Novel disposable biochip platform employing supercritical angle fluorescence for enhanced fluorescence collection
- First Online:
- Cite this article as:
- Hill, D., McDonnell, B., Hearty, S. et al. Biomed Microdevices (2011) 13: 759. doi:10.1007/s10544-011-9546-2
- 181 Views
This paper presents an overview of development of a novel disposable plastic biochip for multiplexed clinical diagnostic applications. The disposable biochip is manufactured using a low-cost, rapid turn- around injection moulding process and consists of nine parabolic elements on a planar substrate. The optical elements are based on supercritical angle fluorescence (SAF) which provides substantial enhancement of the fluorescence collection efficiency but also confines the fluorescence detection volume strictly to the immediate proximity of the biochip surface, thereby having the potential to discriminate against background fluorescence from the analyte solution. An optical reader is also described that enables interrogation and fluorescence collection from the nine optical elements on the chip. The sensitivity of the system was determined with a biotin-avidin assay while its clinical utility was demonstrated in an assay for C-reactive protein (CRP), an inflammation marker.
KeywordsLab on a chipSupercritical angle fluorescenceImmunoassayCardiac marker
There is a strong demand for the development of disposable biochips for point-of-care (POC) diagnostic assay platforms, for use in a broad range of settings. Such miniaturised and inexpensive biochips are designed to migrate the analysis from the hospital laboratory to the POC setting, where the complicated sequences of clinical tests can be performed quickly by relatively un-skilled personnel (Ligler 2009). Common transduction technologies employed in POC immunoassay platforms have traditionally been optically-, or electrochemically-mediated (Soper et al. 2006). Fluorescence-based systems are the most widely used due primarily to their superior sensitivity compared with alternative approaches. In such systems, fluorescent labels are employed to transduce the biomolecular binding event (Schult et al. 1999). In addition, fluorescence-based methodologies are ideally suited to parallel detection, often referred to as array or imaging detection (Cho and Bright 2002). The use of arrays enables the detection of multiple analytes in a single test. Such tests generally consist of multiple bio-recognition elements patterned onto a surface, which is typically planar or has planar elements to facilitate both patterning and detection. There is an increasing requirement to detect low levels of analyte in small volumes and this necessitates the use of high sensitivity detection techniques. Signal enhancement strategies that may be employed for optical biochips include plasmonic enhancement (Nooney et al. 2010; Stranik et al. 2007), high-brightness nanoparticles (Nooney et al., 2009) and the use of high collection efficiency optics.
Supercritical Angle Fluorescence (SAF) refers to the anisotropic emission of dipoles near a dielectric interface characterised by a preferential emission into angles above the critical angle in a substrate with higher refractive index than the superstrate (Lukosz and Kunz 1977; Enderlein et al. 1999; Ruckstuhl et al. 1999; Polerecky et al. 2002; Polerecky et al. 2000; Blue et al. 2005). Exploitation of this near-field optical effect enables surface-confined fluorescent species, bound or in close proximity (~ λ) to a substrate, to be collected efficiently by customised optics onto a detector. In particular, it has been demonstrated that paraboloid elements enable detection of fluorescence down to ultra-low concentrations (Verdes et al. 2007; Ruckstuhl et al. 2000). Typically a paraboloid element is built into an optical system, which, when optically-coupled to a planar substrate, enables efficient collection of surface-localised fluorescence. For example, efficient multiplexed assay measurements on plastic biochips via a scanning paraboloid format have been reported (Kurzbuch et al. 2009). The combination of the total internal reflection (TIR) excitation technique with a parabolic lens for SAF collection has been shown to give an attoliter detection volume and single molecule sensitivity (Ruckstuhl and Seeger 2004). It has also been shown that with only SAF collection using a parabolic element (and an aspheric lens below or an objective lens above focusing excitation light) one can achieve diffraction-limited high aperture collection of fluorescence down to the level of even a single molecule (Verdes et al. 2007; Ruckstuhl and Verdes 2004).
The work presented here is the first report of a novel embodiment for low-cost disposable plastic array biochips whereby the paraboloid elements are integrated into the biochips. In the particular example presented, the biochip comprises an array of nine paraboloid elements. The nine elements are designed for highly localised excitation (via TIRF) and highly efficient collection of light (SAF) from the planar interface. A simple optical reader employing efficient excitation and detection is also presented. The relative sensitivity of the system was determined using a dye-labelled biotin-neutravidin direct binding assay. As a proof of principle, the clinical utility of the biochip was demonstrated with an assay for the cardiac marker C-reactive protein (CRP), an indicator for inflammation (McDonnell et al. 2009).
1.1 Super-critical angle fluorescence (SAF)
2 Materials and methods
2.1 Paraboloid biochip
The chips were coated with a proprietary surface chemistry comprising an aldehyde-activated dextran surface coating for antibody immobilisation (Åmic AB, Uppsala, Sweden). The resultant aldehyde surface enabled immobilization of amino-functionalized molecules, such as proteins, onto the surface. Reaction vessels used in surface discrimination experiments were made from Poly(methylmethacrylate) (PMMA). Nine wells, 2 mm in diameter, were micro-milled into a 2 mm thick piece of PMMA. Each well was aligned to a matching paraboloid element. The PMMA was attached to the chip via 50 μm thick pressure-sensitive-adhesive (ARcare®8890, Adhesives Research, Ireland).
2.2 Biochip reader
2.3 Sensitivity assay
Commercial neutravidin (Pierce, #31000) was diluted in 50 mM sodium phosphate buffer, pH 7.5 containing 1% (w/v) milk powder (spotting solution) to a final concentration of 200 μg/mL and deposited onto each paraboloid element using a sciFLEXARRAYER™ piezo-dispensing system (Scienion, Germany). Following 1 h incubation in 70% humidity at room temperature, the spotted chips were blocked in a solution of phosphate buffered saline (PBS) containing 5% (w/v) milk powder overnight at 4°C. Excess blocking solution from the incubation was removed by washing with PBS. Atto 647 Biotin (Sigma, #30700), a fluorescently-labelled biotin conjugate, was then diluted in PBS containing 0.05% (v/v) tween20 and 0.5% (w/v) milk powder to concentrations ranging from 10 μM to 10 M. A 350 μL aliquot of each of the dilutions was applied to the chips and incubated for 1 h at room temperature. Next, any unbound biotin was removed by washing in PBS containing 0.05% (v/v) Tween20 (PBST) three times and then in PBS three times, followed by a wash with deionised water to remove any excess salt. The chips were then read in the custom-built biochip reader.
2.4 Anti-CRP polyclonal antibody production, purification and labelling
A New Zealand White rabbit was subjected to a comprehensive immunisation regimen with CRP as the antigen (AMS Biotechnogy, A97201H). Following several boosts, a significant response to the CRP antigen was confirmed via titration of the rabbit serum. The polyclonal antibody (pAb) was then purified from the rabbit serum using protein A affinity chromatography. A sample of the purified CRP-specific polyclonal antibody was fluorescently labelled using a commercially-available kit (DyLight 647 monoclonal antibody labelling kit, Pierce #53015).
2.5 CRP assay
Purified rabbit anti-CRP pAb (capture antibody) was diluted in spotting buffer to a final concentration of 200 μg/mL. The pAb solution was then accurately deposited as described above. Following a 1 h incubation in 70% humidity at room temperature, the spotted chips were blocked and washed as described above for the biotin capture assay. Next, varying concentrations of CRP, diluted in CRP-depleted serum (HyTest, 8CFS), were pipetted onto each of the washed chip surfaces (330 μL per chip). Following a 1 h incubation at room temperature, another wash step was performed comprising PBST (x3) and PBS (x3) to remove excess unbound CRP antigen. A 1/400 dilution of fluorescently labelled anti-CRP pAb in 1% (w/v) milk-PBST was then pipetted onto the chip surface and left to incubate for 1 h at room temperature. A final wash step was then performed as described earlier, followed by a wash with deionised water, to remove any excess salt. Finally, the washed chips read using the custom-built biochip reader.
3 Results and discussion
3.1 Chip and reader characterisation
A series of 10 aldehyde-dextran-coated biochips were spotted with Cy5-labelled human-IgG as described above. A 1 μg/mL concentration of Cy5-labelled IgG antibody was spotted into the centre of the paraboloids on all ten chips and then washed and blocked leaving only surface-immobilised Cy5-labelled antibodies on the surface. These chips were scanned in a conventional fluorescence scanner, GMS 418 ArrayScanner (Genetic MicroSystems) and the fluorescence spots quantified to ensure the uniformity of immobilised antibody on the top surface of each parabola. The standard deviation of fluorescence measured from each parabola across all ten chips was found to be within 2%, showing that uniform deposition of protein onto the tops of the paraboloids is possible.
A reference chip was made by spotting a 1 μg/mL of Cy5-labelled human IgG onto the tops of all parabolas. The SAF images captured from this reference chip show that the outer rings are distorted (Fig. 5(a)). This is due to spherical aberration caused by the achromatic doublet used for focusing the ring array down onto the CCD sensor. Consistent with POC diagnostics, low cost optical elements have been used in this system. This effect can be eliminated by using higher cost, larger numerical aperture lenses. The integrated fluorescent intensities of the outer rings are also lower than that of the centre ring. This is due to the Gaussian beam profile of the laser beam incident on the chip resulting in non-uniform excitation of the fluorescent dye on the surface of the differing paraboloids. In order to correct for this, all the grey-scale images were processed and analysed using Cell^f fluorescence imaging software (Olympus, UK). Each ring was isolated and the average fluorescence intensity was measured. Normalization coefficients for each paraboloid were determined from the differences in fluorescence, measured from each paraboloid element, with respect to the central ring, which was the brightest ring. These coefficients were applied to all subsequent measurements.
The background signal in each image has two principal contributory elements; a residual capacitance in the CCD and thermal noise in the CCD; the former is invariant to around 0.1% and the latter dependent upon integration time. To account for this, the signal from a blank aldehyde-dextran-coated chip was measured before each experimental run in order to provide a systematic background measurement. This average background signal was then subtracted from all measured data before further processing.
To demonstrate the enhanced collection efficiency of the paraboloid elements, signals from the central paraboloid of chips used in the biotin-avidin assay were measured in the optical reader with and without the annular masks in place. This allowed both direct illumination of selected paraboloid surfaces in addition to TIRF excitation but also enabled fluorescence emitted into the lower angles (non-SAF) pass through to the detector via the pair of achromatic lenses. Ray-tracing analysis (TracePro, Lambda Research Corp., USA) of the system showed that the non-SAF light was collected in a small spot at the centre of the SAF image rings. While all of the light from the SAF component was collected due to the paraboloid element, the non-SAF component remained divergent and, thus, only a portion of the non-SAF light was collected, equating to about 10% of the total non-SAF component. Figure 5(b) shows an example of the resultant images collected. Visually it is obvious that the SAF signal from the rings are much greater than the non-SAF signal coming from the central spot. The SAF and non-SAF components could be isolated and analysed as described above. The signal was integrated for each ring and spot, the normalisation coefficient applied and the background signal was subtracted from the final value. The analysis shows that the paraboloid elements were able to collect between 80 and 100 times more SAF fluorescence than the non-SAF fluorescence. This is in agreement with theory previously reported [Enderlein et al., 1999] which outlines the highly efficient light collection of SAF compared to non-SAF. High quality microscope objective lens with high numerical apertures (NA) could be used to capture more of the non-SAF but this leads to more complicated multiple lens systems with increasing optical losses due to multiple interfaces. This would not suit a POC application. The increase in collection efficiency also indicates that cheaper and less sensitive detectors could be used with this platform.
3.2 Sensitivity assay—biotin-avidin assay
3.3 Clinical utility—CRP assay
3.4 Discrimination from bulk fluorescence
We have described a low-cost novel biochip platform consisting of an array of integrated SAF collection elements. The demonstrated high collection efficiency and background discrimination, together with the simplicity, low cost and mass-producibility of the system, makes this biochip platform eminently suitable for POC diagnostic applications. The biochip shows good repeatability and ‘ease-of-use’ using a dedicated biochip reader, both of which allow for low-cost mass production. Each paraboloid can be interrogated individually thereby facilitating multi-analyte sensing when appropriate microfluidics are implemented. The ‘generic sensitivity’ of the system was characterised using a high affinity biotin-neutravidin assay and LOD and LOQ values were found to be 4.52 × 1010 and 6.3 × 1010 dye molecules/mL, respectively. The enhancement of fluorescence capture in a biotin-avidin assay is of the order of 80-100-fold. The system design allows for collection of fluorescence from the surface of the chip while discriminating against fluorescence from the bulk thereby negating the requirement for a wash step. The biochip was validated by measuring the inflammatory marker CRP over the clinical range in serum, giving a LOD of 7.6 ng/mL. Low cost fabrication, dedicated optical interrogation instrumentation and an antibody-compatible surface, allied with the clear amenability for incorporation of a microfluidic sample delivery component, indicate that this device offers a very real alternative to traditional diagnostic assay platforms and has significant potential for application in a POC setting.
This material is based upon works supported by the Science Foundation Ireland under Grant No. 05/CE3/B754