Bioprospecting Antimicrobials from Lactiplantibacillus plantarum: Key Factors Underlying Its Probiotic Action

Lactiplantibacillus plantarum (L. plantarum) is a well-studied and versatile species of lactobacilli. It is found in several niches, including human mucosal surfaces, and it is largely employed in the food industry and boasts a millenary tradition of safe use, sharing a long-lasting relationship with humans. L. plantarum is generally recognised as safe and exhibits a strong probiotic character, so that several strains are commercialised as health-promoting supplements and functional food products. For these reasons, L. plantarum represents a valuable model to gain insight into the nature and mechanisms of antimicrobials as key factors underlying the probiotic action of health-promoting microbes. Probiotic antimicrobials can inhibit the growth of pathogens in the gut ensuring the intestinal homeostasis and contributing to the host health. Furthermore, they may be attractive alternatives to conventional antibiotics, holding potential in several biomedical applications. The aim of this review is to investigate the most relevant papers published in the last ten years, bioprospecting the antimicrobial activity of characterised probiotic L. plantarum strains. Specifically, it focuses on the different chemical nature, the action spectra and the mechanisms underlying the bioactivity of their antibacterial and antiviral agents. Emerging trends in postbiotics, some in vivo applications of L. plantarum antimicrobials, including strengths and limitations of their therapeutic potential, are addressed and discussed.


Introduction
Lactiplantibacillus plantarum (L. plantarum), previously known as Lactobacillus plantarum, is a versatile species of lactobacilli. This subgroup of lactic acid bacteria (LAB) encompasses prokaryotes present in a range of diverse environments, including the gastrointestinal tract (GIT) of mammals, vaginal mucosa, food matrices, soil, and vegetable-associated niches. In this light, L. plantarum, in reason of a flexible behaviour, is found in association with all these niches (in several cases, with a dominant character) and is largely employed as starter cultures in the food industry. Selected L. plantarum strains are used to promote and/or carry-over fermentation processes that are functional to food production addressed to human consumption [1]. For its capacity to ferment and preserve food, enhancing its sensory properties and nutritional value, L. plantarum boasts a millenary tradition of safe use and thus shares a long-lasting relationship with humans. Moreover, like most Figure 1. Conceptualisation of the intake of probiotic-containing foods/beverages and of the most investigated beneficial properties exerted by probiotics, with a specific emphasis on the antimicrobial chemicals produced by lactobacilli. The indicated probiotic mechanisms are putative, not always sufficiently proven, and may depend on bacterial number, host gut microbiome composition and the specific probiotic strain. Image partially reproduced from Pop et al. [30] (copyright 2020 MDPI). In grey, an original picture of Lactiplantibacillus plantarum WCFS1 cells imaged by scanning electron microscope (SEM).

Emerging Trends in Probiosis, Postbiotics and Antimicrobials
While the terms probiotics, prebiotics, and synbiotics have been extensively defined in the last two decades and consensus documents have been provided by experts for each of them [5,31,32], the new concept of postbiotic is taking shape as an important microorganism-derived tool to promote health [10][11][12]. The need for a more precise terminology derived from observing the potential beneficial effect of non-viable microbial cells or effectors molecules contained, for example, in fer-mented foods (especially after prolonged storage) or in probiotic preparations (especially at the end of shelf life). The antimicrobial and antiviral activity exerted by beneficial microorganisms, such as L. plantarum, depends on a variety of molecules (organic acids, peptides, short-chain fatty acids (SCFA) and other antagonistic metabolites) that act with different action mechanisms and, sometimes, synergically produce the final bactericidal or bacteriostatic effects against the target microbes. Likewise, bacterial lysates have been shown to have some health benefits [33]. However, a precise distinction between the efficacy of nonviable beneficial bacteria, their growth products and their end-products on overall host health is not yet well defined, as each could have a beneficial role individually or in combination with the others. Very recently, postbiotics have been defined as a "preparation of inanimate microorganisms and/or their components that confers a health benefit on the hosts" [10], including in this concept the killed microbial cells with or without metabolites and excluding purified products (i.e., proteins, peptides, exopolysaccharides (EPS), SCFAs) [10]. The specific killed microorganism, the matrix and the inactivation method should be accurately indicated in the postbiotic definition. Until recently, the term postbiotics referred also to soluble factors secreted by live bacteria or Figure 1. Conceptualisation of the intake of probiotic-containing foods/beverages and of the most investigated beneficial properties exerted by probiotics, with a specific emphasis on the antimicrobial chemicals produced by lactobacilli. The indicated probiotic mechanisms are putative, not always sufficiently proven, and may depend on bacterial number, host gut microbiome composition and the specific probiotic strain. Image partially reproduced from Pop et al. [30] (copyright 2020 MDPI). In grey, an original picture of Lactiplantibacillus plantarum WCFS1 cells imaged by scanning electron microscope (SEM).

Emerging Trends in Probiosis, Postbiotics and Antimicrobials
While the terms probiotics, prebiotics, and synbiotics have been extensively defined in the last two decades and consensus documents have been provided by experts for each of them [5,31,32], the new concept of postbiotic is taking shape as an important microorganism-derived tool to promote health [10][11][12]. The need for a more precise terminology derived from observing the potential beneficial effect of non-viable microbial cells or effectors molecules contained, for example, in fer-mented foods (especially after prolonged storage) or in probiotic preparations (especially at the end of shelf life). The antimicrobial and antiviral activity exerted by beneficial microorganisms, such as L. plantarum, depends on a variety of molecules (organic acids, peptides, short-chain fatty acids (SCFA) and other antagonistic metabolites) that act with different action mechanisms and, sometimes, synergically produce the final bactericidal or bacteriostatic effects against the target microbes. Likewise, bacterial lysates have been shown to have some health benefits [33]. However, a precise distinction between the efficacy of non-viable beneficial bacteria, their growth products and their end-products on overall host health is not yet well defined, as each could have a beneficial role individually or in combination with the others. Very recently, postbiotics have been defined as a "preparation of inanimate microorganisms and/or their components that confers a health benefit on the hosts" [10], including in this concept the killed microbial cells with or without metabolites and excluding purified products (i.e., proteins, peptides, exopolysaccharides (EPS), SCFAs) [10]. The specific killed microorganism, the matrix and the inactivation method should be accurately indicated in the postbiotic definition. Until recently, the term postbiotics referred also to soluble factors secreted by live bacteria or released after bacterial lysis, also known simply as cell-free supernatants (CFS), i.e., quite heterogeneous mixtures including SCFAs, cellular enzymes, peptides, teichoic acids, peptidoglycan-derived muropeptides, EPS, cell surface proteins, vitamins, plasmalogens, and organic acids, which could have the same health beneficial effects of the strain that generated them [34]. However, a precise boundary line between what is currently defined postbiotic and what is not (i.e., CFS) is actually very difficult to delineate because some chemically synthesised compounds and/or metabolites might be present in both viable and not viable microbial cell preparations ( Figure 2). Likewise, other scientists have pointed to some inconsistencies and ambiguities associated with the recently proposed re-definition of postbiotics [11]. Therefore, we wonder whether microbiologists should coin a new term to correctly define a probiotic bacterial CFS, which in most research work, is filtered to eliminate cells and cellular debris. Nonetheless, according to the International Scientific Association of Probiotics and Prebiotics (ISAPP), the term CFS is sufficiently defined as such and further definitions are not deemed as necessary [12].
peptidoglycan-derived muropeptides, EPS, cell surface proteins, vitamins, plasmalogens, and organic acids, which could have the same health beneficial effects of the strain that generated them [34]. However, a precise boundary line between what is currently defined postbiotic and what is not (i.e., CFS) is actually very difficult to delineate because some chemically synthesised compounds and/or metabolites might be present in both viable and not viable microbial cell preparations ( Figure 2). Likewise, other scientists have pointed to some inconsistencies and ambiguities associated with the recently proposed re-definition of postbiotics [11]. Therefore, we wonder whether microbiologists should coin a new term to correctly define a probiotic bacterial CFS, which in most research work, is filtered to eliminate cells and cellular debris. Nonetheless, according to the International Scientific Association of Probiotics and Prebiotics (ISAPP), the term CFS is sufficiently defined as such and further definitions are not deemed as necessary [12].
In this context, not yet clearly defined, we gathered and reported in temporal order (Table 1) the data from the last decade on the chemical nature of compounds with antibacterial and/or antiviral activity as found in the CFS of probiotic L. plantarum strains. In addition, the investigated action mechanisms of the single antimicrobial substances are reported. We realise from these studies that, despite trying to understand and dissect the action mechanism of the single antimicrobial compound, the overall mechanisms of complex mixtures underlying food preservation and human/animal health are still far from being fully understood.  In this context, not yet clearly defined, we gathered and reported in temporal order (Table 1) the data from the last decade on the chemical nature of compounds with antibacterial and/or antiviral activity as found in the CFS of probiotic L. plantarum strains. In addition, the investigated action mechanisms of the single antimicrobial substances are reported. We realise from these studies that, despite trying to understand and dissect the action mechanism of the single antimicrobial compound, the overall mechanisms of complex mixtures underlying food preservation and human/animal health are still far from being fully understood.

Nature and Mechanisms of L. plantarum Antimicrobials
Two modalities are basically involved in the antimicrobial action: (i) bacteriostatic, which means that the antimicrobial agent prevents the growth of target microbe, and (ii) bactericidal, which means that the agent kills the target cell. However, the exact microbiological bactericidal or bacteriostatic setting of bacterial-related agents may be influenced by growth conditions, bacterial density, duration of the test used, and measure of reduction in bacterial numbers [35].
The antimicrobials produced by lactobacilli are quite diverse and fall within two main chemical categories, proteinaceous and non-proteinaceous substances. Data reported in Tables 1 and 2 show that the antimicrobial activity of L. plantarum is mainly exerted by bacteriocins (~60% of the reported strains) or partially characterised proteinaceous compounds, followed by organic acids or acidic conditions (i.e., acid CFS [25]), and biosurfactants (BS) such as glycoproteins and EPS.

Bacteriocins
Bacteriocins are a heterogeneous group of ribosomally synthesised, gene-encoded peptides with specific antimicrobial activity towards a spectrum of target microbes, which can be narrow or broad depending on the producing strain [23]. The bacteriocins produced by L. plantarum spp. are referred to as plantaricins; generally, they are small, heat-stable, frequently very potent, being active at nanomolar concentrations, and exert their killing effect mostly through membrane permeabilization, through pore formation and subsequent leakage of cytoplasmic compounds. For their high antimicrobial versatility, they have been considered for use as bio-preservatives, antibiotic alternatives, health-promoting gut modulators and animal growth promoters [36]. As observed for other bacteriocins, the production of plantaricins is regulated through a quorum-sensing based network and seems to be switched on by specific signals that characterise the natural niche of the strain [37]. Stress conditions and co-culture with other specific inducing bacteria usually enhance plantaricin gene expression [38][39][40]. Likewise, the accumulation of signalling molecules, such as the PlnA peptide, outside the bacterial cells, can modulate bacteriocin production [40]. Intriguingly, transcriptomic and genomic studies in the reference strain L. plantarum WCFS1 pointed out that plantaricins synthesis may be triggered in vivo, during transit along the gut [41,42], thereby providing competitive advantages over other intestinal resident microbes, contributing to microbiota balance and possibly playing a role in the molecular interplay with the host immune system [43,44].
Like bacteriocins, even plantaricins can be roughly divided into two classes, which, in turn, include subclasses: Class I, containing post-translationally modified peptides, and Class II, containing predominantly unmodified peptides. Being amphiphilic, plantaricins interact electrostatically with the negatively charged bacterial cell surface (mainly constituted by phospholipids and lipopolysaccharides) by their positively charged group (cationic plantaricins) [45], while their hydrophobic portion passes through the membrane lipid bilayer of target bacteria. After internalisation, the peptides aggregate to form polymers or complexes, which create holes in the cell wall and the membrane, causing a change in permeability with consequent leakage and depletion of intracellular compounds (e.g., genetic and proteinaceous material), eventually leading to cell membrane lysis [46][47][48][49].
A few studies reported in Tables 1 and 2 investigated the mechanism of action of plantaricins; among them, Kim and co-workers showed that plantaricins 3 and 5, from L. plantarum NIBR97, inhibited S. Enteritidis growth, causing its cellular lysis by damaging the membrane via pore formation [48]. Synthetic plantaricins 3 and 5 were further investigated for antiviral activity against GFP-labelled lentiviruses and, interestingly, scanning electron micrography (SEM) revealed that plantaricin 3 caused lentiviral lysis in human host cells through the collapse of their envelopes, while plantaricin 5 did not, implying two different antiviral mechanisms [50]. Tenea and co-workers investigated the mechanism by which Gt2 peptides and Cys5-4 peptides, from L. plantarum UTNGt2 and UTNCys5-4, respectively, target Gram-negative bacteria (Table 1). Gt2 and Cys5-4 peptides comprise, in turn, two or more post-translationally modified peptides forming one aggregate forming one functional inhibitory unit, which altered the cellular membrane permeability of E. coli and Salmonella, causing the leakage of cytoplasmic contents, followed by cellular death [50,51]. In a later study, the same authors investigated deeply the changes in Salmonella cells treated with Gt2 and Cys5-4 extracts, highlighting the occurrence of direct interaction between Salmonella genomic DNA and the peptides [46]. Such interaction has also been observed for antibacterial compounds from other species and was suggested to concur to the antimicrobial effects, i.e., by inducing DNA damages and genomic instability [52].
Data collected so far suggest that the action mode for plantaricins is usually bactericidal, through the induction of ion-selective pores in the target cell membrane, which causes the dissipation of intracellular ATP and depletion of the proton motive force, leakage of intracellular substances, followed by cell death. However, bacteriostatic effects have been reported for plantaricin W [53] and plantaricin LpU4 [54], as well as for other plantaricins previously studied [55,56]. In these cases, it has been suggested that plantaricin binds to target cell surface without reaching the specific receptors needed to achieve the killing effect [53]. In addition, certain L. plantarum strains produce more than one bacteriocin, either acting individually, e.g., bacteriocins F1 and F2 which exhibit different levels of efficacy against Staphylococcus aureus [57], or whose combination/synergic action results in higher antibacterial efficacy, e.g., plantaricin KL-1Y [58]. It is also worth mentioning that some plantaricins, such as KL-1Y [58], ZJ008 [59] and BM-1 [60], exert a bactericidal action without apparent cell lysis, whereas some, e.g., BM-1, can inhibit Gram-negative bacteria growth by bacteriostatic action, by influencing metabolic pathways and affecting the cell wall structure, eventually causing its collapse [61].
In addition to plantaricins, other L. plantarum proteinaceous compounds have been ascribed antibacterial activity. In most studies, such antimicrobials are generically referred to as bacteriocin-like peptides. For instance, the extracellular anti-staphylococcal protein fractions produced by a strain isolated from meat were recently identified as a couple of enzymes (i.e., transglycosylase and glyceraldehyde-3-phosphate dehydrogenase (GADPH)), which inhibit S. aureus growth through different mechanisms [62]. Extracellular transglycosylase binds to S. aureus peptidoglycan, thereby degrading the cell wall until cell lysis occurs; once the membrane has been damaged by transglycosylase, GADPH enters the cell and up-regulates S. aureus autolysis genes [62].
Considering the key role of the membrane surface charge and fluidity in the action mode of bacteriocins, it is easy to guess that the manipulation of these two bacterial properties may render the bacteriocins ineffective, resulting in bacteriocin resistance [63,64], or, on the other hand, in higher efficacy, which might be achieved for instance by bioengineering [65]. That being said, this represents an important mechanism of bacterial adaptation [66] that deserves to be studied and treated in depth separately, as it goes beyond the focus of this review.

Organic Acids
L. plantarum antimicrobial activity also often depends on the production and release of different types of organic acids (primarily, lactic and acetic acids, to follow tartaric, citric, malic, oxalic, and succinic acid) and on the associated pH lowering of the surrounding environment. Both factors concur to hinder the survival of acid-sensitive microorganisms [67]. Even short-chain fatty acids (SCFAs), e.g., butyric, propionic and valeric acids, and their modified derivatives have been ascribed antibacterial activity.
Since the growth of important food-spoilage and food-poisoning microorganisms is inhibited at low pH (<4), L. plantarum, which is one of the lactobacilli with the highest lactic acid production rate [25,68], can be added to many fermented foods as a natural preservative [69,70]. LAB, including L. plantarum, may be homofermentative or heterofermentative depending on the pathway used for glucose fermentation. Homofermentative bacteria produced more lactic acid through glycolysis compared to heterofermentative bacteria, which ferment glucose via the 6-phosphogluconate/phosphoketolase pathway [71]. Because different strains produce different types, amounts and combinations of organic acids, the resulting overall inhibitory actions are quite variable. There is much here to indicate that the antimicrobial mechanism resulting from pH acidification is species-and strain-specific [70,72]. The hydrophobic, undissociated form of the acid permeates the cell membrane and dissociates inside the target cell as a consequence of the intracellular neutral pH conditions, thereby acidifying the cytoplasm. The acid pH and the neutralisation of the electrochemical proton gradient stops various pH-dependent transport mechanisms causing bacteriostasis and eventually cell death [73,74]. The pH variation of the cytoplasmic environment depends on the specific pKa values of the organic acids produced by the L. plantarum strain, which explains the variability of their antimicrobial action [68]. Furthermore, by their chelating properties, organic acids can capture essential growth elements, such as iron [75].
Some modified acids have also been ascribed antibacterial activity. A derivative of propionic acid, i.e. 2-(2-1 mino-1-hydroxyethoxy) ethyl 2-methylpropanoate (LPB 102), was found to be the antimicrobial agent produced by L. plantarum NTU 102, with inhibitory action against Vibrio parahaemolyticus, a bacterium that is frequently associated with foodborne outbreaks of disease [76]. The authors attributed the inhibitory effects of LPB 102 to the suppression of specific V. parahaemolyticus genes that underlie its intrinsic resistance to various antimicrobial agents [77].
3-Phenyl lactic acid (PLA), a metabolite produced by some LAB from phenylalanine catabolism [78], is quite a new type of powerful and broad-spectrum antimicrobial compound that is active against both bacteria and fungi [79]. For its capacity to contrast food spoilage microorganisms, it is considered a valuable natural food preservative. Therefore, PLA biosynthesis, and strategies to increase its yield in starter LAB, have attracted much research effort [78,80]. PLA occurs in two enantiomers, L-PLA and D-PLA, whose difference in antibacterial capacity is still debated [81,82]. In a few pathogenic species, among those that are sensitive to PLA, this compound has been demonstrated to function by targeting the bacterial membrane [83,84], i.e., affecting its charge distribution and hydrophobic properties [85]. In Listeria monocytogenes, a mixture of the two isomers, obtained from a fermented vegetable L. plantarum isolate, was found to disrupt the cell membrane, and induce pore formation and leakage of intracellular material by interacting with cell membrane proteins [86]. Such mechanism of action has been recently confirmed also for the anti-Salmonella activity of PLA chemically characterised and purified from an infant faeces-isolated L. plantarum strain [26]. Intriguingly, the authors observed that, besides destroying the cell membrane, the purified compound was able to intercalate genomic DNA, suggesting a further mode of action for this molecule [26].
A few studies have characterised the action mechanism of proteinaceous BS from L. plantarum spp. Crude BS extracted from an L. plantarum cheese isolate were found to counteract, dose-dependently, biofilm formation of S. aureus, apparently by affecting the expressions of biofilm-related genes and by interfering with quorum-sensing signalling [94]. The BS produced by L. plantarum 60FHE was structurally characterised and identified as a mixture of glycoproteins, which exert antimicrobial activity against some foodborne pathogens (Table 1), possibly through penetration into the cell and by rupturing the membrane, leading to cell lysis [91]. Interestingly, the biosurfactant produced from this strain was also shown to be a potential anti-cancer agent [91].
Some L. plantarum exopolysaccharides (EPS) have BS-related antimicrobial properties. EPS, i.e., hydrophilic extracellular high-molecular-mass polymers, are produced by different LAB and exhibit high structural diversity in terms of sugar compositions, type of bonds between the repeating units, chain length, branching, and non-sugar modifications [87]. In L. plantarum the ability to synthesise EPS is a strain-specific trait and requires the presence of specific gene clusters encoding for regulatory factors and enzymes that enable biosynthesis and assembly of the sugar monomers and secretion of the polysaccharide [98]. In addition to the antimicrobial [99] and antibiofilm activities [100][101][102][103][104], L. plantarum EPS have been ascribed other properties that may impact the interaction with the host and can account for the health benefits provided by probiotic lactobacilli, such as immune-stimulating [98,105,106], antioxidant [107] and anti-cancer activities [108,109].
Generally, EPS from LAB, including L. plantarum, exert their antibacterial activity by interfering with the adhesion to surfaces and with cell adhesion/recognition mechanisms, thereby contrasting the formation of biofilm [99,103]. Biofilms are surface-associated, complex microbial communities, embedded in a self-synthesised polymeric matrix. These multicellular, three-dimensional structures develop thanks to inter-cellular signalling and through modulation of cell adhesion properties, and can confer to microbes a greater resistance to antibiotics [110]. Song and co-workers reported that EPS produced by L. plantarum 12 exert antibiofilm activity against Shigella flexneri, a foodborne enteric pathogen that can induce bacillary dysentery [101]. The authors found that the active form of its EPS (i. e., L-EPS) decreased polysaccharide production in the extracellular polymeric matrix of S. flexneri only by direct contact with the pathogen and without affecting its growth. L-EPS were hypothesised to disturb the signalling involved in biofilm formation and to interfere with the extracellular polymeric structures of the pathogen, which is crucial for maintaining the integrity of its biofilm [101]. Within biofilms, which can colonise the surfaces of medical equipment or food, microbial cells gain a greater resistance to disinfectants and conventional drugs, thus representing a serious global health concern [110].
Interestingly, EPS from a cheese-isolated L. plantarum strain were found both to inhibit E. coli biofilm formation and to reduce the activity of efflux pumps implicated in drug resistance [100]. The authors ascribed the antibiofilm effect to a decreased production of indole, i.e., a metabolite putatively involved in virulence and in the quorum-sensing systems sustaining biofilm development, and to a reduction of cell surface hydrophobicity, as observed in EPS-treated E. coli cells [100]. Figure 3 sums up the different chemical nature of the principal antimicrobial agents produced by L. plantarum strains, and the suggested mechanisms underlying their antimicrobial effect.
What we are looking at reinforces the hypothesis that the antimicrobial action of a probiotic such as L. plantarum may not be due to a single molecule but to the synergic action of several molecules produced by the strain and released into the environment (i.e., growth media, food matrix or gut, according to application and niche of the strain). This is further supported by the observation that in most of the works on the isolation of the antimicrobial agents, CFS show a wider antimicrobial inhibition spectrum compared to the isolated antimicrobial agents [40,48,76,111]. The question is whether antimicrobial action of the isolated compound has been tested, or not, on all target pathogens used to test also the corresponding CFS. Namely, in most published papers, the antibacterial activity of CFS is usually tested on a broad spectrum of pathogenic bacteria, while the single CFS-derived antimicrobial compound is assayed on a single target. This can be a choice dictated by many reasons and does not necessarily mean that the isolated compound is active only against that specific target pathogen (i.e., the ones reported in Table 1). Then, perhaps, in view of applications in the food industry and human medicine, as bio-preservatives and bio-therapeutics, we should rethink our research of the single antibacterial agent (spending time and money in using complicated technologies) and focus more on the entire bacterial product (growth media), rather standardising times and methods of CFS collection and processing. implicated in drug resistance [100]. The authors ascribed the antibiofilm effect to a decreased production of indole, i.e., a metabolite putatively involved in virulence and in the quorum-sensing systems sustaining biofilm development, and to a reduction of cell surface hydrophobicity, as observed in EPS-treated E. coli cells [100]. Figure 3 sums up the different chemical nature of the principal antimicrobial agents produced by L. plantarum strains, and the suggested mechanisms underlying their antimicrobial effect. What we are looking at reinforces the hypothesis that the antimicrobial action of a probiotic such as L. plantarum may not be due to a single molecule but to the synergic action of several molecules produced by the strain and released into the environment (i.e., growth media, food matrix or gut, according to application and niche of the strain). This is further supported by the observation that in most of the works on the isolation of the antimicrobial agents, CFS show a wider antimicrobial inhibition spectrum compared to the isolated antimicrobial agents [40,48,76,111]. The question is whether antimicrobial action of the isolated compound has been tested, or not, on all target pathogens used to test also the corresponding CFS. Namely, in most published papers, the antibacterial activity of CFS is usually tested on a broad spectrum of pathogenic bacteria, while the single CFS-derived antimicrobial compound is assayed on a single target. This can be a choice dictated by many reasons and does not necessarily mean that the isolated compound is active only against that specific target pathogen (i.e., the ones reported in Table 1). Then, perhaps, in view of applications in the food industry and human medicine, as bio-preservatives and bio-therapeutics, we should rethink our research of the single antibacterial agent (spending time and money in using complicated technologies) and focus more on the entire bacterial product (growth media), rather standardising times and methods of CFS collection and processing.

Antibacterial and Antiviral Spectrum of L. plantarum Extracellular Compounds
A key role of probiotics is that of preventing infections in the host, maintaining a healthy and balanced intestinal microbiota; likewise, microbes intended for use as starter and food preservatives should enhance food quality and safety, limiting contamination by fastidious and potentially dangerous microbial species. Therefore, a powerful, broadspectrum antibacterial and antiviral activity against pathogens is strongly desirable, both as whole cells and as growth products/metabolites (intra and extracellular). Table 1 shows that L. plantarum bacteriocins are effective against several pathogenic bacteria, including clinically relevant pathogens such as L. monocytogenes (~50% of the reported bacteriocins), a Gram-positive species which is widespread in nature (i.e., soil, vegetation, mammalian cells), robust, able to grow at refrigeration temperatures, and also recognised for a long time as a cause of human disease. Indeed, listeriosis can determine sepsis in immunocompromised patients, meningoencephalitis and febrile gastroenteritis [112]. The activity of these proteins against L. monocytogenes, sometimes regardless of pH [113], besides probiotic properties and safety of the producing strain, allow the development of novel bio-preservatives, with potential use in the food industry. At present, only two bacteriocins have been given the GRAS status, being approved for use as natural food preservatives, both produced by LAB (i.e., nisin, from Lactococcus, and pediocin PA-1D, from Pediococcus genus) [114]. Organic acids, mainly lactic acid, produced from L. plantarum extracted from kimchi [115] and bean [73], also showed to inhibit L. monocytogenes as well as, almost to the same extent, other pathogens, making the corresponding strains potentially useful as starter culture [73,115].
L. monocytogenes, along with S. aureus and E. coli, is also considered a foodborne pathogen, as these bacteria can produce enterotoxins in contaminated food. Several different plantaricins were found to be active also against Listeria innocua, the closely related but non-pathogenic Listeria species, often used in laboratories as a surrogate organism for a better understanding of the behaviour of the pathogen during food processing [116].
Quite a few bacteriocins from L. plantarum inhibit S. aureus (~60% of the studies reported in Table 1), one of the most common pathogens that can colonise intestine, skin tissues and perineal regions of the human host, causing severe infectious diseases, such as osteomyelitis, endocarditis, pneumonia, septicaemia, and health hazardous effects worldwide [117]. Indeed, S. aureus represents the most common microorganism causing infections in communities with very high economic burden at the social level; furthermore, it can develop considerable resistance towards conventional antimicrobial agents, with major prevalence of methicillin-resistant S. aureus (MRSA) and vancomycin-resistant S. aureus (VRSA) forms. For instance, MRSA accounted for 16% of necrotising soft tissue infections worldwide, although overall mortality is declining over the last ten years [117,118]. Examples of antimicrobials against these resistant forms comprise the CFS of L. plantarum strains extracted from sauerkraut [113] and from faecal microbiota [102], plantaricins LpU4 and ZJ008 from L. plantarum strains isolated from milk [54,59]. These extracellular compounds and their strains may represent an alternative bio-control strategy against skin infections. In addition, L. plantarum enzymes, i.e., transglycosylase and glyceraldehyde-3-phosphate dehydrogenase (GADPH), biosurfactants [90,94] and EPS [104] (in a dose-dependent manner), and, to a lesser extent, organic acids [73], all showed antagonistic activity against S. aureus [62].
Together with L. monocytogenes and S. aureus, E. coli is the most frequently inhibited by plantaricins (more than 50% of the L. plantarum strains/studies as reported in Table 1). This Gram-negative species represents the most prevalent commensal inhabitant of the human gastrointestinal tract, as well as one of the most common human and animal pathogens, being acknowledged as the causative agent of multiple clinical syndromes such as diarrhoeal diseases, meningitis and urinary tract infections [119]. In fact, although this bacterium is usually a benign gut commensal, some strains can acquire virulence, becoming able to cause diarrhoea in humans and other animals, and making E. coli one of the most widely studied etiologic agents worldwide [120]. Pathogenic E. coli forms causing diarrhoea have been classified into different pathotypes, including, among others, enterotoxigenic E. coli (ETEC), enteropathogenic E. coli (EPEC), Shiga toxin-producing E. coli (STEC), enteroinvasive E. coli (EIEC), enteroaggregative E. coli (EAEC), and diffusely adherent E. coli (DAEC) [120]. Interestingly, as shown in Table 1, ETEC and EPEC are inhibited mostly by organic acids and other unidentified extracellular compounds produced by several L. plantarum strains. Furthermore, EPS purified from a breast milk L. plantarum isolate demonstrated an excellent capacity to inhibit the adhesion of E. coli to epithelial human cells [107].
Bacillus cereus is another common food contaminant with highly variable pathogenic potential ranging from strains that show little or no cytotoxic in vitro activity, to forms that are highly cytotoxic [128]. B. cereus can be responsible for two types of poisonings, depending on the toxin it produces, resulting in diarrhoea and emesis [129] (that in severe cases require hospitalisation and are sometimes fatal). B. cereus is also recognised as an aetiological agent of localised wound, eye and systemic infections [128]. Almost all plantaricins reported in Table 1 have been found active against B. cereus (corresponding to approximately 20% of the L. plantarum strains reported in Table 1) and in some cases, as for plantaricin GZ1-27, time-and dose-dependent activity was demonstrated [111]. Organic acids from L. plantarum S0/7 [73] and EPS from L. plantarum R315 [99] showed also an inhibitory activity against B. cereus. In addition, EPS from L. plantarum R315 were reported to inhibit B. cereus, other foodborne pathogens described above and Cronobacter sakazakii, an opportunistic Gram-negative bacterium that survives in very dry niches, and can contaminate food such as powdered infant milk, causing neonatal infections with high fatality rates [130,131].
Pseudomonas aeruginosa is another clinically relevant species, i.e., a Gram-negative, opportunistic pathogen with a high intrinsic resistance to a wide variety of antibiotics. P. aeruginosa is often found in medical equipment, such as inhalers, dialysis equipment, respirators, vaporisers, in toilets and sinks [132] and, consequently, it is the cause of several kinds of hospital-acquired infections, such as catheter-associated urinary tract infections [57], ventilator-associated pneumonia, gastrointestinal infections, dermatitis, skin infections, bacteraemia, bone and joint infections, and other infections, particularly in patients with severe burns and in immunocompromised subjects (i.e., suffering from cancer or AIDS) [133]. As shown in Table 1, the antimicrobial activity of L. plantarum against P. aeruginosa is mainly due to plantaricins. In addition, EPS produced by L. plantarum isolated from human breast milk showed a very strong inhibition for P. aeruginosa, higher compared to inhibition of other foodborne pathogens described so far [107].
Researchers have focused mainly on the antibacterial and antifungal properties of L. plantarum compounds, whereas their antiviral action has been much neglected. Table 2 reports studies, mostly in vitro, documenting such activities. Plantaricins 3 and 5, produced by L. plantarum NIBR97, were found to exhibit antibacterial activities against a broad range of pathogens (Table 1), plus significant antiviral activities against the human pathogen influenza A virus (H3N2) ( Table 2) [48]. Therefore, they were recently suggested as potential natural disinfectants, which might be an alternative to the chemical ones (alcohol-or chlorine-based preparations), for the disinfection of hands and surfaces in conditions of pandemics [48]. The proliferation of the influenza A virus was also found to be inhibited by proteinaceous compounds from L. plantarum LBP-K10 [147].
Other poorly defined extracellular metabolites from various L. plantarum strains (as reported in Table 3) could inhibit in vitro (i) Echovirus, enteroviruses isolates recovered from acute flaccid paralysis cases [148]; (ii) transmissible gastroenteritis virus (TGEV), which causes many gastrointestinal infections in piglets, characterised by diarrhoea and high mortality [149]; (iii) enterovirus Coxsackievirus B4, a challenging virus, infections of which have been linked to the onset of type 1 diabetes [150]; (iv) porcine epidemic diarrhoea virus (PEDV), a coronavirus responsible of one of the highly contagious viral diseases in the pig industry, causing severe (sometimes fatal) diarrhoea in piglets [151]; (v) human rotavirus, i.e., the causative agent of severe diarrhoea in newborns and children worldwide [152].
Notably, the addition of prebiotics, such as those derived from microalgae, was reported to enhance both viability and antiviral effects of probiotics, as was observed for L. plantarum ATCC LP299v [153]. The anti-rotavirus action could account for beneficial effects of probiotics (mainly bifidobacteria and lactobacilli, including L. plantarum) in preventing enteric infections and alleviating diarrhoea symptoms [154]. In fact, dietary intake of L. plantarum LRCC5310, whose EPS were shown to inhibit the growth of rotavirus in vitro and in mice model [152], was subsequently found to be effective and safe in patients with rotaviral enteritis [154]. Table 3. L. plantarum strains with antimicrobial activity, whose probiotic and antipathogenic efficacy was tested in vivo (clinical and/or preclinical investigations). The spreading drug resistance in all cited microbial pathogens makes it difficult to treat and eradicate them and represents a severe problem for public health, requiring the development of alternative antimicrobial strategies. In this regard, the characterisation of antibacterial extracellular compounds produced by L. plantarum species opens new horizons in managing drug resistance. Particularly, bacteriocins might help to tackle antibiotic-resistant bacterial pathogens, a phenomenon that has become a worldwide threat, considering that the number of deaths per year due to antimicrobial resistance is predicted to exceed that of people who die from cancer [155,156].

In Vivo Studies on L. plantarum Strains Whose Antibacterial Activity Was Earlier Characterised In Vitro
In the medical field, L. plantarum is being investigated for an increasing number of applications such as: healing of skin wounds and burn infections [102,157,158]; treatment of mucosal infections [159,160]; protection from environmental mutagens [161,162]; amelioration of acute and chronic GIT infections [163,164], gut inflammatory disorders and urinary tract infections [165]; cholesterol level-lowering properties [166]; and beneficial effects on obesity [167], diabetes [168], colon cancer [169] and cognitive impairments [170]. Such broad range of possible utilisations reflects the genomic diversity of L. plantarum, which entails its large phenotypic diversity, versatility and flexibility [171].
This paragraph briefly discusses only the studies (reported in Table 3) which combine both in vitro and in vivo approaches to assess the antimicrobial ability of L. plantarum strains/compounds and those that investigate in vivo L. plantarum strains and/or related metabolites, whose antimicrobial effectiveness was earlier characterised in vitro (i.e., reported in Table 1). Noticeably, the number of in vitro studies (Table 1) far exceeds those where such strains and/or their isolated compounds are tested through in vivo experiments (Table 3). Most likely, we have to consider that, for example, the path from the discovery of the antibacterial activity of bacteriocins (by in vitro models) to their application as therapeutic agents is long, and involves many crucial steps to advance into clinical trials, such as the use of animal models and studies on toxicity and biosafety in vivo [172].
Moreover, it is worth highlighting that, in several of the studies reported in Table 3, it is not possible to conclude whether the observed effect is caused by antimicrobials alone or by a combination of factors, e.g., by antimicrobials or the L. plantarum strain per se, by nutrient competition, or through the induction of host antimicrobial proteins.
A few preclinical studies have prospected the use of L. plantarum-derived bacteriocins as a promising tool to control post-operative infections. In two independent studies, systemic treatments based on the intravenous injection of L. plantarum bacteriocins were shown to contrast S. aureus bone fracture-associated infections. Using two different animal models (i.e., rabbit and mice), and upon bacteriocin treatment, the authors observed a reduction of pathogen biofilm and a decreased serum level of pro-inflammatory markers [173,174]. Coherent findings were also reported by similar investigations using bacteriocins from other lactobacilli [175]. The above-mentioned preclinical studies rely on systemic administration of the isolated antimicrobials. However, a higher therapeutic effect could be probably achieved by topic application of the antimicrobial compounds, i.e., by their direct administration at the host surface, such as oral cavity, gut, skin, and urogenital mucosa, as was also recommended for postbiotics [10]. For instance, antimicrobials could be helpful to promote healing and prevent infection at wounds, ulcers and burn sites [176][177][178]. However, so far, most of such studies concerning L. plantarum cell extracts and/or secreted metabolites have been carried out mainly in vitro [158].
Recently, an elegant study in mice demonstrated the key role of bacteriocin for the anti-infective action of probiotics in the host gut [179]. This study provides clear molecular evidence that protective and anti-listerial effects of a bacteriocin-producing L. plantarum strain depend just on its ability to synthesise the bacteriocin, in situ, i.e., within the gut. The authors found that oral intake of bacteriocin-negative mutants of L. plantarum 423 failed to exclude L. monocytogenes from the gastrointestinal tract of mice, while administration of wild type L. plantarum could not inhibit gut colonisation by recombinant L. monocytogenes strains expressing the plantaricin immunity proteins.
Other animal studies indicate that oral intake of L. plantarum strains endowed with antimicrobial properties can contribute to (preserve or restore) gut microbiota balance and thus support future approaches to combat enteric infections and associated GIT inflammations. For instance, Choi et al. observed antiobesity effects in mice fed with a L. plantarum strain isolated from Kimchi (a traditional Korean fermented food), and ascribed these to its in vitro inhibitory activity against obesity-inducing bacteria (i.e., Enterobacter cloacae) [146]. Likewise, a plantaricin-like heat-stable antimicrobial was isolated and partially charac-terised by a food isolate L. plantarum and oral administration of such strain resulted in a healthy recovery of mice infected by S. Typhimurium [180]. Moreover, a L. plantarum strain, whose supernatants inhibited Helicobacter pylori growth in vitro, could attenuate H. pylori-induced gastric inflammation in rats [181].
L. plantarum with antimicrobial properties may have applications even in veterinary science and livestock industry, e.g., for the management of seafood farming, as studied by Chomwong and his co-workers [182]. These authors investigated the antipathogenic effect of L. plantarum SGLAB01, a strain isolated from the gut of shrimp. Dietary supplementation of such strain was found to enhance resistance to infections by Vibrio parahaemolyticus, which is responsible for acute hepatopancreatic necrosis, a disease implying troubles in intensive shrimp aquaculture [182]. This work also demonstrates the successful use of host-derived probiotics, i.e., microbes isolated from the digestive tract of the animal that is itself the target of the microbe-based therapeutic strategy. Such an approach reflects an increasing and up-to-date awareness that the animal and, specifically, human gut provides rich and as yet poorly explored reservoirs of potentially curative microorganisms, i.e., the so-called next-generation probiotics [183]. In very recent preclinical and clinical studies, (alterations of) some of these next-generation probiotics, identified as members of the human gut microbiota, have been associated with diverse diseases and hence have been indicated to own promising therapeutic potentials, e.g., for the treatment of diabetes and other endocrine/metabolism-related diseases [184].

Conclusions
As supported by the papers discussed in the present review, it is undoubtful that antimicrobial properties are of utmost relevance for the health-promoting effect of probiotics. Although several in vitro studies suggest a vast therapeutic potential for the antimicrobials from L. plantarum, to date, only a handful of investigations have explored the feasibility of their applications in vivo. Even if L. plantarum antimicrobials were found to be active against a broad spectrum of pathogens, just a few of them have been subsequently proven to prevent or ameliorate disease phenotypes in animal models. The numerical discrepancy between in vitro studies on L. plantarum antimicrobials and in vivo investigations might depend on the demanding organisation and management of the protocols required for experiments on animals, and then, in humans. A limiting factor may also be the isolation or the synthesis of adequate amounts of antimicrobial compounds to conduct in vivo investigations. In addition, comparative studies to assess the effectiveness of the isolated antimicrobial compound and the corresponding strain, in the form of viable cells, would be very useful and could indicate the way forward for probiotic research. Therapies based on probiotics, e.g., L. plantarum and probiotic-derived factors, have a high potential for the treatment of disorders, especially infections and gastrointestinal tract (GIT) diseases, and much research is still needed to define and consolidate it.