Simultaneous determination of selected flavour enhancers in beef pho broth samples by chemometrics - Assisted spectrophotometry

JOURNAL OF SCIENCE OF HNUE DOI: 10.18173/2354-1059.2016-0062 Natural Sci. 2016, Vol. 61, No. 9, pp. 104-112 This paper is available online at 104 SIMULTANEOUS DETERMINATION OF SELECTED FLAVOUR ENHANCERS IN BEEF PHO BROTH SAMPLES BY CHEMOMETRICS - ASSISTED SPECTROPHOTOMETRY Nguyen Thi Hop and Ta Thi Thao Department of Analytical Chemistry, Faculty of Chemistry, VNU Uninersity of Science Abstract. The chemometrics-assisted spectrophotometry has been studied for simultaneous determination of monosodium glutamate (MSG), disodium guanylate (DSG) and disodium inosinate (DSI) in beef Pho broth. The effect of protein was eliminated by precipitation with ammonium sulfate. Due to the overlap of spectra of three analytes in UV range (214 nm - 280 nm), the multiple linear regressions including classical least square (CLS), inverse least square (CLS), principal component regression (PCR) and partial least square (PLS) had been applied. The experimental results showed that PCR with 4 principal components was the best algorithm which gave the lowest error (less than 2.5%). LOD of MSG, DSG and DSI were 44 ppm, 0.3 ppm and 0.2 ppm, respectively. The recovery and repeatability of the method were in the ranges of AOAC criteria. The procedure has been sucessfully applied to determine concentration of MSG, DSG and DSI in three beef Pho broth samples collected from street restaurants in Hanoi. The concentrations of MSG in samples were 5 to 8 times higher than the maximum concentration limit (10 g/kg), whereas those of DSG and DSI were below the maximum concentration limit (500 mg/kg). Keywords: Monosodium glutamate, disodium guanylate, disodium inosinate, multiple linear regressions, beef pho broth. 1. Introduction Nowadays, monosodium glutamate (MSG), disodium guanylate (DSG) and disodium inosinate (DSI) are the most flavor enhancers produced to make food more tasteful. Because of their synergistic effect, three of them are often added together in food products, which allows decreasing amount of them in food but still ensures the taste, hence, reduces the cost for them. Nevertheless, the improper use of these food additives as regulated may lead to harmful effect on human health. An intake of high concentrations of MSG causes the occurrence of neurological diseases including Parkinson and Alzheimer [1], and since MSG could damage the nervous system, it is not permitted in foods for infants and young children [2]. As salts of guanosine 5'- monophosphate (GMP) and inosine 5'-monophosphate (IMP), DSG and DSI metabolize to GMP and IMP purine nucleotides synthesized in the human body and play diverse roles in the cellular metabolism [3]. Lo´pez Navarro et al. suggested that dietary nucleotides are energetically advantageous to fulfill the liver need for nucleotides [4]. Received May 10, 2016. Accepted October 10, 2016. Contact Ta Thi Thao, e-mail address: tathithao@hus.edu.vn Simultaneous determination of selected flavour enhancers in beef pho broth samples 105 However, people with high levels of uric acid in blood and urine must avoid food with this kind of compounds because the degradation of purine nucleotides leads to the formation of uric acid [3]. Therefore, concentrations of MSG, DSG and DSI in foodstuffs must be controlled. Since 2008, the European Commission has regulated the maximum level of MSG in food expressed as glutamic acid at 10 g/kg, individually or in combination; and for DSG, DSI expressed as guanylic acid [5] at 500 mg/kg, relatively. Up to now, there have been various analytical methods to determine solely MSG in foodstuffs, from high performance liquid chromatography (HPLC) with the help of a U.V visible detector [6] or HPLC in combination with fluorescence detection [7], even with the use of biosensor [8]. SDG and SDI are also simultaneously analyzed in food products by methods such as HPLC [9] or UV-spectrophotometry with the help of multiple linear regressions [10]. However, only few founded studies for simultaneous determination of these MSG, SDG and SDI in foodstuffs used spectrophotometric data combined with chemo-metrics applied by Carolina C. Acebal et al. in Argentina [11] and Durán-Merás I et al. in Spain [12]. And no other studies have been conducted for the simultaneous determination of DSG, DSI and MSG in popular food subjects in Vietnam yet. Hence, in this study, a spectrophotometric method combined with some multiple linear regressions (MLRs) including the Classical Least Squares (CLS), the Inverse Least Squares (ILS), the Partially Least Squares Regression (PLS) and the Principal Component Regression (PCR) has been developed for simultaneous determination of contents of MSG, DSG and DSI in some broth samples of a very popular and favorite dish in Vietnamese – beef Pho. 2. Content 2.1. Experiment 2.1.1. Reagents and aparatus All reagents used in this work were analytical reagents. Aqueous stock standard solutions of 10.0 mg/mL MSG, 0.500 mg/mL SDG, 0.500 mg/mL SDI were prepared by dissolving appropriate amount of their solid forms (Sigma-Aldrich, Germany) in distilled water. Besides, solutions of 1.00 mg/mL Bovine serum albumine (BSA), 1% CuSO4, 2% Na3C6H5O7, a solution of 2% Na2CO3 and 0.10 N NaOH and Folin-Ciocalteu reagent were prepared for protein determination by Lowry’s method. Three analytical samples of beef Pho broth were collected from three sellers in Hanoi. Beef broth was cooked in the following steps: simmering beef bones for 10 hours; then, the broth was decanted and the bones were discarded. All samples were contained in plastic bottles and stored at 4 o C for use within one week, and at -18 o C for longer time of use. Absorption spectra were obtained by using UV-VIS Spectrophotometer 1650 PC - Shimazu (Japan). 2.1.2. Sample preparations In this work, beef Pho broth was the matrix where both the analytes of MSG, DSG and DSI and the major interferent, protein, are in soluble forms. Thus, protein precipitation via salting out using ammonium sulfate was chosen for sample treatment. Nguyen Thi Hop and Ta Thi Thao 106 Each of control samples was prepared by adding 1.0 mL of the matrix, 10.00 mL of buffer at pH of 8 and different volume of stock solutions of MSG, DSG and DSI into a 50.0 mL volumetric flask so that their known concentrations were as shown in table after diluting with double distilled water. 2.1.3. Methodology MSG, DSG and DSI absorb light in ultraviolet range with overlapping bands at pH of 8. Hence, ultra-violet molecular absorption spectrometry combined with multiple linear regressions including CLS, ILS, PLS and PCR were used as methods for simultaneous determination of these three flavor enhancers. Besides, protein determination by Lowry’s method was used to assess the efficiency of protein prepitation using ammonium sulphate. Data were processed with the aid of computer softwares such as Minitab 16 Statistical Software, and MATLAB R2012b. Determination of each analyte in analytical samples was carried out in the following procedures: add 3.50 g (NH4)2SO4 and 5.00 mL sample to a centrifuge tube and shake well; centrifug at 6000 rpm in 30 minnutes; filter with filtering paper to obtain the filtrate; then, add 1.50 mL filtrate and 10.00 mL buffer (pH 8) to a 50.0 mL volumetric flask; dilute with double distilled water; measure absorbance from 214 nm to 280 nm and finally, process the obtained data by PCR model with 4 PCs which gave out concentrations of 3 analytes. 2.2. Results and discussions 2.2.1. Spectra of MSG, DSG, DSI and their mixture Figure 1. Absorption spectra of MSG, DSG, DSI and their mixture at pH 8 Figure 1 shows that at pH 8, all three of analytes had the maximum absorption of wave- length at 206 nm, while DSG and DSI had two more maximum absorption ones at 252 nm and 250 nm, respectively. The frequency of the most overlapping spectra from 214 nm to 280 nm was selected for later determination of analytes in analytical samples. 2.2.2. Influence of protein and efficiency of protein precipitation by ammonium sulphate Determination of protein concentrations in the matrix prior and after sample treatment was used to quantitatively assess influence of protein concentrations and efficiency of protein precipitation by ammonium sulphate to minimize the influence. Simultaneous determination of selected flavour enhancers in beef pho broth samples 107 Figure 2 shows that absorbances of the matrix prior to sample treatment were much higher than those of the matrix after sample treatment in the recorded region from 200 nm to 232 nm. Efficiency of the protein precipitation method by salting out using ammonium sulphate was 90.9% (Figure 2). Figure 2. Absorption spectra of matrix before (1-3) and after (4-6) sample treatment However, high concentrations of proteins (from 24.0 ppm to 95.8 ppm) had great influence on simultaneous determination of MSG, DSG and DSI (Two-way ANOVA, Pvalue < 0.05). 2.2.3. Effect of pH The optimal pH condition was selected based on two criteria: high absorbances and the best collinearity of the three flavour enhancers in a wave-length range between 200 nm and 300 nm. The investigated pH range was from 2 to 11. In order to obviously illustrate the dependence of absorbances of MSG, DSG, and DSI on pH, absorbances at two wave-lengths of 214 nm (instead of 206 nm because absorbance in the wavelength was shorter than 214 nm, and unstable due to adjacency to the near-UV region) and 250 nm were picked up from data sets and graphed in Figures 3. Figure 3. Dependence of absorbances of MSG, DSG and DSI on pH at 214 nm and 250 nm Nguyen Thi Hop and Ta Thi Thao 108 Figures 3 led to two following findings: - At 214 nm, absorbance of three analytes of interest, in general speaking, tended to increase coresponding to increase pH values (except the case that absorbance of DSI decreased when pH changed from 10 to 11). The same tendency was founded for absorbance of their mixtures. - At 250 nm, with increasing values of pH from 2 to 11, absorbance of MSG decreased slightly, however, in fact, could be considered negligible. That of DSI also went up insignificantly and that of DSG increased at from 2 to 8 pH, then went down to pH of 11. The trend in independence of absorbance of DSI on pH was also in absorbance of the mixture. Based on these two findings, pH values of 8, 9, 10 and 11 were chosen for further consideration of collinearity in absorbances of MSG, DSG and DSI. This property was investigated based on the relative error between absorbances of the mixture containing three analytes of interest, and the sum of their individual absorbances. It turned out that at pH 8, absorbances of three analytes were the most collinear in 214 nm - 280 nm range (relative errors within 1.2 % at pH 8, some higher ones up to 4.5 % at pH 9; and 5.4 % at pH 10, relatively, and even more higher at pH 11 with the maximum value of 16.8 %). Hence, pH 8 was the optimal pH condition for simultaneous determination of MSG, DSG and DSI in their mixtures. 2.2.4. Validation of the analytical method * Assessment of the utility of MLRs The utility of the MLRs was validated based on relative errors between known concentrations, andcalculated ones of three analytes in 3 control samples from the MLRs which were established based on absorbances of 25 standard solutions containing different known ones of three analytes at 25 selected wave-lengths in 214 nm - 280 nm range. Table 1. Collected data on concentrations of analytes in control samples and relative errors calculated from MLRs Analyte Known conc. (ppm) Relative error (%) CLS ILS PLS PCR 200 3.3 31.3 6.1 2.4 MSG 400 1.7 -15.4 3.4 1.0 600 -0.8 -17.1 -1.2 -1.3 8 -0.4 1.8 0.1 -0.6 DSG 10 0.2 -7.3 0.7 0.0 6 -1.1 -10.1 -1.3 -1.4 12 -1.0 -2.1 -1.5 -1.0 DSI 10 -0.4 2.8 -1.1 -0.5 8 1.3 1.2 1.4 1.1 Simultaneous determination of selected flavour enhancers in beef pho broth samples 109 Results in Table 1 proved the utility of MLRs of CLS, PLS and PCR with 4 PCs in simultaneous determination of MSG, DSG and DSI in mixtures. The relative errors between known and calculated concentrations of these three analytes were smaller than 5%, the accepted error in analytical chemistry, except only one of 6.1 % calculated from PLS in the case of low concentration of MSG (200.0 ppm). High relative errors (up to 31.3 %) in some samples calculated from ILS could be explained due to failure in selecting the least number of wave- lengths at which signals of all analytes contributed to the total one of the mixture. The least errors calculated from PCR resulted from the fact that using this model could minimize influence of the interferent - protein in the matrix. The selection of 4 PCs gave the smallest relative errors compared to other numbers of PCs from 1 to 25 (because 25 standard solutions were used to establish MLRs). Hence, PCR model with 4 PCs was selected for simultaneous determination of MSG, DSG and DSI in three analytical samples. * Limit of detection (LOD), limit of quantitation (LOQ), calibration sensitivity (SEN), analytical sensitivity (γ), recovery and repeatability The LOD and LOQ, SEN and γ in multiple linear regressions are determined according to the following equations: |||| ||||3 bk LOD   , |||| ||||10 bk LOQ   , SEN = 1 || ||bk and γ = 1 || || , where ||Ɛ|| is the normalized vector of Euclidian vector Z, a matrix of signals of replicate blank samples; ||bk|| is the normalized vector of of Euclidian vector bk, the vector of regression co- eficients corresponding to k-constituents of analytes, determined in methods of MLRs. The reciprocal of γ, i.e., γ-1 is the smallest concentration which is significant when there are no systematic errors in the analytical method. Table 2. Collected data on figures of merit for simultaneous determination of analytes Figures of merit MSG DSG DSI LOD (ppm) 44 0.3 0.2 LOQ (ppm) 150 1.0 0.7 SEN 1800 11 10 γ-1 (ppm) 4.3x10-6 6.9x10-4 7.9x10-4 The analytical procedure for determination of recoveries in spiked samples is shown in Figure 4. Recoveries of individual MSG, DSG and DSI at 3 concentration levels were between 88.7 % and 98.4%, allowable as regulated by AOAC [13]. Nguyen Thi Hop and Ta Thi Thao 110 2.2.5. Simultaneous determination of MSG, DSG and DSI in real samples Concentrations of MSG, DSG and DSI in three analytical samples S1, S2 and S3 were determined using the procedure as mentioned in section 2.1.3. Each sample was conducted three times and the results obtained by running PCR (4 PCs) with relative standard derivations shown in Table 3. The obtained relative standard deviations of repeatability were within 1.6 %, allowable as regulated by AOAC [11]. Based on results in Table 3 and values of LOD and LOQ of MSG, DSG and DSI presented previously in Table 2, concentrations of three analytes in three analytical samples were calculated, and reported in Table 4. Figure 4. Spike procedure to calculate recoveries of MSG, DSG and DSI Table 3. Collected data on mean concentrations of analytes in analytical samples obtained from PCR (4 PCs) and relative standard deviations Sample Mean con. (ppm) RSD (%) MSG DSG DSI MSG DSG DSI S1 1400 -1.8 9.9 0.3 -0.9 0.3 S2 1300 -4.7 12 0.3 -0.9 0.1 S3 1900 2.0 4.4 0.0 1.6 0.3 Simultaneous determination of selected flavour enhancers in beef pho broth samples 111 Table 4. Collected data on concentrations of analytes in analytical samples Sample Conc. MSG (g/kg) DSG (mg/kg) DSI (mg/kg) S1 57 not detected 410 S2 54 not detected 490 S3 80 83 180 Table 4 indicates that concentrations of MSG in these analytical samples were much higher (5 to 8 times) than the accepted maximum level regulated by EC (10 g/kg) and concentrations of both DSG and DSI in three analytical samples were lower than the accepted maximum level (500 mg/kg). 3. Conclusion A UV-spectrophotometric method of analysis was successfully researched and developed for the simultaneous determination of MSG, DSG and DSI in beef Pho broth samples. Multi-variate calibration models were built from the spectral data matrices on the basis of CLS, ILS, PLS and PCR regressions. Verification of the calibrations were conducted with the help of a set of mixtures containing the three compounds and the results showed that ILS model performed poorly in comparison with PCR, PLS and CLS ones, and the PCR with 4 PCs was the best model. Thus, PCR with 4 PCs was selected, combined with spectrophotometric data, for simultaneous determination of the three flavour enhancers in three beef Pho broth samples. REFERENCES [1] Arruda N. J., Filho J. L., Montenegro M. C., Arau´jo A. N. and Silva V. L., 2003. Simple and inexpensive flow L-glutamate determination using pumpkin tissue, Journal of Agricultural and Food Chemistry, 51, 6945-6948. [2] European Parliament and Council Directive 95/2/EC of 20 February 1995 on food additives other than colours and sweeteners, achieved on May 19 th , 2015 from: [3] Devlin T., 1997. Purine and pyrimidine nucleotide. Wiley-Liss, New York. [4] Lo´pez Navarro A. T., Bueno J. D., Gil A. and Sanchez P. A., 1996. Morphological changes in hepatocytes of rats deprived of dietary Nucleotides. British Journal of Nutrition, 76(4), 579-589. [5] Regulation (EC) No. 1333/2008 of the European Parliament and of the Council of 16 December 2008 on food additives, achieved on May 19th 2015 from: lex.europa.eu/legalcontent/EN/TXT/PDF/?uri=CELEX:32008R1333&from=EN [6] Rodriguez M. S., Gonzalez M. E. and Centurion M. E., 2003. Determination of Monoglutamate in Meat Products. The Journal of Argentine Chemical Society, 91(4), 41-45. Nguyen Thi Hop and Ta Thi Thao 112 [7] Garcla S. F., Aguilar G. A., 1992. Liquid chromatographic and spectrofluorimetric determination of aspartame and glutamate in foodstuffs following fluorescamine fluorigenic labeling. Analytica Chimica Acta, 270, 45-53. [8] Basu A. K., Chattopadhyay P., Roychudhuriand U. and Chakraborty R., 2006. Development of biosensor based on immobilized L-glutamate oxidase for determination of monosodium glutamate in food. Indian Journal of Experimental Biology, 44(5), 392-398. [9] Xiang L. X., Yang D. X., Wei Y. M., and Yan B. X., 2011. Quantitative Determination of Disodium 5′-Inosinate and Disodium 5′-Guanylate in Chicken Extract by HPLC. Journal of Beijing Technology and Business University, 6. [10] Shan Y. E., 2010. Simultaneous UV-Spectrophotometric Determination of Disodium 5′- inosinate and Disodium 5′-guanylate in Tasty Flavourings with Calibration by the Method of Principal Component Regression. Physical Testing and Chemical Analysis, 7, 193-206. [11] Acebal C. C., Lista A. G. and Fernandez B. B. S., 2008. Simultaneous determination offlavor enhancers in stock cube samples by using spectrophotometric data and multivariate calibration. Food Chemistry, 106, 811-815. [12] Durán M. I., Muńoz P. A., Espinosa M. A. and Salinas F., 1993. Multicomponent determination of flavour enhancers in food preparations bypartial least squares and principal component regression modelling of spectrophotometric data. Analyst, 118(7), 807-813. [13] AOAC Guidelines for Single Laboratory Validation of Chemical Methods for Dietary Supplements and Botanicals, achieved on May 19th. 2015, from: Dietary_Supplements.pdf