Hydroxyl Radical Overproduction in the Envelope: an Achilles' Heel in Peptidoglycan Synthesis

Abstract

While many mechanisms governing bacterial envelope homeostasis have been identified, others remain poorly understood. To decipher these processes, we previously developed an assay in the Gram-negative model Escherichia coli to identify genes involved in maintenance of envelope integrity. One such gene was ElyC, which was shown to be required for envelope integrity and peptidoglycan synthesis at room temperature. ElyC is predicted to be an integral inner membrane protein with a highly conserved domain of unknown function (DUF218). In this study, and stemming from a further characterization of the role of ElyC in maintaining cell envelope integrity, we serendipitously discovered an unappreciated form of oxidative stress in the bacterial envelope. We found that cells lacking ElyC overproduce hydroxyl radicals (HO^•) in their envelope compartment and that HO^• overproduction is directly or indirectly responsible for the peptidoglycan synthesis arrest, cell envelope integrity defects, and cell lysis of the ΔelyC mutant. Consistent with these observations, we show that the ΔelyC mutant defect is suppressed during anaerobiosis. HO^• is known to cause DNA damage but to our knowledge has not been shown to interfere with peptidoglycan synthesis. Thus, our work implicates oxidative stress as an important stressor in the bacterial cell envelope and opens the door to future studies deciphering the mechanisms that render peptidoglycan synthesis sensitive to oxidative stress. IMPORTANCE Oxidative stress is caused by the production and excessive accumulation of oxygen reactive species. In bacterial cells, oxidative stress mediated by hydroxyl radicals is typically associated with DNA damage in the cytoplasm. Here, we reveal the existence of a pathway for oxidative stress in the envelope of Gram-negative bacteria. Stemming from the characterization of a poorly characterized gene, we found that HO^• overproduction specifically in the envelope compartment causes inhibition of peptidoglycan synthesis and eventually bacterial cell lysis.

Keywords: Fenton reaction; bacterial envelope biology; hydroxyl radical; iron homeostasis; oxidative stress; peptidoglycan; peptidoglycan synthesis; reactive oxygen species.

FIG 1

The envelope defect and lysis phenotype of ΔelyC cells depend on medium aeration. (A to C) Growth curves of wild-type (WT) (black curves) and ΔelyC (red curves) cells at 21°C. The values represented correspond to the mean of the optical density at 600 nm (OD_600nm) measurements of at least three biological replicates ± standard deviation (SD). The cultures were grown as follows: 25 mL in 250-mL Erlenmeyer flasks (A); 5 mL in narrow 10-mL culture tubes (B); and in anaerobic conditions (−O₂) (C). Shown is a representative data set from experiments performed in biological triplicates. (D) Phase contrast microscopy imaging at 100× showing cell morphology observed after 12 h of culture in flasks and culture tubes or after 24 h for anaerobic growth. The cells were grown under the same conditions as in panels A to C. The conditions are indicated by pictograms; respectively from left to right: Erlenmeyer flask, culture tube, and anaerobic conditions. Solid arrows point to misshaped, bulging, or lysing cells, and dotted arrows point bacterial ghosts. Bar = 3 μm. (E) Chlorophenol red-β-d-galactopyranoside (CPRG) assay plate for the WT, ΔelyC, and ΔmrcB strains incubated at 21°C in aerobic (+O₂) and anaerobic conditions (−O₂).

FIG 2

ΔelyC mutant cells produce high levels of HO• at 21°C. (A) Simplified molecular scheme representing the production of reactive oxygen species (ROS), superoxide radical (O₂^•−), hydrogen peroxide (H₂O₂), and hydroxyl radical (HO^•). Successive univalent electron (e⁻) leakages to oxygen form O₂^•− and H₂O₂ in the presence of protons and free ferrous iron (Fe²⁺). HO^• are generated from H₂O₂ in the presence of Fe²⁺ through the Fenton reaction, releasing a hydroxyl ion (HO⁻) and ferric iron (Fe³⁺). Chromium at its III or V (Crⁿ) oxidation state induces HO^• production through a Fenton-like reaction (35). (B, C) Flow cytometry histograms showing fluorescein detection at 515 nm (±50 nm) by the number of events from the hydroxyphenyl fluorescein (HPF) probe used to evaluate HO^• levels at 21°C (B) and 37°C (C). WT cells (black curves), ΔelyC cells (red curves), and WT cells were grown with 120 μM potassium chromate as a positive control on the assay for increased HO^• production (orange curves).

FIG 3

HO•-generating free iron is necessary for ΔelyC peptidoglycan (PG) synthesis inhibition and cell lysis at 21°C. (A, D, G, J) Growth curves of WT (black curves) and ΔelyC (red curves) cells grown in flasks. The values represented correspond to the mean of OD_600nm measurements of at least three biological replicates ± SD. (B, E, H, K) Flow cytometry histograms reporting HPF probe fluorescence signal to measure HO^• levels in WT (black curves) and ΔelyC (red curves) cells grown until OD_600nm = 0.35. (C, F, I, L) Phase contrast microscopy imaging at 100× showing cell morphology observed after 12 h of culture. Solid arrows point to lysing or misshaped cells. Dotted arrows point to bacterial ghosts. Bar = 3 μm. Specific growth conditions are indicated on the left of each serial figure line. (A to C) Control conditions. (D to F) Fe-depleted conditions (cultures supplemented with 375 μM 2,2′-dipyridyl). (G to I) Fe-undepleted conditions (cultures supplemented with 375 μM 2,2′-dipyridyl and 100 μM FeSO₄). (J to L) Fe-redepleted conditions (cultures supplemented with 600 μM 2,2′-dipyridyl and 100 μM FeSO₄). (M) Bar graph showing the relative PG amount of WT and ΔelyC cells grown with or without 375 μM 2,2′-dipyridyl (Dip) until OD_600nm = 0.5. The total PG amount of WT cells was defined as the 100% reference point. The data are expressed as means ± standard error of the mean (SEM). The asterisk indicates a significant difference from the reference condition (P < 0.05). ns, not significant.

FIG 4

The ΔelyC mutant does not show elevated background SOS levels. (A) Expression levels of the SOS-regulated sulA promoter, reported by LacZ β-galactosidase activity (measured in Miller units) in WT and ΔelyC (EM9)-derived strains: SG24 and SG25, respectively. SG24 (MG1655, Δ(lacI-lacZp)Φ(Kan^R-sulAp-lacZ)) and SG25 (EM9, Δ(lacI-lacZp)Φ(Kan^R-sulAp-lacZ)). The values represented correspond to the mean of OD_600nm measurements of at least three biological replicates ± SD. Promoter fusion is represented on top. The cells were grown under control conditions and with potassium chromate (chromate) at 125 μM as a positive control for SOS response activation and was added upon culture set-up. The cells were harvested at OD_600nm = 0.35, except for chromate-treated ΔelyC cells, which were collected at OD_600nm = 0.15. Asterisks indicate a significant difference for chromate-treated cells compared to control WT cells (P < 0.05). (B) Growth curves of WT (black curves) and ΔelyC (red curves) cells. The cells were grown in control condition (dotted faint lines) or with 125 μM chromate (full lines) added after 6 h of incubation (T + 6 h). The cells were also grown without (black square) or with (red square) 375 μM 2,2′-dipyridyl at growth initiation (Fe-depleted; see Materials and Methods). (C) Phase contrast microscopy imaging at 100× of WT and ΔelyC cells from growth curves in B harvested after 14 h of growth. Solid arrows point to lysing or misshaped cells. Dotted arrows point to bacterial ghosts. Bar = 3 μm. Chromate or 2,2′-dipyridyl were added at growth initiation. (D) Model representing the impact of oxidative stress in WT and ΔelyC cells in control, Fe-depleted, chromate, and Fe-depleted + chromate conditions. The skull represents lethal PG damages. Blue cages represent 375 μM 2,2′-dipyridyl chelating labile pool of free ferrous iron (Fe²⁺). Cr(V) represents chromium salt at his fifth oxidation state produced from oxidation of chromate (Cr(VI)) occurring in cells (it could also be represented as Cr(III)). Lightning symbols represent SOS response activation (SOS). Under control conditions, the overproduction of HO^• in the envelope of ΔelyC cells through the Fenton reaction induces lethal PG defects. In the Fe-depleted condition, HO^• overproduction is inhibited by iron chelation, relieving the ΔelyC mutant from the PG defect and lysis. When chromate is added to ΔelyC mutant cells, HO^• produced from Fenton and Fenton-like reactions adds up in the envelope, causing more damage than under control conditions and accelerating the lysis phenotype. In the cytoplasm of WT and ΔelyC cells, HO^• generated by the Fenton-like reaction induced by chromate causes DNA damages leading to SOS response activation and cell division inhibition. Adding chromate in the Fe-depleted condition leads to cell division inhibition in ΔelyC cells. As the Fenton reaction is inhibited by the iron chelator, the native production of high levels of toxic HO^• in the envelope of ΔelyC cells is suppressed. The ΔelyC cells then phenocopy WT cells in reaction to chromate treatment. OM, outer membrane; PG, peptidoglycan; IM, inner membrane.

FIG 5

Extracytoplasmic free iron is necessary for ΔelyC mutant PG defect and cell lysis at 21°C. (A, D, G) Growth curves of WT (black curves) and ΔelyC (red curves) at 21°C. The values represented correspond to the means of OD_600nm measurements of at least three biological replicates ± SD. Solid growth curves represent bacteria grown under the condition annotated on the left of the figure. Bacteria grown in control conditions are represented as dotted lines. (B, E, H) Flow cytometry histograms using HPF probe for measuring HO^• production in cells grown at 21°C at an OD_600nm of 0.35, with WT in black and ΔelyC in red. (C, F, I, K) Microscopy photographs of WT and ΔelyC cells at ∼14 h of culture. Solid arrows point to lysing or misshaped cells. Dotted arrows point to bacterial ghosts (100× phase contrast objective). Bar = 3 μm. The corresponding growth conditions are indicated on the left: ethylenediamine-N,N′-bis(2-hydroxyphenylacetic) acid (EDDHA) Fe-depleted condition (culture supplemented with 250 μM EDDHA) (A to C); EDDHA Fe-undepleted condition (culture supplemented with 250 μM EDDHA and 100 μM FeSO₄) (D–F); EDDHA Fe-redepleted condition (culture supplemented with 100 μM FeSO₄ and 600 μM EDDHA) (G–I); Relative PG in amount of WT and ΔelyC cells grown under control conditions or in EDDHA Fe-depleted condition (EDDHA) (J); and control conditions (K). The values are the means ± SEM (of three biological replicates). The asterisk indicates significant differences to WT grown without EDDHA (P < 0.05). ns, not significant.