Content of Biogenic Amines in Table Olives

Transcription

1 111 Journal of Food Protection, Vol. 63, No. 1, 2, Pages Copyright, International Association for Food Protection Content of Biogenic Amines in Table Olives P. GARCÍA-GARCÍA, M. BRENES-BALBUENA, D. HORNERO-MÉNDEZ, A. GARCÍA-BORREGO, AND A. GARRIDO-FERNÁNDEZ* Food Biotechnology Department, Instituto de la Grasa (C.S.I.C.), Avda. Padre García Tejero, 4, 4112 Seville, Spain MS : Received 27 April 1999/Accepted 17 July 1999 ABSTRACT Content of biogenic amines in flesh and brines of table olives was determined by high-pressure liquid chromatography analysis of their benzoyl derivatives. No biogenic amines were found in the flesh of fresh fruits at any stage of ripeness. Contents of biogenic amines in green or stored olives increased throughout the brining period but were always higher in the former. Putrescine was the amine found in the highest concentration. Small quantities of cadaverine were found in the samples taken after 3 months of brining. This compound and histamine, tyramine, and tryptamine were also found in samples taken after 12 months. Gordal cultivar showed the highest contents, followed by and Hojiblanca. No relationship was found between contents of biogenic amines and lactic acid production or table olive spoilages, although zapatera olives had considerably higher amounts than those brines that had undergone a normal process. Concentrations in directly brined olives were markedly lower than contents in olives. With respect to partition between flesh and brine, there was equilibrium between both media in the case of olives, whereas the contents in directly brined olives were higher in flesh than brine. Biogenic amines are low-molecular-weight organic bases that possess biological activity. They can be formed and degraded during normal metabolic activity in animals, plants, and microorganisms and are usually produced by the decarboxylation of amino acids. Diamines and polyamines are known to be generated as the result of endogenous amino acid decarboxylase activity in raw food material (19) or by the growth of decarboxylase-positive microorganisms under conditions favorable to enzymatic activity (9). In fact, the presence of large amounts of these amines has been reported only in aged, fermented, or spoiled products. The production of cadaverine, histamine, putrescine, spermidine, and tyramine in concentrations ranging from 1 to 1 mg/kg in lactic acid fermented vegetables (such as carrot and red beet) is well known (1). Kuensch et al. (15) found considerable amounts of putrescine (up to 15 mg/ kg) during the initial stage of sauerkraut fermentation, with histamine and tyramine appearing toward the end of the process. Putrescine, spermidine, and spermine are present at very low concentrations in developing olive tissues, although salt stress has been reported to cause an accumulation of spermidine and spermine in the fruit (21). Increased fruit set has been reported in olives after application of putrescine at high concentration during flowering (3). In Spain, table olives (Olea europaea L.) are produced as two main commercial presentations: green and ripe (black) olives in brine. The procedure for preparing the first consists of treating the fruits with dilute NaOH solutions, followed by one or two water washes to eliminate * Author for correspondence. Tel: ; Fax: ; garfer@cica.es. the excess alkali; the fruits are then put into a NaCl solution (brine) in which they undergo a lactic fermentation (2, 6, 7). Sometimes spoilage appears during the fermentation. The most frequent spoilages are palmiche, butyric, and zapatera spoilages. nd butyric spoilages usually appear during the first phases of the process and are related to the absence of a proper growth of lactic acid bacteria. Zapatera spoilage may appear at the end of the process due to the growth of microorganisms of the genera Propionibacterium and Clostridium in a specific sequence (5 7, 18). Propionic acid bacteria appear first and cause a rise in ph, which allows Clostridium species to grow and produce the malodorous substances typical of zapatera. Butyric spoilage is characterized by the presence of butyric acid (4, 6, 7), whereas the unpleasant odor in zapatera spoilage has been related to the formation of cyclohexanecarboxylic acid (16). Processing of ripe (black) olives includes a prior stage of storage in an aqueous solution. The olives are put into a 4 to 1% (wt/vol) NaCl brine that contains acetic acid added in a concentration ranging from.3 to 1.% (vol/ vol) to control the ph. Sometimes salt is completely absent (5 7). The most common spoilage that affects the fruits during this period is softening, which has been related to cellulolytic bacteria (17) and/or fermentative yeasts (22). In fermented green table olive brines, Iñigo and Bravo-Abad (12) found very low concentrations of biogenic amines (.5 to 2. ppm, expressed as histamine), although this concentration increased to 5. ppm in zapatera spoiled olives. The colorimetric method used was not selective and did not distinguish between specific amines. More recently, Hornero-Méndez and Garrido-Fernández (1, 11) developed a high-pressure liquid chromatography

2 112 GARCÍA-GARCÍA ET AL. J. Food Prot., Vol. 63, No. 1 FIGURE 1. HPLC chromatogram for a standard mixture of benzoylated amines. Peak identity: 1, putrescine; 2, cadaverine; 3, tryptamine; 4, -phenylethylamine; 5, spermine; 6, 1,8-diamineoctano (internal standard); 7, spermidine; 8, histamine; 9, tyramine; and 1, agmantine. Detection wavelength: 225 nm. (HPLC) method that allows the detection and quantification of eight biogenic amines found in olive brine samples. Putrescine and cadaverine, as well as low concentrations of tryptamine, -phenylethylamine, and histamine, were observed in zapatera green olives. Agmantine, spermine, and spermidine were not detected. No information is available on the content of biogenic amines in processed ripe olives. The aim of this work was to study the content of biogenic amines in fresh table olive fruits and to follow its changes during the fermentation process of green olives and the storage stage of ripe olives. Correlations between biogenic amine contents in brines and flesh and biogenic amine concentrations in brines of spoiled olives have also been investigated. MATERIALS AND METHODS Fresh olives. Olives (O. europaea L.) of, Gordal, and Hojiblanca cultivars were used. They were harvested at two ripeness stages. The first samples were taken when fruits had reached complete development but still had a green surface color; the others were taken when the olives were completely black. Four samples of each ripeness stage and cultivar were collected in different orchards of Seville and Córdoba, Spain. Brines and fermented or stored olives. Samples of both brines and olives were taken from industrial fermenters (16,- liter capacity) (6, 7) in selected table olive processing companies. Samples of green olives were from Gordal,, and Hojiblanca cultivars. Those from the preservation stage of ripe olives were of Hojiblanca and cultivars. Sampling was performed two times after brining: at 3 and 12 months. Local processors supplied spoiled brines. Biogenic amines analysis. The simultaneous determination of putrescine, cadaverine, tryptamine, -phenylethylamine, spermine, spermidine, histamine, tyramine, and agmantine was made following the benzoyl derivative method of Hornero-Méndez and Garrido-Fernández (11) for table olive brines, with slight modifications. FIGURE 2. Changes of biogenic amines content in brines of green table olives of Gordal,, and Hojiblanca cultivars during processing. The values are the means (and standard deviations) of 3 samples. See Materials and Methods for equivalences between symbols and biogenic amines. Reagents. Reference samples of tryptamine hydrochloride, -phenylethylamine hydrochloride, putrescine dihydrochloride, cadaverine dihydrochloride, histamine dihydrochloride, tyramine hydrochloride, spermidine trihydrochloride, and spermine tetrahydrochloride were purchased from Sigma Chemical Company (Madrid, Spain); agmantine sulfate, 1,8-diamineoctano (internal standard), and benzoyl chloride (ACS) were purchased from Fluka (Buchs, Switzerland). Methanol was HPLC grade (Teknokroma, Barcelona, Spain) and used without further purification. Deionized water was obtained with a MilliQ 5 system (Millipore, Marlborough, Mass.). Extraction of biogenic amines from olives. The procedure used was based on that described by Hornero-Méndez and Garrido-Fernández (1). Olives were pitted and triturated to produce a paste. Ten grams of the homogenized paste, added with 1 l of the internal standard solution (1 g/ml), were extracted by shaking with 25 ml of 5% (wt/vol) trichloroacetic acid. The suspension was then filtered into a standard flask, and the residue washed with 5% trichloroacetic acid solution to bring the volume up to 5 ml. Derivatization procedure. One milliliter of brine (and 2 l of internal standard solution, 1 g/ml) or 5 ml of the flesh-extracted solution was taken for the reaction. The procedure was that described by Hornero-Méndez and Garrido-Fernández (11), except that the solid residue was dissolved in 1 ml of methanol. Apparatus. The HPLC analyses were performed using a Waters 717 plus autosampler, a Waters 6 E pump, and a Waters 996 photodiode array detector operated with Millennium 21 software (Waters Inc., Milford, Mass.). The column, purchased from Hewlett-Packard (Avondale, Pa.), was packed with reversed-

3 J. Food Prot., Vol. 63, No. 1 BIOGENIC AMINES IN TABLE OLIVES 113 FIGURE 3. Distribution of putrescine and cadaverine contents in the 3 sample brines of green table olives of Gordal,, and Hojiblanca cultivars during fermentation. phase C18 Spherisorb ODS-2 (5 m particle size, 12.5 by.4 cm). A bench-top centrifuge, model Micro-Centaur (MSE Scientific Instruments, Sussex, UK), was used to preclean the HPLC samples. Chromatographic conditions. Chromatographic separation was achieved using a binary gradient elution with methanol (eluent A) and deionized water (eluent B). To improve separation, the gradient used by Hornero-Méndez and Garrido-Fernández was slightly changed (11). The gradient started at 6% A, was increased linearly to reach 7% in 4 min, and then changed to 85% in 5 min. These conditions were maintained for 6 min. The flow rate was 1. ml/min, and the injection volume was 5 l. After each analysis, the column was flushed with 1% methanol for 2 min to clean it before the next injection. The initial conditions were reached again in 1 min. Figure 1 shows a typical chromatogram from the analysis of a standard mixture of amines. Chemical analysis of brines. Both lactic and acetic acids were analyzed by HPLC following the method described by Montaño et al. (16). Experimental design and statistical analysis of results. Distribution of biogenic amines in stored and fermented olives was determined by selecting six representative table olive industries and five randomly taken samples of each style and cultivar. Means, standard deviations, and histograms were used to estimate distribution of biogenic amines. Data were also analyzed by two-way analysis of variance, considering cultivars and industries as fixed and random factors, respectively. The Statistica software package (2) was used to achieve the above-mentioned studies. RESULTS AND DISCUSSION Fresh olives. No biogenic amines were found in the flesh of fresh fruits at any stage of ripeness. Thus, the raw material of or black (ripe) table olives does not contribute to the content of the fermentation or storage brines or of their respective fruits. Changes of amines in fermentation or storage brines and fruits. The average biogenic amine composition for 3 industrial samples of Gordal,, and Hojiblanca cultivars was diverse (Fig. 2). Biogenic amines content increased through the brining period. The highest concentrations were always for putrescine. In addition to putrescine, small quantities of cadaverine were found in some samples taken after 3 months of brining. As fermentation progressed, biogenic amines content increased. Thus, in samples collected after 12 months of brining, concentrations of putrescine and cadaverine were higher than in the first sampling. By that time, the presence of histamine, tyramine, and tryptamine was also detected, although in amounts that were always lower than 3 mg/liter and statistically the same (P.5) for the three studied cultivars. The same phenomenon has been reported in sauerkraut, in which histamine and tyramine concentration increased at the end of fermentation (15). Putrescine content varied greatly in all cultivars and in both samplings (Fig. 2). This was probably due to the different processing conditions in the various processing plants and to the different microorganisms and populations that can be observed in the spontaneous lactic acid fermentation process (6, 7). Concentration of putrescine in Gordal brines was very high with respect to and Hojiblanca after 3 months of brining. Later, the same tendency was maintained, although differences between Gordal and were not significant after 12 months of brining. The lowest concentration of this amine was always observed in Hojiblanca brines. Formation of biogenic amines during beer production has been demonstrated to be due to the growth of lactic acid bacteria (13, 14). The mechanism could be similar in table olives. Differences between Gordal and Hojiblanca cultivars might be related to their respective amino acid

4 114 GARCÍA-GARCÍA ET AL. J. Food Prot., Vol. 63, No. 1 TABLE 1. Putrescine content in green olive brines and flesh of Gordal cultivar after 12 months of brining a Olive flesh moisture (%, wt/wt) 71.2 (.5) 72.3 (.6) 7.1 (.6) 7.7 (.3) (.3) 7.1 (.7) 7.1 (.6) 71.2 (.8) 71.5 (.2) Brine (mg/liter) 58.9 (.8) 57. (.7) 62.8 (2.5) 55.7 (2.5) 55.1 (2.2) 57.1 (1.4) 55.5 (.3) 56.9 (.8) 64.4 (.5) Putrescine Olive flesh (mg/kg) 42.4 (.4) 41. (.6) 43.9 (.8) 39.6 (.5) 4.6 (1.5) 39.9 (.8) 4.1 (1.88) 4.55 (2.1) 46.3 (1.6) Theoretical in flesh (mg/kg) b a Standard deviation is given in parentheses. b (Theoretical putrescine in flesh) (Putrescine in brine) (Flesh moisture/1). compositions and/or fermentation intensities. Usually, the process is more active in the former due to its higher sugar concentration (6) and the higher environmental temperatures during the initial phases of its fermentation. The Gordal is the first cultivar to reach the appropriate green ripeness stage, and its processing usually begins in early September. Hojiblanca ripens later, and its picking period is at the end of October or even in November, when temperatures are lower (1 to 15 C). However, irrespective of the cultivar, no correlation between biogenic amines content and lactic acid concentrations has been found in samples taken 3 months after brining. Furthermore, production of lactic acid had practically finished by that time, but formation of biogenic amines apparently continued. So, despite the above-mentioned relationship observed in other products, biogenic amines can hardly be related to the growth of lactic acid bacteria in table olives. Late biogenic amine formation might come from nitrogenous material produced during microbial growth or lysis. Distribution of putrescine and cadaverine in the green olive brines was diverse (Fig. 3). Gordal cultivar showed the widest distribution in the first sampling, with the highest frequency in the range of 25 to 4 mg/ liter. and Hojiblanca cultivars had a very simple distribution more than 5% of samples had very low amounts of amines ( to 1 mg/liter) and the rest had low amounts (1 to 25 mg/liter). After 12 months of brining, the distribution changed, and more samples had higher contents. Hojiblanca samples contained between 1 and 4 mg/ liter; was similar to Gordal, although it showed a higher frequency in the 1 to 25 mg/liter range of concentrations. The highest frequency for Gordal was in the range of 4 to 6 mg/liter. Cadaverine was practically absent in the first sampling (Fig. 3), but small quantities appeared after 12 months. There were no important differences between cultivars with respect to the content of this amine, most of the samples being in the range of to 1 mg/liter. In green table olives, equilibrium between flesh and brine is usually found for solutes such as acetic or lactic acids, salt, and other compounds (6, 7). The same behavior is observed with biogenic amines. Considering the flesh moisture, calculated theoretical putrescine concentrations agree fairly well with those found analytically (Table 1). TABLE 2. Putrescine and cadaverine concentrations in brines of spoiled olives Spoilage Process Cultivar Putrescine (mg/kg) Cadaverine (mg/kg) Butyric b Butyric b Butyric b Gordal Gordal Gordal a Samples collected at 3 months. b Samples collected after 1 months.

5 J. Food Prot., Vol. 63, No. 1 BIOGENIC AMINES IN TABLE OLIVES 115 The relationship between spoilages of green olives and biogenic amines was diverse. Spoiled fruit brines contained mainly putrescine and cadaverine (Table 2). In palmiche spoilage, only small amounts of putrescine were observed, and in one sample no biogenic amines were detected. In butyric spoilage, concentrations of putrescine and cadaverine, as well as the other amines found (histamine, tyramine, and tryptamine), were similar to those observed in normal fermentations (Figs. 2 and 3). Thus, palmiche and butyric spoilages are apparently not related to abnormal formation of biogenic amines. Zapatera spoilage had the highest contents of cadaverine (above 9 mg/liter). Such levels are considerably higher than those found in proper fermentation processes (Figs. 2 and 3). No differences with respect to concentrations of the other biogenic amines were found. Thus, it can be stated that the presence of zapatera spoilage can be revealed by the high production of cadaverine, in addition to the other previously established characteristics (2, 14, 22). FIGURE 4. Changes of biogenic amines content in the storage stage brines of ripe olives of Hojiblanca and cultivars. Values are the means (and standard deviations) of 3 samples. See Materials and Methods for equivalences between symbols and biogenic amines. Storage stage brines of ripe olives. Concentrations of biogenic amines found in the brines of stored black (ripe) olives were always lower than those in green olives (Fig. 4). The effect of processing can be observed by comparing the amine content for the same cultivar (Hojiblanca) in the two processing systems (Figs. 2 and 4). The storage brines of ripe olives not only had low amounts of putrescine and cadaverine, but also histamine and tyramine were completely absent. This might be due to the different characteristics of the respective processes. Spontaneous or induced lactic fermentation is usual in green olives but rarely observed in the storage stage of ripe olives (2, 6, 7). Furthermore, high acetic acid concentration usually prevents any kind of fermentation (2). In the first sampling (after 3 months of brining), no biogenic amines were detected in Hojiblanca cultivar brines., however, had moderate amounts of putrescine and small quantities of cadaverine a similar pattern to that found in green olives. In the second sampling (12 months of brining), the amine content always increased, and in cultivar, the presence of histamine and tyramine was also observed. The categorized histogram for putrescine (Fig. 5) shows that generally had the highest contents, although the maxima were similar. No specific distribution was observed. Thus, in general, the presence of biogenic FIGURE 5. Distribution of putrescine content in the storage stage brines of ripe olives of Hojiblanca and cultivars.

6 116 GARCÍA-GARCÍA ET AL. J. Food Prot., Vol. 63, No. 1 TABLE 3. Putrescine contents in the storage brines and flesh of cultivar ripe olives after 12 months of brining a Brine (mg/liter) 6.2 (.1) 6.1 (.3) 5.1 (.7) 7. (.2) 5.8 (.2) 6.6 (1.3) 6.5 (.1) 5.6 (.7) 6.5 (.1) a Standard deviation is given in parentheses. Olive flesh (mg/kg) 8.1 (.8) 8.2 (1.6) 7.2 (1.9) 8.8 (1.3) 9.1 (.5) 1. (.2) 7.2 (.4) 5.9 (.2) 7.2 (.4) amines in normal processes does not follow any pattern, and it is not yet possible to relate their formation to a particular characteristic. Contents of biogenic amines in stored ripe olives were higher in flesh than brine (Table 3). Thus, there was no equilibrium between fruits and the surrounding medium in this case. Apparently, endogenous enzyme or some microorganisms that colonized the pulp produced the amines. Retention within the fruit was also favored by the low permeability of stored olives, whose skins are still intact at this stage. Lye treatments, which partially destroy the skin, are applied later, during the darkening (oxidation) process. This lack of equilibrium between flesh and brine is another characteristic that distinguishes the amine behavior pattern in stored ripe olives and green olives. The brines of spoiled (soft centers) preserved olives had higher levels of putrescine ( 19 ppm) and cadaverine ( 3 ppm) (Table 2) than samples without spoilage (2.1 and.4 ppm, respectively) (Fig. 4) for the same brining period. Thus, the abnormal fermentation responsible for the appearance of soft centers might eventually cause an increase in the concentration of such amines. This study has established that biogenic amines can be formed during the spontaneous fermentation of Spanishstyle green olives or the storage period of directly brined fruits. Their formation is apparently affected by diverse factors. Lactic acid bacteria can be one of them. Selection of starter cultures for the proper biocontrol of the table olive processing must include tests to avoid using those strains that might produce such compounds. Additionally, analysis of biogenic amines in processed olives can be used as complementary information to confirm possible spoilages. ACKNOWLEDGMENTS The authors express their gratitude to the Spanish government (CI- CYT, ALI and ALI CE) and to the European community (FAIR ) for the financial support. REFERENCES 1. Anderson, R. E Biogenic amines in lactic acid fermented vegetables. Lebensm. Wiss. U. Technol. 21: Durán Quintana, M. C., C. Romero Barranco, P. García García, M. Brenes Balbuena, and A. Garrido Fernández Bacterias del acido láctico en la fermentación de aceitunas de mesa. Grasas Aceites 48: Evans, P. T., and R. L. Malmberg Do polyamines have roles in plant development? Annu. Rev. Plant Physiol. Plant Mol. Biol. 4: Fleming, H. P., M. A. Daeschel, R. F. McFeeters, and M. D. Pierson Butyric acid spoilage of fermented cucumbers. J. Food Sci. 54: Garrido-Fernández, A., M. Brenes-Balbuena, P. García-García, and M. C. Durán-Quintana Conservación de aceitunas verdes o de color cambiante en salmuera. Grasas Aceites 47: Garrido-Fernández, A., M. J. Fernández Díez, and M. R. Adams Table olive: production and processing. Chapman & Hall, London. 7. Garrido-Fernández, A., P. García-García, and M. Brenes-Balbuena Olive fermentation, p In H. J. Rehm and G. Reed (ed.), Biotechnology, vol. 9. Enzymes, biomass, food and feed, G. Reed and T. W. Naagodawithana (vol. ed.). VCH Weinheim, Germany. 8. González Cancho, F., L. Rejano Navarro, and J. M. R. de la Borbolla y Alcalá La formación de ácido propiónico durante la conservación de aceitunas verdes de mesa, III: microorganismos responsables. Grasas Aceites 31: Halász, A., Barath, L. Simon-sarkidi, and W. Holzapfel Biogenic amines and their production by microorganisms in food. Trends Food Sci. Technol. 5: Hornero-Méndez, D., and A. Garrido-Fernández Biogenic amines in table olives: analysis by high-performance liquid chromatography. Analyst 119: Hornero-Méndez, D., and A. Garrido-Fernández Rapid highperformance liquid chromatography analysis of biogenic amines in fermented vegetables brines. J. Food Prot. 6: Iñigo, R. M., and F. Bravo-Abad Fermentación y nivel de histamina en aceitunas verdes. Alimentaria 141: Izquierdo-Pulido, M., J. M. Carceller-Rosa, A. Mariné-Font, and C. Vidal-Carou Tyramine formation by Pediococcus spp. during beer fermentation. J. Food Prot. 6: Izquierdo-Pulido, M., J. Font-Frabegas, J. M. Carceller-Rosa, A. Mariné-Font, and C. Vidal Carou Biogenic amine changes related to lactic acid bacteria during brewing. J. Food Prot. 59: Kuensch, U., H. Schaerer, and A. Temperli Study on the formation of biogenic amines during sauerkraut fermentation, p In P. Zeuten, J. C. Cheftel, C. Eriksson, T. R. Gormley, P. Linko, and K. Paulus (ed.), Processing and quality of foods, vol 2. Food biotechnology: avenues to healthy and nutritious products. Elsevier Applied Science, London. 16. Montaño, A., A. de Castro, L. Rejano, and A. H. Sánchez Analysis of zapatera olive by gas and high-performance liquid chromatography. J. Chromatogr. 594: Patel, I. B., and R. H. Vaughn Cellulolytic bacteria associated with sloughing of California ripe olives. Appl. Microbiol. 25: Plastourgos, S., and R. H. Vaughn Species of Propionibacterium associated with zapatera spoilage of olives. Appl. Microbiol. 5: Smith, T. A Polyamines. Annu. Rev. Plant Physiol. Plant Mol. Biol. 3: StatSoft Inc Statistica for Windows (computer program manual). StatSoft Inc., Tulsa, Okla. 21. Tattini, M., D. Heimler, M. L. Traversi, and A. Pieroni Polyamine analysis in salt stressed plants of olive (Olea europaea, L.) J. Hort. Sci. 68: Vaughn, R. H., K. E. Stevenson, B. A. Dave, and H. C. Park Fermenting yeast associated with softening and gas-pocket formation in olives. Appl. Microbiol. 23: