Making the SepF-FtsZ interaction crystal clear

Revealing the essential dynamic interdependence of SepF and FtsZ for early Z-ring formation in Corynebacterium glutamicum

Like Comment
Read the paper

In the lab, we are interested in understanding the molecular mechanisms governing actinobacterial cell physiology. Actinobacteria represent one of the largest phyla on earth that include important human pathogens such as Mycobacterium tuberculosis and Corynebacterium diphtheriae but also species of industrial interest such as Streptomyces (that produce almost 2/3 of the antibiotics we use) or Corynebacterium glutamicum (in charge of the monosodium glutamate production (> 3 tons / year) vastly used as a food flavour enhancer). Despite their biotechnological and clinical importance, very little is known about the exact molecular mechanisms of many essential cellular functions, starting with one of the most fundamental: cell division.

It all started in 2016 when I obtained the international Pasteur-PPU PhD fellowship and joined the Alzari lab at the Institut Pasteur in Paris, where together with Anne Marie Wehenkel, we started to work side by side on actinobacterial cell division. FtsZ is the prokaryotic homologue of tubulin and plays a central role in orchestrating cytokinesis. Interestingly, many of the well-characterized FtsZ interactors in Escherichia coli or Bacillus subtilis were missing in Actinobacteria, among them FtsA and ZipA, which link FtsZ to the membrane. In B. subtilis, the deletion of ftsA can be rescued by the overexpression of a non-essential gene, sepF. As SepF is the only membrane-binding cell division protein known to directly interact with the conserved C-terminal domain of FtsZ (FtsZCTD), we decided to structurally and functionally characterize this protein in the actinobacterial model organism C. glutamicum.

Polymerization is a recurrent feature in cell division proteins, for instance, FtsZ polymerizes forming filaments that generate a ring -about the same diameter as the cell -marking the future division site at mid-cell. SepF also presents this trait, (i.e. B. subtilis SepF forms rings in vitro). In our study, we found that the presence of lipid membranes stimulates SepF polymerization producing tubulated structures on the surface of liposomes, whereas FtsZ filaments have the opposite effect.

Since structural information at atomic resolution of how SepF recognizes FtsZ was missing, we tried hard to obtain the crystal structure of the interacting complex. The structure revealed a hydrophobic pocket on SepF that accommodates the FtsZ C-terminus in an extended hook-like conformation. The 2:2 stoichiometry of the complex defined the SepF homodimer as the functional unit and explained why SepF is able to bundle FtsZ filaments. Last but not least, the crystal structure also revealed a small alpha-helix (α3) at the C-terminus of SepF that caps the interface that is supposed to be responsible for SepF polymerization, thus we suggest that it could serve as a mechanical switch modulating the SepF polymerization state.

After shedding light on some of the atomic details regarding SepF-FtsZ interaction, we aimed to understand the physiological role of SepF and how these two proteins interplay in vivo during cell division. Thus, I decided to take a train to Jülich (Germany) and spend almost a month in the Bott lab, where I learned how to genetically manipulate Corynebacterium. It was a lovely experience and I am really thankful to all the team for helping me out and sharing their knowledge. After screening many (many) colonies, I succeeded to replace the native promoter of sepF by a repressible one to be able to shutdown sepF expression. Under depletion conditions, cells became elongated – with branches along the later cell wall – and eventually died. In these cells, FtsZ localization was completely lost, indicating that FtsZ needs to interact with SepF to reach mid-cell and assemble a functional divisome. Intriguingly, previous work on Mycobacterium smegmatis had shown that in the absence of FtsZ, SepF also losses its mid-cell localization, thus uncovering an interdependence between the two proteins for the right placement at the future division site.

Our work combined biophysics data, structure-driven hypothesis and bacterial genetics to show that SepF is an essential component of the actinobacterial divisome. The protein has a complex role at the division site, participating in FtsZ membrane tethering, FtsZ bundling, and membrane remodeling. Our data suggest that SepF could act as a putative cell division checkpoint protein, participating in membrane constriction only once the divisome is fully functional. More importantly, this work opens new interesting questions: Why in the absence of SepF the cell poles are misplaced (branches)? Why does FtsZ present a non-random distribution in the same conditions, i.e. without its membrane tethering factor? What is the physiological role of SepF-α3? Does it modulate protein-protein interactions? We do not have clear answers for these questions. However, several observations point towards the existence of Actinobacteria-specific FtsZ interactors waiting to be discovered.

To see the full story check out the publication in Nature communications :

Adrià Sogues

Postdoc, VIB-VUB Center for Structural Biology

Postdoctoral researcher in the Remaut lab at VIB-VUB (Belgium), studying cell surface and its link with cell division. Using genetics, biochemistry, X-ray crystallography and electron microscopy. Formerly a PhD student in the Alzari lab (supervised by Anne Marie Wehenkel) at the Institut Pasteur studying the early steps of the divisome assembly.