Analysing single live cells by scanning electrochemical microscopy

Fraser P. Filice and Zhifeng Ding *
Department of Chemistry, The University of Western Ontario, 1151 Richmond Street, London, Ontario N6A 5B7, Canada. E-mail: zfding@uwo.ca

Received 3rd August 2018 , Accepted 16th September 2018

First published on 17th September 2018


Single live cell analysis methods provide information on the characteristics of individual cells, yielding not only bulk population averages but also their heterogeneity. Scanning electrochemical microscopy (SECM) offers single live cell activities along its topography with high accuracy probe tip positioning. Both intracellular and extracellular processes can be electrochemically examined through the use of SECM. This non-invasive technique allows for high resolution mapping of electrochemical measurements in or around the cell sample of interest. Reactive oxygen species and reactive nitrogen species can be determined in a non-invasive label-free method and utilized as a probe for cellular pathology and physiology. Membrane permeability and rate of membrane species transport can be quantified in SECM. The cell response to external stressors can be monitored and modelled. SECM is able to offer nanoscale mapping and low concentration detection, providing a powerful bioanalytical tool for live cell studies. Herein we present an overview of recent progress in the imaging and characterization of single live cells using SECM.


image file: c8an01490f-p1.tif

Fraser P. Filice

Fraser P. Filice received his B.Sc. from the University of Western Ontario in 2014 studying Chemistry and Computer Science. He is currently working as a PhD candidate in Dr Zhifeng Ding's Lab at the University of Western Ontario. His expertise is in scanning electrochemical microscopy and related 3D finite element simulations.

image file: c8an01490f-p2.tif

Zhifeng Ding

Zhifeng Ding obtained his degrees of BSc at Southeast University, MSc at Nanjing University and PhD at EPFL. After his postdoc research position at the University of Texas at Austin, he joined the University of Western Ontario as a faculty member in 2002, where he has been promoted to full professor. His research focuses on scanning electrochemical microscopy, electrochemiluminescence/electroluminescence, solar cells, and room-temperature ionic liquid electrochemistry.


1. Introduction

Numerous analytical tools have been developed for the analysis of live cells and their functions. However, the majority of these focus on analyzing the cell bulk population. While they allow for rapid characterization, these techniques often are incapable of detecting the heterogeneity of the cells under study and giving chemical reactivity information. Single cell techniques facilitate meeting the research demands on an individual basis. Among these, scanning electrochemical microscopy (SECM) has attracted great research attention in recent years.

In fact, SECM is a powerful non-invasive analytical technique of the scanning probe microscopy family.1–3 This technique utilizes a small diameter (≤25 μm) electrochemical probe to image a region of space with extreme precision. Faradaic current is recorded with the electrode position to reconstruct detailed maps of topography and sample reactivity. SECM utilizes an electrochemically active redox species, either introduced into the system or intrinsic to the sample under study. This electrochemical mediator is either oxidized or reduced at the electrode tip in a diffusion-controlled process yielding a steady state current.

The SECM concept was introduced in 1986 by Bard et al.4 and Engstrom et al.5 and showed great promise as a new electroanalytical technique. Significant early SECM development and experimentation was performed by Bard et al. in 1989, which laid the groundwork for its application to a diverse range of research fields.6,7 SECM has since been employed in a vast number of applications,8 including but not limited to kinetics studies,9,10 surface and interface studies,11–13 and microstructure fabrication.14,15 SECM has also found applications as a bioanalytical tool for cellular imaging,16–25 membrane transport,17,26–32 neurotransmitter release,33,34 multidrug resistance studies,18,19 reactive oxygen species (ROS) and reactive nitrogen species (RNS) mapping,24,35–40 and cellular redox processes.22,32,41–45

SECM has also been employed in combination with other complementary bioanalytical tools. These include SECM–AFM,46 SECM–NSOM,47,48 SECM–SICM,49 SECM-Raman imaging,50 SECM-fluorescence microscopy,51 SECM-microfluidic systems,52 and many more. SECM is also commonly used in association with finite element method (FEM) simulations to quantify additional physical characteristics of the sample under study.53–56

The focus of this review will be on the biological applications of SECM for the study of single live cells. Cell related SECM advances such as instrumentation, reactive oxygen and nitrogen species, membrane transport, heavy metal stress, finite element simulations and nanoscale images will be highlighted. For a survey of the general applications of SECM, please see the 2016 review “Scanning Electrochemical Microscopy: A Comprehensive Review of Experimental Parameters from 1989 to 2015” by Mauzeroll's group at McGill University.8

2. Analytical and instrumental methodology

A typical SECM workstation is composed of a low current electrochemical potentiostat, a high-resolution positioning system, and a small diameter SECM probe (Fig. 1). SECM workstations now often include an optical imaging system to aid in cell and electrode positioning and relocation.20 The electrochemical system is normally a low current potentiostat. Current response is dependent on probe size and the concentration of species being monitored. As a result, the requirements for the selected potentiostat can range from low nanoamps (nA) down to femtoamps (fA). Lower current systems are also susceptible to electrical noise from external sources. It is important to consider instrument grounding and isolation, instrument shielding (faraday cage), the quality of shielded analog cabling, and physical isolation for vibration.
image file: c8an01490f-f1.tif
Fig. 1 Typical setup of a bio-SECM workstation: the cell is mounted on a stage below the ultramicroelectrode probe. The UME tip-to-sample position is controlled by an XYZ positioning system. The electrochemical cell consists of a 2 or 3 electrode system (WE, CE/RE or WE, CE, RE, respectively). An inverted objective microscope is often used for electrode positioning and optical imaging. Data output from the potentiostat, the positioning system, and the camera are logged simultaneously by a computer.

Accurate movement of the electrode tip relative to the cell is essential. These positioning systems are commonly piezoelectric positioners or high accuracy stepper motors. Many early designs of SECM positioning systems utilized open loop positioning systems, which rely on logging the number of steps and a known step size to determine the electrode position. These systems while being incredibly accurate are susceptible to minor drifts in electrode position depending on the tolerance specification for each step. Over the course of long duration scans, the accumulation of drifts from each individual step can cause significant deviations. Closed loop positioning systems as the common practice nowadays provide a solution to this problem, using active monitoring of the positioning35,57 and a feedback loop by which the compensation for probe drift can be applied.

The SECM probe is arguably the most important part of the instrument. SECM resolution is directly dependent on the size of the electrochemical probe. Various methods of SECM electrode fabrication have been developed including encasing a metal or other conductors in glass,58 laser pulling both metal and glass,59,60 chemical vapor deposition (CVD), pyrolytic carbon deposition,61 and many more specialized electrodes.55,62 Electrodes can also be modified in a molecule specific coating to create specialized sensors.63

A close relative of the SECM imaging method is Scanning Ion Conductance Microscopy (SICM).64 SICM imaging utilizes an electrolyte filled nanopipette in place of an SECM disk electrode. Whereas the resolution of an SECM comes from the critical radius of the UME disk, SICM yields a similar result through the radius of the nanopipette tip. SICM probes are often gravity pulled or laser pulled from glass capillaries and filled with the electrolyte when performing the experiment.

Various imaging modes for SECM probe movement exist. Simple 1D line scans such as the constant height scan in Fig. 2A fix the height of the electrode and scan horizontally across the sample. Constant distance imaging, however, moves the probe up and down as it passes over the substrate, in an attempt to keep the tip-to-cell distance constant.24,65 Various methods have been used to control for SECM tip-to-cell distance including SECM–AFM systems,46,66 shear force,67,68 impedance,24,65 faradaic current,69 and ion current.49,70 Each of these methods has its own advantages and disadvantages. Both the constant height and the many forms of constant distance scans provide excellent information on topography and variation in surface reactivity. Another simple 1D scan is the probe approach curve (PAC, Fig. 2B), where the electrode is moved linearly toward the substrate, providing detailed information about a specific point on the sample. PACs provide excellent tip-to-substrate information, and more detailed reactivity information at a specific location. As a result, PACs are more commonly employed for the detection of membrane permeability, or surface emission/consumption of species.


image file: c8an01490f-f2.tif
Fig. 2 (A) Horizontal 1D scanning methods for SECM imaging of a cell. (B) 1D electrode tip movement in the z direction to yield the probe approach curve. (C) Common 2D scanning methods across a cell. (D) Full 3D scanning method over a cell.

2D scanning modes such as depth scanning38 and surface mapping have the advantage of providing information on a much larger region of space over the cell (Fig. 2C). They are commonly composed of hundreds of constant height, constant distance, or PACs at fixed distances from each other. Depth scan mode operates in the xz or yz plane over a sample, where surface maps are generally taken as an xy map over the cell. A full 3D map over a cell provides information on a volume of space, which is composed of numerous surface maps and depth scans with a fixed spacing (Fig. 2D). These provide a significant amount of data and the ability to extract any needed scan in post processing. This scanning method does, however, come at the expense of long experiment run times.

The SECM hopping mode can be performed as 2D or 3D imaging.71 When scanned in 2D, the electrode is moved in a similar plane to the depth scanning mode. However, the order by which the two axes are scanned changes. Depth scanning performs a single horizontal line scan above the sample (x or y axis) before moving the electrode one step down (z axis) and repeating the line scan. In the hopping mode, a single PAC is performed (z axis) to the sample until a predefined feedback is detected at the probe tip. The electrode is then repositioned horizontally (x or y axis) before the PAC scan is repeated. Similarly, when performed as a 3D scan, a 2D array of PACs is performed over a sample, imaging a volume of space over the sample surface.

In 2010, Schuhmann's group reported the 4D shear force-based constant distance mode.67 Using a 4D shear force, multiple constant distance surface images are obtained above the sample. The tip-to-sample distance is determined using the probe shear force, and the tip is retracted to the desired distance.67 This generates a comprehensive 4D data set of the electrode tip current and shear force in a volume of space over the sample.

High precision electrochemical characterization of live cells by SECM can often carry significant costs. However, low-cost solutions to SECM imaging have been shown to be viable for basic uses. In 2017 it was shown that a SECM could be built from low cost consumer available parts and 3D printing for $100 (USD, UME probe not included).72 The low-cost SECM instrument utilized an Arduino UNO microcontroller for position control, signal generation for the potentiostat, and data logging. The potentiostat was custom built and included a summing amplifier, potential control circuitry, and a transimpedance amplifier (Fig. 3). Off the shelf stepper controllers and NEMA 17 stepper motors were utilized with 500 μm per revolution threaded rods. This provides a 2.5 μm step size for the positioning system. The SECM body was assembled from custom prepared 3D printed parts. The software was created in Sketch, and incorporated multiple scanning techniques, including line scans and hopping mode. This provides an excellent low-cost option for individuals and schools interested in SECM based sample analysis.


image file: c8an01490f-f3.tif
Fig. 3 Diagram of the electronic components of the EC-SPM setup showing the microcontroller board, the custom built potentiostat, and “off-the-shelf” stepper controller boards. Highlighted in the potentiostat circuit diagram are (A) the summing amplifier, (B) the potential control circuit and the electrochemical cell and (C) the transimpedance amplifier.72

3. Reactive oxygen species and reactive nitrogen species detection

Oxidative stress is an area of significant interest,73 as it is an important marker for carcinogenic cell behavior,74 in cell signalling and inflammatory response,75 and has been linked to long term genetic damage.76 Excessive concentrations of ROS or RNS can overwhelm a cell's antioxidant defenses, triggering apoptosis.77 Due to the electrochemical reactivity of both ROS and RNS species, a label free method of characterization of cell oxidative species production is possible with SECM. This technique provides a non-invasive, highly localized method for performing live cell analysis, allowing for species detection and mapping in the region of interest.78 Significant success has been obtained using this method to map inflammatory ROS production in the extracellular space over live cells.24,38,40 Combination SECM-fluorescence microscopy techniques have also been developed to simultaneously monitor extracellular and intracellular ROS.51

SECM has also shown promise for use in directly monitoring intracellular ROS species. Examination of intracellular ROS can be performed by utilizing small scale (<300 nm diameter) electrodes. Mirkin et al. showed the characterization of internal cell ROS and RNS species using 20, 150, and 300 nm Pt black electrodes.79 Production rates of O2˙ and NO˙ were monitored in two metastatic breast cancer cell lines (MDA-MB-231 and MDA-MB-468) as well as control human breast epithelial cells (MCF-10A). Probe approach curves, which included tip insertion into the cell interior, were examined with all studied electrode sizes. ROS and RNS production was also induced in healthy MCF-10A cells through the introduction of diacylglycerol-lactone (DAG-lactone), increasing the activation of protein kinase C. Following cell treatment with DAG-lactone and DMSO, no measurable ROS/RNS oxidation was detected for the first 25 minutes of study (Fig. 4B). Following this initial period, multiple sharp bursts of oxidative stress were observed electrochemically in the cell interior. This observation is in agreement with previous studies that show DAG-lactone induced PKC activation after 25 minutes.80


image file: c8an01490f-f4.tif
Fig. 4 Oxidative stress detection in an MCF-10A cell treated with DAG-lactone. (A) Optical images of two immobilized MCF-10A cells in PBS before (1 and 3) and 30 min after adding DAG-lactone dissolved in DMSO (2) or an equal volume of DMSO (0.1%, v/v) (4). (B) Current recordings at the 40 nm platinized tip inside an MCF-10A cell treated with DAG-lactone (green curve) or with DMSO (black curve). ET = 850 mV.79

4. Cell membrane transport

SECM provides highly localized characterization of electrochemical processes. By positioning an electrode in close proximity to a cell membrane, the transport of electrochemically active biological molecules can be quantified.53,81,82 External nontoxic electrochemical mediators can also be added to characterize the physical properties of a cell if no intrinsic redox source is present. This can allow for the detection of cell emissions, uptake of resources, and general membrane permeability (Fig. 5). Imaging can be performed over ion channels,30,83 or pores,35,84 or even used to measure the flux of the cell membrane.53,85
image file: c8an01490f-f5.tif
Fig. 5 Simplistic representation of (A) the control, (B) decreased, and (C) increased membrane permeability to FcMeOH via passive diffusion.86

Membrane permeability is often characterized on live cells by introducing an electrochemical mediator with partial permeability to the cell membrane.21,53,87 Alternatively, a mediator that is generally excluded from the cell can be used to determine severe damage to the live cell and serves as a good indicator of cell viability. In 2016, Li et al. showed the membrane permeability characterization of T24 live cells, stressed with Cd2+, with various redox mediators.88 Hydrophilic redox mediators such as ferrocene carboxylate (FcCOO), 1,1′-ferrocene dicarboxylate (Fc(COO)22−) and hexaamineruthenium(III) (Ru(NH3)63+) were examined by SECM depth scanning mode (Fig. 6A–C). PACs were extracted from these depth scans and membrane permeability was quantified with the aid of membrane permeability FEA simulations (Fig. 6D–F). The charged nature of these redox mediators normally makes them impermeable to the cell membrane. However, when exposed to sufficient Cd2+ (4.4 mM for 1 h) permeability was induced in the cells exposed to FcCOO and Fc(COO)22− (Fig. 6G and H, respectively). This concentration also corresponded to the beginning of a decrease in cell population viability, monitored by the MTT assay. This indicates significant damage to the cell membrane or nonspecific diffusion channels, permitting mediator flux. Ru(NH3)63+ was found to show an increase in membrane permeability much earlier, at concentrations as low as 2 mM Cd2+ (Fig. 6I). A time dependent examination of the hydrophobic mediator ferrocenemethanol (FcMeOH) was also performed, which is partially permeable to the cell membrane. This mediator shows an initial decrease in membrane permeability between Cd2+ injection (25 μM) and 2 h exposure. Membrane permeability reaches a minimum at this time, where it is one-fifth the value observed before injection. Following this, the membrane is observed to increase to the double the permeability (4 h) observed before injection and reaches a plateau at this value (6 h total study time).


image file: c8an01490f-f6.tif
Fig. 6 SECM depth scan images of T24 cells treated with 4.4 mM Cd2+ for 1 h detecting using: (A) FcCOO, (B) Fc(COO)22− and (C) Ru(NH3)63+. The scale bar in each image represents 10 μm. The arrows in (A)–(C) indicate where the cross section was taken to extract the PACs (D)–(F), respectively. Experimental PACs were overlaid onto simulated curves to quantify membrane permeability. Graphical change trends of the average membrane permeability after acute Cd2+ concentrations (G)–(I).88

In 2016, Bondarenko et al. investigated the effects of fixation and induced permeabilization on cells (Fig. 7A).89 Fixation is a process that halts biological activity in a cell while preserving cell proteins, carbohydrates, and other structures by cross-linking proteins inside the cell.89 It was found that cell topography scans (using SECM and AFM) showed an increase in cell height following fixation. Constant height SECM scans with a 25 μm UME were taken over a variety of different cell lines. Cells were patterned into clusters using a polyimide mask, and variations in cell concentration as well as UME scan speed were examined. Investigation of cell clusters using an FcMeOH (hydrophobic) electrochemical mediator showed a significant change in the current profile depending on the UME tip speed (Fig. 7B). Scans taken over fixed cells (formaldehyde) showed a less drastic change in the current profile and a simplified electrochemical signal (Fig. 7C). The UME tip speed was found to have a less significant effect over fixed cells as well. The SECM signal was found not to be drastically affected by the permeabilization (Triton X-100) of the cells (Fig. 7D). The use of the hydrophilic mediator FcCOOH showed a significant current decrease over both living and fixed cells and was minimally affected by the scan rate (Fig. 7E and F). Permeabilized cells were observed to have a significant decrease in current feedback, similar to that seen with FcMeOH (Fig. 7G). The proteins bovine serum albumin (BSA) and the intracellular melanoma biomarker tyrosinase (TyR) both contain electrochemically active amino acids. These proteins were examined in several cell lines by SECM and immunoassay staining.


image file: c8an01490f-f7.tif
Fig. 7 Schematic representation of alive, fixed, and permeabilized cells (a). Influence of the UME translation speed on the SECM response (normalized current) provided by alive (b and e), fixed (c and f), and permeabilized (d and g) adherent WM-115 melanoma cells in the presence of noncharged (FcMeOH, b–d) and charged (FcCOOH, e–g) redox mediators.89

In 2017, Soldà et al. showed the glucose uptake of single live cells using custom biosensor probes.90 Biosensor materials included the enzymes glucose oxidase (GOx) or lactate oxidase (LOx), which are sensitive to the presence of glucose and lactate, respectively. Various forms of application were investigated for the preparation of biosensor electrodes, including cross-linking, electropolymerization, and adsorption by physical and electrostatic interactions. It was found that for all biosensors, the sensitivity increases with coating thickness up to a maximum of 20 μm in thickness on 10 μm Pt UMEs. It was also determined that the electropolymerization methodology produced probes with ∼4 times the spatial resolution of the other techniques. This is due to the biosensor material being spatially confined to the Pt disk, while in other methods it spread across the glass sheathing of the UME in addition to the Pt. Once optimized, these biosensor electrodes were applied to the imaging of MCF10A human breast epithelial cells by SECM (Fig. 8A). 2D images of these cells were acquired, showing the current decrease over the surface of live cells, corresponding to the metabolic consumption of glucose by the cell (Fig. 8B). SECM voltage switching91 was employed to compare the topographic and electrochemical responses of glucose to the topography only using Ru(NH3)63+ as an electrochemical mediator (Fig. 8C). The average glucose uptake was determined to be 17 ± 3 μM using this method.


image file: c8an01490f-f8.tif
Fig. 8 (A) Optical image of the scanned MCF10A cells. Scale bar: 30 μm. (B) Representative SECM image of glucose uptake of several MCF10A cells. SECM experiments were carried out in the constant height mode using an electropolymerized 10 μm Pt GOx-UME biosensors in PBS buffer containing 0.1 mM glucose, 10 μM insulin, and 2.5% horse serum. Scan speed 15 μm s−1. E = +0.65 V vs. Ag/AgCl, 3 M KCl. Comparison between the two profiles: the black line contains both functional and topographical contributions, while the red line corresponds only to the topographical contribution. The currents were normalized by the respective steady-state bulk current values and subtracted by the normalized currents on the Petri dish. The measurements were performed with the same electropolymerized 10 μm Pt GOx-based UME biosensors in PBS buffer containing 0.1 mM glucose, 10 μM insulin, and 500 μM of [Ru(NH3)6]3+.90

In 2017 Polcari et al. showed the activity of HeLa and HeLa-R MRP1 activity via SECM.92 Cells were cultured on custom patterned Zeonor cell culture substrates. Oxygen plasma was used to treat the patterned spots, which promotes the attachment, proliferation, and growth of HeLa cells. The spot size was optimized to 50 μm, where they exhibit a typical morphology and no increase in stress. A dual redox mediator methodology was used, employing both FcMeOH and RuHex mediators. The FcMeOH travels into the cells, inducing an increase in glutathione (GSH) release from HeLa cells from the MRP1 pumps embedded in the membrane. RuHex, however, does not interact with expelled GSH, providing only information on topography. Using a combination of the FcMeOH and RuHex images, as well as an extracted map of the normalized tip-to-substrate distance, it was possible to deconvolute the heterogeneous rate constant profile (k). The k scan profile provides an indication of the MRP1 functional activity in the imaged cell. The anticancer drug doxorubicin was used to treat HeLa and HeLa-R cells, at 0.05 μM and 0.10 μM (1/10 and 1/5 of the LD50, respectively). Constant height SECM images of these cells were taken, and the k map was calculated (Fig. 9). Cells were also characterized by western blotting and flow cytometry, allowing for the quantification of the MRP1 expression factor in each of these studied cell populations. HeLa-R populations were observed to have 6.7× the expression of MRP1 than HeLa populations. However, a correlation between the expression factor and activity was not found.


image file: c8an01490f-f9.tif
Fig. 9 MRP1 functional activity of six different cell populations obtained through a doxorubicin drug challenge. Top panel: Optical micrographs of the cell populations during SECM imaging. Bottom panel: Extracted apparent heterogeneous rate constant profile. A single scale is used for all populations to visualize differences in contrast. Values are presented as 10−3 cm s−1. (Imaging conditions: a = 3.5 μm; v = 10 μm s−1.)92

5. Heavy metal exposure

Many heavy metals are present in nature or are common in industrial processes, which are toxic to humans if consumed, and some are necessary in trace quantities but are toxic at high concentration. Heavy metal toxicity has been shown to induce stress in live cells, and in high enough quantities activate apoptotic and necrotic pathways.93,94 They have been shown to affect membrane fluidity through oxidative damage and lipid peroxidation.95,96 Changes in membrane permeability induced by oxidative stress have been shown to be detectable by SECM, making it an excellent tool for localized analysis of membrane damage.21,53 SECM provides a non-destructive, non-invasive method of cell analysis. When applied to the imaging and characterization of live cells, this provides a method of repeated rapid characterization for the cell sample. As a result, SECM can be used to repeatedly image the same cell and provide time dependent analysis of cell homeostasis.

SECM with liquid–liquid interface probes has also been used to greatly detect uptake of heavy metal ions with complex redox activity, such as antimicrobial Ag+.97,98 In the case of these electrodes, an inorganic solution phase exists inside the electrode capillary, which interfaces with the water based aqueous phase containing live cells at the electrode tip. The transport of the Ag+ ion across the solution phase boundary is electrochemically detectable, allowing for the quantification of ion availability and the mapping of cell uptake.

In 2014, Li et al. demonstrated the time dependent characterization of single live cell membranes by SECM.17 Live cells were exposed to various concentrations (1 mM to 10 mM) of the toxic heavy metal Cd2+ and were monitored every 5 minutes following injection. Optical images of the cells were taken before injection (Fig. 10A), as well as after 50 minutes of exposure (Fig. 10B) for each concentration. The SECM depth scanning mode was used to image an xz plane of space over the center of the cell (Fig. 10C). Probe approach curves were extracted from the cell center position on the depth scans, and were matched with theoretically generated curves, allowing the quantification of cell membrane permeability (Fig. 10D). Exposing the cells to 1 mM of Cd2+ provides negligible changes in membrane permeability over the 45 minutes of SECM analysis. This concentration also yielded very little morphological change either optically or visible in the depth scan. As higher concentrations were examined, it was found that with higher concentration cells would contract vertically, requiring repositioning of the electrode. It was also determined that higher concentrations induced a gradual increase in membrane permeability up to a new maximum. At the highest tested concentration of 10 mM Cd2+ (Fig. 10D), an increase in membrane permeability from 2.5 × 10−5 m s−1 at t = 0 to 2.0 × 10−4 m s−1 at t = 45 min is observed.


image file: c8an01490f-f10.tif
Fig. 10 Optical micrographs of a T24 cell (A) before Cd2+ stimulation and (B) after 45 min of 10.0 mM Cd2+; (C) SECM depth scan image of a T24 cell at t = 15 min; (D) initial and final PACs overlaid onto theoretical ones.17

In 2016, live T24 human bladder cancer cells were incubated with a wide range of toxic Cd2+ concentrations.86 Incubations occurred for 1 h, indicating acute response to the Cd2+. SECM scans were performed using a 10 μm Pt electrode and a membrane permeable FcMeOH electrochemical mediator. Cells were examined using the SECM depth scanning mode, and PACs were extracted above the cell center. Quantification of cell membrane permeability was performed using 2D axially symmetric FEM simulations. Incubation with low concentrations of Cd2+ (1.0 μM–7.5 μM) showed no deviation from untreated control scans (Fig. 11A). Increasing the concentration (between 10 μM and 100 μM) yielded a decrease in membrane permeability (Fig. 11B). The minimum membrane permeability observed occurred at 75 μM and 100 μM, where the membrane permeability dropped to 25 μm s−1. Incubations with higher concentrations than this (250 μM–1000 μM), yielded an increase in membrane permeability (Fig. 11C).


image file: c8an01490f-f11.tif
Fig. 11 Experimental PACs overlaid onto theoretically generated ones, displaying regions of (A) stable, (B) decreasing, and (C) increasing membrane permeability.86

In 2017, Henderson et al. examined the membrane permeability response to incubation with various concentrations of K2Cr2O7.99 This oxidation state of chromium serves no biological role and is known to be toxic to live cells.93 SECM depth scan imaging and 3D finite element method simulations were used to quantify cell membrane permeability. Using FcMeOH as the electrochemical mediator, 1 h incubation with low concentrations of K2Cr2O7 (1.0 μM–25 μM) were found to be indistinguishable from untreated control cells (Fig. 12A, region a). At these low concentrations, over the 1 h incubation, the cellular defense mechanisms are able to cope with the toxic effects of K2Cr2O7. By increasing the concentration of the exposure (50 μM–500 μM), the membrane permeability is observed to decrease to a minimum value at 500 μM (Fig. 12A, region b). Incubation with higher concentrations was observed to yield an increase in membrane permeability (Fig. 12A, region c). Using the impermeable mediator FcCOOH, membrane permeability was observed to be impermeable between 0 μM and 250 μM. When exposed to 500 μM of K2Cr2O7 or higher, the membrane permeability is observed to start to increase (Fig. 12B, region c). This trend shows similarities to the FcMeOH study and shows that the cell membrane is sufficiently compromised so that it is no longer capable of excluding charged hydrophilic species. This study was coupled with an MTT cell viability study. It was observed that population viability was indistinguishable from control between 1.0 μM and 10 μM. Incubation with 25 μM–3000 μM showed a gradual decrease in cell viability. This occurs at a similar concentration to the change membrane permeability trend from region a to region b (Fig. 12A) observed by SECM imaging with FcMeOH. This study provided insight into the effects of acute exposure to K2Cr2O7.


image file: c8an01490f-f12.tif
Fig. 12 Graphical representation of the membrane permeability change vs. K2Cr2O7 incubation concentration, acquired with (A) FcMeOH and (B) FcCOOH mediators. The three regions of (a) stable, (b) decreasing, and (c) increasing permeability coefficients are labeled and separated by the grey dashed lines.99

To analyze the more chronic effects of K2Cr2O7 on live cells, examination of lower concentrations needed to be explored over a longer timescale. In 2018, a long duration study (5 days) of the effects of K2Cr2O7 was performed.100 Low concentrations of as K2Cr2O7 were examined over this timeframe (2.0, 5.0 and 10.0 μM) with SECM. The T24 cell line studied has a doubling time of 18–19 h, allowing study of K2Cr2O7 over several generations of cell growth. This allows the examination of membrane permeability effects associated with long duration chronic exposure to low concentrations of K2Cr2O7. When exposed to 2.0 μM of K2Cr2O7, an initial decrease in membrane permeability is observed between 1 h and 6 h of study (Fig. 13A). Further incubation of the cells out to two days showed a rapid increase in membrane permeability. Continuing incubation for 3 and 5 days showed a gradual increase and plateau in membrane permeability. At 5 days of incubation a significant reduction in total cell population was noted, and incubation was discontinued. Performing long duration incubations of the T24 cells with 5.0 μM and 10.0 μM (Fig. 13B and C respectively) produced a permeability minimum at 3 h of incubation. The two concentrations exhibit very similar behavior out to 12 h of incubation. The cells incubated with 5.0 μM K2Cr2O7 show a gradual rise in permeability for the 5 day period (Fig. 13B). The 10.0 μM K2Cr2O7 cells however rapidly reach a maximum permeability after 1 day, and then begin a steady decline in permeability (Fig. 13C). A significant decrease in the cell population was reported for the 5.0 μM cells following 5 days of incubation, and 3 days of incubation for the 10.0 μM cells. MTT studies of the cell populations show a rapid decrease in cell viability for the 5.0 μM and 10.0 μM to 20% between day 1 and day 2. The 2.0 μM K2Cr2O7 cells are able to maintain population during this timeframe. However, a massive drop is observed between day 3 and 5 from 72% viability to <10%.


image file: c8an01490f-f13.tif
Fig. 13 Graphical representation of the time-lapse membrane permeability change vs. K2Cr2O7 incubation concentration, acquired following incubations with K2Cr2O7 at concentrations of (A) 2 μM, (B) 5 μM, and (C) 10 μM.100

6. SECM analysis of live cells assisted by simulations

SECM is commonly paired with finite element method (FEM) simulations, as it allows the quantification of additional information on the cell of interest. Physiological characteristics such as membrane permeability17,82,87 and molecular transport of species28 can be simulated in this fashion. One of the most common ways these techniques are coupled is through the experimental acquisition and later the simulation of SECM probe approach curves (PACs). The experimental curve is overlaid onto simulated theoretical curves. Replicating the experimental geometry and physical parameters is important, as simulation results can vary significantly if these are not accounted for in some systems.

Early FEM simulations commonly used 2D-axially symmetric models for simulation. These models are created in 2D, with a linear axis of symmetry as part of the model. At time of computation, the 2D geometry is effectively rotated around the axis of symmetry constructing a pseudo 3D model. This methodology provides simplicity of design, low memory and storage utilization, and fast compute times. In the case of SECM, this axis of symmetry is positioned in the center of the electrode, and the electrode is positioned over the center of an axially symmetric sample.83 To simulate a simple PAC over a single round cell, this is often adequate. However, nonuniform cells or electrodes cannot be accurately simulated in this way. The electrode is also bound to the central axis of the cell, limiting the versatility of the simulation method.

In 2016, the creation of a 3D model for the off axis simulation of PACs to a single live cell was explored.54 The 3D geometry was designed to mirror that of previous 2D axially symmetric models. A comparison between conductor and insulator PACs was performed to validate the new 3D model (Fig. 14). Simulation of cell membrane permeability was performed at the central cell axis, as well as 2 other locations over the cell, and a location 2× the cell radius. These simulated PACs were compared to experimentally acquired PACs, which were extracted from an SECM depth scan (Fig. 15). The variation in geometry across the cell membrane caused significant changes in the PAC curve shape, however those changes mirrored the experimental PACs. The experimental cells were artificially stressed through incubation with various concentrations of Cd2+. This heavy metal stressor has been shown to yield changes in membrane permeability.17,86 Membrane permeability was found to be affected uniformly across the cell membrane by each tested concentration of Cd2+.


image file: c8an01490f-f14.tif
Fig. 14 (A) Meshed 3D model of the SECM approach to a T24 cell. (B) Meshed 3D SECM probe tip. (C) Labelled 2D axially symmetric model geometry of the SECM approach to the T24 cell. (D) Meshed 2D SECM model with axial symmetry. (E) Comparison of SECM PACs to an insulating membrane in 2D axial symmetry and 3D models. (F) Comparison of SECM PACs to a conductor in 2D axial symmetry and 3D models.54

image file: c8an01490f-f15.tif
Fig. 15 (A) Extracted experimental PACs along the symmetrical axis to single live cells incubated with 0, 50 and 100 mM Cd2+, with which a set of simulated PACs are overlaid. (B) Simulated and extracted experimental PACs at half cell radius. (C) Simulated and extracted experimental PACs at full cell radius. (D) Simulated and experimental PACs to the Petri dish at 2× cell radius.54

In 2016, the simulation of full 2D depth scans was performed.101 Modelling the SECM probe approach to a single live cell in 3D allowed for the complete control of electrode position and cell surface geometry. This allowed for the generation of off axis PACs, horizontal line scans at various heights over the cell, as well as the simulation of full 2D depth scans (Fig. 16). Quantification of average T24 cell height by SECM was performed, using 2 different PAC methods. The accuracy of these methods was compared. The topography of the cells was also compared, allowing for characterization of the full cell profile at various tip to cell distances. Simulation of full SECM depth scans allowed for the direct comparison of experimental and theoretical probe approach data. PAC or horizontal line scan extractions can be taken from any location in these data sets and rapidly compared.


image file: c8an01490f-f16.tif
Fig. 16 (A) Experimental and (B) simulated depth scans of an approach to a single cell with a height of 8 μm and a diameter of 20 μm (5 μm UME, RG 3). The theoretical UME approaches to 2.0 μm from the highest point on the cell during its closest horizontal sweep. The experimental depth scan was determined to approach to a height of 2.6 μm above the cell.101

7. Imaging live cells in nanoscale

SECM resolution is directly dependant on the size of the electrochemical probe. As a result there has been a push to fabricate smaller probes since the inception of the technique. A recent review by Bard et al. in Chemical Communications (2018) provides an excellent examination of general SECM techniques and applications at the nanometer scale.102 This section seeks to focus more exclusively on SECM applications towards single live cell analysis. Note that nanoscale probes103 are commonly used to electrochemically analyze intracellular processes, due to the less invasive nature of their small size.32,36

Nanoscale imaging comes with many technical challenges. Smaller scale electrodes require positioning closer to the sample of interest to detect electrochemical feedback. As a result, variation in sample height, and sample tilt play a much larger role at smaller scale. Depth based scanning techniques are a common way to avoid this limitation, though do not guarantee that collision with the substrate can be avoided. Many combination methods such as 4D shear force,67 hopping mode,71 SECM–SICM,49 and SECM–AFM46 show their strengths when imaging nanoscale. Soft probes have also been prepared that can be scanned across the sample surface without causing significant damage or stress to the sample or probe.104,105 Voltage switching methods have been explored for live cell study at the nanoscale that allow for the simultaneous acquisition of SECM data with multiple electrochemically active mediators.91

In 2017, Takahashi et al. showed the effective use of combination SECM–SICM electrode probes to image single 3T3-L1 fixed cells (Fig. 17A).106 Dual conductor SECM–SICM electrode probes were created in house using a CO2 laser puller and the pyrolytic carbon deposition method.49 This fabrication method allows for nanoscale electrodes with small RGs (a = 278 nm, RG = 1.1). The SECM probe is used with an FcMeOH electrochemical mediator to determine the membrane flux of the 3T3-L1 cells (Fig. 17B). The simultaneous SICM scan is used to determine cell topography (Fig. 17C). The comparison of the SECM and SICM scans shows sites of inhomogeneous current response across the cell surface (labelled as white arrows on the images). The first derivative of the topography image was also calculated providing a plot of the slope to characterize surface roughness (Fig. 17D). Regions with a steep vertical slope are observed at the positions of inhomogeneous current feedback, indicating a difference in current–distance responses between SECM and SICM based on geometry. The SECM–SICM method was also applied to image regions of the cell where intracellular lipid droplets were present. The presence of these lipid droplets yielded no significant difference in the faradaic current when comparing cells with and without the droplets. This indicates that the lipid droplets are not close to the membrane surface but are deeper inside the cell.


image file: c8an01490f-f17.tif
Fig. 17 (a) Optical image and SECM–SICM, (b) faradaic current, (c) topography, and (d) slope images of adipocyte in 0.5 mM FcCH2OH + PBS. The SECM electrode radius is 117 nm with RG = 1.1. The SECM and SICM electrodes were held at 500 and 200 mV vs. Ag/AgCl, respectively. The hopping amplitude was 3.0 mm. The surface area of the captured images was 80 mm × 80 mm.106

In 2018, Shen et al. showed SECM imaging of neurotransmission at the synaptic cleft between living Aplysia neurons.107 Imaging within the synaptic cleft has been difficult in the past due to its extremely small size (300 nm x < 100 nm). A 30 nm diameter nanoITES pipette electrode was used to selectively image acetylcholine release. The direct measurement of acetylcholine in the synaptic cleft reduces the dilution of the neurotransmitter, as it diffuses away from the site of release into the extracellular medium. This increases the performance of the low concentration electrochemical measurement. Synaptic neurotransmitter release was observed to be composed of singlet, doublet and multiplet profiles (Fig. 18A–C). The singlet profile was observed 50% of the time with lower frequency for doublet and multiplet. The frequency of synaptic acetylcholine release was found to be consistent with the measured norepinephrine release (measured by a carbon nanofiber electrode). The average number of acetylcholine molecules released for the doublet was determined to be the double that observed in the singlet. Quantitative analysis of the amperometric peak half widths and the number of molecules suggests the release of multiple vesicles, or partial release of a single vesicle (Fig. 18D and E).


image file: c8an01490f-f18.tif
Fig. 18 Single synaptic cholinergic neurotransmission measured in situ. (A–C) Current–time trace (amperometry) representing synaptic transmitter concentration and release dynamics simultaneously, where diverse cholinergic concentration dynamics were observed that consisted of singlet (panel A), doublet (panel B) and multiplet (panel C). Single current (concentration) maxima occur during singlet release (50% occurrence frequency out of 16 events in total); a second current maximum occurs before the first current peak decreases to the base value for the doublet events (∼30% occurrence frequency); multiple concentration peaks (more than two) were observed for multiplets with lower occurrence frequency (∼20%). (D) Proposed mechanism of the variation in synaptic transmitter release dynamics. The neurotransmitter is released into the synaptic cleft from a single vesicle (left). The neurotransmitter is released into the synaptic cleft from two vesicles, V1 and V2, simultaneously (middle) or multiple vesicles simultaneously, which are going through either different stages of exocytosis as shown here, or similar stages of exocytosis (Right). (E) An alternative mechanism is possible for explaining doublets and multiplets based on the phenomenon of partial release.54–67 A vesicle goes through partial release twice, generating a doublet (middle); the two individual peaks (Peak 1 and Peak 2) correspond to each partial release event.107

In 2017, Page et al. characterized the uptake of [Ru(NH3)6]3+ by live Zea mays root hair cells using SECM–SICM.70 SECM–SICM dual electrode probes were custom prepared with a total probe diameter of 500 nm, composed of a SICM nanopipette and a pyrolytically deposited carbon SECM electrode. The SICM nanopipette was filled with a 10 mM [Ru(NH3)6]3+ (RuHex) solution, with no RuHex in bulk solution. This provides a localized source of RuHex, which is electrochemically reduced at the carbon electrode tip. It was found that the concentration at the nanopipette orifice was 2 mM. The SICM electrode measures resistance between itself and an Ag/AgCl electrode in bulk solution providing topographic information and does not react with the RuHex. This imaging methodology was validated experimentally on root cells as well as simulated using the FEM method. A single Zea mays root hair cell was imaged at two locations, the cell tip and body (Fig. 19A). Variation in cell height was observed at the two locations on the cell (Fig. 19B). Monitoring the uptake of RuHex by the live cell shows faster uptake at the root hair tip, with uptake rates of 0.27 ± 0.05 cm s−1 for the cell tip and 0.22 ± 0.05 cm s−1 for the cell body being observed (Fig. 19C and D).


image file: c8an01490f-f19.tif
Fig. 19 SICM-SECM topographical and [Ru(NH3)6]3+ uptake mapping of two regions of a single Zea mays root hair cell. (a) Optical image of the scanned root hair cell; the scan area is denoted by a dashed rectangle. (b) Substrate topography extracted from the z-position at the point of closest approach from the SICM channel. (c) Normalized SECM current map showing a clear difference in uptake between the root hair cell body (higher current, lower uptake) and the root hair cell tip (lower current, higher uptake). “Normalized current” is the ratio of the [Ru(NH3)6]3+ reduction current at the point of closest approach to the same reduction current in bulk. (d) Histograms of the normalized SECM current across the two different regions of the root hair cell, “tip” and “body” (see part b).70

8. Summary and future directions

Recent advances in single live cell analysis using SECM have shown great promise for increased adoption as a common bioanalytical technique. The trend toward coupling SECM with other strong bioanalytical tools provides interesting options for the simultaneous acquisition of data from multiple techniques on the same cell. The development of advanced scanning methods to avoid probe-to-sample crash offers an increased ease of use for new users, and well documented electrode fabrication techniques allow for greater availability of specialized probes. Low cost options for SECM also allow users to perform basic forms of sample analysis with advanced features trickling down in software or higher precision parts, providing an excellent introduction to the analytical technique. SECM has demonstrated strength in its ability to detect extracellular and intracellular ROS and RNS as a label free analytical technique. This provides localized single cell analysis of reactive oxygen species, leading to the quantification of these compounds and their correlation to oxidative damage. SECM also provides an excellent method for detecting localized membrane transport. Bulk transport across a large region of the cell membrane can be performed, or mapping of variation in membrane transport can be carried out with small scale electrodes. Analysis can also be performed over specific cell features such as ion channels to quantify the rate of species flux into and out of the cell. This has provided an exceptional tool for the analysis of membrane permeability change induced by heavy metal stress. FEM simulations of the SECM analysis of single live cells have also become more advanced as computing power has become more available and lower cost. Highly demanding full 3D simulation models are now capable of running on consumer available gaming or workstation grade PC hardware. Increased CPU core count provides faster computation, and the availability of memory and storage space makes larger more complex models viable. This allows for the creation of tailored geometry, simultaneous simulation of multiple electrochemical species, and the incorporation of additional physical calculations into the simulation. In recent years, there has been an increased prevalence of SECM studies involving nanoscale imaging. This small scale technique provides excellent bioanalytical tools for the electrochemical imaging of intracellular processes, localized membrane permeability, membrane imbedded proteins, and small scale cellular structures. SECM is anticipated to be a much stronger bioanalytical tool for single live cell analysis, offering great promise for the imaging and quantification of cellular physiological processes.

Author contributions

The manuscript was written through the contributions of both authors. Both authors have given approval to the final version of the manuscript.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC, DG RGPIN-2013-201697), Canada Foundation of Innovation/Ontario Innovation Trust (CFI/OIT, 9040), Premier's Research Excellence Award (PREA, 2003), Canada Institute of Photonics Innovation (2005), Ontario Photonics Innovation (2002), and the University of Western Ontario. We would also like to thank the Electronic Shop and ChemBio Store in the Department of Chemistry at Western for the high-quality service. Jonathan Wong's assistance in graphic design is very much appreciated.

References

  1. A. J. Bard and L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, John Wiley & Sons, Inc., 2 edn, 2001 Search PubMed.
  2. A. J. Bard and M. V. Mirkin, Scanning electrochemical microscopy, CRC Press, 2 edn, 2012 Search PubMed.
  3. C. G. Zoski, J. Electrochem. Soc., 2016, 163, H3088–H3100 CrossRef CAS.
  4. H. Y. Liu, F. R. F. Fan, C. W. Lin and A. J. Bard, J. Am. Chem. Soc., 1986, 108, 3838–3839 CrossRef CAS.
  5. R. C. Engstrom, M. Weber, D. J. Wunder, R. Burgess and S. Winquist, Anal. Chem., 1986, 58, 844–848 CrossRef CAS.
  6. A. J. Bard, G. Denuault, C. Lee, D. Mandler and D. O. Wipf, Acc. Chem. Res., 1990, 23, 357–363 CrossRef CAS.
  7. A. J. Bard, F. R. F. Fan, J. Kwak and O. Lev, Anal. Chem., 1989, 61, 132–138 CrossRef CAS.
  8. D. Polcari, P. Dauphin-Ducharme and J. Mauzeroll, Chem. Rev., 2016, 116, 13234–13278 CrossRef CAS PubMed.
  9. Z. F. Ding, B. M. Quinn and A. J. Bard, J. Phys. Chem. B, 2001, 105, 6367–6374 CrossRef CAS.
  10. G. Wittstock, M. Burchardt, S. E. Pust, Y. Shen and C. Zhao, Angew. Chem., Int. Ed., 2007, 46, 1584–1617 CrossRef CAS PubMed.
  11. C. Nowierski, J. J. Noel, D. W. Shoesmith and Z. F. Ding, Electrochem. Commun., 2009, 11, 1234–1236 CrossRef CAS.
  12. Y. H. Yin, L. Niu, M. Lu, W. K. Guo and S. H. Chen, Appl. Surf. Sci., 2009, 255, 9193–9199 CrossRef CAS.
  13. C. Combellas, A. Fuchs and F. Kanoufi, Anal. Chem., 2004, 76, 3612–3618 CrossRef CAS PubMed.
  14. J. Tang, J. J. Zheng, Y. T. Yu, L. N. Chen, N. Zhang and Z. W. Tian, Electrochim. Acta, 2012, 83, 247–252 CrossRef CAS.
  15. A. Eifert, B. Mizaikoff and C. Kranz, Micron, 2015, 68, 27–35 CrossRef CAS PubMed.
  16. S. Amemiya, J. Guo, H. Xiong and D. A. Gross, Anal. Bioanal. Chem., 2006, 386, 458–471 CrossRef CAS PubMed.
  17. M. S. Li, F. P. Filice and Z. Ding, J. Inorg. Biochem., 2014, 136, 177–183 CrossRef CAS PubMed.
  18. S. Kuss, D. Polcari, M. Geissler, D. Brassard and J. Mauzeroll, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 9249–9254 CrossRef CAS PubMed.
  19. S. Kuss, R. Cornut, I. Beaulieu, M. A. Mezour, B. Annabi and J. Mauzeroll, Bioelectrochemistry, 2011, 82, 29–37 CrossRef CAS PubMed.
  20. I. Beaulieu, S. Kuss, J. Mauzeroll and M. Geissler, Anal. Chem., 2011, 83, 1485–1492 CrossRef CAS PubMed.
  21. D. Koley and A. J. Bard, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 16783–16787 CrossRef CAS PubMed.
  22. B. Liu, S. A. Rotenberg and M. V. Mirkin, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 9855–9860 CrossRef CAS.
  23. A. J. Bard, X. Li and W. Zhan, Biosens. Bioelectron., 2006, 22, 461–472 CrossRef CAS PubMed.
  24. X. Zhao, P. M. Diakowski and Z. Ding, Anal. Chem., 2010, 82, 8371–8373 CrossRef CAS PubMed.
  25. C. Lee, J. Kwak and A. J. Bard, Proc. Natl. Acad. Sci. U. S. A., 1990, 87, 1740–1743 CrossRef CAS.
  26. N. Nioradze, J. Kim and S. Amemiya, Anal. Chem., 2011, 83, 828–835 CrossRef CAS PubMed.
  27. S. Cannan, J. Cervera, R. J. Steliaros, E. Bitziou, A. L. Whitworth and P. R. Unwin, Phys. Chem. Chem. Phys., 2011, 13, 5403–5412 RSC.
  28. T. Matrab, F. Hauquier, C. Combellas and F. Kanoufi, ChemPhysChem, 2010, 11, 670–682 CrossRef CAS PubMed.
  29. D. Koley and A. J. Bard, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 11522–11527 CrossRef CAS PubMed.
  30. J. Kim, A. Izadyar, M. Shen, R. Ishimatsu and S. Amemiya, Anal. Chem., 2014, 86, 2090–2098 CrossRef CAS PubMed.
  31. S. Bergner, J. Wegener and F. M. Matysik, Anal. Chem., 2011, 83, 169–174 CrossRef CAS PubMed.
  32. P. Sun, F. O. Laforge, T. P. Abeyweera, S. A. Rotenberg, J. Carpino and M. V. Mirkin, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 443–448 CrossRef CAS PubMed.
  33. J. M. Liebetrau, H. M. Miller, J. E. Baur, S. A. Takacs, V. Anupunpisit, P. A. Garris and D. O. Wipf, Anal. Chem., 2003, 75, 563–571 CrossRef CAS PubMed.
  34. R. T. Kurulugama, D. O. Wipf, S. A. Takacs, S. Pongmayteegul, P. A. Garris and J. E. Baur, Anal. Chem., 2005, 77, 1111–1117 CrossRef CAS PubMed.
  35. R. Zhu, S. M. Macfie and Z. Ding, J. Exp. Bot., 2005, 56, 2831–2838 CrossRef CAS PubMed.
  36. Y. Wang, J. M. Noel, J. Velmurugan, W. Nogala, M. V. Mirkin, C. Lu, M. Guille Collignon, F. Lemaitre and C. Amatore, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 11534–11539 CrossRef CAS PubMed.
  37. X. C. Zhao, N. O. Petersen and Z. F. Ding, Can. J. Chem., 2007, 85, 175–183 CrossRef.
  38. M. M. Zhang, Y. T. Long and Z. Ding, J. Inorg. Biochem., 2012, 108, 115–122 CrossRef CAS PubMed.
  39. X. C. Zhao, M. N. Zhang, Y. T. Long and Z. F. Ding, Can. J. Chem., 2010, 88, 569–576 CrossRef CAS.
  40. X. C. Zhao, S. Lam, J. Jass and Z. F. Ding, Electrochem. Commun., 2010, 12, 773–776 CrossRef CAS.
  41. D. P. Zhan, X. Li, A. B. Nepomnyashchii, M. A. Alpuche-Aviles, F. R. F. Fan and A. J. Bard, J. Electroanal. Chem., 2013, 688, 61–68 CrossRef CAS.
  42. X. Li and A. J. Bard, J. Electroanal. Chem., 2009, 628, 35–42 CrossRef CAS.
  43. C. Cai, B. Liu, M. V. Mirkin, H. A. Frank and J. F. Rusling, Anal. Chem., 2002, 74, 114–119 CrossRef CAS PubMed.
  44. W. Feng, S. A. Rotenberg and M. V. Mirkin, Anal. Chem., 2003, 75, 4148–4154 CrossRef CAS PubMed.
  45. B. Liu, W. Cheng, S. A. Rotenberg and M. V. Mirkin, J. Electroanal. Chem., 2001, 500, 590–597 CrossRef CAS.
  46. E. Casero, L. Vazquez, A. M. Parra-Alfambra and E. Lorenzo, Analyst, 2010, 135, 1878–1903 RSC.
  47. Y. B. Zu, Z. F. Ding, J. F. Zhou, Y. M. Lee and A. J. Bard, Anal. Chem., 2001, 73, 2153–2156 CrossRef CAS PubMed.
  48. Y. Lee, Z. Ding and A. J. Bard, Anal. Chem., 2002, 74, 3634–3643 CrossRef CAS PubMed.
  49. Y. Takahashi, A. I. Shevchuk, P. Novak, Y. Zhang, N. Ebejer, J. V. Macpherson, P. R. Unwin, A. J. Pollard, D. Roy, C. A. Clifford, H. Shiku, T. Matsue, D. Klenerman and Y. E. Korchev, Angew. Chem., Int. Ed., 2011, 50, 9638–9642 CrossRef CAS PubMed.
  50. N. B. Schorr, A. G. Jiang and J. Rodriguez-Lopez, Anal. Chem., 2018, 90, 7848–7854 CrossRef CAS PubMed.
  51. S. E. Salamifar and R. Y. Lai, Anal. Chem., 2013, 85, 9417–9421 CrossRef CAS PubMed.
  52. D. Momotenko, F. Cortes-Salazar, A. Lesch, G. Wittstock and H. H. Girault, Anal. Chem., 2011, 83, 5275–5282 CrossRef CAS PubMed.
  53. J. Guo and S. Amemiya, Anal. Chem., 2005, 77, 2147–2156 CrossRef CAS PubMed.
  54. F. P. Filice, M. S. Li, J. D. Henderson and Z. Ding, Anal. Chim. Acta, 2016, 908, 85–94 CrossRef CAS PubMed.
  55. Y. Lee, S. Amemiya and A. J. Bard, Anal. Chem., 2001, 73, 2261–2267 CrossRef CAS PubMed.
  56. O. Sklyar, A. Kueng, C. Kranz, B. Mizaikoff, A. Lugstein, E. Bertagnolli and G. Wittstock, Anal. Chem., 2005, 77, 764–771 CrossRef CAS PubMed.
  57. R. Zhu and Z. Ding, Can. J. Chem., 2005, 83, 1779–1791 CrossRef CAS.
  58. C. Wei and A. J. Bard, J. Electrochem. Soc., 1995, 142, 2523–2527 CrossRef CAS.
  59. L. Danis, D. Polcari, A. Kwan, S. M. Gateman and J. Mauzeroll, Anal. Chem., 2015, 87, 2565–2569 CrossRef CAS PubMed.
  60. P. Actis, S. Tokar, J. Clausmeyer, B. Babakinejad, S. Mikhaleva, R. Cornut, Y. Takahashi, A. Lopez Cordoba, P. Novak, A. I. Shevchuck, J. A. Dougan, S. G. Kazarian, P. V. Gorelkin, A. S. Erofeev, I. V. Yaminsky, P. R. Unwin, W. Schuhmann, D. Klenerman, D. A. Rusakov, E. V. Sviderskaya and Y. E. Korchev, ACS Nano, 2014, 8, 875–884 CrossRef CAS PubMed.
  61. K. McKelvey, B. P. Nadappuram, P. Actis, Y. Takahashi, Y. E. Korchev, T. Matsue, C. Robinson and P. R. Unwin, Anal. Chem., 2013, 85, 7519–7526 CrossRef CAS PubMed.
  62. L. Danis, M. E. Snowden, U. M. Tefashe, C. N. Heinemann and J. Mauzeroll, Electrochim. Acta, 2014, 136, 121–129 CrossRef CAS.
  63. D. Polcari, A. Kwan, M. R. Van Horn, L. Danis, L. Pollegioni, E. S. Ruthazer and J. Mauzeroll, Anal. Chem., 2014, 86, 3501–3507 CrossRef CAS PubMed.
  64. A. Page, D. Perry and P. R. Unwin, Proc. R. Soc. A, 2017, 473, 20160889 CrossRef PubMed.
  65. P. M. Diakowski and Z. Ding, Phys. Chem. Chem. Phys., 2007, 9, 5966–5974 RSC.
  66. J. V. Macpherson, P. R. Unwin, A. C. Hillier and A. J. Bard, J. Am. Chem. Soc., 1996, 118, 6445–6452 CrossRef CAS.
  67. M. Nebel, K. Eckhard, T. Erichsen, A. Schulte and W. Schuhmann, Anal. Chem., 2010, 82, 7842–7848 CrossRef CAS PubMed.
  68. A. Hengstenberg, A. Blochl, I. D. Dietzel and W. Schuhmann, Angew. Chem., Int. Ed., 2001, 40, 905–908 CrossRef CAS PubMed.
  69. F. R. Fan and A. J. Bard, Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 14222–14227 CrossRef CAS.
  70. A. Page, M. Kang, A. Armitstead, D. Perry and P. R. Unwin, Anal. Chem., 2017, 89, 3021–3028 CrossRef CAS PubMed.
  71. R. A. Lazenby, K. McKelvey and P. R. Unwin, Anal. Chem., 2013, 85, 2937–2944 CrossRef CAS PubMed.
  72. G. N. Meloni, Anal. Chem., 2017, 89, 8643–8649 CrossRef CAS PubMed.
  73. S. Di Meo, T. T. Reed, P. Venditti and V. M. Victor, Oxid. Med. Cell. Longevity, 2016, 2016, 1245049 Search PubMed.
  74. G. Waris and H. Ahsan, J. Carcinog., 2006, 5, 14 CrossRef PubMed.
  75. V. J. Thannickal and B. L. Fanburg, Am. J. Physiol.: Lung Cell. Mol. Physiol., 2000, 279, L1005–L1028 CrossRef CAS PubMed.
  76. M. Valko, M. Izakovic, M. Mazur, C. J. Rhodes and J. Telser, Mol. Cell. Biochem., 2004, 266, 37–56 CrossRef CAS PubMed.
  77. A. Valencia and J. Moran, Free Radicals Biol. Med., 2004, 36, 1112–1125 CrossRef CAS PubMed.
  78. C. Amatore, S. Arbault, C. Bouton, K. Coffi, J. C. Drapier, H. Ghandour and Y. Tong, ChemBioChem, 2006, 7, 653–661 CrossRef CAS PubMed.
  79. Y. Li, K. Hu, Y. Yu, S. A. Rotenberg, C. Amatore and M. V. Mirkin, J. Am. Chem. Soc., 2017, 139, 13055–13062 CrossRef CAS PubMed.
  80. S. F. Steinberg, Physiol. Rev., 2008, 88, 1341–1378 CrossRef CAS PubMed.
  81. T. Yasukawa, Y. Kondo, I. Uchida and T. Matsue, Chem. Lett., 1998, 27, 767–768 CrossRef.
  82. T. Yasukawa, I. Uchida and T. Matsue, Biochim. Biophys. Acta, Biomembr., 1998, 1369, 152–158 CrossRef CAS.
  83. J. Kim, A. Izadyar, N. Nioradze and S. Amemiya, J. Am. Chem. Soc., 2013, 135, 2321–2329 CrossRef CAS PubMed.
  84. M. Tsionsky, Z. G. Cardon, A. J. Bard and R. B. Jackson, Plant Physiol., 1997, 113, 895–901 CrossRef CAS PubMed.
  85. Z. Chen, S. Xie, L. Shen, Y. Du, S. He, Q. Li, Z. Liang, X. Meng, B. Li, X. Xu, H. Ma, Y. Huang and Y. Shao, Analyst, 2008, 133, 1221–1228 RSC.
  86. J. D. Henderson, F. P. Filice, M. S. M. Li and Z. Ding, J. Inorg. Biochem., 2016, 158, 92–98 CrossRef CAS PubMed.
  87. M. N. Zhang, Z. Ding and Y. T. Long, Analyst, 2015, 140, 6054–6060 RSC.
  88. M. S. M. Li, F. P. Filice, J. D. Henderson and Z. F. Ding, J. Phys. Chem. C, 2016, 120, 6094–6103 CrossRef CAS.
  89. A. Bondarenko, T. E. Lin, P. Stupar, A. Lesch, F. Cortes-Salazar, H. H. Girault and H. Pick, Anal. Chem., 2016, 88, 11436–11443 CrossRef CAS PubMed.
  90. A. Soldà, G. Valenti, M. Marcaccio, M. Giorgio, P. G. Pelicci, F. Paolucci and S. Rapino, ACS Sens., 2017, 2, 1310–1318 CrossRef PubMed.
  91. Y. Takahashi, A. I. Shevchuk, P. Novak, B. Babakinejad, J. Macpherson, P. R. Unwin, H. Shiku, J. Gorelik, D. Klenerman, Y. E. Korchev and T. Matsue, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 11540–11545 CrossRef CAS PubMed.
  92. D. Polcari, J. A. Hernandez-Castro, K. Li, M. Geissler and J. Mauzeroll, Anal. Chem., 2017, 89, 8988–8994 CrossRef CAS PubMed.
  93. A. Hartwig, Free Radicals Biol. Med., 2013, 55, 63–72 CrossRef CAS PubMed.
  94. K. Jomova and M. Valko, Toxicol., 2011, 283, 65–87 CrossRef CAS PubMed.
  95. J. J. Garcia, E. Martinez-Ballarin, S. Millan-Plano, J. L. Allue, C. Albendea, L. Fuentes and J. F. Escanero, J. Trace Elem. Med. Biol., 2005, 19, 19–22 CrossRef CAS PubMed.
  96. J. Van der Paal, E. C. Neyts, C. C. W. Verlackt and A. Bogaerts, Chem. Sci., 2016, 7, 489–498 RSC.
  97. D. Zhan, F. R. Fan and A. J. Bard, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 12118–12122 CrossRef CAS PubMed.
  98. D. Zhan, X. Li, W. Zhan, F.-R. F. Fan and A. J. Bard, Anal. Chem., 2007, 79, 5225–5231 CrossRef CAS PubMed.
  99. J. D. Henderson, F. P. Filice, M. S. M. Li and Z. F. Ding, ChemElectroChem, 2017, 4, 856–863 CrossRef CAS.
  100. F. P. Filice, M. S. M. Li, J. M. Wong and Z. Ding, J. Inorg. Biochem., 2018, 182, 222–229 CrossRef CAS PubMed.
  101. M. S. M. Li, F. P. Filice and Z. F. Ding, J. Electroanal. Chem., 2016, 779, 176–186 CrossRef CAS.
  102. T. Kai, C. G. Zoski and A. J. Bard, Chem. Commun., 2018, 54, 1934–1947 RSC.
  103. Y.-L. Ying, Z. Ding, D. Zhan and Y.-T. Long, Chem. Sci., 2017, 8, 3338–3348 RSC.
  104. F. Cortes-Salazar, M. Trauble, F. Li, J. M. Busnel, A. L. Gassner, M. Hojeij, G. Wittstock and H. H. Girault, Anal. Chem., 2009, 81, 6889–6896 CrossRef CAS PubMed.
  105. F. Cortes-Salazar, D. Momotenko, A. Lesch, G. Wittstock and H. H. Girault, Anal. Chem., 2010, 82, 10037–10044 CrossRef CAS PubMed.
  106. Y. Takahashi, H. Ida, Y. Matsumae, H. Komaki, Y. Zhou, A. Kumatani, M. Kanzaki, H. Shiku and T. Matsue, Phys. Chem. Chem. Phys., 2017, 19, 26728–26733 RSC.
  107. M. Shen, Z. Qu, J. DesLaurier, T. M. Welle, J. V. Sweedler and R. Chen, J. Am. Chem. Soc., 2018, 140, 7764–7768 CrossRef CAS PubMed.

This journal is © The Royal Society of Chemistry 2019