Downstream Simultaneous Electrochemical ... - ACS Publications

Yun Li , Catherine Sella , Frédéric Lemaître , Manon Guille-Collignon , Christian Amatore , and Laurent Thouin*. PASTEUR, Département de chimie, Ã...
1 downloads 0 Views 3MB Size
Subscriber access provided by Technical University of Munich University Library

Article

Downstream Simultaneous Electrochemical Detection of Primary Reactive Oxygen and Nitrogen Species Released by Cell Populations in an Integrated Microfluidic Device Yun Li, Catherine Sella, Frederic Lemaître, Manon Guille Collignon, Christian Amatore, and Laurent Thouin Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b02039 • Publication Date (Web): 06 Jul 2018 Downloaded from http://pubs.acs.org on July 7, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Downstream

Simultaneous

Electrochemical

Detection

of

Primary Reactive Oxygen and Nitrogen Species Released by Cell Populations in an Integrated Microfluidic Device Yun Li, Catherine Sella, Frédéric Lemaître, Manon Guille-Collignon, Christian Amatore and Laurent Thouin* PASTEUR, Département de chimie, École normale supérieure, PSL Université, Sorbonne Université, CNRS, 75005 Paris, France

Abstract An innovative microfluidic platform was designed to monitor electrochemically four primary reactive oxygen (ROS) and reactive nitrogen species (RNS) released by aerobic cells. Taking advantage of the space confinement and electrode performances under flow conditions, only few experiments were sufficient to directly provide significant statistical data relative to the average behavior of cells during oxidative stress bursts. The microfluidic platform comprised an upstream microchamber for cell culture and four parallel microchannels located downstream for detecting separately H2O2, ONOO-, NO● and NO2. Amperometric measurements were performed at the highly-sensitive Pt-black electrodes implemented in microchannels. RAW 264.7 macrophages secretion triggered by a calcium ionophore were used as an important cell type for assessing the performance, sensitivity and specificity of the integrated microfluidic device. In comparison to some previous evaluations achieved from single-cell measurements, reproducible and relevant determinations validated the proof of concept of this microfluidic platform for analyzing statistically significant oxidative stress responses of various cell types.

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

INTRODUCTION

Investigation of cellular responses linked to oxidative stress mechanisms is of major importance for understanding many complex physiological and pathological life processes.1,2 A challenge in this area is aimed at monitoring in real time and in situ the release of the primary reactive oxygen species (ROS) or reactive nitrogen species (RNS) from biological organisms, explants, population of cells, single cells and even cellular organelles. This can be achieved by using fluorescent techniques with relatively good quantifications of the global oxidative stress behavior.3 However, in such a case, the selectivity among the primary ROS and RNS production is generally poor due to the use of redox fluorescent markers which react with many ROS and RNS simultaneously. For this reason, bioelectroanalytical techniques still remain the techniques of choice for monitoring fluxes of oxidative species that exhibit electroactive properties. Good analytical performances of amperometric detections can be achieved by combining high sensitivity and high selectivity microelectrode responses through adequate surface modifications and precise control of electrode potentials. This is the case for the four key oxidative species H2O2, ONOO-, NO● and NO2-. In previous studies, these species could been selectively detected at the single-cell level with Pt-black modified electrodes after surface modification of either carbon-fiber ultramicroelectrodes4-6 or platinum microband electrodes.7-9 Furthermore, good spatio-temporal resolutions were attained using several electrode configurations.4-6,9-12 Single-cell investigations were performed according to a so-called semi-artificial synapse configuration.4 Indeed, positioning an electrode at micrometric or submicrometric distances from a living cell allows restriction of the extracellular volume in which species are produced. This ensures both an adequate signal-to-noise ratio and a quantitative collection of released species. Under these conditions, the composition of individual cellular oxidative bursts can be monitored with high sensitivity and selectivity.5 However, investigation at single-cell levels only reports on the behavior of the particular cell that is monitored, and not about a cell type due to the normal high variability even within a same cell culture. Numerous individual experiments must then be performed in order to afford a statistically significant and representative set of data. To reduce the number of experiments, alternative

2 ACS Paragon Plus Environment

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

strategies consist in performing measurements on a size-representative cell population rather than on a series of single cells. In a recent study,9 cells were preliminary loaded in four independent microwells and cultured on Pt / Pt black microband electrodes located at the bottom of each microwell. Each four working electrodes were poised at a distinct specific potential to monitor the release of one ROS / RNS after simultaneous stimulations of quite large cell populations. This investigation showed that reproducible and statistically significant quantifications of the four main ROS / RNS could be reached during oxidative stress from much less measurements than with the single-cell approach. Nevertheless, this basal detection implied with this configuration was suspected to (i) lose some apical releases of unstable RNS leading to lower detections and (ii) introduce some biases due to the fact that cells were placed in contact with the electrodes surface. Hence cell responses can be altered by their reaction to the production of oxidized ROS and RNS (OH., NO+, etc). In order to eliminate these drawbacks, a new approach was considered by taking advantage of multi-chamber microfluidic devices. Indeed, microfluidic systems have become of great interests for cell capture and isolation,13,14 cell manipulations,13,15-17 cell culture based assays,18-21 angiogenesis research,22 neurotransmission and tissue culture.23,24 Microfluidic channels provide new abilities to control parameters of cellular microenvironments at relevant length and time scales16,18,21,22,25,26 down to the single cell resolution.13,27,28 Using integrated sensors, many biomarkers13,27,29-33 secreted from cell populations including oxidative stress ones may then be investigated providing new insights into cellular functions. In the following, a specifically designed microfluidic platform is described for monitoring cellular oxidative stress under conditions that overcome potential drawbacks incurred in microwell experiments.9 This novel approach consists in separating the cell culture chamber from the detection area of ROS / RNS.25,34 Four parallel microfluidic channels are thus connected downstream to a same culture microchamber and one working electrode integrated in each channel for detecting one specific ROS / RNS. This microfluidic platform is expected to concentrate the species released in the microchamber into the microchannels and transfer them sufficiently rapidly to the detection area to avoid their chemical decomposition. The confined space of the microchannels and the hydrodynamic regime established over 3 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

the electrodes during ROS / RNS detections ensure high collection efficiencies and high analytical performances. Pt / Pt black channel electrodes for optimized H2O2, NO2−, NO• and ONOO− detections7,8 are used under controlled flow conditions. Macrophages are chosen as an important cellular model for oxidative stress owing to their crucial involvement in immune responses thank to abundant enzymatic pools enabling the generation of large amounts of ROS / RNS. They are stimulated with a calcium ionophore for fast, ubiquitous and reproducible activation of oxidative bursts. The results are discussed and compared with those obtained from single-cell experiments and cell populations in microwells.

EXPERIMENTAL SECTION

Materials for cell culture and platinum black deposition. All aqueous solutions were prepared from deionized water (Milli-Q, Millipore Corp.). Phosphate buffered saline solutions (PBS; pH 7.4; 0.140 M NaCl, 0.01 M Na2HPO4, and 2.68 mM KCl) were prepared by dissolving tablets (Sigma) in water and used for experiments with ROS / RNS. The calcium ionophore (A23187, Sigma) was diluted in BioUltra ethanol (Sigma) to prepare a 1 mM stock solution preserved in a freezer at -20°C. The LockeX1 buffer (LockeX1; pH 7.4; 5.6 mM glucose, 154 mM NaCl, 5.6 mM KCl, 15 mM Hepes, 1.2 mM MgCl2, 3.6 mM NaHCO3, 2.5 mM CaCl2) was used as physiological solution during electrochemical measurements. A solution of 50 µg mL-1 fibronectin (Sigma-Aldrich) was prepared in LockeX1 buffer and stored at 20°C in a sterile safe-lock tube (Eppendorf). Pt-black deposition solutions were prepared from 1 mL hydrogen hexachloroplatinate (IV) solution (8% wt in water; Sigma-Aldrich) and 1.6 mg lead (II) acetate trihydrate (99.8%; Sigma) added into 6.36 mL PBS. Cells culture. Murine macrophages RAW 264.7 (American Type Culture Collection) cell line was cultured at 37°C under a 5 % CO2 atmosphere in Dulbecco’s modified Eagle’s medium (DMEM) containing 1.0 g L-1 D-glucose and sodium pyruvate (Invitrogen). The medium was supplemented with 5% fetal bovine serum (Invitrogen) and 20 µg mL-1 gentamicin (Sigma). Almost confluent monolayers (8090%) of RAW 264.7 cells in flasks were re-suspended through trypsinisation, centrifugated at 1000 rpm, and diluted into 2 mL of medium (density of 107 cells mL-1). Before seeding cells in the microfluidic 4 ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

device, the entire device was UV-sterilized during 20 min and its culture compartment treated with 50 µg mL-1 fibronectin during 20 min to promote fast and firm cells attachment on the glass floor of the chamber and to allow long-term perfusion inside the microsystem. The cell suspension solution was driven into the device with an external syringe pump at 2 µL min-1 flow rate. After resting for 1 h under stop-flow condition in a cell incubator, macrophages showed a good adhesion before being perfused with a DMEM solution flowing at 4.0 µL min-1 to maintain constant nutrition replenishment and wastes removal. Electrochemical studies were performed after 12 to 24 h perfusion culture when the immobilized macrophages reached suitable density and viability. Same conditions were used for experiments with LName (N-nitro l-arginine methyl ester, Sigma Aldrich) used as the nitric oxide synthase inhibitor. In this case, macrophages were pre incubated with 1 mM L-Name in DMEM solution. In every experiments, the cell density was checked using a cell counting chamber (Malassez) under inverted microscopy (Observer D1, Zeiss). Before performing electrochemical measurements, the medium culture in the microsystem was replaced by pumping LockeX1 buffer. Stimulation was triggered by injecting in a continuous buffer flow an appropriate volume of 10 µM calcium ionophore A23187 solution through a peek sample loop (Rheodyne; 5 µL for stop-flow stimulation and 100 µL for non-stop-flow stimulation). The input volumetric flow rate was then fixed at 8 µL min-1 to maintain a relatively low hydrodynamic shear stress23,35,36 of about 0.6 dyn cm-2 in the cell chamber with a flow velocity of 0.53 mm s-1. At the same time, a higher flow velocity of 8.3 mm s-1 resulted in each downstream microchannel due to their reduced cross-section vs. that of the culture chamber. The microdevice was cleaned and reusable after biological measurements. Cells detachment was achieved

by

continuously

rinsing

the

microchamber

with

a

0.05%

trypsin-EDTA

(ethylenediaminetetraacetic acid) solution during at least 5 minutes at 4 µL min-1. The flow was imposed in the inverse direction of that used during measurements to facilitate the cells removal. Microsystems. The microsystems consisted in PDMS-glass hybrid microdevices specially designed to separate cell culture chamber from the electrochemical detection area. As described in Figure 1, it comprised upstream one microchannel (2.8 mm length, 400 µm width, 50 µm height, 0.056 µL 5 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

total volume), a cell microchamber (10 mm length, 5 mm width, 50 µm height, 2.5 µL total volume) connected downstream to four identical microchannels (4.5 mm length, 200 µm width, 20 µm height, 0.018 µL total volume). The volume of the microchannels and their connections was 0.25 µL, i.e., 1 % of that of the microchamber. These channels were designed to selectively detect the four ROS and RNS primary species amperometrically using four different potentials at the same time (one potential per channel). The upper layer was made by casting a 10:1 mixture (polymer-to-curing agent ratio) of polydimethylsiloxane (PDMS, RTV-615; Momentive Performance Materials) onto a SU-8 patterned mold. The under layer was a glass substrate on which several microband electrodes were patterned by soft lithography and deposited in a sputtering coater (K675XD; Emitech). The top-view design is reported in Figure 1B. Four sets of paired platinized working electrodes WE1-4 (WE1 to WE4, 197 µm width) and WE’1-4 (WE1’ to WE4’, 197 µm width) were implemented (a pair per channel). WE1-4 and WE’1-4 were located 3.4 mm and 4.3 mm downstream from the cell chamber respectively. Four counter electrodes (CE1 to CE4, 300 µm width, one per channel) were in Ti/Pt with 8 nm/25 nm respective thickness. Four reference electrodes (REF1 to REF4, 200 µm width, one per channel) were in Ti/Pt/Ag with 8 nm/25 nm/50 nm respective thickness. The WEs were further platinized before pasting the PDMS under layer through reduction of Pt (IV) according to a procedure previously optimized.7,8 A constant current density of 8 mAcm-2 was applied to each WE immersed in a PtCl62- solution so as to reach a charge density of 240 mC cm-2 for adequate sensitivity and robustness towards the detection of the four ROS / RNS species. Ag/AgCl reference electrodes were finally produced by oxidizing the Ag upper layer by 5 mM FeCl3. The two parts of the microdevice were aligned with precision as depicted in Figure 1A and pasted irreversibly after oxygen plasma treatment (Harrick). Since the microchannels were positioned perpendicularly to the WEs microbands, the effective WEs lengths were delimited by the microchannel widths. Only one set of the WEs pairs was connected during one experiment (WE1-4 or WE’1-4). Because the upstream chamber and downstream microchannels had not the same height, the SU-8 mold consisted of a multi-layer microstructure generated by using two masks. Within the chamber mold, circle structures were incorporated (i) sparsely as support pillars (100 µm diameter) to prevent any 6 ACS Paragon Plus Environment

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

subsidence of the PDMS chamber ceiling and (ii) along a dense row at the downstream end of the culture chamber to act as a barrier (pillars of 50 µm diameter; 10 µm gap) preventing cell access to downstream channels. This two-layer structure was transferred to silicon wafer after two-step photolithography involving an accurate alignment of both parts. The mold of the chamber was constituted with SU-8 3050 while that of the thinner channels was made by SU-8 2015. The whole mold was baked at 150°C during 30 min to complete the joint interface. After fabrication, the microstructures were checked optically under microscope. Electrochemical measurements. Experiments were performed at room temperature using a multipotentiostat (VSP-300 Multi Channels Potentiostat, Biologic). In each microchannel, a threeelectrode configuration was used by connecting one REF, one of the two WEs and one CE (as sequentially-localized along each microchannel according to the flow direction) to one of the four channel boards of the multipotentiostat. The flow was pressure-driven by a syringe pump (Pump 11 Elite; Harvard Apparatus). The device was preliminary tested and its electrochemical stability assessed by recording simultaneously anodic voltammograms at the four WEs with 0.1 mM flowing H2O2 solution. Highly comparable responses for all WEs were obtained demonstrating high reproducibility, stable potentials supported by the REFs located at each channel entry and identical convective transport (2 µL min-1) within each microchannel. After microfabrication of the device and between experiments, the WEs were cleaned and activated by applying alternative potential pulses (+0.2 V/REF, 1s; -0.5 V/REF, 1s; 30 cycles). This procedure was optimized to activate the electrode and to eliminate surface biofouling during experiments. It allowed the initial electrode performance to be retrieved after little surface passivation during the fibronectin pretreatment. The electrodes were subsequently polarized for 1 minute at their operating potential (see below) to stabilize the background current before any measurements. During macrophages stimulation, chronoamperometric experiments were performed simultaneously in each channel by applying one potential (300, 450, 620 or 850 mV vs. Ag/AgCl) to each WE to allow monitoring H2O2,

7 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

ONOO-, NO• and NO2- in the downstream flows according to the potential. Each electrode response was preliminary calibrated after baseline correction as described in our previous works.7,8 Stimulation procedures. Two stimulation modes were considered with continuous-flow and stop-flow stimulations involving 10 µM injection of calcium ionophore A23187. In continuous-flow mode, the content of the sample loop (100 µL of A23187) was delivered at the inlet of the microdevice at constant flow rate during the time required (at least 10 min) to detect a complete oxidative stress response from the population of macrophages held in the chamber. In stop-flow mode, the content of the sample loop (5 µL of A23187) was first delivered to fill the microchamber. Then, the flow was stopped during 10 min and was restarted after this time delay to push the accumulated ROS / RNS species from the microchamber towards the microchannels by a continuous flow of buffer.

RESULTS AND DISCUSSION

This study was initiated taking advantage of the knowledge developed during previous investigations carried out on the simultaneous electrochemical detections of the four important ROS and RNS species released from cells in response to oxidative stress.4-6 The original design of the microdevice described in Figure 1 provided important benefits versus previous ones reported by our group with the aim of offering an alternative and complementary methodology to single-cell measurements.9 The concept was based on a non-invasive electrochemical detection for fast and reliable determinations of H2O2, ONOO-, NO• and NO2- through their oxidation. Microfluidic conditions are expected to remove any bias related to the detection. The microdevice comprised an upstream microchamber designed to immobilize a population of living cells and connected downstream to four parallel microchannels integrating Pt-black microband working electrodes. These electrodes were optimized to reach high performances for detecting selectively the four ROS and RNS primary species in their mixtures.7-9 Full details about device fabrication and cell culture are reported in experimental section. The device geometry and flow velocity were chosen in order to detect reactive species very close from their production sites, under physiological conditions and with no shear stress applied onto the cells cultured in the microchamber. The size of the 8 ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

culture microchamber was defined according to the cell density (about 105 cells cm-2) and minimal number of cells (about 3.5 104 cells) required to detect statistically significant production of ROS / RNS (ca. 5 to 10 fmol per cell and per species) after stimulation by a calcium ionophore. Downstream concentrations were expected to range between 0.1 and 0.2 mM. As indicated in the experimental section and detailed below, two stimulation procedures were used and tested. Cell culture. The protocol was adapted from our former study9 to ensure fast and robust cell adhesion in the microchamber (see experimental section). Before seeding cells, the entire device was UVsterilized and treated by fibronectin. After seeding the cells were perfused with a constant nutrition solution flow. This also allowed wastes removal. Flow rate for cell culture was sufficient to provide effective transport of growth medium while imposing shear stress below critical values that might have caused cell damages. As shown in Figure 2, macrophages displayed good morphology and brightness after 24 h perfusion, indicating adequate viability. Under these conditions, long-term cell culture could be achieved with cell proliferation and confluence all over the microchamber floor. However, cell culture was interrupted after 12 to 24 h perfusion to afford a homogeneous cell distributions with all densities suitable for experiments. Indeed, 48 h to 96 h perfusion led to cell aggregation preventing measurements of biologically significant data for macrophages (Figure 2A). This value was estimated from average cells density derived from at least 20 cells culture assays. A mean cell number of ca. 3.5 104 cells was determined from an average cell density of (0.7 ± 0.2) 105 cells cm-2. The 10 µm gap distance between micropillars at the output of the microchamber allowed an effective confinement of macrophages within the microchamber preventing cells proliferation towards the detection area during the culture step (Figure 2B). In absence of micropillars, it was noted that cells tended to aggregate at the entry of microchannels due to the local volume reduction (data not shown). Experiments based on a continuous-flow stimulation. Using this procedure, stimulation of macrophages by calcium ionophore and electrochemical detection took place simultaneously. Under these experimental conditions, the mixtures of ROS and RNS produced by the cells were delivered from the output of the microchamber to each of the downstream electrodes with a transition delay estimated to be 9 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

less than 1 s, i.e., sufficiently small to avoid significant decomposition of ONOO- and reaction of NO● with the aerated medium. In each experiment, independent but simultaneous chronoamperometric responses were recorded in each channel using one of the pair of working electrodes while a continuous flow of 10 µM calcium ionophore was applied at the inlet of the chamber. An example of typical oxidative stress responses is given in Figure 3A during one stimulation. The signals monitored at each electrode poised at a different selected potential featured similar shapes but were different in magnitude. These responses were indicative of well-defined release kinetics.9,10 A fast current rise was achieved within 3 minutes at each electrode, evidencing the rapid activation of cells enzymatic systems and fast release of ROS / RNS cocktail. This release lasted during ca. 10 minutes before the currents returned to their initial baseline level. It was possible to evaluate the individual contributions of each species. At first glance, the quantity of each species Nj produced at each time is deduced from the following equation:  =  

(1)

where [Cj] is the local concentration of species j in the flowing solution at a given instant and Vplug the volume of flowing solution containing the species (plug of solution). Cj can be readily converted into corresponding current intensity ij based on a pre-calibration between ij and Cj under the same experimental conditions. The pre-calibration allowed to account for the outcome of mass transport regime established at the microchannel electrode. Thus, if Sj is the slope resulting from the calibration curve at the flow rate uflow, the quantity Nj is directly given by: 

 = 4       

(2)

The factor 4 refers to the number of microchannels in which the total quantity Nj leaving the chamber is dispatched. It must be underlined that the current ij is relative to the contribution of one single ROS or RNS species j that had to be derived from the overall current iE monitored at potential E. In some previous studies,7,8 we have carefully established the contributions of each of the four species H2O2, ONOO-, NO• and NO2- at each potential (300, 450, 620 and 850 mV vs Ag/AgCl) individually applied to microchannel electrodes (Eqs. 3 to 6): 10 ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

 = . ! − .#$ !

(3)

• = .#$ ! − .& !

(4)

 = 2.02 .& ! − 2.38 .+ !

(5)

, - = 2.38 .+ . − 1.02 .& .

(6)

Pre calibration curves were established under the same hydrodynamic conditions (electrode width 200 µm, channel height 20 µm and flow rate 2 µL min-1) leading to slopes equal to 0, - = 2.41 10-3 A M-1; 0-1-- = 1.04 10-3 A M-1; 01-• = 1.13 10-3 A M-1 and 01- = 2.22 10-3 A M-1. Thus, the corresponding mean quantities evaluated per cell could be easily evaluated under these specific experimental conditions by introducing these values into Eq. 2 and by taking into account the number of cells estimated in the microchamber (see experimental section). The average overall quantities of RNS / ROS released per living cell in the microchamber during one stimulation were respectively: 5.1 ± 0.8 fmol H2O2, 6.4 ± 1.8 fmol ONOO–, 7.9 ± 1.7 fmol NO● and 6.8 ± 1.1 fmol NO2-. Figures 3B and 3C show the time dependence of the proportion of the four species along the entire process. Error bars in these figures correspond to five different batches of cells. In Figure 3B, stable kinetic fluxes ratios were derived evidencing the constant composition of oxidative bursts, as well as the reliability of the determinations using this microfluidic device and stimulation procedure. Concentrations corresponding to the data in Figure 3B were reported in Figure 3C displaying concentration variation of the species in the micromolar range with NO● predominance. These important results were further confirmed by measuring the outcome of three successive stimulations from a same population of macrophages (Figure 4) separated by a 10 min time delay between each stimulation. Within the accuracy of the measurements, all curves almost superimposed demonstrating the reproducibility of the experiments. At this level of investigation, no obvious perturbation in cells behavior or biological process could be observed so that an average behavior could then be readily afforded based on only one experiment. All the data demonstrated concomitantly the benefits resulting from the procedure with accurate and highly reproducible performances of the microchannel electrodes. Note here that the current

11 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

magnitude recorded (a few tens of nanoamperes, see Figure 3A) corresponded to a concentration of species released in the chamber volume being in the micromolar range. Experiments based on a stop-flow stimulation. A second stimulation procedure with two successive steps was tested. During the first step, macrophages were stimulated by a static 10 µM calcium ionophore solution for 10 min in order to achieve complete delivery of ROS / RNS. This was followed by a second step during which a continuous flow of buffer was imposed in order to elute the released species that accumulated within the microchamber towards the microchannels electrodes. Typical oxidative stress responses recorded at each microchannel electrode are presented in Figure 5A after a first stimulation. As expected, much higher faradic currents with shorter durations were monitored as compared to the continuous-flow procedure. Indeed, during the stimulation period in static solution, all electroactive species released by cells accumulate in the chamber leading to higher local concentrations (up to 10 times those estimated using the continuous-flow procedure). The duration of the responses is determined mostly by the time required to elute the total volume of the microchamber. At 8 µL min-1 flow rate, the experimental average signal duration was about 50 s being in agreement with the 20 s theoretical duration required to elute the total volume of the microchamber. This difference is coherent with the slight expected dispersion of the RNS /ROS species into the microchamber that imposes a residence time during elution. The mean quantities of species released per living cell could be estimated from several batches of cell population (experiments operated from only one stimulation). Following the same calculations, the corresponding values were respectively: 4.1 ± 0.9 fmol H2O2, 2.5 ± 1.4 fmol ONOO–, 3.0 ± 1.5 fmol NO● and 10.2 ± 1.2 fmol NO2- (see below for comparison).The main surprising difference from the continuousflow procedure was observed during successive stimulations. A drastic decrease of current magnitude by a factor of 5 was observed after the second response even after performing a cleaning step for electrodes between the measurements (Figure 5B). This points out the fact that the accumulated species within the microchamber during the static step influenced the normal cellular metabolism, impacting the release of ROS / RNS species. Furthermore, the experiments were not as reproducible as in the previous procedure. When compared to the continuous-flow procedure these data evidenced that as expected both ONOO- and 12 ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

NO● suffered some decomposition during the static step, presumably into NO2- (whose quantity increased) and NO3- which is not detectable. Average production of the primary ROS and RNS species. The average productions of H2O2, ONOO-, NO● and NO2- based on the two stimulation procedures are compared in Figure 6 as well as to previous data based on two other configurations. They correspond to experiments performed under similar biological conditions using macrophages Raw 264.7 stimulated by addition of the same calcium ionophore (10 µM A23187 solution). One of these configurations is the so-called artificial synaptic configuration based on single-cell measurements5 and the other is the basal detection from wells containing cell population.9 All these data are average values that have same statistical relevance. They resulted from new determinations with different numbers of experiments. This comparison shows that the synaptic configuration / single-cell measurements provided the best collection efficiency. Though the microfluidic configurations reported ROS and RNS quantities per living cell of same orders of magnitude than the apical or basal measurements (Figure 6A), they exhibited obvious difficulties in detecting the short-lived species ONOO- and NO● that, under our conditions, have half-life times of 1 s and a few seconds respectively. Indeed, the delay needed for the eluting solution to cross over the whole chamber length is at minimum 20 s. This can also explain the differences between the two stimulation procedures performed under microfluidic conditions. In stop-flow stimulation, the time delay (10 min) was undoubtedly too long as compared to the short life-times of ONOO– and NO●. Under these conditions, reaction of NO● with O2 (about 0.25 mM O2 solubility in aerobic solution) led to an increased detected quantity of NO2- and spontaneous decomposition of ONOO– into NO2- and possibly NO3-. These competitive reactions explain why higher NO2- but lower ONOO– and NO● levels were determined in stop-flow stimulation as compared to continuous-flow stimulation. This comment is also valid for comparison with the synaptic configuration. In basal detection, the lower collection of ONOOand NO● might be ascribed both to adhesion of cells on working electrodes and to loss by apical diffusion. It can be due also to possible blasting of the whole cells basal poles by products of ROS and RNS oxidation. 13 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

Therefore, an underestimation of the relative composition of oxidative bursts are inevitable under microfluidic conditions (Figure 6B). Despite some deviations, quantifications of ROS and RNS are still relevant. In particular the peak-shaped amperometric responses in continuous-flow stimulation characterized very well the kinetics of oxidative-stress release as in synaptic configuration. Disadvantage of the microfluidic procedures is compensated by far by reproducible and statistically significant results with only one or a few measurements due to a large number of cells involved. Influence of drugs or chemicals can be easily investigated. As proof of concept of the method, we wish to report the analysis of the influence of a NO synthase inhibitor (L-Name) during the perfusion culture. In this case, macrophages were pre incubated with 1 mM L-Name solution in DMEM during 12 to 18 h. L-NAME is well known to act as a prodrug leading to selective NO synthase inhibitory activity. After perfusion, the macrophages presented no obvious difference under optical observation compared to cells cultured without L-Name. However, after stimulation of macrophages by calcium ionophore, electrochemical measurements performed under continuous-flow mode showed a drastic decrease of ONOO-, NO● and NO2- releases with a simultaneous increase of H2O2 (Figure 6C). H2O2 production that results mainly from superoxide dismutation resulted higher since much less NO● was available to react with superoxide under these conditions. The amount of H2O2 increased by 30% which was in good agreement with expected compositions of cellular oxidative bursts.4 One must underline here that only 5 experiments were sufficient to obtain statistically significant quantitative results with cells population using the continuousflow stimulation under microfluidic conditions.

CONCLUSION

We established that simultaneous amperometric determinations of the four key ROS and RNS, H2O2, ONOO−, NO• and NO2− were achievable during oxidative bursts using an innovative microfluidic platform integrating both a cell microchamber and microchannel electrodes. The overall experiments implementing cell culture, stimulation and electrochemical detection were successively achieved using a 14 ACS Paragon Plus Environment

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

same microdevice. The design of the platform led to effective and non-invasive detections of ROS and RNS from a large population of cells. Separating emitting cells from detection area seems a relevant approach to avoid any perturbations of cell behaviors by electrode operations. In comparison to the socalled artificial synaptic configuration, collection efficiencies of short-lived species NO• and ONOO− were slightly lower due to their kinetic decomposition. However, highly reproducible and relevant measurements could be obtained statistically from only few experiments. These results validated successfully the proof of concept of this microfluidic platform which is aimed at getting rapid insights about biologically-relevant effects of oxidative stress on average cellular behaviors. In the future, better performances will be expected by optimizing the geometry and hydrodynamic conditions. Indeed, the coupling between confined space and flow rate is definitely the strategy to shorten the delivery time towards the channel electrodes in order to improve the overall performances. Other operating conditions can be also envisaged to mimic natural perfusion with appropriate tuning of flowing modes adapted to different cellular events. Finally, the design can be further improved by incorporating chemicallymodified electrodes. This preliminary study paves the way to a very broad field of bioanalytical applications for high-throughput monitoring of cells behaviors.

Acknowledgments

This work was supported in parts by CNRS UMR 8640, ENS (Ecole normale superieure), PSL Université and Sorbonne Université as well. Y.L. thanks the Chinese Scholarship Council for her financial support with Ph. D. grant. C.A. thanks the support by LIA CNRS NanoBioCatEchem and by ANR grant “ChemCatNanoTech” (ANR-AAP-CE06).

15 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

References

(1)

Kohen, R.; Nyska, A. Toxicol Pathol 2002, 30, 620-50.

(2)

Steven, S.; Daiber, A.; Dopheide, J. F.; Munzel, T.; Espinola-Klein, C. Redox Biol 2017, 12, 787797.

(3)

Winterbourn, C. C. Biochimica et biophysica acta 2014, 1840, 730-8.

(4)

Amatore, C.; Arbault, S.; Bruce, D.; de Oliveira, P.; Erard, M.; Sojic, N.; Vuillaume, M. Analusis 2000, 28, 506-517.

(5)

Amatore, C.; Arbault, S.; Guille, M.; Lemaıtre, F. Chem. Rev. 2008, 108, 2585-2621.

(6)

Amatore, C.; Arbault, S.; Koh, A. C. W. Anal. Chem. 2010, 82, 1411-1419.

(7)

Li, Y.; Sella, C.; Lemaître, F.; Guille Collignon, M.; Thouin, L.; Amatore, C. Electroanalysis 2013, 25, 895-902.

(8)

Li, Y.; Sella, C.; Lemaître, F.; Guille-Collignon, M.; Thouin, L.; Amatore, C. Electrochim. Acta 2014, 144, 111-118.

(9)

Li, Y.; Meunier, A.; Fulcrand, R.; Sella, C.; Amatore, C.; Thouin, L.; Lemaître, F.; GuilleCollignon, M. Electroanalysis 2016, 28, 1865-1872.

(10)

Amatore, C.; Arbault, S.; Chen, Y.; Crozatier, C.; Tapsoba, I. Lab Chip 2007, 7, 233-8.

(11)

Bedioui, F.; Quinton, D.; Griveau, S.; Nyokong, T. Phys Chem Chem Phys 2010, 12, 9976-88.

(12)

Quinton, D.; Girard, A.; Thi Kim, L. T.; Raimbault, V.; Griscom, L.; Razan, F.; Griveau, S.; Bedioui, F. Lab Chip 2011, 11, 1342-50.

(13)

Armbrecht, L.; Dittrich, P. S. Anal Chem 2017, 89, 2-21.

(14)

Valihrach, L.; Androvic, P.; Kubista, M. Int J Mol Sci 2018, 19.

(15)

Whitesides, G. M. Nature 2006, 442, 368-73.

(16)

Velve-Casquillas, G.; Le Berre, M.; Piel, M.; Tran, P. T. Nano Today 2010, 5, 28-47.

(17)

Gao, C.; Sun, X.; Gillis, K. D. Biomed. Microdevices 2013, 15, 445-51.

16 ACS Paragon Plus Environment

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

(18)

Paguirigan, A. L.; Beebe, D. J. Integr. Biol. 2009, 1, 182-95.

(19)

Halldorsson, S.; Lucumi, E.; Gomez-Sjoberg, R.; Fleming, R. M. Biosens. Bioelectron. 2015, 63, 218-31.

(20)

Jiang, H.; Jiang, D.; Zhu, P.; Pi, F.; Ji, J.; Sun, C.; Sun, J.; Sun, X. Biosens. Bioelectron. 2016, 83, 126-33.

(21)

Zuchowska, A.; Kwapiszewska, K.; Chudy, M.; Dybko, A.; Brzozka, Z. Electrophoresis 2017, 38, 1206-1216.

(22)

Young, E. W. J. Lab. Autom. 2013, 18, 427-36.

(23)

Leclerc, E.; David, B.; Griscom, L.; Lepioufle, B.; Fujii, T.; Layrolle, P.; Legallaisa, C. Biomaterials 2006, 27, 586-95.

(24)

Miura, T.; Yokokawa, R. Develop. Growth Differ. 2016, 58, 505-15.

(25)

Young, E. W.; Beebe, D. J. Chem. Soc. Rev. 2010, 39, 1036-48.

(26)

Riahi, R.; Shaegh, S. A.; Ghaderi, M.; Zhang, Y. S.; Shin, S. R.; Aleman, J.; Massa, S.; Kim, D.; Dokmeci, M. R.; Khademhosseini, A. Sci. Rep. 2016, 6, 24598.

(27)

Fan, Y.; Dong, D.; Li, Q.; Si, H.; Pei, H.; Li, L.; Tang, B. Lab Chip 2018, 18, 1151-1173.

(28)

Shukla, V. C.; Kuang, T. R.; Senthilvelan, A.; Higuita-Castro, N.; Duarte-Sanmiguel, S.; Ghadiali, S. N.; Gallego-Perez, D. Trends Biotechnol 2018, 36, 549-561.

(29)

Wang, W.; Zhang, S. H.; Li, L. M.; Wang, Z. L.; Cheng, J. K.; Huang, W. H. Anal Bioanal Chem 2009, 394, 17-32.

(30)

Ges, I. A.; Currie, K. P.; Baudenbacher, F. Biosens. Bioelectron. 2012, 34, 30-6.

(31)

Ges, I. A.; Brindley, R. L.; Currie, K. P.; Baudenbacher, F. J. Lab Chip 2013, 13, 4663-73.

(32)

Vlaski-Lafarge, M.; Ivanovic, Z. J. Cell Sci. 2015, 128, 3849-60.

(33)

Shin, S. R.; Kilic, T.; Zhang, Y. S.; Avci, H.; Hu, N.; Kim, D.; Branco, C.; Aleman, J.; Massa, S.; Silvestri, A.; Kang, J.; Desalvo, A.; Hussaini, M. A.; Chae, S. K.; Polini, A.; Bhise, N.; Hussain, M. A.; Lee, H.; Dokmeci, M. R.; Khademhosseini, A. Adv Sci (Weinh) 2017, 4, 1600522.

(34)

Fan, X.; Jia, C.; Yang, J.; Li, G.; Mao, H.; Jin, Q.; Zhao, J. Biosens. Bioelectron. 2015, 71, 380-6. 17 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

(35)

Kim, L.; Toh, Y. C.; Voldman, J.; Yu, H. Lab Chip 2007, 7, 681-94.

(36)

Yeh, C. C.; Chang, S. F.; Huang, T. Y.; Chang, H. I.; Kuo, H. C.; Wu, Y. C.; Hsieh, C. H.; Shi, C. S.; Chen, C. N. Arthritis Res Ther 2013, 15, R53.

18 ACS Paragon Plus Environment

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 1: A) 3D scheme of the microfluidic platform integrating upstream a microchamber for cell culture and downstream 4 parallel channels with microband electrodes. B) Top view of the microdevice with its microfluidic circuit and electrode paths for electrical connection.

19 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

Figure 2: Optical images of cell cultures of Raw 264.7 macrophages inside the microchamber. A) Views after 4 h, 24 h, 60 h and 96 h continuous medium perfusion. B) View after 24 h perfusion in the proximity of the micropillars arrays at the exit of the culture chamber.

20 ACS Paragon Plus Environment

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 3: A) Amperometric responses monitored at the four working electrodes during continuous-flow stimulation of raw 264.7 macrophage population by 10 µm A23187 calcium ionophore solution. Curves recorded respectively at 0.3, 0.45, 0.62 and 0.85 V vs/ Ag/AgCl. The black arrow indicates the time when the A23187 solution is pumped across the microchamber. Flow rate 2 µL min-1 in each channel. B) Flux ratios of the four detected species during continuous-flow stimulation. C) Corresponding concentration variations. In B and C, average values and standard deviations were evaluated from 5 experiments (5 different batches of cells). 21 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

Figure 4: Amperometric responses monitored during 3 successive continuous-flow stimulations of a same macrophage population. A 10 min time interval was set between each stimulation to rest the activated cells. Other conditions were similar to Figure 3.

22 ACS Paragon Plus Environment

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 5: Amperometric responses monitored during 2 successive stop-flow stimulations of a same macrophage population. Responses during the first (A) and second (B) stimulation. The black arrows indicate the time when the flow is restarted after a 10 min stimulation under static solution of calcium ionophore. A 10 min time interval was set between the two stimulations to rest the activated cells. Other conditions were similar to Figure 3. 23 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

Figure 6: Comparison between quantities of primary ROS and RNS species collected during one stimulation in single-cell detection (SC), cells population in basal detection (BD), continuous-flow mode (CF) and stop-flow mode (SF). A) Average quantities in SC, BD, CF and SF. B) Total quantities collected in each configuration from data in A. C) Influence of L-Name pre incubation on average quantities in CF. The figures on the graph indicate the differences in N values per living cell after L-Name pre incubation. ttests, p < 0.05 (H2O2), p < 0.001 (ONOO-), p < 0.001 (NO•) and p < 0.001 (NO2-). In A to C, average values were calculated from 20 to 30 experiments in SC and 5 experiments in BD, CF and SF.

24 ACS Paragon Plus Environment

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

For TOC only

25 ACS Paragon Plus Environment