• Users Online: 437
  • Print this page
  • Email this page

 Table of Contents  
ORIGINAL ARTICLE
Year : 2021  |  Volume : 5  |  Issue : 2  |  Page : 81-89

Analysis of the microbiome based on 16S rRNA gene signature in women with preterm versus term birth


1 Obstetrics and Gynecology Hospital of Fudan University, Shanghai 200090, China
2 Shiji Medical Laboratory, Shanghai 200093, China

Date of Submission13-Aug-2020
Date of Decision08-Oct-2020
Date of Acceptance26-Feb-2021
Date of Web Publication08-Jul-2021

Correspondence Address:
Rong Hu
Obstetrics and Gynecology Hospital of Fudan University, 128 Shenyang Road, Shanghai 200090
China
Xiao-Tian Li
Obstetrics and Gynecology Hospital of Fudan University, 128 Shenyang Road, Shanghai 200090
China
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/2096-2924.320887

Rights and Permissions
  Abstract 


Objective: To characterize and compare the microbiome signature in the maternal, intrauterine, and fetal environments and the associated bacterial species in women who experienced preterm birth and term birth.
Methods: A total of 140 women with singleton pregnancies were enrolled in this study. Among them, 31 experienced spontaneous preterm delivery (gestational age < 37 weeks), and 28 of them experienced vaginal delivery at term. Maternal peripheral blood, saliva, and vaginal discharge samples and fetal membrane, amniotic fluid, and cord blood samples were collected immediately after delivery under sterile conditions. DNA was isolated from the fetal membrane and umbilical cord blood samples, and the V3–V4 region of the bacterial 16S rRNA gene was sequenced. The sequence data were quality-filtered, chimera-checked, and organized into operational taxonomic units (OTUs) based on phylogeny. Principal coordinate analysis of beta diversity measures was used for visualization. The linear discriminant analysis effect size (LEfSe) algorithm and Wilcoxon test were used to differentiate the microbiomes found in the fetal membranes and cord blood in the cases of preterm birth.
Results: OTU analysis based on the 16S rRNA gene showed similar microbiomes in the maternal peripheral blood, amniotic fluid, fetal membranes, and cord blood. However, the LEfSe algorithm revealed significantly different bacterial compositions in the fetal environment between the preterm and term groups, with some of the bacterial species originating from the maternal peripheral blood or saliva.
Conclusions: The bacteria in the intrauterine and fetal environments may originate from other body sites through hematogenous transmission, and may cause the occurrence of preterm birth.

Keywords: Hematogenous Transmission; Microbiome; Preterm Birth; 16S rRNA Gene


How to cite this article:
Yu J, Peng T, Lu J, Li XT, Hu R. Analysis of the microbiome based on 16S rRNA gene signature in women with preterm versus term birth. Reprod Dev Med 2021;5:81-9

How to cite this URL:
Yu J, Peng T, Lu J, Li XT, Hu R. Analysis of the microbiome based on 16S rRNA gene signature in women with preterm versus term birth. Reprod Dev Med [serial online] 2021 [cited 2021 Jul 31];5:81-9. Available from: https://www.repdevmed.org/text.asp?2021/5/2/81/320887




  Introduction Top


Preterm deliveries are defined as those that occur at less than 37 weeks of gestational age; they are associated with 75% of perinatal mortality and more than 50% of long-term morbidity.[1] The preterm delivery rate is generally 5%–13%.[1] Infection is the only pathological process for which both a firm causal link with preterm birth has been established and a molecular pathophysiology is defined.[2],[3] It is estimated that infection is associated with at least 40% of all preterm deliveries.[4] However, the gold standard method for microbiological testing in clinical settings is bacterial culture. Traditional culture methods are often unable to detect bacteria that are challenging to culture or that cannot be cultured (especially anaerobes).[5] This limitation may hinder the identification of some bacteria associated with preterm birth and their possible sources. In addition, culture methods cannot truly reflect the original ecological status of microorganisms; therefore, they do not allow for an accurate analysis of the overall microbial community.

Recent advances in 16S rRNA gene amplicon high-throughput sequencing have enabled researchers to further identify preterm birth-related bacteria and their origin. The human body is home to millions of microorganisms.[6] The species of bacteria that have been observed in the intrauterine environment of women who experienced preterm birth include Ureaplasma parvum, Ureaplasma urealyticum, Mycoplasma hominis, Gardnerella vaginalis, Peptostreptococcus sp., Enterococcus sp., Streptococcus sp., Fusobacterium nucleatum, Leptotrichia sp., Sneathia sanguinegens, Haemophilus influenzae, and Escherichia coli.[7] You et al. reported that Bacteroides, Lactobacillus, Sphingomonas, Fastidiosipila, Weissella, and Butyricicoccus were more abundant in preterm birth samples than in term birth samples.[8] These bacteria were hypothesized to be the cause of preterm birth.[9],[10] Furthermore, excluding transmission from the ascending genital tract, Aagaard et al. reported that the placenta-harbored microbiome may originate from the oral cavity.[11] Some of these oral microbes, such as F. nucleatum, may spread from the maternal bloodstream.[7],[12]

Although it has been suggested that microbes in the intrauterine environment may originate from other body sites through hematogenous transmission,[6] studies comparing the microbiomes of the maternal peripheral blood and other maternal body sites with those of the amniotic fluid, fetal membranes, and cord blood are lacking. In addition, whether the intrauterine microbiome in pregnant Chinese women differs from that in pregnant women from other countries has not been investigated. The objective of this study was to determine the microbiome signature in the maternal, intrauterine, and fetal environments. In addition, we aimed to compare the intrauterine microbiome in healthy term pregnancies with that in preterm pregnancies to identify preterm birth-specific bacteria and their probable transmission pathways using a metagenomic approach based on 16S rRNA genes.


  Methods Top


Study approval and sample collection

This case-control study was approved by the Obstetrics and Gynecology Hospital of Fudan University (Ethical Review of Obstetrics and Gynecology (2013-26). All samples were procured after obtaining written informed consent from the participants.

We enrolled gravida from the labor room of the Obstetrics and Gynecology Hospital of Fudan University between March 2014 and December 2014. All participants conceived naturally and received regular prenatal care. The eligibility criteria included the successful acquisition of cord blood. Spontaneous preterm birth was defined as a documented gestational age >20 and < 37 weeks. The gestational age was confirmed via a first trimester sonogram. Subjects with evidence of maternal or fetal complications (e.g., multiple gestations, placenta previa, preeclampsia, and fetal anomalies) were excluded because of the probability of iatrogenic preterm birth. A control group, consisting of women who experienced spontaneous onset of labor or labor induction at term without suspected intrauterine infection, was also recruited.

After receiving informed consent, the clinical information was obtained directly from the medical records. Data entry into the database was completed by a research staff member. For this study, the extracted clinical metadata included the maternal age, gestational age at delivery, duration between membrane rupture and delivery, presence of maternal infection, mode of delivery, and neonatal condition.

Maternal peripheral blood, vaginal discharge, and saliva samples were collected from participants after enrollment in the study. Amniotic fluid (during C-section) and placental and cord blood samples were collected following delivery.

To collect the vaginal discharge samples, our research coordinator performed a sterile speculum examination and inserted a swab into the vagina of the patient, swirled it six times, followed by withdrawal without contamination. The swabs were stored in the collection tubes. Saliva samples were collected from the pharynx using a swab, following the same method.

Following delivery, the placentas were placed on a sterile operating platform. Three 0.5 cm × 0.5 cm × 0.5 cm chorion samples were collected from each placenta. Before collecting the membrane, sterile forceps were used to remove the amnion to decrease the potential contaminants. Biopsies were then collected via sharp dissection: one from the umbilical cord insertion site and two from the lateral edges of the placenta. Some fetal membrane samples were obtained by swirling the swab on the surface of the chorion, and these were labeled with a capital J. Each fetal membrane sample was placed in a sterile, labeled vial.

Amniotic fluid was sampled under sterile conditions after direct visualization of the intact amniotic bag through the uterine incision. Umbilical vein cord blood was collected under sterile conditions immediately after delivery and placed in ethylenediaminetetraacetic acid anticoagulant tubes.

All samples were immediately stored at −20°C and then transferred to a −80°C freezer within 48 h for storage until DNA extraction.

DNA extraction, polymerase chain reaction amplification, quantification, and sequencing

DNA was extracted under strict sterile conditions from the fetal membranes, amniotic fluid, cord blood, and maternal peripheral blood using the QIAamp DNA Blood Mini and QIAamp DNA Mini Kits (Qiagen, Hong Kong, China) according to the manufacturer's protocols. All extracted DNA samples were stored at −20°C.

The V3–V4 region of the 16S rRNA gene from the microbial genomic DNA was amplified using polymerase chain reaction (PCR) by using appropriate primers (forward primer: 5′-ACTCCTACGGGAGGCAGCAG-3′; reverse primer: 5′-GGACTACHVGGGTWTCTAAT-3′). The PCR products were detected using dual-indexing amplification, normalized, pooled, and sequenced using an Illumina MiSeq desktop sequencer (250 bp and 300 bp paired-end reads).[13] The PCR conditions used were as follows: 50°C for 2 min, 95°C for 10 min; 95°C for 15 s, 56°C for 30 s, and 72°C for 1 min, repeated for 40 cycle; and then 72°C for 10 min.

Sequence read processing was conducted using QIIME (version 1.9.0, http://bio.cug.edu.cn/qiime/), which included additional quality trimming, demultiplexing, and taxonomic assignments. Profiling of microbiota was conducted using PICRUSt (https://github.com/picrust/picrust/) based on the August 13, 2013, Greengenes database. The output file was further analyzed using Statistical Analysis of Metagenomic Profiles software package version 2.1.3 (https://beikolab.cs.dal.ca/software/STAMP).[14],[15] The raw sequencing paired-end reads were overlapped to form contiguous reads. Trimmomatic and FLASH software (http://ccb.jhu.edu/software/FLASH/) were used for quality control and filtering. The sequences were then aggregated to operational taxonomic units (OTUs) using Usearch 7.143 (http://qiime.org/) based on a 97% pairwise identity using QIIME's open-reference OTU picking strategy. Taxonomic classification of the representative sequences of fungal OTUs was conducted using the Ribosomal Database Project classifier (Release 11.1, http://rdp.cme.msu.edu/) against the fungal ITS database using Unite (Release 5.0, http://unite.ut.ee/index.php). Chimeric OTUs were detected using UCHIME (version 4.2.40, http://drive5.com/usearch/manual/uchimealgo.html) and removed from the OTU table.[16]

To assess alpha diversity, we employed Sobs, Shannon, Simpson, Abundance-based Coverage Estimator (ACE), and Chao indices as we analyzed the groups of samples. Beta diversity was calculated using principal component analysis (PCA) and principal coordination analysis (PCoA), using the jackknife support pipeline provided by QIIME.[17]

Statistical analysis

The Chi-square test was used to understand the significance of the categorical metadata for the study subjects. Student's t-test was used to understand the significance of continuous metadata. The threshold of statistical significance was set at P < 0.05. To exclude the possibility that the significant difference was caused by the difference in dispersions, the linear discriminant analysis effect size (LEfSe) algorithm was used to detect the significant features that differentiated the groups and to rank these features based on effect size. The threshold for the logarithmic linear discriminant analysis score for discriminative features was 2.0. The P value threshold used for the Wilcoxon test was set at 0.05.


  Results Top


Subject demographics and sampling types

Between March 2014 and December 2014, 4,266 gravida delivered at the hospital. Of these, 140 women with singleton gestations were enrolled. Women from whom we were unable to acquire cord blood samples were excluded. A total of 385 samples were collected. The specimen types included cord blood (140 samples), fetal membranes (129 samples), amniotic fluid (48 samples), maternal peripheral blood (45 samples), saliva (11 samples), and vaginal discharge (12 samples). A number of additional samples were excluded because of contamination during sampling. An extra speculum examination was refused by most participants. Baseline subject characteristics are shown in [Supplementary Table 1]. The incidence of premature rupture of the membranes was 50.0%, and the average duration from membrane rupture to delivery was 12.3 ± 15.2 h. The incidence of prenatal fever was 9.3%. The cesarean section rate was 50.0%. The four instances of perinatal mortality were attributed to extreme preterm births.



Differentiation of the microbiota among tissues

When analyzing the alpha diversity among samples from different tissues, variable levels of differentiation were observed. The vaginal and oral cavity microbiomes in participants were found to be less diverse than those found in maternal peripheral blood, cord blood, fetal membranes, and amniotic fluid [Supplementary Figure 1].



We calculated the PCA at the OTU level of abundance to reflect the differences among the six body sites in terms of beta diversity. [Figure 1] shows the niche differences based on the sample site. The placental microbiome was similar to that of the maternal peripheral blood, amniotic fluid, and fetal cord blood. In contrast, the microbiomes of saliva and vaginal discharge samples were distinct from those of the placenta [Figure 1].
Figure 1: Beta diversity analysis of bacterial 16S rRNA genes was calculated at the operational taxonomic unit level to reflect the differentiation among the six body sites. Each point corresponds to a sample from a given body site. The percentage of variation explained by the plotted principal coordinates is indicated on the axes. The vaginal and oral cavity microbiomes were different from the communities of the maternal peripheral blood, cord blood, fetal membranes, and amniotic fluid. A: Amniotic fluid; B: Maternal peripheral blood; C: Cord blood; M and J: Fetal membranes; S: Saliva; V: Vaginal discharge.

Click here to view


We used the Ribosomal Database Project classifier to perform taxonomic classification of the microbial community of each sample. At the phylum level, each body niche was represented by one or a few signature phyla, such as Firmicutes and Actinobacteria in the vagina and Firmicutes and Fusobacteria in saliva. Firmicutes and Proteobacteria were present in the maternal blood and intrauterine environments. In agreement with diversity analysis, a strong phylum-level similarity was observed in the maternal peripheral blood, placenta, amniotic fluid, and cord blood taxonomic profiles, which differed from those of the saliva and vaginal discharge samples [Figure 2]. We compared the mutual OTUs in the maternal peripheral blood, cord blood, fetal membranes, and amniotic fluid and found that approximately 25% of OTUs were identified in all four body sites [Figure 3].
Figure 2: Microbial abundance at the phylum level among each sample. Each sample was arranged according to their tissue origin on the y-axis. The relative abundance of microbes at the phylum level is represented by a horizontal bar on the x-axis. A: Amniotic fluid; B: Maternal peripheral blood; C: Cord blood; M and J: Fetal membranes; S: Saliva; V: Vaginal discharge.

Click here to view
Figure 3: Mutual operational taxonomic units in the maternal peripheral blood, cord blood, fetal membranes, and amniotic fluid. (a) The mutual operational taxonomic units in the maternal peripheral blood, cord blood, fetal membranes, and amniotic fluid. (b) The number of operational taxonomic units in each of the body site groups. (c) The number of operational taxonomic units shared by body sites. A total of 6,585 operational taxonomic units were shared by all four body sites. A total of 3,228 operational taxonomic units were shared by any three of the four groups. A: Amniotic fluid; B: Maternal peripheral blood; C: Cord blood; M and J: Fetal membranes.

Click here to view


Although phylum-level similarity was observed in the maternal peripheral blood, placenta, amniotic fluid, and cord blood taxonomic profiles, there were still differences in the microbial composition. We compared the bacteria at the phylum level among these four body sites using the Wilcoxon test and found diversity regarding Bacteroidetes, Chloroflexi, Acidobacteria, Gemmatimonadetes, Chlorobi, Tenericutes, and Nitrospirae [Figure 4].
Figure 4: Differentiation of microbial abundance at the phylum level among the maternal peripheral blood, cord blood, fetal membrane, and amniotic fluid samples. The y-axis represents bacterial abundance at the phylum level, and the x-axis represents their mean proportion in each tissue. A: Amniotic fluid; B: Maternal peripheral blood; C: Cord blood; M: Fetal membranes. *P ≤ 0.05; - P ≤ 0.01; - P ≤ 0.001.

Click here to view


To identify the specific taxonomic biomarkers in each body site, we additionally applied the LEfSe algorithm to identify the microbial species that were significantly more prevalent in each body niche than in all other niches. The specific microbes in each of the six body niches are shown in [Supplementary Table 2].



Variation in the intrauterine microbiota associated with preterm delivery

Among the 140 subjects, 31 experienced spontaneous preterm births, and 28 of them experienced vaginal delivery at term without pregnancy complications. Two cases of iatrogenic preterm delivery were excluded. Consistent with the principles of a case-control design, we analyzed the alpha and beta diversity in the placental and cord blood specimens derived from either healthy gravida who experienced vaginal delivery at term as the control group (n = 28) or women who experienced a spontaneous preterm birth (n = 31). The baseline characteristics of the study subjects are presented in [Supplementary Table 3]. In the preterm birth group, there was a significantly longer duration between the rupture of the membranes and delivery, higher white blood cell counts, lower 1- and 5-min Apgar scores, and higher neonatal admission and infection rates than the term birth group [Supplementary Table 3].



As many maternal peripheral blood, saliva, vaginal discharge, and amniotic fluid samples could not be used due to contamination, we only had a sufficient number of fetal membranes and cord blood samples to complete the comparisons. For the alpha diversity analysis, Sobs, Shannon, Simpson, ACE, and Chao indices were calculated for the preterm and term delivery groups. As shown in [Supplementary Table 4] and [Supplementary Table 5], we observed no significant differences between the two groups.



In terms of beta diversity, microbial community clustering was observed in PcoA plots using the unweighted UniFrac distance. PcoA plots revealed similar communities in the fetal membrane and cord blood samples that were structured between the preterm and term delivery groups [Figure 5].
Figure 5: Beta diversity analysis of bacterial 16S rRNA genes in the similar microbiome communities of fetal membrane (a) and cord blood (b) samples structured between preterm and term delivery. Each point represents a sample from a subject with either preterm delivery (blue) or term delivery (red). The beta diversity of the microbial community clustering corresponds to principal coordination analysis plots using unweighted UniFrac distance. The percentage of variation explained by the plotted principal coordinates is indicated on the axes.

Click here to view


While our beta diversity analysis revealed similarity regarding the fetal membrane and cord blood microbiomes between women who experienced a term delivery and those who experienced a preterm delivery, there could be differences in the microbial composition. Therefore, to identify the specific fetal membrane and cord blood microbiota in gravida who experienced spontaneous preterm birth, we applied the LEfSe algorithm. We observed that the abundance of Burkholderiaceae, Proteobacteria, Limnobacter, Intestinibacter, Intestinibacter bartlettii DSM 16795, Ruminococcaceae UCG 013, Megasphaera elsdenii, and Nesterenkonia was significantly higher in the fetal membrane of the preterm group than in that of the term group [Figure 6]a. In contrast, Lactobacillus iners, Dokdonella, Ideonella, Megasphaera, Coprococcus, M. elsdenii, Finegoldia, and Eubacterium rectale were significantly more abundant in the cord blood samples of the preterm group than in the same samples of the term group [Figure 6]b.
Figure 6: Bacterial taxa contributing to the differentiation of the fetal membrane (a) and cord blood (b) microbiome communities in women who experienced preterm birth. Bacterial taxa were selected as significantly differentially abundant between preterma delivery (gray bar) and term delivery (red bar) microbial communities using the linear discriminant analysis effect size algorithm and were sorted based on the degree of difference. The threshold used on the linear discriminant analysis effect size algorithm for the discriminative feature was 2.0.

Click here to view


To determine the origin of these bacteria, which is potentially relevant for analyzing spontaneous preterm birth, we compared the obtained bacteria to the specific microbes observed in each body niche, as shown in [Supplementary Table 2]. We observed that Burkholderiaceae was significantly more abundant in the amniotic fluid and saliva samples, while Megasphaera was significantly more abundant in the saliva samples than other sample sites. The abundance of both L. iners and Finegoldia was significantly higher in the vaginal discharge samples than in the other samples. Ruminococcaceae UCG 013 was prevalent in the saliva, maternal peripheral blood, fetal membrane, and cord blood samples. Dokdonella, Nesterenkonia, and Ideonella were significantly more abundant in the fetal membrane samples than in the other sample sites.


  Discussion Top


Our 16S rRNA gene-based OTU analysis of samples from the maternal peripheral blood, amniotic fluid, fetal membranes, cord blood, and other maternal body sites revealed that the maternal peripheral blood, intrauterine environment, and cord blood harbor similar microbiomes. Furthermore, the LEfSe algorithm showed a significantly different bacterial composition in the preterm group compared to the term group; some of these bacterial species may originate from the saliva or maternal peripheral blood. This suggests a hematogenous pathway of microbes in the intrauterine and fetal environment, which may be associated with preterm birth.

In our study, only saliva and vaginal discharge samples could be differentiated from the samples of the other body sites using alpha and beta diversity analyses. The placental microbiota at the phylum level also showed high levels of similarity in the maternal peripheral blood, amniotic fluid, and fetal blood, and these sample types shared approximately 25% of their OTUs. Although some microbes (of the 25% OTUs) that are common in all four body sites might represent systemic contamination, the purpose of this study was to show the possibility that these sites harbor similar microbiomes. To mitigate this impact, we analyzed the outputs. The workload required to find where all these bacterial species are commonly found is extremely large; thus, we only focused on the bacteria of interest achieved from the comparative analysis. Burkholderiaceae, Proteobacteria, Limnobacter, Intestinibacter, Ruminococcaceae UCG 013, and Megasphaera were significantly more abundant in the fetal membrane samples from the preterm group, whereas L. iners, Dokdonella, Ideonella, Megasphaera, Coprococcus, Finegoldia, and E. rectale were significantly more abundant in the cord blood samples from the preterm group than term group.

Many of these microorganisms have not been previously reported to be associated with preterm delivery. Eubacterium has been isolated from various human sites, including the gastrointestinal tract, female genital tract, oral cavity, thoracic cavity, and prostate,[18] and it has also been found in newborn and preterm infants with meningitis.[19] To the best of our knowledge, this is the first study to report its association with preterm birth. Finegoldia, previously related to bone and joint infection, abscess, and infectious endocarditis,[20],[21],[22] was found to be significantly more abundant in the cord blood of the preterm group than term group our study, which is also the first report of its relationship with preterm birth.

Megasphaera, a common bacterium associated with bacterial vaginosis,[23] is found in a high proportion in the vaginal environment of women with a history of spontaneous preterm birth. The incidence of recurrent spontaneous preterm birth is also increased in patients with a high proportion of Megasphaera in vaginal discharge before 16 weeks of gestation.[24] According to our study, the abundance of Megasphaera was high in the placenta and umbilical cord of women in the preterm birth group, as well as in their saliva samples, which suggested the impact of oral bacteria on preterm birth and the possibility of its spread through the blood stream. Therefore, Burkholderiaceae, a bacterial genus previously associated with preterm birth,[25] which was significantly more abundant in the amniotic fluid and saliva samples, might have originated from the oral cavity. Ruminococcaceae UCG 013, a bacterial group that is challenging to culture, was prevalent in the saliva, maternal peripheral blood, fetal membrane, and cord blood samples, and it may have originated from the maternal peripheral blood and oral cavity. These outcomes suggest that the microbiota in the intrauterine and fetal environments may be established through hematogenous transmission.

Our findings regarding the relevance of the oral and intrauterine microbiomes and the associated transmission pathways are consistent with previously published observations. Aagaard et al. found that a remote history of antenatal infection could cause changes in the intrauterine microbial community, leading to an increase in the abundance of Streptococcus, Acinetobacter, Thioalkalivibrio, and Pseudomonas.[11] Bearfield et al. found that the presence of Streptococcus and F. nucleatum in amniotic fluid may be of oral origin, which also suggests a hematogenous transmission route of intrauterine infection.[26] Fardini et al. further supported this hypothesis by injecting gingival plaque samples from patients with periodontal disease into the veins of mice. By using 16S rRNA gene analysis, they found that the majority of the placental microbiota originated from oral microorganisms. In addition, they found that some of these bacterial species were enriched compared to those in gingival samples.[27] The underlying mechanism may involve Fusobacterium adhesion factor A (FadA), which binds to vascular endothelial cadherin (VE-cadherin), thereby causing VE-cadherin to migrate away from the cell-cell junction, which increases the endothelial permeability and enables microbes to penetrate the endothelial barrier.[28],[29]

Our study has several limitations, the primary one being the limited number of samples from several body sites, especially saliva and vaginal samples. As there was only one sample collector involved in this study, some samples may not have been collected from patients who were admitted or who delivered outside working hours. As most of the participants had eaten before entering the labor room, we did not take their saliva samples to avoid food particle interference. Furthermore, an extra speculum examination was refused by most participants. During labor, the amnion may separate from the chorion, which can cause contamination of the placental sample; therefore, we excluded the placental samples collected in these conditions.

Another limitation was that the placenta, blood, and amniotic fluid samples were relatively free of microbes; therefore, the results might be more heavily influenced by contaminants introduced during the experimental procedures.[30],[31] Although the concept of using a negative or positive control exists since 2018,[30],[31] there are other researchers, such as Aagaard et al., who question if the fact that these believed “contaminants” do not belong to the uterus should be confirmed. Another issue is associated with how to decrease the impact of systemic contamination, especially in samples containing fewer microbes.[32] To overcome this problem, we used comparative analysis methods (LEfSe algorithm and Wilcoxon test) to minimize its impact. However, in tissue comparisons, the LEfSe algorithm was used to identify the microbes that were significantly enriched in one tissue compared to other tissues, which may have resulted in some microbial species being missed due to the algorithm. This may explain why preterm birth-related microbes were scarce in the maternal peripheral blood samples.

Our study involves observational research, which limits the conclusions that can be drawn regarding the causal impact of microbiota on preterm birth. Whether these bacteria cause preterm birth and their impact on newborns need to be further investigated in future experiments.

Maternal peripheral blood, the intrauterine environment, and cord blood were shown to contain similarly structured microbiomes, indicating that the microbes in the intrauterine and fetal environments may have a hematogenous origin. Among these, Megasphaera, Burkholderiaceae, Ruminococcaceae UCG 013, Coprococcus, Eubacterium, and Finegoldia were associated with preterm birth, and some of these bacteria may have originated from saliva through hematogenous transmission.

Acknowledgments

We are grateful to all study participants and technical laboratory support at the Obstetrics and Gynecology Hospital of Fudan University and Shiji Medical Laboratory. We wish to thank the staff in the labor room of the Obstetrics and Gynecology Hospital of Fudan University.

Financial support and sponsorship

This study was supported by the Program of Shanghai Leading Talent (2012), Shanghai Key Program of Clinical Science and Technology Innovation (17411950500; 17411950501; 18511105602), National Science Foundation of China (81741047; 81971411; 81571460), Shanghai Medical Center of Key Programs for Female Reproductive Diseases (2017ZZ01016), National Key Basic Research Plan of China (973 Plan) (2015CB943300), National Key R&D Program of China (2016YFC1000403), and The Major Program of the National 13th Five-Year Plan of China (2016YFC1000400).

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet 2008;371:75-84. doi: 10.1016/s0140-6736(08)60074-4.  Back to cited text no. 1
    
2.
Vaisbuch E, Romero R, Erez O, Mazaki-Tovi S, Pedro KJ, Soto E, et al. Fragment Bb in amniotic fluid: Evidence for complement activation by the alternative pathway in women with intra-amniotic infection/inflammation. J Matern Fetal Neonatal 2009;22:905-16. doi: 10.1080/14767050902994663.  Back to cited text no. 2
    
3.
Romero R, Espinoza J, Gonçalves LF, Kusanovic JP, Friel L, Hassan S. The role of inflammation and infection in preterm birth. Semin Reprod Med 2007;25:21-39. doi: 10.1055/s-2006-956773.  Back to cited text no. 3
    
4.
Simhan HN, Canavan TP. Preterm premature rupture of membranes: Diagnosis, evaluation and management strategies. BJOG 2005;112:32-7. doi: 10.1111/j.1471-0528.2005.00582.x.  Back to cited text no. 4
    
5.
Han YW, Shen T, Chung P, Buhimschi IA, Buhimschi CS. Uncultivated bacteria as etiologic agents of intra-amniotic inflammation leading to preterm birth. J Clin Microbiol 2009;47:38-47. doi: 10.1128/jcm. 01206-08.  Back to cited text no. 5
    
6.
Parnell LA, Briggs CM, Mysorekar IU. Maternal microbiomes in preterm birth: Recent progress and analytical pipelines. Semin Perinatol 2017;41:392-400. doi: 10.1053/j.semperi.2017.07.010.  Back to cited text no. 6
    
7.
Payne MS, Bayatibojakhi S. Exploring preterm birth as a polymicrobial disease: An overview of the uterine microbiome. Front Immunol 2014;5:595. doi: 10.1053/j.semperi.2017.07.010.  Back to cited text no. 7
    
8.
You YA, Yoo JY, Kwon EJ, Kim YJ. Blood microbial communities during pregnancy are associated with preterm birth. Front Microbiol 2019;10:1122. doi: 10.3389/fmicb.2019.01122.  Back to cited text no. 8
    
9.
Holst D, Garnier Y. Preterm birth and inflammation – The role of genetic polymorphisms. Eur J Obstet Gynecol Reprod Biol 2008;141:3-9. doi: 10.1016/j.ejogrb.2008.07.020.  Back to cited text no. 9
    
10.
Kacerovsky M, Pliskova L, Menon R, Kutova R, Musilova I, Maly J, et al. Microbial load of umbilical cord blood Ureaplasma species and Mycoplasma hominis in preterm prelabor rupture of membranes. J Matern Fetal Neonatal 2014;27:1627-32. doi: 10.3109/14767058.2014.887068.  Back to cited text no. 10
    
11.
Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, Versalovic J. The placenta harbors a unique microbiome. Sci Transl Med 2014;6:237ra65. doi: 10.1126/scitranslmed.3008599.  Back to cited text no. 11
    
12.
Han YW, Ikegami A, Bissada NF, Herbst M, Redline RW, Ashmead GG. Transmission of an uncultivated Bergeyella strain from the oral cavity to amniotic fluid in a case of preterm birth. J Clin Microbiol 2006;44:1475-83. doi: 10.1128/jcm.44.4.1475-1483.2006.  Back to cited text no. 12
    
13.
Fadrosh DW, Ma B, Gajer P, Sengamalay N, Ott S, Brotman RM, et al. An improved dual-indexing approach for multiplexed 16S rRNA gene sequencing on the Illumina MiSeq platform. Microbiome 2014;2:6. doi: 10.1186/2049-2618-2-6.  Back to cited text no. 13
    
14.
Parks DH, Tyson GW, Hugenholtz P, Beiko RG. STAMP: Statistical analysis of taxonomic and functional profiles. Bioinformatics 2014;30:3123-4. doi: 10.1093/bioinformatics/btu494.  Back to cited text no. 14
    
15.
Langille MG, Zaneveld J, Caporaso JG, McDonald D, Knights D, Reyes JA, et al. Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nat Biotechnol 2013;31:814-21. doi: 10.1038/nbt.2676.  Back to cited text no. 15
    
16.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods 2010;7:335-6. doi: 10.1038/nmeth.f.303.  Back to cited text no. 16
    
17.
Aagaard K, Riehle K, Ma J, Segata N, Mistretta TA, Coarfa C, et al. A metagenomic approach to characterization of the vaginal microbiome signature in pregnancy. PLoS One 2012;7:e36466. doi: 10.1371/journal.pone.0036466.  Back to cited text no. 17
    
18.
Lee MR, Huang YT, Liao CH, Chuang TY, Wang WJ, Lee SW, et al. Clinical and microbiological characteristics of bacteremia caused by Eggerthella, Paraeggerthella, and Eubacterium species at a university hospital in Taiwan from 2001 to 2010. J Cli Microbiol 2012;50:2053-5. doi: 10.1128/JCM.00548-12.  Back to cited text no. 18
    
19.
Brook I. Meningitis and shunt infection caused by anaerobic bacteria in children. Pediatr Neurol 2002;26:99-105. doi: 10.1016/s0887-8994(01) 00330-7.  Back to cited text no. 19
    
20.
Walter G, Vernier M, Pinelli PO, Million M, Coulange M, Seng P, et al. Bone and joint infections due to anaerobic bacteria: An analysis of 61 cases and review of the literature. Eur J Cli Microbiol Infect Dis 2014;33:1355-64. doi: 10.1007/s10096-014-2073-3.  Back to cited text no. 20
    
21.
Fournier PE, La MV, Casalta JP, Richet H, Collart F, Raoult D. Finegoldia magna, an early post-operative cause of infectious endocarditis: Report of two cases and review of the literature. Anaerobe 2008;14:310-2. doi: 10.1016/j.anaerobe.2008.11.001.  Back to cited text no. 21
    
22.
Castello L, Bou M, Bazzana MS, Predari SC. Nonpuerperal breast abscess caused by Finegoldia magna. Rev Argent Microbiol 2007;39:95-8.  Back to cited text no. 22
    
23.
Tamrakar R, Yamada T, Furuta I, Cho K, Morikawa M, Yamada H, et al. Association between Lactobacillus species and bacterial vaginosis-related bacteria, and bacterial vaginosis scores in pregnant Japanese women. BMC Infect Dis 2007;7:128. doi: 10.1186/1471-2334-7-128.  Back to cited text no. 23
    
24.
Nelson DB, Hanlon A, Nachamkin I, Haggerty C, Mastrogiannis DS, Liu CZ, et al. Early pregnancy changes in bacterial vaginosis-associated bacteria and preterm delivery. Paediatr Perinat Epidemiol 2014;28:88-96. doi: 10.1111/ppe.12106.  Back to cited text no. 24
    
25.
Porter MC, Pennell CE, Woods P, Dyer J, Merritt AJ, Currie BJ. Case report: Chorioamnionitis and premature delivery due to Burkholderia pseudomallei infection in pregnancy. Am J Trop Med Hyg 2018;98:797-9. doi: 10.4269/ajtmh.17-0789.  Back to cited text no. 25
    
26.
Bearfield C, Davenport ES, Sivapathasundaram V, Allaker RP. Possible association between amniotic fluid micro-organism infection and microflora in the mouth. BJOG 2002;109:527-33. doi: 10.1111/j. 1471-0528.2002.01349.x.  Back to cited text no. 26
    
27.
Fardini Y, Chung P, Dumm R, Joshi N, Han YW. Transmission of diverse oral bacteria to murine placenta: Evidence for the oral microbiome as a potential source of intrauterine infection. Infect Immun 2010;78:1789-96. doi: 10.1128/iai.01395-09.  Back to cited text no. 27
    
28.
Vander Haar EL, So J, Gyamfi-Bannerman C, Han YW. Fusobacterium nucleatum and adverse pregnancy outcomes: Epidemiological and mechanistic evidence. Anaerobe 2018;50:55-9. doi: 10.1016/j.anaerobe.2018.01.008.  Back to cited text no. 28
    
29.
Solt I. The human microbiome and the great obstetrical syndromes: A new frontier in maternal-fetal medicine. Best Pract Res Clin Obstet Gynaecol 2015;29:165-75. doi: 10.1016/j.bpobgyn.2014.04.024.  Back to cited text no. 29
    
30.
Leon LJ, Doyle R, Diez-Benavente E, Clark TG, Klein N, Stanier P, et al. Enrichment of clinically relevant organisms in spontaneous preterm delivered placenta and reagent contamination across all clinical groups in a large UK pregnancy cohort. Appl Environ Microbiol 2018;84:e00483-18. doi:10.1128/AEM.00483-18.  Back to cited text no. 30
    
31.
de Goffau MC, Lager S, Sovio U, Gaccioli F, Cook E, Peacock SJ, et al. Human placenta has no microbiome but can contain potential pathogens. Nature 2019;572:329-34. doi: 10.1038/s41586-019-1451-5.  Back to cited text no. 31
    
32.
Bawagan J. Babies get critical gut bacteria from their mother at birth, not from placenta, study suggests. Science 2019. doi: 10.1126/science.aay9546. Available from: https://www.sciencemag.org/news/2019/07/bacteria-free-placentas-suggest-babies-pick-microbiome-birth. [Last accessed on 2020 Aug 13].  Back to cited text no. 32
    


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
Abstract
Introduction
Methods
Results
Discussion
References
Article Figures

 Article Access Statistics
    Viewed454    
    Printed2    
    Emailed0    
    PDF Downloaded56    
    Comments [Add]    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]