Oncogene (2009) 28, 2784–2795 & 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00 www.nature.com/onc ORIGINAL ARTICLE Cyclophilins contribute to Stat3 signaling and survival of multiple myeloma cells

K Bauer1,2,7, AK Kretzschmar1,6,7, H Cvijic1, C Blumert1,DLo¨ffler1,2, K Brocke-Heidrich1,2, C Schiene-Fischer3, G Fischer3, A Sinz4, CV Clevenger5 and F Horn1,2,6

1Institute of Clinical Immunology and Transfusion Medicine, Medical Faculty, University of Leipzig, Leipzig, Germany; 2Interdisciplinary Center for Clinical Research, University of Leipzig, Leipzig, Germany; 3Max Planck Research Unit for Enzymology of Folding, Halle (Saale), Germany; 4Department of Pharmaceutical Chemistry and Bioanalytics, Institute of Pharmacy, Martin Luther University Halle-Wittenberg, Halle (Saale), Germany; 5Department of Pathology, Northwestern University, Chicago, IL, USA and 6Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany

Signal transducer and activator of transcription 3 (Stat3) controlled cellular processes such as differentiation, is the major mediator of interleukin-6 (IL-6) family proliferation and cell survival. Stat3 has been first cytokines. In addition, Stat3 is known to be involved in the described by us as a central component of interleukin-6 pathophysiology of many malignancies. Here, we show (IL-6) signaling, which mediates the IL-6-induced that the cis-trans peptidyl-prolyl cyclophilin production of acute phase in hepatocytes (Cyp) B specifically interacts with Stat3, whereas the (Wegenka et al., 1993; Yuan et al., 1994). Activation highly related CypA does not. CypB knockdown inhibited of Stat3 occurs through phosphorylation of tyrosine the IL-6-induced transactivation potential but not the 705, mediated by receptor-associated janus (JAK) tyrosine phosphorylation of Stat3. Binding of CypB to protein tyrosine kinases (Lu¨tticken et al., 1994). The Stat3 target promoters and alteration of the intranuclear factor translocates to the nucleus where it binds to localization of Stat3 on CypB depletion suggested a regulatory elements of target genes, controlling their nuclear function of Stat3/CypB interaction. By contrast, transcription (Yuan et al., 1994). In addition to IL-6, CypA knockdown inhibited Stat3 IL-6-induced tyrosine other cytokines, growth factors and hormones have phosphorylation and nuclear translocation. The Cyp been shown to activate Stat3, which is now known to be inhibitor cyclosporine A (CsA) caused similar effects. involved in a variety of physiological processes, includ- However, Stat1 activation in response to IL-6 or ing functions during development, in the immune interferon-c was not affected by Cyp silencing or CsA system, and homeostasis in many organs and tissues. treatment. As a result, Cyp knockdown shifted IL-6 Dysregulation of Stat3 signaling is involved in etiology signaling to a Stat1-dominated pathway. Furthermore, and/or progression of various diseases, for example, Cyp depletion or treatment with CsA induced apoptosis in autoimmune and chronic inflammatory diseases, as well IL-6-dependent multiple myeloma cells, whereas an IL-6- as cancer. In fact, being activated in many cancers, Stat3 independent line was not affected. Thus, Cyps support the is now widely regarded as an oncoprotein (Hodge et al., anti-apoptotic action of Stat3. Taken together, CypA and 2005). Stat3 is essential for transformation induced by CypB both play pivotal roles, yet at different signaling various oncogenes, and a constitutively active Stat3 levels, for Stat3 activation and function. These data also mutant causes cellular transformation (Bromberg et al., suggest a novel mechanism of CsA action. 1999). In many cancer cells, Stat3 signaling has been Oncogene (2009) 28, 2784–2795; doi:10.1038/onc.2009.142; recognized as a pivotal pathway supporting survival and published online 8 June 2009 growth (Schwarze and Hawley, 1995; Catlett-Falcone et al., 1999; Rawat et al., 2000; Puthier et al., 2001). Keywords: Stat3; cyclophilin B; cyclophilin A; cyclo- The Stat3 signaling is modulated, both positively and sporine A; interleukin-6; multiple myeloma negatively, by its interaction with numerous other proteins, and crosstalk occurs with various other signaling cascades, including the NF-kB, AP-1 or PI- Introduction 3K pathways. Here, we show that Stat3 activation and function are supported by the cyclophilins (Cyps) A and Signal transducer and activator of transcription-3 B, both targets of the cyclos- (Stat3) has a key regulatory role in various cytokine- porine A (CsA). Cyps are members of the immunophilin family of proteins that show peptidyl-prolyl cis-trans Correspondence: Dr F Horn, Institute of Clinical Immunology and isomerase (PPIase) activity. They are highly conserved Transfusion Medicine, University of Leipzig, Johannisallee 30, 04103 in the phylogenetic tree, from archaebacteria to verte- Leipzig, Germany. brates, suggesting a crucial role in the function of living E-mail: [email protected] 7These authors contributed equally to this work. cells (Wang and Heitman, 2005). Received 20 October 2008; revised 21 February 2009; accepted 31 March Cyclophilin A is best known for its role in inhibiting 2009; published online 8 June 2009 in a complex with CsA (Wang and Heitman, Interaction of Stat3 with cyclophilin B K Bauer et al 2785 2005). Furthermore, CypA is involved in multiple signaling events, for example, it regulates IL-2 tyrosine kinase activity (Brazin et al., 2002), is required for retinoic acid-induced neuronal differentiation (Song et al., 2004) and modulates prolactin receptor signaling (Syed et al., 2003). CypA also affects gene expression by physically interacting with transcription factors and histone-modifying (Arevalo-Rodriguez et al., 2000; Cui et al., 2002). CypB was also described to be involved in signal transduction pathways and transcrip- tional control (Rycyzyn et al., 2000; Rycyzyn and Clevenger, 2002; Obata et al., 2005). Furthermore, both CypA and CypB are secreted from cells in response to inflammatory stimuli (Sherry et al., 1992; De Ceuninck et al., 2003). Secreted CypA has multiple functions in chemotaxis and cell signaling through its cellular receptor, CD147 (Yang et al., 2008). Both CypA and CypB are found in high concentrations in the synovial fluid of arthritic joints, suggesting their role in chronic Figure 1 Identification and specificity of Stat3–CypB binding. inflammatory diseases (Billich et al., 1997; De Ceuninck (a) HepG2 cells treated without or with 10 ng/ml interleukin-6 et al., 2003). (IL-6) for 30 min were crosslinked, lysed and immunoprecipitated Here, we show that both CypA and CypB support (IP) with Stat3 or cyclophilin (Cyp) B antibodies. As a control (ctr), Stat3 signaling, yet at different levels of the pathway. immunoprecipitation procedure was performed without antibody. Immunoblotting (IB) was performed with the antibodies as While CypB/Stat3 interaction seems functionally rele- indicated for the immunoprecipitates or lysate (input). (b) HepG2 vant mainly in the nucleus, CypA is required to allow cell lysates were immunoprecipitated with Stat3 antibody as above maximal Stat3 tyrosine phosphorylation and nuclear and subjected to immunoblot analysis using antibodies to either translocation. Both Cyps proved to be required for the CypB or CypA. (c) HepG2 cells were treated with cyclosporine A (CsA) (10 mg/ml). Immunoprecipitation from cell lysates and survival of Stat3-dependent multiple myeloma cells. immunoblotting was as above. Stat3, signal transducer and activator of transcription 3.

Results quently, intracellular CypB is mainly found associated with the compartment (Wang Stat3 interacts specifically with CypB and Heitman, 2005). By contrast, Stat3 is preferentially To identify proteins that interact with Stat3, we cytoplasmic in untreated HepG2 cells and translocates analysed proteins co-immunoprecipitated with Stat3 to the nucleus on activation in response to IL-6. from lysates of HepG2 human hepatocellular carcinoma To analyse in which cellular compartment the cells using mass spectrometry as described in Materials interaction of Stat3 and CypB may take place, CypB and methods. To also identify protein interactions that intracellular localization in untreated or IL-6-stimulated are not sufficiently stable to withstand lysis and HepG2 cells was studied by indirect immunofluores- precipitation procedures, we chose to crosslink the cence. In untreated cells, CypB was found preferentially living cells beforehand using a chemical crosslinker. in a perinuclear distribution and in lower amounts in the This approach identified, among others, CypB as a nucleus and (Figure 2a). Interestingly, after novel Stat3 interactor. As shown in Figure 1a, the treatment with IL-6, CypB redistributed to a cytoplas- interaction of Stat3 and CypB could be verified in mic and nuclear localization in approximately 10% of HepG2 lysates by immunoprecipitation and subsequent the cell population. To characterize this process further, immunoblotting in both directions. The interaction was a carboxy-terminal fusion of CypB with yellow fluor- detected in untreated HepG2 cells and slightly increased escent protein (YFP) was constructed and analysed in on IL-6 stimulation. This suggests that Stat3 activation, HepG2 cells. Co-transfection of the endoplasmic though not a prerequisite, may support the association reticulum marker DsRed-ER showed that the bulk of of the two proteins. In contrast to CypB, the highly CypB-YFP colocalized with that compartment related CypA was not found to interact with Stat3 (Figure 2b–d, upper row). Similar to endogenous CypB, (Figure 1b). Treatment with the immunosuppressant CypB-YFP redistributed in a fraction of the cells after CsA, known to bind and inhibit the enzymatic activity the addition of IL-6 (Figure 2b–d, lower row). of Cyps, reduced but did not abolish Stat3/CypB interaction (Figure 1c). The CypA and CypB are highly homologous within Functional importance of Cyps for Stat3 activation their catalytic core domains. The latter, however, and function possesses additional amino- and carboxy-terminal se- To analyse a physiological function of the CypB/Stat3 quences comprising a signal peptide and an endoplasmic interaction in living cells, we investigated the influence reticulum retention sequence, respectively. Conse- of CypB on the expression of the Stat3-specific gene

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2786

Figure 2 Redistribution of cyclophilin (Cyp) B in response to interleukin-6 (IL-6). (a) Immunofluorescence images of HepG2 cells subjected to CypB antisera and stained with FITC secondary antibody before and after IL-6 addition (10 ng/ml). (b) HepG2 cells co-transfected with DsRed-ER and CypB-YFP were treated with IL-6 (10 ng/ml) for 30 min as indicated. YFP, yellow fluorescent protein.

regulation using reporter gene analysis. Overexpression To investigate the level of Stat3 signaling the Cyps act of CypB in HepG2 cells yielded a small but significant on, we next studied whether their knockdown interferes increase in the Stat3-dependent promoter activity of the with IL-6-induced activation of Stat3, that is, phosphor- a1-antichymotrypsin (ACT) gene (Figure 3a). ylation of tyrosine residue 705. To this purpose, HepG2 Unexpectedly, overexpression of CypA caused a cells were transiently transfected with small hairpin comparable increase in ACT promoter activity. By RNA vectors as above, together with an expression contrast, the IL-6-unresponsive promoter of the herpes vector for a Stat3-YFP fusion protein that we showed virus thymidine kinase (tk) gene was not supported earlier to be fully functional (Kretzschmar et al., 2004). by Cyp overexpression, suggesting that the observed This fusion protein allowed us to selectively monitor effect is specific for the Stat3-regulated promoter. Stat3 tyrosine phosphorylation in the transfected cells. In contrast to wild-type CypB, a catalytically compro- As shown in Figure 4a, Stat3 phosphorylation was not mised mutant (CypB-DPPI) did not show any effect on affected by a knockdown of CypB but was clearly ACT promoter activity (Figure 3b), suggesting that inhibited by CypA depletion. PPIase activity is required for CypB to act on Stat3 The IL-6 activates not only Stat3 but also the highly signaling. related Stat1. This raised the question whether the As reported earlier, extracellular CypB supports observed inhibitory effect of CypA silencing represents a prolactin/prolactin receptor/Stat5 signaling, is interna- general block of IL-6 signaling or whether it may be lized together with prolactin, and was shown to support specific for Stat3 activation. As shown in Figure 4b, the transactivating potential of Stat5 in the nucleus neither CypB nor CypA knockdown inhibited Stat1 (Rycyzyn et al., 2000; Rycyzyn and Clevenger, 2002). As activation. Likewise, Stat1 phosphorylation in response of CypB is enhanced under inflammatory to interferon-g remained unchanged (Figure 4c). Taken circumstances, we figured that its IL-6-driven redistri- together, these data suggest that CypA specifically bution as observed above might reflect a secretion/ contributes to Stat3 signaling at the level of tyrosine reuptake pathway, and that the effects of CypB on Stat3 phosphorylation, whereas the supportive effect of CypB function might be due to internalized CypB. In fact, acts on a level distal from Stat3 activation. recombinant CypB, when added to the medium of The ACT promoter is specifically transactivated by HepG2 cells, increased the IL-6-induced activity of the Stat3 but not by Stat1 (FH, unpublished results). Given ACT promoter to a comparable extent, as did intracel- our observation that CypA depletion inhibited Stat3, lular overexpression (data not shown). but not Stat1 activation, the question arises how genes The moderate CypB-induced increase in ACT pro- that are targets for either factor respond to a CypA moter activity likely reflects the fact that most cells knockdown. Using reporter gene assays with the express rather high protein levels of CypB. Therefore, promoter of the intercellular adhesion molecule-1 we next used RNA interference to investigate the (ICAM-1) gene that is strongly induced by both IL-6 influence of a Cyp knockdown on Stat3. Expression of and interferon-g, we observed an increased rather than small hairpin RNAs targeting either CypA or CypB decreased luciferase activity on Cyp silencing mRNAs yielded a reduction but not complete depletion (Figure 4d). It is noted that this was true for both of Cyp protein levels (see Supplementary Figure 1). This Cyps, suggesting that the supportive role of CypB is is most likely due to the rather slow turnover of Cyp Stat3-specific as well. proteins. When tested by luciferase reporter assays in The Stat3 tyrosine phosphorylation is followed by HepG2 cells, knockdown of either Cyp significantly translocation of the factor to the nucleus. We therefore inhibited Stat3-driven ACT promoter activity next studied whether the interaction with CypB might (Figure 3c). This leads to the conclusion that, in influence this translocation process. Stat3-YFP was accordance with the overexpression studies above, both expressed by transient transfection in HepG2 cells and CypA and CypB support Stat3 function. its subcellular distribution visualized by fluorescence

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2787 microscopy (for a complete representation of the results for 48 or 72 h prevented nuclear translocation of Stat3- obtained, see Supplementary Figure 2). YFP, as was expected from its inhibitory effect on In mock-transfected cells, Stat3 was uniformly dis- tyrosine phosphorylation (Figure 5 and Supplementary tributed throughout the cytoplasm and nucleus in the Figure 2, part A). By contrast, CypB knockdown for absence of IL-6. After 30 min of IL-6 treatment, 48 h did not affect nuclear translocation (Supplementary however, Stat3 preferentially localized to the nucleus, Figure 2, part A). However, when cells were studied 72 h as apparent by its colocalization with the nuclear after transfection (a time point where CypB knockdown marker DS-Red-Nuc (Figure 5). Depletion of CypA was more effective), the situation changed: Although the cytoplasmic distribution of nonactivated Stat3 remained unaffected, Stat3 accumulated in the nucleus of a majority of cells in a dotlike pattern. This result shows that CypB silencing does not impede nuclear accumula- tion of Stat3 but affects its intranuclear distribution (Figure 5). The subcellular localization of a Stat1-YFP fusion protein was studied in human embryonic kidney 293 cells stimulated through either the interferon-g or the IL-6 pathway. Owing to the low levels of IL-6 receptor a subunits in these cells, stimulation was performed with hyper-IL-6, a fusion protein of IL-6 and the soluble IL-6 receptor a that acts directly on the signal transducing gp130 subunit of the IL-6 receptor (Peters et al., 1998). Neither CypA nor CypB silencing blocked Stat1 nuclear translocation. This was obvious both 48 h (Supplemen- tary Figure 2, part B) and 72 h post-transfection (not shown). As evident from the above data, CypB/Stat3 interac- tion affects nuclear functions and localization of the transcription factor. Hence, we reasoned that CypB might interact with the carboxy-terminal transactivation domain of Stat3 and/or might associate with Stat3 target promoter regions. Co-immunoprecipitation experiments using a truncated Stat3 mutant lacking the transactivation domain (Stat3-D715) showed that it is not required for the Stat3/CypB interaction (Figure 6a). However, chromatin immunoprecipitation (ChIP) studies in HepG2 cells indicated an association of CypB with the Stat3-binding region of the ACT gene (Figure 6b). ChIP analysis of the endogenous ACT gene generally produced low signals in HepG2 cells. There- fore, we extended the studies to additional Stat3 target genes, that is, the junB and haptoglobin genes (Figure 6c). The data show strong binding of CypB even in the absence of IL-6 stimulation. On IL-6

Figure 3 The transactivation potential of Stat3 is supported by cyclophilins (Cyps). (a, b) Luciferase reporter assays in HepG2 cells: cells were transiently transfected with expression vectors qfor CypA, CypB, CypB-DPPI or with a control vector (ctr). a1-antichymotrypsin (ACT) promoter luciferase construct (pACT-Luc) and b-galactosidase (b-gal) expression vector under the control of the early SV40 T antigen promoter were co- transfected. At 48 h after transfection, cells were treated without or with interleukin-6 (IL-6) (10 ng/ml) for 4 h and harvested thereafter. Luciferase activities as determined in the cell lysates were normalized to b-gal activities to account for transfection efficiency. (c) HepG2 cells were transfected with pSUPER- shCypA, pSUPER-shCypB or an empty pSUPER (ctr). After 72 h, the cells were treated without or with IL-6 for 4 h and the luciferase and b-gal activities were measured. All values represent the means of three independent experiments, ±standard devia- tion. tk, tyrosin kinase.

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2788

Figure 4 Cyclophilin (Cyp) A knockdown inhibits Stat3 tyrosine phosphorylation. (a) HepG2 cells were transiently transfected with a Stat3-YFP expression vector and with pSUPER-shCypA, pSUPER-shCypB or empty pSUPER (ctr). At 72 h after transfection, cells were treated without or with interleukin-6 (IL-6) for 30 min, harvested and the cell lysates were analysed by immunoblotting with antibodies to tyrosine-phosphorylated Stat3 (anti-pY-Stat3), Stat3 or b-actin. (b) HepG2 cells transfected with a Stat1-YFP expression vector and pSUPER vectors as above were IL-6-treated and analysed by immunoblotting as in (a). (c) Human embryonic kidney 293 cells were transfected with a Stat1-YFP expression vector and pSUPER vectors as above. After 72 h, the cells were treated with interferon-g (IFN-g, 1000 U/ml) for 30 min or left untreated, and immunoblot analysis was performed. (d) HepG2 cells were transfected with pSUPER-shCypA, pSUPER-shCypB or empty pSUPER (ctr), together with pICAM-1-Luc and b-gal expression vector. At 72 h post-transfection, cells were treated for 4 h with IL-6 or IFN-g and the luciferase and b-gal activities were measured. Values represent the means of three independent experiments, ±s.d. Stat3, signal transducer and activator of transcription 3; YFP, yellow fluorescent protein.

Figure 5 Influence of cyclophilin (Cyp) knockdown on Stat3 nuclear translocation. HepG2 cells were transiently transfected with expression vectors encoding Stat3-YFP and DS-Red-Nuc and with pSUPER-shCypA, pSUPER-shCypB or empty pSUPER (ctr). At 72 h after transfection, cells were treated with interleukin-6 (IL-6) for 30 min or left untreated, followed by monitoring of the subcellular localization of Stat3 using fluorescence microscopy. Stat3, signal transducer and activator of transcription 3; YFP, yellow fluorescent protein.

addition, binding of CypB decreased slightly. By when nonspecific immunoglobulins were used contrast, a control gene (b-actin) that is not bound by (Figure 6c). In contrast to CypB, we could not detect Stat3 did not show substantial CypB binding. Specificity any considerable association of CypA with the promo- of the ChIP assay was verified by the absence of signals ters tested.

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2789

Figure 6 Cyclophilin (Cyp) B associates with Stat3 target promoters. (a) Human embryonic kidney 293 cells were transiently transfected with expression vectors encoding either wild-type Stat3 or carboxy-terminally truncated Stat3-D715, crosslinked and lysed. Immunoprecipitation and immunoblotting were performed with antibodies to CypB and Stat3, respectively. (b, c) HepG2 cells, untreated or treated with interleukin-6 (IL-6) for 30 min, were subjected to a chromatin immunoprecipitation (ChIP) assay using CypB, CypA or Stat3 antibodies, or to control immunoglobulins. The isolated DNA was analysed by PCR using primers specific for the Stat3-binding regions of the a1-antichymotrypsin (ACT) (b) or junB and haptoglobin promoters (c). As a control, the upstream region of the b-actin gene was used. Total chromatin (input) was used as a template for haptoglobin PCR to show equal amounts of the starting material. Stat3, signal transducer and activator of transcription 3.

Role of Cyps for the Stat3-dependent survival and growth (Supplementary Figure 3d), the only Bcl-2 family of multiple myeloma cells member found previously to be upregulated in We next investigated whether the supportive role of these cells in response to IL-6 (Brocke-Heidrich et al., CypA and CypB for Stat3 activation and function, 2004). respectively, translates into an influence on Stat3- To further substantiate a role for Cyps in myeloma dependent cell physiological responses. To this purpose, cell survival, we next silenced CypA and CypB by we chose the Stat3-dependent survival and growth of transfecting the IL-6-dependent myeloma lines INA-6 multiple myeloma cells. As we had shown earlier, the and XG-1 with small hairpin RNA vectors as described human myeloma cell lines INA-6 and XG-1 strictly above for HepG2 cells. As a control, we included the depend on IL-6-induced Stat3 activation to grow and to IL-6-independent myeloma line MM1.S in the study. be protected against apoptosis (Brocke-Heidrich et al., Cells were co-transfected with an expression vector 2004; Lo¨ffler et al., 2007). encoding enhanced green fluorescent protein to allow Inhibition of CypA and CypB by treatment with CsA gating for transfected cells during flow cytometric in the presence of IL-6 rapidly induced apoptosis in analyses. As evident from Figure 8, knockdown of INA-6 cells and caused a G1 cell cycle arrest in XG-1 either CypA or CypB yielded similar rates of apoptosis cells, as determined by flow cytometric annexin V induction as did a Stat3 knockdown in both INA-6 and apoptosis and propidium iodine cell-cycle assays, XG-1 cells. However, neither Stat3 nor Cyp knockdown respectively (Figure 7a). These responses were induced apoptosis in MM1.S cells. Therefore, there is a equivalent to those observed when the cell lines were close correlation between the anti-apoptotic actions of withdrawn from IL-6 (Figure 7a and (Brocke-Heidrich Cyps and Stat3 in myeloma cells. et al., 2004)). Furthermore, CsA treatment induced As discussed above, CypB might at least partially act rapid inhibition of Stat3 but not Stat1 tyrosine on Stat3 signaling as the secreted form after its reuptake phosphorylation in these myeloma cells (Figure 7b and by the cell. In fact, when CypB-depleted XG-1 cells were c). To determine whether CsA-induced apoptosis reconstituted by the addition of recombinant CypB to occurred through the extrinsic or intrinsic pathway, we the medium, apoptosis induction was significantly measured caspase 8 and 9 cleavage in INA-6 cells (see reduced (Figure 8d). The catalytically inactive mutant Supplementary Figure 3a and b). Both caspases were CypB-PPI was completely ineffective in that respect, found to be inducibly cleaved, indicating that both the proving the requirement of PPIase activity for this CypB receptor-dependent and the mitochondrial apoptosis action. Likewise, when INA-6 cells were cultivated pathways are activated by CsA. Studies with inhibitors under reduced levels of IL-6 (0.1 ng/ml), resulting in specific for caspase 8 or 9 supported this view a moderate induction of apoptosis, treatment with (Supplementary Figure 3c). Furthermore, we observed recombinant CypB protein caused a significant decrease reduced levels of Mcl-1 on CsA treatment in INA-6 cells of the apoptotic poly-(ADP-ribose) polymerase

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2790 Discussion

There is growing evidence for a role of PPIases in transcriptional control. Various PPIases were found to interact with transcription factors and affect their activity. This includes the inhibitory action of Cyp-40 and FKBP52 on cMyb and IRF-4, respectively, and the supportive interaction of CypB with IRF-3 (Leverson and Ness, 1998; Mamane et al., 2000; Obata et al., 2005). Furthermore, pin1 interacts with NF-kB p65, the tumor suppressor p53, as well as Stat3 supporting their transactivating potential (Zacchi et al., 2002; Ryo et al., 2003; Lufei et al., 2007). CypB was shown to interact with another member of the STAT family, Stat5 (Rycyzyn and Clevenger, 2002). In breast cancer cells, this interaction enhances the transcriptional activity of Stat5, most likely by supporting Stat5 DNA binding and release of the inhibitor PIAS3. Hence, the interaction of CypB and Stat3 as reported here is not entirely unexpected. Binding of CypB to Stat3 was observed only when cells were crosslinked, indicating either low- affinity binding or a high on/off rate. In fact, we were recently able to verify Stat3/CypB interaction in vitro by surface plasmon resonance studies, and these data indicate a limited but significant affinity between the two proteins and substantiate the requirement of the CypB PPIase activity for their interaction (manuscript in preparation). In contrast to pin1 that binds to the transactivation domain of Stat3, CypB association did not depend on that domain. This is in line with the observation that CypB interacts with an amino-terminal part of Stat5 (Rycyzyn and Clevenger, 2002). As observed for Stat5, CypB supported the transactivating potential of Stat3 toward a Stat3-specific target promoter. This was shown by both ectopic overexpression and silencing of CypB. Although the underlying mechanism is not yet clear, our data provide evidence that Stat3/CypB interaction is a Figure 7 Cyclosporine A (CsA) inhibits myeloma cell survival and nuclear phenomenon. This view is further supported by growth. (a) INA-6 (left) myeloma cells were cultivated in the the presence of CypB at Stat3 target promoters. As this presence or absence of interleukin-6 (IL-6) (1 ng/ml) for 12 h, without or with CsA (80 mM) as indicated. Apoptosis was association was most significant in the absence of IL-6 determined by flow cytometric annexin V assay. Values indicate stimulation, it needs to be clarified how CypB is directed viable annexin V and propidium iodide-negative cells. XG-1 cells specifically to such promoter regions. CypB silencing did (right) were cultivated in the presence or absence of IL-6 (2 ng/ml) not inhibit Stat3 tyrosine phosphorylation and nuclear for 72 h, without or with CsA (80 mM) as indicated. Cell cycle was measured using flow cytometric propidium iodide staining as translocation but caused Stat3 to localize at punctuate described in Materials and methods. All data represent means±s.d. structures within the nucleus reminiscent of promyelo- from three independent experiments. (b) INA-6 cells cultivated in cytic leukemia bodies. As promyelocytic leukemia the presence of IL-6 (1 ng/ml) were incubated with CsA (10 mM)or bodies are associated with PIAS proteins and SUMOy- solvent control (ctr) for the time period indicated. Thereafter, cell lation activity, this observation suggests that—as in the lysates were subjected to immunoblotting using antibodies to phospho-Stat3, Stat3 and b-actin. (c) INA-6 cells cultivated with case of Stat5—CypB might release a PIAS factor from IL-6 (1 ng/ml) were incubated with interferon g (1000 U/ml) for the Stat3. time period indicated, in the presence of CsA (10 mM) or solvent The CypB is expressed in many tissue types and control (ctr). Cell lysates were then analysed by immunoblotting resides mostly within the endoplasmic reticulum. Secre- using antibodies to phospho-Stat1, Stat1 and b-actin. Stat3, signal transducer and activator of transcription 3. tion of CypB can yield appreciable protein levels in blood and milk, and is significantly increased by inflammatory stimuli (Mariller et al., 1996; Bukrinsky, fragments, again indicative of a protective effect of 2002). Subsequently, it is immobilized on the cell surface CypB treatment (Supplementary Figure 3e). through interaction with proteoglycans. The fact that Taken together, we conclude that CypA and CypB CypB is an intrinsic component of cartilage, is found in support myeloma cell survival, most likely through their high concentrations in the rheumatoid arthritis synovial action on Stat3 signaling. fluid, and that it can be released by matrix metallopro-

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2791

Figure 8 Apoptosis induction by cyclophilin (Cyp) A and CypB silencing in interleukin-6 (IL-6)-dependent myeloma cells. (a–c) Various myeloma cell lines were transfected with an enhanced green fluorescent protein (EGFP) expression vector and with pSUPER- shCypA, pSUPER-shCypB, pSUPER-shStat3 or empty pSUPER (ctr). The IL-6-dependent lines INA-6 (a) and XG-1 (c) were cultivated in the presence of IL-6, the independent line MM1.S without cytokine. After the time indicated, the cells were subjected to flow cytometric annexin V apoptosis analysis. Significance of data obtained from sh-expressing versus control cells were calculated by Student’s t-test, yielding P values o0.01 at 72 h for all cases. (d) XG-1 cells were transfected with an EGFP expression vector and with either pSUPER-shCypB or empty pSUPER (ctr). Where indicated, recombinant CypB or CypB-PPI was added twice (24 and 48 h after transfection) to the medium. Apoptosis was measured by flow cytometric annexin V analysis. Statistical evaluation (t-test) for the effect of recombinant CypB on sh-CypB-treated cells yielded Po 0.01 (**). All data represent means±s.d. of three independent experiments. Stat3, signal transducer and activator of transcription 3. teinases from the surface of chondrocytes, suggests that accelerated by IL-6 stimulation of HepG2 cells (KB, it has a role in the pathogenesis of arthritic diseases FH, unpublished results). (De Ceuninck et al., 2003) and possibly other inflam- The physiological relevance of our observations was matory and autoimmune diseases. Here, recombinant further studied in myeloma cell lines, which require CypB added to the culture medium showed protective activated Stat3 for proliferation and survival (Brocke- effects toward myeloma cells and was as potent a Heidrich et al., 2004). CypB knockdown induced supporter of transcriptional Stat3 function as was apoptosis in these cells to a comparable extent as did intracellular overexpression. Therefore, it seems possible Stat3 silencing, suggesting a critical role for the CypB/ that nuclear CypB, which interacts with Stat3, princi- Stat3 interaction in this process. It will be interesting to pally derives from such a secretion/reuptake pathway see whether this represents a global mechanism, that is, and that the redistribution of CypB, which we observe in whether other Stat3-dependent tumor cell lines require response to IL-6, reflects such a pathway. This is CypB for their growth and survival as well. supported by the finding that CypB supports Stat5- The involvement of Cyps in myeloma growth and mediated signal transduction and enters the cell in a survival as reported here is in line with a growing complex with prolactin (Rycyzyn et al., 2000; Rycyzyn number of observations on the oncological relevance of and Clevenger, 2002). Similar to the IL-6 receptor, the CypA and CypB. CypA was found to be overexpressed prolactin receptor belongs to the type-I cytokine in many cancer cells, including the pancreas (Shen et al., receptor family. Furthermore, CypB is internalized by 2004) and non-small cell lung carcinomas, (Howard T lymphocytes, and the internalization of CypB in et al., 2004) and to render cancer cells resistant to complex with CsA induced tolerance of these cells hypoxia-induced cell death (Choi et al., 2007). CypA against organ transplants (Allain et al., 1996). In fact, knockdown by siRNAs reduced non-small cell lung preliminary data suggest that the uptake of CypB is tumor growth in vivo (Howard et al., 2005). Likewise,

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2792 CypB expression is associated with malignant progres- pcDNA3.1( þ ) (Invitrogen, Karlsruhe, Germany), yielding sion of breast cancer (Fang et al., 2009). pcDNA-CypA. Likewise, pcDNA-CypB was amplified by Our data show that CypA and CypB act on the Stat3 PCR from pcDNA3.1/V5-His-CypB (Rycyzyn et al., 2000) pathway at different levels. CypA depletion inhibited and inserted into the EcoRI and NotI sites of pcDNA3.1( þ ). Stat3 tyrosine phosphorylation and nuclear transloca- A vector encoding CypB-YFP was generated by amplifying the tion. Hence, CypA might interact with the IL-6 receptor CypB cDNA and by inserting it into the BglII/Asp718 site of the vector pEYFP-N1 (Clontech, Saint-Germain-en-Laye, complex or indirectly modulate the activity of the France). receptor-associated JAK tyrosine kinases. In fact, CypA was found to functionally interact with both the prolactin receptor and the JAK2 kinase (Syed et al., Preparation of recombinant CypB 2003), and this interaction correlates with mammary Escherichia coli, strain BL21, were transformed with the carcinoma progression (Zheng et al., 2008). As the IL-6 expression constructs GST-CypB or GST-CypB-PPI (Rycyzyn and Clevenger, 2002) and cultured in Luria Bertani medium receptor gp130 subunit is related to the prolactin until an optical density of 0.7 was reached. Protein expression receptor and also binds JAK2, it seems quite possible was induced by the addition of 1 mM isopropyl b-D-1- that an analogous interaction also takes place at the IL- thiogalactopyranoside (Sigma Aldrich, Taufkirchen, 6 receptor. In addition, IL-6 signaling requires the Germany) to the medium for 4 h. The bacteria were solubilized localization of IL-6 receptors in caveolae (Podar et al., using a french press as described (Kofron et al., 1991; Rycyzyn 2003), and caveolin-1, a protein found in these and Clevenger, 2002). Recombinant proteins were purified structures, is also known to interact with CypA through coupling to glutathione-Sepharose (Amersham (Uittenbogaard et al., 1998). However, a general Bioscience, Uppsala, Sweden) and cleavage by 50 units inhibition of IL-6 receptor and JAK kinase function thrombin (Calbiochem, Darmstadt, Germany). The protein by CypA depletion cannot be expected, as we observe was then eluted in phosphate-buffered saline. that Stat1 phosphorylation is not affected. Stat1 recruitment to gp130 occurs through tyrosine motifs Cell culture, transfection and reporter assays that also recruit Stat3 (Gerhartz et al., 1996). Therefore, Human embryonic kidney 293 cells or HepG2 human if receptor and/or JAK kinase functions are modulated hepatoma cells were grown in DMEM F-12 containing 10% by CypA, it seems to occur in a target-specific manner. FCS and 1% penicillin/streptomycin (all from Life Techno- Alternatively, one might also consider that CypA does logies, Carlsbad, CA, USA). INA-6 and XG-1 myeloma cell not act on the receptor level but rather modulates the lines were cultured in RPMI medium containing 10% FCS, 1% penicillin/streptomycin and IL-6 (1 and 2 ng/ml, respec- Stat3 dephosphorylation rate by protein tyrosine tively). MM1.S myeloma cells were grown in the same medium . without the addition of IL-6. Our findings suggest that the action of CypA is Transfections were carried out using the phosphate crucial to maintain the dominance of Stat3 over Stat1 co-precipitation technique according to standard protocols. signaling in response to IL-6. This might be of particular The medium was replaced 6 h after transfection. At 24 h before importance for cells that rely on Stat3 for their stimulation, cells were placed in a medium containing 0.5% survival, as Stat1 is known to preferentially induce FCS. For stimulation of human embryonic kidney 293 and pro-apoptotic signals. This conclusion also has HepG2 cells, 10 ng/ml Hyper-IL-6 and 10 ng/ml IL-6, respec- interesting implications for the action of CsA. This tively, were used. Both were gifts from Stefan Rose-John immunosuppressive drug is successfully used in the (University Kiel, Germany). Interferon-g was from Peprotech (Hamburg, Germany). treatment of RA (Kitahara and Kawai, 2007). However, For reporter studies, HepG2 cells were grown in six-well the anti-inflammatory effect of CsA cannot be suffi- tissue culture dishes, transfected with reporter vectors, and ciently explained by its immunosuppressive function. luciferase and b-galactosidase assays were conducted as Therefore, our observation that CsA specifically blocks previously described (Brocke-Heidrich et al., 2006). the Stat3 but not the Stat1 pathway might represent a Myeloma cells were transfected using the Nucleofector missing link explaining the action of CsA on inflamma- technology (Amaxa Inc., Cologne, Germany) according to tory processes. optimized protocols provided by Amaxa.

Immunoblot analysis Cells were lysed as described (Bellido et al., 1998). Equal Materials and methods amounts of cellular protein were separated by electrophoresis through an sodium dodecyl sulfate-polyacrylamide gel, pro- Vector construction teins transferred to polyvinyldifluoride membrane (Amersham, Construction of pSUPER-siStat3, encoding a small-hairpin Buckinghamshire, UK) by semi-dry electroblotting, and RNA targeting Stat3, as well as of a control vector (pSUPER- immunodetection was performed by standard techniques using scrambled), has been recently described (Brocke-Heidrich chemiluminescence (SuperSignal, West Dura Substrate, Pierce, et al., 2006). Sequences encoding a small-hairpin RNA Rockfort, IL, USA). Anti-phospho-Stat3 (Y705), Stat3, Stat1 targeting CypA and CypB have been described by Watashi and caspase antibodies were obtained from Cell Signaling et al. (2005) and were cloned into the BglII and HinDIII sites (Boston, MA, USA) and BD Transduction Laboratories (San of pSUPER (Brummelkamp et al., 2002). Jose, CA, USA), respectively. Polyclonal antibodies against The CypA cDNA was amplified by PCR from pET11c- CypB and CypA were from Alexis (Lo¨rrach, Germany). Mcl-1 CypA, a gift from Hans-Georg Kra¨usslich (University Heidel- antibody was from Santa Cruz Biotechnology (Santa Cruz, berg, Germany), and inserted into the EcoRI and NotI sites of CA, USA).

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2793 Co-immunoprecipitation and mass spectrometry analysis sulfate lysis buffer. Lysates were sonicated four times for 15 s HepG2 cells were grown to 70% confluency, stimulated with at 60% amplitude (Sonifier 250, Branson, Danbury, CT, 1 ng/ml IL-6, treated for 30 min at 4 1C with 1 mM 3,30- USA). ChIP was performed according to the protocol of ChIP Dithiodipropionic acid di(N-succinimidyl-)ester (Lomant’s assay kit (Upstate Biotechnology, Lake Placid, NY, USA). Reagent), and harvested. Polyclonal antibodies to Stat3 Input DNA or DNA enriched by immunoprecipitation with (R&D Systems, Minneapolis, MN, USA) or to CypB (Alexis) anti-CypB, anti-CypA or anti-Stat3 was amplified by PCR. were added and immunoprecipitation was carried out with Primer sequences are available on request. mMACS Protein A magnetic beads (Miltenyi, Bergisch Gladbach, Germany) following the manufacturer’s instruc- Apoptosis and cell cycle analysis tions. Immunoprecipitate was eluted in Laemmli buffer Apoptosis was assessed using poly-(ADP-ribose) polymerase containing b-mercaptoethanol to reverse crosslinking. antibodies (Zymed, Carlsbad, CA, USA) or the annexin The immunoprecipitates were loaded onto a 10% sodium V-FITC apoptosis detection kit II (BD Biosciences, San Jose, dodecyl sulfate-polyacrylamide gel. Coomassie blue-stained CA, USA). Flow cytometric analysis was performed with an bands of interest were excised and digested in gel with trypsin FACScan flow cytometer using the CellQuest software (BD (Sequencing Grade, Roche, Mannheim, Germany) at 37 1C for Biosciences). 16 h, as previously described (Jensen et al., 1997). The resulting Cell cycle distribution was measured by flow cytometric peptide mixtures were analysed by nano-HPLC/nano-ESI- analysis using propidium iodide staining. In total, 3 Â 104 cells FTICR mass spectrometry, using an Ultimate Nano-LC were washed with phosphate-buffered saline and fixed in 80% system (Dionex, Idstein, Germany) coupled online to an Apex ethanol for 20 min at À20 1C. After rehydrating in phosphate- II FTICR mass spectrometer equipped with a 7 Tesla buffered saline for 15 min, the cells were incubated with the superconducting magnet (Bruker Daltonics, Leipzig, same solution containing 100 mg/ml RNase, 0.05% Tween 20 Germany) and a nano-electrospray ionization source (Agilent and 10 mg/ml propidium iodide. Fluorescence of FL2-A was Technologies, Santa Clara, CA, USA) according to a protocol analysed on a flow cytometer as above. Histograms were fitted described previously (Ihling et al., 2003). Protein identification using ModFit Software (BD Biosciences). was performed with the Mascot software (http://www. Caspase-8 inhibitor (Z-IETD-FMK) and Caspase-9 inhibi- matrixscience.com). tor (Z-LEHD-FMK), both from R&D Systems (Wiesbaden- Nordenstadt, Germany) were reconstituted using Dimethyl Fluorescence microscopy sulfoxide. The final concentration in cell culture medium was Images were captured at 37 1C on a heatable table using a DM 50 mM. IRB microscope with an HCX PL APO 63 Â /1.32 oil objective (Leica, Solms, Germany) and an ORCA-ER camera (Hama- matsu Photonics, Bridgewater, NJ, USA) under the control of Openlab software (Improvision, Tu¨bingen, Germany). CFP, Conflict of interest YFP and DsRed were detected using Polychrome IV (TILL Photonics, Gra¨felfing, Germany) at excitation wavelengths The authors declare no conflict of interest. 433, 510 and 550 nm, respectively. Expression plasmids for nuclear markers DsRed-Nuc and ECFP-Nuc and the marker Acknowledgements for endoplasmic reticulum DsRed-ER were purchased from Clontech. We thank Hans-Georg Kra¨usslich, University Heidelberg, Germany, and Stefan Rose-John, University Kiel, Germany, Chromatin immunoprecipitation for CypA vector and recombinant IL-6, respectively. INA-6 HepG2 cells were serum-starved for 2 days. After IL-6 cells were kindly provided by Renate Burger and Martin stimulation (10 ng/ml) for 30 min, HepG2 cells (1.5 Â 107) were Gramatzki, University Kiel, Germany. This work was crosslinked by 1% formaldehyde for 10 min at room tempera- supported by SFB610 (project C2) from the Deutsche ture. Cells were harvested and resuspended in sodium dodecyl Forschungsgemeinschaft.

References

Allain F, Denys A, Spik G. (1996). Cyclophilin B mediates cyclosporin isomerase cyclophilin A. Proc Natl Acad Sci USA 99: A incorporation in human blood T-lymphocytes through the 1899–1904. specific binding of complexed drug to the cell surface. Biochem Brocke-Heidrich K, Ge B, Cvijic H, Pfeifer G, Lo¨ffler D, Henze C J 317(Pt 2): 565–570. et al. (2006). BCL3 is induced by IL-6 via Stat3 binding to intronic Arevalo-Rodriguez M, Cardenas ME, Wu X, Hanes SD, Heitman J. enhancer HS4 and represses its own transcription. Oncogene 25: (2000). Cyclophilin A and Ess1 interact with and regulate silencing 7297–7304. by the Sin3-Rpd3 histone deacetylase. EMBO J 19: 3739–3749. Brocke-Heidrich K, Kretzschmar AK, Pfeifer G, Henze C, Lo¨ffler D, Bellido T, O’Brien CA, Roberson PK, Manolagas SC. (1998). Koczan D et al. (2004). Interleukin-6-dependent gene expression Transcriptional activation of the p21(WAF1,CIP1,SDI1) gene by profiles in multiple myeloma INA-6 cells reveal a Bcl-2 family- interleukin-6 type cytokines. A prerequisite for their pro- differ- independent survival pathway closely associated with Stat3 activa- entiating and anti-apoptotic effects on human osteoblastic cells. tion. Blood 103: 242–251. J Biol Chem 273: 21137–21144. Bromberg JF, Wrzeszczynska MH, Devgan G, Zhao Y, Billich A, Winkler G, Aschauer H, Rot A, Peichl P. (1997). Presence of Pestell RG, Albanese C et al. (1999). Stat3 as an oncogene. Cell cyclophilin A in synovial fluids of patients with rheumatoid 98: 295–303. arthritis. J Exp Med 185: 975–980. Brummelkamp TR, Bernards R, Agami R. (2002). A system for stable Brazin KN, Mallis RJ, Fulton DB, Andreotti AH. (2002). expression of short interfering RNAs in mammalian cells. Science Regulation of the tyrosine kinase Itk by the peptidyl-prolyl 296: 550–553.

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2794 Bukrinsky MI. (2002). Cyclophilins: unexpected messengers in Lu¨tticken C, Wegenka UM, Yuan J, Buschmann J, Schindler C, intercellular communications. Trends Immunol 23: 323–325. Ziemiecki A et al. (1994). Association of transcription factor APRF Catlett-Falcone R, Landowski TH, Oshiro MM, Turkson J, Levitzki and protein kinase JAK1 with the IL-6 signal transducer gp130. A, Savino R et al. (1999). Constitutive activation of Stat3 signaling Science 263: 89–92. confers resistance to apoptosis in human U266 myeloma cells. Mamane Y, Sharma S, Petropoulos L, Lin R, Hiscott J. (2000). Immunity 10: 105–115. Posttranslational regulation of IRF-4 activity by the immunophilin Choi KJ, Piao YJ, Lim MJ, Kim JH, Ha J, Choe W et al. (2007). FKBP52. Immunity 12: 129–140. Overexpressed cyclophilin A in cancer cells renders resistance Mariller C, Allain F, Kouach M, Spik G. (1996). Evidence that human to hypoxia- and cisplatin-induced cell death. Cancer Res 67: milk isolated cyclophilin B corresponds to a truncated form. 3654–3662. Biochim Biophys Acta 1293: 31–38. Cui Y, Mirkia K, Florence Fu YH, Zhu L, Yokoyama KK, Chiu R. Obata Y, Yamamoto K, Miyazaki M, Shimotohno K, Kohno S, (2002). Interaction of the retinoblastoma gene product, RB, with Matsuyama T. (2005). Role of cyclophilin B in activation of cyclophilin A negatively affects cyclosporin-inhibited NFAT interferon regulatory factor-3. J Biol Chem 280: 18355–18360. signaling. J Cell Biochem 86: 630–641. Peters M, Blinn G, Solem F, Fischer M, Meyer zum Buschenfelde KH, De Ceuninck F, Allain F, Caliez A, Spik G, Vanhoutte PM. (2003). Rose-John S. (1998). in vivo and in vitro activities of the High binding capacity of cyclophilin B to chondrocyte heparan gp130-stimulating designer cytokine Hyper-IL-6. J Immunol 161: sulfate proteoglycans and its release from the cell surface by matrix 3575–3581. metalloproteinases: possible role as a proinflammatory mediator in Podar K, Tai YT, Cole CE, Hideshima T, Sattler M, Hamblin A et al. arthritis. Arthritis Rheum 48: 2197–2206. (2003). Essential role of caveolae in interleukin-6- and insulin-like Fang F, Flegler AJ, Du P, Lin S, Clevenger CV. (2009). Expression of growth factor I-triggered Akt-1-mediated survival of multiple cyclophilin B is associated with malignant progression and myeloma cells. J Biol Chem 278: 5794–5801. regulation of genes implicated in the pathogenesis of breast cancer. Puthier D, Thabard W, Rapp M, Etrillard M, Harousseau J, Bataille Am J Pathol 174: 297–308. R et al. (2001). Interferon alpha extends the survival of human Gerhartz C, Heesel B, Sasse J, Hemmann U, Landgraf C, Schneider- myeloma cells through an upregulation of the Mcl-1 anti-apoptotic Mergener J et al. (1996). Differential activation of acute phase molecule. Br J Haematol 112: 358–363. response factor/STAT3 and STAT1 via the cytoplasmic domain of Rawat R, Rainey GJ, Thompson CD, Frazier-Jessen MR, Brown RT, the interleukin 6 signal transducer gp130. I. Definition of a novel Nordan RP. (2000). Constitutive activation of STAT3 is associated phosphotyrosine motif mediating STAT1 activation. J Biol Chem with the acquisition of an interleukin 6-independent phenotype by 271: 12991–12998. murine plasmacytomas and hybridomas. Blood 96: 3514–3521. Hodge DR, Hurt EM, Farrar WL. (2005). The role of IL-6 and STAT3 Rycyzyn MA, Clevenger CV. (2002). The intranuclear prolactin/ in inflammation and cancer. Eur J Cancer. cyclophilin B complex as a transcriptional inducer. Proc Natl Acad Howard BA, Furumai R, Campa MJ, Rabbani ZN, Vujaskovic Z, Sci USA 99: 6790–6795. Wang XF et al. (2005). Stable RNA interference-mediated suppres- Rycyzyn MA, Reilly SC, O’Malley K, Clevenger CV. (2000). Role of sion of cyclophilin A diminishes non-small-cell lung tumor growth cyclophilin B in prolactin signal transduction and nuclear retro- in vivo. Cancer Res 65: 8853–8860. translocation. Mol Endocrinol 14: 1175–1186. Howard BA, Zheng Z, Campa MJ, Wang MZ, Sharma A, Haura E Ryo A, Suizu F, Yoshida Y, Perrem K, Liou YC, Wulf G et al. (2003). et al. (2004). Translating biomarkers into clinical practice: Regulation of NF-kappaB signaling by Pin1-dependent prolyl prognostic implications of cyclophilin A and macrophage migratory isomerization and ubiquitin-mediated proteolysis of p65/RelA. inhibitory factor identified from protein expression profiles in non- Mol Cell 12: 1413–1426. small cell lung cancer. Lung Cancer 46: 313–323. Schwarze MM, Hawley RG. (1995). Prevention of myeloma cell Ihling C, Berger K, Hofliger MM, Fuhrer D, Beck-Sickinger AG, apoptosis by ectopic bcl-2 expression or interleukin 6-mediated up- Sinz A. (2003). Nano-high-performance liquid chromatography in regulation of bcl-xL. Cancer Res 55: 2262–2265. combination with nano-electrospray ionization Fourier transform Shen J, Person MD, Zhu J, Abbruzzese JL, Li D. (2004). Protein ion-cyclotron resonance mass spectrometry for proteome analysis. expression profiles in pancreatic adenocarcinoma compared with Rapid Commun Mass Spectrom 17: 1240–1246. normal pancreatic tissue and tissue affected by pancreatitis as Jensen O, Shevchenko A, Mann M. (1997) In: Creighton TE (ed). detected by two-dimensional gel electrophoresis and mass spectro- Protein Structure, A Practical Approach. Oxford University Press: metry. Cancer Res 64: 9018–9026. Oxford, UK, pp 48. Sherry B, Yarlett N, Strupp A, Cerami A. (1992). Identification of Kitahara K, Kawai S. (2007). Cyclosporine and for the cyclophilin as a proinflammatory secretory product of lipopolysac- treatment of rheumatoid arthritis. Curr Opin Rheumatol 19: charide-activated macrophages. Proc Natl Acad Sci USA 89: 238–245. 3511–3515. Kofron JL, Kuzmic P, Kishore V, Colon-Bonilla E, Rich DH. (1991). Song J, Lu YC, Yokoyama K, Rossi J, Chiu R. (2004). Cyclophilin A Determination of kinetic constants for peptidyl prolyl cis-trans is required for retinoic acid-induced neuronal differentiation in p19 by an improved spectrophotometric assay. Biochemistry cells. J Biol Chem 279: 24414–24419. 30: 6127–6134. Syed F, Rycyzyn MA, Westgate L, Clevenger CV. (2003). A novel and Kretzschmar AK, Dinger MC, Henze C, Brocke-Heidrich K, Horn F. functional interaction between cyclophilin A and prolactin receptor. (2004). Analysis of Stat3 (signal transducer and activator of Endocrine 20: 83–90. transcription 3) dimerization by fluorescence resonance energy Uittenbogaard A, Ying Y, Smart EJ. (1998). Characterization of a transfer in living cells. Biochem J 377: 289–297. cytosolic heat-shock protein-caveolin chaperone complex. Involve- Leverson JD, Ness SA. (1998). Point mutations in v-Myb disrupt a ment in cholesterol trafficking. J Biol Chem 273: 6525–6532. cyclophilin-catalyzed negative regulatory mechanism. Mol Cell 1: Wang P, Heitman J. (2005). The cyclophilins. Genome Biol 6: 226. 203–211. Watashi K, Ishii N, Hijikata M, Inoue D, Murata T, Miyanari Y et al. Lo¨ffler D, Brocke-Heidrich K, Pfeifer G, Stocsits C, Hackermu¨ller J, (2005). Cyclophilin B is a functional regulator of Kretzschmar AK et al. (2007). Interleukin-6-dependent RNA polymerase. Mol Cell 19: 111–122. survival of multiple myeloma cells involves the Stat3-mediated Wegenka UM, Buschmann J, Lu¨tticken C, Heinrich PC, Horn F. induction of microRNA-21 through a highly conserved enhancer. (1993). Acute-phase response factor, a nuclear factor binding to Blood. acute-phase response elements, is rapidly activated by interleukin-6 Lufei C, Koh TH, Uchida T, Cao X. (2007). Pin1 is required for the at the posttranslational level. Mol Cell Biol 13: 276–288. Ser727 phosphorylation-dependent Stat3 activity. Oncogene 26: Yang Y, Lu N, Zhou J, Chen ZN, Zhu P. (2008). Cyclophilin A up- 7656–7664. regulates MMP-9 expression and adhesion of monocytes/macro-

Oncogene Interaction of Stat3 with cyclophilin B K Bauer et al 2795 phages via CD147 signalling pathway in rheumatoid arthritis. Zacchi P, Gostissa M, Uchida T, Salvagno C, Avolio F, Volinia S et al. Rheumatology (Oxford) 47: 1299–1310. (2002). The Pin1 reveals a mechanism to control Yuan J, Wegenka UM, Lu¨tticken C, Buschmann J, Decker T, p53 functions after genotoxic insults. Nature 419: 853–857. Schindler C et al. (1994). The signalling pathways of interleukin-6 Zheng J, Koblinski JE, Dutson LV, Feeney YB, Clevenger CV. (2008). and gamma interferon converge by the activation of different Prolyl isomerase cyclophilin A regulation of Janus-activated kinase transcription factors which bind to common responsive DNA 2 and the progression of human breast cancer. Cancer Res 68: elements. Mol Cell Biol 14: 1657–1668. 7769–7778.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene