Final Activity Report Summary - ARTIZYME CATALYSIS (Development of late transition metalloenzymes and metallo-DNAzymes for highly efficient catalytic processes)
We aimed to create artificial metalloenzymes, using the molecular recognition properties of proteins, for the use in well-defined and engineered catalytic transformations. After engineering of the protein sequence, the proteins were site-specifically functionalised with phosphine-ligands and bound to transition metals. Adjusting the structures of the proteins and the phosphine-ligands was expected to optimize the catalytic performance of these artificial enzymes.
Selective functionalisation of Photoactive Yellow Protein has been achieved by reacting the protein with activated carboxylic acid derivatives of diphenylphosphine-ligands. Binding a palladium-precursor to the ligand prior to coupling of the ligand to the protein does yield a palladium-ligand-protein complex. Characterisation by mass-spec and NMR shows the palladium is coordinated to the phosphine-ligand. This complex is active in the allylic amination of diphenylallyl acetate with benzylamine. Unfortunately this method proved not be generally applicable for other proteins.
In our search for a general and robust method for protein modification, we developed a two-step method for the introduction of free phosphine-ligands in protein hosts, which shows excellent efficiency and selectivity for all the proteins tested. The method is based on the formation of a hydrazone linkage between a hydrazine- and an aldehyde moiety. The hydrazine moiety is introduced via a maleimide-group, using the commercially available maleimide propionic acid hydrazide, which is subsequently allowed to react with an aldehyde-functionalized phosphine.
Using this approach, we have introduced several phosphines in structurally very different proteins. This approach is the first method for the site-selective covalent introduction of free phosphine-ligands in structurally diverse proteins and thus constitutes a major breakthrough in the development of artificial metalloenzymes. With this tool in hand, the stage is now set to explore all our proteins in asymmetric catalysis.
A second approach involves modification of ligand containing oligonucleotide strands combined with rational design of the ligand structure, ultimately yielding transition metal catalysts, tailor-made for the desired selective conversion. DNA is a chiral molecule that can be engineered to form highly selective binding cavities for a vast number of different molecules.
In the method employed by our group the modified nucleotide introduced into DNA strands is 5-iodo-2-deoxyuridine. This nucleotide can be selectively modified with a phosphine, which in turn can be coordinated to a transition metal. Phosphine-modified mono- and trinucleotides synthesised this way have been tested in asymmetric palladium catalysed allylic amination. When the monodentate ligand dppdU was used enantomeric excesses up to 82% were obtained. The solvent significantly influenced the selectivity, e.g. in THF the absolute configuration is S whereas in DCM it is R). Unfortunately, the trinucleotides dAdppdUdT and dCdppdUdG induced lower selectivity. A phosphine modified nonanucleotide was also obtained in this way, but the resulting compound proved to be very oxygen sensitive.
Therefore, we decided to modify 5-iodo-2-deoxyuridine with different spacers prior to modification with a phosphine. Amine and alkyne linkers were introduced via routine coupling methods. Using these modifiers we are able to introduce several transition metal phosphine complexes into oligonucleotides enabling the exploration of DNA based transition metal catalysis.
A DNA 15mer containing the central modifier monomer, synthesised using a standard DNA synthesiser, was coupled to diphenylphosphinobenzoic acid using EDC as activating agent. Presently, the coupling procedure is being optimised and the catalytic performance will be tested.
Selective functionalisation of Photoactive Yellow Protein has been achieved by reacting the protein with activated carboxylic acid derivatives of diphenylphosphine-ligands. Binding a palladium-precursor to the ligand prior to coupling of the ligand to the protein does yield a palladium-ligand-protein complex. Characterisation by mass-spec and NMR shows the palladium is coordinated to the phosphine-ligand. This complex is active in the allylic amination of diphenylallyl acetate with benzylamine. Unfortunately this method proved not be generally applicable for other proteins.
In our search for a general and robust method for protein modification, we developed a two-step method for the introduction of free phosphine-ligands in protein hosts, which shows excellent efficiency and selectivity for all the proteins tested. The method is based on the formation of a hydrazone linkage between a hydrazine- and an aldehyde moiety. The hydrazine moiety is introduced via a maleimide-group, using the commercially available maleimide propionic acid hydrazide, which is subsequently allowed to react with an aldehyde-functionalized phosphine.
Using this approach, we have introduced several phosphines in structurally very different proteins. This approach is the first method for the site-selective covalent introduction of free phosphine-ligands in structurally diverse proteins and thus constitutes a major breakthrough in the development of artificial metalloenzymes. With this tool in hand, the stage is now set to explore all our proteins in asymmetric catalysis.
A second approach involves modification of ligand containing oligonucleotide strands combined with rational design of the ligand structure, ultimately yielding transition metal catalysts, tailor-made for the desired selective conversion. DNA is a chiral molecule that can be engineered to form highly selective binding cavities for a vast number of different molecules.
In the method employed by our group the modified nucleotide introduced into DNA strands is 5-iodo-2-deoxyuridine. This nucleotide can be selectively modified with a phosphine, which in turn can be coordinated to a transition metal. Phosphine-modified mono- and trinucleotides synthesised this way have been tested in asymmetric palladium catalysed allylic amination. When the monodentate ligand dppdU was used enantomeric excesses up to 82% were obtained. The solvent significantly influenced the selectivity, e.g. in THF the absolute configuration is S whereas in DCM it is R). Unfortunately, the trinucleotides dAdppdUdT and dCdppdUdG induced lower selectivity. A phosphine modified nonanucleotide was also obtained in this way, but the resulting compound proved to be very oxygen sensitive.
Therefore, we decided to modify 5-iodo-2-deoxyuridine with different spacers prior to modification with a phosphine. Amine and alkyne linkers were introduced via routine coupling methods. Using these modifiers we are able to introduce several transition metal phosphine complexes into oligonucleotides enabling the exploration of DNA based transition metal catalysis.
A DNA 15mer containing the central modifier monomer, synthesised using a standard DNA synthesiser, was coupled to diphenylphosphinobenzoic acid using EDC as activating agent. Presently, the coupling procedure is being optimised and the catalytic performance will be tested.