Summary
A biomembrane force probe (BFP) is an in situ dynamic force spectroscopy (DFS) technique. BFP can be used to measure the spring constant of molecular interactions on living cells. This protocol presents spring constant analysis for molecular bonds detected by BFP.
Abstract
A biomembrane force probe (BFP) has recently emerged as a native-cell-surface or in situ dynamic force spectroscopy (DFS) nanotool that can measure single-molecular binding kinetics, assess mechanical properties of ligand-receptor interactions, visualize protein dynamic conformational changes and more excitingly elucidate receptor mediated cell mechanosensing mechanisms. More recently, BFP has been used to measure the spring constant of molecular bonds. This protocol describes the step-by-step procedure to perform molecular spring constant DFS analysis. Specifically, two BFP operation modes are discussed, namely the Bead-Cell and Bead-Bead modes. This protocol focuses on deriving spring constants of the molecular bond and cell from DFS raw data.
Introduction
As a live-cell DFS technique, BFP engineers a human red blood cell (RBC; Figure 1) into an ultrasensitive and tunable force transducer with a compatible spring constant range at 0.1-3 pN/nm1,2,3. To probe ligand-receptor interaction, BFP enables DFS measurements at ~1 pN (10-12 N), ~3 nm (10-9 m), and ~0.5 ms (10-3 s) in force, spatial, and temporal resolution4,5. Its experimental configuration consists of two opposing micropipettes, namely the Probe and the target. The Probe micropipette aspirates a RBC and a bead is glued at its apex via a biotin-streptavidin interaction. The bead is coated with the ligand of interest (Figure 1A). The Target micropipette aspirates either a cell or a bead bearing the receptor of interest, corresponding to the Bead-Cell (Figure 1B) and Bead-Bead (Figure 1C) modes, respectively5.
BFP construction, assembly and the DFS experimental protocols were described in detail previously1,6. Briefly, a BFP touch cycle consists of 5 stages: Approach, Impinge, Contact, Retract and Dissociate (Figure 1D). The horizontal RBC apex position is denoted as ΔxRBC. At the beginning, the unstressed (zero-force) RBC deformation ΔxRBC is 0 (Table 1). The Target is then driven by a piezotranslator to impinge on and retract from the Probe bead (Figure 1D). The RBC probe is first compressed by the Target with negative RBC deformation ΔxRBC < 0. In a Bond event, the Retract stage transitions from a compressive to a tensile phase with positive RBC deformation ΔxRBC > 0 (Figure 2C and D). According to Hooke's law, the BFP bearing force is able to be measured as F = kRBC × ΔxRBC, where kRBC (Table 1) is the RBC spring constant of the BFP. Upon bond rupture and the completion of one touch cycle, the probe bead returns to zero-force position with ΔxRBC = 0 (Figure 1D).
To determine the kRBC, we measure and record the radii of the probe micropipette inner orifice (Rp), the RBC (R0) and the circular contact area (Rc) between the RBC and the probe bead (Figure 1A). Then kRBC is calculated according to the Evan's model (Eq. 1)7,8 using a LabVIEW program that acts as a virtual instrument (VI) to operate the BFP (Figure S1A)8,9.
(Eq. 1)
With a BFP established and DFS raw data obtained, hereby we present how to analyze the spring constant of ligand-receptor pair or cells. The DFS raw data on the interaction of the glycosylated protein Thy-1 and K562 cell bearing integrin α5β1 (Thy-1-α5β1; Figures 3A and 3B)10 and that of the fibrinogen and bead coated integrin αIIbβ3 (FGN-αIIbβ3; Figure 3C)11,12 have been used to demonstrate the Bead-Cell and Bead-Bead analysis modes, respectively.
BFP Experimental Preparation
For details of BFP experimental preparation and instrumentation, please refer to the previously published protocols3. In brief, human RBC has been biotinylated using the Biotin-PEG3500-NHS in the carbon/bicarbonate buffer. Proteins of interest have been covalently coupled to the borosilicate glass beads using MAL-PEG3500-NHS in the phosphate buffer. To attach to the biotinylated RBC, the probe bead is also coated with streptavidin (SA) using the MAL-SA. Please see the Table of Materials and Table 2.
To assemble the BFP (Figure 1, left), the third micropipette termed 'Helper' will be used to deliver the probe bead and glue it to the RBC's apex1,3. The covalent interaction between the SA coated probe bead and biotinylated RBC is much stronger than the ligand-receptor bond of interest. Thus, the Dissociate stage can be interpreted as the ligand-receptor bond rupture rather than the detachment of Probe bead from the RBC.
Subscription Required. Please recommend JoVE to your librarian.
Protocol
1. Obtain Analyzable DFS Events
- Start the experiment in the software (e.g., LabVIEW VI) for the BFP control and parameter setting (Figure S1A).
- Observe the repetitive probe bead-target bead/cell touches in the software for BFP Monitor (Figure S1B).
- Test and achieve the adhesion frequency ≤ 20% within the first 50 touches by tuning the impingement force and contact time, by which it ensures that ≥ 89% of DFS adhesion event are mediated by single bonds12,13,14.
NOTE: For each Bead-Cell/Bead pair, we perform 200 repetitive touch cycles. To obtain publishable data quality, we usually perform n ≥ 3 Bead-Bead or Bead-Cell pairs.- Save data, in the form of Force vs. Time, to the user directed folder by the end of each pair, prompted by the software for BFP control and parameter setting.
- Collect the Force vs. Time raw data of Bond events, as exemplified in the Figure 2A, using the BFP acquisition platform (Figure S1C).
- Open the BFP data analysis software. Click on the yellow folder icon and select the corresponding raw data file by double clicking on them.
- Run the program, and then click the up and down button to switch between events. Use the outlier exclusion criteria (Figure S2) to screen out invalid events. Select the exporting data type as Force vs. Time format and click on the Export Plots Data button.
2. Convert the Force vs. Time Curve to the Force vs. Displacement Curve
- Export the data segment corresponding to the Retract stage to a spreadsheet (Figure 2A, square marquee), which is relevant to the spring constant analysis.
- Plot the Force vs. Time Curve using spreadsheet software. To obtain the Force vs. Displacment curve, convert the time values (Figure 2A, x-axis) to the total displacement values (Δxtot) by multiplying time values with piezo movement velocity (i.e., 4,000 nm/s by preset).
- Zero the first data point by subtracting the smallest displacement value from each acquired displacement value. This horizontal transformation does not affect the ascending slopes of the Retract stage nor the subsequent spring constant calculation.
- Notably, the BFP is considered as a serial spring system in which Δxtot (Table 1) sum deformations of the RBC, ΔxRBC (Table 1), the molecular bond, Δxmol (Table 1), and the Target cell, Δxcell (Table 1), as the Eq. 2:
(Eq. 2) - Plot the Force (F) vs. Displacement (Δxtot) curve as shown in the Figure 2B.
(Eq. 2)3. Spring Cnstant Analysis of Bead- Cell Mode
- In the Force vs. Displacement curve, two distinct slop can be identified, where each can represent the compressive phase and the tensile phase. Fit a regression line to each data group (Figure 2B), where the larger linear fit slope represents the total spring constant at compressive phase (Figure 2B, red), denoted as k1 (Table 1); and the smaller linear fit slope represents the total spring constant at tenslile phase (Figure 2B, blue), denotated as k2 (Table 1).
- For springs connected in series per step 2.2 description, express the reciprocal of the total spring constant, ktot (Table 1), as the sum of the spring constant inverses of RBC, kRBC (Table 1), the molecular bond, kmol (Table 1), and the Target cell, kcell (Table 1). During the compressive phase of the Bead-Cell mode, the molecular bond is not stretched, therefore kmol is not taken into consideration. The reciprocal of the ktot in this scenario (1/k1) is expressed as
(Eq. 3).
In the example data, kRBC is pre-determined (0.25 pN/nm by default). kcell can be derived from the Eq. 3 with the acquired k1 and kRBC (Figure 3B). - During the tensile phase, adhesion is formed between the ligand-receptor pair. Express the reciprocal of the ktot in this scenario (1/k2) as
(Eq. 4)
where k2 (Table 1) represents the total spring constant during the tensile phase. - Derive kmol from subtracting 1/k1 from 1/k2 (compare Eq. 3 vs. Eq. 4).
(Eq. 3).
(Eq. 4)4. Spring Constant Analysis of Bead- Bead Mode
- Fit a regression line to the compressive phase data to obtain k1 (similar to the Figure 2B, red). Of note, in the Bead-Bead mode, the Target cell is replaced by a glass bead coated with the receptor of interest (Figure 1C). Since bead deformation is negligible, the 1/kcell term can be removed from the Eq. 3 and Eq. 4 accordingly. The reciprocal ktot of the compressive phase (1/k1) can be expressed as:
(Eq. 5) - Fit a regression line to the tensile phase data to obtain k2 (similar to the Figure 2B, blue). The reciprocal ktot of the tensile phase (1/k2) can be expressed as:
(Eq. 6) - Derive kmol from subtracting 1/k1 from 1/k2 (compare Eq. 5 vs. Eq. 6).
(Eq. 5)
(Eq. 6)Subscription Required. Please recommend JoVE to your librarian.
Representative Results
In this work, we have demonstrated the protocol of the BFP spring constant analysis. For the Bead-Cell analysis mode, we analyzed the kmol of the molecular bond between the glycosylated protein Thy-1 coated on the Probe bead and the integrin α5β1 expressed on the Target K562 cell (Thy-1-integrin α5β1; Figure 3A)10. The kcell is also derived from the Bead-Cell mode (K562 Cell; Figure 3B). For the Bead-Bead mode, the molecular bond formed between fibrinogen and integrin αIIbβ3 (FGN-integrin αIIbβ3; Figure 3C)11,12 is used to demonstrate the Bead-Bead analysis mode.
For the Bead-Cell mode, we measured the spring constants of Thy-1-integrin α5β1 bond and K562 cell as kmol = 0.45 ± 0.28 pN/nm (Figure 3A) and kcell = 0.18 ± 0.07 pN/nm (Figure 3B) from 27 pre-screened analyzable events. For the Bead-Bead mode, we measured the spring constants of FGN-integrin αIIbβ3 bond as kmol = 0.53 ± 0.29 pN/nm (Figure 3C) from 33 pre-screened analyzable events.
| Symbol | Definition | Symbol | Definition |
| Δxtot | The total displacement of the piezo, which can also be interpreted as the total deformation of RBC, target cell and molecular bond. | ΔxRBC | The RBC deformation, which can also be interpreted as the displacement of the Probe bead. |
| ktot | The total spring constant of the entire BFP serial spring system. | kRBC | The spring constant of the aspirated RBC by the Probe micropipette. |
| kmol | The spring constant of the BFP detected molecular bond | kcell | The spring constant of the Target cell. |
| k1 | ktot of the compressive phase in the Retract stage. | k2 | ktot of the tensile phase in the Retract stage. |
| ΔF1 | The increment of force sensed by the Probe bead in the compressive phase. | ΔF2 | The increment of force sensed by the Probe bead in the tensile phase. |
| Δx1 | The increment of displacement in the compressive phase. | Δx2 | The increment of displacement in the tensile phase. |
Table 1. Symbol definitions for the BFP molecular spring constant analysis. Horizontal positions of all objects are defined as x, while Δx [nm] refers to the deformation relative to the original position. ΔF [pN] refers to the force increment measured by BFP. k [pN/nm] refers to the spring constant. Subscripts 1 and 2 correspond to the compressive and tensile phases, respectively. The molecular spring constant is derived from the Force (F) vs. Displacement (Δxtot) curve.
Figure 1: BFP configuration and DFS touch cycle. (A) The BFP system assembles two opposing micropipettes, namely the Probe (left) and the Target (right). The Probe micropipette aspirates a RBC (red) with a glass bead glued at its apex to serve as a force transducer. The Target micropipette aspirates a receptor-bearing cell (blue). The RBC spring constant (kRBC) is determined by the aspiration pressure (Δp) and the radii of aspirated RBC (R0), Probe micropipette (Rp) and circular contact area (Rc) between the RBC and the Probe bead. (B and C) Micrographs of the Bead-Cell (B) and Bead-Bead (C) BFP modes. Scale bars = 5 µm. (D) A BFP touch cycle that consists of Approach, Impinge, Contact, Retract and Dissociate stages. The ΔxRBC = 0 dash line indicates the BFP unstressed, or zero-force, position. The Retract stage includes compressive phase (ΔxRBC < 0, red), zero-force position (ΔxRBC = 0, black) and tensile phase (ΔxRBC > 0, blue) in sequence. Black arrow indicates the position of Bond event. Please click here to view a larger version of this figure.
Figure 2: Derive molecular spring constants from DFS raw data. (A) Representative Force (F) vs. Time (t) curve of DFS raw data in one BFP touch cycle. (B) Representative converted Force (F) vs. Displacement (Δxtot) curve that depicts the Retract stage. k1 and k2 represent the fit slope of the consecutive compressive and tensile phases respectively. ΔF1 and ΔF2 represent the increments of force in the compressive phase and the tensile phase data, respectively, where Δx1 and Δx2 represent the increments of displacement in the compressive phase and tensile phase data, respectively. R2 values for spring constant during compressive stage (R12) and tensile phase (R22) are labelled on the graph to indicate good statistical fitness. (C and D) Illustrations of the Retract stage in the Bead-Cell (C) and Bead-Bead (D) experimental modes. kRBC represents the spring constant of RBC; kcell and kmol represent the spring constants of the Target cell and the molecular bond, respectively. During the tensile phase, adhesion is formed between the ligand-receptor pair, the RBC deflects in the same direction as piezo retracts beyond zero-force position (ΔxRBC > 0). Please click here to view a larger version of this figure.
Figure 3: Representative histograms of BFP measured spring constants. The event number (left y-axis) and frequency distribution (right y-axis) of measured spring constants for Thy-1-integrin α5β1 bond (A) and K562 Target cell (B) in the Bead-Cell mode and FGN-integrin αIIbβ3 bond (C) in the Bead-Bead mode. Histograms are fit with Gaussian distribution curve (pink) and the statistical parameter, R2, is used to indicate the strength of fitness. Please click here to view a larger version of this figure.
Figure S1: The homemade BFP interface. (A) BFP control and parameter setting interface. The parameters to determine RBC spring constant are entered from the panel of biophysical parameters. (B) BFP monitoring. Live BFP touch cycles will be observed from this camera view. (C) BFP DFS analysis interface where the Force (F) vs. Time (t) curves are offline reviewed and pre-processed for subsequent molecular spring constant analysis. Please click here to download this File.
Figure S2. BFP analyzable data quality control and pre-screening criteria. (A) Good DFS events with high quality: (i) Bead-Cell rupture force event; (ii) Bead-Cell lifetime event; (iii) Bead-Bead rupture force event; (iv) Bead-Bead lifetime event. (B) Acceptable DFS events with some noise: (i) Data drifting but the zoom-in Retract stage remains valid; (ii)Slight data drifting after bond dissociation; (iii) Data kink at the zero-force regime; (iv) Holding force is small (< 10 pN). (C)Poor-quality events that should be discarded: (i) No adhesion; (ii) Data oscillation; (iii) Data drifting all the time; (iv) Discontinuous data; (v) Compressive force too small (≈ 0 pN); (vi) Multiple bonds; (vii) Invalid data with derived kmol < 0; (viii) Signal error. Zero force is indicated by the grey intermittent line. Please click here to download this File.
Subscription Required. Please recommend JoVE to your librarian.
Discussion
In summary, we have provided a detailed data analysis protocol for preprocessing the DFS raw data and deriving molecular spring constants in the BFP Bead-Bead and Bead-Cell analysis modes. Biomechanical models and equations required for determining molecular and cellular spring constants are presented. Albeit different integrins are studied, the kmol measured by the Bead-Bead mode and the Bead-Cell mode possesses significant range differences (Figure 3A vs. Figure 3C). Of note, with the Bead-Bead mode, the receptor is covalently linked to the glass bead. In contrast, with the Bead-Cell mode, the surface receptor is adapted by the underlying plasma membrane and cytoskeletons, which most likely influence the measured kmol.
Data quality control is crucial to ensure the reproducibility. To this end, we have implemented the DFS data pre-screening and outlier exclusion criteria on the Force vs. Time plots. To demonstrate this, a representative dataset was selected, in which we categorized the DFS raw data into three levels of quality: Good (Figure S2A), Acceptable with noise (Figure S2B), and Poor unacceptable (Figure S2C). For beginners of using the BFP, we recommend the strict criteria to pre-screen the data with the Good quality (Figure S2A). Of note, based on the data pre-screening criteria, the regression fit line of the compressive phase should be steeper than that of the tensile phase, specifically k1 > k2 (Figure S2C, vii). When measured k1 < k2 (Figure S2C, vii), the derived kmol < 0 is against the rationale per calculation in the step 4. Such events should then be considered as the invalid outliers and discarded.
To favor the BFP measurement on single-molecular level during data acquisition, multiple experimental configurations have been implemented according to previous study12. Firstly, protein coating density on beads is usually titrated down to a minimal level (e.g. 60 µm-2) by strictly control the solution concentration, quantity of the protein and reaction conditions15. The average spatial distance between proteins on the bead is thereby estimated much larger than the linear dimensions of the protein, favoring our measurements on single-molecular level12,13,14. Secondly, we control the adhesion frequency for each ligand-receptor pair ≤ 20%, under which molecular binding events will follow Poisson distribution predicting ≥ 89% of events would be single-molecular binding14,15. To achieve so, impingement force and contact time are set accordingly and need to be consistent throughout the experiment12. Nevertheless, it is still possible that multiple bonds occur sequentially (Figure S2C, vi). In such cases, we will discard the events with signatures of multiple bonds. Last but not least, negative control experiments will be performed with beads coated with bovine serum albumin (Table of Materials) or SA alone to ensure the non-specific adhesion frequency is ≤ 2%16,17.
Although BFP is powerful to investigate protein dynamics on living cell surface10,11,12, there are technical limitations. In BFP, only one ligand-receptor pair can be investigated at a time. It would be time consuming to obtain sufficient data with statistical significance. Besides, the experimental procedures are labor intensive with steep learning curves. Implementation wise, the current BFP system is susceptible to the environmental drifting and surrounding mechanical vibration. As a result, continuous manual adjustment is required to ensure the DFS data quality. To this end, one of our recent studies has introduced ultra-stable BFP feedback control algorithms to improve the stability of the BFP force-clamp DFS assays4. This technical advance enables measurements of stronger molecular interaction such as antigen-antibody binding with ultra-long bond lifetime (>50 s). Nevertheless, we foresee future efforts will be made to automate and integrate the BFP data acquisition and DFS analysis into one computerized program, making the entire BFP operation and data analysis more user friendly and high-throughput.
Subscription Required. Please recommend JoVE to your librarian.
Disclosures
The authors declare that they have no competing interests to report regarding the present study.
Acknowledgments
We thank Guillaume Troadec for helpful discussion, Zihao Wang for hardware consultation, and Sydney Manufacturing Hub, Gregg Suaning and Simon Ringer for support of our lab startup. This work was supported by Australian Research Council Discovery Project (DP200101970 - L.A.J.), NSW Cardiovascular Capacity Building Program (Early-Mid Career Researcher Grant - L.A.J.), Sydney Research Accelerator prize (SOAR - L.A.J.), Ramaciotti Foundations Health Investment Grant (2020HIG76 - L.A.J.), National Health and Medical Research Council Ideas Grant (APP2003904 - L.A.J.), and The University of Sydney Faculty of Engineering Startup Fund and Major Equipment Scheme (L.A.J.). Lining Arnold Ju is an Australian Research Council DECRA fellow (DE190100609).
Materials
| Name | Company | Catalog Number | Comments |
| 3-Mercaptopropyltrimethoxysilane (MPTMS) | Uct, Specialties, llc | 4420-74-0 | Glass bead functionalization |
| Anhy. Sodium Phosphate Dibasic (Na2HPO4) | Sigma-Aldrich | S7907 | Phosphate buffer preparation |
| BFP data acquisition VI | LabVIEW | BFP control and parameter setting | |
| BFP data analysis VI | LabVIEW | BFP raw data analysis | |
| Biotin-PEG3500-NHS | JenKem | A5026-1 | RBC biotinylation |
| Borosilicate Glass beads | Distrilab Particle Technology, Netherlands | 9002 | Glass bead functionalization |
| Bovine serum albumin | Sigma-Aldrich | A0336 | Ligand functionalization |
| Camera VI | LabVIEW | BFP monitoring | |
| D-glucose | Sigma-Aldrich | G7021 | Tyrode’s buffer preparation |
| Hepes | Sigma-Aldrich | H3375 | Tyrode’s buffer preparation |
| MAL-PEG3500-NHS | JenKem | A5002-1 | Glass bead functionalization |
| Potassium Chloride (KCl) | Sigma-Aldrich | P9541 | Tyrode’s buffer preparation |
| Sodium Bicarbonate (NaHCO3) | Sigma-Aldrich | S5761 | Carbonate/bicarbonate buffer preparation; Tyrode’s buffer preparation |
| Sodium Carbonate (Na2CO3) | Sigma-Aldrich | S2127 | Carbonate/bicarbonate buffer preparation |
| Sodium Chloride (NaCl) | Sigma-Aldrich | S7653 | Tyrode’s buffer preparation |
| Sodium Phosphate Monobasic Monohydrate (NaH2PO4•H2O) | Sigma-Aldrich | S9638 | Phosphate buffer preparation |
| Streptavidin-Maleimide | Sigma-Aldrich | S9415 | Glass bead functionalization |
References
- Chen, Y., et al. Fluorescence Biomembrane Force Probe: Concurrent Quantitation of Receptor-ligand Kinetics and Binding-induced Intracellular Signaling on a Single Cell. The Journal of Visualized Experiments. (102), e52975 (2015).
- Su, Q. P., Ju, L. A. Biophysical nanotools for single-molecule dynamics. Biophysics Reviews. 10 (5), 1349-1357 (2018).
- Ju, L. Dynamic Force Spectroscopy Analysis on the Redox States of Protein Disulphide Bonds. Methods in Molecular Biology. 1967, 115-131 (2019).
- An, C., et al. Ultra-stable Biomembrane Force Probe for Accurately Determining Slow Dissociation Kinetics of PD-1 Blockade Antibodies on Single Living Cells. Nano Letters. 20 (7), 5133-5140 (2020).
- Chen, Y., Ju, L., Rushdi, M., Ge, C., Zhu, C. Receptor-mediated cell mechanosensing. Molecular Biology of the Cell. 28 (23), 3134-3155 (2017).
- Ju, L., Chen, Y., Rushdi, M. N., Chen, W., Zhu, C. Two-Dimensional Analysis of Cross-Junctional Molecular Interaction by Force Probes. Methods in Molecular Biology. 1584, 231-258 (2017).
- Evans, E., Ritchie, K., Merkel, R. Sensitive force technique to probe molecular adhesion and structural linkages at biological interfaces. Biophysical Journal. 68 (6), 2580-2587 (1995).
- Ju, L., Zhu, C. Benchmarks of Biomembrane Force Probe Spring Constant Models. Biophysical Journal. 113 (12), 2842-2845 (2017).
- Evans, E., Ritchie, K., Merkel, R. Sensitive Force Technique to Probe Molecular Adhesion and Structural Linkages at Biological Interfaces. Biophysical Journal. 68, 2580 (1995).
- Fiore, V. F., Ju, L., Chen, Y., Zhu, C., Barker, T. H. Dynamic catch of a Thy-1-alpha5beta1+syndecan-4 trimolecular complex. Nature Communications. 5, 4886 (2014).
- Passam, F., et al. Mechano-redox control of integrin de-adhesion. Elife. 7, (2018).
- Chen, Y., et al. An integrin alphaIIbbeta3 intermediate affinity state mediates biomechanical platelet aggregation. Nature Materials. 18 (7), 760-769 (2019).
- Chen, Y., Lee, H., Tong, H., Schwartz, M., Zhu, C. Force regulated conformational change of integrin αVβ3. Matrix Biology. 60, 70-85 (2017).
- Liu, B., Chen, W., Zhu, C. Molecular force spectroscopy on cells. Annual Review of Physical Chemistry. 66, 427-451 (2015).
- Piper, J. W., Swerlick, R. A., Zhu, C. Determining force dependence of two-dimensional receptor-ligand binding affinity by centrifugation. Biophysical Journal. 74 (1), 492-513 (1998).
- Ju, L., Dong, J. -f, Cruz, M. A., Zhu, C. The N-terminal flanking region of the A1 domain regulates the force-dependent binding of von Willebrand factor to platelet glycoprotein Ibα. Journal of Biological Chemistry. 288 (45), 32289-32301 (2013).
- Ju, L., Chen, Y., Xue, L., Du, X., Zhu, C. Cooperative unfolding of distinctive mechanoreceptor domains transduces force into signals. Elife. 5, 15447 (2016).
