- Email: [email protected]
Positioning of bacterial chemoreceptors
TIMI-1189; No. of Pages 10
Review
Special issue: Microbial translocation
Positioning of bacterial chemoreceptors Christopher W. Jones and Judith P. Armitage Department of Biochemistry, University of Oxford, Oxford, OX1 3QU, UK
For optimum growth, bacteria must adapt to their environment, and one way that many species do this is by moving towards favourable conditions. To do so requires mechanisms to both physically drive movement and provide directionality to this movement. The pathways that control this directionality comprise chemoreceptors, which, along with an adaptor protein (CheW) and kinase (CheA), form large hexagonal arrays. These arrays can be formed around transmembrane receptors, resulting in arrays embedded in the inner membrane, or they can comprise soluble receptors, forming arrays in the cytoplasm. Across bacterial species, chemoreceptor arrays (both transmembrane and soluble) are localised to a variety of positions within the cell; some species with multiple arrays demonstrate this variety within individual cells. In many cases, the positioning pattern of the arrays is linked to the need for segregation of arrays between daughter cells on division, ensuring the production of chemotactically competent progeny. Multiple mechanisms have evolved to drive this segregation, including stochastic self-assembly, cellular landmarks, and the utilisation of ParA homologues. The variety of mechanisms highlights the importance of chemotaxis to motile species. Bacterial motility and taxis Bacteria are subject to large variations in environmental conditions and, to ensure optimum growth across changing environments, they must be able to respond and adapt to these changes. Such responses include altering patterns of gene expression and cellular behaviour. Many bacteria respond to unfavourable conditions by directed movement in a more favourable direction, a process known as taxis. The importance of motility is demonstrated in both the prevalence of motile species and the variety of approaches that bacteria have evolved to achieve movement [1]. These include various strategies for both moving across solid surfaces and for swimming through liquid media. Flavobacterium johnsoniae glides rapidly over surfaces, mediated by surface attachment and flow of proteins in the outer membrane causing translocation of the cell body [2]. Corresponding author: Armitage, J.P. ([email protected]). Keywords: bacterial chemotaxis; chemoreceptor array; cell division; array segregation; protein localisation. 0966-842X/ ß 2015 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tim.2015.03.004
Myxococcus xanthus uses two mechanisms for surface motility, distinct from each other and that of F. johnsoniae. S-motility depends on the extension, surface attachment, and retraction of type IV pili [3]. A-motility involves focal adhesion complexes attaching to the surface and moving the cell body forward [4]. Pseudomonas aeruginosa has different modes of motility for surface and aqueous movement. Surface movement is mediated by type IV pili [5], while movement through aqueous environments is mediated by rotation of flagella, which are semi-rigid helical filaments projecting from the cell [6]. While there are other mechanisms for swimming through aqueous media, the use of flagella is the most prevalent [1]. Movement is, in general, only useful if there is a mechanism for controlling that movement towards an improving environment. Swimming bacteria are constantly buffeted by their environment and to be able to swim in a positive direction, they need to sense and respond to the environment every second or so. Most bacteria are too small to sense spatial gradients and, therefore, use temporal sensing, comparing current conditions with those a few seconds ago. To allow a comparison, they also need a type of memory in the form of adaptation to current conditions. The best-studied chemosensory pathway is that of Escherichia coli (Figure 1). Escherichia coli is propelled through liquid by a bundle of four to six flagella, each powered by a transmembrane motor rotating in a counterclockwise direction. To reorientate the cell body, one or more of the flagellar motors switch the direction of rotation, causing the flagellar bundle to break apart and the cell to tumble and, thus, stop forward movement [7]. The random tumbling means that, when all the flagella are again rotating counter-clockwise, reforming the bundle, the cell swims off in a new, but random, direction [8]. Modulation of the frequency of these tumbling events in response to changing conditions generates chemotaxis. The switching of flagellar rotation from counter-clockwise to clockwise is controlled by a modified two-component pathway comprising transmembrane receptors (methyl-accepting chemotaxis proteins; MCPs), a histidine protein kinase (CheA), an adaptor protein (CheW), a response regulator (CheY), and two adaptation proteins, CheB and CheR. The MCPs are dimers that form trimers of dimers, which, along with CheW and CheA, pack hexagonally to form large arrays of thousands of proteins [9,10] (Figure 2). The hexagonal arrays of chemoreceptors are formed by the interactions of the CheW and CheA proteins across Trends in Microbiology xx (2015) 1–10
1
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
Repellent
Aractant
OM IM +CH3
CheR
–CH3
–P
MCPs
CheW
CheA CheB
CheY
–P
CheZ
Motor switching IM OM
TRENDS in Microbiology
Figure 1. Escherichia coli chemotaxis pathway. Transmembrane methyl-accepting chemotaxis proteins (MCPs) activate the kinase CheA in response to repellents or in the absence of attractants. CheA transfers phosphoryl groups to the response regulators CheY and CheB. When phosophorylated, CheY diffuses to the flagellar motor and causes switching of rotational direction. Phosphorylated CheB, along with CheR, changes the methylation state of the MCPs. The demethylation of MCPs in response to CheA activity reduces MCP sensitivity and returns the system to the pre-stimulus level. Abbreviations: IM, inner membrane; OM, outer membrane.
the cytoplasmic tips of the receptor trimers of dimers (Figure 2B). The presence of repellents or absence of attractants result in increased CheA activity and, thus, increased levels of phosphorylated CheY (CheY-P). CheY-P diffuses to, and interacts with, the flagellar motor proteins FliM and FliN, and causes motor switching [11,12]. To stop tumbling, the signal must be terminated. In E. coli, this is achieved through the action of CheZ, a CheY-P-specific phosphatase. CheA also phosphorylates the methylesterase CheB, which, once phosphorylated, removes methyl groups from specific glutamates on the MCPs. The action of CheB is countered by the methyltransferase CheR, which is constitutively active. MCPs with higher methylation have a greater ability to activate CheA, which in turn increases the activity of CheB and reduces the methylation state of the MCPs [13]. Thus, CheB and CheR work in concert to 2
generate adaptation, allowing cells to respond to relative changes in chemical concentrations across a range of background concentrations [14]. In E. coli, there are four structurally related MCPs forming mixed trimers of dimers. CheR is anchored by a long tether to the terminal domain of the two high-abundance receptor types (Tsr and Tar), allowing methylation and, thus, adaptation of other MCPs (e.g., Trg) in proximity. The elegant simplicity of the E. coli pathway makes it an attractive paradigm for chemotaxis. However, analysis of the genomes of motile bacterial species has revealed that over 50% contain multiple copies of each of the core chemotaxis genes (cheA, cheB, cheR, cheW, and cheY) [15]. The diversity of chemotaxis systems is also demonstrated by the proteins themselves; many species contain chemotaxis proteins that are not homologues of any of the E. coli proteins, such as the scaffold protein CheV, the deamidase CheD, and phosphatases CheC and CheX [16]. Many species also contain multiple homologues of MCPs [17], often including both transmembrane and cytoplasmic receptors [16]. Recently, it was shown that cytoplasmic receptors from Vibrio cholerae and Rhodobacter sphaeroides formed trimers of dimers packed hexagonally into large cytoplasmic arrays with the same spacing as transmembrane clusters [18]. This suggests that the packing of the chemoreceptors and the signalling proteins into large arrays is critical for correct signal transduction. Indeed, packing of the MCPs into these large arrays been shown to be important in creating gain in signal transduction through cooperative interactions between MCPs [19]. The methylation state of MCPs has been shown to alter their packing within the array, suggesting that adaptation alters the activity of CheA through changes in receptor packing [20]. The rate of adaptation in response to different chemical stimuli is the same, because of the packing of different receptors into the array and allosteric interactions between these receptors [21], further demonstrating the importance of these large arrays for proper chemotactic function. The position of flagella varies among species. Many have several randomly positioned flagella (peritrichous), as in E. coli, while others, such as P. aeruginosa, have a single polar flagellum. To move usefully through the environment, cells need to test the environment every second or so. Data suggest that the time taken for the 15-kDa CheYP to diffuse the length of an average cell is only approximately 100 ms [22] and, therefore, the positioning of the chemosensory array close to the flagellum is not essential for efficient chemosensory signalling. Electron cryotomography shows that, in most species, including E. coli, the membrane arrays are at, or close to, the cell poles, although in some polarly flagellate species, the arrays are found close to the flagellum. Data suggest that being motile and chemotactic is a survival advantage for many bacteria and, therefore, daughter cells need to not only be flagellate, but also inherit a chemosensory array on division to respond to the local environment. Therefore, it is critical that mechanisms exist to ensure that large arrays are inherited on division. If the arrays are localised to only one cell before division, then the daughter cell will
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
(A)
(B) (ii)
(i)
IM
MCPs
CheA/W baseplate
(C) (i)
(ii)
(D) (i)
(ii)
TRENDS in Microbiology
Figure 2. Chemosensory arrays. (A) Schematic of the polar transmembrane chemosensory array of Escherichia coli. (B) (i) Typical transmembrane chemosensory array viewed from the side. Sensory domains of methyl-accepting chemotaxis proteins (MCPs) in the periplasm, which is connected to a coiled-coil domain in the cytoplasm via a transmembrane region spanning the inner membrane (IM). MCPs form trimers of dimers (dark blue), which interact with CheA (red) and CheW (light blue) molecules at their distal tip. (ii) Top-down view of a typical array at the baseplate showing the hexagonal packing of MCP trimers of dimers (dark blue), CheA (red), and CheW (light blue). (C) (i) Electron cryotomography of Rhodobacter sphaeroides cytoplasmic chemosensory cluster (black arrowheads), showing a doublelayer architecture with CheA/W baseplates on the outside. (ii) Schematic of the array shown in (Ci). Colours as in ()B. (D) (i) Electron cryotomography of both cytoplasmic (CA) and transmembrane (MA) arrays in Vibrio cholerae. The transmembrane array shows typical architecture, with the CheA/W base plate distal to the IM. The cytoplasmic array forms a double-layer array with CheA/W baseplates on the outside. Abbreviation: OM, outer membrane. (ii) Schematic of the array shown in (Di). Adapted from [18] (Ci, Di).
not inherit a chemosensory network and, thus, will be nonchemotactic and unable to compete until a new array is expressed and assembled. If clusters are randomly distributed, then chance dictates that a proportion of daughter cells will also not inherit a chemosensory network; therefore, both the numbers and positioning of chemosensory clusters must be controlled. The focus of this review is how bacteria do this and how this relates to the cell cycle. Positioning patterns In many bacterial species, both electron cryotomography and fluorescence microscopy have shown that these large chemosensory arrays preferentially localise to the polar regions of the cell [23,24]. It has been proposed that this may be due to the array structure being more stable in curved membranes [25,26] or preference for the lipid com-
position of the poles [27]. The chemosensory proteins in E. coli tend to be polar and the size of the array appears to increase as the cells grow; therefore, an old pole will have a large cluster and, as the cell grows, additional clusters form, including one at the new pole, such that when the cell divides, each daughter cell has a cluster at its old pole (Figure 3A) [28]. Indeed, the lineage and age of cells can be followed by tracking the size of the polar cluster. In addition to these large polar clusters, E. coli also has smaller lateral clusters, probably formed as new receptors are expressed and inserted into the membrane [29]. By contrast, V. cholerae exhibits a strictly unipolar or bipolar distribution of clusters. The transition from unipolar to bipolar occurs before cell division (Figure 3C) [30]. Rhodobacter sphaeroides localises its transmembrane and soluble receptors to spatially distinct positions within the cell [31]. The transmembrane receptors are positioned in a 3
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
(B) (A)
(C)
(D)
(E)
(F)
1 µm TRENDS in Microbiology
Figure 3. Fluorescent images of bacterial chemosensory arrays. (A) Escherichia coli expressing YFP-CheR. Polar and lateral clusters can be seen. Scale bar = 1 mm. (B) Localisation of methyl-accepting chemotaxis proteins (MCPs) in Pseudomonas aeruginosa determined by immune fluorescence using anti-Tsr antibodies. Fluorescence is in green and the cell body in blue. (C) Vibrio cholerae cells expressing YFP-CheW1. The white line indicates a unipolar positioning pattern, and the red line a bipolar pattern. (D) Myxococcus xanthus cell expressing FrzCD-GFP. This soluble chemoreceptor forms clusters throughout the cytoplasm, excluding the polar regions. (E) Localisation of the transmembrane chemosensory cluster of Rhodobacter sphaeroides (YFP-CheW3, in green) and the Z-ring (FtsZ-CFP, in red). (F) The cytoplasmic chemosensory cluster of R. sphaeroides (TlpT-YFP). Adapted from [25] (A), [36] (B), [30] (C), [42] (D), and [32] (E).
pattern similar to that of E. coli, with predominantly polar clusters and some smaller lateral clusters (Figure 3E) [32]. The soluble receptors form a cluster in the cytoplasm that is positioned approximately mid-cell; as the cell cycle progresses, an additional cluster is formed and the two clusters are positioned at ¼ and 3/ 4 positions, such that each daughter cell inherits one cluster at the mid-cell on division (Figure 3F) [33]. The P. aeruginosa genome contains four sets of chemotaxis-like genes (che, che2, wsp, and pil-chp) and 26 chemoreceptor genes [34]. Gene products from the che locus localise to the polar regions of the cell, with some lateral clusters also present (Figure 3B); in dividing cells, clusters are often found at the old poles of the cells [35]. Che2 proteins also form clusters in the polar regions of the cell; however, co-expression of fluorescent fusions to Che and Che2 proteins (CheA and CheY2, respectively) showed no co-localisation of proteins from these two pathways [35]. Foci were either at opposite poles or adjacent to each other at the same pole; the deletion of the cheW gene from either the che or che2 locus resulted in proteins from that locus being unable to form clusters, but had no effect on proteins from the other locus. McpS is a soluble chemoreceptor; however, unlike the soluble receptors of R. sphaeroides, it does not localise to a site distinct from the membrane chemoreceptors, but localises to the poles, and is presumed to interact with transmembrane receptors and chemosensory proteins of one of the two systems that localise there (Che and Che2) [36]. The Wsp system is a chemosensory-like pathway that initiates biofilm formation in 4
response to surfaces [37]. WspA, the transmembrane receptor, forms clusters that localise throughout the membrane with no specificity for either polar or lateral membranes [38]. The gliding bacterium M. xanthus has eight sets of genes that encode putative chemosensory networks [39,40]. The proteins of these networks adopt a variety of positioning patterns. The soluble receptor FrzCD, which is responsible for controlling the frequency of reversals, forms clusters in the cytoplasm (Figure 3D) [41,42]. A receptor from the che7 operon, Mcp7, and an orphan receptor, McpJ, both localise to polar regions of the cell, even though Mcp7 is soluble and McpJ is transmembrane. Receptors from the other operons (all probably transmembrane) localise to clusters in the membrane along the length of the cell, with receptors from operons che4, che5, and che6 as well as three orphan receptors (McpH, McpM, and McpA) all co-localising [42]. These M. xanthus receptors clearly demonstrate the variety of positioning patterns of both soluble and transmembrane chemosensory clusters. They also demonstrate that arrays can comprise proteins from multiple operons as well as individual operons. In the species with multiple chemosensory pathways, localising them to different parts of the cell may represent an additional layer of regulation by preventing crosstalk. Unlike the above examples, where all the species undergo symmetric cell division, Caulobacter crescentus divides asymmetrically, with one cell remaining stalked and attached to a surface and the other being flagellated and chemotactic [43]. Accordingly, the chemosensory array only localises to the flagellate pole of the cell, resulting in
TIMI-1189; No. of Pages 10
Review only the swimmer cell inheriting a chemosensory network after division [44]. Across these species, it is clear that many transmembrane arrays preferentially localise to the poles of the cell, although often not exclusively. Soluble arrays have been found to localise both throughout the cell body and at the poles. It is currently unclear whether these positioning patterns are functionally important or simply a consequence of the systems used to segregate them on cell division. Mechanisms generating positioning patterns To generate and maintain these positioning patterns, bacteria must have mechanisms (either active or passive) to do so. Given the variety of positioning patterns, it is not unsurprising that several distinct mechanisms exist to generate these patterns. Stochastic self-assembly Escherichia coli has chemosensory clusters at the poles as well as small clusters periodically positioned along the cell length in the lateral membranes. Recent research has suggested that these clusters are formed and positioned by a process known as stochastic self-assembly [45,46]. Receptors are inserted individually into the membrane throughout the cell [47] and, once in the membrane, have two possible fates: either join an existing cluster or nucleate a new one. If the receptor is inserted close to an existing cluster, then there is a high probability that it will diffuse, encounter, and be absorbed into that cluster before it is able to nucleate a new one. Conversely, if it inserted far from the existing cluster, then probability favours nucleation, because the time required for a nucleation event to occur will be less than the time needed to encounter the existing cluster. Thus, the number of clusters in a given cell is dependent on the size of the cell, the concentration of receptors, and the diffusion coefficient of the receptors. Photoactivated localisation microscopy was used to position all molecules of the MCP Tar in individual E. coli cells. This revealed that, as well as the large clusters often found at the poles and smaller lateral clusters previously described, there are many smaller clusters and single proteins in the membrane [46]. The distribution of cluster sizes is continuous, suggesting that large clusters are formed by the gradual absorption of smaller clusters and single proteins rather than a sudden formation event from a pool of free nonoligomeric receptors. This study also revealed that the size of clusters increased with increasing distance from the large clusters at the poles, further supporting a stochastic self-assembly model in which the formation of large clusters is prevented in regions near an existing cluster. Simulations of stochastic self-assembly systems have shown that they are capable of producing positioning patterns that closely resemble those seen for chemosensory clusters in E. coli [48]. It has also been suggested that, once clusters have been formed, they are held in position by anchoring to unidentified structures at future division sites to maintain the spatial separation. Having a cluster at the division site ensures that, after division, the new poles of the daughter cells inherit a small cluster (Figure 4A) [45]. This transition from lateral to
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
polar clusters through cell division, along with the potential stabilising effect of the curved membrane [26,49], explains the positioning of the oldest and, therefore, largest clusters at the poles. This provides an elegant link between the cell cycle and cluster numbers and positioning; as the cell nears division, it will have approximately doubled in size, allowing for an increase in cluster numbers, which are formed spatially distant from each other, thus ensuring that both daughter cells inherit clusters on division. The positioning of the transmembrane chemosensory cluster of R. sphaeroides at first glance seems similar to that of E. coli, with polar and lateral clusters; however, recent work shows that, unlike E. coli, the clusters appear excluded from the division sites characterised by the Z-ring (a ring of polymerising FtsZ, which forms the septation site) (Figure 4B) [32]. The lateral clusters of E. coli show limited diffusion [25], but those of R. sphaeroides diffuse; indeed, in sphaeroplasts, the clusters show no organised localisation pattern and are present in random patches all over the cell. This suggests that polar clusters are formed by the movement of lateral clusters that are trapped in the curved membrane of the poles. While clusters merge with other diffusing clusters suggesting that stochastic selfassembly has a role in R. sphaeroides cluster formation and positioning, clusters also move away from arrays. The use of stochastic self-assembly in conjunction with other mechanisms represents a way in which to diversify the positioning patterns generated by stochastic self-assembly, while still linking array numbers and cell size. Cellular landmarks Only nonmotile stalked C. crescentus cells are able to initiate DNA replication and cell division. Cell division results in one stalked cell plus one motile and chemotactic swarmer cell [50]. For C. crescentus to produce two distinct progeny, one motile and one nonmotile, it must use a strategy for positioning flagella and chemosensory pathways that is different from that of E. coli. As the cell cycle progresses, the stalked cell elongates, with asymmetry of the progeny imposed during this predivisional stage. The new pole is marked by the presence of the protein TipN (Figure 4D), a transmembrane protein with several coiled-coil domains, which recruits TipF, a transmembrane c-di-GMP phosphodiesterase required for flagellum assembly [51,52]. During this early predivisional stage, TipN and TipF mediate the placement of the flagellum, pili, and chemosensory array at the new (nonstalked) pole. In late predivisional cells, TipN moves from the new pole to the site of constriction, followed by TipF, resulting in both these proteins being present at the new poles of the progeny [52]. Localisation of the flagellum, pili, and the chemosensory array to the new pole of the dividing cell ensures that they are inherited only by the swarmer cell. However, both the swarmer and stalked cells inherit TipN and TipF at their new poles. To ensure that the stalked cell does not have either a flagellum or a chemosensory system until it enters the predivisional stage, the expression of their genes is controlled temporally [53] and the activity of TipF is controlled by cell cycledependent c-di-GMP levels. 5
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
(A)
(B)
E. coli
R. sphaeroides
Key:
Chemoreceptor arrays Soluble chemoreceptor array Z-ring
(C)
(D)
V. cholerae Key:
HubP ParC ParP
C. crescentus Key:
TipN
TRENDS in Microbiology
Figure 4. Systems used to create positioning patterns. (A) Newborn Escherichia coli cells have one or more chemosensory arrays, the largest of which localises to the old pole. As the cell grows, arrays are formed at the new pole, as well as in the lateral membranes. On cell division, the polar clusters are positioned at the old poles of the daughter cells. The positioning of the lateral arrays is such that, on division, they will be positioned at the new poles of the daughter cells. Here, the preferential localisation at the poles, as well as the spatial limit on new cluster formation, ensures that at the least each daughter cell will inherit an array at its old pole. (B) Rhodobacter sphaeroides has two distinct chemosensory arrays, one polar and transmembrane (green), the other soluble (blue). The transmembrane array forms a large array at the old pole of the cell, with lateral clusters and a new pole array forming as the cell elongates. As the cell divides, lateral clusters are excluded from the area surround the Z-ring (red). The soluble array is positioned at mid-cell; as the cell grows, the array is split and the two resultant arrays are positioned at ¼ and 3/4 positions, so that, when the cell divides, both daughter cell inherit a single cluster at mid-cell. (C) Newborn Vibrio cholerae cells have a single flagella and a chemosensory array at their old pole. The polarorganising protein HubP is present at both poles, but, as the cell grows, ParC (a ParA homologue) is recruited to the HubP at the new pole. ParC in turn recruits ParP, which is able to stabilise the formation of a second chemosensory array at this pole. As the cell divides, HubP also localises to the site of constriction, resulting in daughter cells with HubP at both new and old poles. HubP recruits multiple ParA homologues, including FlhG, which ensures that the daughter cells have polar flagella. (D) The new pole of Caulobacter crescentus is marked by the protein TipN. When a swimming C. crescentus cell attaches to a surface, it sheds its flagellum, degrades its chemosensory array, and forms a stalk. As the stalked cell initiates cell division, the chemosensory genes are expressed and their localisation is directed by TipN to the new pole; this is also true of the flagellum. Just before division, TipN moves to the mid-cell, ensuring that both daughter cells will have TipN present at their new poles.
6
TIMI-1189; No. of Pages 10
Review When a swarmer cell encounters a surface, it sheds its flagellum and forms a stalk at the previously flagellate pole (the old pole) [50]. During this swarmer–stalked transition, the protease ClpXP degrades the chemosensory proteins [54,55]. Thus, once the cell has differentiated fully into a stalked cell and started cell division, any newly expressed chemotaxis proteins will form a de novo array at the swarmer (new) pole. While the exact nature of the interaction between the chemotaxis array and TipN/TipF remains unclear, TipN/ TipF might act as an anchor about which the array forms using a diffusion capture mechanism. Although this anchor generates the asymmetry in the predivisional cell, additional levels of control, through repression of expression plus degradation by proteases, exist to ensure that chemosensory arrays are not present in stalked cells postdivision. ParA homologues Several recent studies have shown that positioning patterns of chemosensory arrays in some bacterial species are dependent on homologues of ParA proteins. ParA proteins were initially identified as part of a tripartite system used to segregate low copy-number plasmids [56]. These systems comprise: ParA, a ‘deviant’ Walker A ATPase [57]; a parS nucleotide sequence in the plasmid; and ParB, which binds parS and stimulates the ATPase activity of ParA [58]. ATP-bound ParA dimerises [59] and subsequently nonspecifically binds DNA, coating the surface of the chromosome [60]. ParB, bound to the plasmid via the parS site, stimulates the ATPase activity of ParA and, thus, releases it from the DNA; the progression to the next chromosomebound dimer moves ParB and the associated plasmid over the chromosome surface [61]. Several models have been put forward for actual mechanisms by which this process generates segregation of the plasmids, including filament pulling [62] and diffusion ratchet [61]. Although discussion of these models is beyond the scope of this review, the core mechanism that underpins the models relies on a ParA protein with an ATP/ADP switch controlling localisation behaviour and a partner protein (here ParB) that modulates this switch. ParA proteins have recently been shown to be involved in the positioning and segregation of many macromolecular assemblies within bacterial cells (reviewed in [63]). These include the origins of replication of chromosomes in many species as well as several protein complexes, including the Z-ring, carbon fixation complexes, cellulose biosynthesis machinery, and chemosensory arrays in Vibrio spp. and R. sphaeroides. Vibrio spp. Both V. cholerae and Vibrio parahaemolyticus chemosensory clusters localise close to the single polar flagellum, with a cell cycle-dependent shift from unipolar to bipolar positioning. Both the uni- and bipolar positioning is dependent on a ParA homologue, termed ParC, whose gene is encoded within the chemotaxis operon [30]. The use of fluorescent fusions revealed that ParC switches from a unipolar pattern to a bipolar one, and is then followed by the chemotaxis proteins CheW1 and CheY3, which form
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
the second cluster de novo. ATP-locked ParC mutants form polar foci, as well as some nonpolar foci, whereas a mutant defective in ATP binding is diffuse throughout the cytoplasm and unable to form foci. Both of these mutants resulted in aberrant positioning of the chemotaxis proteins, indicating that the cycle of ATP hydrolysis and ADP/ATP nucleotide exchange is required. The identification of a pole-organising factor, HubP, helps explain ParC localisation, with ParC then determining the positioning of chemotaxis array [64]. Yamaichi et al. showed that HubP was required for the correct localisation pattern of ParC; it is also required for the polar localisation of ParA of chromosome I (ParA1) and a ParA homologue involved in the regulation of flagellum assembly (FlhG). Fluorescent fusions to HubP localised to the poles as well as the septation site before division to mark the new poles after division. Mutants lacking hubP are unable to sequester the origin of replication of chromosome I at the poles, cannot correctly position the chemosensory cluster, and produce multiple mislocalised flagella, rather than a single one. In these DhubP cells, mislocalised CheY3 remains colocalised with ParC, indicating that HubP recruits ParC, which in turn interacts with the chemosensory array. HubP is a large protein comprising multiple domains, including a single transmembrane domain in the N-terminal region and a large C-terminal cytoplasmic region. Interestingly, truncations and deletions of domains revealed that different domains of HubP are required for the polar localisation of ParA1, FlhG, and ParC. Bacterial two-hydrid assays and co-expression of fluorescent fusions in E. coli showed that both ParA1 and FlhG directly interact with HubP; however, ParC does not, indicating an intermediary protein(s) not present in E. coli. This suggests that interactions that tether these ParA proteins to HubP are distinct and that each system evolved separately to use HubP as a polar anchor. A subsequent study identified a further protein involved in the polar localisation of the chemosensory clusters in both V. cholerae and V. parahaemolyticus [65]. Plasmid parA genes are encoded immediately upstream of their partner proteins, parB. Ringgaard et al. identified an unannotated gene immediately downstream of parC termed parP. ParP is required for the correct positioning of chemotaxis proteins (CheW, CheA, and an MCP in V. parahaemolyticus; CheY3 and CheW1 in V. cholerae). ParP contains a CheW-like domain and a proline-rich N-terminal region, and was shown to interact with both CheA and ParC in bacterial two hybrid assays. These assays further revealed that ParP specifically interacted with a localisation and inheritance domain (LID) within CheA. Bioinformatic analysis of cheA genes showed that LID domains were only present in cheAs that were associated with a parC and a parP [66]. ParP co-localises with both ParC and chemotaxis proteins following the same cell cycle-dependent unipolar/bipolar positioning pattern. Fluorescence recovery after photobleaching (FRAP) experiments indicated that ParP is able to stabilise CheA within the array, and the co-expression of a membrane-targeted ParP and CheA in E. coli demonstrated that ParP is sufficient to induce CheA clustering. 7
TIMI-1189; No. of Pages 10
Review Together, these studies suggest a system in which HubP acts as a polar marker and anchor for several ParA proteins, including ParC. As the cell cycle progresses, ParC is recruited to HubP at the new pole, bringing with it ParP and resulting in bipolar localisation of both of these proteins. By interacting with CheA, ParP appears to drive nucleation and stabilises a new second chemosensory cluster at the new pole (Figure 4C). Thus, this creates a bipolar pattern of positioning that ensures that each daughter cell inherits a chemosensory network on division. The mechanism by which HubP localises to the pole has yet to be determined and is of interest, given the number of proteins that rely on its positioning to determine their own. The relation between ParC and ParP is also of interest, specifically whether ParP can drive ParC ATPase activity or if another protein is required for this and ParP is just the cargo. Rhodobacter sphaeroides The soluble chemosensory cluster of R. sphaeroides also shows cell cycle-dependent changes in number and position. Newly born cells have a single cluster positioned approximately at the mid-cell; before cell division, a second cluster is formed and these two clusters are positioned at the ¼ and 3/ 4 positions (Figure 4B) [33]. Both the positioning of the clusters and the formation of a second cluster are dependent on a ParA ATPase homologue, PpfA, encoded in the same operon as that of the proteins that comprise the cluster. Although PpfA localises to the DNA of the chromosome surface, it also co-localises with the cluster [67]. This contrasts with ParC, which segregates first and then recruits a second chemosensory cluster. As with ParA proteins from plasmid segregation systems, PpfA binds the nucleoid in an ATP-dependent manner, and mutants unable to bind either ATP or DNA are unable to co-localise with the chemosensory cluster. Fluorescent fusions to a mutant unable to hydrolyse ATP and, thus, locked in an ATP-bound state form brighter foci co-localising with the cluster; however, they are unable to duplicate or segregate the cluster, demonstrating that the ATP hydrolysis cycle is required for segregation. This suggests that the R. sphaeroides chemosensory cluster uses the segregating chromosomes as the mechanism for ensuring that each daughter inherits a cytoplasmic array. Roberts et al. also identified the partner protein for PpfA, a soluble chemoreceptor, TlpT, which forms part of the soluble cluster [67]. TlpT, along with the parB and parP genes, is encoded immediately downstream of its partner parA gene ppfA. TlpT is required for cluster formation; fluorescent fusions of the other proteins in the cytoplasmic chemosensory cluster display diffuse fluorescence in a DtlpT background, while cephalexin-treated filamentous cells that normally show evenly spaced chemosensory clusters had one large, wandering cluster in a DppfA background [68]. TlpT lacking the N-terminal 120 amino acids is still able to form clusters, but loses the co-localisation with PpfA and the ability to form multiple clusters or segregate, indicating that this region is responsible for interaction with PpfA [67]. While this suggests that the N-terminal domain of TlpT acts as a ParB homologue to drive PpfA ATPase activity and segregation, this remains to be unexplored, 8
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
as does the mechanisms involved in splitting the chemosensory array and segregating one with the segregating nucleoid. Interestingly, both the segregation systems for V. cholerae and R. sphaeroides depend not only on a ParA protein, but also adapted chemotaxis proteins (TlpT and ParP) to function at the very least as binding partners and possibly performing the same function as ParB in activating the ATPase activity of ParA. Given that over 35% of sequenced genomes appear to encode a ParA homologue in one of their chemotaxis operons [15,30], it would be interesting to determine whether this adaptation of chemotaxis proteins to partner ParA proteins is common. These two examples also illustrate that the use of ParA proteins can generate different positioning patterns and do so by different mechanisms, ParC creating a landmark for the formation of a second de novo array and PpfA splitting a single array into two. Concluding remarks It is clear that, for effective chemotaxis, allowing sensing of changes across a range of background concentrations, the proteins of the chemosensory pathways need to assemble into hexagonal arrays of many hundreds to thousands of sensory and signal transducing proteins, and this arrangement is critical regardless of whether the signalling pathway is transmembrane or soluble. Given that an ability to sense and respond is essential for motile bacteria to compete after division, each daughter must inherit a chemosensory cluster on division. It takes tens of minutes to express the proteins to assemble a new cluster, and the chemosensory behaviour of populations is measurably impaired where a cluster is not inherited and half the population need to synthesise de novo. It is likely that the absolute number of proteins is not critical, but the ratio of proteins and their assembly into the interacting array is, otherwise all the proteins could freely diffuse and each cell would simply inherit a percentage on division, which is not the case. The variety of mechanisms that bacterial species have evolved to ensure that progeny inherit chemotaxis arrays and, thus, are chemotactic after division demonstrates its importance. Although many transmembrane arrays have an apparent preference for the polar regions of the cells, this may reflect the curvature and phospholipid properties of the membrane, trapping the developing array. Concentrating the developing array at the poles ensures each daughter will inherit on division, but clearly some species appear to have no preference for the poles, suggesting that preference for the poles is not an innate property of all MCPs. Where species have more than one chemosensory pathway, these also show variation in positioning, possibly reflecting the function of the different pathways. Pathways that colocalise might allow direct integration of responses from multiple signals, while pathways that are clearly separate will allow the isolation of pathways and prevent crosstalk. In some species, there is a clear link between the numbers and positioning of flagella, and of chemosensory arrays. Escherichia coli have multiple flagella randomly positioned and there is no correlation between flagella position and array position. This is also the case for R. sphaeroides, which randomly positions its single flagellum, and the
TIMI-1189; No. of Pages 10
Review chemosensory membrane arrays, while predominantly polar, diffuse as large arrays around the membrane. In both cases, the formation of randomly positioned arrays ensures inheritance. However, species that are monotrichous, such as V. cholerae and C. crescentus, use polar markers to position both the array and the flagellum. It is worth noting that, while the distance between the receptor array and the flagellar motor should have no effect of chemosensory signalling, given the bacterial size, there is evidence that in Vibrio spp. the loss of positioning of the array in ParP mutants does have some effect of chemotactic ability for reasons that are still unclear. This ParA-like system has also been adapted to position the large soluble chemosensory arrays in species such as R. sphaeroides. The evolution of the tripartite ParA-like ATPase system to position proteins and plasmids in bacterial cells ensuring inheritance on division reveals how critical spatial organisation is in bacterial cells: they are not membrane sacs of diffusing enzymes, but ordered and organised organisms. This is neatly illustrated by the chemosensory pathways: the proteins need to be in very large hexagonal arrays to allow the sensitivity required with changing gradients, but the time taken to synthesise an array on division would mean that a daughter cell would not compete. Each cell must inherit a large cluster on division and, while for some species a diffusion capture system ensures a distribution of arrays, ParA-like systems have evolved for other species and for soluble arrays. Increasing numbers of proteins are being identified that arrange in large complexes and arrays and it will be interesting to see whether ensure inheritance in these similar systems. Acknowledgements The Rhodobacter sphaeroides work performed in J.P.A.’s group is funded by the Biotechnology and Biological Sciences Research Council (BBSRC).
References 1 Jarrell, K.F. and McBride, M.J. (2008) The surprisingly diverse ways that prokaryotes move. Nat. Rev. Microbiol. 6, 466–476 2 Nakane, D. et al. (2013) Helical flow of surface protein required for bacterial gliding motility. Proc. Natl. Acad. Sci. U.S.A. 110, 11145–11150 3 Bulyha, I. et al. (2009) Regulation of the type IV pili molecular machine by dynamic localization of two motor proteins. Mol. Microbiol. 74, 691– 706 4 Mignot, T. et al. (2007) Evidence that focal adhesion complexes power bacterial gliding motility. Science 315, 853–856 5 Mattick, J.S. (2002) Type IV pili and twitching motility. Annu. Rev. Microbiol. 56, 289–314 6 Deforet, M. et al. (2014) Hyperswarming adaptations in a bacterium improve collective motility without enhancing single cell motility. Soft Matter 10, 2405–2413 7 Darnton, N.C. et al. (2007) On torque and tumbling in swimming Escherichia coli. J. Bacteriol. 189, 1756–1764 8 Berg, H.C. and Brown, D.A. (1972) Chemotaxis in Escherichia coli analysed by three-dimensional tracking. Nature 239, 500–504 9 Zhang, P. et al. (2007) Direct visualization of Escherichia coli chemotaxis receptor arrays using cryo-electron microscopy. Proc. Natl. Acad. Sci. U.S.A. 104, 3777–3781 10 Briegel, A. et al. (2012) Bacterial chemoreceptor arrays are hexagonally packed trimers of receptor dimers networked by rings of kinase and coupling proteins. Proc. Natl. Acad. Sci. U.S.A. 109, 3766–3771 11 Welch, M. et al. (1993) Phosphorylation-dependent binding of a signal molecule to the flagellar switch of bacteria. Proc. Natl. Acad. Sci. U.S.A. 90, 8787–8791 12 Sarkar, M.K. et al. (2010) Chemotaxis signaling protein CheY binds to the rotor protein FliN to control the direction of flagellar rotation in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 107, 9370–9375
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
13 Hazelbauer, G.L. et al. (2008) Bacterial chemoreceptors: highperformance signaling in networked arrays. Trends Biochem. Sci. 33, 9–19 14 Lazova, M.D. et al. (2011) Response rescaling in bacterial chemotaxis. Proc. Natl. Acad. Sci. U.S.A. 108, 13870–13875 15 Hamer, R. et al. (2010) Deciphering chemotaxis pathways using cross species comparisons. BMC Syst. Biol. 4, 3 16 Wuichet, K. and Zhulin, I.B. (2010) Origins and diversification of a complex signal transduction system in prokaryotes. Sci. Signal. 3, ra50 17 Alexander, R.P. and Zhulin, I.B. (2007) Evolutionary genomics reveals conserved structural determinants of signaling and adaptation in microbial chemoreceptors. Proc. Natl. Acad. Sci. U.S.A. 104, 2885– 2890 18 Briegel, A. et al. (2014) Structure of bacterial cytoplasmic chemoreceptor arrays and implications for chemotactic signaling. Elife 3, e02151 19 Sourjik, V. (2004) Receptor clustering and signal processing in E. coli chemotaxis. Trends Microbiol. 12, 569–576 20 Irieda, H. et al. (2006) Control of chemotactic signal gain via modulation of a pre-formed receptor array. J. Biol. Chem. 281, 23880–23886 21 Krembel, A.K. et al. (2015) Universal response-adaptation relation in bacterial chemotaxis. J. Bacteriol. 197, 307–313 22 Segall, J.E. et al. (1985) Chemotactic signaling in filamentous cells of Escherichia coli. J. Bacteriol. 161, 51–59 23 Briegel, A. et al. (2009) Universal architecture of bacterial chemoreceptor arrays. Proc. Natl. Acad. Sci. U.S.A. 106, 17181–17186 24 Sourjik, V. and Armitage, J.P. (2010) Spatial organization in bacterial chemotaxis. EMBO J. 29, 2724–2733 25 Thiem, S. et al. (2007) Positioning of chemosensory clusters in E. coli and its relation to cell division. EMBO J. 26, 1615–1623 26 Hall, B. et al. (2012) Mechanism of bacterial signal transduction revealed by molecular dynamics of Tsr dimers and trimers of dimers in lipid vesicles. PLoS Comput. Biol. 8, e1002685 27 Mileykovskaya, E. and Dowhan, W. (2000) Visualization of phospholipid domains in Escherichia coli by using the cardiolipinspecific fluorescent dye 10-N-nonyl acridine orange. J. Bacteriol. 182, 1172–1175 28 Ping, L. et al. (2008) Tsr-GFP accumulates linearly with time at cell poles, and can be used to differentiate ‘old’ versus ‘new’ poles, in Escherichia coli. Mol. Microbiol. 69, 1427–1438 29 Sourjik, V. and Berg, H.C. (2000) Localization of components of the chemotaxis machinery of Escherichia coli using fluorescent protein fusions. Mol. Microbiol. 37, 740–751 30 Ringgaard, S. et al. (2011) A family of ParA-like ATPases promotes cell pole maturation by facilitating polar localization of chemotaxis proteins. Genes Dev. 25, 1544–1555 31 Wadhams, G.H. et al. (2003) Targeting of two signal transduction pathways to different regions of the bacterial cell. Mol. Microbiol. 50, 763–770 32 Chiu, S-W et al. (2013) Positioning of chemosensory proteins and FtsZ through the Rhodobacter sphaeroides cell cycle. Mol. Microbiol. 90, 322–337 33 Thompson, S.R. et al. (2006) The positioning of cytoplasmic protein clusters in bacteria. Proc. Natl. Acad. Sci.U.S.A. 103, 8209 34 Stover, C.K. et al. (2000) Complete genome sequence of Pseudomonas aeruginosa PAO1, an opportunistic pathogen. Nature 406, 959–964 35 Gu¨vener, Z.T. et al. (2006) Two different Pseudomonas aeruginosa chemosensory signal transduction complexes localize to cell poles and form and remould in stationary phase. Mol. Microbiol. 61, 106–118 36 Bardy, S.L. and Maddock, J.R. (2005) Polar localization of a soluble methyl-accepting protein of Pseudomonas aeruginosa. J. Bacteriol. 187, 7840–7844 37 Hickman, J.W. et al. (2005) A chemosensory system that regulates biofilm formation through modulation of cyclic diguanylate levels. Proc. Natl. Acad. Sci. U.S.A. 102, 14422–14427 38 O’Connor, J.R. et al. (2012) Surface sensing and lateral subcellular localization of WspA, the receptor in a chemosensory-like system leading to c-di-GMP production. Mol. Microbiol. 86, 720–729 39 Shi, X. et al. (2008) Bioinformatics and experimental analysis of proteins of two-component systems in Myxococcus xanthus. J. Bacteriol. 190, 613–624 9
TIMI-1189; No. of Pages 10
Review 40 Goldman, B.S. et al. (2006) Evolution of sensory complexity recorded in a myxobacterial genome. Proc. Natl. Acad. Sci. U.S.A. 103, 15200– 15205 41 Mauriello, E.M.F. et al. (2009) Localization of a bacterial cytoplasmic receptor is dynamic and changes with cell-cell contacts. Proc. Natl. Acad. Sci. U.S.A. 106, 4852–4857 42 Moine, A. et al. (2014) Functional organization of a multimodular bacterial chemosensory apparatus. PLoS Genet. 10, e1004164 43 Gomes, S.L. and Shapiro, L. (1984) Differential expression and positioning of chemotaxis methylation proteins in Caulobacter. J. Mol. Biol. 178, 551–568 44 Alley, M.R. et al. (1992) Polar localization of a bacterial chemoreceptor. Genes Dev. 6, 825–836 45 Thiem, S. and Sourjik, V. (2008) Stochastic assembly of chemoreceptor clusters in Escherichia coli. Mol. Microbiol. 68, 1228–1236 46 Greenfield, D. et al. (2009) Self-organization of the Escherichia coli chemotaxis network imaged with super-resolution light microscopy. PLoS Biol. 7, e1000137 47 Shiomi, D. et al. (2006) Helical distribution of the bacterial chemoreceptor via colocalization with the Sec protein translocation machinery. Mol. Microbiol. 60, 894–906 48 Wang, H. et al. (2008) Self-organized periodicity of protein clusters in growing bacteria. Phys. Rev. Lett. 101, 218101 49 Kim, K.K. et al. (1999) Four-helical-bundle structure of the cytoplasmic domain of a serine chemotaxis receptor. Nature 400, 787–792 50 Jensen, R.B. et al. (2002) Dynamic localization of proteins and DNA during a bacterial cell cycle. Nat. Rev. Mol. Cell Biol. 3, 167–176 51 Lam, H. et al. (2006) A landmark protein essential for establishing and perpetuating the polarity of a bacterial cell. Cell 124, 1011–1023 52 Huitema, E. et al. (2006) Bacterial birth scar proteins mark future flagellum assembly site. Cell 124, 1025–1037 53 Laub, M.T. et al. (2000) Global analysis of the genetic network controlling a bacterial cell cycle. Science 290, 2144–2148 54 Potocka, I. et al. (2002) Degradation of a Caulobacter soluble cytoplasmic chemoreceptor is ClpX dependent. J. Bacteriol. 184, 6635–6641 55 Tsai, J.W. and Alley, M.R. (2001) Proteolysis of the Caulobacter McpA chemoreceptor is cell cycle regulated by a ClpX-dependent pathway. J. Bacteriol. 183, 5001–5007
10
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
56 Austin, S. and Abeles, A. (1983) Partition of unit-copy miniplasmids to daughter cells: I. P1 and F miniplasmids contain discrete, interchangeable sequences sufficient to promote equipartition. J. Mol. Biol. 169, 353–372 57 Leipe, D.D. et al. (2002) Classification and evolution of P-loop GTPases and related ATPases. J. Mol. Biol. 317, 41–72 58 Bouet, J.Y. and Funnell, B.E. (1999) P1 ParA interacts with the P1 partition complex at parS and an ATP-ADP switch controls ParA activities. EMBO J. 18, 1415–1424 59 Leonard, T. et al. (2005) Bacterial chromosome segregation: structure and DNA binding of the Soj dimer: a conserved biological switch. EMBO J. 24, 270–282 60 Hester, C.M. and Lutkenhaus, J. (2007) Soj (ParA) DNA binding is mediated by conserved arginines and is essential for plasmid segregation. Proc. Natl. Acad. Sci. U.S.A. 104, 20326–32031 61 Vecchiarelli, A.G. et al. (2010) ATP control of dynamic P1 ParA-DNA interactions: a key role for the nucleoid in plasmid partition. Mol. Microbiol. 78, 78–91 62 Ringgaard, S. et al. (2009) Movement and equipositioning of plasmids by ParA filament disassembly. Proc. Natl. Acad. Sci. U.S.A. 106, 19369–19374 63 Lutkenhaus, J. (2012) The ParA/MinD family puts things in their place. Trends Microbiol. 20, 411–418 64 Yamaichi, Y. et al. (2012) A multidomain hub anchors the chromosome segregation and chemotactic machinery to the bacterial pole. Genes Dev. 26, 2348–2360 65 Ringgaard, S. et al. (2014) ParP prevents dissociation of CheA from chemotactic signaling arrays and tethers them to a polar anchor. Proc. Natl. Acad. Sci. U.S.A. 111, E255–E264 66 Davis, N.J. et al. (2013) De- and repolarization mechanism of flagellar morphogenesis during a bacterial cell cycle. Genes Dev. 27, 2049–2062 67 Roberts, M.J. et al. (2012) ParA-like protein uses nonspecific chromosomal DNA binding to partition protein complexes. Proc. Natl. Acad. Sci. U.S.A. 109, 6698–6703 68 Wadhams, G.H. et al. (2005) Requirements for chemotaxis protein localization in Rhodobacter sphaeroides. Mol. Microbiol. 58, 895–902
Review
Special issue: Microbial translocation
Positioning of bacterial chemoreceptors Christopher W. Jones and Judith P. Armitage Department of Biochemistry, University of Oxford, Oxford, OX1 3QU, UK
For optimum growth, bacteria must adapt to their environment, and one way that many species do this is by moving towards favourable conditions. To do so requires mechanisms to both physically drive movement and provide directionality to this movement. The pathways that control this directionality comprise chemoreceptors, which, along with an adaptor protein (CheW) and kinase (CheA), form large hexagonal arrays. These arrays can be formed around transmembrane receptors, resulting in arrays embedded in the inner membrane, or they can comprise soluble receptors, forming arrays in the cytoplasm. Across bacterial species, chemoreceptor arrays (both transmembrane and soluble) are localised to a variety of positions within the cell; some species with multiple arrays demonstrate this variety within individual cells. In many cases, the positioning pattern of the arrays is linked to the need for segregation of arrays between daughter cells on division, ensuring the production of chemotactically competent progeny. Multiple mechanisms have evolved to drive this segregation, including stochastic self-assembly, cellular landmarks, and the utilisation of ParA homologues. The variety of mechanisms highlights the importance of chemotaxis to motile species. Bacterial motility and taxis Bacteria are subject to large variations in environmental conditions and, to ensure optimum growth across changing environments, they must be able to respond and adapt to these changes. Such responses include altering patterns of gene expression and cellular behaviour. Many bacteria respond to unfavourable conditions by directed movement in a more favourable direction, a process known as taxis. The importance of motility is demonstrated in both the prevalence of motile species and the variety of approaches that bacteria have evolved to achieve movement [1]. These include various strategies for both moving across solid surfaces and for swimming through liquid media. Flavobacterium johnsoniae glides rapidly over surfaces, mediated by surface attachment and flow of proteins in the outer membrane causing translocation of the cell body [2]. Corresponding author: Armitage, J.P. ([email protected]). Keywords: bacterial chemotaxis; chemoreceptor array; cell division; array segregation; protein localisation. 0966-842X/ ß 2015 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tim.2015.03.004
Myxococcus xanthus uses two mechanisms for surface motility, distinct from each other and that of F. johnsoniae. S-motility depends on the extension, surface attachment, and retraction of type IV pili [3]. A-motility involves focal adhesion complexes attaching to the surface and moving the cell body forward [4]. Pseudomonas aeruginosa has different modes of motility for surface and aqueous movement. Surface movement is mediated by type IV pili [5], while movement through aqueous environments is mediated by rotation of flagella, which are semi-rigid helical filaments projecting from the cell [6]. While there are other mechanisms for swimming through aqueous media, the use of flagella is the most prevalent [1]. Movement is, in general, only useful if there is a mechanism for controlling that movement towards an improving environment. Swimming bacteria are constantly buffeted by their environment and to be able to swim in a positive direction, they need to sense and respond to the environment every second or so. Most bacteria are too small to sense spatial gradients and, therefore, use temporal sensing, comparing current conditions with those a few seconds ago. To allow a comparison, they also need a type of memory in the form of adaptation to current conditions. The best-studied chemosensory pathway is that of Escherichia coli (Figure 1). Escherichia coli is propelled through liquid by a bundle of four to six flagella, each powered by a transmembrane motor rotating in a counterclockwise direction. To reorientate the cell body, one or more of the flagellar motors switch the direction of rotation, causing the flagellar bundle to break apart and the cell to tumble and, thus, stop forward movement [7]. The random tumbling means that, when all the flagella are again rotating counter-clockwise, reforming the bundle, the cell swims off in a new, but random, direction [8]. Modulation of the frequency of these tumbling events in response to changing conditions generates chemotaxis. The switching of flagellar rotation from counter-clockwise to clockwise is controlled by a modified two-component pathway comprising transmembrane receptors (methyl-accepting chemotaxis proteins; MCPs), a histidine protein kinase (CheA), an adaptor protein (CheW), a response regulator (CheY), and two adaptation proteins, CheB and CheR. The MCPs are dimers that form trimers of dimers, which, along with CheW and CheA, pack hexagonally to form large arrays of thousands of proteins [9,10] (Figure 2). The hexagonal arrays of chemoreceptors are formed by the interactions of the CheW and CheA proteins across Trends in Microbiology xx (2015) 1–10
1
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
Repellent
Aractant
OM IM +CH3
CheR
–CH3
–P
MCPs
CheW
CheA CheB
CheY
–P
CheZ
Motor switching IM OM
TRENDS in Microbiology
Figure 1. Escherichia coli chemotaxis pathway. Transmembrane methyl-accepting chemotaxis proteins (MCPs) activate the kinase CheA in response to repellents or in the absence of attractants. CheA transfers phosphoryl groups to the response regulators CheY and CheB. When phosophorylated, CheY diffuses to the flagellar motor and causes switching of rotational direction. Phosphorylated CheB, along with CheR, changes the methylation state of the MCPs. The demethylation of MCPs in response to CheA activity reduces MCP sensitivity and returns the system to the pre-stimulus level. Abbreviations: IM, inner membrane; OM, outer membrane.
the cytoplasmic tips of the receptor trimers of dimers (Figure 2B). The presence of repellents or absence of attractants result in increased CheA activity and, thus, increased levels of phosphorylated CheY (CheY-P). CheY-P diffuses to, and interacts with, the flagellar motor proteins FliM and FliN, and causes motor switching [11,12]. To stop tumbling, the signal must be terminated. In E. coli, this is achieved through the action of CheZ, a CheY-P-specific phosphatase. CheA also phosphorylates the methylesterase CheB, which, once phosphorylated, removes methyl groups from specific glutamates on the MCPs. The action of CheB is countered by the methyltransferase CheR, which is constitutively active. MCPs with higher methylation have a greater ability to activate CheA, which in turn increases the activity of CheB and reduces the methylation state of the MCPs [13]. Thus, CheB and CheR work in concert to 2
generate adaptation, allowing cells to respond to relative changes in chemical concentrations across a range of background concentrations [14]. In E. coli, there are four structurally related MCPs forming mixed trimers of dimers. CheR is anchored by a long tether to the terminal domain of the two high-abundance receptor types (Tsr and Tar), allowing methylation and, thus, adaptation of other MCPs (e.g., Trg) in proximity. The elegant simplicity of the E. coli pathway makes it an attractive paradigm for chemotaxis. However, analysis of the genomes of motile bacterial species has revealed that over 50% contain multiple copies of each of the core chemotaxis genes (cheA, cheB, cheR, cheW, and cheY) [15]. The diversity of chemotaxis systems is also demonstrated by the proteins themselves; many species contain chemotaxis proteins that are not homologues of any of the E. coli proteins, such as the scaffold protein CheV, the deamidase CheD, and phosphatases CheC and CheX [16]. Many species also contain multiple homologues of MCPs [17], often including both transmembrane and cytoplasmic receptors [16]. Recently, it was shown that cytoplasmic receptors from Vibrio cholerae and Rhodobacter sphaeroides formed trimers of dimers packed hexagonally into large cytoplasmic arrays with the same spacing as transmembrane clusters [18]. This suggests that the packing of the chemoreceptors and the signalling proteins into large arrays is critical for correct signal transduction. Indeed, packing of the MCPs into these large arrays been shown to be important in creating gain in signal transduction through cooperative interactions between MCPs [19]. The methylation state of MCPs has been shown to alter their packing within the array, suggesting that adaptation alters the activity of CheA through changes in receptor packing [20]. The rate of adaptation in response to different chemical stimuli is the same, because of the packing of different receptors into the array and allosteric interactions between these receptors [21], further demonstrating the importance of these large arrays for proper chemotactic function. The position of flagella varies among species. Many have several randomly positioned flagella (peritrichous), as in E. coli, while others, such as P. aeruginosa, have a single polar flagellum. To move usefully through the environment, cells need to test the environment every second or so. Data suggest that the time taken for the 15-kDa CheYP to diffuse the length of an average cell is only approximately 100 ms [22] and, therefore, the positioning of the chemosensory array close to the flagellum is not essential for efficient chemosensory signalling. Electron cryotomography shows that, in most species, including E. coli, the membrane arrays are at, or close to, the cell poles, although in some polarly flagellate species, the arrays are found close to the flagellum. Data suggest that being motile and chemotactic is a survival advantage for many bacteria and, therefore, daughter cells need to not only be flagellate, but also inherit a chemosensory array on division to respond to the local environment. Therefore, it is critical that mechanisms exist to ensure that large arrays are inherited on division. If the arrays are localised to only one cell before division, then the daughter cell will
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
(A)
(B) (ii)
(i)
IM
MCPs
CheA/W baseplate
(C) (i)
(ii)
(D) (i)
(ii)
TRENDS in Microbiology
Figure 2. Chemosensory arrays. (A) Schematic of the polar transmembrane chemosensory array of Escherichia coli. (B) (i) Typical transmembrane chemosensory array viewed from the side. Sensory domains of methyl-accepting chemotaxis proteins (MCPs) in the periplasm, which is connected to a coiled-coil domain in the cytoplasm via a transmembrane region spanning the inner membrane (IM). MCPs form trimers of dimers (dark blue), which interact with CheA (red) and CheW (light blue) molecules at their distal tip. (ii) Top-down view of a typical array at the baseplate showing the hexagonal packing of MCP trimers of dimers (dark blue), CheA (red), and CheW (light blue). (C) (i) Electron cryotomography of Rhodobacter sphaeroides cytoplasmic chemosensory cluster (black arrowheads), showing a doublelayer architecture with CheA/W baseplates on the outside. (ii) Schematic of the array shown in (Ci). Colours as in ()B. (D) (i) Electron cryotomography of both cytoplasmic (CA) and transmembrane (MA) arrays in Vibrio cholerae. The transmembrane array shows typical architecture, with the CheA/W base plate distal to the IM. The cytoplasmic array forms a double-layer array with CheA/W baseplates on the outside. Abbreviation: OM, outer membrane. (ii) Schematic of the array shown in (Di). Adapted from [18] (Ci, Di).
not inherit a chemosensory network and, thus, will be nonchemotactic and unable to compete until a new array is expressed and assembled. If clusters are randomly distributed, then chance dictates that a proportion of daughter cells will also not inherit a chemosensory network; therefore, both the numbers and positioning of chemosensory clusters must be controlled. The focus of this review is how bacteria do this and how this relates to the cell cycle. Positioning patterns In many bacterial species, both electron cryotomography and fluorescence microscopy have shown that these large chemosensory arrays preferentially localise to the polar regions of the cell [23,24]. It has been proposed that this may be due to the array structure being more stable in curved membranes [25,26] or preference for the lipid com-
position of the poles [27]. The chemosensory proteins in E. coli tend to be polar and the size of the array appears to increase as the cells grow; therefore, an old pole will have a large cluster and, as the cell grows, additional clusters form, including one at the new pole, such that when the cell divides, each daughter cell has a cluster at its old pole (Figure 3A) [28]. Indeed, the lineage and age of cells can be followed by tracking the size of the polar cluster. In addition to these large polar clusters, E. coli also has smaller lateral clusters, probably formed as new receptors are expressed and inserted into the membrane [29]. By contrast, V. cholerae exhibits a strictly unipolar or bipolar distribution of clusters. The transition from unipolar to bipolar occurs before cell division (Figure 3C) [30]. Rhodobacter sphaeroides localises its transmembrane and soluble receptors to spatially distinct positions within the cell [31]. The transmembrane receptors are positioned in a 3
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
(B) (A)
(C)
(D)
(E)
(F)
1 µm TRENDS in Microbiology
Figure 3. Fluorescent images of bacterial chemosensory arrays. (A) Escherichia coli expressing YFP-CheR. Polar and lateral clusters can be seen. Scale bar = 1 mm. (B) Localisation of methyl-accepting chemotaxis proteins (MCPs) in Pseudomonas aeruginosa determined by immune fluorescence using anti-Tsr antibodies. Fluorescence is in green and the cell body in blue. (C) Vibrio cholerae cells expressing YFP-CheW1. The white line indicates a unipolar positioning pattern, and the red line a bipolar pattern. (D) Myxococcus xanthus cell expressing FrzCD-GFP. This soluble chemoreceptor forms clusters throughout the cytoplasm, excluding the polar regions. (E) Localisation of the transmembrane chemosensory cluster of Rhodobacter sphaeroides (YFP-CheW3, in green) and the Z-ring (FtsZ-CFP, in red). (F) The cytoplasmic chemosensory cluster of R. sphaeroides (TlpT-YFP). Adapted from [25] (A), [36] (B), [30] (C), [42] (D), and [32] (E).
pattern similar to that of E. coli, with predominantly polar clusters and some smaller lateral clusters (Figure 3E) [32]. The soluble receptors form a cluster in the cytoplasm that is positioned approximately mid-cell; as the cell cycle progresses, an additional cluster is formed and the two clusters are positioned at ¼ and 3/ 4 positions, such that each daughter cell inherits one cluster at the mid-cell on division (Figure 3F) [33]. The P. aeruginosa genome contains four sets of chemotaxis-like genes (che, che2, wsp, and pil-chp) and 26 chemoreceptor genes [34]. Gene products from the che locus localise to the polar regions of the cell, with some lateral clusters also present (Figure 3B); in dividing cells, clusters are often found at the old poles of the cells [35]. Che2 proteins also form clusters in the polar regions of the cell; however, co-expression of fluorescent fusions to Che and Che2 proteins (CheA and CheY2, respectively) showed no co-localisation of proteins from these two pathways [35]. Foci were either at opposite poles or adjacent to each other at the same pole; the deletion of the cheW gene from either the che or che2 locus resulted in proteins from that locus being unable to form clusters, but had no effect on proteins from the other locus. McpS is a soluble chemoreceptor; however, unlike the soluble receptors of R. sphaeroides, it does not localise to a site distinct from the membrane chemoreceptors, but localises to the poles, and is presumed to interact with transmembrane receptors and chemosensory proteins of one of the two systems that localise there (Che and Che2) [36]. The Wsp system is a chemosensory-like pathway that initiates biofilm formation in 4
response to surfaces [37]. WspA, the transmembrane receptor, forms clusters that localise throughout the membrane with no specificity for either polar or lateral membranes [38]. The gliding bacterium M. xanthus has eight sets of genes that encode putative chemosensory networks [39,40]. The proteins of these networks adopt a variety of positioning patterns. The soluble receptor FrzCD, which is responsible for controlling the frequency of reversals, forms clusters in the cytoplasm (Figure 3D) [41,42]. A receptor from the che7 operon, Mcp7, and an orphan receptor, McpJ, both localise to polar regions of the cell, even though Mcp7 is soluble and McpJ is transmembrane. Receptors from the other operons (all probably transmembrane) localise to clusters in the membrane along the length of the cell, with receptors from operons che4, che5, and che6 as well as three orphan receptors (McpH, McpM, and McpA) all co-localising [42]. These M. xanthus receptors clearly demonstrate the variety of positioning patterns of both soluble and transmembrane chemosensory clusters. They also demonstrate that arrays can comprise proteins from multiple operons as well as individual operons. In the species with multiple chemosensory pathways, localising them to different parts of the cell may represent an additional layer of regulation by preventing crosstalk. Unlike the above examples, where all the species undergo symmetric cell division, Caulobacter crescentus divides asymmetrically, with one cell remaining stalked and attached to a surface and the other being flagellated and chemotactic [43]. Accordingly, the chemosensory array only localises to the flagellate pole of the cell, resulting in
TIMI-1189; No. of Pages 10
Review only the swimmer cell inheriting a chemosensory network after division [44]. Across these species, it is clear that many transmembrane arrays preferentially localise to the poles of the cell, although often not exclusively. Soluble arrays have been found to localise both throughout the cell body and at the poles. It is currently unclear whether these positioning patterns are functionally important or simply a consequence of the systems used to segregate them on cell division. Mechanisms generating positioning patterns To generate and maintain these positioning patterns, bacteria must have mechanisms (either active or passive) to do so. Given the variety of positioning patterns, it is not unsurprising that several distinct mechanisms exist to generate these patterns. Stochastic self-assembly Escherichia coli has chemosensory clusters at the poles as well as small clusters periodically positioned along the cell length in the lateral membranes. Recent research has suggested that these clusters are formed and positioned by a process known as stochastic self-assembly [45,46]. Receptors are inserted individually into the membrane throughout the cell [47] and, once in the membrane, have two possible fates: either join an existing cluster or nucleate a new one. If the receptor is inserted close to an existing cluster, then there is a high probability that it will diffuse, encounter, and be absorbed into that cluster before it is able to nucleate a new one. Conversely, if it inserted far from the existing cluster, then probability favours nucleation, because the time required for a nucleation event to occur will be less than the time needed to encounter the existing cluster. Thus, the number of clusters in a given cell is dependent on the size of the cell, the concentration of receptors, and the diffusion coefficient of the receptors. Photoactivated localisation microscopy was used to position all molecules of the MCP Tar in individual E. coli cells. This revealed that, as well as the large clusters often found at the poles and smaller lateral clusters previously described, there are many smaller clusters and single proteins in the membrane [46]. The distribution of cluster sizes is continuous, suggesting that large clusters are formed by the gradual absorption of smaller clusters and single proteins rather than a sudden formation event from a pool of free nonoligomeric receptors. This study also revealed that the size of clusters increased with increasing distance from the large clusters at the poles, further supporting a stochastic self-assembly model in which the formation of large clusters is prevented in regions near an existing cluster. Simulations of stochastic self-assembly systems have shown that they are capable of producing positioning patterns that closely resemble those seen for chemosensory clusters in E. coli [48]. It has also been suggested that, once clusters have been formed, they are held in position by anchoring to unidentified structures at future division sites to maintain the spatial separation. Having a cluster at the division site ensures that, after division, the new poles of the daughter cells inherit a small cluster (Figure 4A) [45]. This transition from lateral to
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
polar clusters through cell division, along with the potential stabilising effect of the curved membrane [26,49], explains the positioning of the oldest and, therefore, largest clusters at the poles. This provides an elegant link between the cell cycle and cluster numbers and positioning; as the cell nears division, it will have approximately doubled in size, allowing for an increase in cluster numbers, which are formed spatially distant from each other, thus ensuring that both daughter cells inherit clusters on division. The positioning of the transmembrane chemosensory cluster of R. sphaeroides at first glance seems similar to that of E. coli, with polar and lateral clusters; however, recent work shows that, unlike E. coli, the clusters appear excluded from the division sites characterised by the Z-ring (a ring of polymerising FtsZ, which forms the septation site) (Figure 4B) [32]. The lateral clusters of E. coli show limited diffusion [25], but those of R. sphaeroides diffuse; indeed, in sphaeroplasts, the clusters show no organised localisation pattern and are present in random patches all over the cell. This suggests that polar clusters are formed by the movement of lateral clusters that are trapped in the curved membrane of the poles. While clusters merge with other diffusing clusters suggesting that stochastic selfassembly has a role in R. sphaeroides cluster formation and positioning, clusters also move away from arrays. The use of stochastic self-assembly in conjunction with other mechanisms represents a way in which to diversify the positioning patterns generated by stochastic self-assembly, while still linking array numbers and cell size. Cellular landmarks Only nonmotile stalked C. crescentus cells are able to initiate DNA replication and cell division. Cell division results in one stalked cell plus one motile and chemotactic swarmer cell [50]. For C. crescentus to produce two distinct progeny, one motile and one nonmotile, it must use a strategy for positioning flagella and chemosensory pathways that is different from that of E. coli. As the cell cycle progresses, the stalked cell elongates, with asymmetry of the progeny imposed during this predivisional stage. The new pole is marked by the presence of the protein TipN (Figure 4D), a transmembrane protein with several coiled-coil domains, which recruits TipF, a transmembrane c-di-GMP phosphodiesterase required for flagellum assembly [51,52]. During this early predivisional stage, TipN and TipF mediate the placement of the flagellum, pili, and chemosensory array at the new (nonstalked) pole. In late predivisional cells, TipN moves from the new pole to the site of constriction, followed by TipF, resulting in both these proteins being present at the new poles of the progeny [52]. Localisation of the flagellum, pili, and the chemosensory array to the new pole of the dividing cell ensures that they are inherited only by the swarmer cell. However, both the swarmer and stalked cells inherit TipN and TipF at their new poles. To ensure that the stalked cell does not have either a flagellum or a chemosensory system until it enters the predivisional stage, the expression of their genes is controlled temporally [53] and the activity of TipF is controlled by cell cycledependent c-di-GMP levels. 5
TIMI-1189; No. of Pages 10
Review
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
(A)
(B)
E. coli
R. sphaeroides
Key:
Chemoreceptor arrays Soluble chemoreceptor array Z-ring
(C)
(D)
V. cholerae Key:
HubP ParC ParP
C. crescentus Key:
TipN
TRENDS in Microbiology
Figure 4. Systems used to create positioning patterns. (A) Newborn Escherichia coli cells have one or more chemosensory arrays, the largest of which localises to the old pole. As the cell grows, arrays are formed at the new pole, as well as in the lateral membranes. On cell division, the polar clusters are positioned at the old poles of the daughter cells. The positioning of the lateral arrays is such that, on division, they will be positioned at the new poles of the daughter cells. Here, the preferential localisation at the poles, as well as the spatial limit on new cluster formation, ensures that at the least each daughter cell will inherit an array at its old pole. (B) Rhodobacter sphaeroides has two distinct chemosensory arrays, one polar and transmembrane (green), the other soluble (blue). The transmembrane array forms a large array at the old pole of the cell, with lateral clusters and a new pole array forming as the cell elongates. As the cell divides, lateral clusters are excluded from the area surround the Z-ring (red). The soluble array is positioned at mid-cell; as the cell grows, the array is split and the two resultant arrays are positioned at ¼ and 3/4 positions, so that, when the cell divides, both daughter cell inherit a single cluster at mid-cell. (C) Newborn Vibrio cholerae cells have a single flagella and a chemosensory array at their old pole. The polarorganising protein HubP is present at both poles, but, as the cell grows, ParC (a ParA homologue) is recruited to the HubP at the new pole. ParC in turn recruits ParP, which is able to stabilise the formation of a second chemosensory array at this pole. As the cell divides, HubP also localises to the site of constriction, resulting in daughter cells with HubP at both new and old poles. HubP recruits multiple ParA homologues, including FlhG, which ensures that the daughter cells have polar flagella. (D) The new pole of Caulobacter crescentus is marked by the protein TipN. When a swimming C. crescentus cell attaches to a surface, it sheds its flagellum, degrades its chemosensory array, and forms a stalk. As the stalked cell initiates cell division, the chemosensory genes are expressed and their localisation is directed by TipN to the new pole; this is also true of the flagellum. Just before division, TipN moves to the mid-cell, ensuring that both daughter cells will have TipN present at their new poles.
6
TIMI-1189; No. of Pages 10
Review When a swarmer cell encounters a surface, it sheds its flagellum and forms a stalk at the previously flagellate pole (the old pole) [50]. During this swarmer–stalked transition, the protease ClpXP degrades the chemosensory proteins [54,55]. Thus, once the cell has differentiated fully into a stalked cell and started cell division, any newly expressed chemotaxis proteins will form a de novo array at the swarmer (new) pole. While the exact nature of the interaction between the chemotaxis array and TipN/TipF remains unclear, TipN/ TipF might act as an anchor about which the array forms using a diffusion capture mechanism. Although this anchor generates the asymmetry in the predivisional cell, additional levels of control, through repression of expression plus degradation by proteases, exist to ensure that chemosensory arrays are not present in stalked cells postdivision. ParA homologues Several recent studies have shown that positioning patterns of chemosensory arrays in some bacterial species are dependent on homologues of ParA proteins. ParA proteins were initially identified as part of a tripartite system used to segregate low copy-number plasmids [56]. These systems comprise: ParA, a ‘deviant’ Walker A ATPase [57]; a parS nucleotide sequence in the plasmid; and ParB, which binds parS and stimulates the ATPase activity of ParA [58]. ATP-bound ParA dimerises [59] and subsequently nonspecifically binds DNA, coating the surface of the chromosome [60]. ParB, bound to the plasmid via the parS site, stimulates the ATPase activity of ParA and, thus, releases it from the DNA; the progression to the next chromosomebound dimer moves ParB and the associated plasmid over the chromosome surface [61]. Several models have been put forward for actual mechanisms by which this process generates segregation of the plasmids, including filament pulling [62] and diffusion ratchet [61]. Although discussion of these models is beyond the scope of this review, the core mechanism that underpins the models relies on a ParA protein with an ATP/ADP switch controlling localisation behaviour and a partner protein (here ParB) that modulates this switch. ParA proteins have recently been shown to be involved in the positioning and segregation of many macromolecular assemblies within bacterial cells (reviewed in [63]). These include the origins of replication of chromosomes in many species as well as several protein complexes, including the Z-ring, carbon fixation complexes, cellulose biosynthesis machinery, and chemosensory arrays in Vibrio spp. and R. sphaeroides. Vibrio spp. Both V. cholerae and Vibrio parahaemolyticus chemosensory clusters localise close to the single polar flagellum, with a cell cycle-dependent shift from unipolar to bipolar positioning. Both the uni- and bipolar positioning is dependent on a ParA homologue, termed ParC, whose gene is encoded within the chemotaxis operon [30]. The use of fluorescent fusions revealed that ParC switches from a unipolar pattern to a bipolar one, and is then followed by the chemotaxis proteins CheW1 and CheY3, which form
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
the second cluster de novo. ATP-locked ParC mutants form polar foci, as well as some nonpolar foci, whereas a mutant defective in ATP binding is diffuse throughout the cytoplasm and unable to form foci. Both of these mutants resulted in aberrant positioning of the chemotaxis proteins, indicating that the cycle of ATP hydrolysis and ADP/ATP nucleotide exchange is required. The identification of a pole-organising factor, HubP, helps explain ParC localisation, with ParC then determining the positioning of chemotaxis array [64]. Yamaichi et al. showed that HubP was required for the correct localisation pattern of ParC; it is also required for the polar localisation of ParA of chromosome I (ParA1) and a ParA homologue involved in the regulation of flagellum assembly (FlhG). Fluorescent fusions to HubP localised to the poles as well as the septation site before division to mark the new poles after division. Mutants lacking hubP are unable to sequester the origin of replication of chromosome I at the poles, cannot correctly position the chemosensory cluster, and produce multiple mislocalised flagella, rather than a single one. In these DhubP cells, mislocalised CheY3 remains colocalised with ParC, indicating that HubP recruits ParC, which in turn interacts with the chemosensory array. HubP is a large protein comprising multiple domains, including a single transmembrane domain in the N-terminal region and a large C-terminal cytoplasmic region. Interestingly, truncations and deletions of domains revealed that different domains of HubP are required for the polar localisation of ParA1, FlhG, and ParC. Bacterial two-hydrid assays and co-expression of fluorescent fusions in E. coli showed that both ParA1 and FlhG directly interact with HubP; however, ParC does not, indicating an intermediary protein(s) not present in E. coli. This suggests that interactions that tether these ParA proteins to HubP are distinct and that each system evolved separately to use HubP as a polar anchor. A subsequent study identified a further protein involved in the polar localisation of the chemosensory clusters in both V. cholerae and V. parahaemolyticus [65]. Plasmid parA genes are encoded immediately upstream of their partner proteins, parB. Ringgaard et al. identified an unannotated gene immediately downstream of parC termed parP. ParP is required for the correct positioning of chemotaxis proteins (CheW, CheA, and an MCP in V. parahaemolyticus; CheY3 and CheW1 in V. cholerae). ParP contains a CheW-like domain and a proline-rich N-terminal region, and was shown to interact with both CheA and ParC in bacterial two hybrid assays. These assays further revealed that ParP specifically interacted with a localisation and inheritance domain (LID) within CheA. Bioinformatic analysis of cheA genes showed that LID domains were only present in cheAs that were associated with a parC and a parP [66]. ParP co-localises with both ParC and chemotaxis proteins following the same cell cycle-dependent unipolar/bipolar positioning pattern. Fluorescence recovery after photobleaching (FRAP) experiments indicated that ParP is able to stabilise CheA within the array, and the co-expression of a membrane-targeted ParP and CheA in E. coli demonstrated that ParP is sufficient to induce CheA clustering. 7
TIMI-1189; No. of Pages 10
Review Together, these studies suggest a system in which HubP acts as a polar marker and anchor for several ParA proteins, including ParC. As the cell cycle progresses, ParC is recruited to HubP at the new pole, bringing with it ParP and resulting in bipolar localisation of both of these proteins. By interacting with CheA, ParP appears to drive nucleation and stabilises a new second chemosensory cluster at the new pole (Figure 4C). Thus, this creates a bipolar pattern of positioning that ensures that each daughter cell inherits a chemosensory network on division. The mechanism by which HubP localises to the pole has yet to be determined and is of interest, given the number of proteins that rely on its positioning to determine their own. The relation between ParC and ParP is also of interest, specifically whether ParP can drive ParC ATPase activity or if another protein is required for this and ParP is just the cargo. Rhodobacter sphaeroides The soluble chemosensory cluster of R. sphaeroides also shows cell cycle-dependent changes in number and position. Newly born cells have a single cluster positioned approximately at the mid-cell; before cell division, a second cluster is formed and these two clusters are positioned at the ¼ and 3/ 4 positions (Figure 4B) [33]. Both the positioning of the clusters and the formation of a second cluster are dependent on a ParA ATPase homologue, PpfA, encoded in the same operon as that of the proteins that comprise the cluster. Although PpfA localises to the DNA of the chromosome surface, it also co-localises with the cluster [67]. This contrasts with ParC, which segregates first and then recruits a second chemosensory cluster. As with ParA proteins from plasmid segregation systems, PpfA binds the nucleoid in an ATP-dependent manner, and mutants unable to bind either ATP or DNA are unable to co-localise with the chemosensory cluster. Fluorescent fusions to a mutant unable to hydrolyse ATP and, thus, locked in an ATP-bound state form brighter foci co-localising with the cluster; however, they are unable to duplicate or segregate the cluster, demonstrating that the ATP hydrolysis cycle is required for segregation. This suggests that the R. sphaeroides chemosensory cluster uses the segregating chromosomes as the mechanism for ensuring that each daughter inherits a cytoplasmic array. Roberts et al. also identified the partner protein for PpfA, a soluble chemoreceptor, TlpT, which forms part of the soluble cluster [67]. TlpT, along with the parB and parP genes, is encoded immediately downstream of its partner parA gene ppfA. TlpT is required for cluster formation; fluorescent fusions of the other proteins in the cytoplasmic chemosensory cluster display diffuse fluorescence in a DtlpT background, while cephalexin-treated filamentous cells that normally show evenly spaced chemosensory clusters had one large, wandering cluster in a DppfA background [68]. TlpT lacking the N-terminal 120 amino acids is still able to form clusters, but loses the co-localisation with PpfA and the ability to form multiple clusters or segregate, indicating that this region is responsible for interaction with PpfA [67]. While this suggests that the N-terminal domain of TlpT acts as a ParB homologue to drive PpfA ATPase activity and segregation, this remains to be unexplored, 8
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
as does the mechanisms involved in splitting the chemosensory array and segregating one with the segregating nucleoid. Interestingly, both the segregation systems for V. cholerae and R. sphaeroides depend not only on a ParA protein, but also adapted chemotaxis proteins (TlpT and ParP) to function at the very least as binding partners and possibly performing the same function as ParB in activating the ATPase activity of ParA. Given that over 35% of sequenced genomes appear to encode a ParA homologue in one of their chemotaxis operons [15,30], it would be interesting to determine whether this adaptation of chemotaxis proteins to partner ParA proteins is common. These two examples also illustrate that the use of ParA proteins can generate different positioning patterns and do so by different mechanisms, ParC creating a landmark for the formation of a second de novo array and PpfA splitting a single array into two. Concluding remarks It is clear that, for effective chemotaxis, allowing sensing of changes across a range of background concentrations, the proteins of the chemosensory pathways need to assemble into hexagonal arrays of many hundreds to thousands of sensory and signal transducing proteins, and this arrangement is critical regardless of whether the signalling pathway is transmembrane or soluble. Given that an ability to sense and respond is essential for motile bacteria to compete after division, each daughter must inherit a chemosensory cluster on division. It takes tens of minutes to express the proteins to assemble a new cluster, and the chemosensory behaviour of populations is measurably impaired where a cluster is not inherited and half the population need to synthesise de novo. It is likely that the absolute number of proteins is not critical, but the ratio of proteins and their assembly into the interacting array is, otherwise all the proteins could freely diffuse and each cell would simply inherit a percentage on division, which is not the case. The variety of mechanisms that bacterial species have evolved to ensure that progeny inherit chemotaxis arrays and, thus, are chemotactic after division demonstrates its importance. Although many transmembrane arrays have an apparent preference for the polar regions of the cells, this may reflect the curvature and phospholipid properties of the membrane, trapping the developing array. Concentrating the developing array at the poles ensures each daughter will inherit on division, but clearly some species appear to have no preference for the poles, suggesting that preference for the poles is not an innate property of all MCPs. Where species have more than one chemosensory pathway, these also show variation in positioning, possibly reflecting the function of the different pathways. Pathways that colocalise might allow direct integration of responses from multiple signals, while pathways that are clearly separate will allow the isolation of pathways and prevent crosstalk. In some species, there is a clear link between the numbers and positioning of flagella, and of chemosensory arrays. Escherichia coli have multiple flagella randomly positioned and there is no correlation between flagella position and array position. This is also the case for R. sphaeroides, which randomly positions its single flagellum, and the
TIMI-1189; No. of Pages 10
Review chemosensory membrane arrays, while predominantly polar, diffuse as large arrays around the membrane. In both cases, the formation of randomly positioned arrays ensures inheritance. However, species that are monotrichous, such as V. cholerae and C. crescentus, use polar markers to position both the array and the flagellum. It is worth noting that, while the distance between the receptor array and the flagellar motor should have no effect of chemosensory signalling, given the bacterial size, there is evidence that in Vibrio spp. the loss of positioning of the array in ParP mutants does have some effect of chemotactic ability for reasons that are still unclear. This ParA-like system has also been adapted to position the large soluble chemosensory arrays in species such as R. sphaeroides. The evolution of the tripartite ParA-like ATPase system to position proteins and plasmids in bacterial cells ensuring inheritance on division reveals how critical spatial organisation is in bacterial cells: they are not membrane sacs of diffusing enzymes, but ordered and organised organisms. This is neatly illustrated by the chemosensory pathways: the proteins need to be in very large hexagonal arrays to allow the sensitivity required with changing gradients, but the time taken to synthesise an array on division would mean that a daughter cell would not compete. Each cell must inherit a large cluster on division and, while for some species a diffusion capture system ensures a distribution of arrays, ParA-like systems have evolved for other species and for soluble arrays. Increasing numbers of proteins are being identified that arrange in large complexes and arrays and it will be interesting to see whether ensure inheritance in these similar systems. Acknowledgements The Rhodobacter sphaeroides work performed in J.P.A.’s group is funded by the Biotechnology and Biological Sciences Research Council (BBSRC).
References 1 Jarrell, K.F. and McBride, M.J. (2008) The surprisingly diverse ways that prokaryotes move. Nat. Rev. Microbiol. 6, 466–476 2 Nakane, D. et al. (2013) Helical flow of surface protein required for bacterial gliding motility. Proc. Natl. Acad. Sci. U.S.A. 110, 11145–11150 3 Bulyha, I. et al. (2009) Regulation of the type IV pili molecular machine by dynamic localization of two motor proteins. Mol. Microbiol. 74, 691– 706 4 Mignot, T. et al. (2007) Evidence that focal adhesion complexes power bacterial gliding motility. Science 315, 853–856 5 Mattick, J.S. (2002) Type IV pili and twitching motility. Annu. Rev. Microbiol. 56, 289–314 6 Deforet, M. et al. (2014) Hyperswarming adaptations in a bacterium improve collective motility without enhancing single cell motility. Soft Matter 10, 2405–2413 7 Darnton, N.C. et al. (2007) On torque and tumbling in swimming Escherichia coli. J. Bacteriol. 189, 1756–1764 8 Berg, H.C. and Brown, D.A. (1972) Chemotaxis in Escherichia coli analysed by three-dimensional tracking. Nature 239, 500–504 9 Zhang, P. et al. (2007) Direct visualization of Escherichia coli chemotaxis receptor arrays using cryo-electron microscopy. Proc. Natl. Acad. Sci. U.S.A. 104, 3777–3781 10 Briegel, A. et al. (2012) Bacterial chemoreceptor arrays are hexagonally packed trimers of receptor dimers networked by rings of kinase and coupling proteins. Proc. Natl. Acad. Sci. U.S.A. 109, 3766–3771 11 Welch, M. et al. (1993) Phosphorylation-dependent binding of a signal molecule to the flagellar switch of bacteria. Proc. Natl. Acad. Sci. U.S.A. 90, 8787–8791 12 Sarkar, M.K. et al. (2010) Chemotaxis signaling protein CheY binds to the rotor protein FliN to control the direction of flagellar rotation in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 107, 9370–9375
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
13 Hazelbauer, G.L. et al. (2008) Bacterial chemoreceptors: highperformance signaling in networked arrays. Trends Biochem. Sci. 33, 9–19 14 Lazova, M.D. et al. (2011) Response rescaling in bacterial chemotaxis. Proc. Natl. Acad. Sci. U.S.A. 108, 13870–13875 15 Hamer, R. et al. (2010) Deciphering chemotaxis pathways using cross species comparisons. BMC Syst. Biol. 4, 3 16 Wuichet, K. and Zhulin, I.B. (2010) Origins and diversification of a complex signal transduction system in prokaryotes. Sci. Signal. 3, ra50 17 Alexander, R.P. and Zhulin, I.B. (2007) Evolutionary genomics reveals conserved structural determinants of signaling and adaptation in microbial chemoreceptors. Proc. Natl. Acad. Sci. U.S.A. 104, 2885– 2890 18 Briegel, A. et al. (2014) Structure of bacterial cytoplasmic chemoreceptor arrays and implications for chemotactic signaling. Elife 3, e02151 19 Sourjik, V. (2004) Receptor clustering and signal processing in E. coli chemotaxis. Trends Microbiol. 12, 569–576 20 Irieda, H. et al. (2006) Control of chemotactic signal gain via modulation of a pre-formed receptor array. J. Biol. Chem. 281, 23880–23886 21 Krembel, A.K. et al. (2015) Universal response-adaptation relation in bacterial chemotaxis. J. Bacteriol. 197, 307–313 22 Segall, J.E. et al. (1985) Chemotactic signaling in filamentous cells of Escherichia coli. J. Bacteriol. 161, 51–59 23 Briegel, A. et al. (2009) Universal architecture of bacterial chemoreceptor arrays. Proc. Natl. Acad. Sci. U.S.A. 106, 17181–17186 24 Sourjik, V. and Armitage, J.P. (2010) Spatial organization in bacterial chemotaxis. EMBO J. 29, 2724–2733 25 Thiem, S. et al. (2007) Positioning of chemosensory clusters in E. coli and its relation to cell division. EMBO J. 26, 1615–1623 26 Hall, B. et al. (2012) Mechanism of bacterial signal transduction revealed by molecular dynamics of Tsr dimers and trimers of dimers in lipid vesicles. PLoS Comput. Biol. 8, e1002685 27 Mileykovskaya, E. and Dowhan, W. (2000) Visualization of phospholipid domains in Escherichia coli by using the cardiolipinspecific fluorescent dye 10-N-nonyl acridine orange. J. Bacteriol. 182, 1172–1175 28 Ping, L. et al. (2008) Tsr-GFP accumulates linearly with time at cell poles, and can be used to differentiate ‘old’ versus ‘new’ poles, in Escherichia coli. Mol. Microbiol. 69, 1427–1438 29 Sourjik, V. and Berg, H.C. (2000) Localization of components of the chemotaxis machinery of Escherichia coli using fluorescent protein fusions. Mol. Microbiol. 37, 740–751 30 Ringgaard, S. et al. (2011) A family of ParA-like ATPases promotes cell pole maturation by facilitating polar localization of chemotaxis proteins. Genes Dev. 25, 1544–1555 31 Wadhams, G.H. et al. (2003) Targeting of two signal transduction pathways to different regions of the bacterial cell. Mol. Microbiol. 50, 763–770 32 Chiu, S-W et al. (2013) Positioning of chemosensory proteins and FtsZ through the Rhodobacter sphaeroides cell cycle. Mol. Microbiol. 90, 322–337 33 Thompson, S.R. et al. (2006) The positioning of cytoplasmic protein clusters in bacteria. Proc. Natl. Acad. Sci.U.S.A. 103, 8209 34 Stover, C.K. et al. (2000) Complete genome sequence of Pseudomonas aeruginosa PAO1, an opportunistic pathogen. Nature 406, 959–964 35 Gu¨vener, Z.T. et al. (2006) Two different Pseudomonas aeruginosa chemosensory signal transduction complexes localize to cell poles and form and remould in stationary phase. Mol. Microbiol. 61, 106–118 36 Bardy, S.L. and Maddock, J.R. (2005) Polar localization of a soluble methyl-accepting protein of Pseudomonas aeruginosa. J. Bacteriol. 187, 7840–7844 37 Hickman, J.W. et al. (2005) A chemosensory system that regulates biofilm formation through modulation of cyclic diguanylate levels. Proc. Natl. Acad. Sci. U.S.A. 102, 14422–14427 38 O’Connor, J.R. et al. (2012) Surface sensing and lateral subcellular localization of WspA, the receptor in a chemosensory-like system leading to c-di-GMP production. Mol. Microbiol. 86, 720–729 39 Shi, X. et al. (2008) Bioinformatics and experimental analysis of proteins of two-component systems in Myxococcus xanthus. J. Bacteriol. 190, 613–624 9
TIMI-1189; No. of Pages 10
Review 40 Goldman, B.S. et al. (2006) Evolution of sensory complexity recorded in a myxobacterial genome. Proc. Natl. Acad. Sci. U.S.A. 103, 15200– 15205 41 Mauriello, E.M.F. et al. (2009) Localization of a bacterial cytoplasmic receptor is dynamic and changes with cell-cell contacts. Proc. Natl. Acad. Sci. U.S.A. 106, 4852–4857 42 Moine, A. et al. (2014) Functional organization of a multimodular bacterial chemosensory apparatus. PLoS Genet. 10, e1004164 43 Gomes, S.L. and Shapiro, L. (1984) Differential expression and positioning of chemotaxis methylation proteins in Caulobacter. J. Mol. Biol. 178, 551–568 44 Alley, M.R. et al. (1992) Polar localization of a bacterial chemoreceptor. Genes Dev. 6, 825–836 45 Thiem, S. and Sourjik, V. (2008) Stochastic assembly of chemoreceptor clusters in Escherichia coli. Mol. Microbiol. 68, 1228–1236 46 Greenfield, D. et al. (2009) Self-organization of the Escherichia coli chemotaxis network imaged with super-resolution light microscopy. PLoS Biol. 7, e1000137 47 Shiomi, D. et al. (2006) Helical distribution of the bacterial chemoreceptor via colocalization with the Sec protein translocation machinery. Mol. Microbiol. 60, 894–906 48 Wang, H. et al. (2008) Self-organized periodicity of protein clusters in growing bacteria. Phys. Rev. Lett. 101, 218101 49 Kim, K.K. et al. (1999) Four-helical-bundle structure of the cytoplasmic domain of a serine chemotaxis receptor. Nature 400, 787–792 50 Jensen, R.B. et al. (2002) Dynamic localization of proteins and DNA during a bacterial cell cycle. Nat. Rev. Mol. Cell Biol. 3, 167–176 51 Lam, H. et al. (2006) A landmark protein essential for establishing and perpetuating the polarity of a bacterial cell. Cell 124, 1011–1023 52 Huitema, E. et al. (2006) Bacterial birth scar proteins mark future flagellum assembly site. Cell 124, 1025–1037 53 Laub, M.T. et al. (2000) Global analysis of the genetic network controlling a bacterial cell cycle. Science 290, 2144–2148 54 Potocka, I. et al. (2002) Degradation of a Caulobacter soluble cytoplasmic chemoreceptor is ClpX dependent. J. Bacteriol. 184, 6635–6641 55 Tsai, J.W. and Alley, M.R. (2001) Proteolysis of the Caulobacter McpA chemoreceptor is cell cycle regulated by a ClpX-dependent pathway. J. Bacteriol. 183, 5001–5007
10
Trends in Microbiology xxx xxxx, Vol. xxx, No. x
56 Austin, S. and Abeles, A. (1983) Partition of unit-copy miniplasmids to daughter cells: I. P1 and F miniplasmids contain discrete, interchangeable sequences sufficient to promote equipartition. J. Mol. Biol. 169, 353–372 57 Leipe, D.D. et al. (2002) Classification and evolution of P-loop GTPases and related ATPases. J. Mol. Biol. 317, 41–72 58 Bouet, J.Y. and Funnell, B.E. (1999) P1 ParA interacts with the P1 partition complex at parS and an ATP-ADP switch controls ParA activities. EMBO J. 18, 1415–1424 59 Leonard, T. et al. (2005) Bacterial chromosome segregation: structure and DNA binding of the Soj dimer: a conserved biological switch. EMBO J. 24, 270–282 60 Hester, C.M. and Lutkenhaus, J. (2007) Soj (ParA) DNA binding is mediated by conserved arginines and is essential for plasmid segregation. Proc. Natl. Acad. Sci. U.S.A. 104, 20326–32031 61 Vecchiarelli, A.G. et al. (2010) ATP control of dynamic P1 ParA-DNA interactions: a key role for the nucleoid in plasmid partition. Mol. Microbiol. 78, 78–91 62 Ringgaard, S. et al. (2009) Movement and equipositioning of plasmids by ParA filament disassembly. Proc. Natl. Acad. Sci. U.S.A. 106, 19369–19374 63 Lutkenhaus, J. (2012) The ParA/MinD family puts things in their place. Trends Microbiol. 20, 411–418 64 Yamaichi, Y. et al. (2012) A multidomain hub anchors the chromosome segregation and chemotactic machinery to the bacterial pole. Genes Dev. 26, 2348–2360 65 Ringgaard, S. et al. (2014) ParP prevents dissociation of CheA from chemotactic signaling arrays and tethers them to a polar anchor. Proc. Natl. Acad. Sci. U.S.A. 111, E255–E264 66 Davis, N.J. et al. (2013) De- and repolarization mechanism of flagellar morphogenesis during a bacterial cell cycle. Genes Dev. 27, 2049–2062 67 Roberts, M.J. et al. (2012) ParA-like protein uses nonspecific chromosomal DNA binding to partition protein complexes. Proc. Natl. Acad. Sci. U.S.A. 109, 6698–6703 68 Wadhams, G.H. et al. (2005) Requirements for chemotaxis protein localization in Rhodobacter sphaeroides. Mol. Microbiol. 58, 895–902