Final Report Summary - COMPDESIGN (Statistical physics-based computational design of protein-RNA complexes)
In the course of the present project, a method has been developed to computationally design biomolecular complexes. Existing computational design techniques of such complexes usually neglect the role of flexibility during the binding, which undermines the reliability of these approaches. Therefore, the design methods developed here explicitly take entropy into account, thereby overcoming shortcomings of existing techniques.
The central focus of the project was placed on the computational design of RNA-switches, which are RNA molecules which have the ability to adopt alternative folds depending on complex formation with an effector RNA. This structural flexibility allows bacteria to regulate protein expression and to efficiently react to environmental changes. Rationally designing RNA-switches not only allows to gain insight into their functioning but also expands their functionality for applications in e.g. synthetic biology.
While several numerical approaches to design RNAs of a single desired secondary structure exist (e.g. [1]), the design of RNA-switches is more intricate: sequences which depending on interaction with an effector will exhibit two different foldings have to be found, leading to additional design constraints [2]. A design algorithm was developed which comprises two steps: i) sequence optimisation via repeated mutations and evaluations by a cost function that favours RNAs capable of showing the two desired folds while disfavouring competing folds [3]; ii) evaluation of the sequences' folding properties under presence and absence of the effector RNA via secondary structure prediction algorithms incorporating the entropic effects of folding [4].
This algorithm allowed to design activators and repressors of protein expression. Analysis of the RNA-switches with best folding abilities showed little unwanted sequence complementarity within the RNA-switches even at short helical lengths, thereby proving the success of the design algorithm. The functioning of the switches was experimentally confirmed in collaboration with the Universite Joseph Fourier in Grenoble [5]. In summary, the algorithm allows to design arbitrary RNA-switches to be used in biotechnological and medical applications.
In a second step, a more general design method was developed to design arbitrary biomolecular complexes. This algorithm takes entropical effects of the binding process into account by exploring conformational and sequence space via two superimposed Monte-Carlo simulation schemes: structural relaxations are alternated with random single-point mutations of amino acids of the ligand. This algorithm has been implemented for protein-peptide complexes on a recently developed lattice model [6] which allowed the proof of concept [7]. The method can straight-forwardly be generalised to RNA-protein complexes once reliable parametrisations of the interactions between RNA bases and amino acids become available.
This project allowed the researcher to broaden her expertise, establish collaborations with two European universities, and to publish two papers (one as senior author) in high-profile journals. The project further offered her the freedom to apply for permanent positions, leading to interviews in Germany, France, the Netherlands and Austria. She was ranked 3rd for a junior professorship in Austria and offered a permanent research position in France. Finally, the researcher decided to take an opportunity to work in industry, leading to the early termination of the present project.
[1] Hofacker, I. L. et al. (1994) Monatsh. Chem. 125, 167; Andronescu, M. et al. (2004) J. Mol. Biol. 336, 607
[2] Flamm, C., Hofacker, I. L., Maurer-Stroh, S., Stadler, P. F., and Zehl, M. (2001) RNA 7, 254
[3] Seeman, N. C., and Kallenbach, N. R. (1983) Biophys. J. 44, 201
[4] Xayaphoummine, A., Bucher, T., and Isambert, H. (2005) Nucleic Acids Res. 33, W605.
[5] Mladek, B. M. and Dawid, A., Rationally designing RNA switches (under preparation).
[6] Abeln, S., and Frenkel, D. (2011) Biophys. J. 100, 693
[7] Mladek, B. M. and Abeln, S., Protein-peptide complexes: a de novo design strategy (under preparation)
The central focus of the project was placed on the computational design of RNA-switches, which are RNA molecules which have the ability to adopt alternative folds depending on complex formation with an effector RNA. This structural flexibility allows bacteria to regulate protein expression and to efficiently react to environmental changes. Rationally designing RNA-switches not only allows to gain insight into their functioning but also expands their functionality for applications in e.g. synthetic biology.
While several numerical approaches to design RNAs of a single desired secondary structure exist (e.g. [1]), the design of RNA-switches is more intricate: sequences which depending on interaction with an effector will exhibit two different foldings have to be found, leading to additional design constraints [2]. A design algorithm was developed which comprises two steps: i) sequence optimisation via repeated mutations and evaluations by a cost function that favours RNAs capable of showing the two desired folds while disfavouring competing folds [3]; ii) evaluation of the sequences' folding properties under presence and absence of the effector RNA via secondary structure prediction algorithms incorporating the entropic effects of folding [4].
This algorithm allowed to design activators and repressors of protein expression. Analysis of the RNA-switches with best folding abilities showed little unwanted sequence complementarity within the RNA-switches even at short helical lengths, thereby proving the success of the design algorithm. The functioning of the switches was experimentally confirmed in collaboration with the Universite Joseph Fourier in Grenoble [5]. In summary, the algorithm allows to design arbitrary RNA-switches to be used in biotechnological and medical applications.
In a second step, a more general design method was developed to design arbitrary biomolecular complexes. This algorithm takes entropical effects of the binding process into account by exploring conformational and sequence space via two superimposed Monte-Carlo simulation schemes: structural relaxations are alternated with random single-point mutations of amino acids of the ligand. This algorithm has been implemented for protein-peptide complexes on a recently developed lattice model [6] which allowed the proof of concept [7]. The method can straight-forwardly be generalised to RNA-protein complexes once reliable parametrisations of the interactions between RNA bases and amino acids become available.
This project allowed the researcher to broaden her expertise, establish collaborations with two European universities, and to publish two papers (one as senior author) in high-profile journals. The project further offered her the freedom to apply for permanent positions, leading to interviews in Germany, France, the Netherlands and Austria. She was ranked 3rd for a junior professorship in Austria and offered a permanent research position in France. Finally, the researcher decided to take an opportunity to work in industry, leading to the early termination of the present project.
[1] Hofacker, I. L. et al. (1994) Monatsh. Chem. 125, 167; Andronescu, M. et al. (2004) J. Mol. Biol. 336, 607
[2] Flamm, C., Hofacker, I. L., Maurer-Stroh, S., Stadler, P. F., and Zehl, M. (2001) RNA 7, 254
[3] Seeman, N. C., and Kallenbach, N. R. (1983) Biophys. J. 44, 201
[4] Xayaphoummine, A., Bucher, T., and Isambert, H. (2005) Nucleic Acids Res. 33, W605.
[5] Mladek, B. M. and Dawid, A., Rationally designing RNA switches (under preparation).
[6] Abeln, S., and Frenkel, D. (2011) Biophys. J. 100, 693
[7] Mladek, B. M. and Abeln, S., Protein-peptide complexes: a de novo design strategy (under preparation)