Ni

FAG

a

ARRAA

KSFGNSA

1

btSsbBSnrrpd

fa

0d

Biosensors and Bioelectronics 34 (2012) 151– 158

Contents lists available at SciVerse ScienceDirect

Biosensors and Bioelectronics

j our na l ho me page: www.elsev ier .com/ locate /b ios

anogold probe enhanced Surface Plasmon Resonance immunosensor formproved detection of antibiotic residues

átima Fernández, Francisco Sánchez-Baeza, M.-Pilar Marco ∗

pplied Molecular Receptors Group (AMRg), CIBER de Bioingeniería, Biomateriales y Nanomedicina, Department of Chemical and Biomolecular Nanotechnology, IQAC-CSIC, Jorgeirona, 18-26, 08034 Barcelona, Spain

r t i c l e i n f o

rticle history:eceived 13 November 2011eceived in revised form 2 January 2012ccepted 27 January 2012vailable online 6 February 2012

eywords:urface Plasmon Resonance immunosensorluoroquinolone antibioticsold nanoparticlesanogold probesignal enhancementntibody

a b s t r a c t

An exhaustive study is reported on the effect that antibody nanogold probes produce on the perfor-mance of a Surface Plasmon Resonance (SPR) immunosensor. The paper studies the improvement thatdifferent nanogold probes prepared at different antibody:gold nanoparticle (IgG:AuNP) ratios and AuNPsizes produce on the maximum signal and detectability of a simple SPR immunosensor developed toanalyze fluoroquinolone (FQ) antibiotic residues (SPReeta system). The investigation compares the fea-tures of sensor enhanced formats using both, secondary and primary nanogold probes (anti-IgG and IgGcoupled to AuNP, on double and single-antibody immunochemical assay steps, respectively), in respectto the unenhanced format. For this purpose, a reproducible bioconjugation procedure for preparing goldbiohybrid nanoparticles has been established, involving the formation of a mixed self-assembled mono-layer (m-SAM) with PEGylated cross-linkers around the AuNP followed by the covalent attachment ofthe antibodies. The procedure allows controlling the IgG:AuNP ratio of the nanogold probes on a repro-

ducible manner and the functionalized NPs have been found to be stable during assay and storage. Bothformats, using secondary and primary nanogold probes, are excellent strategies to improve immunosen-sor detectability. Thus, using anti-IgG-AuNP, the detectability could be improved by a factor of 14 (LOD0.07 ± 0.01 �g L−1 vs. 0.98 ± 0.38 �g L−1) reducing at the same time the amount of primary antibody used(30,000 vs. 1000 dilution factor). Likewise, the format using IgG-AuNP also allows improving detectability(LOD 0.11 ± 0.01 �g L−1), but reducing the number of needed steps.

. Introduction

In last years, SPR (Surface Plasmon Resonance) biosensors haveeen extensively applied in different fields like clinical diagnos-ics, environmental monitoring and food safety (Homola, 2008;itu et al., 2010; Shankaran et al., 2007). Depending on the SPRystem, the detectability reached can be lower than microplate-ased ELISA methods or than high performance SPR devices (ex.iacore). In previous works, we proved that simple and portablePR immunosensors (SPReeta and SPRCD systems) may reach theecessary detectability to determine antibiotic residues (fluo-oquinolones and sulfonamides) in milk samples according UEequirements (Fernández et al., 2010, 2011). But for chloram-henicol, an antibiotic with very restrictive required level, theetectability was slightly limited (Fernández et al., 2010).

Improvement of SPR detectability can be accomplished by dif-erent strategies. LOD has been improved by using secondaryntibodies (Yuan et al., 2007; Mizuta et al., 2008; Gobi, 2007).

∗ Corresponding author. Tel.: +34 934 006 100; fax: +34 932 045 904.E-mail addresses: pilar.marco@cid.csic.es, pilar.marco@cid.csic.es (M.-P. Marco).

956-5663/$ – see front matter © 2012 Elsevier B.V. All rights reserved.oi:10.1016/j.bios.2012.01.036

© 2012 Elsevier B.V. All rights reserved.

Immunoreagents may also be coupled to other elements thatincrease their mass and refractive index, such as liposomes (Winket al., 1998), enzymes (Mitchell and Lowe, 2009; Yang and Kang,2008), polystyrene NPs (He, 2004), magnetic NPs (Teramura et al.,2006), or AuNPs (Lyon et al., 1998; Mitchell et al., 2005; Cao andSim, 2007). The use of AuNPs has motivated great interest dueto the possibility of control their size and to use known chemi-cal functionalization approaches to generate biohybrid NPs, alsoknown as nanogold probes (Boisselier and Astruc, 2009; Daniel andAstruc, 2004; Li-Na et al., 2010). It has been postulated that AuNPsmay also enhance detectability thanks to their own SPR proper-ties (LSPR, localized Surface Plasmon Resonance) reinforcing thesignal by means of electronic coupling of surface and the NP plas-mons (Lyon et al., 1998; Lyon, 1999; Chah et al., 2001). Natan’sgroup, was one of the first research teams demonstrating the poten-tial of nanogold probes for signal enhancement. They employedsecondary nanogold probes (anti-IgG coupled to AuNP) to detecthuman IgG in a sandwich format (Lyon et al., 1998). The same strat-

egy has been used to enhance the signal of SPR sensors to detectdisease biomarkers (Choi et al., 2008; Lee et al., 2009). Regardinglow molecular weight analytes, few papers report signal enhance-ment using 10–40 nm secondary nanogold probes. The enhancement

1 nd Bio

a10e2npmprm0t

peerghhotgiptahiite

nopph

2

2

((p

2

agh(aouAaM(iac

52 F. Fernández et al. / Biosensors a

llowed reducing the concentration of primary antibody (i.e. from0 to 1 �g mL−1) and improved the detectability (i.e. LOD from.1 to 0.007 �g L−1 for benzaldehyde) (Yuan et al., 2007; Mitchellt al., 2005; Mitchell and Lowe, 2009; Gobi et al., 2008; Yuan et al.,008). However, the authors do not discuss the effect related to theumber of IgG molecules attached to the nanoparticles, which wasrobably greater for the bigger AuNPs (Yuan et al., 2008). Enhance-ent using primary nanogold probes is also possible and allows

erforming analysis with less steps, but their use has only beeneported in scarce occasions. Jiang et al. (2009) conjugated the pri-ary antibody to AuNP allowing to improve the LOD of estriol from

.2 to 0.03 �g L−1, but no data is shown comparing performance ofhis strategy in respect to the use of secondary gold probes.

Regarding synthesis, most of IgG-AuNP bioconjugates have beenrepared by passive adsorption of the antibodies on gold (Lyont al., 1998; Mitchell et al., 2005; Pieper-Fürst et al., 2005; Xiulant al., 2005; Jung et al., 2009), but the antibody can be leached oreplaced by other molecules. In last years, a variety of bioconju-ation methods for covalent binding of biomolecules on surfacesave emerged, which allows synthesizing nanogold probes withigher stability and more defined composition. These methods areften based on the use of heterobifunctional alkanes with reactivityowards gold (i.e. thiol groups), and towards proteins (i.e. carboxylroups) (Yu and Irudayaraj, 2007; Eum et al., 2010). Since few yearst is known that PEG (polyethileneglycol) thiol monolayers preventrotein adsorption over gold surfaces (Harder et al., 1998). Fur-her, Sakura et al. reported that PEG thiol reagents can also be selfssembled on the surface of AuNPs (Sakura et al., 2005). Moreover, itas been demonstrated that PEG reagents improve biocompatibil-

ty and stability of nanogold probes in aqueous environment due tots hydrophilic character (Akiyama et al., 2000), which has favoredheir use to prepare biofunctional AuNP (Kumar et al., 2008; Eckt al., 2008; Chen et al., 2005).

In this work, we report the effect of secondary and primaryanogold probes on the detectability of enrofloxacin (ERFX), a flu-roquinolone (FQ) used as model target analyte. For it, differentrobes were synthesized employing PEG-thiol linkers. Influence ofrobe size and IgG:AuNP ratio are also discussed. The evaluationas been made using a simple and portable SPR sensor device.

. Experimental

.1. SPR sensor set-up

The employed system was SPReeta Evaluation Kit-SPR3Nomadics, Inc., Stillwater, OK) using gold chips model TSPR1K23Texas Instruments, Inc., Dallas, TX). More details can be found in arevious work (Fernández et al., 2011).

.2. Chemicals, immunochemicals and bioconjugates

The heterobifunctional reagents used for AuNP function-lization O-(2-carboxyethyl)-O′-(2-mercaptoethyl)-heptaethylenelycol (acid-PEG-thiol) and O-(methyl)-O′-(2-mercaptoethyl)-exaethylene glycol (m-PEG-thiol) were supplied by PolypureOslo, Norway). The commercial secondary nanogold probes, goatnti-rabbit IgG conjugated to 10 nm AuNP (anti-IgG-AuNP-1), usedn certain experiments, the goat anti-rabbit IgG (affinity isolated),sed to prepare in-house made secondary nanogold probes (antiIgG-uNP-2) and, unless otherwise indicated, any other chemicalsnd biochemicals were acquired from Sigma–Aldrich (St. Louis,O). The preparation of the anti-fluoroquinolone antibody Ab171

IgG fraction isolated by (NH4)2SO4 precipitation, ∼10% specificmmunoglobulins), used to prepare the primary nanogold probes,nd the FQ-BSA bioconjugate used to biofunctionalize the sensorhips has been described before (Marco et al., 2010). Unspecific

electronics 34 (2012) 151– 158

IgGs to test nonspecific interactions and to prepare negative con-trol nanogold probes were isolated from rabbit pre-immune sera(IgG fraction obtained in the same manner). Bio-Rad protein assay(Bio-Rad Laboratories GmbH) was employed for the Bradford test.Enrofloxacin (EFRX) used as standard was supplied by Biochemika,Fluka (Buch, Switzerland).

2.3. Buffers and solutions

Borax buffer is 50 mM in boric acid and 5.6 mM in sodium borate(pH 8.7). The solution used to form a self-assembled monolayer(PEG-SAM solution, 1 mL) on the AuNP was a freshly preparedmixture of acid-PEG-thiol and m-PEG-thiol (250 and 750 mM,respectively) in milliQ water. The carboxylic acids were activatedusing a solution of NHS (20 mM) and ECD (40 mM) prepared in PBS(10 mM phosphate buffer with 0.8% saline solution, pH 7.5) justprior use. The antibody solutions employed for AuNP functionaliza-tion were prepared in borax buffer in the 5–40 �g mL−1 range foranti-rabbit IgG (for secondary nanogold probes preparation) and inthe 30–240 �g mL−1 range for Ab171 (for primary nanogold probespreparation). These solutions provide an initial IgG:AuNp ratio of∼3–25 and ∼20–150 range respectively. The nanogold probe stor-age solution contained 20% glycerol, 1% BSA, 0.05% NaN3 and 20 mMTris base in water. For the ERFX used as standard, a 10 mM stock wasprepared in 50 mM NaOH aqueous solution and stored at 4 ◦C. Dif-ferent ERFX concentration solutions were prepared in PBSCa buffer(PBS with 1 mM in CaCl2) for the calibration of the sensor in therange 3500–0.04 �g L−1.

2.4. Synthesis of nanogold probes

A scheme of the process is showed in Fig. 1.

2.4.1. Synthesis of AuNP100 mL of AuNPs (10 mg of HAuCl4 + 95 mL water + 5 mL of 1%

sodium citrate solution) were prepared following the proceduredeveloped by Turkevich et al. (1951) (Enustun and Turkevich,1963). The AuNP were characterized in solution by UV–vis spec-trophotometry and by transmission electron microscopy (TEM,Philips CM30). Their size was determined by measuring the diam-eters of 30 AuNPs on the TEM images.

2.4.2. Preparation of IgG-AuNP bioconjugates(i) Preparation of PEGylated particles: PEG-SAM solution (1 mL)

was added to the AuNP colloid solution (100 mL). The mixture wasemplaced under N2 atmosphere and kept overnight under stirring.The solution was distributed in 10 tubes (10 mL each) and cen-trifuged at 18,000 × g for 30 min at 4 ◦C. The pellets of each tubewere resuspended in water (0.5 mL), combined and splited againin aliquots (1 mL). (ii) Activation: NHS/ECD solution (1 mL) wasadded to each vial containing the PEGylated particles (1 mL) andthe mixture was stirred for 30 min and subsequently centrifugedfor 30 min at 4 ◦C at 20,000 × g in the same tubes. (iii) Bindingto IgG: different antibody solutions (see Section 2.3) were added(2 mL) to the activated AuNP vials containing the pellet. The vialswere kept under magnetic stirring for 4 h at room temperatureand subsequently centrifuged (20,000 × g, 30 min, 4 ◦C). The super-natants were analyzed by Bradford test to know the concentrationof unbound anti-IgG. Subtraction of this value from the initial con-centration allowed to estimate the amount of IgG bound to theAuNP. The pellets of each vial were resuspended in water (2 mL)and centrifuged for a second time to ensure that the probes were

IgG free. (iv) Final reconstitution: the pellets were reconstitutedwith the storage solution (200 �L each). The content of the vialswere collected in a final working aliquot (∼5 × 1013 AuNP mL−1). Allsteps were monitored by recording the UV–vis spectra in order to

F. Fernández et al. / Biosensors and Bioelectronics 34 (2012) 151– 158 153

Fig. 1. Scheme showing the process used to prepare nanogold probes by covalent attachment of the antibodies to the PEGylated nanoparticles via the active ester method.T ed alkU ± 2.4

ctpc

2

d

2

fiAspmnt(iA(woflcrueI7Aw

he AuNP are synthesized by citrate reduction and functionalized with two PEGylatV-vis spectra of the AuNP and the nanogold probes. The average AuNP size is 18.8

ontrol the AuNP concentration and potential losses after each cen-rifugation/reconstitution step. Negative controls were prepared inarallel by conjugating IgGs isolated from preimmune serum. Theseontrols were used to test unspecific signal.

.5. Biofunctionalization of the SPR chips

The biofunctionalization of the gold surface was performed asescribed before (Fernández et al., 2011). See Fig. 2A for detail.

.6. SPR immunosensor protocol

The evaluation of the probes was made using different con-gurations (see Fig. 2): format A, single antibody step (antiserums171), without nanogold probes; format B, two antibody steps, theecond one using anti-IgG coupled to AuNP (secondary nanogoldrobes, antiIgG-AuNP), format C, single antibody step, using pri-ary antibodies (antibody Ab171) coupled to AuNP (primary

anogold probes, Ab171-AuNP). The table below Fig. 2 shows addi-ional details for the different formats. Briefly, a 1:1 pre-incubated10 min) mixture of the ERFX standard solutions (3500–0 �g L−1

n 8–9 calibration points) and the FQ-antibody (As171 for formats and B or Ab171-AuNP for format C) was flowed into the sensor15 min in formats A and C, 10 min in format B) and then, the chipas washed with running buffer (5 min). In format B, a solution

f secondary nanogold probes (anti-IgG-AuNP) was subsequentlyowed through the cell (5 min), and washed again. Finally, thehip was regenerated by passing 0.3 M NaOH (5 min) followed byunning buffer (10 min) to recover the baseline level. The flow ratesed was 33 �L min−1. The concentration of the probes used tovaluate the detectability was ∼4–5 × 1012 AuNP mL−1, and the

gG:AuNP ratio was ∼3 for the antiIgG-AuNP-1 used in format B1,–8 for the antiIgG-AuNP-2 used in format B2 and 10–12 for theb171-AuNP used in format C. The response, change in RI (RIUs),as determined subtracting the signal before the binding of the

anes. The inserted white box at the bottom shows the TEM image of the AuNP andnm (n = 30 particles).

primary antibody to the signal level just before regeneration. Asimilar maximum response was intended to accomplish for allthe formats for comparison purposes. The maximum signal range(600–1000 RIUs) was chosen as a signal that is not too affectedby baseline fluctuations, temperature changes or the drift of thesystem. The comparison between the formats was made as afunction of the parameters of the equation used to fit standardcurves. The equation is Y = [(A − B)/1 − (C/x)D̂] + B, where A is themaximal signal, B is the minimum signal, C is the concentrationproducing 50% of the maximal signal, and D is the slope at theinflection point of the curve. LOD (concentration producing 90% ofthe maximum signal) and coefficient or regression (R2) were calcu-lated. Calibration curves were analyzed using at least n = 2 channelsin n′ = 2 different days. Variability inter-days and inter-chips areincluded in the results. To assess the reproducibility, the averageand standard deviation (X ± S) were calculated for all parameters.

3. Results and discussion

3.1. Synthesis of nanogold probes

To have reproducible and stable nanogold probes, the useof a robust procedure to bind the IgG to AuNP on a covalentmanner and with a controlled IgG:AuNP ratio is crucial. Often,procedures for the preparation of nanogold probes relay on passiveadsorption of the antibody molecules on the AuNP surface (Lyonet al., 1998; Mitchell et al., 2005; Pieper-Fürst et al., 2005; Xiulanet al., 2005; Jung et al., 2009). This kind of procedures can leadto high biomolecule:AuNP ratio forming multiple protein layers(Tkachenko et al., 2005). Moreover, there is a risk of protein leach-ing phenomena during storage or while using them on the assay,

or even replacement of the antibodies by other molecules presentin the samples. Thus, with this objective, monodispersed sphericalAuNP were synthesized by reducing the gold salt (HAu3+Cl4) withcitrate as described (Enustun and Turkevich, 1963; Turkevich

154 F. Fernández et al. / Biosensors and Bioelectronics 34 (2012) 151– 158

Fig. 2. Configurations assayed in the FQ SPR immunosensor: format A, single antibody step, without nanogold probes; format B, two antibody steps, the second, using anti-IgGcoupled to AuNP (secondary nanogold probes); and format C, single antibody step, using primary antibodies coupled to AuNP (primary nanogold probes). Performance of the SPRusing format B was investigated with two different AuNPs: 10 nm AuNP (B1) and 19 nm AuNP (B2). In format C, the size of AuNP was 19 nm. Below each graph, the optimizedc ratiosB e prei

eU1ttaeoolhcS(e(nctcmttFci5w

onditions for FQ analysis with the SPR immunosensor are shown. The IgG:AuNPs

2 and C, respectively. Previously to the binding step (I), the standard solutions wer

t al., 1951). The resulting AuNP were characterized by TEM andV–vis spectrophotometry resulting in AuNP of a medium size8.8 ± 2.4 nm, and �max 522 nm (see Fig. 1). Taking into accounthe amount of gold used (50 �g mL−1 of Au3+), the mean size ofhe AuNP and assuming a quantitative reduction of the gold salt,

concentration of 7.4 × 1011 AuNPs mL−1 (1.2 × 10−6 mM) wasstimated in the colloidal solution (100 mL). For the preparationf the nanogold probes, the AuNPs were treated with a mixturef PEG-thiols to form m-SAM around the nanoparticles. PEGinkers are excellent reagents to stabilize gold colloid due to theirydrophilic character, in addition to providing the necessaryhemical functional groups for bioconjugation. Although oftenAM PEG-layers on the AuNP are prepared with just the reactiveusually COOH ended) thiolated linker (Cao and Sim, 2007; Eckt al., 2008; Kumar et al., 2008), we used a mixture of reactiveacid-PEG-thiol) and unreactive (CH3-PEG-thiol) linkers simulta-eously in order to get a suitable charge distribution for successfulonjugation, since a low density of carboxylic groups has proveno favor functionalization of gold surfaces (Gobi, 2007). Thus, bothross-linkers were used at a 1:3 (acid-PEG-thiol:CH3-PEG-thiol)olar ratio and antibody coupling was accomplished by activating

he carboxylic groups of the m-SAM layer followed by the forma-ion of amide bonds with the lysine groups of the antibody (seeig. 1). At the end of the procedure, the obtained solution had a

oncentration of 4–5 × 1013 AuNP mL−1 (about 70 times highern respect to starting colloidal solution) and showed a �max of24 nm (see Fig. 1), indicating that no significant loses of AuNPere produced during the distinct centrifugation/reconstitution

of the nanogold probes in the final protocol were 3, 7–8 and 10–12 for formats B1,ncubated for 10′ with As171 (formats A and B) or with Ab171-AuNP (format C).

steps. The small shift of the �max (522 → 524 nm) and the slightpeak broadening observed on the spectra has also been reportedby other authors (Cao and Sim, 2007; Kumar et al., 2008).

The established procedure was applied to synthesize primary(Ab171-AuNP) and secondary nanogold probes (anti-IgG-AuNP),by slightly changing the experimental conditions based on thedifferent purity degree of the antibodies used. Thus, for Ab171 (iso-lated by precipitation, ≥10% specific IgGs), higher IgG:AuNP initialratios were employed (20–150 IgG:AuNP) than for antiIgG (3–25IgG:AuNP, affinity purified, 100% specific IgGs according to the sup-plier). The final IgG:AuNP ratio achieved is seldom mentioned inthe papers, and in the few occasions where that datum is reported,the value given is based on the use of fluorescence labeled of IgGsto prepare the probes (Kumar et al., 2008). In our case, the esti-mation of IgG molecules bound to the AuNP has been estimatedby measuring the protein concentration remaining in the super-natant using Bradford test. A clear correlation (R2 = 0.980, 0.941)could be observed between the initial IgG:AuNP ratio employedand the IgG:AuNP ratio achieved on the nanogold probe as it isshown in Fig. 3a and c for the conjugation with anti-IgG and Ab171,respectively. Moreover, the %CV between the different conjugationbatches was lower than 13% (n = 6) indicating the excellent repeata-bility of the procedure established. In some papers (Kumar et al.,2008; Eghtedari et al., 2009) is reported the necessity to block the

remaining surface of the gold particle after IgG coupling, to improvestability of the nanogold probes. However, the PEG m-SAM layerinitially formed around the AuNP has been found to provide to thesenanogold probes with sufficient stability.

F. Fernández et al. / Biosensors and Bioelectronics 34 (2012) 151– 158 155

(c) Conjugation of IgG to AuNPs

0 25 50 75 100 125 150

0

5

10

15

20

25

30

y = 0.20x + 1.50R = 0.941

Ab171 :AuNP initial ratio

Ab

17

1:A

uN

P

ratio

ach

ieve

d

(d) Response of primary nangogold probes

10 15 20 25 300

1

2

3

4

5

0.00

0.55

1.10

1.65

2.20

2.75

y = 0.23x - 1.85R = 0.979

Ab171:AuNP ratio achieved

En

ha

nc

em

en

t fa

cto

r (E

.F.)

RIU

s·1

0-3

(a) Conjugation of anti-IgG to AuNPs

0 5 10 15 20 250

3

6

9

12

15

y = 0.64x - 0.54R = 0.980

anti-IgG:AuNP initial ratio

an

ti-I

gG

:Au

NP

ratio

ach

ieve

d

(b) Response of secondary nanogold probes

2 4 6 8 10 12 14 160

6

12

18

24

y = 1.40x + 2.52R = 0.924

0.0

3.3

6.6

9.9

13.2

anti-IgG:AuNP ratio achieved

En

ha

nc

em

en

t F

ac

tor

(E.F

)

RIU

s·1

0-3

Total Signal (As171 + anti-IgG-AuNP)

Sign al As171E.F =

Signal Ab17 1-AuNP

Signal As171E.F =

Fig. 3. Data related to the bioconjugation yield (a and c) and the effect of the antibody:AuNP ratio achieved on the SPR response (b and d) using either secondary nanogoldprobes (anti-IgG-AuNP) or primary nanogold probes (Ab171-AuNP). The response graphs show the raw signal (RIUs, right axis) and the enhancement factor (left axis) vs.a ilutions m). Thp

3i

dr(wp(

FIc(1

ntibody:AuNP ratio. The experiments of graphic B were developed using a 1/1000 decondary nanogold probes (anti-IgG-AuNP, 4–5 × 1012 AuNPs mL−1, AuNp size 19 nrimary nanogold probes (Ab171-AuNP, 4–5 × 1012 AuNPs mL−1, AuNP size 19 nm).

.2. Evaluation of the nanogold probes on the fq SPRmmunosensor signal

The probes were evaluated on the FQ SPReeta immunosensoreveloped before (Fernández et al., 2011) using different configu-ations as shown in Fig. 2. The procedure using one antibody-step

format A) was compared with two enhanced formats: format B,ith two binding steps (As171 followed by the secondary nanogold

robes, anti-IgG-AuNP), and format C, with a single antibody stepprimary nanogold probes, Ab171-AuNP). As observed in Fig. 4a, the

ig. 4. Sensograms showing the enhancement on the SPR signal achieved with differentgG-AuNP) for format B1 (3 anti-IgGs per AuNP, AuNP size 10 nm) and B2 (7–8 anti-IgGsell followed by the solutions of the nanogold probes (dilution 1/10). On the third channequivalent to concentration of anti-IgG in format B1). (b) Enhancement achieved employ9 nm). The black line shows the achieved signal for free Ab171 (25 �g mL−1). The concen

of As171 (flowed for 15 min) and 1/10 dilution (flowed for 15 min) of the differente experiments of graphic C were developed with a 1/10 dilution of the different

use of secondary probes (formats B1 and B2) produced a significantincrease of the signal, in comparison to the signal achieved withoutany enhancement (format A) or just using anti-IgG as amplifier.As reported by other authors (Lyon, 1999; Yuan et al., 2008)our results also indicate that size of AuNP, have a direct effecton the SPR signal. Thus, format B1 (10 nm AuNP) and format B2

(19 nm AuNP) produced an enhancement of about 7 and 11 times,respectively, in respect to the signal of the primary antibody(As171, 1/1000 dilution). However, on the overall interpretationof the signal enhancement observed, it should also be taken into

strategies. (a) Enhancement achieved employing secondary nanogold probes (anti- per AuNP, AuNP size 19 nm). A 1/1000 dilution of As171 was flowed through theel (black line) the second step was developed with a free anti-IgG at ∼1.3 �g mL−1

ing primary nanogold probes (Ab171-AuNP) in format C (20 IgGs per GNP, AuNP sizetration of bioconjugates was 4–5 × 1012 AuNPs mL−1.

156 F. Fernández et al. / Biosensors and Bioelectronics 34 (2012) 151– 158

Fig. 5. (a) Calibration curves obtained for enrofloxacin (ERFX) using formats A, B1, B2 and C. The calibration points correspond to a mean value of change in RI (RIUs) ofmeasurements made in, at least, n = 2 channels. The features of these assays and curves are summarized in Table 1 and Fig. 2. (b) Sensogram of the ERFX calibration curveg 5 × 10(

cpah(dws1bu

enerated in format B (anti-IgG-AuNP, 7–8 anti-IgGs per AuNP, AuNP size 19 nm, 4–Ab171-AuNP, 11–12 IgGs per AuNP, AuNP size 19 nm, 4–5 × 1012 AuNPs mL−1).

onsideration the availability of antibody recognition sites in bothrobes (see effect of the IgG:AuNP below). On the other hand, someuthors have mentioned that small size nanogold probes couldave better diffusion kinetics, favoring their access to the surfaceYuan et al., 2008). In this respect, we did not observe significantifferences in the analysis time when using nanoprobes preparedith either 10 or 19 nm AuNP. In contrast, the second binding

tep with no labeled anti-IgG produced an enhancement of just.6 times (Fig. 4a), pointing to the great contribution producedy the AuNP. Similarly, as it is shown in Fig. 4b in format C, these of Ab171-AuNP (19 nm AuNP) also produce a significant

12 AuNPs mL−1). (c) Sensogram of the ERFX calibration curve generated in format C

enhancement of the response in comparison with unconjugatedAb171. The achieved SPR signal was 3 times higher. It is worthnoticing that no negative effect was observed in respect to theregeneration of the surface, being possible to recover the samebaseline after each analysis. Furthermore, experiments performedusing unspecific gold nanoprobes, did not produce any significantincrease in the signal (<90 RIUs) pointing to a low unspecific

binding of the AuNP to the biofunctionalized SPR chip.

The effect of different IgG:AuNP ratios on the signal enhance-ment was also investigated. As shown in Fig. 3b, using format B2 theenhancement factor produced by the secondary probes correlated

F. Fernández et al. / Biosensors and Bioelectronics 34 (2012) 151– 158 157

Table 1Features of the ERFX calibration curves in no-enhancement (format A) and enhancement conditions (formats B and C).

Analysis features Format A Format B Format C

ImmRc Dil min B1 B2 ImmR Dil min

ImmR Dil min ImmR Dil min

SPR immunosensor conditionsFirst step As171 1000 15 As171 20,000 10 As171 30,000 10 Ab171-AuNP 10 15Second step – aIgG-AuNP 10 5 aIgG-AuNP 10 5 –AuNP size, nm – 10 19 19Ab:AuNP – 3 7–8 10–12Anal. time, mind 33 40 40 35Assay parametersa

Smax, RIUs 672 ± 78 1069 ± 38 580 ± 41 946 ± 129Smin,RIUs 104 ± 7 207 ± 114 80 ± 20 96 ± 48Slope −1.0 ± 0.2 −0.9 ± 0.1 −0.7 ± 0.1 −0.9 ± 0.1IC50b 10.17 ± 2.69 2.03 ± 0.16 0.85 ± 0.28 1.18 ± 0.52IC90 (LOD) 0.98 ± 0.38 0.23 ± 0.14 0.07 ± 0.01 0.11 ± 0.01R2 0.98 0.98 0.97 0.99

a All the parameters are expressed as the average and standard deviation (at least n = 2 channels and n = 2 days).

qwAtmgctismt

3

ifippvsYtff(dB6eC(o

dwwFlitde

b All the concentrations are expressed in �g L−1.c ImmR is immunoreagent.d Total analysis time including regeneration of the chip surface.

uite well (R2 = 0.924) with their anti-IgG:AuNP ratio. The sameas observed in format C when assaying primary probes at differentb171:AuNP ratios (see Fig. 3d, R2 = 0.979). These results indicate

hat signal enhancement is due to the AuNP but also by the IgGolecules around the NP, since higher IgG:AuNP ratios mean also

reater size. Moreover, IgG:AuNP ratios may also increase thehances of the probes for binding to the surface, contributing tohe signal enhancement observed. From Fig. 3b and d, it can also benferred that primary probes require higher IgG:AuNP ratios thanecondary probes to reach the same enhancement. As it has beenentioned above, this fact can be explained by the lower purity of

he conjugated IgGs (10% vs. 100%).

.3. Effect on the SPR detectability

Due to the great signal enhancement achieved in format B, thencubation times of each step could be reduced to 10 min in therst step (As171) and to just 5 min on the second step (secondaryrobes). Moreover, As171 could be used at higher dilution, in com-arison to format A, which allowed reaching better detectabilityalues. Similar results have been reported by other authors usingecondary nanogold probes (Gobi et al., 2008; Mitchell et al., 2005;uan et al., 2007). Thus, the As171 solution could be diluted 20,000imes in format B1 (10 nm, antiIgG:AuNP ∼ 3) and 30,000 times forormat B2 (19 nm, antiIgG:AuNP ∼ 7, a intermediate ratio), while inormat A it was diluted just 1000 times to reach the same signal600–1000 RIUs). Higher dilutions could be employed in format B2ue to the higher signal amplification achieved in respect to format1 (see Fig. 4a). Other authors have also reported dilution factors–30 times greater by employing secondary nanogold probes (Gobit al., 2008; Mitchell et al., 2005; Yuan et al., 2007, 2008). In format, primary nanogold probes with a IgG:AuNP ratios of about 11–1215′ of binding) provide comparable signal to format A as it can bebserved in Fig. 3d.

Using the above-mentioned conditions (see Fig. 2) theetectability of ERFX, a representative FQ antibiotic congener,as assessed through the preparation of calibration curves thatere run with the different formats. As it can be observed in

ig. 5, the curves obtained in formats B and C are shifted to theeft side in respect that obtained using format A (non enhanced),

ndicating that those formats are able to detect lower ERFX concen-rations. The parameters of the curves are shown in Table 1. A goodetectability (IC50: 10.17 ± 2.69 �g L−1; LOD: 0.98 ± 0.4 �g L−1 fornrofloxacin in buffer) was already achieved by the FQ SPReeta

immunosensor when run in format A (Fernández et al., 2011), butit could be improved about 14 times (IC50: 0.85 ± 0.28 �g L−1;LOD: 0.07 ± 0.01 �g L−1) in format B2 and about 5 times (IC50:2.03 ± 0.16 �g L−1; LOD: 0.23 ± 0.14 �g L−1) in format B1. As dis-cussed above, this difference is related to the different sizeof the AuNP and the distinct IgG:AuNP ratio of the secondarynanogold probes used in these two formats. Using format C, thedetectability was 8 times better (IC50: 1.18 ± 0.52 �g L−1; LOD:0.11 ± 0.01 �g L−1) than format A. An improvement of about 5–6times has been reported in the scarce studies found in the litera-ture in respect to the use of primary nanogold probes (Jiang et al.,2009).

4. Conclusions

The PEGylated thiols are excellent reagents for the synthesis ofnanogold probes (IgG-AuNP), providing high stability in aqueoussolution for long time. The bioconjugation method developed isreproducible and allows a carefull control of the IgG:AuNP ratio.The signal achieved with nanogold probes could be controlled byjust selecting the appropriate IgG:AuNP ratio.

The extraordinary signal enhancement produced by the AuNPshas allowed reducing the concentration of specific immunore-agents used and, consequently, leading to a significant increase inthe detectability. The use of secondary nanogold probes (format B)improved the detectability up to 14 times, reaching LOD values forERFX down to 0.07 �g L−1 (0.98 without enhancement, format A).The use of primary nanogold probes (format C) also lead to a similarimprovement reaching a LOD value of 0.11 �g L−1 in just one stepassay.

Acknowledgments

This work has been supported by the Spanish Ministry of Sci-ence and Innovation in the frame of the Detecta and OligoCODEsprojects (AGL2008-05578-C05-01, MAT2011-29335-C03-01). TheAMR group is a consolidated research group (Grup de Recerca)of the Generalitat de Catalunya and has support from the Depar-

tament d’Universitats, Recerca i Societat de la Informació laGeneralitat de Catalunya (expedient 2009SGR 1343). CIBER-BBN isan initiative funded by the VI National R&D&i Plan 2008–2011, Ini-ciativa Ingenio 2010, Consolider Program, CIBER Actions and financed

1 nd Bio

bR

R

A

BCC

C

C

DE

E

E

E

F

F

GG

H

HHJ

J

58 F. Fernández et al. / Biosensors a

y the Instituto de Salud Carlos III with assistance from the Europeanegional Development Fund.

eferences

kiyama, Y., Otsuka, H., Nagasaki, Y., Kato, M., Kataoka, K., 2000. Bioconjugate Chem-istry 11 (6), 947–950.

oisselier, E., Astruc, D., 2009. Chemical Society Reviews, doi:10.1039/b806051g.ao, C., Sim, S.J., 2007. Biosensors and Bioelectronics 22 (9–10), 1874–1880.hah, S., Hutter, E., Roy, D., Fendler, J.H., Yi, J., 2001. Chemical Physics 272 (1),

127–136.hen, J., Saeki, F., Wiley, B.J., Cang, H., Cobb, M.J., Li, Z., Au, L., Li, B.K.M., Younan, X.X.,

2005. Nano Letters 5 (3), 473–477.hoi, J.W., Kang, D.Y., Jang, Y.H., Kim, H.H., Min, J., Oh, B.K., 2008. Colloids and Surfaces

A: Physicochemical and Engineering Aspects 313, 655–659.aniel, M.-C., Astruc, D., 2004. Chemical Reviews 104 (1), 293–346.ck, W., Craig, G., Sigdel, A., Ritter, G., Old, L.J., Tang, L., Brennan, M.F., Allen, P.J.,

Mason, M.D., 2008. ACS Nano 2 (11), 2263–2272.ghtedari, M., Liopo, A.V., Copland, J.A., Oraevsky, A.A., Motamedi, M., 2009. Nano

Letters 9 (1), 287–291.nustun, B.V., Turkevich, J., 1963. Journal of the American Chemical Society 21 (8),

3317.um, N.-S., Yeom, S.-H., Kwon, D.-H., Kim, H.-R., Kang, S.-W., 2010. Sensors and

Actuators B: Chemical 143 (2), 784–788.ernández, F., Hegnerová, K., Piliarik, M., Sanchez-Baeza, F., Homola, J., Marco, M.P.,

2010. Biosensors and Bioelectronics 26 (4), 1231–1238.ernández, F., Pinacho, D.G., Sánchez-Baeza, F., Marco, M.P., 2011. Journal of Agri-

cultural and Food Chemistry 59 (9), 5036–5043.obi, K.V., 2007. Analytical and Bioanalytical Chemistry 387 (8), 2727–2735.obi, K.V., Matsumoto, K., Toko, K., Miura, N., 2008. Sensing and Instrumentation for

Food Quality and Safety, doi:10.1007/s11694-008-9033-5.arder, P., Grunze, M., Dahint, R., Whitesides, G.M., Laibinis, P.E., 1998. Journal of

Physical Chemistry B 102 (2), 426–436.

e, L., 2004. The Journal of Physical Chemistry B 108 (30), 10973–10980.omola, J., 2008. Chemical Reviews 108 (2), 462–493.

iang, X., Waterland, M., Blackwell, L., Wu, Y., Jayasundera, K.P., Partridge, A., 2009.Steroids 74 (10–11), 819–824.

ung, J., Na, K., Lee, J., Kim, K.W., Hyun, J., 2009. Analytica Chimica Acta 651 (1), 91–97.

electronics 34 (2012) 151– 158

Kumar, S., Aaron, J., Sokolov, K., 2008. Nature Protocols 3 (2), 314–320.Lee, J.H., Kang, D.Y., Lee, T., Kim, S.U., Oh, B.K., Choi, J.W., 2009. Journal of Nanoscience

and Nanotechnology 9 (12), 7155–7160.Li-Na, M., Dian-Jun, L., Zhen-Xin, W., 2010. Chinese Journal of Analytical Chemistry

38 (1), 1–7.Lyon, L.A., 1999. The Journal of Physical Chemistry B 103 (28), 5826–5831.Lyon, L.A., Natan, M.D., Musick, M.J., 1998. Analytical Chemistry 70 (4),

5177–5183.Marco, M.-P., Sánchez-Baeza, F., Pinacho, D. G., 2010. P201031721, OEPM, Madrid.Mitchell, J.S., Lowe, T.E., 2009. Biosensors and Bioelectronics 24 (7), 2177–2183.Mitchell, J.S., Wu, Y., Cook, C.J., Main, L., 2005. Analytical Biochemistry 343 (1),

125–135.Mizuta, Y., Onodera, T., Singh, P., Matsumoto, K., Miura, N., Toko, K., 2008. Biosensors

and Bioelectronics 24 (2), 191–197.Pieper-Fürst, U., Stöcklein, W.F.M., Warsinke, A., 2005. Analytica Chimica Acta 550

(1–2), 69–76.Sakura, T., Takahashi, T., Kataoka, K., Nagasaki, Y., 2005. Colloid & Polymer Science

284 (1), 97–101.Shankaran, D.R., Gobi, K.V., Miura, N., 2007. Sensors and Actuators B: Chemical 121

(1), 158–177.Situ, C., Mooney, M.H., Elliott, C.T., Buijs, J., 2010. TrAC Trends in Analytical Chemistry

29 (11), 1305–1315.Teramura, Y., Arima, Y., Iwata, H., 2006. Analytical Biochemistry 357 (2),

208–215.Tkachenko, A., Xie, H., Franzen, S., Feldheim, D.L., 2005. In: Rosenthal, S.J., Wright,

D.W. (Eds.), Nanobiotechnology Protocols. Humana Press, Totowa, New Jersey,pp. 85–99.

Turkevich, J., Stevenson, P.C., Hillier, J., 1951. Discussions of the Faraday Society 11,55–75.

Wink, T., van Zuilen, S.J., Bult, A., van Bennekom, W.P., 1998. Analytical Chemistry70 (5), 827–832.

Xiulan, S., Xiaolian, Z., Jian, T., Zhou, J., Chu, F.S., 2005. International Journal of FoodMicrobiology 99 (2), 185–194.

Yang, G., Kang, S.W., 2008. Food Science and Biotechnology 17 (3), 547–552.

Yu, C.X., Irudayaraj, J., 2007. Analytical Chemistry 79 (2), 572–579.Yuan, J., Oliver, R., Aguilar, M.I., Wu, Y.Q., 2008. Analytical Chemistry 80 (21),

8329–8333.Yuan, J., Oliver, R., Li, J., Lee, J., Aguilar, M., Wu, Y., 2007. Biosensors and Bioelectronics

23 (1), 144–148.