DNA Length Evaluation Using Cyanine Dye and ... - ACS Publications

Various lengths of dsDNA fragments were prepared and mixed with YOYO prior to FCS, and the dependence of the diffusion time of a dsDNA−YOYO complex ...
0 downloads 0 Views 64KB Size
Download PDF
Biomacromolecules 2005, 6, 2703-2707

2703

DNA Length Evaluation Using Cyanine Dye and Fluorescence Correlation Spectroscopy Masafumi Shimizu, Satoshi Sasaki, and Makoto Tsuruoka* School of Bionics, Tokyo University of Technology, 1404-1 Katakura, Hachioji, Tokyo 192-0982, Japan Received March 9, 2005; Revised Manuscript Received May 11, 2005

To develop a high-performance method for measuring the length of double-stranded DNA (dsDNA) fragments, the capability of fluorescence correlation spectroscopy (FCS) was examined. To omit troublesome and timeconsuming labeling operations such as PCR with fluorescently labeled mononucleotides or primers, intercalation of dimeric cyanine dye YOYO-1 iodide (YOYO) to dsDNA was utilized as a simple labeling method. Various lengths of dsDNA fragments were prepared and mixed with YOYO prior to FCS, and the dependence of the diffusion time of a dsDNA-YOYO complex on the length of dsDNA fragment and the dsDNA/YOYO ratio was investigated. It was successfully demonstrated that the dsDNA length can be measured using YOYO and FCS, and the calibration curve was developed taking into account the rewinding and expansion of the dsDNA fragment caused by YOYO intercalation. Introduction At the forefront of molecular biology research, it is becoming increasingly important to be able to make highperformance measurements of the lengths of many DNA samples. Furthermore, the quantity of sample required should be as small as possible. Many samples of DNA are handled day after day, and the DNA size determination is the routine work for the molecular biologists. One of the most popular techniques for DNA length measurement is electrophoresis. However, this technique is not satisfactory for measuring hundreds of samples in an operation because of its run time and the quantity of sample required. In this paper, we focused on fluorescence correlation spectroscopy (FCS) as a method for measuring DNA length with high performance in samples with very small quantities of DNA. FCS is known to be a powerful tool for investigating the dynamic properties of fluorescently labeled molecules. In FCS, the fluorescence intensity of molecules, which fluctuates due to variations in the number density of the molecules in a confocally defined volume element, is recorded as photon counts and correlated in time.1-19 Several dynamic properties such as the number of fluorescent molecules in the volume element and the diffusion time of fluorescent molecules can be obtained. Furthermore, the fluorescent properties, such as photon count rate, photon count rate per fluorescent molecule, fractional population, and decay time of the triplet state, can be obtained simultaneously.4,13 There have been several reports that have studied the dynamic properties of DNA fragments using FCS;3,5-10 however, it has been necessary to introduce a fluorescent label to the DNA fragments using fluorescent probe hybridization5 or using PCR with fluorescently labeled mononucleotides6-8 or primers.3,9,10 These time-consuming methods for introducing fluorescent labels to DNA fragments * To whom correspondence should be addressed. Phone: +81-042637-2405; fax: +81-0426-37-2118; e-mail: [email protected].

are inadequate for measuring the lengths of DNA fragments with high performance. Here, we focused on the fact that cyanine dyes such as YO-PRO-1 iodide and YOYO-1 iodide (YOYO) have large molar absorptivities and are virtually nonfluorescent in the free form but show very strong fluorescence when complexed with double-stranded DNA (dsDNA).20-35 Many monomeric or dimeric cyanine dyes have been developed to detect nucleic acids,21,22 and they play very important roles in molecular biology,20,21,23 cytology,25,30 and so on.32,33,36 Because it is possible to label DNA fragments by simply mixing these dyes with DNA solutions, the need for troublesome and time-consuming labeling operations such as PCR can be completely omitted. In this paper, we investigate the potential of FCS as a method for measuring the length of a dsDNA fragment. For this purpose, various lengths of dsDNA fragments were prepared and mixed with the dimeric cyanine dye YOYO. We carried out the FCS analyses on these samples and investigated the dependence of the diffusion time of the dsDNA-YOYO complex on the length of the dsDNA fragment and the dsDNA/YOYO ratio. We also developed a calibration curve for different lengths of dsDNA fragments using FCS, taking into account the rewinding and expansion of dsDNA fragments caused by YOYO intercalation. Materials and Methods Materials. Dimeric cyanine dye YOYO-1 iodide (YOYO) was purchased from Molecular Probes. Sequence grade oligodeoxyribonucleotides were purchased from Date Concept, Japan. Unless otherwise mentioned, other chemicals were purchased from Wako Chemicals, Japan. All materials were of analytical grade and used without further purification. Preparation of dsDNA Fragments. The 20 and 50 bp dsDNA fragments were prepared by annealing. For the

10.1021/bm0501813 CCC: $30.25 © 2005 American Chemical Society Published on Web 07/06/2005

2704

Biomacromolecules, Vol. 6, No. 5, 2005

preparation of the 20 bp dsDNA fragment, an oligo (dGdC)10 fragment was used. For the 50 bp dsDNA fragment, an arbitrary sequence (5′-TCT ACT GGG ACG GAA CAG CTT TGA GGT GCG TGT TTG TGC CTG TCC TGG GA3′) and its complementary single-stranded DNA oligo were used. The synthesized single-stranded DNA oligos were dissolved and mixed stoichiometrically in a dilution buffer (10 mM Tris, 1 mM EDTA, 50 mM NaCl, pH 7.4). The solutions were heated to 94 °C and gradually cooled to 4 °C using a thermal cycler PTC-200 (MJ Research). These dsDNA solutions were used without further purification. For the dsDNA fragments longer than 100 bp, a Forever 100 bp Ladder Personalize Kit (Seegene, Korea) was used, which contains plasmids enabling us to amplify 100-1500 bp dsDNA fragments with PCR. After PCR using Ex Taq polymerase (TaKaRa, Japan), the dsDNA samples were purified with a QIAquick PCR Purification Kit (QIAGEN). The concentrations of the dsDNA samples were determined using a UV spectrometer (Ultrospec 2100 pro, Amersham Biosciences, Sweden). Preparation of dsDNA-YOYO Complex Solutions. The purchased YOYO solution, which is supplied as a 1 mM solution in DMSO, was diluted with DMSO (Nacalai Tesque, Japan) to prepare stock solutions of 20 µM YOYO. A working solution of 100 nM YOYO was freshly prepared immediately prior to use by dilution with DMSO. The solutions of dsDNA fragments were freshly diluted to desired concentrations prior to use with the aforementioned dilution buffer. Each diluted dsDNA solution was mixed with a 5 vol % 100 nM YOYO solution, followed by vigorous mixing with a vortex mixer. FCS Measurements. FCS measurements were performed on a commercial combination system (ConfoCor 2, Carl Zeiss, Germany) equipped with a water-immersion microscope objective (×40, 1.2 NA) and a 488 nm Ar laser for the excitation, of which the power was fixed at 150 µW. All the FCS measurements were carried out using 20 µL of a sample solution placed on a NEO microcover glass, 25 × 55 mm, No. 1 (Matsunami Glass, Japan). All spectroscopic measurements were carried out over a period of 20 s at room temperature (around 25 °C), and were taken 5 times. For calibration, an aqueous solution of 10 nM Rhodamine 6G (Aldrich) was studied as a control before each data acquisition series. Each set of correlation data obtained was fitted to the FCS autocorrelation function taking into account the triplet state buildup of the fluorophore4 and neglecting the sample volume distortions G(τ) ) 1 +

( ) [ () ]

τ -1 1 (1 - T + Te-τ/τT) 1 + × τD N(1 - T) w0 2 τ -1/2 1+ (1) z0 τ D

where τ is the correlation time, N is the average number of fluorescent molecules in the detection volume, T and τT are the fractional population and decay time of the triplet state, respectively, τD is the characteristic diffusion time, and w0 and z0 are the distances from the center of the laser beam

Shimizu et al.

Figure 1. Typical correlation curves of various lengths of the dsDNAYOYO complex. The base pair/YOYO ratio is 20 for all lengths of the dsDNA-YOYO complex. All the prepared DNA solutions were mixed with 5 nM YOYO.

focus in the radial and axial directions, respectively. τD corresponds to the time for a fluorescent molecule to diffuse across the confocal volume element and is related to the translational diffusion constant D of the fluorescent species by τD ) w02/4D. The diffusion time D is related to the frictional coefficient. Kovacic et al.37 have derived an expression representing the relationship between the frictional coefficient and the length of a dsDNA fragment. Furthermore, Bjo¨urling et al. have utilized this equation to express diffusion time.8 In this paper, we modified the expression to take into account the effect of the expansion caused by the YOYO intercalation, as follows: τD,L ∝

bsf bsf +γ ln d

(2)

where b is the number of base pairs, s is the distance between bases on a strand of DNA, f is an expansion factor, d is the diameter of hydrated dsDNA fragments, and γ is an endeffect correction.37 In this paper, we fixed s and γ at 0.34 nm and 0.39, respectively. For the diameter of the dsDNA fragment, we fixed d at 2.7 nm considering the effect of solvation. Here, we assume that neither d nor γ is influenced by the rewinding and expansion of dsDNA fragments caused by the YOYO intercalation. Results and Discussion Dependence of Diffusion Time of dsDNA-YOYO Complex on dsDNA Length and Base Pair/YOYO Ratio. The different lengths of dsDNA fragments were diluted to various concentrations, mixed with 5 nM YOYO, and subjected to FCS analysis. The normalized typical correlation functions obtained by the experiments are shown in Figure 1. Longer dsDNA-YOYO complexes exhibit a longer persistence of correlation, which means a longer diffusion time. The diffusion times are calculated by fitting the obtained correlation curves to eq 1, with which the diffusion time and decay time of the triplet state of fluorophore can be elucidated

DNA Length Evaluation Using Cyanine Dye and FCS

Figure 2. Relationships between the diffusion times of various lengths of dsDNA-YOYO complexes and the theoretical curves obtained using eq 1. The numbers in the legend represent the base pair/YOYO ratios. All diffusion time values were normalized to that of a 20 bp dsDNA-YOYO complex. For each plot, the prepared DNA solutions were mixed with 5 nM YOYO, and then FCS measurements were carried out, followed by fitting of the correlation data to eq 1. Solid line: theoretical curve calculated using eq 2 with the expansion factor d set to 1.3336 and dashed line: without any expansion consideration. For the derivation of both theoretical curves, the distance between bases on a strand of DNA s, the diameter of hydrated DNA fragment d, and the end-effect γ were fixed at 0.34, 2.7 nm, and 0.39,37 respectively.

separately. To elucidate the relationships between the diffusion time and both the length of dsDNA fragment and the base pair/YOYO ratio, the diffusion times of dsDNAYOYO complexes, normalized to those of a 20 bp dsDNA fragment, are plotted against the length of dsDNA in Figure 2. The diffusion time of dsDNA-YOYO complexes correlates positively with the length of dsDNA fragment and negatively with the base pair/YOYO ratio at a constant length of dsDNA. We think that the diffusion time of dsDNA-YOYO complexes is dependent on the base pair/YOYO ratio due to the rewinding and expansion effects caused by the intercalation of the chromophores. It has already been observed with fluorescence microscopy that a lambda DNA molecule 16.2 µm in length extends to 21.5 µm on intercalation with YOYO,36 which corresponds to an expansion factor of 1.33. To discuss the dependence of diffusion time on dsDNA length and the base pair/YOYO ratio, two types of theoretical diffusion time profiles were calculated using eq 2, and these are shown in Figure 2. One of the theoretical curves was derived with the assumption that a dsDNA molecule extends by a factor of 1.33 (solid line), and the other was derived without assuming any dsDNA extension (dashed line). When the base pair/YOYO ratio is lower than 20, the plots show good agreement with the theoretical curve that assumes expansion. On the other hand, when the base pair/YOYO ratio is as high as 200, the plots are in accordance with the theoretical curve that does not assume any expansion. However, in both cases, the plots for dsDNA longer than 800 bp are irregular because there were so few fluorescent molecules in the confocally defined volume element that it was difficult to obtain a low-noise correlation curve. These results suggest that it is possible to accurately determine the dsDNA length in the 20-600 bp range, even when the dsDNA concentration is as low as 1 nM base pairs,

Biomacromolecules, Vol. 6, No. 5, 2005 2705

which corresponds to 0.66 pg/µL, if the dsDNA concentration is controlled. Because the confocal volume element for FCS measurement is as small as 0.1 fL, the sample solution on the order of 1 fL, which corresponds to only 1 zeptogram (1 × 10-21 g) of dsDNA, should be enough for the measurement. It is noteworthy that this ultrasensitive measurement requires only the mixing of dsDNA and YOYO solutions, which enables us to fully omit troublesome amplification operations such as PCR using fluorescently labeled mononucleotides6-8 or primers.3,9,10 Furthermore, the required sample concentration can be reduced by optimization of the measurement conditions, such as the duration time and repetition of the measurement. On the other hand, the diffusion time coincides with the theoretical curve derived without the assumption of any expansion only for dsDNA fragments shorter than 100 bp, when the base pair/YOYO ratio is higher than 2000. This result suggests that if the base pair/YOYO ratio is high enough, the dsDNA-YOYO complex should contain only one YOYO molecule, and the dsDNA with a single YOYO molecule behaves as a rodlike molecule only if the dsDNA is shorter than 100 bp. This is consistent with the report that the persistence length of dsDNA is around 450-500 Å, which corresponds to 130-150 bp, under moderate salt conditions.38 Incidentally, the diffusion time profile of the dsDNA fragment labeled with fluorescent mononucleotides agrees with the theoretical diffusion time profile, even when the dsDNA fragment is as long as 500 bp.8 We think that this inconsistency may occur because the dsDNA fragments with fluorescently labeled mononucleotides should be considered to be thicker than intact dsDNA fragments due to the size of the fluorescently labeled moiety and/or because the fluorescently labeled moiety hinders the internal motion of dsDNA. Our new method for DNA length evaluation has been only applied for the sample containing a single length of the DNA fragment. We consider that it is important to improve the applicability of this technique for the sample containing multiple lengths of dsDNA fragments in the future work, utilizing the component analysis method.8 Fluorescence of dsDNA-YOYO and Its Relationship to the Base Pair/YOYO Ratio. It was important to verify that the fluorescence intensity is not affected by differences in the lengths of the dsDNA fragments. Therefore, the photon count rate, which corresponds to the fluorescence intensity, was plotted against the base pair/YOYO ratio, and this is shown in Figure 3. For all lengths of dsDNA fragments, the photon count rate increased with an increasing base pair/ YOYO ratio. When FCS was carried out on dsDNA fragments without YOYO, only a background level of the photon count rate was observed (data not shown). These results are consistent with the report that YOYO exhibits high fluorescence when it binds to dsDNA but shows little fluorescence when it is free.20-35 This result confirms that the fluorescence intensity is affected not by differences in the dsDNA length but by the base pair/YOYO ratio. In contrast, the photon count per fluorescent molecule was greatly affected by not only the length of dsDNA fragment but also by the base pair/YOYO ratio.39 The photon count

2706

Biomacromolecules, Vol. 6, No. 5, 2005

Shimizu et al.

YOYO molecules. Incidentally, the base pair/YOYO ratio of the peak region is consistent with that YOYO molecules intercalate with 165 kbp dsDNA up to a base pair/YOYO ratio of 8,26 although the lengths of dsDNA fragments differ by a factor of over 100. Conclusion

Figure 3. Photon count rates of various lengths of the dsDNA-YOYO complex with respect to the base pair/YOYO ratio. All the prepared DNA solutions were mixed with 5 nM YOYO.

Various lengths of dsDNA fragments were prepared and mixed with the dimeric cyanine dye YOYO. FCS analysis was carried out on these samples. Several properties, such as the dependence of the diffusion time, the photon count rate, and the photon count rate per fluorescent molecule on the dsDNA/YOYO ratio, were investigated. It was successfully demonstrated that the dsDNA length can be measured using cyanine dye and FCS, and the calibration curve was developed taking into account the rewinding and expansion of the dsDNA fragments caused by YOYO intercalation. These results open the way for the high-performance measurement of dsDNA lengths using small sample quantities, without requiring time-consuming and troublesome amplification operations such as PCR using fluorescently labeled mononucleotides or primers. References and Notes

Figure 4. Photon count rate per fluorescent molecule for various lengths of the dsDNA-YOYO complex with respect to the base pair/ YOYO ratio. All prepared DNA solutions were mixed with 5 nM YOYO, FCS measurements were carried out, and the correlation data were fitted to eq 1.

rate values per fluorescent molecule are shown in Figure 4. These plots exhibit peaks at base pair/YOYO ratios between 5 and 10 and plateaus of approximately 10 kHz at base pair/ YOYO ratios higher than 2000 for all lengths of dsDNA fragments. This indicates that the number of YOYO molecules held in a dsDNA-YOYO complex changes depending on the base pair/YOYO ratio. The range of base pair/YOYO ratios in the plateau region is in agreement with that indicated by the previously discussed dependence of the diffusion time on this ratio (Figure 2). Although the dependence at the higher base pair/YOYO ratio is understandable, the decrease in fluorescence at the lower base pair/YOYO ratio seems strange. The YOYO dye might be undergoing self-quenching at the lower base pair/YOYO ratio. Larsson and co-workers reported that a second nonintercalative binding mode appears at base pair/YOYO ratio below 4,26 similar to the point at which the fluorescence begins to decrease in Figure 4. Thus, as the DNA is saturated by the bisintercalating dye, excess dye can bind through electrostatics or in one of the grooves. The close proximity of the nonintercalated dye to the intercalated dye could cause quenching of the latter, resulting in the observed trend. The profile of the 50 bp dsDNA fragment shows a very weak peak because the dsDNA fragment is too short to bind with a multiple number of

(1) Rigler, R.; Widengren, J.; Mets, U ¨ . Interactions and Kinetics of Single Molecules as obserVed by Fluorescence Correlation Spectroscopy; Fluorescence Spectroscopy. New methods and applications; Springer: Berlin, 1992. (2) Rigler, R.; Mets, U ¨ .; Widengren, J.; Kask, P. Eur. Biophys. J. 1993, 22, 169-175. (3) Kinjo, M.; Rigler, R. Nucleic Acids Res. 1995, 23, 1795-1799. (4) Widengren, J.; Mets, U ¨ .; Rigler, R. J. Phys. Chem. 1995, 99, 1336813379. (5) Castro, A.; Williams, J. G. K. Anal. Chem. 1997, 69, 3915-3920. (6) Kinjo, M.; Nishimura, G. Bioimaging 1997, 5, 134-138. (7) Nishimura, G.; Rigler, R.; Kinjo, M. Bioimaging 1997, 5, 129-133. (8) Bjo¨rling, S.; Kinjo, M.; Fo¨ldes-Papp, Z.; Hagman, E.; Thyberg, P.; Rigler, R. Biochemistry (Moscow) 1998, 37, 12971-12978. (9) Kettling, U.; Koltermann, A.; Schwille, P.; Eigen, M. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 1416-1420. (10) Rigler, R.; Fo¨ldes-Papp, Z.; Meyer-Alme, F. J.; Sammet, C.; Vo¨lcker, M.; Schnetz, A. J. Biotechnol. 1998, 63, 97-109. (11) Orden, A. V.; Keller, R. A. Anal. Chem. 1998, 70, 4463-4471. (12) Schwille, P.; Haupts, U.; Maiti, S.; Webb, W. W. Biophys. J. 1999, 77, 2251-2265. (13) Tessier, A.; Me´allet-Renault, R.; Denjean, P.; Miller, D.; Pansu, R. B. Phys. Chem. Chem. Phys. 1999, 1, 5767-5769. (14) Bacia, K.; Majoul, I. V.; Schwille, P. Biophys. J. 2002, 83, 11841193. (15) Doi, N.; Takashima, H.; Kinjo, M.; Sakata, K.; Kawahashi, Y.; Oishi, Y.; Oyama, R.; Miyamoto-Sato, E.; Sawasaki, T.; Endo, Y.; Yanagawa, H. Genome Res. 2002, 12, 487-492. (16) LeCaptain, D. J.; Orden, A. V. Anal. Chem. 2002, 74, 1171-1176. (17) Tsuchiya, N.; Ohashi, J.; Tokunaga, K. Immunol. ReV. 2002, 190, 169-181. (18) Hori, K.; Shin, W. S.; Hemmi, C.; Toyo-oka, T.; Makino, T. Curr. Pharm. Biotechnol. 2003, 4, 477-484. (19) Bannai, M.; Higuchi, K.; Akesaka, T.; Furukawa, M.; Yamaoka, M.; Sato, K.; Tokunaga, K. Anal. Biochem. 2004, 327, 215-221. (20) Glazer, A. N.; Rye, H. S. Nature 1992, 359, 859-861. (21) Rye, H. S.; Yue, S.; Wemmer, D. E.; Quesada, M. A.; Haugland, R. P.; Mathies, R. A.; Glazer, A. N. Nucleic Acids Res. 1992, 20, 28032812. (22) Rye, H. S.; Dabora, J. M.; Quesada, M. A.; Mathies, R. A.; Glazer, A. N. Anal. Biochem. 1993, 208, 144-150. (23) Rye, H. S.; Drees, B. L.; Nelson, H. C.; Glazer, A. N. J. Biol. Chem. 1993, 268, 25229-25238. (24) Carlsson, C.; Larsson, A.; Jonsson, M.; Albinsson, B.; Norde´n, B. J. Phys. Chem. 1994, 98, 10313-10321.

DNA Length Evaluation Using Cyanine Dye and FCS (25) Hirons, G. T.; Fawcett, J. J.; Crissman, H. A. Cytometry 1994, 15, 129-140. (26) Larsson, A.; Carlsson, C.; Jonsson, M.; Albinsson, B. J. Am. Chem. Soc. 1994, 116, 8459-8465. (27) Larsson, A.; Carlsson, C.; Jonsson, M. Biopolymers 1995, 36, 153167. (28) Netzel, T. L.; Nafisi, K.; Zhao, M.; Lenhard, J.; Johnson, I. J. Phys. Chem. 1995, 99, 17936-17947. (29) Rye, H. S.; Glazer, A. N. Nucleic Acids Res. 1995, 23, 1215-1222. (30) Guindulain, T.; Comas, J.; Vives-Rego, J. Appl. EnViron. Microbiol. 1997, 63, 4608-4611. (31) Petty, J. T.; Bordelon, J. A.; Robertson, M. E. J. Phys. Chem. B 2000, 104, 7221-7227. (32) Namasivayam, V.; Larson, R. G.; Burke, D. T.; Burns, M. A. Anal. Chem. 2002, 74, 3378-3385. (33) Ferree, S.; Blanch, H. W. Biophys. J. 2003, 85, 2539-2546.

Biomacromolecules, Vol. 6, No. 5, 2005 2707 (34) Sovenyhazy, K. M.; Bordelon, J. A.; Petty, J. T. Nucleic Acids Res. 2003, 31, 2561-2569. (35) Ghasemi, J.; Niazi, A.; Westman, G.; Kubista, M. Talanta 2004, 62, 835-841. (36) Bensimon, A.; Simon, A.; Chiffaudel, A.; Croquette, V.; Heslot, F.; Bensimon, D. Science 1994, 265, 2096-2098. (37) Kovacic, R. T.; van Holde, K. E. Biochemistry (Moscow) 1977, 16, 1490-1498. (38) Hagerman, P. J. Annu. ReV. Biophys. Biophys. Chem. 1988, 17, 265286. (39) Kii, H.; Tamura, M.; Kinjo, M. Analysis of DNA nanostructures by Fluorescence Correlation Spectroscopy; 42nd Annual Meeting of The Biophysical Society of Japan, 2004; Vol. 42, poster 2P108.

BM0501813