TL;DR
Researchers have identified novel inhibitors for a key enzyme in Mycobacterium tuberculosis (Mtb) essential for its survival. By combining fragment-based drug discovery with CRISPR interference, they've validated this enzyme as a promising drug target and developed a lead compound with on-target activity against Mtb.
What Problem Does This Paper Address?
Tuberculosis (TB) remains a major global health threat, exacerbated by the rise of multidrug-resistant strains. The development of new drugs with novel mechanisms of action is crucial. The coenzyme A (CoA) biosynthesis pathway in Mtb has been identified as a potential target for new TB therapies. Specifically, the enzyme phosphopantetheine adenylyltransferase (MtbPPAT), which catalyzes a critical step in CoA synthesis, is being investigated. While this enzyme is essential for Mtb growth, no inhibitors had been reported, and its viability as a drug target needed further validation.
Methods
The study employed a multi-pronged approach:
- Fragment-Based Drug Discovery (FBDD): A library of 1265 small molecules ("fragments") was screened against MtbPPAT using a cascade of biophysical techniques, including fluorescence-based thermal shift assays, saturation transfer difference (STD) NMR, and WaterLOGSY NMR. This identified initial "hits" that weakly bind to the enzyme.
- X-ray Crystallography: Co-crystal structures of MtbPPAT bound to identified fragments were determined. This provided crucial atomic-level information about how the fragments interact with the enzyme's active site, revealing distinct binding pockets.
- Fragment Linking: Based on the crystal structures, fragments binding in close proximity were chemically linked together using flexible ether-based linkers. This strategy aimed to combine the binding affinities of individual fragments to create more potent inhibitors.
- CRISPR Interference (CRISPRi): To validate the enzyme as a drug target and confirm the on-target activity of the developed compounds, a CRISPRi system was used to genetically silence the coaD gene (which encodes MtbPPAT) in Mtb. This created a "knockdown" strain that was then tested for susceptibility to the lead compound.
Datasets
Not explicitly discussed as "datasets" in the traditional machine learning sense. The study utilized:
- A library of 1265 rule-of-three compliant fragments for screening.
- Existing knowledge of the MtbPPAT active site and natural substrates (ATP and phosphopantetheine).
- Crystal structures of apo MtbPPAT and MtbPPAT complexed with various fragments and synthesized compounds.
- A genetically engineered Mtb strain with a coaD gene knockdown.
Key Insights and Ideas
- MtbPPAT as a druggable target: The study successfully identified novel, non-natural ligands for MtbPPAT, demonstrating its potential as a target for anti-TB drug development.
- Fragment linking is effective for MtbPPAT: By strategically linking fragments that bind to adjacent sites in the MtbPPAT active site, researchers were able to significantly improve binding affinity and create potent inhibitors.
- Distinct binding pockets: The X-ray crystallography revealed that different fragments bind to distinct regions within the MtbPPAT active site, including both the phosphopantetheine and ATP binding sites, offering diverse opportunities for inhibitor design.
- CRISPRi as a powerful validation tool: The use of CRISPRi to silence the coaD gene provided strong evidence that MtbPPAT is essential for Mtb survival and that the developed compound acts on-target.
Key Results
- A fragment screen identified three chemically distinct fragment hits that bind to MtbPPAT.
- X-ray crystallography revealed the binding modes of these fragments, showing they occupy different regions of the active site.
- Fragment linking led to the development of compound 20, which showed a ~65-fold increase in affinity compared to its parent indole fragment, with a K<sub>D</sub> of 15 ± 4 μM.
- Compound 20 demonstrated on-target anti-proliferative activity against Mtb in vitro, as confirmed by CRISPRi-mediated knockdown of the coaD gene. The coaD knockdown strain showed enhanced susceptibility to compound 20.
Limitations and Future Work
- Fragment affinity: The initial fragments had very low affinities, making quantitative binding measurements challenging.
- Linker flexibility: The flexible ether linker in compound 20, while effective, might not be optimal for maximizing affinity. Future work could involve rigidifying the linker to reduce the entropic cost of binding.
- Specificity: While the study focused on MtbPPAT, further work may be needed to ensure specificity against human orthologs.
- Further optimization: Future studies will focus on modifying the phenyl and indole moieties of the inhibitors to enhance interactions with surrounding amino acid residues and potentially link with the benzophenone fragment for improved specificity. The authors also suggest exploring modifications to the linker to improve affinity.
Summary updated