Seamless Signal Transduction from Live Cells to an NO Sensor via a

Feb 1, 2008 - Ibaraki, 305-8506, Japan. A smart live-cell assay was developed as a cellular bio- sensing system. This system is based on novel tactics...
0 downloads 0 Views 226KB Size
Download PDF
Anal. Chem. 2008, 80, 1505-1511

Seamless Signal Transduction from Live Cells to an NO Sensor via a Cell-Adhesive Sensing Matrix Hitoshi Asakawa,† Katsumi Mochitate,‡ and Tetsuya Haruyama*,†

Department of Biological Functions and Engineering, Kyushu Institute of Technology, Kitakyushu Science and Research Park, Fukuoka 808-0196, Japan, and National Institute for Environmental Studies, 16-2, Onogawa, Tsukuba-Shi, Ibaraki, 305-8506, Japan

A smart live-cell assay was developed as a cellular biosensing system. This system is based on novel tactics: the direct assembly of human cultured cells onto a celladhesive sensing matrix. This novel design provides considerable advantages, among them the possibility of capturing molecular signals immediately after they are secreted from living cells. The design also helps preserve all cellular characteristics intact. In this study, a celladhesive NO sensing matrix, acting as both an NOpermeable membrane and a cell-adhesive scaffold, was designed using functional polymers and a short peptide sequence derived from extracellular matrix (ECM) proteins. Using the cell-adhesive NO sensing matrix, we constructed a cellular biosensing system based on in situ monitoring of NO released from a human umbilical vein endothelial cell (HUVEC) layer. HUVECs were employed as an organ-functional model of a blood vessel in view of screening vasodilatory substances for clinical purposes. In our novel system, the electrochemical NO sensor is adjacent to the NO-producing cells, which allows the sensing device to achieve superior sensitivity and precise response to a very low number of NO molecules. Our design enables the fixing of the exact distance between the organ-functional model and the chemical sensor without cumbersome manipulations. Consequently, this cellular biosensing system may be readily applicable to high-throughput analysis in the field of drug screening. High-throughput analysis (HTA) is an essential assay process in various fields, especially clinical drug screening. In the pharmaceutical industry, the rapid and systematic evaluation of the biological effects (e.g., drug potency and toxicity) is imperative when screening candidate drugs (lead substances) from a massive chemical library.1,2 Other fields, such as chemical safety, also need a sophisticated HTA method. In clinical drug research and development, the development of a smart HTA tool is desired to further advance the analytical technologies used for the discovery * To whom correspondence should be addressed. E-mail: haruyama@ life.kyutech.ac.jp. Phone and Fax: +81-(0)93-695-6065. † Kyushu Institute of Technology. ‡ National Institute for Environmental Studies. (1) Drews, J. Science 2000, 287, 1960-1964. (2) Hertzbeng, R. P.; Pope, A. J. Curr. Opin. Chem. Biol. 2000, 4, 445-451. 10.1021/ac702001u CCC: $40.75 Published on Web 02/01/2008

© 2008 American Chemical Society

of next-generation clinical drugs and maintenance of chemical safety. The authors developed a high-throughput analytical method based on a novel analytical approach of qualified analysis. The conventional analytical concepts of qualitative and quantitative analysis always target specific molecules. In our application of qualified analysis, the efficacy, influence, or commonality of external stimulation of a given biological system, e.g., the human organism, is the target. Currently, the evaluation of biological effects is based on the monitoring of specific biological response (mostly molecular production or secretion) of organ-functional models. The concept of qualified analysis would serve to refine existing assessment technologies according to actual social demands. Based on the qualified analysis, the authors have developed a conjugative cellular biosensing technology system comprised of an organ-functional model with a chemical sensor. We have previously reported cellular biosensing as a practical technology for surveying drug efficacy profiles and chemical safety.3-7 An organ-functional model consisting of cultured cells is constructed in order to judge the effects of external stimuli on specific objects (e.g., a human organ). Smart HTA is a viable alternative to animal experimentation. Organ-functional models transmit a specific signal in response to various external stimuli. Such cellular response is a part of the native activity of a living cell and is based on the function of a specific biological cascade. The transmitted responses can take many forms, including gene expressions, morphological change, and specific types molecular secretion. Recent advances in cell biology have allowed researchers to obtain detailed information about the link between external stimuli and cellular response signals. Most of these cellular responses can be monitored if an adequate sensor system is coupled with a livecell/tissue culture component prepared as an organ-functional model. In the present study, the authors focused on nitric oxide (NO) as a cellular response signal in cellular biosensing. It is well-known (3) Haruyama, T. Adv. Drug Delivery Rev. 2003, 55, 393-401. (4) Haruyama, T. Anal. Chim. Acta 2006, 568, 211-216. (5) Haruyama, T.; Kobatake, E.; Aizawa, M. Biosens. Bioelectron. 2002, 17, 209215. (6) Haruyama, T.; Bongsebandhu-Phubhakdi, S.; Nakamura, I.; Mottershead, D.; Keinanen, K.; Kobatake, E.; Aizawa, M. Anal. Chem. 2003, 75, 918921. (7) Kamei, K.; Mie, M.; Yanagida, Y.; Aizawa, M.; Kobatake, E. Sens. Actuators, B 2004, 99, 106-112.

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008 1505

Figure 1. Schematic diagram of the constructed cellular biosensing device. (A) Experimental setup of the electrochemical measurement. (B) Cross section of the cellular biosensing device. In this device, a human umbilical vein endothelial cell (HUVEC) layer is assembled directly on a cell-adhesive NO sensing matrix. The cell-adhesive NO sensing matrix consists of a polyion complex and MAST-peptide.

that NO is a key molecular signal in a variety of cellular functions, including blood vessel relaxation,8 neurotransmission,9 and immune response.10 One of the functions of the endothelial cells of blood vessels is to secrete NO molecules in response to external stimulation by biological vasodilatory substances (e.g., acetylcholine and bradykinin).11 The secreted NO induces blood vessel relaxation as it is diffused into smooth muscle cell, thereby affecting blood pressure. This fact about the biological function of NO molecules suggests that we can evaluate the vasodilatory actions of external stimuli based on the accurate monitoring of NO secretion from endothelial cells as an organ-functional model of blood vessels. Several techniques of detecting biological NO molecules have been reported, e.g., chemiluminecence,12 the Griess method,13 and the use of fluorescence probes.14 The advantages of electrochemical NO sensing for the in situ monitoring of biological NO are also well-studied.15,16 Therefore, the authors have developed a novel amperometric NO sensor using several sensing materials.17,18 However, the in situ monitoring of the cellular NO signal is practically difficult, because NO is rapidly metabolized and degraded under cell-culture conditions. Due to the instability of NO, the diffusion distance of NO is very limited. Furthermore, cellular NO molecules are considerably diluted by diffusion into the cell-culture media. Isik et al. have reported that the response of an NO sensor is significantly affected by a distance between the NO sensor and the NO-producing cells.19 These observations suggest that the NO sensor should be adjacent to the NOproducing cells, at an exact distance; however, the traditional (8) Ignarro, L. J.; Buga, G. M.; Wood, K. S.; Byrns, R. E.; Chaudhuri, G. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 9265-9269. (9) Savchenko, A.; Barnes, S.; Kramer, R. H. Nature 1997, 390, 694-698. (10) Fujihara, M.; Connolly, N.; Ito, N.; Suzuki, T. J. Immunol. 1994, 152, 18981906. (11) Cohen, R. A.; Vanhoutte, P. M. Circulation 1995, 92, 3337-3349. (12) Kikuchi, K.; Nagano, T.; Hayakawa, H.; Hirata, Y.; Hirobe, M. Anal. Chem. 1993, 65, 1794-1799. (13) Guevara, I.; Iwanejko, J.; Dembinska-Kiec, A.; Pankiewicz, J.; Wanat, A.; Anna, P.; Golabek, I.; Bartus, S.; Malczewska-Malec, M.; Szczudlik, A. Clin. Chim. Acta 1998, 274, 177-188. (14) Kojima, H.; Nakatsubo, N.; Kikuchi, K.; Kawahara, S.; Kirino, Y.; Nagoshi, H.; Hirata, Y.; Nagano, T. Anal. Chem. 1998, 70, 2446-2453. (15) Taha, Z. H. Talanta 2003, 61, 3-10. (16) Bedioui, F.; Villeneuve, N. Electroanalysis 2003, 15, 5-18. (17) Haruyama, T.; Shiino, S.; Yanagida, Y.; Kobatake, E.; Aizawa, M. Biosens. Bioelectron. 1998, 13, 763-769. (18) Asakawa, H.; Ikeno, S.; Haruyama, T. Sens. Actuators, B 2005, 108, 646650. (19) Isik, S.; Etienne, M.; Oni, J.; Blochl, A.; Reiter, S.; Schuhmann, W. Anal. Chem. 2004, 76, 6389-6394.

1506

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

procedure of physically manipulating the NO sensor electrode to achieve the optimal distance to the cell is unsuitable in the HTA process. To overcome the limitation of the traditional method, the authors invented an original tactic, i.e., the direct assembly of an organ-functional model on the surface of an NO sensing matrix. This means the organ-functional model is constructed on the NO sensor surface using as a scaffold. The resulting cellular biosensing system may provide superior sensitivity and precise response to NO molecules, because the produced NO is immediately diffused and captured by the NO sensing matrix. Furthermore, since this cellular biosensing system does not require the manipulation of the NO sensor in order to minimize the distance from the NO-producing cells, it will be readily applicable for the HTA process. However, it is difficult to construct an organfunctional model on an NO sensing matrix, because many types of cells are anchorage-dependent: they require specific substrate attachment for growth and regulation of cellular functions. In order to realize the construction of an organ-functional model and its direct and simultaneous assembly on the NO sensor surface, we designed a cell-adhesive NO sensing matrix using a short peptide sequence derived from extracellular matrix (ECM) proteins. The designed NO sensing matrix allows cultured cells to adhere and grow on the surface of the NO sensor, thereby enabling the direct assembly of organ-function model on the sensor. In order to appraise the potentiality of our tactics, we built a cellular biosensing device using directly assembled human umbilical vein endothelial cells (HUVECs) as the organ-functional model of a blood vessel in view of screening vasodilatory chemicals for clinical drugs. (Figure 1). Using the constructed cellular biosensing system, we demonstrated the seamless signal transduction from live HUVECs to the NO sensor via the designed cell-adhesive NO sensing matrix. EXPERIMENTAL SECTION Chemicals and Reagents. Poly-L-lysine hydrobromide (average molecular weight, 84 000) and poly(sodium 4-styrenesulfonate) (average molecular weight, 70 000) were purchased from Sigma Aldrich (St. Louis, MO). Sodium nitrite and acetylcholine chloride were purchased form Wako Chemicals Co. (Tokyo, Japan). Calcein-AM solution and NG-nitro-L-arginine methyl ester (LNAME) were purchased from Dojin Chemical Co. (Kumamoto, Japan). Hoechst 33342 was obtained from Invitrogen Corp. (California). Poly(maleic anhydride-co-styrene) (MAST) and 1-ethyl3-(3-dimethylaminopropyl)-carbodiimide hydrochloride, a water-

Figure 2. Molecular structures of the two polymers, poly-L-lysine (PLL) and polystyrene sulfonate (PSS), and that of the cell-adhesion element, cell-binding peptide-derivatized poly(maleic acid-co-styrene) (MAST-peptide). R indicates the cell-binding peptide.

soluble condensation agent (WSC), were purchased from Celagix (Yokohama, Japan) and Wako Chemicals Co. (Tokyo, Japan). Peptides at the guaranteed purity of more than 95% were obtained from American Peptide (Sunnyvale, CA). The other chemicals were guaranteed grade. Preparation of Cell-Adhesion Elements. MAST powder (200 mg) was first suspended in water and completely hydrolyzed with 1 N NaOH. The hydrolyzed MAST (MAST-OH) was dialyzed against water and lyophilized. The MAST-OH powder (200 mg) was mixed and activated by WSC (50 mg) in the buffer at pH 4.7 under stirring for 2 h at room temperature. Cell-binding peptide (20 mg) was added to activated the MAST-OH solution, and the mixture was stirred for another 2 h. After synthesis, the reaction mixture was thoroughly dialyzed against water to remove free peptide, etc., and lyophilized to obtain a powder of peptide-tagged MAST-OH polymer. This MAST-peptide, i.e., MAST-OH tagged with the cell-binding peptide, was employed as the cell-adhesion element. The basic structure of cell-adhesion elements is illustrated in Figure 2. We prepared four different cell-adhesion elements using three different peptides: MAST-FIB1, MASTAG73, and MAST-A4G78, and MAST-OH as the negative control. Cell culturing, Counting, and Staining. Human umbilical vein endothelial cells (KURABO, Japan) were cultured in human endothelial-SFM (GIBCO, Invitrogen) supplemented with human recombinant basic fibroblast growth factor (hrbFGF) (20 ng/mL), human recombinant epidermal growth factor (hrEGF) (10 ng/ mL), and fibronectin (10 µg/mL) at 37 °C in a humidified incubator (95% air with 5% CO2). The living cell number was determined by CellTiter-Glo luminescent cell viability assay (Promega, Madison, WI), which is based on luminescence analysis of the cellular ATP level. HUVECs were plated at a density of 3 × 104 cells/cm2 and allowed to adhere for 24 h. Luminescence intensity was measured using a Wallac 1420 ARVOsx multilabel counter (Perkin-Elmer). Since the NO sensing matrix-coated plate was opaque, the adherent cells were observed by staining with the fluorescent dyes calcein-AM and Hoechst 33342. The cells were stained with 1 µM calcein-AM and 1 µM Hoechst 33342 for 15 min in Hank’s balanced salt solution (HBSS) at 37 °C. After

incubation, the stained cells were assayed using a fluorescence microscope (TE2000, Nikon, Japan). Construction of Cellular Biosensing Device. A gold electrode (2 cm × 2 cm) was constructed by vapor deposition of Cr (thickness ) 50 Å) and Au (thickness ) 3000 Å) on the surface of a Corning glass piece (thickness ) 0.5 mm). A glass ring (i.d. ) 15.6 mm, o.d. ) 18 mm, height ) 15 mm) was attached on the gold electrode surface with a silicon adhesive (KE42T, Shin-Etsu Kagaku Co., Japan). The electrode-bottomed ring was employed as both the electrochemical bath and cell-culture chamber. The experimental setup of the electrochemical measurement is illustrated in Figure 1A. The gold electrode (bottom of the chamber) is employed as the sensor electrode (working electrode, WE). Coiled Pt wire (counter electrode, CE) is immersed in the chamber from above. A Ag/AgCl electrode (reference electrode, RE) was immersed into 0.1 M KCl solution, which was connected to the chamber through a KCl-agar salt bridge. NO sensing matrix is prepared as follows. A 600 µL volume of 25 mM polyL-lysine hydrobromide aqueous solution (in monomer units) and a 300 µL volume of 25 mM polystyrene sulfonate (in monomer units) were mixed and poured into the cellular biosensing chamber and allowed to dry for 24 h at room temperature. The polyion layer acts as an NO sensing matrix with NO selectivity. The MAST-peptide solution drops were poured into the cellular biosensing chamber and incubated for 12 h to allow the absorption of the MAST-peptide on the surface of the NO sensing matrix. The unadsorbed MAST-peptide was removed by washing twice with PBS and culture medium. Electrochemical Measurement of NO. Double potential step chronoamperometry (DPSCA), which is an electrochemical measurement method, was employed for the monitoring of NO in the constructed cellular biosensing system. The double-pulse step potential was applied to the sensor device after maintaining the potential at 0 mV for 30 s. The first and the second potential of DPSCA were 700 mV (500 ms) and 750 mV (250 ms) versus Ag/ AgCl, respectively. The hold time length was enough for decaying the charging current. The current response (∆I) was defined as the difference between the first potential and second potential, because NO molecules are oxidized selectively at this potential. The step potentials were applied at the intervals of 30 s in the in situ experiment of cellular NO monitoring. A standard NO solution was prepared by dissolving pure NO in HBSS. Pure standard NO gas (99.7%) was purchased by Sumitomo Seika Chemicals (Osaka, Japan). Pure NO gas and Ar gas (99.9995%, Taiyo Toyo Sanso, Osaka, Japan) were mixed using a computerized gas mass flow blender (STEC Inc., Kyoto, Japan) to obtain the desired mixed ratio, and the gas mixture was bubbled in 5 mL of degassed HBSS for 30 min. The standard NO solution had a concentration of 95 µM (5% NO + 95% Ar/total dissolved gas in solution). In Situ Monitoring of Biological NO. To assemble the organfunctional model of blood vessels on the cellular biosensing device, HUVECs were cultivated in the cell-culture chamber of the sensing device for 24 h. The in situ cellular NO monitoring experiments were performed based on DPSCA in HBSS at 37 °C. After a stable DPSCA background response was obtained, acetylcholine stimulating solution, a drug which models a vasodilator, was injected into the chamber of the sensing device. L-NAME, an inhibitor of nitric oxide synthetase (NOS), was employed to clarify whether Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1507

or not the sensor response was caused by biological NO from the organ-functional model. RESULTS AND DISCUSSION To actualize our tactics of direct assembly of human cultured cells on an NO sensor surface using interfacial matrix design, we constructed the cellular biosensing device shown in Figure 1. In this study, we used HUVECs in the cellular biosensing as an organ-functional model of a blood vessel for the HTA of vasodilatators. The chamber formed by attaching the glass ring to the gold electrode served as both the electrochemical bath and the cell-culture well. The cell-adhesive NO sensing matrix layer uniformly covered the bottom of the chamber, which was goldplated to serve as the working electrode. The authors have previously reported that the polyion complex layer consisting of poly-L-lysine (PLL) and polystyrene sulfonate (PSS) (Figure 2) acts as an NO-permeable membrane for NO selectivity.20 Although the polyion complex layer can be employed as an NO sensing matrix for performing electrochemical NO measurement in aqueous condition, it is difficult to achieve the long-term cultivation of HUVECs on the polyion complex surface. Therefore, in this study, we developed a cell-adhesive NO sensing matrix to act as both the NO-permeable membrane and the cell-adhesive scaffold simultaneously. Molecular Design of the Cell-Adhesion Element. Previously, various kinds of methods have been employed for the design of cell-adhesive materials in the field of biomaterial research. Surface modification with cell-binding peptide is a sophisticated way of endowing material surfaces with cell-adhesive activity.21 Based on these techniques, we tried to induce cellular adhesiveness in the NO sensing matrix surface designing and synthesizing cell-adhesion elements that consist of poly(styreneco-maleic acid) (MAST-OH) and cell-binding peptide, as illustrated in Figure 2. The cell-binding peptides were conjugated with MAST-OH via an amide bond between the N-terminal amino groups of the peptides and the carboxyl groups of MAST-OH. We employed the amphiphilic alternating copolymer MAST-OH for the immobilization of cell-binding peptide. The monolayer formation of MAST-OH at the interface due to its amphiphilic property has been reported.22 MAST-OH and MAST-peptide polymers bind strongly to the PLL-PSS complex through their phenyl group to form a monolayer film at the interface. The peptide moiety of the MAST-peptide that mimics the cell-binding sequence of ECM proteins can protrude to the outer surface of the cellular membrane, anchoring itself in the interface. Four different cell-adhesion elements, including one negative control, are listed in Table 1: the three actual cell-adhesion elements, MAST-FIB-1, MAST-AG73, and MAST-A4G78, and the negative control, MAST-OH. The FIB-1 peptide contains an Arg-GlyAsp (RGD) sequence, which has been identified as a cell-binding domain of fibronectin via integrin recaptors. The AG73 and A4G78 peptides of the G4-domein of laminin-R1 and -R4 chains have several biological activities involved in cell adhesion and proliferation.23 The cell-adhesive activity of each cell-adhesion element in (20) Kamei, K.; Haruyama, T.; Mie, M.; Yanagida, Y.; Kobatake, E.; Aizawa, M. Biotechnol. Lett. 2003, 25, 321-325. (21) Hersel, U.; Dahmen, C.; Kessler, H. Biomaterials 2003, 24, 4385-4415. (22) Garnier, G.; Duskova-Smrckova, M; Vyhnalkova, R. Langmuir 2000, 16, 3757-3763.

1508

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

Figure 3. Assay of HUVEC adhesion on the cell-adhesion elements coated polystyrene dishes. The number of adherent cells was determined at 24 (white bar) and 58 h (black bar) after plating. The initial plating cell density (3 × 104 cells/cm2) was assigned a relative value of 100. Table 1. Prepared Cell-Adhesion Elements

HUVECs was evaluated (Figure 3). HUVECs were seeded on polystyrene dishes coated with 10 µg/mL MAST-FIB-1, MASTAG73, MAST-A4G78, or MAST-OH solution. The initial cell density was 3 × 104 cells/cm2. Figure 3 shows the relative number of adherent cells on the surface-coated culture dishes. At 24 h after the cell seeding, there was apparently no significant difference in the number of adherent cells among these dishes coated with four different cell-adhesion elements. However, HUVECs could proliferate on the MAST-FIB-1-coated surface, while the adherent cells on the MAST-AG73, -A4G78, and -OH coats decreased gradually during the long-term cultivation. The morphology of the adherent cells was observed simultaneously with a phase-contrast microscope. The HUVECs spread and proliferated more widely on the MAST-FIB-1-coated dishes than on other coats (data not shown). These results suggest that HUVECs are able to proliferate on the MAST-FIB-1-coated surface by anchoring themselves to the FIB-1 peptides via the integrin receptors. After the adhesion experiment, MAST-FIB-1 was employed as the best cell-adhesion element for the subsequent construction of cell-adhesive NO sensing matrixes. Construction of the Cell-Adhesive NO Sensing Matrix. In order to construct a cell-adhesive NO sensing matrix, the coating of the NO sensing matrix with the cell-adhesion element composed of the polyion PLL-PSS complex was performed. The MAST-FIB-1 cell-adhesion element was adsorbed on the NO (23) Okazaki, I.; Suzuki, N.; Nishi, N.; Utani, A.; Matsuura, H.; Shinkai, H.; Yamashita, H.; Kitagawa, Y.; Nomizu, M. J. Biol. Chem. 2002, 277, 3707037078.

Figure 4. Assembly of an HUVEC layer on the NO sensing matrix via a polyion complex (a) without a cell-adhesion element coat, (b) with a MAST-OH coat (100 µg/mL), and (c) with a MAST-FIB-1 coat (10, 50, 100 µg/mL).

Figure 5. Visualization of the adherent HUVECs stained with calcein-AM (green) and Hoechst 33342 (blue) on the cell-adhesive NO sensing matrix: (A) on the polyion complex with no cell-adhesion element coat, (B) with a MAST-OH coat, and (C) with a MASTFIB-1 coat. Scale bar: 50 µm.

sensing matrix surface via electrostatic and hydrophobic interaction. We investigated the concentration of the MAST-FIB-1 coat for the adherence of HUVECs on the surface of the NO sensing matrix (Figure 4). HUVECs were cultured in wells coated with the polyion complex and then with MAST-FIB-1. A CellTiterGlo luminescent cell viability assay kit (Promega) was used for counting the adherent cells on these substrata. As shown in Figure 4, the HUVECs adhesion is increased with coating of MASTFIB-1. The increment of adherent cells at 100 mg/mL reached about 10 times compared to without coat and with MAST-OH coat. The results suggest that the HUVECs adhered on the NO sensing matrix indirectly, by binding to the FIB-1 ligand of the MAST-FIB-1 element via integrin receptors. Since the culture plates were opaque after being coated with the cell-adhesive NO sensing matrix containing the MAST-FIB-1 element, the adherent cells were observed by fluorescence microscopy after staining with calcein-AM (green-fluorescent livecell stain) and Hoechst 33342(blue-fluorescent cell nuclei stain) (Figure 5). In the Figure 5A, we can see that the apparently adhering HUVECs on the NO sensing matrix lacking the celladhesion element were not viable. Even when coating with MAST-OH (Figure 5B), the HUVECs could not adhere well to the NO sensing matrix. This suggests that the amphiphilic alternating copolymer MAST-OH formed a thin layer on the surface of the NO sensing matrix. In contrast, the HUVECs could spread most widely and maintain the highest viability on the MAST-FIB-1 coat (Figure 5C). The HUVECs reached confluence when seeded at a cell density of 2 × 105 cells/cm2 on the MAST-

FIB-1-coated polyion complex (PLL and PSS) at a concentration of 100 µg/mL. In the field of cellular biosensing, it is crucial not only to maintain viability but also to restore the original functions expressed in vivo, in blood vessel tissue in the case of HUVECs. Electrochemical Characterization of the Cellular Biosensing Device. The cell-adhesive NO sensing matrix-coated gold electrode was connected to a computerized electrochemical analyzer (HZ3000, HOKUTO Denko, Japan) with a three-electrode system for DPSCA. The experimental setup is presented in Figure 1A. The DPSCA method is a well-known electrochemical pulse technique of obtaining selectable current response in heterogeneous solution,24 such as a cell-culture medium. Furthermore, the dynamic time-course behavior of biological NO production can be monitored by applying a double-pulse potential at a regular interval, as shown in Figure 6A. For the characterization of the NO sensor performance, the cellular biosensing device was connected to the electrochemical analyzer without HUVECs. The well formed by the glass ring, which was used as the cell-culture chamber where the HUVECs were assembled on the NO sensing matrix, becomes an electrochemical bath in the NO measurement experiment. The typical DPSCA recorded current for dissolved NO is shown in Figure 6B, and the current response (∆I) was defined as difference between the first potential and the second potential. Figure 6C shows the logarithmic plots of the response current versus the standard NO and interference concentrations in HBSS. Nitrite, ammonium, and ascorbic acid were major interfering agents in the amperometric NO detection under cellculture conditions. In particular, the interference of nitrite affected the NO sensor response due to similar oxidation potentials of nitrite and NO. The current response plots show that the NO sensing matrix could achieve prominent NO sensitivity, 2 figures higher than that against nitrite. The sensor responses were associated with the increment of the dissolved NO concentration in the nanomolar (nM)-to-micromolar (µM) range. These results indicate that the sensor device can determine NO concentrations based on the magnitude of sensor response (∆I). The NO detection limit of the constructed device was 200 nM. The detection limit is defined as a signal/noise ratio ) 5. Conventional sensing devices have required a lower detection limit (typically subnanomolar) to monitor biological NO, because the produced NO is degraded momentarily under biological condition. In contrast, the present seamless signal transduceable sensor device can capture and detect biological NO immediately. As appeared in the results, the present cellular biosensing system can monitor the cellular NO production successfully by the seamless signal transduction system. In order to sophisticate the cellular biosensing system, the performance of the sensing matrix has been improved in the author’s group. Our molecular design and synthetic process is producing some new sensing materials, e.g., an artificial enzyme membrane.25 In Situ Monitoring of Cellular NO Production. Recently, the construction of a sensing device for the in situ monitoring of cellular NO secretion has been reported.7,19 As evidenced by such reports, the efficient capture of NO molecules requires the close proximity of the NO sensor to the NO-producing cells. However, the applicability of a cellular biosensing system to HTA is reduced (24) Lantoine, F.; Trevin, S.; Bedioui, F.; Devynck, J. J. Electroanal. Chem. 1995, 392, 85-89. (25) Ikeno, S.; Asakawa, H.; Haruyama, T. Anal. Chem. 2007, 79, 5540-5546.

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1509

Figure 6. Electrochemical measurement of NO by double potential step chronoamperometry (DPSCA): (A) set point of DPSCA, (B) typical DPSCA spectra of the NO sensor for the dissolved NO, and (C) the logarithmic plots of the response current vs the standard NO and major interference concentrations.

if the procedure requires the mechanical manipulation of the NO sensor for every assay. To overcome these limitations, we developed a cellular biosensing system in which the organfunctional model is assembled directly on the NO sensing matrix, as schematically illustrated in Figure 1. In this study, HUVECs were assembled as an organ-functional model of blood vessels on the surface of a cell-adhesive NO sensing matrix. Analysis of the cell-adhesive NO sensing matrix by scanning electron microscopy (SEM) showed that the thickness of the matrix layer was approximately 3 µm and that it covered the covered the gold electrode surface uniformly. The extreme thinness of the NO sensing matrix layer can minimize the degradation of NO caused by diffusion to the surface of the NO sensor electrode. These results strongly suggest that the cellular biosensing device can detect NO efficiently because the produced NO molecules diffuse directly into the NO sensing matrix. With the use of the constructed cellular biosensing system, this study demonstrated the in situ monitoring of cellular NO secretion induced by acetylcholine (Ach) stimulation. We selected Ach as a drug which models vasodilators in this study, because the Ach signaling pathway to NO generation has been well cleared.11 The Ach signal is transmitted to the intercellular signaling cascades via muscarinic acetylcholine receptor (mAchR). In particular, the prevalent expression of M3 subtype mAchR, among the five subtypes of mAchR (M1-M5), in endothelial cells has been reported.26 M3-subtype mAchR, which is a G-proteincoupled receptor, mediates the rapid increment of intercellular Ca2+ concentration. Endothelial NO synthase (eNOS) is then activated by the binding of the Ca2+-calmodulin complex, and NO is subsequently generated. The dynamic time-course behaviors of NO release from HUVEC were measured using the constructed system. Figure 7 gives the typical DPSCA sensor profiles. The stimulating solution was applied to the chamber at the time indicated by the downward (26) Walch, L.; Brink, C.; Norel, X. Therapie 2001, 56, 223-226.

1510 Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

arrow in Figure 7. As a control experiment, HUVECs were treated with 1 mM L-NAME for 1 h before the application of the stimulating solution. L-NAME is a NOS-specific inhibitor. In the control experiment, no increment of the sensor output was observed (blue plots). In contrast, the NO sensor output increased 5 min after Ach application and onward in a concentrationdependent manner (red and green plots). These results clearly indicate that the profiling of the sensor output represents the dynamic time-course behavior of NO release by the HUVECs, because L-NAME selectively inhibited NO generation by NOS. Stimulation with 1 mM Ach caused a 5 µA increment of sensor output (red plots). We estimated that the NO concentration reached 400 nM as a result of the Ach application, based on the data of the NO sensor performance (Figure 6C). This quantitative result suggests that NO concentration was very high in the vicinity of the cell surface. This study demonstrated an example of qualified analysis. It can evaluate the efficacy of stimulus for vasodilatation on the organ-functional model of the human blood vessel. With the use of the system, candidate drugs (lead substances for drug discovery for cardiovascular diseases) will be recognized from a large

Figure 7. In situ DPSCA monitoring of NO released from the HUVEC layer. Acetylcholine stimulating solution was injected into the sensing device at the time indicated by downward arrow.

number of substances library. The qualified analysis-based sensor can be employed as a potential tool for clinical drug screening, safety surveillance, and other purposes. CONCLUSION In order to evaluate the effect of external stimuli (chemical, biological, or physical) on biological cascades in an HTA manner, we developed a cellular biosensing system based on the concept of qualified analysis. Analytical processes would be more sophisticated if they were based on cellular biosensing. Such analytical tools will be undoubtedly spread to various fields, including drug screening. This is the first report of a cell-adhesive NO sensing matrix which allows seamless signal transduction from live HUVECs to an NO sensor. Using a cell-adhesive NO sensing matrix, we developed a cellular biosensing device in which an HUVEC layer is assembled as an organ-functional model of blood vessels directly on the NO sensor. We demonstrated that the cellular biosensing device could perform the in situ monitoring of NO released from the HUVEC layer externally stimulated by a drug which models

vasodilators. These results demonstrate the potential of the present cellular biosensing system as a powerful analytical tool for screening vasodilatory substances. In conventional cell-based sensor systems, the spatial relationship between the cell/tissues and the sensor is always erratic, and accuracy and reproducibility cannot be guaranteed. Our tactic provides a clear solution to these serious problems of current cellular biosensing. Our cell-adhesive sensing matrix may improve the performance of live-cell assays for HTA. ACKNOWLEDGMENT The present research was financially supported in part by an R&D Project for Environmental Nanotechnology Japan. H.A. acknowledges support of a JSPS Research Fellowship (Project No. 18-7790).

Received for review September 24, 2007. Accepted December 1, 2007. AC702001U

Analytical Chemistry, Vol. 80, No. 5, March 1, 2008

1511