Immuno-affinity Purification of Mammalian Protein Complexes

Yoshihiro Nakatani1 and Vasily Ogryzko2

1

Dana-Farber Cancer Institute and Harvard Medical School, Boston, Massachusetts

02115, USA

2

Laboratoire Oncogenese, Differenciation et Transduction du Signal, CNRS UPR

9079, Institut Andre Lwoff, Villejuif, France

Introduction

Identification of the interaction partners of a particular protein became a valuable way to gain insight into the protein’s physiological role. Often, when studying a gene of unknown molecular function, the knowledge of the molecular complexes that include the encoded protein as a subunit can be a starting point for understanding the function of the protein. Many proteins appear to exist and function as stable multimeric protein complexes. For instance, the multimeric RNA polymerase II protein complex consists of 12 polypeptides [1]. It transcribes DNA templates as a complex, while individual subunits have no such activity, thus making it impossible to predict “bona-fide” functions of an RNA polymerase II from a single subunit. Likewise, if the protein of interest functions as a complex, it is crucial to purify it as a complex to determine what it does in the cell. The most widely used system to identify inter-molecular interactive partners of a particular protein of interest is the yeast two-hybrid system [2!, 3]. However, given that this system is designed to detect mostly binary interactions, it is not suitable for investigating the intra-molecular interaction partners of a protein if the latter is a component of a multimeric protein complex. The other approach is affinity purification of interactive partners using the protein of interest as the ligand. Generally, the protein of interest is immobilized on matrix and incubated with an extract in vitro, allowing one to purify inter-molecular interaction partners [3]. An advantage of this method over the yeast two-hybrid system is that, if the purification is performed under native conditions, inter-molecular interaction partners can be purified as native forms. However, if the protein of interest exists in a stable multimeric protein complex,

identification of intra-molecular interaction partners of the immobilized protein would not be possible because the liganded protein would not be able to gain access and thus efficiently replace the corresponding endogenous protein in the complex. For identification of intra-molecular interaction partners, purification of the native protein complexes from the cells would be the best method. This would provide information on how the endogenous protein interacts with intra-molecular interaction partners. However, in general, purification of native complexes is technically difficult and time-consuming, especially when the protein of interest is scarce. These limitations can be addressed with epitope tagging [4, 5!, 6-9]. It has been shown that stably expressed exogenous proteins with epitope tags are integrated into the complexes when the complex is synthesized de novo. Such a protein complex can be rapidly purified by immunoprecipitation with antibodies against epitopes. We took advantage of the epitope-tagging technique to develop a standardized approach to purifying multimeric protein complexes from mammalian cells. We put two different epitope tags tandemly (i.e., FLAG and HA) on either the Nor C-terminus of the protein of interest. For rapid generation of cell lines that express the epitope-tagged protein of interest, we use retroviral transduction. The cells are magnetically sorted after transduction. Once stably transduced cell lines are established, the protein complex can be purified from cells in less than 24 hr. Here we describe both the generation of recombinant cells (molecular cloning and retroviral transduction) and the technique for purification of the complex (large-scale cell growth and affinity chromatography). Thus, the paper is divided into two parts that correspond to these different steps in the procedure.

Generation of epitope-tagged protein–expressing cells

To generate stable cells that express double-epitope–tagged versions of the proteins of interest, we constructed two retroviral vectors: pOZ-FH-N and pOZ-FH-C (Fig 1). These vectors are derivatives of the MMLV-based pOZ vector, constructed by Bruce Howard and colleagues (Ziran, N., Ogryzko, V., Hirai, T., Russanova, V. and Howard, B. H., unpublished). pOZ contains a bicistronic transcriptional unit that allows expression of two proteins from a single transcript. This design ensures tight coupling between expression of the gene of interest and the selection marker, the interleukin-2 receptor a-chain (IL2Ra) [10]. This approach is particularly advantageous when the expression of proteins of interest causes growth retardation. pOZ-FH-N and pOZ-FH-C allow expression of the protein of interest in, respectively, the N-terminally and C-terminally tagged form. Generally speaking, the epitope tags are flexible and accessible to antibodies without disrupting the tagged complexes. Therefore, the position of the epitope tags has proved not to be an important issue in most proteins we have tested, except in those cases in which we found the nature of the complexes purified or the functional integrity of the protein to depend on the position of the epitope. Thus, it would be worth testing both an Nterminally and a C-terminally tagged protein. Upstream of the XhoI site, pOZ-FH-N contains a Kozak sequence [11], an initiation methionine, and the FLAG [12] and HA [5] tags. The open-reading frame (from the second codon to the stop codon) of protein of interest should be subcloned into the XhoI and NotI sites. We typically amplify the open-reading frame by PCR as

an XhoI-NotI fragment and subclone it into pOZ-FH-N. If XhoI sites are present in the open-reading frame, an alternative enzyme that produces an overhang comparable to XhoI (e.g., SalI) can be used. pOZ-FH-C contains the FLAG and HA tags and a stop codon downstream of the NotI site. If one amplifies the open-reading frame by PCR, the 5' PCR oligo should include a Kozak sequence [11] and an initiating methionine after the XhoI sequence. Note that the pOZ vector has a cryptic translation start site upstream of the cloning sites. To avoid formation of a fusion protein, the oligo should be designed so that the protein of interest is out of frame from the XhoI (CTC–GAG) site. A limitation of the retroviral system is the size of protein to be expressed. Retroviruses can package RNAs as large as 10 kb, measuring the distance from the start of the 5' LTR to the end of the 3' LTR. Thus, the maximum size of the insert in this system is ~6 kb. Although this limit has not been systematically tested, we have successfully expressed several proteins with molecular weights of about 100–130 kDa [8]. Once the retroviral vector is constructed, we recommend checking expression of the epitope-tagged protein by transient transfection. We routinely transfect the retroviral plasmid into 293T fibroblasts. 48–72 hours after transfection, the epitopetagged protein is immunoprecipitated from cell lysates with anti-FLAG antibody–conjugated beads, and is subsequently detected by immunoblotting with appropriate antibodies, e.g., antibodies against FLAG, HA and/or the protein of interest.

Transfection of the retroviral vector into packaging cells

In order to generate transduction-competent retrovirus, the retroviral vector is transfected into a cell line that expresses retroviral structural gene products (gag, pol and env) necessary for packaging of retroviral RNA into infectious particles [13, 14]. The packaging cells routinely used for generation of amphotropic retrovirus, necessary to infect HeLa cells, are Bing or Phoenix A cells. Although the retroviral system employed is replication-deficient, a safer alternative is to produce ecotropic retroviruses by transfecting the retroviral plasmid DNA into an ecotropic packaging cell line, such as Phoenix E. HeLa cells that stably express murine ecotropic receptors can be infected with the resulting viruses. In any event, work with amphotropic and ecotropic retroviruses should be done using appropriate safety precautions. In the following, we describe a calcium phosphate transfection method [13, 14]. Although this method is widely employed for introduction of retroviral vectors, lipid-mediated transfection methods can be also used. We found that transfection with FuGENE 6 (Roche) works as efficiently as does the calcium phosphate method.

Reagents

2 X HBS: 50 mM Hepes, 10 mM KCl, 270 mM NaCl, 11 mM Dextrose (glucose), 1.5 mM Na2HPO4. Adjust pH to 7.05 at room temperature with NaOH, filtrate through a 0.45 micron filter, aliquot, and store at –20°C.

Procedure

1. Plate packaging cells (Bing or Phoenix A) at the density of 6 x 106 per 100 mm–diameter tissue culture dish in 10 ml of DMEM/10% FCS. Grow overnight in a 37°C incubator (5% CO2).

2a. Mix plasmids to be transfected and CaCl2 solution as follows. First, dilute 2 M CaCl2 with H2O in a 1.5 ml Eppendorf tube. Then add the plasmids indicated below to the calcium solution. Although packaging cells are supposed to express gag, pol, and env, these genes are relatively unstable. Thus, we co-transfect the retroviral vector along with expression plasmids for gag-pol and env.

24.8 ml

2 M CaCl2

5 mg

Retroviral plasmid

2.5 mg

gag-pol expression plasmid

2.5 mg to the total volume of 200 ml

env expression plasmid H2O

2b. Place 200 ml of 2 X HBS in a 14 ml round bottom tube (Falcon 2059 or equivalent). Add the DNA/calcium mixture (from Step 2a) into 2 X HBS dropwise, mixing after each addition.

3. Add the transfection mixture (from Step 2b) dropwise to the packaging cells in a dish (from Step 1), stirring well after each addition, and incubate the plate in a 37°C incubator (5% CO2).

4. 12 hrs after transfection, replace the medium with fresh DMEM/10% FCS and incubate for an additional 48 hrs.

5. Centrifuge the culture supernatant at 2,000 x g for 5 min. To remove the packaging cells completely, filter the supernatant through a cellulose acetate membrane with 0.45 mm pore size, which yields a retroviral stock. We recommend using the retroviral stock for transduction the same day.

Retroviral transduction

The next step is introduction of recombinant retrovirus into cells that are used for expression and purification of the epitope-tagged protein. We prefer to use HeLa S3 cells for large-scale protein purification since they grow as a suspension culture in a medium containing calf serum, which is significantly less expensive than fetal calf serum. However, many other cell lines can be used if required. We have successfully purified complexes from NIH3T3 mouse fibroblasts and BJ-1 human fibroblasts.

We use HeLa S3 cells so that we can employ spinner culture. HeLa S3 cells can be grown in either DMEM/CS in a tissue culture flask or Joklik’s medium/CS in a spinner bottle [15]. In general, we maintain HeLa cells in tissue culture flasks and grow them in spinner bottles for a large-scale culture. The amount of CS required for optimum cell growth depends on serum lots (ranging between 5% and 7%). Before purchasing, we carefully select a serum lot that supports optimum cell growth at less than 5%. When HeLa cells are grown in a tissue culture flask, close the cap loosely and grow them in a 37°C tissue culture incubator with 5% CO2. However, when using a spinner bottle, close the cap tightly and grow them on a magnetic stirrer in a normal 37°C incubator. The density of HeLa cells should be kept between 2–6 x 105/ml.

Procedure

1. Grow HeLa cells in DMEM/5–7% CS in appropriate flasks. 5 x 105 cells per transduction would be required. When HeLa S3 cells are grown in DMEM/CS, they grow as a mixed population of floating and attached cells. When passaging the cells, tap the flask gently and transfer the cell suspension to a 50-ml conical tube (Falcon 2070 or equivalent). Cells that remain attached to the flask are washed with PBS, trypsinized, and harvested. Combine these cells with the cell suspension in the Falcon tube.

2. Transfer the cell suspension containing 5 X 105 cells to a 15-ml conical tube and centrifuge at ~2,000 g for 5 min. While spinning, add 0.1 ml of 400 mg/ml Polybrene to 10 ml viral stock.

3. After centrifugation, discard the supernatant and resuspend the cell pellet into 10 ml viral stock/Polybrene mixture. Suspend thoroughly by pipetting up and down, transfer the suspension to T75 flask, and place in a 37°C incubator with 5% CO2.

4. When the cell density reaches ~5 x 105/ml (or 80% confluency), transfer the cells to 25 ml of the fresh medium in a T175 flask, and allow them to continue to grow in a 37°C incubator with 5% CO2.

5. When the cell density reaches ~5 x 105/ml (or 80% confluency) in the T175 flask, they are ready for sorting.

Sorting of transduced cells

The next step is selection of the transduced cells by magnetic affinity sorting with antibody against the surface selection marker [10]. Usually, two to three rounds of sorting are sufficient to achieve a pure population. However, the number of sorting rounds depends on protein of interest, especially when expression of the protein of interest causes growth retardation; given that small contaminants of non-transduced

cells would, in such a case, eventually overtake the population, additional rounds of sorting may be required.

Materials

Dynabeads M-450 goat anti-mouse IgG (Dynabeads, Catalog number: 110.06) PBS/BSA: 0.1% BSA in PBS Storage solution: PBS/BSA in 50% glycerol Mouse anti-IL2Ra antibody (Upstate Biotechnology, Inc., #05-170) Magnets for separation!Biomag flask separator (Polysciences, # 41015)

Preparation of anti-IL2Ra antibody–conjugated magnetic beads

1. Wash 100 mg of Dynabeads 3 times with 10 ml PBS/BSA in a T25 flask by magnetic separation. After the final wash, suspend the beads in 3.2 ml PBS/BSA.

2. Dissolve 200 mg anti-IL2Ra antibody in 0.8 ml PBS/BSA.

3. Transfer 0.8 ml anti-IL2Ra antibody (Step 2) to the tube containing the bead suspension.

4. Rotate or shake the tube at 4°C overnight.

5. Wash 4 times with 10 ml PBS/BSA by magnetic separation.

6. Resuspend the beads in 4 ml storage solution. Make aliquots and store at -20°C.

Sorting procedure

1. Add anti-IL2Ra antibody–conjugated magnetic beads (5 ml/sample) to DMEM/5–7% CS (1 ml/sample), and suspend thoroughly by vortexing. If you have multiple samples, prepare enough suspension (e.g., 1.1 ml x sample number) for the last aliquot.

2. Add 1 ml bead suspension to the cell culture in the T175 flask. Incubate the flask in a 37°C incubator (5% CO2) for 30 min with occasional shaking. Observe binding of the beads to the cells under a microscope.

3. Transfer the suspension of cells and beads to a T25 flask as follows. Tap the T175 flask gently and transfer the suspension to a new T25 flask. To recover tightly attached cells from the T175 flask, wash them with PBS, and then treat with 2 ml 0.05% trypsin solution. To avoid degradation of the surface marker, exposure to trypsin should be kept to a minimum. Thus, when cells start to be detached, tap the flask several times and add 5 ml of DMEM/CS immediately to neutralize the trypsin. Combine these cells with the suspension of those that have already been transferred to the T25 flask. Close the cap tightly.

4. Sandwich the T25 flask between magnetic plates (Biomag flask separator) and secure with rubber bands.

5. Stand the T25 flask/magnetic plate assembly vertically (the cap uppermost) in a tissue culture hood. Gently rock the flask occasionally to prevent the cells from accumulating at the bottom.

6. After 30 min, gently aspirate the medium containing the unbound cells.

7. To wash the positive cells and remove negative cells, add 50 ml DMEM/5–7% CS to the T25 flask and close the cap tightly. Gently rock the flask for a few minutes, and aspirate the medium. Repeat this washing step two or three times.

8. After the final washing, remove the flask from the magnetic plates. Add 5 ml DMEM/5–7% CS and suspend the cells by pipetting up and down. Carefully wash walls of the flask to recover the cells. Observe binding of the beads to the cells under a microscope and incubate the flask in a 37°C incubator with 5% CO2. If the cell number is less than ~1 x 105, transfer to a smaller flask.

9. When the cell density reaches ~5 x 105/ml (or 80% confluency), add 1 ml bead suspension (see Step 1) to the flask. Repeat Step 2–8. For a further sorting, allow the sorted cells to grow in a 37°C incubator with 5% CO2.

10. Repeat cell sorting until ~100% cells bind to anti-IL2Ra–conjugated magnetic beads, as judged by viewing under a microscope.

Detection of expression of the epitope-tagged protein

Once ~100% of the cells become IL2Ra-positive, propagate the cells in DMEM/5–7% CS in flasks. We routinely propagate cells to ~8 x 107 in four T175 flasks. These cells are used for: 1) checking expression of the epitope-tagged protein; 2) preparing frozen stocks; and 3) large-scale cell culture. We determine expression of the epitope-tagged protein by immunoblotting and

in situ immunofluorescence microscopy with anti-FLAG and/or HA antibody. The former analysis provides information about the expression level and molecular weight of the epitope-tagged protein, whereas the latter shows expression level, subcellular localization, and variation of expression levels among various cells. Epitope-tagged proteins may be detected by direct immunoblotting of cell lysates. However, given that the sensitivity and specificity of the FLAG and HA antibodies for immunoblotting is relatively low, we first concentrate epitope-tagged proteins by immunoprecipitation. In our experience, all tagged proteins tested so far are detected by immunoprecipitation followed by immunoblotting (see below). We also detect tagged proteins by in situ immunofluorescence microscopy [16, 17]. However, if the copy number of a tagged protein is relatively low, they may not be detectable by the latter method. If a specific antibody that recognizes the endogenous protein is available, we also compare expression levels of the protein of interest in transduced (the

endogenous and tagged protein) and non-transduced (the endogenous protein) HeLa cells. In our experience, transduced and non-transduced HeLa cells do not differ with regard to most proteins that we have tested. If the epitope-tagged protein can be separated from the endogenous one by SDS–PAGE, the endogenous and tagged protein levels in transduced cells can be analyzed.

Materials

RIPA buffer: 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% NP40, 0.5% DOC, 0.1% SDS. Store at 4°C.

M2 anti-FLAG antibody-conjugated agarose (Sigma A 2220)

Procedure

1. Spin down 1 x 107 cells and wash with PBS.

2. Suspend the cell pellet in 0.8 ml RIPA buffer and transfer to a 1.5 ml Eppendorf tube.

3. Vortex 1 min and rotate in a cold room for 30 min.

4. While extracting proteins, wash M2 anti-FLAG antibody–conjugated agarose beads to remove unconjugated antibody. Place ~10 ml (packed volume) antibody

beads in a 0.5 ml Eppendorf tube and add 200 ml 100mM glycine-HCl (pH 2.5). Vortex gently and leave for 2 min (do not expose to the acidic buffer for any longer than this). Spin down at 2,000 x g for 1 min and remove the supernatant with care. Suspend the beads in 200 ml 200 mM Tris-HCl (pH 8.0)/0.1% Tween 20. Spin down at 2,000 x g for 1 min and remove 180 ml supernatant.

5. Centrifuge the cell lysate (Step 3) at 15,000 x g for 5 min and transfer the supernatant to a new Eppendorf tube. 400 ml cell lysate is used for immunoprecipitation, whereas the rest is saved for other experiments, such as detection of protein of interest with a specific antibody.

6. Add 400 ml cell lysate to the washed beads (Step 4). Rotate or shake the tube for 1 hr in a cold room.

7. Wash 3 times with 200 ml PBS/0.1% Tween 20 by centrifugation at 2,000 x g for 1 min.

8. After the final wash, remove the supernatant. Spin again at 2,000 x g for 1 min and carefully remove residual buffer.

9. Resuspend in 10 ml elution buffer and incubate at room temperature for 30 min with occasional gentle vortexing (vortex very gently so that beads and buffer remain at the bottom of the tube).

10. Spin again at 2,000 x g for 1 min and carefully recover the supernatant, which can be used for immunoblotting with antibodies against FLAG, HA and/or protein of interest. Alternatively recover the supernatant by passing through a microspin column (BioRad Micro Bio-Spin or equivalent).

Large-scale cell growth and complex purification

This section describes large-scale cell growth, cell fractionation, and affinity purification of the double-tagged protein. The scale of purification depends on the level of expression and purposes of the experiments. We purify complexes from as few as 2 x 107 cells (a single T175 flask) to as many as 8 x 1010 cells (100 liter culture). If identification of the complex by mass spectrometry is a goal and the complex is relatively abundant, purification of the complex from ~6 x 109 cells (8 liter culture) would be sufficient. However, amounts required for identification of proteins vary depending on equipment, methodologies, investigator’s skill, character of the proteins, etc. We describe here the protocol for tissue culture at the 8-liter scale. HeLa cells can be grown in Joklik medium as a suspension culture. For large-scale spinner culture, we place heavy-duty magnetic stirrers (Bellco #7785-D1108 or equivalent) in a large 37°C incubator (Forma Scientific #S33665-SC3 or equivalent). If a large incubator is not available, HeLa cells can be grown in a 37°C warm room. Note that caps of the spinner flask should be tightly closed and the density of HeLa cells kept between 2–6 x 105/ml.

Growth of the cells

Materials

1 liter Spinner bottle (Bellco # 1965-01000) 12 liter Spinner bottle (Bellco # 1965-12000)

Procedure

1. Spin down 4 x 107 cells (see Detection of expression of the epitope-tagged

protein) and resuspend in 10 ml Joklik medium containing 5-7% CS. Transfer to 200 ml Joklik medium containing 5-7% CS in a 1-liter spinner bottle (2 x 105 cells/ml). Put the spinner bottle on a magnetic stirrer in a 37°C incubator and stir at 60 rpm.

2. Monitor cell density every 24 hrs and add Joklik medium containing 5-7% CS pre-warmed to 37°C to a final density of 2 x 105 cells/ml.

3. After 72 hrs, the cells should have grown to yield ~3.2 x 108 in ~800 ml. Transfer the whole culture to a 12-liter spinner bottle by decantation and add Joklik medium containing 5-7% CS pre-warmed to 37°C to a final density of 2 x 105 cells/ml.

4. Monitor cell density every 24 hrs and add Joklik medium as described above.

5. After 72 hrs, the cells should have grown to yield ~2.5 x 109 in ~6.4 liters. Add Joklik medium with 5-7% CS pre-warmed to 37°C to the final volume of 8 liters. Continue to grow until the density reaches 5~8 x 105 cells/ml.

6. Harvest cells by centrifugation.

Cell fractionation and complex purification

An method of cell fractionation should be chosen that is appropriate for the subcellular localization and characterization of the protein of interest. We typically prepare cytoplasmic and nuclear extracts as well as nuclear pellet fractions from fresh HeLa cells [18]. These fractions can be stored at –80°C. Here we describe our purification method from nuclear extracts. At least in the beginning, we recommend performing parallel mock purification from non-transduced HeLa cells. HeLa nuclear extracts contain proteins that specifically bind to M2 antiFLAG antibody. Such “contaminants” can be detected in samples purified from both transduced and non-transduced cells. The “contaminants” can be removed by the second immuno-affinity purification, since they do not bind to 12CA5 anti-HA antibody. Thus, complexes can be purified almost completely by the sequential immuno-affinity purification.

Anti-FLAG antibody immuno-affinity purification

Materials

M2 anti-FLAG antibody-conjugated agarose (Sigma A 2220) 12CA5 anti-HA antibody-conjugated Protein A–Sepharose: Immobilize 12CA5 anti-HA antibody to Protein A Sepharose 4 Fast Flow (Amersham Biosciences) at xx mg/ml and crosslink with xxx using a standard protocol [17].

Washing buffer: 20 mM Tris HCl (pH 8.0), 100 mM KCl, 5 mM MgCl2, 0.2 mM EDTA, 10% glycerol, 0.1 % Tween. Just prior to use, add 10 mM 2-mercaptoethanol and 0.25 mM PMSF.* *Prepare 250 mM PMSF solution of PMSF in DMSO and store at –20°C.

FLAG-elution buffer: 50 mg/ml FLAG peptide in Washing buffer. To prepare the stock solution of FLAG peptide, reconstitute lyophilized FLAG peptide (Sigma, # F3290) with 0.2 M Tris-HCl (pH 8.0) to 5 mg/ml, make aliquots, and store at -20°C.

HA-elution buffer: 50 mg/ml HA peptide in Washing buffer. Prepare the stock solution of HA peptide from reconstitute lyophilized HA peptide (Roshe, # 1666975) as described for the FLAG peptide.

Procedure

1. Typically, an 8-liter culture yields ~15 ml (~8 mg protein/ml) nuclear extract. If a nuclear extract is stored at –80°C, thaw it rapidly in 37°C water bath, shaking occasionally. Put the tube on ice just before thawing the extract completely.

2. Centrifuge the extract at 50,000 x g for 30 min at 4°C. Immediately after centrifugation, transfer the supernatant to a new centrifugation tube with a 10-ml pipette. Since the pellet is relatively soft and the lipid layer on the top diffuses rapidly, it is hard to obtain a clear supernatant, but this is not a cause for concern.

3. Centrifuge again at 50,000 x g for 30 min at 4°C.

4. While spinning the extract, wash M2 anti-FLAG antibody–conjugated agarose beads. The amount of the beads required for purification depends on the level of the epitope-tagged protein in the extract and the accessibility of the epitopetagged protein to the antibody. Thus, we recommend preliminary experiments with each protein to determine the optimum amount of beads. In our experience, 200 ml beads per 10 ml nuclear extract may be used for most proteins, although this may not be optimal in all cases. Pipette the beads onto a column (BioRad Poly Prep #731-1550 or equivalent). Wash a sufficient amount of beads so that there is enough for the last aliquot, e.g., volume per sample x (sample number + 0.5). The packed volume of the beads can be measured by the scale on the

column. Wash the beads with 10 volumes of 100 mM glycine-HCl (pH 2.5) to remove uncrosslinked antibody.

5. Wash the beads with 10 bed volumes of 0.2 M Tris-HCl (pH 8) and then 10 bed volumes of washing buffer.

6. After the last drop from the column, close the bottom with the cap (supplied with columns) and add 1 volume of washing buffer to the column.

7. With an Eppendorf Pipetman P-1000 (or equivalent), transfer the bead suspension (50% slurry) to a 15-ml conical tube (Falcon 2096 or equivalent). To transfer the beads completely, cut the end of the tip with a clean razor blade. Suspend thoroughly by pipetting up and down with P-1000 (or equivalent), and transfer the aliquot to the tube. After the beads settle, read the packed volume of the beads off the scale on the tube. If necessary, adjust the volume by adding or removing the beads. Keep the tube on ice.

8. Immediately after centrifugation (from Step 3), transfer the supernatant to the tube containing M2-antibody beads (from Step 7).

9. Close the cap tightly, and rotate or shake the tube for 3 hrs at 4°C.

10. While rotating the tube, set up a column (BioRad Poly-Prep Column or equivalent) over a 50-ml conical tube using a plastic column holder (such as the one provided with QUIAGEN Plasmid Midi Kit). Chill the column in a cold room.

11. Load the sample onto the column. Allow it to drain completely by gravity flow. Transfer the flow-through to an appropriate tube and store it at –80°C, if necessary.

12. Wash the beads by filling the column to the top with washing buffer, and allowing it to drain completely by gravity flow. Repeat the washing step for a total of three washes. Empty the 50 ml Falcon tube before each wash.

13. After the final wash, set up the column over a 15-ml conical tube. To remove liquid from the beads, spin the column at 1,200 x g for 5 min in a Beckman J6-M1 centrifuge with a JS-4.2 rotor or equivalent.

14. Immediately after centrifugation, close the bottom of the column with the cap. Load 200 ml (or volume equal to the packed volume of the beads) FLAG-elution buffer to the beads, and suspend well by vortexing gently. Incubate at room temperature for 30 min, mixing occasionally. Alternatively, incubate at 4°C for 1 hr. In most cases, we elute proteins at room temperature rather than at 4°C, because it is faster.

15. Remove the bottom cap and spin the column on a new 15-ml conical tube at 1,200 x g for 5 min. Transfer the purified sample to a 1.5 ml Eppendorf tube. 80%~90% of the complex can be recovered in the first elution. If necessary, repeat the elution step. When the sample is immediately further purified by antiHA antibody immunoprecipitation, keep the sample on ice. Otherwise, freeze it on powdered dry ice and store at –80°C.

Anti-HA antibody immuno-affinity purification

1. Wash anti-HA 12CA5 antibody-conjugated beads as follows. Place ~20 ml (packed volume) antibody beads in a 0.5 ml Eppendorf tube and add 200 ml glycine-HCl (pH 2.5). Vortex mildly and leave for 2 min (do not expose to the acidic buffer an longer than this). Spin down at 2,000 x g for 1 min and remove supernatant with care. Suspend the beads in 200 ml 200 mM Tris-HCl (pH 8.0)/0.1% Tween 20. Spin down at 2,000 x g for 1 min, remove the supernatant, and resuspend in 200 ml of washing buffer.

2. Spin the beads at 2,000 x g for 1 min, remove the supernatant, and load ~200 ml anti-FLAG antibody–immunopurified material. Save 10~20 ml anti-FLAG antibody–purified material for SDS-PAGE analysis. In general, we find that ~20 ml of the antibody beads provides a large excess and is more than sufficient for ~200 ml sample. However, we often find protein complexes with poor binding to anti-HA antibody, perhaps due to poor accessibility to the epitope. If you find a significant amount of the unbound complex, increase the amount of the antibody beads.

3. Spin down at 2,000 x g for 1 min and transfer the supernatant to a new 500 ml Eppendorf tube, saving the unbound material for SDS-PAGE analysis.

4. Resuspend in 200 ml washing buffer and transfer to a microspin column (BioRad Micro Bio-Spin or equivalent), set up over a 1.5-ml Eppendorf tube. Spin at 2,000 x g for 1 min.

5. Wash again with 200 ml washing buffer by centrifugation at 2,000 x g for 1 min.

6. After centrifugation, close the bottom with the cap (supplied with columns). Load 40 ml (or 2 bed volume) of the HA-elution buffer to the packed beads. Incubate at room temperature for 1 hr, mixing occasionally.

7. Recover the eluant by centrifugation at 2,000 x g for 2 min.

8. For the second elution, repeat Steps 6 and 7.

Analysis of the affinity-purified materials

We typically analyze 2–10 ml anti-FLAG antibody–immunopurified material, 2–10 ml flow through of anti-HA antibody–immunopurification, 0.4–2 ml each of the first and second eluates from anti-HA antibody–beads on an SDS-PAGE gel. Although silver staining is not quantitative, we analyze the samples along with a series dilution of a

molecular weight marker for rough estimate of the amount of the purified materials. We routinely run 4–20% Tris-glycine SDS-PAGE or 4–12% Bis-Tris NuPAGE gels (Invitogen), which are suitable for analysis of wide-range of molecular weight proteins ranging from less than 10 kDa to over 500 kDa.

Acknowledgements

We would like to thank members in the Nakatani and Ogryzko labs for valuable comments.

References:

1.

R.H. Ebright, J Mol Biol, 304, 687, (2000).

2.

S. Fields and O. Song, Nature, 340, 245, (1989).

3.

A. Pandey and M. Mann, Nature, 405, 837, (2000).

4.

C.M. Chiang, et al., Embo J, 12, 2749, (1993).

5.

J. Field, et al., Mol Cell Biol, 8, 2159, (1988).

6.

T. Ikura, et al., Cell, 102, 463, (2000).

7.

H. Ogawa, et al., Science, 296, 1132, (2002).

8.

V.V. Ogryzko, et al., Cell, 94, 35, (1998).

9.

Q. Zhou, T.G. Boyer, and A.J. Berk, Genes Dev, 7, 180, (1993).

10.

R. Padmanabhan, et al., Anal Biochem, 170, 341, (1988).

11.

M. Kozak, J Cell Biol, 115, 887, (1991).

12.

T.P. Hopp, et al., Bio/Technology, 6, 1204, (1988).

13.

Current Protocols in Molecular Biology, ed. F. Ausubel, et al., New York: John Wiley & Sons.

14.

Retroviruses, ed. J. Coffin, S. Hughes, and H. Varmus. 1997, Cold Spring Harbor Laboratory, NY: Cold Spring Harbor Laboratory Press.

15.

W.F. McLimans, Methods Enzymol, 58, 149, (1979).

16.

Cells : A Laboratory Manual, ed. D. Spector, B. Goldman, and L. LA. 1998, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.

17.

E. Harlow and D. Lane, Using Antibodies: A Laboratory Manual. 1988, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.

18.

J.D. Dignam, et al., Methods Enzymol, 101, 582, (1983).

Figure Legend

Figure 1.

A. Restriction map of the retroviral vector pOZ-FH-N.

B. Cloning sites of pOZ-FH-N (top) and pOZ-FH-C (bottom), represented by the gray box (FH) for pOZ-FH-N in A.

A

BglII XhoI/NotI

FH

EM

C IR

ES

NcoI

IL 2R

HindIII

or

BamHI

3' LT R

5' LT R

α

pOZ-FH-N 7.9 kb i

a mp

B pOZ-FH-N BglII AGATCTTCCGCTCGATGAACCATGGACTACAAGGACGACGATGACAAGCTCGATGGAGGATACCCCTACGACGTGCCCGACTAC M D Y K D D D D K L D G G Y P Y D V P D Y XhoI NotI GCCGGAGGACTCGAGGAATTCGCGGCCGCT A G G L E E F A A A

FLAG

HA

pOZ-FH-C BglII XhoI NotI AGATCTTCCGCTCGAGGAATTCGCGGCCGCTGGAGGAGACTACAAGGACGACGATGACAAGTCGGCCGCTGGAGGATACCCC D L P L E E F A A A G G D Y K D D D D K S A A G G Y P FLAG TACGACGTGCCCGACTACGCCTAG Y D V P D Y A * HA

Fig. 1