Monday, 18 March 2019

Microbial Growth on Food

Microbial Growth on Food

Microbial Growth on Food

Foods are in general dispersed systems and most of them exhibit a structure. The latter is provided by the presence of vegetal or meat tissues or by the inclusion of hydrocolloids and lipids in order to get viscous, gelled or emulsified food products. Some examples of the main structuring agents used in food products are detailed in.

The ability of microorganisms to grow on foods depends on storage conditions, food composition, presence of additives and food structure (Wilson et al., 2002). Food structure modifies water mobility and distribution of solutes such as acidulants, aw depressors and preservatives (Brocklehurst et al., 1993; Castro et al., 2003; Wimpenny et al., 1995). Furthermore, it affects the mobility of microorganisms. It is well known that the site of microbial growth is the aqueous phase and that in liquids, it occurs planktonically. The medium surrounding microorganisms is uniform; transport of nutrients to the cell occurs freely and also the metabolites produced during growth are able to diffuse into the medium (Brocklehurst 2004). On the contrary, in structured foods the mobility is restricted; microorganisms are immobilized and grow as colonies. As a result of the close spatial distribution, colonies can compete for nutrients and oxygen; besides, their metabolic end products can be accumulated near colonies affecting growth (Wilson et al., 2002). Moreover, susceptibility to stress factors is modified (Castro et al., 2009; Skandanamis et al., 2000).

As previously mentioned, most food products present some degree of structure, such as the case of emulsions, gels and solid foods. However, the effects of structure on microbial growth and on the effectiveness of stress factors have been partially evaluated specially when dealing with spoilage flora. Briefly, trends reported about the effect structure on microbial growth are diverse. Many studies postulate that structure acts as an additional stress factor and therefore lower growth is expected. For example, Meldrum et al. (2003) found that Listeria monocytogenes in a gel made with gelatin grew more slowly than in broth.

Brocklehurst (2004) reported that aw depression with sodium chloride in a gelatin gel was more effective than in broth for decreasing the growth rate of Bacillus cereus. Conversely, other studies show that structure increases growth probability. For example, Wilson et al. (2002) reported that the addition of sucrose to broth promoted a decrease in Staphylococcus aureus growth, but when sucrose was added to a gelatin gel, S. aureus growth rate remained unaffected. The main objective of this article is to review the bibliography concerning the effect of structure on microbial growth and on the activity of stress factors with special emphasis on aw depressors, pH adjustment and on antimicrobial agents. This information will help to choose the conditions that assure food microbial stability and contribute to improve the safety and quality of foods.

Gels represent one of the simplest structured systems. Gelling agents are incorporated into food formulations in order to modify their microstructure and texture (Antwi et al., 2006). However, the effects may be wider. It must be noted that in gelled products, microorganisms are immobilized and grow as micro-colonies (Robins and Wison 1994; Wilson et al., 2002). Mentioned immobilization affects bacterial growth rate, as well as microbial response to environmental conditions, as a consequence of the reduced metabolic activity found in some regions of the colony (Koutsoumanis et al., 2004; Robins and Wilson 1994). Compared to cells in suspensions, bacteria immobilized in solid media have to overcome an additional stress to be able to initiate their growth. The lower microbial growth observed in solid than in liquid media may be related to nutrient diffusion, oxygen availability, the rate and profile of end-products production and cell to cell communication (Koutsoumanis et al., 2004; Skandamis et al., 2000).

Gelled foods are obtained by the addition of hydrocolloids which are polysaccharides or proteins (Antwi et al., 2006; Wilson et al., 2002). In addition to their uses as texturizing agents in the food industry, they are used to model solid food products for research purposes. The gelling agents most frequently selected are agar and gelatin. Agar is a polysaccharide obtained from seaweed and gelatin is a fibrous protein derived from collagen. Gelatin has a low melting point (25-37°C), as a consequence, it is possible to inoculate a microorganism at low temperature, without its inactivation. In addition, it is easy to remelt the gel for subsequent analysis. On the other hand, agar solidifies at 45°C and melts at 85°C. The high temperature at the inoculation step may be stressful for microorganisms, but the obtained gel is less influenced by temperature changes. As an advantage, agar is nontoxic and physiologically inert toward microorganisms (Mertens et al., 2009; Theys et al., 2010). Both gelling agents are generally recognized as safe (GRAS) food ingredients (FDA, 2014). To evaluate microbial growth in model gelled systems, the inoculated gelling agent (mainly agar or gelatin) is placed in Petri dishes, microplates or in a Gel Cassette. The latter consists of a frame holding a layer of gel. Mentioned frame is sealed with a plastic film, which is gaspermeable (Wilson et al., 2002). This system can be used to study microorganism‘s growth on the gel surface as well as within the gel matrix by measuring microbial growth by plate count (Brocklehurst et al., 1997; Mertens et al., 2009). Plate count is a time consuming technique. To solve up this problem, Mertens et al. (2011) develop the well scan method, which involves the determination of the optical density of the inoculated gel contained in microplate wells at nine different positions in each well. The average of the nine values is calculated and it is used to construct growth curves.

The role of structure on the inhibition of microorganisms‘ growth was widely demonstrated. A compilation of the studies on this subject is shown in Antwi et al. (2006) observed that L. innocua growth rate decreased as gelatin concentration increased and that micro-colonies morphology changed as gelatin concentration was varied. However, gelling could not be used as the principal stress factor in food. The combination with other preservation factors is necessary to guaranty food safety. In reference to aw, it was reported that its depression decreased the growth rate or the maximum population reached by bacteria and that the magnitude of the effect was dependent on the type of solute used. Although the effect of aw was greater in liquid media, the presence of solutes produced significant changes on bacterial development in gelled systems (Brocklehurst et al., 1997; Meldrum et al., 2003; Theys et al., 2008). As an example, Meldrum et al. (2003) reported that sucrose produced a decrease in L. monocytogenes Scott A growth rate in gelcassettes. In addition, Brocklehurst et al. (1997) observed that the maximum population reached by Salmonella Typhimurium during growth decreased as the concentration of NaCl or sucrose increased. Furthermore, it was shown that lowering aw from 0.990 to 0.970 produced the elongation of lag phase and the decrease of growth rate and the maximum population reached by S. Typhimurium at different pH values or gelatin concentrations (Theys et al., 2008). Concerning the effect of pH, Meldrum et al. (2003) observed that the minimum pH at which L. monocytogenes Scott A was able to initiate its growth was higher for an immobilized culture than for a planktonic one. Similar results were obtained by Brocklehurst et al. (1997). Moreover, this trend is enhanced at low aw values (Theys et al., 2008). This effect is related to the fact that the presence of gelling agents increases the buffering capacity of the medium, which offers protection to microorganisms (Antwi et al., 2006).

As regards the combined effect of mentioned stress factors, Koutsoumanis et al. (2004) studied the effect of structure, pH and aw on bacterial growth and observed that the minimum values that allowed the growth in agar were higher than in broth, being even higher when temperature was decreased. Moreover, it has been observed that refrigerated incubation, at low pH, low aw and immobilization may prevent L. monocytogenes Scott A growth and even cause the loss of cell viability (Meldrum et al., 2003).

Antimicrobial effectiveness in solid media depends on the spatial relationship between colonies since microorganisms‘ growth is developed as discrete colonies. In general, antimicrobial action is lower in structured than in liquid media. This may be related with the proximity of the antimicrobial agent to bacteria or antimicrobial agent diffusion both limited by the solid medium (Gliemmo et al., 2015; Hammer et al., 1999; Skandamis et al., 2000). In the case of essential oils, different trends were reported. In the presence of oregano essential oil, the growth rate and the maximum population level of S. Typhimurium in gelatin gel, supplemented with yeast extract, were higher than in Tryptone soya broth (Skandamis et al., 2000). A level of 6% cilantro oil immersed in gelatin gel covering a ham, allowed the reduction of 1.3 log CFU/cm2 on L. monocytogenes growth. This level is higher than 0.074% that is the minimal inhibitory concentrations (MIC) value of cilantro oil in broth (Gill et al., 2002). Hammer et al. (1999) obtained lower values of MIC of diverse essential oils against various Gram-positive and Gram-negative bacteria in broth than in agar. However, they observed the opposite trend for patchouli and sandalwood against C. albicans growth. These discrepancies could be due to the different sensitivity of microorganisms to the action of the essential oil, differences in solubility of essentials oils and the nature of the emulsifier used for oils dispersion. Other hypothesis would be that the oil droplet size in gel may limit oil diffusion toward microorganism colonies resulting in the reduction of antimicrobial action or that the growth as a colony protects cells located within the colonies from antimicrobial action (Burt 2004; Skandamis et al., 2000).


Oil in Water Emulsions
Many foods are oil in water emulsions from the physical point of view. The oil concentration varies between 3-5% in the case of milk, from 10 to 40% for salad dressings and can be as high as 85% for mayonnaise. The oil phase consists of polydispersed droplets with a diameter of 0.15-10 μm (Brocklehurst and Wilson 2000). When droplets concentration is high, the space between droplets can be of the same order as the diameter of the droplets. This trend limits the space available for microorganism to grow (Brocklehurst et al., 1995).

The study of microbial growth in emulsions is more difficult than in gels due to the fact that emulsions are opaque and microscopy and absorbance based methods cannot be applied. To solve up this problem, Parker et al. (1995) developed a method that removes the oil phase of the emulsion with a mixture of methanol and chloroform prior to scanning electron microscopy, light microscopy or transmission electronic microscopy evaluation. Using these techniques, they found that L. monocytogenes and Y. entercolitica grew in the form of colonies in emulsions containing hexadecane (30-83%) or double cream (33% fat) with or without agarose. In the case of dairy cream, it was found that bacteria were associated with particulate material, probably with casein and that little fat content was trapped among bacteria. Many regions of the sample were sterile and bacteria were concentrated in small areas. This fact has to be taken into account when evaluating microbial stability of these

Brocklehurst el al. (1995) evaluated the effect of hexadecane and sunflower oil concentrations and droplet diameter on the form of L. monocytogenes and Y. entercolitica growth using the previously described method. They found that when mean droplet was around 2 μm and concentration of oil was high -in order to have close packed droplets-, bacteria grew as colonies and growth rates were lower than in liquid media. The increase in oil droplets to 15-25 μm removed the inhibitory effect on growth rate but the population at the stationary phase remained lower than the one found in liquid media.

Castro et al. (2003, 2009) studied the effect of oil concentration (11-23 and 46%) on rheological properties and on Z. bailii and Lactobacillus fructivorans survival in model salad dressings and they reported that the emulsions containing 11 or 23% oil behaved as viscoelastic systems and the ones with 46% oil behaved as a gel. Z. bailii was able to grow in emulsions containing 11 or 23% oil but it was inactivated in the presence of 46% oil, suggesting that the solid character of the media could affect growth and decrease nutrient availability. However, in the emulsions inoculated with L. fructivorans, the population remained constant for all oil levels suggesting that the change in structure produced by the use of 46% oil did modify L. fructivorans growth or survival.

Prachaiyo and McLandsborough (2003) studied the growth of E. coli in emulsions with 5 to 40% of hexadecane. They found that generation time of bacteria increased with the level of hexadecane, also the population reached at the stationary phase decreased as the level of hexadecane increased. They ascribed these trends to the space limitations between the droplets since their size (1-2 μm) was similar to the size of bacterial cells. Naïtali et al. (2009) evaluated the growth of two strains of L. monocytogenes in emulsions containing 80% oil simulating a mayonnaise and in a liquid media at acidic and neutral pH. They found no effect of the emulsion on the growth rate and maximum population at the stationary phase of both Listeria strains independently of pH. The authors attributed
this lack of effect of oil to the fact that the studied emulsions presented a fluid structure despite their high content of oil.

It must be stressed that independently of its role as a structuring agent, the presence of oil affects the physiology of microorganisms. For example, Prachaiyo and McLandsborough (2003) reported that E. coli cells grown in emulsions with 40% hexadecane produced extracellular filaments as it could be seen by scanning electron microscopy images. This type of formation is usually induced when cells were exposed to stress factors such as low temperature or nutrient starvation (Prigent-Combarent et al., 2001). Moreover, thermal resistance studies showed that bacteria grown in 40% hexadecane emulsions exhibited high tolerance to heat since D value at 55 ºC increased from 9.9 ± 1.3 to 18.6 ± 0.9 when cells submitted to the thermal treatment had been grown in an emulsion. Naïtali et al. (2009).reported that L. monocytogenes grown in a neutral emulsion exhibited an increase in hydrophobicity. As a consequence a weak affinity for dichlororoisocyanuric acid -hydrophilic compound- and a higher affinity for didecyldimethylammonium bromide -cationic amphiphile compound- were verified. These tendencies have to be considered when selecting a disinfectant.

Stress factors applied for the preservation of oil in water emulsions such as salad dressings, sauces and mayonnaises are mainly aw depression by the addition of NaCl or glucose, depression of pH by the addition of acidulants (organic acids such as acetic, citric and lactic) and preservatives such as sorbic and benzoic acids. It is known that these compounds interact with the lipid phase and, consequently, their distribution is modified. Undissociated organic acids tend to partition between the aqueous and the lipid phase, hence the amount of the organic acid in the aqueous phase decreases and the magnitude of this depends on partition coefficient value. In the case of food preservatives such as sorbic and benzoic acids, partition coefficients are in the range of 3-6 and therefore most of the preservative is present in the lipid phase, fact that can compromised microbial stability (Brocklehurst and Wilson 2000). Moreover, emulsifiers are added to stabilize the emulsion; they are located at the interfacial phase and can also interact with the preservative (McClements and Decker 2000). In summary, the preservative added to an emulsion will not be entirely available for antimicrobial activity, some of the preservative will partition into the oil phase whereas some will interact with the interfacial phase. Thus, the knowledge of the distribution of the preservative is necessary to determine its required amount in order to protect food emulsions from microbial contamination since only the preservative which exists in its free form in the aqueous phase will be active (Wedzicha et al., 1991).

The most common emulsifiers used in the food industry are amphiphilic proteins, phospholipids, and small surfactant molecules. The nature of the interfacial membrane formed by these emulsifiers can have a large impact on the antimicrobial activity of preservatives. The characteristics of the interface depend on the type and concentration of the surface-active molecules present. Generally, this interface consists of a narrow region -few nanometers thick- surrounding each oil droplet, though the spatial fraction it occupies within an emulsion depends on the droplet size. A significant portion of preservative molecules may therefore partition into this interfacial region, particularly if they are present at relatively low concentrations. Moreover, when surfactants are used to stabilize emulsions, their amount in the aqueous phase far exceeds its critical micellar concentration, thus only a fraction of them actually surround the droplets, with the rest remaining in the aqueous phase, usually as surfactant micelles (Kurup et al., 1991a; Wedzicha and Ahmed 1994). These micelles entrap some of the preservative, resulting in less free preservative available to act against microorganisms. The presence of surfactants also reduces the interfacial tension between the oil and aqueous phases. Controversially, Kurup et al. (1991b) suggested that both the preservative and the microorganisms would be attracted to the interface where the accumulated preservative molecules could destroy microorganisms at a faster rate.

The effect of the interaction between potassium sorbate, oil and the surfactant Tween® 20 on the growth and inhibition of Z. bailii in salad dressings was studied by Castro et al. (2003). The presence of potassium sorbate (500 ppm) inhibited Z. bailii growth in emulsions containing 11 and 23% (v/v) of oil. When the oil content was raised to 46% (v/v), a sharp death rate curve was observed and potassium sorbate antimicrobial activity was overlapped by the inhibitory action of the high oil level. Tween® 20 effect depended on the amount of oil; in emulsions containing 11 or 23% (v/v) a depression in the activity of potassium sorbate was observed. This fact was attributed to the decrease in the free form of the preservative due to its partition between surfactant micelles and water. However, when the oil content was 46% (v/v), potassium sorbate activity was enhanced in spite of the decrease in the amount of its free form. These results highlight the significant impact of the distribution of molecules within an emulsion on ingredient interactions, and hence, on its preservation.

Moreover, a decrease in the antimicrobial activity of the bacteriocin nisin was verified in milk with varying fat content (Bhatti et al., 2004; Jung et al., 1992; Zapico et al., 1999). Jung et al. (1992) and Bhatti et al. (2004) also found that the non-ionic emulsifier, Tween® 80, partially counteracted the decreases of nisin activity in milk caused by the high fat content. Nevertheless, Henning et al. (1986) confirmed an antagonistic effect of emulsifiers on the antimicrobial efficacy of nisin. The interaction of nisin with surfactants in a food matrix is relatively ambiguous and it still needs to be elucidated.

A research conducted by Castro et al. (2009) made this ambiguity even more obvious. They formulated emulsions with different oil contents with potassium sorbate and nisin as preservatives. When both antimicrobial agents were added together to an emulsion containing 110 g/kg of oil, the presence of potassium sorbate exerted an antagonistic action on nisin effectiveness, while in emulsions with higher levels of oil, a synergistic action on the bacteriocin activity against Lactobacillus fructivorans was verified. However, when the bacteriocin was exclusively present, it produced different effects which depended on system composition. Addition of Tween® 20 did not affect bacterial survival for emulsions free of additives or containing only potassium sorbate as preservative. Conversely, when nisin was present, the emulsifier effect was entirely dependent on oil content. It is noteworthy what these authors postulated regarding emulsion formulation. The addition of the surfactant turned the emulsions more fluid or solid, depending on the oil content of the systems. Consequently, they found that the structure of the food matrix appeared as an additional factor which could influence either the growth of the microorganisms or the functionality of the preservatives. As previously commented, food emulsions comprise an abundant sort of products; ranging from liquids to solids, many are the examples to be mentioned: milk, fruit beverages, soup, cream, mayonnaise, salad dressings, ice-cream, butter, margarine, Frankfurters (Vienatype sausages), structured ham, etc. Antimicrobial agents used to preserve them showed to be strongly dependent on system composition. Outcomes shown herein highlight the importance of considering the effect of the structure, ingredient interactions, and composition, when evaluating microbial stability of these food systems.

Water in Oil Emulsions
Several foods such as butter, margarine and fat spreads are water in oil emulsions in which droplets of aqueous phase (0.30-30 μm) are dispersed in the oil phase (Brocklehurst and Wilson 2000). Microbial growth takes place in water droplets therefore space and nutrient availability restrict growth (Verrips and Zaalberg 1980). A smaller droplet size and its separate distribution enhanced microbial stability (Brocklehurst and Wilson 2000). Based on commented trends, mechanistic models were develop and applied successfully to predict the potential for bacterial growth based on droplet size and bacterial energy demands (ter Steeg et al., 1995; Verrips et al., 1980). Mentioned models show that bacteria growth is under control if emulsion structure is stable but if droplet coalescence takes place bacteria can grow. Regarding moulds, whose growth is not only restricted to the droplets, ter Steeg et al. (2001) proposed another model to evaluate the effect of green antifungal, droplet size distribution and temperature on mould outgrowth in fat spread taking into account movement of moulds to the lipid phase.

It must be stressed that microbial growth is not the main deteriorative reaction in water in oil emulsions but, if microbial growth takes place, high number of microorganisms can be found within the droplets and probably they died when concentration of metabolic end products become toxic or nutrients were not available (Brocklehurst and Wilson 2000).

In meat and vegetable tissues growth occurs at the surface. Different model systems manufactured with gelling agents were used to study this issue. As an example, a solid surface made with agar was used to evaluate the effect of gas atmosphere composition on the growth of food-borne pathogens (Bennik et al., 1995). In surfaces, growth occurs in colonies and constraints on growth were the same as in gels. But, diffusion limitations and accumulation of protons under the colony are greater than in gels; as a consequence, microorganism growth rate becomes lower in surface colonies than in immersed colonies   (Wilson et al., 2002). Furthermore, many foods contain micro-architectures and growth of microorganisms can take place plancktonically, in colonies -immersed or at the surface depending on the localization of the microorganisms. Water is located within the micro structure and its distribution is not uniform, in this way different aw values can be found providing a heterogeneous environment for microorganism (Møller et al., 2013). As aw gives information of the global available water, its determination must be complemented with the information of the different populations of water within the structure. The latter can be evaluated using proton nuclear magnetic resonance (NMR). These measurements were done and used to evaluate their relationship with food structure in cheese (Møller et al., 2013). In the following sections, the effect of structure on microbial growth in vegetables, dairy products, meat and meat products are discussed.

Increased consumption of minimally processed fruits and vegetables has lead to an increase in the number of outbreaks related to these products. The outbreaks were specifically associated with leafy green commodities and with Salmonella (Patel and Sharma 2010), B. cereus (Elhariry 2011), E. coli O157:H7 (Kroupitski et al., 2011) and L. monocytogenes (Ells and Hansen 2006). Since these products are often consumed raw or minimally processed, it is essential to understand the initial stages of pathogen attachment to vegetables in order to apply strategies to avoid it. Attachment ability depends on the pathogen, the surface morphology of the vegetables, the temperature and the integrity of the tissue (Ells and Hansen 2006). It is a complex mechanism that is linked to physicochemical properties of both bacterium and plant surfaces (Hirano and Upper 2000; Ukuku and Fett 2002). A correlation between the negative surface charge and hydrophobicity of several bacterial pathogens and the strength of their attachment to cantaloupe rind surfaces was previously stated by Ukuku and Fett (2002). For example, when materials possess pits and cavities on surfaces such as asparagus, spores can penetrate in these areas (Park and Beuchat 1999). Moreover, bacteria with greater hydrophobic membrane may attach to the cuticle or plant surfaces (Patel and Sharma 2010). Main results of some studies about the topic will be commented in the next paragraphs.

Patel and Sharma (2010) evaluated the ability of five Salmonella serovars to attach to and colonize intact and cut lettuce (iceberg, romaine) and cabbage surfaces. They found that all Salmonella serovars attached rapidly on intact and cut produce surfaces. But, Salmonella spp. attached to romaine lettuce at significantly higher numbers than those attached to iceberg lettuce or cabbage. Attachment strength was significantly lower on cabbage followed by Iceberg and Romaine lettuce. Cabbage, intact or cut, did not support attachment of Salmonella as well as romaine lettuce.

Elhariry (2011) investigated the ability of six B. cereus strains to attach and form biofilm on cabbage and lettuce surfaces. The highest biofilm formation on cabbage and lettuce surfaces was obtained by spores and vegetative cells of all tested strains on the 4th hour of the incubation period. This trend highlights the importance of hygienic preparation and handling to avoid attachment and assure safety of green-leafy vegetables. The strength attachment of both spores and vegetative cells of the strains to the lettuce surface was higher than that of the cabbage surface confirming the dependence of surface morphology on attachment. Kroupitski et al. (2011) analyzed the distribution of green-fluorescent protein-labeled S. Typhimurium on artificially contaminated romaine lettuce leaves in order to understand initial pathogen–leaf interactions. They reported that bacteria attachment to different leaf regions was highly variable and a higher attachment level was observed on leaf regions localized close to the petiole compared to surfaces at the far-end region of the leaf blade. Finally, attachment to surfaces located at a central leaf region demonstrated intermediate attachment level. Moreover, Salmonella were also visualized underneath stomata within the parenchymal tissue, suggesting that the bacteria can also internalize romaine lettuce leaves. Comparison of attachment to leaves of different ages showed that Salmonella displayed higher affinity to older compared to younger leaves. Scanning electron microscopy revealed a more complex topography on the surface of older leaves, as well as on the abaxial side of the examined leaf tissue supporting the notion that a higher attachment level might be correlated with a more composite leaf landscape.

Ells and Hansen (2006) evaluated the ability of different Listeria strains to attach and colonize intact or cut cabbage tissue which were exposed to different temperatures. Results showed that all strains exhibited more attachment to cut tissues compared to intact leaf surfaces. Furthermore, scanning electron microscopy revealed the presence of increased cell numbers on the cut edges with numerous cells located within folds and crevices. Cells found on the intact surfaces were randomly distributed with no apparent affinity for specialized surface structures. The culture growth temperature significantly affected the strength of attachment during the first 4 h of exposure to intact surfaces, being cells cultivated at 37°C more easily removed from leaf surfaces than those cultivated at 10 or 22°C. However, after 24 h, binding was not significantly different between temperatures and increasing exposure time to the cabbage resulted in increased attached cell numbers as well as increased binding strength.

Dairy Products
Microorganisms are present in dairy products as natural microflora or as starter cultures. In fermented products flocculation of casein induced the formation of a gel and, according to the product, different microstructures can be found. Location of microorganisms during the processing of cheese and yoghurt is of interest due to the role of the microorganisms and their enzymes on ripening (Belitz and Grosch 2009). Non destructive microscopic methods such as confocal laser scanning and scanning electron microscopy are very useful to visualize the location of microorganisms (Hickey et al., 2015).

Different studies evaluated the distribution of microorganism in cheese. Hickey et al. (2015) reported that microorganisms are entrapped in the protein matrix being necessary the diffusion of nutrients to bacterial colony an also the diffusion of metabolic product to assure bacterial growth. Microscopy shows that bacteria are preferentially located at the fat-protein interface and sometimes within whey pockets. Parker et al. (1998), using light and electronic microscopy, found four populations of microorganisms in mature Serra Cheese depending on the position within the cheese and these trends were related to changes in flavor and texture. Modification in dairy products formulations -such as the substitution of lipids by proteins- could modify the potential of growth of bacteria. It is known that microbial development during ripening or storage can be limited to aqueous micro-zone within a lipid protein matrix. However, in light dairy products the aqueous phase is the continuous phase and microbial growth can take place. Guerzoni et al. (1994) modeled the growth of L. monocytogenes and Y. enterocolitica in food model systems with different levels of NaCl and lipids and in dairy products and compared the results obtained. They observed that individual or interactive effects of pH, salt and lipid content were not enough to predict bacteria growth and that microstructure played an important role.

Meat and Meat Products
Meat tissue consists of long, thin, parallel cells arranged into fiber bundles which are surrounded by elastin fibers. Its surface was defined by Wilson et al. (2002) as being the simplest form of micro-architecture affecting the growth of microorganisms. The aqueous phase is structured among the network of meat fibers and the crevices of the surrounded tissue. Hence, microorganisms are immobilized and constrained to grow as colonies. This fact limits nutrients diffusion and produces depletion of oxygen and accumulation of protons beneath the colony; this results in local changes in the concentration of substrates and acidic metabolic end-products, which leads to decreasing growth rates. Despite this overall unfavorable scenario, pathogens and spoilage microorganisms manage to colonize, survive and proliferate in meat surfaces. Bacteria that are trapped and attached firmly to these structures are hard to remove by decontamination or rinsing (Noriega et al., 2010). In general,
surface composition, roughness, charge and hydrophobicity of both surface and cells influence the adhesion of bacteria to a surface (Treese et al., 2007). Bacterial cells attach to and colonize surfaces which are more elastic, porous and rougher surfaces as compared to dense and smooth surfaces (Katsikogianni and Missirlis 2004; Wan Norhana et al., 2009). It should also be considered that microorganisms will tend to form microcolonies in a solid or particulate food, e.g., minced meat, or grow in the form of slime-covered biofilms on meat surfaces. Kumar and Anand (1998) made an exhaustive revision of the protective effect of biofilms against antimicrobial substances in the food industry.

Weak organic acids -mostly acetic and lactic acids- are frequently used as a cheap and effective means to reduce number and prevalence of bacterial pathogens on meat. Application of organic acids might be done as spray wash or dipping solution. The application of 2% lactic acid spray solution on beef carcasses and chicken breasts reduced E. coli O157:H7 population for more than 1.5 log CFU/cm2 (Anang et al., 2007; Bosilevac et al., 2006; Kalchayanand et al., 2008). When used as carcasses or meat cuts decontamination rinses, possible color discoloration, flavour and/or odour due to the treatment should be evaluated since these could affect sensory quality of meat (Pipek et al., 2005; Theron and Lues 2007). Blagojevic et al. (2015) applied hot lactic acid (HLA) for the decontamination of incoming beef in the manufacturing of dry sausages. The process significantly reduced E. coli O157:H7 and Salmonella counts (but not L. monocytogenes); however, sensorial quality of finished sausages produced with HLA-decontaminated beef was somewhat reduced. Apparently, the effectiveness of the treatments depends on the extent of cell attachment; cells can be both actively attached to -and also potentially trapped in- the irregular structure and crevices on skin or the network of collagen and reticular fibers in meat surfaces (Gill et al., 1984; Thomas and McMeekin 1980). Thomas et al. (1987) suggested that the cells can be drawn in with water into the fiber tissues and migrate to a depth of about 25 mm. Furthermore, the buffering capacity of meat, i.e., its ability to modulate acidic or alkaline pH towards neutral and to withstand rapid pH fluctuations, has a great influence on these treatments (Goli et al., 2007). Tan et al. (2014) suggested that the difference in fat composition and other components, such as proteins, may increase the vulnerability of attached cells to acetic acid.

The use of bacteriocinogenic lactic acid bacteria (LAB) for the biopreservation of meat surfaces has been one of the novel techniques developed in the last decades. The efficacy of bacteriocins in meat and meat products critically depends on interactions with meat constituents (mainly fat, proteins and enzymes), and the distribution of bacteriocin molecules in the food matrix. For instance, application of nisin in meat products faces several limitations resulting from its interaction with phospholipid emulsifiers and other food components (Aasen et al., 2003; Henning et al., 1986; Jung et al., 1992), poor solubility at pH above 6.0, and inactivation by formation of a nisin-glutathione adduct (Ross et al., 2003). Nevertheless, inactivation is lower in cooked meats due to the loss of free sulphydryl groups during cooking as a result of the reaction of glutathione with proteins (Stergiou et al., 2006). Antilisterial activity of sakacin Q produced by Lb. curvatus ACU-1 on the surface of cooked pork meat was assessed by Rivas et al. (2014). The use of freeze-dried reconstituted cell free supernatant (CFS) was effective to control Listeria growth reducing its population to undetectable levels after two weeks of refrigerated storage. Despite this effectiveness, the CFS suspension showed a lower antimicrobial potency. The authors attributed this phenomenon to the adsorption of the bacteriocin to fat and meat tissues since hydrophobicity of bacteriocins from LAB could lead to an unspecified union of bacteriocin molecules to the hydrophobic surface of fat particles (Holzapfel et al., 1995). Furthermore, Kouakou et al. (2009) found that highfat content meat antagonizes the antilisterial effect of bacteriocinogenic Lb. curvatus CWBI-B28wt in pork meat systems. The latter trends could be explained by the study of Aasen et al. (2003) which showed that the activity of the bacteriocins in a medium depends not only on the fat content and type of bacteriocin, but also on the physical state of the medium; i.e., liquid or solid.

Meat in its natural condition might differ from disintegrated states. Considering just meat composition would give us only limited information regarding the physical state, structure or engineering properties of meats. Ground meats have a chemical composition similar to their sources but quite different physical properties, rheological behavior and sensorial attributes (Aguilera 2005). Besides, fermentation, emulsification, or cooking -among other processing operations- will condition meat structure and its general behavior.

Comminuted Meat Batters - the So-Called “Meat Emulsions”
composed of water, protein, fat and salt; being the most representative products, Frankfurter sausages and bologna. The aqueous phase of the meat emulsion is the one that holds up microbial growth, as such is the case of other food emulsions. The availability of water determines how and where microbial proliferation will proceed in the meat batter. As a good example, the work of Terjung et al. (2015) can be cited. These authors found that the antimicrobial effectiveness of a spice ferment against L. innocua LTH 3096 and Lb. curvatus LTH 683 was higher on restructured ham compared to emulsion type sausages. They attributed this effect to the lower fat content - hence, higher aw value – of restructured ham. Microbial contamination of these products occurs during post-processing steps, i.e., handling, packaging, storage, etc.; hence, the colonization of the food matrix mainly takes place on the surface. The type of preservation method to be used is focused mainly on the protection of the outer layer of the products. Consequently, antimicrobial agents are added to rinsing or spray solutions, protective coatings, or packaging. The effectiveness of these preservatives is influenced by the solid food structure (Wilson 2000), which provokes a chemical redistribution of food preservatives (Brocklehurst et al.  , 1993; Brocklehurst and Wilson 2000). In addition, concentration gradients of nutrients and metabolites, and also O2 and pH, within and around the colonies affect microbial growth, as a consequence of diffusional limitations in solid systems. Baka et al. (2015) found differences on the lag estimates of L. monocytogenes at 4°C on Frankfurters with different fat concentrations (‗type 1‘: 19%; ‗type 2‘: 14.5%). They assumed that the undissociated organic acids (preservatives) -being lipophilic compounds- migrated to the lipid phase of the product; which was driven by the presence of solutes, such as NaCl, in the water phase (Gooding et al., 1955; Sofos and Busta 1981). Therefore, in the ‗type 1‘ sausages, with higher fat concentration, the action of undissociated organic acids might have been diminished. This would explain the more extended lag phase found on the ‗type 2‘ sausages.

High-pressure processing (HPP) is an attractive preservation technology which has a good potential for the meat industry in particular. The resistance of the microorganisms is variable depending on the strain and the meat matrix to be treated (Garriga et al., 2004). As a consequence, antimicrobial agents are used together with HPP in ready-to-eat (RTE) meat products so as to contribute with their preservation. In a research conducted by Pal et al. (2008) on refrigerated RTE frankfurters, potassium lactate and sodium diacetate were added to the meat batter and then final products were vacuum-packaged and subjected to HPP (400 MPa, 15 minutes). Although both antimicrobial agents had a bactericidal effect on background microbiota, a bacteriostatic effect had merely been displayed against L. monocytogenes. It is presumed that the application of HPP could have structurally degraded some muscle cell components and released additional nutrients for L. monocytogenes growth, contributing to its survival along shelf-life of the product.

The results showed herein comprise a condensed group of examples chosen to illustrate the wide-ranging influence of meat structure on several preservation methods. This evidences that: i) microbial growth data from liquid media could only characterize the isolated effect of the preservatives‘ levels, but not the stress from the ‗solid environment‘; ii) chemical nature of preservative agents -amphiphilic molecules (microbial peptides such as bacteriocins); lipophilic organic acids (e.g., sorbic and benzoic acids)- condition the distribution of these substances on the different phases of the meat matrix (aqueous-proteinaceous phase or lipoid phase); iii) processing mechanisms associated with changes (disrupters) of the cells within meat tissues could result in nutritional changes favorable for microbial growth. On the whole, meat and meat products have to be considered as complex architectures regarding food matrices, and therefore should be addressed as such when choosing the best way to preserve them.

Predictive microbiology pretends to reduce time-consuming and costs involved in challenge tests through the model of microbial growth as a function of time (primary models) and as a function of a few environmental factors (secondary models). The last factors include the traditional ones such as temperature, pH, aw, and others like antimicrobials, organic acids and oxygen. However, sometimes microbial growth cannot be predicted by these models since some factors such as background flora, microbial competition, stress factors, medium structure and environmental changes produced by microbial growth are not taken into account. This omission is the main source of error in predictive microbiology and it is called as the completeness error (McMeeckin and Ross 2002).

Concerning about this, Boons et al. (2014, 2015) studied the effect of increase complexity of the structured medium on E. coli and S. cerevisiae growth. They included heterogeneous systems and NaCl as a stress factor mimicking the inhomogeneous composition and structure of foods. Microbial dynamics was affected by medium structure complexity since the microorganisms showed higher growth in complex than in liquid medium. However, the behavior of both microorganisms was different in the same structured medium. A secondary model including the effect of medium structure on S. Typhimurium growth rate, previously developed by Theys et al. (2008), was successfully validated in pasteurized milk and cheese (Theys et al., 2009a). Also, this model described L. innocua and Lb. lactis growth as a function of gelatin level.

Different models such as Fickian diffusion model, to predict the diffusion of nutrients and metabolites, and Buffering Theory, to predict local pH changes, have being developed to be incorporated into an integrated modeling methodology to predict growth in structured systems (Van Impe et al., 2005; Wilson et al., 2002). Regarding to this, Van Impe et al. (2013) explain that traditional predictive models consider the behavior of average population and fails in the description of colony dynamics since the local competition for nutrients causes a different behavior of the individual colony cells and does not have a normal distribution. They propose considering the effects of environmental conditions on cell metabolism and growth dynamics by using Metabolic Flux Analysis. They suggest that this information will allow improving precision of predictive models for more complex systems like structured media. More recently, Tack et al. (2015) applied an individual based model on E. coli growth in gel from cellular parameters reported in bibliography. The model included the local nutrient competition, individual cell differences and intercolony interaction. It successfully reproduced single colony dynamics, simulated interaction between colonies and demonstrated that nutrient diffusion and local cellular glucose competition produced emergence of a starvation zone in the center of the colony. From the knowledge of cellular parameters of microorganisms in structured media, this model contributes to microbiological quality and safety of structured foods.

On the other hand, there are several methods that combine microscopy and image analysis that pretend to be used in predictive food microbiology. In this regard, several authors developed microscopy techniques for monitoring individual colony dynamics. Skandamis et al. (2007) found the correlation between growth rate values of E. coli O157:H7 in gel cassette obtained by plate count and those obtained by light microscopy and image analysis. Also, they could fit a secondary model with growth rates obtained via image analysis. Theys et al. (2009b) compared colony volume dynamics and cell number of S. Typhimurium colonies in gel cassette and they found that there were dead and growing cells in stationary phase. However, these methodologies have some disadvantages since they
require the design of the matrix to control experimental conditions (gel cassette in these cases) and a complex data processing procedure. This led to the search of faster and more versatile methodologies to study the structure effect on the colony growth. In this regard, Mertens et al. (2011) studied the effect of pH, aw, acetic acid concentration, temperature and medium structure (Carbopol or xanthan gum as thickening or gelling agents) on the growth/no growth boundaries of Z. bailii in model systems of acidic sauces through optical density measurements. They could observe differences in growth probability of yeast between liquid and structured media. In connection with the use of optical density measurements, a novel method was developed by Mertens et al. (2012) for monitoring the individual dynamics of colonies growth in solid medium as a function of time by scanning the area of growth of an individual colony through optical density measurements in microtiter plates. Particularly, they studied E. coli growth in Brain Heart Infusion broth plus 5% agar but the methodology could be extended to other microorganisms. This methodology is promising because it allows to minimize the difficulties related to inoculation and sampling and to obtain rapid data collection. However, the transparency of the medium is required, which is a disadvantage when applying to other food matrices. Therefore, more research is needed to predict the effect of the structure on microbial growth in solid media by using methodologies as fast and versatile as this.

No comments:

Post a comment