Infertility is a global public health and social issue and has affected about 48.5 to 72.4 million couples worldwide (Boivin et al., 2007; Mascarenhas et al., 2012). The main causes of infertility include tubal and pelvic factors, anovulation, male factors, and unexplained infertility. In vitro fertilization-embryo transfer (IVF-ET) is currently used as an effective assistive reproductive therapy for infertility (Stewart et al., 2011). However, many patients fail to conceive even after IVF-ET, possibly due to endometrial factors and poor embryo quality, but some patients experience unexplained implantation failure.

Microorganisms inhabit all organs of the human body in huge orders of magnitude (Sender et al., 2016). The Human Microbiome Project revealed significant differences in diversity and abundance of the human microbiota even in healthy individuals (Human Microbiome Project Consortium, 2012). The urogenital tract represents 9% of the whole human microbiota (Peterson et al., 2009). Recent advancements in high-throughput sequencing technology and the increase in microbial studies have shed light on the microbiota of the reproductive tract. Microbial communities of the vagina, cervical canal, uterus, fallopian tubes, and peritoneal fluid in women of reproductive age have been identified. This revealed the bacterial colonization of the upper genital tract and a continuum of microbiota along the reproductive tract (Chen et al., 2017). The presence of a microbial continuum emphasizes the importance of healthy microbiota in the reproductive process (Franasiak and Scott, 2015). Thus, the role of genital tract microbiota in embryo implantation has attracted immense attention in recent times.

A number of studies suggested that vaginal dysbiosis may have a negative impact on the outcome of IVF-ET. Collection of vaginal secretions prior to ET revealed that women with Lactobacillus-dominated vaginal microbiota were more likely to get pregnant (Bernabeu et al., 2019; Riganelli et al., 2020). Women with lower proportions of Lactobacillus crispatus in the vagina had a decreased chance of successful pregnancy, and the vaginal microbiota could be a good predictor of IVF outcomes (Haahr et al., 2019a; Koedooder et al., 2019). An imbalance in the vaginal microbiota, as in bacterial vaginosis, may reduce clinical pregnancy rates in IVF patients (Haahr et al., 2016). Moreover, vaginal microbial composition, especially the relative abundance of Lactobacillus, may be associated with recurrent implantation failure (RIF; Fu et al., 2020). However, some studies also found no effect of vaginal microbiota on pregnancy or live birth rates after IVF treatment (Haahr et al., 2019b; Vergaro et al., 2019). Whether and how vaginal microbiota affects the outcome of ET remains to be elucidated.

For cervical microbiota, earlier studies based on bacterial culture identification showed decreased implantation and clinical pregnancy rates in the group with a positive culture of Escherichia coli and Streptococcus on the tip of the transplanted catheter as well as the cervical canal swab (Egbase et al., 1996; Fanchin et al., 1998). Infertile women had fewer Lactobacillus and a higher diversity index in the cervical microbiota than fertile women (Graspeuntner et al., 2018). Several studies have also investigated the effects of the upper reproductive tract microbiota on ET. Some studies have suggested that intrauterine microbiota, such as Lactobacillus, might have an impact on embryo implantation and pregnancy maintenance (Moreno et al., 2016). Another study found that although Lactobacillus was the dominant genus, there was no link between Lactobacillus and pregnancy outcome in IVF patients (Franasiak et al., 2016). Riganelli et al. found that Kocuria dechangensis was significantly more abundant in non-pregnant women undergoing IVF (Riganelli et al., 2020). The role of the uterine microbial composition in reproductive outcomes remains controversial.

There are only limited high-throughput sequencing studies on the impact of cervical microbiota on ET. The correlation between cervical and vaginal microbiota of patients undergoing IVF-ET has also been less studied. Cervical mucus sampling can be used to investigate the uterine microenvironment prior to ET due to unfeasible endometrial sampling and the intraindividual correlation between microbiota in each reproductive tract site (Chen et al., 2017). Therefore, a study on the impact of vaginal and cervical microbiota on ET in a large sample of IVF patients is needed. In this study, we characterized the microbial composition in the vagina and cervical canal samples from infertile women who underwent IVF, and explored the potential genera contributing to pregnancy failure based on 16S rRNA gene amplicon sequencing.

This study included 150 women who underwent their first fresh cycle of IVF or ICSI, and compared the vaginal and cervical microbial composition between pregnant and non-pregnant patients to explore the effect of lower genital microbial composition on the outcome of ET.

In our study, we found that many genera were abundant in both the cervix and vagina, and the Sorenson index confirmed intraindividual correlation between cervical and vaginal microbiota in the same patient. The similarity of cervical and vaginal microbiota in infertile patients undergoing IVF was related to the anatomical location of the cervix and vagina, and the colonizing microbes might migrate between the two sites. However, a higher Shannon index suggested a more diverse cervical microbiota. The PCoA plot also demonstrated that the microbial communities in the cervix and vagina were significantly different. The vaginal microbiota was dominated by Lactobacillus, resulting in a significant decrease in diversity. The variation in the microbial composition between the two sites may be due to the blocking of cervical mucus leading to changes during migration and colonization, or changes in the environment caused by large amounts of lactic acid produced by abundant Lactobacillus in the vagina (Aldunate et al., 2015).

Ravel et al. clustered vaginal microbial communities into five groups, known as community state types (CSTs). CST I, II, III, and V were dominated by Lactobacillus, while CST IV included more abundant strictly anaerobic bacteria (Ravel et al., 2011). Lactobacillus, one of the dominant genera in the vaginal microbiota, is capable of producing lactic acid (Aldunate et al., 2015) and bacteriocins (Barbés and Boris, 1999) to maintain a healthy environment and prevent pathogen invasion. The relative abundance of Lactobacillus has been suggested to be related to infertility (Zhao et al., 2020) and RIF (Fu et al., 2020). In our study, Lactobacillus and Gardnerella, although abundant, might have no effect on the outcome of ET. This result is similar to that of a previous study (Bernabeu et al., 2019). Lactobacillus crispatus and L. iners are the main Lactobacillus species in the vagina (Ravel et al., 2011). A previous study suggested that L. crispatus might have a positive impact on pregnancy rates (Koedooder et al., 2019; Vergaro et al., 2019), and L. iners was more likely associated with vaginal dysbiosis (Petrova et al., 2017). Similar to previous studies, we observed that L. crispatus was more abundant and L. iners was less abundant in the pregnant group. Although the differences were not statistically significant, we found a trend. L. iners≥5.69% was negatively correlated with pregnancy. L. iners, as the abundant species in the vagina, may be indicative of non-pregnancy. Larger samples may be needed to verify the positive effects of L. crispatus on pregnancy.

Some studies suggested that vaginal microbiota did not affect live birth rates or clinical pregnancy rates (Haahr et al., 2019b; Vergaro et al., 2019). A previous study found that Bifidobacterium might have a negative effect on pregnancy (Zhao et al., 2020). Prevotella is anaerobic and it has been shown to be associated with increased levels of cytokines, such as IL-1β, IL-8, and IL-17 (Larsen, 2017). In our study, we found an increase in Bifidobacterium and Prevotella in the vaginal microbiota of patients who failed to conceive, and they were also indicative of non-pregnancy. Vaginal microbial composition, such as the increase in Prevotella, Bifidobacterium, and L. iners, might play an important role in embryo implantation.

Fanchin et al. (1998) reported that positive cultures of E. coli and Streptococcus in the cervical microbiota on ET catheters led to decreased pregnancy rates. Salim et al. (2002) found a significant association between cervical colonization and pregnancy achievement. In our study, we found that an increase in S. torvum, Streptococcus, and Fusobacterium might be associated with conception failure. Meanwhile, it is possible that L. crispatus in the cervical microbiota could be favorable for the achievement of pregnancy, but the correlation was not confirmed. The inconsistency between our study and previous studies on the effect of cervical and vaginal microbiota might be due to the study population, sample size, research method, or sequencing fragments. However, S. torvum, Streptococcus, and Fusobacterium in the cervix, as well as Bifidobacterium and Prevotella in the vagina, were not so abundant. In particular, the ROC curve suggested that the predictive value of S. torvum and Fusobacterium in the cervix for non-pregnancy is not very high. Meanwhile, the mechanism of their influence on pregnancy outcomes after IVF was not determined in this study. More appropriate samples are therefore needed to validate their negative association with pregnancy and whether they affect pregnancy outcomes through metabolites (Fu et al., 2020) or immune factors (Benner et al., 2018).

Previous studies have confirmed that ovulation induction had no effect on the vaginal microbiota of infertile women who underwent IVF (Zhao et al., 2020). In this study, we fixed the sampling on the day of day 3 cleavage-stage ET in the fresh cycle to minimize the effect of ART protocols on the genital microbiota for convincing results. Although baseline parameters such as age, duration of infertility, ART protocols, embryo quality, and endometrial thickness did not differ between pregnant and non-pregnant groups in this study, there were large within-group differences due to differences in microbiota between individuals. A larger multicenter study is needed to validate these results. In addition, a larger sample will allow subgroup analysis of patients with biochemical pregnancy or abortion to investigate the effects of reproductive microbiota on early embryo implantation failure and spontaneous abortion after clinical pregnancy.

With the development of sequencing technology, the impact of reproductive tract microbial profiles on human reproduction has been increasingly well understood. This study explored the effect of the lower genital tract microbiota on fresh ET. Prevotella, Bifidobacterium, and L. iners in the vagina, as well as S. torvum, Streptococcus, and Fusobacterium in the cervical canal, may affect embryo implantation and have predictive value for IVF-ET outcomes. However, studies on the microbiota in the upper genital tract are lacking. Therefore, more large-scale studies on the microbiota of the reproductive tract continuum and its impact on ET, as well as interventional treatments such as probiotic or antibiotic therapies to improve the microecology of the reproductive tract, are needed.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA741994.

The studies involving human participants were reviewed and approved by the Ethics Committee of West China Second University Hospital of Sichuan University. The patients/participants provided their written informed consent to participate in this study.

RW and GZ designed the study and collected the samples. RW analyzed the data and wrote the manuscript. LW, XH, and YL contributed to the collection, arrangement, and transfer of the samples. BL provided statistical advice and contributed to the revision of the manuscript. HZ contributed to the collection of samples and guided the study. WH designed and guided the study and revised the manuscript. All authors read and approved the final version of the manuscript for submission.

This study was supported by a grant from West China Second University Hospital of Sichuan University (KL029).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.