Introduction, research interests

Peter Tompa contributed significantly to the recognition and description of structural disorder in proteins from the very beginning of this research field. Studying and explaining this structural phenomenon remains the main focus of the Researc Group of Intrinsically Disordered Proteins, with the aim to understand better the structure-function relations of proteins. Intrinsic protein disorder represents a new functionality that is fundamentally different from the traditional structure-function paradigm. Understanding how intrinsically disordered proteins (IDPs) function requires new conceptual approaches where the aim is not only to determine one structure, but rather describing how function arises from the multitude of different structural states. In order to extend the structure-function paradigm, we use the combination of bioinformatics, theoretical, in vitro and in vivo techniques and interpret the results comprehensively. Our studies include general, genome-level investigations as well as detailed structure-function analysis of selected proteins.

Main collaborators

Kyu Hoon Han, KRIBB, South Korea

Richard G. Pestell, Thomas Jefferson University, USA; Sidney Kimmel Cancer Center, USA

Laszlo Poppe, Budapest University of Technology and Economics

Jozsef Kardos, Eötvös Loránd University

Ongoing research

Role of Phase Transitions of Fusion Proteins in Cancer

Cancer, a leading cause of death in the world, is characterized by a loss of control over cell growth caused by the accumulation of genetic defects aggravated by deteriorating genome stability. Often, this manifests itself in the appearance of fusion proteins, primarily in hematological (non-solid) cancers, but fusion events are increasingly recognized in solid tumors too.

The goal of this project is to take a novel look at cancer-related fusion proteins, primarily the ones in which anaplastic lymphoma kinase (ALK) forms protein chimeras with partners, such as NPM. Such fusion events may lead to solid tumors, for example non-small cell lung cancer (NSCLC), traditionally targeted by tyrosine kinase inhibitors (TKIs), such as crizotinib. We take a radically different and provocative look at this phenomenon, by addressing if cancer can be caused by aberrant protein-protein interactions, exemplified by homodimerization, liquid-liquid phase separation and, ultimately, aggregation, of fusion proteins. By the combination of bioinformatics, structural molecular biology, polymer physics and cell biology, we identify the most important ALK-fusion proteins in NSCLC, characterize their structure and function in molecular detail and point to their tendency to engage in various types of pathological interactions.

Structural and functional characterisation of the protein-RNA complexes in the epithelial-mesenchymal transition

During the process of epithelial mesenchymal transition (EMT), the differentiated epithelial cells gain migratory and invasive properties and become mesenchymal cells. It is important during embryonic development, but in the case of tumor tissues, it represents the first step in metastasis. Many details of the molecular processes in the background of EMT are known, but many open questions remain still. One of the most important markers of EMT is the loss of E-cadherin, a main component of the extracellular matrix holding epithelial cells together. The level of E-cadherin is regulated by the Polycomb Repressive Complex 2 (PRC2) with the aid of Snail1 transcription factor. Recent scientific results point to the involvement of HOTAIR, a long noncoding RNA molecule in the regulation of E-cadherin metabolism. Cellular experiments proved that physical contact is established between Snail1, the PRC2 and HOTAIR, but the structural background of these interactions is not known. The picture has become even more complex since it was shown that not only one, but two PRC2 components (Ezh2 and Suz12) are able to bind HOTAIR. The main goal of the project is the structural and biophysical characterisation of the Snail1-HOTAIR, the Ezh2-HOTAIR and the Suz12-HOTAIR complexes in vitro and to verify the physiological relevance of the findings in cellular context.

Long-distance molecular recognition and complex assembly by IDPs

Current models fail to explain how long disordered regions (LDRs) promote complex assembly and scaffolding functions, since they were built on freely diffusing IDPs, which does not hold for most assembly proteins. Based on the architectures of known assembly proteins we propose a new model, termed Real Fly-Casting (RFC). We claim that complex assembly and long-distance molecular communication require modular proteins that are anchored to large and slowly-diffusing macromolecular entities of the cell, like membranes, DNA or RNA, by dedicated protein domains, but also harbour tentacle-like, often terminal LDRs rich in binding motifs. This arrangement could eliminate the problem of slow translational diffusion that was proposed to reduce the capture rate of freely diffusing IDPs and confer increased capture radius coupled with the efficient reduction of search space dimensions to facilitate productive recognition of distant partners. Upon recognition, conformational compaction of the LDR may exert a pulling force on the partner, thereby enforcing proximity to other components for a productive complex assembly. In previous computational studies, we raised the idea that such a mechanism may operate in NMD and vesicle coat assembly.  By fine-tuning the length, dynamics or motif composition of LDRs, alternative splicing, PTMs and evolution provide versatile means to control assembly dynamics and the subunit composition of the resulting complexes. We currently work on validating this model through a multidisciplinary approach that involves bioinformatics analyses, molecular biology, genetics and biophysics techniques, and mathematical modelling.

The interplay of alternative splicing and the assembly of membraneless organelles

Liquid-liquid phase separation (LLPS) processes that lead to the formation of different membraneless organelles mediating crucial functions in eukaryotic cells are driven by multivalent weak interactions between low complexity, intrinsically disordered protein regions (IDRs). This relationship is currently intensively researched. Even though alternative splicing (AS) of interaction-prone disordered regions is known to confer huge functional diversity onto protein isoforms, the impact of AS on LLPS regions has never been investigated. Furthermore, despite the fact that mutations affecting human LLPS regions are associated with devastating neurological disorders, muscular atrophies and cancers, the cancer-associated splicing perturbations of such regions have also not been studied. Therefore, we first collected known LLPS proteins through comprehensive literature mining and now build a dedicated, public database covering all information relevant for LLPS. Through an advanced data integration approach we identify the alternative and tissue-specific splice isoforms, as well as cancer-associated splicing perturbations of the collected human LLPS regions, estimate the impact of these variations on their conformations and select some interesting and medically relevant variants of a few proteins for experiments. With recombinant wild type and variant LLPS protein regions we study phase separation behaviour in vitro. Based on the acquired results we plan to elucidate and compare the impact of physiological AS variations and cancer-associated splicing perturbations on the LLPS processes of human cells.

The evolution of disordered regions and their coevolution with partner proteins

Through bioinformatics approaches we study the evolutionary changes affecting disordered regions. We detect signatures of coevolution between IDPs and their folded partners through correlated mutation analyses and investigate the types of amino acids and chemical bonds maintaining the identified residue-residue contacts. Also, using a comprehensive collection of positively selected human coding codons, we investigate if positive selection preferentially affects disordered regions and short linear motif-mediated interactions.

Important publications

Boeynaems S, Alberti S, Fawzi NL, Mittag T, Polymenidou M, Rousseau F, Schymkowitz J, Shorter J, Wolozin B, Van Den Bosch L, Tompa P, Fuxreiter M

Protein Phase Separation: A New Phase in Cell Biology


Varadi M, De Baets G, Vranken WF, Tompa P, Pancsa R

AmyPro: a database of proteins with validated amyloidogenic regions.


Lazar T, Schad E, Szabo B, Horvath T, Meszaros A, Tompa P, Tantos A

Intrinsic protein disorder in histone lysine methylation


Tompa P, Schad E, Tantos A, Kalmar L

Intrinsically disordered proteins: Emerging interaction specialists


Kalmar L, Acs V, Silhavy D, Tompa P

Long-range interactions in nonsense-mediated mRNA decay are mediated by intrinsically disordered protein regions.



Ágnes Tantos