Extraction and Visualization of Protein Aggregates after Treatment of Escherichia coli with a Proteotoxic Stressor

Bacteria are inevitably exposed to a myriad of environmental stresses, including low pH (e.g., in the mammalian stomach)^1,2, reactive oxygen and chlorine species (ROS/RCS) (e.g., during oxidative burst in phagocytes)^3,4,5, elevated temperatures (e.g., in hot springs or during heat-shock)^6,7, and several potent antimicrobials (e.g., AGXX used in this protocol)⁸. Proteins are particularly vulnerable to any of these stressors, and exposure can provoke protein un-/misfolding that then seeds aggregation. All organisms employ protective systems that allow them to cope with protein misfolding⁹. However, severe stress can overwhelm the protein quality control machinery and disrupt the secondary and/or tertiary structure of proteins, which ultimately inactivates proteins. As a consequence, protein aggregates can severely impair critical cellular functions required for bacterial growth and survival, stress resistance, and virulence¹⁰. Therefore, research focusing on protein aggregation and its consequences in bacteria is a relevant topic due to its potential impact on infectious disease control.

Heat-induced protein unfolding and aggregation are often reversible⁷. In contrast, other proteotoxic stresses, such as oxidative stress, can cause irreversible protein modifications through the oxidation of specific amino acid side chains resulting in protein un-/misfolding and, eventually, protein aggregation⁴. Stress-induced formation of insoluble protein aggregates has been extensively studied in the context of molecular chaperones and their protective functions in yeast and bacteria^11,12,13. Several protocols have been published that utilize a variety of biochemical techniques for the isolation and analysis of insoluble protein aggregates^14,15,16,17. The existing protocols have mainly been used to study bacterial protein aggregation upon heat-shock and/or identification of molecular chaperones. While these protocols have certainly been an advancement to the field, there are some major inconveniences in the experimental procedures because they require (i) a large bacterial culture volume of up to 10 L^14,17, (ii) complicated physical disruption processes, including the use of cell disruptors, French press, and/or sonication^14,15,17, or (iii) time-consuming repeated washing and incubation steps^15,16,17.

This paper describes a modified protocol that aims to address the limitations of the previous approaches and allows the analysis of the amount of protein aggregates formed in two different Escherichia coli strains after treatment with a proteotoxic antimicrobial surface coating. The coating is composed of metal-silver (Ag) and ruthenium (Ru)-conditioned with ascorbic acid, and its antimicrobial activity is achieved by the generation of reactive oxygen species^8,18. Herein is a detailed description of the preparation of the bacterial culture after treatment with the antimicrobial compound and a comparison of protein aggregation status upon exposure of two E. coli strains with distinct susceptibility profiles to increasing concentration of the antimicrobial. The described method is inexpensive, fast, and reproducible and can be used to study protein aggregation in the presence of other proteotoxic compounds. In addition, the protocol can be modified to analyze the impact that specific gene deletions have on protein aggregation in a variety of different bacteria.

1. Stress treatment of E. coli strains MG1655 and CFT073

1. Inoculate 5 mL of lysogeny broth (LB) medium with a single colony of commensal E. coli strain MG1655 and uropathogenic E. coli (UPEC) strain CFT073, respectively, and incubate for 14-16 h (overnight) at 37 °C and 300 rpm.
   NOTE: Escherichia coli CFT073 is a human pathogen. Handling of CFT073 must be performed with appropriate biosafety measures in a Biosafety Level-2 certified lab.
2. Dilute each strain into a 500 mL flask containing 70 mL of 3-(N-morpholino)propanesulfonic acid (MOPS)-glucose (MOPS-g) (Table 1) medium to an optical density at 600 nm (OD₆₀₀) value of 0.1. Incubate at 37 °C and 300 rpm until mid-log phase is reached (OD₆₀₀ = 0.5-0.55).
3. Transfer 20 mL of each culture into three prewarmed 125 mL flasks and incubate at 37 °C and 300 rpm for 2 min.
   NOTE: As timely processing of the samples is required, handle no more than 6 cultures at a time.
4. Prepare an antimicrobial compound solution in MOPS-g medium at a concentration of 2 mg/mL. Add the antimicrobial to each culture to reach the indicated concentrations. For the untreated control, add the required volume of MOPS-g medium.
   NOTE: Vortex the 2 mg/mL antimicrobial solution to allow an even distribution of the compound particles and avoid sedimentation.
5. Incubate the cultures for 45 min at 37 °C and 300 rpm.

Figure 1: Escherichia coli stress treatment. Bacterial cultures are grown in MOPS-g and treated with the indicated concentrations of the silver-ruthenium-containing antimicrobial when the mid-log phase is reached. Abbreviations: LB = lysogeny broth; Ag-Ru = silver-ruthenium; MOPS-g = 3-(N-morpholino)propanesulfonic acid (MOPS)-glucose. Please click here to view a larger version of this figure.

2. Collecting bacterial cell samples

1. After 45 min of stress treatment, determine the OD₆₀₀ of each culture. For each sample, harvest cells equivalent to 4 mL of OD₆₀₀ = 1 in 15 mL centrifuge tubes by centrifugation for 15 min at 3,000 × g and 4 °C.
2. Completely remove the supernatant and resuspend the cell pellets in 50 µL of ice-cold lysis buffer (Table 1). Incubate the samples for 30 min on ice.
   NOTE: This lysis step degrades the peptidoglycan layer. Always use freshly prepared lysis buffer.
3. Transfer the samples into 1.7 mL microcentrifuge tubes. Freeze at -80 °C until further use.

Figure 2: Bacterial sample collection. Cell samples are harvested by centrifugation and resuspended in lysis buffer followed by storage at -80 °C. Please click here to view a larger version of this figure.

3. Extracting the insoluble protein aggregates

1. Thaw samples on ice.
   NOTE: The freeze-thaw cycle contributes to cell lysis.
2. Add 360 µL of ice-cold buffer A (Table 1) and mix gently by pipetting.
   NOTE: The osmotic shock will also contribute to cell lysis.
3. Transfer the sample to a 2 mL microcentrifuge tube containing ~200 µL of 0.5 mm glass beads. Incubate for 30 min at 8 °C in a thermomixer with shaking at 1,400 rpm.
   NOTE: This step results in the physical disruption of the cell. A protease inhibitor can be used to minimize protein degradation. The disruption can be performed at 4 °C. Note that the use of glass beads has been reported to induce aggregation of a small subset of proteins in yeast¹⁹.
4. Incubate for 5 min on ice without shaking to settle the glass beads. Transfer 200 µL of the cell lysate into 1.7 mL microcentrifuge tubes.
   NOTE: Avoid the transfer of the glass beads.
5. Centrifuge at 16,000 × g and 4 °C for 20 min. Collect the supernatant, which contains soluble proteins, and proceed to section 4.
6. Resuspend the pellet in 200 µL of ice-cold buffer A (Table 1) using the pipette. Centrifuge at 16,000 × g and 4 °C for 20 min. Carefully remove the supernatant altogether.
7. Add 200 µL of ice-cold buffer B (see Table 1 and the Table of Materials) and carefully resuspend the pellet by pipetting.
   NOTE: The non-ionic detergent solubilizes membrane protein.
8. Repeat the centrifugation at 16,000 × g and 4 °C for 20 min. Carefully remove the supernatant.
9. Resuspend the pellet in 200 µL of cold buffer A (Table 1) by pipetting. Centrifuge at 16,000 × g and 4 °C for 20 min. Completely remove the supernatant.
10. Resuspend the pellet in 100 µL of 1x reducing SDS sample buffer (Table 1) and boil for 5 min at 95 °C in a thermomixer.
11. Store the sample at -20 °C to proceed later or immediately load on an SDS polyacrylamide gel for separation.

Figure 3: Extraction of insoluble protein aggregates. The extraction of protein aggregates involves a series of steps including cell disruption, the separation of protein aggregates from soluble proteins, the solubilization of membrane proteins, and washing. Abbreviation: SDS = sodium dodecyl sulfate. Please click here to view a larger version of this figure.

4. Soluble protein sample preparation

1. Mix 1 volume of 100% trichloroacetic acid (TCA) with 4 volumes of soluble protein sample from step 3.6.
   NOTE: Handling of TCA requires a fume hood and personal protective equipment and an approved waste disposal procedure.
2. Incubate for 10 min at 4 °C to allow for protein precipitation.
   NOTE: White precipitate will appear very soon.
3. Centrifuge to precipitate at 21,000 × g and 4 °C for 5 min and remove the supernatant. Wash the pellet with 200 µL of ice-cold acetone to remove cellular debris. Centrifuge at 21,000 × g and 4 °C for 5 min and remove the supernatant. Repeat these actions in step 4.3 a total of three times.
4. Place the microcentrifuge tubes with open lids in a thermomixer at 37 °C to remove the remaining acetone from the pellet.
   NOTE: Incubation of more than 5 min may reduce the solubility of the protein pellet.
5. Add 100 µL of 1x reducing SDS buffer (Table 1) and completely dissolve the pellet. Boil the sample for 5 min at 95 °C.
6. Store sample at -20 °C to proceed later or immediately load on an SDS polyacrylamide gel for separation.

Figure 4: Preparation of soluble proteins. The preparation of soluble protein involves a precipitation step with trichloroacetic acid and repeated washing with ice-cold acetone. Abbreviations: TCA = trichloroacetic acid; SDS = sodium dodecyl sulfate. Please click here to view a larger version of this figure.

5. Separation and visualization of extracted protein aggregates using SDS-PAGE

1. Prepare a 12% SDS-polyacrylamide gel.
   1. For two separating gels, pipette 5.1 mL of double-distilled water (ddH₂O), 3.75 mL of Tris-HCl (pH 8.8), 7.5 mL of 20% (w/v) SDS, 6 mL of 30% acrylamide/bisacrylamide (29:1) solution, 75 mL of 10% w/v ammonium persulfate, and 10 mL of tetramethylethylenediamine (TEMED) into a 15 mL centrifuge tube and mix gently without introducing air bubbles. Pour the gel using a 1 mL pipet within the glass plates, leaving the upper 2 cm free of the mixture. Add 70% ethanol on the top of the separating gel and allow an even interface between the two layers.
   2. After polymerization of the separating gel, prepare the stacking gel by pipetting 1.535 mL of ddH₂O, 625 mL of Tris-HCl (pH 6.8), 12.5 mL of 20% (w/v) SDS, 335 mL of 30% acrylamide/bisacrylamide (29:1) solution, 12.5 mL of 10% w/v ammonium persulfate, and 2.5 mL of TEMED. Remove the ethanol from the separating gels and add the stacking gel solution. Insert a comb with the desired number of pockets without introducing air bubbles. Allow polymerization for 20-30 min.
2. Load 4 µL of each sample and protein ladder into separate wells and run the gel(s) in Tris-Glycine running buffer (Table 1) at 144 V for 45 min at room temperature.
   NOTE: Stop the gel when the bromophenol band is about to migrate out of the gel.
3. Stain the gel(s) in a prewarmed Fairbanks solution A (Table 1) for 30 min on a rocker.
4. Decolor the gel(s) in a prewarmed Fairbanks solution D (Table 1) until the desired background (e.g., overnight) on a rocker.

Figure 5: Protein separation and visualization. The samples are separated by SDS-PAGE and visualized by Coomassie staining. Abbreviation: SDS-PAGE = sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Please click here to view a larger version of this figure.

Figure 6: Representative results of antimicrobial-induced protein aggregation in commensal Escherichia coli strain MG1655 and UPEC strain CFT073. E. coli strains MG1655 and CFT073 were grown at 37 °C and 300 rpm to OD₆₀₀= 0.5-0.55 in MOPS-g media before they were treated with the indicated concentrations (-, 0 mg/mL; +, 175 mg/mL; ++, 200 mg/mL) of the antimicrobial for 45 min. Soluble and insoluble protein samples were prepared as described in the protocol and Figure 1, Figure 2, Figure 3, and Figure 4 and visualized on a 12% SDS polyacrylamide gel (Figure 5). Protein aggregate formation (insoluble fraction) was increased in both strains in the presence of the antimicrobial while the amounts of soluble proteins were decreased. Overall, the antimicrobial had a much more potent effect on MG1655 than on CFT073. Abbreviations: M= Protein marker; UPEC = uropathogenic E. coli; MOPS-g = 3-(N-morpholino)propanesulfonic acid (MOPS)-glucose; SDS = sodium dodecyl sulfate. Please click here to view a larger version of this figure.

Here, two E. coli strains were used that differ in their susceptibility to a proteotoxic silver-ruthenium-containing antimicrobial to demonstrate this protocol. Preliminary survival data revealed that the commensal E. coli strain MG1655 is significantly more sensitive to the ROS-generating antimicrobial than the UPEC strain CFT073 (data not shown). Both strains were grown in MOPS-g media at 37 °C and 300 rpm. At the mid-log phase, the cells were either left untreated or treated with 175 µg/mL and 200 µg/mL of the antimicrobial, respectively, and incubated for 45 min. Subsequently, the cells were lysed, and cellular protein aggregates separated from the soluble proteins. Proteins in both fractions were then separated by SDS-PAGE and visualized by Coomassie staining. The insoluble fraction shown in Figure 6 represents the amount of protein aggregates formed, which was increased when cells were incubated in the presence of the antimicrobial compared to untreated cells. The increase in protein aggregate formation was independent of the strain background, although a much more pronounced increase in aggregate formation was detected in the more sensitive strain MG1655. Conversely, lower amounts of soluble proteins (soluble fraction) were observed after antimicrobial treatment of the cells compared to the untreated counterpart. This result was expected given the preliminary data that showed a substantially higher tolerance of the antimicrobial in CFT073 than MG1655.

Table 1: Buffer, Media, and Solutions. Recipes for buffer, media, and solutions used in this protocol.

This protocol describes an optimized methodology for the analysis of protein aggregate formation after treatment of different E. coli strains with a proteotoxic antimicrobial. The protocol allows the simultaneous extraction of insoluble and soluble protein fractions from treated and untreated E. coli cells. Compared to existing protocols for protein aggregate isolation from cells^14,15,16,20, this method has several advantages: (i) only small culture volumes (4-8 mL) are needed; (ii) the cell disruption process does not rely on special equipment such as a French press, cell disruptor, or sonicator; and (iii) the protocol is easy to follow even for relatively early-career scientists in the field.

Bacteria encounter a myriad of stresses in their natural environment, and many of them represent a threat, particularly to proteins, the most abundant macromolecule in the cell²¹. The described methodology offers many potential applications. One of them is the possibility to investigate the efficacy by which a wide range of stresses (e.g., elevated temperatures, reactive oxygen species, reactive chlorine species) and chemical compounds affect protein homeostasis in bacteria, archaea, and even eukaryotic cells^5,12,22,23. Our extraction was performed with the two distantly related E. coli strains, MG1655 and CFT073. However, it has also been successfully applied to study the role of specific gene products for protein homeostasis by comparing protein aggregate formation in wild-type and mutant strains^11,12.

After considering appropriate modifications and troubleshooting of growth conditions and stressor concentrations, this protocol can also be used to determine protein aggregate formation in other Gram-negative⁵ and Gram-positive bacteria⁸. Notably, the gel bands separated after SDS-PAGE can be densitometrically analyzed (i.e., using ImageJ). This approach can also be used to analyze the impact of additional therapeutic compounds such as inhibitors of molecular chaperones²⁴. Most intriguingly, it can be combined with mass spectrometric approaches to provide information on the type of protein species that are aggregation-sensitive under a specific stress condition, simply by determining their presence or absence in the insoluble protein fraction^5,15.

The success of this protocol requires careful consideration of the stressor concentrations that cells are exposed to and the exposure. We, therefore, recommend performing a preliminary survival assay with increasing stressor concentrations to determine the proteotoxic concentration. Moreover, the stressor solutions should be prepared freshly before each experiment, and the sample collection times be kept consistent. One limitation of this method is the low number of samples that can concurrently be processed. This is mainly due to the time-sensitive handling of samples in section 1, which involves significant vortexing steps of the antimicrobial solution in between the addition of the compound to the cultures to avoid sedimentation. However, this may not be such an issue when different soluble stressors are applied. Moreover, the described procedure does not provide any time-resolved information on protein aggregate location and trajectory, which would require more advanced techniques such as fluorescence microscopy in combination with time-lapse microscopy²⁵. In summary, the improved methodology is simple, easy to follow, inexpensive, and offers the potential for additional modification that allows a tailored approach for identifying proteotoxic compounds or bacterial stress response genes.