Desorption Behaviors

Jan 30, 2009 - spirillum magneticum AMB-1, were used as solid supports for silanization as previously described.26 Electron microscopy has revealed th...
3 downloads 0 Views 242KB Size
Download PDF
2956

Langmuir 2009, 25, 2956-2961

Contributions of Phosphate to DNA Adsorption/Desorption Behaviors on Aminosilane-Modified Magnetic Nanoparticles Tsuyoshi Tanaka, Ririko Sakai, Ryosuke Kobayashi, Keiichi Hatakeyama, and Tadashi Matsunaga* Department of Biotechnology, Tokyo UniVersity of Agriculture and Technology, 2-24-16, Naka-cho, Koganei, Tokyo 184-8588, Japan ReceiVed March 10, 2008. ReVised Manuscript ReceiVed NoVember 23, 2008 The adsorption and desorption behaviors of DNA on aminosilane-modified magnetic nanoparticles were investigated by altering both type of anions and solvation state to achieve eficient recovery of DNA useful for subsequent polymerase chain reaction (PCR) analysis. The effects of multiple anions in accordance with the Hofmeister ion series were determined to clarify the contribution of phosphate ions on the effective desorption of DNA from aminosilane surfaces. Efficient DNA desorption (85% recovery) occurred in the presence of 1 M phosphate buffer, however, little DNA desorption was observed using any other anions. This phenomenon indicates that desorption originated from the replacement of DNA by phosphate ions. Furthermore, the adsorption and desorption were significantly affected by the addition of both protic and aprotic solvents. Efficient recovery of adsorbed DNA was attained using deoxynucleotide triphosphates (dNTPs) in place of phosphate buffer and was suitable for subsequent PCR analyses. Therefore, the DNA adsorption/desorption process proposed in this study will be a promising, novel approach for DNA purification with a high recovery ratio that is suitable for subsequent enzymatic reactions, such as PCR or restriction enzyme digestion.

Introduction Solid-phase DNA purification has been commonly used for molecular biology due to its quick processing time, reduced chemical requirements, and ease of implementation. Silica-based DNA purification was first established in 1979.1 The addition of chaotropic ions, which destroy the structure of water molecules that surround DNA,2 especially guanidine thiocyanate (a strong chaotrope in the Hofmeister series), promotes the binding of DNA to silica.3 Recently, interaction between DNA and chaotropes (e.g., guanidinium chloride, urea) has been precisely analyzed by the differential scanning calorimetry,4 suggesting the binding of guanidinium ions in the minor groove of DNA, and the binding to amide-like groups of single stranded (or melted) DNA. DNA is easily released from silica under low ionic conditions. Solid-phase reversible immobilization (SPRI) using carboxyl group-coated solid supports has also been proposed.5-7 PEG-DNA precipitates under high NaCl conditions reversibly bind to these supports. These approaches circumvent the use of organic solvents that may influence subsequent applications, such as gene amplification and enzymatic digestion of DNA, especially, in microfluidic device. Recently, various microfabrications, such as silica pillar,8 silica monolith9-12 or silica (magnetic) beads,13,14 * Corresponding author. Telephone: +81-42-388-7020. Fax: +81-42385-7713. E-mail: [email protected]. (1) Vogelstein, B.; Gillespie, D. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 615– 619. (2) Cacace, M. G.; Landau, E. M.; Ramsden, J. J. Q. ReV. Biophys. 1997, 30, 241–277. (3) Boom, R.; Sol, C. J.; Salimans, M. M.; Jansen, C. L.; Wertheim-van Dillen, P. M.; van der Noordaa, J. J. Clin. Microbiol. 1990, 28, 495–503. (4) Tabuashvili, E.; Getashvili, G.; Makharadze, M.; Tsulukidze, L.; Sujashvili, R.; Khoshtariya, D. E. J. Biol. Phys. Chem. 2006, 6, 25–29. (5) Lis, J. T. Methods Enzymol. 1980, 65, 347–353. (6) Hawkins, T. L.; O’Connor-Morin, T.; Roy, A.; Santillan, C. Nucleic Acids Res. 1994, 22, 4543–4544. (7) Krizova, J.; Spanova, A.; Rittich, B.; Horak, D. J. Chromatogr. A 2005, 1064, 247–253. (8) Christel, L. A.; Petersen, K.; McMillan, W.; Northrup, M. A. J. Biomech. Eng. 1999, 121, 22–27. (9) Wen, J.; Guillo, C.; Ferrance, J. P.; Landers, J. P. Anal. Chem. 2007, 79, 6135–6142.

have been developed in order to enhance the efficiency of DNA trapping on a chip. These researches have attained to minimize the effect of organic solvents in microfludic device. An alternative approach to overcome the above problem is organic solvent-free DNA extraction. Our research group has demonstrated a novel DNA recovery method using aminosilane-modified solid supports15-18 that is based on electrostatic interactions between the amino groups on the solid supports and nucleic acids, as well as subsequent DNA release with changing ionic strength. A similar platform was achieved using the amino groups on a chitosan polymer.19,20 On the other hand, amine-coated solid supports have been widely used as substrates for DNA microarrays since 1995.21 Aminosilane-modified microarray glass slides are more popular than poly-L-lysine-coated slides because of the relatively high stability of aminosilane.22 At present, limited knowledge of DNA adsorption and desorption behaviors on aminosilane-modified (10) Wu, Q.; Bienvenue, J. M.; Hassan, B. J.; Kwok, Y. C.; Giordano, B. C.; Norris, P. M.; Landers, J. P.; Ferrance, J. P. Anal. Chem. 2006, 78, 5704–5710. (11) Legendre, L. A.; Bienvenue, J. M.; Roper, M. G.; Ferrance, J. P.; Landers, J. P. Anal. Chem. 2006, 78, 1444–1451. (12) Wolfe, K. A.; Breadmore, M. C.; Ferrance, J. P.; Power, M. E.; Conroy, J. F.; Norris, P. M.; Landers, J. P. Electrophoresis 2002, 23, 727–733. (13) Breadmore, M. C.; Wolfe, K. A.; Arcibal, I. G.; Leung, W. K.; Dickson, D.; Giordano, B. C.; Power, M. E.; Ferrance, J. P.; Feldman, S. H.; Norris, P. M.; Landers, J. P. Anal. Chem. 2003, 75, 1880–1886. (14) Hong, J. W.; Studer, V.; Hang, G.; Anderson, W. F.; Quake, S. R. Nat. Biotechnol. 2004, 22, 435–439. (15) Yoza, B.; Arakaki, A.; Matsunaga, T. J. Biotechnol. 2003, 101, 219–228. (16) Yoza, B.; Matsumoto, M.; Matsunaga, T. J. Biotechnol. 2002, 94, 217– 224. (17) Nakagawa, T.; Tanaka, T.; Niwa, D.; Osaka, T.; Takeyama, H.; Matsunaga, T. J. Biotechnol. 2005, 116, 105–111. (18) Yoza, B.; Arakaki, A.; Maruyama, K.; Takeyama, H.; Matsunaga, T. J. Biosci. Bioeng. 2003, 95, 21–26. (19) Cao, W.; Easley, C. J.; Ferrance, J. P.; Landers, J. P. Anal. Chem. 2006, 78, 7222–7228. (20) Hourfar, M. K.; Michelsen, U.; Schmidt, M.; Berger, A.; Seifried, E.; Roth, W. K. Clin. Chem. 2005, 51, 1217–1222. (21) Schena, M.; Shalon, D.; Davis, R. W.; Brown, P. O. Science 1995, 270, 467–470. (22) Chiu, S. K.; Hsu, M.; Ku, W. C.; Tu, C. Y.; Tseng, Y. T.; Lau, W. K.; Yan, R. Y.; Ma, J. T.; Tzeng, C. M. Biochem. J. 2003, 374, 625–632.

10.1021/la8032397 CCC: $40.75  2009 American Chemical Society Published on Web 01/30/2009

DNA Adsorption/Desorption

substrates has been obtained.23 The main driving forces of oligodeoxyribonucleotide (ODN) adsorption onto aminopropylsilane-modified substrates have been explained by a Langmuir reversible adsorption model.24,25 The equilibrium constant K (ml/mol), the ratio between the adsorption rate constant (kads: 0.56 × 10-6 cm/s) and the desorption rate constant (kdes: 0.69 × 10-15 mol/cm2/s)24, had a very high value (0.846 × 109 cm3/ mol), indicating very rapid adsorption (saturated within 60 min) and very slow desorption of ODN. Therefore, aminosilanemodified substrates are preferable for DNA immobilization but not for DNA extraction. Recently, suitable conditions for DNA desorption from aminosilane-modified solid supports were discovered by our research group. Double-stranded DNA (dsDNA; λDNA) adsorbed on 3-[2-(2-aminoethylamino)ethylamino]propyltrimethoxysilane(AEEA-) modified magnetic nanoparticles was effectively released by the addition of phosphate ion (53% release of captured DNA), but not by a solution of 2 M NaCl (5% release of captured DNA).26 Furthermore, desorption was performed in 20 min. Possible explanations for this rapid desorption are the replacement of phosphate ions with adsorbed DNA on the aminosilanemodified magnetic nanoparticles or the effects of multiple anions (i.e., PO42- ions). If multiple anions contribute to the desorption, another divalent anions (e.g., sulfate ion) would be useful for the DNA desorption. These DNA desorption behaviors should be followed by the Hofmeister ion series, which is an opposite phenomenon of chaotropy (i.e., kosmotropy)2 and shows the graduated effects on the structuring of macromolecules. Even though the use of phosphate ions for DNA desorption is a promising approach for DNA purification, investigations toward a systematic understanding of the efficient recovery of DNA from aminosilane-modified surfaces have not been performed. In this study, the adsorption and desorption behaviors of λDNA on aminosilane-modified magnetic nanoparticles was investigated by altering both type of anion and solvation state to achieve the efficient recovery of DNA. The effect of multiple anions in accordance with the Hofmeister ion series was determined to clarify the contribution of phosphate ions to the effective desorption of DNA from aminosilane-modified magnetic nanoparticles. These analyses provide new insights into the mechanism of DNA recovery using aminosilane-modified solid supports. On the basis of these results, gene amplification by PCR using desorbed DNA from aminosilane-modified magnetic nanoparticles was demonstrated.

Experimental Section Materials. 3-[2-(2-aminoethylamino)ethylamino]propyltrimethoxysilane (AEEA) was obtained from Fluka Chemical (Buchs SG, Switzerland). PicoGreen used for DNA quantification was purchased from Molecular Probes (Eugene, OR). λDNA (48 502 bp) was purchased from Takara Bio (Shiga, Japan). N,N-Dimethylformamide (DMF, dehydrated) was obtained from Wako Pure Chemical Industries (Osaka, Japan). Sulfosuccinimidyl 6-[3′-(2pyridyldithio)-propionamido] hexanoate (Sulfo-LC-SPDP) was purchased from Pierce Chemical (Rockford, IL). 2′-Deoxyadenosine 5′-triphosphate (dATP) was purchased from Invitrogen (Carlsbad, CA). Adenosine was obtained from Sigma-Aldrich (St. Louis, MO). Other commercially available reagents were either of analytical or laboratory grade. Magnetic nanoparticles (magnetite; average 80 (23) Fang, Y.; Hoh, J. H. Nucleic Acids Res. 1998, 26, 588–593. (24) Balladur, V. V.; Theretz, A.; Mandrand, B. J. Colloid Interface Sci. 1997, 194, 408–418. (25) Chan, V.; Graves, D. J.; Fortina, P.; McKenzie, S. E. Langmuir 1997, 13, 320–329. (26) Nakagawa, T.; Hashimoto, R.; Maruyama, K.; Tanaka, T.; Takeyama, H.; Matsunaga, T. Biotechnol. Bioeng. 2006, 94, 862–868.

Langmuir, Vol. 25, No. 5, 2009 2957 nm in diameter) purified from magnetotactic bacterium, Magnetospirillum magneticum AMB-1, were used as solid supports for silanization as previously described.26 Electron microscopy has revealed that the magnetic nanoparticle consists of a single crystal of magnetite (Fe3O4) with a single-domain magnetic structure. The magnetite crystals were 50-100 nm in size and exhibited cuboctahedral morphology. Preparation of Aminosilane-Modified Magnetic Nanoparticles. Magnetic nanoparticles (10 mg) were washed with distilled water and then treated with H2O:NH3:H2O2 (5:1:1) at room temperature for 10 min to activate surface hydroxyl groups. After washing with methanol (MeOH), the magnetic nanoparticles were incubated with 20 mL of 2% AEEA solution in ethanol (EtOH) at room temperature for 10 min. After washing with DMF, aminosilane-modified magnetic nanoparticles were baked at 120 °C in 20 mL of DMF for 30 min with sonication every 10 min. After this treatment, the particles were washed with MeOH three times and stored in MeOH at 4 °C. The concentration of magnetic nanoparticles was estimated from absorbance at 660 nm. The relationship between absorbance and dry weight of magnetic nanoparticle was estimated by a standard curve: 1.0 × Abs660 corresponded to 213 µg of magnetic nanoparticles/ml. To determine the number of amine moieties on their surfaces, aminosilane-modified magnetic nanoparticles (250 µg) were incubated in 10 mM Sulfo-LC-SPDP in 8.1 mM phosphate buffer containing 137 mM NaCl (phosphate buffered saline; PBS) at room temperature for 30 min. Sulfo-LC-SPDP-conjugated magnetic nanoparticles were washed three times with PBS and incubated in 200 µL of 20 mM dithiothreitol in PBS to release 2-pyridylthiol. The absorbance of 2-pyridylthiol at 343 nm was measured in a spectrophotometer. The concentrations of Sulfo-LC-SPDP reacted with primary amine groups on nanoparticles were determined using a standard curve generated with different concentrations of SulfoLC-SPDP in 20 mM dithiothreitol. The ξ potential for aminosilanemodified magnetic nanoparticles was calculated from their electrophoretic mobilities that were determined using a laser particle analyzer (ELS-8000, Otsuka Electronics, Osaka, Japan). Adsorption of λDNA onto AEEA-Magnetic Nanoparticles. Aminosilane-modified magnetic nanoparticles (50 µg) were mixed with appropriate amounts of λDNA in 40 µL of buffer and incubated at room temperature. The DNA-magnetic nanoparticle complexes were collected magnetically and washed three times with the same buffer. The adsorbed DNA was quantified as the total added DNA amount minus the amount of the initial supernatant fraction. To investigate the effect of pH, a mixture of 25 mM MES buffer (pH 5 or 6), 25 mM Tris-HCl buffer (pH 7 or 8), and 25 mM carbonate buffer (pH 9 or 10) was used for DNA adsorption experiments. Additionally, 5% of MeOH, EtOH, acetonitrile, or DMF was added to the DNA and aminosilane-modified magnetic nanoparticle suspensions to evaluate solvation effects. The amount of DNA was determined using the DNA intercalator, PicoGreen (excitation at 502 nm, emission at 523 nm). The fluorescent intensity was measured using a fluorescence microplate reader (FLUO Star Galaxy, BMG Labtechnologies Inc., Offernburg, Germany). Calibration curves for DNA quantification were prepared using λDNA as a standard in appropriate buffers. Desorption of λDNA from Aminosilane-Modified Magnetic Nanoparticles. To investigate DNA desorption behaviors, λDNAmagnetic nanoparticle complexes prepared under the conditions described above were resuspended in Tris-HCl buffer (pH 8.0) containing 1 M sodium salts of citric acid, tartaric acid, sulfuric acid, and nitric acid, sodium chloride, or 1 M phosphate buffer (pH 8.0) for 40 min at 80 °C. The effects of increasing MeOH concentrations in aqueous solutions as well as increasing dNTP concentrations on DNA desorption were also examined. dNTP was dissolved in Tris-HCl buffer at pH 8.0. After magnetic separation, the supernatants were collected as for the DNA-desorbed fractions. The amount of DNA was determined in the same manner described above. Gene Amplification Using DNA Released from AminosilaneModified Magnetic Nanoparticles. Aminosilane-modified magnetic nanoparticles (50 µg) were mixed with 1 µg genomic DNA in 40

2958 Langmuir, Vol. 25, No. 5, 2009

Tanaka et al.

Table 1. Effects of Protic and Aprotic Solventsa on DNA Adsorption added solvents

captured DNA (ng/50 µg-particles)

MeOH EtOH DMF acetonitrile no addition

129 ( 45 157 ( 1 207 ( 40 235 ( 20 68 ( 23

a 5% of organic solvents were added in Tris-HCl at pH8.0. Data were means of triplicate experiments ( SD.

µL of 25 mM Tris-HCl buffer (pH 8.0) containing 5% MeOH and incubated at room temperature. The genomic DNA was extracted from Escherichia coli K-12. The DNA-magnetic nanoparticle complexes were collected magnetically and washed three times with Tris-HCl buffer and incubated in a 40-µL dNTP (2.5-100 mM) solution for 40 min at 80 °C to release the DNA. The genomic DNA desorbed from the aminosilane-modified magnetic nanoparticles was used as a template for PCR. The 184-bp fragment of the gyrB gene was amplified using the following primers: forward, 5′-TGCGTGAGTTGTCGTTCCTC-3′, and reverse, 5′-ACGCCAATACCGTCTTTTTCA-3′. PCR amplification was performed with the following protocol: denaturation (94 °C for 5 min), 35 cycles at 94 °C for 20 s, 55 °C for 30 s, and 72 °C for 40 s, and a final extension at 72 °C for 3 min. The reaction mixture contained 1 µL DNA released from the aminosilane-modified magnetic nanoparticles, 0.75 U AmpliTaq DNA polymerase LD (Applied Biosystems, Foster City, CA), 100 nM primers, and 7.5 mM MgCl2 in a final volume of 25 µL. As a control, gene amplification using DNA released by 1 M phosphate was performed in the above reaction mixture containing 250 µM dNTPs. Amplified PCR products were analyzed by 2% (w/v) agarose gel electrophoresis and stained with ethidium bromide.

Results Characterization of Aminosilane-Modified Magnetic Nanoparticles. The number of amine groups on the magnetic nanoparticles was estimated at 2.7 × 104 ((0.5 × 104) amine groups/particle or 1.1 ( 0.2 groups/nm2 (n ) 3). The surface area of a single magnetic particle, assuming a 50 × 50 × 100nm3 rectangular box model, was 2.5 × 104 nm2. The amine density was similar to our previous report of 2.1 × 104 amine groups/particle,15 although a slight modification was made in the silanization process in this study (e.g., baking in 20 mL DMF at 120 °C for 30 min). On the other hand, the -OH saturation coverage of Fe3O4(111) was estimated at 2.0 groups/nm2 by thermal desorption spectroscopy as described previously.27 If silicon triethoxides in AEEA are completely reacted with -OH groups on the Fe3O4 surface, the maximum amine coverage value is assumed as the same, i.e. 2.0 groups/nm2. Because the experimental amine coverage value was 1.1 amino groups/nm2, the coverage of the amino groups per magnetic nanoparticle was expected to be approximately 55% of the maximum amine coverage value. DNA Adsorption Behavior on Aminosilane-Modified Magnetic Nanoparticles. The effects of solvation on DNA adsorption onto aminosilane-modified magnetic nanoparticles was investigated using protic (MeOH or EtOH) and aprotic (acetonitrile or DMF) solvents. Because the aminosilane-modified magnetic nanoparticles were resuspended in 100% MeOH, MeOH was replaced by Tris-HCl buffer prior to the DNA adsorption experiments. DNA adsorption onto aminosilane-modified magnetic nanoparticles was enhanced when organic solvents were added to the aqueous solutions (Table 1). (27) Joseph, Y.; Kuhrs, C.; Ranke, W.; M, R.; Weiss, W. Chem. Phys. Lett. 1999, 314, 195–202.

Figure 1. Effects of binding buffer pH on λDNA adsorbed (closed circle) and desorbed (open circle) on aminosilane-modified magnetic nanoparticles (50 µg/40 µL). Binding buffers: 25 mM MES buffer (pH 5 or 6), 25 mM Tris-HCl buffer (pH 7 or 8), 25 mM carbonate buffer (pH 9 or 10). Data were means of triplicate experiments ( SD.

Figure 2. Amount of λDNA adsorbed on aminosilane-modified magnetic nanoparticles (50 µg/40 µL) as a function of DNA concentration at 25 °C. λDNA solution was diluted to the appropriate concentration with 25 mM Tris-HCl buffer (pH 8.0). MeOH concentration: 5%. Data were means of triplicate experiments ( SD.

Greater adsorption in aprotic solvents was observed compared with protic solvents. To avoid the use of highly toxic reagents, 5% MeOH was added to the aqueous solutions in the subsequent experiments for DNA adsorption. Effects of pH on DNA adsorption onto aminosilane-modified magnetic nanoparticles were observed between pHs 5 and 10 (Figure 1, closed circle). Efficient DNA adsorption occurred at pH 8.0, while a remarkable decrease in DNA adsorption occurred at pHs greater than 9.0. Adsorption at pH 10.0 decreased to approximately 10% of the maximum adsorption. Amino groups present at pH 9.0 are expected to be deprotonated because the pKa of aminosilane is estimated at 9-10.28,29 In contrast, the protonation of amino groups on aminosilane-modified magnetic nanoparticles is expected to occur at pHs less than 8. These results indicate that DNA adsorption is correlated with surface positive charges. The DNA-binding capacity of aminosilane-modified magnetic nanoparticles was examined using λDNA at different concentrations and reaction times. Previous work indicated that the DNA capturing amount was saturated at 600 ng when the amine number on magnetic nanoparticles was 1.1 × 105 amino groups/particle26 (amount of magnetic nanoparticles: 10 µg). In contrast, the DNA adsorption was saturated at approximately 110 ng (Figure 2) when the amine n umber was 2.7 × 104 amino groups/particle (amount of magnetic nanoparticles: 50 µg). These results suggest (28) Pan, B. F.; Gao, F.; Gu, H. C. J. Colloid Interface Sci. 2005, 284, 1–6. (29) Metwalli, E.; Haines, D.; Becker, O.; Conzone, S.; Pantano, C. G. J. Colloid Interface Sci. 2006, 298, 825–831.

DNA Adsorption/Desorption

Figure 3. Amount of λDNA adsorbed on aminosilane-modified magnetic nanoparticles (50 µg/40 µL) at 25 °C as a function of time (5-80 min). MeOH concentration: 5%. Data were means of triplicate experiments ( SD.

Figure 4. Effects of anions and MeOH concentrations on DNA desorbed from aminosilane-modified magnetic nanoparticles (50 µg/40 µL) at 80 °C. Sodium citrate, sodium tartrate, sodium sulfate, and sodium chloride (1 M) in 25 mM Tris-HCl buffer (pH 8.0), and 1 M phosphate buffer (pH 8.0) were used as eluates. The recovery ratio was defined as the amount of desorbed λDNA/the amount of adsorbed λDNA × 100 (%). Data were means of triplicate experiments ( SD.

that the binding capacity of magnetic nanoparticles depends on the surface amine number. DNA adsorption reached a plateau at greater than 40 min (Figure 3) when an excess concentration of λDNA (1000 ng/mL) was added to the suspension of aminosilane-modified magnetic nanoparticles. This result agrees with a previous report that investigated oligonucleotide adsorption onto aminopropylsilane-modified glass slides fitted by the Langmuir reversible adsorption model,24 indicating a very rapid adsorption and a very slow desorption of DNA. The maximal DNA adsorbed value (0.85 fmol/cm2) was more similar to a previous result (1.1 fmol/cm2) for high-molecular weight dsDNA (2961 bp plasmid)30 than to that (0.5 or 19 pmol/cm2) for singlestranded oligonucleotides (20 or 21 bases).24,25 DNA Desorption Behavior from Aminosilane-Modified Magnetic Nanoparticles. Negligible DNA desorption occurred between pHs 6 and 11 although pH-dependent desorption was observed between pHs 9 and 11 (Figure 1, open circle). The effect of multiple anions was then investigated. Figure 4 shows the amount of DNA desorbed from aminosilane-modified magnetic nanoparticles using Tris-HCl buffer containing different anions in accordance with the Hofmeister ion series:

sulfate > phosphate > citrate > tartrate > chloride > nitrate The anions above are presented in terms of their abilities to precipitate egg globulin.31 Although slight DNA desorption from (30) Bezanilla, M.; Manne, S.; Laney, D. E.; Lyubchenko, Y. L.; Hansma, H. G. Langmuir 1995, 11, 655–659. (31) Kunz, W.; Henle, J.; Ninham, B. W. Curr. Opin. Colloid Interface Sci. 2004, 9, 19–37.

Langmuir, Vol. 25, No. 5, 2009 2959

Figure 5. DNA desorption from aminosilane-modified magnetic nanoparticles (50 µg/40 µL) with time. Elution buffer: 1 M phosphate buffer (pH 8.0) at 80 °C. The recovery ratio was defined as the amount of desorbed λDNA/the amount of adsorbed λDNA × 100 (%). Data were means of triplicate experiments ( SD.

aminosilane-modified magnetic nanoparticles was detected when either sodium citrate or sodium tartrate was used as an elution buffer, a remarkable desorption of DNA was not observed by the addition of the other anions except for phosphate. The amount of DNA desorbed remarkably increased with increasing MeOH concentration when sodium phosphate was used. The maximum DNA recovery was 85%. These results suggest that efficient DNA desorption occurred in the presence of phosphate ion regardless of the Hofmeister series. The addition of MeOH may play a role in the interactions of phosphate ions with surface amines with positive charges. Figure 5 shows DNA desorption kinetics from aminosilane-modified magnetic nanoparticles. DNA desorption reached a plateau at times greater than 40 min. Rapid desorption behavior was incompatible with the Langmuir reversible adsorption model, that is, very slow desorption of DNA occurred from aminosilane surfaces. This unusual phenomenon indicates that desorption originated from the replacement of DNA by phosphate ions. To further confirm the contribution of phosphate ions to the DNA adsorption/desorption behaviors, DNA adsorption in the presence of various concentrations of phosphate was investigated. As expected, DNA adsorption was inhibited with increasing phosphate concentrations and was completely inhibited in the presence of 400 mM phosphate (data not shown). Furthermore, DNA desorption using dATP was examined in place of sodium phosphate (Figure 6). Adenosine, which lacks triphosphate, was used as a control. Significant DNA desorption was confirmed only when dATP was added to λDNA-magnetic nanoparticle complexes. We concluded from these results that the main driving force for DNA desorption from aminosilane-modified magnetic nanoparticles was the replacement of adsorbed DNA with phosphate ion on the aminosilane-modified magnetic nanoparticles. DNA desorption increased with increasing dNTP concentrations (Figure 7A). Complete recovery of adsorbed DNA was attained using 100 mM dNTPs compared with 1 M phosphate buffer (85%) (Figure 4). The dNTP concentration was 10-times lower than that of phosphate ions. Because each nucleotide contains three phosphate groups, dNTPs would be more effective for DNA desorption compared with phosphate ions at the same molarity. Furthermore, the use of mononucleotide (dNTP) would be one of the most suitable replacements for polynucleotide on aminosilane-modified magnetic nanoparticles. PCR Amplification Using DNA Desorbed from Aminosilane-Modified Magnetic Nanoparticles. On the basis of the above results, PCR amplification of the gyrB gene was ex-

2960 Langmuir, Vol. 25, No. 5, 2009

Figure 6. Desorption of λDNA from aminosilane-modified magnetic nanoparticles (50 µg/40 µL) at 80 °C using either 2 mM dATP or 2 mM adenosine (in Tris-HCl buffer, pH 8.0) as an eluate. The relative recovery ratio was defined as the amount of desorbed λDNA/the amount of desorbed λDNA using dATP × 100 (%). The recovery ratio using dATP was defined as 100%. Data were means of triplicate experiments ( SD.

Tanaka et al.

Figure 8. Quantitative analysis of desorbed DNA from aminosilanemodified magnetic nanoparticles by real-time PCR. The recovery ratio was defined as the amount of desorbed E. coli genomic DNA/the amount of adsorbed E. coli genomic DNA × 100 (%). Data were means of triplicate experiments ( SD.

4 mM). Lower concentrations of MgCl2 should be used for optimal amplification using DNA desorbed by lower concentrations of dNTPs. In contrast, no amplification was observed when 1 M phosphate was used as an elutant. It was necessary to dilute the 1-M phosphate buffer more than 10-fold for successful PCR amplification (data not shown) because phosphate ions inhibit PCR. Furthermore, quantitative analysis of desorbed DNA was examined by real-time PCR when 0.5 mM to 100 mM dNTPs were used as elutants (Figure 8). Complete recovery was attained at more than 25 mM dNTPs. Therefore, DNA desorption by dNTPs enables highly effective DNA recovery and allows for subsequent PCR analysis.

Discussion

Figure 7. (A) Comparison of the recover ratios of λDNA from aminosilane-modified magnetic nanoparticles (50 µg/40 µL) at 80 °C using various concentrations of dNTPs or 1 M phosphate buffer. The relative recovery ratio was defined as the amount of desorbed λDNA/ the amount of desorbed λDNA using 1 M phosphate buffer × 100 (%). The recovery ratio using 1 M phosphate was defined as 100%. (B) Agarose gel showing PCR amplification of the gyrB gene in E. coli genomic DNA desorbed from aminosilane-modified magnetic nanoparticles using various concentrations of dNTPs or 1 M phosphate buffer. Key: M, ladder; P, positive control (1 ng of genomic DNA was used); N, negative control (no template).

amined using E. coli genomic DNA, which was desorbed from aminosilane-modified magnetic nanoparticles using 2.5 to 100 mM dNTPs or 1 M phosphate buffer. The released DNA solution (1 µL) including dNTPs was added to a 24-µL PCR mixture without dNTPs. The final dNTP concentration ranged from 0.1 to 4 mM. The gyrB gene (184 bp) was successfully amplified with each dNTP concentration (Figure 7B). The PCR products using DNA desorbed with 50 mM and 100 mM dNTPs were similar to those amplified using purified genomic DNA as a template. Smear bands were evident in mixtures containing either 2.5 mM or 25 mM dNTPs due to the relatively high concentrations of MgCl2. In PCR experiments, MgCl2 was fixed at 7.5 mM for optimal gene amplification in 100 mM dNTPs (final concentration:

The adsorption and desorption of DNA on aminosilanemodified magnetic nanoparticles were significantly affected by the addition of organic solvents (Table 1 and Figure 4). Especially, greater adsorption in aprotic solvents was observed compared with protic solvents (Table 1). One explanation for this might be that organic solvents interact with the propyl linkers on aminosilane-modified magnetic nanoparticles, and cause the amino groups to be pushed away from the surface, resulting in making them more accessible for DNA adsorption. Similarly, the adsorbed DNA on aminosilane-modified magnetic nanoparticles could be efficiently released by making a Hofmeister ion, phosphate more accessible for DNA desorption in the presence of organic solvents. On the other hand, efficient DNA desorption from aminosilane-modified magnetic nanoparticles was performed at 80 °C. Therefore, λDNA on aminosilanemodified magnetic nanoparticles could be partly denatured. A calorimetric analysis suggests that the binding manner of cationic ions to DNA changes with temperature, that is, the binding of guanidinium ion in the minor groove of double stranded DNA, and the binding to amide-like groups of single stranded (or melted) DNA.4 Therefore, further analysis should be performed to evaluate the effect of temperature on DNA desorption. On the other hand, the phosphodiester in DNA is a strong acid with pKa at less than 1. Most proteins, which are a possible interfering biomolecule for DNA adsorption, shows relatively higher pI value (>4). On the basis of these understandings, the selective adsorption and desorption of DNA based on electrostatic interaction has been developed. Several cationic solid supports for efficient DNA purification (DNA adsorption/desorption) based on changing pH have been proposed using magnetic particles (ChargeSwitch Technology; Invitrogen Co.) and polymers

DNA Adsorption/Desorption

(chitosan oligosaccharide lactate).19 For example, ChargeSwitch Technology includes DNA adsorption at pHs less than 6.5 and desorption at pHs greater than 8.5 due to the protonation/ deprotonation of ions on the particle surface. Our proposed method, that is, DNA adsorption by electrostatic interactions and subsequent desorption by the replacement of DNA by phosphate ions, showed similar protocols in terms of DNA purification. DNA recovery from human whole blood using aminosilane-modified magnetic nanoparticles has been demonstrated in our previous work.26 Selective DNA recovery or little protein binding was observed when blood lysate was used as a sample. However, it remains possible that acidic compounds (including acidic proteins) adsorbed onto aminosilane-modified magnetic nanoparticles. However, proteins adsorbed by electrostatic interactions, could not be desorbed from magnetic nanoparticles because DNA desorption is based on the replacement with phosphate ion. DNA purification based on electrostatic interactions between a cationic solid support and nucleic acids is an alternative approach to organic solvent-free DNA extraction, which could be useful especially in microfluidic devices. Furthermore, the Boom method is not adequate for DNA extraction from plant seeds because the addition of chaotropic agents increases solution viscosity due to aggregation of proteoglycans, resulting in poor separation of the

Langmuir, Vol. 25, No. 5, 2009 2961

magnetic beads.32 Therefore, DNA purification based on electrostatic interactions is also superior to conventional solidbased DNA extraction on this point. In this study, the addition of at least 5% MeOH was required for efficient DNA desorption. The addition of MeOH can inhibit PCR although the presence of MeOH promotes both DNA absorption and DNA desorption. However, 5% MeOH could be negligible in gene amplification because this MeOH concentration is diluted in the final PCR mixture. To eliminate the effect of MeOH on PCR, the proportion of at least ten for PCR mixture to one for DNA sample (containing 5% MeOH) should be required for successful PCR. Furthermore, the use of extremely high concentrations of phosphate ions in the elution buffer inhibits PCR amplification due to magnesium phosphate precipitation.26 Magnesium concentration is a crucial factor in PCR amplification. Therefore, the use of dNTPs may be better for DNA recovery than sodium phosphate (Figure 7) because dNTPs are integral components of PCR amplification. In conclusion, the DNA adsorption/desorption process proposed in this study will be a novel approach for DNA purification with a high recovery ratio. LA8032397 (32) Ota, H.; Lim, T. K.; Tanaka, T.; Yoshino, T.; Harada, M.; Matsunaga, T. J. Biotechnol. 2006, 125, 361–368.