Highly sensitive and selective odorant sensor

Highly sensitive and selective odorant sensor using living cells expressing insect olfactory receptors Nobuo Misawaa,b,1, Hidefumi Mitsunob,c,1, Ryohei Kanzakic, and Shoji Takeuchia,b,2 a Institute of Industrial Science, University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan; bLife Bio Electro-mechanical Autonomous Nano Systems (Life BEANS) Center, The BEANS Project, University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan; and cResearch Center for Advanced Science and Technology, University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8904, Japan

Edited by Katepalli R. Sreenivasan, New York University, New York, NY, and approved July 20, 2010 (received for review March 31, 2010)

This paper describes a highly sensitive and selective chemical sensor using living cells (Xenopus laevis oocytes) within a portable fluidic device. We constructed an odorant sensor whose sensitivity is a few parts per billion in solution and can simultaneously distinguish different types of chemicals that have only a slight difference in double bond isomerism or functional group such as ─OH, ─CHO and ─Cð═OÞ─. We developed a semiautomatic method to install cells to the fluidic device and achieved stable and reproducible odorant sensing. In addition, we found that the sensor worked for multiple-target chemicals and can be integrated with a robotic system without any noise reduction systems. Our developed sensor is compact and easy to replace in the system. We believe that the sensor can potentially be incorporated into a portable system for monitoring environmental and physical conditions. fluidic channel ∣ odorant receptor ∣ robot ∣ Xenopus oocyte

V

arious approaches to developing artificial odorant detectors have been reported. Presently, chemical sensors are mainly fabricated based on metal-oxide semiconductors, quartz crystal microbalances, surface plasmon resonators, or surface acoustic wave devices (1–5). These sensors were developed with the aim of achieving good portability while demonstrating high sensitivity and selectivity, in order that they find application as mobile chemical sensors in the monitoring of environmental or physical conditions (6). However, it has been difficult to develop an adequate chemical sensor that combines all of these properties. Natural living systems, on the other hand, show reactions unique to several chemicals at the molecular level. Such systems are able to distinguish between different chemical moieties found in odorants and tastants (7, 8). The specificity of an organism’s response to certain chemicals comes mainly from membrane proteins associated with the organism. For example, when a certain ion channel activating a single ligand molecule opens for 1 s under a −100 mV-membrane potential, the number of monovalent ions that are transferred through the ion channel reaches the 106 –107 level and generates a few picoamperes (9). This reaction means that chemical signals are converted to amplified current signals in living systems; the system can be regarded as a transistor with an excellent amplifier. Thus, living cells represent attractive tools for realizing highly selective and sensitive chemical sensors (10, 11). To exploit the properties of living cells for application in chemical sensors, electrical measuring methods of cells including the patch clamp technique have been suggested (12–14). However, these methods are unsuited for a portable device because laborious cell handling using micromanipulators under a microscope and noise reduction processes are required. Here, to overcome these problems, we have combined the use of Xenopus laevis oocytes and fluidic devices to develop a highly sensitive and selective odorant sensor using oocytes capable of expressing insect olfactory receptors. Xenopus oocytes are not only useful for the expression of many membrane proteins, but they are also easy to handle owing to their cell size of about 1 mm in diameter (15–18). However, these oocytes have not 15340–15344 ∣ PNAS ∣ August 31, 2010 ∣ vol. 107 ∣ no. 35

previously been considered for application as sensors. A fluidic device is considered a suitable cartridge-type platform for use as an automatic cell array (19), and it can be integrated with electrical measuring systems (20). Furthermore, closed fluidic channels are expected to reduce noise due to low fluctuation of the channel flow. For construction of a reproducible system, we investigated the flow rate and layout of glass electrodes in the fluidic channel. We also demonstrated that the resultant fluidic device could distinguish chemicals due to the high specificity of the receptor used, and it could be connected with a servo motor. Fig. 1A shows the concept image of the odorant sensor based on a fluidic device employing oocytes to express insect olfactory receptors. The setup is composed of the sensor device, an amplifier, a computer, and actuators or robots. In our proposed system, the fluidic channels have oocyte-trap regions (Fig. 1 B and C) and electrodes for measurement of the oocytes’ response to each chemical by the two-electrode voltage clamping (TEVC) method (Fig. 1D) (21). We used pheromone and odorant receptors of moths or flies as model chemical receptors; the expression procedures of these receptors are widely studied and sufficiently established for use in this work. Unlike mammalian odorant receptors that belong to the G-protein coupled receptor family, insect olfactory receptors work as an ionotropic receptor (22). This type of receptor combines both the functions of ligand recognition (receptor) and ion flux regulation (ion channel); therefore, the recognition of odorants by the insect receptors is closely followed by a conversion of the stimuli to current flow. Although the mechanisms for the recognition of odorants are unclear, the oocytes expressing the insect olfactory receptor “sniff out” the odorants with high sensitivity. Our used receptors include the BmOR1, BmOR3, PxOR1, and DOr85b receptors, which come from the silkmoth (Bombyx mori), the diamondback moth (Plutella xylostella), and the fruit fly (Drosophila melanogaster) (23–26). Especially, the availability of DOr85b implies that the system can be applied to detect odorants as well as pheromones. These receptors are known to be involved in the regulation of ion flow through plasma membranes by selective recognition of the corresponding pheromones or odorants, bombykol, bombykal, Z11-16:Ald, and 2-heptanone, respectively (Fig. 2). Results and Discussion Optimal Conditions for Oocyte Installation. We designed a fluidic channel that could trap an oocyte with two glass capillary electrodes along a streamline in the channel. Here, we evaluated the optimum conformation of the fluidic device for Xenopus oocyte Author contributions: N.M., H.M., R.K., and S.T. designed research; N.M. and H.M. performed research; N.M. and H.M. contributed new reagents/analytic tools; N.M., H.M., R.K., and S.T. analyzed data; and N.M., H.M., R.K., and S.T. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1

N.M. and H.M. contributed equally to this work.

2

To whom correspondence should be addressed. E-mail: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1004334107/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1004334107

Fig. 2. Chemical structures of the receptors used in the study. BmOR1 and BmOR3 are the sex pheromone receptors of the silkmoth (Bombyx mori), and these receptors respond to (E,Z)-10,12-hexadecadien-1-ol (bombykol) and (E,Z)-10,12-hexadecadienal (bombykal), respectively. PxOR1 is a sex pheromone receptor of the diamondback moth, (Plutella xylostella), which responds to (Z)-11-hexadecenal (Z11-16:Ald). DOr85b is a receptor of the fruit fly (Drosophila melanogaster), and it responds to 2-heptanone as an odorant.

ENGINEERING

oocytes to detach from the capillaries. It was also found that a flow rate of more than 2 mL∕ min facilitated the glass electrode insertion into the oocytes. However, it was found unfavorable Fig. 1. (A) A schematic illustration of the proposed multichannel chemical sensor using Xenopus oocytes cells and a fluidic system that can be connected with a mechanical system. (B) A schematic view of a Xenopus oocyte trapped and inserted by two glass capillary electrodes in the fluidic channel. The two Ag/AgCl electrodes connect inside the oocyte through 3-M KCl solution in glass tubes. (C) A picture of an oocyte connected with two glass electrodes. (D) A principle image of a “two-electrode voltage clamp method (TEVC)” to detect changes in membrane potential. Command voltage for clamping the membrane potential is shown as V m . The ion channel is shown as a cation channel in this image.

trapping and electrophysiological measurements. The fluidic device was made of acrylic material. Two glass capillary electrodes were symmetrically placed in the device along the channel (Fig. 1 B and C). In the case of multichannel detection, they were installed in series as described in Materials and Methods. First, we investigated the appropriate configuration of the two glass capillary electrodes because the geometrical parameters of the setup including the glass capillary electrodes will influence the current recordings; intensities of cell responses vary according to the position of the current electrode in the cell (27). Fig. 3A shows that a correlation exists between the “distance between the tips of the two glass capillaries” and the “relative ratio of signal current.” Each signal current recording was normalized to the distance of 570 μm. We found that the optimal values for the signal current recordings were in the range between 500–600 μm for efficient measurements of the signal current. Second, we examined the optimum flow rate and the angle between the two glass capillaries for stable oocyte trapping. Fig. 3B shows that the success rate of glass electrode insertion into the oocytes depends on the flow rate (less than 1, 1–1.9, and 2–2.3 mL∕ min) and the angle (15°, 30°, 45°, and 60°) between the two glass capillaries as seen in Fig. 3B Inset. The result means that the lower the angle is between the two glass capillaries, the easier it is to insert glass capillaries into the oocytes and for the Misawa et al.

Fig. 3. (A) Correlation between the distance between the tips of the two glass capillary electrodes (Inset) and the relative ratio of signal current. Each signal current is normalized by comparing with the maximum current as 1.0 in the case of a distance of 570 μm. Error bar: SD. Each data point is denoted as n. The rate in the case of using a θ tube is shown as a reference. (B) Bar graph indicating the success rate of glass electrode insertion into oocytes depends on the flow rate (less than 1, 1–1.9, and 2–2.3 mL∕ min) and the angle (15°, 30°, 45°, and 60°) between two glass tubes as seen in inset. Each data point is more than 10 measurements. Line graph indicating the minimum backflow rate when oocytes escape from the gap between the tips of the two glass capillaries. Error bar: SD. Each data point is more than 8 measurements. PNAS ∣ August 31, 2010 ∣

vol. 107 ∣

no. 35 ∣

15341

that oocytes detached from the glass capillaries due to the backflow that was generated occasionally from the pulsing motion caused by pumping. And, the glass capillaries penetrated through the oocyte when the flow rate was too high (> about 2.3 mL∕ min). As a result, the optimal values for the angle between the two glass capillaries and the flow rate for stable trapping of oocytes were determined to be 30° and ∼2 mL∕ min, respectively. Under these conditions, it is evident that in our proposed system, conventional electrophysiological setups (e.g., micromanipulators and a microscope) are unnecessary. The semiautomatic measurements are thus accomplished due to the high reproducibility of the oocyte installation system. Quantitative Sensing. Fig. 4A shows the differences in signal current value obtained at low concentrations of bombykal. The signal current values increased as the concentration of bombykal in the solution increased when an oocyte expressing BmOR3 was used. Our proposed device achieved sensitivity of 10 nM that means about 2 ppb in solution (t test, n ¼ 6, P < 0.001). Furthermore, we checked the quantitative performance of the sensor using four different chemicals, bombykol, bombykal, Z1116:Ald, and 2-heptanone for four different chemical receptors, BmOR1, BmOR3, PxOR1, and DOr85b, respectively. Fig. 4B shows the measurement results of the dose-dependent response for these four receptors. The results indicate that the sensor can quantitatively detect several chemicals with high reproducibility

in the dynamic ranges of about 10 nM–1 μM. We could record stable signal currents in the case of using single oocytes (Fig. 4A). However, there was variation depending on the individual differences of the oocytes and the kinds of odorant receptors. We think that the differences of signal currents are mainly due to each receptor’s expression level (28), which is supported by our Western blotting result (Fig. S1). In our experiment, the EC50 values of BmOR1, BmOR3, PxOR1, and DOr85b were 0.25, 0.38, 2.52, and 45.6 μM, respectively; these values are basically reasonable when compared to the reported data (24–26). For practical use, reproducible measurements will be achieved by construction of stable expression systems and generating a standard curve. Currently, more than 50 insect odorant receptors have been analyzed using Xenopus oocytes (29), and engineered olfactory receptors for chemical sensing systems are widely studied (10). These receptors can be essentially applied to our proposed device to detect various chemicals. Hence, we believe that the utilization of the Xenopus oocyte has high versatility for the development of chemical sensors for various odorants. Multichannel Sensing with High Selectivity. The proposed fluidic system is considered suitable for making an array of cells combined with an electrical measuring system. We examined the selectivity of the sensor for target chemicals by simultaneously detecting bombykol and bombykal that are ligands for BmOR1 and BmOR3, respectively (23, 24) (Fig. 2). We initially added 10 μL of a 100-nM bombykal solution and then 10 μL of a 3μM bombykol solution to the sensor device that had two oocytes expressing BmOR1 and BmOR3, respectively (Fig. 5). The result shows that the sensor can recognize bombykol and bombykal in the same fluidic channel, indicating that our method allows the detection of chemicals with only a slight difference in their respective chemical structures, such as double bond isomerism and different types of functional groups such as ─OH, ─CHO and ─Cð═OÞ─. The responses that were checked by using the ligand cocktails also indicated the high specificity (Fig. S2). The discrimination has been difficult for conventional portable odorant sensors. Moreover, these differences are conventionally detected using techniques that require large instrumentation such as gas chromatography mass spectrometry (30). Since the proposed fluidic device is compact (5 mm × 30 mm × 45 mm), we think that the device has the potential to be used as a portable cartridge-type multichannel chemical sensor for application in determining either changes in the chemical composition of chemicals in the environment or as a health monitoring device. Integration of the Sensor with a Robot. Sensor systems for measuring cell responses (e.g., electrophysiological recording), such as

Fig. 4. (A) Bar graph indicating the differences in the signal current value in cases of buffer solution (0 nM bombykal), 3 nM, 10 nM, and 30 nM bombykal for one oocyte expressing BmOR3. Error bar: SD. Each data point was obtained by 6 times experiments. Statistical significance was assessed by the t test, one asterisk: 0.1 < P < 1; two asterisks: P < 0.001. (B) Dose-dependent increases in amplitude of each ligand-induced current. Current recordings at different doses were measured on different oocytes. Error bar: SD. Each data point was obtained by 5 times experiments. (Inset) Current traces of oocytes expressing BmOR3 as a typical example. 15342 ∣


Fig. 5. Simultaneous current recordings using two oocytes that express receptors (A) BmOR3 and (B) BmOR1. We first added 100 nM bombykal solution 10 μL and second 3 μM bombykol solution 10 μL. Each solution was added at the time indicated by arrows.

Misawa et al.

Conclusions We have constructed a highly sensitive and selective odorant sensor using Xenopus oocytes expressing insect olfactory receptors. It was found that this sensor could detect different types of chemicals that have only a slight difference in chemical composition with sensitivity of a few parts per billion in solution. Moreover, we found that the sensor worked for multiple-target chemicals and can be integrated with a robotic system without noise reduction systems. Hence, this biohybrid sensor device provides an innovative platform to equip robots with odorant sensors. We believe that the sensor can potentially be incorporated into a machine that can be useful for several applications, including food administration, environmental monitoring, and health management.

Materials and Methods Device Fabrication and Oocyte Trapping. The device was composed of fabricated acrylic material, glass capillaries (B100F-3, World Precision Instruments, Inc.), and Ag/AgCl electrodes (0.4 mm in diameter AG-401355, The Nilaco Co.) made in advance by dipping in a commercial bleach agent overnight. The fluidic device was fabricated by machining a 5-mm-thick acrylic plate (Mitsubishi rayon, Acrylite L) with an automated computer-aided-design/ computer-aided-manufacturing modeling machine (Modia systems, MM100) with a 1-mm diameter drill. The glass capillary electrodes were fabricated from glass capillaries with filaments, with an outer diameter of 1.0 mm and an inner diameter of 0.75 mm (TW100F-3; World Precision Instruments, Inc.). The capillary was pulled using a P-2000 laser puller (Sutter Instrument Co.) preprogrammed to fabricate pipettes with an inner diameter of less than 50 μm. Parameters used were Heat: 260, Fil: 4, Vel: 100, Del: 150, and Pul: 90. The tips of the resulting capillaries had inner diameters of less than ∼50 μm. The Ag/AgCl electrodes were inserted into the glass capillaries for the measurements of the oocyte responses. The capillaries were initially filled with a 3-M KCl solution. The fluidic channel was filled with Barth’s solution (88 mM NaCl, 1 mM KCl, 0.3 mM CaðNO3 Þ2 , 0.4 mM CaCl2 , 0.8 mM MgSO4 , 2.4 mM NaHCO3 , 15 mM Hepes, pH 7.6). The solution was perfused constantly using a peristaltic pump. The solution flows in the channel at a flow rate of 1–2 mL∕ min. For the installation of an oocyte, the oocyte is first introduced and flowed into the fluidic channel, and it can be then trapped by insertion of two glass capillary electrodes embedded in the fluidic channel. In the case of multichannel measurements, we installed the multiple oocytes to the trapping area one by one before sealing with the lid of the fluidic channel; they were then captured after sealing with the lid (Fig. 5 Inset). The trapped oocytes are exposed to the solution containing chemicals solubilized by 1% DMSO. Even though a difference in protein expression has been reported between the vegetal pole and the animal pole of the oocytes (15, 31), the point of insertion of the electrodes into the oocytes is not considered to cause a significant difference in measurements using the TEVC method. Therefore, we did not control the orientation of the oocytes in our proposed system. Receptor Expression in Xenopus laevis Oocyte and Electrophysiological Recording. Stage V to VI oocytes were treated with 1.5 mg∕mL collagenase (Wako Pure Chemical Industries, Ltd.) in Ca2þ -free saline solution (82.5 mM NaCl, 2 mM KCl, 1 mM MgCl2 , 5 mM Hepes, pH7.5) for 1.5 h at 20 °C. The oocytes were then microinjected with 25 ng each of the odorant receptor gene (BmOR1 or BmOR3 or PxOR1 or DOr85b) RNA and coreceptor protein gene (Or83b family gene) RNA synthesized by mMESSAGE mMACHINE (Ambion Inc.). The injected oocytes were incubated for a few days at 20 °C in Barth’s solution supplemented with 10 μg∕mL penicillin and streptomycin. We used oocytes that were nearly the same size as each other [1.3 0.1 mm (SD), (n > 100)] and showed no significant differences in their responses. The signal currents of the oocytes were recorded with a two-electrode voltage clamping method using a custom-built multichannel amplifier (Triton, Tecella, LLC). The currents were monitored every 100 μs at a hold voltage of −80 mV. In the

Fig. 6. (A) Images of an odorant sensor using a Xenopus oocyte trapped and inserted between two electrodes in a fluidic device connected to the electric motor of a head-shaking robotic system outside of a Faraday cage. The head of a robot mannequin was made of styrofoam. (B) Sequential images and the current trace of the robot’s head shaking, which was triggered by an olfactory stimulus using Z11-16:Ald recognized by PxOR1. Each number in the sequential image corresponds to the equivalent number on the current trace. The threshold level is represented as a dashed line.

Misawa et al.

PNAS ∣ August 31, 2010 ∣

vol. 107 ∣

no. 35 ∣

15343

ENGINEERING

the patch clamp technique, typically employ a noise reduction system such as a Faraday cage to cancel out background noise and to maximize the signal-to-noise ratio. This configuration is usually indispensable when the sensor is used under noisy conditions. We mounted the sensor device onto a robotic system without any noise reduction system. As shown in Fig. 6A, we installed the sensor device into the head of a robot mannequin and connected it to an electric motor via an amplifier. The sensor device was configured with an oocyte expressing PxOR1, whereas the electric motor was configured such that when the sensor device detected a threshold value of Z11-16:Ald, this triggered actuation of the electric motor causing the robot’s head to move back and forth (i.e., shake) twice. We initially added 100 μL of buffer solution as a negative control and subsequently added 100 μL of a 10 μM Z11-16: Ald solution to the sensor. The system successfully showed a high tolerance to noise, even during shaking (i.e., back-and-forth motion) of the robot’s head (Movie S1). We evaluated the noise level determined by standard deviation of the peak values in the noise. As a result, the noise induced by the head shaking of the robot was 150 nA (SD) and the background noise was 10 nA (SD). On the other hand, the background noise was significantly high [260 nA (SD)] when we used a conventional open chamber that was positioned outside of the Faraday cage. We attribute the noise tolerance to (i) the closed and small fluidic system and (ii) the oocyte that is firmly impaled by the glass capillaries, which restrains its movement during the shaking. This demonstration implies that the sensor can be directly mounted on a moving robot programmed to react chemicals in the environment.

case of using oocytes expressing BmOR3, the response remained viable in the device for about 12 h at room temperature (Fig. S3). Chemicals were prepared in Barth’s solution including 1% DMSO. Concentrated stock solutions (20 μL) of each of the chemicals were added to the recording bath to provide the indicated final concentrations. Additionally, we think that systems for the solubilization of ligands into the device will help our system. Although ligands are dissolved by DMSO in this study, gas permeable materials and usages of odorant-binding proteins will be useful for incorporation of volatile chemicals into the fluidic device (32).

electric motor was controlled by the microcomputer to move back and forth two times when the signal was over the threshold. We simultaneously monitored the shaking of the robot head and the current tracing using the computer attached to the amplifier.

Robot Integration. The sensor device was set inside the head of a robot mannequin made of styrofoam, and ligand solutions were injected into the fluidic channel connected with a plastic tube passing through the head of the robot. The Ag/AgCl electrodes were connected to an electric motor through the amplifier and a programmed microcomputer (Coron, Techno-road Co.). The

ACKNOWLEDGMENTS. The authors thank Prof. I. Shimoyama and his colleagues at Information and Robot Technology Research Initiative of the University of Tokyo for helpful discussions, and Prof. M. Asashima of the University of Tokyo for kindly providing the Xenopus laevis oocytes. We also appreciate the generous support by Y. Tanaka of Tecella, LLC for customization of the multichannel amplifier and the substantial contribution by H. Ishihara, S. Takaya, and S. Sugimoto of the University of Tokyo for the robot integration. This work was partly supported by the New Energy and Industrial Technology Development Organization and by Tokyo Electron Ltd.

1. Mizuta Y, et al. (2008) Development of an oligo(ethylene glycol)-based SPR immunosensor for TNT detection. Biosens Bioelectron 24:191–197. 2. Lee JY, Ko HJ, Lee SH, Park TH (2006) Cell-based measurement of odorant molecules using surface plasmon resonance. Enzyme Microb Tech 39:375–380. 3. Kohl D, Heinert L, Bock J, Hofmann T, Schieberle P (2000) Systematic studies on responses of metal-oxide sensor surfaces to straight chain alkanes, alcohols, ketones, acids and esters using the SOMMSA approach. Sensor Actuat B-Chem 70:43–50. 4. Yano K, et al. (1998) Development of an odorant sensor using polymer-coated quartz crystals modified with unusual lipids. Biosens Bioelectron 13:397–405. 5. Wohltjen H, Dessy R (1979) Surface acoustic-wave probe for chemical-analysis. I. Introduction and instrument description. Anal Chem 51:1458–1464. 6. Ciosek P, Wroblewski W (2007) Sensor arrays for liquid sensing—electronic tongue systems. Analyst 132:963–978. 7. Touhara K, Vosshall LB (2009) Sensing odorants and pheromones with chemosensory receptors. Annu Rev Physiol 71:307–332. 8. Dahanukar A, Hallem EA, Carlson JR (2005) Insect chemoreception. Curr Opin Neurobiol 15:423–430. 9. Moczydlowski E (1986) Ion Channel Reconstitution, ed C Miller (Plenum, New York), pp 75–113. 10. Radhika V, et al. (2007) Chemical sensing of DNT by engineered olfactory yeast strain. Nat Chem Biol 3:325–330. 11. Figueroa XA, Cooksey GA, Votaw SV, Horowitz LF, Folch A (2010) Large-scale investigation of the olfactory receptor space using a microfluidic microwell array. Lab Chip 10:1120–1127. 12. Pantoja R, et al. (2004) Silicon chip-based patch-clamp electrodes integrated with PDMS microfluidics. Biosens Bioelectron 20:509–517. 13. Feng XJ, et al. (2007) A living cell-based biosensor utilizing G-protein coupled receptors: Principles and detection methods. Biosens Bioelectron 22:3230–3237. 14. Sakata T, Miyahara Y (2008) Noninvasive monitoring of transporter-substrate interaction at cell membrane. Anal Chem 80:1493–1496. 15. Sigel E (1990) Use of Xenopus oocytes for the functional expression of plasma membrane proteins. J Membrane Biol 117:201–221. 16. Klemic KG, Klemic JF, Reed MA, Sigworth FJ (2002) Micromolded PDMS planar electrode allows patch clamp electrical recordings from cells. Biosens Bioelectron 17:597–604. 17. Joshi PR, Suryanarayanan A, Schulte MK (2004) A vertical flow chamber for Xenopus oocyte electrophysiology and automated drug screening. J Neurosci Methods 132:69–79.

18. Dahan E, Bize V, Lehnert T, Horisberger JD, Gijs MAM (2007) Integrated microsystem for non-invasive electrophysiological measurements on Xenopus oocytes. Biosens Bioelectron 22:3196–3202. 19. Tan WH, Takeuchi S (2007) A trap-and-release integrated microfluidic system for dynamic microarray applications. Proc Natl Acad Sci USA 104:1146–1151. 20. Suzuki H, Pioufle BL, Takeuhci S (2009) Ninety-six-well planar lipid bilayer chip for ion channel recording fabricated by hybrid stereolithography. Biomed Microdevices 11:17–22. 21. Stuhmer W (1992) Electrophysiological recording from Xenopus oocytes. Methods Enzymol 207:319–339. 22. Sato K, et al. (2008) Insect olfactory receptors are heteromeric ligand-gated ion channels. Nature 452:1002–1006. 23. Sakurai T, et al. (2004) Identification and functional characterization of a sex pheromone receptor in the silkmoth Bombyx mori. Proc Natl Acad Sci USA 101:16653–16658. 24. Nakagawa T, Sakurai T, Nishioka T, Touhara K (2005) Insect sex-pheromone signals mediated by specific combinations of olfactory receptors. Science 307:1638–1642. 25. Mitsuno H, et al. (2008) Identification of receptors of main sex-pheromone components of three Lepidopteran species. Eur J Neurosci 28:893–902. 26. Nichols AS, Luetje CW (2010) Transmembrane segment 3 of Drosophila melanogaster odorant receptor subunit 85B contributes to ligand-receptor interactions. J Biol Chem 285:11854–11862. 27. Baumgartner W, Islas L, Sigworth FJ (1999) Two-microelectrode voltage clamp of Xenopus oocytes: Voltage errors and compensation for local current flow. Biophys J 77:1980–1991. 28. Strutz-Seebohm N, et al. (2006) Additive regulation of GluR1 by stargazin and serumand glucocorticoid-inducible kinase isoform SGK3. Eur J Physiol 452:276–282. 29. Wang G, Carey AF, Carlson JR, Zwiebel LJ (2010) Molecular basis of odor coding in the malaria vector mosquito Anopheles gambiae . Proc Natl Acad Sci USA 107:4418–4423. 30. Galdamez JR, Danner RP, Duda JL (2007) Application of mass spectrometer-inverse gas chromatography to study polymer-solvent diffusivity and solubility. J Chromatogr A 1157:399–407. 31. Miledi R, Sumikawa K (1982) Synthesis of cat muscle acetylcholine receptors by Xenopus oocytes. Biomed Res 3:390–399. 32. Vidic J, et al. (2008) On a chip demonstration of a functional role for odorant binding protein in the preservation of olfactory receptor activity at high odorant concentration. Lab Chip 8:678–688.

15344 ∣


Misawa et al.