Abstract Alkyl hydroperoxide reductase (Ahp) is the primary scavenger of endogenous hydrogen peroxide in Escherichia coli (E. coli). Ahp-deficient strains have been found to have high reactive oxygen species (ROS) levels, sufficient to cause cell damage. However, the exact role and underlying mechanisms of Ahp deficiency-induced cell damage remain largely unknown. Here, the E. coli MG1655 ΔAhp mutant strain was constructed as a model of deficiency to assess its role. The cells of the ΔAhp strain were found to be significantly longer than those of the wild strain, with elevated ROS and hydrogen peroxide (H[2]O[2]) levels. Proteome, redox proteome and metabolome analyses were performed to systematically present a global and quantitative profile and delineate the redox signaling and metabolic alterations at the proteome, metabolome, and cysteine oxidation site levels. The multiomics data revealed that Ahp deficiency disrupted the redox balance, activated the OxyR system, upregulated oxidative defense proteins and inhibited the TCA cycle to some extent. Surprisingly, the mutant strain shifted from aerobic respiration to anaerobic respiration and fermentation during the logarithmic phase in the presence of sufficient O[2.] The acid resistance system was activated to mitigate the effect of excessive acid produced by fermentation. Taken together, the results of this study demonstrated that Ahp deficiency triggered cellular redox imbalance and regulated metabolic pathways to confer resistance to submicromolar intracellular H[2]O[2] levels in E. coli. Keywords: Alkyl hydroperoxide reductase, Label-free proteomics, Metabolic alterations, Redox proteomics, Metabolomics Graphical abstract Image 1 [29]Open in a new tab 1. Introduction Reactive oxygen species (ROS) include superoxide anion (O[2]^−), hydrogen peroxide (H[2]O[2]), and hydroxyl radicals (OH·) [[30]1]. It has been demonstrated that 87% of endogenous H[2]O[2] in Escherichia coli (E. coli) is produced by the respiratory chain reaction [[31]2]. Excessive ROS cause oxidative damage to proteins, nucleic acids, and cysteine residues in cells. Thus, E. coli has evolved enzymatic and nonenzymatic ROS-scavenging systems [[32]3]. The main antioxidant enzymes that scavenge H[2]O[2] are alkyl hydroperoxide reductase (Ahp), catalase G (KatG), and catalase E (KatE) [[33]4]. Ahp was demonstrated to be the primary scavenging enzyme for endogenous low-concentration H[2]O[2] in E. coli [[34]5]. The mutant strain lacking katG and katE grew like the wild-type strain [[35]6]. However, H[2]O[2] accumulated in ahp and katE and katG knockout E. coli, which was enough to cause damage to DNA and proteins as well as cell growth and metabolism [[36]7,[37]8]. Ahp is composed of two subunits, the hydroperoxide reductase AhpC and flavoprotein AhpF. AhpC is among the ten most abundant proteins in E. coli [[38]9] and is a member of the typical 2-Cys antioxidant protein family. AhpF reduces oxidized AhpC using NADH/NAD(P)H as an electron donor. AhpC and AhpF jointly catalyze the conversion of endogenous H[2]O[2] to two molecules of water [[39]10]. It was demonstrated that Ahp scavenged most of the endogenous H[2]O[2] under regular growth conditions, with only a small fraction scavenged by catalase [[40]11]. AhpC also plays a key role in iron metabolism in E. coli, and AhpC deficiency leads to slow cell growth and reduced iron accumulation [[41]12]. Additionally, overexpression of AhpF has been found to suppress protein aggregation upon antibiotic stimulation, suggesting that cellular defense against H[2]O[2] can reduce the cell damage [[42]13]. However, how Ahp deficiency-induced ROS modulate the redox imbalance and metabolic alterations atill warrant investigation and needs to be further clarified. In this study, the E. coli MG1655 ΔAhp mutant strain was constructed first, and this Ahp-deficient strain exhibited aggregation and filamentation along with elevated ROS and H[2]O[2] levels. Then, we aimed to characterize the proteome, redox proteome and metabolome of both the E. coli MG1655 (wild type, WT) and E. coli MG1655/ΔAhp (ΔAhp) strains ([43]Fig. 1). These muliomic studies not only provided more comprehensive insights into Ahp function but also delineated the redox signaling and downstream pathways. Ahp deficiency disrupted the redox balance, activated the OxyR system, upregulated oxidative defense proteins and altered energy metabolism. The results of this study demonstrated that Ahp knockout triggered redox imbalance and metabolic alterations in the mutant strain. This study provides a useful dataset with interesting implications for understanding E. coli physiology. Fig. 1. [44]Fig. 1 [45]Open in a new tab Overall experimental flowchart. Both WT and ΔAhp strains were subjected to analyses using proteomics, redox proteomics, untargeted metabolomics, and targeted metabolomics. 2. Materials and methods 2.1. Chemical reagents Urea, dithiothreitol (DTT), iodoacetamide-^13C[2],2-D[2], and ammonium bicarbonate were purchased from Sigma‒Aldrich (Steinheim, Germany). Trypsin was purchased from Promega (Madison, WI). Formic acid and acetonitrile (ACN) were purchased from Fisher Scientific Canada (Edmonton, Canada). The protease inhibitor was supplied by Roche (Mannheim, Germany). The BCA protein assay kit and BCA peptide assay kit were purchased from Thermo Fisher Scientific (USA). 2.2. Bacterial strains and growth curves The WT strain E. coli MG1655 was maintained by our laboratory. The mutant strain E. coli MG1655/ΔAhp was constructed by homologous recombination with plasmids [[46]8,[47]14]. The primers were shown in [48]Table 1. The 50 bp sequences on both sides of the ahpC/F gene were selected as the upstream and downstream homologous recombination arms. Then, the gene target fragments were obtained by PCR using the plasmid pKD3 as template. Gene targeting fragments were used to electroporate the WT strain carrying the helper plasmid pKD46, and positive clones were selected based on the associated antibiotic resistance. Finally, the plasmid pCP20 was used to remove the chloramphenicol resistance gene, and the mutant strain ΔAhp was obtained. The mutation was confirmed by PCR followed by agarose gel electrophoresis of the PCR products. The two strains were transferred to sterile LB liquid medium and cultured at 37 °C with shaking at 150 rpm until they reached log phase (OD[600] ~ 0.8) for further morphology observation and multiomics analysis. Table 1. Sequence list of primers used for ahpC/F knockout and verification. Gene to be knocked out Primer sequence ahpC/F （to be replaced by the chloramphenicol resistance gene） UP：CCAGGTTTGACCCCGGCGGCTAAGCAATTGCAGGTGAATCTTACTTCTTCgagctgcttcgaagttcc ta Down：GGCAAAAATTGGTTACCTTACATCTCATCGAAAACACGGAGGAAGTATAGcatatgaatatcctcc ttagttcctattc Chloramphenicol resistance gene Up：CTAACGCCTCGAGCGTGATC Down：GTTCACAAAGTTGTCTTACGAAGGTTG [49]Open in a new tab For plotting the growth curves, both the WT and ΔAhp strains were cultured in sterilized fresh LB liquid medium at 37 °C with shaking at 150 rpm. The absorbance at OD[600] for each strain was measured every hour in three biological replicates. Based on the OD[600] values, the growth curves of both the WT and ΔAhp strains were plotted. 2.3. Observation of strain morphology The two strains were observed using a Leica TCS-SP5 microscope and field emission scanning electron microscopy to observe the effect of gene knockout on the phenotype of the strains. The Leica TCS-SP5 microscope (Leica Microsystems, Mannheim, Germany) was used to illuminate the field using a 10x eyepiece and a 20x objective for observation, and the cell morphology was imaged by scanning electron microscopy using the critical point drying sample preparation method, mainly to observe the differences between the WT and ΔAhp strains in terms of shape, size, and the degree of colony aggregation. 2.4. Assay of H[2]O[2] and ROS In the presence of H[2]O[2], horseradish peroxidase (HRP) can oxidize amplex red (AR) to the fluorescent product resorufin [[50]15]. AR and HRP were diluted with 50 mM potassium phosphate (KPi, pH 7.8) to prepare stock solutions. AR, HRP and the standards/samples were mixed, and the fluorescence was immediately measured with a fluorescence-based microplate reader with excitation at 530 nm and detection at 585 nm. The concentration of H[2]O[2] in each sample was calculated using a standard curve prepared with known concentrations of H[2]O[2]. ROS were detected by the fluorescent probe DCFH-DA, which was hydrolyzed to DCFH and then oxidized to fluorescent DCF by intracellular ROS [[51]16]. Both strains were resuspended in fresh medium, the DCFH-DA probe was added, and the samples were incubated at 37 °C. The excess probe was washed away with PBS. The fluorescence of each sample was detected by flow cytometry. 2.5. Sample preparation for multiomics analysis The WT and ΔAhp strains were transferred to sterile LB liquid medium and cultured at 37 °C with shaking at 150 rpm until they reached log phase (OD[600] ~ 0.8). The two cultures were centrifuged at 10000 g for 10 min at 4 °C, and the pellets were collected and washed twice with 50 mM PBS. Proteomics: The bacterial pellet was resuspended in lysis buffer (8 M urea, EDTA, protease inhibitor mixture, 50 mM NH[4]HCO[3]). Then, the cells were disrupted on ice by sonication for 4 min. Afterward, the sample was centrifuged at 12000 g for 30 min at 4 °C, and the supernatant was transferred into a new tube. Each sample was reduced with 10 mM DTT for 30 min at 56 °C, alkylated with 50 mM IAA in the dark for 30 min, and then diluted and digested for 18 h at 37 °C by trypsin. The residual trypsin activity was quenched by the addition of 1% formic acid (v/v). The peptides were desalted using a C[18] solid-phase extraction (SPE) column (Supelco, Bellefonte, PA) and dried using a vacuum centrifuge. Metabolomics: The bacterial pellets were snap-frozen with liquid nitrogen quickly, and then 1 mL precooled methanol/acetonitrile/water solution (2:2:1, v/v/v) was added. Then, each sample was sonicated on ice for 30 min and stored at −20 °C for 1 h. After centrifugation at 14000g for 20 min, the supernatant was collected for LC-MS/MS analysis. Redox proteomics: The pellets of the WT and ΔAhp strains were resuspended in lysis buffer containing iodoacetamide (^12C[2]H[4]INO) and stable isotope-labeled heavy iodoacetamide (^13C[2]D[2]H[2]INO) [[52]17], respectively. After protein extraction, equal amounts of total protein extracts (50 μg) were mixed using a tag exchange replication strategy. Following treatment of the protein samples with 10 mM DTT, reduced thiols were blocked with N-ethylmaleimide (NEM). Tryptic digests were subjected to reverse-phase chromatography. The peptides were resuspended in buffer A (98% water, 0.02% ammonium hydroxide; pH 10) and separated by an Agilent 1100 HPLC system with a Durashell column (2.1 × 150 mm, 5 μm) at a flow rate of 0.2 mL/min. The linear gradient was 5% B (98% ACN, 0.02 ammonium hydroxide; pH 10) for 5 min, followed by 5–27% B for 25 min, 27–40% B for 10 min, and 44–100% B for 5 min. The gradient was then held at 100% for 5 min, and finally at 5% B for another 10 min. The eluate was collected at 1 min intervals. The fractions were combined into 10 fractions according to the UV spectroscopy results and evaporated to dryness by Speed Vac. 2.6. LC-MS/MS analysis The peptide mixture for proteomics and redox proteomics was dissolved in water containing 0.1% FA and analyzed using an online U3000-nano coupled with an Orbitrap Q Exactive HFX mass spectrometer (Thermo Fisher Scientific, Massachusetts, USA). Proteomic samples were separated using a 15 cm house-made C[18] reversed-phase column (100-μm inner diameter, 1.9 μm resin) and a 110 min elution gradient (mobile phase B: 5% at 0 min, 5% at 4 min, 10% at 20 min, 22% at 64 min, 35% at 94 min, 99% at 99 min, 99% at 104 min, 5% at 105 min, 0% at 110 min). The data were acquired in a data-independent mode. For mass spectrometry parameters, the m/z scan range was set to 400–1200, and 41 isolation windows were set. The redox proteomic samples were separated with an 80 min elution gradient at a flow rate of 300 nl/min (mobile phase B: 5% at 0 min, 5% at 4 min, 35% at 64 min, 99% at 69 min, 99% at 74 min, 5% at 75 min, 0% at 80 min). The data were acquired in a data-dependent mode. The m/z range was set to 350–2000 for the MS scan. The 20 most intense ions in MS1 were selected for MS/MS analysis. For metabolomic analysis, untargeted metabolomics samples were analyzed using an UHPLC (1290 Infinity LC, Agilent Technologies) coupled to a quadrupole time-of-flight instrument (AB Sciex TripleTOF 6600) at Shanghai Applied Protein Technology Co., Ltd. Each sample was analyzed using a 2.1 mm × 100 mm ACQUIY UPLC BEH 1.7 μm column (Waters, Ireland). In both ESI positive and negative modes, mobile phase A contained 25 mM ammonium acetate and 25 mM ammonium hydroxide in water, and mobile phase B was acetonitrile. The gradient was 85% B for 1 min, which was linearly reduced to 65% in 11 min and then to 40% in 0.1 min, which was maintained for 4 min, followed by an increase to 85% in 0.1 min. The ESI source conditions were set as follows: Ion Source Gas1 (Gas1): 60, Ion Source Gas2 (Gas2): 60, curtain gas (CUR): 30, source temperature: 600 °C, IonSpray Voltage Floating (ISVF): ±5500 V. In MS only acquisition, the instrument was set to acquire over the m/z range 60–1000 Da, and the accumulation time for the TOF MS scan was set at 0.20 s/spectrum. In auto MS/MS acquisition, the instrument was set to acquire over the m/z range 25–1000 Da. The parameters were set as follows: the collision energy (CE) was fixed at 35 V with ±15 eV; the declustering potential (DP) was 60 V (+) and −60 V (−); isotopes within 4 Da were excluded; and 10 candidate ions were monitored per cycle. Targeted metabolomics samples were separated using the same chromatographic system and analyzed on a 5500 QTRAP mass spectrometer. The samples were analyzed using an ACQUITY UPLC BEH Amide column (2.1*100 mm, 1.7 μm; Waters MS Technologies, Manchester, UK). Mobile phase A contained 15 mM CH[3]COONH[4] in water, and mobile phase B was acetonitrile. The column temperatures were kept constant at 45 °C. The gradients were implemented at a flow rate of 300 μL/min, and a 2 μL aliquot of each sample was injected. The gradient was as follows: 90% B, linearly reduced to 40% in 0–18 min, increased to 90% in 0.1 min, increased to 85% in 0.1 min, and then maintained for 18.1–23 min. The QC samples, prepared from the pooled samples, were placed into the column at regular intervals in the analysis sequence (one QC after every 5 samples) to monitor the precision and stability of the method during operation. In ESI negative mode, the conditions were set as follows: source temperature 450 °C, ion source gas 1: 45 psi, ion source gas 2: 45 psi, and curtain gas: 30 psi. 2.7. Data processing and bioinformatics analysis For proteomics samples, the raw data were processed using DIA-NN software with the UniProt E. coli database (2019/10/29, taxonomy ID: 83333, 4391 sequences). In particular, FASTA digestion for library-free search/library generation and deep learning-based spectra and RT and IM prediction were enabled for precursor ion generation. The search parameters were as follows: trypsin digestion with a maximum of two missed cleavages, carbamidomethyl (C) as a fixed modification, oxidation (M) as a variable modification, peptide length of 7–30, precursor charge range of 2–4, peptide mass range of 400–1200 Da, and fragment m/z range of 200–1800 Da. The RAW mass spectrometry files from redox proteomics were processed using MaxQuant with an integrated Andromeda search engine (FDR<0.01). Tandem mass spectra were searched against the UniProt E. coli database (2019/10/29, taxonomy ID: 83333, 4391 sequences) concatenated with a reverse decoy database. The following parameters were used for the proteomics search: trypsin enzyme specificity, a maximum of two missed cleavages; fixed modification: carbamidomethyl (C), variable modification: oxidation (M). The following parameters were used for the redox proteomics search: trypsin enzyme specificity; label mode: carbamidomethyl (C) as the light label and carbamidomethyl +4 Da (C) as the heavy label. The other parameters in MaxQuant were set at default values. Then, the retrieval data were imported into Perseus software for statistical analysis, including principal component analysis (PCA) and cluster heatmap analysis. The identified peptides containing at least one cysteine residue with Log[2] (Ratio Heavy/Light) values greater than 0.5 or less than −0.5 were considered significantly oxidized or reduced peptides, respectively. The DAVID 6.8 bioinformatics tools were used for functional categorization of differentially expressed proteins, including by GO (gene ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) enrichment. In addition, Cytoscape (version 3.6.1) plugin ClueGO (version 2.5.4) and CluePedia (version 1.5.4) were used to visualize the KEGG pathways and PPIs (protein‒protein interactions) of related proteins (p values < 0.05; kappa score 0.7), and the PPI information was downloaded from the STRING database [[53]18]. Metabolic data were converted into mzXML and RAW file formats using ProteoWizard software, and peak identification alignment, retention time correction, and extracted peak area determination were performed using XCMS software. Untargeted metabolomics data were analyzed using the Shanghai Applied protein Technology Bioinformatic Platform for subsequent information generation. Targeted metabolomics results were analyzed using MultiQuant software to extract chromatographic peak areas and retention times, and after absolute quantification, histograms were plotted using GraphPad. 2.8. Assessment of ATP and GSH ATP, adenosine triphosphate, is responsible for the transfer and storage of energy. ATP was measured using an ATP Assay Kit (Beyotime, China). Glutathione (GSH) can produce yellow TNB and GSSG when reacting with the chromogenic substrate DTNB [[54]19]. After the two strains were cultured to logarithmic phase, GSH and GSSG detection kits (Beyotime Biotechnology, China) [[55]20] were used for quantification. The content of reduced GSH was calculated by measuring the total GSH and GSSG levels. 2.9. Parallel reaction monitoring analysis The differentially expressed proteins between the WT and the ΔAhp strains were selected for further targeted quantification by parallel reaction monitoring (PRM). Skyline software (MacCross Laboratory, University of Washington) was used to analyze the PRM data and calculate the normalized peak area generated from precursor ions in the two strains. Relative protein quantification was performed by calculating the ratio of the peptide peak area of the mutant strain to that of the wild strain. 3. Results 3.1. ΔAhp strain construction and growth curve The PCR amplification products of MG1655/ΔAhp:Cm were 1368 bp shorter than those of the control strain ([56]Fig. S1A). When the Cm gene was successfully deleted in MG1655/ΔAhp::Cm, the length of the PCR-amplified products obtained with the outer primers of ahp was 374 bp ([57]Fig. S1B). We also compared the expression level of the ahp gene between the WT and ΔAhp strains to confirm that the gene was knocked out successfully ([58]Fig. S1C). The growth status of the two strains was continuously measured for 18 h ([59]Fig. S1D). The ΔAhp strain grew as well as the wild-type strain for the initial 5 h. However, it grew slower than the wild-type strain in the logarithmic phase, which may reflect a shift in the carbon source being used and was probably due to the accumulation of H[2]O[2] in the ΔAhp strain. 3.2. H[2]O[2] and ROS levels and morphology of the two strains To investigate the levels of oxidative stress and whether H[2]O[2] accumulates in the ΔAhp strain, the ROS and H[2]O[2] concentrations were measured. The intracellular ROS level was significantly higher in the ΔAhp strain than in the WT strain ([60]Fig. 2A). Furthermore, the H[2]O[2] concentration of the ΔAhp strain was significantly higher than that of the WT strain ([61]Fig. 2B). The intracellular H[2]O[2] levels were submicromolar in both strains. Fig. 2. [62]Fig. 2 [63]Open in a new tab (AB). Comparison of ROS and H[2]O[2] concentrations between the two strains. The ROS and H[2]O[2] levels were significantly higher in the ΔAhp (red) strain than in the WT (gray) strain. (CD). Morphology of the WT and ΔAhp strains under optical microscopy. (EF). Cell morphology of the WT and ΔAhp strains under scanning electron microscopy. (G). The strain length was calculated between two strains (n = 30). The length of the ΔAhp strain was found to be significantly longer than that of the WT strain. (For interpretation of the references to color in this