The overall aim of the laboratory is to design innovative strategies in organic synthesis and bring them to bear towards the design of molecules with function valuable for chemical biology. The main three topics are: profile of enzymatic activity using small molecule microarrays, diversity oriented synthesis based on natural product scaffolds and template-accelerate synthesis.
Small molecule microarrays to screen for inhibitors and profile enzymatic activity in complex proteomes
A cell's ability to rapidly turn on and off protein function at the post-translational level has been known for a long time. However, robust methods to measure enzymatic activities on a proteomic scale are still lacking. Amongst post-translationally regulated enzymes, proteases, kinases and phosophatases play crucial roles in signal transduction and key regulatory mechanisms. There is thus widespread interest in methods to correlate their activity to a phenotype in an analogous fashion to gene expression profiling. From a chemistry perspective, enzymatic activity can be measure with small molecule substrates. For this purpose, we are developing a method based on peptide nucleic acid (PNA) encoding which allows for libraries of small molecules present as a mixture in solution to be reformatted to a microarray (see adjacent figure). This technique has two advantages over microarrays prepared by direct spotting; first, the library can be used in solution prior to the readout (conversion to the microarray format) which may minimize non specific interactions of proteins with surfaces and second, it offers the possibility to select for transformed library members and as such, provides a detection method that would not be possible with spotted arrays (vide infra). In order to measure enzymatic activity, we use libraries of small molecules substrate designed to query the activity state of enzymes on a protein-class basis. We have shown that we could synthesize libraries of fluorogenic substrates based on rhodamine (see figure below) to profile proteases.[1] Libraries of substrates can be used to determine the preferred substrate of a given protease or, knowing the substrate fingerprint of a protease, to measure its activity in crude lysates. More significantly, we showed that this method was sensitive enough to measure the difference in proteolytic activity between crude cell lysates from healthy cells and apoptotic cells. Despite the complexity of the proteomic mixture, we could clearly identify the activity of caspase-3 in the apoptotic sample. In a more diagnostic orientated application, we showed that we could measure the difference in thrombin activity between serum from patients on anticoagulation therapy and healthy individuals. With regards to kinases, we have taken a similar approach making a library of substrates. For tyrosine kinases, detection of the phosphorylated peptides was achieved with a very specific antiphosphotyrosine antibody (see figure above) and we have shown that we could assess the preferred substrate of a given tyrosine kinase1 or evaluate the selectivity of a given inhibitor against a panel of kinases (unpublished results). While these preliminary results with kinases are encouraging, a more general method that would be applicable simultaneously to both tyrosine and serine/ threonine kinases is being developed. The method is based on the unique chemical reactivity of the phosphate group which is chemoselectively labeled. Importantly, this method also allows us to enrich the phosphorylated substrates thereby increasing sensitivity (signal to noise ratio).molecular diversity combined with selection and amplification will be explored.
Diversity oriented synthesis based on natural product scaffolds
It can be argued that secondary metabolites, having been selected by nature for a specific biological activity, are privileged scaffolds in diversity space. While the target oriented synthesis of natural products has been the testing ground for new methodologies in chemistry and has reach a certain maturity, the development of synthetic strategies which are amenable to diversification offers new challenges. The first family of natural products that we have targeted is the resorcyclic macrolide which was selected based on the high number of ATPase and kinase inhibitors amongst its natural members (there are 22 reported natural products with the resorcyclide scaffold, see below for selected examples).Radicicol was in fact shown to be a selective ligand for HSP90's ATP-binding pocket despite its lack of structural similarity to ATP. Interestingly, closely related family members such as LL-Z1640-2 and hypothemycin are potent kinase inhibitors. While it may seem suspicious that small structural differences such as the level of saturation at the benzylic position can account for a change of selectivity between different kinases, or that opening of an epoxide to a halohydrin can change the selectivity from an HSP90 inhibitor to a helicase inhibitor, we have shown that such small modifications lead to dramatic differences in the conformational landscape of the macrocycles. Thus, an interesting aspect of these macrocycles is that structurally
similar compounds can be topologically quite different. We have developed diversity oriented solid phase and polymer assisted syntheses of radicicol, pochonins, as well as aigialomycins based on the use of a thioether linker (see figure below). The tioether not only provides an attachement point to the resin but also offers rich chemistry to construct adjacent carbon-carbon bonds. In the case of radicicol and pochonin C, the thioether allowed the use of a Pummerer allylation to introduce the necessary susbtitution of the Weinreb amide while for aigialomycin D, the thioether linker further acidified the benzylic position thereby facilitating the alkylation chemistry. In both cases, masking the final olefin in the form of the thioether was also found to be important to avoid undesired conjugate addition in the case of radicicol and undesired ring closing metathesis in the case of aigialomycin D. For the purpose of molecular diversity, we have shown that the compounds could be released from the resin under oxidative conditions to induce an elimination or under reductive conditions to afford the corresponding alkane (see figure below). We are currently exploring the utility of the thioether linker to acces benzylic epoxide as present in hypothemycin and aigialomycin A as well as benzylic ketones.
Modified oligonucleotides are used to direct designed chemical reactions. The potential of such methodologies for the chemical evolution of molecular function through the generation of molecular diversity combined with selection and amplification will be explored.
A cell's ability to rapidly turn on and off protein function at the post-translational level has been known for a long time. However, robust methods to measure enzymatic activities on a proteomic scale are still lacking. Amongst post-translationally regulated enzymes, proteases, kinases and phosophatases play crucial roles in signal transduction and key regulatory mechanisms. There is thus widespread interest in methods to correlate their activity to a phenotype in an analogous fashion to gene expression profiling. From a chemistry perspective, enzymatic activity can be measure with small molecule substrates. For this purpose, we are developing a method based on peptide nucleic acid (PNA) encoding which allows for libraries of small molecules present as a mixture in solution to be reformatted to a microarray (see adjacent figure). This technique has two advantages over microarrays prepared by direct spotting; first, the library can be used in solution prior to the readout (conversion to the microarray format) which may minimize non specific interactions of proteins with surfaces and second, it offers the possibility to select for transformed library members and as such, provides a detection method that would not be possible with spotted arrays (vide infra). In order to measure enzymatic activity, we use libraries of small molecules substrate designed to query the activity state of enzymes on a protein-class basis. We have shown that we could synthesize libraries of fluorogenic substrates based on rhodamine (see figure below) to profile proteases.[1] Libraries of substrates can be used to determine the preferred substrate of a given protease or, knowing the substrate fingerprint of a protease, to measure its activity in crude lysates. More significantly, we showed that this method was sensitive enough to measure the difference in proteolytic activity between crude cell lysates from healthy cells and apoptotic cells. Despite the complexity of the proteomic mixture, we could clearly identify the activity of caspase-3 in the apoptotic sample. In a more diagnostic orientated application, we showed that we could measure the difference in thrombin activity between serum from patients on anticoagulation therapy and healthy individuals. With regards to kinases, we have taken a similar approach making a library of substrates. For tyrosine kinases, detection of the phosphorylated peptides was achieved with a very specific antiphosphotyrosine antibody (see figure above) and we have shown that we could assess the preferred substrate of a given tyrosine kinase1 or evaluate the selectivity of a given inhibitor against a panel of kinases (unpublished results). While these preliminary results with kinases are encouraging, a more general method that would be applicable simultaneously to both tyrosine and serine/ threonine kinases is being developed. The method is based on the unique chemical reactivity of the phosphate group which is chemoselectively labeled. Importantly, this method also allows us to enrich the phosphorylated substrates thereby increasing sensitivity (signal to noise ratio).molecular diversity combined with selection and amplification will be explored.
It can be argued that secondary metabolites, having been selected by nature for a specific biological activity, are privileged scaffolds in diversity space. While the target oriented synthesis of natural products has been the testing ground for new methodologies in chemistry and has reach a certain maturity, the development of synthetic strategies which are amenable to diversification offers new challenges. The first family of natural products that we have targeted is the resorcyclic macrolide which was selected based on the high number of ATPase and kinase inhibitors amongst its natural members (there are 22 reported natural products with the resorcyclide scaffold, see below for selected examples).Radicicol was in fact shown to be a selective ligand for HSP90's ATP-binding pocket despite its lack of structural similarity to ATP. Interestingly, closely related family members such as LL-Z1640-2 and hypothemycin are potent kinase inhibitors. While it may seem suspicious that small structural differences such as the level of saturation at the benzylic position can account for a change of selectivity between different kinases, or that opening of an epoxide to a halohydrin can change the selectivity from an HSP90 inhibitor to a helicase inhibitor, we have shown that such small modifications lead to dramatic differences in the conformational landscape of the macrocycles. Thus, an interesting aspect of these macrocycles is that structurally
similar compounds can be topologically quite different. We have developed diversity oriented solid phase and polymer assisted syntheses of radicicol, pochonins, as well as aigialomycins based on the use of a thioether linker (see figure below). The tioether not only provides an attachement point to the resin but also offers rich chemistry to construct adjacent carbon-carbon bonds. In the case of radicicol and pochonin C, the thioether allowed the use of a Pummerer allylation to introduce the necessary susbtitution of the Weinreb amide while for aigialomycin D, the thioether linker further acidified the benzylic position thereby facilitating the alkylation chemistry. In both cases, masking the final olefin in the form of the thioether was also found to be important to avoid undesired conjugate addition in the case of radicicol and undesired ring closing metathesis in the case of aigialomycin D. For the purpose of molecular diversity, we have shown that the compounds could be released from the resin under oxidative conditions to induce an elimination or under reductive conditions to afford the corresponding alkane (see figure below). We are currently exploring the utility of the thioether linker to acces benzylic epoxide as present in hypothemycin and aigialomycin A as well as benzylic ketones.
Modified oligonucleotides are used to direct designed chemical reactions. The potential of such methodologies for the chemical evolution of molecular function through the generation of molecular diversity combined with selection and amplification will be explored.
