In Vitro Antagonistic Effect of Lactobacillus on Organisms Associated with Bacterial Vaginosis - Magdalena Strus, Ph.D., Magdalena Malinowska, M.D., and Piotr B. Heczko,

OBJECTIVE: To assess antagonistic properties of Lactobacillus strains isolated from the vaginas of healthy women as compared to the most common bacteri­al agents related to vagi­nosis.

STUDY DESIGN: Antagonistic activity of different Lactobacillus strains isolated from the vaginas of healthy women not treated for infections with an antibi­otic for the previous three

months was screened using an agar slab method. The ac­tivity was tested against test organisms associated with bacterial vaginosis and/or urinary tract infections:

Staphylococcus aureus, Enterococcus faecalis, Streptococcus agalactiae, Escherichia coli, Gard­nerella vaginalis, Peptostreptococcus anaerobius and Prevotella bivia.

RESULTS: Many of the 146 Lactobacillus strains tested exerted apparent antagonistic activities against

grant-positive aerobic cocci and gram-negative rods, such as S aureus and E coli, and a marked number of Lacto­bacillus strains inhibited facultative bacteria, such as Gardnerclla vaginalis and the anaerobes P anaerobius and P bivia. Only a fezv lac­tobacilli were able to inhibit growth of E faecalis and S agalactiae. Indicator bac­teria growth inhibition prob­ably relies upon several different complementary

mechanisms. The specific indicator bacteria species deter-mines which mechanism predominates. CONCLUSION: Lactobacillus strains taken from nor-mal vaginal flora demonstrated antagonistic activity against a variety of bacteria related to vaginal and uri-nary tract infections. The specific occurrence rates of ac­tive Lactobacillus strains are different, and this differ­ence is dependent on the indicator bacteria species. (J

Reprod Med 2002;47:41—46)

Our results confirm the specific
antagonistic features of LAB, which
most probably are present and active
in normal human flora

From the Department of Bacteriology, Chair of Microbiology, Medical College of Jagiellonian University, and Department of Obstetrics and Gynecology, Health Care Trust of the Ministry of Internal Affairs and Administration, Krakow, Poland

Dr. Strus is Head, Laboratory of Probiotics and Normal Flora, Chair of Microbiology, Medical College of Jagiellonian University.

Dr. Malinowska is Assistant, Department of Obstetrics and Gynecology, Health Care Trust of the Ministry of Internal Affairs and Ad-,nizzistralion.

Dr. Heczko is Professor and Chairman, Department of Bacteriology, Institute of Microbiology, Medical College of Jagiellonian University.

Supported by Scientific Research Committee grant 4 P05F 001 C/2737 from the State Committee for Scientific Research.

Address reprint requests to: Piotr B. Heczko, M.D., I'h.D., Department of Bacteriology, Institute of Microbiology, Medical College of Jagiellonian University, UI. Czysta 18, 31-121 Krakow, Poland (mbheczko6cyf-kr.edu.pl).

Financial Disclosure: The authors have no connection to any companies or products mentioned in this article.

0024-7758/02/4701-(0)41 /515.00/0 D Journal of Reproductive Medicine'", Inc.

The journal of Reproductive Medicine

Keywords: bacterial vaginosis, vaginitis, Lactobacilli introduction

Vaginitis is a common disorder in adult women; its major three forms are distinguished according to It is possible that lactobacillus inhabiting complex microorganism ecosystems actively controls the growth of other microorganisms by multiple complementary mechanisms.

etiology: candidiasis, trichomoniasis and bacterial vaginosis.'" Until recently, it was thought that bac¬terial vaginosis was caused by a single etiologic factor (either Gardnerella vaginalis, Prevotella bivia, Bacteroidc's spp., Peptostreptococcus or Mobilunclus). Now it is known that bacterial vaginosis results from the uncontrolled, sequential growth of bacteria that are consistently found in the vaginas of healthy women.' The pathogenicity of this group of bacteria is supported by the fact that they are isolated from the genital tracts of women with acute pelvic inflammatory disease.3,'0

Lactobacillus bacteria (LAB) in the normal vaginal flora are represented usually by two or three species. The most common organisms are Lactobacilluus plantarum, Lactobacillus fermentum, and Lactobacillus acidophilus. Using molecular biology methods, it has been found that L acidophilus is not a predominant component of vaginal flora!' The diversity and density of Lactobacillus populations in the vagina can be directly correlated with patient age and health status. The presence of Lactobacillus is also necessary to maintain a microbiologic balance owing to its antagonistic and adherence capabilities. LAB have never been found to be an etiologic agent in genital tract infections.

Antagonism is a one of the mechanisms used by LAB to establish their stable populations in the vagina. This antagonism is based upon Lactobacillus ability to control potentially pathogenic, mucosa-colonizing microorganisms by one or a combination of the following mechanisms: a pro-found decrease in vaginal pH caused mainly by products of carbohydrate fermentation (lactic, acetic, propionic and pyroglutamic acid) the production of specific antibacterial bacteriocin like substances, such as lactocin (Lactobacillus spp) and acidophilin (L acidophilus)'; and the production of hydrogen peroxide.'

The antagonistic and adherence properties of Lactobacillus are necessary to maintain normal, constant composition of the vaginal flora. However, not all Lactobacillus strains possess these properties equally, and thus coexistence of bacteria related to vaginosis with several Lactobacillus strains at various ratios is observed in some vaginal infections.'

The purpose of this study was to compare the antagonistic properties of LAB isolated from the vaginas of healthy women to bacteria related to vaginal and/or urinary tract infections, such as P bivia, Peptostreptococcus anaerobitus, G. vaginalis, Streptococcus agalactiae, Enterococcus faecalis, Escherichia coli and Staphylococcus aureus.

Materials and Methods Bacterial Strains

Lactobacillus strains were isolated from the genital tracts of healthy women. These strains were initially characterized according to morphologic features, ability to grow at 15°C, catalase production, arginine degradation and glucose fermentation. After detailed phenotypic identification using API 50 CHL tests (bio-Merieux, Marcy-l'Etoile, France), the Lactobacillus strains were stored at -70°C on glass beads in a 10%, glycerol medium. Before testing, the glass beads, coated with pure strain culture, were incubated in 2 mL MRS liquid medium (Oxoid, Basindstoke, U.K.) at 37°C for 18 hours under anaerobic conditions (MACS MG500, Don Whitley Scientific Ltd, Shipley, England).

The organisms used as indicator bacteria were G vaginal is ATCC 4018, S agalactiae)S 11508 (Chair of Microbiology's own strain), E coli (strain 4 with P type fimbriae from the PZH collection (National institute of Hygiene, Warsaw, Poland]), Staphylococcus aureus JS 16317 (Institute of Microbiology's own strain), P bivia ATCC 29303 and P anaerobius ATCC 27337. These organisms were obtained from inter-national microorganism collections or were isolated from specimens taken from patients with genital tract infections. The indicator strains were kept frozen at — 70°C in a culture broth with 10'!';, glycerol coated on glass beads.

Slab Method for Antagonism Testing

For testing antagonism, the indicator bacteria were cultured on appropriate broth media. Aerobic species were cultivated in Tryptic Soy Broth (Difco, Detroit, Michigan) for 24 hours at 37°C. Anaerobes were propagated in a special broth (ATCC 593) for 48 hours at 37°C in anaerobic conditions (as above), and then G vaginalis was subcultured on Columbia agar (Difco) plates enriched with human blood and P llivia and P anaerobiirs on plates containing Brucella agar (Difco) enriched with human blood, hemine and phylloquinone (vitamin K1). The plates were incubated in an anaerobic atmosphere (as above) for 48 hours at 37°C. To prepare test plates, one loop of the aerobic broth culture was transferred to 5 mL of fresh broth and incubated again for four to eight hours at 37°C in a water bath. The cultures were then diluted with normal saline to 0.5 density ac-cording to McFarland's score. Of this suspension, 0.1 mL was swabbed onto the surface of 20 mL of an appropriate solid medium. Mueller-Hinton's (Difco) medium was used for both E coli and S aurelrs, and the same medium containing 5% sheep blood was used for streptococci. Anaerobic indicator bacteria grown on agar plates were removed under anaerobic conditions by swabbing and transferred to tubes containing special, prereduced broth (ATCC 593) to obtain dense suspension (2.0, McFarland's score). Then the suspensions were plated on adequate solid media by swabbing 0.1 mL on the surface. Plates containing 20 mL of Columbia agar (Difco) enriched with human blood were used for G vaginalis, and plates containing 20 mL of Brucella agar (Difco) enriched with human blood, hemine and phylloquinone were used for P biviri and P anaerobius

A slab method described in detail in a previous paper9 was used for antagonism testing. The 48-hour-old MRS broth cultures (Oxoid) of the test Lac¬tobacillus strains were streaked with a swab onto plates containing 40 mL of MRS agar (Oxoid) and incubated for 48 hours at 37°C under anaerobic con¬ditions. Round slabs 9 mm in diameter were then cut out of the agar plates with a cutter. These slabs with bacterial growth on top were placed on the surface of plates previously inoculated with indica-tor strains, as described above. After incubation in conditions adjusted to the demands of the specific indicator bacteria, as specified above, indicator strain growth inhibition zone diameters were mea¬sured (in millimeters) (Figure 1). Slabs cut off the uninoculated MRS agar were placed on each cul¬ture of the indicator strains as controls.

Results

The following observations can be drawn from the data presented in Figures 2 and 3. These figures show the antagonistic activity of 146 Lactobacillus strains upon S agalactiae and E faecalis. The initial pH value of the MRS medium used for culture of Lactobacillus was 6.2 ±0.2. All Lactobacillus strains uniformly lowered the pH of the MRS medium on which they were grown to between 4.0 and 4.5. However, only certain Lactobacillus strains inhibited the growth of S agalactiae and E faecalis (four and two strains, respectively). Thus, we conclude that the mechanism of antagonistic activity of Lactobacil¬lus on S ag>aleic tint' and E fat'calis is not connected with a decrease in pH by Lactobacillus but by anoth¬er mechanism not directly connected with pH.

Figures 4 and 5 show the antagonistic action of Lactobacillus on unrelated indicator bacteria, S a1r­reus and E coli; however, the activity patterns look similar. Comparing this action with that on strepto­cocci and enterococci (Figures 2 and 3) demon­strates that many more Lactobacillus strains showed moderately and markedly increased activity (inhi­bition zone diameters 15—18 mm). The resulting plot of these inhibition zones have a Gaussian dis­tribution.

Figure 6 shows that of the 146 tested strains of Lactobacillus, 116 markedly inhibited the growth of G vaginalis.

Lactobacillus' inhibitory activity against the anaer­obes P hivia and P anaerobius was then tested. Strains of Lactobacillus were selected for their marked inhi­bition of aerobes. Twenty-four selected LAB strains were tested against P anaerobins (Figure 7), and 23 of them caused marked inhibition. In the case of P bivia (Figure 8), 22 LAB strains were selected, and 19 caused marked inhibition.

Discussion

It was not a purpose of this study to analyze exact mechanisms of the antagonistic activity of the test-ed Lactobacillus strains but to observe the occur­rence and range of this phenomenon among vagi­nal lactobacilli. However, it is possible to speculate that high acidity of the test medium caused by the growth of lactobacilli is not enough to cause growth inhibition of many different indicator bacteria. All the tested lactobacilli decreased the pH of the test medium up to 4.0^—1-.5, and in spite of this there are evident differences between indicator bacteria in their susceptibility to Lactobacillus action.

Bacteriocinlike substances are unlikely candi­dates for this inhibition since their spectrum of ac­tivity is very narrow.' Their activity could be associated with growth inhibition of E faecalis since en­terococci belong also to lactic acid bacteria, and their cell wall structure is similar to that of Lacto­bacillus. McGroarty and Reid^s detected decreased Lactobacillus activity on enterococci and suggested that this activity may be associated with substances similar to bacteriocins. However, it is also possible that a mechanism of antagonistic action of Lacto­bacillus on these indicator strains is based upon the production of acids other than acetic or lactic acid, as described by Yang and coworkers' ^I; that may act directly upon the bacterial cells of S ag>alactiae and E J _`tecaliS.

The occurrence of antagonistic activity in vaginal Lactobacillus strains against two such systematically different bacteria as S aureus and E coli was almost identical. This allows us to suggest that the mecha­nisms of Lactobacillus' antagonistic action against both bacterial species are similar. To our knowl­edge, S aureus strains have never been used before in studies of Lactobacillus's antibacterial activity. E coli inhibition has been shown only for one strain, Lactobacillus rhantnosus GR-1, and it was ascribed to the presence of unknown low-molecular-weight substances.'

Although we think that several mechanisms overlapped in the case of Gardnerella growth inhibi­tion, decreased pH secondary to lactic acid forma­tion seems to be the dominant mechanism since nearly all tested lactobacilli were active against these bacteria, and acid production is a very com­mon feature of the genus Lactobacillus. Growth inhi­bition of strict anaerobes by LAB is most probably dependent upon mechanisms similar to those seen when Lactobacillus inhibits G vaginalis: acid produc­tion. As shown in Figures 7 and 8, standard collec­tion strains of Lactobacillus, acidophilus ATCC 4356 and Lactobacillus plantarum NCDO 1752, not known for any antibacterial activity were added to a set of test lactobacilli for comparison and also caused in­hibition. This hypothesis is supported by Pybus and coworkers,' who analyzed the anaerobe growth rate in the presence of low vaginal pH and found that anaerobes are extremely sensitive to an acidic environment.

LAB antibacterial activity against anaerobic bacteria has never been tested because of the lack of an adequate method. The slab method that we developed and tested" allows direct examination of bacteria with different oxygen tolerances in one agar system. Our results confirm the specific antagonistic features of LAB, which most probably are present and active in normal human flora. It is possible that Lactobacillus inhabiting complex microorganism ecosystems actively controls the growth of other microorganisms by multiple complementary mechanisms.

References

1. Dahiya RS, Speck ML: Hydrogen peroxide formation by lac¬tobacilli and its effect on Staphylococcus aureus. J Dairy Sci 1968;51:1568-1572

2. Hillier SL: Diagnostic bacteriology of vaginosis. Am J Obstet Gynecol 1993;169:445-459

3. Hillier SL, Holmes KK: Bacterial vaginosis. In Sexually Transmitted Diseases. Second edition. Edited by KK Holmes, PA MArdh, PF Sparling, et al. New York, McGraw-Hill, 1984, pp 547-559

4. Hoover DC, Steenson L (editors): Bacteriocins of Lactic Acid Bacteria. San Diego, Academic Press, 1993

5. McGroarty J, Reid C: Inhibition of Enterococci by Lactobacil¬lus species in vitro. Microb Ecol HIth Dis 1988;1:215-219

6. McGroarty J: Probiotic use of lactobacilli in the human fe¬male urogenital tract. HEMS Immunol Med Microbial 1993;6:251-264

7. Pybus V, Onderdonk AB: The effect of pH on the growth and succinate production by Prer'otc'lla bivia. Microb Eco) Hlth Dis 1996;9:19••25

8. Reid G, McGroarty J, Angotti R, et al: Lactobacillus inhibitor production against Esclrerichia coli and coaggregation ability with uropathogens. Can J Microbiol 1988;34:344-351

9. Strus M: A new method of invitro evaluation of the antago¬nistic effect of Lactobacillus on selected disease-producing in¬dicator bacteria (in Polish). Med Dosw Mikrobiol 1998;50

questa pagina è aggiornata al 23/06/2005