MB49 Murine Urothelial Carcinoma: Molecular and Phenotypic Comparison to Human Cell Lines as a Model of the Direct Tumor Response to Bacillus Calmette-Guerin

MB49 Murine Urothelial Carcinoma: Molecular and Phenotypic Comparison to Human Cell Lines as a Model of the Direct Tumor Response to Bacillus Calmette-Guerin Fanghong Chen,* Guangjian Zhang,* Yanli Cao, Martin J. Hessner and William A. See†,‡ From the Department of Urology, Medical College of Wisconsin and Max McGee National Research Center for Juvenile Diabetes, Department of Pediatrics at Medical College of Wisconsin and Children’s Research Institute of Children’s Hospital of Wisconsin (MJH), Milwaukee, Wisconsin

Abbreviations and Acronyms AP1 ⫽ activator protein 1 BCG ⫽ bacillus Calmette-Guerin C/EBP ⫽ CCAAT-enhancer-binding protein FBS ⫽ fetal bovine serum GO ⫽ gene ontology HMGB1 ⫽ high molecular group box protein 1 IL-6 ⫽ interleukin-6 NF-␬B ⫽ nuclear factor ␬B p21 ⫽ cyclin dependent kinase inhibitor p21 (cip1; waf1) RAGE ⫽ receptor for advanced glycation of end products TLR ⫽ toll-like receptor UC ⫽ urothelial carcinoma Submitted for publication February 26, 2009. Supported by a grant from the Department of Veterans Affairs and Milwaukee Veterans Affairs Medical Center. * Equal study contribution. † Correspondence: Department of Urology, Medical College of Wisconsin, 9200 West Wisconsin Ave., Milwaukee, Wisconsin 53226 (telephone: 414-456-6950; FAX: 414-456-6217). ‡ Financial interest and/or other relationship with AstraZeneca.

2932

www.jurology.com

Purpose: The mouse urothelial carcinoma cell line MB49 is widely used as an in vitro and in vivo model of urothelial carcinoma. Little comparative data exist on the molecular and phenotypic responses of this cell line relative to human cell lines. We compared the effect of bacillus Calmette-Guerin on the MB49 cell line relative to responses previously observed in the human urothelial carcinoma lines T24 (ATCC®) and 253J. Materials and Methods: Molecular end points in MB49 cells after bacillus Calmette-Guerin exposure were signaling pathway activation (NF-␬B, AP1 and C/EBP), gene expression (IL-6 and p21), HMGB1 release/responsiveness and gene expression profiling at 6 hours. Phenotypic response end points were direct cytotoxicity using dye exclusion, viability on MTT assay, apoptotic sensitivity and cell cycle compartmentalization. Results: NF-␬B, AP1, C/EBP, IL-6 and p21 reporter constructs were activated in MB49 cells in response to bacillus Calmette-Guerin. Gene expression profiles showed an inflammatory/immune clustering response. Bacillus Calmette-Guerin decreased cell viability and induced G1 cell cycle arrest. Treatment of MB49 cells with bacillus Calmette-Guerin induced caspase independent cell death while simultaneously decreasing sensitivity to pro-apoptotic agents. Cell death was associated with release of the necrotic cell death marker HMGB1. MB49 cells expressed HMGB1 receptors and activated intracellular NF-␬B signaling pathways in response to bacillus Calmette-Guerin. Conclusions: MB49 cells show molecular and phenotypic responses to bacillus Calmette-Guerin that replicate those observed in human urothelial carcinoma lines. MB49 cells appear to be an excellent model in which to study bacillus Calmette-Guerin as an antitumor agent for urothelial carcinoma. Key Words: urinary bladder; Mycobacterium bovis; urinary bladder neoplasms; urothelium; models, animal EXPOSURE of human UC cells to BCG has a direct effect on the biology of the exposed cells. BCG exposure alters gene expression and the biological phenotype of human UC cells. Our group reported that BCG acti-

vates multiple intracellular signaling pathways (NF-␬B, AP1 and C/EBP) that regulate downstream gene expression (p21 and IL-6).1–3 Phenotypic changes associated with BCG exposure include cell cycle arrest in G1, loss of

0022-5347/09/1826-2932/0 THE JOURNAL OF UROLOGY® Copyright © 2009 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 182, 2932-2937, December 2009 Printed in U.S.A. DOI:10.1016/j.juro.2009.08.018

MURINE UROTHELIAL CANCER AND HUMAN BACILLUS CALMETTE-GUERIN RESPONSE

apoptotic sensitivity and cell death associated with release of the potent chemokine HMGB1 in a subpopulation of cells.4 – 6 While the contribution of these direct effects to the antitumor activity of BCG is unclear, the nature of BCG induced changes supports a potential role. In vitro studies have proved useful for studying the direct effect of BCG on UC biology. However, the complex interplay of host, tumor and BCG limits the usefulness of in vitro model systems for evaluating the contribution of direct BCG/tumor interactions to the antitumor response. In contrast, animal models have the potential to recapitulate the human clinical setting. Consequently in vivo systems have the potential to provide more robust, clinically relevant insight into the BCG mechanism of action. Key to any in vivo model is the reproducible induction/implantation of a relevant tumor. Ideally relevance would imply that the tumor used in these models would manifest a response to BCG that was similar if not identical to that of human UC lines. The murine MB49 tumor line has been used extensively as an implantable orthotopic animal model for studying BCG. To our knowledge no studies to date have assessed the extent to which the direct response of the MB49 cell line to BCG mirrors that in human UC cell lines. We compared the molecular and phenotypic effects of BCG on MB49 cells to those previously observed in the human UC cell lines T24 and 253J to determine the appropriateness of MB49 as an in vivo model in which to study BCG.

MATERIALS AND METHODS Cell Lines and BCG The murine UC cell line MB49, and the human UC cell lines T24 and 253J were maintained at 37C in 5% CO2 in RPMI 1640 (Gibco®) supplemented with 10% FBS, penicillin and streptomycin (complete medium). TICE® BCG, living organisms of an attenuated bacillus of the Calmette and Guerin strain of Mycobacterium bovis, were used in the experiments. Freeze-dried BCG was reconstituted in complete medium at an estimated concentration of 2.5 ⫻ 107 viable organisms per ml. Dilution assumed an average viability of 4 ⫻ 108 organisms per vial based on the manufacturer specified range of 1 to 8 ⫻ 108 per vial.

Luciferase Reporter Assay Activation of NF-␬B, AP1, C/EBP, p21 and IL-6 in MB49 cells in response to BCG was determined using luciferase reporter constructs, as previously described.1 Cells were transfected for 6 hours with 1 ␮g constructed cDNA using Lipofectamine 2000 reagent (Gibco). Cells were subsequently stimulated for 6 hours with 1 ⫻ 107 BCG. Cell lysates were harvested and analyzed sequentially for luciferase activity.

Gene Expression Profiling/Microarray Analysis We performed gene expression profiling of MB49, T24 and 253J UC cells 6 hours after BCG exposure using the

2933

Affymetrix® system, including the GeneChip® Mouse Genome 430 2.0 or Human Genome U133 Plus 2.0 array. Details of the methodology used for microarray analysis are described. RNA isolation. Cell line RNA was isolated using TRIzol® reagent and further purified using the RNeasy™ Mini Kit according to the manufacturer protocol. RNA was eluted from the columns with 100 ␮l water and then quantified spectrophotometrically. The 260/280 nm ratio for all RNA exceeded 1.9. cRNA preparation. Purified RNA samples were processed according to Affymetrix instructions. For microarray hybridization and data capture fragmented cRNA samples were hybridized onto the species appropriate GeneChip, washed and scanned according to the manufacturer protocol for eukaryotic target hybridization. Data were analyzed using GeneChip Operating System, versions 1.2– 1.4, followed by statistical analysis. Briefly, hybridization cocktails (250 ␮l) containing heat fragmented and biotin labeled cRNA at a concentration of 0.05 ␮g/␮l were injected into GeneChips and incubated at 45C on a rotator in a hybridization oven 640 (Affymetrix) overnight at 60 rpm. Arrays were washed and stained with phycoerythrin conjugated streptavidin (Molecular Probes®). Arrays were scanned immediately after the wash and staining procedure. Image data were analyzed with Affymetrix GeneChip Operating Software and normalized with BioConductor Robust Multichip Analysis to determine signal log ratios. Data processing and statistical analysis. Raw data were preprocessed with background adjustment, normalization and probe set summarization using the BioConductor Robust Multiarray Analysis package, as described by Schlecht et al.7 The statistical significance of differential gene expression was derived using Student’s t test and the percent of false-positive predictions, as determined with the Bioconductor RankProd package.

Flow Cytometry for Cell Cycle Analysis Flow cytometry was performed to quantify cell cycle compartmentalization of UC cell lines after BCG exposure, as previously described.5 Cells (5 ⫻ 106/6 cm dish) were synchronized using serum starvation for 60 hours before the start of the experiment. At time zero cells were exposed to BCG and 10% FBS at a ratio of 50:1 BCG to cells or to 10% FBS alone. Subsequently cells were trypsinized, centrifuged at 1,500 rpm for 5 minutes, washed with phosphate buffered saline and treated with 50 ␮g/ml RNase A (Sigma®). DNA was stained with 50 ␮g/ml propidium iodide for 10 minutes at room temperature and cell cycle analysis was done.

Apoptosis Assay The influence of BCG treatment on UC cell apoptotic sensitivity was determined by measuring activation of the caspase-3 pathway after cell treatment with the DNA topoisomerase inhibitor camptothecin, as previously described.4 Cells were treated with BCG at a 1:50 ratio for 6 hours or left untreated to serve as controls. Cells were washed and then cultured for 18 hours in the presence of the apoptotic inducing agent camptothecin at a concentration of 4 ␮M or left as untreated controls. Subsequently cells were harvested with trypsinization,

2934

MURINE UROTHELIAL CANCER AND HUMAN BACILLUS CALMETTE-GUERIN RESPONSE

3

Fold Induction

2.5 2 1.5 1 0.5 0 NFkB

AP-1

CEBP

p21

IL-6

Figure 1. Exposure of MB49 cells to BCG initiated intracellular signaling through NF-␬B, AP1 and C/EBP, and transactivated promoter reporter constructs for IL-6 and p21. Results are shown as fold induction vs to control cells. Activation of each reporter construct was significantly increased vs controls (Wilcoxon signed rank test p ⬍0.05).

washed once in phosphate buffered saline, fixed and permeabilized using the Cytofix/Cytoperm™ Kit for 20 minutes at room temperature and then pelleted and washed with Perm/Wash™ buffer. Cells were stained with anti-active caspase-3 mAb using 20 ␮g/1 ⫻ 106 cells for 60 minutes at room temperature in the dark. After incubation with Ab cells were washed in Perm/ Wash Buffer and analyzed by flow cytometry.

Dye Exclusion Assay for Cell Viability Prior studies show that exposure to BCG results in caspase independent cell death in a subpopulation of cells.4 This series of experiments used vital dye exclusion to study the direct cytotoxic effect of BCG on MB49 cells, as previously described.6 BCG was added to cultures at a ratio of 50:1 BCG to cells. Subsequently cell viability in attached and detached cell populations was evaluated using trypan blue dye (Invitrogen™). Cells were incubated with 0.4% trypan blue for 3 minutes before counting. Cells were counted in 4 fields and the number of dead cells was calculated as a percent of the total cell population. Cells with loss of membrane integrity as evidenced by trypan blue uptake were counted as dead.

Western Analysis for HMGB1 Protein and Receptors Analysis of HMGB1 in conditioned medium from MB49 cells was performed as previously described.6 Cells (1 ⫻ 106) were

seeded in 6 cm dishes and grown overnight at 37C in complete medium supplemented with 10% FBS. Medium was replaced with 5 ml fresh complete medium with or without BCG at a ratio of 50:1 BCG to cells. At 24 hours the conditioned medium from each dish was transferred into 1.5 ml tubes and centrifuged at 2,000 rpm at 4C for 15 minutes. Supernatant (4.5 ml) was transferred into Amicon® Ultra centrifugal filter devices and centrifuged at 7,000 rpm at 4C for 15 minutes to concentrate the supernatant. After centrifugation 150 ␮l concentrated medium were transferred into a 1.5 ml tube. The protein concentration in the medium was quantified using the BCA™ Protein Assay Kit. Equal amounts of total protein were loaded into each lane of a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel. After electrophoretic separation proteins were transferred onto Immobilon™ Transfer Membrane. Membranes were blocked with 5% milk at room temperature for 1 hour and washed 3 times with 1 ⫻ tris buffered saline with 0.1% Tween 20. The membrane was probed with anti-HMGB1 rabbit polyclonal IgG (Upstate Biotechnology, Lake Placid, New York) (2 ␮g/ml) at 4C overnight. The membrane was washed 3 times and incubated with peroxidase labeled antirabbit antibody (Amersham, Piscataway, New Jersey) (1: 4,000) at room temperature for 1 hour. After 3 washes bound antibodies were developed using the Enhanced Chemiluminescence System (Amersham). Western blotting for the HMGB1 receptors TLR-2, TLR-4 and RAGE was done as described with rabbit polyclonal antibodies against TLR-2, TLR-4 and RAGE (Santa Cruz Biotechnology, Santa Cruz, California), respectively (1:1,000). Experiments were performed a minimum of 3 times. The exception to this triplicate replication was microarray analysis, which was done in duplicate for all 3 cell lines. Results are shown as the average ⫾ 1 SE. Statistical comparisons were made using the Student t or Wilcoxon signed rank test, or multifactorial ANOVA for repeat measures. Microarray statistical analysis for gene up-regulation in replicate experiments was done using the rank products method.8

RESULTS MB49 Cell Response to BCG Intracellular signaling pathways were activated. Exposure of MB49 cells to BCG resulted in the activation of signaling pathway/transcription factors previously shown to be activated in the hu-

Table 1. GO terms associated with genes up-regulated in response to BCG T24 GO Identifier 0005125 0006952 0005615 0002376 0006954 0009611 0008009 0006935

GO Biological Process Name Cytokine activity Defense response Extracellular space Immune system process Inflammatory response Response to wounding Chemokine activity Chemotaxis

No. Genes Identified/ No. Mapped* 15/237 17/553 16/508 19/830 13/299 14/423 9/49 11/144

253J p Value ⫺18

1.60 ⫻ 10 3.43 ⫻ 10⫺16 3.41 ⫻ 10⫺15 2.90 ⫻ 10⫺16 1.33 ⫻ 10⫺13 2.83 ⫻ 10⫺13 2.46 ⫻ 10⫺14 9.59 ⫻ 10⫺14

* Using Database for Annotation, Visualization and Integrated Discovery.

No. Genes Identified/ No. Mapped* 4/237 7/553 7/508 5/830 7/299 7/423 3/49 3/144

MB49 p Value ⫺04

4.40 ⫻ 10 2.10 ⫻ 10⫺06 6.20 ⫻ 10⫺03 4.60 ⫻ 10⫺03 4.50 ⫻ 10⫺08 3.80 ⫻ 10⫺07 5.30 ⫻ 10⫺04 5.80 ⫻ 10⫺03

No. Genes Identified/ No. Mapped*

p Value

18/228 13/466 23/2084 17/741 13/231 13/322 12/42 8/113

1.50 ⫻ 10⫺22 1.29 ⫻ 10⫺10 7.90 ⫻ 10⫺12 3.81 ⫻ 10⫺13 3.28 ⫻ 10⫺14 1.71 ⫻ 10⫺12 6.87 ⫻ 10⫺21 4.54 ⫻ 10⫺09

MURINE UROTHELIAL CANCER AND HUMAN BACILLUS CALMETTE-GUERIN RESPONSE

sis revealed significantly identified GO term molecular functions associated with the changes in gene expression after BCG exposure (p ⬍0.001). Table 1 lists GO terms associated with upregulated gene expression after 6 hours of BCG treatment in each cell line. Table 2 shows a partial list of specific genes common to the mouse and human cell lines with up-regulated expression in response to BCG.

Table 2. Partial list of genes common to mouse and human cell lines with significantly increased expression in response to BCG Mean ⫾ SE Fold Induction vs Controls Gene Chemokine: (C-X-C motif) ligand 1 (C-X-C motif) ligand 2 (C-X-C motif) ligand 3 (C-X-C motif) ligand 5 (C-C motif) ligand 20 Colony-stimulating factor: 2 3 IL-6 Vascular cell adhesion molecule 1

MB49

T24

253J

4.2 ⫾ 0.5 5.3 ⫾ 0.4 1.9 ⫾ 0.2 1.8 ⫾ 0.1 3.2 ⫾ 1.4

2.3 ⫾ 1.3 1.3 ⫾ 0.7 3.2 ⫾ 1.4 1.4 ⫾ 0.6 4.3 ⫾ 3.6

2.0 ⫾ 0.8

2.7 ⫾ .06 3.9 ⫾ 0.2 5.4 ⫾ 0.1 2.5 ⫾ 0.2

1.2 ⫾ 0.8 1.6 ⫾ 0.8 3.2 ⫾ 1.3 1.5 ⫾ 1.3

2.4 ⫾ 1.4

BCG Decreased MB49 Cell Viability and Induced Cell Cycle Arrest Treatment of human UC cell lines with BCG decreased viablity.5 Exposure of MB49 cells to BCG significantly decreased cell viability as a function of time in MB49 cells, as measured by MTT (3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay (2-factor ANOVA p ⬍0.0001). Figure 2, A shows relative cell numbers at 1 and 3 days in BCG treated and control groups. Consistent with the reported effect of BCG on cell cycle compartmentalization in human cell lines, treatment of the MB49 cell line induced cell cycle arrest in G1.5 Figure 2, B shows the percent of cells in G1, S and G2 in control and BCG treated cells. BCG exposure significantly increased the percent of cells in G1 associated with a similar percent decrease in the G2 compartment (Student’s t test p ⬍0.05).

3.6 ⫾ 2.8

2.4 ⫾ 0.4

man cell lines 253J and T24.1 Figure 1 shows the fold induction of NF-␬B, AP-1 and C/EBP in MB49 cells 6 hours after BCG treatment. Activation of each luciferase reporter construct was significantly increased vs that in controls (Wilcoxon signed rank test p ⬍0.05). IL-6 and p21 were transactivated. The acute phase cytokine IL-6 and the cyclin-dependent kinase inhibitor p21 are transactivated in human UC cell lines in response to BCG.1,2 Promoter-reporter constructs for these proteins were activated in MB49 cells in response to BCG treatment. Figure 1 shows the fold induction of each reporter construct relative to controls (Wilcoxon signed rank test p ⬍0.05).

BCG Pretreatment of MB49 Cells Conferred Apoptotic Resistance Exposure of MB49 cells to the DNA topoisomerase inhibitor camptothecin significantly increased apoptosis, as measured by caspase-3 activation. The mean percent of caspase-3 positive cells was 4.8% and 19.8% in control and camptothecin treated cells, respectively (p ⫽ 0.001). BCG pretreatment before camptothecin exposure significantly decreased the percent of caspase-3 positive cells from 19.8% for camptothecin to 12.1% for BCG

Gene expression profiles were similar to those of human UC cell lines. The rank product method was used to identify up-regulated gene expression in response to BCG. With significance considered at p ⬍0.0001 ontological analysis of significantly BCG regulated gene expression was done with the Database for Annotation, Visualization and Integrated Discovery (http://david.abcc.ncifcrf.gov).9 Ontological analy-

B

A 1.2

90

Control 1

BCG

0.8

80

Untreated

70

BCG

60 Percent

Absorbance

2935

0.6 0.4

50 40 30 20

0.2

10

0

0

Day1

Day3

G1

S

G2

Figure 2. MB49 cells. A, MTT assay 1 and 3 days after BCG treatment shows that BCG exposure significantly decreased number of viable cells as function of time (2-factor ANOVA p ⬍0.0001). B, BCG induced cell cycle arrest. BCG exposure resulted in significant change in cell cycle compartmentalization vs controls and significantly increased percent of cells in G1 associated with similar percent decrease in G2 compartment (Student’s t test p ⬍0.05).

2936

MURINE UROTHELIAL CANCER AND HUMAN BACILLUS CALMETTE-GUERIN RESPONSE

A

B 16

20

Untreated

14

BCG

12

Caspase Inhibitor 1

10

BCG

Camptothesin(4uM)+BCG

15

Camptothesin (4 uM)

10

Dead Cells (%)

Percent Caspase 3 Positive

25 Untreated

BCG+Caspase Inhibitor 1

8 6 4

5 2 0

0

Figure 3. MB49 cells. A, pretreatment with BCG conferred apoptotic resistance. Exposure to DNA topoisomerase inhibitor camptothecin significantly increased apoptosis, as measured by caspase-3 activation (p ⫽ 0.001). BCG pretreatment before camptothecin exposure significantly decreased percent of caspase-3 positive cells (p ⫽ 0.01). B, caspase independent cell death in response to BCG. Exposure to BCG significantly increased number of nonviable cells, as measured by vital dye exclusion (Student’s t test p ⬍0.009). Effect was independent caspase pathway inhibition (controls vs BCG plus inhibitor p ⬍0.05 and BCG vs BCG plus inhibitor p ⫽ 0.69).

plus camptothecin (p ⫽ 0.01). Figure 3, A shows these results, which are identical to those previously reported for T24 and 253J cell lines.4

MB49 cells with BCG resulted in HMGB1 release into the cell culture supernatant, as measured by Western blot (fig. 4, A).

MB49 Cells Underwent Caspase Independent Cell Death After in vitro BCG exposure a percent of human UC cells undergo caspase independent cell death.4 Exposure of MB49 cells to BCG significantly increased the percent of dead cells on vital dye exclusion assay (Student’s t test p ⬍0.009). Pretreatment of cells with the broad-spectrum caspase inhibitor Z-VAD (OMe)-FMK failed to decrease BCG induced cytotoxicity (controls vs BCG plus inhibitor p ⬍0.05 and BCG vs BCG plus inhibitor p ⫽ 0.69, fig. 3, B).

MB49 Cells Expressed Cell Surface Receptors for HMGB1 and Activated NF-␬B Signaling Pathways After HMGB1 Exposure Human UC cell lines express HMGB1 receptors and activate HMGB1 responsive signaling pathways after HMGB1 exposure.6 Western blot analysis revealed that MB49 cells express the same HMGB1 receptors identified on human UC cell lines. Figure 4, B shows a composite Western blot using MB49 cells with size appropriate bands for TLR2, TLR4 and RAGE. Intact HMGB1 signaling in MB49 cells was confirmed using an NF-␬B reporter construct. Figure 5 shows the dose-response relationship between HMGB1 concentration and NF-␬B activation. NF-␬B activation significantly increased as a function of an HMGB1 concentration of 0 to 1 ␮g/ml (ANOVA p ⬍0.05).

BCG Treatment of MB49 Cells Was Associated with HMGB1 Release The nonhistone nuclear protein HMGB1 functions as a potent chemokine when released into the pericellular space. HMGB1 release is associated with necrotic cell death.10 A prior study by our group showed that BCG treatment increases HMGB1 release in vitro and in vivo.6 Treatment of

1.6 1.4

B

1.2 Fold Induction

A

1 0.8 0.6 0.4 0.2 0

Figure 4. Western blot analysis reveals that MB49 cells expressed HMGB1 receptors and released HMGB1 in response to BCG. A, HMGB1 receptors TLR2, TLR4 and RAGE in cell membrane proteins. B, binary increase from essentially nondetectable to apparent of nonhistone nuclear protein HMGB1 in supernatant.

Untreated

0.1ug/ml

0.5ug/ml

1ug/ml

Figure 5. In MB49 cells NF-␬B luciferase reporter construct activation was significantly increased as function of HMGB1 concentration (ANOVA p ⬍0.05). ug, ␮g.

MURINE UROTHELIAL CANCER AND HUMAN BACILLUS CALMETTE-GUERIN RESPONSE

DISCUSSION In patients undergoing treatment for UC the intravesical administration of BCG elicits a complex humoral and cellular host immune response. Beyond its impact on the host a mounting body of evidence demonstrates that BCG treatment stimulates a similarly complex response in the tumor. Up-regulation and secretion of acute phase cytokines and chemokine release coupled with fundamental changes in tumor biology suggest that the tumor response to BCG has the potential to influence the host response and tumor biology. Simultaneously the host immune response to BCG may alter tumor biology as well as the tumor response to subsequent BCG treatments. The protein HMGB1 is a useful example with which to illustrate the potential of the tumor response to BCG to have a broad impact on the in vivo system. HMGB1 released by BCG exposed tumor cells as consequence of nonapoptotic cell death can bind to and activate signaling through tumor associated cell surface receptors.11,12 Downstream signaling pathways, including NF- ␬ B, have the potential to regulate the expression of multiple genes involved in proliferation and apoptosis. These alterations in gene expression are translated into changes in tumor biology. In parallel tumor derived HMGB1 has the potential to influence the host response. The effect of HMGB1 on the host compartment has been demonstrated in multiple experimental models. HMGB1 serves as a chemoattractant for specific immune cell populations. Also, HMGB1 regulates immune cell differentiation and the nature of the humoral response of immune effector cells.13,14 Given the ability of urine to serve as a vehicle for remote delivery of soluble substances, the

2937

response at a single site has the potential to influence the biology of the entire bladder. Success in understanding how the interplay among BCG, tumor and host results in an antitumor response requires valid in vivo model systems. The host response to BCG has been extensively studied in the murine system. Humoral and cellular immune responses appear to parallel those in the clinical setting. Less well studied are the syngeneic murine tumor cell lines that have been used in mouse species to study BCG in orthotopic models.

CONCLUSIONS The MB49 tumor cell line has been used as a model in studies of various aspects of BCG antitumor activity. An Ovid MEDLINE® search combining the search terms MB49 and BCG provided 20 citations. By comparing the response of BCG exposed MB49 cells to the BCG response observed in human UC cell lines our study was designed to determine whether the MB49 line is a representative model system. We studied the direct effect of BCG treatment on signaling, gene expression and phenotypic response in the MB49 cell line compared to those in prior studies in human UC cell lines. Our results revealed that the response of MB49 cells to BCG is remarkably similar to that of human UC cell lines. Based on these findings we conclude that the MB49 cell line is an excellent model in which to assess the complex in vivo interrelationships among tumor, host and BCG.

ACKNOWLEDGMENTS Dr. Timothy Ratliff, Purdue, Indiana, provided the murine MB49 cell line. Dr. Richard Williams, Iowa City, Iowa, provided the human 253J UC cell line.

REFERENCES 1. Chen FH, Crist SA, Zhang GJ et al: Interleukin-6 production by human bladder tumor cell lines is upregulated by BCG through NF-␬b and AP-1 via an immediate early pathway. J Urol 2002; 168: 786. 2. Zhang G, Chen F, Cao Y et al: Bacillus Calmette-Guerin induces p21 expression in human transitional carcinoma cell lines via an immediate early, p53 independent pathway. Urol Oncol 2007; 25: 221.

carcinoma cell lines as a consequence of integrin cross-linking. BMC Urol 2005; 5: 8.

and Integrated Discovery. Genome Biol 2003; 4: P3.

6. See WA, Zhang G, Chen F et al: BCG induces caspase-independent cell death in urothelial carcinoma cells together with release of the necrosisassociated chemokine high molecular group box protein 1. BJU Int 2008; 103: 1714.

10. Scaffidi P, Misteli T and Bianchi ME: Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 2002; 418: 191.

3. Chen F, Zhang G, Iwamoto Y et al: Bacillus Calmette-Guerin initiates intracellular signaling in a transitional carcinoma cell line by cross-linking alpha 5 beta 1 integrin. J Urol 2003; 170: 605.

7. Schlecht U, Demougin P, Koch R et al: Expression profiling of mammalian male meiosis and gametogenesis identifies novel candidate genes for roles in the regulation of fertility. Mol Biol Cell 2004; 15: 1031.

4. Chen F, Zhang G, Cao Y et al: Bacillus CalmetteGuerin inhibits apoptosis in human urothelial carcinoma cell lines in response to cytotoxic injury. J Urol 2007; 178: 2166.

8. Breitling R, Armengaud P, Amtmann A et al: Pank products: a simple yet powerful new method to detect differentially regulated genes in replicated microarray experiments. FEBS Lett 2004; p 57383.

5. Chen F, Zhang G, Iwamoto Y et al: BCG directly induces cell cycle arrest in human transitional

9. Dennis G Jr, Sherman BT, Hosack DA et al: DAVID: Database for Annotation, Visualization,

11. Park JS, Svetkauskaite D, He Q et al: Involvement of toll-like receptors 2 and 4 in cellular activation by high mobility group box 1 protein. J Biol Chem 2004; 279: 7370. 12. Park JS, Gamboni-Robertson F, He Q et al: High mobility group box 1 protein interacts with multiple Toll-like receptors. Am J Physiol Cell Physiol 2006; 290: C917. 13. Yang H, Wang H, Czura CJ et al: The cytokine activity of HMGB1. J Leuk Biol 2005; 78: 1. 14. Dumitriu IE, Baruah P, Manfredi AA et al: HMGB1: guiding immunity from within. Trends Immunol 2005; 26: 381.