Membrane translocation process revealed by in situ structures of type II secretion system secretins

Bacterial secretion systems are crucial for virulence and survival. The type II secretion system (T2SS) is widespread among Proteobacteria, secreting diverse substrates impacting nutrient uptake, symbiosis, and pathogenesis. The secretin, a major outer membrane channel component of T2SS, is essential for substrate transport. Phylogenetic analysis reveals two secretin types: *Klebsiella*-type and *Vibrio*-type, differing in their transmembrane regions and proposed membrane interactions. While previous studies have visualized the in situ T2SS architecture at limited resolution, understanding the detailed translocation mechanism of secretins from the inner to the outer membrane remains unclear. Two potential pathways are known: one involving scaffolding proteins GspA and GspB that increase peptidoglycan pore size, and another utilizing a small pilotin protein, GspS, which translocates to the outer membrane via the Lol pathway. However, direct visualization of this translocation process in living cells is lacking. This study uses *E. coli* as a model, focusing on GspDα (*Klebsiella*-type) and GspDβ (*Vibrio*-type) secretins to elucidate the translocation mechanism through in situ cryo-ET studies.

Prior research has established the importance of T2SS in bacterial pathogenesis and various biological processes. Studies have characterized the structure and function of secretins, revealing their cylindrical channel architecture with N0-N3 domains, a central gate region, and an S domain. However, the *Klebsiella*- and *Vibrio*-type secretins differ in their transmembrane regions, specifically the presence of a cap gate in the *Vibrio*-type, impacting their predicted membrane interactions. Previous in vitro structural studies have provided insights, but in situ studies are crucial to understand secretin behavior in the cellular environment. Biochemical experiments have proposed two pathways for secretin translocation, one involving scaffolding proteins and peptidoglycan remodeling, and another utilizing pilotins and the Lol pathway. Nevertheless, these processes have not been directly visualized within living cells.

The study employed a *E. coli* minicell system to obtain thin cells suitable for cryo-ET imaging, allowing for improved contrast. Overexpression of GspDα and GspDβ, with and without GspS, was induced in different *E. coli* strains (including a Δ*gspS* strain). Cryo-ET was performed, collecting numerous tilt series. Subtomogram averaging, using EMAN2, was performed on manually picked particles. A novel algorithm was developed to improve particle orientation determination by incorporating cell membrane location as a constraint, particularly for distinguishing top and bottom views of GspDα particles. Symmetry determination involved analysis of particle radii and refinement with various symmetries (C12, C14, C15, C16). To visualize membrane interactions more clearly, symmetry relaxation was implemented to generate asymmetric structures. Targeted refinement was also employed to focus on the protein region independent of the membrane. D-methionine was utilized to reduce peptidoglycan crosslinking to observe the effects on GspDα translocation. Membrane separation experiments using sucrose density gradient centrifugation and western blotting with anti-His tag antibodies verified the location of GspDα and GspDβ in the inner and outer membranes.

The study reveals four in situ structures of T2SS secretins: GspDα on the inner membrane, GspDα on the outer membrane, GspDβ-GspS complex on the outer membrane, and GspDβ on the inner membrane. GspDα on the inner membrane exists in a flexible, partially connected state, exhibiting conformational flexibility and swinging motion around its membrane contact site. Adding D-methionine, which reduces peptidoglycan crosslinking, facilitated GspDα translocation to the outer membrane, where it adopts a more stable, evenly connected conformation. GspDβ, when co-expressed with GspS, forms a stable complex on the outer membrane, with GspS densities clearly visible. Even without GspS overexpression, endogenous GspS is sufficient for GspDβ outer membrane localization, though a small fraction remains on the inner membrane. In the absence of GspS, GspDβ forms stable multimers on the inner membrane. Notably, the transmembrane region of the secretins only interacts with one leaflet of the membrane, a finding that contrasts previous expectations.

The findings support two distinct translocation models. For GspDα, the initial inner membrane-bound state is unstable and facilitates translocation when the peptidoglycan pore size is increased. For GspDβ, the pilotin GspS plays a key role, binding to GspDβ, likely facilitating monomeric translocation and subsequent reassembly on the outer membrane. The observed interaction of secretins with only one leaflet of the membrane warrants further investigation, possibly pointing towards an extended in vivo conformation. The differences between GspDα and GspDβ translocation mechanisms, particularly the involvement of GspS, might explain the evolutionary preference for the *Vibrio*-type T2SS in many bacteria.

This study provides high-resolution in situ structures of T2SS secretins, revealing distinct translocation mechanisms for *Klebsiella*- and *Vibrio*-type secretins. The findings highlight the roles of peptidoglycan pore size and the pilotin protein GspS in secretin targeting. Future research could focus on visualizing the intermediate steps of translocation, clarifying the unexpected transmembrane interaction with only one leaflet, and exploring the precise role of GspS in monomeric translocation.

The study primarily uses overexpression systems, which might not fully reflect the native conditions. The resolution of some structures could be further improved to allow for more precise modeling of protein-membrane interactions. While the models propose mechanistic insights, direct visualization of all intermediate steps remains challenging.