Ea

H

a

b

c

a

ARR1AA

KIQBCN

1

adCiotisbeihbt

0d

Analytica Chimica Acta 713 (2012) 50– 55

Contents lists available at SciVerse ScienceDirect

Analytica Chimica Acta

jo u rn al hom epa ge: www.elsev ier .com/ locate /aca

lectrochemical immunoassay of cotinine in serum based on nanoparticle probend immunochromatographic strip

ungchi Niana,b, Jun Wanga, Hong Wua, Jiunn-Guang Lob, Kong-Hwa Chiuc, Joel G. Poundsa, Yuehe Lina,∗

Pacific Norwest National Laboratory, Richland, WA 99352, USADepartment of Biomedical Engineering and Environmental Sciences, National Tsing Hua University, Hsinchu, 30013, TaiwanDepartment of Applied Science, National DongHwa University, Hualien, 970, 30013, Taiwan

r t i c l e i n f o

rticle history:eceived 4 August 2011eceived in revised form0 November 2011ccepted 12 November 2011vailable online 22 November 2011

a b s t r a c t

A disposable sensor for the determination of cotinine in human serum was developed based onimmunochromatographic test strip and quantum dot label. In this assay, cotinine linked with quantumdot competes with cotinine in sample to bind to anti-cotinine antibody in the test strip and the quantumdots serve as signal vehicles for electrochemical readout. Some parameters governing the performance ofthe sensor were optimized. The sensor shows a wide linear range from 1 ng mL−1 to 100 ng mL−1 cotininewith a detection limit of 1.0 ng mL−1. The sensor was validated with spiked human serum samples and

eywords:mmunochromatographic stripuantum dotsiomarkerotinineanotechnology

it was found that this method was reliable in measuring cotinine in human serum. The results demon-strate that this sensor is rapid, accurate, and less expensive and has the potential for point of care (POC)detection of cotinine and fast screening of tobacco smoke exposure.

© 2011 Elsevier B.V. All rights reserved.

. Introduction

Cotinine (1-methyl-5-(3-pyridinyl)-2-pyrrolidinone, Fig. 1) is metabolite of nicotine (3-[(2S)-1-methylpyrrolidin-2-yl] pyri-ine), which can serve as a biomarker of tobacco smoke exposure.otinine has an in vivo half life of approximately 20 h, and is typ-

cally detectable from several days up to one week after the usef tobacco [1]. The level of cotinine in the blood is proportionalo the amount of exposure to tobacco smoke, so it is a valuablendicator of tobacco smoke exposure, including secondary (pas-ive) tobacco smoke exposure. Active smokers almost always havelood cotinine levels higher than 10 ng mL−1, while non-smokersxposed to low levels of environmental tobacco smoke (ETS) typ-cally have blood concentrations less than 1 ng mL−1. Following

eavy exposure to ETS, non-smokers can have blood cotinine levelsetween 1 and 10 ng mL−1 [2]. In previous studies, it is speculatedhat serum cotinine level was the predictor of lung cancer risk

∗ Corresponding author. Tel.: +1 509 371 6241; fax: +1 509 371 6955.E-mail address: yuehe.lin@pnnl.gov (Y. Lin).

003-2670/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.aca.2011.11.028

[3,4]. Compared with subjects with a cotinine level of less than5 ng mL−1, the odds ratio of lung cancer was increasing linearly,reaching 55.1 among individuals with a serum cotinine level ofmore than 378 ng mL−1 [4]. Therefore, there is a growing interestto develop fast, simple, sensitive, low-cost, and effective analyti-cal methods or devices to analyze cotinine in human body fluids.Analysis of cotinine in human body fluid samples is routinelycarried out using analytical techniques, such as gas chromatogra-phy [5], high performance liquid chromatography [5,6], capillaryelectrophoresis [6], gas chromatography–mass spectrometry [7],high performance liquid chromatography–mass spectrometry [3].Such analyses are generally performed at centralized laboratories,requiring extensive labor and analytical resources, and often resultin a lengthy analytical time. These analysis methods have a numberof disadvantages that limit their applications primarily to labora-tory settings and prohibit their use for rapid analyses in the field.Immunoassays are gaining a position in the analysis of food, agro-

chemicals, and environmental samples because of their simplicity,cost-effectiveness, and high sample throughout [8]. Traditionalenzyme-linked immunosorbent assays (ELISA) [5,9] and ELISA kits(Calbiotech) have been used to detect cotinine. A conventional

H. Nian et al. / Analytica Chimic

EacnTtEtai[iica[swueaoacotcteit

subrsuqsr

mttlpoeds

resulting solution was shaked for 30 min and concentrated with

Fig. 1. Structure of cotinine and trans-4-cotinine-carboxylic acid.

LISA assay for cotinine is based on competitive immunoreactionsmong cotinine, cotinine–enzyme conjugate, and immobilizedotinine antibodies on plates. Cotinine–enzyme and analyte coti-ine bind competitively to a fixed amount of cotinine antibodies.he activity of the captured enzyme tracers was monitored elec-rochemically or optically [5,9,10]. The major disadvantages ofLISA are the relatively long analysis time and complexity due tohe multiple-step processes involving antibody–antigen complex-tion, washing, and the addition of reagents [11]. Therefore, themmunosensor is an alternative tool to replace the traditional ELISA11]. On the basis of a specific reaction of the antibody and antigen,mmunosensors provide a sensitive and selective tool for determin-ng immunoreagents. Immunochromatographic strip tests (ISTs)oupled chromatography with conventional immunoassay offer

new analytical tool for protein analysis and clinical diagnosis12–16]. It has been widely used as an in-field and POC diagno-is tool to detect and identify biomarkers of diseases and biologicalarfare agents [17–20]. The advantages of such ISTs include theser-friendly format, the very short analysis time, less interfer-nce due to chromatographic separation, long-term stability over

wide range of climates, and a relatively low cost. The early stagef ISTs uses visible judgment for qualitative or semi-quantitativenalysis of the analytes. However, some of these tests are not suffi-iently sensitive or specific for accurate POC use. The visual readoutf the strip is usually limited to qualitative detection. Therefore,he quantitative ISTs are highly desirable and can offer accurateoncentration information for the analytes. The electrochemicalechniques have employed to develop the quantitative ISTs becauselectrochemical readers were less expensive, highly sensitive, anddeally suited for meeting the portability requirements of decen-ralized POC testing or field detection of bioagents [13,14].

Recently, nanomaterial-based electrochemical immunosen-ors/bioassays have attracted considerable interest because of theirnique physical and chemical properties and various nanomaterial-ased amplification approaches/platforms have been recentlyeported [21–26]. Among them, quantum dots (QDs), a core–shelltructure with CdS as the core and ZnS as the shell, have beensed as labels for immunosensors/bioassays [27–29]. However,uantum dot-based electrochemical immunoassay for detection ofmall molecules such as cotinine combined with IST has not beeneported so far.

In this paper, we present a disposable competitive immunochro-atographic sensor for sensitive detection of cotinine in serum. In

his approach, the assay combines conventional ISTs with an elec-rochemical readout technique, and highly sensitive quantum dotabels. The QD–cotinine conjugates were synthesized and somearameters governing the performance of the sensor have beenptimized (e.g. the fabrication of strip and experimental param-ters). This sensor shows a wide dynamic range and has low

etection limit for cotinine. The sensor was validated with spikederum samples and obtained satisfactory recovery.

a Acta 713 (2012) 50– 55 51

2. Materials and methods

2.1. Materials

Sheep polyclonal cotinine antibody was obtained from Ameri-can Research Products, Inc (Belmont, MA, USA). Cotinine, cotinine-carboxylic acid, phosphate buffer saline (PBS), bovin serum albumin(BSA), N-Hydroxy-succinimide (NHS), 2-morpholinoethane sul-fonic acid (MES), Tween-20, glycine, and atomic absorptionstardand solution were purchased from Sigma–Aldrich. 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and polyethyleneglycol 6000 (PEG 6000) were purchased from Fluka. Nitrocellulosemembrane, absorbent pad and glass fiber were purchased fromMillipore. PD-10 columns were purchased from GE Healthcare UKLimited. QD 655 (CdS@ZnS) was purchased from Molecular Probes,Inc (Eugene, OR, USA). Standard human serum was purchased fromGolden West Biologicals, Inc (Temecula, CA, USA). All chemicalsused in this study were analytical reagent grade. All stock solutionswere prepared using deionized water purified with the NanopureSystem (Barnstead, Kirkland, WA, USA).

2.2. Instruments and methods

Square-wave voltammetric (SWV) measurements were per-formed with an electrochemical analyzer CHI 660A (CH Instru-ments, Austin, TX, USA) operated with a personal computer. Allelectrochemical measurements were carried out with a dispos-able screen-printed electrode (SPE) consisting of a carbon workingelectrode, a carbon counter electrode, and an Ag/AgCl referenceelectrode purchased from Pine Research Instrumentation (Raleigh,NC, USA). A sensor connector (Alderon Biosciences, Inc.) was used toconnect the SPE to the CHI electrochemical analyzer. Prior to use,the SPE was pretreated electrochemically by cyclic voltammetricscanning 10 times at potential window of 0–1.2 V. The treated SPEwas washed with distilled water and dried in air. SWV measure-ments were performed with a screen-printed electrode. 50 �L ofsample solution was dropped on the sensing area of the three elec-trodes to form an electrochemical cell. The released cadmium ionswere measured with SWV using in situ-plated mercury film on theSPE. The electrode was preconditioned at 0.6 V for 65 s and −1.4 Vfor 120 s (accumulation). Subsequent SWV measurement was per-formed from −0.9 to −0.5 V with a step potential of 4 mV, amplitudeof 25 mV, and a frequency of 15 Hz. A baseline correction of theresulting voltammogram was performed using CHI 660A software.

2.3. Procedures

2.3.1. Preparation of QD–cotinine conjugatesQD–cotinine conjugate was prepared using an EDC-NHS

crosslinker. Briefly, 0.038 g EDC, 0.005 g trans-4-cotinine-carboxylic acid, and 0.0024 g NHS were dissolved in 2 mL of0.1 M MES buffer. Then the solution was shaked for 1 h. Aftershaking, 60 �L of 8 �M QD in dimethyl sulfoxide solution wasadded to the previous solution and the solution was shaked for2 h. Then the resulting mixture was filtered with PD-10 column.PD-10 columns were prepacked, disposable columns containingSephadexTM G-25 Medium for group separation of high (Mr > 5000)from low molecular weight substances (Mr < 1000) by desaltingand buffer exchange. Cotinine was eluted firstly from the column.The first flow-through drops were clear. Then the red flow-through drops including QD–cotinine conjugate were collected.Then the collected solution was added with 0.1058 g glycine. The

30 k centrifugal tube (7000 rpm, 4000 g) for 15 min. Finally, theQD–cotinine was reconstituted with 100 �L of 0.5% BSA buffersolution and stored at 4 ◦C.

52 H. Nian et al. / Analytica Chimica Acta 713 (2012) 50– 55

F hromi e QD–O

2

aw1it0(atAefFecTs

22c6sw

2s

ig. 2. (A) Schematic illustration of disposable QD-based electrochemical immunoctive immunoassay on the strip. With lower cotinine concentration in samples, morn the other hand, the signal will be smaller.

.3.2. The binding affinity test of QD–cotinineThe binding affinity of QD–cotinine conjugate to the cotinine

ntibody was studied by incubating the QD–cotinine conjugatesith the cotinine antibody on a microplate. Briefly, 20 �L of

.0 mg mL−1 cotinine antibody was loaded into a microplate andncubated overnight at 4 ◦C. Then the wells were washed threeimes with washing buffer containing 0.1% (w/v) Tween-20 in.05 M PBS solution. 250 �L of the blocking solution containing 3%w/v) BSA and 1% PEG 6000 dissolved in 0.05 M PBS buffer wasdded into each well in the plate and incubated at room tempera-ure for 2 h. Then the wells were washed with the washing buffer.bout 20 �L of diluted QD–cotinine suspension was added intoach well in the previous plate and incubated at room temperatureor 1 h. The wells were washed three times with the washing buffer.or electrochemical measurement, 20 �L of 1 M HCl was added intoach well and incubated for 3 min. 50 �L of 0.2 M acetate bufferontaining 10 ppm Hg and 500 ppb Bi was added into each well.he resulting solution was transferred to a screen-printed electrodeurface for square ware voltammetric (SWV) measurement.

.3.3. Preparation of immunochromatographic test strip

.3.3.1. Preparation of sample application pad. The sample appli-ation pad was made from glass fiber. The pad was cut out to

× 20 mm and saturated with a buffer (pH 8.0) containing 20 mModium borate, 2% (w/v) sucrose, 2% BSA, and 0.1% NaN3. Then it

as dried and stored in a desiccator at room temperature.

.3.3.2. Preparation of conjugate pad. The stock conjugate suspen-ion was diluted with PBS buffer solution containing 0.5% (w/v)

atographic sensor. (B) Schematic illustration of QD-based electrochemical compet-cotinine will be captured by antibodies, leading to a larger electrochemical signal.

BSA. The conjugate pad was prepared by adding 3 �L of dilutedQD–cotinine conjugate solution onto the glass fiber pad and thendrying it. The pad was stored in a desiccator at 4 ◦C.

2.3.3.3. Preparation of capturing antibody pad in nitrocellulose mem-brane. The capturing antibody pad was located on a completenitrocellulose membrane. About 1 �L of desired amount of coti-nine antibody was loaded onto the capturing antibody zone ofnitrocellulose membrane and dried overnight at 4 ◦C. Then thenitrocellulose membrane containing the capturing antibody padwas treated with blocking regent (a solution of 3% BSA and 0.05%Tween-20 in 0.01 M PBS buffer) for 2 h at 4 ◦C and washed mem-brane with 0.05% Tween-20 in PBS buffer for 5 min and repeatedthree times. The membrane was dried in a nitrogen box for 1 h andstored at 4 ◦C in a dry state.

2.3.3.4. Assembling of the test strip. The sample application pad,conjugate pad, nitrocellulose membrane containing capturing-antibody, and absorption pad were laminated on a plastic backingplate.

2.3.3.5. Sample assay procedure with a test strip. 100 �L of sam-ple solution containing a desired concentration of cotinine wasapplied to the sample application zone. After waiting for a desired

time (for example, 10 min), the capturing antibody zone was cutoff and transferred into a 0.5 mL plastic vial. 20 �L of 1 M HCl wasadded into the vail and waited for 3 min for complete dissolution ofquantum dots. 50 �L of 0.2 M acetate buffer with 10 ppm Hg2+ and

H. Nian et al. / Analytica Chimica Acta 713 (2012) 50– 55 53

e for

5f

3

3i

ptbfQtcitbTwTEc

3a

ccatpriwvawwtmsiQQpfQ

Fig. 3. Schematic illustration of the procedur

00 ppb Bi3+ was added into vial. The resulting solution was trans-erred to a screen-printed electrode surface for SWV measurement.

. Results and discussion

.1. Principle of electrochemical competitivemmunochromatographic strip

In the current study, the principle of the assay is based on com-etitive binding between QD–cotinine and cotinine in a sampleo the limited amount of cotinine antibody on the capturing anti-ody zone (Fig. 2A). Briefly, the cotinine in the sample would flowrom sample application pad to QD–cotinine conjugate pad whereD–cotinine and cotinine would mix together. The mixed solu-

ion would continue to flow to capturing antibody zone throughapillary force and QD–cotinine and cotinine would bind compet-tively to the cotinine antibody on capturing antibody zone. Withhe increase of the amount of cotinine in the sample, there woulde less QD–cotinine attached onto the membrane in the test zone.herefore, when test zone was dissolved in HCl, these cadmium ionsould be generated. The equation was CdS + HCl → CdCl2 + H2S.

hen cadmium ions would be measured at working electrode.lectrochemical signal would be inversely proportional to cotinineoncentration in the samples as shown in Fig. 2B.

.2. Characterization of QD–cotinine conjugate and testingntibody binding affinity to cotinine-QD conjugates

Trans-4-cotinine-carboxylic acid has the similar structure withotinine and is used as a cotinine derivative to prepare QD–cotinineonjugates. The carboxylic group of trans-4-cotinine-carboxyliccid facilitates the conjugation of trans-4-cotinine-carboxylic acido the amine functionalized QD surface. The conjugation wasroceeded by using a conventional procedure with the couplingeagent EDC (detailed in the Section 2). The synthetic procedures illustrated in Fig. 3. The as-prepared QD–cotinine conjugates

ere characterized with the electrochemical technique. A typicaloltammogram of the QD–cotinine conjugate was recorded by SWVfter dissolution of the quantum dots with 1 M HCl (Fig. 4A). Aell-defined peak was observed with peak potential at −0.75 V,hich is consistent with that of the standard cadmium. Therefore,

he stripping peak is ascribed to cadmium. This favorable voltam-etric peak was used to electrochemically quantify cotinine in

amples using competitive cotinine immunoassay in the follow-ng sections. The binding affinity of the cotinine antibody to theD–cotinine conjugate was further studied with QD–cotinine and

D (a control) using an anti-cotinine antibody coated microplatelatform. The experiments with QD–cotinine and QD were per-ormed under the same conditions to confirm that the capturedDs were derived from the immunoreactions between cotinine

the synthesis of the QD–cotinine conjugates.

antibody and QD–cotinine. The captured QDs on the cotinineantibody in the microplate well were detected by SWV after anacid dissolution. Fig. 4B presents the electrochemical responsesobtained by incubating QD and QD–cotinine conjugates on the anti-cotinine antibody coated microplate, respectively. It can be seenthat a large electrochemical response was obtained by incubat-ing QD–cotinine with cotinine antibody. This result demonstratesthat the QD–cotinine conjugate can be specifically recognized bythe cotinine antibody, indicating that QDs do not generate sterichindrance to the binding events. While a very small response wasobserved from the control experiment, indicating there is the non-specific interaction of QD nanoparticle conjugates on the platewalls. More serious blocking and washing would greatly reducesuch non-specific binding in the microplate.

3.3. Optimization of parameters of the electrochemicalimmunochromatographic sensor

Certain parameters of the disposable electrochemical compet-itive immunochromatographic sensor would affect the electro-chemical response. To minimize the non-specific interaction andachieve better performance, optimization of experimental param-eters and blocking conditions were critical. It was found that theaddition of BSA to the quantum-cotinine conjugate suspensionwith final concentration of 0.5% BSA can seriously reduce the non-specific binding of QD–cotinine conjugates on the test strip. Toobtain a maximum response using a minimum concentration ofQD–cotinine conjugate, the dilution of stock QD–cotinine conjugatesuspension for applying to the conjugate pad was optimized by fix-ing desired amount of cotinine antibody on the test zone. Fig. 5Ashows the corresponding electrochemical responses to the differ-ent concentrations of QD–cotinine conjugate suspensions, whichwere prepared by diluting the stock suspension of QD–cotinine.As shown in the figure, the electrochemical response increasedwith increasing the conjugate concentration and then reached aplateau at the 5-fold dilution. Consequently, 3 �L of 5-fold dilutionof QD–cotinine conjugate suspension was used in the conjugatepad.

The amount of cotinine antibody immobilized in the test zonealso influences the performance of the sensor. We optimized theconcentration of cotinine antibody with the optimized amount ofQD–cotinine conjugate. Fig. 5B shows the relationship between theelectrochemical responses and the concentration of cotinine anti-body applied to the test zone. It can be seen that the responsecurrent increases with the increase of the concentration of coti-nine antibody and it starts to level off at the concentration of

1.0 mg mL−1 of cotinine antibody. So 1.0 mg mL−1 of cotinine anti-body was used for the following experiments.

Another parameter affecting the electrochemical response ofthe immunosensor is the immunoreaction time. Fig. 5C presents the

54 H. Nian et al. / Analytica Chimica Acta 713 (2012) 50– 55

Fig. 4. (A) Typical square-wave voltammogram of the dissolved QD–cotinine conjugate. The released cadmium ions were measured with SWV using in situ-plated mercuryfi + QDp ical d

eaastot

Fd5

lm on the SPE. (B) Electrochemical responses of the incubation of cotinine antibodylate were dissolved with 20 �L of 1 M HCl to release cadmium ions for electrochem

lectrochemical responses of the sensor with different immunore-ction time. The immunoreactions proceeded at different time suchs 7 min, 10 min and 12 min with the same conditions. It can be

een from Fig. 5C that the electrochemical response increases withhe increase of the immunoreaction time, and then tends to levelff at 10 min. Therefore, 10 min of immunoreaction time was usedhroughout this study.

ig. 5. Optimization of parameters of immunochromatographic strip. (A) Effect of dilutionilution of stock QD–cotinine suspension. (B) Effect of the concentration of continine an-fold dilution of QD–cotinine conjugates: (C) Effect of immunoreaction time using 3 �L of

and cotinine antibody + QD–cotinine. After washing steps, the captured QDs on theetection.

3.4. Analytical performance of the sensor

Under the optimal conditions, the performance of the electro-

chemical competitive immunochromatographic sensor for cotininewas further evaluated. Fig. 6 shows the typical square-wavevoltammograms of the sensor with increasing cotinine analyte con-centrations (from top to bottom, 0, 1, 5, 10, 50, and 100 ng mL−1).

of QD–cotinine; immunoreactions were performed for 15 min with 3 �L of differenttibody. Immunoreaction time: 15 min; 1 �L of cotinine antibody applied; 3 �L of

5-fold dilution of QD–cotinine conjugate suspension and 1 �L of cotinine antibody.

H. Nian et al. / Analytica Chimic

Fig. 6. Typical SWV responses of the sensor with increasing cotinine concentrations(

WmcIcm1ptrcf9c

3

msIspattnsfotc12s4a1ibps

[

[[[

[

[[

[

[[[[

[[

[[[[

435–446.[28] J. Wang, G.D. Liu, H. Wu, Y.H. Lin, Small 4 (2008) 82–86.[29] Y. Xing, J.H. Rao, Cancer Biomarkers 4 (2008) 307–319.

from top to bottom 0–100 ng mL−1).

ell-defined voltammetric peaks (cadmium) are observed and theagnitude of the peak currents decreases with the increase of

otinine concentrations. The normalized signals are expressed as I0−1(%) (where I and I0 are the peak current obtained with theotinine analyte and the blank sample, respectively) and the nor-alized signals of cotinine concentrations (from 0, 1, 5, 10, 50, and

00 ng mL−1) are 95.2%, 81.8%, 76.0%, 62.5%,and 56.7%. The data waslotted versus the logarithm of cotinine concentration. The equa-ion of the linear calibration curve of cotinine at the concentrationange of 1–100 ng mL−1 was [I I0−1(%)] = 95.2–19.2 × log [cotinineoncentration] (R2 = 0.9990). The detection limit was estimatedrom competitive curves as the normalized signal was assigned at5% referred to Seccareccia’s study [30]. The detection limit underurrent conditions is about 1.0 ng mL−1.

.5. Serum sample analysis

The applicability of the QD-based electrochemical immunochro-atographic sensor for biomonitoring of cotinine in real biological

amples was preliminarily assessed using spiked human serum.t is reported the concentration of cotinine in active smokererum was more than 200 ng mL−1[30]. The human serum sam-les spiked with final concentrations of 100, 200, 300, 400,nd 500 ng mL−1 cotinine were tested. After 10-fold dilution,hese samples were dropped into sample application pad ofhe sensor directly. The serum sample without spiking coti-ine served as a control. Cotinine concentrations of the spikedamples were quantified based on the electrochemical signalrom these samples and the calibration curve. The equationf the linear calibration curve of cotinine at the concentra-ion range of 100–500 ng mL−1 was [(I I0−1(%)] = 2580.1 × [cotinineoncentration]−0.741(R2 = 0.9926). The spiked concentrations of00, 200, 300, 400, and 500 ng mL−1 cotinine measured 95.4, 219.4,93.4, 400.9, and 486.4 ng mL−1. The recoveries of these humanerum samples spiked with final concentrations of 100, 200, 300,00, and 500 ng mL−1 cotinine were 95.4%, 109.7%, 97.8%, 100.2%,nd 97.3%. The average recoveries for these spiked samples were00.1% with a relative standard deviation of 5.65%. The results

ndicate the sensor can be used for biomonitoring of continue iniological samples. Such results demonstrate that this sensor has a

otential for sensitive, rapid, and in field biomonitoring for tobaccomoke exposure.

[

a Acta 713 (2012) 50– 55 55

4. Conclusion

We have successfully developed the disposable competitiveimmunochromatographic sensor for rapid, sensitive, and quanti-tative detection of cotinine based on the immunochromatographicstrip and nanoparticle label. The disposable sensor in this studytakes advantage of the speed and low cost of conventionalimmunochromatographic strip test and high sensitivity of thenanoparticle-based electrochemical immunoassay. Under opti-mal conditions, the sensor is capable of detecting a minimum1 ng mL−1 cotinine. The sensor provides a rapid, accurate, andquantitative tool for cotinine detection and shows great promisefor in-field and POC quantitative screening of tobacco smokeexposure.

Acknowledgments

The work was performed at Pacific Northwest National Lab-oratory (PNNL) and supported by a NIH grant (U54 ES16015)from the National Institute of Environmental Health Sciences(NIEHS). The contents of this publication are solely the responsi-bility of the authors and do not necessarily represent the officialviews of the NIH. The work was performed at the Environmen-tal Molecular Sciences Laboratory, a national scientific user facilitysponsored by the U.S. Department of Energy (DOE) and locatedat PNNL. PNNL is operated by Battelle for DOE under contractDE-AC05-76RL01830.

References

[1] D.W. Sepkovic, N.J. Haley, D. Hoffmann, JAMA-J. Am. Med. Assoc. 256 (1986),pp. 863–863.

[2] CDC, NHANES Analytic Guidelines, 2005.[3] J.T. Bernert, W.E. Turner, J.L. Pirkle, C.S. Sosnoff, J.R. Akins, M.K. Waldrep, Q.

Ann, T.R. Covey, W.E. Whitfield, E.W. Gunter, B.B. Miller, D.G. Patterson, L.L.Needham, W.H. Hannon, E.J. Sampson, Clin. Chem. 43 (1997) 2281–2291.

[4] P. Boffetta, S. Clark, M. Shen, R. Gislefoss, R. Peto, A. Andersen, Cancer Epidemiol.Biomarkers Prev. 15 (2006) 1184–1188.

[5] H.W. Kuo, J.S. Yang, M.C. Chiu, J. Chromatogr. B – Anal. Technol. Biomed. LifeSci. 768 (2002) 297–303.

[6] P. Kowalski, M. Marszall, I. Oledzka, W. Czarnowski, Chromatographia 66 (2007)357–361.

[7] I.B. Karaconji, L. Skender, V. Karacic, Acta Chim. Slov. 54 (2007) 74–78.[8] J.P. Sherry, Criti. Rev. Anal. Chem. 23 (1992) 217–300.[9] S. Benkirane, A. Nicolas, M.M. Galteau, G. Siest, Eur. J. Clin. Chem. Clin. Biochem.

29 (1991) 405–410.10] R.I. Stefan, R.G. Bokretsion, J.F. van Staden, H.Y. Aboul-Enein, Talanta 59 (2003)

883–887.11] C.A. Marquette, L.J. Blum, Biosens. Bioelectron. 21 (2006) 1424–1433.12] J. Kohn, Immunology 15 (1968) 863–870.13] Y.Y. Lin, J. Wang, G.D. Liu, H. Wu, C.M. Wai, Y.H. Lin, Biosens. Bioelectron. 23

(2008) 1659–1665.14] G.D. Liu, Y.Y. Lin, J. Wang, H. Wu, C.M. Wai, Y. Lin, Anal. Chem. 79 (2007)

7644–7653.15] C.M. Plotz, J.M. Singer, Am. J. Med. 21 (1956) 888–892.16] P. Yager, T. Edwards, E. Fu, K. Helton, K. Nelson, M.R. Tam, B.H. Weigl, Nature

442 (2006) 412–418.17] M.A. Cousino, T.B. Jarbawi, H.B. Halsall, W.R. Heineman, Anal. Chem. 69 (1997)

A544–A549.18] L.F. Peruski, A.H. Peruski, Biotechniques 35 (2003) 840–846.19] M.G. Weller, Fresenius J. Anal. Chem. 366 (2000) 635–645.20] M.Z. Zheng, J.L. Richard, J. Binder, Mycopathologia 161 (2006) 261–273.21] K. Kerman, M. Saito, S. Yamamura, Y. Takamura, E. Tamiya, Trac-Trends Anal.

Chem. 27 (7) (2008) 585–592.22] G.D. Liu, J. Wang, H. Wu, Y.H. Lin, Anal. Chem. 78 (2006) 7417–7423.23] (a) Y. Xie, A. Chen, D. Du, Y. Lin, Anal. Chim. Acta 699 (2011) 44–48;

(b) Y. Wang, Z. Li, J. Wang, J. Li, Y. Lin, Trends Biotechnol. 29 (2011) 205–212.24] J. Wang, Analyst 130 (2005) 421–426.25] J. Wang, Y.H. Lin, Trac-Trends Anal. Chem. 27 (2008) 619–626.26] J. Wang, G.D. Liu, M.H. Engelhard, Y.H. Lin, Anal. Chem. 78 (2006) 6974–6979.27] I.L. Medintz, H.T. Uyeda, E.R. Goldman, H. Mattoussi, Nat. Mater. 4 (2005)

30] F. Seccareccia, P. Zuccaro, R. Pacifici, P. Meli, F. Pannozzo, K.M. Freeman, A.Santaquilani, S. Giampaoli, M. Project, Eur. J. Epidemiol. 18 (2003) 487–492.