In our previously reported LPAD for histidine, the enzymatic reaction was needed to be performed outside the LPAD: the reaction mixture with HisRS, histidine, ATP, and MgCl2 in a microtube was heated using an aluminum heating block at 80 °C for 30 min and cooled on ice for 5 min. Subsequently, the reaction mixture was loaded onto the LPAD. The factitious step with heating of the reaction mixture and pipetting of the sample were necessary19. In this study, an LPAD that works consecutively for both enzymatic and colorimetric reactions in one step.
Evaluation of sizes filtration papers for microfluidics
The enzymatic reaction occurs when the enzymatic reaction mixture penetrates the detection area; consequently, the enzymatic reaction time is important and will affect the LPAD response. The type and shape of the filtration papers used for the microfluidics were evaluated using Advantec Grade No. 1, Advantec Grade No. 5B, and MN616G. Table 1 shows the sets of lengths between the enzymatic reaction and detection areas and the width of the microfluidic paths. A length between the enzymatic reaction and detection areas of 15 mm and width of the microfluidic paths of 3.0 mm was determined to be the preferred size for the LPAD. A short length and/ or narrow width path showed no or only a scarce response as the reaction mixture reached the detection area instantly. If the length was longer (20 mm), the loaded enzymatic reaction mixture could not reach the detection area. Therefore, the provision of sufficient aaRS enzymatic reaction times during penetration of the reaction mixture into the filtration paper is important to consider in designing LPADs. The density of the filtration paper fiber is also an important factor. The high-density filtration paper Advantec Grade No. 1 showed better performance in the point of the density of fiber because the reaction mixture penetrated gradually and this was sufficient for the enzymatic reaction times. The high-density filtration papers also sufficiently retained the enzyme and reagent solutions to enable the enzymatic reaction in the detection area, indicating that the stable fabrication of LPADs could be possible. Moreover, at the point of the colorimetric detection which was performed based on the molybdenum blue reaction, the depth of the color changed in a time dependent manner and became saturated. Evaluation of the color of the detection areas at 15 min after the deposition of the samples was preferred (data not shown).
Assay of the LPAD
Photos obtained after assaying the LPAD using 0–100 μM for each amino acid are shown in Fig. 3a. The color of the LPAD detection area for glycine (upper right corner of the LPAD) after loading of glycine changed from yellow to blue, whereas the detection areas for tryptophan (upper left corner), histidine (lower left corner) and lysine (lower right corner) after loading of glycine displayed no change in color. In the same manner, the color of the LPAD detection area when only the tryptophan, histidine, or lysine were loaded respectively, changed from yellow to blue, and no reactions were observed for the discordant amino acids.
Photos of the laminated paper-based analytical devices (LPADs) after the loading of each amino acid. (a) Original image of each LPAD after interaction with 0–100 μM tryptophan, glycine, histidine, and lysine. Color change was observed only in the detection area of substrate amino acid. (b) The images were color-inverted using the GNU Image Manipulation Program.
The inverted images obtained using the GNU Image Manipulation Program are shown in Fig. 3b.
Figure 4 shows the calibration curves for tryptophan, glycine, histidine and lysine detection (filled circle in each graph). The horizontal axis represents the initial concentration of each amino acid, and the vertical axis represents the integration signal (arbitrary unit), which is calculated as the product of the brightness and detection area. The integration signal increased in response to the substrate amino acid addition, and good linearity ranges between 3.6 and 100 μM were obtained for tryptophan, with a detection limit of 1.1 μM (r = 0.9717, Fig. 4a), 10.1–100 μM for glycine, with a detection limit of 3.3 μM (r = 0.9722, Fig. 4b), 5.9–100 μM for histidine, with a detection limit of 1.9 μM (r = 0.9816, Fig. 4c), and 5.6–100 μM for lysine, with a detection limit of 1.8 μM (r = 0.9756, Fig. 4d).
Calibration curves for tryptophan, glycine, histidine, and lysine sensing. The filled circle in each graph represents the substrate amino acid, whereas the open circles indicate the average of the integration signals of three non-substrate amino acids. Data represent the average of three measurements, and the error bars indicate standard deviations.
The limit of detection (LOD) of the conventional HPLC (Hitachi Amino Acid Analyzer L-8900) is approximately 0.5 μM6, and slightly superior to the LOD of our LPAD. However, the measurable concentrations of each amino acid of the LPADs were within the approximate range of the amino acid levels found in the blood.
Figure 4 also shows the selectivity of the LPAD. The open circles in each graph represent the average of the integration signal of three non-substrate amino acids; the open circle in Fig. 4a (tryptophan detection area) indicates the average of the integration signal of histidine, lysine, and glycine. Each calibration curve was non-leaning, and the values were almost the same as those for 0 μM substrate amino acid; therefore, no response was observed for non-substrate amino acids. Owing to the substrate specificities of TrpRS, GlyRS, HisRS, and LysRS, these enzymes specifically bind to their corresponding substrate amino acids. Hence, the LPAD could selectively analyze the amino acids. In our previous paper, no interference was observed in the binding of the substrate amino acid to aaRS. The binding activity of aaRS to the solo substrate amino acid and the 20 amino acid mixture was almost same value; therefore, the existence of another 19 amino acids in the reaction mixture would not interfere the binding of the substrate amino acid to aaRS14.
Validation of the LPAD
The reproducibility of LPAD responses to 100 μM of each amino acid among three different fabrication (3 days) and assay dates were evaluated (Table 2). Each entry was repeated three times. The coefficient of variation [CV (%)] was approximately less than 2%, and the CV values were low. These findings suggest that fabrication of the LPADs, including the cutting of the filtration papers and films, as well as the coating of the reagents, can be reproduced precisely and consistently. The LPADs showed sufficient reproducibility for each amino acid. Furthermore, as described above, they required only several micromoles of each amino acid to function, and this is consistent with the levels of amino acids in the blood.
