Da

Ta

b

a

ARRAA

KBTEN

1

iatcacroHcnce

tm

t2

0d

Biosensors and Bioelectronics 25 (2010) 2686–2689

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

journa l homepage: www.e lsev ier .com/ locate /b ios

etection of breast cancer cells specially and accurately byn electrochemical method

ing Lia, Qi Fana, Tao Liub, Xiaoli Zhua, Jing Zhaoa,b, Genxi Li a,b,∗

Laboratory of Biosensing Technology, School of Life Sciences, Shanghai University, Shanghai 200444, PR ChinaDepartment of Biochemistry and National Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, Nanjing 210093, PR China

r t i c l e i n f o

rticle history:eceived 25 January 2010eceived in revised form 1 April 2010ccepted 4 May 2010vailable online 7 May 2010

eywords:

a b s t r a c t

Breast cancer is one of the most common cancers to cause death in the world, and the accurate diagnosisis of great importance to determine the stage of the disease and then to design the suitable therapy.Compared with the traditional detection methods relying on the recognition of only one tumor marker, weherein propose a sensitive electrochemical immunoassay to detect breast cancer cells by simultaneouslymeasuring two co-expressing tumor markers, human mucin-1 and carcinoembryonic antigen on thesurface of the cancer cells, which may efficiently improve the accuracy of the detection as well as facilitate

reast cancer cellsumor markerslectrochemistryanoparticles

the classification of the cancer cells. The experimental results have revealed that well electrochemicalresponse can be observed only under the condition that both of the tumor markers are identified on thesurface of the tumor cells. With this method, breast cancer cell MCF-7 can be easily distinguished fromother kinds of cells, such as acute leukemia cells CCRF-CEM and normal cells islet beta cells. Moreover,the prepared cytosensor can specially monitor breast cancer cell MCF-7 in a wide range from 104 to107 cell mL−1 with well reproduction and low detection limit, which may have great potential in clinicalapplications.

. Introduction

Breast cancer is a kind of cancer that mainly takes place in thenner lining of the milk ducts or lobules with different spread,ggressiveness and genetic makeup. Worldwide, breast cancer ishe second common type of cancer after lung cancer and the fifthommon cause of cancer death. The distant metastases are regardeds the major reason to cause cancer death (Lacroix, 2006), so theancer treatment at an early stage is highly expected, and accu-ate diagnose of breast cancer is essential to determine the extentf disease and to plan appropriate therapies (Berois et al., 2000).owever, conventional histological methods are inaccurate, timeonsuming and lack of efficiency. Therefore, the development ofew techniques to unambiguously, simply and rapidly identify andharacterize the tumor cells has been of great importance (Ehrhart

t al., 2008; Liu et al., 2007).

Tumor marker is a substance abnormally expressed in responseo cancer as well as certain benign conditions. Different tumor

arkers may indicate different types of cancers with altered dis-

∗ Corresponding author at: Department of Biochemistry and National Key Labora-ory of Pharmaceutical Biotechnology, Nanjing University, 22 Hankou Road, Nanjing10093, PR China. Tel.: +86 25 83593596; fax: +86 25 83592510.

E-mail address: genxili@nju.edu.cn (G. Li).

956-5663/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.bios.2010.05.004

© 2010 Elsevier B.V. All rights reserved.

ease process, while one kind of cancer may have more than onemarker (Koepke, 2006). To breast cancer, human mucin-1(MUC1)and carcinoembryonic antigen (CEA) are the most common markersto monitor the metastatic breast tumors (Molina and Gion, 1998).On the one hand, since highly overexpressed MUC1 is frequentlyfound in breast cancer, either throughout the cytosol or aroundthe plasma membrane, and is associated with a poor prognosisand increased lymph node metastases, elevated level of MUC1 hasbeen considered to be an important indicator in the diagnose ofbreast cancer (Cheung et al., 2000; Rahn et al., 2001; Shen et al.,2008). On the other hand, lots of studies have revealed the posi-tive correlation between changes in CEA and therapeutic responsein patients with metastatic breast cancer (Haagensen et al., 1978;Jezersek et al., 1996; Kammerer et al., 2003). And, some studies havepointed out that the 5-year and 10-year survival rates for breastcancer patients with CEA-positive tumors are significantly worsethan those with CEA negative ones (Shousha et al., 1979). There-fore, compared with the detection by evaluating MUC1 only, theadditional detection of CEA may provide more precise prognosticinformation, and be helpful in classifying tumors in more detail,

so that a more appropriate cancer treatment can be planed outto maximize efficacy and minimize toxicity according to the dis-tinct tumor cell type. In the meantime, since previous studies haverevealed that the tumor markers can be also found in normal cellsof peripheral blood, lymph nodes or bone marrow, and molecular

T. Li et al. / Biosensors and Bioelectronics 25 (2010) 2686–2689 2687

cells

mctd

bmtcmsbutcobg

2

2

nHwbd6paacwp>

2

wbfta

Fig. 1. Schematic illustration of the method to detect breast cancer

arkers are also found to express frequently in normal epithelialells (Taback et al., 2001), the traditional detection based on oneumor marker is not reliable, which may yield false-positive resultue to the low specificity.

In this work, we have proposed a new assay method to screenreast cancer cells by simultaneously recognizing the two tumorarkers on the breast cancer cell surfaces. As is illustrated in Fig. 1,

he breast cancer cells MCF-7 are firstly recognized due to the spe-ific interaction between MUC1 on the cell surface and its aptamerolecules, which are immobilized on a gold electrode surface. Sub-

equently, another tumor marker CEA on the cell surface is capturedy CdS nanoparticles (CdS NPs) labeled anti-CEA. Therefore, onlynder the condition that both the two markers are expressed byhe cells, the well electrochemical wave can be observed, so breastancer cells MCF-7 can be detected accurately and specially. More-ver, due to the merits of electrochemical techniques, the assay cane operated easily and rapidly, so this proposed method may havereat potential in future application.

. Experimental

.1. Materials

The thiolated DNA were manufactured by Invitrogen Biotech-ology Co., Ltd. The sequence of MUC 1 aptamer was 5′-S-GCAGTTGATCCTTTGGATACCCTGG-3′. The control sequenceas 5′-HS-CACGACGTTGTAAAACGACGGCCAG-3′. Albumin from

ovine serum (BSA), N-hydroxysuccinimide (MHS), 1-ethyl-3-(3-imethyllaminopropyl) carbodiimide hydrochloride (EDC HCl),-mercapto-1-hexanol (MCH), and tris (2-carboxyethyl) phos-hine hydrochloride (TCEP) were purchased from Sigma. CdCl2nd thioacetamide were purchased from J&K Chemical Ltd. CEAntibody (anti-CEA) was from Touching (Shanghai, China). Otherhemicals were all of analytical grade. All solutions were preparedith doubly distilled water, which was purified with a Milli-Qurification system (Branstead, USA) to a specific resistance of18 M� cm.

.2. Preparations of CdS NPs

CdS NPs were synthesized according to our previous reports

ith some modification (Huang et al., 2005). Briefly, 100 mL dou-

ly distilled water was firstly deoxygenated by bubbling nitrogenor 1 h. Then CdCl2 (1 mM) and MPA (2.4 mM) were added intohe aqueous solution, the pH of which was adjusted to 8.0–8.5 bydding 1 M NaOH solution. Nitrogen was further bubbled through

through simultaneous recognition of two different tumor markers.

the solution for another 30 min. Subsequently, 0.625 mM thioac-etamide was added into the above CdCl2-MPA solution, and thefinal molar ratios of Cd2+/MPA/S2− were 1:2.4:0.625. The result-ing mixture solution was refluxed under nitrogen flow at 100 ◦Cfor 10 h with vigorously stirring to obtain a final yellow solutionof MPA-modified CdS NPs solution. The solution was then ultrafil-tered to remove excessive MPA by using Vivaspin 500 (Sartorius,50,000 MW) at 12,000 rpm for 10 min at 4 ◦C. The upper phase waswashed twice and then dissolved with pH 7.4 PBS.

2.3. Bioconjugation of CdS NPs with the antibody

A 250 �L solution of the above purified CdS NPs was firstlymixed with 50 �L of 0.1 mg/mL anti-CEA solution, EDC and NHS(50 mg/mL), the performance of which was carried out under mod-erately shaking in the dark and kept overnight at 4 ◦C. The resultingsample was then ultrafiltered to remove the non-conjugatednanoparticles and by-product by using Vivaspin500 (Sartorius,100,000 MW) at 5000 rpm for 20 min at 4 ◦C, and subsequentlywashed with 10 mM pH 7.4 PBS buffer for three times by ultra-filtration. Thus, the anti-CEA-CdS NPs conjugation was prepared.The product was diluted into 10 mM pH 7.4 PBS buffer and kept at4 ◦C. For control experiments, BSA-CdS NPs was prepared with thesimilar procedure.

2.4. Cell lines and cell culture

The MCF-7, CCRF-CEM and islet beta cells were provided by theInstitute of Biochemistry and Cell Biology, Chinese Academy of Sci-ence. MCF-7 cells were maintained in DMEM supplemented with10% FBS, and CCRF-CEM and islet beta cell were cultured in RPMI-1640 supplemented with 15% FBS. The cells were incubated at 37 ◦Cin a humidified incubator (5% CO2–95% air).

2.5. Preparation of MUC 1 aptamer modified electrode

A substrate gold electrode was prepared by inserting a gold rodinto a glass tube and fixing it with epoxy resin. Electrical contactwas made by wrapping copper wire to the rod with the help ofwoods alloy. The gold electrode was firstly polished on fine sandpapers and alumina (particle size of about 0.05 �m)/water slurry

on silk. Then it was thoroughly ultrasonicated in ethanol and dou-bly distilled water for about 5 min, separately. Finally, the electrodewas electrochemically cleaned to remove any remaining impuritiesin 1 M H2SO4. After drying with nitrogen, the gold electrode wasimmersed in a solution containing 1 �M MUC1 aptamer, 10 mM

2 oelectronics 25 (2010) 2686–2689

TfMgr

2

miftciatePc

2

w0Ttit0

2

Ht0aibfidaNa2At(csgwmE5t

3

aM

Fig. 2. Electrochemical impedance spectra (Nyquist plots) of the bare gold electrode

688 T. Li et al. / Biosensors and Bi

ris–HCl, 1 mM EDTA, 1.0 M NaCl, and 1 mM TCEP (pH 8.0) for 16 h,ollowed by a 2 h treatment with an aqueous solution of 1 mM

CH. After thoroughly rinsed with pure water and dried with nitro-en, MUC1 aptamer modified electrode was prepared and could beeady for use.

.6. Assembly of cells onto MUC1 aptamer modified electrode

After culture for 48 h, the cells were firstly separated from theedium by centrifugation at 1000 rpm for 10 min, and were then

ncubated in 100 �L of PBS containing 0.5% Tween-20 and 1% BSAor 30 min. After that, the cells were separated from the PBS solu-ion by centrifugation and resuspended into a sterile pH 7.4 PBSontaining 1 mM Ca2+ and 1 mM Mg2+. The MUC1 aptamer mod-fied electrode was then immersed into a 100 �L cell suspensiont a certain concentration and incubated at 37 ◦C for 2 h. Afterhe immobilization of the cells onto the MUC1 aptamer modifiedlectrode, the electrode was carefully rinsed with 0.01 M pH 7.4BS containing 0.5% Tween-20 to remove the nonspecific adsorbedells.

.7. Immobilization of anti-CEA-CdS NPs

The MUC1 aptamer modified electrode which was assembledith MCF-7 cells was firstly immersed in a 100 �L PBS containing

.5% Tween-20 and 1% BSA at the room temperature for 30 min.hen it was separately rinsed twice with 0.01 M pH 7.4 PBS con-aining 0.5% Tween-20 and pH 7.4 PBS. After that, the electrode wasncubated with the above anti-CEA-CdS NPs solution for 2 h. Finally,he electrode were rinsed twice with 0.01 M pH 7.4 PBS containing.5% Tween-20 to remove the nonspecific adsorbed conjuncture.

.8. Electrochemical measurements

The above electrode was firstly immersed in a 100 �L 0.1 MNO3 for 2 h to dissolve anti-CEA-CdS NPs loaded on the elec-

rode surfaces. The resulting solution was then mixed with 1.9 mL.2 M pH 5.2 HAc–NaAc buffer for electrochemical analysis by usingnodic stripping voltammetric technique with a mercury film mod-fied glassy carbon electrode. The working electrode was preparedy 4 cycles of alternate deposition at −1.0 V for 40 s and scanrom −0.9 to −0.2 V in 0.2 M pH 5.2 HAc–NaAc buffer contain-ng 40 �g/mL Hg2+ under N2 atmosphere. The anodic strippingetection was carried out by first electrodeposition of cadmiumt −1.1 V for 4 min and then stripping from −0.9 to −0.2 V under2 atmosphere using a square-wave voltammetric waveform, with4 mV potential step, a 15 Hz frequency, and an amplitude of

5 mV. The instrumentation was a model 660C Electrochemicalnalyzer (CH Instruments). A three-electrode system consisting of

he modified gold electrode, saturated calomel reference electrodeSCE) and platinum counter electrode was used for all the electro-hemical measurements. Before the measurement, the electrolytehould be thoroughly deoxygenated by bubbling high-purity nitro-en through the solution for at least 10 min. A stream of nitrogenas then blown gently across the surface of the solution in order toaintain the solution anaerobic throughout all the experiments.

lectrochemical impedance spectrum (EIS) was obtained with a0 mM PBS (pH 7.0) buffer containing 5 mM [Fe(CN)6]3−/4− as elec-rolyte.

. Results and discussion

EIS has been firstly obtained to characterize the surface alter-tion of the gold electrode after subsequent modification withUC1 aptamer, MCF-7 cells and anti-CEA labeled with CdS NPs

(a), subsequently modified with MUC1 aptamer (b), MCF-7 cells (c) and anti-CEA-CdS NPs (d), for a 50 mM Tris–HCl buffer solution containing 5 mM [Fe(CN)6]3−/4−

with pH 7.0, in a frequency range of 1 Hz to 100 kHz. Alternative voltage: 10 mV. Cellconcentration: 107 cell mL−1.

(Fig. 2). Since ssDNA with abundant negative charges is unfavor-able for the electron transfer between the electrochemical probe[Fe(CN)6]3−/4− and the electrode surface as a result of electrostaticrepulsion, after the modification of MUC1 aptamer onto the goldelectrode surface, an increase of the electron-transfer resistancecan be observed. Furthermore, after the modified electrode is incu-bated with MCF-7 cells, due to the recognition between the MUC1aptamer and the MUC1 mucin exposed on the surface of the cells,the cells are then immobilized on the electrode surface, which willcause the further increase of electrical impedance. On the otherhand, if the modified electrode is further incubated with an anti-CEA-CdS NPs solution, because of the specific interaction betweenthe antibody and the antigen, CdS NPs labeled anti-CEA will be fur-ther loaded onto the electrode surface since there will be also CEA,another marker molecules, on the surface of the cells. Therefore,the electron transfer between the electrochemical probe and theelectrode surface is further prohibited. As is shown in Fig. 2, the EISdata can well support the modification procedure.

We have then made use of anodic stripping voltammetry toobtain the electrochemical signal of cadmium component withinthe anti-CEA-CdS NPs complex which has been immobilized on thesurface of the MCF-7 cells due to the specific interaction betweenthe antibody and CEA which is overexpressed by the cancer cells, sothat more useful electrochemical response can be employed for thisstudy. As is shown by the solid line in Fig. 3, after an acid-dissolutionstep, released Cd2+ from the complex can be measured by a mer-cury film coated glassy carbon electrode with 0.2 M acetate buffer(pH 5.2), which may present a well-defined voltammetric peak at−0.708 V.

What is more interesting, from the results revealed by the twocontrol experiments (the dotted and dash line in Fig. 3), we canknow that only under the condition that both MUC1 and CEAare present, well-defined voltammetric wave can be observed.Certainly, small signals can be also observed ascribing to theunremoved physically species adsorbed on the electrode surface.Nevertheless, compared with the electrochemical signals obtainedfor the case that the two tumor markers of both MUC1 and CEA arepresent, the peak values of the control experiments are obviously

much lower.

Specificity is of critical importance in cancer detection, so wehave examined the specificity of our method by detecting other twokinds of cells, islet beta cells and acute leukemia cells CCRF-CEM.According to the mechanism of our proposed method, lack of MUC1

T. Li et al. / Biosensors and Bioelectr

Fig. 3. Square-wave voltammograms for strong acid treated MUC1 aptamer/MCF-7/anti-CEA-CdS NPs which has been modified on a gold electrode (solid line). Thedotted and dash lines are separately the cases that random DNA or BSA is employed.A baseline correction of the resulting voltammogram was performed using the linearbaseline correction mode of the CHI 660C software. Buffer solution: 0.2 M acetatebuffer (pH 5.2). Cell concentration: 107 cell mL−1; potential step: 4 mV; amplitude:25 mV; frequency: 15 Hz.

FF

otNniapdst

tvt

ig. 4. Square-wave voltammograms for different kinds of cells. Others same as inig. 3.

n the cell surface will prohibit the immobilization of the cells ontohe surface of MUC1 aptamer modified electrode, while the CdSPs labeled antibody cannot bind with the antigen if the cells doot express CEA. Actually, as is shown in Fig. 4, this method can

ndeed work very well. Although the nonspecific adsorption maylso result in small electrochemical signal, the comparison of theeaks can clearly indicate that the proposed method can efficientlyistinguish the breast cancer cells from other control cells with highpecificity. So, the reliability and accuracy of the detection of the

umor cells proposed in this work has been greatly enhanced.

We have also conducted studies by using different concentra-ions of MCF-7 cells. Experimental results reveal that well-definedoltammetric peaks can be observed even at a low cell concen-ration, and the value of the peak current can increase linearly

onics 25 (2010) 2686–2689 2689

with cell concentration in a range from 104 to 107 cell mL−1. Theresulting linear equation is: y = −4.925 − 0.00202x, R = 0.999. Theminimum detectable concentration of the cancer cells by usingthis method is 3.3 × 102 cell mL−1, which is lower than the previ-ous reports (Shen et al., 2007; Ruan et al., 2002). Besides, a series offive repetitive measurements with cell concentrations of 104 and107 cell mL−1 may have a relative standard deviation of 4.6% and6.4%, respectively, so the reproducibility of the detection can bealso satisfactory.

4. Conclusions

In summary, we here have proposed an efficient electrochemicalmethod to detect breast cancer cells via recognition of two differenttumor markers expressed on the surface of the cancer cells at thesame time. This method has offered great promise for rapid, simpleand cost-effective analysis of the cancer with specificity, sensitivityand reproducibility. Due to the reliability and accuracy of the detec-tion, it can be also helpful to monitor the spread of breast tumoras well as the response to therapy with more dependable resultsand enhanced efficiency. Based on the precise assay of cancer cells,more appropriate treatments may be carried out to raise the curerate and the survival rates of patients with tumors. Moreover, sincethis method can be further expanded to screen more kinds of tumorcells by altering the related aptamers and the nanoparticles forthe different kinds of tumor markers in accordance with the clinicdemands, this work will be followed by more studies in this andother labs in this community.

Acknowledgements

This work is supported by the National Natural Science Founda-tion of China (Grant Nos. 90406005, 20575028) and the Program forNew Century Excellent Talents in University, the Chinese Ministryof Education (NCET-04-0452). The authors also thank the supportfrom the Science Foundation of Jiangsu Province, China (Grant No.BK2008268).

References

Berois, N., Varangot, M., Aizen, B., Estrugo, R., Zarantonelli, L., FernaÂndez, P., Krygier,G., Simonet, F., Barrios, E., MuseÂ, I., Osinaga, E., 2000. Eur. J. Cancer 36, 717–723.

Cheung, K.L., Graves, C.R.L., Robertson, J.F.R., 2000. Cancer Treat. Rev. 26, 91–102.Ehrhart, J.C., Bennetau, B., Renaud, L., Madrange, J.P., Thomas, L., Morisot, J., Brosseau,

A., Allano, S., Tauc, P., Tran, P.L., 2008. Biosens. Bioelectron. 24, 467–474.Haagensen, D.E., Kister, S.J., Vandevoorde, J.P., Gates, J.B., Smart, E.K., Hansen, H.J.,

Wells, S.A., 1978. Cancer 42, 1512–1519.Huang, Y., Zhang, W., Xiao, H., Li, G., 2005. Biosens. Bioelectron. 21, 817–821.Jezersek, B., Cervek, J., Rudolf, Z., Novakovic, S., 1996. Cancer Lett. 110, 137–144.Kammerer, U., Thanner, F., Kapp, M., Dietl, J., Sutterlin, M., 2003. Anticancer Res. 23,

1051–1055.Koepke, J.A., 2006. Cancer 69, 1578–1581.Lacroix, M., 2006. Endocr. Relat. Cancer 13, 1033–1067.Liu, L., Zhu, X., Zhang, D., Huang, J., Li, G., 2007. Electrochem. Commun. 9, 2547–2550.Molina, R., Gion, M., 1998. Breast 7, 187–189.Rahn, J.J., Dabbagh, L., Pasdar, M., Hugh, J.C., 2001. Cancer 91, 1973–1982.Ruan, C.M., Yang, L.J., Li, Y.B., 2002. Anal. Chem. 74, 4814–4820.Shen, Q., Rahn, J.J., Zhang, J., Gunasekera, N., Sun, X., Shaw, A.R.E., Hendzel, M.J.,

Hoffman, P., Bernier, A., Hugh, J.C., 2008. Mol. Cancer Res. 6, 555–567.Shen, Z.H., Huang, M.C., Xiao, C.D., Zhang, Y., Zeng, X.Q., Wang, P.G., 2007. Anal. Chem.

79, 2312–2319.Shousha, S., Lyssiotis, T., Godfrey, V.M., Scheuer, P.J., 1979. Br. Med. J. 1, 777–779.Taback, B., Chan, A.D., Kuo, C.T., Bostick, P.J., Wang, H., Giuliano, A.E., Hoon, D.S.B.,

2001. Cancer Res. 61, 8845–8850.