Lactobacillus rhamnosus GG inhibits invasion of cultured human respiratory cells by prtF1-positive macrolide-resistant group A streptococci


Bruna Facinelli, Institute of Microbiology and Biomedical Sciences, Polytechnic University of Marche Medical School, Ancona, Italy, Via Tronto 10/A – 60020 Ancona, Italy. E-mail:


Aims:  This study was designed to determine whether the probiotic strain Lactobacillus GG, which is extensively used in the treatment and prevention of intestinal disorders, is able to inhibit invasion of cultured human respiratory cells by macrolide-resistant group A streptococci (GAS) carrying the prtF1 gene, which encodes the fibronectin (Fn)-binding invasin F1.

Methods and Results:  Eight prtF1-positive erythromycin-resistant GAS strains were used to infect A549 monolayers in competition and displacement assays with Lactobacillus GG. Live (L-LGG) and heat-killed (HK-LGG) lactobacilli and their spent culture supernatant (SCS) significantly reduced (< 0·001) GAS invasion efficiency in both assays. No antibacterial activity of Lactobacillus GG against GAS was detected. Both L-LGG and HK-LGG and all prtF1-positive GAS induced a strong agglutination reaction using Fn-coated particles.

Conclusions: Lactobacillus GG exerts an antagonistic action against GAS by inhibiting cell invasion. Competitive binding of Lactobacillus GG and GAS to Fn might be involved in the inhibition process.

Significance and Impact of the Study:  The finding that Lactobacillus GG can prevent in vitro invasion of respiratory cells by GAS suggests new applications for this probiotic strain and warrants further studies of its capacity to prevent GAS throat infections.


Probiotics are living micro-organisms that, when ingested in adequate amounts, exert health benefits towards the host (FAO/WHO 2002). Most probiotics belong to the genus Lactobacillus, whose species are widely distributed in the gastrointestinal (GI), genitourinary and oral tracts. Lactobacillus rhamnosus GG (Lactobacillus GG), originally isolated from healthy human GI tract, is by far one of the most important probiotic strains. Lactobacillus GG is especially active in the GI tract and is extensively used in the treatment and prevention of a variety of intestinal disorders (Gorbach 2000). Several mechanisms have been suggested to contribute to the probiotic action of Lactobacillus GG in the GI tract, including competition with pathogens for binding sites and production of antimicrobial components (Servin 2004). An antagonistic activity of Lactobacillus GG against human intestinal cell invasion by Salmonella typhimurium has also been reported (Hudault et al. 1997).

Group A streptococci [(GAS), Streptococcus pyogenes] are the principal aetiologic agents of bacterial pharyngotonsillitis in children; they also cause a variety of diseases ranging from mild, self-limiting to life-threatening infections and poststreptococcal sequelae (Cunningham 2000). GAS can be efficiently internalized by, and survive within, human respiratory cells (Cleary and Cue 2000); a fibronectin (Fn)-binding invasin [protein F1 (prtF1)], is required for efficient entry into epithelial cells (Okada et al. 1998; Molinari and Chhatwal 1999). A significant association between erythromycin resistance and cell invasiveness has been documented in Italy in GAS isolated from children with pharyngotonsillitis (Facinelli et al. 2001). GAS strains combining macrolide resistance and cell invasiveness may be able to escape penicillin and other β-lactams due to intracellular location and macrolides due to resistance, resulting in failure of eradication and persistent throat carriage (Facinelli et al. 2001).

The global increase of bacterial resistance along with a renewed interest in alternative methods to prevent infections makes probiotics an interesting research field (Reid et al. 2003). This study was designed to determine whether Lactobacillus GG is able to inhibit invasion of cultured human respiratory cells by highly invasive, macrolide-resistant GAS.

Materials and methods

Bacterial strains and culture conditions

Eight prtF1-positive GAS strains isolated from children with pharyngotonsillitis and Lactobacillus GG (ATCC 53103) were used. The GAS strains represented different clones recently identified in Italy: all were macrolide-resistant [erythromycin minimum inhibitory concentration, ≥1 μg ml−1); erm(A), erm(B), and/or mef(A)], and highly invasive (>10% of the inoculum) for human epithelial respiratory (A549) cells (Facinelli et al. 2001; Spinaci et al. 2004, 2006). The characteristics of prtF1-positive GAS are listed in Table 1. Enterococcus faecalis ATCC 29212 and a prtF1-negative GAS strain (Facinelli et al. 2001) were used as controls.

Table 1.   Characteristics of prtF1-positive GAS strains and RR of cell invasion (95% CIs) in competition and displacement assays with L-LGG
StrainGenotype/phenotype of Ery resistanceRR of GAS cell invasion (95% CIs) in assays with L-LGG
  1. *Strains SP1188, SP1900 and SP1951 are representatives of major clones detected in Italy (Spinaci et al. 2004).

  2. †SP1900 belongs to clone IMC-77 [erm(A)/iMLS-B tet(O) emm77; Sequence Type (ST) 369], recently detected also in Norway (Palmieri et al. 2006).

  3. GAS, group A streptococci; RR, relative risk; L-LGG, live Lactobacillus rhamnosus GG.

SP9707erm(B)mef(A)/cMLS0·051 (0·047–0·055)0·0035 (0·0034–0·0037)
SP1003erm(B)/cMLS0·046 (0·044–0·047)0·882 (0·876–0·889)
SP114erm(B) mef(A)/iMLS-A0·035 (0·034–0·036)0·542 (0·538–0·545)
SP1188*erm(B)/iMLS-A0·135 (0·133–0·138)0·507 (0·504–0·510)
SP1900*†erm(A)/iMLS-B0·085 (0·084–0·086)0·305 (0·303–0·307)
SP1161erm(A)/iMLS-C0·080 (0·079–0·081)0·622 (0·620–0·625)
SP1951*mef(A)/M0·017 (0·0168–0·0173)0·079 (0·076–0·082)
SP1013mef(A)/M0·047 (0·044–0·049)0·286 (0·279–0·293)

Streptococci were routinely grown in blood agar base (Oxoid, Basingstoke, UK) supplemented with 5% defibrinated sheep blood and Brain Heart (BH; Oxoid) broth and agar. Streptococci were maintained in glycerol at −70°C and subcultured twice on blood agar before testing. Lactobacillus GG was grown in de Man, Rogosa and Sharpe (MRS; Merck, Darmstadt, Germany) broth and agar at 37°C in presence of 5% CO2. For preparation of spent culture supernatant (SCS), lactobacilli were grown overnight in RPMI-1640 (Euroclone, West York, UK) in 5% CO2 at 37°C; then, SCS (pH 6·2) was sterilized through a sterile 0·22-μm pore size filter unit (Millipore, Molsheim, France). Heat-killed Lactobacillus GG (HK-LGG) was prepared by growing lactobacilli in MRS broth, followed by boiling (2 h), harvesting and resuspension in RPMI-1640.

Detection of antimicrobial activity

The antimicrobial activity of Lactobacillus GG against GAS was examined by the agar spot test (Jacobsen et al. 1999). In other experiments, after overnight growth, streptococci were washed once with PBS, resuspended in SCS (final counts approx. 1 × 107 CFU ml−1) and incubated in 5% CO2 at 37°C. At fixed times of incubation, aliquots were removed, serially diluted and plated on BH agar to determine the number of CFU. Control experiments were performed by incubating the same amount of streptococci with BH broth and RPMI-1640. All experiments were also performed with E. faecalis ATCC 29212 as control. All assays were performed twice.

Invasion inhibition assays

A549 cells (ATCC CCL 185) were used, as previously described (Facinelli et al. 2001). Cells were routinely grown in 25-cm2 plastic tissue-culture flasks (Corning Costar, Milan, Italy) at 37°C in 5% CO2. The culture medium was RPMI-1640 supplemented with 10% foetal calf serum (FCS; Euroclone) and containing penicillin (5 μg ml−1) and streptomycin (100 μg ml−1). GAS cell invasion efficiency was established as described previously (Facinelli et al. 2001). GAS inocula to be used in monolayer infection assays were prepared by resuspending overnight grown streptococci [iMLS strains in presence of 1 μg ml−1 erythromycin (Sigma, St Louis, MO, USA) to induce resistance] in RPMI-1640 with FCS; they were then incubated for 1 h at 37°C and resuspended in RPMI-1640 (1–3 × 105 CFU ml−1). Lactobacillus GG inoculum was prepared by resuspending overnight grown lactobacilli in RPMI-1640 (1–3 × 109 CFU ml−1). Previously described competition and displacement assays were used (Forestier et al. 2001). In competition assays, A549 cells were infected with: (i) a 1 : 1 mixture containing GAS and live lactobacilli (L-LGG); (ii) a 1 : 1 mixture containing streptococci and HK-LGG or (iii) SCS-resuspended GAS. In displacement assays, A549 cells were infected with streptococci, then washed and incubated with L-LGG, HK-LGG or SCS. Control experiments were run; in experiments with SCS, GAS were resuspended in RPMI-1640 adjusted to pH 6·2.

In each assay, infected monolayers were incubated in 5% CO2 for 1 h at 37°C, washed, and then covered with RPMI-1640 containing penicillin (5 μg ml−1) and gentamicin (100 μg ml−1) (antibiotic protection assay) (Facinelli et al. 2001). After 2 h, cells were extensively washed and lysed in cold distilled water. The CFU of intracellular GAS were counted after suitable dilutions of the lysates were plated on BH agar supplemented with erythromycin (1 μg ml−1), to inhibit lactobacilli and incubated for 36–48 h at 37°C. Bacteria were tested in three separate assays on different days; each assay represented the average of triplicate wells.

Fn-binding assays

Evaluation of Lactobacillus GG and GAS binding to Fn was done essentially as described (Naidu et al. 1988; Štyriak et al. 1999). Briefly, 1 ml of latex bead suspension (diameter = 0·8 μm; Sigma) was mixed with 3 ml of 0·17 mol l−1 glycine-NaOH (gly) buffer (pH 8·2) and centrifuged at 4500 g for 5 min; the pellet was resuspended in 3 ml of the same buffer. Bovine Fn (100 μg; Sigma) was added and the mixture was kept at 30°C overnight on a horizontal shaker at 50 rev min−1. The mixture was centrifuged at 9200 g for 5 min at 20°C; the beads were resuspended in 2 ml of gly buffer containing BSA 0·05% and merthiolate 0·01% and kept at 4°C for 12 h. Fn-coated latex beads (20 μl) were mixed on a glass slide with an equal volume of a bacterial cell suspension of 1010 CFU ml−1 in PBS. In some experiments, Fn-coated latex beads were mixed with bacterial suspension pretreated with Fn (100 μg ml−1) for 1 h at room temperature. The agglutination reaction was read after 2 min. Strains were checked for autoaggregation by mixing bacterial cell suspension with gly buffer. A negative control was performed by mixing Fn-coated latex beads with PBS.

Statistical analysis

In competition and displacement experiments, the proportions of intracellular streptococci were estimated both in presence (pL) and absence (pC, control) of lactobacilli (L-LGG, HK-LGG or SCS). The effect on GAS invasion efficiency compared with control cultures was quantified according to the relative risk (RR) of invasion (RR = pL/pC) and asymptotic 95% confidence intervals (CIs) (Altman 1991). The RR of invasion is the ratio of the probability of cell invasion occurring in presence vs absence of lactobacilli. RR = 1 indicates that there is no difference in invasion efficiency between assays performed in presence and absence of lactobacilli; RR < 1 means that cell invasion is less likely to occur in presence than in absence, whereas RR > 1 means that cell invasion is more likely to occur in presence than in absence. The natural logarithm of RR has an approximately normal distribution; this property was used to test the hypothesis that inhibition was equal in competition/displacement experiments vs the hypothesis that it was less pronounced in the control assays.


Inhibition of GAS invasion

The cell invasion efficiency of GAS was first established in an antibiotic protection assay. In these experiments, the percentage of intracellular streptococci with respect to initial inoculum ranged from 11·2% (SP1900) to 35·1% (SP1003) (Fig. 1). In inhibition experiments, the invasion efficiency of all GAS in presence of L-LGG, quantified according to the RR of invasion, was significantly reduced (< 0·001) compared with control cultures. In particular, it ranged from RR = 0·017 (SP1951) to RR = 0·135 (SP1188) in competition experiments, and from RR = 0·0035 (SP9707) to RR = 0·882 (SP1003) in displacement assays (Table 1). Strain SP1900 was chosen to test the inhibition ability of SCS and HK-LGG. Invasion of SP1900 compared with control cultures was significantly reduced (< 0·001) in both competition and displacement assays, the respective RR of invasion being 0·3860 (CI 0·3828–0·3892) and 0·4970 (CI 0·4905–0·5036) in presence of SCS and 0·4099 (CI 0·4066–0·4132) and 0·4552 (CI 0·4491–0·4614) in presence of HK-LGG.

Figure 1.

 Entry of GAS strains into A549 cells. Data are expressed as % (of initial inoculum) of viable bacteria recovered after 1 h incubation with penicillin (5 μg ml−1) and gentamicin (100 μg ml−1). Each column is the mean (±SD) of three independent experiments performed in triplicate.

Lack of antimicrobial activity of Lactobacillus GG against GAS

No inhibition was demonstrated against any of the eight GAS strains in the agar spot test, whereas a clear inhibition zone (>1 mm) was observed with E. faecalis ATCC 29212 (not shown). When SP 1900 and E. faecalis ATCC 29212 were cultured for 6 h in SCS, growth inhibition was observed only in enterococci (not shown).

Agglutination of L-LGG, HK-LGG and GAS by Fn-coated particles

When L-LGG, HK-LGG and prtF1-positive GAS were mixed with Fn-coated particles, all strains showed an identical strong agglutination reaction (Fig. 2). A weak reaction was noted with the prtF1-negative strain. Autoaggregation was not detected. Pre-incubation of bacteria with Fn prevented the agglutination reaction (data not shown), indicating the specifity of the assay.

Figure 2.

 Strain agglutination by Fn-coated particles. (a) Positive, SP1900; (b) positive, L-LGG; (c) negative, control.


Our data demonstrate that Lactobacillus GG significantly inhibits invasion of cultured respiratory cells by prtF1-positive, macrolide-resistant GAS, both in competition and in displacement assays. Similar results have previously been reported by Hudault et al., who described an antagonistic activity exerted by Lactobacillus GG against S. typhimurium intestinal cells infection. L-LGG significantly reduced the RR of invasion of A549 cells by all GAS tested, irrespective of their genotype/phenotype of macrolide resistance. Inhibition was uniformly high in competition assays, the RR of invasion of all, but one GAS being ≤0·085. These results suggest that Lactobacillus GG is able to compete with GAS during co-infection. In displacement experiments, the RR of invasion of all, but two GAS (i.e. SP9707 and SP1951) was much higher than in competition experiments; interestingly, a RR > 0·5 was displayed by strains with invasion efficiency ≥30%. These findings suggest that the ability of Lactobacillus GG to displace GAS is inversely related to their invasion efficiency. Invasion was also significantly inhibited both by killed Lactobacillus GG and its supernatant, although less efficiently compared with L-LGG.

The lack of detectable antibacterial activity of Lactobacillus GG against GAS points to the involvement of competition for binding sites in the inhibition of invasion. Lactobacilli, including Lactobacillus GG, express cell surface proteins, also secreted in the supernatant, which mediate binding to extracellular matrix molecules such as Fn (Lorca et al. 2002). Given that in GAS invasion Fn functions as a bridging molecule between Fn-binding proteins and the cell receptor (Ozeri et al. 1998), competitive binding of Lactobacillus GG and GAS for Fn might be involved in the inhibition process. This hypothesis was explored by comparing the ability of Lactobacillus GG (L-LGG and HK-LGG) and of prtF1-positive GAS to bind to Fn-coated particles. A strong, specific, agglutination reaction was observed with all strains, demonstrating that Lactobacillus GG is able to bind Fn as well as prtF1-positive GAS. These findings seem to support the hypothesis of competitive binding of Lactobacillus GG and GAS for Fn in the inhibition process.

To our knowledge, this is the first report describing the antagonistic action of Lactobacillus GG against GAS. Further work is required to understand the exact mechanism(s) of this activity. The finding that Lactobacillus GG can prevent in vitro invasion of respiratory cells by GAS also suggests new applications for this probiotic strain and warrants further studies of its ability to prevent GAS throat infections.


This work was supported in part by a grant from the Italian Ministry of Education, University and Research.