Watching individual proteins acting on single molecules of DNA

Abstract

In traditional biochemical experiments, the behavior of individual proteins is obscured by ensemble averaging. To better understand the behavior of proteins that bind to and/or translocate on DNA, we have developed instrumentation that uses optical trapping, microfluidic solution delivery, and fluorescent microscopy to visualize either individual proteins or assemblies of proteins acting on single molecules of DNA. The general experimental design involves attaching a single DNA molecule to a polystyrene microsphere that is then used as a microscopic handle to manipulate individual DNA molecules with a laser trap. Visualization is achieved by fluorescently labeling either the DNA or the protein of interest, followed by direct imaging using high-sensitivity fluorescence microscopy. We describe the sample preparation and instrumentation used to visualize the interaction of individual proteins with single molecules of DNA. As examples, we describe the application of these methods to the study of proteins involved in recombination-mediated DNA repair, a process essential for the maintenance of genomic integrity.

Figure 13.1

Illustration of the three-channel flow cells used in the single-molecule experiments. (A) Photograph of a three-channel flow cell. The flow cell is fabricated using the process described in Section 4.2. To demonstrate the flow path, green dye flows through channels I and III, whereas yellow dye flows through channel 2. (B) Schematic (drawing not to scale) of a three-channel flow cell showing typical dimensions; magnification shows the detail at the end of channel divider. The divider is ∼100 μm wide with a semicircular end of radius of ∼50 μm. The gray area to the right of the divider illustrates the region inside the mean diffusion length boundary. Experiments are conducted at a point ∼200 μm downstream of the divider, ∼750 μm into each channel, and ∼35 μm from the surface where the effects of diffusion are minimal.

Figure 13.2

The calculated concentrations of Mg²⁺, D = 10⁻⁵ cm ²/s (A, left panel), and ATP, D = 10⁻⁶ cm ²/s (B, left panel) as a function of position for diffusion from channel I into channel II. The flow cell has the same dimensions as described in Fig. 13.1, and the flow rate, v, is 50 μm/s from left to right; the end of the divider is at the origin of the plot. The cross section of concentration as a function of the distance downstream of the flow cell is also shown (right panels). The solid white line in the left panels indicates the mean diffusion distance of the solute from and into each channel. The calculations were performed in MATLAB (Math Works).

Figure 13.3

Flow diagram for the microfabrication of a three-channel flow cell (see text for details).

Figure 13.4

Schematic diagram of the microscope, optical trap, and flow cell. The trapping IR laser initially passes through a 20× beam expander, and is then further collimated and steered by lenses L1 and L2 an electronic shutter (S) is in-between. A high-pass IR dichroic mirror (DM1) directs laser beam into the objective (OBJ). The flow cell is mounted on an x–y translocation stage that is controlled by a computer (PC) solutions are delivered to the flow cell using a multisyringe pump. A high-pressure mercury arc lamp is used for illumination (fluorescence and bright field). A second dichroic mirror (DM2) is used to image the fluorescent protein–DNA–bead complex onto an electron bombardment camera; the real-time image is displayed on a monitor.

Figure 13.5

Schematic diagram for a dual laser-trap microscope. Lenses L1 and L2 initially collimate and expand the laser. The first beam path (black line) passes through an AOM which is imaged on the back aperture of the objective lens by lenses L3, L4, L5, and L9. The second beam path (gray line) is reflected off a movable mirror which is imaged onto the back aperture of the objective by lenses L6, L7, L8, and L9. The image from the objective (dashed line) is split between a camera that images the fluorescent protein–DNA–bead complex, and a quadrant photodiode (QPD) for position detection of the bead in the first trap by means of a half-mirror (HM). The signal from the QPD passes through an amplifier and a low-pass filter before being processed by a PC which uses the information to control the AOM, thus providing feedback control on the position of the first trap with nm resolution. Mirrors (M) and beam splitters (BS) serve to direct the beam path. A dichroic mirror (DM) is used to direct the trapping lasers into the objective and to pass light from the Xenon lamp to the camera and QPD; the real-time image is displayed on a monitor. Lenses L10 and L11 image the trapped bead onto the camera and QPD.

Figure 13.6

Schematic illustration of the components used for temperature measurement and control. Top: the flow cell used for temperature determination is made of a glass slide, a coverslip, and a layer of Parafilm sandwiched in between. A thermocouple is placed inside the channel. Bottom: an illustration of the temperature controller (side view; not to scale).

Figure 13.7

Measurement of IR laser-induced temperature changes. (A) Relative fluorescence as a function of temperature. The ratio of the background-corrected fluorescence of RhB and Alexa-488 is normalized to that at 25 °C. The relative fluorescence intensity decreases 2% per °C increase. (B) Relationship between induced temperature change and the IR laser power delivered at the focus, at a starting temperature of 25 °C.

Figure 13.8

Thermal gradient in a flow cell due to infrared heating by an optical trap. (A) Example showing the measured temperature distribution in an optical trap with an IR laser power of 587 m W when the temperature is set at a starting temperature of 24 °C using only the microscope slide temperature controller. The circle indicates the position of the trap. The rectangle indicates the region where the temperature shown in B is measured. (B) Plot of the temperature distribution in the horizontal direction of the region shown in (A). Black: measured temperature; gray: linear fitting from 0.5 to 14.5 μm away from the trap center. Linear fitting gives a temperature gradient of 0.06 °C/μm.

Figure 13.9

RecBCD translocating through, and unwinding, an individual λ DNA molecule. (A) Kymograph showing a YOYO-1 stained λ dsDNA molecule being unwound by a RecBCD molecule bound to the free DNA end. The drawing to the left of the kymograph depicts the optically trapped bead–YOYO-1-DNA–RecBCD complex. (B) Plot of DNA length versus time. Black line shows the fit to a straight line. (C) Kymograph showing translocation by a fluorescent nanoparticle-labeled RecBCD molecule on λ dsDNA. The drawing on the left side of the kymograph depicts the optically trapped bead–DNA–RecBCD–nanoparticle complex. (D) Plot of the position of the RecBCD molecule, indicated by the nanoparticle, versus time. The black line shows the fit to a straight line. Note that the difference in unwinding rates in (B) and (D) is not due to a difference in the techniques, but rather reflects the intrinsic heterogeneity of individual RecBCD enzyme behavior.

Figure 13.10

Rad54 translocating on a single dsDNA molecule. (A) Schematic illustration of the optically trapped λ DNA–bead complex with a bound FITC–Rad54 complex. (B) Kymographs depicting upstream translocation (in the direction opposite to flow) of Rad54 on the dsDNA. (C) Plot of FITC–Rad54 position relative to the bead versus time.

Figure 13.11

Rad51 assembling onto, and dissociating from, a single dsDNA molecule. (A) Kymograph of Rad51 assembly on Cy3-end-labeled λ DNA. The schematic on the left side of kymograph depicts the optically trapped bead, initial position of Cy3-end-label of the DNA (star), and DNA (solid line); the schematic on the right side depicts the extended Rad51 nucleoprotein filament. DNA length is measured from the center of the bead to the Cy3-end-label. (B) Plot of DNA length versus time for the assembly of Rad51 on DNA analyzed using two-dimensional Gaussian fitting of the end-label position. (C) Kymograph of Rad51 disassembly from Cy3-end-labeled λ DNA. The schematic on the left side of kymograph depicts the optically trapped bead, initial position of Cy3-end-label of the DNA (star), Rad51 (filled circles), and DNA (solid line). (D) Plot of DNA length versus time for the disassembly of Rad51 from DNA.

Figure 13.12

Direct visualization of nucleation and growth of RecA-RFP nucleoprotein filaments on individual molecules of dsDNA. (A) Illustration of the nucleation of RecA-RFP clusters on DNA. (B) Representative video frames showing nucleation at four different RecA-RFP concentrations. Flow is left to right. Each vertical strip represents the same DNA molecule repeatedly dipped into the RecA-RFP protein solution for the incubation times indicated. The trapped bead position is indicated by an arrow; the bead is fluorescent due to the nonspecific binding of the RecA-RFP.