Quantitative Evaluation of the Pyruvic Acid Content in Onion Samples with a Fully Validated High-Performance Liquid Chromatography Method

Abstract

Onion cultivars with low pungency have been increasing in popularity because they are more attractive for fresh, uncooked use. The pyruvic acid amount is commonly used as a measure of onion pungency. For the first time the validation of a reversed-phase high-performance liquid chromatography method for the quantitative analysis of pyruvic acid in onion extracts was carried out. The established chromatography method based on the use of a 50 mM phosphate buffer (pH 2.5) eluent system provided good precision (RSD% in the range of 1.25–1.29, in the long-period) and accuracy (recovery% in the range of 97.70–111.62, in the long-period) as well as appreciably low limit of detection and limit of quantitation values (11.03 and 33.41 ng/mL, respectively). The validated method proved successful in the analysis of the onion extract from Dorata di Parma cultivar after a preliminary solid-phase extraction step. A significant adherence with results from previous spectrophotometric determinations turned out.

Keywords:

• Onion extract
• Reversed-phase high-performance liquid chromatography (RP-HPLC)
• Method validation
• Pungency
• Pyruvic acid
• Solid-phase extraction

INTRODUCTION

The recent literature is rich in both in vivo and in vitro studies reporting the anti-thrombotic-,^[1] hypolipidemic-,^[2,3] anti-diabetic-,^[4,5] anti-obesity-,^[6] antioxidant-,^[7–9] anti-inflammatory-,^[10] cancer chemopreventive-,^[11] and antiparasitic-^[12,13] properties of onion extracts. Moreover, some epidemiological studies suggest that a diet rich in onions may have a favorable effect on the risk of acute myocardial infarction^[14] and benign prostatic hyperplasia.^[15]

Onion’s health benefits have been mainly ascribed to two chemical classes of compounds, namely flavonoids and alk(en)yl cysteine sulfoxides (ACSOs).^[16] ACSOs are non-proteinogenic sulfur amino acids which are peculiar to the vegetative tissues of all Allium species and are indirectly responsible for their characteristic odor and flavor.^[17] (+)-S-trans-1-propenyl-L-cysteine sulfoxide (isoalliin)^[18] is the most abundant ACSO present in onions, and also represents the precursor of the lachrymatory factor (LF). ACSOs are stable in intact tissues, but undergo a rapid enzymatic hydrolysis by the enzyme alliinase (EC 4.4.1.4) as soon as the cells are damaged by mechanical chopping or maceration. Sulfenic acids (1) and α-iminopropionic acid (2) are the compounds first produced in alliinase-mediated ACSOs transformation (Fig. 1). Then (2), spontaneously hydrolyzes to give ammonia and pyruvic acid (PA), whereas (1) condenses to form thiosulfinates (TS), in turn generating thiosulphonates, mono-, di- and tri-sulphides as well as the LF.^[19] The consumer perception of the typical onion “pungency,” which is closely related to palatability, depends on these organosulfur metabolites, but since PA is also a final product of this metabolic cascade, its content is commonly used as a measure of onion pungency.

FIGURE 1 Enzymatic generation of pyruvic acid (PA) in onion.

A vast number of onion varieties are known exhibiting a great diversity in terms of pungency.^[16] Furthermore, the climatic conditions, the agronomic practices and soil type may affect the genetic information of the onion cultivar, thus producing a variety of locally present ecotypes endowed with particular organoleptic and nutritional characteristics. Recently, onion cultivars with low pungency (the so-called “sweet” onions) have been increased in popularity among the consumers because they are more attractive for fresh, uncooked use. To address these new market demands onion-growers and traders advertise often their products as low-pungency onions without, however, the support of any experimental data.

Several studies have been carried out along the years to quantitatively assay chemical components in onions. In this scenario, RP-HPLC methods have been successfully used by other authors to measure the PA levels in extracts from different varieties of onion.^[20–22] However, to the best of the author’s knowledge, none of them neither validated the developed method, nor employed solid-phase extraction (SPE) cartridges for sample clean-up, recovery, and concentration necessary for accurate quantitative measurements.

In the present article we describe for the first time the validation of a RP-HPLC method for the quantitative analysis of PA in onion extracts. We demonstrate herein that the fully validated method can be successfully applied in the analysis of real sample after a preliminary SPE step.

MATERIALS AND METHODS

Reagents, Solvents, and Standards for HPLC Method

All the reagents used to build-up the calibration curve were of analytical grade. PA, phosphoric acid, methanol (MeOH), and ammonium acetate (NH₄OAc) were purchased from Sigma-Aldrich (Milano, Italy). Sodium dihydrogen phosphate was purchased from J.T. Baker (Milano, Italy). HPLC-grade water was obtained from a New Human Power I Scholar water purification system (Human Corporation, Seoul, Korea).

Instrumentation

The analytical HPLC measurements were made on a Shimadzu (Kyoto, Japan) LC-Workstation Class LC-10 equipped with a CBM-10A system controller, two LC-10AD high-pressure binary gradient delivery systems, an SPD-10A variable wavelength ultra violet (UV)-visible detector and a Rheodyne 7725i injector (Rheodyne, Cotati, CA, USA) with a 20 μL stainless-steel loop. A Luna C-18 (Phenomenex, Torrance, CA, USA) 250 × 4.6 mm i.d., 5 μm, 100 Å was used as the analytical column. The column temperature was controlled through a Grace (Sedriano, Italy) heather/chiller (Model 7956R) thermostat. The Grace Pure SPE Anion-X 500 mg/3 mL solid phase extraction cartridges were purchased from Grace (Grace, Loreken, Belgio). The employed eluent system was degassed with 20 min sonication before use.

Onion Characteristics

The nine samples of onion (of the Rossa di Toscana variety) were provided by Cannara Onion Producers’ Union (Consorzio dei Produttori della Cipolla di Cannara). All the samples were cultivated in 2012 at a farm belonging to Cannara Onion Producers’ Union in the municipality of Cannara, a small town in the Umbria region of Italy, using the same environmental and agronomic conditions and harvested in July 2012 when 50% of the foliage had collapsed. Afterward, they were stored at an average ambient temperature of 20–25°C until analyzed. For the HPLC analysis, the normally edible parts of the onions were used, that is the bulb after discard of the outer dry skins, neck, and base. Onions were weighted and homogenized without any additional water in a domestic electric blender for 2 min. After 30 min at room temperature, the purée was filtered through paper filter by the aid of a vacuum. The total volume of the juice obtained was measured and then freeze dried before being submitted to SPE procedure. All the analyses were performed in triplicate and the results were expressed as mean ± standard deviation. The data were expressed as µmol of PA/g fresh weight (FW).

Sample Purification Via Solid Phase Extraction

The SPE cartridge was conditioned with MeOH (1 mL) and water (3 × 1 mL). The sample aqueous solution was loaded onto the cartridge. A washing step with a MeOH/water 10/90 (v/v) + 0.1 M NH₄OAc solution (2 × 1 mL) was carried out to remove interferences while retaining the acidic components of the onion extract. Elution was performed with 5% formic acid MeOH solution (8 × 0.5 mL). The purified samples were dried and stored at 4°C before the HPLC analysis, which was carried out by re-solubilizing the sample in mobile phase components.

RESULTS AND DISCUSSION

In the frame of a wide multidisciplinary study aimed at enlightening the peculiarities of three onion cultivars (Dorata di Parma, Borettana di Rovato, and Rossa di Toscana) when harvested in Cannara, we recently evaluated their pungency in terms of PA content via spectrophotometric analysis.^[23] Rossa di Toscana was found to contain the lowest PA amount (4.24 μmol/g FW), whereas no statistically significant differences emerged between the other two varieties (5.41 μmol/g FW for Dorata di Parma and 4.99 μmol/g FW for Borettana di Rovato). For the present study, the Dorata di Parma cultivar was selected as the real case example to assay the goodness of the validated HPLC method, by comparing the result with that previously obtained with the alternative analytical technique.

Method Validation

The calibration curve for PA was built by running the analyses with an eluent system consisting of 50 mM phosphate buffer (pH 2.5), flowed through the column at 0.2 mL/min and with a column temperature fixed at 15 °C. The wavelength of detection was set at 205 nm.

Selectivity

In order to identify the presence of interference peaks within the investigated analysis time, three repeated injections of the selected solvent blank (namely the eluent system) were consecutively run. The peaks clearly distinguishable from the base-line noise (with very small area values in arbitrary units) did not overlap that corresponding to PA. On this basis, the established method can be regarded as highly selective for the purpose of the present study.

Linearity

Five calibration standards, having PA concentration spanning within the range 0.01–1.00 mg/mL, were used. All the standard solutions were run in triplicate and the average of the peak area values was employed to build-up the regression line (Table 1). Specifically, the calibration curve (Eq. 1, Table 1) was constructed by plotting the concentration value of each standard solution (as independent variable) against the corresponding peak area value (in arbitrary units). The obtained mathematical model was characterized by a very appreciable linearity, as indicated by the correlation coefficient value (R² = 0.9997). The established method was then validated using an external set of two control solutions whose concentration were included in the above concentration range (Tables 2 and 3).

TABLE 1 Calibration data for pyruvic acid: regression equation, correlation coefficient (R²) value, explored linearity range, LOD, and LOQ values

Cpd	Regression equation	R^2	Eq. #	Linearity range (mg/mL)	LOD (ng/mL)	LOQ (ng/mL)
PA	y = 55069094.26 (523397.65) x + 557611.15 (30430.71)	0.9997	1	0.01–1.00	11.03	33.42

TABLE 2 Statistical analysis for pyruvic acid in the short period (intra-day precision and accuracy values)

Solution	Theoretical concentration (mg/mL)	Day	Mean observed concentration (mg/mL)	n^a	Precision (RSD%)	Accuracy (recovery %)
1	0.05	1	0.06	3	0.84	112.10
		2	0.06		0.52	111.72
		3	0.06		2.15	111.05
2	0.35	1	0.34	3	0.42	97.41
		2	0.34		1.41	96.78
		3	0.35		0.90	98.90

^aNumber of replicates.

TABLE 3 Statistical analysis for pyruvic aci in the long period (inter-day precision and accuracy values)

Solution	Theoretical concentration (mg/mL)	Mean observed concentration (mg/mL)	n^a	Precision (RSD%)	Accuracy (recovery %)
1	0.05	0.06	9	1.25	111.62
2	0.35	0.34		1.29	97.70

^aNumber of replicates.

LOD and LOQ

Very appreciable limit of detection (LOD) and limit of quantitation (LOQ) values were obtained by utilizing the mathematical model reported in Table 1. The LOD and LOQ values were derived from the following Eq. (2) and Eq. (3), respectively:^[24]

(2)

(3)

where C_LOD and C_LOQ refer to the sample concentration corresponding to the LOD and LOQ, respectively, σ_y is the standard deviation of the peak area values produced by repeated eluent (blank) injections, in the correspondence of the characteristic PA retention time, b is the slope of the calibration equation (Eq. 1, Table 1).

Intra-day and inter-day precision

The intra-day precision was assessed by running the two control solutions in triplicate (n = 3) within a period of approximately 2 h. This procedure was repeated for three consecutive days. The previously obtained mathematical model (Eq. 1, Table 1) was then used to calculate the concentration of the control solutions (mean observed concentration, Table 2).

The intra-day precision was determined as the relative standard deviation (RSD%) among the concentration values achieved from consecutive injections. For each control solution, the variation within replicate injections performed in a time-frame of three consecutive days (n = 9) was employed to determine the inter-day precision (Table 3). As shown in Table 2, a very appreciable variation in terms of RSD% values was maintained during the consecutive three days of analysis. This, in turn, indicates a high reproducibility of the adopted RP-HPLC method in the short-period. A satisfactory precision was also found when the long-term (inter-day) precision was considered (Table 3).

Intra-day and inter-day recovery

The recovery percentage (recovery%) approach was selected to estimate the accuracy of the adopted method. For each determination, the recovery% value was calculated through the following, Eq. (4).

(4)

where C_measured represents the sample concentration as calculated through the regression equation in (mean observed concentration), while C_theoretical corresponds to the concentration of the employed external test solution (theoretical concentration) Table 2.

Analogously to the estimation of the short- and long-term precision, the two control solutions of the external set were also used to estimate the intra-day and the inter-day accuracy. Accordingly, the former was determined by considering the three runs for each control solution obtained in a single day (n = 3), while the latter was calculated in the time-frame of the three consecutive days of analysis (n = 9).

The recovery% value was found to be generally higher for solution 1 than for solution 2 (Table 2), with more appreciable values obtained during day 3. Accuracy was also achieved in the long-period (Table 3).

In Fig. 2, the chromatographic profiles of the standard PA (Fig. 2a) and of the extracts from the Dorata di Parma cultivar before (Fig. 2b) and after (Fig. 2c) SPE-based purification are shown. The SPE sample preparation method was successful in purifying the compound and improving the quantitative measurement of the species of interest. Accordingly, an average PA concentration equal to 0.87 ± 0.02 mg/mL, corresponding to 5.34 ± 0.13 μmol/g, was estimated with the validated HPLC method, thus revealing a strict accordance with the previously developed and validated spectrophotometric method. The experimental evidence is in line with other studies in which significant adherence between results from spectrophotometric and HPLC determinations of the PA level in different types of onion bulbs were described.^[23]

FIGURE 2 Chromatographic trace of the A: standard PA, and of the extract from the Dorata di Parma cultivar B: before and C: after SPE.

CONCLUSIONS

The RP-HPLC developed method for the quantitative assay of the PA content in onion extracts can be useful from an industrial point of view. Indeed, the method could be fruitfully applied for the proper selection of the onion varieties and cultivars to be used for definite manufactory processes, such as the preparation of caramelized onion marmalades, ice creams, and liqueurs, to cite but a few. Moreover, the assessment of the characteristic onion sweetness with the use of a validated chromatography method, can promote and facilitate the commercialization of the vegetable, thus contributing to ameliorate the economic asset of the corresponding production area.

FUNDING

The authors gratefully acknowledge Fondazione Cassa di Risparmio di Perugia (Italy) for the financial support with grant 2013.0222.021.

ACKNOWLEDGMENTS

The authors are grateful to Cannara Onion Producers’ Union (Consorzio dei Produttori della Cipolla di Cannara) for the supply of the onion samples.

Additional information

Funding

The authors gratefully acknowledge Fondazione Cassa di Risparmio di Perugia (Italy) for the financial support with grant 2013.0222.021.