Acquisition of Resistance to Butyrate Enhances Survival After Stress and Induces Malignancy of Human Colon Carcinoma Cells

[CANCER RESEARCH 64, 4593–4600, July 1, 2004] Acquisition of Resistance to Butyrate Enhances Survival after Stress and Induces Malignancy of Human Colon Carcinoma Cells

Isabel Lo´pez de Silanes,1,3 Nieves Olmo,1 Javier Turnay,1 Gonzalo Gonza´lez de Buitrago,2 Pablo Pe´rez-Ramos,1 Ana Guzmań-Arańguez,1 Marta Garcıá-Dıéz,1 Emilio Lecona,1 Myriam Gorospe,3 and M. Antonia Lizarbe1 1Departamento de Bioquõ«mica y Biologõ«a Molecular, Facultad de Ciencias Quõ«micas, Universidad Complutense, Madrid, Spain; 2Departamento de Inmunologõ«a y Oncologõ«a, Centro Nacional de Biotecnologõ«a, Consejo Superior de Investigaciones Cientõ«ficas, Madrid, Spain; and 3Laboratory of Cellular and Molecular Biology, National Institute on AgingÐIntramural Research Program, National Institutes of Health, Baltimore, Maryland

ABSTRACT proposed to be a consequence of histone hyperacetylation because such modifications facilitate endonuclease access to chromatin and Acquired resistance to apoptosis by tumor cells remains a major ob- induce or repress the expression of genes encoding proteins such as stacle for cancer treatment, and hence the analysis of resistance to apo- caspase-3 and Bcl-2 that are involved in cell death programs (7–9, 13, ptosis constitutes a major goal in the development of antitumoral drugs. We have established a butyrate-resistant human colon adenocarcinoma 14). Other butyrate-driven modifications altering gene expression cell line (BCS-TC2.BR2) from nontumorigenic BCS-TC2 cells to analyze include alterations in histone phosphorylation, DNA methylation, whether the acquisition of such phenotype confers resistance to apoptosis transcriptional activation of gene promoters bearing butyrate-response and stress. Although BCS-TC2.BR2 cells exhibited a more differentiated elements, and the expression of genes downstream of the ␤-catenin phenotype than the parental BCS-TC2 cells, higher butyrate concentra- and Tcf-signaling pathways (4, 15, 16). tions remained capable of additionally enhancing their differentiation Because of its cytostatic and cytotoxic effects on tumor cells, there without inducing apoptosis. Survival rates of BCS-TC2.BR2 cells after has been much interest in exploring the use of butyrate and its glucose deprivation and heat shock were higher than those of parental analogues as anticancer drugs (9, 17, 18). In addition, butyrate has cells, revealing a stress-resistant phenotype. These findings were accom- been used to sensitize multidrug-resistant tumor cell lines (19), al- panied by key differences between parental and butyrate-resistant cells in though in certain instances, treatment with butyrate might itself in- gene expression profiles and the acquisition of in vivo tumorigenicity. In conclusion, cells gaining resistance to an endogenous physiological mod- crease the multidrug-resistant phenotype (20, 21). ulator of growth, differentiation, and apoptosis concurrently acquired Increased proliferation and decreased apoptosis are key features of resistance to other agents that influence cell survival. neoplasia. In the normal colonic epithelium, butyrate is a major player in the regulation of these processes. However, some cells can escape from butyrate-triggered apoptosis and become resistant to this agent, INTRODUCTION as evidenced by the existance of butyrate-resistant cell lines (21–24). Therefore, the identification of the mechanisms responsible for the A growing number of epidemiological studies suggest that a fiber- acquisition of resistance to butyrate-triggered apoptosis is of great rich diet exerts a protective role against colorectal cancer develop- potential value because it could reveal additional targets for therapeu- ment, the third most common cancer and the second leading cause of tic intervention and new markers for colorectal cancer detection and cancer deaths. It is generally accepted that the end-products of colonic diagnosis. fiber fermentation, particularly butyrate, play an important role in In the present study, we have established butyrate-resistant cells protection against this disease. Butyrate is essential for maintaining from a poorly differentiated, nontumorigenic human colon adenocar- the physical and functional integrity of the intestinal mucosa through cinoma cell line (BCS-TC2) (25) to analyze their biological properties its stimulatory influence on cell proliferation, differentiation, cation and their resistance to chemical and environmental stress. The degree absorption, and colonic blood flow (1–4). Butyrate increases prolif- of induction of apoptosis after exposure to different types of stresses eration of nondifferentiated colonocytes from the crypt base, where was found to be markedly elevated in the parental cells, suggesting the butyrate concentration is low, whereas it induces differentiation and existence of apoptosis-resistant mechanisms in the butyrate-resistant apoptosis of cells at the neck and top of the crypts, exposed to the cells. Importantly, this increase in apoptosis resistance was accompa- higher butyrate concentrations of the intestinal lumen (5). Alterations nied by the acquisition of a malignant phenotype in vivo. in butyrate metabolism can lead to inflammatory bowel disease, atrophic colonic epithelium, and colorectal cancer. Butyrate inhibits cell proliferation and induces differentiation and MATERIALS AND METHODS first apoptosis in transformed cells from colon and other tissues (5–7). It first imposes cell cycle arrest (8, 9), afterward triggering a differ- Cell Culture, Proliferation, and Treatment. Human colon adenocarci- entiation program reflected in changes in specific markers such as noma cells were cultured in DMEM (Sigma, Alcobendas, Spain) supplemented alkaline phosphatase (ALP) activity, which can subsequently induce with 5% FCS (Bio-Whittaker, Verviers, Belgium). Cell proliferation and cells to undergo apoptosis. Butyrate strongly influences transcrip- viability were measured as described previously (26). Spheroids were obtained by seeding the cells on nonadherent dishes coated with 1% (w/v) agar (Difco, tional activity, causing extensive modifications in gene expression Detroit, MI), their growth monitored using an ocular with a grid (Olimpus, patterns (10, 11). Although the precise mechanisms underlying such Melville, NY), and their size calculated according to (a ϫ b)1/2, where a and alterations have not been fully elucidated, the inhibitory effect of b are the largest diameter and its perpendicular diameter, respectively. Heat butyrate on histone deacetylase activity appears to be a major influ- shock was performed at 42°C for the indicated times. Glucose-free DMEM encing factor (3, 12). Butyrate-triggered apoptosis has also been was from Biochrom AG (Berlin, Germany). Butyrate (Sigma) was prepared in standard medium and used at the indicated final concentrations. After estab- Received 2/26/04; revised 4/30/04; accepted 5/13/04. lishment of the butyrate-resistant cell line BCS-TC2.BR2, cultures were rou- Grant support: Direccioń General de Ensenãnza Superior Grant PM98-0083 and tinely maintained in standard growth medium in the presence of 2 mM butyrate. Direccioń General de Investigacioń Grant BMC2002-01407. Unless otherwise specified, all experiments comparing parental and butyrate- The costs of publication of this article were defrayed in part by the payment of page resistant cells were performed after trypsinization and reseeding of the cells charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. without butyrate. Plating densities were chosen to achieve exponential cell Requests for reprints: Maria Antonia Lizarbe. E-mail: [email protected]. growth by 2–3 days in culture, whereupon experiments were performed. 4593

Preparation of Cell Lysates and Western Blot Analysis. Cell lysates viability, causing the detachment of cells from the monolayer, and were prepared as previously described (27) and protein content determined induced profound morphological changes, as evidenced by the pres- using the DC Protein Assay (Bio-Rad, Madrid, Spain). For Western blotting, ence of multinuclear cells and by increases in cell volume and vacu- monoclonal antibodies recognizing the 70-kDa constitutive (HSC70) and in- olization (Fig. 1A). The decrease in cell viability was a consequence ducible (HSP70) heat-shock proteins (Stressgen, Victoria, British Columbia, of the induction of apoptosis, as revealed by the detection of internu- Canada), p21 (Neomarkers, Fremont, CA), or vinculin (Sigma) were used. cleosomal DNA fragmentation (Fig. 1B), the exposure of phosphati- Detection was performed using horseradish peroxidase-labeled secondary antibodies and enhanced chemiluminescence (Amersham-Pharmacia-Biotech, dylserine (Fig. 1C), and the dispersion and loss of mitochondrial Buckinghamshire, United Kingdom). Vinculin was used as loading control for membrane potential (Fig. 1D). All of these changes documented a Western blotting data, and densitometric analysis was performed using a time-dependent induction of apoptosis in BCS-TC2 cells. However, photodocumentation system (UVItec, Cambridge, United Kingdom) and the after treatment with butyrate for 30 days (Fig. 1A), the number of cells UVIBand V.97 software (28). released from the monolayer began to decrease, and the remaining Flow Cytometry, DNA Fragmentation, and Caspase and ALP Activity cells retained high levels of viability and eventually resumed prolif- Assays. For flow cytometry, 106 cells resuspended in 10 mM HEPES (pH 7.4), eration. The resulting population was initially heterogeneous, display- containing 150 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 1.8 mM CaCl2 were ing both large and small cells. After reaching confluence, the cells incubated for 15 min in the presence of 2 ␮g/ml annexin A5-FITC (25) or 1 were subcultured in the presence of 2 mM butyrate and cultured for 30 ␮g/ml rhodamine 123. To discern necrotic cells, 0.005% (w/v) propidium additional days under these conditions. From this time on, the cul- iodide was added. Analyses were performed in a FACScan cytometer (Becton- Dickinson, San Jose, CA), as previously reported (29). ture was composed of a homogeneous population of small cells with DNA laddering was assessed as described previously (6). Caspase activity stable morphology and growth rates, displaying no significant DNA was measured in cell lysates obtained as previously described (27) using 10 ␮M fragmentation (Fig. 1B) or other apoptotic features. The resulting of either acetyl-DEVD- or acetyl-LEHD-7-amino-4-methyl-coumarin (Calbio- butyrate-resistant cell population presented cytometric characteristics chem, Darmstadt, Germany) as substrates to detect caspase-3 and caspase-9 similar to those of the parental cells regarding membrane asymmetry activities, respectively. ALP activity was measured with the ALP 10 kit and mitochondrial membrane potential (Fig. 1, C and D). The new (Sigma) using 20 ␮l of cell lysates. Absorbance at 405 nm was monitored butyrate-resistant cells maintained these characteristics along subse- using a DU 640 spectrophotometer (Beckman Coulter, Inc., Fullerton, CA). quent subculture in the presence of 2 mM butyrate and were then Transmission Electron Microscopy. BCS-TC2 and BCS-TC2.BR2 cells deemed an established cell line and named BCS-TC2.BR2. BCS- were seeded without butyrate and cultured for 1 week until near-confluence. TC2.BR2 cells were routinely cultured in the presence of 2 mM Cells were then fixed in situ with 0.2% (v/v) glutaraldehyde in PBS for 30 min at 4°C. After scraping and centrifugation, cell pellets were resuspended in 0.1 butyrate, and their overall characteristics remained stable. BCS- TC2.BR2 cells cultured in standard growth medium without butyrate M glucose in PBS and postfixed in 1% (w/v) OsO4 for 30 min, dehydrated with ethanol, and embedded in Epon 812 resin. Ultrathin sections were contrasted retained apoptotic-resistant features as shown in routine monthly tests with uranyl acetate and lead citrate and visualized in a JEOL 101-O electron- that verified their continued ability to grow in the presence of 2 mM microscope at 80 kV. butyrate. cDNA Array Analyses. Total RNA was prepared from each established Characterization of Butyrate-Resistant BCS-TC2.BR2 Cells. cell line growing under standard culture conditions in the absence of butyrate First, we compared the proliferation rates of butyrate-resistant BCS- and were collected at ϳ80% confluence. Cells were lysed, and total RNA was TC2.BR2 cells with those of parental BCS-TC2 cells in the absence or isolated using the ToTALLY RNA kit (Ambion, Austin, TX). RNA samples presence of butyrate (Fig. 2A). The doubling time for BCS-TC2 cells were reverse transcribed in the presence of [␣-33P]dCTP and the radiolabeled was moderately long (35–40 h) in the absence of butyrate, but this products used to hybridize cDNA arrays (MGC arrays, 9600 genes),4 using previously reported methodologies (30, 31). Z ratios were considered signifi- agent dramatically decreased their growth rate. By contrast, BCS- cant when Ն ϩ1.5 or Յ Ϫ1.5; only Z scores from BCS-TC2 and BCS- TC2.BR2 cells presented a 1.6-fold shorter doubling time than the TC2.BR2 cells in which average was Ն0 were included in the analysis. parental cells, and the addition of butyrate affected proliferation only Significant values were also tested using a two-tailed test (Z values Ն 2.5; Z slightly. In this regard, expression of the cell cycle inhibitor p21 after value is equal to Z diff divided by the SD of each Z diff) and P Յ 0.01. The butyrate treatment was found to increase in both cell lines, but this data reflect three independent experiments. elevation was more pronounced in the parental cells, likely contrib- Tumor Development in Nude Mice. Athymic nude male mice (8 weeks uting to their slower growth after butyrate treatment (Fig. 2B). Cleav- old) were from Harlan Ibe´rica (Barcelona, Spain). In vivo tumorigenicity of age of p21 in the parental cells was detected after treatment with 4 and BCS-TC2 and BCS-TC2.BR2 cells was assessed by s.c. inoculation of 106 8mM butyrate. cells in 0.2 ml of DMEM in the lumbar region. Experiments were carried out The morphological characteristics of the butyrate-resistant cells in triplicate with n ϭ 12 for each cell line. Mice were surveyed daily and measured twice weekly after tumors became visible, as previously reported were examined by transmission electron microscopy (Fig. 2C). Com- (29). Tumor size was calculated as the average of the longest diameter and its pared with the parental cells (Fig. 2C, top panel), BCS-TC2.BR2 cells perpendicular diameter, as measured with a Vernier caliper. exhibited features of enhanced cellular differentiation. Some cells Statistical Analysis. The differences between the mean values were ana- showed cellular polarization with an apical pole and a basolateral lyzed with SigmaPlot version 8.02 (SPSS, Chicago, IL) and using Student’s t membrane (Fig. 2C, a) but without the characteristic epithelial brush- test. Statistical significance was considered to be achieved at the P Ͻ 0.05 border. They also displayed intracellular cysts (Fig. 2C, b) and in- level. creased intermediate filaments and desmosomes (Fig. 2C, c and d). The presence or absence of 2 mM butyrate in the growth medium of RESULTS BCS-TC2.BR2 cells did not modify the morphological and ultrastruc- tural characteristics of these cells. Establishment of Butyrate-Resistant BCS-TC2.BR2 Cells from A well-established colon differentiation marker, ALP activity, is the Human Colon Adenocarcinoma BCS-TC2 Cell Line. BCS- higher and increases with cell proliferation in differentiated cells. TC2 cells were cultured in continuous presence of 2 mM butyrate, and Basal ALP activity in BCS-TC2.BR2 cells, determined in the absence changes in cell morphology and viability were monitored (Fig. 1). of butyrate, was found to be 2.9-fold higher than in BCS-TC2 cells, Treatment of parental BCS-TC2 cells with butyrate decreased cell increasing 14-fold after being cultured for up to 11 days (Fig. 2D). Taken together, these results revealed a higher degree of differentia- 4 Internet address: http://www.grc.nia.nih.gov/branches/rrb/dna/index/dnapubs.htm#2. tion of BCS-TC2.BR2 cells compared with the parental cells. 4594

Fig. 1. Establishment of butyrate-resistant BCS-TC2.BR2 cells. A, representative micrographs of either untreated BCS- TC2 cells (top left) or BCS-TC2 cells that were continuously treated with 2 mM butyrate for up to 60 days. Increasing vacu- olarization (arrows) or increasing cellular volume (arrowheads) are indicated. Magnification: ϫ100 (10-day micrograph: ϫ200). B, internucleosomal DNA fragmentation analysis after treatment of BSC-TC2 cells with 2 mM butyrate for the indicated times. C, apoptosis in BCS-TC2 cells was monitored throughout butyrate treatment by flow cytometry using annexin A5-FITC; percentages of apoptotic cells are shown in the bot- tom right of each quadrant. D, changes in mitochondrial membrane potential were monitored using rhodamine 123.

Effect of Butyrate on BCS-TC2.BR2 Cell Differentiation and ods.” Of the 126 genes that showed the most significant differences in Apoptosis. It was important to assess whether BCS-TC2.BR2 cells expression levels, 54 were up-regulated and 72 down-regulated in remained responsive to higher butyrate concentrations regarding cell BCS-TC2.BR2 cells relative to BCS-TC2 cells (Table 1). The com- differentiation and induction of apoptosis. We first analyzed the plete array analysis is available online.5 influence of 4 and 8 mM butyrate on ALP activity (Fig. 3A). Butyrate A selection of the genes that exhibited the greatest differences in increased ALP activity in both cell types in a time- and concentration- expression levels in butyrate-resistant cells, as well as those associated dependent manner, but BCS-TC2.BR2 cells consistently displayed with the effects of butyrate, resistance to butyrate or apoptosis, and higher ALP activity (after 4 days in 8 mM butyrate, ALP activity was colon tumorigenesis, are shown in Table 2. Interestingly, additional 9-fold higher in BCS-TC2.BR2 cells and 33-fold higher in BCS-TC2 genes involved in glucose metabolism such as glyceraldehyde-3- cells). phosphate dehydrogenase, aldolase A, and pyruvate kinase were We detected a significant number of floating cells in butyrate- found to be down-regulated in BCS-TC2.BR2 cells. treated BCS-TC2 cultures only, possibly indicating the activation of The different gene expression profiles and the strikingly distinct apoptotic pathways in these cells (6). To confirm this hypothesis, we apoptotic responses after butyrate treatment supported the notion that evaluated DNA fragmentation and caspase-3 activity after butyrate BCS-TC2.BR2 cells might have a broad resistance against different treatment. As expected, increasing concentrations of butyrate markedly enhanced apoptosis in BCS-TC2 cells, as revealed by the pres- stresses, and thus, we set out to test this possibility directly. ence of DNA fragmentation products (Fig. 3B) and the increase in Effect of Glucose Deprivation on BCS-TC2 and BCS-TC2.BR2 caspase-3 activity in floating cells (Fig. 3C). By contrast, BCS- Cells. To analyze the response of these cells to a metabolic stress, we TC2.BR2 cells showed neither DNA fragmentation nor increased cultured the cells in glucose-free medium. Glucose depletion caused a caspase-3 activity even at butyrate concentrations as high as 50 mM. loss of cell viability (Fig. 4A), but the toxic effect was more pro- Differences between Gene Expression Profiles in BCS- nounced in BCS-TC2 cells than in BCS-TC2.BR2 cells (45 versus TC2.BR2 and BCS-TC2 Cells. Additional characterization of the 90% viable cells after 24 h, respectively). This reduction in cell butyrate-resistant cells was carried out using cDNA array analysis. viability was caused by the induction of apoptosis as shown in Fig. Genes were considered differentially expressed if the change in expression fulfilled the four criteria outlined in “Materials and Meth- 5 Internet address: http://www.grc.nia.nih.gov/branches/rrb/dna/index/dnapubs.htm. 4595

Fig. 2. Characterization of butyrate-resistant BCS- TC.BR2 cells. A, proliferation curves of BCS-TC2 and BCS-TC.BR2 cells in the absence (Control) or presence of either 2 or 4 mM butyrate. Data represent the mean values of normalized absorptions at 570 nm; error bars,SDof eight different measurements from three independent experiments. B, Western blot analysis of p21 expression in BCS-TC2 and BCS-TC2.BR2 cells after butyrate treatment. Cells were seeded without butyrate and cultured for 3 days, whereupon fresh butyrate-containing medium was added, and cells were cultured for an additional 24 h before lysis for Western blot analysis. Vinculin signals (Vin), obtained after stripping of the membranes and reprobing, demonstrate the evenness in loading and transfer of the protein samples. C, transmission electron micrographs of BCS- TC2.BR2 cells showing differentiation features: a, colum- nar cells; b, intracellular cysts; c, intermediate filaments; and d, desmosomes. A micrograph of BCS-TC2 cells maintained under identical culture conditions is shown in the top panel (bars, 2.5 ␮m). D, basal and proliferation-associated alkaline phosphatase (ALP) activity in BCS-TC2 and BCS- TC2.BR2 cells in the absence of butyrate. Data represent mean values Ϯ SD.

4B. Whereas the percentage of apoptotic cells, determined by flow at 42°C for different periods of time (Fig. 5C) revealed marked cytometry using annexin A5-FITC, rose significantly in BCS-TC2 differences in HSP70 basal levels (2.2-fold higher in BCS-TC2.BR2 cells after glucose deprivation (2.3- and 4.8-fold after 12 or 24 h, cells) and distinct patterns of induction: BCS-TC2.BR2 cells dis- respectively), only a small increase was detected in the butyrate- played a time-dependent rise (up to 2.5-fold) in HSP70 expression resistant cells. Additionally, no significant caspase-9 or caspase-3 levels, whereas parental cells exhibited maximum HSP70 levels after activation was observed in BCS-TC2.BR2 cells after 24 h of glucose 4–6 h of treatment (4.9-fold increase), decreasing thereafter but deprivation, but a significant caspase-9 activation was detected (2.4- maintaining 2.5-fold higher levels by 10 h in continuous heat shock. fold) in the parental cells (Fig. 4C). BCS-TC2 and BCS-TC2.BR2 Spheroids and Effect of Butyrate. Effect of Heat Shock on BCS-TC2 and BCS-TC2.BR2 Cells. Culture of adherent cells on nonadherent surfaces stimulates the Additional characterization of the stress-resistant phenotype of BCS- establishment of cell-cell contacts and the three-dimensional growth TC2.BR2 cells was carried out by studying their response to heat of cells as multicellular spheroids. Spheroids formed from both BCS- shock. Cells were incubated at 42°C for up to 24 h, whereupon TC2 and BCS-TC2.BR2 cells cultured on agar, but the behavior of cellular viability was measured (Fig. 5A). In both cell types, heat BCS-TC2.BR2 spheroids and BCS-TC2 spheroids was markedly dif- shock initially decreased cell viability (by 25% after 5 h), but no ferent upon subculture onto adherent surfaces. First, although BCS- additional decline was detected in BCS-TC2.BR2 cells for up to 24 h, TC2.BR2 spheroids were capable of maintaining growth in suspen- whereas cell viability decreased down to Ͻ50% in BCS-TC2. Anal- sion, BCS-TC2 spheroids attached to the culture plate surface, and ysis of the expression of the constitutive (HSC70) and the inducible cells gradually returned to growth in monolayer (Fig. 6A). Second, (HSP70) 70-kDa HSPs before (Fig. 5B) and after incubation of cells within the spheroids, the influence of butyrate treatment on cell 4596

reaching sizes of ϳ0.5–0.6 cm by 30 days after inoculation (Fig. 7) with a 100% tumor growth incidence (12 of 12). In contrast, no tumor growth was observed in animals injected with BCS-TC2 cells, in agreement with previous data (29). These findings demonstrated that the butyrate-resistant BCS-TC2.BR2 cells have a tumorigenic phenotype.

DISCUSSION Homeostasis of the normal colonic mucosa relies on a tightly controlled balance between cell proliferation, differentiation, and apoptosis. Butyrate, an end-product of colonic fiber fermentation, is a major natural regulator of these processes (1, 3), exerting a trophic effect on nondifferentiated cells at the base of the colonic crypts while promoting differentiation and subsequently apoptosis of cells at the top of the crypts (5). However, even under normal physiological conditions, some cells can escape this regulation and become resistant to the induction of apoptosis by butyrate. Evidence of progressive inhibition of apoptosis during the adenoma-to-carcinoma progression stages supports this idea (32). In vitro studies using different colorectal cell lines have demonstrated that butyrate is able to inhibit cell proliferation, induce cell differentiation and apoptosis, and enhance immunosurveillance and anti-inflammatory responses (6, 7, 13, 22, 33). However, the molecular bases of the chemopreventive effects of butyrate remain largely unknown. Moreover, the resistance to butyrate has been analyzed in vitro in cells from different sources, but only few actual butyrate- resistant cell lines have been established, and in some cases, these cells were found to revert their phenotype after butyrate removal (21–24). Here, we describe the establishment of the butyrate-resistant BCS-TC2.BR2 cell line that was derived from a butyrate-sensitive human colon adenocarcinoma cell line (BCS-TC2; Ref. 6) through continuous culture in the presence of 2 mM butyrate. How can cells acquire this type of resistant phenotype? Mariadason et al. (22) have suggested a correlation between the degree of differentiation along the absorptive lineage of colonic epithelial cells and the resistance to butyrate. Caco-2 and HT29cl.19A cells, which differentiated spontaneously with time in culture, showed increased

Table 1 Classification of genes differentially expressed between BCS-TC2.BR2 and BCS-TC2 cells Up-regulated genes Down-regulated genes

Fig. 3. Effect of butyrate on BCS-TC2.BR2 cell differentiation and apoptosis. Cells Z ratio Z ratio were seeded without butyrate and cultured for 3 days, whereupon fresh medium contain- a Ն a Յ Ϫ ing the indicated butyrate concentrations was added. A, alkaline phosphatase (ALP) Biological function Total 2 Total 2 activity in cells that were incubated in the absence or presence of 4 or 8 mM butyrate for A. Cell cycle control and proliferation 7 6 5 2 P Ͻ 0.01). B, internucleosomal DNA fragmentation analysis after B. Apoptosis and stress response 2 1 6 2 ,ءء) either 2 or 4 days treatment for 3 days with the butyrate concentrations shown. C, caspase-3 activity of C. Protein biosynthesis, folding and 73 6 0 BCS-TC2 and BCS-TC2.BR2 cells (a.u., arbitrary units) after 3 days of culture in the degradation presence of butyrate (0–50 mM). Data represent mean values Ϯ SD. D. Transcriptional and post- 73 8 2 transcriptional processes E. Signal reception and transduction 7 2 4 2 F. Metabolic pathways 5 1 7 2 differentiation also differed between the two cell lines. Whereas ALP G. DNA replication and repair 2 1 0 0 H. Ion transport and physiology, and 4 1 11 6 activity in BCS-TC2 spheroids was similar to what was observed in intracellular membrane trafficking cells growing in monolayer, reduced ALP activity was detected in I. Cytoskeleton structure and mobility. 30 4 1 BCS-TC2.BR2 spheroids (Fig. 6B). Cell-matrix interactions J. Differentiation and development 1 0 4 0 In Vivo Tumorigenicity of BCS-TC2.BR2 Cells in Nude Mice. K. Other functions 9 0 17 2 BCS-TC2.BR2 cells displayed important differences with the parental Total 54 18 72 19 cells regarding gene expression patterns and increased resistance to a Total RNA was obtained and processed as described in “Materials and Methods.” apoptosis after exposure to different forms of stress. Accordingly, Intensity data from the different genes were analyzed in both cell lines, and only those with significant differences according to the four criteria outlined in the “Materials and these cells might be predicted to exhibit a more malignant phenotype. Methods” section (Z ratio, Z value, average Z score, and P) were considered. Among these To investigate this possibility, the ability of BCS-TC2.BR2 and BCS- genes, those showing a Z ratio Ն2orՅϪ2 are also indicated. Genes without a clearly assigned biological function, as well as those coding hypothetical proteins, are grouped in TC2 cells to form s.c. tumors in nude mice was assessed. After 14 K. Each gene appears only in one group, although several biological functions may be days of latency, tumors from BCS-TC2.BR2 cells started to grow, assigned to some genes. 4597

Table 2 Partial list of genes exhibiting significantly different expression in BCS- cDNA array analysis (Table 2). Another possible route of acquisition TC2.BR2 compared to parental cells of the resistant phenotype is through the selection of a preexisting cell The table reflects only a partial selection of genes showing significant differences between both cell lines, ordered according to Z ratio values (29). subpopulation that either spontaneously differentiated to a butyrate- oxidizing phenotype or carried permanent alterations in their genome. Z ratio The selection of such a subpopulation during the establishment of Up- Down- Biological BCS-TC2.BR2 cells was quite plausible, particularly because BCS- a Gene product Symbol regulation regulation involvement TC2 parental cells are heterogeneous and have previously allowed the Annexin A1 ANXA1 3.12 b selection of a tumorigenic cell subpopulation from tumors induced Calmodulin 2 (phosphorylase CALM2 3.05 kinase ␦) after s.c. coinjection into nude mice of the nontumorigenic parental Nucleophosmin (nucleolar NPM1 2.94 a,b,c cells with specific extracellular matrix components (29, 34). phosphoprotein B23) EBNA1 binding protein 2 EBNA1BP2 2.72 The most prominent feature of BCS-TC2.BR2 cells is their overall Clusterin CLU 2.51 a,b resistance to stress-induced apoptosis. It is interesting to note that cells Heat-shock 70 kDa protein 8 HSPA8 2.43 a,b gaining resistance to an endogenous physiological modulator of Thyroid hormone receptor ␣ THRA 2.40 (oncogene ERBA1) growth, differentiation, and apoptosis were found to concurrently High mobility group AT-hook 1 HMGA1 2.19 c exhibit increased survival in response to other stress conditions such Proliferation-associated 2G4 PA2G4 2.19 as heat shock, glucose deprivation, or attachment-independent growth. Cytochrome c (somatic) HCS 1.94 Unactive progesterone receptor TEBP 1.90 b It is well known that HSPs confer protection to intestinal epithelial Zinc finger protein 266 ZNF266 1.89 cells and improve cell survival under a variety of stress conditions Asparagine synthetase ASNS 1.84 b Oncogene DJ-1 DJ1 1.84 b Ring finger protein 7 RNF7 1.76 b Sorcin SRI 1.75 b ATP-dependent DNA helicase II XRCC5 1.71 b Thymosin ␤4, X-linked TMSB4X 1.62 c Actin ␥ 1 ACTG1 1.55 Nucleoside-diphosphate kinase 1 NME1 1.52 c Zinc finger protein 174 ZNF174 Ϫ1.51 CDC28 protein kinase regulatory CKS2 Ϫ1.52 subunit 2 Cofilin 1 (nonmuscle) CFL1 Ϫ1.55 b,c Tumor protein, translationally TPT1 Ϫ1.57 b,c controlled 1 Calgizzarin S100A11 Ϫ1.66 c APR-1 protein MAGEH1 Ϫ1.70 b BCL2-related protein A1 BCL2A1 Ϫ1.71 b Hypoxia-inducible factor 1- RTP801 Ϫ1.73 a,b responsive RTP801 Glyceraldehyde-3-phosphate GAPD Ϫ1.75 a,c dehydrogenase Epithelial cell-transforming ECT2 Ϫ1.77 sequence 2 oncogene Aldolase A ALDOA Ϫ1.77 Fatty-acid CoA ligase, long- FACL5 Ϫ2.03 chain 5 Caspase recruitment domain CARD14 Ϫ2.04 b family, member 14 Inhibitor of DNA binding 3 ID3 Ϫ2.53 c Adrenomedullin ADM Ϫ2.57 b Stratifin SFN Ϫ3.13 b,c IFN-␣-inducible protein G1P2 Ϫ3.21 Pyruvate kinase, muscle 2 PKM2 Ϫ3.40 b S100A2 S100A2 Ϫ3.41 a,b,c Calbindin 1 CALB1 Ϫ3.44 a Thyroid hormone receptor TRIP10 Ϫ3.70 interactor 10 Calcyclin S100A6 Ϫ3.87 c a The biological involvement refers to genes whose expression changes with (a) butyrate treatment or resistance, (b) resistance to apoptosis and stress, and (c) colon tumorigenesis. resistance to butyrate because of their enhanced ability to oxidize it, thereby decreasing its intracellular levels. Consequently, histone deacetylase was not inhibited as potently as it was in the nonresistant undifferentiated Caco-2 cells, which showed histone hyperacetylation. Our data are in accordance with this model because BCS-TC2.BR2 cells not only showed a higher degree of differentiation than did the poorly differentiated parental cells but were also able to further differentiate either with time in culture or in the presence of higher Fig. 4. Effect of glucose deprivation in BCS-TC2 and BCS-TC2.BR2 cells. A, repre- ϩ - concentrations of butyrate. These cells were also able to grow under sentative micrographs of cells cultured in the presence (Glu ) or absence (Glu ) of glucose for 24 h; cellular viability is indicated in each panel. B, apoptosis detection after culture conditions of glucose deprivation probably because they can effi- in the presence or absence of glucose for 12 or 24 h, determined by flow cytometry using ciently use butyrate as an alternative energy source. This hypothesis is annexin A5-FITC. Data are normalized to the percentage of apoptotic cells in the Gluϩ controls. C, caspase-9 and caspase-3 activities in the presence or absence of glucose for supported by the fact that several genes linked to the glycolytic 24 h. Data correspond to mean values (ϮSD) from two independent experiments with .(P Ͻ 0.01 ,ءء ;P Ͻ 0.05 ,ء) pathway were down-regulated in BCS-TC2.BR2 cells according to triplicate samples 4598

Fig. 6. Effect of butyrate on BCS-TC2 and BCS-TC2.BR2 cell spheroids. A, representative micrographs showing multicellular spheroids growing on adherent surfaces. B, alkaline phosphatase (ALP) activity in monolayers (M) or spheroids (S) from cells treated with 4 mM butyrate for 7 days. Data correspond to mean values (ϮSD) from three independent experiments.

the acquisition of tumorigenic capacity, as revealed by the ability of BCS-TC2.BR2 cells to develop tumors after inoculation into nude Fig. 5. Influence of heat shock on BCS-TC2 and BCS-TC2.BR2 cells. A, cell viability mice. A study by Williams et al. (36) supports this notion by provid- after incubation at 42°C for up to 24 h (100% corresponds to ϳ1.5 ϫ 105 cells/cm2 in both cell lines). Western blot analysis of the expression of 70-kDa heat-shock cognate (HSC70) ing experimental evidence that during the adenoma3carcinoma pro- and heat-shock protein 70 (HSP70) in cell lysates: basal expression (B) and expression (C) gression sequence in vitro, butyrate contributed to the selection of after incubation for different time periods at 42°C. Vinculin signals are shown to demonstrate the evenness in loading and transfer of the protein samples (Vin). Protein cells with a more malignant phenotype. Butyrate did not directly expression (in a.u., arbitrary units) is normalized to the intensity values of untreated induce tumorigenicity in the butyrate-resistant cells, but it did increase BCS-TC2 cells. Data correspond to mean values (ϮSD) from three different experiments their susceptibility to transformation by carcinogens. These findings .P Ͻ 0.01). Representative Western blot analyses are shown ,ءء) are particularly interesting in light of increasing in vivo evidence of a progressive inhibition of apoptosis during malignant transformation. In this regard, in the adenoma3carcinoma transition model of colon through the stabilization of cellular components and processes (pro- carcinogenesis, tissue specimens of increasing malignancy showed teins, cytoskeleton, mitochondrial function, and so forth) or the inhi- decreased rates of apoptosis (32). bition of apoptotic pathways (35). Among these proteins, inducible The molecular events underlying the resistance of BCS-TC2.BR2 HSP70 has been shown to be strongly involved in cellular protection cells to apoptosis are likely to be complex, as suggested by the against stress. Heat shock induced the expression of this protein in extensive differences in gene expression profiles between the parental both cell lines, but this increase appeared to be insufficient for and the resistant cell lines. These changes in gene expression patterns maintaining cell viability in the parental cells probably due to their lower basal HSP70 levels. Moreover, BCS-TC2.BR2 cells are also resistant to metabolic stress induced by glucose starvation. Whereas the parental cells showed clear apoptotic features after glucose deprivation, including loss of membrane asymmetry and caspase-9 activation, butyrate-resistant cells were almost unaffected by these culture conditions. These results, taken together with the lack of caspase-3 activation even at high butyrate concentrations, suggest that BCS- TC2.BR2 cells presented an impairment or blockage in the apoptotic pathway upstream of the activation of effector caspases. In addition, when cultured on nonadherent substrates, butyrate-resistant cells were able to form more stable multicellular aggregates than parental cells. This type of growth additionally increased the resistance of these cells to the effects of butyrate, probably by restricting the accessibility of this agent into the inner cell layers of the spheroids, suggesting that cell-cell interactions between these cells were tighter than in the Fig. 7. In vivo tumorigenicity of BCS-TC2 and BCS-TC2.BR2 cells in nude mice. Tumorigenicity was assessed by s.c. injection of 106 cells in the lumbar region of nude parental cell aggregates. mice. Tumor size measurements were carried out with a caliper at the time points shown; Most notably, attaining a resistant phenotype was accompanied by data represent mean values (ϮSD). 4599

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2004 American Association for Cancer Research. ANTIAPOPTOTIC PHENOTYPE OF BUTYRATE-RESISTANT CELLS have been described after short-term butyrate treatment (10, 11, 37), 11. Della Ragione F, Criniti V, Della Pietra V, et al. Genes modulated by histone but to our knowledge, this is the first time that the gene expression acetylation as new effectors of butyrate activity. FEBS Lett 2001;499(4):199Ð204. 12. Davie JR. Inhibition of histone deacetylase activity by butyrate. J Nutr 2003;133(7): profiles of butyrate-resistant and butyrate-sensitive colon carcinoma 2485SÐ93S. cells have been systematically compared. Detailed examination of the 13. Litvak DA, Evers BM, Hwang KO, Hellmich MR, Ko TC, Townsend CM. Butyrate- multiple biochemical pathways that may contribute to the resistant induced differentiation of Caco-2 cells is associated with apoptosis and early induction of p21Waf1/Cip1 and p27Kip1. Surgery (St. Louis) 1998;124(2):161Ð9. phenotype of BCS-TC2.BR2 cells await future investigation. 14. Medina V, Edmonds B, Young GP, James R, Appleton S, Zalewski PD. Induction of Inhibition of apoptosis in tumors subjected to antineoplastic ther- caspase-3 protease activity and apoptosis by butyrate and trichostatin A (inhibitors of apies is recognized as a major problem in the outcome of patients with histone deacetylase): dependence on protein synthesis and synergy with a mitochondrial/cytochrome c-dependent pathway. Cancer Res 1997;57(17):3697Ð707. cancer. Therefore, the elucidation of novel and more specific apo- 15. Bordonaro M, Lazarova DL, Augenlicht LH, Sartorelli AC. Cell type- and promoter- ptotic triggers, as well as the identification of mechanisms of resist- dependent modulation of the Wnt signaling pathway by sodium butyrate. Int J Cancer ance to apoptosis, constitutes a major objective in the current efforts 2002;97(1):42Ð51. 16. Germann A, Dihlmann S, Hengerhahn M, Doeberitz MK, Koesters R. Expression to develop antitumoral drugs (38). Here, we have developed butyrate- profiling of CC531 colon carcinoma cells reveals similar regulation of beta-catenin resistant cells from the butyrate-sensitive human colon adenocarci- target genes by both butyrate and aspirin. Int J Cancer 2003;106(2):187Ð97. noma BCS-TC2 cell line. Butyrate-resistant BCS-TC2.BR2 cells grew 17. Patnaik A, Rowinsky EK, Villalona MA, et al. A Phase I study of pivaloyloxymethyl butyrate, a prodrug of the differentiating agent butyric acid, in patients with advanced regularly in the presence of 2 mM butyrate, exhibited a higher degree solid malignancies. Clin Cancer Res 2002;8(7):2142Ð8. of differentiation than did parental cells, and butyrate concentrations 18. Edelman MJ, Bauer K, Khanwani S, et al. Clinical and pharmacologic study of tributyrin: an oral butyrate prodrug. Cancer Chemother Pharmacol 2003;51(5): of up to 50 mM additionally enhanced their differentiation without 439Ð44. inducing apoptosis. Gene expression profiling provided additional 19. Shack S, Miller A, Liu L, Prasanna P, Thibault A, Samid D. Vulnerability of support for the notion that BCS-TC2.BR2 cells were overall more multidrug-resistant tumor cells to the aromatic fatty acids phenylacetate and phenyl- butyrate. Clin Cancer Res 1996;2(5):865Ð72. resistant to apoptosis induced by exogenous stimuli. The higher sur- 20. Vilaboa NE, Galan A, Troyano A, de Blas E, Aller P. Regulation of multidrug vival rates of BCS-TC2.BR2 cells observed after glucose deprivation resistance 1 (MDR1)/P-glycoprotein gene expression and activity by heat-shock and heat-shock treatment corroborated this hypothesis. These find- transcription factor 1 (HSF1). J Biol Chem 2000;275(32):24970Ð6. 21. Ho SB, Yan P, Dahiya R, Neuschwander-Tetri BA, Basbaum C, Kim YS. Stable ings, taken together with the observation that BCS-TC2.BR2 cells differentiation of a human colon adenocarcinoma cell line by sodium butyrate is displayed an enhanced ability to grow without attachment to a sub- associated with multidrug resistance. J Cell Physiol 1994;160(2):213Ð26. strate and were capable of forming tumors after s.c. inoculation into 22. Mariadason JM, Velcich A, Wilson AJ, Augenlicht LH, Gibson PR. Resistance to butyrate-induced cell differentiation and apoptosis during spontaneous Caco-2 cell nude mice, led us to propose that these cells had acquired both a differentiation. Gastroenterology 2001;120(4):889Ð99. stress-resistant phenotype and a more malignant phenotype. The cell 23. Derjuga A, Richard C, Crosato M, et al. Expression of p21Waf1/Cip1 and cyclin D1 is model herein reported provides a useful tool for dissecting and further increased in butyrate-resistant HeLa cells. J Biol Chem 2001;276(41):37815Ð20. 24. Chai F, Evdokiou A, Young GP, Zalewski PD. Involvement of p21(Waf1/Cip1) and investigating the mechanisms of resistance to apoptosis, as well as to its cleavage by DEVD-caspase during apoptosis of colorectal cancer cells induced by help devise rational strategies to specifically target butyrate-resistant butyrate. Carcinogenesis (Lond) 2000;21(1):7Ð14. cancers. 25. Turnay J, Olmo N, Gavilanes JG, Benõ«tez J, Lizarbe MA. Establishment and characterization of a new human colon-adenocarcinoma cell line: BCS-TC2. Cytotech- nology 1990;3(1):75Ð88. ACKNOWLEDGMENTS 26. Turnay J, Olmo N, Jime«nez A, Lizarbe MA, Gavilanes JG. Kinetic study of the cytotoxic effect of alpha-sarcin, a ribosome inactivating protein from Aspergillus We thank Dr. Alberto A´ lvarez and Agustõ«n Ferna«ndez for their assistance giganteus, on tumour cell lines: protein biosynthesis inhibition and cell binding. Mol Cell Biochem 1993;122(1):39Ð47. with the flow cytometry and electron microscopy analyses and Dr. Kevin G. 27. Olmo N, Turnay J, Gonza«lez de Buitrago G, Lo«pez de Silanes I, Gavilanes JG, Becker for his contributions to the microarray analysis. We also thank the Lizarbe MA. Cytotoxic mechanism of the ribotoxin alpha-sarcin. Induction of cell Microscopy and Flow Cytometry Center and the Animal Care Laboratory from death via apoptosis. Eur J Biochem 2001;268(7):2113Ð23. the Complutense University of Madrid (Spain). 28. Turnay J, Olmo N, Lizarbe MA, von der Mark K. Changes in the expression of annexin A5 gene during in vitro chondrocyte differentiation: influence of cell attachment. J Cell Biochem 2002;84(1):132Ð42. REFERENCES 29. Lo«pez-Conejo T, Olmo N, Turnay J, Navarro J, Lizarbe MA. Characterization of tumorigenic sub-lines from a poorly tumorigenic human colon-adenocarcinoma cell 1. Velazquez OC, Lederer HM, Rombeau JL. Butyrate and the colonocyte. Production, line. Int J Cancer 1996;67(5):668Ð75. absorption, metabolism, and therapeutic implications. Adv Exp Med Biol 1997;427: 30. Fan J, Yang X, Wang W, Wood WH, Becker KG, Gorospe M. Global analysis of 123Ð34. stress-regulated mRNA turnover by using cDNA arrays. Proc Natl Acad Sci USA 2. Topping DL, Clifton PM. Short-chain fatty acids and human colonic function: roles 2002;99(16):10611Ð6. of resistant starch and nonstarch polysaccharides. Physiol Rev 2001;81(3):1031Ð64. 31. Cheadle C, Vawter MP, Freed WJ, Becker KG. Analysis of microarray data using z 3. Andoh A, Tsujikawa T, Fujiyama Y. Role of dietary fiber and short-chain fatty acids score transformation. J Mol Diagn 2003;5(2):73Ð81. in the colon. Curr Pharm Des 2003;9(4):347Ð58. 32. Sinicrope FA, Roddey G, McDonnell TJ, Shen Y, Cleary KR, Stephens LC. Increased 4. Augenlicht LH, Mariadason JM, Wilson A, et al. Short chain fatty acids and colon apoptosis accompanies neoplastic development in the human colorectum. Clin Cancer cancer. J Nutr 2002;132(12):3804SÐ8S. Res 1996;2(12):1999Ð2006. 5. Velazquez OC, Zhou D, Seto RW, et al. In vivo crypt surface hyperproliferation is ϩ decreased by butyrate and increased by deoxycholate in normal rat colon: associated 33. Forest V, Pierre F, Bassonga E, Meflah K, Olivier C, Menanteau J. Apc /Min in vivo effects on c-Fos and c-Jun expression. J Parenter Enteral Nutr 1996;20(4): colonic epithelial cells express TNF receptors and ICAM-1 when they are co-cultured 243Ð50. with large intestine intraepithelial lymphocytes. Cell Immunol 2003;223(1):70Ð6. 6. Navarro JM, Olmo N, Turnay J, Lo«pez-Conejo MT, Lizarbe MA. Differentiation of 34. Lo«pez-Conejo MT, Olmo N, Turnay J, Lo«pez de Silanes I, Lizarbe MA. Interaction BCS-TC2 human colon adenocarcinoma cells by sodium butyrate: increase of 5Ј- of fibronectin with human colon adenocarcinoma cells: effect on the in vivo tumor- nucleotidase activity. Eur J Clin Invest 1997;27(7):620Ð8. igenic capacity. Oncology 2002;62(4):371Ð80. 7. Saito H, Ebinuma H, Takahashi M, et al. Loss of butyrate-induced apoptosis in human 35. Musch MW, Sugi K, Straus D, Chang EB. Heat-shock protein 72 protects against hepatoma cell lines HCC-M and HCC-T having substantial Bcl-2 expression. Hepa- oxidant-induced injury of barrier function of human colonic epithelial Caco2/bbe tology 1998;27(5):1233Ð40. cells. Gastroenterology 1999;117(1):115Ð22. 8. Gaschott T, Stein J. Short-chain fatty acids and colon cancer cells: the vitamin D 36. Williams AC, Harper SJ, Paraskeva CH. Neoplastic transformation of a human receptor-butyrate connection. Recent Res Cancer Res 2003;164:247Ð57. colonic epithelial cell line: in vitro evidence for the adenoma to carcinoma sequence. 9. Chen JS, Faller DV, Spanjaard RA. Short-chain fatty acid inhibitors of histone Cancer Res 1990;50(15):4724Ð30. deacetylases: promising anticancer therapeutics? Curr Cancer Drug Targets 2003; 37. Iacomino G, Tecce MF, Grimaldi C, Tosto M, Russo GL. Transcriptional response of 3(3):219Ð36. a human colon adenocarcinoma cell line to sodium butyrate. Biochem Biophys Res 10. Tabuchi Y, Arai Y, Kondo T, Takeguchi N, Asano S. Identification of genes Commun 2001;285(5):1280Ð9. responsive to sodium butyrate in colon epithelial cells. Biochem Biophys Res Com- 38. Tsuruo T, Naito M, Tomida A, et al. Molecular targeting therapy of cancer: drug mun 2002;293:1287Ð94. resistance, apoptosis and survival signal. Cancer Sci 2003;94(1):15Ð21.

4600

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2004 American Association for Cancer Research. Acquisition of Resistance to Butyrate Enhances Survival after Stress and Induces Malignancy of Human Colon Carcinoma Cells

Isabel López de Silanes, Nieves Olmo, Javier Turnay, et al.

Cancer Res 2004;64:4593-4600.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/13/4593

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/64/13/4593. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.