Document not found! Please try again

Signal transduction in bacterial chemotaxis - FEBS Press - Wiley

47 downloads 0 Views 394KB Size Report
electrochemical energy to rotate a long helical flagella filament that propels the ...... Conclusions. In the course of evolution, the bacterial chemotaxis system has.
Review articles

Signal transduction in bacterial chemotaxis Melinda D. Baker,1 Peter M. Wolanin,2 and Jeffry B. Stock1,2* Summary Motile bacteria respond to environmental cues to move to more favorable locations. The components of the chemotaxis signal transduction systems that mediate these responses are highly conserved among prokaryotes including both eubacterial and archael species. The best-studied system is that found in Escherichia coli. Attractant and repellant chemicals are sensed through their interactions with transmembrane chemoreceptor proteins that are localized in multimeric assemblies at one or both cell poles together with a histidine protein kinase, CheA, an SH3-like adaptor protein, CheW, and a phosphoprotein phosphatase, CheZ. These multimeric protein assemblies act to control the level of phosphorylation of a response regulator, CheY, which dictates flagellar motion. Bacterial chemotaxis is one of the mostunderstood signal transduction systems, and many biochemical and structural details of this system have been elucidated. This is an exciting field of study because the depth of knowledge now allows the detailed molecular mechanisms of transmembrane signaling and signal processing to be investigated. BioEssays 28:9–22, 2006. ß 2005 Wiley Periodicals, Inc. Introduction Microbiology began with the advent of light microscopy in the 17th century. Minute particles were observed that were clearly 1

Department of Chemistry, Princeton University. Department of Molecular Biology, Princeton University. Funding agency: Melinda D. Baker is funded by a Science Research Dissertation Fellowship provided by the UNCF-Merck Science Initiative and Merck Research Labs, and Jeffry B. Stock is supported by the National Institutes of Health, Institute of General Medicine: Grant number R01GM057773. *Correspondence to: Jeffry B. Stock, Princeton University, Department of Molecular Biology, Lewis Thomas Laboratory, Princeton, NJ 08544. E-mail: [email protected]. DOI 10.1002/bies.20343 Published online in Wiley InterScience (www.interscience.wiley.com). 2

Abbreviations: HPK, histidine protein kinase; ATP, adenosine triphosphate; ADP, adenosine diphosphate; HAMP, histidine kinases, adenylyl cyclases, methyl-binding proteins and phosphatases’ domain; MH1, methylated helix 1; MH2, methylated helix 2; HPt, histidine phosphotransfer; NMR, nuclear magnetic resonance; EM, electron microscopy; MCP, methyl-accepting chemotaxis protein domain; PAS, Drosophila protein period clock (PER), the Ah receptor nuclear translocator (ARNT) and the Drosophila single-minded (SIM) domain.

BioEssays 28:9–22, ß 2005 Wiley Periodicals, Inc.

living organisms because of their purposeful motions (for a review see Berg(1)). By the end of the 19th century, the motor responses of bacteria had been thoroughly characterized by numerous investigators including the great German physiologist, Wilhelm Pfeffer. This research established that bacteria move in response to changes in temperature (thermotaxis), light (phototaxis), salinity (osmotaxis) and oxygen (aerotaxis), and to specific metabolites and other signaling molecules (chemotaxis). It was not until the end of the 20th century, however, that the molecular mechanisms that underlie bacterial sensory-motor regulation had been established. In the 1960s, Julius Adler began to elucidate the mechanism of chemotaxis in Escherichia coli.(2) Adler and his colleagues established that E. coli chemotaxis responses to amino acids and sugars are mediated by receptors at the cell surface that relay information via an intracellular signal transduction network to effect appropriate changes in motor behavior. The components of the intracellular signal transduction machinery were defined through the isolation and mapping of hundreds of different che (chemotaxis) mutants. Molecular genetic approaches initiated by Silverman and Simon in the 1970s established a bridge from che genes to Che proteins.(3) By the 1980s, it was possible to reconstitute the entire E. coli chemotaxis signal transduction system in vitro from its purified component parts (Table 1).(4) Atomic resolution structures are now available for several receptor fragments and all six Che proteins of the E. coli chemotaxis system. Recent research has largely focused on the way that these components are organized in the bacterial cell and how signals are transmitted across the membrane. It had generally been assumed that each membrane receptor interacted with a small complement of Che proteins to produce its own signal. The motor output was thought to represent a summation of the inputs from several thousand independent receptor–signaling units scattered over the surface of the cell. This view was initially brought into question in the 1990s through results obtained by Maddock, Shapiro and colleagues, which established that E. coli receptors and Che proteins tend to be localized together in tight clusters at one or both polls.(5) Each of these polar assemblies contains thousands of receptors together with their associated Che proteins. This review focuses first on how this information-processing organelle is assembled from its component parts. We then go on to discuss how these higher order complexes may serve a decision

BioEssays 28.1

9

Review articles

Table 1. E. coli chemotaxis signal transduction proteins Protein

Conserved protein domain

tsr

Tsr

MCP, HAMP

tar

Tar

MCP, HAMP

trg tap aer

Trg Tap Aer Tsr þ Tar CheA CheW CheY CheZ CheR CheB

MCP, HAMP MCP, HAMP MCP, HAMP, PAS

Gene

cheA cheW cheY cheZ cheR cheB

HPK, HPt, CheW-like CheW-like Response regulator Methyltransferase Methylesterase, response regulator

making function that coordinates numerous sensory inputs to produce a single control. Chemotaxis behavior It has become customary to regard living systems as machines. The E. coli flagellar motor fits well with this type of analogy. It is a nanoscale device that operates at close to 100% efficiency. Embedded in the bacterial cell, each motor uses electrochemical energy to rotate a long helical flagella filament that propels the bacterium.(6,7) A typical cell has a complement of half a dozen or more flagella anchored to independently rotating motors randomly distributed over the surface of the cell. Each motor alternates between clockwise or counterclockwise rotation with switching frequencies that exhibit the stochastic features of a two-state thermal equilibrium. Hydrodynamic drag causes counterclockwise rotating flagella to come together to form a bundle that acts cooperatively to push the cell body at speeds of approximately 20 microns per second. This behavior is termed smooth swimming or running. If one or more motors switch to rotate clockwise, the flagella become uncoordinated and the bacterium tumbles in place. In a uniform environment, cells move in a random walk: running for about a second, then tumbling for about a tenth of a second, then running in a random new direction. If a cell detects increasing concentrations of attractants or decreasing concentrations of repellents, its tendency to tumble is reduced, biasing its overall motion towards attractants and away from repellents. In a sense, chemotaxis towards attractants and away from repellents is determined by the cumulative effects of the second-to-second decisions of each individual to continue swimming or to tumble and change direction (for general reviews of bacterial chemotaxis see Berg,(8) Bourret,(9) and Wadhams(10)). The large polar assemblies of receptors and Che proteins function to control the probability that a cell will tumble and

10

BioEssays 28.1

Null phenotype

Protein data bank structure

Blind to L-serine and related amino acids Blind to L-aspartate, related amino acids, and maltose Blind to ribose and galactose Blind to dipeptides Deficient in aerotaxis Smooth swimming Smooth swimming Smooth swimming Smooth swimming Tumbly Smooth swimming Tumbly

1QU7 1VLS, 1VLT

1I5N, 1FWP, 1B3Q 1K0S 2CHY 1KMI 1AF7 1CHD

change direction. There is no simple relationship between the enormous quantities of sensory information received by these structures and the signals that they generate to control motility. Research has focused on properties of bacterial behavioral responses that are widely shared by other organisms. (1) Bacteria adapt. To a first approximation, they respond only to changes in the concentration of an attractant or repellent chemical rather than to absolute levels. After a short period (seconds to minutes) of continuous exposure, they behave as if no stimulus were present. (2) Bacteria have memory and can learn. The response of each bacterium to a given stimulus is entirely dependent on the history of that particular cell. (3) Bacteria are individuals, each with a unique character.(11) Some are more generally tumbly than others. Some are more smooth swimming. Responses to stimuli vary enormously from cell to cell. The molecular mechanisms that underlie these behaviors are now beginning to be understood in some detail. Overview of E. coli chemotaxis signaling Virtually all, motile prokaryotes use a two-component signal transduction system with conserved components to regulate motor activity.(12,13) In general, a two-component system includes a histidine protein kinase (HPK) that catalyzes the transfer of phosphoryl groups from ATP to one of its own histidine residues and a response regulator that catalyzes transfer of phosphoryl groups from the HPK-histidine to an aspartate residue on the response regulator.(14) In the chemotaxis system, the histidine kinase, CheA, associates with a distinct class of transmembrane receptor proteins, termed chemoreceptors, which interact with chemicals in the surrounding environment. Together chemoreceptors, CheA,

Review articles

Figure 1. The Escherichia coli chemotaxis system. Environmental cues are sensed by transmembrane chemoreceptors that are organized in a higher-order array at one or both of the cell poles. Receptor–signaling complexes consisting of chemoreceptors, CheW and CheA, control the transfer of phosphoryl groups from ATP to CheY by regulating the autophosphorylation of CheA. Phosphorylated CheY interacts with flagellar motors anchored in the cell wall. The phosphorelay is terminated by the dephosphorylation of CheY by CheZ releasing inorganic phosphate. CheR and CheB covalently modify chemoreceptors. Glutamyl modification, methylation or amidation, are depicted by a yellow star on the cytoplasmic domain of chemoreceptors. Abbreviations: W, CheW; A, CheA; P, phosphoryl group; Z, CheZ; Pi, inorganic phosphate; R, CheR; B, CheB; SAM, S-adenosylmethionine; CH3OH, methanol.

and a third protein, CheW, form large receptor–signaling complexes that integrate sensory information to control CheA kinase activity (Fig. 1). By regulating CheA autophosphorylation, receptor–signaling complexes control the phosphorylation of the chemotaxis response regulator, CheY. CheY reversibly binds CheA, dissociates from CheA upon phosphorylation, and rapidly diffuses to flagellar motors.(15–19) At the motor, phospho-CheY acts as an allosteric regulator to promote clockwise rotation and tumbling.(20) The primary output of the E. coli chemosensory apparatus is phospho-CheY. Chemotaxis results from the modulation of the concentration of phospho-CheY present in the bacterial cells that are swimming in gradients of attractant and repellant chemicals. Attractant stimuli suppress tumbles by interacting with chemoreceptors to inhibit CheA kinase activity and thereby decrease phospho-CheY.(14) The concentration of phosphoCheY is also affected by three soluble enzymes that are peripheral components of the sensory system: CheZ, CheR and CheB. CheZ is a protein phosphatase that associates with the receptor–signaling complex where it acts to enhance the rate of phospho-CheY dephosphorylation.(21–23) CheR and CheB are enzymes that methylate and demethylate the cytoplasmic portion of each chemoreceptor.(24,25) CheR is an S-adenosylmethionine-dependent methyltransferase that methylates specific glutamate side chains, converting these carboxylate anions into uncharged methyl esters. CheB

is an esterase that hydrolyzes the methyl esters formed by CheR to restore negatively charged glutamates. CheB also deamidates specific glutamine groups to produce glutamates that are then subject to esterification by CheR. The activities of CheR and CheB are regulated by the activity of the receptor– signaling complexes to generate changes in the chemoreceptor methylation and amidation levels that play a critical role in adaptation.(26–28) The CheR and CheB modifications also provide a memory mechanism that alters behavioral responses to subsequent stimuli. The stochastic nature of these modifying activities also ensures that no two cells will have precisely the same complement of receptor sensitivities. The six essential Che proteins, CheA, CheW, CheY, CheZ, CheR and CheB together with five chemoreceptors, Tsr, Tar, Tap, Trg and Aer, collectively constitute the E. coli chemotaxis system (Table 1). Differences in the number of copies of these genes and fusions and deletions of Che proteins represent numerous variations on this theme in different bacterial and archael species.(10,29) Nevertheless, in virtually all motile prokaryotes, receptor–signaling complexes composed of homologous chemoreceptor proteins, CheWs and CheAs act together with homologous CheYs to control sensory-motor activities. In order to understand the function and regulation of these conserved receptor–signaling systems, first we provide a discussion of the individual structures and functions of the chemoreceptors, CheA, CheW and CheY found in E. coli.

BioEssays 28.1

11

Review articles

Structures of conserved chemotaxis signaling components—chemoreceptors CheA, CheW and CheY

Chemoreceptors E. coli sense attractant and repellant stimuli via five chemoreceptor proteins: Tsr, Tar, Tap, Trg and Aer.(30,31) These transmembrane proteins are composed of highly variable periplasmic sensing domains that interact with stimulatory ligands, and a conserved cytoplasmic domain that provides a scaffold for CheW and CheA binding (Fig. 2).(32,33)

Figure 2. Bacterial chemotaxis chemoreceptor. A schematic of the domain architecture of transmembrane chemoreceptors (left). The sensing domain forms an up-down-up-down four helix bundle. The cytoplasmic domain is divided into three subdomains: the HAMP domain, the methylated helices, and the signaling domain. Molecular models of the dimeric sensing domain of Tar and a dimer of the truncated cytoplasmic domain of Tsr (residues 290–514) are shown to the right. Models were generated using coordinates taken from Yeh et al. and Kim et al. using Swiss PDB Viewer.(35,41,106)

12

BioEssays 28.1

The sensing domain of the aspartate receptor, Tar, has been expressed from the corresponding fragment of the tar gene and purified as a soluble homodimer.(34,35) The X-ray crystal structure indicates that each monomer is composed of an up-down-up-down four-helix bundle. In the cell, the first and last helices of this four-helix bundle extend across the membrane into the cytoplasm. The C-terminal end of the last helix (the second transmembrane helix) is linked to the signaling domain in the cytoplasm. The Tar sensing domain homodimer has two symmetrical, non-overlapping aspartate binding sites at the dimer interface.(36,37) Binding of aspartate to either symmetric site causes a conformational change that precludes binding at the second site. Further ligand-induced conformational changes such as a downward piston-like movement of the second transmembrane helix with respect to the first transmembrane helix and a rotation of dimer subunits with respect to one another may be the source of signaling across the cell membrane for inhibition of CheA kinase activity in the cytoplasm.(38–40) Although the chemoreceptor sensory domains are variable and specialized for ligand binding, they are all linked to a conserved cytoplasmic domain that extends away from the membrane and then bends back on itself via a hairpin turn (Fig. 2).(41) The degree of sequence identity is at a maximum in the hairpin turn region and decreases away from the center giving variable sequences.(32,33) The cytoplasmic domain structure can be divided into four subdomains beginning at the N terminus: (1) the Histidine kinases, adenylyl cyclases, methyl-binding proteins and phosphatases’ domain (HAMP), (2) methylated helix 1 (MH1), (3) the highly conserved domain or signaling domain, and (4) methylated helix 2 (MH2). The HAMP domain sequences have very little primary sequence identity but biochemical investigations of Tar support sequence-based structure predictions that the HAMP domain consists of two amphipathic helices connected by a non-helical or globular structure.(42–44) The HAMP domain is followed by a long, antiparallel alpha-helical coiled coil with MH1 and MH2 juxtaposed above the conserved signaling domain.(45) Together the methylated helices (MH1 and MH2) contain four or more glutamyl residues that are substrates for CheR and CheB modification.(46,47) These residues are spaced in heptad repeats along one face of each helix.(48) CheA and CheW interact with chemoreceptors in the region of the highly conserved signaling domain.(45) The structure of a soluble fragment of Tsr, encompassing the methylated helices and the highly conserved domain, has been solved by X-ray crystallography (Fig. 2).(41) The solved structure confirms experiments and predictions that the long protein fiber is an antiparallel coiled coil that forms a four-helix bundle when dimerized. This crystallographic data, in addition to biochemical crosslinking data, show that multiple four-helix bundles come together into trimers of dimers contacting one another within the signaling domain.(41,49,50)

Review articles

CheA CheA is the largest and most-complex component of the chemotaxis system. It is divided into five structurally and functionally distinct domains: the histidine phosphotransfer domain (P1), the response regulator binding domain (P2), the dimerization domain (P3), the histidine protein kinase catalytic domain (P4), and the regulatory domain (P5) (Fig. 3). These five domains are numbered in order from the amino to the carboxy terminus. The P1 domain belongs to the histidine phosphotransfer (HPt) family of proteins that function as intermediates in the transfer of phosphoryl groups between ATP and the phosphoaccepting aspartate side chains of response regulators. Other HPt domains of known structure include Ypd1 from Saccaromyces cerevisiae and ArcB from E. coli.(51) All of these HPt proteins consist of an up-down-up-down four-helix bundle (Fig. 3). Despite their structural and functional similarities, the sequences of HPt proteins are poorly conserved and difficult to detect by sequence alignment. The phosphorylated histidine, however, is invariably located in a solvent exposed position on the second helix of the four-helix bundle, and conserved glutamate and lysine residues surround the active site. In CheA, phosphorylation occurs on the N3 nitrogen of the imidazole side chain of His48.(52,53) P1 can be expressed from

the corresponding fragment of the cheA gene and purified to yield a soluble monomeric protein.(51) The isolated P1 domain can be phosphorylated by the HPK catalytic core (i.e. the P3 and P4 domains), and the phosphorylated product, phosphoP1, retains its CheY-phosphotransfer activity. Even though the active site glutamate and lysine residues are essential for ATPdependent phosphorylation of P1, they are not required for phosphotransfer between P1 and CheY. The response regulator binding domain, P2, is flanked by two flexible linker sequences connecting it to P1 and P3.(54) Like P1, P2 can be produced as an independent monomeric protein. The structure of P2 shows four antiparallel betasheets and two oppositely oriented alpha-helices (Fig. 3).(55,56) When P2 is in complex with CheY, the CheY active site undergoes a conformational change that increases the accessibility of the phospho-acceptor aspartate, Asp57. More importantly, P2 binds CheY in close proximity to the phosphoP1 domain and increases its effective concentration.(16,57) Sequences that are homologous to P3 and P4 have been identified in over a thousand different signal transduction proteins.(58) These two domains constitute the histidine protein kinase (HPK) catalytic core. Expression of P3–P4 from the corresponding portion of the cheA gene produces a protein that phosphorylates P1 at rates comparable to those obtained

Figure 3. A molecular model of the histidine protein kinase—CheA. The histidine phosphotransfer domain (P1) and the response regulator CheY/CheB-binding domain (P2) are depicted as monomers connected to one another and the remainder of CheA via flexible linkers. The dimerization domain (P3), ATP-binding phosphotransfer domain (P4), and the receptor-binding domain (P5) are all depicted within a CheA dimer. Models were generated using coordinates taken from Mourey et al, McEvoy et al, and Bilwes et al using Swiss PDB Viewer.(51,55,61,106)

BioEssays 28.1

13

Review articles

by full length CheA.(59) The P3 domain is a long antiparallel coiled coil that forms a symmetric up-down-up-down four-helix bundle in the CheA dimer; hence, the P3 four-helix bundle has essentially the same fold as the chemoreceptor cytoplasmic domain dimer (Fig. 3).(60,61) In solution, CheA homodimers and monomers are in equilibrium (KD  0.2 mM).(52) Whereas the monomers are inactive, dimers exhibit a basal rate of ATPdependent histidine phosphorylation (kcat  0.2 s1). In most HPKs, a dimerization domain corresponding to P3 contains the site of histidine phosphorylation within a conserved sequence that has been termed the H-box.(58) The X-ray crystal stucture of the CheA-P3 domain is very similar to the nuclear magnetic resonance (NMR) solution structure of the phosphoaccepting dimerization domain of the archetypal HPK, EnvZ. Although some of the conserved H-box residues in EnvZ are retained in CheA, the CheA dimerization domain is not phosphorylated. The residue corresponding to the phospho-accepting histidine in the EnvZ H-box is a glycine in CheA. The only site of phosphorylation in CheA is the His48 side chain in P1. Nevertheless, the CheA dimerization domain appears to play almost as important a role in CheA histidine kinase activity as it does in HPKs like EnvZ. Although the catalytic ATP-binding P4 domain appears to be an independent unit that does not participate in dimeric interactions, the CheA-catalytic core must be dimeric to phosphorylate P1.(62,63) It seems likely that, in CheA, the dimeric P3 domain fuctions to bind P1 and position it for phosphorylation by ATP bound to P4. The P1 domain of a CheA subunit that has a defective kinase catalytic domain is readily phosphorylated in trans by a CheA subunit that has a defective P1 domain and an active kinase catalytic domain. Trans phosphorylation, which has also been shown for a number of HPKs, does not, however, explain the need for dimerization. P4 is attached directly to the dimerization domain. Similar direct linkages between dimerization and ATP-binding phosphotransfer domains are seen in most HPK catalytic core structures.(61,64) Aside from the modified H-box, the CheA HPK catalytic core seems rather typical. The P4 domain contains all the conserved motifs found in most HPKs including the N, D (G1), F and G (G2) boxes. The HPK catalytic domain is homologous to the ATPase domains of the GyrB, MutL and Hsp90 families.(65) In these proteins as well as in the HPKs, ATP binding is accompanied by a large conformational change that involves the movement of a relatively unstructured loop (the ATP binding lid) into a fixed position over the bound nucleotide.(64,65) This creates a new protein surface that serves as a binding site for different, specialized protein domains. Binding of the second domain triggers ATP hydrolysis, which leads to domain dissociation and the release of ADP. The ATP-dependent binding and release of the auxiliary domain drives the DNA remodeling (GyrB), mutational repair (MutL), or chaperon (Hsp90) activities of these proteins. Structural and kinetics studies of CheA indicate that histidine

14

BioEssays 28.1

kinase activity follows a similar mechanism of ATP binding followed by closure of the ATP binding lid, and binding of a second domain of protein structure. In the case of CheA, the auxiliary domain that binds to the ATP-bound catalytic domain is P1. Instead of triggering the transfer of a phosphoryl group from ATP to water, P1 binding triggers phosphotransfer to His48. This leads to the dissociation of phospho-P1 and the release of ADP. To complete the cycle, the released phosphoP1 domain passes its phosphoryl group to CheY. P5 is connected directly to P4 with no obvious linker. P5 is homologous over its entire length to CheW.(61) P5 is composed of tandem SH3-like subdomains. This portion of CheA mediates binding to the chemoreceptor signaling domains. It was originally hypothesized that the association of CheA with chemoreceptors was mediated by CheW, which was sandwiched between the two proteins. It has been shown, however, that CheA binds in vitro directly to Tar or Tsr in membranes in the complete absence of CheW.(66) Moreover, high levels of CheW binding displace bound CheA and vice versa. Under saturating conditions, each CheA dimer that binds to membranes takes the place of four CheW monomers. While CheW binds stoichiometrically to Tsr subunits in the membrane, CheA binding is substoichiometric. The maximum stoichiometry in the absence of CheW is about one CheA dimer per ten Tsr dimers. In the presence of substoichiometric levels of CheW, the maximum level of in vitro CheA binding increases to about one CheA dimer per six Tsr dimers. Unlike CheW, CheA does not appear to bind to the isolated monomeric receptor signaling domains.(45) These results are consistent with the notion that CheA-P5 monomers and CheW bind to overlapping sites but CheA with a lower affinity; therefore, effective associations between P5 and chemoreceptors may require binary interactions that are afforded by the dimeric nature of CheA. The P5 domains in CheA dimers are far apart, separated by a center-to-center distance of approximately 7 nm. This means that CheA binding depends on the clustering of numerous chemoreceptor dimers together in the membrane so as to provide a multiplicity of appropriately spaced P5 binding sites. The enhanced stoichiometry of CheA binding in the presence of subsaturating CheW can be understood in terms of a chaperone-like function for CheW in facilitating the transition of independent chemoreceptor dimers into a relatively well-organized multimeric array. At the same time, the relatively long-range bridging of CheA between distant chemoreceptor proteins seems to facilitate the organization of the large receptor–signaling complexes that are seen in vivo.(5)

CheW CheW is a soluble monomeric protein that is commonly referred to as an adaptor protein. The NMR structure of CheW reveals two tandem SH3 domains that exhibit a strand exchange fold that is essentially the same as that seen in the

Review articles

CheA P5 domain (Fig. 4).(67,68) CheW binds to the highly conserved chemoreceptor signaling domain (KD  10 mM; 1 mol/mol stoichiometry).(45) Despite the structural similarities between the SH3 domains of eukaryotic signal transduction proteins and CheW, there is no sequence similarity that would indicate an evolutionary relationship.(61,69,70) Moreover, there are no prototypical polyproline SH3-binding sequences in the chemoreceptor proteins or any other components of the chemotaxis system. A preference for binding to proline-rich PxP or PxxP sequences is not, however, an invariant feature of eukaryotic SH3 domains. Monomers of the Eps8 protein, a substrate for the epidermal growth factor tyrosine kinase, binds PxxDY sequences.(71,72) The SH3 domain of Eps8 dimerizes via beta-strand exchange, generating an SH3-like fold composed of secondary structural elements from each monomer. The tandem SH3 subdomains of CheW and P5 also exchange beta-strands creating a similar hydrophobic surface for protein–protein interactions.

Figure 4. The SH3-like adaptor protein—CheW. A: Topology diagram of CheW. B: Molecular model of CheW. Molecular models were generated using coordinates taken from Griswold et al using Swiss PDB Viewer.(68,106)

CheY CheY, the chemotaxis response regulator, is a soluble monomeric protein.(73) The CheY phospho-receiver domain is conserved across motile bacteria and appears in some species as hybrid proteins with other chemotaxis protein domains such as those found in CheA and CheW.(29) The catalytic domain of CheY is a doubly wound five-stranded parallel beta-sheet.(73) CheY localizes to receptor–signaling complexes via interactions with the CheA response regulator binding domain, P2.(74) Phospho-CheY has a substantially decreased affinity for CheA and diffuses away from CheA in order to interact with FliM at the flagellar motor.(15–19) Despite the essential roles P1 and P2 play in enhancing the rate and specificity of phosphotransfer from phospho-P1, CheY is far from a passive participant in the phosphotransfer catalytic process. Like most response regulators, CheY is autophosphorylated by small molecule phosphodonors such as phosphoramidate or acetylphosphate and actively transfers the phosphate from CheA to its conserved aspartate.(75,76) Receptor–signaling complexes Determination of the structures of the chemoreceptors and all the Che proteins has provided a foundation for understanding the assembly of receptor–signaling complexes. It was initially assumed that CheW monomers bound to receptor–signaling dimers, and then CheA dimers bound to the receptorassociated CheWs to form 2:2:2 complexes.(45,77,78) Each 2:2:2 complex was thought to work independently to modulate the rate of CheA autophosphorylation. The outputs from the thousands of these complexes in a single cell were thought to be summarily integrated through their effects on a common pool of phospho-CheY in the cytoplasm. More recent findings indicate that this model is incorrect. Now it is clear that thousands of chemoreceptor proteins in a single E. coli come together to form one or two large interconnected arrays at one or both poles of the cell.(5) CheW and CheA both closely associate with these interconnected receptor arrays at the membrane. In E. coli the five chemoreceptors, of varying sensory specificities, interact cooperatively to regulate CheA kinase activity and are expressed together with all of the che genes at roughly fixed ratios of one to another.(79–81) In vivo cross-linking studies indicate that these different receptors are juxtaposed within signaling complexes.(49,50) Before addressing the implications of this structural organization, we first need to consider the problem of its assembly from core components: the chemoreceptors, CheW and CheA. Immunoelectron microscopy (EM) studies of mutant E. coli deleted for individual chemotaxis proteins indicate that only chemoreceptor proteins, CheW and CheA, are required for the formation of polar receptor–signaling clusters.(5) In contrast, cells lacking the covalent modification enzymes, CheR and CheB, exhibit polar clusters similar to wild-type cells.(82) In general, localization and aggregation of high abundance

BioEssays 28.1

15

Review articles

receptors, Tsr and Tar, are insensitive to the signaling state of the receptor–signaling domain.(83) When expressed alone, low abundance receptors, Trg and Tap, are deficient in clustering if locked in a state in which CheA is fully inhibited, but polar localization is not altered.(84) This suggests that localization and cluster formation are independent. There have been numerous studies of complex formation in vitro. Tar or Tsr at high concentrations in membranes readily bind CheW and CheA to form functional receptor–signaling complexes.(66) CheW and CheA binding fit simple adsorption isotherms. There is no evidence for any cooperativity between CheW and CheA. Although relatively low concentrations of CheW significantly increase the maximum stoichiometry of CheA bound per chemoreceptor subunit, CheW does not affect the affinity for CheA. Attractants or changes in glutamyl modification, which have dramatic affects on kinase activity, do not substantially affect the affinity for, or stoichiometry of, CheW and CheA binding. Soluble fragments of Tar or Tsr consisting of the conserved signaling domain also bind CheA and CheW to form soluble receptor–signaling complexes.(85–87) Structural studies of these complexes indicate that each structure contains 12 receptor coiled-coil dimers, approximately 6 CheW monomers, and 2 CheA dimers.(87) The structures have a cylindrical

shape with CheW and CheA binding around the center and 6 four-helix bundles extending from each side of this central region (Fig. 5). Soluble complex formation is critically dependent on the nature of the cytoplasmic domain fragment, including the level of glutamyl methylation or amidation.(85) The repulsive effect of negatively charged glutamate residues, up to 8 per four-helix bundle, is probably too difficult to overcome to allow soluble complex assembly. Most Tar cytoplasmic domain fragments do not tend to form complexes unless a coiled-coil dimerization domain from a leucine zipper transcription factor is fused to their N termini.(45,88) Similar results are obtained by deletion of MH2, the anti-parallel carboxy-terminal partner of the methylated coiled-coil region.(45) Presumably this forces dimerization of the unpaired MH1 helix to produce an effect similar to that obtained with the leucine zipper fusion. Several cytoplasmic domain fragments of Tsr have been found to participate in soluble complex formation without requiring the fusion of a leucine zipper dimerization domain or cleaving off a carboxy-terminal portion of the anti-parallel coiled coil.(89) Despite differences in their propensities for complex assembly, complexes formed from different soluble Tar and Tsr constructs all appear to have a similar structural organization (unpublished observations Wolanin P, Francis N, Thomas D, Stock J and Derosier D).

Figure 5. A structural model of soluble receptor–signaling complexes.(87) A: A density map of the reconstruction of soluble lzTarc – CheW–CheA complexes from electron microscopy images aligned at the middle. B: Twenty-four receptor monomers can be accommodated in the electron density map using the atomic model of Tsrc four-helix bundles representing Tarc. Six chemoreceptor four-helix bundles on each side project from the center. C: A cross-sectional view through the center of the density map. D: A cross-sectional view through the center of the electron density model with the atomic models of Tsrc, CheA (cyan) and CheW (magenta). Abbreviation: lz, leucine zipper coiled-coil dimerization domain.

16

BioEssays 28.1

Review articles

The X-ray crystal structure of the Tsr cytoplasmic domain indicates a trimer of dimer receptor arrangement with tight interdimer contacts in the hairpin turn regions, and crosslinking studies with receptors in membranes are consistent with this type of structure.(41,49,50) It seems likely that different packing interactions pertain in the presence of CheA and CheW. The close contact between receptor dimers is not seen in EM images of receptor–signaling complexes in vitro.(87) In these complexes, the hairpin turn regions at the ends of the signaling domain four-helix bundles are splayed apart, to accommodate CheW and CheA binding to this region. Nevertheless, a trimeric unit of organization appears to prevail, with two trimeric units coming together to form 6 four-helix bundles that extend from the central hairpin turn region. The trimer of dimers seen in the crystal structure may represent a folding intermediate that comes apart to accommodate CheW and CheA binding. Complex assembly may first involve monomers coming together to form dimers, then, dimers associating to form trimers, and finally CheW binding to the trimeric structure to force the dimers apart in the hairpin turn region. Unlike CheW, CheA binding requires a binary interaction at two relatively distant hairpin turns,(90,91) and this requirement probably cannot be satisfied within the context of a single trimer of dimer structure. The cytoplasmic domain fragments with leucine zipper dimerization domains that have been used to generate soluble complexes are not dimeric, but are long, rod-like tetramers.(85,88) We have hypothesized that this tetrameric structure might be generated by an exchange of carboxyterminal anti-parallel helices between two dimers.(86) The only significant difference in structure that would be expected from a domain exchange of this type would be in the hairpin turn region. Instead of a tight turn folding back on itself, the polypeptide chain would need to extend a distance of a few nanometers to engage antiparallel helices of adjacent dimers. An intriguing possibility is that CheW and CheA binding to the hairpin turn region causes it to undergo a conformational change that promotes domain swapping. Thus, binding of CheW to signaling domain dimers might cause them to exchange helices, and thereby promote interactions with other dimers to form higher-order arrays. Once the arrays are sufficiently large, CheA binding can occur and assist in the organization into still larger associations. Structure–function relationships in signal transduction

Signaling across the membrane There have been many theories as to how stimuli from the extracellular environment are perceived and how this information is used to effect excitatory and adaptive responses. Hypotheses about the mechanism of transmembrane signaling in the E. coli chemotaxis system were initially derived from

detailed analyses of the conformational changes that occur when aspartate binds to Tar. It is apparent that aspartate binding causes a significant movement of one subunit with respect to the other within a receptor dimer.(38,39) In addition to these small inter-subunit displacements, there are substantial changes over the entire solvent-exposed surface of the dimer as well as changes in the orientation of the dimer with respect to the plane of the membrane.(40) All of these perturbations would be expected to promote disorder within the sensory array favoring expansion of the array of signaling domains and their associated CheAs and CheWs-on the other side of the membrane. Aspartate binds along one side at the juncture between dimers. Aspartate binding at either of the two equivalent sites obstructs the other site. The KD for aspartate binding to receptor domain dimers in solution is approximately 1 mM. Assuming the rate of binding is diffusion limited, the half-life of an individual aspartate-bound receptor is about a millisecond. Once an aspartate molecule enters a sensory domain array, it will tend to bind numerous times to numerous different receptors. EM images of receptors in membranes indicate tightly packed highly ordered structures.(92) Each aspartatebinding event would tend to disrupt such organized arrangements. Chemotaxis signal transduction can be approximated by a two-state formalism.(80,93,94) Thus the structure of the sensing domain array may be considered in terms of an equilibrium between two states—an ordered, tense or T state and a disordered, relaxed or R state (Fig. 6), with aspartate binding with slightly higher affinity to sensory domains in the R state. The structure of the receptor–signaling complex in the cytoplasm may also be considered in terms of an equilibrium between two states—a highly condensed T state where CheA is fully active, and a relatively diffuse R state where CheA is inactive. As a first approximation, one can assume a one-toone correspondence between the R ! T equilibrium of the input sensing array and the R ! T equilibrium of the signaling-complex output. Using such a formulation, one can obtain a good fit of receptor-mediated stimulus–response coupling.(80)

Mechanism of CheA regulation The histidine kinase activity of CheA in receptor–signaling complexes can be several hundred-fold higher than that of isolated CheA dimers.(85) This activation is completely dependent on CheW. Kinetic studies indicate that the effect of CheW is to stabilize a form of CheA wherein P1 and ATP are bound together at the CheA active site.(95) High concentrations of CheA and CheW have been shown to interact in solution, but in the absence of receptors, CheW has no affect on CheA activity.(66) For activation to occur, both proteins must be associated with chemoreceptors. Under these conditions, the activity per molecule of bound CheA varies linearly with the concentration of bound CheW. This linear relationship even

BioEssays 28.1

17

Review articles

Figure 6. A higher-order array of chemoreceptor-sensing domains may exist in equilibrium between an ordered, tense state (T) and a disordered, relaxed state (R). Top-down view from the periplasm. Aspartate, for example, would have a slightly higher affinity for the R state. A similar relationship between a T state and R state could also exist for the signaling domains in the cytoplasm. While receptor–signaling complexes would remain in the presence of ligand, direct contacts or the distances between CheA and CheW binding sites in the cytoplasm would be perturbed.

extends to concentrations of CheW that are sufficiently high to begin to displace CheA, in which case the total kinase activity decreases as the activity per mol increases. Signaling complexes with a subunit stoichiometry of 6 receptors: 4 CheWs: 1 CheA give close to optimal kinase stimulation. This molar protein ratio has not been seen in vivo; the ratio of total proteins is 3.4 receptors dimers: 1.6 CheW monomers: 1 CheA dimer.(81) However, in vivo stoichiometries of receptor–signaling complexes would be expected to be optimized for kinase regulation rather than maximal kinase activation. One hypothesis for the mechanism of chemotaxis signal transduction postulates that CheW and CheA reversibly bind to adjacent sites within the receptor–signaling array in an orientation relative to one another such that CheW and CheA interact to stimulate CheA autophosphorylation. Binding of CheW and CheA to receptor–signaling domains in other orientations would render the kinase essentially inactive. According to this model, regulation of CheA activation is determined by inputs such as methylation or amidation that influence the probability that CheW and CheA are properly positioned relative to one another within a receptor–signaling array. Expansion or reorganization of the signaling array, due to electrostatic repulsion between chemoreceptor four-helix bundles, for example, would preclude close encounters between receptor-associated CheW and CheA proteins and thereby inhibit kinase activity.

Role of methylation in adaptation The kinase activity of receptor–signaling complexes in E. coli is generally maintained such that the level of phospho-CheY is at or near 50%. This optimizes the sensitivity of the system to both positive and negative fluctuations in kinase activation. The feedback methylation system acts to preserve this

18

BioEssays 28.1

delicately balanced state. Attractant stimuli like aspartate and serine cause increases in CheR-dependent methylation and decreases in CheB-dependent demethylation and deamidation.(4,26,27) The attractant-induced increases in methylation promote CheA activity precisely enough to counteract the inhibitory effect of attractant-sensory domain binding. Methylation acts, at least in part, by reducing the net negative charge at the surface of signaling domain four-helix bundles to favor a more condensed signaling domain lattice. Repellent stimuli have opposing effects—causing increased CheA activity that is inactivated by decreases in methylation associated with increased CheB and decreased CheR activities. The CheR/ CheB-regulated belt of net negative charge that lies in the MH1-MH2 regions of the cytoplasmic four-helix bundles of receptor–signaling arrays acts to maintain the system at a constant steady-state rate of CheY phosphorylation. Increases in aspartate or serine initially suppress tumbling by binding to Tar or Tsr, inhibiting CheA kinase activity and causing a substantial decrease in the level of phospho-CheY. Increased levels of Tar or Tsr methylation causes the level of phospho-CheY to return back to its pre-stimulus level despite the continued presence of aspartate or serine. Tsr and Tar are by far the most abundant receptors in E. coli with the level of Tsr being approximately twice that of Tar.(81) Trg, a minor receptor, is expressed at 3% of the level of total receptor proteins; the other minor receptor receptors, Tap and Aer, are presumably expressed at similarly low levels. Tsr or Tar work fairly well on their own in cells deficient in all other receptors, but the minor receptors require the presence of Tsr or Tar.(96,97) Increasing levels of expression do not correct this dependence, which stems from a defect in the ability of the minor receptors to tune their sensitivity by modulating levels of glutamate methylation. Tsr and Tar have a conserved

Review articles

carboxy-terminus that serves as a binding site for CheR and CheB.(26) If this sequence, which is missing from the minor receptors, is appended to their carboxy-termini, their abilities to mediate chemotaxis responses in the absence of Tsr or Tar are restored.(98,99) Adaptation assistance provides strong evidence for cross talk between major and minor receptors.

Phosphatase regulation CheZ controls the rate of dephosphorylation of phospho-CheY by greatly enhancing the rate of CheY phosphatase activity. CheZ is a peripheral component of receptor–signaling complexes. Most of the CheZ in a cell binds to an N-terminally truncated, short form of CheA called CheA short (CheAS) that is expressed from an alternative translational initiation site at levels comparable to full-length CheA.(23,100,101) A large fraction of CheA dimers in E. coli are CheA/CheAS heterodimers. CheA preferentially binds CheY, phosphorylates it and releases phospho-CheY; in the same way, CheZ-bound to CheAS binds phospho-CheY, dephosphorylates it and releases CheY. Fluorescence microscopy of GFP-CheY indicates that a large fraction of the total CheY in a cell is associated with the polar receptor–signaling complexes.(74) This would mean that both CheZ and CheA are potentially working near Vmax at rates that may not be dependent on CheY or phospho-CheY concentrations. When this is the case, one expects to see wide swings in the balance between phosphoCheYand CheY because of a phenomenon termed zero-order ultrasensitivity.(102) Zero-order ultrasensitivities that stem from phosphorylation-induced changes in CheY binding to CheA and CheZ may function to amplify signals that exceed a certain threshold. While CheZ activity in E. coli has never been shown to be regulated by chemoreceptor signaling, changes in CheZ activity could regulate phospho-CheY just as effectively as changes in CheA kinase activity. The chemotaxis systems in many species lack CheZ and regulate phospho-CheY dephosphorylation via other mechanisms.(29) Conclusions In the course of evolution, the bacterial chemotaxis system has emerged as the mechanism for motor regulation in virtually all motile prokaryotes. Almost identical components control the behaviors of bacterial and archael species that are thought to have diverged billions of years ago. The signal transduction mechanism in each cell involves huge organelle-like assemblies composed of thousands of interacting protein subunits. In the past, membrane receptor signaling has generally been seen as simply a mechanism to link a particular stimulus to an associated response. Bacterial chemotaxis receptor–signaling complexes function more as ‘go–no go’ decision-making devices. In a sense, the polar chemotaxis receptor complexes are the microbial counterparts of a brain. The chemotaxis mechanism clearly has an enormous computational capacity considering it is only a few hundred nanometers in diameter.

The realization that individual receptor dimers do not work independently, but are fixed within large arrays, has necessitated a reassessment of the information-processing capacity of the system. In the past, arguments were advanced that the rather large number of chemoreceptors per cell was essential for the efficient capture of low abundance stimulatory ligands. This type of analysis makes sense if one imagines that the primary function of the chemosensory apparatus is stimulus detection. The existence of higher order chemoreceptor assemblies suggests that information processing and memory storage may be more central to the chemotaxis mechanism than excitation and adaptation per se. After all, the chemotaxis problem from a bacterial perspective primarily involves command and control issues rather than simply the acquisition and transmission of sensory information. Recognition of the almost brain-like computational capacity of the bacterial chemotaxis system leads one to question what it could be used to accomplish. In the past, most theoretical discussions of bacterial chemotaxis have advanced molecular mechanisms to explain responses to simple experimental stimulus–response paradigms. Cells have generally been treated experimentally as if they were simple chemical automatons. To this day, it is routine to grow E. coli on very complex media, dilute them into minimal salts, suddenly expose them to high concentrations of a single attractant or repellent stimulus, and measure the average time that it takes them to adapt. Whatever measure of bacterial behavior is provided by such procedures clearly has little direct relevance to the complex problems that are critical to bacterial chemotaxis and survival in the natural environment. As structural studies have begun to reveal the enormous potential complexity of the molecular interactions that may actually be involved, the need for more complex experimental paradigms is becoming evident. This is leading toward a much more thorough investigation of the roles of chemotaxis within the broader context of the physiology of bacterial adaptations to environmental stress. One consequence of this new emphasis on bacterial physiology has been a greater appreciation of the important roles played by signaling interactions within and between microbial populations. Recent studies indicate that the sensory modalities of the E. coli system may be attuned more for communication between individual cells than to simply measure the nutritive content of the surrounding media.(103,104) The major E. coli receptors detect amino acids such as aspartate and glutamate (Tar) and serine, alanine and glycine (Tsr) that are directly connected with intermediary metabolism and are frequently secreted as overflow metabolites. The most-reliable indicator of a healthy environment for E. coli growth and survival is most likely the presence of a substantial population of adequately fed E. coli. In light of this, it is not surprising that, in complex environments, E. coli and other bacteria tend to congregate and even form organized

BioEssays 28.1

19

Review articles

communities.(105) Just as the synergy between interacting receptors provides an enormous increase in informationprocessing capability for an individual cell, complex social interactions between individuals can generate a tremendous increase in the information-processing capabilities of a population of cells. Acknowledgments The authors thank Daniel Webre for critical review of the manuscript. References 1. Berg HC. 1975. Chemotaxis in bacteria. Annu Rev Biophys Bioeng 4:119–136. 2. Adler J. 1975. Chemotaxis in bacteria. Annu Rev Biochem 44:341– 356. 3. Silverman M, Simon M. 1977. Chemotaxis in Escherichia coli: methylation of che gene products. Proc Natl Acad Sci USA 74:3317–3321. 4. Ninfa EG, Stock A, Mowbray S, Stock J. 1991. Reconstitution of the bacterial chemotaxis signal transduction system from purified components. J Biol Chem 266:9764–9770. 5. Maddock JR, Shapiro L. 1993. Polar location of the chemoreceptor complex in the Escherichia coli cell. Science 259:1717–1723. 6. Macnab RM. 1996. Flagella and Motility. In: Neidhardt FC, Curtiss R III, Ingraham JL, Lin ECC, Low KB Jr, et al. editors. Escherichia coli and Salmonella: cellular and molecular biology. 2nd ed. Washington, DC: ASM Press. p 123–145. 7. Berg HC. 2003. The rotary motor of bacterial flagella. Annu Rev Biochem 72:19–54. 8. Berg HC. 2000. Motile behavior of bacteria. Physics Today 53:24–29. 9. Bourret RB, Stock AM. 2002. Molecular information processing: Lessons from bacterial chemotaxis. Journal of Biological Chemistry 277:9625–9628. 10. Wadhams GH, Armitage JP. 2004. Making sense of it all: Bacterial chemotaxis. Nature Reviews Molecular Cell Biology 5:1024–1037. 11. Koshland DE Jr. 1980. Bacterial chemotaxis in relation to neurobiology. Annu Rev Neurosci 3:43–75. 12. Stock AM, Robinson VL, Goudreau PN. 2000. Two-component signal transduction. Annu Rev Biochem 69:183–215. 13. Wolanin PM, Thomason PA, Stock JB. 2002. Histidine protein kinases: key signal transducers outside the animal kingdom. Genome Biol 3:REVIEWS3013. 14. Borkovich KA, Kaplan N, Hess JF, Simon MI. 1989. Transmembrane signal transduction in bacterial chemotaxis involves ligand-dependent activation of phosphate group transfer. Proc Natl Acad Sci USA 86:1208–1212. 15. Li J, Swanson RV, Simon MI, Weis RM. 1995. The response regulators CheB and CheY exhibit competitive binding to the kinase CheA. Biochemistry 34:14626–14636. 16. Stewart RC. 1997. Kinetic characterization of phosphotransfer between CheA and CheY in the bacterial chemotaxis signal transduction pathway. Biochemistry 36:2030–2040. 17. Scharf BE, Fahrner KA, Turner L, Berg HC. 1998. Control of direction of flagellar rotation in bacterial chemotaxis. Proc Natl Acad Sci USA 95:201–206. 18. Elowitz MB, Surette MG, Wolf PE, Stock JB, Leibler S. 1999. Protein mobility in the cytoplasm of Escherichia coli. J Bacteriol 181:197–203. 19. Sourjik V, Berg HC. 2002. Binding of the Escherichia coli response regulator CheY to its target measured in vivo by fluorescence resonance energy transfer. Proc Natl Acad Sci USA 99:12669– 12674. 20. Alon U, Camarena L, Surette MG, Aguera y Arcas B, Liu Y, et al. 1998. Response regulator output in bacterial chemotaxis. Embo J 17:4238– 4248. 21. Stock AM, Stock JB. 1987. Purification and characterization of the CheZ protein of bacterial chemotaxis. J Bacteriol 169:3301–3311.

20

BioEssays 28.1

22. Zhao R, Collins EJ, Bourret RB, Silversmith RE. 2002. Structure and catalytic mechanism of the E. coli chemotaxis phosphatase CheZ. Nat Struct Biol 9:570–575. 23. Wang H, Matsumura P. 1996. Characterization of the CheAS/CheZ complex: a specific interaction resulting in enhanced dephosphorylating activity on CheY-phosphate. Mol Microbiol 19:695–703. 24. Djordjevic S, Stock AM. 1997. Crystal structure of the chemotaxis receptor methyltransferase CheR suggests a conserved structural motif for binding S-adenosylmethionine. Structure 5:545–558. 25. West AH, Martinez-Hackert E, Stock AM. 1995. Crystal structure of the catalytic domain of the chemotaxis receptor methylesterase, CheB. J Mol Biol 250:276–290. 26. Okumura H, Nishiyama S, Sasaki A, Homma M, Kawagishi I. 1998. Chemotactic adaptation is altered by changes in the carboxy-terminal sequence conserved among the major methyl-accepting chemoreceptors. J Bacteriol 180:1862–1868. 27. Levit MN, Stock JB. 2002. Receptor methylation controls the magnitude of stimulus-response coupling in bacterial chemotaxis. J Biol Chem 277:36760–36765. 28. Sourjik V, Berg HC. 2002. Receptor sensitivity in bacterial chemotaxis. Proc Natl Acad Sci USA 99:123–127. 29. Szurmant H, Ordal GW. 2004. Diversity in chemotaxis mechanisms among the bacteria and archaea. Microbiol Mol Biol Rev 68:301–319. 30. Stock JB, Surette MG. 1996. Chemotaxis. In: Neidhardt FC, Curtiss R III, Ingraham JL, Lin ECC, Low KB Jr, et al. editors. Escherichia coli and Salmonella: cellular and molecular biology. 2nd ed. Washington, D.C.: ASM Press. p 1103–1129. 31. Bibikov SI, Biran R, Rudd KE, Parkinson JS. 1997. A signal transducer for aerotaxis in Escherichia coli. J Bacteriol 179:4075–4079. 32. Le Moual H, Koshland DE Jr. 1996. Molecular evolution of the C-terminal cytoplasmic domain of a superfamily of bacterial receptors involved in taxis. J Mol Biol 261:568–585. 33. Zhulin IB. 2001. The superfamily of chemotaxis transducers: from physiology to genomics and back. Adv Microb Physiol 45:157– 198. 34. Yeh JI, Biemann HP, Pandit J, Koshland DE, Kim SH. 1993. The threedimensional structure of the ligand-binding domain of a wild-type bacterial chemotaxis receptor. Structural comparison to the crosslinked mutant forms and conformational changes upon ligand binding. J Biol Chem 268:9787–9792. 35. Yeh JI, Biemann HP, Prive GG, Pandit J, Koshland DE Jr, et al. 1996. High-resolution structures of the ligand binding domain of the wild-type bacterial aspartate receptor. J Mol Biol 262:186–201. 36. Milburn MV, Prive GG, Milligan DL, Scott WG, Yeh J, et al. 1991. Threedimensional structures of the ligand-binding domain of the bacterial aspartate receptor with and without a ligand. Science 254:1342–1347. 37. Biemann HP, Koshland DE Jr. 1994. Aspartate receptors of Escherichia coli and Salmonella typhimurium bind ligand with negative and halfof-the-sites cooperativity. Biochemistry 33:629–634. 38. Ottemann KM, Thorgeirsson TE, Kolodziej AF, Shin YK, Koshland DE Jr. 1998. Direct measurement of small ligand-induced conformational changes in the aspartate chemoreceptor using EPR. Biochemistry 37:7062–7069. 39. Ottemann KM, Xiao W, Shin YK, Koshland DE Jr. 1999. A piston model for transmembrane signaling of the aspartate receptor. Science 285:1751–1754. 40. Mowbray SL, Sandgren MO. 1998. Chemotaxis receptors: a progress report on structure and function. J Struct Biol 124:257–275. 41. Kim KK, Yokota H, Kim SH. 1999. Four-helical-bundle structure of the cytoplasmic domain of a serine chemotaxis receptor. Nature 400:787– 792. 42. Singh M, Berger B, Kim PS, Berger JM, Cochran AG. 1998. Computational learning reveals coiled coil-like motifs in histidine kinase linker domains. Proc Natl Acad Sci USA 95:2738–2743. 43. Appleman JA, Chen LL, Stewart V. 2003. Probing conservation of HAMP linker structure and signal transduction mechanism through analysis of hybrid sensor kinases. J Bacteriol 185:4872–4882. 44. Mehan RS, White NC, Falke JJ. 2003. Mapping out regions on the surface of the aspartate receptor that are essential for kinase activation. Biochemistry 42:2952–2959.

Review articles

45. Surette MG, Stock JB. 1996. Role of alpha-helical coiled-coil interactions in receptor dimerization, signaling, and adaptation during bacterial chemotaxis. J Biol Chem 271:17966–17973. 46. Terwilliger TC, Bogonez E, Wang EA, Koshland DE Jr. 1983. Sites of methyl esterification on the aspartate receptor involved in bacterial chemotaxis. J Biol Chem 258:9608–9611. 47. Terwilliger TC, Koshland DE Jr. 1984. Sites of methyl esterification and deamination on the aspartate receptor involved in chemotaxis. J Biol Chem 259:7719–7725. 48. Lupas A, Van Dyke M, Stock J. 1991. Predicting coiled coils from protein sequences. Science 252:1162–1164. 49. Ames P, Studdert CA, Reiser RH, Parkinson JS. 2002. Collaborative signaling by mixed chemoreceptor teams in Escherichia coli. Proc Natl Acad Sci USA 99:7060–7065. 50. Studdert CA, Parkinson JS. 2004. Crosslinking snapshots of bacterial chemoreceptor squads. Proc Natl Acad Sci USA 101:2117– 2122. 51. Mourey L, Da Re S, Pedelacq JD, Tolstykh T, Faurie C, et al. 2001. Crystal structure of the CheA histidine phosphotransfer domain that mediates response regulator phosphorylation in bacterial chemotaxis. J Biol Chem 276:31074–31082. 52. Surette MG, Levit M, Liu Y, Lukat G, Ninfa EG, et al. 1996. Dimerization is required for the activity of the protein histidine kinase CheA that mediates signal transduction in bacterial chemotaxis. J Biol Chem 271:939–945. 53. Zhou H, Dahlquist FW. 1997. Phosphotransfer site of the chemotaxisspecific protein kinase CheA as revealed by NMR. Biochemistry 36: 699–710. 54. Zhou H, McEvoy MM, Lowry DF, Swanson RV, Simon MI, et al. 1996. Phosphotransfer and CheY-binding domains of the histidine autokinase CheA are joined by a flexible linker. Biochemistry 35:433– 443. 55. McEvoy MM, Zhou H, Roth AF, Lowry DF, Morrison TB, et al. 1995. Nuclear magnetic resonance assignments and global fold of a CheYbinding domain in CheA, the chemotaxis-specific kinase of Escherichia coli. Biochemistry 34:13871–13880. 56. Welch M, Chinardet N, Mourey L, Birck C, Samama JP. 1998. Structure of the CheY-binding domain of histidine kinase CheA in complex with CheY. Nat Struct Biol 5:25–29. 57. Stewart RC, Jahreis K, Parkinson JS. 2000. Rapid phosphotransfer to CheY from a CheA protein lacking the CheY-binding domain. Biochemistry 39:13157–13165. 58. Grebe TW, Stock JB. 1999. The histidine protein kinase superfamily. Adv Microb Physiol 41:139–227. 59. Garzon A, Parkinson JS. 1996. Chemotactic signaling by the P1 phosphorylation domain liberated from the CheA histidine kinase of Escherichia coli. J Bacteriol 178:6752–6758. 60. Varughese KI, Madhusudan, Zhou XZ, Whiteley JM, Hoch JA. 1998. Formation of a novel four-helix bundle and molecular recognition sites by dimerization of a response regulator phosphotransferase. Mol Cell 2:485–493. 61. Bilwes AM, Alex LA, Crane BR, Simon MI. 1999. Structure of CheA, a signal-transducing histidine kinase. Cell 96:131–141. 62. Levit M, Liu Y, Surette M, Stock J. 1996. Active site interference and asymmetric activation in the chemotaxis protein histidine kinase CheA. J Biol Chem 271:32057–32063. 63. Ellefson DD, Weber U, Wolfe AJ. 1997. Genetic analysis of the catalytic domain of the chemotaxis-associated histidine kinase CheA. J Bacteriol 179:825–830. 64. Bilwes AM, Quezada CM, Croal LR, Crane BR, Simon MI. 2001. Nucleotide binding by the histidine kinase CheA. Nat Struct Biol 8:353–360. 65. Stock J. 1999. Signal transduction: Gyrating protein kinases. Curr Biol 9:R364–R367. 66. Levit MN, Grebe TW, Stock JB. 2002. Organization of the receptorkinase signaling array that regulates Escherichia coli chemotaxis. J Biol Chem 277:36748–36754. 67. Griswold IJ, Dahlquist FW. 2002. The dynamic behavior of CheW from Thermotoga maritima in solution, as determined by nuclear magnetic resonance: implications for potential protein-protein interaction sites. Biophys Chem 101–102:359–373.

68. Griswold IJ, Zhou H, Matison M, Swanson RV, McIntosh LP, et al. 2002. The solution structure and interactions of CheW from Thermotoga maritima. Nat Struct Biol 9:121–125. 69. D’Aquino JA, Ringe D. 2003. Determinants of the SRC homology domain 3-like fold. J Bacteriol 185:4081–4086. 70. Zarrinpar A, Bhattacharyya RP, Lim WA. 2003. The structure and function of proline recognition domains. Sci STKE 2003:RE8. 71. Kishan KV, Newcomer ME, Rhodes TH, Guilliot SD. 2001. Effect of pH and salt bridges on structural assembly: molecular structures of the monomer and intertwined dimer of the Eps8 SH3 domain. Protein Sci 10:1046–1055. 72. Kishan KV, Scita G, Wong WT, Di Fiore PP, Newcomer ME. 1997. The SH3 domain of Eps8 exists as a novel intertwined dimer. Nat Struct Biol 4:739–743. 73. Stock AM, Mottonen JM, Stock JB, Schutt CE. 1989. Three-dimensional structure of CheY, the response regulator of bacterial chemotaxis. Nature 337:745–749. 74. Sourjik V, Berg HC. 2000. Localization of components of the chemotaxis machinery of Escherichia coli using fluorescent protein fusions. Mol Microbiol 37:740–751. 75. Mayover TL, Halkides CJ, Stewart RC. 1999. Kinetic characterization of CheY phosphorylation reactions: comparison of P-CheA and smallmolecule phosphodonors. Biochemistry 38:2259–2271. 76. Da Re SS, Deville-Bonne D, Tolstykh T, Ve´ron M, Stock JB. 1999. Kinetics of CheY phosphorylation by small molecule phosphodonors. FEBS Lett 457:323–326. 77. Gegner JA, Dahlquist FW. 1991. Signal transduction in bacteria: CheW forms a reversible complex with the protein kinase CheA. Proc Natl Acad Sci USA 88:750–754. 78. Gegner JA, Graham DR, Roth AF, Dahlquist FW. 1992. Assembly of an MCP receptor, CheW, and kinase CheA complex in the bacterial chemotaxis signal transduction pathway. Cell 70:975–982. 79. Mello BA, Tu Y. 2003. Quantitative modeling of sensitivity in bacterial chemotaxis: the role of coupling among different chemoreceptor species. Proc Natl Acad Sci USA 100:8223–8228. 80. Sourjik V, Berg HC. 2004. Functional interactions between receptors in bacterial chemotaxis. Nature 428:437–441. 81. Li M, Hazelbauer GL. 2004. Cellular stoichiometry of the components of the chemotaxis signaling complex. J Bacteriol 186:3687–3694. 82. Lybarger SR, Maddock JR. 1999. Clustering of the chemoreceptor complex in Escherichia coli is independent of the methyltransferase CheR and the methylesterase CheB. J Bacteriol 181:5527–5529. 83. Lybarger SR, Maddock JR. 2000. Differences in the polar clustering of the high- and low-abundance chemoreceptors of Escherichia coli. Proc Natl Acad Sci USA 97:8057–8062. 84. Lybarger SR, Nair U, Lilly AA, Hazelbauer GL, Maddock JR. 2005. Clustering requires modified methyl-accepting sites in low-abundance but not high-abundance chemoreceptors of Escherichia coli. Mol Microbiol 56:1078–1086. 85. Liu Y, Levit M, Lurz R, Surette MG, Stock JB. 1997. Receptor-mediated protein kinase activation and the mechanism of transmembrane signaling in bacterial chemotaxis. Embo J 16:7231–7240. 86. Francis NR, Levit MN, Shaikh TR, Melanson LA, Stock JB, et al. 2002. Subunit organization in a soluble complex of tar, CheW, and CheA by electron microscopy. J Biol Chem 277:36755–36759. 87. Francis NR, Wolanin PM, Stock JB, Derosier DJ, Thomas DR. 2004. Three-dimensional structure and organization of a receptor/signaling complex. Proc Natl Acad Sci USA 101:17480–17485. 88. Cochran AG, Kim PS. 1996. Imitation of Escherichia coli aspartate receptor signaling in engineered dimers of the cytoplasmic domain. Science 271:1113–1116. 89. Ames P, Parkinson JS. 1994. Constitutively signaling fragments of Tsr, the Escherichia coli serine chemoreceptor. J Bacteriol 176:6340– 6348. 90. Gardina PJ, Manson MD. 1996. Attractant signaling by an aspartate chemoreceptor dimer with a single cytoplasmic domain. Science 274: 425–426. 91. Tatsuno I, Homma M, Oosawa K, Kawagishi I. 1996. Signaling by the Escherichia coli aspartate chemoreceptor Tar with a single cytoplasmic domain per dimer. Science 274:423–425.

BioEssays 28.1

21

Review articles

92. Weis RM, Hirai T, Chalah A, Kessel M, Peters PJ, et al. 2003. Electron microscopic analysis of membrane assemblies formed by the bacterial chemotaxis receptor Tsr. J Bacteriol 185:3636–3643. 93. Asakura S, Honda H. 1984. Two-state model for bacterial chemoreceptor proteins. The role of multiple methylation. J Mol Biol 176:349– 367. 94. Barkai N, Leibler S. 1997. Robustness in simple biochemical networks. Nature 387:913–917. 95. Levit MN, Liu Y, Stock JB. 1999. Mechanism of CheA protein kinase activation in receptor signaling complexes. Biochemistry 38:6651–6658. 96. Barnakov AN, Barnakova LA, Hazelbauer GL. 1998. Comparison in vitro of a high- and a low-abundance chemoreceptor of Escherichia coli: similar kinase activation but different methyl-accepting activities. J Bacteriol 180:6713–6718. 97. Barnakov AN, Barnakova LA, Hazelbauer GL. 1999. Efficient adaptational demethylation of chemoreceptors requires the same enzymedocking site as efficient methylation. Proc Natl Acad Sci USA 96:10667–10672. 98. Weerasuriya S, Schneider BM, Manson MD. 1998. Chimeric chemoreceptors in Escherichia coli: signaling properties of Tar-Tap and TapTar hybrids. J Bacteriol 180:914–920.

22

BioEssays 28.1

99. Feng X, Lilly AA, Hazelbauer GL. 1999. Enhanced function conferred on low-abundance chemoreceptor Trg by a methyltransferase-docking site. J Bacteriol 181:3164–3171. 100. Kofoid EC, Parkinson JS. 1991. Tandem translation starts in the cheA locus of Escherichia coli. J Bacteriol 173:2116–2119. 101. Cantwell BJ, Draheim RR, Weart RB, Nguyen C, Stewart RC, et al. 2003. CheZ phosphatase localizes to chemoreceptor patches via CheA-short. J Bacteriol 185:2354–2361. 102. Koshland DE Jr, Goldbeter A, Stock JB. 1982. Amplification and adaptation in regulatory and sensory systems. Science 217:220– 225. 103. Park S, Wolanin PM, Yuzbashyan EA, Lin H, Darnton NC, et al. 2003. Influence of topology on bacterial social interaction. Proc Natl Acad Sci USA 100:13910–13915. 104. Park S, Wolanin PM, Yuzbashyan EA, Silberzan P, Stock JB, et al. 2003. Motion to form a quorum. Science 301:188. 105. Shapiro JA. 1998. Thinking about bacterial populations as multicellular organisms. Annu Rev Microbiol 52:81–104. 106. Guex N, Peitsch MC. 1997. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714–2723.