ABSTRACT
The bacterial flagellar motor is driven by an ion flux that is converted to torque by motor-attendant complexes known as stators. The dynamics of stator assembly around the motor in response to external stimuli have been the subject of much recent research, but less is known about the evolutionary origins of stator complexes and how they select for specific ions. Here, we review the latest structural and biochemical data for the stator complexes and compare these with other ion transporters and microbial motors to examine possible evolutionary origins of the stator complex.
INTRODUCTION
The machinery behind microbial swimming is an astounding example of natural nanotechnology. The Escherichia coli flagellar motor is currently the best-understood system for bacterial locomotion. It is the product of ∼50 genes distributed over 10 operons—of which half encode motor-incorporated proteins and half encode proteins with auxiliary and regulatory functions (1). The bacterial flagellar motor (BFM) structure and amino acid sequence of constituent proteins are conserved across a diverse range of taxa that colonize diverse terrestrial, marine, and commensal environments (2–4). The presence of the BFM across such a wide range of bacterial taxa and environments suggests both an early origin and a significant fitness advantage that is maintained even in nonaqueous environments (5).
The stator complexes are motor-associated protein complexes which convert potential energy, in the form of an ion motive force, to torque. The majority of BFMs are powered by protons or sodium ions flowing across the plasma membrane, but the ion used to power rotation varies across different species, and the mechanism whereby the BFM selectively filters ions is yet to be fully determined.
When investigating the origins of microbial motility, it behooves us also to consider the rotary motors that drive swimming in some species of archaea. Recently, much progress has been made in structural and biophysical characterization of archaeal flagellar motors (AFMs) in Thermococcus kodakarensis, Pyrococcus furiosus, Sulfolobus acidocaldarius, Methanocaldococcus jannaschii, Halobacterium salinarum, and Haloferax volcanii (6). These motors share homology to type IV pili and the homologous type II secretion system (T2SS), rather than the type III secretion system (T3SS), which shares homology with the BFM. The AFM differs from the BFM in the structure of the rotor and filament and, notably, in being powered by ATP hydrolysis rather than ion transit. However, both motors are coupled to conserved chemotaxis machinery which relies on the phosphorylation of CheY (7). The surprising outcome that the chemotaxis machinery is highly conserved, while large structural and functional differences exist between the motors, presents rotary motility as an excellent case study in convergent evolution.
Here, we review the origin and diversity of the BFM, focusing specifically on the stator complexes. We summarize what is known about the genetic and structural factors that govern ion selectivity of the stators and their evolutionary relationship to other ion-coupling-dependent proteins.
STRUCTURE
The structure of the bacterial flagellar motor.Although an intact BFM in E. coli contains 22 different proteins, only five are directly involved in rotation and direction switching. These include three proteins in the rotor, FliG, FliM, and FliN, and two proteins, MotA and MotB, forming the MotAB stator complex (Fig. 1). Different stator complexes, such as PomAB, MotCD, and MotPS, have been found in other species (8). Energy for torque generation comes from an influx of cations across the plasma membrane, through the stator complex, driven by the electrochemical gradient, which is subsequently converted into mechanical rotation by a rotor substructure known as the switch complex. The switch complex is comprised of dozens of copies of FliG, FliM, and FliN, which form a large bell-like structure ∼45 nm in diameter (the C-ring) which is responsible for torque generation, motor reversal, and regulation of rotation speed (9). Charged residues on or around a single helix in the C-terminal domain of FliG interact with at least two residues on the stator protein MotA to generate torque. However, the most significant interactions for torque occur between FliG-R298/MotA-E98 and FliG-D289/MotA-R90 (Fig. 2A) (10, 11). Motor reversal from the counterclockwise (CCW) direction to the clockwise (CW) direction is mediated by the binding between the phosphorylated form of CheY (CheY-P) and FliM (12–14). Binding of CheY-P to FliM triggers a structural change in FliM which in turn moderates interaction of FliG with the MotAB stator complex (15, 16).
Schematic structures of the bacterial flagellar motor and archaeal motor. (A) The core of the flagellar motor is composed of the switch complex, containing FliG, FliM, and FliN, which forms an ∼45-nm-diameter C-ring, and the stator complex, a heterodimer of MotA and MotB. Torque is generated by a proton influx through the stator complex, which triggers interactions between MotA and FliG. Directional switching occurs via interaction of the chemotaxis protein CheY-P and FliM in the switch complex, which in turn moderates interactions between FliM and FliG and between FliG and MotA. (B) Archaellins FlaJ, FlaI, FlaH, FlaF, and FlaG are conserved in the euryarchaeal (left) and crenarchaeal (right) archaella. FlaJ, FlaI, and FlaH form the archaellum core and are involved in assembly and torque. FlaJ is known to be membrane bound, buts its structure is unknown. FlaI ATPase assembles as a hexamer and interacts with the FlaH ATPase regulator. Torque occurs through hydrolysis of ATP via the assembled FlaI and FlaJ, but the exact mechanism is unknown. FlaF is postulated to anchor the archaellum to the S-layer, similar to MotB conferring motor and cell surface stability by anchoring to the peptidoglycan layer. The structure and function of FlaG are unknown. In the Euryarchaeota, FlaC, FlaD, and FlaE form the switch complex. The structural interactions of these archaellins are unknown, but these proteins are postulated to encircle FlaI and FlaH. Directional switching occurs through the interaction of CheY-P/CheF with FlaC, FlaD, and FlaE. In the Crenarchaeota, FlaX assembles as an ∼30-nm-diameter ring around FlaJ, FlaI, and FlaH. FlaX has homology to chemotaxis proteins. However, the chemotaxis machinery employed is unknown.
Topology and ion transport mechanism of the stator. (A) Schematic showing the topology of stator components MotA and MotB (monomers) also shared by PomA/PomB. Each MotA contains 4 TM helices arranged in a bundle, and the cytoplasmic loop between TM2 and TM3 displays the charged amino acid residues interacting with FliG (dashed box). Residues interacting at the MotA-FliG interface include R90/D288-D289 and E98/R281 (data from reference 10). MotB displays a single TM helix followed by a plug domain and an OmpA-like PG-anchoring domain at its C terminus. PDB identifier, 2ZVY (105). See also reference 40. (B) Schematic of the mechanochemical gating cycle of the stator. (I) When the cation (yellow circle) binds to aspartate on subunit B (red circle), this triggers a conformational change in subunit A which promotes its interaction with FliG (dashed circle). (II) The rearrangement of subunit A perturbs the ion-binding site by the action of additional side chains in the pore (blue circle), destabilizing the interaction with Asp. (III) After the release of the cation into the cytoplasm, TM3 and TM4 return to their original position while producing torque at the interface with FliG. Figure adapted from Scientific Reports (90).
The structure of the archaeal flagellum (archaellum).Overall, the archaellum is a rotary device powered by ATP hydrolysis that is structurally similar to the bacterial type IV pili (Fig. 1). The archaellum is composed of 7 to 15 proteins (archaellins), depending on species, which are homologous to bacterial type IV pilus proteins and which form the general assembly machinery and motor complex (6). Unlike bacterial type IV pili that extend and retract, archaeal filaments rotate like flagellar filaments (17).
Of core importance to archaellum function is FlaI, a cytosolic ATPase enzyme responsible for archaellum assembly and motility (18, 19). FlaI, along with FlaJ and FlaH, forms the main motor structure in the AFM (Fig. 1B). The membrane-bound platform protein FlaJ interacts with the N-terminal domain of FlaI, and a regulator protein, FlaH, interacts with the C-terminal domain of FlaI and regulates the ATPase activity of FlaI (19–21). ATP binding and hydrolysis likely drive the switch between the open and closed conformation and rotation of the FlaI hexamer, which assembles and rotates the archaellum (19).
Most recently, functional studies on rotating archaella in H. salinarum (22, 23) have replicated similar measurements that had been executed earlier on the BFM using optical trapping and interferometry. These measurements have led to the current estimation for the number of ATP molecules consumed per revolution to be 6 ATPs (22).
The core FlaJ, FlaI, and FlaH motor complex is conserved across archaea. Other conserved archaellins essential for function include the membrane proteins FlaF and FlaG. The exact FlaG function is unknown. FlaF binds as a dimer to S-layer protein and anchors the archaellum in the cell envelope (24). This enables FlaF to serve as a stator and also acts to prevent membrane rupture during archaellum filament rotation. This function is similar to MotB anchoring the MotAB stator of the BFM to the peptidoglycan layer.
(i) Torque generation in archaea. In the Euryarchaeota, the switch complex consists of FlaC, FlaD, and FlaE (6) and is postulated to wrap around FlaI and FlaH as a cytosolic ring (Fig. 1B). However, the structural complex formed by FlaC, FlaD, and FlaE and their interaction with FlaI and FlaH are currently unknown (25). Crenarchaeota do not contain FlaC, FlaD, or FlaE but contain FlaX, which has a transmembrane N-terminal domain and a cytoplasmic C-terminal domain. The C-terminal domain of FlaX forms a 30-nm-diameter oligomer ring that possibly serves as the assembly scaffold for the Crenarchaeota motor complex (26). FlaX interacts with both FlaI (26) and FlaH (27). The FlaX ring encases both the FlaI hexamer and a ring of FlaH proteins, and the N-terminal domain of FlaX is also thought to encase FlaJ embedded in the membrane (27).
(ii) Commonality in microbial chemotaxis. The core genes of the bacterial chemotaxis machinery are conserved in the Euryarchaeota (28, 29). In fact, the change of rotational direction in the archaella of H. salinarum and H. volcanii is also mediated by CheY-P (30, 31). However, in the Euryarchaeota, CheY-P complexes with an adaptor protein, CheF, and CheF is what binds to the switch complex FlaC/D/E (30). This chemotaxis system is not found in the Crenarchaeota, and while FlaX has similarities to methyl-accepting chemotaxis proteins (32), the type of chemotaxis employed, and whether rotation switching occurs, remains to be shown. However, the common use of rotary motility for microbial locomotion, via very different structural, functional, and energetic pathways, is an excellent case study in convergent evolution, particularly given the conservation of CheY. This conservation of chemotactic machinery could imply that the infrastructure for chemotaxis arose once and was subsequently shared by horizontal gene transfer (HGT). However, how separate motors adapted to respond to these sensing pathways is unknown.
STATORS
The stators are where the motor transduces chemical energy from ion transit into mechanical force, and they are also the site of ion selectivity in the BFM (1). Stator complexes mainly include H+-coupled MotAB and MotCD and Na+-coupled PomAB and MotPS (4), but complexes coupling K+, Rb+, Mg2+, and Ca2+ have also been recently discovered (33, 34). Further stator proteins exist whose primary function is to stabilize stator complexes. In Vibrio alginolyticus, MotXY anchors PomAB in the motor (35), and in Sinorhizobium meliloti, MotE maintains the stability of MotC, which in turn stabilizes the MotAB complex (36, 37).
The functional stators in the majority of the BFM are motor-associated transmembrane (TM) complexes made up of A and B subunits—with a stoichiometry of A(4)B(2). These form an ion channel spanning the periplasmic space and the cytoplasm. Two transmembrane regions of MotA (A-TM3 and A-TM4 [Fig. 2]) form the ion-binding channel with subunit B, with one of the remaining TM regions (A-TM2) forming one membrane anchor of the cytoplasmic loop and the other (A-TM1) having an unconfirmed function (38).
Association of the entire stator complex with the motor is accompanied by conformational changes mediated by electrostatic interactions with FliG (39, 40). Simultaneously with motor association, the stator complex attaches to the peptidoglycan layer via an OmpA-like binding domain (41–43). When this association is complete, a plug domain in the B subunit opens to allow ion influx and selectively transports cations from the periplasmic space into the cytoplasm (44).
Stator complexes likely function as a loosely coupled Brownian ratchet, where the stators can execute power strokes on the rotor driven by either one or two energizing ions (39, 45). Conformational changes in MotA are induced by ion binding in the TM region of MotB. Ion flow is obstructed by the essential aspartate residue D32 in MotB, at which point it has been proposed that protonation of D32 induces a bend in A-TM3 which is then transmitted via the A-cytoplasmic loop to FliG to allow final passage of the ion into the cytoplasm (Fig. 2B) (46). Molecular dynamics modeling suggests that hydrogen bonding between protonated D32 in MotB and D170 in A-TM3 induces the conformational change which pivots around P173 in A-TM3 (47).
Loss-of-function mutants of MotB that cause a misalignment of the stator relative to FliG could be rescued by mutations in MotA at the pore interface to restore proton conduction, although with some loss in motor performance (48, 49). Furthermore, the efficiency of the force transfer from the stator to FliG appears to depend on the presence of prolines (P173 and P222) at the cytoplasmic end of MotA TM helices that presumably restrict the motion of the stator during the cascade of conformational changes to optimize the power stroke. In fact, the replacement of an arginine residue with a proline (R109P) in the α-helix connecting the transmembrane and peptidoglycan binding domains in a nonfunctional chimeric PotB construct, consisting of a C-terminal PomB and N-terminal MotB, restored Na+ conductivity in the chimera by conferring a proper arrangement of transmembrane helices (50). These examples highlight how the structural requirements for coupling ion passage through the stator to force generation can be met by several amino acid configurations displayed from either pore subunit.
Binding to the conserved aspartate residue (D32) within the TM region of the B subunit causes conformational changes that drive torque generation (39, 51). In V. alginolyticus, seven residues in PomA and six residues in FliG contribute to torque generation, with the interaction between PomA-E97 and FliG-K284 being critical for function (40). The equivalent conserved glutamate residue, E96, in Aquifex aeolicus MotA, along with R88, is important in the function of the hyperthermophile’s flagellar motor (40, 52–54). These residues are exposed on the cytoplasmic loop surface between A-TM2 and A-TM3 which forms the interface with FliG. Chimeras of A. aeolicus and E. coli stator proteins are functional in E. coli flagellar motors, suggesting that the interaction between the A subunit and the rotor is highly conserved and ancient (55). However, there are differences between E. coli MotA and V. alginolyticus PomA in their interactions with FliG, as the cytoplasmic loop of PomA has more charged residue interactions with FliG than the loop of MotA (40).
SELECTIVITY
The cations selected by the stator complex are most commonly protons or sodium ions, as seen in enteric bacteria and Vibrio spp. (1). However, stators that utilize other ions of the alkali metal group have been discovered, such as rubidium and potassium in alkaliphilic Bacillus (34). Some bacterial species even have the ability to couple multiple ions via the use of multiple sets of stator genes encoding different stator proteins, e.g., Bacillus subtilis and Shewanella oneidensis (56, 57), or single, dual-purpose complexes such as in Pseudomonas aeruginosa and alkaliphilic Bacillus (34, 58). Dual ion use can be introduced into single-use stators with mutations in the ion-binding transmembrane region of the B unit, as seen in Bacillus subtilis (59). A Paenibacillus species has been found that couples divalent magnesium and calcium ions, which are abundant in its environment, for motility, providing the possibility of flagellar rotation powered by multivalent ions (33).
Attempts to obtain information about the TM regions of MotA and MotB have been hindered by the difficulty of generating crystals (here in Helicobacter pylori) (42). Recently, a combination of electron microscopy single-particle imaging in A. aeolicus (60); molecular dynamics (47); and small-angle X-ray scattering (SAXS) in H. pylori, Salmonella enterica, and V. alginolyticus (61, 62) has refined models for MotA and MotB, but an atomic structure, perhaps via improved single-particle cryo-electron microscopy (cryo-EM), is yet to be determined.
Filter.Molecular dynamics simulations of the ion channel region have proffered size exclusion as a mechanism for ion selectivity (47). The radius of the ion channel created by MotB residues 19 to 54 was predicted to be 1.0 to 2.3 Å in a nonprotonated state, with the narrowest point occurring at L46. Protonation of critical residue D32 expands the channel near the cytoplasm and the periplasm to promote entry and ejection of the translocating ion, but the channel remains narrow at L46 to maintain selectivity. Increases in radius of the ion channel typically improve sodium flux over proton flux as seen in Na+-coupled ATP synthases (63). Thus, ion selectivity in the stator complex does not appear to rely solely on size exclusion, nor to be controlled by a single residue. However, in Vibrio cholerae, PomB-S26, located on the same face of the B-TM helix as the conserved aspartate D23, is known to be important for optimal swimming and is thought to disrupt the hydration shell of Na+, allowing the ion to move through a constriction in the channel (64).
The E. coli MotAB complex and the V. alginolyticus PomAB complex are the best-characterized H+- and Na+-coupled stators, respectively. A chimeric B unit, PotB, consisting of the TM helix and cytoplasmic domain of V. alginolyticus PomB (residues 1 to 50) and the peptidoglycan-binding (PGB) domain and periplasmic linker of E. coli MotB (residues 59 to 300), has been used extensively to study underenergized motors at low sodium concentration in E. coli (1).
PotB forms a chimeric stator complex with PomA (PomA4PotB2) and restores motility in E. coli ΔmotAB as an Na+-coupled PomA/PotB stator complex (65). In the PotB chimera, the ion channel is made exclusively of the V. alginolyticus PomA and PotB TM region, with the N-terminal E. coli component of the B subunit allowing the complex to bind to the peptidoglycan layer. This further verifies that B-TM, A-TM3, and A-TM4 determine ion selectivity, as anticipated by previous research with stator chimeras in Rhodobacter sphaeroides and V. alginolyticus (65).
EVOLUTION
The BFM shares core structural components with the pathogenic injectisome of proteobacteria, the nonflagellar type III secretion system (T3SS) (66). It is likely that the T3SS evolved through the exaptation of genes of the BFM, as evidenced by modern phylogenies of flagellar and nonflagellar T3SS (67). The evolutionary primacy of the BFM is also supported by the T3SS’s constrained existence in newer bacterial phyla relative to the broader spread of flagellar motors (68). The oldest forms of the BFM reside in early-branched hyperthermophilic bacteria (55). By looking at B subunit chimeras of A. aeolicus and E. coli in E. coli, we can understand which interactions between stators and motors have been conserved. A chimera combining the TM region of A. aeolicus MotB and the PGB domain of E. coli MotB forms a functional stator complex with A. aeolicus MotA in E. coli. The interaction between A. aeolicus stators and E. coli rotors suggests that the interactions between FliG and MotA, which drive torque generation in the BFM (Fig. 1), must be highly conserved, even among phylogenetically distant taxa.
Of the ∼50 gene products necessary to produce the BFM, perhaps as few as two proteins both are indispensable to function in all known varieties of motors and do not share a known sequence homology with another extant protein (69). Therefore, far from being irreducibly complex, the BFM contains an overwhelmingly large amount of modular, bifunctional, and noncritical mass that points to a common ancestral beginning.
Origins of stators.MotA and MotB share sequence homology with other proton-coupled TM transport systems. The Tol-Pal system in E. coli is implicated in membrane stability and protein transport, and it depends on the proton motive force (PMF) for function (70). This system consists of TolQ, a polytopic cytoplasmic membrane-spanning protein analogous to MotA, and component TolR, a MotB analogue, that transmits a signal to TolA in active colicin uptake. TolQR shares TM structural and functional homology with MotAB, as does ExbBD of the TonB system (71). ExbBD uses the PMF to energize transport of siderophores and vitamin B12 in Gram-negative bacteria. In ExbD, a homologue of MotB, a mutation of a conserved aspartate residue in the TM domain abolishes the function of the complex (72), as it does when the conserved aspartate D32 in MotB is mutated.
There is also evidence of horizontal gene transfer (HGT) of flagellar motor and stator genes occurring between bacterial taxa; the alphaproteobacterium R. sphaeroides contains a second set of flagellar motor genes acquired through HGT from a gammaproteobacterium (73), and Rhodospirillum centenum acquired its second set of stator genes from a Vibrio species (74). S. oneidensis, the only member of its genus to live in freshwater, is predicted to have received its H+-coupled MotAB stator from a sympatric Aeromonas species, whereas its Na+-coupled PomAB stator is homologous to marine members of its genus (75). Takekawa et al. (55) demonstrated that the hyperthermophilic A. aeolicus MotA and MotB conduct Na+, suggesting that filament rotation in ancient bacteria was driven by Na+, as these bacteria are thought to possess some of the first motor proteins to have evolved. Furthermore, due to the low sequence similarity between Na+- and H+-coupled stators in bacteria that have both types of stator, it was suggested that Aquifex and Thermotoga are common recipients of ancestral sodium-coupled flagellar stator genes (55), as both genera are predominantly hyperthermophilic and share significant genome homology with archaea (76). It is possible that A. aeolicus shares a MotB lineage with Epsilonproteobacteria (77), a class of bacteria with members that colonize hydrothermal vents typical of Aquificae, Thermotogae, and other early-branched bacterial and archaeal lineages.
Although phylogenetic data do not support the assumption that Na+-coupled stators arose first, the assumption that ancestral microbes in high-salt environments would be powered by sodium motive force aligns with contemporary theories for ancient membrane bioenergetics (78). Na+ coupling is thought to have arisen before H+ coupling in F1Fo-ATP synthetases (79). Generally, the hypothesis that Na+ selection predates H+ selection in TM protein complexes is well supported (80) but only circumstantially supported in stators by looking at the ion-coupling preferences of contemporary bacteria and comparing this to their evolutionary age (55).
Insights from other bacterial ion transporters.The proton transport mechanism through the stator complex is more akin to ion transporters than typical ion channels. For ion channels, opening produces a continuous water-filled pore through which water and ions pass, often in single file through the narrowest region, while in transporters one or a few ions are transferred across the bilayer as a result of a series of dedicated conformational changes in the protein (81). This makes ion transporters orders of magnitude slower than channels at moving ions. Theoretical calculations of ionic flow rates estimate ∼2 × 105 H+/s through an E. coli stator, whereas the observed rate for voltage-gated sodium channels is on the order of 107 Na+/s (82, 83). Similarly to the N-type inactivation observed in voltage-gated sodium channels (84), the flow of ions through the MotAB stator complex can be inactivated by the action of the plug domain of MotB (44). In the BFM, this inactivation is maintained before the stator complex assembles onto the rotor to prevent leakage of ions prior to motor assembly rather than to cease flow after sustained ion conduction, as in other channels. Activation of the stator depends on binding of the C-terminal OmpA-like domain of MotB to the peptidoglycan layer (85).
The rate of proton flux at the MotA/B interface is limited by the rate of conformational change that is induced via the binding and release of H+ to the aspartic acid residue D32 in the stator (86, 87). As per an ion transporter, this mechanochemical process relies on the sequential binding of H+ to an acidic amino acid (e.g., D32 in E. coli MotB), followed by a conformational change and the subsequent release of H+ from the amino acid to regenerate the proton-binding site, thus limiting the ion flux to one proton per torque-generating cycle of conformational change at each pore (∼1,200 protons per revolution per motor in Streptococcus spp. and E. coli) (39, 88, 89). This mode of transport appears to be conserved in Na+-powered stators, which employ additional polar amino acid residues along the pore to accommodate the binding and release of sodium ions (90).
It has been speculated that substantial transmembrane helical movements are required for ion channels to transition between open and closed states (91). Such movements might be inefficient or infeasible in a single protein with an ion conduction pathway through its center in comparison with the interface of separate subunits. Thus, ion conduction at protein interfaces possibly provides an essential flexible pathway that enables rapid conformational changes (91).
Helix swiveling has been proposed as a critical consequence of aspartate protonation during H+ transport in the Fo subunit of the E. coli ATP synthase (92). A similar mechanism might be involved in the force generation step in the BFM stator (90). In contrast to the BFM, however, H+ selectivity in the bacterial ATP synthases has been suggested to be determined by the c-ring rotor of Fo rather than the interfaces in the stator a subunit (93). The binding of H+ to R61 on the c-ring directly mediates a 36° rotation of the c-ring by modulation of the electrostatic interactions at the stator-rotor interface and the twisting of the protonated helix (93).
Other homologous complexes for the stator exist in prokaryotes that exploit PMF to couple a “power stroke” or conformational change to various outputs (Fig. 3). The TonB-ExbB-ExbD complex plays an essential role in siderophore acquisition in E. coli, transducing PMF into translocation (94, 95). The TolQ-TolR stator complex of E. coli is recruited to cell division sites, where it is coupled to proteins dedicated to maintaining the integrity of the outer membrane (96). Last, the AglR/Q stator of Myxococcus xanthus drives a distinct bacterial motor involved in gliding motility (97). In flagellar stators, the selectivity for protons, and presumably for sodium ions, appears to be dependent on interfacial residues from both A and B subunits, suggesting a coevolutionary linkage underlying the mechanism of PMF-driven torque generation. This may have been relevant for the adaptation of motility machinery after the oxygenation of the atmosphere in ancient times, when the enzymes involved in membrane bioenergetics, to maintain the Na+/K+ disequilibrium across the membrane, gradually implemented Na+/H+ transport mechanisms (80, 98). With the concomitant evolution of membrane lipids to cope with the new environment (99), prokaryotic membranes became largely impermeable to both sodium ions and protons, and proton-dependent bioenergetics became prevalent (78).
PMF-coupled systems in prokaryotes. (A) Cartoon diagram of prokaryotic membrane machinery employing stator homologues to couple PMF to various biological outputs. Proton transport at the stator subunit interface (straight downward arrow) generates a mechanical interaction with downstream effectors (curved red arrow) to provide torque to the flagellar motor (MotA/MotB), to activate iron transport (ExbB/ExbD), to promote outer membrane remodeling (TolQ/TolR), or to drive gliding propulsion (AglQ/AglR/AglS) in each system, respectively. All complexes are shown with 1:1 stoichiometries. The stoichiometry of MotAB is 4:2, that of ExbBD is 6:3 (95), that of TolQR is 6:2 (96), and that of AglQR is not known, albeit speculated to be homologous to TolQR (97). (Figure modified from reference 106 with permission of the publisher.) (B to D) Schematic view of the stators from the extracellular side. (B) The BFM stator subunit A contains 4 TM helices, while subunit B contains only a single TM. The two B subunit TMs and TM3 and TM4 of two A subunits assemble to form two ion-conducting pathways (yellow area). (C) ExbB contains 3 TM helices per subunit and assembles with ExbD to form a single ion conduction pathway per stator. (D) TolQ/AglR is thought to contain 3 TMs per subunit and to assemble with TolR/AglQ in a 6:2 ratio to form two proton-conducting pathways.
CONCLUSIONS
The high structural diversity of the flagellar motor complex across contemporary bacterial taxa and recent improvements in cryotomography have enabled the first predictions of the quaternary structure of the ancestral complex. In particular, recent papers have revealed the extent of contemporary diversity (100–103)—and have used high-resolution imaging of larger rotors in species such as Borrelia burgdorferi to elucidate torque-generating interactions between the stator and rotor (104). Thus, the discovery of these new BFM basal-body structures has shed new light on the motor, even in long-studied model organisms. The inability to resolve an atomic structure of the transmembrane domains of the stator complex, however, means that we do not yet have a full understanding of the physical processes behind torque generation and ion selectivity.
Further structural understanding of the stators may shed light on the conformational rearrangements accompanying torque generation and the pore profile and residues involved in ion selectivity. New experiments in directed and experimental evolution will inform the molecular basis of evolution and adaptation in complex multicomponent systems.
ACKNOWLEDGMENTS
M. M. Tanaka is supported by Australian Research Council Future Fellowship FT140100398. M. A. B. Baker is supported by Australian Research Council Discovery Project DP190100497 and a CSIRO Future Platform in Synthetic Biology 2018 Project Grant.
- Copyright © 2020 American Society for Microbiology.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵