- Email: [email protected]
Ultrasound-assisted extraction of amino acids from grapes
Ultrasonics Sonochemistry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultson
Ultrasound-assisted extraction of amino acids from grapes Ceferino Carrera a, Ana Ruiz-Rodríguez b, Miguel Palma b,⇑, Carmelo G. Barroso b a b
Andalusian Center for Wine Research, Agrifood Campus of International Excellence (ceiA3), University of Cádiz, Campus de Puerto Real, 11510 Puerto Real, Cádiz, Spain Department of Analytical Chemistry, Faculty of Sciences, Agrifood Campus of International Excellence (ceiA3), University of Cádiz, Spain
a r t i c l e
i n f o
Article history: Received 11 November 2013 Received in revised form 7 February 2014 Accepted 26 May 2014 Available online xxxx Keywords: Amino acids Ultrasound-assisted extraction Grapes
a b s t r a c t Recent cultivar techniques on vineyards can have a marked influence on the final nitrogen content of grapes, specifically individual amino acid contents. Furthermore, individual amino acid contents in grapes are related to the final aromatic composition of wines. A new ultrasound-assisted method for the extraction of amino acids from grapes has been developed. Several extraction variables, including solvent (water/ethanol mixtures), solvent pH (2–7), temperature (10–70 °C), ultrasonic power (20–70%) and ultrasonic frequency (0.2–1.0 s1), were optimized to guarantee full recovery of the amino acids from grapes. An experimental design was employed to optimize the extraction parameters. The surface response methodology was used to evaluate the effects of the extraction variables. The analytical properties of the new method were established, including limit of detection (average value 1.4 mmol kg1), limit of quantification (average value 2.6 mmol kg1), repeatability (average RSD = 12.9%) and reproducibility (average RSD = 15.7%). Finally, the new method was applied to three cultivars of white grape throughout the ripening period. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Total free amino acid contents in grapes and also the amino acid profiles can vary dramatically depending on the grape cultivar [1], cultivar practices [2], ripening degree [3], vine nutrient/water/fungicide management and health status, including fungal infections [4]. Additionally, grape cultivar using different rootstocks usually present significant differences for the final composition of grapes. It has been demonstrated that different rootstocks for the same grape variety, for example, produce different amino acid profiles in the final grapes [3]. The Cabernet Sauvignon grape variety with a high vigor rootstock, produces much higher levels of most amino acids in the grapes than with a low vigor rootstock [1]. Cultivar practices, such as the use of cover crops, have different effects on amino acid levels depending on the variety. For example, changes were not observed for Cabernet Sauvignon [1] but a dramatic reduction in amino acid levels was observed for the Pinot Noir grape variety on using cover crops [2]. Vine health status and fungicide application usually affect the levels of amino acids in grapes. The application of fungicide close to the harvest date leads to lower
⇑ Corresponding author. Address: Department of Analytical Chemistry, Faculty of Sciences, Agrifood Campus of International Excellence (ceiA3), University of Cádiz, 11510 Puerto Real, Spain. Tel.: +34 956016775; fax: +34 956016590. E-mail address: [email protected] (M. Palma).
levels for all amino acids [4] and this makes the alcoholic fermentation process more difficult. It has been proved that there is a direct relationship between the amino acid levels in grape must and their consumption during the first half of alcoholic fermentation [5]. This means that the starting levels of nitrogen compounds, specifically amino acids, will determine the evolution of the first step in the alcoholic fermentation, therefore affecting the final composition of wines. It has been demonstrated in several previous studies that there is a correlation between amino acid contents in grapes/must and the aroma of the final wines, specifically for young wines [6]. Free nitrogen levels, including amino acids, are also associated with some problems in the winemaking process. Insufficient nitrogen levels can stop fermentation and this leads to the production of numerous secondary metabolites that are not of interest for wine production [7]. Furthermore, some amino acids are directly related to certain volatile compounds that contribute to the wine aroma, including fusel oils, aldehydes and esters [8]. Wines with very high levels of nitrogen compounds can also have higher microbiological instability, including those processes related to biogenic amines and ethyl carbamate generation [9]. Given the information outlined above, it is advantageous to determine amino acids in grapes during ripening in an effort to identify the effects of cultivar practices and cultivar conditions. It is also of interest to determine amino acid levels in grapes just prior to alcoholic fermentation in order to establish more suitable winemaking practices.
http://dx.doi.org/10.1016/j.ultsonch.2014.05.021 1350-4177/Ó 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
2
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
There are several methods for the determination of free amino nitrogen [10] and amino acids in must. Most of these approaches involve chromatographic techniques [11], but colorimetric methods are also employed [12]. In many cases, the final determination requires some kind of derivatization step to increase the analytical signal, including by fluorescence detection systems [11]. However, the problem of the determination of amino acids in grapes has not yet been solved satisfactorily. Differences in individual free amino acid profiles and concentrations have been observed between the different sample preparation methods usually applied, i.e. determination in grape juice or chemically extracted grapes, before amino acid determination [1]. There are three main reasons for these differences. Firstly, extraction – rather than grape juice preparation – will include amino acids from grape skins in the final level. Proline, for example, has been reported to be found in grape skins [1]. Secondly, extraction usually leads to dilution of the sample due to the presence of extraction solvents and this means that some levels could go below the limits of quantification. Finally, quantification in grape juice will be expressed as mg per liter whilst quantification in grapes will be expressed as mg per kg. The relationship between kg of grape and liter of grape juice is dependent on the winemaking process. Additionally, during the red winemaking process, grape skins are in contact with grape must for several days and amino acids will therefore be extracted from the skins to some extent. As a consequence, determination of the total amount of amino acids in grapes would be more useful than the amount in the grape juice alone. The aim of the work described here was to develop and validate a new ultrasound-assisted extraction method for amino acids in grapes. The overall goal was to provide a rapid and reliable method for winemakers to determine amino acids in grapes during grape ripening and harvest. Ultrasound assisted extraction was used because of this technique usually produce faster extraction methods, lower solvent volumes and higher extraction yields [13]. 2. Materials and methods 2.1. Chemicals and solvents Ethanol and methanol (Panreac, Barcelona, Spain) were HPLC grade. Ultra pure water was supplied by a Milli-Q water purification system from Millipore (Bedford, MA, USA). 2.2. Grape samples The white grape (var. Verdejo) was employed in the development of the ultrasound-assisted extraction method. Samples were obtained from local vineyards in the Jerez region (Spain). The full berry, i.e. skin, pulp and seeds, was studied. The berries were triturated with a conventional beater until a homogeneous sample was obtained for analysis. The triturated sample was stored in a freezer at 20 °C prior to analysis. 2.3. Extraction procedure The extraction of amino acids from grapes by means of ultrasound was performed by employing water/ethanol mixtures as solvent. The effects of the extraction solvent (0–25% EtOH in water), temperature (10–70 °C), output amplitude of the nominal amplitude of the transducer (30–70%), duty cycle (0.2–0.7 s), pH of the extraction solvent (2–7) and the sample/solvent ratio (1 g 10 mL1–1 g 20 mL1) were studied. Ultrasonic irradiation was carried out using a UP200S sonifier (200 W, 24 kHz) (Hielscher Ultrasonics, Teltow, Germany), with
the sample immersed in a water bath coupled to a temperature controller (Frigiterm, J.P. Selecta, Barcelona, Spain). 2.4. Experimental design for the evaluation of the effects of extraction variables Optimization of extraction variables was performed using the Box–Behnken statistical methodology by considering the total amount of amino acids extracted. The results for the 54 extractions carried out in duplicate for the seven extraction variables (each variable has three levels: low, medium and high) including 6 center points are shown in Table 1 along with the respective responses. The results for the total amount of amino acids (mmol kg1 of sample) determined by HPLC were used as the response variable. The responses obtained from the various extractions were entered into to a second-order polynomial equation into which each of the various parameters was introduced. The polynomial equation is as follows:
Y ¼ b0 þ b1X1 þ b2X2 þ b3X3 þ b4X4 þ b5X5 þ b6X6 þ b7X7 þ b12X1X2 þ b13X1X3 þ b14X1X4 þ b15X1X5 þ b16X1X6 þ b23X2X3 þ b24X2X4 þ b25X2X5 þ b26X2X6 þ b34X3X4 þ b35X3X5 þ b36X3X6 þ b45X4X5 þ b46X4X6 þ b56X5X6 þ b11X12 þ b22X22 þ b33X32 þ b44X42 þ b55X52 þ b66X62 In this equation Y is the aforementioned response, b0 is the ordinate at the origin; X1 [percentage of EtOH in the extraction solvent], X2 [temperature (°C)], X3 [ultrasound amplitude], X4 [duty cycle], X5 [pH], X6 [ratio solid sample (g)/extraction volume (mL)] are the independent variables; bi are the linear coefficients; bij are the cross product coefficients and bii are the quadratic coefficients. The analysis of data for the Box–Behnken design was carried out using Unscrambler X (Camo, No). This software was used to estimate the effects of the variables on the final response, the variance analysis, the second order mathematical model, the optimum levels of the significant variables and the surface graphs. 2.5. Amino acids determination The AccQ-Tag method was used for derivatization and chromatographic determination. The AccQTag Reagent Kit was purchased from Waters (Milford, Massachusetts, USA). The reagent kit consists of Waters AccQFluor Borate Buffer, Waters AccQFluor Reagent Powder (6-aminoquinolyl-N-hydroxysuccinimidyl carbamate – AQC), Waters AccQFluor Reagent Diluent, Waters AccQTag Amino Acid Analysing Column (Nova-Pak C18, 4 lm, 150 3.9 mm) and Waters Amino Acid Hydrolysate Standard (each ampoule contains a 2.5 mM mixture of the 17 amino acids with the exception of cystine – 1.25 mM. The following amino acids were found and quantified in the samples: aspartic acid, serine, glutamic acid, glycine, histidine, arginine, threonine, alanine, proline, cysteine, tyrosine, valine, methionine, lysine, isoleucine and leucine. 3. Results and discussion 3.1. Optimization of the extraction method A Box–Behnken design was used to optimize the extraction conditions for amino acids from homogenized grape samples. Six different extraction variables were studied in the following ranges: ethanol in water between 0% and 50%, temperature values between
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
3
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx Table 1 Results from the experimental design. Extraction yield (mmol of total amino acids kg1 of grape) for the evaluation of extraction variables in the experimental design. Experiment
% EtOH in water
Temperature
Amplitude
Cycle
pH
Ratio
mmol kg1 obtained
mmol kg1 estimated by the model
Difference real vs estimated (%)
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 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 cp01 cp02 cp03 cp04 cp05 cp06
25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 0 50 0 50 0 50 0 50 25 25 25 25 25 25 25 25 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 25 25 25 25 25 25
40 40 40 40 40 40 40 40 10 70 10 70 10 70 10 70 40 40 40 40 40 40 40 40 10 70 10 70 10 70 10 70 10 10 70 70 10 10 70 70 40 40 40 40 40 40 40 40 40 40 40 40 40 40
30 70 30 70 30 70 30 70 50 50 50 50 50 50 50 50 30 30 70 70 30 30 70 70 30 30 70 70 30 30 70 70 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 2 2 7 7 2 2 7 7 2 2 2 2 7 7 7 7 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 2 2 7 7 2 2 7 7 4.5 4.5 4.5 4.5 4.5 4.5
3 3 7 7 3 3 7 7 3 3 3 3 7 7 7 7 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 3 3 3 3 7 7 7 7 5 5 5 5 5 5 5 5 5 5 5 5 5 5
10 10 10 10 20 20 20 20 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 10 10 10 10 20 20 20 20 15 15 15 15 15 15 15 15 10 10 10 10 20 20 20 20 15 15 15 15 15 15
14.4 14.2 15.9 14.3 12.8 12.6 13.7 14.5 11.7 12.7 12.5 15.8 11.0 13.9 11.5 14.1 18.7 7.0 13.1 8.0 17.6 6.3 13.5 12.4 13.4 18.1 12.1 13.1 14.0 16.1 10.4 19.5 12.3 6.9 13.3 14.1 14.1 6.4 16.8 13.2 10.1 8.4 15.0 9.4 14.8 14.7 10.2 9.9 11.9 13.2 12.8 13.4 12.6 13.6
13.5 11.7 13.8 11.7 12.5 12.8 13.5 13.6 10.6 13.5 11.2 15.1 11.6 15.1 10.6 15.2 18.3 9.7 11.6 11.3 16.9 8.4 13.5 13.4 12.9 14.4 10.1 13.3 10.8 15.1 10.3 16.2 13.6 8.3 14.4 14.2 15.5 6.8 16.9 13.4 13.7 7.5 17.8 11.7 16.2 13.5 12.7 10.0 12.9 12.9 12.9 12.9 12.9 12.9
6 18 13 18 2 2 2 7 9 6 10 4 6 9 7 7 2 38 11 42 4 33 0 8 4 20 16 1 22 6 2 17 11 21 8 1 10 7 1 2 36 11 19 25 9 9 24 1 8 2 1 4 2 5
10 and 70 °C, ultrasound amplitude between 30% and 70%, ultrasound duty cycle between 0.2 and 0.7 s1, solvent pH between 2 and 7 and sample-to-solvent ratio between 0.1 and 0.05 g mL1 (Table 1). Extraction time was fixed at 5 min. All experiments (a total of 54) were run in duplicate for the extraction. The recovery of amino acids was correlated with experimental conditions by a second-order polynomial equation. The resulting coefficients for the second-order polynomial equation from the Box–Behnken design and their significance (p-value) are presented in Table 2. Data related to the fitting properties of the resulting model, expressed as percentage difference between the experimental values for the amino acid recoveries and the calculated ones, are presented in Table 1. The resulting average difference was 10%, ranging from 0% to 42%, with only nine experiments showing differences greater than 20% between experimental and
calculated values, i.e. 92% of the experimental data were fitted by the model with an error of less than 20%. It has to be noted that most experiments showing high relative differences between real and predicted values were those showing low level for total amino acids, experiments 18, 20, 22, 34, 41, 44 and 47. For those cases, relative differences are obviously high. The resulting model therefore explains the recovery results for the extracts based on the experimental conditions used during the extraction process. This means that extraction variables control the final extract composition in terms of amino acid recovery and that the extraction conditions can therefore be managed to optimize the extraction of amino acids from grapes. Analysis of the model clearly shows that, among the linear terms, the most influential variable was the solvent (p-value = 0.00001). The solvent has a negative coefficient, which
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
4
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
Table 2 Coefficients for the second-order polynomial equation from the Box–Behnken design.
B0 Solvent Temp Amplitude Cycle pH Ratio Solvent * temp Solvent * amplitude Temp * amplitude Solvent * cycle Temp * cycle Amplitude * cycle Solvent * pH Temp * pH Amplitude * pH Cycle * pH Solvent * ratio Temp * ratio Amplitude * ratio Cycle * ratio pH * ratio
Coefficient
p-Value
386.60 65.96 55.60 12.83 5.07 7.66 6.15 38.70 63.27 12.37 0.68 7.88 25.19 25.25 5.04 1.76 11.70 25.50 20.52 16.36 57.51 5.39
0.00001 0.00013 0.32180 0.69367 0.55219 0.63305 0.08927 0.00730 0.57932 0.96552 0.72366 0.26239 0.26132 0.74885 0.93692 0.59991 0.25666 0.35966 0.30253 0.01384 0.80872 Fig. 1. Surface plot for the interaction between solvent and temperature on the amino acid recovery.
indicates that the use of a lower proportion of ethanol gives rise to higher levels of amino acids in the extract, i.e. higher recoveries. If results in Table 1 are grouped by the level of ethanol in the solvent, it can be found that the average total amino acid content in the extracts obtained using 0% ethanol was 14.1 mmol kg1, whereas the average total amino acid content in the extracts obtained using 50% ethanol was 9.7 mmol kg1. Therefore decreasing the level of ethanol in the solvent, amino acid recovery can be increase as much as 45%. It can also be seen from the results in Table 2 that temperature has a significant effect on the recovery. The effect of temperature was positive, with higher temperatures leading to a higher recovery of amino acids. The average recovery obtained at 70 °C was 15.0 mmol kg1, whereas the average total amino acid content in the extracts obtained at 10 °C was 11.3 mmol kg1. Analysis of interactions between extraction variables usually shows interesting information. In this case, a significant interaction was found for solvent composition and temperature. The resulting surface plot for this interaction is shown in Fig. 1. It can be seen that there is a very rapid increase in the amino acid recovery as extraction conditions move from low to high temperatures when 50% ethanol was used as the extraction solvent (recoveries changed from around 7.3 mmol kg1 to 12.7 mmol kg1). However, if the same changes in temperature are produced using 100% water as the extraction solvent, the recovery only increased from 14.0 mmol kg1 to 16.0 mmol kg1, indicating a much lower effect due to temperature changes in this case. It means the positive effect of increasing temperature was more significant if a low extraction power solvent is used, i.e. high level of ethanol. The positive effect was lower if using low levels of ethanol in the extraction solvent. None of the other individual extraction variables had a significant effect on the recovery. However, experimental variables related to ultrasound properties showed significant interaction with the solvent composition (p-value: 0.007). The interaction found for US amplitude and extraction solvent is represented in Fig. 2. It can be seen that the recoveries are much lower on using mixtures of 50% ethanol in water than on using either pure water or 25% ethanol in water, although the reduction in the recovery was lower for high amplitude ultrasound (70%) than for low amplitude (20%). This finding indicates that ultrasound can, to some extent, overcome the effect of a less efficient extraction solvent than water. The pH of the solvent and the sample-to-solvent ratio did not have significant effects on the recovery, with p-values greater than
0.05. In this way, the average recovery for extractions run at different pH values were 12.8 mmol kg1 at pH = 2, 1 12.8 mmol kg at pH = 5 and 13.0 mmol kg1 at pH = 7. The average recovery for extractions run at different sample-to-solvent ratios were 13.2 mmol kg1 for sample-to-solvent ratio = 0.1 g mL1, 12.5 mmol kg1 for sample-to-solvent ratio = 0.067 g mL1 and 13.6 mmol kg1 for sample-to-solvent ratio = 0.05 g mL1. Having evaluated the experimental design, it can be concluded that pure water should be used as solvent, with an extraction temperature of 70 °C or higher, an output amplitude of 70% or higher, any duty cycle between 2 and 7 s1, any pH between 2 and 7 and also any sample-to-solvent ratio between 0.1 g mL1 and 0.05 g mL1. As water was chosen as the extraction solvent, temperatures above 70 °C could be applied and therefore temperatures of 80 °C and 90 °C were assayed. Output amplitude values higher than 70% were not assayed because they usually produce liquid losses due to liquid splashes during the extraction. For the other variables, 7 s1 was used for cycle, pH was not controlled and a sample-to-solvent ratio of 0.1 g mL1 was selected in order to produce more concentrated extracts. The results obtained for higher extraction temperatures are presented in Fig. 3. It can be seen that non-significant differences were found for recoveries at higher temperatures (80 or 90 °C) in comparison to the recoveries found for 70 °C. Therefore 70 °C was set as the extraction temperature. This result is related with the interaction previously presented between solvent composition and extraction temperature, as explained before the positive effect of higher temperatures was really significant for low extraction power solvents, much less important for solvents with low ethanol percentages. 3.2. Kinetics of the extraction process In order to assess the kinetics of the extraction process, several extractions were run using extraction times between 3 and 15 min. Results for the recovery of total amino acids are presented in Fig. 4. Significant differences were not found for extractions run between 3 and 12 min, but a much higher standard deviation was found for extraction times of 3 min, a result that is probably related to the time needed for the sample to reach 70 °C. As a consequence, 6 min was chosen as the extraction time.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
5
Fig. 2. Average amino acid recoveries for extracts obtained with specific solvent compositions and ultrasound amplitudes.
Levels of individual amino acids found in the extracts were always higher for the extracts obtained using ultrasound, differences reached +48% for alanine, +15% for histidine and +12% for proline, some of the major amino acids in the sample. Non-significant differences were found for arginine. 3.4. Analytical properties
Fig. 3. Average amino acid recoveries for extracts obtained at 70 °C, 80 °C and 90 °C.
Fig. 4. Average amino acid recoveries for extracts obtained at using different extraction times.
3.3. Effect of ultrasound In order to assess the effect of ultrasound on the recovery, eight extractions were carried out on the same sample both with and without ultrasound but using magnetic stirring. The same solvent (water), extraction temperature (70 °C) and extraction time (6 min) were used.
The repeatability and reproducibility of the developed method were studied. A total of 15 extractions were performed and these are distributed as follows: 9 extractions performed on the first day of the study and 3 more extractions on each of the two following days. Evaluation of both repeatability and reproducibility was carried out using individual values for the amino acids in the sample. Repeatability results (RSD) ranged from 8.6% for alanine to 19.3% for glycine, with an average value of 12.9%. Reproducibility results ranged from 9.7% for histidine to 30.5% for proline, with an average value of 15.7%. High RSD values for reproducibility and repeatability for some specific amino acids could be related to sample stability. The same homogenized grape sample was used to check the reproducibility, although it should be noted that the sample was thawed and re-frozen every day. An amino acid standard solution was used for recovery calculations. Recoveries of individual amino acids ranged from 63.1% for histidine to 105.2% for threonine, with an average recovery of 89.1% and only four amino acids with recoveries below 90%, i.e. histidine (63.1%), methionine (70.7%), isoleucine (77.5%) and valine (78.6%). Low recoveries for those compounds could be related to oxidative degradation processes. It has been demonstrate that ultrasounds produce oxidative degradation for some compounds in foods, specifically for lipids [14], lycopene [15] or some anthocyanins showing some hydroxylation reactions [16]. Pingret et al. [17] has recently reported several examples of degradation reactions in ultrasound assisted process. Regarding amino acids, there is not previous information about their degradation reactions in these conditions. Additionally, Fig. 4 shows that recovery started to decrease after 15 min. Therefore, low recoveries found using only 6 min as extraction time, could be related to some adsorption or reaction with other compounds in the matrix instead of degradation reactions. The limits of detection and quantification were established after running the extraction of a blank six times. The LOD values (n = 6) and LOQ values (n = 6) ranged from 0.7 and 2.2 mmol kg1, respectively, for glutamic acid and 1.0 and 3.3 mmol kg1, respectively, for methionine.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
6
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
3.5. Application to real samples The suitability of this method for real samples during the ripening period was evaluated by monitoring the ripening process for three cultivars [irrigate (denoted by the letter I), regular cultivar (R) and nitrogen treated vines (N)] for the same grape variety. The results found for the three cultivars are shown in Fig. 5. Three sampling points were used (1, 2 and 3). The first
sampling point was used as a reference and differences were not found between R, I and N samples because both irrigation and nitrogen addition in vines started after the sampling date. It can be seen that only histidine showed differences between the amino acids found in the samples. Values for I and N samples were around 30% higher than for grapes from untreated vines. The second sampling point was run 2 weeks after both the irrigation and nitrogen supplement started. Differences were
Fig. 5. Amino acid levels for Verdejo grapes under three different cultivar conditions: R (regular cultivar), I (irrigated cultivar) and N (nitrogen supplement).
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
not significant between the three cultivars. Only a clear reduction of aspartic acid was found for both R and I samples. In this second sample point a decrease in histidine was found and the levels of this amino acid were lower for both I and N grapes and very similar to those for the R grapes. Finally, the third sampling point was run two days before the harvest day, i.e. with mature grapes. It can be seen that a reduction in levels for most amino acids was detected. However, for some of the amino acids, including histidine, arginine and proline, the reduction produced for grapes from vines treated with nitrogen were clearly lower, it means these grapes have maintained higher levels for those amino acids than R grapes. R samples also showed a lower reduction for proline. Isoleucine levels were also increasing for N samples vs I and R samples. Therefore the final grapes from vines treated with nitrogen showed higher levels than both R and I grapes for the amino acids mentioned above. These results indicate that nitrogen supplements in vines affect the final levels of amino acids in grapes, and this in turn will condition the final levels of amino acids in the grape must, with clearly higher levels of histidine, arginine and proline.
4. Conclusions Ultrasound-assisted extraction, using the method developed in this study, allows the quantitative and reproducible extraction of the amino acids present in grapes in a short time (6 min) with water employed as the extraction solvent at 70 °C. Comparison of the extraction using maceration with the developed ultrasound assisted extraction method shows that the new method provides higher recoveries for most amino acids in grapes. Given its low instrumental requirements, its simplicity and its analytical capabilities, the method developed here is suitable for the routine analysis of amino acids in grapes during ripening at the winemaking companies.
Acknowledgements This work was carried out with funding received for the project AGR 6874 funded by the Junta de Andalucía.
7
References [1] J. Lee, K.L. Steenwerth, Rootstock and vineyard floor management influence on ‘Cabernet Sauvignon’ grape yeast assimilable nitrogen (YAN), Food Chem. 127 (2011) 926–933. [2] S. Gouthu, P.A. Skinkis, J. Morre, C.S. Maier, L.G. Deluc, Berry nitrogen status altered by cover cropping: effects on berry hormone dynamics, growth and amino acid composition of Pinot Noir, Food Chem. 135 (2012) 1–8. [3] S. Jogaiah, D.P. Oulkar, K. Banerjee, P. Raveendran, N.P. Rokade, Amino acid profile of ‘Thompson Seedless’ grapes grafted on different rootstocks at various stages of berry development, Int. J. Fruit Sci. 10 (2010) 323–340. [4] J. Oliva, T. Garde-Cerdán, A.M. Martínez-Gil, M. Rosario Salinas, A. Barba, Fungicide effects on ammonium and amino acids of Monastrell grapes, Food Chem. 129 (2011) 1676–1680. [5] T. Garde-Cerdán, A.M. Martínez-Gil, C. Lorenzo, J.F. Lara, F. Pardo, M.R. Salinas, Implications of nitrogen compounds during alcoholic fermentation from some grape varieties at different maturation stages and cultivation systems, Food Chem. 124 (2011) 106–116. [6] S.J. Bell, P.A. Henschke, Implications of nitrogen nutrition for grapes, fermentation and wine, Aust. J. Grape Wine Res. 11 (2005) 242–295. [7] T. Garde-Cerdán, C. Ancín-Azpilicueta, Effect of the addition of different quantities of amino acids to nitrogen-deficient must on the formation of esters, alcohols, and acids during wine alcoholic fermentation, LWT – Food Sci. Technol. 41 (2008) 501–510. [8] P. Hernandez-Orte, M. Bely, J. Cacho, V. Ferreira, Impact of ammonium additions on volatile acidity, ethanol, and aromatic compound production by different Saccharomyces cerevisiae strains during fermentation in controlled synthetic media, Aust. J. Grape Wine Res. 12 (2006) 150–160. [9] R. Jackson, Wine Science, Academic Press, 2008. [10] E. Casalta, J.M. Sablayrolles, J.M. Salmon, Comparison of different methods for the determination of assimilable nitrogen in grape musts, LWT – Food Sci. Technol. 54 (2013) 271–277. [11] R.M. Callejón, A.M. Troncoso, M.L. Morales, Determination of amino acids in grape-derived products: a review, Talanta 81 (2010) 1143–1152. [12] D. Long, K.L. Wilkinson, K. Poole, D.K. Taylor, T. Warren, A.M. Astorga, V. Jiranek, Rapid method for proline determination in grape juice and wine, J. Agric. Food Chem. 60 (2012) 4259–4264. [13] D. Pingret, A.-S. Fabiano-Tixier, F. Chemat, Ultrasound-assisted extraction, in: M. Rostagno, J. Prado (Eds.), Natural Product Extraction: Principles and Applications, RSC Publishing, Cambridge, UK, 213, 2013, pp. 89–112. [14] D. Pingret, G. Durand, A.-S. Fabiano-Tixier, A. Rockenbauer, C. Ginies, F. Chemat, Degradation of edible oil during food processing by ultrasound: electron paramagnetic resonance, physicochemical, and sensory appreciation, J. Agric. Food Chem. 60 (2012) 7761–7768. [15] A.L.-S. Eh, S.-G. Teoh, Novel modified ultrasonication technique for the extraction of lycopene from tomatoes, Ultrason. Sonochem. 19 (2012) 151– 159. [16] M. Ashokkumar, D. Sunartio, S. Kentish, R. Mawson, L. Simons, K. Vilkhu, C. Versteeg, Modification of food ingredients by ultrasound to improve functionality: a preliminary study on a model system, Innov. Food Sci. Emerg. 9 (2008) 155–160. [17] D. Pingret, A.-S. Fabiano-Tixier, F. Chemat, Degradation during application of ultrasound in food processing: a review, Food Control 31 (2013) 593–606.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
Contents lists available at ScienceDirect
Ultrasonics Sonochemistry journal homepage: www.elsevier.com/locate/ultson
Ultrasound-assisted extraction of amino acids from grapes Ceferino Carrera a, Ana Ruiz-Rodríguez b, Miguel Palma b,⇑, Carmelo G. Barroso b a b
Andalusian Center for Wine Research, Agrifood Campus of International Excellence (ceiA3), University of Cádiz, Campus de Puerto Real, 11510 Puerto Real, Cádiz, Spain Department of Analytical Chemistry, Faculty of Sciences, Agrifood Campus of International Excellence (ceiA3), University of Cádiz, Spain
a r t i c l e
i n f o
Article history: Received 11 November 2013 Received in revised form 7 February 2014 Accepted 26 May 2014 Available online xxxx Keywords: Amino acids Ultrasound-assisted extraction Grapes
a b s t r a c t Recent cultivar techniques on vineyards can have a marked influence on the final nitrogen content of grapes, specifically individual amino acid contents. Furthermore, individual amino acid contents in grapes are related to the final aromatic composition of wines. A new ultrasound-assisted method for the extraction of amino acids from grapes has been developed. Several extraction variables, including solvent (water/ethanol mixtures), solvent pH (2–7), temperature (10–70 °C), ultrasonic power (20–70%) and ultrasonic frequency (0.2–1.0 s1), were optimized to guarantee full recovery of the amino acids from grapes. An experimental design was employed to optimize the extraction parameters. The surface response methodology was used to evaluate the effects of the extraction variables. The analytical properties of the new method were established, including limit of detection (average value 1.4 mmol kg1), limit of quantification (average value 2.6 mmol kg1), repeatability (average RSD = 12.9%) and reproducibility (average RSD = 15.7%). Finally, the new method was applied to three cultivars of white grape throughout the ripening period. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Total free amino acid contents in grapes and also the amino acid profiles can vary dramatically depending on the grape cultivar [1], cultivar practices [2], ripening degree [3], vine nutrient/water/fungicide management and health status, including fungal infections [4]. Additionally, grape cultivar using different rootstocks usually present significant differences for the final composition of grapes. It has been demonstrated that different rootstocks for the same grape variety, for example, produce different amino acid profiles in the final grapes [3]. The Cabernet Sauvignon grape variety with a high vigor rootstock, produces much higher levels of most amino acids in the grapes than with a low vigor rootstock [1]. Cultivar practices, such as the use of cover crops, have different effects on amino acid levels depending on the variety. For example, changes were not observed for Cabernet Sauvignon [1] but a dramatic reduction in amino acid levels was observed for the Pinot Noir grape variety on using cover crops [2]. Vine health status and fungicide application usually affect the levels of amino acids in grapes. The application of fungicide close to the harvest date leads to lower
⇑ Corresponding author. Address: Department of Analytical Chemistry, Faculty of Sciences, Agrifood Campus of International Excellence (ceiA3), University of Cádiz, 11510 Puerto Real, Spain. Tel.: +34 956016775; fax: +34 956016590. E-mail address: [email protected] (M. Palma).
levels for all amino acids [4] and this makes the alcoholic fermentation process more difficult. It has been proved that there is a direct relationship between the amino acid levels in grape must and their consumption during the first half of alcoholic fermentation [5]. This means that the starting levels of nitrogen compounds, specifically amino acids, will determine the evolution of the first step in the alcoholic fermentation, therefore affecting the final composition of wines. It has been demonstrated in several previous studies that there is a correlation between amino acid contents in grapes/must and the aroma of the final wines, specifically for young wines [6]. Free nitrogen levels, including amino acids, are also associated with some problems in the winemaking process. Insufficient nitrogen levels can stop fermentation and this leads to the production of numerous secondary metabolites that are not of interest for wine production [7]. Furthermore, some amino acids are directly related to certain volatile compounds that contribute to the wine aroma, including fusel oils, aldehydes and esters [8]. Wines with very high levels of nitrogen compounds can also have higher microbiological instability, including those processes related to biogenic amines and ethyl carbamate generation [9]. Given the information outlined above, it is advantageous to determine amino acids in grapes during ripening in an effort to identify the effects of cultivar practices and cultivar conditions. It is also of interest to determine amino acid levels in grapes just prior to alcoholic fermentation in order to establish more suitable winemaking practices.
http://dx.doi.org/10.1016/j.ultsonch.2014.05.021 1350-4177/Ó 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
2
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
There are several methods for the determination of free amino nitrogen [10] and amino acids in must. Most of these approaches involve chromatographic techniques [11], but colorimetric methods are also employed [12]. In many cases, the final determination requires some kind of derivatization step to increase the analytical signal, including by fluorescence detection systems [11]. However, the problem of the determination of amino acids in grapes has not yet been solved satisfactorily. Differences in individual free amino acid profiles and concentrations have been observed between the different sample preparation methods usually applied, i.e. determination in grape juice or chemically extracted grapes, before amino acid determination [1]. There are three main reasons for these differences. Firstly, extraction – rather than grape juice preparation – will include amino acids from grape skins in the final level. Proline, for example, has been reported to be found in grape skins [1]. Secondly, extraction usually leads to dilution of the sample due to the presence of extraction solvents and this means that some levels could go below the limits of quantification. Finally, quantification in grape juice will be expressed as mg per liter whilst quantification in grapes will be expressed as mg per kg. The relationship between kg of grape and liter of grape juice is dependent on the winemaking process. Additionally, during the red winemaking process, grape skins are in contact with grape must for several days and amino acids will therefore be extracted from the skins to some extent. As a consequence, determination of the total amount of amino acids in grapes would be more useful than the amount in the grape juice alone. The aim of the work described here was to develop and validate a new ultrasound-assisted extraction method for amino acids in grapes. The overall goal was to provide a rapid and reliable method for winemakers to determine amino acids in grapes during grape ripening and harvest. Ultrasound assisted extraction was used because of this technique usually produce faster extraction methods, lower solvent volumes and higher extraction yields [13]. 2. Materials and methods 2.1. Chemicals and solvents Ethanol and methanol (Panreac, Barcelona, Spain) were HPLC grade. Ultra pure water was supplied by a Milli-Q water purification system from Millipore (Bedford, MA, USA). 2.2. Grape samples The white grape (var. Verdejo) was employed in the development of the ultrasound-assisted extraction method. Samples were obtained from local vineyards in the Jerez region (Spain). The full berry, i.e. skin, pulp and seeds, was studied. The berries were triturated with a conventional beater until a homogeneous sample was obtained for analysis. The triturated sample was stored in a freezer at 20 °C prior to analysis. 2.3. Extraction procedure The extraction of amino acids from grapes by means of ultrasound was performed by employing water/ethanol mixtures as solvent. The effects of the extraction solvent (0–25% EtOH in water), temperature (10–70 °C), output amplitude of the nominal amplitude of the transducer (30–70%), duty cycle (0.2–0.7 s), pH of the extraction solvent (2–7) and the sample/solvent ratio (1 g 10 mL1–1 g 20 mL1) were studied. Ultrasonic irradiation was carried out using a UP200S sonifier (200 W, 24 kHz) (Hielscher Ultrasonics, Teltow, Germany), with
the sample immersed in a water bath coupled to a temperature controller (Frigiterm, J.P. Selecta, Barcelona, Spain). 2.4. Experimental design for the evaluation of the effects of extraction variables Optimization of extraction variables was performed using the Box–Behnken statistical methodology by considering the total amount of amino acids extracted. The results for the 54 extractions carried out in duplicate for the seven extraction variables (each variable has three levels: low, medium and high) including 6 center points are shown in Table 1 along with the respective responses. The results for the total amount of amino acids (mmol kg1 of sample) determined by HPLC were used as the response variable. The responses obtained from the various extractions were entered into to a second-order polynomial equation into which each of the various parameters was introduced. The polynomial equation is as follows:
Y ¼ b0 þ b1X1 þ b2X2 þ b3X3 þ b4X4 þ b5X5 þ b6X6 þ b7X7 þ b12X1X2 þ b13X1X3 þ b14X1X4 þ b15X1X5 þ b16X1X6 þ b23X2X3 þ b24X2X4 þ b25X2X5 þ b26X2X6 þ b34X3X4 þ b35X3X5 þ b36X3X6 þ b45X4X5 þ b46X4X6 þ b56X5X6 þ b11X12 þ b22X22 þ b33X32 þ b44X42 þ b55X52 þ b66X62 In this equation Y is the aforementioned response, b0 is the ordinate at the origin; X1 [percentage of EtOH in the extraction solvent], X2 [temperature (°C)], X3 [ultrasound amplitude], X4 [duty cycle], X5 [pH], X6 [ratio solid sample (g)/extraction volume (mL)] are the independent variables; bi are the linear coefficients; bij are the cross product coefficients and bii are the quadratic coefficients. The analysis of data for the Box–Behnken design was carried out using Unscrambler X (Camo, No). This software was used to estimate the effects of the variables on the final response, the variance analysis, the second order mathematical model, the optimum levels of the significant variables and the surface graphs. 2.5. Amino acids determination The AccQ-Tag method was used for derivatization and chromatographic determination. The AccQTag Reagent Kit was purchased from Waters (Milford, Massachusetts, USA). The reagent kit consists of Waters AccQFluor Borate Buffer, Waters AccQFluor Reagent Powder (6-aminoquinolyl-N-hydroxysuccinimidyl carbamate – AQC), Waters AccQFluor Reagent Diluent, Waters AccQTag Amino Acid Analysing Column (Nova-Pak C18, 4 lm, 150 3.9 mm) and Waters Amino Acid Hydrolysate Standard (each ampoule contains a 2.5 mM mixture of the 17 amino acids with the exception of cystine – 1.25 mM. The following amino acids were found and quantified in the samples: aspartic acid, serine, glutamic acid, glycine, histidine, arginine, threonine, alanine, proline, cysteine, tyrosine, valine, methionine, lysine, isoleucine and leucine. 3. Results and discussion 3.1. Optimization of the extraction method A Box–Behnken design was used to optimize the extraction conditions for amino acids from homogenized grape samples. Six different extraction variables were studied in the following ranges: ethanol in water between 0% and 50%, temperature values between
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
3
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx Table 1 Results from the experimental design. Extraction yield (mmol of total amino acids kg1 of grape) for the evaluation of extraction variables in the experimental design. Experiment
% EtOH in water
Temperature
Amplitude
Cycle
pH
Ratio
mmol kg1 obtained
mmol kg1 estimated by the model
Difference real vs estimated (%)
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 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 cp01 cp02 cp03 cp04 cp05 cp06
25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 0 50 0 50 0 50 0 50 25 25 25 25 25 25 25 25 0 50 0 50 0 50 0 50 0 50 0 50 0 50 0 50 25 25 25 25 25 25
40 40 40 40 40 40 40 40 10 70 10 70 10 70 10 70 40 40 40 40 40 40 40 40 10 70 10 70 10 70 10 70 10 10 70 70 10 10 70 70 40 40 40 40 40 40 40 40 40 40 40 40 40 40
30 70 30 70 30 70 30 70 50 50 50 50 50 50 50 50 30 30 70 70 30 30 70 70 30 30 70 70 30 30 70 70 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 2 2 7 7 2 2 7 7 2 2 2 2 7 7 7 7 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 2 2 7 7 2 2 7 7 4.5 4.5 4.5 4.5 4.5 4.5
3 3 7 7 3 3 7 7 3 3 3 3 7 7 7 7 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 3 3 3 3 7 7 7 7 5 5 5 5 5 5 5 5 5 5 5 5 5 5
10 10 10 10 20 20 20 20 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 10 10 10 10 20 20 20 20 15 15 15 15 15 15 15 15 10 10 10 10 20 20 20 20 15 15 15 15 15 15
14.4 14.2 15.9 14.3 12.8 12.6 13.7 14.5 11.7 12.7 12.5 15.8 11.0 13.9 11.5 14.1 18.7 7.0 13.1 8.0 17.6 6.3 13.5 12.4 13.4 18.1 12.1 13.1 14.0 16.1 10.4 19.5 12.3 6.9 13.3 14.1 14.1 6.4 16.8 13.2 10.1 8.4 15.0 9.4 14.8 14.7 10.2 9.9 11.9 13.2 12.8 13.4 12.6 13.6
13.5 11.7 13.8 11.7 12.5 12.8 13.5 13.6 10.6 13.5 11.2 15.1 11.6 15.1 10.6 15.2 18.3 9.7 11.6 11.3 16.9 8.4 13.5 13.4 12.9 14.4 10.1 13.3 10.8 15.1 10.3 16.2 13.6 8.3 14.4 14.2 15.5 6.8 16.9 13.4 13.7 7.5 17.8 11.7 16.2 13.5 12.7 10.0 12.9 12.9 12.9 12.9 12.9 12.9
6 18 13 18 2 2 2 7 9 6 10 4 6 9 7 7 2 38 11 42 4 33 0 8 4 20 16 1 22 6 2 17 11 21 8 1 10 7 1 2 36 11 19 25 9 9 24 1 8 2 1 4 2 5
10 and 70 °C, ultrasound amplitude between 30% and 70%, ultrasound duty cycle between 0.2 and 0.7 s1, solvent pH between 2 and 7 and sample-to-solvent ratio between 0.1 and 0.05 g mL1 (Table 1). Extraction time was fixed at 5 min. All experiments (a total of 54) were run in duplicate for the extraction. The recovery of amino acids was correlated with experimental conditions by a second-order polynomial equation. The resulting coefficients for the second-order polynomial equation from the Box–Behnken design and their significance (p-value) are presented in Table 2. Data related to the fitting properties of the resulting model, expressed as percentage difference between the experimental values for the amino acid recoveries and the calculated ones, are presented in Table 1. The resulting average difference was 10%, ranging from 0% to 42%, with only nine experiments showing differences greater than 20% between experimental and
calculated values, i.e. 92% of the experimental data were fitted by the model with an error of less than 20%. It has to be noted that most experiments showing high relative differences between real and predicted values were those showing low level for total amino acids, experiments 18, 20, 22, 34, 41, 44 and 47. For those cases, relative differences are obviously high. The resulting model therefore explains the recovery results for the extracts based on the experimental conditions used during the extraction process. This means that extraction variables control the final extract composition in terms of amino acid recovery and that the extraction conditions can therefore be managed to optimize the extraction of amino acids from grapes. Analysis of the model clearly shows that, among the linear terms, the most influential variable was the solvent (p-value = 0.00001). The solvent has a negative coefficient, which
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
4
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
Table 2 Coefficients for the second-order polynomial equation from the Box–Behnken design.
B0 Solvent Temp Amplitude Cycle pH Ratio Solvent * temp Solvent * amplitude Temp * amplitude Solvent * cycle Temp * cycle Amplitude * cycle Solvent * pH Temp * pH Amplitude * pH Cycle * pH Solvent * ratio Temp * ratio Amplitude * ratio Cycle * ratio pH * ratio
Coefficient
p-Value
386.60 65.96 55.60 12.83 5.07 7.66 6.15 38.70 63.27 12.37 0.68 7.88 25.19 25.25 5.04 1.76 11.70 25.50 20.52 16.36 57.51 5.39
0.00001 0.00013 0.32180 0.69367 0.55219 0.63305 0.08927 0.00730 0.57932 0.96552 0.72366 0.26239 0.26132 0.74885 0.93692 0.59991 0.25666 0.35966 0.30253 0.01384 0.80872 Fig. 1. Surface plot for the interaction between solvent and temperature on the amino acid recovery.
indicates that the use of a lower proportion of ethanol gives rise to higher levels of amino acids in the extract, i.e. higher recoveries. If results in Table 1 are grouped by the level of ethanol in the solvent, it can be found that the average total amino acid content in the extracts obtained using 0% ethanol was 14.1 mmol kg1, whereas the average total amino acid content in the extracts obtained using 50% ethanol was 9.7 mmol kg1. Therefore decreasing the level of ethanol in the solvent, amino acid recovery can be increase as much as 45%. It can also be seen from the results in Table 2 that temperature has a significant effect on the recovery. The effect of temperature was positive, with higher temperatures leading to a higher recovery of amino acids. The average recovery obtained at 70 °C was 15.0 mmol kg1, whereas the average total amino acid content in the extracts obtained at 10 °C was 11.3 mmol kg1. Analysis of interactions between extraction variables usually shows interesting information. In this case, a significant interaction was found for solvent composition and temperature. The resulting surface plot for this interaction is shown in Fig. 1. It can be seen that there is a very rapid increase in the amino acid recovery as extraction conditions move from low to high temperatures when 50% ethanol was used as the extraction solvent (recoveries changed from around 7.3 mmol kg1 to 12.7 mmol kg1). However, if the same changes in temperature are produced using 100% water as the extraction solvent, the recovery only increased from 14.0 mmol kg1 to 16.0 mmol kg1, indicating a much lower effect due to temperature changes in this case. It means the positive effect of increasing temperature was more significant if a low extraction power solvent is used, i.e. high level of ethanol. The positive effect was lower if using low levels of ethanol in the extraction solvent. None of the other individual extraction variables had a significant effect on the recovery. However, experimental variables related to ultrasound properties showed significant interaction with the solvent composition (p-value: 0.007). The interaction found for US amplitude and extraction solvent is represented in Fig. 2. It can be seen that the recoveries are much lower on using mixtures of 50% ethanol in water than on using either pure water or 25% ethanol in water, although the reduction in the recovery was lower for high amplitude ultrasound (70%) than for low amplitude (20%). This finding indicates that ultrasound can, to some extent, overcome the effect of a less efficient extraction solvent than water. The pH of the solvent and the sample-to-solvent ratio did not have significant effects on the recovery, with p-values greater than
0.05. In this way, the average recovery for extractions run at different pH values were 12.8 mmol kg1 at pH = 2, 1 12.8 mmol kg at pH = 5 and 13.0 mmol kg1 at pH = 7. The average recovery for extractions run at different sample-to-solvent ratios were 13.2 mmol kg1 for sample-to-solvent ratio = 0.1 g mL1, 12.5 mmol kg1 for sample-to-solvent ratio = 0.067 g mL1 and 13.6 mmol kg1 for sample-to-solvent ratio = 0.05 g mL1. Having evaluated the experimental design, it can be concluded that pure water should be used as solvent, with an extraction temperature of 70 °C or higher, an output amplitude of 70% or higher, any duty cycle between 2 and 7 s1, any pH between 2 and 7 and also any sample-to-solvent ratio between 0.1 g mL1 and 0.05 g mL1. As water was chosen as the extraction solvent, temperatures above 70 °C could be applied and therefore temperatures of 80 °C and 90 °C were assayed. Output amplitude values higher than 70% were not assayed because they usually produce liquid losses due to liquid splashes during the extraction. For the other variables, 7 s1 was used for cycle, pH was not controlled and a sample-to-solvent ratio of 0.1 g mL1 was selected in order to produce more concentrated extracts. The results obtained for higher extraction temperatures are presented in Fig. 3. It can be seen that non-significant differences were found for recoveries at higher temperatures (80 or 90 °C) in comparison to the recoveries found for 70 °C. Therefore 70 °C was set as the extraction temperature. This result is related with the interaction previously presented between solvent composition and extraction temperature, as explained before the positive effect of higher temperatures was really significant for low extraction power solvents, much less important for solvents with low ethanol percentages. 3.2. Kinetics of the extraction process In order to assess the kinetics of the extraction process, several extractions were run using extraction times between 3 and 15 min. Results for the recovery of total amino acids are presented in Fig. 4. Significant differences were not found for extractions run between 3 and 12 min, but a much higher standard deviation was found for extraction times of 3 min, a result that is probably related to the time needed for the sample to reach 70 °C. As a consequence, 6 min was chosen as the extraction time.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
5
Fig. 2. Average amino acid recoveries for extracts obtained with specific solvent compositions and ultrasound amplitudes.
Levels of individual amino acids found in the extracts were always higher for the extracts obtained using ultrasound, differences reached +48% for alanine, +15% for histidine and +12% for proline, some of the major amino acids in the sample. Non-significant differences were found for arginine. 3.4. Analytical properties
Fig. 3. Average amino acid recoveries for extracts obtained at 70 °C, 80 °C and 90 °C.
Fig. 4. Average amino acid recoveries for extracts obtained at using different extraction times.
3.3. Effect of ultrasound In order to assess the effect of ultrasound on the recovery, eight extractions were carried out on the same sample both with and without ultrasound but using magnetic stirring. The same solvent (water), extraction temperature (70 °C) and extraction time (6 min) were used.
The repeatability and reproducibility of the developed method were studied. A total of 15 extractions were performed and these are distributed as follows: 9 extractions performed on the first day of the study and 3 more extractions on each of the two following days. Evaluation of both repeatability and reproducibility was carried out using individual values for the amino acids in the sample. Repeatability results (RSD) ranged from 8.6% for alanine to 19.3% for glycine, with an average value of 12.9%. Reproducibility results ranged from 9.7% for histidine to 30.5% for proline, with an average value of 15.7%. High RSD values for reproducibility and repeatability for some specific amino acids could be related to sample stability. The same homogenized grape sample was used to check the reproducibility, although it should be noted that the sample was thawed and re-frozen every day. An amino acid standard solution was used for recovery calculations. Recoveries of individual amino acids ranged from 63.1% for histidine to 105.2% for threonine, with an average recovery of 89.1% and only four amino acids with recoveries below 90%, i.e. histidine (63.1%), methionine (70.7%), isoleucine (77.5%) and valine (78.6%). Low recoveries for those compounds could be related to oxidative degradation processes. It has been demonstrate that ultrasounds produce oxidative degradation for some compounds in foods, specifically for lipids [14], lycopene [15] or some anthocyanins showing some hydroxylation reactions [16]. Pingret et al. [17] has recently reported several examples of degradation reactions in ultrasound assisted process. Regarding amino acids, there is not previous information about their degradation reactions in these conditions. Additionally, Fig. 4 shows that recovery started to decrease after 15 min. Therefore, low recoveries found using only 6 min as extraction time, could be related to some adsorption or reaction with other compounds in the matrix instead of degradation reactions. The limits of detection and quantification were established after running the extraction of a blank six times. The LOD values (n = 6) and LOQ values (n = 6) ranged from 0.7 and 2.2 mmol kg1, respectively, for glutamic acid and 1.0 and 3.3 mmol kg1, respectively, for methionine.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
6
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
3.5. Application to real samples The suitability of this method for real samples during the ripening period was evaluated by monitoring the ripening process for three cultivars [irrigate (denoted by the letter I), regular cultivar (R) and nitrogen treated vines (N)] for the same grape variety. The results found for the three cultivars are shown in Fig. 5. Three sampling points were used (1, 2 and 3). The first
sampling point was used as a reference and differences were not found between R, I and N samples because both irrigation and nitrogen addition in vines started after the sampling date. It can be seen that only histidine showed differences between the amino acids found in the samples. Values for I and N samples were around 30% higher than for grapes from untreated vines. The second sampling point was run 2 weeks after both the irrigation and nitrogen supplement started. Differences were
Fig. 5. Amino acid levels for Verdejo grapes under three different cultivar conditions: R (regular cultivar), I (irrigated cultivar) and N (nitrogen supplement).
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021
C. Carrera et al. / Ultrasonics Sonochemistry xxx (2014) xxx–xxx
not significant between the three cultivars. Only a clear reduction of aspartic acid was found for both R and I samples. In this second sample point a decrease in histidine was found and the levels of this amino acid were lower for both I and N grapes and very similar to those for the R grapes. Finally, the third sampling point was run two days before the harvest day, i.e. with mature grapes. It can be seen that a reduction in levels for most amino acids was detected. However, for some of the amino acids, including histidine, arginine and proline, the reduction produced for grapes from vines treated with nitrogen were clearly lower, it means these grapes have maintained higher levels for those amino acids than R grapes. R samples also showed a lower reduction for proline. Isoleucine levels were also increasing for N samples vs I and R samples. Therefore the final grapes from vines treated with nitrogen showed higher levels than both R and I grapes for the amino acids mentioned above. These results indicate that nitrogen supplements in vines affect the final levels of amino acids in grapes, and this in turn will condition the final levels of amino acids in the grape must, with clearly higher levels of histidine, arginine and proline.
4. Conclusions Ultrasound-assisted extraction, using the method developed in this study, allows the quantitative and reproducible extraction of the amino acids present in grapes in a short time (6 min) with water employed as the extraction solvent at 70 °C. Comparison of the extraction using maceration with the developed ultrasound assisted extraction method shows that the new method provides higher recoveries for most amino acids in grapes. Given its low instrumental requirements, its simplicity and its analytical capabilities, the method developed here is suitable for the routine analysis of amino acids in grapes during ripening at the winemaking companies.
Acknowledgements This work was carried out with funding received for the project AGR 6874 funded by the Junta de Andalucía.
7
References [1] J. Lee, K.L. Steenwerth, Rootstock and vineyard floor management influence on ‘Cabernet Sauvignon’ grape yeast assimilable nitrogen (YAN), Food Chem. 127 (2011) 926–933. [2] S. Gouthu, P.A. Skinkis, J. Morre, C.S. Maier, L.G. Deluc, Berry nitrogen status altered by cover cropping: effects on berry hormone dynamics, growth and amino acid composition of Pinot Noir, Food Chem. 135 (2012) 1–8. [3] S. Jogaiah, D.P. Oulkar, K. Banerjee, P. Raveendran, N.P. Rokade, Amino acid profile of ‘Thompson Seedless’ grapes grafted on different rootstocks at various stages of berry development, Int. J. Fruit Sci. 10 (2010) 323–340. [4] J. Oliva, T. Garde-Cerdán, A.M. Martínez-Gil, M. Rosario Salinas, A. Barba, Fungicide effects on ammonium and amino acids of Monastrell grapes, Food Chem. 129 (2011) 1676–1680. [5] T. Garde-Cerdán, A.M. Martínez-Gil, C. Lorenzo, J.F. Lara, F. Pardo, M.R. Salinas, Implications of nitrogen compounds during alcoholic fermentation from some grape varieties at different maturation stages and cultivation systems, Food Chem. 124 (2011) 106–116. [6] S.J. Bell, P.A. Henschke, Implications of nitrogen nutrition for grapes, fermentation and wine, Aust. J. Grape Wine Res. 11 (2005) 242–295. [7] T. Garde-Cerdán, C. Ancín-Azpilicueta, Effect of the addition of different quantities of amino acids to nitrogen-deficient must on the formation of esters, alcohols, and acids during wine alcoholic fermentation, LWT – Food Sci. Technol. 41 (2008) 501–510. [8] P. Hernandez-Orte, M. Bely, J. Cacho, V. Ferreira, Impact of ammonium additions on volatile acidity, ethanol, and aromatic compound production by different Saccharomyces cerevisiae strains during fermentation in controlled synthetic media, Aust. J. Grape Wine Res. 12 (2006) 150–160. [9] R. Jackson, Wine Science, Academic Press, 2008. [10] E. Casalta, J.M. Sablayrolles, J.M. Salmon, Comparison of different methods for the determination of assimilable nitrogen in grape musts, LWT – Food Sci. Technol. 54 (2013) 271–277. [11] R.M. Callejón, A.M. Troncoso, M.L. Morales, Determination of amino acids in grape-derived products: a review, Talanta 81 (2010) 1143–1152. [12] D. Long, K.L. Wilkinson, K. Poole, D.K. Taylor, T. Warren, A.M. Astorga, V. Jiranek, Rapid method for proline determination in grape juice and wine, J. Agric. Food Chem. 60 (2012) 4259–4264. [13] D. Pingret, A.-S. Fabiano-Tixier, F. Chemat, Ultrasound-assisted extraction, in: M. Rostagno, J. Prado (Eds.), Natural Product Extraction: Principles and Applications, RSC Publishing, Cambridge, UK, 213, 2013, pp. 89–112. [14] D. Pingret, G. Durand, A.-S. Fabiano-Tixier, A. Rockenbauer, C. Ginies, F. Chemat, Degradation of edible oil during food processing by ultrasound: electron paramagnetic resonance, physicochemical, and sensory appreciation, J. Agric. Food Chem. 60 (2012) 7761–7768. [15] A.L.-S. Eh, S.-G. Teoh, Novel modified ultrasonication technique for the extraction of lycopene from tomatoes, Ultrason. Sonochem. 19 (2012) 151– 159. [16] M. Ashokkumar, D. Sunartio, S. Kentish, R. Mawson, L. Simons, K. Vilkhu, C. Versteeg, Modification of food ingredients by ultrasound to improve functionality: a preliminary study on a model system, Innov. Food Sci. Emerg. 9 (2008) 155–160. [17] D. Pingret, A.-S. Fabiano-Tixier, F. Chemat, Degradation during application of ultrasound in food processing: a review, Food Control 31 (2013) 593–606.
Please cite this article in press as: C. Carrera et al., Ultrasound-assisted extraction of amino acids from grapes, Ultrason. Sonochem. (2014), http:// dx.doi.org/10.1016/j.ultsonch.2014.05.021