Download Characteristics of Lactobacillus and Gardnerella vaginalis from

Journal of Medical Microbiology (2012), 61, 1074–1081
DOI 10.1099/jmm.0.041962-0
Characteristics of Lactobacillus and Gardnerella
vaginalis from women with or without bacterial
vaginosis and their relationships in gnotobiotic mice
G. S. Teixeira,1 F. P. Carvalho,1 R. M. E. Arantes,2 A. C. Nunes,3
J. L. S. Moreira,3 M. Mendonça,4 R. B. Almeida,4 L. M. Farias,1
M. A. R. Carvalho1 and J. R. Nicoli1
Correspondence
J. R. Nicoli
[email protected]
1
Departamento de Microbiologia, Instituto de Ciências Biológicas, Universidade Federal de Minas
Gerais, Belo Horizonte, MG, Brazil
2
Departamento de Patologia Geral, Instituto de Ciências Biológicas, Universidade Federal de Minas
Gerais, Belo Horizonte, MG, Brazil
3
Departamento de Biologia Geral, Instituto de Ciências Biológicas, Universidade Federal de Minas
Gerais, Belo Horizonte, MG, Brazil
4
Departamento de Ginecologia e Obstetrı́cia, Faculdade de Medicina, Universidade Federal de
Minas Gerais, Belo Horizonte, MG, Brazil
Received 22 December 2011
Accepted 22 April 2012
The objectives of the present study were to evaluate in vitro the production of antagonistic
compounds against Gardnerella vaginalis by Lactobacillus strains isolated from women with or
without bacterial vaginosis (BV), and to select one of the better Lactobacillus producers of such a
substance to be tested in vivo using a gnotobiotic animal model challenged with one of the more
sensitive G. vaginalis isolates. A total of 24 isolates from women with and without BV were identified
as G. vaginalis. A higher frequency (P,0.05) of this bacterium was observed in women with BV
(56.7 %) when compared to healthy women (17.6 %). A total of 86 strains of Lactobacillus were
obtained from healthy women and women with BV. Lactobacillus strains were more frequently
present (P,0.05) in healthy women (97.5 %) than in women with BV (76.7 %). Lactobacillus
crispatus was the predominating strain in both healthy women and women with BV. Lactobacillus
jensenii, Lactobacillus johnsonii, Lactobacillus gasseri and Lactobacillus vaginalis were isolated
with an intermediate frequency in the two groups. In vitro antagonism assays were performed using
as indicators 17 reference strains and the G. vaginalis strains isolated from women with BV and from
healthy women. Lactobacillus isolated from healthy women showed the higher antagonistic activity
against all the indicator strains when compared with isolates from women with BV. Concerning the
indicator strains, G. vaginalis found in women with BV was more resistant to the antagonism,
particularly when Lactobacillus isolates from women with BV were used as producer strains. A high
vaginal population level of G. vaginalis was obtained by intravaginal inoculation of germ-free mice,
and this colonization was accompanied by vaginal histopathological lesions. A tenfold decrease in
vaginal population level of G. vaginalis and a reduction of histological lesions were observed when
the pathogenic challenge was performed in mice previously monoassociated with an L. johnsonii
strain. Concluding, results of the present study suggest that progression of G. vaginalis-associated
BV depends in part on a simultaneous presence of Lactobacillus populations with a low antagonistic
capacity and of a G. vaginalis strain with a high resistance to this antagonism. The results could also
explain why G. vaginalis is frequently found in the vaginal ecosystem of healthy women.
INTRODUCTION
Bacterial vaginosis (BV) is a polymicrobial syndrome where
indigenous Lactobacillus populations, which are usually
Abbreviation: BV, bacterial vaginosis.
1074
dominant in the vagina of healthy women (Hillier, 2005),
are replaced by a mixture of bacteria that generally includes
Gardnerella vaginalis, Prevotella/Bacteroides species, Peptostreptococcus species, Mycoplasma hominis, Ureaplasma
urealyticum and/or Mobiluncus species (Atassi et al., 2006).
All of these bacteria can be present in low populations in
Downloaded from www.microbiologyresearch.org by
041962 G 2012 SGM
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
Printed in Great Britain
Relationships between Lactobacillus and G. vaginalis
healthy women, but during BV their concentrations are
generally increased 100–1000-fold over normal levels
(Eschenbach, 2007).
Factors that lead to the decline of the lactobacilli and the
overgrowth of an atypical microbiota, as well as the sequence
of bacterial population changes leading to BV, remain
unknown. Microbial interactions, including synergism,
commensalism and antagonism, can probably modify the
environment so that it becomes adverse for some microorganisms by production of inhibitory compounds, but
suitable for other organisms by releasing specific growth
factors. Pybus & Onderdonk (1999) described in vitro
nutritional inter-relationships that can explain commensalism between Prevotella bivia and both G. vaginalis and
Peptostreptococcus anaerobius. However, production of
antagonistic substances may also be an important factor in
the competitive colonization of the vaginal ecosystem.
The lactobacilli metabolize glucose essentially into lactic
acid, which contributes to the maintenance of a low vaginal
pH (4.0–4.5) and reduces the growth of most pathogenic
micro-organisms (Aroutcheva et al., 2001). Many isolates of
vaginal lactobacilli also produce H2O2, a compound having
broad antimicrobial activity (Cherpes et al., 2008). Among
other antagonistic mechanisms, bacteriocinogenic activity is
one of the most studied, and seems to contribute to the intraand inter-regulation of the human microbiota, influencing
microbial invasion and defence (Riley & Wertz, 2002).
G. vaginalis is a variable Gram-staining, facultatively anaerobic, non-motile, rod-shaped bacterium (Jarosik et al., 1998),
and is commonly found in the vaginal mucosa of asymptomatic women, but shows high concentrations in BV
(Ingianni et al., 1997). Recently, our laboratory reported
the production of synergistic and/or antagonistic compounds by G. vaginalis (Teixeira et al., 2010). In a second
study, we confirmed the importance of high population
levels of Lactobacillus (essentially Lactobacillus crispatus) and
the quality of these lactobacilli as producers of antagonistic
substances for a healthy vaginal environment (Branco et al.,
2010). These results showed that the quantity of beneficial
bacteria as well as their qualities as producers of antagonistic
compounds seem to be pivotal factors in understanding BV
and the ecological role of these bacteria in the vaginal
environment. However, these results were obtained in in vitro
experiments, and it is not certain whether it occurs in vivo.
The confirmation of the antagonistic phenomenon in an
animal model could lead to the selection and development of
probiotic Lactobacillus strains for the prevention and/or
treatment of BV (Kaewsrichan et al., 2006). The likely
contribution of this mechanism is difficult to determine in
the wider presence of a complex vaginal microbial ecosystem.
For these reasons, the gnotobiotic mouse provides an in vivo
simplified system that allows the observation of ecological
interactions in the vaginal tissue among few microbial strains
inoculated in this ecosystem.
The objectives of the present study were to evaluate in vitro
the production of antagonistic compounds against G.
http://jmm.sgmjournals.org
vaginalis by Lactobacillus strains isolated from women with
or without BV, and to select one of the best Lactobacillus
producers of such a substance to be tested in vivo using a
gnotobiotic animal model challenged with one of the most
sensitive G. vaginalis isolates.
METHODS
Ethical aspects. This project was approved by the Ethics Committee
of the Federal University of Minas Gerais for the experiments with
humans (COEP, protocol 062/03) and written informed consent was
obtained from all subjects before inclusion in the study. The study was
also approved by the Ethics Committee in Animal Experimentation of
the Federal University of Minas Gerais (CETEA/UFMG, protocol no.
227/2009).
Patients. The bacterial samples were isolated from healthy women
(n540) and patients with BV (n530). Patients were screened at the
Serviço de Ginecologia, Hospital das Clı́nicas, Universidade Federal
de Minas Gerais, Belo Horizonte, Brazil. Bacterial vaginosis was
diagnosed when three out of four Amsel criteria were observed
(Amsel et al., 1983). Inclusion criteria were: women 18–45 years old,
eumenorrheic, with normal blood glucose levels and with or without
BV. The use of oral or topical antimicrobials 30 days before the
sampling, pregnancy, menstruation, virginity and women in the
puerperal phase or under immunosuppressive therapy were considered as exclusion criteria.
Vaginal sampling and sample processing. Vaginal samples were
obtained using two 10 ml sterile loops which, after sampling, were
introduced under a CO2 flux in a tube containing 1 ml Ringer-PRAS
solution and in a tube containing 1 ml Gardnerella transport medium,
respectively. The samples were processed in a time range from 2 to 4 h
after collection. The samples were introduced into an anaerobic
chamber (Forma Scientific) containing an atmosphere of 85 % N2,
10 % H2 and 5 % CO2, and submitted to serial decimal dilutions until
1026. Aliquots of 0.1 ml from dilutions were spread onto Vaginalis agar
(using Columbia agar as base; Difco) and de Man, Rogosa and Sharp
agar (MRS; Merck). After 48 h to 7 days of incubation at 37 uC,
bacterial counts were determined and expressed as log10 [c.f.u. (ml
vaginal fluid)21]. Colonies with distinct morphology were isolated on
the respective medium. The bacterial samples were maintained at
280 uC in medium supplemented with 10 % glycerol.
Animals. Germ-free 28-day-old female mice (NIH, Taconic,
Germantown, USA) were used in this study. The animals were
housed in flexible plastic isolators (Standard Safety Equipment) and
handled according to established procedures. Experiments with
gnotobiotic mice were carried out in micro-isolators (UNO
Roestvaststaal). Water and commercial autoclavable diet (Nuvital)
were sterilized by steam and administered ad libitum, and animals
were maintained in an animal house with controlled lighting (12 h
light, 12 h dark). All experimental procedures were carried out
according to the standards set forth in the ‘Colegio Brasileiro de
Experimentação Animal’ rules (COBEA, 2006).
G. vaginalis isolates. The isolates from Vaginalis agar were
identified as G. vaginalis according to Piot et al. (1982), using the
results from Gram stain, catalase and oxidase tests, starch and
hippurate hydrolysis, and a- and b-galactosidase activity. Activity of
fructose-6-phosphate phosphoketolase was also evaluated (Orban &
Patterson, 2000).
Lactobacillus isolates. Bacteria isolated from the subjects were
selected as presumptive Lactobacillus based on the following
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
1075
G. S. Teixeira and others
characteristics: Gram-positive, microaerophilic and catalase-negative
rods isolated on MRS agar (Merck) at 37 uC. The selected strains were
identified by amplified rDNA restriction analysis as described by
Moreira et al. (2005).
DNA extraction. Chromosomal DNA was isolated from overnight
cultures of all isolates in 10 ml MRS broth. After washing the cells
with deionized water, the pellet was obtained by centrifugation at
14 000 g for 5 min at 4 uC, suspended in 1 ml 5 M LiCl, and
incubated for 1 h with constant shaking. After a second washing with
1 ml deionized water, the pellet was suspended in 1 ml protoplasting
buffer (50 mM Tris/HCl, pH 8.0; 10 mM EDTA; 10 mg lysozyme
ml21; 100 mg RNase ml21). After incubation for 1 h at 37 uC and
centrifugation at 14 000 g for 5 min at 4 uC, the pellet was suspended
in 500 ml protoplasting buffer without lysozyme, and 100 ml 10 % SDS
was added to allow cells to lyse. After lysis, the mixture was extracted
once with phenol, phenol–chloroform–isoamyl alcohol (25 : 24 : 1)
and chloroform–isoamyl alcohol (24 : 1). After 2-propanol precipitation, the DNA was dissolved in 100 ml TE buffer (10 mM Tris/HCl,
1 mM EDTA, pH 8.0).
PCR amplification of the 16S–23S rRNA gene intergenic
spacer. The 16S–23S rRNA gene intergenic spacer region was
amplified using the primer 16-1A (GAATCGCTAGTAATCG), which
anneals to a conserved region of the 16S rRNA genes, and primer 23-1B
(GGGTTCCCCCATTCGGA), which anneals to a conserved region of
the 23S rRNA genes, using a PTC-100 Thermal cycler (MJ Research).
The reaction mixture (50 ml) contained 10 pM of each primer, 0.2 mM
of each deoxyribonucleotide triphosphate, reaction buffer containing
1.5 mM MgCl2, 5 U Taq DNA polymerase (Phoneutria Biotecnologia
& Serviços) and 5 ml template DNA solution. The amplification
program was 95 uC for 2 min, 35 cycles of 95 uC for 30 s, 55 uC for
1 min, 72 uC for 1 min, and finally 72 uC for 10 min. PCR products
were electrophoresed on a 1.4 % agarose gel and visualized by UV
transillumination after staining with an ethidium bromide solution
(5 mg ml21).
Amplified rDNA restriction analysis. The amplified 16S–23S rRNA
intergenic spacer regions of lactic acid bacteria were digested with a set
of 12 enzymes chosen after compilation of nucleotide sequences already
deposited at GenBank and in silico restriction digestion using the
Webcutter 2.0 tool (Heiman, 1997; http://rna.lundberg.gu.se/cutter2/).
SphI, NcoI and NheI enzymes hydrolysed inside the 16S rRNA gene;
SspI, SfuI, DraI, VspI, HincII and EcoRI enzymes hydrolysed inside the
intergenic region; and AvrII and HindIII enzymes hydrolysed inside the
23S rRNA gene. EcoRV enzyme hydrolysed inside the spacer region for
the Lactobacillus casei group and in the 23S rRNA gene for the
Lactobacillus acidophilus group. For several lactobacillus species no
spacer nucleotide sequences have been reported and only fragments of
the 16S and/or 23S rRNA genes were found. All restriction enzymes
were purchased from Promega.
In vitro antagonism assay. The isolated lactobacilli were cultured in
MRS broth for 24 h at 37 uC in the anaerobic chamber (Forma
Scientific). After growth, an aliquot (5 ml) of the culture was spotted
onto MRS agar (Difco). After incubation at 37 uC for 48 h under
anaerobic conditions, the cells were killed by exposure to chloroform
for 20 min. Residual chloroform was allowed to evaporate and Petri
dishes were overlaid with 3.5 ml brain heart infusion (BHI) or
Sabouraud soft agar (0.7 %) (Difco), which had been inoculated with
0.2 ml of a 24 h culture of the indicator strain to be tested. All the G.
vaginalis isolates from women with or without BV were used as
indicator strains, as well as Enterococcus faecalis ATCC 19433,
Staphylococcus aureus ATCC 33591, Clostridium difficile ATCC 9689,
Clostridium perfringens ATCC 3624, Listeria monocytogenes ATCC
1531, Bacteroides fragilis ATCC 25923, Bacteroides vulgatus ATCC 8482,
Fusobacterium nucleatum ATCC 10953, Fusobacterium necrophorum
1076
ATCC 25286, Peptostreptococcus anaerobius ATCC 27337, Prevotella
melaninogenica ATCC 24845, Prevotella nigrescens ATCC 33569,
Porphyromonas gingivalis FDC 381, Bifidobacterium bifidum ATCC
29521, Bifidobacterium longum ATCC 15707, G. vaginalis ATCC 14018
and Candida albicans ATCC 18804. After 24 h of incubation at 37 uC
under aerobic or anaerobic conditions depending on the indicator
strain, the antagonistic activity was determined as the presence of a
growth inhibition zone around the spot.
Determination of H2O2 production. To test for the production of
H2O2 by lactobacillus isolates, a qualitative assay on tetramethylbenzidine (TMB; Sigma) agar plates was done as described by Rabe &
Hillier (2003). Lactobacilli were inoculated onto TMB agar plates and
incubated in the anaerobic chamber at 37 uC. After 40 h, the plates
were exposed to ambient air. If the lactobacilli produced H2O2, a
reaction occurred with horseradish peroxidase (Sigma) present in the
medium, which oxidizes TMB, causing the colonies to turn blue.
Lactobacillus acidophilus ATCC 4356 was used as a control.
In vivo antagonism assay. The Lactobacillus johnsonii strain used in
this assay was selected based on its high antagonistic activity, and the
G. vaginalis strain on its high sensitivity to this antagonism. In a first
experiment, two groups of germ-free mice were submitted to the
following protocols: group LJGV received by intravaginal inoculation
a unique dose of 0.1 ml containing 9.0 log c.f.u. L. johnsonii ml21,
and 10 days later the monoxenic animals were inoculated intravaginally with 0.1 ml of a suspension containing 7.0 log c.f.u. G.
vaginalis ml21. The second group, GV, received only sterile saline by
vaginal application 10 days before the pathogenic challenge. In a
second experiment, two groups were submitted to the same protocols
described above, but both were also injected subcutaneously with two
doses of 0.5 ml oestradiol benzoate (0.5 mg in 0.1 ml peanut oil) 72 h
before the challenge and 72 h after the challenge, respectively (LJGVH
and GVH). One week after the challenge, all the mice were sacrificed
by cervical dislocation and vaginal tissues were removed. In a third
experiment, two groups were treated only with the lactobacillus, with
or without the hormonal treatment, and were sacrificed 17 days after
the monoassociation with the lactobacillus (LJ and LJH). Three germfree mice were used in each group.
Part of the tissue was weighed and suspended in a volume of sterile PBS
to obtain, after homogenization, a dilution of 1021 (w/v). Then, this
suspension was introduced into the anaerobic chamber and submitted
to serial decimal dilutions in PBS. A 0.1 ml aliquot of the dilutions was
spread onto Vaginalis agar and/or MRS agar, and incubated for 48 h at
37 uC for bacterial count. All the determinations were performed in
triplicate.
Another part of the vaginal tissue was fixed in buffered 4 %
formaldehyde and processed for paraffin embedding. The histopathological sections (3–5 mm) were stained with haematoxylin and eosin
(H&E). The slides were coded and examined by a single pathologist,
who was unaware of the experimental conditions of each group.
Statistical analysis. Data were analysed using the Minitab Release
14.20 software. Fisher’s exact test and Kruskal–Wallis one-way
analysis of variance followed by pair-wise multiple comparisons
using the Student–Newman–Keuls method were used. P,0.05 was
considered to be statistically significant.
RESULTS
A total of 24 isolates from women with (17 isolates) and
without (7 isolates) BV were identified as G. vaginalis. Table
1 shows a higher frequency (P,0.05) of this bacterium in
women with BV (56.7 %) when compared to healthy women
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
Journal of Medical Microbiology 61
Relationships between Lactobacillus and G. vaginalis
Table 1. Frequency (%) of women with or without BV presenting at least one isolate of Lactobacillus spp. or Gardnerella vaginalis,
and population levels of these bacteria (mean log c.f.u. ml”1±SD) in their respective vaginal ecosystem
Bacterium (total number of isolates)
Gardnerella vaginalis (24)
Total Lactobacillus spp. (86)
L. crispatus (29)
L. jensenii (11)
L. gasseri (10)
L. johnsonii (12)
L. vaginalis (6)
L. fermentum (3)
L. salivarius (4)
L. reuteri (2)
L. acidophilus (3)
L. delbrueckii (1)
L. colehominis (1)
L. hilgardii (1)
Lactobacillus spp. (3)
Women with BV (30)
Healthy women (40)
P
56.7
7.9±1.4
76.7
6.8±1.2
26.7
7.4±0.4
10.0
7.7±2.0
16.6
6.1±0.8
3.4
6.8
6.8
7.9
6.8
6.1
3.4
5.0
0
0
0
0
3.4
5.0
0
0
0
0
3.4
17.6
8.0±1.2
97.5
6.9±1.4
35.0
6.2±1.5
17.5
8.0±1.4
12.5
7.1±1.3
17.5
7.1±1.3
2.5
7.7
0
0
2.5
6.6
2.5
7.1
2.5
6.7
0
0
2.5
6.5
2.5
7.0
5.0
0.001*
0.834D
0.009*
0.644D
0.316*
0.099D
0.298*
0.798D
0.437*
0.198D
0.067*
*Fisher’s exact test.
DStudent t-test.
(17.6 %). However, when present, the vaginal population
levels of G. vaginalis were similar in both groups, and
reached about 8.0 log c.f.u. (ml vaginal fluid)21. A total of 86
strains of Lactobacillus were obtained from healthy women
(58 isolates) and women with BV (28 isolates). Lactobacillus
strains were more frequently present (P,0.05) in healthy
women (97.5 %) than in women with BV (76.7 %), but as
for G. vaginalis, vaginal population levels were similar in
both groups and reached about 6.8 log c.f.u. (ml vaginal
fluid)21. Lactobacillus crispatus was the predominating
strain in both healthy women (22 isolates) and women
with BV (7 isolates). Lactobacillus jensenii, L. johnsonii,
Lactobacillus gasseri and Lactobacillus vaginalis were isolated
with an intermediate frequency in the two groups.
A total of 2549 assays of antagonism were performed in the
present study, using as indicators the reference strains and
the G. vaginalis isolates from women with BV and from
healthy women. The Lactobacillus isolates from healthy
women showed a higher antagonistic activity against the
indicator strains when compared with isolates from women
http://jmm.sgmjournals.org
with BV (Table 2). Additionally, all the 24 G. vaginalis
isolates used as indicator were sensitive to at least one of the
Lactobacillus isolates from healthy women and tested in the
antagonistic assay (data not shown). However, 64.7 % of the
G. vaginalis isolated from patients with BV and 42.8 % of
those from healthy women were resistant to all the 17
Lactobacillus isolates from women with BV (data not
shown). Concerning the indicator strains, G. vaginalis found
in women with BV was more resistant to the antagonism,
particularly when Lactobacillus isolates from women with
BV were used as producer strains. However, a statistical
significance (P,0.05) was observed only when Lactobacillus
species from healthy women were tested. The higher
antagonistic activity frequency (of about 70 %) was observed
when L. johnsonii isolates were assayed against G. vaginalis,
both from healthy women.
Table 3 shows that 57.4 % of the Lactobacillus isolates
produced H2O2, with a slightly higher frequency (45.5 %) for
strains isolated from healthy women than from women with
BV (35.3 %). However, this difference was not statistically
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
1077
G. S. Teixeira and others
Table 2. Frequency (%) of antagonism of Lactobacillus crispatus, Lactobacillus jensenii, Lactobacillus gasseri and Lactobacillus
johnsonii isolated from healthy women and women with BV against the reference strains and the strains of Gardnerella vaginalis
isolated in the present study from healthy women and women with BV
Isolated Lactobacillus
Indicator strains
Reference strains (17)
Healthy women
L. crispatus (22)
L. jensenii (9)
L. gasseri (5)
L. johnsonii (9)
Women with BV
L. crispatus (7)
L. jensenii (2)
L. gasseri (5)
L. johnsonii (3)
P*
Isolated G. vaginalis (24)
Women with BV (17)
Healthy women (7)
24.6
17.0
21.2
30.1
32.6
30.7
22.3
31.4
57.8
44.4
54.3
69.8
1028
0.039
0.001
1027
12.6
8.8
12.9
17.6
7.6
5.9
4.7
3.9
16.3
21.4
8.6
14.3
0.079
0.140
0.332
0.113
*Values of P from Fisher’s exact test for comparisons of antagonism frequencies between the same isolates of one Lactobacillus species against G.
vaginalis isolated from healthy women or from women with BV.
significant (P.0.05), but this can be due to the too low
number of strains tested. L. johnsonii from healthy women
was the highest producer of H2O2 in terms of intensity of
production.
both bacteria reached vaginal population levels ranging from
4.0 to 5.0 log c.f.u. (g tissue)21. Curiously, a vaginal
colonization by L. johnsonii [between 2.0 and 3.0 log c.f.u. (g
vaginal tissue)21] was observed when the monoassociated
mice were posteriorly challenged with G. vaginalis (Fig. 1c).
A tenfold reduction in vaginal population levels of G.
vaginalis was observed when the pathogenic challenge was
performed in mice previously monoassociated with L.
johnsonii, and this reduction was statistically significant in
the mice not treated with hormone (from 5.01 to 3.62 log
c.f.u. g21; P50.009). A tendency for a similar effect was also
noted in animals treated with the hormone (from 4.40 to
3.34 log c.f.u. g21; P50.078).
For the in vivo study of the inter-relationships between
Lactobacillus and G. vaginalis, a highly antagonistic and
H2O2 producer strain of L. johnsonii from a healthy woman
was tested against a G. vaginalis isolate selected for its high
sensitivity to this antagonism. Fig. 1(a) shows that treatment with oestradiol benzoate was necessary for the
colonization of germ-free mice by L. johnsonii, but not for
G. vaginalis (Fig. 1b). When the colonization was obtained,
Table 3. Frequency (%) and intensity of H2O2 production by Lactobacillus crispatus, Lactobacillus jensenii, Lactobacillus gasseri
and Lactobacillus johnsonii isolated from healthy women and women with BV
Production of H2O2: 2, absence; +, low; ++, intermediate; +++, high; ++++, intense.
Presence and intensity of H2O2 production
Isolated Lactobacillus
Healthy women
L. crispatus (22)
L. jensenii (8)
L. gasseri (5)
L. johnsonii (9)
Women with BV
L. crispatus (7)
L. jensenii (2)
L. gasseri (5)
L. johnsonii (3)
Total frequency of positive and negative strains
1078
”
+
++
+++
++++
9
4
3
4
1
9
1
1
1
3
1
2
2
2
1
3
2
1
1
1
3
2
42.6
1
1
1
1
Frequency of positive strain
59.1
50.0
40.0
55.5
85.7
100.0
40.0
33.3
57.4
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
Journal of Medical Microbiology 61
Relationships between Lactobacillus and G. vaginalis
7
7
(b)
Lactobacillus
6
log [c.f.u. (g vaginal tissue)–1]
log [c.f.u. (g vaginal tissue)–1]
(a)
5
4
3
2
1
0
7
2
Mice
4
3
2
1
1
3
(d)
Lactobacillus
6
7
log [c.f.u. (g vaginal tissue)–1]
log [c.f.u. (g vaginal tissue)–1]
5
0
1
(c)
Gardnerella
6
5
4
3
2
1
0
2
Mice
3
Gardnerella
6
5
4
3
2
1
0
1
2
Mice
3
1
2
Mice
3
Fig. 1. Individual vaginal population levels [log c.f.u. (g tissue)”1] of Lactobacillus johnsonii (a, c) and Gardnerella vaginalis
(b, d) in mice only monoassociated (a, b) or diassociated (c, d), with (&) or without (h) hormone treatment. n53.
Histological examination of the vaginal mucosa from germfree mice showed no inflammation or hyperaemia of the
epithelium. In mice only monoassociated with L. johnsonii, a
discrete inflammatory infiltrate was observed, as well as a
direct transition from columnar to mature squamous
epithelium. In mice challenged with G. vaginalis, histopathological examination showed inflammatory alterations
of the ectocervix and lamina propria epithelium. In mice
pre-treated with L. johnsonii before the challenge with G.
vaginalis, the inflammatory infiltrate and oedema were less
pronounced. In these mice, no differences were observed in
relation to the germ-free animal for the endocervical
epithelium aspects. Table 4 presents the results obtained
from the histopathological examination and shows a clear
Table 4. Histopathological score for vaginal mucosa of mice only challenged with Gardnerella vaginalis (treated, GVH, or not, GV,
with hormone), only associated during 17 days with Lactobacillus johnsonii (treated, LJH, or not, LJ, with hormone) or mice
associated during 10 days with Lactobacillus johnsonii and then challenged with Gardnerella vaginalis (treated, LJGVH, or not,
LJGV, with hormone)
Score: 2, absence; +, low; ++, intermediate; +++, high; ++++, intense.
Histopathological aspect of vaginal mucosa
Reactive area
Inflammatory alterations
Eosinophil presence
Interstitial oedema
http://jmm.sgmjournals.org
Score in each experimental group
GVH
GV
LJH
LJ
LJGVH
LJGV
++
++
++
+
+++
+++
++++
++
2
++
2
2
++
2
2
2
2
+
2
+
2
+
+
+
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
1079
G. S. Teixeira and others
protective effect against the G. vaginalis infection when
animals were pre-treated with L. johnsonii.
DISCUSSION
The presence of a high population of Lactobacillus species is
well known as an important factor for a healthy vaginal
ecosystem (Hillier, 2005; Falagas et al., 2007). However,
beyond their high local levels, these lactobacilli must also be
highly antagonistic against potential pathogenic microorganisms to offer protection. There is little information
about the importance of the sensitivity of the pathogenic
target to this antagonism. Additionally, the antagonistic
property is generally demonstrated by in vitro assay, and it is
not certain whether this ability can be extrapolated to in vivo
conditions.
The present study confirms that Lactobacillus species are more
frequently found in the vaginal ecosystem of healthy subjects
than in patients with BV (Mikamo et al., 2000; Eschenbach,
2007; Klomp et al., 2008). Moreover, as described in other
studies (Hellberg et al., 2001; Branco et al., 2010; Srinivasan
et al., 2010), G. vaginalis frequency was higher in patients with
BV when compared to healthy women. However, when
present in patients with BV, Lactobacillus levels were similar to
those found in healthy women, suggesting that a beneficial
effect of these bacteria depends not only on their presence in
high quantities but also on their quality in term of protective
ability. These results were different from those in some studies
which observed higher levels of Lactobacillus in healthy
women (Mikamo et al., 2000; Eschenbach, 2007; Klomp et al.,
2008). Similarly, G. vaginalis population levels when present
in healthy women were as high as in patients with BV,
suggesting on the other hand that pathological symptoms of
infection depend also on the pathogenic capacity of the
bacteria and not only on their quantities.
demonstrated in vitro for vaginal Lactobacillus against
pathogens. For such finality, it is necessary to obtain
animals which can be colonized by human lactobacilli and
infected by human G. vaginalis. The progression of
colonization/infection of body epithelial surfaces by a
micro-organism depends mainly on its resistance to the
local conditions (such as the mucosal barrier integrity, the
indigenous microbiota balance and the immunological
status of the host), its ability to use the available nutrients
and its capacity to adhere to the mucosa. Few studies
describe experimental vaginal inoculation of micro-organisms in animal models, and none of them used germ-free
animals for such finality (Johnson et al., 1984; Fidel et al.,
2000; de Ruiz et al., 2001; Hamad et al., 2004). In these
animal models, the authors recommended an oestrogen
administration to induce a pre-oestral stage, which ensures a
successful colonization/infection. In healthy human vaginas,
it is well known that the presence and number of lactobacilli
are influenced by oestrogen production (Keane et al., 1997).
Oestrogen converts columnar epithelium into a thick layer
of squamous stratified epithelium and increases the glycogen
content and other substrates for bacterial growth. As
expected from this information, vaginal colonization of
germ-free mice by the selected L. johnsonii strain was
observed in the present study only in animals treated with
the hormone. However, this treatment was not necessary for
the G. vaginalis colonization. Interestingly, vaginal colonization by L. johnsonii was obtained without hormonal
treatment when the mice were also challenged with G.
vaginalis. However, the L. johnsonii population levels were
higher in this diassociated model when the mice were treated
with oestradiol when compared to the non-treated group.
We have no strong explanation for this phenomenon, but
the presence of G. vaginalis could provide a synergistic effect
for the growth of L. johnsonii similar to that described for
Peptostreptococcus anaerobius in a recent study in our
laboratory (Teixeira et al., 2010). In any case, the previous
monoassociation with L. johnsonii protected the mice
against the vaginal challenge with G. vaginalis as demonstrated by both lower population levels of the pathogen and
the reduced lesions observed from the histopathological
examination. However, it was not possible to determine in
this animal model whether the protection was due to the
production of antagonistic compounds observed in the in
vitro assays or to other mechanisms such as co-aggregation
or competition for shared adhesion receptors or nutrients
between the lactobacillus and the pathogen.
Many studies have suggested that the presence of H2O2producing vaginal lactobacilli may protect against BV,
although some studies do not support this hypothesis
(Falagas et al., 2007). In the present study, L. johnsonii from
healthy women showed better antagonistic properties and
H2O2 production when compared to the other Lactobacillus
species. In fact, the production of lactic acid by lactobacilli,
which is mainly responsible for the low vaginal pH, could
contribute more than the production of H2O2 to the
inhibition of the G. vaginalis growth (McLean & McGroarty,
1996). However, in the present study, experiments determining the most acidic culture supernatant after growth of
the various Lactobacillus isolates have not been performed to
correlate this information with their antagonistic abilities.
The most interesting information described here is related to
the need for a high Lactobacillus inhibitory capacity combined with a high G. vaginalis susceptibility to direct the
vaginal ecosystem to a healthy status and inversely for the
BV status.
ACKNOWLEDGEMENTS
No reports are available in the literature using animal
models to confirm in vivo the antagonistic properties
The authors are grateful to Clélia Nunes da Silva for valuable
technical help and to Antônio Mesquita Vaz for animal care. This
1080
Concluding, results of the present study suggest that the
success of BV due to G. vaginalis depends, at least in part,
on a simultaneous presence of Lactobacillus populations
with low antagonistic capacity and of a G. vaginalis strain
with a high resistance to this antagonism.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
Journal of Medical Microbiology 61
Relationships between Lactobacillus and G. vaginalis
work was supported by grants from Conselho Nacional de
Desenvolvimento Cientı́fico e Tecnológico (CNPq) and Fundação
de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG).
5041–5047.
REFERENCES
Johnson, A. P., Ison, C. A., Hetherington, C. M., Osborn, M. F.,
Southerton, G., London, W. T., Easmon, C. S. & Taylor-Robinson, D.
(1984). A study of the susceptibility of three species of primate to vaginal
Jarosik, G. P., Land, C. B., Duhon, P., Chandler, R., Jr & Mercer, T.
(1998). Acquisition of iron by Gardnerella vaginalis. Infect Immun 66,
colonization with Gardnerella vaginalis. Br J Exp Pathol 65, 389–396.
Amsel, R., Totten, P. A., Spiegel, C. A., Chen, K. C., Eschenbach, D. &
Holmes, K. K. (1983). Nonspecific vaginitis. Diagnostic criteria and
microbial and epidemiologic associations. Am J Med 74, 14–22.
Aroutcheva, A., Gariti, D., Simon, M., Shott, S., Faro, J., Simoes, J. A.,
Gurguis, A. & Faro, S. (2001). Defense factors of vaginal lactobacilli.
Am J Obstet Gynecol 185, 375–379.
Atassi, F., Brassart, D., Grob, P., Graf, F. & Servin, A. L. (2006).
Lactobacillus strains isolated from the vaginal microbiota of healthy
women inhibit Prevotella bivia and Gardnerella vaginalis in coculture
and cell culture. FEMS Immunol Med Microbiol 48, 424–432.
Branco, K. M. G. R., Nardi, R. M. D., Nunes, A. C., Farias, L. M., Nicoli,
J. R. & Carvalho, M. A. R. (2010). Identification and in vitro
production of Lactobacillus antagonists from women with or without
bacterial vaginosis. Braz J Med Biol Res 43, 338–344.
Cherpes, T. L., Hillier, S. L., Meyn, L. A., Busch, J. L. & Krohn, M. A.
(2008). A delicate balance: risk factors for acquisition of bacterial
Kaewsrichan, J., Peeyananjarassri, K. & Kongprasertkit, J. (2006).
Selection and identification of anaerobic lactobacilli producing
inhibitory compounds against vaginal pathogens. FEMS Immunol
Med Microbiol 48, 75–83.
Keane, F. E. A., Ison, C. A. & Taylor-Robinson, D. (1997). A
longitudinal study of the vaginal flora over a menstrual cycle. Int J
STD AIDS 8, 489–494.
Klomp, J. M., Verbruggen, B. S., Korporaal, H., Boon, M. E., de Jong, P.,
Kramer, G. C., van Haaften, M. & Heintz, A. P. (2008). Gardnerella
vaginalis and Lactobacillus sp in liquid-based cervical samples in healthy
and disturbed vaginal flora using cultivation-independent methods.
Diagn Cytopathol 36, 277–284.
McLean, N. W. & McGroarty, J. A. (1996). Growth inhibition of
metronidazole-susceptible and metronidazole-resistant strains of
Gardnerella vaginalis by Lactobacilli in vitro. Appl Environ Microbiol
62, 1089–1092.
vaginosis include sexual activity, absence of hydrogen peroxideproducing lactobacilli, black race, and positive herpes simplex virus
type 2 serology. Sex Transm Dis 35, 78–83.
Vaginal microflora in healthy women with Gardnerella vaginalis.
J Infect Chemother 6, 173–177.
COBEA (2006). Legislação e Ética. http://www.cobea.org.br/.
Accessed on March, 2009.
Moreira, J. L., Mota, R. M., Horta, M. F., Teixeira, S. M., Neumann, E.,
Nicoli, J. R. & Nunes, A. C. (2005). Identification to the species level of
de Ruiz, C. S., Rey, M. R., de Ruiz Holgado, A. P. & Nader-Macı́as,
M. E. (2001). Experimental administration of estradiol on the
Lactobacillus isolated in probiotic prospecting studies of human,
animal or food origin by 16S-23S rRNA restriction profiling. BMC
Microbiol 5, 15.
colonization of Lactobacillus fermentum and Escherichia coli in the
urogenital tract of mice. Biol Pharm Bull 24, 127–134.
Eschenbach, D. A. (2007). Bacterial vaginosis: resistance, recurrence,
and/or reinfection? Clin Infect Dis 44, 220–221.
Falagas, M. E., Betsi, G. I. & Athanasiou, S. (2007). Probiotics for the
treatment of women with bacterial vaginosis. Clin Microbiol Infect 13,
657–664.
Fidel, P. L., Jr, Cutright, J. & Steele, C. (2000). Effects of reproductive
hormones on experimental vaginal candidiasis. Infect Immun 68, 651–
657.
Hamad, M., Abu-Elteen, K. H. & Ghaleb, M. (2004). Estrogen-
dependent induction of persistent vaginal candidosis in naı̈ve mice.
Mycoses 47, 304–309.
Heiman, M. (1997). Webcutter 2.0. http://rna.lundberg.gu.se/cutter2/.
Hellberg, D., Nilsson, S. & Mårdh, P. A. (2001). The diagnosis of
bacterial vaginosis and vaginal flora changes. Arch Gynecol Obstet 265,
11–15.
Hillier, S. L. (2005). The complexity of microbial diversity in bacterial
vaginosis. N Engl J Med 353, 1886–1887.
Ingianni, A., Petruzzelli, S., Morandotti, G. & Pompei, R. (1997).
Genotypic differentiation of Gardnerella vaginalis by amplified
ribosomal DNA restriction analysis (ARDRA). FEMS Immunol Med
Microbiol 18, 61–66.
http://jmm.sgmjournals.org
Mikamo, H., Sato, Y., Hayasaki, Y., Hua, Y. X. & Tamaya, T. (2000).
Orban, J. I. & Patterson, J. A. (2000). Modification of the
phosphoketolase assay for rapid identification of bifidobacteria.
J Microbiol Methods 40, 221–224.
Piot, P., Van Dyck, E. &, Totten, P. A. & Holmes, K. K. (1982).
Identification of Gardnerella (Haemophilus) vaginalis. J Clin Microbiol
15, 19–24.
Pybus, V. & Onderdonk, A. B. (1999). Microbial interactions in the
vaginal ecosystem, with emphasis on the pathogenesis of bacterial
vaginosis. Microbes Infect 1, 285–292.
Rabe, L. K. & Hillier, S. L. (2003). Optimization of media for detection
of hydrogen peroxide production by Lactobacillus species. J Clin
Microbiol 41, 3260–3264.
Riley, M. A. & Wertz, J. E. (2002). Bacteriocins: evolution, ecology, and
application. Annu Rev Microbiol 56, 117–137.
Srinivasan, S., Liu, C., Mitchell, C. M., Fiedler, T. L., Thomas, K. K.,
Agnew, K. J., Marrazzo, J. M. & Fredricks, D. N. (2010). Temporal
variability of human vaginal bacteria and relationship with bacterial
vaginosis. PLoS ONE 5, e10197.
Teixeira, G. S., Soares-Brandão, K. L. K., Branco, K. M. G. R.,
Sampaio, J. L. M., Nardi, R. M. D., Mendonça, M., Almeida, R. B.,
Farias, L. M., Carvalho, M. A. R. & Nicoli, J. R. (2010). Antagonism and
synergism in Gardnerella vaginalis strains isolated from women with
bacterial vaginosis. J Med Microbiol 59, 891–897.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 05:59:36
1081