Impedance-based cellular assays for regenerative medicine

Abstract

Therapies based on regenerative techniques have the potential to radically improve healthcare in the coming years. As a result, there is an emerging need for non-destructive and label-free technologies to assess the quality of engineered tissues and cell-based products prior to their use in the clinic. In parallel, the emerging regenerative medicine industry that aims to produce stem cells and their progeny on a large scale will benefit from moving away from existing destructive biochemical assays towards data-driven automation and control at the industrial scale. Impedance-based cellular assays (IBCA) have emerged as an alternative approach to study stem-cell properties and cumulative studies, reviewed here, have shown their potential to monitor stem-cell renewal, differentiation and maturation. They offer a novel method to non-destructively assess and quality-control stem-cell cultures. In addition, when combined with in vitro disease models they provide complementary insights as label-free phenotypic assays. IBCA provide quantitative and very sensitive results that can easily be automated and up-scaled in multi-well format. When facing the emerging challenge of real-time monitoring of three-dimensional cell culture dielectric spectroscopy and electrical impedance tomography represent viable alternatives to two-dimensional impedance sensing.

This article is part of the theme issue ‘Designer human tissue: coming to a lab near you’.

1. Introduction

Regenerative medicine aims to re-establish normal function of cells, tissues or organs by cell therapy, tissue engineering or by stimulating endogenous repair. Therapies based on regenerative techniques have the potential to radically improve healthcare in the coming years. A successful translation to the clinical setting will be facilitated by the production of renewable and tunable tissues. These tissues could also address a need in developing stable and robust in vitro assays for pharmacological investigations.

(a). An emerging need for non-destructive label-free approaches in tissue engineering and regenerative medicine

There is an emerging need for non-destructive and label-free technologies to assess the quality of engineered tissue and cell-based products prior to their use in the clinic. Clinicians must be able to assess that the right cells in the right state are being transplanted without interfering with their therapeutic potential, and the integrity of a sample. In parallel, the emerging regenerative medicine industry that develop cell-based products for cell therapy will benefit from moving away from existing destructive biochemical assays to assess online the quality of their production.

The state of the art in imaging tissue engineering and regenerative medicine is based on a combination of different techniques, the majority of which are destructive endpoint tests, such as histology, scanning electron microscopy, fluorescence microscopy, immunohistochemistry and other biochemical assays. They require the use of staining agents and sample processing which should ideally be avoided to limit safety issues.

In parallel to their exploitation for clinical translation, stem-cell technologies have triggered a step-change in the development of human in vitro disease models. Animal-based disease models have been widely used for many years. However, many diseases are species-specific and animal models cannot fully reflect the human behaviour in these cases [1]. Hence, primary human cells are still the preferred cell type for physiologically relevant disease models. However, primary cells are limited, expensive and difficult to obtain and culture [2]. Stem-cell-based disease models address several of these issues and are becoming an essential tool for investigating underlying mechanisms of diseases, and a vital platform for drug development and novel therapies [3].

(b). Impedance-based cellular assays as non-destructive label-free approaches

This review is focused on the recent emergence of impedance-based cellular assays (IBCA) in the field of tissue engineering and regenerative medicine. We define broadly IBCA as a range of methods using microelectrodes to measure the impedance of biological systems to gain information on the cellular behaviour of adherent cell cultures, cell suspensions and three-dimensional (3D) tissue models. Cell sorting and manipulation based on their dielectric properties, as in dielectrophoresis and impedance flow cytometry, is beyond the scope of this review.

IBCAs have emerged as an alternative approach to study stem-cell properties and cumulative studies, reviewed here, have shown their potential to monitor stem-cell renewal, differentiation and maturation. They offer a novel method to non-destructively assess and quality-control stem-cell cultures. In addition, when combined with in vitro disease models they provide complementary insights by allowing real-time monitoring of cell viability and measurement of cell–substrate and cell–cell adhesion parameters. IBCAs provide quantitative and very sensitive results that can easily be automated and up-scaled in multi-well formats.

2. An overview of impedance technology

Here we provide a brief overview of various techniques used in biology to measure impedance. Recent reviews have covered some aspects in more details [4,5].

(a). Passive electrical behaviour of cells

The Coulter counter is likely the most established impedance-based instrument in biology [6]. It measures the resistance—the ratio of measured voltage to input current between two electrodes—of biological cells flowing in narrow channels, to count cells and deduce cell volume.

Another impedance-based assay conventionally used in biology laboratories is the measurement of trans-epithelial electrical resistance (TEER) with chopstick electrodes, or epithelial volt/ohm metres, to estimate cell monolayer integrity [7]. The TEER across an epithelium or endothelium correlates with tight-junction formation between neighbouring cells.

The passive electrical behaviour of cells, as opposed to active electrophysiological potential, is mainly due to the presence of an insulating bi-lipid membrane that separates two ion-rich media. It creates an interfacial polarization under an applied field [8,9]. This results in cells being of an insulative nature at low frequency. Consequently, the impedance increases with increasing number of cells in a system. This response has been exploited to monitor yeast density in the brewing process [10] or biomass in large bioreactors [11–13].

Similarly, when cells are cultured directly on microelectrode systems (figure 1), for improved sensitivity, the impedance increases with cell coverage [14]. In addition, any changes in cell shape [15], spreading on surface [16] and membrane integrity [17] results in a change of single-cell dielectric properties and of the measured impedance [16]. Then, when live cells form layers or more complex tissues, the measured impedance depends on cell–substrate adhesion and cell–cell junction or barrier function [18,19]. Finally, collective cellular micromotion correlated to cell metabolism can be quantified with impedance sensing by analysing the fluctuations generated in the time-course impedance [20–22].

Figure 1.

(a) Schematic showing cells growing on a microelectrode with corresponding stages in impedance and (b) (I) proliferation stage, (II) confluency, (III) optional formation of barrier function and (IV) cell metabolic activity. (Online version in colour.)

(b). Electrical impedance spectroscopy

Electrical impedance spectroscopy (EIS) was applied to biological tissues as early as 1925 by Fricke & Morse [23]. When impedance of tissues is measured with increasing frequencies, a decrease by successive plateau is observed. There are four main transitions to lower plateaus [24,25], i.e. dielectric dispersions. The Beta-dispersion related to interfacial polarization across the cellular plasma membrane is the most informative when studying cell behaviour and occurs in the kilohertz range. Dielectric parameters of cell suspensions can be then retrieved by fitting the frequency-response to effective medium approximations such as the Maxwell–Wagner–Hanai, and are very well described in Asami [9]. However, the cell volume fraction should be sufficiently large to stand out from the highly conductive media contribution. A more detailed analysis is beyond the scope of this review and can be found in Heileman et al. [4]. In general, the impedance is acquired through the use of a four-electrode system to compensate for the double layer effect at low frequency, i.e. the accumulation of charged ions on the measuring electrodes. As mentioned above, EIS has been proved particularly useful to measure the biomass in bioreactors.

With the development of microelectrodes and microfluidics, the field has branched out into impedance flow cytometry where single cells flow in microchannels surrounded by microelectrodes. It has been shown to discriminate cells based on both cell size and intracellular dielectric properties [26,27].

(c). Electric cell–substrate impedance sensing

A major drawback of EIS is the overwhelming contribution of the cell medium to the impedance spectra. It can entirely mask the cell contribution for low cell to media volume ratio. In a pioneering work, Giaever, Keese and co-workers [14,18,28] have been able to free themselves from the contribution of the medium—improving considerably the sensitivity and ease of interpretation. Their technology, electric cell–substrate impedance sensing (ECIS) [15], is a real-time, label-free monitoring technology in which a small non-invasive AC current (approx. 1 µA) is applied through gold microelectrodes. ECIS originally used a two-electrode set-up with a very small sensing electrode (when compared to the counter electrode), onto which adherent cells were directly grown, providing very high sensitivity.

When cells are cultured on top of the microelectrodes (figure 1), they alter the current pathways due to the insulating properties of the cell plasma membranes. The measured impedance increases with cell growth until it reaches a plateau as the cells form a confluent monolayer on top of the electrodes. At low frequencies, the current is forced to flow under and in-between neighbouring cells, and the measured impedance in this case is directly related to the properties of cell–substrate adhesion and cell–cell tight junctions. Only at high frequencies can the current capacitively couple through the plasma membrane and the impedance give an insight into the integrity and dielectric properties of the cell membrane [16,29].

Since this pioneering work, impedance sensing has been extensively used for a wide range of biological assays including cell proliferation [16,30–32], cytotoxicity [33–37], wound healing [38–40], cell signalling [41–43], cell invasion [44] and blood–brain barrier permeability studies [19].

(d). Electrical impedance tomography for cellular assays

Electrical impedance tomography (EIT) was first developed in 1978 [45] and was mainly focused on clinical applications such as thorax imaging [46], brain function monitoring [47] and breast cancer screening [48,49]. EIT reconstructs the conductivity images of an object based on the voltages and the currents at the surface of the object [50].

The EIT measurement system mainly comprises three parts: a current source to stimulate the AC current into the subject, the multiplexer array for switching the current source and the data acquisition unit [50]. The first EIT system for medical research, the Sheffield Mk1, was developed in 1987 [51]. Now, the development of micro technologies has allowed the application of EIT at the cellular scale. One of the first EIT sensors for in vitro assay was developed by Linderholm et al. [52]. It was exemplified by studying cellular migration of human epithelial stem cells (YF 29). A sensor for a single organism, Physarum polycephalum, a slime mould growing on agar gel, was also demonstrated by Sun et al. [53]. These two studies paved the way for further research in the field of IBCA based on EIT.

3. Current progress of impedance-based cellular assays in tissue engineering and regenerative medicine

(a). Real-time label-free monitoring of cell differentiation

(i). Adult stem cells

Differentiation of stem cells is associated with a change in cell morphology, proliferation capacity and dielectric properties. Impedance sensing can detect these changes, defining distinctive impedance profiles for different differentiation paths. Cho & Thielecke [54] were the first to use impedance spectroscopy to study and characterize the growth of human mesenchymal stem cells (hMSCs). Then, an increase in resistance measurements associated with the differentiation of hMSCs towards osteoblasts [55] was reported, using a planar 1 mm diameter platinum electrode-based chip.

In parallel, independent studies have explored the ability to differentiate cell lineages arising from adult stem-cell sources with impedance sensing [56,57]. In our group, adipose-derived stem cells (ADSCs) were differentiated into osteoblasts and adipocytes and monitored throughout their differentiation (figure 2) [57]. We reported an increase in the impedance measurements with the osteo-induced ADSCs that was also accompanied by an increase in tightness of cell–cell junctions. However, the adipo-induced cells showed a drop in the impedance measurements and looseness in cell–cell junctions. A variation in the cell membrane capacitance between undifferentiated stem cell, osteo- and adipo-induced was also measured and pointed out as an early (less than 4 days) marker of stem-cell fate. Similar changes in dielectric properties accompanying stem-cell differentiation towards adipocytes was also reported by Lee et al. [58] and Fu et al. [59], which was related to lipid vacuole accumulation.

Figure 2.

Representative time-course impedance for ADSCs into two distinct lineages, i.e. osteogenesis (n = 3) and adipogenesis (n = 3), and a no cell control recorded with ECIS. Dotted arrows point at feeding time, plain arrow at induction time. (Online version in colour.)

In a similar work, Angstmann et al. [56] used two commercial systems, ECIS and xCELLigence, to examine differences in early phases of MSC differentiation towards adipocytes and osteoblasts, while Kramer et al. [60] have used xCELLigence to study adipogenesis of the preadipocyte 3T3-L1 cells. In agreement with the previous studies, osteo-induced cells showed an increase in impedance when compared to undifferentiated adult stem cells (ASCs), while adipo-induced cells had a marked decrease. Interestingly, they showed that impedance sensing can distinguish differentiation potential between low (p6) and high (p12) passage, paving the way to more recent studies looking at donor variability [61]. Striking differences were found in the impedance profile when comparing human ASCs isolated from different donors, opening the door to predicting osteogenic potential with impedance sensing with direct applications/implication for their translational potential.

The effect of coatings, collagen I, collagen IV, fibronectin and laminin on cell differentiation were also investigated [56,62]. Authors showed an increase of the impedance profile for osteo-lineage differentiated on collagen I and IV and a less pronounced drop in impedance for adipogenesis. These findings correlated well with the previous observation of increased osteogenesis and decreased adipogenesis on collagen coating. Other studies used the xCELLigence system in the context of regenerative medicine as a quantitative measurement of cell adhesion, e.g. for human endometrial MSCs [62], and proliferation of adipose-derived mesenchymal stem cells from ovariectomized mice [59], vascular smooth muscle cells from skin-derived precursors of human Wharton's jelly stem cells [63], and to assess bone marrow-derived stem cells in the context of good manufacturing practice [64]. Reitinger et al. [65] used a radio frequency identification-based sensor platform to monitor impedance wirelessly and in a reusable manner. They found similar impedance profile when differentiating ADSCs to osteoblasts and adipocytes.

The effect of neural differentiation media on MSC growth and differentiation have also been investigated with ECIS [66]. The authors reported a slower increase in impedance, due to lower cell number and not morphological changes, for the differentiated cells that corresponded to the arrest of the cell cycle induced by the differentiation.

(ii). Pluripotent stem cells

Impedance sensing with microelectrodes is not always well suited for monitoring embryonic stem cell or human induced pluripotent stem cell (hIPSC) renewal. They grow in patchy compact cell colonies which expand, and that may not be uniformly sampled by the microelectrodes. Similarly, the use of a feeder layer is clearly not compatible with impedance sensing, although this can be avoided. Recent advances in media and substrate formulation have allowed embryonic cell growth in a single layer and our group has been the first to monitor mouse embryonic stem-cell renewal on ECIS [67]. Impedance sensing was useful to demonstrate in a quantifiable way that the leukaemia inhibitory factor (LIF), a soluble growth factor necessary to the maintenance of pluripotency, could be replaced with LIF encapsulated in a hydrogel-based liposomal system or in poly(lactide-co-glycolic acid) polyester nanoparticles without being detrimental to cell growth.

The human cell line NT2 D1, an embryonic carcinoma cell line, was cultured on ECIS multi-well arrays and differentiated by retinoic acid (RA) into the neural lineage [68]. This study showed that both the impedance and the slope of the impedance increased in a dose-dependent manner with RA addition. The efficiency(state) of differentiation was then quantified by qRT-PCR expression as a decrease in stem-cell factors OCT4 and Nanog and an increase in differentiation markers HOXA1 and SNAP 25, and was found to be correlated in a dose-dependent manner to RA addition. Testing a panel of differentiation-inducing drugs, the authors were able to analyse both the differentiation and the cytotoxicity of the candidates. This study paved the way to using impedance sensing as a screening platform to study molecules that induce differentiation in a quantitative way.

IPSC self-renewal and differentiation to the mesendodermal and ectodermal lineage were recorded with an ad hoc device combining impedance sensing and a quartz microbalance [69]. Impedance was measured between 0.1 Hz and 10 kHz up to 96 h, and an equivalent circuit model, taking into account the gel layer, was used to derive the resistance and capacitance of the cell layer. The authors clearly showed a distinct impedance time course for each lineage that significantly correlated with morphological changes.

Neural stem cells were shown to have differences in time-course capacitance when cultured on interdigitated capacitance sensors and cultured under different conditions [70]. Proliferation and differentiation into either neuronal or astroglial cells could be differentiated in real time. However, neural network formation and the full maturation process were only reported with a biochip layout optimized for monitoring in real-time neural pluripotent stem-cell differentiation by Seidel et al. [71]. The potential of their platform, amenable to a 96-well plate format, for quality control in industrial processes was clearly demonstrated using the gamma secretase inhibitor DAPT to modulate the differentiation process. The same group developed a 384 multi-well microelectrode-array version [72] by solving a key challenge in multiplexing counter electrodes, opening the door to automated screening for pharmaceutical investigations. This was exemplified by establishing the dose- and time-dependent therapeutic effects of a kinase inhibitor (SRN-003-556) on a Sh-SY5SY tauopathy model.

(b). Impedance-based cell phenotypic assays for stem-cell-based in vitro disease models

Generally, label-free assays are promising tools for drug discovery [73–75]. They monitor drug–cell interactions in real time and in living cells. They are particularly well suited to assess drug polypharmacology as they are not based on a particular molecular marker [75]. These assays can quantify whole cell integrated responses which encompass the full complexity of drug–target interactions. This convoluted time-course signal can then be deciphered to determine the mechanism of action of drugs by acquiring the profile of an individual compound from a library. The most prominent label-free approach to assess drug polypharmacology is currently the optical measurement of dynamic mass redistribution with nanogratings and now impedance sensing is rapidly emerging as an alternative method [43]. They are both sensitive to cell density, cell–substrate adhesion and morphological changes.

Abassi et al. [76] led the way in 2012 by demonstrating an impedance sensing system that monitored dynamically the beating periodicity of stem-cell-derived cardiomyocytes. They obtained a dose–response profile for over 60 compounds, assessing simultaneously and non-invasively the periodicity of beating, contractility and overall viability of stem-cell-derived cardiomyocytes. Similarly, Guo et al. [77] used the same system to assess 28 compounds with known cardiac effects on IPSC-derived cardiomyocytes. These two studies, based on the equipment commercialized by Roche and now by ACEA, paved the way for a host of research studies combining impedance sensing and stem-cell-based cardiomyocytes (figure 3) which were reviewed by Peters et al. [78]. Indeed, impedance sensing at high acquisition frequency lended itself very well to the field, yielding quantitative parameters without any post-processing steps.

Figure 3.

(a) IPSCs generated from a patient can be differentiated into cardiomyocytes directly on top of gold microelectrodes. The spontaneous beating, an alternation of contracted (b) and relaxed states (c), results in cycles in the impedance (d). Changes in amplitude and frequency of the cycles can easily be quantified, e.g. for a compound inducing a chronotropic effect (e). (Online version in colour.)

IPS-derived retinal pigment epithelium cells generated from a patient with inherited macular degeneration and from an unaffected sibling were cultured on ECIS [79]. Retinal pigment epithelium (RPE) maturation was monitored in real time for greater than 25 days before reproducible and spatially controlled RPE layer damage was induced by an elevated current pulse to mimic cell loss in age-related macular degeneration. Migration rates between the two cell lines were then studied, and this showed the potential of the platform to quantitatively assess patient-specific RPE cell repair and to screen therapeutic compounds.

4. Conclusion and perspectives

(a). The challenge of monitoring three-dimensional cultures

Increasingly, cells are cultured in 3D environments where they exhibit cellular physiological functions closer to in vivo [80,81]. The regenerative field is increasingly investigating cells in 3D cultures such as organoids, multicellular spheroids, artificial and organotypic tissues. They offer better in vitro models to improve our understanding of cell biology and are regarded as a step-change in drug discovery. They are also developed as potential candidates for replacement and repair therapies.

As pointed out earlier, the state of the art in imaging tissue engineering and regenerative medicine is mostly based on destructive endpoint tests. Optical technologies are progressing rapidly to fill the gap [82], but generally require extra computer-intensive steps to retrieve quantitative parameters.

In this context, IBCAs present themselves as a viable complementary technique to assess cell behaviour in 3D cell culture. Although only few applications have been demonstrated so far, monitoring 3D cell culture is the main perspective for the IBCA field.

There are some technical challenges to translate impedance sensing directly to 3D cell monitoring as the microelectrode needs to be in contact with the sample. However, the application of impedance spectroscopy and EIT has been more straightforward.

Arrays of microcavities have been designed by the Robitzki group to host 3D cardiomyocyte clusters derived from embryonic stem cells and assess them in real time with EIS and field potential measurements [83]. Both chronotropic and action potential duration prolongation effects have been detected with this system. Considering the current trend in the regenerative field this platform could have a great potential to assess stem-cell-based organoids and spheroids for pharmaceutical investigations.

Cell proliferation and differentiation have been investigated with EIS with a dielectric probe that limited the use of the low frequency field [84–86] but was able to monitor large-scale samples. The sensitivity was increased by Daoud et al. [87] by using ad hoc macrochambers lined up with parallel-plate platinum electrodes; they enabled impedance measurements in the beta-dispersion frequency range. The epithelial differentiation processes of Madin–Darby canine kidney cells to hollow cyst-like structures were captured with this technique. The cells were embedded in collagen gels and hosted into 3D printed poly (dl-lactide-co-glycolide) acid scaffolds demonstrating the potential of IS to non-destructively monitor tissue formation in complex tissue engineered products. Recently, vertically aligned pairs of microelectrodes were used to monitor MSC migration in alginate gels. This technology was also demonstrated to measure cell proliferation and death [88].

Large production of stem cells at an industrial and clinical scale will likely lead to the use of bioreactors [89]. EIS is already used in bioreactors [11–13] and will, therefore, represent a viable strategy to monitor stem-cell growth and differentiation state when culturing stem-cell progeny. Mesenchymal stem-cell expansion on microcarriers cultured in bioreactors has been monitored successfully with EIS [90].

Although, impedance 3D imaging could lend itself very well to monitoring 3D cell culture, only few applications have been reported so far. Yang et al. [91] carried out 3D impedance imaging of breast cancer cell spheroids based on 3D-Laplacian and sparsity joint regularization algorithms. Overall, EIT is a relatively novel impedance-based measurement technique which has the potential to monitor tissue engineering products and in vitro disease models in real time with a high temporal resolution. However, further work, in both sensor and algorithm design, needs to be conducted to increase its sensitivity.

(b). Conclusion

IBCA have firmly established themselves in the field of biology as complementary assays providing time-course quantitative values of cell adhesion, cell–cell junctions and proliferation. With the recent demonstration of their ability to monitor stem-cell differentiation (electronic supplementary material, table 1), and as label-free phenotyping assays, IBCAs are becoming attractive for regenerative medicine applications. They could be integrated online to provide data-driven operation and quality control of large-scale culture, and into bench-top equipment to assess the quality of cell therapies products before their use in patients.

Data accessibility

This article has no additional data.

Competing interests

We declare we have no competing interests.

Funding

We received no funding for this study.