ľ¹Ï¸£ÀûÓ°ÊÓ

"How do proteins get in touch?"

A structural bioinformatics and modelling project from the NMR Research group.

contact person:

Dr. A.M.J.J. Bonvin
Bloembergen NMR building, room 1.22, phone:030-2533859

Introduction:

With the presently available amount of genetic information, a lot of attention focuses on systems biology and in particular on biomolecular interactions. Considering the huge number of such interactions, and their often weak and transient nature, conventional experimental methods such as X-ray crystallography and NMR spectroscopy will not be sufficient to gain structural insight into those. A wealth of biochemical and/or biophysical data can however easily be obtained for biomolecular complexes. Next to experimental data, bioinformatic predictions can be made based for example on amino-acid sequence conservation. Combining the predicted and/or experimental information with docking, the process of modeling the 3D structure of a complex from its known constituents, should provide valuable structural information and complement the classical structural methods.

Goals:

The goal of this project is to investigate how already available interface mapping data obtained from mass spectrometry techniques and bioinformatic predictions can help in the modeling of biomolecular complexes. This information will be translated into structural restraints which will be introduced in the data-driven docking approach HADDOCK developed in the NMR research group. The results obtained from data-driven docking should be compared to a purely ab-initio docking approach based only on shape complementarity and energetic consideration.

Methods:

This project involves computational modeling and bioinformatics approaches. It will consists of the following steps.

• Data mining and bioinformatics analysis of the protein sequences to predict and analyze the binding interfaces. This information will be required in order to model the complexes in absence of MS experimental data to provide a reference point. You will use among other interface prediction methods developed in the NMR research group (see )
• Computational modeling of the complexes by docking using the data-driven docking approach HADDOCK developed in the NMR research group (see ). The work will be performed on linux workstations with access to computational Beowulf clusters.
• Computational modeling of the complexes by ab-initio docking using web-based docking servers from other groups.
• Structural analysis of the resulting models with a variety of software

Planning & practical aspects:

For this project, the students will work in two groups: one group will investigate the use of experimental data and/or bioinformatic predictions for the modelling of complexes by docking while the other group will concentrate on ab-initio modelling (thus without data). The results of the two different approaches should at the end be compared in the light of the available experimental information.

Characterisation of protein complexes involved in transcription.

contact person:

G.E. Folkers
Bloembergen NMR building, phone:030-2539930

Scientific background:

The post genome projects, aiming at functional characterization of individual components, and the relationship between them is providing a deeper understanding in regulatory processes of complex biological systems. Key towards understanding such regulatory networks are the availability of quantitative information on the interaction between the individual components. These interaction networks are scale free, where few proteins participate in many interactions (hubs). Especially the so-called date-hubs participate in multiple distinct interactions through an intrinsically unstructured region. Only if it is understood how these date hubs regulate in a spatio-temporal fashion the large number of distinct interactions, we can begin to predict complex biological systems, where the nature of the interaction determines the outcome. To fully understand how proteins interact, both the kinetic and structural information is needed. We study the interaction between transcription activators and coactivators. Transcription activation domains are disordered in solution and for unknown reasons show either interaction with only a few or with many distinct coactivators, resulting in structural changes in the activation domain. We have determined the interaction between a large set of transcriptional activators with coactivators in vitro.

Research proposal:

Before we could study the structural behaviour of these complexes we first need to confirm these interactions and identify minimal interaction domains. We have previously established a high throughput robot controlled in vitro interaction assay. In this project this method will be used to study the interaction of a few complexes (CHD7-CTCF;, MDM2-p53, cMyb-p300) in detail by performing various techniques. First various deletion constructs will be designed, and interaction will be evaluated. Next site-directed mutagenisis of evolutionary conserved residues will be performed to determine the key residues in both domains. Furthermore efforts will be made to develop a genetic screen using a random mutagenesis screen (using e.g. UV). Although challenging, a new interaction method could be developed: a bacterial two-hybrid assay permitting both interaction verification and identification of novel interactions. Finally identified interactions will be verified using various biophysical techniques including thermofluor anlysis, NMR and possibly mass spectrometry.

Methods:

• PCR
• Cloning
• Protein (co-)expression
• Protein purification in small scale using robotics
• Large scale protein purification using HPLC
• HTP protein interaction assays
• Site-directed mutagenisis
• Two hybrid assays
• Bacterial two-hybrid assay
• Biophysical characterisation of proteins using fluorescence techniques

Functional characterisation of DNA repair protein complexes

contact person:

G.E. Folkers
Bloembergen NMR building, phone:030-2539930

Scientific background:

During DNA replication, recombination and DNA repair, double-stranded DNA frequently forms three- or four-way junctions, bubbles, flaps or broken ends with single-stranded extensions. These irregular structures must be corrected to maintain genome stability, integrity and fidelity. This task is accomplished by structure-specific endonucleases. Inactivation or malfunctioning of these enzymes causes genetic defects or cancer. The XPF family, is such a structure-specific endonuclease. In humans seven members (XPF, MUS81, ERCC1, EME1, EME2, FANCM and FAAP24) have been characterized by the presence of the ERCC4 nuclease domain. Only two of them however, XPF and MUS81, have nuclease activity mediated by the conserved core nuclease motif (ERKX3D). Their catalytic function however depends on heterodimer formation with the non-catalytic family members. For example XPF forms an obligate heterodimeric complex with ERCC1 but not with other members of the family and functions primarily in the Nucleotide Excision Repair (NER) pathway. This versatile pathway is able to detect and remove a variety of bulky DNA lesions induced by UV light and environmental carcinogens and thereby maintains genome integrity. We have previously determined the structure of various domains within this family and characterised how these proteins heterodimerize, how they bind to DNA but the catalytic reaction remains elusive and what determines the preferential heterodimer formation is also not known. Recently we have implemented an in vitro assay that functionally resembles the DNA repair reaction and assays to study protein-protein and protein-DNA interactions. These assays, in cooperation with the structural information available, permit us to study the structure-function relationship between the ability to form heterodimers, bind to DNA and perform catalysis.

Research proposal:

On the basis of available structural information and functional models design mutants in the catalytic domain of XPF, the interaction domain between ERCC1 and XPF and the DNA binding domains of ERCC1 and XPF. Clone these mutants and express the protein in a suitable expression hosts. After purification the ability to bind to DNA, to interact with the interaction partner, to perform incision reactions will be determined. For a few mutants biophysical characterisation will be performed using fluorescence methods and NMR spectroscopy. Finally using the obtained information a structural model will be made to explain the biochemical observations.

Methods:

• PCR
• Cloning
• Protein (co-)expression
• Large scale protein purification using HPLC
• Radio nuclide experiments
• Protein-protein interaction assays
• Protein-DNA interaction assays
• Site-directed mutagenisis
• DNA incision assays
• Biophysical characterisation of proteins using NMR spectroscopy
• Biophysical characterisation of proteins using fluorescence techniques