Article pubs.acs.org/ac
Quencher-Free Fluorescent Method for Homogeneously Sensitive Detection of MicroRNAs in Human Lung Tissues Guichi Zhu,† Li Liang,‡ and Chun-yang Zhang*,† †
Single-Molecule Detection and Imaging Laboratory, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, China ‡ Department of Tumor Chemotherapy and Radiation Sickness, Peking University Third Hospital, Beijing 100191, China ABSTRACT: Sensitive and accurate analysis of microRNA (miRNA) expression is imperative for understanding the biological functions of miRNAs and the early diagnosis of human cancers. Here, we develop a quencher-free fluorescent method for homogeneously sensitive detection of let-7a miRNA using the target-triggered recycling signal amplification in combination with a 2-aminopurine probe. The 2-aminopurine probe is characterized by the substitution of 2-aminopurine for adenine in the DNA strand and the quenching of 2-aminopurine fluorescence through its stacking interaction with the adjacent bases. The binding of target miRNA with the 2-aminopurine probe initiates the extension reaction in the presence of polymerase to produce the DNA duplexes. These DNA duplexes can be further cleaved by lambda exonuclease through the recycling digestion to release abundant free 2-aminopurines, leading to an enhanced fluorescence signal. The proposed method exhibits high sensitivity with a detection limit of 0.3 fmol, and it can even discriminate the single-base difference among the miRNA family members. More importantly, this method can accurately distinguish the expression of let-7a miRNA in human lung tissues between ten non small cell lung cancer (NSCLC) patients and ten healthy persons, holding a great potential for further application in early clinical diagnosis.
M
from the low sensitivity and the poor specificity because of the small size, the sequence homology among the family members and the low abundance of miRNAs. To improve the sensitivity and the specificity of miRNA assay, a variety of amplification strategies have been introduced, such as real-time polymerase chain reaction (RT-PCR),13 rolling circle amplification (RCA),14 exponential amplification reaction (EXPAR),15 quadratic isothermal amplification16 and nanoparticle-based signal amplification.17 Among these approaches, the RT-PCR method has gained increasing attention because of its high sensitivity and practicality, but it involves the precise control of temperature cycling and the sophistication of primer design.13 The RCA amplification is a simple enzymatic process to generate long single-stranded DNAs under isothermal conditions, but it is usually time-consuming with the total reaction process duration of 4−8 h.14 Alternatively, the EXPAR method can provide >106-fold amplification within minutes, and can obtain a detection limit at the attomolar level for miRNA assay.15a,b However, this method might be hindered by nonspecific background amplification. The quadratic isothermal amplification method might achieve single-cell sensitivity for miRNA assay,16 but the involvement of doubly labeled fluorescent probe (e.g., molecular beacon) might increase the experimental cost and the complexity of probe design. The nanoparticles possess unique electronic, optical and catalytic
icroRNAs (miRNAs) are short endogenous noncoding RNA molecules with lengths of about 22 nucleotides, which are derived from long primary transcripts by the catalysis of Dicer enzyme.1 More than 450 kinds of miRNAs have been identified in the human genome,2 and most of them play a crucial role in various biological processes including cell development, differentiation, apoptosis and proliferation through imperfect pairing with the target mRNA of proteincoding genes and the transcriptional regulation of their expression.3 Recent research demonstrates that the aberrant expression of miRNAs is associated with many human carcinoma such as lung cancer,4 hepatocellular carcinoma,5 breast cancer6 and gastric carcinoma,7 suggesting that miRNAs might be used as valuable biomarkers for early diagnosis of cancers in asymptomatic individuals.8 Among human cancers, lung cancer is the leading cause of cancer-related death in the world, and its etiology is primarily genetic and epigenetic damage caused by cigarette smoking, air pollution and longterm exposure to the environments containing carcinogenic substances.4 Lung cancer has two major subtypes including small cell lung cancer (SCLC) and nonsmall cell lung cancer (NSCLC), and NSCLC accounts for approximately 80% in all cases.9 Much emerging evidence suggests that miRNAs might function as potential oncogenes or tumor suppressors in lung cancer.10 Consequently, accurate and quantitative analysis of miRNA expression is critical for further understanding the biological functions of miRNAs and the early diagnosis of lung cancer. The Northern blotting11 and microarray-based methods12 are usually used for miRNA detection, but they might suffer © XXXX American Chemical Society
Received: September 7, 2014 Accepted: October 30, 2014
A
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
properties, making them ideal candidates for signal generation and amplification in miRNA assay.17 Some typical examples include the use of electrocatalytic nanoparticle tags to chemically amplify the signal,17a the use ferrocene-capped gold nanoparticle/streptavidin conjugates for amplified voltammetric detection17b and the use of Au nanoparticle for amplified surface plasmon resonance imaging.17c Nevertheless, these methods usually involve the time-consuming synthesis of nanoparticles and the complicated chemical modification of substrates. Recently, several new methods such as strand displacement amplification,18 padlock probe-based exponential RCA assay,19 size-coded ligation chain reaction,20 electrochemical biosensor21,22 and peptide nucleic acid−graphene oxide-based methods23 have been developed for the detection of miRNAs in both the spiked samples and the cell samples. However, the development of simple and highly sensitive methods with the capability of clinical application still remains a great challenge. The nucleotide of 2-aminopurine is a fluorescent analog of adenine, and is usually used as a site-specific probe for the study of nucleic acid structure and dynamics.24 The 2-aminopurine exhibits a weak fluorescence when being incorporated into the DNA strands, but an enhanced fluorescence when being free in the solution.25 In contrast to the conventional molecular beacon, the quenching of 2-aminopurine fluorescence results from its stacking interaction with the adjacent bases24b without the involvement of any extra quenchers. Moreover, 2aminopurine is insensitive to the visible light because its excitation wavelength locates at 310 nm.26 Here, we develop a quencher-free fluorescent method for homogeneously sensitive detection of miRNAs using the target-triggered recycling signal amplification in combination with 2-aminopurine probe. The proposed method allows for sensitive detection of miRNAs with a detection limit of as low as 0.3 fmol, and it can discriminate the single-nucleotide difference among miRNA family members. Importantly, this method can accurately distinguish the expression of let-7a miRNA in human lung tissues between ten NSCLC patients and ten healthy persons, holding great potential for further applications in early clinical diagnosis.
Table 1. Sequences of 2-Aminopurine Probes, Helper DNAs and miRNAsa note 2-aminopurine probe-a 2-aminopurine probe-b helper DNA-1 helper DNA-2 helper DNA-3 let-7a let-7b let-7c
Sequence (5′-3′) AAA ACT↓CAC TGC ACT CGA TCG GAA CTA TAC AAC CTA CTA CCT CA-NH2 AAA ACT↓CAC TGC ACT CGA TCG GAA CTA TAC AAC CTA CTA CCT CA-NH2 AAG ATC GAG TGC C-NH2 AAC GAT CGA GTG CAC C-NH2 AAC CGA TCG AGT GCA GTC C-NH2 UGA GGU AGU AGG UUG UAU AGU U UGA GGU AGU AGG UUG UGU GGU U UGA GGU AGU AGG UUG UAU GGU U
a
The bold regions of 2-aminopurine probes and helper DNAs indicate the phosphorothioate modification. The arrow in 2-aminopurine probes indicates the nicking position of Nb.BtsI. The bold italic region of 2-aminopurine probes indicates the binding sequence of helper DNA-2. The italic region of helper DNAs indicates the binding sequence of 2-aminopurine probes. The underlined region of 2aminopurine probes indicates the 2-aminopurine substitution. The bases in let-7b and let-7c that differ from those in let-7a are marked in bold. There are two base mismatches near the 3′ end of miRNAs between let-7a and let-7b, and one base mismatch near the 3′ end of miRNAs between let-7a and let-7c.
KF polymerase, 5 U Nb.BtsI, 5 U lambda exonuclease, 40 U RNase inhibitor, 1× NEB CutSmart buffer (20 mM Tris− acetate, 50 mM potassium acetate, 10 mM magnesium acetate, 0.1 mg/mL BSA, pH 7.9), DEPC treated water, let-7a miRNA at different concentrations, and incubated at 37 °C for 60 min. At last, the resultant reaction solution was subjected to the fluorescence measurement using a Hitachi F-4500 fluorometer (Tokyo, Japan) equipped with a xenon lamp as the excitation source. The spectra were recorded in the range from 345 to 500 nm at an excitation wavelength of 310 nm. The excitation and emission slits were set for 5.0 and 5.0 nm, respectively. The fluorescence signals were normalized, and the intensity at the emission wavelength of 365 nm was used for data analysis. Gel Electrophoresis. A 10% nondenaturating polyacrylamide gel electrophoresis (PAGE) analysis of reaction products was carried out in 1× TBE (9 mM Tris−HCl, pH 7.9, 9 mM boric acid, 0.2 mM ethylenediaminetetraacetic acid, EDTA) at a 120 V constant voltage for 35 min at room temperature. The gel was stained by SYBR Green I, and scanned by a Kodak Image Station 4000 MM (Woodbridge, CT, USA). Clinical Sample Analysis. The total RNA was extracted from the human lung tissues of NSCLC patients and healthy persons using the miRNeasy FFPE kit (Qiagen, Germany) according to the manufacturer’s procedures, and the concentrations of total RNA were determined by measuring the absorbance at 260 nm with a spectrophotometer. The extracted total RNA concentrations were diluted to 200 ng/μL with DEPC-treated water (RNase free), and 20 μL of total RNA sample (4 μg in total) was added to each 50 μL of reaction solution for measurement.
■
EXPERIMENTAL SECTION Materials. The 2-aminopurine probes, helper DNAs, miRNAs of let-7a, let-7b and let-7c, deoxynucleotide solution mixture (dNTPs), recombinant RNase inhibitor and diethylpyrocarbonate (DEPC) treated water (RNase free) were purchased from TaKaRa Biotechnology Co., Ltd. (Dalian, China). The sequences of oligonucleotides are listed in Table 1. The Klenow fragment polymerase (3′ → 5′exo-, KF polymerase), the nicking enzymes of Nb.BtsI, the lambda exonuclease, NEB CutSmart buffer and lambda buffer were obtained from New England BioLabs (Beverly, MA, USA). SYBR Green I was obtained from Xiamen Bio-Vision Biotechnology Co. Ltd. (Xiamen, China). The miRNA qRT-PCR kit was purchased from Life Technologies (Carlsbad, CA, USA). The paraffinembedding lung tissue samples of NSCLC patients and healthy persons were obtained from the Peking University Third Hospital (Beijing, China). All other chemical reagents were of analytical grade and obtained from Sigma-Aldrich (St. Louis, MO, USA). Detection of let-7a miRNA. The experiments were performed in 50 μL of solution consisting of 200 nM 2aminopurine probe, 200 nM helper DNA, 300 μM dNTPs, 5 U
■
RESULTS AND DISCUSSION Principle of miRNA Assay. The principle of miRNA assay is illustrated in Scheme 1. We designed a 2-aminopurine probe with two substitutions of 2-aminopurine molecules for adenines in a single-stranded DNA (ssDNA). Because the fluorescence of 2-aminopurine in double-stranded DNA (dsDNA) is much lower than that in single-stranded DNA (ssDNA),27 we further
B
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Scheme 1. Schematic Illustration of miRNA Assay Using Target-Triggered Recycling Signal Amplification in Combination with 2-Aminopurine Probe
introduced a helper DNA, which can hybridize with the 2aminopurine probe, to reduce the background fluorescence signal. In the presence of target miRNA, the amplification reaction is initiated with the assistance of KF polymerase, and the helper DNA is replaced by the target extension, producing a complete dsDNA with the recognition sites for nicking enzyme of Nb.BtsI. The subsequent cleavage of 2-aminopurine probe by Nb.BtsI exposes the site of phosphate group (PO4) for the lambda exonuclease digestion, releasing both the free 2aminopurine molecules and the extended targets. The extended target can further hybridize with new 2-aminopurine probe repeatedly, generating the cycle of nicking−digestion−hybridization and consequently the production of abundant free 2aminopurine molecules for significant fluorescence enhancement. While in the absence of target miRNA, the helper DNA is unable to be replaced, and the recognition site of nicking enzyme in 2-aminopurine probe is unavailable. As a result, the cycle of nicking−digestion−hybridization cannot be initiated, and no fluorescence enhancement is observed. To demonstrate the feasibility of the proposed method for miRNA assay, we performed both fluorescence measurement (Figure 1A) and nondenaturating PAGE analysis (Figure 1B). In the presence of let-7a miRNA target and three kinds of enzymes (KF polymerase, Nb.BtsI nicking enzyme and lambda exonuclease), a strong fluorescence signal with the characteristic emission peak of 365 nm is observed (Figure 1A), indicating the release of abundant free 2-aminopurine molecules due to the cycle of nicking−digestion−hybridization. However, no significant fluorescence signal is observed in the absence of either let-7a miRNA or one kind of enzyme (Figure 1A), indicating that the cycle of nicking−digestion−hybridization cannot be initiated. The above results are further confirmed by the nondenaturating PAGE analysis (Figure 1B). Only a weak DNA band is observed in the presence of let-7a miRNA and three kinds of enzymes (Figure 1B, lane 1), indicating the digestion of 2-aminopurine probe by lambda exonuclease due to the cycle of nicking−digestion−hybridization. In contrast, a distinct and strong DNA band is observed in the absence of either let-7a miRNA or one kind of enzyme (Figure 1B, lanes 2−5), indicating that the cycle of nicking− digestion−hybridization cannot be initiated.
Figure 1. (A) Normalized fluorescence emission spectra of reaction products in the presence of let-7a miRNA + three enzymes (red line), control + three enzymes (black line), let-7a miRNA + KF polymerase + Nb.BtsI (blue line), let-7a miRNA + KF polymerase + lambda exonuclease (green line) and let-7a miRNA + Nb.BtsI + lambda exonuclease (purple line), respectively. The concentration of the let-7a miRNA target is 10 nM. (B) Nondenaturating PAGE analysis of reaction products. Lane m is the DNA ladder marker; Lanes 1, 2, 3, 4, and 5 represent let-7a miRNA + three enzymes, control + three enzymes, let-7a miRNA + KF polymerase + Nb.BtsI, let-7a miRNA + KF polymerase + lambda exonuclease, and let-7a miRNA + Nb.BtsI + lambda exonuclease, respectively.
The quenching of 2-aminopurine fluorescence results from its stacking interaction with the adjacent bases,24b and the fluorescence of 2-aminopurine in dsDNA is much lower than that in ssDNA because of the more efficient stacking interaction among the bases in dsDNA.27 To reduce the background fluorescence signal by keeping 2-aminopurine molecules in dsDNA, we introduced a helper DNA to hybridize with the 2aminopurine probe. In both the helper DNA and the 2aminopurine probe, the 5′ end is modified with phosphorothioate to prevent the nonspecific digestion of lambda exonuclease, and the 3′ end is modified with amination to prevent the nonspecific extension of KF polymerase (Table 1). In addition, the length of helper DNA is crucial to the cycle of nicking−digestion−hybridization. When the matched bases between the helper DNA and the 2-aminopurine probe are too short, the DNA duplexes are unstable at the reaction temperature. When the matched bases are too long, the helper DNA is hard to be replaced by the extended targets. To select the proper helper DNA, three helper DNAs with nine matched bases (helper DNA-1), twelve matched bases (helper DNA-2) and fifteen matched bases (helper DNA-3), respectively, were investigated in the presence of 10 nM let-7a miRNA. As shown in Figure 2A, the F/F0 value in the presence of helper DNA-2 is much higher than those in the presence of helper DNA-1 and helper DNA-3, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. Therefore, the helper DNA-2 is the best choice experimentally, C
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
formed by the hybridization of helper DNA-2 with its complementary sequence (48.2 °C) is about 10 °C higher than the reaction temperature of 37 °C, which can not only keep the DNA duplex stable but also ensure the proceeding of subsequent replacement reaction. It should be noted that the introduction of helper DNA does not influence the digestion of 2-aminopurine probe by Nb.BtsI nicking enzyme and the subsequent cycle of nicking−digestion−hybridization. In the presence of let-7a miRNA, no significant difference is observed in the obtained fluorescence signals no matter the helper DNA is present or not (Figure 2B). Nevertheless, the helper DNA induces a significant decrease in the background signal. The signal-to-noise ratio in the presence of helper DNA has increased by as much as 2.5-fold, as compared with that in the absence of helper DNA, greatly benefiting the improvement of detection sensitivity. Optimization of Experimental Condition. To optimize the design of the 2-aminopurine probe, two 2-aminopurine probes (Table 1) including one with two substitutions of 2aminopurine (probe-a) and another with one substitution of 2aminopurine (probe-b) were investigated. As shown in Figure 3A, the F/F0 value of probe-a group is higher than that of probe-b group, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. This result indicates that probe-a with two substitutions of 2aminopurine has higher signal-to-noise ratio than probe-b with one substitution of 2-aminopurine. It should be noted that the mismatched base pair of 2-aminopurine:thymine has lower stability than the natural base pair of adenine:thymine,28 which might lead to the unstable thermodynamic performance and the poor hybridization efficiency for the probe with too many substitutions of 2-aminopurine. Taking into account the design complexity and the synthesis cost of 2-aminopurine probe as well as the hybridization efficiency, probe-a with two
Figure 2. (A) Optimization of helper DNA. (B) Influence of helper DNA on the signal-to-noise ratio. The concentration of let-7a miRNA is 10 nM. Error bars show the standard deviation of three experiments.
consistent with the theoretical calculation as well. The melting temperature (Tm) of DNA duplex formed by the hybridization of helper DNA-1, helper DNA-2 and helper DNA-3 with its complementary sequence is calculated to be 27.6, 48.2 and 58.5 °C, respectively. In theory, the Tm value of DNA duplex
Figure 3. (A) Variance of fluorescence value ratio of F/F0 with the design of 2-aminopurine probe. Probe-a is a 2-aminopurine probe with two substitutions of 2-aminopurine, and probe-b is a 2-aminopurine probe with one substitution of 2-aminopurine. (B) Variance of fluorescence value ratio of F/F0 with different buffers. (C) Variance of the normalized fluorescence intensity with the reaction time. (D) Variance of fluorescence value ratio of F/F0 with the concentration of 2-aminopurine probe. (E) Variance of fluorescence value ratio of F/F0 with the amount of nicking enzyme Nb.BtsI. (F) Variance of fluorescence value ratio of F/F0 with the amount of lambda exonuclease. F and F0 are the fluorescence signals in the presence and in the absence of 10 nM let-7a miRNA, respectively. Error bars show the standard deviation of three experiments. D
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
substitutions of 2-aminopurine is used in the subsequent research. This research involves three kinds of enzymes and two different buffers. The CutSmart buffer (20 mM Tris−acetate, 50 mM potassium acetate, 10 mM magnesium acetate, 100 μg/ mL BSA, pH 7.9) is for KF polymerase and Nb.BtsI, whereas the lambda buffer (67 mM Glycine-KOH, 2.5 mM magnesium chloride, 50 μg/mL BSA, pH 9.4) is for lambda exonuclease. To obtain the optimal reaction buffer, we investigated the detection performance in CutSmart buffer, lambda buffer and the mixture buffer (CutSmart buffer:lambda buffer = 1:1), respectively. As shown in Figure 3B, the highest F/F0 value is obtained in CutSmart buffer, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. Thus, the CutSmart buffer is used in the subsequent research. Notably, this miRNA assay can be simply performed in a one-pot reaction in one tube. We further investigated the influence of reaction time upon the fluorescence signal. As shown in Figure 3C, the fluorescence intensity increases with the reaction time from 0 to 60 min, and levels off at 60 min. Thus, the reaction time of 60 min is used in the subsequent research. The high-concentration 2-aminopurine probe will result in the high hybridization efficiency and consequently the high digestion performance and high fluorescence enhancement, but the high-concentration 2-aminopurine probes are accompanied by the high background signal.29 Therefore, the concentration of 2-aminopurine probe should be optimized as well. Figure 3D shows the variance of the F/F0 value with the concentration of 2-aminopurine probe, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. The F/F0 value increases with the increase of 2aminopurine probe concentration from 50 to 200 nM, but it decreases beyond the concentration of 200 nM. Thus, 200 nM 2-aminopurine probe is used in the subsequent research. Even though the high fluorescence signal can be obtained at high amount of Nb.BtsI, the background signal increases correspondingly due to the unspecific nicking. Therefore, the amount of nicking enzyme Nb.BtsI should be carefully optimized. As shown in Figure 3E, the F/F0 value increases with the increase of Nb.BtsI from 0.5 to 5 U, followed by the decrease beyond the amount of 5 U (F and F0 are the fluorescence intensity in the presence and in the absence of let7a miRNA, respectively). Thus, 5 U Nb.BtsI is used in the subsequent research. Similarly, the amount of lambda exonuclease was optimized as well. Figure 3F shows the variance of F/F0 value with the amount of lambda exonuclease, where F and F0 are the fluorescence intensity in the presence and in the absence of let7a miRNA, respectively. The F/F0 value increases with the increase of lambda exonuclease from 1.25 to 5 U, followed by the decrease beyond the amount of 5 U. Therefore, 5 U lambda exonuclease is used in the subsequent research. Sensitivity and Selectivity of the Proposed Method. To assess the sensitivity of the proposed method, we measured let-7a miRNA at various concentrations under the optimal conditions. Figure 4A shows the variance of fluorescence emission spectra with the concentration of let-7a miRNA. The florescence intensity increases with the increase of let-7a miRNA concentration from 0 to 10 nM. In the logarithmic scales (inset of Figure 4A), the normalized fluorescence intensity has a linear correlation with the concentration of
Figure 4. (A) Variance of the normalized fluorescence intensity as a function of let-7a miRNA concentration. The concentration of let-7a miRNA is 0, 10, 30, 100, 300 pM, 1 nM, 3, and 10 nM from bottom to top, respectively. The inset shows the linear correlation between the logarithm of normalized fluorescence intensity and the logarithm of let-7a miRNA concentration in the range from 10 pM to 10 nM. (B) Selectivity of the proposed method. The concentration of each let-7a, let-7b, let-7c and random sequence is 10 nM. Error bars show the standard deviation of three experiments.
let-7a miRNA over the range from 10 pM to 10 nM. The regression equation is log10 F = −0.334 + 0.317 log10 C with a correlation coefficient of 0.993, where F is the normalized fluorescence intensity and C is the concentration of let-7a miRNA, respectively. The limit of detection is calculated to be 6.5 × 10−12 M (or 0.3 fmol) based on the evaluation of average blank signal plus three times standard deviation. Notably, the sensitivity of the proposed method has improved by as much as 3 orders of magnitude, as compared with that of hairpin ribozyme-based assay (5 nM),30 and more than 2 orders of magnitude, as compared with that of molecular beacon-based method (1 nM),31 and more than 1 order of magnitude, as compared with that of the Q-STAR probe-based RCA assay (0.2 nM).32 The improved sensitivity of the proposed method can be attributed to both the reduced background signal and the recycling signal amplification. A great challenge for miRNA assay is the ability to distinguish the miRNA family members with sequence homology. To evaluate the selectivity of the proposed method, we measured the perfectly matched let-7a miRNA, let-7b miRNA with two-base mismatched, let-7c miRNA with singlebase mismatched and the random sequence, respectively. As shown in Figure 4B, the normalized fluorescence intensity of let-7a is 5.3-fold higher than that of let-7b, and 2.4-fold higher than that of let-7c, even though there is only a single base E
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
lung tissues between the NSCLC patients and the healthy persons, and might be further applied for early clinical diagnosis.
mismatch near the 3′ end of miRNA between let-7a and let-7c (Table 1), suggesting good selectivity of the proposed method for miRNA assay. Detection of let-7a miRNA in the Clinical Sample. Previous research demonstrated that differential expressions of certain miRNAs might be used as a potential predictor of a patient’s overall prognosis.33 Especially, the expression level of let-7a miRNA is frequently reduced in nonsmall cell lung cancer (NSCLC).34,35 To investigate the feasibility of the proposed method in the clinical diagnosis, we quantified the expression of let-7a in total RNA samples extracted from human lung tissues. As shown in Figure 5, the expressions of
■
AUTHOR INFORMATION
Corresponding Author
*C.-y. Zhang. Phone: +86 755 86392211. Fax: +86 755 86392299. E-mail: [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (Grant No. 21325523), the Award for the Hundred Talent Program of the Chinese Academy of Science, and the Fund for Shenzhen Engineering Laboratory of Singlemolecule Detection and Instrument Development (Grant No. (2012) 433).
■
REFERENCES
(1) (a) Kim, V. N.; Han, J.; Siomi, M. C. Nat. Rev. Mol. Cell Biol. 2009, 10, 126−139. (b) Krol, J.; Loedige, I.; Filipowicz, W. Nat. Rev. Genet. 2010, 11, 597−610. (2) Calin, G. A.; Croce, C. M. Nat. Rev. Cancer 2006, 6, 857−866. (3) (a) Bartel, D. P. Cell 2004, 116, 281−297. (b) Bartel, D. P.; Chen, C. Z. Nat. Rev. Genet. 2004, 5, 396−400. (4) (a) Yanaihara, N.; Caplen, N.; Bowman, E.; Seike, M.; Kumamoto, K.; Yi, M.; Stephens, R. M.; Okamoto, A.; Yokota, J.; Tanaka, T.; Calin, G. A.; Liu, C. G.; Croce, C. M.; Harris, C. C. Cancer Cell 2006, 9, 189−198. (b) Pope, C. A.; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Kazuhiko, I.; Thurston, G. D. JAMA, J. Am. Med. Assoc. 2002, 287, 1132−1141. (5) Murakami1, Y.; Yasuda, T.; Saigo, K.; Urashima, T.; Toyoda, H.; Okanoue, T.; Shimotohno, K. Oncogene 2006, 25, 2537−2545. (6) Iorio, M. V.; Ferracin, M.; Liu, C. G.; Veronese, A.; Spizzo, R.; Sabbioni, S.; Magri, E.; Pedriali, M.; Fabbri, M.; Campiglio, M.; Menard, S.; Palazzo, J. P.; Rosenberg, A.; Musiani, P.; Volinia, S.; Nenci, I.; Calin, G. A.; Querzoli, P.; Negrini, M.; Croce, C. M. Cancer Res. 2005, 65, 7065−7070. (7) Tsujiura, M.; Ichikawa, D.; Komatsu, S.; Shiozaki, A.; Takeshita, H.; Kosuga, T.; Konishi, H.; Morimura, R.; Deguchi, K.; Fujiwara, H.; Okamoto, K.; Otsuji, E. Br. J. Cancer 2010, 102, 1174−1179. (8) (a) Raymond, C. K.; Roberts, B. S.; Garrett-Engele, P.; Lim, L. P.; Johnson, J. M. RNA 2005, 11, 1737−1744. (b) Bartels, C. L.; Tsongalis, G. J. Clin. Chem. 2009, 55, 623−631. (9) Jemal, A.; Bray, F.; Center, M. M.; Ferlay, J.; Ward, E.; Forman, D. Ca-Cancer J. Clin. 2011, 61, 69−90. (10) (a) Boeri, M.; Verri, C.; Conte, D.; Roz, L.; Modena, P.; Facchinetti, F.; Calabro, E.; Croce, C. M.; Pastorino, U.; Sozzi, G. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 3713−3718. (b) Chin, L. J.; Ratner, E.; Leng, S.; Zhai, R.; Nallur, S.; Babar, I.; Muller, R. U.; Straka, E.; Su, L.; Burki, E. A.; Crowell, R. E.; Patel, R.; Kulkarni, T.; Homer, R.; Zelterman, D.; Kidd, K. K.; Zhu, Y.; Christiani, D. C.; Belinsky, S. A.; Slack, F. J.; Weidhaas, J. B. Cancer Res. 2008, 68, 8535−8540. (11) Lagos-Quintana, M.; Rauhut, R.; Lendeckel, W.; Tuschl, T. Science 2001, 294, 853−858. (12) (a) Thomson, J. M.; Parker, J.; Perou, C. M.; Hammond, S. M. Nat. Methods 2004, 1, 47−53. (b) Lee, J.; Cho, H.; Jung, Y. Angew. Chem., Int. Ed. 2010, 49, 8662−8665. (13) (a) Chen, C.; Ridzon, D. A.; Broomer, A. J.; Zhou, Z.; Lee, D. H.; Nguyen, J. T.; Barbisin, M.; Xu, N. L.; Mahuvakar, V. R.; Andersen, M. R.; Lao, K. Q.; Livak, K. J.; Guegler, K. J. Nucleic Acids Res. 2005, 33, e179. (b) Li, J.; Yao, B.; Huang, H.; Wang, Z.; Sun, C.; Fan, Y.; Chang, Q.; Li, S.; Wang, X.; Xi, J. Anal. Chem. 2009, 81, 5446−5451. (14) (a) Cheng, Y.; Zhang, X.; Li, Z.; Jiao, X.; Wang, Y.; Zhang, Y. Angew. Chem., Int. Ed. 2009, 48, 3268−3272. (b) Li, Y.; Liang, L.; Zhang, C. Y. Anal. Chem. 2013, 85, 11174−11179.
Figure 5. Analysis of let-7a miRNA in human lung tissues. The bars represent the content of let-7a in total RNA from ten healthy persons and ten NSCLC patients measured by the proposed method (red bars) and qPCR (green bars), respectively. Error bars show the standard deviation of three experiments.
let-7a miRNA obtained from ten NSCLC patients are significantly lower than that obtained from ten healthy persons (unpaired t test, P < 0.0001). Notably, the mean content of let7a from ten NSCLC patients (1.28 × 109 copies/μg) is much lower than that from ten healthy persons (5.57 × 109 copies/ μg), which is consistent with previous researches about the reduced expression of let-7a in NSCLC patients.36,37 In addition, our results (Figure 5, red bars) are consistent with those obtained by the standard qPCR method (Figure 5, green bars), further confirming the accuracy of our method. These results suggest that the proposed method can directly distinguish the expression of let-7a miRNA among NSCLC patients and healthy persons with excellent accuracy, thus holding a great potential for further application in clinical diagnosis.
■
CONCLUSION In summary, we have developed a quencher-free fluorescent method for homogeneously sensitive detection of let-7a miRNA using the target-triggered recycling signal amplification in combination with the 2-aminopurine probe. In contrast to the conventional molecular beacon, the 2-aminopurine probe is quenched through its stacking interaction with the adjacent bases24b without the involvement of any extra quenchers. In addition, the introduction of a helper DNA into 2-aminopurine probe significantly improves the signal-to-noise ratio. The proposed method allows for sensitive detection of let-7a miRNA with a detection limit of as low as 0.3 fmol, and it can even discriminate the single-nucleotide difference among the miRNA family members. More importantly, this method can accurately distinguish the expression of let-7a miRNA in human F
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
(15) (a) Jia, H.; Li, Z.; Liu, C.; Cheng, Y. Angew. Chem., Int. Ed. 2010, 49, 5498−5501. (b) Zhang, Y.; Zhang, C. Y. Anal. Chem. 2012, 84, 224−231. (c) Wang, G. L.; Zhang, C. Y. Anal. Chem. 2012, 84, 7037− 7042. (16) Duan, R.; Zuo, X.; Wang, S.; Quan, X.; Chen, D.; Chen, Z.; Jiang, L.; Fan, C.; Xia, F. J. Am. Chem. Soc. 2013, 135, 4604−4607. (17) (a) Gao, Z.; Yang, Z. Anal. Chem. 2006, 78, 1470−1477. (b) Wang, J.; Yi, X.; Tang, H.; Han, H.; Wu, M.; Zhou, F. Anal. Chem. 2012, 84, 6400−6406. (c) Fang, S.; Lee, H. J.; Wark, A. W.; Corn, R. M. J. Am. Chem. Soc. 2006, 128, 14044−14046. (18) Shi, C.; Liu, Q.; Ma, C.; Zhong, W. Anal. Chem. 2014, 86, 336− 339. (19) Liu, H.; Li, L.; Duan, L.; Wang, X.; Xie, Y.; Tong, L.; Wang, Q.; Tang, B. Anal. Chem. 2013, 85, 7941−7947. (20) Zhang, P.; Zhang, J.; Wang, C.; Liu, C.; Wang, H.; Li, Z. Anal. Chem. 2014, 86, 1076−1082. (21) Cheng, Y.; Lei, J.; Chen, Y.; Ju, H. Biosens. Bioelectron. 2014, 51, 431−436. (22) Ren, Y.; Deng, H.; Shen, W.; Gao, Z. Anal. Chem. 2013, 85, 4784−4789. (23) Ryoo, S. R.; Lee, J.; Yeo, J.; Na, H. K.; Kim, Y. K.; Jang, H.; Lee, J. H.; Han, S. W.; Lee, Y.; Kim, V. N.; Min, D. H. ACS Nano 2013, 7, 5882−5891. (24) (a) Sowers, L. C.; Fazakerly, G.. V.; Eritja, R.; Kaplan, B. E.; Goodman, M. F. Proc. Natl. Acad. Sci. U. S. A. 1986, 83, 5434−5438. (b) Jean, J. M.; Hall, K. B. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 37− 41. (25) Mitsis, P. G.; Kwagh, J. G. Nucleic Acids Res. 1999, 27, 3057− 3063. (26) Xu, Q.; Cao, A.; Zhang, L.; Zhang, C. Y. Anal. Chem. 2012, 84, 10845−10851. (27) Marti, A. A.; Jockusch, S.; Li, Z.; Ju, J.; Turro, N. J. Nucleic Acids Res. 2006, 34, e50. (28) (a) Eritja, R.; Kaplan, B. E.; Mhaskar, D.; Sowers, L. C.; Petruska, J.; Goodman, M. F. Nucleic Acids Res. 1986, 14, 5869−5885. (b) Xu, D. G.; Evans, K. O.; Nordlund, T. M. Biochemistry 1994, 33, 9592−9599. (29) Feng, K. J.; Kong, R. M.; Wang, H.; Zhang, S. F.; Qu, F. L. Biosens. Bioelectron. 2012, 38, 121−125. (30) Hartig, J. S.; Grune, I.; Najafi-Shoushtari, S. H.; Famulok, M. J. Am. Chem. Soc. 2004, 126, 722−726. (31) Baker, M. B.; Bao, G.; Searles, C. D. Nucleic Acids Res. 2012, 40, e13. (32) Harcourt, E. M.; Kool, E. T. Nucleic Acids Res. 2012, 40, e65. (33) Esquela-Kerscher, A.; Slack, F. J. Nat. Rev. Cancer 2006, 6, 259− 269. (34) Kumar, M. S.; Erkeland, S. J.; Pester, R. E.; Chen, C. Y.; Ebert, M. S.; Sharp, P. A.; Jacks, T. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 3903−3908. (35) Johnson, S. M.; Grosshans, H.; Shingara, J.; Byrom, M.; Jarvis, R.; Cheng, A.; Labourier, E.; Reinert, K. L.; Brown, D.; Slack, F. J. Cell 2005, 120, 635−647. (36) Takamizawa, J.; Konishi, H.; Yanagisawa, K.; Tomida, S.; Osada, H.; Endoh, H.; Harano, T.; Yatabe, Y.; Nagino, M.; Nimura, Y.; Mitsudomi, T.; Takahashi, T. Cancer Res. 2004, 64, 3753−3756. (37) Karube, Y.; Tanaka, H.; Osada, H.; Tomida, S.; Tatematsu, Y.; Yanagisawa, K.; Yatabe, Y.; Takamizawa, J.; Miyoshi, S.; Mitsudomi, T.; Takahashi, T. Cancer Sci. 2005, 96, 111−115.
G
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Quencher-Free Fluorescent Method for Homogeneously Sensitive Detection of MicroRNAs in Human Lung Tissues Guichi Zhu,† Li Liang,‡ and Chun-yang Zhang*,† †
Single-Molecule Detection and Imaging Laboratory, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, Guangdong 518055, China ‡ Department of Tumor Chemotherapy and Radiation Sickness, Peking University Third Hospital, Beijing 100191, China ABSTRACT: Sensitive and accurate analysis of microRNA (miRNA) expression is imperative for understanding the biological functions of miRNAs and the early diagnosis of human cancers. Here, we develop a quencher-free fluorescent method for homogeneously sensitive detection of let-7a miRNA using the target-triggered recycling signal amplification in combination with a 2-aminopurine probe. The 2-aminopurine probe is characterized by the substitution of 2-aminopurine for adenine in the DNA strand and the quenching of 2-aminopurine fluorescence through its stacking interaction with the adjacent bases. The binding of target miRNA with the 2-aminopurine probe initiates the extension reaction in the presence of polymerase to produce the DNA duplexes. These DNA duplexes can be further cleaved by lambda exonuclease through the recycling digestion to release abundant free 2-aminopurines, leading to an enhanced fluorescence signal. The proposed method exhibits high sensitivity with a detection limit of 0.3 fmol, and it can even discriminate the single-base difference among the miRNA family members. More importantly, this method can accurately distinguish the expression of let-7a miRNA in human lung tissues between ten non small cell lung cancer (NSCLC) patients and ten healthy persons, holding a great potential for further application in early clinical diagnosis.
M
from the low sensitivity and the poor specificity because of the small size, the sequence homology among the family members and the low abundance of miRNAs. To improve the sensitivity and the specificity of miRNA assay, a variety of amplification strategies have been introduced, such as real-time polymerase chain reaction (RT-PCR),13 rolling circle amplification (RCA),14 exponential amplification reaction (EXPAR),15 quadratic isothermal amplification16 and nanoparticle-based signal amplification.17 Among these approaches, the RT-PCR method has gained increasing attention because of its high sensitivity and practicality, but it involves the precise control of temperature cycling and the sophistication of primer design.13 The RCA amplification is a simple enzymatic process to generate long single-stranded DNAs under isothermal conditions, but it is usually time-consuming with the total reaction process duration of 4−8 h.14 Alternatively, the EXPAR method can provide >106-fold amplification within minutes, and can obtain a detection limit at the attomolar level for miRNA assay.15a,b However, this method might be hindered by nonspecific background amplification. The quadratic isothermal amplification method might achieve single-cell sensitivity for miRNA assay,16 but the involvement of doubly labeled fluorescent probe (e.g., molecular beacon) might increase the experimental cost and the complexity of probe design. The nanoparticles possess unique electronic, optical and catalytic
icroRNAs (miRNAs) are short endogenous noncoding RNA molecules with lengths of about 22 nucleotides, which are derived from long primary transcripts by the catalysis of Dicer enzyme.1 More than 450 kinds of miRNAs have been identified in the human genome,2 and most of them play a crucial role in various biological processes including cell development, differentiation, apoptosis and proliferation through imperfect pairing with the target mRNA of proteincoding genes and the transcriptional regulation of their expression.3 Recent research demonstrates that the aberrant expression of miRNAs is associated with many human carcinoma such as lung cancer,4 hepatocellular carcinoma,5 breast cancer6 and gastric carcinoma,7 suggesting that miRNAs might be used as valuable biomarkers for early diagnosis of cancers in asymptomatic individuals.8 Among human cancers, lung cancer is the leading cause of cancer-related death in the world, and its etiology is primarily genetic and epigenetic damage caused by cigarette smoking, air pollution and longterm exposure to the environments containing carcinogenic substances.4 Lung cancer has two major subtypes including small cell lung cancer (SCLC) and nonsmall cell lung cancer (NSCLC), and NSCLC accounts for approximately 80% in all cases.9 Much emerging evidence suggests that miRNAs might function as potential oncogenes or tumor suppressors in lung cancer.10 Consequently, accurate and quantitative analysis of miRNA expression is critical for further understanding the biological functions of miRNAs and the early diagnosis of lung cancer. The Northern blotting11 and microarray-based methods12 are usually used for miRNA detection, but they might suffer © XXXX American Chemical Society
Received: September 7, 2014 Accepted: October 30, 2014
A
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
properties, making them ideal candidates for signal generation and amplification in miRNA assay.17 Some typical examples include the use of electrocatalytic nanoparticle tags to chemically amplify the signal,17a the use ferrocene-capped gold nanoparticle/streptavidin conjugates for amplified voltammetric detection17b and the use of Au nanoparticle for amplified surface plasmon resonance imaging.17c Nevertheless, these methods usually involve the time-consuming synthesis of nanoparticles and the complicated chemical modification of substrates. Recently, several new methods such as strand displacement amplification,18 padlock probe-based exponential RCA assay,19 size-coded ligation chain reaction,20 electrochemical biosensor21,22 and peptide nucleic acid−graphene oxide-based methods23 have been developed for the detection of miRNAs in both the spiked samples and the cell samples. However, the development of simple and highly sensitive methods with the capability of clinical application still remains a great challenge. The nucleotide of 2-aminopurine is a fluorescent analog of adenine, and is usually used as a site-specific probe for the study of nucleic acid structure and dynamics.24 The 2-aminopurine exhibits a weak fluorescence when being incorporated into the DNA strands, but an enhanced fluorescence when being free in the solution.25 In contrast to the conventional molecular beacon, the quenching of 2-aminopurine fluorescence results from its stacking interaction with the adjacent bases24b without the involvement of any extra quenchers. Moreover, 2aminopurine is insensitive to the visible light because its excitation wavelength locates at 310 nm.26 Here, we develop a quencher-free fluorescent method for homogeneously sensitive detection of miRNAs using the target-triggered recycling signal amplification in combination with 2-aminopurine probe. The proposed method allows for sensitive detection of miRNAs with a detection limit of as low as 0.3 fmol, and it can discriminate the single-nucleotide difference among miRNA family members. Importantly, this method can accurately distinguish the expression of let-7a miRNA in human lung tissues between ten NSCLC patients and ten healthy persons, holding great potential for further applications in early clinical diagnosis.
Table 1. Sequences of 2-Aminopurine Probes, Helper DNAs and miRNAsa note 2-aminopurine probe-a 2-aminopurine probe-b helper DNA-1 helper DNA-2 helper DNA-3 let-7a let-7b let-7c
Sequence (5′-3′) AAA ACT↓CAC TGC ACT CGA TCG GAA CTA TAC AAC CTA CTA CCT CA-NH2 AAA ACT↓CAC TGC ACT CGA TCG GAA CTA TAC AAC CTA CTA CCT CA-NH2 AAG ATC GAG TGC C-NH2 AAC GAT CGA GTG CAC C-NH2 AAC CGA TCG AGT GCA GTC C-NH2 UGA GGU AGU AGG UUG UAU AGU U UGA GGU AGU AGG UUG UGU GGU U UGA GGU AGU AGG UUG UAU GGU U
a
The bold regions of 2-aminopurine probes and helper DNAs indicate the phosphorothioate modification. The arrow in 2-aminopurine probes indicates the nicking position of Nb.BtsI. The bold italic region of 2-aminopurine probes indicates the binding sequence of helper DNA-2. The italic region of helper DNAs indicates the binding sequence of 2-aminopurine probes. The underlined region of 2aminopurine probes indicates the 2-aminopurine substitution. The bases in let-7b and let-7c that differ from those in let-7a are marked in bold. There are two base mismatches near the 3′ end of miRNAs between let-7a and let-7b, and one base mismatch near the 3′ end of miRNAs between let-7a and let-7c.
KF polymerase, 5 U Nb.BtsI, 5 U lambda exonuclease, 40 U RNase inhibitor, 1× NEB CutSmart buffer (20 mM Tris− acetate, 50 mM potassium acetate, 10 mM magnesium acetate, 0.1 mg/mL BSA, pH 7.9), DEPC treated water, let-7a miRNA at different concentrations, and incubated at 37 °C for 60 min. At last, the resultant reaction solution was subjected to the fluorescence measurement using a Hitachi F-4500 fluorometer (Tokyo, Japan) equipped with a xenon lamp as the excitation source. The spectra were recorded in the range from 345 to 500 nm at an excitation wavelength of 310 nm. The excitation and emission slits were set for 5.0 and 5.0 nm, respectively. The fluorescence signals were normalized, and the intensity at the emission wavelength of 365 nm was used for data analysis. Gel Electrophoresis. A 10% nondenaturating polyacrylamide gel electrophoresis (PAGE) analysis of reaction products was carried out in 1× TBE (9 mM Tris−HCl, pH 7.9, 9 mM boric acid, 0.2 mM ethylenediaminetetraacetic acid, EDTA) at a 120 V constant voltage for 35 min at room temperature. The gel was stained by SYBR Green I, and scanned by a Kodak Image Station 4000 MM (Woodbridge, CT, USA). Clinical Sample Analysis. The total RNA was extracted from the human lung tissues of NSCLC patients and healthy persons using the miRNeasy FFPE kit (Qiagen, Germany) according to the manufacturer’s procedures, and the concentrations of total RNA were determined by measuring the absorbance at 260 nm with a spectrophotometer. The extracted total RNA concentrations were diluted to 200 ng/μL with DEPC-treated water (RNase free), and 20 μL of total RNA sample (4 μg in total) was added to each 50 μL of reaction solution for measurement.
■
EXPERIMENTAL SECTION Materials. The 2-aminopurine probes, helper DNAs, miRNAs of let-7a, let-7b and let-7c, deoxynucleotide solution mixture (dNTPs), recombinant RNase inhibitor and diethylpyrocarbonate (DEPC) treated water (RNase free) were purchased from TaKaRa Biotechnology Co., Ltd. (Dalian, China). The sequences of oligonucleotides are listed in Table 1. The Klenow fragment polymerase (3′ → 5′exo-, KF polymerase), the nicking enzymes of Nb.BtsI, the lambda exonuclease, NEB CutSmart buffer and lambda buffer were obtained from New England BioLabs (Beverly, MA, USA). SYBR Green I was obtained from Xiamen Bio-Vision Biotechnology Co. Ltd. (Xiamen, China). The miRNA qRT-PCR kit was purchased from Life Technologies (Carlsbad, CA, USA). The paraffinembedding lung tissue samples of NSCLC patients and healthy persons were obtained from the Peking University Third Hospital (Beijing, China). All other chemical reagents were of analytical grade and obtained from Sigma-Aldrich (St. Louis, MO, USA). Detection of let-7a miRNA. The experiments were performed in 50 μL of solution consisting of 200 nM 2aminopurine probe, 200 nM helper DNA, 300 μM dNTPs, 5 U
■
RESULTS AND DISCUSSION Principle of miRNA Assay. The principle of miRNA assay is illustrated in Scheme 1. We designed a 2-aminopurine probe with two substitutions of 2-aminopurine molecules for adenines in a single-stranded DNA (ssDNA). Because the fluorescence of 2-aminopurine in double-stranded DNA (dsDNA) is much lower than that in single-stranded DNA (ssDNA),27 we further
B
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Scheme 1. Schematic Illustration of miRNA Assay Using Target-Triggered Recycling Signal Amplification in Combination with 2-Aminopurine Probe
introduced a helper DNA, which can hybridize with the 2aminopurine probe, to reduce the background fluorescence signal. In the presence of target miRNA, the amplification reaction is initiated with the assistance of KF polymerase, and the helper DNA is replaced by the target extension, producing a complete dsDNA with the recognition sites for nicking enzyme of Nb.BtsI. The subsequent cleavage of 2-aminopurine probe by Nb.BtsI exposes the site of phosphate group (PO4) for the lambda exonuclease digestion, releasing both the free 2aminopurine molecules and the extended targets. The extended target can further hybridize with new 2-aminopurine probe repeatedly, generating the cycle of nicking−digestion−hybridization and consequently the production of abundant free 2aminopurine molecules for significant fluorescence enhancement. While in the absence of target miRNA, the helper DNA is unable to be replaced, and the recognition site of nicking enzyme in 2-aminopurine probe is unavailable. As a result, the cycle of nicking−digestion−hybridization cannot be initiated, and no fluorescence enhancement is observed. To demonstrate the feasibility of the proposed method for miRNA assay, we performed both fluorescence measurement (Figure 1A) and nondenaturating PAGE analysis (Figure 1B). In the presence of let-7a miRNA target and three kinds of enzymes (KF polymerase, Nb.BtsI nicking enzyme and lambda exonuclease), a strong fluorescence signal with the characteristic emission peak of 365 nm is observed (Figure 1A), indicating the release of abundant free 2-aminopurine molecules due to the cycle of nicking−digestion−hybridization. However, no significant fluorescence signal is observed in the absence of either let-7a miRNA or one kind of enzyme (Figure 1A), indicating that the cycle of nicking−digestion−hybridization cannot be initiated. The above results are further confirmed by the nondenaturating PAGE analysis (Figure 1B). Only a weak DNA band is observed in the presence of let-7a miRNA and three kinds of enzymes (Figure 1B, lane 1), indicating the digestion of 2-aminopurine probe by lambda exonuclease due to the cycle of nicking−digestion−hybridization. In contrast, a distinct and strong DNA band is observed in the absence of either let-7a miRNA or one kind of enzyme (Figure 1B, lanes 2−5), indicating that the cycle of nicking− digestion−hybridization cannot be initiated.
Figure 1. (A) Normalized fluorescence emission spectra of reaction products in the presence of let-7a miRNA + three enzymes (red line), control + three enzymes (black line), let-7a miRNA + KF polymerase + Nb.BtsI (blue line), let-7a miRNA + KF polymerase + lambda exonuclease (green line) and let-7a miRNA + Nb.BtsI + lambda exonuclease (purple line), respectively. The concentration of the let-7a miRNA target is 10 nM. (B) Nondenaturating PAGE analysis of reaction products. Lane m is the DNA ladder marker; Lanes 1, 2, 3, 4, and 5 represent let-7a miRNA + three enzymes, control + three enzymes, let-7a miRNA + KF polymerase + Nb.BtsI, let-7a miRNA + KF polymerase + lambda exonuclease, and let-7a miRNA + Nb.BtsI + lambda exonuclease, respectively.
The quenching of 2-aminopurine fluorescence results from its stacking interaction with the adjacent bases,24b and the fluorescence of 2-aminopurine in dsDNA is much lower than that in ssDNA because of the more efficient stacking interaction among the bases in dsDNA.27 To reduce the background fluorescence signal by keeping 2-aminopurine molecules in dsDNA, we introduced a helper DNA to hybridize with the 2aminopurine probe. In both the helper DNA and the 2aminopurine probe, the 5′ end is modified with phosphorothioate to prevent the nonspecific digestion of lambda exonuclease, and the 3′ end is modified with amination to prevent the nonspecific extension of KF polymerase (Table 1). In addition, the length of helper DNA is crucial to the cycle of nicking−digestion−hybridization. When the matched bases between the helper DNA and the 2-aminopurine probe are too short, the DNA duplexes are unstable at the reaction temperature. When the matched bases are too long, the helper DNA is hard to be replaced by the extended targets. To select the proper helper DNA, three helper DNAs with nine matched bases (helper DNA-1), twelve matched bases (helper DNA-2) and fifteen matched bases (helper DNA-3), respectively, were investigated in the presence of 10 nM let-7a miRNA. As shown in Figure 2A, the F/F0 value in the presence of helper DNA-2 is much higher than those in the presence of helper DNA-1 and helper DNA-3, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. Therefore, the helper DNA-2 is the best choice experimentally, C
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
formed by the hybridization of helper DNA-2 with its complementary sequence (48.2 °C) is about 10 °C higher than the reaction temperature of 37 °C, which can not only keep the DNA duplex stable but also ensure the proceeding of subsequent replacement reaction. It should be noted that the introduction of helper DNA does not influence the digestion of 2-aminopurine probe by Nb.BtsI nicking enzyme and the subsequent cycle of nicking−digestion−hybridization. In the presence of let-7a miRNA, no significant difference is observed in the obtained fluorescence signals no matter the helper DNA is present or not (Figure 2B). Nevertheless, the helper DNA induces a significant decrease in the background signal. The signal-to-noise ratio in the presence of helper DNA has increased by as much as 2.5-fold, as compared with that in the absence of helper DNA, greatly benefiting the improvement of detection sensitivity. Optimization of Experimental Condition. To optimize the design of the 2-aminopurine probe, two 2-aminopurine probes (Table 1) including one with two substitutions of 2aminopurine (probe-a) and another with one substitution of 2aminopurine (probe-b) were investigated. As shown in Figure 3A, the F/F0 value of probe-a group is higher than that of probe-b group, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. This result indicates that probe-a with two substitutions of 2aminopurine has higher signal-to-noise ratio than probe-b with one substitution of 2-aminopurine. It should be noted that the mismatched base pair of 2-aminopurine:thymine has lower stability than the natural base pair of adenine:thymine,28 which might lead to the unstable thermodynamic performance and the poor hybridization efficiency for the probe with too many substitutions of 2-aminopurine. Taking into account the design complexity and the synthesis cost of 2-aminopurine probe as well as the hybridization efficiency, probe-a with two
Figure 2. (A) Optimization of helper DNA. (B) Influence of helper DNA on the signal-to-noise ratio. The concentration of let-7a miRNA is 10 nM. Error bars show the standard deviation of three experiments.
consistent with the theoretical calculation as well. The melting temperature (Tm) of DNA duplex formed by the hybridization of helper DNA-1, helper DNA-2 and helper DNA-3 with its complementary sequence is calculated to be 27.6, 48.2 and 58.5 °C, respectively. In theory, the Tm value of DNA duplex
Figure 3. (A) Variance of fluorescence value ratio of F/F0 with the design of 2-aminopurine probe. Probe-a is a 2-aminopurine probe with two substitutions of 2-aminopurine, and probe-b is a 2-aminopurine probe with one substitution of 2-aminopurine. (B) Variance of fluorescence value ratio of F/F0 with different buffers. (C) Variance of the normalized fluorescence intensity with the reaction time. (D) Variance of fluorescence value ratio of F/F0 with the concentration of 2-aminopurine probe. (E) Variance of fluorescence value ratio of F/F0 with the amount of nicking enzyme Nb.BtsI. (F) Variance of fluorescence value ratio of F/F0 with the amount of lambda exonuclease. F and F0 are the fluorescence signals in the presence and in the absence of 10 nM let-7a miRNA, respectively. Error bars show the standard deviation of three experiments. D
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
substitutions of 2-aminopurine is used in the subsequent research. This research involves three kinds of enzymes and two different buffers. The CutSmart buffer (20 mM Tris−acetate, 50 mM potassium acetate, 10 mM magnesium acetate, 100 μg/ mL BSA, pH 7.9) is for KF polymerase and Nb.BtsI, whereas the lambda buffer (67 mM Glycine-KOH, 2.5 mM magnesium chloride, 50 μg/mL BSA, pH 9.4) is for lambda exonuclease. To obtain the optimal reaction buffer, we investigated the detection performance in CutSmart buffer, lambda buffer and the mixture buffer (CutSmart buffer:lambda buffer = 1:1), respectively. As shown in Figure 3B, the highest F/F0 value is obtained in CutSmart buffer, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. Thus, the CutSmart buffer is used in the subsequent research. Notably, this miRNA assay can be simply performed in a one-pot reaction in one tube. We further investigated the influence of reaction time upon the fluorescence signal. As shown in Figure 3C, the fluorescence intensity increases with the reaction time from 0 to 60 min, and levels off at 60 min. Thus, the reaction time of 60 min is used in the subsequent research. The high-concentration 2-aminopurine probe will result in the high hybridization efficiency and consequently the high digestion performance and high fluorescence enhancement, but the high-concentration 2-aminopurine probes are accompanied by the high background signal.29 Therefore, the concentration of 2-aminopurine probe should be optimized as well. Figure 3D shows the variance of the F/F0 value with the concentration of 2-aminopurine probe, where F and F0 are the fluorescence intensity in the presence and in the absence of let-7a miRNA, respectively. The F/F0 value increases with the increase of 2aminopurine probe concentration from 50 to 200 nM, but it decreases beyond the concentration of 200 nM. Thus, 200 nM 2-aminopurine probe is used in the subsequent research. Even though the high fluorescence signal can be obtained at high amount of Nb.BtsI, the background signal increases correspondingly due to the unspecific nicking. Therefore, the amount of nicking enzyme Nb.BtsI should be carefully optimized. As shown in Figure 3E, the F/F0 value increases with the increase of Nb.BtsI from 0.5 to 5 U, followed by the decrease beyond the amount of 5 U (F and F0 are the fluorescence intensity in the presence and in the absence of let7a miRNA, respectively). Thus, 5 U Nb.BtsI is used in the subsequent research. Similarly, the amount of lambda exonuclease was optimized as well. Figure 3F shows the variance of F/F0 value with the amount of lambda exonuclease, where F and F0 are the fluorescence intensity in the presence and in the absence of let7a miRNA, respectively. The F/F0 value increases with the increase of lambda exonuclease from 1.25 to 5 U, followed by the decrease beyond the amount of 5 U. Therefore, 5 U lambda exonuclease is used in the subsequent research. Sensitivity and Selectivity of the Proposed Method. To assess the sensitivity of the proposed method, we measured let-7a miRNA at various concentrations under the optimal conditions. Figure 4A shows the variance of fluorescence emission spectra with the concentration of let-7a miRNA. The florescence intensity increases with the increase of let-7a miRNA concentration from 0 to 10 nM. In the logarithmic scales (inset of Figure 4A), the normalized fluorescence intensity has a linear correlation with the concentration of
Figure 4. (A) Variance of the normalized fluorescence intensity as a function of let-7a miRNA concentration. The concentration of let-7a miRNA is 0, 10, 30, 100, 300 pM, 1 nM, 3, and 10 nM from bottom to top, respectively. The inset shows the linear correlation between the logarithm of normalized fluorescence intensity and the logarithm of let-7a miRNA concentration in the range from 10 pM to 10 nM. (B) Selectivity of the proposed method. The concentration of each let-7a, let-7b, let-7c and random sequence is 10 nM. Error bars show the standard deviation of three experiments.
let-7a miRNA over the range from 10 pM to 10 nM. The regression equation is log10 F = −0.334 + 0.317 log10 C with a correlation coefficient of 0.993, where F is the normalized fluorescence intensity and C is the concentration of let-7a miRNA, respectively. The limit of detection is calculated to be 6.5 × 10−12 M (or 0.3 fmol) based on the evaluation of average blank signal plus three times standard deviation. Notably, the sensitivity of the proposed method has improved by as much as 3 orders of magnitude, as compared with that of hairpin ribozyme-based assay (5 nM),30 and more than 2 orders of magnitude, as compared with that of molecular beacon-based method (1 nM),31 and more than 1 order of magnitude, as compared with that of the Q-STAR probe-based RCA assay (0.2 nM).32 The improved sensitivity of the proposed method can be attributed to both the reduced background signal and the recycling signal amplification. A great challenge for miRNA assay is the ability to distinguish the miRNA family members with sequence homology. To evaluate the selectivity of the proposed method, we measured the perfectly matched let-7a miRNA, let-7b miRNA with two-base mismatched, let-7c miRNA with singlebase mismatched and the random sequence, respectively. As shown in Figure 4B, the normalized fluorescence intensity of let-7a is 5.3-fold higher than that of let-7b, and 2.4-fold higher than that of let-7c, even though there is only a single base E
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
lung tissues between the NSCLC patients and the healthy persons, and might be further applied for early clinical diagnosis.
mismatch near the 3′ end of miRNA between let-7a and let-7c (Table 1), suggesting good selectivity of the proposed method for miRNA assay. Detection of let-7a miRNA in the Clinical Sample. Previous research demonstrated that differential expressions of certain miRNAs might be used as a potential predictor of a patient’s overall prognosis.33 Especially, the expression level of let-7a miRNA is frequently reduced in nonsmall cell lung cancer (NSCLC).34,35 To investigate the feasibility of the proposed method in the clinical diagnosis, we quantified the expression of let-7a in total RNA samples extracted from human lung tissues. As shown in Figure 5, the expressions of
■
AUTHOR INFORMATION
Corresponding Author
*C.-y. Zhang. Phone: +86 755 86392211. Fax: +86 755 86392299. E-mail: [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (Grant No. 21325523), the Award for the Hundred Talent Program of the Chinese Academy of Science, and the Fund for Shenzhen Engineering Laboratory of Singlemolecule Detection and Instrument Development (Grant No. (2012) 433).
■
REFERENCES
(1) (a) Kim, V. N.; Han, J.; Siomi, M. C. Nat. Rev. Mol. Cell Biol. 2009, 10, 126−139. (b) Krol, J.; Loedige, I.; Filipowicz, W. Nat. Rev. Genet. 2010, 11, 597−610. (2) Calin, G. A.; Croce, C. M. Nat. Rev. Cancer 2006, 6, 857−866. (3) (a) Bartel, D. P. Cell 2004, 116, 281−297. (b) Bartel, D. P.; Chen, C. Z. Nat. Rev. Genet. 2004, 5, 396−400. (4) (a) Yanaihara, N.; Caplen, N.; Bowman, E.; Seike, M.; Kumamoto, K.; Yi, M.; Stephens, R. M.; Okamoto, A.; Yokota, J.; Tanaka, T.; Calin, G. A.; Liu, C. G.; Croce, C. M.; Harris, C. C. Cancer Cell 2006, 9, 189−198. (b) Pope, C. A.; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Kazuhiko, I.; Thurston, G. D. JAMA, J. Am. Med. Assoc. 2002, 287, 1132−1141. (5) Murakami1, Y.; Yasuda, T.; Saigo, K.; Urashima, T.; Toyoda, H.; Okanoue, T.; Shimotohno, K. Oncogene 2006, 25, 2537−2545. (6) Iorio, M. V.; Ferracin, M.; Liu, C. G.; Veronese, A.; Spizzo, R.; Sabbioni, S.; Magri, E.; Pedriali, M.; Fabbri, M.; Campiglio, M.; Menard, S.; Palazzo, J. P.; Rosenberg, A.; Musiani, P.; Volinia, S.; Nenci, I.; Calin, G. A.; Querzoli, P.; Negrini, M.; Croce, C. M. Cancer Res. 2005, 65, 7065−7070. (7) Tsujiura, M.; Ichikawa, D.; Komatsu, S.; Shiozaki, A.; Takeshita, H.; Kosuga, T.; Konishi, H.; Morimura, R.; Deguchi, K.; Fujiwara, H.; Okamoto, K.; Otsuji, E. Br. J. Cancer 2010, 102, 1174−1179. (8) (a) Raymond, C. K.; Roberts, B. S.; Garrett-Engele, P.; Lim, L. P.; Johnson, J. M. RNA 2005, 11, 1737−1744. (b) Bartels, C. L.; Tsongalis, G. J. Clin. Chem. 2009, 55, 623−631. (9) Jemal, A.; Bray, F.; Center, M. M.; Ferlay, J.; Ward, E.; Forman, D. Ca-Cancer J. Clin. 2011, 61, 69−90. (10) (a) Boeri, M.; Verri, C.; Conte, D.; Roz, L.; Modena, P.; Facchinetti, F.; Calabro, E.; Croce, C. M.; Pastorino, U.; Sozzi, G. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 3713−3718. (b) Chin, L. J.; Ratner, E.; Leng, S.; Zhai, R.; Nallur, S.; Babar, I.; Muller, R. U.; Straka, E.; Su, L.; Burki, E. A.; Crowell, R. E.; Patel, R.; Kulkarni, T.; Homer, R.; Zelterman, D.; Kidd, K. K.; Zhu, Y.; Christiani, D. C.; Belinsky, S. A.; Slack, F. J.; Weidhaas, J. B. Cancer Res. 2008, 68, 8535−8540. (11) Lagos-Quintana, M.; Rauhut, R.; Lendeckel, W.; Tuschl, T. Science 2001, 294, 853−858. (12) (a) Thomson, J. M.; Parker, J.; Perou, C. M.; Hammond, S. M. Nat. Methods 2004, 1, 47−53. (b) Lee, J.; Cho, H.; Jung, Y. Angew. Chem., Int. Ed. 2010, 49, 8662−8665. (13) (a) Chen, C.; Ridzon, D. A.; Broomer, A. J.; Zhou, Z.; Lee, D. H.; Nguyen, J. T.; Barbisin, M.; Xu, N. L.; Mahuvakar, V. R.; Andersen, M. R.; Lao, K. Q.; Livak, K. J.; Guegler, K. J. Nucleic Acids Res. 2005, 33, e179. (b) Li, J.; Yao, B.; Huang, H.; Wang, Z.; Sun, C.; Fan, Y.; Chang, Q.; Li, S.; Wang, X.; Xi, J. Anal. Chem. 2009, 81, 5446−5451. (14) (a) Cheng, Y.; Zhang, X.; Li, Z.; Jiao, X.; Wang, Y.; Zhang, Y. Angew. Chem., Int. Ed. 2009, 48, 3268−3272. (b) Li, Y.; Liang, L.; Zhang, C. Y. Anal. Chem. 2013, 85, 11174−11179.
Figure 5. Analysis of let-7a miRNA in human lung tissues. The bars represent the content of let-7a in total RNA from ten healthy persons and ten NSCLC patients measured by the proposed method (red bars) and qPCR (green bars), respectively. Error bars show the standard deviation of three experiments.
let-7a miRNA obtained from ten NSCLC patients are significantly lower than that obtained from ten healthy persons (unpaired t test, P < 0.0001). Notably, the mean content of let7a from ten NSCLC patients (1.28 × 109 copies/μg) is much lower than that from ten healthy persons (5.57 × 109 copies/ μg), which is consistent with previous researches about the reduced expression of let-7a in NSCLC patients.36,37 In addition, our results (Figure 5, red bars) are consistent with those obtained by the standard qPCR method (Figure 5, green bars), further confirming the accuracy of our method. These results suggest that the proposed method can directly distinguish the expression of let-7a miRNA among NSCLC patients and healthy persons with excellent accuracy, thus holding a great potential for further application in clinical diagnosis.
■
CONCLUSION In summary, we have developed a quencher-free fluorescent method for homogeneously sensitive detection of let-7a miRNA using the target-triggered recycling signal amplification in combination with the 2-aminopurine probe. In contrast to the conventional molecular beacon, the 2-aminopurine probe is quenched through its stacking interaction with the adjacent bases24b without the involvement of any extra quenchers. In addition, the introduction of a helper DNA into 2-aminopurine probe significantly improves the signal-to-noise ratio. The proposed method allows for sensitive detection of let-7a miRNA with a detection limit of as low as 0.3 fmol, and it can even discriminate the single-nucleotide difference among the miRNA family members. More importantly, this method can accurately distinguish the expression of let-7a miRNA in human F
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
(15) (a) Jia, H.; Li, Z.; Liu, C.; Cheng, Y. Angew. Chem., Int. Ed. 2010, 49, 5498−5501. (b) Zhang, Y.; Zhang, C. Y. Anal. Chem. 2012, 84, 224−231. (c) Wang, G. L.; Zhang, C. Y. Anal. Chem. 2012, 84, 7037− 7042. (16) Duan, R.; Zuo, X.; Wang, S.; Quan, X.; Chen, D.; Chen, Z.; Jiang, L.; Fan, C.; Xia, F. J. Am. Chem. Soc. 2013, 135, 4604−4607. (17) (a) Gao, Z.; Yang, Z. Anal. Chem. 2006, 78, 1470−1477. (b) Wang, J.; Yi, X.; Tang, H.; Han, H.; Wu, M.; Zhou, F. Anal. Chem. 2012, 84, 6400−6406. (c) Fang, S.; Lee, H. J.; Wark, A. W.; Corn, R. M. J. Am. Chem. Soc. 2006, 128, 14044−14046. (18) Shi, C.; Liu, Q.; Ma, C.; Zhong, W. Anal. Chem. 2014, 86, 336− 339. (19) Liu, H.; Li, L.; Duan, L.; Wang, X.; Xie, Y.; Tong, L.; Wang, Q.; Tang, B. Anal. Chem. 2013, 85, 7941−7947. (20) Zhang, P.; Zhang, J.; Wang, C.; Liu, C.; Wang, H.; Li, Z. Anal. Chem. 2014, 86, 1076−1082. (21) Cheng, Y.; Lei, J.; Chen, Y.; Ju, H. Biosens. Bioelectron. 2014, 51, 431−436. (22) Ren, Y.; Deng, H.; Shen, W.; Gao, Z. Anal. Chem. 2013, 85, 4784−4789. (23) Ryoo, S. R.; Lee, J.; Yeo, J.; Na, H. K.; Kim, Y. K.; Jang, H.; Lee, J. H.; Han, S. W.; Lee, Y.; Kim, V. N.; Min, D. H. ACS Nano 2013, 7, 5882−5891. (24) (a) Sowers, L. C.; Fazakerly, G.. V.; Eritja, R.; Kaplan, B. E.; Goodman, M. F. Proc. Natl. Acad. Sci. U. S. A. 1986, 83, 5434−5438. (b) Jean, J. M.; Hall, K. B. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 37− 41. (25) Mitsis, P. G.; Kwagh, J. G. Nucleic Acids Res. 1999, 27, 3057− 3063. (26) Xu, Q.; Cao, A.; Zhang, L.; Zhang, C. Y. Anal. Chem. 2012, 84, 10845−10851. (27) Marti, A. A.; Jockusch, S.; Li, Z.; Ju, J.; Turro, N. J. Nucleic Acids Res. 2006, 34, e50. (28) (a) Eritja, R.; Kaplan, B. E.; Mhaskar, D.; Sowers, L. C.; Petruska, J.; Goodman, M. F. Nucleic Acids Res. 1986, 14, 5869−5885. (b) Xu, D. G.; Evans, K. O.; Nordlund, T. M. Biochemistry 1994, 33, 9592−9599. (29) Feng, K. J.; Kong, R. M.; Wang, H.; Zhang, S. F.; Qu, F. L. Biosens. Bioelectron. 2012, 38, 121−125. (30) Hartig, J. S.; Grune, I.; Najafi-Shoushtari, S. H.; Famulok, M. J. Am. Chem. Soc. 2004, 126, 722−726. (31) Baker, M. B.; Bao, G.; Searles, C. D. Nucleic Acids Res. 2012, 40, e13. (32) Harcourt, E. M.; Kool, E. T. Nucleic Acids Res. 2012, 40, e65. (33) Esquela-Kerscher, A.; Slack, F. J. Nat. Rev. Cancer 2006, 6, 259− 269. (34) Kumar, M. S.; Erkeland, S. J.; Pester, R. E.; Chen, C. Y.; Ebert, M. S.; Sharp, P. A.; Jacks, T. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 3903−3908. (35) Johnson, S. M.; Grosshans, H.; Shingara, J.; Byrom, M.; Jarvis, R.; Cheng, A.; Labourier, E.; Reinert, K. L.; Brown, D.; Slack, F. J. Cell 2005, 120, 635−647. (36) Takamizawa, J.; Konishi, H.; Yanagisawa, K.; Tomida, S.; Osada, H.; Endoh, H.; Harano, T.; Yatabe, Y.; Nagino, M.; Nimura, Y.; Mitsudomi, T.; Takahashi, T. Cancer Res. 2004, 64, 3753−3756. (37) Karube, Y.; Tanaka, H.; Osada, H.; Tomida, S.; Tatematsu, Y.; Yanagisawa, K.; Yatabe, Y.; Takamizawa, J.; Miyoshi, S.; Mitsudomi, T.; Takahashi, T. Cancer Sci. 2005, 96, 111−115.
G
dx.doi.org/10.1021/ac503365z | Anal. Chem. XXXX, XXX, XXX−XXX
