Oncogene (2004) 23, 7571–7579 & 2004 Nature Publishing Group All rights reserved 0950-9232/04 $30.00 www.nature.com/onc

Interleukin 24 is induced by the RET/PTC3 oncoprotein and is an autocrine for epithelial cells

Shogo Shinohara1 and Jay L Rothstein*,1

1Departments of Microbiology/Immunology and Otolaryngology-Head and Neck Surgery, Kimmel Institute, Thomas Jefferson University, Philadelphia, PA 19107, USA

Thyroid , like hematological , are Keywords: inflammation; 24; Mda7; RET/ commonly associated with chromosomal translocations PTC; thyroid leading to the formation of fusion . Through altered signaling by fusion proteins, cell death and survival pathways are disrupted and the physiological balance of cell–cell communication may be lost. A consequence of this disruption is the release of factors by stressed cells that Introduction alert the host. One type of host response is leukocytic infiltration that may develop into chronic inflammation or The development of thyroid cancer, like other malig- . Although inflammation can be nancies, can be associated with inflammation or auto- associated with neoplastic tissue, the mechanism driving immune disease. Consistent with this notion, 10–40% of this process is largely unknown. Therefore, to address the patients diagnosed with Hashimoto’s thyroiditis (HT) mechanism of cancer inflammation we investigated the may show evidence of concurrent thyroid carcinoma effects of an oncogene in a murine model system. A (Dailey et al., 1955; Ott et al., 1987). Although the role comprehensive genetic analysis revealed several soluble of these immune infiltrates is uncertain, clinical studies factors that were induced by RET/papillary thyroid have shown that patients with papillary thyroid carcinoma (PTC)3 expression including several carcinoma (PTC) that present with thyroiditis (PTC/ proinflammatory , and immunologi- HT) have a better clinical outcome than patients cally relevant costimulatory molecules. Following a large diagnosed with PTC alone (Matsubayashi et al., 1995; genetic screen using RP3-expressing thyroid cells, we Loh et al., 1999; Tamimi, 2002). Understanding the identified a highly abundant transcript and later identified molecular mechanisms that distinguish PTC from PTC/ it as interleukin 24 (Il24), a with diverse tumor HT will be critical if we are to take advantage of the suppressor and inflammatory activities. We show that anticancer responses that likely explain the difference RET/PTC3 induces Il24 expression in rat thyrocytes and between these two pathologies. Along these lines, some that this expression is dependent on the signaling properties information about the early stages of cancer may of its tyrosine kinase. Likewise, RET/PTC3 induces large provide important clues about this difference. RET/ amounts of Il24 following expression in murine thyrocytes, PTC oncogenes are a family of fusion proteins derived but its expression is dramatically reduced in poorly from chromosomal rearrangements involving the tyr- differentiated carcinomas, a finding that parallels the loss osine kinase domain of the c-RET proto-oncogene and of RET/PTC3 expression. Consistent with its behavior as are frequent mutations found early in the development a tumor suppressor, the loss of Il24 coincided with the loss of a variety of differentiated thyroid carcinomas (Jhiang of RET/PTC3 in poorly differentiated mouse tumors. A et al., 1992; Santoro et al., 1992; Fusco et al., 1995; functional role of Il24 in the autocrine growth/survival of Takahashi et al., 1995; Fugazzola et al., 1996; Sugg et al., RET/PTC3-expressing thyroid cells was identified helping 1996; Bongarzone et al., 1997). Whereas RET/PTC to support its role in cellular transformation. These data expression has been frequently observed in patients suggest that the induction of Il24 by oncogenes may affected with PTCs showing concurrent thyroiditis, its support tumor growth at the early stages of cancer. expression was also reported in thyroid tissue of patients Oncogene (2004) 23, 7571–7579. doi:10.1038/sj.onc.1207964 afflicted with HT with no detectable cancer suggesting a Published online 23 August 2004 common etiology of these two diseases (Wirtschafter et al., 1997; Sheils et al., 2000; Mechler et al., 2001; Pasquale et al., 2001). The association of RET/PTC with *Correspondence: JL Rothstein, Departments of Microbiology/Im- HT suggested that oncoproteins may be one component munology and Otolaryngology-Head and Neck Surgery, Kimmel involved in promoting organ-specific inflammation. Cancer Institute, Thomas Jefferson University, 233 South 10th Street, While the mechanism of how RET/PTC contributes to BLSB 909, Philadelphia, PA 19107, USA; E-mail: [email protected] autoimmune thyroiditis is not clear, the activated Received 8April 2004; revised 9 June 2004; accepted 9 June 2004; tyrosine kinase activity associated with RET/PTC has published online 23 August 2004 been shown to support the production of inflammatory RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7572 mediators and induce inflammation (Russell et al., Results 2003). One member of this oncoprotein family RP3 (RP3) can induce nuclear NF-kB activity and the RP3 oncogene expression induces Il24 in thyroid cells downstream effects of this activation including the production of the Gmcsf and Mcp1 from thyroid cells In empirical studies evaluating the role of the thyroid in vitro and in vivo (Russell et al., 2003, 2004). oncoprotein RP3 to alter cell survival pathways, we Furthermore, cytokine synthesis was largely dependent found several inflammatory factors produced from on a tyrosine at position 588 of RET/PTC, which oncogene-expressing transfectants that were absent from et al corresponds to tyrosine 1062 of c-RET, a Shc docking mutant or control cells (Russell ., 2003). To extend site critical for downstream signaling (Russell et al., these studies and to objectively identify additional 2003). factors that are produced following oncoprotein expres- To identify other inflammatory proteins requiring sion, we compared parental PC CL3 thyroid cells and Shc-dependent signaling, we screened mRNAs from those expressing RP3 or the Shc signaling mutant Y588F using a rat gene array containing 10 000 RET/PTC-expressing thyroid cells that were absent RP3 cDNAs. The function of RET/PTC1 to signal mitogen- from cells expressing RP3 Tyr588. This analysis revealed esis is dependent on the phosphorylation of tyrosine 451 additional factors induced by RP3 Tyr588 including and the subsequent binding of the adaptor Shc cytokines, chemokines and costimulatory molecules of et al et al immunological significance. Among these , we (Chiariello ., 1998; Salvatore ., 2000). In our previous report, tyrosine 588 in RP3 corresponding to were especially interested in interleukin 24 (Il24), a Tyr451 in RET/PTC1 was required for the production of member of the IL-10 gene family produced in abun- certain inflammatory cytokines including Gmcsf, Mcp1 dance in RP3-expressing cells. Previous studies have and Kc/Groa (Russell et al., 2003). Comparing the list shown that Il24 behaves like a tumor suppressor gene, of genes activated by RP3, but not by RP3Y588F or vector although little was known about its role in the host controls, revealed the identity of several additional response to cancer. candidate genes whose function includes the induction Il24 was originally identified as an abundant tumor- of inflammation. For these studies, we focused on the specific transcript expressed by human melanoma cells expression of RP3-induced genes that showed a 10-fold treated with interferonb (IFNb) and the protein kinase C activator mezerein (Jiang et al., 1995). However, or greater induction compared with controls (Table 1). One mediator that showed a dramatic induction unlike its expression in normal melanocytes and benign following RP3 expression was the inflammatory cyto- lesions (nevi), advanced and metastatic melanomas lack detectable Il24 expression (Ekmekcioglu et al., 2001). kine Il24. Il24 was induced 80-fold in RP3-expressing cells compared to vector controls. Il24, like many Interestingly, forced expression of Il24 using a recombi- nant adenovirus causes growth inhibition and in a diversity of human cancers, but normal cells are unaffected (Su et al., 1998; Saeki et al., 2000; Ekmek- Table 1 cioglu et al., 2001; Lebedeva et al., 2003; Yacoub et al., Genea RP3/vectorb 2003b). This interesting distinction prompted the devel- opment of several clinical trials for the treatment of MCP1 189.31 cancer. However, while clinical trials involving the use Thymosin b4 162.65 of Il24 as an cancer therapeutic are ongoing, little is Interleukin 24 80.94 known about its biological function in spontaneous disease (Sarkar et al., 2002; Fisher et al., 2003). With Kc/Groa 66.06 regard to its function in the immune system, Il24 was GMCSF 45.85 shown to act as an antagonist of IL10 and induce the Amino-acid transporter system A 27.14 PKCd binding protein 22.16 production of various proinflammatory mediators such Carbonic anhydrase 2 21.44 as TNFa, IFNg, IL6, IL1b, IL12 and GMCSF through Sialoprotein () 20.33 the binding of the heterodimeric receptors, IL20R1/ Alfatoxin B1 aldehyde reductase 18.36 Il20R2 or IL22R1/ Il20R2 (Caudell et al., 2002; Wang Urinary plasminogen activator receoptor 2 17.62 Fibronectin 1 17.56 et al., 2002). Although a direct role of Il24 signaling in Collagen alpha1 type1 15.42 carcinogenesis has not been shown, indirect effects have Matrix metallaproteinase 10 14.79 been implied following the activation of oncogenic RAS Mesothelin 14.64 that can induce Il24 in animal models (Zhang et al., S100 calcium-binding protein A4 14.54 2000). IGE binding protein 13.82 Glutathione-S-transferase, alpha type 11.92 To gain a better understanding of the role of soluble Nuclear receptor subfamily 0, group B 11.56 mediators at the early stages of cellular transformation, we investigated the role of the RP3 oncogene to induce aGene identified by differential binding for a fluorescent cDNA probe I124 in thyroid cells in culture and from thyroids of RP3 onto a glass microarry chip containing 10 000 gene-specific rat transgenic mice. Moreover, using these in vitro and in oligonucleotides. Sequence information was obtained from the NCBI database and genes in gray were verified by RT–PCR in other studies vivo models, we assessed the function of Il24 on (Russell et al., 2003). bRatio of hybridization signal for the listed gene transformed thyroid cells and find a role for this tumor in cells expressing RP3 versus cells transfected with vector alone. cell mediator in the growth of transformed cells. Number reflects the fold increase in

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7573

Figure 2 Thyroid-specific RP3 expression in transgenic mice induces Il24 mRNA. (a) Total RNA from thyroids of wild-type or RP3-expressing transgenic mice was reverse transcribed to produce cDNA, amplified by PCR using primers for the listed genes and PCR products resolved on a 2% agarose gel. Odd numbered lanes from 1 to 11 represent PCR products derived from individual pools of cDNA synthesized from wild-type mice, whereas even numbered lanes from 2 to 12 represent those from RP3-expressing mice. Lane 13 represents the negative control for these experiments (unrelated template). (b) Summary of pooled data shows induced expression of Il24 in RP3-expressing, but not control, mouse thyroid tissue. Il24 is expressed significantly higher in the thyroids of RP3-expressing transgenic mice than controls (*Po0.01). Data are derived from 14 mice/group

confirm the Il24 gene array data, we performed RT– PCR. Although the expression of Il24 was significantly higher in RP3-expressing cells than that in vector Figure 1 RP3 induces Il24 in thyrocytes. (a) A representative RT– PCR experiment is shown. cDNA was amplified by PCR for the controls, detectable levels of Il24 were also found in listed genes using total RNA extracted from PC Cl3 thyroid cells RP3Y588F-expressing cells, consistent with the gene array transfected with vector alone, RP3- or RP3Y588F-encoding vectors. data (Figure 1a and b). This induction was observed No template negative control (NC) is also shown. (b) The EtBr- despite the expression of equal amounts of RP3 or stained Il24 PCR products were quantified by densitometric analysis and the data from a total of six pools of cDNA are mutant RP3 mRNA (Figure 1a) or protein (not shown). shown. Significant increases in Il24 expression was observed in The production of Il24 protein synthesized by thyroid RP3-expressing cells, but the expression was only partially induced cells was verified by Western blot analysis (Figure 1c). in RP3Y588F-expressing cells (*Po0.01). RP3 and RP3Y588F mRNAs The observed molecular weight of rat Il24 is reported were expressed at similar levels since no statistical difference was between 25 and 35 kDa, a value higher than the deduced found between them (P40.5). These data are obtained from six independent pools cDNA. (c) Vector control, RP3- or RP3Y588F- molecular weight of 23 kDa. This difference is likely due expressing PC Cl3 thyroid cells were incubated in medium to post-translational modifications, such as glycosyla- containing Brefeldin A for blocking secretion of cytokines before tion of the endogenously produced molecule when harvest. In all, 90 mg of total protein (cellular lysates) were resolved compared with recombinant Il24 controls (Wang et al., using 16% (b-actin) or 12% (Il24) SDS–PAGE gels, electroblotted onto nitrocellulose and hybridized with a polyclonal antiserum 2002). against b-actin and a polyclonal antiserum against rat Il24, respectively. Since rodent anti-Il24 antibody crossreacts with RP3 oncogene-induced expression of Il24 in vivo human Il24 protein, recombinant human Il24 was included as a positive control Transgenic mice expressing human RP3 exclusively in the thyroid develop thyroid hyperplasia by 4 weeks of mediators induced by RP3, was severely reduced or age and develop solid subtype papillary carcinomas as absent when analysed using the RP3Y588F mutant, they age beyond 8months of age (Powell et al., 1998). suggesting that this docking site was important for Since analysis of Il24 expression was first identified from cytokine signaling (Russell et al., 2003). An exception to cell lines in vitro, we were interested in determining if this was Mcp1 and Kc/Groa that showed higher levels RP3-induced genes were also found in the thyroids of of gene transcription in RP3Y588F-mutated cells when RP3 transgenic mice. To investigate this, RT–PCR compared with vector control containing thyroid cells analysis was performed on freshly isolated thyroid tissue (23- and 10-fold, respectively, not shown), although still from RP3 transgenic mice. These data indicated that not as high as that observed in RP3-expressing cells. To Il24 mRNA is expressed in the thyroid of RP3

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7574

Figure 3 Il24 protein expression in the thyroid of RP3-expressing transgenic mice. (a) A representative thyroid from a 6-month-old wild-type mouse showing normal architecture and absence of staining following treatment with Il24-specific antibody. (b)A representative thyroid tissue from a 6-month-old RP3 transgenic mouse demonstrating differentiated papillary features coordinate with strong Il24-specific staining. (c) No staining was observed when an isotype control antibody was used on RP3 transgenic thyroids. (d) Preabsorbing Il24 antibody with whole recombinant Il24 protein reduced the Il24-specific signal in RP3-expressing tissue. (e) Antibody specific for the RET kinase failed to bind thyroid tissue of wild-type mice but bound strongly to the thyroids of RP3 transgenics as seen in panel f. Luminal (arrows) were frequently found in RP3-expressing thyroids (c and d), a finding consistent with previous reports (Russell et al., 2004). These images are representative of seven different thyroids specimens (magnification  40)

transgenic mice but not in thyroid tissue from wild-type binant Il24 protein prior to staining. This preabsorption mice (Figure 2a and b). Il24 mRNA expression was step confirmed the specificity of antibody binding corroborated by analysis of protein expression using (Figure 3d) and indicated that Il24 reactivity was immunohistochemical methods. Indeed, serial sections responsible for the staining observed in Figure 3b. of embedded mouse RP3-expressing thyroid organs Preabsorption with an unrelated recombinant protein stained brightly using a rodent-specific anti-Il24 anti- failed to neutralize antibody reactivity (data not shown). body (Figure 3). Specifically, the thyroids of the RP3 The presence of inflammatory mediators within the transgenic mice, which developed hyperplasia of the follicle is consistent with the inflammatory nature of follicles including expanded colloid, produced Il24 RP3-expressing tissue (Russell et al., 2003, 2004). protein within the cytoplasm of epithelial cells and also Indeed, luminal monocytes were present in transgenic secreted this factor into the colloid, suggesting that large specimens (Figure 3c and d). Little to no background amounts of Il24 is locally released (Figure 3b). No staining was observed in the colloid of anti-RET-stained staining was detected in the colloid of RP3 transgenic wild-type thyroids (Figure 3e) compared to the strong thyroids treated with isotype control antibody staining observed in positive control RP3 transgenic (Figure 3c). To verify the specificity of these reactions, thyroids stained with anti-RET-specific antibody we preabsorbed the Il24 antibody with purified recom- (Figure 3f).

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7575

Figure 5 Reduced proliferation of RP3-expressing thyroid cells by neutralizing anti-rat Il24. RP3-expressing cells were incubated with a 1 : 50 or 1 : 200 dilution of rat Il24-specific antisera or a 1 : 50 dilution of antisera directed against an unrelated recombinant protein. Shown is the ratio of the cell count before and after Figure 4 Il20R2 is expressed in the thyroids of wild-type and RP3- incubation with antisera at the indicated time points. Total cell expressing transgenic mice. (a) A representative RT–PCR experi- numbers for anti-Il24 antibody-treated cultures were significantly ment is shown. For all experiments, cDNA was synthesized from lower at 24 and 48h time points (* P 0.01). Error bars indicate the RNA of wild-type or RP3-expressing thyroids and amplified by o s.e.m. from eight independent experiments PCR using gene-specific primers for the indicated genes. Lanes 1, 3 and 5 represent PCR products derived from individual wild-type mice, whereas lanes 2, 4 and 6 represent products derived from RP3-expressing mice. The product shown in lane 7 is from an period, we find that proliferation of anti-Il24-treated unrelated mouse (negative control). (b) The relative mRNA thyroid cells in culture is significantly reduced at all time expression of mouse Il20R2 is shown. There is no significant difference in the expression of RNA between wild-type and RP3- points compared with the growth of cells in control expressing thyroids (*P40.5). Data are derived from nine thyroid cultures (Figure 5). specimens Loss of Il24 expression from thyroid carcinomas of Il24 supports the proliferation of RP3-expressing RP3p53À/Àmice correlates with the loss of RP3 expression thyrocytes To better understand the inflammatory changes that Although several studies have demonstrated the proa- occur in gene expression during later stages of tumor poptotic effects of forced IL24 expression in malignant progression, we studied RP3p53À/À mice that develop cells, less is known about its physiological function advanced thyroid malignancies containing regions of during neoplastic development. binding studies both differentiated and undifferentiated carcinoma have shown that IL24 can initiate signaling through (Powell et al., 2001). These tumors were characterized specific heterodimeric receptors belonging to the as poorly differentiated carcinoma due to the presence IL20R1/IL20R2 and IL22R1/IL20R2 gene families of numerous mitotic figures, pleiotropic nuclear mor- (Wang et al., 2002). Expression of these receptors on phology and irregular cell shapes as described (Powell thyroid cells would be a requisite if RP3-induced Il24 et al., 2001). Since the loss of RP3 expression correlates contributes to thyroid pathology. To investigate this with the loss of selected cytokine gene expression, we possibility, we isolated RNA from mouse thyroids and next examined if Il24 expression was also dependent on examined Il24 receptor expression by RT–PCR. Results the presence of functional RP3 within tumors containing from these experiments showed that the common Il24 a mixed histological morphology. In differentiated receptor chain Il20R2 was expressed equally in both regions of tumors, which expressed high amounts of wild-type and RP3 transgenic mice (Figure 4). Even RP3 protein (Figure 6a), specific staining was also though the Il24 receptor was found on wild-type and detected for Il24 (Figure 6b). Interestingly, in undiffer- RP3-expressing thyroid tissue, only thyroids expressing entiated regions of tumors expressing reduced or no RP3 would likely support autocrine growth effects. RP3 protein (Figure 6c), the presence of Il24 signal was Accordingly, we assessed the potential for Il24 to likewise reduced (Figure 6d). influence organ function through alterations in cell survival or by promoting cell death in RP3-expressing cells. To examine this, murine anti-Il24 antibody was Discussion incubated with RP3-expressing thyroid cells to neutra- lize the secreted Il24 in the culture medium. The Thyroid epithelial cells are capable of producing a wide proliferation of antibody-treated cells was compared variety of inflammatory cytokines in larger quantities with untreated RP3-expressing cells or antibody control- than any other endocrine cell type (Aust and Scher- treated cells. Following a short-term (2 day) incubation baum, 1996). Some of these, such as IL1, IL6, IL8and

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7576

Figure 6 Il24 protein expression in the thyroids tumors of RP3p53À/À mice. (a) A representative thyroid specimen removed from a 2- month-old RP3p53À/À mouse, which showed mixed papillary and poorly differentiated features, strongly stained (darkened cells) for the RET kinase. (b) A serial section of the same thyroid specimen shown in panel a stained with anti Il24 antibody. Dark staining is indicative of high amounts of Il24 protein expression within thyroid epithelial cells. (c) Specimens stained with anti-RET antibody or (d) anti Il24-specific antibody revealed that some regions of the tumor, which did not express the RET kinase, also lacked Il24 expression. These findings are representative of 10 different thyroid tumors (magnification  40)

TNFa can be induced even from normal thyroid cells, the receptor for Il24 was also found on normal indicating that cytokines may not only contribute to thyrocytes, one consequence of Il24 production by various thyroid cancers but may also serve to support transformed thyroid cells may be to influence neighbor- the homeostatic control of cell survival at precancerous ing nontransformed cells and, in this manner, contribute stages (Rasmussen, 2000). Tumor tissue or tumor- to a generalized growth-supporting effect in the thyroid derived cell lines from differentiated carcinomas release likely causing aberrant epithelial . Consistent GCSF, GMCSF, IL1a, IL4, IL6, IL8, IL10, MCP1 and/ with these observations, rodent Il24 is implicated in the or TNFa (Tohyama et al., 1992; Kayser et al., 1995; proliferation induced by oncogenic forms of RAS Nakada et al., 1996; Basolo et al., 1998; Kurebayashi (Zhang et al., 2000) and is a critical growth factor for et al., 2000; Scarpino et al., 2000) and undifferentiated fibroblast-like cells at the wound edge during normal thyroid carcinomas produce GROa, GCSF, GMCSF, tissue repair (Soo et al., 1999). In addition, over and IL1a, IL4, IL6, IL8, IL10, and/or TNFa (Enomoto expression of Il24 in human colon carcinomas or in et al., 1990; Yoshida et al., 1992; Aust and Scherbaum, melanocytic nevi (Zhang et al., 2000; Ekmekcioglu et al., 1996; Basolo et al., 1998; Stassi et al., 2003), although, in 2001) corresponds with the expression of oncogenic Ki- either case, very little is known about when in tumor RAS or BRAF in these lesions (Zhang et al., 2000; development these factors are made or what their Pollock et al., 2003); signaling pathways also implicated purpose may be. Whereas no single biochemical path- in the development of thyroid cancer (Kimura et al., way has been identified that explains factor production 2003). by thyroid cells, we previously demonstrated that an In contrast to the findings of this study, direct gene alternative form of activated RET kinase (RP3) can delivery of human Il24 using recombinant adenoviruses induce secretion of large amounts of Mcp1 and Gmcsf, a caused cell death in malignant cells even though property dependent on the function of the tyrosine evidence of ligand–receptor binding was not demon- kinase domain (Russell et al., 2003). In thyroid cells, strated. In addition, the growth inhibition and apoptosis some cytokines can act in an autocrine and/or paracrine observed in a wide variety of human cancer cells was not manner in promoting the transformation of epithelial observed in normal human cells (Su et al., 1998; Saeki cells during tumorigenesis. For example, exposure to et al., 2000; Ekmekcioglu et al., 2001; Lebedeva et al., IL4 and IL10 reduces apoptosis and sensitivity of 2003; Yacoub et al., 2003a). In fact, when injected into thyroid carcinomas to chemotherapeutic agents by animals, Il24 has been shown to reduce , an inducing Bcl-2 and Bcl-XL (Stassi et al., 2003). Using effect that appears to indirectly limit tumor progression a similar approach, we showed that Il24 is induced by (Saeki et al., 2000; Ramesh et al., 2003). Thus, the RP3 to support cell proliferation of RP3-expressing apparent disparity of functions of Il24 may be explained thyrocytes in an autocrine manner. Interestingly, since either, in part, by species differences (Caudell et al.,

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7577 2002) or by differences in the malignant nature of the RT–PCR analysis target cell. Consistent with this, as cancer becomes more Thyroid tissue, resected from transgenic or wild-type mice, was invasive (aggressive), genes expressed at early stages homogenized into 500 ml of TRIZOL (Invitrogen, Carlsbad, may be lost, suggesting a selection towards changing CA, USA). PC CL3 cells were homogenized in 1 ml of signaling pathways during cancer progression (Powell TRIZOL (Invitrogen, Carlsbad, CA, USA) according to et al., 2001). These data suggest that the growth- protocols supplied by the manufacturer. RNA was extracted promoting effects of Il24 are stage specific and that the and purified and treated with RNAse-free DNAse (Ambion, high levels of Il24 expression that may support the Austin, TX, USA). DNA-free total RNA (5 mg) was converted growth of differentiated thyroid malignancies may not into cDNA using Superscript II reverse transcriptase (Invitro- be required for the continued proliferation and out- gen, Carlsbad, CA, USA). The amount of template cDNA was estimated from the approximate 33% conversion rate of RNA growth of poorly differentiated (non-RP3-expressing) to cDNA and amplified using primers specific to glyceralde- carcinomas. hyde-3-phosphase dehydrogenase (G3pdh). The cycling para- In summary, these studies provide a connection meters were chosen to provide linear estimations of products between the expression of cellular oncogenes and the for comparison as described previously (Russell et al., 2003, production of inflammatory factors that influence the 2004). Each test cDNA was normalized to PCR products growth of newly transformed epithelial cells. Although obtained with G3pdh primers and a ratio of products was the effects of Il24 have yet to be fully explored, these reported. PCR cycling conditions consisted of one cycle of studies support the notion that oncogene-specific denaturation at 941C followed by 20–45 cycles of denaturation induction of Il24 is an important mediator of tumor (941C for 30 s), primer annealing (601C for 30 s), extension 1 1 cell survival and growth. The unusual abundance and (72 C for 60 s) and a final single extension cycle of 72 C for 7 min. PCR products were resolved in 2% agarose gels frequency of Il24 in various differentiated thyroid containing 0.5 mg/ml of EtBr and band intensities quantified carcinomas is offset by its apparent loss during tumor using the BioRad Gel Doc system with an integrated progression. Understanding this mechanism will be densitometric analysis program. Optimal cycling conditions important for elucidating methods or therapeutic for each primer set to provide products in the linear portion of strategies to limit the spread of these, and other, the amplification curve were determined by plotting cycle malignancies. numbers varying from 20 to 45 cycles versus band intensities as described previously (Powell et al., 2003). The optimal cycling conditions for G3pdh primer set was 25 cycles, while all other Materials and methods primer sets were determined to be 30 cycles. The following are specific sequences of primers for murine G3pdh, 50 primer Mice GGGCCCCCTTCATTGACCTCAACTAC and 30 primer GGTACCATGACAAGCTTCCCATTCTC; rat Il24, 50 pri- RP3 transgenic mice were generated as previously described mer AGCTTGAGACAACAGATTCTCCCC and 30 primer (Powell et al., 1998). RP3 mice were backcrossed to C3H/HeJ ACTGTCACTAATGGGAAGCATGGC; mouse Il24, 50 pri- mice for greater than five generations. Homozygous RP3 mice mer GGACTACAGATTCTCCCCTGCCTG and 30 primer on C3H/HeJ background mice were crossed to C57BL/6- AGCATGGAATTGTCCTTACTGGGC; the human RP3 Tpr53 mice (R Fishel, Kimmel Cancer Institute) and double breakpoint region, 50 primer CCAGAGCAGAAGTCAGCA p53À/À heterozygous progeny intercrossed to generate RP3 TCC and 30 primer CTCTTTCAGCATCTTCACGGC; and homozygotes. mouse Il20R2, 50 primer CTGCCCCTCAGAACCTCTCTG and 30 primer CCACAAGGAACTCAAACTGGG. Cell culture Western blotting PC CL3 rat thyrocytes were kindly provided by Dr Jeffery Knauf (University of Cincinnati, OH, USA). Vectors were For detection of Il24, a soluble protein, we needed to block its individually transfected into PC CL3 cells. Briefly, the human secretion from the within target cells. RP3 oncogene was ligated into the retroviral vector, pMV7- To accomplish this, each transfected PC CL3 cell lines were Tyr588 (Tyr1062 of c-RET) in and the kinase domain was incubated in medium containing 10 mg/ml of Brefeldin A for mutated to a phenylalanine by using QuickChange Mutagen- 5 h before harvest as described (Winterrowd and Chin, 1999). esis Kit (Stratagene, La Jolla, CA, USA). The constructs, After treatment, cells were homogenized in protein lysis buffer pMV7/RP3, pMV7/RP3Y588F and the pMV7 vector alone were (20 mM Tris-HCl (pH 7.5), 1 mM EDTA, 1 mM EGTA 1% transfected into PC CL3 cells using the Phoenix transfection Triton X-100, 150 mM NaCl with freshly added 1 mg/ml method supplied by the manufacturer. Following DNA aprotinin, 1 mg/ml leupeptin, 1 mM phenylmethylsulfonyl transfer, cell transfectants were selected in thyroid cell growth fluoride and 1 mM orthovanadate) and stored at À701C. media (Coon’s modified F12 medium containing 5% FBS, Proteins were resolved using SDS–PAGE and gels transferred 100 U/ml penicillin/streptomycin, 2 mML-glutamine and the to a nitrocellulose membrane. Membranes were blocked for following growth factors: 10 ng/ml somatostatin, 10 ng/ml 60 min in TBST buffer (20 mM Tris-HCl (pH 7.6), 135 mM glycine-histidine-lysine, 5 mg/ml transferring, 10 nM hydrocor- NaCl, 1% Tween-20) with 5% nonfat dry milk and tisone, 10 mg/ml insulin and 10 mIU/ml bovine thyroid- immunoblotted with a 1 : 1000 dilution of rabbit anti-rat Il24 stimulating hormone) supplemented with G418(400 mg/ml). antisera (GenHunter Corp., Nashville, TN, USA) or with a PC CL3 cells transfected with pMV7/RP3 were grown in 1 : 500 dilution of goat anti-human b actin (I-19) polyclonal separate medium without growth factors, since culture in this antibody for 16 h at 41C. Anti-rat Il24 antisera was precleared medium causes apoptosis and selects against newly expressing using 4 mg/ml of protein from wild-type PC CL3 cells to reduce RP3 þ cells. Phoenix-Eco cells used for transfection were nonspecific binding. Following incubation, membranes were grown in Dulbecco’s modified Eagle’s media supplemented washed three times for 5 min in TBST buffer and treated with a with 10% FBS and 100 U/ml penicillin/streptomycin. 1 : 1000 dilution of horseradish peroxidase-conjugated donkey

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7578 anti-rabbit IgG (Amersham) or a 1 : 5000 dilution of horse- with TBS/1% BSA þ 0.01% Tween-20 and incubated with radish peroxidase-conjugated rabbit anti-goat IgG (Pierce biotinylated secondary antibody for 1 h at room temperature. Biotechnology, Rockford, IL, USA), respectively, for 2 h at Following this incubation, specimens were treated with RT. Membranes were washed again three times in TBST substrate according to the DAB Vectastatin kit (Vector Labs, buffer and protein bands visualized using ECL (Amersham) Burlingame, CA, USA), counterstained with hematoxylin, after exposure to X-ray film for 10 min. Commercially dehydrated and mounted. available human Il24-His protein was used as a control for protein size and mobility (GenHunter Corp., Nashville, TN, USA). Cell proliferation analysis PC CL3 cells transfected with the pMV7/RP3 vector or control Immunohistochemistry vectors were plated into each well of a 12-well tissue culture dish (5 Â 104 cells/well) using 500 ml of culture medium Paraffin-embedded mouse thyroid sections were dewaxed with containing a 1 : 50 or 1 : 200 dilution of anti-rat Il24 rabbit xylene and rehydrated in ethanol as described previously antisera (GenHunter Corp., Nashville, TN, USA) or 1 : 50 (Russell et al., 2000, 2003, 2004). Slides were heated in 10 mM dilution of rabbit antisera directed against an unrelated citrate buffer (pH 6.0) in a microwave for 15 min. Slides, recombinant mouse protein. Then, at designated time points cooled to RT, were washed with PBS and incubated in 3% (0, 24 and 48h) cells in each group were harvested by H2O2/PBS for 10 min. Slides were then blocked with PBS/1% trypsinization and viable cells counted using trypan blue bovine serum albumin (BSA) containing 0.01% Tween-20 for (GIBCO BRL, Grand Island, NY, USA). 5 min, followed by10% normal serum for 15 min (goat serum was used for RET staining, whereas rabbit serum was used for Il24 staining). After blocking, specimens were incubated with a Statistical analysis 1 : 200 dilution of a rabbit anti-human RET antibody (Santa Results are presented as the mean7s.e.m. The data were Cruz Biologicals, Santa Cruz, CA, USA), specific for an analysed using the Student’s t-test with the Analyze-it add-in epitope corresponding to the carboxy-terminal amino acids software program for Microsoft Excel. For these experiments (784–801) of the RET/PTC protein or a 1 : 500 dilution of goat P-values o0.05 were considered statistically significant. anti Il24 antibody (C-16, Santa Cruz Biologicals, Santa Cruz, CA, USA) overnight at RT. Control rabbit IgG (Vector Labs, Burlingame, CA, USA) and goat polyclonal IgG directed Acknowledgements against an unrelated protein (human Bruton’s tyrosine kinase) This work was supported by grants from the National at the same dilution were also utilized for control staining. The Institutes of Health, CA-76259 (JLR) and T32-CA09678 next day, samples were washed with TBS three times, blocked (JPR).

References

Aust G and Scherbaum WA. (1996). Exp. Clin. Endocrinol. Jhiang SM, Caruso DR, Gilmore E, Ishizaka Y, Tahira T, Diabetes, 104, 64–67. Nagao M, Chiu IM and Mazzaferri EL. (1992). Oncogene, 7, Basolo F, Fiore L, Pollina L, Fontanini G, Conaldi PG and 1331–1337. Toniolo A. (1998). Clin. Cancer Res., 4, 381–387. Jiang H, Lin JJ, Su ZZ, Goldstein NI and Fisher PB. (1995). Bongarzone I, Butti MG, Fugazzola L, Pacini F, Pinchera A, Oncogene, 11, 2477–2486. Vorontsova TV, Demidchik EP and Pierotti MA. (1997). Kayser L, Broholm H, Francis D, Perrild H, Olsen BE, Genomics, 42, 252–259. Bendtzen K and Hoyer PE. (1995). Autoimmunity, 20, Caudell EG, Mumm JB, Poindexter N, Ekmekcioglu S, 75–82. Mhashilkar AM, Yang XH, Retter MW, Hill P, Chada S Kimura ET, Nikiforova MN, Zhu Z, Knauf JA, Nikiforov YE and Grimm EA. (2002). J. Immunol., 168, 6041–6046. and Fagin JA. (2003). Cancer Res., 63, 1454–1457. Chiariello M, Visconti R, Carlomagno F, Melillo RM, Kurebayashi J, Tanaka K, Otsuki T, Moriya T, Kunisue H, Bucci C, de Franciscis V, Fox GM, Jing S, Coso OA, Uno M and Sonoo H. (2000). J. Clin. Endocrinol. Metab., Gutkind JS, Fusco A and Santoro M. (1998). Oncogene, 16, 85, 2889–2896. 2435–2445. Lebedeva IV, Su ZZ, Sarkar D and Fisher PB. (2003). Semin. Dailey ME, Lindsay S and Skahen R. (1955). AMA Arch. Cancer Biol., 13, 169–178. Surg., 70, 271–297. Loh KC, Greenspan FS, Dong F, Miller TR and Yeo PP. Ekmekcioglu S, Ellerhorst J, Mhashilkar AM, Sahin AA, (1999). J. Clin. Endocrinol. Metab., 84, 458–463. Read CM, Prieto VG, Chada S and Grimm EA. (2001). Int. Matsubayashi S, Kawai K, Matsumoto Y, Mukuta T, Morita J. Cancer, 94, 54–59. T, Hirai K, Matsuzuka F, Kakudoh K, Kuma K and Tamai Enomoto T, Sugawa H, Inoue D, Miyamoto M, Kosugi S, H. (1995). J. Clin. Endocrinol. Metab., 80, 3421–3424. Takahashi T, Kitamura N, Yamamoto I, Konishi J, Mori T Mechler C, Bounacer A, Suarez H, Saint Frison M, Magois C, and Imura H. (1990). Cancer, 65, 1971–1979. Aillet G and Gaulier A. (2001). Br. J. Cancer, 85, 1831–1837. Fisher PB, Gopalkrishnan RV, Chada S, Ramesh R, Grimm Nakada T, Sato H, Inoue F, Mizorogi F, Nagayama K and EA, Rosengeld MR, Curiel DT and Dent P. (2003). Cancer Tanaka T. (1996). Intern. Med., 35, 815–820. Biol. Ther., 2, S23–37. Ott RA, McCall AR, McHenry C, Jarosz H, Armin A, Fugazzola L, Pierotti MA, Vigano E, Pacini F, Vorontsova Lawrence AM and Paloyan E. (1987). Am. Surg., 53, 442– TV and Bongarzone I. (1996). Oncogene, 13, 1093–1097. 445. Fusco A, Santoro M, Grieko M, Carlomagno F, Pasquale D, Rothstein JL and P PJ. (2001). Hum. Pathol., 32, Dathan N, Fabien N, Berlingieri MT, Li Z, De Franciscis 24–30. V, Salvatore D, Melillo RM, Portella G, Cerrato A, Pollock PM, Harper UL, Hansen KS, Yudt LM, Stark M, Celantuoni V and Vecchio G. (1995). J. Endocrinol. Invest., Robbins CM, Moses TY, Hostetter G, Wagner U, Kakareka 18, 127–129. J, Salem G, Pohida T, Heenan P, Duray P, Kallioniemi O,

Oncogene RET/PTC3 oncoprotein induces Il24 S Shinohara and JL Rothstein 7579 Hayward NK, Trent JM and Meltzer PS. (2003). Nat. Soo C, Shaw WW, Freymiller E, Longaker MT, Bertolami Genet., 33, 19–20. CN, Chiu R, Tieu A and Ting K. (1999). J. Cell. Biochem., Powell Jr DJ, Eisenlohr LC and Rothstein JL. (2003). 74, 1–10. J. Immunol., 170, 861–869. Stassi G, Todaro M, Zerilli M, Ricci-Vitiani L, Di Liberto D, Powell D, Russell J, Li G, Huebner K and Rothstein JL. Patti M, Florena A, Di Gaudio F, Di Gesu G and De Maria (2001). Oncogene, 20, 3235–3246. R. (2003). Cancer Res., 63, 6784–6790. Powell D, Russell J, Nibu K, Li G, Rhee E, Liao M, Santoro Su ZZ, Madireddi MT, Lin JJ, Young CS, Kitada S, Reed JC, M, Fusco A and Rothstein JL. (1998). Cancer Res., 58, Goldstein NI and Fisher PB. (1998). Proc. Natl. Acad. Sci. 5523–5528. USA, 95, 14400–14405. Ramesh R, Mhashilkar AM, Tanaka F, Saito Y, Branch CD, Sugg SL, Zheng L, Rosen IB, Freeman JL, Ezzat S and Asa Sieger K, Mumm JB, Stewart AL, Boquio A, Dumoutier L, SL. (1996). J. Clin. Endocrinol. Metab., 81, 3360–3365. Grimm EA, Renauld JC, Kotenko S and Chada S. (2003). Takahashi K, Kashima K, Daa T, Yokoyama S, Nakayama I Cancer Res., 63, 5105–5113. and Noguchi S. (1995). Pathol. Int., 45, 366–374. Rasmussen AK.. (2000). Dan. Med. Bull., 47, 94–114. Tamimi DM. (2002). Int. J. Surg. Pathol., 10, 141–146. Russell JP, Engiles JB and Rothstein JL. (2004). J. Immunol., Tohyama K, Yoshida Y, Ohashi K, Sano E, Kobayashi H, 172, 4059–4067. Endo K, Naruto M and Nakamura T. (1992). Jpn. J. Cancer Russell JP, Powell DJ, Cunnane M, Greco A, Portella G, Res., 83, 153–158. Santoro M, Fusco A and Rothstein JL. (2000). Oncogene, Wang M, Tan Z, Zhang R, Kotenko SV and Liang P. (2002). 19, 5729–5735. J. Biol. Chem., 277, 7341–7347. Russell JP, Shinohara S, Melillo RM, Castellone MD, Santoro Winterrowd GE and Chin JE. (1999). J. Immunol. Methods, M and Rothstein JL. (2003). Oncogene, 22, 4569. 226, 105–118. Saeki T, Mhashilkar A, Chada S, Branch C, Roth JA and Wirtschafter A, Schmidt R, Rosen D, Kundu N, Santoro M, Ramesh R. (2000). Gene Therapy, 7, 2051–2057. Fusco A, Multhaupt H, Atkins JP, Rosen MR, Keane WM Salvatore D, Barone MV, Salvatore G, Melillo RM, Chiappetta and Rothstein JL. (1997). Laryngoscope, 107, 95–100. G, Mineo A, Fenzi G, Vecchio G, Fusco A and Santoro M. Yacoub A, Mitchell C, Brannon J, Rosenberg E, Qiao L, (2000). J. Clin. Endocrinol. Metab., 85, 3898–3907. McKinstry R, Linehan WM, Su ZS, Sarkar D, Lebedeva IV, Santoro M, Carlomango F, Hay ID, Herrmann MA, Grieco Valerie K, Gopalkrishnan RV, Grant S, Fisher PB and Dent M, Melillo R, Pierotti MA, Bongarzone I, Della Porta G, P. (2003a). Mol. Cancer Ther., 2, 623–632. Berger N, Peix JL, Paulin C, Fabian N, Vecchio G, Jenkins Yacoub A, Mitchell C, Lister A, Lebedeva IV, Sarkar D, RB and Fusco A. (1992). J. Clin. Invest., 89, 1517–1522. Su ZZ, Sigmon C, McKinstry R, Ramakrishnan V, Sarkar D, Su ZZ, Lebedeva IV, Sauane M, Gopalkrishnan Qiao L, Broaddus WC, Gopalkrishnan RV, Grant S, RV, Dent P and Fisher PB. (2002). Biotechniques, 33 Fisher PB and Dent P. (2003b). Clin. Cancer Res., 9, (Suppl), 30–39. 3272–3281. Scarpino S, Stoppacciaro A, Ballerini F, Marchesi M, Prat M, Yoshida M, Matsuzaki H, Sakata K, Takeya M, Kato K, Stella MC, Sozzani S, Allavena P, Mantovani A and Ruco Mizushima S, Kawakita M and Takatsuki K. (1992). Cancer LP. (2000). Am. J. Pathol., 156, 831–837. Res., 52, 464–469. Sheils OM, O’Leary JJ, Uhlmann V, Luttich K and Sweeney Zhang R, Tan Z and Liang P.. (2000). J. Biol. Chem., 275, EC. (2000). Int. J. Surg. Pathol., 8, 185–189. 24436–24443.

Oncogene