CAM-DR - Nature

3 downloads 87 Views 143KB Size Report
Terry H Landowski1,3, Nancy E Olashaw2, Deepak Agrawal2 and William S ... of Cell Biology, H Lee Moffitt Cancer Center and Research Institute, University of ...
Oncogene (2003) 22, 2417–2421

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

Cell adhesion-mediated drug resistance (CAM-DR) is associated with activation of NF-jB (RelB/p50) in myeloma cells Terry H Landowski1,3, Nancy E Olashaw2, Deepak Agrawal2 and William S Dalton1 1 Department of Interdisciplinary Oncology, H Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, FL 33612, USA; 2 Department of Cell Biology, H Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, FL 33612, USA; 3Department of Medicine, Arizona Cancer Center, University of Arizona, Tucson, AZ 85724, USA

The microenvironment has been shown to influence tumor cell phenotype with respect to growth, metastasis, and response to chemotherapy. We have utilized oligonucleotide microarray analysis to identify signal transduction pathways and gene products altered by the interaction of myeloma tumor cells with the extracellular matrix component fibronectin that may contribute to the antiapoptotic phenotype conferred by the microenvironment. Genes with altered expression associated with fibronectin cell adhesion, either induced or repressed, were numerically ranked by fold change. FN adhesion repressed the expression of 469 gene products, while 53 genes with known coding sequences were induced by twofold or more. Of these 53 genes with two fold, or greater increase in expression, 11 have been reported to be regulated by the nuclear factor-kappa B (NF-jB) family of transcription factors. EMSA analysis demonstrated NF-jB binding activity significantly increased in cells adhered to fibronectin compared to cells in suspension. This DNA binding activity consisted primarily of RelBp50 heterodimers, which was distinct from the NF-jB activation of TNFa. These data demonstrate the selectivity of signal transduction from the microenvironment that may contribute to tumor cell resistance to programmed cell death. Oncogene (2003) 22, 2417–2421. doi:10.1038/sj.onc.1206315 Keywords: myeloma; drug resistance; adhesion; NF-kB; microarray

We, and others, have described a multidrug-resistant phenotype that is associated with cell–cell adhesion or adhesion to extracellular matrices, including fibronectin (Damiano et al., 1999; Green et al., 1999). The term cell adhesion-mediated drug resistance’’ or CAM-DR has been used to describe this phenotype. Molecular mechanisms associated with the CAM-DR phenotype that have been described include alteration in intracellular distribution of the drug target (Hazlehurst et al., *Correspondence: WS Dalton, H Lee Moffitt Cancer Center and Research Institute, 12902 Magnolia Dr, Tampa, FL 33612, USA; E-mail: dalton@moffitt.usf.edu Received 14 May 2002; revised 4 December 2002; accepted 10 December 2002

2001), increased p27kip1 levels (St Croix et al., 1996; Hazlehurst et al., 2000), and increased expression of antiapoptotic molecules (Shain et al., 2002). To further elucidate critical molecules involved in the antiapoptotic signaling environment associated with cell adhesion, we used oligonucleotide microarray analysis to examine gene expression when myeloma cells were placed in suspension culture versus those adhered to fibronectin. Understanding basal levels of gene expression between cells in suspension and those adhered to fibronectin may elucidate why adherent cells are more resistant to drugs and other forms of cellular stress. Indeed, using this analytic approach, we found evidence that activation of nuclear factor-kappa B (NF-kB) is associated with CAM-DR. Myeloma cells were adhered to fibronectin for 8 h, and the expression profile of 6000 genes compared to that of cells maintained in suspension. mRNA expression levels were determined using Affymetrix HuGeneFL human gene chips, and analysed with GeneChip 3.3 software. Data calculated from hybridization intensities are expressed as an average difference of sequencespecific hybridization (as measured by a perfect complement probe set) and the nonspecific hybridization (as measured by a corresponding 1 base mismatch probe set) following normalization to internal controls. Cells maintained in suspension were designated as the reference population, and genes with altered expression in cells adhered to FN were ranked in numerical order based on fold change. We identified only 53 known gene products induced by twofold or greater following FN adhesion. In contrast, 469 gene products were decreased greater than two fold. Of the 53 genes induced by adhesion to FN, 11 have been documented as regulated by the NF-kB family of transcription factors (Table 1). NF-kB was originally identified as a DNA binding complex regulating the expression of genes involved in immune and inflammatory responses (rev. in Gugasyan et al., 2000; Karin and Lin, 2002). Subsequent studies demonstrated that the NF-kB family includes five members, p50/p105 (NF-kB1), p52/p100 (NF-kB2), cRel, RelB, and p65 (RelA) with a common Rel homology domain (RHD) required for protein dimerization, nuclear translocation and DNA binding. Inactive NF-kB is sequestered in the cytoplasm as inactive complexes with a family of inhibitory proteins

Adhesion of FN activates RelB in myeloma cells TH Landowski et al

2418 Table 1 NF-kB-regulated genes induced by adhesion to fibronectin GenBank accession

Description

Fold change

M92357 Y00081 X56692 D79206 M69043 X53683 Z11697 U37546 X66867 Y00787 X66365

Tumor necrosis factor, alpha-induced protein 2 Interleukin 6 (interferon, beta 2) C-reactive protein, pentraxin related clg protein, syndecan-4 Nuclear factor of kappa light polyepeptide gene enhancer in B-cell, inhibitor, alpha Small inducible cytokine A4 (homologous to mouse Mip-1b) CD83 antigen (activated B lymphocytes, immunoglobulin superfamily) Apoptosis inhibitor 2 MAX protein Interleukin 8 Cyclin-dependent kinase 6

6.5 4.9 4.4 3.9 2.9 2.8 2.2 2.1 2.1 2.0 2.0

A total of 8226 myeloma cells were adhered to FN or maintained in suspension for 8 h. RNA was isolated and analysed for gene expression using Affymatrix Gene Expression system and GeneChip software. Genes were ranked by fold induction in FN adhered cells versus suspension cells. mRNA expression of selected genes was verified by RNase protection assay using the Riboquant Multitemplate probe (Pharmingen). Of the 53 genes demonstrating twofold or greater induction, 11 have been demonstrated as regulated by NF-kB

called IkB that bind to the RHD and mask the nuclear localization signal. Upon activation by a number of stimuli, the IkB proteins are phosphorylated and targeted for ubiquitination and degradation by the 26S proteasome, allowing NF-kB translocation to the nucleus for the regulation of specific gene expression. The p50/p105 and p52/p100 proteins are synthesized as ankyrin repeat-containing precursors that also function as inhibitory factors. Limited proteolysis of the p105 and p100 proteins by the 26S proteasome is required to generate active DNA binding subunits; however, the specificity of the signal transduction pathway that regulates this limited proteolysis is not entirely understood (Xiao et al., 2001). We utilized gel shift analysis to determine if adhesion to fibronectin induces NF-kB binding activity in myeloma cells. As shown in Figure 1a, myeloma cells frequently contain low levels of constitutively active NFkB DNA binding complexes consisting primarily of p50 homodimers as has been previously reported for terminally differentiated B cells (Liou et al., 1994). Adhesion to fibronectin induces a time-dependent increase in DNA binding activity, reaching a 20-fold increase following 16–24 h of adhesion. The RPMI 8226, ARH77, H929 and MM.1S myeloma cell lines all demonstrate an induction of NF-kB DNA binding activity following adhesion to FN (Figure 1a). In contrast, the plasma cell leukemia-derived cell line, IM-9, shows only very low affinity for fibronectin binding, and does not display NF-kB activity in FN adhered cells. We believe this may be a reflection of the plasma cell leukemia phenotype, which is found in peripheral circulation, as opposed to the bone marrow microenvironment. Supershift analysis with antibodies specific for NF-kB family members demonstrates the majority of the DNA binding activity in FN adhered cells is composed of RelB/p50 heterodimers with very low levels of p65 (Figure 1b). Western blot analysis of nuclear extracts verified the nuclear accumulation of RelB in cells adhered to fibronectin as compared to cells maintained in suspension (Figure 2). Activation of NFkB following adhesion to ECM substrates has preOncogene

viously been demonstrated in fibroblasts and smooth muscle cells (Qwarnstrom et al., 1994), thymic epithelium (Scupoli et al., 2000), Jurkat T cells (Bearz et al., 1998) and THP-1 monocytic cells (Rosales and Juliano, 1996). In these diverse cell types, NF-kB activity was associated with translocation of the p50/p65 heterodimer. In contrast, this report is the first to demonstrate adhesion of myeloma cells to fibronectin resulting in the activation of RelB, suggesting a signal-specific regulation of NF-kB activity in these cells. Reports from other laboratories have demonstrated that adhesion of myeloma cells to bone marrow stromal cells induces the production of TNFa by the myeloma cells and the activation of NF-kB in the stromal cells (Hideshima et al., 2001). To determine if the NF-kB binding activity induced by the adhesion of 8226 myeloma cells was a secondary effect of TNFa secretion, we compared the DNA binding activity of cells treated with TNFa to those adhered to FN (Figure 3). In contrast to the NF-kB activity seen following adhesion to FN, TNFa induces a rapid activation of the canonical p65/p50 subunits within 30 min of treatment. Although we cannot entirely rule out the contribution of soluble factors to NF-kB activity in FN adhered cells, we did not find increases in TNFa mRNA in our system, nor was the NF-kB DNA binding profile compatible with induction by TNFa. Therefore, these data demonstrate that RelB activation by adhesion to FN is not secondary to TNFa secretion. The mechanism of differential NFkB activation in myeloma cells, and the point of bifurcation in this pathway is not currently known. However, previous reports have demonstrated constitutive RelB activation in v-src transformed fibroblasts (Shain et al., 1999) suggesting a role for the protooncogene c-src, or a src family member contributing to adhesion-mediated NF-kB activity in myeloma cells. Over 150 target genes have been identified as regulated by the NF-kB family. In the hematopoietic system, NF-kB is considered a central regulator of inflammation and the immune response, primarily because of the large number of cytokines and growth factors whose expression is under NF-kB control

Adhesion of FN activates RelB in myeloma cells TH Landowski et al

2419

a

8226

ARH77

H929

IM-9

MM.1S

S FN

S FN

S FN

S FN

S FN

NF-KB

l ne 0 2 5 elB -Re no p5 p5 p6 R C

no

Supershift antibody

Fibronectin ne

Suspension

b

l

0 2 5 elB -Re C p5 p5 p6 R

p50/RelB

Figure 1 (a) Adhesion to FN activates NF-kB in myeloma cell lines. Fibronectin-coated plates were prepared by adding 40 mg/ml Fibronectin (Intergen, San Diego, CA, USA) to Nunclon tissue culture dishes and allowing liquid to evaporate overnight (Damiano et al., 1999). Myeloma cell lines 8226, ARH77, H929, IM-9, and MM.1S were adhered to FN for 16 h, nuclear extracts were prepared as previously described (Catlett-Falcone et al., 1999) and assayed for DNA binding activity using a radiolabeled probe corresponding to the NF-kB consensus sequence (Shain et al., 1999). S: Cells maintained in suspension; FN: Cells adhered to fibronectin. (b). NF-kB activity in FN-adhered myeloma cells is primarily RelB/p50 heterodimers. To identify NF-kB subunits, nuclear extracts prepared as described were incubated with Rel-specific antibodies for 60 m on ice prior to EMSA analysis. All antibodies were obtained from Santa Cruz Biotech (Santa Cruz, CA, USA)

2h

4h

S FN S FN

8h

16h

24h

S FN S FN S FN

RelB

actin Figure 2 RelB/p50 DNA binding activity correlates with accumulation of RelB protein in the nucleus. The 8226/S myeloma cells were adhered to FN for the indicated time, harvested by hypotonic lysis, and nuclear extracts prepared. In correlation with DNA binding analysis, Western blot demonstrated an accumulation of RelB in the nucleus following adhesion to FN

(Gugasyan et al., 2000). Recent studies have implicated NF-kB-mediated gene expression in mechanisms of resistance to apoptosis induced by both physiological and cytotoxic stimuli (Wang et al., 1996; Van Antwerp et al., 1996; Feinman et al., 1999; Bian et al., 2001). Although distinct mechanisms of activation have been described, both cytotoxic drugs and death receptors initiate the proteolytic caspase cascade that cleaves essential cell proteins. Caspase activity is regulated at multiple points by proteins that are transcriptional products of NF-kB including Bfl-1/A1 (Zong et al., 1999; Wang et al., 1999a; Grumont et al., 1999), Flice inhibitory protein (Micheau et al., 2001), Bcl-xL (Chen et al., 2000), and the inhibitor of apoptosis (IAP) family (Chu et al., 1997; Wang et al., 1998). Microarray analysis demonstrated a 2.1-fold induction of c-IAP mRNA in myeloma cells adhered to FN for 8 h (see Table 1). We further examined the mRNA and protein expression of c-IAP2 in myeloma cells adhered to FN. Oncogene

Adhesion of FN activates RelB in myeloma cells TH Landowski et al

2420

RelB DNA binding activity has recently been implicated in the survival of B lymphocytes following ligation of the TNF family receptor TACI/BCMA/BAFF-R (Do et al., 2000; Do RK and Chen-Kiang, 2002). This activity was associated with increased expression of the antiapoptotic proteins, Bcl-2 and Bcl-xL, and reduced expression of the proapoptotic factor, Bak. However, in agreement with our previous report demonstrating no changes in the expression levels of the Bcl-2 family following adhesion to FN (Damiano et al., 1999), Bcl-xL expression was not induced by FN adhesion in our myeloma system. Although the mechanism of selectivity is not clear, these data support a differential regulation of gene expression by microenvironmental factors in physiological systems. Microarray technology can be used to provide a global view of gene expression that defines a specific phenotype (Kudoh et al., 2000). Using this approach, we have identified a signal transduction pathway activated by the interaction of tumor cells with their microenvironment. Several of the NF-kB gene products identified by microarray analysis in our study are well-characterized survival factors in multiple myeloma. Although it is tempting to propose any one of these gene products, ‘the CAM-DR resistance factor’, it is unlikely that the microenvironment conferred resistant phenotype is entirely because of a single gene product. Rather, the antiapoptotic phenotype conferred by the microenvironment is likely to reflect a culmination of the gene expression profile, and is likely to involve additional transcriptional regulation in addition to RelB. Identification of the NF-kB signal transduction pathway by the gene expression profile supports the hypothesis that this pathway may represent a novel molecular target for drug development (Van Antwerp et al., 1996; Wang et al., 1999b; Mayo and Baldwin, 2000). Further investigation identifying the specificity of RelB activation by the microenvironment may provide additional targets for chemotherapeutic intervention.

ne

8226/S + TNF no

no

Supershift antibody

ne

Ct

5

p6

0

p5

lB

Re

l

Re

c-

NF-KB

Figure 3 TNFa-mediated activation of NF-kB is distinct from adhesion-mediated activation. The 8226/S myeloma cells were treated with 100 ng/ml TNF for 30 min, nuclear extracts prepared and analysed for NF-kB DNA binding activity as described. Supershift demonstrates the primary DNA binding activity in 8226 myeloma cells treated with TNFa is p50/p65. CT: untreated control

As shown in Figure 4, c-IAP2 protein demonstrated a time-dependent increase, with maximum protein expression approximately twofold in cells adhered to FN for 8 h. This correlated well with the cDNA analysis and mRNA expression examined by RNase protection assay, which demonstrated a mean increase of 1.7-fold (s.d. 0.15, n ¼ 4) (data not shown). The IAP proteins have been shown to bind and potently inhibit procaspase 9, preventing activation of the effector caspases 3, 6, and 7 through the mitochondrial pathway. In addition, c-IAP1, c-IAP2, and XIAP can complex with the active form of caspases 3 and 7, suppressing amplification of death receptor-mediated apoptosis through caspase 8. Thus, induced expression of c-IAP protein by adhesion to fibronectin may contribute to resistance to both physiological- and cytotoxic-induced apoptosis in this phenotype. Further studies are warranted to identify the role of the c-IAP family in the CAM-DR phenotype. Expression of the antiapoptotic family member, Bcl-xL has been reported to be regulated by c-Rel and p65 transactivation in HT 1080 sarcoma cells (Chen et al., 2000). Additionally,

2h S

4h FN

S

Acknowledgements This work was supported by the National Cancer Institute (CA77859 and CA82533 to WSD), and the Peninsula Community Foundation for Myeloma Research. We also thank the H Lee Moffitt Flow Cytometry and Microarray Core Facilities (CA76292).

8h FN

S

FN

24h S FN C-IAP2

actin Figure 4 Adhesion of 8226 myeloma cells to FN induces the expression of c-IAP2. 8226 myeloma cells were adhered to FN for the indicated time, and whole cell lysates assayed for c-IAP2 expression by Western blot. Data shown are representative of three independent experiments Oncogene

Adhesion of FN activates RelB in myeloma cells TH Landowski et al

2421 References Bearz A, Tell G, Colombatti A, Formisano S and Pucillo C. (1998) Biochem. Biophys. Res. Commun., 243, 732–737. Bian X, McAllister-Lucas LM, Shao F, Schumacher KR, Feng Z, Porter AG, Castle VP and Opipari Jr AW. (2001). J. Biol. Chem. 276, 48921–48929. Catlett-Falcone R, Landowski TH, Oshiro MM, Turkson J, Levitzki A, Savino R, Ciliberto G, Moscinski L, FernandezLuna JL, Nunez G, Dalton WS and Jove R. (1999). Immunity, 10, 105–115. Chen C, Edelstein LC and Gelinas C. (2000). Mol. Cell. Biol., 20, 2687–2695. Chu Z-L, McKinsey TA, Liu L, Gentry JJ, Malim MH and Ballard DW. (1997). Proc. Natl. Acad. Sci. USA, 94, 10057– 10062. Damiano JS, Cress AE, Hazlehurst LA, Shtil AA and Dalton WS. (1999). Blood, 93, 1658–1667. Do RK and Chen-Kiang S. (2002). Cytokine Growth Factor Rev., 13, 19–25. Do RK, Hatada E, Lee H, Tourigny MR, Hilbert D and Chen-Kiang S. (2000). J. Exp. Med., 192, 953–964. Feinman R, Koury J, Thames M, Barlogie B, Epstein J and Siegel DS. (1999). Blood, 93, 3044–3052. Green SK, Frankel A and Kerbel RS. (1999). Anticancer Drug Des., 14, 153–168. Grumont RJ, Rourke IJ and Gerondakis S. (1999). Genes Dev. 13, 400–411. Gugasyan R, Grumont R, Grossmann M, Nakamura Y, Pohl T, Nesic D and Gerondakis S. (2000). Immunol. Rev., 176, 134–140. Hazlehurst LA, Damiano JS, Buyuksal I, Pledger WJ and Dalton WS. (2000). Oncogene, 19, 4319–4327. Hazlehurst LA, Valkov N, Wisner L, Storey JA, Boulware D, Sullivan DM and Dalton WS. (2001). Blood, 98, 1897–1903. Hideshima T, Richardson P, Chauhan D, Palombella VJ, Elliott PJ, Adams J and Anderson KC. (2001). Cancer Res., 61, 3071–3076.

Karin M and Lin A. (2002). Natl. Immunol., 3, 221–227. Kudoh K, Ramanna M, Ravatn R, Elkahloun AG, Bittner ML, Meltzer PS, Trent JM, Dalton WS and Chin KV. (2000). Cancer Res., 60, 4161–4166. Liou HC, Sha WC, Scott ML and Baltimore D. (1994). Mol. Cell Biol., 14, 5349–5359. Mayo MW and Baldwin Jr AS. (2000). Biochim. Biophys. Acta, 1470, 55–62. Micheau O, Lens S, Gaide O, Alevizopoulos K and Tschopp J. (2001). Mol. Cell. Biol., 21, 5299–5305. Qwarnstrom EE, Ostberg CO, Turk GL, Richardson CA and Bomsztyk K. (1994). J. Biol. Chem., 269, 30765–30768. Rosales C and Juliano R. (1996). Cancer Res., 56, 2302–2305. Scupoli MT, Fiorini E, Marchisio PC, Poffe O, Tagliabue E, Brentegani M, Tridente G and Ramarli D. (2000). J. Cell Sci., 113, 169–177. Shain KH, Jove R and Olashaw NE. (1999). J. Cell Biochem., 73, 237–247. Shain KH, Landowski TH and Dalton WS. (2002). J. Immunol., 168, 2544–2553. St Croix B, Florenes VA, Rak JW, Flanagan M, Bhattacharya N, Slingerland JM and Kerbel RS. (1996). Nat. Med., 2, 1204–1210. Van Antwerp DJ, Martin SJ, Kafri T, Green DR and Verma IM. (1996). Science, 274, 787. Wang C-Y, Cusack JC, Liu R and Baldwin Jr AS. (1999b). Nat. Med., 5, 412–417. Wang C-Y, Guttridge DC, Mayo MW and Baldwin Jr AS (1999a). Mol. Cell. Biol., 19, 5923–5929. Wang C-Y, Mayo MW and Baldwin Jr AS (1996). Science, 274, 784–787. Wang C-Y, Mayo MW, Korneluk RG, Goeddel DV and Baldwin Jr AS (1998). Science, 281, 1680–1683. Xiao G, Harhaj EW and Sun SC (2001). Mol. Cell, 7, 401–409. Zong W-Z, Edelstein LC, Chen C, Bash J and Gelinas C (1999). Genes Dev., 13, 382–387.

Oncogene