Open Access
Meyer, Matthew Geoffrey
Graduate Program:
Biochemistry, Microbiology, and Molecular Biology
Doctor of Philosophy
Document Type:
Date of Defense:
December 19, 2001
Committee Members:
  • Daniel J Cosgrove, Committee Member
  • Ming Tien, Committee Member
  • B Tracy Nixon, Committee Chair
  • Jean Elnora Brenchley, Committee Member
  • Song Tan, Committee Member
  • x-ray crystallography
  • transcription activation
To understand the structure/function of response regulator proteins of two-component signal transduction regulatory systems, the crystal structure of a fragment bearing the wild-type DctD receiver domain, as well as the structures for a number of single amino acid substitution variants, have been solved using MAD (Multiwavelength Anomalous Dispersion) and molecular replacement. Two of these substitutions, E121K and K122E, constitutively activate full length DctD as an ATPase and transactivator of s54-dependent transcription from the dctA promoter. The structures of these mutant proteins reveal inter- and intramolecular hydrogen bonding rearrangements; however, at room temperature the variant structures are essentially identical to wild-type. N-terminal deletion analysis has shown that only the C-terminal portion of this fragment is necessary to keep DctD inactive. The structures determined in this thesis reveal that this part of the molecule forms a homodimerization surface, and furthermore, mutations that perturb that surface result in constitutively active DctD. Equilibrium and velocity ultracentrifugation experiments confirm that the fragment (aa 2-143) forms a dimer in solution. Calculated KDs of dimer dissociation increase 10-fold for E121K fragment over wild-type; however, phosphorylation of wild-type fragments shows no apparent dimer loosening. Important hypotheses have been proposed as a result of further study of this novel structure by the Nixon lab. CD spectra suggest a conformational change when constitutive amino acid variant E121K is cooled from 30 ºC to 5 ºC. CD spectra suggest a similar temperature-dependent conformational change for the WT 2-143 fragment, but only when it is phosphorylated. Since removing Mg2+ from the phosphorylated E121K receiver domain decreased its dephosphorylation rate to a half-life of ~50 hours, crystallization conditions were developed for rapidly growing crystals from the mixture of phosphorylated and unphosphorylated protein. While this did not yield crystals of phosphorylated protein, the new conditions did reveal E121K crystal structures altered in a way previously not observed for two-component receiver domains, which involve altered contacts between subunits of the DctD receiver domain dimer. This second conformation is hypothesized to be involved in DctD's structural mechanism of activation, though per se it is not believed to represent the active state. Finally, crystals of wild-type and the E121K variant of the DctD receiver domain were soaked in conditions that in solution yield phosphorylated protein. Using protein from single crystals to provide material for mass spectrometry, evidence of carbamylation was obtained but no phosphorylation was detectable. Carbamylation was expected to occur more slowly than phosphorylation from the solution studies. New results from Park and Nixon using BeF3 suggest that a totally different dimer may represent the active state of DctD. In summary, this work defines a novel dimeric structure for two component receiver domains that is used in DctD to maintain the full length protein in its 'off-state'. Together with the new results of others in the lab, this work provides a model for activation in which phosphorylation of the receiver domain dramatically alters partitioning between two distinctly different dimeric structures. Sequence analysis suggests that this mechanism may be relevant to 4.5% of all two component response regulators. This is a major accomplishment, because two-component signal transduction is the most prevalent form of signalling among cells living in the earth's biosphere.