Miniaturization of on-board actuation and powering engenders the proliferation of biohybrid microrobots, which integrate motile bacteria or cells with synthetic functional components to achieve micron-scale actuations. Flagellated bacteria like S. marcescens are among the leading candidates for the actuators of swimming microrobots. However, the high intrinsic stochasticity in bacteria-driven microrobots severely limits their potential applications, such as targeted drug delivery. Taxis behaviors (e.g., chemotaxis), which help free-swimming bacteria to navigate towards favorable environments and away from hazardous ones, may offer an elegant means to control the motion of bacteria-driven microrobots. Therefore, this thesis focuses on: (a) addressing the motion guiding of bacteria-driven microrobots using common bacterial taxis behaviors, specifically chemotaxis and pH-taxis, (b) explaining the physical mechanisms associated with the tactic motions in bacteria-driven microrobots, and (c) developing a biophysical model to describe the bacterial propulsion and the chemotaxis in bacteria-driven microrobots. In order to produce considerable chemotactic motion in bacteria-driven microrobots, an appropriate chemical concentration profile needs to be determined, which requires the knowledge of the chemotaxis response of the integrated bacterial species. Thus, we first propose an experimental and modeling framework to characterize bacterial chemotaxis. The chemotaxis response of a species against a chemoattractant is experimentally quantified under a linear concentration gradient of the attractant. A signaling pathway model is fitted to the experimental measurements over a series of gradients to determine the species-specific parameters in the model, thereby fulfilling an analytical characterization of the chemotaxis. Subsequently, in a multi-bacteria-driven microrobotic system, we quantify the chemotactic drift of the microrobotic swarms towards a potent chemoattractant L-serine and elucidate the physical mechanisms associated with the drift motion by statistical trajectory analysis. It shows that the microrobots have an apparent heading preference for moving up the gradient, which constitutes the major factor that produces the chemotactic drift. The apparent heading bias is caused by a higher persistence in the heading direction when a microrobot moves up the the L-serine gradient compared to traveling down the gradient. Besides chemotaxis, we explore the potential of utilizing ambient pH to guide the motion of the bacteria-driven microrobots. Under three different pH gradients, we demonstrate that the microrobots exhibit both unidirectional and bidirectional pH-tactic behaviors. Two factors, a swimming heading bias and a speed bias, are found to be responsible for the pH-tactic motion while the heading bias contributes more. Like in chemotaxis, the heading directions of the microrobots are also significantly more persistent when they move towards favored pH regions. Finally, a biophysical model is developed to describe the bacterial propulsion and the chemotaxis in an extensively adopted design of bacteria-driven microrobots. The model traces helical trajectories and chemotactic motion that resemble those observed from experiments, which validates the basic correctness of the model. The model simulation also suggests that the seemingly collective chemotaxis among the multiple bacteria attached to a microrobot could be explained by a synchronized signaling pathway response among these bacteria. Furthermore, we investigate the dependencies of the microrobots’ per

No description available.

The formation of dendritic arbors is necessary for the proper establishment of neuronal circuits. The Drosophila transcription factor Spineless has been shown to play an important role in the control of dendritic morphogenesis, although the pathways through which it functions are not completely understood. Here, we show genetic evidence that Spineless interacts with the actin/microtubule cross linking protein Shortstop to control the dendrite arbor development of the dendritic arborization (da) sensory neurons. In addition, we have discovered a novel function for spineless as we show that spineless mutant larvae exhibit an increased sensitivity to specific odorants in the absence of morphological defects of the chemosensory organs. These data show that spineless acts in multiple cell-specific contexts to control the diversification of sensory neuron morphology and function.

This study focuses on the mechanism of transmembrane signaling by Tar, the aspartate chemoreceptor of Escherichia coli. Like other bacterial chemoreceptors, Tar localizes to the cell membrane and relays information about the external chemical environment through the membrane to a cytoplasmic signaling domain. The output of the signaling domain controls the directional bias of the rotary flagellar motors of the cell. Net movement of a cell in a chemical gradient involves temporal comparison of the current concentration with the concentration in the recent (a few seconds) past. The current concentration is measured as the percent occupancy of the extracellular ligand-binding domain of the receptor, and the past is represented by the extent of covalent methylation of four conserved glutamyl residues in the cytoplasmic domain. Under steady-state conditions, the methylation level corresponds to ligand occupancy. Tar is a dimer, and much evidence suggests that dimers associate into trimers of dimers. Higher-order arrays of receptors form in the presence of the cytoplasmic proteins CheA and CheW. The conformational change generated by ligand binding is transmitted through the membrane by one subunit of a dimer. To examine whether this initially asymmetric signal becomes symmetric within the cytoplasmic domain, I examined aspartate-induced adaptive methylation of the two subunits of mutant Tar receptor heterodimers. In the presence of CheA and CheW, adaptive methylation after addition of aspartate was symmetric, but in their absence, although the level of methylation increased, the rates were different for the two subunits. I also found that cross-talk, at the level of adaptive methylation, occurs between different receptor types even in the absence of CheA and CheW. These results provide support for the idea that a tight association of receptor dimers within trimers of dimers allows for an actively signaling receptor to affect the methylation state, and thus presumably the signaling state, of receptors within a trimer that are not bound to an attractant ligand.

The transmembrane serine receptor Tsr associates with a coupling protein, CheW, and a histidine kinase, CheA, to form a ternary complex that regulates the activity of CheA. CheA activity is inhibited by binding of L-serine to Tsr. This work aims to characterize the ligand-binding properties of Tsr and the inhibitory effect of L-serine on CheA activity. The periplasmic domain of Tsr (pTsr) was purified and characterized. Analytical gel filtration and analytical ultracentrifugation indicated that binding of Lserine promotes dimerization. The binding stoichiometry and dissociation constant for binding of L-serine to pTsr were determined by fluorescence spectroscopy. As protein concentration decreased, the dissociation constant increased. A working model was proposed to account for the interactions between L-serine and pTsr. The activity of CheA in a ternary complex with full-length Tsr and CheW was analyzed by measuring the production of [32P]-phospho-CheY. (Phospho-CheY is the product of CheA catalysis.) The results revealed that binding of L-serine decreased CheA activity without changing its affinity for ATP. These findings suggest that the allosteric effect of L-serine on CheA activity might occur through V-type inhibition. Optimization of an alternative, continuous, non-radioactive assay for CheA is underway.

This study focuses on the mechanism of transmembrane signaling by Tar, the aspartate chemoreceptor of Escherichia coli. Like other bacterial chemoreceptors, Tar localizes to the cell membrane and relays information about the external chemical environment through the membrane to a cytoplasmic signaling domain. The output of the signaling domain controls the directional bias of the rotary flagellar motors of the cell. Net movement of a cell in a chemical gradient involves temporal comparison of the current concentration with the concentration in the recent (a few seconds) past. The current concentration is measured as the percent occupancy of the extracellular ligand-binding domain of the receptor, and the past is represented by the extent of covalent methylation of four conserved glutamyl residues in the cytoplasmic domain. Under steady-state conditions, the methylation level corresponds to ligand occupancy. Tar is a dimer, and much evidence suggests that dimers associate into trimers of dimers. Higher-order arrays of receptors form in the presence of the cytoplasmic proteins CheA and CheW. The conformational change generated by ligand binding is transmitted through the membrane by one subunit of a dimer. To examine whether this initially asymmetric signal becomes symmetric within the cytoplasmic domain, I examined aspartate-induced adaptive methylation of the two subunits of mutant Tar receptor heterodimers. In the presence of CheA and CheW, adaptive methylation after addition of aspartate was symmetric, but in their absence, although the level of methylation increased, the rates were different for the two subunits. I also found that cross-talk, at the level of adaptive methylation, occurs between different receptor types even in the absence of CheA and CheW. These results provide support for the idea that a tight association of receptor dimers within trimers of dimers allows for an actively signaling receptor to affect the methylation state, and thus presumably the signaling state, of receptors within a trimer that are not bound to an attractant ligand.

No description available.

No description available.

Three behavioural mutants of A. tumefaciens C58C1 (mot-l, mot-12 and fla-15) generated by transposon (Tn5) mutagenesis were studied. Analysis was initially at the molecular level, as a cosmid, pDUB1900, from a representative genomic library of C58C1 had been isolated that complemented the mutants. A region of 8624 nucleotides to which the Tn5 insertion sites of the three mutants had been mapped was sequenced completely in both directions. The comparison of this sequence with sequence databases and other computer analyses revealed six flagellar gene homologues (flgI,flgH,fliP,flaA,flaB,flaC), three open reading frames (ORFA, B and C) with no significant sequence identity to any open reading frames in the databases and the partial sequence of the flagellar gene homologue flgG. Computer analysis also showed that theflgH,flgI andfliP homologues, and ORFs A, B and C, could form the downstream region of a larger operon involved in chemotactic and motility functions. However putative transcription signals were also found within the operon. A new mutant (MANl) was created in the last gene (fliP) of the putative operon to investigate the function of possible transcription signals in the open reading frame immediately upstream of it (ORFC). The mot-12 mutant phenotype of fully synthesised but paralysed flagella is brought about by the insertion of Tn5 in ORFC. ORFC contains a possible promoter for fliP. The Tn5 insertion in ORFC should have polar effects upon the expression of fliP, unless the putative promoter can cause expression of fliP. The MANl mutant had a flagella-less phenotype. FliP in other bacteria is required early in the synthesis of flagella and the null phenotype is/7a-. Thus for flagella to be present in mot-12 suggests fliP must have a promoter. The ORFC sequence is highly conserved in R. meliloti and the overall regulation of these flagellar gene homologues may be as an operon with other regulatory signals. Evidence from other operons (including motility operons) with multiple transcription signals is discussed. The flaABC homologues were multiple copies of the gene encoding the flagellin protein of the flagellum. The mot-l phenotype of severely truncated filaments was caused by a Tn5 insertion in flaA. Analysis of the sequence showed flaABC to each have transcription signals that could lead to separate transcription. Transcription analysis by Northern blotting showed flaA to be transcribed monocistronically. Flagella were isolated from A. twnefaciens and the flagellins separated by SDS-PAGE. The migrated distances (relative to those of markers) was not as predicted from the nucleotide sequence. This anomaly could be caused by unequivalent binding of SDS or post-translational modification of FlaA. The A. tumefaciens flagellar genes were most similar to those of R. meliloti. However A. tumefaciens flagella do not exhibit the characteristic cross-hatching of the complex flagella of R. meliloti. This study also showed A. tumefaciens flagella not to be dependent on divalent cations for subunit associations unlike R. meliloti. These properties of A. tumefaciens flagella were similar to those of R. leguminosarum.The open reading frames found were isolated, radiolabelled and used as probesagainst Southern blots containing chromosomal DNA from a variety of soil bacteria, and cosmids known to contain motility genes in R. meliloti. Hybridisation revealed homologous DNA sequences in a number of these bacteria. All the A. tumefaciens open reading frames hybridised to homologous DNA in R. meliloti and are found in the same order in both species. This suggests that there are similarities at the molecular level in motility and chemotaxis functions between R. meliloti and A. tumefaciens as well as in the patterns of chemotaxis and motility observed previously.

This study has succeeded in identifying, cloning and sequencing three genes previously unknown in R. sphaeroides, che W, che R and che Y2. These genes are homologous to genes recently discovered in Rhizobium meliloti and also show a lesser homology to genes in other organisms including the well characterised enteric chemotaxis genes. A comparative analysis of the deduced proteins has been made to determine structural and functional similarities between the R. sphaeroides genes and their homologues in R. meliloti and in E. coli. Considerable conservation of functionally important amino acid residues was revealed. In vivo complementation was done using the R. sphaeroides Che W protein to complement Che W^- strain of E. coli. The R. sphaeroides protein complemented the E. coli strain empirically proving similarity of function of the protein between these widely divergent genera. Homologous recombination using transposon-interrupted genes and internal gene fragments was attempted and met with limited success because R. sphaeroides proved permissive to the suicide vectors used. The most significant result of this study has been not in the similarities revealed between the chemotaxis systems of R. sphaeroides and the enterics, but in the differences discovered. It is now known that R. sphaeroides possesses two Che Y proteins, Che Y and Che Y2, and that whereas the R. sphaeroides Che Y2 protein is closely related to the E. coli Che Y protein, the R. sphaeroides Che Y protein shows considerable evolutionary divergence from it's enteric homologue. The significance of a second copy of Che Y protein suggests that these two response regulators act independently at the motor/switch complex and that they represent the final elements of two functionally distinct chemotactic sensory transduction pathways. This dual pathway system is not present in the enteric bacteria (a member of the γ-group of proteobacteria) but, we propose, may be present in a large group of environmentally important bacteria, the α-group of proteobacteria. Caulobacter, Agrobacterium, Rhizobium and Azospirillum species (α-group proteobacteria) all show behavioural similarities, and there are genetic clues to suggest that these organisms possess a dual chemotactic sensory transduction similar to the one we have found in R. sphaeroides. Apart from being fundamentally important, the dual sensory pathway hypothesis explains the wild-type behaviour of R. sphaeroides, as well as the difficulty in obtaining behavioural mutant phenotypes using methods developed during the investigation of E. coli.