Cryo-EM reveals how unrelated phage peptides converge on the same bacterial target

Bacteriophages, the viruses that infect bacteria, carry some of the smallest weapons in biology: single-gene lysis proteins, or Sgl proteins, that sabotage essential bacterial machinery and burst the host cell open. A new structural study published in PLOS Biology has now captured one of these tiny viral proteins in the act of jamming a vital molecular pump in Escherichia coli, and the findings suggest that evolutionarily distant phages have independently arrived at strikingly different ways to disable the very same target.

The target in question is MurJ, an essential membrane protein in gram-negative bacteria that flips lipid II, the building-block carrier for peptidoglycan, from the cytoplasmic side of the inner membrane to the periplasm, where the cell wall is assembled. Because peptidoglycan maintains cell shape and protects bacteria from osmotic stress, and because its biosynthesis pathway is already a proven target for antibiotics, MurJ has long been viewed as a vulnerable chokepoint. The research team, led by Kaito Hosoda, Hidetaka Kohga and Tomoya Tsukazaki of the Nara Institute of Science and Technology in Japan, set out to determine exactly how a 55-residue lysis protein called SglPP7, encoded by the Pseudomonas aeruginosa phage PP7, shuts this transporter down.

Using cryo-electron microscopy, the researchers determined the structure of the MurJ/SglPP7 complex at an overall resolution of 3.2 angstroms. They purified the complex using full-length SglPP7 and E. coli MurJ, reconstituted it into lipid nanodiscs made with E. coli phospholipids, and collected data on a CRYO ARM 300 microscope at SPring-8. The resulting model, deposited in the Protein Data Bank under accession code 9X2N with the corresponding map in EMDB as EMD-66480, shows SglPP7 lodged along the intramembrane cleft of MurJ, with a single transmembrane helix running close to MurJ’s seventh transmembrane segment.

The most distinctive feature of the structure is what the authors call the embedded tail. The C-terminal residues of SglPP7, roughly residues 50 through 55, plunge deeply into a positively charged internal cavity on the C-lobe side of MurJ, a pocket formed by transmembrane helices TM7, TM8, TM10 and TM12. This tail insertion appears to be the crux of the inhibitory mechanism: by occupying the internal cavity, SglPP7 locks MurJ in its outward-facing conformation, the state in which the transporter opens toward the periplasm, and thereby arrests the conformational cycling that lipid II transport requires. Notably, a C-terminally tagged version of SglPP7 lost all lytic activity in bacterial assays, consistent with the idea that a bulky tag sterically blocks the tail from inserting into the cavity.

Mutational analysis sharpened this picture. Successive truncations from the C-terminus showed that the sequence through residue 53 was required for lytic activity, and substitutions at the tryptophan at position 53, which sits deepest inside the C-lobe cavity, revealed a graded pattern: the W53A mutant retained full activity, W53G, W53C and W53S showed slightly reduced activity, and the charged substitutions W53K and W53R abolished lysis altogether. When the same variants were tested in a truncated SglPP7(1-53) background that retains only minimal activity, all of them showed weakened function, indicating that the C-terminal region is a key determinant of SglPP7’s potency, with W53 serving as a central, though partially compensated, residue.

To probe the dynamics of the locked state, the team ran all-atom molecular dynamics simulations of MurJ both alone and bound to SglPP7, each embedded in a mixed POPE/POPG membrane, with three independent simulations per condition lasting 120 nanoseconds. The C-alpha root mean square deviation of MurJ was consistently lower in the complex than in the free protein, and the structural fluctuations observed in MurJ alone were suppressed when SglPP7 was bound. A telling detail emerged from tracking the distance between Y266 on TM8 and L247 on TM7, two residues that flank the embedded tail: in all three simulations of MurJ alone this distance tended to shrink, as if the cavity were closing, whereas in the complex the distance was maintained. The distance between G360 and G415, deeper in the cavity and away from the tail, stayed largely unchanged in both conditions. Together, these simulations support the model that the embedded tail physically props open the outward-facing state.

The structure also revealed an unassigned lipid-like density in the hydrophobic groove of MurJ, wedged between TM1, TM8 and TM14, in the region proposed to accommodate the undecaprenyl chain of lipid II. This density sits near F22, R24 and R270, residues known to be essential for MurJ function, although it does not directly contact them. Because the map lacks features corresponding to lipid II’s hydrophilic headgroup, the authors caution that the molecular identity of this lipid cannot be determined; the observation establishes only that a lipid-like molecule can occupy the putative lipid II-binding region even while SglPP7 holds MurJ in the outward-facing state.

The comparative dimension of the study is where the evolutionary story emerges. LysM, a 37-residue lysis protein from phage M, also targets MurJ, but it shares no detectable amino acid sequence similarity with SglPP7 and is encoded at a different genomic location. When the researchers compared their MurJ/SglPP7 structure with the previously determined MurJ/LysM complex, the MurJ portions aligned closely, with a C-alpha RMSD of 0.816 angstroms, and both showed the transporter in the outward-facing conformation. Yet the binding modes differ sharply. LysM consists of two helices, with its first helix positioned peripherally near TM2 and TM7 and its second helix lying along the membrane surface, interacting with the external face of MurJ. SglPP7, by contrast, adopts a different helix orientation relative to TM7, contacts a broader stretch of that helix, and uses a non-overlapping set of interface residues. Alanine-scanning mutagenesis of SglPP7 found only modest activity reductions for F22A, L30A and S39A, suggesting the helix is supported by distributed, weak interactions across a broad surface rather than a few critical contacts.

Functional tests reinforced the distinction between the two inhibitors. When the researchers challenged a panel of LysM-resistant MurJ mutants with SglPP7, only the Q244P mutant allowed bacterial growth under SglPP7 induction, while every other mutant remained sensitive, confirming that the two proteins attack through different binding interfaces. A third player, SglCJ3, adds further nuance: its transmembrane helix occupies a position similar to SglPP7’s but its C-terminal region only partially extends into the cavity, filling the membrane-embedded cleft formed by TM2 and TM7/TM8 instead. The authors therefore classify the three strategies as cavity-occupying for SglPP7, surface-binding for LysM, and cleft-binding for SglCJ3. Interestingly, despite these differences, the side chains of Y36 in SglPP7, L14 in LysM and Y76 in SglCJ3 all insert into a similar pocket on MurJ, and substitution of Y76 in SglCJ3 markedly reduces its lytic activity.

The convergence extends to the tools of modern structural prediction. AlphaFold3 failed to predict the experimentally observed placement of SglPP7 on MurJ and showed low confidence in that region, and all of its predicted MurJ models adopted the inward-facing conformation rather than the outward-facing state stabilized by the phage protein, a bias the authors attribute to the predominance of inward-facing MurJ structures in the Protein Data Bank. The experimental structure also proved superior to a concurrently reported complex built with a truncated, mutation-stabilized SglPP7 variant: the two structures are nearly identical, with a C-alpha RMSD of 0.747 angstroms, but the new full-length determination reaches higher resolution. Taken together, the findings present a structurally characterized example of sequence-divergent phage proteins independently evolving distinct molecular strategies against the same essential bacterial protein, a pattern the authors describe as consistent with convergent evolution toward MurJ inhibition, and one that may inform the development of protein-based antibacterial approaches against drug-resistant pathogens.

Subject of Research: Structural basis of MurJ inhibition by the phage lysis protein SglPP7 and convergent evolution among phage single-gene lysis peptides

Article Title: Structural basis for inhibition of the lipid flippase MurJ by phage lysis protein Sgl PP7 suggests convergent evolution with other Sgl peptides

Article References: Hosoda, K., Kohga, H., Wu, S., Tanaka, H., Shigematsu, H., Miyazaki, R., Chek, M. F., & Tsukazaki, T. (2026). Structural basis for inhibition of the lipid flippase MurJ by phage lysis protein SglPP7 suggests convergent evolution with other Sgl peptides. PLOS Biology, 24(10), e3004039. https://doi.org/10.1371/journal.pbio.3004039

Image Credits: AI Generated

DOI: 10.1371/journal.pbio.3004039

Keywords: bacteriophage, MurJ, cryo-EM, SglPP7, lipid II flippase, peptidoglycan biosynthesis, single-gene lysis protein, convergent evolution, molecular dynamics, antibacterial targets, LysM, PLOS Biology