Vaginal lactic acid bacteria in healthy and ill bitches and evaluation of in vitro probiotic activity of selected isolates

Abstract

The concentrations of lactic acid bacteria (LAB) in 42 vaginal samples from healthy and ill bitches were determined. Eight isolates belonging to the genera Lactobacillus and Enterococcus were selected and identified and their in vitro antimicrobial activity against canine pathogens and their ability to adhere to canine vaginal epithelial cells were determined. There was no correlation between the concentrations of vaginal LAB and clinical status, body temperature, vaginal pH, or age. Although the animals were either well or suffering from various illnesses, LAB were found in almost every sample, and the selected isolates showed promising probiotic-related features. These findings are significant for the design of strategies for the modulation of vaginal microbiota by vaginal LAB isolates.

Introduction

Normal vaginal microbiota protects the genito-urinary tract against potentially pathogenic bacteria by competing for nutrients or interfering with adhesion to epithelial cell receptors (1). In dogs, changes in the vaginal microbiota are associated with reproductive tract diseases characterized by the proliferation of microorganisms that are normal components of the enteric microbiota (such as, Escherichia coli and Proteus mirabilis) (2). One of these diseases, bacterial vaginitis, is associated with replacement of the native microbiota, primarily by enteric bacteria. Bjurström (3) found that E. coli, beta-hemolytic streptococci, Staphylococcus intermedius, and Pasteurella multocida were the species most often isolated from bitches with pyometra and from those with dead puppies (3).

Probiotics are “live microorganisms which, when administered in adequate amount, confer a health benefit on the host” (4). However, their mode of action is not well understood. Lactic acid bacteria (LAB), particularly lactobacilli, are frequently used in probiotic preparations (5). The traditional use of probiotics in dogs is, as in human beings, mainly for the treatment of diarrheas and prophylactic supplementation to restore imbalances of the intestinal microbiota associated with antibiotic therapy or clostridial diarrheas (6).

Several bacterial features must be taken into account to assess whether a particular microorganism is likely to be effective as a probiotic. Adhesion to epithelial cells may endow probiotic bacteria with the ability to establish in the host and to possibly interfere with colonization by pathogens. In addition, antimicrobial activity due to the production of bacteriocins, hydrogen peroxide, or other compounds is desired in microorganisms selected for probiotic applications. Also, probiotic administration must be safe; the microorganisms should not be potential pathogens (1).

The aim of this study was to quantify vaginal LAB of healthy and ill bitches, to evaluate the relationship of LAB concentrations with various clinical states, and to assess the probiotic potential of selected isolates.

Materials and methods

A total of 42 bitches from the Veterinary Hospital and Barrios Unidos clinic (Faculty of Veterinary Medicine, University of Uruguay) were sampled from October 2004 to July 2005. The dogs were kept individually and did not have contact with each other. Sixty-seven percent (n = 28) of the population had various diseases at the time of the assay. Fifty-seven percent (n = 24) of the bitches were crossbred, while the remaining animals belonged to 11 different breeds. Ages ranged from 3 mo to 14 y (mean = 4.5 y). Clinical condition was evaluated for each bitch and the vaginal pH was measured using indicator strips (Merck, Darmstadt, Germany). Age and antibiotic or hormone treatment during the last month were recorded (Table 1).

Table 1.

Characteristics of the bitches used in the study including body temperature, vaginal pH, and concentrations of bacteria on MRS and Rogosa agar media

Age(years)	Breeda	Clinical status	Treatmentb	Body temperature(°C)	Vaginal pH	MRS agar log (CFU/sample)	Rogosa agar log (CFU/sample)
0.25	PK	Healthy		38.0	6.0	4.5	0.0
1.6	CB	Venereal transmissible tumour	Vincristine	38.7	8.0	3.3	2.0
13	CB	Uremic syndrome	Sulfa/Trimethoprim	38.2	6.5	2.3	0.0
2	CB	Healthy		38.5	7.5	4.5	0.0
7	TE	Healthy		38.2	7.0	4.7	1.9
2	CS	Healthy		39.2	7.0	5.6	1.8
2	CS	Healthy	Progestagens	38.9	6.5	5.9	3.0
2	CS	Healthy	Progestagens	39.0	6.0	3.9	2.2
2	CS	Mammary gland tumor		38.9	7.5	3.9	1.0
2	CS	Healthy		39.6	6.5	2.4	1.0
3	CB	Healthy		39.0	6.5	3.4	2.9
10	YT	Stomatitis		38.6	6.0	3.4	2.9
6	CB	Traumatism		39.5	6.5	2.9	2.0
8	CB	Uremic syndrome	Amoxicillin	38.0	7.5	0.0	0.0
10	CB	Hemorrhagic enteritis		39.1	7.5	4.5	1.3
0.8	CB	Gastroenterocolitis	Sulfa/Trimethoprim	38.4	6.5	5.0	0.0
0.8	CB	Gastroenterocolitis	Sulfa/Trimethoprim	39.3	7.0	4.3	3.9
0.8	CS	Pneumonia	Amoxicillin/Clavulamic acid	39.4	6.5	3.3	0.0
0.7	CB	Canine distemper		39.3	7.0	5.7	0.0
14	CB	Basal cells tumor		38.1	7.5	4.7	3.3
12	CB	Pyometra		39.7	5.5	4.8	2.3
8	B	Mastitis		39.0	6.5	5.9	1.8
1	CB	Canine distemper		38.8	7.0	4.8	2.3
9	GS	Mammary gland tumor		39.4	6.5	3.6	2.4
10	B	Mammary gland tumor	Cephalexin	NDc	7.5	4.9	3.8
4	P	Canine distemper	Amoxicillin/cephradine	39.5	6.5	3.8	2.7
2	CB	Healthy		39.0	7.5	2.4	1.7
7	GS	Wet patch		38.7	6.5	4.3	1.7
3	CB	Healthy		39.0	6.5	5.3	1.6
1.5	CB	Healthy		38.2	7.0	3.2	2.9
1.5	CB	Traumatism		39.4	6.5	4.5	0.0
6	D	Wobbler disease		37.4	6.5	5.7	5.1
0.75	BT	Demodectic mange	Ivermectin	39.0	6.5	6.1	0.0
11	CB	Vaginal tumor		38.6	7.0	5.7	0.0
4	CB	Healthy		38.5	6.5	6.7	6.0
0.5	CB	Healthy		39.5	7.5	5.7	0.0
3.5	CB	Sarcoptic mange		39.6	7.0	4.2	0.0
13	CB	Pyometra		36.6	5.5	5.7	0.0
1.5	CB	Venereal transmissible tumor		39.4	7.5	3.8	0.0
2	CB	Canine distemper		39.0	7.0	8.0	0.0
1	CB	Canine distemper	Amoxicillin/Clavulamic acid	39.8	7.0	6.0	0.0
0.5	GH	Healthy		38.5	6.5	5.0	0.0

^aPK — Pekinese, CB — Crossbreed, TE — Tervueren, CS — Cocker spaniel, YT — Yorkshire terrier, B — Breton epagneul, GS — German shepherd, P — Pinscher, D — Doberman, BT — Bull terrier, GH — Greyhound

^bCases with specific treatment

^cND — Not determined

Samples were obtained by gently swabbing the vaginal wall for 30 s, avoiding contact with the vulva. The swabs were immediately introduced into 1 mL peptone water, kept at 4°C until arrival at the laboratory (<3 h), and then vortexed vigorously for 1 min to suspend attached bacteria (7). The resultant bacterial suspension was decimally diluted with phosphate-buffered saline (PBS) and serial dilutions were plated with a detection limit of 10 colony-forming units (CFU) per sample. Lactic acid bacteria were isolated by plating on de Man Rogosa Sharpe (MRS) agar (Oxoid, UK) supplemented with nystatin (150 U/mL, Urufarma, Uruguay). The concentration of lacto-bacilli was determined by plating on Rogosa agar. The plates were incubated for 48–72 h at 37°C under a microaerophilic atmosphere (8).

Eight different isolates were characterized to determine if individual canine vaginal isolates had potential as probiotics for dogs. Enterococcal isolates were randomly picked from MRS and lactobacilli from Rogosa plates. The isolates were characterized by phenotype, based on Gram stain and cell morphology, oxidase and catalase reactions, and growth in anaerobic jars (Oxoid, UK) (9).

In order to identify the isolates, a molecular approach was used. Polymerase chain reaction (PCR) amplification and sequencing of the almost complete 16S rRNA were carried out with primers 11f 5′-AGAGTTTGAT(C/T)(A/C)TGGCTCAG-3′ and 1492r 5′-TACCTTGTTACGACTT-3′ (8). Genomic DNA was extracted with the Gene elute bacterial genomic DNA extraction kit (Sigma, St. Louis, Missouri, USA) according to the manufacturer’s instructions. The PCR reactions were performed in a T1 thermocycler (Biometra, Germany) as previously described (8). Sequences were obtained using a sequencing service (Macrogen, Seoul, South Korea) with primers 11f, 1492r and internal primer 518 of ′-CCAGCAGCCGCGGTAAT-3′ (10). The sequences were compared with sequences in the GenBank database of the National Center for Biotechnology Information using the BLASTn program (11).

Test pathogens consisting of strains of P. mirabilis (12), S. aureus (13), and E. coli (isolated from canine pyometra, this study) used for antimicrobial assays were cultured in brain heart infusion (BHI) medium.

An agar spot test was performed to evaluate the isolates’ ability to inhibit E. coli, S. aureus, or P. mirabilis as representative gram-positive and gram-negative pathogenic bacteria (8,13). A 2.5 μL volume of a fresh culture of each isolate was spotted onto MRS agar plates and incubated for 24 h at 37°C under microaerophilic conditions. Then the plate surface was covered with BHI semi-solid agar (0.8% w/v) inoculated with the test pathogen (1 × 10⁷ CFU/mL) and aerobically incubated at 37°C. Bacterial growth inhibition was determined by measuring the zones of inhibition surrounding the bacterial spots. This assay was performed in duplicate.

Bacterial adhesion to canine vaginal epithelial cells (CVEC) was assessed following the method described by McLean and Rosenstein, with modifications (14). In order to obtain CVEC, sterile swabs were rubbed against the vaginal walls of 3 healthy bitches, immersed in citric acid-Na₂HPO₄ buffer and stored at 4°C for less than 3 h until they were used. The swabs were vortexed and the supernatants containing the CVEC were pooled. The CVEC were collected by centrifugation at 800 × g for 7 min, washed 3 times, and then suspended at a concentration of 1 × 10⁶ cells/mL in the same buffer. Equal volumes of CVEC and bacteria (1 × 10⁶ CFU/mL) were incubated without shaking for 1 h at 37°C and then the cells with adherent bacteria were collected and washed 3 times in citric acid-Na₂HPO₄ buffer (800 × g, 7 min) to remove unattached bacteria. Smears for light microscopy were prepared and stained with 1% crystal violet and bacteria attached to 50 individual CVEC were counted. Control CVEC smears were made to confirm that the presence of native bacteria was negligible.

A regression model was used to correlate the following variables: decimal logarithms of bacterial counts on MRS and Rogosa agar, age, body temperature, vaginal pH, health status and treatment. The model adjustment was calculated by r ² and the association between variables was considered significant when P < 0.05. Data were analyzed using computer software (STATA, version 8.2; SataCorp, College Station, Texas, USA).

Results

Lactic acid bacteria were detected in 41 of the 42 vaginal samples cultured on MRS agar and their concentrations varied from 2.3 to 8.0 log₁₀ CFU/sample (Table 1). An 8-year-old bitch suffering from uremic syndrome was the only one in which LAB were not detected. Growth of lactobacilli on Rogosa agar plates was observed in 25 of the 42 vaginal samples (59%); bacterial concentrations ranged from 1 to 6 log₁₀ CFU/sample(Table 1).

There were no significant correlations between concentrations of bacteria on MRS and Rogosa agar and age, vaginal pH, body temperature, or clinical condition. When the 16S rRNA gene sequences of the 8 isolates selected for further study were compared with the GenBank (NCBI) database, the sequences belonged to Lactobacillus and Enterococcus genera, including Lactobacillus murinus, Lactobacillus plantarum, and Enterococcus canintestini (Table 2).

Table 2.

Identification of 8 vaginal isolates, their adherence to canine vaginal epithelial cells, and their inhibition of growth of selected pathogens

			Diameter of antibacterial inhibition zones (mm)
Strain	Isolates identity	Adhesion to CVECa	E. coli	P. mirabilis	S. aureus
8B	Enterococcus sp.	17.3 6 18.6	25	15	23
12C1	Lactobacillus sp.	123.5 6 43.1	25	25	25
13C	L. murinus	46.8 6 34.68	17	9	21
14C1	Enterococcus sp.	28.4 6 22.7	7	11	5
21A	Lactobacillus sp.	5.02 6 5.08	21	15	17
23B	E. canintestini	26.54 6 18.93	13	13	15
26D	L. plantarum	10.12 6 7.01	16	17	15
43D	Enterococcus sp.	NDb	17	17	17

^aAdhesion results are expressed as number of attached bacteria per canine epithelial vaginal cell (CVEC, mean ± standard deviation)

^bND — Not determined

All the selected isolates showed antimicrobial activity against the pathogenic bacteria used in the assays (P. mirabilis, S. aureus, and E. coli) (Table 2). Enterococcus sp. 8B and Lactobacillus sp. 12C1 were the isolates that produced the widest zones of inhibition against E. coli isolated from a canine pyometra (25 mm). Lactobacillus sp. 12C1 also showed the widest zones of inhibition against P. mirabilis and S. aureus while Enterococcus sp. 14C1 produced the smallest zones.

Significant differences in bacterial adhesion to CVEC were observed (P < 0.05). The most adhesive strain was Lactobacillus sp. 12C1 (123 ± 43.1 attached bacteria/CVEC) while Lactobacillus sp. 21A showed the lowest adhesion value (5.02 ± 5.08 attached bacteria/CVEC; Table 2).

Discussion

The aim of this study was to quantify vaginal LAB, to evaluate their relationship with various clinical parameters, and to establish if individual LAB strains exhibited potential for use as probiotics in bitches. Lactic acid bacteria were recovered from the vagina of healthy and ill bitches, including animals treated with antibiotics, suggesting that the canine vaginal lactic acid microbiota is stable under a variety of conditions.

Statistical analysis showed no correlation between physiological variables such as vaginal pH and body temperature with the presence of LAB. In our population of bitches, the vaginal pH in the healthy dogs ranged from 6.0 to 7.5, while in humans, where lactobacilli are predominant members of the microbiota, the normal vaginal pH is 4.5 or even lower (15). Hormonal status, physiological changes, or other components of the microbiota not investigated in the present study, could explain the differences in pH values.

Adhesion to the mucosal epithelium is an important property for a strain to be considered as a probiotic. This feature can favor repopulation of the vagina and control of pathogens (16). All the isolates that were studied adhered to CVEC to varying degrees depending on the strain, with Lactobacillus 12C1 being the most adhesive. Antimicrobial activity against common vaginal and urinary pathogens could be explained by diffusible substances produced by this bacterial group (1). All isolates assayed inhibited the in vitro growth of either gram-positive or gram-negative pathogens, suggesting that potential probiotic strains could contribute to the balance of the normal vaginal biota (17).

This is the first study that reports the presence of L. plantarum, L. murinus, and E. canintestini in the canine vagina. These species are components of the normal intestinal biota (18). Their presence in the canine vagina suggests the occurrence of an intestine-to-vagina transfer of the microbiota as has been described in other species (15,19) but further studies are needed to confirm this hypothesis.

The present study represents an approach to understanding aspects of the bitch vaginal lactic microbiota and the potential of the vagina as a source of beneficial organisms with probiotic properties.