Here, we describe a set of methods for characterizing the interaction of proteins with membranes of cells or microvesicles.
In the human body, most of the major physiologic reactions involved in the immune response and blood coagulation proceed on the membranes of cells. An important first step in any membrane-dependent reaction is binding of protein on the phospholipid membrane. An approach to studying protein interaction with lipid membranes has been developed using fluorescently labeled proteins and flow cytometry. This method allows the study of protein-membrane interactions using live cells and natural or artificial phospholipid vesicles. The advantage of this method is the simplicity and availability of reagents and equipment. In this method, proteins are labeled using fluorescent dyes. However, both self-made and commercially available, fluorescently labeled proteins can be used. After conjugation with a fluorescent dye, the proteins are incubated with a source of the phospholipid membrane (microvesicles or cells), and the samples are analyzed by flow cytometry. The obtained data can be used to calculate the kinetic constants and equilibrium Kd. In addition, it is possible to estimate the approximate number of protein binding sites on the phospholipid membrane using special calibration beads.
Biomembranes separate the inner contents of animal cells and extracellular space. Note that membranes also surround microvesicles formed during the cell's life cycle and organelles. The cell membrane is predominantly composed of lipids and proteins. Membrane proteins perform signaling, structural, transport, and adhesive functions. However, the lipid bilayer is also essential for the interrelation of the animal cell with the extracellular space. This paper proposes a method for studying the peripheral interaction of external proteins with the lipid membrane.
The most striking example of reactions occurring on the outer membrane layer of an animal cell is the blood coagulation reaction. An important feature of blood coagulation is that all the main reactions proceed on the phospholipid membranes of cells and microvesicles arising from these cells and not in the plasma1,2,3. Membrane-dependent reactions include the process of starting coagulation (on the cell membranes of the subendothelium, inflamed endothelium, or activated immune cells, with the participation of a tissue factor), all reactions of the main cascade-activation of factors IX, X, prothrombin; activation of factor XI by thrombin (on the membranes of activated platelets, erythrocytes, lipoproteins, and microvesicles); reactions of the protein C pathway; inactivation of coagulation enzymes (on the membranes of endothelial cells with the participation of thrombomodulin cofactors, endothelial protein C receptor, heparan sulfate); and contact pathway reactions (on membranes of platelets and some microvesicles with the participation of unknown cofactors). Thus, it is impossible to investigate blood coagulation without studying the interaction of various plasma proteins with the membrane of blood cells.
This paper describes a flow-cytometry-based method for characterizing the interaction of proteins with lipid membranes of cells or microvesicles. This approach was initially proposed to study the interaction of blood plasma with platelets and artificial phospholipid vesicles. Moreover, most of the studied proteins interact directly with negatively charged membrane phospholipids, particularly with phosphatidylserine4,5. Additionally, there are proteins whose interaction with the membrane is mediated by special receptors6.
An important ability of flow cytometry is discriminating between free and bound ligands without additional separation. This feature of cytometry allows the study of ligand equilibrium binding at the endpoint and helps perform continuous kinetic measurements. The technique is unsophisticated and does not require complex sample preparation. Flow cytometry is actively used to quantitatively study the dynamics of interaction between fluorescent peptides, receptors, and G-proteins in intact and detergent-permeable neutrophils7. This approach is also applicable for exploring protein-DNA interactions and the kinetics of endonuclease activity in real time8. Over time, this method was used to quantitatively study high-affinity protein-protein interactions with purified lipid vesicles9, or, more generally, with membrane proteins expressed in a highly efficient Sf9 cell expression system10. Quantitative methods have also been described for characterizing protein-liposome interactions using flow cytometry for transmembrane proteins11.
This technique uses self-made calibration beads to avoid using commercially available beads7. The calibration beads used previously7 were intended to work with fluorescein, which substantively restricted the assortment of accessible fluorescent ligands on the proteins. In addition, this paper offers a new way to acquire and analyze kinetic data for reasonable time resolution. Although this method is described for artificial phospholipid vesicles, there are no obvious limitations for its adaptability to cells, natural vesicles, or artificial phospholipid vesicles with a different lipid composition. The method described herein allows the estimation of the parameters of interaction (kon, koff) and equilibrium (Kd) and facilitates quantitative characterization of the number of protein binding sites on the membrane. Note that this technique provides an approximate estimate of the number of binding sites. The advantages of the method are its relative simplicity, accessibility, and adaptability to native cells and natural and artificial microvesicles.
1. Fluorescent protein labeling
2. Preparation of phospholipid vesicles
3. Isolation of platelets from whole blood
4. Detection of protein – lipid interaction by flow cytometry
5. Analysis of flow cytometry data
6. Converting fluorescence intensity to the mean number of binding sites
The flow cytometry method described herein is used to characterize the binding of plasma coagulation proteins to activated platelets. In addition, phospholipid vesicles PS:PC 20:80 were applied as a model system. This paper mainly focuses on artificial phospholipid vesicles as an example. The parameters of the cytometer, in particular, the photomultiplier tube (PMT) voltage and the compensation must be selected for each specific device, the object of study (cells, artificial or natural microvesicles), and the dyes used. Figure 1B,C show an example of gating artificial phospholipid vesicles that are ~1 µm in size with the incorporated lipophilic fluorescent dye DiIC16 (3). Large vesicle size and lipophilic fluorescent dye helped detect vesicles using a cytometer. The gate was set based on a sample containing the same-size artificial lipid vesicles but without the fluorescent dye (Figure 1B). Only events inside this gate were used in the analysis.
The kinetics of protein binding to vesicles was analyzed at the first stage. The sample for this was collected continuously as described in step 4.1. A typical dot-plot is shown in Figure 1D–F. The data obtained were analyzed using the flow cytometry software. The resulting curve is shown in Figure 1G. Solid lines show the curves of approximation, from which the kinetic constants of association (kon) and dissociation (koff) were obtained. As factor X binds to phospholipid vesicles reversibly and Ca2+-dependently, samples with EDTA controlled the specificity and reversibility of this binding. The resulting constants are shown in Table 1.
Based on the kinetics of binding, a time of 20 min was chosen for further equilibrium experiments to describe saturation in binding accurately. The mean fluorescence intensity of the factor was subsequently determined in the region of the vesicles. Each sample was analyzed in the presence and absence of EDTA. The fluorescence intensity in the presence of EDTA was taken as background and subtracted from the total signal as the binding of fX to the membrane in the absence of Ca2+ ions is considered nonspecific. The resulting fluorescence was converted to the number of binding sites per vesicle using the calibration beads.
Figure 1: Specific binding of fX to artificial phospholipid vesicles. (A) Scheme of experiment. (B, C) Typical dot plots of phospholipid vesicles without (B) or with (C) lipophilic fluorescent dye DiIC16 (3). (D–F) Typical dot plots of factor X interaction with phospholipid vesicles before (D) or after (E) 20-fold dilution and in presence EDTA (F). (G) Kinetics of FX (250 nM) binding and dissociation to phospholipid vesicles. (H) Equilibrium interaction of factor X to phospholipid vesicles. Results are the means ±SD for n=3 different samples. Abbreviations: FX = Factor X; Ph vesicles = phospholipid vesicles; SSC = side scatter; a.u. = arbitrary unit). Please click here to view a larger version of this figure.
Apparent Kd ± SEM (nM) | kon ± SEM (μM-1s-1) | Koff ± SEM (s-1) | Apparent number of binding sites per vesicle ± SEM |
400 ± 80 | 0.371 ± 0.012 | 0.019 ± 0.004 | 8,000 ± 800 |
Table 1: Parameters of fX interaction with artificial phospholipid vesicles. Parameters were determined from the curves (see Figure 1F,G). Results are the mean ± SEM for n = 3.
The proposed method can be adapted for a rough characterization of the interaction of proteins with phospholipid membranes from various sources and compositions. The quantitative flow cytometry described here concedes to surface plasmon resonance (SPR) in several parameters. In particular, it has a lower sensitivity and time resolution and requires fluorescent labeling of proteins. Fluorescent labeling can lead to a change in conformation and loss of activity for many proteins and therefore requires careful control. However, this technique has significant advantages over others. This method provides an opportunity to explore the interaction of proteins with the native cell membrane, which is not readily implemented using SPR. Moreover, the approach allows the estimation of the number of protein binding sites on the membrane surface and can be efficient for some analysis tasks.
Commercially available beads are available for some fluorescent dyes to count binding sites. However, there are no such beads for many widely used dyes. Hence, self-made beads are the best way to resolve this. The same cells were used to prepare these beads as for all other experiments. However, as the beads require high centrifugation speeds during washing, phospholipid vesicles cannot be used. However, cells or vesicles can be replaced with beads with amino-reactive groups, which can be conjugated to the chosen dye. The sequence of actions will be similar to those described in step 6.1.
The limitations of this quantitative flow cytometry method are related to the technical capabilities of the used cytometer. Three different models of flow cytometers (with the variable lasers, detectors, pumps) were applied to reproduce this technique without any complication. However, the selection of a fluorescent label suitable for the cytometer used must be carefully considered because the set of lasers, detectors, and optical filters differ from device to device, even within the same model. It is necessary to focus on the ability of the cytometer to measure microparticles of a specific diameter; not all instruments are equally capable of detecting particles at resolutions below 200 nm (to determine this, use the fixed-size calibration beads supplied by the instrument’s manufacturer). Additionally, some flow cytometers, which are sampled using a syringe pump, cannot measure continuous binding kinetics in principle. In this case, the kinetics can be recorded only point by point, taking separate samples for measurement at certain points in time4,6.
Flow cytometry is used to investigate the expression of antigens on various cells-the presence/absence of an antigen and the percentage of cell populations expressing and not expressing this antigen. The ability of flow cytometry to concurrently discriminate free and bound ligands without additional separation procedures also affords the opportunity for quantitative assessment of ligand-binding dynamics. The method proposed herein describes the preparation of self-made calibration beads to quantify the binding of a fluorescent ligand to artificial phospholipid vesicles. This approach does not limit the choice of commercially available fluorophores. Beyond that, the technique described here for the acquisition and analysis of kinetic data enables improved time resolution. Thus, the method described herein can be used alone as a first step in characterizing protein-membrane interactions or in combination with other methods (for example, SPR or microscopy) to improve measurement accuracy.
The authors have nothing to disclose.
The authors were supported by a Russian Science Foundation grant 20-74-00133.
A23187 | Sigma Aldrich | C7522-10MG | |
Alexa Fluor 647 NHS Ester (Succinimidyl Ester) | Thermo Fisher Scientific | A37573 | fluorescent dye |
Apyrase from potatoes | Sigma Aldrich | A2230 | |
BD FACSCantoII | BD Bioscience | ||
bovine serum albumin | VWR Life Science AMRESCO | Am-O332-0.1 | |
Calcium chloride, anhydrous, powder, ≥97% | Sigma Aldrich | C4901-100G | |
Cary Eclipse Fluorescence Spectrometer | Agilent | ||
D-(+)-Glucose | Sigma Aldrich | G7528-1KG | |
DiIC16(3) (1,1'-Dihexadecyl-3,3,3',3'-Tetramethylindocarbocyanine Perchlorate) | Thermo Fisher Scientific | D384 | |
DMSO | Sigma Aldrich | D8418 | |
EDTA disodium salt | VWR Life Science AMRESCO | Am-O105B-0.1 | |
FACSDiva | BD Bioscience | cytometry data acquisition software | |
FlowJo | Tree Star | cytometer software for data analysis | |
HEPES | Sigma Aldrich | H4034-500G | |
Human Factor X | Enzyme research | HFX 1010 | |
Hydroxylamine hydrochloride | Panreac | 141914.1209 | |
L-α-phosphatidylcholine (Brain, Porcine) | Avanti Polar Lipids | 840053P | |
L-α-phosphatidylserine (Brain, Porcine) (sodium salt) | Avanti Polar Lipids | 840032P | |
Magnesium chloride | Sigma Aldrich | M8266-100G | |
Mini-Extruder | Avanti Polar Lipids | 610020-1EA | |
OriginPro 8 SR4 v8.0951 | OriginLab Corporation | Statistical software | |
Phosphate Buffered Saline (PBS) Tablets, Biotechnology Grade | VWR Life Science AMRESCO | 97062-732 | |
Potassium chloride | Sigma Aldrich | P9541-500G | |
Prostaglandin E1 | Cayman Chemical | 13010 | |
Sephadex G25 | GE Healthcare | GE17-0033-01 | gel filtration medium for protein purification |
Sepharose CL-2B | Sigma Aldrich | CL2B300-500ML | gel filtration medium for platelet purification |
Sodium bicarbonate | Corning | 61-065-RO | |
Sodium chloride | Sigma Aldrich | S3014-500G | |
Sodium phosphate monobasic | Sigma Aldrich | S3139-250G | |
Spin collumns with membrane 0.2 µm | Sartorius | VS0171 | |
Trisodium citrate dihydrate | Sigma Aldrich | S1804-1KG |