The unholy trinity: inflammation, cytokines and STAT3 shape the cancer microenvironment

Abstract

Tumor-associated inflammation is a consequence and a driver of tumorigenesis. Three papers in this issue of Cancer Cell demonstrate the importance of tumor-elicited inflammation in development and progression of pancreatic ductal adenocarcinoma and esophageal squamous carcinoma. Disruption of tissue homeostasis culminates in activation of STAT3, generating a pro-tumorigenic inflammatory microenvironment.

One hundred and fifty years after its initial description by Virchow, tumor-associated inflammation has been recognized as an important hallmark of cancer (Hanahan and Weinberg, 2011). Inflammation contributes to almost every aspect of tumor development (Grivennikov et al., 2010). Inflammation’s role in tumor promotion has been extensively studied and recent evidence supports its function in tumor recurrence and metastasis (Tan et al., 2011). However, the role of inflammatory signals in tumor initiation has been difficult to address in part due to the paucity of adequate experimental models.

Most inflammatory signals affect tumorigenesis by activating NF-κB and STAT3 (Grivennikov et al., 2010). Persistent STAT3 activation in malignant cells stimulates proliferation, survival, angiogenesis, invasion and tumor-promoting inflammation. Moreover, STAT3 activation within immune cells enables suppression of anti-tumor immunity and promotes differentiation and recruitment of immature myeloid cells (iMCs) and tumor-associated macrophages (TAM) (Yu et al., 2009). Although in many cancers STAT3 is not directly activated by oncogenic mutations, it exerts critical oncogenic functions in both cancer and immune cells within the microenvironment.

Two papers in this issue demonstrate that STAT3 activation is essential for initiation and progression of pancreatic ductal adenocarcinoma (PDAC) (Fukuda et al., 2011; Lesina et al., 2011). In the Kras^G12D mouse model that develops pancreatic intraepithelial neoplasia (PanINs) with high penetrance, with some of which eventually progresses to PDAC (Hingorani et al., 2003), STAT3 in pancreatic epithelial cells promotes tumor development and progression, as its ablation resulted in smaller lesions, lower tumor grade and fewer metastases. Importantly, both groups demonstrate that STAT3 ablation in pancreatic epithelial cells attenuates Kras^G12D-induced PanIN formation, concluding unequivocally that epithelial STAT3 is important for tumor initiation (Fukuda et al., 2011; Lesina et al., 2011).

Effects of STAT3 on tumor initiation or early promotion are likely related to its ability to control expression of anti-apoptotic and proliferative genes, thereby increasing the likelihood that Kras^G12D-expressing cells will survive and proliferate instead of undergoing oncogene-induced senescence or apoptosis. Indeed, STAT3 in pancreatic epithelial cells controls expression of Bcl-X_L, Mcl1, survivin, c-Myc and cyclin D1 and STAT3 deficiency decreased proliferation and increased apoptosis (Fukuda et al., 2011; Lesina et al., 2011). Equally important is STAT3’s ability to promote conversion of quiescent adult pancreatic epithelial cells to cells with a progenitor-like phenotype upon pancreatitis or other insults (Fukuda et al., 2011), a process often referred to as “acinar-ductal metaplasia” (ADM). As a result of ADM, more ductal cells expressing progenitor cell markers, such as Pd×1 or Hes1, are formed and these cells are more susceptible to KRas-mediated transformation (Murtaugh and Leach, 2007). Chronic pancreatitis carries increased risk for pancreatic cancer in humans and significantly accelerates KRas-induced tumorigenesis in mice (Guerra et al., 2007). Hence, reprogramming of normal pancreatic epithelial cells into progenitors that can assume a neoplastic fate may be a key step in PDAC initiation. Importantly, activation of STAT3 was observed during pancreatic regeneration in the context of acute pancreatitis and in Kras^G12D-induced PanIN lesions.

As for STAT3’s role in pancreatic tumor promotion and progression, several mechanisms were identified. The first is the well-described ability of STAT3 to sustain cell proliferation and block apoptosis (Lesina et al., 2011). The second depends on a STAT3 target gene encoding MMP7, whose ablation in the pancreas of Kras^G12D mice limits tumor size, progression and metastasis, but does not affect ADM and PanIN development (Fukuda et al., 2011). Other studies, nevertheless, suggest that MMP7 may also be involved in early tumorigenesis, as it is needed for pancreatitis- and TGFα-induced ADM (Sawey et al., 2007). Furthermore, epithelial cell STAT3 orchestrates tumor-associated inflammation by upregulating chemokines capable of attracting immune and inflammatory cells that further propagate STAT3 activity through production of IL-6, IL-11 and other cytokines (Fukuda et al., 2011; Lesina et al., 2011; Yu et al., 2009). STAT3 ablation in the pancreas prevents macrophage recruitment during pancreatitis and tumor initiation (Lesina et al., 2011) and decreases total inflammatory infiltration and expression of pro-inflammatory cytokines during acute pancreatitis, which often precedes tumor development (Fukuda et al., 2011).

The ability of STAT3 to control and shape the tumor microenvironment directly impacts the question of how STAT3 is activated during pancreatic tumorigenesis. Fukuda et al. found that STAT3 in pancreatic cells directly affects expression of IL-6 and IL-11, both of which are STAT3 activating cytokines (Fukuda et al., 2011). Persistent expression of IL-6 or other STAT3-activating cytokines by epithelial cells may be responsible for KRas-induced STAT3 activation. Lesina et al., however, suggest that bone marrow derived cells, particularly TAM, are the major source of STAT3-activating cytokines during PDAC development. They found that STAT3 is activated in cancer cells and stromal cells in the tumor microenvironment but not in isolated cancer cell lines, suggesting a non-cell-autonomous activation of STAT3. Consistent with this, two STAT3 activating cytokines, IL-6 and LIF, were found to be robustly produced upon KRas activation (Lesina et al., 2011). Ablation of IL-6 in Kras^G12D mice demonstrated that IL-6 is one of the key STAT3 activators during PanIN initiation and progression. IL-6 uses two major modes of signaling: classical signaling through ligation of IL-6 receptor (IL-6R) and gp130 on the surface of the target cell; and trans-signaling through binding of IL-6 to a soluble form of IL-6R (sIL-6R), with subsequent binding of the IL-6-sIL-6R complex to any gp130-expressing cell. Transgenic mice that overexpress a circulating sgp130Fc fusion protein, which inhibits IL-6 trans-signaling but not classical signaling, exhibit attenuated Kras^G12D-driven PanIN progression and STAT3 activation in cancer cells, suggesting that IL-6 trans-signaling drives PanIN progression (Lesina et al., 2011). Conversely, ablation of the gp130 signaling inhibitor SOCS3 increased STAT3 activation and accelerated PanIN progression (Lesina et al., 2011). In human PDAC patients, as in many other cancers, circulating IL-6 is elevated and STAT3^Y705 phosphorylation is observed in PanIN lesions and acinar cells in tumor-adjacent tissue. An interesting and important question is whether SOCS3 expression is downregulated during PDAC development, as observed in liver and colon cancers. As in the mouse model, myeloid cells and TAM appear to be the main sources of IL-6 in human PDAC (Lesina et al., 2011). In conclusion, both epithelial and immune cells produce pro-tumorigenic STAT3-activating cytokines to promote PDAC development (Figure 1).

Figure 1.

A) Fukuda et al. suggest that KRas activation drives epithelial cell expression of cytokines, such as IL-6 and IL-11, which in turn activate STAT3 in an autocrine fashion. STAT3-induced MMP7 is required for tumor progression but not for tumor initiation, which depends on other STAT3 targets. B) Lesina et al. suggest that epithelial cells harboring active KRas recruit immune cells, particularly myeloid cells, which produce IL-6 and soluble IL-6R and activate STAT3 in epithelial cells via IL-6 trans-signaling in a paracrine manner. STAT3 induces anti-apoptotic and pro-proliferative genes, fueling tumor initiation, promotion and progression. C) Stairs et al. show that loss of p120-catenin in epithelial cells alters local tissue homeostasis, generating a pro-tumorigenic microenvironment through recruitment of iMCs and other immune cells. Recruitment of these cells is presumably mediated by cytokines and chemokines produced by mutant epithelial cells, eventually causing activation of Akt, NF-κB and STAT3 in neoplastic cells.

The third paper demonstrates p120-catenin loss in the oral cavity, esophagus and forestomach results in desmoplasia, a local inflammatory response, and leads to development of squamous esophageal cancer (Stairs et al., 2011). p120-catenin loss results in E-cadherin downregulation and activation of NF-κB, Akt and STAT3, pathways that contribute to local inflammation and tumorigenesis. p120-catenin-deficient cancer cells secrete GM-CSF, M-CSF, MCP-1 and TNF, which contribute to pronounced immune cell infiltration and create a tumor-promoting microenvironment. Interestingly, the main populations of immune cells that are recruited to the microenvironment upon p120-catenin deletion are Gr1⁺CD11b⁺ iMC. The authors demonstrate that iMCs are pro-tumorigenic due to their effects on cancer cells and cancer-associated fibroblasts (CAF). There is no evidence, however, that iMC, frequently referred to as myeloid suppressor cells, are engaged in immune suppression. It remains largely unclear how p120-catenin loss leads to activation of NF-κB, Akt and STAT3 and what is the relative contribution of these distinct pathways to tumorigenesis and tumor-associated inflammation in this system. Nevertheless, the recruitment of iMC suggests that p120-catenin loss causes upregulation of chemokines and cytokines which may account for tumor-elicited inflammation (Figure 1).

Altogether, these studies extend our understanding and appreciation of tumor-elicited inflammation in cancer development, and further underscore the notion that STAT3’s role in cancer extends beyond its well-appreciated effects on cell proliferation and survival. The role of STAT3 activation in immune cells, particularly myeloid cells, during PDAC development deserves further exploration, so does the cause of STAT3 activation in nearly all human PDAC cell lines, as it has to rely on cell-autonomous mechanisms.

The paradigm that emerges from these papers and others is that disrupted tissue homeostasis and integrity, oncogene activation or inflammatory insults result in induction of chemokines and cytokines that shape the local microenvironment. Recruited immune cells and CAF as well as altered epithelial and neoplastic cells continue to produce cytokines (such as IL-6) that activate oncogenic transcription factors (i.e. STAT3), further sustaining tumor-associated inflammation. This paradigm applies not only to pancreatic or esophageal cancers, but also to the majority of solid malignancies. This implies that drugs that disrupt tumor-associated inflammation, either by targeting STAT3 activating kinases, e.g. JAK2, or key pro-inflammatory cytokines, such as IL-6, should have significant therapeutic and preventive effects in a variety of cancers regardless of their origin.