Mapping the Contact Sites of the Escherichia coli Division-Initiating Proteins FtsZ and ZapA by BAMG Cross-Linking and Site-Directed Mutagenesis

Abstract

Cell division in bacteria is initiated by the polymerization of FtsZ at midcell in a ring-like structure called the Z-ring. ZapA and other proteins assist Z-ring formation and ZapA binds ZapB, which senses the presence of the nucleoids. The FtsZ–ZapA binding interface was analyzed by chemical cross-linking mass spectrometry (CXMS) under in vitro FtsZ-polymerizing conditions in the presence of GTP. Amino acids residue K42 from ZapA was cross-linked to amino acid residues K51 and K66 from FtsZ, close to the interphase between FtsZ molecules in protofilaments. Five different cross-links confirmed the tetrameric structure of ZapA. A number of FtsZ cross-links suggests that its C-terminal domain of 55 residues, thought to be largely disordered, has a limited freedom to move in space. Site-directed mutagenesis of ZapA reveals an interaction site in the globular head of the protein close to K42. Using the information on the cross-links and the mutants that lost the ability to interact with FtsZ, a model of the FtsZ protofilament–ZapA tetramer complex was obtained by information-driven docking with the HADDOCK2.2 webserver.

ZapA tetramerization is required for midcell localization and ZapB interaction in Escherichia coli

“…Instead ZapB localized polarly in these cells mimicking the pattern of cells without ZapA [7,15]. This sets ZapA I83E apart from the other non-complementing ZapA mutants that did localize to midcell and recruited ZapB [7]. The ZapB localization pattern is a good indication whether ZapA complements at midcell ( Figure S2 ).…”

“…To assess whether ZapA I83E would complement the elongated Δ zapA phenotype and restore wild-type morphology in vivo , a complementation experiment was performed. Expression of WT ZapA, ZapA I83E or a negative empty vector (EV) control was induced from plasmid with 50 μM IPTG in TB28 Δ zapA cells growing in rich medium for ~8 mass doubling as described before [7]. The cells were then fixed, imaged, and average cell lengths were analyzed.…”

“…FtsZ localized regularly at midcell for all cultures but also more diffusely for the Δ zapA and the ZapA I83E cells. Since ZapB midcell localization is dependent on ZapA [7,14] and both proteins are reported to oscillate together between the cell poles [15], it was expected to localize diffusely through the cells expressing dimeric ZapA I83E . Instead ZapB localized polarly in these cells mimicking the pattern of cells without ZapA [7,15].…”

“…Since ZapB midcell localization is dependent on ZapA [7,14] and both proteins are reported to oscillate together between the cell poles [15], it was expected to localize diffusely through the cells expressing dimeric ZapA I83E . Instead ZapB localized polarly in these cells mimicking the pattern of cells without ZapA [7,15]. This sets ZapA I83E apart from the other non-complementing ZapA mutants that did localize to midcell and recruited ZapB [7].…”

“…These effects were only observed for tetrameric ZapA that was able to interact with FtsZ [5]. ZapA can exist as a mixture of dimers and tetramers in vitro but in vivo it is likely to be mostly tetrameric due to molecular crowding conditions [3,6,7]. In fact, the available ZapA structures and in vitro cross-linking showed a tetrameric structure and tetramerization has already been suggested to be required for FtsZ bundling [6,7,8,9].…”