A digital microfluidic device integrated with electrochemical sensor and three-dimensional matrix for detecting PD-L1

Introduction:: PD-L1 is an immune checkpoint protein that can bind to PD-1 receptors on T cells. This interaction can diminish T-cell activities and thus promote tumor growth. Immune checkpoint inhibitors, an revolutionary cancer immunotherapy, can block this interaction, but the effectiveness varies among tumor types and individuals. Meanwhile, drug resistance and immunotoxicities can occur. Therefore, evaluating therapeutic responsiveness in a regular basis is crucial to maximize benefits and minimize adverse health effects. Enzyme-linked immunosorbent assay (ELISA) is the gold standard, however it requires trained technologists to operate equipment in centralized laboratories, creating logistic challenges for frequent therapeutic monitoring. To bridge this gap, this study integrates an electrochemical sensor array into a digital microfluidic (DMF) device to quickly quantify PD-L1 through an immunoassay. A distinct advantage of a DMF platform is its ability to handle sample droplets in an automated and programmable fashion, which allows us to perform the immunoassay with minimally trained hands. To increase the detection sensitivity, we integrated a 3D matrix structure on the sensing electrodes. The 3D matrix is composed of reduced graphene oxide (rGO), bovine serum albumin (BSA), and glutaraldehyde (GA), which can increase binding sites and reduce non-specific binding. This integrated system presents excellent sensitivity and selectivity, and is able to detect as low as 1 pg/mL PD-L1. In the long term, it is possible to detect multiple markers simultaneously in this DMF platform, which makes this platform highly suitable for point-of-care testing.


Materials and Methods:: The DMF platform is a double-plate electrowetting-on-dielectrics (EWOD) device. There are 40 actuation electrodes on the bottom plate for droplet handling, and an array of electrochemical sensors on the top plate (Figure a) [1]. The sensing electrodes were drop-casted with a rGO/BSA/GA layer to accommodate the capture antibody(Figure b). The liquid mixture was prepared using 5 mg/mL of BSA in phosphate buffered saline (PBS) with 5 mg/mL of tetraethylene pentamine functionalized rGO (rGO-TEPA). The resulting composite solution was sonicated for 5 hours, then heat-denatured at 105 °C for 5 min. Any excess black agglomerate of undissolved graphene was removed from the nanomaterial solution via centrifugation (10xg, 15 min). GA was added to the clear nanomaterial supernatant solution in a 70:1 volume ratio before applying the coating to the cleaned sensors.

A sandwich immunoassay was formed by incubating the captured PD-L1 with a secondary antibody labeled with HRP, followed by another incubation with precipitating TMB for electrochemical characterization. The device was operated in a commercial DropBot system, and the sensing arrays were interfaced with an electrochemical workstation (CHI600E). Cyclic voltammetry (CV) detections were conducted between -0.8V and 0.6V under the scan rate of 100 mV/s in an electrolyte solution containing 10 mM K₃[Fe(CN)₆]/K₄[Fe(CN)₆] and a supporting electrolyte (0.1 M KCl). Differential Pulse Voltammetry (DPV) was measured with the potential step, frequency and amplitude set as 0.01V, 20 Hz and 50 mV. Electrochemical Impedance Spectroscopy (EIS) was carried out between 0.1 HZ to 1 MHz under the initial voltage of 50 mV.


Results, Conclusions, and Discussions:: In the 3D matrix, GA crosslinked BSA and rGO nanoflake; this antifouling coating have displayed exceptional 3D porous matrix structure for biomolecule characterization in biofluids [2]. Based on SEM characterization, rGO/BSA/GA material was densely arranged onto the Au electrode surface, resulting in a roughened surface that can increase the accommodation of molecules (Figure c). Electrical property changes in Figure d showed that the coating retained 49.1% (oxidation) and 67.2% (reduction) of current density, and a 1.8-fold increment of charge transfer resistance (Rct). In addition, the current produced during voltammograms were directly proportional to the square root of the scan rate, indicating a diffusion-limited process on the coated surface for electrochemical characterization.

An immunoassay was performed to detect PD-L1 as depicted in Figure f. The DPV characterization were conducted after each treatment, and the current decreased sequentially after surface coating and antibody conjugation due to inhibition of the charge transfer between the sensor surface and the electrolyte solution (Figure g). Current signal decreased further with increased PD-L1 concentrations (μg/mL) captured by antibodies. A calibration curve was plotted with excellent linearity with an R² value of 0.99. To amplify the electrochemical signal, a sandwich immunoassay was formed by adding detection antibodies conjugated with biotin molecules. Following this, poly-HRP-streptavidin was incubated to react with biotin, and precipitation TMB (3,3′,5,5′-Tetramethylbenzidine) was used to catalyze the HRP which precipitates insoluble reaction product locally at the electrode where the enzyme is present. The immunoassay allows the detection of PD-L1 at pg/mL level with great linearity (R²=0.96). In addition, the non-specific binding was characterized with two controls, and the results showed that the current changes caused by non-specific absorption (0.056) and blank control (0.055) were significantly lower than the specific capture (0.236) of PD-L1 at 100 pg/mL (Figure h).

This work demonstrated an integrated platform to detect PD-L1 at pg/mL level in a simple and automated manner. This highly sensitive and selective platform holds the potential to be a promising point-of-care tool that can provide regular evaluation of patients’ immune responsiveness of PD-1/PD-L1 blockade therapy to inform the effectiveness of treatment.


Acknowledgements (Optional): : This work was funded by the Ivan Bowen Family Foundation. The authors thank the Microbiome Program and the Center for Individualized Medicine at Mayo Clinic for their support.


References (Optional): : [1] Y. Zhang and Y. Liu, "A Digital Microfluidic Device Integrated with Electrochemical Impedance Spectroscopy for Cell-Based Immunoassay," Biosensors, vol. 12, no. 5, 2022, doi:https://doi.org/10.3390/bios12050330.

[2] U. Zupančič, P. Jolly, P. Estrela, D. Moschou, and D. E. Ingber, "Graphene Enabled Low-Noise Surface Chemistry for Multiplexed Sepsis Biomarker Detection in Whole Blood," Adv. Funct. Mater., vol. 31, no. 16, p. 2010638, 2021, doi: https://doi.org/10.1002/adfm.202010638.