PROTOCOL

Southwestern blotting in investigating transcriptional regulation Francis K Y Siu, Leo T O Lee & Billy K C Chow School of Biological Sciences, The University of Hong Kong, Pokfulam Road, HKSAR, China. Correspondence should be addressed to: B.K.C.C. ([email protected]).

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

Published online 20 December 2007; doi:10.1038/nprot.2007.492

Southwestern blotting is used to investigate DNA–protein interactions. The advantage of this technique over other related methods such as electrophoretic mobility shift assay (EMSA) and DNA footprinting is that it provides information regarding the molecular weight of unknown protein factor. This method combines the features of Southern and Western blotting techniques; a denaturing SDS-PAGE is first employed to separate proteins electrophoretically based on size, and after transferring the proteins to a membrane support, the membrane-bound proteins are renatured and incubated with a 32P-labeled double-stranded oligonucleotide probe of specific DNA sequence. The interaction of the probe with the protein(s) is later visualized by autoradiography. This technique could be combined with database searching (TransFac, http://www.gene-regulation.com/pub/databases.html#transfac), prediction of potential protein factors binding onto a target motif (e.g., Patch search), in vitro supershift EMSA and in vivo chromatin immunoprecipitation (ChIP) assays for effective identification of protein factors. The whole Southwestern blotting procedure takes B4 d to complete. In this article, a commonly used protocol and expected results are described and discussed.

INTRODUCTION Understanding the spatial and temporal expression patterns of a gene is essential in elucidating its functions. Southwestern blotting is useful in investigating DNA–protein interactions to aid the identification of DNA-binding protein factors. Transcription and hence expression regulation of a gene involves initially highly specific interactions of protein factors with distinct short DNA sequence motifs, 6–8 bp in size, usually located within the promoter or other functional cis-acting elements. Identification of protein factors that bind to genes to turn them on or off is therefore important in investigating gene functions. Southwestern blotting is a technique developed for the detection and characterization of protein factors interacting with distinct DNA motifs. The technique is modified from a previous procedure called ‘protein blotting’ described by Bowen et al.1. In this original protocol, proteins are first separated based on size by SDS-PAGE. The polyacrylamide gel containing the proteins is then sandwiched between two filters so that the proteins can diffuse out and bind onto the filters to produce replicas of the gel. One of these filters is used to visualize proteins in the lanes by staining, and the other one is for detecting protein factors interacting with a labeled probe1. The technique usually requires either heat inactivation2 or partial purification3–5 of crude protein extract. Since then, modifications of the initial protocol have been suggested: first, electrophoretic transfer of proteins from the SDS-PAGE gel to the filter is used instead of diffusion so that renaturation of proteins can occur during the transfer process, as the electric field rapidly removes SDS from the proteins; second, before the application of the probe, the filter is incubated with skim milk to block nonspecific and low-affinity binding sites; and finally, DNA binding is performed under optimal salt concentrations to facilitate specific and high-affinity DNA–protein interactions6–8. A flow chart briefly describing the procedures and time needed for Southwestern blotting is shown in Figure 1. When studying the interactions between transcription factors with DNA motifs, electrophoretic mobility shift assay (EMSA) is one of the most popular methods as it is rapid, simple and sensitive.

The principle behind this technique is that the electrophoretic mobility of the higher molecular weight protein–nucleic acid complex is less than that of the free probe. However, EMSA suffers Harvest cells to prepare nucleus extract by protocol described Steps 1–11 2 h* Anneal two synthesized oligonucleotides Steps 22–25

Separate proteins by SDS-PAGE Steps 12–18

1 h*

3 h*

Label the annealed oligos with P using end-labeling Step 26

32

Transfer proteins to membrane by electroblotting Steps 19–21

20 min 1 h 15 min* Purify the oligos using a G-25 column Steps 27 and 28

Renature the proteins in TNED buffer room temp. Steps 29–30

10 min*

Overnight Hybridize the proteins with radioactively labeled probe Step 31 Overnight

Wash the membrane in TNED buffer Step 32 15 min × 3

Visualize the signal using X-ray film or molecular imager Steps 33–35

1–3 d

Decide the molecular weight of the factors

Figure 1 | A flow chart summarizing the approximate time needed and the procedures involved in Southwestern blotting. Asterisk represents pause points between the steps. NATURE PROTOCOLS | VOL.3 NO.1 | 2008 | 51

PROTOCOL Figure 2 | A flow chart showing a strategy for identifying DNA-binding proteins (e.g., transcription factors) by Southwestern blotting in combination with other methods. Promoter assays are first performed to identify the promoter regions that are important for regulating transcription of the target gene. Information from promoter studies will be used for designing oligonucleotide primers for electrophoretic mobility shift assay (EMSA), which confirm binding of protein factors with the DNA sequence in an in vitro environment. Southwestern blotting is unique as it is the only technique that shows the molecular weight of the protein factor. On the basis of these studies, one can search databases TransFac and Patch search to identify the protein. To confirm, supershift and chromatin immunoprecipitation (ChIP) assay should then be performed to show the in vitro and in vivo interactions of the protein and the DNA motif using specific antibody.

Identify a motif by mutation analysis (promoter assay)

Design oligo and test for DNA–protein interaction by EMSA

Perform Southwestern analysis with the same probe as EMSA

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

Test the specificity of the Southwestern signal using a mutated probe

from many limitations; for example, many transcription factors could bind several related and even unrelated DNA sequences. Also, the electrophoretic mobility shift pattern provides little information regarding the identity of various proteins in the complex(es) and further analysis such as supershift assays are needed. Moreover, supershift assays require prior experience to establish the link between electrophoretic mobility shift pattern and DNA-binding proteins, and also it is difficult to select appropriate antibodies for the study9. Southwestern blotting, on the other hand, could provide information regarding the molecular weights of the DNA-binding proteins as the technique involves the separation of proteins under SDS-PAGE. Hence, with this information, the identity of the protein(s) could be deduced by searching transcription factors databases, for example, TransFac10. It is, similar to EMSA, also a simple technique that does not require special equipment. The major problem with Southwestern blotting is that the separation of proteins under denaturing SDS-PAGE could dissociate polymeric protein factors leading to ineffective binding of DNA. Therefore, Southwestern blotting is not suitable for detecting protein factors that require more than one subunit for efficient DNA binding. In addition, large quantities of protein extract are required in SDSPAGE to increase the sensitivity of the technique. Based on the pros and cons of EMSA and Southwestern techniques, a strategy to identify and confirm DNA–protein interactions, with which to study the involvement of DNA motifs and protein factors in gene regulation is suggested in Figure 2. Most of the techniques employed in identifying protein factors, when used alone, suffer from some limitations. For example, computer-based analysis, such as P-match search11, often results in suggesting too many possible protein factors. Site-directed mutagenesis coupled to transient expression provides information on functional cis-acting motifs in an in vitro cell model. EMSA shows only if there are factors interacting with a putative motif. Supershift EMSA, on the other hand, serves to confirm some already predicted protein factor(s), as well as it depends on the availability of antibodies and often produces ambiguous data. Hence, it is necessary to combine all these techniques to effectively identify functional protein factors/cis-acting motifs in regulating a gene. Experimental design Cell types and sample preparation. A wide range of cells can be used for the preparation of protein extracts for Southwestern analysis (see Steps 1–11). There is no specific limitation to the choice of cell lines, though we recommend performing functional 52 | VOL.3 NO.1 | 2008 | NATURE PROTOCOLS

Determine the molecular weight of the binding protein

Search TransFac with the ‘molwSearch’ utility

Comapare the results from the motif prediction (e.g., Patch search)

List the most possible transcription factors

Use supershift and/or functional analysis (e.g., overexpression and ChIP assay) to confirm such prediction

assays (e.g., EMSA or promoter mutation/deletion assay) beforehand to confirm the presence of functional target protein(s) in the cell. For continuous cell lines, propagation and subculturing conditions should follow instructions from the provider or American Type Culture Collection (http://www.atcc.org). Tissue or primary cell cultures in monolayer or cell suspension can also be used directly for preparing nuclear extracts. The protocol is similar, except that the tissue should be homogenized in nuclear lysis buffer using either Polytron or Dounce homogenizer. Determination of protein content. To determine protein concentration of the nuclear extract, protein assay based on either Bradford’s12 (Bio-Rad) or detergent compatible Lowry’s method13 (Bio-Rad) could be used. Procedure for the protein assay (Bradford’s method) is outlined in Box 1. Optimization of conditions. We recommend that the readers follow our suggested protocol at least for the first time. On the basis of data obtained, they can optimize experimental conditions. There are mainly two parameters to optimize: percentage of acrylamide gel and the amount of protein loaded on one lane. In general, a lower percentage of acrylamide is suitable for resolving larger proteins, while a higher percentage is for separating smaller proteins. Gel percentage should be adjusted only when poor resolution of desired proteins in the SDS-PAGE is observed. One can also try using a gradient gel (8–25%) for resolving proteins in a wide range of molecular weights. With respect to the amount of

PROTOCOL BOX 1 | BIO-RAD PROTEIN ASSAY

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

1. Prepare several dilutions of a protein standard (e.g., 0–10 mg ml 1 BSA) in nuclear extraction buffer. ! CAUTION Presence of detergent in the sample may affect the data; hence, it is best to prepare the standards with the nuclear extraction buffer (which contains only 0.1% detergent). Alternatively, if the concentration of sample protein is high, it could be diluted in a manner that the detergent concentration becomes r0.1%. In our experience, such low concentrations of detergent does not significantly interfere with the assay. 2. Pipette 800 ml of standards and samples into 1.5-ml centrifuge tubes. All the samples should be assayed in duplicate or triplicate. 3. Add 200 ml of dye reagent to each tube and mix thoroughly. 4. Incubate the mixture for 5 min at room temperature. ! CAUTION Absorbance will increase over time; samples should be incubated at room temperature for o1 h. 5. Absorbance is measured at 595 nm to determine concentration of samples by comparison with the standards.

protein, 100 mg of nuclear extract, in our hands, is sufficient for the detection of protein factors. If the target protein has a low expression level, one can try to use more nuclear extract or more radioactive probes. It should, however, be noted that more extract/ probe may lead to overloading of the SDS-PAGE and higher background in detection. When the background of the blot is high, one can reduce the level of nuclear extract or probe. Also, more stringent washing conditions, such as increasing the washing temperature, washing time and number of washes can also be tried. Please refer also to TROUBLESHOOTING for more detailed discussions.

when a probe is longer, it may include other binding motifs as well, which can increase nonspecific binding, while a shorter probe may result in lower binding affinity. It is strongly recommended that the probe should be tested using EMSA to confirm there is binding of protein factor(s) at least in in vitro environment. A double-stranded probe is produced from overlapping complementary oligonucleotides and can be used to identify transcription factors as they bind to double-stranded DNA. In a situation when one wants to study and identify proteins interacting with specific sequences in single-stranded DNA, single-stranded probe may then be used.

DNA probe design and synthesis. A functional target DNA-motif must be identified before considering Southwestern blotting. The probe is designed by including the target DNA sequences (usually 6–8 bp) with 5–10 bp of flanking sequences at both sides to make up an 18–30-nt probe. It should be noted that

Controls. For Southwestern blotting, the following control studies can be considered: (i) loading BSA as a control lane in SDS-PAGE; (ii) loading nuclear extract from cells without the target factor as determined by EMSA; (iii) using a mutated probe for binding and (iv) using an unrelated probe for binding.

MATERIALS REAGENTS . Tris–hydrochloride (Invitrogen, cat. no. 15506-017) . HEPES (Sigma, cat. no. H4034) . Glycerol (USB, cat. no. US16374) . Magnesium chloride anhydrous (MgCl2; Sigma, cat. no. M8266) . Triton X-100 (BDH, cat. no. 306324N) . DTT (USB, cat. no. 15397) . PMSF (USB, cat. no. 20203) . Leupeptin (USB, cat. no. 18413) . Pepstatin A (USB, cat. no. 20037) . Aprotinin (USB, cat. no. 11388) . BSA (Roche Applied Science, cat. no. 10735086001) . Bio-Rad protein assay (based on Bradford12; Bio-Rad, cat. no. 500-0006) . 30% Acrylamide/bis solution (29:1; Bio-Rad, cat. no. 161-0156) ! CAUTION Acrylamide and bisacrylamide are neurotoxic and are readily absorbed through skin. Wear gloves to avoid direct contact with the solution. . Trizma base (Sigma, cat. no. T1503) . SDS (BDH, cat. no. 108073J) . Ammonium persulfate (APS; Sigma, cat. no. 248614) . Tetramethylethylenediamine (TEMED; Invitrogen, cat. no. 15524-010) . Gly (USB, cat. no. 16407) . Bromophenol blue (USB, cat. no. 17-1329-01) . Methanol (Merck KGaA, cat. no. 106009) . Ready-To-Go T4 polynucleotide kinase (GE Healthcare Life Sciences, cat. no. 27-0737-01) . Redivue [g-32P] ATP 250 mCi (GE Healthcare Life Sciences, cat. no. AA0068) ! CAUTION When dealing with radioactive materials, appropriate safety precautions must be followed and National regulations adhered to.

. Skim milk powder (any commercially available low fat milk power is also suitable)

. Oligonucleotides (250 pmol ml 1) . Protein ladder: PageRuler prestained protein ladder (Fermentas, cat. no. SM0671) (other commercially available prestained protein ladders are also suitable) . Tris-buffered saline (TBS) (see REAGENT SETUP) . Cell lysis buffer (see REAGENT SETUP) . Nuclear extract buffer (see REAGENT SETUP) . Separating gel (see REAGENT SETUP) . Stacking gel (see REAGENT SETUP) . Running buffer (see REAGENT SETUP) . Sample buffer (see REAGENT SETUP) . Transfer buffer (see REAGENT SETUP) . TNED buffer (see REAGENT SETUP) . 10 Oligo annealing buffer (see REAGENT SETUP) EQUIPMENT . Cell lifter (Corning) . Centrifuges: Eppendorf 5417R (Eppendorf) and Beckman CS-6R (Beckman) . Rocker platform (Gyro rocker STR-9; Stuart Scientific) (other 3D gyroscopic rocker or rotation rocker could also be used as an alterative) . Spectrophotometer (U-2800, Hitachi) (any spectrophotometers which are able to measure absorbance 595 nm are also suitable) . Vertical electrophoresis apparatus: the Mini-Protean II gel system (Bio-Rad), gel size (8.3  7.3 cm2) (most of the commercially available vertical protein gel electrophoresis systems are also suitable; however, according to the gel sizes of other electrophoresis systems, the volume

NATURE PROTOCOLS | VOL.3 NO.1 | 2008 | 53

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

PROTOCOL of buffer or the amount of probe to be used in the procedure should be adjusted) . Protein transfer system: Mini Trans-Blot Cell (Bio-Rad) (other transfer systems could also be used, but the transfer protocol in Steps 19–21 should be modified accordingly) . Power supply for SDS-PAGE and electroblotting (POWER PAC300; Bio-Rad) (other power supplies with minimum support of 120 V, 500 mA are also suitable) . Boiler (any commercially available water boilers are suitable) . Nitrocellulose membranes (e.g., Hybond-C extra; GE Healthcare Life Sciences) or PVDF membrane (GE Healthcare Life Sciences) . 3-mm Paper (Whatman) . MicroSpin G-25 column (GE Healthcare Life Sciences) . Exposure cassette with intensifying screen (GE Healthcare Life Sciences) . Autoradiographic film: BioMax MR film (Kodak) REAGENT SETUP TBS 10 mM Tris–Cl, 150 mM NaCl (pH 8.0). Cell lysis buffer 20 mM HEPES (pH 7.6), 20% glycerol (vol/vol), 10 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.1% triton X-100 (vol/vol), 1 mM DTT, 1 mM PMSF, 10 mg ml 1 leupeptin, 10 mg ml 1 pepstatin, 100 mg ml 1 aprotinin. m CRITICAL PMSF, leupeptin, pepstatin and aprotinin must be freshly added to the lysis buffer just before use. Nuclear extract buffer Cell lysis buffer with 500 mM NaCl. m CRITICAL PMSF, leupeptin, pepstatin and aprotinin must be freshly added to the nuclear extract buffer just before use.

Separating gel Water (ml) 30% Acrylamide (29:1) (ml) 1 M Tris, pH 8.8 (ml) 10% SDS (ml) 10% APS (ml) TEMED (ml)

8% 2.53 2.00 2.81 75 75 7.5

10% 2.03 2.50 2.81 75 75 7.5

12% 1.53 3.00 2.81 75 75 7.5

The volume is enough for single gel (8.3 cm  B5 cm; W  H) Stacking gel Water (ml) 30% Acrylamide (29:1) (ml) 1 M Tris, pH 6.8 (ml) 10% SDS (ml) 10% APS (ml) TEMED (ml)

2.55 0.625 0.475 37.5 37.5 2.5

The volume is enough for single gel (8.3 cm  B2 cm; W  H) Running buffer 20 mM Tris-base, 200 mM Gly, 0.12% SDS (wt/vol) Sample buffer 30 mM Tris–Cl (pH 6.8), 30% glycerol (vol/vol), 10% SDS (wt/vol), 1 mM DTT, 0.002% (wt/vol) bromophenol blue Transfer buffer 25 mM Tris-base, 190 mM Gly, 20% methanol (vol/vol) TNED buffer 10 mM Tris (pH 7.5), 50 mM NaCl, 0.1 mM EDTA and 1 mM DTT 103 Oligo annealing buffer 10 mM Tris (pH 8.0), 100 mM NaCl, 1 mM EDTA

PROCEDURE Nuclear extract preparation TIMING B2 h 1| Culture cells on a 150-mm diameter dish (see Experimental design for details). Harvest cells when 80–90% confluency is reached. m CRITICAL STEP Avoid overgrowing cells which could lead to apoptosis and affect expression of transcription factors.



2| Remove culture medium and wash cells three times each using 10 ml TBS. In each wash, after adding TBS, shake the dish gently for 1 min before removing the buffer. 3| Add an appropriate amount of cell lysis buffer to the culture dish and proceed to Step 4 immediately. For one 150-mm diameter culture dish, 5-ml cell lysis buffer is recommended. 4| Dislodge the cells by scraping using a cell lifter and transfer the cells into a centrifuge tube. 5| Collect cell nuclei by centrifugation at 2,500g for 5 min at 4 1C. Remove supernatant as much as possible and avoid disrupting of the pellet. 6| Resuspend the pellet in nuclear extract buffer. We recommend using 650 ml nuclear extract buffer for a 150-mm diameter dish. Resuspend the pellet gently and completely by pipetting it up and down. Transfer the solution into a 1.5-ml centrifuge tube. 7| Lyse nuclei by rocking the mixture for 1 h at 4 1C on a rocker platform (50 r.p.m.). 8| Remove debris by centrifugation at 12,000g for 10 min at 4 1C. 9| Transfer the supernatant (nuclear extract) into a fresh centrifuge tube. 10| Measure protein concentration of the nuclear extract by a standard protein assay (e.g., Bio-Rad protein assay; see Box 1 and Experimental design for further details). Use BSA as a standard in the assay. m CRITICAL STEP Usually, 2–5 mg ml 1 nuclear extract concentration could be obtained from a 150-mm culture dish with 80–90% confluency depending upon the type of cell. We recommend the concentration should be at least 2 mg ml 1. 11| Aliquot the extract, quick-freeze the tubes in liquid nitrogen and store at 80 1C. ’ PAUSE POINT Prepare aliquots of the extract to prevent multiple freezing and thawing. The volume of each aliquot depends upon your experiment. We recommend an aliquot size of 100 ml. The extract can be stored at 80 1C for at least 1 month.



SDS-PAGE TIMING B3 h 12| Clean up the gel plates, spacers and comb with 70% ethanol and then rinse thoroughly using distilled water. Set up the apparatus and make sure that there is no leakage according to the manufacturer’s manual of the electrophoresis system. 54 | VOL.3 NO.1 | 2008 | NATURE PROTOCOLS

PROTOCOL If commercially available precast gels are used, proceed to Step 15. Refer to instructions provided by http://www.bio-rad.com for general gel cassette preparation or more specifically while using the Bio-Rad system. ! CAUTION Glass plates and combs must be thoroughly cleaned.

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

13| Pour the separating gel mix into the apparatus and overlay the poured gel with isopropanol (B500 ml). Allow the separating gel to polymerize for 30 min at room temperature (25 1C). In the meantime, prepare the stacking gel mixture as indicated in REAGENT SETUP. m CRITICAL STEP TEMED and APS should be added just before pouring the separating gel. The percentage of acrylamide in the separating gel can be adjusted according to the molecular weight of transcription factor. 14| Remove the isopropanol by aspiration, and pour the stacking gel mix into the apparatus. Insert the comb and allow the gel to polymerize for 20 min. ’ PAUSE POINT Polymerized gels can be stored for several hours at room temperature or overnight at 4 1C. 15| Remove the comb, fill the tank with 1 running buffer. Prerun the gel at 100–120 V for 30 min. 16| Mix the loading buffer with 100 mg nuclear extract in a 1:1 buffer:sample ratio (vol/vol). m CRITICAL STEP The amount of nuclear extract can be adjusted according to the expression level and the affinity of transcription factor with probe. 17| Rinse the wells with 1 running buffer before sample loading as residual unpolymerized acrylamide may affect mobility of proteins in the gel. Boil the samples for 5 min and load the samples and protein ladder into the wells. 18| Run the gel at 100–120 V, until the front dye (bromophenol blue) of the loading buffer reaches the end (B2 h). The progress of electrophoresis could be monitored also through the prestained protein ladder.



Electroblotting TIMING 1 h 15 min 19| Remove the gel from the gel cassette. Discard the stacking gel. Prepare a nitrocellulose membrane and two 3-mm Whatman papers that are the same size as the separating gel. 20| Set up the transfer as recommended by the manufacturer of the protein transfer system (Bio-Rad). m CRITICAL STEP Set up the gel-membrane sandwich under transfer buffer. Bubbles trapped between sandwiched layers must be removed. In addition to Hybond-C membrane, PVDF membrane could also be used as an alternative. However, it is necessary to activate the PVDF membrane by dipping it into methanol before setting up the transfer system. Other electroblotting techniques, such as semi-dry transfer system (Bio-Rad), are also suitable. 21| Electroblot proteins from the gel to the membrane at 100 V at 4 1C for 60 min. m CRITICAL STEP After electroblotting, monitor the intensity of the prestained protein ladder on the membrane to estimate the overall transfer efficiency. Proteins with larger molecular weight (4100 kDa) take a longer time to complete the transfer. Increase the transfer time and/or use freshly prepared transfer buffer to enhance the overall efficiency whenever necessary. ’ PAUSE POINT The blot can be stored at 4 1C until the renaturing step.



Preparation of DNA probe TIMING 1.5 h 22| Synthesize complimentary oligonucleotides containing the functional DNA motif(s) (at least 5-nt long flanking regions on both sides of the motif should be included; totally around 18–30 nt; see Experimental design section for further details of probe design). Prepare stocks (250 pmol ml 1) of the oligos in distilled water. 23| Mix sense and antisense oligos (20 ml each) together. Add 10 oligo annealing buffer (5 ml) and distilled water (5 ml) to top up the reaction volume to 50 ml. 24| Heat to 94 1C for 5 min and gradually cool down the reaction mix to room temperature. m CRITICAL STEP Gradual cooling is beneficial to the annealing process. This could be done by denaturing the mix in a boiler. After denaturation, one can turn off the boiler and let the temperature goes down slowly, which should take at least 60 min. ’ PAUSE POINT The annealed oligonucleotides could be stored at 4 1C or frozen until use. 25| Dilute the annealed oligonucleotides into 25 pmol ml

1

and use 1 ml (i.e., 25 pmol) for labeling reaction.

26| Label the DNA with [g-32P] ATP (50 mCi) by the Ready-To-Go T4 polynucleotide kinase kit according to the manufacturer’s instructions (GE Healthcare Life Sciences). m CRITICAL STEP Other radioactive labeling methods, such as 5¢-end, 3¢-end or internal labeling are also suitable.

NATURE PROTOCOLS | VOL.3 NO.1 | 2008 | 55

PROTOCOL 27| Purify the radioactively labeled DNA probe by a MicroSpin G-25 column to remove the unincorporated radioactive nucleotides. 28| Store the labeled DNA at 4 1C until use. ’ PAUSE POINT The probe should be used as soon as possible or at least within a half-life of 32P (14 d) when stored at 4 1C. For maximum sensitivity, the labeled DNA should be used within 2 d.



Renature the protein extract and probe hybridization TIMING B48 h 29| Place the blot in a small clear plastic box or hybridization chamber.

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

30| Renature the proteins in 5 ml TNED buffer containing 5% skim milk overnight at room temperature with shaking (50 r.p.m.). Protect the blot and buffer from exposure of light. 31| Replace the buffer with 3 ml TNED buffer containing 5% skim milk. Add 20 pmol of 32P-labeled probe (2.0  107 c.p.m.) into the container and shake (50 r.p.m.) the blot overnight at room temperature under light-sealed condition. 32| Wash the membrane three times (15 min each) with TNED buffer at room temperature. After the final wash, seal the membrane by sandwiching it in plastic food wrap.



Autoradiography TIMING 1–3 d 33| To obtain an autoradiogram, insert an X-ray film between the blot (with the gel side face toward the film) in dark room and an intensifying screen and keep the cassette at 80 1C. m CRITICAL STEP The lanes in the blot, in relation to the protein size ladder, could be positioned using phosphorescent ruler, X-ray film marker or tape. 34| Exposure time varies depending upon the specific activity of the probe. Normally exposure time ranges from 1 to 3 d. m CRITICAL STEP Intensifying screen is used to reduce the exposure time. Note that all intensifying screens work optimally at 60 to 80 1C. 35| Develop, fix and dry the film, according to the manufacturer’s or specific laboratory specifications. ? TROUBLESHOOTING



TIMING Steps 1–11, nuclear extract preparation: B2 h Steps 12–18, SDS-PAGE: B3 h Steps 19–21, electroblotting: 1 h 15 min Steps 22–28, preparation of DNA probe: 1.5 h Steps 29–32, renature the protein extract and probe hybridization: B48 h Steps 33–35, autoradiography: 1–3 d ? TROUBLESHOOTING Troubleshooting advice can be found in Table 1. TABLE 1 | Troubleshooting table Problem No bands observed after experiment

Possible reason Poor quality of nuclear extract

Solution Check the protein concentration Confirm the DNA–protein interaction by performing electrophoretic mobility shift assay

Low protein concentration in the extract

Make sure protease inhibitors are added in Steps 3 and 6 Use more cells for nuclear extract isolation Reduce the volume of nuclear extraction buffer in Step 6 Prepare nuclear extract by other methods14

Inefficient labeling or weak radioactivity

Check the labeling efficiency and prepare a new probe Use fresh 32P within the first half-life

Protein not transferred to the membrane

Check electroblotting set-up Extend the transfer time if necessary

56 | VOL.3 NO.1 | 2008 | NATURE PROTOCOLS

PROTOCOL TABLE 1 | Troubleshooting table (continued).

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

Problem

Possible reason Probe is too short

Solution Use a longer probe, but note that too long a probe may cause nonspecific binding

Probe concentration is too dilute

Reduce the volume of TNED buffer in hybridization Increase the amount of probe in the hybridization

The expression level of the target protein is too low

Increase the amount of nuclear extract Use a longer exposure time and use intensifying screen

Nonspecific interaction

Reduce amount of extract and/or probe

Probe is too long

Redesign the probe Use shorter oligos, but not o18 bp

Bands compacted on top of the blot

Size(s) of the target factor(s) is/are large

Use a longer running time in SDS-PAGE Lower the percentage of separating gel in SDS-PAGE

High background

Contamination of membrane

Handle the membrane carefully Wear gloves and clean the containers/chamber before use Filter all solutions

Bubbles trapped during electroblotting

Remove all bubbles trapped in the electroblotting setup

Evaporation of the hybridization buffer

Seal the hybridization chamber/box tightly

TNED buffer with 5% skim milk turned yellow or some precipitate observed

Prepare fresh solution Make sure incubation in light-tight condition

Too long or too short running time

Adjust run time

Poor resolution of the separating gel

Use higher or lower percentage of gel

The color of the protein ladder faded out

Mark the ladder on the blot using pencil after electroblotting

Too many bands are observed

The molecular weight of the band is difficult to determine

ANTICIPATED RESULTS Southwestern blotting is a technique that could provide information on the molecular weight of the protein interacting with specific DNA sequences. With the help of this information, one can dramatically reduce the number of putative transcription factors to be analyzed in a study. Figure 3 shows typical results from Southwestern analysis. Before the experiment shown in Figure 3, synthetic oligonucleotides containing the N1 target motif and a GC-box were designed based on results as suggested in the flow chart (Fig. 2). Southwestern blotting showed two bands on the blot hybridized with the wild-type probe (Fig. 3a), indicating that at least two proteins of sizes B110 and 60 kDa interacted with the probe. To identify the protein specific for the N1 target site, another probe with the N1 motif mutated can be used. After hybridization, only one band with a molecular weight B110 kDa was detected (Fig. 3b), which is consistent with the sizes of GC-box binding Sp proteins (Sp1 and Sp3); thus, it can be established that the N1 target motif interacts specifically with a 60 kDa protein. On the basis of this, the database TransFac can be searched accordingly using the molwSearch utility10. The identity of the transcription factor binding onto the N1 can be then confirmed by in vitro and in vivo supershift and a b ChIP assays, respectively, as well as other functional assays9. M NE BSA M NE 111.4

111.4 79.6

1

79.6

Figure 3 | Southwestern blotting of the N1 motif with SH-SY5Y nuclear extracts. One hundred microgram of nuclear extract from SH-SY5Y cells was separated in a 10% SDS-PAGE. The separated proteins were transferred to Hybond-C extra membrane, which was then incubated with either the (a) wild-type or the (b) mutated probe, and the signals were visualized by autoradiography. The sequences of the probes are shown below. The underlined sequences indicate the locations of the GC-boxes and the N1 target site. The bold-type sequences are the mutated motifs.

61.3

1

61.3

2

49.0 49.0 Wild-type probe

Target site

Mutated probe

GC-box

GC-box

NATURE PROTOCOLS | VOL.3 NO.1 | 2008 | 57

PROTOCOL ACKNOWLEDGMENTS The work was supported by Hong Kong Government RGC Grants HKU7501/05M, HKU7639/07M to B.K.C.C. 7. AUTHOR CONTRIBUTIONS F.K.Y.S. and L.T.O.L. contributed equally to this work, and should be considered as cofirst authors. 8.

© 2007 Nature Publishing Group http://www.nature.com/natureprotocols

Published online at http://www.natureprotocols.com Reprints and permissions information is available online at http://npg.nature.com/ reprintsandpermissions 1. Bowen, B., Steinberg, J., Laemmli, U.K. & Weintraub, H. The detection of DNA-binding proteins by protein blotting. Nucleic Acids Res. 8, 1–20 (1980). 2. Jack, R.S., Gehring, W.J. & Brack, C. Protein component from Drosophila larval nuclei showing sequence specificity for a short region near a major heat-shock protein gene. Cell 24, 321–331 (1981). 3. Anachkova, B. & Russev, G. Differential binding of nonhistone chromosomal proteins to the putative mouse origin of replication. Biochim. Biophys. Acta 740, 369–372 (1983). 4. Triadou, P., Crepin, M., Gros, F. & Lelong, J.C. Tissue-specific binding of total and beta-globin genomic deoxyribonucleic acid to non-histone chromosomal proteins from mouse erythropoietic cells. Biochemistry 21, 6060–6065 (1982). 5. Watt, R.A., Shatzman, A.R. & Rosenberg, M. Expression and characterization of the human c-myc DNA-binding protein. Mol. Cell Biol. 5, 448–456 (1985). 6. Miskimins, W.K., Roberts, M.P., McClelland, A. & Ruddle, F.H. Use of a proteinblotting procedure and a specific DNA probe to identify nuclear proteins that

58 | VOL.3 NO.1 | 2008 | NATURE PROTOCOLS

9.

10.

11.

12.

13. 14.

recognize the promoter region of the transferrin receptor gene. Proc. Natl. Acad. Sci. USA 82, 6741–6744 (1985). Cheng, C.K., Yeung, C.M., Hoo, R.L., Chow, B.K. & Leung, P.C. Oct-1 is involved in the transcriptional repression of the gonadotropin-releasing hormone receptor gene. Endocrinology 143, 4693–4701 (2002). Lee, L.T., Tan-Un, K.C., Lin, M.C. & Chow, B.K. Retinoic acid activates human secretin gene expression by Sp proteins and nuclear factor I in neuronal SH-SY5Y cells. J. Neurochem. 93, 339–350 (2005). Hellman, L.M. & Fried, M.G. Electrophoretic mobility shift assay (EMSA) for detecting protein-nucleic acid interactions. Nat. Protoc. 2, 1849–1861 (2007). Matys, V. et al. TRANSFAC and its module TRANSCompel: transcriptional gene regulation in eukaryotes. Nucleic Acids Res. 34, D108–D110 (2006). Chekmenev, D.S., Haid, C. & Kel, A.E. P-Match: transcription factor binding site search by combining patterns and weight matrices. Nucleic Acids Res. 33, W432–W437 (2005). Bradford, M.M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254 (1976). Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–275 (1951). Lee, K.A., Bindereif, A. & Green, M.R. A small-scale procedure for preparation of nuclear extracts that support efficient transcription and pre-mRNA splicing. Gene Anal. Tech. 5, 22–31 (1988).