Antimicrobial Activity of Fermented Vegetable Byproduct Extracts for Food Applications

To prevent foodborne diseases and extend shelf-life, antimicrobial agents may be used in food to inhibit the growth of undesired microorganisms. In addition to the prevention of foodborne diseases, another huge concern of our time is the recovery of agri-food byproducts. In compliance with these challenges, the aim of this work was to more deeply investigate the antimicrobial activity of extracts derived from fermented tomato, melon, and carrot byproducts, previously studied. All the fermented extracts had antimicrobial activity both in vitro and in foodstuff, showing even higher activity than commercial preservatives, tested for comparison against spoilage microorganisms and foodborne pathogens such as Salmonella spp., L. monocytogenes, and B. cereus. These promising results highlight an unstudied aspect for the production of innovative natural preservatives, exploitable to improve the safety and shelf-life of various categories of foodstuff.


Introduction
Foodborne diseases have been an important health problem for all populations since the beginning of humanity. The types, severity, and impacts of these illnesses have changed over time and widely differ across regions, countries, and communities. All over the world, foodborne diseases are an important cause of morbidity and mortality, as well as an important impediment to socioeconomic development. Billions of people are at risk, and millions fall ill every year as the result of consuming unsafe food [1]. Among the microorganisms most relevant and most frequently notified by the RASFF in the last 40 years can be found Salmonella spp., Listeria monocytogenes, Escherichia coli, Staphylococcus aureus, and Bacillus cereus [2,3]. Moreover, nonpathogenic microorganisms, such as Pseudomonas spp., can be considered a serious problem because they can reduce the quality of contaminated food products [4]. To avoid foodborne disease risk and food deterioration, different methods for preserving food have been applied until today. Among them, antimicrobials may be used to upgrade food product safety by inhibiting/inactivating pathogenic microorganisms or by improving the shelf-life of food perishable by spoilage microorganisms [5]. Currently, consumers have concerns about the synthetic preservatives used in food [6,7]. Accordingly, an increasing number of consumers are demanding minimally processed foods and "cleaner" labels, leading to a great interest in natural antimicrobial utilization [5]. In the same direction, in recent years, there has been a growing increase in research concerning the investigation of antimicrobials from different natural sources. In fact, natural antimicrobials Lacticaseibacillus casei 2240, and Lacticaseibacillus casei 2246, belonging to the collection of the Food and Drug Department (University of Parma, Parma, Italy), were used as starters for tomato, melon, and carrot byproduct fermentation, respectively, as reported by   [15]. Briefly, the three sterilized substrates (tomato, melon, and carrot byproducts) were inoculated individually with the mentioned strain suspensions in order to obtain a final concentration of 7 log 10 CFU/mL. The inoculated byproducts were then incubated for 72 h at 37 • C. Fermented byproducts were then lyophilized and subjected to water/ethanol 50/50 (v/v) extraction as previously reported [15]. The extracts obtained were freeze-dried in order to obtain a powder and stored at −80 • C until their use.

Microorganisms Tested
The antimicrobial activity of extracts was tested toward 16  , and the collection of the Food and Drug Department (University of Parma, Parma, Italy). All strains were kept at −80 • C in tryptic soy broth (TSB) (Oxoid, Milan, Italy) supplemented with 12.5% glycerol (v/v). Before use, they were cultured twice for 24 h at 37 • C (Salmonella spp., L. monocytogenes, E. coli, S. aureus, and B. cereus) and 30 • C (Pseudomonas spp.) with a 3% v/v inoculum in TSB added with 0.6% yeast extract (Oxoid, Milan, Italy). Each revitalized culture was used to test the susceptibility to the fermented byproduct extracts.

Susceptibility Tests
The minimum inhibitory concentration (MIC) of fermented byproduct extracts was determined using the microdilution method according to Ricci et al. (2016) [21] with some modifications. Briefly, TSB was used, and microplates were incubated for 24 h at the optimal growth temperature tested. Due to the turbidity of extracts, the MIC determination was prone to error; therefore, all the dilutions were re-cultured (10 µL) on Tryptic Soy Agar (TSA) (Oxoid, Milan, Italy) and incubated at the same temperature for 24 h to determine the MBC [22]. Different microbial concentrations were tested (approximately 8 log 10 CFU/mL, 6 log 10 CFU/mL, 4 log 10 CFU/mL, and 2 log 10 CFU/mL).

Effect of Fermented Byproduct Extracts on Food Preservation
The antimicrobial activity against spoilage microorganisms exerted by fermented byproduct extracts was tested using minced meat and ready-to-eat sliced vegetal products as food models. These food models were selected according to their origin, both animal and vegetal, the high perishability that could make the use of an antimicrobial attractive, and the simplicity of the formulation that allowed a homogeneous distribution of the extract on a laboratory scale. Minced pork meat was bought from a local market and used to test the antimicrobial activity of fermented tomato byproduct extract. Different concentrations of extract were tested (0.8% (w/w), 1.2% (w/w), 1.6% (w/w), and 2.4% (w/w)) and, at the same time, compared with the most applied preservatives in meat (sodium lactate 1.2% (w/w) and sodium lactate/sodium diacetate (96:4) 2.5% (w/w)). The concentrations tested "in situ" were chosen in a range near the concentrations of common preservatives tested, lower than the MBC. The extract/preservative was added to the minced meat and mixed. Small hamburgers of 10 g were prepared and stored at refrigerated temperature. Total microbial count was analyzed immediately after the addition of extract/preservative (T0) and after 3 (T3), 6 (T6), and 9 (T9) days of storage. On the other hand, a ready-to-eat sliced vegetal product, mainly composed of water, melon, and thickeners, produced by a local food industry (Salumificio San Paolo S.r.l, Traversetolo, PR, Italy), was used to test the antimicrobial activity of fermented melon and carrot byproduct extracts. The extracts were added separately at 1.2% (w/w) during the preparation of the vegetal product. Afterward, it was sliced and stored under vacuum in refrigerated conditions. Total bacterial load was detected after product preparation and solidification (T0) and after 15 (T15) and 30 (T30) days of storage. For both products, the analyses were conducted by homogenizing the samples with ringer solution (VWR, Milan, Italy); then, decimal dilutions were spread on plate count agar (PCA) (VWR, Milan, Italy) and incubated for 72 h at 30 • C (UNI EN ISO 4833-1:2013). In both experiments, the same process was followed without adding extracts, and these samples were used as control. All experiments were performed in triplicate.

Effect of Fermented Tomato Byproduct Extracts on Foodborne Pathogens in Minced Meat
Minced pork meat was also used as a food model to simulate, through microbiological challenge test, the behavior of different foodborne pathogens (Salmonella spp., L. monocytogenes, and B. cereus) in case of eventual contamination. Before artificial contamination of the samples, revitalized strains belonging to the same species were mixed in equal concentration. Minced pork meat was added with fermented tomato byproduct extract 1.6%, sodium lactate 1.2%, or sodium lactate/sodium diacetate (96:4) 2.5%, while one sample was used as a control without the addition of extract/preservative. The contamination was carried out under sterile conditions with high microbial concentrations, approximatively ranging between 5 and 6 log 10 CFU/g in the product. All samples were stored in refrigerated conditions for 9 days, and the microbial contamination was checked just after the preparation of samples (T0) and after 3 (T3), 6 (T6), and 9 (T9) days. The experiments were performed in triplicate. Quantification of Salmonella spp., L. monocytogenes, and B. cereus was carried out at the scheduled time. The concentration of Salmonella spp. was detected using the medium Rambach agar (Merck, Milan, Italy), incubating the plates at 37 • C for 24 h. The determination of L. monocytogenes was performed by applying the indication reported in ISO 11290-2:2017 using Listeria selective agar base according to Ottaviani and Agosti (VWR, Milan, Italy). Enumeration of B. cereus was determined following the ISO 7932:2005 using a mannitol egg yolk polymyxin agar base (MYP) (VWR, Milan, Italy).

Statistical Analysis
One-way ANOVA was used to evaluate differences among treatments (addition or not of fermented byproduct extracts or preservatives) at the same storage time. The analyses were conducted with SPSS, version 21.0 software (IBM Corp., Armonk, NY, USA) using Tukey's test; a p-value < 0.05 was considered statistically significant.

Minimum Bactericidal Concentration
The MBC of fermented tomato, melon, and carrot byproduct extracts, which is the lowest concentration of extract responsible for the death of inoculated cells, was evaluated by considering four concentrations of the target microorganisms (Table 1). Regarding tomato byproduct extract, it had the highest activity against E. coli, S. aureus, B. cereus, and Pseudomonas spp. Overall, the tested strains had an MBC value of 25 mg/mL which decreased to 12.5 mg/mL at the lowest concentration of pathogen tested (2 log 10 CFU/mL). Salmonella spp. showed the same susceptibility with the exception of the suini strain, which was less affected than the other strains analyzed. L. monocytogenes seemed to be the species least inhibited, especially at high concentration, while, at 2 log 10 CFU/mL, the MBC was comparable with the other species studied. Melon byproduct extract, compared to the other extracts, revealed a more homogenous activity. Practically all the tested strains considered at diverse concentrations had the same MBC value (25 mg/mL) with the exception of two strains of Pseudomonas spp. (5003 and 5004), having the highest MBC at the highest concentration. The carrot byproduct fermented extract was the most effective (6.25 and 3.125 mg/mL) against all pathogens tested at the concentrations of 6, 4, and 2 log 10 CFU/mL.

Effect of Fermented Byproduct Extracts on Food Preservation
Considering a possible application in food, we decided to evaluate the effect of fermented tomato byproduct extract in minced pork meat and of fermented melon and carrot byproduct extracts in a ready-to-eat sliced vegetal product. Total bacterial load was evaluated in minced pork meat with the presence or absence of tomato byproduct fermented extract, at different concentrations, or with diverse preservatives. The changes in microbial concentration were evaluated over 9 days in refrigerated storage ( Figure 1). After 3 days (T 3 ), the highest decrease in microbial concentration was observed in minced meat with 1.2%, 1.6%, and 2.4% tomato by-product extract compared to T 0 . In minced meat, untouched or with sodium lactate, an increase in total bacterial load of 1.57 log 10 CFU/g and 1.15 log 10 CFU/g was observed, respectively. On the other hand, the adjunction of 0.8% extract and sodium lactate/sodium diacetate showed little to no changes in the initial total microbial count ( Figure 1). After 6 days (T 6 ), an increase in microbial concentration was observed for all the treatments done, except for sodium lactate/sodium diacetate. However, the total bacterial load of meat treated with sodium lactate/sodium diacetate was not statistically different (Table S1, Supplementary Materials) from that observed in meat with 1.6% and 2.4% of tomato byproduct extract. After 9 days, all the extract concentrations, with the exception of 0.8%, had an effect statistically comparable with sodium lactate/sodium diacetate, while meat not containing extract or sodium lactate showed a further bacterial load increase.  The efficacy of fermented melon and carrot byproduct extracts on food preservation was tested on a ready-to-eat sliced vegetal product mainly composed of melon. As reported in Table 2, it is clear that, just after the preparation, there was a positive antimicrobial effect of the fermented byproduct extracts. A difference in microbial load was clear between the ready-to-eat sliced vegetal product not containing extracts and that containing byproduct fermented extracts. Indeed, in the control sample (untouched ready-to-eat sliced vegetal product), an increase in the total microbial load was observed, especially after 15 days of storage. On the contrary, the employment of fermented melon and carrot byproduct extracts at 4 log 10 CFU/mL reduced the microbial load just after the preparation of the product, maintaining the microbial concentration throughout the storage period. Table 2. Total microbial load ± standard deviation (log 10 CFU/g) in ready-to-eat sliced vegetal product with (1.2%) or without fermented melon and carrot byproduct extracts after its preparation (T 0 ) and after 7 (T 7 ), 15 (T 15 ), and 30 (T 30 ) days of storage.

Effect of Fermented Tomato Byproduct Extract on Foodborne Pathogens in Minced Meat
Minced pork meat was also used as a food model to simulate, through microbiological challenge test, the behavior of different foodborne pathogens (Salmonella spp., L. monocytogenes, and B. cereus) in case of an eventual contamination. Salmonella spp. concentration increased in control samples over time until reaching the highest concentration after 9 days of refrigerated storage. The addition of sodium lactate did not affect Salmonella growth, while the addition of sodium lactate/sodium diacetate and fermented tomato byproduct extract slowed it down. The inhibitory effect of the extract was already noticeable after 3 days, whereas, after 9 days, a decrease of almost 1 log 10 CFU/g was observed compared to T 0 ( Figure 2). After 9 days, the difference between the control sample and meat with extract was 2.55 log 10 CFU/g. Furthermore, a significant difference in activity could be observed between fermented tomato byproduct extract and sodium lactate/sodium diacetate at the end of storage ( Ready-to-eat sliced vegetal product + fermented melon byproduct extract <1 <1 <1 <1 Ready-to-eat sliced vegetal product + fermented carrot byproduct extract <1 <1 1.83 ± 0.40 1.44 ± 0.62

Effect of Fermented Tomato Byproduct Extract on Foodborne Pathogens in Minced Meat
Minced pork meat was also used as a food model to simulate, through microbiological challenge test, the behavior of different foodborne pathogens (Salmonella spp., L. monocytogenes, and B. cereus) in case of an eventual contamination. Salmonella spp. concentration increased in control samples over time until reaching the highest concentration after 9 days of refrigerated storage. The addition of sodium lactate did not affect Salmonella growth, while the addition of sodium lactate/sodium diacetate and fermented tomato byproduct extract slowed it down. The inhibitory effect of the extract was already noticeable after 3 days, whereas, after 9 days, a decrease of almost 1 log10 CFU/g was observed compared to T0 (Figure 2). After 9 days, the difference between the control sample and meat with extract was 2.55 log10 CFU/g. Furthermore, a significant difference in activity could be observed between fermented tomato byproduct extract and sodium lactate/sodium diacetate at the end of storage (Table S2, Supplementary Materials). Behavior of Salmonella spp. in minced pork meat with or without fermented tomato byproduct extract/preservatives determined through microbial challenge test. Salmonella spp. concentration in minced pork meat (blue line), minced pork meat with 1.6% fermented tomato byproduct extract (red line), minced pork meat with 1.2% sodium lactate (green line), and minced pork meat with 2.5% sodium lactate/sodium diacetate (96:4) (purple line). Concentration is expressed as log10 CFU/g. A similar trend was followed by L. monocytogenes. The extract showed an inhibitory effect on L. monocytogenes after 9 days with a decrease in its concentration of 1.38 log10 CFU/g compared to T0. Sodium lactate/sodium diacetate was less effective against this foodborne pathogen species, while sodium lactate showed no significant effect ( Figure 3, Table S3, Supplementary Materials). Overall, at the end of storage, the difference between the control sample and meat with the extract was 1.76 log10 CFU/g. A similar trend was followed by L. monocytogenes. The extract showed an inhibitory effect on L. monocytogenes after 9 days with a decrease in its concentration of 1.38 log 10 CFU/g compared to T 0 . Sodium lactate/sodium diacetate was less effective against this foodborne pathogen species, while sodium lactate showed no significant effect ( Figure 3, Table S3, Supplementary Materials). Overall, at the end of storage, the difference between the control sample and meat with the extract was 1.76 log 10 CFU/g. B. cereus was able to grow in refrigerated conditions in minced pork meat. Sodium lactate did not stop its development at the tested concentration, while sodium lactate/sodium diacetate and extract inhibited the growth, maintaining the cell number until 6 days and allowing a slight growth between days 6 and 9. At the end of storage, the difference between the control sample and minced pork meat with extract was 2.24 log 10 CFU/g ( Figure 4, Table S4, Supplementary Materials). B. cereus was able to grow in refrigerated conditions in minced pork meat. Sodium lactate did not stop its development at the tested concentration, while sodium lactate/sodium diacetate and extract inhibited the growth, maintaining the cell number until 6 days and allowing a slight growth between days 6 and 9. At the end of storage, the difference between the control sample and minced pork meat with extract was 2.24 log10 CFU/g ( Figure 4, Table S4, Supplementary Materials).
.   B. cereus was able to grow in refrigerated conditions in minced pork meat. Sodium lactate did not stop its development at the tested concentration, while sodium lactate/sodium diacetate and extract inhibited the growth, maintaining the cell number until 6 days and allowing a slight growth between days 6 and 9. At the end of storage, the difference between the control sample and minced pork meat with extract was 2.24 log10 CFU/g ( Figure 4, Table S4, Supplementary Materials). .

Discussion
Consumers are increasingly concerned about synthetic preservatives used in foodstuff [6,7], perceived as not natural. As a consequence, the demand for minimally processed foods and "cleaner" labels is rising, and the research into new and efficient natural antimicrobials is growing in importance [5,23]. During lactic acid fermentation, LAB are able to produce different compounds exerting antimicrobial activity [15][16][17] such as hydrogen peroxide, CO 2 , peptides or proteins, bacteriocins, ethanol, acids such as acid, benzoic, sorbic, and citric lactic, and diacetyl, and their antimicrobial activity may be exerted via the combined action of different metabolites on undesirable bacteria [24]. Furthermore, some phenolic compounds, well documented in the literature for their antimicrobial activity, are naturally present in raw substrates, but their content can be implemented during lactic acid fermentation [25]. LAB are also able to produce phenyllactic acids [26], compounds with a widely documented activity on pathogenic microorganisms [27,28].
Considering this background and our previous work [15], in which the extracts obtained from fermented tomato, melon, and carrot byproducts exhibited antimicrobial activity, in the present work, we wanted to further verify the efficacy of the extracts directly in food. Initially, the MBC of the same extracts was determined considering 16 strains belonging to Salmonella spp., L. monocytogenes, E. coli, S. aureus, B. cereus, and Pseudomonas spp. Using the susceptibility test, we identified the concentration at which each extract caused the death of microorganisms tested. Fermented tomato, melon, and carrot byproduct extracts exhibited bactericidal activity against all tested strains. Overall, the highest efficacy (ranging from 25 to 3.13 mg/mL) was observed at the lowest bacterial concentration tested (2 log 10 CFU/mL). However, for most strains, the same extract concentration also resulted bactericidal at higher bacterial concentrations (4, 6, and 8 log 10 CFU/mL). Different studies reported in the literature focused their attention on natural antimicrobial activity, often based on essential oils. However, fewer studies took into consideration plant extracts [29][30][31][32][33], with even fewer considering the antimicrobial activity of fermented extracts [34][35][36] and only one exploiting the fermentation of byproducts to produce extracts [37], outside of our previous work [15]. These studies reported the effective antimicrobial activities of fermented extracts. Di Onofrio et al. (2019) [34] reported the effect of spontaneous fermented garlic extract on two strains of Pseudomonas aeruginosa. They reported an MIC of 4%, a value that was higher than the MBC of fermented tomato and carrot extracts observed in the present study for strains of the same genus (MBC ranged from 25 to 6.25 mg/mL). Pine needles were spontaneously fermented by Yim et al. (2006), showing activity against Gram-positive (S. aureus and B. cereus) and Gram-negative (Salmonella Typhimurium and P. aeruginosa) bacteria [35]. Codonopsis lanceolata was fermented with LAB by He et al. (2011) [36], and, after fermentation, a huge reduction in minimum inhibitory concentration was observed compared to the nonfermented plant, confirming that lactic acid fermentation positively affects the antimicrobial activity of different raw plant materials. Moreover, ginger marc was fermented with Lactiplantibacillus plantarum by Eom et al. (2017) [37], and the fermented product showed antimicrobial activity against S. aureus and E. coli. Considering a potential application to extend the shelf-life of foods, the efficacy of fermented tomato byproduct extract was tested in minced pork meat at different concentrations (0.8%, 1.2%, 1.6%, and 2.4%). We observed that the extract maintained a total microbial load lower than the control (minced pork meat without extract), and its activity was comparable (at 1.6% and 2.4%) with that of sodium lactate/sodium diacetate, a common preservative used for meat preservation [38,39]. Instead, the activity of fermented melon and carrot byproduct extracts (1.2%) was tested on a ready-to-eat sliced vegetal product throughout its shelf life. In this case, we also observed that extracts maintained a markedly lower total microbial load than the control, for all 30 days monitored. After the evaluation of the positive effect against spoilage microorganisms, we wanted to test in situ the effect against foodborne pathogens. In particular, we determined the efficacy of fermented tomato byproduct extract at 1.6% in minced pork meat on Salmonella, L. monocytogenes, and B. cereus, during refrigerated storage. Results of the microbial challenge test highlighted that Salmonella and L. monocytogenes were not able to grow when the extract was present, decreasing over 9 days, while B. cereus was inhibited until 6 days before a slight increase was observed between days 6 and 9 of storage. To the best of our knowledge, no study has reported the activity in pork meat of extracts obtained after lactic acid fermentation. However, different plant extracts were recently tested for pork meat preservation. Swamp cranberry fruit and pomace extracts were used at 2.5% in minced pork meat, showing a decrease in L. monocytogenes and Salmonella enteritidis of 4 log 10 CFU/g after 4 days of storage [40]. Another study [41] tested oakwood extract as a preservative at different concentrations (0.05%, 0.5%, and 1%) in pork patties, which did not have an effect on the total microbial count, unlike the fermented tomato, melon, and carrot byproduct extracts in this study, which exerted an inhibitory effect. In 2016, Novak and colleagues tested the antimicrobial activity of cherry and blackcurrant leaf extracts (0.5% and 1%, respectively) in vacuum-packed pork sausages in refrigerated storage conditions [42]. Mesophilic bacteria, psychotropic bacteria, LAB, and Enterobacteriaceae were not affected by the presence of the two extracts, whereas, after only 28 days of storage, Pseudomonas growth was lower compared to the control without extracts.

Conclusions
From the different reports available in literature, it is recognized that plants, as well as microorganisms, can be sources of antimicrobial compounds which can be applied for food preservation. However, the employment of both, exploiting the lactic acid fermentation process, is practically unexplored, with very few studies considering this opportunity [15,[34][35][36]. Despite the low number of papers on this topic, evidence suggests an improvement of antimicrobial activity during fermentation. To the best of our knowledge, no one, except for Ricci et al. (2019) [15], has used tomato, melon, and carrot byproducts as raw matrices fermentable for the recovery of antimicrobial compounds. Therefore, the present paper suggests a novel idea for the exploitation of these byproducts by coupling them with lacto-fermentation, recovering extracts characterized by antimicrobial activity. All these extracts demonstrated activity both "in vitro" and in foodstuff, showing that byproduct lacto-fermentation can be a strategy for the recovery of antimicrobials which can be applied for food preservation. Starting from the results achieved in this study, these extracts can be employed as natural preservatives to improve the safety and shelf-life of different categories of foodstuff, in line with the demand of consumers for cleaner and simpler labels replacing artificial preservatives perceived as "chemicals".
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/foods10051092/s1: Table S1. Effect of fermented extracts on food preservation; Table S2. Effect of fermented tomato extracts on Salmonella spp.; Table S3. Effect of fermented tomato extracts on L. monocytogenes; Table S4. Effect of fermented tomato extracts on B. cereus.