Evaluation of characteristics of Pediococcus spp. to be used as a vaginal probiotic


  • S. Borges,

    1. CBQF – Centro de Biotecnologia e Química Fina, Escola Superior de Biotecnologia, Centro Regional do Porto da Universidade Católica Portuguesa, Porto, Portugal
    Search for more papers by this author
  • J. Barbosa,

    1. CBQF – Centro de Biotecnologia e Química Fina, Escola Superior de Biotecnologia, Centro Regional do Porto da Universidade Católica Portuguesa, Porto, Portugal
    Search for more papers by this author
  • J. Silva,

    1. CBQF – Centro de Biotecnologia e Química Fina, Escola Superior de Biotecnologia, Centro Regional do Porto da Universidade Católica Portuguesa, Porto, Portugal
    Search for more papers by this author
  • P. Teixeira

    Corresponding author
    • CBQF – Centro de Biotecnologia e Química Fina, Escola Superior de Biotecnologia, Centro Regional do Porto da Universidade Católica Portuguesa, Porto, Portugal
    Search for more papers by this author


Paula Teixeira, CBQF – Centro de Biotecnologia e Química Fina, Escola Superior de Biotecnologia, Centro Regional do Porto da Universidade Católica Portuguesa, Rua Dr. António Bernardino Almeida, 4200-072 Porto, Portugal.

E-mail: pcteixeira@porto.ucp.pt



The aim of our research was to select, identify and characterize an isolate of lactic acid bacteria to be considered as a vaginal probiotic.

Methods and Results

Thirty-five isolates of Pediococcus spp. showed bacteriocinogenic activity against Listeria monocytogenes and the ability to survive in simulated vaginal fluid (SVF) at pH 4·2. One isolate of Pediococcus spp. was selected and characterized to evaluate its safety before the use as vaginal probiotic. Pediococcus pentosaceus SB83 did not show the presence of virulence factors such as the production of gelatinase, lipase and DNase, haemolytic activity, nor the presence of virulence genes (genes esp, agg, gelE, efaAfm, efaAfs, cylA, cylB and cylM). Pediococcus pentosaceus SB83 was considered sensitive to chloramphenicol, gentamicin, streptomycin, kanamycin, erythromycin and ampicillin. This strain was considered resistant to tetracycline and vancomycin. Pediococcus pentosaceus SB83 was a biofilm producer at different pH values (4·2, 5·5 and 6·5) in SVF and in de Man, Rogosa and Sharpe medium.


The in vitro results provide a basis for the use of P. pentosaceus SB83 as a vaginal probiotic, to prevent colonization by L. monocytogenes in pregnant women.

Significance and Impact of Study

The application of vaginal probiotics could have the potential for preventing vaginal infections and consequently reduce abortion and neonatal infections.


The normal vaginal microbiota is dominated by lactobacilli, at 107–108 CFU ml−1 of vaginal fluid in healthy premenopausal women (Boris and Barbés 2000; Dicks et al. 2000; Bolton et al. 2008). Lactobacilli contribute to the prevention of genital infections and play a role in the maintenance of a healthy state. The capacity that lactobacilli have to adhere and compete for adhesion sites in the vaginal epithelium (Zárate and Nader-Macias 2006; Coudeyras et al. 2008) and the capacity to produce antimicrobial compounds, for example hydrogen peroxide (Aslim and Kilic 2006; Wasiela et al. 2008), lactic acid (Fraga et al. 2008), bacteriocin-like substances (Aroutcheva et al. 2001; Dover et al. 2007; Turovskiy et al. 2009) and biosurfactants, are important in the impairment of colonization by pathogens. Furthermore, the production of lactic acid may help to maintain a low vaginal pH, approximately 4–4·5, which makes the vaginal environment more conducive to lactobacilli growth (Boris and Barbés 2000; Reid and Bocking 2003; Bolton et al. 2008). Low vaginal pH is also beneficial for other antimicrobials; H2O2 is stable in these conditions and bacteriocins are highly active. An increase in the vaginal pH leads to a decrease in the lactobacilli-associated antimicrobial activity (Dover et al. 2008).

Maternal urinary tract infections increase the risk of puerperal sepsis and the associated risk of neonatal sepsis (Ganatra and Zaidi 2010). In pregnancy, infections can be transmitted from mother to child causing adverse sequelae, such as abortion, stillbirth, neonatal infection. Infections of the newborn may be acquired in utero (congenital), around the time of delivery (intrapartum infection) or in the neonatal period (postpartum infection).

There is often a loss of colonization by lactobacilli due to antibiotic therapy, douching, sexual activity, hormone deficiency and contraceptive measures (Barrons and Tassone 2008; Bolton et al. 2008). The selection and use of ‘vaginal probiotics’ can be important to restore a healthy vaginal microbiota (Pascual et al. 2008).

Probiotics have been defined as ‘live microorganisms, which when administered in adequate amounts, confer a health benefit on the host’ (World Health Organization 2005). Lactic acid bacteria (LAB) may play a major role in preventing illness of the host, including bacterial vaginosis, yeast vaginitis, urinary tract infection and sexually transmitted diseases. The administration of probiotics by mouth and intravaginally has been shown to be safe, and for pregnant women, this restoration could be important to lower the risk of preterm labour (Reid and Bocking 2003; Wilks et al. 2004). The capacity of LAB to colonize the vaginal mucosa depends on the route of delivery; oral formulations must be capable of maintaining their structural integrity (viability) during passage through the gastrointestinal tract and delivery to the rectal area for ascension and colonization of the vaginal tract (Barrons and Tassone 2008).

There are more than 80 known species of lactobacilli in the intestines and vagina; therefore, in terms of probiotic effects, individual species may differ in their ability to restore normal microbiota and regulate the overgrowth of pathogens (Barrons and Tassone 2008).

Vaginal colonization by pathogens can result in transmission to the foetus/neonate by vertical transmission. Therefore, the vaginal application of LAB could be a preventative strategy to reduce the global burden of neonatal infections.

The aim of this study was to select, identify and characterize an isolate of LAB to be considered as a vaginal probiotic candidate.

Materials and methods


Lactic acid bacteria isolates

Fifty-six food isolates (Escola Superior de Biotecnologia culture collection) and 19 vaginal isolates (Hospital São Marcos; Braga, Portugal) of lactic acid bacteria were selected for this study. Stock cultures were maintained in de Man, Rogosa and Sharpe (MRS) broth (Lab M; Bury, UK) supplemented with 30% (v/v) of glycerol (Panreac; Barcelona, Spain) and stored at −80°C. Cultures were recovered on MRS agar (Lab M) and were subcultured twice in MRS broth at 37°C for 24 h before use in all tests.

Pathogenic strains

As target bacteria for the inhibitory effects of LAB, 29 clinical isolates of L. monocytogenes from the Listeria culture collection of Escola Superior de Biotecnologia (Table 1) and 10 vaginal isolates of Group B Streptococcus (GBS, Hospital São Marcos; Braga, Portugal) were used. Cultures were grown on tryptone soya agar with yeast extract (Lab M) 0·6% (w/v) (TSA-YE) and were subcultured in tryptone soya broth with yeast extract (TSB-YE, Lab M) at 37°C for 24 h. Stock cultures were maintained in TSB-YE containing 30% (v/v) of glycerol and stored at −80°C.

Table 1. Listeria monocytogenes isolates used in this study
IsolateSerotypeIsolated from
9991/2bCerebrospinal fluid
17611/2aCerebrospinal fluid
18911/2aCerebrospinal fluid
20744bBlood (mother and neonate)
2085/14bCerebrospinal fluid
2085/24bCerebrospinal fluid
20911/2aCerebrospinal fluid
20991/2aCerebrospinal fluid

More than 95% of infections by L. monocytogenes in humans are caused by serotypes 1/2a, 1/2b and 4b, and the majority of listeriosis outbreaks are caused by strains of serotype 4b (Swaminathan and Gerner-Smidt 2007). Therefore, in this study, we used clinical isolates of L. monocytogenes of these three different serotypes.

Screening for antimicrobial activity of LAB

The antimicrobial activities of LAB isolates were tested against pathogenic bacteria by using the agar-spot test method. TSA-YE plates were spread with overnight suspensions of each of the target bacteria, and drops (10 μl) of overnight cultures of the LAB isolates were spotted on the lawns of pathogens (L. monocytogenes and GBS) and incubated for 24 h at 37°C. Translucent halo zones observed around the spots were registered as positive for antimicrobial activity. For the positive strains, in order to determine the nature of inhibition, the pH of the supernatant, obtained by centrifugation at 8877 × g, 10 min, 4°C (Rotina 35R; Hettich, Germany), was adjusted to six with sterile 1 mol l−1 NaOH (Pronalab; Lisbon, Portugal) and then treated with catalase (0·1 mg ml−1; Sigma-Aldrich; Steinheim, Germany) and trypsin (0·1 mg ml−1; Sigma-Aldrich). After each of these treatments, supernatants were spotted against target bacteria (L. monocytogenes and GBS). As a control, Pediococcus acidilactici HA-6111-2 was used (Albano et al. 2007).

Quantification of antibacterial activity

For those LAB strains that showed antimicrobial activity by bacteriocin production (i.e. no reduction in activity after pH adjustment and treatment with catalase, but total loss of activity after trypsin treatment), this activity was quantified. Antimicrobial activity was expressed as arbitrary units (AU) per ml. AU is defined as the reciprocal of the highest serial twofold dilution showing a clear zone of growth inhibition of the target strains (Van Reenen et al. 1998). Six isolates of L. monocytogenes of different serotypes (1/2b, 4b and 1/2a) were used as target strains.

Identification of bacteriocin-producing lactic acid bacteria

Total (genomic and plasmid) DNA isolation was performed using a Gen Elute Bacterial Genomic DNA kit (Sigma-Aldrich). Amplification of the 16S rDNA was carried out with the primers 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) (MWG Biotech AG; Ebersberg, Germany) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) (MWG Biotech AG). Amplification reactions were performed in a Thermocycler (MyCycler™, Thermocycler Firmware, Bio-Rad; Richmond, CA, USA) as follows: initial denaturation at 95°C for 5 min, 30 cycles of 1 min at 94°C and 1 min at 55°C, followed by an increase to 72°C over 1·5 min. Extension of the amplified product was at 72°C for 10 min. Following amplification, 5 μl of product was separated at 90 V for 50 min in a 1% (w/v) agarose gel in 1× TAE buffer (4·84 g Tris base, 1·09 g glacial acetic acid, 0·29 g ethylenediaminetetraacetic acid, 1 l distilled water) and then stained with 0·5 μg ml−1 of ethidium bromide. A 100-bp DNA ladder (Bio-Rad) was used as a molecular weight marker. PCR products, used as templates, were then purified with the GFX PCR DNA and Band Purification kit (GE HealthCare, Amersham Biosciences; Amersham, UK) and sent for sequencing. Sequences obtained were aligned with sequences in GenBank using the BLAST program.

Random amplified polymorphic DNA (RAPD)

The genetic heterogeneity of isolates was determined by numerical analysis of DNA profiles obtained by RAPD-PCR.

DNA primers M13 (5′-GAG GGT GGC GGT TCT-3′) (MWG Biotech AG) and D8635 (5′-GAG CGG CCA AAG GGA GCA GAC-3′) (MWG Biotech AG) were used. RAPD-PCR was performed on total DNA, according to Andrighetto et al. (2001). The 25-μl reaction volume composed of the primer M13: 0·99 pmol l−1 primer, 150 mmol l−1 of dNTPs (ABGene; Surrey, UK), 1 × PCR buffer (MBI Fermentas; Mundolsheim, France), 2·5 mmol l−1 of MgCl2 (MBI Fermentas), 2 U of Taq DNA polymerase (MBI Fermentas) and 1 μl of extracted solution of DNA. For the primer D8635, the mixture contained 0·88 pmol l−1 of primer, 200 mmol l−1 of dNTPs, 1 × PCR buffer, 2·5 mmol l−1 of MgCl2, 2 U Taq polymerase and 1 μl of extracted solution of DNA.

Amplification was performed in a Thermocycler by using the following programme: initial denaturation at 94°C for 2 min, 35 cycles of 1 min per cycle at 94°C, annealing temperature of 46·9°C for 1 min, followed by an increase to 72°C over 1·5 min. Extension of the amplified product was at 72°C for 10 min. Amplified products were separated by electrophoresis in 1·2% (w/v) agarose gels in 1× TAE buffer at 80 V for 2 h. Gels were stained in TAE buffer containing 0·5 μg ml−1 of ethidium bromide. A 100-bp DNA ladder was used as molecular weight marker.

The gels were photographed on a UV transilluminator (GelDoc2000, Bio-Rad), and image analysis was accomplished using Quantity One® software (Bio-Rad).

Random amplified polymorphic-PCR profiles were subsequently analysed using the Gel Compar II Software (Applied Maths; Kortrijk, Belgium). Calculation of the similarity of the bands' profile and grouping of the RAPD-PCR patterns were based on the Pearson's coefficient, and agglomerative clustering was performed with the unweighted pair group matching algorithm.

Survival of LAB in simulated vaginal fluid

For LAB strains that showed antimicrobial activity, survival in simulated vaginal fluid (SVF) was investigated according to Borges et al. (2011). SVF was prepared using components described in Table 2. This mixture was adjusted to a pH of 4·2 using HCl (Pronalab).

Table 2. Composition of simulated vaginal fluid (Owen and Katz 1999)
ComponentConcentration (g l−1)
Bovine serum albumin0·018
Lactic acid2·0
Acetic acid1·0

Lactic acid bacteria cells were harvested by centrifugation at 8877 × g for 10 min at 4°C and washed twice in sterile Ringer's solution (Oxoid; Hampshire, UK). Pellets were then resuspended to the original volume in Ringer's solution.

The SVF was inoculated with 2% (v/v) of harvested cultures of LAB, with a final cell density of 107 CFU ml−1, and incubated at 37°C. At various time intervals (0, 4, 8, 24 and 48 h), aliquots were withdrawn for further enumeration. Two independent replicates of these assays were performed. The enumeration of LAB was performed by serial 10-fold dilutions in Ringer's solution and 20 μl was spread-plated in duplicate on MRS agar. Colony forming units per millilitre (CFU ml−1) were determined after 48 h of incubation at 37°C. Results were expressed as log (N/N0), where N represents the CFU ml−1 of LAB at times after inoculation and N0 represents the initial CFU ml−1.

Simultaneously, 1 ml of each LAB culture was added to 1 ml of SVF. After 24 h of contact, antimicrobial activity was tested against six isolates of L. monocytogenes of different serotypes (1/2b, 4b and 1/2a), using the agar-spot test method (previously described). Translucent halo zones observed around the spots were registered as positive for antimicrobial activity. As a negative control, SVF (without LAB cells) was used.

Characterization of Pediococcus pentosaceus SB83

Based on the bacteriocinogenic activity and survival in SVF, one isolate of LAB (P. pentosaceus SB83) was selected for further studies.

Production of hydrolytic enzymes

The gelatinase and lipase activities were assessed according to Tiago et al. (2004). DNase activity was tested as described by Ben Omar et al. (2004).

All experiments were performed in duplicate, and Staphylococcus aureus ATCC 25213 was used as a positive control.

Haemolytic activity

Production of haemolysin was determined by streaking P. pentosaceus SB83 on a plate of Columbia agar with 5% of sheep blood (BioMérieux; Marcy l'Etoile, France). Plates were incubated at 37°C for 24 h, under aerobic conditions, after which plates were examined for haemolysis. Enterococcus faecalis F2 (from a collection of Tracy Eaton, Division of Food Safety Sciences, Institute of Food Research, Norwich, UK) and E. faecalis DS16 (from a collection of C. B. Clewell, Department of Oral Biology, School of Dentistry, University of Michigan, Ann Arbor, MI, USA) were used as beta-haemolysis controls. The presence or absence of zones of clearing around the colonies was interpreted as beta-haemolysis (positive haemolytic activity) or gamma-haemolysis (negative haemolytic activity), respectively. When observed, greenish zones around the colonies were interpreted as alpha-haemolysis and taken as negative for the assessment of haemolytic activity (Semedo et al. 2003).

Virulence genes

Total DNA isolation was performed using a Gen Elute Bacterial Genomic DNA kit. The primers used for the amplification of genes esp, agg, gelE, efaAfm and efaAfs, cylA, cylB and cylM were described by Eaton and Gasson (2001), and primers of cyl operon: cylLL and cylLS were developed by Semedo et al. (2003). All the primers were purchased from MWG Biotech AG. PCR amplifications were performed in a ThermoCycler in 0·2-ml reaction tubes each with 25 μl of mixtures using 1 × PCR buffer, 2·5 mmol l−1 MgCl2, 0·1 mmol l−1 deoxynucleoside triphosphates (dNTPs), 0·5 mmol l−1 of each primer, 2 U of Taq DNA polymerase and 100 ng μl−1 of DNA. Amplification reactions were as follows: initial cycle of 94°C for 1 min, 35 cycles of 94°C for 1 min, 55°C for 1 min, 72°C for 2 min, a final extension step of 72°C for 7 min and thereafter cooled to 4°C. Amplification products were combined with 3 μl of loading buffer (Bio-Rad), and the preparation was electrophoresed on 0·8% (w/v) agarose gel at 90 V for 2 h. A 100-bp PCR DNA ladder was used as a molecular weight marker. The gels were photographed on an UV transilluminator, and image analysis was accomplished using Quantity One software.

The positive controls used were E. faecalis DS 16 (cylLL+, cylLS+), E. faecalis F2 (cylA+, cylB+, cylM+, efaAfs+), Efaecium F10 (efaAfm+), E. faecalis P1 (agg+, gelE+) and E. faecalis P36 (esp+) (from a collection of Tracy Eaton, Division of Food Safety Sciences, Institute of Food Research, Norwich, UK). For each PCR, a negative control (sample without template) was included.

Antibiotic susceptibility testing

The minimum inhibitory concentration (MIC; μg ml−1) of fourteen antibiotics was determined by the agar microdilution method, according to the National Committee for Clinical Laboratory Standards (NCCLS, 2004). The antibiotics investigated were vancomycin (Fluka; Steinheim, Germany), chloramphenicol (Fluka), nitrofurantoin (Sigma-Aldrich), erythromycin (Labesfal; Tondela, Portugal), tetracycline (Labesfal), ciprofloxacin (Labesfal), rifampicin (Labesfal), gentamicin (Labesfal), streptomycin (Sigma-Aldrich), oxacillin (Sigma-Aldrich), kanamycin (Fluka), ceftazidime (Sigma-Aldrich), penicillin G (Sigma-Aldrich) and ampicillin (Fluka). Antibiotic concentration ranged from 0·03 to 512 μg ml−1.

Minimum inhibitory concentrations of trimethoprim/sulfamethoxazole (SXT, AB Biodisk; Solna, Sweden), meropenem (AB Biodisk) and imipenem (AB Biodisk) were determined using Etest method.

Each test was carried out on Muller-Hinton agar (MHA, BioMérieux) and on MHA with cations adjusted for penicillin G and ampicillin. The inoculum was prepared by suspending a culture in sterile Ringer's solution in order to obtain turbidity equivalent to 0·5 McFarland standards. The experiment was performed in duplicate. Pediococcus pentosaceus SB83 grown on plates of MHA and MHA with cations adjusted with no antibiotic was used as the negative control. Enterococcus faecalis ATCC 29212 and Escherichia coli ATCC 259222 were used as quality control strains.

Biofilm assay

The quantification of biofilm production was performed as described previously by Borges et al. (2011). The wells of a sterile 96-well polystyrene microplate (Brand, Wertheim, Germany) were filled with 180 μl of MRS broth or SVF, both media adjusted to pH values 4·2, 5·5 and 6·5 with HCl or NaOH; 20 μl of overnight grown cells suspended in Ringer's solution was added to each well. The plates were incubated aerobically for 24, 48 and 72 h at 37°C. To quantify the biofilm formation, the wells were gently washed three times with 250 μl of sterile distilled water. The attached bacteria were fixed with 200 μl of methanol (Romyl, Leics, UK) for 15 min, and then, microplates were emptied and dried at room temperature. Subsequently, 200 μl of a 2% (v/v) crystal violet solution (Merck, Darmstadt, Germany) was added to each well and held at ambient temperature for 5 min.

Excess stain was then removed by placing the plate under gently running tap water. Stain was released from adherent cells with 160 μl of 33% (v/v) glacial acetic acid. The optical density (OD) of each well was measured at 630 nm using a plate reader (Microplate reader, Bio-Rad, Hercules; CA, USA).

Each assay was carried out eighteen times, and the negative control was performed in uninoculated MRS broth or SVF. The cut-off (ODc) was defined as the mean OD value of the negative control. Based on the OD, strains were classified as nonbiofilm producers (OD ≤ ODC), weak (OD< OD ≤ 2 × ODC), moderate (2 × ODC < OD ≤ 4 × ODC) or strong biofilm producers (4 × ODC < OD).

Statistical analysis

An analysis of variance (anova) was carried out to test the effect of SVF on survival of LAB strains and to test any significant effect of media, pH and time on biofilm formation of P. pentosaceus SB83. All calculations were made using the software Kaleidagraph (version 4.04, Synergy Software, Reading, USA).


Screening for antimicrobial activity of LAB

Among the food isolates of LAB investigated, 62·5% (35/56) demonstrated antilisterial activity, but did not inhibit GBS.

Clinical isolates of LAB did not show antimicrobial activity against L. monocytogenes or GBS.

Growth inhibition can result from competition, production of lactic acid, hydrogen peroxide or bacteriocins. According to the screening method used, all inhibitory isolates produced proteinaceous compounds that exhibited antimicrobial activity (activity was lost only after the addition of trypsin). This result suggests the antilisterial activity was caused by a bacteriocin.

Quantification of antibacterial activity

Bacteriocin activity of LAB isolates against L. monocytogenes isolates of three different serotypes varied between 400 and 6400 AU ml−1, although values of 1600 and 3200 AU ml−1 (i.e. supernatants diluted 1600-fold and 3200-fold retained significant antilisterial activity) were more common.

One isolate of LAB (P. pentosaceus SB83) had a higher activity for all serotypes compared with the other LAB strains. This isolate showed an antibacterial activity of 6400 AU ml−1 for serotypes 1/2b and 4b and 3200 AU ml−1 for serotype 1/2a.

Identification of bacteriocin-producing lactic acid bacteria and RAPD-PCR

All strains with antilisterial activity were identified as Pediococcus spp. by 16S rDNA sequencing; three isolates were identified as P. acidilactici and the remaining 32 isolates as P. pentosaceus.

A combined dendrogram for both primers M13 and D8635 was achieved, and ten different profiles (clusters at 75%) were obtained for a cophenetic correlation value of 0·85 (data not shown). This value indicates a satisfactory representation of the similarity matrix in the dendrogram, which should be very close to 1 for a high-quality solution.

Survival of LAB in simulated vaginal fluid

The SVF at normal pH (4·2) was used to determine the survival of Pediococcus spp. in these conditions.

In general, the behaviour of strains in SVF was similar (P = 0·31) and the ten different profiles obtained using RAPD-PCR also had a similar behaviour (P = 0·38).

The results demonstrated that in 24 h, there were no great reductions in viable cells (log (N/N0) varied between 0·03 and −0·4), but at 48 h, a greater reduction was observed. After 48 h, six isolates showed a decrease of more than 1·0 log. At this time, the most sensitive isolate had decreased by 1·6 log CFU ml−1 and the most resistant had reduced by only 0·2 log viable cells. Thus, after 48 h, survival was demonstrated to be strain dependent (P < 0·001).

Figure 1 presents the survival in SVF of P. pentosaceus SB83, at different time intervals (4, 8, 24 and 48 h).

Figure 1.

Survival of Pediococcus pentosaceus SB83 in simulated vaginal fluid at pH = 4·2. All points are means ± standard deviations.

All Pediococcus spp. isolates showed antilisterial activity after 24 h of contact with SVF.

Characterization of Pediococcus pentosaceus SB83

Virulence factors (production of gelatinase, lipase and DNase, haemolytic activity and virulence genes).

The production of extracellular enzymes (gelatinase, lipase and DNase) and haemolytic activity were not demonstrated by P. pentosaceus SB83.

The presence of the surface adhesin genes (efaAfs, efaAfm and esp), the aggregation protein gene (agg), the cytolysin genes (cylM, cylB, cylA, cylLL cylLs) and extracellular metallo-endopeptidase gene (gelE) were not present in P. pentosaceus SB83.

Antibiotic susceptibility testing

The antibiotic susceptibility of P. pentosaceus SB83 was evaluated for several antibiotics, in order to cover some of known chemical and functional classes of antibiotics, and MICs (μg ml−1) were determined (Table 3).

Table 3. Minimum Inhibitory Concentration (MIC; μg ml−1) of seventeen antibiotics for Pediococcus pentosaceus SB83
AntibioticImipenemMeropenemSXTAmpicillinPenicillin GCeftazidimeKanamycinOxacillinStreptomycinGentamicinRifampicinCiprofloxacinTetracyclineErythromycinNitrofurantoinChloramphenicolVancomycin
μg ml−10·1251·5>3220·5324880·25816160·1251284512

Biofilm assay

Pediococcus pentosaceus SB83 produced biofilms in both media (MRS and SVF) at different pH values (Fig. 2). In SVF, this strain was classified as a weak or moderate producer; in MRS, P. pentosaceus SB83 was classified as a moderate or strong producer. The quantity of biofilm produced was media dependent (P < 0·001).

Figure 2.

Production of biofilm by Pediococcus pentosaceus SB83 at different time intervals, in SVF and MRS broth at different values of pH. Optical density was measured in SVF at pH 4·2 (image_n/jam12232-gra-0001.png), 5·5 (image_n/jam12232-gra-0002.png) and 6·5 (image_n/jam12232-gra-0003.png) and in MRS at pH 4·2 (image_n/jam12232-gra-0004.png), 5·5 (image_n/jam12232-gra-0005.png) and 6·5 (image_n/jam12232-gra-0005.png) Based on the OD, the strain was classified as a strong, moderate or weak biofilm producer. Error bars represent standard deviation of the mean.

In SVF, P. pentosaceus SB83 produced more biofilm at pH 4·2 (normal vaginal pH) than at higher pH values; at pH 5·5 and 6·5, biofilm production was not significantly different (P = 0·13). In MRS, biofilm production was pH dependent (P < 0·001); biofilm production increased with increasing pH.

At 24 h and 48 h, no significant differences were observed in the production of biofilm at pH 5·5 and 6·5 (in both media). At pH 4·2, in SVF, we did not observe significant differences over time (P = 0·07), whereas in MRS, biofilm formation was higher at 48 h, and no significant differences were observed after 24-h and 72-h incubation (P = 0·28).


Infection with L. monocytogenes can be acquired from animals through direct contact, from contaminated foods, via vertical transmission from mother to foetus/neonate through the placenta or birth canal and via nosocomial cross-infection (Delgado 2008; Posfay-Barbe and Wald 2009). Listeriosis during pregnancy has many complications including preterm labour, chorioamnionitis, spontaneous abortion, stillbirth and neonatal infection (DiMaio 2000). Neonatal disease is classified as either early or late onset. Early onset is more likely in preterm infants, probably resulting from in utero transmission. Late onset is more likely in term infants who acquired L. monocytogenes from the vaginal tract (or birth canal) of asymptomatic mothers. Listeriosis in the neonate is very similar to GBS infection in both course and treatment (Delgado 2008).

Borges et al. (2011) demonstrated that L. monocytogenes can survive and proliferate in conditions of SVF, when women have an increased vaginal pH (pH of 5·5 and 6·5). An increase in pH is caused by a depletion of vaginal lactobacilli (Simhan et al. 2003; Donders et al. 2007), menstruation, unprotected sexual intercourse with the deposition of semen (Tevi-Bénissan et al. 1997) and vaginal medications.

In this study, 35 isolates of LAB showed antimicrobial activity against clinical isolates of L. monocytogenes. Therefore, these isolates of LAB could be used to inhibit L. monocytogenes, preventing the previously mentioned sequelae caused by this pathogen. These LAB isolates were identified as Pediococcus spp., and although they produce lactic acid, their antimicrobial activity was shown to be mainly caused by the production of a bacteriocin. Several authors demonstrated that L. monocytogenes is highly resistant to acidic environments and strains of clinical origins are more resistant to acid than isolates of food origin (Ramalheira et al. 2010; Borges et al. 2011).

Therefore, these isolates of Pediococcus spp. have two fields of action: on one side it inhibits L. monocytogenes by the action of a bacteriocin and on the other side, it maintains the acidic vaginal pH, which will prevent colonization and multiplication by other pathogenic micro-organisms, including GBS.

Many other studies have demonstrated the bacteriocinogenic activity of Pediococcus spp. against isolates of Listeria spp. (Albano et al. 2007; Anastasiadou et al. 2008; Huang et al. 2009; Pinto et al. 2009; Abrams et al. 2011).

Bacteriocin activity of Pediococcus spp. against L. monocytogenes of three different serotypes varied between 400 and 6400 AU ml−1. Albano et al. (2007) reported two pediocins bacHA-6111-2 and bacHA-5692-3 with an activity of 1600 AU ml−1 against L. innocua.

Pediococcus spp. have been used as probiotics in foods (Jonganurakkun et al. 2008; Ruiz-Moyano et al. 2011), although it may also be used in other fields. The antilisterial Pediococcus spp. strains tested could potentially be used as probiotics for vaginal application; consequently, it was important to evaluate their capacity to maintain their viability in a SVF. All isolates of Pediococcus spp. tested were able to survive in SVF at normal pH, and all isolates demonstrated antilisterial activity after exposure to SVF after 24 h. During 24 h, Pediococcus spp. maintained their survival and the greatest reduction was only 0·4 log CFU ml−1. After 48 h, only six isolates demonstrated a decrease of more than 1·0 log.

In our study, we used a SVF proposed by Owen and Katz (1999). This formulation has been approved for research into contraceptive and prophylactic drug delivery, and thus, some compounds present in vaginal fluid are not included (e.g. not included are glycogen or mucin and the glucose concentration is low). However, even with the exclusion of some ingredients, Pediococcus spp. could survive and were shown to be resistant isolates.

These results are in agreement with other reports that demonstrated the survival of Lactobacillus spp. in vaginal conditions (Geshnizgani and Onderdonk 1992; Reid et al. 1998; Valore et al. 2002; Tomás and Nader-Macías 2007).

The vaginal microbiota is normally dominated by lactobacilli; however, many other LAB colonize women. In the study of Jin et al. (2007), a total of 338 vaginal lactic acid bacteria were isolated, including five genera: Lactobacillus, Leuconostoc, Pediococcus, Streptococcus and Weisella. The source of vaginal LAB has not been identified, but one source could be the environment, such as eating fermented food. In a study by Petricevic et al. (2012), 30 pregnant women and 30 postmenopausal women were investigated for LAB colonization; 80% of pregnant women and 40% of postmenopausal women had the same lactobacilli strains in their vagina and rectum. So, these results support the hypothesis that the rectum may play an important role as a reservoir for some strains of LAB that colonize the vagina.

Therefore, a bacterial replacement therapy controlling vaginal infections could be developed with isolates from food. Thus, the Pediococcus spp. used in this study could be used as a probiotic with vaginal application. Our isolates are able to inhibit the growth of L. monocytogenes by the production of a bacteriocin and can survive in SVF. P. pentosaceus SB83 showed a greater bacteriocinogenic activity compared with other isolates. This strain maintained viability in SVF, showing only a slight decrease after 48 h of exposure (0·3 cycle log).

Lactic acid bacteria have acquired the ‘generally regarded as safe’ status by the American Food and Drug Administration. However, the safety of any strain of LAB should be evaluated before consideration for use as a vaginal probiotic.

Pediococcus pentosaceus SB83 did not produce virulence factors such as gelatinase, lipase and DNase, haemolytic activity and nor show the presence of virulence genes (genes of surface adhesion, aggregation protein, cytolysin and extracellular metallo-endopeptidase). Some of these virulence factors, frequently identified in pathogenic bacteria, have been detected in enterococci (Barbosa et al. 2010).

The safety aspects of LAB are of concern, including the presence of potentially transferable antibiotic resistances to pathogenic bacteria (Ammor et al. 2007). Probiotics can be susceptible to the majority of antibiotics or can naturally be, or rendered, multiresistant (Courvalin 2006).

According to the microbiological breakpoints for antimicrobials defined by the Panel on Additives and Products or Substances in Animal Feed (FEEDAP) of the European Food Safety Authority (EFSA, 2008), P. pentosaceus SB83 was considered sensitive to chloramphenicol, gentamicin, streptomycin, kanamycin, erythromycin and ampicillin. This strain was considered resistant to tetracycline and intrinsically resistant to vancomycin. The resistance to tetracycline was found in other studies of Pediococcus spp. (Swenson et al. 1990; Tankovic et al. 1993; Rojo-Bezares et al. 2006; Danielsen et al. 2007).

From the data obtained in the study by Klare et al. (2007), tentative species- or group-specific epidemiological cut-off (ECOFF) values of MICs were defined for recognizing intrinsic and acquired antimicrobial resistances. According to that study, the ECOFF for P. pentosaceus to oxytetracycline (class of tetracycline) was 32 μg ml−1. Therefore, the tetracycline resistance of P. pentosaceus SB83 is likely to be intrinsic. Klare et al. (2007) did not observe an acquired antibiotic resistance in any isolates of the Pediococcus spp. tested.

Vancomycin resistance is due to their possession of D-Ala-D-lactate in their peptidoglycan rather than the D-Ala–DAla dipeptide. Such resistance is thus intrinsic in most LAB because the antibiotic's target is absent and not comparable to the transmissible plasmid-encoded vancomycin resistance found in Enterococcus spp. Intrinsic resistance is not horizontally transferable and poses no risk in nonpathogenic bacteria (Ammor et al. 2007).

Additives and Products or Substances in Animal Feed (FEEDAP) did not define the breakpoints for nitrofurantoin, ciprofloxacin, rifampicin, oxacillin, ceftazidime, penicillin, SXT, meropenem and imipenem.

In general, our results are in agreement with other studies for a broad range of antibiotics although different nutrient media, incubation conditions and susceptibility testing methods were used (Swenson et al. 1990; Tankovic et al. 1993; Zaragaza et al. 1999; Rojo-Bezares et al. 2006; Danielsen et al. 2007; Klare et al. 2007).

A biofilm is defined as a community of micro-organisms attached to a surface, producing extracellular polymeric substances, exhibiting an altered phenotype compared with planktonic cells, especially with regard to gene transcription and interacting with each other (Lindsay and von Holy 2006).

Pediococcus pentosaceus SB83 produced biofilm in MRS and SVF at different pH values. It was demonstrated that P. pentosaceus SB83 formed a greater quantity of biofilm in MRS than in SVF fluid; medium composition influenced the quantity of biofilm production.

In MRS medium, biofilm production increased with increasing pH. In SVF, P. pentosaceus SB83 produced more biofilm at pH 4·2 (normal vaginal pH) than at higher pH values. Van der Veen and Abee (2011) studied the formation of single- and mixed-species biofilms of L. monocytogenes and L. plantarum, and the contribution of each species to the biofilm formation was dependent on the composition of medium. The addition of glucose to the medium (brain heart infusion) decreased the number of L. monocytogenes and increased the contribution of L. plantarum for the production of mixed biofilm. The acidification resulting from metabolism of glucose allowed the formation of biofilm by L. plantarum because this strain is able to grow at low pH. The acid tolerance of LAB provides the opportunity to produce biofilm at acidic pH values; this capacity was evident in P. pentosaceus SB83 in the conditions of SVF at normal pH.

In a study by Guerrieri et al. (2009), LAB biofilms (of strains L. plantarum 35d, E. casseliflavus IM 416K1, L. plantarum 396/1, E. faecalis JH2-2) showed the ability to influence the survival and multiplication of L. monocytogenes. So, the production of biofilm by P. pentosaceus SB83 in SVF can be an advantage in reduction in L. monocytogenes. Lactic acid bacteria biofilms also serve as a protective layer against the colonization by other pathogenic bacteria. These results should be complemented with adhesion assays to vaginal epithelial cells.

Cell adhesion and biofilm production play an important role in the stress resistance of LAB. Three mechanisms could increase the resistance of biofilm cells: the cell membrane becomes more resistant; the biofilm is protected by extracellular polymeric secretions, and the three-dimensional structure of the biofilm protects the inner cells (Kubota et al. 2008).

In conclusion, this study suggests that P. pentosaceus SB83 has the potential to be used as a vaginal probiotic, to prevent colonization of L. monocytogenes in pregnant women and consequently to reduce neonatal infections.


We thank Hospital S. Marcos (Braga) for providing isolates of Group B Streptococcus and vaginal Lactobacillus. This work was supported by National Funds from FCT – Fundação para a Ciência e a Tecnologia through project PEst-OE/EQB/LA0016/2011. Financial support for author S. Borges was provided by PhD fellowship, SFRH/BD/45496/2008 (FCT – Fundação para a Ciência e a Tecnologia).