Biosensors and Bioelectronics 74 (2015) 849–855

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

http://d0956-56

n CorrE-m1 Th

journal homepage: www.elsevier.com/locate/bios

Comprehensive detection of diverse exon 19 deletion mutations ofEGFR in lung Cancer by a single probe set

Jin Ho Bae 1, Seong-Min Jo 1, Hak-Sung Kim n

Department of Biological Sciences, Korea Advanced Institute of Science and Technology (KAIST), Daejeon 305-701, Republic of Korea

a r t i c l e i n f o

Article history:Received 21 April 2015Received in revised form12 July 2015Accepted 20 July 2015Available online 21 July 2015

Keywords:Exon 19 deletion mutationsOligo-quencherMolecular beaconLung cancerEGFR

x.doi.org/10.1016/j.bios.2015.07.04363/& 2015 Elsevier B.V. All rights reserved.

esponding author.ail address: hskim76@kaist.ac.kr (H.-S. Kim).ese authors equally contributed to the work.

a b s t r a c t

Detection of exon 19 deletion mutation of EGFR, one of the most frequently occurring mutations in lungcancer, provides the crucial information for diagnosis and treatment guideline in non-small-cell lungcancer (NSCLC). Here, we demonstrate a simple and efficient method to detect various exon 19 deletionmutations of EGFR using a single probe set comprising of an oligo-quencher (oligo-Q) and a molecularbeacon (MB). While the MB hybridizes to both the wild and mutant target DNA, the oligo-Q only binds tothe wild target DNA, leading to a fluorescent signal in case of deletion mutation. This enables thecomprehensive detection of the diverse exon 19 deletion mutations using a single probe set. We de-monstrated the utility and efficiency of the approach by detecting the frequent exon 19 deletion mu-tations of EGFR through a real-time PCR and in situ fluorescence imaging. Our approach enabled thedetection of genomic DNA as low as 0.02 ng, showing a detection limit of 2% in a heterogeneous DNAmixture, and could be used for detecting mutations in a single cell level. The present MB and oligo-Q dualprobe system can be used for diagnosis and treatment guideline in NSCLC.

& 2015 Elsevier B.V. All rights reserved.

1. Introduction

Lung cancer is one of leading causes of death, giving rise toaround 1.3 million deaths annually with a 5-year survival rate atonly around 15% (Jemal et al., 2004). Two types of lung cancer,namely non-small-cell and small-cell lung cancers, have beenidentified, and non-small-cell lung cancer (NSCLC) has beenknown to be more common and caused by mutations on theepidermal growth factor receptor (EGFR). One of the most fre-quently occurring EGFR mutations is the in-frame deletion muta-tions on exon 19, accounting for approximately 48% of all EGFR-mutated NSCLC (Mitsudomi and Yatabe, 2010). Various anticanceragents, such as EGFR tyrosine kinase inhibitors (EGFR-TKI) in-cluding gefitnib and erlotnib, have been developed and clinicallyused for treating EGFR-mutated NSCLC. Extensive clinical trialswith such therapeutics have revealed that therapeutic benefits areclosely associated with the exon 19 deletion mutations of EGFR,leading to overall response rates higher than 82% (Fukuoka et al.,2011; Lynch et al., 2004; Maemondo et al., 2010; Shepherd et al.,2005).

Various methods have been developed to detect exon 19 de-letion mutations, including PNA–LNA-PCR clamp, PCR invader and

Scorpion-ARMS analysis (Ellison et al., 2013; Janne et al., 2006;Kimura et al., 2006; Yatabe et al., 2006). Nonetheless, a simple andefficient detection of exon 19 deletion mutations of EGFR remainsa challenge because diverse types of deletion mutations exist(Sharma et al., 2007). The use of conventional methods usuallyrequires respective probes specific for each deletion mutations.Therefore, to identify all exon 19 deletion mutations of EGFR, di-rect sequencing of EGFR fragments is used following amplificationof genomic DNA from tumor tissue or circulating tumor cells(Schwarzenbach et al., 2011). However, the amount of DNA sampleavailable from cancer patients is limited, which restricts directsequencing of DNA. Furthermore, if the portion of tumor tissue isless than 20% in a sample, direct sequencing data may be in-accurate (Yatabe et al., 2006). Therefore, a rapid and accurate de-tection of diverse exon 19 deletion mutations of EGFR is of greatsignificance for diagnosis and treatment guideline in NSCLC.

Here we demonstrate a comprehensive and simple method todetect diverse exon 19 deletion mutations of EGFR by using asingle set of a molecular beacon (MB) and an oligo-quencher(oligo-Q). The MB binds to an intact DNA fragment in both thewild-type EGFR and its deletion mutants, whereas the oligo-Q isdesigned to bind only to the wild-type DNA target, but not to thetargets with deletion mutations (Scheme 1). Consequently, ourapproach leads to a fluorescent signal from the MB only whenexon 19 deletion mutations are present, enabling the detection ofdiverse exon 19 deletion mutations of EGFR using a single probeset. We demonstrated the utility and simplicity of the approach by

Scheme 1. Schematic representation for detecting diverse exon 19 deletion mutations of EGFR using a single probe set. As for wild-type template, molecular beacon (MB)binds to the upstream region of exon 19 deletion site, and an oligo-quencher (oligo-Q) hybridizes with the deletion sites, which results in no fluorescence signal owing to thequenching of fluorescence dye. In contrast, in the presence of exon 19 deletion mutations, MB hybridizes with the upstream region of target DNA, but oligo-Q does not bindto the target DNA owing to in-frame deletion mutations. Consequently, MB emits the fluorescence signals.

J.H. Bae et al. / Biosensors and Bioelectronics 74 (2015) 849–855850

efficiently detecting most frequent exon 19 deletion mutations ofEGFR through a real-time PCR and in situ fluorescence imaging.The present approach can be used for clinical diagnosis andtreatment guideline in NSCLC. Details are reported herein.

2. Materials and methods

2.1. Reagents

nPfu-Forte DNA polymerase, dNTP, and its buffer were obtainedfrom Enzynomics (Daejeon, Korea). Synthetic templates and pri-mers were synthesized by Bioneer (Daejeon, Korea).

2.2. Lung cancer cell lines

H1975, H460, HCC827, and PC9 cell lines were from SamsungMedical Center (Seoul, Korea). H1975, and H460 contain the wildtype exon 19 of EGFR, whereas HCC827, and PC9 possess muta-tions. All the cell lines were cultured in RPMI 1640 medium sup-plemented with 10% fetal bovine serum (GE Healthcare Hyclone,United Kingdom). All cells were cultured in 5% CO2 at 37 °C (MCO-5AC; Sanyo, Japan).

2.3. Design of molecular beacon and oligo-quencher

Molecular beacon (MB) for the detection of various exon 19deletion mutations of EGFR was designed according to theguidelines described elsewhere (Marras et al., 2006). Secondarystructure of the molecular beacon was predicted using Quikfoldwebserver program (Zuker, 2003). FAM dye was used as a reporterand DABCYL as a quencher, and conjugated to 5′- and 3′-ends ofMB, respectively. Oligo-Q was designed to bind only to the wild-type DNA target, but not to the targets with deletion mutations bythe similar manner, and BHQ1 was conjugated to 5′-end of oli-gonucleotide as a quencher (Xiang et al., 2014). Both the molecularbeacon and oligo-Q were synthesized by Bioneer (Daejon, Korea).

2.4. Thermal profiling of molecular beacon and oligo-quencher

The thermal profiling was carried out using Rotor Gene Q(Qiagen, CA, USA). The pre-melt was done at 95 °C for 90 s, fol-lowed by a cooling from 95 °C to 25 °C at a rate of 1 °C/30 s.Fluorescence was measured at each degree during the 30 s hold.The reaction mixture contained a total of 10 pmole of molecularbeacon and 50 pmole of target or random template DNA in a totalof 25 μL of 10X nPfu-Forte buffer. The amount of oligo-Q usedranged from 0 to 250 pmole.

2.5. Real time PCR using a molecular beacon and an oligo-quencher

Real time PCR was carried out using CFX96 Realtime System(BioRad, CA, USA) under the following condition: 95 °C for 5 min,followed by 45 cycles of 95 °C for 30 s, 52 °C for 60 s, and 72 °C for30 s. Fluorescence intensity was measured at the end of the 52 °Cextension period. The total cycling time was around 2 h and30 min. The reaction mixture for real-time PCR using syntheticDNA contained 5 pmole of MB, 2 pmole of oligo-Q, 10 pmole ofprimer, 10 pmole of target DNA, 2 pmole of dNTP, and 0.625 unitsof nPfu-Forte DNA polymerase in a total of 25 μL of 10X nPfu-Fortebuffer. A range of 0.01–1000 pmole of template DNA, 0–5 pmole ofmolecular beacon, and 0–40 pmole of oligo-Q were used if other-wise stated. The primers used were: forward 5′-CTCTCTGTCA-TAGGGACT-3′ and reverse 5′-CATCGAGGATTTCCTTGTTC-3′. ThePCR products were subjected to 2% agarose gel electrophoresis.

2.6. In situ fluorescence imaging using a molecular beacon and anoligo-quencher

H1975, H460, HCC827 and PC9 cells were cultured on eight-channel culture slides for 2 days. Cells were fixed using coldacetone for 7 min followed by a three-time wash with Dulbecco'sphosphate-buffered saline (DPBS) (Jo et al., 2013). The cells wereincubated with 400 nM of a MB and various concentrations of anoligo-Q ranging from 0 to 4 μM. Cells were washed again withDPBS, followed by DAPI staining, subjected to imaging with aconfocal laser microscope (LSM NLO 710, Carl Zeiss, Germany).

Fig. 1. Thermal profiling of MB and oligo-Q. (A) Thermal profile of MB in thepresence of a target, random template, or in the absence of DNA templates.(B) Thermal profiles of MB and oligo-Q with only MB target template, MB andoligo-Q target template, and random template. The ratios of MB to oligo-Q were 1:1and 1:5. (C) Thermal profiles of MB and oligo-Q for the wild-type and 10 differentexon 19 deletion mutations. nInsertion of nucleic sequence AAGA.

J.H. Bae et al. / Biosensors and Bioelectronics 74 (2015) 849–855 851

3. Results

3.1. Design of a molecular beacon and an oligo-quencher

It has been known that a variety of exon 19 deletion mutationsof EGFR exist in NSCLC. The use of the conventional molecularbeacon (MB) system will require respective MBs for each exon 19deletion mutations (Bonnet et al., 1999; Tan et al., 2000). We at-tempted to develop a method to detect such diverse exon 19 de-letion mutations of EGFR using a single probe set. To this end, an‘oligo-quencher’ (oligo-Q) was employed and used together with aMB for detecting various deletion mutations. Based on the se-quences of the most frequent exon 19 deletion mutations, we firstdesigned a MB that is complementary to the intact DNA fragment,

namely the upstream region of the actual exon 19 deletion mu-tations of EGFR as depicted in Scheme 1. The designed MB wascomposed of 25 nucleotides forming a stem and a loop, and a FAMdye was tethered at the 5′ end and a Dabcyl quencher at the 3′end. When the MB remains unbound, no fluorescent signal isemitted owing to a fluorescence resonance energy transfer be-tween a FAM dye and a Dabcyl quencher. In contrast, when the MBhybridizes to the complementary region on the target DNA, astrong fluorescence signal is emitted.

We next designed an oligo-Q complementary to the regionwhere the various deletion mutations occur. Consisting of 23 nu-cleotide sequences with a black-hole-quencher (BHQ1) at the 3′end, the oligo-Q thus only binds to the wild-type DNA of EGFR, butnot to the DNA with deletion mutations (Scheme 1). The site ofhybridization of MB and oligo-Q was designed based on the report(Marras, 2006) to be three nucleotides apart (physical distance isestimated to be 10.2 Å), an effective distance for the BHQ1 toquench the FAM of the MB when both the MB and oligo-Q bind tothe DNA. On the other hand, as the oligo-Q only binds to the wild-type, the MB will emit a fluorescence signal only when the exon 19deletion mutations are present on the target DNA. Therefore, thedesigned dual probe system comprising the MB and oligo-Q en-ables the detection of various exon 19 deletion mutations of EGFR(Santangelo et al., 2004). The nucleotide sequences of the designedMB and oligo-Q are shown in Table S1.

To determine the optimal hybridization temperatures for thedesigned MB, we examined the thermal profiles. The difference influorescence signals was measured at temperatures ranging from25 °C to 90 °C using synthetic templates consisting of (1) a tem-plate with the MB complementary sequence and (2) a randomsequence template (Table S2). As the temperature decreased fromaround 66 °C, the fluorescence signal of the MB increased as itbound to the synthetic target (Fig. 1a). As for the random sequencetemplate, the fluorescence signal decreased, clearly differentiatingthe signals between the target and random sequences, which leadsto the ‘window of discrimination’ between 25 °C and 66 °C. Thesignal level from the random sequence template were almostnegligible as the control without any DNA template, indicating ahighly specific binding of the MB to the target. Based on thethermal denaturing profiles, the optimal temperatures for the MBwere determined to range from 20 °C to 66 °C.

We also examined the optical hybridization temperature of theoligo-Q for the exon 19 deletion mutations. At temperature ran-ging from 25 °C to 90 °C, difference in fluorescence signals wasmeasured using synthetic templates consisting of (1) both MB andoligo-Q complementary template, (2) only MB and random oligo-Qcomplementary template, and (3) both random sequence template(Table S3). From around 67 °C, the difference in fluorescence sig-nals increased as the temperature decreased (Fig. 1b). Compared tothe only MB and random oligo-Q template, the fluorescence signaldropped proportionally as the amount of oligo-Q increased forboth MB and oligo-Q template. This result indicates that the oligo-Q quenched the FAM dye of the MB only when the oligo-Q boundto the target. Based on the thermal denaturing profiles for theoligo-Q, the optimal temperatures for the hybridization of oligo-Qwere determined to be between 25 °C and 67 °C, similar to that ofthe MB thermal profile. Taken together, the optimal temperatureduring real-time PCR was fixed at 52 °C.

3.2. Detection of the various deletion types in exon 19

With the optimized thermal profiles of the MB and oligo-Q, wetested if the dual probe system can detect the various exon 19deletion mutations. Using the sequence data from the exon 19deletions of EGFR in lung cancer patients, 10 different re-presentative templates for deletion mutations were synthesized

Table 1Sequences of exon 19 deletion mutations tested using a single probe set.

Mutation type Sequence

Wildn GAAAGTTAAAATTCCCGTCGCTATCAAGGAATTAAGAGAAGCAACATCTCCGAAAΔE746-A750 (1) GAAAGTTAAAATTCCCGTCGCTATCAAG—————————————ACATCTCCGAAAΔE746-A750 (2) GAAAGTTAAAATTCCCGTCGCTATCAA—————————————AACATCTCCGAAAΔL747-S752 ins S GAAAGTTAAAATTCCCGTCGCTATCAAGGAAT—————————————CTCCGAAAΔL747-P753 ins S GAAAGTTAAAATTCCCGTCGCTATCAAGGAAT————————————————CGAAAΔE749-K754 ins E GAAAGTTAAAATTCCCGTCGCTATCAAGGAATCAAGAG—————————————AAΔL747-T751 ins P GAAAGTTAAAATTCCCGTCGCTATCAAGGAA————————C–CATCTCCGAAAΔE746-S752 ins V GAAAGTTAAAATTCCCGTCGCTATCAAGG———————————————T-TCCGAAAΔK745-T751nn GAAAGTTAAAATTCCCGTCGCTATCA—————AAGA—————CATCTCCGAAAΔL747-K754 ins EE GAAAGTTAAAATTCCCGTCGCTATCAAGGAA————————CAAC——————AAΔI744-T751 ins IKT GAAAGTTAAAATTCCCGTCGCT——————ATTAAGA—————CATCTCCGAAA

n The bold and underline sequences indicate the hybridization sites for MB and oligo-Q, respectively.nn Insertion of nucleic sequence AAGA.

The ratio of MB to oligo-Q was 1:5.

J.H. Bae et al. / Biosensors and Bioelectronics 74 (2015) 849–855852

(Marchetti et al., 2012), and thermal profiling was carried out forthe ten templates and the wild-type. The MB and the oligo-Q wereshown to give rise to distinct fluorescence signals for all of 10mutation templates (Table 1), whereas the wild-type templateresulted in a negligible signal (Fig. 1c). This result indicates thatour dual probe system could detect the 10 exon 19 deletion mu-tations, strongly implying that our system can also be used forother deletion mutations which are known to differ only slightlyfrom the tested deletion mutations.

3.3. Optimization of the molecular beacon and oligo-quencher byreal-time PCR

We then attempted to detect the exon 19 deletion by real-timePCR. The wild-type and the common deletion mutant (ΔE746-A750 (1)) template were used (Table S4). We used differentamounts of wild type template for real-time PCR at a fixed amountof MB without oligo-Q. For a range from 0.01 to 1000 pmole of thetemplate, the threshold cycle, which is a required cycle to obtainthe threshold fluorescence intensity for positive detection, showeda linear relationship (Fig. S1a). This result indicates the utility ofthe real-time PCR as well as the specificity of the MB in exon 19deletion detection. Based on the results, the optimal amount of thetemplate was determined to be 10 pmole for MB and oligo-Q op-timization tests. To optimize the amounts of MB and oligo-Q inreal-time PCR, different amounts of MB from 0 to 5 pmoles weretested, and the fluorescence signal was shown to increase linearlywith the amount of MB (Fig. S1b). As the fluorescence intensityincreased with the amount of MB, we used 5 pmoles of MB forreal-time PCR tests. Next, the amount of oligo-Q was optimized toaccomplish a maximum signal-to-background ratio without com-promising the yield. We conducted real-time PCR for the wild-typeand mutant template with increasing oligo-Q amounts. As a result,the signal for the wild-type decreased to almost zero level with2 pmole of oligo-Q, whereas a distinct fluorescence signal from themutant template remained constant (Fig. S1c). Increasing theamount of oligo-Q resulted in a decrease in the fluorescence fromthe mutant template, compromising the yield. The decrease offluorescence probably occurred due to random energy transferbetween the MB and the free oligo-Q in the solution as the amountof oligo-Q increased. Based on the result, the optimal amounts ofthe MB and oligo-Q were fixed at 5 pmole and 2 pmole,respectively.

3.4. Detection of exon 19 deletion mutation through real-time PCR

We tested the dual probe system for genomic DNA extractedfrom lung cancer cell lines. PC9 with the deletion mutation

(ΔE746-A750) was used as a mutant cell type, and H460 was usedas the wild-type. Real-time PCR was conducted using 20 ng ofgDNA from each cell line. As shown in Fig. 2a, gDNA from PC9showed a strong fluorescence signal in the absence and presenceof oligo-Q, but wild-type gDNA from H460 gave rise to a signalonly in the absence of oligo-Q. Wild-type gDNA resulted in anegligible fluorescence signal in the presence of oligo-Q, similar tothe control with no template. A slight fluctuation in the fluores-cence signal was observed at the start of the real-time PCR cycle.We next determined the sensitivity in the detection of exon 19deletion mutations using gDNA from mutant cell line PC9. As aresult, the threshold cycle showed a linear relationship within therange from 20 ng to 0.02 ng of genomic DNA (Fig. S2), and thedetection limit was estimated to be 0.02 ng of gDNA (Fig. 2b). Wefurther compared the detection sensitivity of the dual probe sys-tem with the conventional sequencing method for a mixture ofdifferent ratios between wild-type and mutant DNA. As shown inFig. 2c, the use of the dual probe system enabled the detection ofheterogeneous gDNA as low as 2%, showing much higher sensi-tivity than the conventional sequencing method. Direct sequen-cing was shown to have heterogeneous sensitivity of around 10%in our previous work (Oh et al., 2010). Furthermore, we tested theMB and oligo-Q for two additional cancer cell lines, including amutant cell line (HCC827) and a wild type cell line (H1975). Si-milar to PC9 and H460, distinct fluorescence signals were observedonly from gDNA of the mutant cell lines (Fig. S3). This resultstrongly supports that the present method can be used for de-tecting exon 19 deletion mutations with high sensitivity in het-erogeneous cell types by real-time PCR.

3.5. Detection of the exon 19 deletion mutation through in situfluorescence imaging

We attempted to detect the exon 19 deletion mutations byfluorescence imaging using the MB and oligo-Q. PC9 and H460cells were used as the mutant and wild-type cell line, respectively.In the case of in situ hybridization, the MB and an oligo-Q hy-bridize to mRNA of EGFR exon 19 in the cells. We tested variousconcentrations of the MB and oligo-Q. As shown in Fig. 3, in theabsence of an oligo-Q, both PC9 and H460 showed green fluores-cence signals. However, green fluorescence signal for H460 wassignificantly reduced at the MB to oligo-Q ratio of 1:5, but stillexhibited a weak fluorescence signal. At the ratio of 1:10, no greenfluorescence signal was observed for H460, whereas PC9 exhibiteda distinct green fluorescence signal. Based on the result, the op-timized ratio for the MB to oligo-Q was chosen as 1:10. The opti-mized ratios of fluorescence imaging and real-time PCR differowing to different test conditions. As for the in situ cell imaging,

Fig. 2. Detection of exon 19 deletion mutation in genomic DNA by real-time PCR.Genomic DNA extracted from PC9 (Mutant) and H460 (Wild) cell lines were used.(A) Real-time PCR using 20 ng of genomic DNA with or without oligo-Q.(B) Fluorescence intensity for various amounts of mutant and wild-type genomicDNA. (C) Fluorescence intensity for different proportions of mutant and wild-typegenomic DNA in a total of 20 ng of genomic DNA.

J.H. Bae et al. / Biosensors and Bioelectronics 74 (2015) 849–855 853

the probes hybridized to the fixed mRNA in the cells and anyunbound oligo-Q was washed thoroughly. In contrast, for the real-time PCR, free unbound oligo-Q was present in the solution. With

the optimized ratio, we tested two additional cancer cell lines,including a mutant cell line (HCC827) and a wild-type cell line(H1975) (Fig. S4). Similar to PC9 and H460, only the mutant cellline was shown to display distinct green fluorescence signals inthe presence of the MB and oligo-Q. Blue spheres indicate thenucleus stained by DAPI. This result indicates that the MB andoligo-Q bind specifically to the mRNAs of the wild-type and mu-tant EGFR exon 19 in cells, and exon 19 deletion mutations couldbe detected through a fluorescence imaging.

4. Discussion

We have demonstrated a comprehensive and simple method todetect diverse exon 19 deletion mutations of EGFR using a singleprobe set comprising of an oligo-Q and a MB. Although there havebeen many reports regarding MBs associated with dual FRET, uti-lizing MB with oligo-Q to detect various mutation targets withpositive fluorescence in case of a mutation is novel. The presentapproach allowed for the detection of 10 frequently occurringdeletion mutations, clearly showing its utility and simplicity. Themajor barrier in detecting the exon 19 deletion mutations of EGFRis the existence of numerous deletion types. Several methods, in-cluding PNA–LNA PCR clamp, PCR invader, and Scorpion-ARMSanalysis, have been developed, but have some limitations. Thesemethods require specific probes for respective deletion mutations,making it difficult to detect all of the deletion mutations at a singleanalysis. Thus, direct sequencing is most commonly used for de-tecting exon 19 deletion mutations, requiring a long time and highcost. Furthermore, direct sequencing usually needs at least 20% oftarget DNA in a sample to obtain reliable data. Based on the re-sults, our approach can be widely used for detecting various exon19 deletion mutations in genomic DNA from lung cancer patients.

We used two different approaches, real-time PCR and in situfluorescence imaging, to detect the exon 19 deletion mutations ofEGFR. The real-time PCR enabled specific and sensitive detection ofthe mutations even in a highly heterogeneous DNA sample. Thedetection sensitivity of the real-time PCR was estimated to be0.02 ng of genomic DNA. The serum genomic DNA collected frompatients has been known to range from 1 ng/μl to 10 ng/μl (Asanoet al., 2006; Hoshi et al., 2007; Janne et al., 2006; Kimura et al.,2006; Takano et al., 2007). Therefore, our dual probe system islikely to allow for the detection of exon 19 deletion mutationseven at much lower concentrations of genomic DNA in serum.Consequently, the minimum amount of target DNA in a hetero-geneous DNA mixture was as low as 2%. The dual probe systemtherefore led to a much higher sensitivity than the conventionaldirect sequencing method, showing better or comparable sensi-tivity to other methods (Table 2). Furthermore, the fluorescenceimaging using a single probe set enabled the detection of exon 19deletion mutation in a single cell level, requiring no DNA pre-paration or amplification step. Recently, WIP-QP method has beendeveloped for detecting various exon 19 deletion mutations usinga single probe (Nakamura et al., 2012). However, this method isdifficult to use as a real-time PCR tool as well as a microscopy toolto detect mutation in intact cells. The present method can be usedfor the detection of exon 19 deletion mutations of EGFR for a smallnumber of tumor cells, such as circulating tumor cells isolatedfrom the blood of lung cancer patients.

5. Conclusion

In summary, we have shown that the single probe set com-prising the MB and oligo-Q can be effectively used for compre-hensive detection of the exon 19 deletion mutations of EGFR,

Fig. 3. Detection of exon 19 deletion mutations in cancer cells by in situ fluorescence imaging. Images of PC9 (A, B, C) and H460 (D, E, F) cell lines are shown for various ratioof MB to oligo-Q. In the cases of (A) and (D), only MB was added in the absence of oligo-Q. As for (B) and (E), the ration of MB to oligo-Q ratio was 1:5. Finally, for (C) and (F),the ratio of MB to oligo-Q was 1:10. The fixed cells were incubated with 400 nM of MB for 30 min at 60 °C. Magnification is 400 fold. (For interpretation of the references tocolor in this figure, the reader is referred to the web version of this article.)

Table 2Sensitivity of mutation analysis methods.

Method Limit of detection (%) Detection methodn References

Direct sequencing 10–25 X (Roma et al., 2013)Fragment analysis 5 X (Roma et al., 2013)Real-time PCR (allelic discrimination) 5 X (Roma et al., 2013)Dual Probe System 2 Fluorescence imaging NAPCR invader 1 X (Goto et al., 2012)PNA-LNA PCR clamp 1 X (Goto et al., 2012)ARMS 1 X (Roma et al., 2013)CastPCR 0.1 X (Roma et al., 2013)WIP-QP 0.1 X (Nakamura et al., 2012)

n Method of detection exon 19 deletion mutation of EGFR other than PCR.

J.H. Bae et al. / Biosensors and Bioelectronics 74 (2015) 849–855854

providing the crucial information for diagnosis and therapeuticguideline in NSCLC. Based on the results, the present method en-abled the detection of numerous exon 19 deletion mutations withhigh sensitivity through real-time PCR and in situ fluorescenceimaging. The simplicity of the present approach allows for easygenotyping, offering some advantages over conventional methods.The principle of the present approach can be applied to the de-tection of other targets with different deletion or even additionmutations.

Acknowledgment

This research was supported by the Bio & Medical TechnologyDevelopment Program (NRF-2013M3A9D6076530), Mid-ca-reer Researcher Program (NRF-2014R1A2A1A01004198) of theNational Research Foundation (NRF) funded by the Ministry ofScience, ICT & Future Planning, and Brain Korea 21 funded by theMinistry of Education.

Appendix A. Supplementary material

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.bios.2015.07.043.

References

Asano, H., et al., 2006. Clin. Cancer Res. 12, 43–48.Bonnet, G., et al., 1991. Proc. Natl. Acad. Sci. USA 96, 6171–6176.Ellison, G., et al., 2013. J. Clin. Pathol. 66, 79–89.Fukuoka, M., et al., 2011. J. Clin. Oncol. 29, 2866–2874.Goto, K., et al., 2012. Ann. Oncol. 23, 2914–2919.Hoshi, K., et al., 2007. Clin. Cancer Res. 13, 4974–4983.Janne, P.A., et al., 2006. Clin. Cancer Res. 12, 751–758.Jemal, A., et al., 2004. Cancer 101, 3–27.Jo, S.M., et al., 2013. Biosens. Bioelectron. 46, 142–149.Kimura, H., et al., 2006. Cancer Sci. 97, 642–648.Lynch, T.J., et al., 2004. N. Engl. J. Med. 350, 2129–2139.Maemondo, M., et al., 2010. N. Engl. J. Med. 362, 2380–2388.Marchetti, A., et al., 2012. PLoS One 7, e42164.Marras, S.A., 2006. Methods Mol. Biol. 335, 3–16.

J.H. Bae et al. / Biosensors and Bioelectronics 74 (2015) 849–855 855

Marras, S.A., et al., 2006. Clin. Chim. Acta 363, 48–60.Mitsudomi, T., Yatabe, Y., 2010. FEBS J. 277, 301–308.Nakamura, T., et al., 2012. J. Thorac. Oncol. 7, 1369–1381.Roma, C., et al., 2013. Biomed. Res. Int. 2013, 385087.Oh, Y.H., et al., 2010. J. Mol. Diagn. 12, 644–652.Santangelo, P.J., et al., 2004. Nucleic Acids Res. 32, e57.Schwarzenbach, H., et al., 2011. Nat. Rev. Cancer 11, 426–437.

Sharma, S.V., et al., 2007. Nat. Rev. Cancer 7, 169–181.Shepherd, F.A., et al., 2005. N. Engl. J. Med. 353, 123–132.Takano, T., et al., 2007. Clin. Cancer Res. 13, 5385–5390.Tan, W., et al., 2000. Chemistry 6, 1107–1111.Xiang, D., et al., 2014. Talanta 129, 249–253.Yatabe, Y., et al., 2006. J. Mol. Diagn. 8, 335–341.Zuker, M., 2003. Nucleic Acids Res. 31, 3406–3415.