Sizing of Single Globular DNA Molecules by Using a Circular

May 20, 2008 - Health Technology Research Center, National Institute of Advanced Industrial ... and Department of Applied Chemistry, Graduate School o...
1 downloads 0 Views 248KB Size
Download PDF
Anal. Chem. 2008, 80, 5197–5202

Sizing of Single Globular DNA Molecules by Using a Circular Acceleration Technique with Laser Trapping Ken Hirano,*,† Hideya Nagata,† Tomomi Ishido,† Yoshio Tanaka,† Yoshinobu Baba,†,‡ and Mitsuru Ishikawa† Health Technology Research Center, National Institute of Advanced Industrial Science and Technology (AIST), 2217-14, Hayashi-cho, Takamatsu, Kagawa, 761-0395, Japan, and Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan We describe a method for in situ sizing individual huge DNA molecules by laser trapping. Single DNA molecules are reversibly transformed, without mechanical fragmentation of fragile huge-sized DNA, from their random coil state into their globular state induced by condensing agents poly(ethylene glycol) and Mg2+. With the use of a globular DNA molecule folded by condensation, the critical velocity of the circularly accelerated single globular DNA molecule by laser trapping was found to be proportional to the size of the DNA. Yeast, Saccharomyces cerevisiae, chromosome III (285 kbp) was successfully sized (281 ( 40 kbp) from a calibration curve scaled using λ, T4, and yeast chromosome VI (48.5, 166, and 385 kbp, respectively). The use of critical velocity as a sizing parameter makes it possible to size single DNA molecules without prior conformational information, i.e., the radius of a single globular huge DNA molecule as a nanoparticle. A sized single globular DNA molecule could be trapped again for subsequent manipulation, such as transportation of it anywhere. We also investigated a possibility of reusing the globular DNA molecules condensed by PEG and Mg2+ for PCR and found that PCR efficiency was not deteriorated in the presence of the condensation agents. The laser-trapping technique has numerous applications in studies of biological and biophysical systems that involve the manipulation of single DNA molecules, such as the analysis of physical and chemical properties of a single DNA molecule,1,2 the observationofbiologicaleventsinvolvingDNA-proteininteractions,3,4 and the measurement of DNA-nanostructure mechanical interactions.5 Manipulation of single huge DNA molecules, over several * To whom correspondence should be addressed. E-mail: [email protected]. Phone and Fax: +81-87-869-3569. † National Institute of Advanced Industrial Science and Technology. ‡ Nagoya University. (1) Perkins, T. T.; Quake, S. R.; Smith, D. E.; Chu, S. Science 1994, 264, 822– 826. (2) Smith, S. B.; Cui, Y.; Bustamante, C. Science 1996, 271, 795–799. (3) Comish, P. V.; Ha, T. ACS Chem. Biol. 2007, 2, 53–61. (4) Abbondanzieri, E. A.; Greenleaf, W. J.; Shaevitz, J. W.; Landick, R.; Block, S. M. Nature 2005, 438, 460–465. (5) Keyser, U. F.; van den Dose, J.; Dekker, C.; Dekker, N. H. Rev. Sci. Instrum. 2006, 77, 105105. 10.1021/ac8003538 CCC: $40.75  2008 American Chemical Society Published on Web 05/20/2008

tens of thousands of base pairs (kbp) long, is an important technique in molecular biology and nanotechnology.6–9 However, we have a limitation of length in manipulating random coil DNA molecules in solution, because a hydrodynamic shear force produced randomly by large velocity gradients in solution generates tension against DNA molecules and causes fragmentation of such huge DNA molecules. Levinthal and Davison found that T2 phage DNA (166 kbp) is fragmented at a critical flow rate of 46 µL/s using a capillary in 125 µm radius for quantitative analysis of fragmentation.10 Indeed, to avoid shearing during the extraction of intact huge DNA molecules, cells are lysed and biochemically treated further in situ in an agarose gel plug.11 The use of a DNA-bead complex or laser trapping of several nanospheres encircling a single DNA molecule permits randomly coiled DNA molecules to be manipulated without fragmentation.12 However, there are limits on the size of DNA (several tens of kbp) that can be manipulated without fragmentation by the DNA-bead complex method. On the other hand, DNA molecules have simply and artificially been folded by mixing with chemical reagents 13,14 or proteins.7,15 These folded globular DNA molecules were subsequently manipulated without fragmentation by laser trapping in solution. For instance, the use of poly(ethylene glycol) (PEG) with lowmolecular weight inorganic cations condenses a single DNA molecule from its random coil state to its globular state, being dependent upon concentrations of PEG and inorganic cations.14 In this condensation, the DNA strand is formed into a hexagonal (6) Hirano, K.; Matsuzawa, Y.; Yasuda, H.; Katsura, S.; Mizuno, A. Proceedings of the µ-TAS 2000 Symposium, Enschede, The Netherlands, May 14-18, 2000; pp 439-442. (7) Oana, H.; Kubo, K.; Yoshikawa, K.; Atomi, H.; Imanaka, T. Appl. Phys. Lett. 2004, 85, 5090–5092. (8) Rhee, M.; Burns, M. A. Trends Biotechnol. 2007, 25, 174–181. (9) Michalet, X.; Ekong, R.; Fougerousse, F.; Rousseaux, S.; Schurra, C.; Hornigold, N.; van Slegtenhorst, M.; Wolfe, J.; Povey, S.; Beckmann, J. S.; Bensimon, A. Science 1997, 277, 1518–1523. (10) Levinthal, C.; Davison, P. F. J. Mol. Biol. 1961, 3, 674–683. (11) Sambrook, J.; Russell, D. W. Molecular Cloning, 3rd ed.; Cold Spring Harbor Laboratory Press: New York, 2001; pp 5.61-5.73. (12) Hirano, K.; Baba, Y.; Matsuzawa, Y.; Mizuno, A. Appl. Phys. Lett. 2002, 80, 515–517. (13) Matsuzawa, Y.; Hirano, K.; Mizuno, A.; Ichikawa, M.; Yoshikawa, K. Appl. Phys. Lett. 2002, 81, 3494–3496. (14) Matsuzawa, Y.; Hirano, K.; Mori, K.; Katsura, S.; Yoshikawa, K.; Mizuno, A. J. Am. Chem. Soc. 1999, 121, 11581–11582. (15) Laurence, R. B.; Corzett, M.; Balhorn, R. Science 1999, 286, 120–123.

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5197

arrangement and then shaped into a toroidal or a rod structure by condensing agents.16–18 This condensation dramatically reduces fragmentation of huge DNA molecules because the physical size of globular DNA is very much smaller than that of random coil DNA; indeed, shear force produced by velocity gradients decreases with reducing length of polymers.10 Thus, a fragile huge DNA can be manipulated without its fragmentation. Moreover, the refractive index of this highly compacted DNA molecule will be higher than that of the surrounding media, thereby allowing laser trapping by the enhanced trapping force.13,19 This technique of manipulating single globular DNA molecules by laser trapping is a powerful tool not only in biology but also in physical chemistry and nanotechnology.20,21 However, no information on the size of the globular DNA molecule was obtained from this technique, because of the lack of a physical relationship between the laser trapping force and the size of the DNA molecule. If the size can be obtained from the globular DNA molecules reversibly transformable to a random coil state, a new in situ sizing technique of huge-sized DNA, such as chromosomal DNA, is achieved at the single-molecule level without the mechanical fragmentation. Indeed, information on the size of single DNA molecules in situ is important for subsequent biophysical and biochemical analyses of single DNA molecules in single-molecule manipulation. For instance, future singlemolecule genome analyses, such as single-molecule haplotyping and sequencing, might require the selection of target DNA molecules from a known DNA mixture in a genome library or an unknown DNA mixture extracted from a cell. For this requirement, it is also essential that single target huge DNA molecules remain capable of subsequent reuse and manipulation after sizing, for example, by transportation by laser trapping followed by biochemical treatment, such as single-molecule polymerase chain reactions (PCR).22,23 Here we demonstrate a technique of in situ sizing single huge DNA molecules in the globular state by circular acceleration through laser trapping. From the hydrodynamic drag force generated by the rotation, we evaluated the critical velocity above which a globular single DNA molecule, folded by PEG and MgCl2 as condensing agents, will be released from the laser focal point balanced by the laser trapping force and the hydrodynamic drag force. Furthermore, for future application of the current proposed sizing method, we also investigated a possibility of reusing the globular DNA molecules condensed by PEG and MgCl2 treatment for PCR. EXPERIMENTAL SECTION Sample Preparation. The DNA samples used in the present study were λ and T4 phage DNA (Nippon Gene, 48.5 and 166 kbp, respectively) and two sections of chromosomal DNA from (16) Schellman, J. A.; Parthasarathy, N. J. Mol. Biol. 1984, 175, 313–329. (17) Maniatis, T.; Venable, J. H.; Lerman, L. S. J. Mol. Biol. 1974, 84, 37–64. (18) Me’lnikov, S. M.; Sergeyev, V. G.; Yoshikawa, K.; Takahashi, H.; Hatta, L. J. Chem. Phys. 1997, 107, 6917–6924. (19) Chiu, D. T.; Zare, R. N. J. Am. Chem. Soc. 1996, 118, 6512–6513. (20) Ichikawa, M.; Matsuzawa, Y.; Koyama, Y.; Yoshikawa, K. Langmuir 2003, 19, 5444–5447. (21) Mayama, H.; Nomura, S. M.; Oana, H.; Yoshikawa, K. Chem. Phys. Lett. 2000, 330, 361–367. (22) Lagally, E. T.; Medintz, I.; Mathies, R. A. Anal. Chem. 2001, 73, 565–570. (23) Leamon, J. H.; Link, D. R.; Egholm, M.; Rothberg, J. M. Nat. Methods 2006, 3, 541–543.

5198

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

the yeast Saccharomyces cerevisiae (285 kbp and 365 kbp from chromosome VI and chromosome III, respectively). The yeast DNA samples were separated and extracted in their sizes from gel plugs of intact yeast chromosomal DNA (strain YNN295, BioRad) by pulsed-field gel electrophoresis (PFGE) with a CHEFDRII system (BioRad). PFGE was performed on a 1.0% certified megabase agarose gel (BioRad) in 0.5× TBE buffer (45 mM Tris, 45 mM boric acid, and 1 mM EDTA, pH 8.3) at 14 °C. The run time was 24 h at 6 V/cm with a 60-120 s (initial to final) switch time ramp at 120° pulsed-field angle. The slices of agarose containing the bands of each section size of chromosomal DNA were cut out from a gel after PFGE. The chromosomal DNA divided into each section was extracted from a cut gel slice by an electroelution method with a Maxyield-NP (Atto). An electroelution was carried out in 0.5× TBE buffer under the following conditions: electric field strength, 50 V; running time, 60 min; dialysis, membrane 500 molecular weight cutoff. An eluted DNA solution was dialyzed to exchange with TE buffer (10 mM Tris and 1 mM EDTA, pH 7.2) using a dialysis membrane (molecular weight cutoff, 500) at 4 °C for 12 h. The globular DNA was prepared by treatment with condensing agents PEG and MgCl2. This condensation method exclusively generates individual globular DNA molecules.14,24 The morphology of globular DNA molecules was analyzed from their fluorescence images. In the current experiment, The critical concentrations required to fold the DNA were 100 mg/mL for PEG6000 (Hampton Research) and 40 mM for MgCl2. The final sample solution was consisted of 0.3 µM DNA as nucleotides in 20 mM MOPS buffer (pH 7.0), 0.6 µM 4′,6-diamidino-2-phenylindole (DAPI) as a fluorescent dye, 100 mg/mL PEG, and 40 mM MgCl2. The sample was incubated for 15 min at room temperature (23.8-24.7 °C) to form a globular DNA. Fluorescent polystyrene beads of 0.1, 0.2, 1.0, and 2.0 µm in diameter (Molecular Probes) were used for optimizing the laser power for trapping and were diluted to 5.7 × 107 particles/mL: the composition of the bead solution was identical to that of the sample solution except for DNA; thus, the viscosities of the two solutions were equal to each other. The viscosity of the sample solution, which was a parameter for estimation of trapping force, was measured with a Visconic ELD viscometer (Tokimec) at 22 ± 0.2 °C. A stretched single DNA molecule was immobilized on an amine-modified coverslip (Matsunami Glass) using a method of dynamic molecular combing to find a difference in DNA physical sizes before and after condensation.9 To immobilize T4 DNA molecules, a 10 µL drop of a DNA solution (2.6 ng/µL in nucleotide in 10 mM MOPS buffer, pH 7.0) was placed on an amine-modified coverslip. The coverslip was spun and dried to immobilize stretched single DNA molecules using a dc motordriven rotation stage at 700 rpm. The immobilized single DNA molecules were stained with a 0.6 µM DAPI solution in 20 mM MOPS buffer, pH 7.0, for 5 min to be fluorescently visualized for microscopic observation. Instrumentation. Fluorescence images of single DNA molecules and beads were obtained using an inverted fluorescence microscope (Olympus IX-70) equipped with a 100×, NA1.30 oilimmersion objective lens and a highly sensitive EB-CCD camera (24) Yoshikawa, K.; Matsuzawa, Y. J. Am. Chem. Soc. 1996, 118, 929–930.

Figure 2. (a) Stretched single T4 DNA molecule (166 kbp) immobilized on an amine-modified coverslip by using a method of dynamic molecular combing (ref 9). (b) A single T4 DNA molecule condensed by PEG and MgCl2. The scale bar is 10 µm for both images.

Figure 1. Schematic of in situ sizing a single huge-sized DNA molecule using circularly accelerating motion with laser trapping. A trapped single globular DNA molecule receives a hydrodynamic force during circular motion. Once accelerated circular motion released a trapped single DNA molecule from the laser focus (trapping point) above a critical velocity, at which a trapping force is balanced by the hydrodynamic force. The measured critical velocity was proportional to the size of a single DNA molecule.

(Hamamatsu Photonics). Fluorescence images were recorded on digital videotape in real time. An apparatus of laser trapping in the current study was the same apparatus as used in our previous report.25 A beam from a Nd:YAG laser (1064 nm, CW, Spectron Laser Systems) was introduced into the microscope and focused on the plane of the fluorescence image using the objective lens to trap optically single DNA molecules and fluorescent beads. To measure a critical velocity of single globular DNA molecules, the laser focus, i.e., trapping point, was circularly scanned in on the observation plane at an arbitrary speed. The laser beam for trapping was scanned using a pair of galvano mirrors to control x- and y-axis scanning. Each mirror was driven by the following functions: x(t) ) A cos θ(t), y(t) ) A sin θ(t), where θ(t) is an angle of a mirror oscillation and A is an amplitude of a mirror oscillation. Then, a beam of light reflected from the mirrors traced a circle, i.e., Lissajous’ figure x2 + y2 ) A2 to adjust the radius of circular scanning. The radius of the circle drawn by a laser trapping beam was 10 µm on the observation plane in the current experiment. The laser beam was circularly scanned at intervals of 1.0° using computer-controlled galvano mirrors. The step of 1.0° is equivalent to a step of 175 nm. Sizing Method. Figure 1 shows a schematic of sizing a single DNA molecule. First, a target single globular DNA molecule is trapped using a laser, and then it is circularly moved to size it. A single DNA molecule moving in an aqueous solution is exerted by the hydrodynamic force (Stokes’s drag force), which is proportional to the moving velocity, the particle size, and the viscosity of a solution. An accelerated circle motion released a trapped single globular DNA molecule from the laser focus (trapping point) above the critical velocity, at which a laser trapping force is balanced by the hydrodynamic force. The size of a single DNA molecule was estimated by measuring the critical velocity, which is related to its size. To measure the critical velocity, accelerating a circle motion was controlled from 0 to 900 µm/s at intervals of 1 µm/s. A single globular DNA molecule was (25) Tanaka, Y.; Hirano, K.; Nagata, H.; Ishikawa, M. Electron. Lett. 2007, 43, 412–413.

manipulated using laser power from 200 to 800 mW. A 20 µL droplet of a sample solution was sandwiched with two coverslips sealed with glue to prevent the sample solution from evaporating. PCR of Globular DNA. To investigate the influence of PEG and Mg2+ on PCR, the product of PCR amplification was examined using the DNA samples treated with PEG and Mg2+. Three kinds of λ phage DNA as the templates were prepared; native DNA template as positive control (0.22 ng/µL DNA in TE buffer), globular DNA template without fluorescent dye staining (0.22 ng/ µL DNA in 20 mM MOPS buffer, pH 7.0, 100 mg/mL PEG, and 40 mM MgCl2), and globular DNA template with fluorescent dye staining (0.22 ng/µL DNA in 20 mM MOPS buffer, pH 7.0, 0.6 µM DAPI, 100 mg/mL PEG, and 40 mM MgCl2). The composition of globular DNA solutions was identical to that of sample solutions for sizing, which was described in the Sample Preparation section, except for the presence of DAPI as a fluorescent dye. A target fragment of DNA template, which is 3980 bp in length, was amplified by PCR using primers located at 7801-11781 in λ DNA. The sequences of the sense and antisense primers are TCCCTGTTTGTCCGGACTGA and TCCTGACGGGCGGTATATTT, respectively. A 3 µL aliquot (6.6 pg) of each template DNA was amplified using 2.5 U TaKaRa LA Taq DNA polymerase under a recommended PCR condition in 50 µL total volume. A cocktail including 0.4 mM each dNTP, 2.5 mM MgCl2, and 0.2 µM each primer was treated in the following thermal cycle sequence: preheating at 94 °C for 1 min; 25 cycles of heating, cooling, and heating at 98 °C for 10 s, at 65 °C for 3 min, and at 72 °C for 5 min, respectively; finally additional extension at 72 °C for 5 min. After the PCR amplification, an aliquot of 0.5 µL of reacted solution was added to a 10 µL sample loading buffer (10 mM Tris-HCl, pH 8.0, 5% glycerol, 1 mM EDTA, 0.04 wt % bromophenol blue, 0.04 wt % xylene cyanol FF) for analysis with 1.0 wt % agarose gel electrophoresis. After the electrophoresis, DNA bands were fluorescently visualized using ethidium bromide. The stained gel image was acquired using a digital camera system (FluorChem, Alpha Innotech). The captured gel image was then analyzed using Gel-Pro Analyzer software (Media Cybernetics). RESULTS AND DISCUSSION Condensation of a single DNA molecule into a globular form enhances the refractive index of a single DNA molecule compared with that of the surrounding medium, thereby allowing laser trapping of a globular DNA due to the enhanced trapping force favored by the enhanced refractive index.14,19 Parts a and b of Figure 2 show fluorescence images of single Τ4 DNA molecules (166 kbp) to compare their sizes before and after globular folding. Figure 2a shows that a stretched Τ4 DNA molecule measured 60.3 µm long (approximately 3 kbp/µm in B-form DNA) before Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5199

folding. It has been reported that the conformation of DNA can be changed from the coiled to the globular state with increasing concentration of PEG and MgCl2.14 Figure 2b shows a single globular DNA molecule condensed using 100 mg/mL PEG and 40 mM MgCl2. This dramatically condensed globular DNA molecule forms a toroidal structure, which is approximately 100 nm in radius and is composed of hexagonally close-packed DNA strings.16–18 Thus, the effective volume of the DNA is reduced by a factor of ∼10-4. Hydrodynamic force exerted on a globular DNA molecule or a fluorescent bead moving in a solution is given by Stokes’s drag law: Fdrag ) 6πηRv

(1)

where η is the viscosity of the hydrodynamic solution, R is the hydrodynamic radius of the particle, and vc is its critical velocity. Acceleration of circular rotation allows a trapped particle to be released from the laser focal point as a result of the hydrodynamic drag force, because the hydrodynamic drag force Fdrag is balanced by the laser trapping force Ftrap at the critical velocity vc. We obtain Ftrap ) 6πηRvc

(2)

The application of eq 2 to evaluating a relationship between Ftrap and vc for a globular DNA molecule is an unrealistic task. Indeed, the radius of globular huge DNA molecules is generally unknown, although it has been estimated theoretically and evaluated experimentally by electron microscopy for specific samples of DNA, such as T4 phage DNA.26,27 Moreover, it is necessary to estimate a suitable laser power for trapping a single globular DNA molecule and to find the actual applicability of eq 2. For these two reasons we evaluated the relationship between the known size of nanospheres and their critical velocity. The diameter of beads was selected from 100 nm to 2 µm to cover the expected sizes of globular DNA molecules. Indeed, the diameter of globular DNA is expected to be 100 nm or larger on the basis of previous experiments.16,17 Figure 3a shows the critical velocity for a polystyrene bead determined by accelerating the circular motion of a trapped bead at a laser power from 200 to 800 mW. The critical velocity increased with increasing bead diameter and was strongly saturated using bead diameters above 1.0 µm. Moreover, the slope of a plot of the critical velocity versus the bead diameter increased with increasing laser power before saturation. The strong saturation may arouse our concern for malfunction in the current experimental setup. Figure 3b shows the dependence of the trapping force on the diameter of beads. The trapping force was calculated from eq 2 using the known parameters: the radius of spherical polystyrene beads R, the critical velocity vc, and the viscosity of a solution η. The viscosity of a mixture of 100 mg/ mL PEG and 40 mM MgCl2 was 4.53 mPa · s. The trapping force increased with increasing bead size and laser power. In contrast to the relationship between bead diameter and critical velocity, (26) Plum, G. E.; Arscott, P. G.; Bloomfield, V. A. Biopolymers 1990, 30, 631– 643. (27) Takenaka, Y.; Yoshikawa, K.; Yoshikawa, Y.; Koyama, Y.; Kanbe, T. J. Chem. Phys. 2005, 123, 014902.

5200

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

Figure 3. Dependence of (a) the critical velocity and (b) the trapping force on the diameter of polystyrene beads. Laser trapping was performed at 200, 400, 600, and 800 mW.

no strong saturation is found in the relationship between bead diameter and trapping force, thus removing the concern of the malfunction of the experimental setup. A particle much smaller than the laser wavelength, that is, a Rayleigh particle, is equivalent to a simple dipole. In general, the trapping force toward the center of the laser beam (Ftrap) for an object with permittivity (εp) and volume (V) in an electric field (E) within the surrounding media with permittivity (εs) is given by14 Ftrap ∝ (εp - εs )V grad E2

(3)

The critical velocity vc in eq 2 is proportional to a square of the particle radius R2 because Stokes’s drag force Ftrap in eq 1 is proportional to the particle radius R, whereas a laser trapping force is proportional to the particle volume, that is, R3. In Figure 3, parts a and b, we find deviation from ideal R2 dependence of vc on 2R smaller than 1.0 µm and ideal R3 dependence of Ftrap on 2R smaller than 2.0 µm; indeed each dependence on 2R looks linear with increasing laser power. This deviation is likely due to departure of actual experimental conditions from the ideal requirements used to obtain eq 1. Although we have experimentally evaluated dependence of critical velocity and trapping force on particle diameter using known-size polystyrene beads, the use of critical velocity, without the radius of a nanoparticle, is favorable for sizing single globular huge DNA molecules with unknown radii. Moreover, the slope of critical velocity increased with increasing laser power. Thus, the critical velocity at a laser power of 800 mW provides the optimum condition for sizing single globular huge DNA molecules under the current experimental conditions.

Figure 4. (a) Schematic of the expected trajectory of a globular single λ DNA molecule (48.5 kbp) in circular motion induced by laser trapping. (b) A folded single DNA molecule was trapped at the initial position indicated by the open arrowhead, (c) accelerated along a circle, and (d) finally released from the laser focal point at critical velocity and diffused from the circle by Brownian motion. The scale bar is 10 µm.

Figure 5. (a) Histograms of critical velocities for three globular DNA molecules, 48.5, 166, and 365 kbp. Each histogram was fitted into a Gaussian curve (solid line). (b) Calibration curve for the critical velocity vs the size of DNA in kbp plotted from the three histograms in (a). This calibration was used for the sizing of 285 kbp DNA molecules (see the text).

Figure 4a shows an experimental demonstration of the laser trapping of a single globular λ DNA molecule (48.5 kbp) in a circular motion. The single globular DNA molecule was trapped at the focal point of the laser in Figure 4b, and then scanning of the laser beam was accelerated along a circle 10 µm in radius in Figure 4c. The laser beam was circularly scanned at intervals of 1.0°. This step is equivalent to 175 nm, which was smaller than the diffraction-limited spot size of the laser we used (∼681 nm).

The trapped DNA molecule was released from the trapping point at a critical velocity of 0-900 µm/s when the trapping force and the hydrodynamic drag force for a single globular DNA molecule were balanced in Figure 4d. The released single globular DNA molecule was diffused by its Brownian motion and could be trapped again for subsequent manipulation, such as sizing again or transportation of it anywhere. With the use of the current circular acceleration technique, the change in the critical velocity with the size of the DNA was investigated. Figure 5a shows a histogram of the critical velocity of λ DNA (48.5 kbp), T4 DNA (166 kbp), and a fragment of the yeast chromosomal DNA (365 kbp). To obtain adequate statistics, each histogram represents 50 readings for individual molecules of each DNA sample. The three different DNA sizes were clearly resolved. Figure 5b shows a calibration for the critical velocity against the DNA size plotted using average values of the critical velocities from Figure 5a. This calibration is well fitted by a straight line with a slope of 0.336 µm/s · kbp and is useful for sizing globular single DNA molecules. With the use of this calibration, yeast chromosomal DNA (285 kbp) was sized from its critical velocity. The average critical velocity for seven single globular chromosomal DNA molecules was found to be 94.4 ± 13.4 µm/s. From the calibration in Figure 5b, this value corresponds to a DNA size of 281 ± 40 kbp. Thus, the size of the DNA molecule predicted from the calibration agreed well with its actual size. In its present state, our sizing achievements are undeniably less accurate than those by PFGE. However, the throughput of our technique is more rapid than that of PFGE, which takes more than 24 h in sizing DNA. Moreover, it is important to evaluate our technique as the one that permits both sizing single huge DNA molecules to use folded globular DNA and subsequently using the same target single DNA molecule after sizing for biochemical treatment, such as singlemolecule PCR. This single-molecule PCR combined with our current technique might contribute to the development of future single-molecule genome analysis. For such combination, we also investigated a possibility of reusing the globular DNA molecules condensed by PEG and Mg2+ for PCR. The efficiency of PCR was evaluated by analyzing amplified 3980 bp fragments. The composition of the globular DNA sample solution was the same as that of the sizing sample solution. Figure 6a shows the results of PCR amplification from native and two kinds of globular λ DNA templates without and with DAPI staining. The PCR efficiency was evaluated using a ratio of the fluorescence intensity of each PCR product from two kinds of globular DNA templates without and with DAPI staining to that of PCR product from native DNA template. From this measurement in Figure 6b, PCR efficiency was not reduced in two globular templates without and with DAPI staining and with PEG and Mg2+ treatment. From this analysis, globular DNA molecules look unfolded by the dilution of PEG and Mg2+ in PCR solution and undergo PCR amplification in a native DNA conformation. DNA conformation is reversibly switched between the random coiled state and globular state by changing the concentration of PEG and Mg2+ 14 and is controllable to transport from one region to another region with different concentrations.28 A single-molecule PCR from the single globular (28) Yoshikawa, Y.; Nomura, S. M.; Kanbe, T.; Yoshikawa, K. Chem. Phys. Lett. 2000, 330, 77–82.

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

5201

Furthermore, the relationship between critical velocity and huge DNA size was fitted with a linear line. Thus, this result suggests that the radius of a single globular huge DNA molecule is proportional to its DNA size although the radius of huge-sized globular DNA molecule is unknown. Our technique of sizing single globular huge DNA molecules not only provides a method of in situ sizing but also may contribute to a research technique of analyzing a huge-sized folded DNA molecule with the help of various condensation systems. CONCLUSIONS In conclusion, we propose a method for in situ sizing single huge DNA molecules (48.5-365 kbp) in a reusable manner by laser trapping in the globular state. The method provides a simple way for sizing single DNA molecules by measuring the critical velocity of circle motion of a trapping laser and does not require any prior knowledge of the radius of the single huge DNA molecule as a globular nanoparticle. We have demonstrated rapid, in situ sizing of single yeast chromosomal DNA molecules, which are usually analyzed by a PFGE technique more than 24 h. We also conclude that PCR efficiency is not affected by PEG and Mg2+. Thus, a single target globular DNA molecule after sizing can be reusable for subsequent biochemical treatment such as single-molecule PCR. The technique of laser trapping combined with condensation of a target DNA molecule into the globular state has potential for future single-genome analysis, which is now under development in our laboratory. Figure 6. (a) Effect of condensed λ DNA (48.5 kbp) templates condensed using PEG and Mg2+ on PCR amplification of λ DNA. Lane 1, native DNA template; lanes 2 and 3, globular DNA templates without and with DAPI staining, respectively; M, marker of 1 kbp ladder. (b) Ratio of the fluorescence intensity of 3980 bp product bands (lanes 1-3) to that of the native DNA template product band (lane 1) provides PCR efficiency.

DNA molecules after sizing and selecting by laser trapping looks possible using such a transportation technique to exchange a surrounding buffer solution on a microfluidic device.

5202

Analytical Chemistry, Vol. 80, No. 13, July 1, 2008

ACKNOWLEDGMENT We thank Professor A. Mizuno and Dr. Y. Matsuzawa, Toyohashi University of Technology, Aichi, Japan, for valuable discussions. This study was partly supported by a Grant-in-Aid (No. C17560241) for Scientific Research from the Japan Society for the Promotion of Science (JSPS). K.H. thanks PRESTO, Japan Science and Technology Agency (JST). Received for review February 20, 2008. Accepted April 25, 2008. AC8003538