Name: Agnes Rinaldo-Matthis
PI: Nordlund (SU)
Project: Structural studies on protein phosphatases
Gender: F
Mentors: A. Liljas, Ladenstein
Actual starting date: 1 February, 1997
Administrative starting date: 1 January, 1997
In vitro evolution as a tool for crystallising proteins
Many of the most interesting structural projects are stalled, or severely slowed down, by the problems of obtaining protein samples which yield high quality crystals. General approaches for the successful preparation of such samples are therefore of utmost importance.
The minimal requirements for the crystallisation of a protein are; 1) a highly structurally homogeneous protein sample which is: 2) soluble at relatively high protein concentrations. The success of the crystallisation experiments, and quality of the crystals obtained thereof, are in most cases highly correlated to 1) and 2).
The solubility of proteins can be particularly bad in two cases: 1) when only single domains/subunits have been expressed of a multi-domain/subunit protein resulting in relatively hydrophobic protein-protein interaction surfaces being exposed to the solvent and 2) highly hydrophobic trans-membrane proteins (1). The present revolution in protein engineering techniques has the potential to provide solutions to these problems (2,3).
Two main concepts, which could be viewed as general approaches to the crystallisation problem have already been applied in a few cases. These are: 1) the addition of extra charged residues on predicted surface loop structures using site directed mutagenesis, in order to make the protein more soluble and 2) the generation of antibody Fab/Fc complexes, giving more soluble and stable proteins as well as allowing the purification of more structurally homogeneous samples (the antibody can often differentiate between different conformers/modifications of a protein).
However in the cases where these concepts have been used so far, the powerful combinatorial chemistry and in vitro evolution techniques made available in recent years have not been applied. Random mutagenesis and DNA-shuffling methods have now allowed the direct evolution of e.g. a significantly more soluble green fluorescent protein (2). Combinatorial protein chemistry, combined with phage display or ribosome based display, has allowed the direct evolution of antibodies/affibodies and peptides, selected at controlled conditions for binding, to specific forms of proteins (2,3). In the present project we are implementing/developing such techniques, to allow the efficient structure determinations of problematic proteins.
We are using two "model systems" for these developments. One is the protein phosphatase SpoIIE, involved in regulating asymmetric differentiation during sporulation (4). Three different expression constructs have been made for SpoIIE containing one or two of the soluble domains of this membrane linked protein. Presently SpoIIE has a very low solubility and can only be obtained as inclusion bodies and refolded. The other "model system" is photosystem II (PSII), the water splitting complex of the PSII membrane protein complex, which has turned out to be notoriously difficult to crystallise (5). Structural information on both of these proteins should be of great interest.
Agnes Rinaldo-Matthis, the student financed by SBNet, is applying the phage display techniques and in order to obtain phage display antibodies selected at suitable conditions. A rabbit was immunised with a small peptide containing a sequences from the D1 protein of PSII. The rabbit library has been cloned into a M13 phage and contains some 105 different genes. A mouse was immunised with the whole core of PS II. The spleen of the mouse was harvested and a library of PCR fragments was amplified. Ligations of the library in to the phage system is under way.
The rabbit library, known to contain excellent antibodies recognising the 3-d structure of PSII, will be used as a model system to optimise selection conditions. The mouse library, potentially containing clones to different epitopes and forms of PSII, will be used to select antibodies directed to the different sides of the membrane as well as different phosphorylated and complex forms of PSII. Naive (non-immunised) antibody or affibody phage libraries will in the next step be used. Biotinylated PSII had aslso been prepared for the selection steps. This work has not yet led to any publications.
1. Ostermeir, C. and Michel, H., Curr. Op in Struc. Biol., 5, 697-701 (1996).
2. Patten P.A., Howard, R.J. and Stemmer, W.P., Curr. Opin in Biotech.,8, 724-743 (1997).
3. Nygren, P.A. and Uhlen, M., Curr. Opin. in Struct. Biol., 7,463-470 (1997).
4. L. Duncan, S. Alper, F. Arigoni, R. Losick, P., Science, 270, 641-4 (1995).
5. Rutherford A.W., TIBS, 14, 227-232 (1993).
Latest update at 9 March, 1998.