Tuberculosis remains a leading cause of death worldwide due to infection by Mycobacterium tuberculosis (Mtb). The continuing global burden of disease underscores the need for new therapeutic strategies and novel mechanisms of action to inhibit Mtb growth and improve patient outcomes. Structural approaches that reveal ligandable sites on essential bacterial proteins can inform development of new antibiotics or chemical probes.
FtsZ is an essential bacterial GTPase that polymerizes into protofilaments at the division site of a replicating cell and coordinates septation to produce two daughter cells. Because of its central role in bacterial cytokinesis, FtsZ is an attractive target for anti-infective drug discovery. Structural knowledge of where small molecules can bind FtsZ — including sites that are conformationally dynamic — is valuable for guiding medicinal chemistry efforts.
The study describes an X-ray crystallographic fragment screening campaign against MtbFtsZ. A total of 1,070 crystals were soaked with fragment compounds, and diffraction data were collected. Of those crystals, 714 datasets were selected for downstream analysis. The large-scale soaking and data-collection effort provided the dataset necessary to search for low-occupancy fragment binding across many parallel crystals.
The authors used PanDDA, a multi-dataset analysis method designed to detect weak or partial-occupancy ligand-binding events by comparing many related electron-density maps. PanDDA-generated event maps were produced from the dataset pool; 149 datasets exhibited event-map density considered sufficient to support modeling of fragment binding. These event densities represent candidate ligand-binding occurrences identified across the crystallographic screen.
From the fragment-screening results, the authors describe 15 novel ligand-binding sites on MtbFtsZ. The preprint reports these newly observed pockets as potential starting points for structure-guided elaboration. The identified sites expand the map of possible small-molecule interaction locations on the protein and may include pockets that are only apparent in certain conformations or under fragment-bound conditions.
The crystallographic models captured both major conformational states of FtsZ described in the study, namely the ON and the OFF conformations. Notably, both conformations were present together within the asymmetric unit of the crystals. The co-existence of these states in the crystal lattice provides direct structural evidence of conformational heterogeneity and offers the opportunity to relate fragment binding to specific FtsZ conformations.
The dataset revealed asymmetric binding of fragments to different chains in the crystallographic models. That is, fragment molecules were observed to bind to one chain in a multimeric asymmetric unit but not necessarily to the corresponding site on another chain. This asymmetric occupancy highlights the conformationally specific nature of some fragment interactions and suggests that ligand binding may preferentially stabilize particular chain states or conformers.
The fragment hits and mapped binding sites provide routes for medicinal-chemistry follow-up. The authors note that the crystallographic fragment-binding data offer opportunities for fragment growing and fragment merging strategies to develop initial fragments into higher-affinity, more drug-like molecules or into chemical probes that are specific for particular FtsZ conformations. These structure-based optimization approaches rely directly on the binding-mode information captured in the PanDDA event maps and resulting models.
This work is reported as a preprint and has not been certified by peer review. Funding information declared in the preprint includes support from the National Institutes of Health (P30GM133893) and the United States Department of Energy. Supplementary materials and links to data and code are indicated in the preprint record; the authors also reference an RCSB group summary entry. No competing interests were declared by the authors in the source.
Note: All factual statements above are drawn from the source preprint. The preprint provides the numbers of crystals, datasets analyzed, PanDDA event hits, and the count of 15 novel sites, and reports the presence of ON and OFF conformations and asymmetric fragment binding. Details beyond those reported in the source (for example, chemical structures of fragments, detailed pocket descriptions, or quantitative binding affinities) were not reported in the source document and are not included here.