Super-resolution imaging of bacterial division – DIVinHD

We have combined genetics, biochemistry, structural biology and microscopy methods to investigate the assembly, the positioning and the PG synthesis activity of protein machineries dedicated to cell elongation and division. Our most remarkable methodological development relies on biorthogonal PG labeling coupled to dSTORM (direct STochastic Optical Reconstruction Microscopy) imaging and in silico simulation. More specifically, we incorporated clickable probes into the peptide chain of the newly synthesized PG and conjugated them to small fluorescent dyes. Labeled PG regions were observed by dSTORM and their dimensions were measured to obtain the kinetics of PG synthesis and remodeling. Eventually, we created a geometrical model of the elongation and division of an ovoid cell, in which we added a labeling event to interpret in 3D the experimental patterns observed in 2D. This model, which is governed by parameters extracted from our experimental data, allowed identifying the most relevant hypotheses regarding the dynamics of PG synthesis and remodeling along S. pneumoniae cell cycle.

The study published in (Hosek et al., 2020) shows that the majority of the cytoplasmic domain of the MapZ protein, which is a substrate of the Ser/Thr kinase StkP, is intrinsically disordered. However, two regions could form helical secondary structures upon interaction with the major division protein FtsZ and/or with membrane lipids. MapZ interacts with FtsZ filaments and its monomer, but has no effect on protein polymerization. We also showed an interaction between MapZ and liposomes, which was stronger under experimental conditions mimicking its phosphorylation. These observations suggest that the recruitment and stability of the PG machineries at the division site involve not only protein interactions, but also interactions with membrane lipids and StkP-mediated phosphorylation.
The work by (Jacq et al., 2018) indicates that Pmp23 has a lysozyme activity that is important for S. pneumoniae morphogenesis and division. We also showed that the active form of Pmp23 is required for MapZ and FtsZ localization, but not for PG synthesis activity. Pmp23 localizes to midcell, where it interacts directly with the PG synthase PBP2x. Finally, inactivation of the hydrolase activity of Pmp23 increases the proportion of complex peptide crosslinking. These observations support the hypothesis that Pmp23 degrades glycan chains incorrectly incorporated by PBP2x into the PG network. Uncorrected, these incorporation «errors« would lead to partial delocalization of MapZ, FtsZ and PBPs, which would be rapidly accentuated by PG synthesis, leading to severe shape defects. This work thus suggests a new «quality control« function for a PG hydrolase.
Finally, the methods and study published in (Trouve et al., 2021a, 2021b) provided unprecedented tools and information for PG synthesis and remodeling in ovococci. Data obtained at nanoscale resolution allowed analysis of the architecture of cellular regions in which PG is assembled and remodeled, and extraction of kinetic parameters inaccessible with conventional microscopy methods. These experimental data were exploited in silico to model the morphogenesis of an ovoid cell and to test several hypotheses concerning the activities of the PG machineries. This work shows that in the pneumococcus, divisome and elongasome activities start concomitantly in the septal region, but progress separately during the cell cycle, with elongation persisting after the end of the division. Finally, our data suggest that septal PG synthesized by the divisome is remodeled by PG hydrolases, which partition it, and by the elongasome, which inserts PG into the cleaved regions at the periphery of the septum.

The study by Hosek et al. supports the common idea that early division proteins, such as MapZ, drive the assembly of the PG machineries, which include cell wall synthases and hydrolases (Hosek et al., 2020). The work of Jacq et al. brings an additional level of complexity to this model by showing that, in turn, the enzymatic activity of a PG hydrolase can influence the localization of early division proteins (Jacq et al., 2018). These two studies thus suggest that the assembly of the PG machineries relies on a subtle balance between regulatory mechanisms taking place in the cytoplasm (such as interactions between MapZ, FtsZ and lipids) and at the cell surface (such as interactions between MapZ and the PG). The involvement of lipids in these mechanisms is particularly interesting, and remains a largely unexplored field. Partner 2 has already identified proteins that might influence the fluidity of the cytoplasmic membrane, and thus possibly the recruitment and/or activity of the PG synthases, which are all membrane proteins. Partner 1 has also set up fluorescent labeling methods to localize the fluid and rigid phases of the membrane, as well as the newly synthesized PG. Therefore, one of our future goals will be to exploit these different elements to understand how the nature and organization of lipids influence the assembly and activity of the PG synthesis machineries.
Other exciting research perspectives have been brought by the development of clickable probes to label PG synthesis activity without disrupting the physiological process, and by the study of this activity using dSTORM super-resolution fluorescence microscopy (Trouve, et al., 2021a, 2021b). This work not only opens new concepts regarding PG synthesis and morphogenesis in bacteria, but also provides a new methodological approach that can be exploited in different ways. For example, it can be used on mutant strains to understand the function of proteins involved in cell elongation and division, or in combination with antibiotics targeting the PG synthesis pathway to understand their mode of action. More generally, other specific biorthogonal probes can be developed to study different cellular processes in all fields of life. For example in the pneumococcus, we will tackle the combined study of PG synthesis with synthesis of the teichoic acids, the other major cell wall component in Gram-positive bacteria. The metabolism of these complex glycan polymers is also essential for cell survival and dissecting its relationship with PG metabolism might provide new avenues for the discovery of antibacterial targets.

The DIVinHD project led to 4 research articles published in peer-reviewed journals (Hosek et al., 2020; Jacq et al., 2018; Trouve et al., 2021a, 2021b), with associated highlights written for a broad audience, 26 oral communications including 16 as invited speakers for C. Morlot or C. Grangeasse. During the course of the project, C. Morlot received the Bronze Medal (2018) and C. Grangeasse the Silver Medal of the CNRS (2020) for their work in bacterial cell wall synthesis and cell division, which has been partly funded by this ANR project.

Worldwide infectious diseases are still the predominant cause of death over cancers. In this context, it is paradoxical that our knowledge of prokaryotic cell division, which is a promising source of antibacterial targets, is less advanced than that of eukaryotes.

Bacterial division results from the combination of membrane constriction with expansion and remodeling of the cell wall. To ensure cell integrity and maintenance of cell shape, these processes are coordinated within a large protein machinery known as the divisome. The identity of most divisome components is known, several molecular interactions have been described, enzymatic activity assays and crystallographic structures are available, but we still do not understand how these proteins assemble into a functional macromolecular complex in the cell, and how the assembly and the activity of this complex are regulated in space and time along the bacterial cell cycle. Deciphering these mechanisms is however essential to understand the mode of action of current antibiotics, identify new antibacterial targets and develop new drugs.

Investigating the assembly and regulation mechanisms of protein complexes in vivo using fluorescence microscopy has remained limited for a long time by the diffraction of light (~250 nm), which approximates the fourth of a bacterial cell diameter (~1 µm). In the last years, the emergence of super-resolution imaging has provided very powerful methods to determine how proteins assemble into physiologically relevant structures in vivo. The DIVinHD (DIVision in High Definition) consortium has contributed to this important breakthrough by applying 3D-SIM (Structured Illumination Microcopy, ~110 nm resolution) and pioneering the use of PALM (PhotoActivated Localization Microscopy, ~20 nm resolution) in the human pathogen Streptococcus pneumoniae. These studies provided new fundamental information regarding i) the physiological structure formed by the cytoskeletal protein FtsZ, which assembles into a ring-like structure (the Z-ring) that serves as a scaffolding platform for the divisome, and ii) the positioning of the Z-ring at midcell. The DIVinHD project will exploit this recent technological development to understand how the divisome assembles onto the Z-ring platform and decipher the role of the Ser/Thr kinase StkP in this assembly, StkP playing a major role in the regulation of pneumococcal growth and division. Our consortium is in a privileged position to tackle these questions through a pluridisciplinar approach combining bacterial genetics, biochemistry, biophysics, super-resolution imaging and structural biology.

In wild-type S. pneumoniae and mutant strains impaired for cell division or for StkP-dependent phosphorylation, we will determine in vivo the PALM nano-structure of FtsZ, StkP, the cell wall synthases PBP2x and PBP2b, as well as EzrA, DivIVA and GpsB, three proteins that connect the Z-ring to the PBPs. Because PBPs are present at low copy number in the cell, dSTORM (direct STochastic Optical Reconstruction Microscopy) and 3D-SIM alternative strategies are proposed to bypass potential problems of PBP detection by PALM. The dynamics of all these proteins along the cell cycle will be studied using FRAP (Fluorescence Recovery After Photobleaching), pulse-chase imaging and sptPALM (single-particle tracking PALM). The interactions between FtsZ, EzrA, DivIVA and GpsB will be investigated in vitro using biophysics and electron microscopy. We will finally solve the crystallographic structure of the FtsZ-EzrA complex, which is made of two essential proteins, and determine whether this interaction is essential for pneumococcal survival. Altogether, our work will provide fundamental insights into the mechanisms of bacterial division, helping in the long term the development of novel antibacterial strategies, in particular by inhibiting essential protein-protein interactions.