Dumbbell-Shaped DNA Analytes Amplified by Polymerase Chain

This means that the peak resolution is rationally controlled by the base number of the .... Next, the running buffer solution of the analytes (WT and ...
0 downloads 0 Views 1004KB Size
Download PDF
Technical Note pubs.acs.org/ac

Dumbbell-Shaped DNA Analytes Amplified by Polymerase Chain Reaction for Robust Single-Nucleotide Polymorphism Genotyping by Affinity Capillary Electrophoresis Hideaki Shibata, Atsushi Ogawa,† Naoki Kanayama, Tohru Takarada,* and Mizuo Maeda Bioengineering Laboratory, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan S Supporting Information *

ABSTRACT: A sample preparation method was developed for single-nucleotide polymorphism (SNP) genotyping based on hybridization between a single-stranded DNA (ssDNA) analyte and an allele-specific oligonucleotide (ASO) probe. When the SNP site is located in the stable secondary structure, the folding of this analyte imposes kinetic penalties on the hybridization with the ASO probe. To address this issue, the sequence of the ssDNA analyte was converted from the original one so that the analyte exhibited a clear dumbbell-shaped structure composed of two stem−loop moieties and an unfolded probe-binding site. The asprepared analyte was structurally favorable for hybridization with the ASO probe, irrespective of the original sequence and secondary structure of the analyte. The sequence conversion was easily achieved by polymerase chain reaction using forward and reverse primers having an additional sequence at the 5′-terminus. These ssDNA analytes were subjected to affinity capillary electrophoresis using a diblock copolymer probe composed of an ASO segment and a poly(ethylene glycol) segment. The 70-base dumbbell-shaped analytes with a single-base difference were clearly separated within 12 min, although the original ones exhibited almost no separation due to the undesired folding of the probebinding site. This sample preparation method should open up a wide range of applications for the ASO probes in genetic analysis.

S

ssDNA analyte, this target analyte migrates more slowly than its single-base-substituted mutant in the presence of the affinity polymer probe, because the wild-type analyte encounters a large amount of hydrodynamic friction during electrophoresis by forming a reversible complex with the probe. In contrast, a negligibly weak interaction takes place between a single-basesubstituted mutant and the probe owing to a single-base mismatch. Consequently, two distinctive peaks are observed on the electropherogram; a peak for the mutant appears earlier than that for the wild-type analyte. The difference in migration time between the two peaks depends on the thermodynamic stability of the complex between the wild-type analyte and the probe and also depends on the amount of the hydrodynamic friction that the complex faces. This means that the peak resolution is rationally controlled by the base number of the ASO segment and by the molecular weight of the synthetic polymer segment.6−8 However, an ssDNA analyte often bears a folded structure via intramolecular base pairing under usual analytical conditions. The folding of the analyte imposes kinetic penalties on the hybridization with the affinity probe when the SNP site of interest is located in a stable secondary structure.9−11 In this

ingle-nucleotide polymorphisms (SNPs) are genetic variations resulting from single-base substitution in genomes. The single-base difference often accounts for phenotypic variance in living organisms. For example, crop pathogens produce mutants highly resistant to a fungicide by developing a single-base mutation in their gene, which codes the target protein of the fungicide.1 Most of the SNP detection methods thus far developed utilize a hybridization step between a single-stranded DNA (ssDNA) analyte and an allele-specific oligonucleotide (ASO) probe. They can discriminate between a wild-type gene and its single-base mutant by taking advantage of the fact that the thermodynamic stability of the fully matched double-stranded DNA (dsDNA) between the target ssDNA analyte and the ASO probe is higher than the stability of the single-base-mismatched dsDNA between the nontarget analyte and the probe. The ASO probe works well in various analytical platforms, such as microarrays, mass spectrometry, real-time polymerase chain reaction, and capillary electrophoresis.2 We have demonstrated that a diblock copolymer composed of an ASO segment and a synthetic polymer segment served as a good probe in affinity capillary electrophoresis (ACE) for highly reliable SNP genotyping.3−8 Water-soluble and electrically neutral polymers, such as poly(ethylene glycol) (PEG)3−7 and polyacrylamide,7,8 are appropriate for the polymer segment, since they can act as a molecular drag-tag as follows. When the ASO segment is designed to be complementary to a wild-type © XXXX American Chemical Society

Received: March 26, 2013 Accepted: May 9, 2013

A

dx.doi.org/10.1021/ac400900b | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Technical Note

base sequence of the additional fragment was properly designed, the ssDNA analyte obtained with enzymatic degradation of its complementary strand exhibited a dumbbell-shaped structure having a flexible probe-binding site. In the present study, this analyte was subjected to the ACE analysis using the PEG-b-DNA probe for SNP genotyping.

case, peptide nucleic acid (PNA), which is an electrically neutral oligonucleotide analogue, can work well as the ASO segment. We have recently reported that a diblock copolymer probe composed of an allele-specific PNA and PEG (PEG-bPNA) was able to retard a peak for a target wild-type ssDNA analyte with a stable hairpin structure by unwinding a folded region of the analyte to form the affinity complex.12,13 However, the PEG-b-PNA probe hardly exhibited sufficient water-solubility when the base number was larger than 8. This led to inefficient preparation of affinity probes and insufficient reproducibility in affinity separation. In the current study, we took an alternative and more versatile approach to affinity electrophoretic separation of folded ssDNA analytes; we developed a novel sample preparation method (Figure 1a). For this purpose, we

■

MATERIALS AND METHODS Reagents. All reagents were purchased from Wako Pure Chemical Industries (Osaka, Japan) unless otherwise noted. Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) was obtained from Pierce (Rockford, IL). All chemically synthesized DNAs used in this study were purchased from Tsukuba Oligo Service (Ibaraki, Japan). The 5′-thiol-terminated DNA was purified through a NAP-5 column from GE Healthcare (Buckinghamshire, UK). The DNA concentration was determined by measuring the absorbance at 260 nm with a UV-2550 spectrophotometer from Shimadzu Scientific Instruments (Kyoto, Japan). Maleimide-terminated methoxy PEG (mPEG-MAL) with a nominal molecular weight of 5000 (5K) or 20 000 (20K) was obtained from Nektar Therapeutics (San Carlos, CA). Deionized water (>18.1 MΩ) purified with a Milli-Q instrument from Millipore (Billerica, MA) was used for all experiments except PCR amplification of DNA samples. DNA Samples. Chemically synthesized 5′-fluorescein isothiocyanate (FITC)-labeled DNA strands (WT1, WT2, and WT3) and their single-base-substituted mutants (MT1, MT2, and MT3) were used for analytes. The base sequences are shown in Figure 2. WT1 and MT1 correspond to a fragment of the mitochondrial cytochrome b gene of cucumber powdery mildew and its single-base mutant at codon 143 (GGT to GCT), respectively. Similarly, WT2 and MT2 correspond to a fragment of the same gene of wheat powdery mildew and its mutant, respectively. Both of the single-base substitutions induce strobilurin-related fungicide resistance.14 The base sequences of WT3 and MT3 were identical to those of WT2 and MT2, respectively, differing only in the terminal regions. The sequence substitutions were made in order for WT3 and MT3 to bear a dumbbell-shaped secondary structure. We also prepared 5′-FITC-labeled 70-base analytes using PCR amplification of a fragment of the mitochondrial cytochrome b gene of wheat powdery mildew and its mutant (WT4 and MT4). Chemically synthesized dsDNA templates (89 base pairs) were employed as a model of the wild-type gene and the mutant gene. 5′-FITC-labeled forward primers and 5′-

Figure 1. Schematic representation of (a) the present method for preparing dumbbell-shaped ssDNA analytes and (b) the conventional method for preparing ssDNA analytes through PCR amplification and enzymatic digestion. The forward primer is 5′-FITC-labeled (green circle) while the reverse primer is 5′-phosphorylated (orange circle). The sense and antisense strands of the template are shown as a black line and gray line, respectively. The SNP sites of the sense and antisense strands are shown as an open circle and closed circle, respectively. The primer’s additional sequence for preparing the dumbbell-shaped analyte is shown as a green line (for the forward primer) or an orange line (for the reverse primer). Their complementary sequences are drawn in the corresponding light color.

performed polymerase chain reaction (PCR) amplification of a target gene using forward and reverse primers with an additional DNA fragment at the 5′-terminus. Given that the

Figure 2. Sequences of the ssDNA analytes and the ODN segments of the affinity probes. All analytes are FITC-labeled at the 5′-terminus. The probe-binding site of the analytes is shown in bold. The SNP site is shown in blue (for the wild-type analyte) or red (for the single-base-substituted analyte). The terminal base sequence changed from the original one is underlined. B

dx.doi.org/10.1021/ac400900b | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Technical Note

Figure 3. Electropherograms showing the separation of (a) WT1 and MT1, (b) WT2 and MT2, (c) WT3 and MT3, and (d) WT4 and MT4 with PEG(20K)-b-ODNW and PEG(5K)-b-ODNM. Peaks marked with an asterisk (∗) represent impurities. Conditions: running buffer, 50 mM Trisborate (pH 7.4) containing 7.0 mM MgCl2; capillary temperature, 35 °C; applied voltage, −15 kV; [probe] = 5.0 μM each; [analyte] = 50 nM each.

ODNM, were designed to be complementary to a part of WT and MT, including a single-base-substitution site, respectively (Figure 2). The affinity probes used in this study were PEG(20K)-b-ODNW and PEG(5K)-b-ODNM. The weightaveraged molecular weight (Mw), the number-averaged molecular weight (Mn), and the polydispersity index (Mw/ Mn) of the affinity probes were determined with aqueous size exclusion chromatography (Supporting Information). The obtained values are given in Table S1 (Supporting Information). The observed Mn values agreed well with the calculated ones. Capillary Electrophoresis. All experiments were performed using a P/ACE MDQ capillary electrophoresis system with a laser-induced fluorescent detector from BeckmanCoulter (Fullerton, CA). A CEP-coated capillary tube (internal diameter, 75 μm; total length, 60.5 cm; effective length, 40.5 cm) with a precoated inner surface was purchased from Agilent Technologies (Wilmington, DE). This capillary was employed to suppress the electroosmotic flow. As a running buffer, 50 mM Tris-borate buffer (pH 7.4) containing 7.0 mM MgCl2 was used.5 The fluorescence of the FITC-labeled analyte was measured with excitation at 488 nm and recording emission at 520 nm. The capillary temperature was held constant (35 °C) with a recirculating liquid coolant system. Prior to electrophoresis, a solution of PEG(20K)-b-ODNW and PEG(5K)-bODNM (5.0 μM each) in the running buffer was injected into the capillary from the cathode end by positive pressure (20 psi for 45 s). Next, the running buffer solution of the analytes (WT and MT, 50 nM each) was introduced into the capillary using a similar method (0.50 psi for 10 s). The analytes were electrophoresed under reversed polarity with a constant voltage of −15 kV for 20 min. Between measurements, the capillary was washed with deionized water (20 psi for 60 s) and then

phosphorylated reverse primers were used for the PCR amplification. All sequences are shown in Figure S1 (Supporting Information). All PCR reactions were performed on a 20 μL scale. Each reaction mixture was composed of 1.0 nM of each template, 1.0 μM of each primer, and 10 μL of 2× PrimeSTAR MAX Premix (Takara Bio, Ohtsu, Japan). Nuclease-free water was used for preparing the reaction solutions. Five thermal cycles using steps of 98 °C for 10 s, 55 °C for 15 s, and 72 °C for 15 s, followed by 25 cycles using steps of 98 °C for 10 s, 55 °C for 5 s, and 72 °C for 15 s, were performed with a thermal cycler (iCycler; Bio-Rad Laboratories, Hercules, CA). After all cycles, the reaction solution was kept at 4 °C. The PCR products were purified using a MinElute PCR Purification Kit from Qiagen (Hilden, Germany). Then, the 5′-phosphorylated antisense strands of the obtained PCR products were digested by λ exonuclease on a 20 μL scale. Each reaction mixture was composed of 10 μL of the PCR solutions, 5 U of λ exonuclease, and its reaction buffer (New England Biolabs Japan, Tokyo, Japan). The mixture was incubated at 37 °C for 1.0 h and at 75 °C for 10 min using the thermal cycler. After the digestion, the residual 5′-FITC-labeled sense strands were purified using a QIAquick Nucleotide Removal Kit (Qiagen) to yield a mixture of almost equal amounts of WT4 and MT4. For all DNA analytes, the folded structures and their thermal and thermodynamic parameters under the electrophoresis conditions were calculated using mfold.15 Affinity Probes. Diblock copolymers between oligodeoxyribonucleotide (ODN) and PEG (PEG-b-ODN) were synthesized by the previously reported method.3 The 5′-thiolterminated ODN was coupled with the mPEG-MAL via the Michael addition reaction in the presence of TCEP as a reductant. The nominal molecular weight of PEG was 5000 (5K) or 20 000 (20K). The ODN segments, ODNW and C

dx.doi.org/10.1021/ac400900b | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Technical Note

filled with the running buffer containing the probe (20 psi for 45 s). Each electrophoresis was conducted in triplicate to confirm the reproducibility.

that in Figure 3a, even though the probe-binding sites of WT2 and MT2 had sequences identical to those of WT1 and MT1, respectively (Figure 2). The migration times of WT2 and MT2 were reduced compared to those of WT1 and MT1, respectively. The results indicate that the affinity interaction between the analytes (WT2 and MT2) and the corresponding probes was suppressed during electrophoresis. The thermodynamic stability of the folded structures of WT2 and MT2 accounted for this undesired suppression. Figure 5a shows the most stable secondary structure around the probe-binding site of WT2 and MT2, which was predicted by using a nearestneighbor base-pairing model (mfold)15 under the electrophoresis conditions.12,13 The SNP site of WT2 and MT2 was buried in the stem−loop structure. The calculated melting temperature (Tm) values for this local hairpin structure of WT2 and MT2 were 50 and 76 °C, respectively. These results suggest that both the secondary structures should be maintained at the capillary temperature (35 °C). Although one could perform the accurate SNP analysis with careful deconvolution of the overlapping peaks in Figure 3b, it is generally difficult to achieve reliable SNP genotyping based on the electropherogram exhibiting multiple peaks without baseline separation. Design of Dumbbell-Shaped Analytes. To make this SNP genotyping method more robust, we need to completely exclude the possibility that such a stochastic block might occur. For this purpose, we decided to prepare dumbbell-shaped ssDNA analytes, which should exclusively expose their probebinding site, by introducing artificial 5′- and 3′-terminal sequences with PCR (Figure 1a). Our research group has previously reported the effects of prehybridization between a folded ssDNA analyte and two 20-base ODNs termed “helpers,” which were complementary to the upstream and downstream positions of the probe-binding site.17 Formation of this ternary complex enabled us to overcome the problem with respect to the emergence of undesired secondary structures for SNP genotyping. The current method corresponds to PCRbased integration of the helpers into the terminal regions of ssDNA analytes. This incorporation would eliminate a timeconsuming annealing step between the analyte and the externally added helpers and also would increase the fraction of the analyte with the desired conformation due to the intramolecular base pairing.

■

RESULTS AND DISCUSSION Separation of Unfolded DNA. Figure 3a shows that a mixture of WT1 and MT1 was separated by simultaneously using PEG(20K)-b-ODNW and PEG(5K)-b-ODNM as the affinity probes.5 The migration time for WT1 was longer than that for MT1. This is because the hydrodynamic friction that WT1 encountered when it hybridized to the PEG(20K)-bODNW probe during electrophoresis was stronger than that MT1 received when hybridized to PEG(5K)-b-ODNM. The peaks observed at approximately 7 min were assigned to impurities involved in the ssDNA analytes. This migration time agreed well with that of ssDNA with a scrambled sequence. The coincidence led us to conjecture that the impurities were 5′FITC-labeled DNA strands with incompletely deprotected nucleic acid bases.16 A certain fraction of these byproducts should have a negligibly small interaction with the probe. The diagram for the affinity separation is depicted in Figure 4.

Figure 4. Schematic diagram illustrating affinity capillary electrophoresis of wild-type and mutant analytes when simultaneously using two kinds of the PEG-b-ODN probes. For the probe, the gray and green wavy lines represent the ODN and PEG segment, respectively.

Separation of Natively Folded DNA. Figure 3b shows the results of the electrophoresis of WT2 and MT2 under the same conditions. The electropherogram was clearly different from

Figure 5. Secondary structure of (a) the probe-binding region of WT2 and MT2, (b) the overall region of WT3 and MT3, and (c) the overall region of WT4 and MT4 under the electrophoresis conditions ([MgCl2] = 7.0 mM, at 35 °C) predicted by using mfold.15 The calculated Tm values are also shown in parentheses. The probe-binding site is shown in bold. The SNP site is shown in blue (WT) or red (MT). The black and red dashed lines indicate a Watson−Crick and mismatched base pair, respectively. D

dx.doi.org/10.1021/ac400900b | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Technical Note

bases long; this total length was long enough for efficient purification by the spin column used in this study. Considering all these steps together, we designed the base sequences of WT4 and MT4 as shown in Figure 5c. PCR amplification was successfully accomplished by using the newly designed primers (Figure S1b, Supporting Information). Separation of Dumbbell-Shaped Analytes. We first confirmed that a mixture of chemically synthesized WT4 and MT4 was separated by simultaneously using PEG(20K)-bODNW and PEG(5K)-b-ODNM (Figure S4, Supporting Information). We then conducted similar ACE of WT4 and MT4 obtained from the PCR amplification. Good separation was achieved as shown in Figure 3d. The migration times of the PCR-amplified WT4 and MT4 were in accordance with those of the chemically synthesized WT4 and MT4, respectively (Figure S4, Supporting Information). These results indicate that the current separation system worked well for the PCRamplified DNA analytes, which adopted the deliberate secondary structure. In contrast, a mixture of the 5′-FITClabeled wild-type sense strand and the 5′-FITC-labeled singlebase-mutated sense strand, both of which were employed as a part of the PCR template, could not be directly separated by the same ACE; an electropherogram almost identical to that in Figure 3b was obtained (data not shown). Therefore, the present PCR-based sample preparation was essential for achieving a clear separation. In Figure 3d, the main peak observed at approximately 7 min was assigned to the FITC-labeled forward primers and/or the FITC-labeled dsDNA that were left undigested by the nuclease, either or both of which could have been incompletely excluded from the final ssDNA products by the current purification procedure. As shown in Figure S5 (Supporting Information), a single peak for the PCR product (dsDNA) without the enzymatic digestion was actually observed at ca. 7 min under the identical electrophoresis conditions. When we used either PEG(20K)-b-ODNW or PEG(5K)-b-ODNM, a peak for the nontarget ssDNA analyte (MT4 for PEG(20K)-b-ODNW or WT4 for PEG(5K)-b-ODNM) completely overlapped the peak for these impurities on the electropherogram (data not shown). This is because the electrophoretic mobility of DNA is independent of the sequence and length, provided that the base number is larger than 10.18 Thus, it is also demonstrated that simultaneous use of two different probes is of practical importance for the affinity separation of ssDNA analytes prepared from PCR products, which might contain FITClabeled primers and duplexes as impurities.

For the preliminary ACE experiments, we initially used chemically synthesized ssDNA analytes that were expected to exhibit dumbbell-shaped structures (WT3 and MT3). Their secondary structures are shown in Figure 5b. The dumbbellshaped analytes that we first designed had two stem−loop moieties composed of a 5-base loop and an 11-base pair stem, which were connected to each other by an unfolded and flexible probe-binding site. Figure 3c shows that a mixture of WT3 and MT3 was separated by simultaneously using PEG(20K)-bODNW and PEG(5K)-b-ODNM. As expected, the peak for MT3 appeared earlier than that for WT3, indicating that the separation principle shown in Figure 4 held true also in this case. It should be noted that the migration times of WT3 and MT3 were in fair agreement with those of WT1 and MT1, respectively (Figure 3a). This result means that the stem−loop moieties neighboring the probe-binding site caused a negligibly small steric hindrance for the hybridization with the affinity polymer probes. The 2-base gap between the probe-binding site and each of the stem−loop moieties might have been enough to alleviate the steric repulsion between the analyte’s stem and the probe’s PEG segment (Figure 5b). PCR Amplification of Dumbbell-Shaped Analytes. We then moved on to the PCR-based preparation of WT3 and MT3 using the templates and primers shown in Figure S1a (Supporting Information). The present sample preparation method is schematically outlined in Figure 1a. First, we performed PCR using the 5′-FITC-labeled forward primer and the 5′-phosphorylated reverse primer. The PCR product was then subjected to digestion of the 5′-phosphorylated antisense strand by λ exonuclease to yield the 5′-FITC-labeled ssDNA analyte. The base sequence of the additional nucleotide involved in the primers was designed in such a way that the terminal regions of the obtained ssDNA analyte could undergo intramolecular base pairing with flanking regions of the probebinding site. As a result, the ssDNA analyte thus obtained should exhibit the dumbbell-shaped structure, which is structurally favorable for hybridization with the ASO probe irrespective of the analyte’s original sequence and secondary structure. However, the PCR to obtain WT3 and MT3 showed extremely slow progression (data not shown). The resistance to amplification was likely attributable to two phenomena. First, the forward and reverse primers may have exhibited a rigid hairpin structure (Figure S2a, Supporting Information), thereby limiting their hybridization to the template during the PCR cycles. To overcome this issue, we reduced the thermodynamic stability of the hairpin structure; we decreased the stem’s basepair number from 11 to 8 and also substituted a G−T mismatched pair for a Watson−Crick base pair at the middle of each stem. These alterations reduced the calculated Tm values for folding of the forward and reverse primer from 77 to 54 °C and 68 to 45 °C, respectively (Figure S2, Supporting Information). Both of the new Tm values were approximately equal to or lower than the primer-binding temperature for PCR (55 °C). Second, partially folded sense and antisense strands may have been formed during the PCR cycles, and these strands could have been subjected to an undesired extension reaction at the 3′-terminus (Figure S3, Supporting Information). These byproducts should prohibit the further amplification. To exclude this possibility, we added seven nucleotides (A7) to the 5′- and 3′-termini of WT3 and MT3 as a protecting group. The base number of this additional sequence was determined such that the final ssDNA products would be 70

■

CONCLUSION We developed a PCR-based sample preparation method for easily achieving sequence-selective separation of folded ssDNA analytes by ACE. To our knowledge, this is the first example of the preparation of ssDNA analytes that were designed to exhibit a favorable secondary structure for affinity interaction with the ASO probe for SNP genotyping, despite the use of chemically synthesized dsDNA for PCR templates. The ssDNA analytes obtained from living organisms are currently being subjected to the same ACE analysis. Since the reaction efficiencies of PCR and enzymatic digestion for a wild-type sample could be slightly different from those for a single-basemutated sample, highly quantitative SNP analysis based on the present method might require appropriate corrections to obtained values. These investigations will be reported elsewhere. Importantly, the ssDNA analytes thus prepared could E

dx.doi.org/10.1021/ac400900b | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Technical Note

also be used in other DNA detection methods, such as DNA microarrays and mass spectrometry. The present sample preparation method should open up a wide range of applications for the ASO probes in gene diagnosis.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Address †

Proteo-Science Center, Ehime University, 3 Bunkyo-cho, Matsuyama, Ehime 790-8577, Japan. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by a Grant for Ecomolecular Science Research provided by RIKEN and by a Grant-in-Aid for Scientific Research (A) (No. 20245020) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

■

REFERENCES

(1) Ma, Z.; Michailides, T. J. Crop Prot. 2005, 24, 853−863. (2) Syvänen, A. C. Nat. Rev. Genet. 2001, 2, 930−942. (3) Kanayama, N.; Takarada, T.; Kimura, A.; Shibata, H.; Maeda, M. React. Funct. Polym. 2007, 67, 1373−1380. (4) Kanayama, N.; Takarada, T.; Kimura, A.; Shibata, H.; Maeda, M. J. Sep. Sci. 2008, 31, 837−844. (5) Kanayama, N.; Takarada, T.; Shibata, H.; Kimura, A.; Maeda, M. Anal. Chim. Acta 2008, 619, 101−109. (6) Tsukada, H.; Watanabe, T.; Kanayama, N.; Takarada, T.; Maeda, M. Electrophoresis 2012, 33, 2122−2129. (7) Takarada, T.; Maeda, M. Bull. Chem. Soc. Jpn. 2013, 86, 547−556. (8) Kanayama, N.; Shibata, H.; Kimura, A.; Miyamoto, D.; Takarada, T.; Maeda, M. Biomacromolecules 2009, 10, 805−813. (9) Sekar, M. M. A.; Bloch, W.; St John, P. M. Nucleic Acids Res. 2005, 33, 366−375. (10) Gao, Y.; Wolf, L. K.; Georgiadis, R. M. Nucleic Acids Res. 2006, 34, 3370−3377. (11) Chen, C.; Wang, W.; Wang, Z.; Wei, F.; Zhao, X. S. Nucleic Acids Res. 2007, 35, 2875−2884. (12) Kundu, L. M.; Tsukada, H.; Matsuoka, Y.; Kanayama, N.; Takarada, T.; Maeda, M. Anal. Chem. 2012, 84, 5204−5209. (13) Tsukada, H.; Kundu, L. M.; Matsuoka, Y.; Kanayama, N.; Takarada, T.; Maeda, M. Anal. Biochem. 2013, 433, 150−152. (14) Ishii, H.; Fraajie, B. A.; Sugiyama, T.; Noguchi, K.; Nishimura, K.; Takeda, T.; Amano, T.; Hollomon, D. W. Phytopathology 2001, 91, 1166−1171. (15) Zuker, M. Nucleic Acids Res. 2003, 31, 3406−3415. (16) Barley-Maloney, L.; Agris, P. F. Anal. Biochem. 2002, 306, 135− 143. (17) Han, A.; Inoue, A.; Hosokawa, K.; Maeda, M. Anal. Biochem. 2007, 371, 124−126. (18) Stellwagen, E.; Lu, Y.; Stellwagen, N. C. Biochemistry 2003, 42, 11745−11750.

F

dx.doi.org/10.1021/ac400900b | Anal. Chem. XXXX, XXX, XXX−XXX