Continuous differential impedance spectroscopy of single cells
- 1.9k Downloads
A device for continuous differential impedance analysis of single cells held by a hydrodynamic cell trapping is presented. Measurements are accomplished by recording the current from two closely-situated electrode pairs, one empty (reference) and one containing a cell. We demonstrate time-dependent measurement of single cell impedance produced in response to dynamic chemical perturbations. First, the system is used to assay the response of HeLa cells to the effects of the surfactant Tween, which reduces the impedance of the trapped cells in a concentration dependent way and is interpreted as gradual lysis of the cell membrane. Second, the effects of the bacterial pore-forming toxin, Streptolysin-O are measured: a transient exponential decay in the impedance is recorded as the cell membrane becomes increasingly permeable. The decay time constant is inversely proportional to toxin concentration (482, 150, and 30 s for 0.1, 1, and 10 kU/ml, respectively).
KeywordsImpedance spectroscopy Single-cell analysis Microfluidic Microfabrication
Single-cell analysis and their response to chemical and biologic stimuli can provide unique insights into cellular behavior, including dynamics and localization of cellular processes. However, due to heterogeneous behavior among a population, the varied behavior is often difficult to interpret using quantitative models. Current high-throughput methods often lack the ability to track dynamic processes occurring in space and time (Di Carlo et al. 2006a, b).
Single cells can be used as sensors for detecting chemical and biological toxins or mutagens, and are also widely used for screening pharmacologically active compounds. Cell analysis systems have applications in drug discovery, bio-threat detection, and environmental pollutant identification (Asphahani and Zhang 2007). Most biosensors require analysis of a label attached to a molecule. The amount of label is measured and correlated to the number of bound targets. Labels can be fluorophores, magnetic beads, radioactive tags, enzymes that produce an easily detectable product (optical or electrochemical), or nanoparticles.
Impedance spectroscopy is a noninvasive, label-free analytical method that can provide information on the response of cells to their environment. This technique is used in many areas including the analysis of fluids (Nwankwo and Durning 1999), polymers (Fernandez-Sanchez et al. 2005), ion cells (Liaw et al. 2004), batteries (Xia et al. 1997), corrosion phenomena (Walter 1986), electrochemical kinetics/mechanisms, and biologic systems at the tissue (Kedar et al. 1994) and whole-organism level (Kushner and Schoeller 1986). Impedance spectroscopy has been used to measure the passive electrical properties of biologic cells for many years, both in bulk suspensions (Bordi et al. 2002; Markx et al. 1991; Davey et al. 1996; Schwan 1993; Merla and D’inzeo 2006; Fuhr et al. 1994; Lisin et al. 1996) and on substrates (Luong et al. 2004; Xiao et al. 2002a, b; Ceriotti et al. 2007a, b; Ciambrone 2004; Linderholm et al. 2006; Wegener et al. 1999, 2002). Traditionally, impedance measurements have been performed on suspensions of cells (Asami 1996; Gheorghiu and Asami 1998), but this method is insensitive to rare events and leads to temporal averaging: fast, time-dependent transitions occurring at the single-cell level may spread slowly through the population and could be interpreted as a gradual change (Di Carlo et al. 2006a). Experimental platforms that analyze large number of individual cells overcome this problem, whereby any heterogeneity within a cell population is measured (Lidstrom and Meldrum 2003).
The patch-clamp technique is a powerful method for the measurement of electrophysiologic activity of single cells; however, it is a labor-intensive method that conventionally requires a micromanipulator, micropipettes, and a skilled operator. To this end, high-throughput single-cell patch-clamp analysis systems are being developed within microfluidic lab-on-a-chip devices (Ionescu-Zanetti et al. 2005; Seo et al. 2004), with great promise of higher-throughput systems (Chen and Folch 2006). However, forming effective (i.e., giga-ohm) seals on arrays of cells is still problematic, the technique is invasive as the cellular membrane is intentionally disrupted, and chip fabrication is often complicated or requires exceptionally expensive equipment or materials. In this article, we describe a device for performing continuous differential impedance analysis of single cells that are hydrodynamically captured and held in traps within a microfluidic channel without the use of labels.
The design of the trapping structures ensures that the reference trap remains empty because it faces downstream, as shown in a top view micrograph in Fig. 1c and in a cross-sectional diagram in Fig. 1d. Figure 1c shows the microfluidic channel with six sets of traps, and the electrode pairs, one sample and one reference. Also shown in Fig. 1e and f are examples of a trapped fluorescent cell-sized beads and HeLa cell labeled with Celltracker dye, respectively.
As “proof of concept,” individual HeLa cells were trapped and continuously monitored over an extended time period during exposure to Tween and Streptolysin-O (SLO). Tween is a surfactant that is often used to lyse mammalian cells by compromising the cellular membrane. SLO is a pore-forming bacterial toxin classified as a member of the cholesterol-dependent cytolysin family, a large group of proteins that attack cholesterol containing membranes to form ring-shaped pores that mediate cell death (Tilley et al. 2005; Tilley and Saibil 2006).
The device here enables measurement of toxin activity at the single cell level in a noninvasive and label-free manner. The impedance of captured single cells perfused with solutions of SLO was measured, and the effect on the electrical conductivity of the cell membrane extrapolated to determine the effect of pore formation.
The chip was held on a PCB with spring-loaded connectors (SAMTEC SEI series) to contact the electrode pads and connected to a pump and impedance analyzer as shown in Fig. 1a. Cells were observed with an upright microscope. Impedance signals were acquired using two 8-way integrated multiplexers (ADG608, Analog Devices), controlled by MATLAB (The Mathworks, Natick, MA) via a USB interface. The impedance of each trap was measured using an impedance analyzer (Novocontrol Alpha-N) controlled by MATLAB via an IEEE-488 interface. Microfluidic ports were made by punching small casts of PDMS and plasma bonding the ports to the upper glass surface. HeLa cells were cultured in Dulbecco’s Modified Eagle’s Medium (Sigma) with 5% fetal calf serum (Gibco) and 100 μg/ml Penicillin/Streptomycin (Gibco). Cells were harvested and resuspended in Phosphate Buffer Saline (PBS).
Tween solutions were prepared by diluting Polyoxyethylenesorbitanmonolaurate (Sigma) in PBS at concentrations ranging from 0.01 to 1% w/w. SLO toxin (Sigma) was prepared at concentrations of 100, 10, and 100 U/ml in PBS and mixed with 100 mM dithiothreitol (DTT). DTT activates the toxin by creating a reducing environment for cysteine residues.
2.3 FEM simulations
Simulations were performed using both three-dimensional and two-dimensional axi-symmetric models, as shown in Fig. 3II). The system is not strictly axi-symmetric; however, the two-dimensional model provides good approximations (not shown) to the significantly more computationally intensive three-dimensional simulations. Changes in impedance spectra due to variations in cell membrane conductance, cell size, and position in the trap were simulated.
Figure 3 shows the electric field and potential for traps with and without the SU8 boundary as a function of frequency. The field simulations demonstrate that this design of SU8 structure confines the electric field to the region where the cell sits, increasing the effective volume fraction and maximizing the sensitivity of the measurement. This design is relatively insensitive to the position of the trapped cell: variations in the position of the cell within the channel by up to ± 12 μm (for a channel height of 25 μm) result in a maximum change in the magnitude of the impedance of 3% at 100 kHz.
In order to estimate the sensitivity of the system to changes in cell parameters, simulations were performed for different values of cell membrane conductivity and cell dimensions. At frequencies lower than 50 kHz, any changes are masked by the double layer. Although changes in cell size cannot be differentiated from changes in cell membrane conductivity, the simulations showed that a change of 0.1 mS/m in membrane conductivity results in a 5% change in the magnitude of the impedance, at 100 kHz.
The conductivity of SLO pores has not been reported (e.g., from patch clamp). However, measurements of Perfringolysin-O (PFO), also a member of the cholesterol-dependent cytolysin family of toxins, demonstrate a single pore conductance of 4.5 nS (Shepard et al. 2000). This is equal to a single pore conductivity of 31.8 nS/m for the SLO pores, which have diameters as large as 30 nm and span membranes 5 nm thick (Bhakdi et al. 1985; Alouf and Geoffroy 1988; Bhakdi et al. 1984). Multiple pores increase the membrane conductivity linearly. However, the effect on the impedance of the cell is not simple. FEM simulations were used with analytical calculations to estimate the number of open pores from the measured impedance response. The simulation results (shown in Supplementary Fig. 3) illustrate that the impedance response is quite insensitive to the insertion of a small number of pores, but can be used to quantify the effect of many thousands of pores, with some confidence, assuming that other cellular parameters (such as cell shape and size) do not change. Simulations suggest that measurable changes (1%) in the impedance magnitude occur for approximately 1,000 pores, when the decrease is in the range of 2–9%.
3 Results and discussions
In a typical experiment, the impedance spectrum was continuously recorded over a frequency range from 100 Hz to 2 MHz. The signal was multiplexed from eight active trapping site and the eight reference electrodes.
Exponential decay time constants are fitted to the data for the impedance response of single cells to SLO toxin
Time constant (s)
Interestingly, it appears that although the insert rate was significantly faster for higher concentrations (10 kU/ml in 20 s) than for the lower concentration (100 U/ml in 500 s), the final number of pores inserted into the membrane is of the range 10,000–15,000 regardless of the toxin concentration in the bathing solution.
A single-cell recording device has been designed, fabricated, and used to noninvasively quantify the effect of a surfactant and a pore-forming toxin on captured cells. The platform allows multiplexed recording of continuous differential impedance spectra from individual cells held in an array of hydrodynamic traps. The system was used to assay the transient response of HeLa cells to the lysing effects of the surfactant Tween and the kinetic pore-forming effect of SLO. Tween was found to change the impedance of trapped cells, with the change correlating with concentration. Perfusion with SLO toxin caused an exponential decay in the impedance with time constants inversely proportional to toxin concentration. The combination of single hydrodynamic cell trapping with single cell impedance analysis provides a scalable label-free cell analysis system. The detection limit of the platform was determined to be between approximately 1,000 pores. Although this is a much lower sensitivity of that afforded by patch-clamp techniques, the method is quick and noninvasive. Therefore, there is the potential to create vast two-dimensional arrays of single-cell traps, each individually addressable to create an automated platform for cell screening. Further developments include the capability to electroporate or electrically lyse single cells using DC potentials, either through the direct action of a high local electric field or by generating a localized hydroxide-rich environment, which disrupts the cellular membrane (Nevill et al. 2007).
The authors would like to acknowledge the support from a UK IRC in Bio-nanotechnology (DM), National Science Foundation (Career Award: Biomolecular Nanoelectronic Junctions), National Institutes of Health (NIH) Nanomedicine Development Centers Funding (JTN), NDSEG Fellowship (JTN), and GlaxoSmithKline for financial contributions, Dino Di Carlo for help, and the UC Berkeley Microlab and the University of Southampton ORC for clean room facilities.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
- Alouf JE, Geoffroy C (1988) Production, purification and assay of Streptolysin O. Method Enzymol 165:52–59Google Scholar
- Asami K (1996) Dielectric spectroscopy of biological cells. Bioelectrochem Bioenerg 40:141–145Google Scholar
- Asphahani F, Zhang M (2007) Cellular impedance biosensors for drug screening and toxin detection. Analyst 132(9):835–841Google Scholar
- Bhakdi S, Roth M, Sziegoleit A, Tranum-Jensen J (1984) Isolation and identification of two hemolytic forms of streptolysin-O. Infect Immun 46(2):394–400Google Scholar
- Bhakdi S, Tranum-Jensen J, Sziegoleit A (1985) Mechanism of membrane damage by Streptolysin-O. Infect Immun 47(1):52–60Google Scholar
- Bordi F, Cametti C, Gili T (2002) Dielectric spectroscopy of erythrocyte cell suspensions. A comparison between Looyenga and Maxwell. J Non-Cryst Solids 305:278–284Google Scholar
- Ceriotti L, Ponti J, Broggi F, Kob A, Drechsler S, Thedinga E, Colpo P, Sabbioni E, Ehret R, Rossi F (2007a) Real-time assessment of cytotoxicity by impedance measurement on a 96-well plate. Sens Actuators B 123:769–778Google Scholar
- Cho S, Thielecke H (2007) Micro hole-based cell chip with impedance spectroscopy. Biosens Bioelectron 22(8):1764–1768Google Scholar
- Davey HM, Davey CL, Woodward AM, Edmonds AN, Lee AW, Kell DB (1996) Oscillatory, stochastic and chaotic growth rate fluctuations in permittistatically controlled yeast cultures. Biosystems 39(1):43–61Google Scholar
- Di Carlo D, Aghdam N, Lee LP (2006a) Single-cell enzyme concentrations, kinetics, and inhibition analysis using high-density hydrodynamic cell isolation arrays. Anal Chem 78(14):4925–4930Google Scholar
- Di Carlo D, Wu LY, Lee LP (2006b) Dynamic single cell culture array. Lab Chip 6(11):1445–1449Google Scholar
- Fernandez-Sanchez C, Mcneil CJ, Rawson K (2005) Electrochemical impedance spectroscopy studies of polymer degradation: application to biosensor development. TRAC-Trend Anal Chem 24(1):37–48Google Scholar
- Fuhr G, Glasser H, Muller T, Schnelle T (1994) Cell manipulation and cultivation under AC electric-field influence in highly conductive culture media. Biochimica et Biophysica Acta 1201(3):353–360Google Scholar
- Gheorghiu E, Asami K (1998) Monitoring cell cycle by impedance spectroscopy: experimental and theoretical aspects. Bioelectrochem Bioenerg 45:139–143Google Scholar
- Jang LS, Wang MH (2007) Microfluidic device for cell capture and impedance measurement. Biomed Microdevices 9(5):737–743Google Scholar
- Kedar RP, Bourne TH, Powles TJ, Collins WP, Ashley SE, Cosgrove DO, Campbell S (1994) Effects of tamoxifen on uterus and ovaries of postmenopausal women in a randomized breast-cancer prevention trial. Lancet 343(8909):1318–1321Google Scholar
- Kushner RF, Schoeller DA (1986) Estimation of total-body water by bioelectrical impedance analysis. Am J Clin Nutr 44:417–424Google Scholar
- Liaw BY, Nagasubramanian G, Jungst RG, Doughty DH (2004) Modeling of lithium ion cells—a simple equivalent-circuit model approach. Solid State Ionics 175:835–839Google Scholar
- Lisin R, Ginzburg BZ, Schlesinger M, Feldman Y (1996) Time domain dielectric spectroscopy study of human cells. 1. Erythrocytes and ghosts. Biochim Biophys Acta 1280(1):34–40Google Scholar
- Luong JHT, Xiao C, Lachance B, Leabu SM, Li X, Al E (2004) Extended applications of electric cell-substrate impedance sensing for assessment of the structure. Anal Chim Acta 501:61–69Google Scholar
- Markx GH, Tenhopen HJG, Meijer JJ, Vinke KL (1991) Dielectric-spectroscopy as a novel and convenient tool for the study of the shear sensitivity of plant-cells in suspension-Culture. J Biotechnol 19:145–158Google Scholar
- Merla C, D’inzeo G (2006) Dielectric spectroscopy of blood cells suspensions: study on geometrical structure of biological cells. Proceedings of the 28th IEEE EMBS annual international conferenceGoogle Scholar
- Morgan H, Green NG (2003) AC electrokinetics: colloids and nanoparticles. Research Studies Press, BaldockGoogle Scholar
- Nevill JT, Cooper R, Dueck M, Breslauer DN, Lee LP (2007) Integrated microfluidic cell culture and lysis on a chip. Lab Chip 7(12):1689–1695Google Scholar
- Nwankwo E, Durning CJ (1999) Fluid property investigation by impedance characterization of quartz crystal resonators—Part I: methodology, crystal screening, and Newtonian fluids. Sens Actuators A 72:99–109Google Scholar
- Schwan HP (1993) Mechanisms responsible for electrical-properties of tissues and cell-suspensions. Med Prog Technol 19(4):163–165Google Scholar
- Walter GW (1986) A review of impedance plot methods used for corrosion performance analysis of painted metals. Corros Sci 26(9):681–703Google Scholar
- Wegener J, Keese CR, Giaever I (2002) Recovery of adherent cells after in situ electroporation monitored electrically. Biotechniques 33(2):348, 350, 352 passimGoogle Scholar
- Xia YY, Zhou YH, Yoshio M (1997) Capacity fading on cycling of 4V Li/LiMn2O4 cells. J Electrochem Soc 144(8):2593–2560Google Scholar