Loyola University Chicago Loyola eCommons
Dissertations Theses and Dissertations
1997
Structure-Function Analysis of the Catalytic Domain of the Histidine Kinase Chea
Dolph David Ellefson Loyola University Chicago
Follow this and additional works at: https://ecommons.luc.edu/luc_diss
Part of the Microbiology Commons
Recommended Citation Ellefson, Dolph David, "Structure-Function Analysis of the Catalytic Domain of the Histidine Kinase Chea" (1997). Dissertations. 3425. https://ecommons.luc.edu/luc_diss/3425
This Dissertation is brought to you for free and open access by the Theses and Dissertations at Loyola eCommons. It has been accepted for inclusion in Dissertations by an authorized administrator of Loyola eCommons. For more information, please contact [email protected].
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License. Copyright © 1997 Dolph David Ellefson LOYOLA UNIVERSITY MEDICAL CENTER LIBRARY
LOYOLA UNIVERSITY OF CHICAGO
STRUCTURE-FUNCTION ANALYSIS OF THE
CATALYTIC DOMAIN OF THE HISTIDINE
KINASE CHEA
A DISSERTATION SUBMITTED TO
THE FACULTY OF THE GRADUATE SCHOOL
IN CANDIDACY FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
DEPARTMENT OF MICROBIOLOGY AND IMMUNOLOGY
BY
DOLPH DAVID ELLEFSON
CHICAGO, ILLINOIS
MAY, 1997 Copyright by Dolph David Ellefson, 1997 All Rights Reserved
ii ACKNOWLEDGEMENTS
I would like to thank my director, Dr. Alan J. Wolfe, for his support, advice,
and encouragment during the many years in his laboratory. In his laboratory, I was given a rare opportunity to explore a new arena of science and interact with a field of gifted researchers who I would not known otherwise.
I would also like to thank the members of my committee, Ors. Charles
Lange, Hans-Martin Jack, Tom Gallagher, and Rick Stewart for their valuable advice and criticism. I would also like to thank Dr. Katharine Knight for her support and her example of excellence in science. I would like to thank Dr.
Herb Mathews for his valuable advice and friendship while at Loyola.
Finally, I would like to thank my family. My sisters JoAnn, Laurie, and
Kara. My brothers Greg, Tim, Jim, and Sean. My father Woody, and my mother
Rose. I did not walk alone through all of this. Without their love and constant support, this would not have been possible. All I have done and all I have yet to do will be as much theirs as mine. Thank you all so much for letting me do this selfish thing.
iii TABLE OF CONTENTS
ACKNOWLEDGEMENTS .iii
LIST OF ILLUSTRATIONS • • • . . . . • . • . • • • • • • • • • • . • • . • • . • • . • • • • • • ix
LIST OF TABLES ...... X
Chapter
1. INTRODUCTION...... 1
2. LITERATURE REVIEW • • • • • • • • • • • • • • • • • • • • • . • • • • • • • • • • • • . 4
Sense and Response ...... 4
Chemotaxis and the Flagellum in Eschericia coli ...... 5
The Flagellar Complex. • • • • • • • • • • • . • • • • • • • • • • • • • • • • • • • 8
The basal body ...... 1 0
MotA and MotB ...... 11
The hook ...... 11
The filament ...... 12
The Chemotaxis Pathway. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 1 2
The Excitatory Response. . • • • • • . • • • • • • • • • • • • • • • • • . . . • . 1 3
iv The chemoreceptors ...... 1 3
CheW ...... 15
Che Y ...... 15
CheZ ...... 16
The Adaptation Response . . • . . . • . . . • . . • ...... 1 6
CheR ...... 18
CheB ...... 1 9
Functional Organization of CheA • ...... • • . . • . • ...... 20
Histidine Kinases ...... 24
3. MATERIALS AND METHODS ...... 28
Chemicals and Reagents • • • . • • . • • • ...... • . . • . • . . . 28
Bacterial Strains and Alleles • • . • • • • • • • . • . • . . . • . . . . • • • 29
Media and Growth Conditions...... 33
Swarm Assays ...... 33
Plasmid Construction . • • • . . • • . . • . • • • • • • • . . • • . • . • . . . • 34
Competence Procedure. . • . . . . . • • • • . • • . . . . . • . . • . • . . . . 37
Site-Directed Mutagenesis . . • . • . • • . . . . • ...... • . . . . . 37
Homologous Recombination • . • ...... • ...... 40
Purification of His-tagged proteins • . . . • . • . . • ...... 41
V In Vitro Phosphorylation of Proteins, SOS-PAGE, Autoradiography and lmmunoblot Analysis . . . . • • • • • . • • 43
Sequencing ...... 43
Primer End-Labeling ...... 44
Genomic DNA Preparation • • • • • • • • • • • • • • • • • • • • • • • • • • • 45
Western Analysis ...... 46
Generation of CheAL-Specific and CheA antisera • • • • • • 4 7
ATP Binding Assay ...... 49
4. RESULTS ...... 51
In Vivo Analysis of CheA Putative catalytic domain Mutants ...... 51
Rationale for design of catalytic mutants • • • • • • • • • • • 51
Mutant Alleles Defective in the Putative Catalytic Domain do not Support Chemotaxisln Vivo • ...... 53
The Inability to Perform Chemotaxis is not due to a Metabolic Defect ...... 55
CheAs, but not CheAL48HQ Restores of Chemotactic Ability to Kinase-Deficient CheAL Mutants • • • • • • • • • • • • 58
Dominance of Various CheA Catalytic Mutants • • • . . . . . 61
In Vitro Autophosphorylation Analysis of CheA Variants Defective in G 1, F or G2. • • • • . • • • • • • • • • • • • • • • • • • • • . . . 64
vi ATP Binding ...... 68
The Non-Phosphorylatable Variants CheAs or CheAL48HQ can Mediate Transphosphorylation of Kinase-Deficient CheAL Mutants In Vitro...... 69
Structure Averaging Algorithms • • • • • . • . • • • • • • • . • • • • • • • • 73
CheA237-519 ...... 77
5. DISCUSSION AND SUMMARY •••••••••••••••••••••••••• 79
Mutagenesis strategy: The CheA catalytic domain • • • • • • • 79
Chemotaxis Requires Intact G1, F, and G2 Subdomains • • 81
Autophosphorylation Requires Intact G1, F, and G2 Subdomains ...... 82
G1 Residue G422 Plays an Intimate Role in Binding of ATP: G2 Residues G470, G472, and G474 do not • • • • • • •• 83
G2 Residues G470, G472, and/or G474 Play a Role in either the Docking of P1 and/or the Subsequent Phosphotransfer; G 1 Residue G422 and F Residues F455 and F459 do not • • • • • • • • • • • • • • • • • • • • • • • • . • • • • • • • 85
Clustering of Ternary Complexes Requires an Intact G2 Subdomain, but not a G1 Subdomain • • • • • • • • • • • • • • • • • • 86
G 1 and G2 Function in cis . . • • ...... • ...... 88
Secondary Structure Prediction and Model for the Structure of the Catalytic Domain of the Histidine Kinase CheA...... •• • . • 90
vii Significance of other Histidine Kinases that Lack Some of the Conserved Regions • • • • • • • • • • . . . • • • • . • • • • 92
WORKS CITED ...... 95
VITA ...... 113
viii LIST OF ILLUSTRATIONS
Figure Page
1. Th e " run " an d ''t um bl e " re s p o n se ••.•...•...... •...... •...••. 7
2. The flagellar complex ...... 9
3. Circuitry of the chemotactic signal transduction pathway .••••...• 14
4. Transfer of phosphate (P*) from the sensor CheA to the response regulator Che Y...... 17
5. The domain organization of the two forms of CheA. • • • • • • • • • • . . . 22
6. Amino acid alignment of the N, G1, F, and G2 region of CheA homo logs...... 25
7. Comparison of the N, G1, F, and G2 consensus sequences between CheA and orthodox TR domains...... • • . • . • • . • . . • . . 26
8. Functional organization of CheA and schematic of cheA alleles used in this study...... 32
9. Photographs of representative swarms. . • • • . . • . . • • . • • . . • . • . • • • 54
10. Photographs of representative swarm patterns. • • • • • • • • • • • • • • • . 56
11. Growthcurv-es ...... 57
12. Photographas of representative swarm patterns demonstrating that CheAs restores chemotaxis to cells that express the kinase- deficient mutants CheAL G 1 or CheAL G2 but not CheAL G 1/G2. • • 60
ix 13. Photograph of swarm plates demonstrating that allele cheAG2 confers a dominant-negative phenotype in a wild-type background ...... •...... •...... 62
14. Displacement of the outermost serine band plotted as a function of time demonstrating the effect of overexpressing plasmid-borne CheA variants transformed into wild-type cells •••••••••••••••••• 63
15. Purified variant CheA proteins •••••••••••••••••••••••••••••••• 66
16. In vitro autophosphorylation of wild-type and mutant CheA proteins ...... 67
17. In vitro assays demonstrating that the non-phosphorylatable proteins CheAs and CheAL 48HQ can mediate transphosphorylation of kinase deficient mutants of CheAL •••••• 71
18. Phosphorylation of CheAL variants ••••••••••••••••••••••••••• 74
19. lmmunoblot analysis of using antisera specific for CheAL and
CheA8 or CheAL alone ...... 75
20. Computer algorithm analysis of secondary structure of the transmitter domain of CheA ••••••••••••••••••••••••••••••••• 76
21. Model for the structure of the G1-G2 region of the transmitter domain of CheA ...... 93
X LIST OF TABLES
Table Page
1. Ligands of the four known methyl-accepting chemotaxis proteins...... 16
2. Bacterial strains used in this study. . • • . . • • • • . . . . . • • • . . • • • • • • • • 30
3. Mutagenesis and sequencing primers...... 31
4. Plasmids used in this study. . . • • • • • • . . . . • • • ...... • • . • • . • • . • • • 36
5. TNP-ADP binding affinities of CheAL proteins...... 70
xi INTRODUCTION
This dissertation project sought to determine the relationship between function and structure of the transmitter domain of the chemotaxis-associated histidine autokinase CheA of Eschericia coli. For some time, the transmitter domain of CheA has been proposed as the site of ATP binding and hydrolysis on the basis of homology to eukaryotic serine/threonine kinases. Here, I have tested this hypothesis, establishing that the transmitter region of CheA does act as the catalytic site.
First, I used site-directed mutagenesis to introduce conservative amino acid changes in the highly conserved regions of the putative catalytic domain of
CheA. I then examined their ability to mediate chemotaxis and in the process demonstrated that changes in the most conserved residues of this domain completely eliminated chemotaxis.
Second, I purified these mutant CheA proteins and tested their ability to autophosphorylate (a measure of their capacity to bind and hydrolyze ATP).
None of the CheA mutants tested autophosphorylated despite prolonged exposure to ATP. In collaboration with Rick Stewart (University of Maryland), we next determined that a CheA protein mutated in the first of two glycine-rich
1 2 regions bound the ADP analog TNP-ADP poorly ie., with about a 10-fold decrease in affinity relative to that of the wild-type protein. In contrast, a CheA protein mutated in the second glycine-rich regions bound TNP-ADP about as well as did the wild-type CheA protein. Although point mutations within this second glycine-rich region exerted no effect upon its ability to bind TNP-ADP, eliminating the entire region caused about a 3-fold decrease in affinity. A protein mutated in both glycine-rich regions bound TNP-ADP about 2 times less efficiently than did a protein mutated in the first glycine-rich region.Thus, although this region may not play a direct role in binding TNP-ADP, it does influence binding by the first glycine-rich region.
Third, I examined each of the mutants for their ability to be complemented
, with an intact putative catalytic domain supplied in trans by CheA8 a natural variant of CheA that lacks the site of autophosphorylation. In vitro, CheA 8 complemented all of the CheA mutants tested. In vivo, CheA complemented 8
CheA proteins mutated in either of the glycine-rich regions, but not those defective in both of the glycine-rich regions or deleted for the second one.
Since transphosphorylation in vitro of these latter two mutant proteins was significantly reduced relative to that of mutants defective for one glycine-rich region or the other, these results support the hypothesis that CheA autophosphorylation must reach a certain threshold level before chemotaxis can occur in vivo. CheA proteins mutated in G 1 or G2 cannot complement each other in vivo or transphosphorylate each other in vitro suggesting that the G 1- 3
F-G2 regions in a CheA dimer function independently and in cis within each
protomer.
Fourth, I examined the putative catalytic domain of CheA with computer algorithms designed to predict secondary structure. On the basis of these structure-averaging algorithms, I propose that the catalytic region might best be
represented as a cleft in which the F subdomain lies at the base and acts as a flexible turn or hinge flanked by two short alpha helices. This arrangement would serve to approximate G1 and G2, predicted to be unstructured
ATP-binding loops. The binding of ligand by the chemoreceptor would initiate a cascade through the ternary complex with the end result being the movement of the G 1 and G2 loops apart, rendering CheA in an "open" or inactive conformation (Fig 20). To test this hypothesis, NMR analysis and x-ray crystallographic analysis is required. CHAPTER2
LITERATURE REVIEW
Sense and Response
All of life is connected by common needs. Among life's enormous divergence, there exist fundamental systems and behaviors which bind the simplest to the most complex. For instance, to survive and prosper, all cells, from the simplest form of bacteria to the most complicated eukaryote, must perform two tasks very well. First, they must sense their environment. Second, they must be able to respond to what they sense.
A eukaryotic cell senses its environment primarily through the use of proteins located on its surface and within its interior. Messages from the exterior of the cell are conveyed to the interior by physical interaction of the cell surface proteins with cytoplasmic proteins and with other cell surface proteins.
Although simple in comparison to eukaryotic cells, bacteria sense their environment in much the same way.
The life of a bacterial cell is a precarious one. It must live in an environment which is often hostile and dangerous, requiring that it respond rapidly to unexpected extremes of nutrient supply, temperature, pH, and acidity.
4 5
Bacteria are particularly adept at adapting to extremes. In fact, bacteria have been so successful at adaptation, that they survive in a wide range of environments that would be much too extreme for most eukaryotic cells. For example, bacterial cells have been found thriving in the cooling tanks of nuclear reactors and in super hot volcanic vents. To cope with such extremes, bacteria have developed a wide array of specialized systems for sensing, interpreting and responding to changes in their extracellular environment. They have evolved a range of complex responses to environmental stresses. One such response occurs when an E. coli cell encounters an increase in osmolarity of the surrounding environment. In response, an inner membrane protein called
EnvZ autophosphorylates before passing that phosphate to a cytoplasmic protein (OmpR). In its phosphorylated state, OmpR binds DNA, increasing transcription of one outer membrane porin and decreasing that of another. This response allows the cell to optimize its internal salt concentration (Alphen and
Lugtengerg, 1977; Forst and Inouye, 1988). Phosphorylated OmpR also influences the expression of the entire flagellar regulon (Shin and Park, 1995).
The response could also be far more complex, such as the formation of spores by Bacillus subtilis (Burbulys et al., 1991 ), fruiting body formation in microorganisms such as Myxococcus xanthus (Zusman and McBride, 1991), or the chemotactic response by Eschericia coli (Mutoh and Simon, 1986).
Chemotaxis and the Flagellum in Eschericia coli
Chemotaxis is the process whereby a bacterium senses the presence of attractants or repellents in its environment and responds by altering its 6 swimming behavior. Some cells sense and respond to the concentration of attractant in a highly directional manner, basing its direction of movement upon where the stimulus occurred on the surface of the cell. Lymphocytes of the mammalian immune system move in this fashion, altering their direction of movement toward the point of the stimulus in the cellular membrane. This type of "spatial" sensing relies upon the cells ability to calculate the difference in the concentration of attractant at the point of the stimulus from the concentration of attractant at a point opposite the stimulus and modify its direction accordingly.
A bacterium such as E. coli , however, senses the concentration of attractants continuously in a temporal rather than spatial manner (Macnab and Koshland,
1972). An E. coli cell must sense temporally both because of its small size and because it can sense changes in ligand concentration even when the concentration of ligand is spatially homogeneous (Berg and Purcell, 1977;
Macnab and Koshland, 1972). This type of sensing is non-directional: it does not matter if the attractant is being sensed at one end of the cell versus the other. If attractants were sensed spatially by E. coli, the origin of the stimulus would always point in the direction of swimming.
E. coli swims in a fluid environment by the use of propellor-like structures called flagella. At the base of each flagellum is a motor that rotates alternately counterclockwise (CCW) and clockwise (CW). Because of the shape and periodicity of its flagella, E. coli is able to bundle all of its flagella into a single helical unit which works in concert to propel the cell forward in a smooth gently 7
ccw RUN
cw TUVIBLE
( I
Fig 1. The "run" and "tumble" response. Illustration of the relationship between direction of flagellar rotation and swimming behaviors displayed by cells of E. coli. Counterclockwise (CCW) rotation corresponds to runs; clockwise (CW) _rotation corresponds to tumbles. 8
curving path called a "run". This behavior occurs when the motors rotate CCW.
The cell changes direction by simply reversing some of its motors, causing them to rotate in the CW direction. This action causes the flagellar bundle to jam, the flagella fly apart and, working at cross-purposes, the cell executes a flailing behavior called a ''tumble" until the motors re-assume the CCW direction of rotation causing the bundle to reform (Fig. 1).
How does an E coli cell use seemingly random tumbles to swim toward a better life? E. coli is an "optimist" and is constantly in search of a "better life"
(Berg 1983). When swimming in a liquid environment, et is constantly sensing and calculating the concentration of attractants and repellents it encounters
(Berg and Purcell 1977). When the cell swims up a gradient of attractant, it suppresses the next tumble. By contrast, if it swims down that gradient, the next tumble occurs. Thus, runs in favorable directions are twice as long as those in unfavorable directions. The Flagellar Complex
E. coli is an enteric organism of the human gut, where it exists in a commensal relationship with its host. The digestive tract of humans is limited in its ability to break down many nutrients. Bacteria serve as an aid to promote conversion of many indigestible nutrients into forms the host can metabolize.
Chemotaxis is an expensive undertaking in the context of energy consumption.
The energy expenditure is, however, justifiable under a variety of conditions. In the environment external to the host, the flagellar complex may be needed for 9
Fig 2. The flagellar complex. (Adapted from Macnab, 1996) 10
the purpose of finding nutrients and escaping toxins. But, in the nutrient-rich environment of the gut, a flagellated cell may act as a virulence factor, facilitating attachment and entry into preferred environments. When it becomes necessary to produce a flagellum, E. coli initiates, through a number of operons, the coordinate synthesis of all of the components necessary for its construction. The flagellum is a complex structure that can be divided into three main structural components; the extended basal body, the hook, and the filament. Expression of the flagellar genes is performed in a hierarchical fashion (Kutsukake and lino, 1994., Macnab, 1992). The actual assembly of the flagellum is conducted sequentially with the basal body being constructed first, followed by the hook, and then the filament (Fig 2).
The basal body : The basal body traverses the inner and outer membranes and is anchored to the rigid cell wall. Elegant isolation techniques and electron microscopic studies have revealed a structurally complex organelle composed of a number of distinct structural features. The extended basal body is the portion of the organelle lying on the inside of the inner membrane and is called the "C" ring. This ring, composed of three proteins called FliG, FliM, and FliN, is the switching apparatus responsible for controlling rotational direction of the motor. This switch in direction is made in response to interaction with the phosphorylated chemotaxis protein CheY, described below (Francis et al., 1994).
The parts of the flagellar motor that comprise the actual turning apparatus of the flagellar motor is composed of a number of interacting structural 1 1 components Above the C ring and traversing the inner membrane, cell wall, and outer membrane are the MS ring (FliF), the proximal rod (FliB, FlgC, and
FlgF), the P and L rings (Flgl and FlgH), and the distal rod. The MS ring and the
Rod are spinning regions of the motor, while the L and P rings are thought to act as a sort of flange, anchoring the basal body to the cell wall.
MotA and MotB: The flagellar motor is driven by proton motive force.
The protons needed to rotate the motor are supplied via a proton channel composed of two proteins called MotA and MotB (Zhou et al., 1995). Mutants in the Mot proteins are characterized as "paralyzed", meaning the motor does not spin. This suggests that these proteins are directly involved in conduction of protons across the membrane to drive the motor (Blair and Berg, 1990; Block and Berg, 1984). Although not an integral part of the basal body structure, MotA and MotB surround the basal body. They act as the stator for the flagellar motor and are essential for motility. Both proteins span the entire inner membrane.
Whereas MotA contains a substantial cytoplasmic interface, MotB contains a peptidoglycan binding domain (Dean et al., 1984). The precise mechanism of proton movement through the channel is not known. However, several residues critical to channeling protons through MotA and MotB have been identified
(Sharp et al., 1995; De Mot and Vanderleyden, 1994).
The hook: The hook (FlgE) assembly is a highly flexible link between the basal body and the filament. Its function is similar to that of a U-joint on a car. The hook both allows the filament to rotate on a different plane than the basal body motor and allows several flagella to bundle into a uniform complex. 12
The helical wavelength of the hook is different from that of the filament despite having the same nine fibril structure (Morgan et al., 1993). This allows flexibility not possible with the more rigid filament structure (Block et al., 1989; Block et al., 1991 ). Without this important component, all of the flagellar components would have to be located within a small region to enable the flagella to form a bundle.
The filament: The filament is the structural assembly which acts like a propeller, giving the cell the ability to swim through a fluid environment. This structure is composed entirely of a single protein subunit called flagellin (FliC)
(Kuwajima et al., 1986). When observed at high magnification, the filament can be seen to possess a periodicity (a sine wave pattern) of about 0.5 µm (Kamiya et al., 1982). When the motor is rotating in its default CCW direction, the flagella group together in a single bundle and allow the cell to swim in a forward direction.
The Chemotaxis Pathway
Chemotaxis in E coli and other microorganisms is dependent upon the ability of chemoreceptors to communicate with switch components of flagellar motors to modulate swimming behavior in response to the cells' chemical environment. This communication requires the cooperative effort of the cytoplasmic gene products of six genes, cheW, cheR, cheB, cheY, cheZ and cheA (Mutoh and Simon 1986; Stock and Stock, 1987). Additionally, E coli encodes 5 transmembrane chemoreceptors also known as transducers or methyl-accepting chemotaxis proteins (MCP's). They are Tar (Clark and 13
Koshland, 1979), Tsr (Hedblom and Adler, 1980), Tap (Manson et al., 1986),
Trg (Adler et al., 1973), and the recently described Aer (aerotaxis) (Bibikov and
Parkinson, Johnson and Taylor, personal communication). Tar and Tsr, the two most extensively studied chemoreceptors, sense the amino acids aspartate and serine, respectively.
The Excitatory Response
Chemotaxis is typically discussed in terms of two response mechanisms, excitation and adaptation. When cells determine that the concentration of attractant has increased or the concentration of repellent has decreased, they respond by increasing the length of their runs. This behavior is called the excitatory response. Even though the stimulus might persist, the excitatory response lasts only seconds before the cell returns to its baseline behavior.
The return to baseline behavior occurs because of the adaptation response.
Both excitation and adaptation occur in response to signals generated by an inner membrane-bound ternary complex composed of an MCP dimer, a
CheA dimer, and two CheW monomers (Fig 3). This ternary complex relays information concerning ligand concentration from the outside to the inside of the cell.
The chemoreceptors: The MCPs are located within the inner membrane. They are composed of a highly variable periplasmic domain, a transmembrane region, and a largely conserved cytoplasmic domain. The degree of conservation among these-different domains speaks to their respective functions. The periplasmic domains are variable since they bind 14
Fig 3. Orcuitry of the chemotactc signal transduction pathway. Chemoreceptors (MCP) located in the inner membrane transduce signals from the periplagn to OleAL (AL) or CheAs (As) through the roupling protein CheW ~). CheAautopho~horylates in response to a decrease in concentration of ligand (Table 1), and, in turn, pho~horylates the response regulator CheY (Y). Alospho-CheY then interacts with tte flag3llar motor to change the drection of 10tation from CCW to CW. The phmphatase CheZ (Z) dephosphorylates CheY. The chemoreceptor adapts to lgand roncentration via receptor methylation/demethylation. The93 reactions involve the meth~transferase CheR (R) and the meth~esterase CheB (8). 15 different ligands (Table 1). The conservation of the cytoplasmic domain is necessary because the same internal signal is initiated in response to different stimuli. In addition, there are two transmembrane regions termed TM1 and TM2 which flank a central, periplasmic ligand binding domain. TM2 is followed by a linker region which is connected to signaling domain by a short, helical methylated segment called MH1. The signaling domain is, in turn, followed by another methylated region called MH2. MH1 and MH2 serve as the methyl-accepting regions involved in the adaptation response described below.
The cytoplasmic domain is also known as the signalling domain (Ames and
Parkinson, 1994). It is thought that this region transfers the ligand signal from the periplasmic domain into a coupling protein called CheW which then transfers this signal into the histidine kinase CheA.
CheW: CheW (14 kDa) couples the signal from the transducer to the histidine kinase CheA (described in detail below). While the nature of CheW's
MCP-CheA coupling remains elusive, the sites of interaction with CheA and Tsr have been well characterized (Bourret et al., 1993). Analysis of second-site repressors of mutations in the chemoreceptor suggest that the signaling region of the chemoreceptor is the site of interaction with CheW ( Liu and Parkinson,
1991 ).
CheY: Signal transduction through the chemotaxis pathway is accomplished via the phosphorylation cascade CheA-P + CheY-> CheA +
CheY-P (Fig 4). CheY (14 kDa), the response regulator of the CheA-CheY 16
Table 1.--Ligands of the four known methyl-accepting chemotaxis proteins. MCP Attractant/Repellent
Tsr Serine, Alanine, Glycerol/Leucine
Tar Aspartate, Glutamate/ Ni2+,Co2+
Tap Di peptides/Phenol
Trg Galactose, Glucose
Aer Oxygen
two-component chemotaxis signal transduction pathway, is a phospho-protein which, in its phosphorylated form, interacts with the flagellar protein FliM, eliciting a change in direction of the flagellar motor.
CheZ: CheY is an autophosphatase, with a phospho-aspartate half-life of only a few seconds. Despite this instability, the rate of spontaneous dephosphorylation roughly equals that of the rate of phosphorylation. Clearly,
CheY needs a phosphatase to increase the rate of CheY-P degradation. CheZ
(24 kDa) dramatically enhances CheY dephosphorylation. It binds CheY and presumably forces a conformational change in CheY that facilitates its transition to a dephosphoryated state (Blat and Eisenbach, 1994; Lukat et al., 1991 ).
The Adaptation Response
The adaptation response functions to maintain the MCPs in a constant state of readiness so that they may respond to slight alterations in the 17
p•~oo+ ~ o-o-c=]][[JOO + ~p•-~ ~ CheV Che A CheY-P*
Fig 4. Transfer of phosphate (P*) from the sensor CheA to the response regulator CheY. For details concerning the domain organization of CheA, see Fig 5. 18
concentration of attractant or repellent. For instance, when an E. coli cell is swimming, the surrounding medium contains fluctuations in the concentration of attractant. In order to swim in a direction in which life is "better'', E. coli must be able to sense the presence of a gradiant of the attractant. By fluctuating between a methylated and unmethylated state, the chemoreceptor "adapts" to the homogeneity of attractant or repellant in the surrounding medium. In this way, the chemoreceptor "fine tunes" itself and filters out extraneous noise caused by homogeneous concentrations of attractant or repellant. The MCP adapts to attractant or repellant concentration by the action of two proteins, a methyltransferase (CheR) and a methylesterase (CheB).
CheR: CheR (30kDa) is a methyltransferase which functions in the chemotactic adaptive response. Its purpose is to methylate glutamate residues located in the MH1 and MH2 regions of the cytoplasmic side of the chemoreceptors. CheR accomplishes this by transfer of a methyl group from the methyl donor AdoMet to the glutamate residues. The degree of strength of signal through the chemoreceptor correlates with the degree of methylation at the glutamate residues. Increases in the amount of methylation causes an increase in rate of CheA autophosphorylation within the ternary complex
(Borczuk et al., 1986; Ninfa et al., 1991). Methylation seems to help regulate the signaling domain of the chemoreceptor by conformational changes in the
MH1/MH2 regions. The region containing the glutamate residues has been elegantly shown to be a coiled structure whose torsional state has a tremendous effect upon the ability of the transducer to transfer a signal to CheA 19
(Cochran and Kim, 1996). In other words, as CheR-mediated methylation proceeds from zero residues to one residue and so on, the magnitude of signal amplification through the transducer increases.
CheB: CheB (37kDa) is a methylesterase which also functions within the chemotactic adaptation response. Its purpose is to de-methylate/de-amidate the glutamate residues and counter methyltransferase-mediated signal enhancement by CheR. In addition to CheY and the periplasmic domain of Tar,
CheB is the only other chemotaxis protein for which crystallographic data exists.
In addition to CheY, CheB is the only other known protein to be phosphorylated by CheA. The donation of a phosphate to CheB results in activation of this protein. An increase in autophosphorylation of CheA leads to an increase in phosphorylation of CheB and subsequent demethylation of the chemoreceptors.
Structurally, CheB is a monomer with regulatory and catalytic domains separated by a linker. The regulatory domain of CheB contains the phospho-acceptor site and is homologous to the response regulator CheY
(Lupas and Stock, 1989; Stock et al., Cold Spring Harbor Symposium 1988).
Phosphorylation of this domain increases the methylesterase activity of CheB over 10 fold (Stock et al., 1985). The catalytic domain of CheB has been revealed to be a wound a/B structure with seven a/B strands (West et al., 1995)
The adaptation response in chemotaxis functions to ''fine tune" the ternary complex, dampening the effects of extremes of ligand availability. High levels of methylation activate CheA and low levels inactivate CheA. High levels of 20 methylation decrease the affinity of the chemoreceptor for its ligand (Yonekawa and Hayashi, 1986). So, in low levels of ligand concentration, CheR mediates methylation of the chemoreceptor which brings about activation of CheA and decreased affinity for ligand. Activated CheA phosphorylates CheB, which, in turn, demethylates the chemoreceptor, bringing about an increased affinity for ligand. Extreme levels of demethylation causes a decrease in CheA activity, again decreasing ligand affinity and shutting off the ternary complex. A middle ground, in which the chemoreceptor is only partially methylated is the ideal balance sought by the adaptation system. The ternary complex of
CheA/CheW/chemoreceptor exists in a flux between an "on" and an "off' position. The methylation/demethylation system serves to stabilize the ternary complex into a state which can respond appropriately to subtle changes in ligand concentration (Stock and Surrette, 1994).
Functional Organization of CheA
Of the six chemotaxis genes, cheA, the primary focus of this study, is the most central because it serves as a phospho-donor to both Che Y, the excitatory signal, and CheB, the adaptive signal. The cheA gene encodes two proteins,
CheAL (78 kDa) and CheA (69 kDa), translated in-frame from two different 8 initiation sites (Parkinson and Kofoid, 1992; Smith and Parkinson, 1980) (Fig 5).
Both proteins are histidine kinases of the two-component family (Stock et al.,
1985).
Like other two-component histidine kinases, CheA functions as a dimer
(Gegner and Dahlquist 1991 ). The two-component histidine kinase CheAL 21 autophosphorylates a conserved histidine residue (His-48) located near its amino-terminus (Hess et al., 1988). Phosphorylated-CheAL' in turn, acts as a phospho-donor for two other chemotaxis proteins: CheY, required for clockwise signal generation, and CheB, required for adaptation (reviewed in Bourrett et al., 1991; Stock et al., 1991; Stock and Surette, 1994). CheAL homodimers form ternary complexes with CheW and chemoreceptors, e.g., Tar (Gegner et al.,
1992). Within these complexes, the autokinase activity of CheAL is significantly accelerated relative to that of CheAL homodimers alone. Binding of ligand to the transducer protein, e.g., the Tar-specific attractant L-aspartate, greatly diminishes CheAL autokinase activity (Borkovich et al., 1989; Borkovich and Simon 1991; Ninfa et al., 1991 ). Thus, CheW and chemoreceptors modulate CheAL autokinase activity in a ligand-dependent manner.
Although CheA8 lacks the site of autophosphorylation, it retains domains that in CheAL are required for kinase activity, ternary complex formation and phospho-transfer to CheY and CheB (Bourret et al., 1993; Hess et al., 1988;
Kofoid and Parkinson 1988; Morrison and Parkinson 1994; Swanson et al.,
1993). Evidence, obtained in vivo and in vitro, indicates that the CheAL domains responsible for kinase activity and ternary complex formation also
. function in CheA8 In vitro, CheA8 mediates receptor-modulated transphosphorylation of mutant CheAL proteins either deficient for kinase 22
CheA._ N- C
Che,¾, C
Fig 5. The domain organization of the two forms of CheA. Both forms of CheA are expressed in-frame from alternate translational start sites. P1 contains the site of autophosphorylation (H), P2 is the CheY binding domain, TL is the site of CheA dimerization, TR contains the putative catalytic subunits, M and C couple CheA to CheW and chemoreceptor. The rectangles in TR represent highly conserved sequences. 23
activity or truncated for the carboxy-terminal segments required for ternary complex formation (Swanson et al., 1993; Wolfe et al., 1994; Wolfe and Stewart
1993). In vivo, CheA restores chemotactic ability to cells that express either 8 kinase-deficient or truncated CheAL mutant proteins (Wolfe et al., 1994; Wolfe and Stewart 1993).
CheAL and CheA8 are composed of several distinct functional domains
(reviewed in Kofoid and Parkinson 1988; Fig 5). Two of these domains, P1 and
P2, located at the amino-terminus of the protein, are separated by a stretch of amino acids that resemble charge-rich linkers found in other regulatory and sensory transduction proteins (Parkinson and Kofoid 1992; Wooten and
Drummond, 1989.). P1 contains the site of autophosphorylation, His-48 (Hess et al., 1988). This domain, present in CheAL but not in CheA , is required for 8 phospho-transfer to CheY and CheB (Swanson et al., 1993). P2 likely assists in the interaction between the phospho-donor site in P1 and amino acid sequences located in CheY (Morrison and Parkinson 1994; Swanson et al.,
1993). Two other domains, M and C, located at the carboxy-terminus, appear to play important roles in enabling CheAL and/or CheA to receive sensory 8 information from the chemoreceptors (Bourret et al., 1993; Parkinson and
Kofoid, 1992).
Histidine Kinases
Over 100 histidine kinases have been identified to date (Russo and 24
Silhavy, 1993; Parkinson and Kofoid, 1992; and Stock et al., 1989). The histidine kinases are linked into a family by virtue of 5 highly conserved subdomains termed H, N, G1 (D), F, and G2(G). CheA is atypical of most histidine kinases in that the H subdomain is located in P1, far removed from the other conserved subdomains. The N, G1, F, and G2 subdomains, located in the
TR region of the transmitter domain of CheA are characterized by the consensus sequences illustrated in Fig 6.
The G 1 and G2 subdomains (Fig 6) resemble the glycine-rich regions of the nucleotide regions of other nucleotide binding domains (Saraste et al.,
1990; Taylor et al., 1992; Schulz 1992; Muller 1993). These subdomains have previously been proposed to be the site of Mg-ATP binding and catalysis
(Parkinson and Kofoid, 1992). Mutations affecting the ability of CheA to autophosphorylate have been mapped to this region, suggesting a role for this region in ATP hydrolysis (Oosawa et al., 1988; Swanson et al., 1994; Wolfe and
Stewart, 1993). The G1 region is of particular interest because the first D and first G of the DDGAG motif are thought to form a Mg-binding backbone which would enable proper positioning of of a Mg-ATP complex (Taylor et al., 1992).
To date, no detailed studies have been performed to elicit the function(s) of the highly conserved residues contained within the putative catalytic site of CheA. By analysis of mutants defective in selected amino acids within this region of CheA, this study sought to address the following questions concerning the function of the conserved subdomains of the CheA transmitter 25
N E. C. 370 iidpLtHlvR NslDHGiElp ekRlaaGKns vgnlilsaeh s. t. 387 iidpLtHlvR NslDHGiEmp ekRleaGKnv vgnlilsaeh E.a...... R.m. 467 ladpLtHliR NavDHGlEtp ekRvaaGKnp egtvrltakh R. s. 399 ladpLtHmiR NavDHGlEpa drRlhlGKpp vgliyltaah B. s. 399 igdpLvHliR NsiDHGlEap etRlgkGKpe sgkvvlkayh L. m. 346 igdpLvHliR NsvDHGaEtv evRrknGKne tatinlkafh H.s. 399 isdpLrnHllR NavDHGiEkp avRednGKDar egtitlsaer M.x. 367 vrdaLvHllR NsvDHGvEsp dtRgglGKpl ngririrvrv Consensus X-OZL-HXXR NZXDHGXE-Z O-R---GK-- -z-x-x-z-J Gl E.c. 411 gggnicievt DDGaGlnrer ilak.Aasggl . tvse .. nrns s.t. 428 gggnicievt DDGaGlnrer ilak.Amsggm .avne .. nrnt E.a...... •••..... ek ilak.Aaaggl .avtd .. trns R.m. 408 rsqriviela DDGaGinrek vrgk.Aidndl iaada .. nls R.s. 440 rsgrvlieik DDGaGinrpr vleiAggkgl vaqda .. qlt B.s. 440 sgnhvfieve DDGaGinr.k ilekpler.v itekeattle L.m. 387 sgnnvvieia DDGaGinkrk vlek.Aiaknv vtraestkrnt H.s. 440 drdhvligvr DDGaGidhdt rnrek.Aiekgv ktreevqdrnp M.x. 408 dgdrnlhieve DDGrGidper lrqaAiskrl inavqaaals Consensus -Z--X-IOX- D~ZGXOJOJ XX-JA----X -z--o---xz F G2 E.c. 452 ddevarnLiFa pGFSTaeqv tdvSGRGVGMD WkrniqkrnG s.t. 469 ddevgmLiFa pGFSTaeqv tdvSGRGVGMD WkrniqernG E.a. deevgmliFa pGFSTaeqv tdvSGRGVGMD WkrniqernG R.m. 449 deevdnLiFh aGFSTadki sdiSGRGVGMD WkrsiqalG R.s. 481 eeeidgLlFrn pGFSTasvv sdlSGRGVGMD WksaieslG B.s. 481 dnqiyeLiFa pGFSTadqi sdiSGRGVGMD WknkleslG L.m. 428 daeifdLl.d sGFSTadqv sdlSGRGVGlD WrntilkiG H. s. 481 dddvedLvFh pGFSTndev tdvSGRGVGrnD Wrdtvtrld M.x. 449 ereaieLiFr pGFSTrdqv selSGRGVGrnD WkrkvetlG Consensus OOOX--LXF- ZGFSTZOOX ZOXSGRGVGXD WJ--XO-XG
Fig 6. Amino acid alignment of the N, G1, F, and G2 region of CheA homologs. Abbreviations are:Eschericia coli (E.c., Kofoid and Parkinson, 1991), Salmonella typhimurium (S.t., Stock et al., 1988), Enterobacter aerogenes ( E.a., Dahl et al., 1989), Rhizobium meli/oti (R.m., Greek et al., 1995), Rhodobacter sphaeroides (R.s., Ward et al., 1995), Bacillus subtilus (B.s., Fuhrer and Ordal, 1991 ), Listeria monocytogenes (L.m., Dons, et al., 1994), Halicobacter salinarium (H.s., Rudolph and Oesterhelt, 1995) and Myxococcus xanthis (M.x., McCleary and Zusman, 1990). Numbers at left respresent amino acid residues. Strictly conserved residues are capitalized. X refers to residues at which 70% of the CheA family have a nonpolar amino acid (I,M,L,or V). Z refers to residues at which at least 70% of the CheA family have a polar amino acid (A,G,P,S, or T). J refers to residues at which at least 70% of the CheA family have a basic amino acid (R, K, or H). 0 refers to residues at which at least 70% of the CheA family have an acidic or amidic amino acid (D,E,Q, or N). Wild-card positions with less than 70% conservation among the kinase family are indicated with dashes. The N, G1, F, and G2 subdomains are underlined and strictly conserved residues are in bold. 26
N Gl F G2
CheA consensus X-HXXRNZXDHZ X-DDGZGXO XF-ZGF GJGXGX Orthodox kinase X-QXXXNXX-NZ X-DDGZGXZ XFOPF GZGXGX
Fig 7. Comparison of the N, G1, F, and G2 consensus sequences between CheA and orthodox TR domains X refers to residues at which 70% of the CheA family have a nonpolar amino acid {l,M,L,or V). Z refers to residues at which at least 70% of the CheA family have a polar amino acid (A,G,P,S, or T). J refers to residues at which at least 70% of the CheA family have a basic amino acid (R, K, or H). 0 refers to residues at which at least 70% of the CheA family have an acidic or amidic amino acid (D,E,Q, or N). Wild-card positions with less than 70% conservation among the kinase family are indicated with dashes. The strictly conserved residues are in bold. 27
domain:
•First, do the subdomains play a role in chemotaxis in vivo?
•Second, do the subdomains play a role in the ability of CheA to autophosphorylate?
•Third, are the subdomains involved in formation and maintainance of the ternary complex?
•Fourth, are the subdomains involved in the binding of ATP?
•Fifth, within a CheA dimer, do the conserved subdomains operate independently and in cis, or interdependently and In trans ?
•Sixth, what can we learn about the structure of the putative catalytic
region of CheA? CHAPTER3
MATERIALS AND METHODS
Chemicals and Reagents
Bacto-tryptone and bacto-yeast extract were obtained from Difeo (Detroit,
Ml). The restriction enzymes Ndel, BamHI, Munl, Saeli, Xmal, BssHII and M/ul, the Wizard™ minpreps kit , the Altered Sites TM in vitro Mutagenesis System, and the Wizard™ M13 DNA Purification System were obtained from Promega
(Madison, WI). BstBI, agarose, glycine, DNA ligase, DNA polymerase, T4 polynucleotide kinase, proteinase K and 5x forward reaction buffer were obtained from Gibco-BRL (Gaithersburg, MD). RNase-A, alkaline phosphatase goat anti-rabbit lg, Tris-HCI, isopropyl-B-D-thiogalactopyranoside (IPTG), nitro blue tetrazolium (NBT), Na-ampicillin, tetracycline, chloramphenicol, KCI,
, MgCl2 dithiothreitol (OTT), ethylenediaminetetracetic acid (EDTA), acrylamide, and Freund's incomplete adjuvant were obtained from Sigma Chemical
Company (St. Louis, MO). Glacial acetic acid, glycerol, methanol, sodium
, dodecyl sulfate (SOS), chloroform, isopropanol, Na2PO 4 NaCl, Long Ranger gel solution {50% acrylamide solution), and dimethyl sulfoxide (DMSO) were obtained from J. T. Baker Inc. (Phillipsburg, NJ). The bromochloro indolyl
28 29
phosphate (BCIP) bicinchoninic assay (BCA) kit was obtained from Boehringer
Mannheim (Indianapolis, IN). The lmject Supercarrier™ System for Peptides
was obtained from Pierce Chemicals (Rockford, IL). The Amplicycle™
Sequencing Kit was obtained from Perkin Elmer (Foster City, CA). Biomax
centrifuge filters were obtained from Millipore (Bedford, MA). [y-32P] ATP and
[a-32P] ATP were obtained from Amersham (Arlington Heights, IL).
Bacterial Strains and Alleles
All strains were isogenic derivatives of E. coli K-12 and are listed in Table 2.
Alleles CheA (Wolfe and Stewart, 1993), cheA, and cheA48HQ (Oosawa et al., 8
1988) encode wild-type CheA alone, both wild-type CheA and wild-type 8 8
CheAL, and both wild-type CheA8 and the mutant protein CheAL 48HQ,
respectively. This mutant protein cannot autophosphorylate because of a
His-Gin substitution at position 48. Alleles cheA422GA (herein called cheAG1),
cheA470GA472GA474GA ( cheAG2), cheA422GA470GA472GA4 74GA
( cheAG 1/G2) and cheA455FY459FY ( cheA2FY) encode CheA8 and CheAL proteins disrupted in G 1 alone, G2 alone, both G 1 and G2 or F alone,
respectively. Each mutation and unique restriction sites used to track these mutations during various in vitro manipulations were generated using standard oligonucleotide-directed mutagenesis procedures (Kunkel, 1985).
Oligonucleotides used for mutagenesis are illustrated in Table 3 and the 30
Table 2. --Bacterial strains used in this study
Strain Relevant Genotype Source AJW531 RP437 cheA470GA472GA474GA recA::cml eda:+- This Study AJW953 RP437 cheA422GA470GA472GA474GA This sr/::Tn 10 eda+ Study AJW955 RP437 cheA 422GA recA sr/::Tn 10 eda+ This Study AJW1071 RP9535 recA::cml This BL21(DE3) F- ompT hsdSs (rs- ms-) Ion dcm (').Jmm21 lacl Studier et /acUV al. (1990) T7 polymerase int) BMH71-18 thi supE A(lac-proAB) mutS::Tn 10/F' proA+ £3+ Kunkel mutS (1985) lac/OZ AM 15 CP366 po/A(Ts) Park and Hazelbau er(1986) DH5cx endA 1 hsdR17 (rK-mK-) supE44 thi-1 recA 1 gyrA Wood- (NalR) re/A 1 A(lac/ZYA-argF)U169 deoR cock et al.(1989) ( JM109 F' traD36 /acP A(lacZ)M15 proA+£3+ I e14- (McrA) Yannisch- Perron et A(/ac-proAB) thi gyrA96 (NalR) endA 1 hsdR17 al. (1985) (rK-mK-) re/A 1 supE44 recA 1 KO685 ArecA (Sstll-EcoRI) sr/::Tn 10 Hess et al. (1987) MH6 recA::cml A. Binney a RP437 thr(Am) 1 /eu86 his-4 metF(Am) 159 eda-50 Parkinson rpsL 136 thi-1/acY1 ara-14 mtl-1 xy/-5 tonA31 and Houts tsx-78 (1982) RP9535 RP437 AcheA 1643 recA:: cm/ eda+ Liu (1990) a A. Binney, Purdue University, West Lafayette, Indiana 31 Table 3.--Mutagenesis and sequencing primers Primer# M/S8 Allele Sequence 1 M G1 5'-ACCGACGATGCCGCGGGGCTAA-3' 2 M G2 5'-GACGTCTCCGCGCGCGCCGTCGCCATGGACGTC-3' 3 M F 5'-ATGCTGATATACGCACCCGGGTACCCACGGCAG-3' 4 M CheA709 5'-CCACCCCACATATGCTTAAAC-3' 5 s CheA 5'-AGAACGCATT ATCGACCCGC-3' 6 s CheA 5'-ATTCCAT AACAGCA TTCAGCG-3' 7 s T7 5'-AACAGCTATGACCATG-3' 8 s CheA 5'-CCTGGCTTCTCCACGGCAGAGC-3' a: M refers to mutagenesis primers and S refers to sequencing primers alleles presented in Fig 8. The following oligonucleotides were used: (i) to generate the allele cheAG1, 5'-ACCGACGATGCCGCGGGGCTAA-3' which substitutes an alanine residue for a glycine at position 422 (Fig 6 and 8) and which introduces a unique Saeli restriction site into G1; (ii) to generate the allele cheAG2, 5'-GACGTCTCC GCGCGTCGCCATGGACGTC-3' which substitutes alanine residues for glycines at positions 470, 472, and 474 (Fig 6 and 8) and which introduces a unique BssHII restriction site into G2; and (iii) to generate the allele cheA2FY 5'-ATGCTGATATACG CACC CGGGTACCCACGGCAG-3' which substitutes tyrosine residues for phenylalanines at positions 455 and 459 (Fig 6 and 8) and introduces a unique Xmal restriction site in F. Allele cheA.1(468-477) (herein called cheA~G2) encodes CheA and CheAL proteins 8 32 - - -- Allele G1 F ---G2 WT DDGAG FAPGF GRGVG cheA- G1 DDAAG FAPGF GRGVG cheA-G2 DDGAG FAPGF ;RAVA cheA- G1/G2 DDAAG FAPGF /P.AVA cheA-t:,G2 DDGAG FAPGF 6 cheA-2FY DDGAG YAPGY GRGVG Fig 8. Functional organization of CheA and schematic of cheA alleles used in this study. The shaded bars within TR represent the N, G1, F, and G2 blocks of highly conserved amino acids shared by histidine kinases of the family of two-component signal transduction systems. Mutant amino acids are italicized and underlined. 33 deleted for nine amino acids including those that comprise the entire G2 region (Fig 6 and 8). Media and Growth Conditions Cells were grown with aeration in tryptone broth (TB, 1% (wt/vol) Bacto-tryptone, 0.5% (wt/vol) sodium chloride) or in Luria-Bertani broth (LB, TB supplemented with 0.5% (wt/vol) Bacto-yeast extract). The optical density at 61 O nm (OD ) was monitored (Spectronic 20D, Milton Roy Company). 610 Bacteria strains containing plasmids to be made single-stranded for site-directed mutagenesis were prepared for R408 phage lysis by growth in TYP broth (1.6%(wt/vol) Bacto-tryptone, 1.6% (wt/vol) Bacto-yeast extract, 0.5%(wt/vol) NaCl and 0.25% (wt/vol) K HPO ). Competent cells were 2 4 prepared in Yb medium (2% Bacto-tryptone, 0.5% Bacto-yeast extract, and , 0.5% MgSO4 pH 7.6). Swarm Assays Cells were grown at 32°C to mid-exponential phase (OD610 ~ 0.3) in TB supplemented with the requisite antibiotics. Tryptone swarm plates (100 x 15 mm, Falcon) were constructed by adding 15 ml of 0.20% agar in tryptone broth. Antibiotics were not added. A 5 µI aliquot of the culture (106-107 cells) was placed on the surface of the swarm plate near its center and the plate incubated at 32°C in a humid environment. Swarm plates were handled and measured as described previously (Wolfe and Berg, 1989). Plasmids used for in vivo swarm plate analyses were derivatives of pAR1 34 (Gegner and Dahlquist, 1991 ), which controlled the expression of the cheA alleles. Plasmid Construction All expression plasmids used to purify proteins were constructed using the pET14b (Novagen) his tag purification vector and for the purpose of clarity were designated as pDE followed by the allele name Table 4. All expression plasmids used for behavioral studies are based on the pAR1 vector (Gegner and Dahlquist, 1991) and are designated as pAR1 followed by the allele name (Table 4). To construct the CheAL purification plasmid pDE.cheA, the CheA expression plasmid pAR1 .cheA (Wolfe and Stewart, 1993) was digested with Ndel and BamHI. The 2.1-kb fragment containing the cheA gene was ligated into an Ndel - BamHI digest of pET14b. To form the CheA purification plasmid 8 pDE.CheA , the CheA expression plasmid pAR1 .CheA (Wolfe and Stewart, 8 8 8 1993) was digested with Ndel and BamHI. The 1.8-kb fragment containing the CheA8 gene was ligated into an Ndel - BamHI digest of pET14b. The mutant allele CheMG2 was generated by digestion of pAR1 .cheA with Aatll. The 27 bp Aatll-Aatll fragment was discarded and the remainder of the plasmid ligated to create an in-frame deletion of 9 amino acids encompassing the G2 region. The resulting CheA~G2 expression plasmid was named pAR1 .cheA~G2. To construct the purification plasmid pDE.cheA~G2, pAR1 .cheAL1G2 was digested 35 with Ndel and BamHI and the 2.09-kb fragment ligated into a pET14b plasmid which had also been digested with Ndel and BamHI. pDE.cheA48HQ a purification plasmid, encoding the phosphorylation-deficient CheAL 48HQ, was generated by removal of a 1 .43-kb Munl - Mlul fragment from the CheAL 48HQ expression plasmid pCS30. This fragment was ligated into a 5.22-kb fragment of a Munl - Mlul digest of pDE.cheA. pAR1 .cheA48HQ, a cheA48HQ expression plasmid, was constructed by removal of a 2.1 kb Ndel - BamHI fragment from pDE.cheA48HQ, removing the entire cheA48HQ allele. This fragment was ligated into a pAR1 vector derived by digestion of pAR1 .CheA8 with Ndel and BamHI. To construct the purification plasmid pDE.cheAG1, a 2.1 kb Ndel - BamHI digest of the expression plasmid pJN.cheAG1(pJN19) was ligated into an Ndel - BamHI digest of pET14b. To construct the expression plasmid pJN.cheAG1, a 2.1-kb Ndel - BamHI digest of pJN001.cheAG1 (a pAlter™ plasmid containing cheAG1) was ligated into a vector resulting from an Ndel -BamHI digest of pAR1 .cheA . To construct pAR1 .cheAG2, a 692 bp BstE II - BamHI of pDE.cheAG1/G2 was ligated into a vector resulting from a BstBI BamHI digest of pAR1.cheA. To construct pDE.cheAG2, a 2.1-kb Ndel-BamHI digest of pAR1 .cheAG2 was ligated into an Ndel - BamHI digest of pET14b. The expression plasmid pDE.cheAG1/G2, was constructed by removal of a 2.1-kb Ndel -BamHI fragment of pJN001.cheAG1/G2 (a pAlter based plasmid 36 Table 4.--Plasmids used in this study Expression Plasmids to Express Mutation and/a r Allele Aasmids his-tagged Proteins cheA(WT) pAR1.cheA pDE.cheA cheA~WT) pAR1.cheAs pDE.cheAs cha4422GA pAR1 .cheAG 1 pDE.cheAG1 cheA470GA/472GN47 4GA pAR1 .cheAG2 pDE.cheAG2 c~A422 GA/ 470GA/472GA/ifl 4GA pAR1 .cheAG1/G2 pDE.cheAG1iG2 cha4481-0 pAR1 .cheA4BHQ pDE.cheA48HQ cheAA(468-476) pAR1 .cheAAG2 pDE.cheAAG2 containing the cheAG1/G2 mutation). This fragment was ligated into a vector resulting from an Ndel - BamHI digest of pAR1 .cheA. To construct pDE.cheAG1/G2, a 2.1-kb Ndel-BamHI digest of pJN.cheAG1/G2 was ligated into an Ndel-BamHI digest of pET14b. To construct pAR1 .cheAG2, a 692 bp BstEII-BamHI digest of pDE.cheAG1/G2 was ligated into a vector resulting from a BstEII-BamHI digest of pAR1 .cheA. This fragment was ligated into a vector resulting from an Ndel - BamHI digest of pAR1 .cheA. To construct pDE1.cheAG1/G2, a 2.1-kb Ndel - BamHI digest of pJN.cheAG1/G2 was ligated into an Ndel - BamHI digest of pET14b. To construct pAR1 .cheAG2, a 692 bp 37 BstBI - BamHI digest of pDE.cheAG1/G2 was ligated into a vector resulting from a BstE II - BamH I digest of pAR1 .cheA. Competence Procedure To make cells competent for transformation, a single freshly isolated colony was inoculated into 125 mis of Yb medium and aerated at 37°C until the culture reached an optical density (OD ) of 0.3. This culture was diluted 1:20 in 550 pre-warmed Yb medium and grown at 37°C with aeration until the culture reached an OD of 0.48, whereupon it was immediately placed on ice for 5 550 min and then pelleted by centrifugation at 3000 x g in a Beckman tabletop The cell pellet was resuspended in 40 mis cold (4°C) Tbf1 buffer (30 mM KOAc (pH , , 5.8), 100 mM KCI, 10 mM CaCl2 50 mM MnCl2 15% glycerol) for 5 min and pelleted by centrifugation at 3000 x g. The supernatant was decanted and the , pellet resuspended in 4 mis of Tbf2 (1 O mM MOPS (pH 6.5), 75 mM CaCl2 1O mM KCI, 15% glycerol) and left at 4°C for 15 min before use. Cells were divided into 250 µI aliquots and stored frozen at -70°C until needed (approx. 0.5 x 109 cells/ml). Site-Directed Mutagenesis Mutagenesis of the putative catalytic domain was performed using the Altered Sites™ in vitro Mutagenesis System (Promega). A BamHI digest of pAR1 .cheA which included both cheA and flanking sequences was ligated into a calf alkaline phosphatase (CIP, Gibco-BRL)-treated BamHI digest of the 38 mutagenesis vector pAlter™. The ligation reaction was transformed into DH5a, plated onto LB plates containing tetracycline (15 µg/ml), and incubated at 37°C for 16 hrs. Tetracycline resistant colonies were inoculated into 5 mis of LB containing tetracycline (15 µg/ml) and allowed to grow to saturation overnight at 37°C. The overnight cultures were pelleted by centrifugation and the plasmids isolated by use of the Wizard™ miniprep DNA purification system. The insert was mapped by restriction analysis for a 5' - 3' orientation relative to the T7 promotor present in pAlter™. A plasmid containing the proper insert in the correct orientation was selected and named pJN001. Single-stranded DNA used as a template in all site-directed mutageneses was obtained by transforming pJN001 into strain JM109. One freshly isolated colony was inoculated into 2 mis of TYP broth and incubated overnight at 37°C with aeration. 100 µI of the overnight culture was inoculated into 5 mis of TYP broth containing 15 µg/ml of tetracycline and aerated at 37°C. After 30 min, the culture was infected with the helper phage R408 at a multiplicity of infection (MOI) of approximately 10, shaken for an additional 6 hrs, and the culture was then centrifuged at 3000 x g to remove whole cells and debris. Single-stranded DNA was isolated from the supernatant by use of the Wizard™ M13 DNA purification system. Mutagenesis of pJN001 was performed according to the guidelines outlined by Promega . Phosphorylated oligonucleotides ( one designed to 39 introduce a mutation into CheA and one designed to repair the ~-lactamase gene in pAlter) were annealed to the single-stranded pJN001 template (0.05 pmol pJN001, 0.25 pmol ampicillin repair oligonucleotide, 1.25 pmol cheA mutagenic oligonucleotide, 2 µI 1OX annealing buffer, final volume 20 µI) by incubation at 70°C for 5 min and slow cooling to room temperature for 30 min. The annealing mix was allowed to extend and repair (3 µI 1OX synthesis buffer, 1O U T 4 DNA polymerase, 2 U T 4 DNA ligase, final volume 30 µI) at 37°C for 90 min. Following the extension reaction, the entire mix was transformed into competent BMH71-18 mutS cells. The BMH71-18 mutS containing the mutagenesis mix was inoculated into 4 mis of LB containing 100 µg/ml of ampicillin and incubated overnight at 37°C with aeration. The overnight culture was pelleted by centrifugation at 3000 x g and plasmids isolated by the miniprep system described previously. The plasmid preparation then was transformed into competent DH5a cells and plated on LB agar containing ampicillin (100 µg/ml) and tetracycline (15 µg/ml). Ampicillin resistant colonies were inoculated into 5 mis of LB broth containing ampicillin (100 µg/ml) and grown overnight at 37°C with aeration. Plasmids isolated from these overnight cultures were mapped for the presence of the unique restriction sites introduced as markers for each mutation. All mutations generated were confirmed by sequence analysis. 40 Homologous Recombination Alleles cheAG1, cheAG2 and cheAG1/G2 were introduced into the chromosome by means of homologous recombination in the po/A {Ts) strain CP366 (Park and Hazelbauer, 1986). Briefly, plasmids containing the mutant alleles were transformed into CP366, plated onto LB plates supplemented with ampicillin (100 µg/ml), and incubated overnight at 42°C. Ampicillin-resistant co integrants were selected, inoculated into LB without antibiotics and passaged repeatedly (1 0 X) at 37°C. A sample of the final passaged culture was serially diluted such that about 200 colonies were plated onto LB plates without antibiotic and grown overnight at 37°C. Ampicillin-sensitive recombinants were identified by replica-plating onto LB plates either containing ampicillin (100 µg/ml) or not. Chromosomal preparations were prepared from 6 ampicillin sensitive colonies and the existence of each mutation verified by restriction analysis of polymerase chain reaction (PCR) products using primers designed to amplify the entire G1-F-G2 region. Each mutation was marked by a silent mutation that introduced a unique restriction site that flanked the mutation. P1 transduction (Silhavy et al., 1984) was used to transfer each mutant allele out of its PolA(Ts) host into the PolA+RecA+Che+ recipient strain RP437 and to make the resultant transductants recA by using either strain MHS- or KO685 as the donor. The recipient strain RP437 carries an eda- phenotype. Recombinants of the chemotaxis operon (42.5 min) of RP437 can be selected by virtue of repair of the eda locus (41.5 min). Transductants that recombined 41 the mutated cheA alleles into RP437 were selected by their eda+ phenotype. The eda+ recombinants were screened for nonchemotactic behavior. Those that proved nonchemotactic were then tested for their ability to perform wild-type chemotactic behavior when transformed (Silhavy et al., 1984) with pAR1 .cheA, a plasmid that encodes both CheA8 and CheAL. Chromosomal preparations were prepared of each putative recombinant and the entire G 1-F-G2 region amplified by PCR. The presence of the mutation was verified by restriction analysis and sequencing of the amplified product. The mutant alleles are presented in Fig 8. Purification of His-tagged proteins To purify CheAL, the 2.1-kb Ndel-BamHI fragment from pAR1 .cheA (Wolfe et al., 1994) was subcloned into pET14b (Novagen Inc., Madison, WI), generating the plasmid pDE1. This plasmid permits expression, by means of the bacteriophage T71 O promoter, of a 6XHis sequence fused to the amino-terminus of CheAL. To purify mutant CheAL proteins, the 5.9-kb Stul-M/ul restriction fragment of pDE1 was ligated to 1.2-kb Stul-Mlul fragments from each of the respective pAR1 derivatives. To purify CheA , the 1.8-kb Ndel-BamHI 8 fragment from pAR 1.CheA (Wolfe et al., 1994) was subcloned into pET14b 8 (Novagen Inc., Madison, WI). To purify the TMC fragment of wild-type CheA, an Ndel site was engineered at cheA nucleotides 706 - 711 and the resultant 1.4-kb Ndel-BamHI fragment was subcloned into pET14b. Each protein used in 42 this study was purified by transforming the plasmid which expressed the appropriate 6X his-CheA variant fusion protein under control of the T7 promoter, into BL21 (DE3), a strain which expresses T7 RNA polymerase under the control of lacP {Studier et al., 1991 ). Cells of the resultant transformant were grown overnight in 5 ml of LB in the presence of ampicillin (100 mg/ml). The overnight culture was diluted 1: 100 in LB containing ampicillin (200 mg/ml) and grown to an OD of~ 0.4 at 30°C before adding 0.1 mM IPTG and more ampicillin (200 610 µg/ml). After a 3 hr incubation under inducing conditions, the cells were harvested by centrifugation at 3,000 x g for 20 min at 4°C. Purification of the fusion protein and cleavage of the 6XHis sequence were performed according to the manufacturer's instructions (Novagen [Hoffman and Roeder, 1991 ]), except that the thrombin:protein ratio was 1: 1000. Protein concentrations were determined by the BCA assay. To purify the 30-kDa fragment that results from spontaneous degradation of the mutant protein CheAL G 1, we filtered the mixed degradative pool of that protein through a SOK nominal molecular weight limit (NMWL) Biomax filter (Millipore). Because the protein was fused to an amino-terminal 6X his sequence, we reabsorbed the resultant Biomax filtrate over His-bind resin (Novagen), eluted those fragments that bound, and equilibrated them in TKMD buffer (50 mM Tris-HCI, 50 mM KCI, 5 mM MgC1 , 0.5 mM dithiothreitol 2 [pH 7.51). To purify spontaneous degradative fragments of CheAL G1 that include the 43 TMC domains, we first absorbed a mixed degradative pool of CheAL G 1 fragments to His-bind resin to remove any fragments containing an amino-terminal 6XHis sequence. We then filtered the unbound fragments sequentially through SOK NMWL and 30K NMWL filters. This process resulted in a mixed population of fragments ranging in size from 30- to 50-kOa,size range which includes TMC-containing fragments. In Vitro Phosphorylation of Proteins. SOS-PAGE. Autoradiography and lmmunoblot Analysis Phosphorylation reactions contained the indicated proteins in TKMO buffer 32 1 and 50 µM [y- P]ATP (~ 2500 cpm pmol" ) at 25°C for the times indicated. Reactions were terminated by the addition of 3X SOS-PAGE sample buffer. Samples were subjected to SOS-PAGE (10% polyacrylamide gels) at 10 to 15°C in a Hoefer (San Francisco, Ca) SE600 gel apparatus under the conditions of Laemmli (1971 ). Following electrophoresis, the gels were incubated for 2X at 25°C for 1 hr in drying buffer (45% methanol, 7.0% glacial acetic acid, 10% glycerol, 0.5% dimethyl sulfoxide), dried for 2 hrs and visualized by autoradiography. Sequencing Mutations in the catalytic domain were confirmed by direct cycle sequencing using the Amplicycle sequencing kit (Perkin-Elmer). Templates for sequencing were derived either from plasmids containing mutant alleles in the catalytic domain of cheA or chromosome preparations from E. coli cells which 44 had cheA catalytic site mutations introduced into the chromosome by homologous recombination. For each allele sequenced, 6 µI of the template mix (4 µI of cycling mix, 1 µI of [y2P]-ATP, 15 µI of template (100-150 µg/ml), and 1O µI of sterile H 0 ) was added to 0.25 ml microfuge tubes containing 2 µI each 2 of the G, A, T, and C termination nucleotides. 7-deaza-GTP was added to inhibit compressions. The sequencing mix was subjected to 30 cycles in an an Omn-E, (Hybaid) thermocycler under the following conditions. Pre-warm cycle (5 min, 95°C), melting cycle (1 min, 95°C), annealing cycle (1 min, 62°C), extension cycle (1 min, 72°C). Primer End-Labeling Primers used for sequencing were end-labeled according to the method of Donis-Keller et al. (1977). A reaction mix was prepared consisting of 5 pmol primer, 5 µI of 5X forward reaction buffer, 1 µI (10 U) of T4 polynucleotide kinase, 2.5 µI of [ y-32P]ATP (10 µCi/ul, 3000 Ci/mmol), and sterile H 0 to a total 2 volume of 25 µI. The contents were mixed thoroughly and incubated at 37°C for 10 min. The reaction was stopped by the addition of 4 µI of 500 mM EDTA. Alternatively, a reaction mix was prepared consisting of 5 µI [y-32P]ATP (10 µCi/ul, 3000Ci/mmol), 1 µI of 10x reaction buffer, 1 µI of 100 mM DTT, 3 µI of T 4 polynucleotide kinase and 0.8 µI of 20 µM primer. The contents were mixed 45 thoroughly and incubated at 37°C for 10 min. The reaction was terminated by incubation of the mixture at 94°C for 5 min. Genomic DNA Preparation To confirm chromosomal mutations in the putative catalytic domain of CheA, genomic DNA was prepared as described (Ellefson and Wolfe 1997). 100 mis of E. coli cells were grown overnight in LB. The overnight culture was pelleted by centrifugation and washed once in 1X SET buffer (20 mM Tris pH 7.5, 75 mM NaCl, 25 mM EDTA). The washed pellet was resuspended in 5 ml of SET buffer and transfered to a 15 ml polypropylene tube. Lysozyme was added to 1 mg/ml and the solution incubated at 37°C for 30 min. RNase A was added to a concentration of 0.4 mg/ml and the solution incubated for an additional 30 min at 37°C. Following the 60 minute incubation, 0.1 volumes of 10% SOS and 1 mg/ml of proteinase K was added and the entire solution mixed and allowed to incubate in a 56°C water bath. After 2 hrs, 0.33 volumes of 6M NaCl was added to the solution and mixed gently. One volume of chloroform was then added and the entire solution rotated for 1 hr at 23°C. The solution was then centrifuged for 1O min at 3000 X g to separate the organic phase from the aqueous phase. The aqueous phase was transferred to a fresh 15 ml polypropylene tube. One volume of isopropanol was added and the solution was inverted frequently to precipitate the chromosomal DNA. The DNA was transferred with a glass pipet to a 1.5 ml microfuge tube, rinsed with 70% EtOH, and pelleted by centrifugation. The supernatant was decanted from the pellet which was air dried to remove extraneous EtOH. The dried pellet was 46 resuspended in cold H 0 and stored at -20°C. 2 Western Analysis lmmunoblot analyses, using affinity-purified rabbit anti-CheA antibodies, were performed by standard procedures (Harlow and Lane, 1988). 40 µg or 5 µg of a total protein lysate were separated by SOS-PAGE analysis. The proteins were transferred to nitrocellulose (0.2 µm, Schleicher and Schull) in a Trans-Blot cell (Bio-Rad, Hercules) overnight at 4°C and 40V. The transfer buffer contained the following: Tris 3 g/I, Glycine 14.4 g/1, Methanol 200 ml/I. Unreacted sites on the nitocellulose blot were blocked by immersion for 1 hr in 100 mis of blotto (Tris-buffered saline/ 0.1% Tween-20 (TBS-T), 5% nonfat dry milk). The blots were rinsed twice and then washed three times each for 15 min in TBS-T on a rotating platform. Pre-absorbed anti-CheA antiserum or anti-CheAL serum were diluted to a concentration of 1 : 10,000 in TBS-T and incubated with the appropriate blot for 1 hr at room temperature. Following incubation with primary antibody, the blots were rinsed and washed as described above. The blots were incubated for 1 hr at 23°C with alkaline-phosphatase conjugated anti-rabbit lgG (Sigma) diluted in TBS-T to a concentration of 1 : 15,000. The blots were washed with TBS-T once for 15 min and three times for 5 min to remove unbound antibody. Following the wash, the blots were rinsed twice with an equilibrating buffer consisting of 150 mM NaCl, 50 mM Tris-HCI, [pH 7.5]. Reactive bands were visualized by adding 47 66 ml of nitro blue tetrazolium (NBT, 50 mg/ml in 70% dimethylformamide, Sigma) and 33 ml of bromochloro indolyl phosphate (BCIP, 50 mg/ml in 100% dimethylformamide, Sigma) into 1O ml of alkaline phosphatase buffer (100 mM ) Tris-HCI pH9.5, 5 mM MgCl2 and incubating the blot until color development occurs. The reaction was stopped by the addition of 5 mM EDTA. Blots were washed in distilled H 0 and dried between Whatman filters. 2 Generation of CheAL-Specific and CheA antisera High-titer antiserum recognizing only the long form of CheA was generated in rabbits by immunization with a conjugate of a CheAL-specific peptide and bovine serum albumin (BSA) carrier. A CheAL-specific peptide, IKGGAGTFGFS, corresponding to amino acids 50-60 in CheAL was coupled to BSA and keyhole limpet hemocyanin (KLH) by use of the lmject Supercarrier™ EDC system for Peptides (Pierce). The peptide/KLH conjugate was used as an irrelevant carrier to screen the resulting antiserum for peptide reactive antibody by ELISA. Each carrier was reconstituted into 200 µI of deionized water and incubated with 2 mg of peptide dissolved into 500 µI of conjugation buffer (pH 4.7). The peptide/carrier mix was added to a vial containing 1-ethyl-3-(3-dimethylaminopropyl carbodiimide) hydrochloride (EDC). The peptide/carrier/EDC mix was allowed to incubate at 23°C for 2 hrs. The peptide/carrier conjugates were purified over a cross-linked dextran sulfate column pre-equilibrated with purification buffer (83 mM Na PO 2 4 48 pH 7.2, 900 mM NaCl). 0.5 ml aliquots of purification buffer were absorbed over the column and collected in separate tubes for analysis. The hapten-carrier conjugate, eluted in fractions 4 - 6. These fractions were pooled and the coupling efficiency calculated by comparing of peptide (by absorbance at 280nm) before conjugation with the peptide concentration in the eluate arising from the coupling column. The coupling efficiency of the peptide to both the BSA and KLH carrier conjugates was estimated at approximately 78% (approx. 10 peptides/molecule BSA). The BSA/peptide conjugate was dialyzed against phosphate buffered saline (PBS), quantitated by BCA analysis (a protein determination kit from Pierce), and 100 µg resuspended in 1 ml of Fruend's incomplete adjuvant. The mixture was shaken for 1 min in a model 5100 mixer/shaker (Spex) and drawn into a 3 ml disposable syringe. Pre-immune serum was collected from a female New Zealand White rabbit (approx. 5 kg). The rabbit was injected subcutaneously at 5 separate sites in the dorsal-lumbar region. After 4 weeks, the rabbit was boosted with 50 µg of the peptide/ BSA conjugate in a manner similar to that outlined above. Two weeks after boosting, 35 mis of blood was taken by ear puncture and the serum collected by allowing the blood to coagulate overnight at 4°C. Serum reactivity against the peptide was analyzed by indirect detection of peptide titer by enzyme-linked immunosorbant assay (ELISA) using the peptide-KLH conjugate as the antigen according to the method of Van Regenmortel (1988) . Briefly, 50 µg/ml of peptide/KLH conjugate was dialyzed against carbonate buffer (15mM Na CO , 2 3 49 35 mM NaHCO , pH 9.6, 4°C). Each well of a a 96-well plate was absorbed 3 overnight at 4°C with 100 µI of the dialyzed peptide/KLH conjugate mixture. The plates were washed three times with PBS/0.1 % tween-20 and the unreacted sites blocked by incubating the plate at 37°C for 1 hr with a 100 µI of 10 mg/ml of KLH in PBS/0.1 % tween-20. The unreacted KLH was removed by washing 3 times in PBS/0.1 % tween-20. Peptide-reactive antiserum was quantitated by aliquoting 100 µI/well of 2-fold serial dilutions of serum ranging from 1:100 to 1:10,000 and incubating the plate at 37°C for 2 hr. The plate was washed 3 times in PBS/tween-20 and incubated with alkaline-phosphatase conjugated goat anti-rabbit lg (Sigma, 1:10,000) for 2 hrs at 37°C. The wells were washed 3 times with PBS/tween-20 and incubated for 2 hr at 37°C in substrate (3 mM p-nitrophenyl phosphate, 0.05 M Na CO , 0.5 mM MgCl ). 2 3 2 The absorbance was read at 405 nm with a spectrophotometer. Generation of antiserum against both the long and short forms of CheA was performed in a manner similar to the one outlined above except his-tagged CheA protein was used as the primary immunogen instead of a peptide-conjugate mixture. ATP Binding Assay The binding of ATP to wild-type and mutant CheA was performed according to the method of Ninfa et. al (1993). Each wild-type and mutant CheA protein was diluted to 5 µM in TKMG buffer (50 mM Tris-HCI pH 8.0, 60 mM 50 32 KCI, 5.0 mM MgCl , and 5.0% glycerol), and incubated with [a- P]ATP. A 40 2 µI aliquot of each protein was dispensed into separate wells of a 96 well microtiter plate and exposed to UV light (254 nm) for 3 min with a UV Stratalinker model 1800 (Stratagene, LaJolla, California). The reactions were stopped by the addition of 3X sample buffer and subjected to SOS-PAGE and autoradiography. Quantitation of ADP binding was performed as described (Thomas et al. 1992; Hiratsuka 1982). Progressively higher concentrations of the fluorescent ADP analog TNP-ADP were titrated into a 1O µM concentration of wild-type or mutant CheA. By itself, TNP-ADP has a low level of fluorescence, but there is a significant increase in fluorescence when the analog is bound to a kinase. This increase was quantified by measurement of the absorption spectra in an Applied Photophysics SX19MV spectrofluorimeter (model SK.1 E). Data were plotted as fluorescent intensity versus TNP-ADP concentration. These titration profiles were fitted with a least squares program to calculate the binding affinity of the analog. CHAPTER4 RESULTS In Vivo Analysis of CheA Putative catalytic domain Mutants Rationale for design of catalytic mutants: Parkinson and Kofoid (1992) proposed, on the basis of sequence homology to the catalytic sites of eukaryotic serine/threonine kinases, that the transmitter domain of CheA is the site of ATP catalysis. To test this hypothesis, ideally one would first derive structural information from x-ray crystallography or multidimensional NMR analyses and then obtain information via functional analysis of selected mutants. Unfortunately, there have been no successful structural studies of the putative catalytic domain of CheA or that of any other two-component histidine kinase. Thus, I have chosen to use mutagenesis to obtain functional and structural information concerning the putative catalytic domain of CheA. To design a mutagenic strategy, I first compared the sequence of the catalytic domains of serine/threonine kinases and with that of several histidine kinases (Saraste et al. 1990, Schultz 1992, Taylor et al. 1992, Ninfa et al. 1993, Yang and Inouye 1993). I chose three highly conserved regions (G1, F, and G2) for 51 52 mutant analysis of catalytic function. Since G 1 and G2 are glycine-rich regions, I decided to generate a series of conservative glycine to alanine changes because researchers have traditionally used alanine substitutions to facilitate functional analysis of the catalytic regions of other kinases (Atkinson and Ninfa; 1993, Ninfa et al., 1993) and because a variety of structure averaging algorithms predicted that G to A substitutions would cause no obvious alterations in the secondary structure of those regions. Thus, I expected that such mutations had a good chance of preserving the general architecture of the mutated protein. In the F region, which is phenylalanine-rich, I introduced F to Y and F to W mutations. Because tyrosine (Y) differs from phenylalanine (F) only in the presence of a reactive hydroxyl group on the aromatic ring, I expected F to Y to be the most conservative of the two substitutions. In contrast, I expected that a tryptophan (W) substitution might cause a more severe phenotype since it contains an additional ring that may cause steric problems (Sharp et al. 1995, Sharp, et al., 1995). Using these mutations, I sought to address the following questions concerning the function of the conserved subdomains within the transmitter domain of CheA: •First, do these subdomains play a role in chemotaxis in vivo? •Second, do these subdomains play a role in the ability of CheA to autophosphorylate? •Third, are the subdomains involved in formation and maintainance of the ternary complex? •Fourth, are the subdomains involved in the binding of ATP? 53 • Fifth, within a CheA dimer, do the conserved subdomains operate independently and in cis, or interdependently and in trans ? •Sixth, can we make predictions concerning the structure of the putative catalytic region of CheA? Mutant Alleles Defective in the Putative Catalytic Domain do not Support Chemotaxis/n Vivo A deletion of cheA eliminates the ability of the cell to perform chemotaxis (Liu and Parkinson, 1989). Since CheA is a histidine kinase, mutations designed to disrupt the binding of ATP should affect the ability of CheA to autophosphorylate. Such mutations would disrupt the phospho-relay and, thus, affect chemotaxis. Thus, I inoculated cells carrying either the wild-type allele cheA (strain RP437) or the mutant allele cheAG1 (strain AJW955), cheAG2 (strain AJW531),cheAG1/G2 (strain AJW 953) or cheAt1G2 (strain AJW415) at the center of tryptone swarm plates and, after 6 hr of incubation, compared the displacements of the outermost edges of their resultant swarms (Fig 9). Wild- type cells (Fig 9A) formed two rapidly migrating concentric bands (0.72 cm h,1 for the outermost band), indicating that such cells are chemotactic for both L serine and L-aspartate (Adler 1966). In contrast, cells of each mutant strain formed dense irregular swarms that migrated at a rate (0.02 cm hr·1) characteristic of nonchemotactic cells that cannot tumble, e.g., the t1cheA strain AJW1071 (Fig 9B). To determine whether increased expression of the mutant proteins could confer chemotactic ability, I transformed the mutant alleles cheAG1, cheAG2, 54 WT 11cheA A B Fig 9. Pho tog raphs of representative swarms demonstrating that wild-type cells (A) form two concentric rings whereas cells deleted for cheA (B) do not. Cells were inoculated near the center of replicate trypto ne swarm plates containing 0.2% agar, and the plates were incubated at 32°C for 6 hr. Experiment is representative of repetitive experiments ( >1 0). 55 cheAG1/G2, cheA G2, or cheA2FY or cheA2FW (each carried by the IPTG inducable vector pAR 1) into the !lcheA strain AJW1071 . For positive and negative controls, I also transformed this !lcheA strain with the alleles cheA and CheA (also carried by the vector pAR1). I inoculated cells of the resultant 8 transformants at the center of tryptone swarm plates containing 25 µM I PTG and, after 6 hrs of incubation, compared the displacement of outermost edges of the swarms. As shown in Fig 10, only cells that expressed both CheAL and CheA formed two rapidly migrating concentric bands (Panel A). Those that 8 expressed variants of CheA mutated in G1 (Panel 8), G2 (Panel C), G1/G2 (Panel D), or F (Panel E) formed dense nonchemotactic swarms indistinguishable from those produced by the cells of AJW1071 alone (Fig 98). The Inability to Perform Chemotaxis is not due to a Metabolic Defect The inability of mutants of the putative catalytic domain of CheA to perform chemotaxis could be due solely to reduced catalytic activity. Alternatively, it could be due to reduced ability to metabolize serine. A reduction in the ability to metabolize an amino acid such as serine would manifest itself as a reduced growth rate. To distinguish between these possibilities, I transformed the !lcheA strain AJW1071with pAR1 derivatives that carried the alleles cheA, cheA , 8 cheAG1, cheAG2 or cheAG1/G2 (each carried by the IPTG-inducible vector pAR1 ). I then measured the growth rate of each transformant in TB at 32°C either in the absence of IPTG or in the presence of 50 µM IPTG. All transformants grew at about the same rate, regardless of IPTG 56 A cheA 8 cheAG1 C cheAG1 I G2 D cheAG2 E cheA2FY F cheA2FW Fig 10. Photographs of representative swarm patterns demonstrating that increased expression of plasmid-borne cheA mutants does not restore chemotaxis to cells deleted for cheA: cells deleted for cheA and transformed with either pAR1 .cheA (A}, pAR1 .cheAG1 (B), pAR1 .cheAG2 (C), pAR1 .cheAG1/G2 (D), pAR1 .cheA2FY (E), or pAR1 .cheA2FW. All cells were inoculated in duplicate onto replicate (x 5)TB swarm plates supplemented with 25 mM IPTG to induce expression of the plasmid-borne alleles. 57 A B 1.0 1 .0 -.....0 -.....0 (0 (0 ci ci d d -Q) -Q) ni 0.1 ni 0.1 a: a: .c .c -3: -3: s s 0.01 0.01 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Time ( hrs) Time (hrs) Fig 11. Growth curves demonstrating that cells deleted for cheA ( v) or deleted for cheA and transformed with pAR 1 derivatives that carry the alleles cheA ( ■), cheAG1 ('~ ), cheAG2 ( e), or cheAG1/G2 ( □)grow at about the same rate. Cells were inoculated into TB supplemented with no IPTG (A) or 50 µM IPTG (B) and incubated at 32oC. This experiment was performed twice. 58 induction (Fig 11). CheAs. but not CheA 48HQ Restores of Chemotactic Ability to Kinase-Detident 6 CheAL Mutants CheA8 lacks the N-terminal 97 amino acids including the site of phosphorylation (48H). However, it still possesses a functional catalytic domain. In contrast, variants of the putative catalytic domain of CheAL possess the site of phosphorylation, but presumably lack kinase activity. CheA functions as a dimer (Surette et al., 1996). Furthermore, the expression of CheA 8 restores chemotaxis to cells that express the kinase-deficient mutant CheAL- 470GK, presumably by forming a heterodimer (Swanson et al., 1993; Wolfe and Stewart 1993). Therefore, to test whether the non-phosphorylatable variant CheA8 can also restore chemotactic ability to mutant cells that express other CheA variants defective in the glycine-rich regions, I transformed the allele cheA8 (carried by the vector pAR1) into cells carrying chromosomal copies of the mutant alleles cheAG1, cheAG2 or cheAG1/G2 (strains AJW955, AJW531 or AJW953, respectively). Cells of the resultant transformants were inoculated at the center of tryptone swarm plates containing 25 µM IPTG and, after 6 hrs of incubation, the displacements of their swarm fronts were compared (Fig 12). Transformants of strains AJW955 and AJW531 (alleles cheAG1 and cheAG2, respectively) formed two concentric bands (Panels D and E) that resembled those of the wild-type strain RP437 (Panel A). The outermost bands, formed by 59 these transformants in response to L-serine, migrated about 0.28 cm h 1 and 0.33 cm h-1, respectively. In comparison, pAR1 .cheA transformants of the ilcheA strain AJW1071 formed an outermost band that migrated at a rate of 0.49 cm h- 1 (Panel C). In contrast, transformants of strain AJW953 (allele cheAG 1/G2) formed dense non-chemotactic swarms (Panel F) that migrated as 1 slowly (0.02 cm h- ) as cells of the ilcheA strain AJW 1071 (Panel B), even when the concentration of IPTG was increased to as much as 100 µM. Expression of CheA8 restores chemotactic ability to cells that express CheA proteins mutated in either the G1 or G2 regions but not both. Likely, this occurs because CheA forms a functional heterodimer with each 8 of the catalytic mutants. The non-phosphorylatable mutant CheAL 48HQ was chosen as a control because, like CheA , it possesses an intact putative 8 catalytic domain and no site of phosphorylation. Since the putative catalytic domains of CheA8 and CheAL 48HQ, are identical, the expression of CheAL 48HQ should also restore chemotaxis to a CheAL protein mutated in the conserved G 1 and G2 subdomains. I transformed the plasmid pAR 1 . cheA48HQ into cells carrying the mutant alleles cheAG1 (strain AJW955) or cheAG2 (AJW531 ), inoculated cells of the resultant transformants at the center of tryptone swarm plates containing 25 µM IPTG and, after 6 hrs of incubation, compared the displacement of their swarms. Transformants of both strains formed dense, non-chemotactic swarms indistinquishable from those produced 60 cheA LlCheA LlCheA + cheA A B C cheAG1 cheAG2 cheAG1/ G2 + + + cheA5 cheA5 cheA5 D E F Fig 12. Photographs of representative swarm patterns demonstrating that CheAs restores chemotaxis to cells that express the kinase-deficient mutants CheALG1 or CheALG2 but not CheALG1/G2. Wild-type cells (A, strain RP437), cells deleted for cheA (B, strain AJW1071 ), cells deleted for cheA and transformed with pAR1 .cheA (C),cells carrying the allele cheAG1 and transformed with pAR1 .cheAs (D), cells carrying the allele cheAG2 and transformed with pAR1 .cheAs (E), and cells carrying the allele cheAG1/G2 and transformed with pAR1 .cheAs (F). Cells were inoculated at the center of tryptone swarm plates containing 0.2% agar, and the plates incubated at 32°C for 6 hr (replicate, x5). Expression of CheA variants was induced with 25 µM IPTG. 61 by the parent cells or those of the !!,.cheA strain, AJW1071 (Fig 128). Thus, in , contrast to that of CheA8 expression of CheAL 48HQ does not restore chemotactic ability to cells that express CheA proteins defective in either the G 1 or G2 regions. Similarly, cells that carried the mutant allele cheAG1 transformed with the allele cheAG2 (carried by the IPTG-inducible vector pAR1) and cells that carried the mutant allele cheAG2 transformed with the allele cheAG1 (also carried by pAR1) formed only non-chemotactic swarms that migrated slowly regardless of IPTG concentration. Thus, CheAL G1 and CheAL G2 variants cannot complement each other in vivo. Dominance of Various CheA Catalytic Mutants To determine the effect exerted by the putative catalytic mutants upon the ability of wild-type CheA to perform its proper function, I explored the possibility that the phenotypes resulting from these mutations could be explained by an acquired ability of the mutants to exert a dominant negative effect. I transformed wild-type cells with the following alleles: cheA, cheAG1, cheAG2, cheAG1/G2, or cheAAG2 (each carried by the IPTG-inducible vector pAR1 ). Each transformant was inoculated in duplicate onto the center of replicate swarm plates containing 25 µM IPTG and their migration rate followed over 6 hrs (Fig 13). Wild-type cells transformed with cheA (panel A), cheA AG2 (panel C), cheAG1 (panel D), or cheAG1/G2 (panel E) produced a serine band that 62 A cheA B cheAG2 cheAt1.G2 cheAG1 E cheAG1/G2 Fig 13. Photographs of swarm plates demonstrating that allele cheAG2 confers a dominant-negative phenotype in a wild-type background. Cells of the wild-type strain RP437 were transformed with: (A) pAR1 .cheA, (B) pAR1 .cheAG2, (C) pAR1 .cheMG2, (D) pAR1 .cheAG1, and pAR1 .cheAG1/G2 (E). The transformants were grown to mid-exponential phase in TB and spotted ontothe center of replicate (x5) tryptone agar plates containing 0.2% agar, 25 µM, IPTG and incubated at 320c. The displacement of their outermost swarms were compared after 6 hrs. 63 A 8 C 80 80 80 70 E 70 -E 70 E E 60 E 60 E 60 ... 50 50 50 Q) D E F 80 80 80 E 70 E 70 E 70 E 60 E 60 E 60 ... 50 50 50 Q) G 80 E 70 E 60 0 1 2 3 4 5 6 7 8 9 10 Time (hrs) Fig 14. Displacement of the outermost serine band plotted as a function of time demonstrating the effect of overexpressing plasmid-borne CheA variants transformed into wild-type cells. cheA mutant catalytic alleles (each carried by the IPTG-inducible vector pAR1) were transformed into the wild-type background RP437. Cultures of each strain were inoculated into TB and grown at 320c with aeration to an O.D.610 of 0.3. Cells were inoculated onto replicate (x5) TB plates containing ( 6) 0, ( •) 25, ( 'v) 50, or ( ♦) 100 µM IPTG and the displacement of their outermost bands compared over time. Wild-type cells transformed with the IPTG inducible vector expressing: (A) cheA, (B) cheAG1, (C) cheA.1G2, (D) cheAG2, (E) cheAG1/G2, (F) cheA2FW, and (G) cheA2FY. 64 migrated at rates of approximately 0.22 cm hr -1, 0.28 cm hr1, 0.23 cm hr1, or 0.28 cm hr1 respectively. In contrast, cells transformed with allele cheAG2· 1 (Panel B) migrated very slowly relative to the other mutants (0.08 cm hr ). To determine whether increased expression of any of the CheA variants could either produce or enhance enhance a dominant negative effect, this experiment was repeated using a range of IPTG concentrations ranging from 0 to 100 µM. Overexpression of the wild-type cheA allele in a wild-type background decreases the migration rate in a dose-dependent fashion (Panel A). The mutant cheA alleles cheAG1 (Panel B) and cheAG1/G2 (Panel E), when overexpressed in a wild-type background, decrease migration rate in a manner similar to that observed for overexpression of wild-type cheA. In contrast, overexpression of the mutant allele cheAG2 (Panel D) exerts a more pronounced effect on migration rate than observed with overexpression of either cheAG1, cheAG1/G2 or cheA. CheA~G2 reduces migration rate less than overexpressed wild-type cheA. Finally, allelescheA2FY and cheA2FW seem to exert no effect at all upon migration rate, even when induced at 100 µM IPTG. Thus, the allele cheAG2 has a dominant-negative effect upon chemotaxis. In contrast, both cheA2FY and cheA2FW are recessive mutations. Alleles cheAG1, cheAG1/G2, and cheA~G2 exert a limited effect upon chemotaxis. 65 In Vitro Autophosphorylation Analysis of CheA Variants Defective in G1, F orG2 Mutations in the transmitter domain of CheA, which alter the conserved subdomains comprising the putative catalytic site, eliminate the ability to perform chemotaxis. A catalytic site is distinguished by its ability to bind and hydrolyze ATP. If the G1, F, and G2 subdomains comprise the catalytic site of CheA, then CheA variants defective in these regions should fail to autophosphorylate. To test this hypothesis, I purified each CheA variant and analyzed them for their ability to autophosphorylate in vitro. I performed autophosphorylation experiments on wild-type CheAL' wild- type CheA , and each mutant protein. Wild-type CheAL, wild-type CheA and 5 5 the CheAL mutant proteins CheAL G 1, CheAL G 1/G2, CheAL G2, CheAi:1G2, CheAL2FY, and CheAL48HQ, were purified to near homogeneity, as judged by SOS/PAGE (Fig 15). To ensure that the purified proteins were variants of CheA, western analysis was performed with CheA-specific antibody (data not shown). To assess the ability of the CheA variants to autophosphorylate, I added 32 [y- P]ATP to purified preparations of wild-type CheAL' wild-type CheA5 and each mutant protein. At the conclusion of a 5 min labeling period at room temperature, the reaction was terminated by the addition of 3X SOS/PAGE sample-loading buffer, the proteins separated by SOS/PAGE, and the radioactive bands detected by autoradiography (Fig 16). As expected, the reaction mixture containing only wild-type CheAL (lane 1) resulted in a 66 N C, 0 N >, ,....; ,....; N C, ::r:: C, - C, ri. 00 '-'...:i <;b N...:i <( Cl'.l 97.4 - 68 - 43 - 29 - Fig 15. Purified variant CheA proteins. Proteins were isolated from cells of strain BL21 (OE3) transformed with various CheA expression vectors, bound and eluted from His-bind resin, separated by SOS-PAGE (10% gel), and detected by Coomassie blue staining. Lane (1) wild-type CheAL, lane (2) wild-type CheAs, lane (3) CheALG1 , lane (4) CheALG1/G2, lane (5) CheALG2, lane (6) CheALL\G2, lane (7) CheAL2FY, and lane (8) CheAL48HQ. Following each purification, all proteins were confirmed for purity by SOS-PAGE analysis. The image was derived by scanning a photograph of the resultant gel with a Hewlett-Packard Scanjet model lie. 67 N d Cl ---E------E--- ,_.. ,_.. ~ ::c N N 00 d ~ ~ -d ll...... l en d....l d....l N....l <] ....l ------~ ~""" <(l) <(l) <(l) <(l) (l) <(l) (l) <(l) ..c ..c ..c ..c ..c ..c ..c ..c u u u u u u u u kD 1 2 3 4 5 6 7 8 97.4 - 68- 43- 29 - Fig 16. In vitro autophosphorylation of wild-type and catalytic mutant CheA proteins. CheA samples (5 µM final concentration of each protein) were incubated with 50µM [y-32P]ATP (~2500 cpm pmol-1) in TKMD buffer (see Materials and Methods), the reactions were terminated by the addition of 3X SOS-PAGE sample buffer, and the samples were analyzed by SDS/10% PAGE followed by autoradiography for 60 min. Lane 1, wild-type CheAL; lane 2, wild-type CheAs; lane 3, CheALG1; lane 4, CheALG2; lane 5, CheALG1/G2; lane 6, CheAL2FY; lane 7; CheAL48HQ; and lane 8, CheAL L\G2. The image was derived by scanning a photograph of the resultant gel with a Hewlett-Packard Scanjet model lie. 68 radiolabeled protein of MW about 78-kDa. in contrast, the reactions containing the following CheA variants did not result in such a band: CheA8 (lane 2), CheALG1 (lane 3), CheALG2 (lane 4), CheALG1/G2 (lane 5), CheAL2FY (lane6), CheAL 48HQ (lane 7), or CheAL LiG2 (lane 8). Thus, unlike wild-type CheAL, wild-type CheA and the mutant proteins did not autophosphorylate. 8 ATP Binding On the basis of sequence homology with ATP binding motifs identified in eukaryotic serine/threonine kinases, Parkinson and Kofoid (1992) proposed the glycine-rich regions of the transmitter domain to be the site of ATP binding. However, no study has identified conclusively that this region actually binds ATP. My first attempt to address this problem ended in failure. I attempted to cross-link [a-32P]ATP to CheA using UV light. The resulting adducts were separated by SOS-PAGE and visualized by autoradiography. All of the proteins, both wild-type and mutant forms, seemed to bind ATP ( data not shown). To determine if such binding was specific or not, I attempted to interfere with the ability of wild-type CheA to bind [a-32P]ATP by pre-incubating CheA with 100-fold excess non-radioactive ATP. This excess unlabeled ATP failed to eliminate labeled ATP binding. I concluded that this method may encourage non-specific binding of ATP. Subsequently, we established a collaboration with Richard Stewart (University of Maryland). To determine if mutagenesis of the putative catalytic domain had an effect on nucleotide binding, the fluorescent ADP analog, TNP- 69 ADP, was titrated against 10 µM of wild-type CheA and that of CheAL G 1, CheALG2, CheALG1/G2 or CheAL~G2. (according to the method of White and Dewey, 1987). The increase in fluorescence was measured and the data plotted as fluorescent intensity as a function of TNP-ADP concentration of the K0 calculated (Table 4). The mutants CheALG1 or CheALG2, CheALG1/G2, and CheAL~G2 bound TNP-ADP with a significantly reduced affinity relative to that of wild-type CheAL (K of 60 µM, 110 µM, 16 µM, and 5 µM respectively). In 0 contrast, CheAL G2 bound TNP-ADP about as well as wild-type CheA (K0 of 6 µM). Thus, the G 1 region of CheA appears to be intimately involved in nucleotide binding whereas the G2 region does not. The Non-Phosphorylatable Variants CheA:S and CheA.L 48HQ can Mediate Transphosphorylation of Kinase-Deficient CheA.L Mutants In Vitro To assess the ability of the non-phosphorylatable CheA variants, CheA8 or CheAL 48HQ, to promote phosphorylation of the CheA putative catalytic mutants CheAL G 1, CheAL G2, CheAL G 1/G2 and CheAL ~G2, [y-32P]ATP was added to purified preparations of wild-type CheAL or wild-type CheA8 or each mutant protein which had been pre-incubated for 30 min in the presence of either CheA8 or CheAL 48HQ. At the conclusion of a 30-min labeling period at room temperature, the reaction was terminated by adding 3X SOS-PAGE 70 Table 5.--TNP-ADP binding affinities of CheAL proteins Protein Ko CheAL(WT) 5µM CheALG1 60µM CheALG2 6µM QieALG1/G2 110 µM CheAL~G2 16µM sample-loading buffer and the results analyzed by SOS-PAGE and autoradiography. Reactions containing wild-type CheAL alone or both CheA8 and any of the following CheA mutants resulted in a radiolabeled protein of MW about 78-kDa: CheALG1 (lane 1), CheALG2 (lane 2), CheALG1/G2 (lane 3) and CheAL~G2 (lane 4). In contrast, reactions containing only CheA did not. 8 Interestingly, reactions performed with CheA repeatedly resulted in more 8 radiolabeled CheALG1 or CheALG2 (about 6% of wild-type levels) than radiolabeled CheAL G 1/G2 or CheAL ~G2 (about 1.5% of wild-type levels). 71 + CheA5 (5 µM) 5µM A + CheAL 48HQ (5 µM) 5µM B Fig 17. In vitro assays demonstrating that the non-phosphorylatable proteins CheAs (A) and CheAL 48HQ (B) can mediate transphosphorylation of kinase deficient mutants of CheAL. CheA samples containing 5µM [final co-ncentration] of either CheAL or CheAs or both CheAs and each mutant protein or CheAL 48HQ and each mutant protein were pre-incubated for 30 min in TKMO buffer (see Materials and Methods) before the addition of 50µM [y-32 P]ATP (~2500 cpm pmol-1). The reactions were terminated after 30 min by the addition of 3X SOS-PAGE sample buffer and the samples were analyzed by SOS-PAGE (10%) followed by autoradiography for 60 min. Two radiolabeled bands were observed: one of about 78 kOa and a second, of about 30 kOa. Only the 78 kOa band is shown. Counts within this 78 kOa band were collected for 1 hr with a Betascope 603 blot analyzer within the linear range of the instrument. The experiment was performed 3 times. 72 Similarly, reactions containing both CheAL48HQ (Panel B) and each of the cheA mutants resulted in a radiolabeled protein MW 78 kOa. The intensity of the signal however, was substantially reduced compared to the levels observed for CheA . Again, the amount of radiolabelled CheAL G1 or CheAL G2 was 8 greater than the amount of radiolabelled CheAL G 1/G2 or CheAL ~G2 To assess the ability of the mutant proteins CheAL G1 and CheAL G2 to promote phosphorylation of each other, I added [y-32P]ATP to purified preparations of wild-type CheAL alone, CheAL G1 alone, CheAL G2 alone, both CheAL G 1 and CheAL G2, or CheAL2FY alone that had been preincubated for 30 min (Fig 18). At the conclusion of a 30-min labelling period at room temperature, the reaction was terminated by adding 3X SOS-PAGE sample loading buffer and the results analyzed by SOS-PAGE and autoradiography. In contrast to reactions containing wild-type CheAL alone (lane 1), those containing CheAL G1 alone (lane 2), CheAL G2 alone (lane 3), both mutant proteins together (lane 4), or CheAL 2FY alone did not result in a radiolabeled 78-kOa protein. Reactions performed with both CheAL G 1 and CheAL G2 together (Lane 4) resulted in very poor phosphorylation of a protein that migrated more slowly than wild-type CheAL. For several reasons, it seems unlikely that this band corresponds to a phosphorylated form of either CheAL G 1 or CheAL G2 or of their 6XHis-tagged variants. First, the phosphorylated forms of CheAL G1 and CheAL G2 migrate with the same mobility as phosphorylated 73 wild-type CheAL. Second, after cleavage with thrombin, we detected no 6Xhis CheA variant by immunoblot analysis. Third, we detected no phosphorylation of the 6X-His tag itself when fused to either of the nonphosphorylatable proteins, CheA and CheAL 48HQ (data not shown). 8 These results are consistent with a model by which the G1 and G2 subdomains within the TR of one protomer of a CheA dimer form a unit that mediates transphosphorylation of the other protomer within that dimer. Interestingly, reactions involving CheAL G1 (lanes 2 and 4) yielded a radiolabeled protein fragment of MW about 30-kDa. We observed a phosphorylated fragment of similar molecular weight in reactions involving wild-type CheAL alone (lane 1) or the F mutant, CheAL 2FY (lane 5). To determine whether the observed fragments included the site of phosphorylation, His-48, we performed immunoblot analyses using anti-CheA (Fig 19a) or anti-CheAL antisera (Fig 19b). Wheras CheAL and the 30-kDa fragment reacted with both antisera (lanes 1 and 3 respectively), CheA reacted with only 8 the anti-CheA serum (lanes 2). On the basis of its size (30kDa), immunoreactivity with antisera specific for an epitope unique to the P1 domain, and the ability to accept a phosphoryl group, I have concluded that this degradation product corresponds to a previously reported CheA fragment composed of the N-terminal P1 and P2 domains (Oosawa et al., 1988). Thus, although the G1 and F mutants cannot autophosphorylate, they seem to be able to mediate transphosphorylation of the P1 -P2 fragment. · In contrast, 74 I- - T- T- >- ;: C\J C\J LL CJ CJ CJ CJ C\J <( <( <( <( + 97.4 - 68 - 43 - 29 - • Fig 18. Phosphorylation of CheAL variants. CheA samples (each protein at 5 µM [final concentration]) were pre-incubated for 30 min in TKMO buffer (see Materials and Methods) before the addition of 50µM [y-32P]ATP ( ~2500 cpm pmol-1). The reactions were terminated by the addition of 3X SOS-PAGE sample buffer and the samples were analyzed by SOS-PAGE (10% gel) followed by autoradiography for 60 min. Lane 1, wild-type CheAL; lane 2, CheAL G 1; lane 3, CheAL G2; lane 4, CheAL G 1 and CheAL G2; lane 5, CheAL2FY. Arrow indicates position of 30 kOa degradation product. The image was derived by scanning autoradiographs with a Hewlett-Packard Scanjet model lie. 75 a b kD 1 2 3 1 2 3 97.5 - 68 30 Fig 19. lmmunoblot analyses using antisera specific for CheAL and CheAs (A) or CheAL alone (B) as primary antibodies. (A) lane (1 ), 6XHis-CheAL; lane (2), 6XHis-CheAs; lane (3), 6XHis-CheAL incubated at room temperature for 5 days at room temperature. (B) Lane (1 ), 6XHis-CheAL; lane (2), 6XHis-CheAs, lane (3), 6XHis-CheAL incubated at room temperature for 5 days to permit degradation and reabsorbed over His-bind resin (Novagen) to enrich for P1 containing fragments. The image was derived by scanning the immunoblots with a Hewlett-Packard Scanjet model lie. r 76 CheA(252) REKTTRSNESTSIRVAVEKVDQLINLVGELVITQSMLAQRSSELDPVNHG Gilbrat EEEEECCCCCCEEEEHHHHHHHHHHHHHHHHHHHHHHHCCCCCCCHECCC Levin HHCCCCTTCCHCHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCCCCCCCTT DPM HHHCCTTTCTCCEEEHEHHEHHEECEECHEEEEEEHHHHHTCCCTCCTTC SOPMA CCCEECCCCCCCCCEHHHHHCCCCCCCCCCCCCHHHHCCCCCHHHEEECC Consens us HHCCCCCCCCCCEEEHHHHHHHHHCHHCHHHHHHHHHHCCCCCCCCCCCC CheA DLITSMGQLQRNARDLQESVMSIRMMPMEYVFSRYPRLVRDLAGKLGKQV Gilbrat CEEEHHHHHHCCCCCHHHHHHHHHCCCEEEECCCCCHHHHHHHHHHHHHH Levin CHEEESCCCCTTCCHHHHHHHHHCCCCHHHHHHCCHHHHHHHHHHCCTHH DPM CCECCCCCHHCTHHCHHHEEEEEHHHCHHEEECCCCCEEHHHCCCCCCHE SOPMA CHHHHHCCHHHHHHHHHHHHHHHHCCCCCCCCCCCCHHHHHHHHHHHHHH Consensus CHEEHCCCHHCTCCCHHHHHHHHHCCCHHEECCCCCHHHHHHHHHCCHHH CheA ELTLVGSSTELDKSLIERIIDPLTHLVRNSLDHGIELPEKRLAAGKNSVG Gilbrat EEEEECCCHHHHHHHHHHHHCCHHHEEECCCCCCHHHHHHHHHHCCCCHH Levin EEEEECCCCCHHHHHHHHHHHHHHHHHHHHHCTCCCCCHHHCCCCCCHHH DPM HEEEETCCCCCTCCHEHHEECCCCCEECCCTCCCCCHCHHHHHCTCTTCC SOPMA CEEEECCCCCCCCCCCEEEECCCCCCHHHHHHHCCHHHHHHHHHHHHHHC Consensus EEEEECCCCCCHCCHHHHHHCCCCCEHHCCHCCCCCHCHHHHHCCCCHHC G1 CheA NLILSAEHQGGNICIEV'lll■LNRERILAKAASQGLTVSENMSDDEVA Gilbrat EEEEEHHCCCCCEEEEEECCCCCHHHHHHHHHHHHCCEEEECCCCHHHHH Levin HEEEEHCCTTCCCEEEECCSTSCCCHHHHHHHHHHCCCEEEECCCTSCEE DPM CEEHHHCTTTCCEEEHECCCTCCCTHHHHHHHHHTCCCCETCTTTCCHHH SOPMA CEEECCCCCCCCEEEEEECCCCCCHHHHHHHHHCCCCEEECCCCCCCCHE PhD LLL.EEEEE . LLLLLLLHHHHHHHHHH . LL.E .. LLL . .... Consensus CEEEEHCCCCCCEEEEECCCCCCCHHHHHHHHHHHCCCEEECCCCCCCHH F G2 CheA ML8BsTAEQVTDv,lii~li'1DwKRNIQKMGGHVEIQSKQGTGTTI Gilgrat HHHHCCCCCCCHEEEEECCCCCCEEEEEHCHHHCCCEEEEEEECCCCCEE Levin EEEECSSCCCTHTCHCCTTSEECEEECHHSCHTTTSCCEECESCCCCCEE DPM HHEHHCCCCCHHHECCETTCCCTECEEHHCECTCCCCEHECTTCTCCCEE SOPMA EEEEEECCCCCEEEECCCCCCCCCCHHHHHHHHHHHHHHHHCCCCCCEEE PhD . EEE.LLLLL.HH . ... LLLLLL .. HHH .. LL Consensus HHEHCCCCCCCHEEECCCCCCCCECEEHHCHHHCCCCEHECECCCCCCEE Fig 20: Computer algorithm analysis of secondary structure of the transmitter domain of CheA. Secondary structure is expressed as (E) sheet, (C) coil , (L) loop, (H) helix, (T,S) turn. 77 the G 1 mutant seemingly cannot. Structure Averaging Algorithms In contrast to a number of serine/threonine kinases (Lima, et al. 1996, Owen, et.al. 1995; Emerson, et al. 1995; Owen, et al. 1995), there is little or no structural information about the putative catalytic domain of CheA. In the absence of such structural information, I made use of several structure averaging algorithms to derive information about the possible secondary structure of this domain (Fig 20). One of the algorithms, the PhD method, was used to test the accuracy of the program by comparison of the published NMR structures of the P1 and P2 domains with the predicted structure. The PhD method revealed a level of agreement of about 64%. CheA237-519 The observation that the G2 mutation confers stability and solubility to CheA led to me to believe that this mutation might increase the solubility of the isolated transmitter domain, thereby enabling NMR spectroscopic analysis of this highly insoluble domain. To test this possibility, I generated a truncated CheA protein, by introducing an Nde I site into the linker region between domains P2 and T. This involved two amino acid substitutions (P236H, V237M) which, interestingly, had no effect upon chemotaxis when compared to wild-type (data not shown). Into this construct, I then introduced both the G2 mutation and a premature stop codon in the linker region between T and M. This construction, called pDECheA237-519 (pDE36), was subcloned into pET14b, expressed, purified, and concentrated to 6 78 mg/ml. Solubility, in the context of NMR spectroscopy, refers to the ability of the protein, in high concentration and in an appropriate buffer, to resist precipitation when incubated at room temperature. CheA237-519 was judged soluble by virtue of its ability to stay in solution at both room temperature and at 4°C overnight when at a concentration of 6 mg/ml. The pDE36 construction was sent to Rick Dahlquist (University of Oregon) and is currently being assessed by Megan McEvoy (University of Oregon) as a candidate for NMR spectroscopy. If successful, this would be the first NMR spectroscopic analysis of a histidine kinase transmitter domain. Other, more conservative mutations in G2 are being generated in an attempt to mimic the degree of solubility observed in the G2 mutant. The enhanced stability and solubility of CheAL G2 and its ability to bind ATP with about wild-type affinity, has led us to consider it as a candidate for x-ray crystallography. CHAPTER 5 DISCUSSION AND SUMMARY To identify regions of the CheA transmitter domain required to perform its catalytic activity, I constructed a select group of CheA mutants. I tested each of the mutants for their ability to perform chemotaxis in vivo and to autophosphorylate and/or to be transphosphorylated in vitro. Finally, I measured the ability of a subset of these mutants to bind ATP and/or to facilitate clustering of ternary complexes. Mutagenesis Strategy There are few studies that explore the relationship between structure and function of the putative catalytic regions of two-component histidine kinases. Since the putative catalytic domain of such kinases are insoluble at high concentrations, x-ray crystallographic and NMR spectroscopic studies to determine the structure have not been possible. Therefore, I used mutations as a tool to hypothesize about what this region of CheA might look like and how it functions. The G 1 and G2 regions of CheA resemble the kinase 2 and kinase 1 regions of several bacterial kinases (Matte et al., 1996; Muller and Schulz, 79 80 1993; Reinstein et al, 1990; and Rose et al., 1991 ). Glycines are known to react with the nucleotides AMP and ATP (Yoneya et al., 1989). The consensus kinase 1 motif of E. coli phosphoenolpyruvate kinase (PCK) and adenylate kinase is G-X-X-G-X-G-K-T. This closely resembles the GXGXG motif of the G2 subdomain of CheA. The kinase 2 region contains the consensus motif X-X-X-X-D (L-I-G-D-D in PCK) (Traut, 1994). The conserved "D" forms a Mg2+ binding site in conjunction with one of the hydrophobic "X" residues (Matte et al., 1996). The kinase 2 region of these proteins closely resemble the G1 region of CheA (DDGAG), except that the "D" residue in CheA is N-terminal rather than C-terminal. We decided that the best mutation was one which should exert little structural impact on the region of interest, while changing the ability of that reactive amino acid to interact with either its nucleotide or other functional components of the putative catalytic domain. For instance, changing a phenylalanine to a tyrosine maintains the aromatic nature of the amino acid but adds a reactive hydroxyl group to the aromatic ring. This type of strategy has been used previously to determine the identity of amino acids reactive to the a, ~. and yphosphoryl groups of ATP (Bazaes et al., 1993, 1995). Mutations in ATP binding regions which change glycine to some other residue have repeatedly been shown to alter catalytic function in kinases (Yoneya et al., 1989; Ninfa et al., 1993). We chose to change conserved glycines in the G 1 and G2 subdomains of CheA because this change presented 81 the least possibility of causing structural alterations in the region. A change from glycine to alanine replaces the reactive H with a methyl group but maintains the aliphatic identity of this residue. Despite proline being "aliphatic like", it was eliminated as a possible replacement for the glycine residues because the G1 and G2 regions are predicted to be flexible loops and the rigid backbone of proline could have perturbed the interaction of these putative loops with neighboring structures. Three crystallographic studies and one functional study performed on conserved glycine residues in the nucleotide binding regions of E. coli adenylate kinase reinforce the decision to replace glycine residues with alanines. Muller and Schulz (1993) demonstrated that replacement of glycine-10 with valine hinders the induced fit movements of the active site. This suggests that this type of substitution could perturb structure in an ATP binding loop. Reinstein et al. (1990) clearly showed that a glycine-10 to valine substitution has an increased dissociation constant for substrate and structural alterations in the nucleotide binding site. Rose et al. (1991) showed that a glycine 85 to valine substitution in the kinase-2 region of E. coli adenylate kinase caused a decrease in the thermodynamic stability of the protein. Finally, Yoneya et al. demonstrated that a glycine 15 to alanine or a glycine-20 to alanine substitution in E. coli adenylate kinase decreased the affinity of the kinase-1 region for AMP and ATP. Chemotaxis Requires Intact G1. F. and G2 Subdomains The mutations 422GA, 455FY/459FY, 455FW/459FW, 470GA/472GA/474GA and ~(468-476), located within the highly conserved G1, 82 F, and G2 regions of the histidine kinase CheA, each eliminated the ability of cells to perform chemotaxis. The G2 mutation 470GK also eliminates chemotaxis (Oosawa et al., 1988). Mutations in G1, F, and G2 in other histidine kinases yield similar results. For example, a mutation in the G2 region of NRII reduces the ability of cells to activate expression of glnA (Ninfa et al., 1993). Similarly, a mutation that deletes the F and G2 regions of DctB eliminates the ability of that protein to functon in vivo (Giblin et al., 1996). Autophosphorylation Requires Intact G1, F. and G2 Subdomains To determine why these mutant alleles failed to support chemotaxis in vivo, I measured the ability of each purified mutant protein to autophosphorylate. In contrast to wild-type CheAL' none of the putative catalytic mutants tested autophosphorylated when incubated in the presence of [y-32P]ATP for up to 30 minutes. Other researchers also have shown that mutations in the conserved subdomains inhibit the ability of CheA and other histidine kinases to autophosphorylate. Mutations which truncate the catalytic domain at either the 503'rd position or change residue G470 to a K eliminate the ability to autophosphorylate (Wolfe and Stewart, 1993; Swanson et al., 1994). Similarly, mutations that truncate the F and G2 regions of DctB (Giblin et al., 1996) and those that alter the G2 domain of NRII severely perturb the ability of these proteins to autophosphorylate (Ninfa et al., 1993). The CheALG2, CheAL~G2, and CheALG1/G2 mutants were more stable than the wild-type CheAL protein. In contrast, the CheAL G1 mutant was less 83 stable than wild-type CheAL. Thus, the lack of stability in these mutants could account for the inability of CheAL G 1 to autophosphorylate and mutant cells expressing this protein to perform chemotaxis. In the case of the CheAL G 1 mutant however, this is unlikely because CheA8 can mediate transphosphorylation of CheAL G 1. The same cannot be said for the F mutants. In contrast to CheAL G 1 and in particular CheAL G2 and CheALLiG2 the F mutants were very unstable in vitro. The lack of autophosphorylation may be an artifact of such instability. Since purification of CheA and variants of CheA involves the absorption of a total cell lysate of BL21 (DE3), an E. coli strain, phosphatases to CheA such as CheB or CheY could have co-precipitated with CheA in the eluate, despite stringent washing of the Ni2+ absorption column. Autophosphorylation experiments performed with fresh preparations of wild-typeCheA, however, revealed no radiolabelled bands corresponding to the molecular weights of CheB or CheY. Similarly, no contaminating bands were ever observed upon SOS-PAGE analysis of freshly purified preparations of CheA and variants of CheA. Despite these observations, it is possible that very small quantities of CheY and CheB contaminants were in the purified preparations of CheA and CheA variants. The level of disparity of phosphorylation between wild-type CheA and CheA variants, however, demonstrates that whatever influence the possible contaminants had on the ability of the CheA variants to 84 autophosphorylate was minimal. Thus, we conclude that the inability of G1 and G2 mutants to perform chemotaxis likely results from their inability to autophosphorylate. G1 Residue G422 Plays an Intimate Role in Binding of ATP: G2 Residues G470, G472, and G474 do not To autophosphorylate, CheAL must first bind ATP. Therefore, quantitative nucleotide binding studies were performed on the G 1 , G2. G 1/G2 and ~G2 mutants of CheAL. Intriguingly, CheAG2 bound TNP-ADP about as well as wild type. In contrast. CheAL G1 bound TNP-ADP poorly. However, we conclude that G 1 is intimately involved in the binding of adenine-containing nucleotides. Conversely, we can conclude that G2 is not. Evidence exists, however, that this latter region may play a secondary role in such binding. The CheAL ~G2 mutant has had all of the conserved residues removed, including those flanking the conserved glycines. In contrast. the CheALG1/G2 mutant retains the conserved residues flanking G2. If the G1 region was solely responsible for binding TNP-ADP, then one might expect that the level of nucleotide binding in the ~G2 region would approximate that observed in the G2 mutant. In addition, the double CheAL G1/G2 mutant might also be expected to bind nucleotide with about the same affinity as CheAL G 1. This is not the case. Both CheAL~G2 and CheALG1/G2 bound TNP-ADP with a 85 lower affinity than predicted. These reduced binding affinities suggest two possibilities. Either the G2 region contains other conserved residues (perha·ps those flanking it) which could be involved in nucleotide binding or these mutations cause structural perturbations by not allowing nucleotide-binding residues in G2 to interact with the remaining nucleotide-binding residues in G 1. Secondary structure prediction methods (see below) suggest that the in-frame removal of nine amino acids in the CheAL ~G2 mutant would eliminate a predicted G2 loop structure and render the region from F through the the G2 region a helix. This suggests that a structural perturbation and the removal of critical nucleotide-binding residues may be responsible for the reduced nucleotide-binding observed in this mutant. The interpretation of these results must take into account that the affinity of CheA for TNP-ADP is significantly greater than the affinity of CheA for ADP or ATP (K 5 µM, 50 µM, or 300 µM respectively). This fact must be taken into 0 account when considering this data. In the absence of ATP competition binding experiments, we cannot draw any conclusions concerning binding affinity of these mutants for nucleotide. G2 Residues G470, G472. and/or G474 Play a Role in either the Docking of P1 and/or the Subsequent Phosphotransfer: G1 Residue G422 and F Residues F455 and F459 do not To autophosphorylate, CheA must facilitate the interaction between bound ATP, the site of phosphorylation, and phospho-transfer. The P1 domain 86 receives its phosphate from the catalytic site. Autophosphorylation experiments demonstrated that the catalytic domains of the unstable G 1 and F mutants efficiently transphosphorylated a 30 kDa degradation product comprised of the P1 and P2 domains even though full length protein could not autophosphorylate. Since the G 1 and F mutant proteins transphosphorylate the P1-P2 fragment, the ability of these proteins to interact with P1 must not be significantly affected. In contrast, the G2 mutant could not transphosphorylate the degradation product. The apparent inability of this mutant protein to transphosphorylate the 30 kDa degradation product could be explained if the G2 mutation conferred gross structural defects. This is unlikely, however, since CheAL G2 binds ATP as efficiently as wild-type CheA. Thus, the G2 region may play a role in docking of the P1 domain to the catalytic site. Alternatively, it could function in subsequent phosphotransfer. At present, I cannot distinguish between these two possibilities. However, the fact that G2 seemingly is not required for nucleotide binding yet seemingly is necessary for subsequent events suggests that G1 and G2 each play different roles in the catalytic process. Clustering of Ternary Complexes Requires an Intact G2 Subdomain, but not a G 1 Subdomain Intriguingly, mutants defective in G2 do not form polar aggregates (J. Maddock, personal communication). The lack of aggregates resembles that of mutants entirely lacking CheA, CheW, or MCPs. The simplest explanation for these results is that G2 does not form ternary complexes. At present, this is only 87 a hypothesis. Careful in vitro binding studies will be required to establish whether CheAL G2 can couple into a ternary complex. In contrast, mutants in the G1 subdomain do form aggregates. Therefore, the G1 mutant must form ternary complexes. Ternary complex formation in vivo must not be closely associated with the ability to bind ATP. The G1 mutant binds TNP-ADP poorly yet forms ternary complexes, while the G2 mutant binds TNP-ADP well yet does not form ternary complexes. These in vivo results fit well with the observation by Gegner et al. (1992) that ATP is not necessary for ternary complex formation in vitro. These observations could be explained if the G2 subdomain is one of the sites through which the ternary complex affects the hydrolysis of ATP and the subsequent phosphotransfer by bringing G2 close to G 1 to form a functional catalytic site. If G2 is a docking site for the P1 domain, then this arrangement would make sense. This is only speculation, however, and there could be alternate explanations for these results. This hypothesis could be tested, in part, by performing studies which quantitate the binding affinity of the 30 kDa P1/P2 domains for each of the mutants and comparing that to the binding affinity of the P1/P2 domains for wild-type CheA. An interesting experiment might be to do the binding studies in a ternary complex composed of each catalytic mutant, CheW, and chemoreceptor cytoplasmic domains which have their methylation sites altered to place the ternary complex in the "on" or "off " states. If G2 is a docking site for the P1 domain, then a chemoreceptor in the "off" state might have a lower affinity for the P1 domain since the docking site would be 88 sequestered. G1 and G2 Function in cis CheA functions as a dimer (Surrette et al., 1996). The P1 domains of each protomer obtain their phosphate in trans from their dimer partner (Wolfe and Stewart, 1993). There are two possible ways in which the G1, F, and G2 subdomains of the transmitter domains in a CheA dimer could function. First, they could function in trans. That is, they could each contribute subdomains to form one functional catalytic unit. Second, they could function in cis as structurally independent catalytic subunits. I performed a series of in vitro complementation experiments in which purified preparations of several CheA catalytic mutant proteins were mixed together and their respective levels of phosphorylation assayed in the presence of [y-32P]ATP. Additionally, I performed a series of in vivo experiments in which various catalytic mutants of CheA , placed in single copy on the chromosome, were complemented with a wild-type catalytic domain provided by CheA . 5 In vitro, the short form of CheA transphosphorylated all of the CheAL catalytic domain. The level of transphosphorylation by CheA5 of the CheALG1 and CheAL G2 mutants was about 6% that of wild-type CheAL autophosphorylated under the same conditions. In contrast, the level of transphosphorylation of CheAL G 1/G2 or CheAL ~G2 was about 1% that of wild-type CheAL. In vivo, CheA5 complemented only variants mutated in G1 or 89 G2 alone. It could not complement the double mutation (CheAL G 1/G2) or the mutation in which the entire G2 region has been removed in-frame. The inability of these latter mutants to be complemented in vivo is likely to be due to the low level of CheA -mediated phosphorylation observed in vitro. The 8 disparity between the levels of phosphorylation required for CheA8-mediated chemotaxis and the wild-type levels of phosphorylation in vivo seems to indicate that a minimum threshold of phosphorylation is required to drive chemotaxis. Evidence exists for such a phenomenon: a P1 deletion mutant can be complemented in trans only if P1 is provided in excess (Garzon and Parkinson, 1996). If the level of phosphorylation drops below that minimum threshold, then the efficiency of phospho-transfer to CheY, even when the mutant protein is bound into the ternary complex, is too low to mediate chemotaxis. If the level of phospho-CheY is too low to overcome spontaneous degradation or CheZ-mediated dephosphorylation, the cells will rarely tumble. The predominately smooth swimming behavior of these mutants has been confirmed by light microscopy under a variety of induction conditions. In vivo, CheAL G1 and CheAL G2 could not complement each other. In vitro, the combination of CheAL G 1 and CheAL G2 resulted in a poorly phosphorylated product that migrated slightly slower than the 78 kDa full length wild-type CheA protein. For several reasons, it seems unlikely that this band corresponds to a phosphorylated form of either CheAL G 1 90 or CheAL G2. First, the phosphorylated forms of CheAL G1 and CheAL G2 migrated with the same mobility as wild-type CheAL. Second, after cleavage with thrombin, no 6XHis-CheA variant was detected by immunoblot analysis. Third, we detected no phosphorylation of the 6XHis tag itself when fused to either of the non-phosphorylatable proteins, CheAs, or CheA48L HQ. These results are consistent with a mechanism by which the G1 and G2 subdomains within the TR of one protomer of a CheA dimer form a unit that mediates transphosphorylation of the other protomer within that dimer. This model easily reconciles with that of Tawa and Stewart (1994) who demonstrated that at least one residue (i.e., 337P) in the TL of the protomer that becomes phosphorylated helps to determine the binding affinity of CheA for ATP. One can envision a model in which the TL of one protomer interacts with the TR of the second protomer within the dimer to modulate ATP-binding and hydrolysis. Secondary Structure Prediction and Model for the Structure of the Catalytic Domain of the Histidine Kinase CheA Secondary structure algorithms, while far from perfect as a replacement for structural visualization methods such as NMR spectroscopy and x-ray-crystallography, can be helpful when trying to correlate data derived from in vitro and in vivo analysis to functional models. With this kind of knowledge in hand, an initial test can be performed which allows a researcher to generalize 91 about what the secondary structure of any particular protein may look like. As an initial analysis, these algorithms take short stretches of amino acids (approx. · 30 aa or smaller) and attempt to find similar structures within proteins that have already been crystallized. The protein data bank (PDB) contains the three dimensional structures of several hundred proteins. Many proteins are highly flexible, assuming different shapes in the presence or absence of antigen, ligands, metals, or nucleotides. This type of structural mobility has been observed when crystallographic structures were compared between different binding states. When protein motifs on file in the PDB are found to be homologous to the protein being analyzed, the predictions derived from algorithms such as these can be quite accurate. Using the CheA primary amino acid sequence as a template, I used several algorithms in an attempt to derive secondary structure information over the entire protein. Since it is the focus of this study, the putative ATP binding region of transmitter domain was of particular interest. To predict the secondary structure of this region, 5 different algorithms were utilized (Gilbrat et al., 1987, Levin et al., 1986, Deleage and Roux, 1987, Rost and Sander, 1994, Geourjon and Deleage, 1995). Based upon such factors as sequence homology, side chain associations, and solvent accessibility, a consensus of the methods predicts that G1, F, and G2 resemble coils and turns (loops). In a separate secondary structure analysis of this region using only the PhD prediction method, these regions are predicted to be unstructured loops, consistent with observations of the crystallographic structure of several different eukaryotic ATP binding regions. F is flanked by alpha 92 helices and sheets that suggest a bend, As a reference for the accuracy of prediction of each of these methods, the predictions were compared to the deduced structure of the N-terminal domains P1 and P2 of CheA. These domains have undergone rigorous analysis by NMR and their 2° and 3° structures are known. On the basis of these structure-averaging algorithms, we propose that the catalytic region is best represented as a cleft in which the F subdomain lies at the base and acts as a flexible turn or hinge flanked by two short alpha helices (Fig 20). This arrangement would serve to approximate G1 and G2, predicted to be unstructured ATP-binding loops (Rossman and Liljas, 1974). The binding of ligand by the chemoreceptor would initiate a cascade through the ternary complex with the end result being the movement of the G1 and G2 loops apart, rendering CheA in an inactive conformation. Release of ligand would remove this repression permitting the catalytic domain to return to the active conformation in which G 1 and G2 properly align relative to one another. Significance of Other Histidine Kinases that lack Some of the Conserved Regions In undertaking this study, I was interested in defining the structure and function of a little understood region of the histidine kinase CheA. Although this study was concerned with the putative catalytic domain of one protein, the motifs which define CheA as a histidine kinase are shared by many other proteins. Over 100 two-component histidine kinases have been identified in bacteria. Several have also been identified in eukaryotes and plants (Ota and 93 G2 Fig 21. Model for the structure of the G1 - G2 region of the transmitter domain of CheA. The catalytic site is represented in this model as two ATP binding loops (G1 and G2) separated by two beta-pleated sheets and a bend at the F box. The site is in an "active" form when the two loops are allowed to approximate into a single, functional catalytic site. The catalytic site is in an "inactive" form upon dissociation of G 1 and G2. 94 Varshavsky, 1993 Chang, et.al., 1993) suggesting that this type of kinase is not unique to either prokaryotes or even to single-celled organisms. Structural information about the catalytic region of any histidine kinase is a much desired goal of many different groups of researchers. In addition to its obvious intellectual value, catalytic domain structural information has enormous potential forpractical value as well. Since the general catalytic motif of histidine kinases has mostly been observed in bacteria, a number of which are pathogenic to humans and animals, it has important potential as a therapeutic target. Several companies, including one our lab has been in collaboration with, are interested in designing drugs which would interfere with the function of this domain. Not all putative catalytic domains of histidine kinases have the N, G1, F, G2 arrangement of all four catalytic subdomains. The nitrate/nitrite sensor protein NarX, for instance, lacks both the F and G2 regions. Yet, it mediates phosphorylation and dephosphorylation of the response regulator NarL (Schroder et al., 1994; Chiang et al., 1992). The Bacillus subti/us histidine kinase DegS also lacks the G2 region. In addition, its F subdomain seems to be in the orientation opposite to that of the conventional histidine kinase motif. It too can mediate phosphorylation of a response regulator DegU (Hanner et al., 1988). The absence of catalytic motifs in these other histidine kinases should be noted since they may help us understand the function of particular subdomains. Clearly, further structure-function investigations such as this are necessary. WORKS CITED 1. Adler, J. 1966. Chemotaxis in bacteria. Science 153:708-16. 2. Adler, J., G. L. Hazelbauer and M. M. Dahl. 1973. Chemotaxis toward sugars in Escherichia coli. Journal of Bacteriology 115:824-47. 3. Alley, M. R., J. R. Maddock and L. Shapiro. 1993. Requirement of the carboxyl terminus of a bacterial chemoreceptor for its targeted proteolysis [see comments]. Science 259:1754-7. 4. Alphen, W. V. and B. Lugtenberg. 1977. Influence of osmolarity of the growth medium on the outer membrane protein pattern of Escherichia coli. Journal of Bacteriology 131 :623-30. 5. Ames, P. and J.S. Parkinson. 1994. Constitutively signaling fragments of Tsr, the Eschericia coli serine receptor. Journal of Bacteriology 176: 6340-48. 6. Atkinson, M.R. and A.J. Ninfa. 1993. Mutational analysis of the bacterial signal-transducing protein kinase/phosphatase nitrogen regulator II (NRII or NtrB). Journal of Bacteriology 175(21): 7016-23. 7. Bazeas, S., H. Goldie, E. Cardemil, and A.M. Jabalquinto. 1995. Identification of reactive lysines in phosphoenolpyruvate carboxylkinases from Eschericia coli and Saccharomyces cereviseae. FEBS Letters 360(2): 207-10. 8. Bazeas, S., R. Silva, H. Goldie, E. Cardemil, and A.M. Jabalquinto. 1993. Reactivity of cysteinyl, arginyl, and lysyl residues of Eschericia coli phosphoenolpyruvate carboxykinase against group-specific chemical reagents. Journal of Protein Chemistry 12(5): 571-77. 9. Bellsolell, L., J. Prieto, L. Serrano, and M. Coll. 1994. Magnesium binding to the bacterial chemotaxis protein CheY results in large conformational changes involving its functional surface. Journal of Molecular 95 96 Biology 238: 489-95. 10. Berg, H. C. 1988. A physicist looks at bacterial chemotaxis. [Review]: Cold Spring Harbor Symposia on Quantitative Biology 1 : 1-9. 11. Berg, H. C. and E. M. Purcell. 1977. Physics of chemoreception.Biophysical Journal 20: 193-219. 12. Blair, D. F. 1993. Testing the limits of flagellar motors. Biophysical Journal 65:1751-2. 13. Blair, D. F. 1995. How bacteria sense and swim. [Review]. Annual Review of Microbiology 49:489-522. 14. Blair, D. F. and H. C. Berg. 1990. The MotA protein of E. coli is a proton-conducting component of the flagellar motor. Cell 60:439-49. 15. Blair, D. F. and H. C. Berg. 1991. Mutations in the MotA protein of Escherichia coli reveal domains critical for proton conduction. Journal of Molecular Biology 221: 1433-42. 16. Blair, D. F. and S. K. Dutcher. 1992. Flagella in prokaryotes and lower eukaryotes. [Review]. Current Opinion in Genetics & Development 2:756-67. 17. Blat, Y. and M. Eisenbach. 1994. Phosphorylation-dependent binding of the chemotaxis signal molecule CheY to its phosphatase, CheZ. Biochemistry 33: 902-6. 18. Block, S. M. and H. C. Berg. 1984. Successive incorporation of force-generating units in the bacterial rotary motor. Nature 309:470-2. 19. Block, S.M., D.F. Blair, and HC Berg. 1989. Compliance of bacterial flagella measured with optical tweezers. Nature 338: 514-18. 20. Block, S. M., K. A. Fahrner and H. C. Berg. 1991. Visualization of bacterial flagella by video-enhanced light microscopy. Journal of Bacteriology 173:933-6. 21. Block, S. M., J. E. Segall and H. C. Berg. 1983. Adaptation kinetics in bacterial chemotaxis. Journal of Bacteriology 154:312-23. 22. Borczuk, A., A. Staub and J. Stock. 1986. Demethylation of bacterial chemoreceptors is inhibited by attractant stimuli in the complete 97 absence of the regulatory domain of the demethylating enzyme. Biochemical & Biophysical Research Communications 141 :918-23. 23. Berkovich, K. A., N. Kaplan, J. F. Hess and M. I. Simon. 1989. Transmembrane signal transduction in bacterial chemotaxis involves ligand-dependent activation of phosphate group transfer. Proceedings of the National Academy of Sciences of the United States of America 86: 1208-12. 24. Berkovich, K. A. and M. I. Simon. 1990. The dynamics of protein phosphorylation in bacterial chemotaxis. Cell 63: 1339-48. 25. Berkovich, K. A. and M. I. Simon. 1991. Coupling of receptor function to phosphate-transfer reactions in bacterial chemotaxis. Methods in Enzymology 200:205-14. 26. Bourret, R. B., K. A. Berkovich and M. I. Simon. 1991. Signal transduction pathways involving protein phosphorylation in prokaryotes. [Review]. Annual Review of Biochemistry 60:401-41. 27. Bourret, R. B., J. Davagnino and M. I. Simon. 1993. The carboxy-terminal portion of the CheA kinase mediates regulation of autophosphorylation by transducer and CheW. Journal of Bacteriology 175:2097-101. 28. Bourret, R. B., J. F. Hess, K. A. Berkovich, A. A. Pakula and M. I. Simon. 1989. Protein phosphorylation in chemotaxis and two-component regulatory systems of bacteria. [Review]. Journal of Biological Chemistry 264:7085-8. 29. Burbulys, D., K. A. Trach and J. A. Hoch. 1991. Initiation of sporulation in B. subtilis is controlled by a multicomponent phosphorelay. Cell 64:545-52. 30. Chang, C., S. F. Kwok, A. B. Bleecker and E. M. Meyerowitz. 1993. Arabidopsis ethylene-response gene ETR1: similarity of product to two-component regulators [see comments]. Science 262:539-44. 31. Chiang, R.C., C.R. Cavicchioli, and R.P. Gunsalus. 1992. Identification and characterization of NarQ, a second nitrate sensor for nitrate-dependent gene regulation in Eschericia coli. Molecular Microbiology 6:1913-23. 32. Clarke, S. and D. J. Koshland. 1979. The effect of protein modification reagents on the chemotactic response in Salmonella typhimurium. 98 Canadian Journal of Biochemistry 57:1331-6. 33. Clarke, S. and D. J. Koshland. 1979. Membrane receptors for aspartate and serine in bacterial chemotaxis. Journal of Biological Chemistry 254:9695-702. 34. Cochran, A. G. and P. S. Kim. 1996. Imitation of Escherichia coli aspartate receptor signaling in engineered dimers of the cytoplasmic domain. Science 271: 1113-6. 35. Dahl, M.K., W. Boos, and M.D. Manson. 1989. Evolution of chemotactic-signal transducers in enteric bacteria. Journal of Bacteriology 171: 2361-71. 36. Deleage, G. and B Roux. 1987. An algorithm for protein secondary structure prediction based on class prediction. Protein Engineering 1(4): 289-94. 37. DeMot, R. and J. Vanderleyden. 1994. The C-terminal sequence conservation between OmpA-related outer membrane proteins and MotB suggests a common function in both gram-positive and gram-negative bacteria, possibly in the interaction of these domains with peptidoglycan [letter] [see comments]. Molecular Microbiology 12:333-4. 38. DeMot, R. and J. Vanderleyden. 1994. A conserved surface-exposed domain in major outer membrane proteins of pathogenic Pseudomonas and Branhamella species shares sequence homology with the calcium-binding repeats of the eukaryotic extracellular matrix protein thrombospondin [letter]. Molecular Microbiology 13:379-80. 39. Dean, G. E., R. M. Macnab, J. Stader, P. Matsumura and C. Burks. 1984. Gene sequence and predicted amino acid sequence of the motA protein, a membrane-associated protein required for flagellar rotation in Escherichia coli. Journal of Bacteriology 159:991-9. 40. Donis-Keller, H., A. M. Maxam and W. Gilbert. 1977. Mapping adenines, guanines, and pyrimidines in RNA. Nucleic Acids Research 4:2527-38. 41. Dons, L., J.E. Olsen, and O.F. Rasmussen. 1994. Characterization of two putative Listeria monocytogenes genes encoding polypeptides homologous to the sensor protein CheA and the response regulator CheY. DNA Sequence 4(5): 301-11. 99 42. Emerson, S.D., D.S. Waugh, J.E. Scheffler, K.L. Tsao, K.M. Prinzo, and D.C. Fry. 1994. Chemical shift assignments and folding topology of the Ras-binding domain of human Raf-1 as determined by heteronuclear three-dimensional NMR spectroscopy. Biochemistry 33: 7745-52. 43. Emerson, S.D., V.S. Madison, R.E. Palermo, D.S. Waugh, J.E. Scheffler, K.L. Tsao, S.E. Kiefer, S.P. Liu, and D.C. Fry. 1995. Solution structure of the Ras-binding domain of c-Raf-1 and identification of its interaction surface. Biochemistry 34: 6911-6918. 44. Falke, J. J., D. F. Blair, T. J. Silhavy and R. Schmitt. 1995. BLAST 1995: international conference on bacterial locomotion and signal transduction. Molecular Microbiology 16:1037-50. 45. Forst, S. and M. Inouye. 1988. Environmentally regulated gene expression for membrane proteins in Escherichia coli. [Review]. Annual Review of Cell Biology 4:21-42. 46. Francis, N. R., V. M. lrikura, S. Yamaguchi, D. J. DeRosier and R. M. Macnab. 1992. Localization of the Salmonella typhimurium flagellar switch protein FliG to the cytoplasmic M-ring face of the basal body. Proceedings of the National Academy of Sciences of the United States of America 89:6304-8. 47. Francis, N.R., G.E. Sosinsky, D. Thomas, and D.J. DeRosier. 1994. Isolation, characterization and structure of bacterial flagellar motors containing the switch complex. Journal of Molecular Biology 235: 1261-70. 48. Frank, R., M. Trosin, A.G. Tomasselli, L. Noda, R.L. Krauth-Siegel, and R.H. Schirmer. 1986. Mitochondrial adenylate kinase (AK2) from bovine heart. The complete primary structure. European Journal of Biochemistry 154: 205-11 . 49. Fuhrer, D.K. and G.W. Ordal. 1991. Bacillus subtilus cheN, a homolog of cheA, the central regulator of chemotaxis in Eschericia coli. Journal of Bacteriology 173: 7443-48. 50. Garzon, A. and J.S. Parkinson. 1996. Chemotactic signalling by the P1 phosphorylation domain liberated from the CheA histidine kinase of Eschericia coli. Journal of Bacteriology 178: 6752-58. 51. Gegner, J. A. and F. W. Dahlquist. 1991. Signal transduction in bacteria: CheW forms a reversible complex with the protein kinase CheA. 100 Proceedings of the National Academy of Sciences of the United States of America 88:750-4. 52. Gegner, J. A., D. R. Graham, A. F. Roth and F. W. Dahlquist. 1992. Assembly of an MCP receptor, CheW, and kinase CheA complex in the bacterial chemotaxis signal transduction pathway. Cell 70:975-82. 53. Geourjon, C. and G. Deleage. 1995. SOPMA: significant improvements in protein secondary structure prediction by consensus prediction from multiple sequence alignments. Computer Applications in Bioscience 11 (6): 681-84. 54. Giblin, L., J. Archdeacon, and F. O'Gara. 1996. Modular structure of the Rhizobium meliloti DctB protein. FEMS Microbiology Letters 139: 19-25. 55. Gilbrat, J.F., J. Garnier, and B. Robson. 1987. Further developments of protein secondary structure prediction using imformation theory. New parameters and consideration of residue pairs. Journal of Molecular Biology 198(3): 425-43. 56. Greek, M., J. Platzer, V. Sourjik, and R. Schmitt. 1995. Analysis of the chemotaxis operon in Rhizobium meliloti. Molecular Microbiology 15(6): 898-1000. 57. Hanks, S. K., A. M. Quinn and T. Hunter. 1988. The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. [Review]. Science 241 :42-52. 58. Harlow, E. and D. Lane. 1988. Antibodies: a laboratory manual. Cold Spring Harbor Laboratories Cold Spring Harbor, N.Y. 59. Hedblom, M. L. and J. Adler. 1980. Genetic and biochemical properties of Escherichia coli mutants with defects in serine chemotaxis. Journal of Bacteriology 144:1048-60. 60. Hedblom, M. L. and J. Adler. 1983. Chemotactic response of Escherichia coli to chemically synthesized amino acids. Journal of Bacteriology 155:1463-6. 61. Henner, D.J., M. Yang, and E. Ferrari. 1988. Localization of Bacillus subtilus sacU(Hy) mutations to two linked genes with similarities to the conserved prokaryotic family of two-component signalling systems. Journal of Bacteriology 170: 51 02-9. 101 62. Herschlog, D. and W.P. Jencks. 1990. Catalysis of the hydrolysis of phosphorylated pyridines by Mg(OH)+: a possible model for enzymatic phosphorylal transfer. Biochemistry 29: 5172-79. 63. Hess, J. F., R. B. Bourret and M. I. Simon. 1988. Histidine phosphorylation and phosphoryl group transfer in bacterial chemotaxis. Nature 336: 139-43. 64. Hess, J. F., R. B. Bourret and M. I. Simon. 1991. Phosphorylation assays for proteins of the two-component regulatory system controlling chemotaxis in Escherichia coli. Methods in Enzymology 200:188-204. 65. Hess, J. F., K. Oosawa, N. Kaplan and M. I. Simon. 1988. Phosphorylation of three proteins in the signaling pathway of bacterial chemotaxis. Cell 53:79-87. 66. Hess, J. F., K. Oosawa, P. Matsumura and M. I. Simon. 1987. Protein phosphorylation is involved in bacterial chemotaxis. Proceedings of the National Academy of Sciences of the United States of America 84:7609-13. 67. Hiratsuka, T. 1982. New fluorescent analogs of cAMP and cGMP available as substrates for cyclic nucleotide phosphodiesterase. Journal of Biological Chemistry 257(22): 13354-58. 68. Hoffmann, A. and R. G. Roeder. 1991. Purification of his-tagged proteins in non-denaturing conditions suggests a convenient method for protein interaction studies. Nucleic Acids Research 19:6337-8. 69. Hotani, H. and H. Kagawa. 1974. Unidirectional melting of Salmonella flagella in vitro. Journal of Molecular Biology 90:169-80. 70. Jin, T. and M. Inouye. 1993. Ligand binding to the receptor domain regulates the ratio of kinase to phosphatase activities of the signaling domain of the hybrid Eschericia coli transmembrane receptor, Taz1. Journal of Molecular Biology 232: 484-492. 71. Kamiya, R., H. Hotani and S. Asakura. 1982. Polymorphic transition in bacterial flagella. [Review]. Symposia of the Society for Experimental Biology 35:53-76. 72. Kishi, F. , V. Tanizawa, and A. Nakazawa. 1987. Isolation and characterization of two types of cDNA for mitochondrial adenylate kinase and their characterization in Eschericia coli. 262: 11785-9. 102 73. Kofoid, E. C. and J. S. Parkinson. 1988. Transmitter and receiver modules in bacterial signaling proteins. Proceedings of the National Academy of Sciences of the United States of America 85:4981-5. 74. Kofoid, E. C. and J. S. Parkinson. 1991. Tandem translation starts in the cheA locus of Escherichia coli. Journal of Bacteriology 173:2116-9. 75. Komeda, V., K. Kutsukake and T. lino. 1980. Definition of additional flagellar genes in Escherichia coli K12. Genetics 94:277-90. 76. Koshland, D. J. 1993. The two-component pathway comes to eukaryotes [comment]. Science 262: 532. 77. Kunkel T.A. 1985. Rapid and efficient site-specific mutagenesis without phenotype selection. Proceedings of the national Academy of Sciences U.S.A. 82: 488-92. 78. Kutsukake, K. and T. lino. 1985. Refined genetic analysis of the region II che mutants in Salmonella typhimurium. Molecular & General Genetics 199:406-9. 79. Kutsukake, K. and T. lino. 1994. Role of the FliA-FlgM regulatory system on the transcriptional control of the flagellar regulon and flagellar formation in Salmonella typhimurium. Journal of Bacteriology 176:3598-605. 80. Kuwajima, G., J. Asaka, T. Fujiwara, K. Node, and E. Kondo. 1986. Sequence of the hag gene encoding flagellin of Eschericia coli. Journal of Bacteriology 168: 1479-83. 81. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-5. 82. Laemmli, U. K., E. Molbert, M. Showe and E. Kellenberger. 1970. Form-determining function of the genes required for the assembly of the head of bacteriophage T4. Journal of Molecular Biology 49:99-113. 83. Levin, J.M., B. Robson, and J. Garnier. 1986. An algorithm for secondary structure determination in proteins based on sequence similarity. FESS Letters 205(2): 303-8. 84. Levit, M., V. Liu, M. Surette, and J. Stock. 1996. Active site interference and asymetric activation in the chemotaxis protein histidine kinase CheA. Journal of Biological Chemistry 271: 32057-63. 103 85. Lima, C.D., M.G. Klein, 1.8. Weinstein, and W.A. Hendrickson. 1996. Three-dimensional structure of human protein kinase C interacting protein 1, a member of the HIT family of proteins. Proc. Nat. Acad. Sci. USA 93:5357-5362. 86. Liu, J. D. and J. S. Parkinson. 1989. Role of CheW protein in coupling membrane receptors to the intracellular signaling system of bacterial chemotaxis. Proceedings of the National Academy of Sciences of the United States of America 86:8703-7. 87. Liu, J. D. and J. S. Parkinson. 1991. Genetic evidence for interaction between the CheW and Tsr proteins during chemoreceptor signaling by Escherichia coli. Journal of Bacteriology 173:4941-51 . 88. Lloyd, S. A., H. Tang, X. Wang, S. Billings and D. F. Blair. 1996. Torque generation in the flagellar motor of Escherichia coli: evidence of a direct role for FliG but not for FliM or FliN. Journal of Bacteriology 178:223-31. 89. Lukat, G.S., B.H. Lee, J.M. Mottonen, A.M. Stock, and J.B. Stock. 1991. Roles of the highly conserved aspartate and lysine residues in the response regulator of bacterial chemotaxis. Journal of Biological Chemistry 266: 8348-54. 90. Lupas, A. and J. Stock. 1989. Phosphorylation of an N-terminal regulatory domain activates the CheB methylesterase in bacterial chemotaxis. Journal of Biological Chemistry 264:17337-42. 91. Macnab, R.M. and D.E. Koshland. 1972. The gradient-sensing mechanism in bacterial chemotaxis. Proceedings of the National Academy of Sciences. 69: 2509-12. 92. Macnab, R. M. 1992. Genetics and biogenesis of bacterial flagella. [Review]. Annual Review of Genetics 26:131-58. 93. Maddock, J. R. and L. Shapiro. 1993. Polar location of the chemoreceptor complex in the Escherichia coli cell. Science 259:1717-23. 94. Manson, M. D., V. Blank, G. Brade and C. F. Higgins. 1986. Peptide chemotaxis in E.coli involves the Tap signal transducer and the dipeptide permease. Nature 321 :253-6. 95. Manson, M. D. and M. Kossmann. 1986. Mutations in tar suppress defects in maltose chemotaxis caused by specific malE mutations. Journal of Bacteriology 165:34-40. 104 96. Matte, A., H. Goldie, R.M. Sweet, and L.T.J. Delbaere. 1996. Crystal structure of Eschericia coli phosphoenolpyruvate carboxykinase: A new structural family with the P-loop nucleoside triphosphate hydrolase fold. Journal of Molecular Biology 256: 126-43. 97. McCleary, W.R. and D.R. Zusman. 1990. FrzE of Myxococcus xanthus is homologous to both CheA and CheY of Salmonella typhimurium. Proceeding of the National Academy of Sciences U.S.A. 87: 5898-5902. 98. Morgan, D. G., R. M. Macnab, N. R. Francis and D. J. DeRosier. 1993. Domain organization of the subunit of the Salmonella typhimurium flagellar hook. Journal of Molecular Biology 229:79-84. 99. Morrison, T. B. and J. S. Parkinson. 1994. Liberation of an interaction domain from the phosphotransfer region of CheA, a signaling kinase of Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America 91 :5485-9. 100. Muller, C.W. and G.E. Schulz. 1993. Crystal structures of two mutants of adenylate kinase from Eschericia coli that modify the Gly-loop. Proteins 15: 42-9. 101. Mutoh, N. and M. I. Simon. 1986. Nucleotide sequence corresponding to five chemotaxis genes in Escherichia coli. Journal of Bacteriology 165: 161-6. 102. Ninfa, A. J. 1991. Protein phosphorylation and the regulation of cellular processes by the homologous two-component regulatory systems of bacteria. [Review]. Genetic Engineering 13:39-72. 103. Ninfa, E. G., M. R. Atkinson, E. S. Kamberov and A. J. Ninfa. 1993. Mechanism of autophosphorylation of Escherichia coli nitrogen regulator II (NRII or NtrB): trans-phosphorylation between subunits. Journal of Bacteriology 175:7024-32. 104. Ninfa, E.G. A. Stock, S. Mowbray, and J Stock. 1991. Reconstitution of the bacterial chemotaxis signal transduction system from purified components. Journal of Biological Chemistry 266: 9764-70. 105. Oosawa, K., J. F. Hess and M. I. Simon. 1988. Mutants defective in bacterial chemotaxis show modified protein phosphorylation. Cell 53:89-96. 106. Ota, I. M. and A. Varshavsky. 1993. A yeast protein similar to 105 bacterial two-component regulators [see comments]. Science 262:566-9. 107. Owen, D.J., M.E. Noble, E.F. Garman, A.C. Papageorgiou, and L.N. Johnson. 1995. Two structures of the catalytic domain of phosphorylase kinase: and active protein kinase complexed with substrate analogue and product. Structure 3:476-482. 108. Owen, D.J., A.C. Papageorgiou, E.F. Garman, M.E. Noble, L.N. Johnson. 1995. Expression, purification and crystallisation of phosphorylase kinase catalytic domain. Journal of Molecular Biology 246: 374-381. 109. Park, C. and G. L. Hazelbauer. 1986. Mutation plus amplification of a transducer gene disrupts general chemotactic behavior in Escherichia coli. Journal of Bacteriology 168:1378-83. 11 0. Park, C. and G. L. Hazelbauer. 1986. Mutations specifically affecting ligand interaction of the Trg chemosensory transducer. Journal of Bacteriology 167:101-9. 111. Park, C. and G. L. Hazelbauer. 1986. Transfer of chromosomal mutations to plasmids via Hfr-mediated conduction. Journal of Bacteriology 165:312-4. 112. Parkinson, J. S. and D. F. Blair. 1993. Does E. coli have a nose? [comment]. Science 259:1701-2. 113. Parkinson, J.S. and S.E. Houts. 1982. Isolation and behavior of Eschericia coli deletion mutants lacking chemotaxis functions. Journal of Bacteriology 151 : 106-13. 114. Parkinson, J. S. and E. C. Kofoid. 1992. Communication modules in bacterial signaling proteins. [Review]. Annual Review of Genetics 26:71-112. 115. Reinstein, J., I. Schlichting, and A. Wittinghofer. 1990. Structural! and catalytically important residues in the phosphate binding loop of adenylate kinase of Eschericia coli. Biochemistry 29: 7451-59. 116. Rose, T., P. Glaser, W.K. Surwicz, H.H. Mantsch, J. Reinstein, K. Le Blay, A.M. Gilles, and 0. Barzu. 1991. Structural and functional consequences of amino acid substitutions in the second conserved loop of Eschericia coli adenylate kinase. Journal of Biological Chemistry 266: 23654-59. 117. Rost, B. and C. Sander. 1994. Combining evolutionary information 106 and neural networks fo predict protein secondary structure. Proteins 19(1 }: 55-72. 118. Rossman, M. G. and A. Liljas. 1974. Letter: Recognition of structural domains in globular proteins. Journal of Molecular Biology 85:177-81. 119. Russo, F.D. and T.J. Silhavy. 1993. The essential tension: opposed reactions in bacterial two-component regulatory systems. Trends in Microbiology 1: 306-10. 120. Rudolph, J. and D. Oesterhelt. 1995. Chemotaxis and phototaxis require a CheA histidine kinase in the archeaon Halobacterium salinarium. EMBO Journal 14(4}: 667-73. 121. Sanders, D.A. B.L. Gillece-Castro, A.M. Stock, A.L. Burlingame, and D.E. Koshland. 1989. Identification of the site of phosphorylation of the chemotaxis response regulator protein, CheY. Journal of Biological Chemistry 264: 21770-78. 122. Saraste, M., P.R. Sibbald, and A. Wittinghofer. 1990. The P-loop--a common motif in ATP- and GTP- binding proteins. Trends in Biochemical Science 15: 430-4. 123. Scheffler, J.E., D.S. Waugh, E. Bekesi, S.E. Kiefer, J.E. Losardo, A. Neri, K.M. Prinzo, K.L. Tsao, B. Wegrzynski, S.D. Emerson, and D.C. Fry. 1994. Characterization of a 78-residue fragment of c-raf-1 that comprises a minimal binding domain for the interaction with Ras-GTP. Journal of Biological Chemistry 269: 22340-22346. 124. Schirmer, T. and P.R. Evans. 1990. Structural basis of the allosteric behaviour of phosphofructokinase. Nature 343: 140-5. 125. Shapiro, L. 1993. Protein localization and asymmetry in the bacterial cell. [Review]. Cell 73:841-55. 126. Schroder, I., C.D. Wolin, R. Cavicchioli, and R.P. Gunsalus. 1994. Phosphorylation and dephosphorylation of the NarQ, NarX, and NarL proteins in the nitrate-dependent two-component regulatory system of Eschericia coli. Journal of Bacteriology 176: 4985-92. 127. Schulz, G.E. 1992. Binding of nucleotides by proteins. Current Opinions in Structural Biology 2: 61-7. 128. Shapiro, L. 1995. The bacterial flagellum: from genetic network to 107 complex architecture. [Review]. Cell 80:525-7. 129. Sharp, L. L., J. Zhou and D. F. Blair. 1995. Features of MotA proton channel structure revealed by tryptophan-scanning mutagenesis. Proceedings of the National Academy of Sciences of the United States of America 92:7946-50. 130. Sharp, L. L., J. Zhou and D. F. Blair. 1995. Tryptophan-scanning mutagenesis of MotB, an integral membrane protein essential for flagellar rotation in Escherichia coli. Biochemistry 34:9166-71. 131. Shin, S. and C. Park. 1995. Modulation of flagellar expression in Eschericia coli by acetyl phosphate and the osmoregulator OmpR. Journal of Bacteriology 177: 4696-4702. 132. Shioi, J., R.C. Tribhuwan, S.T. Berg, and B.L. Taylor. 1988. Signal transduction in chemotaxis to oxygen in Eschericia coli and Salmonella typhimurium. Journal of Bacteriology 170: 5507-11. 133. Silhavy, T.J., M.L. Berman, and L.W. Enquist. 1984. Experiments with gene fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 134. Simon, M. I., K. A. Borkovich, R. B. Bourret and J. F. Hess. 1989. Protein phosphorylation in the bacterial chemotaxis system. [Review]. Biochimie 71: 1013-9. 135. Smith, R. A. and J. S. Parkinson. 1980. Overlapping genes at the cheA locus of Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America 77:5370-4. 136. Soderling, T. R. 1990. Protein kinases. Regulation by autoinhibitory domains. [Review]. Journal of Biological Chemistry 265: 1823-6. 137. Stewart, V. and R.S. Rabin. 1995. Dual response refulators interact to control nitrate- and nitrite-responsive gene expression in Eschericia coli. p 233-262. in J.A. Hoch and T.J. Silhavy (ed), Two-Component Signal Transduction. American Society for Microbiology, Washington D.C. 138. Stock, A., T. Chen, D. Welsh, and J. Stock. 1988. CheA protein, a central regulator of bacterial chemotaxis, belongs to a family of proteins that control gene expression in response to changing environmental conditions. Proceedings of the national Academy of Sciences U.S.A. 85: 1403-7. 139. Stock, A. M. and J. B. Stock. 1987. Purification and characterization 108 of the CheZ protein of bacterial chemotaxis. Journal of Bacteriology 169:3301-11. 140. Stock, A. M., D. C. Wylie, J. M. Mottonen, A. N. Lupas, E. G. Ninfa, A. J. Ninfa, C. E. Schutt and J. B. Stock. 1988. Phosphoproteins involved in bacterial signal transduction. Cold Spring Harbor Symposia on Quantitative Biology 1 :49-57. 141. Stock, A.M., J.M. Mottonen, J.B. Stock, and C.E. Schutt. 1989. Three-dimensional structure of CheY, the response regulator of bacterial chemotaxis. Nature 337: 745-49. 142. Stock, J., A. Borczuk, F. Chiou and J. E. Burchenal. 1985. Compensatory mutations in receptor function: a reevaluation of the role of methylation in bacterial chemotaxis. Proceedings of the National Academy of Sciences of the United States of America 82:8364-8. 143. Stock, J., G. Kersulis and D. J. Koshland. 1985. Neither methylating nor demethylating enzymes are required for bacterial chemotaxis. Cell 42:683-90. 144. Stock, A.M., D.E. Koshland, and J.B. Stock. 1985. Homologies between the Salmonella typhimurium CheY protein and proteins involved in the regulation of chemotaxis, membrane protein synthesis, and sporulation. Proceedings of the National Academy of Sciences U.S.A. 82: 7989-93. 145. Stock, A.M., E. Martinez-Hackert, B.F. Rasmussen, A.H. West, J.B. Stock, D. Ainge, and G.A. Petsko. 1993. Structure of the Mg(2+) bound form of CheY and mechanism of phosphoryla transfer in bacterial chemotaxis. Biochemistry 32: 13375-80. 146. Stock, J. and M. Surette. 1994. Bacterial chemotaxis. The motor connection. [Review]. Current Biology 4:143-4. 147. Stock, J., M. Surette, and P. Park. 1994. Chemosensing and signal transduction in bacteria. Current Opinions in Neurobiology 4: 474-80. 148. Stock, J.B., G.S. Lukat, and A.M. Stock. 1991. Bacterial chemotaxis and themolecular logic of intracellular signal transduction networks. Annual Revue of Biophysics and Biophysical Chemistry 20: 109-36. 149. Stock, J. B. and D. J. Koshland. 1981. Changing reactivity of receptor carboxyl groups during bacterial sensing. Journal of Biological Chemistry 256: 10826-33. 109 150. Stock, J. B., A. M. Maderis and D. J. Koshland. 1981. Bacterial chemotaxis in the absence of receptor carboxylmethylation. Cell · 151 . Studier, F. W. 1991 . Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. Journal of Molecular Biology 219:37-44. 152. Studier, F. W. and B. A. Moffatt. 1986. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. Journal of Molecular Biology 189: 113-30. 153. Surette, M.G., M. Levit, Y. Liu, G. Lukat, E.G. Ninfa, A. Ninfa, and J.B. Stock. 1996. Dimerization is required for the activity of the protein histidine kinase CheA that mediate signal transduction in bacterial chemotaxis. Journal of Biological Chemistry 271: 939-45. 154. Swanson, R.V., L.A. Alex, and M.I. Simon. 1994. Histidine and aspartate phosphorylation: two-component systems and the limits of homology. Trends in Biochemical Science 19(11 ): 495-90. 155. Swanson, R. V., R. B. Bourret and M. I. Simon. 1993. Intermolecular complementation of the kinase activity of CheA. Molecular Microbiology 8:435-41 . 156. Swanson, R. V ., S. C. Schuster and M. I. Simon. 1993. Expression of CheA fragments which define domains encoding kinase, phosphotransfer, and CheY binding activities. Biochemistry 32:7623-9. 157. Tang, H., S. Billings, X. Wang, L. Sharp and D. F. Blair. 1995. Regulated underexpression and overexpression of the FliN protein of Escherichia coli and evidence for an interaction between FliN and FliM in the flagellar motor. Journal of Bacteriology 177:3496-503. 158. Tang, H. and D. F. Blair. 1995. Regulated underexpression of the FliM protein of Escherichia coli and evidence for a location in the flagellar motor distinct from the MotA/MotB torque generators. Journal of Bacteriology 177:3485-95. 159. Tang, H., T. F. Braun and D. F. Blair. 1996. Motility protein complexes in the bacterial flagellar motor. Journal of Molecular Biology 261 :209-21. 160. Tawa, P. and R. C. Stewart. 1994. Kinetics of CheA autophosphorylation and dephosphorylation reactions. Biochemistry 110 33:7917-24. 161 . Tawa, P. and R. C. Stewart. 1994. Mutational activation of CheA; the protein kinase in the chemotaxis system of Escherichia coli. Journal of Bacteriology 176:4210-8. 162. Taylor, S.S., D.R. Knighton, J. Sheng, L.F. Ten Eych, and J.M. Sowadski. 1992. Structural framework for the protein kinase family. Annual Reviews of Cell Biology 8: 429-62. 163. Thomas, P.J., Garboczi, D.N., and P.L. Pederson. 1992. Mutational analysis of the consensus nucleotide binding sequences in the rat liver mitochondrial ATP synthase beta-subunit. Journal of Biological Chemistry 267(28): 20331-38. 164. Tomasselli, A.G. and L.H. Noda. 1980. Motochondrial ATP:AMP phosphotransferase from beef heart: purification and properties. European Journal of Biochemistry 103: 481-91. 165. Traut, T.W. 1994. The functions and consensus motifs of nine types of peptide segments that form different types os nucleotide-binding sites. European Journal of Biochemistry 222: 9-19. 166. Van Regenmortel, M.H.V., J.P Briand, S. Muller, and S. Plaue. 1988. Synthetic polypeptides as antigens. In Laboratory Techniques in Biochemistry and Molecular Biology, Vol. 19 (R.H. Burdon and P.H. Knippenberg, eds) Elsevier, New York and Amsterdam. 167. Volz, K. and P. Matsumura. 1991. Crystal structure of Eschericia coli CheY refined at 1.7-A resolution. Journal of Biological Chemistry 266: 15511-19. 168. Vonrhein, C., G.J. Schlauderer, and G.E. Schulz. 1995. Movie of the structural changes during a catalytic cycle of nucleoside monophosphate kinases. Structure 3:483-490. 169. Ward, M.J., A.W. Bell, P.A. Hamblin, H.L. Packer, and J.P. Armitage. 1995. Identification of a chemotaxis operon with two cheY genes in Rodobacter sphaeroides. Molecular Microbiology 17(2): 357-66. 170. West, A. H., S. Djordjevic, H. E. Martinez and A. M. Stock. 1995. Purification, crystallization, and preliminary X-ray diffraction analyses of the bacterial chemotaxis receptor modifying enzymes. Proteins 21 :345-50. 1 1 1 171. West, A. H., H. E. Martinez and A. M. Stock. 1995. Crystal structure of the catalytic domain of the chemotaxis receptor methylesterase, CheB. Journal of Molecular Biology 250:276-90. 172. White, T.E. and T.G. Dewey. 1987. A fluorescence investigation of the nucleotide binding sites of the Ca ATPase. Membrane Biochemistry 7(1 }: 67-72. 173. Wolfe, A. J. and H. C. Berg. 1989. Migration of bacteria in semisolid agar. Proceedings of the National Academy of Sciences of the United States of America 86:6973-7. 174. Wolfe, A. J., M. P. Conley, T. J. Kramer and H. C. Berg. 1987. Reconstitution of signaling in bacterial chemotaxis. Journal of Bacteriology 169: 1878-85. 175. Wolfe, A. J., B. P. McNamara and R. C. Stewart. 1994. The short form of CheA couples chemoreception to CheA phosphorylation. Journal of Bacteriology 176:4483-91. 176. Wolfe, A. J. and R. C. Stewart. 1993. The short form of the CheA protein restores kinase activity and chemotactic ability to kinase-deficient mutants. Proceedings of the National Academy of Sciences of the United States of America 90: 1518-22. 177. Woodcock, D.M., P.J. Crowther, J. Doherty, S. Jefferson, E. DeCruz, M. Noyer-Weidner, S.S. Smith, M.Z. Micheal, and M.W. Graham. 1989. Quantitative evaluation of Escherichia coli host strains for tolerance to cytosine methylation in plasmid and phage recombinants. Nucleic Acids Research 17(9): 3469-78. 178. Wootton, J. C. and M. H. Drummond. 1989. The Q-linker: a class of interdomain sequences found in bacterial multidomain regulatory proteins. Protein Engineering 2:535-43. 179. Yang, V. and M. Inouye. 1993. Requirement of both kinase and phosphatase activities of an Eschericia coli receptor (Taz1) for ligand-dependent signal transduction. Journal of Molecular Biology 231: 335-342. 180. Vanisch-Perron, C., J. Viera, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33(1): 103-19. 112 181. Yonekawa, H. and H. Hayashi. 1986. Desensitization by covalent modification of the chemoreceptor of Escherichia coli. Febs Letters 198:21-4. 182. Yoneya, T., M. Tagaya, F. Kishi, A. Nakazawa, and T. Fukui. 1989. Site-directed mutagenesis of Gly-15 and Gly-20 in the glycine-rich region of adenylate kinase. Journal of Biochemistry (Tokyo) 105(2): 158-60. 183. Zhou, J., R. T. Fazzio and D. F. Blair. 1995. Membrane topology of the MotA protein of Escherichia coli. Journal of Molecular Biology 251 :237-42. 184. Zusman, D. R. and M. J. McBride. 1991. Sensory tran~duction in the gliding bacterium Myxococcus xanthus. [Review]. Molecular Microbiology 5:2323-9. VITA The author, Dolph David Ellefson, was born in Duluth, Minnesota to Rose and Elwood Ellefson. He received his secondary education at San Gorgonio High School in San Bernadino, California in 1974. His undergraduate education was at California Polytechnic State University where he received a bachelor of science degree in 1978. In August 1990, the author entered the Department of Microbiology and Immunology in the Graduate School of Loyola University of Chicago. In 1991, he joined the laboratory of Alan Wolfe where he began his study of signal transduction in Eschericia coli. Upon completion of his work in Alan Wolfe's lab, the author begins work at Oregon Health Sciences University in the laboratory of Fred Heffron in the Department of Molecular Microbiology and Immunology. While there, he will pursue work on the development of a breast cancer vaccine using the bacterium Salmonella as a carrier of the Her2/Neu oncogene. The author is a recipient of a National Institutes of Health training grant. 113 APPROVAL SHEET The dissertation/thesis submitted by Dolph David Ellefson has been read and approved by the following committee: Dr. Alan Wolfe, Director Associate Professor Department of Microbiology and Immunology Loyola University, Chicago Dr. Charles Lange Professor Department of Microbiology and Immunology Loyola University, Chicago Dr. Hans-Martin Jack Associate Professor Department of Microbiology and Immunology Loyola University, Chicago Dr. Thomas Gallagher Assistant Professor Department of Microbiology and Immunology Loyola University, Chicago Dr. Richard Stewart Assistant Professor Department of Microbiology University of Maryland The final copies have been examined by the director of the dissertation/thesis and the signature which appears below verifies the fact that any necessary changes have been incorporated and that the dissertation/thesis is now given final approval by the committee with reference to content and form. The dissertation/thesis is therefore accepted in partial fulfillment of the requirements for the degree of PhD. Date Direct