|
The "domain concept" of the Max-Planck-Institute for Molecular Physiology increases efficiency in the search for drug candidates.
A team of scientists led by Prof. Waldmann of the Dortmund Max-Planck-Institute for Molecular Physiology has developed a new concept for a more efficient search for drug candidates. The so called "domain concept" aims at the generation of compound libraries with a significantly increased hit rate. The concept is based on structurally conserved yet genetically mobile protein domains and corresponding natural products selected during evolution. The researchers detail and support their hypothesis in several articles in the internationally leading journal "Angewandte Chemie" (see for instance Angew. Chemie Int. Ed. 2002, 41, 307-311).
After the deciphering of the entire human genome the worldwide hunt is on for genes that hold the promise for biomedical success stories. Such success stories can be expected when the genes code for proteins which are promising drug targets. A multitude of new potential protein targets for medical therapies is hidden and slumbering in the genome. But how does one rapidly and efficiently find the right drug candidate? This endeavour resembles the search for a needle in a hay-stack. Currently, some 10,000 chemical compounds have to be synthesized for each molecular target to finally hit a hot spot. The researchers of the Max-Planck-Institute for Molecular Physiology want to enhance this search by means of the combinatorial chemistry of evolutionary selected and therefore biologically relevant natural products. In the currently pursued approach, the target molecules are dissected theoretically into individual building blocks. During the actual chemical synthesis these building blocks are varied widely and linked to each other in such a way that as many combinations as possible are created. Thereby, so-called "compound libraries" with thousands to hundreds of thousands and in some cases even millions of chemical compounds can be generated. Up to now the major criteria for the design of the libraries and the planning of the syntheses were chemical feasibility, accessibility of suitable building blocks and robustness and reliability of synthesis methods, which should not fail even if applied many thousand fold. However, the intensive combination of chemical building blocks alone in a sort of "number game" does not solve the core problem. In too many cases the hit rates in the subsequent biological tests are disappointingly low. This indicates that the simple generation of as many compounds as possible is not decisive. Rather, a sufficiently large number of the "right" drug candidates has to be synthesized that are relevant to the corresponding biological system.
|
Picture 1: Modular structure of multi-domain proteins
Picture: MPI of molecular
Physiology/Schulte/Herter
The concept is based on the realization that Nature builds the targets for the drug candidates - that is the proteins - from modules. Proteins are composed of protein domains which often are structurally conserved but genetically mobile and which are repeatedly used in the course of evolution. In the assembly of new functional proteins Nature goes back to modules available from a "building set" (Figure 1). Consequently within the same but also among different species many proteins are found which often harbour very similar domains while they have different functions. While the similarity may be high, usually the differences between the domains are pronounced enough to allow for selective binding of particular partner molecules, so-called ligands. This is a prerequisite for the development of drugs.
The researchers furthermore realized that also in naturally occurring ligands for proteins, that is in biologically active natural products, certain underlying structures occur repeatedly. For this reason the Max-Planck researchers propose to employ the underlying structure of such natural products, which obviously were selected during evolution for binding to particular protein domains, as basis for the design and synthesis of compound libraries. Such libraries should yield biologically active "hits" with a frequency that is significantly increased when compared to the conventional hit finding process that is built in particular on "combining numbers" and chemical feasibility.
Picture 2: Combinatorial synthesis of a compound library. The target compound is dissected into individual building blocks which are then combined with each other in different combinations on a solid support. The dashed lines indicate the theoretical cuts, the spheres symbolize the support on which the synthesis is carried out. It can be viewed at http://www.mpg.de/news02/news0204_bild2.gif
Picture: MPI of molecular
Physiology/Schulte/Herter
The "domain concept" regards a natural product class as a starting point for the drug finding process and the development of compound libraries. This starting point has been confirmed by evolution and is therefore biologically relevant. The scientists now present their concept in three studies in the internationally leading science journal "Angewandte Chemie". The authors describe the fundamental idea of the "domain concept" employing a natural inhibitor of the protein phosphatase Cdc 25 as example (Angew. Chem. Int. Ed. 2002, 41, 307-311). In this study they could successfully overcome one of the major hurdles in the realization of the concept. They delineated a compound library from a complex natural product and synthesized the library in multistep and a very demanding synthetic sequence (Figures 2 and 3). The protein phosphatase Cdc 25 is decisively involved in the control of the cell cycle and is regarded as a starting point for the development of new anti-tumour drugs.
In the second study the synthesis of a small compound library is described which is based on a natural inhibitor of the protein tyrosine kinase Her-2/Neu. This library yielded inhibitors for the Tie-2, the VEGFR-3 and the IGF1 receptor tyrosine kinases with a high hit rate. Tie-2 and VEGFR-3 play key roles in the development of new blood and lymph vessels and also offer new promising starting points for tumour therapy (Angew. Chem. Ind. Et. 2002, in press). A third publication which is in press details the concept (Angew. Chem. Ind. Et. 2002, in press).
Picture 3: Synthesis of a dysidiolide library on a solid support. By means of a complex multi-step solid phase synthesis a library of analogs of the Cdc25-protein phosphatase inhibitor dysidiolide was generated. This synthesis proves that natural products and compound libraries derived therefrom can be synthesized on polymeric supports. The Figure shows electron microscopic photographs of polymer beads and tumour cells. The tumour cells floating in front of the polymer beads were used in a cytotoxicity assay. The result of the assay is visible on the also shown microtiter plates to the naked eye: living cells convert a yellow into a violet dye. The Figure also shows the structure of the target protein Cdc25.
Picture: MPI of molecular
Physiology/Schulte/Herter
As the next step the Dortmund scientists will analyse the major classes of protein domains and their natural ligands. Additionally, they will search among the known classes of biologically active natural products for basic repeatedly occurring structural motifs.
On the basis of these analyses they intend to design and produce new compound libraries, to substantiate and refine their concept and to search for new drug candidates.
The Dortmund researchers also introduced "combinatorial natural product chemistry" into the foundation of a "start-up" company. Herbert Waldmann and Alfred Wittinghofer from the Max Planck Institute for Molecular Physiology Dortmund, Walter Birchmeier from the Max Delbrück Center Berlin, Haus Bos and Hans Clevers both from the University Medical Center Utrecht and the experienced pharma manager Rian de Jonge founded the company Semaia Pharmaceuticals (www.semaia.com).
Semaia unites tumour biology, structural biology and medicinal chemistry under one roof and develops drugs for diseases related to biological signal transduction, in particular cancer.
Contact:
Prof. Herbert Waldmann
Max Planck Institute of molecular Physiology, Dortmund
Phone: 02 31 / 1 33 - 24 00
Fax: 02 31 / 1 33 - 24 99
E-Mail: herbert.waldmann@mpi-dortmund.mpg.de
Journal
Angewandte Chemie