Label‐Free and Regenerative Electrochemical Microfluidic Biosensors for Continual Monitoring of Cell Secretomes

Development of an efficient sensing platform capable of continual monitoring of biomarkers is needed to assess the functionality of the in vitro organoids and to evaluate their biological responses toward pharmaceutical compounds or chemical species over extended periods of time. Here, a novel label‐free microfluidic electrochemical (EC) biosensor with a unique built‐in on‐chip regeneration capability for continual measurement of cell‐secreted soluble biomarkers from an organoid culture in a fully automated manner without attenuating the sensor sensitivity is reported. The microfluidic EC biosensors are integrated with a human liver‐on‐a‐chip platform for continual monitoring of the metabolic activity of the organoids by measuring the levels of secreted biomarkers for up to 7 d, where the metabolic activity of the organoids is altered by a systemically applied drug. The variations in the biomarker levels are successfully measured by the microfluidic regenerative EC biosensors and agree well with cellular viability and enzyme‐linked immunosorbent assay analyses, validating the accuracy of the unique sensing platform. It is believed that this versatile and robust microfluidic EC biosensor that is capable of automated and continual detection of soluble biomarkers will find widespread use for long‐term monitoring of human organoids during drug toxicity studies or efficacy assessments of in vitro platforms.


Functionalization of microelectrode bonded PDMS chip:
The biotinylated anti-albumin and anti-GST-α were obtained from ELISA kits which was purchased from Abcam (USA) and MyBiosource (USA), respectively. The biotinylated antibodies (biot-Ab) of interested biomarker were immobilized to the surface of microelectrode as detailed below. After two-step EC cleaning, the microelectrode was rinsed by anhydrous ethanol and 11-mercaptoundodecanoicacid (MUA, 10 mM prepared in anhydrous ethanol, Sigma). After one hour incubation of MUA, rinsing with ethanol and DPBS to remove unbounded thiols, the carboxylicacid terminated alkyl surface was converted to an active NHS ester via 7 min washing and 30 minute incubation of freshly Unbounded SPV molecules were removed via 7 min washing by washing buffer (provided by ELISA kits). The biot-Ab were immobilized to the surface via SPB-Biotin interaction upon 7minute washing of biot-Ab (1x dilution in dilution buffer, Abcam) prior to one-hour incubation.
Having rinsed the biot-Ab modified microelectrode surface by the washing buffer, the surface of microelectrode was blocked via biomarker-free cell culture media for 1 hour to prevent nonspecific binding of proteins.

EC detection of albumin and GST-: Varying concentrations of human albumin and GST-
 were analyzed via EC impedance spectroscopy [1] performed on a CHI 660E EC workstation (Shanghai CH Instruments Co., China) with three-electrode system fabricated on microelectrode; Au working electrode, Ag reference electrode and Au counter electrode. EIS measurements were conducted in the frequency range varying from 10 -1 Hz to 10 5 Hz under a potential value of 0.10 V and modulation amplitude of 5 mV in 50 mM K 3 Fe(CN) 6 . The charge transfer resistance (R ct ) of samples were determined via fitting of the raw data by CHI software based on the equiavalent circuit composed of the circuit elements; R ct , solution resistance R s , double-layer capacitance, C dl , and Warburg impedance, W. All impedance data collected for various conditions were fitted using the same model ( Figure S1(e)) and had goodness of fit values within the range of 10 -6 and the errors associated with the fitting parameters were 2%-6%. The limit of detection (LOD) value was calculated by three times the standard deviation of the media signal.

Regeneration of microelectrode:
In order to achieve the long-term monitoring, the 1.6. Characterization of microelectrode surface: To characterize the surface of microfabricated Au electrode before and after modification via proteins and also two-step regeneration, Atomic force microscopy (AFM) and Energy Dispersive X-ray spectroscopy (EDX) have been done. AFM was performed by using Bruker Dimension Icon AFM (Bruker Corporation, Germany). The images were analyzed using the offline AFM software (NanoScope Analysis, version 1.5) in tapping-mode in air at room temperature.

Bioreactor construction, bioprinting and drug treatment:
The bioreactor was assembled as formerly described [3] . A 20 dot microarray of GelMA hydrogel embedded spheroid droplets were printed by a NovoGen MMX bioprinter (CA, US). Each dot contained average 18 spheroids.
The dots were printed on a tetramethylsilane (TMSPMA) treated glass and exposed under UV light exposure (850mV at 8.5 cm) for 18 s, and then GelMA dots were successfully crosslinked and attached onto the culture glass in the bioreactor. Subsequently the bioreactor was closed and perfused with a peristaltic pump. Medium was pumped from a reservoir, passed through the bioreactor and collected on days 1, 3, 5 and 7 for ELISA and electrochemical (EC) sensing, respectively. APAP was dissolved in cell culture media and was continuously administrated as described in our previous work [4] . APAP doses of 5 mM and 10 mM were used in different bioreactors for drug toxicity evaluation on 1, 3, 5 and 7 days.

Cell counts:
To count the total number of cells in the printed sample, the GelMA hydrogel embedded spheroids were treated by 1.0 mg/ml collagenase II solution at 37 o C for 10 min until the GelMA hydrogel fully degraded. We then added 1X trypsin solution for 5 min at 37 o C to obtain the single cell suspension. The obtained cell suspension was washed by the fresh culture media and then the cells were counted with a hemocytometer.

Live and dead assay and microscopy: LIVE/DEAD® viability (Life Technologies)
assays were performed to assess viability on the control and APAP treated bioreactors on 1, 3, 5 and 7 days. Fluorescent images were acquired by Zeiss Axio Observer D1 Fluorescence Microscope (Carl Zeiss, Germany) was used for acquisition. Images were analyzed by ImageJ analysis software [5] .

Statistical analysis:
To analyze statistical significance, we used a one-way ANOVA where appropriate (GraphPad Prism 6.0, GraphPad Software). Error bar represents the mean ± standard deviation (SD) of measurements performed on each sample groups. To determine whether a significant difference exists between specific treatments, we used Tukey's multiple comparison tests (p<0.05).              for control, 5mM and 10 mM APAP exposed bioreactor samples (n=3).