The aim of this study was a challenge testing the effect of lower concentrations of micronized benzoic acid against two strains of Alicyclobacillus.
The aim of this study was a challenge testing the effect of lower concentrations of micronized benzoic acid against two strains of Alicyclobacillus.
The effect of micronized benzoic acid was compared with the usual levels of untreated commercial sodium benzoate and benzoic acid, at the challenge temperature of 45°C. The size of the benzoic acid particles was determined by scanning electron microscopy. The diameter of the micronized particles was around 10 μm with a maximum length of 200 μm, while the untreated preservative structures were irregular with lengths up to 500 μm. A continuous bactericidal effect against two Alicyclobacillus strains, throughout the 28-day period, was observed with 50 mg l−1 of micronized benzoic acid, but when the untreated preservative was used, the same lethal effect was not achieved even after doubling its concentration.
The antimicrobial activity of benzoic acid was improved by micronization. The process proved to be an effective alternative to reduce the benzoic acid concentration necessary to ensure stability of an orange juice matrix.
The results proved that the micronization process represents an alternative to reduce the required food preservative concentration; this method increased the stability of the compound, which maintains its bioavailability.
High acidity in food is normally enough to prevent the proliferation of non-spore-forming bacteria such as lactic acid bacteria, moulds and yeasts; the majority being eliminated by pasteurization. However, this concept of acid product spoilage began to change with reports of spoiled pasteurized apple juice caused by an acid- and temperature-tolerant bacterium, which later was classified in the genus Alicyclobacillus (Chang and Kang 2004).
Alicyclobacillus inhabit natural environments, including soil and hot springs (Hiraishi et al. 1997). However, they also thrive in various fruits, fruit juices, sugar syrups and other food stuffs (Jensen 2000; Steyn et al. 2011). Orange juice contains such compounds and is one of the fruit juices most prone to spoilage. Thermo-acidophilic bacteria are Gram variable, spore-forming bacteria, which grow in the ranges of 20–60°C and pH 2·5–6·0. Alicyclobacillus produce heat-resistant spores, making them extremely difficult to eliminate by conventional heat pasteurization methods. The presence of ω-alicyclic fatty acids in the cell membrane contributes to its spore thermal resistance, as they reduce the permeability of the membrane (Jensen 1999).
Although Alicyclobacillus are nonpathogenic, their sporadic colonization of food and beverages is problematic, as besides being thermo-tolerant and capable of surviving pasteurization, some species are also capable of producing volatile, odorous compounds (Jensen and Whitifield 2003). Guaiacol and halophenols were identified as the offensive smelling agent in many Alicyclobacillus spp.-related spoilage. Alicyclobacillus acidoterrestris was the first identified and is considered the main off-flavour-producing Alicyclobacillus (Hiraishi et al. 1997; Witthuhn et al. 2012).
One of the principal control measures to prevent spoilage by A. acidoterrestris includes, when permitted, addition of a chemical preservative to the product (Bevilacqua and Sinigaglia 2008). The use of benzoic acid in carbonated and noncarbonated soft drinks is allowed by the Brazilian law (Brazil, 2007) at a maximum concentration of 500 mg l−1, whereas in the European Union, the maximum allowed amount is 150 mg l−1 (Walker et al. 2011).
The sodium and potassium salt forms of this preservative are preferred due to its higher water solubility. However, both forms, sodium benzoate and benzoic acid, are considered generally recognized as safe (GRAS) at the recommended concentrations (Davidson and Branen 1993). Benzoic acid is considered effective in acid foods, as a result of the ability of the undissociated molecule entering the living cell by passive transport, lowers the intracellular pH, weakens the gradient of the membrane and adversely affects key enzyme activities. Depletion of ATP was found to be the main mechanism of action of benzoic acid in Zygosaccharomyces bailii, which causes a general energy loss, that is, ATP depletion (Warth 1991).
A new category of supercritical fluid–based micronization techniques has been introduced that can be even more effective in producing particles with a good control of particle size and particle size distributions. These processes are an evolution of the traditional atomization methods, in which supercritical CO2 is used as one of the process fluids.
The micronization process represents an alternative for reducing the concentration of the preservative to be used and would also be an alternative to satisfy consumer demands for foods with lower levels of additives, besides the advantage of improving the benzoic acid water solubility.
Conventional methods of nanostructures' preparation apply toxic solvents (chloroform and dichloromethane), while only carbon dioxide is used in the supercritical fluid process (Fu et al. 2011); consequently, the RESS (rapid expansion of supercritical solutions) process is an attractive alternative to the conventional methods. One of the fundamental properties in such cases is the solid solubility in supercritical fluid. The application of supercritical fluid (SCF) solvents is based on the experimental observation that many gases exhibit enhanced solvating power when compressed to conditions above the critical point.
The solution in a high pressure vessel is expanded through an ejector, forming a jet that causes an abrupt change in density according to the amendment of the solubility of solute/solvent due to variations of pressure and temperature. This process provides particulate microscale products, with increased bioavailability due to greater ability to penetrate through the cell membrane (Giese 1994). Moreover, the micronization processes lead to a purified compound and dissolution rate, which is enhanced by the RESS process, and this might be an important factor to provide early microbial control.
Supercritical fluid extraction (SFE) is a process that is growing in importance as an alternative to conventional separation processes and is already considered practical and possible. As SFE systems are now commonly used in many industries, safety rules have been defined to assure that the pressure vessels as well as the main components of a high pressure unit are perfectly safe for continuous operation (Soares and Coelho 2012). Behaviour of materials under high pressures and compressibility of liquids and gases are known in more and more detail, which allows to design and build equipment for much high pressure and temperatures than what is probably needed in the food industry.
Nowadays, the importance of the technologies applied to obtain a high aggregated value product, like alternative preservatives, has been increasing. The engineering principles of commercial high pressure technology (supercritical fluid extraction) are well established. The challenge to produce in large scale and the commercial applications in the food industry lie in finding technical solutions that are economically feasible.
The aim of this work was a challenge testing the effect of lower concentrations of micronized benzoic acid (25 and 50 mg l−1) as compared to the usual levels of commercial sodium benzoate and benzoic acid (50 and 100 mg l−1) against two strains of Alicyclobacillus, Alicyclobacillus isolated from pasteurized orange juice and A. acidoterrestris (DSM 2498) in an orange juice matrix.
The experimental unit is shown in Fig. 1. The solute (benzoic acid, VETEC, Brazil) was placed in a 300-ml pressure vessel (Carl ROTH model, Germany). Subsequently, carbon dioxide was added in the pressure vessel until no more changes in internal pressure were detected. A thermostatic bath was used to control the temperature. The experiment was conducted at approximately 40°C, 160 bar (16 MPa) using 0·35 g mass of benzoic acid. After opening the expansion valve, the sample was taken through a capillary nozzle of 1·58 mm o.d. (outside diameter) and 0·254 mm i.d. (inside diameter) and the unit coated with a mantle for temperature control. The material was collected in a trap after the sudden change in solubility. The solid material was deposited on the walls of the trap separating the fluid.
The characterization of the micronized material was performed using optical microscopy and scanning electron microscopy.
Samples of benzoic acid (before and after micronization) were fixed in a metallic stub using double-faced adhesive carbon tape and placed in a glass desiccator containing silica gel (with vacuum) for 5 min to allow complete removal of humidity. After this, the samples were immediately covered with 40 mÅ of gold for 180 s using a sputter coater (Bal-Tec SCD 0150, Liechtenstein). The visualization was made with the Stereoscan 200 scanning electron microscope (Cambridge Instruments Ltd, England), working at a high voltage of 15 kV.
Samples of benzoic acid (before and after micronization) were placed on glass slides and observed in an inverted Axiovert microscope, with objective LD Achroplan 40×/0·60 Korr (Carl Zeiss) by high-resolution differential interference contrast (DIC). The samples were photographed with the digital camera (Motic Mc Camera, China).
Alicyclobacillus acidoterrestris CCT 4384 (equivalent to DSM 2498) and a strain of Alicyclobacillus isolated from a commercial pasteurized orange juice were maintained in orange serum agar (OSA) (HiMedia, Mumbai, India) at 4°C.
The isolated strain was identified to the genus level by morphological and physiological tests (catalase, oxidase and indole production, citrate fermentation and nitrate reduction). In addition, Fourier-transform infrared (FT-IR) spectroscopy was used to identify the isolate, comparing the FT-IR spectral obtained with the reference strain, A. acidoterrestris DSM 2498. The method used was described by Lin et al. (2005) with modifications. FT-IR transmittance spectra were collected using a Thermo Nicolet 6700 FT-IR spectrometer (USA). Infrared spectra, in the range 5000–400 cm−1, were obtained with KBr pellets with the inoculum of Alicyclobacillus sp. at room temperature.
PCR-based amplification of an approximately 1500-bp fragment of the 16S rDNA gene was performed using the universal primers AMP1 (5′ GAG AGT TTG ATY CTG GCT CAG 3′) and AMP2 (5′ AAG GAG GTG ATC CAR CCG CA) on a Mastercycler gradient thermocycler (Eppendorf). PCR mixture comprised 1 μl of DNA, 2·5 μl of 10X reaction buffer (Invitrogen), 3 mmol l−1; MgSO4 (Invitrogen), 200 μmol l−1 of each dNTP (Promega), 3·75 pmol of each primer and 0·5 units of Platinum® Taq Polymerase High Fidelity (Invitrogen™); molecular biology grade water (Sigma) was used to adjust the final reaction volume to 25 μl. Negative controls containing 1 μl of molecular biology grade water, in the place of template, were included in each assay. The cycling conditions were as follows: an initial denaturation at 95°C for 150 s, followed by 33 cycles of 95°C for 15 s, 60°C for 45 s and 68°C for 120 s; with a final extension at 68°C for 5 min.
An aliquot (5 μl) of each PCR was examined by agarose gel electrophoresis to confirm the presence of the appropriate amplicon. Remaining PCR products were treated with Exo-Sap-IT (USB) according to the manufacturer's protocol and sequenced in both directions, using the amplification primers and an addition six internal primers, to provide unambiguous sequence data, by use of the BigDye Ready Reaction mix (ABI Corp); reaction products were analysed on a Prism 3700 automated DNA analyser (ABI Corp). Sequence alignments were performed using Sequencher (version 4.10.1, Genecodes Corporation). All sequences were entered into the BLAST search algorithm and the NCBI nucleotide database to determine gene identity.
Concentrated juice (59·9 ºBrix) composed of orange juice concentrate, orange essential oil and natural aroma was reconstituted (1 : 10) with sterile water and used for the growth of Alicyclobacillus.
Stock cultures of Alicyclobacillus spp. were cultured four times, at 45°C for 24 h, in Bacillus acidocaldarius medium (BAM) with pH 3·5, according to modifications of Silva et al. (1999). Then, the cultures were grown in BAM broth for 6 days at 45°C and subsequently centrifuged at 3000 g for 10 min at 10°C. The supernatant was discarded, and pellet was resuspended in BAM broth with 15% glycerol. After shaking vigorously, the inoculum suspension was distributed in small vials, which were stored in a freezer at −18°C until the moment of use. Spore numbers were determined by plate counting in OSA after heat shock procedure (80°C for 45 min).
The matrices were supplemented with sodium benzoate (50 and 100 mg l−1), commercial benzoic acid (50 and 100 mg l−1) and micronized benzoic acid (25 and 50 mg l−1). Afterwards, they were inoculated with a suspension of spores to a final concentration of 103–104 spores ml−1 and incubated at 45°C up to 28 days. Samples were collected initially (0 day) and at 7-day intervals for Alicyclobacillus cell counting by surface plating on OSA with incubation at 45°C for 36 h.
Microstructural alterations in Alicyclobacilli cells by benzoic acid were evaluated by transmission electron microscopy (TEM). Two millilitre of a suspension containing approximately 2 × 106 spores was added to 1·0 g of benzoic acid and incubated at 45°C for 4 h. After this, 1·5 ml of the sample was collected in Eppendorf tubes and centrifuged at 10 000 g for 10 min. The pellet was fixed for 24 h with 2·5% glutaraldehyde in 0·01 mol l−1 phosphate buffer at pH 7·0, washed in phosphate buffer and finally postfixed in 20 g l−1 osmium tetroxide for 2 h. After dehydration using ethanol (10–100% with intervals of 1 min), the sample was embedded in White resin for 7 days and then individually placed in gelatin capsules with new resin and polymerized in an oven at 55°C for 18 h. The sections obtained with the stage of the microtome were collected on copper screens and contrasted with uranyl acetate for 20 min and lead citrate for 2 min. A. acidoterrestris spores treated with benzoic acid were visualized on transmission electron microscope (Carl Zeiss, model 900).
The experiments were replicated on three different occasions, and data were evaluated by the analysis of variance, with mean separation achieved using Tukey's statistical test, with reliability at a level of 0·05.
The crystals of benzoic acid obtained from the RESS process using a capillary nozzle under the conditions used for pre-expansion of around 40°C and 160 bar (16 MPa) were compared with the untreated commercial preservative as shown in Fig. 2(a,b). The diameter of the micronized benzoic acid particles was around 10 μm determined by scanning electron microscopy (SEM) (Fig. 2a). The elongated crystals and their length were measured in an optical microscope; the length of some particles reached 200 μm (Fig. 2b). The structure of untreated commercial benzoic acid was irregular, sizing up to 500 μm (Fig. 3a,b).
FT-IR transmittance spectra of A. acidoterrestris DSM 2498 (Fig. 4a) showed similar profile, but more infrared absorption peaks than the isolate (Fig. 4b), that was identified as belonging to the genera Alicyclobacillus.
Sequencing of the near full-length gene encoding 16S rRNA identified the isolated strain as A. acidoterrestris. The aligned sequencing data showed 99% homology (two nucleotide differences over 1484 nucleotides), with the sequence deposited for A. acidoterrestris DSM 2498 (GenBank accession number AB059675). The novel sequence was deposited in the NCBI nucleotide database with the accession number KC783431.
The results of the performance of the different concentrations of commercial and micronized preservatives against Alicyclobacilli are shown in Figs 5 and 6. During the first weeks, the differences in the effect of the different commercial preservative treatments were not significant (P > 0·05), but at the 4th week, 50 mg l−1 sodium benzoate treatments were clearly ineffective (Figs 5a and 6a).
However, the effect was maximized when micronized benzoic acid was used. A consistent bactericidal effect was observed with 50 mg l−1 of micronized benzoic acid along the entire incubation period, and it was significantly more effective than other treatments (P ≤ 0·05). When the nonmicronized preservative was used, the same lethal effect was not achieved even doubling its concentration (Figs 5 and 6).
The lethal effect of benzoic acid showing a complete lysis of Alicyclobacillus cell is shown in Fig. 7.
Preservatives such as benzoic acid inhibit cell metabolism by interference with enzyme systems. Dissociated weak acids are charged molecules and (in most cases) cannot freely pass the plasma membrane by passive diffusion alone. On the other hand, undissociated weak acids like benzoic acid are neutral molecules and can therefore freely pass the microbial plasma membrane through passive diffusion. The pH of orange juice was 3·8, lower than the benzoic acid pKa (pKa 4·19).
Organic acids may act by perturbing intracellular membrane dynamics under acidic conditions. In addition, preservatives often alter cell membrane permeability causing leakage of cell constituents (partial lysis, complete lysis, cytoplasmic leakage and/or coagulation of cytoplasmic constituents (protein precipitation) (Hazan et al. 2004; Remington 2006).
Crystal particles of smaller size can be obtained through changes in the capillary nozzle, in which the relationship between length and diameter (L per D) is an important factor (Domingo et al. 1996; Türk 1999) to improve the expansion time to 10−4–10−6 s (Bevilacqua and Sinigaglia 2008) and to avoid obstructing the particles' flow. Furthermore, the temperature of the capillary nozzle has to be controlled to around 403 K in order to maintain its internal heat and avoid early depletion of the solute during the expansion, which would lead to the early growth of particles (Domingo et al. 1996). The irregular distribution of particle size can be improved by avoiding temperature variations in the capillary nozzle and using collectors at low temperatures (Domingo et al. 1997).
Micronization processes using the unique properties of supercritical fluids have reached a state, where entering industrial application. To date, the majority of micronized products are pharmaceutical drugs or other material of high value. To produce 1 kg of micronized particles of substances like carotenoids, fat-soluble vitamins, it requires 10 to some 100 kg of CO2, which has an economical impact (Weidner 2009). For food applications, the RESS process might become economically feasible if SCF is applied that dissolves higher concentrations of the substance to be powdered, which is the case of benzoic acid.
According to Lin et al. (2005), the slight difference in the FT-IR spectra of A. acidoterrestris and Alicyclobacillus sp. is due to very subtle compositional differences in the bacterial cell constituents.
Absorption peak at 3420 cm−1 corresponds to the 0 → 1 vibrational transition water-bonded OH groups (Omta et al. 2003). Peaks at 2853 and 2923 cm−1 represented C–H stretches; however, these peaks could be indicative of amides associated with proteins (Braissant et al. 2009). A doublet at 2343 and 2359 cm−1 could be attributed to the O-H stretching vibrations from sulfinic or sulfonic acids (Braissant et al. 2007). The peak at 1650 cm−1 can also be attributed to the C=O stretching vibrations of amides of proteins (Xiao and Jiao 2011). Symmetric deformation of CH3 and CH2 of proteins and symmetric stretch of C–O of COO groups were also indicated by peaks around 1400 cm−1 (Lin et al. 2005). A peak at 1110 cm−1 could be attributed to S=O, stretching vibration from sulfinic or sulfonic acid (Braissant et al. 2007), and absorption peaks at 1000 and 1050 cm−1 were assigned to carbohydrate C–O stretching vibrations (Xiao and Jiao 2011).
Benzoic acid has good solubility mainly in nonpolar solvents (Beerbower et al. 1984). The micronization process increases the preservative contact surface area with a consequent rise in solubility and bioavailability of the compounds (Perrut et al. 2005). The morphology of the particles, shape, size and surface of the crystal structure are important characteristics in the activity of drugs (Pasquali et al. 2006). The bioavailability is influenced mainly by the size and distribution of particles (Martín and Cocero 2008). Under different crystallization conditions, the different crystalline structures (polymorphism) may strongly affect the properties of the substance. The surface and mechanical properties, among others, are different for different physical forms (Pasquali et al. 2006). This information corroborates to explain the consistent bactericidal action observed with the micronized benzoic acid during 28 days, even at lower concentration (50 mg l−1) when compared to the untreated preservatives.
Despite the bacteriostatic action observed along 21 days in the samples inoculated with A. acidoterrestris DSM 2498 strain treated with 25 mg l−1 micronized benzoic acid, growth was observed afterwards, reaching two cycles of log. (Figs 5b and 6b). These findings suggest the possibility of a microbial adaptation to this lower concentration of the preservative. In addition, it is well known that microbial cells are more sensitive when metabolically active (Emin 1992), so after reaching the stationary phase, the inhibitory effect had ceased.
According to Yeh et al. (2004), A. acidoterrestris cells and spores can recover despite the presence of preservatives at high concentrations and no initial growth detected, indicating that in case of the application of incorrect subsequent control procedures such as abusive temperature, bacteria will multiply and the product will consequently spoil over storage.
Alicyclobacillus acidoterrestris DSM 2498 was more resistant to the preservatives as compared to the isolated strain (Figs 5 and 6). Even after 28 days, the preservatives were more effective against the novel A. acidoterrestris (Figs 5 and 6). Bactericidal effect against the strain DSM 2498 was observed with 50 mg ml−1 of the micronized benzoic acid, but even a double concentration of the nonmicronized preservative resulted in just a bacteriostatic activity (Fig. 6).
Although benzoic acid is a more effective preservative, sodium benzoate is more commonly used as a food additive because benzoic acid does not dissolve well in water (Adams and Mos 2008). Corroborating this statement, benzoic acid was more effective than sodium benzoate against both Alicyclobacillus strains (Figs 5 and 6). It is concluded that 50 mg l−1 of micronized benzoic acid resulted in continuous bactericidal effect throughout 28 days against A. acidoterrestris, because the population counts were always lower than that the amount initially inoculated, without recovering. On the other hand, the bactericidal effect against the novel A. acidoterrestris strain was even more effective and increased significantly (P < 0·05) with time (Figs 5b and 6b).
The same or even a double concentration of the untreated preservative was less effective. In addition, the effect of 50 mg l−1 of the untreated preservative against the strain DSM 2498 decreases significantly at the end of incubation period, suggesting that the cells of A. acidoterrestris DSM 2498 started recovering (Fig. 6a). The bactericidal effect of benzoic acid against Alicyclobacillus was confirmed by the cell lysis and exposure of intracellular material observed in TEM photomicrographs (Fig. 7).
It is concluded that the applications the RESS process for micronization of food additives like benzoic acid is feasible as SCF dissolves higher concentration of the preservative and satisfies the consumer demands for safe and quality foods with lower levels of additives.
This work received financial support from CAPES (Coordination of Training of Higher Education Graduate Foundation). We are grateful to Dr Douglas McIntosh, Multi-user Molecular Biology Laboratory, Department of Animal Parasitology, Federal Rural University of Rio de Janeiro, for performing the molecular and bioinformatical procedures.