TFT sensor array for real-time cellular characterization, stimulation, impedance measurement and optical imaging of in-vitro neural cells

https://doi.org/10.1016/j.bios.2020.112546Get rights and content

Highlights

  • Multimodal transparent TFT biosensor chip is developed for neuronal ensemble investigation.

  • Extracellular recording, electrical and chemical stimulation, and calcium imaging are performed.

  • Impedance spectroscopy for cell detection, mapping and cellular status monitoring is performed.

  • Any TFT sensor is accessible by all six modalities using switch matrix.

  • In-vitro neuron culture study is possible by electrical, chemical, and optical characterizations.

Abstract

Real-time in-vitro multi-modality characterization of neuronal cell ensemble involves highly complex interdependent phenomena and processes. Although a variety of microelectrode arrays (MEAs) have been reported, diagnosis techniques are limited in term of sensing area, optical transparency, resolution and number of modalities. This paper presents an optically transparent thin-film-transistor (TFT) array biosensor chip for neuronal ensemble investigation, in which TFT electrodes are used for six modalities including extracellular voltage recording of both action potential (AP) and local field potential (LFP), current or voltage stimulation, chemical stimulation, electrical impedance measurement, and optical imaging. The sensor incorporates a large sensing area (15.6 mm × 15.6 mm) with a 200 × 150 array of indium-tin-oxide (ITO) electrodes placed at a 50 μm or 100 μm pixel pitch and with 10 ms temporal resolution; these performances are comparable to the state-of-the-art MEA devices. The TFT electrode array is designed based on the switch matrix architecture. The reliability and stability of TFTs are examined by measuring their electrical characteristics. Impedance spectroscopy function is verified by mapping the neuron position and the status (cells alive or dead, contamination) on the electrodes, which facilitates the biochemical studies in electrical domain that adds quantitative views to visual observation of cells through the optical microscopy. An in-vitro neuron culture is studied using electrophysiological, electrochemical, and optical characterization. Detailed signal analysis is demonstrated to prove the capability of bioassay.

Introduction

Demands for investigating the neuronal networks and the functions of the nervous system are escalating in the field of advanced in-vitro drug screening (Park et al., 2015) to cope with the central nervous system diseases such as neurodevelopmental, psychiatric, and neurodegenerative disorders (Grskovic et al., 2011). The electrogenic cells generating electrical signals facilitate the study in the electrical domain under in-vitro or in-vivo conditions (Azim et al., 2019; Cohen-Karni et al., 2010; Hess et al., 2011; Santoro et al., 2014; Shaik et al., 2020). The in-vivo study of complex neuronal networks is challenging because of the physiological complications, ethical dilemmas, and additional filtration. In-vitro neuronal study, being further classified as either extracellular or intracellular, uses measurement tools such as patch-clamp and sharp microneedle probes. Although intracellular study has an advantage of a high signal-to-noise ratio (Angle et al., 2015), it has several major drawbacks such as mechanical instability, incompatibility with multisite recording, and a risk of fluid contamination or dilution with the electrode fluid in the case of patch-clamp studies, all of these resulting in inaccurate diagnosis (Pottosin and Dobrovinskaya, 2015). Understanding the brain functions, for instance, depends mostly on the extracellular recording, which consists of direct measurement of electrical activities from recording electrodes placed close to the neuronal cell culture without causing damage to the neuron's plasma membrane (Aziz et al., 2009; Kim et al., 2015; Ruther and Paul, 2015; Shen et al., 2015). The resulting voltage signals are composed of the local field potentials (LFPs) and the spikes emitted by one or more neurons (Zanos et al., 2011).

To develop an in-vitro model through the study of neural dynamics, an ideal readout device needs to have the following features: biocompatibility, optical transparency (for fluoroscopy analyses), high fill factor, large sensor area, on-chip multiplexing, bidirectional interface, fast temporal response, and high spatial resolution. Traditional multi-electrode-array (MEA) devices are electrically passive and planar in shape with metal electrodes on top of a glass substrate (Abdelfattah et al., 2016; Foidl et al., 2018; Franke et al., 2017). However, due to the congestion problem of electrical interconnection, MEAs usually have limited number of independent electrodes, resulting in a low density and a low throughput of measurement (Cabello et al., 2018; Hood et al., 2009; Lu et al., 2016; Spira and Hai, 2013). On the contrary, complementary metal-oxide semiconductor (CMOS) arrays can accommodate a large number of electrodes at a high density. Nonetheless, silicon substrate for CMOS technology is opaque, and it would not allow visual observation by the inverted microscope (Bertotti et al., 2014; Cabello et al., 2018; Heer et al., 2006; Park et al., 2018; Snell et al., 1981).

Besides MEA and CMOS technologies, thin film transistor (TFT) technology is expected to play an increasing role for the electrophysiology. TFT is a well-known technology for flat panel display (Snell et al., 1981). Several milestones have been attained (Brody et al., 1973; Le Comber et al., 1979; Spear and Le Comber, 1993; Yamamoto, 2012) since the first TFTs were made of a compound semiconductor called cadmium sulfide (CdS), which was introduced in 1962 by Weimer (1962). Hosono's group with the Tokyo Institute of Technology, Japan, first explored the semiconducting indium gallium zinc oxide (IGZO) as a TFT channel material, and demonstrated crystalline and amorphous IGZO-TFTs in 2003 and 2004, respectively (Nomura et al., 2004, 2003).

In this paper, we present a transparent high-definition crystalline-IGZO-TFT system as a bio-measurement platform. The device chip is made by the same fabrication technology for the liquid-crystal display (LCD) fabrication. The crystalline-IGZO is an optically transparent semiconductor material with electron mobility of 20–50 times higher than that of amorphous silicon. In this study, crystalline-IGZO-TFTs are used as a transparent substrate for see-through observation of bio-tissue, and the high electron mobility is adopted for high-speed recording of the ionic behaviors. The system based on the TFT technology thus benefits from the optical transparency and high-density electronics to produce a sensor array of various cell analysis.

The purpose of this paper is to demonstrate all the potentials of TFT systems applicable to a neuron culture. Previous reports from the authors’ group demonstrated a singular modality (neuron sensing only, or impedance sensing). The different aspects of the systems are described in the following sections. The system architecture, corresponding functional units, and their integration are described in Sections 2. The electrical characterization and measurements as well as the biological measurements are described in Section 3. Sections 4 Comparison with the state-of-the-art, 5 Conclusion respectively give a detailed summary and comparison to the states-of-the-art measurement methods.

Section snippets

System architecture

Crystalline-IGZO-TFT in this work is designed to perform multiple stimulation/measurement functions as schematically shown in Fig. 1(a) by appropriately choosing the electrode sites (Shaik et al., 2018, 2017). The block diagram in Fig. 1(b) illustrates the system architecture, where all the functional units are depicted. Typical chip has 175 × 175 TFT microelectrodes over an area of 243 mm2. TFTs plates of several cm2 or more are achievable, which is useful for organ-on-chip study (Wikswo et

Device characterization

Characterization of the TFT electrodes has been performed to ensure precise and selective stimulation of individual neurons. Typical performances are shown in Fig. 3, including (a) drain voltage response, (b) gate voltage response, (c) frequency response, (d) total harmonic distortion (THD), (e) drain current as a function of transistor gate width, (f) electrolysis test, (g) stress characterization, and (f) optical characterization. The output curves give qualitative information such as the

Comparison with the state-of-the-art

The performance of the developed TFT chip has been summarized and compared with other states-of-the-art devices used for electrogenic cells study as shown in Table 1. In the past, a large number of studies have been carried out on opaque substrates such as silicon-based MOSFET and ion-sensitive field-effect transistor. A trade-off exists in the conventional MEAs between the substrate transparency and the electrode density; achieving the high transparency required a sparse electrode density and

Conclusion

An arrayed biosensor chip with six modalities for neuronal ensemble investigation has been developed using the thin-film-transistor (TFT) technology, and characterized from electrophysiological, electrochemical and optical points of view. The electrodes are designed in such a way that they can accommodate all the necessary features as an ideal readout system to understand the complex neuronal behaviors. Besides CMOS, the high definition TFT system includes the largest number of measurement or

CRediT authorship contribution statement

Faruk Azam Shaik: Writing- original draft. Hiroshi Toshiyoshi: Writing - review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This work was supported by JSPS Grant-in-Aid for Scientific Research (B) Grant Number 15H03984.

References (58)

  • M.R. Angle et al.

    Curr. Opin. Neurobiol.

    (2015)
  • M. Cabello et al.

    Sensor. Actuator. B Chem.

    (2018)
  • S.B. Dunnett et al.

    Brain Res. Protoc.

    (1997)
  • B.M. Foidl et al.

    J. Neurosci. Methods

    (2018)
  • C. Grienberger et al.

    Neuron

    (2012)
  • P. Ruther et al.

    Curr. Opin. Neurobiol.

    (2015)
  • Y. Song et al.

    Biosens. Bioelectron.

    (2012)
  • W.E. Spear et al.

    Solid State Commun.

    (1993)
  • A.S. Abdelfattah et al.

    J. Neurosci.

    (2016)
  • N. Azim et al.

    J. Microelectromechanical Syst.

    (2019)
  • J.N.Y. Aziz et al.

    IEEE J. Solid State Circ.

    (2009)
  • V. Benfenati et al.

    Nat. Mater.

    (2013)
  • G. Bertotti et al.
  • G.J. Brewer et al.

    J. Neurosci. Res.

    (1993)
  • T.P. Brody et al.

    IEEE Trans. Electron. Dev.

    (1973)
  • G.A. Cathcart et al.

    IEEJ Trans. Sensors Micromachines

    (2020)
  • G.A. Cathcart et al.
  • G.A. Cathcart et al.
  • G. Cellot et al.

    J. Biomed. Nanotechnol.

    (2017)
  • T. Chi et al.

    IEEE Trans. Biomed. Circuits Syst.

    (2015)
  • C.H. Chu et al.

    Sci. Rep.

    (2017)
  • T. Cohen-Karni et al.

    Nano Lett.

    (2010)
  • A.P.P. Correia et al.
  • A. Dietzel

    Microsystems for Pharmatechnology

    (2016)
  • J. Dragas et al.

    IEEE J. Solid State Circ.

    (2017)
  • K. Franke et al.

    Nature

    (2017)
  • M. Grskovic et al.

    Nat. Rev. Drug Discov.

    (2011)
  • M. Gulino et al.

    Front. Neurosci.

    (2019)
  • K.M. Harris

    Some typical dimensions of dendrites for a few types of neurons

  • Cited by (20)

    • a-IGZO thin-film transistors with transparent ultrathin Al/Ag bilayer source and drain for active neural interfaces

      2023, Materials Science in Semiconductor Processing
      Citation Excerpt :

      The shortcomings of the transparency of neural interfaces are surmountable by thin-film transistors (TFTs). Fabrication of high-density electrodes and usage of transparent glass substrates are possible with TFT-based active-matrix systems [9–11]. Moreover, the TFT itself can be transparent and flexible when wide bandgap semiconducting materials such as a-IGZO are used, which is well known for high optical transparency and mobility in display applications.

    • Recent trends of biomaterials and biosensors for organ-on-chip platforms

      2022, Bioprinting
      Citation Excerpt :

      The focused ultrasound was integrated vertically on the microfluidic device so the pressure waves were delivered perpendicularly. Shaik et al. [100] presented a thin-film-transistor (TFT) array biosensor chip for neuronal ensemble investigation. The biosensor allowed the recording of local field and action potentials, as well as optical imaging, chemical stimulation, current or voltage stimulation and electrical impedance measurement.

    • A real-time mirror-LAPS mini system for dynamic chemical imaging and cell acidification monitoring

      2021, Sensors and Actuators, B: Chemical
      Citation Excerpt :

      Nonoptical sensors with electrical signal detection for noninvasive and long-term monitoring were applied to measure pH changes since protons, lactic acid or CO2 products of cellular metabolism are generated in glycolysis and respiration processes [1,6,7]. With the scaling of microfabrication technology, nonoptical sensors with dimensions down to the micrometer level, including microelectrode arrays (MEAs) [8,9], ion-sensitive field effect transistors (ISFETs) [6,7,10–12] and light-addressable potentiometric sensors (LAPSs) [13–15], have been developed, which could be applied in the biochemical measurement of living cells and tissues. All of them can achieve two-dimensional (2D) electrical measurements, but the resolution resulting from the dimensions and spacing of the sensing points of MEAs and ISFETs is still limited by the complexity of fabrication and high cost of arrays.

    View all citing articles on Scopus
    View full text