Optimization of antibody immobilization for sensing using piezoelectrically excited-millimeter-sized cantilever (PEMC) sensors

https://doi.org/10.1016/j.snb.2006.09.034Get rights and content

Abstract

The effectiveness of antibody immobilization on sensor performance was evaluated using a piezoelectric-excited millimeter-sized cantilever (PEMC) sensor and a model protein, bovine serum albumin (BSA). The immobilization parameters—antibody and activation reagent concentration and reaction time were systematically varied and the resultant sensor response to 1 pg/mL BSA was measured.

The highest frequency shift due to BSA attachment (1931 ± 60 Hz) was obtained when 100 μg/mL antibody solution was activated with EDC (2 mM) and sulfo-NHS (5 mM) for 30 min prior to reaction with primary amine on the sensor surface. Increasing activation time from 30 to 60 min resulted in an 8% decrease in sensor response, while activation at 0.2 mM EDC and 0.5 mM NHS resulted in a 65% decrease. The logarithmic correlation between antibody concentration and sensor response suggests that a lower antibody concentration of 10 μg/mL is sufficient for BSA at 1 pg/mL. With 10 μg/mL antibody, the PEMC sensor response was 725 ± 50 Hz (n = 2) for 1 pg/mL BSA with a signal to noise ratio of 41.

Introduction

Biosensor-based immunoassays have been developed for various applications including medical diagnostics, detection of bio-threat agents, environmental monitoring and pathogen and toxin detection in food samples. The immuno-sensor function is based on selective binding of antigens to antibodies by molecular recognition. The antibody is immobilized on the sensor enabling transduction of antigen binding into electrical, mechanical, optical, magnetic or other signal generation.

Previous work has examined the reaction chemistry used to immobilize antibodies to glass sensor surfaces [1], [2], [3], [4]. The primary focus of these reports has been on optimizing the antibody reaction to the functionalized surface which was evaluated by measuring the amount of antibody immobilized [5], [6], [7]. While a sensor's performance is strongly dependent on antibody surface density and its availability to react with the antigen of interest, the amount of antibody that results on the sensor may not directly translate into sensor performance. This is because immobilization can cause some of the antibody to lose binding activity due to one of the following three factors: (a) direct chemical modification of the antigen-binding site, (b) steric hindrance by the surface itself and (c) steric hindrance by adjacently immobilized antibodies [8]. Few studies address biosensor performance as a function of reaction conditions used to prepare the antibody-immobilized surface. Therefore, we report here optimizing the sensing performance of immobilized antibody on a cantilever biosensor.

Recently, we showed that piezoelectric-excited millimeter-sized cantilever (PEMC) sensors, comprised of a piezoelectric layer acting as an actuating and a sensing element, and a borosilicate glass surface for antibody attachment, provided a sensitive platform for detection of a model protein [9] and pathogens [10]. PEMC sensors respond to changes in mass that occur due to binding of the target molecules to the sensor surface. When the antigen binds to the antibody immobilized on the sensor surface, the effective mass of the cantilever increases which alters the cantilever's resonant frequency. Monitoring the resonant frequency change with time provides quantitative measures of the target antigen in the sample.

In the above applications, we immobilized antibody covalently to surface amine groups on the glass surface. In these studies, we demonstrate the detection of large antigens while keeping the procedure for antibody immobilization constant. To evaluate the effectiveness of antibody immobilization methods on sensor performance, in this study, we use a 66 kDa model protein, bovine serum albumin, as the target antigen. There is limited work demonstrating the impact of antibody immobilization on a sensor's ability to detect proteins in solution [1], [5]. In this short note, we investigate the effects of systematically varying three parameters during surface functionalization and antibody activation, to determine the change in sensor response when it is subjected to a solution of target antigen.

Section snippets

Antibody immobilization

To immobilize an antibody to the surface, two essential preparations are required. First, the glass surface must be functionalized so that covalent bonding with the antibody can be facilitated. Secondly, the functional group on the antibody must be activated to expedite the reaction with a surface functional group.

In order to functionalize the glass surface prior to reaction with an antibody, several techniques may be employed. While many methods are available for immobilizing antibody to

PEMC sensor

Details of PEMC fabrication and its response characteristics have been previously reported [9]. The cantilever tips were designed with the PZT layer, 4 mm × 1 mm (length × width), bonded to the glass layer, 2 mm × 1 mm (length × width), such that the 2 mm of glass is available as a surface for antibody immobilization and antigen detection. Several cantilevers were fabricated and used in the various experiments reported in this paper. Since the cantilevers are made individually and manually, no two have

Immobilization procedure

The procedural steps used to functionalize the cantilever sensor surface and activate the target antibody are schematically represented in Fig. 1. The glass surface was sequentially cleaned with methanol–hydrochloric acid solution (1:1, v/v), concentrated sulfuric acid, hot sodium hydroxide and finally boiling water (Fig. 1, Step I). After cleaning, the glass surface was silanylated with 3-aminopropyl-triethoxysilane (APTES) in deionized water at pH 3.0 (adjusted by hydrochloric acid, 0.1N) and

Results and discussion

Approximately 20 cantilever sensors were fabricated and used in the a-BSA/BSA detection experiments. Of these, 15 had resonance characteristics in air between 930 and 990 kHz and Q values from 20 to 35 indicating a high degree of reproducibility (within ∼6%) from sensor to sensor. The sensors used in the experimental plan were chosen from the subset of fifteen exhibiting similar characteristics. For brevity and quantitative performance comparison of the various surface preparations, the data

Conclusion

It was shown that an antibody concentration of 100 μg/mL activated with EDC (2 mM) and sulfo-NHS (5 mM) for 30 min gave the largest sensor response. The logarithmic correlation between antibody concentration and sensor response suggests that a lower concentration is sufficient to detect BSA at 1 pg/mL. Immobilization with 10 μg/mL antibody gave a response for 1 pg/mL BSA with a signal to noise ratio of 15.

David Maraldo received his BS in chemical engineering from Widener University in 1999 and his MS in chemical engineering from Drexel University in 2003. He is currently working on his PhD degree in chemical engineering at Drexel University.

Cited by (28)

  • Highly sensitive and selective electrochemical cortisol sensor using bifunctional protein interlayer-modified graphene electrodes

    2017, Sensors and Actuators, B: Chemical
    Citation Excerpt :

    Protein aggregates were observed on the d-BSA/rGO surface if higher concentrations of the antibody were used. It should be noted that the use of excessive amount of protein is detrimental to the passivation of active domain-limited surface [33]. On the other hand, the incubation with anti-cortisol antibody at 1 μg/mL achieved insufficient surface coverage.

  • Optimization and characterization of a homogeneous carboxylic surface functionalization for silicon-based biosensing

    2016, Colloids and Surfaces B: Biointerfaces
    Citation Excerpt :

    The outcomes highlight that the mean number of Ab-HRP grafted to a non-activated surface is around 1 × 1011 molecules/cm2. This amount increases up to 2.34 ± 0.26 × 1011, 3.04 ± 0.20 × 1011 and 3.30 ± 0.23 × 1011 molecules/cm2 if the concentrations of EDC/NHS used are 2/5 mM (typically suggested for bioconjugation by suppliers and researchers [19]), 4/10 mM or 260/23 mM, respectively. The highest value of grafted biomolecules was attained with the highest concentration of crosslinker reactants (260/23 mM), also confirmed by other authors for silicon and silica functionalization [16].

  • Graphene Functionalization for Biosensor Applications

    2016, Silicon Carbide Biotechnology: A Biocompatible Semiconductor for Advanced Biomedical Devices and Applications: Second Edition
  • Current aspects in immunosensors

    2014, Biosensors and Bioelectronics
View all citing articles on Scopus

David Maraldo received his BS in chemical engineering from Widener University in 1999 and his MS in chemical engineering from Drexel University in 2003. He is currently working on his PhD degree in chemical engineering at Drexel University.

Raj Mutharasan received his BS degree in chemical engineering from IIT Madras (India) and a PhD in chemical engineering from Drexel University. He has been at Drexel University since 1974 where he is the Frank A. Fletcher professor of chemical and biological engineering. His research interests are in biosensors (cantilever, fiber optic, magnetoelastic), and process biotechnology. He has published extensively and has several patents. He is a fellow of American Institute of Chemical Engineers.

View full text