This protocol combines the characterization of a protein sample by capillary gel electrophoresis and a fast-binding screening for charged ligands by affinity capillary electrophoresis. It is recommended for proteins with a flexible structure, such as intrinsically disordered proteins, to determine any differences in binding for different conformers.
Plants are strongly dependent on their environment. In order to adjust to stressful changes (e.g., drought and high salinity), higher plants evolve classes of intrinsically disordered proteins (IDPs) to reduce oxidative and osmotic stress. This article uses a combination of capillary gel electrophoresis (CGE) and mobility shift affinity electrophoresis (ACE) in order to describe the binding behavior of different conformers of the IDP AtHIRD11 from Arabidopsis thaliana. CGE is used to confirm the purity of AtHIRD11 and to exclude fragments, posttranslational modifications, and other impurities as reasons for complex peak patterns. In this part of the experiment, the different sample components are separated by a viscous gel inside a capillary by their different masses and detected with a diode array detector. Afterward, the binding behavior of the sample towards various metal ions is investigated by ACE. In this case, the ligand is added to the buffer solution and the shift in migration time is measured in order to determine whether a binding event has occurred or not. One of the advantages of using the combination of CGE and ACE to determine the binding behavior of an IDP is the possibility to automate the gel electrophoresis and the binding assay. Furthermore, CGE shows a lower limit of detection than the classical gel electrophoresis and ACE is able to determine the manner of binding a ligand in a fast manner. In addition, ACE can also be applied to other charged species than metal ions. However, the use of this method for binding experiments is limited in its ability to determine the number of binding sites. Nevertheless, the combination of CGE and ACE can be adapted for characterizing the binding behavior of any protein sample towards numerous charged ligands.
Plants are more dependent on their environment than many other life forms. Since plants cannot move to other places, they have to adjust to changes in their surroundings (e.g., drought, cold and high salt concentrations). Consequently, higher plants developed specialized stress proteins like dehydrins, which fulfill manifold tasks to reduce cell stress related to high salinity. These proteins bind water and ions inside cells, reduce oxidative stress by binding Cu2+-ions, and interact with phospholipids as well as the cytoskeletons. Moreover, binding Zn2+-ions allows these proteins to act as transcription factors. Their ability to bind Ca2+-ions after phosphorylation has also been reported1.
The multifunctional behavior of these proteins is related to the absence of hydrophobic amino acid residues. Consequently, they lack any hydrophobic interactions inside the peptide chain and also a constrained structure. However, because these proteins lack a restrictive structure, they can occupy different conformers under the same conditions. Therefore, they can be described best as an ensemble of structures rather than as a single conformation. Proteins with these properties are known as intrinsically disordered proteins (IDPs) and are a widely used concept for stress proteins and crosstalk between different pathways in eukaryotic cells2.
One of these stress-related IDPs is AtHIRD11. It is one of Arabidopsis thaliana's most highly drought-expressed IDPs. Hence, the different conformers can be separated by their effective radius to charge ratio, and capillary electrophoresis (CE) has been used for further investigations. Previous ACE experiments demonstrated the interactions between AtHIRD11 and transition metal ions such as Cu2+-, Zn2+-, Co2+, and Ni2+-ions. The detailed results can be found in Hara et al.3 and Nachbar et al.4.
The ACE method that will be used here is based on our earlier published works6. However, the addition of the EOF marker acetanilide to the protein sample is not suitable. AtHIRD11 shows broad peak patterns, and adding the EOF marker to the sample would disguise two peaks. Therefore, the marker is used in a separate run. Before the binding behavior is examined, it is confirmed that the peaks found during previous experiments are from different conformers. Thus, CGE is used to distinguish between the protein conformers, the post-translational modified protein, and impurities, such as fragments of AtHIRD11, by their different masses. Subsequently, the characterized AtHIRD11 sample's binding behavior towards various different metal ions is investigated.
The purpose of this article is to describe an experimental setup to distinguish between an IDP and other components of a sample in order to evaluate the differences in the binding behavior of different conformers.
1. Preparation of the Capillary Electrophoresis Instruments
2. Prepare the Solutions
3. Capillary Gel Electrophoretic Separation
Note: Prepare the separation according to the method described in previous work by Nachbar et al.4.
4. Affinity Capillary Electrophoretic Analysis
Note: Prepare the separation according to the method described in previous works by Nachbar et al.4 and Alhazmi et al.5.
Figure 1 shows the electropherogram for the AtHIRD11 sample obtained during the CGE experiments. The peptide size increases from left to right. Peak number 4 has the largest mass and indicates the intact protein. The smaller peaks 2 and 3 represent smaller impurities (e.g., degradation products). The first peak and the inconsistency of the baseline before could also be reproduced without the protein sample. Therefore, it is related to the SDS gel itself and does not represent any sample-associated impurities.
Figure 2 represents the electropherogram for acetanilide during the ACE experiments. The acetanilide solution shows only 1 high peak since it should have no impurities. The detection time for the peak maximum indicates the migration time of the electroosmotic flow (tEOF) and is used for the calculation of the interaction.
Figure 3 is the electropherogram of the AtHIRD11 sample during the ACE experiment in the absence of SDS and metal ions. It shows, besides the 2 impurities from the CGE electropherogram, at least 5 peaks which are related to the protein itself. Peak 1 and 2 can be assigned to impurities since their migration times indicate that they are small or highly positively charged. This assumption is supported by the increase of the peak height and the area over time. Since AtHIRD11 peaks 3 and 4 are close to the EOF, they are hardly charged and cannot be separated in every run. They almost show no interactions and represent conformations without accessible residues essential for the interactions with the metal ions. AtHIRD11 shows several different charge-to-size-ratios due to different effective radii of different conformers. These conformers can merge continuously. Subsequently, peaks 5 – 7 are broader than the usual protein peaks. The broad peaks give a hint on the kinetics of the interactions. Since the peaks are not narrow, a fast and a very slow conversion between the different conformers can be excluded. In both cases, the peaks would be baseline-separated4. In this case, a pre-equilibration step can change the peak pattern and would give a better insight into the interactions with slower kinetics.
Figure 4 shows the graphical evaluation of the measured migration time shifts in the presence of the various metal ions for peak 6. It is represented by the calculated value ΔR/Rf. This value indicates how strong a shift is (by its value) and the overall change of the protein-metal ion complexes charge (by its sign). In order to differentiate between a significant interaction and a coincidental shift, the confidence intervals for α = 0.05 are calculated as well. In cases where the confidence interval is not intersecting with the zero line, the interaction is considered as significant. Results are taken into account if the peak was present in at least 6 out of 10 electropherograms. Peak 6 was not present in any run with 500 µM Co2+. The complete list of the interactions for each AtHIRD11 peak is published in an earlier work by Nachbar et al.4.
ΔR/Rf is calculated using the migration time ratios Ri (in the presence of the ligand) and the migration time ratios Rf (in the absence of the ligand):
Ri and Rf are calculated using the peak top times tprotein for the protein sample peaks and the corresponding peak top time for the EOF-marker tEOF:
For Ri, the times in the presence of the ligand are used and for Rf, the times in absence.
Figure 1: Electropherogram for the CGE separation of the AtHIRD11 sample in gel buffer. Peak 1 and the peaks on its left can be observed even without a sample injection, so they are artefacts. Since the masses and the overall peak areas of peak 2 and 3 are smaller than that of peak 4, they are likely impurities, such as degradation products. Please click here to view a larger version of this figure.
Figure 2: Electropherogram of the EOF-marker acetanilide run. The figure shows only 1 peak for the compound without any metal ions and SDS gel buffer. The time of the peak top marks the electroosmotic flow. Please click here to view a larger version of this figure.
Figure 3: Example for an electropherogram of the sample separation during the ACE experiments showing a complex peak pattern in the absence of any metal ions. At least 7 peaks could be identified and related to the protein and its 2 impurities. Please click here to view a larger version of this figure.
Figure 4: Graphical evaluation of the binding behavior towards the investigated metal ions using a concentration of 500 µM for peak 6. The values of ΔR/Rf give a hint on the intensity of the migration time shift and therefore on the strength of the conformational changes on metal ion binding. The sign of ΔR/Rf indicates if the complex is more positively or negatively charged compared to the unbound protein. The error bars indicate the confidence intervals for α = 0.05. Subsequently, they show the area where the true ΔR/Rf-value can be found with a certainty of 95%. The red bars indicate the calculated results obtained by insufficient data (the peak was present in less than 6 electropherograms). Please click here to view a larger version of this figure.
For all CE experiments, the preparation of the capillary is a critical step. Since it is a glass tube with a small diameter, it can easily break at the spots where the coating is removed when it is being handled. The installation of the capillary into the instrument should be done very carefully.
The critical steps involved in using the ACE method mainly pertain to the experimental setup. Another critical step is finding the right sample injection parameters. The injected amount of the sample has to be enough for high peaks. However, overloading it would result in a poorly resolved peak pattern with overlapping peaks. When decreasing the injected volume, it is recommended to decrease the injection pressure instead of the injection time since it takes some time for the instrument to reach the programmed injection pressure.
Unusually broad peaks can be found for different protein samples. This pattern can be influenced by increasing the voltage. Consequently, the peaks get narrower and the separation time is reduced. Nevertheless, the shorter separation time can result in not-resolved peaks and the higher voltage increases the Joule heating and limits the effect of narrow peaks by electrophoretic dispersion7,8. Therefore, the right conditions have to be determined for each new protein examined.
Combining CGE and ACE to gather information about the intrinsically disordered character of a protein and its binding behavior is a novel approach. The characterization of the sample by gel electrophoresis is a necessary step to prove that any multiple peaks observed are not related to impurities. The advantage of CGE is the possibility to automate it, the shorter preparation time, and a higher resolution compared to the classical gel electrophoresis9. Using ACE for IDP binding experiments shows its superiority towards other assays (e.g., immobilized metal affinity chromatography in the results) since ACE not only shows binding events. Moreover, it separates different conformers of a protein and can reveal differences in binding ligands of diverse conformers. This behavior cannot be detected by methods without a separation during the binding experiment. In addition, the analysis time is only 6 min; therefore, the results can be calculated in a short amount of time10. On the other hand, the fast analysis comes with limitations in estimating the number of binding sites11.
The future of this approach lies in the possibility to screen potential binding partners for proteins, especially IDPs, in a fast manner. It gives a quick overview of the different binding behavior of the various conformers present in a sample. The use of a size-dependent separation (e.g., CGE) for the sample characterization is mandatory since ACE itself cannot distinguish between impurities and the protein itself. However, CGE gives an idea about the number of different-sized species. The application of ACE as a binding assay has been extended already from metal ion ligands to any charged ligand12. Therefore, it can add additional information about the interactions to any characterized binding, since the combination of the separation and the binding assay in one method can give evidence of any altered binding behavior of different conformers and posttranslational modifications, respectively.
The authors have nothing to disclose.
We grateful thank Masakuza Hara (the Research Institute of Green Science and Technology, University of Shizuoka, Japan) for providing the AtHIRD11 protein samples.
AtHIRD11 sample | Shizuoka University (Group Prof. M. Hara) | – | Dehydrin Protein from Arabidopsis thaliana expressed in Escherichia coli |
Barefused silica capillary | Polymicro Technologies (Phoenix, USA) | 106815-0017 | TSP050375, 50 μm inner diameter, 363 μm outer diameter, polyimide coating |
Agilent 1600A | Agilent Technologies (Waldbronn, Germany) | comercially not available anymore | Capillary electrophoresis instrument; Agilent 7100 CE can be used instead |
Agilent 7100 CE | Agilent Technologies (Waldbronn, Germany) | G7100A | Capillary electrophoresis instrument |
Injekt 2 mL | B. Braun (Melsungen, Germany) | 4606051V | Syringe for filtration |
Rotilabo-syringe filters | Carl Roth GmbH + Co. KG (Karlsruhe, Germany) | KY62.1 | PVDF membrane filter for solution filtration |
Eppendorf Research plus 10 μL | Eppendorf (Wesseling-Berzdorf, Germany) | 3121 000.023 | Micro pipette for sample handling |
Eppendorf Research plus 10 μL | Eppendorf (Wesseling-Berzdorf, Germany) | 3121 000.120 | Micro pipette for handling the ligand solutions |
Bulb pipette 10 mL | Duran Group GmbH(Mainz, Germany) | 24 338 08 | Preparing theNaOH solution |
Bulb pipette 25 mL | Duran Group GmbH(Mainz, Germany) | 24 338 14 | Preparing the ligand solution |
Duran glas volumetric flask 25 mL | Duran Group GmbH(Mainz, Germany) | 24 671 1457 | Preparing the ligand stock solution |
Duran glas volumetric flask 10 mL | Duran Group GmbH(Mainz, Germany) | 24 671 1054 | Preparing the ligand stock solution |
Proteome Lab SDS MW Gel Buffer | Beckman Coulter (Brea, USA) | comercially not available anymore | Separation during capillary gel electrophoresis / Alternative SDS buffer: CE-SDS run buffer from Bio-Rad Laboratories (München, Germany) Catalog Number: 1485032 |
Acetanilide | Sigma-Aldrich (Steinheim, Germany) | 397229-5G | Electroosmotic flow marker |
Manganese(II) chloride | Sigma-Aldrich (Steinheim, Germany) | 13217 | Ligand |
Ethylenediaminetetraacetic acid (EDTA) | Sigma-Aldrich (Steinheim, Germany) | 431788-100G | Rinsing ingredient |
Bariumchloride | Sigma-Aldrich (Steinheim, Germany) | 202738-5G | Ligand |
Sodium dodecyl sulfate | Sigma-Aldrich (Steinheim, Germany) | 71729-100G | Solublizing protein for capillary gel electrophoresis |
Nickel(II) chloride hexahydrate | Sigma-Aldrich (Steinheim, Germany) | 654507-5G | Ligand |
Selenium(IV) chloride | Sigma-Aldrich (Steinheim, Germany) | 323527-10G | Ligand |
2-amino-2-hydroxy-methylpropane-1.3-diol (Tris) | Sigma-Aldrich (Steinheim, Germany) | 252859-100G | Buffer ingredient |
Zinc(II) chloride | Merck Millipore ( Darmstadt, Germany) | 1088160250 | Ligand |
Strontium nitrate | Merck Millipore ( Darmstadt, Germany) | 1078720250 | Ligand |
Calcium chloride dihydrate | Merck Millipore ( Darmstadt, Germany) | 1371015000 | Ligand |
37% hydrochloric acid | Merck Millipore ( Darmstadt, Germany) | 1003171000 | Adjusting pH |
Copper(II) chloride dihydrate | Riedel-de Haën (Seelze, Germany) | 31286 | Ligand |
Sonorex Longlife RK 1028 CH 45L | Allpax (Papenburg, Germany) | 10000084;0 | Ultrasonic bath |
Agilent ChemStation Rev. 8.04.03-SP1 | Agilent Technologies (Waldbronn, Germany) | G2070-91126 | Software packages to operate the CE instruments, acquisite data and evaluate it |