- Email: [email protected]
Conducting elastomer surface texturing: a path to electrode spotting
Biosensors and Bioelectronics 20 (2004) 197–203
Conducting elastomer surface texturing: a path to electrode spotting Application to the biochip production Christophe A. Marquette∗ , Lo¨ıc J. Blum Laboratoire de Génie Enzymatique et Biomoléculaire, Université Claude Bernard Lyon 1, UMR 5013 EMB2-CNRS, Bˆat. CPE, Building 43, 11 November 1918, Villeurbanne 69622, Cedex, France Received 17 September 2003; received in revised form 20 January 2004; accepted 20 January 2004 Available online 10 May 2004
Abstract A new active support for electro-chemiluminescent biochip preparation has been developed. This material was based on an original material composed of graphite modified polydimethyl siloxane (PDMS). The addressed inclusion of Sepharose beads at the surface of this elastomeric electrode generated interesting local high specific surface. The electrode was characterised by electrochemical (cyclic voltametry, chronoamperomatry) and imaging (scanning electron microscopy (SEM)) methods, and a surface area increase factor of 50 was found, linked to the texturing of the surface generated by the presence of the Sepharose beads. The consequence of this increase was shown to be a jump of the local electrochemical activity which induced a well defined and localised electro-chemiluminescent signal. The new material was used to design biochips based on the electro-chemiluminescent reaction of luminol with enzymatically produced hydrogen peroxide. Thus, when using beads bearing bio-molecules such nucleic acid or human IgG, in conjunction with glucose oxidas-labelled DNA or antibody, sensitive biochips could be obtained with detection limits of 1011 and 1010 molecules, respectively. Multi-parameter enzyme-based biochips could also be achieved by locally adsorbing, at the PDMS-graphite surface, either glucose oxidase, lactate oxidase or choline oxidase. Detection limits of 10 M for lactate and choline and 20 M for glucose were found, with detection ranging over two decades at least. © 2004 Elsevier B.V. All rights reserved. Keywords: Biochip; Conducting elastomer; DNA; Electro-chemiluminescence; Enzyme; Immuno; Microarray; PDMS
1. Introduction Biochip technology has become in the last 5 years a large field of investigation, by both the academic and the industrial laboratories. As a consequence, a large number of arraying systems as well as transduction and detection technologies have been developed. Nevertheless, those biochips could be separated in two main categories: (i) the passive arrays (Fig. 1A), composed of a solid support solely dedicated to the immobilisation of bio-molecules (Charles et al., 2003; Salin et al., 2002), and (ii) the active arrays (Fig. 1B and C) in which the support acts in the transduction and/or in the immobilisation steps (Epstein et al., 2002; Patolsky et al., 2002; Marquette et al., 2002). The latter represent some interesting systems since they could enable the ad∗ Corresponding author. Tel.: +33-472-44-8214; fax: +33-472-44-7970. E-mail address: [email protected] (C.A. Marquette).
0956-5663/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2004.01.033
dressing of the analytical signal, but also the use of transduction systems incompatible with the passive arrays. Indeed, a number of detection methods, based on electrochemical processes (amperometry, potentiometry and electro-generated luminescence), known to be sensitive systems, required the use of an active conducting support. However, amperometry and potentiometry are based on electrochemical measurements, and designing such active array involves the production of a matrix of individually wired electrodes (Sullivan et al., 1999; Yamaguchi et al., 2002), with number of technical difficulties. On the contrary, electro-generated luminescent biochips could be prepared at the surface of a unique conducting support such a glassy carbon square (Marquette et al., 2003). Indeed, the electrochemical process generating the light emission is identical for each spot present at the surface of the support. The electro-luminescent system involved, could then be a ruthenium complex (Szunerits et al., 2003), a luminol-based system (Marquette and Blum, 1998; Yang et al., 2002) or an
198
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
signal (A)
e-
(B)
signal signal
(C)
1.1.3.17, from Alcanigenes species), glucose oxidase (grade I, EC 1.1.3.4, from Aspergillus niger), human IgG, lactate oxidase (EC number not available, from Pediococcus), luminol (3-aminophtalhydrazide), Sepharose 4B beads (mean diameter: 100 m), Sepharose 4B beads bearing human IgG, Sepharose 4B beads bearing d(A)22 were purchased from Sigma (France). Glucose oxidase-labelled avidin was obtained from Jackson Immuno Research. PDMS precursor and curing agent (Sylgard 184) were supplied by Dow Corning. All buffers and aqueous solutions were made with distilled de-mineralised water.
e-
2.2. Biochip preparation Fig. 1. Schematic representation of the different biochip categories. (A) Passive support biochip, (B) active and individually wired electrode array, (C) active support biochip.
hydrogen peroxide producing system (oxidase type enzymes) coupled to the luminol reaction (Marquette et al., 2000; Marquette et al., 2003). This latter is of particular interest since it represents an enhancement of the luminol-based system. Moreover, the use of oxidase-tagged DNA, could enable the achievement of nucleic acid electro-chemiluminescent biochip, the use of oxidase-labelled antibodies, the production of immuno electro-chemiluminescent biochip (Marquette and Blum, 1998) and the immobilisation of oxidases lead to enzyme-based electro-chemiluminescent biochip (Marquette et al., 2003). Unfortunately, geographical limitation problems of the signal are classically pointed out (Suzuki and Akaguma, 2000) for enzyme-based electro-chemiluminescent biochip, caused by the lateral diffusion of the hydrogen peroxide produced. The present study proposes to solve such cross-talk problems. Indeed, since the generated luminescent signal is triggered by an electrochemical process, we investigate the possibility of locally increasing this process by generating addressed high specific surface, at an active support interface. The active support is composed of an original elastomer/graphite organised solid, at the surface of which, micro-beads could be physically entrapped (Marquette and Blum, 2004). Group of beads could then be spotted and are used to locally increase the specific surface of the active support. When beads bearing bio-molecules are spotted, they could be used not only to immobilise the bio-molecules, but also to increase the active specific surface. Such approach results in the achievement of electro-chemiluminescent biochips.
The biochips were prepared by spotting 0.3 l drops (Fig. 2) of a solution prepared by mixing 45 l of an aqueous solution composed of glycerol 2.5% and NaCl 0.05 M, with 5 l of wet beads (360 mg/ml of wet beads). The 0.3 l drops were deposited on a flat PVC substrate every 3 mm and dried under a tungsten lamp during 30 min. The presence of glycerol in the spotting solution enabled the achievement of homogeneous and smooth dry bead spots. The dry bead spots were then covered with graphite by spreading over the beads a 0.1 g/ml solution of graphite (TIMREX T15 from TIMCAL, particle mean diameter: 15 m) in chloroform. After a few seconds drying under a tungsten lamp a thin graphite layer (50–100 m thickness) is obtained. The
2. Experimental 2.1. Reagents Biotin labelled anti-human IgG (␥-chain specific), biotin labelled oligonucleotides d(T)22 , choline oxidase (EC
Fig. 2. Steps for the preparation of the addressed texturing of the PDMS-graphite conducting elastomer.
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
2.3. Electrochemical and physical characterisation The electrochemical measurements were performed with a PGZ-100 VoltaLab potentiostat (Radiometer, France). The experiments were carried out in a 2 l volume drop, localised on a bead spot, in which the reference electrode was dipped. The cyclic voltametry and chronoamperometry experiments were performed in an electrolyte solution composed of LiClO4 0.1 M, containing in addition potassium ferricyanide at a concentration of 10 mM. The voltamograms were recorded at 100 mV s−1 between −600 and +600 mV versus a silver pseudo-reference electrode. The chronoamperometry measurements were performed at a +500 mV applied potential during 10 s. The SEM images were obtained with a S800 FEG (Hitachi) electron microscope at 15 kV. 2.4. Nucleic acid assay As a model of oligonucleotide hybridisation process, a d(T)22 –d(A)22 couple was used. Briefly, the biochip was saturated during 30 min with a solution of buffer A (30 mM Veronal (diethylbarbiturate), pH 8.5 containing in addition 30 mM KCl, 0.2 M NaCl, 0.1% Tween and 1% BSA), then incubated for 1 h with biotinylated d(T)22 at a particular concentration in buffer A. The biochip was then washed 10 min with buffer B (30 mM Veronal, pH 8.5 containing 30 mM KCl, 0.2 M NaCl), afterwards saturated with buffer A during 20 min and then incubated with a glucose oxidase-labelled streptavidin solution at a 1 g/ml concentration during 30 min. The array was finally washed 20 min with buffer B and was ready to be measured.
a CCD light measurement system (Intelligent Dark Box II, Fuji Film) and the emitted signal integrated during 3 min. The electro-chemiluminescent reaction was triggered by applying a +850 mV potential between the beads modified conducting elastomer and a bare conducting elastomer, both integrated in the biochip. The measurement solution was composed of glucose 20 mM and luminol 50 M in Veronal buffer 30 mM, pH 8.5, added of KCl 30 mM. In the particular case of the enzyme-based biochip, the oxidase substrate was either glucose, lactate or choline, depending of the experiment carried out.
3. Results and discussion 3.1. Electrochemical and physical properties of the PDMS/graphite elastomer The electrochemical properties of the PDMS/graphite elastomer were first studied by cyclic voltametry. The Fig. 3A presents the voltamograms obtained by cycling the bare and the Sepharose bead-modified conducting elastomer in the presence of 10 mM potassium ferricyanide. As it can be seen, the bare elastomer presents an acceptable conductivity but no electrocatalytic properties (black line) against the Fe(CN)6 3− . On the contrary, the bead-modified 8e-5 4e-5 Current (A/cm²)
dried layers (spots and graphite) were then transferred to the PDMS interface by pouring a mixture of precursor and curing agent onto the PVC substrate and heating for 90 min. The biochip preparation was then terminated by peeling off the PDMS-graphite polymer. At that time, the beads are definitely entrapped at the interface, and could not be detached easily.
199
0e+0 -4e-5 -8e-5 -1
-0,5
(A)
0
0,5
1
Potential (V) 1e-4
40
2.5. Immuno biochip assay In a way similar to that used for the nucleic acid biochip, the immuno biochip was saturated during 30 min in buffer A, then incubated for 1 h with biotinylated anti-IgG antibody at a particular concentration in buffer A. The biochip was then washed 10 min with buffer B, saturated with buffer A during 20 min and then incubated with glucose oxidase-labelled streptavidin at a 1 g/ml concentration during 30 min. The array was then washed 20 min with buffer B and was ready to be measured. 2.6. Electro-chemiluminescent biochip measurement The different electro-chemiluminescent biochips were measured in an identical way. The array was disposed under
Current (A/cm²)
8e-5
10 4 0
0e+0 -4e-5 -8e-5 -1e-4
(B)
20
4e-5
-1
-0,5
0
0,5
1
Potential (V)
Fig. 3. (A) Cyclic voltamograms obtained at a bare PDMS-graphite electrode (black line) and at a textured PDMS-graphite electrode (grey line). (B) Cyclic voltamograms obtained at a bare PDMS-graphite electrode (grey line) and at a textured PDMS-graphite electrode (black line) prepared with different amounts of beads per spot. Voltamograms recorded at 100 mV s−1 between −600 and +600 mV in Fe(CN)6 3− 10 mM.
200
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
elastomer exhibits a good electrocatalytic activity (grey line) shown by the oxidation and reduction peaks of the redox couple. These distinct electrocatalytic activities can be attributed to a difference of specific surface between the bare and the modified elastomer. Indeed, the electrode roughness is well known to influence the efficiency of the electron transfer, keeping the geometrical surface constant (Elliott et al., 1999; Pekala et al., 1998; Evans et al., 2002). Chronoamperometry was used to estimate the active surface of the bead-modified elastomer. The Cottrell equation (Bard, 1996) was used to calculate the ferricyanide diffusion coefficient at the bare elastomer electrode. An estimated active surface was then subsequently calculated from the chronoamperometry curve obtained at the bead-modified polymer. For an identical geometric area (0.07 cm2 ), a fifty-fold increase of the specific surface was obtained with the modified elastomer. The SEM image (Fig. 4A) of one island of beads at the interface of the elastomer demonstrates this local surface roughness. Sepharose beads introduce heterogeneity in the conducting surface, increasing the surface area. The 3D representation of one Sepharose bead (Fig. 4B) shows how it is inserted in the polymer and how it emerges at the surface. A schematic diagram of the layer organisation is presented
in the Fig. 4C. The presence of the porous Sepharose beads during the graphite deposition creates deformations of the conducting surface, generating then a higher exposed surface area; the area under the beads remaining electroactive, without detectable diffusion problem. As an evidence of the effect of the presence of Sepharose beads on the electrochemical activity, the Fig. 3B presents the cyclic voltamograms obtained with spots containing different amounts of beads, from 40 to 0. As it can be seen, the electrode activity decreases with the number of beads in the spots. The higher the number of beads, the higher the specific surface at the spot location and consequently the higher the electrochemical activity. 3.2. Biochip applications The Fig. 5 summarises the different analytical systems used: (i) an enzyme-based biochip (Fig. 5A) composed of oxidase type enzymes physically adsorbed, co-localised with the spot of beads, (ii) a nucleic acid biochip (Fig. 5B) in which the spotted beads are bearing single strand d(A)22 oligonucleotide, (iii) an immuno biochip (Fig. 5C) prepared by spotting human IgG-immobilised beads. The last two systems are reacting with biotinylated d(T)22 and anti-human
Fig. 4. (A) Electron microscopy image (15 kV) of Sepharose beads spot at the surface of the PDMS-graphite elastomer, (B) 3D representation of the SEM image of a particular bead, (C) schematic representation of the layer organisation.
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
201
Fig. 5. The different electro-chemiluminescent biochip formats. (A) Oxidase type enzyme-based biochip, (B) nucleic acid-based biochip, (C) immuno biochip.
IgG, respectively, and affinity interactions with targets were detected with an avidin-glucose oxidase conjugate. 3.2.1. Enzymes-based electro-chemiluminescent biochip The biochip is composed of 18 spots of beads organised as 3 columns of 6 spots. Each row is composed of one spot of adsorbed lactate oxidase, one spot of adsorbed choline oxidase and one spot of adsorbed glucose oxidase. By covering the biochip with a six channels PDMS block polymer, six samples could be injected and analysed in a single measurement.
The Fig. 6 presents the performances of the enzyme-based biochip, each row of the micrographs corresponds to a particular channel in which a particular concentration of analyte was injected. The micrographs show how a single biochip could be used to detect three different analytes, without cross-reactivity problems. Moreover, the spatial resolution of the signal appears to be satisfying, whereas, the different reacting species, that is, the hydrogen peroxide produced and the luminol, are free to diffuse at the surface of the conducting elastomer without generating contamination problem. The local high surface area permits then to avoid the
202
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
Fig. 6. Electro-chemiluminescent enzyme-based biochip. (A) Micrograph and calibration curve of the lactate detection, (B) micrograph and calibration curve of the choline detection, (C) micrograph and calibration curve of the glucose detection.
glowing of the analytical signal, leading to more accurate quantitative results. The graphs corresponding to the calibration curves obtained after treatment of the numerised pictures illustrate the performances of the biochip. Detection limits of 10 M for lactate and choline and 20 M for glucose were found, with detection ranging over two decades at least. Regarding sensitivity, these results compared well those with previously obtained with electro-chemiluminescent biosensors (Marquette et al., 2003).
Fig. 7. Electro-chemiluminescent affinity biochip. (A) Micrograph of nucleic acid biochip, (B) calibration curve of nucleic acid detection, (C) immuno biochip competition calibration curve (Logit representation).
ing the affinity reaction with an anti-IgG antibody. First of all, a study of the saturation of the affinity binding sites present in each spot was carried out by applying to each spots a 2 l drop containing various concentrations of anti-IgG antibody. The results obtained (data not show) were in agreement with the typical expected saturation curve. The curve was a sigmoid type with a plateau, levelling off at high antibody concentrations. The part of the curve exhibiting the higher slope enables to access the optimum anti-IgG concentration to be use in the competitive assay, that is, 5 g/ml. The Fig. 7C presents the Logit representation of the competition curve obtained, B (B/B0 ) Logit = log B0 1 − (B/B0 )
3.2.2. Nucleic acid electro-chemiluminescent biochip The nucleic acid biochip is composed of 24 spots of d(A)22 bearing beads. Each spot could be incubated separately with a 2 l drop of the complementary strand. The micrograph presented in Fig. 7B shows how various amounts of biotinylated single strand could be detected with the electro-chemiluminescent biochip. The quantification (Fig. 7A) of the triplicated spots for each concentration lead to obtain the calibration curve presented. A detection limit of 20 fmol of target analyte in the sample was found, with a two decade detection range. This detection limit corresponds to a number of target molecules of 1011 . Compared to the detection limits obtained with fluorescence (Trau et al., 2002; Epstein et al., 2002) or radiolabelled-based (Salin et al., 2002) systems, the present method exhibits a slightly lower sensitivity but a larger detection range.
4. Conclusions
3.2.3. Immuno electro-chemiluminescent biochip The electro-chemiluminescent immuno biochip was used in a competitive format in which free human IgG from the sample were in competition with immobilised antigen, dur-
A new biochip design has been developed and validated. Addressed texturing of the surface of an original conducting elastomer, based on the entrapment of Sepharose beads at the interface, was performed and enabled a local fifty-fold
where B is the electro-chemiluminescent signal at a particular free IgG concentration and B0 the signal in the absence of free IgG. As it can be seen, a limit of detection of 0.2 ng could be reached with a detection ranging over four decades at least. This limit corresponds to a number of molecules below 1010 .
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
increase of the specific surface of the conducting support. The consequence of this increase was shown to be a jump of the local electrochemical activity which induced a well defined and localised electro-chemiluminescent signal. Applied to enzyme, nucleic acid and antibody-based electro-chemiluminescent biochips, such system proved its analytical potentialities. First, with all the different biochip systems, the quantitations were performed with the same light measurement apparatus (CCD camera) under identical conditions. Moreover, clearly distinguished light spots could be observed at the surface of a unique material electrode, without glowing of the signals, due to lateral diffusion of the reacting species. Finally, acceptable detection limits were obtained for the three studied biochips. The extension of the present system to other bio-molecules is now under study.
References Bard, A.J., 1996. Electroanalytical Chemistry, vol. 19. Marcel Dekker, New York. Charles, P.T., Vora, G.J., Andreadis, J.D., Fortney, A.J., Meador, C.E., Dulcey, C.S., Stenger, D.A., 2003. Fabrication and surface characterisation of DNA microarrays using amine- and thiol-terminated oligonucleotide probes. Langmuir 19 (5), 1586–1591. Elliott, J.M., Birkin, P.R., Barlett, P.N., Attard, G.S., 1999. Platinum microelectrode with unique high surface areas. Langmuir 15, 7411– 7415. Epstein, J.R., Lee, M., Walt, D.R., 2002. High-density fiber-optic genosensor microsphere array capable of zeptomole detection limits. Anal. Chem. 8, 1836–1840. Evans, U., Colavita, P.E., Doescher, M.S., Schiza, M., Myrick, M.L., 2002. Construction and characterisation of a nanowell electrode array. Nano Lett. 2 (6), 641–645. Marquette, C.A., Blum, L.J., 1998. Electrochemiluminescence of luminol for 2,4-D optical immunosensing in a flow injection analysis system. Sens. Actuators, B 51, 100–106.
203
Marquette, C.A., Raveau, S., Blum, L.J., 2000. Luminol electrochemiluminescence-based biosensor for total cholesterol determination in natural samples. Anal. Lett. 33 (9), 1779–1796. Marquette, C.A., Lawrence, I., Polychronakos, C., Lawrence, M.F., 2002. Impedance-based DNA chip for direct Tm measurement. Talanta 56, 763–768. Marquette, C.A., Thomas, D., Degiuli, A., Blum, L.J., 2003. Design of luminescent biochips based on enzyme, antibody or DNA composite layer. Anal. Bioanal. Chem. 377 (5), 922–928. Marquette, C.A., Blum, L.J., 2004. Direct immobilisation in PDMS for DNA, protein and enzyme fluidic biochips. Anal. Chim. Acta 506 (2), 127–132. Patolsky, F., Weizmann, Y., Willner, I., 2002. Redox-active nucleic-acid replica for the amplified bioelectrocatalytic detection of viral DNA. J. Am. Chem. Soc. 124 (5), 770–772. Pekala, R.W., Farmer, J.C., Alviso, C.T., Tran, T.D., Mayer, S.T., Miller, J.M., Dunn, B., 1998. Carbon aerogels for electrochemical applications. J. Non-Cryst. Solids, 74–80. Salin, H., Vujasinovic, T., Mazurie, A., Maitrejean, S., Menini, C., Mallet, J., Dumas, S., 2002. A novel sensitive microarray approach for differential screening using probes labelled with two different radioelements. Nucleic Acids Res. 30 (4), 4–17. Sullivan, M.G., Utomo, H., Fagan, P.J., Ward, M.D., 1999. Automated electrochemical analysis with combinatorial electrode arrays. Anal. Chem. 71 (19), 4369–4375. Suzuki, M., Akaguma, H., 2000. Chemical cross-talk in flow-type integrated enzyme sensors. Sens. Actuators, B 64, 136–141. Szunerits, S., Tam, J.M., Amatore, C., Walt, D.R., 2003. Spatially resolved electrochemiluminescence on an array of electrode tips. Anal. Chem. 75 (17), 4382–4388. Trau, D., Lee, T.M.H., Lao, A.I.K., Lenigk, R., Hsing, I., Ip, N.Y., Carles, M.C., Sucher, N.J., 2002. Genotyping on a complementary metal oxide semiconductor silicon polymerase chain reaction chip integrated DNA microarray. Anal. Chem. 74 (13), 3168–3173. Yamaguchi, A., Jin, P., Tsuchiyama, H., Masuda, T., Sun, K., Matsuo, S., Misawa, H., 2002. Rapid fabrication of electrochemical enzyme sensor chip using polydimethylsiloxane microfluidic channel. Anal. Chim. Acta 468, 143–152. Yang, M., Liu, C., Qian, K., He, P., Fang, Y., 2002. Study on the electrochemiluminescence behavior of ABEI and its application in DNA hybridisation analysis. Analyst 127 (9), 1267–1271.
Conducting elastomer surface texturing: a path to electrode spotting Application to the biochip production Christophe A. Marquette∗ , Lo¨ıc J. Blum Laboratoire de Génie Enzymatique et Biomoléculaire, Université Claude Bernard Lyon 1, UMR 5013 EMB2-CNRS, Bˆat. CPE, Building 43, 11 November 1918, Villeurbanne 69622, Cedex, France Received 17 September 2003; received in revised form 20 January 2004; accepted 20 January 2004 Available online 10 May 2004
Abstract A new active support for electro-chemiluminescent biochip preparation has been developed. This material was based on an original material composed of graphite modified polydimethyl siloxane (PDMS). The addressed inclusion of Sepharose beads at the surface of this elastomeric electrode generated interesting local high specific surface. The electrode was characterised by electrochemical (cyclic voltametry, chronoamperomatry) and imaging (scanning electron microscopy (SEM)) methods, and a surface area increase factor of 50 was found, linked to the texturing of the surface generated by the presence of the Sepharose beads. The consequence of this increase was shown to be a jump of the local electrochemical activity which induced a well defined and localised electro-chemiluminescent signal. The new material was used to design biochips based on the electro-chemiluminescent reaction of luminol with enzymatically produced hydrogen peroxide. Thus, when using beads bearing bio-molecules such nucleic acid or human IgG, in conjunction with glucose oxidas-labelled DNA or antibody, sensitive biochips could be obtained with detection limits of 1011 and 1010 molecules, respectively. Multi-parameter enzyme-based biochips could also be achieved by locally adsorbing, at the PDMS-graphite surface, either glucose oxidase, lactate oxidase or choline oxidase. Detection limits of 10 M for lactate and choline and 20 M for glucose were found, with detection ranging over two decades at least. © 2004 Elsevier B.V. All rights reserved. Keywords: Biochip; Conducting elastomer; DNA; Electro-chemiluminescence; Enzyme; Immuno; Microarray; PDMS
1. Introduction Biochip technology has become in the last 5 years a large field of investigation, by both the academic and the industrial laboratories. As a consequence, a large number of arraying systems as well as transduction and detection technologies have been developed. Nevertheless, those biochips could be separated in two main categories: (i) the passive arrays (Fig. 1A), composed of a solid support solely dedicated to the immobilisation of bio-molecules (Charles et al., 2003; Salin et al., 2002), and (ii) the active arrays (Fig. 1B and C) in which the support acts in the transduction and/or in the immobilisation steps (Epstein et al., 2002; Patolsky et al., 2002; Marquette et al., 2002). The latter represent some interesting systems since they could enable the ad∗ Corresponding author. Tel.: +33-472-44-8214; fax: +33-472-44-7970. E-mail address: [email protected] (C.A. Marquette).
0956-5663/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2004.01.033
dressing of the analytical signal, but also the use of transduction systems incompatible with the passive arrays. Indeed, a number of detection methods, based on electrochemical processes (amperometry, potentiometry and electro-generated luminescence), known to be sensitive systems, required the use of an active conducting support. However, amperometry and potentiometry are based on electrochemical measurements, and designing such active array involves the production of a matrix of individually wired electrodes (Sullivan et al., 1999; Yamaguchi et al., 2002), with number of technical difficulties. On the contrary, electro-generated luminescent biochips could be prepared at the surface of a unique conducting support such a glassy carbon square (Marquette et al., 2003). Indeed, the electrochemical process generating the light emission is identical for each spot present at the surface of the support. The electro-luminescent system involved, could then be a ruthenium complex (Szunerits et al., 2003), a luminol-based system (Marquette and Blum, 1998; Yang et al., 2002) or an
198
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
signal (A)
e-
(B)
signal signal
(C)
1.1.3.17, from Alcanigenes species), glucose oxidase (grade I, EC 1.1.3.4, from Aspergillus niger), human IgG, lactate oxidase (EC number not available, from Pediococcus), luminol (3-aminophtalhydrazide), Sepharose 4B beads (mean diameter: 100 m), Sepharose 4B beads bearing human IgG, Sepharose 4B beads bearing d(A)22 were purchased from Sigma (France). Glucose oxidase-labelled avidin was obtained from Jackson Immuno Research. PDMS precursor and curing agent (Sylgard 184) were supplied by Dow Corning. All buffers and aqueous solutions were made with distilled de-mineralised water.
e-
2.2. Biochip preparation Fig. 1. Schematic representation of the different biochip categories. (A) Passive support biochip, (B) active and individually wired electrode array, (C) active support biochip.
hydrogen peroxide producing system (oxidase type enzymes) coupled to the luminol reaction (Marquette et al., 2000; Marquette et al., 2003). This latter is of particular interest since it represents an enhancement of the luminol-based system. Moreover, the use of oxidase-tagged DNA, could enable the achievement of nucleic acid electro-chemiluminescent biochip, the use of oxidase-labelled antibodies, the production of immuno electro-chemiluminescent biochip (Marquette and Blum, 1998) and the immobilisation of oxidases lead to enzyme-based electro-chemiluminescent biochip (Marquette et al., 2003). Unfortunately, geographical limitation problems of the signal are classically pointed out (Suzuki and Akaguma, 2000) for enzyme-based electro-chemiluminescent biochip, caused by the lateral diffusion of the hydrogen peroxide produced. The present study proposes to solve such cross-talk problems. Indeed, since the generated luminescent signal is triggered by an electrochemical process, we investigate the possibility of locally increasing this process by generating addressed high specific surface, at an active support interface. The active support is composed of an original elastomer/graphite organised solid, at the surface of which, micro-beads could be physically entrapped (Marquette and Blum, 2004). Group of beads could then be spotted and are used to locally increase the specific surface of the active support. When beads bearing bio-molecules are spotted, they could be used not only to immobilise the bio-molecules, but also to increase the active specific surface. Such approach results in the achievement of electro-chemiluminescent biochips.
The biochips were prepared by spotting 0.3 l drops (Fig. 2) of a solution prepared by mixing 45 l of an aqueous solution composed of glycerol 2.5% and NaCl 0.05 M, with 5 l of wet beads (360 mg/ml of wet beads). The 0.3 l drops were deposited on a flat PVC substrate every 3 mm and dried under a tungsten lamp during 30 min. The presence of glycerol in the spotting solution enabled the achievement of homogeneous and smooth dry bead spots. The dry bead spots were then covered with graphite by spreading over the beads a 0.1 g/ml solution of graphite (TIMREX T15 from TIMCAL, particle mean diameter: 15 m) in chloroform. After a few seconds drying under a tungsten lamp a thin graphite layer (50–100 m thickness) is obtained. The
2. Experimental 2.1. Reagents Biotin labelled anti-human IgG (␥-chain specific), biotin labelled oligonucleotides d(T)22 , choline oxidase (EC
Fig. 2. Steps for the preparation of the addressed texturing of the PDMS-graphite conducting elastomer.
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
2.3. Electrochemical and physical characterisation The electrochemical measurements were performed with a PGZ-100 VoltaLab potentiostat (Radiometer, France). The experiments were carried out in a 2 l volume drop, localised on a bead spot, in which the reference electrode was dipped. The cyclic voltametry and chronoamperometry experiments were performed in an electrolyte solution composed of LiClO4 0.1 M, containing in addition potassium ferricyanide at a concentration of 10 mM. The voltamograms were recorded at 100 mV s−1 between −600 and +600 mV versus a silver pseudo-reference electrode. The chronoamperometry measurements were performed at a +500 mV applied potential during 10 s. The SEM images were obtained with a S800 FEG (Hitachi) electron microscope at 15 kV. 2.4. Nucleic acid assay As a model of oligonucleotide hybridisation process, a d(T)22 –d(A)22 couple was used. Briefly, the biochip was saturated during 30 min with a solution of buffer A (30 mM Veronal (diethylbarbiturate), pH 8.5 containing in addition 30 mM KCl, 0.2 M NaCl, 0.1% Tween and 1% BSA), then incubated for 1 h with biotinylated d(T)22 at a particular concentration in buffer A. The biochip was then washed 10 min with buffer B (30 mM Veronal, pH 8.5 containing 30 mM KCl, 0.2 M NaCl), afterwards saturated with buffer A during 20 min and then incubated with a glucose oxidase-labelled streptavidin solution at a 1 g/ml concentration during 30 min. The array was finally washed 20 min with buffer B and was ready to be measured.
a CCD light measurement system (Intelligent Dark Box II, Fuji Film) and the emitted signal integrated during 3 min. The electro-chemiluminescent reaction was triggered by applying a +850 mV potential between the beads modified conducting elastomer and a bare conducting elastomer, both integrated in the biochip. The measurement solution was composed of glucose 20 mM and luminol 50 M in Veronal buffer 30 mM, pH 8.5, added of KCl 30 mM. In the particular case of the enzyme-based biochip, the oxidase substrate was either glucose, lactate or choline, depending of the experiment carried out.
3. Results and discussion 3.1. Electrochemical and physical properties of the PDMS/graphite elastomer The electrochemical properties of the PDMS/graphite elastomer were first studied by cyclic voltametry. The Fig. 3A presents the voltamograms obtained by cycling the bare and the Sepharose bead-modified conducting elastomer in the presence of 10 mM potassium ferricyanide. As it can be seen, the bare elastomer presents an acceptable conductivity but no electrocatalytic properties (black line) against the Fe(CN)6 3− . On the contrary, the bead-modified 8e-5 4e-5 Current (A/cm²)
dried layers (spots and graphite) were then transferred to the PDMS interface by pouring a mixture of precursor and curing agent onto the PVC substrate and heating for 90 min. The biochip preparation was then terminated by peeling off the PDMS-graphite polymer. At that time, the beads are definitely entrapped at the interface, and could not be detached easily.
199
0e+0 -4e-5 -8e-5 -1
-0,5
(A)
0
0,5
1
Potential (V) 1e-4
40
2.5. Immuno biochip assay In a way similar to that used for the nucleic acid biochip, the immuno biochip was saturated during 30 min in buffer A, then incubated for 1 h with biotinylated anti-IgG antibody at a particular concentration in buffer A. The biochip was then washed 10 min with buffer B, saturated with buffer A during 20 min and then incubated with glucose oxidase-labelled streptavidin at a 1 g/ml concentration during 30 min. The array was then washed 20 min with buffer B and was ready to be measured. 2.6. Electro-chemiluminescent biochip measurement The different electro-chemiluminescent biochips were measured in an identical way. The array was disposed under
Current (A/cm²)
8e-5
10 4 0
0e+0 -4e-5 -8e-5 -1e-4
(B)
20
4e-5
-1
-0,5
0
0,5
1
Potential (V)
Fig. 3. (A) Cyclic voltamograms obtained at a bare PDMS-graphite electrode (black line) and at a textured PDMS-graphite electrode (grey line). (B) Cyclic voltamograms obtained at a bare PDMS-graphite electrode (grey line) and at a textured PDMS-graphite electrode (black line) prepared with different amounts of beads per spot. Voltamograms recorded at 100 mV s−1 between −600 and +600 mV in Fe(CN)6 3− 10 mM.
200
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
elastomer exhibits a good electrocatalytic activity (grey line) shown by the oxidation and reduction peaks of the redox couple. These distinct electrocatalytic activities can be attributed to a difference of specific surface between the bare and the modified elastomer. Indeed, the electrode roughness is well known to influence the efficiency of the electron transfer, keeping the geometrical surface constant (Elliott et al., 1999; Pekala et al., 1998; Evans et al., 2002). Chronoamperometry was used to estimate the active surface of the bead-modified elastomer. The Cottrell equation (Bard, 1996) was used to calculate the ferricyanide diffusion coefficient at the bare elastomer electrode. An estimated active surface was then subsequently calculated from the chronoamperometry curve obtained at the bead-modified polymer. For an identical geometric area (0.07 cm2 ), a fifty-fold increase of the specific surface was obtained with the modified elastomer. The SEM image (Fig. 4A) of one island of beads at the interface of the elastomer demonstrates this local surface roughness. Sepharose beads introduce heterogeneity in the conducting surface, increasing the surface area. The 3D representation of one Sepharose bead (Fig. 4B) shows how it is inserted in the polymer and how it emerges at the surface. A schematic diagram of the layer organisation is presented
in the Fig. 4C. The presence of the porous Sepharose beads during the graphite deposition creates deformations of the conducting surface, generating then a higher exposed surface area; the area under the beads remaining electroactive, without detectable diffusion problem. As an evidence of the effect of the presence of Sepharose beads on the electrochemical activity, the Fig. 3B presents the cyclic voltamograms obtained with spots containing different amounts of beads, from 40 to 0. As it can be seen, the electrode activity decreases with the number of beads in the spots. The higher the number of beads, the higher the specific surface at the spot location and consequently the higher the electrochemical activity. 3.2. Biochip applications The Fig. 5 summarises the different analytical systems used: (i) an enzyme-based biochip (Fig. 5A) composed of oxidase type enzymes physically adsorbed, co-localised with the spot of beads, (ii) a nucleic acid biochip (Fig. 5B) in which the spotted beads are bearing single strand d(A)22 oligonucleotide, (iii) an immuno biochip (Fig. 5C) prepared by spotting human IgG-immobilised beads. The last two systems are reacting with biotinylated d(T)22 and anti-human
Fig. 4. (A) Electron microscopy image (15 kV) of Sepharose beads spot at the surface of the PDMS-graphite elastomer, (B) 3D representation of the SEM image of a particular bead, (C) schematic representation of the layer organisation.
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
201
Fig. 5. The different electro-chemiluminescent biochip formats. (A) Oxidase type enzyme-based biochip, (B) nucleic acid-based biochip, (C) immuno biochip.
IgG, respectively, and affinity interactions with targets were detected with an avidin-glucose oxidase conjugate. 3.2.1. Enzymes-based electro-chemiluminescent biochip The biochip is composed of 18 spots of beads organised as 3 columns of 6 spots. Each row is composed of one spot of adsorbed lactate oxidase, one spot of adsorbed choline oxidase and one spot of adsorbed glucose oxidase. By covering the biochip with a six channels PDMS block polymer, six samples could be injected and analysed in a single measurement.
The Fig. 6 presents the performances of the enzyme-based biochip, each row of the micrographs corresponds to a particular channel in which a particular concentration of analyte was injected. The micrographs show how a single biochip could be used to detect three different analytes, without cross-reactivity problems. Moreover, the spatial resolution of the signal appears to be satisfying, whereas, the different reacting species, that is, the hydrogen peroxide produced and the luminol, are free to diffuse at the surface of the conducting elastomer without generating contamination problem. The local high surface area permits then to avoid the
202
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
Fig. 6. Electro-chemiluminescent enzyme-based biochip. (A) Micrograph and calibration curve of the lactate detection, (B) micrograph and calibration curve of the choline detection, (C) micrograph and calibration curve of the glucose detection.
glowing of the analytical signal, leading to more accurate quantitative results. The graphs corresponding to the calibration curves obtained after treatment of the numerised pictures illustrate the performances of the biochip. Detection limits of 10 M for lactate and choline and 20 M for glucose were found, with detection ranging over two decades at least. Regarding sensitivity, these results compared well those with previously obtained with electro-chemiluminescent biosensors (Marquette et al., 2003).
Fig. 7. Electro-chemiluminescent affinity biochip. (A) Micrograph of nucleic acid biochip, (B) calibration curve of nucleic acid detection, (C) immuno biochip competition calibration curve (Logit representation).
ing the affinity reaction with an anti-IgG antibody. First of all, a study of the saturation of the affinity binding sites present in each spot was carried out by applying to each spots a 2 l drop containing various concentrations of anti-IgG antibody. The results obtained (data not show) were in agreement with the typical expected saturation curve. The curve was a sigmoid type with a plateau, levelling off at high antibody concentrations. The part of the curve exhibiting the higher slope enables to access the optimum anti-IgG concentration to be use in the competitive assay, that is, 5 g/ml. The Fig. 7C presents the Logit representation of the competition curve obtained, B (B/B0 ) Logit = log B0 1 − (B/B0 )
3.2.2. Nucleic acid electro-chemiluminescent biochip The nucleic acid biochip is composed of 24 spots of d(A)22 bearing beads. Each spot could be incubated separately with a 2 l drop of the complementary strand. The micrograph presented in Fig. 7B shows how various amounts of biotinylated single strand could be detected with the electro-chemiluminescent biochip. The quantification (Fig. 7A) of the triplicated spots for each concentration lead to obtain the calibration curve presented. A detection limit of 20 fmol of target analyte in the sample was found, with a two decade detection range. This detection limit corresponds to a number of target molecules of 1011 . Compared to the detection limits obtained with fluorescence (Trau et al., 2002; Epstein et al., 2002) or radiolabelled-based (Salin et al., 2002) systems, the present method exhibits a slightly lower sensitivity but a larger detection range.
4. Conclusions
3.2.3. Immuno electro-chemiluminescent biochip The electro-chemiluminescent immuno biochip was used in a competitive format in which free human IgG from the sample were in competition with immobilised antigen, dur-
A new biochip design has been developed and validated. Addressed texturing of the surface of an original conducting elastomer, based on the entrapment of Sepharose beads at the interface, was performed and enabled a local fifty-fold
where B is the electro-chemiluminescent signal at a particular free IgG concentration and B0 the signal in the absence of free IgG. As it can be seen, a limit of detection of 0.2 ng could be reached with a detection ranging over four decades at least. This limit corresponds to a number of molecules below 1010 .
C.A. Marquette, L.J. Blum / Biosensors and Bioelectronics 20 (2004) 197–203
increase of the specific surface of the conducting support. The consequence of this increase was shown to be a jump of the local electrochemical activity which induced a well defined and localised electro-chemiluminescent signal. Applied to enzyme, nucleic acid and antibody-based electro-chemiluminescent biochips, such system proved its analytical potentialities. First, with all the different biochip systems, the quantitations were performed with the same light measurement apparatus (CCD camera) under identical conditions. Moreover, clearly distinguished light spots could be observed at the surface of a unique material electrode, without glowing of the signals, due to lateral diffusion of the reacting species. Finally, acceptable detection limits were obtained for the three studied biochips. The extension of the present system to other bio-molecules is now under study.
References Bard, A.J., 1996. Electroanalytical Chemistry, vol. 19. Marcel Dekker, New York. Charles, P.T., Vora, G.J., Andreadis, J.D., Fortney, A.J., Meador, C.E., Dulcey, C.S., Stenger, D.A., 2003. Fabrication and surface characterisation of DNA microarrays using amine- and thiol-terminated oligonucleotide probes. Langmuir 19 (5), 1586–1591. Elliott, J.M., Birkin, P.R., Barlett, P.N., Attard, G.S., 1999. Platinum microelectrode with unique high surface areas. Langmuir 15, 7411– 7415. Epstein, J.R., Lee, M., Walt, D.R., 2002. High-density fiber-optic genosensor microsphere array capable of zeptomole detection limits. Anal. Chem. 8, 1836–1840. Evans, U., Colavita, P.E., Doescher, M.S., Schiza, M., Myrick, M.L., 2002. Construction and characterisation of a nanowell electrode array. Nano Lett. 2 (6), 641–645. Marquette, C.A., Blum, L.J., 1998. Electrochemiluminescence of luminol for 2,4-D optical immunosensing in a flow injection analysis system. Sens. Actuators, B 51, 100–106.
203
Marquette, C.A., Raveau, S., Blum, L.J., 2000. Luminol electrochemiluminescence-based biosensor for total cholesterol determination in natural samples. Anal. Lett. 33 (9), 1779–1796. Marquette, C.A., Lawrence, I., Polychronakos, C., Lawrence, M.F., 2002. Impedance-based DNA chip for direct Tm measurement. Talanta 56, 763–768. Marquette, C.A., Thomas, D., Degiuli, A., Blum, L.J., 2003. Design of luminescent biochips based on enzyme, antibody or DNA composite layer. Anal. Bioanal. Chem. 377 (5), 922–928. Marquette, C.A., Blum, L.J., 2004. Direct immobilisation in PDMS for DNA, protein and enzyme fluidic biochips. Anal. Chim. Acta 506 (2), 127–132. Patolsky, F., Weizmann, Y., Willner, I., 2002. Redox-active nucleic-acid replica for the amplified bioelectrocatalytic detection of viral DNA. J. Am. Chem. Soc. 124 (5), 770–772. Pekala, R.W., Farmer, J.C., Alviso, C.T., Tran, T.D., Mayer, S.T., Miller, J.M., Dunn, B., 1998. Carbon aerogels for electrochemical applications. J. Non-Cryst. Solids, 74–80. Salin, H., Vujasinovic, T., Mazurie, A., Maitrejean, S., Menini, C., Mallet, J., Dumas, S., 2002. A novel sensitive microarray approach for differential screening using probes labelled with two different radioelements. Nucleic Acids Res. 30 (4), 4–17. Sullivan, M.G., Utomo, H., Fagan, P.J., Ward, M.D., 1999. Automated electrochemical analysis with combinatorial electrode arrays. Anal. Chem. 71 (19), 4369–4375. Suzuki, M., Akaguma, H., 2000. Chemical cross-talk in flow-type integrated enzyme sensors. Sens. Actuators, B 64, 136–141. Szunerits, S., Tam, J.M., Amatore, C., Walt, D.R., 2003. Spatially resolved electrochemiluminescence on an array of electrode tips. Anal. Chem. 75 (17), 4382–4388. Trau, D., Lee, T.M.H., Lao, A.I.K., Lenigk, R., Hsing, I., Ip, N.Y., Carles, M.C., Sucher, N.J., 2002. Genotyping on a complementary metal oxide semiconductor silicon polymerase chain reaction chip integrated DNA microarray. Anal. Chem. 74 (13), 3168–3173. Yamaguchi, A., Jin, P., Tsuchiyama, H., Masuda, T., Sun, K., Matsuo, S., Misawa, H., 2002. Rapid fabrication of electrochemical enzyme sensor chip using polydimethylsiloxane microfluidic channel. Anal. Chim. Acta 468, 143–152. Yang, M., Liu, C., Qian, K., He, P., Fang, Y., 2002. Study on the electrochemiluminescence behavior of ABEI and its application in DNA hybridisation analysis. Analyst 127 (9), 1267–1271.