Cyclic di-GMP-mediated repression of swarming motility by Pseudomonas aeruginosa PA14 requires the MotAB stator.

The second messenger cyclic diguanylate (c-di-GMP) plays a critical role in the regulation of motility. In Pseudomonas aeruginosa PA14, c-di-GMP inversely controls biofilm formation and surface swarming motility, with high levels of this dinucleotide signal stimulating biofilm formation and repressing swarming. P. aeruginosa encodes two stator complexes, MotAB and MotCD, that participate in the function of its single polar flagellum. Here we show that the repression of swarming motility requires a functional MotAB stator complex. Mutating the motAB genes restores swarming motility to a strain with artificially elevated levels of c-di-GMP as well as stimulates swarming in the wild-type strain, while overexpression of MotA from a plasmid represses swarming motility. Using point mutations in MotA and the FliG rotor protein of the motor supports the conclusion that MotA-FliG interactions are critical for c-di-GMP-mediated swarming inhibition. Finally, we show that high c-di-GMP levels affect the localization of a green fluorescent protein (GFP)-MotD fusion, indicating a mechanism whereby this second messenger has an impact on MotCD function. We propose that when c-di-GMP level is high, the MotAB stator can displace MotCD from the motor, thereby affecting motor function. Our data suggest a newly identified means of c-di-GMP-mediated control of surface motility, perhaps conserved among Pseudomonas, Xanthomonas, and other organisms that encode two stator systems.

Spatiotemporal modelling of CheY complexes in Escherichia coli chemotaxis.

The chemotaxis pathway of Escherichia coli is one of the best studied and modelled biological signalling pathways. Here we extend existing modelling approaches by explicitly including a description of the formation and subcellular localization of intermediary complexes in the phosphotransfer pathway. The inclusion of these complexes shows that only about 60% of the total output response regulator (CheY) is uncomplexed at any moment and hence free to interact with its target, the flagellar motor. A clear strength of this model is its ability to predict the experimentally observable subcellular localization of CheY throughout a chemotactic response. We have found good agreement between the model output and experimentally determined CheY localization patterns.

Overview of mathematical approaches used to model bacterial chemotaxis II: bacterial populations.

We review the application of mathematical modeling to understanding the behavior of populations of chemotactic bacteria. The application of continuum mathematical models, in particular generalized Keller-Segel models, is discussed along with attempts to incorporate the microscale (individual) behavior on the macroscale, modeling the interaction between different species of bacteria, the interaction of bacteria with their environment, and methods used to obtain experimentally verified parameter values. We allude briefly to the role of modeling pattern formation in understanding collective behavior within bacterial populations. Various aspects of each model are discussed and areas for possible future research are postulated.

Overview of mathematical approaches used to model bacterial chemotaxis I: the single cell.

Mathematical modeling of bacterial chemotaxis systems has been influential and insightful in helping to understand experimental observations. We provide here a comprehensive overview of the range of mathematical approaches used for modeling, within a single bacterium, chemotactic processes caused by changes to external gradients in its environment. Specific areas of the bacterial system which have been studied and modeled are discussed in detail, including the modeling of adaptation in response to attractant gradients, the intracellular phosphorylation cascade, membrane receptor clustering, and spatial modeling of intracellular protein signal transduction. The importance of producing robust models that address adaptation, gain, and sensitivity are also discussed. This review highlights that while mathematical modeling has aided in understanding bacterial chemotaxis on the individual cell scale and guiding experimental design, no single model succeeds in robustly describing all of the basic elements of the cell. We conclude by discussing the importance of this and the future of modeling in this area.

Targeting of two signal transduction pathways to different regions of the bacterial cell.

Components of bacterial chemosensory pathways which sense via transmembrane receptors have been shown to localize to the cell poles. Many species, however, have operons encoding multiple putative chemosensory pathways, some including putative cytoplasmic receptors. In-genome fusions to single or multiple genes encoding components of two chemosensory pathways in Rhodobacter sphaeroides, cheOp2 and cheOp3, revealed that while sensory transducing proteins associated with transmembrane receptors and encoded on cheOp2 were targeted to the cell poles, the proteins associated with putative cytoplasmic receptors and encoded on cheOp3 were all targeted to a cytoplasmic cluster. No proteins were localized to both sites. These data show that bacteria target components of related pathways to different sites in the cell, presumably preventing direct cross-talk between the different pathways, but allowing a balanced response between extracellular and cytoplasmic signals. It also indicates that there is intracellular organization in bacterial cells, with specific proteins targeted and localized to cytoplasmic regions.

TlpC, a novel chemotaxis protein in Rhodobacter sphaeroides, localizes to a discrete region in the cytoplasm.

TlpC is encoded in the second chemotaxis operon of Rhodobacter sphaeroides. This protein shows some homology to membrane-spanning chemoreceptors of many bacterial species but, unlike these, is essential for R. sphaeroides chemotaxis to all compounds tested. Genomic replacement of tlpC with a C-terminal gfp fusion demonstrated that TlpC localized to a discrete cluster within the cytoplasm. Immunogold electron microscopy also showed that TlpC localized to a cytoplasmic electron-dense region. Correct TlpC-GFP localization depended on the downstream signalling proteins, CheW3, CheW4 and CheA2, and was tightly linked to cell division. Newly divided cells contained a single cluster but, as the cell cycle progressed, a second cluster appeared close to the initial cluster. As elongation continued, these clusters moved apart so that, on septation, each daughter cell contained a single TlpC cluster. The data presented suggest that TlpC is either a cytoplasmic chemoreceptor responding to or integrating global signals of metabolic state or a novel and essential component of the chemotaxis signalling pathway. These data also suggest that clustering is essential for signalling and that a mechanism may exist for targeting and localizing proteins within the bacterial cytoplasm.

CheR- and CheB-dependent chemosensory adaptation system of Rhodobacter sphaeroides.

Rhodobacter sphaeroides has multiple homologues of most of the Escherichia coli chemotaxis genes, organized in three major operons and other, unlinked, loci. These include cheA(1) and cheR(1) (che Op(1)) and cheA(2), cheR(2), and cheB(1) (che Op(2)). In-frame deletions of these cheR and cheB homologues were constructed and the chemosensory behaviour of the resultant mutants examined on swarm plates and in tethered cell assays. Under the conditions tested, CheR(2) and CheB(1) were essential for normal chemotaxis, whereas CheR(1) was not. cheR(2) and cheB(1), but not cheR(1), were also able to complement the equivalent E. coli mutants. However, none of the proteins were required for the correct polar localization of the chemoreceptor McpG in R. sphaeroides. In E. coli, CheR binds to the NWETF motif on the high-abundance receptors, allowing methylation of both high- and low-abundance receptors. This motif is not contained on any R. sphaeroides chemoreceptors thus far identified, although 2 of the 13 putative chemoreceptors, McpA and TlpT, do have similar sequences. This suggests that CheR(2) either interacts with the NWETF motif of E. coli methyl-accepting chemotaxis proteins (MCPs), even though its native motif may be slightly different, or with another conserved region of the MCPs. Methanol release measurements show that R. sphaeroides has an adaptation system that is different from that of Bacillus subtilis and E. coli, with methanol release measurable on the addition of attractant but not on its removal. Intriguingly, CheA(2), but not CheA(1), is able to phosphorylate CheB(1), suggesting that signaling through CheA(1) cannot initiate feedback receptor adaptation via CheB(1)-P.

The roles of the multiple CheW and CheA homologues in chemotaxis and in chemoreceptor localization in Rhodobacter sphaeroides.

Rhodobacter sphaeroides has multiple homologues of most of the Escherichia coli chemotaxis genes, organized in two major operons and other, unlinked, loci. These include cheA1 and cheW1 (che Op1) and cheA2, cheW2 and cheW3 (che Op2). We have deleted each of these cheA and cheW homologues in-frame and examined the chemosensory behaviour of these strains on swarm plates and in tethered cell assays. In addition, we have examined the effect of these deletions on the polar localization of the chemoreceptor McpG. In E. coli, deletion of either cheA or cheW results in a non-chemotactic phenotype, and these strains also show no receptor clustering. Here, we demonstrate that CheW2 and CheA2 are required for the normal localization of McpG and for normal chemotactic responses under both aerobic and photoheterotrophic conditions. Under aerobic conditions, deletion of cheW3 has no significant effect on McpG localization and only has an effect on chemotaxis to shallow gradients in swarm plates. Under photoheterotrophic conditions, however, CheW3 is required for McpG localization and also for chemotaxis both on swarm plates and in the tethered cell assay. These phenotypes are not a direct result of delocalization of McpG, as this chemoreceptor does not mediate chemotaxis to any of the compounds tested and can therefore be considered a marker for general methyl-accepting chemotaxis protein (MCP) clustering. Thus, there is a correlation between the normal localization of McpG (and presumably other chemoreceptors) and chemotaxis. We propose a model in which the multiple different MCPs in R. sphaeroides are contained within a polar chemoreceptor cluster. Deletion of cheW2 and cheA2 under both aerobic and photoheterotrophic conditions, and cheW3 under photoheterotrophic conditions, disrupts the cluster and hence reduces chemotaxis to any compound sensed by these MCPs.

The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1.

Rhodobacter sphaeroides 2.4.1 is an alpha-3 purple nonsulfur eubacterium with an extensive metabolic repertoire. Under anaerobic conditions, it is able to grow by photosynthesis, respiration and fermentation. Photosynthesis may be photoheterotrophic using organic compounds as both a carbon and a reducing source, or photoautotrophic using carbon dioxide as the sole carbon source and hydrogen as the source of reducing power. In addition, R. sphaeroides can grow both chemoheterotrophically and chemoautotrophically. The structural components of this metabolically diverse organism and their modes of integrated regulation are encoded by a genome of approximately 4.5 Mb in size. The genome comprises two chromosomes CI and CII (2.9 and 0.9 Mb, respectively) and five other replicons. Sequencing of the genome has been carried out by two groups, the Joint Genome Institute, which carried out shotgun-sequencing of the entire genome and The University of Texas-Houston Medical School, which carried out a targeted sequencing strategy of CII. Here we describe our current understanding of the genome when data from both of these groups are combined. Previous work had suggested that the two chromosomes are equal partners sharing responsibilities for fundamental cellular processes. This view has been reinforced by our preliminary analysis of the virtually completed genome sequence. We also have some evidence to suggest that two of the plasmids, pRS241a and pRS241b encode chromosomal type functions and their role may be more than that of accessory elements, perhaps representing replicons in a transition state.

Behavioral responses of Rhodobacter sphaeroides to linear gradients of the nutrients succinate and acetate.

Rhodobacter sphaeroides cells were tethered by their flagella and subjected to increasing and decreasing nutrient gradients. Using motion analysis, changes in flagellar motor rotation were measured and the responses of the cells to the chemotactic gradients were determined. The steepness and concentration ranges of increasing and decreasing gradients were varied, and the bacterial responses were measured. This allowed the limits of gradients that would invoke changes in flagellar behavior to be determined and thus predicts the nature of gradients that would evoke chemotaxis in the environment. The sensory threshold was measured at 30 nM, and the response showed saturation at 150 microM. The study determined that cells detected and responded to changing concentration rates as low as 1 nM/s for acetate and 5 nM/s for succinate. The complex sensory system of R. sphaeroides responded to both increasing and decreasing concentration gradients of attractant with different sensitivities. In addition, transition phases involving changes in the motor speed and the smoothness of motor rotation were found.

Fine tuning bacterial chemotaxis: analysis of Rhodobacter sphaeroides behaviour under aerobic and anaerobic conditions by mutation of the major chemotaxis operons and cheY genes.

Rhodobacter sphaeroides chemotaxis is significantly more complex than that of enteric bacteria. Rhodobacter sphaeroides has multiple copies of chemotaxis genes (two cheA, one cheB, two cheR, three cheW, five cheY but no cheZ), controlling a single 'stop-start' flagellum. The growth environment controls the level of expression of different groups of genes. Tethered cell analysis of mutants suggests that CheY(4) and CheY(5) are the motor-binding response regulators. The histidine protein kinase CheA(2) mediates an attractant ('normal') response via CheY(4), while CheA(1) and CheY(5) appear to mediate a repellent ('inverted') response. CheY(3) facilitates signal termination, possibly acting as a phosphate sink, although CheY(1) and CheY(2) can substitute. The normal and inverted responses may be initiated by separate sets of chemoreceptors with their relative strength dependent on growth conditions. Rhodobacter sphaeroides may use antagonistic responses through two chemosensory pathways, expressed at different levels in different environments, to maintain their position in a currently optimum environment. Complex chemotaxis systems are increasingly being identified and the strategy adopted by R.sphaeroides may be common in the bacterial kingdom.

Inverted behavioural responses in wild-type Rhodobacter sphaeroides to temporal stimuli.

Both aerobically and photosynthetically grown wild-type Rhodobacter sphaeroides swarmed through soft nutrient agar. However, individual aerobically and photosynthetically grown tethered cells showed different responses to steps in concentrations of some attractants. Photosynthetically grown cells showed little response to a step-up in attractant, but large response to a step-down. Aerobically grown cells showed a large but opposite response to a step-up of chemoeffectors such as succinate and aspartate. The responses in che operon deletion mutants were also investigated and indicated that the aerobic response may depend on the protein products of che operon 1.

Identification and localization of a methyl-accepting chemotaxis protein in Rhodobacter sphaeroides.

Genes coding for a classical membrane spanning chemoreceptor (mcpG) and a response regulator (cheY4) were identified in a region of Rhodobacter sphaeroides DNA unlinked to either of the two previously identified chemosensory operons. Immunogold electron microscopy had shown that the expression of chemoreceptors in R. sphaeroides varies with growth conditions. Using GFP fused to the newly identified McpG, we examined the targeting of this single methyl-accepting chemotaxis protein (MCP) under different growth conditions. The gene encoding the C-terminal McpG-GFP fusion was introduced by homologous recombination into the chromosome, replacing the wild-type gene. The resultant protein localized to the poles of the cell under aerobic, photoheterotrophic and anaerobic dark conditions, demonstrating that this MCP is expressed under all three growth conditions. More protein was always found at one pole than the other. The polar fluorescence increased during the cell cycle, with protein becoming evident at the second pole around the time of septation. At division, each daughter cell had a label at one pole, but the intensity of fluorescence was higher in the daughter cell containing the original labelled pole. McpG localization was not altered in a che Operon 1 deletion strain, lacking CheW1 and CheA1, but a che Operon 2 deletion strain, lacking CheW2, CheW3 and CheA2, showed significantly reduced polar localization. This observation indicates that polar localization of McpG depends on Che proteins encoded by Operon 2, but not homologues encoded by Operon 1.

Response kinetics of tethered Rhodobacter sphaeroides to changes in light intensity.

Rhodobacter sphaeroides can swim toward a wide range of attractants (a process known as taxis), propelled by a single rotating flagellum. The reversals of motor direction that cause tumbles in Eschericia coli taxis are replaced by brief motor stops, and taxis is controlled by a complex sensory system with multiple homologues of the E. coli sensory proteins. We tethered photosynthetically grown cells of R. sphaeroides by their flagella and measured the response of the flagellar motor to changes in light intensity. The unstimulated bias (probability of not being stopped) was significantly larger than the bias of tethered E. coli but similar to the probability of not tumbling in swimming E. coli. Otherwise, the step and impulse responses were the same as those of tethered E. coli to chemical attractants. This indicates that the single motor and multiple sensory signaling pathways in R. sphaeroides generate the same swimming response as several motors and a single pathway in E. coli, and that the response of the single motor is directly observable in the swimming pattern. Photo-responses were larger in the presence of cyanide or the uncoupler carbonyl cyanide 4-trifluoromethoxyphenylhydrazone (FCCP), consistent with the photo-response being detected via changes in the rate of electron transport.

Identification of a fourth cheY gene in Rhodobacter sphaeroides and interspecies interaction within the bacterial chemotaxis signal transduction pathway.

The Escherichia coli chemotaxis signal transduction pathway has: CheA, a histidine protein kinase; CheW, a linker between CheA and sensory proteins; CheY, the effector; and CheZ, a signal terminator. Rhodobacter sphaeroides has multiple copies of these proteins (2 x CheA, 3 x CheW and 3 x CheY, but no CheZ). In this study, we found a fourth cheY and expressed these R. sphaeroides proteins in E. coli. CheA2 (but not CheA1) restored swarming to an E. coli cheA mutant (RP9535). CheW3 (but not CheW2) restored swarming to a cheW mutant of E. coli (RP4606). R. sphaeroides CheYs did not affect E. coli lacking CheY, but restored swarming to a cheZ strain (RP1616), indicating that they can act as signal terminators in E. coli. An E. coli CheY, which is phosphorylated but cannot bind the motor (CheY109KR), was expressed in RP1616 but had no effect. Overexpression of CheA2, CheW2, CheW3, CheY1, CheY3 and CheY4 inhibited chemotaxis of wild-type E. coli (RP437) by increasing its smooth-swimming bias. While some R. sphaeroides proteins restore tumbling to smooth-swimming E. coli mutants, their activity is not controlled by the chemosensory receptors. R. sphaeroides possesses a phosphorelay cascade compatible with that of E. coli, but has additional incompatible homologues.

Bacterial tactic responses.

Many, if not most, bacterial species swim. The synthesis and operation of the flagellum, the most complex organelle of a bacterium, takes a significant percentage of cellular energy, particularly in the nutrient limited environments in which many motile species are found. It is obvious that motility accords cells a survival advantage over non-motile mutants under normal, poorly mixed conditions and is an important determinant in the development of many associations between bacteria and other organisms, whether as pathogens or symbionts and in colonization of niches and the development of biofilms. This survival advantage is the result of sensory control of swimming behaviour. Although too small to sense a gradient along the length of the cell, and unable to swim great distances because of buffetting by Brownian motion and the curvature resulting from a rotating flagellum, bacteria can bias their random swimming direction towards a more favourable environment. The favourable environment will vary from species to species and there is now evidence that in many species this can change depending on the current physiological growth state of the cell. In general, bacteria sense changes in a range of nutrients and toxins, compounds altering electron transport, acceptors or donors into the electron transport chain, pH, temperature and even the magnetic field of the Earth. The sensory signals are balanced, and may be balanced with other sensory pathways such as quorum sensing, to identify the optimum current environment. The central sensory pathway in this process is common to most bacteria and most effectors. The environmental change is sensed by a sensory protein. In most species examined this is a transmembrane protein, sensing the external environment, but there is increasing evidence for additional cytoplasmic receptors in many species. All receptors, whether sensing sugars, amino acids or oxygen, share a cytoplasmic signalling domain that controls the activity of a histidine protein kinase, CheA, via a linker protein, CheW. A reduction in an attractant generally leads to the increased autophosphorylation of CheA. CheA passes its phosphate to a small, single domain response regulator, CheY. CheY-P can interact with the flagellar motor to cause it to change rotational direction or stop. Signal termination either via a protein, CheZ, which increases the dephosphorylation rate of CheY-P or via a second CheY which acts as a phosphate sink, allows the cell to swim off again, usually in a new direction. In addition to signal termination the receptor must be reset, and this occurs via methylation of the receptor to return it to a non-signalling conformation. The way in which bacteria use these systems to move to optimum environments and the interaction of the different sensory pathways to produce species-specific behavioural response will be the subject of this review.

The bacterial flagella motor.

The bacterial flagellum is probably the most complex organelle found in bacteria. Although the ribosome may be made of slightly more subunits, the bacterial flagellum is a more organized and complex structure. The limited number of flagella must be targeted to the correct place on the cell membrane and a structure with cytoplasmic, cytoplasmic membrane, outer membrane and extracellular components must be assembled. The process of controlled transcription and assembly is still not fully understood. Once assembled, the motor complex in the cytoplasmic membrane rotates, driven by the transmembrane ion gradient, at speeds that can reach many 100 Hz, driving the bacterial cell at several body lengths a second. This coupling of an electrochemical gradient to mechanical rotational work is another fascinating feature of the bacterial motor. A significant percentage of a bacterium's energy may be used in synthesizing the complex structure of the flagellum and driving its rotation. Although patterns of swimming may be random in uniform environments, in the natural environment, where cells are confronted with gradients of metabolites and toxins, motility is used to move bacteria towards their optimum environment for growth and survival. A sensory system therefore controls the switching frequency of the rotating flagellum. This review deals primarily with the structure and operation of the bacterial flagellum. There has been a great deal of research in this area over the past 20 years and only some of this has been included. We apologize in advance if certain areas are covered rather thinly, but hope that interested readers will look at the excellent detailed reviews on those areas cited at those points.

Transformations in flagellar structure of Rhodobacter sphaeroides and possible relationship to changes in swimming speed.

Rhodobacter sphaeroides is a photosynthetic bacterium which swims by rotating a single flagellum in one direction, periodically stopping, and reorienting during these stops. Free-swimming R. sphaeroides was examined by both differential interference contrast (DIC) microscopy, which allows the flagella of swimming cells to be seen in vivo, and tracking microscopy, which tracks swimming patterns in three dimensions. DIC microscopy showed that when rotation stopped, the helical flagellum relaxed into a high-amplitude, short-wavelength coiled form, confirming previous observations. However, DIC microscopy also revealed that the coiled filament could rotate slowly, reorienting the cell before a transition back to the functional helix. The time taken to reform a functional helix depended on the rate of rotation of the helix and the length of the filament. In addition to these coiled and helical forms, a third conformation was observed: a rapidly rotating, apparently straight form. This form took shape from the cell body out and was seen to form directly from flagella that were initially in either the coiled or the helical conformation. This form was always significantly longer than the coiled or helical form from which it was derived. The resolution of DIC microscopy made it impossible to identify whether this form was genuinely in a straight conformation or was a low-amplitude, long-wavelength helix. Examination of the three-dimensional swimming pattern showed that R. sphaeroides changed speed while swimming, sometimes doubling the swimming speed between stops. The rate of acceleration out of stops was also variable. The transformations in waveform are assumed to be torsionally driven and may be related to the changes in speed measured in free-swimming cells. The roles of and mechanisms that may be involved in the transformations of filament conformations and changes in swimming speed are discussed.

Localization and environmental regulation of MCP-like proteins in Rhodobacter sphaeroides.

Chemotaxis to many compounds by Rhodobacter sphaeroides requires transport and at least partial metabolism of the chemoeffector. Previous investigations using phototrophically grown cells have failed to find any homologues of the MCP chemoreceptors identified in Escherichia coli. However, using an antibody raised against the highly conserved domain of E. coli Tsr, MCP-like proteins were identified in R. sphaeroides WS8N. Analysis using Western blotting and immunogold electron microscopy showed that expression of these MCP-like proteins is environmentally regulated and that receptors are targeted to two different cellular locations: the poles of the cells and the cytoplasm. In aerobically grown cells, these proteins were shown by immunoelectron microscopy to localize predominantly to the cell poles and to an electron-dense body in the cytoplasm. Western blot analysis indicated a 17-fold reduction in protein concentration when cells were grown in the light. The number of immunogold particles was also dramatically reduced in anaerobically light-grown cells and their cellular distribution was altered. Fewer receptors localized to the cell poles and more particles randomly distributed within the cell, but the cytoplasmic cluster remained. These trends were more pronounced in cells grown anaerobically under dim light than in those grown anaerobically under bright light, suggesting that expression is controlled by redox state and either light intensity or the extent of photosynthetic membrane synthesis. Recent work on E. coli chemosensing suggests that oligomerization of receptors and chemosensory proteins is important for sensory signalling. The data presented here suggest that this oligomerization can occur with cytoplasmic receptors and also provides an explanation for the multiple copies of chemosensory proteins in R. sphaeroides.

Roles of chemosensory pathways in transient changes in swimming speed of Rhodobacter sphaeroides induced by changes in photosynthetic electron transport.

The response of free-swimming Rhodobacter sphaeroides to increases and decreases in the intensity of light of different wavelengths was analyzed. There was a transient (1 to 2 s) increase in swimming speed in response to an increase in light intensity, and there was a similar transient stop when the light intensity decreased. Measurement of changes in membrane potential and the use of electron transport inhibitors showed that the transient increase in swimming speed, following an increase in light intensity, and the stop following its decrease were the result of changes in photosynthetic electron transport. R. sphaeroides has two operons coding for multiple homologs of the enteric chemosensory genes. Mutants in the first chemosensory operon showed wild-type photoresponses. Mutants with the cheA gene of the second operon (cheAII) deleted, either with or without the first operon present, showed inverted photoresponses, with free-swimming cells stopping on an increase in light intensity and increasing swimming speed on a decrease. These mutants also lacked adaptation. Transposon mutants with mutations in cheAII, which also reduced expression of downstream genes, however, showed no photoresponses. These results show that (i) free-swimming cells respond to both an increase and a decrease in light intensity (tethered cells only show the stopping on a step down in light intensity), (ii) the signal comes from photosynthetic electron transfer, and (iii) the signal is primarily channelled through the second chemosensory pathway. The different responses shown by the cheAII deletion and insertion mutants suggest that CheWII is required for photoresponses, and a third sensory pathway can substitute for CheAII as long as CheWII is present. The inverted response suggests that transducers are involved in photoresponses as well as chemotactic responses.