CHAPTER FIFTEEN
Standard In Vitro Assays for Protein–Nucleic Acid Interactions – Gel Shift Assays for RNA and DNA Binding Sarah F. Mitchell, Jon R. Lorsch1 Department of Biophysics and Biophysical Chemistry, Johns Hopkins University School of Medicine, Baltimore, MD, USA 1 Corresponding author: e-mail address: [email protected]
Contents 1. Theory 2. Equipment 3. Materials 3.1 Solutions & buffers 4. Protocol 4.1 Preparation 4.2 Duration 4.3 Caution 5. Step 1 Radiolabeling the Nucleic Acid Probe 5.1 Overview 5.2 Duration 5.3 Tip 5.4 Tip 6. Step 2 Bind Protein and Nucleic Acid 6.1 Overview 6.2 Duration 6.3 Tip 6.4 Tip 6.5 Tip 7. Step 3 Preparation of Polyacrylamide Gel 7.1 Overview 7.2 Duration 7.3 Tip 7.4 Tip 7.5 Tip 7.6 Tip
Methods in Enzymology, Volume 541 ISSN 0076-6879 http://dx.doi.org/10.1016/B978-0-12-420119-4.00015-X
180 184 184 185 186 186 186 186 187 187 187 188 188 188 188 188 189 190 190 190 190 190 191 191 191 191
#
2014 Elsevier Inc. All rights reserved.
179
180
Sarah F. Mitchell and Jon R. Lorsch
8. Step 4 Loading and Running Gel 8.1 Overview 8.2 Duration 8.3 Tip 8.4 Tip 9. Step 5 Analysis of Gel 9.1 Overview 9.2 Duration 9.3 Tip 9.4 Tip 9.5 Tip 9.6 Tip 9.7 Tip 9.8 Tip References Source References
191 191 192 192 192 192 192 193 193 194 194 194 194 194 195 196
Abstract The characterization of protein–nucleic acid interactions is necessary for the study of a wide variety of biological processes. One straightforward and widely used approach to this problem is the electrophoretic mobility shift assay (EMSA), in which the binding of a nucleic acid to one or more proteins changes its mobility through a nondenaturing gel matrix. Usually, the mobility of the nucleic acid is reduced, but examples of increased mobility do exist. This type of assay can be used to investigate the affinity of the interaction between the protein and nucleic acid, the specificity of the interaction, the minimal binding site, and the kinetics of the interaction. One particular advantage of EMSA is the ability to analyze multiple proteins, or protein complexes, binding to nucleic acids. This assay is relatively quick and easy and utilizes equipment available in most laboratories; however, there are many variables that can only be determined empirically; therefore, optimization is necessary and can be highly dependent upon the system. The protocol described here is for the poly(A)-binding protein (PABP) binding to an unstructured RNA probe of 43 bases. While this may be a useful protocol for some additional assays, it is recommended that both reaction conditions and gel running conditions be tailored to the individual interaction to be probed.
1. THEORY The mobility of a protein–nucleic acid complex through a gel matrix is determined by several factors: the ratio of the mass of the protein to that of the nucleic acid, the shape of the complex (particularly any conformational change in the nucleic acid), and the charge of the complex. Because binding
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
181
of proteins to free nucleic acid results in a complex of greater mass, as well as a change in the charge to mass ratio in most cases, and occasionally a conformational change, these complexes can frequently be separated from the free nucleic acid using nondenaturing (‘native’) gel electrophoresis. EMSA can be utilized to obtain a great deal of information about the protein– nucleic acid complex; however, this information can only be deemed accurate if the two complexes can be resolved into separate bands and the ratio of these two bands reflects the ratio of free to bound nucleic acids in solution. Fortunately, because of the ability of the gel matrix to keep dissociated proteins trapped in a small area (‘caging’), thus promoting reassociation, the gel matrix reduces dissociation, frequently allowing the amount of bound complex to remain relatively stable in the gel (Lane et al., 1992; Fried, 1989). Obtaining maximal resolution of the bands of free and bound nucleic acids can be challenging and is influenced by a number of variables that depend upon the identity of the protein and nucleic acid used in the assay. Much has been written about optimization of pH, salt concentration, gel matrix, temperature, voltage, cofactors, and nucleic acid probe choice (Fried, 1989; Garner and Revzin, 1986; Kerr, 1995; Lane et al., 1992; Ryder et al., 2008). This protocol will briefly address those subjects, but additional reading may be necessary for optimization. Two types of gel matrix are commonly used in EMSA, acrylamide and agarose gels. Of the two, acrylamide is the more widely used. It is able to better resolve complexes of lower molecular weight. Acrylamide concentrations of 4–10% are commonly found to give good resolution, depending upon the sizes of the nucleic acid and the protein. For very large nucleic acids, proteins or both, a low percentage (0.5–3%) agarose gel may be an option. Composite gels made up of both agarose and polyacrylamide have also been used. The most commonly altered variable in EMSA is the composition of the running buffer (and nonmatrix gel components). Salt concentration is a key factor. Low salt concentration strengthens the ionic interactions that are involved in protein–nucleic acid binding, which is important for reducing dissociation during running. Such dissociation can blur bands. However, low salt concentrations can also lead to an increased nonspecific binding. High salt concentrations, in addition to weakening protein–nucleic acid interactions, impede the progress of the complexes through the gel and increase the heat produced while the gel is running. Addition of divalent cations (such as Mg2þ) often stabilizes complexes involving nucleic acids and can be used in both the binding and running buffers.
182
Sarah F. Mitchell and Jon R. Lorsch
Another key factor to consider is pH. Proteins of neutral or positive charge will reduce the mobility of the nucleic acid upon binding, but strongly negatively charged proteins may bind with little retardation of mobility. Reducing the pH in order to neutralize anionic proteins can increase the separation of free and bound nucleic acids. Additional molecules such as cofactors, detergents (see Explanatory Chapter: Choosing the right detergent), or reducing agents can also stabilize the complex, giving a better-defined band. The power (watts) used when running the gel can be varied in order to reduce the heat produced, as heat can disrupt complexes. Running the gel at a lower temperature, either by cooling the buffer or by running the gel in a cold room, can allow increased power to be used without overheating the gel. Reducing the time of electrophoresis can also minimize dissociation and reduce smearing. In order to reduce nonspecific binding, which can also lead to smearing of bands, additional DNA or RNA is sometimes added in the form of tRNA, salmon sperm DNA, or small polynucleotide oligos (e.g., poly (UC)). Heparin, a charged polymer, may also be added to reduce nonspecific binding, although it can also occasionally disrupt protein–nucleic acid complexes and thus should be used with caution. Reducing the size of the nucleic acid probe to the minimal binding site may reduce nonspecific binding, but the ends of the oligonucleotide might interfere with binding. Shorter nucleic acid probes have the additional advantage of producing a larger shift between the bound and unbound bands. Finally, during quantification of the bands, it may be possible to calculate the fraction bound by comparing the amount of free nucleic acid to the total nucleic acid in the lane, thereby reducing the problems caused by smearing of bands. Some commonly used buffers are 1 TBE (45 mM Tris, 45 mM boric acid, pH 8.3, 1 mM EDTA), 1 TG (25 mM Tris, pH 7.9, 200 mM glycine), 1 TAE (40 mM Tris–acetate, 1 mM EDTA, pH 7.9), and 1 TE (20 mM Tris, 1 mM EDTA). For our studies with ribosomal complexes, 1 THEM (34 mM Tris base, 57 mM Hepes, 0.1 mM EDTA, 2.5 mM MgCl2) has proved to be the best buffer. In order for the ratio of the free to the bound nucleic acid observed in the gel to accurately reflect the ratio in solution, the reaction must run into the wells before a significant amount of dissociation can occur. This can be accomplished in several ways. After loading, the gel may be run at a higher voltage and then turned down after the samples have entered the gel. The use of small sample volumes and wide wells will also lead to quick absorption and compact bands (Fried, 1989). Adding glycerol to 5% final concentration
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
183
or sucrose to 8% will increase the sample density, causing it to fall more quickly to the bottom of the wells. This protocol calls for a radiolabeled nucleic acid; however, gels can also be stained (e.g., with methylene blue, ethidium bromide, or Coomassie blue (see Coomassie Blue Staining)) to visualize either proteins or nucleic acids. Alternatively, the protein can be radiolabeled. Fluorescent groups can be attached to either the protein (see Labeling of a Protein with Fluorophores using Maleimide Derivitization) or the nucleic acid. In general, radiolabeling allows for a low concentration of nucleic acid to be used in the reactions, saving reagents and allowing for better determination of weak dissociation constants. For a further discussion of this, see Adams and Fried (2007). The ability to perform EMSA with small amounts of material and commonly found lab equipment is a great advantage. However, EMSA may not give accurate binding constants for weak interactions (Kd > 3 mM). It may also be difficult to find conditions in which the complexes are stable during running of the gel and in which well-resolved bands are produced. Several alternative methods to study protein–nucleic acid interactions are available. Nitrocellulose filter binding is a simple method to study the binding of proteins and nucleic acids (see Protein Filter Binding). It requires relatively small amounts of materials, but there are concerns that interaction with the filter can perturb the protein’s interaction with the nucleic acid, and a great deal of rinsing is required, during which the complexes can dissociate. Unlike EMSA, filter binding does not differentiate between monomers and oligomers of protein on the nucleic acid, and can therefore be less useful for complicated systems. Fluorescence anisotropy or FRET studies have the advantage of taking place in solution but require fluorescent groups to be placed on one or both binding partners and generally require larger amounts of reagents. This last problem is also true of isothermal titration calorimetry. Footprinting is a technique that has frequently been used to study protein–nucleic acid interactions. In this assay, the protein protects a region of the nucleic acid from digestion or modification, revealing the protein’s binding site. This tool is commonly used in conjunction with EMSA to establish the binding sequence of a protein. EMSA can be performed in a variety of ways to yield different types of information. A simple reaction can determine whether a protein interacts at all with a particular nucleic acid. Varying the concentration of the protein can yield a dissociation constant for the interaction. Cooperative binding can be investigated. Competition between a labeled nucleic acid probe and varying concentrations of unlabeled probes can provide information about specificity and the minimal size of the binding site. Various methods can be used
184
Sarah F. Mitchell and Jon R. Lorsch
to determine stoichiometry of binding (Adams and Fried, 2007). Loading reactions onto a running gel can yield a rate constant of association if reactions are begun by adding the protein to the nucleic acid, or a dissociation rate constant if the reaction is begun by dissociating a complex (usually by adding excess unlabeled nucleic acid). While frequently performed with purified protein, EMSA can also be used with crude lysates when significant amounts of protein are unavailable or the protein of interest has yet to be identified (Varshavsky, 1987).
2. EQUIPMENT Circulating water bath 20 C freezer Microcentrifuge Radioactivity general equipment Power supply Gel dryer Phosphor screen Phosphorimager Sequencing gel apparatus Plates, spacers, and comb for gel 1.5-ml polypropylene tubes Micropipettors Pipettor tips G-50 spin column Syringe Gel loading pipettor tips Whatman 3MM chromatography paper Plastic wrap
3. MATERIALS Protein of interest RNA or DNA probe Calf intestinal phosphatase (CIP) (and reaction buffer) T4 polynucleotide kinase (and buffer) [g-32P]-ATP Phenol Chloroform
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
185
Sodium acetate (NaOAc) Ethanol HEPES Potassium hydroxide (KOH) Potassium acetate (KOAc) Magnesium acetate (MgOAc) Dithiothreitol (DTT) Glycerol Tris base Boric acid (H3BO3) EDTA Sucrose Bromophenol blue Xylene cyanol RNase inhibitor 40% acrylamide/bisacrylamide (19:1) Ammonium persulfate (APS) TEMED
3.1. Solutions & buffers Step 2 10 Reaction buffer Component
Final concentration
Stock
Amount
HepesKOH, pH 7.4
300 mM
1M
300 ml
KOAc
1M
2M
500 ml
MgOAc
30 mM
2M
15 ml
DTT
20 mM
1M
20 ml
Component
Final concentration
Stock
Amount
HepesKOH, pH 7.4
20 mM
1M
20 ml
KOAc
100 mM
2M
50 ml
Glycerol
10%
100%
100 ml
DTT
2 mM
1M
2 ml
Add water to 1 ml
Protein dilution buffer
Add water to 1 ml
186
Sarah F. Mitchell and Jon R. Lorsch
Step 3 10 TBE Component
Final concentration
Stock
Tris base
0.9 M
108 g
Boric acid
0.9 M
55 g
EDTA, pH 8.0
20 mM
0.5 M
Amount
40 ml
Add water to 1 l
Gel loading dye Component
Final concentration
Stock
Amount
Sucrose
50%
50%
1 ml
Bromophenol blue
0.02%
0.2 mg
Xylene cyanol
0.02%
0.2 mg
Add water to 1 ml
4. PROTOCOL 4.1. Preparation If using a purified protein, purify the protein as close to homogeneity (see Explanatory Chapter: Troubleshooting protein expression: what to do when the protein is not soluble) as possible. Verify that the preparation is free of nucleases. If using a cell lysate, prepare lysate as appropriate (see Lysis of mammalian and Sf9 cells). Obtain the RNA or DNA probe of interest by purchasing, synthesizing, using PCR, or transcribing it (see in vitro Transcription from Plasmid or PCR amplified DNA) and purify it as necessary (see Purification of DNA Oligos by Denaturing Polyacrylamide Gel Electrophoresis (PAGE)).
4.2. Duration Preparation
Varies with system
Protocol
About 2 days
4.3. Caution Consult your institute’s Radiation Safety Officer for proper ordering, handling, and disposal of radioactive materials.
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
187
Figure 15.1 Flowchart of the complete protocol, including preparation.
See Fig. 15.1 for the flowchart of the complete protocol, including preparation.
5. STEP 1 RADIOLABELING THE NUCLEIC ACID PROBE 5.1. Overview Label the 50 end of the DNA or RNA probe using T4 polynucleotide kinase (see RNA Radiolabeling)
5.2. Duration 5 h 1.1 If your probe has a 50 phosphate group, it will be necessary to remove it by treating the nucleic acid with phosphatase. Dephosphorylate 50 pmol of nucleic acid probe in a 25-ml reaction according to the supplier’s instructions. Heat-inactivate the enzyme before proceeding with the phosphorylation reaction. 1.2 Phosphorylate the probe using [g-32P]-ATP and polynucleotide kinase. Set up the following reaction:
188
Sarah F. Mitchell and Jon R. Lorsch
25 ml nucleic acid (1–50 pmol) 50 pmol [g-32P]-ATP 5 ml 10 PNK buffer 20 units T4 polynucleotide kinase ddH2O to 50 ml Incubate at 37 C for 1 h. 1.3 Pass the phosphorylated nucleic acid over a G-50 spin column to remove unincorporated ATP, extract with phenol and chloroform, and ethanol precipitate the nucleic acid. Alternatively, if the nucleic acid is large (>200 nucleotides), it can be separated from free ATP on an agarose gel and then extracted from the gel.
5.3. Tip Calf intestinal phosphatase may not be completely heat-inactivated. If it is not, purify the probe using phenol and chloroform extraction followed by ethanol precipitation or from a gel.
5.4. Tip Whenever working with radioactive reagents, be careful to avoid contamination of pipettors, gloves, and bench. Wear gloves and a lab coat while handling radioactivity. Frequently check the work area with a Geiger counter and clean any contamination. Keep radioactivity behind an appropriate shield. When finished with the protocol, check all equiptment, bench space, and nearby floors for any contamination. See Fig. 15.2 for the flowchart of Step 1.
6. STEP 2 BIND PROTEIN AND NUCLEIC ACID 6.1. Overview In this step, you will incubate your protein and nucleic acid of interest together in the appropriate buffer to allow them to bind.
6.2. Duration 30 min 2.1 As this probe is unstructured, no annealing step is required. If your probe is structured or a duplex, heat the probe at 95 C for 5 min and then slowly cool to allow it to properly fold or anneal the two strands (reduce temperature by 0.1 C s1 to 4 C, then store on ice until use).
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
189
Figure 15.2 Flowchart of Step 1.
2.2 Make a master mix. For each reaction, add to a 1.5-ml tube: 1.2 ml 10 reaction buffer 1.2 ml 100 nM 32P-labeled probe 1.2 ml of 1 U ml1 RNase inhibitor 7.2 ml of ddH2O Aliquot the mix into the number of tubes needed for your experiment (protein concentration points, time points, etc.). In this example, PABP is being titrated between 1 nM and 1 mM and ten points will be observed. 2.3 Add 1.2 ml of a 10 stock of protein, in this case, PABP. The protein is diluted to 10 for each point by mixing with the appropriate amount of Protein Dilution Buffer. 2.4 Incubate reactions at 26 C for 30 min.
6.3. Tip Divalent cations (i.e., Mg2þ) may help with folding, but avoid incubating RNA at high temperatures with Mg2þ as this can lead to hydrolysis.
190
Sarah F. Mitchell and Jon R. Lorsch
6.4. Tip When titrating the protein to find equilibrium binding constants, it is generally recommended to use a 1000-fold range of concentrations. Taking points at half-log intervals is a good starting point.
6.5. Tip The assay can be carried out as an equilibrium binding assay or a kinetic assay. The incubation time will vary depending on the experiment. See Fig. 15.3 for the flowchart of Step 2.
7. STEP 3 PREPARATION OF POLYACRYLAMIDE GEL 7.1. Overview In this step, you will pour your gel, set it up in the gel running apparatus, clean the wells and prerun the gel.
7.2. Duration 1.5 h 3.1 Mix an 8% acrylamide gel in 1 TBE buffer. For the 10 cm by 10 cm 0.75 mm thick gels used here, 10 ml of acrylamide solution is more than adequate.
Figure 15.3 Flowchart of Step 2.
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
Combine:
191
1 ml of 10 TBE 2 ml of 40% acrylamide:bisacrylamide (19:1) 6.9 ml of ddH2O
When ready to pour the gel, add 100 ml of 10% APS and 10 ml of TEMED. Mix and pour into the prepared gel plates. Insert a 10-well comb before it polymerizes. 3.2 Attach the gel to the gel running apparatus. Fill the reservoirs with 1 TBE. 3.3 Using a syringe or pipettor, clean the wells to remove bits of gel and unpolymerized acrylamide. 3.4 Prerun the gel for 30 min at 200 V, with cooling to 20 C, to bring it to its running temperature.
7.3. Tip Acrylamide is a neurotoxin. Take care to avoid touching or inhaling it. Working with acrylamide in solution rather than powder form reduces the risk of inhalation.
7.4. Tip The amount of bisacrylamide (the cross-linking reagent) can be varied in order to change the pore size of the gel. Large complexes will run into the gel more easily with less bisacrylamide (i.e., a higher acrylamide:bisacrylamide ratio).
7.5. Tip To assemble the gel plates, place the spacers between the two plates, tape them together, and pour the gel at an angle. Alternatively, you can lay the untaped assembled gel plates nearly flat and pour slowly across the top of the gel while the air exits the bottom.
7.6. Tip It may not be necessary to prerun the gel in all systems. See Fig. 15.4 for the flowchart of Step 3.
8. STEP 4 LOADING AND RUNNING GEL 8.1. Overview In this step, you will mix your binding reactions with a loading dye, load them onto the gel, and run the gel for an appropriate amount of time.
192
Sarah F. Mitchell and Jon R. Lorsch
Figure 15.4 Flowchart of Step 3.
8.2. Duration 1h 4.1 Add 10 ml of reaction to 2 ml of 6 gel loading dye. 4.2 Load your reaction into a well on the gel. A flat-tipped gel loading pipette tip often makes loading easier. 4.3 Run the gel at 200 V for 45 min, with cooling to 20 C.
8.3. Tip If loading a running gel, avoid putting your fingers into the buffer as this could result in electrocution.
8.4. Tip Cool the gel by running it in the cold room or using an attached circulating, refrigerated water bath. This may help reduce smearing of bands. See Fig. 15.5 for the flowchart of Step 4.
9. STEP 5 ANALYSIS OF GEL 9.1. Overview After running the gel, expose it to a phosphor screen. The screen is then scanned, the bands quantified, and data are fit to the appropriate equation.
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
193
Figure 15.5 Flowchart of Step 4.
9.2. Duration Overnight 5.1 Remove the gel from the glass plates by carefully prying them open with a wedge or spatula. 5.2 Place the gel with one side against plastic wrap, the other side against Whatman 3MM paper. 5.3 Place on gel dryer, filter paper side down. Heat at 80 C, under vacuum, until dry, usually 1–2 h. Allow dryer to cool to room temperature before breaking the vacuum to prevent shattering of the dried gel. 5.4 Wrap the gel in plastic wrap. Make sure the plastic wrap is completely dry. 5.5 Expose the gel to a phosphor screen overnight. 5.6 Scan the screen using a phosphorimager. Quantify the counts and calculate the fraction bound in each lane. 5.7 Plot the fraction bound against the appropriate value (protein concentration, competitor concentration, time, etc.) and use a least squares fitting program to fit the data.
9.3. Tip Opening the gel plates and removing the gel can be challenging depending upon the percentage of acrylamide in the gel. Low percentage gels can easily stretch or deform. It is possible for the gel to stick to both plates in different regions, resulting in the gel folding over onto itself possibly wrinkling or tearing. This risk can be minimized by siliconizing one plate (the plate that you lift off of the gel) and roughing the opposite plate using steel wool. One should always begin prying the plates apart from a bottom corner and proceeding slowly toward the opposite side. A razor blade can be useful in detaching the gel from the upper plate if it does stick. If the gel wrinkles, it can
194
Sarah F. Mitchell and Jon R. Lorsch
sometimes be flattened out by squirting water underneath the wrinkle and gently smoothing it with your hands (while wearing gloves, of course).
9.4. Tip Low percentage acrylamide gels (
Standard In Vitro Assays for Protein–Nucleic Acid Interactions – Gel Shift Assays for RNA and DNA Binding Sarah F. Mitchell, Jon R. Lorsch1 Department of Biophysics and Biophysical Chemistry, Johns Hopkins University School of Medicine, Baltimore, MD, USA 1 Corresponding author: e-mail address: [email protected]
Contents 1. Theory 2. Equipment 3. Materials 3.1 Solutions & buffers 4. Protocol 4.1 Preparation 4.2 Duration 4.3 Caution 5. Step 1 Radiolabeling the Nucleic Acid Probe 5.1 Overview 5.2 Duration 5.3 Tip 5.4 Tip 6. Step 2 Bind Protein and Nucleic Acid 6.1 Overview 6.2 Duration 6.3 Tip 6.4 Tip 6.5 Tip 7. Step 3 Preparation of Polyacrylamide Gel 7.1 Overview 7.2 Duration 7.3 Tip 7.4 Tip 7.5 Tip 7.6 Tip
Methods in Enzymology, Volume 541 ISSN 0076-6879 http://dx.doi.org/10.1016/B978-0-12-420119-4.00015-X
180 184 184 185 186 186 186 186 187 187 187 188 188 188 188 188 189 190 190 190 190 190 191 191 191 191
#
2014 Elsevier Inc. All rights reserved.
179
180
Sarah F. Mitchell and Jon R. Lorsch
8. Step 4 Loading and Running Gel 8.1 Overview 8.2 Duration 8.3 Tip 8.4 Tip 9. Step 5 Analysis of Gel 9.1 Overview 9.2 Duration 9.3 Tip 9.4 Tip 9.5 Tip 9.6 Tip 9.7 Tip 9.8 Tip References Source References
191 191 192 192 192 192 192 193 193 194 194 194 194 194 195 196
Abstract The characterization of protein–nucleic acid interactions is necessary for the study of a wide variety of biological processes. One straightforward and widely used approach to this problem is the electrophoretic mobility shift assay (EMSA), in which the binding of a nucleic acid to one or more proteins changes its mobility through a nondenaturing gel matrix. Usually, the mobility of the nucleic acid is reduced, but examples of increased mobility do exist. This type of assay can be used to investigate the affinity of the interaction between the protein and nucleic acid, the specificity of the interaction, the minimal binding site, and the kinetics of the interaction. One particular advantage of EMSA is the ability to analyze multiple proteins, or protein complexes, binding to nucleic acids. This assay is relatively quick and easy and utilizes equipment available in most laboratories; however, there are many variables that can only be determined empirically; therefore, optimization is necessary and can be highly dependent upon the system. The protocol described here is for the poly(A)-binding protein (PABP) binding to an unstructured RNA probe of 43 bases. While this may be a useful protocol for some additional assays, it is recommended that both reaction conditions and gel running conditions be tailored to the individual interaction to be probed.
1. THEORY The mobility of a protein–nucleic acid complex through a gel matrix is determined by several factors: the ratio of the mass of the protein to that of the nucleic acid, the shape of the complex (particularly any conformational change in the nucleic acid), and the charge of the complex. Because binding
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
181
of proteins to free nucleic acid results in a complex of greater mass, as well as a change in the charge to mass ratio in most cases, and occasionally a conformational change, these complexes can frequently be separated from the free nucleic acid using nondenaturing (‘native’) gel electrophoresis. EMSA can be utilized to obtain a great deal of information about the protein– nucleic acid complex; however, this information can only be deemed accurate if the two complexes can be resolved into separate bands and the ratio of these two bands reflects the ratio of free to bound nucleic acids in solution. Fortunately, because of the ability of the gel matrix to keep dissociated proteins trapped in a small area (‘caging’), thus promoting reassociation, the gel matrix reduces dissociation, frequently allowing the amount of bound complex to remain relatively stable in the gel (Lane et al., 1992; Fried, 1989). Obtaining maximal resolution of the bands of free and bound nucleic acids can be challenging and is influenced by a number of variables that depend upon the identity of the protein and nucleic acid used in the assay. Much has been written about optimization of pH, salt concentration, gel matrix, temperature, voltage, cofactors, and nucleic acid probe choice (Fried, 1989; Garner and Revzin, 1986; Kerr, 1995; Lane et al., 1992; Ryder et al., 2008). This protocol will briefly address those subjects, but additional reading may be necessary for optimization. Two types of gel matrix are commonly used in EMSA, acrylamide and agarose gels. Of the two, acrylamide is the more widely used. It is able to better resolve complexes of lower molecular weight. Acrylamide concentrations of 4–10% are commonly found to give good resolution, depending upon the sizes of the nucleic acid and the protein. For very large nucleic acids, proteins or both, a low percentage (0.5–3%) agarose gel may be an option. Composite gels made up of both agarose and polyacrylamide have also been used. The most commonly altered variable in EMSA is the composition of the running buffer (and nonmatrix gel components). Salt concentration is a key factor. Low salt concentration strengthens the ionic interactions that are involved in protein–nucleic acid binding, which is important for reducing dissociation during running. Such dissociation can blur bands. However, low salt concentrations can also lead to an increased nonspecific binding. High salt concentrations, in addition to weakening protein–nucleic acid interactions, impede the progress of the complexes through the gel and increase the heat produced while the gel is running. Addition of divalent cations (such as Mg2þ) often stabilizes complexes involving nucleic acids and can be used in both the binding and running buffers.
182
Sarah F. Mitchell and Jon R. Lorsch
Another key factor to consider is pH. Proteins of neutral or positive charge will reduce the mobility of the nucleic acid upon binding, but strongly negatively charged proteins may bind with little retardation of mobility. Reducing the pH in order to neutralize anionic proteins can increase the separation of free and bound nucleic acids. Additional molecules such as cofactors, detergents (see Explanatory Chapter: Choosing the right detergent), or reducing agents can also stabilize the complex, giving a better-defined band. The power (watts) used when running the gel can be varied in order to reduce the heat produced, as heat can disrupt complexes. Running the gel at a lower temperature, either by cooling the buffer or by running the gel in a cold room, can allow increased power to be used without overheating the gel. Reducing the time of electrophoresis can also minimize dissociation and reduce smearing. In order to reduce nonspecific binding, which can also lead to smearing of bands, additional DNA or RNA is sometimes added in the form of tRNA, salmon sperm DNA, or small polynucleotide oligos (e.g., poly (UC)). Heparin, a charged polymer, may also be added to reduce nonspecific binding, although it can also occasionally disrupt protein–nucleic acid complexes and thus should be used with caution. Reducing the size of the nucleic acid probe to the minimal binding site may reduce nonspecific binding, but the ends of the oligonucleotide might interfere with binding. Shorter nucleic acid probes have the additional advantage of producing a larger shift between the bound and unbound bands. Finally, during quantification of the bands, it may be possible to calculate the fraction bound by comparing the amount of free nucleic acid to the total nucleic acid in the lane, thereby reducing the problems caused by smearing of bands. Some commonly used buffers are 1 TBE (45 mM Tris, 45 mM boric acid, pH 8.3, 1 mM EDTA), 1 TG (25 mM Tris, pH 7.9, 200 mM glycine), 1 TAE (40 mM Tris–acetate, 1 mM EDTA, pH 7.9), and 1 TE (20 mM Tris, 1 mM EDTA). For our studies with ribosomal complexes, 1 THEM (34 mM Tris base, 57 mM Hepes, 0.1 mM EDTA, 2.5 mM MgCl2) has proved to be the best buffer. In order for the ratio of the free to the bound nucleic acid observed in the gel to accurately reflect the ratio in solution, the reaction must run into the wells before a significant amount of dissociation can occur. This can be accomplished in several ways. After loading, the gel may be run at a higher voltage and then turned down after the samples have entered the gel. The use of small sample volumes and wide wells will also lead to quick absorption and compact bands (Fried, 1989). Adding glycerol to 5% final concentration
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
183
or sucrose to 8% will increase the sample density, causing it to fall more quickly to the bottom of the wells. This protocol calls for a radiolabeled nucleic acid; however, gels can also be stained (e.g., with methylene blue, ethidium bromide, or Coomassie blue (see Coomassie Blue Staining)) to visualize either proteins or nucleic acids. Alternatively, the protein can be radiolabeled. Fluorescent groups can be attached to either the protein (see Labeling of a Protein with Fluorophores using Maleimide Derivitization) or the nucleic acid. In general, radiolabeling allows for a low concentration of nucleic acid to be used in the reactions, saving reagents and allowing for better determination of weak dissociation constants. For a further discussion of this, see Adams and Fried (2007). The ability to perform EMSA with small amounts of material and commonly found lab equipment is a great advantage. However, EMSA may not give accurate binding constants for weak interactions (Kd > 3 mM). It may also be difficult to find conditions in which the complexes are stable during running of the gel and in which well-resolved bands are produced. Several alternative methods to study protein–nucleic acid interactions are available. Nitrocellulose filter binding is a simple method to study the binding of proteins and nucleic acids (see Protein Filter Binding). It requires relatively small amounts of materials, but there are concerns that interaction with the filter can perturb the protein’s interaction with the nucleic acid, and a great deal of rinsing is required, during which the complexes can dissociate. Unlike EMSA, filter binding does not differentiate between monomers and oligomers of protein on the nucleic acid, and can therefore be less useful for complicated systems. Fluorescence anisotropy or FRET studies have the advantage of taking place in solution but require fluorescent groups to be placed on one or both binding partners and generally require larger amounts of reagents. This last problem is also true of isothermal titration calorimetry. Footprinting is a technique that has frequently been used to study protein–nucleic acid interactions. In this assay, the protein protects a region of the nucleic acid from digestion or modification, revealing the protein’s binding site. This tool is commonly used in conjunction with EMSA to establish the binding sequence of a protein. EMSA can be performed in a variety of ways to yield different types of information. A simple reaction can determine whether a protein interacts at all with a particular nucleic acid. Varying the concentration of the protein can yield a dissociation constant for the interaction. Cooperative binding can be investigated. Competition between a labeled nucleic acid probe and varying concentrations of unlabeled probes can provide information about specificity and the minimal size of the binding site. Various methods can be used
184
Sarah F. Mitchell and Jon R. Lorsch
to determine stoichiometry of binding (Adams and Fried, 2007). Loading reactions onto a running gel can yield a rate constant of association if reactions are begun by adding the protein to the nucleic acid, or a dissociation rate constant if the reaction is begun by dissociating a complex (usually by adding excess unlabeled nucleic acid). While frequently performed with purified protein, EMSA can also be used with crude lysates when significant amounts of protein are unavailable or the protein of interest has yet to be identified (Varshavsky, 1987).
2. EQUIPMENT Circulating water bath 20 C freezer Microcentrifuge Radioactivity general equipment Power supply Gel dryer Phosphor screen Phosphorimager Sequencing gel apparatus Plates, spacers, and comb for gel 1.5-ml polypropylene tubes Micropipettors Pipettor tips G-50 spin column Syringe Gel loading pipettor tips Whatman 3MM chromatography paper Plastic wrap
3. MATERIALS Protein of interest RNA or DNA probe Calf intestinal phosphatase (CIP) (and reaction buffer) T4 polynucleotide kinase (and buffer) [g-32P]-ATP Phenol Chloroform
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
185
Sodium acetate (NaOAc) Ethanol HEPES Potassium hydroxide (KOH) Potassium acetate (KOAc) Magnesium acetate (MgOAc) Dithiothreitol (DTT) Glycerol Tris base Boric acid (H3BO3) EDTA Sucrose Bromophenol blue Xylene cyanol RNase inhibitor 40% acrylamide/bisacrylamide (19:1) Ammonium persulfate (APS) TEMED
3.1. Solutions & buffers Step 2 10 Reaction buffer Component
Final concentration
Stock
Amount
HepesKOH, pH 7.4
300 mM
1M
300 ml
KOAc
1M
2M
500 ml
MgOAc
30 mM
2M
15 ml
DTT
20 mM
1M
20 ml
Component
Final concentration
Stock
Amount
HepesKOH, pH 7.4
20 mM
1M
20 ml
KOAc
100 mM
2M
50 ml
Glycerol
10%
100%
100 ml
DTT
2 mM
1M
2 ml
Add water to 1 ml
Protein dilution buffer
Add water to 1 ml
186
Sarah F. Mitchell and Jon R. Lorsch
Step 3 10 TBE Component
Final concentration
Stock
Tris base
0.9 M
108 g
Boric acid
0.9 M
55 g
EDTA, pH 8.0
20 mM
0.5 M
Amount
40 ml
Add water to 1 l
Gel loading dye Component
Final concentration
Stock
Amount
Sucrose
50%
50%
1 ml
Bromophenol blue
0.02%
0.2 mg
Xylene cyanol
0.02%
0.2 mg
Add water to 1 ml
4. PROTOCOL 4.1. Preparation If using a purified protein, purify the protein as close to homogeneity (see Explanatory Chapter: Troubleshooting protein expression: what to do when the protein is not soluble) as possible. Verify that the preparation is free of nucleases. If using a cell lysate, prepare lysate as appropriate (see Lysis of mammalian and Sf9 cells). Obtain the RNA or DNA probe of interest by purchasing, synthesizing, using PCR, or transcribing it (see in vitro Transcription from Plasmid or PCR amplified DNA) and purify it as necessary (see Purification of DNA Oligos by Denaturing Polyacrylamide Gel Electrophoresis (PAGE)).
4.2. Duration Preparation
Varies with system
Protocol
About 2 days
4.3. Caution Consult your institute’s Radiation Safety Officer for proper ordering, handling, and disposal of radioactive materials.
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
187
Figure 15.1 Flowchart of the complete protocol, including preparation.
See Fig. 15.1 for the flowchart of the complete protocol, including preparation.
5. STEP 1 RADIOLABELING THE NUCLEIC ACID PROBE 5.1. Overview Label the 50 end of the DNA or RNA probe using T4 polynucleotide kinase (see RNA Radiolabeling)
5.2. Duration 5 h 1.1 If your probe has a 50 phosphate group, it will be necessary to remove it by treating the nucleic acid with phosphatase. Dephosphorylate 50 pmol of nucleic acid probe in a 25-ml reaction according to the supplier’s instructions. Heat-inactivate the enzyme before proceeding with the phosphorylation reaction. 1.2 Phosphorylate the probe using [g-32P]-ATP and polynucleotide kinase. Set up the following reaction:
188
Sarah F. Mitchell and Jon R. Lorsch
25 ml nucleic acid (1–50 pmol) 50 pmol [g-32P]-ATP 5 ml 10 PNK buffer 20 units T4 polynucleotide kinase ddH2O to 50 ml Incubate at 37 C for 1 h. 1.3 Pass the phosphorylated nucleic acid over a G-50 spin column to remove unincorporated ATP, extract with phenol and chloroform, and ethanol precipitate the nucleic acid. Alternatively, if the nucleic acid is large (>200 nucleotides), it can be separated from free ATP on an agarose gel and then extracted from the gel.
5.3. Tip Calf intestinal phosphatase may not be completely heat-inactivated. If it is not, purify the probe using phenol and chloroform extraction followed by ethanol precipitation or from a gel.
5.4. Tip Whenever working with radioactive reagents, be careful to avoid contamination of pipettors, gloves, and bench. Wear gloves and a lab coat while handling radioactivity. Frequently check the work area with a Geiger counter and clean any contamination. Keep radioactivity behind an appropriate shield. When finished with the protocol, check all equiptment, bench space, and nearby floors for any contamination. See Fig. 15.2 for the flowchart of Step 1.
6. STEP 2 BIND PROTEIN AND NUCLEIC ACID 6.1. Overview In this step, you will incubate your protein and nucleic acid of interest together in the appropriate buffer to allow them to bind.
6.2. Duration 30 min 2.1 As this probe is unstructured, no annealing step is required. If your probe is structured or a duplex, heat the probe at 95 C for 5 min and then slowly cool to allow it to properly fold or anneal the two strands (reduce temperature by 0.1 C s1 to 4 C, then store on ice until use).
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
189
Figure 15.2 Flowchart of Step 1.
2.2 Make a master mix. For each reaction, add to a 1.5-ml tube: 1.2 ml 10 reaction buffer 1.2 ml 100 nM 32P-labeled probe 1.2 ml of 1 U ml1 RNase inhibitor 7.2 ml of ddH2O Aliquot the mix into the number of tubes needed for your experiment (protein concentration points, time points, etc.). In this example, PABP is being titrated between 1 nM and 1 mM and ten points will be observed. 2.3 Add 1.2 ml of a 10 stock of protein, in this case, PABP. The protein is diluted to 10 for each point by mixing with the appropriate amount of Protein Dilution Buffer. 2.4 Incubate reactions at 26 C for 30 min.
6.3. Tip Divalent cations (i.e., Mg2þ) may help with folding, but avoid incubating RNA at high temperatures with Mg2þ as this can lead to hydrolysis.
190
Sarah F. Mitchell and Jon R. Lorsch
6.4. Tip When titrating the protein to find equilibrium binding constants, it is generally recommended to use a 1000-fold range of concentrations. Taking points at half-log intervals is a good starting point.
6.5. Tip The assay can be carried out as an equilibrium binding assay or a kinetic assay. The incubation time will vary depending on the experiment. See Fig. 15.3 for the flowchart of Step 2.
7. STEP 3 PREPARATION OF POLYACRYLAMIDE GEL 7.1. Overview In this step, you will pour your gel, set it up in the gel running apparatus, clean the wells and prerun the gel.
7.2. Duration 1.5 h 3.1 Mix an 8% acrylamide gel in 1 TBE buffer. For the 10 cm by 10 cm 0.75 mm thick gels used here, 10 ml of acrylamide solution is more than adequate.
Figure 15.3 Flowchart of Step 2.
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
Combine:
191
1 ml of 10 TBE 2 ml of 40% acrylamide:bisacrylamide (19:1) 6.9 ml of ddH2O
When ready to pour the gel, add 100 ml of 10% APS and 10 ml of TEMED. Mix and pour into the prepared gel plates. Insert a 10-well comb before it polymerizes. 3.2 Attach the gel to the gel running apparatus. Fill the reservoirs with 1 TBE. 3.3 Using a syringe or pipettor, clean the wells to remove bits of gel and unpolymerized acrylamide. 3.4 Prerun the gel for 30 min at 200 V, with cooling to 20 C, to bring it to its running temperature.
7.3. Tip Acrylamide is a neurotoxin. Take care to avoid touching or inhaling it. Working with acrylamide in solution rather than powder form reduces the risk of inhalation.
7.4. Tip The amount of bisacrylamide (the cross-linking reagent) can be varied in order to change the pore size of the gel. Large complexes will run into the gel more easily with less bisacrylamide (i.e., a higher acrylamide:bisacrylamide ratio).
7.5. Tip To assemble the gel plates, place the spacers between the two plates, tape them together, and pour the gel at an angle. Alternatively, you can lay the untaped assembled gel plates nearly flat and pour slowly across the top of the gel while the air exits the bottom.
7.6. Tip It may not be necessary to prerun the gel in all systems. See Fig. 15.4 for the flowchart of Step 3.
8. STEP 4 LOADING AND RUNNING GEL 8.1. Overview In this step, you will mix your binding reactions with a loading dye, load them onto the gel, and run the gel for an appropriate amount of time.
192
Sarah F. Mitchell and Jon R. Lorsch
Figure 15.4 Flowchart of Step 3.
8.2. Duration 1h 4.1 Add 10 ml of reaction to 2 ml of 6 gel loading dye. 4.2 Load your reaction into a well on the gel. A flat-tipped gel loading pipette tip often makes loading easier. 4.3 Run the gel at 200 V for 45 min, with cooling to 20 C.
8.3. Tip If loading a running gel, avoid putting your fingers into the buffer as this could result in electrocution.
8.4. Tip Cool the gel by running it in the cold room or using an attached circulating, refrigerated water bath. This may help reduce smearing of bands. See Fig. 15.5 for the flowchart of Step 4.
9. STEP 5 ANALYSIS OF GEL 9.1. Overview After running the gel, expose it to a phosphor screen. The screen is then scanned, the bands quantified, and data are fit to the appropriate equation.
Standard In Vitro Assays for Protein–Nucleic Acid Interactions
193
Figure 15.5 Flowchart of Step 4.
9.2. Duration Overnight 5.1 Remove the gel from the glass plates by carefully prying them open with a wedge or spatula. 5.2 Place the gel with one side against plastic wrap, the other side against Whatman 3MM paper. 5.3 Place on gel dryer, filter paper side down. Heat at 80 C, under vacuum, until dry, usually 1–2 h. Allow dryer to cool to room temperature before breaking the vacuum to prevent shattering of the dried gel. 5.4 Wrap the gel in plastic wrap. Make sure the plastic wrap is completely dry. 5.5 Expose the gel to a phosphor screen overnight. 5.6 Scan the screen using a phosphorimager. Quantify the counts and calculate the fraction bound in each lane. 5.7 Plot the fraction bound against the appropriate value (protein concentration, competitor concentration, time, etc.) and use a least squares fitting program to fit the data.
9.3. Tip Opening the gel plates and removing the gel can be challenging depending upon the percentage of acrylamide in the gel. Low percentage gels can easily stretch or deform. It is possible for the gel to stick to both plates in different regions, resulting in the gel folding over onto itself possibly wrinkling or tearing. This risk can be minimized by siliconizing one plate (the plate that you lift off of the gel) and roughing the opposite plate using steel wool. One should always begin prying the plates apart from a bottom corner and proceeding slowly toward the opposite side. A razor blade can be useful in detaching the gel from the upper plate if it does stick. If the gel wrinkles, it can
194
Sarah F. Mitchell and Jon R. Lorsch
sometimes be flattened out by squirting water underneath the wrinkle and gently smoothing it with your hands (while wearing gloves, of course).
9.4. Tip Low percentage acrylamide gels (
