The gut bacterium Escherichia coli has the remarkable ability to rotate its flagella either counterclockwise, which results in forward swimming, or clockwise, which causes cell tumbling. The direction of flagellar rotation at any moment is determined by the degree of phosphorylation (addition of a phosphoryl group [-PO_3^2-]) to the chemotaxis signaling protein CheY, which binds to the base of the flagella. CheY is phosphorylated by the histidine kinase CheA and dephosphorylated by the CheZ phosphatase, the subject of our X-ray crystallographic studies.

Our initial efforts using crystals containing only CheZ were unsuccessful due to the poor diffraction of these crystals. We then shifted strategies and exploited a recent discovery that beryllium fluoride (BeF_3^-) binds to CheY to form a stable replica of phosphorylated CheY, the substrate of CheZ. We obtained crystals containing CheZ and CheY-BeF_3^--Mg²⁺, and a cryoprotectant of 50% sucrose enabled us to improve the diffraction of these crystals from 8 angstrom (obtained with standard cryoprotectants) to 3.5 angstrom.

The CheZ dimer features an impressively long (105 angstrom) four-helix bundle composed of two helices from each chain, as shown in figure 1. About 130 (of the 214) residues from each chain form an extended amphipathic (having both water-loving and water-repelling groups) alpha helix which folds on itself with a single hairpin turn. Two of these hairpins assemble to form the bundle. The amino-terminal of each chain forms an additional helix, as does the carboxyl-terminal, which interacts with a surface on CheY. The linker region between the carboxyl-terminal helix and the four-helix bundle was not visible in the electron density maps, implying that this region was disordered in the crystal.

For decades, scientists have debated whether CheZ acts as a positive allosteric effector of CheY’s own autodephosphorylation activity by binding at a region of CheY away from the active site, or as a traditional phosphatase, which dephosphorylates CheY by binding to its active site. Our results showed two surfaces of interaction between CheY and CheZ. The carboxyl-terminal α-helix of CheZ binds to a surface on CheY that is distant from its active site, but a region of the four-helix bundle of CheZ interacts directly with the active site of CheY.

Strikingly, one residue from CheZ, called glutamine 147, inserts directly into the CheY active site. Mutagenesis studies showed that this residue was critical for catalytic activity. We now believe that glutamine 147 stimulates the CheY autodephosphorylation reaction by positioning a water molecule in an appropriate position to remove phosphoryl groups, as shown in figure 2. So CheZ activity cannot be classified simply as an allosteric effector or true phosphatase, but instead contains features of both.

Our work provides structural information on the last of the seven proteins involved in E. coli chemotaxis. This new description of the CheZ structure and mechanism will enhance our understanding of bacterial chemotaxis.

BEAMLINE
X4A, X25

FUNDING
National Institutes of Health
Cancer Research Institute

PUBLICATION
Rui Zhao, Edward J. Collins, Robert B. Bourret, Ruth E. Silversmith. “Structure and catalytic mechanism of the Escherichia coli chemotaxis phosphatase CheZ,” Nature Struct. Biol. 9, 570 (2002)

Philip Matsumura, “Last but not least,” Nature Struct. Biol. 9, 563 (2002) (Commentary accompanying Rui Zhao et al.’s article)

FOR MORE INFORMATION
Ruth Silversmith
Research Assistant Professor
Department of Microbiology and Immunology
University of North Carolina at Chapel Hill
Email: silversr@med.unc.edu
Web:
University of North Carolina at Chapel Hill: http://www.unc.edu/~gpguanga/
University of Colorado Health Science Center in Denver: http://biomol.uchsc.edu/researchFacilities/XRayCore/contact.html