Using FRET-Based Biosensor Cells to Study the Seeding Activity of Tau and α-Synuclein

Abstract

Two fluorescence resonance energy transfer (FRET)-based biosensor cell lines developed several years ago by the Diamond group (University of Texas, Southwestern) have allowed convenient, sensitive, and specific measurement of the intracellular aggregation of tau and α-synuclein following the addition of oligomer or small-aggregate “seeds” of these proteins from various sources, and an advancement relative to similar single-fluorophore systems. These biosensor cell lines allow researchers to both visualize the intracellular aggregates of tau or α-synuclein and measure intracellular aggregation with high sensitivity using a FRET signal in flow cytometry. Here we provide detailed protocols for generating seeds, culturing the biosensor cells, measuring intracellular aggregates by flow cytometry, and analyzing the results and discuss the utility of the technique with the aim of characterizing factors involved in the regulation of intracellular tau and α-synuclein aggregation.

1. Introduction

Neurodegenerative diseases affect millions of people worldwide. Many of these diseases are proteinopathies, caused by misfolding, oligomerization, aggregation, and spreading of specific proteins [1]. Prominent examples are tauopathies, such as Alzheimer’s disease, frontotemporal dementia (FTD), and progressive supranuclear palsy [2], and synucleinopathies, including Parkinson’s disease, dementia with Lewy bodies, and multiple system atrophy [1, 3, 4]. Tauopathies are caused by misfolding of the microtubule-associated protein tau into neurotoxic oligomers and fibrillar aggregates that deposit in the tissue as neurofibrillary tangles, Pick bodies, or other pathological forms [1]. Tau has six isoforms, which differ by the inclusion of 0, 1, or 2 exons in the N-terminal domain and three (3R) versus four (4R) repeats in the microtubule-binding domain. Disruption of the normal 3R/4R-tau ratio is associated with tauopathy. In addition, in certain tauopathies, the deposits consist of only the 3R or 4R form [2, 5, 6]. Similarly, synucleinopathies are caused by misfolding and self-assembly of the presynaptic neuronal protein α-synuclein into toxic oligomers and aggregates.

Tau and α-synuclein are natively unstructured proteins, which are thought to adopt stable structures in a healthy brain when they are associated with microtubules or synaptic vesicles, respectively. In different neurodegenerative diseases or central nervous system injuries, these proteins tend to self-assemble first into neurotoxic oligomers and eventually into β-sheet-rich fibrils and higher-order structures. The molecular mechanisms governing these processes in vivo are not well understood. Interestingly, in both cases, aberrant posttranslational modification, particularly phosphorylation and truncation, appear to be involved [2, 7, 8]. The oligomers are thought to be the most toxic form of the proteins, whereas the larger, fibrillar aggregates that deposit in the tissue may be, in fact, the body’s way of containing and sequestering the offending proteins when the normal clearance mechanisms fail to degrade them [9]. In recent years, it has become evident that the pathological forms of these proteins spread throughout the brain in a prion-like manner as the respective diseases progress, yet the mechanisms involved in the spreading are only partially understood [1, 10–12]. The spreading agents are referred to as “seeds” (oligomers or small aggregates of the misfolded protein), which are taken up by healthy cells, come into contact with the normal form of the endogenous protein inside the host cell, and “corrupt” it by inducing its misfolding into the toxic form (see Fig. 1) [13].

Fig. 1.

Schematic representation of tau seeding. A tau seed enters a host cell expressing normal tau. The seed comes in contact with the physiologic form of the protein inside a host cell and “corrupts” it by acting as a template to induce misfolding of the endogenous tau into the same structure as that of the seed. The misfolded tau self-assembles into oligomers and higher-order structures, similar to the original structure of the seed. New seeds form, some of which may leave the cell and go on to repeat the same process in other “naïve” cells

Studying the aggregation and spreading of toxic protein assemblies is important for understanding the molecular mechanisms governing these processes and for developing diagnostic testing, progression markers, and cures for proteinopathies. In the cases of tauopathies and synucleinopathies, the development of fluorescence resonance energy transfer (FRET)-based biosensor cell lines by Diamond and coworkers has provided a significant technical advancement to the field [13, 14]. Previously, such systems used single fluorophores, which allowed observation and counting of bright puncta formed upon protein aggregation using a fluorescent microscope [15, 16]. Though these were useful, FRET-based quantification of the signal increases the sensitivity of the method substantially and allows detection of seeding as the earliest sign of pathology [13, 14].

The FRET-based biosensor cell lines for detection of tau or α-synuclein have been made available to the research community by the Diamond group. Both the tau and α-synuclein biosensor lines are HEK 293 cells stably expressing aggregation-prone versions of the respective proteins. The tau biosensor cell line expresses the tau repeat domain (RD) containing the FTD-linked P301S variant of tau (substitution of a proline in position 301 by a serine) fused to either cyan fluorescent protein (CFP) or yellow fluorescent protein (YFP). Both fusion proteins are expressed at approximately equal concentration levels in each cell [13]. Similarly, the α-synuclein biosensor cells express the familial Parkinson’s disease-linked A53T variant of α-synuclein fused to either CFP or YFP at approximately equal levels in each cell. Under nascent conditions, diffuse fluorescence can be observed in the cells expressing these proteins, whereas upon introduction of the appropriate seeds, the aggregation-prone proteins aggregate, bringing the two fluorescent proteins into close proximity (<10 nm) and allowing FRET to occur (see Fig. 2), which can be conveniently quantified with high sensitivity using flow cytometry [4, 13].

Fig. 2.

Schematic representation of the FRET-based biosensor system. Tau RD containing the P301S substitution and conjugated to CFP or YFP are expressed in the same cell. When the two proteins are far from each other, each fluorescent protein fluoresces at its own typical wavelength and there is no FRET. Upon addition of seeds to the biosensor cells, the endogenously expressed P301S-tau RD proteins aggregate, bringing the two fluorophores into close proximity, allowing FRET to occur. When CFP absorbs light at 440 nm, rather than emitting fluorescence at 490 nm, the light is absorbed by YFP and then emitted at 527 nm

A practical challenge in this technique is introduction of the seeds, which are too large to go through the cell membrane, into the biosensor cells. To overcome this issue, in a typical experiment, the seeds are mixed with the transfection agent Lipofectamine 2000 to facilitate their direct transduction into the cells. Mixing with Lipofectamine 2000 is unlikely to affect the cross-β structure of amyloid fibril fragments but would be expected to alter the structure of metastable oligomers [17]. Therefore, experiments using oligomers as seeds should be interpreted with caution. Interestingly, recently, Shin et al. demonstrated that in the case of tau seeding, the use of a transfection agent could be avoided by incubating the cells with sub-toxic concentrations of amyloid β-protein oligomers 24 h prior to introduction of tau seeds [18]. In addition to the technical difficulty in introducing the seeds into the cells as discussed above, another important limitation of this technique is that the biosensor cells are human kidney cells, whereas the diseases of interest affect primarily the central nervous system [13, 14]. Thus, although seeding can be conveniently tested and quantified using these cell lines, in many cases the results need to be validated in neurons to ensure that the seeding phenomenon is relevant to the central nervous system.

To induce FRET, after the seeds cause misfolding and aggregation of the endogenous CFP/YFP-tagged protein, the cells are exposed to a 440-nm laser light (see Fig. 2), which is absorbed by CFP, leading to fluorescence at 490 nm (cyan). This wavelength overlaps with the excitation wavelength of YFP, which absorbs the 490-nm light and emits light at 527 nm (yellow) [19]. Following incubation of the cells with the seeds, fluorescent puncta can be observed using a fluorescence microscope and the FRET signal can be quantified using a flow cytometer. In the flow cytometer, the cells pass in a single file through a 440-nm laser beam, allowing sorting of the cells that produce the FRET signal at 527 nm and separating them from all the other cells. The signal is then quantified as integrated FRET density, which multiplies the number of FRET-producing cells by the intensity of the emitted light (determined by the number and size of the aggregates in each cell). The technique allows for a quick readout of data, is highly sensitive thanks to a low signal-to-noise ratio, requires minimal manipulation of samples, and can be easily automated [14, 19].

This chapter discusses the general methodology for measuring tau or α-synuclein seeding using the respective biosensor cell lines and outlines the cell culture techniques, flow cytometry, and analysis of results. Because the techniques are nearly identical for tau and α-synuclein, we illustrate the process for tau only.

2. Materials

1. HEK-293 T cells untagged, ATCC-CRL-3216

2. Tau RD P301S-YFP cell line, available from Dr. Diamond, University of Texas, Southwestern

3. Tau RD P301S-CFP cell line, available from Dr. Diamond, University of Texas, Southwestern

4. Tau RD P301S-CFP/YFP FRET biosensor cell line, ATCC-CRL-3275

5. Cell culture medium: 89% Dulbecco’s Modified Eagle’s Medium (DMEM), 10% Fetal Bovine Serum (FBS), 1% penicillin-streptomycin. In the example below, we have used the following:

   1. DMEM containing high glucose, GlutaMAX supplement, and pyruvate
   2. Australia Origin FBS
   3. Penicillin-streptomycin (10,000 U/mL)
6. Flow cytometry buffer: Hank’s Balanced Salt Solution (HBSS), 1% FBS, 1 mM ethylenediaminetetraacetic acid (EDTA)

   1. HBSS, no calcium, no magnesium, no phenol red
   2. Ethylenediaminetetraacetic acid disodium salt dihydrate

7. Trypsin (TrypLE Express Enzyme Reagent (1X) phenol red)

8. 96-well, clear, flat-bottom, tissue-culture plate with lid

9. 96-well, clear, round-bottom, polystyrene plate not treated

10. 37 °C water bath

11. 15-mL tube and 96-well plate swinging bucket rotor centrifuge

12. 1.5-mL microcentrifuge tubes

13. Attune NxT flow cytometer (see Note 1)

14. 9-in plastic Pasteur pipettes individually wrapped, sterile

15. 15-mL centrifuge tubes

16. 50-mL conical tubes

17. CryoGrinder

18. Extraction buffer: 1X Tris-buffered saline, Halt Protease/Phosphatase Inhibitor Cocktail, diH₂O

19. Hippocampus, cortex, or whole brain of PS19 mouse (see Note 2)

20. Dry ice

21. Liquid nitrogen

22. Disposable pipetting reservoirs, 25 mL, sterile, individually wrapped

23. Opti-MEM reduced-serum medium

24. Lipofectamine 2000 transfection reagent

25. Keyence BZ-X710 fluorescence microscope (see Note 3)

26. Cell-counting plate INCYTO C-Chip, disposable hemacytometers, Neubauer improved grid, 100-μm chamber depth SKC (see Note 4)

3. Methods

In this section, for brevity, the tau RD P301S-CFP/YFP FRET biosensor cell line is referred to as “biosensor cells.”

3.1. Preparing Seeds

There are many different sources of seeds, including fibrils made of the recombinant proteins or brain extracts from patients or animal models. Here, we illustrate the method using brain extracts from a tauopathy mouse model, the P301S-tau (line PS-19) mouse [13].

1. Make the extraction buffer by adding 5 mL of 10X Tris-buffered saline, 500 μL of 100X Halt Protease/Phosphatase Inhibitor, to a 50-mL conical tube and fill up to 50 mL with diH₂O.

2. Submerge the cryogrinding mortars, drill bits, and metal scraper in a bucket with liquid nitrogen and allow them to cool for 5–10 min.

3. Remove the mortar and place it in a bucket with dry ice.

4. Place the frozen PS19 mouse brain in the mortar, and using the CryoGrinder, begin forcefully grinding the brain into a powder.

5. Transfer the powdered brain into a labeled 1.5-mL microcentrifuge tube and weigh the material.

6. Add the extraction buffer to the powered brain in a 1:10 mass (mg) to volume (μL) ratio.

7. Spin the homogenized sample in a centrifuge at 100,000 × g for 20 min at 4 °C.

8. Collect the supernatant in a new labeled tube and store it at −80 °C until use.

3.2. Cell Culture

3.2.1. Counting and Replating Cells

Four cell lines, HEK-293-T cells, tau RD P301S-CFP cells, tau RD P301S-YFP cells, and biosensor cells, should be started and passaged as necessary.

1. Warm up the cell culture medium and trypsin in a 37 °C water bath (see Note 5).

2. Working in a sterile, laminar flow hood, use a 9-in plastic Pasteur pipette to aspirate the cell media from the cell culture flask (see Note 6).

3. Add 1 mL of 37 °C trypsin to the cells (see Note 7).

4. Repeat steps 2 and 3 for each cell line.

5. Incubate the flasks at 37 °C in the cell culture incubator for 2 min.

6. Check if the cells have detached from the bottom surface of the flask by tapping the sides of the flask (see Note 8).

7. If the cells have detached, add 4 mL of warmed cell culture medium (see Note 5). Wash the flask thrice with medium and transfer the cells to a labeled 15-mL centrifuge tube.

8. Centrifuge the cells at 125 g for 5–7 min at room temperature.

9. In the hood, aspirate the supernatant medium from the 15-mL tubes using a plastic Pasteur pipette.

10. Resuspend the cells gently in 5 mL of warmed cell culture medium and mix gently by pipetting up and down.

11. Count the cells and dilute each line to a concentration of 15 × 10⁴ cells/mL (see Notes 4 and 9).

12. Add the HEK-293-T control cell suspension to the wells of the 96-well plate (hereon referred to as experiment plate) (130 μL of cell suspension per well) (see Notes 10, 11, and 12). Repeat this step for all cell line suspensions.

13. Place a cover on the experiment plate and gently tap all four edges of the plate to ensure that the cells spread evenly. Incubate for 24 h.

3.2.2. Seed Preparation and Transduction

• 1Calculate the amount of opti-MEM, Lipofectamine, and seed sample needed for the experiment (opti-MEM 8.75 μL/well, Lipofectamine 1.25 μL/well). In this example, using P301S-tau mouse brain extracts, we have found that 1-μg total protein per well, measured using a bicinchoninic acid protein assay, yields robust results.
• 2In a sterile tissue culture hood, combine the opti-MEM (8.75 μL/well) and Lipofectamine (1.25 μL/well) together in a 1.5-mL microcentrifuge tube. Add the opti-MEM first. Vortex for 10 s. Label this tube “O + L.”
• 3Next, make up the tubes containing the seeds (“sample” tubes). The total volume is 10 μL per well consisting of the required seed volume at the desired concentration and opti-MEM. Add the opti-MEM first and then the seeds, and mix well (see Note 1).
• 4Prepare the biosensor cell control tube, which consists of 10 μL opti-MEM per well, and label this tube “control.”
• 5Add 10 μL per well of the O + L solution to each “sample” tube including the “control” tube. Mix the contents of each tube by pipetting up and down ten times (see Notes 13 and 14).
• 6Incubate at RT for 20 min (see Note 15).
• 7Add 20 μL of each sample to the appropriate wells and 20 μL of the control solution to the control biosensor cell wells (see Notes 16 and 17).
• 8Incubate for 48 h in a tissue culture incubator (see Note 18).

3.2.3. Preparing Cells for Flow Cytometry

Roughly 1 h before performing flow cytometry, the cells should be prepared. If desired, imaging should take place at this stage.

• 9Using a fluorescence microscope, take images of cells (see an example in Fig. 3 and Note 3).
• 10Warm the medium and trypsin in a 37 °C water bath (see Note 5).
• 11Place three reservoirs and a 96-well, round-bottom plate in the tissue culture hood.
• 12In the hood, add 50 μL of trypsin per each well of the experiment plate to a reservoir.
• 13Remove the old media from the experiment plate by inverting the plate and shaking it over a biohazard waste bin (see Note 18).
• 14Add 50 μL of trypsin to each cell-containing well of the experiment plate using a multichannel pipette.
• 15Place the experiment plate in a tissue culture incubator for 1–2 min.
• 16Observe the cells under a light microscope to ensure they are detached from the plate’s surface (see Note 8).
• 17In the tissue culture hood, add the desired amount of 37 °C medium to a clean reservoir.
• 18Add 150 μL of the 37 °C medium to each cell-containing well of the experiment plate using a multichannel pipette and mix well by pipetting up and down.
• 19Individually collect the cells from each well of the experiment plate and transfer them to a new, 96-well, round-bottom plate.
• 20Centrifuge the plate at 125 g for 5–7 min at room temperature.
• 21Carefully aspirate the supernatant (medium and trypsin; see Note 19).
• 22Add the desired amount of flow cytometry buffer to a clean reservoir.
• 23Add 200 μL/well of flow cytometry buffer to each cell-containing well using a multichannel pipette and mix by pipetting up and down seven times to resuspend the cells (see Note 20).
• 24Add 30 μL of the solution in the HEK-293-T control cell well to two of the wells containing the tau RD P301S-CFP cells. Mix well by pipetting up and down.
• 25The plate is now ready for flow cytometry measurement. If the facility is not in the same lab and the plate needs to be transferred there, wrap it in tin foil for the duration of the transit to prevent photobleaching of the fluorophores before the assay.
• 26Perform the FRET flow cytometry as soon as possible (within 1 h) after preparing the plate as cell viability and fluorescence may reduce over time.

Fig. 3.

Fluorescent images of seeded and unseeded biosensor cells. (a) Unseeded CFP/YFP-conjugated P301S-tau RD cells in the absence of seeds. (b) CFP/YFP-conjugated P301S-tau RD cells seeded with 1-μg total protein per well PS19 mouse brain extract

3.3. Flow Cytometry (The Experiment Is Described for the Attune System and Should Be Adjusted to the System Used by Each Lab)

• 27In the Attune software, create a new experiment. In the heatmap tab, add wells to a new group until the plate on the screen matches your experimental plate (see Note 21).
• 28Set up the instrument so that the correct lasers are selected to measure 440 nm, 490 nm, and 527 nm light (in the Attune instrument settings tab, check the “BL1” slot and select “YFP” for the target and label it “FITC.” In the “VL1” slot, select “CFP” for the target and label “Pacific Blue.” In the “VL2” slot, select FRET for the target and label “Brilliant Violet 510”).
• 29Draw five graphs (see Note 22). The first is a scatterplot of forward scatter area (FSC-A) versus side scatter area (SSC-A; x,y). Draw a rough polygon gate (this will separate the live cells from dead cells and debris); the Attune program will label this gate R1 (see Fig. 4a). The next graph is a scatterplot of FSC-A, FSC-H (x,y) from the FSC-A versus SSC-A graph (R1). Draw a rough gate; this will be the gate for the single cells (the Attune program will label this R2, see Fig. 4b). Add a histogram of the CFP laser (VL1-A-CFP-Pacific Blue) from the FSC-A versus forward scatter height (FSC-H) graph (R2), a histogram of the YFP laser (BL1-A-YFP-FITC) from the FSC-A versus FSC-H graph (R2), and a histogram of FRET signal (VL2-A-FRET-Brilliant Violet 510) from the FSC-A versus FSC-H graph (R2).
• 30Next, set the voltages for FSC, SSC, YFP laser (BL1), CFP laser (VL1), and FRET signal (VL2). Follow your FRET flow cytometry facility guide for setting up the voltages (see Note 23).
• 31Apply 50 μL of HEK-293-T control cells + tau RD P301S-CFP cells using the slowest flow settings to gate the FSC versus SSC graph for cells. The gate should be between 200 × 10³ and the right border of the FSC-A axis and between 0.01 × 10³ and 500 × 10³ on the SSC-A axis (in Attune, this gate is labeled R1). Then, on the FSC-A versus FSC-H graph, gate the single cells (between 200 × 10³ and 800 × 10³ on the FSC-A axis and between 100 × 10³ and 500 × 10³ on the FSC-H axis) (in Attune, this gate is R2). Check to make sure the gates are where they need to be (see Note 24).
• 32Once the voltage and gates are set properly, the collection panel will need to be set. Select “collect entire plate from beginning,” and set acquisition volume to 150 μL and total sample volume to 200 μL. Set the desired speed and stop options for the machine to run (see Note 25, in Attune, be sure to click “apply to experiment”). Then run the experiment, save it, and export the FSC files to a labeled folder.

Fig. 4.

Initial gating. The initial gating separates living cells from debris and dead cells and subsequently single cells from clumps or clusters of cells. The separation is based on size and density. In the Attune flow cytometer software, the x-axis is the forward scatter area (FSC-A), indicating the size of the particle recorded, and the y-axis is the side scatter area (SSC-A), indicating the density of the object recorded. The title of the graph indicates that these are all the events recorded by the instrument (a) and all events in R1 (b) for well C4. (a) Gate the events that correspond to the parameters (size and density) of your cells. This gate will be automatically labeled “R1”. (b) Gate the events recorded by the instrument that correspond to the parameters (size and density) of single cells from R1. The single cell gate will be automatically labeled R2

3.4. Analysis of Flow Cytometry Results (We Use the FlowJO Software)

3.4.1. In Flow Cytometry Software (Instructions Using i, ii etc. Are for FlowJO)

• 33Import the folder containing the FRET flow cytometry FSC experiment files (see Note 26).
• 34
  Ensure the changes made to one gate will update on all the wells that gate applies to.

  1. Double-click on the file folder at the top of the screen. This will open a settings window for the experiment. Click <live-group> and <sync> and then <apply> and close the window (see Note 27).
• 35
  Gate the cells (see Fig. 5a).

  1. Double-click on any well. This will open up a window containing a heatmap graph of the data gathered from that well, where red areas show a high density of events and blue areas show a lower density of events. Verify that the y-axis is SSC-A and the x-axis is FSC-A. Click the polygon gating option at the top of the window and gate around the live-cell population excluding debris and dead cells (see Note 28). Extend the gate all the way to the right edge of the graph. Label this gate “Cells” (see Note 29).
  2. Drag and drop the “Cells” gate under the file folder. This applies the “Cells” gate to all of the wells (see Note 30).
• 36
  Gate single cells (see Fig. 5b).

  1. Under any well, double-click the “Cells” gate. On the graph that opens, set the y-axis to FSC-H and the x-axis to FSC-A. Using the polygon gating option, gate the single-cell population (see Notes 31 and 32).
  2. Drag and drop the “Single Cell” gate under the “Cells” gate at the top of the screen under the file folder to apply it to all of the wells.
• 37
  CFP/YPF compensation (see Note 21 and Fig. 6, which shows CFP/FRET compensation, but the steps are essentially the same).

  1. On the well that contains both the tau RD P301S-CFP cells and the HEK-293-T control cells, double-click on the “Single cells” gate. On the graph that opens, set the y-axis to CompBL1-YFP-FITC-A and the x-axis to CompVL1-CFP-Pacific Blue-A.
  2. There will be two groups of data points on the graph. The first group represents the HEK-293-T control cells, which show no CFP fluorescence and no YFP fluorescence. This group usually is circular with a red center and a blue “aura” around it (see Fig. 6a). The second group represents the tau RD P301S-CFP cells, which are positive for CFP fluorescence and slightly positive for YFP fluorescence. This residual YFP fluorescence is what we want to eliminate. This group typically is oblong and has a green/yellow center and blue aura. Use the “quad gating” option to gate the cells so that the y-axis of the quad gate is between the two groups of data points (in the blue area between the red dot and the green/yellow oval) and the x-axis is above both groups of data points (see Fig. 6a). On the main screen, four data lines will appear below the well: Q1, Q2, Q3, and Q4. HEK-293-T control cells will be in Q3 and the tau RD P301S-CFP cells will be in Q4. At the top of the screen, under workspace, add a median of compBL1A-YFP-FITC-A to Q3 and Q4 (YFP-FITC-A-).
  3. On the main screen, double-click on the matrix grid beside the same well. In the matrix window, click in the CFP, YFP cell and change the compensation coefficient to compensate between CFP and YFP so that the medians of Q3 and Q4 are within 10 points of each other (see Note 33).
• 38
  CFP/FRET compensation (see Fig. 6).

  1. Return to the window displaying the graph of CompBL1-YFP-FITC-A versus CompVL1-CFP-Pacific Blue-A and change the y-axis to CompVL2-FRET-Brilliant Violet 510-A. The x-axis remains the same at CompVL1-CFP-Pacific Blue-A.
  2. There will be two groups of data points on the graph. Using the “quad” gating option, gate the cells so that the y-axis of the quad gate is between the two groups of data points and the x-axis is above both groups (see Fig. 6a). On the main screen, four data lines will appear below the well: Q5, Q6, Q7, and Q8. At the top of the screen, under workspace, add a median of CompVL2-FRET-Brilliant Violet 510-A to Q7 and Q8 (FRET-Brilliant Violet 510-A).
  3. On the main screen, double-click on the matrix grid beside the same well. In the matrix window, click in the CFP, FRET cell and change the compensation coefficient to compensate between CFP and FRET so that the medians of Q7 and Q8 are within 10 points of each other (see Fig. 6b, Note 34).
• 39
  Gate false FRET (see Fig. 6c).

  1. On the well containing only the tau RD P301S-YFP cell, click on the “Single Cells” gate.
  2. In the window that opens, set the graph axes so that the y-axis is CompVL2-FRET-Brilliant Violet 510-A and the x-axis is CompBL1-YFP-FITC-A.
  3. Use the polygon gating option to gate the cells so that the gate excludes most cells. Angle the gate up and to the right at the right end of the graph. Ensure that 1% of the population of cells is inside the gate.
  4. Label the gate “False Fret.”
  5. Drag and drop the “False Fret” gate under the “Single Cell” gate at the top of the screen under the file folder to apply it to all of the wells.
• 40
  Gate FRET signal (see Fig. 6d).

  1. On the well containing only biosensor cells with no seeds, double-click on the “Single cells” gate.
  2. In the window that opens, set the graph axes so that the y-axis is CompVL2-FRET-Brilliant Violet 510-A and the x-axis is CompVL1-CFP-Pacific Blue-A.
  3. Use the polygon gating option to gate the cells so that the gate excludes most cells. Angle the gate up and to the right at the right end of the graph. Ensure that 1% of the population of cells is inside the gate.
  4. Label the gate “FRET.”
  5. Drag and drop the “FRET” gate under the “False Fret” gate at the top of the main screen under the file folder to apply the gate to all the wells.
• 41
  Inspect the FRET Signal.

  1. On the main screen, double-click on the “Single Cell” gate under any well that contains seeds. In the window that opens, set the graph axes so that y-axis is CompVL2-FRET-Brilliant Violet 510-A and the -axis is CompVL1-CFP-Pacific Blue-A. The points inside the gate represent the FRET signal.
• 42
  Calculate the median of the FRET signal and “Frequency of Grandparent” (see Note 35).

  1. On the main screen, select the “FRET” gate under any well and add a median of CompVL2-FRET-Brilliant Violet-A. Drag and drop the median under the “FRET” gate at the top of the screen under the file folder.
  2. On the main screen, select the “FRET” gate under any well. Under the menu bar at the top of the window, click <Add Statistic> under the “workspace” tab. Choose “Freq. of Grandparent” and click <apply>.
  3. Drag and drop the “Freq. of Grandparent” statistics under “FRET” at the top of the screen under the file folder.
• 43
  For further analysis, the results can be saved and exported to external programs used for processing. Here, we provide examples for using Microsoft Excel and GraphPad Prism for subsequent analyses.

  1. Save workspace and export the file to Microsoft Excel (File > Save As >Export to Excel).

Fig. 5.

Analysis in FlowJO. (a) Gating the events that correspond to the parameters (size and density) of your cells. (b) Gating the events that correspond to the parameters (size and density) for single cells

Fig. 6.

CFP-FRET compensation. (a) A tau RD P301S-CFP + HEK-293-T control cells well before CFP-FRET compensation. The inset shows the HEK-293-T control cell group in Q8. The y-axis of the “quad gate” to the right of the group separates the HEK-293-T control cell group in Q8 from the tau RD P301S-CFP group in Q7. (b) A tau RD P301S-CFP + HEK-293-T control cells well after a CFP-FRET compensation coefficient was added. (c) A YFP well gating false FRET. (d) A tau biosensor control well gating FRET signal. (e) A tau biosensor well seeded using mouse brain extract. The events inside the gate labeled “FRET” is the FRET signal

3.4.2. Multiply Frequency of Grandparent by Median of FRET

• 44
  The example is given for Microsoft Excel.

  1. Open the Excel file.
  2. In Excel, multiply “freq. of grandparent” × “median of FRET” for all wells with seeds and the control wells (biosensor cells only, exclude tau RD P301S-CFP cells, tau RD P301S-YFP cells, and HEK-293-T control cells).

3.4.3. Analyzing Data by Subtracting Baseline Values from Data Set

• 45The example is given using GraphPad Prism.
• 46In Prism, create a new, grouped, three-replicate file.
• 47Copy the product of “freq. of grandparent” × “median of FRET” into Prism.
• 48Ensure that the values for the biosensor cell control wells are in row 1.
• 49
  Analyze the data and create a graph.

  1. Remove baseline and column math (see Fig. 7).

     1. Click Analyze; select remove baseline and column math. Define baseline as first row (biosensor cell control wells). Set calculation to “ratio: value/baseline.” Select “Replicates. No matching. Average the baseline replicates and do calculations with the average” for subcolumns.
     2. Insert a new graph of the existing data.

Fig. 7.

Tau biosensor cells seeded with PS19 mouse brain extracts. The left hemisphere of five different ~1-year-old PS19 mouse brains (from “m1” to “m5” bars) was cryoground and extracted using Tris-buffered saline supplemented with protease and phosphatase inhibitors. Tau biosensor cells were seeded using 1- or 2-μg total protein per well. The analysis shows substantial differences in integrated FRET density among the mice. Seeding using 1- or 2-μg total protein per well yielded essentially the same result. Control wells were incubated with Lipofectamine and opti-MEM in the absence of brain extract. The baseline was subtracted from the raw values using the “remove baseline and column math” function in Prism

4. Notes

1. We have found that the Attune flow cytometer works well but any flow cytometer equipped with filters for measuring fluorescence emissions from tau RD P301S-CFP cells and tau RD P301S-YFP cells will work.

2. Detergents, protease/phosphatase inhibitors, or any cytotoxic compounds should be avoided in the preparation of the seeds. Recombinant tau or α-synuclein or other animal and human tissue extracts can be used as seeds. We have tested different ways of preparing brain extracts, including sonication, cryogrinding, and digestion with papain.

3. We used a Keyence BZ-X710 fluorescence microscope in this example, but any fluorescent microscope will work as long as the appropriate filters are available. Taking fluorescent images is optional, and although it provides additional qualitative data, it is not necessary for the FRET-based flow cytometry analysis.

4. Count cells per your lab’s protocol. We count cells using a cell-counting plate INCYTO C-Chip and microscope, but any method of cell counting will suffice.

5. Ensure that the media and trypsin are at 37 °C before adding them to the cells.

6. Make sure the tip of the pipette does not touch the surface of the cells as this will disturb and possibly damage them.

7. When adding the trypsin solution to the cells, add it slowly to the wall of the flask, not directly onto the cells.

8. If the cells have not detached completely from the surface of the flask, place them back in the incubator for another 1–2 min.

9. The HEK-293-T cell line grows faster than the biosensor cell line. Therefore, it is best to dilute the HEK-293-T cells a bit more than 15 × 10³. Typically, we dilute the HEK-293-T cells, tau RD P301S-CFP cells, and tau RD P301S-YFP cells to a concentration of 18–20 × 10³.

10. It is best to run experiments in triplicates to validate reproducibility. For the HEK-293-T cell, tau RD P301S-CFP cell, and tau RD P301S-YFP cell conditions, plate three wells each. Each well contains 130 μL of cells. For the HEK-293-T control cells, tau RD P301S-CFP cells, and tau RD P301S-YFP cell wells, you will need 390 μL total volume of the cell suspension.

11. For the first FRET-based flow cytometry experiments, it is advisable to plate a few extra wells for each cell line. The extra wells help with programing the settings of the flow cytometer. Once the settings are adjusted, they can be saved for future experiments and only three replicates of each well are needed.

12. When adding the cells to the well, add them along one side of the well instead of in the center to get an even spread. If cells are added to the center of the well, they tend to collect around the edge of the well making it more difficult to take fluorescent images.

13. We recommend preparing samples at least in triplicates so make a total volume of 30 μL of each of the sample tubes.

14. If you are doing your experiment in triplicates, you will add 30 μL of the O + L mixture to each sample tube including the control tube.

15. It takes ~20 min for the Lipofectamine to form liposomes around the seeds. This is not a precise time. We have found no difference between incubating the seeds and Lipofectamine for 20 min versus 1 h.

16. Samples can stick to the walls of the microcentrifuge tubes. To avoid running out of sample, make enough for one more well than needed.

17. The incubation time can be changed based on the experiment. When changing the incubation time, take into consideration the time needed for the seeds to induce intracellular aggregation and the growth rate of the cells. If you are incubating for a longer period of time than described here, the concentration of cells plated may need to be adjusted.

18. Although aspirating the medium out using a pipet is more sterile, it may damage the cells. We have found that discarding the media by shaking over a biohazard bin is sufficient.

19. We have found that attaching a 200-μL unfiltered pipette tip to the vacuum tube works better than using a plastic Pasteur pipette. The pipette tip allows for more precise aspiration of the supernatant and is less likely to damage the cells than a Pasteur pipette.

20. When resuspending the cells, pipette slowly so as not to agitate the solution and create air bubbles. If the well has air bubbles on the surface, the flow cytometer will not read the well. If air bubbles are accidentally introduced, try to gently aspirate them off.

21. We have found that the Attune flow cytometry machine requires the least compensation for CFP bleeding into FRET than older machines such as the LSR. The Attune has a laser with a FRET channel that is farther away in wavelength from the CFP channel (Pacific Blue 450 nm). The Attune machine allows you to use the brilliant violet 510 channel (512/25 nm) for FRET, whereas older machines such as the LRS only has the AmCyan-a channel (500 nm) as an option for FRET. The phenomenon of “bleeding” is when CFP gives a false FRET signal. This is because the detector that measures CFP-emitted light detects wavelengths of 450 ± 40 nm, which overlaps with the detector that measures the FRET signal, which detects wavelengths of 512 ± 25 nm. This overlap of detectable emitted light between the CFP channel and the FRET channel means that CFP will produce a false FRET signal which is compensated for during analysis. We also compensate for CFP bleeding into YFP. Even though the lasers (CFP laser 405 nm, YPF laser 488 nm) and channels (CFP-VL1-Pacific Blue 450 ± 40 nm, YFP BL1 FIT-C 530 ± 30 nm) are farther apart than the CFP and FRET lasers and channels, we still compensate for the small amount of bleeding of CFP into YFP.

22. The CFP channel (VL1-Pacific Blue) graph shows the events recorded at a 450 ± 40 nm when a 405-nm light is applied. The YFP channel (BL1 FIT-C) graph shows the events recorded at 530 ± 30 nm when a 488-nm light is applied. The FRET channel (VL2-Brilliant Violet) graph shows events detected at 512 ± 25 nm wavelength when a 405-nm light is shown.

23. We have found that FSC is ~30, SSC is ~275, BL1 is ~150, VL1 is ~220, and VL2 is ~190 for HEK-293-T cells and biosensor cells. However, these numbers depend on the size and density of the cells and the machine.

24. When the voltages and gates are set correctly, the following will be true: For the control wells there will be no signal for the tau RD P301S-CFP, tau RD P301S-YFP, or FRET histogram (meaning that the peak will be at or below 10³). For the tau RD P301S-CFP + HEK-293-T control wells, there will be a small peak at or below 10³ and a larger one above 10³ for the tau RD P301S-CFP histogram, a peak at or below 10³ for the tau RD P301S-YFP histogram, and two peaks for the FRET histogram – one at or below 10³ and one that is slightly over 10³ (compensating in FlowJO afterwards will fix this bleeding). For the tau RD P301S-YFP cells, there will be a peak at or below 10³ for the tau RD P301S-CFP and FRET histograms. For the tau RD P301S-YFP histogram, there should be one peak above 10³. Lastly, for the biosensor cells, there should be two peaks for both tau RD P301S-CFP and tau RD P301S-YFP histograms – one small at or below 10³ and one slightly above 10³. There should be a peak at or below 10³ for the FRET histogram.

25. We usually set the machine to run at 500 μL/min. The faster the machine runs, the more pressure it puts on the cells. For stop options, we typically set it to stop at 10,000 events on R2.

26. Details provided are for FlowJO version 10.6.2.

27. This ensures that the changes made to one gate will update on all of the wells that gate applies to.

28. This population should be around (450 K, 200 K).

29. The “Cells” gate will appear in black under the well you selected.

30. The gate will turn blue under each well.

31. The single-cell population is approximately 350–800 K on the x-axis (FSC) and approximately 180–400 K on the y-axis (SSC).

32. The program will automatically name the gate “Single Cells.” If not, check to make sure the axes are correct.

33. If you are using the Attune instrument, the compensation coefficient between CFP and YFP is usually between 0.4 and 1.5. This can vary based on the cells, seeds, and specific experiment.

34. If you are using the Attune instrument, the compensation coefficient between CFP and FRET is usually between 45 and 80. This can vary based on the cells, seeds, and specific experiment.

35. The statistic “Frequency of Grandparent” is calculated by dividing the number of events in the subpopulation by the number of events in its ancestor population.