J S S
ISSN 1615-9306 · JSSCCJ 38 (12) 2007–2192 (2015) · Vol. 38 · No. 12 · June 2015 · D 10609
JOURNAL OF
SEPARATION SCIENCE
Methods Chromatography · Electroseparation Applications Biomedicine · Foods · Environment
12 15
www.jss-journal.com
2053
J. Sep. Sci. 2015, 38, 2053–2058
Yihang Zeng Wensheng Cai Xueguang Shao Research Center for Analytical Sciences, College of Chemistry, Nankai University, Tianjin Key Laboratory of Biosensing and Molecular Recognition, State Key Laboratory of Medicinal Chemical Biology, Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin, China Received January 22, 2015 Revised March 27, 2015 Accepted March 30, 2015
Research Article
Quantitative analysis of 17 amino acids in tobacco leaves using an amino acid analyzer and chemometric resolution A method was developed for quantifying 17 amino acids in tobacco leaves by using an A300 amino acid analyzer and chemometric resolution. In the method, amino acids were eluted by the buffer solution on an ion-exchange column. After reacting with ninhydrin, the derivatives of amino acids were detected by ultraviolet detection. Most amino acids are separated by the elution program. However, five peaks of the derivatives are still overlapping. A non-negative immune algorithm was employed to extract the profiles of the derivatives from the overlapping signals, and then peak areas were adopted for quantitative analysis of the amino acids. The method was validated by the determination of amino acids in tobacco leaves. The relative standard deviations (n = 5) are all less than 2.54% and the recoveries of the spiked samples are in a range of 94.62–108.21%. The feasibility of the method was proved by analyzing the 17 amino acids in 30 tobacco leaf samples. Keywords: Amino acids / Chemometrics / Non-negative immune algorithm / Tobacco leaves DOI 10.1002/jssc.201500090
1 Introduction In recent years, there has been an increasing interest for accurate quantitative analysis of amino acids in biological research [1]. Several methods can be used to analyze amino acids, such as HPLC [2, 3], GC–MS [4], and CE [5]. The most commonly used method is postcolumn derivatization with ninhydrin using LC [6, 7], which has shown good reliability. Based on the method, an automatic amino acid analyzer (AAA) was developed and widely used to quantify amino acids automatically [8, 9]. It uses the classic method of postcolumn derivatization with ninhydrin [10], known as “the Gold Standard” [11], to analyze amino acids. The AAA system realizes automation by combining autosampler with a separation column unit and reactor unit [12]. It makes the analysis of amino acids automatic, fast, and simple [13]. However, AAA system still has the problem of peak overlap when analyzing multicomponent samples, such as tobacco leaf and ginseng [14,15]. Therefore, more efficient methods are necessary to be developed.
Correspondence: Professor Xueguang Shao, Research Center for Analytical Sciences, College of Chemistry, Nankai University, Tianjin 300071, P. R. China Fax: +86-22-23502458 E-mail: [email protected]
Abbreviations: AAA, automatic amino acid analyzer; Ala, alanine; Asn, asparagine; Asp, aspartic acid; CFA, chemical factor analysis; EFA, evolving factor analysis; Glu, glutamic acid; Gly, glycine; HELP, heuristic evolving latent projections; His, histidine; IA, immune algorithm; Mehis, methylhistidine; NNIA, nonnegative immune algorithm; Pro, proline
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chemometrics provides new opportunities for the resolution of overlapping chromatographic signals [16–18]. To analyze overlapping 1D signals, curve fitting techniques, convolution, and deconvolution methods, and derivation calculation have been developed [19, 20]. For overlapping 2D signals, a variety of chemometric methods such as evolving factor analysis (EFA) [21], subwindow factor analysis [22] and HELP [23], etc. based on chemical factor analysis (CFA) have been developed. CFA is a branch of multivariate analysis and shows obvious advantage when it comes to multivariate problems [24, 25]. Furthermore, alternative methods like multivariate curve resolution-alternating least squares [26], based on bilinear models, are developed and employed in the analysis of environmental and biological samples [27, 28]. In our previous studies, an immune algorithm (IA) has been developed for resolving overlapping signals in HPLC, spectroscopy, and GC–MS analyses [29]. IA extracts information of each component from the total signal by projection and subtraction with the help of the provided standard signals of all the possible components [30]. In the primary IA, therefore, the information of all the components possibly contained in the analyzing sample must be provided [31]. For retrieving the information of a specific component from an overlapping signal, non-negative immune algorithm (NNIA) was proposed and had been proved to be an efficient way for determining an interested component in a complex sample [32]. In this work, a method for quantifying amino acids in tobacco leaves by AAA with chemometric resolution was developed. Most amino acids are separated on the ion-exchange column of AAA system. However, there are five peaks of amino acid derivatives still overlapping. NNIA is used to resolve the overlapping signals of amino acids. With the www.jss-journal.com
2054
Y. Zeng et al.
J. Sep. Sci. 2015, 38, 2053–2058
Figure 1. Chromatograms of mixed standard solution and tobacco sample.
resolved signals, the amino acids can be quantitatively determined with an acceptable precision and accuracy. Then the feasibility of the method is proved by analyzing the 17 amino acids in 30 tobacco samples.
2 Materials and methods 2.1 Materials and reagents Hydrochloric acid (HPLC grade) and the mixed standard solution of amino acids (100 nmol/L) were purchased from Sinopharm Chemical Reagent (Beijing, China). The single standard solutions (100 nmol/L), including aspartic acid (Asp), threonine (Thr), serine (Ser), asparagine (Asn), glutamic acid (Glu), glycine (Gly), alanine (Ala), valine (Val), isoleucine (Ile), leucine (Leu), tyrosine (Tyr), phenylalanine (Phe), ␥-amino butyric acid (g-ABA), histidine (His), lysine (Lys), arginine (Arg), and proline (Pro), were also obtained from Sinopharm Chemical Reagent (Beijing, China). Lithium buffer system and ninhydrin solution were purchased from membraPure (Bodenheim, Germany). Thirty burley tobacco samples selected from three different regions (R1, R2, R3) were provided by Beijing Cigarette Factory (Beijing, China). To investigate the content of amino acids in different sampling position of tobacco leaves, 15 samples were selected from upper (U) parts of leaves and the other 15 samples were collected from the middle (M) parts.
centrifuged at 15 000 rpm for 15 min, then filtered with 0.22 m membrane, and stored at 4⬚C for AAA analysis. 2.3 Chromatographic analysis The analysis of amino acids was carried out by using A300 amino acid analyzer (membraPure Bodenheim, Germany). The ion-exchange chromatographic column (60–0041, membraPure Bodenheim, Germany) was adopted, where the amino acids were eluted by lithium buffer system of AAA. After reacting with ninhydrin, the resultant derivatives were measured by UV detection at two wavelengths of 440 and 570 nm simultaneously. Due to a low response at 570 nm, the derivative of Pro is detected at 440 nm. The chromatographic profiles of the mixed standard solution and tobacco sample were obtained by AAA. An example of the result is shown in Fig. 1. Figure. 1 (A1) and (A2) shows the chromatographic profiles of a mixed standard solution detected at 570 and 440 nm, respectively. The chromatographic profiles of a tobacco sample are shown in Fig. 1 (B1) and (B2). It is clear that, although 106.8 min elution is needed to separate the amino acids, there are still overlapping peaks in both the chromatographic profiles of mixed standard solution and tobacco samples. There are three groups of overlapping peaks, including Asn, Glu, Gly, Ala, His, and methylhistidine (Mehis). The chromatographic profiles of Asn and Gly are overlapping respectively with Glu and Ala, and the signal of His is overlapping with Mehis. Due to the low content of Mehis in tobacco samples, it is not considered as the target amino acid in the calculation.
2.2 Samples preparation A 0.5000 g tobacco sample powder was weighed into the flask and extracted with 50 mL hydrochloric acid (0.005 mol/L) by ultrasonic cleaner for 35 min at 40⬚C. The extract was C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.4 Chemometric calculations To resolve the overlapping signals, NNIA is employed for extracting the information of the amino acids. NNIA is www.jss-journal.com
J. Sep. Sci. 2015, 38, 2053–2058
Liquid Chromatography
2055
Figure 2. Resolved results by NNIA from the chromatogram of the mixed standard solution: the solid lines in black are the experimental signals of the mixed standard solution; the dash lines in orange are the residual signals after the resolution; the other dot lines in different colors are the extracted signals of Asn, Glu, Gly, Ala, and His.
developed on the basis of IA, which extracts information by projection and an iteration of subtraction. The iteration works with the help of the provided standard signals. The calculation will stop until the residual approaches zero, which means that the information of the components has been completely subtracted. A non-negative correction is added in NNIA, which makes the method able to extract the information of the interested components one by one from the measured data matrix without the requirement of providing the standard signals of all the components. In the calculation, the chromatographic profiles of the standard amino acids, including Asn, Glu, Gly, Ala, and His were measured and taken as the standard signals of the resolving components, respectively. Therefore the chromatograms of the five amino acids were obtained. Then peak areas were adopted for quantitative analysis of all the amino acids.
3 Results and discussion 3.1 Resolving the overlapping peaks of standard solutions To validate the accuracy of the chemometric method, the signal of mixed standard solution is resolved by NNIA. As shown in Fig. 1 (A1), it is difficult to separate all the components in the mixed standard solution. Therefore, NNIA was used to extract the information of the amino acids. The chromatographic profiles of the standard amino acids were taken as the input, respectively. Because the five overlapping peaks of amino acids, including Asn, Glu, Gly, Ala, and His, are located in the retention time around 25.2, 26.0, 36.2, 37.4, and 79.0 min, the data in the region of 23.8–28.1, 35.2–42.5, and 78.2–81.2 min are used in the calculation. The resolved results are shown in Fig. 2. The solid line in black is the total experimental signal and the resolved C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
chromatograms of the five amino acids are plotted by the dot lines in different colors. The dash line in orange is the residual signal after the resolution. From the shapes of the resolved signals and the profile of the residual signal, it is indicated that the decomposition is reasonable. To demonstrate the reproducibility of the method, five repeated measurements were performed. Then relative concentrations of the amino acids were calculated by the chemometric resolution and peak area integral and the RSDs (n = 5) of the results are less than 1.57%.
3.2 Resolving the overlapping peaks of tobacco samples and validation of the proposed method To investigate the feasibility of the proposed method, resolution of the overlapping signals measured with the tobacco samples is performed. The chromatographic profiles of the amino acids in tobacco leaves can be obtained with NNIA. An example of the resolved chromatographic profiles for tobacco samples are shown in Fig. 3. In the figure, the solid line in black shows the total chromatographic profile of the experimental signal and the resolved results of the five amino acids (Asn, Glu, Gly, Ala, His) are plotted by the dashed lines in different colors. The curve in orange shows the residual after the resolution. Investigating the differences between the experimental and the resolved signal, the rationality of the results can be proved by the shapes of the resolved results. From the residual signals, it is clear that almost all the information of the amino acids in tobacco samples is extracted by the resolution. Validation of the assay included the LOD, LOQ, linearity, precision, and accuracy. The LODs range from 0.18 to 0.58 mol/L and the LOQs range from 0.68 to 1.99 mol/L. For quantification, the calibration curves of the 17 amino acids were established with the peak areas of the standard www.jss-journal.com
2056
J. Sep. Sci. 2015, 38, 2053–2058
Y. Zeng et al.
Figure 3. Resolved results by NNIA from the chromatogram of the tobacco sample: the solid lines in black are the experimental signals of the tobacco sample; the dash lines in orange are the residual signals after the resolution; the other dot lines in different colors are the extracted signals of Asn, Glu, Gly, Ala, and His. Table 1. Accuracy of the proposed method and conventional method
Analyte
Calculated (RSD, %)
Added
Found (RSD, %)
Recovery-NNIA (%)
Recovery-conventional method (%)
Asp Thr Ser Asn Glu Gly Ala Val Ile Leu Tyr Phe g-ABA His Lys Arg Pro
0.98 (0.63) 0.17 (1.68) 0.24 (1.21) 0.44 (0.53) 0.32 (0.75) 0.20 (0.86) 0.28 (0.54) 0.20 (0.71) 0.13 (1.78) 0.14 (1.93) 0.14 (1.48) 0.21 (2.04) 0.19 (1.40) 0.16 (1.41) 0.15 (1.27) 0.14 (0.86) 0.27 (2.10)
0.98 0.17 0.24 0.44 0.32 0.20 0.28 0.20 0.13 0.14 0.14 0.21 0.19 0.16 0.15 0.14 0.27
1.95 (1.52) 0.35 (0.80) 0.50 (1.03) 0.91 (0.17) 0.66 (1.35) 0.39 (0.70) 0.57 (1.26) 0.40 (0.54) 0.26 (0.89) 0.28 (1.51) 0.28 (0.29) 0.44 (1.81) 0.37 (2.54) 0.31 (1.57) 0.31 (1.76) 0.27 (1.55) 0.54 (1.94)
99.21 102.31 105.59 106.55 105.33 98.20 107.24 104.25 96.87 98.23 100.14 108.21 94.62 97.29 104.25 94.84 95.26
99.15 105.40 107.22 109.23 108.17 98.15 104.37 105.91 96.38 97.56 101.19 109.05 94.53 94.39 106.81 92.77 91.80
samples with five different concentrations. Good linearity (r ࣙ 0.9990) was observed for all the amino acids. Furthermore, to validate the accuracy of the resolution, standard addition method was employed. Five repetitions of three spiked samples were performed. An example of the quantitative results obtained by the proposed method is shown in Table 1. The calculated concentrations are listed in the second column of the table. The recoveries listed in the fifth column of the table are obtained by dividing the calculated result of spiked standard solution by the theoretical value. It can be seen that the values of recoveries range from 94.62 to 108.21%, demonstrating a good accuracy. Then, the precision is expressed as RSD that is the ratio of SD to average value. They are calculated and listed in the fourth column. It can be found that all the values of RSDs are less than 2.54%, which suggests a very good reproducibility of the proposed method. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Furthermore, to compare with conventional method, the recoveries of amino acids were calculated by the method that accompany the instrument and the results were listed in the last column of Table 1. It can be found that the recoveries calculated by NNIA are better than conventional method. The results demonstrate a better accuracy and a good correctness of the proposed method.
3.3 Quantitative results of 17 amino acids in tobacco samples The analyses of 17 amino acids in the 30 tobacco samples were performed by the proposed method. All the samples were extracted and analyzed under the same condition. With the proposed method, both chromatographic profiles www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2053–2058
2057
Table 2. Quantitative results (mg/g) of 17 amino acids in 30 tobacco samples Amino acid No.
Sample information
Asp
Thr
Ser
Asn
Glu
Gly
Ala
Val
Ile
Leu
Tyr
Phe
g-ABA
His
Lys
Arg
Pro
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
R1-1-U R1-1-M R1-2-U R1-2-M R1-3-U R1-3-M R1-4-U R1-4-M R1-5-U R1-5-M R2-6-U R2-6-M R2-7-U R2-7-M R2-8-U R2-8-M R2-9-U R2-9-M R2-10-U R2-10-M R3-11-U R3-11-M R3-12-U R3-12-M R3-13-U R3-13-M R3-14-U R3-14-M R3-15-U R3-15-M
6.72 5.44 7.72 6.41 5.51 4.53 1.72 7.61 11.16 9.23 6.5 6.45 4.8 3.89 11.23 7.95 14.3 6.14 9.38 6.75 14.30 6.98 6.64 6.45 9.38 4.80 6.75 7.61 11.16 9.23
0.15 0.11 0.09 0.11 0.47 0.24 0.38 0.28 0.42 0.37 0.23 0.17 0.15 0.29 0.37 0.22 0.36 0.54 0.23 0.26 0.36 0.22 0.54 0.17 0.23 0.15 0.26 0.28 0.42 0.37
0.15 0.24 0.21 0.15 0.86 0.57 0.32 0.73 1.11 0.45 0.25 0.22 0.22 0.49 0.45 0.36 0.89 0.98 0.82 0.65 0.89 0.48 0.98 0.22 0.82 0.22 0.65 0.73 1.11 0.45
20.05 18.02 16.13 13.57 47.81 18.96 30.76 21.54 40.72 30.22 33.65 22.58 10.72 23.6 30.22 11.54 4.74 36.67 16.31 23.62 4.74 7.84 36.67 22.58 16.31 10.72 23.62 21.54 40.72 30.22
0.78 0.78 0.29 0.56 1.4 1.01 1.18 0.89 1.71 0.88 1.17 0.67 0.56 1.03 0.88 1.03 1.76 1.43 0.23 0.55 1.76 1.10 1.43 0.67 0.23 0.26 0.55 0.89 1.71 0.88
0.11 0.09 0.09 0.1 0.15 0.14 0.19 0.19 0.2 0.21 0.08 0.13 0.14 0.16 0.21 0.15 0.17 0.15 0.14 0.17 0.17 0.13 0.15 0.13 0.14 0.14 0.17 0.19 0.20 0.21
0.87 0.45 0.31 0.55 1.02 0.66 2.2 0.81 0.99 0.9 1 0.7 0.39 0.69 0.9 0.51 0.48 0.96 0.36 0.67 0.48 0.35 0.96 0.70 0.46 0.39 0.67 0.81 0.99 0.90
0.11 0.12 0.02 0.09 0.18 0.22 0.19 0.2 0.31 0.26 0.13 0.17 0.12 0.2 0.26 0.2 0.25 0.36 0.19 0.04 0.25 0.12 0.36 0.17 0.19 0.12 0.04 0.20 0.31 0.26
0.04 0.04 0.03 0.04 0.07 0.08 0.08 0.08 0.1 0.04 0.05 0.05 0.04 0.09 0.04 0.01 0.08 0.08 0.04 0.12 0.08 0.06 0.08 0.05 0.04 0.04 0.12 0.08 0.10 0.04
0.1 0.09 0.07 0.09 0.15 0.14 0.14 0.18 0.15 0.07 0.11 0.11 0.09 0.06 0.07 0.06 0.14 0.15 0.02 0.12 0.14 0.10 0.15 0.11 0.02 0.09 0.12 0.18 0.15 0.07
0.04 0.06 0.02 0.03 0.03 0.11 0.2 0.07 0.13 0.25 0.08 0.05 0.03 0.13 0.25 0.1 0.13 0.32 0.04 0.14 0.13 0.04 0.32 0.05 0.04 0.03 0.14 0.07 0.13 0.25
0.56 0.53 0.3 0.52 1.36 0.82 1.24 0.18 1.47 0.06 0.91 0.93 0.37 0.05 0.06 0.04 0.91 1.61 0.08 0.1 0.91 0.49 1.61 0.93 0.08 0.37 0.10 0.18 1.47 0.06
0.51 0.39 0.32 0.37 0.62 0.35 1.16 0.62 0.84 0.63 0.5 0.45 0.32 0.45 0.63 0.11 0.64 0.88 0.33 0.29 0.64 0.40 0.88 0.45 0.33 0.32 0.59 0.62 0.84 0.63
0.38 0.39 0.29 0.29 0.88 0.63 0.8 0.73 0.97 1.05 0.61 0.78 0.39 0.63 1.05 0.47 0.44 0.93 0.52 0.73 0.44 0.37 0.93 0.78 0.52 0.39 0.73 0.73 0.97 1.05
0.26 0.13 0.17 0.22 0.42 0.21 0.81 0.28 0.4 0.66 0.34 0.32 0.22 0.26 0.66 0.34 0.23 0.39 0.15 0.61 0.23 0.26 0.39 0.32 0.35 0.22 0.61 0.28 0.40 0.66
0.17 0.1 0.14 0.17 0.27 0.16 0.32 0.18 0.21 0.42 0.23 0.24 0.21 0.27 0.42 0.19 0.16 0.3 0.11 0.18 0.16 0.17 0.30 0.24 0.21 0.21 0.28 0.18 0.21 0.42
8.57 6.24 1.34 3.52 4.49 3.61 16.94 2.55 7.69 2.67 6.24 2.1 1.39 2.29 2.67 0.94 2.16 14.34 0.54 2.02 2.16 0.72 4.34 2.10 1.54 1.39 2.32 2.55 7.69 2.67
and relative concentrations can be obtained. The contents of amino acids in measured samples are calculated and listed in Table 2. The sample information listed in the second column containing the regions, numbers of the sample and the sampling positions in leaves. For example, R1-2-U means the tobacco sample was numbered 2 and collected from the upper parts of the leaves from region 1. As shown in Table 2, Asn is the richest amino acid component in the burley tobacco sample. The result is in agreement with that reported in the literature [33]. Comparing the content of amino acids in different regions, the total content of all 17 kinds of amino acids in R1 is higher than the other two regions. It can be inferred that the content of free amino acids in tobacco leaves are significantly influenced by growth environment [34, 35]. Comparing the content of amino acids in different sampling positions, it can be observed that the content of amino acids in the upper parts of leaves is higher than that in the middle parts, which will be helpful to improve the utilization rate of amino acids in tobacco leaves.
overlapping signals in measured chromatographic data, NNIA was employed for extracting the information of amino acids in tobacco leaves. The results showed that both the chromatogram and the relative concentration of each amino acid can be obtained by using the proposed method. With five repetitions of three spiked sample, the recoveries of the spiked samples were range from 94.62 to 108.21%. Then by analyzing the 17 amino acids in 30 tobacco leaf samples, the conclusions can be drawn that Asp, Asn, and Pro are the main components of burley tobacco leaves. The content of all 17 kinds of amino acids in upper parts of leaves from region 1 is higher than others.
4 Conclusion
5 References
A method was developed and used for quantifying the amount of amino acids in tobacco leaves. Considering the C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
This study was supported by National Natural Science Foundation of China (Nos. 21175074 and 21475068). Thanks also go to Beijing Third Class Tobacco Supervision Station for providing the samples and the experimental measurements. The authors have declared no conflict of interest.
[1] Zhang, J. J., Zhao, C. X., Chang, Y. W., Zhao, Y. N., Li, Q. H., Lu, X., Xu, G. W., J. Sep. Sci. 2013, 36, 2686–2877.
www.jss-journal.com
2058
Y. Zeng et al.
J. Sep. Sci. 2015, 38, 2053–2058
[2] Gwatidzo, L., Botha, B. M., McCrindle, R. I., Food Chem. 2013, 141, 2163–2169.
[19] Goodman, K. J., Brenna, J. T., Anal. Chem. 1994, 66, 1294–1301.
[3] Avino, P., Campanella, L., Mario, V. R., J. Sep. Sci. 2003, 26, 392–396.
[20] Layer, E., Tomczyk, K., Signal Transforms in Dynamic Measurements, Springer International Publishing, Berlin 2015, 16, pp. 169–188.
[4] Pereira, D. M., Valentao, P., Teixeira, N., Andrade, P. B., Food Chem. 2013, 141, 2412–2417.
[21] Maeder, M., Anal. Chem. 1987, 59, 527–530.
[5] Chiu, T.-C., Anal. Bioanal. Chem. 2013, 405, 7919– 7930.
[22] Manne, R., Shen, H. L., Liang, Y. Z., Chemom. Intell. Lab. Syst. 1999, 45, 171–176.
[6] Friedman, M., J. Arg. Food Chem. 2004, 52, 385– 406.
[23] Liang, Y. Z., Kvalheim, O. M., Keller, H. R., Massart, D. L., Kiechle, P., Erni, F., Anal. Chem. 1992, 64, 946–953.
[7] Sun, S. W., Lin, Y. C., Weng, Y. M., Chen, M. J., J. Food Compos. Anal. 2006, 19, 112–117.
[24] Zhao, Z. M., Malinowski, E. R., Anal. Chem. 1999, 71, 602–608.
[8] Zuraini, A., Somchit, M. N., Solihah, M. H., Goh, Y. M., Arifah, A. K., Zakaria, M. S., Somchit, N., Rajion, M. A., Zakaria, Z. A., Mat Jais A. M., Food Chem. 2006, 97, 674– 678.
¨ [25] Schworer1, R., Rothl, J., J. Appl. Phys. 1995, 77, 3812– 3817.
¨ [9] Nilsson, M., Holst, J. J., Bjorck, I. M. E., Am. J. Clin. Nutr. 2007, 85, 996–1004.
[27] Navea, S., de Juan, A., Tauler, R., Anal. Chim. Acta 2001, 446, 185–195.
[10] Lund, E., Thomsen, J., Brunfeldt, K., J. Chromatogr. 1977, 130, 51–54.
[28] Hantao, L. W., Aleme, H. G., Pedroso, M. P., Sabin, G. P., Poppi, R. J., Augusto, F., Anal. Chim. Acta 2012, 731, 11–23.
[11] Kaspar, H., Dettmer, K., Gronwald, W., Oefner, P. J., Anal. Bio. Chem. 2009, 393, 445–452. [12] Durham, B., Geren, C. R., Anal. Biochem. 1981, 116, 331– 334. [13] Alonzo, N., Hirs, C. H. W., Anal. Biochem. 1968, 23, 272– 288. [14] Zhang, W., Cao, Y. Z., Liu, Y. M., Cui, Chemistry 2002, 6, 418–421. [15] Han, J., Li, P., Cai, W. S., Shao, X. G., J. Sep. Sci. 2014, 37, 2126–2130. [16] Zhang, Q. L., Ni, Y. N., Kokot, S., J. Pharm. Biomed. Anal. 2010, 52, 280–288.
˜ ˜ [26] Gusmao, R., Arino, C., D´ıaz-Cruz, J. M., Esteban, M., Anal. Bioanal. Chem. 2009, 394, 1137–1145.
[29] Shao, X. G., Yu, Z. L., Sun, L., TrAC, Trends Anal. Chem. 2003, 22, 59–69. [30] Shao, X. G., Wang, G. Q., Wang, S. F., Su, Q. D., Anal. Chem. 2004, 76, 5143–5148. [31] Liu, Z. C., Cai, W. S., Shao, X. G., J. Chromatogr. A 2009, 1216, 1469–1475. [32] Shao, X. G., Liu, Z. C., Cai, W. S., TrAC, Trends Anal. Chem. 2009, 28, 1312–1321. [33] Li, G., Wu, D., Xie, W. Y., Sha, Y. F., Lin, H. Q., Liu, B. Z., J. Chromatogr. A 2013, 1296, 243–247.
[17] Ni, Y. N., Wang, Y., S. Kokot, Talanta 2009, 78, 432–441.
[34] Wilkinson, R. E., Kasperbauer, M. J., Young, C. T., J. Agric. Food Chem. 1981, 29, 658–660.
[18] Miao, L. N., Cai, W. S., Shao, X. G., Talanta 2011, 83, 1247–1253.
[35] Zhang, J. S., Jia, C. X., Li, Y. Q., Mao, D. B., Zhang, W. Y., Tob. Sci. Technol. 2004, 205, 26–32.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
ISSN 1615-9306 · JSSCCJ 38 (12) 2007–2192 (2015) · Vol. 38 · No. 12 · June 2015 · D 10609
JOURNAL OF
SEPARATION SCIENCE
Methods Chromatography · Electroseparation Applications Biomedicine · Foods · Environment
12 15
www.jss-journal.com
2053
J. Sep. Sci. 2015, 38, 2053–2058
Yihang Zeng Wensheng Cai Xueguang Shao Research Center for Analytical Sciences, College of Chemistry, Nankai University, Tianjin Key Laboratory of Biosensing and Molecular Recognition, State Key Laboratory of Medicinal Chemical Biology, Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin, China Received January 22, 2015 Revised March 27, 2015 Accepted March 30, 2015
Research Article
Quantitative analysis of 17 amino acids in tobacco leaves using an amino acid analyzer and chemometric resolution A method was developed for quantifying 17 amino acids in tobacco leaves by using an A300 amino acid analyzer and chemometric resolution. In the method, amino acids were eluted by the buffer solution on an ion-exchange column. After reacting with ninhydrin, the derivatives of amino acids were detected by ultraviolet detection. Most amino acids are separated by the elution program. However, five peaks of the derivatives are still overlapping. A non-negative immune algorithm was employed to extract the profiles of the derivatives from the overlapping signals, and then peak areas were adopted for quantitative analysis of the amino acids. The method was validated by the determination of amino acids in tobacco leaves. The relative standard deviations (n = 5) are all less than 2.54% and the recoveries of the spiked samples are in a range of 94.62–108.21%. The feasibility of the method was proved by analyzing the 17 amino acids in 30 tobacco leaf samples. Keywords: Amino acids / Chemometrics / Non-negative immune algorithm / Tobacco leaves DOI 10.1002/jssc.201500090
1 Introduction In recent years, there has been an increasing interest for accurate quantitative analysis of amino acids in biological research [1]. Several methods can be used to analyze amino acids, such as HPLC [2, 3], GC–MS [4], and CE [5]. The most commonly used method is postcolumn derivatization with ninhydrin using LC [6, 7], which has shown good reliability. Based on the method, an automatic amino acid analyzer (AAA) was developed and widely used to quantify amino acids automatically [8, 9]. It uses the classic method of postcolumn derivatization with ninhydrin [10], known as “the Gold Standard” [11], to analyze amino acids. The AAA system realizes automation by combining autosampler with a separation column unit and reactor unit [12]. It makes the analysis of amino acids automatic, fast, and simple [13]. However, AAA system still has the problem of peak overlap when analyzing multicomponent samples, such as tobacco leaf and ginseng [14,15]. Therefore, more efficient methods are necessary to be developed.
Correspondence: Professor Xueguang Shao, Research Center for Analytical Sciences, College of Chemistry, Nankai University, Tianjin 300071, P. R. China Fax: +86-22-23502458 E-mail: [email protected]
Abbreviations: AAA, automatic amino acid analyzer; Ala, alanine; Asn, asparagine; Asp, aspartic acid; CFA, chemical factor analysis; EFA, evolving factor analysis; Glu, glutamic acid; Gly, glycine; HELP, heuristic evolving latent projections; His, histidine; IA, immune algorithm; Mehis, methylhistidine; NNIA, nonnegative immune algorithm; Pro, proline
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Chemometrics provides new opportunities for the resolution of overlapping chromatographic signals [16–18]. To analyze overlapping 1D signals, curve fitting techniques, convolution, and deconvolution methods, and derivation calculation have been developed [19, 20]. For overlapping 2D signals, a variety of chemometric methods such as evolving factor analysis (EFA) [21], subwindow factor analysis [22] and HELP [23], etc. based on chemical factor analysis (CFA) have been developed. CFA is a branch of multivariate analysis and shows obvious advantage when it comes to multivariate problems [24, 25]. Furthermore, alternative methods like multivariate curve resolution-alternating least squares [26], based on bilinear models, are developed and employed in the analysis of environmental and biological samples [27, 28]. In our previous studies, an immune algorithm (IA) has been developed for resolving overlapping signals in HPLC, spectroscopy, and GC–MS analyses [29]. IA extracts information of each component from the total signal by projection and subtraction with the help of the provided standard signals of all the possible components [30]. In the primary IA, therefore, the information of all the components possibly contained in the analyzing sample must be provided [31]. For retrieving the information of a specific component from an overlapping signal, non-negative immune algorithm (NNIA) was proposed and had been proved to be an efficient way for determining an interested component in a complex sample [32]. In this work, a method for quantifying amino acids in tobacco leaves by AAA with chemometric resolution was developed. Most amino acids are separated on the ion-exchange column of AAA system. However, there are five peaks of amino acid derivatives still overlapping. NNIA is used to resolve the overlapping signals of amino acids. With the www.jss-journal.com
2054
Y. Zeng et al.
J. Sep. Sci. 2015, 38, 2053–2058
Figure 1. Chromatograms of mixed standard solution and tobacco sample.
resolved signals, the amino acids can be quantitatively determined with an acceptable precision and accuracy. Then the feasibility of the method is proved by analyzing the 17 amino acids in 30 tobacco samples.
2 Materials and methods 2.1 Materials and reagents Hydrochloric acid (HPLC grade) and the mixed standard solution of amino acids (100 nmol/L) were purchased from Sinopharm Chemical Reagent (Beijing, China). The single standard solutions (100 nmol/L), including aspartic acid (Asp), threonine (Thr), serine (Ser), asparagine (Asn), glutamic acid (Glu), glycine (Gly), alanine (Ala), valine (Val), isoleucine (Ile), leucine (Leu), tyrosine (Tyr), phenylalanine (Phe), ␥-amino butyric acid (g-ABA), histidine (His), lysine (Lys), arginine (Arg), and proline (Pro), were also obtained from Sinopharm Chemical Reagent (Beijing, China). Lithium buffer system and ninhydrin solution were purchased from membraPure (Bodenheim, Germany). Thirty burley tobacco samples selected from three different regions (R1, R2, R3) were provided by Beijing Cigarette Factory (Beijing, China). To investigate the content of amino acids in different sampling position of tobacco leaves, 15 samples were selected from upper (U) parts of leaves and the other 15 samples were collected from the middle (M) parts.
centrifuged at 15 000 rpm for 15 min, then filtered with 0.22 m membrane, and stored at 4⬚C for AAA analysis. 2.3 Chromatographic analysis The analysis of amino acids was carried out by using A300 amino acid analyzer (membraPure Bodenheim, Germany). The ion-exchange chromatographic column (60–0041, membraPure Bodenheim, Germany) was adopted, where the amino acids were eluted by lithium buffer system of AAA. After reacting with ninhydrin, the resultant derivatives were measured by UV detection at two wavelengths of 440 and 570 nm simultaneously. Due to a low response at 570 nm, the derivative of Pro is detected at 440 nm. The chromatographic profiles of the mixed standard solution and tobacco sample were obtained by AAA. An example of the result is shown in Fig. 1. Figure. 1 (A1) and (A2) shows the chromatographic profiles of a mixed standard solution detected at 570 and 440 nm, respectively. The chromatographic profiles of a tobacco sample are shown in Fig. 1 (B1) and (B2). It is clear that, although 106.8 min elution is needed to separate the amino acids, there are still overlapping peaks in both the chromatographic profiles of mixed standard solution and tobacco samples. There are three groups of overlapping peaks, including Asn, Glu, Gly, Ala, His, and methylhistidine (Mehis). The chromatographic profiles of Asn and Gly are overlapping respectively with Glu and Ala, and the signal of His is overlapping with Mehis. Due to the low content of Mehis in tobacco samples, it is not considered as the target amino acid in the calculation.
2.2 Samples preparation A 0.5000 g tobacco sample powder was weighed into the flask and extracted with 50 mL hydrochloric acid (0.005 mol/L) by ultrasonic cleaner for 35 min at 40⬚C. The extract was C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.4 Chemometric calculations To resolve the overlapping signals, NNIA is employed for extracting the information of the amino acids. NNIA is www.jss-journal.com
J. Sep. Sci. 2015, 38, 2053–2058
Liquid Chromatography
2055
Figure 2. Resolved results by NNIA from the chromatogram of the mixed standard solution: the solid lines in black are the experimental signals of the mixed standard solution; the dash lines in orange are the residual signals after the resolution; the other dot lines in different colors are the extracted signals of Asn, Glu, Gly, Ala, and His.
developed on the basis of IA, which extracts information by projection and an iteration of subtraction. The iteration works with the help of the provided standard signals. The calculation will stop until the residual approaches zero, which means that the information of the components has been completely subtracted. A non-negative correction is added in NNIA, which makes the method able to extract the information of the interested components one by one from the measured data matrix without the requirement of providing the standard signals of all the components. In the calculation, the chromatographic profiles of the standard amino acids, including Asn, Glu, Gly, Ala, and His were measured and taken as the standard signals of the resolving components, respectively. Therefore the chromatograms of the five amino acids were obtained. Then peak areas were adopted for quantitative analysis of all the amino acids.
3 Results and discussion 3.1 Resolving the overlapping peaks of standard solutions To validate the accuracy of the chemometric method, the signal of mixed standard solution is resolved by NNIA. As shown in Fig. 1 (A1), it is difficult to separate all the components in the mixed standard solution. Therefore, NNIA was used to extract the information of the amino acids. The chromatographic profiles of the standard amino acids were taken as the input, respectively. Because the five overlapping peaks of amino acids, including Asn, Glu, Gly, Ala, and His, are located in the retention time around 25.2, 26.0, 36.2, 37.4, and 79.0 min, the data in the region of 23.8–28.1, 35.2–42.5, and 78.2–81.2 min are used in the calculation. The resolved results are shown in Fig. 2. The solid line in black is the total experimental signal and the resolved C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
chromatograms of the five amino acids are plotted by the dot lines in different colors. The dash line in orange is the residual signal after the resolution. From the shapes of the resolved signals and the profile of the residual signal, it is indicated that the decomposition is reasonable. To demonstrate the reproducibility of the method, five repeated measurements were performed. Then relative concentrations of the amino acids were calculated by the chemometric resolution and peak area integral and the RSDs (n = 5) of the results are less than 1.57%.
3.2 Resolving the overlapping peaks of tobacco samples and validation of the proposed method To investigate the feasibility of the proposed method, resolution of the overlapping signals measured with the tobacco samples is performed. The chromatographic profiles of the amino acids in tobacco leaves can be obtained with NNIA. An example of the resolved chromatographic profiles for tobacco samples are shown in Fig. 3. In the figure, the solid line in black shows the total chromatographic profile of the experimental signal and the resolved results of the five amino acids (Asn, Glu, Gly, Ala, His) are plotted by the dashed lines in different colors. The curve in orange shows the residual after the resolution. Investigating the differences between the experimental and the resolved signal, the rationality of the results can be proved by the shapes of the resolved results. From the residual signals, it is clear that almost all the information of the amino acids in tobacco samples is extracted by the resolution. Validation of the assay included the LOD, LOQ, linearity, precision, and accuracy. The LODs range from 0.18 to 0.58 mol/L and the LOQs range from 0.68 to 1.99 mol/L. For quantification, the calibration curves of the 17 amino acids were established with the peak areas of the standard www.jss-journal.com
2056
J. Sep. Sci. 2015, 38, 2053–2058
Y. Zeng et al.
Figure 3. Resolved results by NNIA from the chromatogram of the tobacco sample: the solid lines in black are the experimental signals of the tobacco sample; the dash lines in orange are the residual signals after the resolution; the other dot lines in different colors are the extracted signals of Asn, Glu, Gly, Ala, and His. Table 1. Accuracy of the proposed method and conventional method
Analyte
Calculated (RSD, %)
Added
Found (RSD, %)
Recovery-NNIA (%)
Recovery-conventional method (%)
Asp Thr Ser Asn Glu Gly Ala Val Ile Leu Tyr Phe g-ABA His Lys Arg Pro
0.98 (0.63) 0.17 (1.68) 0.24 (1.21) 0.44 (0.53) 0.32 (0.75) 0.20 (0.86) 0.28 (0.54) 0.20 (0.71) 0.13 (1.78) 0.14 (1.93) 0.14 (1.48) 0.21 (2.04) 0.19 (1.40) 0.16 (1.41) 0.15 (1.27) 0.14 (0.86) 0.27 (2.10)
0.98 0.17 0.24 0.44 0.32 0.20 0.28 0.20 0.13 0.14 0.14 0.21 0.19 0.16 0.15 0.14 0.27
1.95 (1.52) 0.35 (0.80) 0.50 (1.03) 0.91 (0.17) 0.66 (1.35) 0.39 (0.70) 0.57 (1.26) 0.40 (0.54) 0.26 (0.89) 0.28 (1.51) 0.28 (0.29) 0.44 (1.81) 0.37 (2.54) 0.31 (1.57) 0.31 (1.76) 0.27 (1.55) 0.54 (1.94)
99.21 102.31 105.59 106.55 105.33 98.20 107.24 104.25 96.87 98.23 100.14 108.21 94.62 97.29 104.25 94.84 95.26
99.15 105.40 107.22 109.23 108.17 98.15 104.37 105.91 96.38 97.56 101.19 109.05 94.53 94.39 106.81 92.77 91.80
samples with five different concentrations. Good linearity (r ࣙ 0.9990) was observed for all the amino acids. Furthermore, to validate the accuracy of the resolution, standard addition method was employed. Five repetitions of three spiked samples were performed. An example of the quantitative results obtained by the proposed method is shown in Table 1. The calculated concentrations are listed in the second column of the table. The recoveries listed in the fifth column of the table are obtained by dividing the calculated result of spiked standard solution by the theoretical value. It can be seen that the values of recoveries range from 94.62 to 108.21%, demonstrating a good accuracy. Then, the precision is expressed as RSD that is the ratio of SD to average value. They are calculated and listed in the fourth column. It can be found that all the values of RSDs are less than 2.54%, which suggests a very good reproducibility of the proposed method. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Furthermore, to compare with conventional method, the recoveries of amino acids were calculated by the method that accompany the instrument and the results were listed in the last column of Table 1. It can be found that the recoveries calculated by NNIA are better than conventional method. The results demonstrate a better accuracy and a good correctness of the proposed method.
3.3 Quantitative results of 17 amino acids in tobacco samples The analyses of 17 amino acids in the 30 tobacco samples were performed by the proposed method. All the samples were extracted and analyzed under the same condition. With the proposed method, both chromatographic profiles www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2015, 38, 2053–2058
2057
Table 2. Quantitative results (mg/g) of 17 amino acids in 30 tobacco samples Amino acid No.
Sample information
Asp
Thr
Ser
Asn
Glu
Gly
Ala
Val
Ile
Leu
Tyr
Phe
g-ABA
His
Lys
Arg
Pro
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
R1-1-U R1-1-M R1-2-U R1-2-M R1-3-U R1-3-M R1-4-U R1-4-M R1-5-U R1-5-M R2-6-U R2-6-M R2-7-U R2-7-M R2-8-U R2-8-M R2-9-U R2-9-M R2-10-U R2-10-M R3-11-U R3-11-M R3-12-U R3-12-M R3-13-U R3-13-M R3-14-U R3-14-M R3-15-U R3-15-M
6.72 5.44 7.72 6.41 5.51 4.53 1.72 7.61 11.16 9.23 6.5 6.45 4.8 3.89 11.23 7.95 14.3 6.14 9.38 6.75 14.30 6.98 6.64 6.45 9.38 4.80 6.75 7.61 11.16 9.23
0.15 0.11 0.09 0.11 0.47 0.24 0.38 0.28 0.42 0.37 0.23 0.17 0.15 0.29 0.37 0.22 0.36 0.54 0.23 0.26 0.36 0.22 0.54 0.17 0.23 0.15 0.26 0.28 0.42 0.37
0.15 0.24 0.21 0.15 0.86 0.57 0.32 0.73 1.11 0.45 0.25 0.22 0.22 0.49 0.45 0.36 0.89 0.98 0.82 0.65 0.89 0.48 0.98 0.22 0.82 0.22 0.65 0.73 1.11 0.45
20.05 18.02 16.13 13.57 47.81 18.96 30.76 21.54 40.72 30.22 33.65 22.58 10.72 23.6 30.22 11.54 4.74 36.67 16.31 23.62 4.74 7.84 36.67 22.58 16.31 10.72 23.62 21.54 40.72 30.22
0.78 0.78 0.29 0.56 1.4 1.01 1.18 0.89 1.71 0.88 1.17 0.67 0.56 1.03 0.88 1.03 1.76 1.43 0.23 0.55 1.76 1.10 1.43 0.67 0.23 0.26 0.55 0.89 1.71 0.88
0.11 0.09 0.09 0.1 0.15 0.14 0.19 0.19 0.2 0.21 0.08 0.13 0.14 0.16 0.21 0.15 0.17 0.15 0.14 0.17 0.17 0.13 0.15 0.13 0.14 0.14 0.17 0.19 0.20 0.21
0.87 0.45 0.31 0.55 1.02 0.66 2.2 0.81 0.99 0.9 1 0.7 0.39 0.69 0.9 0.51 0.48 0.96 0.36 0.67 0.48 0.35 0.96 0.70 0.46 0.39 0.67 0.81 0.99 0.90
0.11 0.12 0.02 0.09 0.18 0.22 0.19 0.2 0.31 0.26 0.13 0.17 0.12 0.2 0.26 0.2 0.25 0.36 0.19 0.04 0.25 0.12 0.36 0.17 0.19 0.12 0.04 0.20 0.31 0.26
0.04 0.04 0.03 0.04 0.07 0.08 0.08 0.08 0.1 0.04 0.05 0.05 0.04 0.09 0.04 0.01 0.08 0.08 0.04 0.12 0.08 0.06 0.08 0.05 0.04 0.04 0.12 0.08 0.10 0.04
0.1 0.09 0.07 0.09 0.15 0.14 0.14 0.18 0.15 0.07 0.11 0.11 0.09 0.06 0.07 0.06 0.14 0.15 0.02 0.12 0.14 0.10 0.15 0.11 0.02 0.09 0.12 0.18 0.15 0.07
0.04 0.06 0.02 0.03 0.03 0.11 0.2 0.07 0.13 0.25 0.08 0.05 0.03 0.13 0.25 0.1 0.13 0.32 0.04 0.14 0.13 0.04 0.32 0.05 0.04 0.03 0.14 0.07 0.13 0.25
0.56 0.53 0.3 0.52 1.36 0.82 1.24 0.18 1.47 0.06 0.91 0.93 0.37 0.05 0.06 0.04 0.91 1.61 0.08 0.1 0.91 0.49 1.61 0.93 0.08 0.37 0.10 0.18 1.47 0.06
0.51 0.39 0.32 0.37 0.62 0.35 1.16 0.62 0.84 0.63 0.5 0.45 0.32 0.45 0.63 0.11 0.64 0.88 0.33 0.29 0.64 0.40 0.88 0.45 0.33 0.32 0.59 0.62 0.84 0.63
0.38 0.39 0.29 0.29 0.88 0.63 0.8 0.73 0.97 1.05 0.61 0.78 0.39 0.63 1.05 0.47 0.44 0.93 0.52 0.73 0.44 0.37 0.93 0.78 0.52 0.39 0.73 0.73 0.97 1.05
0.26 0.13 0.17 0.22 0.42 0.21 0.81 0.28 0.4 0.66 0.34 0.32 0.22 0.26 0.66 0.34 0.23 0.39 0.15 0.61 0.23 0.26 0.39 0.32 0.35 0.22 0.61 0.28 0.40 0.66
0.17 0.1 0.14 0.17 0.27 0.16 0.32 0.18 0.21 0.42 0.23 0.24 0.21 0.27 0.42 0.19 0.16 0.3 0.11 0.18 0.16 0.17 0.30 0.24 0.21 0.21 0.28 0.18 0.21 0.42
8.57 6.24 1.34 3.52 4.49 3.61 16.94 2.55 7.69 2.67 6.24 2.1 1.39 2.29 2.67 0.94 2.16 14.34 0.54 2.02 2.16 0.72 4.34 2.10 1.54 1.39 2.32 2.55 7.69 2.67
and relative concentrations can be obtained. The contents of amino acids in measured samples are calculated and listed in Table 2. The sample information listed in the second column containing the regions, numbers of the sample and the sampling positions in leaves. For example, R1-2-U means the tobacco sample was numbered 2 and collected from the upper parts of the leaves from region 1. As shown in Table 2, Asn is the richest amino acid component in the burley tobacco sample. The result is in agreement with that reported in the literature [33]. Comparing the content of amino acids in different regions, the total content of all 17 kinds of amino acids in R1 is higher than the other two regions. It can be inferred that the content of free amino acids in tobacco leaves are significantly influenced by growth environment [34, 35]. Comparing the content of amino acids in different sampling positions, it can be observed that the content of amino acids in the upper parts of leaves is higher than that in the middle parts, which will be helpful to improve the utilization rate of amino acids in tobacco leaves.
overlapping signals in measured chromatographic data, NNIA was employed for extracting the information of amino acids in tobacco leaves. The results showed that both the chromatogram and the relative concentration of each amino acid can be obtained by using the proposed method. With five repetitions of three spiked sample, the recoveries of the spiked samples were range from 94.62 to 108.21%. Then by analyzing the 17 amino acids in 30 tobacco leaf samples, the conclusions can be drawn that Asp, Asn, and Pro are the main components of burley tobacco leaves. The content of all 17 kinds of amino acids in upper parts of leaves from region 1 is higher than others.
4 Conclusion
5 References
A method was developed and used for quantifying the amount of amino acids in tobacco leaves. Considering the C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
This study was supported by National Natural Science Foundation of China (Nos. 21175074 and 21475068). Thanks also go to Beijing Third Class Tobacco Supervision Station for providing the samples and the experimental measurements. The authors have declared no conflict of interest.
[1] Zhang, J. J., Zhao, C. X., Chang, Y. W., Zhao, Y. N., Li, Q. H., Lu, X., Xu, G. W., J. Sep. Sci. 2013, 36, 2686–2877.
www.jss-journal.com
2058
Y. Zeng et al.
J. Sep. Sci. 2015, 38, 2053–2058
[2] Gwatidzo, L., Botha, B. M., McCrindle, R. I., Food Chem. 2013, 141, 2163–2169.
[19] Goodman, K. J., Brenna, J. T., Anal. Chem. 1994, 66, 1294–1301.
[3] Avino, P., Campanella, L., Mario, V. R., J. Sep. Sci. 2003, 26, 392–396.
[20] Layer, E., Tomczyk, K., Signal Transforms in Dynamic Measurements, Springer International Publishing, Berlin 2015, 16, pp. 169–188.
[4] Pereira, D. M., Valentao, P., Teixeira, N., Andrade, P. B., Food Chem. 2013, 141, 2412–2417.
[21] Maeder, M., Anal. Chem. 1987, 59, 527–530.
[5] Chiu, T.-C., Anal. Bioanal. Chem. 2013, 405, 7919– 7930.
[22] Manne, R., Shen, H. L., Liang, Y. Z., Chemom. Intell. Lab. Syst. 1999, 45, 171–176.
[6] Friedman, M., J. Arg. Food Chem. 2004, 52, 385– 406.
[23] Liang, Y. Z., Kvalheim, O. M., Keller, H. R., Massart, D. L., Kiechle, P., Erni, F., Anal. Chem. 1992, 64, 946–953.
[7] Sun, S. W., Lin, Y. C., Weng, Y. M., Chen, M. J., J. Food Compos. Anal. 2006, 19, 112–117.
[24] Zhao, Z. M., Malinowski, E. R., Anal. Chem. 1999, 71, 602–608.
[8] Zuraini, A., Somchit, M. N., Solihah, M. H., Goh, Y. M., Arifah, A. K., Zakaria, M. S., Somchit, N., Rajion, M. A., Zakaria, Z. A., Mat Jais A. M., Food Chem. 2006, 97, 674– 678.
¨ [25] Schworer1, R., Rothl, J., J. Appl. Phys. 1995, 77, 3812– 3817.
¨ [9] Nilsson, M., Holst, J. J., Bjorck, I. M. E., Am. J. Clin. Nutr. 2007, 85, 996–1004.
[27] Navea, S., de Juan, A., Tauler, R., Anal. Chim. Acta 2001, 446, 185–195.
[10] Lund, E., Thomsen, J., Brunfeldt, K., J. Chromatogr. 1977, 130, 51–54.
[28] Hantao, L. W., Aleme, H. G., Pedroso, M. P., Sabin, G. P., Poppi, R. J., Augusto, F., Anal. Chim. Acta 2012, 731, 11–23.
[11] Kaspar, H., Dettmer, K., Gronwald, W., Oefner, P. J., Anal. Bio. Chem. 2009, 393, 445–452. [12] Durham, B., Geren, C. R., Anal. Biochem. 1981, 116, 331– 334. [13] Alonzo, N., Hirs, C. H. W., Anal. Biochem. 1968, 23, 272– 288. [14] Zhang, W., Cao, Y. Z., Liu, Y. M., Cui, Chemistry 2002, 6, 418–421. [15] Han, J., Li, P., Cai, W. S., Shao, X. G., J. Sep. Sci. 2014, 37, 2126–2130. [16] Zhang, Q. L., Ni, Y. N., Kokot, S., J. Pharm. Biomed. Anal. 2010, 52, 280–288.
˜ ˜ [26] Gusmao, R., Arino, C., D´ıaz-Cruz, J. M., Esteban, M., Anal. Bioanal. Chem. 2009, 394, 1137–1145.
[29] Shao, X. G., Yu, Z. L., Sun, L., TrAC, Trends Anal. Chem. 2003, 22, 59–69. [30] Shao, X. G., Wang, G. Q., Wang, S. F., Su, Q. D., Anal. Chem. 2004, 76, 5143–5148. [31] Liu, Z. C., Cai, W. S., Shao, X. G., J. Chromatogr. A 2009, 1216, 1469–1475. [32] Shao, X. G., Liu, Z. C., Cai, W. S., TrAC, Trends Anal. Chem. 2009, 28, 1312–1321. [33] Li, G., Wu, D., Xie, W. Y., Sha, Y. F., Lin, H. Q., Liu, B. Z., J. Chromatogr. A 2013, 1296, 243–247.
[17] Ni, Y. N., Wang, Y., S. Kokot, Talanta 2009, 78, 432–441.
[34] Wilkinson, R. E., Kasperbauer, M. J., Young, C. T., J. Agric. Food Chem. 1981, 29, 658–660.
[18] Miao, L. N., Cai, W. S., Shao, X. G., Talanta 2011, 83, 1247–1253.
[35] Zhang, J. S., Jia, C. X., Li, Y. Q., Mao, D. B., Zhang, W. Y., Tob. Sci. Technol. 2004, 205, 26–32.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
