Oncogene (2000) 19, 2577 ± 2584 ã 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc The biology of Stat4 and Stat6

Andrea L Wurster1, Takashi Tanaka1 and Michael J Grusby*,1,2

1Department of Immunology and Infectious Diseases, Harvard School of Public Health, Boston, Massachusetts, MA 02115, USA; 2Department of Medicine, Harvard Medical School, Boston, Massachusetts, MA 02115, USA

IL-4 and IL-12 are cytokines that are important Structural characteristics of Stat6 regulators of the proliferation, di€erentiation and functional capacity of lymphocytes. STATs (signal IL-4 is a cytokine that can a€ect the proliferation, transducers and activators of transcription) are tran- viability, gene expression and di€erentiation of lym- scription factors that provide a direct link between the phocytes (for review see Paul, 1997). Early work on IL- cytokine receptors and cytokine induced gene transcrip- 4 signal transduction suggested that IL-4 signaling tion. Stat6 and Stat4 are two STAT family members would parallel the general STAT pathway described that speci®cally mediate signals that emanate from the above (Kohler and Rieber, 1993; Kotanides and Reich, IL-4 and IL-12 receptors, respectively. Recently a great 1993; Schindler et al., 1994). For example, IL-4, like deal of progress has been made in understanding the other cytokines, had been found to induce rapid and speci®c roles that Stat6 and Stat4 play in lymphocyte speci®c gene activation events within hours of cytokine function through in vitro as well as in vivo studies using stimulation. Additionally there was evidence that a Stat6 and Stat4-de®cient mice. This report will summar- latent cytoplasmic factor was tyrosine phosphorylated ize and describe the recent advances made in under- after IL-4 stimulation and subsequently redistributed standing the activation and regulation of Stat6 and Stat4 to the nucleus. In order to clone the factor, an IL-4 as well as their roles in the development of an immune responsive DNA element from the FcgRI promoter response. Oncogene (2000) 19, 2577 ± 2584. was used to purify this activity from protein extracts derived from an IL-4 treated cell line (Hou et al., Keywords: Stat4; Stat6; T helper cells 1994). The protein that bound to the site was tyrosine phosphorylated in response to IL-4 and shared homology with other known STAT proteins in its Introduction proposed DNA binding, SH3 and SH2 domains. The protein was termed by several groups STF-IL-4, IL-4 STATs (signal transducers and activators of transcrip- STAT, IL-4 NAF and ultimately Stat6 (Hou et al., tion) are members of a recently identi®ed family of 1994; Kotanides and Reich, 1993; Quelle et al., 1995; transcription factors that activate gene transcription in Schindler et al., 1994). response to a number of di€erent cytokines (for review Although much of the structure-function analysis of see Hoey and Grusby, 1999; Leonard and O'Shea, STAT proteins has been performed with Stat1 as the 1998; O'Shea, 1997). The STATs are latent cytoplasmic prototype, Stat6 and Stat4 share a number of the proteins that are promptly activated by tyrosine functional domains characteristic of STAT proteins, phosphorylation by the cytokine receptor associated such as a central DNA-binding domain, a conserved JAK (Janus) kinases after cytokine exposure. STAT SH2 domain for dimerization, an SH3 domain and a phosphorylation allows the dimerization of individual C-terminal transactivation domain (Hoey and Grusby, STAT proteins via their SH2 (src homology 2) 1999; Leonard and O'Shea, 1998; O'Shea, 1997). The domains. The resulting functional STAT dimer is then NH2-terminal 130 amino acids are conserved among capable of migrating directly to the nucleus where it STAT family members and this region is essential for can bind DNA and directly activate cytokine respon- tetramerization of dimerized STAT molecules, which sive gene transcription. To date seven di€erent STAT enables cooperative DNA binding on the promoters proteins have been described each activated by speci®c containing multiple potential STAT recognition sites cytokine/cytokine receptor combinations. (Vinkemeier et al., 1998; Xu et al., 1996). In this review, we will focus on two STAT proteins, Like other STAT proteins, Stat6 must be tyrosine Stat6 and Stat4, which are activated speci®cally by IL- phosphorylated by JAK kinases in order for dimeriza- 4 and IL-12, respectively. Both of these STATs have tion via the STAT SH2 domains and subsequent been found to be critical for the function and nuclear translocation to occur. The speci®c tyrosine development of T helper cells. We will discuss the on Stat6 that is phosphorylated has been mapped to Y- speci®c characteristics that set Stat6 and Stat4 apart 641, just C-terminal of the SH2 domain (Mikita et al., from the other STAT proteins, how Stat6 and Stat4 1996). Additionally, other residues surrounding the Y- are activated and regulated as well as the regulation of 641 were shown to be capable of in¯uencing dimer potential target genes by Stat6 and Stat4. We will also formation in vivo (Mikita et al., 1998a). Recently, a discuss the respective roles of Stat6 and Stat4 in the conditionally active form of Stat6 was generated by generation of a productive immune response. fusing Stat6 with a form of the estrogen receptor (ER) (Kamogawa et al., 1998; Kurata et al., 1999). In the presence of an estrogen analog, the Stat6-ER chimera eciently dimerized and translocated to the nucleus to *Correspondence: MJ Grusby activate transcription. This occurred in the absence of Stat4 and Stat6 AL Wurster et al 2578 IL-4 stimulation and tyrosine phosphorylation of the ob receptor transfected into COS cells (Ghilardi et al., Stat6-ER protein, suggesting that when dimerization is 1996). It is unclear at this point what role Stat6 is achieved by other means, the phospho-tyrosine is playing in these alternative pathways, and experiments dispensable and does not a€ect the subsequent events using Stat6-de®cient mice have not yet revealed a clear of nuclear translocation and transcriptional activation. biological role for these observations. Experiments using deletion mutants and GAL4 DNA binding domain fusions of Stat6 have also Transcriptional regulation by Stat6 revealed a potent transactivation domain in the C terminus of the Stat6 protein (Lu et al., 1997; Mikita et IL-4 is the primary cytokine that can activate Stat6 al., 1996; Moriggl et al., 1997). This region is rich in and additionally, IL-4 activates the expression of a prolines, similar to transactivation domains of other very speci®c gene program that regulates the viability, transcription factors. Interestingly the Stat6 activation proliferation and di€erentiation of the responding cell. domain bears very little sequence similarity to the This precise speci®city is controlled at two levels. First, activation domains of other STAT proteins. One study the STAT proteins bind via their SH2 domains to indicated that the e€ectiveness of the Stat6 activation speci®c docking sites on the corresponding activated domain itself could be increased by IL-4 stimulation, cytokine receptor (Hou et al., 1994; Mikita et al., suggesting an additional post-translational modi®ca- 1998a; Schindler et al., 1995). For example, Stat6 only tion of this region mediated by IL-4 (Moriggl et al., recognizes phosphorylated tyrosines on the cytoplasmic 1997). Other STAT proteins have been shown to be tail of IL-4 receptor a chain and does not interact with phosphorylated on speci®c serine residues and this the phospho-tyrosines on the IFNg receptor. Con- modi®cation can in¯uence their activation capabilities versely, Stat1, which is activated by IFN signaling, (Wen et al., 1995). Thus far though, no evidence exists does not recognize a phospho-peptide derived from the that Stat6 is modi®ed in this manner although the IL-4 receptor (Schindler et al., 1995). Therefore a part notion certainly deserves further attention. of the speci®c cytokine response is controlled at the Stat6 can also exist in di€erent isoforms depending level of the STAT/cytokine receptor interaction. on the tissue or cell type studied. Two naturally Additionally, while a number of the STAT proteins occurring isoforms (Stat6b and Stat6c) of human Stat6 recognize very similar DNA binding sites in the appear to arise from alternate splicing of the Stat6 promoters of cytokine responsive genes, Stat6 stands mRNA (Patel et al., 1998). The Stat6b isoform has a out as having a unique binding site preference. Most of deletion at the amino terminus of the full length Stat6 the STAT proteins recognize an IFNg Activating protein while Stat6c lacks a portion of the SH2 Sequence (GAS), originally identi®ed as a high anity domain. Both of these isoforms can attenuate various Stat1 binding site. This site consists of a palindromic IL-4 responses when overexpressed. The Stat6c isoform sequence separated by a 3 bp spacer (TTCNNNG- in particular behaves e€ectively as a dominant negative AA)(N3 site). Stat6 is also able to bind the GAS site to Stat6 by reducing Stat6 tyrosine phosphorylation, but only at a low anity (Schindler et al., 1995). Stat6 dampening IL-4 induced mitogenesis, and blocking the instead displays a much greater preference for a site upregulation of IL-4 inducible genes. Stat6c appears to with a 4 bp spacer (TTCNNNNGAA)(N4 site) function as a dominant negative by blocking the dimer (Schindler et al., 1995). None of the other STAT formation of endogenous Stat6. Another form of a proteins tested are capable of recognizing the N4 site in truncated Stat6 was recently shown to be coexpressed vitro (Mikita et al., 1996). Indeed, most of the Stat6 with native Stat6 speci®cally in mast cells (Sherman et binding sites mapped in IL-4 responsive promoters al., 1999). The origin of the truncated form (alternative have consisted of N4 sites (Delphin and Stavnezer, splicing of mRNA or proteolysis of the native protein) 1995; Kotanides and Reich, 1996). Therefore, Stat6 is is not clear, but it appears to lack the carboxy terminal able to selectively regulate IL-4 signal transduction not portion of conventional Stat6. The mast cell speci®c only by speci®cally interacting with the IL-4 receptor isoform binds avidly to the Stat6 DNA binding site but but also by recognizing a unique subset of STAT its role functionally is not clear at this time. Based on binding sites. these observations it appears that an additional level of In order to understand mechanistically how Stat6 Stat6 regulation could be asserted by the presence or induces gene expression in response to IL-4, a great absence of these functionally distinct isoforms ex- deal of e€ort has been focused on the analysis of the pressed in various tissues. promoters of IL-4 inducible genes. The best character- Although IL-4, and the similar cytokine IL-13, is the ized thus far is the promoter for the IL-4 induced Ig primary activator of Stat6, there are a few reports of germline transcript for the Ce gene. In the Ce other stimuli that can result in Stat6 activation. In promoter, an IL-4 response region (IL4RR) had been primary B lymphocytes both Ig and CD40 crosslinking mapped by transient transfection and reporter assays can induce Stat6 tyrosine phosphorylation (Karras et (Delphin and Stavnezer, 1995). Within the IL4RR is an al., 1996, 1997). In the case of Ig crosslinking, there N4 Stat6 site and this site is required for the IL-4 appears to be an additional dependence on PKC to responsiveness of the promoter. Interestingly, unlike mediate this signal (Karras et al., 1996). In rat cardiac other STAT proteins, Stat6 is unable to transactivate myocytes, angiotensin II was shown to be able to this site when it is removed from its promoter context, induce Stat6 phosphorylation and nuclear translocation even though Stat6 is still capable of binding the site in (Mascareno et al., 1998). PDGF was also able to vitro (Mikita et al., 1996). Directly adjacent to the Ce activate Stat6 in NIH3T3 cells and the addition of IL-4 Stat6 site is a binding site for a C/EBP transcription could potentiate this e€ect (Patel et al., 1996). factor (Delphin & Stavnezer, 1995). C/EBP consists of Additionally, leptin was able to activate several STAT a family of nuclear factors that are constitutively proteins, including Stat6, through the long form of the expressed and not induced by IL-4. When the Stat6 site

Oncogene Stat4 and Stat6 AL Wurster et al 2579 is now combined with the C/EBP site, IL-4 induced Zurawski et al., 1995). Unlike IL-4, the IL-4Ra chain is reporter gene activity is restored (Mikita et al., 1996). unable to recognize IL-13 by itself, but when combined This functional synergy between Stat6 and C/EBP is with the IL-13Ra chain the complex binds IL-13 with a dependent on the transactivation domains of both high anity (Aman et al., 1996; Hilton et al., 1996). factors although no evidence exists that Stat6 and C/ The IL-4Ra chain is tyrosine phosphorylated on ®ve EBP directly interact (Mikita et al., 1998b). Addition- tyrosine residues, providing potential binding sites for ally, the presence of C/EBP on the promoter fragment SH2 and PTB domain proteins, and a great deal of appears to be able to reduce the dissociation rate of e€ort has been expended to elucidate the importance of Stat6 bound to the IL4RR and this may help explain these sites in IL-4 signaling (Smerz-Bertling and the functional synergy observed in the reporter assays Duschl, 1995). Receptor mutagenesis and cell transfec- (Mikita et al., 1998b). tion studies have suggested that the tyrosines can be The Ce promoter is also activated in B cells that are functionally subdivided into two categories: a gene stimulated through CD40, a surface molecule on B activation domain and a growth domain (Keegan et cells that receives signals of T cell help. CD40 induces al., 1994; Ryan et al., 1996). The amino acids 437 ± 557 the activation of NFkB family members and NFkB that contain one of the phosphotyrosines appear to be sites were also mapped within the IL4RR (Berberich et important for providing a growth signal to transfected al., 1994; Delphin and Stavnezer, 1995). Functional cells after IL-4 stimulation, and amino acids 575 ± 657 studies indicated that optimal IL-4 induction of the Ce which include three of the phosphotyrosines transmit promoter required intact Stat6 and NFkB sites and signals that stimulate IL-4 induced gene expression. these sites also activated transcription synergistically Additionally, it was demonstrated that Stat6 activation (Delphin and Stavnezer, 1995; Iciek et al., 1997; correlated with the phosphorylated tyrosines in the Messner et al., 1997). Unlike the situation with C/ gene regulation domain while the phosphorylation and EBP, it has been recently demonstrated by coimmu- activation of the large adaptor molecule, IRS-2, noprecipitation and GST pulldown studies that NFkB corresponded to the growth domain (Keegan et al., family members and Stat6 can physically interact and 1994). From these studies it was concluded that signals bind cooperatively to the IL4RR (Shen and Stavnezer, generated through Stat6 would activate a gene 1998). The presence of NFkB proteins appears to program induced by IL-4 while IRS-2 would mediate increase the anity of Stat6 to the IL4RR which could the mitogenic signal through the activation of various in part explain the functional synergy. Moreover, the downstream regulators such as PI3 kinase. p50 NFkB family member, which does not have its own transactivation domain, was able to augment the Stat6-deficient mice transactivation potential of Stat6 when the two were complexed together on the binding site (Shen and Studies of mice made de®cient for Stat6 by homo- Stavnezer, 1998). Functional NFkB binding sites have logous recombination in ES cells supported the notion also been mapped near Stat6 sites in the promoters of that the Stat6 pathway provided the gene activation other IL-4 inducible genes including CD23 and the signal in response to IL-4 stimulation. Indeed, human Sg3 gene (Richards and Katz, 1994; Scha€er et lymphocytes from Stat6-de®cient mice were unable to al., 1999). From the studies on the Ce promoter it can upregulate MHC Class II, CD23, and IL-4Ra expres- be concluded that Stat6 has an acute dependence on sion in response to IL-4, supporting the importance of the presence of other transcription factors in an IL-4 Stat6 in the activation of IL-4 inducible genes (Kaplan responsive promoter. Only with the proper combina- et al., 1996a; Shimoda et al., 1996; Takeda et al., tion of these factors does IL-4 induced transcription 1996b). Stat6 de®cient B lymphocytes were also unable occur in a physiological manner. to switch to the immunoglobulin isotypes IgG1 and IgE in response to IL-4, a ®nding that correlated with an inability to transcribe the corresponding germline Stat6 and the IL-4 receptor transcripts for these genes (Linehan et al., 1998). Stat6 is activated primarily via the ligation of the IL4 Changes in gene expression in B cells and macrophages receptor (IL-4R) (for review see Nelms et al., 1999). after IL-13 stimulation were also demonstrated to be The IL-4R is composed of two chains, the 140 kD a defective in Stat6-de®cient mice supporting the notion chain (IL-4Ra) which has high anity for IL-4 and the that Stat6 activation is also an important downstream common g chain which is shared among a number of e€ector of IL-13 signaling (Takeda et al., 1996a). Stat6- hematopoietic cytokine receptors. The homodimeriza- de®cient T helper cells were also unable to di€erentiate tion of IL-4Ra itself is capable of leading to Stat6 into Th2 cells in vitro or in vivo, an observation that activation in cultured cells, but the most likely will be discussed further below (Kaplan et al., 1996a; physiological stimulus is through a combination of Shimoda et al., 1996; Takeda et al., 1996b). the a and g chains. The JAK kinases, Jak 1 and 3, are able to bind the a and g chains respectively and are Stat6 and cellular proliferation responsible for the subsequent phosphorylation of the IL-4 receptor itself as well as of Stat6. One surprising ®nding that came out of the studies on IL-13, a cytokine secreted by activated T cells, is also the Stat6-de®cient mice was that Stat6-de®cient capable of activating Stat6 under some circumstances lymphocytes did not proliferate to IL-4 as well as and shares many of the biological properties of IL-4 wildtype controls (Kaplan et al., 1996a; Shimoda et al., (Kohler et al., 1994; Zurawski and de Vries, 1994). The 1996; Takeda et al., 1996b). This is in spite of the fact overlapping functional and biochemical activities of IL- that IRS-2 was shown to be activated normally by IL-4 4 and IL-13 can be explained by the fact that their in the Stat6-de®cient lymphocytes (Kaplan et al., receptors share the IL-4Ra chain (Lin et al., 1995; 1998a). This observation is in con¯ict with the in vitro

Oncogene Stat4 and Stat6 AL Wurster et al 2580 work discussed above that suggested that Stat6 did not IFNg has also been shown to be a negative regulator play a role in proliferation but was only involved in of Stat6 dependent transcription of target genes. gene activation. The proliferation defect in the Stat6- Recently the IFNg dependent inhibition was shown de®cient cells was determined to be due to a block in to require Stat1 activation (Venkataraman et al., 1999). the progression from the G1 to S phase of the cell cycle IFNg treatment reduced Stat6 phosphorylation, DNA (Kaplan et al., 1998a). The disregulation of the cdk binding and transactivation after IL-4 stimulation. The inhibitor, p27kip, after IL-4 stimulation in Stat6- inhibitory e€ects of IFNg were not apparent immedi- de®cient cells has been implicated in the cell cycle ately but required several hours of IFNg treatment block (Kaplan et al., 1998a). Under normal circum- suggesting a requirement for de novo protein synthesis stances p27kip protein levels are decreased in IL-4 of a Stat1 target gene (Venkataraman et al., 1999). stimulated cells, relieving the inhibition on the cell SOCS1 expression, in particular, is also induced by cycle dependent kinases. In the absence of Stat6, IFNg and its induction after pretreatment has been p27kip levels remain high for an extended period of proposed to explain the negative regulation of IFNg on time after IL-4 stimulation but are reduced normally in IL-4 signal transduction. response to other cytokines, such as IL-2, which Bcl-6 is another negative regulator of Stat6 activity. activates Stat5. This defect in p27kip expression does Bcl-6 is a transcriptional repressor normally expressed not appear to be due to a direct transcriptional e€ect in germinal center B cells and some T cells (Cattoretti of Stat6 on the p27kip gene since p27kip message levels et al., 1995; Seyfert et al., 1996). Although Bcl-6 does are normal in Stat6-de®cient lymphocytes. Instead, not resemble Stat6 structurally, Bcl-6 is capable of Stat6 is probably activating another regulator(s) that recognizing the Stat6-speci®c DNA binding site (Dent modulates the expression levels of cell cycle regulators. et al., 1997). When Bcl-6 is overexpressed it can block The notion that Stat6 may be an important regulator the ability of Stat6 to transactivate the CD23 promoter of mitogenesis when cells respond normally to IL-4 by presumably competing for the Stat6 binding site leads to the question of whether or not inappropriate (Dent et al., 1997). These in vitro observations are activation of Stat6 could also be involved in uncon- supported by the in vivo observations made in Bcl-6- trolled cell growth in an oncogenic state. There are a de®cient mice (Dent et al., 1997). Bcl-6-de®cient mice few examples of constitutive Stat6 activation in are characterized by a Th2 in¯ammatory disease of the transformed cell lines. For example, one of the heart and lungs. Moreover, T cells from Bcl-6-de®cient products of the Philadelphia chromosome transloca- mice predominantly express Th2 cytokines. These tion, p190BCR/ABL, is capable of inducing prominent results suggest that in the absence of Bcl-6, Stat6 Stat6 activation in the absence of cytokine (Ilaria and may be activating transcription in an unregulated Van Etten, 1996). Additionally, HTLV-1 transformed manner resulting in an overactive Th2 response. T cells also constitutively activate the JAK/Stat6 Therefore, Bcl-6 normally may play an important role pathway (Takemoto et al., 1997). These are of course in regulating Stat6 activity and modulating a Th2 only correlative observations, and a direct role for immune response. Interestingly, in mice made de®cient Stat6 in cellular transformation has not yet been for both Stat6 and Bcl-6, the Th2 in¯ammatory disease demonstrated. and cytokine production still develop revealing an additional Stat6-independent pathway for Th2 di€er- entiation in vivo (Dent et al., 1998). Negative regulation of Stat6 activity Stat4 A great deal has been discovered about Stat6 activation in response to IL-4, but now e€orts are Stat4 was ®rst isolated by two groups using degen- underway to understand how the activity of Stat6 is erative-PCR or low-stringency hybridization, both of modulated or turned o€. A recently identi®ed family of which were based on homology with the SH2 domain proteins, suppressor of cytokine signaling (SOCS or of other STAT proteins (Yamamoto et al., 1994; SSI or JAB), is believed to regulate the JAK/STAT Zhong et al., 1994). To date, most studies of Stat4 pathway by interfering with the function of the JAK have been focused on the functional analysis of this kinases (Endo et al., 1997; Naka et al., 1997; Starr et molecule in T cells, because Stat4 is mainly phos- al., 1997). Overexpression of one family member, phorylated by IL-12-mediated signaling pathway in T SOCS1, in a murine B cell line resulted in decreased cells. Here we will ®rst describe the signal transduction Jak1 phosphorylation, decreased Stat6 phosphorylation pathway of IL-12 brie¯y, and then discuss the role of and inhibition of IL-4 induced CD23 expression Stat4 in the functional di€erentiation and proliferation (Losman et al., 1999). SOCS mRNA is induced by of T cells. the cytokines themselves and SOCS1 expression in Stat4 expression is restricted in myeloid cells, thymus particular can be induced by IL-4 in some cell types and testis (Zhong et al., 1994). In human T cells, Stat4 (Naka et al., 1997; Starr et al., 1997). This leads to a expression is dramatically induced by activation with model where IL-4 signaling via Stat6 initially occurs PHA (phytohemaglutinin), although its expression is unopposed but is subsequently dampened by a negative very low in resting T cells (Bacon et al., 1995). IL-12 is feedback mechanism through the IL-4/Stat6 dependent the major cytokine that can activate Stat4, resulting in induction of SOCS1 expression. This is supported by its tyrosine phosphorylation, in both human and the observation that Stat6 phosphorylation is pro- mouse (Leonard and O'Shea, 1998). The IL-12 receptor longed in SOCS1-de®cient lymphocytes (Naka et al., is composed of two chains, termed IL-12Rb1 and IL- 1998; Starr et al., 1998). SOCS proteins therefore may 12Rb2, and ligand binding results in heterodimer be important regulators for the termination of an formation and activation of the receptor associated immune or cytokine response. JAK kinases, Jak2 and Tyk2 (Sinigaglia et al., 1999).

Oncogene Stat4 and Stat6 AL Wurster et al 2581 Stat4 is phosphorylated by these tyrosine kinases, (Kopf et al., 1993; Kuhn et al., 1991; Magram et al., homodimerizes via its SH2 domain, and translocates 1996; Noben-Trauth et al., 1997; Wu et al., 1997). into nucleus where it can recognize traditional N3 STAT target sequences in IL-12 responsive genes Stat4 and T helper cell differentiation (Bacon et al., 1995; Cho et al., 1996; Jacobson et al., 1995). Although Stat4 is expressed both in Th1 and Th2 cells, Although IL-12 appears to be the predominant Stat4 can only be phosphorylated by IL-12 in Th1 cells activator of Stat4, Stat4 can also be phosphorylated due to the downregulation of IL-12Rb speci®cally in in response to IFN-g stimulation through activation of Th2 cells (Szabo et al., 1997). This suggested that Stat4 Jak1 and Tyk2, but this has only been observed in plays an important role in Th1 cell function or human cells (Cho et al., 1996). Additionally, recent development. In fact, mice lacking Stat4 clearly studies in human vascular smooth muscle and demonstrated that Stat4 is necessary for the generation endothelial cells demonstrated that the urokinase-type of Th1 cells (Kaplan et al., 1996b; Thierfelder et al., plasminogen activator (uPA) receptor is associated 1996). Lymphocytes from Stat4-de®cient mice were with Jak1 and Tyk2, and uPA stimulation of these cells unable to make high levels of IFNg, characteristic of resulted in Stat4 phosphorylation (Dumler et al., 1999). Th1 cells, after IL-12 priming. Additionally, NK cell IL-17 can also activate Stat4 in human monocytic cytolytic functions were severely impaired in Stat4 leukemia cell lines (Subramaniam et al., 1999). knockout mice. Moreover, abrogation of Stat4 signal- Furthermore, IL-2 can induce Jak2 and Stat4 activa- ing in a murine diabetes model signi®cantly reduced tion in NK cells but not in T cells (Wang et al., 1999). the development of CD4+ T cell-dependent diabetes, IL-12 is capable of inducing a mitogenic response in but not CD4+ T cell-independent diabetes (Holz et al., some cell types. A role for Stat4 in transducing this 1999). These data highlight the importance of Stat4 signal was demonstrated in Stat4-de®cient mice and Th1 responses in pathogenesis of autoimmune (Kaplan et al., 1996b). Stat4-de®cient lymphocytes diseases, and provide exciting therapeutic options to were speci®cally defective in their ability to proliferate treat these diseases. in response to IL-12. Additionally, Stat-4-de®cient However, little is known about the exact mechanism lymphocytes, like Stat6-de®cient lymphocytes, were by which Stat4 activation leads to Th1 di€erentiation, unable to downregulate p27kip protein levels after including the target genes of Stat4. Recently, one of the cytokine stimulation (Kaplan et al., 1998a). These data potential target genes of Stat4, termed ERM, was suggest that STAT proteins, as a general mechanism, identi®ed (Ouyang et al., 1999). ERM is a transcription may control cytokine-mediated cell proliferation by factor that belongs to the Ets family of transcription regulating the expression of cell cycle inhibitors which factors. The expression of ERM is speci®cally induced are critically involved in cell cycle progression. by IL-12 in wild-type Th1 cells, but not in Stat4- de®cient T cells. However, the role of this factor in Th1 di€erentiation is still unknown. The IFN-inducible T helper cell differentiation factor-1 (IRF-1) is also considered to be a Stat4 target CD4+ T helper lymphocytes di€erentiate into two gene, since Stat4 can bind to and transactivate the distinct subsets, de®ned by their unique cytokine IRF-1 promoter (Galon et al., 1999). Interestingly pro®les and distinct functional abilities. T helper 1 IRF-1-de®cient mice were defective in their Th1 (Th1) cells produce IL-2 and IFNg, whereas Th2 cells responses. It is unclear at this point whether or not secrete IL-4, IL-5, IL-6 and IL-13 (for review see the defect is intrinsic to the T cells though, since Abbas et al., 1996; O'Garra, 1998). Th1 cells are antigen presenting cells from IRF-1-de®cient mice do responsible for cell-mediated/in¯ammatory immunity not function normally as well (Loho€ et al., 1997; Taki and can enhance defenses against infectious agents and et al., 1997). cancer, while Th2 cells are essential for humoral Stat4 is expressed in both CD4+ T cells and CD8+ T immunity and the clearance of parasitic antigens. cells, both of which can produce IFN-g, although these Tumor immunity is often mediated by cytotoxic T cells have distinct functions. A recent report showed lymphocytes whose activation is supported by Th1 that the requirement for Stat4 is di€erent in these cells. cytokines. However, the unbalanced Th1/Th2 response Only CD4+ cells require Stat4 for IFN-g production can result in pathological conditions. For example, a after the cells are activated through the T cell receptor, polarized Th1 response can cause in¯ammatory or while Stat4 is required in both lineages for IL-12- autoimmune diseases such as fulminant hepatitis, mediated IFN-g production (Carter and Murphy, insulin-dependent diabetes mellitus, ulcerative colitis 1999). and graft-versus host diseases. In contrast, an exaggerated Th2 response can lead to allergic diseases Stat6 and T helper cell differentiation such as atopic dermatitis and bronchial asthma (Andre et al., 1996; de Pater-Huijsen et al., 1997; Kapsenberg The IL-4 dependent aspect of Th2 di€erentiation also et al., 1991; Kuchroo et al., 1995; Racke et al., 1994). appears to require the activation of Stat6 (Kaplan et The most potent factors that can promote Th1 and al., 1996a; Shimoda et al., 1996; Takeda et al., 1996b). Th2 di€erentiation are the cytokines IL-12 and IL-4 Stat6-de®cient T lymphocytes are unable to di€erenti- respectively (Abbas et al., 1996; O'Garra, 1998). This ate into IL-4 producing Th2 e€ector cells when has been clearly demonstrated in mice de®cient for the cultured with antigen and IL-4 in vitro or when cytokines themselves or their receptor subunits. For stimulated by a Th2 inducing antigen (anti-IgD) in example, mice de®cient for IL-12 or IL-12Rb1 chain vivo. Additionally, Stat6-de®cient animals are unable to are unable to generate Th1 cells, and mice lacking IL-4 mount an immune response to helminthic parasites and or the IL-4Ra chain are de®cient for Th2 responses therefore are unable to clear the parasitic infections

Oncogene Stat4 and Stat6 AL Wurster et al 2582 (Kaplan et al., 1998b). IL-4 signaling and Stat6 appear cing cells was similar between the double-de®cient and to play an important role in the development of an the wildtype Th1 cultures but the amount of IFNg allergic asthma response as well. In a murine asthma produced on a per cell basis was reduced in the cells model, Stat6-de®cient mice did not develop airway that lacked both Stat4 and Stat6. These results suggest hyperresponsiveness after allergen sensitization like that Stat4 is not absolutely required for the generation their wildtype littermates and were protected from of IFNg producing Th1 cells but is instead necessary allergic asthma (Akimoto et al., 1998; Kuperman et al., for the upregulation of IFNg production from Th1 1998). Clearly, the ability to modulate the activation of cells. This Stat4 independent pathway for Th1 cell the Stat6 pathway by pharmacological means could be development is revealed only in the absence of any e€ective for directing the immune response from Stat6 and Stat4 signaling. pathological to protective. The molecular mechanisms of how Stat6 promotes Th2 di€erentiation are unclear as of now. Several Conclusions and future directions developmentally regulated transcription factors, such as GATA-3 and c-maf, have been shown to be critical The analysis of mice de®cient for Stat4 and Stat6 has for the generation of IL-4 producing Th2 cells (Ho et provided a great deal of insight into the roles of these al., 1996; Kim et al., 1999; Zheng and Flavell, 1997). proteins in the regulation of an immune response The expression of these transcription factors is induced (Figure 1). Still, a number of important issues remain as Th precursor cells become Th2 e€ectors. Interest- to be resolved. Of primary importance is the identi®ca- ingly, the induction of GATA-3 expression in Th2 cells tion of transcriptional targets for Stat4 and Stat6 in has been shown to be dependent on IL-4 stimulated lymphocytes. The characterization of target genes Stat6 activation, although it is unclear whether or not would hopefully lead to a mechanistic understanding Stat6 activates GATA-3 transcription directly (Ouyang of how Stat4 and Stat6 a€ect cellular proliferation and et al., 1998). Ectopic expression of activated Stat6 in di€erentiation events. Additionally, it is also clear that Th1 cells can also result in the expression of GATA-3 the transcriptional potency of STATs is in¯uenced a and c-maf in these cells (Kurata et al., 1999). great deal by their association with other proteins, Additionally, a recent report demonstrated that large either through direct interaction on a promoter (Stat6 scale chromatin remodeling of the IL-4 gene occurs as and NFkB) or through post-translational modi®cation cells di€erentiate into Th2 e€ectors (Agarwal and Rao, (serine phosphorylation of Stat1). Future experiments 1998). This process was also shown to be Stat6 aimed at the identi®cation and characterization of dependent. Once again, it is unclear whether or not Stat4 and Stat6 interacting partners could potentially Stat6 is directly responsible for the chromatin remodel- reveal new mechanisms for STAT regulation. Regard- ing or instead is activating a subset of genes, including less, the critical and exquisitely speci®c roles of Stat4 downstream e€ectors like GATA-3 and c-maf, that and Stat6 in T helper cell di€erentiation suggest that could be responsible for the chromatin changes at the these proteins would be ideal therapeutic targets for IL-4 locus. The identi®cation and characterization of modulating an immune response. Stat6 target genes in di€erentiating Th2 cells would greatly aid in answering these questions.

Th1 cell differentiation in the absence of Stat4 and Stat6 It had been clearly demonstrated in the Stat4 and Stat6-de®cient mice that these transcription factors play important roles in Th1 and Th2 di€erentiation, respectively. Interestingly, mice made de®cient for both Stat4 and Stat6 (double-de®cient) revealed a STAT- independent pathway for Th1 development (Kaplan et al., 1998c). Lymphocytes from double-de®cient mice were unable to generate IL-4 producing Th2 cells like their Stat6-de®cient littermates, supporting the notion that Stat6 is critical for the di€erentiation of Th2 cells. Interestingly, double-de®cient lymphocytes were cap- able of generating IFNg producing Th1 cells, unlike the Stat4-de®cient lymphocytes. The double-de®cient lym- phocytes produce signi®cantly more IFNg than the Stat4-de®cient lymphocytes although the level is still approximately fourfold less than IL-12 stimulated wildtype Th1 cells. Using ELISPOT analysis, it was determined that the absolute number of IFNg produ- Figure 1 Biological consequences of Stat6 and Stat4 activation

References

Abbas AK, Murphy KM and Sher A. (1996). Nature, 383, Akimoto T, Numata F, Tamura M, Takata Y, Higashida N, 787 ± 793. Takashi T, Takeda K and Akira S. (1998). J. Exp. Med., Agarwal S and Rao A. (1998). Immunity, 9, 765 ± 775. 187, 1537 ± 1542.

Oncogene Stat4 and Stat6 AL Wurster et al 2583 AmanMJ,TayebiN,ObiriNI,PuriRK,ModiWSand Karras JG, Wang Z, Huo L, Frank DA and Rothstein TL. Leonard WJ. (1996). J. Biol. Chem., 271, 29265 ± 29270. (1997). J. Immunol., 159, 4350 ± 4355. AndreI,GonzalezA,WangB,KatzJ,BenoistCandMathis Keegan AD, Nelms K, White M, Wang LM, Pierce JH and D. (1996). Proc. Natl. Acad. Sci. USA, 93, 2260 ± 2263. Paul WE. (1994). Cell, 76, 811 ± 820. Bacon CM, Petricoin III EF, Ortaldo JR, Rees RC, Larner Kim JJ, Ho IC, Grusby MJ and Glimcher LH. (1999). AC, Johnston JA and O'Shea JJ. (1995). Proc. Natl. Acad. Immunity, 10, 745 ± 751. Sci. USA, 92, 7307 ± 7311. Kohler I, Alliger P, Minty A, Caput D, Ferrara P, Holl- Berberich I, Shu GL and Clark EA. (1994). J. Immunol., 153, Neugebauer B, Rank G and Rieber EP. (1994). FEBS 4357 ± 4366. Lett., 345, 187 ± 192. Carter LL and Murphy KM. (1999). J. Exp. Med., 189, Kohler I and Rieber EP. (1993). Eur. J. Immunol., 23, 3066 ± 1355 ± 1360. 3071. Cattoretti G, Chang CC, Cechova K, Zhang J, Ye BH, Falini Kopf M, Le Gros G, Bachmann M, Lamers MC, Blue- B, Louie DC, Ot K, Chaganti RS and Dalla-Favera R. thmann H and Kohler G. (1993). Nature, 362, 245 ± 248. (1995). Blood, 86, 45 ± 53. Kotanides H and Reich NC. (1993). Science, 262, 1265 ± Cho SS, Bacon CM, Sudarshan C, Rees RC, Finbloom D, 1267. Pine R and O'Shea JJ. (1996). J. Immunol., 157, 4781 ± Kotanides H and Reich NC. (1996). J. Biol. Chem., 271, 4789. 25555 ± 25561. de Pater-Huijsen FL, Pompen M, Jansen HM and Out TA. Kuchroo VK, Das MP, Brown JA, Ranger AM, Zamvil SS, (1997). Immunol. Lett., 57, 47 ± 51. SobelRA,WeinerHL,NabaviNandGlimcherLH. Delphin S and Stavnezer J. (1995). J. Exp. Med., 181, 181 ± (1995). Cell, 80, 707 ± 718. 192. Kuhn R, Rajewsky K and Muller W. (1991). Science, 254, Dent AL, Hu-Li J, Paul WE and Staudt LM. (1998). Proc. 707 ± 710. Natl. Acad. Sci. USA, 95, 13823 ± 13828. Kuperman D, Scho®eld B, Wills-Karp M and Grusby MJ. Dent AL, Sha€er AL, Yu X, Allman D and Staudt LM. (1998). J. Exp. Med., 187, 939 ± 948. (1997). Science, 276, 589 ± 592. Kurata H, Lee H, O'Garra A and Arai N. (1999). Immunity, Dumler I, Kopmann A, Wagner K, Mayboroda OA, Jerke 11, 677 ± 688. U, Dietz R, Haller H and Gulba DC. (1999). J. Biol. Leonard WJ and O'Shea JJ. (1998). Annu. Rev. Immunol., 16, Chem., 274, 24059 ± 24065. 293 ± 322. Endo TA, Masuhara M, Yokouchi M, Suzuki R, Sakamoto Lin JX, Migone TS, Tsang M, Friedmann M, Weatherbee H, Mitsui K, Matsumoto A, Tanimura S, Ohtsubo M, JA, Zhou L, Yamauchi A, Bloom ET, Mietz J and John S. Misawa H, Miyazaki T, Leonor N, Taniguchi T, Fujita T, (1995). Immunity, 2, 331 ± 339. Kanakura Y, Komiya S and Yoshimura A. (1997). Nature, Linehan LA, Warren WD, Thompson PA, Grusby MJ and 387, 921 ± 924. Berton MT. (1998). J. Immunol., 161, 302 ± 310. GalonJ,SudarshanC,ItoS,FinbloomDandO'SheaJJ. Loho€ M, Ferrick D, Mittrucker HW, Duncan GS, Bischof (1999). J. Immunol., 162, 7256 ± 7262. S, Rollingho€ M and Mak TW. (1997). Immunity, 6, 681 ± Ghilardi N, Ziegler S, Wiestner A, Sto€el R, Heim MH and 689. Skoda RC. (1996). Proc. Natl. Acad. Sci. USA, 93, 6231 ± Losman JA, Chen XP, Hilton D and Rothman P. (1999). J. 6235. Immunol., 162, 3770 ± 3774. Hilton DJ, Zhang JG, Metcalf D, Alexander WS, Nicola NA Lu B, Reichel M, Fisher DA, Smith JF and Rothman P. and Willson TA. (1996). Proc. Natl. Acad. Sci. USA, 93, (1997). J. Immunol., 159, 1255 ± 1264. 497 ± 501. Magram J, Connaughton SE, Warrier RR, Carvajal DM, Ho IC, Hodge MR, Rooney JW and Glimcher LH. (1996). Wu CY, Ferrante J, Stewart C, Sarmiento U, Faherty DA Cell, 85, 973 ± 983. and Gately MK. (1996). Immunity, 4, 471 ± 481. Hoey T and Grusby MJ. (1999). Adv. Immunol., 71, 145 ± Mascareno E, Dhar M and Siddiqui MA. (1998). Proc. Natl. 162. Acad.Sci.USA,95, 5590 ± 5594. Holz A, Bot A, Coon B, Wolfe T, Grusby MJ and von Messner B, Stutz AM, Albrecht B, Peiritsch S and Herrath MG. (1999). J. Immunol., 163, 5374 ± 5382. Woisetschlager M. (1997). J. Immunol., 159, 3330 ± 3337. Hou J, Schindler U, Henzel WJ, Ho TC, Brasseur M and Mikita T, Campbell D, Wu P, Williamson K and Schindler McKnight SL. (1994). Science, 265, 1701 ± 1706. U. (1996). Mol. Cell Biol., 16, 5811 ± 5820. Iciek LA, Delphin SA and Stavnezer J. (1997). J. Immunol., Mikita T, Daniel C, Wu P and Schindler U. (1998a). J. Biol. 158, 4769 ± 4779. Chem., 273, 17634 ± 17642. Ilaria Jr RL and Van Etten RA.(1996). J. Biol. Chem., 271, Mikita T, Kurama M and Schindler U. (1998b). J. Immunol., 31704 ± 31710. 161, 1822 ± 1828. Jacobson NG, Szabo SJ, Weber-Nordt RM, Zhong Z, Moriggl R, Berchtold S, Friedrich K, Standke GJ, Kammer Schreiber RD, Darnell Jr JE and Murphy KM. (1995). J. W, Heim M, Wissler M, Stocklin E, Gouilleux F and Exp. Med., 181, 1755 ± 1762. Groner B. (1997). Mol. Cell. Biol., 17, 3663 ± 3678. Kamogawa Y, Lee HJ, Johnston JA, McMahon M, O'Garra Naka T, Matsumoto T, Narazaki M, Fujimoto M, Morita Y, A and Arai N. (1998). J. Immunol., 161, 1074 ± 1077. Ohsawa Y, Saito H, Nagasawa T, Uchiyama Y and Kaplan MH, Daniel C, Schindler U and Grusby MJ. (1998a). Kishimoto T. (1998). Proc. Natl. Acad. Sci. USA, 95, Mol. Cell. Biol., 18, 1996 ± 2003. 15577 ± 15582. Kaplan MH, Schindler U, Smiley ST and Grusby MJ. Naka T, Narazaki M, Hirata M, Matsumoto T, Minamoto S, (1996a). Immunity, 4, 313 ± 319. Aono A, Nishimoto N, Kajita T, Taga T, Yoshizaki K, Kaplan MH, Sun YL, Hoey T and Grusby MJ. (1996b). Akira S and Kishimoto T. (1997). Nature, 387, 924 ± 929. Nature, 382, 174 ± 177. Nelms K, Keegan AD, Zamorano J, Ryan JJ and Paul WE. Kaplan MH, Whit®eld JR, Boros DL and Grusby MJ. (1999). Annu. Rev. Immunol., 17, 701 ± 738. (1998b). J. Immunol., 160, 1850 ± 1856. Noben-Trauth N, Shultz LD, Brombacher F, Urban Jr JF, Kaplan MH, Wurster AL and Grusby MJ. (1998c). J. Exp. Gu H and Paul WE. (1997). Proc. Natl. Acad. Sci. USA, Med., 188, 1191 ± 1196. 94, 10838 ± 10843. Kapsenberg ML, Wierenga EA, Bos JD and Jansen HM. O'Garra A. (1998). Immunity, 8, 275 ± 283. (1991). Immunol Today, 12, 392 ± 395. O'Shea JJ. (1997). Immunity, 7, 1±11. Karras JG, Wang Z, Coniglio SJ, Frank DA and Rothstein TL. (1996). J. Immunol., 157, 39 ± 47.

Oncogene Stat4 and Stat6 AL Wurster et al 2584 Ouyang W, Jacobson NG, Bhattacharya D, Gorham JD, Starr R, Willson TA, Viney EM, Murray LJ, Rayner JR, Fenoglio D, Sha WC, Murphy TL and Murphy KM. Jenkins BJ, Gonda TJ, Alexander WS, Metcalf D, Nicola (1999). Proc. Natl. Acad. Sci. USA, 96, 3888 ± 3893. NA and Hilton DJ. (1997). Nature, 387, 917 ± 921. Ouyang W, Ranganath SH, Weindel K, Bhattacharya D, Subramaniam SV, Cooper RS and Adunyah SE. (1999). Murphy TL, Sha WC and Murphy KM. (1998). Immunity, Biochem. Biophys. Res. Commun., 262, 14 ± 19. 9, 745 ± 755. Szabo SJ, Dighe AS, Gubler U and Murphy KM. (1997). J. PatelBK,WangLM,LeeCC,TaylorWG,PierceJHand Exp. Med., 185, 817 ± 824. LaRochelle WJ. (1996). J. Biol. Chem., 271, 22175 ± Takeda K, Kamanaka M, Tanaka T, Kishimoto T and Akira 222182. S. (1996a). J. Immunol., 157, 3220 ± 3222. Patel BKR, Pierce JH and LaRochelle WJ. (1998). Proc. Takeda K, Tanaka T, Shi W, Matsumoto M, Minami M, Natl. Acad. Sci. USA, 95, 172 ± 177. Kashiwamura S, Nakanishi K, Yoshida N, Kishimoto T Paul WE. (1997). Ciba Found. Symp., 204, 208 ± 219. and Akira S. (1996b). Nature, 380, 627 ± 630. Quelle FW, Shimoda K, Thierfelder W, Fischer C, Kim A, Takemoto S, Mulloy JC, Cereseto A, Migone TS, Patel BK, Ruben SM, Cleveland JL, Pierce JH, Keegan AD and Matsuoka M, Yamaguchi K, Takatsuki K, Kamihira S, Nelms K. (1995). Mol. Cell Biol., 15, 3336 ± 3343. White JD, Leonard WJ, Waldmann T and Franchini G. Racke MK, Bonomo A, Scott DE, Cannella B, Levine A, (1997). Proc. Natl. Acad. Sci. USA, 94, 13897 ± 13902. Raine CS, Shevach EM and Rocken M. (1994). J. Exp. Taki S, Sato T, Ogasawara K, Fukuda T, Sato M, Hida S, Med., 180, 1961 ± 1966. Suzuki G, Mitsuyama M, Shin EH, Kogima S, Taniguchi Richards ML and Katz DH. (1994). J. Immunol., 152, 3453 ± T and Asano Y. (1997). Immunity, 6, 673 ± 679. 3466. Thierfelder WE, van Duersen JM, Yamamoto K, Tripp RA, Ryan JJ, McReynolds LJ, Keegan A, Wang LH, Garfein E, Sarawar SR, Carson RT, Sangster MY, Vignali DA, Rothman P, Nelms K and Paul WE. (1996). Immunity, 4, Doherty PC, Grosveld GC and Ihle JN. (1996). Nature, 123 ± 132. 382, 171 ± 174. Scha€er A, Cerutti A, Shah S, Zan H and Casali P. (1999). J. Venkataraman C, Leung S, Salvekar A, Mano H and Immunol., 162, 5327 ± 5336. Schindler U. (1999). J. Immunol., 162, 4053 ± 4061. Schindler C, Kashleva H, Pernis A, Pine R and Rothman P. Vinkemeier U, Moare® I, Darnell Jr JE and Kuriyan J. (1994). EMBO J., 13, 1350 ± 1356. (1998). Science, 279, 1048 ± 1052. Schindler U, Wu P, Rothe M, Brasseur M and McKnight SL. Wang KS, Ritz J and Frank DA. (1999). J. Immunol., 162, (1995). Immunity, 2, 689 ± 697. 299 ± 304. Seyfert VL, Allman D, He Y and Staudt LM. (1996). Wen Z, Zhong Z and Darnell Jr JE. (1995). Cell, 82, 241 ± Oncogene, 12, 2331 ± 2342. 250. Shen CH and Stavnezer J. (1998). Mol. Cell Biol., 18, 3395 ± Wu C, Ferrante J, Gately MK and Magram J. (1997). J. 3404. Immunol., 159, 1658 ± 1665. Sherman MA, Secor VH and Brown MA. (1999). J. Xu X, Sun YL and Hoey T. (1996). Science, 273, 794 ± 797. Immunol., 162, 2703 ± 2708. Yamamoto K, Quelle FW, Thierfelder WE, Kreider BL, Shimoda K, van Deursen J, Sangster MY, Sarawar SR, Gilbert DJ, Jenkins NA, Copeland NG, Silvennoinen O CarsonRT,TrippRA,ChuC,QuelleFW,NosakaT, and Ihle JN. (1994). Mol. Cell. Biol., 14, 4342 ± 4349. Vignali DA, Doherty PC, Grosveld G, Paul WE and Ihle Zheng W and Flavell RA. (1997). Cell, 89, 587 ± 596. JN. (1996). Nature, 380, 630 ± 633. Zhong Z, Wen Z and Darnell Jr JE. (1994). Proc. Natl. Acad. Sinigaglia F, D'Ambrosio D, Panina-Bordignon P and Sci. USA, 91, 4806 ± 4810. Rogge L. (1999). Immunol. Rev., 170, 65 ± 72. Zurawski G and de Vries JE. (1994). Immunol. Today, 15, Smerz-Bertling C and Duschl A. (1995). J. Biol. Chem., 270, 19 ± 26. 966 ± 970. ZurawskiSM,ChomaratP,DjossouO,BidaudC,McKenzie Starr R, Metcalf D, Elefanty AG, Brysha M, Willson TA, AN, Miossec P, Banchereau J and Zurawski G. (1995). J. Nicola NA, Hilton DJ and Alexander WS. (1998). Proc. Biol. Chem., 270, 13869 ± 13878. Natl. Acad. Sci. USA, 95, 14395 ± 14399.

Oncogene