| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dairy Products Technology Center California Polytechnic State University San Luis Obispo, CA 93407
Corresponding author:
R. Jiménez-Flores; e-mail:
rjimenez{at}calpoly.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Abbreviation key: BGE = background electrolyte, CE = capillary electrophoresis, CTAB = hexadecyltrimethylammonium bromide, EOF = electroosmotic flow, f = response factor, i.d. = internal diameter, IS = internal standard, LAB = lactic acid bacteria, NSLAB = nonstarter LAB, PDC = 2,6-pyridinedicarboxylic acid, p.s.i. = pound square inch
Key Words: organic acids and amino acids • lactose • cheese • capillary electrophoresis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Organic acids may appear in cheese as a result of hydrolysis of milk fat during lipolysis (FFA such as acetic or butyric), normal bovine biochemical metabolism (citric, orotic, and uric), or bacterial growth (lactic, acetic, pyruvic, propionic, and formic). These are the major products of carbohydrate catabolism of lactic acid bacteria (LAB). The resulting acidity prevents the development of spoilage and pathogenic microorganisms, improving the hygienic quality of cheese. However, the ability of LAB to inhibit undesirable bacteria depends not only on the reduction of pH, but also on the sort of organic acids they produce (Izco et al., 2002). Quantitative determination of organic acids is important for monitoring bacterial growth and activity and for nutritional reasons and because organic acids contribute to the flavor and aroma characteristics of most cheeses (Fox, 1993; González de Llano et al., 1996).
Proteolysis is the most complex of the three events during ripening. It is also the major contributor to the changes taking place in the cheese matrix and occurs in most pressed-curd cheeses (Irigoyen at al., 2001). Casein breakdown liberates amino acids and peptides that directly contribute to cheese flavor. Because nonstarter LAB (NSLAB) possess a wide range of hydrolytic enzymes, they affect proteolysis and lipolysis during ripening and, therefore, the flavor of the cheese (Lane and Fox, 1996). Also, NSLAB adjuncts have been shown to accelerate the production of amino acids (Lynch et al., 1996). Furthermore, some amino acids (e.g., Glu, Leu, and Lys) have been linearly correlated with ripening time in Idiazábal and Ossau-Iraty cheese and are very helpful indicators of the degree of ripening (Izco et al., 2000).
These types of compounds have been analyzed by a variety of different analytical techniques. Until a few years ago, organic acids were commonly analyzed by chromatographic techniques. Most methods developed to analyze organic acids in dairy products are based on HPLC (Fernández-García and McGregor, 1994; Mullin and Emmons, 1997). Amino acids are analyzed by HPLC methods with pre- or postcolumns derivatization and ultraviolet or fluorescence detection (Izco et al., 2000). On the other hand, the analysis of carbohydrates normally is performed by HPLC with a refractive index detector, whereas the analysis of lactose in milk may be carried out by polarimetric, gravimetric or enzymatic methods (AOAC, 1995). Because analytical instrumentation, separation columns, mobile phases, and detectors are different, at least three different systems and methods are necessary for a complete analysis. Therefore, an easy, efficient, and rapid simultaneous analysis method for these compounds would be a great benefit.
Capillary electrophoresis (CE) has emerged as a powerful separation technique that can provide high resolution and efficiency, offering great potential for rapid detection and quantification. Recently we developed a CE method to analyze 11 metabolically important organic acids (Izco et al., 2002). The fact that the amino acids and lactose can be charged (using pH values higher than their respective pKa values would have a net negative charge) suggests that they could be separated by CE. By applying a negative voltage, the anions migrate toward the detector situated at the anode end of the capillary. Addition of modifiers, such as hexadecyltrimethylammonium bromide (CTAB), reverse the direction of the electroosmotic flow (EOF) inside the capillary and promote the comigration of the analytes with the EOF, thus speeding up the analysis. Since organic acids, amino acids, and lactose have little or no UV absorbance, detection could be accomplished by indirect UV by using a background electrolyte (BGE).
The aim of this work was to establish a CE method for the simultaneous separation and quantification of organic acids, lactose, and several amino acids in cheese and in other dairy products.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Reagents
Oxalic, formic, succinic, orotic, uric, pyruvic, and lactic acids, Asp, Glu, Tyr, Gly, Ala, Ser, Leu, Phe, Lys, Trp, sodium propionate, and lactose monohydrate, were obtained from Sigma (St. Louis, MO); citric, acetic, and butyric acids were purchased from Fisher Scientific (Pittsburgh, PA). Ultrapure water (18.2 M
) prepared by treating deionized water with a Barnstead/Thermolyne System (Dubuque, IA) was used to prepare all solutions. Individual stock solutions of each compound at 1000 ppm (10,000 ppm in case of lactose) were prepared by disolving the proper quantity in 4.5 mM H2SO4 containing 15 ppm of boric acid as an internal standard (IS), except orotic and uric acids, Asp, Glu, and Tyr, which were prepared in 0.1 N NaOH. The reagents used to prepare the running buffer were analytical or reagent grade: CTAB was purchased from Sigma and 2,6-pyridinedicarboxylic acid (PDC) was obtained from Aldrich (Milwaukee, WI).
Electrophoretic Procedures and Conditions
The conditions used to analyze the samples were based on the results obtained previously (Izco et al., 2002) with some modifications. The BGE was prepared daily with 20 mM PDC and 0.5 mM CTAB. The pH of the buffer was adjusted at 12.15 with 1 M NaOH. The separations were carried out on fused-silica capillaries with 105 cm of effective length x 75 µm i.d. Before first use, a new capillary was pretreated with 0.25 M NaOH for 10 min, followed by water for 10 min and BGE for 10 min. Before each run, the capillary was washed for 3 min with 0.25 M NaOH and preconditioned with run electrolyte for 3 min at 60 p.s.i. Before storing it, the capillary was rinsed for 3 min with water and dried by passing air for 1 min at 60 p.s.i.
The sample was injected by hydrodynamic injection for 10s at 1 p.s.i.. The separation was performed at –25 kV and the capillary was thermostated at 25°C. The wavelength for indirect UV detection was selected at 230 nm, and the signal with negative peaks was inverted to obtain a more familiar electropherogram to integrate and process. However, orotic and uric acids absorb at 230 nm, drastically decreasing the sensitivity for orotic acid, and uric acid appeared as a negative peak. Nevertheless, our CE System is equipped with a photo-diode array detector that allows select different wavelengths. Because these two compounds can be easily detected at 300 nm, we set the detector at both 
= 230 and 300 nm. In this form, for the same analysis we obtained one electropherogram at 300 nm to quantify orotic and uric acids and another at 230 nm for the rest of compounds (Figure 1
).
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
). An unknown peak appears in the electropherogram due to the running buffer. This peak (sometimes detected as a jump of the baseline, see Figure 3, below, between peaks 2 and 3) appears only when high pH values running buffers are used, due to the high quantity of NaOH added to adjust the pH of the electrolyte. We have previously detected this anomaly (Izco et al., 2002). Butyric acid and Ala were not totally resolved in commercial samples (Figure 2), and it seems that Ala comigrated with an unknown peak detected in the milk after the IS (Figure 3). Also, other important amino acids such as Pro, Gln, Asn, Val, Met, or His were tested. However, these compounds appeared after the peak corresponding to Ser, and some of them comigrated with a negative unknown peak that appears in dairy samples, making their quantification impossible (see Figure 2). Also, the resolution for some of these amino acids was poorer; therefore, we decided to validate the method for the 10 amino acids selected, which are perfectly separated and identified. Some of the amino acids selected are present in different varieties of cheese, e.g., Glu, Leu, Phe, and Lys have been shown to be major in Ossau Iraty cheese (Izco et al., 2000) with a clear relationship between their concentration and the ripening time.
|
|
show the results obtained in the validation of the analytical technique. The calibration curves were calculated by analyzing in duplicate 5, 10, 20, 50, and 100 ppm of each compound, except in the case of lactose, which was calculated using 50, 100, 200, 500, 1000, 5000, and 10,000 ppm in order to cover the normal concentration of lactose in milk (Table 1). Boric acid was used as an IS, and the data points from calibration curves were subjected to a least square regression analysis. The slope (a), intercept (b), and coefficient of determination (r2) were calculated. Detection limits were estimated as (3 x a/b) x 1/
n, where a is the independent term of the curve, b the slope, and n is the number of replicates. The r2 values calculated (Table 1) were better than 0.99 for all of the analytes (except for uric acid, 0.9858), showing that the linearity of the method was good and, therefore, suitable for quantification, with detection limits as low as 0.2•10–2 mM in the case of the citric acid and Gly.
|
|
). However, the CV values for peak areas were better than 5% for all of them (including lactose), except oxalic acid (6.4%). That higher CV calculated for oxalic acid could be because this acid migrates close to the big peak of sulfate, which makes its quantification more difficult. The results obtained are better than those previously obtained in the analysis of organic acids (Izco et al., 2002), where CV values around 10% for uric and citric acid, and better than 7% for the rest, were calculated. Consequently, the analytical conditions used in this work have clearly improved the CE method. Similar results have been reported by Wu et al. (1995) or Galli et al. (2000), using phosphoric acid as BGE to analyze some of these organic acids in different matrixes, and slightly better than those recorded by Chen et al. (1997), who calculated CV values of 6% using phthalate as electrolyte.
Since the recovery in the extraction of organic acids from milk and cheese with 4.5 mM H2SO4 has been reported in the literature (Fernández-García and McGregor, 1994; González de Llano et al., 1996), our proposal was not to validate the method of extraction, but to estimate the accuracy of the analytical technique. Blue cheese was extracted, and the sample obtained was spiked with 10 and 20 ppm of five organic acids (formic, citric, uric, acetic, and lactic) and five amino acids (Asp, Tyr, Leu, Phe, and Trp). Recovery was calculated for both cases (10 and 20 ppm) using the responses of the peaks and the average of the percentage recovery calculated are shown in Table 2. Since no lactose was detected in Blue cheese, cream cheese was used to calculate recovery by spiking lactose at 2500 and 1250 ppm. Recovery was close to 100% for the compounds indicated. These results are similar to the recovery calculated when extracting organic acids from milk, cheese, or yogurt with H2SO4 (Fernández-García and McGregor, 1994; González de Llano et al., 1996). Also, by using the high shear blender (Ultra Turrax), we shortened the extraction time from 1 h (González de Llano et al., 1996) to 15 min.
Application for Commercial Dairy Samples
For all of the commercial samples, one extract was prepared from each sample, and duplicate analysis was conducted on each extract. Figure 2 shows the electropherograms corresponding to some of the commercial samples tested. The results obtained are shown in Table 3. Two electropherograms ( = 230 nm) corresponding to pasteurized whole milk and milk ripened for 30 min with starter bacteria during manufacture of Cheddar cheese are shown in Figure 3. To simplify the figure, only the electropherogram at = 300 nm corresponding to pasteurized milk have been included (the electropherograms were similar). Besides lactose (not shown in the figure), several organic acids (oxalic, formic, citric, orotic, uric, acetic, pyruvic, propionic, and lactic), three amino acids (Asp, Glu, and Gly), and several unknown peaks can be simultaneously detected and quantified in whole milk by this CE technique. As observed in Figure 3, we detected an increase of acetic and lactic acids produced by the metabolism of starter culture bacteria after growing for 30 min by this technique (peaks 10 and 14, respectively).
|
= 230 nm obtained when analyzing Blue cheese is shown in Figure 4. Twenty-one compounds were identified. A big peak corresponding to phosphate appears in the electropherogram, making the quantification of succinic acid difficult. Because the concentration of inorganic phosphate in milk is 19 to 23 mM (Walstra et al., 1999), it is logical to find phosphate in cheese. Several small peaks and one big negative peak have been detected, but they could not be identified. On the other hand, there are amino acids appearing between the peaks 18 and 19 that have not been labeled because some of those amino acids (Asn, Val, His, Pro, Met) coeluted as a single peak, and we could not ensure their identification.
|
80 mg/100 g DM) was slightly higher than the values recorded by Fernández-García and McGregor (1994) when using HPLC to quantify organic acids in yogurt. The quantity of orotic acid in milk depends on the cow’s origin, diet, and lactation (Gaia et al., 2000). It is an intermediate product in the synthesis of nucleotides and a growth factor for yogurt starter cultures; a decrease of up to 48% in orotic acid content during manufacturing and storage of yogurt has been reported (Arla, 1982). Concentration of citric acid remaining in yogurt (1386.6 mg/100 g of DM) was similar to other results (2.3 mg/g) recorded in the literature (Fernández-García and McGregor, 1994).
Acetic acid is another important organic acid detected in yogurt, probably formed as product of the fermentation of lactose and citric acid. As observed in Table 3, lactose has not been completely consumed by the starter bacteria in yogurt. In the same way, an important quantity remains intact in fresh cheese (2298.7 mg/100 g of DM). To test the sensibility of this method, the same fresh cheese was stored at 32°C for 2 d, and that period was enough to clearly detect a decrease of 15% of lactose and 24% of citric acid, whereas concentration of acetic acid increased approximately threefold. One advantage of the proposed CE method with respect to HPLC is that this method allows simultaneous analysis of lactose and those organic acids in cheese (see Figure 2). In the case of HPLC, those compounds must be analyzed separately to take advantage of the specificity and sensitivity of the detectors, and to avoid problems of coelution of some of them (Mullin and Emmons, 1997).
Lactose was not present in the aged cheese varieties analyzed (Table 3). In fact, 98% of lactose is removed in the whey as lactose or lactate, and the complete and rapid metabolism of residual lactose remaining in the curd (0.8 to 1.5%) and its component monosaccharides is essential to produce good quality cheese (Fox, 1993).
Lactic acid is the organic acid most abundant in all cheeses tested (Table 3). Acetate may be produced by starter LAB from lactose or lactate or citrate or amino acids and must make some contribution to Cheddar cheese flavor. Nevertheless, the oxidation of lactate to acetate in cheese can be carried out by NSLAB (Fox, 1993). Because Roncal cheese is made with raw milk and the population of NSLAB is abundant, the concentration of acetic acid found in Roncal cheese (239.7 mg/100 g of DM) may be higher than the rest. Acetic acid has been previously detected in Roncal cheese, and the importance of NSLAB for the development of the characteristic flavor of Roncal cheese has been reported also (Ortigosa et al., 2001). Perhaps, this method could be useful for the characterization of cheese made with raw milk from the same type of cheese, but manufactured with pasteurized milk. In the same form, citrate can be metabolized by starter bacteria to generate acetoin, which is reduced to butan-2,3-diol (diacetyl). In Roncal cheese, these components are reduced to butan-2-ona and butan-2-ol by microorganisms present in raw milk. All these compounds are very important flavor components of Roncal cheese (Ortigosa et al., 2001). On the other hand, due to CO2 production, rapid citrate metabolism can be responsible for the undesirable openness and floating curd in Cheedar cheese. Citric acid is not the first energy source of bacteria, but can be metabolized very rapidly by Lactococcus lactis subsp. diacetylactis or Leuconostoc spp. in Cheddar cheese. Depending on the starter used, citrate can remain constant at 2% (wt/wt) up to 3 mo of ripening, and decrease to 0.1% (wt/wt) at 6 mo (Fox, 1993). Citrate in cheese presumably reflects the concentration of colloidal citrate in milk. The concentration of citric acid in the sample of Cheddar cheese was 195.5 mg/100 g of DM, which is into the normal range (0.2 to 0.5%, wt/wt) of citrate content in Cheddar cheese (Fox et al., 1993).
Proteolysis is the major contributor to the changes taking place in the cheese matrix and occurs in most pressed-curd cheeses. The main proteolytic agents are: the natural proteases of the milk, milk-clotting enzymes retained in the curd, and proteases and peptidases from starter and nonstarter bacteria. However, FAA are released mainly by the action of microbial enzymes from the starter culture (Irigoyen et al., 2001). Soluble Tyr has been used as rapid method to monitor cheese ripening because its concentration is higher in cheeses in which ß-casein is extensively hydrolyzed, e.g., Blue, Cheddar, and Parmesan (Marcos and Esteban, 1993). The concentrations of this amino acid in these cheeses were 70.6, 56.0, and 197.7 mg/100 g of DM, respectively, whereas 45.5 mg/100 g of DM was found in Roncal cheese (See Table 3). In fact, the two fraction of ß-casein are not extensively hydrolyzed during ripening of Roncal cheese (Irigoyen et al., 2000), perhaps due to higher resistance of ovine ß-casein fractions to proteolysis (Izco et al., 1999).
Many of the major amino acids quantified in the cheeses tested have been also reported previously, e.g., Glu, Leu, Phe, Leu, and Lys are some of the major amino acids in Parmiagiano Reggiano cheese, a high proteolytic cheese very similar to Parmesan cheese (Battistotti and Corradini, 1993); they are also major in Idiazabal and Ossau-Iraty cheese (Izco et al., 2000). Compared to the proteolysis in other cheeses, that in mold-ripened cheeses (particularly in Blue cheeses) is higher (Gripon, 1993). Extreme proteolysis results in Blue cheese and Parmesan having higher concentration of total free amino acids (measured as the sum of the free amino acids individually quantified) than the rest of the cheeses.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The proposed method appears to be an alternative to other analytical techniques, with the additional advantages of low solvent consumption (milliliters per day vs. liters per day for HPLC), short time of analysis, no hazardous solvents, and low costs compared with others (more than 1000 analysis were run with the same capillary without lost of resolution; in the case of HPLC, columns must be deeply cleaned and regenerated after a few analysis). Nevertheless, the principal advantage is that the great versatility of the method allows analysis of all those compounds simultaneously, whereas to analyze them by HPLC different columns, buffers, and chromatographic systems are necessary. The procedure offers faster and simpler sample preparation for the analysis, e.g., HPLC analysis of amino acids requires precipitation of the protein and pre- or postcolumn derivatization of the previously extracted free amino acids to detect them.
It has been demonstrated that the proposed technique yields different CE patterns that can be used as "reference fingerprints" for cheese varieties. The lactose and organic acid content can be used to monitor the fermentation process, providing additional information about the type of fermentation, whereas the amino acids contents are reliable indices of the proteolysis and therefore, of the ripening of cheese.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication March 6, 2002. Accepted for publication March 21, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. W. Park, J. H. Lee, and S. J. Lee Effects of frozen and refrigerated storage on organic acid profiles of goat milk plain soft and Monterey Jack cheeses. J Dairy Sci, March 1, 2006; 89(3): 862 - 871. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ji, V. B. Alvarez, and W. J. Harper Influence of Starter Culture Ratios and Warm Room Treatment on Free Fatty Acid and Amino Acid in Swiss Cheese J Dairy Sci, July 1, 2004; 87(7): 1986 - 1992. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
