The NovInDXS project started with establishing the first DCC experiment for identification of novel ligands for the anti-infective target DXS. For this study, we used acylhydrazone formation as a reversible reaction of choice and the DXS enzyme from Deinococcus radiodurans (drDXS) as a target protein.
To maximise the chances of finding potent ligands for DXS, we decided to use heterocyclic building blocks for acylhydrazone formation in DCC, as they are fundamental structural motifs in a plethora of bioactive compounds. The initial choice of heterocyclic building blocks was based on their structural similarities with the existing hits and/or with the cofactor thiamine diphosphate (ThDP) of the enzyme DXS. The molecular docking study of all possible acylhydrazones in the active site of the enzyme drDXS suggested that these products are accommodated in the active site of the enzyme drDXS.
We performed an adaptive tdDCC experiment (DCC-1) by reacting several aldehydes with hydrazides in the presence (protein-templated) and absence (blank) of the target drDXS. The comparison of these tdDCC experiments revealed the amplification of five acylhydrazone products in the protein-templated DCC. These five hits showed moderate binding affinity (KD) ranging from 55−270 μM for the enzyme drDXS and 50−152 μM for the truncated homologue from Mycobacterium tuberculosis dmtDXS. Instead of a traditional medicinal-chemistry approach, we decided to employ tdDCC to optimise these hits. We assumed that using the common structural motifs from the first hits as building blocks in the second round of tdDCC (DCC-2), along with new aldehyde and/or hydrazide counterparts would provide better chances of identifying improved hits. Using this innovative approach we were able to identify five new hits with improved binding affinities (KD, = 7–150 μM for drDXS and 10−110 μM for dmtDXS). To further support our hypothesis that ‘tdDCC can be used for the optimisation of inhibitors/hits’, we performed the third round of tdDCC (DCC-3 and DCC-4) using the same approach. We aimed at studying to some extent the traditional medicinal chemistry approach driven by structure−activity relationships, by including a range of heterocyclic building blocks to cover a wide chemical space with good structural diversity. On the contrary to the traditional medicinal chemistry approach, where all possible acylhydrazone products from these three DCLs should be synthesised and tested for their biological activity, we let the protein select its best binders and synthesised only the amplified derivatives. When tested the eleven hits from these tailored libraries for their binding affinities for drDXS and dmtDXS, most of the hits show substantial improvements in binding affinity compared to the hits from the first and second round of tdDCC experiments. Five hits showed single-digit micromolar affinities (2–8 μM) for drDXS. Similarly, two hits showed 6 and 7 μM binding affinity for dmtDXS. Furthermore, we tested these hits for their enzymatic activity against drDXS and selected hits against Mycobacterium tuberculosis DXS, as well as evaluated their antibacterial activity against Escherichia coli TolC. We observed significant improvement in enzymatic activity of hits from IC50 value of 51 ± 3 μM in the first round of DCC to IC50 = 34 ± 4 μM in the third round of DCC, and antibacterial-activity from no inhibition in the first round of DCC to MIC value of 14 ± 4 μM in the third round of DCC.
We also confirmed the binding mode of the identified hits by molecular docking, competition assays with cofactor and substrates, and competition DCC experiments with cofactor and known substrate inhibitor. These findings suggest that most of the identified hits from DCL-1 are cofactor ThDP and substrate(s) competitors.
These results are published recently in the preprint [R. P. Jumde, et al,
https://doi.org/10.26434/chemrxiv.12681761.v1](opens in new window) and submitted for publication in an internationally renowned peer-reviewed scientific journal.
