Extracellular electrophysiological based sensor to monitor cancer cells cooperative migration and cell-cell connections

Highlights

• A sensor monitors and quantifies in real-time collective migration of cancer cells.

• The bioelectrical signal arises from a synchronized cell activity.

• The power of the signal provides information about cell-cell connectivity.

• The sensor can distinguish individual from collective cell migration.

Abstract

Herein, we describe an electrophysiological based sensor that reproducibly monitors and quantifies in real-time collective migration and the formation of cell-cell junctions by C6 glioma cells seeded on top of electrodes. The signal amplitude and frequency generated by the migrating cells changed over time and these parameters were used to accurately calculate the migration speed. Electrophysiological measurements could also distinguish individual from collective cell migration. The migration of densely packed cells generated strong signals, while dispersed cells showed weak bioelectrical activity. We propose this electrophysiological technique as a cell-based biosensor to gain insight into the mechanisms of cooperative migration of cancer cells. Possible applications include screening for anti-migratory compounds, which may lead to the development of novel strategies for antineoplastic chemotherapy.

Introduction

Cell-based biosensors have been utilized for cellular physiological analysis, pharmaceutical evaluation, environmental monitoring and medical diagnosis (Banerjee and Bhunia, 2010; DeBusschere and Kovacs, 2001; Johnstone et al., 2010; Pan et al., 2019; Pancrazio and Borkholder, 1999; Ye et al., 2019). Electrogenic cells or tissues are cultured on top of microelectrode arrays known as MEAS (Hierlemann et al., 2011). Extracellular potential changes can be monitored to assess the influence of drugs on cellular activities. Following the pioneering work of Gross (Gross et al., 1995) who used neuronal networks seeded on MEAS, some other electrogenic cells and tissues, such as cardiomyocytes, brain slices and retinal networks, are also gradually becoming acceptable as pharmacological models (Spira and Hai, 2013). Recently, we (Medeiros et al., 2016; Rocha et al., 2016) and others (Cabello et al., 2019) reported that populations of non-electrogenic or non-excitable cells such as cancer cells and astrocytes (Mestre et al., 2017b, 2017a) synchronize their activity and generate discrete bioelectrical signals. This type of sensing device differs from conventional MEAS because they are comparatively large area devices and therefore do not have the spatial resolution to measure individual action potentials. Furthermore, these sensing electrodes together with non-invasive electrophysiological measuring techniques can be used as cell-based biosensors to study cooperative activity in non-excitable cells.

A highly relevant biological problem in which cell-cell cooperation occurs is cell migration and in particular cell migration during cancer metastasis. Cancer cells are capable of synchronizing their signaling apparatus, thus cooperating to migrate and invade. Ion channels and ionic waves may have an important role when cells move as coherent groups. This type of migration is known as collective cell migration (Gov, 2014; Jiang et al., 2015; Theveneau and Mayor, 2012). The underlying cellular and molecular mechanisms of collective migration include cell-cell adhesion, force generation and orientation. The regulation of these interactions requires the coordination of a multiplicity of signals, both spatially and temporally. Despite some molecular mechanistic insights, it remains unclear how the various, chemical, mechanical and, potentially, electrical signals from neighboring cells become integrated to allow coordinated multicellular movements. There is evidence that this coordination may involve multicellular Ca²⁺ waves transmitted from cell to cell (Giannone et al., 2002; Klepeis et al., 2001). Ion channels (migration-associated transportomes) are instrumental in cell migration (Schwab and Stock, 2014), playing multiple roles; i) they act as a sensor for extracellular guidance cues (Gurnett and Hedera, 2007), ii) they mediate the influx of Ca²⁺ that controls actin polymerization (Veksler and Gov, 2009), and iii) they induce cell shape changes and membrane depolarization (Schwab et al., 2012). The study of ion channels essentially relies on the use of patch clamp techniques (Chen et al., 2009). This technology requires precision micromanipulation and operators with large experience and skills. The integration of patch clamp methods with a migration experiment is also not trivial. In this context extracellular measurements of the ionic fluctuation during migration of cell populations are a simpler technique and may provide direct insight into the way the cells coordinate their collective movement.

Herein we present a cell-based biosensor that relies on extracellular electrophysiological cooperative cell activity. The method provides a set of relevant parameters about cell migration, namely speed and the kinetics of the establishment of cell-cell connections.

In order to validate the electrophysiological based method, we use a standard and well-accepted technique to measure cell migration. This method know as electrochemical cell impedance spectroscopy (ECIS) relies on the change of the electrode impedance with cell coverage (Giaever and Keese, 1991; Wegener et al., 2000). Cell migration experiments using the ECIS method were carried out in parallel with electrophysiological time traces. The spectral properties of the recorded signals, namely the frequency and the signal duration, were correlated with cell confluence estimated by the ECIS method. We used C6 glioma cells as a model system. The C6 system has similarities with primary brain tumors, gliomas, and the role of ion channels in the promotion of cell motility has been particularly well studied using patch clamp (Bowman and Lohr, 1996) and optical fluorescence methods (Pollak et al., 2017; Bose et al., 2015; Cuddapah et al., 2014; Mcferrin and Sontheimer, 2006; Pollak et al., 2017; Wang et al., 2015).