LuxS-mediated quorum sensing system in Lactobacillus plantarum NMD-17 from koumiss: induction of plantaricin MX in co-cultivation with certain lactic acid bacteria

A bacteriocin termed plantaricin MX with a broad antimicrobial spectrum was produced by Lactobacillus plantarum NMD-17, which was isolated from Inner Mongolia traditional koumiss of china. Among 300 strains of lactic acid bacteria (LAB) belonging to the genera Lactococcus, Lactobacillus, Streptococcus, Leuconostoc, and Enterococcus, five strains including Lactobacillus reuteri NMD-86, Lactobacillus helveticus NMD-137, Lactococcus lactis NMD-152, Enterococcus faecalis NMD-178, and Enterococcus faecium NMD-219 were revealed to significantly induce the bacteriocin synthesis and greatly increase the cell numbers of Lactobacillus plantarum NMD-17 and activity of AI-2 signaling molecule. Bacteriocin synthesis was not increased by cell-free supernatants and autoclaved cultures of inducing strains, demonstrating that intact cells of inducing strains were essential to the induction of bacteriocin synthesis. The existence of bacteriocin structural plnEF genes and the plnD and luxS genes involved in quorum sensing was confirmed by PCR, and the presence of plnB gene encoding histidine protein kinase was determined by single oligonucleotide nested PCR (Son-PCR). Quantitative real-time PCR demonstrated that plnB, plnD, luxS, plnE, and plnF genes of L. plantarum NMD-17 were upregulated significantly (P < 0.01) in co-cultivation with L. reuteri NMD-86. The results showed that the bacteriocin synthesis of L. plantarum NMD-17 in co-cultivation might have a close relationship with LuxS-mediated quorum sensing system.


Introduction
Bacteriocins from lactic acid bacteria (LAB) strains as biopreservative play an important role in food application because of their safety (De Montijo-Prieto et al. 2019;Man et al. 2019;Paramithiotis et al. 2019;Lu et al. 2020;Qiao et al. 2020). Nevertheless, narrow antibacterial spectrum, poor stability, or low yield are the main reasons for limiting the application of bacteriocin, excavating of new bacteriocins with a wide spectrum, good stability, and high yield is a challenge that food industry and researchers are faced with (Chanos and Mygind 2016;De Gianil et al. 2019;Man and Xiang 2019).
A bacteriocin called plantaricin MX from koumiss L. plantarum NMD-17 with a wide antibacterial spectrum could inhibit gram-negative and gram-positive bacteria such as Salmonella Typhimurium, Escherichia coli, Pseudomonas fluorescens, Staphylococcus aureus, Bacillus subtilis, Listeria monocytogenes, and so on. The plantaricin MX had excellent temperature tolerance (121 °C for 30 min; 4 °C and −20 °C for 30 days) and pH tolerance (2.0-10.0). However, the amount of bacteriocin synthesis is to be further improved.
Until now, the research on increasing the bacteriocin synthesis had mainly focused on the screening of high yield strains, the optimization of culture conditions, mutagenesis breeding, and heterogeneic expression, but there were some problems, such as poor specificity, limited increase, genetic instability, low safety, and unclear mechanism (O' Shea et al. 2013;Zhang et al. 2014). Previous researches indicated that co-cultivation with specific gram-positive bacteria was an effective way to increase the bacteriocin synthesis of L. plantarum such as L. plantarum J23, NC8, and DC400 (Maldonado et al. 2004;Rojo-Bezares et al. 2008;Di Cagno et al. 2010), and the induction of bacteriocin synthesis in co-cultivation was regulated by the quorum sensing (QS) system (Diep et al. 2009).
QS is a cell-cell communication system within and between species that regulates the gene expression in a growth phase-, cell density-, and co-ordinated-dependent manner ( Li et al. 2010;Wu et al. 2020). The regulatory mechanism of QS system is realized by signal molecules, a histidine protein kinase (HPK), and a response regulator (RR) with an invariant aspartate residue (Gobbetti et al. 2007). Signal molecules include acylated homoserine lactone (AI-1) (Tsai and Winans 2010), furanosylborate-diester (autoinducer-2, AI-2) (Han and Lu 2009), and oligopeptide (Freeman and Bassler 1999). LuxS AI synthase encoded by luxS gene is a key enzyme for the synthesis of AI-2, and the widespread presence of luxS gene also indicates that AI-2 is considered a universal language of interspecies communication (Kozlova et al. 2008). HPK and RR reported as twocomponent regulatory system (2CRS) include characteristic domains, called transmitters and receivers. The functions of HPK in class II bacteriocin synthesis are the receptors of signal molecules. It is generally thought that the HPK modulates the activity of the cognate RR by phosphorylation in a signal molecule-dependent manner, and the phosphorylated RR dominates the transcription of target gene in bacteriocin synthesis (Quadri 2003). The 2CRS associated with bacteriocin synthesis of L. plantarum include the following three categories: the first is PlnB, PlnC, and PlnD in L. plantarum WCFS1, C11, ST-III, V90, BFE5092, ATCC14917, and J51; the second is PlNC8HK and PlnD in L. plantarum J23, NC8, LZ206, 5-2, LPT70/3, PCS20, UL4, 163, and LB6; and the third is PlnB and PlnD in L. plantarum JDM1, UCMA3037, and 16 (Diep et al. 2009;Maldonado-Barragán et al. 2009;Tai et al. 2015;Li et al. 2016).
Therefore, this study was conducted to determine how the presence of certain LAB strains influenced the bacteriocin synthesis, cell numbers, and AI-2 activity of L. plantarum NMD-17. The bacteriocin synthesis of L. plantarum NMD-17 might be regulated by LuxS-mediated QS system in cocultivation with specific LAB strains. Therefore, the second aim was to amplify bacteriocin structural gene and the genes encoding the QS system by PCR and Son-PCR and to analyze the distinct expression by Quantitative real-time PCR.

Strains and media used in this study
The plantaricin MX-producer L. plantarum NMD-17 was originally separated from "koumiss" in Inner Mongolia. Three hundred LAB strains tested as inducers of plantaricin MX production belonged to Lactobacillus, Lactococcus, Leuconostoc, Streptococcus, and Enterococcus, which were identified previously in our laboratory by the traditional and molecular identification method. All the LAB strains were incubated in De Man-Rogosa-Sharp (MRS) (Oxoid, England) broth at 37 °C. Salmonella Typhimurium ATCC14028 was used as indicator strain to detect the antimicrobial activity and incubated in nutrient broth medium (NB) (Oxoid, England) at 37 °C. Vibrio harveyi BB170 and V. harveyi BB120 were incubated in Zobell 2216 broth and shaken at 160 rpm at 30 °C for 24 h. One liter of Zobell 2216 broth contained 5 g peptone, 1 g yeast extract, 19.45 g NaCl, 4.98 g MgCl 2 , 3.24 g Na 2 SO 4 , 1.8 g CaCl 2 , 0.55 g KCl, 0.16 g NaHCO 3 , 0.08 g KBr, 0.01 g FeSO 4 , 0.0024 g NaF, 0.024 g CsCl, 0.022 g boric acid, 0.008 g Na 2 HPO 4 , and 0.0016 g NH 4 NO 3 . All the strains were stored at −80 °C in 40% (v/v) glycerol.

Determination of antibacterial activity
Cell-free supernatants (CFS) were obtained by centrifuging the cultures at 15,000 × g for 15 min at 4 °C, after which they were adjusted to pH 7.0 using 1 M NaOH and treated with hydrogen peroxidase to eliminate the antibacterial influence of acids and hydrogen peroxide and then filter-sterilized using a filter with a pore size of 0.22 μm.
The antibacterial activity was measured by agar well diffusion assay (AWDA). Briefly, nutrient broth medium agar (0.7%) inoculated with S. Typhimurium ATCC14028 (10 7 CFU/mL) was overlaid onto 10 mL of 1.2% agar. Wells with a 6-mm diameter were cut out of the plate and filled with 50 μL of CFS. The plates were held at room temperature for 3 h in a laminar airflow hood, then cultivated at 37 °C for 12 h, after which the inhibition diameters on the plate were measured using a digital vernier caliper. Inhibition diameters were expressed as the mean ± standard deviation (SD) in mm (n = 3).

Determination of AI-2 activity
The detection of autoinducer-2 (AI-2) activity was performed as described by Bassler et al. (1997), with some modifications. Pretreatment of V. harveyi reporter strain BB170 (positive control): V. harveyi BB170 was grown overnight at 30 °C with aeration in AB broth, diluted 1:5000 into fresh AB broth. Pretreatment of V. harveyi BB120 culture and samples: V. harveyi BB120 was grown overnight at 30 °C with aeration in AB broth. Vibrio harveyi BB120 culture and samples were centrifuged at 15,000 × g for 15 min at 4 °C, adjusted to pH 7.0, and filtered through 0.22-μm filter. Determination of AI-2 activity: 90 µL of diluted V. harveyi BB170 culture were added to microtiter wells containing 10 μL of supernatant of V. harveyi BB120 culture or sample to be determined for AI-2 activity. The microtiter dishes were shaken in a rotary shaker at 160 rpm at 30 °C. Light production was measured using F4500 fluorescence spectrophotometer (Japan). AI-2 activity was expressed as the fold-induction of V. harveyi BB120 over background.

Screening of bacteriocin-inducing strains
Production of plantaricin MX by L. plantarum NMD-17 in co-cultivation with LAB strains was measured as described below: MRS broth was added with 1% of L. plantarum NMD-17 (about 10 9 CFU/mL) plus 0.5% of co-culture strain (about 10 8 CFU/mL) to analyze the induction effect, and the mixed cultures were grown for 6 h at 30 °C. Next, 50 μL of CFS was applied to measure the antimicrobial activity. The following controls were included the following: (1) inhibition diameter of bacteriocin produced by L. plantarum NMD-17, (2) inhibition diameter of bacteriocin produced by co-culture strain, and (3) inhibition diameter of mixed CFS produced by co-culture strain and L. plantarum NMD-17 in mono-cultivation (1:1). A strain was considered to be an inducer when the inhibition diameter of the CFS from its co-cultivation with L. plantarum NMD-17 was higher than each of the respective ones of the 3 controls.

Testing of cell numbers and AI-2 activity in mono-cultivation and co-cultivation
Cell numbers of L. plantarum NMD-17 in mono-cultivation and co-cultivation with inducing strains or non-inducing strains were measured as described below. The cultures were serially diluted to ascertain the colony counting (CFU/mL) on MRS agar for 36 h at 37 °C. Lactobacillus plantarum NMD-17 and co-culture strains were differentiated by colonial and morphological characteristics. AI-2 activities in mono-cultivation and co-cultivation with inducing strains or non-inducing strains were detected as described above, and the relationships between bacteriocin synthesis, cell numbers, and AI-2 activity were analyzed.

Effect of CFS and autoclaved culture of inducing strain on bacteriocin synthesis of L. plantarum NMD-17
Overnight cultures of the inducing strains were split into two aliquots, one was centrifuged at 15,000 × g for 15 min at 4 °C and filter-sterilized, and the other was autoclaved at 121 °C for 20 min. One hundred microliters of aliquots of the CFS or autoclaved cultures were co-cultivated with 200 μL of L. plantarum NMD-17 (about 10 9 CFU/mL) in 20 mL of fresh MRS broth. The co-cultivation cultures were grown for 6 h at 30 °C, then the inhibition diameter of the CFS, cell numbers, and AI-2 activity was detected as described above.
The inhibition diameter of the CFS, cell numbers, and AI-2 activity in mono-cultivation of L. plantarum NMD-17 was applied as the control.
DNA extraction, polymerase chain reaction, and DNA sequencing.

DNA extraction
Preparation of total genomic DNA from L. plantarum NMD-17 was accomplished using a TaKaRa MiniBEST Bacteria Genomic DNA Extraction Kit (Takara, Dalian, China).

PCR analysis
The existences of plnD gene encoding response regulator, luxS gene encoding LuxS AI-2 synthase, and bacteriocin structural genes (plnEF,plnA,plNC8,plnW,plnS,plnJ,plnK,pln423,plnN,pln1.25β,plnASM,and plnPCS20) in L. plantarum NMD-17 were determined by PCR; the primers based on known sequences of corresponding genes in L. plantarum from GenBank were designed in our study ( Table 1). The following conditions were applied for PCRs: denaturation at 94 °C for 4 min, followed by 30 cycles of 94 °C for 1 min, annealing (Table 1) for 45 s, polymerization at 72 °C for 90 s, and then final elongation at 72 °C for 5 min. The amplicons obtained were identified by agarose gel electrophoresis and ligated into the cloning vector PMD18-T.

Amplification of upstream gene of plnD
The single oligonucleotide nested PCR (Son-PCR) was applied to acquire the full-length sequence of regulatory gene as reported by Antal (2004) with appropriate modifications. According to the plnD sequence obtained in this study, primers Son1 and Son2 (Table 1) were applied to amplify the regulatory gene. The first reaction of Son-PCR was performed in 50 μL reaction volume containing 5 μL of 10 × reaction buffer, 4 μL of 2.5 mM dNTPs, 2 μL of 10 μM primer son1, 1 μL of 100 ng genomic DNA, 0.25 μL of 5U/μL Taq DNA polymerase, and 37.75 μL of ddH 2 O. The second reaction of Son-PCR was performed in 50 μL reaction volume containing 5 μL of 10 × reaction buffer, 4 μL of 2.5 mM dNTPs, 2 μL of 10 μM primer son1, 2 μL of 10 μM primer son2, 1 μL of 1: 25 PCR product dilution of the first reaction, 0.25 μL of 5U/μL Taq DNA polymerase, and 35.75 μL of ddH 2 O. The first reaction was started as 4 min at 94 °C, then five cycles of amplification (30 s at 94 °C, 45 s at annealing temperature, 2.5 min at 72 °C) followed by one ramping step (30 s at 94 °C, 3 min at 29 °C, 30 s at 35 °C, 30 s at 40 °C, 30 s at 50 °C, 1 min at 60 °C, 2.5 min at 72 °C) and 60 new amplification cycles (30 s at 94 °C, 1 min at 54 °C, 2.5 min at 72 °C). The second reaction was started as 4 min at 94 °C, then ten cycles of amplification (30 s at 94 °C, 45 s at 56 °C, 1.5 min at 72 °C), 2 min at 94 °C, and 35 new amplification cycles (30 s at 94 °C, 45 s at 54, 1.5 min at 72 °C). All the reactions finished with a ultimate elongation step of 10 min at 72 °C. The amplicons were detected by gel electrophoresis and ligated into the cloning vector PMD18-T. Primers plnB-NMD-F and plnB-NMD-R were used to analyze the correctness of the obtained gene sequence.

DNA sequencing and analysis
The sequencing was done by Shanghai Sangon Biotech Co., Ltd. Blast was used to determine the sequence homology analysis at http:// www. ncbi. nlm. nih. gov/ BLAST. The primer 6.0 software was used to estimate the protein sequence encoded by the open reading frame of the corresponding gene. The prediction of conserved domains and transmembrane domains was respectively proceeded by CDD (http:// www. ncbi. nlm. nih. gov/ Struc ture/ cdd/ wrpsb. cgi) and TMHMM (http:// www. cbs. dtu. dk/ servi ces/ TMHMM/). Phylogeny tree was constructed by ClustalX and MEGA 4.1.

Distinct expression analysis of plnB, plnD, luxS, plnE, and plnF genes
Quantitative real-time PCR (qRT-PCR) was applied to determine the distinct expression of plnB, plnD, luxS, plnE, and plnF genes in mono-cultivation and co-cultivation with Lactobacillus reuteri NMD-86; the primers used in this study were described in Table 2. Total RNAs were obtained from 1 mL of L. plantarum NMD-17 grown in monoculture and co-culture with L. reuteri NMD-86 at 30 °C for 6 h, respectively. Samples were centrifuged at 15,000 × g for 2 min at 4 °C, and RNA isolation was performed with the RNAprep pure Bacteria Kit as recommended by the manufacturer (Tiangen, Beijing, China). Quality control of RNA was detected by 2% agarose gel electrophoresis and NanoDrop detection method. cDNA was synthesized using the PrimeScript® RT Reagent Kit (Takara, Dalian, China) as described by the manufacturer. For quantitative  TTG TTT CCA ATT TAT TTA TAC GAG G  CTA GTT GTC TCT CAA CAA CTT ATT C  ACC ATC TTG CTG TTG ATT CAA GTC T  ATC TCT TTG   Performing melting curve analysis of the product can help to determine the specificity of amplification. Quantitative real-time PCR product was further loaded on 2% agarose gel and sequenced, RNA digested with 10 mg mL −1 RNAse at 37 °C for 45 min served as a negative control in the quantitative real-time PCR experiment. Data were normalized to levels of 16S rDNA gene and analyzed using a comparative cycle threshold. The expression level of plnB, plnD, luxS, plnE, and plnF genes in co-culture and mono-culture was compared using the relative quantification method (Di Cagno et al. 2009). Real-time data were presented as relative change compared to L. plantarum NMD-17 grown in mono-culture.

Statistical analysis
All the experiments were carried out in triplicate, and the results were expressed as the mean ± standard deviation. One-way analysis of variance (ANOVA) was used to analyze the statistical significance for data comparison by SPSS 19.0 software.

Induction of plantaricin MX in co-cultivation
Bacteriocin synthesis of L. plantarum NMD-17 in cocultivation was induced by 5 out of 300 LAB strains (data are not shown). Five inducing strains were Lactobacillus reuteri NMD-86, Lactobacillus helveticus NMD-137, Lactococcus lactis NMD-152, Enterococcus faecalis NMD-178, and Enterococcus faecium NMD-219. When L. plantarum NMD-17 co-cultivated with inducing strains, the inhibition diameter was significantly higher than all the controls (P < 0.01), and L. reuteri NMD-86 induced the highest production of plantaricin MX (Fig. 1). On the other hand, the inhibition diameters in co-cultivation with non-inducing strains were lower than the controls (data are not shown).
The conserved domains of PlnD protein were predicted by CDD (Fig. 6B); PlnD protein was a response regulator transcription factor belonging to the LytTR/AlgR family with signal receiver (REC domain) and DNA-binding domains (LytTR domain), that might be part of a two-component regulatory system. REC domain (8-124 aa) including phosphorylation site (Asp59), intermolecular recognition site, and dimerization interface could receive the signal from the sensor partner in a two-component system. LytTR domain (151-244 aa) was found in a variety of bacterial transcriptional regulators and could bind to a specific DNA sequence pattern.

Cloning and analysis of regulatory gene plnB encoding HPK
According to the known plnD sequence, primers Son1 and Son2 were designed (Fig. 7A), and two Son-PCR reactions were performed to acquire the full-length sequence of regulatory gene. About 1600-bp amplicon from the second reaction was acquired and sequenced (Fig. 7B). The sequencing result demonstrated that the amplicon comprised of 1593 bp. The sequence analysis proceeded by BLAST, the results indicated that the complete sequence of regulatory gene was acquired by Son-PCR, the open reading frame comprised of 1341 bp encoding 446 amino acids, and the nucleotide sequence had a maximum 99.85% homologous to plnB of L. plantarum JDM1 (CP001617.1). The sequence was available from GenBank under the accession number (MW415975). Primers plnB-NMD-F and plnB-NMD-R designed according to the obtained sequence were applied to determine the correctness of plnB sequence, and only one specific band of about 1300 bp was obtained (Fig. 8A).
The conserved domains of PlnB protein were predicted by CDD (Fig. 8B); PlnB protein was a histidine kinase with CitA domain and histidine kinase-like ATPase domain. CitA domain (212-421 aa) had a close relationship with signal transduction. Histidine kinase-like ATPase (343-442 aa) contained Mg-binding site, ATP lid, and many ATP-binding proteins. Chlor_Arch_YYY (13-206 aa) at the N-terminal was a highly hydrophobic domain belonging to the member of probable integral membrane family. As shown in Fig. 7C, PlnB protein including six transmembrane helices at the N-terminal was a cytoplasmic membrane protein.

Cloning and analysis of luxS gene
The luxS gene of 477 bp encoding 158 aa was detected in L. plantarum NMD-17 by PCR and sequence analysis (Fig. 9A). The sequence was submitted to GenBank under the accession number (MW415976). As shown in Fig. 9B, LuxS protein belonging to LuxS superfamily was S-ribosylhomocysteinase, which had an intimate relationship with signal transduction mechanism.
As shown in Fig. 9C, the result of phylogeny tree showed that LuxS protein of L. plantarum NMD-17 had 100% homology to that of L. plantarum JDM1, 1.0391, and ST-III, and 77-82% homology to that of other Lactobacillus. LuxS proteins of L. plantarum NMD-17, JDM1, 1.0391, and ST-III were clustered into one group.

Distinct expression analysis of plnB, plnD, luxS, plnE, and plnF genes
The agarose gel electrophoresis of total RNA showed that 5S rRNA, 16S rRNA, and 23S rRNA were clearly visible and no Fig. 7 Schematic diagram of the localization of the Son-PCR specific primers (A) and analysis of Son-PCR products by agarose gel electrophoresis (B). Primers son1 and son2 were applied to acquire DNA fragments upstream of plnD by Son-PCR. The products of the first (I) and second (II) reactions were separated by gel electrophoresis genomic contamination (Fig. 10), indicating that the integrity and purity of RNA met the requirements of reverse transcription. The OD 260/280 of RNA was found to be between 1.8 and 2.0 by NanoDrop detection, which proved that the purity of the RNA met the experimental requirements.

Discussion
Previous studies have shown that L. plantarum NC8 and J23 were unable to synthesize bacteriocin if cultured alone in liquid media; co-cultivations can induce the bacteriocin synthesis (Maldonado et al. 2004;Rojo-Bezares et al. 2007). However, the bacteriocin synthesis of L. plantarum NC8C was suppressed in co-cultivation with specific bacteriocinproducing strains (Domínguez-Manzano and Jiménez-Díaz 2013). Unlike the above results, co-cultivation with specific LAB strains could significantly increase the bacteriocin synthesis of L. plantarum NMD-17 over a normal level in our study, and the same result has been obtained in L. plantarum KLDS1.0391 (Man et al. 2012). The emergence of inducing strains might be considered an environmental stimulus, and L. plantarum NMD-17 thus initiated its own defense strategy to sense the inducing strains and finished a subsequent response involving bacteriocin synthesis by QS system.
In our study, inducing strains could significantly increase the cell numbers of L. plantarum NMD-17 and AI-2 activity in co-cultivation, but non-inducing strains decreased the cell numbers of L. plantarum NMD-17 as well as AI-2 activity. The result was similar to previously studied co-culture of L. plantarum KLDS1.0391 or NC8 (Ruiz- Barba et al. 2010;Man et al. 2014), and inconsistent with L. plantarum DC400 (Di Cagno et al. 2009). The cell numbers of L. plantarum DC400 were not influenced in co-cultivation with Lactobacillus rossiae A7 or Lactobacillus sanfranciscensis DPPMA174. Other studies rarely analyzed the changes of cell numbers of L. plantarum and AI-2 activity in co-cultivation with non-inducing strain. Cell numbers of inducing strains or non-inducing strains in co-cultivation decreased to varying degrees compared to mono-cultivation in this study. The above results strongly suggested that bacteriocin synthesis was positively correlated with cell numbers of L. plantarum NMD-17 in co-cultivation with inducing strains, which was consistent with the densitydependent manner of QS regulatory system. Meanwhile, the bacteriocin synthesis and AI-2 activity showed the same increasing trend in co-cultivation, which demonstrated that the bacteriocin synthesis was closely related to AI-2 activity.
Our results suggested that intact cells of inducing strains might be necessary to induce the plantaricin MX production, as autoclaved cultures and CFS of inducing strains were unable to do so, demonstrating that the induction of bacteriocin synthesis might be not achieved by metabolites from intact cells and dead cells of inducing strain, or inducing substances were absorbed or inactivated during membrane filtration and sterilization. These results were similar to those previously studied for other plantaricin (Rojo-Bezares et al. 2007;Man et al. 2012).
In order to amplify the genes encoding 2CRS, a large number of primers were designed according to the known sequences of L. plantarum in GenBank, the expected amplicon of plnD was acquired, but the specific band of gene encoding HPK was not obtained by PCR. Son-PCR is the newest "primer-walking" PCR method, which was applied to gain the full sequence of gene encoding HPK according to the known sequence of plnD gene in L. plantarum NMD-17. Two nested sequence-specific primers were used to acquire the specific amplicon of plnB gene. Similar with the L. plantarum JDM1, UCMA3037, and 16L, 2CRS was discovered in our L. plantarum NMD-17, which is composed of a HPK (PlnB) and a RR (PlnD). The 2CRS of L. plantarum NMD-17 was different from those of L. plantarum WCFS1, C11, ST-III, V90, BFE5092, ATCC14917, and J51 including a HPK (PlnB) and two RRs (PlnC and PlnD), and those of L. plantarum J23, NC8, LZ206, 5-2, LPT70/3, PCS20, UL4, 163, and LB6 including a HPK (PLNC8HK) and a RR (PlnD) (Kleerebezem et al. 2003;Navarro et al. 2008;Diep et al. 2009). Sensation of the extracellular stimulus for bacteriocin synthesis was accomplished by the HPK, it had been confirmed that it was the N-terminal transmembrane structural domain of HPK that contacted with the signaling molecule or extracellular stimulus (Johnsborg et al. 2003;Chanos and Mygind 2016), PlnB protein had the same N-terminal transmembrane structural domain in L. plantarum NMD-17 by CDD analysis, the results indicated that PlnB protein had the potential to carry out the signal transduction in 2CRS. PlnD protein including signal receiver domain and DNA-binding domains belonged to the LytTR/AlgR family by CDD analysis, and it was possible to receive signals from HPK and activate the genes related to bacteriocin synthesis.
LuxS-mediated QS system producing the universal signaling molecule (AI-2) by the activity of LuxS enzyme (S-ribosylhomocysteinase) existed in many G + and G − bacteria and was applied to carrying out the interspecies communication (Winzer et al. 2002;Federle and Bassler 2003; Kaper and Sperandio 2005). luxS gene encoding LuxS protein has been shown to be highly conserved among bacterial species and was a necessary gene for the synthesis of AI-2. The current research on the luxS gene mainly concentrated on its function in the pathogenic bacteria (Asanuma et al. 2004;Kim et al. 2007;Zhu et al. 2007;Kozlova et al. 2008;Sztajer et al. 2008;Medellin-Peña and Griffiths 2009). Nevertheless, the function of luxS gene on bacteriocin synthesis in L. plantarum was less known. In this study, the luxS gene of L. plantarum NMD-17 was relatively easy to obtain by PCR method because of its high conservation, which provided a prerequisite for the qRT-PCR and gene knockout to analyze the influence of the luxS gene on bacteriocin synthesis. The result of phylogeny tree showed that the LuxS protein in Lactobacillus evolved from a common ancestor, confirming the high conservation of luxS gene to a certain extent.
To determine the influence of LuxS-mediated QS system on the bacteriocin synthesis of L. plantarum NMD-17 at transcription level, the transcription level of the luxS, plnB, plnD, plnE, and plnF gene was measured in co-cultivation with L. reuteri NMD-86 by quantitative real-time PCR. A single peak was observed in the melting curve generated from plnB, plnD, luxS, plnE, and plnF. The above result indicated the specificity of the amplified PCR products. Consistent with the study of Di Cagno (2009) and Man et al. (2014), the housekeeping gene 16S rDNA was applied as reference gene. Our previous studies showed that 16S rDNA as reference gene had high PCR efficiency and stability in co-culture or in mono-culture. The transcription levels of the encoding genes of LuxS-mediated QS system (luxS, plnB, plnD) in co-cultivation were significantly higher than those of monocultivation, accompanied by the upregulation of bacteriocin structural genes (plnE, plnF). Meanwhile, the bacteriocin synthesis, cell density, and AI-2 activity during co-culture were significantly higher than those of mono-culture. The above results further indicated that the bacteriocin synthesis of L. plantarum NMD-17 in co-cultivation might have a close relationship with LuxS-mediated QS system.
Based on the results of this study, it is assumed that inducing bacteria might be considered an environmental stimulus by L. plantarum NMD-17, activating the defense strategy of L. plantarum NMD-17 to increase the cell numbers, accompanied by the increase of signal molecule AI-2. When the concentration of AI-2 reached the threshold value, the genes related to bacteriocin synthesis were activated. Nevertheless, non-inducing strains were unable to increase the cell numbers of L. plantarum NMD-17 and AI-2 activity and trigger the LuxS-mediated QS system to regulate the bacteriocin synthesis.
In summary, the bacteriocin synthesis of L. plantarum NMD-17 was only induced by the intact cells of specific LAB strains. Co-cultivation with bacteriocin-inducing strains was beneficial to obtain higher cell numbers of L. plantarum NMD-17 and AI-2 activity. The encoding genes of LuxS-mediated QS system (plnB, plnD, and luxS) and bacteriocin structural gene (plnE and plnF) were detected in L. plantarum NMD-17 and significantly upregulated (P < 0.01) in co-cultivation with L. reuteri NMD-86 by qRT-PCR. Future study will concentrate on gene knockout and the addition of exogenous AI-2 to indicate their influence on bacteriocin synthesis of L. plantarum NMD-17.