Proteins and Exopolysaccharides of Lactic Acid Bacteria

Abstract

Bacteriocin is peptide produced by bacteria to inhibit or kill other bacteria, exopolysaccharides are long chain polysaccharides composed of repeating sugar units. In the past decade, interest in bacteriocin and polysaccharides research from lactic acid bacteria have obtained great momentum due to their potential functions. This chapter will summarize current literature on the biological characteristics and functions of protein and exopolysaccharide produced by lactic acid bacteria, and discuss their potential applications.

3.1 Introduction

This chapter will summarize the current literature on the biological characteristics and functions of protein and exopolysaccharide produced by lactic acid bacteria and discuss their potential applications.

3.2 Bacteriocins

Bacteriocin is a ribosomally synthesized peptide produced by bacteria to inhibit or kill other bacteria in competition for nutrients or habitats. In the past decade, interest in bacteriocin research, especially lactic acid bacteria (LAB), has obtained great momentum due to its potential function both as a natural food preservative and therapeutic antibiotics (Cotter et al. 2005a; Van Heel et al. 2011). Bacteriocin has a number of advantages over conventional antibiotics. The spread of antibiotic-resistant bacteria poses a great threat to public health and is growing worse since the current progress in developing of novel antibiotics is limited (Brown and Wright 2016). There is a requirement for new antimicrobials that can be used as alternatives to conventional antibiotics.

3.2.1 The Classification and Chemical Structure of Bacteriocins

3.2.1.1 The Classification of Bacteriocins

The classification previously established by Klaenhammer grouped bacteriocins into four distinct classes with further subclasses. Before this classification, there were several classification approaches of bacteriocin. For example, early classification methods based on LAB bacteriocins’ characteristics classified the individual bacteriocins into eight groups, such as host range, host combinations, trypsin sensitivity, heat resistance, and the degree of cross-reactivity between various bacteriocins (Drider and Drider 2011a). After Klaenhammer, reclassifications or adaptations were performed by van Belkum and Stiles (2000) and Abriouel H et al. (2010) for the enterocin bacteriocins. In addition, the Cotter proposal has just two major categories: the lantibiotics (Class I) and the non-lantibiotic bacteriocins (Class II). As recently recommended by Kemperman et al. (Riley and Wertz 2002), Klaenhammer’s classes IIc and IV were canceled, since the two classes are unproven entities and they have novel posttranslationally modified cyclic peptides (Tagg et al. 1976; Riley and Wertz 2002). As assigned by Klaenhammer, Class IV was relegated to Class IIc. Heng NCK et al. have proposed a classification scheme, which is in line with Cotter et al. The Klaenhammer classes IIc and IV were eliminated. However, reverse to Cotter, Class III (large bacteriocins) is reserved and separated into Types IIIa (bacteriolytic) and IIIb (non-lytic), and the cyclic posttranslationally modified bacteriocins (Class IIc) are upgraded to Class IV (Heng and Tagg 2006). They believe that this classification scheme can be applied into most bacteriocins, if not all, disregarding the Gram status of the producer strain. And they acknowledged that this scheme will be continually completed as the development of our knowledge of microbial diversity (Ahmad et al. 2017).

The classification can be summarized as follows (Fig. 3.1):

Fig. 3.1

The classification of bacteriocin. (Ahmad et al. 2017)

Class I: Class I is named as lantibiotics, because these bacteriocins have one common characteristic that they all contain lanthionine, and nisin is the most typical bacteriocin in the Class I bacteriocins. The relative molecular mass of this category is less than 50,000, including more than 19–38 amino acid residues. Such bacteriocin-active site contains a lot of rare amino acid molecules, including lanthionine (Lan), β-methylmethionine (MeLan), dehydroalanine (Dha), and dehydrobutyrine (Dhb), and other noncoding amino acids.

Class I can be divided into four subcategories: type A is elongated, with positive charge and amphipathicity, which can form potential dependence holes in bacterial plasmids, like subtilin, ericin S, and ericin A. Type B is globular, without charge or negative charge, mainly through destroying enzymes’ function to work. Type B also includes other lantibiotics such as paenibacillin and sublancin 168. Type C is a two-component lantibiotics, which contains two polypeptide chains, and they are synergistic with each other to act as bacteriostatic agent, such as lichenicidin and haloduracin. Type D includes the unique cyclic peptide subtilosin A that contains a head-to-tail peptide bond as well as special sulfide bridges formed between cysteine groups and dehydrated amino acid residues (Abriouel et al. 2011).

Class II: defined as small-molecule (relative molecular weight < 10kDA), thermal-stable, and membranous-active polypeptide, non-lanthionine, such as tablet (pediocin pa-1). These were further subdivided into four subgroups:

Type IIa class bacteriocin is the largest subclass of Class II bacteriocin. All of the type IIa bacteriocins discovered so far have strong resistance activity to Liszt; therefore, it was also called class piece of Staphylococcus aureus, also known as anti-Lister active polypeptide.

Type IIb bacteriocin is a kind of bicomponent bacteriocin, which is formed by two peptide oligomers and requires the interaction of two peptides to produce a complete activity. Type IIb bacteriocin can be divided into two categories: one is the cooperative type (S type), and the other is the enhanced type (E type). Lactococcin Q, a new two-peptide bacteriocin from an LAB called L. lactis QU isolated from corn, is comprised of peptides Qa and Qb (Zendo et al. 2006). These two peptides have showed high comparability to two peptides Ga and Gb, which comprise the bacteriocin lactococcin G (Nissen-Meyer et al. 1992). The antimicrobial spectra of lactococcins G and Q are very specific, and they only have antimicrobial activity for strains derived from L. lactis.

Type IIc bacteriocin, including those who belong to neither type IIa class nor type IIb non-lantibiotics, is thought to be polypeptides containing the activity of sulfhydryl and the signaling peptide encoding mechanism. The bacterial strains of type IIc are diverse; therefore, the species diversity of this kind of bacteria is also more complex (Cotter et al. 2005a).

Type IId: linear, unmodified, non-pediocin-like bacteriocins. This type of bacteriocins is typically considered to be synthesized without an N-terminal leader sequence. Leaderless bacteriocins are often relatively small, usually 30–50 amino acids, and they do not like other bacteriocins go through posttranslational modifications. Therefore, they contain an N-terminal formylmethionine (Patricia et al. 2016). These characteristics make them relatively easy to obtain through synthetic methods, which opens up new possibilities for detailed bacteriocin research and design of new peptides with improved properties (Ovchinnikov et al. 2014).

In general, these peptides show broad-spectrum activity against Gram-positive bacteria. And occasionally, if the outer membrane of the bacteria is destroyed, they also can show some activity in Gram-negative bacteria (Towle and Vederas 2017).

Class III: bacteriocins in this class (formerly Class III bacteriocins) are large, heat-labile antimicrobial proteins. The relative molecular weight of Class III bacteriocin was found to be larger (usually greater than 30 kDa), and the bacteriocin is thermolabile; it is usually inactive in less than 30 min of 100 °C heating. Such bacteriocins can also be subdivided into two categories: one is lysozyme (lytic bacteriocins) that works through dissolving the cell, and the other is antibacterial protein (non-lytic bacteriocins) (Joerger and Klaenhammer 1986). Enterolysin A cleaves within the peptidoglycan of target cells between L-alanine and D-glutamic acid of the stem peptide and between L-lysine of the stem peptide and D-aspartic acid of the interpeptide bridge. Not only lytic bacteriocins but also some heat-labile, high-molecular-weight bacteriocins without a lytic mode of action have also been found, such as helveticin J from Lactobacillus helveticus 481, dysgalacticin from Streptococcus dysgalactiae subsp. equisimilis W2580, and streptococcin A-M57 (Vaughan et al. 1992; Eijsink et al. 1998). The action mode of dysgalacticin has been studied, and it has been determined that this bacteriocin interferes with glucose transport or metabolism by conjoining the phosphoenolpyruvate-dependent glucose or mannose phosphotransferase transport system (Eijsink et al. 1998).

Class IV: the fourth type of bacteriocin consists of an indeterminate constituent, a mixture of proteins, lipids, and carbohydrates. The existence of the fourth category is mainly supported by the observation that some bacteriocin activities (e.g., Lb. plantarum LPCO10) are obtained in cell-free supernatants, which are not only eliminated by protease treatment but also by glycolysis and lipolysis enzyme. They cannot only restrain Gram-positive bacteria but also have inhibition effects to Gram-negative bacteria and fungi. Therefore, antagonistic substances that do not meet or completely meet the bacterial protein definition are called bacteriocin-like substance. AS-48 (a circular bacteriocin) produced by Enterococcus faecalis is classified into this new class (Sánchezhidalgo et al. 2011a). AS-48 has attracted much attention due to its broad spectrum against Gram-positive bacteria including Clostridium tyrobutyricum, Enterococcus faecalis, Listeria monocytogenes, and some strains of Staphylococcus aureus (Towle and Vederas 2017).

Ahmad Cheikhyoussef, Wei Chen, and Hao Zhang et al. presented a review of the antimicrobial proteinaceous compounds produced by various Bifidobacterium strains, including bifidin (B. bifidum), biflong (B. longum), bifidocin B (B. bifidum NCFB 1454), bifilact Bb-12, bifilong Bb-46, etc., and summarized the identification, characterization, and possible applications of these compounds (Cheikhyoussef et al. 2008). Antimicrobial activity spectrum of the proteinaceous inhibitory compounds obtained from bifidobacteria can be summarized in Table 3.1.

Table 3.1 Antimicrobial activity spectrum of the proteinaceous inhibitory compounds obtained from bifidobacteria (Cheikhyoussef et al. 2008, 2009)

Four strains of Bifidobacterium showed different degrees of antagonistic action toward the indicator strain. The maximum degree of inhibition was from Bifidobacterium infantis (96%) and Bifidobacterium longum (92%). We found that substances or factors other than sole organic acids may contribute to the antimicrobial activity of the supernatants from the bifidobacteria studied (Cheikhyoussef et al. 2007).

In our laboratory, Bifidobacterium infants BCRC 14602 was found to produce a bacteriocin-like inhibitory substance (BLIS), which have a wide antimicrobial spectrum against Gram-positive and Gram-negative bacteria. Firstly, the purification of BLIS has been studied. BLIS was partially purified by a two-step purification scheme resulting in a specific activity of 31,605 AU/mg and a purification fold of 120. Then, a series of characteristics of BCRC 14602 BLIS were recognized. The size of BLIS is approximately 3.0 kDa based on tricine–SDS–PAGE, and it is sensitive to proteolytic enzymes but insensitive to catalase, lipase, and α-amylase. BLIS has high temperature stability and high pH stability in the range of 4–10. The adsorption of BLIS to production cells is strongly influenced by the pH of the broth culture that occurs 100% adsorption on dead cells between pH 6.0 and 7.0 (Cheikhyoussef et al. 2009).

In addition to BLIS, a new bacteriocin bifidin I from Bifidobacterium infantis BCRC 14602 was studied. Similar methods were used to get the characteristics of bifidin I. Unlike BLIS, the purification process of bifidin I was more complicated, in which three steps were needed. The purification scheme resulted in a purification fold of 1390 with a specific activity of 365,714 AU/mg and a yield of 25.6%. We can know that the mass of bifidin I is approximately 2879.7 Da through mass spectrometry analysis. Bifidin I was recovered by adsorption–desorption onto/from silicic acid (ADSA). The pH of the broth culture strongly influences the adsorption of the bifidin I to silicic acid (5%), and 100% adsorption occurs between pH 5.0 and 7.0; otherwise the adsorption ratio decreased from 67% to 45%. Only part of NH2-terminal amino acid sequence consisting eight amino acid residues was obtained: NH2-Lys-Tyr-Gly-Ser-Val-Pro-Leu-Gly. Curing experiments which resulted in Bif variants incapable of producing bifidin I but retaining immunity indicated that production of bifidin I is plasmid associated (Cheikhyoussef et al. 2010).

Another novel bacteriocin studied in our laboratory is sakacin P which is from Lactobacillus sakei and showed strong anti-activity against foodborne pathogens such as Listeria monocytogenes. And the genetic studies were carried out. In L. sakei, the structural gene (sppA) encoding sakacin P is commanded by a strict regulatory mechanism, and the amount of sakacin P secreted is limited. To obtain much more sakacin P, the sppA gene was transformed into E. coli and inducted with isopropyl-β-D-thiogalactopyranoside. Finally, the recombinant sakacin P was successfully expressed. Through agar diffusion method, the activity of sakacin P can be improved in a low temperature (Chen et al. 2012a).

Another bacteriocin-soluble NB-C1 fusion protein has been successfully obtained in the E. coli cell-free protein synthesis system. The NB-C1 gene is a novel potential Class IIa bacteriocin gene with the characteristic of YGNGVxC cluster. We built the expression vector pIVEX2.4d- GFP-NB-C1 both in the continuous exchange cell-free (CECF) systems and batch mode. The amount of soluble fusion protein attained from the CECF system was 2.2 mg/ml, which is around 3 times higher than that in the batch mode (0.73 mg/ml). The soluble fusion protein was purified via Ni–NTA affinity chromatography, resulting in a concentration of 0.26 mg/ml and a purity of 95%. The purified NB-C1 fusion protein was proved to have a strong inhibitory effect on the growth of L. monocytogenes (Chen et al. 2012b).

3.2.1.2 The Chemical Structure of Bacteriocins

The primary structures of lantibiotics and sactibiotics have been identified by some strategies and techniques. A major challenge we need to solve is to identify the topology of thioether bridges in these peptides (i.e., which amino acid residues are involved in which bridges). NMR spectroscopy, Edman degradation, and tandem MS have all been commonly used to characterize these bacteriocins, but they are antipathic with the posttranslational modification protein. Some chemical modifications, such as reduction and desulfurization, can assist the treated bacteriocins more compatible to these standard peptide analytical techniques (Perez et al. 2014).

A rapid system for screening novel bacteriocins from various sources has been developed in the early stages of screening and isolation of bacteriocin-producing strains. This system uses electrospray ionization liquid chromatography/mass spectrometry (ESI-LC/MS) combined with principal component analysis (PCA) to perform molecular weight analysis on the antimicrobial activity spectrum of each bacteriocin produced by the LAB strain. This allows us to save time, energy, and money and quickly track the discovery of more novel bacteriocins.

Class I: Class I bacteriocins or lantibiotics (lanthionine-containing antibiotics) are small peptides (<5 kDa) that contain abnormal posttranslational modifications such as lanthionine or 3-methyllanthionine. These residues are capable of forming covalent bonds between amino acids, forming an internal ring structure and giving unique structural features (Chen et al. 2012b; Cotter et al. 2005b). Bacteriocins are usually synthesized as non-bioactive prepeptides, including N-terminal precursor peptides connected to C-terminal peptides (And and Hoover 2003; Klaenhammer 1993; Cotter et al. 2005b).

Class II: Bacteriocin type IIa class typically contains 37–48 amino acid residues, and the N-terminal consists of conservative YGNGVXaaC groups, which is usually thought to be membrane receptor-binding protein recognition sequence. As more and more new type IIa class of bacteriocin was found, the N-terminal groups can also be shown as YGNGVXaaCXaa XaaXaaCXaaV (K/N) (N/D) (W/R/K) Xaa (G/A/S) (A/N) (the content in the parentheses is conservative residue, meaning the residue can be replaced; Xaa means high change frequency residues).

Another important feature of type IIa class bacteriocin is that N-terminal conservative contains at least two cysteines that formed disulfide bond with each other. The existence of the disulfide bond allows a hydrophilic oleophilic structure in the N-terminal, which is necessary for antimicrobial activity. In general, peptide fragment that contains two disulfide bonds has a wider bacteriostatic spectrum than peptide fragment that contains only one disulfide bond (Eijsink et al. 1998). Compared with the N-terminal conservative property, the similarity of C-terminal sequence is only 34–80%, and an α-helix is formed. The spiral structure served as cross-membrane component when holes are formed on the sensitive bacteria cell membranes.

Enterococcus faecium NKR-5-3, an LAB isolated from the Thai fermented fish (Himeno et al. 2012; Naoki et al. 2012), can produce a kind of bacteriocin named enterocin NKR-5-3C (Ent53C). Ent53C is a typical Class IIa bacteriocin, which displayed extremely strong antimicrobial activity (in nanomolar range) against Listeria spp. and other Gram-positive species (Drider et al. 2006). The number of disulfide bridges in Class IIa bacteriocins directly relates with the intensity of their antimicrobial activity and stability (Drider et al. 2006; Richard et al. 2006).

Class IIb bacteriocins are two-peptide bacteriocins that require the cooperation of two peptides to be fully activated (Oppegård et al. 2007; Nissenmeyer et al. 2010). The novel two-peptide bacteriocin Lactococcus Q, isolated from Lactococcus lactis QU4, a lactis isolated from corn, consists of two peptides: Qa and Qb. Structural analysis of the Lactococcus Q peptide showed it has the same position of α-helical structure as Lactococcus G. This suggested that these bacteriocins have similar modes of action (Zendo et al. 2006).

The cyclic bacteriocins (Class IIc) are characterized by their unique structural feature of a head-to-tail cyclization of their backbones (Maqueda et al. 2008; van Belkum et al. 2011). The circular bacteriocins compared to their linear counterparts usually have superior structural stability, greater thermal stress resistance, and larger stability against proteolytic digestion (Conlan et al. 2011; Craik et al. 2010), which are mainly determined by the nature of their structure. The N- and C-termini of Class IIc bacteriocins are covalently linked giving the peptide an extremely stable structure (Maqueda et al. 2008; van Belkum et al. 2011).

There are some similar structures in Class IIc bacteriocin, such as leader peptide with GG sequence, ABC transport protein, and some related immune protein. Some Class IIc bacteriocins have no identifiable N-terminal signal peptide, such as enterocin Q from excrement Enterococcus, aureocin A53 produced by Streptomyces aureus, and BHT-B produced by S. rattus. Bacteriocin gassercin A isolated from LA39 strains of lactobacillus in human baby feces, which contains a rare loop structure, has inhibitory activity on many food pathogenic bacteria, including Listeria, waxy Bacillus, and Staphylococcus aureus.

Both lactocyclicin Q (LycQ) and leucocyclicin Q (LcyQ) are circular bacteriocins composed of 61 amino acid residues (Sawa et al. 2009; Masuda et al. 2011), and their precursor peptides contain a leader sequence of two amino acid residues in which cyclization occurs between L3 and W63 (Masuda et al. 2011). The prediction of the secondary structure of the two bacteriocins reveals that the four identical α-helices, each with subtle amphiphilic properties, are thought to play an important role in their antimicrobial action.

On the other hand, Class IId bacteriocins contain the remaining bacteriocins, which bind in the wrong combination or bind as one-peptide non-pediocin linear groups (Cotter et al. 2005b). Sec-dependent bacteriocins (Cintas et al. 2000) and leaderless bacteriocins (Fujita et al. 2007) belong to this class. Lacticins Q and Z only have differences in three amino acid residues at the positions of 10, 33, and 44. In addition, they both have a formylated methionine at the starting residue. They are both 53 amino acid highly cationic peptides, exhibiting extremely strong antimicrobial activity (at nanomolar concentrations) and high stability against various stresses (Fujita et al. 2007; Iwatani et al. 2007). The antibacterial activity of lacticin Q is mainly attributed to the two amphiphilic helices, which play a major role in it (Yoneyama et al. 2009a, b, 2011). Since the homology of these two leaderless bacteriocins is very high and their activity spectra are comparable, it was inferred that they have the same mode of antimicrobial action (Iwatani et al. 2007).

Class III: Class III bacteriocins are heat-labile and usually consist of different domains. For example, on the basis of sequence analysis, enterolysin A comprises an N-terminal endopeptidase domain and a C-terminal substrate recognition domain, which is similar to zoocin A (Lai et al. 2002; Nilsen et al. 2003). The zif gene encoding the immune protein, which is close to zooA, adds L-alanine into the cross bridges of peptidoglycan, reducing the ability of animal protein A to degrade the polysaccharide layer (O’Rourke et al. 2009).

Class IV: circular and linear leaderless bacteriocins whose backbone structure assumes a saposin-like fold or a related α-helical bundle in the context of their possible mechanisms and their interactions with lipids.

3.2.2 The Genetics and Regulation of Bacteriocins

Lactic acid bacteria will secrete some bacteriocins in the logarithmic phase and transport bacteriocin to the external medium by the cell membrane permeability, but some types of bacteriocin will stay in cells under certain conditions. In general, the production and synthesis of bacteriocin is synchronized with growth, and the yield is closely related to the number of producing bacteria. The optimization of culture conditions can effectively increase the production of bacteriocin, such as the supplement of sugar, vitamin, nitrogen source, or stimulating factors such as pH, temperature to their best range, etc. (Abbasiliasi et al. 2011; Espeche et al. 2014). The pH of medium is one of the important factors which influences the production of bacteriocin, but the requirement of pH of different kinds of bacteriocin are not identical (Arauz et al. 2012).

According to Drider et al. (2006), at least four key genes are contributed to the secretion and production of bacteriocins. In particular, they are (i) the structural bacteriocin gene, encoding a prebacteriocin; (ii) an immune gene that protects bacteriocin producers from their own bacteriocins; (iii) a gene encoding an ABC transporter necessary for secretion; and (iv) a gene encoding an accessory protein of unknown function (Sabo et al. 2014).

Bacteriocin biosynthesis was regulated by the regulation genes; these genes are generally located on the chromosome or plasmid (Nes et al. 1996). Sulfide antibiotic bacteriocin and Class II bacteriocin operon are mostly located on chromosome, such as bacteriocin plantaricins EF and NC8 produced by plant lactobacillus 8p-A3. These two kinds of bacteriocin production regulation genes are located on chromosome, and the gene fragment size is about 20 kb. Besides these gene cluster located on chromosome, some gene clusters regulating bacteriocin production are located on plasmid. Mandal reported a new bacteriocin pediocin NV5 produced by lactic acid piece Staphylococcus aureus LAB5 whose gene cluster in a size of about 5 KB was proved to be located on plasmid by eliminated plasmid experiments, etc. (Mandal et al. 2011).

Bacteriocins’ expression needs at least two genes, structure gene and immune genes. The former codes express precursor peptide, while the latter encode some immune protein to protect bacteria from the attack on its own. The bacteria have no specific output protein for bacteriocin. The bacteriocin outputs the cells through the normal output pathway, such as bacteriocin 31 produced by Enterococcus YI717, whose structure genes and immune genes were BacA and BacB, respectively. These gene clusters were on removable plasmid pYI17, and the size is about 57.5 KB (Tomita et al. 1996).

Bacterial structural gene encoding a bacteriocin containing N-terminal leader sequence (referred to as double-glycine leader sequence) of the preform element, which appears to function (a) preventing the bacteriocin from being biologically active when detained inside the producer and (b) providing the identification signal used in the transportation system. The length of the bio-glycine leader ranges from 14 residues to about 30 residues (Klaenhammer 1993).

Common elements found in the double-glycine leader sequence include two glycine residues at the C-terminus of the cleavage site, conserved hydrophobic and hydrophilic residues separated by distances determined in the reserved residues. In addition, the minimum length of the double-glycine leader sequence appears to be 14 amino acids. Of the consensus residues of the leader sequence, only the glycine residue at position 2 is completely conserved. The mature bacteriocins identified so far range in size from less than 30 residues to more than 100 residues (Nes et al. 1996).

Class I: Genes involved in lantibiotic synthesis are generally settled in clusters, which can be organized on a transposon (nisin), on a plasmid (epidermin), or on the chromosome (subtilin) (Drider and Drider 2011b). As for the biosynthesis of lantibiotic, it usually includes the following steps. Firstly, the process starts with translation of a leader and modifiable propeptide moiety, and then the prepeptide undergoes modification. Next, the prepeptide is translocated through the cytoplasmic membrane, and the leader peptide is cleaved proteolytically. Genes coding immunity proteins are normally sited in a cluster around the bacteriocin structural gene. The conserved sequences of PTMs and core peptides facilitate high-level analysis by analyzing the genomic data. This can help centralize screening efforts to discover new molecules using different alternative methods (Montalbánlópez et al. 2012; Hegemann et al. 2015; Rutledge and Challis 2015).

Class II: Like lantibiotics, the second class of bacteriocins are synthesized as an inactive prepeptide that usually contain a typical double-glycine proteolytic processing position (Ennahar et al. 2000; Drider et al. 2006). However, some type II bacteriocins are typically Sec-dependent N-terminal signal sequence synthesized and secreted by the general secretory pathway.

Unlike lantibiotics, Class II bacteriocins do not experience extensive posttranslational modifications. Specific enzymes cleave off the leader peptide concomitant when the prepeptide completed translation, and it’s translocated to the extracellular space by a specific ABC transporter that occasionally requires an auxiliary protein (Drider et al. 2006; Ennahar et al. 2000). However, there are a rising number of newly reported bacteriocins that lack leader sequences; these are of interest as they take effect directly after translation (Cintas et al. 2000; Fujita et al. 2007; Masuda et al. 2012).

L. lactis QU 5 and L. lactis QU 14 can, respectively, produce lacticin Q and its homologue lacticin Z, which, respectively, correspond to the biosynthetic gene clusters lnqBCDEF and lnzBCDEF that are related to bacteriocins’ secretion and self-immunity. In L. lactis NZ9000, overexpression of lnqQ resulted in the intracellular accumulation of lacticin Q (Iwatani et al. 2012). Through further analysis of the function of the nqBCDEF gene product, it was found that this gene cluster strictly controls the secretion of lacticin Q into the extracellular space, whereas the control of the self-immunity system is relatively weak (Perez et al. 2014). The ABC transporter-type immunity LnqEF is sufficient to confer minimum immunity, whereas LnqBCD are thought to be accessory proteins that provide activity for LnqEF (Iwatani et al. 2013).

Class III: Both structural and immunity genes of megacins A-216 and A-19213 are plasmid encoded. Megacin A-216 contains 293 amino acid residues and shows a native molecular weight of 66 kDa (Kiss et al. 2008). The biologically active portion includes three parts called g, a, and b chains (32,855, 21,018, and 11,855 Da), corresponding to the full-length protein and two decomposition products, respectively (Kiss et al. 2008).

The genetic determinant of A-216 is encoded by the 5494-bp plasmid region. This plasmid region also contains the structural gene of A-216, which controls a 293 amino acid protein with sequence similarity to proteins with phospholipase A2 activity. The ORF gene near the megA gene encodes a 91 amino acid protein responsible for immunity of producer strain against megA-216. At least two other genes, including ORF 73 and a gene encoding a 188 amino acid protein, are required for the induction of megacin A-216 expression. Sequences encoding the 188 amino acid protein are highly similar to the RNA polymerase σ factor (Abriouel et al. 2011).

Class IV: AS-48 is a typical kind of Class IV bacteriocin. The AS-48-related gene clusters consist of ten major genes, including the structural gene AS-48A; AS-48B relates to a putative cyclase, AS-48C for a DUF95 protein related to immunity and production (Mu et al. 2014), AS-48C1D represents putative ABC transporter associated production, AS-48D1 is associated with typical immune protein production, and AS-48EFGH represents production of immune-related additional ABC transporters (Maqueda et al. 2008).The expression of AS-48 requires the expression of a large transcript that involves AS-48ABC which is posttranscriptionally processed, a second transcript including AS-48C1DD1EFGH, and a third transcript with a weak promoter that transcribes AS-48D1EFGH (Sánchezhidalgo et al. 2011a; Cebrián et al. 2014).

3.2.3 The Applications of Bacteriocins Derived from LAB

An overview of the applications of LAB and their bacteriocins is shown in Fig. 3.2, emphasizing the role of LAB bacteriocins in medicines and other industries.

Fig. 3.2

The applications of LAB and their bacteriocins. (Liong 2015)

• 1. The Application in Food

As for food, it’s not enough to just consider about food’s edibility, which means the shelf life. There are other things that are more important we need to pay attention to, such as nutrition, organoleptic, and, most importantly, consumer acceptability. The currently existing preservatives are not entirely satisfactory to the consumers. The foods which are alleged to be “free of additive” are often more popular to consumers. Bacteriocins as biopreservatives are more natural than chemosynthetic preservatives. Lactic acid bacteria bacteriocin is a kind of polypeptide, which can be digested by proteases secreted by digestive tract in the process of the body’s digestion and will not result in any residues. On the other hand, bacteriocin can inhibit most corruption bacteria and foodborne pathogenic bacteria. In addition, bacteriocins are nontoxic, odorless, and colorless, which makes them likely to be an ideal protectant. The most widely used bacteriocin is lactic Streptococci bacteriocin. Nisin has been widely used in food production, and it can effectively inhibit the growth of harmful bacteria, improve the processing quality of the food, and prolong food storage time. Espitia et al. evaluated physical–mechanical and antimicrobial properties of nanocomposite film with pediocin and ZnO nanoparticles against Staphylococcus aureus and Listeria 7 monocytogenes (Espitia et al. 2013). Gassericin A, as food preservative from Lactobacillus gasseri LA39, was described as stable at 4 °C for 3 months, at 37 °C for 2 months, at 60 °C for 5 h, and at 100 °C for 30 min (Ahmad et al. 2017).

As for LAB bacteriocins, the bacteriocins from food-grade lactic acid bacteria (LAB) were qualified as an ideal food biopreservative predominantly because:

• (i) It has been proven nontoxic to humans.

• (ii) Does not alter the nutrients contained in the food.

• (iii) Effective at low concentration.

• (iv) It is still active under refrigeration conditions (Ahmad et al. 2017).

The preservative effect of bacteriocins can be used in the following aspects.

• (1) Meat are usually perishable and susceptible to microbial contamination. Nisin can effectively control the growth of Clostridium botulinum, and nisin is acidic, so it can reduce the pH of the surrounding medium to reduce the residue of nitrite content and the formation of nitrosamines.

• (2) In ancient times, LAB has begun to be applied to the fermentation food, which was mainly related to the beneficial effect of LAB on nutrition, color, flavor, and shelf life (Richard et al. 2006). Nisin which is added 30–50 mg/kg in dairy products can extend the shelf life and usually doubles the products’ shelf life under 35 °C. Adding 80~100 mg/kg nisin in canned condensed milk can reduce the time of sterilization by 10 min. 20 mg/kg of nisin added in UHT milk can completely inhibit the growth of spore bacteria in sterilized milk (Perez et al. 2014). Yogurt post-acidification process could be delayed for 3 days if 40 IU/mL of nisin is added, and the number of living bacterium stays above 107/mL, and the yogurt sensory quality is good. Nisin is the earliest preservatives used in the cheese. Using the mixture of Streptococcus acid-resistant bacteria and nisin as cheese starter can make cheese merit factor above 90% compared to the conventional method which is only 41% (Maisnier-Patin et al. 1992).

Until now, only nisin and pediocin PA-1 have been commercialized as food additives. However, other LAB bacteriocins like enterocin AS-48 (Sánchezhidalgo et al. 2011b) and lacticin 3147 (Suda et al. 2012) also showed promising perspectives to be used as biopreservatives in food. AS-48 is stable and soluble at wide range of pH and temperatures. This feature makes them able to be widely used in the food industry (Sánchezhidalgo et al. 2011b).

• 2. The Application in Medical Treatment

In the past decade, with the frequent occurrence of clinical drug resistance in the world, peoples’ concerns to multidrug resistance have increased (Carlet et al. 2014). Particularly in Third World countries, most of the effective drugs have now turn out to be almost useless against a large number of the organisms. Microorganisms that are involved in serious infections have developed resistance to one or more than one common broad-spectrum antibiotics (Alanis 2005). The problem is caused not only by the microorganisms that resist effective antibiotics in different ways but also by the increase overprescribing and inadequate medications. Compared with other antibacterial compounds, bacteriocins are target-specific, broad-spectrum, and more effective antagonists. Bacteriocins can be used as alternative therapeutics against resistant pathogens, and there are several promising applications for controlling potential health risk factors (Papagianni 2003). The applications of bacteriocins in medical treatment can be summarized as follows:

• (1) Skin and soft tissue infections are caused by Staphylococcus aureus, Propionibacterium acnes, Staphylococcus epidermidis, B. cereus, B. subtilis, and Listeria monocytogenes. Nisin and some other bacteriocins have been described that they can effectively treat these skin diseases (Manosroi et al. 2010; Bowe et al. 2006; Kang et al. 2009; Izquierdo et al. 2009). For instance, the bacteriocin subpeptin JM4B produced by B. subtilis JM4 is very active against Staphylococcus aureus, Pseudomonas aeruginosa, Salmonella spp., and Enterococcus faecalis (Wu et al. 2005).

• (2) The mouth is a necessary place which pathogens in food must pass. Thus some microorganisms will live in the oral cavity. Besides, oral cavity special environment easily leads to tooth decay, so oral rinse containing bacteriocins offers new possibilities for the development of oral rinse products. Porphyromonas gingivalis, Prevotella intermedia, and Aggregatibacter actinomycetemcomitans are regarded as the principal periodontal pathogens. Many bacteriocins, such as subtilosin A against Porphyromonas gingivalis (MIC = 3.125–6.25 μg/mL) (Shelburne et al. 2007; Siegers and Entian 1995) and lacticin 3147, BLIS K12, and a bacteriocin of 114 kDa from Lactobacillus paracasei strain HL32 against Porphyromonas gingivalis, have been recorded to inhibit the growth of major oral pathogens (Hammami et al. 2011; Howell et al. 1993; Tagg 2004).

• (3) The intestinal flora of humans and animals is closely related to LAB. Recent studies of LAB have demonstrated their efficacy in treating disorders of the intestinal flora. Lacticin L3147 has been shown to inhibit the growth of Bacillus difficile at low concentrations without affecting the normal human flora (Rea et al. 2007). This lantibiotic could be used as an effective oral therapeutic for overcoming C. difficile-induced diarrhea and gastric acid. Helicobacter pylori and antral gastritis are closely related to the occurrence of duodenal ulcer and gastric ulcer, and Nisin can inhibit Helicobacter pylori.

• (4) In hospital settings, the main disease-causing pathogens are Staphylococcus aureus, pneumococci, enterococci, Klebsiella pneumoniae, A. baumannii, Citrobacter freundii, E. coli, and Proteus spp. (Ghodhbane et al. 2015; Michalet et al. 2007). Nisin and lacticin 3147 have significant inhibitory effect on various pathogens in the liver, kidneys, and spleen. But it’s difficult to treat Staphylococcus aureus, enterococci, and pneumococci strains which are resistant to six or more antibiotics. Bacteriocins such as lacticin 3147 and nisin A have been found effective against MRSA and VRE (Piper et al. 2009).

Currently bacteriocin for the treatment of bacterial infection disease is usually an auxiliary treatment of bacterial intervention therapy trials; so far, no pure bacteriocin can be directly used as drugs in clinical therapy.

In addition to the above common application of the bacteriocins, bacteriocins can also be used for the following potential purposes. G. vaginalis, Mycoplasma hominis, Prevotella bivia, and Mobiluncus curtisii can lead to female vaginal diseases, but the antibiotics such as clindamycin and metronidazole we conventionally use can kill beneficial flora of the vagina. However, certain bacteriocins can help women against these bacteria, for example, bacteriocin lactocin 160 against G. vaginalis and subtilin A against B. amyloliquefaciens (Ahmad et al. 2017).

In addition to antimicrobial activity, bacteriocins also have the effect of killing spermatozoa. For example, fermenticin HV6b isolated from Lactobacillus fermentum HV6b (MTCC10770) shows strong function of sperm fixation and spermicidal activity. This characteristic makes it an attractive proposition for formulating antibacterial vaginosis and contraceptive products (Kaur et al. 2013).

Bacteriocins can not only make effects on human and animals, they can also be used in treating plant bacterial diseases. For example, ericin S is active against C. michiganensis, the causative agent of tomato bacterial canker. Purified ericin or its producer strain could be developed as a bioprotectant on tomato plants (Abriouel et al. 2011). In addition to the function of bacteriocins, Abriouel H puts forward that bacteriocins may have a potential to inhibit biofilm formation, thus can be used in pipeline cleaning technologies and consequently reduce biocorrosion, which may be beneficial for the environment.

In the future, we believe that as the more research unfolds, the application of LAB bacteriocins in health control will be more prosperous. Several bioengineered strategies will also be engaged to enhance the commercial potential of LAB and their metabolites in medical and health (O’ Shea et al. 2013).

3.3 Exopolysaccharide (EPS)

Exopolysaccharides (EPSs) are long-chain polysaccharides composed of repeating sugar units in different ratios which mainly include glucose, galactose, rhamnose, and so on (De Vuyst and Degeest 1999a). Instead of permanently attaching to surface, these polysaccharides are secreted into their surrounding environments. This feature differentiates exopolysaccharides from capsular polysaccharides which always attach to the surface of the cell (Laws et al. 2001).

The most valuable application of EPSs from lactic acid bacteria (LAB) was the rheology and texture enhancement of yogurt products. Recently a higher demand for consumption of smooth and creamy products was evoked. It is usually catered by the increase of fat, sugars, and stabilizers. Despite of the cost, the consumers’ demand for products with low fat or sugar and low levels of additives makes EPSs a viable alternative for yogurt products (Jolly et al. 2002). Although LAB EPSs have no own taste and flavor, the time the fermented milk product spends in the mouth was increased by the LAB EPSs, so an elevated perception of the human sense of taste is imparted (Duboc and Mollet 2001). In addition, EPSs will remain longer in the gastrointestinal tract (GIT), by which the colonization of probiotic bacteria is enhanced (German et al. 1999). Moreover, the LAB EPSs have been investigated to have the potential of antitumor effects Kitazawa et al. 1998), immunostimulatory activities (Hosono et al. 1997; Chabot et al. 2001; Shao et al. 2014), antihypertensive effects (Ai et al. 2008a, b), and functions to lower blood cholesterol (Nakajima et al. 1992a).

3.3.1 The Classification and Chemical Structure of EPS

For the EPSs from LABs, two categories can be summarized. The first category is homopolysaccharide which is composed of four different subgroups (i.e., α-D-glucans, β-D-glucans, fructans, and others). α-D-glucans or dextrans consist mostly of α-1,6-linked glucose monomers with various branched chains at position 3 but sometimes at positions 2 and 4 at a relatively lower probability, such as the EPSs from Leuconostoc mesenteroides subsp. mesenteroides and dextranicum. However, the EPSs form Leuc. mesenteroides (i.e., alternan), and Streptococcus mutans and sobrinus (i.e., mutans) are composed of both α-1,3- and α-1,6 linkages. β-D-Glucans consist of β-1,3-linked glucose monomers with β-1,2-branched chains, such as the EPSs from Pediococcus spp. and Streptococcus spp. Fructans consist mainly of β-2,6-linked D-fructoses. For example, the S. salivarius EPSs (i.e., levan) have this structure and some β-2,1 side chains at the O₁ site. Similarly, other EPSs in the first category all consist of structurally identical repetitive elements with various glycosidic links, such as polygalactan.

The second category is heteropolysaccharides. This kind of EPSs are usually produced by LAB strains which have mesophilic and thermophilic life history, such as the mesophilic Lactococcus lactis subsp. lactis and cremoris, Lactobacillus casei, Lactobacillus rhamnosus and thermophilic Lactobacillus acidophilus, Lactobacillus delbrueckii subsp. bulgaricus, and Streptococcus thermophiles.

Interest has been evoked to the latter group of EPSs, for the reason that they play an important part in the mouthfeel of yogurt products. For example, as one of the crucial quality aspects, yogurt texture will become more creamy and smooth even though only tiny amounts of the EPSs are secreted. Moreover, LAB EPSs have its unique promising technical potential for novel fermented milk product development such as low-milk-solid, low-calorie, low-milk-fat, and creamier yogurt products. In addition, as nondigestible food fraction (Gibson and Roberfroid 1995), some polysaccharides may contribute to human health because of their antitumoral (Oda et al. 1983), antiulcer (Nagaoka et al. 1994), immunomodulating (Kitazawa et al. 1993), or cholesterol-lowering (Nakajima et al. 1992a) activity. Therefore, with both health and economic benefits, LAB EPSs have the potential for developing and exploiting functional food ingredients.

The chemical composition of LAB EPSs has long been controversial. (Sundman et al. (1953)) and (Nilsson and Nilsson (1958)) firstly investigated the nature of slime materials form LAB, which was found to be protein-like material. After this, however, some researchers speculated that glycoprotein or carbohydrate–protein complex are produced when milk fermentation would confer the ropy characteristics to yogurt (Macura and Townsley 1984; Garcia-Garibay and Marshall 2008; Toba et al. 1991), while other researchers isolated exo-polymer material which is enriched in carbohydrate material after further purification (Nakajima et al. 1990; Cerning et al. 1986, 1988,1992, 1994; Kojic et al. 1992; Norris et al. 1954; Wang et al. 1963; Bouzar et al. 1996; Cerning 1995). Finally, an agreement is achieved that the LAB exo-polymers are polysaccharides made up of repetitive elements and that many different categories are secreted Cerning 1995). However, the compositions of monosaccharide residues are very similar in appearance. Frequently D-galactose, D-glucose, and L-rhamnose are present but in different ratios (Ariga et al. 1992; Gamar et al. 1997; Doco et al. 1990; Nakajima et al. 1992b; Gruter et al. 1992, 1993; Yamamoto et al. 1994, 1995; Marshall and Cowie 1995; Robijn et al. 1995a, b, 1996a, b; Grobben et al. 1995, 1996, 1997; Mozzi et al. 1996; Stingele et al. 1996; Bubb et al. 1997; Lemoine et al. 1997; De Vuyst et al. 1998). Nevertheless, some special cases exist, for example, EPSs from Lb. acidophilus LMG 9433 (Robijn et al. 1996b), Lb. helveticus TYl-2 (Yamamoto et al. 1994) and NCDO 766 (Robijn et al. 1995b), and Lb. rhamnosus C83 (Gamar et al. 1997), S. thermophilus Sfi20 (Doco et al. 1990; Stingele et al. 1996), Sfi32 (Lemoine et al. 1997), LY03, BTC, and 480 (De Vuyst et al. 1998) lack rhamnose, EPSs from Lb. paracasei 34–1 contains only galactose (Robijn et al. 1996a), EPSs from S. thermophilus OR-901 contains only galactose and rhamnose (Ariga et al. 1992; Bubb et al. 1997), and EPSs from Lb. sake 0–1 only consists of glucose and rhamnose (Robijn et al. 1995a). Other residues, such as N-acetyl-aminosugars and sn-glycerol-3-phosphate, as well as the phosphate and acetyl groups could also be present (Nakajima et al. 1990, 1992b; Doco et al. 1990; Yamamoto et al. 1994, 1995; Marshall and Cowie 1995; Robijn et al. 1995a).

The different LAB EPS compositions found in the past may be to some extent because of the limits in isolation and purification techniques and consequently their poor performance, especially in the situation where complex media are selected, and the possibility that more than one kind of polysaccharides could be secreted by one strain. Grobben et al. once reported that different compositions of exopolysaccharides were characterized from the same Lb. delbrueckii subsp. bulgaricus strain, where different isolation and purification techniques were used under different fermentation conditions (Grobben et al. 1995, 1996, 1997). Marshall et al. isolated two different EPSs from the same L. lactis subsp. cremoris strain having different monosaccharide compositions and molecular masses (Marshall and Cowie 1995). Isolation of high- and low-molecular-mass EPS fractions was also performed form the same Lb. delbrueckii subsp. bulgaricus (Grobben et al. 1997) and S. thermophilus strain, but the monomeric composition of these EPSs does not differ. Ai et al. isolated the crude exopolysaccharides (LCP) from skim milk fermented by Lb. casei LC2W and fractionated these polysaccharides into three fractions (LCP1, LCP2, and LCP3) with the ratios of 35.74%, 12.61%, and 33.34% (w/w), respectively (Fig. 3.3) (Ai et al. 2008a, b).

Fig. 3.3

(a) High-performance size exclusion chromatograms and (b) Fourier transform infrared spectra of EPS fractions LCP1, LCP2, and LCP3. (Ai et al. 2008a, b)

LCP1 and LCP2 were composed of glucose, rhamnose, and galactose in molar ratios of 4.0:1.9:1.0 and 5.0:16.2:1.0, respectively. Besides these three sugars, LCP3 contained a given amount of mannose, the monosaccharide molar ratio of Glu:Rha:Gal:Man is 1.0:2.1:3.8:3.5 Ai et al. 2008a, b). Shao et al. separated two different kinds of EPS fractions from the Lb. rhamnosus KF5 in fermented skim milk by removing proteins, ethanol precipitation, anion exchange, and gel permeation chromatography. Fraction S1 was composed of glucose, arabinose, glucosamine, galactosamine, and galactose in an approximate molar ratio of 2.03:1.29:1.25:0.72:0.61, while fraction S2 contained rhamnose, glucose, and galactose in a molar ratio of approximately 1.73:1.47:1.00 (Fig. 3.4a) (Shao et al. 2014).

Fig. 3.4

(a) HPAEC profile of monosaccharide composition of EPS fractions (S1 and S2) from Lb. rhamnosus KF5 (Shao et al. 2014) and (b) structure of exopolysaccharides EPS-1 and EPS-2 from Lb. johnsonii FI9785 (Dertli et al. 2013)

Dertli et al. isolated and purified exopolysaccharides from Lactobacillus johnsonii FI9785 which has previously been shown to act as a competitive exclusion agent to control Clostridium perfringens in poultry. Structural analysis by NMR spectroscopy revealed that Lb. johnsonii FI9785 can produce two types of exopolysaccharide, EPS-1 and EPS-2 (Dertli et al. 2013) (Fig. 3.4b).

LAB EPSs have molecular masses approximately ranging from 4.0 × 10⁴ to 6.0 × 10⁶ although EPS molecular mass is one of the factors affecting its processing characteristics (Cerning et al. 1992; Vandenberg et al. 1995). The three-dimensional conformation of a polysaccharide in solution is also closely correlated to its physical and rheological properties. In solution polysaccharide chains will experience some kind of topological rearrangements for the sake of transforming into a well-organized conformation in favor of intermolecular interactions and associations. So other factors, such as the intermolecular association abilities of polysaccharide molecules, could be important for a comprehensive understanding of the solution behavior. The secondary and tertiary conformation of EPSs is strongly dependent on the primary structure. Relatively small alteration in primary sequence might result in a tremendous effect on the three-dimensional conformation and corresponding processing characteristics of a EPS molecular mass. Doco et al. firstly determined the structure of the repeating unit of S. thermophilus heteropolysaccharide (Doco et al. 1990). Other structures of the repeating unit in branched LAB EPSs have been confirmed recently via methylation, acetolysis, periodate oxidation, acid hydrolysis, enzymatic digestion, Smith degradation, 1D,2D ¹H-NMR spectroscopy techniques, etc. (Doco et al. 1990; Nakajima et al. 1992b; Gruter et al. 1992, 1993; Robijn et al. 1995a, b, 1996a, b; Stingele et al. 1996; Bubb et al. 1997; Lemoine et al. 1997; Dertli et al. 2013).

The size of repeating unit may range from a disaccharide to a heptasaccharide. Few common profiles are further shown to people, which introduce a question upon the relationship between EPS structures and the their texturizing properties. Exploring this relationship and exploiting the techniques modifying biopolymers that influence the properties of the native polysaccharides would be interesting. Specific enzymes can be applied to tailor the chemical structure of LAB EPSs and their functional properties. Integrating the biochemical and molecular biological researches on the LAB EPSs, future engineering of polysaccharide and perhaps oligosaccharide could be of great value given the explosive tendency of current functional food market.

3.3.2 The Biosynthesis and Genetics of EPS in LAB

The biosynthesis of a fraction of homopolysaccharides, such as dextrans, alternan, mutans, and levans, takes place outside the cell and thus requires specific substrates. For dextran and levan, highly specific glycosyl transferase enzymes dextran and levan sucrase are involved in polymerization reaction, where the energy required for polymerization is provided by hydrolysis of sucrose. Such kind of EPSs can be produced either using bacterial cells or cell-free systems (Cerning 1990).

On the other hand, the synthesis of heteropolysaccharides is in another way where polymerization of the repeating unit precursors takes place in the microbial cytoplasm (Cerning 1990, 1995), and several enzymes and proteins which are not necessarily unique to EPS formation are involved in the biosynthesis and secretion process. In the heteropolysaccharide biosynthesis process, polymerization and interconversions are both crucial steps, where interconversion further includes epimerization, decarboxylation, and dehydrogenation. Correspondingly, sugar activation and modification enzymes inevitably play an important part in this process because of their contributions in the sugar nucleotide (i.e., building blocks) formation.

For example, the UDP-glucose pyrophosphorylase was found to be correlated with EPS production in a ropy S. thermophilus strain but not in the non-ropy strain. Nevertheless, Escalante et al. found that it appears to have nothing to do with the EPS biosynthesis in any S. thermophilus strain examined (Escalante et al. 1998). Meanwhile, in Lb. delbrueckii subsp. bulgaricus NCFB 2772, the UDP-glucose pyrophosphorylase (responsible for the biosynthesis of UDP-glucose and UDP-galactose) was much more active in glucose-grown cultures than that in fructose-grown cultures. However, in the fructose-grown cultures, no enzyme activities were detected which led to dTDP-rhamnose biosynthesis (Grobben et al. 1996), and the produced EPSs possessed a smaller proportion of galactose where the activity of UDP-galactose-4-epimerase was only faintly lower in these cells. So this enzyme doesn’t cut within the sugar composition of the NCFB 2772 EPSs produced (Grobben et al. 1996). However, a correlationship between the activity of UDP-galactose 4-epimerase and EPS production was found in L. lactis (Forsén and Häivä 1981).

Glucose or the glucose element which comes from the hydrolysis reaction of lactose probably was the sugar source for the LAB heteropolysaccharide biosynthesis. It was also found that UDP-glucose pyrophosphorylase activities were high in these bacteria. Glucose-1-phosphate is likely a precursor for polysaccharide formation (Sjöberg and Hahnhägerdal 1989), where phosphoglucomutase is a hub point linking lactose degradation and EPS biosynthesis. If this branching point connects sugar catabolism and sugar anabolism, it will be interesting to engineer for overproduction of EPSs, of which the galactose element could be catabolized in glycolysis pathway, while the glucose element could be used for EPS synthesis. Therefore, the phosphoglucomutase downstream flux should be made sufficiently high (de Vos 1996). However, the experiment on the Gal⁻ (galactokinase) phenotype of S. thermophilus suggested that the function of UDP-galactose-4-epimerase and galactose 1-phosphate uridyltransferase is precursor biosynthesis for the EPSs.

Although the sugar nucleotide concentration in cell is a key factor on the monomeric composition of EPSs, the assembly of repetitive elements is also another important influencing factor. Hundreds to thousands of repetitive elements were assembled by sequential addition of carbohydrate elements via peculiar glycosyl transferases, coupled with the undecaprenyl phosphate carrier. For the first sugar residue, this isoprenoid glycosyl lipid carrier will act as recipient molecule. However, in Gram-negative LABs, only preliminary evidence for this kind of lipid carrier was found (Sutherland 1972, 1982, 1985, 1990). Moreover, the structural diversity of LAB EPSs implies that there must exist a vast set of glycosyl transferases involved in the assembly of the repetitive elements. However, these details have not been exploited thoroughly. Nevertheless, the heterologous expression of different combinations of glycosyl transferase genes will provide a new way of LAB polysaccharide engineering. Translocation of the assembled polysaccharide across the membrane to exterior and excretion in the environment or attachment to the cell are the last steps of EPS biosynthesis. Consequently, both the polymerization and transport processes will affect final amount of EPSs.

The nature and composition of EPSs are influenced by medium composition, growth phase, and generation time of bacterial growth, which are crucial factors in EPS biosynthesis and secretion. In order to increase exopolysaccharide production by Lb. casei LC2W, Ai et al. optimized the culture conditions (incubation temperature, incubation time, and inoculum concentration) which had significant influence on exopolysaccharide production (Ai et al. 2006).

The results showed that the optimal combination of the culture conditions for exopolysaccharide production was incubation temperature of 32.5 °C, incubation time of 26 h, and inoculum concentration of 4%. The maximum exopolysaccharide yield was 140 mg/L. Upon utilizing the optimal culture condition, the exopolysaccharide concentration obtained by experiment was 137 mg/L, with no statistical difference with the predicted value at 5% level of significance. The fermentation by control of pH was further investigated to enhance the exopolysaccharide yield, and the results showed that the optimal pH for exopolysaccharide production was about pH 6.0 and the yield was increased by 15%, i.e., 160 mg/L (Ai et al. 2006). Yan et al. analyzed the relationship between EPS production and tolerance to artificial gastric and intestinal juices. The exopolysaccharide production of several Bifidobacterium longum strains isolated from infant and elder feces was determined, and the priming glycosyltransferase (pGT) gene fragments of these strains were amplified and sequenced. Results indicate that their tolerance correlated well with EPS production. The phylogenetic tree of the pGT gene sequence fragments showed that the pGT genes of infant-originated strains had better homology than those of elder-originated strains (Fig. 3.5) (Yan et al. 2017).

Fig. 3.5

Phylogenetic trees of the internal fragments of priming glycosyltransferases genes obtained by PCR amplifications (A) and the corresponding amino acid sequences (B)

3.3.3 The Physiological Functionality of EPS

The probiotic functional food is now an expanding market. Probiotics are defined as live microbial food ingredients that are beneficial to health (Salminen et al. 1998). The food industry currently uses predominantly lactobacilli and bifidobacteria as probiotic bacteria, some of which produce EPSs. The health-promoting effect of EPS-secreting bacteria may have more or less relationship to the biological and physiological profiles of these secretory biopolymers. In these situations exopolysaccharides could be beneficial to human health in the way as employed by commercially available prebiotics (De Vuyst and Degeest 1999b).

The anticarcinogenic activity of LAB in fermented dairy products has been investigated. Kitazawa et al. reported that in mice if the lyophilized Lactococcus lactis subsp. cremoris KVS 20 was injected intraperitoneally, the growth of sarcoma-180 tumors would be suppressed. However, this strain did not emerge cytotoxicity to the sarcoma-180 tumor cells in vitro (Kitazawa et al. 1991). This implies that the antitumoral effect was mediated through the immune system where the slime material might be the principal factor in the antitumoral effect. In addition, a later study found that the slime material would largely induce an increase of the B-cell-dependent mitogenic activity (Kitazawa et al. 1992). Nakajima et al. found that the intraperitoneal administration of EPSs from Lactococcus lactis subsp. cremoris SBT 0495 in mice enhanced the production of specific antibodies (Nakajima et al. 1995). The starter commonly used in fermented milk products Lb. delbrueckii subsp. bulgaricus OLL 1073R-1, which produces EPSs, has been once reported to possess host-mediated antitumor activities (Kitazawa et al. 1998). The LAB EPSs were also reported to improve other immunological functions such as proliferation of T lymphocytes (Forsén et al. 1987), macrophage activation, and induction of cytokine production (Kitazawa et al. 1996). The water-soluble EPSs from kefir grains were reported to retard tumor growth when administrated orally, which is induced probably through T-cell and B-cell participation (Zubillaga et al. 2001). Dertli et al. assessed the effect of changes in cell surface characteristics on EPS production that may affect the ability of Lb. johnsonii colonizing the poultry host and exclude pathogens. Analysis of physicochemical cell surface characteristics reflected by zeta potential and adhesion to hexadecane showed that an increase in EPSs gave a less negative, more hydrophilic surface and reduced auto-aggregation (Fig. 3.6).

Fig. 3.6

Physiochemical characteristics of L. johnsonii FI9785 and mutant strains. (Dertli et al. 2015) (出版社BMC Microbiol 已成功索要,具体见邮件)

Auto-aggregation happened more frequently in mutants that have reduced EPSs, indicating that EPSs can mask surface structures responsible for cell–cell interactions. EPSs also affected biofilm formation, but here the quantity of EPSs produced was not the only determinant. A reduction in EPS production increased bacterial adhesion to chicken gut explants but made the bacteria more vulnerable to some stresses (Dertli et al. 2015). Further research is necessary to employ the EPSs or EPS-producing LAB into functional foods.

The susceptibility of EPSs to host digestion system might be a crucial point for its possible prebiotic characteristics. For the Lactococcus lactis subsp. cremoris NZ4010, its EPSs did not exert any protection role through the gastrointestinal transit of this bacterium (Looijesteijn et al. 2000). Also the resistance of EPSs to digestion was tested in vivo using rats. The rats were intervened by an EPS-rich diet for 2 weeks. Results showed that no EPSs were digested through the intestinal transit for the reason that the recovery of EPSs in the feces was nearly 100%. The resistance to biodegradability of this EPS was also studied by Ruijssenaars et al. who tested the biological breakdown susceptibility of several LAB EPSs (Ruijssenaars et al. 2000). Gut microbiota could ferment S. thermophilus SFi39 and SFi12 EPSs, which is in contrary to EPSs from Lactococcus lactis spp. cremoris B40, Lactobacillus sake 0–1, Streptococcus thermophilus SFi20, and Lactobacillus helveticus Lh59. The resistance to biodegradability was related to the EPS primary sequences, where EPSs SFi39 and SFi12 had a single β-galactosyl element in the side chains, but EPSs B40 and 0–1 own two kinds of residues, one charged and the other uncharged, making these EPSs less accessible to hydrolases. Moreover, van Casteren et al. tested several hydrolases to B40 EPS, and no activity was found, except for the Trichoderma viride crude cellulase that showed activity to the galactosylphosphate residue (Casteren et al. 1998). More studies on microbial degradability, healthy effect to hosts, and the effects toward beneficial colonic bacteria proliferation are required.

3.3.4 The Technological Functionality of EPS

Firmness and water-holding power are the most important parameters of yogurt, and the structure of the gel might be related with these parameters. With the presence of increased EPSs, the cohesiveness and firmness of ropy strains fermented yogurts decreased (Hassan et al. 1996; Marshall and Rawson 1999). The casein micelles’ association could be disturbed by EPSs, which will lead to a less firm coagulum. Studies on the microstructure of yogurt show that surrounding the EPS-producing microbes, there are some void inches which in turn would affect the matrix integrity. Yogurts made by ropy cultures had the highest ability to retain water and decrease the syneresis susceptibility (Hassan et al. 1996).

Though there was no very clear correlationship between the viscosity of yogurt and EPS concentration in it, in stirred yogurt the LAB EPSs’ contribution to the rheological properties is a common proposed assumption (Hess et al. 1997; Rawson and Marshall 2003; Marle et al. 1999). When using the non-ropy LL yogurt culture, even though the polymer amount produced was nearly the same as that of ropy RR culture, a great difference in viscosity was found. In this situation the apparent viscosity of yogurt was profoundly influenced by the spatial structure of the protein network (Marle and Zoon 1995). Wacher-Rodarte et al. and Sebastiani et al. once reported that thermophilic LAB yogurts’ viscosity values were not positively correlated with the amount of EPSs present (Wacher-Rodarte et al. 1993). Unless the production of a given type of EPSs is increased, the viscosity value also enhanced (Sebastiani and Zelger 1998). Through adding peptone to the milk, S. thermophilus could increase the EPS production, and then a concomitant increase in viscosity was observed.

In fact, the differences in viscosity improvement via different strains are principally a consequence of the differences in EPS intrinsic viscosity. Tuinier et al. described the physical traits of Lactococcus lactis subsp. cremoris B40 EPSs, of which the intrinsic viscosity and the concentration as well as shear rate could be predicted from the hydrodynamic radius and molar mass (Tuinier et al. 1999a). In the studies exploring the physical properties of aqueous solutions from L. lactis subsp. cremoris SBT 0495 EPSs, similar findings were also obtained, where the EPSs have the same repeating units as EPS B40 (Higashimura et al. 2000; Oba et al. 1999).

Normally speaking, the molar mass that is relatively high and the side chain that is relatively stiff are needed for the sake of getting a high yogurt viscosity. The effect of molar mass and radius of EPSs is also deduced by Faber et al. (1998). In the case of Lactobacillus sakei 0–1 EPSs, its average molar mass is at the same order of magnitude compared to that of xanthan gum, but its intrinsic viscosity is higher (Van den Berg et al. 1995). Monosaccharides connected by β(1–4) bonds produce stiffer chains compared with α(1–4) or β(1–3) bonds (Tuinier 1999). Meanwhile extent of branches and side chain groups play an important part in the EPSs’ stiffness. In the case of Lactobacillus helveticus K16 EPSs, the special branching pattern of its molecular structure may be the reason why in aqueous solution these EPSs showed such high viscosity (Yang et al. 2000). In another case if researchers terminally removed the galactosyl residues from Lactococcus lactis subsp. cremoris B39 and B891 EPSs, these polysaccharide molecules were apt to be given a decreased chain stiffness and thickening efficiency (Tuinier et al. 2015). On the other hand, if researchers removed off the acetyl group of B891 EPS, this modification did not affect its stiffness. In addition, depending on the solution ionic strength, the EPS intramolecular repulsion forces could be increased by the presence of charges, such as the negative charge of EPS phosphate group, which results in an enhancement of the hydrodynamic volume and thus the intrinsic viscosity. Several electron microscopy studies have been implemented (Skriver et al. 1995), and Tuinier et al. found that the depletion effects of EPSs with casein micelles happened in natural milk pH conditions (Tuinier et al. 1999b).

By means of metabolic engineering, biosynthesis of desired structure of EPSs could be accomplished. Some success has been achieved to overexpress the EPSs. The epsD gene was expressed under control of the NICE system. Compared to the wild-type L. lactis strain, the induction with nisin A results in a higher EPS production (van Kranenburg et al. 1999). Meanwhile, the attempts to produce new kind of EPSs have also been reported to be successful. When a non-EPS-producing strain was transferred into a gene cluster which encoded the EPSs from S. thermophilus SFi6, the strain would secrete another kind of EPS, unexpectedly, with a different structure, where the backbone N-acetylgalactosamine was substituted by galactose, and at position 6 of the glucose, the galactose side chain disappeared (Stingele et al. 1999). Horn et al. found that the epsC gene from the smooth mutant of Lb. johnsonii FI9785 had a single substitution (G–A) in the coding strand. Another gene in the cluster, epsE, plays a role in cell aggregation and a reduction in exopolysaccharide content (Horn et al. 2013).

However, due to low production levels, currently most LAB strains are not suitable for the commercially profitable production of EPSs. These kinds of microbes are more appropriate as functional additives, where the biopolymers are synthesized in situ, conferring natural fermented products with an improvement of rheological and prebiotic characteristics.