Integrated imaging instrument for self-calibrated fluorescence protein microarrays

Abstract

Protein microarrays, or multiplexed and high-throughput assays, monitor multiple protein binding events to facilitate the understanding of disease progression and cell physiology. Fluorescence imaging is a popular method to detect proteins captured by immobilized probes with high sensitivity and specificity. Reliability of fluorescence assays depends on achieving minimal inter- and intra-assay probe immobilization variation, an ongoing challenge for protein microarrays. Therefore, it is desirable to establish a label-free method to quantify the probe density prior to target incubation to calibrate the fluorescence readout. Previously, a silicon oxide on silicon chip design was introduced to enhance the fluorescence signal and enable interferometric imaging to self-calibrate the signal with the immobilized probe density. In this paper, an integrated interferometric reflectance imaging sensor and wide-field fluorescence instrument is introduced for sensitive and calibrated microarray measurements. This platform is able to analyze a 2.5 mm × 3.4 mm area, or 200 spots (100 μm diameter with 200 μm pitch), in a single field-of-view.

Figure 1

Summary of the IRIS technique. (a) A schematic of the IRIS platform. LEDs illuminate the sample surface in a reflection mode setup. The surface reflection is imaged to a camera through a 4-f system. (b) Each pixel is recording the spectral intensity information of the surface. The intensity is modulated according to the reflectance corresponding to the SiO₂ thickness. (c) Plotting a line profile across a cropped fitted image shows the SiO₂ profile of several different proteins.

Figure 2

Monte Carlo simulations of IRIS accuracy for a 1 nm thickness change while increasing NA (N = 100). (a) The paraxial approximation model becomes more erroneous at higher NA leading to a model breakdown. (b) However, the ASR model shows a significantly reduced and constant error. The vertical error bars are ±1 standard deviation.

Figure 3

Integrated instrument design. (a) Schematic of the optical layout. For fluorescence, the second beamsplitter is swapped for a dichroic mirror and the emission filters are inserted. (b) Model of instrument with key components labeled. (c) Image of the constructed instrument.

Figure 4

Comparison of the IRIS platform and the calibrated fluorescence reader's label-free modality. Diminishing concentrations of probe were spotted onto 4 slides for ONC and 4 slides for CLC. The resultant average and standard deviation of the 10 replicate spots at each spotted concentration for both calibration methods are shown here for both instruments. An R-squared value of >0.995 in both cases exemplifies the validity of the ASR model for high NA IRIS measurements. Error bars are ±1 standard deviation.

Figure 5

Validation of the integrated system. (a) The fluorescence signal shows a highly linear relationship to the measured probe density for both the ONC and CLC configurations on the calibrated fluorescence reader. (b) Comparing the integrated reader response to the IRIS and Genepix 4000B instruments shows a close relationship in calibration response. In both plots, the fluorescence is normalized by the maximum measured signal.