Introduction

The study of protein–protein interactions is vital to understand how proteins function within a cell. A variety of approaches for detecting protein–protein interactions have been developed in the past two decades^{1, 2, 3}. Generally, they are divided into in vivo and in vitro approaches. Commonly used in vivo approaches include two-hybrid systems, using bacteria, yeast or mammalian cells as host^{4, 5, 6}; fluorescence resonance energy transfer (FRET)⁷ and bioluminescence resonance energy transfer (BRET)⁸, using living cells; and the synthetic lethality method in yeast⁹. There are numerous in vitro approaches, such as tandem affinity purification and mass spectrometry¹⁰, surface plasmon resonance (SPR)¹¹, protein microarray¹², dot blotting¹³, co-immunoprecipitation (co-IP) using cultured cells or tissues and pull-down assays using glutathione S-transferase (GST), His or FLAG tags¹⁴. A two-hybrid system is based on the activation of reporter gene(s) once two proteins associate with each other. It is mostly applied to soluble proteins (as is GST pull-down) though systems for membrane proteins do exist¹⁵; however, a major disadvantage with this technique is the presence of false-positives and false-negatives. FRET and BRET use living cells but require the construction of expression vectors and need specific equipment^{7, 8}. In addition, as seen with co-IP, these methods cannot determine whether two proteins directly bind to each other. Both SPR and protein microarray also need specific instruments. Dot blotting is as simple as Far western blotting (WB), but it does not provide the information on the size of binding partner(s).

Far WB was originally developed to screen protein expression libraries with ³²P-labeled GST-fusion proteins^{16, 17}. The technique has now been used to study protein–protein interactions, for example, receptor–ligand interactions, and to screen interacting partners in a library¹⁸. The major advantages of Far WB are that (i) it allows a prey protein to be endogenously expressed in cells; this is particularly meaningful when prey proteins are difficult to purify, and (ii) it enables one to determine whether two proteins directly bind to each other and, if not, whether a third protein mediates the physical complexing between them. The major disadvantages are that at least one protein should be purified to certain amounts and that there may be some nonspecific binding. Far WB is widely used to (i) identify binding partners of a bait protein^{19, 20}; (ii) confirm one-to-one protein–protein interactions resulting from high-throughput screening or other methods^{21, 22}. It can also be used to study the effect of post-translational modifications on protein–protein interactions, examine interacting domain using synthetic peptides as probes and identify protein–protein interactions without using antigen-specific antibodies. Far WB can be performed in a laboratory in which facilities for protein purification and standard WB are available.

Figure 1 shows a flowchart of a Far WB protocol. Instead of detecting a blotted protein in WB based on its molecular weight, Far WB primarily detects the bait protein(s) that is immobilized in the positions where its binding partners (prey) are distributed according to their molecular weights. During the preparation of SDS-PAGE, proteins are typically reduced and denatured by treatment with the Laemmli sample buffer. Because many protein–protein interactions depend on protein secondary and tertiary structures, which are disrupted under reducing and denaturing conditions, it is likely that few, if any, protein complexes could persist following the treatment with an SDS-PAGE sample buffer. In Far WB, proteins are usually denatured with guanidine or urea, and then renatured with gradient-reducing guanidine or urea, which allows proteins to recover their secondary and tertiary structures.

Figure 1: A flowchart for Far western blotting (WB).

Full size image (36 KB)

To illustrate the Far WB protocol, we here utilize our studied proteins in our laboratory, polycystin-2 (PC2) and fibrocystin (FPC), which are mutated in autosomal dominant polycystic kidney disease (PKD) and autosomal recessive PKD, respectively. They do not bind directly to each other but in the presence of a third, intermediate protein, KIF3B, a motor subunit of kinesin-2, they form a complex²².

Materials

Reagents

• Purified (bait) proteins (see EXPERIMENTAL DESIGN for further details of sample preparation)
• Prey protein (in crude extracts or purified, see EXPERIMENTAL DESIGN for further details of sample preparation)
• Acrylamide:bis-acrylamide (29:1) (Bio-Rad, cat. no. 161-0156)
• Guanidine hydrochloride (Sigma, cat. no. G4505-500G)
• N,N,N',N'-Tetramethylethylenediamine (TEMED; Sigma, cat. no. T9281-25ML)
  Caution Harmful if inhaled or swallowed; can cause skin and eye irritation.
• Ammonium persulfate (APS), 10% solution (Sigma, cat. no. A3678-100G)
  Caution Harmful if swallowed.
  Critical Prepare fresh as it decomposes on exposure to water.
• Nitrocellulose or PVDF membrane (Bio-Rad, cat. no. 162-0094; see Experimental design for further details of membrane choice)
• Primary (against bait or prey protein) and secondary (HRP conjugate) antibodies (primary antibodies monoclonal anti-GST and anti-PC2: custom made; rabbit polyclonal anti-GFP: Clontech, cat. no. 632460; secondary antibodies goat anti-mouse and goat anti-rabbit: Chemicon, cat. no. AQ502P and cat. no. AQ132P, respectively.)
• ECL detection kit (Amersham Biosciences, cat. no. RPN-2106)
• Pre-stained protein ladder (Bio-Rad, cat. no.161-0374)
• Tris base (Invitrogen, cat. no. 15504-020-1KG)
• Glycine (Bio-Rad, cat. no. 161-0724-2KG)
• SDS (Bio-Rad, cat. no. 161-0302-1KG)
  Caution Harmful if inhaled or swallowed; can cause skin and eye irritation.
• Skim milk powder (local supermarket)
• Tween-20 (Sigma, cat. no. P7949-100ML)
• DTT (Bio-Rad, cat. no. 161-0611-5G)
• Glycerol (Invitrogen, cat. no. 15514-011)
• -Mercaptoethanol (Sigma, cat. no. M7154-100ML)
  Caution Highly toxic. Wear suitable protective clothing, gloves and eye/face protection. It is harmful if inhaled or swallowed, and toxic when in contact with skin.
• Bromophenol blue (Sigma, cat. no. B0126-25G)
• Methanol (Fisher Scientific, cat. no. A935-4)
  Caution Flammable.
• KH₂PO₄ (Sigma, cat. no. P9791-1KG)
• Na₂HPO₄ (Sigma, cat. no. S3264-1KG)
• Ponceau S (Sigma, cat. no. 81462-1L)
• Separating gel (see REAGENT SETUP)
• Stacking gel (see REAGENT SETUP)
• Protein loading dye (see REAGENT SETUP)
• Tris–Glycine running buffer (see REAGENT SETUP)
• Transfer buffer (see REAGENT SETUP)
• Denaturing and renaturing buffers (see REAGENT SETUP)
• Protein-binding buffer (see REAGENT SETUP)
• PBST buffer (see REAGENT SETUP)
• Re-blot Plus Strong Solution (Chemicon, cat. no. 2504-50ML)

Equipment

• Vertical electrophoresis apparatus (Mini-PROTEAN 3; Bio-Rad, cat. no. 165-3301)
• Power pack (Powerpac 3000; Bio-Rad, cat. no. 164-5070) or equivalent
• Transfer system (Mini Trans-Blot Cell; Bio-Rad, cat. no. 170-3930)
• Petri dish (100 15 13 mm³; BD Biosciences, cat. no. 351112)
• Whatman chromatography paper (Fisher Scientific, cat. no. 05-717-3A)
• X-ray film (Kodak, cat. no. 8377616) or equivalent, developer or phosphor imaging plate and detection system (Fuji BAS-MS; Fuji) or equivalent

Reagent setup

• Separating gel 8% Separating SDS-PAGE gel mixture (29:1) acrylamide:bis-acrylamide, 375 mM Tris–Cl (pH 8.8), 0.1% SDS, 0.1% APS and 0.06% TEMED. The gel mixture should be freshly prepared.
  Critical The choice of the percentage of the separating gel depends on the molecular weight(s) of prey protein(s); generally, the higher the molecular mass, the lower the percentage of the separating gel used.
  Caution Acrylamide and bis-acrylamide are highly toxic and are potential human carcinogens and teratogens. Purchase as stock solutions to avoid exposure to powder, which could easily be dispersed during weighing, for example.
  Caution SDS is a respiratory, skin and eye irritant. It is harmful if swallowed, inhaled or absorbed through the skin.
• Stacking gel 4% Acrylamide:bis-acrylamide (29:1), 125 mM Tris–Cl (pH 6.8), 0.1% SDS, 0.1% APS and 0.1% TEMED. The gel mixture should be freshly prepared.
  Caution Acrylamide and bis-acrylamide are highly toxic and are potential human carcinogens and teratogens. Purchase as stock solutions to avoid exposure to powder, which could easily be dispersed during weighing, for example.
  Caution SDS is a respiratory, skin and eye irritant. It is harmful if swallowed, inhaled or absorbed through the skin.
• Protein loading dye 50 mM Tris–HCl (pH 6.8), 2% SDS, 10% glycerol, 1% -mercaptoethanol, 12.5 mM EDTA and 0.02% bromophenol blue. This loading dye can be stored at room temperature (RT, 20–23 °C) for up to 6 months.
• Tris–Glycine running buffer 25 mM Tris–Cl (pH 8.3), 190 mM Glycine, 1% SDS. This buffer can be stored at RT for up to 3 months.
• Transfer buffer 10% Methanol, 24 mM Tris, 194 mM Glycine. The buffer should be freshly prepared and kept refrigerated before use.
• Denaturing and renaturing buffers: AC buffer 100 mM NaCl, 20 mM Tris (pH 7.6), 0.5 mM EDTA, 10% glycerol, 0.1% Tween-20, 2% skim milk powder and 1 mM DTT (see Table 1). These solutions should be freshly prepared.

  Table 1: AC buffer with varying concentrations of guanidine–HCl used in denaturing and renaturing.

  

  
  Full table
  
  
• Protein-binding buffer 100 mM NaCl, 20 mM Tris (pH 7.6), 0.5 mM EDTA, 10% glycerol, 0.1% Tween-20, 2% skim milk powder and 1 mM DTT. The solution should be freshly prepared.
• PBST buffer 4 mM KH₂PO₄, 16 mM Na₂HPO₄, 115 mM NaCl (pH 7.4) and 0.05% Tween-20. Usually prepare 10 PBST, and store at RT for up to 3 months, and dilute to 1 before use.

Procedure

1. Points from here (point 1) up to and including point 1 are related to Sample preparationSamples from any source can be used in this experiment (See Experimental design for further details of sample preparation). We used lysates of cultured mammalian cells over-expressing a prey protein (the C terminus of FPC with a GFP tag (GFP–FPCC)), and bacterially purified bait proteins (the C termini of KIF3A (KIF3AC) and KIF3B (KIF3BC) proteins with His tags, and those of PC2 and FPC with GST tags; see Fig. 2).

   Figure 2: Expression and purification of polypeptides.

   

   The soluble C termini of KIF3A (KIF3AC, aa 403–702), of KIF3B (KIF3BC, aa 407–747), of polycystin-2 (PC2C, aa 682–968) and of fibrocystin (FPCC, aa 3882–4074) were used in this study. (a) His-tagged KIF3BC and KIF3AC proteins were purified from bacteria and stained by Coomassie brilliant blue (CBB). (b) Escherichia coli extracts expressing glutathione S-transferase (GST) alone, GST-PC2C or GST-FPCC, were subjected to western blotting (WB) with an antibody against GST. (c) GFP and GFP–FPCC were stably expressed in MDCK cells. Shown are WB data using an anti-GFP antibody. BSA acts as a negative control.

   Full size image (20 KB)
   
   
   Critical step The most critical step is to have sufficient amounts of prey proteins (expressed either in crude lysates (10–100 g) or in purified forms (0.2–1 g)) and of the bait proteins in purified forms (1–10 g).Pause Point Cell lysates and purified proteins can be stored at -80 °C for several months.
2. Points from here (point 2) up to and including point 9 are related to SDS-PAGE separation of prey proteinsPour the separating gel (see Reagent setup), using 10 ml per gel, leave enough room for a stacking gel, overlay with water and let it polymerize for 30 min.
3. Remove the aqueous layer from the separating gel and pour the stacking gel (see Reagent setup), 3 ml per gel, on top of the separating gel.
4. Insert a comb, let it polymerize for at least 30 min.
5. Clamp the gel onto the electrophoresis tank.
   Critical step Commercially ready-to-use gel can be alternatively used.
6. Adjust prey protein concentration so that a suitable amount of protein is loaded onto the gel. Usually, 0.2–1 g of purified protein is required, or 10–100 g of cell lysates.
   Critical step Concentrating the samples by SpeedVac may be necessary if the volume is too large for loading.
7. Apply the loading buffer (final 1 , see Reagent setup) to the prey protein samples (maximum volume 50 l), boil the samples for 5 min, and set on ice right after. In our experiments, we used 0.5 g BSA and 50 g cell lysates over-expressing GFP as negative controls or GFP–FPCC as an experimental prey protein (Fig. 2c). Since KIF3AC and KIF3BC associate with each other through their coiled-coil structure, we have used them as a positive control where we used purified KIF 3AC (0.5 g) as prey and separated on gel, and purified KIF3BC (5 g) as bait.
   Critical step Optionally, duplicate or triplicate gels/membranes can be prepared to allow the use of one each for Far WB to detect bait protein (Step 22) and standard WB with an antibody against the prey protein (Step 23), which helps in determining whether bait and prey are on the same spot (rather than using the same membrane for both these procedures). A third membrane may also be useful for staining with Ponceau S (Step 12).
8. Load samples onto the gel (sample volume should not be >50 l for a 1.5 mm 10-well gel) along with positive and negative controls, and protein ladder.
9. Run gel at 80 V for 20 min, then increase to 125 V for a period of time depending on the position of the prey proteins on the gel, which is monitored by a prestained protein ladder. The ideal pattern is that prey proteins stay in the middle of the gel.
10. Points from here (point 10) up to and including point 12 are related to Transfer of proteins onto the membranePrechill the transfer buffer and submerge the membrane in the transfer buffer before use.Troubleshooting
11. Use the traditional sandwich method²⁴ to transfer proteins from gel to membrane at 100 V for 2 h at 4 °C.Troubleshooting
12. Stain the membrane with Ponceau S²⁴ to check whether proteins have been transferred from the gel to the membrane. This step will not affect the following assay.
    Critical step Alternatively, standard WB²⁴ with an anti-prey antibody can be performed to check whether proteins are equally transferred to the membrane. Use equal amounts of BSA (negative control) and cell lysates to run SDS-PAGE, transfer to the nitrocellulose membrane and then detect with WB (Fig. 2c). This must be done on duplicate membranes for standard WB and Far WB. It is best to carry out standard WB before the rest of the Far WB procedure to ensure the proteins have been transferred to the membrane. Alternatively, perform in Step 23, as described.Troubleshooting
13. Points from here (point 13) up to and including point 13 are related to Denaturing/renaturing of proteins on the membraneDenature and renature proteins on the membrane in AC buffer by gradually reducing the guanidine–HCl concentration (see Table 1). Briefly, denature proteins by incubating the membrane in the AC buffer containing 6 M guanidine–HCl for 30 min at RT. Then wash with the AC buffer containing 3 M guanidine–HCl for 30 min at RT. This is followed by washing with the AC buffer containing 0.1 M and no guanidine–HCl AC buffer at 4 °C, for 30 min and 1 h, respectively.
    Critical step Completely renaturing proteins on the membrane is critical for subsequent protein–protein binding assays. In order to keep proteins in proper renaturing states, incubation of the membrane with 0.1 M and no guanidine–HCl AC buffers at 4 °C is strongly recommended. Also, the time of incubation with the guanidine–HCl-free AC buffer should be at least 1 h.Pause Point The last renaturing step with the guanidine–HCl-free AC buffer can last overnight at 4 °C.Troubleshooting
14. Points from here (point 14) up to and including point 14 are related to BlockingBlock the membrane with 5% milk in the PBST buffer for 1 h at RT.
    Critical step After transferring and renaturing proteins on the membrane, use 5% milk to block nonspecific proteins on the membrane, in contrast to 2% used in the previous steps. Determination of an effective blocking buffer should firstly be made empirically. Often, skim milk and BSA are used as a starting point. Insufficient blocking may result in high background, whereas prolonged blocking could result in a weak or masked signal. Protein renaturing also appears to occur during the blocking step, so it is important to optimize the blocking conditions to obtain the best signal:noise ratio for each application and not to deviate from the method.
15. Points from here (point 15) up to and including point 15 are related to Incubation of the membrane with purified interacting (bait) protein(s)Incubate the membrane with purified bait protein(s) (e.g., total 5 g, 1 g ml^-1) in the protein-binding buffer (or an equivalent buffer) overnight at 4 °C. In our experiments, we applied 5 g purified GST-PC2C, His-KIF3AC or His-KIF3BC in a 5 ml reaction buffer.
    Critical step The strength of protein–protein interaction may vary as a function of pH, salt concentration and certain co-factors during the incubation of bait protein(s). It is important that conditions do not change throughout the procedure to maintain the interaction until it is detected. This may affect the formulation of a washing buffer used. To save the proteins used, minimize the binding buffer volume.Pause Point Incubation with purified interacting proteins can be carried out overnight at 4 °C if time is allowed. We have compared different incubation conditions (1 h at 37 °C, 3 h at RT and overnight at 4 °C) and found no significant difference. However, incubation overnight at 4 °C is recommended, as the result is more reproducible under this condition between probing steps.
16. Points from here (point 16) up to and including point 23 are related to Detection of bait proteins bound to prey proteins on the blotWash off unbound bait protein(s) with PBST buffer three times, each for 10 min.
17. Incubate the blot with an appropriate diluted primary antibody for 1 h at RT in 3% milk in the PBST buffer. In our experiments, we used an antibody against PC2 (1A11) at 1:2,000 dilution.Pause Point Incubation with the primary antibody can be carried out overnight at 4 °C, which has been found to be reliable and reproducible in our laboratory.
18. Wash with the PBST buffer three times, each for 10 min.
19. Incubate with a secondary antibody for 1 h at RT. In our experiments, we used an HRP-conjugated anti-mouse IgG antibody at 1:5,000 dilution.
20. Wash with the PBST buffer three times, each for 10 min, and then rinse with PBS buffer for 5 min.
21. Perform chemiluminescent detection of the bound bait protein using an ECL kit, according to the manufacturer's instruction.
    Critical step See Experimental design for details of alternative detection systems.
22. Develop X-ray film immediately to visualize the position of the bait protein.Troubleshooting
23. Strip the probe with the Re-Blot Plus Strong solution for 20 min at RT, and use the same membrane to perform a standard WB²⁴, using an antibody against the prey protein, to determine whether bait and prey are at the same position on the membrane.
24. Points from here (point 24) up to and including point 24 are related to Results analysisCompare the results of Far WB (Step 22) and standard WB (Step 23), and judge whether the bait and prey proteins are located on the same spots. For example, we compared Figures 2c and 3, which show the positions of the prey and/or bait proteins, and concluded that FPCC associates with PC2C in the presence of KIF3BC.

    Figure 3: Far western blotting (WB):KIF3BC bridges the association between PC2C and FPCC.

    

    Lysates of MDCK cells expressing GFP–FPCC or GFP and BSA were separated by SDS-PAGE and transferred onto nitrocellulose membranes. Proteins were denatured, renatured and then incubated with purified GST-PC2C together with (a) none, (b) KIF3AC or (c) KIF3BC. After washing, bound proteins were detected with a PC2 antibody. BSA and GFP vector (lanes 1 and 2) serve as negative controls. The arrow points to an extra band (lane 3, c) not seen in the panels of (a) and (b), corresponding to the binding of PC2C to the site of FPCC (shown in Fig. 2c). GST, glutathione S-transferase.

    Full size image (25 KB)
    
    

Troubleshooting

Troubleshooting advice can be found in Table 2.

Table 2: Troubleshooting table.

Full table

Timing

Timing: Steps 1–8, sample preparation and loading: 30 min to 1 h
Timing: Step 9, electrophoresis: 3 h
Timing: Steps 10–12, electroblotting: 2 h
Timing: Steps 13, denaturing/renaturing: 3 h–overnight
Timing: Step 14, blocking: 1 h
Timing: Step 15, protein binding: 3 h-overnight
Timing: Steps 16–23, detection of proteins: 4 h–overnight
Timing: Step 24, results analysis: 30 min

Anticipated results

A successful Far WB appears to be similar to a standard WB. However, the bait protein(s), as well as a prey protein, is detected on the spots where the prey proteins are located (according to the molecular weight of the prey but not of the bait). In addition, there are often some 'nonspecific' bands on the blot, which probably correspond to some other binding partners of the bait protein(s) that are present in the cell lysates. As shown in Figure 3, we found that our studied protein PC2 associates with fibrocystin through the intermediate protein KIF3B, but not directly or through KIF3A.

Far WB procedures must be performed with care to preserve as much as possible the native conformation of the proteins under study. Denatured proteins may not be able to interact, resulting in a failure to identify an interaction. On the other hand, proteins in non-native conformations may interact with nonspecific proteins, resulting in 'false-positive' interactions. Denaturing may have more impact on prey proteins, because they undergo preparative processing steps for Far WB. This is not to imply that identification of valid interactions is not possible but only to stress the importance of an appropriate validation and the use of controls. As discussed earlier, experimental conditions should be optimized individually.