Background

Lactobacilli are the predominant microbes in the vaginal microenvironment of healthy females [1]. A key feature that favours colonization of Lactobacillus is low vaginal pH due to metabolism of glycogen to lactic acid; glycogen deposits in vaginal epithelial cells in response to rise in circulating estrogen [2, 3]. The predominance of Lactobacillus promotes vaginal health by inhibiting colonization of pathogenic microbes at acidic pH and also by producing antimicrobial bacteriocins [4]. However, the vaginal pH and Lactobacillus levels vary over the life course of women [5]. During bacterial vaginosis (BV), Lactobacillus predominance is lost and replaced by pathogenic microbes, which adversely affects women health making susceptible to HIV, HSV-2, pathogenic microorganisms such as Neisseria gonorrhoeae, Gardnerella vaginalis, Escherchia coli, Staphylococcus aureus, Peptostreptococcus anaerobius, P. bivia, and Chlamydia trachomatis, pelvic inflammatory disease and pre-term delivery [6,7,8,9,10,11,12,13]. Therefore, Lactobacilli are considered gatekeepers of vaginal ecosystem and understanding their qualitative and quantitative diversity is an emphasized area of research [1].

The human microbiome project set out vaginal microenvironment as one of the priority site related to human health and disease and reported around 60 bacterial species that colonizes lower urinogenital tract of female dominated by Lactobacillus crispatus, L. iners, L. gasseri, L. jensenii [1, 14, 15]. Apparently, the diversity and dominance of Lactobacillus varies among different populations, ethnic groups, regions etc. Ravel et al [16] studied Lactobacillus composition in four ethnic groups and found that Asian and White women have higher vaginal Lactobacillus composition, dominated by L. iners and L. crispatus respectively, than Hispanic and Black women, who have higher proportion of strictly anaerobic bacteria and dominant Lactobacillus species is L. iners. In some other studies Lactobacillus composition has been further reported to vary among different populations and as such L. iners, L. crispatus, and L. gasseri were reported predominant among Chinese women [17], L. acidophilus among Mexicans and Swedish [18, 19], L. crispatus, L. iners, L. gasseri, L. jenseni, and L. vaginalis among South Africans [20]. A comparative study on Korean and Uganda women reported that L. fermentum was common in Korean, and L. gasseri, L. reuteri, and L. vaginalis in Ugandan women [21]. In India the biogeographical condition and population varies across different regions and as such the vaginal Lactobacillus species L. iners, L. crispatus, L. reuteri, L. gasseri, and L. jensenii have been found predominant in Mumbai (southwest region of India) [22], L. reuteri, L. fermentum, and L. salivarius in Delhi (central region of India) [23], L. crispatus, L. gasseri, and L. jenseni in Mysore (southern region of India) [24] and many of the regions still remain unexplored. A consensus of all studies explains four Lactobacillus species viz., L. crispatus, L. jenseni, L. gasseri, and L. iners [2] are predominant in vagina and among them L. crispatus and L. jenseni are considered most beneficial [25, 26] and their colonization corresponds to higher vaginal glycogen content [2].

Lactobacillus also restores vaginal health during pregnancy and its imbalance may cause BV and subsequent post-abortal infection, early and late miscarriage, histological chorioamnionitis, postpartum endometritis, preterm premature rupture of membranes and preterm birth [27,28,29]. In this context, a number of studies [30,31,32,33,34,35] had characterized vaginal microbiota during different stages of pregnancy in different biogeographical condition. However, such important studies are lacking from India.

Therefore, it is quite clear that the composition of Lactobacillus in vaginal ecosystem varies across population and dependent on many factors such as hormonal de-regulation, poor genital hygiene, improper vaginal douching, random sexual behaviour and in labouring women. Also different species of Lactobacillus offer different level of protection under different vaginal condition. Hence, structuring of vaginal microbiota in general and Lactobacillus in particular in different populations and their link with female health in normal and pregnant condition is highly appealing.

The present study evaluated the prevalence and the quantities of vaginal microbiota in the vagina of healthy fertile and pregnant women of Northeast India using culture-dependent techniques and subsequent 16S rRNA gene sequencing and analysis. Northeast India is genetically distinct from other parts of India and comprises of many tribes mainly of Austro-Asiatic and Tibeto-Burman background [36]. Further, they have unique dietary habits and lifestyle and these features in combination provide unique opportunity and challenges to decipher microbial composition in different habitats of human body [37] and we focussed on vaginal microbiota.

Results

Screening and characterization of samples, culture of bacterial isolates and sequencing

Vaginal swab samples were initially collected from 83 number of reproductive age females having age range 18–35. Among them 14 cases either had history of urinary tract infection, under antibiotic treatment, or has ongoing vaginal infection and excluded from the study (Additional file 1: Table S1). The remaining 69 samples according to age may be categorized as; 23 cases in age group 18–23, 25 cases in age group 24–29, and 21 cases in age group 30–35. The 69 samples further may be categorized as 29 non-pregnant and 40 pregnant cases and the pregnant cases may be categorized as; 12 first (I) trimester, 16 s (II) trimester, and 12 third ((III) trimester. The swab samples were streak plated in Nutrient Agar (NA), de Man, Rogosa and Sharpe (MRS), and Blood Agar (BA) media for microbial growth under ambient condition. While bacterial colonies appeared in NA and MRS media and a total of 154 bacterial isolates were obtained, no growth was observed in BA media. The number of isolates obtained from each sample, their gram staining and colony morphology, sample and colony code and corresponding sequencing code and accession numbers were detailed in Additional file 2: Table S2.

Vaginal microbial diversity of females of Northeast India

The 16S rRNA gene based identification of the samples were based on a dual approach of homology search with database (Additional file 3: Table S3) and characteristic ML clustering (Additional file 6: Figure S1) of derived and their corresponding closely match database sequence. ML clustering was done with a total 270 sequences, that includes 154 derived and 116 database sequences (Additional file 4: Table S4), modelled by Kimura’s 2 parameter with gamma distributed rate across sites that describes best substitution pattern for the dataset according to Bayesian information criterion [38] (Additional file 5: Table S5). The homology search and ML clustering results are presented in a manageable form in Table 1.

Table 1 Identification of the bacterial isolates based on 16 s rRNA gene sequence
Full size table

Consensus match of derived sequences with a particular microbe in both the databases (NCBI and EZ-Taxon) in the range of similarity 97–100% and cohesive clustering of conspecific derived and database sequences confirmed straight forward identification of 80 sequences into 19 species. This includes Bacillus subtilis, Bifidobacterium breve, Escherichia coli, Janthinobacterium lividum, Kocuria kristinae, Lactobacillus brevis, L. fermentum, L. gasseri, L. ingluviei, L. jensenii, L. mucosae, L. parafarraginis, L. reuteri, Propionibacterium avidum, Streptococcus agalactiae, S. anginosus, S. gallolyticus, S. infantarius, and Weissella confusa. Among them Lactobacillus mucosae showed deep intraspecies split. One derived sequence (KP775927) that showed high similarity (98–99%) with Lactobacillus gallinarum and 3 sequences (KT361211, KT361209, and KP775926) that showed high similarity (97–99%) with L. helveticus based on homology search, clustered together with database sequences of L. gallinarum and L. helveticus. Similarly, 48 sequences identified as four species of Enterococcus (E. avium, E. faecalis, E. faecium, and E. hirae) based on homology search, and database sequences of the four Enterococcus species clustered together. Also, 22 sequences identified as 6 species of Staphylococcus (S. capitis, S. caprae, S. epidermidis, S. haemolyticus, S. hominis, and S. kloosii) and database sequences of the six Staphylococcus species clustered together.

In summary Enterococcus faecalis (25.3%) and Lactbacillus mucosae (16.2%) were found dominant species among the study population. 37.7% of the bacterial isolates were Lactobacilli, that includes 10 species, and their relative dominance is; L. mucosae (16.2%) L. fermentum (6.4%), L. gasseri (3.9%), L. brevis (3.2%), L helveticus and L. jensenii each 2%, L. ingluviei and L. reuteri each 1.3%, and L. gallinarum and L. parafarraginis each 0.6%. Clearly Lactobacilli percentage significantly varied among age group (p < 0.001) and we found in age group 18–23, 87% of females harbour Lactobacillus, in age group 24–29, 76% females harbour Lactobacillus, and in age group 30–35, 71.4% females harbour Lactobacillus.

Lactobacillus composition in non-pregnant and pregnant women

One major objective of the present study is to categorize Lactobacillus diversity in non-pregnant and pregnant women and in three trimester of pregnant women. It is clearly evident in this study that the diversity of Lactobacillus increases in labouring women. As such we found only L. reuteri colonized non-pregnant women; 4 species L. brevis, L. fermentum, L. gasseri, and L. mucosae colonized both non-pregnant and pregnant women; and 5 species L. gallinarum, L. helveticus, L. ingluviei, L. jensenii, and L. parafarraginis colonized only pregnant women (Fig. 1). Trimester-wise the diversity of Lactobacillus is as follows: L. fermentum and L. mucosae overlapped in all three trimester, L. brevis and L. ingluviei overlapped in I and III trimester, L. gasseri overlapped in I and II trimester, L. helveticus overlapped in II and III trimester, while 3 species L. gallinarum, L. jensenii, and L. parafarraginis were found in only II trimester (Fig. 2). Again Lactobacilli percentage significantly varied among non-pregnant and pregnant women (p < 0.0002) and we found 58.6% of studied non-pregnant females harbour Lactobacillus while 90% of studied pregnant females harbour Lactobacillus, which may be further categorized as 83.3% trimester I, 93.8% trimester II, and 91.7% trimester III.

Fig. 1
figure 1

Bacterial colonization in non-pregnant and pregnant women. The numbers under each circle corresponds to number of species found under each condition and in parenthesis relative percentage shown. Clearly colonization is more in pregnant women than non-pregnant women

Full size image
Fig. 2
figure 2

Bacterial colonization in three trimester of pregnant women. The numbers under each circle corresponds to number of species found under each trimester of pregnant women and in parenthesis relative percentage shown. Clearly bacterial colonization is higher in trimester II

Full size image

Discussion

In women the vaginal ecosystem is dominated by Lactobacilli and this system has evolved to protect females health by inhibiting growth of pathogenic microbes. However, the characteristic microbial diversity varies across population [4, 16, 39,40,41,42]. Women in different settings face different geographical condition, dietary habits, and lifestyle. In particular, peoples of Northeast India are from Austro-Asiatic and Tibeto-Burman background and are genetically different from other parts of India [36, 43, 44]. Therefore, it is tempting to screen and compare how the vaginal microbiota is structured in females in Northeast India and how it differs from other population.

In this study altogether we isolated 154 bacterial isolates in NA and MRS media from 69 females; which includes 40 pregnant and 29 non-pregnant cases. BA was used as a selective media for Lactobacilli, especially Lactobacillus iners, however, no colonies appeared in BA. Previous studies [45,46,47] advocates that L. iners grows in transitional phase of BV and healthy state, which is not the case in this study, and this tentatively is the reason for no growth in BV. However, we believe a detail sampling including healthy and BV cases in future studies may provide a comprehensive insight on Lactobacillus diversity among the females of Northeast India.

A dual approach of homology search and ML clustering confirmed the 154 bacterial isolates belonging to 31 species and 37.7% of the bacterial isolates are Lactobacillus. Interestingly, we found L. mucosae is the dominant species and L. fermentum, L. gasseri, L. brevis, L. helveticus, L. jensenii, L. ingluviei, L. reuteri, L. gallinarum, and L. parafarraginis together forms the Lactobacilli complex that colonizes vaginal microenvironment among the females of Northeast India. This is the first report of its type from Northeast India and has implications for global comparison of Lactobacilli composition. Clearly, vaginal Lactobacillus composition among Northeast India population is different from other parts of India [22,23,24] and various regions/countries of the World [16,17,18,19,20,21]. Lactobacillus mucosae has been reported as less common vaginal microflora in previous studies [41, 48], however dominance of L. mucosae than L. crispatus and L. jensenii (as observed in most of the previous studies [2, 16]) observed in this study in Northeast Indian females is an intriguing finding and tentatively reflects different vaginal set-up. In particular, different Lactobacillus species possess unique repertoire of protein families which helps in specific community adaptations [49, 50]. Lactobacillus mucosae has been reported to express Lam29 protein in intestinal epithelial cells, that significantly improves its adhesion for human blood group A and B antigens and out competes pathogens to protect intestinal mucosa [51]. In human vagina, the epithelial layer similarly acts as a barrier for mucosal layer, which the pathogen wants to penetrate and Lactobacillus wants to protect by adhesion and outcompeting pathogens [52,53,54,55]. However, this adhesion is dependent on genetic repertoire of both the host and the bacteria [49, 50]. Further, the adhesion and colonization of Lactobacillus is also dependent on hormonal level (especially estrogen), glycogen content of epithelial cell, and availability of nutrients [3]. We hypothesize Lactobacillus mucosae outcompetes other organisms for receptors and nutrients in the vaginal epithelial cells of females of Northeast India and this claim needs more insight from future studies. Nevertheless, this finding is also tempting to explore probiotic potential of L. mucosae than L. crispatus and L. jensenii among women of Northeast India and correlation with estrogen level and deposition of glycogen in vaginal epithelium.

Besides Lactobacillus that constituted 37.7% of bacterial isolates, we also found 10 other lactic acid bacteria species (39.61%) viz., Bifidobacterium breve, Enterococcus avium, E. faecalis, E. faecium, E. hirae, Streptococcus agalactiae, S. anginosus, S. gallolyticus, S. infantarius, and Weissella confusa with predominance of Enterococcus faecalis (25.3%). Among them S. infantarius, S. gallolyticus, S. agalactiae, S. anginosus, E. avium, E. hirae are opportunistic pathogens and causes severe invasive infections [56,57,58,59,60].

The study also reports 4 species of Enterococcus viz., E. avium, E. faecalis, E. faecium, and E. hirae, 6 species of Staphylococcus (14.28%) viz., S. capitis, S. caprae, S. epidermidis, S. haemolyticus, S. hominis, and S. kloosii and 2 species of Lactobacillus viz., L. gallinarum, and L. helveticus with cohesive congeneric clustering. This clearly reveals congeneric exchange of 16S rRNA gene segment, may be due to horizontal gene transfer or homologous recombination, which is common among microbes [61]. These key findings together with the observation of deep intraspecies split in Lactobacillus mucosae warrants genome-wide study of Enterococcus, Staphylococcus and Lactobacillus in future studies to understand intraspecies genomic variability, possible emergence of new strains/variants and its impact on female health of Northeast India. Nevertheless, Propionibacterium avidum (3.90%), Bacillus subtilis, Escherchia coli, Janthinobacterium lividum, and Kocuria kristinae (each 0.64%) represented renaming of the microbiota found in this study. Lactobacillus ingluviei and L. parafarraginis found in this study have never been previously reported from human vagina [62, 63] and so this is the first case of their report from human vagina.

The vaginal microbiota have a major impact during different stages of pregnancy [30] and therefore we assessed how the microbes, and Lactobacillus in particular, are structured among non-pregnant and different stages of pregnant women. We found Lactobacillus percentage is highest in the age group 18–23 (87%) and successively decreases (71–76%). This finding is consistent with previous studies [64] and therefore as female attains age and Lactobacillus composition decreases, the pathogenic microbes have higher chance of colonization and thus extra precaution must be followed. In particular, aged women may be supplemented with Lactobacillus probiotic, which is predominant in their population, since it is observed herein and some previous studies [4, 16, 39,40,41,42] that predominance of Lactobacillus species varies across population. The Lactobacillus composition in pregnant women showed a sharp increase (90%) compared to non-pregnant women (58.6%) similar to earlier study [65] and this perhaps is the best way to protect vaginal health in labour periods. We further evidenced higher colonization of Lactobacilli in trimester II followed by trimester III and I. These features must be explored in more details for specific functional characterization of different species of Lactobacillus in different stages of pregnancy.

Conclusion

The study clearly highlighted importance of assessment of vaginal microbiota in different populations, groups, ethnics, religions etc. The vaginal microbial composition studied herein revealed some interesting new information’s such as: i) Lactobacillus mucosae rather than L. crispatus or L. iners is dominant among females of Northeast India and this may be related to Lacotbacillus adaptation in particular population mainly due to genetic variability, ii) deep intraspecies split among L. mucosae revealed population level divergence and may be due to evolution of various strains active under specific conditions, iii) Lactobacilli percentage among Northeast India females is less and tentatively compensated by other LABs, iv) parallel to L. mucosae, Enterococcus faecalis is also dominant among females of Northeast India and its specific role must be explored. The Lactobacilli composition in non-pregnant and pregnant cases is consistent to previous studies however specific trimester-wise colonization of particular Lactobacillus species needs to be explored in more details. This study is the first report of its type from Northeast India and although we used culture-dependent technique, we identified around 31 vaginal microbial species and importantly revealed unique Lactobacilli diversity and dominance. We believe our study provided an interesting clue on vaginal Lactobacillus composition. This is tempting for future studies to include broad spectrum culture independent technique together with culture dependent technique for better understanding deficiency and dominance of particular vaginal Lactobacillus species and its relation to female health of Northeast India; Northeast India females have high rate of cervical cancer and Lactobacilli have major role in acquisition and persistence of Human Papilloma Virus [66] and regulation of cervical cancer [67].

Methods

Sample collection and culture of microbes

Vaginal swab samples were collected from 83 numbers of healthy pregnant and non-pregnant women during the period from 23 July 2012 to 29 November 2015 (Additional file 1: Table S1). A conscious effort was made to collect samples from volunteers who did not have any apparent vaginal infection. The present experiment excluded the volunteers who had antibiotic treatment in last one and half months, and who has history of urinary tract infection in last 6 months, any homogenous white discharge from vagina, and whose pH higher than 4.5, and who have the problem of itching, bleeding from vagina prior to sampling. All women were at reproductive age i.e. 18–35. Necessary clearance from the Institutional Ethical Committee, Karimganj College (KC/IEC/2012/M-1/10, dated: 23 Jun 2012) before the commencement of the study was obtained. The samples were collected aseptically by a practicing gynaecologist of the Red Cross Society of India, Karimganj, Assam (India). Initially, the experimental procedure was explained to the volunteers and when they agreed to participate in the study, they were asked to sign a written consent which was prepared as per the recommendation of the Institutional Ethical Committee. The collection of the samples was made after obtaining formal consent. A sterile swab (Hi Media Laboratory Pvt. Ltd., India) was inserted into the vaginal lumen and the samples were collected from the posterior fornix of vagina. The swabs were collected separately in sterile tubes. The swabs were used within 3 h of collection. The microbes from the swabs were transferred to sterile NA, MRS Agar, and BA media by streak-plate method. It is to be mentioned that while MRS and BA are selective media for Lactobacilli, NA was used as a general starter media to get a general view on microbes present apart from Lactobacillus in vaginal microenvironment of females of NE India. The plates were incubated for 48 h at 37 °C both under aerobic and anaerobic condition to allow the growth of microbial colonies. During collection of samples a statistical approach called convenience method/sampling was applied [68] (Additional file 1: Table S1).

DNA isolation and PCR

The microbial genomic DNA was isolated following the protocol as mentioned in our previous paper [69]. The amplification of the 16S rRNA gene was done using published universal primers 27F Forward primer (AGAGTTTGATCMTGGCTCAG) and 1492R Reverse primer (5ˈ-GGTTACCTTGTTACGACTT-3ˈ) [70]. The 10 μl PCR reaction mixture contained template DNA 1 μl, 20 picomole of each primer, Master Mix 2X (HiMedia, India) 5 μl, and nuclease-free water 2 μl. The PCR thermal condition includes initial denaturation at 94 °C for 2 min, 35 cycles of 94 °C for 30 s, 49.3 °C for 30 s, 72 °C for 1 min and final extension at 72 °C for 3 min in S1100 thermal cycler (BioRad, USA). The PCR products were visualized in 1% agarose gel and images were captured by a Gel DOC machine (BioRad, USA).

16S rRNA gene sequencing

The 1.5 Kb amplicons were sequenced using the facilities provided by the MERK, Bangalore, India and Xcelris, Hyderabad, India. After obtaining the sequencing data, these were analysed and edited using Sequence scanner software v1.0 [57] (Applied Biosystem Inc., CA, USA). Alignments are prepared by ClustalX [71] and DNA contigs. Thereafter, the sequence was matched using BLASTn (NCBI database) [72] to identify its closest relatives. The sequences were submitted in the NCBI database using the submission tool BankIt and the accession numbers were obtained.

Identification of the microbes

For Identification of the microbes we followed a dual approach of sequence based homology search and clustering (Additional file 7: Figure S2). For homology search we compared a particular sequence for similarity with two separate databases NCBI using BLASTn approach [72] and EZ-Taxon (www.eztaxon.org/). For clustering we used Maximum Likelihood method [73] structured by Kimura’s 2 parameter [57] model and the computation platform used was MEGAX [74]. The high similarity (97–100%) of a sequence with database sequences of a particular microbe coupled with cohesive clustering of developed and database sequence of the particular microbe and distinct cluster with respect of other developed and database congeners sequences confirmed identification of the microbe with high precision.

The Lactobacilli percentage across different age group and pregnant and non-pregnant women were compared using Student’s t-test.