A Mutant Isoform of ObgE Causes Cell Death by Interfering with Cell Division

Abstract

Cell division is a vital part of the cell cycle that is fundamental to all life. Despite decades of intense investigation, this process is still incompletely understood. Previously, the essential GTPase ObgE, which plays a role in a myriad of basic cellular processes (such as initiation of DNA replication, chromosome segregation, and ribosome assembly), was proposed to act as a cell cycle checkpoint in Escherichia coli by licensing chromosome segregation. We here describe the effect of a mutant isoform of ObgE (ObgE^∗) that causes cell death by irreversible arrest of the cell cycle at the stage of cell division. Notably, chromosome segregation is allowed to proceed normally in the presence of ObgE^∗, after which cell division is blocked. Under conditions of rapid growth, ongoing cell cycles are completed before cell cycle arrest by ObgE^∗ becomes effective. However, cell division defects caused by ObgE^∗ then elicit lysis through formation of membrane blebs at aberrant division sites. Based on our results, and because ObgE was previously implicated in cell cycle regulation, we hypothesize that the mutation in ObgE^∗ disrupts the normal role of ObgE in cell division. We discuss how ObgE^∗ could reveal more about the intricate role of wild-type ObgE in division and cell cycle control. Moreover, since Obg is widely conserved and essential for viability, also in eukaryotes, our findings might be applicable to other organisms as well.

Keywords: Obg; ObgE; cell cycle; cell cycle checkpoint; cell division; cell separation; lysis.

FIGURE 1

Characterization of ObgE^∗-mediated cell death. (A) Exponential-phase cultures of Escherichia coli pBAD33, E. coli pBAD33-obgE or E. coli pBAD33-obgE^∗ were induced at time 0. At several time points before and after induction, the number of viable cells was determined by plate counting. Error bars represent the standard error of the mean, n ≥ 3. (B) Time lapse observations of E. coli pBAD33-obgE^∗ seeded on a lysogeny broth (LB) agar pad containing the inducer of ObgE^∗ expression and propidium iodide (PI). Pictures were taken over a time course of 12 h. Scale bar, 1 μm. (C) Exponential-phase cultures of E. coli pBAD33, E. coli pBAD33-obgE or E. coli pBAD33-obgE^∗ were induced at time 0. At several time points after induction, cultures were stained with PI and the percentage of PI-negative and thus intact cells in the population was measured by flow cytometry. Data are represented as mean ± SEM, n ≥ 3. In every repeat 100,000 cells were collected.

FIGURE 2

Lysis is caused by rupturing of membrane blebs that consist of both outer and inner membrane. (A) Scanning electron microscopy images of E. coli overexpressing ObgE^∗ show the formation of membrane blebs. (B) Microscopy images of E. coli pBAD33, E. coli pBAD33-obgE or E. coli pBAD33-obgE^∗ with a cytoplasmic GFP label encoded by pQE80L-gfp. Cells were stained with HADA for visualization of peptidoglycan and FM4-64 for labeling of membranes. PH, phase contrast. (C) 3D-model of a bleb constructed by focused ion beam-scanning electron microscopy (FIB-SEM). The outer membrane is shown in white, inner membrane in blue and accessory blebs that are not in direct contact with the major bleb are shown in green. Additionally, sections from this bleb from three different angles are shown. (D) Time lapse images showing the formation of blebs and lysis of E. coli pBAD33-obgE^∗ pQE80L-gfp upon induction with arabinose. Blebs that cause lysis are indicated with arrows. Scale bars, 1 μm.

FIGURE 3

Blebs are located at sites of cell division. (A) Quantitative localization of blebs in E. coli pBAD33-obgE^∗. Cultures were induced in the presence of 100 mM MgSO₄ to increase bleb size and lifetime and thus improve visibility under the microscope. Membranes were stained with FM4-64 before visualization. Data are represented as mean ± SEM, n = 3. In every repeat ± 200 blebs were counted. (B) Violin plot of the distribution of blebs from the category ‘Other’ within the cell. To account for the random distribution of bleb formation in either the left hand or right hand side of the cell, all data points were duplicated and mirrored around midcell position. Scale bars, 1 μm.

FIGURE 4

Cell division is necessary for lysis but does not affect loss of viability. (A) Correlation curve showing the expected fraction of intact cells of E. coli pBAD33-obgE^∗ in function of the intracellular ObgE^∗ concentration. (B) Correlation curve showing the expected level of viability of E. coli pBAD33-obgE^∗ in comparison to E. coli pBAD33-obgE in function of the intracellular ObgE^∗ concentration. Colored data points were collected from conditions that inhibit or slow down cell division. ObgE^∗ concentration was determined by measuring fluorescence of an ObgE^∗-Venus fusion by flow cytometry. Gray bands around the expected value represent 99% prediction intervals. Data are represented as mean ± SEM, n ≥ 3, error bars are mostly too small to be visible.

FIGURE 5

Amidases and the Tol-Pal system do not directly cause ObgE^∗-mediated lysis. (A) Induced cultures of E. coli (wt), E. coli ΔamiA ΔamiB ΔamiC (ΔamiABC), E. coli ΔtolQ and E. coliΔtolR with plasmids pBAD33, pBAD33-obgE or pBAD33-obgE^∗ were stained with PI and the percentage of PI-negative and thus intact cells in the population was measured by flow cytometry. (B) The percentage of intact cells upon ObgE^∗ expression was divided by the fraction of intact cells upon expression of ObgE to correct for the differences in baseline levels of membrane integrity in different strains. Data are represented as mean ± SEM, n ≥ 3. In every repeat 100,000 cells were collected (one-way ANOVA, Bonferroni correction: ^∗p < 0.05, ^∗∗∗∗p < 0.0001).

FIGURE 6

ObgE^∗ causes irreversible cell cycle arrest. (A) After ObgE or ObgE^∗ was expressed in stationary phase, cells were seeded on an agarose pad without the inducer of expression. Resumption of growth on the pad was monitored by time lapse microscopy. Insets are enlargements indicated by red lines. (B) Violin plot and box plot of the cell length of E. coli pBAD33-obgE^∗ at the start of the time lapse experiment and 6 h later. Cell length was quantified in ±220 cells spread over three independent repeats. (C) After ObgE^∗ was expressed in E. coli hupA-venus in stationary phase, cells were seeded on an agarose pad without the inducer of expression. Chromosome segregation was monitored by time lapse microscopy. (D) Quantitative analysis of chromosome segregation in E. coli hupA-venus pBAD33Gm-obgE^∗ after induction in stationary phase. Data are represented as mean ± SEM, n = 4. In every repeat ± 100 cells were analyzed. Scale bars, 10 μm.