Introduction

The adherence of bacteria to the surface of intestinal mucosa is a key process for colonization and persistence in gastrointestinal tract (GIT). Therefore, adhesion is an important property for both pathogenic bacteria and bacteria belonging to normal gut microbiota [1, 2]. Several diseases of GIT are caused by Salmonella enterica serotype Typhimurium and Escherichia coli [3, 4]. To cause infection, adhesion to epithelium is essential; preventing pathogenic bacterial adhesion to the epithelial cells is an effective strategy to reduce pathogen-associated illness. The epithelium of GIT can be protected from colonization of pathogen by a number of mechanisms including antibiotic treatment. Probiotic therapy is seen as potential attractive alternative to antibiotic treatment to overcome problems such as emergence of multidrug-resistant bacteria and imbalance of resident gut microbiota [5, 6]. Several Lactobacillus strains are being explored as probiotic bacteria to restore and maintain normal gut microbiota as well as treatment of gastrointestinal diseases. Out of these, L. rhamnosus GG is the best studied and widely accepted standard probiotic strain [7]. This strain that has originated from the intestinal tract of a healthy human is being studied for the treatment of acute diarrhea and prevention of inflammatory bowel diseases [8, 9].

Adherence of different probiotic bacteria has been studied using various eukaryotic cell lines of human origin such as Caco-2 and HT-29 [10–12]. Using these cell lines, several Lactobacillus strains have been shown to reduce the colonization of pathogenic bacteria through competition for adherence. It is believed while inhibition of pathogens is through soluble effectors; competitive exclusion of pathogenic bacteria is through direct competition for common attachment sites on gut epithelium [13–15]. The adhesive mechanism of pathogenic bacteria is well studied; however, the knowledge about surface molecules mediating the adhesion of lactobacilli to the intestinal epithelium is scant, as only few of them have been identified and characterized. The adhesion proteins that have been characterized from different Lactobacillus strains include mucus-binding protein (Mub/CyuL; L. acidophilus NCFM), mucus adhesion–promoting protein (MapA; L. reuteri 104R), surface layer proteins (Slp; L. helveticus R0052), and elongation factor Tu (EF-Tu; L. johnsonii NCC533) [16–19]. EF-Tu is a G-protein that facilitates the transfer of aminoacyl-tRNA to the acceptor site of the ribosome. Despite being a cytoplasmic and an anchorless protein, EF-Tu has also been found at the surface of L. johnsonii NCC 533 (La1) and it mediates the colonization of the bacteria by attachment to mucus and intestinal epithelium [16]. However, the mechanism by which EF-Tu interacts with intestinal cell is not well understood. The presence of EF-Tu at the surface of pathogenic E. coli and periplasm of Neisseria gonorrhoeae also has been reported [20, 21].

The aim of the present study was to investigate the interaction of two Lactobacillus strains, isolate L. plantarum CS24.2 and standard L. rhamnosus GG, with pathogenic strain of E. coli and Salmonella enterica serovar Typhi (S. Typhi). The effect of different co-culture conditions on pathogen adhesion to intestinal epithelial cell line Caco-2 has been analyzed. The applicability of EF-Tu in preventing the adhesion of pathogenic bacteria to gut epithelium was also examined by competitive exclusion of pathogens from Caco-2 cells, using partially purified recombinant EF-Tu derived from L. plantarum CS24.2.

Materials and methods

Bacterial strains and culture conditions

Lactobacillus plantarum CS24.2 used in this study was isolated from human child fecal sample. The isolate L. plantarum CS24.2 has shown significant adhesion to intestinal epithelial cell line—19 bacteria per Caco-2 cells in 24-well tissue culture plate, tolerances toward bile and acid (pH-2) at the survival rate of 12.7 and 4.73%, respectively. The isolate has also shown broad antimicrobial activity against both Gram-positive bacteria (Staphylococcus aureus) and Gram-negative bacteria (Shigella dysentery, Pseudomonas aeruginosa, and Escherichia coli) (unpublished data). The isolate has shown the ability to stimulate Caco-2 leading to an alteration in the levels of transcript for IL-8 and TGF-β, hence demonstrating its immunomodulatory potential. The probiotic properties analyzed are comparable with other isolates and standard strains previously reported [22]. The probiotic strain L. rhamnosus GG (LGG) was obtained as a kind gift from Dr. Shira Doron, MD, Department of Medicine, Tufts–New England Medical Center, USA [8]. The enteropathogenic E. coli serotype O26:H11 (EPEC) and S. Typhi (MTCC 733; IMTECH, Chandigarh, India) were obtained from culture collection of our department. Lactobacilli were grown on de Man–Rogosa–Sharpe (MRS; Himedia, Mumbai, India) broth at 37 °C for 16–18 h before study. E. coli and S. Typhi were grown aerobically at 37 °C on Luria broth (Himedia) for 16–18 h.

Caco-2 cell culture

The Caco-2 cell line was obtained from National Centre for Cell Science (NCCS), Pune, India, and was routinely cultured in Dulbecco’s modified Eagle’s minimal essential medium (DMEM; Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% (v/v) fetal bovine serum (FBS; Sigma-Aldrich), 10 mM nonessential amino acids, 1 mM Na pyruvate, and 50 μg/mL gentamycin, at 37 °C temperature in a humidified atmosphere containing 5% CO2–95% air atmosphere. The media lacked gentamycin whenever antibiotic-free medium was used. To obtain monolayers for adhesion assays, the Caco-2 cells were seeded at a density of 104 cells/well in 24-well standard tissue culture plates (Corning Incorporated, NY, USA) and maintained for 2 weeks after confluence. The culture medium was replaced after every 2 days.

In vitro adhesion assays

For adhesion assays, postconfluence Caco-2 cells were pre-incubated with antibiotic-free medium for 4 h. The pH of the media used for adhesion assay was adjusted to 6.5 before use with 1 N HCl. Lactobacilli and pathogenic strains were harvested by centrifugation (10,000g, 2 min, and 4 °C) and washed twice with Dulbecco’s phosphate-buffered saline (PBS), pH 7.0 (Sigma-Aldrich), and the cell density was adjusted by measuring absorbance at 600 nm to get 1 × 108 cfu/well in 50 μL cell suspension for each culture.

For each combination of Lactobacillus with enteropathogens, three competition conditions were tested: competitive inhibition, adhesion inhibition, and displacements. In competitive adhesion assay, lactobacilli and pathogenic strains were added to the wells with Caco-2 monolayers in same number at the same time and co-incubated for 90 min. In adhesion inhibition assay, lactobacilli were allowed to adhere to Caco-2 cells for 90 min. Un-adhered bacterial cells were then withdrawn from the wells and the Caco-2 monolayers were washed twice with 1 mL PBS each. The pathogenic strains were then added and incubated for an additional 90 min. In displacement assay, pathogenic strains were added first to wells with Caco-2 cells for 90 min followed by lactobacilli adherence after removing unbound pathogenic cells as mentioned above. At the end of each assay, Caco-2 monolayers were washed twice with 1 mL PBS each. To release adhered cells, Caco-2 cells were lysed by treatment with 0.5 mL 0.05% (v/v) Triton X-100 in PBS for 20 min at 37 °C. The Caco-2 lysate including bound bacterial cells was plated after appropriate dilution on Luria agar plate, and the enumeration was done after 24-h incubation at 37 °C. Co-incubations of Caco-2 cells with pathogenic strains alone were taken as control, and the number of bacteria adhering to Caco-2 cells was considered as 100%. It was also determined that a 30-min treatment with 0.05% (v/v) Triton X-100 in PBS at 37 °C did not affect the viability of lactobacilli and pathogenic strains (data not shown). The experiments were repeated twice in duplicates in two successive passages.

Expression of EF-Tu in E. coli

The EF-Tu gene was amplified from L. plantarum CS24.2 by colony PCR. The PCR primers F: 5′ GTCAGCATATGATGGCAAAAGAACATTAT 3′ and R: 5′ GTATAGGATCCGTCATCAATTTCTGAAAC 3′, containing recognition sequences (underlined) for restriction enzymes Nde1 and BamH1 sites, respectively, were designed using sequence of EF-Tu gene (NC_004567) from L. plantarum WCFS1. Amplification was performed using AccuTaq (Sigma-Aldrich) under the following conditions: incubation for 6 min at 94 °C followed by 30 cycles of 30 s at 94 °C, 30 s at 55 °C and 1 min at 72 °C, and final incubation for 10 min at 72 °C before holding at 4 °C. The expected 1.2-Kb amplicon was observed on agarose gel electrophoresis. The amplicon was cloned initially into pJET1.2 blunt-end cloning vector (Fermentas, Maryland, USA). The cloned EF-Tu gene was excised from pJET-EFTu using restriction enzymes Nde1 and BamH1 and subcloned by directional cloning into expression vector pET30(a) (Novagen, EMD Biosciences, Darmstadt, Germany) giving rise to pET-EFTu. This clone was confirmed by sequencing. For expression analysis, E. coli BL21(DE3) transformed with pET-EFTu was inoculated with overnight-grown culture in 5 mL of Luria broth containing 1% glucose and 40 μg/mL of kanamycin and incubated at 37 °C till log phase was achieved. Likewise, E. coli BL-21(DE3) transformed with control vector pET30(a) was also inoculated and used as a negative control. Isopropyl-β-D-thiogalactopyranoside (IPTG; Fermentas) was added to a final concentration of 5 mM and culture was further incubated at 37 °C and aliquots were withdrawn at different time points to check for the optimal expression of recombinant protein. Induced cells of E. coli BL-21(DE3) transformed with pET-EF-Tu and pET30(a) were harvested by centrifugation at 10,000g × 10 min, 4 °C, and pellet was resuspended in 1 × gel loading buffer, boiled for 5 min, and analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) using 10% polyacrylamide gels. Protein bands were visualized by staining the gel with Coomassie brilliant blue R-250 (Sigma-Aldrich).

Partial purification of recombinant EF-Tu

The cells were grown in 25 mL of Luria broth containing 1% glucose and 40 μg/mL of kanamycin and induced with 5 mM IPTG at log phase. They were harvested after 14-h incubation by centrifugation at 10,000g for 10 min at 4 °C. The cells were resuspended in 3 mL PBS and lysed by sonication (9.9 s ON, 9.9 s OFF, 3 min 30 s). The lysate was centrifuged at 10,000g for 10 min at 4 °C. The pellet obtained therefrom was resuspended in 3 mL PBS. EF-Tu was purified from semi-denaturing SDS-PAGE using an in-house electroelution apparatus. Briefly, postsonication resuspended pellet was incubated overnight at 37 °C with gel loading buffer without β-mercaptoethanol and thereafter resolved on 10% SDS-PAGE. The band corresponding to 50 kDa was excised from the gel and was electroeluted into a dialysis membrane by overnight transfer (50 V, 4 °C) in transfer buffer containing Tris-glycine. The eluted protein was dialyzed twice against 2.5% Triton X-100 in 50 mM Tris (pH 8.8) and twice against 50 mM Tris (pH 8.8). Subsequently, the eluted sample was run on 10% SDS-PAGE along with various concentrations of bovine serum albumin (BSA). The gel was then stained with silver nitrate and EF-Tu estimated by densitometric analysis employing AlphaEaseFC 4.0 software (data not shown).

Adhesion of Lactobacillus strains and enteropathogens to Caco-2 cells in the presence of recombinant EF-Tu

Postconfluent Caco-2 cells in 24-well tissue culture plate were pre-incubated with 50 μl of recombinant EF-Tu (8 μg/mL) and 900 μL of DMEM to give a final concentration of 400 ng/mL EF-Tu. The cells were incubated for 15 min at 37 °C followed by addition of 50 μL suspension of lactobacilli and pathogenic strains in respective wells to get 1 × 108 cfu/well. The electroeluted sample from the corresponding region of the SDS-PAGE gel lane bearing induced E. coli BL21(DE3) transformed with pET30(a) vector was taken as negative control for the effect of EF-Tu. After 90-min incubation in a 5% CO2–95% air atmosphere, adhered lactobacilli and pathogenic strains from respective wells were counted by plating on MRS and Luria agar plates, respectively. The experiment was carried out in duplicate in two successive passages.

Antibody-mediated adhesion inhibition of Lactobacillus strains

The antibody-mediated inhibition assay was performed with Caco-2 monolayers in 24-well tissue culture plate according to the method previously described [23]. Different concentrations of rabbit anti-EF-Tu polyclonal antibody (2 mg/mL)—5, 10, and 15% in DMEM—were initially used for adhesion inhibition. Based on the initial observation, 0.5 ml of DMEM containing 0 and 5% rabbit anti-EF-Tu polyclonal antibody and approximately 0.5 × 108 cfu of lactobacilli were added to Caco-2 monolayers. Caco-2 monolayers incubated with negative serum and lactobacilli were used as controls. After 90-min incubation in a 5% CO2–95% air atmosphere, adhered lactobacilli from respective wells were counted by plating on MRS agar plate. The experiment was carried out in duplicate in two successive passages.

Statistics

Values are given as mean along with the standard deviation (SD) from two independent experiments in duplicate. Significant ANOVAs were followed by Dunnett test in the case of different adhesion assay to Caco-2 cells and compared with the respective control (p < 0.05). The statistical analysis was conducted using SigmaStat 3.5 software.

Results

Antagonistic effect of Lactobacillus strains on adhesion of E. coli and S. Typhi to Caco-2 cells

To mimic various in vivo conditions, adhesion assays were designed to include competitive adhesion, adhesion inhibition, and displacement, using probiotic Lactobacillus strains, L. plantarum CS24.2 and L. rhamnosus GG, and enteropathogens, E. coli and S. Typhi. The numbers of enteropathogens bound to Caco-2 monolayer under different competitive adhesion conditions are given in Fig. 1a, b for L. plantarum CS24.2 and in Fig. 1c, d for L. rhamnosus GG.

Fig. 1
figure 1

Adhesion of Escherichia coli O26:H11 and Salmonella Typhi MTCC 733 to Caco-2 cells following competition with, inhibition from, and displacement by L. plantarum CS24.2 (a and b) and L. rhamnosus GG (c and d). Adhesion of E. coli and S. Typhi in the absence of lactobacilli is denoted as control. Each bar represents mean value and standard deviation as error bar. The square box above each bar shows the percentage reduction in adhesion when compared with control. Significant ANOVAs were followed by Dunnett test for multiple comparisons versus control group. *Mean value of adhesion was significantly lower than that of control (p < 0.05)

Full size image

In competitive adhesion assay, lactobacilli and enteropathogens were provided with a chance for binding at an equal ratio. L. rhamnosus competes better with both enteropathogens compared with L. plantarum. An adhesion inhibition assay was performed to investigate the role of lactobacilli growth and its ability to protect intestinal cells from being colonized by the pathogens. The pre-colonization of Caco-2 monolayers with L. rhamnosus reduced the adhesion of E. coli and S. Typhi. In contrast, no significant changes in pathogen adhesion could be detected in case of Caco-2 monolayers pre-incubated with L. plantarum. To investigate the ability of lactobacilli to displace colonized pathogens from intestinal epithelium, the pathogens were allowed to adhere first to Caco-2 monolayers before lactobacilli adhesion. The degree of displacement of S. Typhi by Lactobacillus strains was higher compared with that of E. coli.

Overall, the good adhesive strain L. rhamnosus GG was found to exhibit significant antagonistic activity under different adhesion assay conditions toward both pathogenic strains.

Cloning, expression, and purification of recombinant EF-Tu

The sequencing of cloned EF-Tu gene has shown 99% identity with EF-Tu gene of L. plantarum WCFS1. The amino acid sequence deduced from the nucleotide sequence was analyzed using Pfam database for adhesive domain search. The analysis showed that the L. plantarum CS24.2 EF-Tu has three adhesive domains that are an EF-Tu GTP-binding domain (PF00009), an EF-Tu domain 2 (PF03144), and an EF-Tu C-terminal domain (PF03143). Optimum time for the induction of EF-Tu in E. coli BL21(DE3) was observed to be 14 h (Fig. 2a). A 50-kDa band corresponding to EF-Tu was observed in the lane with E. coli BL21(DE3) transformed with recombinant pET-EF-Tu plasmid in contrast to that in the negative control that contained the similarly induced sample of E. coli BL21(DE3) transformed with control plasmid pET30(a) alone. The partially purified EF-Tu preparation was analyzed on 10% SDS-PAGE employing silver staining (Fig. 2b).

Fig. 2
figure 2

a Coomassie-stained 10% SDS-PAGE gel with total cell lysate of E. coli BL21(DE3) transformed with pET30(a) (vector control; lane 1) and pET-EF-Tu (lane 2). b Silver-stained 10% SDS-PAGE gel with partially purified EF-Tu (lane 1) and protein marker weight marker (lane 2). Open right arrow Denotes band of 50-kDa EF-Tu

Full size image

Competition assays between EF-Tu and lactobacilli for adhesion to Caco-2 cells

To evaluate the contribution of EF-Tu in adhesion of Lactobacillus strains to Caco-2 cells, competition assays were performed in the presence of partially purified recombinant EF-Tu preparation. In competition assays, there was significant reduction (p < 0.05) in adhesion of L. plantarum CS24.2 to Caco-2 cells, when the EF-Tu was co-incubated with the bacterial cells (Fig. 3a). Adhesion of L. plantarum CS24.2 was reduced by 50% in the presence of the recombinant EF-Tu when compared with control. In contrast, there was no effect on the adhesion of L. rhamnosus GG when co-incubated with EF-Tu, under similar assay conditions.

Fig. 3
figure 3

a Adhesion of L. plantarum CS24.2 and L. rhamnosus GG to Caco-2 cells in the presence of partially purified recombinant EF-Tu (filled square box). Adhesion of lactobacilli in the absence of EF-Tu is denoted as control (opensquare box). b Effect of EF-Tu on the adhesion of L. plantarum CS24.2 and L. rhamnosus GG to Caco-2 cells determined by antibody-mediated adhesion inhibition assay (filled square box). Adhesion of lactobacilli in the absence of rabbit anti-EF-Tu polyclonal antibody is considered as absolute binding and denoted as control (opensquare box). c Adhesion of E. coli O26:H11 and S. Typhi MTCC 733 to Caco-2 cells in presence of partially purified recombinant EF-Tu (filled square box). Adhesion of E. coli and S. Typhi in the absence of EF-Tu is denoted as control (opensquare box). Each bar represents mean value and standard deviation as error bar. The square box above bar shows the percentage reduction in adhesion when compared with control. Significant ANOVA was followed by Dunnett test for comparison versus control group. *Mean value of adhesion was significantly lower than that of control (p < 0.05)

Full size image

Antibody-mediated inhibition assay of Lactobacillus strains

To further confirm the role of EF-Tu in adhesion, antibody-mediated adhesion inhibition assay was performed in the presence of rabbit anti-EF-Tu polyclonal antibody. Co-incubation with negative serum showed no effect on adhesion of lactobacilli compared with control where no serum was added. There was significant decrease (p < 0.05) in adhesion of L. plantarum CS24.2 by addition of 5% polyclonal antibody (Fig. 3b). Adhesion of L. plantarum CS24.2 was reduced by 22% in the presence of polyclonal antibody when compared with control. In contrast, there was no significant effect on the adhesion of L. rhamnosus GG when co-incubated with polyclonal antibody, under similar assay conditions.

Competition assays between EF-Tu and enteropathogens for adhesion to Caco-2 cells

To study the effect of EF-Tu on the adhesion of enteropathogens to Caco-2 cells, the bacterial cells were co-incubated with partially purified recombinant EF-Tu preparation of L. plantarum CS24.2 in competitive adhesion assays as described above. There was significant decrease (p < 0.05) in the adhesion of both enteropathogens in the presence of EF-Tu compared with control. The effect of EF-Tu addition to assay was more prominent on the adhesion of S. Typhi compared with that of E. coli. Adhesion of E. coli and S. Typhi to Caco-2 cells was strongly inhibited by addition of recombinant EF-Tu, as it was reduced by 35.3 and 47.7%, respectively (Fig. 3c).

Discussion

Lactobacillus strains have been widely explored as therapeutic agents, as many of them possess health-promoting and corrective properties [24, 25]. Adhesion to intestinal epithelium is a desirable criterion for Lactobacillus colonization and persistence in gastrointestinal tract before it can confer its probiotic effects to the host [26, 27]. Caco-2 cell line is widely used and accepted for study related to probiotic and pathogens adhesion to intestinal epithelium [28]. In order to study the role played by probiotic lactobacilli to protect against pathogen colonization, the competition of two Lactobacillus strains (L. plantarum CS24.2 and L. rhamnosus GG) and adhesive enteropathogens (E. coli O26:H11 and S. Typhi MTCC 733) for adhesion to Caco-2 cells was studied. When incubated along with pathogens, both lactobacilli showed good competitive inhibition and pathogen adhesion to Caco-2 cells decreased by around 40–60%. Lee et al. [29] under similar assay condition reported around 20–50% inhibition of strains of E. coli and S. Typhimurium adhesion to Caco-2 cells by L. rhamnosus GG. It is suggested that the degree of competition is strain dependent and can be determined by the affinity of adhesion molecules present on surface of respective bacteria for the receptor-binding sites that they are competing for [30].

In adhesion inhibition studies, the capacity of lactobacilli to prevent pathogen adhesion was analyzed. Only L. rhamnosus GG was found to exclude pathogen adhesion by around 30% when Caco-2 monolayer was pre-incubated with lactobacilli. Whether the exclusion of pathogens is due to competition for common adhesion receptors or steric hindrance of adhered lactobacilli needs to be understood. Collado et al. [31] have shown 30.3 and 27.9% adhesion inhibition of E. coli and S. Typhimurium by L. rhamnosus GG. Furthermore, Lee and Poung [30] have shown that the degree of adhesion inhibition depends on the relative position of the hydrophobic surface and adhesion receptors. In the present study, both lactobacilli strains examined exerted strong displacement ability toward S. Typhi and significant displacement toward E. coli. The displacement activity exerted by probiotic bacteria depends on the production of antimicrobial compounds or antiadhesion factors rather than mere competition for common adhesion receptors [32]. Candela et al. [33] showed 40–90% displacement of E. coli H10407 and S. Typhimurium from Caco-2 monolayers by strains of L. acidophilus. Overall, L. rhamnosus GG effectively reduces the adhesion of pathogens in all three adhesion assays but L. plantarum CS24.2 is less effective in adhesion inhibition. As per our earlier report, the adhesion ability of L. rhamnosus GG is higher compared with L. plantarum CS24.2 and, on the other hand, the antimicrobial activity is lower when compared with L. plantarum CS24.2 [22]. So it is indicative of the possibility that the adhesion property of L. rhamnosus GG may have a dominant role to play over other properties such as antimicrobial activity.

Most lactobacilli established as probiotic have been selected by their superior phenotypic properties such as adhesion to Caco-2 cells, bile and acid tolerance, and/or antimicrobial activity [8, 34, 35]. However, studying the underlying molecular mechanisms for these properties can assist in exploring the respective molecules for direct application. A variety of adhesive molecules have been identified from different strains of lactobacilli. In case of L. plantarum and L. rhamnosus strains, few of the known adhesive molecules are alfa-enolase, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and mucus via mannose binding (Msa) and modulator of adhesion and biofilm (MabA) and secreted LPXTG-like pilin (SpaC), respectively [36–40]. Some highly conserved cytoplasmic proteins such as EF-Tu, GAPDH, and GroEL, which lack typical signal sequence and membrane anchoring mechanism for surface expression, have been characterized [16, 37, 41]. These proteins are normally referred as anchorless multifunctional proteins or moonlighting proteins. The published genome database of the lactobacilli denotes the presence of EF-Tu gene in all those Lactobacillus strains which is not surprising since the elongation factor Tu (EF-Tu) is a G-protein that plays an important role in protein synthesis. Granato et al. [16] had first demonstrated the presence of EF-Tu on the surface of L. johnsonii NCC 533 (La1) and reported it as an adhesive molecule mediating attachment to mucin and intestinal epithelial cells.

To assess the role of EF-Tu as an adhesive molecule, we carried out heterologous expression of EF-Tu from L. plantarum CS24.2 into E. coli, followed by partial purification. Adhesion of L. plantarum CS24.2 to Caco-2 cells in the presence of partially purified recombinant EF-Tu was reduced by around 50%, which supports the adhesive role of EF-Tu. Similarly, Granato et al. [16] had also shown that La1 EF-Tu could prevent up to 40% adhesion of La1 bacteria to mucin. Under the present study, there was no effect on the adhesion of LGG to Caco-2 cells, which suggests that EF-Tu may not have an exclusive role in the adhesion of LGG to Caco-2. The unequivocal contribution of EF-Tu as adhesin, at least in case of L. plantarum CS24.2, has been reiterated by the experiment carried out using polyclonal antibody to EF-Tu. It may be interesting to note that the recently published genome sequence of LGG reveals the presence of EF-Tu (YP_003171088) and shows 95.7% similarity to La1 EF-Tu. The sequence information also suggests the presence of all 3 adhesive domains reported for La1 EF-Tu. However, it has been reported by others that in the case of fibronectin-binding protein of isogenic mutants, two lactobacilli strains with similar fibronectin-binding protein did not show similar results in adhesion to Caco-2 cells. L. acidophilus NCFM fbpA− isogenic mutant showed reduction in adhesion to Caco-2 cells when compared with wild-type strain, while such mutation in L. casei did not show any reduction in adhesion to Caco-2 cells [17, 42].

Moonlighting function for many housekeeping proteins of pathogens is well documented, and many of them act as enhancers of pathogen virulence. For example, when GAPDH is displayed on streptococcal surface, it has been reported to function as fibronectin-binding protein [43]. The mechanism of pathogen inhibition by lactobacilli is still under investigation, but Nesser et al. [44] have shown that L. johnsonii La1 bind to intestinal cell membrane molecules through mechanisms that are shared by pathogenic bacteria. The data presented in this study show that EF-Tu as an adhesive molecule of lactobacilli, provides beneficial effects against enteropathogen infection in vitro. The competitive adhesion analysis using recombinant EF-Tu against enteropathogens demonstrated a reduction in pathogen adhesion by 35–50%. However, it is also necessary to know whether pathogen exclusion is mediated through steric hindrance or by interacting with specific receptors. The role of EF-Tu in pathogen virulence is not yet studied but the presence of EF-Tu at the surface of pathogenic E. coli and membrane association is well documented [20]. Bioinformatics analysis of EF-Tu from several pathogenic strains of E. coli and S. Typhi using NCBI and Pfam databases shows around 83% similarity with La1 EF-Tu and the presence of all three adhesive domains (PF00009, PF03144, and PF03143) reported for La1 EF-Tu. So the role of these EF-Tu-like molecules in the adhesion of pathogen can be further investigated to establish EF-Tu-mediated inhibition of enteropathogens by lactobacilli.

The selection of a lactobacilli strain on the basis of its ability to antagonize specific pathogens is an important step in the development of a probiotic product for the prevention and treatment of infection caused by that pathogen. In this work, we demonstrated that probiotic strain of L. plantarum CS24.2 and L. rhamnosus GG is able to antagonize the adhesion of pathogenic strain of E. coli and S. Typhi in vitro. Earlier, Izquierdo et al. [45] demonstrated the overexpression of EF-Tu in the cell wall proteome of a highly adhesive strain L. plantarum WHE 92. This work further supports the role of EF-Tu in Lactobacillus adhesion. In conclusion, this is the first report stating that the moonlighting protein EF-Tu from probiotic lactobacilli antagonizes the adhesion of enteropathogens to intestinal cells. The study should serve as template for screening and selection of effective probiotic adjuncts as advancement in the field of food technology.