Identification of Small Molecule-binding Proteins in a Native Cellular Environment by Live-cell Photoaffinity Labeling

Summary

We describe here a method for identification of small molecule-binding proteins using photoaffinity labeling. The advantage of this technique is that binding and covalent labeling of the target proteins occurs within the live cellular environment, removing the risk of disrupting native protein structure and binding conditions upon cell lysis.

Abstract

Identifying the molecular target(s) of small molecules is a challenging but necessary step towards understanding their mechanism of action. While several target identification methods have been developed and used to successfully elucidate the binding proteins of a variety of small molecules, these techniques have drawbacks that make them unsuitable for detecting certain types of small molecule-target interactions. In particular, non-covalent interactions that depend on native cellular conditions, such as those of membrane proteins whose structures may be perturbed upon cell lysis, are often not amenable to affinity-based target identification methods. Here, we demonstrate a method wherein a probe containing a photolabile group is used to covalently crosslink to the small molecule binding protein within the environment of the live cell, allowing the detection and isolation of the target protein without the need for maintenance of the interaction after cell lysis. This technique is a valuable tool for studying biologically interesting small molecules with unknown mechanisms, both in the context of basic biology as well as drug discovery.

Introduction

Bioactive small molecules fundamentally work by interacting with and altering the function of one or more "target" molecules, most commonly proteins, in the cell. In drug discovery, when an active compound is discovered through phenotypic screening, identification of the molecular target(s) of that compound is crucial, not just for understanding the mechanism of action and potential side-effects of the compound, but also for potentially discovering new biology underlying the disease model and paving the way for development of new mechanistic classes of therapeutics¹. Although target identification is not required for a drug to be used therapeutically, in recent years there has been an increasing recognition that novel drug candidates are more likely to succeed in clinical trials, and therefore yield better returns on investment, if a validated target is known². Thus, there has been a growing interest in methods for identifying small molecule target proteins.

A classical target identification experiment typically relies on affinity purification, where the small molecule of interest is immobilized onto a resin and incubated with whole cell lysates, after which unbound proteins are washed away and the remaining proteins are eluted and identified³. While this technique has been used to identify the targets of many small molecules⁴, it is unsuitable as a universal target ID method for several reasons. First, the target protein must retain its native conformation upon cell lysis in order to retain its ability to bind to the small molecule. This can be particularly problematic for membrane proteins, which often undergo conformational changes after being removed from their native environment, or simply aggregate and precipitate out of solution. Second, the small molecule must be chemically modified in such a way that it can be immobilized onto the resin while maintaining its ability to bind the target protein. Deep binding pockets may therefore become inaccessible to a small molecule once it is fixed to the resin. Third, the binding affinity must be sufficiently high that the interaction is maintained during the washing steps, making identification of lower affinity interactions challenging. Fourth, environmental conditions such as pH, ion concentration, or the presence of other endogenous molecules can vary spatially within the cell and are sometimes prerequisites for drug-target interactions. Thus, finding the correct conditions to allow and maintain binding outside of the cell can require a significant amount of trial and error.

Photoaffinity labeling circumvents these issues by allowing the covalent binding of a small molecule and its target within the native context of a cell. Rather than immobilizing the small molecule to a large bulky resin, the molecule is instead chemically modified to install two small functional groups: a photoactivatable moiety which allows covalent crosslinking to the target protein when irradiated with a particular wavelength of light, and a reporter group that allows the target protein to be detected and subsequently isolated. Live cells are treated with the photoaffinity probe, the probe binds and covalently crosslinks to the target protein, and the probe-protein complex is then isolated intact. The specificity of probe binding to the target is demonstrated by performing a competition experiment in parallel, where an excess of the parent compound is used to compete away binding of the probe to the target protein.

The design and synthesis of photoaffinity probes varies greatly from one small molecule to another, and will not be covered in this protocol; however, several excellent discussions on the subject have been published^5-9. The main consideration is that the probe retains the bioactivity of the parent compound, therefore presumably binding to the same target(s). Structure-activity relationship (SAR) studies must be performed to determine which parts of the molecule can be modified without loss of bioactivity. A variety of different chemical groups have been used as photoactivatable crosslinkers, including diazirine, benzophenone, and aryl azide, which each have advantages and disadvantages¹⁰. Likewise, there are multiple reporter tags that have been used to isolate probe-binding proteins. Reporter groups may be functional on their own, such as the commonly used biotin or fluorescent tags, or may be precursors that require further functionalization subsequent to the photocrosslinking step, which have the advantage of being smaller and thus less likely to compromise bioactivity¹¹.

In this protocol, we have used a photoaffinity probe containing a diazirine photocrosslinking group, and a terminal alkyne for the attachment of a reporter group through a Cu(I)-catalyzed Azide-Alkyne Sharpless-Hüisgen cycloaddition (or click) reaction^12-15. The SAR studies, probe design and synthesis, and results of these studies have been published elsewhere^16-18.

Protocol

NOTE: This protocol was adapted from MacKinnon and Taunton¹⁰ for use in live cells.

1. Preparation of Cultured Cells

1. Prepare sterile 6-cm cell culture dishes for the number of samples desired (see below).
   NOTE: One dish of cells is used per treatment condition, but if more protein is required 2 or 3 dishes can be prepared per condition and combined after UV irradiation.
   1. Prepare at least 3 dishes of cells; Negative control with DMSO only (D), Probe treatment only (P), and probe + competitor (C).
   2. Optionally, prepare a 4^th dish as a no UV control, to be treated with probe but not subjected to UV light.
      NOTE: If multiple competitor molecules are to be tested, such as different analogs of the small molecule of interest, prepare additional competitor plates as necessary.
2. Add HEK293T cells to culture dishes in 4 ml culture media. Use 3.5 million cells per plate.
   NOTE: The number of cells should be determined for each cell type so that cells are almost 100% confluent at the beginning of the experiment, in order to achieve the maximum protein yield. For HUVEC, use 0.3 million cells per plate. A good approximation is 1/10 of the cells in a confluent 15 cm dish. HEK293T cells should be cultured in Dulbecco's Modified Eagles Medium supplemented with 10% Fetal Bovine Serum (FBS) and 1% Penicillin/Streptomycin.
3. Return the cells to the culture incubator (37 °C, 5% CO₂) overnight.

2. Treatment of Cells and Photocrosslinking

1. The next day, pre-aliquot the following drugs into 1.5 ml microcentrifuge tubes:
   1. For the first treatment, add 4 µl of DMSO to 2 tubes and 4 µl of 10 mM competitor (i.e., the unmodified parent compound) to 1 tube.
   2. For the second treatment, add 16 µl of DMSO to 1 tube and 16 µl of 50 µM photoaffinity probe to 2 tubes.
      NOTE: The optimal concentration may be different for different probes (see discussion), but between 200-500 nM is a good starting point (here we use 200 nM). The concentration of competitor should exceed that of the probe at least 20 fold (here we use 10 µM). The final concentration of DMSO should be equal in all plates and should not exceed 0.5% (20 µl volume).
2. Bring the cell plates into the cell culture hood and add the pre-aliquoted drugs from step 2.1.1 as follows: Aspirate 1 ml of culture media, resuspending the pre-aliquoted drug in the media, and gently adding it back to the plate drop wise. Add the DMSO to the DMSO (D) plate and the probe (P) plate, and add the competitor to the competition (C) plate. Return the plates to the culture incubator for 30 min.
3. With lights in the culture hood dimmed, add the pre-aliquoted probe from step 2.1.2 to probe (P) and competition (C) plates and add the DMSO to the DMSO (D) plate, as above. Return the plates to the culture incubator for 1 hr.
4. During the incubation, prepare the following items; A tray of ice that can fit all of the plates; ice-cold lysis buffer (PBS [pH 8.5] + 1x protease inhibitor cocktail [complete EDTA-free, diluted from 50x stock solution made in water]; 200 µl/plate); microcentrifuge tubes labeled D, P, or C for each of the different treatment conditions, on ice.
5. 15 min before the end of the 1 hr incubation, set up the UV lamp (365 nm) in the cold room and turn it on to warm up the bulb.
6. After 1 hr incubation with the probe, place dishes on ice. Wash the cells gently with 5 ml ice-cold PBS (pH 7.4) to remove excess probe. Re-cover the cells with 4 ml ice-cold PBS.
7. Place the dish of cells centered 3 cm under the UV lamp on top of an ice pack to minimize heating from the lamp. Irradiate for 3 min. Remove the dish to the tray of ice and repeat for all samples.
8. After irradiation, aspirate the PBS from the cells and add 200 µl of the ice-cold PBS (pH 8.5) with protease inhibitors to each plate. Detach the cells from the plate using a rubber scraper and transfer to the pre-labeled microcentrifuge tubes on ice.
9. Add SDS to a final concentration of 0.4% (10 µl of 10% SDS into 250 µl of sample).
   CAUTION: SDS powder is hazardous. Avoid inhaling SDS powder by wearing a mask over nose and mouth.
10. Lyse the cells by sonicating the suspension for 10 pulses (output 1, duty cycle 30%) and incubate on ice for 1 min before a second round of 10 pulses.
11. If needed, remove 2 µl of suspension and check under a light microscope to ensure complete cell lysis.
12. Boil the samples on a hot plate set to 95 °C for 5 min to complete cell lysis and denature all the proteins.
13. Measure the protein concentration in each sample using a spectrophotometric detergent-compatible protein assay, such as the Dc protein assay¹⁹, according to the manufacturer's instructions, with a spectrophotometer set to measure absorbance at 750 nm.
14. Normalize the protein concentration to 2.5 mg/ml (or if concentrations are lower than 2.5 mg/ml, normalize to the sample with the lowest protein concentration) by adding PBS pH 8.5 + 0.4% SDS as needed (for example, if samples are 5 mg/ml in 250 µl, add 250 µl of PBS pH 8.5 + 0.4% SDS to obtain a final concentration of 2.5 mg/ml).

3. Attachment of Fluorescent Tag by Click Chemistry for Visualization of Labeled Proteins

1. Remove 40 µl of the cell lysate and transfer to a new microcentrifuge tube. Keep the remaining lysates for reaction with biotin-azide (step 4).
2. Add the following reagents in order: 0.2 µl fluor-azide (1 mM stock solution in DMSO), 0.58 µl TCEP (100 mM stock in water with 4 equivalents NaOH added, prepared fresh), and 3.38 µl TBTA (1.7 mM stock in a 4:1 ratio of t-butanol to DMSO). Vortex to mix.
3. Add 1.14 µl CuSO₄-5H₂O (50 mM in water) to start the reaction. Vortex briefly and incubate at room temperature for 30 min in the dark.
4. Add 50 µl of 2x SDS sample buffer to quench the reaction. At this point, run the samples directly on an SDS-PAGE gel²⁰ for best results, or store at -20 °C overnight if necessary.
5. If freezing samples, reheat for 5 min at 95 °C before loading onto the gel.
   NOTE: Since the molecular weight of the target protein is typically not known at this stage, it is advisable to run 2 gels with different acrylamide concentrations (i.e., 8% and 15%), or a wide molecular weight range gradient gel (i.e., 4-15%), to ensure complete molecular weight coverage.
6. When the dye front has reached the end of the gel, continue running the gel for an additional 5 min to ensure all of the excess unreacted fluor-azide has completely exited the gel.
7. Remove the gel to a container with ddH₂O and incubate for 10 min with gentle agitation, to wash away all excess fluor-azide.
8. Place the gel onto a glass plate and scan the gel using a Typhoon fluorescent gel scanner according to the manufacturer's instructions.

4. Attachment of Biotin Tag by Click Chemistry for Affinity Purification of Labeled Proteins

1. Of the remaining lysates from step 3.1, use the maximum amount after protein normalization such that all samples are the same volume (for example, if after normalization to 2.5 mg/ml the resulting sample volumes are 500, 550, and 600 µl, use 500 µl of each).
2. Pre-clear the lysates by adding to 50 µl high-capacity streptavidin agarose beads, pre-washed 2x with PBS (pH 7.4). Incubate 1 hr at 4 °C with rotation.
3. Pellet the beads by centrifugation at 1,000 x g for 3 min. Remove the supernatant to a new microcentrifuge tube on ice and discard the beads.
4. Remove 1-2% of the DMSO sample to a new tube labeled "input". Add an equivalent volume of 2x SDS sample buffer and store at -20 °C.
5. Per 500 µl of lysate, add the following reagents: 1.38 µl biotin-azide (10 mM stock in DMSO), 5.5 µl TCEP, and 32.5 µl TBTA. Vortex to mix.
6. Add 11 µl CuSO₄-5H₂O per 500 µl of lysate and vortex briefly. Incubate at room temperature for 30 min.
7. Add 4 sample volumes of acetone chilled to −20 °C. Vortex the samples and incubate overnight at −80 °C to completely precipitate the proteins and remove unreacted biotin-azide.
8. Centrifuge the samples at 17,000 x g for 15 min at 4 °C to pellet the precipitated proteins.
9. Aspirate the supernatant completely, and resolubilize the proteins by sonication in 150 µl PBS (pH 7.4) + 1% SDS.
10. Add 600 µl of PBS (pH 7.4) to dilute the concentration of SDS to 0.2%.
11. Add the samples to 30 µl pre-washed high-capacity streptavidin agarose beads and incubate 1 hr at 4 °C with rotation.
12. Pellet the beads by centrifugation at 1,000 x g for 3 min. Aspirate the supernatant containing unbound proteins and discard.
13. Add 1 ml of wash buffer (400 mM NaCl, 50 mM Tris, 0.2% SDS, pH 7.4) to the beads and incubate 5 min at room temperature with rotation.
14. Repeat steps 4.12 and 4.13 3 times.
15. Aspirate the wash buffer completely from the beads, and add 30 µl of 2x SDS sample buffer.
16. Incubate for 5 min on a 95 °C heat block to release the proteins from the beads.
17. Centrifuge the beads for 1 min at 13,000 x g at room temperature.
    NOTE: At this point, the samples can be stored at -20 °C until ready to run SDS-PAGE. If freezing samples, reheat for 5 min at 95 °C before use.
18. Carefully pipette the sample buffer containing proteins off of the beads, and load onto SDS-PAGE²⁰ (see important considerations below). The beads can be saved for reboiling later if necessary.
    NOTE: For protein detection by silver stain, better results are obtained using a 1 mm thick gel. If a band is to be cut out of the silver-stained gel for mass spectrometry (MS) analysis, prepare all reagents fresh from sterile solutions and leave an empty well in between samples on the gel. For every band cut from the probe lane, a parallel slice should be taken from the DMSO lane so that background proteins can be subtracted. For validation of protein ID by western blot, it is critical to also load the input sample from step 4.4 when running the SDS-PAGE gel.
19. Follow a standard protocol for either silver staining²¹ or Western blot²² to detect the target protein(s).

Results

The results shown here were obtained with a photo-affinity probe of the antifungal drug itraconazole, the use of which has been previously published¹⁶. These results demonstrate the use of the live-cell photoaffinity labeling technique to successfully identify a major itraconazole-binding protein as the 35 kDa membrane protein Voltage-Dependent Anion Channel 1 (VDAC1).

The above protocol was performed in HEK293T cells...

Discussion

Different approaches to identifying the targets of small molecules can be broadly grouped into two categories: top-down, where the cellular phenotype of the drug is used to narrow down its potential targets based on their known functions, or bottom-up, where the target is identified directly by chemical or genetic means³. Top-down or phenotypic studies can identify certain cellular processes affected by the drug (e.g., transcription/translation/DNA synthesis, cell cycle block, signaling pathway activa...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Tris(2-carboxyethyl)phosphine (TCEP)	Life Technologies	20490	Make fresh day of use. Prepare 100 mM stock in water with 4 eq NaOH.
Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl] amine (TBTA)	AnaSpec	63360-50	Prepare 1.7 mM stock in a 4:1 ratio of t-butanol to DMSO, store at -20 °C.
Copper Sulfate (CuSO4•5H2O)	LabChem, Inc.	LC13440-1	Prepare 50 mM stock in water, store at room temperature.
Biotin-azide	Click Chemistry Tools	AZ104-100	Prepare 10 mM stock in DMSO, store at -20 °C.
Alexa Fluor 647-azide	Life Technologies	A10277	Prepare 1 mM stock in DMSO, store at -20 °C.
365 nm UV lamp	Spectroline	FC100	UV-blocking glasses should be worn while operating.
Protease inhibitor tablets, EDTA-free	Roche Life Science	11873580001	Prepare 50x solution in water and store at -20 °C.
Sonicator	Branson	Sonifier 250	Set to output 1, duty 30%.
Fluorescent gel scanner	GE Healthcare Life Sciences	FLA 9500	Use red laser to detect Alexa-fluor 647.
Detergent-compatible Dc protein assay kit	Bio-Rad	5000112
High Capacity Streptavidin Agarose beads	Life Technologies	20359
Dulbecco's Modified Eagles Medium, low glucose	ThermoFisher Scientific	11885092
Fetal Bovine Serum, qualified	ThermoFisher Scientific	26140079
Penicillin/Streptomycin solution	ThermoFisher Scientific	15140122
SDS Sample Buffer (2x)	ThermoFisher Scientific	LC2676