Back

Enabling antibiotic research: towards selective peptide deformylase inhibitors

Bachmann, N.; Schultz, A.; Zouatom, M.; Zhou, T.; Degenhart, C.; Koch, U.; Schäkermann, S.; Hüning, J.; Heinrich, S.; Dal Molin, M.; Zischinsky, M.-L.; Tourel, S.; Klebl, A.-K.; Rybniker, J.; Klebl, B. M.; Scherkenbeck, J.; Bandow, J. E.

2025-09-18 microbiology
10.1101/2025.09.18.677056 bioRxiv
Show abstract

Peptide deformylase plays a crucial role in prokaryotic translation and constitutes an antibiotic target previously addressed in clinical trials. In eukaryotes, mitochondrial translation also relies on peptide deformylase, necessitating antibiotic development to aim for selective inhibition of the bacterial enzymes. In the present study, we investigated two compound series: derivatives of actinonin and compounds containing a 5-bromoindole scaffold. Antibacterial activity was evaluated by microdilution-based minimal inhibitory concentration assay and selectivity investigated using human peripheral blood mononuclear cells. In vitro peptide deformylase inhibition was compared for the Escherichia coli and human enzyme. To validate peptide deformylase inhibition in vivo, a mass spectrometric analysis directly coupled to the minimal inhibitory concentration assay was developed for the model organism Bacillus subtilis. Two compounds originating from this work (ZHO-119, ZHO-197) showed antibacterial activity comparable to actinonin, and for the comparator compound BB-3497 superior anti-gram-negative and anti-tubercular activity was confirmed. The three compounds displayed no cytotoxicity and were equally selective in vitro for the bacterial enzyme. The mass spectrometry-based analysis indicates that in addition to peptide deformylase, ZHO-197 very effectively inhibits bacterial methionine aminopeptidase, the metallo-enzyme that removes the deformylated N-terminal methionine.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.