- About MBL
- For Contributors
Molecular and Cellular Microbiology (MCM) | Microbiome
Microbiol. Biotechnol. Lett. 2023; 51(1): 109-123
Sudeepti Kulshreshtha1†, Priyanka Narad1†, Brojen Singh2 , Deepak Modi3, and Abhishek Sengupta1*
†These authors contributed equally to this work.
1Amity Institute of Biotechnology, Amity University, Uttar Pradesh 201301, India
2School of Computational and Integrative Sciences, Jawaharlal Nehru University, New Delhi 110067, India
3Department of Molecular and Cellular Biology, National Institute for Research in Reproductive Health and Child Health, Mumbai 400012, India
Correspondence to :
Abhishek Sengupta, firstname.lastname@example.org
†These authors contributed equally to this work.
Preterm birth (PTB) is defined as giving birth prior to the 37th week of pregnancy and is a major cause of infant mortality. Studies have indicated that the vaginal microbiota's composition and its dysbiosis, particularly during pregnancy, may play a major role in PTB. While previous research work concentrated on well-studied microorganisms such as Lactobacillus, Prevotella, Gardnerella, various other microbes, and their significance in the vaginal microbiota's stability remain unknown. Moreover, current studies have focused primarily on the relative abundances of the microbes found, without considering their interactions with other members of the vaginal microbiota. In this work, we developed a novel computational approach and performed taxonomic classification of vaginal microbiota samples stratified longitudinally (Term/PTB) to observe compositional disparities and find underexamined microbes that may be contributing to PTB. Furthermore, we carried out a correlational analysis to build a microbial co-interaction network and investigated the functional implications of the genes present in both Term and PTB samples. The co-occurrence network revealed that Lactobacillus acts in solidarity to maintain the stability of the vaginal microbiota and did not have strong co-interactions with any of the other microbes. Similarly, microbes with strong interactions with Atopobium, a well-known marker microbe of PTB, were also observed. Additionally, several genes such as PTXA, FANCM, GPX, and DUSP were found to be playing an important role in the occurrence of PTB. This study provides a novel conceptual framework revealing distinct vaginal microbiota signatures that could be potential therapeutic targets for the prevention of PTB.
Keywords: Vaginal microbiota, preterm birth, pregnancy-related complications, bioinformatics, reproductive medicine
Every year, around 15 million Preterm births (PTB) occur at less than 37 weeks of pregnancy . Although maternal genetics plays a role in gestation duration, environmental influences like the microbiota, are the most significant contributors to PTB . An increasing amount of research is now focusing on the role of the vaginal microbiota in determining PTB risk. Reports indicate that the vaginal microbial populations are critical in the etiology and pathogenesis of PTB transmission  with microbes such as
A majority of the previous studies have focussed on taxonomic classification to identify microbial signatures associated with healthy and diseased states. However, comprehending the complexities and involvement of the entire microbial community that inhabits the host is far more perplexing than studying individual microbes that cause a specific disease or infection. To begin with, one of the major challenges in microbiome research is the diversity of methodologies used to detect individual microorganisms and characterize the microbiota. Such discrepancies have the potential to result in significant disparities in analysis and data interpretation. Additionally, the majority of current research focuses on certain microorganisms (for example,
In this study, we develop a novel computational approach for functional profiling and correlational analysis to establish a co-occurrence network among different microbes present in the vaginal microbiota. The representative results in this study were produced using data from a 2020 Alabama study on the vaginal microbiota of pregnant women aged 15−34. 39 vaginal microbiota samples were obtained, divided into two broad groups based on ethnicity (White = 20N, Black = 19N). The samples were further stratified based on their birth timing (White = 10 Term, 10 PTB, and Black = 9 Term, 10 PTB). We performed taxonomic classification to investigate the differences between the two vaginal microbiota environments by comparing the taxonomic profiles of Term Birth Delivery vs PTB Delivery samples. The resultant microbial relative abundances were utilized to establish a microbial co-interaction network by performing a correlational analysis among the various microorganisms that comprise the vaginal microbiota. Subsequently, we performed functional profiling to estimate the functional composition of the samples based on 16S rRNA gene OTUs to gain a deeper insight into the functions and pathways that differentiate the Term Birth Delivery state from the PTB Delivery condition. This approach can be applied to numerous microbiome samples to articulate the mechanisms and complex interactions in the vaginal microbiome unique to a dysbiotic condition, which culminates in an elevated risk of PTB in pregnant women.
The computational approach followed for this study is summarised in Fig. 1.
Vaginal microbiota samples of women stratified by birth timing (Term/Preterm) and race (black/white) were downloaded from NCBI Bioproject Project ID: PRJNA600021. The vaginal swabs considered for this study were originally collected at 21 to 25 weeks (stored at -80℃). The V4 region of the 16S rDNA gene was amplified from individual samples to produce an “amplicon library” . In our work, a total of 39 raw samples were downloaded in FASTQ file format in Ubuntu 20.04.4 , stratified by ethnicity (black/white) and birth timing (Preterm/term), as shown in Fig. 2. The quality of the samples was checked using FASTQC software.
The pre-processed samples were subjected to taxonomic classification using the Parallel-Meta 3 tool (version 3.3.2) , which generates comparable findings to QIIME (version 1.9.0)  and PICRUSt (version 1.0) , but at a considerably quicker speed and with much less memory use, demonstrating its capacity to decipher taxonomic and functional dynamics patterns across huge datasets and reveal ecological linkages between the microbiota and environment . Parallel-Meta 3 uses Bowtie2 to align all gene sequences to its reference database, to parse out an Operational Taxonomic Units (OTUs) table along with their relative abundances. To produce the OTU/taxa feature table files, the vaginal microbiota samples (FASTQ files) were taxonomically profiled using the “PM-select-taxa” command used for multi-sample feature selection (with a specified taxonomical level). Greengenes (version 13-8)  was used as the reference database (16s rRNA, 97% level identity), with the following parameters: ASV denoising, Chimera removal, level of taxonomic classification desired (level 5 for Genus), and sequence alignment threshold of 0.99, parsing out relative abundances and count files of the microbes present in each sample as the output. The taxonomic classification results were visualized using the ggplot2 library  in R (version 4.1.2). DiVenn 2.0  was used to visualize the microbes unique to Term Birth Delivery and PTB Delivery conditions.
For functional profiling, Parallel-META 3 re-implements the PICRUSt algorithm using the KEGG database to estimate all of the functional genes found in a microbiota using 16S rRNA gene OTUs. The functional genes are annotated by KEGG Ontology (KO) and the KEGG pathway database (Release 103.0) . So, the “PMselect- func” command was used to carry out 16S rRNAbased functional profiling on the vaginal samples. The PM-select-func command further extracts the sequence counts and normalized relative abundances (from 0– 100%) for all OTUs and estimates the same information for identified genes and pathways based on the specified KEGG pathway level parameter.
The results are formatted into tables (with KEGG Gene IDs, their names and functions, their abundance, and the pathways the genes are involved in) that are appropriate for further analysis and manual user investigation.
Based on the results of the PM-select-taxa command used earlier, Parallel-Meta 3 can evaluate correlations between taxa using Spearman's correlation metrics (
The correlation abundance table was reformatted as a list with all the co-interactions and their scores enumerated individually. The list was then imported by the “Network from File” option in Cytoscape 3.9.0  to visualize the microbial co-interactions. The microbes present in the samples are visualized as nodes, and the interactions that occur between them are visualized as the edges. ClusterViz (version 1.0.3)  was used to elucidate the clusters of the PTB condition for both ethnicities using the MCODE algorithm. The cluster with the highest score indicated the highest degree of interconnectivity and was visualized using the yFiles Radial Layout option of the yFiles plug-in (version 1.1.2)  in Cytoscape.
The results of the PM-select-func command used earlier have information related to the KEGG Gene IDs, their names and functions, their abundance, and the pathways the genes are involved in. These results were manually examined for both Term Birth Delivery and PTB Delivery samples to probe further into the genes and pathways that are unique to each condition and to understand the phenotypic implications of the unique genes found in PTB Delivery samples. Subsequently, the gene counts of common genes were compared in both circumstances to determine if there was any upregulation or downregulation of essential genes that resulted in substantial changes in the phenotype of the vaginal microbiota, contributing to PTB Delivery. DiVenn 2.0  was used to visualize the genes unique to Term Birth Delivery and PTB Delivery conditions, the common genes, as well as the upregulated and downregulated genes based on their gene counts.
Taxonomic classification using parallel-meta 3 revealed the microbial composition of the vaginal microbiota in terms of the microbial genus identified and their relative abundances. A comprehensive list of microbes found in the vaginal microbiota samples of women stratified by birth timing (Term/Preterm) and race (black/white), along with their relative abundances can be found in Table 1 of the Supplementary File.
A total of 67 genera (Table S1) were found in Term Birth Delivery samples (Nblack = 33; Nwhite = 34).
Similarly, a total of 95 genera (Table S1) were found in PTB Delivery samples (Nblack = 47; Nwhite = 48). 36 microbes were found to be unique to the Preterm Birth Delivery condition (Nblack = 19; Nwhite = 17), which mostly belonged to the
Amongst the microbes common in both delivery conditions, the relative abundances of
Upon comparing the relative abundances of Black women with White women, we discover the key differences between the two ethnicities (Fig. 3C). To begin with,
Functional profiling using the Parallel-META 3 tool yielded the list of genes found in the samples for both Term Birth Delivery and PTB Delivery vaginal microbiota samples in both ethnicities. Functional profiles were generated for each sample and compiled to remove any redundant or duplicate genes. A comprehensive list of various pathways found in vaginal microbiota samples stratified by birth timing (Term/Preterm) and race (black/white), along with their abundances, detected for both conditions and both ethnicities can be found in Table 2 of the Supplementary File.
More than 6000 KO annotated genes were found in Term Birth Delivery samples (Nblack = 6324; Nwhite = 6518), along with gene descriptions, relative abundance in the samples, and pathways in which the genes are present. The genes unique to Term Birth Delivery were identified (Nblack = 1120; Nwhite = 60). This indicates that in the PTB Delivery condition, several genes that might play essential functions are absent.
Meanwhile, more than 5000 KO annotated genes were found in Preterm Birth Delivery samples (Nblack = 5279; Nwhite = 6864), along with gene descriptions, relative abundance in the samples, and pathways in which the genes are present. The genes distinctive to Preterm Birth Delivery were identified (Nblack = 75; Nwhite = 406). Similarly, common genes between Term Birth Delivery and PTB Delivery conditions were also identified (Nblack = 5203; Nwhite = 6458). When comparing PTB Delivery samples to Term Birth Delivery samples, there is a considerable difference in the number of genes discovered, suggesting that many pathways and functions are suppressed or overexpressed in the PTB Delivery condition.
Significant differences were observed between the ethnicities during the co-occurrence abundance network building for PTB Delivery samples. The correlational analysis of microbes present in the vaginal microbiota in PTB Delivery conditions reveals the strength of the cooccurrence amongst microbes. We visualized the cooccurrence network in Cytoscape 3.9.0, wherein the nodes represent the microbes, and the edges represent the co-interactions between them. The size of the node is directly proportional to the relative abundance of the microbe as found in the samples. The negative cooccurrences have been depicted with red dashed edges, while positive interactions have been depicted with blue edges. The colour gradient depicts the strength of the interaction.
The microbial co-occurrence network for Black women in PTB Delivery condition can be seen in Fig. 5. The network has 50 nodes, and 560 edges (network diameter = 4, the average number of neighbours each node has = 12).
The microbial co-occurrence network for White women in PTB Delivery conditions can be seen in Fig. 6. The microbial co-occurrence network in White women is seen to be much denser with a higher number of edges, than the network for Black women. The network has 50 nodes, and 1100 edges (network diameter = 6, average number of neighbours each node has = 21).
The unique, as well as the common genes that are present in both the Term Birth Delivery and PTB Delivery conditions, were identified to comprehend the functional implications of the genes present, based on the gene count found in the conditions. Amongst the unique genes found in Term Birth Delivery samples, the top 40 genes (considering their gene counts) were observed. The observed 40 genes were associated with signaling and cellular pathways (Nblack = 15; Nwhite = 8), polyketide biosynthesis (Nblack = 0; Nwhite = 9), carbohydrate metabolism (Nblack = 5; Nwhite = 4), quorum sensing (Nblack = 4; Nwhite = 2), amino acid metabolism (Nblack = 4; Nwhite = 3), disease and infection (Nblack = 5; Nwhite = 3), and more. Similarly, amongst the unique genes found in PTB Delivery samples, the top 40 genes (considering their gene counts) were observed. These genes belonged to pathways associated with lipid metabolism (Nblack = 9; Nwhite = 4), signaling and cellular pathways (Nblack = 6; Nwhite = 11), and energy metabolism (Nblack = 1; Nwhite = 6).
Several disparities were observed even between the ethnicities. Firstly, in Black PTB Delivery samples, we found several bacteria-specific genes like
Genes and their relative abundances eventually contribute toward pathways and bring about physiological impacts on the hosts. We evaluated and compared the pathway abundances between Term Birth Delivery and PTB Delivery conditions to understand the pathways that might be getting inhibited or overstimulated due to differences in gene expression counts/levels (Table S2). As expected, significant differences were exhibited between the ethnicities.
It was observed that in Black women, all the major pathways like signaling and cellular processes, carbohydrate metabolism, DNA repair and recombination, biosynthesis of secondary metabolites, nucleotide metabolism, and amino acid metabolism are getting severely inhibited in PTB conditions suggesting a general decline in their overall health, leading to a more diseased condition and being more prone to having a PTB Delivery. Key pathways that involve carbohydrate metabolism, nucleotide metabolism, signal transduction, replication and repair, and more, have been seriously impacted in the PTB Delivery condition when compared to the Term Birth Delivery condition. Because metabolic pathways have a significant impact on a pregnant woman's overall reproductive health, their inhibition is not a good indication and may be a contributing factor to PTB. This helps us in focusing on specific pathways that are contributing toward a PTB Delivery condition in Black women. On the contrary, in White women, several major pathways like biosynthesis of secondary metabolites, endocrine system, carbohydrate metabolism, cell growth and death, lipid metabolism, metabolism of cofactors and vitamins, signal transduction, and more were shown to be repressed in the PTB Delivery condition, while other pathways involving the excretory system, infectious disease, neurodegenerative disease, replication and repair, circulatory system, and more were found to be overexpressed, resulting in a heterogeneous pattern of pathway expression in White women. This shows that the genes engaged in these pathways are present in higher abundances as compared to Term Birth Delivery conditions and that they have a significant influence in shaping the pregnancy's final outcome. It can also be inferred that PTB may be induced by environmental causes other than vaginal infections. Inhibition of the important pathways in pregnant women suggests a general deterioration of a woman's reproductive health, which increases her risk of PTB. Fig. 8 depicts the main pathways and their expression levels in Term Birth Delivery and PTB Delivery conditions for both ethnicities by considering the average pathway abundance through all the samples.
Present work compared the composition of the vaginal microbiota between women who had a Term Birth Delivery and women who had a Preterm Birth Delivery (before 37 weeks) amongst both black and white women. Akila Subramaniam
Pathogenic microbes such as
Even in white women, we observed that microbes from the
Therefore, we investigated the variations in the composition of the vaginal microbiota between term and PTB Delivery samples, while discovering key microbes other than those that have previously been well studied. We also discovered a slew of distinct but critical genes in PTB Delivery samples, each of which contributes to a disturbed vaginal microenvironment and perhaps PTB in its way.
In conclusion, our work showcased a novel computational approach that gave crucial new insights into the modification of the vaginal microbiota in Term Birth and PTB Delivery for both white and black women. Our findings would provide a solid framework for future research to create management techniques that will aid in achieving a microbial community structure that favours safe delivery without pregnancy-related problems. In this work, we have demonstrated the necessity of investigating the microbial interactions that exist in the vaginal microbiota to better understand its involvement in pregnancy- related complications such as PTB. The keystone hub microbial species discovered have the potential to open new paths for developing microbiota modification techniques to mitigate pregnancy-related complications. Furthermore, the genes we discovered that were specific and unique to the PTB condition, as well as the genes and pathways that were over or under-expressed in the PTB condition, could be targeted for drug development and PTB prevention strategies. We also highlighted the key differences between the vaginal microbiotas of white and black women, since the vaginal microbiota is known to vary in women of different ethnicities .
We would like to acknowledge Dr. Ashok K. Chauhan, Founder, and President, Amity University, Uttar Pradesh for providing us the opportunity to conduct research. We would also like to thank Centre for Computational Biology and Bioinformatics, Amity Institute of Biotechnology, Amity University for providing us with necessary resources.
Designed the research and method: SK, PN, AS; Performed the research: SK, PN, AS; Supervised and validated the results: PN, AS, DM, BS; Conducted data compilation: SK and PN; Supervised the work and edited manuscript: AS, PN, DM, BS; Wrote the paper: PN, SK. All the authors have read and approved the final manuscript.
The authors have no financial conflicts of interest to declare.
Payal Gupta, Shriya Dube, Payal Priyadarshini, Shanvi Singh, Anasuya Pravallika R, Vijay Lakshmi Srivastava, Abhishek Sengupta, and Priyanka NaradMicrobiol. Biotechnol. Lett. 2023; 51(3): 317-324 https://doi.org/10.48022/mbl.2304.04008