Selection of Transporter-Targeted Inhibitory Nanobodies by Solid-Supported-Membrane (SSM)-Based Electrophysiology

Summary

Nanobodies are important tools in structural biology and pose a great potential for the development of therapies. However, the selection of nanobodies with inhibitory properties can be challenging. Here we demonstrate the use of solid-supported-membrane (SSM)-based electrophysiology for the classification of inhibitory and non-inhibitory nanobodies targeting electrogenic membrane transporters.

Abstract

Single domain antibodies (nanobodies) have been extensively used in mechanistic and structural studies of proteins and they pose an enormous potential as tools for developing clinical therapies, many of which depend on the inhibition of membrane proteins such as transporters. However, most of the methods used to determine the inhibition of transport activity are difficult to perform in high-throughput routines and depend on labeled substrates availability thereby complicating the screening of large nanobody libraries. Solid-supported membrane (SSM) electrophysiology is a high-throughput method, used for characterizing electrogenic transporters and measuring their transport kinetics and inhibition. Here we show the implementation of SSM-based electrophysiology to select inhibitory and non-inhibitory nanobodies targeting an electrogenic secondary transporter and to calculate nanobodies inhibitory constants. This technique may be especially useful for selecting inhibitory nanobodies targeting transporters for which labeled substrates are not available.

Introduction

Antibodies are composed of two identical heavy chains and two light chains that are responsible for the antigen binding. Camelids have heavy-chain only antibodies that exhibit similar affinity for their cognate antigen compared to conventional antibodies^1,2. The single variable domain (VHH) of heavy-chain only antibodies retain the full antigen-binding potential and has been shown to be very stable^1,2. These isolated VHH molecules or "nanobodies" have been implemented in studies related to membrane proteins biochemistry as tools for stabilizing conformations^3,4, as inhibitors^5,6, as stabilization agents⁷, and as gadgets for structure determination^8,9,10. Nanobodies can be generated by the immunization of camelids for the pre-enrichment of B-cells that encode target-specific nanobodies and subsequent isolation of B cells, followed by cloning of the nanobody library and selection by phage display^11,12,13. An alternate way to generate nanobodies is based on in vitro selection methods that rely on the construction of libraries and selection by phage display, ribosome display, or yeast display^{14,15,16,17,18,19,20}. These in vitro methods require large library sizes but benefit from avoiding animal immunization and favor the selection of nanobodies targeting proteins with relatively low stability.

The small size of nanobodies, their high stability and solubility, strong antigen affinity, low immunogenicity, and relatively easy production, make them strong candidates for the development of therapeutics^21,22,23. In particular, nanobodies inhibiting the activity of multiple membrane proteins are potential assets for clinical applications^5,24,25,26. In the case of membrane transporters, to evaluate whether a nanobody has inhibitory activity, it is necessary to develop an assay that allows the detection of transported substrates and/or co-substrates. Such assays usually involve labeled molecules or the design of substrate-specific detection methods, which may lack a universal application. Furthermore, the identification of inhibitory nanobodies generally requires the screening of large numbers of binders. Thus, a method that can be used in a high-throughput mode and that does not rely on labeled substrates is essential for this selection.

SSM-based electrophysiology is an extremely sensitive, highly time-resolved technique that allows the detection of movement of charges across membranes (e.g., ion binding/transport)^27,28. This technique has been applied to characterize electrogenic transporters, which are difficult to study using other electrophysiology techniques due to the relative low turnover of these proteins^{29,30,31,32,33,34,35}. SSM electrophysiology does not require the use of labeled substrates, it is suitable for high-throughput screening, and either proteoliposomes or membrane vesicles containing the transporter of interest can be used. Here, we demonstrate that SSM-based electrophysiology can be used to classify transporter-targeted nanobodies with inhibitory and non-inhibitory properties. As a proof-of-principle, we describe the reconstitution of a bacterial choline transporter into liposomes, followed by detailed steps for immobilization of proteoliposomes on the SSM sensors. We next describe how to perform SSM-based electrophysiology measurements of choline transport and how to determine the half-maximal effective concentration (EC₅₀). We then show how to use SSM-based electrophysiology to screen multiple nanobodies and to identify inhibitors of choline transport. Finally, we describe how to determine the half maximal inhibitory concentrations (IC₅₀) of selected inhibitory nanobodies.

Protocol

1. Membrane protein reconstitution

1. Mix 3 mL of E. coli polar lipids with 1 mL of phosphatidylcholine in a round bottom flask under a ventilated hood.
2. Dry the lipid mixture for 20 min under vacuum using a rotary evaporator and a water bath at 37 °C to remove chloroform. If needed, dry further under nitrogen or argon gas.
3. Using TS buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl) containing 2 mM β-mercaptoethanol, resuspend lipids to 25 mg/mL.
4. Aliquot lipids in 500 µL aliquots, flash freeze in liquid nitrogen, and store at -80 °C.
5. Thaw one 500 µL aliquot of lipids and dilute 1:1 using TS buffer containing 2 mM β-mercaptoethanol.
6. Extrude the lipid suspension 15 times using a 400 nm membrane.
7. Dilute the lipid suspension to have a final lipid concentration of 4.4 mg/mL.
8. Add n-dodecyl-β-D-maltoside (DDM) to have a final concentration of 0.2% and leave it rotating for 1 h at 200 rpm at room temperature (RT).
9. Add the purified protein to the lipids using a lipid-to-protein ratio between 1:10 and 1:100 (w:w).
   NOTE: The ratio needs to be adjusted depending on the strength of the signal detected in SSM-electrophysiology measurements of substrate transport (see below). To obtain larger signals, use smaller lipid-to-protein ratios.
10. Incubate the mixture rotating at 200 rpm for 1 h at RT.
11. Add 30 mg/mL of polystyrene adsorbent beads, pre-washed in TS buffer.
    NOTE: Add polystyrene beads stepwise.
12. Incubate the beads-lipids mixture for 30 min at RT under slow stirring.
13. To remove the beads, let the beads-lipids mixture stand so that the beads settle down. Transfer the solution to a new tube and leave the beads behind. Add 30 mg/mL of fresh polystyrene adsorbent beads to the separated lipid mixture.
14. Incubate the mixture for 1 h at 4 °C under slow stirring.
15. Separate the beads from the mixture as described in step 1.13 and add 30 mg/mL of fresh polystyrene adsorbent beads.
16. Incubate the mixture for 16 h at 4 °C.
17. Separate the beads from the mixture as described in step 13 and add 30 mg/mL of fresh polystyrene adsorbent beads.
18. Incubate the mixture for 2 h at 4 °C for a fourth and final wash.
19. Centrifuge at 110,000 x g for 30 min at 4 °C.
20. Wash the pellet with 500 µL of TS buffer containing 2 mM β-mercaptoethanol.
21. Centrifuge again at 110,000 x g for 30 min at 4 °C.
22. Resuspend the pellet to a final lipid concentration of 25 mg/mL in TS buffer with 2 mM β-mercaptoethanol.
23. Estimate the protein concentration using an in-gel or amido black assay³⁶.
24. Aliquot the proteoliposomes, flash freeze in liquid nitrogen, and store at -80 °C.

2. Chip preparation

1. Fill a single sensor chip with 50-100 µL of 0.5 mM 1-octadecanethiol solution (resuspended in isopropanol).
2. Incubate the chip with the solution for 30 min at RT.
3. Remove the thiol solution by tapping the chip on a tissue.
4. Rinse the sensor 3 times with 5 mL of pure isopropanol.
5. Rinse the sensor 3 times with 5 mL of double distilled water.
6. Dry the sensor by tapping on a tissue paper.
7. Apply 1.5 µL of 7.5 µg/µL 1,2-diphytanoyl-sn-glycero-3-phosphocholine (lipids dried in a rotatory evaporator and resuspended in n-decane).
8. Immediately after, fill the sensor with 50 µL of non-activating SSM buffer, which does not contain the substrate. This will lead to a spontaneous formation of the SSM layer.
   NOTE: The SSM buffer should be optimized beforehand to determine optimal conditions that have low background noise. A general buffer containing 30 mM HEPES pH 7.4, 5 mM MgCl₂, 140 mM NaCl, can be used as a starting point. The SSM buffer without the substrate is used for washing before and after the measurement (non-activating buffer). To avoid buffer mismatch, use the non-activating buffer to prepare the buffer containing the substrate (activating buffer). The substrate can be added either directly as a powder or in a small volume from a high concentration stock to avoid dilutions.
9. Thaw proteoliposomes from step 1.24 at RT.
10. Dilute proteoliposomes between 1:5 and 1:100 (proteoliposomes:buffer, (v:v)) in the non-activating SSM buffer (here 1:20).
11. Sonicate proteoliposomes for 20-30 s or 3 times for 10 s, placing on ice in between sonication, if necessary. Here a water bath sonicator at 45 kHz was used.
12. Apply 5-10 µL of the diluted sonicated proteoliposomes sample on the surface of the sensor without touching it.
13. Centrifuge the chips with the solution at RT for 30 min using a speed between 2,000 and 3,000 x g.
    NOTE: Use 50 mL tubes with a flat bottom. Carefully place the sensor chips upright using tweezers. 6-well plates and a centrifuge with a plate holder can also be used.
14. Use the sensor chips on the same day.

3. Measuring the solute transportation: determination of saturation conditions

NOTE: As proof-of-principle, these experiments were performed using a bacterial choline transporter reconstituted in liposomes following the protocol described above. The step-by-step process of determining saturating conditions of the substrate choline prior to the measurement of inhibition by nanobodies is shown here.

1. Prepare 1-2 L of the non-activating SSM buffer.
   NOTE: Prepare and use the same SSM buffer stock for all activating and non-activating buffers throughout all measurements.
2. Take 10 clean tubes and transfer 10 mL of the non-activating SSM buffer into each.
3. Add the substrate into the tubes from step 3.2 using a series of concentrations around the expected half maximum concentration (here 15, 10, 5, 1, 0.5, 0.1, 0.05, 0.01, 0.005, 0.001 mM of choline) to prepare the activating SSM buffers. Use a high concentration stock to avoid dilutions.
4. Switch on the SSM machine.
5. Start the SSM software and let the machine initialize automatically. Set the saving path for data and confirm by hitting the OK button. Select the standard Initial Cleaning Protocol in the workflow options and click Run.
6. Mount the proteoliposome coated chip on the socket, move the arm to lock the chip, and close the mounted chip with the cap.
7. Select the program CapCom in the workflow and let it Run to determine the conductivity and capacitance. Confirm that the conductivity is below 5 nS and the capacitance is between 15 and 35 nF before using it for the measurement.
   NOTE: A capacitance value of 15-35 nF and conductance below 5 nS are recommended by the manufacturer when using a 3 mm chip.
8. Transfer the activating solutions into vials and position the buffers in the probe sampler.
9. Transfer the non-activating buffer into a reservoir and position it next to the chip holder at the reservoir position on the right.
10. Create a protocol for the workflow using a sequence of non-activating (B), activating (A), and non-activating (B) solutions (B-A-B sequence) and a loop that performs three measurements and moves to the next activating buffer for all 10 buffers prepared in step 3.3. Use the default flow rate at 200 µL/s using 1 s - 1 s - 1 s flow times for the B-A-B sequence. Click Play to start the measurement.
    NOTE: A typical experiment consists of the sequential flow of non-activating (B), activating (A), and non-activating (B) solutions (written as B-A-B; see Figure 1A). The immobilized proteoliposomes on the sensor will be washed with the solutions. Therefore, the B-A solution exchange generates a substrate concentration gradient, which drives the electrogenic transport reaction.
11. Save the protocol and let the workflow run by clicking on the Play button. Perform the same type of experiment but using protein-free liposomes. This is highly important as it would show the intensity of background currents. This should be considered when analyzing data of electrogenic transport measured with proteoliposomes (Figure 1C).
12. Use any preferred software for data analysis to plot the measured current versus time. Read out the peak current manually, or if using the software, use the function for peak height estimation in the range of the addition of the activating buffer.
13. Plot the peak current against the substrate concentration to determine the EC₅₀ of the substrate via the nonlinear regression (Figure 1B,C). Read out the lowest concentration at which the peak current reaches a maximum value, this concentration corresponds to saturating conditions.
    ​NOTE: It is important to consider that the number of proteoliposomes that remain immobilized varies from chip-to-chip. This variation is evident as the peak currents at identical measurement conditions will show different amplitudes. Therefore, it is necessary to normalize the current amplitudes of measurements performed on each chip separately before comparing measurements among different chip preparations.

4. Serial classification of inhibitory and non-inhibitory nanobodies

NOTE: This section shows how to measure choline transport in the presence of nanobodies that bind specifically to the bacterial choline transporter. Smaller peak currents in the presence of nanobodies indicate transport inhibition. Non-inhibitory nanobodies will not impact substrate transport, i.e., no decrease of the peak current signal.

1. Prepare 1-2 L of non-activating SSM buffer.
2. Transfer 50 mL of the non-activating SSM buffer into a clean tube. Add the substrate choline to a final concentration of 5 mM (saturating conditions). Use this for a positive control measurement.
3. Transfer 10 mL of the non-activating SSM buffer into a clean tube. Add the substrate choline to a final concentration of 5 mM (saturating conditions) and add nanobody to a final concentration of 500 nM.
4. Repeat step 4.3 for each nanobody to prepare the activating solutions.
   NOTE: If the purified nanobodies are resuspended in a different buffer than the SSM buffer, their addition to the activating and non-activating buffers will lead to a buffer mismatch. A buffer mismatch should be avoided since it can lead to high noise. Exchanging the buffer of the purified nanobodies with the SSM buffer can help to avoid this issue. Furthermore, using nanobody concentrations that allow to reach saturating conditions is recommended. Considering that the binding constants of nanobodies are generally below 100 nM, a nanobody concentration of 500 nM is recommended for this experiment. However, it is important to pre-screen for optimal concentrations.
5. Start the SSM machine and measure the capacitance and conductivity of the proteoliposome coated chip as described in steps 3.4-3.7.
6. Transfer the activating solution without a nanobody into a vial and place the buffer in the probe sampler. Transfer the non-activating buffer without a nanobody into a reservoir and position it in the probe sampler.
7. Transfer the activating solutions containing nanobodies into vials and position the buffers in the probe sampler. Transfer the non-activating buffers containing nanobodies into vials and position the buffers in the probe sampler.
8. Create a protocol for the workflow using a sequence of non-activating (B), activating (A), and non-activating (B) solutions (B-A-B sequence).
9. Create a loop that performs the following: three measurements of the B-A-B sequence using buffers without nanobody, two measurements of the B-A-B sequence with buffers containing a nanobody, 120 s delay time for incubation with the nanobody, then 3 measurements of the B-A-B sequence with buffers containing the nanobody.
10. Save the workflow and let it run by clicking the Play button.
    NOTE: This workflow will measure the initial conditions of the transport without inhibition by running the B-A-B protocol 3 times using non-activating (B) and activating buffers (A) without the nanobody (Figure 1B,C), followed by the measurement of the nanobody effect on the transport by running the B-A-B protocol 5 times with the non-activating and activating buffers containing nanobodies (Figure 2A). The second order binding kinetics dictate the interaction of nanobodies and their target proteins. Therefore, it is important to use a time delay in the B-A step in order to give enough time for nanobodies to bind to transporters in proteoliposomes on the chip. Optimal times depend on the nanobody concentration i.e., at lower concentrations longer times are required. The first two measurements are required to adapt the system to the new conditions and the second run should be performed after a delay time of 120 s. Only the following three measurements should be used for data analysis.
11. Create a new protocol for the workflow using a sequence of non-activating (B), activating (A), and non-activating (B) solutions (B-A-B sequence) and a loop of 5 measurements to wash out the reversibly bound nanobody.
    NOTE: Optionally, include an incubation step to allow dissociation of nanobodies with slow kinetics.
12. Save the workflow and let it run by clicking the Play button.
13. Compare the last peak current of the measurements with the initial substrate-only measurement in step 4.10. The nanobody has been successfully washed out and the initial conditions have been reestablished if the peak current reaches the initial value, otherwise repeat steps 4.12-4.13 or change to a new chip.
14. Repeat steps 4.6-4.13 and use individual chips for each nanobody screen (Figure 2C) or repeat with multiple nanobodies using the same chip (Figure 2D).
15. Use any preferred software for data analysis to plot the measured current versus time. Read out the peak current manually, or if available, in the used software, automatically select the function for peak height estimation in the range of the addition of the activation buffer.
16. Normalize the peak current in presence of the nanobody, based on the preceding substrate-only measurement. Plot the peak currents in a histogram and compare the peak currents of the substrate only measurements to the peak currents measured in the presence of nanobodies (Figure 2C,D) to identify inhibitory nanobodies.
    NOTE: Normalization of the determined peak currents of each individual run with a nanobody should be performed considering the peak current in the absence of the nanobody from the preceding measurement. Also, since the number of proteoliposomes that remain immobilized vary from chip-to-chip, it is important to normalize the current amplitudes of measurements performed on each chip separately before comparing measurements among different chip preparations.

5. IC₅₀ measurement with inhibitory nanobodies

NOTE: After identifying inhibitory nanobodies, it is possible to determine their half maximal inhibitory concentration (IC₅₀). This is done by measuring the transport of choline at constant concentration, while varying concentrations of the inhibitory nanobody.

1. Prepare 1-2 L of the non-activating SSM buffer.
2. Transfer 50 mL of the non-activating SSM buffer into a clean tube. Add the substrate choline to a final concentration of 5 mM (saturating conditions). Use this as the activating solution for positive control.
3. Take 8 clean tubes and add 5 mL of the non-activating solution into each. Add the substrate choline to a final concentration of 5 mM (saturating conditions). Add the inhibitory nanobody to the tubes at concentrations in the expected IC₅₀ range (here 500 nM - 1 nM).
4. Take 8 clean tubes and add 10 mL of the non-activating solution into each. Add inhibitory nanobody to each tube individually at the same concentration as in step 5.3. This corresponds to the non-activating buffer.
   NOTE: This will generate a series of activating and non-activating buffer pairs at different concentrations of the same inhibitory nanobody.
5. Start the SSM setup and measure the capacitance and conductivity of the proteoliposome coated chip as described in steps 3.4-3.7.
6. Transfer the activating solution without nanobody into a vial and place it in the probe sampler. Transfer the non-activating buffer without nanobody into a reservoir and position it at the reservoir position next to the chip holder on the right.
7. Transfer the activating solutions containing nanobodies into vials and position the buffers in the probe sampler. Transfer the non-activating buffers containing nanobodies into vials and position the buffers in the probe sampler.
8. Create a protocol for the workflow using a sequence of non-activating (B), activating (A), and non-activating (B) solutions (B-A-B sequence). Include a loop to measure each concentration 2 times, incubate for 120 s and measure 3 more times. The workflow will start with the positive control of substrate-only followed by the lowest concentration of the nanobody. Each nanobody measurement will be followed by a subsequent measurement of the positive control with substrate-only to restore the initial peak amplitude, before moving to the next higher nanobody concentration.
   NOTE: The second order kinetics dictate the binding of nanobodies. Therefore, it is important to use a time delay in the B-A step. Optimal times depend on the nanobody concentration i.e., at lower concentrations longer times are required; here 120 s was used with satisfactory results.
9. Use any preferred software for data analysis to plot the measured current versus time. Read out the peak current manually, or if using software, select the function for the peak height estimation in the range of the addition of the activation buffer (Figure 3A). Plot the peak currents against the nanobody concentration to determine the IC₅₀ via non-linear regression (Figure 3B).
   ​NOTE: Normalize the current amplitudes of measurements performed for each individual chip before comparing measurements among different chip preparations.

6. Cleaning of sensors

1. Rinse the single sensor chips after use with 10 mL of distilled water.
2. Dry the chip by tapping it on a tissue paper.
3. Fill the sensor cavity of the chip with 100 µL of pure isopropanol and incubate for 10 min at RT.
4. Place a cotton swab in pure isopropanol and incubate for 1-3 min.
5. Use the presoaked cotton swabs and gently rotate on the sensor's surface without pressure to remove residues.
6. Rinse the sensor with 5 mL of pure isopropanol.
7. Rinse the sensor with 10 mL of distilled water.
8. Dry the sensor by tapping the chip on a tissue.
9. Let the sensor dry overnight at RT, and store at RT under dry conditions.
   NOTE: Sensors can be re-used up to 4-5 times when cleaned and stored properly.

Results

SSM-based electrophysiology has been extensively used for the characterization of electrogenic transporters. In the protocol presented here, we show how to use SSM-based electrophysiology to classify nanobodies targeting a secondary transporter (here a bacterial choline symporter) based on their inhibitory and non-inhibitory properties. One of the most useful features of this technique is that it allows for the high-throughput screening of multiple buffer conditions. This particular characteristic is beneficial for the a...

Discussion

The technique presented here classifies nanobodies with inhibitory and non-inhibitory properties targeting electrogenic transporters. Assessing the substrate transport is possible due to the detection of the movement of charges through the transporter embedded in the membrane of proteoliposomes. Some of the critical steps during the setup of an experiment are reconstitution of active protein in liposomes, preparation of stable monolayers on SSM chips, and recovering of initial conditions after the application of the wash...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1-octadecanethiol solution	Sigma Aldrich	O1858-25ML
1,2-diphytanoyl-sn-glycero-3-phosphocholine	Avanti Polar Lipids	850356C-25mg
Bio-Beads SM-2 Adsorbent (Polystyrene adsorbent beads)	BioRad	#152-3920
PD 10 Desalting Columns	GE Healthcare	GE17-0851-01
Filter 200 nm membrane	Whatman Nucleopore	WHA800282
2-Propanol	Merck	33539-1L-R
n-Decane	Sigma Aldrich	8034051000
n-dodecyl-ß-D-maltoside (DDM)	Avanti Polar Lipids	850520P-25g
Sodium Chloride	AppliChem	131659.1211
(SSM setup) SURFE2R N1	Nanion	-----
SURFE2R N1 Single Sensor Chips	Nanion	# 161001
Trizma Base	Sigma Aldrich	T1503
E. coli Polar Lipid Extract	Avanti Polar Lipids	100600C
Egg PC L-α-phosphatidylcholine	Avanti Polar Lipids	840051C