Abstract
As signal transducers and transcription factors, signal transducer and activator of transcriptions (STATs) exert a critical role in cell growth, differentiation, and cell death. Their dysregulation that may lead to constitutive activation, no activation, or functional change, is associated with many different cancers. In this review, the contribution of STATs to cancer growth and progression is presented, and then the context in which interferon regulation of STAT activation provides a mechanism for therapeutic intervention for specific cancers is discussed.
Introduction
S
STAT1, STAT3, and STAT5A/B are activated in response to a number of cytokines, whereas STAT4 and STAT6 are activated primarily in response to interleukin (IL)-12 and IL-23, or IL-4, respectively (Kisseleva and others 2002; Lim and Cao 2006). All STAT proteins are activated in response to type I and type III interferons (IFNs), and STAT1 is activated by IFN-γ (Kisseleva and others 2002; Platanias 2005; Lim and Cao 2006). In response to type I and III IFNs, STAT1–STAT2 heterodimers associate with IFN regulatory factor 9 to form IFN-stimulated gene factor 3, which recognizes IFN-stimulated response element motifs, inducing IFN-stimulated gene (ISG) expression (Platanias 2005; Maher and others 2008). The activation of STAT proteins is regulated by suppressor of cytokine signaling (SOCS) protein, protein inhibitor of activated STATs (PIAS), tyrosine phosphatases such as SHP-2, and by ubiquitination (You and others 1999; Wormald and Hilton 2004; Tanaka and others 2005; Lim and Cao 2006). Aberrant STAT protein expression and activity are directly associated with cancer and disease progression (Buettner and other 2002). In this review, we highlight the complex role of STAT proteins in oncogenesis and IFN-based cancer therapies.
STAT Proteins in Cancer
Knockout studies have shown that STAT1 is important for growth inhibition and apoptosis—STAT1-null mice are more susceptible to spontaneous tumor development (Chan and others 2012)—while STAT3 promotes cell survival and is crucial for early stages of development—STAT3 knockout is embryonic lethal in mice (Takeda and others 1997; Akira 1999), and constitutive STAT3 activation is a signature of malignant transformation (Schlessinger and Levy 2005). Similar to STAT3, constitutive STAT5 activation also contributes to cell proliferation and tumorigenesis (Calò and others 2003). Meanwhile, STAT2 knockout mice are less susceptible to tumor development in the presence of chemical carcinogens (Gamero and others 2010). However, the role of each STAT protein in oncogenesis and disease progression continues to be complex due to tissue specificity (Akira 1999; Akira 2000; Schindler and Plumlee 2008; Yang and Stark 2008) and the pleiotropic effects of STAT-inducible genes (Lim and Cao 2006; Schindler and Plumlee 2008).
Contributions of STAT activation to oncogenesis
Constitutive STAT activation has been linked with the oncogenesis of carcinomas, lymphomas, leukemias, and sarcomas (Turkson and Jove 2000; Punjabi and others 2007). STAT3 and STAT5A/B mediate cell survival responses to cytokine and growth factor signaling (Kisseleva and others 2002; Lim and Cao 2006), and STAT3 has also been shown to antagonize STAT1-mediated antitumor responses (Ho and Ivashkiv 2006; Yu and others 2009). Further, activated STAT3 directly downregulates the expression of p53, an inhibitor of aberrant cell growth and tumorigenesis, by binding to the p53 promoter (Niu and others 2005). It is not, therefore, unexpected that constitutive activation of certain STAT proteins, namely, STAT3 and STAT5A/B, are associated with oncogenesis.
STAT proteins can become constitutively activated through a variety of mechanisms. STAT3 is activated in response to IL-6 to activate transcription of cell cycle-regulating proto-oncogenes and antiapoptosis genes (Gritsko and others 2006; Lim and Cao 2006) to enhance cell survival. Mutations in the IL-6 receptor, gp130, can result in aberrant downstream signaling and STAT3 activation (Rebouissou and others 2009). Further, STAT somatic mutations, such as a single-amino-acid substitution in the STAT protein SH2 domain, can cause constitutive activation independent of the upstream receptor by inducing dimerization (Scarzello and others 2007; Pilati and others 2011; Koskela and others 2012). Dysregulation of the STAT negative regulatory factor, PIAS1, has been associated with colon cancer, contrasting with normal colon cells and benign adenomas that have high PIAS1 expression (Coppola and others 2009). STAT proteins can also be activated in response to virus infection and contribute directly to malignancies such as Epstein-Barr virus-associated nasopharyngeal cancer (Liu and others 2008), hepatitis C virus-associated hepatocellular carcinoma (Yoshida and others 2002), and human papillomavirus-associated cervical carcinoma (Shukla and others 2010).
Constitutive STAT3 activation contributes to the survival and proliferation of the lung (Song and others 2011), breast (Garcia and others 2001; Diaz and others 2006; Gritsko and others 2006), colon (Corvinus and others 2005; Lin and others 2011), prostate (Mora and others 2002; Barton and others 2004), pancreas (Lian and others 2004), kidney (Guo and others 2009), skin (Grandis and others 2000; Pedranzini and others 2004) cancers, and osteosarcoma (Fossey and others 2009; Tu and others 2012), as well as leukemias (Holtick and others 2005; Redell and others 2011; Koskela and others 2012) and lymphomas (Sommer and others 2004; Ding and others 2008). Notably, STAT3 activation has also been linked with metastasis of many of the same types of malignancies (Dauer and others 2005; Hsieh and others 2005; Xie and others 2006; Abdulghani and others 2008; Barbieri and others 2010; Chiu and others 2011; Xiong and others 2012). Immunohistochemistry analysis of human lung cancer tissues revealed an increase in STAT3 activation at the invading edge of early-stage tumors, whereas STAT3 activation in A549 lung adenocarcinomic epithelial cells is associated with an upregulation in genes regulating cell migration and angiogenesis (Dauer and others 2005). In a study of invasive breast cancer, phosphorylated STAT3 was detected in 35% of tumors and correlated with lymph node metastasis (Hsieh and others 2005). Introduction of a constitutively active STAT3 mutant in place of normal wild-type STAT3 in mice enhances oncogene-dependent mammary tumor development, invasion, and metastasis through upregulation of the adhesion molecule Cten (Barbieri and others 2010). Moreover, STAT3 activation in breast cancer and melanoma has been associated with brain metastases (Xie and others 2006; Chiu and others 2011). Constitutive STAT3 activation downregulates the expression of a lipid raft protein, Caveolin-1, in human breast cancer (Chiu and others 2011) and upregulates the expression of proangiogenic genes: vascular endothelial growth factor (VEGF), basic fibroblast growth factor, and matrix metalloproteinase-2 in human melanoma (Xie and others 2006). Similarly, STAT3 activation increases the invasiveness of colon cancer metastasis by downregulating E-cadherin expression (Xiong and others 2012). In vivo and in vitro metastasis models of prostate cancer provide further evidence for the role of STAT3 in metastases, with STAT3 activation increasing cancer cell migration and lung metastases (Abdulghani and others 2008). Indeed, the same group also reported phosphorylated STAT3 in the nucleus of 77% of regional lymph nodes and 67% of bone metastases, and 56% of metastases to distant organs in tissue samples collected from prostate cancer patients (Abdulghani and others 2008).
Among the other STAT family proteins, elevated STAT6 activation can be detected in primary human prostate cancer tissues (Das and others 2007). Similar to STAT3, STAT5 activation has also been shown to promote metastasis of prostate cancer cells (Gu and others 2010). Both STAT5 and STAT6 are significantly activated in leukemias and lymphomas due to their roles in hematopoiesis and immune cell differentiation (Bruns and Kaplan 2006; Lim and Cao 2006; Malin and others 2010). Constitutively activated STAT5 localizes to the adaptor protein Gab2 in the cytoplasm of transformed bone marrow cells, to induce cell proliferation and survival by activating phosphoinositide 3-kinase and protein kinase B (Harir and others 2007). There is also evidence to suggest that STAT1 activation can promote the development of B-cell leukemias through regulation of major histocompatibility complex (MHC) class I expression, where reduced MHC class I expression on leukemic cells in STAT1-null mice leads to natural killer (NK) cell-mediated lysis and tumor clearance (Kovacic and others 2006). STAT1 is also overexpressed in human renal cell carcinoma tissues (Zhu and others 2012).
In addition to direct prosurvival and antiapoptotic functions, STAT3 activation leads to the expression of proinflammatory cytokines and chemokines, which have been shown to be important in establishing procarcinogenic inflammatory microenvironments (Yu and others 2009). These STAT3-dependent inflammatory conditions can arise as a result of infection (Bronte-Tinkew and others 2009), stress (Landen Jr. and others 2007), and exposure to UV light (Aziz and others 2007) and carcinogenic substances such as those found in cigarettes (Arredondo and others 2006), contributing to the development of several cancer types (Yu and others 2009).
Contributions of reduced STAT activation to oncogenesis
In contrast to elevated and sometimes constitutive activation of STAT proteins, a lack of STAT phosphorylation can also contribute to malignancy. Reduced STAT1 expression has been shown to enhance breast (Widschwendter and others 2002; Klover and others 2010; Zhan and others 2011; Chan and others 2012), prostate (Shen and Lentsch 2004), head and neck (Leibowitz and others 2011) cancers, soft tissue sarcoma (Zimmerman and others 2012), and acute myeloid leukemia (AML) (Jiang and others 2011). In a mouse model of Neu oncogene-induced mammary tumorigenesis, the loss of STAT1 expression from mammary epithelium increases early tumor formation (Klover and others 2010). Recent studies have also shown that STAT1-null mice are susceptible to spontaneous estrogen receptor-α (ER-α)- and progesterone receptor-positive mammary tumors, which is reflected in a subset of ER-α-positive human breast cancer tissues that have reduced STAT1 expression (Chan and others 2012). Based on survival analysis of breast cancer patients, low STAT1 activation is associated with disease relapse, whereas elevated STAT1 activation is associated with relapse-free survival (Widschwendter and others 2002). There is also evidence that STAT1 activation in MCF-7 cancer-initiating cells can confer resistance to radiotherapy (Zhan and others 2011), reflecting the complex role of STAT proteins in oncogenesis. While STAT1 activation is comparable in both normal prostate epithelial cells and primary tumor cells, STAT1 activation is significantly decreased in bone metastases from prostate adenocarcinoma and correlates with angiogenic chemokine production, suggesting that low STAT1 activation can result in tumor metastases (Shen and Lentsch 2004). Reduced STAT1 activation in cancer is also linked with resistance of tumors to the immune response (Leibowitz and others 2011). For instance, reduced STAT1 phosphorylation in head and neck squamous cell carcinoma cells limits TAP1 expression, which is required for MHC class I presentation of tumor antigens (Chen and others 1996) and recognition by tumor antigen-specific cytotoxic T cells (Leibowitz and others 2011). Opposite to B-cell leukemia, where STAT1 activation may promote tumorigenesis (Kovacic and others 2006), STAT1 activation in AML can inhibit leukemia cell proliferation in response to retinoic acid-inducible gene I (RIG-I), an RNA helicase generally associated with recognition of the viral genomic material (Yoneyama and others 2004; Jiang and others 2011). Earlier studies indicate that RIG-I expression is upregulated during myelopoiesis, and that RIG-I null mice experience granulocytosis with an accumulation of intermediate- and late-stage granulocytes in the bone marrow, spleen, and peripheral blood (Zhang and others 2008).
Contributions of STAT isoforms and U-STAT to oncogenesis
Alternative splicing of STAT mRNAs and post-translational protease modification of STAT proteins generate full-length α-isoforms, and truncated β-, γ-, and δ-isoforms, which have variable activity (Lim and Cao 2006; Benekli and others 2009). Alternative splicing of STAT6 mRNA also produces STAT6b and STAT6c isoforms (Lim and Cao 2006). Truncated STAT3β and STAT5A/Bβ are expressed in the bone marrow of AML patients (Xia and others 2001a, 2001b), although their exact role in malignancy is unclear. In an early study of AML patients, those with constitutive bone marrow STAT3 activation had reduced survival in comparison to patients who did not have constitutive STAT3 activation, and that overall survival was further decreased in patients with constitutive STAT3β expression (Benekli and others 2002). In this study, the median survival time of patients constitutively expressing STAT3β was 12.9 months, significantly less than the 18.9 months for other patients (Benekli and others 2002). However, more recent studies using several cancer cell types suggest that the expression of truncated STAT3β can inhibit tumor growth, given that STAT3β, which lacks a serine phosphorylation site and transactivation domain, can compete with full-length STAT3α to downregulate transcriptional activation (Xu and others 2009; Zammarchi and others 2011).
Studies have revealed that nuclear U-STAT proteins are able to regulate gene expression (Yang and Stark 2008). U-STAT1 and U-STAT3 expression are increased in response to IFN-γ and IFN-β, and IL-6, respectively, a consequence of their promoter regions containing GAS motifs (Yang and others 2007; Cheon and Stark 2009). Increased U-STAT3 expression upregulates M-Ras and c-Met oncogenes and could, therefore, play a role in oncogenesis (Yang and others 2005). Moreover, accumulation of U-STAT1 correlates with reduced survival in some human cases of soft tissue sarcoma (Zimmerman and others 2012). By contrast, in other human soft tissue sarcoma specimens, high STAT1 activation correlates with increased survival, attributed to upregulation of Fas-mediated apoptosis (Zimmerman and others 2012).
Targeting STAT activation
The critical contributions of STAT proteins to oncogenesis have led to the development of several cancer therapies targeting STAT protein activation (Benekli and others 2009). These include the use of antibodies and tyrosine kinase inhibitors to block ligand binding and phosphorylation of STAT proteins, respectively (Arora and Scholar 2005; Benekli and others 2009). The monoclonal anti-CD20 antibody, Rituximab, used to treat B-cell malignancies inhibits STAT3 activation by blocking IL-10 expression and the IL-10 autocrine/paracrine loop (Alas and Bonavida 2001; Alas and others 2001). By blocking the BCR-ABL1 kinase activity, the ABL1 cytoplasmic tyrosine kinase inhibitor, imatinib mesylate, inhibits STAT3 and STAT5 phosphorylation and is approved for the treatment of chronic myeloid leukemia (CML) and metastatic gastrointestinal stromal tumors, which are associated with elevated STAT3 activation (Paner and others 2003; Nair and others 2012). Notably, enhanced STAT3 and STAT5 activation in tumor cells has been linked with resistance to imatinib mesylate treatment (Bewry and others 2008; Warsch and others 2011; Yamada and others 2011). Other strategies for regulating STAT activation include the use of demethylating agents to upregulate the expression of SOCS family proteins and PIAS in cancers (Khong and others 2008; Kluge and others 2011; Wilop and others 2011). The demethylating agent, Azacitidine, upregulates SOCS3 expression and reduces STAT3 activation to promote apoptosis in multiple myeloma cell lines, where SOCS3 expression is normally limited by methylation (Khong and others 2008; Wilop and others 2011). There are also several strategies that directly target STAT protein expression and their function as transcription factors. These include small interfering RNAs to block STAT protein expression (Lee and others 2004; Kaymaz and others 2012), small-molecule inhibitors that prevent STAT dimerization (Turkson and others 2004; Zhang and others 2010, 2012), and oligonucleotides that compete for STAT binding (Jing and others 2005; Weerasinghe and others 2007). Small-molecule inhibitors of STAT protein dimerization target the SH2 domain to prevent dimerization and subsequent nuclear translocation of phosphorylated STAT proteins, thereby increasing apoptosis of malignant cells (Turkson and others 2004; Zhang and others 2010, 2012). Most recently, the small-molecule inhibitor of STAT3, BP-1–102, has been shown to inhibit STAT3 dimerization in multiple STAT3-dependent cancer cell lines, including MDA-MB-231 (breast carcinoma), DU145 (prostate carcinoma), NIH3T3/v-Src (transformed fibroblast), Panc-1 (pancreatic carcinoma), and A549 (lung adenocarcinoma), as well as inhibiting the growth of human breast and lung tumor xenografts when given orally or intravenously to graft-recipient mice (Zhang and others 2012). In this study, loss of STAT3 activity due to BP-1–102 resulted in reduced proliferation, survival, and migration, marked by a decrease in c-Myc, B-cell lymphoma-extra large, Survivin, Cyclin D1, VEGF, and Krüppel-like factor 8, and an increase in E-cadherin (Zhang and others 2012).
Role of STAT Activation in the Antineoplastic Effects of IFNs
Even though STAT activation contributes to the pathogenesis of several cancers, there is evidence to support the contribution of IFN-inducible STAT phosphorylation to tumor suppression (Sun and others 1998; Landolfo and others 2000; Zhou and others 2001; Connett and others 2003; Kovarik and others 2005; Eriksen and others 2009; Vitale and others 2012). These studies further emphasize the complex roles of STAT proteins in oncogenesis.
An early study of carcinoid tumor biopsies collected from patients before and during IFN-α treatment provided evidence for a correlation between STAT1 and STAT2 activation and survival (Zhou and others 2001). Immunohistochemistry revealed that there was a significant upregulation in STAT1 and STAT2 expression after IFN-α treatment in patients that responded to treatment—characterized by a >50% reduction in the detectable biomarker U-5HIAA and/or tumor size—and in patients who were nonprogressers (Zhou and others 2001). However, STAT1 and STAT2 expression was not significantly upregulated in patients who experienced a progressive disease, characterized by a >25% increase in a detectable biomarker and/or tumor size (Zhou and others 2001). STAT1 expression is also a predictive factor for responsiveness of CML patients to IFN-α treatment (Landolfo and others 2000). In a study of CML cells collected from patients, cells from those who went on to have a complete clinical response to IFN-α expressed STAT1, whereas cells from each of the analyzed nonresponders did not express STAT1 (Landolfo and others 2000). Similarly, cutaneous T-cell lymphomas that are resistant to IFN-α also lack STAT1 expression (Sun and others 1998). IFN-α is also effective for the treatment of hepatocellular carcinoma in a STAT1-dependent manner (Radaeva and others 2002; Inamura and others 2005; Ikeda and others 2010). IFN-α treatment inhibits the growth of multiple hepatocellular carcinoma lines, but in STAT1 and JAK1 knockdowns, growth inhibition is reduced (Inamura and others 2005). Likewise, IFN-α treatment of malignant lymphoma cells and subsequent prolonged STAT1 activation cause growth arrest (Grimley and others 1998). In this early study, lymphoma cells that were resistant to the antiproliferative effects of IFN-α had reduced STAT1 activity (Grimley and others 1998). Additionally, IFN-β and IFN-γ upregulate STAT1 expression in pancreatic cells (Vitale and others 2012), and in human melanoma (Kovarik and others 2005) and breast cancer cells (Connett and others 2003), respectively, resulting in cell cycle arrest and apoptosis. In human melanoma, IFN-γ treatment also induces SOCS3 expression (Kovarik and others 2005), thereby affecting STAT activity.
There is also evidence to support the role of enhanced IFN-inducible STAT3 phosphorylation in tumor regression (Eriksen and others 2009). IFN-α and IL-21 combination therapy activates STAT3, independent of STAT1 and STAT2, in CD8+ and in CD4+ T cells, which correlates with upregulated MHC class I expression (Eriksen and others 2009). Further examination of the effects of IFN-α treatment on T cells isolated from Sezary Syndrome T-cell lymphoma patients revealed that IFN-α alone is able to suppress the malignant T-cell growth (Eriksen and others 2009). Moreover, IFN-α and IL-21 together are able to enhance CD8+ T-cell and NK-cell cytotoxicity and limit tumor growth by up to 42% in mice that have received murine renal cell carcinoma cells subcutaneously (Eriksen and others 2009). However, IFN-α-inducible STAT3 activation may trigger disease progression, presenting a risk factor to consider for IFN-based cancer therapies (Humpoliková-Adámková and others 2009). STAT3 phosphorylation after IFN-α treatment in primary lymph node metastasis cells from melanoma patients was associated with short disease-free time (Humpoliková-Adámková and others 2009). In contrast, a lack of STAT3 activation correlated with longer survival endpoints in the same study (Humpoliková-Adámková and others 2009).
Viewed altogether, the preceding highlights the important contribution that activation of STATs has to tumorigenesis, whether in certain cancers this relates to enhancing progression and metastasis or in other circumstances, to limiting tumor growth. The efficacy of IFNs as antineoplastic agents is directly related to the activation of STATs that inhibit tumor growth—for specific cancers. The challenge continues to be to consider where the application of IFN therapy is relevant, in the context of STAT activation.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
