DNA molecule stretching through thermo-electrophoresis and thermal convection in a heated converging-diverging microchannel
© Hsieh et al.; licensee Springer. 2013
Received: 9 January 2013
Accepted: 28 January 2013
Published: 18 February 2013
A novel DNA molecule stretching technique is developed and tested herein. Through a heated converging-diverging microchannel, thermal convection and thermophoresis induced by regional heating are shown to significantly elongate single DNA molecules; they are visualized via a confocal laser scanning microscopy. In addition, electrophoretic stretching is also implemented to examine the hybrid effect on the conformation and dynamics of single DNA molecules. The physical properties of the DNA molecules are secured via experimental measurements.
Keywordssingle DNA molecule stretching CLSM thermo-electrophoresis converging–diverging microchannel
The past two decades has witnessed a tremendous growth in knowledge regarding the mechanical properties of DNA and its polymeric behavior. In addition, developments in molecular biology and micro- or nanotechnology have increased the interest of scientists and engineers in the mechanical manipulation of single DNA molecules. In fact, engineering DNA stretching would be a key step in the development of the next generation of biological microfluidic devices .
The ability to directly manipulate and visualize single DNA molecules has led to a number of advances in our current understanding of the physical and biological properties of DNA. Two general approaches to DNA stretching are in common use: DNA is stretched in a solution as it flows through a microchannel or it is stretched on a solid surface. Both approaches have their own advantages/disadvantages which depend on the particular application. For the former, with fluorescently labeled DNA molecules, it is possible to visualize the change in the conformation of a single DNA molecule under an optical microscope [2, 3].
Recently, Ichikawa et al.  have presented a novel DNA extension technique via laser heating. They proved that the new stretching technique was promising and could work in selected applications. Thermophoresis has also been found to play an important role in DNA molecule stretching.
The thermal convection induced in this study was similar to the convection that is inferred for the well-known Earth's mantle convection/or Bernard cell convection. Such convection produced the horizontal flow which caused the movement of the solution. Following , the governing equations of thermal convection in the study are the conservation equations of mass, momentum, and energy with the major dimensionless parameter of the Rayleigh number, indicating the vigor of convection and nondimensionalized heat flux.
PDMS flow cell fabrication
Channel total length, Lt
Channel test section length, Ls
Channel contraction length, Lm
Channel main width, Wm
Channel contraction width, Wc
Channel depth, H
Channel hydraulic diameter, Dh
66.67 ~ 160 μm
Channel contraction ratio
Channel expansion ratio
Electric field (kV/m), Ex
5, 7.5, 10
DNA concentration, μg/ml
Viscosity (cP), μ
Reynolds number, Re
0.032 ~ 0.064
λ-DNA contour length (μm) (labeled with YOYO-1)
Radius of λ-DNA gyration (μm)
Temperature ( C), T
Relaxation time (s), τr (Rouse model)
Relaxation time (s), τe (Experiment)
1.2 ~ 2.3
Velocity vector distribution
For the tested channels, precise information on the channel dimensions was extremely important in order to make an accurate evaluation. The depth, width, and length were measured optically within an accuracy of ±0.2%. To understand the surface condition of the present device, the roughness of the channel was measured along its center with a surface profilometer.
Each of the reservoirs (up/downstream plenum) had a volume of 0.15 ml. The channel had a total length of 30 mm, with a length of 800 μm for the test section. The detailed values of the test cells are listed in Table 1.
CLSM/μPIV and μLIF setup
The setup shown in Figure 3 was based on two pulsed Nd:YAG lasers (New Wave SoloII, New Wave Research, Fremont, CA, USA; 30 mJ, double cavity) firing on the second harmonic SoloII (green, 532 nm). The laser provided a laser beam with a measured area. The light was positioned so as to illuminate the entire inlet, outlet, and midsection of the channel. The laser pulse duration was 4 to 80 ms, based on the velocity magnitude. The test system was mounted on a movable xz stage on an inverted epifluorescence microscope (DMILM, Leica, Solms, Germany) with ×10 magnification, 0.25-numerical aperture panchromatic objective, and a field view of 800 × 600 μm. The measurement plane (i.e., the object plane) was precisely positioned relative to the test section by vertically moving the objective lens in the y direction and by horizontally moving the table in the x and z directions. The concentration of stained DNA molecules based on the interrogation volume was 8 × 107 particles/ml.
The images were recorded using a Dantec 80C77 HiSense PIV (Dantec Dynamics, Ulm, Germany) 1,344 × 1,024 × 12 bit interline transfer camera. Five images were taken for each flow field, with a spatial resolution of 32 × 32 pixels. The interrogation cell overlay was 50%. Background-noise influence was removed by subtracting the background intensity from the captured images. In addition, an ensemble averaging 20 images consecutively captured for 4 s was used to obtain the velocity measurements. The calculated measurement depth of the present μPIV was 50 μm. Each measurement was repeated at least three times under specified conditions. The measurements were conducted in the middle region at both the inlet and exit regions of the microchannel. The flow was found to have reached full hydrodynamic development at the middle region of the microchannel.
Visualization of the local buffer solution temperature was achieved with the same apparatus used for flow visualization and measurements (see Figure 3). However, instead of using stained DNA molecules, the channel was filled with a solution of rhodamine B, a fluorescent dye which shows a temperature-sensitive quantum yield in the range of 0°C to 100°C [5, 6]. Experiments were conducted with a fluorescence microscope equipped with a long-working distance ×10 objective lens. The images were recorded with the same equipment used for the μPIV measurements. From the captured images, the detailed temperature distribution could be extracted. Following , the intensity values of the captured images were converted to temperature using intensity-versus-temperature calibration; calibration of the intensity of temperature was made for each solution.
In the electro-osmotically driven flows, a 30-mm-long converging (8:1)-diverging (1:8) microchannel with a cross section of 100 × 400 μm and two reservoirs (up/downstream plenum) was used to supply a buffer of stained DNA molecules for the channel. Before use, the microchannel and entire flow loop were rinsed with DI water for at least 1 h to remove any contaminants. The transparent nature of the microchannel surfaces allowed visual examination of the channels to ensure that no bubbles were left. The buffer solution used was 1× Tris-borate with ethylenediaminetetraacetic acid (EDTA) (TBE) with pH 8.3. A schematic diagram showing the flow cell and the auxiliary system is given in Figure 3.
During each measurement, the microchannel was connected to small reservoirs. Current data were recorded from the power source by a personal computer-based data acquisition system. μPIV measurements were taken through a viewing window at midplane (y = 0) between the two cylindrical reservoirs with a diameter of 5 mm. The potential was applied via platinum electrodes immersed in the two 0.15-ml open reservoirs. The distance between the two reservoirs was 30 mm. When electric field was >10 kV/m, the EOF velocity of the solution will increase, and the mobility would be dependent on the electric strength [6, 7]. In order to avoid joule heating, electric field strengths of 5, 7.5, and 10 kV/m were thus applied.
The μPIV measurement system included visualization and the capture of images, the calculation of two-dimensional velocity vectors, and post-processing for data analysis. The vector field of the flow velocity within the measurement plane of the light sheet was determined by measuring the displacement of the tracer particles and the time durations of two laser pulses. A PIV 2100 processor (Dantec Inc.) was optimized to process the μPIV images into a raw vector map in real time and to transfer the map to a database in the PC. The processor employed cross-correlation to calculate the velocity vectors. A total of 800 sets of data was taken at each location for a specified Reynolds number (Re; i.e., the ratio of inertial forces to viscous forces). The selection of 800 data sets was based on the examination of the data convergence. One set of data consisted of five PIV vector data for a 32 × 32 pixel interrogation area. These data were statistically averaged, and the mean vector fields were obtained and used for the examination of the flow structure. The measurements were performed in a clean room at the University Microsystem Laboratory at a controlled ambient temperature of 298 K.
Methodology used (for electrophoretic mobility of DNA molecules and buffer solution EOF velocity) and temperature visualization
where is the total velocity of the seed particles (i.e., DNA molecules) by μPIV in treated PDMS channels, and is the electrophoretic velocity of the DNA molecules in the untreated PDMS channel.
With respect to measurement uncertainties, the most significant source of error was expected to be the measurements at the wall, and the biggest physical error in the μPIV data was the Brownian diffusion of the stained DNA molecules. Out-of-plane Brownian diffusion causes a reduction of the signal-to-noise ratio of the cross-correlation peak, and such an error was estimated. Errors due to in-plane Brownian diffusion were essentially eliminated by temporally averaging the results in the steady flow. In fact, experimental errors due to the limiting spatial resolution of the CCD camera, as well as errors in determining magnification, were therefore the major source of error in these results and found to be within ±15%.
Visualization of the local fluid temperature was achieved with the same apparatus used for flow visualization and measurements (see Figure 3). Instead of using fluorescent particles, however, the channel was filled with a solution of rhodamine B, a fluorescent dye which shows a temperature-sensitive quantum yield in the range of 0°C to 100°C . Experiments were conducted with a fluorescence microscope equipped with a long-working distance ×10 objective lens. The images were recorded with the same equipment used for the μPIV measurements. From the captured images, the detailed temperature distribution could be extracted. The intensity values of the captured images were converted to temperature using the intensity-versus-temperature calibration . A calibration of the intensity of temperature was made for each solution.
On the basis of standard lithography techniques, we constructed a 30-mm-long, 400-μm-wide, and 100-μm-high PDMS microchannel with a sudden contraction/expansion (a ratio of 8:1:8) test section 20 mm in length. Reservoirs (4 × 4 mm) were cut at each end of the curved PDMS microchannel with a scalpel, and the channels were soaked for 12 h at 45°C in 1× TBE (1× TBE contains, in 1 l, 108 g of Tris base, 55 g of boric acid, and 40 ml of 0.5 M EDTA, pH 8.3) to eliminate permeation-driven flow .
λ-phage double-strand DNA (dsDNA) from New England Biolabs (Ipswich, MA, USA) was used as the tracer in the present study. The DNA was stained, with respect to the backbone, with a fluorescent dye (YOYO-1, 4.7:1 bp/dye molecule), for a total length of 48.5 kbp DNA molecules, and diluted in 1× TBE. The dyed λ-DNA had a contour length (Lc) of 21 μm , and the longest relaxation time (τe) of 0.6 s (from uncoiled maximum length to coiled state) was measured and found in the present study.
Results and discussion
DNA molecule velocity profile with/without temperature effect
Local velocity map (μm/s) with different heating temperatures
Local velocity map (μm/s) with different heating temperatures and electric fields
Again, Figure 5 shows the velocity of the buffer solution convection observed for four different heating temperatures at the up, middle, and downstream locations, respectively (right half). The convection rates were approximately linear with the heating power and coincided with those found in Mao et al. , but they were strongly affected by the location where the velocity was measured. It was found that the convection effect became more dominant as the flow proceeded downstream, which was in good agreement with those of the temperature distributions, namely, the temperature gradient became steeper downstream than upstream.
DNA electrophoretic mobility and diffusion coefficient
Diffusion in the present study could be classified as translational diffusion or rotational diffusion. Only translational diffusion, i.e., diffusion of the center of the mass of DNA molecules, was considered. The translational diffusion was proportional to the thermal energy and, thus, proportional to kBT, as well as the effective viscous mobility, μ. Following , we approximated the DNA diffusion coefficient as D = kBT /6πηrg, where η is the buffer viscosity, rg is the gyration radius of the DNA molecule, and kB is the Boltzmann constant, which was 3.12 to 3.43 × 10−1 μm2/s in the temperature range of 25°C to 55°C, as shown in Figure 6b. Further comparisons were made with those of previous studies for μep and diffusion coefficient D, and the results are shown in Figure 6a,b, respectively. Given the different buffer solutions at different temperatures and the shorter gyration radius of the present study, as expected, the diffusion coefficient D was lower, as illustrated in Figure 6b.
Heating effect on DNA molecule stretching
Using detailed μLIF observations, thermophoresis, often called the Ludwig-Soret effect (thermal diffusion), was considered . The investigation of the Soret effect in the buffer solution was based on the determination of the following transport coefficient: Dmd, mutual diffusion coefficient; DT, thermal diffusion coefficient; and ST, Soret coefficient. Detailed calculation of the values of the above-stated parameters improved our basic understanding of the exact stretching mechanisms involved in this study. However, due to the limitation of the measurements, several physical quantities above were not available at this stage. Further study could include this aspect. Nevertheless, thermal convection, as well as diffusion, was still noted.
Heating effect on the histogram of DNA stretch ratio
Stretching force distribution
As Figure 11b shows, the present experimental data could be approximately fitted by applying the well-known WLC model. The hydrodynamic force measured and calculated through the Stokes formula was found to be a power law function of the forces via the WLC model with different exponents of 3.05 to 4.56 and different coefficients (C1 to C4) with different temperatures. Obviously, the stretching forces were greater than those predicted by the WLC. Furthermore, the temperature effect was again noted; otherwise the exponent found would have been nominally the same.
A decrease in the stretching force was observed as the solution temperature increased, which was in good agreement with that in Williams et al. .
Although thermophoretic stretching was not clearly noted, the effect still seemed to occur and to increase as the temperature increased.
In addition to electrophoretic stretching, thermal convection made an equal contribution in terms of DNA molecule stretching.
As a result of points 2 and 3, when the buffer solution temperature increased, the stretch ratio was three to four times that of the isothermal buffer solution.
DNA molecule thermal expansion played a significant role (≥50%) in DNA molecule stretching. Therefore, the present stretching mechanism included thermal expansion, thermal diffusion (thermophoresis), and thermal convection.
Electrophoretic mobility and the translational diffusion coefficient of the DNA molecules were obtained and compared with those of existing data.
SSH is a professor at the Department of Mechanical and Electro Mechanical Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan, Republic of China. JHC is currently working towards a PhD degree at the Department of Mechanical and Electro Mechanical Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan, Republic of China. CFT is a student working towards a master's degree at the Department of Mechanical and Electro Mechanical Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan, Republic of China.
confocal laser scanning microscopy
micro-particle image velocimetry
This work was supported by the National Science Council (NSC) of Taiwan under contract number NSC 101-2221-E-110-043-MY3.
- Randall GC, Schultz KM, Doyle PS: Methods to electrophoretically stretch DNA: microcontractions, gels, and hybrid gel-microcontraction devices. Lab Chip 2006, 6: 516–525. 10.1039/b515326cView ArticleGoogle Scholar
- Hsieh SS, Liu CH, Liou JH: Dynamics of DNA molecules in a cross-slot microchannels. Meas Sci Technol 2007, 18: 2907–2915. 10.1088/0957-0233/18/9/021View ArticleGoogle Scholar
- Hsieh SS, Liou JH: DNA molecules in converging–diverging microchannels. Biotechnol Appl Biochem 2009, 52: 29–40. 10.1042/BA20070219View ArticleGoogle Scholar
- Ichikawa M, Ichikawa H, Yoshikawa K, Kimara Y: Extension of a DNA molecule by local heating with a laser. Phys Rev Lett 2007, 99: 148104.View ArticleGoogle Scholar
- Ross D, Gaitan M, Locascio LE: Temperature measurement in microfluidic systems using a temperature-dependent fluorescent dye. Anal Chem 2001, 73: 4117–4123. 10.1021/ac010370lView ArticleGoogle Scholar
- Hsieh SS, Yang TK: Electroosmotic flow in rectangular microchannels with joule heating effects. J Micromech Microeng 2008, 18: 025025. 10.1088/0960-1317/18/2/025025View ArticleGoogle Scholar
- Hsieh SS, Lin HC, Lin CY: Electroosmotic flow velocity measurements in a square microchannel. Colloid Polym Sci 2006, 284: 1275–1286. 10.1007/s00396-006-1508-5View ArticleGoogle Scholar
- Mao H, Arias-Gonzalez JR, Smith SB, Tinoco JI, Bustamante C: Temperature control methods in a laser tweezers system. Biophys J 2005, 89: 1308–1316. 10.1529/biophysj.104.054536View ArticleGoogle Scholar
- Kirby BJ: Micro-and nanoscale fluid mechanics-transport in micro fluidic devices. New York: Cambridge University Press; 1979.Google Scholar
- Nkodo AE, Garnier JM, Tinland B, Ren H, Desruisseaux C, McCormick LC, Drouin G, Slater GW: Diffusion coefficient of DNA molecules during free solution electrophoresis. Electrophoresis 2001, 22: 2424–2432. 10.1002/1522-2683(200107)22:12<2424::AID-ELPS2424>3.0.CO;2-1View ArticleGoogle Scholar
- Sato H, Masubuchi Y, Watanabe H: DNA diffusion in aqueous solution in presence of suspended particles. J Polymer Sci, Part B: Polymer Phys 2009, 47: 1103–1111. 10.1002/polb.21715View ArticleGoogle Scholar
- Schallhorn K, Kim M, Ke PC: A single-molecule study on the structural damage of ultraviolet radiated DNA. Int J Mol Sci 2008, 9: 662–667. 10.3390/ijms9040662View ArticleGoogle Scholar
- Smith DE, Perkins TT, Chu S: Dynamical scaling of DNA diffusion coefficients. Macromolecules 1996, 2: 1372–1373.View ArticleGoogle Scholar
- Braun D, Libchaber A: Trapping of DNA by thermophoretic depletion and convection. Phys Rev Lett 2002, 89: 188103.View ArticleGoogle Scholar
- Williams MC, Wenner JR, Rouzina I, Bloomfield VA: Entropy and heat capacity of DNA melting from temperature dependence of single molecule sketching. Biophys J 2001, 80: 1932–1939. 10.1016/S0006-3495(01)76163-2View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.