Elsevier

Biosensors and Bioelectronics

Volume 24, Issue 12, 15 August 2009, Pages 3645-3651
Biosensors and Bioelectronics

Immobilization of bacteriophages on gold surfaces for the specific capture of pathogens

https://doi.org/10.1016/j.bios.2009.05.028Get rights and content

Abstract

Techniques for the chemical attachment of wild-type bacteriophages onto gold surfaces and the subsequent capture of their host bacteria have been developed. The surfaces were modified with sugars (dextrose and sucrose) as well as amino acids (histidine and cysteine) to facilitate such attachment. Non-specific attachment was prevented by using bovine serum albumin as blocking layer. Surfaces modified with cysteine (and cysteamine) followed by activation using 2% gluteraldehyde resulted in an attachment density of 18 ± 0.15 phages/μm2. This represented a 37-fold improvement compared to simply applying physisorption. Subsequently, the phage immobilized surfaces were exposed to the host E. coli EC12 bacteria and capture was confirmed by fluorescence microscopy. We obtained a bacterial capture density of 11.9 ± 0.2/100 μm2, a 9-fold improvement when compared to those on physically adsorbed phages. The specificity of recognition was confirmed by exposing similar surfaces to three strains of non-host bacteria. These negative control experiments do not show any bacterial capture. In addition, no capture of the host was observed in the absence of the phages.

Introduction

Detection of pathogenic bacteria has been an area of prime interest in the field of food and water safety, public health and bio-terrorism. Conventional microbiological techniques take several days in order to culture small loads of bacteria from a sample to a detectable number. In addition, identifying the specific signature of a bacterium requires further biochemical and serological tests, which are costly, time-consuming and labour intensive. Thus, there has been a sustained interest in the development of bacteria sensing platforms that would overcome these limitations.

Polymerase chain reaction (PCR) has been used for the amplification of small quantities of bacterial DNA obtained from the samples, allowing its subsequent detection using established DNA analysis techniques (Steffan and Atlas, 1988, Olive, 1989). Conventional PCR detects DNA from both live as well as dead cells while RT-PCR is used for mRNA detection. Enzyme-linked immunosorbent assay (ELISA) has also been explored. Binding of antigen (pathogens or toxins) to primary antibody is measured in microtiter plate by using a secondary antibody conjugated to an enzyme. Here, bacterial-antigen specific antibodies are used for the detection of a specific bacterium in a sample (Janyapoon et al., 2000). This technique offers the advantage of allowing the detection of pathogens and their toxins (Downes et al., 1989, Basta et al., 1989). Efforts have also been made to combine PCR and ELISA in order to improve their respective performance. Polymerase chain reaction-enzyme-linked immunosorbent assay (PCR-ELISA) has successfully been shown to detect bacteria (Jones et al., 1998, Hong et al., 2003) and their toxins (Baez et al., 1996, Gilligan et al., 2000). However, all these approaches still require pre-enrichment and selective enrichment steps since the bacterial load in samples is usually smaller than detection limits which ranges from 106 to 107 bacterial cells/ml. In ELISA, the problem is further compounded by the low sample volume held in the microtiter plates which requires concentration/enrichment of the sample prior to analysis (Wyatt, 1995). Thus, these methods are still time-consuming, labour intensive and require skill for successful execution. They also suffer from problems associated with DNA extraction, fidelity of DNA replication, enzyme inhibition, antibody stability and specificity.

Biosensing platforms have received increased attention as alternative method for bacterial detection (Ivnitski et al., 1999, Lazcka et al., 2007). Such platforms usually consist of three components: a biological recognition mechanism, a physical transduction platform, and a system to read the transduced signal. The transduction phenomenon can be optical, magnetic, thermoelectric, piezoelectric, electrochemical or mechanical in nature. A wide range of techniques such as quartz crystal microbalance (QCM) (Minunni et al., 1996, Fung and Wong, 2001), surface plasmon resonance (SPR) (Taylor et al., 2005, Oh et al., 2005), flow cytometry (Abdel-Hamid et al., 1998, Abdel-Hamid et al., 1999), amperometry (Gau et al., 2001), and micromechanical resonators (Ilic et al., 2001, Ilic et al., 2004) have been extensively researched. Microcantilevers have specifically received increased attention due to their high mass sensitivities. These devices have been used in the deflection as well as resonance modes for the detection of individual bacteria (Ilic et al., 2001), bacterial toxins (Liu et al., 2003), antigens (Wu et al., 2001a), DNA (Wu et al., 2001b), carbohydrates (Pei et al., 2004), proteins (Savran et al., 2004, Wee et al., 2005) and viruses (Gupta et al., 2004). The specificity of the recognition in such systems is usually provided by a biological probe such as a nucleic acid, an antibody, an enzyme, a cell or an artificial receptor. Antibodies, especially monoclonal antibodies, have been particularly exploited as they offer a fair degree of specificity and selectivity towards their target. The production, isolation and purification of an antibody is however time-consuming, requires high skill and is not particularly cost-effective. Polyclonal antibodies are cheaper and easier to produce but suffer from cross-reactivity and interference problems. Thus, an alternative, robust, inexpensive and specific probing system would be beneficial for such application.

Virulent bacteriophages (or phages) are viruses that bind to specific receptors on the bacterial surface in order to inject their genetic material inside the bacteria. These entities are typically of 20–200 nm in size. The injection of the phage nucleic acid into the bacterial cells allows the phages to propagate inside the host using the host's own replication machinery. The replicated virions are eventually released, killing the bacterium and allowing the infection of more host cells. Phages recognize the bacterial receptors through its tail spike proteins (Kutter and Sulakvelidze, 2004). This recognition is highly specific and has been employed for the typing of bacteria. However, this level of specificity and selectivity also opens avenues for the development of specific pathogen detection technologies. Such use of phages for the creation of pathogenic sensing platforms has been reported by using quartz crystal microbalance (QCM), flow cytometry, complementary metal-oxide semiconductor (CMOS) and surface plasmon resonance (SPR) as transduction platform (Balasubramanian et al., 2007, Edgar et al., 2006, Nanduri et al., 2007, Olsen et al., 2006, Yao et al., 2008). These early reports have mainly relied on physical adsorption for the attachment of the phages on sensor surface. This approach results in poor phage surface coverage, severely inhibiting the sensitivity of the platform. Chemical attachment of the phages onto surface would yield better coverage and thus significantly improve the performance of these sensors.

We recently reported the streptavidin-mediated attachment of bacteriophages that were genetically modified to directly express biotin on their capsid (Gervais et al., 2007). This approach yielded attachment densities of 4.4 phages/μm2, a 15-fold increase over simple physisorption. However, the need for genetic modification of a given phage significantly increases the cost and efforts associated with the development of a sensor system designed to detect a given pathogen. Thus, a simpler attachment approach that would rather leverage the natural proteins of unmodified phages would be desired.

We here report the development of a general procedure allowing the efficient chemical attachment of wild-type bacteriophages onto gold surfaces. While the T4 bacteriophage has been used as a model system, this procedure would be directly applicable to phages associated with other pathogens, including Salmonella, Campylobacter, Listeria, etc. The T4 phages are typically are 200 nm long and 90 nm wide, and perform lytic cycles inside the host bacterium E. coli. In this work, gold surfaces were first modified with different chemical agents such as sugars and amino acids to provide different terminal functional groups and enable the chemical anchoring of the phage through hydrogen bonds. The sugars are believed to get adsorbed on the gold surface through hydrophilic interaction or their carbonyl oxygen which is known to interact with Au (0) (Shafai et al., 2007). The histidine is immobilized through weak binding of amine group to gold surface while the cysteine/cysteamine attach through strong thiol linkage. In addition, gold surfaces modified with amino acids were further activated using gluteraldehyde (Schiff's reaction). Gluteraldehyde is indeed known to efficiently capture proteins and has been exploited extensively for immobilization of antibodies onto a surface (Nugaeva et al., 2007). Such gluteraldehyde activation significantly improved the density of phages on the amino acid modified surface. Attachment densities as high as 18 ± 0.15 phages/μm2 have been obtained in that case, a 4-fold improvement over the values obtained with biotinylated phages, and a 37-fold improvement over simple physisorption. These surfaces were then exposed to the host E. coli EC12 to assess their bacterial capture efficiency. Non-host E. coli strains such as 6M1N1, NP10 and NP30 were also used as negative controls. Bovine serum albumin (BSA) was used as a surface blocking layer to inhibit non-specific binding. Bacterial capture was quantified using fluorescence microscopy. Thus, this work presents a generalized approach to immobilize bacteriophages covalently onto sensor substrates for specific capture of host bacterial strains. This approach can be easily employed into numerous other biosensing platforms such as microresonators, surface plasmon resonance, amperometric sensors or quartz crystal microbalance.

Section snippets

Materials

l-Histidine, l-cysteine, cysteamine hydrochloride, gluteraldehyde and bovine serum albumin (BSA) were purchased from Sigma–Aldrich (St. Louis, MO, USA). Sucrose and dextrose were purchased from Merck & Co., Inc. (NJ, USA). E. coli were fluorescently labelled using the SYTO 12 BC bacteria stain purchased from Invitrogen (Carlsbad, CA, USA). Tween-20 was bought from MP Biomedicals, Inc. (OH, USA). E. coli bacterial strains and the wild-type T4 bacteriophage were obtained from University of Guelph

Phage immobilization using sugars as binding agents

Fig. 1(a–c) shows micrographs of gold surfaces with phages immobilized on them after dextrose and sucrose modification. The immobilization was performed at room temperature to avoid crystallization of the sugars. The bare gold surface showed a phage density of 0.49 ± 0.15 phage particles per μm−2 (Fig. 1a). In turn, the densities obtained on the dextrose- and sucrose-covered surfaces were 2.4 ± 1.3 μm−2 and 3.67 ± 1.1 μm−2, respectively, a 5–7-fold increase over physical absorption (Fig. 1b and c).

Summary

Fig. 6 shows a histogram of the phage densities on the gold surfaces using the different surface modification approaches explored here. A significant increase in the phage density was achieved when compared to a bare gold surface. The gold substrates functionalized by cysteine or cysteamine followed by gluteraldehyde activation yielded the best results. The phage density obtained was 37 times better than that achieved by physical adsorption, and 4 times better than that obtained by chemical

References (41)

  • I. Abdel-Hamid et al.

    Sens. Actuators B

    (1998)
  • I. Abdel-Hamid et al.

    Biosens. Bioelectron.

    (1999)
  • L.A. Baez et al.

    Int. J. Food Microbiol.

    (1996)
  • S. Balasubramanian et al.

    Biosens. Bioelectron.

    (2007)
  • J.J. Gau et al.

    Biosens. Bioelectron.

    (2001)
  • L. Gervais et al.

    Sens. Actuators B

    (2007)
  • K. Gilligan et al.

    Mol. Cell. Probes

    (2000)
  • D. Ivnitski et al.

    Biosens. Bioelectron.

    (1999)
  • O. Lazcka et al.

    Biosens. Bioelectron.

    (2007)
  • M. Minunni et al.

    Anal. Chim. Acta

    (1996)
  • V. Nanduri et al.

    Biosens. Bioelectron.

    (2007)
  • B.K. Oh et al.

    Biosens. Bioelectron.

    (2005)
  • E.V. Olsen et al.

    Biosens. Bioelectron.

    (2006)
  • A.D. Taylor et al.

    Sens. Actuators B

    (2005)
  • D.R. Walt et al.

    Trends Anal. Chem.

    (1994)
  • K.W. Wee et al.

    Biosens. Bioelectron.

    (2005)
  • M. Basta et al.

    J. Clin. Microbiol.

    (1989)
  • F.P. Downes et al.

    J. Clin. Microbiol.

    (1989)
  • R. Edgar et al.

    Proc. Natl. Acad. Sci. U.S.A.

    (2006)
  • Y.S. Fung et al.

    Anal. Chem.

    (2001)
  • Cited by (110)

    View all citing articles on Scopus
    View full text