Ultradense electrochemical chip, microfluidics and machine learning to monitoring cell viability: a promising platform of cancer drug screening.

• Grant number: 24/10870-7
• Support Opportunities: Scholarships in Brazil - Doctorate (Direct)
• Start date: November 01, 2024
• End date: October 31, 2028
• Field of knowledge: Physical Sciences and Mathematics - Chemistry - Analytical Chemistry
• Principal Investigator: Renato Sousa Lima
• Grantee: Paula Cristine Rocha Corsato
• Host Institution: Centro Nacional de Pesquisa em Energia e Materiais (CNPEM). Ministério da Ciência, Tecnologia e Inovação (Brasil). Campinas , SP, Brazil
• Associated research grant: 23/00246-1 - Miniaturized large-scale devices for in-situ analysis: fabrication, characterization and applications, AP.TEM

Abstract

The pharmaceutical industry faces critical challenges in developing successful pharmaceutical drugs. This process takes 10 to 15 years and requires around US$1 billion. Furthermore, more than 95% of drug candidates fail to get approval from regulatory agencies. On average, the preclinical drug screening takes from 5 to 7 years. Along this journey, the candidates are challenged with cell cultures to predict their in vivo efficacy and toxicity. The in vitro effects of drugs are usually gauged through sensing methods such as MTT and Alamar blue assay, which are incapable of providing the large-scale and fast assessment of cell viability (i.e., the number of live cells in response, e.g., to a pharmaceutical drug). In this way, electrochemical sensors have been devised as a promising solution to cope with these limitations. In particular, the electrochemical devices can indirectly assess cell viability because the cells are detached from the electrode surface as they die, hence implying drug-evoked changes in the impedance or faradaic current signals. Although electrochemical sensors are promising for conducting drug susceptibility tests, producing these analyses with (1) high throughput, (2) simplicity, and (3) accuracy remains a challenge for their commercial development. In this context, we propose in this project the amalgamation of a miniaturized ultradense electrochemical chip with microfluidics and machine learning (ML) aiming to generate these analyses that are new calibrations free for determining the viability of 2D tumor cells and the half-maximal lethal concentration of the drug (LC50).The analyses will be performed in a single microfabricated chip that contains a high density of vertical, grid-patterned arrayed sensors (33 to 870). This new sensor array was developed by our group (Adv. Healthcare Mater. 2024, 13, 2303509, and ACS Appl. Mater. Interfaces 2024, DOI: 10.1021/acsami.4c01159) and combines high reproducibility and resolution with low cost due to the large number of sensors on a single chip. Another advantage of these sensors is the low number of conductive lines, which allows for the high integration capacity of several sensors in a compact chip. Thus, these sensors are compatible with microfluidics, a tool that increases reproducibility and decreases reagent consumption, among other advantages. In this project, microfluidic devices will be obtained by sealing the chip with polydimethylsiloxane (PDMS) channels. The device will have 9 channels, with 5 sensors each (45 sensors on a chip). Therefore, up to 9 different samples can be analyzed in series with measurements in quintuplicate for each case. The chip allows serial analyses by simply changing the contact of an electrode. Based on this property, several square wave voltammograms (SWV) can be quickly obtained using a handheld, one-channel workstation, thus enabling drug susceptibility analyses with high throughput and simplicity. Since each SWV measurement will last 7 seconds, the analyses of all 45 sensors on the chip will be completed in just 315 seconds. Cell viability will be determined through SWV analyses of the ion Ru(NH3)63+. The SWV technique will generate not only fast analyses but also multivariate data that will be explored to enhance the accuracy of cell viability prediction using ML. After application to different tumor cells and drugs, ML will also be adopted to obtain a unique equation for predicting cell viability. With this, we aim to allow viability determination through a universal equation, eliminating the need to calibrate the sensor for each new cell or drug analyzed. Ensuring the use of the method without requiring new calibrations will be essential for its widespread applicability. In summary, the solutions addressed here could help accelerate the transformation of drugs into safe and effective medicines for society.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DORETTO, DANIEL S.; CORSATO, PAULA C. R.; SILVA, CHRISTIAN O.; PESSOA, JAMES C.; VIEIRA, LUIS C. S.; DE ARAUJO, WILLIAM R.; SHIMIZU, FLAVIO M.; PIAZZETTA, MARIA H. O.; GOBBI, ANGELO L.; RIBEIRO, IRIS R. S.; et al. Ultradense Electrochemical Chip and Machine Learning for High-Throughput, Accurate Anticancer Drug Screening. ACS SENSORS, v. N/A, p. 12-pg., 2024-11-29. (23/00246-1, 24/10870-7)