Clinica Chimica Acta 425 (2013) 128–138

Contents lists available at ScienceDirect

Clinica Chimica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /c l inch im

Invited critical review

Electrochemical immunosensors in breast and ovarian cancer

Iulia Diaconu a,b,c,d, Cecilia Cristea e, Veronica Hârceagă e, Giovanna Marrazza f,Ioana Berindan-Neagoe e,g, Robert Săndulescu e,⁎a Cancer Gene Therapy Group, University of Helsinki, Helsinki, Finlandb Molecular Cancer Biology Program, University of Helsinki, Helsinki, Finlandc Finnish Institute for Molecular Medicine, University of Helsinki, Helsinki, Finlandd Center for Cell and Gene Therapy, Baylor College of Medicine, Houston, 77030 TX, USAe “Iuliu Haţieganu” University of Medicine and Pharmacy Cluj-Napoca, Faculty of Pharmacy, Analytical Chemistry Department, Romaniaf Universita degli studi di Firenze, Department of Chemistry Ugo Schiff, 3 Via della Lastruccia, Florence, Italyg “Ion Chiricuţă” Cancer Institute, Cluj-Napoca, Romania

⁎ Corresponding author at: 4 Pasteur St., 400349 Cluj-NE-mail address: rsandulescu@umfcluj.ro (R. Săndulesc

0009-8981/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.cca.2013.07.017

a b s t r a c t

a r t i c l e i n f o

Article history:Received 10 April 2013Received in revised form 19 July 2013Accepted 20 July 2013Available online 6 August 2013

Keywords:Electrochemical immunosensorsBreast and ovarian cancerBiomarkers

During the last decades the incidence of cancer increased dramatically especially in developed countries. In spiteof the fact that the immunochemical methods allowed the diagnosis in early stages, the biopsies are generallyinvasive methods that create discomfort to patients. The need for fast, sensitive, easy to use and noninvasivediagnosis tools is actually of great interest for many research groups all over the world.Immunosensors (ISs) are miniaturized measuring devices, which selectively detect their targets by means ofantibodies (Abs) and provide concentration-dependent signals. Ab binding leads to a variation in electric charge,mass, heat or optical properties, which can be detected directly or indirectly by a variety of transducers.A great number of proteins could be considered as recognition element. In this review the attention was focusedon main cancer biomarkers, currently used by immunological methods (immunohistochemistry, ELISA, flowcytometry, Western blot, immunofluorescence etc) and in the development of electrochemical immunoassaysthat could be used in cancer diagnosis, prognosis and therapy monitoring.

© 2013 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1291.1. Electrochemical immunosensors — definitions and different approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1301.2. Immunoassays with biomarkers for ovarian cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1311.3. Mucin proteins (MUC proteins) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

1.3.1. Cancer antigen CA125 (MUC16) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1311.3.2. Cancer antigen CA15-3 (MUC1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

1.4. Human epididymis protein 4 (HE4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1332. Biomarkers for breast cancer and their use in the development of immunoassay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

2.1. Carcinoembryonic antigen (CEA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1342.2. BRCA1 (breast cancer type 1 susceptibility protein) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362.3. Human epidermal growth factor receptor 2 (HER2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

3. Multiplex immunosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

apoca, Romania. Tel.: +40 745770514.u).

ghts reserved.

129I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

1. Introduction

Cancermust be defined as a diseasewhich is characterized by abnor-mal growth and development of normal cells beyond their naturalboundaries. In the last 50 years despite the global efforts to limit the in-cidence of this disease, cancer has become the leading cause of death.The breast cancer is the most common malignancy in women and thesecondmost common cause of cancer-relatedmortality [1]. For this rea-son, the diagnosis at early stage needs specific and sensitive biomarkersand analytical tools [1]. Because diagnoses based on symptoms are notacceptable for cancer, considering the fact that symptomsusually appearwhen tumors are sufficiently large, other useful diagnostic toolsmust bedeveloped. By consequence, a screening for the early detection of canceris needed. This early detection in asymptomatic populations requires anon-invasive (or a minimally invasive) procedure for the assay, whichhas to be performed using a small amount of samples (physiologicalfluids or tissues). Furthermore, an optimal screening assay has to besite-specific (able to detect cancer in different organs and establish siteof tumor formation). In addition, the analytical assay needs to operateat the level of differential diagnosis and to be sufficiently specific to pro-duce an acceptable limit of false-positive results, depending on the prev-alence of the disease in the population [1,2].

Case studies revealed that the early detection still remains the mostpromising approach to improve long-term survival of cancer patients[2]. The precise diagnosis of cancer is nowadays relied on histologicalevaluation of tissues using immunohistochemistry, enzyme-linked im-munosorbent assay (ELISA), radioimmunoassay (RIA), immunofluores-cence (IF), Western blot etc. All these techniques use biomarkers. Abiomarker could be defined as a quantifiable laboratorymeasure of a dis-ease specific biologically relevant molecule that can act as an indicator ofa current or future disease state [1]. The ideal cancer biomarker wouldbe a protein or protein fragment, which can be easily detected in thepatient's blood or urine, but not detected, in healthy patient. Today, themost common use of cancer biomarkers is for the detection of the recur-rent disease and monitoring the cancer therapy [2].

In the last decade, immunology and cancer research has identifiedseveral molecules potentially involved in well-known biological mech-anisms affected by cancer. These new biomarkers present a strikingdissimilarity from the classical “tumor markers”. While the latter werediscovered through their prevalent expression in cancer and are there-fore more indicators of tumor extension, the former are being discov-ered via their relation to the known biological mechanisms of cancercells, which are potentially related to the clinical behavior of the malig-nancy (i.e., aggressiveness, expression of metabolic pathways). Thesepathways may be targets for diverse anticancer agents [3,4].

As an example, the early stage ovarian cancer has an excellent prog-nosis if treated, but advanced stage ovarian cancer, which is diagnosedin approximately 70% of patients, is associated with a poor survivalrate of only 10–30% [5]. Given the limitations of treatment for advancedovarian cancer and the success of treatment for early stage disease, ascreening test is demanded. The ability to accurately detect early stagedisease would potentially improve ovarian cancer survival. However,the low prevalence of ovarian cancer (30–50 cases/100,000 women)limits the achievable sensitivity and specificity of any screening test[6]. Therefore, clinical outcomeand possibly survivalmay be significant-ly improved by the identification of stage I disease without the need tochange surgical or chemotherapeutic approaches.

Usually different biomarkers are probably important for differenttasks. For instance, CA125 is a useful serum biomarker for monitoringovarian cancer patients and possibly for prediction of the patient'sresponse to therapy, but has insufficient sensitivity for diagnosis [7].Unfortunately, CA125, a high molecular weight glycoprotein andserum biomarker [7], is approved for monitoring recurrence of disease[4,5,8]. Clinically significant elevations in CA125 occur in only 80–85%of women diagnosed with early stage disease and 50–60% of late stagedisease. Furthermore, CA125 can also be increased in a variety of other

conditions, both benign and malignant, such as the first trimester ofpregnancy, breast cancer and endometriosis, as well as lesions that pro-mote any type of peritoneal irritation [9].

It is well known that the distinctive characteristic of immune systemis its ability to distinguish self from nonself. An extraordinarily selectiveand versatile reagent is provided by nature in the form of the antibody(Ab). As a part of the immune defense system in animals, antibodies ofhigh specificity can be synthesized by an organism in reasonable quantitywithinweeks of injecting a foreign species called an antigen (Ag). Organ-isms recognize the presence of nonself and respond rapidly by synthesiz-ing Ab that exhibits high binding constants for the different surfacechemical features of the Ag. This is a remarkable feat of life that can beused to great advantage in analytical chemistry. Immunoreactions arerecognized for their high sensitivity and selectivity. This is the main rea-son to select immunochemical methods for clinical analysis. The use ofimmunosensors instead of other immunochemical analyses simplifiesconsiderably the analysis, making it rapid and reliable. Molecules gener-ally designed as antibodies (Abs) include a number of classes and sub-classes of the immunoglobulin whose physiological sites of action,specificity and even molecular weights vary widely.

Cancer is one of the leading causes of mortality. Given this, earlyclinical diagnosis is crucial for successful treatment of the disease.Many immunosensors and immunoassay methods have been devel-oped to isolate single tumor marker, whose concentration in humanserum to be associated with the stages of tumors [5–7].

To improve the sensitivity and the limit of detection of theimmunosensors, researchers started to work with new promisingmaterials like carbon nanotubes (single and multi walls SWCNT,MWCNT), graphene sheets or nanoparticules (magnetic MNP and goldnanoparticles GNP).

Carbon nanotubes have many structures, differing in length, thick-ness, type of helicity and number of layers. Although they are formed es-sentially from the same graphite sheet, their electrical characteristicsdiffer depending on these variations, acting either as metals or assemiconductors. Overall, carbon nanotubes show a unique combinationof stiffness, strength, and tenacity compared to other fiber materialswhich usually lack oneormore of theseproperties. Thermal and electricalconductivity are also very high and comparable to other conductive ma-terials. Their electronic properties suggest that carbon nanotubes havethe ability to mediate electron-transfer reactions with electroactive spe-cies when used as an electrode [8,9]. It has been suggested that the elec-trocatalytic properties of the CNTs originate from the open ends. Theproperties of carbon nanotubes also make them extremely attractive forchemical and biochemical sensor assembly.

Belonging to the same carbon family, graphene is the basic structuralelement of some carbon allotropes including graphite, charcoal, carbonnanotubes and fullerenes. Its structure is one-atom-thick planar sheetsof sp2-bonded carbon atoms that are densely packed in a honeycombcrystal lattice [10].

The magnetic nanoparticles, generally less than 50 nm in diameter,are made up of an iron oxide core stabilized by an organic shell. Theyare called superparamagnetic iron oxide (SPIO) nanoparticles becauseof their magnetic properties. Magnetic nanoparticles can be used forthedetection of cancer. In this case, themagnetic nanoparticles are coat-ed with antibodies targeting cancer cells or proteins. Patient's blood canbe inserted into a microfluidic chip with magnetic nanoparticles in it.These magnetic nanoparticles are trapped inside due to an externallyappliedmagnetic field as the blood is free to flow through. Themagneticnanoparticles can be recovered and the attached cancer-associatedmol-ecules can be assayed for their existence [11]. Another configuration ofan electrochemical immunoassayusingmagnetic nanoparticles and car-bonbased screen printed electrodes is designed to be used as disposabledevice. At the surface of the nanoparticles an ELISA like sandwich assaycan attached a primary antibody recognizing the antigen (like MUC 1 orCA125) that binds to a secondary followed by a third enzyme labeledantibody. The formed complex will transform the substrate into

130 I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

recognizable product (by reduction or oxidation). A general schematicrepresentation of such a device is presented in Fig. 1 [12].

A quantumdot (QD) is amaterialwhich has electronic properties in-termediate between those of bulk semiconductors and those of discretemolecules [13]. Quantum dots are particularly significant for opticalapplications due to their high extinction co-efficient. One of their manyapplications that take advantage of the extraordinary photostability isthe real-time tracking of molecules and cells over extended periods oftime. Antibodies, streptavidin, peptides, nucleic acid aptamers, or small-molecule ligands can be used to target quantum dots to specific proteinson cells [13,14].

A critical review of recent researches towards the developmentof electrochemical immunosensors for the detection of breast and ovar-ian cancer biomarkers is presented in the following chapters. Because ofthe variety of researches that has been published in this area, thisreview is limited to recent publications within the past five years.Most reviews have been organized around different techniques tomake immunosensors or on the type of analyte used for detection.Herein, we present different biomarkers used in several approaches ofimmunosensors for breast and ovarian cancer.

1.1. Electrochemical immunosensors— definitions and different approaches

J. Lin et al. mentioned that immunosensors are important analyticaltools designed to detect the binding event between antibody and

Fig. 1. Schemeof two sandwich formats of the immunoassay based onmagnetic beads [12]: dualthemagnetic beads, (b) blocking of free sites of the beads, (c) reactionwithMUC1 protein, (d) bused to provide the electrochemical signal), (f) electrochemical measurements.

antigen, without need for separation and washing steps. They combinethe inherent specificity of immunoreactions with high sensitivity andconvenience of various physical transducers [15].

There are four main types of immunosensor detection devices: elec-trochemical (potentiometric, amperometric, conductimetric/capacitiveand impedimetric), optical, microgravimetric, and thermometric. Re-cently, the role of electrochemical immunosensors in clinical analysishas increased significantly. Important developments have been recordedespecially in sensor design concerning the type of the membrane select-ed, of transducers and the immobilization technique. New approachesusing nanomagnetic particles (NMP) and quantum dots (QDs) increasedthe number of electrochemical immunosensors and their applications[14]. Their high selectivity and sensitivity as well as the possibility tominiaturize these systems havemade possible the use of electrochemicalimmunosensors for in vivo analyses.

All types can either be run as direct (unlabeled) or as indirect(labeled) immunosensors. The direct sensors are able to detect thephysical changes during the immune complex formation, whereas thelatter sensors use signal-generating labels which allow more sensitiveand versatile detection modes, when incorporated into the complex.There is a great variety of different labels which can be used for indirectimmunosensors. Among the most valuable labels are enzymes such asperoxidase, glucose oxidase, alkalinephosphatase, catalase or luciferase,electroactive compounds such as ferrocene or In2+ salts, and a series offluorescent labels (rhodamine, fluorescein, Cy5, ruthenium diimine

antibody assays (A), aptamer-antibody assay (B); (a) immobilization of the bioreceptor oninding of Ab2/Apt2, (e) interactionwith the enzymatic conjugate(enzyme substrate is then

131I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

complexes, phosphorescent porphyrin dyes, etc) [16]. Although indirectimmunosensors are highly sensitive due to the analytical characteristicsof the label applied, the concept of a direct sensor device is still fascinat-ing and represents a true alternative development to immunoassaysystems. Its potential simplicity holds multiple advantages [17].

Up to date in the literature a large plethora of immunosensors canbe found, described for real sample analysis. The two main classes ofelectrochemical immunosensors utilized in clinical analysis are: amper-ometric and potentiometric immunosensors [18]. Lately, impedimetricand capacitive immunosensors started to gain interest due to theirdirect use to determine the antibody–antigen interaction without theneeds of other reagents and the separation step. Their sensitivities arestill limited, especially for clinical applications.

A central issue of the immunosensors applied in clinical diagnosisor in screening and monitoring is that the biomarker used should beexpressed in detecting limits in biological fluids. A list of approved can-cer biomarkers delivered by FDA [7] suggests that some biomarkers forbreast tumors, ovarian andprostate cancer could beused formonitoring(such as CA15-3, CA27-29, CA125, Her-2/neu) other for screening andmonitoring (PSA, PSA total and PSA complex), prognosis (cytokeratinsin breast tumor), staging (human chorionic gonadotropin-β in testicularcancer) or in drug development. The combination between two or sev-eral biomarkers could lead to the correctly identification of cancer pres-ence in early stages instead of using one biomarker. There are newbiomarkers not yet used in clinical diagnosis butwith promising results,such as; human epididymis protein 4HE4 (for early detection of ovariancancer from serum samples), MUC 1 (CA15-3 in ovarian cancer) or thecombination of CEA and Il-6 for early detection of breast cancer[19,20]. Some important tumor markers, such as carcinoembryonic an-tigen (CEA), carcinoma antigen 125 (CA125), alpha-fetoprotein (AFP),CA15-3 and human chorionic gonadotropin (hCG) have been widelyused for the diagnosis of breast cancer, epithelial ovarian tumors andendometriosis. They are suitable for assessing the activity and complica-tions of the disease and for monitoring therapy in the tumor preventionstage, as well as for the follow-up examination during therapy. In themanagement of cancer patients, they are also used in noninvasivetests for relapse detection [22]. Thus, the detection of tumor markerlevels in human serum is absolutely necessary in clinical assay [15](Table 1).

Several approaches have been used to predict the development ofbreast and ovarian cancer. The assays for several commercial availablebiomarkers are presented in the following chapter.

1.2. Immunoassays with biomarkers for ovarian cancer

Ovarian cancer ranks closely behind pancreatic cancer as the fifthleading cause of death from cancer in U.S. women and is themost lethalof the gynecologic cancers. The majority of women with ovarian cancerare diagnosed when they have distant disease, and the proportionsurviving after 5 years is around 28%. Alternatively, for the minority ofwomen diagnosedwith the disease confined to the ovaries, the propor-tion surviving after 5 years is about 90% (depending on the tumorgrade). Thus, ovarian cancer is an obvious target for better approachesto early detection, including the identification of appropriate molecularmarkers [23].

In a review published in 2010 by A. Goy and co-workers several bio-markers found in serumwerementioned for the diagnosis and followup

Table 1A list of several approved cancer biomarkers for ovarian and breast cancer [8].

Biomarker Disease Cut off value

Her2/neu Breast/stage IV 15 ng mL−1

CA125 Ovarian/breast 95 IU mL−1

CA15-3 Breast 40 U mL−1

Leptin, prolactin Ovarian cancer NA

[24]. Besides CA125and HER4 two other seem important: prostasin andimmunosuppressive acidic protein — IAP.

1.3. Mucin proteins (MUC proteins)

Glycoproteins of the MUC family have been used to develop severaltests for breast and ovarian cancer detection. Mucins are components ofmammary cell–cell junctions andmediate ICAM-1- initiated signal trans-duction. Assays based onMUC1 (different epitopes CA15-3 andCA27-29)and MUC16 (CA125) are probably the best known [19–21], although thelatter is more frequently used for ovarian than for breast cancer [25].MUC1 protein is also expressed in the glandular epithelia of the uterineglands that are the origin of the ectopic lesions of the endometriosis.Moreover, the expression pattern of the protein differs from normal tomalignant tissue. In normal tissue, MUC1 is low or negative, expressedas a hyperglycosilated molecule, and meanwhile in malignant tissue isover-expressed as hyperglycosilated as well as hypoglycosilated protein(Fig. 2).

1.3.1. Cancer antigen CA125 (MUC16)CA125 (Mucin 16) is a high molecular weight protein andmost com-

monly used biomarker for ovarian cancers. It is the onlymarker approvedfor monitoring ovarian cancer progression and treatment response.CA125 has a very low sensitivity for early stage ovarian cancer [19].CA125 is produced by coelomic epithelium which includes mesothelialcells and Mullerian tissues. It is present on the cell surface of ovarian tu-mors and can be detected with monoclonal antibodies such as CA125.This marker is elevated in 80% of epithelial ovarian cancer, yet lacks ex-pression in approximately 20% of patients [26]. CA125 production and re-lease appear to be related to cellular growth.

CA125, even that is a marker discovered 30 years ago, is still the topmarker for ovarian cancers [29].

There are several methodologies for the CA125 immunosensorsdevelopment (Fig. 3).

New detection system was developed using the cyclic voltammetryand a new strategy for constructing a sensitive mediator-type electro-chemical immunosensor, [30]. For this strategy, mediator tris (2,2′-bipyridyl) cobalt (III)[Co(bpy)3]3+) was incorporated into the multi-wall carbon nanotubes–Nafion (MWNTs–Nafion) composite film, via a

Fig. 2. Normal and tumor associated mucin [26] Different statuses of glycosylation of theprotein during disease pathogenesis could be of benefit for early diagnosis and therapymonitoring. Other tested members of the MUC family (e.g. CA 549) [27] have insufficientsensitivity for proper diagnosis [28]. Considering their roles in cell–cell interactions andcancer development, it is possible that other mucins may eventually emerge as diagnosticbiomarkers.

Fig. 3. Scheme of CA125 immunosensor developed: (A) electrodepositing of AuNPs on SPGE, (B) functionality of gold nanoparticles with MUDA, (C) mixed SAM formation with MCH,(D) activation of COOH groups with EDAC/NHS and Ab-CA125 immobilization, (E) blocking step with rIgG, (F) Ab–Ag affinity reaction [33].

132 I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

simple ion-exchange route. Then, gold colloidal nanoparticles (nano-Au) were attached onto [Co(bpy)3]3+/MWNTs–Nafion film throughelectrostatic interaction between negatively charged nano-Au andpositively charged [Co(bpy)3]3+. Finally, CA125 monoclonal antibody(anti-CA125), used as a model antibody, was assembled onto thesurface of nano-Au to achieve an immunosensor for the determinationof CA125 antigen. The detection limit was estimated to be 0.36 U mL−1

at a 3-time signal-to-noise ratio [30].Another approach used the surface plasmon resonance (SPR)

immunosensor for a flow injection label-free detection of CA125 inhuman serum. It was reported to have a detection limit of 0.1 U mL−1

while 0.05 U mL−1 was obtained for the capacitive system. The SPRimmunosensor provided advantages as fast response and real-timemonitoring, while capacitive immunosensor offered a sensitive andcost-effective method for CA125 detection [31].

Another sensor for CA125 detection was developed using anti-CA125 gold hollow microspheres and porous polythionine modifiedGCEs. The gold hollow microspheres greatly amplified the coverage ofanti-CA125 molecules on the electrode surface. Electrochemical detec-tion was accomplished by the amperometric changes occurring beforeand after the antigen–antibody interaction. The CA125 concentrationwas detected from 4.5 to 36.5 UmL−1 [32].

Marrazza et al developed a label-free impedimetric immunosensorfor the detection of tumor marker CA125 based on gold nanoparticlesmodified screen-printed graphite electrode. Under optimized conditions,the biosensor presented a linear range between 0 and 100 U mL−1 and adetection limit of 6 U mL−1 [33].

1.3.2. Cancer antigen CA15-3 (MUC1)Mucin 1 biomarker has been used in the last years for diagnosis of

endometriosis and ovarian cancer [25].Serum CA15-3 levels are elevated in 30% of patients with endometrial

carcinoma and could be correlated with tumor stage. Gallo et al. [34]detected CA15-3 levels N30 U mL−1 in 47% of patients with stage III dis-ease, comparedwith 18% of thosewith stages I–II disease (p = 0.01), andfound a significant relationship between serum CA15-3 positivity(N30 U mL−1 and N50 U mL−1) and shorter survival (p = 0.0004 andp = 0.00025, respectively). Not only the over expression of CA15-3 is

characteristic for the ovarian cancer but also the hyperglycosylation ofit [35].

For CA15-3 detection a gold electrode was modified with ferrocenecarboxylic (Fc-COOH)-doped silica nanoparticles (SNPs) and used asimmobilization support for CA15-3 antibodies. The detection wasbased on the change of the amperometric response (Δip) before andafter antigen–antibody reaction. The developed immunosensor showedunder optimal conditions a detection limit of 0.64 U mL−1 [36].

Another original strategy was proposed for the construction ofreagentless and mediatorless immunosensors based on the direct elec-trochemistry of glucose oxidase (GOD). Firstly, a composite materialcontaining carbon nanotubes (CNTs) and core-shell organosilica andchitosan nanospheres was successfully prepared and cast on the glassycarbon electrode surface directly. Then, Pt nanoclusters (Pt NCs) as anelectron relay were deposited on it to form the interface of biocompat-ibility and huge surface free energy for the adsorption of the first GODlayer. Subsequently, the second Pt NCs layer was deposited on the sur-face of GOD to capture CA15-3 antibodies (anti-CA15-3). Finally, GODas a blocking reagent instead of bovine serum albumin was employedto block the possible remaining active sites of the Pt NCs and avoid thenonspecific adsorption. The immobilized GOD showed direct electrontransfer with a rate constant of 4.89 s−1 and the peak current decreasedlinearly with increasing logarithm of CA15-3 concentration from 0.1 to160 U mL−1 with a relatively low limit of detection of 0.04 U mL−1 at3σ [37].

CA15-3 is also used in the diagnosis of breast cancer.A novel electrochemical immunosensor capable of sensitive and

label-free detection of CA15-3 based on highly conductive grapheneexhibited an increased electron transfer and a very high sensitivity to-wards CA15-3. This novel immunosensor, with a low detection limit of0.012 U mL−1, worked well over a broad linear range of 0.1–20U mL−1

[38].Another two interesting approaches for theMUC1 detectionwere de-

veloped by our research group. An electrochemical aptamer-basedbiosensing assay for MUC1 protein detection by using methylene blue(MB) as electrochemical indicator and modifying the electrode surfaceusing a functionalized conductive polymerwas studied. The optimizationsteps start with electropolymerization of o-aminobenzoic acid (o-ABA)

133I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

onto graphite based screen printed electrodes (SPEs). Immobilization ofprimary antibody as the capturing probe was performed directly onpoly o-ABA (PABA) modified electrodes. Then, a sandwich like structurewas fabricated upon MUC1 -aptamer complex formation, exploitingaptamer as the detection probe and methylene blue as the electrochem-ical activemarker intercalating in the aptamer without previous labeling.The aptamer instead of antibody was successfully used for the electro-chemical detection (Fig. 4).

The recognition of immunoreactions and aptamer binding event wasidentified via monitoring the interfacial electron transfer resistance withelectrochemical impedance spectroscopy (EIS) and cyclic voltammetry(CV). CV and differential pulse voltammetry (DPV)were employed to de-tect the change of MB oxidization peak current related with the humanMUC1 protein concentration. DPV detection showed a reliable andmore sensitive quantification of MUC1 with a detection range of 1–12 ppb and a lower detection limit (0.62 ppb) [39].

Another electrochemical sandwich immunoassay for the ultra-sensitive detection of human MUC1 cancer biomarker using Protein G-functionalizedmagnetic beads (MBs) and graphite based screen printedelectrodes (SPEs)was developed.Magnetic beadswere employed as theplatforms for the immobilization and immunoreaction process. A pair ofprimary and secondary antibodies was used to capture the MUC1 pro-tein. After labeling with a third antibody conjugated with Horseradishperoxidase (HRP), the resulting conjugate was trapped on the graphitebased SPEs and MUC1 determination was carried out by differentialpulse voltammetry (DPV) at 0.4 V uponH2O2 additions usingAcetamin-ophen (APAP) as the redox mediator (Fig. 5).

In another assay alkaline phosphatase (AP) was used as a labelinstead of HRP, detecting in this case 1-naphtol produced during theenzymatic reaction of the substrate 1-naphtyl phosphate, by DPV. Thesensitivity of the two immunosensors was compared. A linear relation-ship was obtained for the detection of human MUC1 over a range of 2–

Fig. 4. Sandwich aptasensor preparation for MUC1 detection: (a) o-ABA polymerization; (b, c)carboxyl activation; (d) PABA free binding-sites blocked with 10 mM ethanolamine; (e) incubhuman MUC1 antigens (g) interaction of MB with G bases in aptamers; (h) DPV measurement

25 ppb with a lowest detection limit of 1.34 ppb when HRP was appliedas a label. When AP was used for the labeling of third antibody, a linearrange of 0–10 ppb was obtained, with a detection limit of−0.62, show-ing that this format of the assay can detect any concentration of MUC1starting from 0 ppb. Preliminary experiments were performed using dis-posable electrochemical sensors in order to optimize some parameters(i.e. incubation times, concentrations and blocking agent).

1.4. Human epididymis protein 4 (HE4)

HE4 is a promising biomarker detectable in serum. It could be usedfor the diagnosis and followup samples. Some reports concluded thatHE4 nearly matched CA125 if not better for early detection of ovariancancer [40].

HE4 is a lowmolecularweight glycoprotein and amember of the sta-ble 4-disulfide family. The exact functions of HE4 have not been charac-terized. HE4 is expressed primarily in epithelia of normal female genitaltissues [41,42] and is overexpressed in 93% of serious, 100% ofendometrioid and 50% of clear cell but not in mucinous or germ-cellovarian cancers [42]. Compared to CA125, HE4 is less frequentlyoverexpressed in some benign ovarian diseases, such as endometriosis[43–45].

There are several conventional methods used for HE4 quantificationincluding enzyme-linked immunosorbent assay (ELISA), immuno-radiometric assay (IRMA) and enzyme immunoassay (EIA). Despiteexhibiting high sensitivity and selectivity, these traditional immunoas-says still have some drawbacks, such as time consuming, the depen-dence of sophisticated and expensive equipment, and the demand forskilled professionals [46]. Although these methods have been proposedfor tumor markers detection, it is still a great challenge to use thesetraditional approaches for the detection of tumor markers at traceamount. Electrochemical immunoassays, which combine the specificity

MUC1 monoclonal mouse antibody immobilization onto PABA modified graphite SPEs byation with human MUC1 at different concentrations; (f) specific binding of aptamer withs [39].

Fig. 5. Sandwich immunosensor preparation for MUC1 detection: (a) primary antibody attachment on MBs surface; (b) MBs free binding-sites blocking using BSA or Milk powder;(c) incubation with MUC1 antigen solutions and Ab1/MUC1 Ag complex formation on the MBs; (d) reaction with Ab2 (e) (immunosensor) incubation with Ab3 labeled with HRP orAP; (f) DPV measurements of MBs-bound MUC1 in presence of acetaminophen/H2O2 and 1-naphtyl-phosphate, respectively.

134 I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

of immunoreactionswith the sensitivity of the electrochemical method,have been considered suitable for the determination of proteins. Com-pared with the conventional immunoassay methods, electrochemicalimmunosensors exhibit several advantages, such as high sensitivity,rapid analysis time, small analyte volume, and simple pretreatment.

For the detection of human epididymis-specific protein 4 (HE4)an immunosensor was developed by using a chitosan–titanium carbide(TiC) film thatwas first electrodeposited onto a tin-doped indiumoxide(ITO) electrode at a constant potential. Gold (Au) nanoparticleswere then electrodeposited on the surface of the chitosan–TiC filmby cyclic voltammetry (CV). The capture antibody (anti-HE4) wasadsorbed onto the Au and TiC nanoparticles. After a specific sandwichimmunoreaction among the capture antibodies, HE4 and biotinylatedsecondary antibody, biotinylated primer DNA was immobilized onthe secondary antibody by biotin–streptavidin system. Appropriateamounts of circular template DNA and biotinylated primer DNA wereused for rolling circle amplification (RCA) under optimal conditions.The RCA products provided a large number of sites to link DNA detec-tion probes. Doxorubicin hydrochloride intercalated the CG–GC stepsbetween the RCA products and the DNA detection probes, which wasmonitored by differential pulse voltammetry (DPV) based on the cur-rent signal of doxorubicin hydrochloride. With the above-mentionedamplification factors, the current responded to HE4 linearly in theconcentration range of 3–300 pM under optimal detection conditions,with a detection limit of 0.06 pM [47]. A combination of HE4 withCA125 and carcinoembryonic antigen (CEA) in an assay panel has beenproposed for detecting early stage ovarian cancer versus benign tumorsowing to its sensitivity of 86% [48]. For themoment only immunosensorsfor each biomarker were developed.

Interestingly, a study by Anastasi et al. showed that HE4 increased5–8 months before CA125 in relapsed ovarian cancer patients, indicat-ing that HE4 might be a better marker for monitoring ovarian cancerrelapse [49].

Prostasin is overexpressed in epithelial ovarian cancer and should beinvestigated further as a screening or tumor marker, alone and in com-bination with CA125.

Prostasin, also known as channel activating protease 1, is a serineprotease with trypsin-like substrate specificity. The preproenzymepossesses a C-terminalmembrane-spanningdomain that can beproteo-lytically processed to generate a secreted form of the enzyme. The pep-tidase activity of prostasin is involved in the regulation of epithelialsodium channels.

Immunosuppressive acidic protein assay is a potentially useful toolin the prognostic characterization of advanced ovarian cancer [50].

To our knowledge, none of those two biomarkers were used in thedevelopment of electrochemical immunoassays.

2. Biomarkers for breast cancer and their use in the developmentof immunoassay

2.1. Carcinoembryonic antigen (CEA)

Carcinoembryonic antigen (CEA), described in 1965, was among thefirst identified tumor antigens. CEA is a glycoprotein belonging to theimmunoglobulin family, which is detected in the serum of cancerpatients using radioimmunoassay or enzyme-linked immunosorbentassay. The clinical value of CEA detection is limited by a high false-positive rate in normal populations and by low diagnostic sensitivityand specificity. An elevated level of CEA is not specific for breast cancersince CEA can be found inmany different types of neoplasia. In breast tu-mors CEA is more prevalent in ductal compared to lobular carcinomas.CEA was found in patients with ductal carcinoma in situ, suggestingthat CEA is an earlymarker of the tumorigenic process [25]. FDA reportedalso CEA as approved biomarker from serum for the colon cancer in themonitoring phase [7].

Several approaches for the construction of immunosensors wereproposed. Mainly amperometric immunosensors were reported.

Ying Zhou et al. [51] reported a reagentless amperometric immuno-sensor based on antibody-embedded gold nanoparticles (nano-Au) andSiO2/Thionine nanocomposite self-assembled layers. L-cysteine (Cys)wasfirstmodified on a bare gold electrode in order to prepare a uniformorientation self-assembled monolayer (SAM) with functional groups of−NH2. Then, they used a linker, SiO2/Thionine nanocomposite, to con-struct two-double gold nanoparticles (nano-Au) layers. Lastly, theyadded carcinoembryonic antibody (anti-CEA) as an immunoreagent,immobilized onto the nano-Au, SiO2/Thionine nanocomposition andnano-Au self-assembled sandwiched layers. Thus, the detection modewas based on the change in the current response before and after the spe-cific binding of anti-CEA to CEA, due to the immunocomplex inhibitingthe access of redox probe [Fe(CN)6]4−/3− to electrode. In this con-figuration, the detection limit for carcinoembryonic antigen (CEA) was0.34 ng/mL [51]. Yan-Ru Yuan et al. [52] published another approachtowards the development of novel immunosensor based on goldnanoparticles (nano-Au) and nickel hexacyanoferrates nanoparticles(NiHCFNPs) for determination of CEA in clinical immunoassay. Thefabrication steps of the immunosensor described by authors were asfollows: firstly, nano-Au was immobilized on the surface of bare glassycarbon electrode (GCE) by using a simple method — electrochemicalreduction of HAuCl4 solution; secondly, NiHCFNPs as an electroactivesubstance were immobilized on the layer of gold nanoparticles; thirdly,nano-Au was again immobilized on the surface of NiHCFNPs, whichoffered a favorable microenvironment and biocompatibility to ad-sorb anti-CEA by chemical adsorption between nano-Au and −NH2 of

135I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

anti-CEA. The detection limit for CEAwas of 0.1 ng mL−1 at three timesbackground noise [52].

By using magnetic core/shell particles coated with self-assembledmultilayer of nanogold, a quick and reproducible electrochemical-based immunosensor technique has been developed.Magnetic particlesstructured from Au/Fe3O4 core–shells were prepared and aminatedafter a reaction between gold and thiourea. They performed additionalmultilayered coatings of gold nanoparticles on the surface of the core/shell particles. Furthermore, they attached, on the surface of solid paraf-fin carbon paste electrode (SPCE) containing an external magnetic field,modified magnetic particles containing anti-CEA antibody. The linearrange for the detection of CEA was from 0.005 to 50 ng mL−1 and thelimit of detection (LOD) was 0.001 ng mL−1. The LOD was approxi-mately 500 times more sensitive than that of the traditional enzyme-linked immunosorbent assay for CEA detection [53].

With a nanosilver-doped DNA polyion complex membrane (PIC) asa sensing interface it was possible to detect CEA at a concentrationof 10 pg mL−1. In order to develop this immunosensor configuration,a double-stranded DNA was initially assembled onto the surface ofThionine/Nafion-modified screen-printed carbon electrode to adsorbsilver ions with positive charges, then silver ions were reduced tonanosilver particleswith the aid of NaBH4, and then anti-CEA antibodieswere immobilized on the nanosilver surface. Detection of CEA with asandwich-type assay format used gold nanoparticles conjugated withhorseradish peroxidase-labeled with anti-CEA as signal antibodies.Under optimal conditions, the immunosensor exhibited a dynamicrange of 0.03–32 ng mL−1 with a low detection limit of CEA. Intra- andinter-assay variation coefficients (imprecision) were b9.5% and 6.5%, re-spectively. The response could remain 90.1% of the original current at30th day. 50 real samples were evaluated using the immunosensor andthe enzyme-linked immunosorbent assay respectively and received inaccordance with those two methods [54].

Usingmodified gold nanoparticlewith reduced graphene oxide nano-composite (GNP–THi–GR) a label-free immunosensor for the sensitivedetection of CEA was prepared. The nanocomposites with good biocom-patibility, excellent redox electrochemical activity and large surface areawere coated onto the glassy carbon electrode (GCE) surface and thenCEA antibody (anti-CEA) was immobilized on the electrode to constructthe immunosensor. CV and differential pulse voltammetry (DPV) studiesdemonstrated that the formation of antibody–antigen complexes de-creased the peak current of THi in the GNP–THi–GR nanocomposites.The decreased currents were proportional to the CEA concentration inthe range of 10–500 pg mL−1 with a detection limit of 4 pg mL−1 [55].Another immunoassay used protonated L-cysteine entrapped in Nafion(Nf) membrane by cation exchange function, forming Nf–Cys (cysteine)composite membrane, which was more stable, compact, biocompatibleand favorable formass and electron transfer comparedwithNffilm solely[56]. Gold (Au) nanoparticles were adsorbed onto the electrode surfaceby thiol groups on the composite membrane. Like this a nano-Au mono-layer was formed, onto which carcinoembryonic antibody was loaded toprepare carcinoembryonic antigen (CEA) immunosensor. In that case alinear concentration range of 0.01 to 100 ng mL−1 with a detectionlimit of 3.3 pg mL−1 (signal/noise = 3)was observed. The performancesof modified electrodes were investigated by cyclic voltammograms andelectrochemical impendence spectroscopy.

In another approach the glucose oxidase (GOD)-functionalizedhollow gold nanospheres encapsulating glucose oxidase was designedto label the ferrocene monocarboxylic-grafted secondary antibodiesfor highly sensitive detection of tumor marker using carboxyl groupfunctionalized multiwall carbon nanotubes as platform [57]. The ferro-cene used to label antibodies acted as a mediator of electron transferbetween GOD and electrode surface. Using CEA the designed tracershowed linear range from 0.02 to 5.0 ng mL−1 with the detectionlimit down to 6.7 pg mL−1 [57].

A sandwich-type immunoassay format that used bimetallic AuPtnanochains, synthesized through a mild chemical method, with anti-

horseradish peroxidase-conjugated anti-carcinoembryonic antigen(HRP-anti-CEA-NCAuPt) was developed for electrochemical detectionof carcinoembryonic antigen (CEA). The alloyed nanocrystals exhibitnot only sound signal amplification effect of Au nanoparticles, butalso electrical and structural properties arising from the disparateAuPt components. As a result, the electrochemical signal was signif-icantly amplified by using theHRP-anti-CEA-NCAuPt as tracer andhydro-gen peroxide as enzyme substrate. The linear range of the developedimmunosensor is 0.01–200 ng mL−1 and the detection limit is0.11 pg mL−1 of CEA [58].

The continuing discovery of cancer biomarkers necessitates im-proved methods for their detection. Molecular imprinting using artifi-cial materials provides an alternative to the detection of a wide rangeof substances. The surface molecular imprinting using self-assembledmonolayers to design sensing elements for the detection of cancer bio-markers and other proteins is of interest lately and many researchteams used MI materials for the development of electrochemical sen-sors. One way to create this type of elements consist of a gold-coatedsilicon chip onto which hydroxyl-terminated alkanethiol moleculesand template biomolecule are co-adsorbed, where the thiol moleculesare chemically bound to the metal substrate and self-assembled intohighly orderedmonolayers. The biomolecules can be removed, creatingthe foot-print cavities in the monolayer matrix for this kind of templatemolecules. Re-adsorption of the biomolecules to the sensing chipchanges its potential, which can be measured potentiometrically. Thismethodwas applied to the detection of CEA in both solutions of purifiedCEA and in the culturemediumof a CEA-producing human colon cancercell line. The CEA assay, validated also against a standard immunoassay,was both sensitive (detection range 2.5–250 ng mL−1) and specific(no cross-reactivity with hemoglobin; no response by a non-imprintedsensor) [59].

Au–TiO2 nanoparticles and multiple horseradish peroxidase (HRP)-labeled antibodies (HRP-Ab2) functionalized hollow Pt nanospheres(HPtNPs) (abbreviated as HRP-Ab2–HPtNPs) were used as the sensorplatform by M. Liang et al. [61]. Three-dimensional (3D) Au–TiO2

nanoparticles were uniformly assembled on a glassy carbon electrode.After capturing the target CEA, the HRP-Ab2–HPtNPs bioconjugateswere bound to the electrode surface via the formation of a sandwichcomplex. The electrochemical signals were amplified by HPtNPs andthe carried HRP towards the reduction of H2O2 using hydroquinone(H2Q) as the electron mediator. Under optimized conditions, the pro-posed immunosensor showed a high sensitivity and a wide linear rangefrom0.02 to 120 ng mL−1with a lowdetection limit of 12 pg mL−1 [60].

Besides the gold or magnetic nanoparticles, carbon based materialslike carbon nanotubes or graphene were used for the immobilization ofthe antibodies anti-CEA. Monoclonal anti-CEA antibodies covalentlyimmobilized on polyethylenaminewrappedmultiwall carbon nanotubesscreen-printed electrode were reported. CEA and αCEA tagged ferrocenecarboxylic acid encapsulated liposomes were used to achieve a sandwichimmunoassay and faradic redox responses of the released ferrocenecarboxylic acid from the immunoconjugated liposomes on the electrodesurface were analyzed by using square wave voltammetry. The magni-tude of the SWV peak current was directly related to the concentrationof CEA, with a detection limit of 10−12 g mL−1 (S/N = 3) [61].

Y. Xiang [62] describes the preparation and use of multi-enzymelayer-by-layer (LBL) assembled single wall carbon nanotube (SWCNT)composite labels to amplify ultrasensitive electrochemical detection ofCEA. The target protein was sandwiched between an electrodesurface-confined capture anti-CEA antibody and the secondary signalanti-CEA/enzyme-LBL/SWCNT bioconjugate. In this case the detectionlimitwas 0.04 pg mL−1. Another feasible and practicable amperometricimmunoassay strategy for sensitive screening of CEA in human serumused carbon nanotube (CNT)-based symbiotic coaxial nanocables as la-bels. To construct such a nanocable, a thin layer of silica nanoparticleswas coated on the CNT surface by sonication and sol–gel methods, andthen colloidal gold nanoparticles were assembled on the amino-

136 I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

functionalized SiO2/CNTs, which were used for the label of horseradishperoxidase-anti-CEA conjugates (HRP-anti-CEA-Au/SiO2/CNT). On ananti-CEA/Au/Thionine/Nafion-modified glassy carbon electrode usingHRP-anti-CEA-Au/SiO2/CNTs as detection antibodies they detected theformation of the sandwich-type immunocomplex in the presence of an-alyte CEA. Under optimal conditions, the cathodic peak currents of theelectrochemical immunosensor were proportional to the logarithm ofCEA concentration over the range from 0.01 to 12 ng mL−1 in pH 5.5HAc-NaAc containing 5 mMH2O2. At a signal-to-noise ratio of 3, the de-tection limit was 5 pg mL−1 CEA. In addition, the technique was evalu-ated by spiking CEA standards in pH 7.4 PBS and with 35 clinical serumspecimens, the resultswere in excellent accordancewith those obtainedwith the commercially available electrochemiluminescent enzyme-linked immunoassay [63].

The mixture between gold nanoparticles (GNPs), Azure and multi-wall carbon nanotubes (MWCNT) to form a self-assembling nanocom-posite was used by J.B. Zheng [64]. In order to optimize theimmunosensor, MWCNT was first dispersed in Nafion (Nf) to obtain ahomogeneous solution and then it was dropped on the surface of thegold electrode (Au). Then the positively-charged redox molecule, AzureI, was entrapped into MWCNT–Nf film to form a redox nanostructuralmembrane. Next, the negatively charged gold nanoparticles (GNPs)were assembled to the interface through the electrostatic force. Finally,carcinoembryonic antibody molecules could be absorbed into the GNPs/Azure /MWCNT–Nf immobilization matrix [64].

A sandwich-type immunoassay format was designed by Zh. Li et al.,in order to quantify carcinoembryonic antigen by using nanogold-entrapped graphene nanocomposites (NGGNs) as trace labels in clinicalimmunoassays [65]. The device consisted of a glassy carbon electrodecoated with Prussian Blue (PB) on whose surface gold nanoparticleswere electrochemically deposited to the further modified with the spe-cific analyte-capturing molecule, anti-CEA antibodies. The immunoas-say was performed using horseradish peroxidase (HRP)-conjugatedanti-CEA as secondary antibodies attached on the NGGN surface (HRP-anti-CEA-NGGN). The method using HRP-anti-CEA-NGGNs as detectionantibodies exhibits a dynamic working range of 0.05–350 ng mL−1

with a low detection limit of 0.01 ng mL−1 CEA (at 3 s) [66].An interesting example is the label-free immunoassay based on

impedimetric measurements developed by Tang et al. [67] for CEA detec-tion. Carcinoembryonic antibody was covalently attached to AuNPs andthe composite immobilized on Au electrode by electrocopolymerizationwith o-aminophenol. Electrochemical impedance spectroscopy studiesdemonstrated that the formation of antibody–antigen complexes in-creased the electron-transfer resistance of [Fe(CN)6]3−/4− redoxprobe at the poly-o-aminophenol/carcinoembryonic antibody-AuNPs/Au electrode, thus monitoring of carcinoembryonic antigen concentra-tion could be performed with a detection limit of 0.1 ng mL−1.

An interesting approach used multilayer films of Prussian blue(PB) and multiwalled-carbon nanotube/polyethylenimine/Au (MWNT–PEI–Au) nanocomposite deposited directly onto the surface of a glassycarbon electrode. Then a layer of chitosan mixed with gold nanoparticleswas cast onto the surface of the electrode. Subsequently, the electrodewas coated with a primary antibody (Ab1) and blocked with BSA.The fabricated immunosensor exhibited a good response to CEA, with adetection range from 0.5 to 160 ng mL−1 and a detection limit of0.08 ng mL−1 [68].

2.2. BRCA1 (breast cancer type 1 susceptibility protein)

BRCAl is an anti-oncogene inwomenwho are genetically predisposedto breast and ovarian cancer. The detection of BRCA1 can offer an oppor-tunity to characterize the function of genetic features in breast and ovar-ian cancer and to screen breast or ovarian cancer patients. BRCA1promoter methylation has been extensively studied in ovarian cancer.Studies have demonstrated that BRCA1 hypermethylation occurs in10–15% of sporadic cases and is associated with loss of expression and

with the serous histotype [69]. In addition, BRCA1 methylation is signifi-cantly correlated with malignancy. Taken together, these studies suggestthat BRCA1 promoter methylation may serve as a biomarker for aggres-sive EOC and response to chemotherapeutic intervention. For the mo-ment the interest in developing immunosensors for BRCA1 is not high.

Horseradish peroxidase was entrapped in the pores of amino-groupfunctionalized SBA-15 and the secondary antibody (Ab2) combinedwith SBA-15 by covalent bond. Ionic liquid (IL) was added into themixed solution of SBA-15/HRP/Ab2 and application of IL increased theelectrochemical activity of HRP and promoted electron transport.Under optimal conditions, the electrochemical immunoassay exhibiteda wide working range from 0.01 to 15 ng mL−1 with a detection limitof 4.86 pg mL−1 BRCA1. The precision, reproducibility, and stability ofthe immunoassay were acceptable [70].

2.3. Human epidermal growth factor receptor 2 (HER2)

HER2/neu protein, a product of c-erbB-2 oncogene, belongs to thefamily of four transmembrane receptor tyrosine kinases, called theepidermal growth factor receptor (EGFR) family (by the best knownrepresentative of the family). The main characteristics of receptor tyro-sine kinases are transmembrane location and obligatory interactionwith an appropriate ligand for the realization of kinase activity and sub-sequent biological effects. EGFR family receptors form homo- andheterodimers upon activation. In many cases, structures containingtype 2 receptor HER2/neu, which have no specific ligands, are most ac-tive. Hence, HER2/neu is a unique dispatcher receptor that does not in-teract with any of the known growth factors activating relatedreceptors, but is a key component in transmission of mitogenic signalsof all EGF-like peptides [71].

The HER2/neu proto-oncogene is amplified and/or overexpressed inapproximately 20 to 25% of invasive breast cancers [72]. HER2/neuoverexpression has been associated with a poor rate of disease-freesurvival. The role of HER2/neu as a predictive marker of response tohormone therapy and chemotherapy is controversial. The extracellulardomain (ECD) of the HER2/neu protein is frequently cleaved andreleased into the circulation. Rising serum HER2/neu concentrationshave been associated with progressive metastatic disease and poorresponse to chemotherapy and hormonal therapy [73].

Extracellular domain of HER2 has been the target of several mono-clonal antibodies created in order to inhibit the proliferation of humancancer cells. The most popular trastuzumab is a humanizedmonoclonalantibody consisting of two antigen-specific sites that bind to the just amembrane portion of the extracellular domain of the HER2 receptorpreventing the activation of its intracellular tyrosine kinase.

Current diagnostic tests for HER2 involve analysis of tumor cellsfor either amplification of the HER2 gene using fluorescent in situhybridization (FISH) or immunohistochemistry (IHC) to determine theexpression of the receptor within the cell membrane [74]. FISH usesfluorescent labeled probes for regionswithin theHER2 gene and follow-ing observation with a fluorescent microscope, amplification of thegene can be deduced based on the number of signals seen. Both proce-dures are complex, involve time consuming steps and require speciallytrained personnel to carry out the procedures. A very recent review hasappeared on available technologies for the detection of HER2 [75].

Electrochemical immunosensors have been recently reported for themeasurement of the ECD of HER2.

Gohring et al. developed sensitive immunosensor based on piezo-electric microcantilever sensors (PEMS) [76]. They have functionalizedPEMS with antibodies that specifically bind to HER2. The function andsensitivity of these anti-HER2 PEMS biosensors was initially assessedusing recombinant HER2 spiked into human serum. Their ability to de-tect native HER2 present in the serumof breast cancer patientswas thendetermined. The anti-HER2 PEMS were able to accurately detect bothrecombinant and naturally occurring HER2 at clinically relevant levels

137I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

(N2 ng mL−1). This indicates that PEMS-based biosensors provide apotentially effective tool for biomarker detection.

An electrochemical immunosensor for the detection ofHER2has beenproposed based on gold nanoelectrode ensembles (NEE). NEEs werefunctionalized with trastuzumab which interacts specifically with HER2.The biorecognition process was completed by adding a primary antibodyand a secondary antibody labeledwith horseradishperoxidase. Hydrogenperoxide was added to modulate the label electroactivity; methylenebluewas the redoxmediator generating voltammetric signals. NEEs func-tionalized with trastuzumab were tested to detect small amounts ofHER2 in diluted cell lysates and tumor lysates [77].

Marrazza et al. developed an immunoassay based on a sandwich for-mat in which a primary monoclonal antibody anti-HER2 is coupledto protein A modified magnetic beads. The modified beads are used tocapture the protein from the sample solution and a sandwich assay isperformed by adding a secondary monoclonal antibody anti-HER2 la-beled with biotin. The enzyme alkaline phosphatase (AP) conjugatedwith streptavidin and its substrate (1-naphthyl-phosphate) are thenused for the electrochemical detection by differential pulse voltammetry(DPV). The serum samples from hospital patients were analyzed byimmunosensor developed [78].

Table 2 presents an overview of themain biomarkers for ovarian andbreast cancer and the developed immunoassays.

3. Multiplex immunosensors

Multiple biomarkers design became lately of interest especially fordiagnosis of cancer.

Tang et al. have reported a novel method to detect tumor markerssuch as α-fetoprotein (AFP), carcinoembryonic antigen (CEA), cancerantigen 125 (CA125), and CA15-3, that can be found in the body (usu-ally blood or urine) in cancer patients. They synthesized magnet core/shell NiFe2O4/SiO2 nanoparticles and fabricated an electrochemical

Table 2A list of biomarkers used in the development of electrochemical immunoassay for ovarian and

Biomarker Disease Type of immunoassay

CA125 Ovarian [Co(bpy)3]3+/MWNTs–NafionAu microspheres and porous polySPE with Au nanoparticles

CA 15-3 OvarianBreast

(Fc-COOH)-doped silica nanopartCNTs and core-shell organosilicaaGrapheneSandwich type with MNP and APAu NP with aptamers

HE4 Ovarian chitosan–titanium carbide (TiC) fiCEA Breast nano-Au and SiO2/Thionine nano

nano-Au and nickel hexacyanofeAu MP attached on a CPEnanosilver-doped DNA polyion coAu NP –grapheneprotonated L-cysteine entrappedAu NP functionalized with GOD aHRP-anti-CEA-NCAuPtMIPAu–TiO2 nanoparticles and multi(HRP-Ab2) functionalized hollowPolyethylene-MWCNT- ferroceneEnzyme LBL-SWCNTCEA/Au/Thionine/Nafion-modifieAu NP – azure I-MWCNTAu NP and aminophenol label freMulti layer of Prussian Blue and m

BRCA 1 Breast/ovarian IL-HRPHer2 Breast Piezoelectroc cantilever

Au NP with trastuzumabProtein A magnetic beads

Abbreviations: MWCNT-multiwall carbon nanotubes, GCEs — glassy carbon electrodes, SPE —

phosphatase, HRP — horse radish peroxidase, ITO — tin-doped indium oxide, CPE — carbon pnanotubes, IL— ionic liquid.

magnetic controlled microfluidic device. The immunoassay systemconsisted of 5 working electrodes and an Ag/AgCl reference electrodeintegrated on a glass substrate. Each working electrode contained adifferent antibody immobilized on the NiFe2O4/SiO2 nanoparticlesurface and was capable of measuring a specific tumor markerusingnoncompetitive electrochemical immunoassay. Under optimal con-ditions, the multiplex immunoassay enabled the simultaneous detectionof 4 tumor markers. The sensor detection limit was b0.5 μg L−1 for mostanalytes [79].

A gold interdigitated (GID) capacitor transducer wasmodified usingmagnetic beads (MB) ant tha obtained platform was initially testedusing C-reactive protein (CRP) as the model analyte. The methodologywas then transferred for multiple marker detection with the aim ofprecise disease diagnostics. For the first time, the protein biomarkersof lung cancer including carcinoembryonic antigen (CEA), epidermalgrowth factor receptor (hEGFR) and cancer antigen 15-3 (CA15-3)were investigated with a capacitive sensor. The threshold levels of themarkers to indicate the cancer are higher than 5 ng mL−1 (CEA),64 ng mL−1 (hEGFR) and 50 U mL−1 (CA15-3), respectively. CEAand hEGFR could successfully be detected in the concentration rangeof 5 pg mL−1 to 1 ng mL−1 while CA15-3 was detected in the rangeof 1–200 U mL−1 with a high specificity. The study demonstrated ahighly specific capacitive immunoassay, presenting a potentialalternative tool for early and precise diagnosis of cancer disease [80].

Taking into consideration the complexity of the analyzed matrix(blood or serum) the detection of several cancer biomarkers by usingthe same analytical device it will simplify the diagnosis and would pro-vide more information in less time. But the differences in the thresholdlevels of cancer biomarkers in the serum make for the moment a littlebeat difficult the development of a sensors arrays for 4, 8 or 12biomarkers. It will be the challenge of the researchers in the future.

The same team of D. Tang et al. designed a multiplexed strippingvoltammetric immunoassay for the simultaneous detection of three

breast cancer.

References

thionine modified GCEs[30][32][33]

icles (SNPs)nd chitosan nanospheres

/HRP

[36][37][38][39]

lm that was first electrodeposited onto ITO [47]compositerrates nanoparticles

mplex membrane (PIC)

in Nafion (Nf) membranend ferrocene monocarboxylic-grafted secondary antibodies

ple horseradish peroxidase (HRP)-labeled antibodiesPt nanospheres (HPtNPs)carboxylic acid encapsulated liposomes

d glassy carbon electrode

eultiwalled-carbon nanotube/polyethylenimine/Au

[51][52][53][54][55][56][57][58][59][60][61][62][63][64][67][68]

[70][76][77][78]

screen printed electrodes, Fc — ferrocene, MP — mangnetic nanoparticles, AP — alkalyneaste electrode, GOD — glucose oxide, LBL — layer by layer, SWCNT — single wall carbon

138 I. Diaconu et al. / Clinica Chimica Acta 425 (2013) 128–138

mucins involved in the breast cancer: CA125, CA15-3 and CA19-9. Theprimary monoclonal antibodies for the three mentioned mucins wereimmobilized on a single magnetic bead and the CdS, ZbS and PbS wereused as labels for the detection antibodies. Themultiplexed immunoas-saywas able to detect simultaneous in a single run the three biomarkerswith a limit of detection of 0.005 U mL−1 [81].

4. Conclusions

Recent advances in molecular biology elucidate that cancer bio-markers play an important role in diagnosis, prognosis and providinginsights into the etiology of cancer. The widespread use of tumormarkers in healthcare will ultimately depend upon the detection ofmany tumor markers with high selectivity and sensitivity. Despite themultitude of different parameters currentlymeasured in the clinical lab-oratory, only aminor part of these aremeasured bymeans of biosensor-basedmethods. Analyzers are dominatingwithin the group of biochem-ical sensors, enzyme ormetabolic sensors as integrated devices in clinic.

Modern electrochemical bioaffinity sensors, such as DNA- orimmunosensors, offer remarkable sensitivity, essential for early cancerdetection. The coupling of electrochemical devices with nanoscalematerials offers a unique multiplexing capability for simultaneousmeasurements of multiple cancer markers. The attractive properties ofelectrochemical devices are extremely promising for improving theefficiency of cancer diagnostics and therapy monitoring. With furtherdevelopment and resources, such portable devices are expected tospeed up the diagnosis of cancer, making analytical results available atpatient bedside or physician office within few minutes.

In this review we present the multitude of assays that are availablefor cancer biomarkerswith the aimof diagnosing the breast and ovariancancer. It is for sure that the researches in this dynamic field involveother biomarkers that are not commercially available yet. Many re-searchers investigate the possibility of using arrays of immunosensorsin order to create a diagnosis tool with a high degree of confidence.

Conflict of interest

There are no competing financial interests.

Acknowledgments

The authors are grateful for the financial support from the RomanianNational Authority for Scientific Research, CNCS–UEFISCDI, projectnumber PN-II-ID-PCE-2011-3-0355.

References

[1] Ullah MF, Aatif M. Cancer Treat Rev 2009;35(3):193–200.[2] Liang S-L, Chan DW. Clin Chim Acta 2007;381:93–7.[3] Gion M, Daidone MG. Eur J Cancer 2004;40:2613–22.[4] Dunn BK, Wagner PD, Anderson D, Greenwald P. Semin Oncol 2010;37(3):224–42.[5] Rasooly A, Jacobson J. Biosens Bioelectron 2006;21:1851–8.[6] Havrilesky LJ, Whitehead CM, Rubatt JM, et al. Gynecol Oncol 2008;110:374–82.[7] Soper SA, Brown K, Ellington A, et al. Biosens Bioelectron 2006;21:1932–42.[8] Polanski M, Anderson NL. Biomark Insights 2006;1:1–48.[9] Laschi S, Bulukin E, Palchetti I, Cristea C, Mascini M. ITBM-RBM 2008;29(2-3):202–7.

[10] Wang J. Biosens Bioelectron 2006;21:1887–92.[11] Geim AK, Novoselov KS. Nat Mater 2007;6(3):183–91.[12] Gupta AK, Gupta M. Biomaterials 2005;26(18):3995–4021.[13] Howarth M, Liu W, Puthenveetil S, et al. Nat Methods 2008;5(5):397–9.[14] Lin JH, Ju HX. Biosens Bioelectron 2005;20(8):1461–70.[15] Dwarakanath S, Bruno JG, Shastry A, et al. BiochemBiophys Res Commun2004;325(3):

739–43.[16] Oswald B, Lehmann F, Simon I, Terpetschnig E,Wolfbeis OS. Anal Biochem 2000;280:

272–7.[17] Luppa BPB, Sokoll LJ, Chan DW. Clin Chim Acta 2001;314:1–26.[18] Stefan RI, Van Staden JF, Aboul-Enein HY. Fresenius J Anal Chem 2000;366:659–68.

[19] Vlad AM, Diaconu I, Gantt KJ. Immunol Res 2006;36(1–3):229–36.[20] Budiu RA, Diaconu I, Chrissluis R, Dricu A, Edwards RP, Vlad AM. Dis Model Mech

2009;2(11–12):593–603.[21] Deng J, Wang L, Chen H, et al. Cancer Metastasis Rev 2013. http://dx.doi.org/

10.1007/s10555-013-9423-y.[22] Magdelenat H, Gerbault-Seureau M, Laine-Bidron C, Prieur M, Dutrillaux B. Breast

Cancer Res Treat 1992;22:119–27.[23] Mok S, Chao C, Skates JS, et al. J Natl Cancer Inst 2001;93(19):1458–64.[24] Suh KS, Park SW, Castro A, et al. Expert Rev Mol Diagn 2010;10(8):1069–83.[25] Levenson VV. Biochim Biophys Acta Gen Subj 2007;1770(7):847–56.[26] Jarrard JA, Linnoila RI, Lee H, Steinberg SM, Witschi H, Szabo E. Cancer Res 1998;58:

5582–9.[27] Husseinzadeh N. Gynecol Oncol 2011;120:152–7.[28] Gadducci A, Cosio S, Tana R, Riccardo Genazzani A. Crit Rev Oncol Hematol 2009;69:

12–27.[29] Su Z, Graybill WS, Zhu Y. Clin Chim Acta 2013;415:341–5.[30] Chen SH, Yuan R, Chai YQ, Min LG, Li WJ, Xu Y. Electrochim Acta 2009;54:7242–7.[31] Suwansa-ard S, Kanatharana P, Asawatreratanakul P, Wongkittisuksa B, Limsakul C,

Thavarungkul P. Biosens Bioelectron 2009;24:3436–41.[32] Fu XH. Electroanalysis 2007;19:1831–9.[33] Ravalli A, Pilon dos Santos G, Yamanaka H, Marrazza G. Sens Actuators B 2013;179:

194–200.[34] de Stefano I, Zannoni GF, Prisco MG, et al. Gynecol Oncol 2011;122(3):573–9.[35] Gadducci A, Cosio S, Riccardo Genazzani A. Crit Rev Oncol Hematol 2011;80:181–92.[36] Hong CL, Yuan R, Chai YQ, Zhuo Y. Anal Chim Acta 2009;633:244–9.[37] Li WJ, Yuan R, Chai YQ, Chen SH. Biosens Bioelectron 2010;25(11):2548–52.[38] Li H, He J, Li S, Turner APF. Biosens Bioelectron 2013;43:25–9.[39] Florea A, Taleat Z, Cristea C, Marrazza G, Mazloum-Ardakani M, Săndulescu R.

Electrochem Commun 2013;33:127–30.[40] Nguyen L, Cardenas-Goicoechea SJ, Gordon P, et al. Womens Health (Lond Engl)

2013;9(2):171–85.[41] Hellstrom I, Raycraft J, Hayden-Ledbetter M. Cancer Res 2003;63:3695–700.[42] Drapkin R, von Horsten HH, Lin Y. Cancer Res 2005;65:2162–9.[43] Moore RG, Miller MC, Steinhoff MM, et al. Am J Obstet Gynecol 2012;206(4):351–8.[44] Ono K, Tanaka T, Tsunoda T. Cancer Res 2000;60:5007–11.[45] Welsh JB, Zarrinkar PP, Sapinoso LM. Proc Natl Acad Sci U S A 2001;98:1176–81.[46] Kwon H-R, Lee KW, Dong ZG, Lee KB, Oh S-M. Biochem Biophys Res Commun

2010;391(1):830–4.[47] Lu LS, Liu B, Zhao ZhH, et al. Biosens Bioelectron 2012;33(1):216–21.[48] Yurkovetsky Z, Skates S, Lomakin A, et al. J Clin Oncol 2010;28(13):2159–66.[49] Anastasi E, Marchei GG, Viggiani V, Gennarini G, Frati L, Reale MG. Tumour Biol

2010;31:113–9.[50] Scambia G, Foti E, Ferrrandina G, et al. Am J Obstet Gynecol 1996;175:1606–10.[51] Zhuo Y, Yu RJ, Yuan R, Chai YQ, Hong CL. J Electroanal Chem 2009;628:90–6.[52] Yuan Y-R, Yuan R, Chai Y-Q, Zhuo Y, Miao X-M. J Electroanal Chem 2009;626:6–13.[53] Li JP, Gao HL, Chen ZhQ, Wei XP, Yang CF. Anal Chim Acta 2010;665(1):98–104.[54] Wu W, Yi P, He P, et al. Anal Chim Acta 2010;673(2):126–32.[55] Kong F-Y, Xu M-T, Xu J-J, Chen H-Y. Talanta 2011;85(5):2620–5.[56] Liao YH, Yuan R, Chai YQ, Zhuo Y, Yang X. Anal Biochem 2010;1:47–53.[57] Song ZhJ, Yuan R, Chai YQ, et al. Biosens Bioelectron 2011;26(5):2776–80.[58] Cao X, Wang N, Jia S, Guo L, Ke L. Biosens Bioelectron 2013;39:226–30.[59] Wang YT, Zhang ZhQ, Jain V, et al. Sensors Actuators B Chem 2010;146(1):381–8.[60] Yang HC, Yuan R, Chai YQ, et al. Biochem Eng J 2011;56(3):116–24.[61] Viswanathan S, Rani C, Anand AV, Ho JA. Biosens Bioelectron 2009;24:1984–9.[62] Xiang Y, Zhang YY, Jiang BY, Chai YQ, Yuan R. Sensors Actuators B Chem 2011;155(1):

317–22.[63] Li QF, Tang DP, Tang J, Su BL, Huang JX, Chen GN. Talanta 2011;84(2):538–46.[64] Sun A-L, Chen G-R, Sheng Q-L, Zheng J-B. Biochem Eng J 2011;57:1–6.[65] Li Zh, Tang Zh, Tang DP, et al. Biosens Bioelectron 2010;25(10):2379–83.[66] An HZ, Yuan RT, Tang DP, Chai Y, Li N. Electroanalysis 2007;19:479–86.[67] Tang H, Chen JH, Nie LH, Kuang YF, Yao SZ. Biosens Bioelectron 2007;22:1061–7.[68] Zhang Y, Chen H, Gao X, Chen Z, Lin X. Biosens Bioelectron 2012;35(1):277–83.[69] Gloss BS, Samimi G. Cancer Lett 2013 Available on line 12 January 2012. http://dx.

doi.org/10.1016/j.canlet.2011.12.036 [in press].[70] Cai YY, Li H, Du B, et al. Biomaterials 2011;32(8):2117–23.[71] Kushlinskii NE, Shirokii VP, Gershtein ES, Yermilova VD, Chemeris Yu, G, Letyagin VP.

Bull Exp Biol Med 2007;143(4):449–51.[72] Pauletti G, Dandekar S, Rong H, et al. J Clin Oncol 2000;18(21):3652–4.[73] Breast Cancer Res 2005;7:R436–43. http://dx.doi.org/10.1186/bcr1020 [http://

breast-cancer-research.com/content/7/4/R436].[74] Tse C, Brault D, Gligorov J, et al. Clin Chem 2005;51:1093–101.[75] Moelans CB, de Weger RA, Van der Wall E, van Diest PJ. Crit Rev Oncol Hematol

2011;80(3):380–92.[76] Loo L, Capobianco JA, Wu W, et al. Anal Chem 2011;83:3392–7.[77] Pozzi Mucelli S, Zamuner M, Tormen M, Stanta G, Ugo P. Biosens Bioelectron

2008;23:1900–3.[78] Al-Khafaji QAM, Harris M, Tombelli S, et al. Electroanalysis 2012;24(4):735–42.[79] Tang D, Yuan R, Chai Y. Clin Chem 2007;53(7):1323–9.[80] Altintas Z, Saravan Kallempudi S, Sezerman U, Gurbuz Y. Sens Actuators B 2012;174:

187–94.[81] TangD,Hou L, Niessner R, XuM,Gao Z, KnoppD. Biosens Bioelectron 2013;46:37–43.