Cell Stress

The Cell Stress team focuses on understanding the genetic program for pancreatic cancer initiation, progression and metastasis formation.

Why we study this cancer?

Because, i/ although it has the 9th position in frequency, among the solid cancers, it is the 4th cause of death by cancer; ii/ pancreatic cancer has a ratio « incidence/death » of almost 1, that signify that every new patient diagnosed of this pathology will death of this disease; iii/ the 5 years survival is below 3%; iv/ the median survival of patients with this tumor is 6 months after its diagnosis; v/ surgery, which is the only therapeutic possibility to increases the survival, is available for only 15% of patients.

Why are these data so negative ? Probably, because the own characteristics of the PDAC’s. Our projects are covering several complementary aspects of the pancreatic cancer.

In fact, we are :

  • identifying personalized treatments against panceatic cancer by investigating the molecular heterogeneity of this disease; Nelson Dusetti
  • studying proteins with tumor suppressive activity; Alice Carrier
  • analyzing the metabolic changes occurring in transformed pancreatic cells; Sophie Vasseur
  • screening novel antitumoral compounds synthesized by us and studying their mechanism of action; Juan Iovanna
  • studying the role of the over-expressed stress proteins in pancreatic cancer cells; Juan Iovanna
  • dissecting the molecular dialogue established between the stroma components and the transformed pancreatic cancer cells; Richard Tomasini
  • analyzing the mechanism of the endothelial-to-mesenchymal transition that occurs in pancreatic cancer; Roselyne Tournaire
  • developing  all necessary tools for studying all known post translational modifications in pancreatic cancer cells; Philippe Soubeyran
  • studying the molecular mechanisms involved in the formation of aberrant glycoconjugates and their clinical impacts ; Eric Mas

To these ends, we have developed several sophisticated molecular, cellular and animal tolls such as vectors of expression (plasmids, retrovirus, letivirus, etc), complex co-culture systems mimicking the tumor microenvironment, some transgenic, KO, KI and immunodeficient animal models of pancreatic cancer, a platform for middle throughput screening of protein-protein interaction based on an original yeast two hybrid system, a platform for high throughput screening of inhibitor of protein-protein interaction based on the BRET approach and a cell culture platform on ISO9001 standard.

Our laboratory is about 40 people, including scientists, oncologists, gastroenterologists, surgeons, pathologists, informaticians, several engineers and technicians and post doc and PhD students.

Proteins with Tumor Suppressive Activity | Alice Carrier

Tumor suppression is a complex network of molecular and cellular processes leading to the reduction of pro-tumoral cell damage and/or removal of damaged or tumoral cells. These processes are activated in stressful conditions such as nutrient, oxidative or inflammatory stress.

My group is interested in studying the mechanisms underlying tumor suppression. In this context, our research has highlighted the tumor suppressive activity of two proteins initially identified as highly expressed in the thymus: the stress factor TP53INP1 (Tumor Protein 53-Induced Nuclear Protein 1) and the protease TSSP (Thymus-Specific Serine Protease).

Our functional genomics studies aimed at exploring the function of both proteins at the molecular and cellular level. In particular, we analyzed their impact on the cellular dysfunctions sustaining tumoral development, by the use of pancreatic and colorectal induced tumorigenesis models in deficient mice for each of these proteins (generated in our laboratory). Our studies demonstrated that these proteins are involved in tumor suppression through distinct mechanisms. In the case of TP53INP1, we evidenced its activity in biological processes that are affected in cancer cells (proliferation, apoptosis, migration, angiogenesis, genetic instability), tightly linked with its contribution in cell redox status regulation. Two major mechanisms by which TP53INP1 plays this regulatory function have been unveiled: (1) TP53INP1 regulates the transcriptional activity of p53 in the nucleus, and (2) TP53INP1 contributes to mitophagy, a process that ensures the quality control of the mitochondrial compartment by avoiding the accumulation of damaged mitochondria that produce an excess of pro-tumoral ROS (Reactive Oxygen Species). 





In the case of TSSP, we demonstrated its involvement in MHC classII antigen presentation triggering CD4+ T cell differentiation in the thymus, thus impacting the generation of a fully efficient T lymphocytes repertoire against tumoral development.





These studies improve the full understanding of cancer development, which is required for developing novel avenues in cancer prevention.

 Currently, we are developing a project which aims at deciphering the impact of mitochondrial alterations in pancreatic cancer cells aggressiveness and resistance to chemo- and radio-therapy. The Pancreatic Ductal AdenoCarcinoma (PDAC) is resistant to currently available treatments. Despite progress in our understanding of the physiopathology of this cancer, many aspects remain poorly understood, which impedes the development of novel therapeutic strategies. Among these strategies, targeting metabolic dysregulations of cancer cells is promising. In particular, the mitochondrial metabolism remains poorly explored in cancers despite the central role of these organelles in cell bioenergetics and apoptosis. The aims of our ongoing project are to decipher mitochondrial dysregulations in PDAC tumors, to correlate these alterations to tumor growth and therapeutic resistance, and to develop preclinical assays targeting mitochondria. 

 To top

Development of personalized treatments for patients with pancreatic adenocarcinoma | Nelson Dusetti

Pancreatic cancer arises due to several factors. Among these, mutations, selection of resistant cells and clonal expansion play a pivotal role. As a result of these phenomena, an important intra and inter-tumor heterogeneity is generated and is directly linked to the harmful prognosis of this cancer. Despite being well-documented, the heterogeneity of pancreatic tumors remains poorly addressed.

Two main issues have hampered a thorough characterization of this heterogeneity:

1-Variation of the stromal content (between 15 to 85%), which is significant between the pancreatic tumors and even stronger between different regions of the same tumor. Most studies did not account for this variation and considered the tumor to be a homogeneous mixture of tumor cells and stroma.

2-Investigations were restricted to surgical specimens to ensure a sufficient amount of tissue. However, this might lead to sample biases because only patients with more advanced tumors (15 to 20%) undergo surgical procedures.

Recently, we have developed a strategy that improves the study of the molecular mechanisms underlying pancreatic cancer. We constructed a "breathing" tumor biobank on the in vivo xenograft model. This method allows us to obtain an almost unlimited amount of cancerous epithelial cells. The stromal cells represent only 10 to 15% of the xenograft and are distributed homogeneously. In addition all patients with pancreatic adenocarcinoma can be included in our studies, 1 metastatic or locally advanced patients for whom we obtain very small tumor samples by ultrasound and fine needle aspiration (EUS-FNA) and 2-operated patients. A consortium was created for this purpose composed of clinicians, gastroenterologists, oncologists and surgeons from the Nord Hospital, La Timone and the Paoli Calmettes Institute of Marseille as well as biologists from the Cancer Research Center of Marseille.

How to personalize the treatments?

Pancreatic cancer is a systematically fatal disease. The chemotherapy given to fight pancreatic adenocarcinoma is indicated according to the general state of the patients and the stage of their disease progression, but not according to the tumor biology. At a time when personalized medicine seems to be the better alternative for the treatment of most cancers, it becomes urgent to adapt the treatment to the subtypes of pancreatic adenocarcinoma and their sensitivity to available drugs. We envision that detailed knowledge on the molecular mechanisms underlying pancreatic tumors could allow to predict the clinical course and chemo-sensitivity of this disease. To test this hypothesis, we have developed omic studies and high throughput techniques that led us to the identification of molecular signatures. These signatures allow the realization of highly accurate and sensitive biological tests that can be done on tumors through the detection of a reduced number of biological molecules. Molecular signatures bring us a new view of tumors beyond histopathological types, grade, and classical stage parameters. They have the appeal of novelty over conventional approaches and can help us tailor treatments to the types of cancer cells that make up the tumor.

Personalized medicine has the potential to revolutionize cancer treatment in the future. This therapeutic strategy is based on the identification of molecular signatures that characterize pancreatic cancer of specific patients, which could lead to effective therapeutic strategies adapted to individual patients.


 To top

Cell Stress | Juan Iovanna

The protein Nupr1/p8:

In the past we have shown that expression of this gene is necessary for tumor development. Recently, we were able to assign a major role in the regulation of epithelial-to-mesenchymal transition (EMT).

In particular, we demonstrated that expression of Nupr1/p8 is involved in a novel mechanism of cell death called entosis or cellular cannibalism. We have partially dissected the molecular mechanism of this new function of Nupr1/p8 and explored the involved pathways.

Finally, we demonstrated that expression of Nupr1/p8 is essential for the cellular response to nutrient deprivation or hypoxia (metabolic stress), the mechanism involved is dependent of the NFkB alternative (RelB), but not of the classical NFkB pathway (p65). We are currently fully characterizing this very original mechanism.


The PAP protein:

We have demonstrated that the interaction of PAP with its receptor activates, on one hand, the Jak-STAT3-SOCS3 and, on the other hand, inhibits the activation of NFkB. These two pathways appear to be involved in the anti-inflammatory effect of PAP.

Because PAP is a secretory protein and we demonstrated the existence of a specific receptor, the PAP-receptor could be a new target for treating certain inflammatory diseases. We are characterizing and trying to clone it.


VMP1 protein:

We demonstrated that expression of VMP1 is “necessary and sufficient” to trigger autophagy. Overexpression of VMP1 induces vacuole formation while the use of a siRNA against VMP1 prevents the recruitment of LC3 and vacuole formation induced by nutrient deprivation or rapamycin treatment.

Then, we developed a transgenic mouse in which overexpression of VMP1 was induced in the pancreas and validated our in vitro studies.


Development of new small anticancer compounds:

In collaboration with a team of chemists from the CNRS we are interested in developing new compounds with antitumor activity in pancreatic cancer. Chemists have synthesized hundreds of novel compounds and we have systematically screened their activity in vitro and in vivo.

Among these molecules, we identified two compounds, structurally related, whose activity is higher than that of gemcitabine in vitro as well as in vivo. Our work is currently focusing on optimizing the structure of these molecules and actively researching on their mechanism of action.

 To top

Regulation of Signaling Pathways | Philippe Soubeyran


Post-translational modifications and therapeutic escape:

Characterized by a 5 years survival rate of less than 5% the pancreatic cancer (pancreatic adenocarcinoma) remains one of the cancers with the worst outcome. This situation is notably due to the exceptional resistance of pancreatic tumors to conventional anticancer therapies. We think that we have to use original approaches to discover new fighting tools.

Ubiquitination is a post-translational modification process known for several tenths of years. However, this is only these last years that the great potential of ubiquitination to regulate cellular functions has been revealed. We think, and some data confirm it, that defects in ubiquitination play an important role in tumor development and resistance to chemo and radiotherapies.

We are trying to identify the proteins which ubiquitination is modified when pancreatic cancer cells undergo anti-cancer treatments. Then we are studying all candidate proteins by evaluating their importance, and the importance of their ubiquitination, regarding the resistance of pancreatic cancer cells.

Then we try to depict the molecular mechanisms responsible for the changing in ubiquitination of these candidate proteins.This should enable us to develop molecules capable to inhibit these modifications in vitro and importantly in vivo.

Our goal will be then to test these molecules upon the resistance of pancreatic cancer cells.

 To top

Microenvironment and Pancreatic Tumorigenesis | Richard Tomasini

Pancreatic ductal carcinoma (PDA) is, at present, the fifth cause of cancer death but predicted to become the second leading cause of cancer death by 20201. On top of those dramatic epidemiological data2,3 , PDA figures, with a median survival of 6 months and overall 3-year survival rate of only ~6%, as the solid cancer with the worst prognosis. While clinical studies, with the development of combined therapies as Folfirinox, the gain in imaging based diagnosis as well as the improvement of surgery quality based on surgeon formation and centralization policy4 led to significant enhancement of patients management, PDA-related fundamental research lack behind and fails to transfer onto the clinic the tremendous enhancements and novel insights done in the last decades. Until now, this strategy has met with rather limited success, overall survival improving by “only” 4 months in the last decade, and only for a limited number of patients. The poor outcome for PDA patients is mainly due to: (1) a late diagnosis, with only 15% of patients eligible for surgery; (2) the lack of unbiased biomarkers, preventing the screening of at risk patient populations and/or leading to patients stratification which would open the field of personalized or adapted therapies; (3) the resistance of the disease to current mono or combined therapies, without any second line choices other than palliative treatments; and (4) the by necessity restrictive use of the best chemotherapies available as most patients failing, since diagnosis, to meet the global health criteria required to withstand their associated toxicity.

In trying to understand “the” reason for such negative outlook, one has to consider that the fundamental and clinical research conducted during the last 30 years has focused on characterizing and understanding the properties of the PDA cancer cells, neglecting the specific cellular context of those tumors. Indeed, as mentioned by pathologists since decades, PDA is a solid tumor with a most prominent desmoplastic reaction (also called intra-tumoral microenvironment or stroma compartment)5. This stroma is composed of host-derived non-tumor cells, such as cancer-associated fibroblasts (CAF), immune cells (M2 macrophages, mast cells, regulatory T-lymphocytes, etc.), endothelial cells and nerve fibers, as well as a dense acellular matrix mainly made of collagen or hyaluronic acid. Recent clinical studies as the CONKO-001 study further reported the specific clinical relevance of stromal cellularity with a negative prognostic impact of high CAFs density in pancreatic cancer patients after curative intended resection6. In PDA the intra-tumoral microenvironment, with a specific focus on CAFs abilities, has been largely associated with tumor-related processes7. Indeed, as the major cell component of the stroma compartment and in association with their phenotype of low-proliferating and highly-productive cells, CAFs were reported to influence tumor initiation and progression, tumor cell aggressiveness, metabolic reprogramming, metastatic dissemination and niche establishment, neural remodelling and chemo-resistance8-10. Besides those strong lines of evidences, some studies, mainly based on genetically engineered mouse models of PDA reported an anti-tumoral potential of CAFs11-12. While controversial, those studies highlighted the necessity for the field to deeply analyse this complex population named CAFs which engulf cells of various origins, with different abilities, known markers and even anatomical localisation within the tumor mass13.

One explanation of the failure in succeeding to transpose fundamental research findings into the clinic for PDA treatment is the low consideration brought to this PDA intra-tumoral microenvironment even if its impact is reported since more than 15 years. In order to considerate then to analyze how the stroma impacts on PDA it becomes inevitable to decipher how those various cells communicate and exchange information to create a reactive network and to coordinate answers towards environmental changes. Cell-cell contacts and soluble factors have been extensively studied and are commonly considered to be central players in the communication between different cell types. More recently, a new player arose into cell-cell communication permitting to exchange ‘complex information’: the extracellular vesicles (EVs)16.

In regards to this context and following our previous publication17, our group intends to decipher the impact of EVs communication between stromal and tumoral cells, within PDA mass, in order to use those knowledges to improve patients’ management. 3 projects are ongoing (see Figure 1); 1/ the impact of stromal-EVs on tumor cells resistance to chemotherapies (PhD Jeremy Nigri), 2/ how cancer cell-EVs shape their environment and stromal cells abilities (PhD Zainab Hussain) and 3/ how inter-stromal cell communication mediated by EVs modulates stromal cells (CAFs and Macrophages) phenotypes and capacities (Post-doctorant Thomas Bertran)



Figure 1.



1. Rahib, L. et al. Cancer Res. 74, 2913-2921 (2014).

2. Bouvier, AM. et al. Pancreas. 39, 1243-1246 (2010).

3. Maire, F. et al. Eur J Gastroenterol Hepatol. 29, 904-908 (2017).

4. Farges, O. et al. Ann Surg. 266, 797-804 (2017).

5. Apte MV. et al. Pancreatology. 15, 32-38 (2015).

6. Sinn, M. et al. Br J Cancer. 111, 1917-1923 (2014).

7. Von Ahrens, D. et al. J Hematol. Oncol. 10, 76 (2017). doi: 10.1186/s13045-017-0448-5.

8. Ireland, L. et al. Cancer Res. 76, 6851-6863 (2016).

9. Franco-Barraza, J. et al. eLife. 6:e20600 (2017). doi: 10.7554/eLife.20600.

10. Secq, V. et al. Cell Death Dis. 6:e1592 (2015). doi: 10.1038/cddis.2014.557.

11. Özdemir, BC. et al. Cancer Cell. 25, 719-734 (2014).

12. Rhim, AD. et al. Cancer Cell. 25, 735-747 (2014).

13. Chiavarina, B. et al. Curr Med Chem. 24, 2846-2859 (2017).

14. Ölhund D. et al. J Exp. Med. 214, 579-596 (2017).

15. Wong, KM. et al. Curr Oncol Rep. 19, 47 (2017).

16. Willms, E. et al. Sci. Rep. 6, 22519 (2016).

17. Leca, J. et al. JCI. 126, 1-17 (2016).

 To top

Angiogenesis, Microenvironment and Cancer | Roselyne Tournaire

Stromal cells provide a microenvironment which is essential for tumoral growth, invasion and metastasis. Fibroblasts are essential components of this microenvironment. Activated fibroblasts (also called myofibroblasts or CAFs for Cancer-Associated Fibroblasts) are localized in close proximity to cancer cells. They are able to modify the phenotype of epithelial cells either by direct cell-cell contact, or through soluble factors, or by modifying the extracellular matrix. The interaction between myofibroblasts and cancer cells is essential for tumor growth and for the formation of metastasis, and the presence of such cells is associated with a poor clinical prognosis. 

Endothelial-mesenchymal transition (EMT) is a process by which endothelial cells disaggregate; endothelial cells change shape and migrate to invade surrounding tissues. EMT is characterized by the loss of endothelial cell markers and the acquisition of mesenchymal cell markers: this process has been reported o occur during embryonic development and in cases of fibrosis.

Mesenchymal cells derived from endothelial cells behave like fibroblasts in damaged tissues.

We are interested in the receptors involved in angiogenesis and more specifically in Tie family receptors, Tie1 and Tie2, which are involved in blood vessel maturation.

We have recently demonsrated, first that the absence of Tie1 in endothelial cells leads to EMT, and second, that EMT occurs in human pancreatic tumors, including in humans. EMT thus contributes to generate the pool of fibroblasts associated with cancer. These data suggest that therapeutic strategies targeting EMT could offer new perspectives for the treatment of pancreatic cancer. Our aims are to analyze the mechanisms involved in EMT and determine the impact of tumor microenvironment on EMT.

Nos objectifs sont d’étudier les mécanismes impliqués au cours de l’EndMT et de déterminer l’impact du microenvironnement présent dans les tumeurs sur la transition endothélio-mésenchymateuse.

 To top

Impact of Hypoxic and Metabolic Stress on Pancreatic Tumor Progression | Sophie Vasseur

Pancreatic ductal adenocarcinoma remains one of the most lethal of all solid tumors with an overall five-year survival rate of only 3–5%. Its aggressive biology and resistance to conventional and targeted therapeutic agents lead to a typical clinical presentation of incurable disease once diagnosed. Presence of a prominent non-tumoral cell compartment within the tumor (a main characteristic of PDAC) is directly impacting on patient clinical outcomes.

In PDAC, cancer cells are “isolated” by a fortress of stromal cells, composed of very few blood vessels, which distorts the normal parenchymal architecture of pancreas and limits the oxygen and nutrient diffusion in the tissue. This severe hypoxic environment at the site of the tumor provides a strong selective pressure able to regulate tumor cell growth and to favor survival of the most aggressive malignant cells.

Another hallmark of PDAC is cachexia, defined as an unintend weight loss of 10% or more of the stable weight over a period of 6 months, and associated with loss of fat and muscle tissue. Pancreatic cancer-patients with cachexia often have an elevated resting energy expenditure (REE), a higher rate of more progressed tumor stages and have significantly reduced survival. Hence, this tumor must harbor metabolic pathways which are probably tied in a complex inter-organ dialogue during its development and illustrate this cancer as a real metabolic disease.

One of the major consequences of intra-tumoral hypoxia, combined with oncogene signaling, is the cell metabolic switch occurring in order to meet the requirements of tumor proliferation under low oxygen and low nutrient supply because of lack of vasculature. Since decades it has been accepted that tumor cells display fundamental changes in pathways of energy metabolism and nutrient uptake, and that alterations in cellular metabolism contribute to malignant phenotype. Cancer cells differ from healthy cells due to a plethora of molecular changes which are mechanistically linked to metabolic reprogramming. Metabolic signature of each type of tumor is then strongly associated to oncogenic mutations occurring in the tumor associated cells.

Our project is based on the concept that pancreatic cancer cells are probably metabolically flexible.

However, molecular changes leading to metabolic adaptations of pancreatic cancer cells remain unclear. We propose innovative strategies to abolish PDAC progression, based on the concept that targeting molecules belonging to a specific « PDAC Metabolic Signature » can be a new relevant therapeutic option to treat PDAC.

Based on the metabolic transcriptomic signature of PDAC, we propose strategies targeting molecules mainly involved in lipid, amino-acid and glucose metabolic pathways, 3 essential metabolic processes for energy production and biosynthesis during tumor growth.

Combined therapies to avoid resistance to a single anti-metabolic drug is challenging but appears to be a good strategy to counteract the tumor cell metabolism plasticity and pancreatic tumor progression.

Integrating hypoxic and metabolic stress in pancreatic tumor development. Activated oncogenes and loss of tumor suppressors induce growth and proliferation of small amount of pancreatic cancer cells leading to formation of tumor mass. In this mass, the tumor cells are rapidly surrounded by a very dense stroma composed of nerve fiber, immune cells, few blood vessels, fibroblasts and stellate cells. Such desmoplasic reaction gradually reduces oxygen and nutrient supply to cancer cells. As a consequence, tumor cells activate the hypoxia-inducible factor 1α (HIF-1α and undergo several metabolic shifts to adapt and survive. Pancreatic tumor cells harboring such changes and able to proliferate under very hypoxic and nutrient deprived conditions, promote epithelial to mesenchymal transition (EMT) and acquire a very aggressive, metastatic phenotype. Although most tumor cells experiencing metabolic stress undergo apoptosis, those surviving will be clonally expanded, and submitted to genomic instability, leading to tumor growth and pancreatic ductal adenocarcinoma formation (PDAC).

 To top

Adhesive networks and invasion | Frédéric André

Glycosylation Process | Éric Mas

Pancreatic adenocarcinoma (PDAC) is a devastating disease progressing asymptomatically until death within months after diagnosis. The predisposition to metastasize is very important in the development of PDAC and implicates complex molecular mechanisms. In particular, aberrant glycosylation of glycosphingolipids and glycoproteins expressed in tumor cells has been implicated as one of essential mechanisms in malignant transformation, cell adhesion and metastatic dissemination. Although it was shown that expression modifications of glycosyltransferases and glycosidases play a key role in the formation of aberrant glycoconjugates, little information is available on the regulation of mechanisms that produce altered glycan structures during pancreatic carcinogenesis and their clinical impact.

Thus, one of objective of our project is to decipher the relationships between glycosyltransferase and glycosidase expression, aberrant glycoconjugate expression and metabolic changes by means of transcriptomic data from PDAC. We also investigate the oncogenic signaling pathways dysregulated in pancreatic carcinogenesis and involved in glyco-enzymes and aberrant glycoconjugate expressions, their functions and their roles in PDAC aggressiveness and metastasis formation. Our goal is to identify and characterize the key-points involved in the modifications of metabolic pathways of aberrant glycoconjugates in particular those of sialylated and/or fucosylated as potential therapeutic targets to treat pancreatic cancer.

In the same time, we try to identify signatures of aberrant glycans and signatures of genes involved in the formation of aberrant glycans by means transcriptomic data from pancreatic tumors, correlated with biological and clinical data of patient. Our goal is to do the proof of concept of signature to classify and predict the outcome of patients in terms of progression and metastatic dissemination.




Figure : Impact of aberrant glycosylation processes in pancreatic cancer