Evaluation of the Impact of Protein Aggregation on Cellular Oxidative Stress in Yeast

Summary

Protein aggregation elicits cellular oxidative stress. This protocol describes a method for monitoring the intracellular states of amyloidogenic proteins and the oxidative stress associated with them, using flow cytometry. The approach is used to study the behavior of soluble and aggregation-prone variants of the amyloid-β peptide.

Abstract

Protein misfolding and aggregation into amyloid conformations have been related to the onset and progression of several neurodegenerative diseases. However, there is still little information about how insoluble protein aggregates exert their toxic effects in vivo. Simple prokaryotic and eukaryotic model organisms, such as bacteria and yeast, have contributed significantly to our present understanding of the mechanisms behind the intracellular amyloid formation, aggregates propagation, and toxicity. In this protocol, the use of yeast is described as a model to dissect the relationship between the formation of protein aggregates and their impact on cellular oxidative stress. The method combines the detection of the intracellular soluble/aggregated state of an amyloidogenic protein with the quantification of the cellular oxidative damage resulting from its expression using flow cytometry (FC). This approach is simple, fast, and quantitative. The study illustrates the technique by correlating the cellular oxidative stress caused by a large set of amyloid-β peptide variants with their respective intrinsic aggregation propensities.

Introduction

Proteostasis is a fundamental determinant of cell fitness and aging processes. In cells, protein homeostasis is maintained by sophisticated protein quality control networks aimed to ensure the correct refolding of misfolded protein conformers by chaperones and/or their targeted proteolysis with several well-conserved mechanisms^1,2,3,4,5. A large number of studies provide support to the link between the onset and progression of a broad range of human diseases and the failure of proteostasis, leading to protein misfolding and aggregation. For instance, the presence of protein deposits is considered a pathological hallmark of many neurodegenerative disorders, such as Alzheimer's, Parkinson's, and Huntington's diseases^6,7,8, prionogenic diseases, and non-degenerative amyloidoses⁹. It has been suggested that early oligomeric and protofibrillar assemblies in the aggregation reaction are the main elicitors of cytotoxicity, establishing aberrant interactions with other proteins in the crowded cellular milieu¹⁰. In addition, protein inclusions (PI) can be transmitted between cells, propagating their toxic effect^11,12. Therefore, it could be that the formation of PI might indeed constitute a detoxifying mechanism that restricts the presence of dangerous aggregated species to specific locations in the cell, where they can be processed or accumulated without major side effects^13,14.

Standard in vitro biochemical approaches have provided important insights into the different species that populate aggregation reactions and their properties^15,16. However, the conditions used in these assays are clearly different from those occurring within the cell and, therefore, question their physiological relevance. Because of the notable conservation of cellular pathways such as protein quality control, autophagy, or the regulation of the cellular redox state^17,18 among eukaryotes^{19,20,21,22,23}, the budding yeast Saccharomyces cerevisiae (S. cerevisiae) has emerged as a privileged simple cellular model to study the molecular determinants of protein aggregation and its associated cytotoxic impact in biologically relevant environments^24,25,26.

Protein aggregation propensity is a feature inherently encoded in the primary sequence. Thus, the formation of amyloid-like structures can be predicted based on the identification and evaluation of the potency of aggregation-promoting regions in polypeptides²⁷. However, despite the success of bioinformatic algorithms to predict the in vitro aggregation properties of protein sequences, they are still far from forecasting how these propensities translate into in vivo cytotoxic impact. Studies that address the link between the aggregated state of a given protein and its associated cellular damage in a systematic manner may help to circumvent this computational limitation. This connection is addressed in the present study, taking advantage of a large set of variants of the amyloid-β peptide Aβ42 differing only in a single residue, but displaying a continuous range of aggregation propensities in vivo²⁸. In particular, an FC-based approach to identify the conformational species accounting for the oxidative damage elicited by aggregation-prone proteins in yeast cells is described. The methodology provides many advantages such as simplicity, high-throughput capability, and accurate quantitative measurement. This approach made it possible to confirm that PI play a protective role against oxidative stress.

Protocol

1. S. cerevisiae Cultures and Protein Expression

Note: Aβ variants exhibit different relative aggregation propensities due to a mutation in a single residue at position 19 (Phe19) of the Aβ42 peptide (Figure 1A). These peptide variants are tagged with green fluorescent protein (GFP), which acts as an aggregation reporter (Figure 1)²⁹.

1. Transform plasmids encoding for 20 Aβ42-GFP variants into yeast cells with a BY4741 parental background (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0)³⁰. Each plasmid encodes for an Aβ42 mutant fused to GFP by a GSSGSSG linker. The plasmids are pESC(-URA) and include the selectable marker URA3²⁹.
   1. Prepare the synthetic complete medium without uracil (SC-URA) with the following components: 1.9 g/L yeast nitrogen base without uracil, 5 g/L ammonium sulfate, and 900 mL of ddH₂O combined with 100 mL of 20% (w/v) glucose.
   2. Put one colony of the transformed yeast cells into 20 mL of the SC-URA medium containing 2% glucose and grow the cultures overnight at 30 °C under agitation at 210 rpm.
2. The next day, grow new yeast cells cultures by inoculating 100 µL of the overnight culture into 5 mL of fresh SC-URA medium.
3. At an OD₅₉₀ of 0.5, centrifuge the cultures at 3,000 x g for 4 min, discard the supernatant, and resuspend the cells in the same volume of fresh SC-URA medium containing 2% raffinose.
4. Incubate at 30 °C under agitation at 210 rpm for 30 min. Then, centrifuge the cells at 3,000 x g for 4 min and discard the supernatant. Resuspend the cells in fresh SC-URA medium containing 2% galactose to induce the recombinant protein expression.
5. After 16 h of inducing protein expression in SC-URA medium containing 2% galactose, harvest 1 mL of the induced cells in sterile microcentrifuge tubes by centrifugation at 3,000 x g for 4 min.
   NOTE: The most relevant differences occur after an induction period of 16 h. Induced yeast cells expressing Aβ42-GFP variants can be visualized by fluorescence microscopy to determine the recombinant protein distribution inside the cells (Figure 2B).

2. Cell Staining

Note: Fresh non-induced cells are required as a negative control to establish a fluorescent threshold during an FC analysis.

1. Take the 16 h-induced yeast cells and determine their optical density at 590 nm (OD₅₉₀)_. Dilute them in a sterile phosphate buffered saline (PBS) to an OD₅₉₀ of 0.1.
   1. Prepare 1x phosphate buffered saline as follows: add 8 g of NaCl, 0.2 g of KCl, 1.44 g of Na₂HPO₄, and 0.24 g of KH₂PO₄ to 800 mL of ddH₂O. Fill the solution up to 1 L of ddH₂O after adjusting the pH to 7.4 with HCl. Filter the buffer through a 0.2 µm filter.
2. Transfer the expressing cell suspensions to appropriately labeled tubes (12 × 75 mm round-bottom polystyrene) and protect them from light. Make sure they are ready to be loaded on the flow cytometer for analysis, together with non-induced and non-stained cells.
3. Add the oxidative stress probe (e.g., CellROX Deep Red) to each sample at a final concentration of 5 µM and incubate the cells at 30 °C for 30 min in the dark.
4. After incubation, wash the cells 3 times with 1x PBS and resuspend them in the same volume of buffer before their analysis by FC.

3. Flow Cytometry Analysis

Note: Use an FC setup with appropriate lasers and filters to detect GFP and oxidative stress probe fluorescence signal. A flow cytometer equipped with a 488 nm blue laser for the detection of GFP and a 635 nm red laser for the detection of oxidative stress probe fluorescence can be used. Acquisition of emission fluorescence of GFP and oxidative stress probe is performed with a 530/30 nm BP filter and a 660/20 BP filter, respectively (Table 1).

1. Click on Open New Worksheet Panel and create the following acquisition dot plots.
   1. Select the Scatter Plot Tool from the toolbar and create a plot with the variables Side Scatter-Area (SSC-A) on the y-axis vs. Forward Scatter-Area (FSC-A) in a linear scale on the x-axis.
   2. Click the Scatter Plot Tool from the toolbar and create a plot with the variables FL1-Area (FITC) on the x-axis vs. FL3-Area (APC) in a logarithmic scale on the y-axis.
2. Click the Instruments Settings icon. Click the Compensation tab and set all compensation levels. Click on the Acquisition tab and select a total number of 20,000 events to be recorded.
3. Click the Flow Rate icon to switch the flow rate to Low.
4. Click the Acquire tab to start running the non-induced, unstained cells and adjust the voltage in the Instrument Settings of the forward and side scatter until the population is distributed in the middle-left quadrant.
   1. Click on the Polygon icon to set a region R1 around the cell population excluding cell debris, and use this gated population P1 = R1 for all fluorescent dot plots and histogram representations (Figure 3).
5. To adjust the PMT voltage of the fluorescence signal, run the unstained cells in the FL1-FL3 dot plot, tuning the gain in the Instrument Setting tab, until the cells are distributed in the lower left quadrant.
6. Change the sample to the induced cells to measure GFP fluorescence. In the Instrument Setting tab, set the gain in the FSC-A vs. FL1 (FITC) when the population is distributed in the lower-right quadrant. Define the positive cell population with a gate (Gate P2).
7. Change the sample to the non-induced stained cells to display oxidative stress by the fluorescence on an FSC-A vs. FL3 (APC) dot plot. Tune the gain until the cell population is distributed in the left-upper quadrant. Gate the positive cell population in P3.
8. Make two histogram plots with the histogram icon to represent the cell fluorescence. To represent the fluorescence intensity, make sure FL1 (FITC) representing the GFP fluorescence is in the x-axis, and FL3 (APC) representing the fluorescence is in the y-axis, applying P2 and P3 populations on a logarithmic scale, respectively.
   NOTE: Histogram overlays are a good way to illustrate differences between the fluorescence profiles of cells expressing different Aβ variants.

4. Recombinant Protein Immunodetection

1. To determine protein expression levels, harvest the yeast cultures induced for 16 h by centrifugation at 4,000 x g for 6 min, and store the cell pellets at -80 °C.
   1. Resuspend the collected cells expressing recombinant protein for 16 h in PBS. Prepare 200 µL cell suspensions of each Aβ42-GFP mutant to an OD₅₉₀ of 20.
2. For total protein fraction analysis, harvest 100 µL of each mutant by centrifugation at 14,000 x g for 20 min and resuspend the cell pellets in the same volume of yeast lysis buffer Y-PER (50 mM Tris-HCl pH 8.0, 1% DMSO, 200 mM NaCl, 1 mM EDTA containing 10 mg/mL SB3-14 (myristyl sulfobetaine) supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF)).
   1. Incubate the samples for 20 min at room temperature under mild agitation.
3. Determine the protein extract concentration by using the Bradford assay. Load up to 5 µg of protein extract of each sample on a 15% acrylamide SDS-PAGE electrophoresis gel and blot onto a polyvinylidene difluoride (PVDF) membrane at 100 V for 60 min.
   1. Prepare a Bradford reagent with 100 mg of Coomassie Brilliant Blue G-250 dissolved in 50 mL of 95% ethanol and add 100 mL of 85% (w/v) phosphoric acid. Then, when the dye is completely dissolved, dilute the mixture with ddH₂O to 1 L and filter it through cellulose filter paper just before use.
      NOTE: The Bradford reagent should be light brown.
   2. Prepare a standard of bovine serum albumin (BSA) ranging from 5 - 100 µg of protein in 1,000 µL of the Bradford reagent.
   3. Add 1 - 10 µL of protein extract to 1 mL of the Bradford reagent and incubate the mixture at room temperature for 5 min.
   4. Measure the absorbance of the standards and the protein extract samples at OD₅₉₅.
   5. For the analysis, make a plot with the absorbance values obtained for the standards vs. µg of protein. Based on the standard curve, determine the concentrations of the original samples from the amount of protein, considering the volume and dilution, if any.
4. Rinse the membrane with 1x Tris-buffered saline (TBS), 0.1% Tween 20 (TTBS), and block it using 5% (w/v) nonfat dry milk in 1x TTBS.
   1. To prepare 1 L of 10x TBS, dissolve 24 g of Tris base and 88 g of NaCl in 900 mL of H₂O_dd and adjust the pH to 7.6. Add ddH₂O to a final volume of 1 L. For a 1x solution, mix 1 part of 10x stock with 9 parts of ddH₂O.
5. Incubate the membrane for 1 h with a primary β-amyloid antibody 6E10 diluted 1:1,000 and for 1 h at room temperature with a secondary antibody goat anti-mouse IgG-HRP conjugate diluted 1:10,000.
6. Develop membranes using 1 mL of Luminata fluorescent reagent and quantify bands by densitometric analysis as described in step 5.2.

5. Data Analysis

1. To analyze the data obtained by FC, create a table displaying the mean fluorescent intensity (MFI) and median fluorescence with its corresponding standard error and/or the coefficient of variance (CV) for GFP fluorescence and oxidative stress levels.
   1. Click the Inspector tab and the Statistics icon to customize the statistical data that are required for each channel.
      NOTE: When comparing protein variants, convert fluorescence values data to a fold-over background or resolution metric (RD):
      
      In this metric, the mean sample is the mean fluorescence of problem samples, the mean control is the mean fluorescence of cells used as a control, and the SD is the standard deviation of the mean fluorescence. RD factor is a more accurate index than MFI value as it not only calculates the difference between samples but also accounts for the spreading of the data. This calculation is specifically relevant when comparing data coming from different experiments carried out over time. Once the RDs have been calculated, the use of a statistical test is advisable to confirm the significance of the observed differences between protein variants.
2. For protein immunodetection, quantify the recombinant protein levels by Western blot using ImageJ software.
   1. Before the densitometry analysis, convert the original raw images of the developed membrane into JPEG file format and greyscale mode.
   2. Adjust the Set Measurements section of the Analyze menu to Mean Grey Value.
   3. Click on the Rectangle tool from ImageJ and define a region of interest (ROI) with a consistent size for the band densitometry analysis. Center each band on the membrane image inside the created rectangular frame and record the measurement one by one using Ctrl+M (the measurement command).
   4. Measure the local background adjacent to each band, applying the same ROI area. After measurement, subtract the corresponding local background from each individual band.

Results

This protocol describes how to employ a collection of 20 variants of the Aβ42 peptide where Phe19 has been mutated to all natural proteinogenic amino acids²⁸. The theoretical aggregation propensities of these proteins can be analyzed using two different bioinformatic algorithms (AGGRESCAN and TANGO^31,32). In both cases, this analysis renders a progressive gradation of aggregation tendencies, ascribing,...

Discussion

A wide range of diseases is linked to the accumulation of misfolded proteins into cellular deposits^6,7,8,33. Many efforts have been made to unravel the molecular mechanisms that trigger the onset of these diseases using computational approaches, which do not take into account protein concentrations, or in vitro approaches, in which the protein concentration remains constant during the ...

Materials

Name	Company	Catalog Number	Comments
Yeast cells BY4741	ATCC	201388	Genotype: MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0
pESC(-Ura) plasmid	Agilent Genomics	217454	Yeast expression plasmid with a Gal promotor. Selectable marker URA3
Yeast Synthetic Drop-out Medium Supplements	Sigma	Y1501	Powder
Yeast Nitrogen Base Without Amino Acids	Sigma	Y0626	Powder
Raffinose	Sigma	R7630	Powder
Glucose	Sigma	G7021	Powder
Galactose	Sigma	G0750	Powder
Phosphate Buffered Saline (PBS)	Fisher Scientific	BP3991	Solution 10X
CellROX Deep Red Reagent	Life Technologies	C10422	Free radical cell-permeant fluorescent sensor, non-fluorescent while in a reduced state, and exhibits bright fluorescence upon oxidation by reactive oxygen species (ROS), with absorption/emission maxima at 644/665 nm.
Y-PER protein extraction reagent	Thermo Scientific	78990	Liquid cell lysis buffer
Acrylamide/Bis-acrylamide	Sigma	A6050	Solution
Bradford dye reagent	Bio-Rad	5000205	Dye reagent for one-step determination of protein concentration
β-amyloid antibody 6E10	BioLegend	803001	Mouse IgG1. The epitope lies within amino acids 3-8 of beta amyloid (EFRHDS).
Goat anti-mouse IgG-HRP conjugate	Bio-Rad	1721011
Membrane Immobilon-P, PVDF	Millipore	IPVH00010
Luminata forte	Merk	WBLUF0100	Premixed, ready to use chemiluminescent HRP detection reagent
Phenylmethanesulfonyl fluoride solution (PMSF)	Sigma	93482	Protease inhibitor. Dissolved at 0.1 M in ethanol
FACSCanto flow cytometer	BD Biosciences	657338	Equipped with a 488 nm blue laser for the detection of GFP, and 635 nm red laser / 530/30 nm BP filter and 660/20 BP filter
Mini Trans-Blot Electrophoresis Transfer cell	Bio-Rad	1703930	Protein transference system
Mini-PROTEAN Tetra Handcast Systems	Bio-Rad	1658000FC	Electrophoresis system