SEARCH

SEARCH BY CITATION

Keywords:

  • organic electronics;
  • epithelial tissue;
  • diagnostics;
  • transepithelial resistance;
  • Salmonella typhimurium

Abstract

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

Ion flow across polarized epithelia is a tightly regulated process. Measurement of the transepithelial resistance is a highly relevant parameter for assessing the function or health of the tissue. Dynamic, electrical measurements of transepithelial ion flow are preferred as they provide the most accurate snapshot of effects of external stimuli. Enteric pathogens such as Salmonella typhimurium are known to disrupt ion flow in gastrointestinal epithelia. Here, for the first time, the use of organic transistors as a powerful potential alternative for front-line, disposable, high-throughput diagnostics of enteric pathogens is demonstrated. The transistors' ability to detect early and subtle changes in transepithelial ion flow is capitalized upon to develop a highly sensitive detector of epithelial integrity. Stable operation of the organic devices under physiological conditions is shown, followed by dynamic, pathogen-specific diagnosis of infection of epithelia. Further, operation of the device is possible in complex matrices, showing particular promise for food and safety applications.

1 Introduction

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

Epithelial tissues play a number of important roles including compartmentalization, protection, selective absorption, and transport.[1] Although one of the primary functions of the epithelium is to block pathogens and toxins from entry, many enteric pathogens have evolved to disrupt epithelial barriers through a variety of different mechanisms. One example is Salmonella typhimurium, one of the leading causes of food-borne illness.[2] The structures between adjacent cells in an epithelial layer that control ion flux are known as tight junctions, which are reported to be directly targeted by some enteric pathogens, or indeed hijacked by the pathogens as receptors.[2, 3] Symptoms of enteric disease such as inflammation and diarrhoea are thought to result primarily from tight junction disruption.[4] Tight junctions are multi-protein complexes containing over 40 proteins including ZO-1, occludin, and claudins.[5, 6] Occludin and claudin are transmembrane proteins containing extracellular loops that form a seal between adjacent cells, assisted by a number of submembrane scaffold proteins such as ZO-1. Also intimately involved is the perijunctional actin cytoskeleton, which binds to ZO-1, thus stabilizing the tight junction.[7] The tightness of this seal is dependent on the composition of the proteins in the complex, particularly the claudins.[8, 9]

A number of techniques are used to investigate epithelial barrier function following infection of the epithelia with enteric pathogens, including measurement of transepithelial resistance (TER), permeability assays, or most frequently immunofluorescence images showing changes in co-localization of tight junction proteins.[2, 10-14] Due in large part to limitations in technology, the vast majority of these reports are either static images, or isolated measurements with a large degree of inhomogeneity.[15] The tight junction has been demonstrated via fluorescence recovery after photobleaching to undergo continuous and rapid remodeling, suggesting a dynamic structure that changes rapidly upon exposure to extracellular stimuli on the scale of seconds to minutes.[16] Steady-state ion flux via the paracellular pathway goes either through the pore pathway formed by transmembrane tight junction proteins or via the non-pore pathway, postulated to be caused by dynamic breaking and resealing of tight junction strands.[9] The distinction between these pathways is often disregarded in literature and indeed in practice, and permeability is measured using tracer molecules, which pass through the non-pore pathway, and thus cannot be considered an accurate assessment of the function of the pore pathway.[5] An accurate assessment of barrier tissue integrity must take into account both pathways and so necessitates a direct measurement of ion flux. Measurement of ion flow across epithelial tissue (the parameter commonly reported is the TER) is an accepted in vitro method for assessing the integrity of epithelia. Measurements of ion flux are also label-free and directly measure a property of the cell layer, rather than indirectly measuring permeability of tracer molecules, which does not always correlate with changes in permeability to ions.[17] The dynamic nature of the structures that control ion flux in epithelia, coupled with the sometimes rapid infection rates of enteric pathogens and toxins, necessitates the development of a dynamic, rapid sampling method, to be able to distinguish between healthy, functioning epithelia, and those disrupted due to the effects of toxins and pathogens.

With the effort to reduce the numbers of animals used in toxicology and in diagnostics, comes also the necessity not only for the development of physiologically relevant in vitro models but also for technology, which can accurately assess them in a low-cost, high-throughput, and dynamic manner. Electronic impedance spectroscopy (EIS) has emerged as a technology to monitor toxicology of cells in vitro.[18, 19] However, in many cases, the devices are best suited to monitoring cell coverage or wound healing assays and are not adapted for measuring subtle changes in ionic flux.[20] One device that has been adapted for dynamic monitoring of changes in TER is the CellZscope;[21] however in its current format, it is not amenable to low-cost, high-throughput toxicology.

Organic electronic devices, over the last decade, have shown considerable promise in interfacing with biological systems due to advantages in terms of processing, flexibility, and low-cost production.[22, 23] In particular, the organic electrochemical transistor (OECT) is a device that has been shown to be particularly suited for the measurement of the electrical properties of cells both in vitro and in vivo,[15, 24] providing a very efficient transduction of ionic signals to electronic ones.[25, 26] In this work, we take advantage of the facile processing and stable operation of the OECT under physiological conditions, to dynamically monitor polarized epithelia in situ by measuring changes in ion flux. We show simultaneous operation of multiple transistors to dynamically assess S. typhimurium infection of an in vitro model of the human intestinal epithelium, both during long-term operation of the device and a focus on early time points, made possible due to rapid sampling. Finally, we show that the device operates stably in cell culture medium, but can also detect the presence of pathogens in complex matrices such as full-fat milk, of particular relevance for diagnostics.

2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

We previously integrated the OECT with epithelia and demonstrated the possibility of measuring intact barrier function, followed by monitoring of the effect of simple toxins such as hydrogen peroxide,[15] with unmatched temporal resolution.[1, 27] These devices were adequate for straightforward toxicology, operated at room temperature.

Limitations of the device included a lack of device-to-device reproducibility when cells were integrated and a loss of cell layer integrity after approximately 90 min at room temperature, precluding testing of more complex toxins or pathogens. To improve reproducibility and move towards high-throughput formats, a multiplex device was adopted, as illustrated in Figure 1a, where three photolithographically defined PEDOT:PSS (poly(3,4-ethylenedioxythiophene): poly(styrenesulfonate)) devices were fabricated on a single glass slide, thus improving considerably device-to-device reproducibility. We maintained the ease of integration with current cell-culturing protocols by culturing Caco-2 cells in a 24-well transwell filter, depicted in Figure 1b. To enable a long-term measurement of the device under physiological conditions, the completed devices were placed inside a portable incubator (Figure 1c). The OECT itself has been shown to be a remarkably stable device, operating in media for up to 5 weeks.[28] Our goal here was to demonstrate integration of the OECT with polarized epithelia for use in diagnostics. The operation of the OECT integrated with barrier tissue has been described previously.[15] Briefly, on application of a positive gate voltage (VGS), cations from the electrolyte drift into the conducting polymer channel thus dedoping it, resulting in a decrease in the source-drain current (IDS). The transient response of the device describes the ion flux between the gate electrode and channel of the OECT and is therefore sensitive to the rate at which ions traverse the barrier tissue layer. Application of a gate voltage pulse is illustrated in Figure 1d, along with the transient response to the gate pulse an OECT alone, OECT + cells, OECT + cells after scratch (where the cell layer is deliberately disrupted by scratching the barrier tissue layer with a needle). The response from the OECT alone and the response from the OECT + cells after scratch are virtually superimposable, while a clear difference can be noted in the OECT + cells. For data analysis of long-term measurements, a single parameter, the exponential time constant associated with the initial transient response to a constant gate voltage pulse, was chosen to monitor the barrier tissue integrity over time (see Methods and Figure S2 for details, Supporting Information). The use of the time constant, τ, is preferred over our previous use of normalized pulse (mid-modulation), as it is a more direct measure of the transient ion flux, is independent of pulse length, and shows lower noise over long-term measurements. In this work, we normalize τ between 0 and 1 as a measure of the integrity of the epithelium, where 0 is the OECT alone and 1 is the OECT + cells.

image

Figure 1. Layout, set-up, and characterization of OECT integrated with polarized epithelia. a) Top view illustration of the layout of three OECT devices fabricated on a single glass slide. b) Side view illustration of the three OECTs fabricated on the same glass slide integrated with Caco-2 cells grown on transwell filters. c) Picture of the multiplex device shown on a Petri dish inside the cell-culture incubator. The cell culture insert is shown suspended in the plastic holder affixed to the glass slide. The Ag/AgCl gate electrode is shown immersed in the apical media, while source and drain cables are attached to their respective positions on the glass slide. d) OECT current response (green) to a square gate pulse (VGS = 0.3 V) without cells, OECT with cells (blue) and OECT with cells after scratch (purple). e) Normalized response (τ) of OECT alone (green) and OECT with cells (blue) for long-term device operation. Results shown are from representative devices.

Download figure to PowerPoint

For long-term device operation, the transient response of the OECT + cells should remain stable. A plot of the normalized response of the time constant for long-term device operation is shown in Figure 1e, where 0 is the OECT alone, and 1 is the OECT + cells, showing that a stable baseline is maintained for approximately 10 h of continuous device operation, after which, the barrier properties of the monolayer seem to deteriorate. In agreement with literature,[29, 30] we have found that there is a toxic effect of the Ag/AgCl electrode over longer time periods (≥10 h). This effect was seen even when the electrode is not in active use, simply passively present in the media (Figure S1, Supporting Information). For acute toxicology or diagnostics purposes, an operation time of up to 10 h is more than sufficient, and Ag/AgCl shows superior performance than other metallic electrode materials such as platinum.[31] However, we are currently investigating other materials for the gate electrode that balance performance with lack of toxicity.

3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

We set out to use the OECT to monitor real-time infection of a polarized epithelial cell model (Caco-2 cells) by the enteric pathogen S. typhimurium. Infections were carried out in the multiplex set-up shown in Figure 1, under physiological conditions, with simultaneous recording possible of up to four devices. It should be noted that the addition of a relatively simple multiplexer can allow operation of many more devices and even higher sampling resolution. Here, the measurement was done at a sampling rate of 60 Hz, to accurately characterize the response time of the transistor to the applied pulse on the gate. Video S1 (Supporting Information) shows the simultaneous recording of four devices in parallel. For the infections we used both a wild-type (WT) S. typhimurium strain and a non-invasive (NI) mutant (see Materials and Methods for details). As illustrated in the cartoon in Figure 2a, the ability of S. typhimurium to infect Caco-2 cells has been established and is known to result in significant decreases in TER over a 6-h period at a similar multiplicity of infection (MOI).[32] Figure 2b shows the normalized response of the OECT upon infection with three different concentrations of bacteria (MOI: 10, 100, 1000; based on the initial seeding of the cells), over a 4-h period. In agreement with literature,[33] the WT S. typhimurium provoked a dramatic increase in ionic flux across the cell monolayer, with a complete destruction of the barrier observed within 1 h at MOI:1000. In the case of the NI bacterium, only negligible changes were observed compared with the control. These observations were confirmed by immunofluorescence analysis of the tight junction proteins claudin-1 and ZO-1 (Figure 2c) showing complete destruction of the monolayer at MOI:1000 with the WT bacteria. In agreement with the OECT data, at MOI:10 or 100, while some disorganization of the tight junction proteins and changes to cell morphology is observed, the cell layer remains more or less intact, corresponding to a less than total disruption of the barrier function. It is well known that the actin depolymerization disrupts tight junction structure and barrier function.[34] S. typhimurium's ability to rearrange host actin is well documented, and indeed we demonstrate significant rearrangement of actin (Figure 2d), with characteristic punctate actin staining previously noted for the WT,[35] but not for the NI.

image

Figure 2. Kinetics of polarized epithelial monolayer infected with Salmonella typhimurium. a) Cartoon illustrating infection with WT (left) and NI S. typhimurium (right). b) Mean normalized response (τ) of the OECT in the presence of WT (left) and NI S. typhimurium (right) at different MOI over 4 h, bacteria were added at t = 0. Non-infected represents OECT + cells with no added bacteria. Non-infected cells are in cyan, MOI:10 in blue, MOI:100 in purple, and MOI:1000 in red. c) For clarity, individual experiments are shown here; however, mean data from multiple experiments with error bars are shown in the Supporting Information (Figures S3,S4). c) Immunofluorescence of tight junction proteins ZO-1 (green) and claudin-1 (red), and the nucleus (blue) after 4 h infection with WT and NI S. typhimurium at different MOI. d) Immunofluorescence of actin cytoskeleton (red) and nucleus (blue) after 4 h infection with WT and NI S. typhimurium at different MOI. Insets show zoom on single cells.

Download figure to PowerPoint

4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

One of the key advantages of the device described here is the possibility to do rapid monitoring of multiple samples. In Figure 2, it is clear that even at the lowest MOI, the infection reaches steady state within the first hour. For this reason, we focused on the first hour of infection (Figure 3a) to monitor the initial kinetics of infection of S. typhimurium. In Figure 3a, a clear relation may be observed between the number of bacteria added and the dramatic decrease in the barrier properties of the epithelial layer within the first hour of infection. We show conclusively that the NI strain does not generate any changes in the ion flux across the epithelial cells, regardless of the MOI (Figure 3b). One interesting observation from the WT infection data is that there is a lag time of approximately 30 min before the onset of the increase in ion flux with the MOI of 10 and 100. In the case of MOI:1000, an initial plateau appears to be reached rapidly, within the first 20–30 min, followed by a second decrease leading to a complete breach of the barrier properties of the tissue within 1 h. To determine whether this effect was related to bacterial replication, we determined the number of bacteria in the basolateral compartment of the device at 30 and 60 min post-infection (Figure 3c). For the WT, the number of bacteria in the basolateral compartment went from approximately 1 × 103 CFU mL−1 at 30 min, to 1 × 105 CFU mL−1 at 60 min. We assert that the second, faster phase of increase in ion flux could be due to effects of the bacteria on the basolateral side, as it has been reported that S. typhimurium can infect polarized epithelial cells via either the apical or basolateral sides.[36] Immunofluorescence analysis of the tight junction proteins showed clear ring-like structures of ZO-1, claudin -1, and actin in control and NI monolayers. In the WT samples, there was a diffuse pattern of claudin-1 and ZO-1, and a constriction of actin, starting at 30 min, but plainly visible at 60 min.

image

Figure 3. Initial kinetics of Salmonella typhimurium infection of polarized epithelial monolayers. a) Mean normalized response of the OECT in the presence of WT S. typhimurium over the first hour of infection. b) Mean normalized response (τ) of the OECT in the presence of NI S. typhimurium over the first hour of infection. Non-infected cells are in cyan, MOI:10 in blue, MOI:100 in purple, and MOI:1000 in red. For clarity, individual experiments are shown here; however, mean data from multiple experiments with error bars are shown in the Supporting Information (Figures S3,S4). c) Number of bacteria present in the basolateral compartment of the epithelial monolayer 30 min and 1 h post-infection. d,e) Immunofluorescence of tight junction protein ZO-1, claudin-1, and actin after d) 30 min and e) 1 h infection. Bacteria were added at t = 0.

Download figure to PowerPoint

5 Kinetics of Salmonella typhimurium Infection in Milk

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

Hitherto we have shown OECTs are promising tools for monitoring the kinetics of infection of epithelia in ideal cell culture environments; however, real-life pathogen diagnostics present a number of additional challenges. Two major problems in diagnostics are specificity; being able to distinguish pathogenic strains from non-pathogenic strains, and, purification and amplification of the target molecule. Often, the sample is a complex matrix containing many different compounds including proteins, fats, and other molecules, which may interfere with downstream analysis.[37, 38] It is clear from the data shown above that the OECT integrated with polarized epithelial cells is capable of distinguishing pathogens such as S. typhimurium from attenuated strains, which are not invasion competent, and by corollary other non-pathogenic bacteria, and is thus pathogen specific. To mimic a “real” sample, we spiked a sample of full-fat milk with WT S. typhimurium (MOI:1000) to assess the effect of a complex matrix on the device operation. Although most analysis would not occur in undiluted milk, we did not dilute the milk, considering this a stringent test of our device (illustrated in Figure 4a). We demonstrate the maintenance of a stable baseline of the OECT over 9 h with full-fat milk in the apical compartment (Figure 4b). Upon addition of S. typhimurium to the milk, we detect a complete destruction of the barrier tissue properties within approximately 4.5 h. This is a slower response than that seen for the same concentration of bacteria in cell culture medium, possibly due to attenuating factors present in the milk or to a change in the bacterial replication kinetics. The same experiment was carried out using a commercially available EIS technique. It can be seen (Figure 4c) that the baseline signal is unstable when the device is operated with milk in the apical compartment, although the infection can be detected within approximately the same time frame.

image

Figure 4. Kinetics of Salmonella typhimurium infection of polarized epithelial cells in milk. a) Cartoon illustrating the integration of cells with OECT with milk in the apical side of the epithelial barrier, with zoom showing cells grown on the filter and indicating the addition of bacteria to the apical side of the polarized epithelium. b) Normalized response (τ) of the OECT in presence of milk (green open circles) and milk spiked with bacteria (red open circles). WT S. typhimurium bacteria were added at t = 0, at an MOI:1000. c) Percentage of the initial TER value measured using a commercially available EIS technique (CellZscope, Nanoanalytics) upon exposure to milk (green open circles) and milk spiked with bacteria (red open circles). WT S. typhimurium bacteria were added at t = 0, at an MOI:1000.

Download figure to PowerPoint

6 Discussion

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

The development of new materials and related technologies has the potential to revolutionize fundamental and applied research in biomedical sciences. One example is the emergence of organic bioelectronics, which should not be thought of as a replacement for traditional electronic materials, but, in certain cases, has been shown to be uniquely suited to interfacing with niche applications in biomedical sciences. The OECT has been repeatedly shown over the last decade to be an excellent transducer of key biological processes.[25, 28, 39-41] Here, we show that the OECT can be integrated with a polarized epithelium and can dynamically measure properties of the tissue in a multiplex fashion, under physiological conditions. Previously, studies reporting TER of infection of S. typhimurium or other enteric pathogens have tended to show TER measurements at isolated time-points, with relatively little attention being paid to the initial infection, most likely due to limitations in technology.[11] One exception is the study by Jepson et al., where MDCK I cells were infected with S. typhimurium and conductance was measured in an Ussing-chamber-type apparatus,[42] a technique that requires a bulky apparatus that could be scaled up with difficulty.[43] We show the first example of a systematic dynamic study on the effect of S. typhimurium infection of human epithelium. The OECT provides a detailed view of early events, directing further investigation of the molecular events involved, using complementary techniques such as immunofluorescence and invasion assays.

We propose that the pathogen-specific diagnostic shown here holds great promise for food and water testing. This is particularly underlined by the excellent stability and long-term operation of our device, and further by its operation in milk. The majority of methods used for diagnosing the presence of a pathogen in a food or water sample rely on techniques such as PCR (polymerase chain reaction) or ELISA (enzyme -linked immunosorbent assay). Such molecular diagnostics methods are capable of detecting very low levels of S. typhimurium; however, they generally require upstream purification and amplification steps, and often suffer from lack of specificity. Previously, studies have shown detection of S. typhimurium in milk,[44] however, to our knowledge, this is the first report of the dynamics of infection of polarized epithelial cells with an enteric pathogen present in a complex matrix. The advantage of a live-cell-based technique is that the disruption of ion flow in the cell layers will be due only to live pathogens and will not be affected by dead bacteria or background flora that often give false positives in current diagnostic methods.[37] The detection of S. typhimurium in milk was found to be slower than in cell culture medium, possibly due to protective effects of milk proteins or sugars. Reduced levels of invasion of Salmonella spp. have been demonstrated in human milk,[45, 46] with components of milk being postulated to improve the innate gastrointestinal immunity to infection due to probiotic function, anti-adhesive antimicrobial activity and also modulation of the intestinal epithelial cells themselves. More specifically for this study, certain bovine whey products have been shown to reduce the initial adhesion of S. typhimurium to Caco-2 cells.[47] Since adhesion is a necessary step for infection, this would certainly affect the rate at which the bacteria can induce changes in the barrier properties of the cells, in agreement with the data shown here.

7 Conclusions

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

A paradigm shift is slowly emerging in the study of barrier tissue with the realization that the tissue is dynamic and undergoes changes very rapidly upon response to stimuli. To fully meet the needs of scientists studying barrier tissue for toxicological purposes, diagnostics and basic research, tools must be made available that provide an exact picture of the response of the tissue. We believe that our OECT-cell-based device fulfills these requirements. Future requirements of the in vitro toxicology community will include a transition to 3D culture systems, with incorporated microfluidics, and importantly, in-line monitoring systems, something that seems infinitely possible given the progress reported here, and the literal flexibility of organic electronic devices. A further consideration is cost and capacity for high-throughput, again achievable with organic electronic devices, which are amenable to large-scale and low-cost production.

8 Experimental Section

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

OECT Fabrication: The conducting polymer formulation consisted of PEDOT:PSS (Heraeus, Clevios PH 1000), supplemented with ethylene glycol (0.25 mL for 1 mL PEDOT:PSS solution; Sigma–Aldrich), 4-dodecylbenzenesulfonic acid (0.5 μL mL−1), and 3-glycidoxypropyltrimethoxysilane (10 mg mL−1). On a clean glass substrate (75 mm × 25 mm), gold source and drain contacts were patterned via lift-off lithography and then thermally evaporated. Photoresist S1813 (MicroChem Corp.) was spin coated at 3000 rpm for 30 s on the glass substrate. Patterns were defined by photolithography (Chrome mask and Mask Aligner). MF-26A was used as developer. Then, 5 nm and 100 nm of chromium and gold, respectively, were evaporated. Finally, the photoresist was lifted-off in an acetone bath for 1 h, which left the substrate with the source and drain Au contacts only. PEDOT:PSS channel dimensions were patterned using a parylene peel-off technique described previously,[25, 28] resulting in a PEDOT:PSS channel width and length of 6 mm and 1 mm, respectively, and a thickness of 460 nm. Following PEDOT:PSS deposition, devices were baked for 1 h at 140 °C under atmospheric conditions. A PDMS (polydimethylsiloxane) well defined the active area, resulting in a channel area of 6 mm2.

Electronics: All of the measurements were done using a Ag/AgCl pellet as a gate electrode (Harvard apparatus) and cell medium (as described below) were the electrolyte. Experiments were performed in a humidified incubator at 37 °C, 5% CO2. Measurement parameters were chosen to avoid exposing the cell layers to a voltage drop above 0.5 V, as high voltages have been shown to damage bilayer membranes.[48] The recording of the OECTs was performed using a National Instruments (NI) PXIe-1062Q system. A four-channel source-measurement unit NI PXIe-4145 was used to bias and measure simultaneously up to four transistor channels. Each device is connected to an independent channel of the source-measurement system. All the channels are triggered through the built-in architecture of the system and are addressable individually. Both gating parameters such as time and voltage of the pulse, and acquisition parameters like the sampling frequency, can be set separately for each transistor. In the experiments presented here, the current in the channel was measured at Vds = −0.2 V and each channel was sampled at 60 Hz. The gate voltage was applied using a NI PXI-6289 modular instrument. A 2 s pulse was applied on the gate every 30 s, at Vgs = 0.3 V. All of the measurements were triggered through the built-in PXI architecture. The recorded signals were saved and analyzed using customized LabVIEW software.

Data Analysis: Data analysis was performed using a customized Matlab program to isolate and fit the time constant for each pulse. The time constant is extracted by performing a least-squares fit of the data from each pulse current response to Equation (1).

  • display math(1)

where α is a constant scaling term describing the magnitude of the current response, t0 is time at which the pulse starts, and τ is the time constant discussed above. Equation (1) is found to fit the experimental data well, both with and without cells (Figure S2, Supporting Information). The second exponential term, described by a time offset and time constant t' and τ' is incorporated to describe the long time evolution of the drain current, likely associated with the OECT—not the barrier tissue. The contribution from the second, slow exponential does not vary significantly over the time scale of the experiment. Similar relative results are attained whether or not the second exponential is used, but result in poor fits to the entire pulse duration (Figure S2, Supporting Information). A background is subtracted by taking into account the time response with no cells and after scratch. The data are then normalized using the following equation: NR = (Tno cells– T)/(Tno cells– Tcells), where Tcells refers to the Tau value in response to the application of the gate voltage of a barrier forming monolayer, and Tno cells refers to the Tau value in response to the application of the gate voltage of no barrier, with the dataset subsequently normalized to a to [0,1] scale.

Cell Culture: Caco-2 cells from ECACC (catalog no. 86010202) were seeded at a density of 1.5 × 104 cells/insert. Cells were routinely maintained at 37 °C in a humidified atmosphere of 5% CO2, in DMEM (Advanced DMEM (Dulbecco's modified Eagle medium) Reduced Serum Medium 1X, Invitrogen) with 2 × 10−3 m Glutamine (Glutamax-1; Invitrogen), 10% fetal bovine serum (Invitrogen), and Pen-strep (5000 [U mL−1] Penicillin–5000 [μg mL−1] Streptomycin; Invitrogen). For all experiments, Caco-2 cell layers were used after 3 weeks in culture, corresponding to a minimum TER of 400 Ω cm2 and a maximum apparent permeability of 1 × 10−6 cm s−1, consistent with literature reports.[49] Cells were cultured on Transwell filters (Millipore) with a 0.4-μm pore size and area of 0.33 cm2.

Bacterial Growth: Salmonella typhimurium strain 12023 and the non-invasive prgH mutant were provided by Prof. Sansonetti of the Pasteur Institute, and were grown in LB medium. The prgH mutant is an established strain, which is unable to invade epithelial cells.[50] After an overnight incubation at 37 °C with shaking, the bacteria were diluted 1:100 in fresh LB and grown for 3.5 h to reach mid-exponential phase for all experiments.

Bacterial Quantitation: Basal media were collected at 30 and 60 min post-infection. The media and serial dilutions were plated in triplicate on LB agar plates and incubated overnight at 37 °C. The bacteria were quantified by counting the number of colony-forming units per milliliter (CFU/mL).

Infection of Polarized Epithelia with S. typhimurium: The day prior to the experiment, cell culture media were exchanged for DMEM without antibiotics. The number of bacteria for each MOI was calculated according to the relation OD600 nm = 6 × 108 bacteria per milliliter. Bacteria were added on the apical side of the monolayer without changing media. For the experiments in milk, before the experiment, apical media were changed for microfiltered full-fat milk and ionic flux measured for 1 h to ensure a stable signal. Bacteria were then added to the apical side of the monolayer.

Immunofluorescence: Cells were infected as above for specified time-points and then fixed with 3%–4% paraformaldehyde in PBS pH 7.4, for 15 min at room temperature. Permeabilization was performed using 0.25% Triton in PBS, for 10 min at room temperature and with a blocking step consisting of 1% BSA in PBST (0.05% Tween 20 in PBS), for 30 min at room temperature. Mouse monoclonal anti-ZO-1 and rabbit polyclonal anti-Claudin-1 were used at 5 μg mL−1 (Invitrogen), in 1% BSA in PBST for 1 h at room temperature. Monolayers were then incubated for 1 h at room temperature with the secondary antibodies Alexa Fluor 488 goat anti-mouse IgG and Alexa Fluor 568 goat anti-rabbit (Molecular Probes). Cells were stained for F-actin using Phallotoxin (Invitrogen) for 20 min. Finally, the cells were incubated for 5 min at room temperature with Fluoroshield with DAPI (Sigma–Aldrich), mounted and examined with a fluorescent microscope.

CellZscope Measurements: The CellZscope (Nanoanalytics) measures the frequency-dependent impedance of barrier-forming cell cultures grown on permeable membranes and provides the TER as output. Impedance of cell layers, maintained as above, was measured in complete DMEM.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

The authors gratefully acknowledge funding from FP7-People-2009-RG, Marie Curie Project No. 256367 (CELLTOX), the European Research Council ERC-2010-StG Proposal No 258966 (IONOSENSE), FP7-People-2011-IIF, a Marie Curie post-doctoral Fellowship (J.R.), as well as a joint grant from the Conseil Regional de Provence Alpes Côte d'Azur and CDL Pharma (S.T.), and funding from ANRT and Microvitae (P.L.). The authors would like to thank Prof. Philippe Sansonetti of the Pasteur Institute for the kind gift of the S. typhimurium strains, and Michel Fiocchi from CMP for technical support during set-up of electronics.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1 Introduction
  4. 2 Multiplexed OECTs for Long-Term Monitoring of Integrity of Polarized Epithelia
  5. 3 Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Cells
  6. 4 Initial Kinetics of Salmonella typhimurium Infection of Polarized Epithelial Monolayers
  7. 5 Kinetics of Salmonella typhimurium Infection in Milk
  8. 6 Discussion
  9. 7 Conclusions
  10. 8 Experimental Section
  11. Acknowledgements
  12. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
adhm201300632-sup-0001-S1.pdf2701KSupplementary

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.