Fluorescence Polarization-based Measurement of Protein-Ligand Interaction in Fungal Cell Lysates

Abstract

Fungi infect over a billion people worldwide and contribute substantially to human morbidity and mortality despite all available therapies. New antifungal drugs are urgently needed. Decades of study have revealed numerous protein targets of potential therapeutic interest that still await discovery and development of potent, fungal selective ligands. To measure the binding of diverse small molecule ligands to their larger protein targets, fluorescence polarization (FP) can provide a robust, inexpensive approach. The protocols in this article provide detailed guidance for developing FP-based assays capable of measuring binding affinity in whole cell lysates without the need for target protein purification. Applications include screening of libraries to identify novel ligands and the definition of structure activity relationships to aid development of compounds with improved target affinity and fungal selectivity.

INTRODUCTION:

Fluorescence polarization (FP) is a sensitive and robust technique for the measurement of protein-protein and protein-ligand interactions (Lea & Simeonov, 2011). FP is based on the principle that excitation of a small fluorescent molecule in solution with plane-polarized light results in the emission of light that is largely depolarized due to rapid rotational reorientation during the lifetime of its excited state. However, if the labeled molecule binds to a larger protein, the rotational diffusion is reduced and an increase in polarized emission signal, proportional to the bound fraction, can be detected (Figure 1A). This principle of FP allows for the measurement of ligand binding (Moerke, 2009). Similarly, competitive displacement of a fluorescent ligand by a non-fluorescent test compound can be monitored to assess relative binding affinity of the test compound (Figure 1B). Such assays are well adapted for high-throughput screening and are widely used in drug development programs (Hall et al., 2016). There are many advantages to an FP-based screening approach including an all-in-one in-solution assay format with no requirement for separation of bound and free ligand (Moerke, 2009). Further, FP experiments can be less susceptible to assay interference problems when compared to other light-based approaches and are relatively low cost (Hall et al., 2016). FP-based assays have been used to investigate a myriad of both common and challenging drug targets including kinases, receptors, ion channels, transcription factors, and epigenetic regulators (Hall et al., 2016). Further, FP provides sufficient sensitivity and specificity to detect target protein binding of ligands in human, bacterial, and fungal whole cell lysates (Sun, Nguyen, Harold Ross, Hollis, & Wynn, 2002; Whitesell et al., 2019).

Figure 1. A. Schematic of the basic principle of FP applied to the measurement of ligand binding.

After passing through an excitation polarizer, plane-polarized light excites the fluorescent tracer. Emitted light intensity is measured after passing through emission polarizers parallel (I ∥) and perpendicular (I⊥) to the excitation light’s plane of polarization. The low molecular weight free tracer exhibits relatively high rotational diffusion during the excited state lifetime (τ) resulting in the emission of depolarized light. The much larger protein-bound tracer complex exhibits reduced rotational diffusion during τ and thus more of the emitted light remains polarized in the same plane as the excitation light. B. FP assay for the assessment of a test compound’s relative binding affinity to a protein target in whole cell lysate. A fluorescently labeled small molecule (tracer) tumbling in solution results in a low measured FP. When the tracer binds with high affinity to a larger protein target, the rotational reorientation is minimized, and the measured FP is high. Addition of a test compound which competes for binding to the same target displaces the tracer in a concentration- dependent manner, reducing the measured FP.

^a Lysates are prepared following Support Protocols 1 and 2. Lysate protein concentration and optimal incubation time are determined by defining the saturation level of polarization as described in Basic Protocol 1.

^b Relative binding affinity of an investigational compound is established by generating competitive displacement curves as described in Basic Protocol 2.

Fungi are greatly underappreciated contributors to human morbidity and mortality (“Stop neglecting fungi,” 2017). Systemic fungal infections are associated with alarming mortality rates, often exceeding 50% and result in the death of an estimated 1.5 million people annually (G. D. Brown et al., 2012). Despite the increase in high-risk immunocompromised populations, including cancer chemotherapy patients, organ transplant recipients, and those infected with HIV, systemic fungal infections remain notoriously difficult to manage, in part due to the frequent emergence of resistance to the limited number of available antifungal treatment options. New antifungal agents are direly needed as the utility of current drugs is also limited by host toxicity, requirement for intravenous administration, or a narrow spectrum of activity (Robbins, Wright, & Cowen, 2016). Extensive research over the last few decades has elucidated a large number of promising antifungal drug targets based on their requirement for growth, virulence or antifungal drug resistance (J. C. Brown et al., 2014; Caplan et al., 2018; Carr et al., 2010; Hu et al., 2007; Lee et al., 2020; Liu et al., 2008; Peng, Zhang, Xu, & Tan, 2018; Roemer et al., 2003; Segal et al., 2018). Biochemical target engagement and structure-guided approaches are now helping fill the antifungal pipeline. Amongst these, fluorescence polarization provides an excellent tool for screening compound libraries against fungal targets, optimizing fungal selectivity and guiding drug development.

The use of whole cell lysates obviates the need for protein tagging which can alter function and recombinant protein expression and purification which can be laborious, especially in a variety of poorly genetically tractable fungal organisms. Further, the use of crude cell extracts allows for the investigation of ligand binding by proteins in their native complexes and in a biologically relevant mix of protein isoforms. Such advantages can be crucial in efforts to optimize selectivity at the whole-cell level.

The protocols detailed in this article describe a general approach to optimizing FP competition-based experiments in fungal and human whole cell lysates, especially in regard to the identification and optimization of inhibitors to fungal drug targets of therapeutic interest. Basic protocol 1 outlines steps to determine optimal tracer and lysate protein concentrations through the generation of fluorescence-polarization saturation binding curves. Once these parameters are defined, Basic protocol 2 walks through steps to establish competition-based assays capable of measuring the relative binding affinity of unlabeled molecules, keeping in mind the caveat that lysate experiments can only provide assay-dependent EC₅₀ values for test compounds. The support protocols provide the necessary details for preparation of whole cell extracts from fungi such as species of Candida and Cryptococcus and human tissue culture cell lines (HepG2) that can be used to assess binding affinity and selectivity with an FP-based approach.

CAUTION:

Follow all biosafety requirements relevant to the fungal species under investigation.

STRATEGIC PLANNING

The FP assays described in this protocol require a fluorescently labeled probe molecule referred to as “tracer”. The choice of compound to be labeled along with an appropriate fluorophore must be carefully considered for the design of a suitable tracer. The probe molecule should have high affinity for the protein of interest (X. Huang, 2003) and must be amenable to fluorescent labelling without loss of target affinity. As such, the location, composition and length of the linker connecting ligand to fluorophore is also an important consideration (described in more detail in Critical Parameters). Ultimately, the optimized tracer should be usable at very low concentrations, in the range of 0.1 – 100 nM, and should not exceed twice the dissociation constant (K_D) of the unlabeled ligand.

Important technical considerations for the choice of the fluorophore are its fluorescence lifetime, which should be short for small molecule probes (<6 ns) (Hall et al., 2016), as well as its excitation and emission wavelengths which must be compatible with the FP-capable plate reader available for use (Table 1). Historically, fluorescein has been the most widely used fluorescent label (Hall et al., 2016). However, red-shifted fluorophores, such as Cy3B and Cy5, are becoming increasingly common as they provide many advantages for fluorescence polarization applications in high throughput drug screening. Red-shifted dyes have longer wavelengths and have been shown to minimize the effect of fluorescence interference on polarization signal (Simeonov et al., 2008). Assay interference can result from the autofluorescence of whole cell lysates or from the test compounds in screening libraries. These molecules can be intrinsically fluorescent, or precipitate to form insoluble crystals which often increase light scattering and alter the polarization signal, thus leading to false positives.

Table 1.

Properties of Fluorophores Commonly Used for FP

Fluorophore	Fluorescence Lifetime (ns)	Excitation (nm)	Emission (nm)
Cy3B	2.8	558	572
Cy5	1.0	646	664
Fluorescein	3.8	494	519
5-TAMRA	2.5*	542	568
Texas Red	4.2	595	615

Data extracted from (Hall et al., 2016) except for * from (Savarese et al., 2012).

BASIC PROTOCOL 1. USE OF SATURATION BINDING CURVES TO OPTIMIZE TRACER AND LYSATE PROTEIN CONCENTRATIONS

This article uses the example of a Cy3B-labeled geldanamycin probe for the measurement of relative binding affinity for experimental inhibitors of Hsp90 in fungal and human whole cell lysates (D. S. Huang et al., 2020; Moulick et al., 2006; Whitesell et al., 2019). Many fluorophores with linkers of various length and chemical reactivities are commercially available. These reagents generally provide a suggestion for the ideal synthetic route to label the desired probe molecule. Depending on the probe molecule and fluorophore, synthesis of the tracer can often be requested commercially. As mentioned, the probe molecule selected should have high target affinity and specificity. It must also be amenable to labeling at a position that does not impair target binding. A wealth of information on antifungal targets of interest can be mined from various fungal genome databases such as the Candida Genome Database (http://www.candidagenome.org/). Some reagents are available commercially for rapid FP assay development, as discussed later in Background Information. For this protocol, our probe molecule geldanamycin binds to the N-terminal ATP-binding domain of Hsp90, inhibiting its ATPase activity. This interaction forms the basis of the fluorescence polarization assay to measure protein-ligand interactions (Prodromou et al., 1997; Panaretou et al., 1998). With the Cy3B-tagged geldanamycin probe synthesized as previously reported (Moulick et al., 2006), relevant whole cell lysates are generated (Support Protocols 1 & 2) in order to determine the binding affinity of the tracer for its target protein, Hsp90, in a lysate.

Most fluorophores are light sensitive and precaution should be taken throughout this protocol to reduce tracer exposure to light, as annotated in relevant steps. If the K_D of the molecule that is fluorescently tagged is known, a concentration of tracer equivalent to the K_D generally provides a good starting point to use in assay development. The concentration should then be lowered, while assuring an appropriate signal window is maintained (fluorescence polarization window should be at least 100 units of millipolarization (mP) between unbound and fully bound tracer). If the K_D is not known, a range of concentrations, from 0.1 nM up to 0.1 μM, can be tested to determine the optimal concentration for the assay. The assay can be set up in 384-well plate format to be amenable for high throughput screening. A titration of lysate protein concentration is first performed at a fixed concentration of tracer to define binding affinity. The goal is to define a plateau or saturation level (maximum polarization) where most of the tracer is bound to the target protein allowing estimation of binding affinity. Thus, unlike classical radioligand binding assays which are typically performed under saturating ligand concentrations, the FP approach relies on use of saturating target conditions. In the early stages of assay development, it may be necessary to test a range of tracer concentrations as well as lysate protein concentration ranges to determine the optimal conditions for signal detection and to maximize the polarization assay window. It is ideal to start at a high concentration of lysate (e.g. 100 μg/well) to ensure that saturation binding of the tracer to the target protein is achieved. The protein concentration of lysate can then be lowered, and additional dilution points added to define the curve as needed for further optimization. The optimal concentration of lysate protein to obtain appropriate tracer binding will depend on the abundance of the particular target of interest in the lysate and may vary between different batches of whole cell lysate due to the variation in target protein expression by cells in culture. Notably, in addition to absolute abundance, structural differences and variant post-translational modifications of a target protein can exist between species which may alter tracer binding affinity. Thus, the optimal lysate concentration must be determined for each species. Finally, depending on the binding kinetics of the tracer to the target protein, assay incubation time may also require optimization to achieve equilibrium binding. The time required can range anywhere from minutes to hours depending on tracer and protein interaction. For example, geldanamycin binding to Hsp90 exhibits unusually slow, tight binding kinetics. To achieve equilibrium binding requires approximately 4 hours (Gooljarsingh et al., 2006). Thus, the same assay plate should be read at multiple time points (e.g. 30 minutes, 1 hour, 4 hours, and 24 hours) during assay development to determine the interval required for binding to reach steady state.

An FP-based assay designed to measure Cy3B-geldanamycin binding to its relatively abundant cytosolic target Hsp90 in human and fungal whole cell lysates is presented as a specific example to illustrate the steps required and the anticipated results of a saturation binding experiment in which average fluorescence polarization (mP) is measured as a function of increasing lysate protein concentration.

Materials

• FP Assay Buffer (see recipe)

• Dimethyl sulfoxide (DMSO) (e.g. Sigma-Aldrich, cat. no. D2438-50ML)

• Whole cell lysates (quantified for protein concentration, e.g. 15 mg/mL HepG2; 10 mg/mL C. albicans strain SC5314; see support protocols)

• Tracer stock solution in DMSO (e.g. 500 nM)

• Microcentrifuge tubes 1.7 mL (e.g. FroggaBio, cat. no. LMCT1.7B)

• Black well/black bottom opaque 384-well microtiter plates (Greiner bio-one, medium binding cat. no. 781076 or low binding plates cat. no. 781900)

• Disposable pipetting reservoirs 25 mL (VWR, cat. no. 89094-662)

• Multichannel pipette (e.g. Mettler Toledo, 2–20 μL LTS)

• Aluminum foil

• Benchtop clinical centrifuge (e.g. Beckman Coulter, Allegra X-14)

• FP-capable plate reader (e.g. Tecan Spark^® multimode microplate reader)

• Excel spreadsheet for data analysis

Prepare samples in duplicate for FP measurements

• 1Prepare sufficient FP Assay Buffer for lysate dilutions and titrations, and tracer dilutions allowing for technical duplicates and a final assay volume of 25 μl per well.

Start lysate titration at 50 μg/well and test fixed tracer concentrations. As an example, for the very potent Cy3B-geldanamcin tracer, examine concentrations of 0.05 nM, 0.1 nM, and 0.2 nM.

• 2Calculate the amount of lysate needed to achieve a starting concentration of lysate protein of 50 μg/well and prepare highest dilution concentration in the FP Assay Buffer in microcentrifuge tubes.

Lysate protein concentration is determined by Bradford reagent (see Support Protocol 1) and used in the assay as μg lysate per well (μg/well). As three different tracer concentrations are being tested in technical duplicates, 6 wells will be filled, but extra should be prepared to ensure enough reagent is available, for instance by calculating for 8 wells. As per the example for HepG2 lysate (15 mg/mL), add 26.7 μl to final 200 μl assay buffer and for C. albicans lysate, add 40 μl to final 200 μl assay buffer. Vortex to ensure proper mixing of lysate and assay buffer.

• 3Set up a series of eleven two-fold lysate dilutions: in duplicate wells in column 1 of the assay plate, pipet 25 μL of the lysate stocks (step 2).
• 4Using multichannel pipette, dispense 12.5 μL of assay buffer to the remaining wells of columns 2–12.
• 5Titer lysate across the plate by pipetting 12.5 μL from column 1 and mixing it well with column 2 then repeating until column 11. Leave the last column 12 as the assay buffer only, to act as control for unbound tracer.

This is a good starting point for assay development; additional dilution points can certainly be added to further define the curve.

• 6Prepare a 2X concentrated dilution in the assay buffer of the tracer at the highest concentration being tested and perform 2-fold serial dilutions to prepare the additional two concentrations in microcentrifuge tubes.

2X-concentration is required as the tracer will be diluted in half once added to the plate containing 12.5 μl of diluted lysate per well. As per the example, add 1.2 μl of 500 nM tracer to final 1.5 mL of assay buffer to achieve 0.4 nM tracer dilution. 750 μl of 0.4 nM dilution to be added to 750 μl of assay buffer to obtain 0.2 nM tracer dilution and lastly, 750 μl of 0.2 nM dilution to 750 μl assay buffer to achieve 0.1 nM tracer dilution. To prevent prolonged exposure to light, tracer dilutions should be prepared in amber microcentrifuge tubes.

• 7Dispense the tracer stock into a disposable pipetting reservoir. Use a multichannel pipette to dispense 12.5 μl of tracer stock into the respective wells in the assay plate for that particular dilution. After the tracer has been added to the assay plate, ensure it is protected from light by covering the plate in foil.

Prepare all the tracer dilutions immediately prior to use as the tracer will start to bind the lysate protein when added to the wells.

• 8Include buffer only wells for background fluorescence measurements.

In the beginning stages of assay development, include wells with lysate only to determine the level of auto fluorescence that could interfere with the polarization signal, usually this is negligible.

• 9Spin the plate in a table-top centrifuge at 1,000 rpm for 1 min at room temperature.
• 10Incubate the plate at room temperature, covered, for the optimized time.

It may be necessary to measure at various timepoints to determine optimal incubation time to reach equilibrium binding. In this example, the incubation period is 4.5 hrs. Optimal incubation temperature can also be evaluated and might vary depending on the stability of the target in lysate, non-specific binding of the probe to other lysate components and the time required for binding to come to equilibrium.

FP measurements

• 11Load the plate into the plate reader. Use the plate reader control software to select an FP acquisition protocol. Set the relevant parameters such as type of plate, excitation and emission wavelengths of the tracer fluorophore (see Table 1). Then select any well to optimize the well height, select the optimal gain option for the plate, and identify the tracer only control wells to determine the G factor (see Background Information).
• 12Save the protocol with the optimized values for well height (referred as Z-position in specified software), gain, and G-factor.

The values determined for G-factor and Z-position will be consistent between assays if the same tracer, plate reader and final volume per well are used. Values for gain will vary depending on the tracer concentration. Note that depending on the plate reader and software used, these parameters could be defined differently.

• 13With the plate layout defined, including the buffer only wells defined as blanks, read the FP signal for the entire plate and export the data to Microsoft Excel. At this time, the same plate can be read at multiple time points to determine the time required to achieve equilibrium binding.

Data analysis

• 14Format the data in a manner that allows determination of average FP measurements in millipolarization (mP) and plot the data with the corresponding lysate concentrations for each tracer concentration tested.

Polarization values are generated with millipolarization (mP) units. Anisotropy values are also provided along with total fluorescence intensity, and fluorescence intensity polarized parallel and perpendicular to the incident light. The software typically corrects for background fluorescence as measured from the buffer only wells.

• 15Determine the appropriate concentration of lysate and tracer that achieves approximately 75% of tracer binding from the saturation binding curves.

Figure 2 shows the expected saturation binding curves for different tracer concentrations for lysates prepared from either fungal or human cells. At this point, it could be necessary to perform the assay again at different lysate and/or tracer concentrations. It is important to perform the binding curves in independent replicates to ensure reproducibility. When testing lysates from multiple species, it is ideal to select a tracer concentration that is consistent and vary the concentration of the lysate to achieve 75% tracer binding for all samples, as indicated by the dotted line. Here, we would choose a tracer concentration of 0.1 nM and lysate protein concentrations of 1.25 μg/well and 3.5 μg/well for HepG2 and C. albicans, respectively.

Figure 2. Saturation binding of the tracer, Cy3B-geldanamycin, to the target protein, Hsp90, in HepG2 and C. albicans lysates.

The average fluorescence polarization (mP) is plotted against increasing lysate protein concentration with three fixed tracer concentrations, either 0.2, 0.1, or 0.05 nM. In this assay, the highest concentration of lysate was 20 μg/well, as binding saturation was observed at this concentration. The dotted line indicates 75% tracer binding at a tracer concentration of 0.1 nM.

BASIC PROTOCOL 2. ESTABLISHMENT OF COMPETITION BINDING EXPERIMENTS

From the saturation binding experiment (Basic Protocol 1), appropriate tracer and protein concentrations can be established for competition binding experiments. For the most robust assay window, optimal conditions include a protein concentration which yields tracer polarization of ~75% of the maximal level. Competition assays measure the decrease in FP as tracer is displaced by an unlabeled ligand. As a control, the same small molecule used in generating the tracer, unlabeled, should be titrated against the pre-determined protein-tracer mix to ensure that FP values decrease to that observed for free (unbound) tracer. The displacement curve that plots average polarization (mP) against log concentration of compound should have a sigmoidal shape (Figure 3). Curve-fitting software can then be used to determine an EC₅₀ for each candidate ligand tested.

Figure 3. Competitive tracer displacement by unlabeled geldanamycin in C. albicans lysate.

The average fluorescence polarization (mP) is plotted against increasing geldanamycin concentration on a log10 scale. The concentration of tracer was 0.1 nM and the lysate protein concentration was 3.5 μg/well.

Materials:

• Small molecule competitors in DMSO stocks

• FP Assay Buffer (see recipe)

• Whole cell lysates (see support protocols)

• Tracer stock in DMSO

• Microcentrifuge tubes 1.7 mL (e.g. FroggaBio, cat. no. LMCT1.7B)

• Black well/black bottom opaque 384-well microtiter plates, medium binding (Greiner bio-one, cat. no. 781076)

• Disposable pipetting reservoirs, 25 mL (VWR, cat. no. 89094-662)

• Multichannel pipette (e.g. Mettler Toledo, 2–20 μL LTS)

• Aluminum foil

• Benchtop centrifuge (e.g. Beckman Coulter, Allegra X-14)

• FP-capable plate reader (e.g. Tecan Spark^® multimode microplate reader)

• GraphPad Prism or other graphing software

Sample preparation for FP measurement

• 1Make 1 mM working dilutions of investigational compounds in DMSO. Always include a positive control compound such as the unlabeled version of the tracer.

These dilutions can be kept at −20°C and thawed for each experiment (freeze-thaw cycles are tolerated).

• 2Make FP Assay Buffer fresh (see recipe).
• 3Dilute compounds to 2X desired highest concentration in FP Assay Buffer in microcentrifuge tubes.

For known potent compounds (e.g. clinical drugs), we recommend a compound dilution series starting at 2 μM. For test compounds, we recommend starting anywhere from 10 to 50 μM depending on the anticipated binding affinity.

• 4Set up a series of eleven two-fold compound dilutions: pipet 25 μL of 2X compound stock in duplicate wells in column 1 of the assay plate. Using a multichannel pipette, dispense 12.5 μL of assay buffer to remaining wells of columns 2–12. Repeat for the second half of plate (ie. 25 μL of 2X compound in duplicates in column 13 and 12.5 μL of assay buffer in wells of columns 14–24). Leave at least one row without compound and include duplicate wells of assay buffer alone at the final concentration per well of 25 μL. A recommended plate layout is presented in Figure 4.
• 5Titer 2X compound stock across plate by pipetting 12.5 μL from column 1 and mixing it with column 2 and repeating until column 11. Leave the last column 12 as the buffer only, to act as control for fully bound tracer.
• 6Prepare lysate dilution (based on Basic Protocol 1) in a conical tube and mix well by gentle vortexing.
• 7Add 2X tracer to lysate dilution, mix well by gentle vortexing and immediately add to all wells, except the duplicate wells for assay buffer only (which serve as blanks for the FP measurements). Keep plates in dark or wrap in foil.

Figure 4. Recommended plate setup for competition-binding experiment.

To assess candidate inhibitors, a control compound (blue wells, ideally the unlabeled tracer molecule) should be used in addition to the investigational compounds (yellow wells) in technical duplicate. Compounds should be titrated in two-fold dilutions across twelve wells to which the pre-determined lysate-tracer mix is added. Duplicate wells of FP assay buffer only serve as blanks (green wells).

The tracer must be added to the lysate immediately before use and quickly added to all the wells as the tracer will begin to bind the lysate protein.

• 8Spin plate(s) at 1,000 rpm for 1 min.
• 9Incubate the plate at room temperature, covered with foil, for the optimized time interval (determined in Basic Protocol 1).

FP measurements should be made when the displacement reaction has reached equilibrium (which can range from minutes to hours). Longer incubation periods are required for tracer-protein complexes of that form with slow kinetics. The assay is determined to be at equilibrium if the EC₅₀ is constant between two incubation intervals. If apparent affinity increases with time, equilibrium has not been reached (Hulme & Trevethick, 2010).

FP measurement and data analysis

• 10As in Basic Protocol 1, set up the plate reader to restore the saved parameters including plate type, measurement height, gain and G-factor; and fill in the plate layout identifying the plate blanks (i.e. wells with assay buffer only).
• 11Read the plate to obtain measurements of fluorescence polarization for all sample wells.
• 12Export the data and plot the average FP measurement of the duplicate wells for each data point as a function of the log10 of the ligand concentration.
• 13Use a graphing software package (e.g. GraphPad Prism) to create a curve fit and determine the EC₅₀ (Figure 3).

If using GraphPad Prism;

• Create a new data table and graph, starting with an empty data table and an XY points only graph.

• Enter the log compound concentration in the first column and the corresponding mean mP values for each compound.

• Select analyze and choose the nonlinear regression (curve fit) under XY analyses.

• Choose the log(inhibitor) vs. response – variable slope and constrain bottom values as > 0 and top values as between 0 and 1.

• The analysis determines an “IC₅₀” value (more appropriately termed EC₅₀ for assays involving whole cell lysates) for each compound which corresponds to the concentration of compound for which 50% of the tracer is displaced. Curve fits generated using Hsp90 inhibitors typically show an R² > 0.99.

SUPPORT PROTOCOL 1. PREPARATION OF FUNGAL CELL LYSATES

The FP assays described in Basic Protocols 1 and 2 require the use of whole cell lysates to determine the affinity of inhibitors to the drug target in native complexes.

Caution: Operation of the French press requires training prior to use of the equipment.

Materials:

• Overnight culture medium (e.g. yeast extract peptone, YPD)

• Dettol or bleach

• 1X Dulbecco’s Phosphate Buffered Saline (dilute from 10X PBS, Sigma-Aldrich, cat. no. D1408)

• Binding Buffer (see recipe)

• Glycerol (BioShop, cat. no. GLY002)

• Sterile 50% glycerol in water

• Dry ice

• Ethanol

• Bradford Reagent for 0.1–1.4 mg/mL protein (Sigma-Aldrich, cat. no. B6916)

• Incubator shaker (e.g. Eppendorf Incubator Shaker Series I 26)

• High-performance centrifuge (e.g. Beckman Coulter Avanti J-26 XP)

• 50mL Polypropylene Conical Centrifuge Tubes (e.g. FroggaBio, cat. no. TB50-500)

• French press (e.g. Thermo FRENCH® pressure cell press) with 1/8” nylon balls

• Refrigerated centrifuge (e.g. Eppendorf 5174)

• 0.2 μm syringe filters (VWR, cat. no. CA28143-310)

Cell culture and preparation:

Day before:

• 1Set up a 50 mL culture for overnight growth in a shaker.

Fungal strains are commonly maintained on solid (2% agar) yeast extract peptone (YPD, 1% yeast extract, 2% bactopeptone, 2% glucose) at 4°C for up to three weeks for use in assays and grown overnight in YPD medium at 30°C.

• 2Prewarm 2 x 1 L of sterile medium in shaker.
• 3Pre-chill 100 mL 1X PBS at 4°C.

Protein Extraction:

• 4Dilute overnight culture to OD₆₀₀ of 0.1 in the 2 x 1 L of pre-warmed medium and grow to OD₆₀₀ of 0.8 – 1.0.

A starting OD₆₀₀ of 0.1 is recommended for C. albicans to reach the desired growth state in approximately 4 hours; for the same length subculture of C. neoformans, a starting OD₆₀₀ of 0.3 is optimal.

• 5Pre-chill centrifuge to 4°C half an hour before use.
• 6Disinfect the metal parts of the French Press (piston, cell including tubing, flow valve, and closure plug) by washing with Dettol or bleach then rinse well with water. Pre-chill the metal parts on ice.
• 7Pellet cultures in large centrifuge at 4,000 rpm for 20mins at 4°C.

Keep samples cold for steps 5–12.

• 8Wash with 50 mL ice cold PBS and pellet again.

For culture conditions that promote hyphal growth, this step may need to be repeated if the pellet is not packed tightly.

• 9Resuspend in 10 mL Binding Buffer.
• 10
  Passage lysate through French press twice.

  • Coat the clean chilled piston with glycerol.
  • Place the piston in the cell to the appropriate volume marking.
  • Place onto the tripod upside down (piston at the bottom, see Figure 5).
  • Coat a French press nylon ball with glycerol and place firmly in the flow valve. Screw into the cell loosely.
  • Fill the cell with sample.
  • Coat the closure plug with glycerol and firmly close the cell while holding down the piston to make sure it doesn’t get pushed out.
  • Holding the assembled unit, place it onto the French press piston side up, between pins. Piston arms should be perpendicular to the support rods (see Figure 5).
  • Prepare for collection of the lysate from the line in a conical tube on ice.
  • Tighten the flow valve. It should be tight now, but you should be prepared to loosen or tighten as necessary to maintain pressure ~ 1200 psi with the lysate coming out of the line drop by drop. The flow valve is very sensitive and should be turned in small increments.
  • Raise the platform until the piston begins pressing down on the cell.
  • Continue raising slowly while maintaining pressure ~ 1200 psi.
  • At the end of the first sample collection, open the flow valve then lower the platform.
  • Repeat the above, replacing the ball in the flow valve.
  • Disinfect and wash the French press equipment after use.
• 11Spin the lysate at 14,000 rpm for 30 mins at 4°C.
• 12Collect the supernatant and filter it through a 0.22 μM filter.
• 13Add sterile glycerol to a final concentration of 15%.
• 14Aliquot the lysate.

Figure 5. Identification of key components of the French Press used in the preparation of whole cell lysate.

A. The piston was inserted in the cell to the appropriate volume marking and the cell was placed upside down on the tripod where the sample was loaded. The flow valve was inserted, and the closure plug sealed the assembled unit. B. The assembled cell was then loaded right-side up onto the French Press platform, ensuring that the piston arms were perpendicular to the support rods. Lysate was collected on ice in blue-capped tube to left.

The lysate should only be thawed once, never refrozen. It is preferable to aliquot in small volumes of 100–200 μL which can be easily arrayed and stored in PCR tubes for single use.

• 15Flash freeze in a dry ice/ethanol bath and store at - 80°C.

In the case of Hsp90, lysates can be stored for up to six months at −80°C without loss of binding activity.

• 16Quantify the protein concentration in an aliquot of the lysate by Bradford protein quantification assay.

Lysate dilutions of 1/10 and 1/20 should be suitable for protein quantification with the Bradford reagent for 0.1 – 1.4 mg/mL protein.

• 17Perform Saturation Binding Curves (Basic Protocol 1).

SUPPORT PROTOCOL 2. PREPARATION OF HUMAN LIVER CARCINOMA (HepG2) CELL LYSATE

Materials

• HepG2 cell line (ATCC Cat# HB-8065)

• Dulbecco’s Modified Eagle’s Medium (DMEM, high glucose; Sigma-Aldrich cat. no. D5796-500ML)

• Fetal Bovine Serum (FBS) (Gibco cat. no. 16000044)

• Phosphate Buffered Saline (PBS) (Sigma-Aldrich cat. no. D8537-500ML)

• 0.05% Trypsin-EDTA (Gibco Bioscience cat. no. 25300054)

• Gibco Trypan Blue Solution, 0.4% (Thermo Fisher Scientific cat. No. 15-250-061)

• Binding Buffer (see recipe)

• Dry ice

• Sterile 75% glycerol in water

• T75 tissue culture flasks (Sarstedt cat. no. 83.3911.002)

• 50mL Polypropylene Conical Centrifuge Tubes (e.g. FroggaBio, cat. no. TB50-500)

• Benchtop centrifuge (e.g. Beckman Coulter, Allegra X-14)

• Microcentrifuge tubes 1.7 mL (e.g. FroggaBio, cat. no. LMCT1.7B)

• Refrigerated centrifuge (e.g. Eppendorf 5174)

• PCR tubes (Thermo Fisher Scientific cat. no. 12-222-262)

• Bradford Reagent for 0.1–1.4 mg/mL protein (Sigma-Aldrich, cat. no. B6916)

Propagate HepG2 cells

• 1Grow HepG2 cells in DMEM supplemented with 10% FBS in T75 flasks under 5% CO₂ at 37°C, ensuring that they are healthy and proliferating as expected. Passage cells at a 1:3 split ratio when they reach ~70% confluence.

A large number of cells are required to generate lysate at a concentration suitable for use in subsequent FP assays. For this purpose, multiple flasks (15 to 20), each home to a confluent cell monolayer are collected for lysate preparation.

Prepare cell lysate

Before starting lysate preparation, cool one table-top centrifuge and one microcentrifuge to 4°C and place one bottle of 1X PBS in the fridge or on ice (cold PBS will be required for washes later in the protocol).

• 2Harvest the cells by pouring off the medium and washing the cells with 5 mL 1X PBS at room temperature.
• 3Add 1 mL 0.05% Trypsin-EDTA to each flask and incubate at 37°C for 5 to 10 minutes allowing cells to detach from the surface of the flask.

HepG2 cells are quite adherent to the plastic and can require more time to release from the flask surface.

• 4Collect the trypsin-treated cells in a 50 mL centrifuge tube. Add cold 1X PBS to bring volume up to approximately 50 mL, then spin balanced tubes at 800 rpm for 30 minutes at 4°C.
• 5Pour off the supernatant and wash the cell pellet with 50 mL cold 1X PBS and spin again at 800 rpm for 30 minutes at 4°C.
• 6Pour off the supernatant and resuspend the pellet in approximately 250 to 500 μL of Binding Buffer, prepared fresh.

Ensure all of the supernatant is removed before resuspending in Binding Buffer. The buffer should be prepared during the second spin and kept on ice. Use the least amount of buffer necessary to properly resuspend the pellet as this will allow for a concentrated lysate.

• 7Transfer cell suspension to a microcentrifuge tube and vortex vigorously for approximately 1 minute, then freeze the cell suspension on dry ice. After 5 minutes on dry ice, thaw the cell suspension on ice followed by vigorous vortexing for another minute. This is considered one freeze-thaw cycle to begin the cell lysis. After every freeze-thaw cycle, check the suspension under the microscope to see the extent of cell lysis. Perform at least 3 freeze-thaw cycles to achieve sufficient cell lysis. Cell fragmentation to smaller particles should be observed due to membrane disruption. As further evidence of disruption, trypan blue stain can be used to assess the extent of cell viability monitored as the ability to exclude the dye. Perform at least 3 freeze-thaw cycles to achieve an extent of cell lysis where at least 80% of the preparation is comprised of lysed cells or cells that stain blue.

Depending on the concentration of cells, it may be necessary to perform an additional free-thaw cycle. Cell membrane disintegration should be visible indicating cell lysis.

• 8Spin at 14,000 rpm for 30 minutes at 4°C to collect lysate and pellet cellular debris.
• 9Transfer the supernatant to another microcentrifuge tube taking care so as to avoid disturbing the pellet.
• 10Add sterile 75% glycerol to the lysate to achieve a final concentration of 15% and mix.
• 11Aliquot approximately 20 to 25 μl into PCR tubes for storage.
• 12Flash freeze the cell lysate using liquid nitrogen or dry ice/ethanol bath and store at −80°C.

The lysate should only be thawed once, never refrozen. Storing lysate in individual tubes and in smaller volumes ensures only the desired amount of lysate is thawed for use in assays.

• 13Determine the amount of protein present in the lysate by Bradford protein quantification assay.

REAGENTS AND SOLUTIONS:

Use Milli-Q purified water or equivalent in all recipes.

Binding Buffer A

Prepare for a total of 500 mL in water:

• 20 mM HEPES pH 7.5 (BioShop, cat. no. HEP001)

• 50 mM KCl (Sigma-Aldrich, cat. no. P5405)

• 5 mM MgCl₂ (Bioshop, cat. no. MAG10)

• 0.01% Triton X100 (BioShop, cat. no. TRX777)

Filter sterilize and store at room temperature.

1 M stocks of HEPES (4°C), KCl (room temperature) and MgCl₂ (room temperature) can be prepared ahead and stored at temperatures indicated.

Binding Buffer

Make fresh before use. To 10 mL of Binding Buffer A add:

• 1 tablet Complete™ mini EDTA-free protease inhibitor cocktail (Roche, cat. no. 11836170001)

• 1 tablet PhosSTOP phosphatase inhibitor cocktail (Roche, cat. no. 04 906837001).

• 20 mM sodium molybdate (Sigma-Aldrich, cat. no. 243655)

• 1 mM diothiothreitol (DTT; BioShop, cat. no. DTT001)

FP Assay Buffer

Make fresh before use. To 10 mL of Binding Buffer A add:

• 20 mM sodium molybdate

• 2 mM DTT

(20 μL from 1 M stock of DTT which can be kept at −20°C)

• 0.1 mg/mL bovine gamma globulin (Sigma-Aldrich, cat. no. G5009)

(20 μL from 50 mg/mL stock should be passed through a 0.2 μm filter and can be kept at −20°C)

COMMENTARY

BACKGROUND INFORMATION:

The degree of polarization of a fluorescent molecule is inversely related to its molecular rotation. As a consequence, the observed fluorescence polarization depends on the tracer’s size and the viscosity of the solution in which it is dissolved (Lea & Simeonov, 2011). Instrumentation was developed in the 1950s to begin applying measurements of FP to the study of proteins (Weber, 1956). The measurement of fluorescence polarization (Figure 1A) is calculated from the difference in the emission light intensity parallel (I ∥) and perpendicular (I⊥) to the plane of excitation light, normalized by the total fluorescence emission intensity, as described by the following equation: $\mathit{F}\mathit{P}=\frac{\mathit{I}\parallel -\mathit{I}\perp }{\mathit{I}\parallel +\mathit{I}\perp }$. Although dimensionless, FP measurements are generally reported in millipolarization units (mP). Fluorescence anisotropy is a closely related and interconvertible measurement to polarization in which values are normalized by the total fluorescence intensity from the sample (Hall et al., 2016).

Instruments typically have unequal sensitivity in their detection of emission light in the parallel and perpendicular orientations and thus a correction factor, called the grating factor or G factor, must be applied to compare FP measurements between instruments (Hall et al., 2016). A value of 1 for the G factor is ideal, indicating no bias towards the parallel or perpendicular channels. In reality, the G factor generally ranges from 0.8 to 1.2. This correction factor is dependent on the polarization filters, the instrument’s dichroic mirrors, as well as the assay plates and assay components such as the buffer. A G-factor should be determined for a specific set of filters (depending on the choice of fluorophore) and assay plates. The use of lysates in FP binding and competition assays has many advantages such as enabling the investigation of small molecule binding in more relevant biological mixtures which may include different protein isoforms, as well as various interacting partners or accessory proteins, each of which may diverge between species. Nevertheless, use of lysates means the binding affinity and species selectivity determined are relative measures and assay dependent. In contrast, the use of purified protein allows for an assay-independent assessment of binding affinity and determination of the biochemical K_D for the tracer and for unlabeled competitive ligands, as previously described in detail with FP data converted to anisotropy (Rossi & Taylor, 2011). For an overview of the purification of recombinant proteins see Current Protocols in Protein Science article: Wingfield 2015.

Finally, while the development of an FP assay requires several optimization steps, some reagents are commercially available for rapid assay development. For instance, Transcreener assay reagents from BellBrook Labs are useful tools to investigate inhibitors to proteins with enzymatic activity coupled to nucleotide hydrolysis. The use of Transcreener UDP and AMP reagents have been reported in the identification of novel antifungal molecules (Perfect, Tenor, Miao, & Brennan, 2017), as well as for the discovery of inhibitors to combat other neglected infectious diseases (Pedró-Rosa et al., 2015).

CRITICAL PARAMETERS:

The protocols in this article are optimized for monitoring the binding of inhibitors to abundant cytosolic protein targets. As such, the use of Hsp90 as a target protein provided an ideal example for the success of these FP-based approaches.

As for the design of the tracer, beyond the choice of ligand and fluorophore (see Strategic Planning), additional factors must be taken into account in the selection of a linker and a location for labeling. Available structure-activity relationship (SAR) information for the ligand can be very useful in selecting a well-tolerated site for fluorophore conjugation. Otherwise, an iterative empirical process may be necessary to obtain a tracer in which the fluorescent labeling does not disturb the affinity of binding to the target protein. In selecting linker length, a balance must be achieved between minimizing steric effects on binding by extending linker length with avoiding excessive local rotational mobility, called the propeller effect. This problematic artifact arises when the fluorophore continues to rotate rapidly even with the probe bound to its larger macromolecular target and can be minimized through the use of shorter, more rigid linkers (Lea & Simeonov, 2011).

The efficiency of labeling and tracer purity are also crucial to a robust FP assay, as any free fluorophore can substantially interfere with the increase in FP value upon tracer binding. Finally, once a tracer has been synthesized, it is important to evaluate its intrinsic polarization properties, if any, under anticipated FP assay conditions, as this can impact the G factor. Ideally, equal parallel and perpendicular polarized light detection occurs if there is no intrinsic polarization of the tracer. If some intrinsic polarization is detected, it may be necessary to recalibrate the value of the G-factor to account for the polarization signal from the tracer alone. Some intrinsic polarization is to be expected, and the tracer can still be used as long as the magnitude of the increase in polarization upon binding of the tracer to its target is great enough to provide a robust assay window. If a high level of intrinsic polarization is observed, and it is difficult to detect further increase in polarization, it is best to re-consider the tracer design for the assay.

TROUBLESHOOTING:

Non-specific binding properties are another aspect that should be considered in FP assay development. In particular, the tracer or the test small molecules might bind non-specifically (high capacity, low affinity) to various components used in the assay. For instance, the tracer or small molecule competitors can bind to the surface of the assay plates. Observing a lack of substantial increase in FP value during generation of saturation binding curves may suggest non-specific binding of the tracer to the plate. This can be minimized by using non-binding plates compatible with the plate reader. Non-specific binding to components in the buffer or in the whole cell lysates can also occur. This is generally observed by unusual, non-sigmoidal or non-saturable binding curves. The buffer recipes for the protocols presented in this article include detergent and bovine gamma globulin (BGG), as well as the reducing agent DTT, that were optimized to prevent non-specific binding in the specific example provided. Modifications to the buffer components can be considered to help prevent the formation of aggregates or non-specific interactions while ensuring that any further variations to the assay buffer do not negatively affect the interaction between tracer and target protein nor the resulting polarization signal. Of note, the use bovine serum albumin (BSA) is not recommended because it may non-specifically bind the tracer and/or test compounds. BGG as a carrier protein shows much less non-specific binding to small molecules and is preferred in FP assays. Moreover, filtering lysate samples and buffers is of great importance to prevent signal interference due to aggregates.

The presence of interfering fluorescent or fluorescence-quenching substances (compounds or lysate components) can be monitored by the measurement of total fluorescence intensity in the FP assay. The total intensity in all assay wells should be similar and substantial variation should be investigated further. Poorly soluble compounds may precipitate and cause light scattering artifacts. Beyond ensuring reagent purity and filtering buffers, compound stocks should be centrifuged briefly prior to use. Additionally, intrinsically fluorescent compounds may be identified by taking measurements before and after addition of the probe. Finally, potentially problematic compounds likely to cause interference in the assay can be flagged using open-source cheminformatic software that aids in the identification of pan-assay interference compounds (PAINS)(Baell & Holloway, 2010), such as the FAF-Drugs4 server (Lagorce, Bouslama, Becot, Miteva, & Villoutreix, 2017).

The most common problems and their solutions during the development of FP-based assays to monitor protein-ligand interactions in whole cell lysates are summarized in the Table 2.

Table 2.

Troubleshooting Guide for Fluorescence Polarization-based Measurement of Protein-Ligand Interaction in Cell Lysates

Problem	Possible Causes	Solutions
Lack of increase in FP signal during the generation of saturation binding curves	- The labeled probe lost binding affinity, the tracer has strong intrinsic polarization, is impure, or the ‘propeller effect’ is occurring	- Measure intrinsic polarization, revisit the tracer purification, and/or re-design the tracer
	- The tracer concentration isn’t appropriate	- Try a higher concentration range
	- The tracer is binding to the assay plates	- Try non-binding assay plates
Saturation binding curve doesn’t reach saturation or appears linear	- Insufficient (target) protein concentration in the lysate	- Try a higher lysate concentration range; decrease tracer concentration if possible - Overexpress the protein of interest in the source organism
Saturation binding curve doesn’t reach saturation or appears linear and the unlabeled probe doesn’t compete off the tracer	- Non-specific binding of the tracer to assay components - Presence in lysate of multiple binding partners of varying affinity for the tracer	- Modify buffer components - Generate tracer with greater target specificity

TIME CONSIDERATIONS:

With all reagents on hand, the procedures in the Basic Protocols can be completed in a day of work or less. The Support Protocols require additional time for the establishment and maintenance of cell cultures, but the lysis procedures can be completed in a day of work.