Real-Time Detection of Nitric Oxide Release in Live Cells Utilizing Fluorinated Xerogel-Derived Nitric Oxide Sensor

Nitric oxide (NO) is an important signaling molecule that regulates a diverse range of physiological and cellular processes in many tissues. Therefore, the accurate de‐ tection of physiological NO concentration is crucial to the understanding of NO sig‐ naling and its biological role. There has been growing interest in the development of electrochemical sensors for direct and real-time monitoring of NO. As the direct electrooxidation of NO requires a relatively high working potential, further surface modification with permselective membranes is required to achieve the desired se‐ lectivity for NO via size exclusion or electrostatic repulsion. Here we reported a pla‐ nar-type NO sensor with a fluorinated xerogel-derived gas permeable membrane for real-time detection of NO release in live cells. First, we evaluated the biocompat‐ ibility of xerogel-derived NO permeable membranes modified with fluorinated functional groups by growing RAW 264.7 macrophages on them. And we per‐ formed the AFM measurements to examine the morphology of RAW 264.7 macro‐ phages on xerogel membrane. Finally, we successfully detected NO release in RAW 264.7 macrophages, using a planar-type xerogel-derived NO sensor. As a result, flu‐ orinated xerogel-derived membrane could be utilized as both NO permeable and cell-adhesive membranes. Besides, planar-type xerogel-based NO sensors can be easily applied to the cellular sensing system, with a simple coating procedure.


Introduction
Nitric oxide (NO) is an important mediator that regulates a diverse range of physiological and cellular processes in many tissues [1,2]. In particular, it plays an important role in a variety of biological processes including neurotransmission, immune defense, regulation of cell death (apoptosis), and cell motility [3][4][5]. The accurate detection of physiological NO concentration is crucial to the understanding of NO signaling and its biological role. However, it is difficult to accurately measure NO levels in vivo because NO is present at nanomolar concentrations in the body and has a half-life of 2-6 s [6]. Additionally, it is highly reactive with numerous endogenous species including free radicals, peroxides, and oxygen [7,8]. Consequently, most methods for NO detection are indirect including spectroscopic approaches such as the Griess assay for nitrite, detection of nitrate and nitrite with reductase enzymes, and detection of methemoglobin after NO reaction with oxyhemoglobin [9][10][11]. Unfortunately, these methods often fail to accurately reflect the dynamics of NO in vivo and in real time [12]. Direct measurement strategies are therefore necessary for examining biological process and diseases related to NO in biological conditions, in particular the action of endogenously produced NO.
There has been growing interest in the development of electrochemical sensors for direct and real-time monitoring of NO. Electrochemical methods provide simplicity, fast response times, good sensitivity, high selectivity, and long-term calibration stability [8,12], and thus have significant advantages over other techniques such as electron paramagnetic resonance (EPR) [13][14][15], chemiluminescence [16][17][18], and fluorescence [19,20]. As the direct electrooxidation of NO requires a relatively high working potential, further surface modification with permselective membranes is required to achieve the desired selectivity for NO via size exclusion or electrostatic repulsion. Many polymeric materials have been reported as NO-permeable and permselective membranes, including polycarbazole [21], collodion [22], Nafion [23], cellulose acetate [24], poly(tetrafluoroethylene) [25], polydimethylsiloxane (silicone rubber) [26], and multilayer hybrids of these polymers. However, the utility of these polymeric membranes for improving sensor performance in vivo or in cells remains limited [8,12].
The sol-gel process is an inorganic polymerization process for the fabrication of materials under mild conditions [27] through the hydrolysis and condensation of suitable alkoxysilane precursors [28]. These sol-gel derived materials (xerogels) have emerged as a class of materials suitable for a wide range of sensing applications [8]. Organically modified xerogels are particularly advantageous for gas sensing applications [29]. The porous inorganic network provides an open and rigid structure for rapid diffusion of gaseous molecules, while the organic groups impart hydrophobicity to the membrane for preventing leakage of the internal electrolyte [30,31]. A previous study reported that sol-gel derived electrochemical sensors showed good sensitivity and selectivity for NO detection in vitro [21,30]. These studies were mainly focused on improving sensor performance such as sensitivity and selectivity, in particular optimization of permselective membrane in vitro.
An amperometric NO sensor for cellular biosensing is significantly affected by the distance between the sensor and the NO-producing cells because NO is rapidly diluted by diffusion under cell-culture conditions [32,33]. However, it is difficult to adjust the exact distance between the needle or disk type sensors and cells without special apparatus [34]. Therefore, it would be better to grow the cells directly on the electrode surface with a NO permeable membrane. Our previous study reported good sensitivity and selectivity of electrochemical NO sensors with a xerogel membrane in standard solution, but it was mainly focused on membrane optimization including composition, permeability, and contact angle to improve the sensor performance [8,12].
In this study, we developed a planar-type NO sensor with a fluorinated xerogel-derived gas permeable membrane for real-time detection of NO release in live cells including RAW 264.7 macrophages. First, we evaluated the cell viability of RAW 264.7 macrophages on xerogelderived NO permeable membranes. And we performed the AFM measurement to examine the morphology of RAW 264.7 macrophages on xerogel membrane. Finally, we performed the real-time electrochemical detection of NO release in RAW 264.7 macrophages by lipopolysaccharide (LPS) stimulation utilizing a planar-type NO sensor.

Preparation and biocompatibility test of fluorinated xerogel membrane
A silane solution was prepared by dissolving 18 μL of MTMOS and 4.5 μL of 17FTMS in 727.5 μL of ethanol, as previously reported [12]. The solution was mixed with 160 μL of water, followed by 10μL of HCl (0.5 M) for 1 h. The ensuing solution was deposited on a culture dish and cured for 1 day under ambient conditions. RAW 264.7 macrophages were cultured in Dulbecco's modified Eagle's medium (DMEM) media with 10% fetal bovine serum (FBS), 1% penicillin, and 1% streptomycin. The cells were then maintained in a humidified atmosphere that contained 5% CO 2 at 37 °C and the medium was changed every second day. Cells were detached by washing and scraping. After washing, cells were resuspended in fresh medium and used for subsequent experiments.
To evaluate the biocompatibility of xerogel membrane, we performed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma, St.Louis, MO, USA) using RAW 264.7 macrophages. The MTT assay, which is one of the most versatile and popular methods to measure cell viability, involves the conversion of water soluble MTT to an insoluble formazan. The amount of formazan in the MTT assay reflects the total mitochondrial enzymatic activity of live cells. Cells (1 × 10 4 cells per well) were seeded in a xerogel membrane coated 96-well plate and incubated for 24, 48, and 72 h. After MTT was added to each group, the cells were incubated for 4 h. The viability was then measured with a microplate reader (Bio-Tek, Winooski, VT, USA) at a test wavelength of 540 nm. The optical density (OD) was calculated as the difference between the reference wavelength and the test wavelength.

AFM measurement of cells on fluorinated xerogel membrane
RAW 264.7 macrophages on xerogel membrane were characterized using atomic force microscopy (AFM). RAW 264.7 macrophages were washed twice with filtered phosphate buffered saline (PBS, 0.1 M, pH 7.4) and fixed for 20 min in 2.5% glutaraldehyde in PBS at room temperature and 5 ml PBS was added to xerogel-modified culture dish containing fixed cells. Contact mode AFM images were obtained using a NANOS N8 NEOS (Bruker, Herzogenrath, Germany), equipped with a 42.5 μm XY/4 μm Z scanner and a Zeiss optical microscope (Epiplan 500x). The AFM was placed on an active vibration isolation table (TS-150, S.I.S., Herzogenrath, Germany), inside a passive vibration isolation table (Pucotech, Seoul, Korea) to eliminate external noise. Data acquisition and image processing were performed with SPIP (Scanning Probe Image Processor Version 4.1, Image Metrology, Denmark).
Silicon cantilevers with the reflective side coated with gold (Budget Sensor, Bulgaria) were used for the measurements under liquid conditions. The AFM probe tips were stabilized with PBS for at least 10 min prior to scanning. Fixed RAW macrophages were scanned in PBS solution at a resolution of 512 × 512 pixels, at a scan speed of 0.5 line/s.

Fabrication of planar-type NO sensor
The planar-type NO sensor with three-configuration (i.e., working, reference, and counter electrodes) was prepared using a screen-printing technique, as shown in Figure 1(A). Briefly, a polyester substrate (500 μm thick; 3 M, St. Paul, MN, USA) was thermally treated at 130 °C for 3 h before screen-printing. The silver (for reference electrode and electrical connections), carbon (for working and counter electrodes), and dielectric insulator patterns were sequentially screen-printed onto the thermally pretreated polyester substrate using a LS-150 semiautomated screen printer (Newlong Seimitsu Kogyo; Tokyo, Japan). Each pattern was cured at 130 °C for 13 min. The carbon working electrode (7.1 mm 2 ) was platinized in 3% chloroplatinic acid hexahydrate and 0.03% lead(II) acetate trihydrate (w/w in water) by cycling the potential from +0.6 to −0.35 V (vs. Ag/AgCl) at a scan rate of 20 mV/s using a CH Instruments 760B bipotentiostat (Austin, TX, USA). The Ag/AgCl reference electrode was prepared by oxidative treatment using 0.3 M FeCl 3 solution on the dried silver paste for 15 minutes. A silane solution for xerogel membrane was deposited onto platinized carbon working electrode and allowed to cure for 24 h under ambient conditions.

Evaluation of planar-type NO sensors
To evaluate the analytical performance of the planar-type NO sensor, amperometric measurements were performed using a CH Instruments 760B bipotentiostat or a portable MicroStat (Dropsens, Spain). The sensor was prepolarized for 100 min and tested in DMEM solution under ambient conditions with constant stirring. Currents were recorded at an applied potential of +0.8 V (vs. Ag/AgCl). For in vitro calibration, a NO standard solution (1.9 mM) was prepared by purging DMEM solution with N 2 gas for 30 min to remove any oxygen, followed by purging with NO (99.5%) for 30 min. The CO solution (0.9 mM) was similarly prepared by consecutively purging DMEM with N 2 for 30 min and CO (99.5%) for another 30 min. (Caution! The NO and CO purging process must be carried out in a fume hood since NO and CO gases are toxic!) The solutions of NO and interfering species (nitrite, ascorbic acid, uric acid, acetaminophen, dopamine, hydrogen peroxide, and carbon monoxide) were freshly prepared every second day and stored at 4 °C.
The selectivity of the xerogel-modified sensors for NO in the presence of interfering species was also evaluated using the separate solution method [12], for which the amperometric selectivity coefficients (logK NO , j amp ) were calculated using the following equation: where ∆I NO and ∆I j are the measured current values for the target analyte (NO) and interfering species (j = nitrite, ascorbic acid, uric acid, acetaminophen, dopamine, hydrogen peroxide, and carbon monoxide), respectively. The concentration of each interfering substance (c j ) was selected to be 1 μM, the same concentration of NO (c NO ).

Real-time detection of NO release in live cells
RAW 264.7 macrophages (1×10 4 cells/electrode) were grown on top of the electrode surface which was covered with sol-gel-derived membrane. Stimulation of RAW macrophages with LPS (1 μg/ml) was performed after stabilizing the background current for 2∼3 h at 37 °C. The experimental scheme is shown in Figure 1(B).

Cytotoxicity assessment of fluorinated xerogel membrane
In general, amorphous, hydrophobic fluoropolymers are highly permeable to many gaseous species including oxygen, nitric oxide, carbon dioxide, ammonia, and volatile organic compounds [12,35]. Indeed, a microporous poly(tetrafluoroethylene) (PTFE) film has been extensively employed to improve a Clark-type oxygen sensor as a gas-permeable membrane [36,37]. The PTFE membrane effectively prevented the biofouling at the interface between the sensor surface and the biological samples (e.g., blood), while enabling high permeation and rapid diffusion of oxygen [36]. However, the fabrication of such sensors remains complicated. The PTFE film must be stretched across the electrode and fixed to the sensor. This process is often irreproducible, impacting both the thickness of the gas-permeable membrane and the sensor's response to oxygen [29,36].
In this study, we employed a fluorinated xerogel gas-permeable membrane to take advantage of both the versatility of sol-gel chemistry and high and fast gas permselectivity like PTFE. Our previous study reported good sensitivity and selectivity of electrochemical NO sensors with a xerogel membrane in standard solution [8,12]. In addition, we had preliminarily evaluated the cell viability, adhesion, and growth of cells on xerogel-derived NO permeable membrane, comparing with that of other cell-adhesive matrices such as collagen and poly Llysine [34].
Here, we performed an MTT assay to reconfirm the cytotoxicity of fluorinated xerogel-derived NO permeable membrane, prior to real-time detection of NO release in live cells. RAW 264.7 macrophages were incubated on fluorinated xerogel-derived NO permeable membrane and collagen for 24, 48, and 72 h. Similar to preliminary result, the viability of RAW 264.7 cells exposed to xerogel membrane for 24 h was 102.64 ± 3.52%. This result showed that fluorinated xerogel-derived NO permeable membrane was not cytotoxic to RAW 264.7 macrophages. Moreover, the proliferation of RAW 264.7 macrophages was not influenced during 48 and 72 h of growth on the xerogel membrane.

Morphology of RAW macrophages on fluorinated xerogel membrane
Macrophages, which was derived from circulating blood monocytes and reside in most tissues, play an important role in host defense and tissue homeostasis as they ingest dead tissue and fight invading pathogens [38,39]. Therefore, the activation of macrophages is a general characteristic for the early stages of pathogens infection. LPS, a major outer membrane component of Gram-negative bacteria, plays an important role in the pathogenesis of Gramnegative bacterial infection [40] and has been served as an important active component for pathogen-induced macrophage inflammation studies [41]. In general, LPS stimulation caused some morphological changes in human macrophages such as the formation of lamellipodia and the extensive cell spreading [42]. But these LPS induced morphological changes in RAW 264.7 macrophages by conventional microscopy could not provide the ultrastructural changes in RAW 264.7 macrophages by LPS stimulation.
AFM has emerged as an imaging technique that provides 3D topographic information and structural details of biological samples including microorganisms, cell membranes, biopolymers, and more. Therefore, AFM is now a common tool for nanotechnology, not only for imaging but also to measure the interaction forces in biological systems [43]. We performed AFM measurement to observe the morphological changes in RAW macrophages on xerogel membrane following LPS treatment. Figure 2 shows representative AFM 3D topography images taken from LPS-stimulated RAW cells for 0, 0.5, 1, 1.5, 6, and 24 h. As shown in Figure  2(A), control RAW macrophage showed a typical spherical shapes with ruffled membranes and a few lamellipodia surround them. But the morphology of RAW cells after 1 h LPS stimulation exhibited the extension of the cytoplasm toward its outside, as shown in Figure  2(C). After the LPS exposure for 1.5 h, they appeared spindle and long-stretched shapes ( Figure  2(D)). The number of microspikes was increased from 1.5 h after stimulation of LPS. As a result, RAW macrophages for 24 h LPS stimulation showed strong lamellipodia with a lot of dendritic pseudopod. In addition, LPS stimulated RAW macrophages for 24 h showed some holes on their surface. Saxena et al. [44] reported that LPS induced differentiation of RAW 264.7 macrophage cells into dendritic like cells. And Pi et al. [39] reported the LPS activated RAW 264.7 macrophages changed to be much bigger than control cells with some holes emerged on cell surface. From AFM results, we suggested that RAW macrophages on fluorinated xerogelderived NO permeable membrane were normally activated by LPS stimulation.

Amperometric response of xerogel-derived planar-type NO sensor
The sensor performance of the resulting NO microelectrode was investigated in DMEM solution. The calibration and dynamic response curves for NO and interfering species (i.e., nitrite) are shown in Figure 3. A sensor coated with fluorinated xerogel membrane did not exhibit serious interference from nitrite concentrations of up to 20 μM, which is a concentration much greater than the highest level present in physiological samples. Indeed, the fluorinated xerogel-modified sensors exhibited excellent selectivity (logK NO , j amp ) of < -5, <-5, <-5, <-5, <-5, -4.38, and -3.85 for j = nitrite, ascorbic acid, uric acid, acetaminophen, dopamine, hydrogen peroxide, and carbon monoxide, respectively. In addition, the xerogel-derived sensor responded to NO concentrations of up to 1,000 nM with respect to sensitivity (0.15 ± 0.012 nA/ nM), linearity (R 2 = 0.9981, 20~1,000 nM NO range), and response time (t 90% < 10 s).

Real-time detection of NO release in live cells utilizing planar-type NO sensor
NO plays an important role in a variety of cellular functions, including blood vessel relaxation, neurotransmission, and immune response [34]. It is produced through the conversion of Larginine into L-citrulline by diverse cell types such as macrophages, neutrophils, neuronal cells, and endothelial cells as both an extra and intracellular signaling molecule [45]. The mechanisms of NO formation by three kinds of nitric oxide synthase (NOS), including endothelial (eNOS), neuronal (nNOS), and inducible NOS (iNOS), are well established [46]. Among them, the expression of iNOS is induced by stimulation of cells with LPS and/or proinflammatory cytokines [46,47]. Macrophage iNOS is responsible for producing large quantities of NO, which are synthesized over the period of several hours after cells are stimulated [34]. NO plays an important role as part of a nonspecific defense mechanism, as a signaling molecule between macrophages and other cells and as an intracellular messenger within the macrophages themselves [45]. Besides, NO produced by endothelium cells plays a critical role in regulation of vascular tone and activity. The continuous detection of NO release from endothelium could provide new and important insights into the modulatory role of NO in cardiovascular system [48].
However, most methods for NO quantification are indirect including spectroscopic approaches such as the Griess assay which determines the major and stable products of NO degradation-nitrite and nitrate, as well as the detection of methemoglobin after NO reaction with oxyhemoglobin [9][10][11]. Unfortunately, these methods often fail to accurately reflect the dynamics of NO in vivo, in vitro, and in real time [12]. In contrast to other electrochemical NO sensor designs, xerogel-based NO sensors are fabricated using a simple coating procedure, with high sensitivity, selectivity, and reproducibility [8]. Figure 4 represents real-time detection monitoring of NO release by a planar-type xerogelderived NO sensor in RAW 264.7 macrophages stimulated by LPS. In control RAW macrophages, there was no increase in current throughout the whole experimental period. But the current resulting from NO production started to rise in 5,500 s after LPS stimulation. After reaching the maximum current in 6,000 s, a slow decline in NO production lasted for the rest of the experiment. The maximum change in NO concentration was about 60 nM, based on in vitro calibration results.

Conclusions
In this study, we evaluated the biocompatibility of sol-gel derived NO permeable membranes modified with fluorinated functional groups with RAW 264.7 macrophages. Our results showed that xerogel-derived NO permeable membrane was nontoxic and could provide good cell adhesion. In addition, the morphology of these adhering RAW 264.7 macrophages did not show any differences compared to those grown on other biocompatible membranes. Therefore, we expected that fluorinated xerogel-derived membrane could be utilized as both NO gas permeable and cell-adhesive membranes on which cells are directly grown for a cellular NO sensor. Besides, in the planar-type sensor with NO permeable membrane, there is no need to adjust the distance between sensor and cells. Overall, in contrast to other electrochemical NO sensor designs, xerogel-based NO sensors can be easily applied to the cellular sensing system, with a simple coating procedure.
In summary, a planar-type NO sensor with fluorinated xerogel membrane has good sensitivity and specificity for detecting biologically released NO in live cells. And we successfully measured the NO release in RAW 264.7 macrophages stimulated by LPS, utilizing our planartype xerogel-derived NO sensor. Therefore, we expect that our sensor can be applied for realtime monitoring of NO dynamics in various cells, tissues, and organs.