Contact
Support Us 

Viral Oncoproteins and Domain-motif networks

Members

Researchers

Post-doctoral fellows

PhD students

Engineers

Viral Oncoproteins and Domain-motif networks

DEPARTMENT

Integrated structural biology

TEAM LEADER

Gilles TRAVE

Most biological functions emerge from highly complex interaction networks involving the three genetically encoded macromolecules of life - DNA, RNA and proteins (the genome, transcriptome and proteome) - together with all other molecular species (the metabolome). At the atomic level, molecular interactions can be depicted by their structure (high-resolution structural biology approach). At the interactomic level, interactions can be depicted by their binding energies, or affinities (quantitative interactomics approach). In our team, we develop innovative approaches to investigate the structure, the binding energies and functional consequences of protein-mediated interactomes at system-wide level. We apply these approaches to study particular thematics of biomedical interest - namely, papillomavirus-induced cancers, rare neurodevelopmental disorders, and steroid nuclear receptors.

Current projects

Our team develops six main lines of research described below.

1. Structural interactomics and inhibition of HPV oncoproteins

Gilles Travé, Alexandra Cousido-Siah, Mariel Donzeau

E6 is a small oncoprotein (ca 150 residues) produced by human papillomaviruses (HPV) responsible for various mucosal and cutaneous cancers. E6 has pro-proliferative and anti-apoptotic properties based on its large interactome with host (human) proteins, which drastically varies depending on the HPV type, tropism (mucosal or cutaneous) or oncogenic risk (high-risk or low-risk). The high-risk mucosal (hrm) HPV E6 proteins, involved in cervical, anal and head and neck cancers, bind in particular to the human Ube3A E3 ubiquitin ligase, the p53 pro-apoptotic tumor suppressor, leading to ubiquitination and subsequent proteasomal degradation of p53. This favors early proliferation of HPV-infected cells, which is beneficial to viral infection. E6 also binds to a variety of other proteins containing either "LxxLL" motifs, often involved in transcription, or "PDZ domains", involved in signal transduction, cell adhesion, cell polarity or apoptosis. Upon phosphorylation, E6 also binds to 14-3-3 phosphoreader proteins.

In this project, we study by crystallography and cryo-EM the structure of E6 bound to Ube3A, p53, LxxLL motifs, PDZ domains, and 14-3-3 proteins. Using our quantitative interactomics approaches, we explore the host interactome of E6, quantify its binding energies, and seek to understand how the E6-host interactome alters the host protein interactome. We design and investigate, in vitro and in cellulo, recombinant inhibitors of E6. We also resolve the structures of E6 proteins in complex with small molecule inhibitors developed by the University of Indiana (US).

2. Structural interactomics of particular neurodevelopmental disorders.

Gilles Travé, Elodie Monsellier

Human brain development and functioning involve complex protein networks whose binding energies are finely tuned for optimal functioning. When, due to gene alterations, a key protein actor in such a network is absent, overexpressed or mutated, this may create a "network imbalance" and lead to a suboptimal brain development or functioning, which can cause neurodevelopment disease (NDD). Hence, we propose to shift the focus from "gene-to-disease" to "interactome-to-disease"  relationship studies.

We are exploring this hypothesis by studies of Ube3A (~850 residues), HERC2 (~4800 residues) and CASK (~930 residues) proteins. Ube3A and HERC2 are both mutually-interacting E3 ubiquitin ligases belonging to the HECT domain family.  Their coding genes are close neighbors in the 15q chromosomal region, and their genetic alteration (mutation, deletion or duplication) can lead to strong NDD such as Angelman, Angelman-like or Dup15q  syndromes. In this project, we study by crystallography and cryo-EM the structures of Ube3A and HERC2, of their binding interface, and of their complexes with a variety of other human proteins. We explore the interactomes of wild-type or mutated Ube3A and HERC2, and quantify their cellular concentrations and binding affinities. We also perform cell localization and co-localization studies, and check whether  best and most relevant binders are also involved in NDD, which appears to be often the case.

CASK is a highly conserved scaffold protein achieving multiple functions in and outside of the central nervous system, thanks to the variety of protein-protein interactions it establishes through its different domains. CASK is mutated in a range of severe NDD, such as X-linked intellectual disability with or without nystagmus (XLID), X-linked intellectual disability with Microcephaly and Pontine and Cerebellar Hypoplasia (MICPCH), amongst others. The wide variety of CASK-associated functions allows for multiple possible pathological mechanisms that remain elusive. Nonetheless, the loss of interaction of CASK with some of its well-known partners has been described for several disease-related mutants. In this project, we integrate different and complementary interactomic methods to generate a contextual and quantitative CASK interactome. We also quantify how CASK pathological variants affect this interactome, and how these interactome perturbations correlate with phenotypic alterations. We expect to define pathological fingerprint of interactome perturbations that will help in assessing the pathogenicity of variants of unknown significance (VUS), a critical question for diagnosing NDDs.

3. Structural interactomics of steroid nuclear receptors

Isabelle Billas

Nuclear receptors are DNA-binding transcription factors that play a crucial role in cell growth, differentiation, embryonic development, and metabolism. Their uniqueness lies in their ability to be activated by small lipophilic ligands, establishing a direct connection between the cellular environment and gene regulation. Our research focuses on unraveling the molecular mechanisms underlying transcriptional regulation and signaling, with a particular emphasis on the tissue-specific actions of steroid nuclear receptors. Our primary subjects of study include the glucocorticoid and androgen receptors, as well as the estrogen-related receptor. Dysregulation of these receptors is linked to metabolic disorders and cancer. We utilize an integrative structural biology approach, coupled with quantitative interactomics, to investigate these processes. Our work is further enriched by functional analyses conducted in cultured cells and animal models, in close collaboration with molecular biologists and clinicians.

In parallel we are dedicated to unravelling the evolutionary forces that dictate protein function. Through collaborations with evolutionary biologists, we have made seminal contributions to understanding the evolution of nuclear receptor dimerization and the steroid hormone-receptor couple.

4. Development of high-throughput approaches for quantification of protein-ligand affinities

Elodie Monsellier

Describing cellular interactomes has been the goal of many high-throughput studies over the last decade, resulting in the identification of hundreds of thousands of binary interactions. Each of these techniques has its own strengths and flaws, resulting in a relatively low overlap between the published interactomes, and a still largely incomplete coverage of the human interactome. In addition, while affinities can span several orders of magnitude and are thus an essential parameter of the interactomes, interactomic data produced by these high-throughput experiments are almost exclusively qualitative ("binds" vs "does not bind"). Thus, the accurate description of protein interaction networks requires new approaches to address interactomes quantitatively, by measuring affinities at a proteome-wide scale.

To fill this gap our team has developed the Holdup, a robust and precise chromatographic assay for high-throughput quantification of protein-ligand affinities. Recently, the coupling of our method to a mass spectrometry readout has allowed to increase its multiplex and measure the interactions between a bait protein of interest and pools of thousands of potential preys in a single assay. Different types of preys have been successfully used: full-length proteins from total cell lysates for a discovery-driven approach, or pool of peptides / protein domains for quantifying interactions between interaction areas (fragmentomic approach). In close collaboration with teams specialized in mass-spectrometry, we are no pushing the limits of our method aiming for the best possible robustness, quantitative precision, proteome coverage, and throughput.  We are also exploring the diversity of protein ligands for which the Holdup can quantify affinities: proteins but also DNA and RNA, as well as small molecules. Finally, we are also interested in assessing the complementarity of the Holdup assay with other interactomic methods, to provide a rigorous analysis of how these different methods complement one another as well as their potential bias.

5. RNA-mediated targeting and interactomics

Mariel Donzeau

RNA-mediated approach for homogeneous and tunable intracellular protein expression. Transient transfection of foreign DNA is the most widely used laboratory technique to study gene function and product in eukaryotic cells. However, the transfection efficiency depends on many parameters, including DNA quantity and quality, transfection methods and target cell lines. We have developed an RNA-based electroporation method, which offers extremely high transfection efficiency.  Protein expression takes place few hours post-transfection, lasts at least for 48 h, and occurs evenly across cells, with a transfection efficiency reaching up to 98% cells in most cell lines tested. Most interestingly, the level of protein expressed can be finely tuned by simply modulating the amounts of mRNAs transfected. We further develop this methodology by improving the stability of the mRNAs using NTP derivatives and by optimizing the 5’ and 3’ UTR regions. mRNA electrotransfer could supplant the conventional DNA-based transient expression system.

Development of the Holdup assay for detection and affinity quantification of RNA-protein complexes and application to ncRNAs.  While 70–80% of the human genome is transcribed, only 2% of it encodes proteins. The vast majority of the human transcriptome consists of noncoding RNAs (ncRNAs). The networks in which ncRNAs act as key regulators influence a wide range of targets to control specific cell biological responses. Of particular relevance, ncRNAs have been identified as oncogenic drivers and tumour suppressors in all major cancer types. Two groups of ncRNAs have been characterized, the small and long ncRNA (sncRNAs and lncRNAs, respectively). Small ncRNAs (sncRNAs), which include among others micro-RNAs, small interference RNAs and piwi-interacting RNAs, have been extensively studied. In contrast, much less is known about the function of long noncoding RNAs (lncRNAs). Therefore, we aim to understand the complex networks of interactions that lncRNAs coordinate in the cellular context. To this end, we are developing a new variation of the Holdup assay to unravel the quantitative protein interactome of selected lncRNA molecules. As a first model, we will study the interactome of the Damage Induced NOncoding lncRNA, DINO. DINO has been characterized as a TP53 transcriptional target and functional modulator that causes TP53 reactivation in HPV-positive cervical cancer cells. The DINO lncRNA interactome may shed light on its potential role in the development of HPV-positive cancers.

6. Quantitative interactomes, affinity variations and functional impacts

Gilles Travé, Elodie Monsellier, Yves Nominé

Living organisms are highly complex systems of interacting molecules, that go through highly reproducible cycles of self-organization and chemical reactions. Intermolecular binding energies (quantified by binding affinity constants, Kd) are the interaction potentials of these systems. Whenever intrinsic genomic sequences vary, some DNA, RNA and protein molecules vary in their extrinsic interaction potentials towards the rest of the system. According to the law of mass action, this should impact the relative concentrations of the various molecular complexes formed, and therefore the global organization and the overall reactions performed by the system should also vary. Ultimately, this should impact the behaviour of the system - in other terms, its biological functions, or phenotypic traits. As described in another paragraph of our team webpage, our lab is developing methods for measuring and quantifying all the binding energies of a given protein, RNA or DNA molecule towards the other molecules. In this way, we obtain "binding profiles" that quantify the intrinsic interaction potential of each protein/RNA/DNA molecule of the studied organism, towards all the other molecules in the system. Additionally, transcriptomic and proteomic approaches provide access to the abundances/concentrations of interacting molecules in particular cell populations (tissues, organs...) of the same organism.

In this project, we measure the binding profiles of large families of protein, RNA or DNA molecules, first using their "wild-type" sequences found most frequently in the human population. Then, we re-measure the same profiles for particular "variant" sequences found across the human population. As variant sequences may be categorized as benign, likely benign, pathogenic, likely pathogenic or of unknown significance, we generally pick various instances across these different categories. Once the profiles are measured, we first apply the law of mass action to predict how the variations of binding energies should affect the proportion of complexes in chosen instances well-characterized laboratory cell lines. Next, we establish cell lines where the wild-type sequences are replaced by the corresponding variants, and measure quantitative phenotypic traits of the variant cell lines, as compared to the wild-type cell lines. The quantitative phenotypic traits that we will systematically measure will be the proteomes and transcriptomes of the different cell lines. Finally, we seek potential correlations between the quantitative variations of binding energies of the variants, and the variations of quantitative traits expressed by the corresponding variant cell lines. The final aim is to investigate whether intrinsic, genetic-based variations of binding profiles can be used to predict extrinsic variations of phenotypes within complex cell populations.

Research subgroup(s)

Collaborations and networks

Renaud Vincentelli, AFMB, Marseille (France)

Christine Carapito, LSMBO, Strasbourg (France)

Etienne Coyaud, PRISM, Lille (France)

Søren Østergaard, Novo Nordisk, Maaloev (Denmark)

Elliot Androphy, Indiana University (USA)

Funding and partners

Ligue nationale contre le cancer (équipe labellisée)

Agence Nationale de la Recherche (ANR)

National Institute of Health (NIH)

Agence nationale de la recherche contre le Sida (ANRS)

Publications

Integrated structural biology - Cancer research