Towards an ultra-rapid smartphone- connected test for infectious diseases

Abstract

The development is reported of an ultra-rapid, point-of-care diagnostic device which harnesses surface acoustic wave (SAW) biochips, to detect HIV in a finger prick of blood within 10 seconds (sample-in-result-out). The disposable quartz biochip, based on microelectronic components found in every consumer smartphone, is extremely fast because no complex labelling, amplification or wash steps are needed. A pocket-sized control box reads out the SAW signal and displays results electronically. High analytical sensitivity and specificity are found with model and real patient blood samples. The findings presented here open up the potential of consumer electronics to cut lengthy test waiting times, giving patients on the spot access to potentially life-saving treatment and supporting more timely public health interventions to prevent disease transmission.

Figure 1

Smartphone-connected SAW test for HIV (a) Schematic to illustrate the laboratory prototype. The control box sends/receives an analogue signal to/from 4 SAW biochips in parallel, and transmits a digital signal to the smartphone (or laptop). An app (or software) processes and analyses the data. (b) Photograph of the hand held SAW development prototype. This figure is not covered by the CC BY licence. [Credits to H Yatsuda of Japan Radio Company]. All rights reserved, used with permission. (c) A schematic to illustrate the principle of SAW generation on biochips via the piezoelectric effect: The SH-SAW is transmitted from the input interdigitated electrode (left IDT) to the output IDT (right) and propagates along the sensing area. The relative phase shift Δφ measured between the time t and the start of the measurement t ₀ is continually measured to provide real-time analysis. (d) Photograph of a disposable SAW biochip measuring 25 mm × 7 mm × 2 mm. (e) A schematic to illustrate the concept of biosensing on SAW biochips: The sensing area (comprising a gold thin film, titanium adhesion layer on quartz) is functionalized with a monolayer of capture ligands (e.g. a protein that binds to the biomarker of interest, blue) using alkanethiol linker chemistry. A sample containing a HIV antibody biomarker (red) is shown in solution binding to the capture ligand. The resultant wave phase shift, Δφ, increases with the amount of capture protein bound to the surface and biomarker specifically bound to the biochip (3). (f) Schematic of HIV. Anti- p24 antibodies are raised against the viral protein p24, which forms the capsid of the virus (shown in orange). These antibodies are raised within two to three months at levels of 10–1000 µg/ml^,. However, seroconversion can take up to six months and therefore recently acquired infections can pass undetected by these tests. During this window, the p24 capsid protein becomes detectable (2 to 3 weeks).

Figure 2

Anti-p24 antibody detection in buffer using a SAW biochip functionalised with HIV p24 recombinant proteins (a) Schematic illustrating the specific binding of anti-p24 HIV antibodies to p24 coated biochips, and a negative control using an antibody with no affinity for p24 (anti-GBP5 antibody) which shows no binding to the biochip. (b) Overlaid raw data plots to show the phase shift recorded between input and output IDTs as a function of time. The sample containing the anti-p24 antibody is injected at t = 0. Each sample trace was normalised with the reference assay (containing only buffer, orange line). (c) Graph showing the total phase shift recorded after 5 minutes, plotted as a function of anti-p24 antibody concentration. Each measurement was repeated 3 times; error bars shows the standard deviation of the mean. Black line: Langmuir adsorption isotherm of equation y  =  47.8 (±3.3)x/(12.2 (±2.7) + x), R ² = 0.994 Inset: Zoom of linear regression in the range 2–50 nM, y = 1.56x + 3.2, R ² = 0.964. The estimated limit of detection (LOD) and the lowest detected concentration (LDC) are marked on the bottom axis by the blue and green arrows, respectively. (d) Plot showing the average phase shift recorded every 5 seconds during the first twenty seconds after injection. Each point represents the average of three measurements, errors bars represent the standard deviation from the mean. Samples of different concentrations (from 50 nM (7.5 µg/ml) and above) can be distinguished from one another 10 seconds after sample injection.

Figure 3

HIV p24 detection in buffer using a SAW biochip functionalised with anti-p24 llama VHH. Schematic showing the immuno-sandwich used to detect HIV p24 (the chip is coated with anti-p24 llama VHH) and the resultant phase shift recorded between the input and the output IDT as a function of time. The numbers shown on the graphs refer to different immuno-sandwich complexes formed using the anti-p24 antibodies listed in Table 1. Samples injected at t = 0. (a) Direct HIV-p24 detection (llama VHH capture + p24 only). (b) Immuno-sandwich p24/anti-p24 complexes are formed in the sample and bind to the functionalised surface. The largest signal is seen for immuno-complex 3. (c) Control samples, where no p24/anti-p24 complexes are formed. (d) Titration of HIV-p24 using optimised immuno-sandwich with NIH-3537 anti-p24 antibody (number 3 in Table 1). The phase shift was recorded 5 minutes after sample injection. Black line shows Langmuir isotherm fit of equation y  =  49.6 ( ± 7.7)x / (45.8 ( ± 19.9) + x), R ² = 0.959. Data shown are the combined results from three measurements and error bars show standard deviation of the mean. (e) Zoom on the 0–40 nM region and fitted with a linear regression model (black line) of equation y = 0.49x + 0.94, R ² = 0.964. Data shown are the combined results from three measurements and error bars show standard deviation of the mean.

Figure 4

Detection of anti-HIV antibodies in a plasma sample from a patient with HIV (a) Schematic of the reference chip functionalized with non-animal protein (NAP), and test chip functionalised with recombinant HIV protein corresponding to the biomarker (HIV p24). Non-specific binding of various plasma proteins and antibodies occurs on both chips, but the biomarker (anti-p24) only binds specifically to the test chip. (b) Anti-p24 detection. Phase shift plotted as a function of time for two different samples (one HIV-positive sample in red and one HIV-negative sample in green). Dashed lines represent the reference chips, and the solid lines the test chips. Sample injected at t = 0. (c) Differential response of test and reference chips for anti-p24 detection in patient samples. The differential test readout represents the change in phase shift due the specific binding of the biomarker to the SAW biochip, and removes the effect of other non-specific perturbations such as the difference in temperature or viscosity between buffer and plasma. Sample injected at t = 0.