Identification of culturable vaginal Lactobacillus species among reproductive age women in Mysore, India

Abstract

A healthy vaginal environment is predominated by certain Lactobacillus species, which lead to the prevention of infections of the reproductive tract. This study examined the characteristics of cultivable Lactobacillus species in both healthy women and women with bacterial vaginosis (BV). Between November 2011 and September 2013, 139 women attending a women's clinic in Mysore, India, were evaluated for BV in a cross-sectional study. BV was diagnosed using Amsel's criteria: homogeneous vaginal discharge, vaginal pH >4.5, production of amines, and presence of “clue” cells. Those with three or more of the characteristics were considered to have BV. Vaginal swabs were then cultured in Rogosa agar and de Man-Rogosa-Sharpe broth. Gram-positive lactobacilli generating 600–800 bp amplicons by16 sRNA were further characterized by sequencing. Cultivable vaginal samples were obtained from 132 women (94.9 %). According to the Amsel criteria, 83 women (62.1 %) were healthy, and 49 (37.1 %) had BV. Eleven different Lactobacillus species were isolated from 47 women. The common lactobacilli species found in this sample included L. crispatus (39.6 %), L. gasseri (45.8 %), and L. jensenii (14.6 %). Lactobacilli were isolated from 39 healthy women and eight with BV. L. gasseri was cultured from 18.8 % of healthy women and 6.1 % with BV. The presence of L. reuteri was significantly associated with normal vaginal microbiota (P-value = 0.026). These results further our understanding of vaginal lactobacilli colonization and richness in this particular population. Our findings showed that lactobacilli species present in the vaginas of healthy women in India do not differ from those reported from other countries.

Introduction

A growing body of research suggests that the vaginal microbiome has evolved to protect women against a variety of pathogens including parasitic, bacterial and viral infections (Nardis et al., 2013; Razzak et al., 2011; Rendón-Maldonado et al., 1998). Genus Lactobacillus, Gram-positive bacteria first identified by Döderlein in 1928, are the best studied of the facultative microbes found in the vaginal environment (van de Wijgert et al., 2014). Research using culture-based methods among women living mostly in industrialized countries first suggested that a healthy vaginal flora was dominated by four species: L. fermentum; L. brevis; L. jensenii; and L. casei (Antonio et al., 1999). More recently, studies using molecular methods in similar populations have narrowed the predominant species seen in the vaginal mucosa of most reproductive age women to L. crispatus, L. iners and L. gasseri (Linhares et al., 2010).

There are only a few studies characterizing the vaginal Lactobacillus species found among healthy women of reproductive age in India. A study among 80 women of reproductive age carried out by Garg et al. (2009) found the predominant species isolated were L. reuteri, present in 32.5 % (n = 26), L. fermentum in 25 % (n = 20), and L. salivarius in 16.3 % (n = 13) of women (Garg et al., 2009). They suggested that this may mean that the species found in the vaginal microbiome of healthy Indian women may vary widely from those found in women from other countries (Antonio et al., 1999). Findings from a pilot study carried out by our group using 16 sRNA in vaginal samples obtained from 40 healthy women of reproductive age (20 from each country) from Mysore, India, and San Francisco, USA, showed instead that the vaginal Lactobacillus species in both groups were similar, with L. jensenii and L. crispatus dominating all other species (Madhivanan et al., 2014).

Growing evidence also shows that bacterial vaginosis (BV) leads to changes in the vaginal microbiome (Ma et al., 2012). BV has been shown to cause the loss of L. crispatus species, and has been associated with a more complex ecology of bacterial communities. Women with BV also have fewer vaginal Lactobacillus species, with the most common being L. iners, BVAB2, L. crispatus, L. jensenii, L. reuteri and L. coleohominis (Madhivanan et al., 2014; Tamrakar et al., 2007). Interestingly, our group found that when Indian women were compared with their counterparts in the USA, the Lactobacillus species found were more commonly obligatively heterofermentative, suggesting a need for more metabolic flexibility and a natural selection process that favoured species protective against a larger number of pathogenic threats (Madhivanan et al., 2014). The present study characterized the cultivatable species of Lactobacillus in a large sample of women of reproductive age with and without BV in Mysore, India.

Methods

Study population

Between November 2011 and September 2013, potential participants were recruited from a consecutive sample of women attending the reproductive health clinic in Mysore, India. This cross-sectional study characterized the vaginal Lactobacillus species in women with and without BV. Detailed methods are described elsewhere (Madhivanan et al., 2014). In brief, women were assessed for eligibility based on a series of questions. To be included in the study, participants had to be between the ages of 18 and 45 years, sexually active (defined as having had vaginal intercourse at least once in the three months prior to enrolment) and having the ability to provide informed consent. Women who were pregnant, having menses or a diagnosis of any reproductive tract infection (except BV) were excluded from the study. In addition, women who had used any antibiotics in the prior 30 days were also excluded. All eligible women were provided details about the study, and the research staff answered any questions they had before obtaining informed consent from each participant in Kannada, the language spoken by the majority of local residents.

Human subjects

The study was reviewed and approved by the Committee for the Protection of Human Subjects at the University of California, Berkeley; Florida International University and the Public Health Research Institute of India's Institutional Review Board, and conducted in compliance with all federal regulations governing the protection of human subjects. Eligible women who gave written informed consent were assigned a study identification number (SID). All subsequent documents were coded with the SID and had no personal identifiers.

Clinical evaluation for bacterial vaginosis

A trained physician conducted physical examinations of all study participants and collected biological specimens for detection of reproductive tract infections (RTIs). The diagnosis of BV was based on the Amsel criteria of having any three of the four following characteristics: a vaginal pH >4.5, the presence of a fishy amine odour upon addition of 10 % potassium hydroxide (KOH), a homogeneous white vaginal discharge and the presence of clue cells on wet-mount microscopy (Amsel et al., 1983). Three vaginal specimens were obtained by placing a cotton swab in the posterior fornix of the vagina. The first swab was used to measure the vaginal pH, smeared onto two microscope slides and then placed in a tube containing four drops of normal saline to prepare a slide for wet-mount microscopic examination. The remaining two swabs were placed into BBL Port-A-Cul tubes (Becton, Dickinson) and stored in a refrigerator until they were transported to the laboratory for further processing. Microscope slides were air-dried, fixed with methanol, and subsequently Gram-stained for Nugent scoring (Nugent et al., 1991).

The vaginal samples were classified according to Nugent scoring of Gram-stained vaginal smears, which is a standardized 0–10 point scoring system based on the presence of three bacterial morphotypes: large Gram-positive rods (Lactobacillus spp.), small Gram-negative or Gram-variable coccobacilli (Gardnerella and anaerobic spp.), and curved Gram-variable rods (Mobiluncus spp.) A score of 0–3 was considered normal, 4–6 was intermediate, and scores ≥ 7 indicated BV. The healthy women according to Nugent score had none or no more than one of the four characteristics based on Amsel criteria, and none of them had clue cells on wet-mount microscopy. Two study personnel carried out the Nugent scoring independently, and the inter-observer agreement was excellent (kappa = 0.84). All discordant slides (n = 8) were read by a third reader as a tiebreaker. This study included women with normal or BV Nugent scores.

Laboratory methods

Laboratory tests were performed by trained research assistants using standardized protocols in the Public Health Research Institute of India (PHRII) laboratory located 10 min driving distance from the clinic. The two vaginal swabs that were stored and transported in Port-A-Cul tubes (Becton Dickinson) from the clinic to the laboratory were then plated on Rogosa and sheep blood agar (HiMedia Laboratories) within 12 h of sample collection. Plates were incubated for 24 to 48 h at 37 °C in 5 % CO₂ anaerobic jars containing AnaeroPacks (Mitsubishi). Only plates containing 10–200 c.f.u. were analysed further. Up to 10 colonies from each plate were randomly selected for processing. The colonies with Gram-positive rods were inoculated in de Man-Rogosa-Sharpe (MRS) broth medium. Individual colonies were cultured in 2.5 ml MRS broth (BD Diagnostics), incubated overnight at 37 °C in 5 % CO₂. Glycerol (10 %) stocks were prepared and stored in 200 μl of the broth culture at − 80 °C. The remainder was used for DNA extraction.

Bacterial DNA extraction

DNA was extracted from the broth cultures for 16S rDNA PCR and sequence analysis. A 1 mL aliquot of the culture was centrifuged for 5 min at 5000 r.p.m. The pellet was resuspended in 500 ml of autoclaved double-distilled water in a sterile cryotube containing 0.5 mm glass beads (Research Products International). Samples were vortexed, and a freeze–thaw method was used for DNA extraction. They were frozen overnight at − 20 °C, then thawed and vortexed again. Finally, they were centrifuged for 30 s at 5000 r.p.m. before PCR amplification.

PCR amplification of 16S rDNA sequences

PCR amplification was carried out with 5.2 μl of DNA template in 30 μl of PCR mixture. Each reaction mixture contained 0.2 mM deoxynucleoside triphosphates (dNTPs), 1 U of Taq polymerase (Genei), 1 ×  Taq reaction buffer, and 1 μM of primers. The 16S rDNA PCR assays were carried out with the primers 16s8F (AGAGTTTGATCCTGGCTCAG) and 16s806R18 (GGACTACCAGGGTATCTAATCC), as described previously (Martinez-Freijo et al., 1998). The 800 bp PCR product was obtained under the following conditions: 94 °C for 5 min, 30–35 cycles of 94 °C for 30 s, 62 °C for 30 s, and 72 °C for 90 s. The PCR products were electrophoresed on 1 % agarose gels and stained with ethidium bromide and visualized under UV transillumination. Escherichia coli ATCC 25922 DNA extracted by the freeze–thaw method was used as a positive control for the 16S rDNA.

Sequencing analysis of 16S rDNA

PCR products were purified by AxiPrep kit (Axigen) or Genejet gel extraction kits (Fermentas). Sequencing was done at SciGenome or Eurofins Genomics India (Bangalore). DNA sequences ranging between 400 and 600 bp or more for the 16S rDNA gene were visually inspected, aligned and compared against sequences deposited in GenBank by blast (National Center for Biotechnology Information) or the Greengenes server with the Align tool (greengenes.lbl.gov). Genus and species were determined by the criteria of 98 % sequence identity for species and 95 % sequence identity for genus (Bäckhed et al., 2005).

Data analysis

Data were entered in an Access database (Microsoft Corporation). Conventional descriptive statistics were used to assess the distribution of Lactobacillus species among the study participants. The Fisher exact test was used to test the relationship between BV status (normal microbiota versus BV) and the presence of any lactobacilli and specific Lactobacillus species. The Pearson chi-squared test was used to examine the association of sociodemographic variables and BV status. Two-tailed P-values < 0.05 were considered statistically significant. Only samples with confirmed blast identity ≥ 95 %, and match length ≥ 500 base pairs were included in the analyses. About 2644 of the 2680 samples had colonies that met the requirements of blast identity ≥ 95 % and match length ≥ 500 base pairs. Data were analysed with SAS version 9.3 (SAS Institute).

Results

Of the 183 women screened, vaginal samples were obtained from 139, and 132 were included in the analyses, as the remaining seven did not have complete information. Eighty-three women had normal microbiota and 49 (37.1 %) were classified as BV by Nugent score. The primary reason for participants to attend the reproductive health clinic was a routine health check-up that was provided at this clinic on a regular basis at no cost. Demographic characteristics are shown in Table 1 for the two most common species of Lactobacillus present in this study population. The majority of the women were married and had a single sex partner. About 87 % of the participants reported their religion as Hindu. None of the women reported any douching practices. While 81 % of the study participants had undergone tubal ligation as the primary family planning method, about 11.9 % reported not using any contraception methods. None of the characteristics including contraceptive methods had any association with BV status. Women with L. crispatus or L. gasseri were significantly older and had more than a primary school education.

Table 1. Sociodemographic characteristics and prevalence of L. crispatus and L. gasseri in 132 women in Mysore, India (November 2011 to September 2013).

	L. crispatus		L. gasseri
Characteristic	N	%	N	%	P-value
Age (years)					0.042
20–29	58	43.9	9	27.3
30–39	67	50.8	23	69.7
40–49	7	5.3	1	3.0
Education (years)					0.016
0	26	19.7	2	6.1
1–7	55	41.7	12	36.4
8–12	49	37.1	19	57.6
>12	2	1.5	0	0.0
Religion					0.560
Hindu	115	87.1	30	90.9
Other	17	12.9	3	9.1
Years married to current partner					0.119
0–11	57	43.2	11	33.3
12–24	69	52.3	22	66.7
25–37	6	4.5	0	0
Method of preventing pregnancy
None	16	11.9	1	3.0	0.118
Condoms	4	3.0	2	6.1	0.254
IUD	1	0.7	1	3.0	0.246
Pill	1	0.7	0	0.0	1.000
Withdrawal	2	1.5	2	6.1	0.059
Tubal ligation	109	81.3	27	81.8	0.936
BV*					0.008
Positive	49	37.4	6	18.2
Negative	82	62.6	27	81.8

BV, Bacterial vaginosis.

* Nugent score was missing for one participant.

Lactobacillus colonies were recovered from 47 (35.6 %) of the 132 women. While 32 (24.2 %) of the 132 women did not have any bacterial growth culturable by either Rogosa or MRS medium, 45 women did show growth, but it was not Lactobacillus. These women were demographically no different than the women who did have Lactobacillus colonies. Another eight women (6.0 %) had growth that produced no bands by aragose gel electrophoresis. Based on the Nugent score of vaginal smears, lactobacilli were isolated from 39/83 (46.9 %) women with normal microbiota and 8/49 (16.3 %) women with BV. The predominant Lactobacillus species cultured in this sample included L. crispatus (40.4 %) and L. gasseri (44.7 %).

Table 2 shows the distribution of colonizing Lactobacillus stratified by BV status in the sample. More than one species of Lactobacillus was cultured from 22 (46.8 %) of the 47 women who had Lactobacillus identified from the vaginal samples. About eight (17.0 %) women had at least three or more different Lactobacillus species simultaneously detected in their vaginal flora. L. coleohominis, L. fermentum, L. iners, L. johnsonii, L. jensenii, L. mucosae, L. oris, L. reuteri and L. thermophilus were only isolated from women with normal microbiota whereas the other predominant Lactobacillus species (L. crispatus, L. gasseri) were isolated from both women with and without BV, and L. ruminis was isolated only from women with BV. Women with normal microbiota had at least one of 11 Lactobacillus species, while women with BV had only three Lactobacillus species present: L. crispatus, L. gasseri and L. ruminis. The Lactobacillus species that we looked for, but were not recovered in this study sample, included L. acidophilus, L. brevis, L. buchneri, L. casei, L. catenaformis, L. confusus, L. parabuchneri, L. rhamnosus, and L. salivarius. Colonization by any Lactobacillus (P = 0.0003) and in particular L. crispatus (P = 0.042), L. jensenii (P = 0.047), L. gasseri (P = 0.015), or L. reuteri (P = 0.026) was significantly associated with a normal vaginal microbiota (Table 3).

Table 2. Bacterial vaginosis (BV) status and distribution of colonizing Lactobacillus species in 132 women in Mysore, India (November 2011 to September 2013).

		BV status
Colonizing species	Women colonized (n)	Healthy (n)	BV+(n)
L. acidophilus	0	–	–
L. coleohominis	4	4	–
L. crispatus	19	16	3
L. fermentum	6	6	–
L. gasseri	21	18	3
L. iners	1	1	–
L. jensenii	7	7	–
L. johnsonii	4	4	–
L. mucosae	2	2	–
L. oris	2	2	–
L. reuteri	9	9	–
L. rhamnosus	0	–	–
L. ruminis	2	–	2
L. salivarius	0	–	–
L. thermophilus	2	2	–
L. vaginalis	0	–	–

Only species that were cultivable are included in the table.

Table 3. Bacterial vaginosis (BV) status and distribution of cultivable Lactobacillus species isolates in 132 Indian women in Mysore, India (November 2011 to September 2013).

	Healthy		BV+		Total
Species	n (%)		n (%)		N (%)		P-value*
Any Lactobacillus
Absent	44	(53.0)	41	(83.7)	85	(63.9)	0.0004
Present	39	(47.0)	8	(16.3)	47	(35.6)
L. coleohominis
Absent	79	(95.2)	49	(37.7)	128	(97.0)	0.296
Present	4	(4.8)	0	(0.0)	4	(3.0)
L. crispatus
Absent	67	(80.7)	46	(93.9)	113	(85.6)	0.036
Present	16	(19.3)	3	(6.1)	19	(14.4)
L. fermentum
Absent	77	(92.8)	49	(100)	126	(95.5)	0.086
Present	6	(7.1)	0	(0)	6	(4.5)
L. gasseri
Absent	65	(77.4)	46	(93.9)	111	(83.5)	0.015
Present	18	(22.6)	3	(6.1)	21	(16.5)
L. iners
Absent	82	(98.8)	49	(100)	131	(99.2)	1.000
Present	1	(1.2)	0	(0)	1	(0.8)
L. jensenii
Absent	76	(91.6)	49	(100.0)	125	(94.7)	0.047
Present	7	(8.3)	0	(0.0)	7	(5.3)
L. johnsonii
Absent	79	(95.2)	49	(100.0)	128	(97.0)	0.296
Present	4	(4.8)	0	(0.0)	4	(3.0)
L. mucosae
Absent	81	(97.6)	49	(100.0)	130	(98.5)	0.530
Present	2	(2.4)	0	(0.0)	2	(1.5)
L. oris
Absent	81	(97.6)	49	(100.0)	130	(98.5)	0.533
Present	2	(2.4)	0	(0)	2	(1.5)
L. reuteri
Absent	74	(89.2)	49	(100.0)	124	(93.2)	0.026
Present	9	(10.8)	0	(0.0)	9	(6.8)
L. ruminis
Absent	81	(100)	47	(95.9)	130	(98.5)	0.134
Present	0	(0)	2	(4.1)	2	(1.5)
L. thermophilus
Absent	81	(97.6)	49	(100)	130	(98.5)	0.530
Present	2	(2.4)	0	(0)	2	(1.5)

* P-values obtained by chi-square tests when expected cell frequency was < 5, otherwise by Fisher's exact tests.

Discussion

Our study shows the vaginal flora of Indian women was dominated by two cultivable Lactobacillus species (L. crispatus and L. gasseri). Research from different parts of the world shows that L. crispatus, L. gasseri and L. jensenii are the most common cultivable Lactobacillus species, and this was true in our study sample as well (Chaban et al., 2014; Mendes-Soares et al., 2014; van de Wijgert et al., 2014). Unlike other research findings, we found frequent co-occurrence of species as opposed to dominance of one species or another. Vaginal colonization predominantly with L. gasseri, L. crispatus, L. reuteri and L. jensenii in our sample shows a distribution similar to that found in women in other regions of the world (Antonio et al., 1999; Damelin et al., 2011; Dong-hui et al., 2009; Song et al., 2000; Vásquez et al., 2002). While L. crispatus and L. gasseri were detected in women with and without BV, they were found in a greater number of women with normal microbiota than in women with BV. The strong association of L. crispatus with normal vaginal microbiota is similar to that found in several other studies (Antonio et al., 1999; Vásquez et al., 2002). In a longitudinal study of vaginal microbiota, Verstraelen et al. (2009) showed that L. crispatus was associated with the stability of the vaginal flora. These findings however, differ greatly from an Indian study by Garg et al. (2009) who found that L. reuteri, L. fermentum and L. salivarius were the main species in the vaginal flora of women in Delhi, northern India. These differences suggest that vaginal Lactobacillus may vary in subpopulations of India.

There are several study limitations that should be considered when interpreting the findings. First, as our methodology used culture first to detect Lactobacillus species, only species that can be cultured were further identified. Our study only provides a limited view of the species of Lactobacillus that may be present in the vaginal environment. Second, the differences found in the diversity and distribution of the lactobacillus species may have resulted from participants being in different phases of their menstrual cycle. While the relationship of hormones to the endogenous vaginal flora is important, it cannot be ascertained from our study. Our negative result for isolation of L. iners does not mean that this species may not be present in the population, as we may have been limited in using a selective media that was not supportive for the growth of this species. Finally, findings from our study cannot be generalized to other populations due to the small sample size, non-probability sample coming from a reproductive health clinic and limited statistical power.

Our study has several strengths as well. Since little is known about the vaginal flora of Indian women, our study contributes to the literature and knowledge base by providing the characteristics of cultivable lactobacillus present among south Indian women. The findings of this study also confirm the prominence and wide carriage of several Lactobacillus species, suggesting that they may be of particular importance in the development of future microbicides and probiotics. The vaginal species of Lactobacillus found in south Indian women from Mysore are similar to those identified in women from other populations. Our findings need to be confirmed in a larger study using molecular methods and longitudinal study designs to better characterize this important vaginal defence.

Abbreviations:

BV

bacterial vaginosis

RTI

reproductive tract infection