Specific Enrichment and Proteomics Analysis of Escherichia coli

Subscriber access provided by REGIS UNIV

Article

Specific Enrichment and Proteomics Analysis of Escherichia coli Persisters from Rifampin Pretreatment Jordy Evan Sulaiman, Chunlin Hao, and Henry Lam J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00625 • Publication Date (Web): 11 Oct 2018 Downloaded from http://pubs.acs.org on October 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Specific Enrichment and Proteomics Analysis of Escherichia coli Persisters from Rifampin Pretreatment Jordy Evan Sulaiman†, Chunlin Hao†, Henry Lam†, * †

Department of Chemical and Biological Engineering, The Hong Kong University of Science &

Technology, Clear Water Bay, Kowloon, Hong Kong *

Corresponding author: [email protected]

Abstract Bacterial persisters, a dormant and multi-drug tolerant subpopulation that are able to resuscitate after antibiotic treatment, have recently received considerable attention as a major cause of relapse of various infectious diseases in the clinic. However, due to their low abundance and inherent transience, it is extremely difficult to study them by proteomics. Here, we developed a magnetic beads-based separation approach to enrich Escherichia coli persisters and then subjected them to shotgun proteomics. Rifampin pretreatment was employed to increase persister formation, and the resulting cells were exposed to high concentration of ampicillin (10× MIC) to remove nonpersisters. The survivors were analyzed by spectral counting-based quantitative proteomics. On average, 710 proteins were identified at a false discovery rate of 0.01 for enriched E. coli persisters. By spectral counting-based quantification, 105 proteins (70 down-regulated, 35 up-regulated) were shown differentially expressed compared to normal cells. Comparison of the differentially expressed proteins between the magnetic beads-enriched persisters and non-enriched persisters (a mixture of persisters and intact dead cells) shows only around half (~58%) overlap and different protein-protein interaction networks. This suggest that persister enrichment is important to eliminate the cumulative effect of dead cells that will obscure the proteome of persisters. As Page 1 of 45 ACS Paragon Plus Environment

Journal of Proteome Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

expected, proteins involved in carbohydrate metabolism, fatty acid and amino acid biosynthesis, and bacterial chemotaxis were found to be down-regulated in the persisters. Interestingly, membrane proteins including some transport proteins were up-regulated, indicating that they might be important for the drug tolerance of persisters. Knockout of the pal gene expressing peptidoglycan-associated lipoprotein – one of the most up-regulated protein detected in persisters – led to 10-fold reduced persister formation under ampicillin treatment.

Keywords: dormancy, resistance, tolerant, persistence, antibiotic, proteomics

Page 2 of 45 ACS Paragon Plus Environment

Page 2 of 45

Page 3 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction When a bacterial culture is treated with lethal doses of antibiotics, there will be a small subpopulation of cells that survives the treatment. These survivors are typically multi-drug tolerant, and exhibited two dormancy states, the viable but non-culturable (VBNC) state and bacterial persistence.1 While persisters will grow immediately upon removal of antibiotics as populations with antibiotic sensitivities similar to the original culture2-5, VBNC cells have lost their ability to grow on media despite being viable.6 Although these two states of existence are conventionally described as two different phenotypes, recent findings have shown that they are actually a part of a shared dormancy continuum and may be better understood as the same bacterial stress state.7,8 Persisters are enriched in biofilms (100 to 1000 folds higher than planktonic cultures)9-11, a slimy extracellular polymeric matrix that protect them from external stresses, and are prevalent during infections.12,13 This signifies a great health concern as the resuscitation of persisters after antibiotic treatment are thought to be one of the reasons for the relapses of biofilm infections.14,15 Unlike resistance which is acquired through genetic mutations that enable growth at higher antibiotic concentrations by decreasing the effectiveness of the drug, persistence is known to be a rare phenotypic feature that allows a small subpopulation of bacteria to survive antibiotics treatment, commonly attributed to transient growth inhibition and the resulting inactivity of essential cell functions.16-20 Even though persistence is different from resistance, it was widely conjectured that persisters could serve as an evolutionary reservoir from which resistance eventually develops.21-23 Persisters could be found in all growth phases, with a significantly higher populations in the stationary phase compared to the exponential phase (103 to 104 times more)24,25, further supporting the idea that persistence is related to a lower metabolic state.26-28 Toxin-antitoxin (TA) systems, which constitute a population maintenance mechanism in bacteria, are perceived to be



the key players in the formation of persisters.29-31 A TA system consists of a stable toxin that targets essential cellular functions and an unstable antitoxin which negates the toxin’s activity.32,33 Efforts in studying molecules related to persistence has made it possible to propose a mechanism for persister formation, where stresses such as antibiotic exposure will induce a cascade of activation and inhibition of signaling molecules, eventually leading to a decrease in the antitoxin level and/or an increase in the toxin level.34,35 The higher toxin activity inside the cells will interfere essential cellular processes and inhibit their growth, thus forming persisters. When the stress is removed, the rise in the antitoxin level will inactivate the excessive toxins and resume cells’ growth. While some molecules responsible for the persisters formation were already discovered, more global knowledge about the overall adaptation of persisters under antibiotic stress is desirable and can best be obtained by proteomics methods. To date, proteomics has been applied mostly to study persisters in biofilm-forming microbes such as M. tuberculosis and C. albicans. Previous proteomic studies on M. tuberculosis has identified several uniquely expressed candidate proteins which could be attractive targets for therapeutic intervention of tuberculosis infection. These proteins are Dop protein that maintain protein degradation during dormant state and DosR regulon proteins that were discovered to be abundant in the dormant state of M. tuberculosis (accounting for 20% of the total protein biomass). 36-38

Similarly in C. albicans persisters, aside from the subdued major metabolic activities, a

specific protein called alkyl hydroperoxide reductase 1 (AHP1) and some other proteins involved in virulence and stress response were found to be greatly up-regulated, indicating their importance for C. albicans tolerance.39,40 Although biofilm persisters have been increasingly a subject of proteomics, proteomic study of persisters in planktonic cultures is currently limited by the technical difficulties in their isolation. Hu et al. has previously attempted to perform proteomics


Page 4 of 45

Page 5 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analysis on E. coli persisters by overproducing YafQ toxin, but a large fraction of cells analyzed were likely not persisters, since they did not apply antibiotic treatment to kill the non-persisters and isolate the persisters for proteomics analysis. Therefore, the profiled proteome might be confounded by the presence of normal growing cells.41 While the isolation of persisters in biofilms is straightforward as they are trapped under a polymeric matrix and could be obtained by scraping and vortexing the biofilms after excessive washing to remove dead cells, isolation of persisters in planktonic cultures is more problematic as they are mixed together with the dead, but still intact, cells. Some previous non-proteomic studies of persisters attempted to isolate planktonic persisters by applying antibiotic stress and then harvest the persisters by centrifugation42,43, but the dead and intact cells would certainly still be pelleted together with the persister cells, possibly affecting the subsequent proteomic analysis in our case. Hence, the goal of this work is to obtain a comprehensive proteome profile of the persisters in non-biofilm forming bacteria, in comparison to normal cells. This information would provide an unprecedented glimpse into the global response of persister populations under antibiotic treatment, which could be used to formulate better models for their multi-drug tolerances and potentially therapeutic intervention in treating non-biofilmforming bacterial infections. As previously mentioned, the inherently transient nature and low abundance of nonbiofilm-forming persisters has thus far made it difficult for cell isolation and analysis by proteomics.44 Recently, Kwan et al. has successfully found a way to induce high level of persistence (10% to 100%) by using chemical pretreatments to mimic the effects of specific toxins inside the cell.45 In this work, we adopt the chemical pretreatment method using rifampin to increase the persister population for proteomic analysis. Then, ampicillin treatment was used to kill the non-persisters. To eliminate the effect of dead cells accumulation caused by antibiotic



treatment, a magnetic beads-based approach has been developed to enrich and isolate the persisters. This technique is more suitable for persisters enrichment compared to other techniques such as fluorescence assisted cell sorting (FACS)46, which needs a long time to accumulate large number of persisters for proteomics study and may give enough time for persister resuscitation. The magnetic-beads method demonstrated in this work provides a quick, high-specificity and highthroughput method to enrich living cells for proteomic analysis. Here, we report a deep quantitative proteomics analysis of the E. coli persisters revealing their proteomic changes, which could shed light on the drug-tolerant mechanisms of persisters.


Page 6 of 45

Page 7 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2. Methods Bacterial Strains and Growth Conditions Bacterial strain used in this study is E. coli K-12 MG1655. Mid-exponential phase cultures were prepared by incubating a 1:1000 diluted overnight culture in Luria-Bertani (LB) broth at 37⁰C at 220 r.p.m until 𝐴600 𝑛𝑚 = 0.7 was reached. LB agar was used for colony counts.

Minimum Inhibitory Concentration (MIC) Assay Single colony isolation was performed by streaking E. coli K-12 from frozen stock onto LB agar and incubated at 37⁰C for 16 h. A randomly selected single colony was inoculated in 3 ml of LB broth and incubated for 4 h until the suspension’s turbidity was similar to that of a McFarland 0.5 𝐵𝑎𝑆𝑂4 turbidity standard (𝐴625 𝑛𝑚 = 0.08-0.13).47 The MICs of E. coli K-12 to rifampin and ampicillin were determined by incubating the cultures for 16 h with various concentrations of rifampin and ampicillin (0.03 – 64 µg/ml), and inhibition of growth was observed based on lack of turbidity. Experiments were performed with three independent cultures.

Persister Viability Assay The schematic of persisters viability assay is shown in Fig. 1. Mid-exponential phase culture of E. coli K-12 was subjected to rifampin (100 µg/ml) pretreatment for 30 min, centrifuged and resuspended in 3 ml of fresh LB broth, followed by ampicillin (100 µg/ml) antibiotic treatment for 3 h. Concentration of ampicillin was chosen to be about 10× the MIC (8 µg/ml) to prevent the emergence of spontaneous resistant E. coli, and the 3 h duration was used to completely eliminate non-persisters. Concentration of rifampin (100 µg/ml) was also significantly higher than the MIC (16 µg/ml) and chosen to produce the highest survivability.45 For prolonged antibiotic treatment, Page 7 of 45 ACS Paragon Plus Environment


ampicillin antibiotic treatment was extended to 7 h, and the number of persisters were determined every 1 h. For comparison, normal E. coli K-12 without any pretreatment was also treated with ampicillin antibiotic. The number of persisters were determined based on an assessment of cell viability after ampicillin antibiotic treatment by serially diluting cultures in LB broth, plating 100 µl on LB agar and spread plates. Experiments were performed with three independent cultures.

Fig. 1. Experimental design of persisters viability assay. Rifampin (100 µg/ml) pretreatment was used for 30 min, and ampicillin (100 µg/ml) treatment was used for 3 h. For prolonged antibiotic exposure, ampicillin treatment was used for 7 h, and plating were done every 1 h to determine the percentage survival of cells.

Epifluorescence Microscopy Cultures were then stained with LIVE/DEAD BacLight bacterial viability kit (Molecular Probes) according to the manufacturer’s standard protocol. Briefly, 3 µl of dye mixture containing SYTO 9 (1.67 mM) and propidium iodide (10 mM) was added to 1 ml culture and incubated in


Page 8 of 45

Page 9 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the dark for 15 min. Stained cells were viewed with fluorescence microscope (Eclipse Ni-U Upright Microscope) with appropriate filter sets. Images were captured with Nikon DS-Fi3 and associated software (NIS-Elements Ver. 5.00).

Persister Isolation and Enrichment MACSflex™ MicroBead Kit (Miltenyi Biotec) was used to isolate persisters from the ampicillin-treated population containing both persisters and dead cells. 1.8 ml of MACSflex™ Reconstitution Buffer was mixed with 200 µl of 1.0 mg/ml propidium iodide (PI), and 2.0 ml of the mixture was added to a glass vial containing 2 mg of lyophilized MACSflex™ MicroBeads. The MicroBeads and PI was incubated at room temperature for 2 hours under dark condition, during which PI was covalently linked to the MicroBeads via a reaction between the amino group in PI and the N-hydroxysuccinimide (NHS) groups on the MicroBeads. The resulting PIfunctionalized MicroBeads were stored at 4 ⁰C overnight before usage. Persisters were formed by exposing mid-exponential phase culture of E. coli K-12 with rifampin (100 µg/ml) pretreatment for 30 min, followed by ampicillin (100 µg/ml) antibiotic treatment for 3 h. The cells were harvested and resuspended in 0.85% NaCl solution. 100 µl of cell suspension was mixed with 100 µl of PI-functionalized MicroBeads, and let to react at room temperature for 15 min. The PIfunctionalized MicroBeads can get inside intact dead cells with compromised cell membranes and bind to the DNAs, but unable to enter the living cells with normal cell membranes. Therefore, only the dead cells should be captured on the MicroBeads. For dead cell removal, a µ Column (Miltenyi Biotec) was placed on the µMACS Separator under a strong magnetic field, and rinsed with 50 µl of MACSflex™ Reconstitution Buffer. The column was first filled with 200 µl of cells and the PI-functionalized MicroBeads mixture and



washed with 120 µl of MACSflex™ Storage Buffer three times. The eluent are the living cells, as the dead cells are retained in the µ Columns due to the binding to the Microbeads through PI. To elute the dead cells, µ Columns was removed from the µMACS Separator and the magnetic field, and the column was washed with 70 µl of MACSflex™ Storage Buffer twice. Performance of the magnetic beads to isolate the living cells population and the dead cells population was investigated using fluorescence microscopy. Cell suspension from the initial population containing both living cells and dead cells, living cells population (which is the eluent from the column under the high magnetic field), and dead cells population (which is the eluent from the column without the magnetic field) were stained with LIVE/DEAD BacLight bacterial viability kit and viewed under fluorescence microscope.

Sample Preparation To generate persisters, mid-exponential phase culture of E. coli K-12 was subjected to rifampin (100 µg/ml) pretreatment for 30 min and ampicillin (100 µg/ml) treatment for 3 h. The persisters were isolated and enriched using MicroBeads to remove the dead cells. The control sample used in this study is the mid-exponential phase culture of E. coli K-12 without any antibiotic treatment. Three biological replicates were performed for each persisters and control sample, and two technical replicates were performed for each biological replicate. The cell pellet was suspended in 300 µL of lysis buffer (0.5% SDS, 50 mM Tris-HCl pH 8.0, and 25 mM dithiothreitol), incubated at 95 ⁰C for 5 minutes, and sonicated for 3 min. The sample was centrifuged to remove cell debris and insoluble materials. Aliquot of the sample was taken for BCA protein assay (Pierce™ BCA Protein Assay Kit). For shotgun proteomics, 200 µg of proteins were mixed with 250 µl of the exchange buffer (6M Urea, 50 mM Tris-HCl pH 8.0,


Page 10 of 45

Page 11 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


600 mM guadinine HCl), transferred to Amicon® filter device (Millipore, Darmstadt, Germany), and centrifuged. Then, 250 µl of the exchange buffer was added once again to the filter device, centrifuged, and the filtrate in the collection tube was discarded. The proteins in the filter device was alkylated with iodoacetamide (IAA, 50 mM in exchange buffer) in dark for 20 min, and then centrifuged. 250 µl of the exchange buffer was added the filter device, centrifuged, and the filtrate in the collection tube was discarded. To dilute the urea concentration, 250 µl of 50 mM ammonium bicarbonate was added to the filter device, centrifuged, and this step was repeated once again. Proteins were digested by sequencing-grade modified trypsin (1:100 w/w, Promega, Madison, WI) for 12 h at 37 ⁰C. Then, the sample was acidified with 10% formic acid to a final concentration of 0.5% (v/v) and desalted by C18 reverse-phase Spin Tip. Finally, samples were dried with SpeedVac (Eppendorf, Hamburg, Germany) and stored at -20 ⁰C before use.

LC-MS/MS We adopt standard shotgun proteomics techniques48 to process the sample through the Thermo Scientific Accela HPLC system coupled to a dual cell linear ion trap MS, LTQ Velos (Thermo Fisher Scientific, Bremen, Germany), which is interfaced to a nano-electrospray ion source. Approximately 10 µg of the protein digest was injected to the HPLC system and separated on a C18 column (Thermo Bio-Basic-18, 150 × 0.1 mm, 300 Å pore size, 5 µm particle size) at a flow rate of 220 µl/min. The mobile phase composition is 0.1% formic acid in water for solvent A, and 0.1% formic acid in acetonitrile for solvent B. The gradient was applied from 2 to 10% of solvent B for 4 min, from 10 to 32% of solvent B for 60 min, from 32 to 60% of solvent B for 6 min, and then from 60 to 80% of solvent B for 2 min. At the end, the mobile phase was kept at 80% of solvent B for 2 min, and then decreased to 2% of solvent B for 2 min. 14 minutes equilibration



Page 12 of 45

with 2% of solvent B was applied before the next injection. The m/z range acquired in the MS full scan was 400 to 2000 Da. The top 10 most intense precursor ions were selected to the MS 2 by using collision-induced-dissociation (CID) with dynamic exclusion of 60s, and the singly charged precursor ions excluded from the MS2. The MS2 isolation width was 2.0 m/z, and the normalized collision energy was set at 30%.

Sequence Database Searching The raw data acquired by the Thermo Scientific LTQ Velos were converted to mzML files by msconvert of the ProteoWizard (version 3.0.11676 64-bit).49 The mzML files were searched using Comet (version 2016.01 rev.2) against the E. coli K-12 protein sequence database obtained from UniProt. The sequences of common contaminants, such as trypsin and human keratins, and decoy sequences generated by shuffling amino acid sequences between tryptic cleavage sites were added to the database. The decoy sequence in the database are used for the false discovery rate (FDR) estimation of the identified peptides. The search parameters criteria were set as follows: 3 amu peptide mass tolerance, monoisotopic mass type, fully digested enzyme termini, 1.005 amu fragment bin tolerance, 0.4 amu fragment bin offset, carbamidomethylated cysteine and oxidated methionine as the fixed and variable modifications respectively. The search results from Comet were processed by PeptideProphet50, iProphet and ProteinProphet of the Trans-Proteomics Pipeline (TPP)51 in the decoy-assisted non-parametric mode. Every mzML run was analyzed independently. Protein identifications were filtered at a false discovery rate of 0.01 as predicted by ProteinProphet.

The

mass

spectrometry

proteomics

data

have

been

deposited

ProteomeXchange52 via the PRIDE53 partner repository with the dataset identifier PXD010590.


to

Page 13 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Label-Free Quantification by Spectral Counting The search results from the two technical replicates of each biological replicate were combined, and the proteins identified in at least two out of three biological replicates were used for label-free quantification by spectral counting. The quantification of proteins were given by the normalized spectral abundance factor (NSAF),54 where the number of peptide-spectrum matches (PSMs) for each protein divided by the length of the corresponding protein is normalized to the total number of PSMs divided by the lengths of protein for all identified proteins. Only proteins with average spectral counts (across all runs) of at least 5 were considered for quantification. Student’s t-test was employed on the NSAF values to detect differential expression between the persister group and the control group. The Benjamini-Hochberg (BH) multiple testing correction was applied to the t-test p-values to control the false discovery rate (FDR) at 10%.55 To further reduce false discoveries and limit our attention to the more highly regulated proteins, only proteins with fold changes higher or lower than ± 2-folds were considered differentially expressed in our subsequent analysis.

Bioinformatics Analysis To reveal the localization distribution of proteins obtained in the persisters, PSORTb v3.056 was used. To highlight potentially important proteins among the differentially expressed proteins, STRING v10.557 was used to predict the protein-protein interactions, and Medusa58 was used to visualize the interactions network. The interaction networks of the differentially expressed proteins between enriched and non-enriched persisters samples were compared.



PCR-targeted Gene Replacement The pal gene deletion in the E. coli K-12 BW25113 was accomplished through λ-red mediated recombination as previously described by Datsenko and Wanner.59 The detailed gene replacement method was described in the Supporting Information. The strains and plasmids used for the gene replacement are listed in the Supplementary Table S1 in the Supporting Information, and the PCR primers used for the gene replacement are listed in the Supplementary Table S2 in the Supporting Information. The Δpal mutant strain was annotated as JES1 and their survival (%) under ampicillin treatment was measured and compared to the wild type strain.


Page 14 of 45

Page 15 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3. Results The overview of the experimental and data analysis procedure was shown in Fig. 2. The persisters were generated by rifampin and ampicillin antibiotic treatment, isolated, and then subjected to standard shotgun proteomic technique. One should note that our method of enrichment and isolation would not distinguish between persisters and viable but non-culturable (VBNC) cells as they were conventionally classified, but we followed Kwan et al.45 in referring to the rifampinpretreated cells as persisters. The Filter Aided Sample Preparation (FASP) method was used for sample preparation60 and the resulting digest was analyzed by LC-MS/MS without any prefractionation. For data analysis, the differentially expressed proteins in enriched persisters in comparison to normal cells were used to provide hints about the antibiotic tolerance mechanism of the persisters. In addition, the comparison between the differentially expressed proteins from enriched persisters (as compared to normal cells) and those from non-enriched persisters (as compared to normal cells) would serve to demonstrate the necessity of the isolation process.

Fig. 2. Schematic diagram of experimental and data analysis procedure.



Characterization of Persisters under Ampicillin Treatment In this study, rifampin was used as the chemical pretreatment to increase the number of persisters, so as to facilitate proteomics analysis. Rifampin is an antibiotic that specifically binds and inhibits bacterial RNA polymerases, which is the enzyme responsible for DNA transcription.61 This mechanism of action is similar to MqsR toxin, the first TA system linked to biofilm and persister formation in E. coli, where MqsR overproduction leads to massive mRNA cleavage and halted transcription.29 Therefore, rifampin pretreatment should in theory increase persistence. Ampicillin, a β-lactam antibiotic, was chosen to be the bactericidal antibiotic for treatment to screen for persister cells. Ampicillin inhibits the enzyme transpeptidase, which is involved in the final stage of cell wall synthesis (peptidoglycan cross linking), leading to cell lysis of nonpersisters.62 The measured MICs of rifampin and ampicillin were 16 µg/ml and 8 µg/ml respectively, and this result is comparable to the one obtained by Kwan et al.45 From the persisters viability assay as shown in Fig. 3a, we could see that the culture pretreated with rifampin exhibit a significant increase in persistence (~15,000-fold increase over untreated culture), with 56% of the population surviving ampicillin treatment. Our result is comparable to that reported by Kwan et al.45 where they observed that the rifampin-induced persisters was consistently between 10 to 100%. This result also illustrated why the pretreatment is necessary, since without it only ~0.004% of the cells would become persisters, too few for proteomic analysis. In Fig. 3b, it was observed that untreated cultures of E. coli K-12 displayed biphasic cell death, where the number of cells initially decreases sharply in the first phase, and the remaining cells (persisters) surviving longer times of ampicillin treatment, showing slow and steady cell death, in the second phase.32,63 To see whether the rifampin-pretreated cultures possess this persisters characteristics, they were exposed to ampicillin (100 µg/ml) for a prolonged period,


Page 16 of 45

Page 17 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and the number of viable cells are quantified every 1 h. Our results show that the rifampinpretreated cultures exhibit slow and steady cell death for up to 7 h of ampicillin treatment, further verifying that our rifampin-induced persisters are “true” persisters. After 0, 1, and 3 h of ampicillin antibiotic exposure, 500 µl of both pretreated and nonpretreated cultures of E. coli K-12 were removed, stained, and observed under fluorescence microscope. As seen from the microscope images in Fig. 3c (1× dilution), normal cells without pretreatment died rapidly upon exposure to ampicillin. From 0 h to 1 h, there is a significant decrease in the number of green (live) cells, consistent with the data from Fig. 3b, where a sharp decrease in cell number happened in the first hour. The number of intact dead cells (red color) also increased with time and far outnumbered living cells by 3 h. On the contrary, the rifampinpretreated culture exhibit high survivability upon ampicillin treatment. Only a small number of dead cells are found after ampicillin exposure, even after 3 h duration. The higher number of dead cells from the non-pretreated cultures could also be seen from the 100× dilutions samples.



Fig. 3. Characterization of E. coli persisters induced with rifampin pretreatment compared to normal cells without pretreatment, both exposed to ampicillin antibiotic. (a) Survival (%) of cells Page 18 of 45 ACS Paragon Plus Environment

Page 18 of 45

Page 19 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


after ampicillin antibiotic (100µg/ml) treatment for 3 h. (b) Time-kill curve upon ampicillin treatment for up to 7 h. Data from three independent cultures are shown along with one standard deviation. (c) Samples were stained with a LIVE/DEAD kit and visualized by epifluorescence microscopy at indicated time points. Two culture dilutions were shown; 1× and 100×. Green cells are living cells while red cells are dead cells.

Persister Isolation and Enrichment After generating persisters using the aforementioned treatment, persister isolation and enrichment is necessary to prevent the dead cells from interfering with the proteomics analysis. Since we generated persisters by exposing the cells to a bactericidal antibiotic, the mixture of cells will contain a significant number of dead, but still intact, cells, which would be pelleted along with the persisters. Besides, the isolation technique used must be able to process a high cell number for proteomic analysis within a short time to prevent persisters resuscitation. The magnetic-beads based approach provide a high-throughput living cells isolation with persister cell recovery of ~30%, and a higher cell recovery of ~53% for normal-growing cells, estimated by CFU counting of cells after the isolation process (3 biological replicates). The loss of persister cells is likely due to cells disruption because of the applied pressure in the column during the isolation process (e.g. repeated washing), as shown by the lower recovery of the more fragile persisters compared to normal-growing cells. Besides, MicroBeads separation also exhibit high specificity in isolating the living cells population. Fig. 4 shows the microscope images of the initial cell population before isolation containing both living and dead cells, isolated living cells population, and the isolated dead cells population. It could be seen in the images that the isolated living cells population is free from dead cells (Fig. 4b). Even though we could see some living cells in the isolated dead cells



population due to nonspecific binding and trapping in the magnetic beads (Fig. 4c), we found out that the number is not significant ( ±1). The reproducibility of the experimental workflow was evaluated using technical replicates obtained from the same biological samples and the same preparation procedures but prepared separately and injected in random order through LC-MS/MS. For all groups of samples, comparison of spectral counts for the protein identified in the two technical replicates in each biological replicate shows a good agreement with R2 > 0.918, slope ~1, and intercept ~0 (Supplementary Figure S1 in the Supporting Information). This indicates that our experimental workflow has good reproducibility. A total of 117 proteins (78 proteins were down-



regulated and 39 proteins were up-regulated) were differentially expressed in non-enriched persisters sample compared to normal cells, whereas 105 proteins (70 proteins were downregulated and 35 proteins were up-regulated) were differentially expressed in enriched persisters sample compared to normal cells. As expected, there are more down-regulated proteins compared to up-regulated proteins due to the inherent nature of persisters. The list of differentially expressed proteins are shown in the Supplementary Table S4 in the Supporting Information. While most of the down-regulated proteins are those involved in carbohydrate metabolism such as glycolysis, pyruvate metabolism, pentose-phosphate pathway and tricarboxylic acid (TCA) cycle, the up-regulated proteins are mostly membrane and transport proteins. As shown in Fig. 6, within the protein-protein interaction network of the up-regulated proteins in enriched persisters, there is a sub-network of membrane proteins (labeled by red circle) which shows a significant upregulation from 2 to 4 folds. Besides membrane proteins, many ribosomal proteins (labeled by the blue circle) were also observed to be up-regulated. The inset figure of Fig. 6b shows the localization distribution of the up-regulated proteins that consists of two major components: membrane proteins (~34%) and cytoplasmic proteins (~60%), the latter of which are mostly ribosomal proteins. It is worth noting that the differentially expressed proteins observed in this study were consistent with previous report by Li et al. on investigating the proteomic profiles of C. albicans biofilm persister fractions by LC-MS.39 They observed that proteins involved in glycolysis, TCA cycle, and protein synthesis were down-regulated, while heat shock proteins, stress response proteins, and membrane proteins, especially those who plays a role for maintaining membrane integrity, are significantly up-regulated.


Page 22 of 45

Page 23 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 5. Protein identification and quantification of non-enriched persisters and enriched persisters compared to normal cells. Venn diagrams for proteome comparison on (a) non-enriched persisters and normal cells, and (b) enriched persisters and normal cells. Volcano plots for (c) non-enriched persisters compared with normal cells and (d) enriched persisters compared with normal cells. Differentially expressed proteins are defined to be those with Benjamini-Hochberg-corrected pvalues below 0.1, and fold change higher or lower than ±2, corresponding to the rectangular regions. Red rectangular regions are the down-regulated proteins and blue rectangular regions are the up-regulated proteins.



Fig. 6. Protein-protein interaction network of the up-regulated proteins in enriched persisters compared with normal cells as predicted by STRING and visualized by Medusa. Each dot represents a protein found to be up-regulated in persisters with different color representing different fold changes. The line between two dots represent the highest probability of the protein interactions as predicted by STRING. Proteins in the red circular region are the membrane proteins, and proteins in the blue circular region are the ribosomal proteins. The inset figure shows the localization distribution of the up-regulated proteins as predicted from PSORTb v3.0.

Fig. 7a and b shows the protein-protein interaction network of the differentially expressed proteins from both non-enriched and enriched persisters compared to normal cells respectively, as predicted by STRING. Notably, the network for proteins involved in glycolysis/gluconeogenesis, pyrimidine metabolism, and some membrane proteins are missing in the non-enriched persisters group. This indicated that the proteins from the dead cells might obscure some important proteome changes in the persisters.


Page 24 of 45

Page 25 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 7. Differentially expressed proteins of (a) non-enriched persisters compared to normal cells, and (b) enriched persisters compared to normal cells, depicted in their interaction networks by STRING v10.5. Each dot is a protein found to be differentially expressed as determined by labelfree quantitative analysis, and the lines represent protein interaction as predicted by STRING (solid lines means high confidence and dotted lines means medium confidence). Dots in different color represents different protein functions. Nodes without function enrichment are hidden from the network. Page 25 of 45 ACS Paragon Plus Environment


4. Discussion Lower Carbohydrate Metabolism in Persisters The most down-regulated protein observed from the persisters is magnesium-transporting ATPase (over 50-folds down-regulation), which mediates magnesium influx from outside the cell to the cytosol, consuming ATP in the process. The lower ATP level in persisters might be the reason of the significant down-regulation of this protein. Furthermore, as Mg2+ is an important cofactor for many enzymes involved in carbohydrate metabolism, the insufficient Mg2+ inside the cells, together with the down-regulation of these proteins, may be a mechanism to achieve a lower cell metabolism in the persisters. Most of the down-regulated proteins are those related to carbohydrate metabolism and mainly involved in four metabolic pathways: pentose-phosphate pathway, glyoxylate cycle, glycolysis, and tricarboxylic acid (TCA) cycle. This result agrees with those obtained in previous proteomics studies on persisters in other biofilm-forming species.39,41 Transaldolase A, a protein that plays an important role in balancing the metabolites in the pentose-phosphate pathway, is found to be down-regulated by ~10 folds. Other down-regulated proteins in persisters are isocitrate lyase and malate synthase A and G, which are enzymes involved in the metabolic adaptation of the cell in response to environmental changes through the glyoxylate cycle. For proteins that plays a role in glycolysis, pyruvate kinase and glucose-6-phosphate isomerase are found to be downregulated. Some down-regulated proteins involved in TCA cycle are citrate synthase, aconitate hydratase A, isocitrate dehydrogenase and succinate dehydrogenase iron-sulfur subunit. We also observed that a protein involved in the biogenesis of iron-sulfur clusters, glutaredoxin 4, is downregulated by around 6 folds. It has been reported previously that iron-sulfur cluster binding proteins


Page 26 of 45

Page 27 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


are mostly down-regulated on persisters.36 In addition, proteins involved in acetyl-CoA biosynthesis are also down-regulated, such as acetate kinase and acetyl-coenzyme A synthetase. Besides carbohydrate metabolism, several proteins involved in bacterial chemotaxis such as ribose import binding protein RbsB, protein phosphatase CheZ, D-galactose-binding periplasmic protein are also down-regulated. The rest of the down-regulated proteins are those involved in fatty acid biosynthesis and amino acid biosynthesis. Because persisters and stationary-phase cells share some similarity in that they are both slow-growing, we compared our observed proteome profile on persisters to the previously acquired transcriptome and proteome profile on stationary-phase E. coli. In general, as E. coli move to the latter stages of growth (stationary phase), it was observed that proteins involved in carbohydrate metabolism, amino acid biosynthesis, iron ion homeostasis, and metal ion binding, continue to decrease.65 Our proteomics analysis revealed that similar to stationary phase E. coli culture, E. coli persisters also down-regulate proteins involved in carbohydrate metabolism, especially those who play a role in the TCA cycle. The proteins for some down-regulated genes on stationary-phase culture such as glcB (encoding malate synthase G) and ansB (encoding L-asparaginase 2) were also observed to be down-regulated in persisters in our experiment.28 However, we also observed some differences between the two groups. While stationary phase E. coli exhibit a significant increase in the expression of acs (encoding acetyl CoA synthetase), aceB (encoding malate synthase A), and aceA (encoding isocitrate lyase), persisters have a significantly reduced expression for these three proteins.66



Persisters Up-regulate Their Membrane Proteins Previous studies have proven that the non-replicating state of persisters alone is not sufficient for their survival and bacteriostatic conditions do not guarantee antibiotic tolerance.67-69 Using fluorescence-activated cell sorting (FACS), Brynildsen et al. have shown that most of dormant cells were actually not antibiotic-tolerant, even though persisters were enriched in the dormant fraction of a population.70 In other words, persisters are not merely non-growing cells and its formation involves specific qualitative changes of their physiology to enable survival under antibiotic stress. Our data provided a glimpse into such physiological changes of persisters. From our data analysis, we found out that most of the up-regulated proteins in persisters are membrane proteins, and these proteins are listed in Table 1. Among the 12 up-regulated membrane proteins, 10 of them are involved in molecules transport. These observed up-regulation in the membrane proteins could be either persisters’ inherent adaptation strategy to better survive antibiotic stress, or some sort of response due to ampicillin treatment. We believe that the former explanation is more likely. In our experimental design, we isolated the persisters, first by enrichment using rifampin pretreatment, and then killing the non-persisters by using ampicillin. It must be noted that the ampicillin treatment was given after the cells were pretreated with rifampin, which should have arrested transcription in these cells. In other words, the ampicillin was given after the persister cells entered a dormant state and should not trigger any specific responses. As shown from Fig. 3b, the rifampin-pretreated cells did not suffer the rapid decline in the first hour under high-dose ampicillin treatment, but rather exhibited slow and steady cell death, indicating that they were not affected by ampicillin from the beginning. Moreover, in the Hu et al. report on the proteomic study on E. coli by overexpressing YafQ toxin, up-regulation in membrane proteins were similarly observed without any antibiotic treatment.41 In addition, the up-regulation of


Page 28 of 45

Page 29 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


membrane proteins on persisters was not only seen on E. coli, but also on other biofilm-forming organism like C. albicans, which was treated using a completely different antibiotic, Amphotericin B.39 These evidences combined suggested that the up-regulation in the membrane proteins are an adaptation strategy of persisters responsible for the generic persister phenotype, rather than any specific response to antibiotics. This apparent relationship between persister formation and up-regulation of membrane proteins is perhaps not surprising. After all, despite the reduced metabolism in persisters, they still require an intact membrane to maintain their viability. In fact, the most common strategy to directly kill persisters is through extensive damage of bacterial membrane.71 It was first suggested by Hurdle et al. in 2011 that proteins integral to membrane function should be the ideal targets for anti-persister therapies72, and this suggestion was followed by the successful identification of membrane-acting anti-persister compounds.73-77 Besides the membrane proteins, we also see a significant increase in chaperone protein Skp, a protein in the periplasm which interact specifically with membrane proteins such as OmpA, OmpC, OmpF, and LamB. Many ribosomal proteins (30S and 50S subunits) are also observed to be up-regulated in persisters. As reported by Wilson et al., E. coli persister cells suppress translation by selectively disassembling their ribosomes, and the ribosomes on E. coli persisters exist largely as inactive ribosomal 30S and 50S subunits instead of functional ribosomal complexes.78 Therefore, persister cells may contain similar quantity of ribosomal proteins as normal cells, despite the former’s lower translational activity. Ribosomal proteins are also among the most abundant in cells, and since we subjected the same total amount of protein for both the control group and persister group to quantitative proteomics, the widespread down-regulation of most other proteins may result in an apparent increase in the relative quantities of the ribosomal proteins observed in our experiments.



Page 30 of 45

Table 1. List of up-regulated periplasmic and membrane proteins in enriched E. coli persisters UniProt Code

Gene

Protein Name

Fold change

P0AAD6

sdaC

Serine transporter

3.81

P0A912

pal

Peptidoglycan-associated lipoprotein

3.38

P69776

lpp

Major outer membrane prolipoprotein Lpp

3.15

P69805

manZ

PTS system mannose-specific EIID component

2.48

P0AAG8

mglA

Galactose/methyl galactoside import ATP-binding

2.47

protein MglA P0AG99

secG

Protein-export membrane protein SecG

2.45

P25714

yidC

Membrane protein insertase YidC

2.41

P02931

ompF

Outer membrane protein F

2.40

P10384

fadL

Long-chain fatty acid transport protein

2.34

P02943

lamB

Maltoporin

2.21

P0ABJ1

cyoA

Cytochrome bo(3) ubiquinol oxidase subunit 2

2.18

P06996

ompC

Outer membrane protein C

2.07

pal Gene Disruption Reduce Persisters Survival under Antibiotic Treatment To see whether membrane integrity is important for persisters, we knocked out the pal gene in E. coli K-12 BW25113 (mutant annotated as JES1). This gene expresses peptidoglycanassociated lipoprotein (Table 1) which was observed to be up-regulated by 3.4 folds in persisters. This protein plays a role in the outer membrane invagination during cell division and is claimed to be important to maintain the outer membrane integrity.79,80 Previous findings have reported that


Page 31 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


mutation in the pal gene confers a defect in outer membrane integrity resulting in hypersensitivity to drugs,81,82 but whether this gene affects the viability of persisters under antibiotic stress was still unknown. We exposed the Δpal mutant strain (JES1) and wild type strain (BW25113) to ampicillin (100µg/ml) antibiotic, and compare their survival (%) after treatment. Under prolonged ampicillin exposure (Fig. 8a), we could see that the normal-growing cells died rapidly in the first hour for both strains, similar to our observation on Fig. 3b which marks the first phase of killing in the biphasic cell death. However, one hour into ampicillin treatment, which is the approximate beginning of the dying phase of persisters, the persisters seems to die more quickly in the JES1 strain compared to wild type strain. After 4 hours of ampicillin treatment, the fraction of persisters in JES1 population is only ~4 ∙ 10−6 , which is around 10 folds less compared to the wild type strain (Fig. 8b). This result indicated that the mutant lacking pal gene is less able to form persisters, or form persisters that are less tolerant to antibiotic stress, compared to the wild type.

Fig. 8. Comparison of Δpal mutant strain (JES1) to the wild type strain (BW25113) under ampicillin treatment. (a) Time-kill curve upon ampicillin treatment (100µg/ml) for up to 7 h. Data from two independent cultures are shown along with one standard deviation. (b) Fraction of



persisters after treatment with ampicillin for 4 h. Data from three independent cultures are shown along with one standard deviation.

Importance of Persisters Enrichment Fig. 9 shows that most of the differentially expressed proteins identified from enriched persisters group are actually different from the non-enriched persisters group. Only around half (~58%) of the differentially expressed proteins are common in both groups. The differentially expressed proteins identified in one or both of the groups are listed in Supplementary Table S5 in the Supporting Information. This result shows that without the isolation process, the proteins identified will be different, since a large fraction of them come from the dead cells and not from the persisters. This difference on the identified proteins explains why the non-enriched persisters have a significant difference in the protein-protein interaction networks compared to the enriched persisters as shown in Fig. 7. Therefore, we propose that in any proteomics experiment involving samples with significant number of dead cells, the living cells need to be firstly isolated from the dead cells prior to subjecting them to sample preparation.


Page 32 of 45

Page 33 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 9. Comparison between differentially expressed proteins from enriched persisters compared to normal cells and non-enriched persisters compared to normal cells.

Conclusion In this study, we use rifampin pretreatment to increase E. coli persisters population, and ampicillin antibiotic to eradicate non-persisters, so as to isolate sufficient persister cells for proteomic analysis. We also developed a magnetic beads-based approach to isolate persisters and remove the effect of dead cells accumulation from the antibiotic treatment. This isolation method is shown to have a high throughput and high specificity in living cells enrichment for proteomic analysis. Although there were previous proteomics studies on persisters in other microbes, experimental difficulty in isolating persisters limited those studies to presumed persisters in biofilms. To our knowledge, this study represented the first investigation into the general phenomenon of persistence in a non-biofilm-forming microbe by proteomics. On average, 710 proteins were identified for enriched E. coli persisters. From spectral counting quantification, 105 proteins were differentially expressed compared to normal E. coli, with 70 down-regulated and 35 up-regulated proteins. Comparison of the differentially expressed proteins between the enriched persisters and non-enriched persisters shows only ~58% overlap, implying that persister isolation is important to minimize the cumulative effect of dead cells. Proteins involved in carbohydrate metabolism, fatty acid and amino acid biosynthesis, and bacterial chemotaxis were found to be down-regulated in the persisters. This lower metabolic activity agrees with the hypothesis that persisters are “dormant” cells. Interestingly, we found that membrane proteins were up-regulated, similar to previous studies on biofilm-forming persisters, which might be responsible for the generic persisters phenotype. We further tested this hypothesis by gene knockout of the pal gene Page 33 of 45 ACS Paragon Plus Environment


(which expresses peptidoglycan-associated lipoprotein, one of the most up-regulated protein in persisters as compared to normal cells), which leads to a 10-fold decrease in persister population, as measured by the survival rate under prolonged and high-dose ampicillin treatment. One should note that the ampicillin treatment, which was a necessary step to kill off non-persisters, may in theory affect the proteome of the survivors. Therefore, our approach to study persisters is dependent on the assumption that the rifampin-pretreated persisters were already dormant and did not respond actively to ampicillin treatment, such that the proteome changes observed in the survivors represent a persister phenotype. We believe this assumption is well-justified in view of the known mechanism of rifampin, as well as the agreement of our observations with previous studies of persisters using completely different antibiotics. Nonetheless, further studies using orthogonal approaches may be warranted to eliminate this potential confounding factor. In conclusion, this study demonstrated a new approach for studying persisters and provided fresh insight into the adaptation strategy of E. coli persisters at the proteome level.

Supporting Information Additional information on the gene replacement technique and proteomics data are provided in the Supporting Information. Method S1: PCR-targeted Gene Replacement technique. Table S1: Strains and plasmids used for gene knockout. Table S2: PCR primers used for gene knockout. Table S3: Unique proteins identified in enriched persisters sample. Table S4: Differentially expressed proteins of enriched persisters sample compared to normal cells. Table S5: Differentially expressed proteins of enriched and non-enriched persisters compared to normal cells. Figure S1: Reproducibility of spectral counts between technical replicates.


Page 34 of 45

Page 35 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Author Information ORCID: J. E. Sulaiman: 0000-0002-0886-8707 H. Lam: 0000-0001-7928-0364 Author Contributions: J. E. S. and C. H. designed the experiments. J. E. S. conducted the experiments and performed the data analysis. J. E. S. and H. L. conceived the study, devised the overall study strategy and wrote the manuscript. Competing interests The authors declare no competing financial interests.

Acknowledgement The authors acknowledge the funding support from Research Grants Council (Grant No. 16100415) of the Hong Kong Special Administrative Region Government, China. The strains and plasmids used for gene knockout in the study were kindly provided by Prof. Yi Yu from the School of Pharmaceutical Sciences, Wuhan University, China. The authors also want to thank the Biosciences Central Research Facility of the Hong Kong University of Science and Technology for access to the mass spectrometer.



References (1)

Ayrapetyan, M.; Williams, T.; Baxter, R.; Oliver, J. Viable but non-culturable and

persister cells coexist stochastically and are induced by human serum. Infect. Immun. 2015, IAI. 00404-00415. (2)

Helaine, S.; Kugelberg, E. Bacterial persisters: formation, eradication, and

experimental systems. Trends Microbiol. 2014, 22, 417-424. (3)

Gerdes, K.; Semsey, S. Microbiology: pumping persisters. Nature 2016, 534, 41.

(4)

Bigger, J. Treatment of Staphylococcal Infections with Penicillin by Intermittent

Sterilisation. Lancet 1944, 497-500. (5)

Bigger, J. W. The bactericidal action of penicillin on Staphylococcus pyogenes. Ir.

J. Med. Sci. 1944, 19, 553-568. (6)

Bamford, R. A.; Smith, A.; Metz, J.; Glover, G.; Titball, R. W.; Pagliara, S.

Investigating the physiology of viable but non-culturable bacteria by microfluidics and timelapse microscopy. BMC Biol. 2017, 15, 121. (7)

Ayrapetyan, M.; Williams, T. C.; Oliver, J. D. Bridging the gap between viable

but non-culturable and antibiotic persistent bacteria. Trends Microbiol. 2015, 23, 7-13. (8)

Kim, J. S.; Chowdhury, N.; Yamasaki, R.; Wood, T. K. Viable but non‐culturable

and persistence describe the same bacterial stress state. Environ. Microbiol. 2018, 20, 20382048. (9)

Fauvart, M.; De Groote, V. N.; Michiels, J. Role of persister cells in chronic

infections: clinical relevance and perspectives on anti-persister therapies. J. Med. Microbiol. 2011, 60, 699-709.


Page 36 of 45

Page 37 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(10)

Maisonneuve, E.; Castro-Camargo, M.; Gerdes, K. (p)ppGpp controls bacterial

persistence by stochastic induction of toxin-antitoxin activity. Cell 2013, 154, 1140-1150. (11)

Spoering, A. L.; Lewis, K. Biofilms and planktonic cells of Pseudomonas

aeruginosa have similar resistance to killing by antimicrobials. J. Bacteriol. 2001, 183, 67466751. (12)

Bjarnsholt, T. The role of bacterial biofilms in chronic infections. APMIS 2013,

121, 1-58. (13)

Lebeaux, D.; Ghigo, J.-M.; Beloin, C. Biofilm-related infections: bridging the gap

between clinical management and fundamental aspects of recalcitrance toward antibiotics. Microbiol. Mol. Biol. Rev. 2014, 78, 510-543. (14)

Lewis, K. Riddle of biofilm resistance. Antimicrob. Agents Chemother. 2001, 45,

999-1007. (15)

Singh, R.; Ray, P.; Das, A.; Sharma, M. Role of persisters and small-colony

variants in antibiotic resistance of planktonic and biofilm-associated Staphylococcus aureus: an in vitro study. J. Med. Microbiol. 2009, 58, 1067-1073. (16)

Balaban, N. Q.; Merrin, J.; Chait, R.; Kowalik, L.; Leibler, S. Bacterial

persistence as a phenotypic switch. Science 2004, 305, 1622-1625. (17)

Lewis, K. Persister cells, dormancy and infectious disease. Nat. Rev. Microbiol.

2007, 5, 48-56. (18)

Du Toit, A. Bacterial physiology: Persisters are under the pump. Nat. Rev.

Microbiol. 2016, 14, 332. (19)

Kaldalu, N.; Hauryliuk, V.; Tenson, T. Persisters-as elusive as ever. Appl.

Microbiol. Biotechnol. 2016, 100, 6545-6553.



(20)

Brauner, A.; Fridman, O.; Gefen, O.; Balaban, N. Q. Distinguishing between

resistance, tolerance and persistence to antibiotic treatment. Nat. Rev. Microbiol. 2016, 14, 320. (21)

Michiels, J. E.; Van den Bergh, B.; Verstraeten, N.; Fauvart, M.; Michiels, J. In

vitro emergence of high persistence upon periodic aminoglycoside challenge in the ESKAPE pathogens. Antimicrob. Agents Chemother. 2016, 60, 4630-4637. (22)

Cohen, N. R.; Lobritz, M. A.; Collins, J. J. Microbial persistence and the road to

drug resistance. Cell Host Microbe 2013, 13, 632-642. (23)

Levin-Reisman, I.; Ronin, I.; Gefen, O.; Braniss, I.; Shoresh, N.; Balaban, N. Q.

Antibiotic tolerance facilitates the evolution of resistance. Science 2017, 355, 826-830. (24)

Lewis, K.: Multidrug tolerance of biofilms and persister cells. In Bacterial

Biofilms; Springer, 2008; pp 107-131. (25)

Goneau, L. W.; Yeoh, N. S.; MacDonald, K. W.; Cadieux, P. A.; Burton, J. P.;

Razvi, H.; Reid, G. Selective target inactivation rather than global metabolic dormancy causes antibiotic tolerance in uropathogens. Antimicrob. Agents Chemother. 2014, 58, 2089-2097. (26)

Prax, M.; Bertram, R. Metabolic aspects of bacterial persisters. Front. Cell. Infect.

Microbiol. 2014, 4, 148. (27)

Amato, S. M.; Fazen, C. H.; Henry, T. C.; Mok, W. W.; Orman, M. A.; Sandvik,

E. L.; Volzing, K. G.; Brynildsen, M. P. The role of metabolism in bacterial persistence. Front Microbiol. 2014, 5, 70. (28)

Smith, A.; Kaczmar, A.; Bamford, R. A.; Smith, C.; Frustaci, S.; Kovacs-Simon,

A.; O’Neill, P.; Moore, K.; Paszkiewicz, K.; Titball, R. W. The culture environment influences both gene regulation and phenotypic heterogeneity in Escherichia coli. Front Microbiol. 2018, 9.


Page 38 of 45

Page 39 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(29)

Wang, X.; Wood, T. K. Toxin-antitoxin systems influence biofilm and persister

cell formation and the general stress response. Appl. Environ. Microbiol. 2011, 77, 5577-5583. (30)

Lewis, K. Persister cells. Annu. Rev. Microbiol. 2010, 64, 357-372.

(31)

Du Toit, A. Bacterial physiology: Persisters running out of energy. Nat. Rev.

Microbiol. 2017, 15, 194. (32)

Kint, C. I.; Verstraeten, N.; Fauvart, M.; Michiels, J. New-found fundamentals of

bacterial persistence. Trends Microbiol. 2012, 20, 577-585. (33)

Jayaraman, R. Bacterial persistence: some new insights into an old phenomenon.

J. Biosci. 2008, 33, 795-805. (34)

Maisonneuve, E.; Gerdes, K. Molecular mechanisms underlying bacterial

persisters. Cell 2014, 157, 539-548. (35)

Van den Bergh, B.; Fauvart, M.; Michiels, J. Formation, physiology, ecology,

evolution and clinical importance of bacterial persisters. FEMS Microbiol. Rev. 2017, 41, 219251. (36)

Albrethsen, J.; Agner, J.; Piersma, S. R.; Hojrup, P.; Pham, T. V.; Weldingh, K.;

Jimenez, C. R.; Andersen, P.; Rosenkrands, I. Proteomic profiling of the Mycobacterium tuberculosis identifies nutrient starvation responsive toxin-antitoxin systems. Mol. Cell. Proteomics 2013, mcp. M112. 018846. (37)

Gopinath, V.; Raghunandanan, S.; Gomez, R. L.; Jose, L.; Surendran, A.;

Ramachandran, R.; Pushparajan, A. R.; Mundayoor, S.; Jaleel, A.; Kumar, R. A. Profiling the proteome of Mycobacterium tuberculosis during dormancy and reactivation. Molecular & Cellular Proteomics 2015, mcp. M115. 051151.



(38)

Schubert, O. T.; Ludwig, C.; Kogadeeva, M.; Zimmermann, M.; Rosenberger, G.;

Gengenbacher, M.; Gillet, L. C.; Collins, B. C.; Röst, H. L.; Kaufmann, S. H. Absolute proteome composition and dynamics during dormancy and resuscitation of Mycobacterium tuberculosis. Cell Host Microbe 2015, 18, 96-108. (39)

Li, P.; Seneviratne, C. J.; Alpi, E.; Vizcaino, J. A.; Jin, L. Delicate metabolic

control and coordinated stress response critically determine antifungal tolerance of Candida albicans biofilm persisters. Antimicrob. Agents Chemother. 2015, AAC. 00543-00515. (40)

Truong, T.; Zeng, G.; Lin, Q.; Lim, T. K.; Cao, T.; Chan, F. Y.; Wang, Y.;

Seneviratne, C. J. Comparative ploidy proteomics of Candida albicans biofilms unraveled the role of AHP1 in the biofilm persistence against amphotericin B. Mol. Cell. Proteomics 2016, mcp. M116. 061523. (41)

Hu, Y.; Kwan, B. W.; Osbourne, D. O.; Benedik, M. J.; Wood, T. K. Toxin YafQ

increases persister cell formation by reducing indole signalling. Environ. Microbiol. 2015, 17, 1275-1285. (42)

Pu, Y.; Zhao, Z.; Li, Y.; Zou, J.; Ma, Q.; Zhao, Y.; Ke, Y.; Zhu, Y.; Chen, H.;

Baker, M. A. Enhanced efflux activity facilitates drug tolerance in dormant bacterial cells. Mol. Cell 2016, 62, 284-294. (43)

Keren, I.; Shah, D.; Spoering, A.; Kaldalu, N.; Lewis, K. Specialized persister

cells and the mechanism of multidrug tolerance in Escherichia coli. J. Bacteriol. 2004, 186, 8172-8180. (44)

Balaban, N. Q.; Gerdes, K.; Lewis, K.; McKinney, J. D. A problem of

persistence: still more questions than answers? Nat. Rev. Microbiol. 2013, 11, 587.


Page 40 of 45

Page 41 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(45)

Kwan, B. W.; Valenta, J. A.; Benedik, M. J.; Wood, T. K. Arrested protein

synthesis increases persister-like cell formation. Antimicrob. Agents Chemother. 2013, 57, 14681473. (46)

Shah, D.; Zhang, Z.; Khodursky, A. B.; Kaldalu, N.; Kurg, K.; Lewis, K.

Persisters: a distinct physiological state of E. coli. BMC Microbiol. 2006, 6, 53. (47)

Wiegand, I.; Hilpert, K.; Hancock, R. E. Agar and broth dilution methods to

determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat. Protoc. 2008, 3, 163-175. (48)

Domon, B.; Aebersold, R. Mass spectrometry and protein analysis. Science 2006,

312, 212-217. (49)

Kessner, D.; Chambers, M.; Burke, R.; Agus, D.; Mallick, P. ProteoWizard: open

source software for rapid proteomics tools development. Bioinformatics 2008, 24, 2534-2536. (50)

Keller, A.; Nesvizhskii, A. I.; Kolker, E.; Aebersold, R. Empirical statistical

model to estimate the accuracy of peptide identifications made by MS/MS and database search. Anal. Chem. 2002, 74, 5383-5392. (51)

Deutsch, E. W.; Mendoza, L.; Shteynberg, D.; Farrah, T.; Lam, H.; Tasman, N.;

Sun, Z.; Nilsson, E.; Pratt, B.; Prazen, B. A guided tour of the Trans‐Proteomic Pipeline. Proteomics 2010, 10, 1150-1159. (52)

Deutsch, E. W.; Csordas, A.; Sun, Z.; Jarnuczak, A.; Perez-Riverol, Y.; Ternent,

T.; Campbell, D. S.; Bernal-Llinares, M.; Okuda, S.; Kawano, S. The ProteomeXchange consortium in 2017: supporting the cultural change in proteomics public data deposition. Nucleic Acids Res. 2016, 54, D1100-D1106.



(53)

Vizcaíno, J. A.; Csordas, A.; Del-Toro, N.; Dianes, J. A.; Griss, J.; Lavidas, I.;

Mayer, G.; Perez-Riverol, Y.; Reisinger, F.; Ternent, T. 2016 update of the PRIDE database and its related tools. Nucleic Acids Res. 2015, 44, D447-D456. (54)

Paoletti, A. C.; Parmely, T. J.; Tomomori-Sato, C.; Sato, S.; Zhu, D.; Conaway,

R. C.; Conaway, J. W.; Florens, L.; Washburn, M. P. Quantitative proteomic analysis of distinct mammalian mediator complexes using normalized spectral abundance factors. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 18928-18933. (55)

Benjamini, Y.; Hochberg, Y. Controlling the false discovery rate: a practical and

powerful approach to multiple testing. J. R. Stat. Soc. Series B Stat. Methodol. 1995, 289-300. (56)

Yu, N. Y.; Wagner, J. R.; Laird, M. R.; Melli, G.; Rey, S.; Lo, R.; Dao, P.;

Sahinalp, S. C.; Ester, M.; Foster, L. J.; Brinkman, F. S. L. PSORTb 3.0: improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes. Bioinformatics 2010, 26, 1608-1615. (57)

Szklarczyk, D.; Franceschini, A.; Kuhn, M.; Simonovic, M.; Roth, A.; Minguez,

P.; Doerks, T.; Stark, M.; Muller, J.; Bork, P.; Jensen, L. J.; Mering, C. v. The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res. 2011, 39, D561-D568. (58)

Hooper, S. D.; Bork, P. Medusa: a simple tool for interaction graph analysis.

Bioinformatics 2005, 21, 4432-4433. (59)

Datsenko, K. A.; Wanner, B. L. One-step inactivation of chromosomal genes in

Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 6640-6645. (60)

Wiśniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Universal sample

preparation method for proteome analysis. Nat. Methods 2009, 6, 359.


Page 42 of 45

Page 43 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(61)

Wehrli, W. Rifampin: mechanisms of action and resistance. Rev. Infect. Dis.

1983, 5, S407-S411. (62)

Fisher, J. F.; Meroueh, S. O.; Mobashery, S. Bacterial resistance to β-lactam

antibiotics: compelling opportunism, compelling opportunity. Chem. Rev. 2005, 105, 395-424. (63)

Keren, I.; Kaldalu, N.; Spoering, A.; Wang, Y.; Lewis, K. Persister cells and

tolerance to antimicrobials. FEMS Microbiol. Lett. 2004, 230, 13-18. (64)

Mali, S.; Mitchell, M.; Havis, S.; Bodunrin, A.; Rangel, J.; Olson, G.; Widger, W.

R.; Bark, S. J. A Proteomic Signature of Dormancy in an Actinobacterium: Micrococcus luteus. J. Bacteriol. 2017, JB. 00206-00217. (65)

Soufi, B.; Krug, K.; Harst, A.; Macek, B. Characterization of the E. coli proteome

and its modifications during growth and ethanol stress. Front Microbiol. 2015, 6, 103. (66)

Bergholz, T. M.; Wick, L. M.; Qi, W.; Riordan, J. T.; Ouellette, L. M.; Whittam,

T. S. Global transcriptional response of Escherichia coli O157: H7 to growth transitions in glucose minimal medium. BMC Microbiol. 2007, 7, 97. (67)

Fung, D. K.; Chan, E. W.; Chin, M. L.; Chan, R. C. Delineation of a bacterial

starvation stress response network which can mediate antibiotic tolerance development. Antimicrob. Agents Chemother. 2010, 54, 1082-1093. (68)

Germain, E.; Roghanian, M.; Gerdes, K.; Maisonneuve, E. Stochastic induction of

persister cells by HipA through (p)ppGpp-mediated activation of mRNA endonucleases. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 5171-5176. (69)

Lewis, K. Persister cells and the riddle of biofilm survival. Biochemistry 2005, 70,

267-274.



(70)

Orman, M. A.; Brynildsen, M. P. Dormancy is not necessary or sufficient for

bacterial persistence. Antimicrob. Agents Chemother. 2013, 57, 3230-3239. (71)

Defraine, V.; Fauvart, M.; Michiels, J. Fighting bacterial persistence: Current and

emerging anti-persister strategies and therapeutics. Drug Resist. Updates 2018, 38, 12-26. (72)

Hurdle, J. G.; O'neill, A. J.; Chopra, I.; Lee, R. E. Targeting bacterial membrane

function: an underexploited mechanism for treating persistent infections. Nat. Rev. Microbiol. 2011, 9, 62. (73)

Mukherjee, D.; Zou, H.; Liu, S.; Beuerman, R.; Dick, T. Membrane-targeting

AM-0016 kills mycobacterial persisters and shows low propensity for resistance development. Future Microbiol. 2016, 11, 643-650. (74)

Feng, J.; Zhang, S.; Shi, W.; Zubcevik, N.; Miklossy, J.; Zhang, Y. Selective

essential oils from spice or culinary herbs have high activity against stationary phase and biofilm Borrelia burgdorferi. Front. Med. 2017, 4, 169. (75)

Bahar, A. A.; Liu, Z.; Totsingan, F.; Buitrago, C.; Kallenbach, N.; Ren, D.

Synthetic dendrimeric peptide active against biofilm and persister cells of Pseudomonas aeruginosa. Appl. Microbiol. Biotechnol. 2015, 99, 8125-8135. (76)

Moreira, W.; Aziz, D. B.; Dick, T. Boromycin kills mycobacterial persisters

without detectable resistance. Front Microbiol. 2016, 7, 199. (77)

Chen, X.; Zhang, M.; Zhou, C.; Kallenbach, N. R.; Ren, D. Control of bacterial

persister cells by Trp/Arg-containing antimicrobial peptides. Appl. Environ. Microbiol. 2011, 77, 4878-4885.


Page 44 of 45

Page 45 of 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(78)

Cho, J.; Rogers, J.; Kearns, M.; Leslie, M.; Hartson, S. D.; Wilson, K. S.

Escherichia coli persister cells suppress translation by selectively disassembling and degrading their ribosomes. Mol. Microbiol. 2015, 95, 352-364. (79)

Cascales, E.; Gavioli, M.; Sturgis, J. N.; Lloubès, R. Proton motive force drives

the interaction of the inner membrane TolA and outer membrane pal proteins in Escherichia coli. Mol. Microbiol. 2000, 38, 904-915. (80)

Gerding, M. A.; Ogata, Y.; Pecora, N. D.; Niki, H.; De Boer, P. A. The trans‐

envelope Tol–Pal complex is part of the cell division machinery and required for proper outer‐ membrane invagination during cell constriction in E. coli. Mol. Microbiol. 2007, 63, 1008-1025. (81)

Lazzaroni, J.-C.; Fognini-Lefebvre, N.; Portalier, R. Cloning of the excC and

excD genes involved in the release of periplasmic proteins by Escherichia coli K12. Mol. Gen. Genet. 1989, 218, 460-464. (82)

Webster, R. The tol gene products and the import of macromolecules into

Escherichia coli. Mol. Microbiol. 1991, 5, 1005-1011.

TOC Graphic


Specific Enrichment and Proteomics Analysis of Escherichia coli

Recommend Documents