DOI:10.1158/0008-5472.

CAN-07-0031

Tumor-Suppressive Activity of the Cell Death Activator GRIM-19 on a Constitutively Active Signal Transducer and Activator of Transcription 3
Sudhakar Kalakonda, Shreeram C. Nallar, Daniel J. Lindner, et al. Cancer Res 2007;67:6212-6220. Published online July 6, 2007.

Updated Version

Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-07-0031

Cited Articles Citing Articles

This article cites 41 articles, 19 of which you can access for free at: http://cancerres.aacrjournals.org/content/67/13/6212.full.html#ref-list-1 This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/67/13/6212.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, contact the AACR Publications Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031

Research Article

Tumor-Suppressive Activity of the Cell Death Activator GRIM-19 on a Constitutively Active Signal Transducer and Activator of Transcription 3
Sudhakar Kalakonda, Shreeram C. Nallar, Daniel J. Lindner, Jiadi Hu, 2 1 Sekhar P. Reddy, and Dhananjaya V. Kalvakolanu
1 2

Department of Microbiology and Immunology, Greenebaum Cancer Center, University of Maryland School of Medicine; Department of Environmental Health Sciences, Johns Hopkins University Bloomberg School of Public Health, Baltimore, Maryland; and 3Taussig Cancer Center, The Cleveland Clinic Foundation, Cleveland, Ohio

Abstract
Signal transducers and activators of transcription 3 (STAT3) was originally identified as a transcription factor that mediates cytokine-induced responses. In these pathways, Janus-activated kinase (JAK)induced transient tyrosine phosphorylation of STAT3 promotes gene expression in response to a number of cytokines, which is inhibited by feedback mechanisms. A number of studies have shown that STAT3 is constitutively activated in human cancer cells, leading to cell proliferation. It is unclear, apart from a chronic tyrosyl phosphorylation of STAT3, what mechanisms contribute to the STAT3 deregulation in tumors. Earlier, we have isolated a novel growth inhibitory gene product, gene associated with retinoid-IFNinduced mortality 19 (GRIM-19), using a genetic approach. GRIM-19 is an IFN/retinoic acidregulated growth suppressor. Subsequent analyses have shown that GRIM-19 binds to STAT3 and prevents interleukin-6induced transcription of cellular genes. However, its effects on a constitutively active STAT3 and cellular transformation are unknown. In this study, we show that GRIM-19 suppresses constitutive STAT3induced cellular transformation in vitro and in vivo by downregulating the expression of a number of cellular genes involved in cell proliferation and apoptosis. [Cancer Res 2007;67(13):621220]

Introduction
Oncogenic proteins alter gene expression patterns during cellular transformation. DNA binding oncoproteins like c-MYC and FRA-1 have been shown to promote cell growth by activating specific sets of genes (1, 2). Antioncogenic proteins reverse these processes to restore normal cell growth. We have been studying the mechanisms of antitumor action of IFN-h and retinoic acid (IFN/ RA) in human tumor cells. IFN/RA exhibited strong antitumor actions in a number of animal models and in clinical studies (3, 4). Using a genetic technique, we have isolated a novel gene product, gene associated with retinoid-IFNinduced mortality 19 (GRIM-19), which promotes IFN/RAinduced cell death (5). More recently, we

have shown that a loss of GRIM-19 expression in human renal cell carcinomas (6) and an inhibition of its proapoptotic activity by viral oncoproteins occur, indicating a potential tumor suppressorlike function for this protein. Overexpression of GRIM-19 in a number of human tumor cell lines induces apoptosis (57). Some esophageal tumors seem to harbor inhibitors of GRIM-19 (8). A portion of the total GRIM-19 in cells is also present in the mitochondrion (911). Very recently, GRIM-19 has been coupled to reactive oxygen species generation in an IFN/RAinduced pathway in HeLa and MCF-7 cells (12). We and others have shown that GRIM-19 binds to signal transducer and activator of transcription 3 (STAT3) transcription factor and inhibits STAT3-dependent cytokine-induced gene transcription (13, 14). However, its role in regulating oncogenic cell proliferation and, in particular, its effects on a ligandindependent, constitutively active STAT3 and cellular transformation are unclear. To address these issues, we have used a cellular system in which a constitutively active STAT3 was sufficient to induce oncogenic transformation (15). We show here that GRIM-19 overrides cellular transformation caused by constitutive STAT3 and suppresses the expression of endogenous genes involved in cell proliferation. GRIM-19 also inhibits injury-induced cell migration and formation of tumors in vivo. These data for the first time show a direct tumor-suppressive effect of GRIM-19 and establish an oncogene-tumor suppressor like relationship between STAT3 and GRIM-19.

Materials and Methods

Plasmids and antibodies. The pRC/CMV flag-STAT3C (A661C and N C) that expresses a constitutively dimerizing version of STAT3 (15) was a gift from Jackie Bromberg (Sloan-Kettering Cancer Center, New York, NY). Mammalian expression vector pCXN2-myc and its derivative pCXN2-GRIM19-myc were described elsewhere. TKS3-Luc, cyclinB1-Luc, and cdc2-Luc were described in our previous publication (14). APRE-Luc contains three copies of acute phase response element (APRE) upstream of the SV40 minimal promoter. c-fos-Luc (16) and Bcl-XL-Luc (17) were described earlier. Lentiviral expression vectors carrying GRIM-19 and a scrambled short hairpin RNA (shRNA) were purchased from Open Biosystems, Inc., and virus stocks were prepared as recommended by the supplier. Antibodies specific for STAT3, phospho-STAT3-Y705, and phospho-STAT3-S727 (Cell Signaling Technology); actin, myc-epitope, and Flag-epitope (SigmaAldrich); and Ki-67 (Oncogene Science) were used in these studies. Establishment of stable cells. Rat fibroblast 3Y1 cells were grown in DMEM supplemented with 10% heat fetal bovine serum (FBS), 100 units/mL penicillin, and 100 Ag/mL streptomycin. Cells were electroporated either individually or in combination with pCXN2-myc, STAT3C, or pCXN2-GRIM 19 (1 Ag of each plasmid) using the Nucleofector technology (Amaxa Biosystems). Transfection efficiency was f45% with this method. Twentyfour hours posttransfection, cells were selected with G418 (750 Ag/mL) for
663

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). S. Kalakonda and S.C. Nallar contributed equally to this study. Requests for reprints: Dhananjaya V. Kalvakolanu, University of Maryland Cancer Center, Howard Hall, Room 324, 660 W. Redwood Street, Baltimore, MD 21201. Phone: 410-328-1396; Fax: 410-328-6559; E-mail: dkalvako@umaryland.edu. I2007 American Association for Cancer Research. doi:10.1158/0008-5472.CAN-07-0031

Cancer Res 2007; 67: (13). July 1, 2007

6212

www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Antitumor Effects of GRIM-19

Table 1. Primers used in this study

Primers used for quantitative real-time PCR Transcript cyclin b1 c-myc pdk 4 sod2 fas rpl 32 Primer sequence Fwd: 5-GGAAACATCTGGATGTGCACCTGC-3 Rev: 5-GGATGGCTCTCATGTTTCCAG-3 Fwd: 5-CGCGCCCAGTGAGGATATC-3 Rev; 5-CCACATACAGTCCTGGATGAT-3 Fwd: 5-GGACTTCGGTTCAGAAAATGC-3 Rev: 5-GCCCAGAAGACCAGAAAG-3 Fwd: 5-GCCTTACGACTATGGCGC-3 Rev: 5-GGCTCAGGTTTGTCCAG-3 Fwd: 5-CTGGAATCCCAAGTCCTGAA-3 Rev: 5-TGATACCAGCACTGGAGCAG-3 Fwd: 5-TTAAGCCTAACTGGCGGAAACC-3 Rev: 5-CAGTAAGATTTGTTGCACATCAGC-3 Primers used for chromatin immunoprecipitation assay Gene bcl-2 bcl-xL cyclin b1 Primer sequence Fwd: 5-ACATCACAGGACTTCTGCGA-3 Rev: 5-TCTCAAAGGTTTCCATTGAA-3 Fwd: 5-GGATTTGAATGTAGGTGGTG-5 Rev: 5-GAAGGGAGAGAAAGAGCTTC-3 Fwd: 5-TGGACCTCTCGTCACTTCGGTC-3 Rev: 5-GAAAGTACTTGATATAATAACG-3 Product size (bp) 321 295 219 Product size (bp) 224 280 212 227 214 200

10 to 12 days. Drug-resistant clones were isolated and expanded. All gene expression studies were conducted using pools of colonies (n = 150) to avoid a clonal bias. Immunoblotting. An equal amount of total protein was resolved by 10% SDS-PAGE. Proteins were transferred to polyvinylidene difluoride membrane and probed with specific antibodies. Blots were developed using enhanced chemiluminescence reagent (Pierce). Gene expression and reporter analyses. Total RNA was extracted using RNA-Bee reagent (Tel-Test, Inc.) as recommended by the manufacturer. Primers for reverse transcription-PCR designed for different genes are presented in the table. About 5 Ag of total RNA were reverse transcribed using a commercially available Superscript enzyme (Invitrogen, Inc.). A portion of cDNA from each reaction was used for PCR using the following conditions: each cycle consisted of denaturation at 95jC for 30 s, annealing at 55jC for 30 s, extension at 72jC for 30 s, and a final extension at 72jC for 5 min. Quantitative real-time PCR was done using a Stratagene MX3005P realtime PCR machine. For all samples, RT PCR amplifications were done in a final volume of 30 AL (20 ng of cDNA, 200 Amol/L of primers, 0.4 unit of thermal DNA polymerase IDPOL, and 0.1 SYBR Green). PCR was carried out using the following cycling conditions: a hot start at 95jC for 10 min, followed by 40 cycles of 95jC for 30 s and 55jC for 30 s. The mRNA coding for the ribosomal protein L32 (rpl 32) was used as an internal control. Expression of different genes was normalized to rpl 32 and C t values (threshold cycle) were estimated. Mean relative abundance of each transcript was calculated. All reactions were done in triplicate. Each experiment was repeated for three separate batches of RNAs. Primers for quantitative real-time PCR were shown in Table 1. Cells were transfected with different luciferase reporters along with a h-actin-h-galactosidase plasmid using the Lipofectamine plus reagent as described earlier (14). Luciferase activity was normalized to that of h-galactosidase. All experiments were repeated at least thrice. Chromatin immunoprecipitation was carried out using commercially available kits (Upstate Biotech) as described earlier (18). Primers used for specific genes are listed in Table 1.

Soft-agar colony formation assay. For soft-agar assays, bottom layer containing 0.7% agarose with growth medium plus 10% fetal bovine serum was poured into six-well plates. Then, 10,000 cells were mixed with 0.35% agarose in DMEM containing 10% FBS and plated onto the existing agar layer. Plates were incubated for f3 weeks in a humidified incubator at 37jC. The colonies were counted in four to five random fields from each of the duplicate samples by using a microscope at 100 magnification. In vitro wound healing assay. 3Y1 cells stably expressing various plasmids were grown in DMEM with 5% FBS and penicillin-streptomycin in a six-well dish. Confluent monolayer was scratched with a 200-AL pipette tip to make a wound. The monolayer was washed twice with PBS and then supplemented with medium and incubated for 4 h at 37jC. Cell migration into wounded area was followed microscopically. Images were captured at the interface of unwounded and wounded areas. At least quadruplicates were used per cell line. These experiments were repeated thrice. Cell migration assays. Cell migration assay was done using Transwell chambers (6.5-mm diameter, 8-Am pores, Transwell; Costar Corp.) containing a polycarbonate membrane. The lower chamber was filled with 0.6 mL of DMEM with 10% FBS; the upper portion of insert is filled with 25,000 cells in 0.1-mL DMEM without serum and incubated at 37jC for 24 h. Nonmigrated cells in the upper membrane surface were removed gently with a cotton swab and the migrated cells attached to bottom surface of the membrane were fixed with ice-cold 100% methanol and stained with 0.01% crystal violet in 20% ethanol. The number of migrated cells was determined by dye elution with 5% acetic acid and 5% methanol; absorbance was measured at 595 nm. In parallel, some membranes were stained similarly without the removal of nonmigrated cells for each cell line. The absorbance data from these controls provided a measure of the total number of cells plated in that well. Absorbance obtained with the migrated cells (experimental wells, where nonmigrated cells were removed) was expressed as a percent of total number of cells in each case. 3Y1 cells without any transfection were used as a baseline for these experiments (100%). All experiments were done thrice in triplicate.

www.aacrjournals.org

6213

Cancer Res 2007; 67: (13). July 1, 2007

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Cancer Research Tumorigenic assays. These assays were conducted as described earlier (3). Three- to four-week-old athymic nude (nu/nu) NCr mice (Taconic) were used in the study. Procedures involving animals and their care were conducted in conformity with the institutional guidelines that are in compliance with national and international laws and policies (EEC Council Directive 86/609, OJL 358, 1 Dec. 1987, and the NIH Guide for the Care and Use of Laboratory Animals, NIH Publication No. 85-23, 1985). Each experimental group contained 10 mice, each bearing one tumor. Cells (1 106) were inoculated into flanks in the mid-axillary line. Tumor volume was calculated using caliper measurements and the formula for a prolate spheroid: V = (4/3) pa 2b, where 2a is the minor axis and 2b is the major axis of prolate spheroid. Students two-tailed t test was used to assess the statistical significance of difference between pairs of means of tumor volume. Tumor specimens were fixed, sectioned, and stained with various antibodies using a commercially available Vectastain ABC kit (Vector Labs).

Results
GRIM-19 inhibits cell growth in cell lines expressing high levels of STAT3. To determine the antitumor effects of GRIM-19, we have used a human prostatic cell line, PC3. This cell line has been shown earlier to express STAT3 at higher levels compared with normal prostatic epithelium (19). In addition, STAT3 in this cell line is constitutively phosphorylated at the Tyr705 residue. An empty mammalian expression vector, pCXN2, or its derivative expressing GRIM-19 as a myc-tagged protein was electroporated into these cells and cells were selected with G418. Following selection for 3 weeks, the surviving cell colonies were fixed and counted. As shown in Fig. 1A and B, GRIM-19 significantly decreased the colony formation by PC3 cells. Similar effects of GRIM-19 were observed when these transfectants were grown on soft agar (data not shown). Similar results were obtained with another prostate cancer cell line, LnCaP, indicating a general antitumor effect of GRIM-19 (data not shown). In addition, we have shown a similar effect of GRIM-19 on MCF-7, BT-20, T47D (breast carcinoma), HeLa, and ACHN (renal cell carcinoma) cells in other reports (57), indicating its global inhibitory effect on a number of tumor cell lines. We transfected these cells with TKS3-Luc, a

Figure 1. Growth-suppressive effect of GRIM-19 on a human prostate tumor cell line, PC3. A, cells (105) were transfected with an expression vector for GRIM-19 (0.5 Ag each) and selected with G418 for 3 wk. The surviving colonies were fixed, stained with sulforhodamine B, and photographed. B, quantification of colonies. Columns, mean of triplicates; bars, SE. Effects of GRIM19 on TKS3-luc (C ) and CMV-Luc (D ) gene expression. Fold induction was calculated based on the luciferase activity (750 RLU) observed with pGL3promoter vector, which does not have any enhancer.

reporter that responds to the transcriptional activity of STAT3, to determine the effect of GRIM-19 on STAT3-induced transcription (Fig. 1C). A significantly lower Luc activity was noted in the presence of GRIM-19 compared with the empty vector control. In contrast, GRIM-19 had no discernible effect on cytomegalovirus (CMV) enhancerdriven expression of luciferase (Fig. 1D). Thus, GRIM-19 seems to ablate constitutive STAT3 activity in these cells. Effect of GRIM-19 on STAT3C-induced cellular transformation. Because the above cancer cell lines may have undergone several genetic changes including a constitutive up-regulation of STAT3, we examined next if these effects were truly directed against STAT3-mediated cellular transformation and downstream biological activities. Therefore, we used the rat fibroblastic cell line 3Y1 for these experiments because it has a lower frequency of spontaneous transformation compared with the NIH 3T3 cell line. This cell line does not form tumors on its own, unless transformed by a potential oncogene. Cells were transfected with a genetically engineered STAT3-mutant, S3C, which could spontaneously dimerize independently of its tyrosine residue. The constitutive dimerization of this protein is mediated by a disulfide bond formed between cysteine residues (15). We also coexpressed GRIM-19 to determine its effect on S3C-induced cellular transformation. Transfection of S3C into cells caused a robust cellular transformation, as revealed by numerous soft-agar colonies (Fig. 2A and B). In the presence of GRIM-19, significantly (P > 0.001) fewer S3Cdependent soft-agar colonies were observed. These colonies were distinctly smaller in size than those formed by S3C transfection. Neither the empty vector (pCXN2) nor GRIM-19 alone had an effect on cellular transformation. Thus, GRIM-19 suppresses S3C-induced cellular transformation. Because the above experiment shows the effects only of cotransfected GRIM-19 on cellular transformation, we stably transfected a GRIM-19 expression vector into a 3Y1 cell line that was previously transformed with S3C. This cell line was a pool of several S3C-expressing colonies. The resultant colonies were used for the formation of soft-agar colonies and cell growth assays. To avoid a clonal bias, all experiments used pools of stable colonies. As expected, the S3C transformants formed large soft-agar colonies. Transfection of GRIM-19, but not the empty vector, into these cells caused a marked reduction in the number of soft-agar colonies (Fig. 2C). We have also ascertained the expression of S3C and GRIM-19 in these cells (Fig. 2D). Because the S3C protein is flag tagged, we detected its expression with a flag-specific antibody. A 95-kDa protein corresponding to S3C was detected only in cells transfected with S3C expression vector but not in those transfected with pCXN2 and GRIM-19 alone. We observed a higher amount of S3C protein in cells carrying both GRIM-19 and S3C. We presume it to be due to an enrichment of high S3C-expressing cells in the pools following transfection with GRIM-19. It is possible that the low S3C expressors in this pool may have died due to the proapoptotic activity of GRIM-19. As expected, the myc-tagged GRIM-19 protein was detected in cells transfected with pCXN2GRIM-19 myc only. A comparable loading of protein was ensured by probing these blots with actin-specific antibodies. Consistent with its inhibitory effect on colony formation, the myc-GRIM-19 protein coimmunoprecipitated the constitutive STAT3 protein. We next checked the status of tyrosine and serine phosphorylation in these cells (Fig. 2E). As expected, there was no constitutive tyrosine phosphorylation of S3C. The positive control [interleukin-6 (IL-6)treated MCF-7 cell extract] showed normal tyrosine phosphorylation. There was no difference in phosphorylation of

Cancer Res 2007; 67: (13). July 1, 2007

6214

www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Antitumor Effects of GRIM-19

Figure 2. 3Y1 cells were transfected with the indicated vector and used for soft-agar growth assays in the presence of G418 (750 Ag/mL). A, photomicrographs of representative fields from various transfections. B, quantification of colony formation. Columns, mean of triplicates; bars, SE. C, a S3C-expressing cell line was transfected with various plasmids and soft-agar colony formation was monitored. D, Western blot (WB ) and immunoprecipitaton (IP ) analyses of the expression of transfected genes in various cells. Actin is an internal control. E, status of STAT3 phosphorylation in the transfectants. Total protein (60 Ag) was used for the blots in D and E.

STAT3 Ser727 among various transfectants. The steady-state serine phosphorylation of STAT3 is also found in many other cell types (20). As anticipated, a higher amount of total STAT3 was seen in the S3C transfectants, which reflects the STAT3 protein derived from endogenous gene and the transfected plasmid. We have also determined its constitutive nuclear presence by a gel-mobility shift assay with a STAT3 binding element as a probe and nuclear extracts from the S3C- and S3C/G19transfected cells and by a Western blot analysis of the same nuclear extracts (Supplementary Figs. S1 and S2). In contrast, the endogenous STAT3 in 3Y1 cells expressing either an empty vector or GRIM-19 translocated to the nucleus only after IL-6 stimulation. Importantly, GRIM-19 did not significantly affect either the IL-6induced nuclear translocation of endogenous STAT3 or the ligand-independent S3C in these cells. GRIM-19 inhibits cell injury and mitogen-driven migration of S3C-expressing cells. We next used these cells for injuryinduced and mitogen-induced transwell migration assays. In the injury-induced migration assays, an incision was introduced into a confluent monolayer of cells growing in a dish. Cell migration into the denuded area was monitored 4 h later. As shown in Fig. 3A, little or no cells migrated into the injured area when cells were expressing GRIM-19 or pCXN2. However, a number of S3C transfectants migrated under the same conditions. Such rapid migration was virtually absent in cells expressing GRIM-19. By 18 h, all cell lines reached confluence (data not shown). These data clearly show that the presence of S3C promotes rapid cell motility.

Cell motility was also assessed using another strategy. Cells were grown in Boyden chambers and their transwell migration toward growth medium was monitored by a colorimetric assay (Fig. 3B). In these experiments, cellular migration obtained with nave 3Y1 cells (control) was considered as 100% and the data obtained with experimental samples were expressed as a percentage of control migration. As shown in the figure, a significant (P > 0.01) percentage of cells expressing S3C migrated in these assays compared with the vector alonetransfected cells. GRIM-19 completely suppressed S3C-dependent migration. GRIM-19 alone did not have any statistically different effect on cell motility. Lastly, cells expressing S3C and GRIM-19 combinations were also used for growth assays using sulforhodamine B staining as previously described. Cell growth was monitored after 3 days (Fig. 3C). On day 0 (7 h after plating the cells), the total numbers of initially plated cells were comparable. However, the S3C-transfected cells grew robustly compared with the pCXN2- or GRIM-19 alone transfected 3Y1 cells. In the presence of GRIM-19, the rapid growth induced by S3C was significantly (P > 0.005) reduced. GRIM-19 suppresses the transcriptional induction of a number of S3C-regulated genes. We next investigated the molecular basis for the loss of cell growth function. Previous studies in human cell lines have shown an up-regulation of a number of genes involved in growth control. We have selected five genes for these studies whose rodent homologues are known. The relative expressions of the mRNAs of these genes were quantified using quantitative real-time PCR. Two of these genes, fas (a death

www.aacrjournals.org

6215

Cancer Res 2007; 67: (13). July 1, 2007

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Cancer Research

receptor) and pyruvate dehydrogenase kinase 4 (pdk4; a protein kinase involved in the regulation of carbohydrate metabolism), are down-regulated in tumor cells expressing high STAT3. Three others, c-myc, cyclin b1, and superoxide dismutase 2 (sod2), are implicated

in tumor cell proliferation. The expression of cyclin B1 (Fig. 4A), c-myc, and sod2 (Supplementary Fig. S4A) was increased over the control levels severalfold (2.5- to 3.5-fold depending on the gene) in the presence of S3C. GRIM-19 inhibited the S3C-dependent increase

Figure 3. Effect of GRIM-19 on S3C-dependent migration of cells to wounded area in various cell lines (A). Magnification, 100. Note the rapid migration of S3C-expressing cells, but not the controls, into the wounded area. These experiments were repeated thrice. Quadruplicate plates were used per cell line. A representative figure is shown. The vertical line in each photograph shows the edge of injured site. B, invasion assay was done using Transwell inserts. Ten percent serum was used as an attractant in the lower chamber of a 24-well tissue culture dish. Cells (25,000) were seeded in the upper chamber of insert and incubated for 24 h; migrated cells were fixed, stained with crystal violet, and the bound dye was eluted and quantified at 595 nm. Absorbance obtained with 3Y1 cells was used as 100%. C, growth of various cell lines as measured by a colorimetric assay with sulforhodamine B. n = 8 per cell line per time point.

Cancer Res 2007; 67: (13). July 1, 2007

6216

www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Antitumor Effects of GRIM-19

Figure 4. Effect of GRIM-19 on S3Cdependent gene expression. Levels of endogenous transcripts cyclin b1 and fas genes were quantified using quantitative real-time PCR. Transcript abundance was measured after normalizing to an internal control, rpl 32. Baseline expression of cyclin B1 gene observed in the pCXN2 vectortransfected cells was subtracted from the experimental samples. A significant repression of fas gene by S3C and its reversal by GRIM-19 (P > 0.01) were observed in these studies. For measuring the effect of GRIM-19 on S3C-driven transcription, the cyclin B1-Luc reporter and a h-galactosidase reporter (0.2 Ag each) were transfected into the indicated 3Y1 stable cell lines. Luciferase activity was normalized to that of h-galactosidase and plotted. Columns, mean luciferase activity; bars, SE. Relative luciferase activity was determined by comparing with the data obtained from pCXN2-3Y1 cells. B, a GRIM-19specific shRNA derepresses S3C-dependent expression of the reporters. S3C/GRIM-19 cells were infected with lentiviral vectors carrying the indicated shRNAs (overnight) and then transfected with cyclin B1-Luc. n = 3 per sample. Experiment was repeated twice. shRNA-mediated knockdown of GRIM-19 in the S3C/GRIM19 cell line was monitored by Western blot analyses. n = 3 per sample. Experiment was repeated twice. Representative data are shown. C, chromatin immunoprecipitation (ChIP ) assays showing the constitutive recruitment of STAT3 to endogenous promoters and IFN/ RAinduced recruitment of GRIM-19 in PC3 cells. STAT3 is constitutively activated in this cell line. Gene promoterspecific primers (Table 1) were used for immunoprecipitating the formaldehyde cross-linked soluble chromatin after treatment with IFN/RA. PCR was carried out with gene promoterspecific primers (indicated on the left) using the Immunoprecipitated genomic DNA as a template (30 cycles). For input controls (column C1 ), soluble chromatin (1/100 of sample used in reactions shown in columns C1 and C3) was used as a template without immunoprecipitation. NR IgG, normal rabbit IgG (control for STAT3 antibodies). IgG2a, mouse IgG2a (control for GRIM-19 antibodies). Western blot and immunoprecipitation analyses of the PC3 cell lysates were conducted with the indicated antibodies.

in the expression of these genes. These levels were comparable to those expressing GRIM-19 alone. In the case of c-myc, GRIM-19 caused a stronger repression of the S3C-induced expression of the gene, even below the baseline level. A converse picture was seen with fas (Fig. 4A) and pdk4 mRNAs (Supplementary Fig. S4A). The expression of both these mRNAs was significantly (P > 0.01) repressed in the presence of S3C. GRIM-19 alone did not have an effect on these genes. However, GRIM-19 reversed the negative regulatory effects of S3C on these genes. Thus, GRIM-19 caused a

reversion of S3C-transformed cells to a pretransformed state with respect to certain gene expression. To show that these are indeed transcriptional effects of GRIM-19, we transfected luciferase reporter plasmids driven by the native promoters of four STAT3 regulated genes, cyclin b1 (Fig. 4A), cdc2, c-fos, and bcl-xL (Supplementary Fig. S4B), into 3Y1 cell lines expressing pCXN2, S3C, GRIM-19, and S3C/GRIM-19 and measured gene expression. Two others reporters, S3-Luc and APRE-Luc, are driven by minimal S3 binding elements to which STAT3 binds

www.aacrjournals.org

6217

Cancer Res 2007; 67: (13). July 1, 2007

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Cancer Research

(Supplementary Fig. S4B). The expression of all reporters was induced severalfold by S3C. In every case, GRIM-19 overrode S3C-induced transcription of the reporter. Notably, it was lowered below the vector alonetransfected lines. To show that this is a true

Figure 5. Effect of GRIM-19 on the in vivo growth of S3C-expressing tumors. A, 3Y1 cell lines expressing various genes were transplanted s.c. into athymic nude mice (n = 10) and tumor growth was monitored. B, immunohistochemical analyses of the tumors grown in nude mice. Tumors were grown for 18 d, excised, fixed, and an immunohistochemical analysis was done with antibodies indicated on top. C, effect of GRIM-19 on the growth of a PC3 prostate tumor in nude mice (n = 10). Tumor cells were transplanted as in (A) and growth was monitored.

effect of GRIM-19, cells were transfected with lentiviral shRNA expression plasmids capable of targeting GRIM-19 into S3C/GRIM19. Two reporters, cyclin B1-Luc (Fig. 4B ) and cdc2-Luc (Supplementary Fig. S4C), were used for these experiments. As expected, the control vector lacking GRIM-19 sequences and scrambled GRIM-19 sequence were unable to reinstate S3C-driven transcription of both reporters (Fig. 4B). However, the lentiviral vector carrying GRIM-19specific shRNA restored S3C-driven transcription of these reporters. These data confirm that no other genetic changes other than that of GRIM-19 are responsible for the repression of S3C-driven transcription. We have also confirmed the knockdown of the expression of transfected GRIM-19, but not that of S3C or actin, in these cells by GRIM-19specific shRNA (Fig. 4B). The scrambled control and empty vector did not have any effect on GRIM-19 expression. We next checked if the gene-repressive effects of GRIM-19 were due to a recruitment of GRIM-19 to the STAT3-regulated gene promoters. We examined these aspects in the human tumor cell line PC3 in which endogenous STAT3 is constitutively active and is located in the nucleus. Because the gene promoters and their sequences of most STAT3-induced genes in rat cells have not been well characterized, we carried out these studies in a human cell line. Furthermore, GRIM-19 inhibits STAT3-driven transcription in these cells (Fig. 1). These cells were treated with IFN/RA, a well-known inducer of GRIM-19 (5, 14), for various lengths of time. Chromatin was cross-linked with formaldehyde, and soluble chromatin was immunoprecipitated with GRIM-19specific or isotype control antibodies. The immunoprecipitation products were used for PCR with gene-specific primers (Fig. 4C). As expected, GRIM-19specific immunoglobulin G (IgG), but not the isotype control IgG2a, immunoprecipitated the bcl-xL, bcl-2, and cyclin b1 gene promoters. More importantly, GRIM-19 recruitment to these promoters was induced in response to IFN/RA. These differences in GRIM-19 recruitment to these gene promoters do not seem to be due to variations in the amounts of total chromatin input into these reactions (Fig. 4C). We also checked the recruitment of STAT3 to these promoters. As expected, STAT3 is constitutively present at these promoters, as revealed by immunoprecipitation of these gene promoters by STAT3-specific IgG but not normal rabbit IgG. Thus, IFN/RAinduced recruitment of GRIM-19 to the STAT3-dependent gene promoters seems to repress these genes. This seems to occur due, in part, to an IFN/RAinduced expression of GRIM-19. STAT3 levels did not significantly change under these conditions. Consistent with these observations, endogenous GRIM-19 interacted with endogenous STAT3 following IFN/RA treatment. In vivo effects of GRIM-19 on tumor suppression. To determine if S3C-transformed cells have differential abilities to form tumors in the presence of GRIM-19, 3Y1 cell lines carrying various expression vectors were transplanted into athymic nude mice and tumor formation was monitored over a period of 25 days. As shown in Fig. 5A, S3C-transformed cells formed large tumors, which grew robustly compared with 3Y1 cells transfected with pCXN2. GRIM-19 significantly knocked down the ability of S3C to form tumors (P > 0.001). Remarkably, GRIM-19 even depressed the baseline tumor formation by 3Y1 cells (P > 0.01). We also examined the cellular bases for the loss of tumor formation by examining the intratumoral levels of proliferation-associated antigen Ki-67, STAT3, and GRIM-19 using specific antibodies to these proteins (Fig. 5B). The S3C tumors were highly positive for Ki-67. The tumors carrying GRIM-19 were weakly positive for STAT3 and Ki67. More importantly, the S3C/GRIM-19 tumors have an extremely

Cancer Res 2007; 67: (13). July 1, 2007

6218

www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Antitumor Effects of GRIM-19

low level of Ki-67 expression, although they have a high STAT3 level. As expected, these tumors are also GRIM-19 positive. Thus, GRIM-19 overrides the S3C-dependent tumor proliferation in vivo. Although the presence of S3C augmented mitogen- and injuryinduced motility in vitro (Fig. 3), tail vein injection of 3Y1 cells carrying S3C did not cause significant metastases in this model (data not shown). It is possible that these cells may not have acquired other metastatic characteristics required for in vivo motility, such as extravasation through blood vessels. To examine the antitumor effect of GRIM-19 to a human tumor, PC3 cells, which express high STAT3, were stably transfected with GRIM-19 and used for tumor formation assays in vivo (Fig. 5C). Unlike 3Y1-tumors, PC3 tumors form at a slower pace. Therefore, we monitored tumor development over a period of 11 weeks. The PC3 line expressing pCXN2 formed large tumors, which significantly differed from GRIM-19expressing tumors in terms of growth pattern as early as 3 weeks after initial transplantation. By the end of 7 weeks, the PC3 tumors were 6-fold bulkier than the GRIM-19 expressing tumors. To examine their metastatic activity, these cells (1 105 per mouse) were injected into mice via the tail vein. In these experiments, 10 mice per group were used. In these assays, the PC3 cells formed a total of 24 and 11 metastases in the liver and kidney tissues, respectively, per group (Supplementary Table S1). Only one mouse did not show any metastases in this group. Metastases occurred in liver and kidney of the same mouse. No metastases were found in other tissues. In contrast, the same number of GRIM-19 transfected PC3 cells formed five and two metastases in the liver and kidney, respectively, per group. About 50% of the mice injected with cells expressing GRIM-19 did not develop any metastases.

Discussion
STAT3 is a major transcription factor for mediating cytokine responses, and its inactivation in experimental animals leads to the development of several pathologies including embryonic lethality (20). In most cases, cytokine-induced transient phosphorylation of STAT3 at Tyr705 and Ser727 is sufficient for driving these responses (20). A feedback inhibition of JAKs by STAT3-induced suppressor of cytokine signaling (SOCS)-3 protein (21), dephosphorylation at Y705 (22), and nuclear export of STAT3 (23) seems to prevent incessant expression of STAT3-dependent genes in normal cells. In contrast to these scenarios, STAT3 is constitutively activated in a number of human tumor cell lines and primary tumors, where activated oncogenes have been suspected to induce uninterrupted tyrosine phosphorylation of STAT3 (24). The most convincing evidence for an oncogenic STAT3 comes from a recent study of npm-alk oncogene, which obligatorily depends on STAT3 for cellular transformation in lymphocytes (25). Whereas excessive tyrosine phosphorylation by oncogenes is one side of the coin, it does not provide a full picture of the events occurring during STAT3dependent tumor growth promotion. It is conceivable that inactivation of STAT3 inhibitors is also a major step involved in pushing the cell into a transformed state, and a mere tyrosine phosphorylation alone may be insufficient. Lastly, the ability of STAT3-Y705F mutant to activate gene transcription by simple overexpression in nontransformed cells also indicates the potential tyrosine phosphorylationindependent mechanisms of cell proliferation (26). Thus, the roles of STAT inhibitors in cellular transformation have been poorly investigated. To date, three STAT3 inhibitors, SOCS3, protein inhibitor of activated STAT3 (PIAS3; ref. 27), and GRIM-19 (5), have been identified. Apart from the

observations that socs3 gene expression is inhibited due to methylation in some nonsmall-cell lung cancers (28), hepatomas (29), and head and neck cancers (30), there is no direct experimental proof that these SOCS proteins counteract the constitutive STAT3 in tumors. PIAS3 seems to act more like a SUMO ligase to regulate the function of steroid receptors than STAT3 itself (3133). Thus, it is conceivable that the transiently (normal) and chronically (tumors) active states of STAT3 can be regulated by different inhibitors. Using a genetic strategy, our lab has isolated a novel gene product, GRIM-19, whose inactivation promotes cell growth, and its overexpression induces growth arrest and apoptosis in a number of cell types (5). GRIM-19 is present in both cytoplasmic and nuclear compartments (5, 14). A first indication that GRIM-19 may be an antioncogenic protein came from observations that DNA viral oncoproteins such as vIRF1, human papillomavirus-E6, and SV40 T antigen inhibit GRIM-19driven apoptosis (34). Subsequent studies by others have shown the presence of GRIM-19 inhibitors in some esophageal tumor cell lines (8) and the occurrence of mutations in the gene coding for GRIM-19 in some thyroid cancers (35). Using mass spectrometric approaches, we have recently shown the loss of GRIM-19 expression in a number of primary renal cell carcinomas (6). We and others have isolated STAT3 as a GRIM-19 binding protein using a yeast two-hybrid screen and showed that GRIM-19 inhibits cytokine-induced STAT3-driven gene expression (13, 14). Whereas most of these studies point out a potential role for GRIM19 as an antioncogenic protein, no direct evidence for it in cellular transformation has been presented. Here, we have used a model in which an experimentally engineered constitutive STAT3, S3C, alone transforms cells in culture. In this model, the single genetic change that transforms cells is S3C. We have used several different methods including softagar growth assays, injury- and mitogen-induced cell migration, and in vivo growth of tumor cells as indices for its direct antitumor actions (Fig. 5 and Supplementary Table S1). More importantly, GRIM-19 was able to reverse the gene expression program in S3Ctransformed cells quite significantly, as evidenced by repression of a number of proliferation-associated genes, identified in gene expression profiling studies as genes regulated by STAT3 (26, 36), and by derepression of antigrowth genes such as Fas (Figs. 4A and Supplementary Fig. S4A). Because STAT3 is known to activate the genes involved in cell proliferation and inhibition of apoptosis, the GRIM-19 effects on STAT3C-induced cell growth are probably exerted through an inhibition of these processes. These results are consistent with other reports that the fas gene is repressed by STAT3 (36) and IFN/RA treatment induces Fas and Fas ligand expression in cells (37), which can suppress tumor growth. Indeed, the presence of GRIM-19 augmented Fas-mediated apoptosis in the S3C-expressing cells (see Supplementary Fig. S3). Interestingly, a metabolic regulatory enzyme, PDK4, is also repressed by S3C, and GRIM-19 restores its expression (Supplementary Fig. S4). PDK4 inactivates the mitochondrial pyruvate dehydrogenase complex (PDHc) via a direct phosphorylation of its subunits (38). At this stage, it is not clear why S3C suppresses this gene. It may have an undefined relationship to altered glucose metabolism in aggressively growing tumors, wherein glycolytic generation of ATP supersedes that of tricarboxylic acid cycledependent oxidative phosphorylation (39). Indeed, c-myc, a target of STAT3, drives up the expression of glycolytic enzymes (40). There are no known STAT3 binding sites in the promoters of human and rodent pdk genes (41); hence, the GRIM-19dependent relief of STAT3-induced inhibition of this gene must occur in an indirect manner. That said, we have also shown

www.aacrjournals.org

6219

Cancer Res 2007; 67: (13). July 1, 2007

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research

DOI:10.1158/0008-5472.CAN-07-0031
Cancer Research

that these effects are a direct result of recruitment of GRIM-19 to several known STAT3-regulated gene promoters. GRIM-19 does not seem to destabilize the promoter-bound STAT3 (Fig. 4C). This would suggest that GRIM-19 associates with constitutively active STAT3 and prevents gene expression. Importantly, the negative effects of GRIM-19 on S3C-driven gene transcription can be eased following introduction of a GRIM-19specific shRNA into the cells, indicating its direct effect on STAT3 (Fig. 4B). These effects, therefore, do not seem to be a result of secondary changes that occurred during selection. Lastly, GRIM-19 binds to the transactivation domain of STAT3. Mutation of the S727 residue prevented the binding of GRIM19 to STAT3 (14). It seems that GRIM-19 prevents the formation of transcriptional coactivator complexes with STAT3.4 Apart from the experimentally engineered S3C, we have also shown that GRIM-19 suppresses cell growth and tumor formation by a human prostate

tumor cell line in which STAT3 is constitutively active (Fig. 5C). Using animal and in vitro models, we have recently shown that plasmid-based shRNAs can directly target STAT3 and, therefore, the expression of STAT3-induced genes and cell proliferation in this cell line (19). Lastly, GRIM-19 can inhibit tumor growth on intratumoral administration of a plasmid coding for it.5 In summary, we conclude that GRIM-19 is a direct inhibitor of chronically active STAT3. Our observations establish a potential oncogene-tumor suppressor like relationship between STAT3 and GRIM-19. That mutations in grim-19 gene (35) and a loss of its expression occur in primary tumors (6) and viral oncoproteins inhibit its proapoptotic function (34) qualifies it to be a novel type of tumor suppressor.

Acknowledgments
Received 1/4/2007; revised 4/6/2007; accepted 5/1/2007. Grant support: NIH grants CA105005 and CA78282 (D.V. Kalvakolanu), CA095020 (D.J. Lindner), and ES011863 (S.P. Reddy). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

S. Kalkonda, S.C. Nallar, and D.V. Kalvakolanu, in preparation. L. Zhang, L. Gao, Y. Li, G. Lin, Y. Shao, K. Ji, H. Yu, D.V. Kalvakolanu, D. J. Kopecko, X. Zhao, and D. Xu, submitted.
5

References
1. Dang CV, ODonnell KA, Zeller KI, et al. The c-Myc target gene network. Semin Cancer Biol 2006;16:25364. 2. Milde-Langosch K. The Fos family of transcription factors and their role in tumourigenesis. Eur J Cancer 2005;41:244961. 3. Lindner DJ, Borden EC, Kalvakolanu DV. Synergistic antitumor effects of a combination of interferons and retinoic acid on human tumor cells in vitro and in vivo . Clin Cancer Res 1997;3:9317. 4. Moore DM, Kalvakolanu DV, Lippman SM, et al. Retinoic acid and interferon in human cancer: mechanistic and clinical studies. Semin Hematol 1994;31:317. 5. Angell JE, Lindner DJ, Shapiro PS, Hofmann ER, Kalvakolanu DV. Identification of GRIM-19, a novel cell death-regulatory gene induced by the interferon-h and retinoic acid combination, using a genetic approach. J Biol Chem 2000;275:3341626. 6. Alchanati I, Nallar SC, Sun P, et al. A proteomic analysis reveals the loss of expression of the cell death regulatory gene GRIM-19 in human renal cell carcinomas. Oncogene 2006;25:713847. 7. Chidambaram NV, Angell JE, Ling W, Hofmann ER, Kalvakolanu DV. Chromosomal localization of human GRIM-19, a novel IFN-h and retinoic acid-activated regulator of cell death. J Interferon Cytokine Res 2000; 20:6615. 8. Zhang X, Huang Q, Yang Z, Li Y, Li CY. GW112, a novel antiapoptotic protein that promotes tumor growth. Cancer Res 2004;64:247481. 9. Fearnley IM, Carroll J, Shannon RJ, et al. GRIM-19, a cell death regulatory gene product, is a subunit of bovine mitochondrial NADH:ubiquinone oxidoreductase (complex I). J Biol Chem 2001;276:383458. 10. Murray JG, Zhang B, Taylor SW, et al. The subunit composition of the human NADH dehydrogenase obtained by rapid one step immunopurification. J Biol Chem 2003;278:1361922. 11. Huang G, Lu H, Hao A, et al. GRIM-19, a cell death regulatory protein, is essential for assembly and function of mitochondrial complex I. Mol Cell Biol 2004;24:844756. 12. Huang G, Chen Y, Lu H, Cao X. Coupling mitochondrial respiratory chain to cell death: an essential role of mitochondrial complex I in the interferon-h and retinoic acid-induced cancer cell death. Cell Death Differ 2007;14:32737. 13. Lufei C, Ma J, Huang G, et al. GRIM-19, a deathregulatory gene product, suppresses Stat3 activity via functional interaction. EMBO J 2003;22:132535. 14. Zhang J, Yang J, Roy SK, et al. The cell death regulator

GRIM-19 is an inhibitor of signal transducer and activator of transcription 3. Proc Natl Acad Sci U S A 2003;100:93427. 15. Bromberg JF, Wrzeszczynska MH, Devgan G, et al. Stat3 as an oncogene. Cell 1999;98:295303. 16. Wang Y, Prywes R. Activation of the c-fos enhancer by the erk MAP kinase pathway through two sequence elements: the c-fos AP-1 and p62TCF sites. Oncogene 2000;19:137985. 17. Kirito K, Watanabe T, Sawada K, et al. Thrombopoietin regulates Bcl-xL gene expression through Stat5 and phosphatidylinositol 3-kinase activation pathways. J Biol Chem 2002;277:832937. 18. Zhang Q, Adiseshaiah P, Kalvakolanu DV, Reddy SP. A Phosphatidylinositol 3-kinase-regulated Akt-independent signaling promotes cigarette smoke-induced FRA-1 expression. J Biol Chem 2006;281:1017481. 19. Gao L, Zhang L, Hu J, et al. Down-regulation of signal transducer and activator of transcription 3 expression using vector-based small interfering RNAs suppresses growth of human prostate tumor in vivo . Clin Cancer Res 2005;11:633341. 20. Levy DE, Darnell JE. STATs: transcriptional control and biological impact. Nat Rev Mol Cell Biol 2002;3: 65162. 21. Alexander WS, Hilton DJ. The role of suppressors of cytokine signaling (SOCS) proteins in regulation of the immune response. Annu Rev Immunol 2004;22:50329. 22. Yamamoto T, Sekine Y, Kashima K, et al. The nuclear isoform of protein-tyrosine phosphatase TC-PTP regulates interleukin-6-mediated signaling pathway through STAT3 dephosphorylation. Biochem Biophys Res Commun 2002;297:8117. 23. Bhattacharya S, Schindler C. Regulation of Stat3 nuclear export. J Clin Invest 2003;111:5539. 24. Buettner R, Mora LB, Jove R. Activated STAT signaling in human tumors provides novel molecular targets for therapeutic intervention. Clin Cancer Res 2002;8:94554. 25. Chiarle R, Simmons WJ, Cai H, et al. Stat3 is required for ALK-mediated lymphomagenesis and provides a possible therapeutic target. Nat Med 2005;11:6239. 26. Yang J, Chatterjee-Kishore M, Staugaitis SM, et al. Novel roles of unphosphorylated STAT3 in oncogenesis and transcriptional regulation. Cancer Res 2005;65:93947. 27. Chung CD, Liao J, Liu B, et al. Specific inhibition of Stat3 signal transduction by PIAS3. Science 1997;278: 18035. 28. He B, You L, Uematsu K, et al. SOCS-3 is frequently silenced by hypermethylation and suppresses cell growth in human lung cancer. Proc Natl Acad Sci U S A 2003;100:141338.

29. Niwa Y, Kanda H, Shikauchi Y, et al. Methylation silencing of SOCS-3 promotes cell growth and migration by enhancing JAK/STAT and FAK signalings in human hepatocellular carcinoma. Oncogene 2005;24:640617. 30. Weber A, Hengge UR, Bardenheuer W, et al. SOCS-3 is frequently methylated in head and neck squamous cell carcinoma and its precursor lesions and causes growth inhibition. Oncogene 2005;24:6699708. 31. Junicho A, Matsuda T, Yamamoto T, et al. Protein inhibitor of activated STAT3 regulates androgen receptor signaling in prostate carcinoma cells. Biochem Biophys Res Commun 2000;278:913. 32. Kotaja N, Aittomaki S, Silvennoinen O, Palvimo JJ, Janne OA. ARIP3 (androgen receptor-interacting protein 3) and other PIAS (protein inhibitor of activated STAT) proteins differ in their ability to modulate steroid receptor-dependent transcriptional activation. Mol Endocrinol 2000;14:19862000. 33. Sentis S, Le Romancer M, Bianchin C, Rostan MC, Corbo L. Sumoylation of the estrogen receptor a hinge region regulates its transcriptional activity. Mol Endocrinol 2005;19:267184. 34. Seo T, Lee D, Shim YS, et al. Viral interferon regulatory factor 1 of Kaposis sarcoma-associated herpesvirus interacts with a cell death regulator, GRIM19, and inhibits interferon/retinoic acid-induced cell death. J Virol 2002;76:8797807. 35. Maximo V, Botelho T, Capela J, et al. Somatic and germline mutation in GRIM-19, a dual function gene involved in mitochondrial metabolism and cell death, is linked to mitochondrion-rich (Hurthle cell) tumours of the thyroid. Br J Cancer 2005;92:18928. 36. Ivanov VN, Bhoumik A, Krasilnikov M, et al. Cooperation between STAT3 and c-jun suppresses Fas transcription. Mol Cell 2001;7:51728. 37. Ma X, Karra S, Guo W, et al. Regulation of interferon and retinoic acid-induced cell death activation through thioredoxin reductase. J Biol Chem 2001; 276:2484354. 38. Patel MS, Korotchkina LG. Regulation of mammalian pyruvate dehydrogenase complex by phosphorylation: complexity of multiple phosphorylation sites and kinases. Exp Mol Med 2001;33:1917. 39. Gatenby RA, Gillies RJ. Why do cancers have high aerobic glycolysis? Nat Rev Cancer 2004;4:8919. 40. Shim H, Dolde C, Lewis BC, et al. c-Myc transactivation of LDH-A: implications for tumor metabolism and growth. Proc Natl Acad Sci U S A 1997;94: 665863. 41. Kwon HS, Harris RA. Mechanisms responsible for regulation of pyruvate dehydrogenase kinase 4 gene expression. Adv Enzyme Regul 2004;44:10921.

Cancer Res 2007; 67: (13). July 1, 2007

6220

www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on June 14, 2012 Copyright 2007 American Association for Cancer Research