Fluorometric Biosniffer Camera “Sniff-Cam” - American Chemical Society

Article pubs.acs.org/ac

Fluorometric Biosniffer Camera “Sniff-Cam” for Direct Imaging of Gaseous Ethanol in Breath and Transdermal Vapor Takahiro Arakawa,† Toshiyuki Sato,‡ Kenta Iitani,‡ Koji Toma,† and Kohji Mitsubayashi*,†,‡ †

Department of Biomedical Devices and Instrumentation, Institute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, 2-3-10 Kanda-Surugadai, Chiyoda-ku, Tokyo 101-0062, Japan ‡ Graduate School of Medical and Dental Sciences, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8549, Japan ABSTRACT: Various volatile organic compounds can be found in human transpiration, breath and body odor. In this paper, a novel two-dimensional fluorometric imaging system, known as a “sniffer-cam” for ethanol vapor released from human breath and palm skin was constructed and validated. This imaging system measures ethanol vapor concentrations as intensities of fluorescence through an enzymatic reaction induced by alcohol dehydrogenase (ADH). The imaging system consisted of multiple ultraviolet light emitting diode (UV-LED) excitation sheet, an ADH enzyme immobilized mesh substrate and a high-sensitive CCD camera. This imaging system uses ADH for recognition of ethanol vapor. It measures ethanol vapor by measuring fluorescence of nicotinamide adenine dinucleotide (NADH), which is produced by an enzymatic reaction on the mesh. This NADH fluorometric imaging system achieved the two-dimensional real-time imaging of ethanol vapor distribution (0.5−200 ppm). The system showed rapid and accurate responses and a visible measurement, which could lead to an analysis of metabolism function at real time in the near future. obel Prize winner Linus Pauling first proposed the concept of diagnosing patients using simultaneous analysis of body metabolites.1 Progress in the analysis of volatile organic compounds, such as those found in the urine and breath, creates an increasing potential for the application of this concept to the medical field.1−6 In recent years, progress in analytical devices has enabled measurement of minute amounts of odorous material and volatile chemical components.2 Bodyderived gases (body gases) in particular, such as gases from halitosis and body odor, contain compounds produced by metabolic processes and ailment-specific components. Highsensitivity analysis or measurement of these volatile biological components is expected to simplify metabolic capacity evaluation, medical diagnostics, and disease screening.7−9 For example, some researches in the application of surfaceenhanced Raman spectroscopy (SERS) to noninvasive biosensing with a focus on in medical diagnostics were reported.10,11 In recent years, examination methods using urine and breath have been put into practice, and research is also being conducted into health monitoring using gaseous compounds emitted with sweat from the skin (transdermal vapor).12−18 However, body-derived volatile chemical compounds see large temporal and spatial variation in concentration, making it difficult to accurately and selectively evaluate the behavior of these spatiotemporal distributions with analytical devices that use sampling like gas chromatography, or existing measurement devices that have weak sensitivity like semiconductor gas

N

© XXXX American Chemical Society

sensors.2 In contrast, gas measurement methods that use biocatalytic enzymes as sensors measure gaseous compounds selectively and with high sensitivity using luminescence and fluorescence.19,20 Furthermore, researchers have devised a visualization system that uses biological luminescence and applied it to the visual measurements and metabolic capacity evaluation of postalcohol administration expired ethanol.21,22 This system has also been applied to the visualization of transdermal vapor emitted from the palm23 but with issues in sensitivity and reactivity because its use of biological luminescence requires two types of enzyme reactions. However, if fluorescence were used instead, gaseous compounds could be induced to fluoresce with just one type of enzyme, with the expectation of increased reactivity and sensitivity. In the present study, we adopted ethanol vapor as the compound of interest and constructed a visualization system that detects the intrinsic fluorescence of coenzymes generated from one type of enzyme reaction and images the distribution of the concentration of ethanol vapor as well as its temporal changes. In addition, we conducted visualization measurements of a body gas, sampling postalcohol consumption breath and transdermal vapor from the human palm. Received: November 25, 2016 Accepted: February 28, 2017

A

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 1. (a) Ethanol vapor visualization measurement system using a biofluorescence method. In the optical system used for NADH fluorescence detection, the designed UV-LED sheet was positioned opposite a high-sensitivity camera (HEED-HARP camera). An excitation source bandpass filter with a central wavelength of 340 nm was set in front of the LED sheet, and a bandpass filter with a central wavelength of 490 nm was set on the imaging side of the camera. (b) Postalcohol-consumption skin-derived ethanol vapor measurement system. To keep the gap between the palm and the enzyme immobilization mesh constant, a transdermal vapor loading plate was created using a 3 mm-thick acrylic board. In line with the expansion of the measurement area, the enzyme immobilization mesh was set at 90 mm × 90 mm. (c) Schematic of ADH enzyme mesh UV-LED sheet (9 × 9 LEDs) for NADH fluorescence.

■

EXPERIMENTAL SECTION Constructing an Ethanol Vapor Fluorescence Visualization Measurement System. The principle of the ethanol vapor visualization measurement system based on the fluorescence of coenzyme nicotinamide adenine dinucleotide (NADH) is shown as follows.

reaction when the gas is loaded. On the basis of this principle, we developed a visualization system that uses a biofluorescence method for human breath and skin gas (Figure 1a,b). First, we constructed an NADH excitation sheet and an optical system for observing fluorescence. Conducting visualization measurements of body-derived volatile components requires excitation over a large area. Thus, we designed specifications for a UV-LED sheet that allowed large-area excitation whereby UV-LEDs with superior wavelength properties were positioned on a plane. Because breath and transdermal vapor emitted from the palm were used as the biosamples for the experiment, the UV-LED sheet needed to allow excitation of the palm. Given that the width of a human palm is approximately 80 mm × 80 mm, we set the area of the sheet at 100 mm × 100 mm and positioned 81 LED elements

ADH

ethanol + NAD+ ⎯⎯⎯⎯→ acetaldehyde + NADH + H+

Ethanol is oxidized to acetaldehyde by the catalysis of alcohol dehydrogenase (ADH), whereupon it generates reduced NADH with the oxidized NAD (NAD+) as an electron acceptor. Because of the fluorescent properties of NADH (excitation 340 nm, fluorescence 490 nm), combining an excitation light source with a high-sensitivity camera allows us to detect and visualize NADH generated by the enzyme B

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

the above-mentioned carriers, drying for 1 h in a 4 °C refrigerator, dripping 32 μL of GA liquid adjusted to 2.5 vol % in PBS into them, and then drying them for 1 h in the 4 °C refrigerator. Next, the enzyme meshes were submersed in NAD+ liquid adjusted to 10 mmol/L in PBS, 105 μL of ethanol (99.5%) was dripped in after the meshes were gently agitated for a fixed period of time to allow them to react, absorbance at a wavelength of 300−400 nm was measured using a spectrophotometer (V-530, JASCO), and the degree of enzyme immobilization toward each carrier was compared. Property Evaluation and Image Analysis of the Visualization Measurement System. Figure 1a gives a schematic of the ethanol vapor fluorescence visualization measurement system. Inside a dark box, we placed a UVLED excitation sheet, a gas atomization nozzle, enzyme mesh, and a high-sensitivity camera, in that order. The enzyme immobilization mesh was washed with PBS and dehydrated, moistened with 300 μL of 5 mmol/L NAD+ liquid, and placed in the dark box. Two types of ADH were used: the abovementioned ADH from yeast as well as ADH from Saccharomyces cerevisiae (369 units/mg solid, 382 mg/protein, Sigma-Aldrich). Ethanol vapor at a specified concentration (100 mL/min) was loaded from the back of the mesh using a standard gas generator (permeator, PD-1B-2, Gastec). The fluorescence generated by the enzyme reaction was consecutively photographed with a high-sensitivity camera. The moving images obtained were processed using multipurpose image analysis software, and fluorescence intensity was calculated. In order to capture changes in fluorescence, differential analysis was conducted on the visualized moving image based on eq 1 using the image analysis software.

in the center in an array of 9 rows and 9 columns, 9 mm apart (see Figure 1c). Because measuring breath does not require as much surface area as measuring transdermal vapor, for this we designed the LED in 9 segments, such that each segment could be used independently. In order to apply a stable current to the whole LED sheet, we used a stabilized dc power supply. To evaluate the performance of the UV-LED sheet, which we had produced externally based on our design specifications (λ = 340 nm, 9 × 9 array, 340X081SFN, Dowa), we used optic fibers (CustomPatch-287254, Ocean Optics) and a spectroscope (USB4000, Ocean Optics) to measure the spectra and study their properties. In the optical system used for NADH fluorescence detection, the designed UV-LED sheet was positioned opposite a highsensitivity camera (HEED-HARP camera, Pioneer) (Figure 1a). This high-sensitive camera consists of high-efficiency electron emission device (HEED) and high-gain avalanche rushing amorphous photoconductor (HARP). An excitation source bandpass filter (BPF) with a central wavelength of 340 nm (λ = 340 ± 42.5 nm, Edmond Optics) was set in front of the LED sheet, and a BPF with a central wavelength of 490 nm (λ = 490 ± 10 nm, Asahi Spectra) was set on the imaging side of the camera. Then, phosphate buffered saline (PBS, pH 8.0, 0.1 mol/L) containing NADH (Oriental Yeast) was dispensed into a cuvette (quartz, Tosoh Quartz), and fluorescence images of NADH were taken using the constructed optical system. The images obtained were processed using multipurpose image analysis software (Cosmos32, Library). Next, we created an enzyme immobilization mesh to be the ethanol vapor recognition element. In the creation of the enzyme mesh, ADH is immobilized using a mesh carrier as the immobilization material, but intrinsic fluorescence contained in these materials acts as noise during NADH measurements. We wished to select materials and methods that would produce less intrinsic fluorescence and be more suitable for enzyme immobilization. First, we considered the following five possible immobilization carriers: polytetrafluoroethylene (PTFE, W.L. Gore & Associates), hydrophilized polytetrafluoroethylene (HPTFE, W.L. Gore & Associates), polylactic acid (PLA fiber, TORAY), cotton (gauze, Ohki), and rayon (Bemcot, Asahi Kasei). PLA fibers were processed into a mesh of 30 mm × 30 mm using 3 warp and 3 woof threads, and all other carriers were cut into 30 mm × 30 mm samples, measured using the above-described fluorescence optical system and the results compared. We also investigated the intrinsic fluorescence of enzyme immobilization materials. Three types of immobilization materials were evaluated: ultraviolet cross-link resin (PVASbQ, Biosurfine SPH, Toyo Gosei), a copolymer (PMEH) of 2methacryloyloxyethyl phosphorylcholine (MPC) and methacrylic acid 2-ethylhexyl (EHMA),24 and glutaraldehyde (GA, 25% solution, Wako Junyaku). First, 240 units of ADH (from yeast, 128 units/mg powder, Oriental Yeast Co.) were immobilized on the cotton carrier (20 mm × 20 mm) using PMEH and GA, moistened using NAD+ liquid prepared with PBS, and intrinsic fluorescence was investigated using the fluorescence optical system. Finally, we created enzyme meshes by GA cross-linkage using each carrier and compared their enzyme immobilization abilities. In the experiment, the enzyme meshes were created by dripping a liquid mixture of 60 units of ADH, 1.5 mg of bovine serum albumin (BSA, Wako), and 300 μL of PBS onto

g (t + Δt ) − g (t ) dl = dt Δt

(1)

I, fluorescence intensity (a.u.), t, time (s), and g(t), time point of the visualized image We also investigated the selectivity of the system for ethanol vapor. In this experiment, we loaded onto the enzyme mesh a standard gas with defined concentrations of representative components contained in breath, chosen based on their concentrations in the breath of healthy people, and measured the system output. Visualization Measurement of Body Gas Using a Biofluorescence Method. We loaded postalcohol-consumption breath and transdermal vapor onto the system and conducted body gas visualizations (Tokyo Medical and Dental University, Authorization No. 2012-6). For the breath experiment, healthy adult males, who consented to having their breath collected, took by mouth an alcoholic drink of 25% concentration over 15 min, such as to consume 0.4 g of alcohol for every 1 kg of their body weight. Their breaths were then loaded directly onto the system through an expiratory flow control unit,22 at standard intervals (every 15 min for the first hour and every 30 min after that) for 3 h after the alcohol administration, and the fluorescence generated by the enzyme reaction was photographed with the high-sensitivity camera. We also conducted imaging of postalcohol-consumption skin-derived ethanol vapor. We constructed a device for visualizing ethanol vapor emitted from the palm (Figure 1b). In order to keep the gap between the palm and the enzyme immobilization mesh constant, a transdermal vapor loading plate was created using a 3 mm-thick acrylic board. In line with the expansion of the measurement area, the enzyme C

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry immobilization mesh was set at 90 mm × 90 mm. The enzyme mesh was attached to the plate, and the enzyme mesh was positioned 4 mm from the UV-LED sheet. The palms of the healthy adult males who had consumed alcohol (0.4 g per 1 kg of body weight) were pressed against the plate from the camera side for 20 s, and fluorescence accompanying the generation of transdermal vapor was photographed with the high-sensitivity camera.

and investigated the enzyme immobilization of the remaining four carriers. Next, we investigated the intrinsic fluorescence of the three types of materials we prepared for enzyme immobilization. Figure 3a shows the results of a comparison of the fluorescence

■

RESULTS AND DISCUSSION Evaluation and Selection of Optical Elements and Parts for the Visualization System. A 60 mA current was applied to each 9-LED segment of the UV-LED sheet: accordingly, a 540 mA current was applied when the nine segments comprising the maximum irradiation area (81 LEDs) were used. Figure 2 shows the ultraviolet intensity of the

Figure 2. Power spectrum of the surface of the UV-LED sheet. The ultraviolet intensity of the excitation light source of wavelength 340 nm at the surface of the UV-LED sheet, as evaluated with a spectroscope at 10 mm intervals.

excitation light source of wavelength 340 nm at the surface of the UV-LED sheet, as evaluated with a spectroscope at 10 mm intervals. Although the excitation light intensity at the edge of the UV-LED sheet was as low as ∼60% that at the center, the central area (60 mm × 60 mm) displayed intensity differences of ±10%, meaning uniform irradiation of excitation light and uniform excitation of the enzyme mesh was possible. When fluorescence of the NADH standard liquid in the cuvette was measured using the stand-alone UV-LEDs and the UV-LED sheet, the fluorescence intensity of the stand-alone UV-LEDs correlated with NADH concentration over the range of 1.0−100 μmol/L, while that of the UV-LED sheet correlated with it over the range of 0.1−320 μmol/L. That is, the UV-LED sheet enabled us to obtain fluorescence intensities 4.6-fold (100 μmol/L) higher than those obtained using a stand-alone UVLED. The fact that having the UV-LED in sheet form increased the light intensity indicates its potential for the high-sensitivity detection of NADH. Next, we investigated the intrinsic fluorescence of the five types of carriers we prepared as enzyme immobilizers. PTFE, H-PTFE, and polylactic acid displayed low intrinsic fluorescence. Naturally derived fibers contain large amounts of conjugated system compounds like lignin, whose amounts differ depending on habitat and cultivation: accordingly, cotton and the semisynthetic fiber rayon displayed more intrinsic fluorescence than the synthetic fibers. Because rayon in particular had a considerably higher intrinsic fluorescence compared with the other four candidates, we excluded rayon

Figure 3. (a) Comparison of intrinsic fluorescence images for each enzyme immobilization material of glutaraldehyde, PVA-SbQ, and PMEH polymer. (b) Comparison of degree of enzyme immobilization by GA cross-linkage for each substrate (PTFE, H-PTFE, polylactic acid, and cotton).

intensities for each of the materials when attached to the cotton mesh. As can be seen in the figure, high intrinsic fluorescence was observed for PVA-SbQ and PMEH. The intrinsic fluorescence of PVA-SbQ is thought to be caused by photofunctional sites in its structure, making it unsuitable for fluorescence measurement. Further, as the synthesis process of D

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry PMEH uses an ethanol solvent, NADH is generated when the enzyme immobilizer is used. This generation of fluorescence makes PMEH particularly unsuitable for visualization of ethanol vapor. GA, however, has no intrinsic fluorescence as its molecular structure is simple, and the GA mesh showed low fluorescence intensity compared with the other immobilization materials (Figure 3b). Thus, we chose a GA-based cross-linkage method for the enzyme immobilization. On the basis of these results, we compared ADH immobilized using GA cross-linkage for the four abovementioned carriers (excluding rayon). The right panel of Figure 3b gives the results of the measurement of the amount of NADH generated by the ADH enzyme reaction for the four types of GA cross-linkage enzyme mesh and the degree of enzyme immobilization. Cotton mesh had the highest fluorescent output and best degree of enzyme immobilization. Therefore, for the subsequent experiments, we decided to create the enzyme mesh using cotton mesh as the carrier and using enzymes immobilized using GA cross-linkage. Evaluating the Properties of the Visualization System Using a Biofluorescence Method. Standard ethanol vapor samples of various concentrations were loaded into the visualization system, which contained ADH derived from Saccharomyces cerevisiae. Fluorescence was observed centering on the point where the ethanol vapor was loaded, and there was confirmation of a gentle stabilization of fluorescence intensity after sample loading (see Figure 4, upper images). The intensity information on 256 gradation was detected for each of all pixels of a fluorescence image. The moving images obtained were analyzed, and the fluorescence intensity measured, giving the change over time. However, because the NADH generated through the enzyme reaction remained on the mesh carrier and the fluorescence intensity did not drop, it was difficult to show

increases or decreases in the distribution of the ethanol vapor load. For this reason, we decided to use differential analysis to find the change in fluorescence intensity for each unit of time, and display this once more, this time as a moving image. First, we considered at which values of Δt in the differential analysis we could obtain the maximum value for the amount of change. Increasing Δt allows the amount of change to be shown as a waveform without varying the derivative. However, as the peak value in the differential analysis of the change decreases with increasing Δt, we decided to smooth and differentiate the change over time in fluorescence intensity using a running average. By setting the running averaged Δt to 5 s, we were able to calculate the derivative with a peak. The peak curve obtained showed increases with loading and decreases when loading stopped. It also showed better reactivity compared with visualization using luminescence with two types of enzymes (90% reactivity at 50 ppm: 35 s (chemiluminescence) → 20 s). The inset images shown are color images at several time points (20, 40, 80 s), taken from the differentiated moving images created with this method (see Figure 4, lower images). These images show that we can determine the spatial distribution of ethanol vapor by conducting a differential analysis of the NADH fluorescence images obtained with this system. We next investigated the system’s quantitative measurements of ethanol concentration at the steady value of fluorescence intensity and at the peak value of the differential analysis. Figure 5 shows

Figure 5. Comparison of calibration formulas based on fluorescence intensity (●) and differential analysis results (red □) against ethanol vapor in the system. The quantitative range was 0.5−150 ppm (from Saccharomyces cerevisiae), and similar quantitative performance (1−150 ppm) was obtained for the peak value from differential analysis.

these two calibration curves. As can be seen from this calibration curve, with fluorescence intensity it was possible to detect ethanol vapor at low concentrations, and the quantitative range was 0.5−150 ppm (from Saccharomyces cerevisiae). Similar quantitative performance (1−150 ppm) was obtained for the peak value from differential analysis, despite detection sensitivity deteriorating at low ethanol concentrations. The two calibration formulas obtained were as follows: peak of Δ avg intensity = 1.73[EtOH(ppm)]0.709 Figure 4. Change over time of fluorescence intensity in the visualization measurements of ethanol vapor (50 ppm) (above) and results of differential analysis after creating a running average from them (below). Inset images: fluorescence images and differential analysis images (fluorescence intensity magnified 2.3-fold and 25-fold, respectively).

(2)

peak slope of Δ avg intensity = 0.0064[EtOH(ppm)]0.633 (3)

We investigated the selectivity of the system using representative components of breath. Figure 6 shows a comparison of the fluorescence output for various components E

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 6. Comparison of NADH generation by several breath components for ADH enzymes derived from yeast and Saccharomyces cerevisiae and effect of acetaldehyde on ethanol catalysts.

Figure 7. Change over time in ethanol concentration calculated based on results of visualization of the breath of each subject after alcohol administration and results of using a detector tube (a, ALDH2(+); b, (−)).

Figure 8. Image of skin-derived ethanol vapor of a subject after alcohol administration (fluorescence intensity magnified 25 times) (a, after 15 min; b, after 45 min).

Fluorescence Visualization Measurement of Human Gas. On the basis of the experimental results above, we applied the visualization system using ADH from Saccharomyces cerevisiae to the fluorescence visualization measurement of breath and transdermal vapor. We conducted the experiment using healthy subjects to whom the purpose of the experiments had been explained and who had already given informed consent. Figure 7a,b shows the change over time in the ethanol concentration calculated from the results obtained by inputting the postalcohol-consumption breath of ALDH2(+) and (−) experimental subjects into the created visualization system. We have included the results obtained using a detector tube for comparative purposes. As seen in this figure, the system enables the visualization measurement of ethanol vapor contained in breath, and an increase in concentration was observed with a peak at 30 min after alcohol administration. Exhaled breath

of two types of ADH (left, yeast; right, Saccharomyces cerevisiae). According to this result, we can see that neither of the enzymes react to breath components other than ethanol, and high selectivity based on the substrate specificity of the enzymes was obtained. As postalcohol consumption breath contains higherthan-normal concentrations of acetaldehyde, we combined acetaldehyde of a concentration 12 times higher than normal with ethanol vapor and investigated the effect of this on fluorescence measurement. The result was that the output of the yeast-derived enzyme fell to below 50%. Both enzymes had similar levels of quantitative accuracy on measuring ethanol vapor. Accordingly, in measurements of postalcohol consumption body gas, we decided to use the enzyme affected less by acetaldehyde, ADH from Saccharomyces cerevisiae, for visualization. F

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

(9) Amann, A.; Mochalski, P.; Ruzsanyi, V.; Broza, Y. Y.; Haick, H. J. Breath Res. 2014, 8, 016003. (10) Henry, A.-I.; Sharma, B.; Fernanda Cardinal, M.; Kurouski, D.; Van Duyne, R. P. Anal. Chem. 2016, 88, 6638−6647. (11) Bocklitz, T. W.; Guo, S.; Ryabchykov, O.; Vogler, N.; Popp, J. Anal. Chem. 2016, 88, 133−151. (12) Kamei, T.; Tsuda, T.; Mibu, Y.; Kitagawa, S.; Wada, H.; Naitoh, K.; Nakashima, K. Anal. Chim. Acta 1998, 365, 259−266. (13) Hawthorne, J. S.; Wojcik, M. H. J. - Can. Soc. Forensic Sci. 2006, 39 (2), 65−71. (14) Dumett, M. A.; Rosen, I. G.; Sabat, J.; Shaman, A.; Tempelman, L.; Wang, C.; Swift, R. M. Applied Mathematics and Computation 2008, 196, 724−743. (15) Diamond, D.; Coyle, S.; Scarmagnani, S.; Hayes, J. Chem. Rev. 2008, 108 (2), 652−679. (16) Dormont, L.; Bessière, J.-M.; Cohuet, A. J. Chem. Ecol. 2013, 39, 569−578. (17) Mochalski, P.; Unterkofler, K.; Hinterhuber, H.; Amann, A. Anal. Chem. 2014, 86, 3915−3923. (18) Yamada, Y.; Hiyama, S.i; Toyooka, T.; Takeuchi, S.; Itabashi, K.; Okubo, T.; Tabata, H. Anal. Chem. 2015, 87, 7588−7594. (19) Kudo, H.; Sawai, M.; Suzuki, Y.; Wang, X.; Gessei, T.; Takahashi, D.; Arakawa, T.; Mitsubayashi, K. Sens. Actuators, B 2010, 147, 676−680. (20) Ye, M.; Chien, P.-J.; Toma, K.; Arakawa, T.; Mitsubayashi, K. Biosens. Bioelectron. 2015, 73, 208−213. (21) Wang, X.; Ando, E.; Takahashi, D.; Arakawa, T.; Kudo, H.; Saito, H.; Mitsubayashi, K. Talanta 2010, 82, 892−898. (22) Arakawa, T.; Wang, X.; Kajiro, T.; Miyajima, K.; Takeuchi, S.; Kudo, H.; Yano, K.; Mitsubayashi, K. Sens. Actuators, B 2013, 186, 27− 33. (23) Arakawa, T.; Kita, K.; Wang, X.; Miyajima, K.; Toma, K.; Mitsubayashi, K. Biosens. Bioelectron. 2015, 67, 570−575. (24) Kudo, H.; Yagi, T.; Chu, M. X.; Saito, H.; Morimoto, N.; Iwasaki, Y.; Akiyoshi, K.; Mitsubayashi, K. Anal. Bioanal. Chem. 2008, 391 (4), 1269−1274.

ethanol concentration rapidly increased after oral administration and the peaks appeared at 30 min and then gradually decreased until the end of sample collection at 180 min. Consistent with previous reports,22 we confirmed higher concentration in ALDH2(−) than in (+). Figure 8 shows visualization images of postalcoholconsumption transdermal vapor. As can be seen from this figure, we were able to visualize the ethanol vapor contained in the transdermal vapor emitted from the palm skin and confirmed the spreading of the ethanol emissions across the palm skin. Using this visualization system based on biofluorescence, we were able to confirm the emission of ethanol vapor 15 min after alcohol administration. We believe that as it offers improved sensitivity and reactivity compared with traditional chemiluminescence, this system can be used for detailed evaluation of the generation of body gases.

■

CONCLUSIONS We developed a two-dimensional fluorometric ethanol imaging system “sniffer camera” using an enzymatic reaction induced by alcohol dehydrogenase. This system measures ethanol concentrations as intensities of fluorescent image by enzyme reaction. The imaging of gaseous ethanol was achieved at 0.5 ppm to detection limit. We applied a high sensitivity gaseous ethanol imaging system for measurement of skin ethanol emission from a human palm. In the near future, we will apply this noninvasive imaging system to visualize a disease screening. The imaging system based on biosensing technologies has a good sensitivity and highly selective for monitoring human gas. Because of these advantages, the proposed method is the potential tool for noninvasive medical care and point-of-care-testing.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +81-3-5280-8091. Fax: +81-3-5280-8094. E-mail: m. [email protected]. ORCID

Kohji Mitsubayashi: 0000-0002-0709-4957 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is partly supported by Japan Society for the Promotion of Science (JSPS) KAKENHI Grant-in-Aid for Scientific Research (B) Grant Number 15H04013 by Ministry of Education, Culture, Sports, Science and Technology (MEXT).

■

REFERENCES

(1) Pauling, L.; Robinson, A. B.; Teranishi, R.; Cary, P. Proc. Natl. Acad. Sci. U. S. A. 1971, 68 (10), 2374−2376. (2) Rattray, N. J. W.; Hamrang, Z.; Trivedi, D. K.; Goodacre, R.; Fowler, S. J. Trends Biotechnol. 2014, 32 (10), 538−548. (3) Ohira, S.; Toda, K. Anal. Chim. Acta 2008, 619, 143−156. (4) Miekisch, W.; Schubert, J. K.; Noeldge-Schomburg, G. F. E. Clin. Chim. Acta 2004, 347, 25−39. (5) DI Natale, C.; Paolesse, R.; Martinelli, E.; Capuano, R. Anal. Chim. Acta 2014, 824, 1−17. (6) Adiguzel, Y.; Kulah, H. Biosens. Bioelectron. 2015, 65, 121−138. (7) Kippenberger, S.; Havlíček, J.; Bernd, A.; Thaçi, D.; Kaufmann, R.; Meissner, M. Exp. Dermatol. 2012, 21, 655−659. (8) Shirasu, M.; Touhara, K. J. Biochem. 2011, 150 (3), 257−266. G

DOI: 10.1021/acs.analchem.6b04676 Anal. Chem. XXXX, XXX, XXX−XXX