Surface techniques for an electrochemical DNA biosensor

Surface techniques for an electrochemical DNA biosensor

Biosensors & Bioelectronics Vol. 12. No. 8, pp. 729–737, 1997  1997 Published by Elsevier Science Limited All rights reserved. Printed in Great Brita...

160KB Sizes 0 Downloads 36 Views

Biosensors & Bioelectronics Vol. 12. No. 8, pp. 729–737, 1997  1997 Published by Elsevier Science Limited All rights reserved. Printed in Great Britain 0956–5663/97/$17.00 PII: S0956–5663(97)00040-7

Surface techniques for an electrochemical DNA biosensor Isabella Moser,*† Thomas Schalkhammer,‡ Fritz Pittner‡ & Gerald Urban† †Technische Universita¨t Wien, Institut fu¨r Allgemeine Elektrotechnik und Elektronik und Ludwig Boltzmann Institut fu¨r Hirnkreislaufforschung, Gusshausstr. 27-29, A-1040, Vienna, Austria ‡Institut fu¨r Biochemie und Molekulare Zellbiologie, Vienna Biocenter, Dr. Bohrg. 9; A-1030 Vienna, Austria

Abstract: The surface of platinum electrodes was modified by a technique that provides binding sites for covalent site-specific immobilization of DNA. In the course of immobilization certain substances such as thiols, iodine and iodide were brought in contact with the metal. The effect on the electrochemical characteristics of thin film platinum electrodes caused by chemisorption of these substances was investigated. Voltage pulsing was very effective in restoring the electrocatalytical properties of the electrodes. A 20 bp oligonucleotide was immobilized via primary amino groups or terminal phosphate and immobilization efficiency of the different immobilized strands was determined. Mercapto-silane coatings and controlled introduction of charged groups in the course of immobilization prevented unspecific DNA binding caused by adsorption. 1997 Elsevier Science Limited Keywords: modified platinum electrodes, DNA immobilization, mercapto-silane, electrochemical electrode characteristic

INTRODUCTION The surface of an amperometric affinity biosensor has to fulfil several important functions: providing immobilization support for the biocomponent; allowing only specific binding of the analyte; and providing an electrochemical active electrode for the detection of the analyte. The first two demands can be fulfilled by surface modification of the electrode, the electro-

*To whom correspondence should be addressed. Current address: Institut fu¨r Mikro-System technik, AlbertLudwigs-Universitat Freiburg, D-79085 Freiburg, Germany. Tel: ++49 761 203 8014 Fax: ++49 761 203 8012 Email: [email protected]

chemical properties of which should not be altered in the ideal case. But it is obvious that in the course of immobilization the chemical treatment of a metal electrode alters the surface by chemical reactions or by chemisorption. The present study focuses on the influence of different immobilization techniques for DNA on the surface of thin film platinum electrodes and on the possibilities of their electrochemical purification. Recently, there has been an increasing demand to find simple and rapid methods for the detection of specific DNA sequences, which can also be used easily in non-specialized laboratories. The detection of specific nucleic acids sequences is of importance because more than 4000 inherited diseases are known, and much effort is needed to identify the underlying mutations. Another 729

Isabella Moser et al.

approach is the rapid detection and identification of micro-organisms in the fields of environmental control, the food industry, the clinical laboratory, and everywhere where only forensic DNA methods can be applied. Most of the traditional techniques in molecular biology are based on hybridization. A singlestranded (ss) DNA probe binds to its complementary strand present in a sample, and a double helix structure is built. The degree of hybridization of a complementary probe to the DNA sample is a measure of the amount of that specific sequence in the sample. For a sandwich hybridization assay a small DNA single strand (probe) is immobilized onto a solid support. After hybridization between the ss-analyte DNA and the probe a second hybridization between the single-stranded end of the sample DNA and a signal labeled DNA provides the signal. In the case of a common sandwich assay the signal generating species is an enzyme such as horseradish peroxidase (Nikiforov & Rogers, 1995). But for most of the applications the amplification by enzyme reaction is not high enough. Polymerase chain reaction can overcome this problem by amplifying the sample DNA up to a range which can be detected by optical or, for example, amperometric measurement methods (Mullis & Faloona, 1987). The complementary structure of DNA offers a precise biorecognitation system suitable for biosensors of the affinity type. In the literature different approaches for DNA-sensors using different types of transducers have been described. Using surface plasmon resonance the hybridization event can be observed directly (Schwarz et al., 1991 and Nilsson et al., 1995). The difference in electrochemical behavior of ss-DNA and hybridized DNA is the measurement principle of Hall et al. (1994), while the intercalation of a redox-active molecule in hybridized DNA is detected by cyclovoltammetry (Millan et al., 1994) or by chronopotentiometry (Wang et al., 1996). In this case peptide nucleic acid probes were used. Common to all of the methods described is the need to immobilize one strand of DNA. Unspecific immobilization distorts the structure of the DNA, which causes incorrect hybridization with a decrease in melting temperature and a broadening of the melting temperature range. Because the binding of only 3% of the bases to the support leads to total inaccessibility of the immobilized DNA (Bu¨nemann et al., 1982), site-specific immobilization of a ss-DNA 730

Biosensors & Bioelectronics

at one end of the strand is recommended. In contrast to incorrect hybridization the adsorption of the signal DNA can also pretend a higher signal. Therefore, this study focuses on special methods for site-specific immobilization of ssDNA, providing simultaneously appropriate surface modifications of the sensor, which additionally prevents adsorption, and enables the electrode to detect a redox signal provided by a label.

EXPERIMENTAL Materials Glucose oxidase (EC was purchased from Sigma (St Louis, MO) or from Boehringer (Mannheim). L-lactate oxidase (EC was purchased from Boehringer. T4 polynucleotide kinase was a product from Sigma. Mercapto-silane (3-Trimethoxysilyl-1propanethiol) was purchased from Fluka (Buchs), the water-soluble CDI N-cyclohexyl-N′-[2-(Nmethylmorpholino)-ethyl]-carbodiimide-4-toluene sulfonate was purchased from Sigma, and the Llactate standard (1 M) solution (125.440) was part of a L-lactate kit from Boehringer. For DNA labeling [␥-32P]ATP and, as a marker, [␣-32P]ATP from Amersham, UK, were used. All other chemicals were obtained from Sigma and were of the highest purity available. A synthetic oligonucleotide with the sequence 5′ ATTGCGGGTTCTAATCCAGA 3′ was used as a DNA probe. Besides a phosphate buffer (pH 7, 1/15 M), a citrate buffer (SSC) and a Tris buffer (kinase I buffer) were used (20 × SSC: 3 M NaCl, 0·3 M sodium citrate, addition of 10 N NaOH till a pH of 7·0 is reached; kinase I buffer: 0·5 M Tris-HCl pH 7·6, 0·1 M MgCl2, 50 mM dithiothreitol, 1 mM spermidine, and 1 mM EDTA). Methods The electrode structures were produced in a photolithographic lift-off process using image reversal photolithography (Moser et al., 1995). As sensor substrates sodium silicate glass sheets of 0·3 mm thickness or aluminum oxide of 0·6 mm were used. Layers of different metals were evaporated by an electron gun (Balzers, Bak 550) under high vacuum conditions (5 × 10−7 mbar). First, an 80 nm titanium layer was

Biosensors & Bioelectronics

evaporated, serving as an adhesion layer. Platinum layers up to a thickness of 60 nm were evaporated on top of the titanium layer. These layers were insulated by a 1000 nm silicon nitride layer and structured by plasma etching in a CF4/O2 plasma. The platinum surface was cleaned by a reactive ion-etched oxygen plasma for 3 min. A surface characterization of the described unmodified electrodes is given in Aschauer et al. (1995). Amperometric measurement For amperometric measurement a three-electrode arrangement in a thermostated flow-through cell was used. As a working electrode a simple oneelectrode structure (A = 0·64 mm2) together with an auxiliary platinum wire electrode and a macroscopic silver/silver chloride electrode was used. All potentials given in the present work relate to the Ag/AgCl (3 M KCl) reference electrode. For electrode surface characterization cyclovoltammetry was performed. The electrolyte used for cyclovoltammetry was a pH 7 potassium phosphate buffer (1/15 M), and the scan speed was 47 mV/s. Measurement of radioactivity The efficiency of [␥-32P]ATP labeled oligonucletide immobilized on the electrodes was measured in a linear analyzer (Berthold). In this way the distribution over the sensor areas of four sensors at the same time could be determined. Radio labeling of the oligonucleotide To a portion of 40 ␮l water containing 50 pM of oligonucleotide 10 ␮l kinase I buffer (10x), 7 ␮lATP (7 ␮Ci), and 10 U (1 ␮l) T4 polynucleotide kinase were added. The enzymatic reaction was performed at 37°C for 30 min, then it was stopped by keeping the reaction mixture at 65°C for 3 min. The so-labeled oligonucleotide was purified on a Sephadex G 25 column (about 0·5 ml). The result of the labeling was checked by gel electrophoreses of the radioactive fractions. Only fractions without free ATP were used for immobilization or hybridization purpose. Pretreatment of the metal surface The electrodes were placed into 1 ml sealed polyethylene reaction vessels. The bottoms of the

Surface techniques for an electrochemical DNA biosensor

vessels tapered off, therefore the electrodes stood upright in the vessels. The vessels were filled with solution volumes up to 120 ␮l, so only the sensing area and insulated parts of the electrodes (and not the bonding bead) were immersed in the solution. Platinum electrodes were reduced in a 1% aqueous sodium-dithionite solution for 30 min at 56°C. After rinsing with water the electrodes were oxidized in a solution of 2·5% potassium dichromate in 15% nitric acid for 30 min at 56°C. Subsequently, the electrodes were rinsed with 0·02 N HCl and water, followed by acetone, then dried and immediately exposed to silane. Coupling with 3-amino-propyl-triethoxysilane The pH of a 10% aqueous solution of 3-aminopropyl-triethoxysilane was adjusted to a pH of 3·5 by using 6 N HCl. The oxidized electrodes were incubated in this solution at 37°C for 30 min. After rinsing with water the electrodes were dried at 110°C for 15 min. The ethoxy moieties act as leaving groups, binding occurs with the platinum oxide as well as intermolecular cross-linking between the silane molecules. The heat treatment at 110°C stabilizes the silane layer by further cross-linking. Electrochemical treatment Electrochemical cleaning procedures were performed as follows: voltage pulsing for 1 min at −500 and 1200 mV and cycling in pH buffer pH 7·0 up to 1200 mV or up to 1300 mV [Ag/AgCl] for sulfur-containing compounds.

RESULTS AND DISCUSSION The phosphate moieties of the DNA cause a negative charge of the whole molecule. Using a support which is derivatized by aminopropyl silane, results in an ion bond between the DNA and the support. Therefore a new derivatization technique with mercapto-silane was employed (see Scheme 1). According to the method of Weetall (1976), the electrodes were treated with 10% mercaptosilane in toluene; but this procedure it not suited for the derivatization of electrodes because it yields thick layers, causing an insulation of the electrodes. The final method used was an aqueous method 731

Isabella Moser et al.

Biosensors & Bioelectronics

Scheme 1. Reaction scheme of DNA immobilization via the amino groups of the bases. (a) Mercapto-silane coating of a platinum surface. (b) Introduction of carboxylic groups by iodoacetic acid. (c) Immobilization of DNA or proteins. The amino groups of the bases of DNA or that of amino acids such as lysine in protein react with the CDI-activated carboxylic groups of the spacer introduced according to (b).

with addition of acetone to an amount where a phase separation could be avoided.

Activation by iodoacetic acid and CDI activation

Mercapto-Silane coating

Mercapto-silane-coated electrodes were incubated with iodoacetic acid (2·8%, in 0·3 M phosphate buffer pH 8, addition of 1 N KOH until pH 8·1 is reached) for 30 min at room temperature in the dark. Sensors were rinsed with water and activated with water-soluble carbodiimides, e.g. by incubating the sensors in an aqueous solution of N-cyclohexyl-N′-[2-(N-methylmorpholino)ethyl]-carbodiimide-4-toluene sulfonate (CDI) (60 mg/ml) for 90 min at room temperature. DNA

A mixture of 4 ml acetate buffer (0·1 M, pH 4·0) with 4·5 ml acetone was prepared. Then 0·5 ml of mercapto-silane was added. The derivatization was performed at 56°C for 30 min. At this point it should be ensured that the whole area of the sensor except the bonding pads is submerged. After rinsing with water and with acetone the sensors were post-baked at 110°C for 10 min. 732

Biosensors & Bioelectronics

Surface techniques for an electrochemical DNA biosensor

or proteins were immobilized to this activated surface. To prove the immobilization technique glucose and lactate biosensors were prepared according to the immobilization procedure described. To the CDI-activated electrodes, glucose oxidase (aqueous solution 10 mg/ml) or lactate oxidase (1 mg in 1 ml phosphate buffer pH 7) were added and incubated at 4°C for 1 h. After an electrochemical cleaning procedure, which is discussed later on, these biosensors were an excellent tool for investigating the electrochemical properties of platinum because of their ability for in situ generation of hydrogen peroxide which was oxidized at the electrode surface at 500 mV.

decreased. It is assumed that the anodic electrode process starting at approximately 600 mV is caused by the oxidation of Pt-S bounds (Krauskopf & Wieckowski, 1992). The second interesting feature is the ease of desorption of the sulfur from the surface—even one cycle between − 600 and 1300 mV is sufficient to obtain unmodified platinum cyclovoltammogram in the second cycle. In the case of glucose sensors the electrochemical cleaning yielded glucose sensors with a glucose response to 4·5 mM glucose of 1·5 ␮A/cm2 and a lactate response of 1·2 ␮A/cm2 for a lactate concentration of 0·7 mM.

Electrochemical properties of mercaptosilane-coated platinum

Iodide treatment of platinum

No detectable anodic current could be observed at a freshly prepared mercapto-silane-coated platinum electrode, which was immersed in an aequeous solution of hydrogen peroxide (10 mmol/l phosphate buffer) and biased at 500 mV. Dong et al. (1995) described the same observation after the treatment of a Pt-electrode with long-chain alkane thiol. A cylovoltammogram with between − 600 and 1300 mV revealed deviations from that of an unmodified platinum electrode (Fig. 1). It is conspicuous that no Pt-O generation could be observed and that the reduction of dioxygen

The second step in the course of immobilization is the modification of sulfhydryl groups on the mercapto-silane by iodoacetic acid with the generation of the leaving group iodide. Immersion of Pt in an aqueous solution of iodide causes an ordered adlayer of iodine atoms (Bommarito et al., 1990). To investigate the pure effect of iodide on electrochemically cleaned (CV − 600/1300 mV) mercapto-silane-coated platinum, the electrodes were submerged in a solution of sodium iodide in phosphate buffer for 60 s. Glucose biosensors were kept under the same conditions. In both cases the hydrogen peroxide response of the electrodes decreased ⬎ 90% of that of the untreated electrode. After applying a voltage of 1000 mV for 30 s followed by a second voltage pulse of − 500 mV the hydrogen peroxide response reached the values obtained before treatment. The same experiment was also conducted with several other chemical compounds. The mercaptosilane-coated electrodes showed a significant decrease in electrochemical response ( ⬎ 90%) when treated with solutions of iodine, iodide, thiocyanate, sulfide and thiols in trace amounts. No effect was observed with bromide, chloride, nitrate or chlorate. Voltage pulsing (1 min at − 500 and 1200 mV) and cycling in phosphate buffer pH 7·0 up to 1200 mV (for iodine) and up to 1300 mV [Ag/AgCl] (for sulfur-containing compounds) enables the platinum biosensor surface to be cleaned without destroying the layers on top.

Fig. 1. Cyclovoltammogram of a mercapto-silane-modified platinum surface. Inset is shown the cyclovoltammogram of the same platinum electrode before the derivation. The observed current of the unmodified electrode was 100 times higher than the current of the mercapto-silane-coated electrode.


Isabella Moser et al.

Biosensors & Bioelectronics

* 1000 = cpm / Sensor

15 14 13 12 11 10 9 8

Sensor 2

7 6 5 4 H20 H20 1/10 SSC 37°C

3 Sensor 1


Sensor 2 Sensor 1

1 0 NH2 - Silane

SH - Silane

1/10 SSC 37°C

S-CH2-COH - Silane

Sensor - Coating

Fig. 2. Adsorption of radiolabeled DNA on bare platinum (—), amino-silane (NH2-Silane), mercapto-silane (SHSilane), and mercapto-silane further derivatized with iodoacetic acid (S-CH2-COOH-Silane) at different washing steps.

cpm/ sensor

7000 6000 5000 4000 3000 2000

Sensor 5 Sensor 4 Sensor 3 Sensor 2 Sensor 1

1000 0 SH




Fig. 3. Comparison of adsorption and immobilization of radiolabeled 20-mer DNA. Only electrodes with CDImodified carboxyl groups bind DNA via the amino groups of the bases. (SH...mercapto-silane coating, S-CH2-COAsp...mercapto-silane coating derivatized with iodoacetic acid and subsequently binding of asparaginic acid to the CDI-activated carboxyl groups at the surface, S-CH2-CO-DNA...immobilization of DNA via the amino groups of the bases of the DNA to the CDI-activated carboxyl groups at the surface, S-ClA-(DNA)...chloranil activated mercapto-silane (Mann-Buxbaum et al., 1990)). 734

Biosensors & Bioelectronics

DNA adsorption on mercapto-silane- and amino-silane-treated platinum To compare unspecific adsorption of the oligonucleotide at different coatings the electrodes were incubated in 150 ␮l DNA solution for 30 min at room temperature. The radioactivity per vessel was 115 000 cpm. For detecting the radioactivity of the sensors a Geiger counter with a geometry of about 150° was used. Four different types of electrodes were tested: oxidized platinum without coatings, amino-silanecoated platinum, mercapto-silane-coated platinum and mercapto-silane-coated platinum, further derivatized by iodoacetic acid, CDI, and asparaginic acid. The sensors were washed twice with water and

Surface techniques for an electrochemical DNA biosensor

then by immersion in 0·1 × SSC at 37°C for 10 min. This procedure was repeated three times. Fig. 2 shows adsorption of DNA on the sensors and the effect of the washing steps. As expected, the amino-silane layers caused the highest nonspecific binding of DNA. The efficacy of the above-described binding of the DNA base amino groups to the support can be seen in Fig. 3. The counts per minute on the sensors are about 6000 (SD = 9·6%, n = 5), while the radioactivity of comparably treated sensors without binding functions was very low (60 cpm). Site-Specific immobilization of the oligonucleotide It is obvious that the kinetics of a hybridization in solution (homogenous hybridization) differ

Scheme 2. Site-specific immobilization of DNA at the 5′ end. The terminal 5′ phosphate group of the oligonucleotide is bound to an amino group on the support. The amino groups were introduced by coupling diaminohexane to carboxylic acids at the support after activation of the mercapto-silane with iodoacetic acid. 735

Isabella Moser et al.

Biosensors & Bioelectronics

from the hybridization of immobilized DNA with mobile DNA (heterogeneous hybridization). The accessibility of the DNA depends on the method of immobilization. The well-known non-specific binding of ss-DNA on nitrocellulose in the course of a Southern blotting procedure yields an accessibility of 40% (Bu¨nemann et al., 1982). By way of contrast the accessibility of the immobilized DNA is optimal for site-specific binding at one of the ends of the DNA, providing tethered strands reaching into the solution. Binding of the 5′ terminus of the oligonucleotide A selective immobilization technique according to Ghosh (Ghosh & Musso, 1987) was used to bind the terminal 5′ phosphate group of the oligonucleotide to an amino group on the sensor surface (Scheme 2). For this purpose the amino groups were introduced by CDI coupling of diaminohexane to carboxylic acid groups at the support. The carboxylic acid groups resulted from an activation of the mercapto-silane with iodoacetic acid, which was performed according to procedures described above. To prove the effect on immobilization efficacy by altering the amino sites on the surface the diaminohexane had been mixed with different amounts of ethanolamine.

The CDI-activated electrodes were incubated with 1,6-diaminohexane/ethanolamine solutions (pH 6·0) at 25°C for 1 h. The following mixtures of diaminohexane with ethanolamine were used to vary the amount of amino groups on the support: (A) 2 M diaminohexane + 0·2 M ethanolamine; (B) 0·2 M for each; (C) 0·2 M diaminohexane + 2 M ethanolamine. The solutions were adjusted to pH 6·0 with HCl. Then the DNA was immobilized at 25°C for 12 h by incubating the electrodes in a mixture containing 0·1 M imidazol (pH 6·0 adjusted with HCl), 0·1 M water soluble carbodiimide and the oligonucleotide (50 pM/ml). The adsorbed DNA could be removed by washing with 0·1 × SSC, 2% SDS at 56°C. As can be seen from Fig. 4, the most homogenous result could be achieved with a high amount of ethanolamine in the coupling solution (method (C) SD = 14·4%; method (B) SD = 31·9%; n = 4 for both experiments). One observed a low unspecific adsorption of DNA on electrode surfaces created by treatment with diaminohexane solution and subsequently according to the protocol for specific binding of the oligonucleotide, except that the CDI was omitted from the coupling solution. Most of the


1200 1000 800 600 400

Sensor 4 Sensor 3


Sensor 2 Sensor 1

0 A




Fig. 4. Binding of the 5′ terminus of DNA to electrode surfaces with increased amino-group content. The ratio of diaminohexane:ethanolamine is (A) 10:1, (B) 1:1, (C) 1:10. For DNA adsorption all steps were performed but CDI was omitted. 736

Biosensors & Bioelectronics

labeled DNA could be removed by washing with 0·1 × SSC with 2% SDS at 56°C. For method (C) a background of 10% remained (see Fig. 4, Row ‘Adsorp.’).

CONCLUSION As a result of this study it can be concluded that the surface of platinum electrodes was modified by a technique, which provides binding sites for covalent site-specific immobilization of DNA. Also, the layers generated during the course of immobilization prevent unspecific immobilization caused by adsorption. The immobilization technique is also convenient for other negative charged biomolecules. The second field of interest was the effect on the electrochemical characteristics of thin film platinum electrodes caused by chemisorption of substances which were brought into contact with the metal during the course of immobilization. It could be shown that electrochemical treatment was very effective in restoring the electrocatalytical properties of the electrodes.

REFERENCES Aschauer, E., Fasching, R., Urban, G., Nicolussi, G. and Husinsky, W. (1995) Surface characterization of thin-film platinum electrodes for biosensors by means of cyclic voltammetry and laser SNMS. Electroanal. Chem. 381, 143–150. Bommarito, G. M., White, J. H. and Abrun˜a, H. D. (1990) Electrosorption of iodide on platinum: packing density and potential-dependent distributional changes observed in situ with x-ray standing waves. J. Phys. Chem. 94, 8280–8288. Bu¨nemann, H., Westhoff, P. and Herrmann, R. G. (1982) Nucleic Acids Res. 10, 7163. Dong, X. D., Lu, J. and Cha, C. (1995) Study of glucose oxidase on Pt electrode modified by alkane thiol. Bioelectrochem. Bioenerg. 36, 73–76. Ghosh, S. S. and Musso, G. F. (1987) Nucleic Acids Res. 15, 5353.

Surface techniques for an electrochemical DNA biosensor Hall, J. M., Moore-Smith, J., Bannister, J. V. and Higgins, I. J. (1994) An electrochemical method for detection of nucleic acid hybridization. Biochem. Molec. Biol. Int. 32, 21–28. Krauskopf, E. K. & Wieckowski, A. (1992). Radiochemical methods to measure adsorption at smooth polycristalline and single crystal surfaces. In Adsorption of Molecules at Metal Electrodes, ed. J. Lipkowski & P. N. Ross. VCH Publishers Inc., N.Y., pp. 143–150. Mann-Buxbaum, E., Pittner, F., Schalkhammer, T., Jachimowicz, A., Jobst, G., Olcaytug, F. and Urban, G. (1990) New microminiaturized glucose sensors using covalent immobilization techniques. Sensors & Actuators B1, 518–522. Millan, K. M., Saraullo, A. and Mikkelsen, S. R. (1994) Voltammetric DNA biosensor for cystic fibrosis based on a modified carbon paste electrode. Anal. Chem. 66, 2943–2948. Moser, I., Jobst, G., Aschauer, E., Svasek, P., Varahram, M., Urban, G., Zanin, V., Tjoutrina, G., Zharikova, A. and Berezov, T. (1995) Miniaturised thin film glutamate and glutamine biosensors. Biosensors & Bioelectronics 10, 527–532. Mullis, K. and Faloona, F. A. (1987) Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Meth. Enzymol. 155, 335–350. Nikiforov, T. T. and Rogers, Y. H. (1995) The use of 96-well polystyrene plates for DNA hybridizationbased assays: an evaluation of different approaches to oligonucleotide immobilization. Anal. Biochem. 227, 201–209. Nilsson, P., Nygren, P-Å., Persson, B. and Uhle´n, M. (1995) Molecular biology techniques monitored in real time with BIA. Analyt. Biochem. 224, 400– 408. Schwarz, T., Yeung, D., Hawkins, E., Heany, P. and McDougall, A. (1991) Detection of nucleic acid hydridization using surface plasmon resonance. TIBTECH 9, 339–340. Wang, J., Palecek, E., Nielsen, P. E., Rivas, G., Cai, X., Shiraishi, H., Dontha, N., Luo, D. and Farias, P. A. M. (1996) Petide nucleic acid probes for sequence-specific DNA biosensors. J. Am. Chem. Soc. 118, 7667–7670. Weetall, H. H. (1976). Covalent coupling methods for inorganic support materials. In Methods in Enzymology, ed. K. Mosbach. Academic Press, London, Vol. 44, p. 134.