RIP4 inhibits STAT3 signaling to sustain lung adenocarcinoma differentiation

Loss of epithelial differentiation and extracellular matrix (ECM) remodeling are known to facilitate cancer progression and are associated with poor prognosis in patients with lung cancer. We have identified Receptor-interacting serine/threonine protein kinase 4 (RIP4) as a regulator of tumor differentiation in lung adenocarcinoma (AC). Bioinformatics analyses of human lung AC samples showed that poorly differentiated tumors express low levels of RIP4, whereas high levels are associated with better overall survival. In vitro, lung tumor cells expressing reduced RIP4 levels showed enhanced activation of STAT3 signaling and had a greater ability to invade through collagen. In contrast, overexpression of RIP4 inhibited STAT3 activation, which abrogated interleukin-6-dependent induction of lysyl oxidase, a collagen cross-linking enzyme. In an autochthonous mouse model of lung AC initiated by Kras(G12D) expression with loss of p53, Rip4 knockdown tumors progressed to a poorly differentiated state marked by an increase in Hmga2, reduced Ttf1, and enrichment of genes regulating extracellular remodeling and Jak-Stat signaling. Tail vein injections of cells overexpressing Rip4 showed a reduced potential to invade and form tumors, which was restored by co-expression of Stat3. Altogether, our work has identified that loss of RIP4 enhances STAT3 signaling in lung cancer cells, promoting the expression of ECM remodeling genes and cancer dedifferentiation.

Role of RIP4 in lung adenocarcinoma differentiation

Jawahar Kopparam

Various studies have attributed NF-kB activation to promote KRAS mediated non-small cell lung cancer progression. In order to identify key genes regulating NF-kB activation and cell survival in lung adenocarcinoma, a synthetic lethal partner screen for KRAS was conducted in cell lines expressing mutant or wildtype KRAS. This screen identified an important role of Receptor-interacting serine/threonine protein kinase 4 (RIP4) in KRAS mediated oncogenesis. But altering RIP4 expression did not affect the viability of KRAS mutant cells, however a novel role of RIP4 in maintaining tumor differentiation was observed. Loss of epithelial differentiation and extracellular matrix remodeling are known to facilitate cancer progression and are associated with poor prognosis in patients with lung cancer. Bioinformatics analyses of human lung adenocarcinoma samples showed that poorly differentiated tumors express low levels of RIP4, whereas high levels are associated with better overall survival. Interleukin 6 (IL6) signaling is activated in lung adenocarcinoma expressing oncogenic KRAS and contributes to cancer invasiveness. Knockdown of RIP4 using shRNA in vitro enhanced activation of IL-6 signaling and also the ability of the cells to invade through collagen. In contrast, overexpression of RIP4 inhibited IL6-mediated Signal Transducer and activator of Transcription 3 (STAT3) activation, which abrogated IL6-dependent induction of lysyl oxidase, a collagen cross-linking enzyme. Tumor-specific Rip4 knockdown in an autochthonous mouse model of lung adenocarcinoma initiated by Kras(G12D) expression with loss of p53, progressed to a poorly differentiated state marked by an increase in Hmga2, reduced Ttf1 and enrichment of genes regulating extracellular matrix (ECM) remodeling and JAK-STAT signaling. Activated STAT3 can enhance NF-kB activation by interacting with NF-kB and promoting its nuclear retention. Tumors with RIP4 knockdown also exhibited enhanced STAT3- NF-kB interaction suggesting a cross talk between the two pathways. Tail vein injections of cells overexpressing RIP4 showed a reduced potential to invade and form tumors. Analysis of factors regulating RIP4 expression revealed that RIP4 is induced in a p53 dependent manner in response to chemotherapy. Altering the induced RIP4 levels in vitro had a modest effect on chemotherapy induced cell death. Our work has identified that loss of RIP4 enhances IL6 signaling in lung cancer cells, promoting the expression of ECM remodeling genes and cancer dedifferentiation.

EPFL2016

The interplay between the VEGF-C/VEGFR-3 axis and CCL21/CCR7 axis in tumor cell invasion

Amine Issa

There are two main routes for tumor spread and metastasis, the blood vasculature and the lymphatic system. A wide range of human carcinomas, including lung, colon and breast cancer, spread to distant sites via the lymphatic system. Preclinical and clinical studies have demonstrated a positive correlation between the incidence of lymph node metastasis and the secretion of vascular endothelial growth factor-C (VEGF-C) by tumor cells. Indeed, VEGF-C is critical for the formation of new lymphatic vessels in a process called lymphangiogenesis, suggesting this mechanism as the main "escape route" for tumor cells. However, many studies have failed to identify functional lymphatic vessels inside tumors, and much research has demonstrated that metastasis can occur in the absence of lymphangiogenesis. Furthermore, there is growing evidence that many tumor cells may express vascular endothelial growth factor receptor-3 (VEGFR-3), the primary receptor for VEGF-C, suggesting that autologous VEGFR-3 signaling loops may enhance the invasive and metastatic potential of tumor cells. Other than lymphangiogenic factors, the lymphatic homing chemokine receptor CCR7 is strongly correlated to lymph node metastasis, suggesting that tumor cells mimic dendritic cells in sensing and homing to lymphatic vessels and lymph nodes. Potential mechanisms underlying these observations have not been elucidated. Using several in vitro 3D models, we examined the interactions between tumor and lymphatic endothelial cells (LECs) to gain insight into the mechanistic roles of VEGF-C and CCR7-mediated tumor cell invasion towards lymphatics as well as elucidate their potential interplay. All the designed experimental setups utilized a 3D matrix for a more physiologically relevant chemokine and growth factor signaling environment. For example, we designed one culture model where tumor cells are seeded in one gel compartment and lymphatic endothelial cells in a horizontally adjacent one, allowing both cell types to establish contact through soluble mediators without being initially mixed while simultaneously allowing direct visualization of their interactions. We first demonstrated that receptor expression patterns by cells are strongly influenced by the extracellular matrix (ECM). Indeed, while we failed to detect tumor expression of VEGFR-3 in 2D conditions, we found both expression and phosphorylation of VEGFR-3 in three of four tumor cell lines tested. Furthermore, in the three VEGFR-3 positive cell lines, invasive capacity and chemoattraction towards lymphatics was demonstrated to be dependent on both VEGF-C secretion and CCR7 expression by the tumor cells. We found that VEGF-C could act in both an autocrine manner, by enhancing the migratory and invasive capacity of tumor cells, and in a paracrine manner where it enhances CCL21 secretion by LECs to promote the chemoinvasion of tumor cells via CCL21 secretion. The same mechanism was also tested in vivo by injecting VEGF-C into mouse skin: VEGF-C, but not saline nor VEGF-A, caused marked upregulation of CCL21 associated with lymphatic vessels. In this way, VEGF-C and CCR7 act synergistically to promote the invasive phenotype of tumor cells. Additionally, we uncovered a new and critical role for VEGFR-3 on mammary epithelial cells. We demonstrated that the receptor is involved in epithelial tubulogenesis in a collagen matrix, and furthermore, it helps to preserve the integrity of a mammary cell layer (since blocking the receptor leads to increased epithelial permeability). In conclusion, this work highlights the importance of the VEGFR-3/VEGF-C and CCR7/CCL21 axis in promoting tumor cell migration and invasion. It also shows that VEGFR-3 may play many roles in the process of tumorigenesis and metastasis.

EPFL2009

In Vivo Studies on the Control of Gene Expression and Protein Degradation

Karolina Bojkowska

Cellular homeostasis is maintained by tightly regulated gene networks, controlled notably at the levels of transcription, mRNA processing, and protein post-translational modification and turnover. This thesis focuses on two of these regulatory steps, namely gene transcription and protein stability. I first explored the impact of the KRAB-ZFP/KAP1 gene regulation system on liver function. KRAB-ZFPs constitute the largest family of transcription factors encoded by mouse and humans, yet their physiological roles and gene targets remain mostly elusive. Conversely, their molecular mechanism of action has been rather well characterized, as they can recognize DNA in a sequence-specific manner and recruit the universal co-factor KAP1 (also known as TRIM28), which in turn serves as a scaffold for various heterochromatin-inducing effectors. Several studies have shed some light on the roles played by this system in embryonic tissues, but its functions in the adult remain almost completely unknown. As chromatin regulation has been shown to be a major factor in the control of metabolism, we decided to investigate the role of KRAB-ZFP/KAP1 in the liver, an organ central to this process. For this, we generated conditional liver-specific Kap1 KO mice, the study of which revealed a strikingly sex-specific phenotype, with early-onset steatosis and age-related tumorigenesis restricted to males, contrasting with the mild metabolic disturbances recorded in Kap1-deleted females. Underlying this phenotype, our combined transcriptome and chromatin analyses revealed that KAP1 controls sexually dimorphic genes notably involved in hormone, drug and xenobiotic metabolism. Finally, by using a recently developed technique for direct RNA quantification, we established the range of KRAB-ZFPs expressed in the liver and likely responsible for at least part of the observed effects. This study thus demonstrates the central role of the KRAB/KAP1 system in control of sexual dimorphism, responses to xenobiotic stress and metabolic control in the liver, and further links disturbances in these processes with hepatic carcinogenesis. In parallel, in an effort to pursue a multidisciplinary training combining skills in biology and chemistry, I worked towards the development of a method to study the half-life of proteins in vivo. Protein turnover critically influences many biological functions, yet methods have been lacking to assess this parameter in vivo. Capitalizing on an in vitro technology previously developed in Kai Johnsson's laboratory, I demonstrated how chemical labeling can be used to measure the half-life of resident intracellular and extracellular proteins in living mice. Our approach is based on labeling of SNAP-tag-fusion proteins with fluorescent probes. First, we verified that SNAP-tag substrates have wide bio-availability in mice and can be used for the specific in vivo labeling of SNAP-tag fusion proteins. We then applied near-infrared probes to perform non-invasive imaging of in vivo labeled tumors. Finally, we used SNAP-mediated chemical pulse-chase labeling to measuring in vivo the half-life of different extra- and intracellular proteins, including KAP1. Importantly, this tagging technology can be applied to any protein amenable to expression as a fusion partner, whether cell-associated or secreted. Furthermore, because the SNAP-tag chemical labeling allows for a large array of probes also suited for experiments aimed at manipulating and monitoring protein function in vivo, such as cross-linking, pull-down or FRET, this methodology opens broad perspectives for studying protein function in living animals.

EPFL2011