Action of a minimal contractile bactericidal nanomachine

Abstract

R-type bacteriocins are minimal contractile nanomachines that hold promise as precision antibiotics^1-4. Each bactericidal complex uses a collar to bridge a hollow tube with a contractile sheath loaded in a metastable state by a baseplate scaffold^1,2. Fine-tuning of such nucleic acid-free protein machines for precision medicine calls for an atomic description of the entire complex and contraction mechanism, which is not available from baseplate structures of the (DNA-containing) T4 bacteriophage⁵. Here we report the atomic model of the complete R2 pyocin in its pre-contraction and post-contraction states, each containing 384 subunits of 11 unique atomic models of 10 gene products. Comparison of these structures suggests the following sequence of events during pyocin contraction: tail fibres trigger lateral dissociation of baseplate triplexes; the dissociation then initiates a cascade of events leading to sheath contraction; and this contraction converts chemical energy into mechanical force to drive the iron-tipped tube across the bacterial cell surface, killing the bacterium.

Extended Data Figure 1.. Resolution assessment and model validation for the cryoEM structures.

Resmap results for the collar and baseplate regions of the pyocin reconstructions in pre- and post-contraction states. Listed in the table are model validation statistics for the collar, baseplate subunits and average.

Extended Data Figure 2.. Model assessment of pre-contraction pyocin subunits.

For each of the collar and baseplate proteins, the cryoEM density map is shown as semi-transparent gray superposed with its atomic model (ribbon). The close-up view of the box region shows the match of the density (wire-frames) and the atomic model (sticks).

Extended Data Figure 3.. Types of sheath handshakes in pyocin.

Ribbon diagrams depicting the 3 types of handshake conformations in pyocin: a. collar-sheath, b. sheath-sheath, and c. sheath-sheath initiator. All handshakes compose of a four-stranded β-sheet.

Extended Data Figure 4.. Trunk transitioning into baseplate.

a. Ribbon diagram depicting lower portion of pyocin. b. Schematic diagram depicting changes in quaternary structure of the sheath subunits approaching the baseplate. The pink circles depict expected positions of the sheath subunits according to helical symmetry of trunk. The blue circles depict actual positions with greater sequential helical turn, 4.4° at the last disk of sheath.

Extended Data Figure 5.. Inspection of the baseplate.

a. Ribbon diagram of ripcord hexamer with tube hexamer. b and c. Ribbon diagram of the hub (b) and the spike (c) (with chelating site of its iron ion highlighted). d. Binding of ripcord into triplexes. e. Ribbon diagram of the baseplate with one sixth of its six-fold symmetric part highlighted in colors as in Fig. 1 showing relative positions of each subunit. f. Baseplate ribbon model superimposed with blurred cryoEM density map of the proximal region of the tail fiber.

Extended Data Figure 6.. Functional and morphogenetic implications of ripcord mutagenesis.

a. Overview of ripcord mutagenesis. b. Co-expression of the WT pyocin and mutant 626TEV with the TEV protease. Pyocin killing activity in the lysates was assessed with the help of a spot assay with P. aeruginosa 13s strain as prey. Both pyocin and protease expression levels are arabinose dependent with the rate of protease production being proportional to arabinose concentration and pyocin expression reaching the maximum at the lowest concentrations of arabinose tested (0.01%). Each experiment was repeated biologically three times, also for c-g. c. Representative negative staining electron microscopy images of the crude lysates shown in the panel b induced with 0.01% arabinose. Despite showing killing activity in the lysates, no extended particles were found in the mutant 626TEV on EM grids. d-f. Temperature dependent sheath contraction rates of the WT pyocins and mutants measured with the help of circular dichroism. g. The rate constants k(T) of WT pyocins, 626ΔWL and 626TEV fitted to the Arrhenius model k(T) = A exp(−E_a/RT) where T is the absolute temperature, A is a temperature independent constant, E_a is the activation energy, R is the ideal gas constant. The dots on the graph are individual values for three biologically independent measurements of ln k(T), and the error bars show 95% confidence interval calculated for them.

Extended Data Figure 7.. Interactions important for triplex formation.

a. Ribbon diagram of the atomic model of the pyocin triplex. b-d. Ribbon model with depicted side chains in the top, stem, and base regions of the triplex. d. Phenylalanine pi-stacking coordination between PA0618a (yellow, Phe89, Phe125) PA0618b (red, Phe89, Phe125) and PA0619 (blue, Phe72). e. Ribbon model diagram of the lateral dimer. f. Close-up highlighting key interacting residues (Phe253 – Phe253, His257 – Ser250, Ser250 – His257).

Extended Data Figure 8.. Electrostatic views of the ripcord handle.

a. Electrostatic surface diagram of ripcord with adjacent triplexes. b-c. Electrostatic properties of the interfaces between ripcord and triplexes 1 and 2, respectively. Positive (blue), neutral (white), negative (red).

Extended Data Figure 9.. Comparison of related protein subunits from Pyocin R2 and T4 phage.

Ribbon diagram of a. pyocin triplex, b. T4 triplex equivalent with subunits marked by corresponding color to (a), c. Pyocin PA0618a (yellow) and T4 gp6A (grey), d. Pyocin PA0619 (blue) and T4 gp7 (grey), e. Pyocin PA0618b (red) and T4 gp6B (grey), f. Pyocin PA0627 (pink) and T4 gp53 (grey), g. Pyocin PA0617 (green) and T4 gp25 (grey).

Figure 1.. CryoEM and overall structure of pyocin in pre- and post-contracted states.

a. Shaded surface representation of the cryoEM reconstructions, colored according to cylindrical radii as shown in the color bar. b. Representative cryoEM micrograph. c. Regions of the cryoEM density map (mesh) superimposed with atomic models (sticks) demonstrating the agreement between the observed and modeled amino acid side chains. d. Atomic models for pyocin in both the pre- and post-contracted states. e and f. Ribbon diagrams of individual subunits of pyocin in the pre-contracted state (e) shown along with their corresponding gene loci (f). See 3 dimensional rendition in Supplementary Videos 1 to 3.

Figure 2.. Architecture of the collar.

a. Top view ribbon diagram of the collar (pink), outer sheath (cyan) and inner tube (grey). b. Space filling model of collar-sheath-tube region. c. Electrostatic surface model of collar sheath handshake with views of the interacting charged surfaces. Positive (blue), negative (red), neutral (same as b). d. ribbon and space filling diagrams of the post-contraction collar-sheath-tube region similar to a and b.

Figure 3.. Triplex expansion and lateral dimer dissociation.

a-e. Ribbon diagrams of triplexes forming an iris ring in pre-contracted state (a), expanded iris in post-contracted state (b), side view of two adjacent triplexes in pre-contracted state (c) and post-contracted state (d), and lateral dimer of PA0618 (e). f. Schematic of the iris ring expansion as result of the tail fiber actuation. g. PA0618 H257F mutant. Left, percentages of pre-contraction pyocins in purified wildtype (WT) and H257F mutant under cryoEM at neutral and acidic pH. (pH 7.4 WT: 185/289, H257F: 118/175; pH 3.5 WT: 46/530, H257F: 64/178. Error bars: standard deviations.) Right, representative cryoEM image for each relevant condition.

Figure 4.. Baseplate transition from the pre- to post-contracted state.

a. Illustration of a pyocin landing on a bacterial cell and firing. Release of the spike and hub following injection is postulated on the basis of the lack of these structures on contracted particles that we observed in vitro. b. Ribbon diagram of the conserved baseplate components and sheath proteins in its pre- and post-contracted states. Ripcord is believed to travel with the inner tube during the power stroke and therefor is not a conserved component of the baseplate after contraction.