Andrzej Stasiak received his PhD in 1981 from the Institute of Biochemistry and Biophysics of Polish Academy of Sciences in Warsaw. From 1981 to 1989 he was a postdoctoral fellow and research associate in the laboratory of Theodor Koller at Institute for Cell Biology, ETHZ, Zurich, Switzerland. In 1989 he joined the Laboratory of Ultra-structural Analysis directed by Jacques Dubochet at the UNIL. In 2007 he joined the Center for Integrative Genomics as Maître d’Enseignement et de Recherche (MER).
The main interest of our group is directed towards understanding the topological aspects of overall organization of genomes starting with the arrangement of DNA in phage heads and ending with chromosome territories in higher eukaryotes. We try to understand how bacterial topoisomerases actively remove knots from crowded DNA molecules and what are the consequences of so called topological exclusion in shaping chromosome territories. We are also interested in the topological aspects of protein folding with the aim of classification of knotted proteins and understanding of particular advantages provided by knotting of the polypeptide chain. Some aspects of our interests go beyond biology, like for example statistical mechanics studies of knotted polymers or simplification pathways of various knots. However, also these studies have biological or biophysical applications in helping to explain, for example, how DNA topoisomerases can efficiently distinguish knotted DNA molecules from unknotted ones. We also maintain our interest in studies of proteins participating in DNA recombination and DNA repair. During recent years numerical methods used to model behavior of DNA molecules and of chromatin fibers became the main method utilized by our group, although biochemistry and electron microscopy, in particular, remained as an essential part of our research.
Organization of nuclear architecture
We are interested in revealing the underlying physical phenomena responsible for the overall organization of chromosomal territories in interphase nuclei. In particular we try to explain why chromatin fibers belonging to different chromosomes do not intermingle with each other but each forms its own chromosome territory. We try to understand how topological domains are formed and what are the factors that are responsible for high contact frequency between sites belonging to the same topological domain, whereas interdomain contacts are relatively infrequent. We try to understand the mechanism of chromatin insulators such as these provided by CTCF bound to chromatin fibers.
Supercoiling and DNA unknotting and decatenation
Normal functioning of DNA requires that intermolecular catenation and intramolecular knotting should be quickly resolved by action of DNA topoisomerases. In bacterial cells the level of DNA knotting and catenation is much lower than this resulting from random DNADNA passages. The question arises then how DNA topoisomerases are directed to act on DNA-DNA juxtapositions where intersegmental passages specifically lead to unknotting and decatenation rather than to act on juxtapositions where passages can produce knots and catenanes. In bacterial cells the DNA is negatively supercoiled and several earlier reports revealed that DNA supercoiling helps DNA topoisomerases to remove DNA knots more efficiently. To shed light on the role of DNA supercoiling in DNA unknotting, we have modeled knotted DNA molecules that are supercoiled. We observed that supercoiling resulted in very strong localization of knotted portions of the molecules in such a way that the local curvature in the knotted portion was significantly higher than the average curvature in the rest of negatively supercoiled DNA molecules. Since it was known already that type II DNA topoisomerases preferentially bind to DNA portions that are highly curved our result provided the missing link needed to understand how DNA supercoiling directs type II DNA topoisomerases to preferentially act on knotted portions of DNA molecules.
Studies of protein knotting
The polypeptide chain of some proteins is knotted. In a collaboration project involving mathematicians and biophysicists (Prof. K. Millett, UCSB, USA, Prof. E. Rawdon, Univ. St. Thomas, USA, Prof. J. Onuchic and Dr. J. Sulkowska, UCSD, USA) we analyzed all deposited protein structures for the presence of knots. Our novel form of analysis permitted us to obtain knotting fingerprints of various proteins. We observed that despite large sequence variance the precise knotting pattern is highly conserved within to the same protein family and sometimes the conservation involves separate families. High conservation of knotting patterns naturally suggests that knots in proteins have a unique function that is hard to achieve without knotting. We continue our study aimed to understand the evolutionarily advantage of knotted proteins.
Electron microscopy studies of DNA structure and of protein -DNA interactions implicated in the process of DNA recombination
In a collaboration project with Dr. T. Lionberger (University of Michigan) we used Cryo-EM to study effects of DNA supercoiling on DNA minicircles. We observed that negative supercoiling in DNA minicircles induces formation of two kinks that are placed 180° apart along the circumference of DNA minicircles. This observation has important implications for gene regulation processes that involve formation of small DNA loops.
Our group has a long time experience in electron microscopy imaging of functional complexes formed with DNA by various proteins participating in the process of homologous recombination. Our studies contributed in a significant way to understanding the role and the mechanism of action of such proteins as RecA, RuvAB, RAD51, RAD52, DMC1, BRCA2 or FANCM. In a new collaboration project involving group of Prof. J.-L. Viovy (Institut Curie, Paris) we used electron microscopy to complement magnetic tweezers studies of RAD51-DNA complexes, in which the DNA was forced to take different twist values. Our study revealed that two different forms of RAD51-DNA complexes, the stretched and non-stretched form are convertible into each other without protein dissociation. This finding brings us closer to complete understanding of molecular mechanisms by which RAD51 protein mediates homologous pairing and DNA strand exchange during double-strand break repair.