Molecular recognition of HER-1 in whole-bloodsamplesIuliana Moldoveanua,b, Camelia Stanciu Gavanc

and Raluca-Ioana Stefan-van Stadena,b*

Multimode sensing was proposed for molecular screening and recognition of HER-1 in whole blood. The tools usedfor molecular recognition were platforms based on nanostructured materials such as the complex of Mn(III) withmeso-tetra (4-carboxyphenyl) porphyrin, and maltodextrin (dextrose equivalence between 4 and 7), immobilizedin diamond paste, graphite paste or C60 fullerene paste. The identification of HER-1 in whole-blood samples, atmolecular level, is performed using stochastic mode and is followed by the quantification of it using stochasticand differential pulse voltammetry modes. HER-1 can be identified in the concentration range between 280 fg/mland 4.86 ng/ml using stochastic mode, this making possible the early detection of cancers such as gastrointestinal,pancreatic and lung cancers. The recovery tests performed using whole-blood samples proved that the platforms canbe used for identification and quantification of HER-1 with high sensitivity and reliability in such samples, thesemaking them good molecular screening tools. Copyright © 2014 John Wiley & Sons, Ltd.

Keywords: molecular recognition; molecular screening; early diagnosis; multimode sensing; HER-1

INTRODUCTION

HER-1 (epidermal growth factor (EGF) receptor-1) or ErbB-1 playsa critical role in cancer biology (Menard et al., 2000), regulatesproliferation and cell fate during epidermal development and isactivated by several EGF-family ligands including heparin-binding EGF-like growth factor (Alroy and Yarden, 1997). It is amitogenic and chemotactic molecule that participates in tissuerepair, tumour growth, and other tissue-modelling phenomena,such as angiogenesis and fibrogenesis (Krampera et al., 2005).HER-1 signalling may be dysregulated through receptoroverexpression and/or mutation. Dysregulated HER-1 signallinghas multiple cellular effects, including increased cell cycleprogression, survival and spread of cancer cells, increasedtranscription of genes, cell migration and metastasis (Fujinoet al., 1996; Prenzel et al., 2001; Herbst and Bunn, 2003; Ettinger,2006). The relationship between HER-1 and prognosis has beenstudied in multiple solid tumour types, including non-smal-celllung carcinoma (Kanematsu et al., 2003; Selvaggi et al., 2004),pancreatic cancer (Ueda et al., 2004; Bloomston et al., 2006; Leeet al., 2007) and breast cancer (DeFazio-Eli et al., 2011).HER-1 expression was assayed using different methods,

including immunohistochemical assays (IHCs; Kanematsu et al.,2003; Leibl and Moinfar, 2005), enzyme-linked immunosorbentassay (Gullick et al., 1986), and electrochemical methods(Elshafey et al., 2013; Vasudev et al., 2013). Stochastic sensingwas proposed previously for molecular screening of whole bloodfor dopamine (Stefan-van Staden et al., 2012), breast cancerantigen (CA15-3; Stefan-van Staden and van Staden, 2013), andhepatitis B biomarkers (Stefan-van Staden et al., 2013; Stefan-vanStaden and Moldoveanu, 2014).The purpose of this work was to develop a new molecular

screening method based on multimode sensing for identificationand quantification of HER-1 in whole-blood samples. Themultimode sensing was perfomed using stochastic and differential

pulse voltametry modes recorded with new analytical platformsbased on different pastes.

Selection of multimode sensing employs materials withcertain characteristics: The presence of pores is essential forstochastic mode, and the capacity of material to be a goodelectrocatalyst is essential for the amperometric–differential pulsevoltammetry mode. Therefore, we select maltodextrin (MD) andthe complex of Mn(III) with meso-tetra (4-carboxyphenyl) porphyrin(Mn(III)P) as modifiers for different carbonmatrices such as pastes ofC60 fullerene, graphite, and diamond. The design of paste-basedsensors is highly reliable and easy to do in any laboratory; therefore,it is preferred to be used as sensing part of the platform.

EXPERIMENTAL

Materials and reagents

HER-1 (EGF receptor), natural diamond powder (DP) having aparticle size of 1μm (99.9%), MD (dextrose equivalent 4–7), the

* Correspondence to: Raluca-Ioana Stefan-van Staden, Laboratory of Electro-chemistry and PATLAB, National Institute of Research for Electrochemistry andCondensedMatter, 202 Splaiul Independentei Str., 060021, Bucharest-6, Romania.E-mail: iustinavanstaden@yahoo.com

a I. Moldoveanu, R.-I. Stefan-van StadenLaboratory of Electrochemistry and PATLAB, National Institute of Research forElectrochemistry and Condensed Matter, 202 Splaiul Independentei Strada,060021 Bucharest-6, Romania

b I. Moldoveanu, R.-I. Stefan-van StadenFaculty of Applied Chemistry and Material Science, Politehnica University ofBucharest, Bucharest, Romania

c C. Stanciu GavanDepartment of Surgery, University of Medicine and Pharmacy ‘Carol Davila’,Bucharest, Romania

Research Article

Received: 10 February 2014, Revised: 7 April 2014, Accepted: 2 May 2014, Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI: 10.1002/jmr.2388

J. Mol. Recognit. 2014; 27: 653–658 Copyright © 2014 John Wiley & Sons, Ltd.

653

Mn(III)P, C60 fullerene, graphite powder and tetrahydrofuran(THF) were purchased from Aldrich (Milwaukee, USA). Paraffinoil was purchased from Fluka (Buchs, Switzerland). Titrisol buffersolution (pH=7.4) was purchased from Merck. Deionized waterobtained from a Millipore Direct-Q 3 System (Molsheim, France)was used for the preparation of all solutions. All standardsolutions were buffered at pH= 7.4 with physiological buffer(saline phosphate buffer, pH= 7.4).

Instrumentation

All measurements were performed using a PGSTAT 12potentiostat/galvanostat connected to a three-electrode celland linked to a computer via an Eco Chemie (Utretch, theNetherlands) software version 4.9. The pH measurements wereperformed using a CyberScan PCD 6500 multiparameter.

Design of multimode sensing platforms

The active sides of the multimode sensing platform were threemodified pastes: C60 fullerene paste modified with MD, graphitepaste modified with MD, and diamond paste modified withMn(III)P. The powder of C60 fullerene, graphite or diamond wasmixed with paraffin oil to form a paste. To 100mg of paste, thesolution of nanostructured materials was added: 50μl ofMn(III)P solution (10�3mol/l in THF, to diamond paste, DP) or25-μl solution of MD (10�3mol/l in water, to graphite paste orC60 fullerene paste) to give the modified paste. The diameter ofthe active site representing the working electrode (WE) in theplatform was 100μm.

The modified paste was placed on platforms containing Ag/AgClink as reference electrodes and Pt ink as an auxiliary electrode.A non-conductive plastic was used for the design of the platform(Figure 1). When not in use, the platforms were stored in a drystate at room temperature.

Stochastic mode

For the stochastic mode, a chronoamperometric technique wasused for the measurements of ton and toff at 125mV. The plat-form was dipped in solutions containing the analyte of differentconcentrations; equations of calibration 1/ton = f(concentrationof HER-1) were determined using statistics. The unknownconcentrations of HER-1 in whole-blood samples were deter-mined from the calibration equations based on the values ofton recorded, while toff values were used as signatures of HER-1for its identification.

Differential pulse votammetry (DPV) mode

For the DPV mode, the applied pulse potential was 25mV/sversus Ag/AgCl. The platform was dipped into the solutioncontaining the analyte; the peak heights were measured at thepotentials given in Table 2 (versus Ag/AgCl) and were plottedagainst the concentrations of HER-1. The unknown concentra-tions of the HER-1 in whole-blood samples were obtained fromthe calibration graphs, based on the peak heights measured atthe same potentials recorded for the standard solutions.The first mode applied was the stochastic one in order to

check if the whole-blood sample is containing HER-1; forthose samples where HER-1 was identified based on the toffvalue (its signature), the quantification was first performedusing the stochastic mode (Figure 2) and then using the DPVmode (Figure 3).

RESULTS

The main advantage of stochastic sensing is that it can be usedfor the qualitative as well as quantitative analysis of HER-1 basedon its signature measured using toff values. Therefore, toff valuesof HER-1 were measured for each platform, and the results areshown in Table 1. Based on the values recorded for ton, for eachplatform, the sensitivities, limits of quantification and the ranges

Figure 1. Platform used for the molecular screening of HER-1.

I. MOLDOVEANU ET AL.

wileyonlinelibrary.com/journal/jmr Copyright © 2014 John Wiley & Sons, Ltd. J. Mol. Recognit. 2014; 27: 653–658

654

of concentration were determined on which HER-1 can bedetermined in whole-blood samples (Table 1). The highestsensitivity was recorded for the platform containing MD/C60paste (1.80 × 105 smg/ml), the same platform being able todetermine lower concentrations of HER-1 in whole-blood on awider concentration range than the other two platformsproposed. For the concentration of 100–1ng/ml, the platformsbased on the MD/graphite and Mn(III)P/DP paste, respectively, canbe used with a sensitivity of magnitude order of 1×103 smg/ml.The lowest limit of determination was recorded for the platformbased on the MD/C60 paste.

The ranges of utilization of the platforms in DPVmode were deter-mined and are shown in Table 2. Using this mode, HER-1 can be de-termined on a concentration range between 1.40pg/ml–3.04μg/ml.The lowest limit of determination (1.40 pg/ml) and the highestsensitivity (3.35x106Amg/ml) were recorded for the platformcontaining the Mn(III)P/DP paste. The platform based on C60/MDcan be used in this mode for quantifying higher concentrationsof HER-1—up to 3.04μg/ml.

The selectivity of the proposed platforms was determined byassessing the values of toff of different biomarkers, for example,tumour -specific biomarkers like CA19-9, CA12-5 and neuron-specific enolase, of some neurotransmitters, for example, dopa-mine, epinephrine, norepinephrine and hormones like estradiol,testosterone and dihydrotestosterone, of interleukin-6, andleptin as well as of antigens specific to hepatitis B. Differentvalues for toff were obtained for all these compounds andHER-1, proving that they do not interfere in the measurementsof HER-1. Furthermore, a standard addition method was usedfor the recovery test of HER-1 in biological fluids; three whole-blood samples from confirmed patients (IHC results negativefor HER-1) were spiked with different amounts of HER-1. Thevalues of toff were found—as determined for standard-buffered(phosphate saline buffer, pH= 7.4) HER-1—in whole-bloodsamples. Based on determined ton values and the equation ofcalibration for each plaltform, the values of HER-1 recoveredwere determined, which in percentage are 99.02 ± 0.05% whenthe paste based on MD/C60 was used, 98.97 ± 0.12% when thepaste based on MD/graphite was used, and 99.32 ± 0.08% whenthe paste based on MnP(III)/DP was used. These recovery testsproved that the proposed platforms are free of interferenceswhen used for the assay of HER-1 in whole-blood samples andthat the matrix of whole blood did not affect the measurements.

Molecular screening of whole-blood samples was performedfor 12 controlled samples (from confirmed patients presentinglung or gastric cancers) received from the University of Medicineand Pharmacy ‘Carol Davila’ Bucharest, accordingly with theEthics Committee approval Nr. 11/2013. From the 12 samplesreceived, only in five of them was HER-1 identified, based ontheir signatures (Figure 2). These results were confirmed by theIHCs performed in the specialized laboratories of the faculty.For the five samples, the quantification of HER-1 was performedusing the platforms in stochastic (Figure 2) and DPV (Figure 3)mode. The results obtained are shown in Tables 3 and 4. As

(c)

(a)

(b)

Figure 2. Diagrams used for HER-1 determination in whole blood usingthe platforms based on (a) MD/C60 paste, (b) MD/graphite paste, and (c)Mn(III)P/DP paste, in stochastic mode.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

I (A

)

0

40x10-9

20x10-9

-20x10-9

-40x10-9

MD/C60

MD/GraphiteMn(III)P/DP

Figure 3. Diagrams used for HER-1 determination in whole blood usingthe platforms based on (a) MD/C60 paste, (b) MD/graphite paste, and (c)Mn(III)P/DP paste, in DPV mode.

MOLECULAR RECOGNITION OF HER-1

J. Mol. Recognit. 2014; 27: 653–658 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jmr

655

Table 1. Response characteristics of the platforms for stochastic mode

Modified pastebased on

Signaturetoff (s)

Linear concentrationrange, mg/ml

Calibrationequationa

Sensitivity s,mg/ml

Limit of determination,mg/ml

MD/C60 5.0 2.80 × 10�10–1.94 × 10�7 1/ton = 0.04 + 1.80(±0.09)× 105 × C r = 0.9982

1.80 × 105 2.80 × 10�10

MD/Graphite 6.0 1.94 × 10�7–4.86 × 10�6 1/ton = 0.07 + 6.40(±0.18)× 103 × C r = 0.9998

6.40 × 103 1.94 × 10�7

Mn(III)P/DP 4.4 1.94 × 10�7–4.86 × 10�6 1/ton = 0.03 + 5.03(±0.13)× 103 × C r = 0.9989

5.03 × 103 1.94 × 10�7

a<1/ton>= /s; <C>= mg/ml.

Table 2. Response characteristics of the platforms in DPV mode

Modified pastebased on

E (mV) Linear concentrationrange, mg/mL

Calibrationequationa

Sensitivity,A mg/mL

Limit of determination,mg/mL

MD/C60 1351 ± 21 2.43 × 10�5–3.04 × 10�3 H = 1.2 + 1.32(±0.11)× 104 × C r = 0.9986

1.32 × 104 2.43 × 10�5

MD/Graphite 500 ± 12 4.86 × 10�6–6.08 × 10�4 H = 2.1 + 3.36(±0.13)× 104 × C r = 0.9998

3.36 × 104 4.86 × 10�6

Mn(III)P/DP 800± 31 1.40 × 10�9–1.94 × 10�7 H = 0.1 + 3.35(±0.15)× 106 × C r = 0.9997

3.35 × 106 1.40 × 10�9

a<H>=nA; <C>=mg/mL.

Table 3. Quantitative analysis of HER-1 from whole blood using stochastic mode

Sample Nr. Multimode sensing IHC

HER-1, mg/ml

Paste based on

MD/C60 MD/graphite Mn(III)P/DP

1 11.10(±0.06) × 10�8 10.70(±0.08) × 10�8 12.50(±0.10) × 10�8 ++2 8.20(±0.02) × 10�8 10.90(±0.10) × 10�8 8.10(±0.03) × 10�8 +3 7.50(±0.03) × 10�8 7.30(±0.07) × 10�8 7.60(±0.05) × 10�8 +4 5.90(±0.03) × 10�8 5.40(±0.07) × 10�8 5.40(±0.04) × 10�8 +5 3.30(±0.01) × 10�8 3.10(±0.02) × 10�8 3.00(±0.02) × 10�8 +

IHC, immunohistochemical assay of tissue samples.

Table 4. Quantitative analysis of HER-1 from whole blood using DPV mode

Sample number Multimode sensing IHC

HER-1, mg/ml

Paste based on

MD/C60 MD/Graphite MnP(III)/DP

1 11.30(±0.23) × 10�8 10.60(±0.21) × 10�8 11.2(±0.19) × 10�8 ++2 8.90(±0.12) × 10�8 1.10(±0.15) × 10�9 8.8(±0.16) × 10�8 +3 6.20(±0.10) × 10�8 6.60(±0.18) × 10�8 6.3(±0.17) × 10�8 +4 5.90(±0.13) × 10�8 5.80(±0.18) × 10�8 5.3(±0.18) × 10�8 +5 3.20(±0.18) × 10�8 3.30(±0.13) × 10�8 2.8(±0.18) × 10�8 +

IHC, immunohistochemical assay of tissue samples.

I. MOLDOVEANU ET AL.

wileyonlinelibrary.com/journal/jmr Copyright © 2014 John Wiley & Sons, Ltd. J. Mol. Recognit. 2014; 27: 653–658

656

can be seen from the tables, there is a good agreement betweenthe results obtained using stochastic mode (Table 3) and DPVmode (Table 4). Therefore, the molecular screening test usingmultimode sensing can be used for early detection of cancersuch as lung cancer and gastric cancer, based on identificationand quantification of HER-1 in whole blood. The screening testcan also be a non-expensive tool for monitoring the patientsduring treatment. The platforms can be used continuously formore than 100 screening tests of whole-blood samples; inbetween two measurements, the surfaces of the WE, Ag/AgCland Pt should be rinsed with distilled water and dried withabsorbent paper. If one should choose the best platform, theplatforms of choice will be those containing the MD/C60 pasteand Mn(III)P/DP paste.

DISCUSSIONS

A molecular screening test is proposed for HER-1 in whole-bloodsamples. The screening test is based on molecular sensing. Formolecular sensing, three pastes are proposed: MD/C60 fullerenepaste, MD/graphite paste, and the Mn(III)P /diamond paste.Signatures of HER-1 were determined using the stochastic modefor each platform; these signatures were used to identify HER-1in whole-blood samples. Quantitative analysis of HER-1 in awhole-blood sample was performed using both stochastic anddifferential pulse voltammetry modes. The results obtainedusing the platforms were in agreement with those provided byIHCs. The platforms can be used continuously for more than100 whole-blood molecular screening tests, making them cheapand reliable tools for early detection of cancers (e.g. lung,gastrointestinal and pancreatic cancer).

The test has great features in oncology for early detection ofcancer by screening of whole blood as well as other biologicalfluids such as saliva. Another feature is monitoring the patientsunder treatment. The main advantages are as follows: Thewhole-blood sample can be used as taken from the patient.one drop of blood being enough for the test, the time of analysisis maximum of 20min, very small concentrations of HER-1 can bedetermined, facilitating early detection of tumour, and the test ishighly reliable. Furthermore, the technique will allow the assay ofa ‘cocktail’ of biomarkers specific to certain tumours togetherwith HER-1, given that in the molecular recognition usingstochastic sensing, the signal is given by the molecularinteraction between a single molecule and the pore at a time.Different values of toff obtained during selectivity studies forbiomarkers like CA-19-9, CA-125, neuron-specific enolase, andother substances of biological importance proved the potentialof the proposed platform for the assay of a ‘cocktail’ ofbiomarkers and other molecules together with HER-1.

CONFLICT OF INTEREST STATEMENT

The authors declare no competing financial interests.

Acknowledgement

This work was supported by PN II Program Capacity, 2012–2014,Contract nr. 3ERC-like/2012. I Moldoveanu and RI Stefan-vanStaden are thankful for the funding from the Sectoral OperationalProgramme Human Resources Development 2007-2013 of theMinistry of European Funds through the Financial AgreementPOSDRU/159/1.5/S/137390.

REFERENCES

Menard S, Tagliabue E, Campiglio M, Pupa SM. 2000. Role of HER-2gene overexpression in breast carcinoma. J. Cell. Physiol. 182:150–162.

Alroy I, Yarden Y. 1997. The ErbB signalling network in embriogenesis andoncogenesis: signal diversification through combinatorial ligand-receptor interactions. FEBS Lett. 410: 83–86.

Krampera M, Pasini A, Rigo A, Scupoli MT, Tecchio C, Malpeli G, Scarpa A,Dazzi F, Pizzolo G, Vinante F. 2005. HB-EGF/HER-1 signaling inbone marrow mesenchymal stem cells: inducing cell expansionand reversibly preventing multilineage differentiation. Blood106: 59–66.

Herbst RS, Bunn PA Jr. 2003. Targeting the epidermal growth factorreceptor in non-small cell lung cancer. Clin. Cancer Res. 9:5813–5824.

Ettinger DS. 2006. Clinical implications of EGFR expression in thedevelopment and progression of solid tumors: focus on non-smallcell lung cancer. Oncologist 11: 358–373.

Prenzel N, Fischer OM, Streit S, Hart S, Ullrich A. 2001. The epidermalgrowth factor receptor family as a central element for cellularsignal transduction and diversification. Endocr. Relat. Cancer 8:11–31.

Fujino S, Enokibori T, Tezuka N, Asada Y, Inoue S, Kato H, Mori A.1996. A comparison of epidermal growth factor receptor levelsand other prognostic parameters in non-small cell lung cancer.Eur. J. Cancer 32A: 2070–2074.

Kanematsu T, Yano S, Uehara H, Bando Y, Sone S. 2003. Phosphorylation,but not overexpression, of epidermal growth factor receptor isassociated with poor prognosis of non-small cell lung cancerpatients. Oncol. Res. 13: 289–298.

Selvaggi G, Novello S, Torri V, Leonardo E, De Giuli P, Borasio P, Mossetti C,Ardissone F, Lausi P, Scagliotti GV. 2004. Epidermal growth factorreceptor overexpression correlates with a poor prognosis incompletely resected non-small-cell lung cancer. Ann. Oncol. 15:28–32.

Ueda S, Ogata S, Tsuda H, Kawarabayashi N, Kimura M, Sugiura Y,Tamai S, Matsubara O, Hatsuse K, Mochizuki H. 2004. The cor-relation between cytoplasmic overexpression of epidermalgrowth factor receptor and tumor aggressiveness: poor progno-sis in patients with pancreatic ductal adenocarcinoma. Pancreas29: e1–8.

Bloomston M, Bhardwaj A, Ellison EC, Frankel WL. 2006. Epidermalgrowth factor receptor expression in pancreatic carcinoma usingtissue microarray technique. Dig. Surg. 23: 74–79.

Lee J, Jang KT, Ki CS, Lim T, Park YS, Lim HY, Choi DW, Kang WK, Park K,Park JO. 2007. Impact of epidermal growth factor receptor(EGFR) kinase mutations, EGFR gene amplifications, and KRASmutations on survival of pancreatic adenocarcinoma. Cancer109: 1561–1569.

DeFazio-Eli L, Strommen K, Dao-Pick T, Parry G, Goodman L, Winslow J.2011. Quantitative assays for the measurement of HER1-HER2heterodimerization and phosphorylation in cell lines and breasttumors: applications for diagnostics and targeted drug mechanismof action. Breast Cancer Res. 13: 1–8.

Leibl S, Moinfar F. 2005. Metaplastic breast carcinomas are negative forHer-2 but frequently express EGFR (Her-1): potential relevance toadjuvant treatment with EGFR tyrosine kinase inhibitors? J. Clin.Pathol. 58: 700–704.

Gullick WJ, Marsden JJ, Whittle N, Ward B, Bobrow L, Waterfield MD. 1986.Expression of epidermal growth factor receptors on human cervical,ovarian, and vulval carcinomas. Cancer Res. 46: 285–292.

Vasudev A, Kaushik A, Bhansali S. 2013. Electrochemical immunosensorfor label free epidermal growth factor receptor (EGFR) detection.Biosens. Bioelectron. 39: 300–305.

Elshafey R, Tavares AC, Siaj M, Zourob M. 2013. Electrochemicalimpedance immunosensor based on gold nanoparticles–proteinG for the detection of cancer marker epidermal growth factorreceptor in human plasma and brain tissue. Biosens. Bioelectron.39: 143–149.

MOLECULAR RECOGNITION OF HER-1

J. Mol. Recognit. 2014; 27: 653–658 Copyright © 2014 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jmr

657

Stefan-van Staden RI, Balasoiu SC, van Staden JF. 2012. Stochastic dotmicrosensors for the assay of dopamine in pharmaceutical samplesand biological fluids. J. Electrochem. Soc. 159: B839–B844.

Stefan-van Staden RI, van Staden JF. 2013. New tool for screening ofwhole blood for early detection of breast cancer antigen (CA153). JMod Med Chem 1: 86–91.

Stefan-van Staden RI, Moldoveanu I, Sava DF, Kapnissi-Christodoulou C,van Staden JF. 2013. Enantioanalysis of pipecolic acid with stochasticand potentiometric microsensors. Chirality 25: 114–118.

Stefan-van Staden RI, Moldoveanu I. 2014. Stochastic sensors based onnanostructured materials used in the screening of whole blood forhepatitis B. J. Electrochem. Soc. 161: B3001–B3005.

I. MOLDOVEANU ET AL.

wileyonlinelibrary.com/journal/jmr Copyright © 2014 John Wiley & Sons, Ltd. J. Mol. Recognit. 2014; 27: 653–658

658