Figure 1.
Immunoblot of cell fractions derived from wt E. coli BL21 (not containing any plasmids).
A steady-state growing culture was disrupted using French-press and fractionated as described in Materials and Methods. The fractions were subjected to SDS-PAGE (10 µg of protein loaded per lane), western-blotted and exposed to anti-DnaA antibodies. A range of amounts of purified His-DnaA [34] was used for quantification (only the 8.5 ng quantity is shown as an internal standard). Numbers listed below are amounts of DnaA in each sample calculated using densitometry.
Figure 2.
DnaA constructs used in this study.
The constructs were engineered as described in Materials and Methods. Amino acids flanking DnaA domains and fragments are shown. Domains I and II are in white, mCherry sequence is colored black, domain III is represented by subdomain IIIa (orange) and IIIb (red) connected by the linker (blue) to domain IV (in green). Point mutation in domain IV, L417P, is also marked.
Figure 3.
Predicted structure of E. coli DnaA Domains III and IV.
The structure was obtained by threading the corresponding primary sequence on the structures of Aquifex aeolicus DnaA, DnaA Domain III of Thermotoga maritima and E. coli domain IV (Protein Data Bank accession codes 1L8Q, 2Z4R and 1J1V, respectively) using I-TASSER server (http://zhanglab.ccmb.med.umich.edu/I-TASSER/). In orange - Domain IIIa, containing the ATP-binding region of AAA+? ATPase-type proteins (Walker motif), in red - Domain IIIb, sensor region 2, in blue - linker segment that connects Domain III and IV, considered to have a membrane interaction function, and in green - the C-terminal Domain 4, DNA binding domain (for domains designation and function see [1]). Amino acid residues chosen as C-terminals of different domains and used to design DnaA fragments (Figure 2) are shown by arrows. Images were generated using Pymol software (www.pymol.org).
Figure 4.
Increase of DnaA-mCherry content in the membrane fraction as a function of its expression level.
E. coli BL21 harboring pBAD24(dnaA-mCherry) was induced by 0.2% arabinose for time periods shown on the graph in the inset. DnaA-mCherry content in cell fractions is presented in mCherry fluorescence intensity units, determined as described in Experimental procedures.
Figure 5.
Combined phase-contrast and fluorescence images of E. coli BL21 expressing different constructs (see
Figure 2) of mCherry-DnaA (Left column) and inner membranes derived from corresponding cells (Right column). Cells expressing mCherry alone are shown in line A. For an example of separate phase-contrast and fluorescence images see Figure S2. Inner membrane vesicles are about 120 nm in size and therefore are hardly visible in homogeneously dispersed sample. However, they tend to aggregate in suspension forming large clumps visible in phase contrast. Scale bar is 2 µm.
Figure 6.
Intracellular distribution and membrane affinity of various mCherry-DnaA constructs.
Panel A. Amounts of different mCherry-DnaA constructs in whole cell lysates (light columns) and in IMV (dark columns) depicted in fluorescence intensity units. The constructs were expressed in E. coli BL21 from pBAD24 by 0.2% arabinose for 2 hours. The cells were harvested, lyzed and fractionated as described in Experimental procedures. Fluorescence intensity was measured in cell lysate and in the IMV fractions brought to the same protein concentration (0.15 mg/ml). Error bars represent SE from three independent experiments. Panel B. Cellular membrane fraction of mCherry-DnaA constructs. Membrane retention of mCherry hybrid proteins on IMV has been calculated as intensity levels in IMV normalized by the steady-state level of the respective proteins. The latter was obtained from fluorescence intensities in cell lysates, taken as a measure for overall expression level, corrected for the extent of degradation of each construct (Figure S3), accounting for only the fraction of the full-size construct. The cellular membrane fraction of DnaA constructs has been calculated assuming that about 7.5% of total cellular protein is in the inner membrane [44] and shown as percent of total cellular mCherry-DnaA content. Error bars represent SE from three independent experiments.
Figure 7.
Fluorescent structures in E. coli BL21 overexpressing mCherry-DnaA(117–378).
Two examples (A and B) of phase-contrast (left image), fluorescence (middle) and their overlay (right) images of BL21 cells expressing mCherry-DnaA(117–378) during 4 hours. The cells visible as non-fluorescent in B actually display a fluorescence level similar to those shown in Figure 5D1, that is just too low relative to intensity of inclusion bodies. Scale bar is 2 µm.
Figure 8.
Electrostatic surface (left) and ribbon representation (right) of Domain III of predicted E. coli DnaA structure (Figure 3).
The electrostatic surface was calculated by PyMol-molecular visualization system (red denotes negatively charged and blue denotes positive charge). A. Overview of Domain III, outlined is the apparent hydrophobic surface continuity extending throughout the “back” of the domain. B. A 90° rotation view, in which the ATP pocket is visible. The bulgy and positively charged Lys327 is shown by short arrows.
Figure 9.
Cartoon illustrating binding of DnaA to the membrane surface.
The primary electrostatic attraction of Lys327 (marked by +) to acidic phospholipids promotes hydrophobic interaction with the membrane of domain IIIb first, further enforced by interaction of domain IIIa (residues forming the contact hydrophobic continuity are colored red). The concerted interaction drives a conformational change in the protein, decreasing the affinity to the nucleotide. ADP is replaced for ATP due to a much higher concentration of the latter in a functional cell.