.9, pre- vs 23.4 � 6.1 ng/mL, post-supplementation, ns, paired t test).

The first 12 patients included in the study had also serum1,25(OH)2D3 concentrations determined at least once [pre- (n ¼ 11)and/or post- (n ¼ 9) calcitriol supplementation), and only onedetermination was below (15 pg/mL) the reference values (15.9e55.6pg/mL). Comparingeightmatchedsamples, nodifferenceswerefoundbetweenpre andpost supplementationvalues (medianvaluespre and post-supplementation: 51.7 vs 57.5 pg/mL; related samplesWilcoxon signed ranks test: p ¼ 0.401). In addition, 1,25(OH)2D3

serum concentration above the reference level (55.6 pg/mL) wasfound in 4/11 (36%) patients before supplementation and in 6/9(67%) patients after calcitriol supplementation. No further analysisconsidering 1,25(OH)2D3 concentrations were performed.

In these patients, tumor dimension evaluated by ultrasonogra-phy, before and after calcitriol supplementation, did not vary(n ¼ 28; 27.0 � 11.5 mm vs 27.5 � 12.4 mm, pre- and post-supplementation, respectively; ns, related samples Wilcoxonsigned rank test).

3.2. Ki67 tumor expression

Ki67 tumor expression was determined in matched samplesfrom each patient (Fig. 1) and in the calcitriol group, a modestreduction of 35%, was detected (Ki67 immunoexpression medianpre- vs post-supplementation: 15.7% vs 10.2%; n ¼ 32, p ¼ 0.032,Wilcoxon test) but not in the historical control group (Ki67 surgery/Ki67 biopsy, n ¼ 23) of patients.

On the other hand, analyzing samples collected in the surgicalprocedure (post-calcitriol supplementation or control groups) andassuming the previously presented cut off value of ln Ki67 < 1, asdefined by other authors (12), for detecting a low proliferative in-dex, no differences between groups were identified, as 5/23 (21.7%)patients from the control group and 10/32 (31.3%) patients from thesupplementation group presented a low proliferative index(p ¼ 0.435; Pearson chi-square) (Fig. 2).

Dividing samples in accordance with this criteria of proliferativeindex (ln Ki67 < 1 or �1 after calcitriol supplementation) therewere no differences in 25(OH)D3 (Fig. 3) serum levels (both beforeand after supplementation) from patients included in both groups.There were neither differences in age of patients and duration ofcalcitriol supplementation nor in tumor dimension, evaluated byultrasonography both pre- and post-calcitriol supplementation(Fig. 3). In addition, a low proliferative index post-supplementationwas not associated with tumor histology, estrogen receptorimmunoexpression, histological grade and TNM pathologicalstaging (pTpNM; stages I, II, III; AJCC Cancer Staging Manual, 6thed., 2003) (Supplementary Table 1).

3.3. Gene expression profile

To evaluate calcitriol effects on gene expression, 15 paired preand post-supplementation tumor samples were analyzed,revealing 21 probes differentially expressed representing 15 knowngenes, all of them over-expressed in post-supplementation samples(SAM paired analysis, FDR < 10) (Table 2). Enriched molecular

Fig. 2. Ki67 expression in paired individual tumors from patients supplemented or not(control) with calcitriol. Tumor samples were collected before (through biopsy) or after(during breast surgery) supplementation and processed for Ki67 immunodetection.Tumors included in the control group were collected from post-menopausal breastcancer patients, who were not supplemented with calcitriol. Each line representsmatched tumor samples from a single patient [pre & post-supplementation or biopsy(bio) & surgery (sur)]. Interrupted lines: Ln (%Ki67þ) >1; Continuous line: Ln (%Ki67þ): <1, in samples collected during surgery.

Table 2Genes potentially regulated by calcitriol supplementation. Fifteen patients whoreceived calcitriol supplementation had matched tumor samples (pre and post-supplementation) available for gene expression analysis by microarray. SAMpaired analysis (FDR < 10) revealed, 15 known genes, over expressed in post-supplementation samples.

Gene_symbol Gene_title Foldchange

FAM13A1 Family with sequence similarity 13, member A1 1.28SEMA5A Sema domain, seven thrombospondin repeats

(type 1 and type 1-like), transmembrane domain(TM) and short cytoplasmic domain,(semaphorin) 5A

1.35

KCNJ2 Potassium inwardly-rectifying channel, subfamily J,member 2

1.36

NA Transcribed locus 1.42ADAMTS3 ADAM metallopeptidase with thrombospondin

type 1 motif, 31.51

ITM2A Integral membrane protein 2A 1.57LAMA3 Laminin, alpha 3 1.62FAM38B Family with sequence similarity 38, member B 1.68PEG3 Paternally expressed 3 1.73PEG3 Paternally expressed 3 1.74ATF3 Activating transcription factor 3 1.79NA NA 1.83NR4A3 Nuclear receptor subfamily 4, group A, member 3 1.90EGR1 Early growth response 1 1.94SCARNA2 Small Cajal body-specific RNA 2 1.96RGS1 Regulator of G-protein signaling 1 2.04RGS1 Regulator of G-protein signaling 1 2.18EGR1 Early growth response 1 2.22EGR1 Early growth response 1 2.34FOS v-fos FBJ murine osteosarcoma viral oncogene

homolog2.58

PBEF1 Pre-B-cell colony enhancing factor 1 2.85

Y.N. Urata et al. / Clinical Nutrition 33 (2014) 136e142 139

functions were sequence-specific DNA binding transcription factoractivity, comprehending 5 genes: NR4A3, nuclear receptor sub-family 4, group A, member; ATF3, activating transcription factor 3;FOS, FBJ murine osteosarcoma viral oncogene homolog; PEG3,paternally expressed 3; EGR1, early growth response 1 (Toppgene,p < 0.05). Unsupervised hierarchical clustering could not segregatesamples according to calcitriol supplementation and individualtumor similarities prevailed, as specimens from the same patientpreferentially grouped in the same branch (Fig. 4).

Fig. 3. Age of patients, duration of calcitriol (D3) supplementation (days), tumordimension evaluated by ultrasonography (mm) and 25(OH)D3 serum concentration(ng/mL) pre- and post-calcitriol supplementation, grouped according to lnKi67 (�1 or<1) in tumor samples after supplementation. No significant differences were foundbetween groups (p > 0.05, Mann Whitney or Student t test, as appropriate).

As a technical validation procedure, expression of four selectedgenes was further evaluated in 14 paired samples, already includedin the microarray analysis, using RT-qPCR, and a positive correla-tion between methods was observed for the expression of all ofthem (Pearson correlation, p < 0.05; Supplementary Fig. 1).

Furthermore, a supplemental evaluation was performed todetermine the expression of candidate genes (EGR1, ATF3, FOS,PBEF1 and RGS1) using RT-qPCR in another four pre and post cal-citriol supplementation samples, which were not previouslyanalyzed by microarray, however we could not find significantdifferences (Fig. 5) (RGS1: n ¼ 3, p:ns, data not shown).

In addition, 15 samples collected before calcitriol supplementa-tion and analyzed by microarray were also tested whether a dif-ferential transcriptional profile could predict a low proliferativeindex after supplementation (ln Ki67 < 1; n ¼ 5 vs ln Ki67 � 1;n ¼ 10). Using SAM, FDR ¼ 0 and all possible permutations, 10known genes and two cDNA probes were identified as moreexpressed in low proliferative samples after calcitriol supplemen-tation (Table 3), with enrichment of extracellular matrix cellularcomponents as KERA, keratocan; CILP, cartilage intermediate layerprotein, nucleotide pyrophosphohydrolase; PAPLN, papilin,proteoglycan-like sulfated glycoprotein; COL14A1, collagen, typeXIV, alpha 1; COL3A1, collagen, type III, alpha 1 (Toppgene, p< 0.05).

4. Discussion

In the present study, the great majority of the patients presented25(OH)D3 insufficiency. These results were not expected for SãoPaulo city, localized in a region of tropical/sub-tropical weather, asin a study conducted in breast cancer patients in Canada, a countrywith a higher latitude, 25(OH)D3 insufficient and deficient levelswas less frequent.3 This data indicates that vitamin D supplemen-tation deserves better attention in our environment.

Fig. 4. Hierarchical clustering of breast cancer samples from post-menopausal patients supplemented with calcitriol. Matched samples (identified by numbers 15e43) werecollected before (pre) and after (post) calcitriol supplementation and gene expression was analyzed by microarray. Each column represents a tumor sample. Colored lines of thedendrogram stand for the support for each clustering, black and gray, more reliable; yellow and red, less reliable. (For interpretation of the references to color in this figure legend,the reader is referred to the web version of this article.)

Y.N. Urata et al. / Clinical Nutrition 33 (2014) 136e142140

This study was planned to evaluate the tumor effects of a shortduration of vitamin D supplementation, without interfering withthe timing of surgery and systemic treatment. Calcitriol was chosenfor two basic reasons: 1,25(OH)2D3 is the biologically active form ofthe hormone and 1,25(OH)2D3 serum concentration in breastcancer patients was reduced as compared with women withoutbreast cancer, in our previous work.2 We assumed that supple-mentation with other formulations, which are not biologicallyactive, would not be appropriate for a short duration study, due to alag time to response, depending on 25(OH)D3 initial serum valuesand loading dose, based on earlier works demonstrating that cu-mulative monthly doses of vitamin D3 of 45,000 IU taken in dailyandweekly oral regimes would result in increased serum 25(OH)D3concentration and steady state values only by day 7 and after onemonth, respectively.15

In accordance with the present results, it was reported thatcalcitriol supplementation for post-menopausal women for 1 year,in the same doses used here, were neither associated with1,25(OH)2D3 nor 25(OH)D3 serum elevations. Despite no in-crements in vitamin D serum levels, a longer period of calcitriolsupplementation was associated with a clinical benefit, repre-sented by gain in bone mass, resulting from reduced bone turnover,indicating that bone tissue effects were present.16 However, it is notpossible to assume that a short treatment using low calcitriol doseswould suffice for breast tissue actions.

We recognize that one weakness of this work is that otherstrategies might have been tried in order to reach higher, but stillsafe, 1,25(OH)2D3 serum concentrations and probably more effi-cient tissue levels. Phase I clinical studies indicate that subcu-taneous doses of calcitriol 2e10 mg given every other day result inpeak 1,25(OH)2D3 serum concentration of 0.25e0.75 nM9 while

weekly pulses of oral calcitriol until 2.8 mg/kg allow higher doseadministration and peak serum concentrations of 1e15 nM.17

Another option would be a large oral bolus of vitamin D3(600.00UI), which is associatedwith a rapid, around 35% increase in1,25(OH)2D3 levels, to 0.15 nM in 24 h.18

In these tumor samples, a clear antiproliferative response couldnot be identified, however a modest reduction in tumor prolifera-tive index of about 35% might be recognized. This small Ki67 vari-ation could not be associated with 25(OH)D3 levels both before andafter supplementation, indicating that serum concentration maynot be used to predict which patients may potentially benefit fromcalcitriol supplementation. In addition, Ki67 variation attributed tocalcitriol was less important than that described in clinicaltrialsreporting neoadjuvant endocrine therapy with tamoxifen, anas-trozol or letrozol, inwhich an early and steep decline in Ki67 tumorexpression, varying from 60 to 75% of the original values, followingjust 2 weeks of administration might be distinguished.12,19

A few samples presented a low proliferative index after sup-plementation and 10 genes might potentially identify them, five ofwhich coding for components of the extracellular matrix, as KERA,CILP, PAPLN, COL14A1 and COL3A1. In addition, four genes may beco-expressed by stromal cells as HMGCS2, 3-hydroxy-3-methylglutaryl-CoA synthase 2 (mitochondrial); CILP, COL14A1and COL3A1. One hypothesis is that tumor fibroblasts, which arealso targets of vitamin D actions, may influence vitamin D tumorresponsiveness.20

Previous works investigating calcitriol effects on gene expres-sion have usually used in vitro systems in which cells were culturedin the presence of large quantities of the hormone for some hoursor a few days. These studies allowed the identification of a largenumber of vitamin D regulated genes in breast cancer cells.5,21

Fig. 5. Expression of ATF3, FOS, EGR1, PBEF1 in tumor samples before and after cal-citriol supplementation. RT-qPCR was performed in paired tumor samples fromanother 4 patients supplemented with calcitriol. Relative expression values are shownon the Y axis. Pre-supplementation samples on the left; post-supplementation sampleson the right. There were no expression differences between groups (p > 0.05, Wilcoxonsigned ranks test).

Y.N. Urata et al. / Clinical Nutrition 33 (2014) 136e142 141

However, in a more physiological model of breast cancer slicesexposed in vitro to 1,25(OH)2D3 in a near physiological concentra-tion (0.5 nM) only a fewgenes were induced by the hormone.10 Thisobservation is in line with the present results indicating that

Table 3Differentially expressed genes in accordance with tumor proliferative status aftercalcitriol supplementation. Ten known genes and two cDNA clones were moreexpressed in pre-supplementation samples attaining a low proliferative index (lnKi67 < 1, n ¼ 5) as compared with those samples not attaining a low proliferativeindex (ln Ki67 � 1; n ¼ 10) after calcitriol supplementation (SAM; FDR ¼ 0; allpossible permutations).

Gene symbol Gene title Fold change

HMGCS2 3-hydroxy-3-methylglutaryl-CoenzymeA synthase 2 (mitochondrial)

12.27

COL14A1 Collagen, type XIV, alpha 1 (undulin) 11.49PAPLN Papilin, proteoglycan-like sulfated glycoprotein 8.35KERA Keratocan 8.27MYRIP Myosin VIIA and Rab interacting protein 7.10NA CDNA FLJ38472 fis, clone FEBRA2022148 5.85F2RL2 Coagulation factor II (thrombin) receptor-like 2 5.79CILP Cartilage intermediate layer protein,

nucleotide pyrophosphohydrolase4.82

SLC25A27 Solute carrier family 25, member 27 4.68NA CDNA FLJ30478 fis, clone BRAWH1000167 4.61COL3A1 Collagen, type III, alpha 1 (Ehlers-Danlos

syndrome type IV, autosomal dominant)4.43

STEAP2 Six transmembrane epithelial antigenof the prostate 2

4.40

calcitriol transcriptional response is concentration dependent, anddismal in near physiological concentrations.

On the other hand, we could identify a trend toward transcrip-tional induction after calcitriol supplementation, including FOS andEGR1, which were previously shown to be regulated by vitamin D.FOS has a critical function in regulating cellular growth anddevelopment and vitamin D responsive elements in the promoterregion of the genewere already described.22 Early growth responsefactor-1 (EGR1) appears to function as a convergence point forsignaling pathways as proliferation, inflammation and apoptosis,through direct regulation of multiple tumor suppressors, includingTGFB1, PTEN and TP53 and up regulation of EGR1 by 1,25(OH)2D3contributing to terminal cell differentiation and growth arrest waspreviously observed in HL60 leukemia cells.23 Another genepotentially regulated was NAMPT/PBEF/visfatin, which codes anenzyme that catalyzes the first step in NAD biosynthesis fromnicotinamide and regulates growth, apoptosis and angiogenesis.Results from our group suggest that NAMPT is also up regulated inbreast cancer associated fibroblasts exposed to vitamin D, indi-cating that this regulation might be taking place, at least partially,in this cell compartment.20 However, calcitriol supplementationtranscriptional response seems to be very discrete, as it could notbe confirmed in another set of samples, examined through RT-PCR.

In conclusion, this was the first study to evaluate calcitriolsupplementation effects in breast cancer samples from post-menopausal patients. Our findings indicate that most post-menopausal breast cancer patients present vitamin D insuffi-ciency, at least in our environment. In these patients, a short periodof calcitriol supplementation in a dose recommended for preven-tion of bone events, was neither effective in reducing Ki67expression to expressively low levels nor eliciting an expressivetranscriptional response. Other regimens of vitamin D supple-mentation may help clarify the potential role of the hormone inbreast cancer and chemoprevention.

Author’s contributions

ECL, MLHK, MAAKF designed the study; ECL, YNU, RB, RAR,APTC, PEZA, RB, SN, JCGSG participated in data acquisition; YNU,ECL, MLHK, MAAKF, participated in data analysis and interpreta-tion; YNU, ECL, MLHK, MMB, MAAKF drafted and revised themanuscript.

Conflict of interest statement

The authors declare that they have no competing interests.

Acknowledgments

This work was supported by Fundação de Amparo à Pesquisa doEstado de São Paulo (FAPESP) and Coordenação de Aperfeiçoa-mento de Pessoal de Nível Superior (CAPES).

The authors would like to thank the helpfull assistence of Dr.Eloá M de Freitas Alves, Dr. João Guidugli Neto, Dr. Rosa MariaRodrigues Pereira, Mrs. Lilian Masako Takayama, Paulo Roberto delValle (for performing microarray analysis), Dr Antonio Hugo JoséFróesMarques Campos and the critical review of Dr. Igor Snitcovskyand Prof. Geertruida Hendrika de Bock.

Appendix A. Supplementary material

Supplementary data related to this article can be found online athttp://dx.doi.org/10.1016/j.clnu.2013.04.001.

Y.N. Urata et al. / Clinical Nutrition 33 (2014) 136e142142

References

1. Abbas S, Linseisen J, Slanger T, Kropp S, Mutschelknauss EJ, Flesch-Janys D, et al.Serum 25-hydroxyvitamin D and risk of post-menopausal breast cancer resultsof a large case-control study. Carcinogenesis 2008;29(1):93e9.

2. de Lyra EC, da Silva IA, Katayama ML, Brentani MM, Nonogaki S, Goes JC, et al.25(OH)D3 and 1,25(OH)2D3 serum concentration and breast tissue expressionof 1alpha-hydroxylase, 24-hydroxylase and vitamin D receptor in womenwith and without breast cancer. J Steroid Biochem Mol Biol 2006;100(4e5):184e92.

3. Goodwin PJ, Ennis M, Pritchard KI, Koo J, Hood N. Prognostic effects of 25-hydroxyvitamin D levels in early breast cancer. J Clin Oncol 2009;27(23):3757e63.

4. Bortman P, Folgueira MA, Katayama ML, Snitcovsky IM, Brentani MM. Anti-proliferative effects of 1,25-dihydroxyvitamin D3 on breast cells: a mini review.Braz J Med Biol Res 2002;35(1):1e9.

5. Swami S, Raghavachari N, Muller UR, Bao YP, Feldman D. Vitamin D growthinhibition of breast cancer cells: gene expression patterns assessed by cDNAmicroarray. Breast Cancer Res Treat 2003;80(1):49e62.

6. Krishnan AV, Swami S, Feldman D. Equivalent anticancer activities of dietaryvitamin D and calcitriol in an animal model of breast cancer: importance ofmammary CYP27B1 for treatment and prevention. J Steroid Biochem Mol Biol2012. http://dx.doi.org/10.1016/j.jsbmb.2012.08.005.

7. Chlebowski RT, Johnson KC, Kooperberg C, Pettinger M, Wactawski-Wende J,Rohan T, et al. Calcium plus vitamin D supplementation and the risk of breastcancer. J Natl Cancer Inst 2008;100(22):1581e91.

8. Bower M, Colston KW, Stein RC, Hedley A, Gazet JC, Ford HT, et al. Topicalcalcipotriol treatment in advanced breast cancer. Lancet 1991;337(8743):701e2.

9. Smith DC, Johnson CS, Freeman CC, Muindi J, Wilson JW, Trump DL. A phase Itrial of calcitriol (1,25-dihydroxycholecalciferol) in patients with advancedmalignancy. Clin Cancer Res 1999;5(6):1339e45.

10. Milani C, Welsh J, Katayama ML, Lyra EC, Maciel MS, Brentani MM, et al.Human breast tumor slices: a model for identification of vitamin D regulatedgenes in the tumor microenvironment. J Steroid Biochem Mol Biol2010;121(1e2):151e5.

11. Tilyard MW, Spears GF, Thomson J, Dovey S. Treatment of postmenopausalosteoporosis with calcitriol or calcium. N Engl J Med 1992;326(6):357e62.

12. Baselga J, Semiglazov V, van Dam P, Manikhas A, Bellet M, Mayordomo J, et al.Phase II randomized study of neoadjuvant everolimus plus letrozole compared

with placebo plus letrozole in patients with estrogen receptor-positive breastcancer. J Clin Oncol 2009;27(16):2630e7.

13. Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N, et al. TM4: a free, open-source system for microarray data management and analysis. Biotechniques2003;34(2):374e8.

14. Chen J, Bardes EE, Aronow BJ, Jegga AG. ToppGene suite for gene list enrich-ment analysis and candidate gene prioritization. Nucleic Acids Res 2009;37:W305e11.

15. Ish-Shalom S, Segal E, Salganik T, Raz B, Bromberg IL, Vieth R. Comparison ofdaily, weekly, and monthly vitamin D3 in ethanol dosing protocols for twomonths in elderly hip fracture patients. J Clin Endocrinol Metab 2008;93(9):3430e5.

16. Sairanen S, Kärkkäinen M, Tähtelä R, Laitinen K, Mäkelä P, Lamberg-Allardt C,et al. Bone mass and markers of bone and calcium metabolism in post-menopausal women treated with 1,25-dihydroxyvitamin D (calcitriol) for fouryears. Calcif Tissue Int 2000;67(2):122e7.

17. Beer TM, Munar M, Henner WD. A phase I trial of pulse calcitriol in patientswith refractory malignancies: pulse dosing permits substantial dose escalation.Cancer 2001;91(12):2431e9.

18. Rossini M, Adami S, Viapiana O, Fracassi E, Idolazzi L, Povino MR, et al. Dose-dependent short-term effects of single high doses of oral vitamin D(3) on boneturnover markers. Calcif Tissue Int 2012;91(6):365e9.

19. Dowsett M, Ebbs SR, Dixon JM, Skene A, Griffith C, Boeddinghaus I, et al.Biomarker changes during neoadjuvant anastrozole, tamoxifen, or the combi-nation: influence of hormonal status and HER-2 in breast cancer e a study fromthe IMPACT trialists. J Clin Oncol 2005;23(11):2477e92.

20. Campos LT, Brentani H, Roela RA, Katayama ML, Lima L, Rolim CF, et al. Dif-ferences in transcriptional effects of 1a, 25 dihydroxyvitamin D3 on fibroblastsassociated to breast carcinomas and from paired normal breast tissues. J SteroidBiochem Mol Biol 2013;133:12e24.

21. Towsend K, Trevino V, Falciani F, Stewart PM, Hewison M, Campbell MJ.Identification of VDR-responsive gene signatures in breast cancer cells.Oncology 2006;71(1e2):111e23.

22. Schrader M, Kahlen JP, Carlberg C. Functional characterization of a novel typeof 1 alpha,25-dihydroxyvitamin D3 response element identified in the mousec-fos promoter. Biochem Biophys Res Commun 1997;230(3):646e51.

23. Chen F, Wang Q, Wang X, Studzinski GP. Up-regulation of Egr1 by 1,25-dihy-droxyvitamin D3 contributes to increased expression of p35 activator of cyclin-dependent kinase 5 and consequent onset of the terminal phase of HL60 celldifferentiation. Cancer Res 2004;64(15):5425e33.