Biochemical, structural and functional analysis of the cell division site determinant MinD(Ng): To divide or not to divide

Abstract

The Min system, constituting MinC, MinD, and MinE proteins plays an essential role in regulating cell division in Gram-negative bacteria. MinD, a peripheral membrane ATPase, functions by recruiting the cell division inhibitor, MinC, to the membrane where it prevents septation. Counteracting this effect is MinE which can bind to MinD, thereby displacing MinC. Formation of the MinE-MinD complex also stimulates the ATPase activity of MinD, and is required for the periodic oscillations of the MinC-MinD complex on and off the membrane required for normal cell division. Our laboratory has shown that MinD self-associates, and this ATP-dependent dimerization is essential for protein function. Recently I and others in the laboratory demonstrated that, unlike wild type MinD, a 3 amino acid (MinDNg-3aaNT) deletion from the N-terminus of MinD is sufficient to abrogate MinDNg-3aaNT self interaction, its interaction with wild type MinDNg, and MinENg (Szeto et al., 2004). In addition, a polar loop on the surface of MinD implicated in its dynamic oscillation and topological specificity (Szeto et al., 2005) was identified. Despite the presence of MinE, hydrophobic mutations at these polar residues (R92, D93, and K94) abolished MinD oscillations. While crystal structures are available for monomeric MinD it is not yet known how dimerization occurs or how interactions with MinE can stimulate MinD ATPase activity. In order to address these questions I developed protocols for the production of MinD samples suitable for analysis by solution NMR and to study ATPase and lipid-binding activity of wild type and mutant MinD proteins in vitro. Since the solubility of wild type MinD was found to be limiting for the solution NMR studies, I designed mutants that exhibit increased solubility and can be obtained in higher yields than previously possible. Preliminary results with the most soluble construct indicated that the protein was folded but that higher concentrations will be required for structural studies. In parallel, functional assays for MinD ATPase and lipid binding activity were also developed. I found that with respect to ATP, MinD displays positive substrate cooperativity both in the presence and absence of MinENg. A Hill analysis of these curves suggests that in the absence of MinE there are at least two interdependent binding sites, while in the presence of MinE the cooperativity increases such that at least eight active sites are functionally linked. Since there can be only one active site per MinD monomer, the presence of cooperativity suggests that MinD must form oligomers that are more active than the monomers. Moreover, the increase in cooperativity in the presence of MinE suggests that MinE promotes the formation of higher order MinD polymers with an extensive network of catalytic sites that positively regulate each other. New insight were also provided into the role of the N-terminus of MinD by these in vitro assays since the mutant MinDNg-3aa NT, had basal ATPase activity that was ∼6-fold higher than wild type and that this activity was not further stimulated by MinE. Thus the N-terminal region of MinD influences ATPase activity and is important for MinE-mediated stimulation of this activity. In contrast I found that a surface exposed loop on MinD would almost completely abrogate ATPase activity if mutated to hydrophobic residues even though it is still able to interact with MinE and phospholipids. Overall, my studies of MinD function, provided insight in which nucleotide binding and hydrolysis control the localization of MinD, and how MinE can promote higher reaction rates in MinD.