Microtubule Dumbbells to Assess the Effect of Force Geometry on Single Kinesin Motors

Abstract

The cytoskeletal motors myosin, kinesin, and dynein and their corresponding tracks, actin and microtubules, are force generating ATPases responsible for motility and morphological changes at the intracellular, cellular, and tissue levels. The pioneering application of optical tweezers to measure the force-producing properties of cytoskeletal motors has provided an unparalleled understanding of their mechanochemistry. The mechanosensitivity of processive, microtubule-based motors has largely been studied in the optical trap using the “single-bead” assay, where a bead-attached motor is held adjacent to a cytoskeletal filament as it processively steps along it. However, because of the geometrical constraints in the conventional single-bead assay, the motor-filament bond is not only loaded parallel to the long axis of the filament, but also perpendicular to the long axis of the filament. This perpendicular force, which is inherent in the conventional single-bead assay, accelerates the motor-filament detachment and has not been carefully considered in prior experiments. An alternative approach is the “three-bead” assay, which was developed for the study of non-processive myosin motors. The vertical force component is minimized in this assay, and the total opposing force is mainly parallel to the microtubule. Experiments with kinesin show that microtubule attachment durations can be highly variable and last for up to tenfold longer times in the three-bead assay, compared to the single-bead assay. Thus, the ability of kinesin to bear mechanical load and remain attached to microtubules depends on the forces in more than one dimension. In this chapter, we provide detailed methods for preparing the proteins, buffers, flow chambers, and bead-filament assemblies for performing the three-bead assay with microtubules and their motors.

1. Introduction

Kinesins are a family of molecular motors (45 genes in humans) that power processive motility (ability to take multiple steps before dissociation) of intracellular organelles along microtubules [1–3]. They also play roles in membrane sorting, formation of membrane tubules, and the contraction of dynamic microtubule networks (e.g., mitotic spindle) [2, 4]. A mechanistic understanding of cytoskeletal motors requires quantitative measurements of the displacements and forces they generate, and experimental knowledge of how stepping behavior is affected by different external force geometries relative to their cytoskeletal track. The development of optical tweezers, and its pioneering application to answer mechanistic biological questions, has revealed a wealth of information on the magnitude of forces developed by single kinesin motors and their stepping kinetics under resisting or assisting forces over the past 25 years [5–17]. Optical trapping is still the only technique that allows study of the mechanochemical cycle of single molecules by very finely controlling the magnitude of external forces with sub-pN resolution and by measuring displacements with sub-nm and sub-ms resolution.

The single-bead optical trapping assay has been the standard in the kinesin field for assessing the activity of processive cytoskeletal motors (Fig. 1a). Motors are attached to a polystyrene bead (~1.0 μm diameter) that is held by a laser trap (1064 nm). The concentration of the motors is titrated so that the probability of a bead carrying two or more motors is <0.05 [14, 17–22], resulting in single-molecule conditions. As the motor moves processively along the microtubule track, it pulls the bead away from the stationary-trap center. The laser trap acts as a soft linear spring (~0.04 pN/nm) that develops opposing forces as the motor advances along the surface-attached microtubule until it detaches. One can record the stepping behavior along with the generated force as a function of time, where force increases with run length (Fig. 1). The single-bead assay can also be performed in feedback mode, where the position of the optical trap follows the bead at a constant force and allows more accurate assessment of the force-velocity and force-run length relationships of the motors [6, 15, 21].

Fig. 1.

Single- and three-bead assays. Cartoon representations (not drawn to scale) are shown of (a) the single-bead assay and (b) the three-bead assay. (a) A bead-bound kinesin molecule is brought into contact via optical tweezers with a surface-immobilized microtubule on a glass coverslip via antibody. (b) A microtubule is attached to two laser-trapped beads via streptavidin-biotin linkage. The microtubule assembly (microtubule dumbbell) is brought in contact with a single kinesin molecule, attached on a surface-immobilized spherical pedestal. Representative raw (gray) and smoothed (black) force and displacement traces of a single kinesin molecule interacting with (c) a surface-immobilized microtubule in the single-bead assay (dia. 0.82 mm) and (d) a microtubule dumbbell in the three-bead assay. The force (F_detach) at which the kinesin detaches (red arrow in (c)) from the microtubule and the duration Δt of the corresponding force ramp (double red arrow in (c)) can be calculated directly from the data for every force ramp. (Figure modified from [18])

Kinesin-1 (KIF5B) is a widely expressed motor crucial for the transport and deformation of a large variety of membranes and organelles, and it is one of the most extensively studied cytoskeletal motors [18, 20, 21]. Biochemical, imaging, and optical trapping experiments have shown that a single, two-headed, KIF5B molecule steps in 8 nm increments, hydrolyzes one ATP molecule per 8 nm step, stalls at forces of ~5 pN, and detaches faster (<1 s) under opposing force rather than in the absence of force (~1 s) [6, 23, 24]. However, discrepancies in studies focused on understanding how teams of kinesin motors work have raised questions about the force-dependent microtubule-detachment kinetics of kinesin obtained from single-bead optical trapping assays [25]. Specifically, it has been proposed that single-bead assays report accelerated microtubule-detachment rates [25].

According to recent theoretical work, in the conventional single-bead assay, in addition to the force component parallel to the microtubule track (horizontal component, F_x), there is a substantial repulsive force component, F_z, vertical to the microtubule track. The magnitude of this force depends on the bead diameter and the motor-tether length for any given trap stiffness and loaded run length of the motor (Fig. 2a) [24, 25].

Fig. 2.

Kinesins detaches faster for larger size beads in the single-bead assay. (a) Cartoon representation of the single-bead assay for two different sizes of beads, R₁ and R₂ = 4R₁, and the same displacement along the microtubule relative to the center of the laser trap. The length L of the attached kinesin construct (inset) is drawn almost to scale relative to R₁ and R₂. The vectors of the total force F and its components (F_x, F_z) that oppose kinesin’s movement are represented by the red arrows. Under mechanical equilibrium, the total force F should be opposite to the force produced by kinesin and lay along the radius of the bead. For the same stiffness of the trap F_1x = F_2x and F_1z < F_2z, (see Eq. (1) and Table 1). (b) Scatter plot of the detachment forces F_detach and attachment durations Δt for beads with three different diameters of 0.51 μm (gray), 0.82 μm (black), and 2.1 μm (red). The cumulative distributions of F_detach and Δt are plotted as solid lines of corresponding colors along the right and top axis of the graph, respectively. (Figure modified from [18])

Importantly, since F_z points away from the filamentous track, it may accelerate the detachment of the motor, ending its processive run. The balance of forces and their associated torques suggests that the force vectors applied on the bead by the laser trap and the motor are along the direction that connects the center of the bead with the point of attachment of the motor on the bead and therefore there are two orthogonal force components, F_x and F_z, for each force vector (Fig. 2a). The magnitude of the vertical force component F_z is expected to scale with the radius R of the bead as:

$$F_z=\frac{F_x}{\sqrt{{\left(1+\frac{L}{R}\right)}^2-1}}$$ (1)

where L is the combined length of the motor and any accessory protein used to anchor the motor on the bead. Representative values of F_x and F_z for three different values of bead radius and assuming that L ~ 35 nm are given in Table 1.

Table 1.

Single-bead assay measurements of forces and attachment durations

Bead diameter (μm)	〈Fdetach〉 (pN)	Median-Δt (s)	Fstall (pN)	% Stall events	aFz/Fx
0.51	3.1 ± 1.2	0.44	4.4 ± 0.81	30	1.8
0.82	2.9 ± 1.0	0.33	4.9 ± 0.43	5	2.4
2.1	2.5 ± 0.74	0.31	NA	0	3.8

^aCalculation of the F_z/F_x according to Eq. (1) assuming F_x = 5 pN and L = 35 nm

We explored the effect of the vertical force F_z on KIF5B microtubule attachment duration during processive movement in a stationary single-bead optical trapping assay [18]. Three different bead diameters (2.1, 0.82, and 0.51 μm) were used in the assay. The distributions of attachment durations and detachment forces F_x shift toward lower values for increasing bead sizes. The differences between the smallest and largest bead size are statistically significant ( p < 0.001, non-parametric Mann-Whitney test) (Fig. 2b). These data therefore support the theoretical prediction that the detachment of kinesin in the single-bead assay is accelerated by the vertical force component, F_z.

In contrast to the single-bead assay, the F_z force component on microtubule-bound kinesin is expected to be negligible when using a dual-beam optical tweezers configuration in the three-bead assay. This assay was initially developed for investigating the mechanochemistry of non-processive myosin–actin interactions [27], and it has been used only less frequently for studying microtubule-based motors [28–30]. The microtubule is suspended between two beads held by two independently controlled optical traps and is brought into contact with single kinesin motors immobilized on pedestal beads affixed to the surface of the experimental chamber (Fig. 1b). The bead-microtubule-bead assembly is referred to as a microtubule dumbbell. The median attachment duration between KIF5B and microtubules in the three-bead assay was strikingly longer than observed in the single-bead assay, under identical assay conditions and similar trap stiffness (Fig. 1). In the three-bead assay, kinesin develops high forces (≥5 pN) more frequently and sustains its microtubule attachment for longer periods of time relative to the single-bead assay (Fig. 3a, b).

Fig. 3.

Kinesin’s attachment durations and detachment forces for single- and three-bead assays. Cartoon representations for each assay are shown on the top panel of plots (a–c). Representative distributions of Δt and F_detach along with their corresponding box statistics in the middle and bottom panels, respectively, for (a) different pairs of single beads (“a” to “g”) and surface-immobilized microtubules, (b) different microtubule dumbbells (“a′” to ”g′”), and (c) different pairs of single beads (“h” to “n”) and microtubules suspended between rectangular ridges. For each box statistics representation, the length corresponds to the interquartile range (IQR), the error bars to SD, the midline to the median value, and the square inside each box to the average value. (d) Shown are the distribution and box statistics of the median-Δt for single kinesin molecules interacting with microtubule dumbbells (n = 50) and surface-immobilized microtubules (n = 20) under resisting load. The green points correspond to the examples shown in plot (b). The black lines, which serve as a guide to the eye, represent a normal distribution with the same mean and SD as the corresponding data. (Figure modified from [18])

To explore the possibility that the surface attachment of the microtubule in the conventional single-bead assay contributes to the different attachment times between the single- and three-bead assays, we performed single-bead experiments on microtubules suspended between parallel rectangular ridges (1 μm tall, 2 μm wide, and spaced 10 μm apart) from each other (Fig. 3c). The median attachment durations and forces were similar to the values measured using the conventional single-bead assay (Fig. 3a, c).

The three-bead assay also revealed substantial variability in kinesin’s attachment durations among different dumbbells (N = 50). Strikingly, durations varied by more than tenfold in the range of 0.1–3.8 s, while the maximum value observed for the single-bead assay was 0.4 s (Fig. 3d). We propose that the variability depends on the relative angular position of the forces developed along the circumference of the microtubule upon engagement with kinesin described in [18]. The mechanism by which the relative angular position of forces around the circumference of the microtubule affects attachment durations is unknown, but we favor a model related to force-induced structural plasticity of the microtubule lattice that leads to allosteric changes in the kinesin binding site [31–35]. This plasticity may be revealed by the variability in the geometry of tensile forces applied across the microtubule since the dumbbell beads can bind to different protofilaments (Fig. 4). Alternatively, clustering of posttranslational modifications of tubulin along different microtubule protofilaments could lead to this variability.

Fig. 4.

Relative geometry of the forces developed in the three-bead assay. (a) Kinesin is stepping toward the plus end (right) of the microtubule and it therefore pulls the dumbbell toward the opposite direction (left) by applying a force F_k on the interacting protofilament and along the microtubule axis. A net opposing restoring force F_trap mainly along the microtubule axis is developed by the stationary laser beams via the trapped beads attached on the microtubule. (b) Cross-sectional view of relative position of the beads and the interacting kinesin with the dumbell in (a). The relative azimuthal position φ between the attachment point of the plus end bead and that of interacting kinesin. The major component of F_k is directed vertically toward the back of the page (⊗) and the major component F_trap toward the front. The variability of azimuthal separation of the opposing forces among various dumbbells reflects the variability in the geometry of forces in the three-bead assay. (Figure modified from [18])

We provide detailed instructions for experimentalists on how to prepare and conduct the single-bead and three-bead experiments on continuous surfaces and nanopatterned rectangular ridges. In the procedure described below, we used a 6xHis-GFP-labeled truncated human kinesin-1 heavy chain construct (aa 1–560 [36]) with a C-terminal AviTag sequence that can be site-specifically biotinylated [37]. For expression and purification of the construct see [38].

2. Materials

2.1. Casein Solution

1. Casein powder (Sigma, Cat No C5890–500G).

2. Ultracentrifuge L8–80 M Beckman and rotor Ti 45 or Sorvall RC 5B Plus and rotor SS-34.

3. Buffer solution of 20 mM HEPES, 200 mM NaCl, pH 8.8 (using NaOH).

4. Nalgene rapid-flow sterile bottle-top filters with 0.2 μm pore size (Thermo Scientific Cat No 09–740-22D).

2.2. Microtubule Preparation

1. Unlabeled porcine tubulin (Cytoskeleton, Cat No T240-B).

2. Biotinylated porcine tubulin (Cytoskeleton, Cat No T333P-B).

3. Rhodamine labeled porcine tubulin (Cytoskeleton, Cat No TL590M-B).

4. Paclitaxel, known as Taxol.

5. High purity (>99.7%) sterile-filtered DMSO (Dimethyl-Sulfoxide).

6. BRB80 buffer: 80 mM PIPES, 2 mM MgCl₂, and 1 mM EGTA, pH 6.9.

7. GTP.

8. GMPPCP.

9. Tabletop Optima TL ultracentrifuge and rotor TLA100 (Beckman).

10. Open-top Thickwall Polycarbonate tubes 7 × 20 mm from Beckman Coulter (Cat No 343775).

11. Water bath at 37 °C.

12. Tabletop centrifuge (e.g., Eppendorf 5430R) at 20–25 °C.

2.3. Coating of Microspheres with Anti-His Antibodies and Kinesin for the Single-Bead Assay

1. Tabletop centrifuge (e.g., Eppendorf 5430R) cooled to 4 °C with rotor for 1.5-mL polypropylene tubes.

2. Low-power bath sonicator (e.g., Ultrasonic Cleaner 50/60 Hz, 80 Watts from Branson, model B12).

3. Streptavidin or Neutravidin coated polystyrene beads of diameter 0.51 μm and 1.0% w/v, dia. 0.82 μm and 1.0% w/v and dia. 2.1 μm and 0.5% w/v (Spherotech, Cat No SVP 05 10, SVP 08 10 and NVP-20–5, respectively).

4. BRB80 buffer: 80 mM PIPES, 2 mM MgCl₂, and 1 mM EGTA, pH 7.4.

5. 10 mg/mL casein solution (see Subheadings 2.1 and 3.1).

6. Biotinylated Anti-His antibody (Qiagen, Cat No 34440) stock solution (0.2 mg/mL), store at 4 °C.

7. d-Biotin (Avidity, Aurora, CO).

8. Syringe-filters 0.2 μm (Millipore, Cat No 09–720-004).

2.4. Glucose Oxidase and Catalase (POC) Oxygen Scavenger Preparation

1. Glucose oxidase from Aspergillus niger, (Sigma Aldrich, Cat No G2133–250KU).

2. Aqueous solution of catalase from bovine liver (Sigma Aldrich, Cat No C100–500MG), store at 4 °C.

3. d-(+)-Glucose.

4. Phosphate Buffer Saline (PBS): 37 mM NaCl, 2.7 mM KCl, 8 mM Na₂HPO₄, and 2 mM KH₂PO₄, pH 7.4.

2.5. Nanopatterning of Rectangular Ridges

1. Chrome master form of 4 × 4 mm with a pattern of rectangular parallel ridges 2 μm wide, 1 μm tall, and separated by 10 μm (purchased from TECHNO-NT, Netherlands).

2. Silicone elastomer base (184 Sylgard Part A), and the corresponding elastomer curing agent (184 Sylgard Part B) product of Dow Corning (Midland, MI) with Cat No 04019862 was purchased by Ellsworth Adhesives.

3. 2% Dimethyldichlorosilane in octamethylcyclooctasilane (PlusOne Repel-Silane ES) (GE Healthcare, Cat No 17–1332-01).

4. Optical adhesive Norland 65 manufactured by Norland Products Inc. (Cranbury, NJ) was purchased from Thorlabs (Cat No NOA65) (see Note 1).

5. Heavy duty glass desiccator vacuum 230 mm diameter connected to an oil vacuum pump in a fume hood.

6. General-purpose heating and drying oven (e.g., Fisherbrand Isotemp).

7. UV transilluminator (e.g., Spectroline model TC-312 E).

8. Micro slides 75 × 25 mm (e.g., Corning Cat No 2948).

9. Scotch Super glue or any similar strong and fast drying glue.

10. Acetone (purity ≥99.5%).

11. 50-mL screw-top cup conical tube (Corning, Tewksbury, MA).

2.6. Silica Sphere Suspension for Use as Pedestals in the Three-Bead Assay

1. Silica spheres of 5 μm in diameter 9.92% solid (Polysciences Inc., Cat No 24332).

2. Methanol (purity ~99.8%).

3. Glass storage vials 12 mL with polypropylene screw cups (Thermo Fisher 03–391-10D).

4. Amyl acetate (Electron Microscopy Sciences Cat No 10815).

2.7. Chamber Preparation for the Optical Tweezers Assay

1. Dual-beam optical tweezers setup with an Nd:YAG laser (1064 nm, Spectra Physics) conjugated to an optical microscope with a 63× water objective, a piezoelectric stage and two quadrant photo detectors (QPD) to detect the deflection of the two laser beams [39].

2. Glass coverslips 22 × 40 × 1.5 mm (Fisher Scientific, Cat No 12–544-B).

3. Sterile collodion (nitrocellulose) 2% in amyl acetate and amyl acetate (Electron Microscopy Sciences, Cat No 12620–50 and 10815, respectively).

4. High-vacuum grease (Dow Corning, Cat No 2021846–0116).

5. Double-sided tape Scotch (e.g., 3M Cat No 34–8716-0599–3).

6. Filter paper (Whatmann, Cat No 10427806).

7. Rubber mat to use as a surface for cutting.

8. Razor blades product (Private Brands, Cat No 94–120-71).

9. 3 mL BD disposable syringe (Fisher Scientific, Cat No 14–823-435) and single use needle (Fisher Scientific, Cat No 14–821-15A).

2.7.1. Beads and Material for the Single-Bead Assay

1. Mouse anti-tubulin b3 antibody (Bio-Rad Laboratories, Cat No MCA2047).

2. Beads and Material with bound kinesin motor (see Subheadings 2.3 and 3.3).

3. Taxol stabilized microtubules made from 98% unlabeled porcine tubulin and 2% TRITC Rhodamine labeled porcine tubulin (see Subheadings 2.2 and 3.2).

4. BRB80 buffer: 80 mM PIPES, 2 mM MgCl₂, and 1 mM EGTA, pH 7.5.

5. ATP, 100 or 200 mM aliquots in DI water, pH 7.5 with KOH (Sigma Aldrich, Cat No A2383).

2.7.2. Beads and Material for the Three-Bead Assay

1. Suspension of silica spheres of 5.0 μm diameter in amyl acetate used for pedestals to anchor kinesin (see Subheadings 2.6 and 3.6).

2. Mouse monoclonal anti-6xHis tag antibody (Abcam, Cat No ab18184, 1 mg/mL).

3. Streptavidin coated beads of 0.82 μm diameter used to make the bead-microtubule-bead assembly (Spherotech, Cat No SVP 08 10).

4. Taxol (or GMPCPP without Taxol) stabilized microtubules made from 48% unlabeled porcine tubulin, 48% biotinylated porcine tubulin and 4% TRITC Rhodamine labeled porcine tubulin (see Subheadings 2.5 and 3.5).

5. BRB80 buffer: 80 mM PIPES, 2 mM MgCl₂, and 1 mM EGTA, pH 7.5.

6. ATP, 100 or 200 mM aliquots in DI H₂O, pH 7.5 with KOH (Sigma Aldrich, Cat No A2383).

3. Methods

3.1. Casein Solution Preparation

1. Make 50 mL of a 20 mg/mL casein solution in 20 mM HEPES, 200 mM NaCl, pH 8.8 (NaOH) at room temperature in a glass beaker with a magnetic stirring bar.

2. Using a stir plate, stir gently overnight (~10 h) at 4 °C or for 6–8 h at RT.

3. Remove insoluble particles by centrifuging at 34,000 × g at 4 °C for 20 min.

4. Collect supernatant avoiding the loose pellet and check pH. If pH is in the range 6.9–7.1, the casein has appropriately dissolved otherwise repeat the procedure ensuring that the initial buffer conditions are correct.

5. Filter the solution using rapid-flow sterile bottle-top filters with 0.2 μm pore size (see Note 2).

6. Measure the concentration using Bradford assay. The final concentration should be 8–14 mg/mL.

7. Store 0.8 mL aliquots at −80 °C (see Note 3).

3.2. Microtubule Preparation

1. Dissolve lyophilized tubulin to 5 mg/mL in BRB80 (pH 6.9) supplemented with 1 mM GTP following product instructions from Cytoskeleton Inc.

2. 20 μL of unlabeled tubulin solution is mixed with 20 μL of biotinylated tubulin and 2 μL of TRITC tubulin. For the single-bead assay, biotinylated tubulin should be replaced with unlabeled tubulin. Tubulin solutions are centrifuged at 300,000 × g at 4 °C for 10 min using a TLA-100 rotor (Beckman Coulter).

3. Tubulin in the supernatant is polymerized by incubating in the dark in a 37 °C water bath for 20 min.

4. Add 1.75 μL Taxol (2 mM) and place back at 37 °C for another 10 min (see Note 4).

5. Add 1 μL Taxol (2 mM) and incubate for another 10 min at 37 °C.

6. Polymerized tubulin solution is centrifuged at 40,000 × g at 25 °C for 20 min.

7. Supernatant is discarded, and the microtubule-containing pellet is dissolved by rigorous pipetting in 50 μL of BRB80 (pH 6.9) supplemented with 40 mM Taxol final concentration.

8. Microtubules are kept at room temperature away from light and used within 5 days of preparation.

9. For GMPCPP-stabilized microtubules, GTP and Taxol are both excluded from the above protocol. Rather, 5 μL of 10 mM GMPPCP is added just before incubation of tubulin solution in a water bath at 37 °C, and tubulin is polymerized for 30 min.

3.3. Coating Microspheres with Anti-6xHis Antibodies and Kinesin for the Single-Bead Assay

1. Mix 10 μL streptavidin beads of 0.82 μm diameter with 40 μL BRB80 (pH 7.5) and bath sonicate for 1 min (see Note 5).

2. Add 0.5 μL biotinylated His-Ab to 49.5 μL BRB80 (pH 7.5), resulting in 50 μL of 2 μg/mL His-Ab. Mix with 50 μL of bath sonicated beads to a final volume of 100 μL. Incubate for at least 1 h at 4 °C on a rotating plate.

3. Add d-Biotin to final concentration of 2.6 μM and incubate for at least 1 h at 4 °C on a rotating plate (see Note 6).

4. Wash three times with 100 μL of 2 mg/mL casein for 5 min, centrifuging each wash at 6000 × g at 4 °C. Mix 9 μL of bead suspension with 1 μL of the appropriate concentration of kinesin (see Note 7) and incubate under constant rotation at 4 °C for at least 4 h or overnight (see Note 8).

5. When ready to do the single-bead assay, dilute 2 μL of suspension to a final volume 50 μL and inject into the experimental chamber (see Subheading 3.7.1).

3.4. Glucose Oxidase and Catalase (GOC) Oxygen Scavenger Preparation

1. Prepare 16 μL aliquots of 21.1 U/μL glucose oxidase in PBS and store at −80 °C.

2. Prepare 50 μL aliquots of 250 mg/mL glucose in DI H₂O and store at −20 °C.

3. The day of the experiment thaw a 16 μL aliquot of glucose oxidase and add 4 μL of catalase stock solution (~702 U/μL) such that the final concentrations of glucose oxidase and catalase become 16.9 U/μL and 140 U/μL, respectively.

4. Glucose oxidase/catalase and glucose solutions should be diluted 100- and 50-fold, respectively, in the final solution introduced in the experimental chamber (see Note 9).

3.5. Nanopatterning of Rectangular Ridges

1. Glue the master mask (Fig. 5a) with the desired pattern of rectangular parallel ridges (4 mm long, 2 μm wide, 1 μm tall, and 10 μm apart) faceup at the middle of a glass micro slide 75 × 25 mm using superglue.

2. Mix nine parts of silicone elastomer base to one part of the corresponding elastomer curing agent in a 50-mL screw-top cup conical tube and gently mix by reversing the tube multiple times (see Note 10). Let the mixture stand undisturbed for 15 min.

3. Place the micro slide with the master mask in a Pyrex borosilicate glass petri plate (Fig. 5b). Working in a fume hood, pour 30 mL of 2% dimethyldichlorosilane solution into a 50 mL glass beaker. Place both the petri dish and the beaker next to each other inside a glass vacuum desiccator under a fume hood. Avoiding bubbles, pipet 5 mL of 2% dimethyldichlorosilane on the mask and apply a hard vacuum for 30 min (see Note 11).

4. After releasing the vacuum, pour gently the elastomer solution on top of the master mask in the glass plate avoiding the formation of bubbles. Apply hard vacuum for another 30 min to remove trapped air (see Note 12).

5. To speed elastomer curing, place the glass plate at 105 °C for 1 h in a general-purpose heating oven.

6. Let cool at room temperature (~20 min) and using a razor blade remove the cured elastomer around the master mask such that a cube or trapezoid of cured elastomer remains on top of the master mask. Use tweezers with flat ends to remove the trapezoid. This trapezoid will serve as a stamp to form rectangular ridges on glass coverslips in the subsequent steps (Fig. 5c) (see Note 13).

7. Elastomer stamps must be exposed to ultraviolet (UV) plasma for no more than 4 s and then placed in a petri dish face-up. Dimethyldichlorosilane (5 μL of 2%) is placed on each stamp and left under the hard vacuum for 30 min in the presence of a glass beaker containing 30 mL of 2% dimethyldichlorosilane solution (see Note 14).

8. Stamps should be placed gently facedown against ~4 μL of optical adhesive on glass coverslips 22 × 40 × 1.5 mm. Push gently against Norland to push any bit of excess from under the stamp and let stand for 5 min. Make sure to stamp the ridges so that their long axes will be perpendicular to the fluid flow channel through the chamber.

9. To cure the optical adhesive, coverslips are exposed to UV light for 10 min using a UV transilluminator.

10. After gently removing the stamps using tweezers, the cured optical adhesive on each coverslip is washed with 0.5 mL of methanol and allowed to dry under a fume hood (see Note 15).

11. Coverslips with cured optical adhesive are used within 24 h of preparation. A differential interference contrast microcopy image of rectangular ridges as well as a fluorescence microscopy image of immobilized microtubules on top of the ridges are shown in Fig. 6.

Fig. 5.

PDMS stamp for rectangular parallel ridges. (a) Image of the master mask (~5 × 5 mm). (b) Master mask is glued faceup on a glass slide (75 × 22 mm) which is placed in a pyrex petri dish. (c) The PDMS trapezoid with the imprinted pattern on it which will serve as a stamp to produce rectangular ridges on Norland 65

Fig. 6.

Rectangular parallel ridges. (a) A differential interference contrast microscopy image of rectangular parallel ridges imprinted in optical adhesive Norland 65 on a glass coverslip. The width of the ridges is 2 μm and spacing between them 10 μm. (b) A fluorescence microscopy image (544 nm) showing surface-attached fluorescently labeled microtubules attached between the ridges shown in (a)

3.6. Preparation of Silica Sphere Suspension To Be Used as Pedestals in the Three-Bead Assay

1. 2.5 mL of 2% silica beads (5 μm diameter) are washed three times in 1 mL methanol at 800 × g for 2 min at room temperature using a benchtop centrifuge.

2. Remove methanol and allow the beads to dry under a fume hood for at least 2 h.

3. Beads are diluted in 5 mL of amyl acetate and transferred to a glass screw cup glass vial for long-term use and stored at 4 °C.

3.7. Chamber Preparation for the Optical Tweezers Assay

1. For the single-bead assay, mix 5 μL of nitrocellullose (2%) with 95 μL of amyl acetate in a 0.5 mL polypropylene tube. For the three-bead assay, in addition to the above solution, mix 5 μL of nitrocellullose (2%) with 85 μL of amyl acetate and 10 μL of 5 μm silica sphere suspension in amyl acetate in a 0.5 mL polypropylene tube (see Note 16).

2. Place 4 μL of the above solution using a pipette across the width of the coverslip (Fig. 7a) and spread it up and down across the coverslip with the tip as shown in Fig. 7b, c (see Note 17).

3. Let the coverslips dry at room temperature either at inclination in an open petri dish or in a ceramic or metal coverslip holder and let them dry for at least 10 min (see Note 18).

4. Place the dried coverslip with nitrocellulose-coated surface faceup on a rubber mat. Cover the two short sides of the coverslip with double-sided tape as shown in Fig. 7d.

5. Use a razor blade to cut the parts of the double-sided tape that extend beyond the coverslip and are adhered on the rubber mat (Fig. 7e) and remove.

6. Remove the plunger of a 3 mL disposable syringe and fill it with vacuum grease. Replace the plunger and add a needle on the syringe. For safety purposes, cut the point from the needle so that it is blunt. The vacuum grease in the syringe can be used for multiple experiments and can be refilled as needed.

7. Use the syringe to apply vacuum grease parallel to the edges of the tape but without touching it (Fig. 7f).

8. On top of the coverslip, place another coverslip with its nitrocellulose-coated surface facedown and with the long axes perpendicular to each other (Fig. 7g).

9. With the blunt side of the metal razor press gently against the top coverslip. Compress the vacuum grease to the point where the top coverslip adheres to the double-sided tape of the bottom coverslip.

10. Flip the assembly to allow injection of solutions into the flow chamber. For the three-bead assay, the chamber can be made by nitrocellulose-coated coverslips with immobilized spherical pedestals on the bottom coverslip or on both sides.

11. In experiments with rectangular ridges, one of the coverslips contains the optical adhesive with the imprinted rectangular ridges. Be sure that the rectangular ridges are oriented in a direction vertical to the direction of the chamber’s flow channel (see Note 19).

Fig. 7.

Preparation of the experimental chamber. (a) Spread 4 μL of 0.1% nitrocellulose in amyl acetate along the shorter dimension of the glass coverslip as indicated by the dashed black line which serves as a guide to the eye. Spread the nitrocellulose solution using the conical surface of the tip and sliding it (b) up and (c) down along the long dimension of the coverslip. Then let it faceup on the side of a petri dish to dry (on the right in photos (b) and (c)). (d) Place the dried coverslip with its nitrocellulose-coated surface faceup on a rubber mat using tweezers or forceps. Immobilize the coverslip (outlined in red color) on the rubber mat with two stripes of double-sided tape (outlined in black color) as shown in the photo. (e) With a razor blade cut the double-sided tape (outlined in black color) at all sides around the coverslip and remove the parts of it that adhere on the rubber mat. (f) Place vacuum grease across the short dimension of the coverslip close, but not right next, to the double-sided tape as shown in the photo. (g) Place another coverslip with its nitrocellulose-coated surface facedown on top of the coverslip with the vacuum grease and the double-sided tape, such than the long axes of the two coverslips are vertical to each other. (h) With the blunt side of the razor blade press gently the top coverslip so that the vacuum grease spreads until the two coverslips are stably connected by the double-sided tape. (i) The formation of the flow channel and its two entries are indicated by the two black arrows

3.7.1. Single-Bead Assay

1. Introduce solutions into the chamber in the following sequence: (1) 20 μL of 0.05 mg/mL anti-tubulin antibody for 5 min; (2) 50 μL of 2 mg/mL casein for 4 min; (3) 50 μL of 125 nM microtubules supplemented with 2 mg/mL casein and 10 μM Taxol for 4 min (see Note 20); (4) wash with 100 μL of 2 mg/mL casein; (5) flow 50 μL of final solution with kinesin-coated beads (diluted at 1:25 from stock, see Subheading 3.3, step 5) containing 2 mM ATP, 2 mM MgCl₂, 50 mM DTT, 20 mM Taxol, 5 mg/mL glucose, 168.8 units/mL of glucose oxidase, and 1404 units/mL of catalase; (6) seal the open ends of the chamber with vacuum grease to prevent evaporation during the experiment (see Note 21).

2. Place the chamber on the microscope and use the camera and fluorescence illumination to find the bottom of the chamber.

3. Wear protective goggles and turn on the 1064-nm laser. Trap a single bead and record the variations in the force signal on the QPDs due to thermal fluctuations at a rate >20 kHz of the bead at least 5 μm away from the coverslip surface to avoid boundary effects (see Note 22).

4. Calculate the power spectrum of the thermal motion, and determine the Voltage to pN conversion factor, the corner frequency, and subsequently the trap stiffness [40–42] (see Note 23).

5. Using the cursor, mark on the monitor the bead position. Dim or block the transmitted light and allow fluorescent illumination (544 nm). Use the piezo stage to bring the bead into contact with a coverslip-bound microtubule (see Note 24). Move the stage in small steps (~25 nm) back and forth along the diameter of the microtubule to screen for interactions. If interactions are not detected within 30 s, trap another bead, and repeat with at least ten different beads until an interaction is detected. For single-molecule interactions, ≤3 out of ten beads should interact with the microtubule to increase the probability of single-molecule conditions. Recording data at 2 kHz and filtering at 1 kHz (Nyquist frequency) is sufficient to reliably detect interactions between kinesin and microtubules.

3.7.2. Three-Bead Assay

1. Introduce solutions into the chamber in the following sequence: (1) 20 μL of 0.2 mg/mL anti-6xHis antibody (Abcam) for 5 min; (2) 50 μL of 2 mg/mL casein for 4 min; (3) 50 μL of kinesin-1 construct ≤1 nM supplemented with 2 mg/mL casein for 5 min (see Note 25); (4) wash with 100 μL of 2 mg/mL casein, and 50 μL of final solution containing ~5 nM 48% biotinylated microtubules, 2 mM ATP, 2 mM MgCl₂, 20 mM DTT, 50 μM Taxol (exclude when GMPCPP microtubules are used), 5 mg/mL glucose, 168.8 units/mL of glucose oxidase, and 1404 units/mL of catalase (see Note 26).

   The volumes of the solutions used correspond to a chamber volume of ~15 μL. For different chamber volumes scale accordingly.

2. Dilute streptavidin coated beads (0.82 mm diameter) at least 1: 80 in final solution without microtubules (~10 μL final volume) and bath sonicate for 1 min. Add 3–5 μL of the bead suspension from one side of the chamber and seal the chamber with vacuum grease from both sides (see Note 27).

3. Place the chamber on the microscope and find the region where beads were introduced. The presence of immobilized spherical pedestals helps to identify the bottom of the chamber. Trap two beads in the region where beads have started diffusing away from the bead injection point. Working in this area prevents trapping multiple beads (see Note 28).

4. Move away from the region of concentrated beads toward the other side of the flow chamber. Record the thermal motion of the two trapped beads while being at least 5 μm from the chamber surface and calculate the power spectrum as in Subheading 3.7.1 (steps 3 and 4).

5. Move the trapped beads to the z-midplane of the chamber, and with the cursor mark the position of the beads on the display of the camera. Dim the transmitted-light illumination and activate fluorescence illumination to visualize microtubules. Search for microtubules and move the stage to attach one of the two beads to the microtubule at a position of 1–3 bead diameters from a free end (see Note 29). Optimally the microtubule is attached to the bead that is toward the opposite side of the chamber from where the beads were introduced. If not, flip positions of the two beads.

6. Smoothly and gradually create a flow by moving the stage along the axis defined by the two trapped beads and in the direction away from the side where beads were introduced (see Note 30). The gradual alignment of the fluorescent microtubule along the direction of the two beads can serve as an indicator of sufficient flow that will eventually allow the microtubule to come in contact and attach on the other trapped bead (see Note 31).

7. Find the bottom of the chamber and check again the focus of the condenser. Center the QPDs such that the average signals in x and y are approximately zero.

8. Apply 5–6 pN tension on the dumbbell beads by moving one of the two traps further away from the other.

9. Center the dumbbell over the top of a spherical pedestal and approach the surface by moving downwards along the z-axis. When the microtubule is touching the pedestal, the x-signal will start shifting due to deformation of the microtubule. By moving in the y-direction, scan the pedestal for 30 s to 1 min to detect the presence of any kinesin molecule (see Note 32).

10. The above screening should be repeated for 10 spherical pedestals. Have 3 or fewer interacting pedestals per 10 tested pedestals to achieve single-molecule conditions.

11. If there are spherical pedestals on the top and bottom coverslips, the dumbbell can be scanned on both. This will require refocusing the condenser (see Note 33).

12. When using GMPCPP-stabilized microtubules, it is preferable to use buffers that do not contain sodium, since hydrolysis of GMPCPP is accelerated and the dumbbells may break under 5–6 pN tensile forces (see Note 34).

13. The total force produced by kinesin, as dictated by the force balance, is given by the sum of the traces from both laser-trapped beads.

4. Notes

1. We found Norland 65 to have lower levels of autofluorescence across the wavelengths used to excite green and red emitting fluorophores, compared to other Norland formulations.

2. Filtering with rapid-flow sterile bottle-top filters is essential. We found 0.2 μm syringe-filters to clog almost immediately.

3. Do not keep stocks of casein solution at 4 °C for extended periods of time due to microbial growth.

4. It is very important not to mix by pipetting up and down because mechanical stress at this stage can disassemble the microtubules. Let Taxol mix passively via diffusion.

5. Stock solutions of beads tend to aggregate over time. An efficient way to disaggregate is to dilute beads in 2 mg/mL casein BRB80 (pH 7.5) the day before and place them on a rotating plate overnight at 4 °C. Next day, bath sonicate for 1 min and proceed to the next step 2. For different size of beads use the manufacturer’s details to calculate the equivalent amount that would correspond to the same amount of biotin binding sites as the beads of diameter 0.82 μm (~3.3·10¹³).

6. This step is crucial when using biotinylated kinesin to ensure the primary attachment is via the antibody.

7. When assaying a new preparation of kinesin, it is recommended to initially make multiple samples of beads mixed with kinesin concentrations in the range of 0.1–100 nM at ~fivefold concentration steps. This will allow for rapid titration to determine conditions for single-molecule experiments. Since this will require serial dilutions of the original kinesin sample, it is important to make the dilutions in ~10 mg/mL casein to prevent kinesin from sticking to the walls of the polypropylene tubes. It is also important to use a new pipette tip (1–10 μL) for each dilution and to ensure the desired volume of solution has entered the tip. Small volumes (0.5–1 μL) can be missed due to defective tips. The above precautions also help for reproducibility after one finds the appropriate range of kinesin concentrations. Due to variabilities in this multi-step process, reproducibility in achieving the exact kinesin concentration needed is not easy to achieve. For this reason, once the appropriate kinesin concentration is found, always prepare three samples of beads on subsequent days: one with the target kinesin concentration, one with half this concentration and one with twice this concentration.

8. Incubation times in the literature vary substantially. In our hands, reproducibility is higher when using this long incubation time. However, one should be mindful of losses of activity for some motors during long incubations [18]. In our case, no change in the activity was observed, even for overnight incubation.

9. The activity units of glucose oxidase and catalase from the manufacturer are usually in U/mg and may therefore be different depending on the lot. The calculations should be modified accordingly such that the final concentrations of glucose oxidase and catalase are 168.8 and 1404 U/mL, respectively.

10. The extent of mixing can be assessed by holding up the sample to the light. Usually, 15–17 reversals are sufficient. Gentle instead of vigorous mixing is essential to avoid air bubble formation that would interfere with uniform mixing.

11. Caution must be taken during the transfer of dimethyldichlorosilane solution. This should take place only under the fume hood. When applying vacuum to the glass desiccator, the dimethyldichlorosilane solution will initially boil due to the abrupt drop in pressure. This is a good indication that hard vacuum is applied.

12. Initially, when applying the hard vacuum, the dissolved and trapped air in the elastomer solution will be released. To avoid overflow of the elastomer, the vacuum should be engaged slowly or intermittently. This process may require repeated rounds of disconnecting and reconnecting the vacuum for the first 15 min. Then the vacuum can remain steadily connected for the remainder of time (15 min). This may take some practice, so be prepared for overflow of the elastomer. Cover the bottom of the desiccator with aluminum foil so spilled elastomer can easily be removed afterwards without having to clean the desiccator.

13. Try to slowly remove the cured elastomer stamp by peeling it off from one of the sides of the square master mask. Failing to produce good separation can be due to poor passivation of the master mask (step 3) or to poor mixing of the silicone elastomer base with its curing agent (steps 2 and 4). Elastomer stamps can be used multiple times to form ridges and do not have to be made every time.

14. Use tweezers to handle the elastomer stamps and avoid long UV-plasma exposure as it may start etching and destroying the imprinted pattern of the stamp. The brief exposure to UV-plasma is sufficient to chemically activate the surface so that it can interact with dimethyldichlorosilane and be passivated. Efficient passivation is essential since the stamp will not be able to be separated by the photo cured Norland adhesive (steps 9 and 10) and successfully form ridges on a glass coverslip.

15. Excess optical glue around the stamp can be removed using a sharp, new razor blade. Cut very gently without applying strong mechanical forces that may either break the glass coverslip or cause the whole imprinted pattern to dissociate from the glass coverslip. Be aware that this step may take practice.

16. The amyl acetate mix, with or without silica beads, should not be stored in polypropylene tubes, as the polymer is not resistant to amyl acetate. Amyl acetate and nitrocellulose stock solutions are not used for more than a month after they are first opened. Beyond that time frame we have found that it is hard to passivate the experimental chamber with Casein protein resulting in non-specific interactions when doing experiments in the three-bead assay.

17. Always vortex the bead suspension before removing 4 μL to coat a glass coverslip. Silica beads settle quickly creating a vertical gradient of bead concentration.

18. The nitrocellulose-coated glass coverslips, with or without beads, can be used for 48 h. Beyond that, nonspecific interactions may occur during experiments similarly as mentioned in Note 16 above.

19. It is preferable the ridges be on the top coverslip to minimize interference with the trapping laser for an upright microscope or the other way around for an inverted microscope.

20. When a chamber with the pattern of rectangular ridges is used, it is better to flow 25 μL of microtubule solution, four times, waiting 1 min in between washes. This procedure increases the probability to get microtubules suspended between the ridges at a direction vertical to the long axis of the rectangular ridges.

21. Pipette solutions on the one side of the chamber and use absorbing filter paper on the other side of the chamber to facilitate the flow. Capillary action should facilitate the entry of the first solution. If the flow toward the other side of the chamber is hindered, gently tap the chamber with the bottom of a 0.5 mL polycarbonate tube. Before sealing the chamber with vacuum grease, check under the microscope to be sure that the beads are not too concentrated. The chamber in this case can be rewashed with 2 mg/mL Casein and 20 μM Taxol, followed by a wash with a decreased density of beads. Beads should be sufficiently dilute such that nearby diffusing beads will not be close enough to be trapped along with the already trapped single bead. Some experimentalists prefer to inject 3–5 μL of dense bead suspension on one side of the chamber. In this case the chamber needs to be prewashed with 50–100 μL of the final solution without the beads before the beads are introduced. A single bead can be trapped and moved away from concentrated collection of beads. If using the eyepiece and not the camera to check for the beads, always wear protective goggles for the IR laser even if the laser source is off or blocked. It is important to develop the habit of never looking through the eyepiece without protective goggles. As an additional protective measure, IR filters can be purchased separately (e.g., by Edmond Optics) and placed in both eyepieces of the microscope. However, goggles must still be worn.

22. Make sure that your signal is filtered at the Nyquist frequency before recording to avoid aliasing. When the distance of the bead from the surface is five times the radius of the bead or more, the effect of the bottom coverslip to the hydrodynamic drag is less than 10% based on Faxen’s law for motion laterally to the surface [41, 42].

23. It is good practice to calculate the power spectrum of the thermal motion for every bead used. The power spectrum should look “flat” in a logarithmic scale for frequency values f lower than the corner frequency f_c (10 Hz < f < f_c). Anomalies can be due to “dirt” in the chamber trapped along with the bead or aberrations caused by variability in the thickness of the glass coverslips. To avoid such issues, all stock solutions should be filtered (pore size 0.2 μm) and the correction collar of the objective should be adjusted to correspond to the right thickness of the coverslip.

24. When the bead is in contact with the surface of the bottom coverslip, the reduction of thermal noise in the force trace of the bead is visible. When interactions are detected between the bead and the microtubule, move along the z-axis away from the microtubule. Find the minimal possible contact distance between the bead and the underlying surface at which interactions are occurring between kinesin and microtubule and start recording the signal.

25. For reproducibility reasons as mentioned in Note 7, serial dilutions from stock solution of kinesin motor should include ~10 mg/mL casein and changing of pipette tips at every dilution step.

26. Before injecting the final solution with the beads, place the chamber on the microscope and examine the density of the fluorescently labeled microtubules. They should be very sparse so that no more than one microtubule will accidentally get bound to the two laser-trapped beads. A low density also ensures minimization of small biotinylated microtubule fragments that may block streptavidin or neutravidin-binding sites on the beads. These fragments result in unstable dumbbell attachments that do not sustain high tension forces required for steps 9 and 10.

27. Hold the chamber against the light while injecting the bead suspension using a small pipette tip (1–10 μL). The scattering of light due to the high density of beads will be visible to the naked eye. Inject õne-third of your chamber volume. More than that amount may overfill the chamber and complicated the ability to isolate the trapped beads. Less than that may result in the inaccessibility of the diffusing front of the beads because the condenser will run into the vacuum grease at the sides of the chamber before the part of the chamber with the beads can come into the field of view.

28. This step takes practice, and one may want to dilute beads even more to prevent trapping multiple beads in the same laser beam. If by accident the bead suspension is too dense, instead of discarding the chamber and creating a new one, aggregates of beads (which are almost always present at high density of beads) can be swept away using the two traps, thus clearing the vicinity of single spotted beads that can be subsequently trapped.

29. If a bead is attached far from the free end of the microtubule, it can be very difficult to form a dumbbell. The reason is that the two segments will develop opposing torques on the attached bead when creating flow in the next step 6, preventing the microtubule to align with the flow sufficiently enough to attach on the other laser-trapped bead and form a dumbbell.

30. The power of the trapping laser can be increased so that the beads will not escape the traps. It should be kept in mind that the presence of the microtubule develops additional high frictional forces. After forming the dumbbell, the power of the laser can be decreased back to its desired level.

31. After the dumbbell is successfully formed, block fluorescence light and use it only when necessary to prevent photodamage.

32. An additional method to check if the microtubule is touching the surface of the spherical pedestal is to monitor the microtubule fluorescence. The microtubule segment that comes in contact with the sphere appears brighter due to dampening of the thermal fluctuations.

33. This requires having an objective with a long working distance (e.g., using a water versus oil objective).

34. When working with GMPCPP-stabilized microtubules, even slight traces of sodium may accelerate the hydrolysis rate of the nucleotide analogue [43] compromising the mechanical stability of dumbbells. In this case, it is better to prepare the casein stock solution in KCl instead of NaCl. However, we have observed that when casein in KCl is used, refractive particles of unknown origin can occasionally be seen in the chambers.