Chapter 21 Microscale Thermophoresis for the Assessment of Nuclear Protein-Binding Affinities Wei Zhang, Stefan Duhr, Philipp Baaske, and Ernest Laue Abstract The rapid advance in our knowledge of cellular regulatory mechanisms, including those involving chromatin-based processes, stems in part from the development of biophysical techniques such as fluorescence spectroscopy, surface plasmon resonance (SPR), and isothermal titration calorimetry (ITC). Despite their widespread utility, each of these techniques has its pros and cons, and new techniques are still required. Here we describe the application of microscale thermophoresis (MST), a novel technique based on thermophoresis, to characterize the binding between histone peptides and a histone chaperone protein, in free solution, with high sensitivity and low sample consumption. Key words Microscale thermophoresis, Protein-peptide interaction, Histone peptide, Histone chaperone, Binding affinity, Capillary, Fluorescence spectroscopy
1
Introduction Microscale thermophoresis (MST) is a new method that enables the quantitative analysis of molecular interactions in free solution using only a few microliters of sample. The technique is based on thermophoresis—the directed motion of molecules along temperature gradients. Thermophoresis is highly sensitive to binding-induced changes of molecular properties, be they changes in size, charge, hydration shell, or conformation. Thermophoresis is an optical technique, where an infrared laser is used to locally heat part of the sample, following which the molecules mobility in the resulting temperature gradient is then analyzed via the monitoring of their fluorescence [1, 2]. In addition to fluorescent labeling of one of the molecules involved, either directly or by expressing a fluorescent fusion protein, intrinsic protein fluorescence can also be utilized for MST, thus allowing for label-free MST analysis [3]. In contrast to other techniques, the interaction of small unlabeled peptides with a larger fluorescent protein can be readily studied by MST, despite the small change in molecular weight involved. The type of buffer
Juan C. Stockert et al. (eds.), Functional Analysis of DNA and Chromatin, Methods in Molecular Biology, vol. 1094, DOI 10.1007/978-1-62703-706-8_21, © Springer Science+Business Media New York 2014
269
270
Wei Zhang et al.
and additives can also be chosen freely. Measurement is even possible in complex bioliquids, such as cell lysates, and thus under close to in vivo conditions without sample purification. In eukaryotic cells, DNA is packaged with proteins to form higher order structures called chromatin. The basic structural unit of chromatin is the nucleosome, which is formed from around 146 base pairs (bp) of DNA wrapped over a histone octamer in 1.65 left-handed helical turns. The histone octamer comprises two copies of each of histones H2A, H2B, H3, and H4 with their unstructured N-terminal tails protruding out from the octameric core [4]. These tails are subjected to various posttranslational modifications (PTMs) such as acetylation, methylation, phosphorylation, and ubiquitination by various enzymes. These modified sites can be utilized as specific codes for either recruiting new effectors or preventing particular interactions in the large number of complex regulatory processes that occur in the cell [5, 6]. During the cell cycle nucleosomes need to be assembled and disassembled, and it is thought that during the cell cycle histone proteins almost never appear freely on their own. Instead they exist in complexes with histone chaperones, proteins that play a key role in nucleosome assembly/disassembly, histone posttranslational modifications, and ATP-dependent nucleosome remodeling [7]. The human RbAp46/RbAp48 (retinoblastoma-associated proteins p46/48) proteins are histone chaperones that are found in a large number of chromatin complexes, where they are thought to function as histone-binding subunits [8]. Analysis of their interactions with histones and histone peptides is therefore of great interest. Although there are several techniques available to assess protein-peptide interactions, they either need considerably larger sample quantities, specifically synthesized labeled peptides, or require studies over a solid surface. For a variety of reasons this is not always convenient or desirable. MST offers a new opportunity to perform studies of p48 histone-peptide interactions without the need for specific labeling and with very small amounts of sample. Here, we show how to carry out the analysis of randomly labeled lysine residues in either the p48 protein (48 kDa) or the histone peptide at low density with a fluorescent dye.
2
Materials
2.1 Peptides and Protein
1. Human histone H4 (amino acid 1–41) peptide was chemically synthesized by Alta Bioscience (http://www.altabioscience.com/). 2. Human histone H3 (amino acid 1–59) peptide was prepared according to Richart et al. [9]. 3. Human p48 protein was prepared according to Lejon et al. [10] (see Note 1).
Microscale Thermophoresis for Histone Studies
2.2 Microscale Thermophoresis Instrument and Reagents
271
(For more details, please see http://www.nanotemper-technologies. com//home/) 1. NanoTemper Monolith NT.115. 2. Dye NT-647: Monolith NT™ Protein Labeling Kit RED-NHS (Cat# L001). 3. Capillaries: NT.115 standard (Cat# K002) and hydrophilic treated (Cat# K004) capillaries. 4. Sephadex G25 gel-filtration column: A component of the Protein Labeling Kit RED-NHS (Cat# L001).
2.3 Binding Buffer (1×)
3
20 mM Hepes (pH 7.5), 100 mM NaCl, 1 mM EDTA, 0.05 % NP-40 (see Note 2). One aliquot of 20 mL was kept on ice for immediate use.
Methods All proteins, peptides, and binding buffer were stored on ice (0–4 °C) throughout the procedures unless otherwise specified. All the measurements were performed on a NanoTemper Monolith NT.115 instrument with settings of 40 % LED, 40 % IR-Laser, 30-s Laser-On, and 5-s Laser-Off unless otherwise specified.
3.1 Labeling of the p48 Protein with NT-647 Fluorescent Dye
1. Prepare 0.1 mL of 20 μM p48 protein in binding buffer (see Note 3). 2. Prepare 0.1 mL of 40 μM NT-647 in binding buffer (see Notes 4–6). 3. Mix 0.1 mL of 20 μM p48 and 0.1 mL of 40 μM NT-647 in a new 1.5 mL Eppendorf tube, and incubate at room temperature for 30 min, protecting the tube from light (see Note 7). 4. Pre-equilibrate a Sephadex G25 gel-filtration column (from the labeling kit) with binding buffer: place the column in a Falcon tube, top up the column with binding buffer and spin at 1,000 × g for 2 min (see Note 8). Discard the flow-through and repeat the procedure until at least 2.5 column volumes of the buffer has been washed through the column. 5. Add 0.3 mL of binding buffer to the tube containing the p48 and NT-647 after the 30-min incubation. Place the preequilibrated column in a new Falcon tube and slowly apply the 0.5 mL of reaction mix to the center of the matrix surface. Spin the column at 1,000 × g for 2 min collecting the labeled p48 at the bottom of the Falcon tube. Transfer the collected 0.5 mL labeled p48 (named NT647–p48 hereafter) into a new 1.5 mL Eppendorf tube (again, protecting the tube from light). Either use it immediately or store at 4 °C for up to 1 week.
272
Wei Zhang et al.
6. Evaluate the labeling efficiency by comparing the absorption at 280 nm (protein) and at 650 nm (dye: molar absorbance 250,000 M−1 cm−1). 3.2 Determining the Working Concentration of NT647–p48 and NT647–H3 (1–59) (See Notes 9 and 10)
1. In 16 low-volume (0.2 mL) reaction tubes (see Note 11), carry out a serial dilution of either 20 μL of NT647-p48 or NT647-H3 (1–59) by a factor of 0.5 at each step from 4 μM to 0.125 nM in binding buffer. 2. Spin the tubes at 15,300 × g for 5 min at 4 °C. 3. Insert a MST grade glass capillary into each sample tube to siphon about 4 μL of sample into the capillary. 4. Measure the fluorescent intensity of every sample. 5. Plot the peak fluorescent intensity of each sample capillary. The most suitable concentrations in the case of NT647–p48 and NT647–H3 (1–59) were 50 nM (see Note 12).
3.3 Testing Sample Capillaries
Some molecules can bind nonspecifically to the surface of the capillary, resulting in a poor quality MST signal (see Note 13). An example of measuring the interaction of NT647–p48 with histone H4 (1–41) peptide using standard treated capillaries and hydrophilic treated capillaries is described below. 1. In 16 low-volume (0.2 mL) reaction tubes, carry out a serial dilution of 20 μL of H4 (1–41) peptide by a factor of 0.5 at each step from 500 μM to 7.5 nM in 50 nM NT647–p48 binding buffer and mix thoroughly (see Note 14). 2. Spin the tubes at 15,300 × g for 5 min at 4 °C. 3. Retrieve samples with standard treated capillaries and measure the MST signals. Split peaks appear when the H4 (1–41) peptide is at a high concentration while single-peaked graphs reappear at a lower concentration of the peptide (Fig. 1a). The split behavior indicates that adsorption of the fluorescent molecule to the capillary surface is occurring. 4. Repeat step 3 with hydrophilic treated capillaries—all peaks appear with the expected height and without the split peak behavior (Fig. 1b) (see Note 15). 5. Plot the MST signal of each sample, and compare the change in thermophoretic mobility to obtain a KD value (Fig. 1c).
3.4 Adjusting the Buffer Conditions
Besides binding of a low molecular weight histone peptides to labeled p48 protein, the binding of a labeled histone H3 (1–59) peptide to unlabeled p48 protein was also tested. The experimental procedure was the same as that described above with one modification as detailed below. The stock of p48 protein was stored in the binding buffer containing 10 % glycerol. To be certain that no buffer effects would
Microscale Thermophoresis for Histone Studies
273
Fig. 1 NT647-labeled p48/histone H4 (1–41) peptide capillary measurements in (a) standard and (b) hydrophilic capillaries. Switching to hydrophilic capillaries prevented nonspecific binding of the labeled protein to the capillary walls. The abscissa in (a) and (b) correspond to the position (mm). (c) The resulting binding curve showed a calculated KD of 0.137 ± 0.062 μM. Inset figure: the normalized MST curves for each sample
274
Wei Zhang et al.
Fig. 2 Binding curve of NT647-labeled histone H3 (1–59) peptide to p48 in hydrophilic treated capillaries: The binding curve represents the data points from three measurements. The calculated KD is 4.2 ± 0.93 μM. Inset figure: the normalized MST curves
be observed (see Note 16), when diluting the p48 protein with 50 nM NT647-labeled histone H3 (1–59) peptide, the binding buffer was prepared with a final glycerol concentration of 5 % in all samples (see Note 17). The capillary scan, the resulting binding curve, and the shape of the MST curves are shown (Fig. 2).
4
Notes 1. The quality and quantity of the protein and the peptides used in these experiments were verified by mass spectrometry and amino acid analysis. 2. Protein (or peptides) labeling was achieved with a reactive dye (NT-647) using N-hydroxysuccinimide (NHS)-ester chemistry, which reacts efficiently with primary amino groups (–NH2) at pH 7–9 to form stable amide bonds. Buffers that contain primary amines (e.g., Tris or glycine) are not compatible with NT-647 dye labeling because they react with the NHS ester moiety attached to the dye. 3. Spin the stock of protein or peptide for 5 min at 15,300 × g in a bench-top centrifuge to remove any large aggregates which are one of the main sources of noise. 4. The labeling of primary amines at low density is of advantage since different molecules of the protein are in most cases labeled at different positions and any effect of the label on the binding interaction is thus minimized. However, site-directed labeling procedures can also be applied, e.g., by labeling single cysteines using maleimide dyes.
Microscale Thermophoresis for Histone Studies
275
5. In cases where an effect as a result of labeling is observed, the Monolith NT.LabelFree instrument, which uses intrinsic tryptophan fluorescence, can be used instead of the NT.115. 6. MST works equally well with most fluorescent dyes and fluorescent fusion proteins. 7. This is to avoid photobleaching. Wrap the tube with a piece of aluminum foil or keep the tube in a dark place. 8. It is necessary to adjust the appropriate speed of centrifugation in order to obtain equal volumes of loading and elution. 9. This procedure for labeling the p48 protein was also applied to label the histone H3 (1–59) peptide. 10. Working concentration was defined as the lowest concentration of labeled molecule giving a sufficient intensity of fluorescence signal (i.e., >100 counts on the NT.115 instrument). 11. Never use too large (>0.5 mL) or too small a volume (
1
Introduction Microscale thermophoresis (MST) is a new method that enables the quantitative analysis of molecular interactions in free solution using only a few microliters of sample. The technique is based on thermophoresis—the directed motion of molecules along temperature gradients. Thermophoresis is highly sensitive to binding-induced changes of molecular properties, be they changes in size, charge, hydration shell, or conformation. Thermophoresis is an optical technique, where an infrared laser is used to locally heat part of the sample, following which the molecules mobility in the resulting temperature gradient is then analyzed via the monitoring of their fluorescence [1, 2]. In addition to fluorescent labeling of one of the molecules involved, either directly or by expressing a fluorescent fusion protein, intrinsic protein fluorescence can also be utilized for MST, thus allowing for label-free MST analysis [3]. In contrast to other techniques, the interaction of small unlabeled peptides with a larger fluorescent protein can be readily studied by MST, despite the small change in molecular weight involved. The type of buffer
Juan C. Stockert et al. (eds.), Functional Analysis of DNA and Chromatin, Methods in Molecular Biology, vol. 1094, DOI 10.1007/978-1-62703-706-8_21, © Springer Science+Business Media New York 2014
269
270
Wei Zhang et al.
and additives can also be chosen freely. Measurement is even possible in complex bioliquids, such as cell lysates, and thus under close to in vivo conditions without sample purification. In eukaryotic cells, DNA is packaged with proteins to form higher order structures called chromatin. The basic structural unit of chromatin is the nucleosome, which is formed from around 146 base pairs (bp) of DNA wrapped over a histone octamer in 1.65 left-handed helical turns. The histone octamer comprises two copies of each of histones H2A, H2B, H3, and H4 with their unstructured N-terminal tails protruding out from the octameric core [4]. These tails are subjected to various posttranslational modifications (PTMs) such as acetylation, methylation, phosphorylation, and ubiquitination by various enzymes. These modified sites can be utilized as specific codes for either recruiting new effectors or preventing particular interactions in the large number of complex regulatory processes that occur in the cell [5, 6]. During the cell cycle nucleosomes need to be assembled and disassembled, and it is thought that during the cell cycle histone proteins almost never appear freely on their own. Instead they exist in complexes with histone chaperones, proteins that play a key role in nucleosome assembly/disassembly, histone posttranslational modifications, and ATP-dependent nucleosome remodeling [7]. The human RbAp46/RbAp48 (retinoblastoma-associated proteins p46/48) proteins are histone chaperones that are found in a large number of chromatin complexes, where they are thought to function as histone-binding subunits [8]. Analysis of their interactions with histones and histone peptides is therefore of great interest. Although there are several techniques available to assess protein-peptide interactions, they either need considerably larger sample quantities, specifically synthesized labeled peptides, or require studies over a solid surface. For a variety of reasons this is not always convenient or desirable. MST offers a new opportunity to perform studies of p48 histone-peptide interactions without the need for specific labeling and with very small amounts of sample. Here, we show how to carry out the analysis of randomly labeled lysine residues in either the p48 protein (48 kDa) or the histone peptide at low density with a fluorescent dye.
2
Materials
2.1 Peptides and Protein
1. Human histone H4 (amino acid 1–41) peptide was chemically synthesized by Alta Bioscience (http://www.altabioscience.com/). 2. Human histone H3 (amino acid 1–59) peptide was prepared according to Richart et al. [9]. 3. Human p48 protein was prepared according to Lejon et al. [10] (see Note 1).
Microscale Thermophoresis for Histone Studies
2.2 Microscale Thermophoresis Instrument and Reagents
271
(For more details, please see http://www.nanotemper-technologies. com//home/) 1. NanoTemper Monolith NT.115. 2. Dye NT-647: Monolith NT™ Protein Labeling Kit RED-NHS (Cat# L001). 3. Capillaries: NT.115 standard (Cat# K002) and hydrophilic treated (Cat# K004) capillaries. 4. Sephadex G25 gel-filtration column: A component of the Protein Labeling Kit RED-NHS (Cat# L001).
2.3 Binding Buffer (1×)
3
20 mM Hepes (pH 7.5), 100 mM NaCl, 1 mM EDTA, 0.05 % NP-40 (see Note 2). One aliquot of 20 mL was kept on ice for immediate use.
Methods All proteins, peptides, and binding buffer were stored on ice (0–4 °C) throughout the procedures unless otherwise specified. All the measurements were performed on a NanoTemper Monolith NT.115 instrument with settings of 40 % LED, 40 % IR-Laser, 30-s Laser-On, and 5-s Laser-Off unless otherwise specified.
3.1 Labeling of the p48 Protein with NT-647 Fluorescent Dye
1. Prepare 0.1 mL of 20 μM p48 protein in binding buffer (see Note 3). 2. Prepare 0.1 mL of 40 μM NT-647 in binding buffer (see Notes 4–6). 3. Mix 0.1 mL of 20 μM p48 and 0.1 mL of 40 μM NT-647 in a new 1.5 mL Eppendorf tube, and incubate at room temperature for 30 min, protecting the tube from light (see Note 7). 4. Pre-equilibrate a Sephadex G25 gel-filtration column (from the labeling kit) with binding buffer: place the column in a Falcon tube, top up the column with binding buffer and spin at 1,000 × g for 2 min (see Note 8). Discard the flow-through and repeat the procedure until at least 2.5 column volumes of the buffer has been washed through the column. 5. Add 0.3 mL of binding buffer to the tube containing the p48 and NT-647 after the 30-min incubation. Place the preequilibrated column in a new Falcon tube and slowly apply the 0.5 mL of reaction mix to the center of the matrix surface. Spin the column at 1,000 × g for 2 min collecting the labeled p48 at the bottom of the Falcon tube. Transfer the collected 0.5 mL labeled p48 (named NT647–p48 hereafter) into a new 1.5 mL Eppendorf tube (again, protecting the tube from light). Either use it immediately or store at 4 °C for up to 1 week.
272
Wei Zhang et al.
6. Evaluate the labeling efficiency by comparing the absorption at 280 nm (protein) and at 650 nm (dye: molar absorbance 250,000 M−1 cm−1). 3.2 Determining the Working Concentration of NT647–p48 and NT647–H3 (1–59) (See Notes 9 and 10)
1. In 16 low-volume (0.2 mL) reaction tubes (see Note 11), carry out a serial dilution of either 20 μL of NT647-p48 or NT647-H3 (1–59) by a factor of 0.5 at each step from 4 μM to 0.125 nM in binding buffer. 2. Spin the tubes at 15,300 × g for 5 min at 4 °C. 3. Insert a MST grade glass capillary into each sample tube to siphon about 4 μL of sample into the capillary. 4. Measure the fluorescent intensity of every sample. 5. Plot the peak fluorescent intensity of each sample capillary. The most suitable concentrations in the case of NT647–p48 and NT647–H3 (1–59) were 50 nM (see Note 12).
3.3 Testing Sample Capillaries
Some molecules can bind nonspecifically to the surface of the capillary, resulting in a poor quality MST signal (see Note 13). An example of measuring the interaction of NT647–p48 with histone H4 (1–41) peptide using standard treated capillaries and hydrophilic treated capillaries is described below. 1. In 16 low-volume (0.2 mL) reaction tubes, carry out a serial dilution of 20 μL of H4 (1–41) peptide by a factor of 0.5 at each step from 500 μM to 7.5 nM in 50 nM NT647–p48 binding buffer and mix thoroughly (see Note 14). 2. Spin the tubes at 15,300 × g for 5 min at 4 °C. 3. Retrieve samples with standard treated capillaries and measure the MST signals. Split peaks appear when the H4 (1–41) peptide is at a high concentration while single-peaked graphs reappear at a lower concentration of the peptide (Fig. 1a). The split behavior indicates that adsorption of the fluorescent molecule to the capillary surface is occurring. 4. Repeat step 3 with hydrophilic treated capillaries—all peaks appear with the expected height and without the split peak behavior (Fig. 1b) (see Note 15). 5. Plot the MST signal of each sample, and compare the change in thermophoretic mobility to obtain a KD value (Fig. 1c).
3.4 Adjusting the Buffer Conditions
Besides binding of a low molecular weight histone peptides to labeled p48 protein, the binding of a labeled histone H3 (1–59) peptide to unlabeled p48 protein was also tested. The experimental procedure was the same as that described above with one modification as detailed below. The stock of p48 protein was stored in the binding buffer containing 10 % glycerol. To be certain that no buffer effects would
Microscale Thermophoresis for Histone Studies
273
Fig. 1 NT647-labeled p48/histone H4 (1–41) peptide capillary measurements in (a) standard and (b) hydrophilic capillaries. Switching to hydrophilic capillaries prevented nonspecific binding of the labeled protein to the capillary walls. The abscissa in (a) and (b) correspond to the position (mm). (c) The resulting binding curve showed a calculated KD of 0.137 ± 0.062 μM. Inset figure: the normalized MST curves for each sample
274
Wei Zhang et al.
Fig. 2 Binding curve of NT647-labeled histone H3 (1–59) peptide to p48 in hydrophilic treated capillaries: The binding curve represents the data points from three measurements. The calculated KD is 4.2 ± 0.93 μM. Inset figure: the normalized MST curves
be observed (see Note 16), when diluting the p48 protein with 50 nM NT647-labeled histone H3 (1–59) peptide, the binding buffer was prepared with a final glycerol concentration of 5 % in all samples (see Note 17). The capillary scan, the resulting binding curve, and the shape of the MST curves are shown (Fig. 2).
4
Notes 1. The quality and quantity of the protein and the peptides used in these experiments were verified by mass spectrometry and amino acid analysis. 2. Protein (or peptides) labeling was achieved with a reactive dye (NT-647) using N-hydroxysuccinimide (NHS)-ester chemistry, which reacts efficiently with primary amino groups (–NH2) at pH 7–9 to form stable amide bonds. Buffers that contain primary amines (e.g., Tris or glycine) are not compatible with NT-647 dye labeling because they react with the NHS ester moiety attached to the dye. 3. Spin the stock of protein or peptide for 5 min at 15,300 × g in a bench-top centrifuge to remove any large aggregates which are one of the main sources of noise. 4. The labeling of primary amines at low density is of advantage since different molecules of the protein are in most cases labeled at different positions and any effect of the label on the binding interaction is thus minimized. However, site-directed labeling procedures can also be applied, e.g., by labeling single cysteines using maleimide dyes.
Microscale Thermophoresis for Histone Studies
275
5. In cases where an effect as a result of labeling is observed, the Monolith NT.LabelFree instrument, which uses intrinsic tryptophan fluorescence, can be used instead of the NT.115. 6. MST works equally well with most fluorescent dyes and fluorescent fusion proteins. 7. This is to avoid photobleaching. Wrap the tube with a piece of aluminum foil or keep the tube in a dark place. 8. It is necessary to adjust the appropriate speed of centrifugation in order to obtain equal volumes of loading and elution. 9. This procedure for labeling the p48 protein was also applied to label the histone H3 (1–59) peptide. 10. Working concentration was defined as the lowest concentration of labeled molecule giving a sufficient intensity of fluorescence signal (i.e., >100 counts on the NT.115 instrument). 11. Never use too large (>0.5 mL) or too small a volume (
