A

hscdpaowteo©

K

1

diReDcea2c

cuH

0d

Journal of Steroid Biochemistry & Molecular Biology 103 (2007) 763–767

Chemopreventive efficacy of 25-hydroxyvitamin D3 in colon cancer

Genoveva Murillo a, Damien Matusiak b, Richard V. Benya b, Rajendra G. Mehta a,∗a Carcinogenesis and Chemoprevention Division, IIT Research Institute, 10 West 35th Street, Chicago, IL 60616, USA

b Department of Medicine, University of Illinois, Chicago, IL 60616, USA

bstract

Recently, it has been reported that 25-hydroxyvitamin D3-1�-hydroxylase [1�(OH)ase, CYP27B1], required to convert non-toxic 25-yxdroxyvitamin D3 [25(OH)D3] to its active metabolite [1�,25(OH)2D3], is present in the epithelial cells of the human colon. In the presenttudy, the potential chemoprotective role of 25(OH)D3 was evaluated for colon cancer using the HT-29, human colon cancer cell line. Colonancer cells were treated with 25(OH)D3 (500 nM or 1 �M), 1�,25(OH)2D3 (500 nM), cholecalciferol (D3, 1 �M) or vehicle and cell numberetermined at days 2 and 5 post-treatment. Results showed that both 25(OH)D3 and 1�,25(OH)2D3 induced dose- and time-dependent anti-roliferative effects on the HT-29 cells, with maximum inhibition noted at day 5. Western blot analyses revealed an up-regulation of VDRnd 1�(OH)ase expression following 24 h of treatment with 25(OH)D3, and 1�,25(OH)2D3. These results are consistent with the expressionf VDR and 1�(OH)ase in samples of normal colonic tissue, aberrant crypt foci (ACFs) and colon adenocarcinomas. The VDR expressionas sequentially increased from normal to pre-cancerous lesions to well-differentiated tumors and then decreased in poorly differentiated

umors. Expression of 1�(OH)ase was equally expressed in normal, pre-cancerous lesions and malignant human colon tissues. The increasedxpression of 1�(OH)ase in colon cancer cells treated with the pro-hormone and its anti-proliferative effects, suggest that 25(OH)D3 mayffer possible therapeutic and chemopreventive option in colon cancer.

2007 Elsevier Ltd. All rights reserved.

eywords: Colon chemoprevention; 25-Hydroxyvitamin D3; VDR; CYP27B1

ntlpeTt2cafs2a

. Introduction

Epidemiological studies have been instrumental inemonstrating an inverse relationship between vitamin Dntake, sunlight exposure and risk for colon cancer [1].ecently, a meta-analysis conducted by Gorham et al. [2]xamined the dose response relationship between vitamin

intake or serum 25(OH)D3 and risk for colon can-er. Dose-gradient curves generated from 18 of the 44xisting observational studies demonstrated that individu-ls with ≥ 1000 IU/day oral vitamin D or ≥ 33 ng/ml serum5(OH)D3 had 50% lower risk for developing colon cancerompared to reference values [2].

Inspite of the extensive evidence demonstrating the anti-

ancer activity of 1,25(OH)2D3 in colon cancer, its clinicalse is limited by its susceptibility to cause hypercalcemia.ence, in order to retain the efficacy but reduce its toxicity

∗ Corresponding author. Tel.: +1 312 567 4970; fax: +1 312 567 4931.E-mail address: RMehta@iitri.org (R.G. Mehta).

a1msic

960-0760/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.oi:10.1016/j.jsbmb.2006.12.074

umerous analogs have been synthesized and evaluated forhe prevention and treatment of colon cancer. Until recently,ittle attention was given to 25(OH)D3 as a potential chemo-reventive agent since it was commonly believed that thenzyme 1�(OH)ase was selectively present in the kidney.he existence of 1�(OH)ase in colon epithelial cells has led

o interest in the potential use of this non-toxic pro-hormone,5(OH)D3, as an effective chemoprotective agent againstolon cancer [3–5]. The possibility of using the 25(OH)D3 aschemopreventive and/or chemotherapeutic agent has been

urther strengthened by in vitro studies which have demon-trated that cells containing 1�(OH)ase are able to convert5(OH)D3 into 1,25(OH)2D3 [6,7]. For example, Bareis etl. [7] demonstrated that the Caco-2 cells, which is a moder-tely differentiated colon cancer cell line, are able to produce,25(OH)2D3 from the pro-hormone. Here we report that nor-

al, aberrant crypt foci (ACF) and malignant human cancer

amples express VDR and 1�(OH)ase and that 25(OH)D3s efficacious as an anti-proliferative agent in human colonancer cells.

7 mistry

2

2

sUID

2

fmGgi

2

stwwb

2

iitsd

2

Sqc

2

eaptwmfhBt

ww

2

a(wmBbwsOtd

3

3t

e1mcmeiVnteI((cpc

h2V2(ctWa

64 G. Murillo et al. / Journal of Steroid Bioche

. Materials and methods

.1. Tumor specimens and histological grading

Colon cancers were randomly selected from the Univer-ity of Illinois at Chicago Gastrointestinal Tumor Bank. Theniversity of Illinois at Chicago and Veterans Administration

nstitutional Review Boards approved use of these tissues.ifferentiation was assessed as previously described [8].

.2. Human colon cancer cell lines

The HT-29, Caco-2 and SW480 cell lines were obtainedrom American Type Culture Collection (Manassas, VA) andaintained in RPMI 1640 media (Life Technologies, Inc.,rand Island, NY) with 10% fetal bovine serum, 2 mM l-lutamine and 1% antibiotic–antimycotic solution and keptn a 37 ◦C humidified atmosphere of 5% CO2.

.3. Analysis of cell proliferation

For determination of proliferation, HT-29 cells wereeeded at a density of 2 × 104 per well in a 12-well cell cul-ure plate and allowed to adhere overnight. After incubationith or without 25(OH)D3 for the appropriate times, cellsere detached with trypsin and cell number was determinedy the Coulter counter.

.4. FACS analysis

Colon cancer cells were seeded at a density of 5.0 × 105

n 25 cm2 flasks and allowed to adhere for 24 h. Follow-ng treatment with or without 1.0 �M 25(OH)D3 for 48 h,hey were harvested with trypsin and washed with PBS. Theamples were then stained with propidium iodide using theetergent-trypsin method described by Vindelov et al. [9].

.5. Measurement of apoptosis

Cells undergoing apoptosis were evaluated using the Initu Cell Death Detection Kit (Roche, Indianapolis, IN). Auantitative assessment was made by determining the per-entage of apoptotic cells.

.6. Western blot analysis

Treated and untreated cells were lysed in freshly preparedxtraction buffer. Protein concentration was determined usingmodified Lowry method (Bio Rad, Hercules, CA). Sam-

les were then separated on 10% polyacrylamide gels andransferred to nitrocellulose membranes. The membranesere blocked and then incubated with appropriate pri-ary and secondary antibodies. Anti-VDR antibody was

rom Neomarkers (Freemont, CA), sheep anti-murine 25-ydroxyvitamin D3-1�-hydroxylase antibody was from theinding Site (San Diego, CA). The chemiluminescence reac-

ion was performed using the ECL system. Bands of interest

1Vf4

& Molecular Biology 103 (2007) 763–767

ere compared to that of actin and relative intensity ratiosere calculated.

.7. Immunofluorescence studies

SW480 cells were seeded on cover slips and allowed todhere overnight. After incubation with or without 25(OH)D31 �M) for 24 h, the cells were fixed in buffered formalin,ashed with PBST (PBS containing 0.1% Tween 20), per-eabilized in 0.2% Triton X-100/PBS, blocked with 1%SA, and then incubated with anti-VDR rat monoclonal anti-ody (1:200) for 1 h at RT. Cells were washed and incubatedith TRIC-labeled anti-rat secondary antibody for 1 h. After

taining nuclei with DAPI, cells were visualized using thelympus BX51 microscope. Tissues were sectioned (4 �m

hick) and processed for immunohistochemistry as previouslyescribed [10].

. Results

.1. Expression of 1α(OH)ase and VDR in human colonissue and cancer cells

The expression patterns of VDR and 1�(OH)ase werevaluated in human colon tissues. As shown in Fig. 1A–E,�(OH)ase showed consistently strong expression in nor-al, premalignant (ACF) and malignant colonic epithelial

ells, irrespective of tumor cell differentiation. In contrast,inimal VDR expression was observed in normal colonic

pithelial cells, with the expression localized predominantlyn the nucleus (Fig. 1F). Samples revealed that the amount ofDR is significantly higher in ACFs than to that of adjacentormal tissue. Moreover, with malignant transformation,otal cellular VDR expression increased markedly and wasspecially high in well-differentiated tumors (Fig. 1H and) but then decreased with tumor cell de-differentiationFig. 1J). Furthermore, these data suggest that 1�(OH)aseFig. 1B) and VDR (Fig. 1G) expression levels in colorectalancer pre-cancerous lesions such as ACFs and polyps mayermit the pro-hormone 25(OH)D3 for use in colorectalancer chemoprevention.

The expression profile of VDR and 1�(OH)ase inuman colon tissue prompted us to investigate whether5(OH)D3 treatment regulated the expression levels ofDR and 1�(OH)ase in the HT-29 cells. For this, HT-9 cells were treated with vehicle (ETOH), 25(OH)D31 �M), 1�,25(OH)2D3 (0.5 �M; positive control) or chole-alciferol (D3, 1 �M; negative control). Following 24 h ofreatment, protein lysate was collected and separated for

estern blot analyses. As shown in Fig. 2A, both 25(OH)D3nd 1,25(OH)2D3 up-regulated the levels of expression of

�(OH)ase by 22 and 30%, respectively. In a similar manner,DR expression levels (Fig. 2B) were significantly increased

ollowing treatment with 25(OH)D3 and 1�,25(OH)2D3 by0 and 48%, respectively. Treatment with the cholecalicerfol

G. Murillo et al. / Journal of Steroid Biochemistry & Molecular Biology 103 (2007) 763–767 765

Fig. 1. Expression of 1�-hydroxylase and VDR in human colon tissue. Similar non-nuclear levels of 1�-hydroxylase expression are noted in normal colonice tumor ct elial celw or cellw (J). Mag

(1i

2lcwaVpac2tcVp2Cscrmia

3a

g

pt5oHa2fff

ctEceme72ti

4

h

pithelial cells (A), aberrant crypt foci (B), as well as, in well-differentiatedumor cells (E). VDR expression levels were low in normal colonic epithell-differentiated tumor cells (H). In contrast, moderately differentiated tumas found to be almost completely lost in poorly differentiated tumor cells

negative control) had no effect on the expression levels of�(OH)ase (Fig. 2A) and showed only a marginal increasen the levels of VDR (Fig. 2B).

In addition to HT-29 cells, we evaluated the effects of5(OH)D3 on SW480 colon cancer cells. The SW480 celline was selected for this study since it is one of the bestharacterized human colon cancer cell lines. Moreover, it isidely used as a model system for vitamin D compounds,

nd is shown to have both VDR negative [VDR(−)] andDR positive [VDR(+)] sub-populations (i.e., VDR(+) flat,olygonal and adherent line; VDR(−) rounded, refractilend less adherent) [11]. For these experiments, SW480ells were plated on cover slips and treated with or without5(OH)D3 (1 �M) for 24, 48 or 72 h. At the appropriateime points, cells were fixed and immunofluorescence wasompleted as described in Section 2. As shown in Fig. 2C,DR expression levels were increased in the VDR(+) flat,olygonal and adherent line by 24 h post-treatment with5(OH)D3 and remained elevated 72 h after treatment.umulatively, these data show that 25(OH)D3 treatment

ignificantly increases the levels of 1�(OH)ase and VDR inolon cancer cells, and support the notion that 25(OH)D3 canegulate the expression of VDR and 1�(OH)ase in a similaranner as the active metabolite in vitro. These results

ndicate that 25(OH)D3 can exhibit similar anti-cancerctivities as have been reported for 1�,25(OH)2D3.

.2. The pro-hormone 25-hydroxyvitamin D3 exerts

nti-proliferative effects on human colon cancer cells

The anti-proliferative effects of 25(OH)D3 were investi-ating using the HT-29 cells. For these studies, cells were

tmcl

ells (C), moderately differentiated tumor cells (D) and poorly differentiatedls (F), increased in aberrant crypt foci (G) with further increase noted ins (I) appear to have less VDR than the well differentiated. VDR expressionnification ×1000.

lated, allowed to adhere for 24 h prior to treatment withhe different vitamin D compounds. As shown in Fig. 3,-day treatment of HT-29 cells with either 1�,25(OH)2D3r 25(OH)D3 significantly inhibited the growth of theT-29 cells. The percent growth inhibition was 14, 65

nd 53% for D3 (1 �M), 1�,25(OH)2D3 (0.5 �M) and5(OH)D3 (0.5 �M), respectively. A dose-dependent effector 25(OH)D3 treatment was also evident (53% versus 77%,or 0.5 and 1.0 �M) for the HT-29 cells. Similar findings wereound in the Caco-2 and SW480 cells (data not shown).

In order to determine the effects of 25(OH)D3 on cellycle progression HT-29, Caco-2 and HCT116 cells werereated with 25(OH)D3 (1 �M), 1�,25(OH)2D3 (0.5 �M) orTOH for 48 h and DNA content was analyzed by flowytometry. Our studies failed to reveal cell cycle arrest withither 1�,25(OH)2D3 or 25(OH)D3 during the 48 h treat-ent period. However, using the tunnel assay, we found

vidence of SW480 cells undergoing apoptosis following2 h of treatment with either 1�,25(OH)2D3 (0.5 �M) or5(OH)D3 (1.0 �M). These results suggest that 25(OH)D3reatment may be inducing growth inhibition in part by caus-ng apoptosis in the colon cancer cells.

. Discussion

Vitamin D3 is biologically inert and requires successiveydroxylations by mitochondrial P450 enzymes present in

he liver and kidney to form the biologically active hor-

one 1�,25(OH)2D3. The final step in metabolism is theonversion of pro-hormone 25(OH)D3 to the active metabo-ite and hormone 1�,25(OH)2D3 by 1�(OH)ase. Due to

766 G. Murillo et al. / Journal of Steroid Biochemistry & Molecular Biology 103 (2007) 763–767

Fig. 2. Expression levels of 1�-hydroxylase and VDR in colon cancer cells treated with 25-hydroxyvitamin D3. Western blot analysis showing 1�-hydroxylase(A) and VDR (B) with appropriate actin control. HT-29 cells were treated with 11�,25(OH)2D3 and 25(OH)D3 were found to up-regulate 1�-hydroxylase and VDRfor 24–72 h. Immunofluorescence revealed an up-regulation of VDR by 24 h and sh

Faag

tnatncrptcite

ig. 3. HT-29 cells were treated with ETOH (control); cholecalciferol (neg-tive control [1 �M]); 1,25(OH)2D3 (positive control [0.5 �M]); 25(OH)D3

t 0.5 or 1.0 �M and cell number determined 5 days post-treatment. Percentrowth inhibition is presented.

tise

�,25(OH)2D3 (0.5 �M), D3 (1 �M), 25(OH)D3 (1.0 �M) for 24 h. Both. (C) SW480 cells were plated and treated with control or 25(OH)3 (1.0 �M)ow evidence that it remains elevated at 72 h following start of treatment.

he hypercalcemic nature of 1�,25(OH)2D3, many relativelyon-calcemic analogs of vitamin D have been synthesizednd evaluated for their efficacy as anti-cancer agents. Ideally,he pro-hormone 25(OH)D3, which is naturally occurring andon-toxic could be developed, however, due to the traditionaloncept of renal hydroxylation of the pro-hormone, untilecently enthusiasm for establishing it as a cancer chemo-rotective agent had been non-significant. In recent years,he expression of 1�(OH)ase, the enzyme that catalyzes thisonversion, has been reported in several extra-renal tissues,ncluding the pancreas, prostate and colon [12] leading tohe hypothesis that 25(OH)D3, may possess chemopreventivefficacy in the organs expressing endogenous 1�(OH)ase. In

he colon, it has been reported that 1�(OH)ase is presentn normal, and malignant colon tumors [3,4]. Results alsohowed that high- to moderately differentiated colon tumorsxpress increased 1�(OH)ase, while poorly differentiated

mistry

tphesiwcsct1Nuoshp

A

o

R

[

[

[

1�-hydroxylase and implications for chemoprevention and treatment,J. Steroid Biochem. Mol. Biol. 97 (2005) 103–109.

[13] K.Z. Guyton, T.W. Kensler, G.H. Posner, Vitamin D and vitamin

G. Murillo et al. / Journal of Steroid Bioche

umors consistently have shown reduced expression or com-lete loss of 1�(OH)ase activity [5]. More recently, weave demonstrated that 1�(OH)ase and VDR are also highlyxpressed in the pre-cancerous lesions (ACF) of human colonamples [10]. Anti-proliferative effects of vitamin D analogsn colon cancer cells have been reported [13]. In this report,e evaluated effects of 25(OH)D3 in several human colon

ancer cell lines. Interestingly 25(OH)D3 enhanced expres-ion of both VDR and 1�(OH)ase in these cells. It is notlear whether 25(OH)D3 directly suppresses cell prolifera-ion of colon cancer cells in vitro or even if its conversion to�,25(OH)2D3 is required for its anti-proliferative actions.onetheless, these results provide the basis for further eval-ation of 25(OH)D3 as a candidate agent for the treatmentf colon cancer. Furthermore, these findings support the pos-ible therapeutic use of 25(OH)D3 in individuals who are atigh risk for developing colon cancer, including those withrior history of colonic polyps or with large numbers of ACFs.

cknowledgement

This work was supported by NCI Mentored Career Devel-pment Award Grant K01CA103861.

eferences

[1] D.M. Harris, V.L. Go, Vitamin D and colon carcinogenesis, J. Nutr. 134(2004) 3463S–3471S.

[2] E.D. Gorham, C.F. Garland, F.C. Garland, W.B. Grant, S.B. Mohr, M.Lipkin, H.L. Newmark, E. Giovannucci, M. Wei, M.F. Holick, Vitamin

D and prevention of colorectal cancer, J. Steroid Biochem. Mol. Biol.97 (2005) 179–194.

[3] H.S. Cross, G. Bises, D. Lechner, T. Manhart, E. Kallay, The vitaminD endocrine system of the gut—its possible role in colorectal cancerprevention, J. Steroid Biochem. Mol. Biol. 97 (2005) 121–128.

& Molecular Biology 103 (2007) 763–767 767

[4] V. Tangpricha, J.N. Flanagan, L.W. Whitlatch, C.C. Tseng, T.C.Chen, P.R. Holt, M.S. Lipkin, M.F. Holick, 25-Hydroxyvitamin D-1�-hydroxylase in normal and malignant colon tissue, Lancet 357 (2001)1673–1674.

[5] G. Bises, E. Kallay, T. Weiland, F. Wrba, E. Wenzl, E. Bonner, S. Kri-wanek, P. Obrist, H.S. Cross, 25-Hydroxyvitamin D3-1�-hydroxylaseexpression in normal and malignant human colon, J. Histochem.Cytochem. 52 (2004) 985–989.

[6] G.G. Schwartz, D. Eads, A. Rao, S.D. Cramer, M.C. Willingham,T.C. Chen, D.P. Jamieson, L. Wang, K.L. Burnstein, M.F. Holick,C. Koumenis, Pancreatic cancer cells express 25-hydroxyvitaminD-1�-hydroxylase and their proliferation is inhibited by the pro-hormone 25-hydroxyvitamin D3, Carcinogenesis 25 (2004) 1015–1026.

[7] P. Bareis, G. Bises, M.G. Bischof, H.S. Cross, M. Peterlik, 25-Hydroxy-vitamin D metabolism in human cancer cells during tumor progression,Biochem. Biophys. Res. Commun. 285 (2001) 1012–1027.

[8] R.E. Carroll, K.A. Matkowskyj, S. Chakrabarti, T.J. McDonald, R.V.Benya, Aberrant expression of gastrin-releasing peptide and its receptorby well differentiated colon cancers in humans, Am. J. Physiol. 276(1999) G655–G665.

[9] L.L. Vindelov, I.J. Christensen, N.I. Nissen, A detergent-trypsin methodfor the preparation of nuclei for flow cytometric DNA analysis, Cytom-etry 3 (1983) 323–327.

10] D. Matusiak, G. Murillo, R.E. Carroll, R.G. Mehta, R.V. Benya, Expres-sion of vitamin D receptor and 25-hydroxyvitamin D3-1�-hydroxylasein normal and malignant human colon, Cancer Epidemiol. BiomarkersPrev. 14 (2005) 2370–2376.

11] H.G. Palmer, J.M. Gonzalez-Sancho, J. Espada, M.T. Berciano, I. Puig,J. Baulida, M. Quintanilla, A. Cano, A. Garcia de Herreros, M. Lafarga,A. Munoz, Vitamin D promotes the differentiation of colon carcinomacells by the induction of E-cadherin and the inhibition of �-cateninsignaling, J. Cell Biol. 154 (2001) 369–387.

12] K. Townsend, K.N. Evans, M.J. Campbell, K.W. Colston, J.S. Adams,M. Hewison, Biological actions of extra-renal 25-hydroxyvitamin D-

D analogs as cancer chemopreventive agents, Nutr. Rev. 61 (2003)227–238.