Immobilization of rhodopsin on a self-assembled multilayer and its specific detection by electrochemical impedance spectroscopy

Biosensors and Bioelectronics 21 (2006) 1393–1402

Immobilization of rhodopsin on a self-assembled multilayer and itsspecific detection by electrochemical impedance spectroscopy

Yanxia Houa,b, Salwa Helalia,c, Aidong Zhangb, Nicole Jaffrezic-Renaulta,∗, Claude Marteleta,Jasmina Minicd, Tatiana Gorojankinad, Marie-Annick Persuyd, Edith Pajot-Augyd,

Roland Salessed, Francois Bessueillee, Josep Samitiere, Abdelhamid Errachide,Vladimir Akimov f, Lino Reggianif, Cecilia Pennettaf, Eleonora Alfinitof

a Centre de Genie Electrique de Lyon (CEGELY), Ecole Centrale de Lyon, B.P.163, 69134 Ecully Cedex, Franceb College of Chemistry, Central China Normal University, Wuhan 430079, PR China

c Unite de Recherches de Physiques des Semiconducteurs et Capteurs, IPEST, La Marsa, Tunisiad Institut National de la Recherche Agronomique (INRA), NOPA-RCC,

Bat. Biotechnologies, Domaine de Vilvert, 78352 Jouy-en-Josas Cedex, Francee Laboratory of NanoBioEngineering, Barcelona Science Park, Barcelona 08028, Spain

f Dipartimento di Ingegneria dell’Innovazione, Universita’ di Lecce and INFM,National Nanotechnology Laboratory, Via Arnesano, 73100 Lecce, Italy

A

as receptore (LB) anda hus, in thiss posed of am inylateda y on self-a vity of thiss contact witho mmobilizede©

K

1

alie

theycep-

aste,ewasPCRsally,d to

0d

Received 16 March 2005; received in revised form 16 May 2005; accepted 13 June 2005Available online 25 July 2005

bstract

Rhodopsin, the G protein-coupled receptor (GPCR) which mediates the sense of vision, was prepared from calf eyes and usednriched membrane fraction. In this study it was immobilized onto gold electrode by two different techniques: Langmuir–Blodgettstrategy based on a self-assembled multilayer. We demonstrated that Langmuir and LB films of rhodopsin are not stable. T

tudy a new protein multilayer was prepared on gold electrode by building up layer-by-layer a self-assembled multilayer. It is comixed self-assembled monolayer formed by MHDA and biotinyl-PE, followed by a biotin–avidin system which allows binding of biotntibody specific to rhodopsin. The immobilization of rhodopsin in membrane fraction, by the specific antibody bound previouslssembled multilayer, was monitored with electrochemical impedance spectroscopy (EIS). In addition, the specificity and sensitielf-assembled multilayer system to the presence of rhodopsin were investigated. No effect was observed when the system was inlfactory receptor I7 in membrane fraction used for control measurements. All these results demonstrate that rhodopsin can be ifficiently, specifically, quantitatively and stably on gold electrode through the self-assembled multilayer.2005 Elsevier B.V. All rights reserved.

eywords: Rhodopsin; Self-assembled multilayer; Langmuir–Blodgett; Electrochemical impedance spectroscopy; Cyclic voltammetry; Biosensor

. Introduction

Membrane proteins, encoded by about 20% of genes inlmost all organisms, including humans, are critical for cel-

ular communication, electrical and ion balances, structuralntegrity of the cells adhesions and other functions (Tellert al., 2001). Among membrane proteins, G protein-coupled

∗ Corresponding author. Tel.: +33 472186243; fax: +33 478331140.E-mail address: [email protected] (N. Jaffrezic-Renault).

receptors (GPCRs) are of special importance becauseform one of the largest and the most diverse groups of retor proteins. They mediate the sense of vision, smell, tand pain (Vaidehi et al., 2002). In our study, rhodopsin, thdim light receptor of rod outer segments of the retina,chosen as representative GPCRs, since it is the only Gwhose crystal structure is known and which is naturhighly available. It contains seven transmembrane�-helicesand its chromophore, 11-cis retinal, is covalently bounlysine 296 through a Schiff base linkage (Lavoie et al., 2002).

956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved.oi:10.1016/j.bios.2005.06.002

1394 Y. Hou et al. / Biosensors and Bioelectronics 21 (2006) 1393–1402

For immobilization of this bioreceptor, two techniques,Langmuir–Blodgett and self-assembled multilayer, wereemployed in this study, since they can provide the requiredcontrol at a molecular level and they have potential appli-cation for the construction of well-ordered and ultra-thinorganic films.

Langmuir–Blodgett technique has already shown its effi-ciency for depositing well-defined films of enzymes (Okahataet al., 1988; Fiol et al., 1992; Chovelon et al., 2000; Dubreuilet al., 1995; Zhang et al., 2002) and antibodies (Tronin etal., 1996; Barraud et al., 1993; Hou et al., 2004) to elaboratebiosensors.

Properties of Langmuir and Langmuir–Blodgett films ofrhodopsin have already been investigated byLavoie et al.(2002)and group of Nicolini (Maxia et al., 1995; Pepe et al.,1996, 1998; Pepe and Nicolini, 1996). Lavoie et al. studiedstructure of rhodopsin in monolayers at the air/water inter-face, and they reported that experimental conditions can befound where the secondary structure of rhodopsin can beretained when spread in monolayers. Nicolini et al. investi-gated the properties of LB films of purified bovine rhodopsinand they found that bleached rhodopsin in LB films has highthermal stability. In this study, we investigated the propertyand stability of monolayer of rhodopsin membrane fractionat the air/water interface. We used the rhodopsin in its mem-brane fraction because the presence of lipids surroundingr ty.

y ofs anes aces( o-l1 ledm ations tiono icu-l ;F

ul-t atedi ors( 03;L anw theb icala eena t of1 pHo tiplew

bledm tohe-x teda asisf teda ch to

anchor membrane protein rhodopsin, using neutravidin anda biotin-labelled antibody specific to rhodopsin.

Among the various transduction techniques, such as quartzcrystal microbalance (QCM), surface plasmon resonance(SPR), ellipsometry, etc., electrochemical impedance spec-troscopy (EIS) is a rapidly developing and effective electro-chemical technique for the characterization of biomaterial-functionalized electrodes and biocatalytic transformations atelectrode surfaces, and specifically for the transduction ofbiosensing events at electrodes. Compared to other electro-chemical techniques, one of the advantages of EIS is thesmall amplitude perturbation from steady state, which makesit possible to treat the response theoretically by linearizedor otherwise simplified current–potential characteristics. Upto now EIS has been extensively performed to characterizethe fabrication of biosensors and to monitor the biomolecu-lar recognition (Katz and Willner, 2003; Guan et al., 2004;Alfonta et al., 2001; Rickert et al., 1996; Pei et al., 2001; Wanget al., 2004). In our study, it was used to monitor the formationof the multilayer structure and the recognition of rhodopsinmembrane fraction by self-assembled multilayer. AFM is apowerful tool in structural biology (Fotiadis et al., 2002) sinceit could gives access to the molecular architecture, and it wasused to characterize the multilayer film in this study.

2

2

g tot tore cteda tateb entm t waso n, itwK -c

unc-t lfac-t s asr eastS ac-t

tionw eres,F

ainstt fromN heaL

l-sn-g t

hodopsin prevents its unfolding and maintains its activiSince Nuzzo and Allara, pioneers in the assembl

ulfur-containing molecules in 1983, noticed that dialkulfides form highly ordered monolayers on metal surfNuzzo and Allara, 1983), the field of self-assembled monayers (SAMs) has witnessed tremendous growth (Ulman,996). The simplicity and adaptability of self-assembonolayers and control over biomolecule surface orient

uggest that SAMs play an important role in the construcf artificial biomolecular recognition surfaces and part

arly in the development of biosensors (Wink et al., 1997erretti et al., 2000; Gooding and Hibbert, 1999).

More recently the construction of self-assembled milayer from biological components has been investigntensively owing to its potential application for biosensSpinke et al., 1993; Cui et al., 2003; Boozer et al., 20add et al., 2004). For this purpose avidin–biotin system cork perfectly as a bridge to anchor bioreceptor, sinceiotinylation of a biomolecule does not affect its biologctivity. And still more, the non-covalent complex betwvidin and biotin is characterized by an affinity constan015 mol−1 l. Once formed, the bond is stable even if thef the solution is changed and it can easily survive mulashing (Storri et al., 1998).In the present study, we used mixed self-assem

onolayers on gold electrodes, formed by 16-mercapadecanoic acid (MHDA) and biotinyl-PE which is insernd bound to MHDA by electrostatic interaction, as the b

or the formation of a multilayer structure. We investigamultilayer system of relevance for biosensor resear

. Materials and methods

.1. Biomaterials and chemicals

Rhodopsin was prepared from calf eyes accordinhe procedure ofPapermaster (1982), and used as recepnriched membrane fraction. In brief, retinae were dissend homogenized in 20% sucrose in 10 mM Tris-aceuffer, pH 7.2. A fraction enriched in rod outer segmembranes containing rhodopsin as a major componenbtained by sucrose gradient centrifugation. After isolatioas resuspended in 137 mM NaCl, 8 mM Na2HPO4, 2.7 mMCl, 1.5 mM KH2PO4, buffer (pH 7.2), with 1.5% octylgluoside, aliquoted and frozen at−80◦C.

For the control of the non-specific response of the fionalized gold electrode, tests were performed using I7 oory receptor, which belongs to the same family of GPCRhodopsin. The I7 olfactory receptor was expressed in yaccharomyces cerevisiae and prepared as membrane frion as previously reported (Minic et al., 2005).

The total protein concentration in membrane fracas determined using the BCA reagent (Pierce, Brebirance) with bovine serum albumin as a standard.

The specific monoclonal Rho-1D4 antibody raised aghe rhodopsin C-terminal nanopeptide was purchasedational Cell Culture Center (Minneapolis, MN, USA). Tntibody was biotinylated using DSB-XTM Biotin Proteinabeling Kit (Molecular Probes, Leiden, Netherlands).

16-Mercaptohexadecanoic acid and 1,2-dioleoylycero-3-phosphoethanolamine-N-(biotinyl) sodium sal

Y. Hou et al. / Biosensors and Bioelectronics 21 (2006) 1393–1402 1395

(biotinyl-PE) were purchased, respectively, from AldrichChemical Company and Avanti. IgG from goat, which wasused as blocking reagent, was bought from Sigma. Neu-travidin was obtained commercially from Pierce. And sol-vents ethanol and acetone were both bought from Fluka(purity > 99.8%).

Potassium dihydrogen phosphate, di-sodium hydrogenphosphate and sodium chloride all from PROLABO, potas-sium chloride from Fluka and sodium hydroxide fromAldrich Chemical Company were used to prepare phos-phate buffer solution (PBS), consisting of 1.8 mM KH2PO4,0.1 mM Na2HPO4, 140 mM NaCl and 2.7 mM KCl, pH 7.0.All of them are analytical grade reagents (>99%). Solutionswere prepared with water that has been purified by an Ultra-pure water system Elga. This buffer solution was used for allexperiments.

2.2. Fabrication and pre-treatment of gold substrate

Gold substrates were provided by Laboratoire d’Analyseet d’Architecture des Systemes (LAAS), Toulouse. Theywere fabricated using standard silicon technologies.〈1 0 0〉-oriented, P-type (3–5� cm) silicon wafers were thermallyoxidized to grow an 800 nm-thick field oxide. Then, a 30 nm-thick titanium layer and a 300 nm-thick gold top layer weredeposited by evaporation under vacuum.

tonei genfl /v,H ica Theyw l andfi

2

uirt d2 er ino tem-p m( red ba haseo

n( thea yer.T ied toc mea-s layerc bilityo ateds me-t erer s att

transferred onto bare gold electrode, which has a hydrophilicsurface after cleaning, by a vertical dipping procedure at tar-get pressure 30 mN m−1.

2.4. Preparation of self-assembled multilayer

After cleaning, gold substrates were immersed immedi-ately into a mixed solution of MHDA and biotinyl-PE inabsolute ethanol at the concentration of 1 and 0.1 mM, respec-tively, for 21 h (Step I). The ratio between components deter-mines the density of specific anchoring sites. Then gold sub-strates were removed from the mixed solution and thoroughlyrinsed with ethanol and dried under N2 flow. Subsequently,gold substrates functionalized with mixed SAMs were putinto 1× 10−7 M goat IgG solution in PBS (pH 7.0) for 2 h forblocking free space inside of the mixed SAMs (Step II). Next,they were placed in a 1× 10−7 M neutravidin solution for90 min (Step III), followed by deposition of biotinylated spe-cific monoclonal anti-rhodopsin antibody (Biot-Rho-1D4) at1× 10−7 M for 1 h (Step IV). Finally, rhodopsin membranefraction was anchored on the self-assembled multilayer sys-tems (Step V).

In this study, neutravidin is used as an alternative to strep-tavidin which is often used to provide avidin–biotin sys-tem, under consideration of several key features, such ascarbohydrate free and neutral isoelectric point that providee ion,n din,t per-m anda tep.

S tor rbedo

2

lec-t non-f werep ninga anal-y wereu Func-t h ane

thep((

ce ofa fre-q tionp V.A ara-d

Before use, gold substrates were cleaned with acen an ultrasonic bath for 10 min and dried under nitroow, followed by immersion in a piranha solution (7:3 v2SO4:H2O2) for 1 min in order to get rid of inorgannd organic contaminants on the substrate surface.ere subsequently rinsed thoroughly in absolute ethanonally dried under nitrogen flow.

.3. Formation and deposition of LB films of rhodopsin

Langmuir experiments were performed with a Langmrough from NIMA (model 611) with 30 cm× 10 cm size an00 ml volume. The trough was equipped with a chambrder to prevent external contamination. The subphaseerature was controlled at 5± 1◦C by a thermostatic systeJulabo-F25, France). The surface pressure was measuWilhelmy balance and the trough was filled with a subpf PBS.

A small amount (40�l) of rhodopsin membrane fractio1.5 mg/ml total protein concentration) was spread atir/water interface for formation of rhodopsin monolahe surface pressure–molecular area isotherm was studharacterize the monolayer at the air/water interface byuring the changes in surface pressure upon the monoompression at a given temperature. In addition, the staf the monolayer at the air/water interface was investigince the stability of the monolayer is an important paraer for the stability of the corresponding LB films. They wecorded as molecular area ratio-time evolution isothermhe target pressure (referred to asA/A0 − t). The film was

y

xceptionally low nonspecific binding properties. In additeutravidin can bind at four binding sites like streptavi

wo binding sites on each of two opposite faces, whichits its binding onto biotin-containing SAMs on one sidenchor Biot-Rho-1D4 on the other side in the following s

After each step, substrates were rinsed well with PBemove the excess of material which was physically adsonto the surface.

.5. Electrochemical measurements

The electrochemical characteristics of the modified erode were measured using cyclic voltammetry (CV) andaradaic impedance. Electrochemical measurementserformed in a conventional electrochemical cell contaithree-electrode system with a Voltalab 80 impedance

ser. A Pt plate and a saturated calomel electrode (SCE)sed as counter and reference electrode, respectively.

ionalized gold electrodes act as working electrodes witffective surface of 0.2122 cm2.

Cyclic voltammetry measurements were performed inresence of 2 mM K4[Fe(CN)6] + 2 mM K3[Fe(CN)6] in PBSpH 7.0) and the potential was swept between−0.3 and 0.6 Vversus SCE) at a rate of 50 mV s−1.

Impedance measurements were carried out in absenny redox probe in the PBS at ambient temperature in auency range from 500 mHz to 100 kHz, at a polarizaotential of 0 V/SCE with a frequency modulation of 10 mll electrochemical measurements were performed in a Fay cage.


2.6. Atomic force microscopy

Atomic force microscopy measurements were performedwith a commercial microscopy instrument (PicoSPM Molec-ular Imaging). A pyramidal silicon tip with spring constant ofapproximately 3 N m−1 was used. The cantilever was oscil-lated at its free resonance frequency (75 kHz) with a freeamplitude,A0 = 1.6 V. All the experiments were performedunder ambient conditions with relative humidity of about60%. The image presented in this paper was taken in tappingmode. Since it minimizes potentially destructive shear andadhesive forces on the sample compared to contact mode,which is a particularly valuable property when imaginingcells or biological molecules. The image was first-order flat-tened.

3. Results and discussion

3.1. Study of Langmuir films of rhodopsin

3.1.1. Characteristics of rhodopsin monolayer at theair/water interface

The characteristics of rhodopsin membrane fractionmonolayer at the air/water interface were studied by mea-suring the changes in surface pressure upon compressingt ept sure0 m-pf ana rt dingb .(

F at thea

a difference in size is probably induced by the procedureof preparation of rhodopsin samples, i.e. different lipidicmolecules or presence of detergent to solubilize rhodopsinmolecules.

3.1.2. Stability of rhodopsin monolayer at the air/waterinterface

Since the stability of the LB films is dependent, to a greatextent, upon the stability of the monolayer at the air/waterinterface, we have investigated the stability of monolayer ofrhodopsin at different surface pressures (30 and 35 mN m−1).They are recorded as molecular area ratio-time evolutionisotherms at constant surface pressure, as shown inFig. 2.It can be seen that the monolayers of rhodopsin are not sta-ble. In particular, at the surface pressure 35 mN m−1, after70 min, molecular area ratioA/A0 reaches nearly 0.5, whichmeans that about 50% rhodopsin molecules were expelledfrom interface and turned into the subphase at such a surfacepressure.

In the previous studies, we showed that an enzyme (Zhanget al., 2002), or an antibody (Hou et al., 2004) can form sta-ble Langmuir and Langmuir–Blodgett films in presence ofamphiphile. They can form protein/surfactant complex withamphiphile after optimizing experimental conditions, espe-cially pH of subphase. Such a complex can strengthen thestability of at the air/water interface. Rhodopsin is a trans-m ndert ateri

sino foreu -c theg ared uousv inedL ility

F twod

he monolayer at a given temperature (5◦C), namely surfacressure–molecular area isotherm (shown inFig. 1). In

his study, the film was spread at an initial surface presmN m−1. It was incubated for 20 min, and finally coressed at the rate of 3 cm2 min−1. Extrapolated to 0 mN m−1

rom the linear portion of the compression curve givespparent surface area of 4500A2/molecule. It is greate

han the molecular area occupied by rhodopsin surrouny lipids in the disc membrane reported byLavoie et al2002) (3400A2) and Pepe et al. (1996)(3710A2). Such

ig. 1. Surface pressure–area isotherm of rhodopsin monolayerir/water interface on a subphase of PBS.

embrane protein, and thus extremely hydrophobic. Uhe conditions used the Langmuir film formed at the air/wnterface is not stable.

In addition, we transferred the monolayer of rhodopnto bare gold electrode, which was freshly cleaned bese, at a target pressure of 30 mN m−1. Then by electrohemical impedance measurements we investigatedold electrode, on which the LB films of rhodopsineposited. It was observed that there was a continariation of impedance with time. It means that the obtaB films are quite unstable, which confirms that the stab

ig. 2. Stability of rhodopsin monolayers at the air/water interface forifferent surface pressures (30 and 35 mN m−1).


Fig. 3. Schematic diagram for self-assembled multilayer formation via biotin/avidin pairs on gold electrode.

of the LB films is dependent upon the stability of themonolayer at the air/water interface. Therefore, it is quitedifficult to elaborate a biosensor with such unstable system.So in this study we investigated another immobilizationtechnique based on a self-assembled multilayer, in whichrhodopsin is anchored and bound on gold electrode byaffinity interaction with its specific antibody which is biotin-labelled.

3.2. Study of self-assembled multilayer formation

Fig. 3 shows the schematic diagram for self-assembledmultilayer formation on gold electrode. Electrochemicalcharacteristics of the stepwise formation of self-assembledmultilayer were studied by CV and EIS.

3.2.1. Electrochemical characteristics of self-assembledmultilayer formation by CV

In the presence of a fairly reversible redox couple[Fe(CN)6]4−/3−, cyclic voltammetry was used to monitorthe formation of self-assembled multilayer. As it is shown inFig. 4A, the pretreated bare gold electrode gives a reversiblecyclic voltammogram (curve a). The formation of mixedSAMs on gold electrode is accompanied by a considerabledecrease in the amperometric response of the electrode and ani thodia firmst tingsa Msw e iso se isp ixedS les.A thet lyc on to

negative redox ions [Fe(CN)6]4−/3−. The increase of currentresponse confirms that goat IgG molecules are indeedinserted inside of the mixed SAMs so that they can blockthe free space between biotinyl-PE molecules in the mixedSAMs. The binding of neutravidin on the electrode reducesslightly the penetration of the redox pair and decreases thecurrent response. In the subsequent step, the immobilizationof Biot-Rho-1D4 decreases the current response slightlyfurther.

3.2.2. Electrochemical characteristics of self-assembledmultilayer formation by EIS

The EIS measurements were performed in PBS eitherin the presence or absence of redox probe [Fe(CN)6]4−/3−.However, we found that the presence of such redox probehas effects on immobilization of rhodopsin, the bindingefficiency of biotinylated specific antibody to rhodopsinseemed to decrease. Such a phenomenon was also observedby Rickert et al. (1996). Therefore, the redox couple[Fe(CN)6]4−/3− is omitted in all electrochemical impedancemeasurements.

Non-faradaic impedance measurements were carried outin the frequency range from 500 mHz to 100 kHz. Nyquistplots of impedance spectra of layer-by-layer self-assembledmultilayer are shown inFig. 4B. A significant difference in

e for-tion.

self-

vail-he-ircuiting

ncrease in the peak to peak separation between the cand anodic waves of the redox probe (curve b). This con

hat the formation of mixed SAMs resulted in an insulaurface and the electron transfer kinetics of [Fe(CN)6]4−/3−re perturbed. After blocking the free space inside SAith goat IgG, a slight increase of current responsbserved (curve c). The increase of current responrobably due to structural rearrangement of the mAMs with insertion and adsorption of goat IgG molecunother possibility is that adsorption of goat IgG on

erminal carboxylic group of MHDA which is negativeharged in the pH 7.0, reduces the electrostatic repulsi

cthe impedance spectra is observed upon the stepwismation of the multilayers and upon rhodopsin deposiMoreover, by EIS measurements we observed that suchassembled multilayers are very stable.

The impedance data were fitted with a commercially aable softwareZplot/Zview (Scibner Associates Inc.). Tequivalent circuit, shown inFig. 4C, was found to fit adequately the data over the entire frequency range. The cfollows a standard Randles cell and includes the followfour elements:

(i) The ohmic resistance of the electrolyte solution,Rs.


Fig. 4. Electrochemical characteristics of self-assembled multilayer for-mation. (A) CVs recorded in PBS solution in the presence of 4 mM[Fe(CN)6]4−/3−. (a) Bare gold electrode; (b) Step I: mixed SAMs modi-fied gold electrode; (c) Step II: blockage with goat IgG; (d) Step III: bindingof neutravidin; (e) Step IV: immobilization of biotinylated antibody Biot-Rho-1D4. The scan rate is 50 mV/s. (B) Nyquist plots of impedance spectrataken in PBS without redox couple in the frequency range from 500 mHzto 100 kHz. (a) Step I: mixed SAMs modified gold electrode; (b) Step II:blockage with goat IgG; (c) Step III: binding of neutravidin; (d) Step IV:immobilization of biotinylated antibody Biot-Rho-1D4; (e) after injectionof 80 ng/ml rhodopsin membrane fraction. Inset: Nyquist plots for bare goldelectrode (1) and gold electrode modified with the mixed SAMs (2). The sym-bols are the experimental data and the line represents the simulated spectrawith the parameters calculated by Zplot/Zview software using the equivalentcircuit shown in (C). The calculated parameters are given inTable 1.

(ii) The constant phase element impedance,ZCPEI, whoseexpression is given by

ZCPEI = 1/[T (i × ω)P ] (1)

wherei is the imaginary unit,ω the circular frequency,and CPEI-T (F(rad/s)1−CPEI-P) and CPEI-P are the twofitting parameters which model the double layer capac-itance and its phase angle, respectively.

(iii) The generalized finite Warburg impedance in the opencircuit model,ZW1o, whose expression is given by

ZW1o = R × ctnh([i × T × ω]P )/(i × T × ω)P (2)

where W1o-R (�), W1o-T, (s), and W1o-P are the asso-ciated three fitting parameters.

(iv) The polarization resistance,Rp.

The fitting values for the stepwise formation of the multi-layers are reported inTable 1. Starting from the bare gold, thedeposition of the mixed SAM monolayer is found to producea significant increase of the polarization resistance togetherwith a significant decrease of the double layer capacitancewhich confirms the good insulating properties of the mixedmonolayer already found in the CV characterization. In allthe successive steps we found a systematic decrease of thep hichi phe-n tivep newl ainpt asureo f thee sfer tof

3c

yerfi ea-s psinm bledmw angei eses stain-i singet telye tingt ther,d frac-t rfacei nt asm

olarization resistance and of the Warburg resistance wndicate a recovery in the efficiency of the mass transferomenon and/or difference in the dielectric or conducroperties of the electrode surface with formation of

ayer. The value of the solution resistance is found to remractically constant. The deviation of CPEI-P and Wo-P from

he ideal values 1 and 0.5, respectively, is taken as a mef the presence of spurious effects like the roughness olectrode surface and some anomalies of the mass tran

ollow the standard diffusion equation.

.2.3. Self-assembled multilayer formationharacterized by AFM

In order to characterize the formation of the multilalm and to obtain information on its architecture, AFM murements were taken in tapping mode. After the rhodoembrane fraction was immobilized on the self-assemultilayers, AFM image was taken, as presented inFig. 5,hich reveals randomly located grainy features that r

n size from 30 to100 nm and are about 5 nm high. Thizes are consistent with those observed by negativeng of the rhodopsin membrane fraction preparation ulectron microscopy (Minic et al., in press). In addition, the

hickness of the rhodopsin-containing layer is approximaquivalent to the thickness of the cell membrane, indica

hat membrane vesicles open upon immobilization. Fururing the scanning, we found that on some areas of

ion membrane, interaction between the tip and the suncreased, which confirms that the rhodopsin is prese

embrane fraction.


Table 1The fitting values of the equivalent circuit elements for stepwise formation of the multilayer films byZplot/Zview software

Rs (�) CPEI-T (�F(rad/s)1−n) CPEI-P (n) Rp (�) Wo-R (�) Wo-T (s) Wo-P χ2 (10−4)

Bare gold 238.8 3.63 0.94 774.5 2433 0.236 0.33 10Step I 230.4 0.744 0.91 3150 4079 0.521 0.40 9.8Step II 233.9 0.655 0.92 2623 2875 0.345 0.40 9.9Step III 219.7 0.681 0.92 2376 2487 0.304 0.40 9.8Step IV 231.5 0.698 0.92 2011 1607 0.172 0.41 8.1After 80 ng/ml rhodopsin 165.8 0.661 0.93 1550 1257 0.149 0.40 9.9

3.2.4. Recognition properties of the self-assembledmultiplayer

In order to evaluate the recognition properties of the sys-tem, in terms of sensitivity and selectivity, we exposed thegold electrode with self-assembled multilayers to variousconcentrations of rhodopsin after immobilization of Biot-Rho-1D4. The corresponding Nyquist plots of impedancespectra are shown inFig. 6A. The fitting values are presentedin Table 2. We found that both the polarization resistanceand the Warburg resistance decrease with the addition ofrhodopsin. At low concentration, the decrease is very sen-sitive in magnitude, and for concentration above 80 ng/ml,the system is prone to be saturated. The solution resistance isalso found to decrease systematically in an analogous but less

Fam

significant way, probably due to the addition of rhodopsinmembrane fraction in the cell. All other parameters deter-mined by the fitting procedure do not exhibit significantchange.

ig. 5. Topographic AFM image (1�m× 1�m) obtained in tapping mode:fter immobilization of rhodopsin membrane fraction on the self-assembledultilayers.

Fc0a0flu

ig. 6. Nyquist plots of impedance spectra taken in PBS without redoxouple (A) with various concentrations of rhodopsin membrane fraction: (a)ng/ml; (b) 10 ng/ml; (c) 20 ng/ml; (d) 30 ng/ml; (e) 50 ng/ml; (f) 80 ng/mlnd (g) 120 ng/ml. (B) with various concentrations of unrelated analyte: (a)ng/ml; (b) 20 ng/ml; (c) 50 ng/ml; (d) 100 ng/ml. Applied frequency was

rom 500 mHz to 100 kHz. The symbols are the experimental data and theine represents the simulated spectra calculated byZplot/Zview softwaresing the same equivalent circuit as shown inFig. 4C.


Table 2The fitting values of the equivalent circuit elements for specific binding of rhodopsin at various concentrations byZplot/Zview software

Concentration (ng/ml) Rp (�) CPEI-T (�F(rad/s)1−n) CPEI-P (n) Rp (�) Wo-R (�) Wo-T (s) Wo-P

0 231.5 0.698 0.92 2011 1607 0.172 0.4110 195.1 0.662 0.92 1825 1515 0.172 0.4020 185.3 0.690 0.92 1736 1351 0.152 0.4030 179.5 0.671 0.92 1669 1315 0.153 0.4050 169.1 0.700 0.92 1615 1231 0.144 0.4080 165.8 0.661 0.93 1550 1257 0.149 0.40

120 158.1 0.693 0.92 1533 1161 0.137 0.40

To confirm that the above-observed impedance changesarise from the specific interaction between Biot-Rho-1D4and rhodopsin, and to reveal the selectivity of the binding,control experiments were performed. After immobilizationof Biot-Rho-1D4, the gold electrode was exposed to var-ious concentrations of an unrelated membrane protein (I7olfactory receptor in its membrane fraction), which is also amember of the GPCR family. It is important note that we usedthe crude membrane fraction containing all membrane pro-teins expressed by the cell, where the I7 receptor is far frombeing the most abundant protein. For instance, this membranefraction contains the Golf proteins as we recently demon-strated (Minic et al., 2005) and many other proteins. Amongthem, the I7 receptor has a structure predicted to be highlyrelated to that of the rhodopsin, but not sequence homol-ogy in the antigen region of rhodopsin. The correspondingNyquist plots of impedance spectra obtained are shown inFig. 6B. It can be seen that there is only a slight variation onthe impedance with the increase of I7 concentration.

Fig. 7shows calibration plots that correspond to the resis-tance change (�R) with different concentrations of the spe-cific analyte (rhodopsin membrane fraction) and non-specific

F e as af s(

analyte (I7 membrane fraction). The changes of resistance arecalculated following the equations:

�R = RAb − RAb-Rho (3)

and

�R = RAb − RAb-17 (4)

whereRAb is the value of the resistance when biotinylatedBiot-Rho-1D4 antibody is immobilized on the electrode,RAb-Rho the value of resistance after injection of specificrhodopsin analyte andRAb-I7 is the value of resistance afterinjection of the unrelated analyte I7.

FromFig. 7 it can be observed that there is only a slightresponse to unrelated analyte addition, which is due to non-specific adsorption. However, compared with the specificresponse, we can see that the response for unrelated analyteI7 at the concentration of 100 ng/ml is even lower than thatfor specific rhodopsin at the concentration of 10 ng/ml. Wethus confirm that impedance change inFig. 6A is not due tonon-specific adsorption but it arises from specific interaction.So we can conclude that the self-assembled multilayer sys-tem that we developed allows distinguishing between specificimmobilization of a given receptor and parasitic adsorptionof the other proteins present, and thus represents an efficientmeans to constitute a biosensor for detection of rhodopsin orw ountt limitf ayeri

4

mo-b elf-a thatL aren stem,b vidinp cienti ea-s tilayers g ofr

ig. 7. Calibration plots showing the change of polarization resistancunction of the different concentrations of specific (�) or unrelated analyte�).

ith rhodopsin as the sensing element. Taking into acche blank and the signal fluctuation (noise), the detectionor the binding of rhodopsin on the self-assembled multils 10 ng/ml.

. Conclusion

In this study, rhodopsin membrane fraction was imilized on gold electrode by Langmuir–Blodgett or a sssembled multilayer method. In a first step, we showedB films of rhodopsin deposited on bare gold electrodeot stable. In a second step, we realized a multilayer syased on mixed self-assembled monolayers and biotin/aairs acting as binding agents, and it proved to be an effi

mmobilization method. Electrochemical impedance murements demonstrated that such self-assembled mulystem is sensitive and selective for the specific graftinhodopsin membrane fraction.


The objective of our project is to explore the possibilityto develop a nanobiosensor array based on the electricalproperties of a single GPCR protein, such as olfactoryreceptors. In the future work we will immobilize olfactoryreceptors in their membrane fraction by this self-assembledmultilayer method on micro- and nano-electrodes to inves-tigate electrical properties of the system in the presence ofodorants.

Acknowledgements

We thank the National Cell Culture Center (Minneapolis,MN, USA) for providing the monoclonal Rho-1D4 antibody.This work was financially supported by the SPOT-NOSEDProject (IST-2001-38739).

References

Alfonta, L., Bardea, A., Khersonsky, O., Katz, E., Willner, I., 2001.Chronopotentiometry and faradaic impedance spectroscopy as sig-nal transduction methods for the biocatalytic precipitation of aninsoluble product on electrode supports: routes for enzyme sensors,immunosensors and DNA sensors. Biosens. Bioelectron. 16, 675–687.

Barraud, A., Perrot, H., Billard, V., Martelet, C., Therasse, J., 1993. Studygett. 8,

B g, S.,monActu-

C 00.ixed

rase.28–

C 03.and. 18,

D 995.gett

98–

F Self-bio-

F del-on

491.F .

M.

G self-m. 18,

G iew).

H on-ixed

Langmuir–Blodgett films of immunoglobulin G/amphiphile and theirapplication for immunosensor engineering. Biosens. Bioelectron. 20,1126–1133.

Katz, E., Willner, I., 2003. Probing biomolecular interaction at conductiveand semiconductive surfaces by impedance spectroscopy: routes toimpedimetric immunosensors, DNA-sensors, and enzyme biosensors.Electroanalysis 15, 913–947.

Ladd, J., Boozer, C., Yu, Q., Chen, S., Homola, J., Jiang, S., 2004. DNA-directed protein immobilization on mixed self-assembled monolayersvia a streptavidin bridge. Langmuir 20, 8090–8095.

Lavoie, H., Desbat, B., Vaknin, D., Salesse, C., 2002. Structureof rhodopsin in monolayers at the air-water interface: a PM-IRRAS and X-ray reflectivity study. Biochemistry 41, 13424–13434.

Maxia, L., Radicchi, G., Pepe, I.M., Nicolini, C., 1995. Characterizationof Langmuir–Blodgett films of rhodopsin: thermal stability studies.Biophys. J. 69, 1440–1446.

Minic, J., Persuy, M.A., Godel, E., Aioun, J., Connerton, I., Salesse, R.,Pajot-Augy, E., 2005. Functional expression of olfactory receptors inyeast and development of a bioassay for odorant screening. FEBS J.272 (2), 524–537.

Minic, J., Grosclaude, J., Aioun, J., Persuy, M.A., Gorojankina, T.,Salesse, R., Pajot-Augy, E., Hou, Y., Helali, S., Jaffrezic, N.,Bessueille, F., Errachid, A., Gomila, G., Ruiz, O., Samitier, J., inpress. Immobilization of native membrane-bound rhodopsin on biosen-sor surfaces. BBA.

Nuzzo, R.G., Allara, D.L., 1983. Adsorption of bifunctional organicdisulfides on gold surfaces. J. Am. Chem. Soc. 105, 4481–4483.

Okahata, Y., Tsuruta, T., Ijiro, K., Ariga, K., 1988. Langmuir–Blodgettfilms of an enzyme–lipid complex for sensor membranes. Langmuir

P Meth.

P en-net-

ctron.

P of85,

P itive

P 98.opic

R . Asen-

S .,the

21–

S odi-Bio-

T mp,ree-pled

T ga-lid

U yers.

V P.,f

of immunoglobulin G thin layers obtained by the Langmuir–Blodmethod: application to immunosensors. Biosens. Bioelectron39–48.

oozer, C., Yu, Q., Chen, S., Lee, C., Homola, J., Yee, S.S., Jian2003. Surface functionalization for self-referencing surface plasresonance (SPR) biosensors by multi-step self-assembly. Sens.ators B 90, 22–30.

hovelon, J.M., Gaillard, F., Wan, K., Jaffrezic-Renault, N., 20Influence of the surface pressure on the organization of mLangmuir–Blodgett films of octadecylamine and butyrylcholineste2. Film transferred onto silica support. Langmuir 16, 626232.

ui, X., Pei, R., Wang, Z., Yang, F., Ma, Y., Dong, S., Yang, X., 20Layer-by layer assembly of multilayer films composed of avidinbiotin-labeled antibody for immunosensing. Biosens. Bioelectron59–67.

ubreuil, N., Alexandre, S., Fiol, C., Sommer, F., Valleton, J.M., 1Characterization of enzymatic structures in mixed Langmuir–Blodfilms by scanning force microscopy. Langmuir 11, 202102.

erretti, S., Paynter, S., Russell, D.A., Sapsford, K.E., 2000.assembled monolayers: a versatile tool for the formulation ofsurfaces. Trends Anal. Chem. 19, 530–540.

iol, C., Valleton, J.M., Delpire, N., Barbey, G., Barraud, A., RuauTeixier, A., 1992. Elaboration of a glucose biosensor basedLangmuir–Blodgett technology. Thin Solid Films 210–211, 489–

otiadis, D., Scheuring, S., Muller, S.A., Engel, A., Muller, D.J., 2002Imaging and manipulation of biological structures with the AFMicron 33, 385–397.

ooding, J.J., Hibbert, D.B., 1999. The application of alkanethiolassembled monolayers to enzyme electrodes. Trends Anal. Che525–533.

uan, J., Miao, Y., Zhang, Q., 2004. Impedimetric biosensors (revJ. Biosci. Bioeng. 97, 219–226.

ou, Y., Tlili, C., Jaffrezic-Renault, N., Zhang, A., Martelet, C., Ponsnet, L., Errachid, A., Samitier, J., Bausells, J., 2004. Study of m

4, 1373–1375.apermaster, D.S., 1982. Preparation of retinal rod outer segments.

Enzymol. 81, 48–52.ei, R., Cheng, Z., Wang, E., Yang, X., 2001. Amplification of antig

antibody interactions based on biotin labeled protein-streptavidinwork complex using impedance spectroscopy. Biosens. Bioele16, 355–361.

epe, I.M., Maxia, L., Nicolini, C., 1996. PropertiesLangmuir–Blodgett films of rhodopsin. Thin Solid Films 284–2739–742.

epe, I.M., Nicolini, C., 1996. Langmuir–Blodgett films of photosensproteins. J. Photochem. Photobiol. B: Biol. 33, 191–200.

epe, I.M., Ram, M.K., Paddeu, S., Nicolini, C., 19Langmuir–Blodgett films of rhodopsin: an infrared spectroscstudy. Thin Solid Films 327–329, 118–122.

ickert, J., Gopel, W., Beck, W., Jung, G., Heiduschka, P., 1996“mixed” self-assembled monolayer for an impedimetric immunosor. Biosens. Bioelectron. 11, 757–768.

pinke, J., Liley, M., Guder, H.J., Angermaier, L., Knoll, W1993. Molecular recognition at self-assembled monolayers:construction of multicomponent multilayers. Langmuir 9, 181825.

torri, S., Santoni, T., Minunni, M., Mascini, M., 1998. Surface mfications for the development of piezoimmunosensors. Biosens.electron. 13, 347–357.

eller, D.C., Okada, T., Behnke, C.A., Palczewski, K., StenkaR.E., 2001. Advances in determination of a high-resolution thdimensional structure of rhodopsin, a model of G-protein-coureceptors (GPCRs). Biochemistry 40, 7761–7772.

ronin, A., Dubrovsky, T., Nicolini, C., 1996. Deposition molecular ornization and functional activity of IgG Langmuir films. Thin SoFilms 284–285, 894–897.

lman, A., 1996. Formation and structure of self-assembled monolaChem. Rev. 96, 1533–1554.

aidehi, N., Floriano, W.B., Trabanino, R., Hall, S.E., Freddolino,Choi, E.J., Zamanakos, G., GoddardШ, W.A., 2002. Prediction o


structure and function of G protein-coupled receptors. Proc. Natl.Acad. Sci. 99, 12622–12627.

Wang, M., Wang, L., Wang, G., Ji, X., Bai, Y., Li, T., Gong, S., Li, J.,2004. Application of impedance spectroscopy for monitoring colloidAu-enhanced antibody immobilization and antibody–antigen reactions.Biosens. Bioelectron. 19, 575–582.

Wink, T., Zuilen, S.J.V., Bult, A., Bennekom, W.P.V., 1997. Self-assembled monolayers for biosensors. Analyst 122, 43R–50R.

Zhang, A., Jaffrezic, N., Wan, J., Hou, Y., Chovelon, J.M., 2002. Opti-mization of the mixed urease/amphiphile Langmuir–Blodgett film andits application for biosensor development. Mater. Sci. Eng. C 21,91–96.

Documents

Immobilization of rhodopsin on a self-assembled multilayer and its specific detection by electrochemical impedance spectroscopy