RECEPTOR-BASED BIOSENSORS

Abstract

A bioelectronic sensor is disclosed for detecting presence of at least one volatile organic compound (VOC), the sensor comprising a single carbon nanotube (CNT) covalently immobilizing a single receptor optionally being a mammalian or an insect receptor, wherein association of the VOC to said receptor allows for a measurable electric field effect.

Claims

1-45. (canceled)

46. A bioelectronic sensor for detecting presence of at least one volatile organic compound (VOC), the sensor comprising a single carbon nanotube (CNT) covalently immobilizing a single insect receptor derived from mosquitos or flies, wherein association of the VOC to said receptor allows for a measurable electric field effect.

47. The sensor according to claim 46, wherein the at least one VOC is indole.

48. The sensor according to claim 46, wherein the receptor is mosquito-derived indolergic odorant receptor (OR).

49. A bioelectronic sensor for detecting presence of at least one indole, the sensor comprising a single carbon nanotube (CNT) covalently immobilizing a single mosquito-derived indolergic odorant receptor (OR), wherein association of the at least one indole to said receptor allows for a measurable electric field effect.

50. The sensor according to claim 49, wherein the OR is selected from OR2, OR9, OR10, OR co-receptor (Orco) or a combination thereof.

51. The sensor according to claim 49, wherein the OR is OR8.

52. The sensor according to claim 50, wherein the OR is OR9 or OR9 co-receptor.

53. The sensor according to claim 46, when assembled into a biochip interfacing a printed circuit board (PCB).

54. The sensor according to claim 53, configured as an array of sensors that are assembled into a biochip interfacing a printed circuit board (PCB).

55. The sensor according to claim 54, wherein the array is segmented into specific biosensing regions enabling multiple ligand bindings to be interrogated in real-time.

56. The sensor according to claim 53, comprising one or more measurement channels configured for simultaneous real-time interrogation.

57. The sensor according to claim 56, wherein each measurement channel is appended with a drain and a source electrodes.

58. The sensor according to claim 53, wherein the PCB further comprises at least one element selected from multiplexers, decoders and analog-to-digital converters.

59. The sensor according to claim 46, wherein the CNT is selected from single-walled carbon nanotube (swCNT), double-walled carbon nanotube (dwCNT) and multi-walled carbon nanotube (dwCNT).

60. The sensor according to claim 46, wherein the receptor is a single OR9 receptor or Orco covalently associated to a single swCNT molecule.

61. The sensor according to claim 47, wherein the indole is selected from indole (1H-indole), skatole, indole-3-butyric acid, indole-3-acetic acid, indole-, 3-carbinol, tryptophan and beta carboline.

62. The sensor according to claim 51, wherein OR8 is selective towards 1-octen-3-ol.

63. A single-molecule field-effect transistor (smFET) device comprising one CNT and a capture probe covalently coupled thereto, wherein the capture probe comprises at least mosquito-derived indolergic odorant receptor (OR)configured to bind or associate to at least one VOC, wherein the smFET further comprises at least one electrode assembly disposed proximate opposing ends of the one CNT to electrically couple the one CNT to a substrate onto which the smFET is disposed, wherein the capture probe is not DNA or a nucleotide.

64. A method for detecting or sensing presence of at least one volatile organic compound (VOC), such as an indole, present in a sample, the method comprising contacting a bioelectronic sensor comprising a single carbon nanotube (CNT) covalently immobilizing a single mosquito-derived indolergic odorant receptor (OR), under conditions permitting association of the VOC to said receptor, wherein association of the VOC to said receptor causes a measurable electric field effect.

65. The method according to claim 64, wherein the sensor is a single-molecule field-effect transistor (smFET) device comprising one CNT and a capture probe covalently coupled thereto, wherein the capture probe comprises at least one receptor configured to bind or associate to at least one VOC, wherein the smFET further comprises at least one electrode assembly disposed proximate opposing ends of the one CNT to electrically couple the one CNT to a substrate onto which the smFET is disposed, wherein the capture probe is not DNA or a nucleotide.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to better understand the subject matter that is disclosed herein and to exemplify how it may be carried out in practice, embodiments will now be described, by way of non-limiting example only, with reference to the accompanying drawings, in which:

[0076] FIG. 1 provides an illustration of an Orco-CNT FET. An amine-terminated lipid is covalently coupled to defects generated on a single-walled carbon nanotube (SWCNT) sidewall via diazonium chemistry. This lipid serves as a nucleation point for the assembly of Orco-containing membrane fragment. The CNT acts as the channel in a field-effect transistor, transducing biomolecular charge into a conductance change in the FET. Binding and dissociation of targets generate random telegraph signals (RTS, shown in inset), which correspond to open and closed states of the ion channel.

[0077] FIGS. 2A-B depict CNT-FET design and fabrication. FIG. 2A- Representative illustration of metal source-drain electrode pattern, with individual CNTs bridging each pair. Also shown (right) is a microscopy image of a fabricated chip. FIG. 2B- an example of CNT-FET process flow. (a) CNTs are either CVD-grown or spun. (b) Lithography patterning for source and drain electrical contacts (e.g., Titanium). (c) Metal (e.g., titanium) deposition and photoresist removal. (d) Patterning for pseudo-reference (e.g., platinum) gate electrodes followed by (e) deposition and photoresist removal. (f) Patterning for metal (e.g., gold) bond pads.

[0078] FIGS. 3A-D provide CNT FET devices characterization. FIG. 3A- SEM image showing a CVD-grown CNT spanning five electrode pairs. Inset: two isolated CNT fragments. FIG. 3B- left: spin-cast CNTs produce single-nanotube crossings. Inset: a zoomed-in of the crossing. Right: AFM image of a CNT fragment capped by a source and drain Titanium electrodes. FIG. 3C- Raman spectra collected from individual CNT fragments demonstrating reproducible Raman features, with ωG.sup.+ and ωG.sup.- located at 1588 cm.sup.-1 and 1566 cm.sup.-1, respectively. Inset shows the RBM spectra of the same nanotubes, located at 170 cm.sup.-1 corresponding to a diameter of 1.45 nm FIG. 3D- I-V curve of a CNT device backgated in air showing ambipolar transport characteristics.

[0079] FIGS. 4A-C demonstrate how octadienal and carvacrol inhibit VUAA1-activated Orco. FIG. 4A- Chemical structures of tested compounds. FIG. 4B- Representative current traces of oocytes expressing Aedes aegypti Orco following exposure to 2.10-4 M VUAA1 alone or in combination with equimolar concentrations of the antagonists octadienal and carvacrol. FIG. 4C- Normalized responses of AaegOrco to VUAA1alone or in combination with octadienal or carvacrol. Odorants effects were statistically significant (one-way ANOVA followed by Tukey’s post-test; P < 0.0001; mean responses ± s.e.m.; VUAA1 alone, n = 7; octadienal & carvacrol n = 6).

[0080] FIGS. 5A-E show electrochemically regulated functionalization of CNT FET.

[0081] FIG. 5A- Complete biofunctionalization scheme includes the voltage bias induced diazonium (FBDP) reaction followed by conjugation of amine-terminated phospholipid (DPPE) via reductive amination. FIGS. 5B-D depict electrochemically-controlled CNT-FET diazonium modification. FIG. 5B- Electrical behavior of CNT devices before and after FBDP sequential exposures, as reflected by I-Vlg measured after each exposure. Decrease in on-state current is a result of carrier scattering due to orbital rehybridization. At 0 V vs. platinum reference electrode, small-band-gap CNT (denoted ‘semi metallic’) show sharp exponential kinetics (circles) while large-band-gap CNT show very slow kinetics with longer saturation times (squares). At a sufficiently high overpotential of -1.5 V, accelerated kinetics is observed resulting in reaction saturated following the first few seconds of exposure. Current values were normalized to the maximum value. FIG. 5C- Representative I-t trace of a device after FBDP introduction. Discrete current drops are observed. The resistance change due to the addition of a single sp.sup.3 defect has been shown to reduce transconductance from h/4e.sup.2 to h/2e.sup.2. FIG. 5D- Raman spectra of CNT devices before and after reaction at -1.5 V exhibited a decrease in G peak and appearance of a D peak. FIG. 5E- Specific CNT coupling of FBDP is indicated by labeling with amine-modified gold nanoparticles. Minimal background adsorption is observed.

[0082] FIG. 6 illustrates an OR-CNT FET. An amine-modified lipid is covalently coupled to defects, generated on a single-walled carbon nanotube (SWCNT) sidewall via diazonium chemistry. This lipid serves as a nucleation point for the assembly of OR8-containing supported lipid bilayer (OR8 receptor is used as an example). The CNT acts as the channel in a field-effect transistor, transducing biomolecular charge into a conductance change in the FET. Binding and dissociation of targets generate random telegraph signals (RTS, shown in inset).

[0083] FIGS. 7A-D illustrate fabrication of OR-integrated CNT FET devices. FIG. 7A-SEM (left) and AFM (right) images showing a homogenous CNT network dispersed on a Si/SiO2 substrate. FIGS. 7B-A scheme of the preliminary chip design showing an array of 12 FET devices and 2 pseudo-reference electrodes. CNT networks are patterned as a FET channel of 50 .Math.m × 100 .Math.m (shown in inset). FIG. 7C- Platinum gate and titanium source and drain electrodes are patterned and deposited by e-beam deposition. FIG. 7D-Each CNT network device is extensively characterized structurally, using SEM and AFM, electronically by using Raman spectroscopy and finally, by measuring its I-V characteristics.

[0084] FIGS. 8A-C wherein FIG. 8A- Chip after fabrication process, FIG. 8B-Microscope image, shows gate metal and source and drain metal in magnification 5X. FIG. 8C- SEM image of device, show the network CNT in scale bar corresponds to 100um.

[0085] FIGS. 9A-B demonstrate FBDP synthesis and characterization. FIG. 9A-Synthesis of FBDP from 4-aminobenzaldehyde. FIG. 9B- FT-IR spectrum of the synthesized FBDP.

[0086] FIGS. 10A-B demonstrate a process for producing OR-containing membrane fragments. FIG. 10A shows crude membrane fractions harvested from Xenopus oocytes. FIG. 10B illustrates the process.

[0087] FIGS. 11A-B show dynamic light scattering analyses, demonstrating reproducibility in generating nanovesicles. FIG. 11A demonsartes size distribution by intensity, FIG. 11B shows sizes distribution by volume.

[0088] FIGS. 12A-C demonstrate OR-containing membrane functionalization of CNT-FET. FIG. 12A- An SEM image of nanovesicles dispersed of a Si/SiO.sub.2 substrate following separation and fragmentation of oocytes membranes. FIG. 12B- AFM image of nanovesicles attached to a CNT network FET device showing the attached nanovesicles (marked with arrows). FIG. 12C- an amine-modified fluorescent probe is covalently attached to a CNT device that was previously modified by diazonium (FBDP) whereas a similar probe is easily removed by rinsing when attached to pristine unmodified CNTs.

[0089] FIG. 13 shows fluorescence response of FITC-NH.sub.2 (fluorescent label) coupled to FBDP-modified CNT FET. The attachment performed with flow system where reagents applied to confined region of 200 (w) x280 (h) .Math.m.sup.2 (marked by dashed lines in the left scheme. The scheme is not in scale).

[0090] FIGS. 14A-B show selective attachment of Cy3-labled protein to CNT FET. FIG. 17A CNT FET device without TMPS, FIG. 14B CNT FET device pretreated with TMPS. The samples were incubated in solution of 0.1 .Math.M protein for 90 min at 4° C. and then in PBS for 15 min at 60° C. Fluorescence emission was collected at 570 nm.

[0091] FIGS. 15A-B demonstrate a CNT FET experimental setup and electrical measurements. FIG. 15A- Each device is electrically characterized by measuring its I-V (transconductance) in air (using the Si substrate as a backgate and applying a gate bias of -10 V to +10 V) or in liquid (electrolytic gate applied by the on-chip Pt gate electrodes from -0.8 V to +0.8 V and source-drain bias of 100 mV). Shown here are backgated I-V plots of four devices with metallic conductance ranging between 60-140 nA. Two devices had ‘opens’ (were disconnected). FIG. 15B- Measurement setup includes a probe station and a source-measure unit that enable a thorough electric characterization of CNT-FET devices and additional biosensing measurements. A microfluidic flow cell is stamped on the chip, defining a flow channel of 200 .Math.m wide. A syringe pump is used to introduce different solutions and reagents.

[0092] FIG. 16 depicts a biofunctionalization scheme. Amine-terminated phospholipids (DPPE) are tethered to the sidewalls of diazonium modified CNTs, functioning as FET channels, via reductive amination. Subsequently, OR-containing membrane fragments, generated from nanovesicles by a combination of ultrasonication and ultracentrifugation, are attached to the covalently bound phospholipid.

[0093] FIGS. 17A-D demonstrate that the mosquito odorant receptor 8 is a highly sensitive and selective octenol receptor. FIG. 17A- OR8 from the mosquito species Aedes aegypti (AaOR8) and Toxorhynchites amboinensis (TaOR8) are highly sensitive to (R)-1-octen-3-ol (parts per billion range, ppb). FIG. 17B- OR8 is an enantioselective octenol receptor, distinguishing between the (R)-enantiomer, which is the most abundant form in nature, and the (S)-enantiomer. FIG. 17C- At low concentration (parts per billion range), OR8 is not activated by other chemical classes. FIG. 17D- At high concentration (parts per thousand), OR8 exhibits a broader chemical receptive range.

DETAILED DESCRIPTION OF EMBODIMENTS

[0094] The majority of methods for single-molecule studies have only become available in the last decade, with developments in fluorescence-based techniques enabling confocal or TIRF measurement of fluorescence resonance energy transfer (FRET) and force-based methods such as atomic force microscopy (AFM) and optical tweezers. The development of single-molecule observation and manipulation techniques has revolutionized the understanding of many biological processes. Studies of ribosome translation elongation and initiation stages; conformational dynamics of water-soluble and even membrane-protein folding; DNA hybridization and replication; enzymatic catalysis; membrane receptor oligomeriztion GPCR interactions are only a few examples of the recent applications of single-molecule methods.

[0095] Established single-molecule methods remain limited by technical obstacles such as difficulties associated with fluorescent labeling, the need to invasively apply external forces to the biochemical system under investigation, limitations in the time resolution with which relatively fast biochemical events can be characterized, and limitations in the total time during which the biochemical reaction under investigation can be observed. For example, Fluorescent probes are fundamentally limited by photobleaching, yielding as few as 10,000 photons for organic dyes and temporal resolution of milliseconds.

[0096] Single-molecule bioelectronics offer an exciting new experimental approach based on direct electronic transduction of specific binding into electrons. Direct electronic transduction avoids the use of optics and light sources and allows low-form-factor devices as well as delivers signal levels that are orders of magnitude higher than those achieved with optical detectors.

[0097] Bioelectronic detectors comprise an electronic transducer functionalized with a bio-recognition element. Among those, single-molecule field-effect transistors (smFETS) have the unprecedented ability to reach submicrosecond time scales in a noninvasive, label-free study of biomolecular interactions and kinetics, overcoming the limitations of existing techniques. Moreover, devices based on electronic platforms are attractive since they are amenable for miniaturization and can be manufactured using conventional microelectronic fabrication techniques.

[0098] Particularly promising are carbon nanotube (CNT) field-effect devices due to their extraordinary properties making them excellent candidates for exposed gate biosensors. CNTs readily form the conducting channel in field-effect transistor (FET) configuration exhibiting an exceptionally high charge carrier mobility and an extremely stable lattice. Most importantly, the density of charge carriers in these 1D materials is sensitive to charges in the environment and therefore the conductance can be modulated by adsorbed molecules. In addition to their biocompatible all-carbon composition, their dimensions are comparable to the size of single biomolecules thus solving the typical problem of ‘form factor mismatch’ between biology and solid-state interfaces. Finally, CNT-FET devices are attractive since they are manufactured using traditional microelectronic fabrication techniques. As discussed, the transduction mechanism of CNT-FET is based on the significant conductance changes induced by the analyte-nanotube interaction, which implies that this detection approach is label-free. The feasibility of CNT-FET biosensors was demonstrated in the detection of nucleic acids and various protein biomarkers. The lack of specificity and sensitivity often associated with nanotube sensors is a direct result of the fact that the entire 1D conductor is uniformly sensitive to local charge density, rendering the sensor susceptible to nonspecific adsorption anywhere on the exposed surface.

[0099] Point-functionalized single-walled carbon nanotube (SWCNT) devices are emerging as an all-electronic, label-free, single-molecule detection platform. This smFET is characterized by a conductance that is sensitive to charges localized within a few Debye lengths of a point defect that is generated on the SWCNT sidewall. This site of functionalization serves as the point of attachment of an SWCNT-tethered probe molecule under study. Under a source-drain voltage bias of several tens to hundreds mV, the current signal level of a typical smFET is tens of nanoampere (nA). smFETs have been successfully used for the study of DNA hybridization and melting kinetics and DNA and protein conformational dynamics.

[0100] The proposed smFET comprises Orco or alternatively OR or alternatively a combination of OR-Orco or any other non-nucleic acid receptor molecule derived from animal, plant, insects, bacteria or fungi, as disclosed herein, as a biorecognition element and applies a single-molecule biophysical approach with a custom-developed bioelectronic assay platform in the form of biofunctionalized CNT-FET arrays assembled into a biochip that is interfacing a printed circuit board (PCB). The array will be segmented into specific biosensing regions enabling multiple ligand bindings to be interrogated in real-time. The board may include several dozen or several hundred independently addressable measurement channels that are simultaneously interrogated in real-time. The drain and source potential for each channel are fully tunable and are composed of at least one gain stage and an anti-aliasing filter topology of high order (at least second-order). An FPGA module may be incorporated to interface the hardware and software. The PCB may also include multiplexers, decoders, and analog-to-digital converters. The smFET biochip will be packaged via a chip-carrier and mounted on the board. A designated software enabling real-time measurements and data acquisition may be used. Alternatively, CMOS integrated biochips may be designed, which contain all the required electronics on-chip and can therefore include several thousand to millions smFET devices on a single chip.

[0101] A designated microfluidic flow channel is designed, which includes inlets and outlets for the introduction of different reagents. The channel may be fabricated from a silicon elastomer such as poly-dimethylsiloxane (PDMS) or any other biocompatible polymer.

[0102] In some implementations of a device of the invention an amine-terminated phospholipid such as 1,2-dipalmitoyl-sn-glycero-3-phosphorylethanolamine (DPPE) is conjugated to the CNT side wall serving as an anchor point, directing the attachment of the membrane fragment, as shown in FIG. 1. By using this new bioelectronic assay, the phase transfer and solubility of volatiles (and their effect on the signal), the electrostatics influence on diffusion-limited capture times and the altering of binding kinetics by an electric field, can be measured. Furthermore, issues of non-specific adsorption and interference are mitigated. The temporal open-closed states distribution of the ion channel can be monitored directly, in real-time. It is suggested that a stronger effect on CNT conductance will be observed due to cationic current in the open state. Despite being incorporated within a lipid bilayer, it is suggested that the channels’ transient conformational changes should affect the electric field and modulate the CNT conductance as well.

[0103] In another implementation, the nanoscale membrane fragment can contain both constituents of the receptor complex, namely Orco-OR, as shown in FIG. 6. Such configuration combines the sensitivity and intrinsic signal amplification of CNT-FET devices with the unique selectivity of mosquito OR.

[0104] Design, fabrication and characterization of CNT-FET array. Chips containing an array of CNT-FET devices are fabricated on a silicon/silicon oxide substrate by either using routinely employed CVD growth techniques or by using drop casting or spin-coating technology combined with photo- or electron beam lithography. Alternatively, devices may be fabricated on any other insulating substrate, such as: metal oxides, polymers etc. said substrate may be of any thickness (provided it is thicker than the deposited metal electrodes, i.e., 90 nanometers) and hardness. As such, substrate may be a flexible material with various degrees of elasticity or plasticity. Spin-cast method is used to randomly place a dilute aqueous single-walled CNT suspension. The nanotube density, spin-cast parameters, nanotube length and electrodes width and gap are optimized in order to maximize the frequency of single CNT crossings. The electrodes pattern (gaps and geometry) are designed to be commensurate with the nanotube length distribution (as shown in FIG. 2A). The optimizations results in single nanotube crossing between each electrode pair. Subsequently, patterning and metal deposition of the source, drain and gate electrodes should result in multiple single CNT-FET devices per chip. In the next step, the devices are characterized using optical and electrical measurements (see FIG. 3). Chips are scanned by SEM and AFM. Raman spectra of individual CNTs are collected as well. Finally, electronic properties are measured by I-V probing in air, using a silicon backgate.

[0105] Chips containing multiple CNT-FET devices are exhibiting relatively high conductance (tens to hundreds of nA), as shown in FIG. 3A. The feasibility of spin-cast method to reproducibly generate a high yield of functional CNT-FETs has been shown. An example of a CVD-grown chip as compared with a spin-cast chip is shown in FIGS. 3A and 3B.

[0106] Also shown are AFM images, which help to confirm the CNT diameter. A CNT resonant Raman spectroscopy analysis measures the ratio of the D band (“disorder mode”) to the tangential G band, which serves as an indicator to the number of intrinsic CNT defect. In addition, the electronic structure is elucidated by analyzing the G mode components. The radial breathing mode (RBM) is collected, as well, indicating the CNT diameter. Only CNTs demonstrating no scattering at the D band (apparent “pristine” nanotubes) are used in the constructions of the devices. Typical Raman spectra of a CNT grown on Si/SiO2 is shown in FIG. 3C.

[0107] A preliminary study was based on CNT network FET devices (CNTN-FET). Such design enables chip production at very high yields. We have optimized the parameters for generating a reproducible and homogenously dispersed CNT layer on a Si/SiO2 chip. Briefly, pure, as-prepared CNT suspension (containing pre-sorted SWCNTs with a narrow size distribution) was used to coat a Si/SiO2 substrate by a brief sonication followed by a spin-casting method while adjusting concentration, solvent type, incubation period, sonication power and time. The layer was characterized by SEM, AFM and Raman, as shown in FIG. 7A. Titanium source and drain electrodes and platinum pseudo-reference gate electrodes were photolithographically patterned and deposited above the CNT layer. An Illustration of the preliminary design is shown in FIG. 7B. For each FET device, a CNT conducting channel was defined (as shown in FIG. 7C) and the remaining CNT layer was plasma etched. The resulting devices demonstrated homogenous CNT density across the chip (as shown in the SEM image of FIG. 7D). Devices were further characterized structurally and electrically as previously mentioned.

[0108] Additional images of a fabricated chip that comprises 90 nm thick titanium source and drain electrodes and platinum gate electrodes, are presented in FIGS. 8A-B. Additional SEM image is presented in FIG. 8C.

[0109] Orco expression, electrophysiology and purification. Supply of Orco-rich membranes will be predicated on the functional validation carried out using the two-electrode voltage clamp of Xenopus oocytes expressing this receptor. Expression and recordation of the pharmacological activity of mosquito odorant receptor complexes (ORx-Orco) in Xenopus oocytes has been studies extensively. This technique to study the agonist activity of Orco agonists (e.g., VUAA1) and antagonists (e.g., octadienal and carvacrol) (as shown in FIG. 4A) or associated with ORx has been achieved. In some cases Orco rather than ORx was used in order to reduce potential pitfalls associated with OR complexes. In some cases, ORx alone is not functional without its co-receptor chaperone. Two types of expression vectors adapted to Xenopus oocyte expression are used. pSP64DV was used to record currents reaching over 1 .Math.A from Orco exposed to VUAA1, indicating high protein levels (FIG. 4B). Considering the objective to extract Orco-rich membrane fragments less than <50 nm in length, receptor density should be sufficient to ensure that these short fragments contain Orco. To evaluate Orco density, this gene was expressed using the pCMV6-AC-mKate expression vectors (Origene), or other tagged version. This plasmid functions well in Xenopus oocytes and adds a constitutively fluorescent mKate2 tag on the C-terminus of the protein of interest (here Orco) that can be directly or indirectly (fluorescence immunohistochemistry) detected by confocal scanning laser microscopy. If the C-terminal tagging compromises Orco function, the N-terminal mKate2-tagging vector instead (pCMV6-AN-mKate, Origene) will be required.

[0110] Orco is synthesized from linearized pSP64DV expression vectors using the mMESSAGE mMACHINE™ T7 ULTRA Transcription Kit (Life Technologies). Capped Orco RNA is injected in stage V-VI oocytes, incubated for 3-5 days and exposed to the agonist VUAA1 based on protocols published elsewhere.

[0111] During a typical recording session, individual oocytes are exposed to a single concentration of 10.sup.-4 M of VUAA1 in order to (i) limit the exposure of ORco to this agonist and (ii) to identify the oocytes with highest Orco expression levels. Qualified oocytes are collected to isolate and evaluate the density of membrane receptor complexes.

[0112] The isolation of small oocyte membrane patches is carried out via lysis, homogenization and a subsequent differential ultracentrifugation in a sucrose gradient. Similar protocols were originally developed as sample preparation methods for both confocal microscopy and high-resolution AFM analysis of Xenopus oocytes membranes. It was shown that patches of solubilized planar oocyte membranes with lateral dimensions <500 nm and thickness of 3-5 nm were generated and retained for prolonged periods.

[0113] Chemical modification and bio-functionalization. Different versions of CNT-FET sensors have attempted to detect biomolecules adsorbed onto pristine and coated CNTs. Transient non-covalent attachment has been pursued using pyrenes or porphyrins exploiting π-π stacking of these molecules with the CNT carbon lattice. The lack of specificity and sensitivity usually associated with nanotube sensors is a direct result of the fact that the entire 1D conductor is uniformly sensitive to local charge density, rendering the sensor susceptible to nonspecific adsorption anywhere on the exposed surface. Introducing a defect onto the CNT surface localizes this sensitivity, making the CNT only sensitive to charge density in the region near the molecule under study. This defect can, in turn, be used to covalently link a biomolecule at the point of charge sensitivity of the transducer. The resulting device can be used to measure time-resolved changes in the conductance of the nanotube that arise from changes in scattering in the 1D channel caused by Coulomb interactions between the biomolecule and the defect. Covalent attachment strategies also offer desirable permanent tethering of the biomolecule. Among all methods, the reaction of CNTs with diazonium salts has been widely studied, becoming one of the most popular routes of CNT covalent functionalization. Covalent modification impart a measurable resistance change in the device by converting carbon bonding from sp.sup.2 to sp.sup.3 orientation. Controlling the reaction electrochemically is feasible by applying solution bias to promote or inhibit electron transfer between the charged diazonium and the CNT lattice (the associated resistance change due to the addition of one sp.sup.3 defect has been argued to be on the order of: h/4e.sup.2). The coupling of 4-formylbenzene diazonium hexafluorophosphate (FBDP), which contains an orthogonal aldehyde functional group, later used for bio-conjugation is electrochemically regulated. A subsequent tethering of a linker molecule (e.g., DPPE) is carried out via reductive amination, as shown in FIG. 5A. This covalently attached lipid serves as an anchor directing the attachment of an Orco-containing membrane fragment. The major advantages of this functionalization strategy are: i) locating the interrogated biomolecule in intimate contact with the charge sensitive region; ii) locating the biomolecule within a Debye sphere around the CNT sidewall (the requisite proximity scales with a Debye screening length) thus enabling electrostatic modulation of Orco binding and conformational changes; iii) ensuring optimal orientation of Orco-containing membrane fragment with the CNT sidewall, perpendicular to the cation flow in the channel; iv) enabling a non-transient, stable Orco-CNT hybrid for prolonged measurements. Studies of SLB (suspended lipid bilayers)-CNT hybrids have shown promise in electrical detection of target binding. The shift in the transistor threshold (ΔV.sub.e), due to additional charges, was shown to be related to its charge density by: σ = 2 ΔV.sub.eℇ.sub.wℇ.sub.0 / λ.sub.d where ℇ.sub.W, is the dielectric constant of water and λ.sub.d is the Debye length.

[0114] The directed attachment of the as-obtained membrane fragments to the covalently bound phospholipid largely depends on the membrane’s planar configuration. It is therefore paramount to understand the conditions allowing for a bilayer membrane fragment to maintain planar configuration in solution. The dynamics of lipid bilayer vesicles has been extensively studied using SLB as a model. Vesiculation or planarization of bilayer membrane is dependent on the free energy of each state, which is in turn, dependent on the entropy of closure and the membrane bending and contour (edge tension) free energies. The bending energy per unit area is given by:

[00001]$e_b=\frac{1}{2}\kappa {\left(c_1+c_2\right)}^2+\overline{\kappa (}{c}_1{c}_2)$

where κ is the bending rigidity and the bending modulus, which indicates the membrane malleability. The principal curvatures, .sub.C1+.sub.C2 are the eigenvalues of a curvature tensor that describes the local shape of the membrane. The structural origin of edge tension (γ, the contour energy per unit length) arises from the deformation of the lipids that occupy the edge. It has been suggested that the stability of a planar or a spherical lipid bilayer can be described by:

[00002]$\alpha =\left(\frac{\gamma }{K_b}\right){\left(\frac{A}{\pi }\right)}^{0.5}$

where y is the edge tension,

[00003]${\kappa }_b=2\kappa +\overline{\kappa }$

describing the bending free energy and A the area of the membrane. Using empirical values of

[00004]${\kappa }_b\sim \text{5-25}{k}_{b}T\left(k_b\right)$

is Boltzmann constant) and γ=1-2 k.sub.bT/l (l is length), it follows that planar lipid bilayers < 50 nm are sufficiently stable. In addition, incubation of the membrane fragments with the CNTs above their transition temperature may further facilitate their correct orientation. Consequently, small planar Orco-containing membrane fragments will be tethered to the CNT sidewall displaying an angular position that is defined (and limited) by the rotational freedom of the CNT-bound lipid tail. It should be noted that although the vesicular configuration is less likely to conjugate to the CNT-phospholipid (since it depends on the vesicle rupture), the closure of a planar bilayer post-conjugation, on the other hand, will not necessarily compromise signal transduction and device performance.

[0115] Moreover, other orientations that result in the positioning of the Orco ion channel parallel to the CNT sidewall (i.e. due to adsorption of the lipid tails to the CNT lattice) should still be able to effectively modulate the device conductance, as long as the ligand-induced ionic current flows within the Debye sphere.

[0116] A single tethered Orco channel affects a measurable conductance change in any given device due to the sensitivity of the 1D channel to electrical charges in the vicinity of the defect as previously described. In fact, the ability of the CNT-FET to resolve the molecular charge of a single nucleotide (~0.1 e) has been demonstrated using a similar device and a tethered nucleic acid.

[0117] In some embodiments, the chemical modification and bio-functionalization steps will be carried out via microfluidic channels (patterned in a Polydimethylsiloxane (PDMS) elastomer stamping).

[0118] Preliminary results. Previously, we have demonstrated our ability to electrochemically control the extent of a CNT-FET diazonium (FBDP) functionalization to a resolution of a single sp.sup.3 defect. Electrochemical control over FBDP functionalization of SWCNT was attained by shifting its Fermi energy (via application of voltage bias), thereby increasing electron density near the surface. The extent of sidewall modification was demonstrated by repeated I-V.sub.1g (liquid gate) measurements following consecutive exposures to FBDP, as shown in FIG. 5B. Moreover, the reaction can be monitored in real-time by measuring the conductance upon FBDP exposure (I-t traces). In this case, as shown in FIG. 5C, current drops originating from the diazonium reaction are clearly visible. These drops are quantized and represent the generation of a single sp.sup.3 defect, as evident, for example, from the altered Raman features and AFM imaging shown in FIGS. 5D and 5E. The rehybridization of sp.sup.2 to sp.sup.3 eliminates current paths in the 1D channel resulting in the observed decreased conductance.

[0119] The diazo compound formylbenzenediazonium hexafluorophosphate (FBDP) was synthesized from 4-aminobenzaldehyde in acidic solution at -10° C. as depicted in FIG. 9A. The product was purified by filtration and characterized by FT-IR (FIG. 9B). The spectrum features the IR stretch bands of the aldehyde group, the para-substituted benzene group, and importantly, the formed diazonium group at 2300 cm.sup.-1.

[0120] In addition, in a preliminary study, OR-containing membrane fragments were obtained by a combination of ultrasonication, ultracentrifugation and detergent treatment, and attached to a chemically modified CNT sidewall. Briefly, crude membrane fractions were harvested from Xenopus oocytes (FIG. 10A). Typically, oocytes can be harvested 2-5 days after RNA injection. Batches of 10-15 oocytes are homogenized in 1 mL of HEDP buffer (100 mM HEPES, 1 mM EDTA, pH 7.6 with NaOH) plus 5 uL of protease inhibitor cocktail performed at 4° C. Homogenates are centrifuged at 5600 rpm for 10 min. the supernatant obtained from step is pipeted into two 7 mL of ice-cold 15% sucrose in HEDO buffer, centrifuged at 175,000 g for 1.5 hours. The process is illustrated in FIG. 10B.

[0121] The inventors’ ability to reproducibly generate nanovesicles (that are ~ 30 nm in size, is proven as indicated by dynamic light scattering analyses, shown in FIGS. 11A-B.

[0122] These generated nanovesicles (Also scanned by using electron microscopy, as shown in FIG. 12A) were subsequently attached to CNT devices (as shown in the AFM image in FIG. 12B). We have further demonstrated the robustness of our covalent bonding strategy by conjugating a fluorescent probe to an FBDP-modified CNT device, as shown in FIG. 12C.

[0123] Receptor attachment with flow system. We proceeded to establish conditions for receptor attachment with flow system. With this setup, the sequential modification of the CNTs with FBDP and receptor coupling via reductive amination are performed by flowing the reagents through 200 (w) x280 (h) .Math.m.sup.2 channel that confines the entire CNTs region and partial areas of the Ti electrodes (see depiction in FIG. 13). As shown in FIG. 13, in addition to the covalent attachment of FITC-NH.sub.2 (fluorescent label) to the CNTs, direct adsorption to the bare SiO.sub.2 is prominent.

[0124] We have previously shown that prolonged incubation at elevated temperature enabled effective clearance of FITC-NH.sub.2 from the bare SiO.sub.2 surface. However, in order to achieve selective attachment without the necessity of rigorous washing and sample heating, we have first activated the SiO.sub.2 surface with propyltrimethoxysilane (TMPS). We expect that in addition to create a physical barrier, formation of hydrophobic surface by the propyl residue will repel incoming polar and/or charged molecules such as ORs. FIG. 14 demonstrates the implementation of our approach for attaching a model protein labeled with Cy3 fluorophore to CNT FET devices. FIG. 14A shows a device that was incubated in protein solution without TMPS activation, and FIG. 14B shows a similar device that was treated with TMPS prior to incubation with the protein. While prominent amount of protein was adsorbed to the bare SiO.sub.2 surface (not treated with TMPS), the TMPS treated device shows selective adsorption to CNTs.

[0125] Experimental setup, PCB design and measurement platform. The feasibility of the proposed bioelectronic constructs can be demonstrated by probing individual devices with a commercially available electronic measurement setup (probe station). The biochips, however, are designed to interface a custom designed PCB enabling a highly efficient and convenient measurement. The PCB may include several dozen or several hundred independently addressable measurement channels that are simultaneously interrogated in real-time. The drain and source potentials for each channel are fully tunable and are composed of at least one gain stage and an anti-aliasing filter topology of high order (at least second-order). An FPGA module may be incorporated to interface the hardware and software. The PCB also includes multiplexers, decoders, and analog-to-digital converters. The biochip will be packaged via a chip-carrier and mounted on the board.

[0126] A designated software enabling real-time measurements and data acquisition may also be used. Alternatively, CMOS integrated biochips may be used, which contain all the required electronics on-chip and can therefore include several thousand to millions smFET devices on a single chip.

[0127] A designated microfluidic flow channel may also be present, which includes inlets and outlets for the introduction of different reagents. The channel may be fabricated from a silicon elastomer such as poly-dimethylsiloxane (PDMS) or any other biocompatible polymer.

[0128] Measurement matrix: a measurement matrix may be formed by filling the flow cell with an electrolytic buffer or an electrically conductive hydrogel or electrically conductive polymer thus providing the Orco-containing membrane with the optimal conditions, ensuring higher activity and sensitivity. Solubilized ligands may be used during the preliminary data collection. Addition of small volume of organic solvent (e.g., DMSO) will enable the dissolution of the rather hydrophobic small organic compounds. The diffusion coefficient of VOCs in air is larger by orders of magnitude compared to water. Mass transfer of a VOC across air-water interface is dependent on its dissolution rate constant (k.sub.d), surface area of the air-water interface (A.sub.aw), the maximal solubility of the VOC (C.sub.wmax) and the water volume (V.sub.w) such that:

[00005]$ln\left(\frac{S_w-C_w\left(t\right)}{S_w}\right)=-\frac{A_{aw}{k}_{d}t}{V_w},$

, where S.sub.w = C.sub.wmax = C.sub.amax/H.sub.c, where C.sub.amax is maximal VOC concentration in the air phase and H.sub.c is Henry’s solubility constant. The VOC dissolution rate constant k.sub.d can hence be estimated from the slope of a ln(S.sub.w-C.sub.w/S.sub.w) versus t plot..

[0129] A designated set up for air sampling will be fitted. A simple air suction system will be installed, comprising a miniature fan and a filter (for dust and airborne small particles). Sampled air will be channeled through tubing into the microfluidic channels to introduce VOC analytes.

[0130] In a preliminary study measurements were performed by individually interrogating each device with an electronic measurement setup. These measurements include I-V and noise characteristics, and time traces. The biochip is stamped with a PDMS microfluidic flow cell placed directly above the devices, which includes inlets and outlets for the introduction of different reagents. A picture of the experimental setup is shown in FIG. 15B.

[0131] A schematic illustration of the chemical modification and biofunctionlaization strategy is presented in FIG. 16: a native OR, embedded in a nanoscale membrane fragment (“nanosomes” is covalently attached (via diazonium defect) to an amine-terminated phospholipid (DPPE) used as an anchor point.

[0132] Disease-causing molds affect the livelihood and food security of millions of people worldwide. With the stakes so high, reducing the impact of these pathogens is paramount. 1-Octen-3-ol or mushroom alcohol is a microbial volatile organic compound (mVOC) released by molds, which is an early warning signals of environmental hazards, neurodegenerative diseases, food contaminations and fruit diseases. It is also a food additive or a flavoring agent that has important uses in process manufacturing. The (R) enantiomer of 1-octen-3-ol is one of the most common and abundant VOC released by fungi such as Penicillium spp. and Aspergillus spp. molds.

[0133] In studies of mosquito OR, the inventors’ have shown that mosquito odorant receptors known as OR8 are activated by (R)-1-octen-3-ol in the parts per billion range (FIG. 17A). Indeed, any modifications of the double bond saturation, side chain length, chiral center position, or functional substitution of the (R)-1-octen-3-ol chemical structure is sufficient to prevent OR8 from being activated (FIG. 17B). At this concentration range, OR8 exhibits a narrow molecular receptive range (FIG. 17C). However, at higher odorant concentrations, OR8 exhibits a broader odorant tuning profile (FIG. 17D), being activated by a wider range of chemical structures.

[0134] These studies established OR8 as a sensitive enantioselective odorant receptor, whose narrow odorant tuning is optimal in the nanomolar range, which is excellent in the field of receptor pharmacology. This receptor is among the most selective reported odorant receptor, whose activation plummets when exposed to any slight modification of the chemical structure of (R)-1-octen-3-ol. OR8 is the epitome of receptor selectivity, being able to reject any slight modification on the (R)-1-octen-3-ol backbone manifesting in complete activation shutdown (FIG. 17B). At a low concentration the receptor is specific to (R)-1-octen-3-ol (FIG. 17C), which is a desirable property for the purpose of using this receptor to detect early signs of microbial contamination or presence. However, OR8 selectivity diminishes at higher odorant concentrations, exhibiting a broader molecular receptive range (FIG. 17D).