From sample to smartphone: Consumer-operable analytical devices for multiplex allergen detection

This thesis described the design and development of prototypical portable analytical devices for multiplex food allergen immunodetection; from sample preparation all the way through to smartphone-based readout. Furthermore, it explored the fundamental binding mechanisms underlying sandwich format immunoassays including antibody-antigen interactions, affinity, cross-reactivity, kinetics and high antigen concentration-dependent effects such as the hook effect. In Chapter 1, the scene was set by introducing the need for disposable allergen immunoassays before providing general information about sample preparation, immunosensing, 3D-printing and smartphone-detection.

In Chapter 2, a comprehensive overview of immunochemical food allergen assays and detectors in the context of their user-friendliness was provided. It summarized traditional laboratory-based methods for food allergen detection such as enzyme-linked-immunosorbent assay, flow cytometry, and SPR, and the potential to modernize these methods by interfacing them with a smartphone readout system, before discussing the emergence of novel smartphone-based food-allergen detection methods that had specifically been designed with the intention of being consumer-friendly. The chapter outlined the criteria for consumer-friendly allergen detection devices as being rapid, affordable, sensitive, simple, multiplex and linked with a smartphone-based detector.

The concepts of assay speed and sensitivity were addressed in Chapter 3, where an SPR-based method was developed for screening and selecting crude anti-hazelnut antibodies based on their relative association rates, cross reactivity and sandwich pairing capabilities, for subsequent application in a rapid LFIA. The method allowed for the selection of antibodies with optimal binding characteristics which were also reflected when applied in sandwich format carbon nanoparticle based LFIAs. One of the developed LFIAs had a time-to-result of 30 seconds and a LOD of 0.1 ppm when detecting hazelnut in a real-life cookie matrix. A smartphone was used to record videos of the developing LFIAs and endpoint images of the developed LFIAs and two freely downloadable smartphone apps were then used to analyze the data.

The antibodies selected in Chapter 3 were applied in three different formats of multiplexed paper-based immunoassay in Chapter 4, namely active and passive flow-through assays, and lateral flow immunoassays with different test line configurations. All three formats of assay formats performed well, detecting total hazeln protein (THP) and total peanut protein (TPP) in the low-ppm range in both spiked buffer and real-life cookie matrix, with the fastest assay time being 1 min and the slowest being 10 min. It was found that the LFIAs were more reproducible and consumer-operable compared with the flow-through immunoassays, and a larger dilution of THP/TPP limited the occurrence of high-concentration dependent effects.  Two different smartphone models were used for the analysis of optimized assays, showing that using an app like OpenCamera to record smartphone images allowed for excellent agreement between the two different models. Additionally, the optimal LFIA configuration was validated as a screening method in spiked matrix extract, blank matrix extract (n = 20) and incurred spiked flour.

The optimized multiplex LFIA that was validated in Chapter 4, was integrated with interconnectable, 3D-printed sample preparation devices in Chapter 5. The chapter described the development and characterization of a novel, compact, inexpensive, and prototype immunosensor combining sample preparation and on-chip reagent storage for multiplex allergen lateral flow immunosensing. The handheld prototype allowed for the total homogenization of solid food samples, 1 minute solid-liquid allergenic protein extraction, 3D-printed sieve-based filtration, ULOC-enabled dilution, mixing, transport, and smartphone-based detection of hazelnut and peanut allergens in solid bakery products with limited operational complexity. A 3D-printed smartphone holder was developed to allow for detection of developing LFIAs under controlled lighting conditions. The multiplex lateral flow immunoassay (LFIA) detected allergens as low as 0.1 ppm in real bakery products; the already consumer-operable system demonstrated its potential for future citizen science approaches by being tested by an untrained user (teenager), proving its usability.

The 3D-printed smartphone holder presented in Chapter 5 was used to enable dynamic data acquisition and false negative monitoring of developing LFIA signals in Chapter 6. This chapter comprehensively studied how high antigen concentrations influence sandwich format immunoassays using LFIA and SPR and developed a smartphone-based video method for dynamic monitoring of high concentration effects in LFIA. Digital analysis of the video data allowed for clear differentiation between highly positive and false negative samples in order to indicate whether the LFIA was operating in the assays dynamic working range or at critically high concentrations. This chapter established that while the endpoint T/C ratio is an appropriate metric for semi-quantification of LFIAs within the dynamic working range, outside of this range when the test or control line is falsely diminished, the final T/C ratio is influenced. 

The research presented in this thesis provides an important advancement in the development of portable analytical devices for integrated consumer-operable allergen detection and a means to monitor for false negative results in LFIA. In Chapter 7, the key themes of sample preparation, immunosensing and smartphone detection were re-examined; the major achievements and challenges of this thesis were dissected and an outlook to the future of disposable analytical devices were discussed.