OUP user menu

Characteristics of the adhesion of PCC®Lactobacillus fermentum VRI 003 to Peyer's patches

Seok-Seong Kang , Patricia L. Conway
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00315.x 19-24 First published online: 1 August 2006

Abstract

The characteristics of the adhesion of PCC®Lactobacillus fermentum VRI 003 to Peyer's patches was studied in vitro. The adhesion of L. fermentum 003 was strongly inhibited in the presence of d-mannose and methyl-α-d-mannoside although other carbohydrates tested, such as N-acetyl-glucosamine, d-galactose, d-glucose and l-fucose, did not affect the adhesion. Lactobacillus fermentum 003 was shown to strongly attach to mannose immobilized on a surface using BSA, suggesting that L. fermentum 003 specifically adhered to mannose-containing molecule(s). Pretreatment of L. fermentum 003 with proteinase K and trypsin decreased the adhesive capacity and bacterial surface extracts diminished adhesion of L. fermentum 003 indicating that cell surface proteins are involved in adhesion to Peyer's patches. It was concluded that a mannose-specific protein mediated adhesion of L. fermentum 003 to the Peyer's patches.

Keywords
  • Lactobacillus
  • Peyer's patches
  • mannose-specific adhesion

Introduction

Lactobacilli have been well studied as probiotic bacteria and their adhesive capacity and colonization of the gastrointestinal tract in humans and animals extensively investigated. Many Lactobacillus strains exert beneficial health effects on the host and they are generally regarded as safe organisms. These health-promoting bacteria are being evaluated as vaccine vectors for foreign antigen delivery (Claassenet al,1995; Pouwelset al,1996).

Peyer's patches are aggregated lymphoid follicles that function as inductive sites of the mucosal immune system. A distinct population of cells, known as M cells, can be mainly found in the follicle-associated epithelium (FAE) that overlays aggregated lymphoid follicles. M cell surface molecules may serve as attachment sites for macromolecules and microbial surface molecules such as lectins (Clarket al,2000). The association of pathogenic bacteria with Peyer's patches has been widely studied (Inmanet al,1983; Owenet al,1986; Clarket al,1994), however, knowledge of the interaction of nonpathogenic bacteria with Peyer's patches is limited. The application of nonpathogenic bacteria such as Lactobacillus strains as a live vector for antigen delivery could eliminate risks of the use of the current vaccines using attenuated pathogenic bacteria. At this stage no recombinant lactobacillus vaccine studies have been carried out in humans, however lactobacilli have the potential for being used as a delivery vehicle in addition to the well recognized benefits of the live organism as a probiotic.

The association of lactobacilli with the Peyer's patches has been studied and it was shown that Lactobacillus fermentum KLD selectively attached to the FAE of the Peyer's patches in mice, and exhibited adjuvant activity (Plant & Conway, 2001; Plant & Conway, 2002).

A number of studies have examined the adhesive properties of Lactobacillus strains to gastrointestinal tissues, mucus or epithelial cells. Surface proteins (Conway & Kjelleberg, 1989; Coconnieret al,1992; Greeneet al,1994; Rojaset al,2002) or carbohydrate-specific molecules (Mukaiet al,1992; Gusilset al,1999) of bacterial cell surface components have been demonstrated to function as mediators of Lactobacillus adhesion. Several carbohydrate determinants on the surface of epithelial cells have been suggested to function as receptors for the attachment of Lactobacillus (Adlerberthet al,1996; Neeseret al,2000). The aim of this study was to investigate the nature of the adhesion of L. fermentum 003 to Peyer's patches and the molecular mechanism of the interaction between the bacterial cell and the Peyer's patches.

Materials and methods

Bacterial strains and growth conditions

Lactobacillus fermentum VRI 003 (PCC®; Probiomics Ltdet al, Sydney, Australia; Australian Government Analytical Laboratories NM02/31704, Sydney, Australia), L. fermentum LMG 8896 (Laboratory of Microbiology, University of Ghent, Belgium) and L. casei ATCC 393 (American Type Culture Collection, Baltimore, MD) were maintained as glycerol stocks and stored at −70°C. The Lactobacillus strains were inoculated into Mann Rogosa Sharpe (MRS) broth (Oxoid, Hampshire, UK) and incubated for 18h at 37°C anaerobically. The overnight cultures were inoculated into fresh Lactobacillus MRS broth, supplemented with [methyl-1′, 2′-3H] thymidine (124Cimmol−1; ICN Biomedicals Australia, Australia) to give a final concentration of 10μCimL−1, and incubated for 18h at 37°C. Cells were collected by centrifugation at 3000g for 15min and the bacterial pellets were washed twice in 0.1M phosphate-buffered saline (PBS; pH 7.2) and resuspended in 1mL of PBS. The bacterial cell suspension was subsequently adjusted to give an optical density of 0.5 at 600nm in 10mL of cold PBS. Saccharomyces cerevisiae UNSW 704000 (The University of New South Wales Culture Collection) was grown on peptone yeast extract glucose broth (PYG; 2% bacteriological peptone +1% yeast extract +2% glucose) at 30°C for 2 days.

Preparation of Peyer's patches

Six-week-old female specific pathogen free BALB/c mice were obtained from the Animal Research Center at the University of New South Wales. Mice were sacrificed using CO2 asphyxiation, and the entire small intestine was aseptically removed. Peyer's patches were trimmed to remove as much as possible of the visible surrounding intestinal villous tissue and washed in cold PBS. Randomly selected sections of Peyer's patches (four pieces per well) were placed into wells of a 24-well tissue culture plate (Nunc, Roskilde, Denmark). The Peyer's patches were kept on ice for not more than 2h before use in the assay. Approval for all animal experiments was obtained from the University of New South Wales animal ethics committee.

Adhesion of Lactobacillus strains to Peyer's patches

The in vitro adhesion assay was performed as described previously (Plant & Conway, 2001). Briefly, radioactively labeled Lactobacillus strains were added to each well containing Peyer's patches and incubated for 20min at 37°C on a platform shaker. The Peyer's patches were then washed three times in PBS to remove unbound and reversibly bound bacteria. Each wash step consisted of agitation of the plate for 5min on the platform shaker at room temperature. The Peyer's patches were transferred to preweighed scintillation vials, weighed and digested with perchloric acid and hydrogen peroxide at 70°C for 18h. Scintillation fluid was added to each vial, and the radioactivity was measured after 10min with a liquid scintillation counter. Lactobacillus suspensions (100μL) were also counted using the scintillation counter, and serially diluted and plated on MRS agar to determine viable count (CFU per mL). Adhesion was expressed as the number of viable bacterial cells adhering per mg wet weight of Peyer's patches tissue.

Inhibition of L. fermentum PC-003 adhesion to Peyer's patches

The inhibition of adhesion of L. fermentum 003 to Peyer's patches was determined in the presence or absence of bacterial cell surface extracts, carbohydrates and Concanavalin A (Con A). To obtain bacterial cell surface extracts, washed bacterial cell pellets were resuspended in PBS and incubated at 37°C for 1h. Following incubation, the bacterial suspension was blended with glass beads (5mm diameter) for 10min. The supernatants were harvested by centrifugation (10000g for 20min) and filtered (0.2μm membrane, Pall Gelman, MI). The residual bacteria were confirmed to be still intact and viable by direct microscopy and enumeration of CFU per mL by spread plating serially diluted aliquots. The radioactively labeled L. fermentum 003 cell suspensions were added to the wells containing the Peyer's patch tissues already mixed with or without the various concentrations of the bacterial cell surface extracts or the carbohydrates (N-acetyl-glucosamine, d-galactose, d-glucose, l-fucose, d-mannose, and methyl-α-d-mannoside) (Sigma, St Louis, MO) used at a concentration of 1, 10, and 25mM and Con A (Sigma). The adhesion assay was performed as described above.

Determination of mannose specificity for L. fermentum 003 adhesion

To determine the binding of bacterial cells to d-mannose, L. fermentum 003 cells were incubated with d-mannose coupled to bovine serum albumin (ρ-aminophenyl-α-d-mannopyranoside-BSA, Man-BSA; Sigma). Wells of a 96-well microtiter plate (Nunc) were coated with Man-BSA [50μLwell−1; 0.001–0.5μgmL−1 in 0.1M sodium carbonate buffer (pH 9.6)]. After washing the wells with PBS, non-specific binding was blocked with 1% BSA in the buffer and incubation carried out for 1h at room temperature. Following the washing as described above, 50μL aliquots of the radioactively labeled L. fermentum 003 (c. 1 × 109CFU per mL) was added to each well, and the plate incubated for 20min at 37°C. After incubation, the plate was washed three times with PBS to remove unbound and reversibly bound bacteria. Aliquots (100μL) of 2% SDS (Sigma) were added to each well to solubilize the bound material. The radioactivity in aliquots of the solubilized material was measured as described above.

Chemical and enzymatic treatments of L. fermentum 003 before adhesion assay

Lactobacillus fermentum 003 which had been washed with PBS was suspended in 0.01M sodium meta-periodate (Sigma) in 0.1M citrate-phosphate buffer (pH 4.5) and incubated at 37°C for 1h. After incubation, bacteria were washed twice and resuspended in PBS. For the enzymatic treatments of L. fermentum 003 with trypsin and proteinase K (Sigma), washed bacteria were suspended in PBS containing 2mg of each enzyme per mL or only PBS as a control, incubated at 37°C for 1h, and then washed twice to remove the enzymes and resupended in PBS. The adhesion of L. fermentum 003 to Peyer's patches was examined as described above.

Agglutination assay

The agglutination assay was performed as described previously (Adlerberthet al,1996), with slight modifications. Briefly, L. fermentum 003 was grown overnight and washed twice in PBS. Washed bacterial cells were resuspended in PBS to one-tenth of the original culture volume. This was further diluted in fourfold steps and aliquots (100μL) of the second dilution step were mixed with 100μL PBS or PBS containing methyl-α-d-mannoside at a final concentration of 25mM in round-bottomed tubes. Two hundred microlitres of 10% (v/v) S. cerevisiae cells (washed in PBS and resuspended in PBS without concentrating) was added and the tubes were shaken for 10min at room temperature. In order to determine whether the bacterial cell surface proteins are specific for the methyl-α-d-mannoside, washed bacterial cells were treated with proteinase K as described above and the agglutination assay was carried out as described above. The agglutination was examined by bright-light microscopy.

Statistical analysis

All results were shown as the mean±standard deviation (SD) of three independent experiments. Statistical differences between control group and test groups were determined using a Student's t-test.

Results and discussion

Characterization of L. fermentum 003 adhesion to Peyer's patches

Three Lactobacillus strains showed different adhesive capacity to the FAE of the Peyer's patches when expressed as the mean log CFU per mg wet weight of Peyer's patches tissue. The extent of the adhesion of Lactobacillus fermentum 003 was greater (6.0±0.3logCFU per mg) than the other two Lactobacillus strains, L. fermentum 8896 (5.1±0.5logCFU per mg) and L. casei 393 (5.4±0.4logCFU per mg), although there was no statistically significant difference in adhesion of three Lactobacillus strains with the Peyer's patches.

The adhesive capacity of L. fermentum 003 to the Peyer's patches was examined in the presence or absence of various carbohydrates used at either 1, 10 or 25mM. Only d-mannose of the six carbohydrates tested inhibited the adhesion of L. fermentum 003 to a significant degree at a concentration of 1mM using the Student's t-test (P<0.05). In contrast, the other four carbohydrates, N-acetyl-glucosamine, d-galactose, d-glucose and L-fucose had no effect on the adhesion. Methyl-α-d-mannoside (1mM) reduced adhesion of L. fermentum 003 but was not significant (P>0.05). At the higher concentration tested (10mM), L. fermentum 003 adhesion to Peyer's patches was significantly inhibited by d-mannose and methyl-α-d-mannoside (P<0.05) while N-acetyl-glucosamine, d-galactose, d-glucose and L-fucose did not inhibit the adhesion. A similar pattern of adhesion was observed at the concentration of 25mM d-mannose and methyl-α-d-mannoside (P<0.01 and P<0.05, respectively). However, no significant effect of 25mM N-acetyl-glucosamine, d-galactose, d-glucose, and L-fucose was observed on the adhesion to Peyer's patches. The result showed that d-mannose and methyl-α-d-mannoside significantly reduced the adhesive capacity of L. fermentum 003. The inhibition of L. fermentum 003 adhesion in the presence of d-mannose or methyl-α-d-mannoside suggests that L. fermentum 003 possesses mannose-specific molecules which mediate adhesion to Peyer's patches. The fact that the inhibition appears to be dependent on the concentration of mannose-containing molecules supports the hypothesis that adhesive molecules of the bacterial cell attach to the mannose-containing molecules added in the assay.

The adhesion of L. fermentum 003 was not blocked by the presence of other carbohydrates such as N-acetyl-glucosamine, d-galactose, d-glucose and L-fucose. It indicates that a specific receptor for L. fermentum 003 adhesion may be present on the FAE of the Peyer's patches. Considering that L. fermentum 003 was specific for mannose-containing molecules, a binding assay with d-mannose coupled to BSA (Man-BSA) was performed. As shown in Fig. 1, the binding of L. fermentum 003 was concentration dependent. Lactobacillus fermentum 003 bound to the Man-BSA with negligible binding to the BSA control without d-mannose. The number of bacterial cells binding increased as the amount of mannose increased until saturation. This finding is consistent with that described for the adhesion of L. plantarum 299 and 299v to HT-29 cells. A mannose-specific adhesion mechanism in L. plantarum 299 and 299v has been confirmed (Adlerberthet al,1996). These workers showed that d-mannose and methyl-α-d-mannoside inhibited the adhesion of bacteria to HT-29 cells and these strains bound directly to d-mannose immobilized on agarose beads. A similar mannose-specific adhesive mechanism has been suggested for L. johnsonii La1 to Caco-2 cells (Neeseret al,2000). The inhibition of adhesion of L. johnsonii La1 was observed in the presence of O-glycosylated yeast mannoprotein, but methyl-α-d-mannoside was not able to inhibit adhesion of this strain to Caco-2 cells. Other workers have identified a gene of L. plantarum that encoded for a mannose-specific adhesion (Pretzeret al,2005). Mannose-specific binding of a variety of Gram-negative bacteria, including Escherichia coli and Salmonella typhimurium, has also been reported (Woldet al,1988; Oyofoet al,1989) and may provide a link between the inhibition by lactobacilli of such pathogens with mannose-specific adhesins in vivo. The pretreatment of Peyer's patches with Con A, a lectin with specificity for d-glucose, d-mannose, d-fructose or d-arabinose, did not lead to a change in L. fermentum 003 adhesion to Peyer's patches (data not shown). This finding demonstrates that these bacteria do not compete with Con A for the same receptor(s), mannose-containing molecules, on the FAE of Peyer's patches, as Con A recognizes complex carbohydrate molecules of glycoconjugates. It seems likely that a fine carbohydrate specificity, which is a mannose-specific adhesion mechanism, is involved in L. fermentum 003 adhesion to Peyer's patches.

1

Adhesion of Lactobacillus fermentum 003 to increasing concentration of d-mannose-BSA (Man-BSA) immobilized in microtiter wells. Adhering bacteria expressed as the disintegrations per minute (DPM) of the radioactive label incorporated into the bacterial cells. The vertical bars represent the standard deviation.

Effect of enzymatic and chemical pretreatment of L. fermentum 003 on the adhesion

Lactobacillus fermentum 003 was subjected to a variety of enzymatic and chemical treatments in order to characterize and compare the adhesive determinants involved in the adhesion to Peyer's patches (Table 1). First, the contribution of protein factors was examined by treating L. fermentum 003 with proteinase K and trypsin. The contribution of carbohydrate moieties was also examined by oxidizing cell surface carbohydrates with sodium meta-periodate. The pretreatment of L. fermentum 003 with proteinase K and trypsin diminished the adhesion to Peyer's patches (34% and 49% of adhesion index compared with 100% as control, respectively). Pretreatment with proteinase K significantly reduced L. fermentum 003 adhesion (P<0.05) and trypsin treatment moderately decreased the adhesive ability of L. fermentum 003, indicating that proteinaceous components of bacterial cells are involved in the adhesion to Peyer's patches. These results reflect that the adhesion of L. fermentum 003 is mediated by a proteinaceous component that is more sensitive to proteinase K than to trypsin.

View this table:
1

Effects of chemical and enzymatic treatments of Lactobacillus fermentum 003 on the adhesion to Peyer's patches

TreatmentAdhesion index (%)
Control (phosphate-buffered saline)100 ± 10.1
Control (0.1M citrate phosphate buffer)100 ± 0.8
Proteinase K33.5 ± 5.7
Trypsin48.7 ± 5.2
Sodium iodate (control)108.6 ± 2.8
Sodium meta-periodate224.1 ± 21.7
  • * The adhesion of bacterial cells to Peyer's patch tissues in the absence of any treatment was taken as 100% adhesion; Mean ± SD.

  • Mean values were significantly different from the control (P<0.05) using Student's t-test.

The cell surface extracts of L. fermentum 003, which presumably interact with mannose-like molecules, inhibited the adhesion to Peyer's patches (Table 2). As the cells were intact after the extraction process when examined by direct microscopy and still viable as shown by unchanged CFU count, the cell extract would have been relatively free of cell wall anchored and intracellular molecules. This confirms that cell surface components play a major role in the specific binding to Peyer's patches and that as protease reduced the binding capacity of L. fermentum 003, it is most probably proteinaceous components that are involved in binding to Peyer's patches. In most cases, bacterial cell surface components have been shown to contribute to the adhesive capacity of lactobacilli (Henriksson & Conway, 1992; Greeneet al,1994; Kapczynskiet al,2000; Tuomolaet al,2000; Lorcaet al,2002). It was suggested that the binding capacity of L. acidophilus strains adherence to collagen and fibronectin was lost after treatment with proteases (Lorcaet al,2002). In addition, pretreatment with proteases has been found to reduce the adhesion of lactobacilli to intestinal epithelium. Pepsin and trypsin treatment of L. acidophilus 1 and L. rhamnosus GG before adhesion assay resulted in a decrease in adhesion to human intestinal glycoproteins (Tuomolaet al,2000).

View this table:
2

Effect of bacterial cell surface extracts on adhesion of Lactobacillus fermentum 003 to Peyer's patches

Bacterial cell surface extracts (μg proteinsmL−1)Adhesion index (%)
Control100 ± 9.2
548.9 ± 18.7
2546.7 ± 15.3
5042.7 ± 12.8
  • * For the bacterial cell surface extraction, cells were incubated in phosphate-buffered saline at 37°C for 1h and blended with glass beads for 10min. The supernatants were harvested by centrifugation and filtered. The bacterial cell surface extracts were used in the adhesion assay described in the Materials and methods. Protein concentrations were determined by using the Bio-Rad protein assay method according to the manufacturer's instruction.

  • The adhesion of bacteria to Peyer's patch tissues in the absence of cell surface extracts was taken as 100% adhesion (Control); Mean ± SD.

  • Mean values were significantly different from the control (P<0.05) using Student's t-test.

Pretreatment of L. fermentum 003 with sodium meta-periodate resulted in a large increase in the adhesion to Peyer's patches (P<0.05) compared with the pretreatment with PBS and an additional control, in which sodium iodate was used at an equivalent concentration. This finding is most likely due to the oxidation of the extracellular polysaccharide (EPS) produced by L. fermentum 003, suggesting that EPS may prevent adhesion to Peyer's patches by steric hindrance of the adhesive components. Oxidation of EPS may have resulted in exposure of proteinaceous structures on the cell surface that interact with carbohydrate moieties. It has been demonstrated that oxidation of EPS promotes adhesive abilities of various lactobacilli (Lorcaet al,2002). Meta-periodate oxidation of L. casei CRL 431 promoted the adhesion to ileal epithelial cells (Morata de Ambrosiniet al,1999). However, the treatment of meta-periodate did not have any effect on the adhesion of some Lactobacillus strains including L. rhamnosus GG and L. acidophilus 1 to mucus glycoproteins (Tuomolaet al,2000). Moreover, others have showed that treatment of Lactobacillus strains with meta-periodate resulted in a decrease in adhesion to Caco-2 cells (Greeneet al,1994). Hence, the different adhesive mechanisms are dependant on particular Lactobacillus strains tested, reflecting the fact that complex adhesive factors are involved and the epithelial surface such as Caco-2 or HT-29 cell lines, mucus or Peyer's patches. Different bacterial strains will rely on multiple mechanisms involving diverse cell surface structures for intestinal colonization of the various components of the tract.

To confirm that the mannose-binding moiety of the L. fermentum was proteinaceous, the agglutination assay using Saccharomyces cerevisiae was utilized. The agglutination of S. cerevisiae was induced by L. fermentum 003 in the absence of 25mM methyl-α-d-mannoside, however in the presence of the methyl-α-d-mannoside, no agglutination occurred, indicating that the agglutination of S. cerevisiae was mediated by the interaction of adhesive factor(s) of L. fermentum 003 with mannose residues on the yeast cells. The cell wall of S. cerevisiae contains mannose-containing polysaccharides and E. coli with the mannose-specific adhesin agglutinates S. cerevisiae in a mannose-specific manner (Ofeket al,1977). The agglutination was not observed following the treatment of L. fermentum 003 with proteinase K for 1h at 37°C in the absence or presence of methyl-α-d-mannoside, confirming strongly that the bacterial cell surface proteins were involved in the mannose-specific adherence of L. fermentum 003.

In this study, the adhesion of L. fermentum 003 to Peyer's patches is shown to be mediated by a proteinaceous component(s) that binds to mannose-containing molecules. Thus, the mannose-specific protein can be implicated as an essential factor in adhesion of L. fermentum 003 to Peyer's patches and may play a role in preventing the attachment of pathogens such as E. coli.

Acknowledgements

This work was supported by Probiomics Ltdet al, Sydney, Australia.

References

View Abstract