Use of Dual Optical Tweezers and Microfluidics for Single-Molecule Studies

Summary

Visual, single-molecule biochemistry studied through microfluidic chambers is greatly facilitated using glass barrel, gas-tight syringes, stable connections of tubing to flow cells, and elimination of bubbles by placing switching valves between the syringes and tubing. The protocol describes dual optical traps that enable visualization of DNA transactions and intermolecular interactions.

Abstract

Visual biochemistry is a powerful technique for observing the stochastic properties of single enzymes or enzyme complexes that are obscured in the averaging that takes place in bulk-phase studies. To achieve visualization, dual optical tweezers, where one trap is fixed and the other is mobile, are focused into one channel of a multi-stream microfluidic chamber positioned on the stage of an inverted fluorescence microscope. The optical tweezers trap single molecules of fluorescently labeled DNA and fluid flow through the chamber and past the trapped beads, stretches the DNA to B-form (under minimal force, i.e., 0 pN) with the nucleic acid being observed as a white string against a black background. DNA molecules are moved from one stream to the next, by translating the stage perpendicular to the flow to enable the initiation of reactions in a controlled manner. To achieve success, microfluidic devices with optically clear channels are mated to glass syringes held in place in a syringe pump. Optimal results use connectors permanently bonded to the flow cell, tubing that is mechanically rigid and chemically resistant, and which is connected to switching valves that eliminate bubbles that prohibit laminar flow.

Introduction

The ability to visualize protein-DNA interactions at the single-molecule level and in real-time has provided significant insight into genome stability^1,2. In addition to working with single molecules of DNA one at a time, the ability to view transactions between individual molecules nearby provides additional insight^3,4,5. The manipulation of additional DNA molecules requires both additional optical traps as well as high-quality, multi-channel, microfluidic flow cells⁶.

There are several methods available to generate more than one optical trap. These include galvanometer scanning mirrors, acoustic optic modulators, and diffractive optics, which generate holographic optical tweezers^4,7,8,9. Often, scanning mirrors and acoustic optic modulators produce traps that timeshare. In the setup described here, the beam of a single Nd:YAG laser is split on polarization, and then galvanometer laser scanning mirrors control the position of what is termed the mobile trap (Figure 1)⁴. To facilitate the positioning of mirrors and filters to direct the trapping beam to the back aperture of the microscope objective, a HeNe laser is used. This makes overall alignment easier as the HeNe beam is visible to the naked eye, whereas infrared beams are not. The HeNe beam is also safer to work with making the positioning of mirrors and other components less stressful. Initially, the beam path for this laser is separate from the 1064 nm beam, but is introduced into the same beam path, and then into the microscope objective. Once physical alignment is achieved, positioning the 1064 nm beam on top of the HeNe beam is done and this is facilitated by the use of an infrared viewer and various beam imaging tools to visualize beam position and quality. Then, the beam expander is introduced, and the resulting expanded infrared beam is aligned onto the back aperture of the objective. Finally, the objective is removed and the power in each polarized beam is measured and adjusted using λ/2 waveplates to be equal (Figure 1C). Power measurements are also done once the objective is returned and typically there is a 53% power loss. There is, however, sufficient power to form stable fixed and moving optical traps in the focal plane (Figure 1D).

To image DNA transactions, microfluidic flow cells play a key role as they permit controlled measurements at the single-molecule level with high spatial and temporal resolution (Figure 2). The term microfluidic refers to the ability to manipulate fluids in one or more channels with dimensions ranging from 5-500 μm^10,11. The term stream refers to the actual fluid within a channel and the channel refers to the physical channel in which a fluid stream or streams move. Single-channel flow cell design has a common, physical channel where reactions are observed and there is typically only one fluid stream present. Thus, these designs are known as single-stream flow cells. In contrast, multi-stream flow cells are defined as a microfluidic device in which two or more entry channels converge into a single, common, physical channel (Figure 2A). Within the common channel, the fluid streams that originate from the individual channels flow parallel to one another, and remain separated with only minimal mixing between them occurring due to diffusion (Figure 2B). In the majority of experimental setups, a single pump pushes fluids into each channel at the same speed. In contrast, when boundary steering is used, three or more, independently controlled pumps push fluids through the channels. However, each pump operates at a different speed but the net flow rate in the common channel is constant¹². This permits the rapid exchange of main channel components simply by altering pump speed.

In addition to laminar flow, another critical factor is the parabolic velocity profile within the laminar fluid stream. The highest flow velocity occurs in the middle of the stream and the slowest occurs next to the surfaces (Figure 2C)¹³. This profile must be considered to fully stretch a DNA molecule that is attached to a bead held in the stream for precise visualization of fluorescent DNA and accurate single-molecule analyses. Here, the DNA is stretched to B-form and is held in place under 0 pN of force. To achieve this, focusing of the optical trap position should be to a position 10-20 μm from the bottom coverslip surface (Figure 2D). Care must be taken so that the DNA molecule is not stretched beyond B-form as this can inhibit enzyme reactions. Under typical buffer conditions, 1 μm = 3,000 bp of DNA¹⁴. Furthermore, by trapping 10-20 μm from the coverslip, the DNA complex is positioned far from the surface thereby minimizing surface interactions.

Many methods have been used to create microfluidic device channels and these can be done in the laboratory or flow cells can be purchased from commercial sources^6,15,16,17. The optimal materials used to construct the flow cells must be mechanically rigid, optically transparent with low fluorescence, and impervious to organic solvents⁶. Frequently, borosilicate float glass, or fused silica are used to provide a stable flow environment for an extended time that is suitable for optical trapping, visualization, and force detection. These materials also permit the use of non-aqueous solvents (e.g., spectrophotometric grade methanol) to simplify surface wetting and removal of air bubbles, and denaturants (e.g., 6M guanidinium hydrochloride) or detergents to clean the flow cell. Finally, the methods used to introduce fluids into flow cells vary from complex vacuum pump systems to single syringe pumps^{14,18,19,20,21,22,23,24,25,26,27}. In the approach described here, a syringe pump that can accommodate up to 10 syringes is used (Figure 3A). This provides flexibility to use single-channel flow cells or flow cells with multiple inlet channels. Here, a three-channel flow cell is used and is mated to syringes held in place on the syringe pump using poly-ether-ether-ketone (PEEK) tubing (Figure 3A-C). The flow of fluids is controlled by four-way switching valves and thus serve to minimize the introduction of bubbles into the flow cell (Figure 3A,D). In addition, Hamilton gastight syringes which have stiff glass walls and polytetrafluoroethylene (PTFE)-coated plungers, are recommended as they provide exceptionally smooth plunger motion that is essential to obtaining a smooth flow^14,27.

In the experimental system described, flow cells with two to five inlet channels have been used. The number of inlet channels is dictated by the experiment being done. For the study of RecBCD and Hop2-Mnd1, two stream channels were sufficient^14,28. For the helicase, the enzyme was bound to the free end of the DNA and translated into a stream containing magnesium and ATP to initiate translocation and unwinding. For Hop2-Mnd1, optically trapped DNA was translated into the adjacent fluid stream containing proteins and buffer ± divalent metal ions. The use of three-channel flow cells enables one to trap DNA in stream 1, translate the DNA into stream 2 to allow protein binding to occur, and then to stream 3 where ATP is present, for example, to initiate reactions. A variation on the above is to use fluorescent-tagged protein in channel 2, which results in the fluid stream being completely white and precludes visualization of the DNA. When this molecule is translated into stream 3, both proteins and DNA are now visible when reactions are initiated.

The use of four-way switching valves to control fluid flow is a critical component of the system to eliminate bubbles in the flow cells. Bubbles are detrimental to stable fluid flow as they contract and expand in unpredictable ways resulting in rapid changes in flow velocity and the introduction of turbulence. When the valves are positioned between syringes and inlet tubing, the flow path is disconnected by switching the valve position when syringes are changed out. When the new syringe is put in place, the plunger can be manually depressed so that >6 μL is ejected (the dead volume of the valve) and this eliminates bubbles almost entirely.

The attachment of connectors to flow cells is frequently the rate-limiting step in flow cell use. We describe the use of two types of connectors: removable known as press-fit and permanent ones (nanoport assemblies). The removable connectors are simple to adhere to a flow cell and different types of flexible tubing in addition to the recommended PTFE can be tested with these connectors. This is a rapid and cost-effective way to test tubing and connectors without sacrificing more expensive glass flow cells. In contrast, nanoport assemblies are attached permanently, withstand pressures up to 1,000 psi, and, in our hands, their use is restricted to PEEK tubing of different diameters. This is not a disadvantage as PEEK tubing is preferably used. A single glass flow cell with permanent assemblies attached can be reused for more than 1 year with careful use.

Protocol

1. Laser trap alignment and testing with polystyrene beads

NOTE: For the setup, refer to Figure 1A,B.

CAUTION: The experimentalist should wear appropriate protective eyewear or laser safety glasses during laser beam alignment. As the optical tweezer system described herein uses both HeNe and IR beams, two separate sets of laser safety glassware are required.

1. HeNe Beam alignment
   1. Position all optical components on the breadboard. Align the breadboard so it is as close to a 90° angle as possible relative to the microscope.
   2. Attach the mating tube containing the objective and telan lens to the microscope (Figure 1A-11,12).
      NOTE: The telan lens is a lens system that is used to focus the image at the intermediate image plane in the eyepieces.
   3. Turn on the HeNe laser and open the shutter (Figure 1A-18). The beam should strike filter (Figure 1A-3) at a height of 54 mm reflecting 50% of the light onto mirror 5 (Figure 1A-5). The remaining 50% of the beam strikes mirror 4 (Figure 1A-4) at the same height as mirror 3.
   4. Adjust mirror 4 (Figure 1A-4) so that the beam strikes mirror 6 (Figure 1A-6) at a height of 51 mm.
   5. Adjust mirror 6 (Figure 1A-6) so that the beam is directed to the lower galvanic mirror at a height of 51 mm. The beam that exits from the upper Galvanic mirror will be at 57.5 mm.
   6. Adjust mirror 3 (Figure 1A-3) so that this beam strikes mirror 5 (Figure 1A-5) at a height of 57.5 mm. Once reflected off mirror 5, the beam should strike mirror 9 (Figure 1A-9) at the same height. The beam is effectively recombined at mirror 9, passes through the objective, and telan lens system onto mirror 21 (Figure 1A-21), which reflects it into the objective.
   7. Place a clean microscope slide onto the stage of the microscope. Turn on the camera and image the beams of the fixed and scanning beams individually on the screen. Starting with mirror 3, then 5 and 9, finally 21, perform small adjustments to position the fixed beam in the center of the screen.
   8. To ensure that the angle of the beam exiting mirror 21 is 90°, alter the Z-height of the microscope and position the beam imaged on the upper and lower surfaces of the slide. When they are identical, the beam is in the correct position, which is perpendicular to the stage. Shutter this beam using shutter 19 (Figure 1A-19).
   9. Starting with mirrors 4 and 6, perform small adjustments to position the fixed beam in the center of the screen. Shutter this beam.
2. Infrared (IR) Beam alignment
   CAUTION: Perform all steps at low laser power for safety reasons.
   1. Adjust wave plate 14 (Figure 1A-14) to adjust the ratio of transmitted and reflected beams being split at the beam splitter 2 (Figure 1A-2). Beam splitter 2 reflects the s-component of the laser beam and transmits the p-component.
   2. The beam that passes through 2, strikes mirror 7 (Figure 1A-7), which deflects the beam onto the lower galvanic mirror. It should follow the path of the HeNe beam onto mirror 21. Perform small movements of mirror 7 to achieve this positioning.
   3. Adjust mirror 2 so that the beam that is reflected by mirror 2, strikes mirror 10 (Figure 1A-10) at a height of 57.5 mm.
   4. The beam is reflected from mirror 10 onto mirror 8 (Figure 1A-8) at a height of 57.5 mm. The subsequent beam should strike mirror 9 at the same height. If not, make small adjustments to mirrors 10 and 8 as required.
   5. Remove the objective and place the head of the power meter over the port in the objective turret.
   6. Use plates 14, 15, and 16 (Figure 1A-14,15,16) to adjust the power of each laser beam so that they are equal as shown in Figure 1C.
   7. Put beam expander 1 (Figure 1A-1) into place. Adjust the expanded laser beam to be circular without any coma or astigmatism.
   8. To test the optical traps, make a solution of 1 μm polystyrene beads suspended in water. Then, place 10 μL of this solution on a microscope slide, place a coverslip on top and seal with nail polish. Place the slide on the microscope stage and test the optical traps by trapping these 1 μm beads. Ensure that the beads are sucked into the traps and held stably when the stage is moved. Trapping should occur equally efficiently from all sides of the trap.

2. Microfluidic chamber preparation

1. If the flow cell is a commercially constructed device made from polydimethylsiloxane (PDMS) on a coverslip, proceed to step 3. If it is a glass device with connectors attached, proceed to step 4.
2. If the flow cell does not have connectors attached, bond them to the entry holes on the microscope slide first. See step 3 for the steps to attach connectors.

3. Flow cell connections using a PDMS flow cell

NOTE: For the setup, refer to Figure 2D.

1. Place the flow cell on a clean flat surface. Hold the PTFE tubing 3 mm from the free end with forceps. Push the tubing into the preformed port (the port can support 2 bar line pressure).
2. Repeat this procedure for each of the remaining ports. Connect the inlet ports to the syringe pump and the outlet to a waste bottle (Figure 4A,B).
3. Fill each glass syringe with 1 mL of spectrophotometric grade methanol. Attach each syringe to a switching valve. Ensure the valve has the outlet directed to waste and purge each line with 50 μL of methanol (Figure 4C, closed position).
4. Switch the outlet position to the flow cell (Figure 4C, open position). Pump 800 μL of methanol through the flow cell at a flow rate of 100 μL/h to wet the surfaces and eliminate bubbles.
5. The next day, repeat this process using 800 μL of ultrapure water. The flow cell is now ready for use.

4. Attachment of press fit tubing connectors

NOTE: For the setup, refer to Figure 2F.

1. If press-fit tubing connectors are to be used, carefully remove the adhesive tape from one side of the connector and place it over the hole in the microscope slide. Press down for a few seconds. Repeat the process for the remaining connectors.
2. Place the flow cell on a clean flat surface. Hold the PTFE tubing 3 mm from the free end with forceps. Push the tubing into the preformed hole in the port and repeat this procedure for each of the remaining ports.
3. Attach tubing from the inlets to the switching valves. Proceed to step 7.

5. Attachment of permanent assemblies

NOTE: For the setup, refer to Figure 2A-C.

1. Perform the attachment either on a clean, flat surface or on a custom-built manifold to hold the flow cell and connectors in place while bonding occurs.
2. Place the flow cell on a clean flat surface or in the recessed section of the manifold (Figure 2B). Place a small amount of glass glue on the bottom of the assembly and insert the seal.
3. Position the nanoport over one of the entry holes on the microscope slide. Gently push down and hold in place with no lateral movement. Repeat the process for the remaining ports.
4. Allow to dry or clamp in place in the manifold. Gently remove the flow cell from the manifold and ensure that the assemblies align well with the entry ports in the flow cell. Proceed to step 7 when ready.

6. Flow cell connections and preparation

NOTE: For the setup, refer to Figure 4A,B.

1. Place a flow cell on the microscope stage. Attach the tubing using the finger-tight connectors.
2. Fill each glass syringe with 1 mL of spectrophotometric grade methanol. Attach each syringe to a switching valve. Ensure the valve has the outlet directed to waste and purge each line with 50 μL of methanol (Figure 4C, closed position).
3. Switch the outlet position to the flow cell (Figure 4C, open position). Pump 800 μL of methanol through the flow cell at a flow rate of 100 μL/h to wet the surfaces and eliminate bubbles.
4. The next day, repeat this process using 800 μL of ultrapure water. The flow cell is now ready for use.

7. Control of fluid flow

1. Turn the switching valves to the closed position (Figure 4C). Remove the syringes, discard the residual water, and fill the syringes with experimental solutions, for example, DNA-beads; enzyme; and ATP. Return syringes to the pump and connect them to the switching valves.
2. In this position, manually force the plungers down 5-10 μL to remove any bubbles. Change the switching valve position to Open.
3. Use the flow rate scheme to achieve a fluid flow speed optimal for trapping and visualization as outlined in Table 1. The optimal fluid flow for optical trapping should enable stable trapping of beads and once the DNA is stretched, it should be B-form (~3,000 bp/μm).

8. Trapping of polystyrene beads with single DNA molecules attached

1. At 10x magnification, locate the border between fluid streams close to the channel entry points. Add oil to the 100x objective and switch to the higher magnification.
2. Open the shutters for the optical traps (either both or one at a time). The focused laser beams should trap beads and DNA should stretch out immediately.
3. Once a complex has been trapped, translate the stage perpendicular to the flow (Figure 3A,C, inset). This will move the trapped complex from the DNA bead solution, into stream 2. Further translation will move the complex to stream 3.

9. Flow cell cleaning

NOTE: Perform this daily.

1. Turn off the pump when the experiment is complete. Turn the switching valves to the closed position (Figure 4C).
2. Remove syringes and empty them. Rinse three times with filtered distilled water.
3. Fill the syringes with cleaning solution. Place syringes in the pump and connect to switching valves.
4. Purge switching valves and change the position to open. Pump 800 μL of cleaning solution at a flow rate of 100 μL/h typically overnight.
5. The next day, close the switching valves, and then remove and rinse syringes with water. Rinse the flow cell and lines with 4-800 μL of water at a flow rate of 400-800 μL/h. If more strenuous cleaning of flow cells is required, replace the cleaning solution with 6 M guanidium hydrochloride.

Representative Results

The initial testing of the trap alignment and strength is done with 1 μm, non-fluorescent polystyrene beads. Since most of the research done in the laboratory uses fluorescence, we further test trap strength using 1 μm, Dragon green polystyrene beads (Figure 1D,E). Thereafter work changes to optical trapping of DNA-bead complexes where the DNA is stained with the bis-intercalating dye YOYO-1^14,29. When these complexes are trapped 10-15 μm above the coverslip surface in a microfluidic flow cell, fluid flow past the bead stretches out the DNA (Figures 3B,C). It is important to measure the length of the fluid-stretched DNA to ensure that it is stretched to its B-form length, which is the expected length. For example, for bacteriophage lambda DNA, the full length is slightly longer than 16 μm, which corresponds to 48,504 bp¹⁴.

To characterize fluid flow, fluorescent beads were introduced into flow cells with a focus height of 15 μm above the coverslip surface using different pump speeds and in solutions of different viscosity. Movies of bead position were captured, and bead position was tracked using image analysis software. The results show that for 1 μm diameter beads, linear velocity can be accurately tracked at a pump flow rate of 100 μL/h in sucrose solutions ranging from 10%-50% (Figure 5A). Under similar conditions, the force on the bead can be calculated and found to range from 10 to 400 pN, depending on the pump flow rate and solution viscosity (Figure 5B). Finally, the effect of bead diameter ranging from 120 to 970 nm on force were evaluated in different concentrations of sucrose. The data reveal that forces ranging from 1-27 pN could be measured by attaching beads of different diameter to a motor protein and asking the motor to pull that bead against the fluid flow (Figure 5C).

Figure 1: Optical layout and demonstration of trapping activity. (A) Schematic of the optical layout with the IR beam in red and HeNe beam in black. (B) Photograph of optical components mounted on a breadboard. The component numbers in panels A and B are as follows: 1), Beam expander; 2), Interference mirror (Rs_max = 1064 nm; Tp_max = 1064 nm); 3), Interference mirrors (633 nm; R = 50%; T = 50%); 4-6), Interference mirrors (R_max = 633 nm); 7), Interference mirror (R_max = 1064 nm; T_max= 633 nm); 8), Interference mirror (Rs_max = 1064 nm; Tp_max = 1064 nm, T_max = 633 nm); 9), Interference mirror (R_max = 1064 nm, 633 nm; T_max = 1064 nm, 633 nm); 10), Interference mirror (R_max = 1064 nm); 11), Objective lens, f = 110 nm; 12), Telan lens; 13), Galvanic mirrors; 14-16), Rotators (/2); 17), IR scanning channel shutter; 18), Visible channel shutter; 19), IR Fixed channel shutter; 20), Beam stops and 21), Interference mirror (R_max = 1064 nm; T_max = 633 nm). (C) Measurement of power output. The power output from the laser is measured before installing the laser head into the optical layout. Once trap alignment is done, the beam power is measured after the 100x objective for each trap. (D) and (E) Images demonstrating stable, optical trapping of 1 μm fluorescent beads. F, fixed trap and M, mobile trap with the mobile trap in scanning mode moving through a small (D) or large (E) circle at 30 mHz. Please click here to view a larger version of this figure.

Figure 2: Microfluidic flow cells. (A) Assembly of a three-channel flow cell made from borosilicate glass. (B) A manifold to stably position flow cells during connector bonding. Top, the base of the aluminum manifold with a 1 mm recessed section to laterally position the flow cell. The four posts (one in each corner) guide the lid of the manifold into position, while the more centrally placed threaded bolts are used to clamp the lid in place. Bottom, manifold lid with drilled holes to match the posts and bolts on the manifold base. (1), nuts that attach to the manifold base bolts; (2) springs that are placed between the nuts and the manifold lid to provide light tension during the bonding process. (C) A close-up view of the convergence of the inlet channels into a common single fluid stream in a glass flow cell with connectors already attached. A quarter is placed next to the flow cell as a reference. (D) An image of the PDMS flow cell. The PDMS layers is 3 mm thick. (E) A fused silica flow cell with luer connectors. (F) A borosilicate glass flow cell with press-fit connectors in place. Please click here to view a larger version of this figure.

Figure 3: Fluid flow characteristics through a glass flow cell. (A) Fluid flow within the flow cell is laminar. The schematic shows the flow cell from the top to demonstrate that inter-channel diffusion is the main source of mixing between adjacent fluid streams. The flow direction is indicated by blue arrows and the individual streams are colored white, light gray, and white. The widening regions of transverse diffusion between channels are indicated in red. The inset shows a fluorescence image of adjacent fluid streams at 10x magnification with fluorescent-labeled DNA-bead complexes in the lower stream and buffer only in the upper stream. The white spots in the upper stream are shot noise from the CCD camera. (B) The flow profile (light purple) in laminar flow cells is parabolic. The flow cell is viewed from the side and the direction of the flow is indicated above each cell. The fastest fluid velocities occur in the flow cell located in the center while the slowest fluid velocities occur next to the flow cell surfaces. Consequently, for an optically trapped DNA-bead complex positioned at the center of the flow cell, the fluid flow stretches the bacteriophage λ DNA molecule (48,504 bp) to 16 μm (B-form) so that clear visualization can be made (left image). In contrast, the same molecule positioned near the surface of the flow cell experience low flow rates and cannot be stretched out. (C) A single DNA bead complex is positioned within a flow cell. Different depth flow cells can be used but, in each case, the beam alignments can be set so that optimal trapping occurs 10-20 μm above the coverslip surface, which eliminates surface effects and has stable flow so that DNA molecules can be stretched to B-form. As indicated in the left image in (C), the DNA is stretched to enable clear visualization. Please click here to view a larger version of this figure.

Figure 4: Flow and reaction control of optically trapped DNA within a flow cell in the microscope stage. (A) Mating of the flow cell to a syringe pump. A photograph of a three-channel flow cell connected to a syringe pump via three-way switching valves is presented. The flow cell is held in place on a motorized stage in an inverted microscope. The flow cell is connected to a waste line (left) and switching valves (dashed cyan circle) via orange, 500 µm tubing (indicated with green tape). The valves are stably mounted onto a custom-built manifold adapted to the syringe pump apparatus, which houses 1 mL gastight syringes (yellow arrows demarcate plungers). The waste lines are used to remove air bubbles that can be trapped in the switching valve during syringe changes. (B) Close-up view of the three-channel flow cell. To facilitate visualization of fluid streams, the outer ones have been colored with food coloring with the central stream containing water only. The zoomed region indicates how an optically trapped bead-DNA molecule complex is translated from one fluid stream to the next (blue arrow) by translating the stage perpendicular to the flow (black arrow). (C) Schematics of the three-way switching valves indicating their operational positions. The letter designations F, S, and W correspond to the direction fluid and can flow to the flow cell, syringe, and waste, respectively. One port is permanently closed as indicated by the black plug on the top of each schematic. In the Open or operational position (left panel), fluid flow passes directly through the switching valve from the syringe (S) to the flow cell (F). In the Closed position (right panel), the dial is turned so that the flow cell is sealed from the environment. In this position, the flow is directed to waste and syringes can be switched out without introducing air bubbles into the flow cell. As the valves have a 6 μL dead volume, very little solution is wasted when purging lines before reverting to the Open configuration. Please click here to view a larger version of this figure.

Figure 5: Bead motion and opposing force are affected by fluid flow. (A) Linear bead velocity is affected by pump speed and sucrose concentration. (B) The opposing force on a bead is affected by solution viscosity. (C) Bead diameter influences the force on beads under flow. In all experiments, fluorescent beads were used (Dragon green). In (A) and (B), 0.97 µm diameter beads were used and in (C), beads of different diameter were used. Sucrose solutions were made in ultrapure water, filtered through 0.2 μm filters and refractive index measured using a handheld refractometer, with viscosity determined in centipoise using ISCO Tables and finally converted to SI units in mPa.s. Please click here to view a larger version of this figure.

Flow rate (μL/h)b	Time per step (min)	Volume dispensed (μL)
800	5	67
400	5	33
200	5	17
100	5	8
50	5	4
25	5	2
15	5 or more	1 or more
a. These are fluid flow rates calculated by the syringe pump software and are unique for each syringe diameter used. The flow rates enable rapid filling of the lines and flow chamber (800 μl/hr), followed by a gradual decrease to the ideal flow rate for trapping and visualization
b. The syringe pump can be controlled from the keypad or using software provided by the vendor.

Table 1: Fluid flow rates to minimize bubbles and achieve stable trapping.

Discussion

The careful assembly of the flow system is critical to the successful outcome of experiments^4,6. One of the most challenging aspects of the protocol is the attachment of connectors to the glass surface. For this, we use the following two approaches: press-fit fit tubing connectors and nanoport assemblies. Press-fit connectors adhere easily to glass followed by pushing of PTFE tubing into the preformed holes using forceps. When a more stable attachment is required, the bonding of nanoport assemblies is preferred. With careful use and cleaning, a single glass flow cell with permanent assemblies can be used for more than 1 year. If flow cells need to be switched out more frequently, PDMS flow cells can be used as they are an excellent and cheaper alternative. They can be reused but PDMS is susceptible to swelling, which can alter the fluid flow behavior over time.

Both glass and PDMS flow cells are not indestructible. In the laboratory, the maximum flow rate used is 800 μL/h. If this is exceeded daily and/or for extended periods of time, flow cell warping (glass flow cells) and distortion of PDMS can occur. While the initial wetting procedure with methanol and water takes time, it is well worth it as all bubbles are eliminated, and also leaks at connectors can be detected. The flow cells described here provide flexibility to the experiments being done. Using these approaches of different flow cells, tubing and attachment of different connectors, experimental conditions can be readily adapted and quickly optimized to ensure optimal data are obtained. Y-shaped channel flow cells are presented but a large range of microfluidic device designs are available from multiple companies both without and with connectors attached. This provides the investigator even greater flexibility in experimental design.

The use of four-way switching valves is a simple and effective method that ensures that once the flow system is assembled and tested, it effectively becomes a closed environment, where contamination is minimal, and bubbles are eliminated. The latter is critical as bubbles tend to compress and decompress randomly during an experiment and this perturbs fluid flow causing trapped DNA molecules to move in unexpected ways and even break. Great care should be taken to ensure they are prevented from forming in the system.

Finally, the system described here is optimized for visual biochemistry where careful analysis of DNA, beads attached to proteins or fluorescent-tagged proteins can be performed. Although the system is not designed to do force measurements with the dual traps and addition of quadrant photodetectors, forces can be measured by attaching beads of different diameter to DNA motor proteins, for example, and asking the motor to pull this bead against fluid flow. By varying solution viscosity, fluid flow, and bead diameter, a large range of opposing forces can be applied to provide detailed insight into motor protein function (Figure 5). This is a simple and inexpensive way to measure forces associated with biological transactions, but it lacks the precision associated with more sensitive approaches. When the visual biochemical approach is combined with sensitive force detection methods, more complete pictures of biochemical reactions can be provided.

Materials

Name	Company	Catalog Number	Comments
100x objective	Leica	506318 or 506038	Oil immersion lenses; Imaging and optical trapping only; Plan APO objectives optimized for fluorescence imaging
10X Objective	Leica	506263	Used to locate laser beams spots during alignment; to find focus and X-Y position in flow cell
1 mm fluorescent beads	Bangs Labs	FSDG004	Used for tap performance, focal position determination
1 mm polystyrene beads	Bangs Labs	CPO1004	Used for trap performance evaluation and binding to biotinylated molecules
63x objective	Leica	506081	Used to locate laser beams spots during alignment and to find focus and X-y position in flow cell; can be used for optical trapping as it has an identical back aperture diameter to the 100X; oil immersion lens
Alignment laser	Lumentum	1100 series	10mW HeNe laser that is visible to the naked eye that is used to position optics
Beam alignment camera	Amscope	MU303	A simple, inexpensive and software controlled camera for imaging of the beam position
Camera control and Image capture software	Hamamatsu	HCImage	Coordinates activities of the Lambda DG4 with the camera to facilitate rapid wavelength switching
Camera; Orca flash 4	Hamamatsu	c13440-20cu	CCD camera for imaging of single-molecule experiments
C-mount for the beam alignment camera	Spot imaging solutions	DE50CMT	Provides optimal positioning of the camera for imaging of laser beams during alignment
C-mount for the Orca Flash 4 camera			Has a retainer ring to hold an IR blocking filter in place. This eliminates reflected IR beam from the optical traps and facilitates clearer imaging of trapped objects.
Cy5  fluorescence filter cube	Semrock	cy5-404a-lsc-zero	Used in conjunction with Lambda DG4 to image Cy5 only
Fitc-Txred  fluorescence filter cube	Semrock	fitc/txred-2x-b-000	Used in conjunction with Lambda DG4 to image FITC and TXRed
Fluidics tubing	Grace Bio	46004	PTFE tubing as an alternate to PEEK; works well on some flow cells. Can be used with PDMS flow cells or glass flow cells when Grace Bio fit tubing connectors are used
GFP fluorescence filter cube	Semrock	gfp-3035b-lsc-zero	Used in conjunction with Lambda DG4 to image GFP only
Glass flow cells	Translume	Custom	Clear flow channels for imaging (Fig. 2E)
Glass glue	Loctite	233841	Securely and easily bonds Nanoport assemblies to glass flow cells
Glass/PDMS sandwich flow cells	CIDRA Precision services	Custom design	Flow cells built according to your specifications; imaging channels are clear (Fig. 2C)
Hamilton Cleaning solution	Hamilton	18311	Gentle but efficient cleaning solution for glass flow cells; does not bubble when used carefully
Illumination system	Sutter Instrument	Lamda DG4	Discontinued so recommend Lambda 721
Illumination system	Sutter Instrument	Lamda DG4	Discontinued so recommend Lambda 721
Image analysis software	Media cybernetics	Image Pro Premiere	Analysis of images and single molecule tracking
Image analysis software	Fiji/NIH Image/Image J	Shareware	Analysis of images and single molecule tracking
Image display card	Melles Griot	06 DLA 001	Alternate product from Thorlabs: VRC5
Immersion oil	Zeiss	444960	Immersol 518 F fluorescence free
Laser beam alignment tools	Thor labs	FMP05/M; dgo5-1500-h1; BHM1	Used to ensure beams are horizontal and at the correct height
Laser beam viewer	Canadian Photonics labs	IR 3150	Used to image IR beam spots on mirrors and  targets
Laser power meter	Thor labs		Measurement of laser output as well as trap strength
Laser safety glasses (HeNe)	Thor labs	LG7 or 8	Blocks >3 OD units of light of wavelengths >600 nm
Laser safety glasses (IR)	Thor labs	LG11	Blocks >7 OD units of light of wavelengths ³1000 nm
Mcherry  fluorescence filter cube	Semrock	mcherry-a-lsc-zero	Used in conjunction with Lambda DG4 to image mcherry only
Microscope	Leica	DMIRE2	DIC port removed to accommodate Dichroic trapping/alignment mirror
Microscope control software	UCSF/shareware	uManager	Controls the microscope, permits focal alignment of objectives as well as stage control
Nanoport assembly	IDEX	N333	Connectors that are bonded to flow cells
Optical table support	Thor Labs	PA52502	Active isolation table support
Optics and lenses	Solar TII	Various	Interference mirrors, telescopes and lenses custom designed for the system
PDMS flow cells	ufluidix	Custom	Flow cells built according to your specifications; imaging channels are clear (Figs. 2B and D)
PEEK tubing	IDEX	1532	Provides excellent connection to flow cells and switching valves
Pinkel fluorescence filter cube	Semrock	lf488/543/635-3x-a-000	Used in conjunction with Lambda DG4 to image multiple fluorophores rapidly
Press fit tubing connectors	GraceBio	46003	Clear silicone connector with adhesive that binds well to glass
Scanning mirrors	GSI Lumonics	VM500	Used to provide control of the second optical trap. GSI Lumonics no longer exists. Similar mirrors can be purchased from Cambridge Scientific
Stage	Leica
Stage micrometer	Electron Microscopy Sciences	68042-08	Provides on screen ruler for positioning of the beam and system calibration
Switching valves	IDEX	V-101T	Control direction of fluid flow and eliminate introduction of bubbles into flow cells
Syringe and valve manifold	Machine shop	None	Custom built
Syringe pump	Harvard Apparatus	PHD 2000	Controls fluid flow through flow cells
Syringe pump software	Harvard Apparatus	70-6000	Flow control provides seamless, programmable control of fluid flow
Syringes	Hamilton	81320	Gas-tight, PTFE Luer Lock, glass barrels with Teflon-coated plungers
Table top	Thor Labs	T36H	Optical table top or breadboard
Trapping laser	Newport/Spectra Physics	J-series; BL106C	Nd:YAG laser; 1064 nm; 5W laser